GAIA Nº 15, LISBOA/LISBON, DEZEMBRO/DECEMBER 1998, pp. 5-61 (ISSN: 0871-5424)
A NEW PHYLOGENY OF THE CARNIVOROUS
DINOSAURS
Thomas R. HOLTZ, Jr.
Department of Geology, University of Maryland. College Park, MARYLAND 20735. USA
E-Mail: tholtz@geol.umd.edu
ABSTRACT: The last several years have seen the discovery of many new theropod dinosaur
taxa. Data obtained from these and from fragmentary forms not previously utilized in cladistic analyses are examined. An analysis of forty one primary ingroup taxa and 386 characters
yielded a set of most parsimonious cladograms which preserves many previously discovered relationships (e.g., a basal split between Ceratosauria and Tetanurae; a carnosaurcoelurosaur clade Avetheropoda outside of more primitive "megalosaur" - grade tetanurines; Dromaeosauridae as the sister taxon to birds, and so forth). The Middle Jurassic
English Proceratosaurus was discovered to be a basal coelurosaur, as was (on less secure
evidence) the Middle Jurassic Chinese Gasosaurus: these are among the oldest coelurosaurs yet described. Several characters previously considered to be restricted to birds and
other advanced coelurosaurs (e.g., furcula, semilunate carpal block) were found to be more
broadly distributed among tetanurines. Other characters, once considered synapomorphies for Avetheropoda (e.g., loss of metacarpal IV, possession of a pubic obturator notch)
were found to be convergent between advanced carnosaurs and advanced coelurosaurs,
lacking in the basal members of both clades. At least three (and possibly four) separate origins for the arctometatarsalian pes were supported in this study. The mosaic of derived
character state distributions for troodontids relative to the dromaeosaurid-bird clade, the
tyrannosaurid-ornithomimosaur clade, and the therizinosauroid-oviraptorosaur clade suggests that relationships alternative to the most parsimonious found here may be supported
in future studies.
The present analysis attempts to synthesize the
proposed phylogenetic data from these studies, beginning as an update of previous work by the present
author (HOLTZ, 1994). Among the changes from that
work include correction of typographical errors in the
character descriptions and elimination or modification of poorly coded characters (see CLARK, PERLE
& NORELL, 1994; HUTT, MARTILL & BARKER, 1996;
CHARIG & MILNER, 1997; NORELL & MAKOVICKY,
1997 for specific examples). Furthermore, many
significant new theropod taxa and more complete
remains of hitherto poorly known forms (therizinosauroids, basal ornithomimosaurs, spinosaurids,
sinraptorids, alvarezsaurids, etc.) which were not
previously used have been incorporated into the
new analysis. Additionally, new characters drawn
from the studies listed previously are included
(some in a modified form) here, as are some previously unused characters.
INTRODUCTION
Since the pioneering work of GAUTHIER (1986),
there has been great scientific interest in the phylogeny of the Theropoda MARSH, 1881. Much of this
interest stems from the recognition that the origin of
birds lies within the theropod dinosaurs, an hypothesis advanced by OSTROM (e.g., 1974, 1975a, 1975b,
1976) primarily from his work on the dromaeosaurid
Deinonychus antirrhopus OSTROM, 1969a (see PADIAN & CHIAPPE, 1998 for a recent review of bird origins). The results of Gauthier's phylogenetic
analysis are shown in Fig. 1A, B.
Numerous authors have proposed phylogenetic
hypotheses subsequent to Gauthier's initial 1986
study: BAKKER, WILLIAMS & CURRIE, 1988; NOVAS,
1992, 1997a; CURRIE & ZHAO, 1993a; RUSSELL &
DONG, 1993a, b; PÉREZ-MORENO et al., 1993, 1994;
HOLTZ, 1994, 1995a, 1996a; SERENO et al., 1994,
1996, 1998; SERENO, 1997, 1998; SUES, 1997; HARRIS, 1998; FORSTER et al. 1998; MAKOVICKY & SUES,
1998. Results of some of these studies are presented in Fig. 1.
To fully describe each of the characters in this
study in detail will require a much longer work (in
preparation by the present author). This paper, how-
5
artigos/papers
T.R. HOLTZ, JR.
(Dromaeosauridae
Fig. 1 - Previously proposed phylogenies of theropod relationships. Taxonomy of listed forms revised to match the
names used here. A - Cladogram of comparatively well known theropods from GAUTHIER (1986). B - Cladogram of all
theropods included in GAUTHIER (1986), dashed line for Ornithomimosauria represents topology from GAUTHIER (1986),
solid line after WILKINSON (1995) (differs from that presented in GAUTHIER (1986) due to incomplete computational analysis in that study: see WILKINSON (1995) for details). C - BAKKER, WILLIAMS & CURRIE (1988). D - NOVAS (1992).
E - RUSSELL & DONG (1993a). F - PÉREZ-MORENO et al. (1993). G - HOLTZ (1994). (Continued)
6
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
Fig. 1 (continued) - Previously proposed phylogenies of theropod relationships. Taxonomy of listed forms revised to
match the names used here. H - PÉREZ-MORENO et al. (1994). I - SERENO (1997, 1998). J - FORSTER et al. (1998).
K - MAKOVICKY & SUES (1998).
peris, and Unenlagia comahuensis) were examined
in subsequent analyses.
To fully describe each of the characters in this
study in detail will require a much longer work (in
preparation by the present author). This paper, however, will serve as an interim study pending that
more detailed phylogenetic analysis.
The sister taxon to those theropods used here is
a matter of some recent debate. NOVAS (1994,
1997b), SERENO (1997, 1998), and SERENO & NOVAS (1992, 1994) have proposed that the Late Triassic taxa Eoraptor lunensis SERENO et al., 1993 of
Argentina and the more globally distributed Herrerasauridae BENEDETTO, 1973 share a more recent
common ancestor with the taxa used in this analysis
than do any other known forms (Fig. 2A). Under this
phylogeny, Eoraptor and the herrerasaurids would
METHODS AND MATERIALS
The operational taxonomic units (OTUs) employed in this study are listed in TABLE I. Forty one ingroup taxa were used in the primary analysis. The
phylogenetic positions of three very fragmentary
forms (Deltadromeus agilis, "Megalosaurus" hes-
7
T.R. HOLTZ, JR.
TABLE I
Theropod taxa used in this analysis.
OPERATIONAL TAXONOMIC UNITS
TAXA INCLUDED IN PRIMARY ANALYSIS
Abelisaurus comahuensis BONAPARTE & NOVAS, 1985
Acrocanthosaurus atokensis STOVALL & LANGSTON, 1950
Afrovenator abakensis SERENO, WILSON, LARSSON, DUTHEIL & SUES, 1994
1
Allosaurus spp. MARSH, 1877
Alvarezsauridae BONAPARTE, 1991
2
Archaeopteryx spp. MEYER, 1861
Bagaraatan ostromi OSMÓLSKA, 1996
Caenagnathidae STERNBERG, 1940
Carcharodontosaurus saharicus (DEPÉRET & SAVORNIN, 1927)
Carnotaurus sastrei BONAPARTE, 1985
Ceratosaurus nasicornis MARSH, 1884
Coelophysidae WELLES, 1984
Coelurus fragilis MARSH, 1879b
Compsognathidae COPE, 1871
Dilophosaurus wetherilli (WELLES, 1954)
Dromaeosauridae RUSSELL, 1969
Dryptosaurus aquilunguis (COPE, 1866)
Elaphrosaurus bambergi JANENSCH, 1920
Eustreptospondylus oxoniensis WALKER, 1964
Gasosaurus constructus DONG & TANG, 1985
Giganotosaurus carolinii CORIA & SALGADO, 1995
Megalosaurus bucklandi MEYER, 1832
Microvenator celer OSTROM, 1970
Monolophosaurus jiangi ZHAO & CURRIE, 1993
Neovenator salierii HUTT, MARTILL & BARKER, 1996
Ornitholestes hermanni OSBORN, 1903
Ornithomimidae MARSH, 1890
Ornithothoraces CHIAPPE & CALVO, 1994
Oviraptoridae BARBOLD, 1976a
Pelecanimimus polyodon PÉREZ-MORENO, SANZ, BUSCALIONI, MORATALLA, ORTÉGA & RASSKIN-GUTMAN, 1994
Piatnitzkysaurus floresi BONAPARTE, 1979
Proceratosaurus bradleyi (WOODWARD, 1910)
Rahonavis ostromi (FORSTER, SAMPSON, CHIAPPE & KRAUSE, 1998)
Scipionyx samniticus DAL SASSO & SIGNORE, 1998
3
Sinraptor spp. CURRIE & ZHAO, 1993a
TROMER
Spinosauridae S
, 1915
Therizinosauroidea RUSSELL & DONG, 1993a
Torvosaurus tanneri GALTON & JENSEN, 1979
Troodontidae GILMORE, 1924
Tyrannosauridae OSBORN, 1906
4
Yangchuanosaurus spp. DONG, CHANG, LI & ZHOU, 1978
TAXA INCLUDED IN SUPPLEMENTARY ANALYSES
Deltadromeus agilis SERENO, DUTHEIL, IAROCHENE, LARSSON, LYON, MAGWENE, SIDOR, VARRICCHIO & WILSON, 1996
“Megalosaurus” hesperis WALDMAN, 1974
Unenlagia comahuensis NOVAS & PUERTA, 1997
1 -PAUL (1988), SMITH (1998), CHURE (1998), HENDERSON (1998) and BAKKER (1998) argue that more than one species (or higher level
taxa) are present in the Morrison Formation genus Allosaurus. 2 - WELLNHOFER (1993) suggests that two species are present in the
Solnhofen Lithographic Limestone genus Archaeopteryx: A. lithographica MEYER, 1861 and A. bavarica WELLNHOFER, 1993. 3 - Sinraptor includes two species, S. dongi CURRIE & ZHAO, 1993a and S. hepingensis (GAO, 1992). 4 - Yangchuanosaurus includes two
species, Y. shangyouensis DONG, CHANG, LI & ZHOU, 1978 and Y. magnus DONG, ZHOU & ZHANG, 1983.
8
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
basal sauropodomorphs (as well as those in common in all three of these and in more distantly related
forms such as ornithischians, basal ornithodirans,
and non-ornithodiran archosauriforms) were considered to the primitive relative to the taxa in the current study. Character states found in the taxa in the
present analysis and in some (but not all three) of
Eoraptor, herrerasaurids, and basal sauropodomorphs were coded as derived for the ingroup
forms: these are discussed below. Finally, character
states found in some or all of the ingroup taxa but
none of the three potential sister taxa are coded as
derived. Based on these codings, an "all zero" outgroup with all primitive states was created to approximate a compromise ancestral condition. The
character states observed in herrerasaurids (primarily Herrerasaurus) and basal sauropodomorphs
("prosauropods") are also included here, and use of
these taxa rather than an "all zero" outgroup are
briefly discussed. In the longer study in preparation
by the present author, differences between the results using this method and various arrangements of
the known potential outgroups will be examined.
Fig. 2 - Alternative phylogenies for sister taxon to theropods used in this study. A - Herrerasaurids and Eoraptor
share a more recent common ancestor with (more advanced) theropods than do sauropodomorphs, after
NOVAS (1994, 1997a), SERENO (1997, 1998), and SERENO
& NOVAS (1992, 1994). B - Sauropodomorphs share a
more recent common ancestor with theropods than do either herrerasaurids or Eoraptor, after HOLTZ & PADIAN
(1995) and BONAPARTE & PUMARES (1995).
The characters employed in this analysis are
listed in APPENDIX I. 386 characters were used (135
craniodental, 75 axial, 74 pectoral and forelimb, and
102 pelvic and hindlimb characters). 301 of the characters are coded as binary; 85 as multistate. All binary characters were considered unordered;
multistate characters were considered unordered
unless otherwise indicated (see APPENDIX I). Descriptors of facial pneumatic structures follow WITMER (1997). In APPENDIX I a brief description of the
primitive and derived state(s) is provided; in work in
preparation, each of these characters will be described in greater detail (as in GRANDE & BEMIS
(1998) or WILSON & SERENO (1998)). However,
some characters of particular phylogenetic significance in this or previous studies will be discussed
below. APPENDIX II is the data matrix analyzed.
phylogeny, Eoraptor and the herrerasaurids would
be considered true theropods (given the definition of
Theropoda following GAUTHIER (1986): birds and all
taxa sharing a more recent common ancestor with
birds than with sauropodomorphs). Among the potential derived characters supporting such an hypothesis are prominent postaxial cervical
epipophyses; greatly reduced manual digits IV and
V; an intramandibular joint; and a distal enlargement
of the pubis (NOVAS, 1994, 1997a; SERENO, 1997).
The resulting data matrix was analyzed using
PAUP 3.1.1 (SWOFFORD, 1993). The unwieldy size
of the data matrix required the use of the Heuristic
search option, as the Branch-and-Bound and Exhaustive search methods would require prohibitively
long run times given current computer calculation
speeds. In all cases, random branch addition was
run with thirty replicates, to reduce the chance of
falsely accepting a local rather than global minimum
(that is, a tree or set of trees with less than maximum
parsimony). The set of most parsimonious trees discovered under these runs was analyzed using MacClade 3.07 (MADDISON & MADDISON, 1997), in order
to determine character state distribution under accelerated and delayed transformation (ACCTRAN
and DELTRAN, respectively) optimizations and to
examine alternate topologies. The tree length, con-
Alternatively, HOLTZ & PADIAN (1995) and BONA& PUMARES (1995) have argued that sauropodomorphs share a more recent common ancestor
with the forms in this analysis than do Eoraptor or
herrerasaurids (Fig. 2B). Derived characters supporting such an hypothesis include pollex ungual
larger than other manual unguals; manual digit II
longest digit in the hand; vertebrae 6-9 longest in the
cervical column; and a distal expansion of the ischium.
PARTE
The question of the sister taxon to the forms used
here is the subject of a separate study by the present
author and PADIAN, in preparation. For this analysis,
a compromise outgroup was used. Character states
shared in common in Eoraptor, herrerasaurids, and
9
T.R. HOLTZ, JR.
TABLE II
Phylogenetic taxonomic definitions used in this study, based primarily on PADIAN, HUTCHINSON & HOLTZ (1999)
TAXON
TYPE
A
B
Theropoda MARSH, 1881
Neotheropoda BAKKER, 1986
Ceratosauria MARSH, 1884
Coelophysoidea HOLTZ, 1994
Neoceratosauria NOVAS, 1992
Abelisauroidea NOVAS, 1992
Abelisauridae BONAPARTE & NOVAS, 1985
Tetanurae GAUTHIER, 1986
Avetheropoda PAUL, 1986
Carnosauria HUENE, 1920
Allosauroidea CURRIE & ZHAO, 1993a
Allosauridae MARSH, 1879a
Sinraptoridae CURRIE & ZHAO, 1993a
Coelurosauria HUENE, 1914
Maniraptoriformes HOLTZ, 1996b
Arctometatarsalia HOLTZ, 1994
Bullatosauria HOLTZ, 1994
Ornithomimosauria BARSBOLD, 1976b
Maniraptora GAUTHIER, 1986
Oviraptorosauria BARSBOLD, 1976b
Paraves SERENO, 1997
Eumaniraptora PADIAN, HUTCHINSON & HOLTZ, 1999
Deinonychosauria COLBERT & RUSSELL, 1969
Avialae GAUTHIER, 1986
Aves LINNE, 1758
Metornithes PERLE, NORELL, CHIAPPE & CLARK, 1993
Stem
Node
Stem
Stem
Stem
Stem
Node
Stem
Node
Stem
Node
Stem
Stem
Stem
Node
Stem
Node
Node
Stem
Node
Stem
Node
Stem
Stem
Node
Node
Neornithes
Ceratosaurus
Ceratosaurus
Coelophysis
Ceratosaurus
Carnotaurus
Abelisaurus
Neornithes
Neornithes
Allosaurus
Allosaurus
Allosaurus
Sinraptor
Neornithes
Ornithomimus
Ornithomimus
Ornithomimus
Ornithomimus
Neornithes
Oviraptor
Neornithes
Deinonychus
Deinonychus
Neornithes
Archaeopteryx
Mononykus
Cetiosaurus
Neornithes
Neornithes
Ceratosaurus
Coelophysis
Ceratosaurus
Carnotaurus
Ceratosaurus
Allosaurus
Neornithes
Sinraptor
Sinraptor
Allosaurus
Allosaurus
Neornithes
Neornithes
Troodon
Pelecanimimus
Ornithomimus
Chirostenotes
Oviraptor
Neornithes
Neornithes
Deinonychus
Neornithes
Neornithes
Taxon, taxon type, and reference taxa for phylogenetic definitions employed in this paper. Definitions and justifications discussed in PADIAN,
HUTCHINSON & HOLTZ (1999). Type: Stem, stem-based; Node, node-based. Stem-based taxon definitions are of the form "refer-
ence taxon A and all taxa sharing a more recent common ancestor with reference taxon A than with reference taxon B". Node-based
taxon definitions are of the form "all descendants of the most recent common ancestor of reference taxa A and B."
sistency index (CI), retention index (RI), and rescaled consistency index (RC) for each tree was calculated on MacClade, while the homoplasy index
(HI) was calculated on PAUP: see pp. 364-368 of
MADDISON & MADDISON (1997) for the differences in
metric calculations between these two programs.
Bremer support values were calculated by means of
the AutoDecay program, version 4.0 (ERIKSSON,
1998).
RESULTS
Names of clades found in this analysis are based
on the standardized phylogenetic definitions provided in PADIAN, HUTCHINSON & HOLTZ (1999), and
are summarized in TABLE II. See the reference
above, SERENO (1998), and the references therein
for a discussion of the principles of phylogenetic taxonomy. SERENO (1997, 1998) provides some different definitions for the same names, and some
different names for the same definitions as used
here: see PADIAN, HUTCHINSON & HOLTZ (1999) for
discussion of these taxonomic conflicts.
10
The primary analysis produced a set of 20
equally parsimonious trees of 1404 steps. The CI for
these trees was 0.442, the RI was 0.618, the RC was
0.273, and the HI was 0.647. The strict consensus of
these trees is presented in Fig. 3A. The normalized
consensus fork index, a measure of consensus tree
resolution (number of nodes in the strict consensus
tree over number of nodes in a fully resolved dichotomous tree of the same number of taxa:
COLLESS, 1980) is 0.825. Use of Herrerasauridae as
an outgroup does not alter tree topology, but instead
produces the same 20 trees at a tree length of 1412,
a CI of 0.439, an RI of 0.614, an RC of 0.269, and an
HI of 0.649. Use of basal sauropodomorphs ("prosauropods") yields 120 trees of tree length 1458, CI
0.433, RI 0.619, RC 0.268, and HI 0.658: the tree topologies are identical to those analyses using the all
zero outgroup or the herrerasaurid outgroup, except
that Megalosaurus, Torvosaurus, Eustreptospondy-
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
lus, and Piatnitzkysaurus are equally parsimoniously placed in six different possible configurations
relative to the Afrovenator-avetheropod clade. The
increase in tree length and decrease in CI, RI, RC,
and HI is explained by homoplasy in some conditions found in herrerasaurids or prosauropods and
various derived theropod clades, but not hypothesized for basal neotheropods under either accelerated or delayed transformation.
dontidae is more closely related to dromaeosaurids
and birds than to ornithomimosaurs was selected for
the summary cladogram, as this topology more
closely reflects the results of most other workers
(e.g., G AUTHIER , 1986; N OVAS , 1992; S ERENO ,
1997, 1998; SUES, 1997; MAKOVICKY & SUES, 1998;
FORSTER et al., 1998; NOVAS & POL, in press). However, the analyses of HOLTZ (1994) and PÉREZMORENO et al. (1994) (and previously the nonnumerical study of THULBORN (1984)) recovered a
troodontid-ornithomimosaur clade to the exclusion
of dromaeosaurids or birds, which represents the al-
Examination of the resulting trees in the primary
analysis (using the all zero outgroup) found that the
polytomies in the strict consensus cladogram could
be decomposed as instability at three different regions. The variability at each these three regions is
independent of the variability at the other three. This
independence (five different possible topologies in
one region of the tree, two in the other two) results in
the 20 different equally parsimonious trees recovered.
In the first of these cases, the instability occurred
because of the incompletely known taxon Proceratosaurus bradleyi, known only from cranial remains
from the Middle Jurassic (Bathonian) of England,
was found to occupy five different possible positions
with respect to the other basalmost coelurosaurs
Gasosaurus and Dryptosaurus without change in
tree length, CI, or other tree metrics (Fig. 3B). The
second region of tree instability concerns two alternative placements for the fragmentary AptianAlbian form Microvenator celer, as either the sister
taxon to Oviraptoridae or the sister taxon to the clade
Oviraptoridae plus Caenagnathidae (Fig. 3C).
The third instability is of greater interest, as it concerns two alternative placements of the well known
taxon Troodontidae as the sister group to two very
different clades: a sister group relationship with the
dromaeosaurid-bird clade on the one hand, and a
sister group relationship with Ornithomimosauria on
the other. Both topologies are equally parsimonious,
and result in the apparent lack of resolution among
maniraptoriform coelurosaurs shown in Fig. 3A. In
fact there is much greater structure than revealed
under strict consensus: all the other taxa have resolved positions relative to each other, with the exception of Troodontidae itself. The structure within
Maniraptoriformes is presented in Fig. 4, which
shows the two alternative positions for troodontids.
Fig. 3 - Relationships among theropod dinosaurs
based on maximum parsimony analysis of 386 morphological characters. A - Tree represents the strict consensus of twenty equally most parsimonious trees (tree
length=1404, CI=0.442, RI=0.618, RC=0.273, HI=0.647,
normalized consensus fork index=0.825). Characters
used in analysis included in Appendix I. B - Dashed lines
indicate the five equally parsimonious alternative positions
of Proceratosaurus (Pr.) relative to Gasosaurus (Gaso.)
and Dryptosaurus (Drypto.). C - Dashed lines indicate the
two equally parsimonious alternative positions of Microvenator (Micro.) relative to Oviraptoridae (Ovir.) and
Caenagnathidae (Caen.).
A summary cladogram (Fig. 5) is used to discuss
the distribution of characters in the present analysis.
This represents one of the twenty most parsimonious trees in the analysis. For this summary cladogram, Proceratosaurus was placed between
Gasosaurus and Dryptosaurus, and Microvenator
placed as outside of an oviraptorid-caenagnathid
clade (the position preferred by the lower stratigraphic position of this form). A tree in which Troo-
11
T.R. HOLTZ, JR.
England) results in 220 trees 1 step longer than the
primary analysis (TL 1405, other metrics identical to
main analysis). "M." hesperis is equally parsimoniously placed as the sister group to nodes G, H, I, J, K,
or L or as the sister group to Spinosauridae, Megalosaurus, Eustreptospondylus, Torvosaurus, or Piatnitzkysaurus.
ternative topology of this study. Characters supporting this position will be discussed below.
Note that this particular configuration (Fig. 5) is
used to facilitate discussion only, and is not preferred by the data analysis over the nineteen other
potential arrangements. Future analyses may help
to resolve the uncertainty with regards to the relationships presented here (if not overturn some or all
of those in the current analysis, pending the addition
of new data).
In the following section, the character states
found at each node are listed. The nodes are listed
by the corresponding letter from the summary cladogram (Fig. 5). Taxon names are listed for some of the
clades mentioned here, following the definitions proposed in PADIAN, HUTCHINSON & HOLTZ (1999), and
listed in TABLE II. Stem-defined names are underlined in the description headings, whereas nodedefined names are not. Where two names are listed,
the first represents the stem-defined taxon name
and the second is the node-defined taxon name. Not
all nodes are named.
Addition of the fragmentary Deltadromeus of the
Cenomanian of northern Africa results in 80 equally
parsimonious trees two steps longer than the primary analysis (tree length 1406, CI 0.441, RI 0.618,
RC 0.272, HI 0.627). This addition does not change
the tree topology: rather, Deltadromeus is equally
parsimoniously placed in five different positions.
These positions are: as the sister taxon to Ornitholestes; as the sister taxon to Coelurus, or as the
sister taxon to nodes ff, gg, or jj. Similarly, inclusion
of the poorly known Unenlagia of the Late Cretaceous of Argentina does not alter overall tree structure. Instead, its presence results in a total of 50
trees eleven steps longer (tree length 1415, CI
0.438, RI 0.619, RC 0.271, HI 0.649) than that in the
primary analysis. Unenlagia is equally parsimoniously placed as the sister taxon to Rahonavis, as the
sister taxon to the Archaeopteryx-Metornithes clade
(node mm), or as the sister group to Ornithothoraces. Inclusion of "Megalosaurus" hesperis (known
only from cranial material of the Middle Jurassic of
For purposes of description of the nodal character state changes, ALL refers to those derived character states present in all optimizations, ACCTRAN
refers to those derived character states present at
that node under accelerated transformation (i.e., at
the basalmost point on the tree where this character
could appear without requiring additional evolutionary steps), and DELTRAN refers to those present at
that node under delayed transformation (i.e., at the
terminal-most point on the tree where this character
could appear without requiring additional evolutionary steps). Characters are listed by the character
number (before the period) and state (after the period): for example, 17.1 refers to the condition "maxillary fenestra present." Characters listed with a
minus sign (-) before them represent a reversal to a
state found at a more basal position in the tree: for
example, -17.0 would indicate "maxillary fenestra
absent" for a form nested within a clade otherwise
characterized by the state 17.1.
NODE A. THEROPODA - NEOTHEROPODA
ALL: 10.1 Premaxilla and nasal do not meet subnarially; 26.1 Narial prominences present; 30.1 Lacrimal broadly exposed on skull roof; 45.1 Orbit oval
or key-shaped, rounded dorsally, constricted ventrally; 54.1 Postorbital frontal process about same
level or slightly higher than squamosal, producing Tshaped postorbital; 73.1 Vomera fused rostrally;
110.1 Reduced overlap of dentary onto postdentary
bones; 111.1 Intramandibular joint; 117.1 Rostral
prong of angular penetrates the dentary-splenial
cavity; 137.1 First intercentrum with large occipital
fossa (two or less times as wide as tall) and small
odontoid notch; 138.1 Second intercentrum cranial
articulation with first intercentrum with broad crescentic fossa; 148.1 Postaxial cervical pleurocoels,
one pair present; 152.1 Caudal cervical epipophy-
Fig. 4 - Maniraptoriform theropod cladogram as recovered in this analysis. Solid lines indicate positions shared
by all twenty most parsimonious trees; dashed line indicates two equally parsimonious positions for Troodontidae
(as bullatosaurian arctometatarsalians or as paravian
maniraptorans). See text for discussion.
12
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
Fig. 5 - Summary cladogram of theropod relationships, representing one of the twenty equally parsimonious trees in
this analysis. See text for character state changes at each node, listed by letter indicated. Note that the topology shown
here is not preferred by the data over other potential topologies, but was instead chose to facilitate discussion of character
distribution. See text for details. A - Ceratosauria and non-maniraptoriform Tetanurae section of tree. B - Maniraptoriformes.
about 50%; 312.1 Pubic blade at least six times as
long as broad; 329.1 Femur shape bowed in convex
arc with less pronounced sigmoidality; 336.1 Proximalmost point of anterior trochanter below femoral
head; 372.1 Metatarsal V vestigial or absent; 378.1
Metatarsal I reduced but retains phalanges; 379.1
Metatarsal I placed near midpoint of metatarsal II
shaft.
ses elongate; 164.1 Longest postaxial cervicals VIIX; 180.1 Cranial and median dorsal pleurocoels,
one pair present; 181.1 Presacral pleurocoels camerate; 185.3 Number of sacrals five; 197.1 Transition point in distal half of tail; 219.1 Coracoid biceps
tubercle conspicuous and well developed; 256.3
Metacarpal V absent; 257.1 Metacarpal IV present,
without ungual; 270.1 Digit II longest in manus;
271.1 Penultimate phalanx longest nonungual phalanx; 286.1 Ilium dolichoiliac; 290.1 Brevis fossa
deep; 301.1 Acetabular height/craniocaudal length
ACCTRAN: 16.1 Promaxillary fenestra present,
visible in lateral view; 24.1 Nasal participates in an-
13
T.R. HOLTZ, JR.
mologous in Herrerasauridae and Neotheropoda,
as SERENO & NOVAS (1994) noted the geometry of
the joint are reversed between the two taxa: in herrerasaurids, the splenial has a concave surface which
slides against the convex ventral margin of the angular, while in neotheropods the splenial has a convex dorsal surface which slides against a concave
depression on the angular (p. 471). No characters
were observed which were present in Eoraptor and
basally within Neotheropoda that were not also
found in herrerasaurids. Under this scheme, the
seven derived characters shared by neotheropods
and basal sauropodomorphs mentioned previously
would either have had to developed independently
in these two lineages, or have been present and subsequently lost in Eoraptor and Herrerasauridae.
torbital cavity; 44.1 Orbit shorter than internal antorbital fenestra length; 58.1 Jugal participates in
internal antorbital fenestra; 67.1 Quadrate foramen
reduced or absent; 81.1 Ventral ectopterygoid recess present and comma-shaped; 253.1 Distal carpal I block does not overlap metacarpal II dorsally,
but does so ventrally; 254.1 Distal carpal I fused to
distal carpal II; 278.1 Pollex larger than other manual unguals; 317.2 Pubic boot rounded, angle between shaft and caudal portion of boot acute; 319.1
Pubic boot present, less than 30% as long as pubic
shaft; 339.1 Muscle scar in craniodistal region of femur present, non-elliptical in shape; 344.1 Ectocondylar tuber proximodistally long, pronounced, and
extends almost to distal end of femur; 350.1 Crista
fibularis present, not well developed.
DELTRAN: None
As defined by GAUTHIER (1986) Theropoda is a
stem-based taxon, comprised of birds and all taxa
sharing a more recent common ancestor with birds
than with sauropodomorphs. The term Neotheropoda BAKKER, 1986 has been used by SERENO et al.
(1993), S ERENO (1997, 1998), and P ADIAN ,
HUTCHINSON & HOLTZ (1999) for the node-defined
taxon comprised of all descendants of the most recent common ancestor of Ceratosauria and Tetanurae (or, more explicitly, of Ceratosaurus and
Neornithes: see P ADIAN , H UTCHINSON & H OLTZ
(1999)).
How and if the diagnosis of Theropoda and Neotheropoda differ hinges on the question of the immediate sister group of Neotheropoda (see above, and
Fig. 2). The following characters from the above list
are shared by basal theropods and basal sauropodomorphs, and would therefore diagnose a clade
more inclusive than Theropoda if herrerasaurids
and Eoraptor are not theropods themselves: 10.1,
67.1, 117.1, 164.1, 270.1, 278.1, and 301.1. Those
shared in common with Herrerasauridae and/or
Eoraptor (see below) would either have evolved
convergently between these Triassic forms and true
theropods or have been present in the common ancestor of these taxa, theropods, and sauropodomorphs, and subsequently lost in the latter. The
remaining characters of the above list would be diagnostic of Theropoda and Neotheropoda (which
would share the same diagnosis, as all known theropods would be neotheropods).
If instead herrerasaurids and Eoraptor are true
theropods, then character 197.1 (Transition point in
distal half of tail) would be considered synapomorphic for Theropoda, and the following for an
herrerasaurid-neotheropod clade: 67.1, 111.1,
117.1, 138.1, and 152.1. It should be noted, however, that the condition represented by character
state 111.1 (Intramandibular joint) may not be ho-
There remain many characters found basally in
Ceratosauria and Tetanurae which are not also
found in basal sauropodomorphs, Eoraptor, or Herrerasauridae. These are diagnostic for Neotheropoda, regardless of either particular sister group
scenario.
NODE B. CERATOSAURIA
ALL: 31.2 Lacrimal prominences comprised of
ridge continuous with raised surface of lateral edge
of nasals; 148.2 Postaxial cervical pleurocoels, two
pairs present; 170.1 Dorsal transverse processes
strongly backturned caudally and triangular in dorsal view; 180.2 Cranial and median dorsal pleurocoels, two pairs present; 193.1 Ventral groove in
cranial caudals; 201.1 Shaft of cervical ribs extremely long (four or more times centrum length) and
slender; 291.2 Brevis fossa distal end broad; 300.1
Supracetabular shelf on ilium present; 309.1 Pubis
orientation propubic, proximal portion of shaft approximately 30 degrees from horizontal; 321.1 Ischial antitrochanter large; 335.1 Anterior trochanter
conical prominence; 338.1 Trochanteric shelf of femur well developed; 340.1 Medial epicondyle (=
mediodistal crest) of femur pronounced, extends
one quarter or more the length of the femoral shaft;
342.1 Groove in lateral condyle of femur; 345.1 Sulcus along medial side of base of crista tibiofibularis;
356.1 Sulcus in proximomedial region of fibula;
359.1 Anterior surface of distal fibula overlaps ascending process of astragalus cranially; 375.1
Metatarsal III dorsal surface area clearly larger than
either metatarsal II or metatarsal IV
ACCTRAN: 189.1 Synsacrum present in adults;
285.1 Pelvic girdle sutures fused in adults; 366.1 Astragalocalcaneum (astragalus fused to calcaneum)
DELTRAN: 24.1 Nasal participates in antorbital
cavity; 44.1 Orbit shorter than internal antorbital fenestra length; 67.1 Quadrate foramen reduced or ab-
14
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
sent; 319.1 Pubis and ischium proximal shafts
narrow
As first proposed by GAUTHIER (1986), Ceratosauria forms a major clade of theropods containing
such forms as Ceratosaurus, Dilophosaurus, and
Coelophysis. As in HOLTZ (1994), Ceratosauria is divided into two primary branches, the relatively
gracile Late Triassic and Early Jurassic Coelophysoidea and the more robust Late Jurassic and Cretaceous Neoceratosauria.
ROWE (1989) and ROWE & GAUTHIER (1990) considered the Early Jurassic (?SinemurianPliensbachian) Sarcosaurus woodi ANDREWS 1921
to be a ceratosaur. This taxon (known only from a
partial pelvis) was not included in the present analysis, so its relationship with other ceratosaurs is not
resolved here.
NODE C. COELOPHYSOIDEA
ALL: 11.1 Subnarial gap; 26.2 Narial prominences comprised of paired ridges along lateral
edges of nasals; 64.1 Dorsal ramus of quadratojugal
does not contact squamosal; 142.1 Axial parapophyses reduced; 143.1 Axial diapophyses absent;
145.1 Axial pleurocoels absent; 151.1 Epipophyses
on cervical vertebrae placed proximally; 234.1 Humeral torsion present; 235.1 Humeral shaft sigmoid;
308.1 Pubic fenestra ventral to obturator foramen
than with some other coelophysoid genus. ROWE
(1989) found evidence that Liliensternus liliensterni
(H UENE , 1934) was a coelophysoid (see also
RAUHUT & HUNGENBÜHLER, 1998), as did SERENO
(1997), who also included Procompsognathus
FRAAS, 1913 and Segisaurus CAMP, 1936 in this
clade. CARPENTER (1997) has described a large
Late Triassic North American coelophysoid Gojirasaurus quayi. These taxa were not included in this
study: future analyses will hopefully clarify the relationships of these taxa to each other and to other
ceratosaurs.
The Late Jurassic African Elaphrosaurus bambergi does share numerous derived features with
coelophysoids in general, and coelophysids in particular (e.g., 154.1, 160.1, 165.1, 376.2). Unlike
PAUL (1988) and NOVAS (1992), Elaphrosaurus was
not found to be member of Coelophysoidea in this
analysis, but rather was hypothesized to share a
more recent common ancestry with abelisauroids
and Ceratosaurus, as in HOLTZ (1994) and SERENO
(1997). However, moving this taxon to a sister group
position with node C or with Coelophysidae requires
only one additional evolutionary step, and additional
details may reveal that Elaphrosaurus was a latesurviving coelophysoid. (See also Discussion).
The newly discovered taxon Genusaurus serus
A CCAIRE et al., 1995 was considered by those
authors to be a ceratosaur closer to Coelophysis
than to Ceratosaurus. If such a position were confirmed, it would indicate the first known Cretaceous
(middle Albian) coelophysoid. However, although
this form demonstrates some ceratosaurian features (pelvic girdle sutures fused, proximal portion of
pubic shaft approximately 30 degrees from the horizontal, trochanteric shelf well developed, sulcus
along medial side of base of crista tibiofibularis), it
does not show any unambiguously coelophysoid
feature. In fact, a well-excavated proximal region of
the fibular medial face is not known in coelophysoids, but is documented in the neoceratosaur Carnotaurus. Additional study may demonstrate
Genusaurus to be a mid-Cretaceous European abelisauroid.
ACCTRAN: 27.1 Paired crescentic crests
formed by nasal and lacrimal prominences; 127.1
Dentary teeth more numerous and smaller than
maxillary teeth; 331.1 Femoral head transversely
elongate
DELTRAN: -317.0 Pubic boot absent; -344.0 Ectocondylar tuber proximodistally short, proximally
placed
As in ROWE (1989), NOVAS (1992), HOLTZ (1994),
and SERENO (1997), a clade comprised of Dilophosaurus and the Coelophysis-Syntarsus clade (Coelophysidae) to the exclusion of other ceratosaurs
was supported.
Paired crescentic crests formed by nasal and lacrimal prominences may be synapomorphic for Coelophysoidea (SERENO, 1997) as they are present in
Dilophosaurus and Syntarsus kayentakatae ROWE
1989 (although not in Coelophysis nor in Syntarsus
rhodesiensis). Such crests were the primary evidence for placing "Dilophosaurus" sinensis HU,
1993 in that genus: since these structures are found
in other coelophysoids, and given certain other derived anatomical differences between these taxa
(tooth row rostral to the orbit in D. wetherilli, five premaxillary teeth in "D." sinensis, among others), it
may be that the Chinese taxon does not share a
more recent common ancestor with Dilophosaurus
NODE D. NEOCERATOSAURIA
ALL: 149.1 Cervical epipophyses powerfully developed and prong-shaped; 185.4 Six sacrals.
ACCTRAN: 4.1 Premaxillary symphyseal region
U-shaped in ventral view; 5.1 Premaxilla subnarially
very deep, main body taller dorsoventrally than long
rostrocaudally; 62.1 Infratemporal fenestra about
twice as large as the area of the orbit in lateral view;
66.1 Quadrate-quadratojugal suture fused; 68.1
Quadrate dorsal ramus greater than height of orbit;
70.1 Quadrate articulation projects well caudal to
15
T.R. HOLTZ, JR.
NODE F. ABELISAUROIDEA - ABELISAURIDAE
the caudal point of the occipital condyle; 83.1 Nuchal
crest pronounced; 84.1 Supraoccipital with very pronounced, strongly demarcated median ridge on occipital surface; 103.1 Occipital condyle constricted
neck; 116.1 Horizontal shelf on lateral surface of
surangular, rostral and ventral to the mandibular
condyle, prominent and extends laterally; 144.1 Axial epipophyses prominent; 181.2 Presacral pleurocoels camellate; 231.1 Ulna/femur length ratio less
than 28%; 232.1 Radius/humerus length ratio less
than 50%; 315.2 Pubic boot rounded, angle between shaft and caudal portion of boot acute; 355.1
Proximal region of fibular medial face shallow and
not conspicuous.
ALL: 15.2 Maxillary antorbital fossa greatly reduced in size, not extending much beyond rim of the
external antorbital fossa; 26.3 Narial prominences
knobby rugosities across dorsal and lateral surface
of nasals, extending onto dorsalmost surface of
maxillae; 35.2 Lacrimal dorsal (= rostral) ramus absent; 37.1 Prefrontals reduced or absent; 41.1
Frontal-frontal suture fused; 42.2 Frontal-parietal
suture on dorsal surface of skull fused, suture indistinguishable; 50.1 Postorbital-lacrimal contact
broad; 53.1 Postorbital suborbital flange
ACCTRAN: -117.0 Rostral prong of angular does
not penetrate the dentary-splenial cavity; 161.1 Midcervical centra greater than 20% broader than tall;
176.1 Dorsal centrum transverse section wider than
high; 188.1 Sacral neural spines fused to form lamina; 211.1 Scapular blade long, slender (four times
or more longer than midshaft width) and strap-like;
212.1 Distal expansion of scapula reduced, less
than width of proximal end of scapula; 213.1 Acromion in scapula reduced; 229.1 Humerus/scapula
length ratio less than 65%; -234.0 Humeral torsion
absent; 237.1 Internal tuberosity on proximal end of
humerus well differentiated and angular; 249.1 Ulnar facet for radius transversely expanded and concave; 269.1 Metacarpal-phalangeal joints not
hyperextensible, extensor pits on metacarpals I-III
reduced; 304.1 Iliac-ischial articulation smaller than
iliac-pubic articulation; -309.0 Pubis orientation propubic, shaft approximately 45 degrees from horizontal; 330.1 Femoral head approximately 90 degrees
from shaft (head directed horizontally); 355.2 Proximal region of fibular medial face well excavated
DELTRAN: 189.1 Synsacrum present in adults;
285.1 Pelvic girdle sutures fused in adults; 366.1 Astragalocalcaneum (astragalus fused to calcaneum);
376.2 Metatarsal III dorsal surface dumbbell shaped
(cranial and (especially) plantar surfaces expanded
to slightly overlap surfaces of metatarsals II and IV).
NODE E.
ALL: 147.1 Cervical centra surfaces markedly
opisthocoelous; 150.1 Cervical epipophyses directed dorsolaterally and taller than neural spine;
185.5 More than six sacrals; 186.1 Sacrals III-V
transversely compressed; 302.1 Ilium about as long
as femur; 334.1 Anterior trochanter present, separated from femoral head by cleft; 337.1 Fourth trochanter of femur present, but little developed.
ACCTRAN: 357.1 Cranial protuberance on fibula
below expansion.
DELTRAN: 4.1 Premaxillary symphyseal region
U-shaped in ventral view; 5.1 Premaxilla subnarially
very deep, main body taller dorsoventrally than long
rostrocaudally; 58.1 Jugal participates in internal
antorbital fenestra; 62.1 Infratemporal fenestra
about twice as large as the area of the orbit in lateral
view; 68.1 Quadrate dorsal ramus greater than
height of orbit; 70.1 Quadrate articulation projects
well caudal to the caudal point of the occipital condyle; 83.1 Nuchal crest pronounced; 84.1 Supraoccipital with very pronounced, strongly demarcated
median ridge on occipital surface; 103.1 Occipital
condyle constricted neck; 116.1 Horizontal shelf on
lateral surface of surangular, rostral and ventral to
the mandibular condyle, prominent and extends laterally; 144.1 Axial epipophyses prominent; 160.2
Midcervical centra length less than twice diameter of
cranial face; 315.2 Pubic boot rounded, angle between shaft and caudal portion of boot acute; 317.2
Pubic boot rounded, angle between shaft and caudal portion of boot acute; 339.1 Muscle scar in
craniodistal region of femur present, non-elliptical in
shape.
DELTRAN: 16.1 Promaxillary fenestra present,
visible in lateral view; 66.1 Quadrate-quadratojugal
suture fused
As in NOVAS (1992), HOLTZ (1994), and SERENO
(1997, 1998), a clade comprised of Ceratosaurus
and abelisaurids was supported here. As in the studies by the latter two authors, Elaphrosaurus was
found to be part of this clade. However, as noted
above, support for a coelophysoid placement of this
taxon is nearly as strong.
As HOLTZ (1994) noted, many features uniting
neoceratosaurs are also found in tetanurines: under
the most parsimonious distributions of derived character states, these are explainable either as convergences between Neoceratosauria and Tetanurae or
as basal neotheropod characters subsequently lost
in Coelophysoidea. Alternatively, Ceratosaurus and
Abelisauridae may share a more recent common ancestor with tetanurines than with Coelophysoidea:
however, such a phylogenetic scenario requires
several additional steps given the present data base
(see Discussion below).
16
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
than long craniocaudally; 225.1 Furcula; 233.1 Manus/(humerus + radius) length ratio greater than
66%; 255.1 Semilunate carpal block fully developed
with transverse trochlea; 257.2 Metacarpal IV present, without phalanges; 261.1 Articular surface between metacarpals I and II extends well into
diaphysis of metacarpal I; 263.1 Metacarpal III
clearly shorter than metacarpal II; 268.1 Metacarpal
IV less than half length of metacarpal II; 279.1 Pollex
ungual greater than three times longer than height of
articular facet; 283.1 Manual ungual length extremely long; 334.1 Anterior trochanter present,
separated from femoral head by cleft; 346.1 Cnemial process arises out of the lateral surface of tibial
shaft; 351.1 Crista fibularis proximally placed; 353.1
Fibula closely appressed to tibia throughout main
shaft; 360.1 Fibula distal end less than twice craniocaudal width at midshaft, and consequently astragalar cup for fibula reduced; 364.1 Astragalar distal
condyles oriented cranioventrally
Although this study did not examine various other
neoceratosaurs, CORIA & SALGADO (1998) describe
a new taxon assignable to this clade. That study details several forms that would belong to the stembased taxon Abelisauroidea (all taxa sharing a more
recent common ancestor with Carnotaurus than with
Ceratosaurus) which were not included in this analysis, as well as additional character evidence for relationships within Neoceratosauria. SAMPSON et al.
(1998) have described excellent, well-preserved remains of the abelisaurid Majungatholus atopus
SUES & TAQUET, 1979, a form which they consider
likely to be the sister taxon to Carnotaurus within
Abelisauridae. This new material will greatly increase our knowledge of neoceratosaur osteology.
NOVAS (1997c) has suggested that Carcharodontosaurus and Giganotosaurus were abelisaurids or abelisaurid relatives. Such a relationship
was not supported here, but is discussed below.
DELTRAN: 16.1 Promaxillary fenestra present,
visible in lateral view; 160.2 Midcervical centra
length less than twice diameter of cranial face; 291.1
Brevis fossa distal end tapered; 338.2 Trochanteric
shelf of femur absent; 339.1 Muscle scar in craniodistal region of femur present, non-elliptical in
shape; 301.1 Crista fibularis present, not well developed; 344.1 Ectocondylar tuber proximodistally
long, pronounced, and extends almost to distal end
of femur.
NODE G. TETANURAE
ALL: 13.1 Rostral ramus of maxilla present, dramatic change in curvature of rostrodorsal surface of
maxilla rostral to dorsal ramus forming concave surface; 14.2 Rostral ramus as long or longer rostrocaudally as dorsoventrally; 34.1 Slot in ventral
process of lacrimal for jugal; 35.1 Lacrimal dorsal
(rostral) ramus dorsoventrally pinched and narrow;
133.1 Caudalmost maxillary tooth position rostral to
orbit; 147.1 Cervical centra surfaces opisthocoelous; 211.1 Scapular blade long, slender (four times
or more longer than midshaft width) and strap-like;
212.1 Distal expansion of scapula reduced, less
than width of proximal end of scapula; 240.1 Humeral ends well expanded, greater than 150% midshaft diameter; 304.1 Iliac-ischial articulation
smaller than iliac-pubic articulation; 305.1 Pubic peduncle of ilium more developed craniocaudally than
mediolaterally; 341.1 Extensor groove in craniodistal region of femur present, but shallow and not conspicuous; 352.1 Tibia distal end expanded to back
calcaneum; 362.1 Astragalar ascending process
mediolaterally reduced, craniocaudally wide, and
proximodistally low ("allosauroid condition")
NODE H.
ALL: 234.1 Humeral torsion present; 249.1 Ulnar
facet for radius transversely expanded and concave; 330.1 Femoral head approximately 90 degrees from shaft (head directed horizontally).
ACCTRAN: 38.1 Prefrontal-frontal peg-insocket suture; 48.1 Postorbital ventral process
broader transversely than rostrocaudally with Ushaped cross-section; 61.1 Jugal recess; 64.2
Broad contact between dorsal ramus of quadratojugal and lateroventral ramus of squamosal; 76.1
Palatine tetraradiate; 78.1 Palatine recesses; 123.1
Retroarticular process of articular faces caudally;
197.2 Transition point in proximal half of tail; 215.1
Scapulacoracoid cranial margin with pronounced
notch between acromial process and coracoid;
221.1 Sternal plates fused medially; 222.1 Sternum
carina present; 223.2 Sternum wider mediolaterally
17
ACCTRAN: 22.1 Pneumatic excavation without
fenestra in cranial portion of maxillary antorbital
fossa; 120.1 Splenial with notch for rostral margin of
internal mandibular fenestra; 132.1 Premaxillary
tooth crowns asymmetrical (strongly convex labially,
relatively flattened lingually); 149.1 Cervical epipophyses powerfully developed and prong-shaped;
205.1 Paired caudal and cranial chevron bases;
315.2 Pubic boot shape rounded, angle between
shaft and caudal portion of boot acute.
DELTRAN: 334.1 Anterior trochanter present,
separated from femoral head by cleft; 335.2 Anterior
trochanter of femur alariform; 346.1 Cnemial process arises out of the lateral surface of tibial shaft;
350.1 Crista fibularis present, not well developed;
351.1 Crista fibularis proximally placed; 376.1 Metatarsal III dorsal surface hourglass shaped.
NODE I.
ALL: 331.1 Femoral head transversely elongate.
T.R. HOLTZ, JR.
surface, with kidney-shaped articular surfaces that
are taller laterally than at midline; 287.2 Anterior trochanter of femur proximalmost point above distal
margin of femoral head.
ACCTRAN: None.
DELTRAN: 22.1 Pneumatic excavation without
fenestra in cranial portion of maxillary antorbital
fossa; 48.1 Postorbital ventral process broader
transversely than rostrocaudally with U-shaped
cross-section; 149.1 Cervical epipophyses powerfully developed and prong-shaped; 315.2 Pubic
boot shape rounded, angle between shaft and caudal portion of boot acute; 317.1 Pubic boot present,
less than 30% as long as pubic shaft; 341.1 Extensor
groove in craniodistal region of femur shallow and
not conspicuous; 353.1 Fibula closely appressed to
tibia throughout main shaft; 360.1 Fibula distal end
less than twice craniocaudal width at midshaft, and
consequently astragalar cup for fibula reduced.
DELTRAN: 319.1 Pubis and ischium proximal
shafts narrow.
As in GAUTHIER (1986), NOVAS (1992), HOLTZ
(1994), SERENO (1997), and most other recent studies of theropod phylogeny, a robustly supported
clade of birds and theropods more closely related to
birds than to Ceratosaurus was discovered. This
clade, GAUTHIER's (1986) Tetanurae, comprises primarily the subdivisions Carnosauria and Coelurosauria, discussed below. However, there are several
forms of theropod which (in the present analysis) lie
outside the carnosaur-coelurosaur clade Avetheropoda, yet were found to share a more recent common ancestor with birds than with Ceratosaurus.
NODE J.
ALL: 31.2 Lacrimal prominences comprised of
ridge continuous with raised surface of lateral edge
of nasals; 336.2 Anterior trochanter of femur proximal most point above distal margin of femoral head;
355.1 Proximal region of fibular medial face slightly
concave; 365.1 Pronounced horizontal groove
across cranial face of astragalar condyles.
The relationships among the basal tetanurines,
informally referred to as "megalosaurs" (in, for example, GAUTHIER (1986): p. 10) have been problematic in most recent studies (HOLTZ, 1994; SERENO et
al., 1994, 1996; SERENO, 1997). Unfortunately, the
present analysis does not provide strong support for
any particular scenario of "megalosaur" phylogeny.
This uncertainty seems to stem from a number of
sources, the most important being: a) the fairly large
number of missing data from some of these taxa,
representing our inadequate knowledge of the osteology of these forms at present; b) the lack of specializations in many of these taxa beyond those
shared by all tetanurines, particularly in the case of
the non-spinosaurid "megalosaurs"; and c) alternatively, the highly apomorphic nature of the skulls of
spinosaurids (CHARIG & MILNER, 1997; SERENO et
al., 1998), in which the rostrum, dentition, palate,
and basicranium are uniquely modified among
theropods.
ACCTRAN: -212.0 Distal expansion of scapula
broad (subequal in width to proximal end of scapula); 339.1 Muscle scar in craniodistal region of femur present, non-elliptical in shape.
DELTRAN: 41.1 Orbit shorter than internal antorbital fenestra length; 61.1 Jugal recess; 64.2 Broad
contact between dorsal ramus of quadratojugal and
lateroventral ramus of squamosal; 205.1 Paired
caudal and cranial chevron bases; 215.1 Scapulacoracoid cranial margin with pronounced notch between acromial process and coracoid; 261.1
Articular surface between metacarpals I and II extends well into diaphysis of metacarpal I.
As previously noted, the fragmentary taxon
"Megalosaurus" hesperis falls within this sector of
the cladogram, but no particular sister group relationship was better supported than the others.
NODE K.
ALL: 139.1 Axial "spine table" (expanded distal
end of neural spine); 146.1 Ventral keel on axial centrum absent; 155.1 Cervical zygapophyses displaced laterally away from centrum in dorsal view;
235.1 Humeral shaft sigmoid.
NOVAS (1992) proposed the name "Avipoda" for a
clade comprised of Eustreptospondylus, Piatnitzkysaurus, and more advanced tetanurines. This corresponds to node I in this study. If additional work on
theropod phylogenetics continues to support a subgroup excluding some basal tetanurines but uniting
Eustreptospondylus, Piatnitzkysaurus, and more
derived taxa, this name would serve as a useful label.
ACCTRAN: 4.1 Premaxillary symphyseal region
U-shaped in ventral view; -10.0 Premaxilla and nasal meet subnarially; 60.2 Jugal ventral quadratojugal process extends further caudally than dorsal
quadratojugal process; 260.1 Metacarpal I one half
to one third metacarpal II length; 264.1 Metacarpal
III very much narrower (less than 50%) than metacarpal II; 266.1 Base of metacarpal III set on palmar
surface of hand below base of metacarpal II; 267.1
Proximal articulation of metacarpal III triangular;
126.1 Cranial cervicals broader than deep on cranial
SERENO et al. (1994, 1996; 1998) and SERENO
(1997) presented evidence that several of these basal tetanurines (in particular, Torvosaurus, Eustreptospondylus, and Spinosauridae) represented a
18
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
ALL: 17.1 Maxillary fenestra; -22.0 Pneumatic
excavation without fenestra in cranial portion of
maxillary antorbital fossa absent; 322.1 Obturator
process of ischium separate, trapezoidal; 323.1 Obturator process of ischium proximally placed; 341.2
Extensor groove in craniodistal region of femur deep
and conspicuous
distinct clade of theropod, exclusive of other carnivorous dinosaurs. CHARIG & MILNER (1997) demonstrated that several of the "synapomorphies" for
this postulated "Torvosauroidea" (later changed to
"Spinosauroidea") are absent in Spinosauridae. In
the present study, some of the other characters suggested to support a monophyletic "Spinosauroidea"
(13.1, 14.1, 22.1, 35.1, and 279.1) are explained in
the most parsimonious distribution of states in this
analysis to be basal tetanurine features lost in some
or all avetheropod taxa.
ACCTRAN: 181.2 Presacral pleurocoels camellate; 217.1 Ventral coracoid process well developed;
355.2 Proximal region of fibular medial face well excavated
Of particular note are conditions related to the
rostral ramus ("anterior ramus" in BAKKER et al.
(1992) and SERENO et al. (1994, 1996)) of the maxilla. The maxillae of basal tetanurines differ from
those of ceratosaurs in the shape of the rostrodorsal
margin (Fig. 6). In ceratosaurs and herrerasaurids
this line is a simple curve, convex dorsally, from the
dorsal ramus to the tooth line. In tetanurines primitively this surface is a more complex curve, with a depression ventral to the external naris. This produces
a rostral ramus to the maxilla, rostral to the dorsal ramus (13.1). The presence of this structure is found in
all non-avetheropods for which skull material is
known (although the condition in spinosaurids is distinct from the other taxa, given their elongate snouts:
CHARIG & MILNER (1997)), and is also present in the
skulls of most carnosaurs (the notable absences being Sinraptor and Yangchuanosaurus) and in basal
coelurosaurs for which the skull is known (Proceratosaurus and Ornitholestes). As with sinraptorid carnosaurs, the clade of coelurosaurs comprised of
Scipionyx and Maniraptoriformes have the primitive
state with a simple dorsally convex curvature of the
rostral portion of the maxilla. This distribution is most
parsimoniously explained as independent reversals
in Sinraptoridae and advanced Coelurosauria.
DELTRAN: 58.1 Jugal participates in internal antorbital fenestra; 197.2 Transition point in proximal
half of tail; 253.1 Distal carpal I block does not overlap metacarpal II dorsally, but does so ventrally;
254.1 Distal carpal I fused to distal carpal II; 255.1
Semilunate carpal block fully developed with transverse trochlea; 260.1 Metacarpal I one half to one
third metacarpal II length; -279.0 Pollex ungual less
than three times longer than height of articular facet;
-283.0 Manual ungual moderate length; 339.2 Muscle scar in craniodistal region of femur present, elliptical in shape; 364.1 Astragalar distal condyles
oriented cranioventrally
NODE M. AVETHEROPODA
ALL: -35.0 Lacrimal dorsal (rostral) ramus dorsoventrally deep; -48.0 Postorbital ventral process
broader rostrocaudally than transversely; 58.1
Squamosal constriction of lateral temporal fenestra;
169.1 Scars for interspinous ligaments terminate
below apex of neural spine; 208.1 Middle chevron
with dramatic bend in distal portion ("L-shaped");
274.1 First phalanx of pollex greater than length of
metacarpal II; 288.1 Iliac preacetabular fossa for M.
cuppedicus
ACCTRAN: 330.2 Femoral head greater than 90
degrees from shaft (head directed dorsally)
SERENO et al. (1994, 1996) specifically recognized a derived condition in which the rostral ramus
was longer rostrocaudally than tall dorsoventrally. In
the present study two different derived states were
recognized to describe the relative proportions of
the rostral ramus: those forms for which the structure is present, but shorter rostrocaudally than tall
dorsoventrally (14.1) and the condition recognized
by SERENO et al. (1994, 1996) (14.2). Rather than
uniting Spinosauridae, Torvosaurus, Eustreptospondylus, and Afrovenator outside of other theropods, however, this condition was found to be the
state at the base of Tetanurae, and subsequently
shortened in various tetanurine taxa.
DELTRAN: 38.1 Prefrontal-frontal peg-in-socket
suture; 76.1 Palatine tetraradiate; 78.1 Palatine recesses; 120.1 Splenial with notch for rostral margin
of internal mandibular fenestra; 123.1 Retroarticular
process of articular faces caudally; 132.1 Premaxillary tooth crowns asymmetrical (strongly convex labially, relatively flattened lingually); 212.1 Distal
expansion of scapula reduced or absent; 217.1 Ventral coracoid process well developed; 225.1 Furcula;
233.1 Manus/(humerus + radius) length ratio
greater than 66%; -249.0 Ulnar facet for radius small
and flat; 264.1 Metacarpal III very much narrower
(less than 50%) than metacarpal II; 266.1 Base of
metacarpal III set on palmar surface of hand below
base of metacarpal II; 267.1 Proximal articulation of
metacarpal III triangular; 268.1 Metacarpal IV less
than half length of metacarpal II; 278.1 Pollex larger
Curiously, all known basal tetanurines represent
fairly large sized taxa (approximately 6 m or longer).
NODE L.
19
T.R. HOLTZ, JR.
Fig. 6 - Cladogram comparing the left maxillae (in left lateral view) of the ceratosaur Dilophosaurus (modified from
WELLES, 1984), the basal tetanurine Afrovenator (modified from SERENO et al., 1994), and the avetheropod carnosaur
Monolophosaurus (modified from ZHAO & CURRIE, 1993). Arrow indicates the presence of the rostral ramus of the maxilla,
a tetanurine synapomorphy (character 13.1). Not to scale.
than other manual unguals; 355.2 Proximal region of
fibular medial face well excavated
derived feature may be synapomorphic for a more
inclusive group (see Discussion below).
As in NOVAS (1992), HOLTZ (1994), SERENO et al.
(1994, 1996), and SERENO (1997), a clade was recognized containing Allosaurus and those taxa closest to it as one branch, and Aves and those taxa
closest to it as the other, outside of the more primitive
tetanurines. Following HOLTZ (1994), and using a
name for this clade first published in PAUL (1988),
this clade comprised of all descendants of the most
recent common ancestor of Allosaurus and Neornithes is termed Avetheropoda. SERENO et al. (1994,
1996) and SERENO (1997, 1998) suggested the alternative name "Neotetanurae": indeed, SERENO
(1997, 1998) uses the same definition as the above,
rendering this term a junior objective synonym of
Avetheropoda (see also PADIAN, HUTCHINSON &
HOLTZ (1999)). Perhaps the name "Neotetanurae"
might be preserved as the name for a more inclusive
taxon (for example, Neornithes and all taxa sharing
a more recent common ancestor with Neornithes
than with Spinosauridae).
The two divisions of Avetheropoda are termed
Carnosauria (Allosaurus and all taxa sharing a more
recent common ancestor with Allosaurus than with
Neornithes) and Coelurosauria (Neornithes and all
taxa sharing a more recent common ancestor with
Neornithes than with Allosaurus): see P ADIAN ,
HUTCHINSON & HOLTZ (1999) for discussion of the
history of these terms. Both carnosaurs and coelurosaurs share the derived character state of fused
clavicles (furculae): however, the lack of preservation of clavicles in more basal tetanurines allows for
the possibility of this character state being synapomorphic for an even more inclusive group (see Discussion below). Twenty four other derived character
states unite carnosaurs and coelurosaurs to the exclusion of other theropod clades, but sixteen of
these cannot be fairly assessed at present in basal
tetanurines, as the skeletal elements concerned are
not recovered for these taxa at present.
NODE N. CARNOSAURIA
The Early Cretaceous (?Barremian) Afrovenator
abakensis was found to share several derived characters with Avetheropoda lacking in other basal tetanurines: in particular, unquestionable presence of a
maxillary fenestra (17.1) and an obturator process
on the ischium (322.1, 323.1). Additionally, the semilunate carpal block (255.1) is present in Afrovenator
as it is in avetheropods, but as the carpus is unknown in other non-avetheropod tetanurines, this
ALL: 23.1 Lateral surface of nasal participates in
antorbital cavity, forming a nasal antorbital fossa;
29.1 Nasal recesses; 75.1 Palatines meet medially;
83.1 Nuchal crest pronounced; 97.1 Distance
across basal tubera less than the transverse width of
occipital condyle; 100.1 Basipterygoid processes
short, but not fused to pterygoids; 166.1 Ventral pro-
20
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
cesses (hypapophyses) on cervicodorsal vertebrae
present as small protrusions.
56.1 Squamosal flange covering quadrate head in
lateral view; 167.1 Neural spines of dorsals equal to
twice centrum height; 357.1 Cranial protuberance
on fibula below expansion.
ACCTRAN: 67.2 Quadrate foramen small and
enclosed within dorsal ramus of quadrate; 113.1
Rostral surangular foramen large, in rostrallyoriented depression; 229.1 Humerus/scapula
length ratio less than 65%; 314.1 Pubic foramen perforating pubic apron in distal half of shaft; 350.2
Crista fibularis well developed; 382.1 Pedal digit I
phalanges 1+2 subequal to pedal digit III phalanx 1.
ACCTRAN: -57.0 Squamosal does not constrict
lateral temporal fenestra.
DELTRAN: -10.0 Premaxilla and nasal meet
subnarially; -14.0 Rostral ramus absent; 113.1 Rostral surangular foramen large, in rostrally-oriented
depression; 175.1 Dorsal centrum "hourglass"
shaped, central section thickness less than 60%
height of cranial face; -330.1 Femoral head approximately 90 degrees from shaft (head directed horizontally).
DELTRAN: 24.1 Nasal participates in antorbital
cavity; 60.2 Jugal ventral quadratojugal process extends further caudally than dorsal quadratojugal
process; 70.1 Quadrate articulation projects well
caudal to the caudalmost point of the occipital condyle.
NODE Q. ALLOSAURIDAE
ALL: -58.0 Jugal does not participate in internal
antorbital fenestra; 103.1 Occipital condyle constricted neck; 115.1 Rostral ramus of surangular
deep; 178.1 Dorsal column subequal to femur
length; 257.3 Metacarpal IV absent; 317.2 Pubic
boot present, greater than 30% as long as pubic
shaft; 318.1 Pubic-ischial contact only narrow region; 349.1 Lateroproximal condyle of tibia with conspicuous waisting between body of condyle and
main body of tibia in proximal view.
NODE O. ALLOSAUROIDEA
ALL: 21.1 Pneumatic excavation of the ascending ramus of the maxilla; 26.2 Narial prominences
comprised of paired ridges along lateral edges of
nasals; 33.1 Lacrimal recess, single opening present; 47.1 Postorbital prominences present; 69.1
Quadrate articulation projects well ventral of the
ventral surface of the maxilla; 116.2 Horizontal shelf
on lateral surface of surangular, rostral and ventral
to the mandibular condyle, prominent and pendant;
179.1 Caudal dorsal neural spines oriented cranially; 315.4 Pubic boot shape triangular (apex caudal) in ventral view and angle between shaft and
caudal portion of boot acute.
ACCTRAN: 16.2 Promaxillary fenestra present,
obscured in lateral view by ascending ramus of maxilla; 87.2 Paroccipital process curving ventrally and
pendant; 112.1 Dentary caudal depth 150-200%
depth of dentary symphysis; -170.0 Apices of dorsal
neural spines unexpanded; 214.1 Caudal margin of
acromial process of scapula forms abrupt change,
perpendicular to blade.
ACCTRAN: 118.1 External mandibular fenestra
reduced; 175.1 Dorsal centrum "hourglass" shaped,
central section depth less than 60% height of cranial
face; 307.1 Obturator foramen of pubis open ventrally to form obturator notch.
DELTRAN: -113.0 Rostral surangular foramen
absent or very small pit; 118.1 External mandibular
fenestra reduced; 229.1 Humerus/scapula length
ratio less than 65%; 307.1 Obturator foramen of pubis open ventrally to form obturator notch; 330.2
Femoral head greater than 90 degrees from shaft
(head directed dorsally).
DELTRAN: 4.1 Premaxillary symphyseal region
U-shaped in ventral view; 67.2 Quadrate foramen
small and enclosed within dorsal ramus of quadrate;
77.1 Jugal process of palatine expanded; 81.1 Ventral ectopterygoid recess present and commashaped; 181.2 Presacral pleurocoels camellate;
257.2 Metacarpal IV present, without phalanges;
314.1 Pubic foramen perforating pubic apron in distal half of shaft; 316.1 Caudal portion of pubic boot
longer than cranial portion, but latter present; 350.2
Crista fibularis well developed; 382.1 Pedal digit I
phalanges 1+2 subequal to pedal digit III phalanx 1.
NODE R.
ALL: 3.2 Five premaxillary teeth; 199.1 Distal
caudal vertebrae with moderate interlocking, prezygapophyses extend more than one half, but less
than one, centrum length; -212.0 Distal expansion of
scapula broad, subequal in width to proximal end.
ACCTRAN: 31.2 Lacrimal prominences triangular hornlets; -78.0 Palatine recesses absent; 96.1
Basioccipital excluded from basal tuber; 195.1
Proximal caudal zygapophyses elongate.
NODE P. SINRAPTORIDAE
ALL: 9.1 External nares with marked inset of the
caudal margin; -13.0 Rostral ramus of maxilla absent, rostrodorsal surface of maxilla forms convex
surface from dorsal ramus to ventral margin; 20.1
Promaxillary fenestra larger than maxillary fenestra;
DELTRAN: 10.1 Premaxilla and nasal do not
meet subnarially; 87.2 Paroccipital process curving
21
T.R. HOLTZ, JR.
the long established name Carnosauria HUENE,
1920 may be conserved as the clade comprised of
Allosaurus and all taxa sharing a more recent common ancestor with this taxon than with Neornithes.
In the present study, the oldest and most basal form
in this clade is the Middle Jurassic Monolophosaurus jiangi ZHAO & CURRIE, 1993 of China. However,
SERENO et al. (1994, 1996) and SERENO (1997)
have proposed that Cryolophosaurus ellioti HAMMER & HICKERSON, 1994 is a member of this clade. If
this hypothesis is correct, the age of this form would
indicate not only a minimum Early Jurassic date for
the origin of Carnosauria but also of the Coelurosauria (see also discussion of maniraptoriforms below),
as well as such an early date for the origin of each of
the various basal tetanurine lineages.
ventrally and pendant; -169.0 Scars for interspinous
ligaments terminate at apex of neural spine; 214.1
Caudal margin of acromial process of scapula forms
abrupt change, perpendicular to blade.
NODE S
ALL: 36.1 Lacrimal suborbital bar; 41.1 Frontalfrontal suture fused; 50.1 Postorbital-lacrimal contact broad; 53.1 Postorbital suborbital flange; -100.0
Basipterygoid processes moderately long; 159.1
Elevation of cranial face of midcervical centra; 191.1
Caudal pleurocoels present in centra.
ACCTRAN: -166.0 Ventral processes (hypapophyses) on cervicodorsal vertebrae absent; 183.1
Sacral pleurocoels; 236.1 Humeral head offset and
emarginated ventrally by groove; 241.1 Deltapectoral crest on humerus expanded and offset from humeral shaft.
HARRIS (1998) observed that palatines that meet
medially are present in both Sinraptor and Allosaurus. Such a geometry appears to be present in
Monolophosaurus (ZHAO & CURRIE, 1993), but is
absent in basal sauropodomorphs, Herrerasaurus,
ceratosaurs, tyrannosaurids, ornithomimids, dromaeosaurids, and other coelurosaurs. In these taxa,
the palatines remain separated medially by the rostral processes of the pterygoids.
DELTRAN: -14.1 Rostral ramus of maxilla
shorter rostrocaudally than dorsoventrally.
NODE T.
ALL: -10.0 Premaxilla and nasal meet subnarially; 15.2 Maxillary antorbital fossa greatly reduced
in size, not extending much beyond rim of the external antorbital fossa; -21.0 Pneumatic excavation of
ascending ramus of maxilla absent; 26.3 Narial
prominences knobby rugosities across dorsal and
lateral surface of nasals, extending onto dorsalmost
surface of maxillae; 52.1 Postorbital bulbous rostrally projecting rugosity; 88.1 Basisphenoid, but not
parasphenoid rostrum, strongly expanded and
pneumatized; 131.1 Lateral surface of teeth with
wrinkles in enamel internal to serrations.
Allosauroidea CURRIE & ZHAO, 1993a has been
proposed as the name for the clade comprised of all
descendants of the most recent common ancestor
of Allosaurus and Sinraptor (see also P ADIAN ,
H UTCHINSON & H OLTZ , 1999). (S ERENO (1997,
1998) uses the same name but with the definition
employed here for Carnosauria). In the present
study, this clade would contain all carnosaurs except
for Monolophosaurus. HARRIS (1998) observed that
allosauroids were characterized by caudal dorsal
vertebrae in which the neural spines were oriented
cranially rather than vertically (as in other theropods). This orientation is not present in Monolophosaurus, but is found in Sinraptor, Allosaurus, and
Acrocanthosaurus. As such, it is supported here as
a synapomorphy of Allosauroidea.
ACCTRAN: 5.1 Premaxilla subnarially very
deep, main body taller dorsoventrally than long rostrocaudally; 8.1 Maxillary process of premaxilla reduced, maxilla participates broadly in ventral
surface of external naris; -17.0 Maxillary fenestra
absent; 37.1 Prefrontals reduced or absent; 42.2
Frontal-parietal suture on dorsal surface of skull
fused, suture indistinguishable; -57.0 Squamosal
does not constrict lateral temporal fenestra; -60.0
Jugal dorsal and ventral quadratojugal processes
subequal in caudalmost extension; 62.1 Jugal recesses; 68.1 Quadrate dorsal ramus height greater
than height of orbit; 105.1 Dentary end squared with
expanded tip; 157.1 Cranial cervicals broader than
deep on cranial surface, with reniform (kidneyshaped) articular surfaces that are taller laterally
than at midline.
Within the allosauroids, a sister group relationship for Sinraptor and Yangchuanosaurus, proposed in the first study of the former taxon (CURRIE &
ZHAO, 1993a), is supported here. Similarly, HUTT,
MARTILL & BARKER (1996) suggested that their
newly described Neovenator (of the Wealden Group
of the Isle of Wight) was closely related to the Late
Jurassic North American genus Allosaurus, a position also retained in the present analysis.
The union of Acrocanthosaurus, Giganotosaurus, and Carcharodontosaurus, first proposed (with
a different topology) by SERENO et al. (1996) was
found in the present analysis. This clade of gigantic
mid-Cretaceous (Aptian-Cenomanian) carnosaurs
was called "Carcharodontosauridae" by SERENO et
DELTRAN: -87.0 Paroccipital process oriented
more caudally than dorsally.
As argued by PADIAN, HUTCHINSON & HOLTZ
(1999) (see also HOLTZ & BRETT-SURMAN, 1997),
22
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
gular deep; 121.1 Coronoid extremely reduced or
absent; 153.1 Cervical prezygapophyses flexed;
157.1 Cranial cervicals broader than deep on cranial
surface, with kidney-shaped articular surfaces that
are taller laterally than at midline; -160.0 Midcervical
centra length about twice diameter of cranial face;
190.1 Number of caudals between 30 and 44; 196.1
Caudal transverse processes only on caudals I-XV
or fewer; 209.1 Distal chevrons with cranial and caudal projections, and more than twice as long craniocaudally as tall dorsoventrally ("boat-shaped"); 215.0 Scapulacoracoid cranial margin smooth;
226.1 Forelimb (humerus+radius+manus)/hindlimb
(femur+tibia+pes) length ratio greater than 50% but
less than 120%; 227.1 Forelimb/presacral vertebral
series length ratio greater than 75% but much less
than 200%; 232.2 Radius/humerus length ratio
greater than 76%; 245.1 Ulnar shaft bowed caudally; 272.1 Length of phalanx 3 of manual digit
III/(sum of lengths of phalanges 1+2 of digit III)
greater than 100%; 330.2 Femoral head greater
than 90 degrees from shaft (head directed dorsally);
332.1 Greater trochanter cleft from femoral head;
347.1 Incisura tibialis cranialis occupies more than
66% of medial surface of proximal tibia; 362.2 Astragalar ascending process craniocaudally reduced
and proximodistally tall, with dorsal margin sigmoid
("ornithomimoid/albertosauroid condition"); -365.0
No pronounced horizontal groove across cranial
face of astragalar condyles; 380.1 Metatarsal I plantar to medial side of metatarsal II; 386.1 Pedal ungual II significantly longer than pedal ungual III.
al. (1996), although (under the phylogenetic taxonomy used here) this clade is part of Allosauridae
(i.e., Allosaurus and all taxa sharing a more recent
common ancestor with Allosaurus than with Sinraptor). Although the analysis of CURRIE & CARPENTER
(in press) found that Acrocanthosaurus was more
closely related to Allosaurus than to Giganotosaurus or Carcharodontosaurus, this study agrees with
SERENO et al. (1996) and HARRIS (1998) in grouping
the three giant mid-Cretaceous taxa exclusive of the
Late Jurassic Allosaurus.
Examining the character states at node T reveals
many derived features also demonstrated by abelisauroids. Indeed NOVAS (1997c) has suggested that
Giganotosaurus and Carcharodontosaurus are
closely related to the abelisauroid neoceratosaurs,
a very different phylogenetic position from that
found here. Although there are several cranial synapomorphies potentially uniting Carcharodontosaurus and Abelisauridae (in general) or Abelisaurus (in
particular), additional features unite the African dinosaur with Giganotosaurus, while further cranial
and postcranial character states group this South
American dinosaur with the unquestioned tetanurine Acrocanthosaurus (see also the Discussion
below).
In the current phylogeny, Carcharodontosaurus
(of Cenomanian age) is the latest known carnosaur.
No fossil evidence presently known indicates the
survival of Carnosauria into the last twenty eight million years of the Cretaceous.
DELTRAN: None.
NODE U. COELUROSAURIA
ALL: -147.0 Cervical centra surfaces amphiplatyan; 315.3 Pubic boot boat-shaped (pointed cranially and caudally) in ventral view and angle between
shaft and caudal portion of boot acute; 354.1 Fibula
proximal end 75% or more proximal width of tibia.
NODE V.
ALL: None
ACCTRAN: 307.1 Obturator foramen of pubis
open ventrally to form obturator notch; 322.2 Obturator process of ischium separate, triangular
shaped; -341.1 Extensor groove in craniodistal region of femur present, but shallow and not conspicuous; 349.1 Lateroproximal condyle of tibia with
conspicuous waisting between body of condyle and
main body of tibia in proximal view.
ACCTRAN: 15.1 Maxillary antorbital fossa
greater than 40% of the rostrocaudal length of the
antorbital cavity; -31.0 Lacrimal prominences absent; -45.0 Orbit shape round; -54.0 Postorbital frontal process sharply upturned; 60.1 Jugal dorsal
quadratojugal process extends further caudally than
ventral quadratojugal process; -61.0 Jugal recesses
absent; 70.2 Quadrate articulation rostral to caudalmost point of occipital condyle; 80.1 Subsidiary fenestra between pterygoid and palatine; 82.1
Endocranial cavity enlarged relative to other dinosaurs, but temporal musculature extends onto frontals; 86.1 Paroccipital process with hollow proximal
portion; 88.1 Basisphenoid, but not parasphenoid
rostrum, strongly expanded and pneumatized; 90.1
Three tympanic recesses; 91.1 Branches of internal
carotid artery enter hypoglosseal fossa through single common foramen; 115.1 Rostral ramus of suran-
DELTRAN: 4.1 Premaxillary symphyseal Ushaped in ventral view; -10.0 Premaxilla and nasal
meet subnarially; 15.1 Maxillary antorbital fossa
greater than 40% of the rostrocaudal length of the
antorbital cavity; -45.0 Orbit shape round; -61.0 Jugal recesses absent; 70.2 Quadrate articulation rostral to caudalmost point of occipital condyle.
NODE W.
ALL: None.
ACCTRAN: 7.1 Medial alae from premaxillae
meet in front of vomera; -44.0 Orbit longer than inter-
23
T.R. HOLTZ, JR.
nal antorbital fenestra length; 116.1 Horizontal shelf
on lateral surface of surangular, rostral and ventral
to the mandibular condyle, prominent and extends
laterally.
26.0 Narial prominences absent; 71.1 Quadrate
pneumaticity well developed; 72.1 Secondary palate well ossified from premaxilla through one-half
the length of the ventral surface of the maxilla; 103.1
Occipital condyle constricted neck; 114.1 Caudal
surangular foramen large opening; 369.1 Metatarsus proportions elongate relative to most other
theropods of same femur length; -376.0 Metatarsal
III dorsal surface shape elliptical.
DELTRAN: 307.1 Obturator foramen of pubis
open ventrally to form obturator notch; 318.1 Pubicischial contact only narrow region; 337.1 Fourth trochanter of femur little developed; -341.1 Extensor
groove in craniodistal region of femur present, but
shallow and not conspicuous; 349.1 Lateroproximal
condyle of tibia with conspicuous waisting between
body of condyle and main body of tibia in proximal
view; 362.2 Astragalar ascending process craniocaudally reduced and proximodistally tall, with dorsal margin sigmoid ("ornithomimoid/albertosauroid
condition"); -365.0 No pronounced horizontal
groove across cranial face of astragalar condyles.
DELTRAN: 263.1 Metacarpal III clearly shorter
than metacarpal II; 272.1 Length of phalanx 3 of
manual digit III/(sum of lengths of phalanges 1 + 2 of
digit III) greater than 100%; 277.1 Manual ungual
palmar and dorsal regions subequal in width; 282.1
Manual ungual cross section blade-like, more than
three times as deep as wide.
NODE Z.
NODE X.
ALL: -235.0 Humeral shaft straight; 257.3 Metacarpal IV absent.
ALL: 333.1 Femoral head transversely elongate.
ACCTRAN: 60.1 Jugal dorsal quadratojugal process extends further caudally than ventral quadratojugal process; 81.2 Ventral ectopterygoid recess
present and subcircular; 277.1 Manual ungual region palmar to ungual groove subequal in width to
region dorsal to ungual groove; 282.1 Manual ungual cross section blade-like, more than three times
as deep as wide; 357.1 Cranial protuberance on fibula below expansion.
ACCTRAN: 157.1 Cranial cervicals broader than
deep on cranial surface, with reniform (kidneyshaped) articular surfaces that are taller laterally
than at midline.
DELTRAN: -13.0, -14.0 Rostral ramus of maxilla
absent, rostrodorsal surface of maxilla forms convex surface from dorsal ramus to ventral margin; 26.0 Narial prominences absent; 153.1 Cervical
prezygapophyses flexed; -215.0 Scapulacoracoid
cranial margin smooth; -221.0 Sternal plates unfused; -222.0 Sternum carina absent; -245.0 Ulnar
shaft bowed caudally.
DELTRAN: -31.0 Lacrimal prominences absent;
-44.0 Orbit longer than internal antorbital fenestra
length; 82.1 Endocranial cavity enlarged relative to
other dinosaurs, but temporal musculature extends
onto frontals; 115.1 Rostral ramus of surangular
deep; -160.0 Midcervical centra length about twice
diameter of cranial face; 181.2 Presacral pleurocoels camellate; 190.1 Number of caudals between
30 and 44; 196.1 Caudal transverse processes only
on caudals I-XV or fewer; 209.1 Distal chevrons with
cranial and caudal projections, and more than twice
as long craniocaudally as tall dorsoventrally ("boatshaped"); 227.1 Forelimb/presacral vertebral series
length ratio greater than 75% but much less than
200%; 322.2 Obturator process of ischium separate, triangular shaped; 332.1 Greater trochanter of
femur cleft from femoral head.
NODE aa.
ALL: 195.1 Proximal caudal zygapophyses elongate.
ACCTRAN: 8.1 Maxillary process of premaxilla
reduced, maxilla participates broadly in ventral surface of external naris; 37.1 Prefrontals reduced or
absent; 118.1 External mandibular fenestra reduced; 158.1 Cranial cervical centra extend beyond
caudal extend of neural arch; 168.1 Apices of dorsal
neural spines expanded transversely to form "spine
table"; 188.1 Sacral neural spines fuse to form lamina; -240.0 Humeral ends little or not expanded;
284.2 Manual unguals straight.
NODE Y.
DELTRAN: 114.1 Caudal surangular foramen a
large opening.
ALL: -149.0 Cervical epipophyses rugosities on
caudal zygapophyses; 154.1 Cervical neural spines
low and craniocaudally short; 186.2 Sacrals III-V
dorsoventrally flattened.
Coelurosauria is a well-supported clade of tetanurine dinosaurs. Most of the traditional coelurosaurs of previous studies (GAUTHIER, 1986; HOLTZ,
1994; SERENO, 1997; SUES, 1997), known from relatively complete material, comprise a clade of derived
forms, the Maniraptoriformes HOLTZ, 1996b: these
ACCTRAN: -13.0, -14.0 Rostral ramus of maxilla
absent, rostrodorsal surface of maxilla forms convex surface from dorsal ramus to ventral margin; -
24
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
neck. Unfortunately, the cervicals of Proceratosaurus are presently unknown.
taxa are discussed below. As in the case of basal
Tetanurae, however, there exist a number of fragmentary forms which demonstrate some shared derived characters with Maniraptoriformes compared
to other theropods, but which were found to lie outside that clade..
Among these, Proceratosaurus bradleyi is the
oldest currently known (Bathonian age, Middle Jurassic). An alleged therizinosauroid maniraptoriform from the Sinemurian age (Early Jurassic) of the
Lower Lufeng Formation, Yunnan, China (ZHAO &
XU, 1998) is even older. If substantiated, this would
indicate that Maniraptoriformes (and indeed, the
various lineages of the oviraptorosaur-Microvenator
clade, Paraves, Compsognathidae, and Arctometatarsalia, the ancestors of the basal coelurosaurs,
and Carnosauria) would date back to at least the Sinemurian. However, the specimen in question is an
isolated dentary, and given the resemblance of the
dentary of therizinosauroids and basal sauropodomorphs (the latter common to the Early Jurassic dinosaurian fauna), this intriguing discovery is
greeted with some caution.
Regardless of the phylogenetic identity of the
Yunnan specimen, Proceratosaurus shares
uniquely with other coelurosaurs several derived
characteristics. The maxillary antorbital fossa (the
wall of bone on the rostral portion of the antorbital
fenestra: WITMER, 1997) forms 40% of the total rostrocaudal length of the antorbital fenestra, a derived
condition first hypothesized as a coelurosaurian
synapomorphy by SERENO et al. (1994, 1996) and
supported here (Fig. 7). In carnosaurs, basal tetanurines, and ceratosaurs, the maxillary antorbital
fossa represents a much smaller fraction of this
length (and, correspondingly, the internal antorbital
fenestra represents a larger fraction of the total
structure). Furthermore, the articulation between
the quadrates and the mandible lie rostral to the caudalmost point of the occipital condyle in Proceratosaurus and most other coelurosaurs. In contrast, the
quadrate articulation lies at the same point or rostral
to the caudalmost point of the occipital condyle in
ceratosaurs, basal tetanurines, and carnosaurs.
Even in the largest skulled of coelurosaurs, large tyrannosaurids such as Tyrannosaurus rex (MOLNAR,
1991) the articulation of the quadrates is slightly rostral to the end of the occipital condyle. Coupled with
the flexion of the cervical prezygapophyses and the
development of kidney-shaped articular surfaces in
the cranial cervicals, the forward placement of the
mandibular joint may indicate specializations towards greater lateral mobility in the necks of coelurosaurs relative to other theropods. In other
theropods, the posterior placement of the mandibular joint may have interfered with lateral motion of the
25
The inclusion of Gasosaurus constructus in Coelurosauria is novel to this study. This poorly known
Middle Jurassic Chinese form is admittedly fragmentary. Its pelvis retains several primitive features
transformed in other, later coelurosaurs: the obturator foramen is present, the ischium is footed, and the
obturator process appears to be trapezoidal. Nevertheless, the hindlimb possesses several coelurosaur characteristics: a femoral head at an angle
greater than 90 degrees to the femoral shaft; a
lesser trochanter cleft from the femoral head, and a
fibula whose proximal portion is greater than 75%
the proximal width of the tibia. Given the incomplete
nature of this taxon, additional character evidence
may reveal it does not belong to Coelurosauria.
Pending such discovery, however, the current analysis suggests that this form does indeed share a more
recent common ancestor with birds than with Allosaurus. (P. Currie, pers. comm. 1998, indicates that
as yet undescribed specimens suggest that Gasosaurus is in fact a primitive carnosaur, perhaps a sinraptorid).
The Late Cretaceous (Campanian) North American taxon Dryptosaurus, previously suggested to be
a coelurosaur by DENTON (1990), was found to be
more derived than Gasosaurus in that it possesses a
true obturator notch, a narrow proximal contact between the pubis and ischium, and a few other
hindlimb characters. Unfortunately, this taxon too is
incompletely known at present.
The Late Jurassic Ornitholestes hermanni OSBORN, 1903 is one of the most completely known basal coelurosaurs. As in HOLTZ (1994), SERENO et al.
(1996), and SERENO (1998), but unlike GAUTHIER
(1986) and MAKOVICKY & SUES (1998), this form
was found to be more distantly related to birds than
is Ornithomimosauria.
Sympatric with Ornitholestes, Coelurus agilis
MARSH, 1879b was found to be somewhat more
closely related to birds than the former taxon, as in
MAKOVICKY & SUES (1998); however, unlike that
study, Coelurus was here found to lie outside the
clade comprised of all descendants of the most recent common ancestor of birds and ornithomimosaurs. Note that as in MAKOVICKY & SUES (1998),
this study follows MILES, CARPENTER & CLOWARD
(1998) in assigning the manual material AMNH 587
to Coelurus rather than Ornitholestes. New, more
complete material of Coelurus currently under study
by those latter authors will greatly aid our understanding of this coelurosaur and its phylogenetic position. Yet another small coelurosaur from the Late
Jurassic Morrison Formation, known only from iso-
T.R. HOLTZ, JR.
Fig. 7 - Cladogram comparing skulls (in right lateral view)of the carnosaur Allosaurus (modified from CURRIE, 1997),
the basal coelurosaur Proceratosaurus (BMNH R4860, new reconstruction based on a photograph by the author), and
the tyrannosaurid Daspletosaurus (modifed from RUSSELL, 1970). Bar above skulls represent total length of antorbital
fenestra; solid portion of bar and number above represents percentage of antorbital fenestra occupied by maxillary antorbital fossa (character 15). Not to scale.
monious position is as the sister taxon of Maniraptoriformes. Like many maniraptoriform groups, this
taxon demonstrates a greatly enlargement of the
caudal surangular foramen (previously thought to
also occur in the carnosaur Acrocanthosaurus (see
STOVALL & LANGSTON, 1950; HOLTZ, 1994; new evidence demonstrates this condition is lacking in that
carnosaur: CURRIE & CARPENTER, in press). Bagaraatan is noteworthy for sharing many features of
the tibia and tarsus with ceratosaurian dinosaurs
and with advanced ornithothoracine birds: these are
best explained here by convergence.
lated vertebrae, was reported by M AKOVICKY
(1997).
Scipionyx samniticus DAL SASSO & SIGNORE,
1998 is a recently described theropod taxon from the
Albian of Italy. It is known from one of the most complete skeletons of any basal coelurosaur. The describers considered it to be a maniraptoriform of
uncertain relationship. In the present analysis it was
found to lie outside of Maniraptoriformes proper.
However, as this specimen is likely a hatchling, it
may be that an adult individual of Scipionyx may
demonstrate maniraptoriform synapomorphies. At
present, this taxon is potentially of great importance
in determining the character distribution in basal
members of the maniraptoriform clade.
NODE bb. MANIRAPTORIFORMES
ALL: 313.1 Pubic apron limited to distal half of pubic shaft; 336.3 Anterior trochanter of femur proximalmost point above proximal margin of femoral
head.
It should be noted that, given the often marked
difference between hatchling and adult dinosaur
body size (CARPENTER, HIRSCH & HORNER, 1994),
the small size of the type and only known specimen
of Scipionyx (237 mm from tip of the premaxilla to the
ninth caudal vertebra: DAL SASSO & SIGNORE, 1998)
may not necessarily reflect a small adult size for this
taxon.
ACCTRAN: None.
DELTRAN: 8.1 Maxillary process of premaxilla
reduced, maxilla participates broadly in ventral surface of external naris; -24.0 Nasal excluded from antorbital cavity; 71.1 Quadrate pneumaticity well
developed; 80.1 Subsidiary fenestra between pterygoid and palatine; 76.1 Paroccipital process with
hollow proximal portion; 90.1 Three tympanic recesses; 118.1 External mandibular fenestra re-
Bagaraatan ostromi OSMÓLSKA, 1996 is known
only from a fragmentary skeleton from the Nemegt
Formation (Late Cretaceous, ?early Maastrichtian
age) of Mongolia. In the present study its most parsi-
26
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
duced; 284.2 Manual unguals extremely curved;
347.1 Incisura tibialis occupies more than 66% of
medial surface of proximal tibia; ; -376.0 Metatarsal
III dorsal surface shape elliptical; 380.1 Metatarsal I
plantar to medial side of metatarsal II.
Following HOLTZ (1996b), the clade comprised of
all descendants of the most recent common ancestor of Ornithomimus and Neornithes is named Maniraptoriformes. As well as ornithomimids and birds,
HOLTZ (1994), SERENO (1997, 1998), and this study
agree that Maniraptoriformes also includes tyrannosaurids, troodontids, therizinosauroids, oviraptorosaurs, and dromaeosaurids (although the
arrangement of these taxa differ among these studies). In MAKOVICKY & SUES (1998), all of the above
except for tyrannosaurids were included within this
clade (there unnamed), as were Ornitholestes and
Coelurus. HOLTZ (1996b) also clarified the phylogenetic definitions of the two main branches of maniraptoriform coelurosaurs: Maniraptora GAUTHIER
1986, Neornithes and all taxa sharing a more recent
common ancestor with neornithines than with Ornithomimus; and its sister taxon by definition, Arctometatarsalia HOLTZ, 1994, Ornithomimus and all
taxa sharing a more recent common ancestor with
Ornithomimus than with Neornithes. Of note, nothing in these phylogenetic taxonomic definitions necessitates the presence of a particular derived
character or suite of characters (e.g., grasping
hands and pinched metatarsi, respectively) nor a
particular combination of taxa. Instead, they simply
describe two complimentary clades within the larger
taxon Maniraptoriformes. As discussed below,
some taxa with "maniraptor" hands may lie within
Arctometatarsalia in the present analysis, while
other taxa with arctometatarsi are found to be members of Maniraptora. See HOLTZ (1996b) and PADIAN, HUTCHINSON & HOLTZ (1999) for discussion.
A character complex of interesting distribution is
the presence of a large ossified secondary palate
formed by medial extensions of the maxillae (72.1)
and by medial alae of the premaxillae contacting
rostral to the vomers (7.1). Because the palate of
several coelurosaurian taxa is unknown or unprepared, the precise distribution of these character
states is unknown. Nevertheless, while therizinosauroids, oviraptorosaurs, dromaeosaurids, avialians, tyrannosaurids, troodontids, and
ornithomimosaurs demonstrably possess these
features, herrerasaurids, ceratosaurs, basal tetanurines, and carnosaurs do not. This pattern is complicated by the lack of a maxillary component to the
secondary palate in Compsognathidae (Sinosauropteryx, pers. observation 1998). An ossified
secondary palate has been demonstrated to be biomechanically advantageous in resisting torsional
forces in the skull (BUSBEY, 1995), which would be
27
consistent with the greater lateral flexibility possible
in the cervical region of coelurosaurs compared to
other theropods (see above). Additionally (and not
mutually exclusive to torsional resistance), a large
ossified secondary palate might allow continued
breathing while the rostrum was occupied with food
manipulation and processing.
In the most parsimonious trees in this analysis all
but one of the maniraptoriform taxa unambiguously
group as either arctometatarsalians or as maniraptorans. However, Troodontidae is equally parsimoniously placed as a paravian maniraptoran (the
sister taxon to the dromaeosaurid-bird clade) or as a
bullatosaurian arctometatarsalian (the sister taxon
to Ornithomimosauria). The summary cladogram
(Fig. 5) follows the first of these options, but character evidence for the second are discussed below.
Furthermore, it requires only one additional evolutionary step to move Troodontidae to a sister group
position to the oviraptorosaur-therizinosauroid
clade within Maniraptora (see Discussion).
NODE cc. ARCTOMETATARSALIA
ALL: 25.1 Nasals narrow caudally behind external nares; 39.1 Rostral portion of frontals relatively
triangular, suture with nasals forms a distinct acute
angle; 55.1 Squamosal recess; 95.1 Three cranial
nerve openings in acoustic fossa; 198.1 Midcaudal
vertebrae with moderately long prezygapophyses,
extending more than one half but less than one centrum length; 199.1 Distal caudal vertebrae with moderate interlocking, prezygapophyses extend more
than one half, but less than one, centrum length;
206.1 Bridge of bone dorsal to haemal canal in distal
chevrons; -278.0 Pollex ungual subequal to unguals
of digits II and III in size; 282.2 Manual ungual crosssection subtriangular, as wide or wider than deep;
287.1 Iliac blades dorsal surface meet along midline; 294.1 Preacetabular process of ilium cranial
margin notched; 326.1 Semicircular scar on caudal
surface of ischium, just distal to iliac process; 341.2
Extensor groove in craniodistal region of femur deep
and conspicuous; 350.2 Crista fibularis well developed; 373.1 Metatarsals II and IV contact at midshaft on the plantar surface; 375.2 Metatarsal III
dorsal surface area clearly smaller than metatarsals
II and IV; 377.1 Arctometatarsus.
ACCTRAN: 251.1 Distal carpal shape flat and
discoidal, no distinct articular surfaces; 283.2 Manual unguals relatively short; 363.1 Round external
fossa at base of ascending process of astragalus;
379.2 Metatarsal I placed distally.
DELTRAN: 7.1 Medial alae from premaxillae
meet in front of vomera; 60.1 Jugal quadratojugal
dorsal process extends further caudally then ventral
quadratojugal process; 72.1 Secondary palate well
T.R. HOLTZ, JR.
mosaurs differs from that of other theropods (except
for troodontids). Also, as WITMER (1997) noted, ornithomimosaurs and tyrannosaurids are the only
two lineages of non-avian theropods known to possess squamosal recesses (55.1).
ossified from premaxilla through one-half the length
of the ventral surface of the maxilla; 132.2 Premaxillary tooth crowns incisiform and reduced in size;
188.1 Sacral neural spines fused to form lamina; 240.0 Humeral ends little or not expanded; -316.0
Caudal portion of pubic boot same length as cranial
portion; 357.1 Cranial protuberance on fibula below
expansion; 369.1 Metatarsus elongate relative to
other theropods of same femoral length; 371.1
Metatarsal cross-section deeper craniocaudally
than mediolaterally at midshaft; -386.0 Pedal ungual
II subequal to pedal ungual III.
As noted by SERENO (1997), there are no possible phylogenetic positions for Troodontidae that do
not result in some level of homoplasy (convergence
or reversals) among maniraptoriform coelurosaurs.
Indeed, as in HOLTZ (1994), troodontids were found
to be arctometatarsalians closer to ornithomimosaurs than to tyrannosaurids in half of the most parsimonious trees in the present analyses. The
following synapomorphies would support such a
placement (only those in all trees or under delayed
transformation are listed): 7.1, 8.1, 25.1, 39.1, 60.1,
71.1, 72.1, 95.1, 132.2, 198.1, 206.1, -278.0, 287.1,
-316.0, 357.1, 363.1, 369.1, 371.1, 373.1, 375.2,
377.1, 379.2. Those characters not in this list but
currently listed as occurring in all optimizations for
node cc are found in Tyrannosauridae and one or
both ornithomimosaur OTUs: these would be regarded as synapomorphies of the expanded Arctometatarsalia (including Troodontidae) under
accelerated transformation. Additionally, tyrannosaurids and troodontids share a well-developed
sagittal crest on the dorsal surface of the parietals
(43.1), unlike ornithomimosaurs and indeed most
other coelurosaurs.
As discussed above, Arctometatarsalia is defined as Ornithomimus and all taxa sharing a more
recent common ancestor with Ornithomimus than
with Neornithes (HOLTZ, 1996b). In half the most
parsimonious trees found in this study (Fig. 3A, 4),
Arctometatarsalia includes only Tyrannosauridae
and Ornithomimosauria; in the other half, Troodontidae is included as well. As HOLTZ (1994) noted,
many of the characters that unite tyrannosaurids
and ornithomimosaurs (and troodontids) are found
in the locomotory apparatus, particularly in the
metatarsus but also in the pelvis and caudal region.
Some of these characters are also found in various
combinations in other taxa (e.g., Elaphrosaurus,
Coelurus, Caenagnathidae, mononykine alvarezsaurids), but no other taxa demonstrate the possession of all of these features in the same animal (with
the exception of Troodontidae). Thus, the majority of
the characters uniting ornithomimosaurs and tyrannosaurids outside of other theropods are associated
with higher cursorial ability (HOLTZ, 1995b) and
might have been convergently acquired by these
two different lineages near the base of Maniraptoriformes (as in SERENO, 1997; MAKOVICKY & SUES,
1998).
Following HOLTZ (1994, 1996b), Bullatosauria is
defined as all descendants of the most recent common ancestor of Troodon and Ornithomimus. Bullatosauria was supported on many (primarily cranial)
characters. W ITMER (1997) and M AKOVICKY &
NORELL (1998) observed several additional cranial
features shared by troodontids and ornithomimosaurs. THULBORN (1984) and PÉREZ-MORENO et al.
(1994) found additional character support for a
troodontid-ornithomimosaur clade. However, as
discussed below, troodontids are supported here as
basal paravian maniraptorans in half of the equally
parsimonious trees in this study. Under this topology, the clade defined by all descendants of the most
recent common ancestor of Troodon and Ornithomimus is exactly the same clade as that defined by the all descendants of the most recent
common ancestor of Neornithes and Ornithomimus.
Thus, Bullatosauria becomes a junior subjective
synonym of Maniraptoriformes under this topology.
Alternatively, when Troodontidae is placed as the
sister group to Ornithomimosauria (and Bullatosauria is thus not the same clade as Maniraptoriformes),
the following synapomorphies would support such a
node (only those found under all optimizations and
under delayed transformation are shown): 1.1, 18.1,
32.1, 46.1, 88.1, 93.1, 94.1, 98.1, 101.1, -116.0,
However, there are other features that unite tyrannosaurids and ornithomimosaurs in the present
study. Some have to do with the forelimb, but could
again be explained by the reduced (or at least altered) grasping ability in both tyrannosaurids and ornithomimosaurs relative to other coelurosaurs
(HOLTZ, 1994). Unlike all other coelurosaurs (again,
except for troodontids), the premaxillary teeth of tyrannosaurids and the basal ornithomimosaur Pelecanimimus are incisiform (132.2), with both carinae
of each tooth being placed along the same plane
perpendicular to the main axis of the skull (differing
from typical avetheropods, in which the premaxillary
teeth are asymmetrical). Furthermore, the morphology of the nasals and of rostral portion of the frontals
(25.1, 39.1) of basal tyrannosaurids and ornithomi-
28
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
126.1, 127.1, 130.1, 135.2, 157.1, 158.1, 163.1, 183.0, -195.0, 201.3, -264.0.
Thus, there is considerable support for a bullatosaurian arctometatarsalian position for Troodontid a e . H o w e v e r, t h e m a n u s o f Tr o o d o n t i d a e
resembles those of non-tyrannosaurid, nonornithomimosaurian coelurosaurs in several features, although these might be explainable as the retention of the primitive state, since lost in other
arctometatarsalians. Similarly, the ischia of troodontids more closely resemble those of oviraptorosaurs, dromaeosaurids, and avialians in being less
than two thirds the length of the pubis (320.1) and
having a distally placed obturator process (323.2).
This might plausibly be the basal condition for Maniraptoriformes, with the condition in Tyrannosauridae and Ornithomimidae representing novel
extensions of the sub-process portion of the ischium, but this ad hoc explanation would require additional evidence (for example, discovery of the
ischia of basal members of the tyrannosaurid and ornithomimosaur lineages).
It should be noted, however, that the most complete skeleton of Troodontidae currently described,
that of the type specimen of Sinornithoides youngi
RUSSELL & DONG, 1993b indicate that some troodontid features once thought to be similar to dromaeosaurids, avialians, and oviraptorosaurs are
now known to resemble the condition in ornithomimosaurs. For example, unlike most other maniraptoriform coelurosaurs, metacarpal III is subequal in
length to metacarpal II (rather than clearly shorter) in
troodontids and ornithomimosaurs (263.0), and (as
in non-maniraptorans) metacarpal III is unbowed
(265.0). While the latter is most parsimoniously explained as the retention of a primitive trait, the former
would best explained a single reversal in the common ancestor of all bullatosaurs if a troodontidornithomimosaur clade were supported.
braincase, may be related to Stokesosaurus and/or
tyrannosaurids (CHURE & MADSEN, 1998) or to dromaeosaurids (CURRIE & ZHAO, 1993b; CURRIE,
1995). Premaxillary teeth with the diagnostic Ushaped horizontal cross-section currently only
known in Tyrannosauridae have been reported from
the ?Aptian-Albian of Japan (M. MANABE, pers. comm. 1997).
NODE dd. ORNITHOMIMOSAURIA
ALL: 1.1 Skull shape elongate and platyrostral,
with obtuse triangular paracoronal cross section;
6.1 Premaxilla long and pointed, with long nasal process; 8.2 Maxillary process of premaxilla extremely
long, extends caudally from the caudal margin of the
external naris for a distance greater than the rostrocaudal length of the external naris; 18.1 Maxillary
fenestra long and low; 19.1 Promaxillary fenestra
dorsal to maxillary fenestra; 32.1 Lacrimal caudal
process at dorsal surface, lacrimal T-shaped; 46.1
Orbit margin with raised rim; 98.1 Parabasisphenoid
bulbous capsule; 101.1 Occipital region directed
ventrocaudally; 136.1 Neck length twice or more
skull length; 141.1 Craniodorsal rim of axial neural
spine convex curve in lateral view; 163.1 Caudal
cervical postzygapophyses elongate; 201.3 Shaft of
cervical ribs short (less than twice centrum length)
and slender; 213.1 Acromion in scapula reduced;
250.1 Ulnar and radial ends closely joined, even by
syndesmosis; -255.0 Semilunate carpal block absent; -263.0 Metacarpal III length subequal to metacarpal II; -264.0 Metacarpal III width not very much
narrower (greater than 50%) than metacarpal II;
269.1 Metacarpal-phalangeal joints not hyperextensible, extensor pits on metacarpals I-III reduced;
276.1 Flexor tubercle of unguals poorly developed
and distally placed; 280.1 Pollex shape stout and robust, dorsoventrally compressed, with proximal articular surface quadrangular.
ACCTRAN: 3.3 Number of premaxillary teeth
seven; 85.1 Orbitosphenoid absent; 93.1 Cranial
tympanic recess invades basisphenoid; 94.1 Internal foramen of facial nerve cranioventral to vestiblocochlear nerve; 107.1 Symphyseal region of
dentary medially recurved; -114.0 Caudal surangular fenestra small pit; 126.1 Number of teeth greater
than 100; 128.2 Tooth serrations absent; 127.1 Dentary teeth more numerous and smaller than maxillary teeth; 130.1 Tooth roots constricted; 135.2
Interdental plates absent in dentary; -168.0 Apices
of dorsal neural spines unexpanded; 182.1 Capitular facet of dorsal ribs situated dorsal to lamina, on
prezygapophyseal base; -195.0 Proximal caudal zygapophyses short; 260.2 Metacarpal I subequal to
metacarpal II in length; 289.1 Fossa for origin of M.
cuppedicus on ilium broad; 348.1 Lateroproximal
Siamotyrannus isanensis BUFFETAUT, SUTEETHORN & TONG, 1996 is a recently discovered form
from the Early Cretaceous (?Barremian) of Thailand, purported to be an ancestral tyrannosaurid, as
is the Late Jurassic North American species Stokesosaurus clevelandi MADSEN, 1976. These taxa
were not included in the present analysis, pending
completion of preparation of additional specimens
and hitherto unreported skeletal elements of these
species (E. B UFFETAUT , pers. commun. 1998;
CHURE & MADSEN, 1998). The former taxon, as well
as the fragmentary Labocania anomala MOLNAR,
1974 are the only theropods other than tyrannosaurids and ornithomimosaurs known to possess a
pronounced semicircular scar on the caudal surface
of the ischium, just distal to the iliac process (326.1):
these taxa thus might be arctometatarsalians. Itemirus medullaris KURZANOV, 1976, known only from a
29
T.R. HOLTZ, JR.
Metatarsals subequal or wider mediolaterally than
craniocaudally at midshaft; 386.1 Pedal ungual II
significantly longer than pedal ungual III.
condyle (fibular condyle) on proximal end of tibia
small and medially situated.
DELTRAN: -37.0 Prefrontals well exposed on
skull roof; -54.0 Postorbital frontal process sharply
upturned; -116.0 Horizontal shelf on lateral surface
of surangular, rostral and ventral to the mandibular
condyle absent or faint ridge; 158.1 Cranial cervical
centra extend beyond caudal extend of neural arch;
251.1 Distal carpal shape flat and discoidal, no distinct articular surfaces.
In the present study, Maniraptora includes not
only birds but also dromaeosaurids, oviraptorosaurs, therizinosauroids, and compsognathids (and
troodontids in half of the equally parsimonious
trees). This composition is similar to that found by
GAUTHIER (1986) (although therizinosauroids, then
called "segnosaurs", were at the time considered
non-theropods). The conclusions here differ from
HOLTZ (1994), where dromaeosaurids were the only
non-avian theropod lineage considered closer to
birds than to Ornithomimus. Unlike SERENO (1997),
tyrannosaurids were not found here to share a more
recent common ancestor with birds than with Ornithomimus.
Following PADIAN, HUTCHINSON & HOLTZ (1999),
Ornithomimosauria is defined as all descendants of
the most recent common ancestor of Pelecanimimus and Ornithomimus.
The recently described Pelecanimus polyodon of
the Early Cretaceous (Barremian) of Spain demonstrates many dental characteristics also found in
troodontids, but lost in the (mostly) edentulous Ornithomimidae. The Early Cretaceous (?AptianAlbian) Harpymimus okladnikovi, coded here within
Ornithomimidae, is more primitive than Pelecanimimus in the retention of a metacarpal I much
shorter than metacarpal II in length. This Asian
taxon is being restudied by Osmólska and PérezMoreno, and may prove to lie outside the
Pelecanimimus-Ornithomimidae (proper) clade in
future analyses. Also of interest is the recent report
of ornithomimosaurian material from the AptianAlbian of Australia (RICH & VICKERS-RICH, 1994)
and from the Barremian of Thailand (BUFFETAUT et
al., 1995; SUTEETHORN et al., 1995).
Compsognathidae is here considered to comprise Compsognathus longipes WAGNER, 1861 and
Sinosauropteryx prima JI & JI, 1996. These taxa
share derived features in an enlarged phalanx 1 of
manual digit I (with shaft diameter greater than that
of the radius) and fan-shaped neural spines on the
dorsal vertebrae (CHEN, DONG & ZHEN, 1998). Both
these taxa are also characterized by tails with
greater than 55 caudal vertebrae (more than those
of all other coelurosaurs) (190.0) and by ischia with a
slight expansion of the distal ischium (327.0). The
former of these traits is most parsimoniously explained by a reversal, while the latter might be a retention of the primitive condition (if the pointed
ischial tip is convergent in tyrannosaurids and advanced maniraptorans) or a reversal (if the condition
in arctometatarsalians and maniraptorans represent a single evolutionary event): see Discussion
below. As with Scipionyx, this taxon is unspecialized
compared to most other advanced coelurosaurs,
and so may be very informative with regard to the
condition of the ancestral maniraptoriform and the
ancestral maniraptoran.
NODE ee. MANIRAPTORA
ALL: 10.1 Premaxilla and nasal do not meet subnarially; -57.0 Squamosal does not constrict lateral
temporal fenestra; 172.1 Caudal edge of dorsal zygapophyses overhangs centrum; 203.1 Medial gastral segment shorter than lateral segment; 298.1
Postacetabular process of ilium with concave caudal margin; 320.1 Ischium less than 66% length of
pubis.
MAKOVICKY & SUES (1998) considered the presence of medial gastral segments shorter than the lateral segments as synapomorphic of a troodontiddromaeosaurid clade, not present in ornithomimosaurs, tyrannosaurids, and non-coelurosaurian
theropods. However, compsognathids appear to
demonstrate the same condition. As such, and given
the current poor understanding of the gastralia in
therizinosauroids, oviraptorosaurs, and basal birds,
this feature (203.1) is most parsimoniously regarded
here as a maniraptoran synapomorphy under the topology placing troodontids as paravians; if troodontids are bullatosaurians, the origin of this derived
state is ambiguous.
ACCTRAN: 89.1 Lateral depression surrounding
opening to middle ear; 192.1 Caudal neural spines
limited to caudals I-IX; 253.2 Distal carpal I block
broadly overlaps metacarpal II dorsally and ventrally; 293.1 Preacetabular portion of ilium significantly longer than postacetabular portion; 381.1
Pedal digit IV larger than II and closer to III in length.
DELTRAN: 37.1 Prefrontals reduced or absent; 54.0 Postorbital frontal process sharply upturned; 116.0 Horizontal shelf on lateral surface of surangular rostral and ventral to the mandibular condyle absent; 158.1 Cranial cervical centra extend beyond
caudal extent of neural arch; -357.0 Cranial protuberance on fibula below expansion absent; -371.0
NODE ff.
30
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
tooth characters 126.1, 127.1, and 128.1, optimized
as present at node ff under accelerated transformation, represent the similarities in the teeth of therizinosauroids and troodontids. Dromaeosaurids lack
these features, and retain the primitive state for
these characters.
ALL: 130.1 Tooth roots constricted; 151.1 Epipophyses on cervical vertebrae placed proximally;
241.1 Deltapectoral crest on humerus expanded
and offset from humeral shaft; 249.1 Ulnar facet for
radius transversely expanded and concave; 296.1
Caudodorsal margin of ilium curves caudoventrally;
297.1 Postacetabular ala of ilium acuminate; 299.1
Supracetabular crest on ilium absent; 323.2 Obturator process of ischium distally placed; 327.1 Ischial
foot absent; 374.1 Metatarsal IV longer than metatarsal II and closer to metatarsal III in length.
In the summary cladogram, the opisthopubic
condition is considered to have evolved independently in therizinosauroids and eumaniraptorans (see
Discussion below). However, if troodontids are
moved to a bullatosaurian arctometatarsalian position (see above), opisthopuby (309.3) and an associated transformation of the ilium (306.1) become
ambiguously polarized. These character states
would be either synapomorphic for node ff, and secondarily lost in Microvenator plus Oviraptorosauria,
or is independently evolved in the dromaeosauridbird clade and in therizinosauroids. MAKOVICKY &
NORELL (1998) also noted the ambiguous nature of
this character. It might be speculated that the convex
curvature of the oviraptorosaurian pubis (310.1)
might represent a "re-propubic" state derived from
the opisthopubic condition: however, additional data
would be needed to support this suggestion.
ACCTRAN: 32.1 Lacrimal caudal process at dorsal surface, lacrimal T-shaped; 126.1 Number of
teeth greater than 100; 127.1 Dentary teeth more
numerous and smaller than maxillary teeth; 128.1
Teeth with large denticles; 163.1 Caudal cervical
postzygapophyses elongate; 166.1 Ventral process
(hypapophyses) on cervicodorsal vertebrae present
as small protrusions; 235.1 Humeral shaft sigmoid;
236.1 Humeral head offset and emarginated ventrally by groove; 243.1 Humeral entepicondyle
prominent; 245.1 Ulnar shaft bowed caudally; 220.1
Metacarpal III bowed laterally; 281.1 Manual unguals II and III with small nubbin proximodistally;
284.1 Manual unguals with extreme curvature;
299.1 Supracetabular crest on ilium absent; 337.2
Fourth trochanter of femur absent; 384.1 Pedal unguals III and IV vertically oval in cross-section.
NODE gg.
ALL: 2.1 Premaxillary teeth absent, presumably
covered with rhamphotheca; 6.1 Premaxilla long
and pointed, with long nasal process; -8.0 Maxillary
process moderately long, premaxilla participates
broadly in ventral surface of external naris; -30.0
Lacrimal not exposed on skull roof; 40.1 Frontal very
broadly exposed on skull roof, postorbital ramus
does not project abruptly laterally from the orbital
margin; 74.1 Vomer limited to rostral region; 79.1
Palatine fenestra (between ectopterygoid and palatine) closed; 100.2 Basipterygoid processes very
short, fused to pterygoids; 107.1 Symphyseal region
of dentary medially recurved; -114.0 Caudal surangular foramen a small pit; 148.2 Postaxial cervical
pleurocoels two pairs present; -197.0 Transition
point in tail absent; 200.2 Distal caudals markedly
shorter than proximal caudals; 201.2 Shaft of cervical ribs short (less than twice centrum length) and
broad; -208.0 Middle chevron shape gentle curvature; -209.0 Distal chevrons lacking cranial and caudal projections; 244.1 Ulnar facet on humerus
expanded, merges with entepicondyle; 292.1 Preacetabular ala of ilium greatly expanded vertically;
335.3 Anterior trochanter of femur cylindrical in
cross section; 367.1 Medial tuber on calcaneum enlarged
DELTRAN: 72.1 Secondary palate well ossified
from premaxilla through one-half the length of the
ventral surface of the maxilla; 103.1 Occipital condyle with constricted neck; -223.0 Sternum shape
relatively round; 240.1 Humeral ends well expanded, greater than 150% midshaft diameter;
253.2 Distal carpal I block broadly overlaps metacarpal II dorsally and ventrally.
This unnamed clade (comparable to the redefined "Maniraptora" of SERENO (1997, 1998)) is
very well supported based on character evidence
(see also GAUTHIER, 1986). Although Troodontidae
is only weakly supported within this clade in the present analysis, and Therizinosauroidea was placed
outside it in the analysis of SERENO (1997, 1998),
most recent workers have recovered a clade comprised of oviraptorosaurs, dromaeosaurids, and
birds to the exclusion of ornithomimosaurs or tyrannosaurids: GAUTHIER (1986); SERENO (1997, 1998);
SUES (1997); MAKOVICKY & SUES (1998); NOVAS &
POL (in press).
Although the derived character state "tooth roots
constricted" (130.1) was found to be synapomorphic
for node ff, this condition is present only in some premaxillary teeth of one member of this clade, Dromaeosauridae (C URRIE & Z HAO , 1993b: fig. 6).
Lateral (maxillary and dentary) teeth of dromaeosaurids do not show this condition. Additionally, the
ACCTRAN: -16.0 Promaxillary fenestra absent; 17.0 Maxillary fenestra absent; -81.0 Ventral ectopterygoid recess absent; 108.1 Rostral half of mandible concave; -120.0 Splenial without notch for
rostral margin of internal mandibular fenestra; 141.1
31
T.R. HOLTZ, JR.
Craniodorsal
Craniodorsal rim
rim of
of axial
axial neural
neural spine
spine convex
convex curve
curve
in
lateral
view;
183.1
Sacral
pleurocoels
in lateral view; 183.1 Sacral pleurocoels present;
present;
187.1
187.1 Caudalmost
Caudalmost sacral
sacral centrum
centrum markedly
markedly smaller
smaller
than
222.1 Sternum
than cranialmost
cranialmost sacral
sacral centrum;
centrum; 222.1
Sternum
carina
-272.0 Length
carina present;
present; -272.0
Length of
of phalanx
phalanx 3
3 of
of manmanual
ual digit
digit III/(sum
III/(sum of
of lengths
lengths of
of phalanges
phalanges 1
1+
+2
2 of
of digit
digit
III)
III) greater
greater than
than 100%;
100%; 275.1
275.1 Manual
Manual ungual,
ungual, dorsal
dorsal
edge
edge of
of articular
articular facet
facet with
with pronounced
pronounced lip
lip on
on dorsal
dorsal
edge;
edge; 289.1
289.1 Fossa
Fossa for
for origin
origin of
of M.
M. cuppedicus
cuppedicus on
on ililium
ium broad
broad
trocaudally; -10.0 Premaxilla and nasal meet subnarially; 12.1 Maxillary teeth absent; 29.1 Nasal recesses present; 42.1 Frontals separated at
medialmost point of suture by rostral process of parietals; 87.1 Paroccipital processes curving ventrally and pendant; 109.1 Dentary rami widely
divergent caudally; -110.0 Overlap of dentary onto
postdentary bones; -111.0 Intramandibular joint absent; 112.2 Dentary caudal depth greater than 220%
depth of dentary symphysis; -115.0 Rostral ramus of
surangular shallow; 122.1 Articular facet for mandibular joint craniocaudally elongate and shallow;
124.1 Retroarticular process elongated and tapering; 178.1 Dorsal column subequal to femur length;
185.4 Number of sacrals six; -195.0 Proximal caudal
zygapophyses short; -196.0 Caudal transverse processes present beyond caudal XV; 262.1 Metacarpal II length about 50% or greater humerus length;
265.1 Metacarpal III bowed laterally; -386.0 Pedal
ungual II subequal to pedal ungual III.
DELTRAN: 7.1 Medial alae from premaxillae
meet in front of vomera; 121.1 Coronoid extremely
reduced or absent; -168.0 Apices of dorsal neural
spines unexpanded; -192.0 Caudal neural spines
present beyond caudal X; 226.1 Forelimb (humerus+radius+manus)/hindlimb (femur+tibia+pes)
length ratio greater than 50% but less than 120%;
243.1 Humeral entepicondyle prominent; 281.1
Manual unguals II and III with small nubbin proximodistally; 293.1 Preacetabular portion of ilium significantly longer than postacetabular portion; 299.1
Supracetabular crest on ilium absent; -316.0 Caudal
portion of pubic boot same length as cranial portion;
337.2 Fourth trochanter on femur absent; 384.1
Pedal unguals III and IV cross section vertically oval
DELTRAN: 141.1 Craniodorsal rim of axial neural spine convex curve in lateral view; 157.1 Cranial
cervicals broader than deep on cranial surface, with
kidney-shaped articular surfaces that are taller laterally than at midline; 187.1 Caudalmost sacral centrum markedly smaller than cranialmost sacral
centrum; 232.2 Radius greater than 76% length of
humerus; 235.1 Humeral shaft sigmoid; 245.1 Ulnar
shaft bowed caudally; 275.1 Manual ungual, dorsal
edge of articular facet with pronounced lip on dorsal
edge; 289.1 Fossa for M. cuppedicus on ilium broad.
As in RUSSELL & DONG (1993a), SUES (1997),
and MAKOVICKY & SUES (1998) but unlike SERENO
(1997), Therizinosauroidea was found to be most
closely related to the oviraptorosaurs among Theropoda. These taxa share numerous cranial and postcranial synapomorphies. Of note is the fact that
many of these derived features are reversals: in
some cases, such as the apparent loss of a transition point in the tail (-159.0) and related caudal transformations (-192.0, 200.2, -208.0, -209.0) and the
exclusion of the lacrimal from the skull roof (-29.0),
are reversals to a pre-neotheropod state. Also as
noted by MAKOVICKY & SUES (1998), the
therizinosauroid-oviraptorosaur clade shares with
Ceratosauria possession of two pleurocoels on the
postaxial cervicals (148.2).
NODE ii. OVIRAPTOROSAURIA
ALL: 106.1 Dentary symphysis fused; 191.2
Caudal pleurocoels present in centrum.
ACCTRAN: None.
DELTRAN: 12.1 Maxillary teeth absent; 87.2 Paroccipital process curving ventrally and pendant; 88.0 Basicranium pneumatization minimal to moderate, but no expansion of basisphenoid; -108.0
Rostral half of mandible ventrally convex or straight;
-110.0 Overlap of dentary onto postdentary bones; 111.0 Intramandibular joint absent; -115.0 Rostral
ramus of surangular shallow; 122.1 Articular facet
for mandibular joint craniocaudally elongate and
shallow; 124.1 Retroarticular process elongated
and tapering; 183.1 Sacral pleurocoels present;
185.4 Number of sacrals six; 186.2 Sacrals III-V dorsoventrally flattened; 188.1 Sacral neural spines
fuse to form lamina; -195.0 Proximal caudal zygapophyses short; 265.1 Metacarpal III bowed laterally.
NODE hh.
ALL: 104.1 Dentary teeth absent; -118.0 External
mandibular fenestra large, horizontally oval; 156.1
Caudal cervical neural arch forms X-shape in dorsal
view; -169.0 Scars for interspinous ligaments terminate at apex of neural spine in dorsal vertebrae;
171.1 Dorsal transverse processes short, wide and
only slightly inclined; 176.1 Dorsal centrum transverse section wider than high; 310.1 Pubic shaft with
marked concave curvature cranially; 358.1 Fibular
tubercle for M. iliofibularis (="anterolateral process")
laterally projecting.
Following the phylogenetic definitions in PADIAN,
HUTCHINSON & HOLTZ (1999), Oviraptorosauria is
defined as all descendants of the most recent common ancestor of Oviraptor and Chirostenotes. As
such, Microvenator is excluded from Oviraptorosau-
ACCTRAN: 5.1 Premaxilla subnarially very
deep, main body taller dorsoventrally than long ros-
32
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
present as small protrusions; 192.1 Caudal neural
spines limited to caudals I-IX; 235.1 Humeral shaft
sigmoid; 236.1 Humeral head offset and emarginated ventrally by groove; 245.1 Ulnar shaft bowed
laterally; 284.1 Manual unguals extremely curved;
381.1 Pedal digit IV larger than II and closer to III in
length.
ria proper in one out of the two topologies of maniraptoriform relationships in this study (Fig. 3C).
Nevertheless, this small Early Cretaceous (?AptianAlbian) taxon is clearly more closely related to Oviraptoridae and Caenagnathidae than to any other
known taxon. As in SUES (1997) and MAKOVICKY &
SUES (1998), this study found very strong support
for uniting Oviraptoridae and Caenagnathidae to the
exclusion of Therizinosauroidea. Note that this
study agrees with CURRIE, GODFREY & NESSOV
(1993), SUES (1997), MAKOVICKY & SUES (1998),
and SERENO (1997), and disagrees with HOLTZ
(1994) in recognizing the oviraptorosaurian nature
of Caenagnathidae ("Elmisauridae" in the latter
study), necessitating an independent evolution of
the arctometatarsus in caenagnathid oviraptorosaurs and in true arctometatarsalians (and in troodontids, if they are not arctometatarsalians, and in
mononykine alvarezsaurids: see also Discussion
below).
SERENO (1997) coined the term "Paraves" for
birds and all theropods sharing a more recent common ancestor with birds than with oviraptorosaurs;
in the present analysis this clade contains Dromaeosauridae (and Troodontidae in half the trees) as well
as birds. A clade containing troodontids, dromaeosaurids, and birds has been recovered by a number
of studies (GAUTHIER, 1986; SERENO, 1997, 1998;
FORSTER et al., 1998; MAKOVICKY & SUES, 1998;
NOVAS & POL, in press). HOLTZ (1994) has been one
of the few explicit analyses to discover a different relationship, with troodontids closer to ornithomimosaurs than to birds.
However, unlike previous studies supporting
troodontid-dromaeosaurid-bird monophyly, this
analysis did not support a troodontiddromaeosaurid clade to the exclusion of birds (as in
GAUTHIER, 1986 (where it was coded as such a priori); SERENO, 1997, 1998; MAKOVICKY & SUES,
1998; NOVAS & POL, in press) nor a troodontid-bird
clade to the exclusion of dromaeosaurids (as in FORSTER et al., 1998). Instead, troodontids were found
to lie outside the dromaeosaurid-bird clade Eumaniraptora in all twenty most parsimonious trees (and
indeed to lie outside Maniraptora in half the trees).
NODE jj. PARAVES
ALL: 162.1 Carotid process on caudal cervical
vertebrae; 194.1 Centra of caudals I-V box-like with
increased flexural capability; 197.3 Transition point
in caudals I-IX; 204.1 Chevron transition between
caudal X and XVII; -274.0 First phalanx of pollex less
than or subequal to length of metacarpal II; 383.1
Pedal digit II hyperextensible; 385.1 Pedal ungual II
sickle-shaped (blade-like cross section and highly
recurved).
ACCTRAN: 92.1 Posttympanic recess confined
to columnar process; 101.1 Occipital region directed
ventrocaudally; 102.1 Foramen magnum taller than
wide; 119.1 Splenial with extensive triangular exposure in lateral view between dentary and angular;
129.1 Posterior serrations much larger than anterior
serrations in maxillary and dentary teeth; 135.2 Interdental plates absent in dentary; 198.1 Midcaudal
vertebrae with moderately long prezygapophyses,
extending more than one half but less than one centrum length; 200.1 Distal caudals more than 130%
length of proximal caudals; 210.1 Distal chevron
cranial and caudal bifurcations; 379.2 Metatarsal I
placed distally.
Nevertheless, several characters do support a
troodontid-dromaeosaurid clade (119.1, 129.1,
210.1) while others support a troodontid-bird grouping (92.1, 101.1, 102.1, 135.2, 198.1, 379.2). It
should be noted here, however, that Troodontidae
was incorrectly coded as possessing ischiadic terminal processes separate (328.1 of this analysis) in
the study of FORSTER et al. (1998), resulting in a basal avialian position for this clade. The terminal processes of the ischia of troodontids observed by the
present author are indeed in contact, as attested by
the flat articular surfaces in some specimens (e.g.,
Troodon formosus, RTMP 86.77.2) and fusion of the
ischiadic symphyses in others (e.g., Saurornithoides mongoliensis, AMNH 6516) (see also NORELL &
M AKOVICKY , 1997). Significantly, as discussed
above, additional character evidence equally
strongly supports a sister group relationship between ornithomimosaurs and troodontids outside of
Maniraptora. Troodontids thus remain one of the
most poorly resolved of maniraptoriform clades with
regards to their phylogenetic position, despite the
fact that this taxon is known from a number of excellent specimens.
DELTRAN: 32.1 Lacrimal caudal process at dorsal surface present, lacrimal T-shaped; -67.0 Quadrate foramen large and situated between quadrate
and quadratojugal; 81.2 Ventral ectopterygoid recess present and subcircular; 91.1 Branches of internal carotid artery enter hypoglosseal fossa
through single common opening; 140.1 Axial neural
spine compressed mediolaterally; 157.1 Cranial
cervicals broader than deep on cranial surface, with
reniform (kidney-shaped) articular surfaces that are
taller laterally than at midline; 166.1 Ventral processes (hypapophyses) on cervicodorsal vertebrae
33
T.R. HOLTZ, JR.
pal II about 50% or greater humerus length; 265.1
Metacarpal III bowed laterally; 306.1 Pubic peduncle of ilium extends more ventrally than ischiadic peduncle; 309.3 Opisthopubic; 334.2 Anterior (=
lesser) trochanter nearly confluent with femoral
head and greater trochanter; 343.1 Adductor fossa
and associated caudodistal crest of distal femur reduced or absent; 348.1 Lateroproximal condyle
(fibular condyle) on proximal end of tibia small and
medially situated; 350.2 Crista fibularis well developed; 351.2 Crista fibularis distally placed.
Curiously, the presence of Troodontidae, Dromaeosauridae, and Rahonavis as serial sister taxa
to Aves strongly supports the presence of a hyperextensible digit II (383.1) with a sickle-shaped claw
(385.1) as the ancestral condition for birds. In trees
where Troodontidae is considered an arctometatarsalian, however, the sickle claw is equally parsimoniously considered as basal to Eumaniraptora
(under accelerated transformation) or as convergently acquired by dromaeosaurids and Rahonavis
(under delayed transformation). A recently described fragmentary taxon, Megaraptor namunhuaiquii NOVAS, 1998, also possesses a sickleshaped pedal digit II ungual. However, N OVAS
(1998) noted that this form lacked derived features
(such as the bowed ulna) that characterize dromaeosaurids and troodontids.
ACCTRAN: 217.2 Coracoid subrectangular, dorsoventral depth more than 130% of craniocaudal
width.
DELTRAN: -89.0 Lateral depression surrounding opening to middle ear absent; -126.0 Number of
teeth less than 100; -127.0 Dentary and maxillary
teeth subequal in number and size; 226.1 Forelimb
(humerus+radius+manus)/hindlimb (femur
+tibia+pes) length ratio greater than 50% but less
than 120%; 232.2 Radius length greater than 76%
humerus length; 243.1 Humeral entepicondyle
prominent; 316.2 Pubic boot cranial portion absent; 384.0 Pedal unguals III and IV cross-section subtriangular.
Related to this, as observed by PAUL (1988),
SERENO (1997), and FORSTER et al. (1998) is the
shared presence of a hyperextensible pedal digit II
(383.1) in troodontids, dromaeosaurids, Rahonavis,
Archaeopteryx, and some basal ornithothoracines.
In the present study, this character state is considered synapomorphic for Paraves, and lost in Alvarezsauridae and advanced ornithothoracine
birds. The presence of this structure in some forms
(such as Archaeopteryx) which lack a trenchant
sickle-claw suggests that the hyperextension of
pedal digit II was employed for purposes other than
predation (OSTROM, 1969a, b) in at least some
paravians. Whether this structure may have originally served a predatory function, and was exapted
into a scansorial function (or vice versa) lie beyond
the scope of the present study.
PADIAN, HUTCHINSON & HOLTZ (1999) proposed
"Eumaniraptora" for the clade comprised of all descendants of the most recent common ancestor of
Deinonychus and Neornithes. Note that this term is
not synonymous with SERENO's (1997) Paraves,
which represents a more inclusive, stem-defined
clade containing the node-defined Eumaniraptora.
In the present study Troodontidae represents a
paravian taxon that was not also a eumaniraptoran
(in half the trees; in the other half, it is an arctometatarsalian clade).
Of minor note: NORELL & MAKOVICKY (1997) correctly identified a typographical error in the description of a character state uniting dromaeosaurids and
birds used in HOLTZ (1994). They correctly observed
that this character state (381.1 in the present study)
is in fact pedal digit IV is longer than pedal digit II.
(The character state incorrectly written in HOLTZ
(1994) as "pedal digit II longer than pedal digit IV" actually describes the condition in Dilophosaurus
(WELLES, 1984)).
NODE kk. EUMANIRAPTORA
ALL: 63.1 Quadratojugal T-shaped; 87.1 Paroccipital process with conspicuous twist in the distal
end orienting distal surface more dorsally than proximal region; 125.1 Vertical columnar process on retroarticular process; 144.1 Axial epipophyses
prominent; 207.1 Proximal chevron shape dorsoventrally depressed; 237.1 Internal tuberosity (= ventral tubercle) on proximal end of humerus well
differentiated and angular; 238.1 Internal tuberosity
on proximal end of humerus craniocaudally compressed and longitudinally elongate; 262.1 Metacar-
34
In the phylogenetic taxonomy of P ADIAN ,
H UTCHINSON & H OLTZ (1999), the two named
branches of Eumaniraptora are Deinonychosauria
COLBERT & RUSSELL, 1969 (Deinonychus and all
taxa sharing a more recent common ancestor with
Deinonychus than with Neornithes) and Avialae
GAUTHIER, 1986 (Neornithes and all taxa sharing a
more recent common ancestor with Neornithes than
with Deinonychus). In the present study, as in HOLTZ
(1994), but unlike SUES (1997), SERENO (1997), and
MAKOVICKY & SUES (1998), Troodontidae was not
found to be a member of Deinonychosauria: see
above. The only OTU within Deinonychosauria in
the present analysis is Dromaeosauridae.
Dromaeosaurids thus remain the closest known
lineage of "traditional" theropod to birds. Placing
troodontids as close or closer to birds than dromaeosaurids would require the shared derived characters
described above to have been either a) convergently acquired in Dromaeosauridae and Avialae or
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
Pollex ends at level of mid-length of phalanx 1 of digit
II; -277.0 Manual ungual, region palmar to ungual
groove wider than region dorsal to ungual groove;
335.4 Anterior trochanter of femur forms trochanteric crest (fusion of greater and anterior trochanters); 360.2 Fibular distal end pinches out less
than half way down tibia length
b) present in basal Paraves and subsequently lost in
Troodontidae. (See Discussion below)
Of additional note, several character state reversals listed in the DELTRAN category for this clade (89.0, -126.0, -127.0) are required in this topology as
the derived state is present in both troodontids on
one branch and therizinosauroids and/or oviraptorosaurs on the other. Thus, under some optimizations
these characters were hypothesized to be present
ancestrally in Paraves, and would then have to be
reversed in Eumaniraptora. However, these conditions might alternatively support a sister group relationship between troodontids and the
therizinosauroid-oviraptorosaur clade to the exclusion of Paraves (see Discussion).
DELTRAN: 163.1 Caudal cervical postzygapophyses elongate; -168.0 Apices of dorsal neural
spines unexpanded; -198.0 Midcaudal vertebrae
with short prezygapophyses, extending less than
one half centrum length; -210.0 Distal chevron cranial and caudal bifurcations absent; 293.1 Preacetabular portion of ilium significantly longer than
postacetabular portion; -369.0 Metatarsus proportions moderate; 379.2 Metatarsal I placed at distal
end of metatarsal II
NODE ll. AVIALAE
ALL: 173.1 Vertebral foramen/cranial articular
facet ratio (vertical diameters) of dorsals 0.4 or
greater; 190.2 Number of caudals less than 25;
216.1 Glenoid oriented laterally; 246.1 Proximal ulnar shaft arched; 293.1 Preacetabular portion of ilium significantly longer than postacetabular portion;
325.1 Ischial proximodorsal process just distal to iliac process; 328.1 Ischiadic terminal processes
separate; -355.0 Proximal region of fibular medial
face flat; -374.0 Metatarsal IV subequal in length to
metatarsal II; 380.2 Metatarsal I completely reverted
NODE mm. AVES
ALL: 178.1 Dorsal column subequal to femur
length; -195.0 Proximal caudal zygapophyses short;
313.2 Pubic apron strongly reduced transversely
and restricted to distal 25% or less of pubic length;
323.3 Obturator process absent, caudoventral margin of ischium smooth from obturator notch to tip; 385.0 Pedal ungual II same shape as other pedal unguals
ACCTRAN: 1.2 Skull shortened and platyrostral,
with acute triangular paracoronal cross-section; 4.0 Premaxillary symphyseal region V-shaped in
ventral view; 6.1 Premaxilla long and pointed with
long nasal process; 51.1 Postorbital-jugal contact
absent; 55.1 Squamosal recess; -64.1 Dorsal ramus
of quadratojugal does not contact squamosal; 74.1
Vomer extends caudally to basicranium; 76.2 Palatine triradiate (no jugal process); 82.2 Endocranial
cavity greatly enlarged, temporal musculature fails
to extend origin onto frontals; -103.0 Occipital condyle constricted neck absent; -114.0 Caudal surangular foramen small pit; 124.1 Retroarticular
process elongated and tapering; 141.1 Craniodorsal rim of axial neural spine convex curve in lateral
view; -152.0 Caudal cervical epipophyses short; 169.0 Scars for interspinous ligaments terminate at
apex of neural spine in dorsal vertebrae; 184.1 First
sacral procoelous; 218.1 Coracoid caudoventral
process length more than twice glenoid diameter;
220.1 Coracoid angle with scapula at level of glenoid
cavity sharp; 221.1 Sternal plates fused; 227.2 Forelimb/presacral vertebral series length ratio about
200% or more; 230.1 Humerus/ulna length ratio less
than or equal to 100%; 247.1 Diameter of ulnar shaft
much thicker than that of radius; -263.0 Metacarpal
III length subequal to metacarpal II; 269.1
Metacarpal-phalangeal joints not hyperextensible,
extensor pits on metacarpals I-III reduced; 273.1
35
ACCTRAN: 171.1 Dorsal transverse processes
short, wide and only slightly inclined; 174.1 Dorsal
hyposphene-hypantrum accessory articulations absent; 177.1 Dorsal centrum ends biconvex
DELTRAN: 1.2 Skull shortened and platyrostral,
with acute triangular paracoronal cross-section; 4.0 Premaxillary symphyseal V-shaped in ventral
view; 6.1 Premaxilla long and pointed with long nasal process; 51.1 Postorbital-jugal contact absent;
55.1 Squamosal recess; -64.1 Dorsal ramus of
quadratojugal does not contact squamosal; 74.1 Vomer extends caudally to basicranium; 76.2 Palatine
triradiate (no jugal process); 92.1 Posttympanic recess confined to columnar process; 101.1 Occipital
region directed ventrocaudally; 102.1 Foramen
magnum taller than wide; -103.0 Occipital condyle
constricted neck absent; -114.0 Caudal surangular
foramen small pit; 121.1 Coronoid extremely reduced or absent; -119.0 Splenial obscured or only
slightly visible in lateral view; -120.0 Splenial without
notch on rostral margin for internal mandibular fenestra; 128.2 Serrations absent; -132.0 Premaxillary
tooth crowns conical; 141.1 Craniodorsal rim of axial
neural spine convex curve in lateral view; -169.0
Scars for interspinous ligaments terminate at apex
of neural spine in dorsal vertebrae; 218.1 Coracoid
caudoventral process length more than twice glenoid diameter; 220.1 Coracoid angle with scapula at
level of glenoid cavity sharp; 221.1 Sternal plates
T.R. HOLTZ, JR.
teryx zoui JI et al., 1998 were not included in the
present analysis (but these spectacularly preserved
taxa are included in work in preparation by the present author). The preliminary phylogenetic analysis
of JI et al. (1998) placed Caudipteryx as a non-avian
avialian (following the taxonomy here), and Protarchaeopteryx as either a non-avian avialian, a deinonychosaur, or as the sister group to Eumaniraptora.
fused; -263.0 Metacarpal III length subequal to
metacarpal II; 273.1 Pollex ends at level of midlength of phalanx 1 of digit II; 299.1 Supracetabular
crest on ilium absent; -386.0 Pedal ungual II subequal in length to pedal ungual III
GAUTHIER (1986) explicitly gave his new taxon
Avialae a stem-based definition (p. 36), although
some authors (NOVAS & PUERTA, 1997; JI et al.,
1998) have subsequently used this term as a nodedefined taxon (all descendants of the most recent
common ancestor of Archaeopteryx and modern
birds). In this study, one taxon was found to share a
more recent common ancestor with neornithine
birds than with Deinonychus, but which were found
to lie outside the clade of Archaeopteryx plus Neornithes. This is the recently discovered species Rahonavis ostromi (F ORSTER et al., 1998) of the
?Campanian of Madagascar. Similarly, the fragmentary form Unenlagia comahuensis NOVAS &
PUERTA, 1997 (of the Turonian of Argentina) was
also found to lie within Avialae but outside of Aves in
some of the most parsimonious trees in which it was
included. If future studies support this topology, or
those which place Unenlagia closer to modern birds
than to Rahonavis, this would support an hypothesis
of secondary (rather than primary) flightlessness in
the large-bodied Unenlagia. However, there is no
phylogenetic evidence found here for hypotheses of
secondary flightlessness in non-avialian maniraptorans (dromaeosaurids, troodontids, oviraptorosaurs, etc.) or other non-avialian coelurosaurs (as
suggested by THULBORN (1984) and PAUL (1988)).
NODE hh. METORNITHES
ALL: -34.0 Slot in ventral process of lacrimal for
jugal absent; 59.1 Jugal postorbital process absent;
65.1 Quadrate articulates with both prootic and
squamosal, and the later contacts neither the quadratojugal nor the postorbital; 222.1 Sternum carina
present; 223.1 Sternum longer craniocaudally than
wide mediolaterally; 224.1 Sternum much greater
than coracoid length; 239.2 Internal tuberosity
(=ventral tubercle) of humerus projected caudally,
separated from humeral head by deep capital incision; 242.1 Humeral distal condyle with only cranial
aspect; 248.1 Ulnar distal condyle subtriangular
shaped in distal view, with a dorsomedial condyle,
and twisted more than 54 degrees with respect to the
proximal end; 252.1 Carpometacarpus (distal carpals fused to each other and to metacarpus); 258.1
Metacarpal III present, without ungual; -350.1 Crista
fibularis present, not well developed; 358.1 Fibular
tubercle for M. iliofibularis (="anterolateral process")
laterally projecting
ACCTRAN: 134.1 Dentary tooth implantation set
in paradental groove; 166.2 Ventral processes (hypapophyses) on cervicodorsal vertebrae very well
developed; 189.1 Synsacrum present in adults;
259.1 Metacarpal II absent; -284.0 Manual ungual
curvature straight; -288.0 Iliac preacetabular fossa
for M. cuppedicus absent; -320.0 Ischium greater
than 75% length of pubis; 366.1 Astragalocalcaneum (astragalus fused to calcaneum); 368.1
Tibiotarsus (astragalocalcaneum fused to tibia); 383.0 Pedal digit II not hyperextensible
As mentioned previously, the pes of Rahonavis
resembles that of troodontids and dromaeosaurids
in the possession of a hyperextensible pedal digit II
terminating in a sickle claw. The pes of Unenlagia is,
regrettably, unknown. Sickle clawed Megaraptor occurs in the same deposits as Unenlagia, but what little is known of the anatomy of Megaraptor suggests
that it is more distantly related to birds than are dromaeosaurids (NOVAS, 1998), while Unenlagia appears to be a basal bird in this analysis and those of
NOVAS & PUERTA (1997) and FORSTER et al. (1998).
DELTRAN: 177.1 Dorsal centrum ends biconvex; 184.1 First sacral procoelous
Because Rahonavis is incompletely known (e.g.,
neither cranial nor manual material has been recovered), there are many characteristics shared by Archaeopteryx and more advanced birds that may at a
later date be found to be synapomorphic for a more
inclusive clade.
Following S ERENO (1997, 1998), P ADIAN &
CHIAPPE (1998), and PADIAN, HUTCHINSON & HOLTZ
(1999), and contra GAUTHIER (1986), the term Aves
is used for the clade comprised of all descendants of
the most recent common ancestor of Archaeopteryx
and Neornithes (see the first four papers for justification). Unlike PAUL (1988), this analysis did not support the hypothesis that dromaeosaurids were more
closely related to Archaeopteryx than either are to
modern birds, nor the hypothesis that ornithomimosaurs, oviraptorosaurs, and troodontids shared a
more recent common ancestor with Neornithes than
did Archaeopteryx.
Unenlagia and Rahonavis share one potential
synapomorphy (vertically oriented pubis, 263.3), although this character state is known in at least one
specimen of Archaeopteryx (WELLNHOFER, 1993;
FORSTER et al., 1998).
The recently described Early Cretaceous taxa
Protarchaeopteryx robusta JI & JI ,1997 and Caudip-
36
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
tive combinations of various theropod taxa, in light of
the character evidence presented here; and the implications for the present new phylogeny of theropod
dinosaurs when mapped onto stratigraphic time.
As in the analyses of PERLE et al. (1993, 1994),
CHIAPPE, NORELL & CLARK (1996, 1998), NOVAS
(1996, 1997a), and FORSTER et al. (1998), the highly
apomorphic Alvarezsauridae were found to be birds
more closely related to Neornithes than is Archaeopteryx (a result surprising to the present author). As
discussed in those previous works, and shown in the
character state lists above, alvarezsaurids were
found to share more derived characters with Ornithothoraces (all descendants of the most recent
common ancestor of Iberomesornis and Neornithes: see PADIAN, HUTCHINSON & HOLTZ, 1999). It
did not support the conclusion of MARTIN (1997),
whose phylogenetic analysis (itself methodologically problematic, as the all-zero outgroup was used
as an OTU in that study) recovered an ornithomimosaurian position for the alvarezsaurid Mononykus
olecranus PERLE et al., 1993. Following PERLE et al.
(1993), the term Metornithes is used for the clade
comprised of all descendants of the most recent
common ancestor of Mononykus and Neornithes.
REVISED DISTRIBUTION OF PREVIOUS
SYNAPOMORPHIES OF THEROPOD SUBCLADES
The inclusion of many new characters and taxa in
the present analysis results, not surprisingly, in new
distributions for character states previously hypothesized to characterize various subclades of
theropods. Some of these characters are discussed
below.
Prior to the extensive revision of archosaur facial
pneumaticity by WITMER (1997), it was hypothesized that the small opening in the rostral region of
the antorbital fossa of non-coelophysid ceratosaurs
was homologous to the maxillary fenestra of tetanurines (GAUTHIER, 1986; HOLTZ, 1994). However,
as WITMER has cogently argued, this structure might
be more appropriately homologized with the promaxillary fenestra, a structure previously thought to
have arisen only among more advanced tetanurines
(HOLTZ, 1994; SERENO et al., 1994, 1996). The promaxillary fenestra (16.1) is present in Spinosauridae, Eustreptospondylus, Piatnitzkysaurus,
Afrovenator, and most avetheropods, but is not
found in Torvosaurus. As such, the most parsimonious set of trees for basal tetanurines suggest that
this condition was present in the ancestor of all tetanurines studied here, and that its lack in Torvosaurus
(and in the derived caenagnathid and therizinosauroid coelurosaurs) are secondary reversals.
As N OVAS (1996, 1997a), N OVAS & P OL (in
press), and SERENO (1997) observed, primitive alvarezsaurids (Alvarezsaurus calvoi BONAPARTE,
1991 and Patagonykus puertai NOVAS, 1996) lack
some derived features found in both Ornithothoraces and the derived mononykine alvarezsaurids
Mononykus, Parvicursor remotus KARHU & RAUTIAN, 1996, and Shuvuuia deserti CHIAPPE, NORELL
& CLARK, 1998: these character states, previously
proposed as synapomorphies of Mononykus and
Ornithothoraces (e.g., PERLE et al., 1993), are best
explained as convergences between advanced alvarezsaurids and ornithothoracines. The present
analysis agrees with NOVAS (1996, 1997a) and disagrees with SERENO (1997) and NOVAS & POL (in
press), however, in proposing that the lack of some
derived features of the postcranium found in Archaeopteryx and Ornithothoraces are reversals to a
pre-avian condition, rather than evidence of a nonavian phylogenetic position of Alvarezsauridae.
Given that the promaxillary fenestra is not present in Ceratosaurus and in Coelophysidae (although
both these forms exhibit a small dimple which does
not perforate the maxilla in approximately the same
location, condition 16.3), but is present in Abelisauridae and Dilophosaurus, this character has an uncertain distribution among Ceratosauria as a whole. As
the most parsimonious trees in the current study resolve this character as present in basal tetanurines,
it is considered to be present in Neotheropoda
ancestrally under accelerated transformation, and
thus the absence of this structure in Ceratosaurus
and Coelophysidae is explained by convergence.
However, under delayed transformation it is optimized as absent in Ceratosauria ancestral, and thus
evolving independently in tetanurines, abelisaurids,
and Dilophosaurus.
The presence of an arctometatarsus even more
specialized than that found in Caenagnathidae or
Arctometatarsalia in Mononykus and Parvicursor
(377.2; see HOLTZ (1995b): p. 511) is interpreted as
convergence with the former two taxa, an hypothesis strongly supported by the lack of this condition in
the more basal alvarezsaurids Patagonykus and Alvarezsaurus or in other eumaniraptorans.
DISCUSSION
The origin of the maxillary fenestra (17.1) is likewise ambiguous, although it is not present in any
known ceratosaurian. This structure can clearly be
identified if the promaxillary fenestra itself can also
be demonstrated in the same specimen, as the
opening between the promaxillary fenestra and the
Although many attributes of theropod phylogenetic history can be examined given the new hypothesis proposed here, three main aspects will be
discussed. These are: new distributions for derived
character states previously hypothesized as synapomorphies of various theropod subclades; alterna-
37
T.R. HOLTZ, JR.
to form a furcula in allosaurids and an unnamed "carcharodontosaurid" (ALCOBER et al., 1998) (neither
separate clavicles nor fused furculae have been reported for other carnosaurian taxa) and various coelurosaurian clades, but are apparently unfused
where known among ceratosaurs (the neoceratosaur Carnotaurus (BONAPARTE, NOVAS & CORIA,
1990); Segisaurus (CAMP, 1936)). This indicates
that, minimally, the clavicles were fused into a furcula in the basalmost avetheropod (i.e., the most recent common ancestor of carnosaurs and
coelurosaurs). However, as neither separate clavicles nor furculae have been reported among the
non-avetheropod tetanurines, it may be that this
structure arose even more basally among theropods
(see also SERENO, 1997).
internal antorbital fenestra. This is clearly the case in
Afrovenator and most carnosaurs and coelurosaurs. In "Megalosaurus" hesperis (BMNH R332)
the lateral aspect of the maxilla is not exposed, so it
is uncertain whether the large opening visible from
the medial side is a maxillary fenestra or a promaxillary fenestra (WITMER, 1997). The very large openings in the rostral part of the antorbital fossa of
Monolophosaurus and Giganotosaurus are difficult
to homologize with other tetanurines: they might be
very large promaxillary fenestrae, true maxillary fenestrae (in which case the promaxillary fenestrae
have been lost), or structures representing the fusion of these two openings (a possibility suggested
by WITMER, 1997). In Carcharodontosaurus only a
single, smaller fenestra is present in the rostral portion of the antorbital fossa: again, it is uncertain if this
is the promaxillary fenestra or the maxillary fenestra.
In this genus (as in the abelisaurids), the maxillary
antorbital fossa as a whole is reduced (15.2), resulting in a very large internal antorbital fenestra.
In any case, the lack of both promaxillary and
maxillary fenestrae in caenagnathid oviraptorosaurs and therizinosauroids are considered reversals. The absence of both fenestrae in these
coelurosaurs might conceivably be synapomorphic
for the oviraptorosaur-therizinosauroid clade, but
this would require the redevelopment of these same
structures in oviraptorid oviraptorosaurs.
In the analysis of HOLTZ (1994), the presence of
the jugal participating in the margin of the internal
antorbital fenestra (58.1) was hypothesized to be
synapomorphic of "Maniraptora" (properly Ornitholestes plus Maniraptoriformes: HOLTZ, 1996b). It
was also noted to be present in the coelophysoid
Dilophosaurus and the abelisaurid Carnotaurus (resulting in a coding of "(0,1)" for this feature, Charact e r 8 6 , f o r A b e l i s a u r i d a e i n H O LT Z , 1 9 9 4 ) .
Subsequent to that study, this condition has been
observed in various other non-maniraptoriform taxa,
including Monolophosaurus (Z HAO & C URRIE ,
1993), Sinraptor (CURRIE & ZHAO, 1993a), and
Afrovenator (SERENO et al., 1994). It is also present
in Herrerasaurus (SERENO & NOVAS, 1994), a form
that may be the sister taxon to the ceratosaurtetanurine clade Neotheropoda. Because of the
greater distribution of this feature than previously
recognized, the current analysis considers it potentially a neotheropod, or even a herrerasaurid plus
neotheropod, synapomorphy under accelerated
transformation.
The furcula (225.1) was once known only in derived coelurosaurs (BARSBOLD, 1983; GAUTHIER,
1986). Its discovery in taxa more distantly related to
birds (CHURE & MADSEN, 1996; MAKOVICKY & CURRIE, 1998) indicates it has a much broader distribution than previously realized. The clavicles are fused
38
The development of the sternum in theropod
phylogeny has a rather perplexing distribution. It
seems to be poorly preserved in many taxa (being
recovered in only one specimen of Archaeopteryx,
for example, despite the excellent preservation of
several of the other specimens: W ELLNHOFER ,
1993). In birds and in some oviraptorids (BARSBOLD,
1983) the sternal plates are fused into a single medial element, while in dromaeosaurids (NORELL &
MAKOVICKY, 1997), tyrannosaurids (LAMBE, 1917),
ornithomimosaurs (PÉREZ-MORENO et al., 1994),
Scipionyx (DAL SASSO & SIGNORE, 1998), and abelisaurids (BONAPARTE, NOVAS & CORIA, 1990) there is
a pair of unfused elements. The recent discovery of
sternal plates fused into a single median element
(221.1) with a ventral ridge (222.1) in spinosaurids
(CHARIG & MILNER, 1997) and sinraptorid carnosaurs (CURRIE & ZHAO, 1993a), similar to the condition found in oviraptorids and birds, may be
convergent with these coelurosaurs. Alternatively,
however, such fusion might be ontogenetically controlled (CURRIE & ZHAO, 1993a; CHARIG & MILNER,
1997), in which case the specimens of theropods
with unfused sternals represent subadult or juvenile
individuals (an hypothesis already proposed for the
type of Scipionyx: DAL SASSO & SIGNORE, 1998).
There has been much discussion of the semilunate carpal block (255.1) and its significance in
theropod phylogeny (O STROM , 1969b, 1975a,
1975b, 1976, 1995, 1997; BARSBOLD, 1983; GAUTHIER & PADIAN, 1985; GAUTHIER, 1986; CHATTERJEE,
1988, 1997; FEDUCCIA, 1996; PADIAN & CHIAPPE,
1997, 1998; CURRIE & CARPENTER, in press). Although OSTROM (1969a, b) identified this structure
as the radiale, PADIAN & CHIAPPE (1997, 1998) and
CURRIE & CARPENTER (in press) argue that this
block is homologous to the fusion of distal carpals 1
and 2, a view accepted here. Indeed, in therizinosauroids (B ARSBOLD , 1983; R USSELL & D ONG ,
1993a) these bones are incompletely fused, demonstrating that this block is composed of two elements.
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
and basal coelurosaurs rather than the condition in
oviraptorosaurs, troodontids, dromaeosaurids, or
basal birds. However, if troodontids are indeed the
sister taxon to ornithomimosaurs, the distribution
becomes more problematic: either the ancestral
maniraptoriform had a large semilunate carpal
block, subsequently reduced in tyrannosaurids and
lost in ornithomimosaurs, or troodontids developed
an expanded semilunate carpal independently of
true maniraptorans.
Previously known only from relatively derived coelurosaurian forms (GAUTHIER, 1986; HOLTZ, 1994; OSTROM , 1997), this structure is now identified in
carnosaurs (Allosaurus sp., Tate Museum unpublished material; Acrocanthosaurus atokensis, CURRIE & CARPENTER, in press) and the basal tetanurine
Afrovenator (SERENO et al., 1994). Based on the distribution of derived character states found in this
analysis, the sequence of transformations of the
theropod wrist can be summarized as: 1) fusion of
distal carpals 1 & 2 among basal neotheropods, as
documented in some (but not all) specimens of coelophysids (COLBERT, 1989; RAATH, 1990) and all
adult tetanurine carpi except for therizinosauroids;
2) development of well developed trochlea and the
semilunate shape (255.1) but the element does not
completely cap the proximal surfaces of metacarpals I and II, among tetanurines, as demonstrated in
Afrovenator, carnosaurs, Coelurus, Scipionyx, and
tyrannosaurids; 3) expansion of the semilunate carpal block to cap the entire proximal surfaces of metacarpals (253.2), as demonstrated in maniraptoran
coelurosaurs (and separately in troodontids, if they
are arctometatarsalians); and 4) fusion of the semilunate carpal block and the metacarpals to form a
carpometacarpus (252.1), as demonstrated in metornithine birds. Unfortunately, the wrist of many
theropods is incompletely known, even for otherwise well preserved and articulated specimens such
as the types of Ceratosaurus (GILMORE, 1920) and
Compsognathus (OSTROM, 1978), so that it is uncertain exactly precisely where within the phylogeny
these transformations took place. For example, the
carpus is currently unknown or undescribed for all
non-avetheropod tetanurines other than Afrovenator, and is only partly known in any neoceratosaur
(Carnotaurus: BONAPARTE, NOVAS & CORIA, 1990).
Tyrannosaurids and ornithomimosaurs also
pose a problem with regards to the distribution of
several other manual characters shared by Maniraptora or its subclades and troodontids (as well as
other postcranial features discussed above). These
might have been present in the ancestral maniraptoriforms, and subsequently lost in the derived manus
of tyrannosaurids and ornithomimosaurs (HOLTZ,
1994); they might have been independently derived
in maniraptorans and troodontids; or they might indicate a closer relationship between troodontids and
the maniraptoran taxa here rather than with tyrannosaurids and ornithomimosaurs. However, this last
hypothesis would require that the various cranial,
pelvic, and hindlimb features shared by troodontids,
ornithomimosaurs, and/or tyrannosaurids to have
evolved more than once, or to have been lost in
Maniraptora (other than Troodontidae).
Some specimens of tyrannosaurids and all ornithomimosaurs lack well developed trochlear surfaces on any of their carpals (251.1), but the most
complete wrists known for these taxa (e.g., Tyrannosaurus (PIN 552-1), Albertosaurus (ROM 807; CMN
11315), Gorgosaurus (CMN 2120); Struthiomimus
(UCMZ(VP) 1980.1), Pelecanimimus (LH 7777))
demonstrate that they possessed a single large carpal element which did not completely cap metacarpals I and II. This most likely represents
modifications of the stage 2 of the evolution of the
theropod carpus mentioned above, and may well be
associated with the loss or modification of more typical advanced tetanurine manual function in these
two derived taxa, rather than derived from a "maniraptoran"-like condition as previously hypothesized
(HOLTZ, 1994). Indeed, the carpus of a subadult
specimen of the tyrannosaurid Albertosaurus sarcophagus (CMN 11315) quite clearly demonstrates
a semilunate carpal block comparable in relative
size and shape to those of Afrovenator, carnosaurs,
39
GAUTHIER (1986) proposed that the absence of a
fourth manual digit beyond embryonic stages
(257.3) was synapomorphic for Tetanurae. This hypothesis was supported in additional analyses (e.g.,
HOLTZ, 1994). However, as Gauthier observed, a
fragment of bone at the proximal end of metacarpal
III in Coelurus (AMNH 619; at the time considered
assignable to Ornitholestes; see also O SBORN ,
1916; OSTROM, 1969b) may be a remnant of metacarpal IV. This suggested that the fourth metacarpal
(although not necessarily the fourth digit) was present in some tetanurines. The presence of metacarpal IV in tetanurines (indeed, in avetheropods) is
supported by the discovery of a probable metacarpal IV in the sinraptorid carnosaur Sinraptor (CURRIE
& ZHAO, 1993a). The relatively flattened distal end of
this element suggests no phalanges were present
(212.2). Nevertheless, if Currie & Zhao have correctly identified this bone (a homology accepted by
the present author), the absence of this element in
allosaurid carnosaurs and maniraptoriform coelurosaurs is best explained by convergence. Unfortunately, the manus of basal carnosaurs (e.g.,
Monolophosaurus) and basal coelurosaurs (Gasosaurus, Proceratosaurus, Dryptosaurus) is unknown or incompletely known at present.
The obturator notch on the pubis (307.1) has previously been used to diagnose Avetheropoda
(HOLTZ, 1994). This structure is present in allo-
T.R. HOLTZ, JR.
saurids, Dryptosaurus, Ornitholestes, Coelurus,
Scipionyx, Bagaraatan, and all the various maniraptoriform subclades, thus including all the tetanurine
taxa employed by G A U T H I E R (1986) and all
avetheropod taxa used by HOLTZ (1994). If the phylogenetic hypothesis presented here is correct,
however, it would indicate that this opening is not a
synapomorphy of Tetanurae or Avetheropoda, but
was instead independently derived in allosaurids
and advanced coelurosaurs (SERENO et al., 1994,
1996). The basal carnosaur Monolophosaurus retains an obturator foramen (307.0) (and thus lacks
an obturator notch) (ZHAO & CURRIE, 1993), as do
the sinraptorids Yangchuanosaurus (DONG, ZHOU &
ZHANG, 1983) and Sinraptor hepingensis (GAO,
1992). It is not apparent whether the lack of contact
between bone ventral to the obturator foramen of
Sinraptor dongi (CURRIE & ZHAO, 1993a) represents
an incipient obturator notch, a pathological condition, or simply a matter of post-mortem damage. Furthermore, the pubis of the basalmost coelurosaurian
taxon in this analysis, Gasosaurus, retains the obturator foramen and thus lacks an obturator notch. The
most parsimonious explanation of the distribution of
this character state in the present analysis is that it
evolved independently in allosaurid carnosaurs and
in advanced coelurosaurs.
As mentioned above, opisthopuby (309.3) is
minimally synapomorphic for Eumaniraptora (dromaeosaurids plus birds). Curiously, although articulated dromaeosaurid remains indicate that these
forms were fully opisthopubic (NORELL & MAKOVICKY, 1997) like metornithine birds, basal avialians
such as some specimens of Archaeopteryx (WELLNHOFER, 1993), and the type and only specimens of
Unenlagia (NOVAS & PUERTA, 1997) and Rahonavis
(FORSTER et al., 1998) had pubes which were vertically oriented rather than fully retroverted (309.2).
Thus, either full retroversion occurred independently in dromaeosaurids and metornithines, or nonmetornithine avialians experienced a partial "deretroversion" from a fully opisthopubic state. A third
alternative, that dromaeosaurids share a more recent common ancestor with metornithine birds than
either does with Archaeopteryx, Rahonavis, and Unenlagia, is not supported by abundant derived character states possessed by these latter three taxa
and later birds which are not present in Dromaeosauridae.
oviraptorosaurs would represent a "deretroversion".
The obturator process of the ischium (322.1-2,
323.1-2) had been hypothesized by G AUTHIER
(1986) to be synapomorphic for Tetanurae, and by
HOLTZ (1994) to be synapomorphic for Avetheropoda. Additionally, HOLTZ (1994) found a triangular
(rather than trapezoidal) shaped obturator process
(323.2) to be synapomorphic for Coelurosauria. In
the present analysis, the distribution of these character states is revised. Although the basalmost carnosaur Monolophosaurus and the sinraptorid
Yangchuanosaurus lacks an obturator process (retaining, instead, the primitive state of an obturator
flange: CHARIG & MILNER, 1997), the sister group to
Avetheropoda in this analysis, Afrovenator, clearly
demonstrates this condition (SERENO et al., 1994).
The basalmost coelurosaur in the present study for
which pelvic material is known, Gasosaurus, has a
carnosaur- or Afrovenator-like trapezoidal obturator
process (323.1) (DONG & TANG, 1985), but all more
derived coelurosaurs have triangular structures (except for some derived birds, which lack the structure
altogether).
HOLTZ (1994) hypothesized that the loss of the ischial foot (327.1) was a synapomorphy of Coelurosauria. This was in part due to inaccurate coding of
the condition in Compsognathus: re-inspection of
this taxon (and its sister taxon, Sinosauropteryx,
coded together here as Compsognathidae) indicates that a small expansion of the distal tip of the ischium is present (327.0). A similar termination of the
ischium is present in Scipionyx (hypothesized here
as the sister taxon to Maniraptoriformes), while
Gasosaurus has a more primitive large expansion of
the distal ischiadic tip. The distal part of the ischium
is not preserved in Dryptosaurus, Proceratosaurus,
or Deltadromeus, while in Ornitholestes the ischium
does terminate in the point.
As discussed by HOLTZ (1994), the distal end of
the ischium in ornithomimids also ends in a small expansion, but as in those trees in this analysis where
troodontids are considered arctometatarsalians, it is
most parsimoniously explained here as a reversal,
as the immediate outgroups to this clade (Troodontidae and Tyrannosauridae) both possess a pointed
tip. When troodontids are considered paravian
maniraptorans, however, the ornithomimid condition is equally parsimoniously regarded as a reversal (if the tyrannosaurid condition is synapomorphic
with that in non-compsognathid maniraptorans) or a
retention of the primitive condition (if the tyrannosaurid condition is independently acquired from advanced maniraptorans). SERENO (1997) postulated
that an ischium terminating in a point was synapomorphic for a clade comprised of Tyrannosauridae
and "Maniraptora" (oviraptorosaurs, deinonycho-
In those trees where Troodontidae is considered
an arctometatarsalian, the sister clade to eumaniraptorans contains one taxon (Therizinosauroidea),
which also possesses a fully retroverted pubis. It is
thus possible (under accelerated transformation)
that opisthopuby is ancestral for the oviraptorosaurtherizinosauroid-eumaniraptoran clade. If this were
the case, the propubic condition in Microvenator and
40
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
saurs (including troodontids), and birds: in fact, by
the definition of the term, tyrannosaurids would be
within Maniraptora in his phylogeny; see PADIAN,
HUTCHINSON & HOLTZ, 1999), and that this clade
was sister taxon to an ornithomimosaurtherizinosauroid clade which retained a primitive ischiadic expansion. The condition in therizinosauroids is variable, however. RUSSELL & DONG
(1993a) considered the Early Cretaceous therizinosauroid Alxasaurus to lack a distal ischiadic point
(based on two flattened strap-like elements they
considered to be the ischia), whereas other therizinosauroids show a number of different conditions: a
terminal point in Enigmosaurus mongoliensis and a
very slight expansion in Segnosaurus galbinensis
(BARSBOLD, 1983) to a greatly expanded flange in
Nanshiungosaurus brevispinus DONG, 1979. For
this reason, this character is coded as multistate
(both present and absent) for therizinosauroids. Depending on the optimization chosen, an ischium
ending in a point is either basal for Maniraptoriformes (and thus reversed in those therizinosauroids
lacking this condition) or is derived independently in
oviraptorosaurs and paravians, in which case therizinosauroids retained the primitive distal ischiadic
expansion primitively, and those forms ending in a
tip represent yet another independent derivation of
this structure.
HOLTZ (1994) postulated that all taxa possessing
the arctometatarsalian condition (377.1) (the
pinched third metatarsal, possibly a specialization
for enhanced cursorial ability: H OLTZ , 1995b)
formed a single clade, and thus this structure was
hypothesized to have arisen only a single time in
theropod history. New data requires that this hypothesis be rejected. Abundant character evidence
supports the hypothesis that caenagnathids ("elmisaurids" in HOLTZ, 1994) are nested in a clade of
theropods lacking the arctometatarsalian condition
(see also SUES (1997) and MAKOVICKY & SUES
(1998)). Although possessing a pinched metatarsal
III, caenagnathid metatarsi differ from those found in
tyrannosaurids and bullatosaurs in that metatarsals
II and IV do not contact each other on the plantar surface (373.0) (whereas these elements do contact on
the plantar surface in ornithomimids, troodontids,
and tyrannosaurids, 373.1). Furthermore, the alvarezsaurid birds Mononykus, Shuvuuia, and Parvicursor possesses an arctometatarsalian pes even
more "pinched" than that found in Arctometatarsalia
proper: the proximal shaft of metatarsal III is entirely
lost, reducing this bone to simply the distal wedge
(377.2). However, other alvarezsaurids (such as
Patagonykus and Alvarezsaurus) lack a pinched
third metatarsal (377.0). The current phylogenetic
hypothesis requires at least three origins of this
structure: in caenagnathid oviraptorosaurs, in advanced alvarezsaurids, and in true arctometatarsali-
ans. In trees where troodontids are considered the
sister taxon to eumaniraptorans, a fourth origin of
the pinched metatarsus is required. Other metatarsal features shared by Caenagnathidae, mononykine alvarezsaurids, and Arctometatarsalia, such as
an elongated metatarsus (369.1) and metatarsals
which have a midshaft cross section deeper craniocaudally than wide mediolaterally (371.1), are also
found in the coelophysoid Elaphrosaurus and the
basal coelurosaur Coelurus.
ALTERNATIVE RELATIONSHIPS BETWEEN
THEROPOD CLADES
As has been noted in previous studies (HOLTZ,
1994; SERENO, 1997), there can be no phylogenetic
solution for theropod interrelationships that does not
result in some homoplasy. Some of this homoplasy
suggests alternative relationships to those found in
the most parsimonious trees in the present study.
Bremer support values or "decay indices" (BREMER, 1988; DONOGHUE et al., 1992), although of limited utility in comparison between data matrices due
to their dependence on data set size (SANDERSON &
DONOGHUE, 1996), were calculated for this study using the AutoDecay program (ERIKSSON, 1998). The
results are plotted on Fig. 8. A large number of nodes
have Bremer support values of 1 or 0: this is due in
part to the insecure phylogenetic position of taxa
known only from very incomplete fossils or with a
high number of plesiomorphic character states
(which can thus assume many alternative positions
on the tree without greatly affecting tree metrics).
For example, the highly incomplete Rahonavis can
assume many positions among Avialae without
greatly increasing tree length.
Nevertheless several topologies were found to
be better supported. In order of increasing support,
these are: monophyly of dromaeosaurids and avialians to the exclusion of other theropods, the inclusion
of tyrannosaurids in Arctometatarsalia, the inclusion
of compsognathids in Maniraptora, the union of the
oviraptorosaur-therizinosauroid and paravian
clades to the exclusion of all other coelurosaurs, and
the sister group position of Troodontidae to Eumaniraptora (the latter four topologies not reflected on
Fig. 8 as drawn); Abelisauridae and the
Ceratosaurus-abelisaurid clades to the exclusion of
other ceratosaurs; Coelurosauria as composed
here and the union of Microvenator, Oviraptoridae,
and Caenagnathidae to the exclusion of other theropods; the union of coelophysoids and neoceratosaurs to the exclusion of other theropods; and the
composition of Tetanurae, Ornithomimosauria, the
therizinosauroid-oviraptorosaur clade, and Avialae
as found here.
41
T.R. HOLTZ, JR.
physis. However, placing it as a basal coelophysoid
(i.e., as the sister taxon to node C) or as the sister
taxon to Coelophysidae require only one additional
step (tree length 1405). Although Elaphrosaurus
was once considered an ornithomimosaur (e.g.,
GAUTHIER, 1986; OSMÓLSKA & BARSBOLD, 1990), it
requires fifty one additional steps (tree length 1455)
to place this taxon as the sister group to Pelecanimimus+Ornithomimidae.
As noted above, Neoceratosauria shares some
features with tetanurines not found in coelophysoid
ceratosaurs. BAKKER (1986) and PAUL (1988) proposed phylogenies (which were not, it must be
noted, the result of numerical cladistic analyses) in
which these taxa shared a more recent common ancestor with birds than with Coelophysis. Moving
Neoceratosauria to a sister group position to node G
requires eleven additional steps (tree length of
1415). As a taxonomic note, the clade comprised of
the traditional "neoceratosaurs" and Tetanurae
would be properly called "Neoceratosauria", as it
would be comprised of Ceratosaurus and all taxa
sharing a more recent common ancestor with that
genus than with Coelophysis (see TABLE II). "Ceratosauria" (i.e., Ceratosaurus and all taxa sharing a
more recent common ancestor with it than with Neornithes) would be comprised solely of Ceratosaurus and Abelisauroidea, and thus Coelophysoidea
would not be considered ceratosaurs. Furthermore,
the node-defined taxon Neotheropoda (i.e., all descendants of the most recent common ancestor of
Ceratosaurus and Neornithes) would become a
subgroup of the stem-defined Neoceratosauria, and
the coelophysoids would not be considered neotheropods.
Fig. 8 - Bremer support ("decay indices") for the most
parsimonious trees found in this analysis. Only values
greater than one are shown. Calculations made using
AutoDecay (ERIKSSON, 1998), employing the summary
cladogram (Fig. 5), but shown here on the strict consensus
of all twenty most parsimonious trees (Fig. 3). Not shown
here due to polytomies on the consensus are a score of
two each for nodes V, W, cc, ee, ff, and jj.
For these calculations, the summary tree (Fig. 5)
was used, and (in most cases) only a single OTU
was moved to the new position.
NOVAS (1997b) proposed that the "carcharodontosaurs" Carcharodontosaurus and Giganotosaurus were closely related to Abelisauridae. SAMPSON
et al. (1998) also recognized several potential synapomorphies between abelisaurids and "carcharodontosaurids", without proposing a sister group
relationship between the two. As noted previously
there are abundant cranial similarities between abelisaurids and carcharodontosaurs as a whole, and
between Carcharodontosaurus and Abelisaurus in
particular. Placing Carcharodontosaurus as the sister taxon to Abelisauridae requires thirty eight additional steps (tree length 1442), and placing it as the
sister taxon to Abelisaurus requires thirty five additional steps (tree length 1439). Moving Carcharodontosaurus+Giganotosaurus to a sister group
relationship with either Abelisauridae as a whole or
Abelisaurus in particular both require forty nine
more steps than the most parsimonious tree (tree
length 1453).
Elaphrosaurus was found here to be a basal neoceratosaur, closer to Ceratosaurus than to Coelo-
SERENO et al. (1994) proposed "Torvosauroidea"
(later emended to "Spinosauroidea": OLSHEVSKY,
An additional method of comparing different tree
topologies is also employed here. Using MacClade,
the differences in tree length (i.e., the number of additional evolutionary steps necessary to explain that
topology compared to the most parsimonious found
here) were evaluated for some previously suggested phylogenies. Although this method lacks
some of the utility of more explicit search methods
(such as calculations of Bremer support indices), it
may shed some insight on the relative strength of
support for various additional tree topologies.
42
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
Allosaurus and coelurosaurs (such as the obturator
notch or the loss of metacarpal IV) are shown here to
be convergent between advanced allosaurid carnosaurs and advanced coelurosaurs. Placing Allosauridae (node Q) as the sister group to the more
advanced coelurosaurs (all those closer to birds
than to Gasosaurus) requires fourteen additional
steps (tree length 1418). Osteological data from Allosaurus alone should therefore be used with some
caution as an outgroup for coelurosaurian studies:
data from the better described of the more primitive
carnosaurs (e.g., Sinraptor and Monolophosaurus)
should also be considered in future analyses.
1995; SERENO et al. (1996, 1998)) in which the "torvosaurids" (Torvosaurus and Eustreptospondylus)
and spinosaurids formed their own clade outside of
all other theropods. CHARIG & MILNER (1997) examined several of the alleged synapomorphies of this
taxon, and found them to be wanting in Baryonyx
and other spinosaurids. Other features proposed for
this grouping were discovered, in the present analysis, to be basal tetanurine characters lost in some
advanced subclades. Restoring Spinosauroidea
with the ingroup topology as in SERENO et al. (1998)
to a sister group relationship to Piatnitzkysaurus+(Afrovenator+Avetheropoda) requires only
seven additional steps (tree length 1411). As the
anatomy of some of these taxa (such as Eustreptospondylus) becomes better known, new support for
this clade might be found.
Although the relatively unspecialized "megalosaur"-grade tetanurines were found to lie outside of
Avetheropoda in this study, it requires only five additional steps to place Afrovenator as a basal carnosaur (tree length 1409) and only six to place
Megalosaurus in that position (tree length 1410).
ELZANOWSKI & WELLNHOFER (1993) and MARet al. (1996) suggested that spinosaurid dinosaurs (or some member taxon thereof) were closely
related to bullatosaurian coelurosaurs, primarily on
cranial and dental features. Moving Spinosauridae
to a sister group position with Bullatosauria requires
thirty three additional steps (tree length 1437); for
trees where troodontids are considered paravians, it
requires thirty four additional steps to place spinosaurids as the sister taxon to Ornithomimosauria
(tree length 1438). Although the highly apomorphic
spinosaurids were found to be the basalmost branch
of Tetanurae in this study, this position is not strongly
supported. Alternative topologies including Spinosauridae within Avetheropoda as either a basal coelurosaur (sister taxon to node U) or a basal
carnosaur (sister taxon to node O) require only
seven or six extra steps (tree lengths 1412 and 1411,
respectively). Similarly, placing Spinosauridae as
the sister group to Avetheropoda requires seven additional steps (tree length 1412), and as the sister
group to the Afrovenator-avetheropod clade requires only six additional steps (tree length 1411).
As with most Cretaceous theropod lineages, spinosaurids are highly specialized and as-yet undiscovered basal members of this taxon would greatly aid
in more strongly establishing their relationship.
TILL
Because its skeletal material is so abundant, and
the osteology is well known, Allosaurus has served
as an outgroup for studies of coelurosaurian relationships (e.g., PÉREZ-MORENO et al., 1994; FORSTER et al., 1998; M AKOVICKY & S UES , 1998).
However, as discussed above, many of the characters previously used as synapomorphies between
In those studies in which Tyrannosauridae is
coded as its own OTU, it has been found to share a
more recent common ancestor with birds and other
typical coelurosaurs than with Allosaurus (Fig. 1).
However, tyrannosaurids were found to occupy several possibly different positions within Coelurosauria in these different studies. PÉREZ-MORENO et al.
(1994) and MAKOVICKY & SUES (1998) placed tyrannosaurids as the sister group to Maniraptoriformes.
This topology is eleven steps longer than the most
parsimonious trees (tree length 1415). Although no
post-Gauthier study has found tyrannosaurids to
share a more recent common ancestor with Allosaurus than with birds, placing the tyrant dinosaurs as
the sister taxon to Allosauridae requires forty six additional steps (1450), as the sister to the Acrocanthosaurus plus (Carcharodontosaurus+Giganotosaurus) clade requires forty eight additional steps
(tree length 1452), and as the sister taxon to Acrocanthosaurus requires fifty one extra steps (tree
lengths 1455).
Although troodontids have been universally regarded as coelurosaurs (Fig. 1), the position of this
clade, too, has varied from analysis to analysis. In
particular, troodontids share many derived features
with avialians, dromaeosaurids, therizinosauroids,
oviraptorosaurs, and various combinations thereof
not present in Tyrannosauridae or Ornithomimosauria. GAUTHIER (1986) considered troodontids a priori
to be the sister taxon of Dromaeosauridae within
Deinonychosauria, a position found in the studies of
SERENO (1997) and MAKOVICKY & SUES (1998).
This requires ten extra steps (tree length 1414).
FORSTER et al. (1998) found troodontids to be basal
avialians (i.e., closer to Archaeopteryx and later
birds than to Dromaeosauridae), a position requiring
only nine steps more than the most parsimonious
(tree length 1413).
Troodontids also share numerous characters
with Therizinosauroidea and Oviraptorosauria (as
mentioned in RUSSELL & DONG, 1993a). Placing
Tr o o d o n t i d a e a s t h e s i s t e r g r o u p t o t h e
therizinosauroid-oviraptorosaur clade (node gg) requires only one additional steps more than the most
43
T.R. HOLTZ, JR.
structed by similar methods. In this case, Compsognathidae was the sister group to a clade comprised
of an ornithomimosaur-therizinosauroid branch and
a branch with tyrannosaurids and oviraptorosaurs
as progressively closer sister taxa to a clade cont a i n i n g D e i n o n y c h o s a u r i a ( D r o m a e o s a u r idae+Troodontidae) and Avialae. This arrangement
is forty two steps longer (tree length 1446) than the
most parsimonious trees found in this study.
parsimonious (tree length 1405), and is thus a very
serious candidate for a potential position for this
taxon in future studies. Placing troodontids as the
sister group to node ff (the clade comprised of oviraptorosaurs, therizinosauroids, and eumaniraptorans) requires only three additional steps (tree
length 1407), and again might well be supported in
future analyses.
Therizinosauroidea has also been a problematic
coelurosaurian taxon. The study here agrees with
the conclusions of SUES (1997) and MAKOVICKY &
SUES (1998) in which oviraptorosaurs are the sister
taxon to these long necked theropods. However,
CLARK, NORELL & PERLE (1994) discussed features
shared with therizinosauroids, troodontids, and ornithomimosaurs. Moving Therizinosauroidea to a
position as the sister taxon to Bullatosauria requires
thirty one extra steps (tree length 1435). SERENO
(1997) proposed a clade comprised of Ornithomimosauria and Therizinosauroidea. Placing therizinosauroids as the sister taxon to ornithomimosaurs
in a configuration using troodontids as maniraptorans requires thirty four additional steps (tree length
1438).
HISTORICAL DISTRIBUTION OF THEROPOD
CLADES
Mapping the summary cladogram (Fig. 5) onto
the stratigraphic time scale (Fig. 9, 10) demonstrates that there remain long durations in the geologic record where, according to the present
analysis, certain taxa should be present but have yet
to be discovered or recognized. For example, no
tetanurines nor neoceratosaurs have been described from the Late Triassic, although the existence of coelophysoid ceratosaurs requires that
both those former taxa must have already diverged
from Ceratosauria and Coelophysoidea, respectively. Similarly, the presence of Allosaurus in the
Late Jurassic requires that the divergence between
the Allosaurus-Neovenator and the "carcharodontosaur" lineages must have occurred by this point in
time, despite the fact that no members of the latter
clade have been identified from units older than the
Aptian (although RAUHUT (1995) suggested that
some isolated teeth from the Upper Jurassic Tendaguru Group of Tanzania may be from this lineage).
H OLTZ (1994) considered Oviraptoridae and
Caenagnathidae ("Elmisauridae" in that study) to be
more closely related to Ornithomimus than to birds,
and thus arctometatarsalians under the revised taxonomy of H OLTZ (1996b). As S UES (1997) and
MAKOVICKY & SUES (1998) have firmly established,
a sister group relationship between Oviraptoridae
and Caenagnathidae is well supported by character
evidence, as is a therizinosauroid-oviraptorosaur
clade (node gg in this study). Moving the whole clade
Therizinosauroidea+(Microvenator+Oviraptorosauria) to a sister taxon position with node cc (Tyrannosauridae+Ornithomimosauria) requires twelve
additional steps (tree length 1416).
Of particular importance is the recent discovery
of an alleged therizinosauroid dentary from the Sinemurian age Lower Lufeng Formation of Yunnan,
China (ZHAO & XU, 1998). If confirmed, this would indicate that the divergence between therizinosauroids and oviraptorosaurs, between the
therizinosauroid-oviraptorosaur clade and the
paravian lineage, between advanced maniraptorans and Compsognathidae, between Maniraptora
and Arctometatarsalia, and all further divergences
within coelurosaurian lines would necessarily have
occurred by the Sinemurian (Fig. 10), if the summary
cladogram of the present analysis is supported. Siamotyrannus may document the presence of Tyrannosauridae in the Barremian (the probable sister
taxon to this clade, Ornithomimosauria, is already
known from that interval in the form of Pelecanimimus). The tooth taxon Koparion CHURE, 1994
may record the presence of Troodontidae in the Kimmeridgian Morrison Formation of Utah, extending
the range of this taxon to the Late Jurassic. Presence of possible maniraptoriforms in the Middle Jurassic had previously been suggested by the
discovery of isolated dromaeosaurid- and
troodontid-like teeth from the Bathonian of England
Although the taxa and characters employed in
this study and that of MAKOVICKY & SUES (1998) do
not wholly overlap, a new arrangement using the
present data matrix was employed to reconstruct a
similar topology among the advanced coelurosaurs
(for their trees, see Fig. 1K). Leaving all other taxa in
the positions of the summary cladogram, the advanced coelurosaurs were rearranged for the following topology: Compsognathidae, Tyrannosauridae, Pelecanimimus+Ornithomimidae, Ornitholestes, and Coelurus were progressively closer
outgroups to node ff, and Troodontidae was placed
as the sister group to Dromaeosauridae within Deinonychosauria. This topology is twenty eight steps
(tree length 1432) than the most parsimonious arrangements found in this analysis.
Similarly, the coelurosaurian arrangement of the
tree of SERENO (1997) (Fig. 1I) was also recon-
44
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
Fig. 9 - Summary cladogram (Fig. 5) superimposed on the geochronologic time scale I: non-maniraptoriform neotheropods. Geochronology follows GRADSTEIN et al. (1995). Solid bars, maximum known duration of suprageneric OTUs;
solid ovals, approximate geochronologic position for generic or specific OTUs; arrow indicates Maniraptoriformes (Fig.
10). Genusaurus and Sarcosaurus are probable ceratosaurs not used in this analysis; Cryolophosaurus is a possible
Early Jurassic carnosaur not used in this analysis: see text for discussion. Abbreviations: Abel., Abelisaurus; Acro., Acrocanthosaurus; Carch., Carcharodontosaurus; Carno., Carnotaurus; Cryolopho., Cryolophosaurus; Drypto., Dryptosaurus; Genu., Genusaurus; Gig., Giganotosaurus; "M.", "Megalosaurus"; Sarco., Sarcosaurus; Spinosaur., Spinosauridae.
Fig. 10 - Summary cladogram (Fig. 5) superimposed on the geochronologic time scale II: maniraptoriform coelurosaurs. Geochronology follows GRADSTEIN et al. (1995). Solid bars, maximum known duration of suprageneric OTUs;
solid ovals, approximate geochronologic position for generic or specific OTUs; gray horizontal bars, approximate geochronologic position for material referred to the OTU immediately above. Solid lines, divergence pattern sufficient to explain branching event without early problematic specimens; dotted lines, divergence pattern necessary to accommodate
problematic early occurrences referable to the OTU. Barremian tyrannosaurid, Siamotyrannus; Sinemurian therizinosauroid, ZHAO & XU (1998); Kimmeridgian troodontid, Koparion; Bathonian troodontid and dromaeosaurid teeth, EVANS
& MILNER, 1994; METCALF & WALKER, 1994. Abbreviations: Caen., Caenagnathidae; Theriz., Therizinosauroidea.
45
T.R. HOLTZ, JR.
sult in any phylogenetic position requires some degree of homoplasy. Identification of Middle Jurassic
carnosaurs and coelurosaurs indicate that the divergence between these lineages was earlier than previously suggested, and renewed search for pre-Late
Jurassic theropod fossils (both in the field and in collections) may prove among the most fruitful discoveries for resolving the phylogeny of the carnivorous
dinosaurs.
ACKNOWLEDGMENTS
Many thanks are offered to the numerous people
who have aided me in my on-going research in
theropod systematics. This text was greatly improved by the comments and corrections by the numerous individuals, both official referees and other
interested parties, who read and reviewed this
manuscript: in particular, I wish to thank Ralph Molnar (Queensland Museum), Phil Currie (Royal Tyrrell Museum of Palaeontology), Hans-Dieter Sues
(Royal Ontario Museum), Scott Sampson (State
University of New York - Stony Brook), and Bernardino Pérez-Moreno (Universidád Autónoma,
Madrid) for their helpful comments. The author is
solely responsible for any errors (and original concepts) presented herein.
(E VANS & M ILNER , 1994; M ETCALF & W ALKER ,
1994).
The data from this new analysis indicates that the
chronological conclusions of HOLTZ (1994), among
others, were premature, and that the primary divergences within Avetheropoda and Coelurosauria
may have been considerably earlier than the Late
Jurassic. Instead, presence of Monolophosaurus,
Proceratosaurus, and Gasosaurus indicate a minimal Middle Jurassic divergence of Carnosauria and
Coelurosauria. Furthermore, Early Jurassic taxa
possibly referable to these clades (Cryolophosaurus and the Yunnan "therizinosauroid", respectively)
not included in the present analysis may indicate
that the primary avetheropod divergence occurred
tens of millions of years early than this study indicates. As such, this predicts that a greater diversity
of theropods (including early members of more typical Late Jurassic and Cretaceous lineages) should
be present in Lower and Middle Jurassic formations
than have currently been recognized.
ABBREVIATIONS
AMNH - American Museum of Natural History,
New York City; BMNH - Natural History Museum,
London; CMN - Canadian Museum of Nature, Ottawa, Alberta; LH - Las Hoyas collection, Museo de
Cuenca, Cuenca (housed in the Unidad de Paleontología, Universidad Autónoma de Madrid); MCZ Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts; PIN - Palaeontological Institute, Moscow; ROM - Royal Ontario
Museum, Toronto, Ontario; RTMP - Royal Tyrrell
Museum of Palaeontology, Drumheller, Alberta;
UCMZ(VP) - Museum of Zoology, University of Calgary, Calgary, Alberta.
APPENDIX I
Morphological characters used in the present study. Multistate characters are considered unordered unless otherwise noted.
Character polarity based on outgroup comparison: see text for explanation. Scoring: 0 = primitive state; 1, 2, 3, 4, or 5 = derived
character states.
1. Skull shape: 0, oreinirostral; 1, elongate and platyrostral,
with obtuse triangular paracoronal cross-section; 2, shortened
and platyrostral, with acute triangular paracoronal cross-section.
CONCLUSION
The present study supports many of the previously discovered phylogenetic relationships among
neotheropod dinosaurs. Due to the fragmentary
and/or plesiomorphic nature of non-avetheropod
tetanurines and non-maniraptoriform coelurosaur
fossil material, however, these regions of the phylogenetic tree remain poorly resolved. Even for taxa
for whom the anatomy is relatively well-known, such
as Troodontidae, the mosaic of derived features re-
UO
2. Premaxillary teeth: 0, present; 1, absent, presumably covered with a rhamphotheca. (RUSSELL & DONG, 1993a)
3. Number of premaxillary teeth: 0, four; 1, three; 2, five; 3, seven; 4, six. UO
4. Premaxillary symphyseal region: 0, V-shaped in ventral
view; 1, U-shaped in ventral view; 2, Y-shaped in ventral view. UO
(HOLTZ, 1994)
46
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
27. Paired crescentic crests formed by nasal and lacrimal prominences: 0, absent; 1, present. (ROWE, 1989)
5. Premaxilla subnarial depth: 0, shallow to moderately deep,
main body as long or longer rostrocaudally than high dorsoventrally; 1, very deep, main body taller dorsoventrally than long rostrocaudally. (HOLTZ, 1994).
28. Nasal fusion:
sals fused together.
6. Premaxilla shape: 0, short and rounded, with short nasal
process; 1, long and pointed, with long nasal process. (GAUTHIER,
1986)
29. Nasal recesses: 0, absent; 1, present. (WITMER, 1997)
30. Lacrimal: 0, not exposed on skull roof; 1, broadly exposed
on skull roof. (GAUTHIER, 1986)
7. Medial alae from premaxilla: 0, absent or separate; 1, meet
in front of vomers. (RUSSELL & DONG, 1993a)
31. Lacrimal prominence: 0, absent; 1, triangular hornlets;
ridge continuous with raised surface of lateral edge of nasals.
8. Maxillary process of premaxilla: 0, moderately long, premaxilla participates broadly in ventral surface of external naris; 1, re-
33. Lacrimal recess: 0, absent; 1, single opening present;
multiple openings present. O (WITMER, 1997)
9. External nares: 0, without marked inset of the caudal margin; 1, with marked inset of the caudal margin. (SERENO et al.,
1994)
34. Slot in ventral process of lacrimal for jugal:
present. (SERENO et al., 1996)
Premaxilla/nasal contact: 0, premaxilla and nasal meet
subnarially; 1, premaxilla and nasal do not meet subnarially.
36. Lacrimal suborbital bar:
CARPENTER, in press)
38. Prefrontal-frontal peg-in-socket suture:
sent. (SERENO et al., 1994)
14. Rostral ramus size: 0, absent; 1, shorter rostrocaudally
than dorsoventrally; 2, as long or longer rostrocaudally as dorsoventrally. UO (SERENO et al., 1996)
40. Frontal shape: 0, narrow or truncated rostrally, postorbital
ramus projects laterally from orbital margin of frontal; 1, very
broadly exposed on skull roof, postorbital ramus does not project
abruptly laterally from the orbital margin. (HOLTZ, 1994)
41. Frontal-frontal suture: 0, unfused; 1, fused.
42. Frontal-parietal suture on dorsal surface of skull: 0, forms
a relatively straight line transversely; 1, frontals separated at medialmost point of suture by rostral process of parietals; 2, frontals
and parietals fused, suture indistinguishable. UO
43. Dorsal surface of parietals: 0, flat, ridge borders supratemporal fenestra; 1, sagittal crest along midline.
44. Orbit length: 0, subequal to or longer than internal antorbital fenestra length; 1, shorter than internal antorbital fenestra
16. Promaxillary fenestra: 0, absent; 1, present, visible in lateral view; 2, present, obscured in lateral view by ascending ramus
of maxilla; 3, small depression in same anatomical position which
does not perforate the maxilla. UO (HOLTZ, 1994)
1,
present. (GAUTHIER,
18. Maxillary fenestra shape: 0, round; 1, long and low. (WI-
TMER,
1997)
19.
Position of the maxillary and promaxillary fenestrae: 0,
promaxillary rostral to maxillary; 1, promaxillary dorsal to maxillary. (WITMER, 1997)
length.
45. Orbit shape: 0, round; 1, oval or key-shaped, rounded dorsally, constricted ventrally. (GAUTHIER, 1986)
20. Relative size of promaxillary and maxillary fenestrae: 0,
maxillary fenestra larger; 1, promaxillary fenestra larger. (WI-
TMER,
0, absent; 1, pre-
39. Rostral portion of frontals: 0, relatively square, suture with
nasals forms a relatively obtuse angle or W; 1, triangular, suture
with nasals form a distinctly acute angle. (HOLTZ, 1994)
15. Maxillary antorbital fossa: 0, approximately 10-25% of the
rostrocaudal length of the antorbital cavity; 1, greater than 40% of
the rostrocaudal length of the antorbital cavity; 2, greatly reduced
in size, not extending much beyond the rim of the external antorbital fenestra. (SERENO et al, 1994)
absent;
0, absent; 1, present. (CURRIE &
37. Prefrontals: 0, well exposed on skull roof; 1, reduced or absent. (GAUTHIER, 1986)
maxilla forms convex surface from dorsal ramus to ventral margin;
1, present, dramatic change in curvature of rostrodorsal surface of
maxilla rostral to dorsal ramus forming concave surface
0,
0, absent; 1,
1996)
11. Subnarial gap: 0, absent; 1, present. (GAUTHIER, 1986)
12. Maxillary teeth: 0, present; 1, absent.
13. Rostral ramus of maxilla: 0, absent, rostrodorsal surface of
Maxillary fenestra:
2,
35. Lacrimal dorsal (= rostral) ramus: 0, dorsoventrally thick;
1, dorsoventrally pinched and narrow; 2, absent. O (SERENO et al.,
10.
17.
2,
32. Lacrimal caudal process at dorsal surface: 0, absent, lacrimal L-shaped or simple shaft; 1, present, lacrimal T-shaped.
(MAKOVICKY & SUES, 1998)
duced, maxilla participates broadly in ventral surface of external
naris; 2, extremely long, extends caudally from the caudal margin
of the external naris for a distance greater than the rostrocaudal
length of the external naris. UO
1986)
0, absent, nasals separate; 1, present, na-
46. Orbit margin: 0, smooth; 1, raised rim. (HOLTZ, 1994)
47. Postorbital prominences: 0, absent; 1, present. (RUSSELL
1997)
xilla:
21. Pneumatic excavation of the ascending ramus of the ma0, absent; 1, present. (SERENO et al., 1996)
22. Pneumatic excavation without fenestra in cranial portion
of maxillary antorbital fossa: 0, absent; 1, present. (SERENO et al.,
48. Postorbital ventral process: 0, broader rostrocaudally than
transversely; 1, broader transversely than rostrocaudally with Ushaped cross-section. (SERENO et al., 1996)
1998)
23. Nasal antorbital fossa: 0, lateral surface of nasal excluded
from antorbital cavity; 1, lateral surface of nasal participates in an-
49. Postorbital ventral process ventralmost extension: 0, dorsal to ventral margin of orbit; 1, ventral to ventral margin of orbit.
(CURRIE & CARPENTER, in press)
& DONG, 1993a)
torbital cavity, forming a nasal antorbital fossa. (WITMER, 1997)
50. Postorbital-lacrimal contact:
tact lacrimal; 1, broad.
24. Nasal participation in antorbital cavity: 0, nasal excluded
from antorbital cavity; 1, nasal participates in antorbital cavity.
25. Nasal expansion behind external nares: 0, broadly expanded; 1, narrow caudally. (HOLTZ, 1994)
26. Narial prominences: 0, absent; 1, median horn or crest; 2,
paired ridges along lateral edges of nasals; 3, knobby rugosities
0, postorbital does not con-
51. Postorbital-jugal contact: 0, present; 1, absent. (CHIAPPE,
NORELL & CLARK, 1998)
52. Postorbital bulbous rostrally projecting rugosity: 0, absent;
1, present. (SERENO et al., 1996)
53. Postorbital suborbital flange: 0, absent; 1, present.
54. Postorbital frontal process: 0, sharply upturned; 1, about
across dorsal and lateral surface of nasals, extending onto dorsalmost surface of maxillae; 4, "bark-like" rugosities, including concave pits separated by crests. UO (CURRIE & CARPENTER, in
press)
same level or slightly higher than squamosal process, producing
T-shaped postorbital. (CURRIE, 1995)
55. Squamosal recess: 0, absent; 1, present. (WITMER, 1997)
47
T.R. HOLTZ, JR.
82. Endocranial cavity: 0, typical size of other bipedal dinosaurs; 1, enlarged relative to other dinosaurs, but temporal musculature extends origin onto frontals; 2, greatly enlarged, temporal
musculature fails to extend origin onto frontals. O (GAUTHIER,
1986)
83. Nuchal crest: 0, small or absent; 1, pronounced. (HOLTZ,
1994)
84. Supraoccipital with very pronounced, strongly demarcated median ridge: 0, absent; 1, present.
85. Orbitosphenoid: 0, present; 1, absent. (HOLTZ, 1994)
86. Paroccipital process pneumatization: 0, solid proximal
portion; 1, hollow proximal portion.
87. Paroccipital process orientation: 0, occipital surface of distal end oriented more caudally than dorsally; 1, conspicuous twist
in the distal end orients distal surface more dorsally than proximal
region; 2, curving ventrally and pendant. UO (CURRIE, 1995)
88. Basicranium pneumatization: 0, minimal to moderate, but
no expansion of basisphenoid; 1, basisphenoid, but not parasphenoid rostrum, strongly expanded and pneumatized. (RUSSELL &
DONG, 1993a)
89. Lateral depression surrounding opening to middle ear: 0,
absent; 1, present. (MAKOVICKY & SUES, 1998)
90. Number of tympanic recesses: 0, two or fewer; 1, three.
(MAKOVICKY & SUES, 1998)
91. Branches of internal carotid artery entering hypoglosseal
fossa: 0, enter separately; 1, enter through single common foramen. (MAKOVICKY & SUES, 1998)
92. Posttympanic recess: 0, invades paroccipital process; 1,
confined to columnar process. (MAKOVICKY & NORELL, 1998)
93. Cranial tympanic recess: 0, excluded from basisphenoid;
1, invades basisphenoid. (MAKOVICKY & NORELL, 1998)
94. Internal foramen of facial nerve: 0, ventral to vestiblocochlear nerve; 1, cranioventral to vestiblocochlear nerve. (MAKOVICKY & NORELL, 1998)
95. Number of cranial nerve openings in acoustic fossa: 0,
two; 1, three. (MAKOVICKY & NORELL, 1998)
96. Basioccipital: 0, not excluded from basal tuber; 1, excluded from basal tuber. (SERENO et al., 1996)
97. Distance across basal tubera: 0, greater than transverse
width of occipital condyle; 1, less than the transverse width of occipital condyle.
98. Parabasisphenoid bulbous capsule: 0, absent; 1, present.
(HOLTZ, 1994)
99. Exoccipital-opisthonic caudoventral limit of contact with
basisphenoid separated from basal tubera by notch: 0, absent; 1,
present. (CURRIE & CARPENTER, in press)
100. Basipterygoid processes: 0, moderately long, not fused
to pterygoids; 1, very short, not fused to pterygoids; 2, very short,
fused to pterygoids. UO (RUSSELL & DONG, 1993a)
101. Occipital region: 0, directed caudally; 1, directed ventrocaudally. (HOLTZ, 1994)
102. Foramen magnum dimensions: 0, subcircular or wider
than tall; 1, taller than wide. (MAKOVICKY & SUES, 1998)
103. Occipital condyle constricted neck: 0, absent; 1, present.
(MAKOVICKY & SUES, 1998)
104. Dentary teeth: 0, present; 1, absent.
105. Dentary end: 0, rounded; 1, squared with expanded tip.
(SERENO et al., 1996)
106. Dentary symphysis: 0, dentaries separate; 1, dentaries
fused.
107. Symphyseal region of dentary: 0, straight; 1, medially recurved. (CLARK, PERLE & NORELL, 1994)
108. Rostral half of mandible: 0, ventrally convex or straight; 1,
concave. (RUSSELL & DONG, 1993a)
56. Squamosal flange covering quadrate head in lateral view;
0, absent; 1, present. (SERENO et al., 1994)
57. Squamosal constriction of lateral temporal fenestra: 0, absent; 1, present. (CURRIE & CARPENTER, in press)
58. Jugal: 0, does not participate in margin of internal antorbital fenestra; 1, participates in internal antorbital fenestra. (HOLTZ,
1994)
59. Jugal postorbital process: 0, present; 1, absent. (CURRIE &
CARPENTER, in press)
60. Jugal quadratojugal processes caudalmost extensions: 0,
dorsal and ventral processes subequal in caudalmost extension;
1, dorsal extends further caudally; 2, ventral extends further caudally. UO. (CURRIE & CARPENTER, in press)
61. Jugal recesses: 0, absent; 1, present. (WITMER, 1997)
62. Infratemporal fenestra: 0, subequal or less in area of orbit
in lateral view; 1, about twice as large as the area of the orbit in lateral view. (HOLTZ, 1994)
63. Quadratojugal: 0, L-shaped; 1, T-shaped. (CURRIE, 1995)
64. Quadratojugal-squamosal contact: 0, tip of dorsal ramus
of quadratojugal contacts tip of lateroventral ramus of squamosal;
1, dorsal ramus of quadratojugal does not contact squamosal; 2,
broad contact between dorsal ramus of quadratojugal and lateroventral ramus of squamosal. UO
65. Articulations of quadrate and squamosal: 0, quadrate articulates only with squamosal and the latter bone contacting both
the quadratojugal and postorbital; 1, quadrate articulates with
both prootic and squamosal, and the latter contacting neither the
quadratojugal nor the postorbital. (CHIAPPE, NORELL & CLARK,
1996)
66. Quadrate-quadratojugal suture: 0, unfused; 1, fused.
67. Quadrate foramen: 0, large and situated between quadrate and quadratojugal; 1, reduced or absent; 2, small and enclosed
within dorsal ramus of quadrate. UO
68. Quadrate dorsal ramus: 0, less than height of orbit; 1, gre-
ater than height of orbit.
69. Depth of quadrate articulation: 0, level with ventral surface
of maxilla in lateral view; 1, projects well ventral of the ventral surface of the maxilla. (HOLTZ, 1994)
70. Length of quadrate articulation: 0, only slightly caudal to
the caudal point of occipital condyle in dorsal view; 1, projects well
caudal to the caudal point of the occipital condyle; 2, rostral to caudal point of occipital condyle. UO
71. Quadrate pneumaticity: 0, absent or poorly developed; 1
well developed.
72. Secondary palate: 0, primarily soft; 1, well ossified from
premaxilla through one-half the length of the ventral surface of the
maxilla.
73. Vomera: 0, separate rostrally; 1, fused rostrally.
(GAUTHIER, 1986)
74. Vomer length: 0, limited to rostral region; 1, extends caudally to basicranium. (RUSSELL & DONG, 1993a)
75. Palatines meet medially: 0, absent, separated by vomera
and/or pterygoids; 1, present. (HARRIS, 1998)
76. Palatine shape: 0, subrectangular or trapezoidal; 1, tetraradiate; 2, triradiate (no jugal processes). UO (HARRIS, 1998)
77. Jugal process of palatine expanded distally: 0, absent; 1,
present. (HARRIS, 1998).
78. Palatine recesses: 0, absent; 1, present. (WITMER, 1997)
79. Palatine fenestra (between ectopterygoid and palatine): 0,
open; 1, closed. (RUSSELL & DONG, 1993a)
80. Subsidiary fenestra between pterygoid and palatine: 0,
absent; 1, present. (GAUTHIER, 1986)
81. Ventral ectopterygoid recess: 0, absent; 1, present and
comma-shaped; 2, present and subcircular. UO (SERENO et al.,
1996)
48
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
109. Dentary rami: 0, subparallel;
dally. (MAKOVICKY & SUES, 1998)
1,
widely divergent cau-
or less times as wide as tall) and small odontoid notch. (GAUTHIER,
1986)
110. Reduced overlap of dentary onto postdentary bones: 0,
1, present. (GAUTHIER, 1986)
111. Intramandibular joint: 0, absent; 1, present. (SERENO &
138. Second intercentrum cranial articulation with first intercentrum: 0, slight concavity; 1, broad crescentic fossa.
(GAUTHIER, 1986)
absent;
139. Axial "spine table" (expanded distal end of neural spine):
0, absent; 1, present. (GAUTHIER, 1986)
140. Axial neural spine shape: 0, flared transversely; 1, com-
NOVAS, 1992)
112. Dentary caudal depth: 0, subequal to 120% depth of dentary symphysis; 1, 150-200% depth of dentary symphysis; 2, grea-
pressed mediolaterally. (MAKOVICKY & SUES, 1998)
ter than 220% depth of dentary symphysis.
141. Craniodorsal rim of axial neural spine: 0, concave in lateral view; 1, convex curve in lateral view. (MAKOVICKY & SUES,
1998)
113. Rostral surangular foramen: 0, absent or very small pit; 1,
larger, in rostrally-oriented depression.
114. Caudal surangular foramen:
ning. (HOLTZ, 1994)
115.
0,
Rostral ramus of surangular:
(GAUTHIER, 1986)
small pit;
0,
1,
shallow;
large ope-
1,
142. Axial parapophyses: 0, prominent; 1, reduced. (ROWE,
1989)
deep.
143. Axial diapophyses: 0, present; 1, absent. (ROWE, 1989)
144. Axial epipophyses: 0, moderately developed; 1, promi-
116. Horizontal shelf on lateral surface of surangular, rostral
and ventral to the mandibular condyle: 0, absent or faint ridge; 1,
prominent and extends laterally; 2, prominent and pendant. UO
117. Rostral prong of angular: 0, does not penetrate the dentary-splenial cavity; 1, penetrates the dentary-splenial cavity.
nent. (GAUTHIER, 1986)
145. Axial pleurocoels: 0, present; 1, absent. (ROWE, 1989)
146. Ventral keel on axial centrum: 0, present; 1, absent.
(MAKOVICKY & SUES, 1998)
(BAKKER, WILLIAMS & CURRIE, 1988)
147. Cervical centra surfaces: 0, amphiplatyan or mildly
opisthocoelous; 1, markedly opisthocoelous.
118. External mandibular fenestra: 0, large, horizontally oval;
1, reduced. (GAUTHIER, 1986)
119. Splenial: 0, obscured or only slightly visible in lateral
view; 1, extensive triangular exposure in lateral view between
148. Postaxial cervical pleurocoels: 0, absent; 1, one pair pre2, two pairs present. O
149. Cervical epipophyseal shape: 0, rugosity on caudal
zygapophyses; 1, powerfully developed and prong shaped. (RUS-
sent;
dentary and angular. (CURRIE, 1995)
120. Splenial with notch on rostral margin of internal mandibular fenestra: 0, absent ; 1, present. (SERENO et al., 1996)
121. Coronoid: 0, present;
(RUSSELL & DONG, 1993a)
1,
SELL
& DONG, 1993a)
150. Cervical epipophyseal height and orientation: 0, directed
caudolaterally and shorter than neural spine; 1, directed dorsola-
extremely reduced or absent.
terally and taller than neural spine. (HOLTZ, 1994)
122. Articular facet for mandibular joint: 0, deeply concave; 1,
151. Epipophyses on cervical vertebrae: 0, placed distally on
postzygapophyses; 1, placed proximally. (MAKOVICKY & SUES,
1998)
craniocaudally elongate and shallow. (MAKOVICKY & SUES, 1998)
123. Retroarticular process of articular: 0, faces dorsocau1, faces caudally. (SERENO et al., 1994)
124. Retroarticular process shape: 0, short and deep; 1, elon-
dally;
152. Caudal cervical epipophyses size: 0, short; 1, elongate.
153. Cervical prezygapophyses: 0, planar; 1, flexed.
gated and tapering. (CURRIE, GODFREY & NESSOV, 1993)
(GAUTHIER, 1986)
125. Vertical columnar process on retroarticular process: 0,
absent; 1, present. (CURRIE, 1995)
126. Number of teeth: 0, less than 100; 1, greater than 100.
(PÉREZ-MORENO et al., 1994)
127. Dentary and maxillary teeth: 0, subequal in number and
size; 1, dentary teeth more numerous and smaller than maxillary
154. Cervical neural spines: 0, tall;
short. (RUSSELL & DONG, 1993a)
1, low and craniocaudally
155. Cervical zygapophyses direction: 0, overhang centrum
parasagittally; 1, displaced laterally away from centrum in dorsal
view. (MAKOVICKY & SUES, 1998)
156. Caudal cervical neural arch forms X-shape in dorsal
0, absent; 1, present. (MAKOVICKY & SUES, 1998)
157. Cranial cervicals: 0, subcircular in cranial view; 1, broa-
teeth. (RUSSELL & DONG, 1993a)
view:
128. Serrations: 0, small serrations; 1, large denticles; 2, abUO
129. Relative serration (or denticle) size of anterior and posterior carinae of maxillary and dentary teeth: 0, subequal; 1, posteri-
sent.
der than deep on cranial surface, with reniform (kidney-shaped)
articular surfaces that are taller laterally than at midline.
(GAUTHIER, 1986)
or serrations much larger than anterior serrations.
158. Cranial cervical centra caudal extent: 0, level with or
shorter than caudal extent of neural arch; 1, extend beyond caudal
extent of neural arch. (MAKOVICKY & SUES, 1998)
130. Tooth roots: 0, unconstricted; 1, constricted.
131. Lateral surface of teeth: 0, smooth; 1, with wrinkles
in
enamel internal to serrations. (CURRIE & CARPENTER, in press)
159. Elevation of cranial face of midcervical centra: 0, present;
1, absent. (SERENO et al., 1996)
160. Midcervical centra length: 0, around twice diameter of
cranial face; 1, four times or more diameter of cranial face; 2, less
than twice diameter of cranial face. UO
161. Midcervical centra breadth: 0, less than 20% broader
than tall; 1, greater than 20% broader than tall. (SERENO et al.,
132. Premaxillary tooth crowns: 0, conical; 1, asymmetrical
(strongly convex labially, relatively flattened lingually); 2, incisiform and reduced in size. UO
133. Caudalmost maxillary tooth position: 0, beneath midpoint
1, rostral to orbit. (GAUTHIER, 1986)
134. Dentary tooth implantation: 0, set in sockets; 1, set in pa-
of orbit;
1996)
radental groove. (CHIAPPE, NORELL & CLARK, 1996)
162. Carotid process on caudal cervical vertebrae: 0, absent;
1, present. (MAKOVICKY & SUES, 1998)
163. Caudal cervical postzygapophyses: 0, short; 1, elongate.
135. Interdental plates: 0, present and separate; 1, fused to2, absent in dentary. UO
136. Neck length: 0, less than twice the length of skull; 1, twice
gether;
(MAKOVICKY & SUES, 1998)
or more the length of skull.
164.
137. First intercentrum: 0, small occipital fossa (three times as
wide as tall) and large odontoid notch; 1, large occipital fossa (two
1994)
49
Longest postaxial cervicals:
0,
III-V;
1,
VI-IX. (NOVAS,
T.R. HOLTZ, JR.
165. Cervical ribs: 0, unfused to centra in adults;
centra in adults. (GAUTHIER, 1986)
191. Caudal pleurocoels:
present in neural arch. UO
1, fused to
192. Caudal neural spines: 0, present beyond caudal X;
mited to caudals I-IX. (GAUTHIER, 1986)
166. Ventral processes (hypapophyses) on cervicodorsal vertebrae: 0, absent; 1, present as small protrusion; 2, very well developed. O (GAUTHIER, 1986)
193. Ventral groove in cranial caudals:
(ROWE & GAUTHIER, 1990)
167. Neural spines of dorsals: 0, less than or equal to centrum
height; 1, equal to twice centrum height; 2, more than twice centrum height. O
168. Apices of dorsal neural spines: 0, unexpanded; 1, expan-
170. Dorsal transverse processes: 0, laterally directed and
subrectangular in dorsal view; 1, strongly backturned caudally and
triangular in dorsal view. (HOLTZ, 1994)
tremely long prezygapophyses, extend more than one centrum
length. O
171. Dorsal transverse processes direction: 0, long and caudodorsally inclined; 1, short, wide and only slightly inclined.
199. Distal caudal vertebrae: 0, only slight interlocking,
prezygapophyses extend less than one half centrum length; 1,
moderately interlocking, prezygapophyses extend more than one
half, but less than one, centrum length; 2, strong interlocking, nonossified structures (cartilaginous extensions of prezygapophyses?) producing rigidity in caudals VII and distal; 3, extremely strong interlocking, bony extensions of prezygapophyses
extending up to 12 centrum lengths; 4, pygostyle. UO
(MAKOVICKY & SUES, 1998)
172. Caudal edge of dorsal postzygapophyses: 0, level with
caudal intracentral articulation; 1, overhangs centrum. (MAKOVI& SUES, 1998)
173.
Vertebral foramen/cranial articular facet ratio (vertical
diameters) of dorsals: 0, around 0.1-0.3 ; 1, 0.4 or greater.
(CHIAPPE, NORELL & CLARK, 1996)
200. Distal caudal length: 0, as long as proximal caudals; 1,
more than 130% length of proximal caudals; 2, markedly shorter
than proximal caudals. UO
174. Dorsal hyposphene-hypantrum accessory articulations:
0, present; 1, absent. (CHIAPPE, NORELL & CLARK, 1996)
175. Dorsal centrum shape: 0, cylindrical, central section thickness greater than 60% height of cranial face; 1, "hourglass"
201. Shaft of cervical ribs: 0, moderately long (two to three times centrum length) and slender; 1, extremely long (four or more
times centrum length) and slender; 2, short (less than twice centrum length) and broad; 3, short (less than twice centrum length)
and slender. UO
shaped, central section thickness less than 60% height of cranial
face. (HOLTZ, 1994)
176. Dorsal centrum transverse section: 0, subcircular or oval;
1, wider than high. (MAKOVICKY & SUES, 1998)
177. Dorsal centrum ends: 0, amphiplatyan; 1, biconvex.
202. Uncinate processes: 0, absent or unossified; 1, ossified.
203. Medial gastral segment: 0, longer than lateral segment;
(CHIAPPE, NORELL & CLARK, 1996)
1, shorter than lateral segment (NORELL & MAKOVICKY, 1997)
204. Chevron transition: 0, beyond caudal XVII; 1, between
0, much longer than femur length; 1, su-
caudal X and XVII. (GAUTHIER, 1986)
179. Caudal dorsal neural spines: 0, oriented vertically or caudally; 1, oriented cranially. (HARRIS, 1998)
180. Cranial and median dorsal pleurocoels: 0, absent; 1, one
pair present; 2, two pairs present. O
181. Presacral pleurocoel structure: 0, absent; 1, camerate; 2,
camellate. O
182. Capitular facet of dorsal ribs: 0, lies on cranioventral lamina from transverse process; 1, situated dorsal to lamina, on
205. Paired caudal and cranial chevron bases:
present. (SERENO
., 1994)
et al
dorsoventrally depressed. (GAUTHIER, 1986)
208. Middle chevron shape: 0, gentle curvature; 1, dramatic
bend in distal portion ("L-shaped"). (SERENO
., 1994)
et al
209. Distal chevrons with cranial and caudal projections, and
183. Sacral pleurocoels: 0, absent; 1, present.
184. First sacral: 0, amphiplatyan; 1, procoelous. (CHIAPPE,
more than twice as long craniocaudally as tall dorsoventrally
("boat-shaped"): 0, absent; 1, present. (HOLTZ, 1994)
NORELL & CLARK, 1996)
210. Distal chevron cranial and caudal bifurcations: 0, absent;
1, present. (SERENO, 1997)
211. Scapular blade: 0, short and broad; 1, long, slender (four
185. Number of sacrals (as determined by number of vertebrae which attach to pelvic girdle): 0, two; 1, three; 2, four; 3, five;
4, six; 5, more than six. O
times or more longer then midshaft width) and strap-like.
(GAUTHIER, 1986)
1, trans-
212. Distal expansion of scapula: 0, broad (subequal in width
to proximal end of scapula); 1, reduced or absent. (CURRIE &
ZHAO, 1993a)
187. Caudalmost sacral centrum size: 0, subequal in width
with cranialmost sacral centrum; 1, markedly smaller than cranialmost sacral centrum. (MAKOVICKY & SUES, 1998)
213.
188. Sacral neural spines fuse to form lamina: 0, absent; 1,
sent.
present.
Acromion on scapula:
0,
prominent;
1,
reduced or ab-
214. Caudal margin of acromial process of scapula: 0, gentle
1, abrupt change, perpendicular to blade.
215. Scapulacoracoid cranial margin: 0, smooth; 1, pronoun-
189. Synsacrum (fusion of sacral centra, neural arches, neural spines, transverse processes, and sacral ribs to ilia): 0, absent;
1, present in adults.
25.
0, absent; 1,
206. Bridge of bone dorsal to haemal canal in distal chevrons:
0, absent; 1, present.
207. Proximal chevron shape: 0, dorsoventrally elongate; 1,
prezygapophyseal base. (RUSSELL & DONG, 1993a)
186. Sacrals III-V: 0, moderately or uncompressed;
versely compressed; 2, dorsoventrally flattened. UO
1, box-like with
195. Proximal caudal zygapophyses: 0, short; 1, elongate.
196. Caudal transverse processes: 0, present beyond caudal
XV; 1, only on caudals I-XV or fewer. (HOLTZ, 1994)
197. Transition point: 0, absent; 1, in distal half of tail; 2, in proximal half of tail; 3, in caudals I-IX. O (GAUTHIER, 1986)
198. Midcaudal vertebrae: 0, short prezygapophyses, extend
less than one half centrum length; 1, moderate prezygapophyses,
extend more than one half but less than one centrum length; 2, ex-
169. Scars for interspinous ligaments: 0, terminate at apex of
neural spine in dorsal vertebrae; 1, terminate below apex of neural
spine. (MAKOVICKY & SUES, 1998)
178. Dorsal column:
bequal to femur length.
1, li-
0, absent; 1, present.
194. Centra of caudals I-V: 0, spool-shaped;
increased flexural capability. (GAUTHIER, 1986)
ded transversely to form "spine table" . (MAKOVICKY & SUES,
1998)
CKY
0, absent; 1, present in centrum; 2,
slope;
ced notch between acromial process and coracoid. (CURRIE &
CARPENTER, in press)
190. Number of caudals: 0, 45 or more; 1, 30-44; 2, less than
O (HOLTZ, 1994)
50
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
244. Ulnar facet on humerus: 0, small or absent; 1, expanded,
merges with entepicondyle. (R
& D , 1993a)
245. Ulnar shaft: 0, straight; 1, bowed caudally. (G
,
1986)
246. Shape of proximal ulnar shaft: 0, straight; 1, arched.
247. Diameterof ulnar shaft: 0, equal to or slightlythickerthan
that or radius; 1, much thicker than that of radius.
248. Ulnar distal condyle: 0, transversely compressed and
craniocaudallyextendedapproximatelyat same plane of humeroulnar flexion-extensionmovement; 1, subtriangularshaped in distalview,withadorsomedialcondyle,andtwistedmorethan54degrees with respect to the proximal end. (N , 1996)
249.Ulnarfacetforradius: 0,smallandflat; 1,transverselyexpanded and concave. (M
& S , 1998)
250. Ulnar and radial distal ends: 0, loosely joined; 1, closely
joined, even with syndesmosis. (P -M
1994)
251. Distal carpal shape: 0, cubic and well formed, with obvious articular surfaces; 1, flat and discoidal, no distinct articular
surfaces. (H , 1994)
252. Carpometacarpus: 0, absent, carpals distinct units; 1,
present, carpals fused to each other and to metacarpus.
253. Distal carpal I block: 0, only overlaps base of metacarpal
I; 1, does not overlap metacarpal II dorsally (does so ventrally); 2,
broadly overlaps metacarpal II dorsally and ventrally. O
254. Distal carpal I: 0, unfused to distal carpal II; 1, fused to
distal carpal II.
255. Semilunate carpal block fully developed with transverse
trochlea: 0, absent; 1, present. (G
, 1986)
256. Metacarpal V: 0, present, with digit; 1, present, without
ungual; 2, present, without phalanges; 3, absent. O
257. Metacarpal IV: 0, present, with digit; 1, present, without
ungual; 2, present, without phalanges; 3, absent. O
258. Metacarpal III: 0, present, with digit; 1, present, without
ungual; 2, present, without phalanges; 3, absent. O
259. Metacarpal II: 0, present, with digit; 1, absent.
260. Metacarpal I size: 0, greater than one half of metacarpal
II length, but less than metacarpalII length; 1, one half to one third
metacarpal II length; 2, subequal to metacarpal II length. UO
261. Articular surface between metacarpals I and II: 0, placed
just at proximal end of metacarpal I; 1, extends well into diaphysis
of metacarpal I. (G
, 1986)
262. Metacarpal II length: 0, much less than humerus length;
1, about 50% or greater humerus length.
263. Metacarpal III length: 0, subequal to metacarpal II; 1,
clearly shorter than metacarpal II; 2, clearly longer than metacarpal II. UO (P -M
., 1993)
264. Metacarpal III width: 0, not very much narrower (greater
than 50%) than metacarpal II; 1, very much narrower (less than
50%) than metacarpal II. (G
, 1986)
265. Metacarpal III shape: 0, straight; 1, bowed laterally.
(G
, 1986)
266. Base of metacarpal III: 0, along same plane as metacarpals I and II; 1, set on palmar surface of hand below base of meta, 1986)
carpal II. (G
267. Proximal articulation of metacarpal III: 0, subquadrilateral; 1, triangular. (R
& D , 1993a)
268. Metacarpal IV length: 0, more than half length of metacarpal II; 1, less than half length of metacarpal II.
269. Metacarpal-phalangeal joints: 0, hyperextensible, deep
extensorpitsonmetacarpalsI-III; 1,nothyperextensible,extensor
pits on metacarpals I-III reduced. (P -M
., 1994)
270.Longestdigitinmanus: 0,digitIII; 1,digitII; 2,digitI. UO
271. Penultimate phalanx: 0, not longest of nonungual phalanges; 1,longestnonungualphalanx.(S
&N ,1992)
216. Glenoid orientation: 0, caudolateral; 1, lateral. (N
&
P
, 1997)
217. Coracoid shape: 0, craniocaudally elongate to subcircular; 1, ventral coracoidal process well developed; 2, subrectangular, dorsoventral depth more than 130% of craniocaudal width; 3,
, 1986)
strut-like. UO (G
218. Coracoidcaudoventralprocesslength: 0, less than twice
glenoid diameter; 1, more than twice glenoid diameter. (C
,
N
& C , 1996)
219. Coracoidbiceps tubercle: 0, absent or poorly developed;
1,conspicuousandwelldeveloped.(P
-M
.,1993)
220. Coracoid angle with scapula at level of glenoid cavity: 0,
moderate; 1, sharp. (C
,N
& C , 1996)
221. Sternal plates: 0, separate; 1, fused.
222. Sternum carina: 0, absent; 1, present.
223. Sternum shape: 0, relatively round; 1, longer craniocaudally than wide mediolaterally; 2, wider mediolaterally than long
craniocaudally. UO
224. Sternum size: 0, craniocaudal length similar to coracoid
length; 1, much greater than coracoid length.
225. Furcula: 0, absent, clavicles unfused; 1, present.
226. Forelimb (humerus + radius + manus)/hindlimb (femur +
tibia+pes)lengthration: 0,lessthan50%; 1,greaterthan50%but
less than 120%; 2, greater than 120%. O (G
, 1986)
227. Forelimb/presacral vertebral series length ratio: 0, less
than 75%; 1, greater than 75% but much less than 200%; 2, about
200% or more. O (G
, 1986)
228.Manus/peslengthratio: 0,muchlessthan100%; 1,greater than 100%.
229. Humerus/scapula length ratio: 0, greater than 65%; 1,
less than 65%. (P -M
., 1993)
230. Humerus/ulna length ratio: 0, greater than 100%; 1, less
than or equal to 100%.
231. Ulna/femurlength ratio: 0, greater than 27%; 1, less than
27%. (S
., 1996)
232. Radius/humerus length ratio: 0, less than 75% but greater than 50%; 1, less than 50%; 2, greater than 76%. UO
233. Manus/(humerus+ radius)lengthratio: 0, less than 66%;
1, greater than 66%. (G
, 1986)
234. Humeral torsion: 0, absent; 1, present. (P
-M
., 1993)
235. Humeral shaft: 0, straight; 1, sigmoid. (H
, 1994)
236. Humeral head: 0, low and confluent with deltapectoral
and bicipital crests; 1, offset and emarginatedventrally by groove.
(M
& S , 1998)
237. Internal tuberosity(= ventral tubercle)on proximalend of
humerus development: 0, not well differentiated; 1, well differentiated and angular. (R
& D , 1993a)
238.Internaltuberosity(=ventraltubercle)ofhumerusshape:
0, conical; 1, craniocaudally compressed and longitudinally elongate.
239.Internaltuberosity(=ventraltubercle)ofhumerusdirection: 0, projected ventrally; 1, projected proximally; 2, projected
caudally, separated from humeral head by deep capital incision.
UO (C
,N
& C , 1996)
240. Humeral ends: 0, little or not expanded; 1, well expanded, greater than 150% midshaft diameter. (P -M
., 1993)
241. Deltapectoralcrest on humerus: 0, low; 1, expandedand
offset from humeral shaft. (M
& S , 1998)
242. Humeraldistalcondyle: 0, mainlyon distalaspect; 1, cranial aspect. (N , 1996)
243. Humeral entepicondyle: 0, small; 1, prominent.(M
& S , 1998)
OVAS
USSELL
UERTA
AUTHIER
AUTHIER
HIAPPE
ORELL
LARK
ÉREZ
HIAPPE
ORELL
ORENO et al
OVAS
LARK
AKOVICKY
AUTHIER
ORENO et al
ERENO et al
AUTHIER
ORENO
AUTHIER
et al
OLTZ
AKOVICKY
UES
USSELL
HIAPPE
ORELL
ÉREZ
ONG
AUTHIER
AUTHIER
USSELL
ORENO et
al
AKOVICKY
UES
ONG
ÉREZ
OVAS
AKOVI-
CKY
ORENO et al
AUTHIER
LARK
ÉREZ
ORENO et al.,
OLTZ
AUTHIER
ÉREZ
UES
ÉREZ
AUTHIER
ÉREZ
ONG
ORENO et al
ERENO
UES
51
OVAS
T.R. HOLTZ, JR.
300. Supracetabular shelf on ilium:
(ROWE & GAUTHIER, 1990)
272. Length of phalanx 3 of manual digit III/(sum of lengths of
phalanges 1 + 2 of digit III): 0, less than 100%; 1, greater than
100%. (GAUTHIER, 1986)
302. Ilium length:
long as femur.
276. Flexor tubercle of unguals: 0, well developed and proximally placed; 1, poorly developed and distally placed. (PÉREZ-MORENO et al., 1994)
1994)
310. Pubic shaft: 0, straight; 1, marked concave curvature cranially; 2, marked convex curvature cranially. UO (MAKOVICKY &
SUES, 1998)
281. Manual unguals II and III: 0, smooth proximodistal surfa1, small nubbin proximodistally. (RUSSELL & DONG, 1993a)
282. Manual ungual cross section: 0, generally oval, two to
three times as deep as wide; 1, blade-like, more than three times
as deep as wide; 2, subtriangular, as wide or wider than deep. UO
283. Manual ungual length: 0, moderate; 1, extremely long; 2,
relatively short. UO
284. Manual ungual curvature: 0, moderate; 1, extremely curved; 2, straight. UO
285. Pelvic girdle sutures: 0, unfused in adults; 1, fused in
311. Pubic apices:
NORELL & CLARK, 1996)
0,
in contact;
1,
separate. (CHIAPPE,
312. Pubic blade: 0, five or less times as long as broad;
least six times as long as broad.
1, at
313. Pubic apron: 0, transversely wide and proximodistally
long, extending more than 50% of pubis length; 1, limited to distal
half of pubis length; 2, strongly reduced transversely and restricted to distal 25% or less of pubic length. (NOVAS, 1996)
314. Pubic foramen perforating pubic apron in distal half of pubic shaft: 0, absent; 1, present. (HARRIS, 1998)
adults. (ROWE & GAUTHIER, 1986)
dolichoiliac. (GAUTHIER,
315. Pubic boot shape: 0, absent; 1, broad triangle, angle
between shaft and caudal portion of boot obtuse; 2, rounded, angle between shaft and caudal portion of boot acute; 3, boat-shaped (pointed cranially and caudally) in ventral view and angle
between shaft and caudal portion of boot acute; 4, triangular (apex
caudal) in ventral view and angle between shaft and caudal portion of boot acute. UO (HOLTZ, 1994)
287. Iliac blades dorsal surface: 0, do not meet along midline;
1, meet along midline. (HOLTZ, 1994)
288. Iliac preacetabular fossa for M. cuppedicus: 0, absent; 1,
present. (GAUTHIER, 1986)
289. Fossa for origin of M. cuppedicus on ilium; 0, narrow or
1, broad.. (MAKOVICKY & SUES, 1998)
290. Brevis fossa (for M. caudofemoralis brevis) depth: 0, absent or poorly developed; 1, pronounced. (GAUTHIER, 1986)
291. Brevis fossa distal end: 0, brevis fossa absent; 1, distal
tapered; 2, broad distal end. UO (SERENO et al., 1996)
292. Preacetabular ala of ilium: 0, not greatly expanded vertically; 1, greatly expanded vertically. (RUSSELL & DONG, 1993a)
293. Preacetabular portion of ilium: 0, subequal in length to
postacetabular portion; 1, significantly longer than postacetabular
316. Pubic boot proportions: 0, caudal portion same length as
cranial portion; 1, caudal portion longer than cranial portion, but
latter present; 2, cranial portion absent. UO (GAUTHIER, 1986)
absent;
317. Pubic boot size: 0, absent; 1, less than 30% as long as
pubic shaft; 2, greater than 30% as long as pubic shaft. O (SERENO et al., 1996)
318. Pubic-ischial contact:
only narrow region. (SERENO et
0,
dorsoventrally deep shelf;
1994)
al.,
319. Pubis and ischium proximal shafts:
(SERENO et al., 1994)
portion.
320. Ischium/pubis length ratio:
than 66%. (GAUTHIER, 1986)
294. Preacetabular process of ilium: 0, cranial margin smooth;
1, cranial margin notched. (SERENO et al., 1996)
295. Median vertical ridge on external surface of ilium: 0, absent; 1, present. (RUSSELL & DONG, 1993a)
296. Caudodorsal margin of ilium: 0, gently arched; 1, curves
1,
0, broad; 1, narrow.
0, greater than 75%; 1, less
321. Ischial antitrochanter: 0, small or absent; 1, large. (ROWE
& GAUTHIER, 1990)
322. Obturator process shape: 0, joined to pubic articular process; 1, separate, trapezoidal; 2, separate, triangular. UO (HOLTZ,
1994)
caudoventrally. (GAUTHIER, 1986)
gin;
et al.,
309. Pubis orientation: 0, propubic, shaft approximately 45
degrees from horizontal; 1, propubic, proximal portion of shaft
approximately 30 degrees from horizontal; 2, vertical; 3, opisthopubic; 4, caudoventrally directed with twisted proximal region. UO
ce;
297. Postacetabular ala of ilium:
minate. (RUSSELL & DONG, 1993a)
about as
308. Pubic fenestra ventral to obturator foramen: 0, absent; 1,
present. (ROWE, 1989)
ventrally compressed, with proximal articular surface quadrangular. (NOVAS, 1996)
1,
1,
307. Obturator foramen of pubis: 0, present; 1, open ventrally
to form obturator notch. (GAUTHIER, 1986)
280. Pollex ungual shape: 0, trenchant, dorsoventrally deep,
with proximal articular surface elliptical; 1, stout and robust, dorso-
brachyiliac;
clearly shorter than femur;
1,
ischiadic peduncle. (GAUTHIER, 1986)
279. Pollex ungual length: 0, less than three times longer than
height of articular facet; 1, greater than three times longer than
height of articular facet. (SERENO et al., 1996)
0,
present.
306. Pubic peduncle of ilium depth: 0, extends ventrally to the
same level as ischiadic peduncle; 1, extends more ventrally than
278. Pollex ungual size: 0, subequal to unguals of digits II and
III in size; 1, larger than other manual unguals.
Ilium shape:
0,
dally than mediolaterally. (SERENO
277. Manual ungual, region palmar to ungual groove: 0, wider
than region dorsal to ungual groove; 1, palmar and dorsal regions
subequal in width. (RUSSELL & DONG, 1993a)
286.
1,
303. Prominent antitrochanter on ilium: 0, absent; 1, present.
304. Iliac-ischial articulation: 0, larger than iliac-pubic articulation; 1, smaller than iliac-pubic articulation. (SERENO et al., 1994)
305. Pubic peduncle of ilium proportions: 0, more developed
mediolaterally than craniocaudally; 1, more developed craniocau-
1993a)
1986)
absent;
301. Acetabular height/craniocaudal length: 0, 27-33%;
about 50%. (SERENO & NOVAS, 1992)
273. Pollex: 0, ends at level of mid-length of phalanx 2 of digit
1, ends at level of mid-length of phalanx 1 of digit II.
274. First phalanx of pollex: 0, Less than or subequal to length
of metacarpal II; 1, greater than length of metacarpal II.
275. Manual ungual, dorsal edge of articular facet: 0, relatively
smooth; 1, pronounced lip on dorsal edge. (RUSSELL & DONG,
II;
0,
0, squared caudally; 1, acu-
323. Obturator process position: 0, joined to pubic articular
process ("obturator flange" of CHARIG & MILNER (1997)); 1, separate, proximally placed; 2, separate, distally placed; 3, absent,
caudoventral margin of ischium smooth from obturator notch to
tip. O
298. Postacetabular process of ilium: 0, straight caudal mar1, concave caudal margin.
299. Supracetabular crest on ilium: 0, present; 1, absent.
(RUSSELL & DONG, 1993a)
52
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
349. Lateroproximal condyle of tibia development in proximal
view: 0, bulge from main surface of tibia; 1, conspicuous waisting
between body of condyle and main body of tibia.
324. Ischium obturator process or flange: 0, not perforated by
foramen; 1, perforated by foramen.
325. Ischial proximodorsal process just distal to iliac process:
0, absent; 1, present. (NOVAS & PUERTA, 1997)
326. Semicircular scar on caudolateral surface of ischium, just
distal to iliac process: 0, absent; 1, present. (HOLTZ, 1994)
327. Ischial foot: 0, present; 1, absent. (HOLTZ, 1994)
328. Ischiadic terminal processes: 0, in contact; 1, separate.
350. Crista fibularis size: 0, absent; 1, not well developed;
well developed. O (PÉREZ-MORENO et al., 1993)
351. Crista fibularis position: 0, absent; 1, proximal; 2, distal.
UO (HOLTZ, 1994)
352. Tibia distal end: 0, not backing calcaneum; 1, expanded
to back calcaneum. (SERENO
(CHIAPPE, NORELL & CLARK, 1996)
et al.,
1996)
353. Fibula: 0, broadly separated from tibia throughout main
shaft; 1, closely appressed to tibia throughout main shaft. (HOLTZ,
329. Femur shape: 0, prominent sigmoid curvature (S-shaped
in two planes); 1, bowed in convex arc with less pronounced sigmoidality. (GAUTHIER, 1986)
1994)
354. Fibula proximal end: 0, less than 75% proximal width of tibia; 1, 75% or more proximal width of tibia. (SERENO et al., 1996)
330. Femoral head angle to shaft: 0, less than 90 degrees
(head directed ventrally); 1, approximately 90 degrees (head directed horizontally); 2, greater than 90 degrees (head directed
dorsally). O (HARRIS, 1998)
355. Proximal region of fibular medial face: 0, flat; 1, slightly
concave; 2, well excavated. O (PÉREZ-MORENO et al., 1993)
331. Femoral head shape: 0, bulky; 1, transversely elongate;
2, rounded. UO
332. Greater trochanter of femur position: 0, confluent with femoral head; 1, cleft from femoral head. (HOLTZ, 1994)
333. Greater trochanter of femur shape: 0, rugosity; 1,
356. Sulcus in proximomedial region of fibula:
present. (SERENO et al., 1994)
334. Anterior (= lesser) trochanter: 0, absent; 1, separated
from femoral head by cleft; 2, nearly confluent with femoral head.
1996)
1,
359. Anterior surface of distal fibula: 0, does not overlap ascending process of astragalus cranially; 1, overlaps ascending
process of astragalus cranially. (ROWE & GAUTHIER, 1990)
335. Anterior trochanter of femur shape: 0, absent; 1, conical
prominence; 2, alariform; 3, cylindrical in cross-section; 4, trochanteric crest (fusion of greater and anterior trochanters). UO
360. Fibular distal end: 0, greater than twice craniocaudal
width at midshaft; 1, less than twice craniocaudal width at
midshaft, and consequently astragalar cup for fibula reduced; 2,
pinches out less than half-way down tibia length. O
336. Anterior trochanter of femur position: 0, absent; 1, proximalmost point below femoral head; 2, proximalmost point above
distal margin of femoral head; 3, proximalmost point extends above proximal margin of femora head. O
361. Fibular fusion with distal tibia: 0, absent; 1, present.
362. Astragalar ascending process: 0, not reduced in horizon-
1,
tal dimension, proximodistally very low ("ceratosauroid condition"); 1, mediolaterally reduced, craniocaudally wide and
proximodistally low ("allosauroid condition"); 2, craniocadually reduced and proximodistally tall, with dorsal margin sigmoid ("ornithomimoid/albertosauroid condition"). O (GAUTHIER, 1986)
338. Trochanteric shelf of femur: 0, moderately developed ridge transversely directed; 1, well developed; 2, absent. UO (PÉREZ-MORENO et al., 1993)
363. Round external fossa at base of ascending process of
astragalus: 0, absent; 1, present. (HOLTZ, 1994)
339. Muscle scar in craniodistal region of femur: 0, absent; 1,
present, non-elliptical in shape; 2, elliptical in shape. UO (PÉREZet al.,
absent;
357. Cranial protuberance on fibula below expansion: 0, ab1, present. (HOLTZ, 1994)
358. Fibular tubercle for M. iliofibularis (= "anterolateral process"): 0, craniolaterally projecting; 1, laterally projecting. (NOVAS,
UO
MORENO
0,
sent;
moundlike eminence. (GAUTHIER, 1986)
337. Fourth trochanter of femur: 0, developed, alariform;
little developed; 2, absent. O (GAUTHIER, 1986)
2,
364. Astragalar distal condyles: 0, oriented ventrally; 1, oriented cranioventrally. (SERENO et al., 1996)
1993)
340.
Medial epicondyle (= mediodistal crest): 0, absent or
weakly developed; 1, pronounced, extends one quarter or more
the length of the femoral shaft.
365. Pronounced horizontal groove across cranial face of astragalar condyles: 0, absent; 1, present. (HOLTZ, 1994)
366. Astragalocalcaneum: 0, absent, astragalus and calcaneum separate; 1, present, astragalus fused to calcaneum.
341. Extensor groove in craniodistal region of femur: 0, absent; 1, shallow and not conspicuous; 2, deep and conspicuous. O
(HARRIS, 1998)
367. Medial tuber on calcaneum: 0, small; 1, enlarged. (RUS-
342. Groove in lateral condyle of femur: 0, absent; 1, present.
SELL
(ROWE & GAUTHIER, 1990)
& DONG, 1993a)
368. Tibiotarsus: 0, absent, proximal tarsals unfused with tibia; 1, present, astragalocalcaneum fused to tibia.
369. Metatarsus proportions: 0, moderate; 1, elongate relative
to most other theropods of same femur length; 2, shortened relative to most other theropods of same femur length. UO (HOLTZ,
343. Adductor fossa and associated caudodistal crest of distal
femur: 0, present, prominent; 1, reduced or absent. (CHIAPPE, NORELL & CLARK, 1996)
344. Ectocondylar tuber: 0, proximodistally short, proximally
placed; 1, proximodistally long, pronounced, and extends almost
to distal end of femur.
1994)
370. Metatarsal ossification: 0, proximally separate; 1, co-ossified proximally; 2, co-ossified throughout shaft. UO
345. Sulcus along medial side of base of crista tibiofibularis: 0,
absent; 1, present. (ROWE & GAUTHIER, 1990)
346. Cnemial process: 0, projects caudally; 1, arises out of the
371. Metatarsal cross-sectional proportions: 0, subequal or
wider mediolaterally than craniocaudally at midshaft; 1, deeper
craniocaudally than mediolaterally at midshaft. (HOLTZ, 1994)
lateral surface of tibial shaft. (HOLTZ, 1994)
372. Metatarsal V:
(GAUTHIER, 1986)
347. Incisura tibialis cranialis: 0, occupies less than 50% of
medial surface of proximal tibia; 1, occupies more than 66% of medial surface of proximal tibia. (HARRIS, 1998)
0,
not reduced;
1,
vestigial or absent.
373. Metatarsals II and IV: 0, separated at midshaft on plantar
surface by metatarsal III; 1, contact at midshaft on the plantar surface. (HOLTZ, 1994)
348. Lateroproximal condyle (fibular condyle) on proximal end
of tibia position: 0, large and posteriorly situated; 1, small and medially situated. (RUSSELL & DONG, 1993a)
374. Metatarsal IV size: 0, subequal in length to metatarsal II;
1, longer than metatarsal II and closer to metatarsal III in length.
(HOLTZ, 1994)
53
T.R. HOLTZ, JR.
375. Metatarsal III dorsal surface area: 0, similar in size to metatarsals II and IV; 1, clearly larger than metatarsals II and IV; 2,
clearly smaller than metatarsals II and IV. UO (GAUTHIER, 1986)
20103
00010
10010
12110
01011
376. Metatarsal III dorsal surface shape: 0, elliptical end; 1,
hourglass shaped; 2, dumbbell shaped (cranial and (especially)
plantar surfaces expanded to slightly overlap surfaces of metatarsals II and IV); 3, dorsal surface not exposed. UO
0????
00100
?????
0??10
?0???
?????
1??00
01110
01011
"proximal shaft" of metatarsal III complete pinched out above distal wedge. O (HOLTZ, 1994)
378. Metatarsal I length: 0, not reduced; 1, reduced but retains
phalanges; 2, absent. O (GAUTHIER, 1986)
379. Metatarsal I vertical position: 0, contacts distal tarsals; 1,
placed near midpoint of metatarsal II shaft; 2, placed at distal end
of metatarsal II. O (GAUTHIER, 1986)
386. Pedal ungual II size: 0, subequal to pedal ungual III; 1, significantly longer than pedal ungual III (HOLTZ, 1994).
APPENDIX II
0????
?????
??000
000?0
?????
1????
0??00
?11??
???11
?????
0?1??
00?0?
?????
???0?
01002
???01
Archaeopteryx
Data matrix used for phylogenetic analysis. Scoring: 0 = primitive state; 1, 2, 3, 4, or 5 = derived character states; ? = data missing due to lack of knowledge of the particular anatomical region of
that taxon or equivocal due to evolutionary transformation of taxon
(e.g., dental characters in edentulous theropods; proportions
between manual digits III and II or I in didactyl theropods).
20000 10101 00001 11000 00000 00001 01010 01?00 00000
00000 10001 00100 00110 00002 10110 20101 ?2000 11?01
11000 000?0 11000 00001 10?01 01100 10111 002?0 10100
0???1 1???? 10100 10111 0?100 ?1110 00000 11??0 0?101
?0??3 ?0002 01010 13021 00111 ?1110 11000 12111 10001
12(01)01 02111 11101 10101 10010 00211 33001 11011 1?111
11100 00000 01010 10101 10100 11110 1011? 110(23)0 01203
21111 02301 01112 ?1122 32220 101?0 11??2 2111? ?0001
02010 00001 01000 ?0122 10100 0
MATRIX
Outgroup
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
20002
?????
?????
?????
?????
?????
?????
?????
11?00
?011?
?????
?????
?????
?????
?????
?????
12111
?00??
?????
?????
?????
?????
?????
?????
00110 11000 10110 20011 20210
01002 10020 02011 ????1 11100
00100 00001 11001 21101 ?0100
11110 01001 00012 00010 02010
10?00
10100
00000
01000
10011
02000
01100
00111
Bagaraatan
?????
?????
?????
?????
?????
?????
?????
??11?
12010
Abelisaurus
????? ????? ????? ????? ?????
????? ????? ????? ????? ?????
????? ???00 0000? ?0?1? 1????
????? ????? ????? ????? ?????
????? 00001 ?2000 ????? ?????
????? ????? ????? ????? ?????
????? ????? 1???1 1???? 00000
0???? ????2 21114 2?200 00011
1?1?? ????? ????? ????? ?
?????
?????
?0100
?????
?????
?????
1?011
00011
?????
?????
??00?
?????
?????
?????
01000
11112
?????
?????
0??00
?????
?????
?????
??00?
01101
?????
??1??
11110
?0000
?????
33000
10011
11011
?????
??00?
?????
10001
01010
1?111
01001
11???
?????
12011
?????
?0??1
?????
1?101
01103
?????
Caenagnathidae
?????
?????
?0000
0????
20104
??0??
10011
01111
02?10
Acrocanthosaurus
00000
0011?
010?0
00000
20111
?????
?????
?0000
?????
??001
10010
?????
20??0 1?101 00000 01000 00000 00001 01000 00?00 00000
00?00 1000? 00010 00??1 01002 ?1??? 20??? ?100? ?1?01
11000 00000 11000 00101 10??? 0?00? ??11? ?02?0 1011?
1???1 1???? 1(01)(01)00 10(01)11 001(01)0 (01)1110
(12)0100 011(01)0 01100 00013 (01)00(01)2 010(01)0
1300(01) ???11 001(01)0 11100 0110(01) 1111? 00010 11100
11011 11100 1(01)1(01)0 01211 3331? ????? ???02 1??10
10111 ?21(02)0 10001 10(01)00 1(01)1(01)0 101(01)0
(01)1030 (01)120(03) (01)(01)110 0?300 01112 (12)11(12)(24)
3(01)200 10110 11111 21110 0010(12) (01)2010 (01)0(01)00
(01)?(01)(01)2 (03)(02)121 (01)0000 (01)
sickle-shaped (blade-like cross-section and highly recurved).
(HOLTZ, 1994)
00010
01001
00000
01110
????? 00120 11000 00?1? ????1
00010 00102 10??0 01000 00???
????? ????? ????? ????? ?????
0???? 1?100 0?001 0?002 00010
???0? 00??? 0200? 00?01 ?0000
???11 0???1 0000? ????? 00111
00000 00000 10001 100?0 00000
01100 00011 20012 10220 20010
00000 0?000 10110 ??000 0
Alvarezsauridae
383. Pedal digit II: 0, not hyperextensible; 1, hyperextensible.
(HOLTZ, 1994)
384. Pedal unguals III and IV cross-section: 0, subtriangular;
1, vertically oval in cross-section. (RUSSELL & DONG, 1993a)
385. Pedal ungual II: 0, same shape as other pedal unguals; 1,
00001 00002 10000 00010 30101
01110 00?0? 010?0 11101 00???
00000 001?? ????? ????? ?????
????? ????? ????? ????? ?????
????? ????? ????? ????? ?????
????? ????? ????? ????? ?????
????? ????? ????? ????? ?????
????? ????? ????? ????? ?????
????? ????? ????? ????? ?
?????
11101
01014
00001
00210 00001 00110 21000 10110 20011 10(12)10 00100 00011
01000 00010 01002 10020 02011 00101 11000 10100 02000
00000 11011 00100 00001 11001 21101 00100 00000 01100
01110 00000 11110 01001 00002 00010 10000 00001 00111
20003 00000 00001 02010 00001 00100 10011 01010 ????1
00010 00111 00001 00000 00000 00111 33001 10010 11101
10010 001(01)0 00000 10101 10000 00000 10011 01000 01014
12110 01100 00012 11012 20220 20010 10012 11102 00001
01011 00000 01000 10110 01000 0
1996)
00011
01001
?????
?????
?????
?????
?????
?????
?????
010?0
10110
01000
11102
Allosaurus
380. Metatarsal I horizontal position: 0, along plane with metatarsals II-IV; 1, plantar to medial side of metatarsal II; 2, completely
reverted. O
381. Pedal digits II and IV: 0, subequal in length and shorter
than digit III; 1, digit IV larger than II and closer to III in length; 2, digit II longer than IV and closer to III in length. UO (GAUTHIER, 1986)
382. Pedal digit I phalanges 1+2: 0, longer than pedal digit III
phalanx 1; 1, subequal to pedal digit III phalanx 1. (SERENO et al.,
00000 00000 00000 00000 00000
00000 00000 00000 00000 00000
00000 00000 00000 00000 00000
00000 00000 00000 00000 00000
00000 00000 00000 00000 00000
00000 00000 00000 00000 00000
00000 00000 00000 00000 00000
00000 00000 00000 00000 00000
00000 00000 00000 00000 0
11011
33001
10011
10012
Afrovenator
377. Arctometatarsus: 0, absent; 1, present; 2, present, with
00000
00000
00000
00000
00000
00000
00000
00000
00000
0000? 10000 02000 00??1 ?0100
00111 10001 10000 00000 00111
00100 00000 101?1 10000 00000
01100 00012 21012 20220 20010
00000 01000 10110 ??0?0 ?
????? ?100? 00000 0000? ?????
????? ????? ????? ????? ?1??0
00002 00110 11000 00000 0100?
????? 10200 ?0?11 11?00 001?0
2110? 2???0 ?0002 20??? ?????
????? ????? ????? ????? ???1?
00000 11000 10111 111?0 11110
02100 01011 11113 32220 00010
0101(01) 11002 31111 0?010 ?
Carcharodontosaurus
0???? ???01 00112 20000 00010 30011 2011? 11?00 12011
01001 0111? ??100 1???? ????? ?0??? ????? ?0100 00100
00000 01000 001?? ????? ????? ????? ????? ??001 0?1?0
54
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
????? ????? ?0110 01000 01?12 100?0 ??000 00001 00?11
00010 (01)1101 01111 00010 10110 11101 01002 11010 10110
?0??? ????? 1??0? ????? 00??1 00??? ????? ????? ?????
01000 00001 10103 20101 01011 13330 01111 01111 11100
????? ????? ????? ????? ????? ????? ????? ????? ?????
02010 00001 11000 02111 11101 10101 00010 00211 33001
????? ????? ????0 101?1 10000 00000 10011 01000 01???
11111 11101 11000 01100 11010 10(01)(01)1 1(01)010 11110
?211? 0?100 0??12 20012 20220 20010 10011 11?0? ???01
10011 11030 01103 (12)1111 02200 01011 11122 3(12)220
0???? ????? ????? ????? ????? ?
10100 11112 21112 00001 02010 00020 01011 00110 10101 1
00011 00000 00002 10000 00010 30001 20002 01?00 12011
????? ????? ?0??? ????? ????? ????? ????? ????? ?????
00001 00110 00100 010?0 11101 00??? ????? ?011? 0000?
????? ????? ????? ????? ????? ?0??? ????? ????? ?????
????? 00000 00100 00001 10000 10000 00000 00000 00000
????? ????? ???0? ??0?? ?0??? ????? ????? ?0000 0??00
01100 00010 01211 01000 00002 10010 00001 00000 10002
????? ????? ????? ????? ????? ????? ????? ????? ?????
?0005 1011? 00?00 0?0?? 10??0 00??? 11100 00010 00000
????? ????? 0???? ?200? ????? ????? ????? ????? ?????
00?10 11010 01000 00000 00010 00000 31000 00000 0001?
????? ???11 00001 ?000? ????? ????? ????? ????? ?????
????? ????? ????1 10001 20000 00001 11010 00000 01002
????0 00?10 ?0100 ????? ????? ????? ????? ?1001 0110?
01010 10000 00011 00012 11111 01011 0000? ????? ?????
??11? ???00 0??12 20012 ?1220 10010 11011 21112 000??
????? ????? ????? ????? ????? ?
02010 00000 0?0?? ?0??? ????? ?
00111 00101 00000 (03)0000 11010 10001 10100 00000 00011
????? ????? ????? ????? ????? ????? ????? ????? ?????
00000 00010 00102 01000 (01)1111 00100 0?0?0 00110 00000
????? ????? ????? ????? ????? ????? ????? ????? ?????
00000 00000 01100 00001 10000 11000 00100 00000 01000
????? ????? ????? ????? ????? ????? ????? ????? ?????
01100 00010 11211 01000 00102 00110 01001 00000 00002
1???? ????? ?0210 01010 00001 0?011 00001 01000 00002
20005 10010 00100 01000 ?0?00 00000 00000 00010 ?????
?0004 00010 00?00 01110 ???00 00??? ????? ????? ?????
???0? ???00 00001 00000 00000 ????? 31000 0?000 0000?
????? ???00 00000 0000? ????? ????? ?1??? 0???? ???0?
????? ????? ????1 10001 20000 00001 11000 00012 01002
????? ????? ????1 10001 20000 00001 10100 0??1? ?????
01010 10000 00010 10011 11111 01001 00000 00001 11010
??01? 10000 00010 00001 20121 01011 00002 10?0? ?00??
00000 10000 0?001 20??? ????0 ?
00000 1?010 0???1 20??? ??0?? ?
Dryptosaurus
Carnotaurus
Elaphrosaurus
Ceratosaurus
Coelophysidae
Eustreptospondylus
00000 00001 10000 (03)0000 00010 (02)(01)001 (02)0000
00000 000?1 00120 10000 01??? ????1 00011 0??00 000??
00000 00010 00000 00000 00000 00010 01000 00100 00000
?0100 00010 00?00 ????0 0?1?? 00??? ????? ?0000 ??000
10000 00000 00000 00000 00000 00000 10000 01000 00000
?0000 00000 0?000 00001 10??? ????? 0???? 00000 0?100
(01)(01)000 00000 11100 01101 10200 11010 00001 00011
0??0? ?0000 ?1110 01000 00002 ??010 00000 00000 00001
00001 00000 00002 10003 00(01)10 00100 01001 10000 00000
?00?? 00000 0??00 ????0 ????? ????? 1100? ????? ?????
00000 0001? ????0 00000 00011 00000 00000 00000 001(01)0
???0? ?0?10 00001 0000? ????0 ????? ????? ????? ???0?
31000 00000 00001 10000 00000 00001 10001 20000 00001
????? ????? ????0 10?01 1???0 00000 1?010 00?0? 010?2
10000 00111 01000 00010 10000 00010 (01)0001 101(01)1
0101? ???0? ???11 20012 10210 10010 10002 11100 00001
01001 00000 00000 10010 (01)0000 10(01)01 01001 20110
01010 00000 0?000 10??? ??00? ?
Gasosaurus
00000 0
Coelurus
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????? ????? ????? ????? ????? ??000 0????
????? ????? ?0??? ????? ????? ????? ?0??0 ??000 ?0???
????? ????? ?1100 10011 00001 000?1 ?0000 00000 00?01
?00?3 0?00? ????? ????? ????? ????? ????? ????? ?????
20??? 2??0? 0?000 ???00 ????? ????? 1???? ????? ?????
????? ???10 ?0001 ?0?0? ????? ????? ????? ????? ?????
???00 0??11 01001 10001 00000 00111 31001 11110 11101
????? ????? ????0 10101 10000 00000 10011 00000 010?3
11??0 01??? 01010 ????? ????? ????? ????? ?1010 01013
11010 01100 00012 21012 2022? 2???? ??000 0101? 00001
1211? ????? ???1? ???12 102?? ????? ????? ?11?? ????1
0???? ???00 0?0?0 ?0??? ????? ?
Giganotosaurus
0???? ??010 1?0?? ?0??? ????? ?
Compsognathidae
00011 00110 00112 ????? 00110 300?1 20210 1???? ???11
001?0 0?101 00001 11000 0000? 00001 00010 01?00 00000
01001 01110 000?? ?10?0 02111 00??? ????? ?0100 ?0100
00000 0000? ?0?11 ?00?? 0?012 ?0??? ????? ??00? ?0???
?0000 ??0?? 00001 00001 ????? ????? ????? 00001 01100
????? ??0?? 0??00 00001 10011 0110? ?0100 00000 00100
0???? ????? ?1110 01001 0?0?? 0?0?? ?00?0 00000 001??
0???0 00000 ?1110 01?01 0?100 ??010 001?0 11??0 0?000
?0?03 ??000 ?0??? 02000 ?0?01 ??1?? 1100? 0???? ?????
?0??3 ????0 01?01 12000 00101 ?0110 11001 01010 ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
00000 00010 00000 00000 00000 ?0?1? 33001 10110 1?102
????? ????? ????0 10101 10000 00000 11011 01000 01014
1?010 0?100 0?020 1010? ?0100 00100 1001? 01000 01??3
12110 01100 00012 20012 20220 20010 10??2 1110? ???01
21111 02100 0001? ???12 302?0 ????? ????? ?111? ?0001
0101? 0?0?? ????? ????? ????? ?
Megalosaurus bucklandi
02?1? 0?000 0100? ?0111 11000 1
Dilophosaurus
????? ????? ?0??? ????? ????? ????? ????? ????? ?????
00000 00001 10000 10000 00010 21001 20000 00000 00011
????? ????? ????? ????? ????? ?0??? ????? ????? ?????
00000 00010 01100 00010 02010 00??? ??0?? ?000? 00000
????? ????? ???00 000?? ?000? ????? ????? 00000 0??00
00000 00000 00000 00001 10001 01100 00000 01000 00100
0???? ????? ????? ????? ????? ????? ????? ????? ?????
01100 01101 00200 11000 00000 00010 00001 00000 00002
??003 0000? ????? ????? ????? ????? 11000 00010 ?????
10002 00000 00?00 01000 ?0?00 00000 00000 00010 ?????
0??00 ???10 00??1 000?0 00010 ????? ????? ????? ?????
00000 00011 00000 00000 00000 00?10 31000 00000 00001
????? ????? ????? 10101 10000 00000 ??011 0???0 ??0??
10000 00100 00000 10001 20000 00001 10000 00110 01000
????? 00000 00011 01012 10210 20010 10011 11??? ?????
00010 10000 00010 10001 10101 01001 00001 00000 ?0010
????? ???00 0?0?0 10??? ???0? ?
Microvenator
00000 00000 01001 00110 2?000 0
Dromaeosauridae
????? ????? ????? ????? ????? ????? ????? ????? ?????
00010 00000 00001 11000 00000 000(01)1 01(01)10 0(01)100
????? ????? ????? ????? ????? ????? ????? ????? ?????
01000 00000 00000 00102 10120 00111 01100 11101 21001
????? ????? ???10 0111? ?2??? ??0?? ????? ????? ?????
11001 10(01)(01)0 00001 00100 00001 10111 11111 00101
????1 10000 11210 11111 11?00 0?1?0 10000 11000 10?01
55
T.R. HOLTZ, JR.
20??? ?1?0? 0???? ???02 ????? ????? ????? ?101? ?????
????? ????? ????? ????? 30??? ????? 1?100 01010 00?0?
????0 02?11 10001 00111 00010 ????? ????? 1???? ???0?
?1??0 ?2110 00000 ?0000 00001 10110 33002 10000 11111
????1 01??? ?1010 ??11? ???0? ???10 1?011 01001 01103
11010 11001 0222? ????? ????? ????? ????? ????? ?????
0111? ????? ???12 21113 32220 10010 11011 11112 0010?
????? ????? ????? ????? ????? ????? ????? ????? ?????
02010 0?0?? ????? ????? ????? ?
????? ????? ????? ????? ????? ?
00000 10001 00120 10000 00111 10111 20010 01?00 00011
0???? ????? 00120 10000 01??? ????? ????? ??0?0 00???
00010 00010 01102 10020 00001 00??1 1???? ?0100 ?0000
????? ????? ????? ????? ????? ?0??? ????? ?0??0 00000
0???? 110?1 00000 00001 10100 01001 00100 00000 0?100
?0??? 100?0 00001 0000? ????? ????? ????? ?0000 0??00
0???0 0???? 11110 01001 00002 00010 10010 00000 00?01
0??10 00000 11110 01101 01000 10010 00000 00000 00001
10003 0000? 0?001 ????? 00??? ????? ????? ????? ?????
10003 ?0??? 0???? ????? ????? ????? 10001 00010 ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
00?00 00?11 00001 01000 0?010 ????? ????? ????? ?????
????? ????? ????? 10101 10000 00000 1?011 00002 010?2
????? ????? ????0 1000? ?00?1 0?000 1001? 00000 010?2
01010 00010 000?? ????? ????? ????? ????? ????? ?????
01010 00000 00011 00012 2?210 100?0 ?0001 11??1 0????
????? ????? ????? ????? ????? ?
0???? ??000 0?000 10??? ????? ?
00210 00101 00120 ????? ?011? ??0?? ????? ??1?? ?????
00010 00000 00121 21000 00??? 1???? ???1? 0???? ???10
????? ????? ????? ????? ????? ?0??? ????? ????? ?2???
0?00? 0?0?? ?1100 00020 01002 ?0?0? ????? ????? ???0?
????? ????? ???00 0000? ?0??? ????? ????? 00000 0?100
????? ????? ?0000 0000? ?00?? 0?0?? ??1?0 00000 01100
0???? ????? ?11?? ????? 0??0? ????? ?0000 00??? 0?1??
????? ????? ????? ????? ????? ????? ????? ????? ?????
??0?? ?0?0? ?0?0? ?2?10 ???0? ?01?? 10011 0???0 ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????0 10101 1???? 000?? 11?11 01000 010?4
????? ????? ????? ????? ????? ????? ????? ????? ?????
12110 01100 00012 20012 20220 2??10 ????? ????? ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ??000 ????? ?0??? ??000 ?
????? ????? ????? ????? ????? ?
00010 0?000 00121 11000 00000 10001 00010 00??0 0?000
????? ????? ????? ????? ????? ????? ????? ????? ?????
00000 0000? 01101 00020 0?012 001?? ???0? 210?? ?0???
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ??0?0 00000 00101 10001 1100? ?0100 00000 01100
????? ????? ????? ????? ????? ????? ????? ????? ?????
0???? ????? ?0110 01101 0?000 000?0 ?0010 00000 00001
????? ????? ????? ????0 0???? ??1?? 100?0 01100 00001
20003 00001 0?000 12010 ???01 ?0110 ????? ????? ?????
2?0?4 00102 01011 13021 ???11 ?1210 11000 1???? ?????
110?0 02111 00001 00101 000?0 ????? ????? ????? ?????
????? 0???? ????? ????1 11?10 ????? ????? ????? ?????
????? ????? ????0 10101 10000 10000 10011 01000 010??
????? ????? ????0 10101 10100 11100 10011 11020 01103
??110 02100 01012 21112 31221 10010 ????? ??1?1 ?????
21111 02201 01112 21124 212?? 1???? 111?2 21110 ?0?02
????? ???00 0?001 10??? ??000 1
02010 00000 01000 00122 ?11?1 ?
Piatnitzkysaurus
Monolophosaurus
Proceratosaurus
Neovenator
Rahonavis
Ornitholestes
Ornithomimidae
Scipionyx
11?10 11200 01001 11110 00001 00001 010(01)0 00?10 00000
?00?0 0?000 00001 11000 0001? 00001 000?0 00??0 0??00
10000 00001 01101 00020 01002 1110? 10101 (01)1001 10101
10?00 0001? 0?100 000?0 0?002 ????? ????? ??0?0 ?0???
00111 00100 10010 01101 10001 01101 10100 ????? ?????
????? ??0?0 0??00 0?001 10??1 ?100? ???0? 00000 0?100
11110 10000 10110 01111 01100 00111 00000 00000 00001
0??10 0???? ???00 01111 0?00? ??010 000?0 00??0 ??00?
2100(34) 20101 00000 12110 30001 10110 11100 01010 ?????
????3 ??00? ???00 ????? ?0??? ?0??? 11000 01010 00?01
11000 00010 00000 00000 00001 10110 3300(12) 10000 11111
?1?00 02110 0??01 000?0 00?00 00111 33001 10110 1?1?1
11010 110(01)1 02(01)(02)0 11111 10010 00000 10011 01000
11010 0?100 0?000 10?01 ????0 00000 100?? 01000 0???3
01103 01110 02100 10012 21112 31220 20010 11112 11112
21?10 0??0? ?001? ???12 212?? ????? ????? ????? ?????
01001 02(01)10 00010 11102 01(12)?? 00000 0
????? ????? ????? ????? ????? ?
Ornithothoraces
Sinraptor
20000 11101 00001 11000 00000 00001 00200 01?00 00000
00010 00010 00000 1(01)001 10110 20011 (12)0210 10100
00000 10001 00110 00111 00002 11110 201?? ?2000 11101
00011 01000 00010 1?102 11020 02011 00101 11100 10100
11000 00000 11000 00001 10001 01100 10101 00200 101(01)2
00000 00000 01011 00000 00001 10100 21101 00100 00000
01110 10010 10100 11111 01100 01111 20000 10110 01101
01100 011?0 00000 11100 01001 00002 00110 11010 00001
2011(345) 10112 01010 13040 0(01)?11 01110 11000 13111
00011 20003 0000? 0?00? ????? 0???? ????? 11001 0????
11111 22101 02011 11111 11101 11110 01211 331(01)1 11011
112?? ????? ????? ????? ????? ????? ????? 3200? 1?010
11111 0?100 01?00 110(01)1 10(01)00 00100 11110 10111
1110? ????0 00??? 00000 10101 10000 00000 10011 01000
11030 (01)120(03) 2(01)11(01) 0?301 01112 21124 32220
01014 11010 01100 00011 20012 20220 20010 10002 11102
10110 11011 21110 00102 02010 10102 01000 00122 10(01)10
01001 01111 00000 01000 10110 ?1000 0
0
Spinosauridae
Oviraptoridae
00320 01001 00120 10000 00000 10101 00011 0010? 00001
01?11 11000 01001 11000 000(01)0 (01)0010 01210 01?01
0???0 ????? ????? ????0 00000 011?? ????? ?0100 00000
01(01)00 00000 000(01)0 00100 00020 01002 11110 10111
????? 00000 00100 00001 10?0? 111?0 0???? 01200 00110
01000 12001 ?0000 00002 00010 11010 02000 0100? 11110
01100 00000 01100 01000 00012 10010 1(02)010 0000(01)
????? ????? 0??10 10000 10200 11111 11100 00010 00000
10001 ?0??? ????? 0?000 ????0 2???0 00??? 11001 01010
11000 10101 20104 2?101 20000 00002 20?01 00000 11000
11?0? 00000 11?00 01001 10101 00000 ????? ????? ?????
01010 (01)1001 11000 02111 00001 10111 00010 00211 33001
????? ????0 00110 00100 10?01 1???0 00000 10011 0??00
11(02)11 1110(12) 1(01)011 01100 11010 10111 11100 11110
01000 ??00? 00000 000?0 100?? ??210 20010 ????? ?1??0
10011 01001 01103 01111 02200 01012 21113 32220 10010
000?? 010?? 000?? ????? ????? ????? ?
11011 11112 00101 02010 01000 01010 10121 1?010 0
Pelecanimimus
Therizinosauroidea
103?0 1?200 00001 11110 0000? 00001 01010 00??? ??000
00000 0000? 00101 10020 00012 01110 00?11 ?100? 10111
10??0 0000? 0?1?? 00??? ??002 ????? ????? ?10?? ?0?0?
?0000 00002 00100 01101 10001 01100 10100 11100 1?100
????? ??1?? 1??00 0?0?1 10??1 0110? ??100 112?0 121?2
1???? ????? ?0200 ?1111 00?00 10010 ?0010 0?000 00000
1??1? 10001 ??100 01111 0?100 0?110 00??0 ????0 ?0??1
?1?03 ??002 00?01 10002 20?01 00000 11000 01010 ?????
11?10 11001 00001 00000 00000 00000 00000 00101 00000
56
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
???0? ???11 11111 1?1?? ????? ????? ????? ????? ?????
????? ????? ????0 10110 00101 11100 10111 11020 01103
22111 02201 0111? ?1112 322?? ????? ????? ????? ?????
????? ????? ????? ????? ????? ?
Herrerasauridae
00010 00000 00002 00000 00000 00000 00000 ?0000 00000
00000 00000 01100 00000 01001 00?00 0?000 ?000? 00000
????? 00000 01000 00000 10000 0100? ?0000 00000 00000
00100 00001 00000 01000 00000 00000 00110 00001 00000
00000 00000 00000 01010 00000 ?0000 11100 00000 ?????
00000 02000 00000 00000 00000 00000 21000 00200 00000
00000 00000 00000 00000 00000 00000 00110 00040 000?1
02000 00000 01000 00000 00000 00000 00000 00010 00000
00000 00000 00000 00000 ?0000 0
Prosauropoda
00(24)00 01(01)01 00000 00000 00010 00000 00000 00000
00000 00000 00000 00000 00000 01102 01000 00010 00000
00000 00000 00000 01000 01(01)00 00000 01000 00010 10100
(01)0000 (01)0000 00000 00000 00000 00000 00010 00000
0000(01) 00000 00001 00000 00000 00000 00000 00000
0000(01) 00000 00000 00(01)00 0(01)000 00000 10100 00010
00000 11000 00000 00001 10000 00100 00000 00000 00000
00000 10000 000(01)0 00000 00000 00000 00000 00000 00000
00000 00000 00000 00000 00000 00020 00000 00000 (01)0000
0
10000 00(01)00 ?1101 10110 000?0 00201 33001 10110 11101
1?0?(01) 011(01)0 11(01)(02)(01) 101?1 11101 11010 10000
11030 011?3 01110 02200 0(01)012 21113 32220 10010 11111
11112 00000 02010 01020 01010 00000 00010 1
Torvosaurus
00100 000?1 00120 00000 010?? ????1 20011 0???? ???11
00100 0001? ??000 1?02? 01?0? 00??? ????? ????? ?????
????? ????? ???00 000?? ?0??? ????? ????? ??000 0??00
01100 00000 01110 01000 00002 00010 00000 00000 00?01
10103 0000? 10000 0?000 ???01 00?00 10101 00000 ?????
?00?1 ?0?10 10001 10100 00010 ????? ??000 10100 00?0?
????0 00?10 ?0100 10001 10000 00000 11011 00000 00002
01000 000(01)0 00011 20012 20220 10010 00001 11101 00001
00001 00000 0?000 10??? ????? ?
Troodontidae
10010 0?101 00001 (12)1100 00001 00001 01010 01?10 00100
10000 00010 0010? 00020 00002 1110? ??101 21000 10111
11111 00100 11100 0(01)001 10?1? 0111? ?0100 11110 12112
01111 00000 10100 11111 01100 01111 10110 01000 00001
20004 20001 01010 13101 30111 10111 11000 01010 ????1
01000 00111 10001 10001 000?0 00211 33001 10(01)00 11101
11000 01000 01010 111?1 10000 11(01)00 10011 01000 011?0
00111 02200 01012 21112 3?220 10010 11011 1111? 0??01
02110 10010 11112 (03)1121 11111 1
Tyrannosauridae
00010 01100 00001 (12)1000 00001 40101 10(12)10 01110
01111 0100(01) 00(01)11 01101 10020 0000(02) 11100 10101
21100 10001 10001 00000 01100 00001 1(01)111 (12)1101
00100 00000 02100 01110 00000 10100 01101 00002 00010
10110 00000 00101 2010(34) (01)0101 00001 12110 00001
10110 11000 01010 00201 00000 11110 00000 00000 00010
(01)0111 33201 10110 11101 1?010 01000 02220 11101 10011
00000 11011 01000 01103 02110 02100 11012 21112 30220
20010 11012 11112 01001 02110 00010 11102 (03)1121 00000
0
REFERENCES
ACCARIE, H.; BEAUDOIN, B.; DEJAX, J.; FRIÈS, G.; MICHARD J.-G. &
TAQUET, P. (1995) - Découverte d'un dinosaure théropode
nouveau (Genusaurus sisteronis n.g., n.sp.) dans l'Albien marin de Sisteron (Alpes de Haute-Provence, France) et extension au Crétacé inférieur de la lignée cératosaurienne. C.R.S.
Acad. Sci. Paris, 212(2): 327-334.
ALCOBER, O.; SERENO, P.C.; LARSSON, H.C.E.; MARTINEZ, R. &
VARRICCHIO, D.J. (1998) - A Late Cretaceous carcharodontosaurid (Theropoda: Allosauroidea) from Argentina. J. Vertebr.
Paleontol., 18 (3): 23A
ANDREWS, C.W. (1921) - On some remains of a theropodous dinosaur from Lower Lias of Barrow-on-Soar. Ann. Mag. Nat. Hist.,
8(9): 570-576.
BAKKER, R.T. (1986) - The Dinosaur Heresies: New Theories Unlocking the Mystery of the Dinosaurs and their Extinction. Wm.
Morrow & Co., New York, 481 pp.
BAKKER, R.T.; WILLIAMS, M. & CURRIE, P.J. (1988) - Nanotyrannus, a new genus of pygmy tyrannosaur, from the latest Cretaceous of Montana. Hunteria, 1(5): 1-30
BAKKER, R.T.; KRALIS, D.; SEIGWARTH, J. & FILLA, J. (1992) Edmarka rex, a new gigantic theropod dinosaur from the
middle Morrison Formation, Late Jurassic of the Como Bluff
outcrop region, with comments on the evolution of the chest
region and shoulder in theropods and birds, and a discussion
of the five cycles of origin and extinction among giant dinosaurian predators. Hunteria, 2(9): 1-24.
BAKKER, R.T. (1998) - Brontosaur killers: late jurassic allosaurids
as sabre-tooth cat analogues. Gaia, 15: this volume.
BARSBOLD, R. (1976a) - On a new Late Cretaceous family of small
theropods (Oviraptoridae fam. n.) of Mongolia. Dokl. Akad.
Nauk S.S.S.R., 226: 685-688. (in Russian)
BARSBOLD, R. (1976b) - On the evolution and systematics of the
late Mesozoic dinosaurs. Sovm. Sov.-Mong. Paleontol. Eksped. Trudy, 3: 68-75. (in Russian)
BARSBOLD, R. (1983) - Carnivorous dinosaurs from the Cretaceous of Mongolia. Sovm. Sov.-Mong. Paleontol. Eksped.
Trudy, 19: 1-117. [In Russian]
BENEDETTO, J.L. (1973) - Herrerasauridae, nueva familia de saurisquios triassicos. Ameghiniana, 10: 89-102.
BONAPARTE, J.F. (1979) - Dinosaurs: a Jurassic assemblage from
Patagonia. Science, 205: 1377-1379.
Yangchuanosaurus
00010 0?010 00000 11001 1011? 20011 20110 00??? ???11
01?00 0001? 10102 00020 0?011 ????? ????? ?01?? ?0???
????? ????? 0??00 00001 10100 20001 ?0100 00000 01100
01110 00000 11110 01001 00002 ??010 ?10?0 00001 0001?
?0003 0000? 00000 0???? 00?01 ?0??? ????1 ????? ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????0 10101 10000 00000 10011 00000 01014
11010 01010 00011 ?0012 20220 200?? 1???2 1110? ?1001
01011 000?? ????? ????? ????? ?
Deltadromeus
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? 0???? ?201? ????1 ??110 ????0 01010 ?????
????? ????? 1???? 1?0?? ????? ????? ????? ????? ?????
????? ????? ????? ????? ????? ????? ????? ????0 ?????
????? ???0? ???12 ???12 21220 10010 ????? ????? ?????
????? ????? ????? ?0??? ????? ?
"Megalosaurus" hesperis
00000 00??? 00110 10000 0?00? ????? ????? ????? ?????
????? ????? ????? ????? ????? ?0??? ????? ????? ?????
????? ????? ???00 000?? ??00? ????? ????? 00000 0??00
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????? ????? ????? ?
Unenlagia
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????? ????? ????? ????? ????? ????? ?????
????? ????? ????? ????? ????? ????? ?10?0 10?00 00?01
?0??4 ???0? ????? ????? ????1 01??? 11000 1???? ?????
57
T.R. HOLTZ, JR.
CHURE, D.J. & MADSEN, J.H. (1996) - On the presence of furculae
in some non-maniraptoran theropods. J. Vertebr. Paleontol.,
16: 573-577.
BONAPARTE, J.F. (1985) - A horned Cretaceous carnosaur from
Patagonia. Natl. Geogr. Res., 1: 149-151.
BONAPARTE, J.F. & NOVAS, F.E. (1985) - Abelisaurus comahuensis, n.g. et n.sp., Carnosauria del Cretácio Tardío de Patagonia. Ameghiniana, 21: 259-265.
CHURE, D.J. (1998) - On the orbit of theropod dinosaurs. Gaia, 15:
this volume.
BONAPARTE, J.F. (1991) - Los vertebrados fósiles de la Formación
Río Colorado, de la ciudad de Neuquén y cercanías, Cretácico Superior, Argentina. Rev. Museo Argent. Cienc. Nat. "B.
Rivadavia", Paleontol., 4: 17-123.
CHURE, D.J. & MADSEN, J.H. (1998) - An unusual braincase
(?Stokesosaurus clevelandi) from the Cleveland-Lloyd Dinosaur Quarry, Utah (Morrison Formation; Late Jurassic). J.
Vertebr. Paleontol., 18: 115-125.
BONAPARTE, J.F. & PUMARES, J.A. (1995) - Notas sobre el primer
cráneo de Riojasaurus incertus (Dinosauria, Prosauropoda,
Melanorosauridae) del Triásico superior de La Rioja, Argentina. Ameghiniana, 32: 341-349.
CLARK, J.M.; PERLE A. & NORELL, M.A. (1994) - The skull of Erlicosaurus andrewsi, a Late Cretaceous "segnosaur" (Theropoda: Therizinosauridae) from Mongolia. Am. Museum Novit.,
3115: 1-39.
BONAPARTE, J.F.; NOVAS, F.E. & CORIA, R.A. (1990) - Carnotaurus sastrei Bonaparte, the horned, lightly built carnosaur from
the Middle Cretaceous of Patagonia. Nat. Hist. Museum Los
Angeles County, Contrib. Sci. 416: 1-41.
COLBERT, E.H. & RUSSELL, D.A. (1969) - The small Cretaceous dinosaur Dromaeosaurus. Am. Museum Novit., 2380: 1-19.
COLBERT, E.H. (1989) - The Triassic dinosaur Coelophysis. Bull.
Museum N. Arizona, 57: 1-160.
BREMER, K. (1988) - The limits of amino acid sequence data in angiosperm phylogenetic reconstruction. Evolution, 42: 795803.
COLLESS, D. H. (1980) - Congruence between morphometric and
allozyme data for Menidia species: a reappraisal. Syst. Zool.,
29: 288-299.
BUFFETAUT, E.; SUTEETHORN, V.; MARTIN, V.; TONG, H.; CHAIMANEE, Y. & TRIAMWICHANON, S. (1995) - New dinosaur discoveries in Thailand. Int. Conf. Geol., Geotech., Min. Res.
Indochina, 1995: 157-161.
COPE, E.D. (1866) - Discovery of a gigantic dinosaur in the Cretaceous of New Jersey. Proc. Acad. Nat. Sci. Philadelphia, 18:
275-279.
BUFFETAUT, E., SUTEETHORN, V. & TONG, H. (1996) - The earliest
known tyrannosaurid from the Lower Cretaceous of Thailand.
Nature, 381: 689-691.
COPE, E.D. (1871) - On the homologies of some of the cranial bones of the Reptilia, and on the systematic arrangement of the
th
class. Proc. Am. Ass. Adv. Sci., 19 Mtg., Troy, 1870: 194247.
BUSBEY, A.B. (1995) - The structural consequences of skull flattening in crocodilians, in THOMASON, J.J. (Ed.), Functional
Morphology in Vertebrate Paleontology, Cambridge Univ.
Press, Cambridge, pp. 173-192.
CORIA, R.A. & SALGADO, L. (1995) - A new giant carnivorous dinosaur from the Cretaceous of Patagonia. Nature, 277: 224-226.
CORIA, R.A. & SALGADO, L. (1998) - A basal Abelisauria, NOVAS,
1992 (Therapoda-Ceratosauria) from the cretaceous of Patagonia, Argentina. Gaia, 15: this volume.
CAMP, C.L. (1936) - A new type of small bipedal dinosaur from the
Navajo Sandstone of Arizona. Univ. California Publ. Geol.
Sci., 24: 39-56.
CURRIE, P.J. (1985) - Cranial anatomy of Stenonychosaurus inequalis (Saurischia, Theropoda) and its bearing on the origin of
birds. Can. J. Earth Sci., 22: 1643-1658.
CARPENTER, K.; HIRSCH, K.F. & HORNER, J.R. (1994) - Dinosaur
Eggs and Babies. Cambridge Univ. Press, Cambridge, 372
pp.
CURRIE, P.J. & ZHAO, X.-J. (1993a) - A new carnosaur (Dinosauria, Theropoda) from the Jurassic of Xinjiang, People's Republic of China. Can. J. Earth Sci., 30: 2037-2081.
CARPENTER, K. (1997) - A giant coelophysoid (Ceratosauria) theropod from the Upper Triassic of New Mexico, USA. N. Jb.
Geol. Paläont., Abhandl., 205: 189-208.
CURRIE, P.J. & ZHAO, X.-J. (1993b) - A new troodontid (Dinosauria, Theropoda) braincase from the Dinosaur Park Formation
(Campanian) of Alberta. Can. J. Earth Sci., 30: 2231-2247.
CHARIG, A.J. & MILNER, A.C. (1997) - Baryonyx walkeri, a fish-eating dinosaur from the Wealden of Surrey. Bull. Nat. Hist.
Museum, Geol. Ser., 53: 11-70.
CURRIE, P.J.; GODFREY, S.J. & NESSOV, L. (1993) - New caenagnathid (Dinosauria: Theropoda) specimens from the Upper
Cretaceous of North America and Asia. Can. J. Earth Sci., 30:
2255-2272.
CHATTERJEE, S. (1988) - Functional significance of the semilunate
carpal in archosaurs and birds. J. Vertebr. Paleontol., 8 (3):
11A.
CURRIE, P.J. (1995) - New information on the anatomy and relationships of Dromaeosaurus albertensis (Dinosauria: Theropoda). J. Vertebr. Paleontol., 15: 576-591.
CHATTERJEE, S. (1997) - The beginnings of avian flight, in WOLBERG, D.L.; STUMP, E. & ROSENBERG, G.D. (Eds.), Dinofest
International: Proceedings of a Symposium Held at Arizona
State University, Academy of Natural Sciences, Philadelphia,
pp. 311-335.
CURRIE, P.J. (1997) - Theropods, in FARLOW, J.O. & BRETT-SURMAN, M. K. (Eds.), The Complete Dinosaur, Indiana Univ.
Press, Bloomington, pp. 216-233.
CHEN, P.-J.; DONG, Z.-M. & ZHEN, S.-N. (1998) - An exceptionally
well-preserved theropod dinosaur from the Yixian Formation
of China. Nature, 391: 147-152.
CURRIE, P.J. & CARPENTER, K. (in press) - A new specimen of
Acrocanthosaurus atokensis from the Lower Cretaceous
Antlers Formation (Lower Cretaceous, Aptian) of Oklahoma,
USA. Geodiversitas.
CHIAPPE, L.M. & CALVO, J.O. (1994) - Neuquenornis volans, a
new Late Cretaceous bird (Enantiornithes: Avisauridae) from
Patagonia, Argentina. J. Vertebr. Paleontol., 14: 230-246.
DAL SASSO,
C. & SIGNORE, M. (1998) - Exceptional soft-tissue preservation in a theropod dinosaur from Italy. Nature, 392: 383387.
CHIAPPE, L.M.; NORELL, M.A. & CLARK, J.M. (1996) - Phylogenetic
position of Mononykus (Aves: Alvarezsauridae) from the Late
Cretaceous of the Gobi Desert. Mem. Queensland Museum,
39: 557-582.
DENTON, R.K., JR. (1990) - A revision of the theropod Dryptosaurus (Laelaps) aquilunguis (COPE, 1869). J. Vertebr. Paleontol., 9(3): 20A.
CHIAPPE, L.M.; NORELL, M.A. & CLARK, J.M. (1998) - The skull of a
relative of the stem-group bird Mononykus. Nature, 392: 275278.
DEPÉRET, C. & SAVORNIN, J. (1927) - Sur la découverte d'une faune de vertébrés albiens à Timimoun (Sahara occidental). C.R.
Acad. Sci. Paris, 181: 1108-1111.
CHURE, D.J. (1994) - Koparion douglassi, a new dinosaur from the
Morrison Formation (Upper Jurassic) of Dinosaur National
Monument; the oldest troodontid (Theropoda: Maniraptora).
BYU Geol. Stud., 40: 11-15.
DONG, Z.-M.; CHANG, Y.; LI, X. & ZHOU, S.-W. (1978) - Note on a
new carnosaur (Yangchuanosaurus shangyouensis gen. et
sp. nov.) from the Jurassic of Yangchuan District, Szechuan
Province. Kexue Tongbao, 23: 298-302. (in Chinese)
58
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
D
H
, J.D. (1998) - A reanalysis of Acrocanthosaurus atoken, its phylogenetic status, and paleobiogeographic implications, based on a new specimen from Texas. Bull. New Mexico Mus. Nat. Hist. Sci., 13: 1-75.
H
, D.M. (1998) - Skull and tooth morfology as indicators
of niche partitioning in sympatric Morrison Formation theropods. Gaia, 15: this volume.
H
, T.R., J . (1994) - The phylogenetic position of the Tyrannosauridae: implications for theropod systematics. J. Paleontol., 68: 1100-1117.
, T.R., J . (1995a) - A new phylogeny of the Theropoda. J.
H
Vertebr. Paleontol., 15(3): 35A.
H
, T.R., J . (1995b) - The arctometatarsalian pes, an unusual structure of the metatarsus of Cretaceous Theropoda (Dinosauria: Saurischia). J. Vertebr. Paleontol., 14: 480-519.
H
, T.R., J . & P
, K. (1995) - Definition and diagnosis of
Theropoda and related taxa. J. Vertebr. Paleontol., 15(3):
35A.
H
, T.R., J . (1996a) - Phylogenetic analysis of the nonavian
tetanurine dinosaurs (Saurischia: Theropoda). J. Vertebr. Paleontol., 16(3): 42A.
H
, T.R., J . (1996b) - Phylogenetic taxonomy of the Coelurosauria (Dinosauria: Theropoda). J. Paleontol., 70: 536-538.
H
, T.R., J . & B
-S
, M.K. (1997) - The taxonomy
and systematics of the dinosaurs, in F
, J.O. & B
S
, M.K. (Eds.), The Complete Dinosaur, Indiana Univ.
Press, Bloomington, pp. 92-106.
H
, T.R., J ; B
, D.L. & C
, C.L. (1998) - Denticle morphometrics and a possibly omnivorous feeding habit
for the theropod dinosaur Troodon. Gaia, 15: this volume.
H , S. (1993) - A new theropod (Dilophosaurus sinensis sp. nov.)
from Yunnan, China. Vertebrata PalAsiatica, 31: 65-69. (In
Chinese)
H
, F. (1914) - Das natürliche System der Saurischia. Abh.
Mineral. Geol. Palaeontol. B1914: 154-158.
H
, F. (1920) - Bemerkungen zur Systematik und Stammesgeschichte einger Reptilien. Z. Indukt. Abstamm. Vereb.
Lehre, 24: 162-166.
H
, F. (1934) - Ein neuer Coelurosaurier in der thüringischen
Trias. Palaeontol. Z., 16: 145-170.
H , S.; M
, D.M. & B
, M.J. (1996) - The first European allosaurid dinosaur (Lower Cretaceous, Wealden
Group, England). N. Jb. Geol. Paläontol., Monat., 10: 635644.
J
, W. (1920) - Über Elaphrosaurus bambergi und die Megalosaurier aus den Tendaguru-Schichten Deutsch-Ostafrikas. Sitzungsber. Ges. Naturforsch. Freunde Berlin, 1920:
225-235.
J , Q. & J , S.-A. (1996) - On discovery of the earliest bird fossil in
China and the origin of birds. Chinese Geol., 233: 30-33. [In
Chinese]
J Q. & J , S.-A. (1997) - Protarchaeopteryx, a new genus of Archaeopterygidae in China. Chinese Geol., 238: 38-41. [In Chinese]
J , Q.; C
, P.J.; N
, M.A. & J , S.-A. (1998) - Two feathered dinosaurs from northeastern China. Nature, 393: 753761.
K
, A.A. & R
, A.S. (1996) - A new family of Maniraptora
(Dinosauria: Saurischia) from the Late Cretaceous of Mongolia. Paleontologich. Zhur., 1996(4): 85-94. (in Russian).
K
, S.M. (1976) - Braincase structure in the carnosaur Itemirus n. gen. and some aspects of the cranial anatomy of dinosaurs. Paleontologich. Zhur., 1976(3): 127-137.
L
, L.M. (1917) - The Cretaceous theropodous dinosaur Gorgosaurus. Mem. Geol. Surv. Can., 100: 1-84.
L , C. (1758) - Systema Natura per Regina Tria Naturae, Se-
, Z.-M. (1979) - The Cretaceous dinosaur fossils in southern
,
na. IVPP and Nanjing Geol. Paleontol. Inst. Sci. Press, Beijing, pp. 342-350. (in Chinese)
, Z.-M.; Z , S.-W. & Z
, Y.-H. (1983) - The dinosaurian remains from Sichuan Basin, China. Palaeontol. Sinica,
162 (C23): 1-145. (in Chinese)
, Z.-M. & T , Z.-L. (1985) - A new mid-Jurassic theropod
(Gasosaurus constructus gen. et sp. nov.) from Dashanpu, Zigong, Sichuan Province, China. Vertebrata PalAsiatica, 23:
77-83. (in Chinese)
, M.J.; O
, R.G.; S
, J.F. & P
, J.D.
(1992) - Phylogenetic relationships of dipsacales based on
rbcL sequences. Ann. Missouri Botan. Garden, 79: 333-345.
, A. & W
, P. (1993) - Skull of Archaeornithoides from the Upper Cretaceous of Mongolia. Am. J. Sci.,
293A: 235-252.
, T. (1998) - AutoDecay, version 4.0 (program distributed by the author). Dept. of Botany, Stockholm Univ.,
Stockholm.
, S.E. & M
, A.R. (1994) - Middle Jurassic microvertebrate assemblages from the British Isles, in F
, N.C. &
S , H.-D. (Eds.), In the Shadow of the Dinosaurs: Early Mesozoic Tetrapods, Cambridge Univ. Press, Cambridge, pp.
303-321.
, A. (1996) - The Origin and Evolution of Birds. Yale
Univ. Press, New Haven, 420 pp.
, C.A., S
, S.D., C
, L.M. & K
, D.W.
(1998) - The theropod ancestry of birds: new evidence from
the Late Cretaceous of Madagascar. Science, 279: 19151919.
, E. (1913) - Die neuesten Dinosaurierfunde in der
schwäbischen Trias. Naturwiss., 45: 1097-1100.
, P.M. & J
, J.A. (1979) - A new large theropod dinosaur from the Upper Jurassic of Colorado. BYU Geol. Stud.,
26: 1-12.
, Y.H. (1992) - Yangchuanosaurus hepingensis - a new species of carnosaur from Zigong, Sichuan. Vertebrata PalAsiatica, 30: 313-324. (in Chinese)
, J.A. & P
, K. (1985) - Phylogenetic, functional,
and aerodynamic analyses of the origin of birds and their
flight, in H
, M.K.; O
, J.H.; V , G. & W
, P. (Eds.), The Beginnings of Birds: Proceedings of the International
Archaeopteryx
Conference
Eichstätt
1984,
Freunde des Jura-Museums Eichstätt, pp. 185-197.
, J.A. (1986) - Saurischian monophyly and the origin of
birds, in P
, K. (Ed.), The Origin of Birds and the Evolution of Flight, Mem. Calif. Acad. Sci., 8: 1-40.
, C.W. (1920) - Osteology of the carnivorous Dinosauria
in the United States National Museum, with special reference
to the genera Antrodemus (Allosaurus) and Ceratosaurus.
Bull. U.S. Nat. Museum, 110: 1-154.
, C.W. (1924) - On Troodon validus, an orthopodous dinosaur from the Belly River Cretaceous of Alberta, Can. Bull.
Dept. Geol., Univ. Alberta, 1: 1-43.
, F.M.; A
, F.P.; O , J.G.; H
, J.;
V , P.; T
, J. & H
, Z. (1995) - A Triassic, Jurassic and Cretaceous time scale, in B
, W.A.; K ,
D.V.; A
, M.-P. & H
, J. (Eds.), Geochronology,
Time Scales and Global Stratigraphic Correlation, SEPM
Special Publication, 54: 95-126.
, L. & B
, W.E. (1998) - A comprehensive phylogenetic study of amiid fishes (Amiidae) based on comparative skeletal anatomy. An empirical search for interconnected
patterns in Natural History. Mem. Soc. Vertebr. Paleontol., 4:
1-690.
, W.R. & H
, W.J. (1994) - A crested theropod dinosaur from Antarctica. Science, 264: 828-830.
ARRIS
ONG
sis
China in Mesozoic and Cenozoic Red Beds in Southern Chi-
D
ONG
D
HOU
ONG
D
ANG
ONOGHUE
E
LMSTEAD
LZANOWSKI
E
ENDERSON
HANG
MITH
ALMER
ELLNHOFER
RIKSSON
E
VANS
ILNER
RASER
UES
F
ORSTER
F
AMPSON
HIAPPE
ALTON
G
RAUSE
STROM
IOHL
VAN
GTERBERG
EEN
HIERRY
GG
G
RANDE
H
AMMER
R
R
R
RETT
URMAN
RINKMAN
ARTILL
I
I
I,
I
I
ARDENBOL
UANG
ERGGREN
UBRY
OLTZ
RETT
HANDLER
ARKER
ANENSCH
ILMORE
RADSTEIN
R
UTT
ELLNHO-
ILMORE
G
OLTZ
ADIAN
UENE
ADIAN
G
R
UENE
AUTHIER
G
OLTZ
UENE
FER
G
R
U
ADIAN
ECHT
OLTZ
OLTZ
ENSEN
AUTHIER
R
ARLOW
AO
G
OLTZ
URMAN
RAAS
G
R
OLTZ
EDUCCIA
F
OLTZ
ENT
ARDENBOL
URRIE
ARHU
EMIS
ORELL
I
AUTIAN
URZANOV
AMBE
ICKERSON
INNÉ
cundum Classes, Ordines, Genera, Species cum Characte-
59
T.R. HOLTZ, JR.
risbus,
Differentiis,
Synonymis,
Locis.
reformata, Tomus I: Regnum Animale.
Editio
NORELL, M.A. & MAKOVICKY, P.J. (1997) - Important features of
the dromaeosaur skeleton: information from a new specimen.
Am. Museum Novit., 3215: 1-28.
decima,
Laurentii Salvii, Hol-
NOVAS, F.E. (1992) - La evolucion de los dinosaurios carnivoros,
in SANZ, J.L. & BUSCALIONI, A.D. (Eds.), Los Dinosaurios y su
Entorno Biotico, Inst. "Juan de Valdes" Excmo. Ayuntamiento
de Cuenca, Cuenca, pp. 125-163.
miae, 824 pp.
MADDISON, W.P. & MADDISON, D.R. (1997) - MacClade: Analysis
of phylogeny and character evolution. Version 3.07. Sinauer
Associates, Sunderland, Massachusetts.
NOVAS, F.E. (1994) - New information on the systematics and
postcranial skeleton of Herrerasaurus ischigualastensis (Theropoda: Herrerasauridae) from the Ischigualasto Formation
(Upper Triassic) of Argentina. J. Vertebr. Paleontol., 13: 400423.
MADSEN, J.H. (1976) - A second new theropod from the Late Jurassic of east Central Utah. Utah Geol., 3: 51-60.
MAKOVICKY, P.J. (1997) - A new small theropod from the Morrison
Formation of Como Bluff, Wyoming. J. Vertebr. Paleontol., 17:
755-757.
NOVAS, F.E. (1996) - Alvarezsauridae, Cretaceous basal birds
from Patagonia and Mongolia. Mem. Queensland Museum,
39: 675-702.
MAKOVICKY, P.J. & CURRIE, P.J. (1998) - The presence of a furcula in tyrannosaurid theropods, and its phylogenetic and functional implications. J. Vertebr. Paleontol., 18: 143-149.
NOVAS, F.E. (1997a) - Anatomy of Patagonykus puertai (Theropoda, Avialae, Alvarezsauridae), from the Late Cretaceous of
Patagonia. J. Vertebr. Paleontol., 17: 137-166.
MAKOVICKY, P.J. & NORELL, M.A. (1998) - A partial ornithomimid
braincase from Ukhaa Tolgod (Upper Cretaceous, Mongolia).
Am. Museum Novit., 3247: 1-16.
NOVAS, F.E. (1997b) - Herrerasauridae, in CURRIE, P.J. & PADIAN,
K. (Eds.), Encyclopedia of Dinosaurs, Academic Press, San
Diego, pp. 303-311.
MAKOVICKY, P.J. & SUES, H.-D. (1998) - Anatomy and phylogenetic relationships of the theropod dinosaur Microvenator celer
from the Lower Cretaceous of Montana. Am. Museum Novit.,
3240: 1-27.
NOVAS, F.E. (1997c) - Abelisauridae, in CURRIE, P.J. & PADIAN, K.
(Eds.), Encyclopedia of Dinosaurs, Academic Press, San Diego, pp. 1-2.
MARSH, O.C. (1877) - Notice of some new dinosaurian reptiles
from the Jurassic Formation. Am. J. Sci., 14(3): 514-516.
NOVAS, F.E. & PUERTA, P.F. (1997) - New evidence concerning
avian origins from the Late Cretaceous of Patagonia. Nature,
387: 390-392.
MARSH, O.C. (1879a) - Principle characters of the American Jurassic dinosaurs. Part II. Am. J. Sci., 17(3): 86-92.
NOVAS, F.E. (1998) - Megaraptor namunhuaiquii, gen. et sp. nov.,
a large-clawed, Late Cretaceous theropod from Patagonia. J.
Vertebr. Paleontol., 18: 4-9.
MARSH, O.C. (1879b) - Notice of new Jurassic reptiles (Coeluria).
Am. J. Sci., 21(3): 339-340.
MARSH, O.C. (1881) - Classification of the Dinosauria. Am. J. Sci.,
23(3): 81-86.
NOVAS, F.E. & POL, D. (in press) - Alvarezsaurid relationships reconsidered, in CHIAPPE, L.M. & WITMER, L.M. (Eds.), Mesozoic Birds: Above the Heads of Dinosaurs. Univ. California
Press, Berkeley.
MARSH, O.C. (1884) - Principle characters of American Jurassic
dinosaurs. Part VIII: the order Theropoda. Am. J. Sci., 27:
329-341.
OLSHEVSKY, G. (1995) - African dinosaur discoveries (Letter). Science, 267: 1750.
MARSH, O.C. (1890) - Description of new dinosaurian reptiles. Am.
J. Sci., 39(3): 81-86.
OSBORN, H.F. (1903) - Ornitholestes hermanni, a new compsognathoid dinosaur from the Upper Jurassic. Bull. Am. Museum
Nat. Hist., 19: 281-296.
MARTILL, D.M.; CRUICKSHANK, A.R.I.; FREY, E.; SMALL, P.G. &
CLARKE, M. (1996) - A new crested maniraptoran dinosaur
from the Santana Formation (Lower Cretaceous) of Brazil. J.
Geol. Soc. London, 153: 5-8.
OSBORN, H.F. (1906) - Tyrannosaurus, Upper Cretaceous carnivorous dinosaur (second comm.). Bull. Am. Museum Nat.
Hist., 22: 281-296.
MARTIN, L.D. (1997) - The difference between dinosaurs and birds
as applied to Mononykus, in WOLBERG, D.L.; STUMP, E. & ROSENBERG, G.D. (Eds.), Dinofest International: Proceedings of
a Symposium Held at Arizona State University, Academy of
Natural Sciences, Philadelphia, pp. 337-343.
OSBORN, H.F. (1916) - Skeletal adaptations of Ornitholestes,
Struthiomimus, Tyrannosaurus. Bull. Am. Museum Nat. Hist.,
35: 733-771.
OSMÓLSKA, H. & BARSBOLD, R. (1990) - Ornithomimosauria, in
WEISHAMPEL, D.B.; DODSON, P. & OSMÓLSKA, H. (Eds.), The
Dinosauria, Univ. California Press, Berkeley, pp. 225-244.
MATTHEW, W.D. & BROWN, B. (1923) - Preliminary notices of skeletons and skulls of Deinodontidae, with notice of a new genus
from the Cretaceous of Alberta. Am. Museum Novit., 89: 1-9.
OSMÓLSKA, H. (1996) - An unusual theropod dinosaur from the
Late Cretaceous Nemegt Formation of Mongolia. Acta Palaeontol. Polonica, 41: 1-38.
METCALF, S.J. & WALKER, R.J. (1994) - A new Bathonian microvertebrate locality in the English Midlands, in FRASER, N.C. &
SUES, H.-D. (Eds.), In the Shadow of the Dinosaurs: Early Mesozoic Tetrapods, Cambridge Univ. Press, Cambridge, pp.
322-331.
MEYER, H. (1832) - Palaeologica zur Geschichte der Erde
ihrer Geshöpfe. Schmerber, Frankfurt an Mein, 540 pp.
OSTROM, J.H. (1969a) - A new theropod dinosaur from the Lower
Cretaceous of Montana. Postilla, 128: 1-17.
OSTROM, J.H. (1969b) - Osteology of Deinonychus antirrhopus,
an unusual theropod from the Lower Cretaceous of Montana.
Bull. Peabody Museum Nat. Hist., 30: 1-165.
und
MEYER, H. (1861) - Archaeopteryx lithographica (Vogel-Feder)
und Pterodactylus von Solnhofen. N. Jb. Min., Geol., Palaeont., 1857: 532-543.
OSTROM, J.H. (1970) - Stratigraphy and paleontology of the Cloverly Formation (Lower Cretaceous) of the Bighorn Basin
area, Wyoming and Montana. Bull. Peabody Museum Nat.
Hist., 35: 1-234.
MILES, C.A.; CARPENTER, K. & CLOWARD, K. (1998) - A new skeleton of Coelurus fragilis from the Morrison Formation of Wyoming. J. Vertebr. Paleontol., 18(3): 64A.
OSTROM, J.H. (1974) - Archaeopteryx and the origin of flight.
Quat. Rev. Biol., 49: 27-47.
MOLNAR, R.E. (1974) - A distinctive theropod dinosaur from the
Upper Cretaceous of Baja California (Mexico). J. Paleontol.,
48:1009-1017.
OSTROM, J.H. (1975a) - The origin of birds. Annu. Rev. Earth Planet. Sci., 3: 55-77.
MOLNAR, R.E. (1991) - The cranial morphology of
rex. Palaeontograph. Abt. A, 217: 137-176.
OSTROM, J.H. (1975b) - On the origin of Archaeopteryx and the
ancestry of birds. Cent. Natl. Rech. Sci., 218: 519-532.
Tyrannosaurus
60
A NEW PHYLOGENY OF THE CARNIVOROUS DINOSAURS
S
O
, J.H. (1976) - Archaeopteryx and the origin of birds. Biol.
., 8: 91-182.
O
, J.H. (1978) - The osteology of Compsognathus longipes
W
. Abhandl. Bayer. Staatssammlung Palaeontol. Hist.
Geol., 4: 73-118.
O
, J.H. (1995) - Wing biomechanics and the origin of bird
flight. N. Jb. Geol. Paläont., Abhandl., 195: 253-266.
O
J.H. (1997) - How bird flight might have come about, in
W
, D.L.; S
, E. & R
, G.D. (Eds.), DinoJ. Linn. Soc
AGNER
S
OSENBERG
S
fest International: Proceedings of a Symposium Held at Arizo-
, Academy of Natural Sciences,
Philadelphia, pp. 301-310.
, K. & C
, L.M. (1997) - Bird origins, in C
, P.J. &
P
, K. (Eds.), Encyclopedia of Dinosaurs, Academic
Press, San Diego, pp. 71-79.
, K. & C
, L.M. (1998) - The origin and early evolution
of birds. Biol. Rev., 73: 1-42.
, K.; H
, J.R. & H
, T.R., J . (1999) - Phylogenetic definitions and nomenclature of the major taxonomic
categories of the carnivorous Dinosauria (Theropoda). J. Vertebr. Paleontol., 19: 69-80.
, G.S. (1988) - Predatory Dinosaurs of the World. Simon &
Schuster, New York. 464 pp.
-M
, B.P.; S , J.L.; S
, J. & S , B. (1993) - A
theropod dinosaur from the Lower Cretaceous of southern
France. Rev. Paléobiol., 7: 173-188.
-M
, B.P.; S , J.L.; B
, A.D.; M
,
J.J.; O
, F. & R
-G
, D. (1994) - A unique
multitoothed ornithomimosaur dinosaur from the Lower Cretaceous of Spain. Nature, 370: 363-367.
A.; N
, M.A.; C
, L.M. & C
, J.M. (1993) Flightless bird from the Cretaceous of Mongolia. Nature, 362:
623-626.
A.; C
, L.M.; B
R.; C
, J.M. & N
,
M.A. (1994) - Skeletal morphology of Mononykus olecranus
(Theropoda: Avialae) from the Late Cretaceous of Mongolia.
Am. Museum Novit., 3105: 1-29.
, M.A. (1990) - Morphological variation in small theropods
and its meaning in systematics: evidence from Syntarsus rhodesiensis, in C
, K. & C
, P.J. (Eds.), Dinosaur
Systematics: Approaches and Perspectives, Cambridge
Univ. Press, Cambridge, pp. 91-105.
, O.W.M. (1995) - Zur systematischen Stellung der afrikanischen Theropoden Carcharodontosaurus Stromer 1931
und Bahariasaurus Stromer 1934. Berlin. Geowissensch.
Abhandl., E16: 357-375.
, O.W.M. & H
, A. (1998) - A review of european triassic theropods. Gaia, 15: this volume.
, T.H. & V
-R , P. (1994) - Neoceratopsians and ornithomimosaurs: dinosaurs of Gondwana origin? Res. Exploration, 10: 129-131.
, T.R. (1989) - A new species of the theropod dinosaur Syntarsus from the Early Jurassic Kayenta Formation of Arizona.
J. Vertebr. Paleontol., 9: 125-136.
, T.R. & G
, J. (1990) - Ceratosauria, in W
, D.B.; D
, P. & O
, H. (Eds.), The Dinosauria, Univ. California Press, Berkeley, pp. 151-168.
, D.A. (1970) - Tyrannosaurs from the Late Cretaceous of
western Canada. Nat. Museum Nat. Sci. Pub. Paleontol., 1: 134.
, D.A. & D , Z.-M. (1993a) - The affinities of a new theropod from the Alxa Desert, Inner Mongolia, People's Republic of China. Can. J. Earth Sci., 30: 2107-2127.
, D.A. & D , Z.-M. (1993b) - A nearly complete skeleton of a new troodontid dinosaur from the Early Cretaceous of
the Ordos Basin, Inner Mongolia, People's Republic of China.
Can. J. Earth Sci., 30: 2163-2173.
na
P
State
ADIAN
ERENO
University
HIAPPE
S
P
ADIAN
P
ADIAN
P
S
UTCHINSON
OLTZ
S
R
ÉREZ
P
ÉREZ
ORENO
ANZ
UDRE
ORENO
ANZ
USCALIONI
RTEGA
P
ERLE
P
ERLE
R
ASSKIN
ORELL
HIAPPE
S
R
S
AUHUT
R
ICH
S
R
S
ERENO
OWE
PEL
ADLEIR
ORELL
R
S
USSELL
R
USSELL
ARRICCHIO
UTHEIL
ADO
ARCOT
IDOR
ARSSON
AUHUT
ARRICCHIO
ILSON
MITH
S
TOVALL
S
ANGSTON
R
TERNBERG
S
TROMER
S
UES
ICH
S
AQUET
UES
S
EISHAM-
UTEETHORN
UFFETAUT
ONG
RIAMWICHANON
SMÓLSKA
S
USSELL
R
ARSSON
ILSON
URRIE
AUTHIER
ODSON
ECK
YON
OWE
R
IDOR
ERENO
UNGERBÜHLER
ICKERS
AROCHENE
AGWENE
ERENO
ORATALLA
AUHUT
R
UTHEIL
ILSON
AATH
ARPENTER
ARSSON
UTHEIL
YON
IGÉ
LARK
ILSON
ERENO
LARK
ARSBOLD
ONETTA
UES
UTMAN
HIAPPE
OGERS
OVAS
ERENO
AUL
P
ORSTER
ERENO
HIAPPE
UFFORD
OVAS
ERENO
URRIE
ADIAN
RAUSE
AVOAVY
ONOGHUE
ANDERSON
STROM,
ORSTER
ODSON
ANDERSON
STROM
TUMP
ITMER
ONNER
STROM
OLBERG
, S.D.; W
, L.M.; F
, C.A.; K
, D.W.;
O'C
, P.M.; D
, P. & R
, F. (1998) - Predatory dinosaur remains from Madagascar: implications for the
Cretaceous biogeography of Gondwana. Science, 280: 10481051.
, M.J. & D
, M.J. (1996) - The relationship
between homoplasy and confidence in a phylogenetic tree, in
S
, M.J. & H
, L. (Eds.), Homoplasy: The Recurrence of Similarity in Evolution, Academic Press, San Diego, pp. 67-89.
, P.C. & N
, F.E. (1992) - The complete skull and skeleton of an early dinosaur. Science, 258: 1137-1140.
, P.C.; F
, C.A.; R
, R.R. & M
, A.M.
(1993) - Primitive dinosaur skeleton from Argentina and the
early evolution of Dinosauria. Nature, 361: 64-66.
, P.C. & N
, F.E. (1994) - The skull and neck of the basal theropod Herrerasaurus ischigualastensis. J. Vertebr. Paleontol., 13: 451-476.
, P.C.; W
, J.A.; L
, H.C.E.; D
, D.B. &
S , H.-D. (1994) - Early Cretaceous dinosaurs from the Sahara. Science, 255: 845-848.
, P.C.; D
, D.B.; I
, M.; L
, H.C.E.;
L , G.H.; M
, P.M.; S
, C.A.; V
, D.J.
&W
, J.A. (1996) - Predatory dinosaurs from the Sahara
and Late Cretaceous faunal differentiation. Science, 272:
986-991.
, P.C. (1997) - The origin and evolution of dinosaurs.
Annu. Rev. Earth Planet. Sci., 25: 435-489.
, P.C. (1998) - A rationale for phylogenetic definitions,
with application to the higher-level taxonomy of Dinosauria. N.
Jb. Geol. Paläontol., Abhandl., 210: 41-83.
, P.C.; B , A.L.; D
, D.B.; G , B.; L
,
H.C.E.; L , G.H.; M
, J.D.; R
, O.W.M.;
S
, R.W.; S
, C.A.; V
, D.D.; W
, G.P.
&W
, J.A. (1998) - A long-snouted predatory dinosaur
from Africa and the evolution of spinosaurids. Science,
282:1298-1302.
, D.K. (1998) - A morphometric analysis of Allosaurus. J.
Vertebr. Paleontol., 18: 126-142.
, J.W. & L
, W., J . (1950) - Acrocanthosaurus
atokensis, a new genus and species of Lower Cretaceous
Theropoda from Oklahoma. Am. Midland Nat., 43: 696-728.
, R.M. (1940) - A toothless bird from the Cretaceous of
Alberta. J. Paleontol., 14: 81-85.
, E. (1915) - Wirbeltier-Reste der Baharije-Stufe (unterstes Cenoman). 3. Das Original des Theropoden Spinosaurus
aegyptiacus nov. gen. nov. spec. Abhandl. K. Bayer. Akad.
Wiss., Math.-Phys. Kl., 28: 1-32..
, H.-D. & T
, P. (1979) - A pachycephalosaurid dinosaur from Madagascar and a Laurasia-Gondwanaland connection in the Cretaceous. Nature, 279: 633-635.
, H.-D. (1997) - On Chirostenotes, a Late Cretaceous oviraptorosaurs (Dinosauria: Theropoda) from western North America. J. Vertebr. Paleontol., 17: 698-716.
, V.; B
, E.; M
, V.; C
, Y.;
T , H. th& T
, S. (1995) - Thai dinosaurs: an
update. 6 Symp. Mesozoic Terrestrial Ecosystems and Biota, Short Papers, pp. 133-136.
, D.L. (1993) - PAUP: Phylogenetic Analysis Using
Parsimony, version 3.1.1. Computer software and documentation distributed by Illinois Natural History Survey, Champaign, Indiana.
, R.A. (1984) - The avian relationships of Archaeopteryx, and the origin of birds. Zool. J. Linnean Soc., 82: 119158.
, A. (1861) - Neue Beiträge zur Kenntnis der urweltlichen
Fauna des lithographischen Schiefers. V. Compsognathus
longipes W
. Abhandl. Bayer. Akad. Wiss., 9: 30-38.
AMPSON
STROM
WOFFORD
ONG
T
HULBORN
ONG
W
AGNER
AGNER
61
ARTIN
HIMANEE