PROCEEDINGS or tue -
GONDWANAN
DINOSAUR SYMPOSIUM

BRISBANE MEMOIRS OF THE VOLUME 39
20 DECEMBER 1996 QUEENSLAND MUSEUM PART 3

KT EVENTS IN INDIA: IMPACT, RIFTING, VOLCANISM AND DINOSAUR
EXTINCTION

SANKAR CHATTERJEE & DHIRAJ K. RUDRA

Chatterjee, S. & Rudra, D.K. 1996 12 20: KT events in India: impact, rifting, volcanism and
dinosaur exticntion. Memoirs of the Queensland Museum 39(3): 489-532. Brisbane. ISSN
0079-8835.

For more than a decade, a number of impact sites have been linked to the mass extinction at
the KT (Cretaceous/Tertiary) boundary. The prime candidate today is the Chicxulub Crater
in Yucatán Peninsula, Mexico. Recently another potential KT impact scar — the Shiva Crater
— has been identified from subsurface data at the India-Seychelles rift margin. The crucial
evidence in support of this impact structure comes from the Bombay High field, a giant
offshore oil basin in India, and associated alkaline intrusives within the Deccan Traps. The
KT boundary age of the crater is inferred from its Deccan lava floor, Palaeocene age of the
overlying sediments, isotope dating (~65Ma) of presumed melt rocks, and the Carlsberg
rifting event (chron 29R) within the basin. Seismic reflection data and India-Seychelles plate
reconstruction at 65Ma reveal a buried oblong crater, 600km long, 450km wide and 12km
deep, carved through Deccan Traps and into underlying Precambrian granite. It represents
the largest impact structure of Phanerozoic age. The crater shows the morphology of a
complex impact structure and basin, with a distinct central uplift in the form of a series of
peaks, an annular trough and a slumped rim. The oblong shape of the crater and the
asymmetric distribution of fluid ejecta indicate oblique impact in a SW-NE trajectory. We
speculate that a 40km diameter meteorite crashed on the western continental shelf of India
around 65Ma, excavating the Shiva Crater, shattering the lithosphere and inducing the
India-Seychelles rifting. The crater appears to narrow in the form of a teardrop to the NE or
downrange where the ejecta melt rocks were emplaced radially outward by the impact shock.
The shape of the Shiva Crater and the asymmetric ejecta distribution mimic those of artificial
craters produced by oblique impacts in laboratory experiments. The synchrony and near-an-
tipodal positions of the Shiva and Chicxulub Craters may indicate two alternative modes of
their origin. Either, both craters originated from splitting of a larger diameter meteorite, or,
large impact on one side of the Earth produced a similar signature on the far side by axial
focusing of seismic waves. Since India was ground zero for both an impact and Deccan
volcanism, their causal relationships and biotic effects were assessed. It appears that Deccan
volcanism began 1Ma before the KT event and was not triggered by the impact. Its origin is
attributed to the Deccan-Reunion hotspot. The extensive areal distribution of Deccan Traps
is owing to intercanyon flows along the drainage of the Narmada, Godavari and the Cambay
basins. During the early stage of Deccan eruption, sauropods, theropods and ankylosaurs
flourished in India, but they died out suddenly at the KT impact boundary. Although both
impact and Deccan volcanism are hypothesized as contributing to the deleterious environ-
mental consequences leading to biotic crisis at the KT boundary, the impact is suggested as
having played the major role as the killing mechanism. [ ] /mpact, Cretaceous-Tertiary
boundary, India, dinosaur, extinction, volcanism.

S. Chatterjee, Museum of Texas Tech University, Lubbock, Texas 79409-3191, USA; D.K.
Rudra, Geology Unit, Indian Statistical Institute, 203 Barrackpore Trunk Road, Calcutta
700035, India; 7 May 1996.

Since its emergence and subsequent eruption of
diversity, life has shown a tenacious and wildly
successful hold on this planet. But the rich history
of life has been repeatedly punctuated by equally
awesome displays of its destruction. It is es-
timated that 99% of plant and animal life that
have ever lived on Earth are now extinct (Wilson,
1992). The history of life is replete with major
episodes of biotic catastrophes or mass extinc-
tions, when 50% or more of the unrelated species

died out fairly rapidly. All mass extinctions, how-
ever, have been followed by at least a partial
evolutionary recovery in which the number of
species on Earth has increased again.

There are five major episodes of mass extinc-
tions during the past 600 million years: Late
Ordovician (440Ma), Late Devonian (365Ma),
Late Permian (245Ma), Late Triassic (210Ma)
and Late Cretaceous (65Ma). Of these mass ex-
tinctions, the one that has captured the greatest

490

TOTAL EXTINCTION
VICTIMS

65 Ma KT BOUNDARY

e. (Pterosaurs)

(Ammonites)
(Nanavian Dinosaurs)

—_ q me

(Large Marine Reptiles)

MAASTRICHTIAN

attention of earth scientists has been the KT
(Cretaceous/Tertiary) extinction when the
dinosaurs and two-thirds of all marine animal
speices were wiped out. The sudden extinction of
dinosaurs has puzzled both scientists and public
for more than a century. Having survived for 160
million years, dinosaurs seemed indestructible.
Not only the dinosaurs died out during that rela-
tively brief period; all land animals weighing
more than 25kg disappeared from the planet. All
pterosaurs, plesiosaurs, mosasaurs as well as
several families of birds and marsupial mammals,
and hundreds of plants were also suddenly wiped
out at this time. The small calcareous plankters
that float at the ocean surface and the ammonites
and rudists from the depths also vanished. The
Earth was devastated. Life was ravaged by one of
the worst catastrophes.

There were survivors, of course, after the KT
disaster (Fig. 1). Neornithine birds, placental
mammals, crocodiles, turtles, lizards and snakes
all survived as groups — despite the extinction of
some species. From this catastrophe, oppor-
tunities arose forthe survivors. The KT extinction
had opened the door for the age of mammals and
the rise of birds and changed the course of evolu-
tionary history. What triggered this catastrophe
that led to such an unprecendented ecological
crisis? Over the years, many theories, some biz-
zare and some plausible, have been offered to
explain the mystery behind the extinction of
dinosaurs. There is no shortage of murder
suspects. Any explanation of the causes of biotic
crises must focus on finding agents of destruction
that affected environments, climates, ecology and
organisms.

(Crocodilas)

MEMOIRS OF THE QUEENSLAND MUSEUM

RECOVERY

ap pe

eae (Snakes)
(Turtles) (Lizards) (Birds)
< ^ . " ium. - {Mammals}

FIG. 1. Victims and survivors after the KT extinction. The primary victims were nonavian dinosaurs, pterosaurs,
large marine reptiles such as plesiosaurs and mosasaurs, and various invertebrates such as ammonites and rudists.
Lizards, snakes, turtles, crocodiles, birds and mammals endured this catastrophe and rebounded. Both birds and
mammals underwent explosive evolutions after this crisis.

By the end of the Cretaceous, harsh changes of
environments were taking place as a result of
plate movements, mountain buildings, volcanic
emissions and sea regressions. Exactly what
caused the biotic crisis remains highly controver-
sial. Currently two competing models have been
proposed to explain this apocalyptic disaster at
the KT boundary: meteorite impact hypothesis
and volcanic hypothesis. The impact theory pos-
tulates that the environments were lethally altered
or destroyed at the end of the Cretaceous by the
collision of a large meteorite leading to biotic
crisis. The volcanic theory argues that the pollu-
tion in the atmosphere and oceans by the massive
outpourings of Deccan flood basalt in India had
devastating effects on ecology.

In 1980, the Alvarez group proposed that the
KT extinction was caused by the impact of a
10km meteorite. This proposal has generated a
great deal of interest among scientists and the
public. But the key piece of evidence was still
missing. If a huge meteorite had indeed crashed
into the Earth, where was the crater? Critics sear-
ched for alternate explanation. The end of the
Cretaceous was also a time of massive continen-
tal flood basalt volcanism, especially the Deccan
Traps in India. Many palaeontologists believe
that such cataclysmic volcanism may have been
the culprit in the KT extinction (McLean, 1985;
Officer et al, 1987). Over the past 15 years,
exciting new insights have poured in from vir-
tually every branch of earth and planetary scien-
ces to understanding the effects of these
catastrophic events — impact and volcanic — on
earth's ecosphere and the evolution of life.
Recently the Chicxulub structure in the northern
coast of Yucatán Peninsula of Mexico has

KT EVENTS IN INDIA

emerged as a prime candidate for the KT impact
site (Hildebrand et al., 1991). In this paper we
describe another buried KT impact structure —
the Shiva Crater at the India-Seychelles rift mar-
gin, and its relevance to the Chicxulub structure,
Deccan volcanism and mass extinction (Chatter-
jee, 1992; Chatterjee & Rudra, 1993).

THE IMPACT MODEL

Like other planets in the solar system, the Earth
resides in a swarm of asteroids and comets. It is
now apparent that the Earth has been heavily
bombarded during its history by meteorites of
various sources, sizes and compositions (Clube &
Napier, 1982). The incontrovertible evidence for
large cosmic collisions is the occurrence of cir-
cular craters associated with considerable local
structural disturbance and shock metamorphism
(French & Short, 1968). Because of the dynamic
nature of the terrestrial lithosphere where such
forces as erosion, volcanism, deposition, orogeny
and plate tectonics constantly restructure the sur-
face, impact craters are often erased or obscured,
unlike the more static surfaces of the Moon, Mer-
cury and Mars. To date over 150 impact craters
have been recognized on the Earth’s surface and
the list is growing. They range in size from ap-
proximately 100m to 200km and in age from
Precambrian to Recent (Grieve, 1987; 1990;
Grieve et al., 1988). The spatial distribution indi-
cates concentrations in cratonic areas. Other
craters may be submerged under oceans and
remain inaccessible or undetected. Scientific in-
terest in the role of impact in geological and
biological evolution has been enhanced by
several developments in recent years. Among the
most prominent of these are the hypotheses of
Alvarez et al. (1980) concerning terminal
Cretaceous extinction and lunar and planetary
exploration by manned and unmanned
spacecraft. As interest in bombardment mounts,
previously unknown or cryptic impact sites are
recognized with increasing frequency. As a grim
reminder that the threat of impact on our planet is
a real possibility, the world’s attention was
focused during late July, 1994, on the spectacular
collision of comet Shoemaker-Levy 9 on the sur-
face of Jupiter, leaving scars the size of Earth on
the giant planet (Levy et al., 1995; Weissman,
1955).

Hypervelocity impacts can have a large range
of effects that depend on the strength and density
of the projectile and the nature of the target
material. The most obvious result of larger col-

491

lisions is seen in the spectrum of crater sizes and
morphologies. The recovery of meteorite frag-
ments and shock effects within or surrounding a
crater are the most persuasive evidence for an
impact origin, but when large craters are deeply
eroded or buried, the evidence of impact is
obscured or blurred. Such evidence may be iden-
tified indirectly from shock-metamorphic effects
on the target rock and ejecta components, as well
as distinctive geochemical signatures attributable
to a particular type of meteoritic projectile
(Grieve, 1990). These signatures may be
preserved locally near the impact site, or globally
at a particular stratigraphic level containing the
ejecta fallout. Together they may provide clues to
the nature of the target material and the impactor.
Impact craters of this obscure nature are the most
controversial and require additional information
for verification.

In 1980, the Alvarez group advanced a startling
theory to explain the sudden demise of dinosaurs
— the most succesful land animals ever to arise
on Earth. They discovered an abnormally high
concentration of iridium (about 30 times more
than the surrounding rocks) at the KT boundary
level of Gubbio, Italy. Soon a comparable iridium
anomaly was found globally at different KT
boundary sections (Orth, 1989). Since iridium is
a very rare element in the earth's crust, but fairly
abundant in chondritic meteorites, the Alvarez
team proposed that the iridium spike at the KT
boundary is cosmic in origin, implying the strike
of a large meteorite. There was enough iridium in
the KT boundary, they calculated, to equal a
10km-diameter asteroid. They proposed that this
giant asteroid crashed into the Earth with a
velocity of 90,000km/hour to cause the
worldwide catastrophic event. This impact lofted
so much debris into Earth's atmosphere as to
creater a ‘nuclear winter'that caused much of the
life on Earth to perish. A blackout of sun would
kill plants and destory the food chain. The global
distribution of the iridium layer was caused by the
impact and vaporization of the bolide. The impact
theory was reinforced by additional evidence
such as shocked quartz (Bohor et al., 1984; 1987;
Owen & Anders, 1988), stishovite (McHone et
al., 1989), micro-diamonds (Carlisle & Braman,
1991), impact glasses (Izett et al., 1990), os-
mimium isotope ratios (Turekian, 1982), Ni-rich
spinels (Robin et al., 1994), rhodium (Bekov et
al., 1988), carbon soots (Wolbach et al., 1988),
tsunami deposits (Bourgeois et al., 1988), and
extraterrestrial amino acids (Carlisle & Braman,
1993) in the KT boundary layer at different sites.

492

Among all this cumulative evidence, the shocked
quartz is a distinctive signature of impact event
as it can form at a force more than 10 gigapascal
(GPa) that travels through quartz-bearing grains
of the target rock to porduce microscopic shock
lamellae (Grieve, 1990). Pressures and tempera-
tures produced by a large body impact are much
greater than those generated by other geologic
processes, such as volcanic activity, mountain
building and earthquakes.

However, the strongest evidence in favor of the
KT impact event would be to locate a crater
marking the point of collision. Such a crater
should be 150km or more in diameter (Grieve,
1982). The search for an impact site of the right
age (~65Ma), and the right size (150km) has
continued, including reassessment of many enig-
matic structures. Now, after a decade-long search,
the Chicxulub structure on the Yucatán Peninsula
of Mexico (Hildebrand et al., 1991) and the Shiva
Crater at the India-Seychelles rift margin (Chat-
terjee, 1992) appear to be two potential can-
didates for the long-sought KT impact scar.

THE VOLCANIC MODEL

Although the impact hypothesis is very com-
pelling, not everybody believes that impacts
killed the dinosaurs and other organisms at the
KT boundary. Critics have advanced a volcanic
alternative. The end of the Cretaceous was also a
time of massive continental flood basalt vol-
canism, especially the Deccan Traps of India.
Recent radiometric dating suggests that the main
pulse of Deccan volcanism may have occured
close to the KT boundary at 65 million years ago
(Duncan & Pyle, 1988; Courtillot, 1990). Many
palaeontologists argue forcefully that such
cataclysmic Deccan volcanism may have been
the main contributing factor for the biotic crisis
at the KT boundary (Clemens, 1982; Officer et
al, 1987; Hallam, 1987; Keller, 1989; Stanley,
1987; Zinsmeister et al., 1989; Courtillot, 1990).
Other contemporary episodes of volcanism at the
KT boundary such as in Cameroon and the Coral
Sea have been linked to the KT event (Sutherland,
1994). The proponents of the volcanic model
argue that the KT extinction was neither global,
nor instantaneous, but occurred over an extended
period of time, because different organisms dis-
appeared at different levels at or near the KT
boundary. Such step-wise extinction pattern
could be best explained by prolonged emissions
of volcanic pollutants. Large amounts of iridium
have been discovered to be spewing from the

MEMOIRS OF THE QUEENSLAND MUSEUM

Hawaiian and Reunion volcanoes, suggesting
that the iridium anomaly at the KT boundary
could have also had a volcanic origin (Olmez et
al., 1986). There is no doubt that such a massive
volcanic outburst over an extended period would
have deleterious environmental consequences.
Proponents of the volcanic model claim that
many ofthe supposed impact signatures at the KT
boundary layer, such as iridium enrichment,
shocked quartz, microspherules, clay mineralogy
and carbon soot, could have volcanic explana-
tions (Officer, at al., 1987; Courtillot, 1990). The
impact proponents disagree. They point out that
the Deccan volcanism was not of an explosive
typeand could not account forthe global distribu-
tion of the iridium anomaly, shocked quartz and
tektites at the KT boundary layer (Alvarez, 1986;
Alvarez & Asaro, 1990). Moreover, the lamellar
features in quartz grains associated with ex-
plosive volcanism show curvatures contrary to
the planar and parallel lamellae in impact-related
shocked quartz recoved from the KT boundary
(Izett, 1990). The gradual extinction pattern seen
among organisms may be an artifact of preserva-
tion and poor sampling quality. Others argue that
because Deccan volcanism had little effect on the
diversity of local Indian biota, its catastrophic
role in global life is questionable (Prasad et al.,
1994).

SEARCH FOR THE KT IMPACT SITE

Since Alvarez et al (1980) presented their
geochemical evidence for an impact event at the
KT boundary, the search for the proposed impact
crater has continued. There are a number of
candidates for the KT impact site, none of which
are very compelling at present. The 35km Man-
son structure in north-central Iowa is such a can-
didate (French, 1984; Anderson & Hartung,
1988), but new work suggests that this crater is
older (~74mya) and played no role in the KT mass
extinction (Izett at al., 1993). Twin impact struc-
tures in the Kara Sea in the former USSR, the
Kara (diameter, 60km) and Ust-Kara (diameter,
25km), have been proposed as possible impact
sites (Koeberl et al., 1988), but recent
geochronologic data suggest that these structures
are also older than the KT boundary event
(Koeberl et al., 1990).

Even the general location of the KT impact,
whether continental or oceanic, remains con-
troversial. Trace element and isotopic studies of
the highly altered KT boundary-layer com-
ponents tend to support the oceanic impact

KT EVENTS IN INDIA

hypothesis (Gilmore et al., 1984; Hildebrand &
Boynton, 1990). On the other hand, the presence
of shocked quartz at several KT boundary sites
would indicate a continental site (Bohor at al.,
1987). The apparent contradiction can be recon-
ciled if the single impact occurred at a continental
margin involving both oceanic and continental
crust, or multiple impacts at different sites.

It was soon realized that the size and abundance
of the ejecta, such as shocked quartz grains and
tektites, may give some clues as to the location of
the impact crater. Bohor et al. (1987) and Izett
(1990) concluded that the largest sizes and
greatest abundance of shocked quartz grains in
KT boundary sediments occur in western North
America, suggesting that the impact occurred on
or near the continent. The discovery of tsunami
deposits at the KT boundary sections on the
Brazos River, Texas (Bourgeois et al., 1988), near
Braggs, Alabama, (Smit et al., 1994), in the
Caribbean (Hildebrand & Boynton, 1988), and
Deep-Sea Drilling Program holes 536 and 540 in
the southwestern Gulf of Mexico (Alvarez et al.,
1992) as well as the identification of tektites at
Beloc, Haiti (Izett et al., 1991; Maurrasse & Sen,
1991) and Arroyo el Mimbral, northeastern
Mexico (Smit et al., 1992), narrowed the search
further to the Caribbean region. Impact breccia
has been recovered from Albion Island of Belize,
near the Mexican border (Ocampo & Pope,
1994). At least four possible Caribbean sites have
been suggested, including the Colombian Basin,
western Cuba, Haiti and the Yucatan Peninsula.

Hildebrand & Boynton (1990) placed the KT
impact location in the Colombian Basin between
Colombia and Haiti on the basis of seismic data
and DSDP core samples, but the putative crater is
not only under water but buried under 1,000m of
sediment and is subject to other interpretations,
such as tectonic origin or a change in the thick-
ness of the oceanic crust. Bohor & Seitz (1990)
speculated that the impact site was near Cuba,
about 1,350km from the site proposed by Hil-
debrand and Boynton, on the basis of a boulder
bed interpreted as ejecta components, but the
boulder bed is found to be of local, weathering
origin and the Cuban site has been rejected (Dietz
& McHone, 1990). The Massif de la Hotte on the
southern peninsula of Haiti, a mountainous
region with Cretaceous sediments, has also been
proposed as the KT impact site (Maurrasse,
1990), but closer examination of the area clearly
indicates that it is not an impact site (Officer et
al., 1992).

493

The most promising KT impact site appears to
be the Chicxulub Crater on the northern margin
of the Yucatán Peninsula, Mexico (Penfield &
Camargo, 1982; Hilderbrand et al., 1991; 1995).
Itis a circular structure about 180km in diameter,
buried under 1,100 m of carbonate strata, extend-
ing out under the Gulf of Mexico, and defined by
magnetic and gravity anomalies (Fig. 2A).

THE CHICXULUB CRATER

The subsurface stratigraphy of the Chicxulub
structure is known primarily from petroleum ex-
ploration bore holes drilled by Pemex, the
Mexican national petroleum company, in the
1950s (Lopez-Ramos, 1975; Meyerhoff et al.
1994). Unfortunately, most of the critical core
samples were destroyed in a warehouse fire. At
present, samples of the Chicxulub structure are
limited; as a result, the subsurface stratigraphy is
open to various interpretations. In hindsight, it is
ironic that drilling and exploration were stopped
as soon as the andesitic bodies at a depth of
1500-2000m were encountered; these may have
provided the critical evidence for the impact.
Penfield & Camargo (1982) suspected an impact
origin for the Chicxulub Crater on the basis of
concentric geophysical anomalies with as-
sociated extrusive material such as andesitic
bodies. Recently located samples from the old
Pemex wells, including brecciated carbonates,
andesites and crystalline basement have been
studied extensively and inferred to support the
impact scenario for this site. For example, Hil-
debrand et al. (1991) reported shocked quartz
within Chicxulub breccias and documented
chemical and isotopic similarities between an-
desites and tektite deposits from the KT boundary
sections of Haiti and Mexico. These findings
indicate that the Chicxulub Crater may be the
source for the Haitian and Mexican tektites.
Kring & Boynton (1992) interpreted the rocks
initially thought to be andesites as probable im-
pact melts, whereas Blum et al (1993) found an
isotopic match between the Haiti glass and the
Chicxulub melt. Subsequently, Sharpton et al.
(1992) recognized that the breccia above the melt
rock is suevite breccia, a distinct signature of an
impact crater. They recognized that the Chic-
xulub melt rocks show high levels of iridium and
their age corresponds well with the KT boundary.
Single crystal “°Ar/*’Ar dating places the melt
rock at 65Ma (Swisher et al, 1992). All this
combined evidence suggests that the Chicxulub

494 MEMOIRS OF THE QUEENSLAND MUSEUM
UNITED STATES
* KT Ejecta Layer
Braggs e ; j
Brazos @ ® Tsunami Deposits
GULF OF MEXICO
Mimbral * J
i tur
MEXICO Chicxulub Struc uro;
JAMAICA
m ae
ies BELIZE
A J
" meet 8S CARIBBEAN SEA
4; GUATEMALA 72mm ee,
; r” HONDURAS $
jt 180 Km >
l |
B | |
[9] : 7
TP C Tertiary Limestone
L I5 â [233
[5 By be rau
E Cretaceous Limestone KT Impact Melt Cretaceous Limestone
PEZ
= Iter +444 Tt
$5 HHIHHBEHÁUBDU pracciaf Central 1455002 RUD
a L HHA pe A1 41443 GrPecia enar Tipo po e eerrebeE tts]
fttt E EE HA eb deny ^ Lens, ++ Uplift + + ey 0 E 324 E47 HAS ERA MOUSE ETH
* +
titrttrttrłtt titt Basement iCERLILE og III as. prises Basementr retried
Lg -bhbtett beet tee tt tet tebe ett tet tet tt+++++++ t+t+ttt+t+t+ ptttt+ tet ttt tt

FIG. 2. A, Location of the Chicxulub structure on the northern edge of the Yucatan Peninsula, Mexico, showing
distribution of proximal impact deposits. B, Cross-section; estmated crater diameter is 180km (simplified from

Hildebrand et al., 1994).

structure may be a prime candidate for the long
sought KT impact crater (Fig. 2B).

It appears from the above discussion that two
lines of evidence support the impact scenario for
the Chicxulub structure: evidence of shock
metamorphism, iridium enrichment, brecciation
and impact melt within the crater itself, and the
distribution of proximal ejecta components at the
KT sections in Haiti, Mexico, Texas, Alabama,
the Caribbean and adjacent areas (Fig. 2A). How-
ever, in recent times both interpretations have
faced strong criticisms. For example, some

workers dispute the impact origin of the Chic-
xulub structure and interpreted it as a buried
volcanic complex of Late Cretaceous age (Of-
ficer et al., 1992; Meyerhoff et al., 1994). Their
counterargument is based on the original subsur-
face correlation proposed by Lopez-Ramos
(1975) and the unpublished well log for the
Yucatan No. 6 well, drilled in 1966 over the
Chicxulub structure. One of the authors, Dr Ar-
thur A. Meyerhoff, was a consulting geologist to
Pemex at the time Yucatan No. 6 was drilled and
had first-hand information on the biostratigraphy

KT EVENTS IN INDIA

of the site. This well penetrated a superimposed
sequence of Pliocene-Miocene, Oligocene,
Eocene-Palaeocene and Maastrichtian-Cam-
panian sediments, and bottomed in andesitic
rocks and Cretaceous limestone, dolomite and
anhydrite. The most damaging evidence against
the impact origin is the inverted stratigraphy of
the Cretaceous horizon in relation to the andesite.
These dissenters point to the presence of 350m of
undisturbed Late Cretaceous sediments with
index fossils (foraminifera) overlying the an-
desitic body. If the andesitic rocks were indeed
impact melt from KT boundary time, the overly-
ing strata must be Palaeocene or younger in age.
Palaeontologic evidence indicates otherwise;
these strata are of Campanian and Maastrichtian
age lying conformably over the andesite. Swisher
etal. (1992) explained this stratigraphic inversion
as fallback breccia of Cretaceous limestone infill-
ing the crater. However these overlying
Cretaceous strata are not disturbed, brecciated
and shocked, and thus pose a problem for the KT
impact age of the andesite body.

Meyerhoff et al. (1994) indicated the following
additional discrepancies in the impact origin
hypothesis of the Chicxulub structure:

1) The Chicxulub structure is too shallow
(~2000m) for an impact of this dimension; ex-
pected excavation depth would be around 10km
or greater (Melosh, 1989).

2) Several layers of bentonitic breccia occur
interbedded with the Cretaceous limestones
without any structural disturbance or obliteration,
so their impact origin is suspect.

3) If the andesite were of impact origin, one
would expect highly homogeneous composition
with appreciable thickness; in contrast, chemical
analysis suggests that the Chicxulub andesite is
thin, chemically inhomegeneous with a wide
range of major oxide compositions (Sharpton et
al., 1992).

4) In an impact structure, the impact melt rep-
resents the aggregate composition of the target
rock (Engelhardt, 1984); in Chicxulub, on the
other hand, the target country rock is metamor-
phosed quartzite and rhyolite, whereas the
presumed impact melt is andesite.

5) The anhydrite at the bottom of the Yucatan
No. 6 well would have been completely
vaporized at the point of collision, if there was an
impact; its presence below the putative melt rock
is anomalous.

6) Unlike the planar and parallel lamellae in
shocked quartz associated with an impact site, the
lamellar features in quartz grains in Chicxulub

495

breccia show curvatures typical of volcanic
origin.

7) Later thermal events in the Chicxulub vol-
canic sequence might have reset the radiometric
age of the andesite; these authors indicated that
out of ten samples analyzed by Sharpton et al.
(1972), nine gave spurious results; accordingly,
the correlation between the KT impact event and
the presumed andesite melt is tenuous at best.

8) The Chicxulub volcanics are not local impact
melt, but part of a well-known Late Cretaceous
igneous province surrounding the Gulf of
Mexico.

Not only the impact origin of the Chicxulub
structure, but also the interpretation of proximal
deposits of ballistic ejecta and impact-wave dis-
turbances in the Caribbean and Gulf Coast is in
dispute. Recent studies have shown that many of
these so called impact-generated deposits may in
fact represent gravity-flow or turbidite deposits
occurring over an extended period of time,
whereas supposed impact droplets are altered
volcanic particles (Lyons & Officer, 1992; Keller
et al., 1993; Stinnesbeck et al., 1993; Beeson et
al, 1994; Adatte et al, 1994).

The presumed connection between the Haitian
glasses and Chicxulub has been questioned by
Koeberl (1993) on the basis of geochemical
evidence. He pointed out that, at the time of
impact, the Chicxulub area was covered by
evaporitic and carbonate deposits several
kilometers thick. Yellow glasses found in the KT
section of Haiti are not tektities and cannot be
linked to evaporitic target rocks of the Chicxulub.
Similarly, the interpretation of some breccias
within Chicxulub as impact breccias may be
wrong. However, proponents of the Chicxulub
Crater have dismissed most of these criticisms
(Alvarez et al, 1994; Hildebrand et al., 1994).

Thus, the Chicxulub impact is open to question,
and more study is needed before a final assess-
ment can be made. The only way to settle this
question decisively is by drilling new holes into
the Chicxulub structure, as in the case of the
Manson Crater, to determine more precisely
when and how it formed.

THE SHIVA CRATER

Although the Chicxulub structure has emerged
as the leading candidate for the KT impact scar,
another promising KT impact site has been iden-
tified at the India-Seychelles rift margin in the
northwest Indian Ocean, almost antipodal to the
Chicxulub structure. Hartnady (1986) suggested

496 MEMOIRS OF THE QUEENSLAND MUSEUM

0°
10°S \ ^X Phenai Mata’ v Y vv
vy
Carlsberg Ridge A RA: z "spun YY
EN A Let gatpuri y Jv vs
Central Peak NR 7 * VY
BAY OF BENGAL
20°S
Deccan-Reunion H
INDIAN OCEAN can-Reunion Hotspot
Positive Gravity Anomaly
Deccan Traps
Precambrian Granite
Impact Melt (Alkali Igneous Complexes)
200 km
30°S

50°E 60°E 70°E

FIG. 3. Location of the Shiva Crater at the India-Seychelles rift margin during KT boundary; arrow indicates
trajectory of meteorite; radial, asymmetric distributions of alkaline igneous complexes (impact melt rocks)
downrange of the Shiva Crater are shown by closed circles (data from Bose, 1980); areas of positive gravity
anomaly (data from Biswas, 1988) coincide with the ejecta melt distribution; asterisk indicates location of the
Deccan-Reunion hotspot.

that the Amirante Basin, south of the Seychelles southwest by the structure of the Amirante Arc.
Island, may be a possible KT impact site. The Sediments from the adjacent Amirante Passage
basin has a subcircular shape of about 300kmin have yielded Late Maastrichtian foraminifera
diameter, bounded on the northeast by the Abathomphalus mayorensis (Johnson et al.,
Seychelles Bank and partially ringed on the 1982), while basalt samples dredged from the

KT EVENTS IN INDIA

Amirante Arc look like Deccan Trap and have a
similar radiometric age (Fisher et al., 1968). Both
palaeontologic and radiometric age indicate that
the arc was formed near the KT boundary. How-
ever, its arcuate structure is enigmatic. It does not
appear to be a recent or ancient trench, as it lacks
volcanic activity, seismicity, and any significant
accretionary sedimentary prism on its ‘landward’
side (Johnson etal., 1982). Thus the interpretation
of the Amirante Arc as a crater rim is a distinct
possibility. The proposed impact also may ex-
plain the puzzling jump of the Carlsberg Ridge at
the KT boundary during the rifting of India and
the Seychelles. Hartnady noticed that the
Carlsberg Ridge between the Seychelles and
Madagascar jumped more than 500km to the
northeast to lie between India and the Seychelles
and initiate rifting between these two landmasses.
He could not find any evidence for plate reor-
ganization in the Atlantic or Pacific Oceans
during this time. He attributed this major plate
tectonic adjustment to the enormous force of a
large meteorite. As additional evidence, he
pointed to massive tsunami deposits in the KT
boundary section of Somalia and Kenya, which
may be linked to this impact event.

Although Hartnady’s model initially had at-
tracted wide attention, one major problem of his
idea is the enigmatic morphology of the Amirante
Basin. It is semicircular in outline, preserving half
of a supposed crater rim. What happened to the
other half of the crater? Although circularity is not
diagnostic of impact origin, there must be
evidence of some structure off the coast of
Seychelles. Is it possible to find the missing rim?

Alt et al. (1988) remedied the deficiency of the
Amirante Basin model as the point of collision.
They argued that during KT boundary, the
Seychelles was adjacent to the west coast of india.
They concurred with Hartanady (1986) that the
western rim of the crater survives in the Amirante
Arc, but the eastern rim lies along the west coast
of India, hidden by the overlying Deccan Traps.
They speculated that the impact was forceful
enough to create not only the enormous crater
approximately 600km in diameter, but also to
cause pressure-release melting in the astheno-
sphere. Basalt then filled the crater basin to form
an immense lava lake, the terrestrial equivalent
of a lunar mare. The synchrony of initiation of
spreading along the Carlsberg Ridge, the
emplacement of the flood basalts at the Deccan
plateau, Saya de Malha bank and Amrante Basin,
as well as close spatial association around the
crater basin, indicate that the array of simul-

497

taneous tectonic and volcanic features might have
been triggered by a single physical cause — an
enormous impact.

Chatterjee (1990; 1992) elaborated upon this
KT impact scenario at the India-Seychelles rift
margin, and identified the eastern rim of the crater
along the Panvel Flexure, near the Bombay coast
(Figs. 3-6). The Panvel Flexure is an arcuate
segment of the crater about 120km long on the
Deccan Traps, and it is difficult to explain in
terms of conventional tectonics. It is marked by a
line of hot springs, dikes, deep crustal faults and
seismicity (Kaila et al., 1981; Powar, 1981).
Since the Indian shield is usually aseismic, the
seismicity along the flexure is unusual, indicating
tectonic instability. The geothermal gradient is
abnormally high along this flexure (36-78°C/km’)
with evidence of thinned lithosphere (31-39km),
suggesting melting conditions at shallow depths
(Negi et al., 1992). The Panvel Flexure may rep-
resent the eastern rim of the crater in the form of
a collapsed rim structure. It exercises tectonic
control on the attitude of the Deccan lavas. To the
east of the flexure, the basaltic flows are horizon-
tal; to the west of the flexure, the basaltic flows
dip west to west-southwest at 50? to 60? toward
the coast. The abrupt change of dip along the
flexure axis may indicate the slope of the eastern
crater wall, which is now concealed by Deccan
lavas. Seismic data indicate that the basement
topography below the Deccan lava west of the
flexure has a crater-like depression (Kaila et al.,
1981). Completing the oval by combining the
Amirante Arc and the Panvel Flexure, the extent
of the crater can be extrapolated. It is a giant oval
crater, 600km long and 450km wide, showing the
morphology of a complex impact scar. Chatterjee
(1992) named this impact structure the Shiva
Crater, after the Hindu god of destruction and
renewal (Fig. 3). However, the Shiva Crater is
difficult to interpret because it is submarine and
largely concealed by the Deccan lava. Many of
the impact signatures are thus erased or obscured.
Moreover, the rifting of the Seychelles from
India, which occurred along the width of the
crater, has obliterated the geomorphology of the
structure. A series of geodynamic and volcanic
events that occurred near KT boundary time must
be untangled and put into proper chronologic
order to unveil the crater morphology. Recent
exploratory data from the Bombay High, a giant
offshore oilfield located 160km west of the city
of Bombay, has produced a wealth of infomation
supporting an impact origin for the Shiva Crater.

498

40°E

ARABIAN

— PENINSULA

ARABIAN
SEA

10?N

60°E

MEMOIRS OF THE QUEENSLAND MUSEUM

70°E 80°E 90°E

INDIA

Panvel
Flexure

Bombay

CARLSBERG RIDGE

0?

SEYCHELLES
/

Amirante Arce

10°S

MADAGASCAR

20°S

30°S

SHIVA CRATER

INDIAN OCEAN

FIG. 4. Present day location of the split Shiva Crater in reference to India and Seychelles on either side of the
Carlsberg Ridge. Today, part of the Shiva Crater is attached to the southern part of the Seychelles, the other-half
to the western part of India. The crater was joined 65 million years ago when the Seychelles was part of India

before the spreading of the Carlsberg Ridge.

INIDA-SEYCHELLES FIT. Today, the Seychel-
les microcontinent is separated from the western
coast of India by 2,800km because of spreading
along the Carlsberg Ridge (Fig. 4). This
midoceanic ridge shows symmetrical magnetic
anomalies of 5, 23, 24, 25, 26, 27, 28 and 29 on
either side of the ridge axis between India and the
Seychelles (Norton & Sclater, 1979; Naini &
Talwani, 1982). Using both palaeomagnetic and
palaeontologic evidence Chatterjee (1992) has
restored the India-Seychelles fit for the KT
boundary time (Figs. 3 & 8). The reconstruction
places the western coast of India against the
Seychelles-Saya de Malha Bank at about the time

of Deccan volcanism and shows matching
geological provinces. The largely submerged
continental block that bears the Seychelles Is-
lands contains enormous flood basalt deposits in
the submarine plateau of the Saya de Malha Bank,
which are inferred to be an extension of the
Deccan volcanism (Meyerhoff & Kamen-Kaye,
1981; Backman et al., 1988), especially the Bushe
Formation of the Upper Deccan Basalt Group
(Devey & Stephens, 1992). The link between KT
magmatism on the Seychelles and India is em-
phasized by matching the geochemistry and
geochronology of alkaline rocks (White & Mc-
Kenzie, 1989; Devey & Stephens, 1992). Similar-

KT EVENTS IN INDIA

ly, the Late Proterozoic Mahe Granite on the
Seychelles (Baker & Miller, 1963) is isochronous
(70050Ma) with the Siwana-Jalor Granite
(Auden, 1974) of western India, the crystalline
basement below the Deccan lava. Various tec-
tonic and volcanic features, when restored for the
Indo-Seychelles block at 65Ma, reveals the
presence of a large oval, oblong structure, the
Shiva Crater (Figs. 3, 5).

ANATOMY OF THE SHIVA CRATER. Since
the Shiva Crater was spilt by the Carlsberg Ridge
and each half is now buried under a thick pile of
lava flows, and because the structure is largely
submarine, geophysical exploration and drilling
data are essential to understanding its morphol-
ogy and structure. Morever, overlying lava flows
and thick sediments obstruct a direct examination
of various impact signatures such as shock
metamorphic effects, breccia and impact melt
that are generally associated with complex
craters. Today, one part of the crater is attached
to the western coast of India, the other to the
Seychelles (Fig. 4), but of course both parts were
joined at the KT boundary (Fig. 3).

The most critical evidence regarding this im-
pact structure comes from recent oil exploration
in the Bombay High offshore basin, which repre-
sents the eastern-half of the crater. Bombay High
is a giant offshore oilfield (~120,000°km) located
160km west of Bombay in the Arabian Sea at a
depth of about 75m (Fig. 5). The structure is 60km
long and 20km wide, trending WNW-ESE with a
faulted eastern flank. The stratigraphy of this
enigmatic structure is known from extensive
drilling and seismic data by the Oil and Natural
Gas Commission (Rao & Talukdar, 1980; Basu et
al., 1982; Bhandari & Jain, 1984). The Bombay
offshore basin shows part of the crater rim (Pan-
vel Flexure and Narmada Fault), annular trough
(Surat Basin, Dahanu Depression, and Panna
Depression) and the central uplift (Bombay High)
(Fig. SA). The morphology of the western-half of
the Shiva Crater around the Seychelles microcon-
tinent is also known from oil exploration data
(Meyerhoff & Kamen-Kaye, 1981; Kamen-
Kaye, 1985; Devey & Stephens, 1992). Here we
see part of the crater rim (Amirante Arc), annular
trough (Amirante basin) and the central peaks
(Mahe and Praslin granitic cores). When the
Shiva structure is restored to KT boundary time
(Fig. 5), it shows the structure and morphology of
a giant, complex crater, oval in outline with: 1, a
collapsed outer ring, 600km long and 450km
wide; 2, an annular trough, presumably filled

499

with KT melt rocks; and 3, a distinct central core
in the form of linear uplifted peaks of older
Precambrian granite.

The Shiva Crater is defined by a collapsed outer
ring, which is partially preserved in the form of
the Amirante Arc, Panvel Flexure and the Nar-
mada Fault (Figs. 3 & 5). The outer rim is sur-
rounded by a gravity high, especially in both the
Panvel Flexure and Narmada Fault areas (Biswas,
1988) and may be linked to the distribution im-
pact melt. Part of the outer ring along the
northwest and southeast edges was obliterated by
spreading of the Carlsberg Ridge. The outer ring
is followed by the annular trough which was
largely filled with ponded Deccan lava (and pos-
sibly by suevite and impact melt). Part of the
trough is preserved in the Surat Basin, Dahanu
Depression and Panna Depression around the
Bombay High, and the Amirante Basin, south-
west of the Seychelles. The uplift in the center is
represented by a large peak followed by two small
irregular peaks, all composed of older
Precambrian granitic cores: the Bombay High on
the western coast of India and the Praslin and
Mahe Islands of the Seychelles. Jansa (1993)
pointed out that the central uplift in oceanic im-
pact sites is generally cylindrical in shape, as is
the Shiva Crater. The lateral continuity of the
central peaks was disrupted by the Carlsberg Rift
and may indicate a collapse structure. The central
uplift associated with the 600km Shiva Crater is
estimated to be 150km wide, about one-fourth of
the crater's final diameter, as is expected in a
complex crater (Melosh, 1989). Figure 5B is a
geologic cross-section of the 600km diameter
Shiva structure along the longitudinal axis. The
depth of the crater can be estimated from the
thickness of the sedimentary basins around the
annular trough as well as from the height of the
central peaks. The sedimentary fill around the
trough of Bombay High consists of shallow
marine Tertiary sediments exceeding 5,000m,
overlying the Deccan basalt floor (Bhandari &
Jain, 1984). Such thick sediments indicate that
the crater basin is more than 5km deep above the
Deccan floor. The thickness of the Deccan lava,
and the presence of impact melt within the crater
are unknown, but Meyerhoff and Kamen-Kaye
(1981) have described a well log on the Saya de
Malha Bank which penetrated 832m of basalt
overlain by 2400m of upper Palaeocene to
Quaternary sediments. Based on their seismic
work, Shor & Pollard (1963) suggested that ap-
proximately 2km of Deccan basalt overlie the
granitic basement some 80km southwest of the

500 MEMOIRS OF THE QUEENSLAND MUSEUM

CRATER RIM
Narmada Fault

A * ró
ad
S Ul CRATER RIM
CARLSBERG 7 ANNULAR TROUGH Panvel Flexure
RIDGE x, # Surat Basin "à

Dahanu Depression

/ ANNULAR TROUGH
/
/ SHIVA N
Seychelles,
/ CRATER Bank Sc :
/ Praslin / Panna Depression
l
I Ex
i ba CENTRAL UPLIFT
7
y^
í A
ANNULAR TROUGH
Amirante Basin sp
CRATER RIM "
Amirante Arc " P EO kn
^
a
A
— ak $
A
nm CRATER DIAMETER >|
(~600 km)
ANNULAR TROUGH COLLAPSE CENTRAL UPLIFT ANNULAR TROUGH
] l T 1
CRATERRIM /
\ CRATER RIM ; ; (Panvel Flexure) / 2 km
\ ? Mahe Praslin Bombay High 2a
MSL (Amirane Arc) EN i EY HSD
-10 km ei NE A | S i ; I. -10 km
-20 km f iov Amp “MA LESS C y EEP CRUST. -20 km
d d VE Te LM ‘ US TG (xl ees d TAS
Bosker Sane PA Hv M DA tanta as : I ME : 36
A EN Impact Melt B
SW v v Deccan Traps NE

++4 Precambrian Granite

FIG. 5. Morphology of the Shiva Crater. A, plan view showing a central uplift (Bombay High, Praslin and Mahe
granitic core); an annular trough (Surat Basin, Dahanu Depression, Panna Depression and Amirante Basin);
and a slumped outer rim (Narmada Fault, Panvel Flexure and Amirante Arc). The oblong crater is about 600km
long, 450km wide and more than 12km deep; it is bisected by the Carlsberg Ridge. B, schematic cross-section
along the line AB; the post-impact Deccan lava flows are removed to show the morphology of the crater and
possible sites of impact melt sheets; (seismic profile data from Rao & Talukdar, 1980; Kaila et al., 1981).

KT EVENTS IN INDIA

Mahe Peak. If we now add to this the 2km thick
lava pile in the Western Ghats section near the
Panvel Flexure, the depth from rim to actual floor
may exceed 10km (Fig. 5B).

AGE OF THE SHIVA CRATER. Although the
age of the Shiva Crater is not precisely known,
combined evidence from various components of
the structure, such as the formation of the rim,
biostratigraphy of the annular trough, age of the
melt sheets, timing of the central uplift, associa-
tion of Deccan volcanics and the Carlberg rifting
event suggests that the crater formed at the KT
boundary. As discussed earlier, the preserved rim
of the Shiva Crater, such as the Amirante Arc
(Hartnady, 1986; Alt et al., 1988), the Panvel
Flexure (Auden, 1949; Chatterjee, 1992) and the
Narmada Fault (Biswas, 1988) evolved at KT
boundary time. Seismic stratigraphy has iden-
tified the basement rock as reflection-free or
chaotic Precambrian granite in the form a central
uplift with a thin veneer of Deccan lavaat the base
(Rao & Talukdar, 1980; Basu et al., 1982; Bhan-
dari & Jain, 1984). The oldest sediment overlying
the Deccan Trap or crystalline basement is the
Panna Formation of Palaeocene age (Fig. 6). The
upper boundary of the Panna Formation coin-
cides with H-4 seismic horizon. The Panna For-
mation is composed of poorly sorted, angular
sandstone, claystone and trap fragments at the
bottom, followed by shale and coal sequence.
This unit is relatively thin on the uplift, but attains
a large thickness (75m) on the flank. Seismic data
indicate that the formation may be as thick as
500m in the annular trough region, but wells have
not penetrated the bottom layer. Although the
formation is mostly unfossiliferous, it has yielded
Globorotalia pseudomenardii from the middle of
the sequence corresponding to P4 planktic
foraminiferal zone of Late Palaeocene, indicating
a short palaeontologic hiatus after the KT event.
Since the impact took place in a shallow-marine
setting, the hiatus may be linked to erosion by a
megatsunami generated by the impact. The KT
boundary section lies farther down at the bottom
of the sequence. Jansa (1993) suggested that in
oceanic impacts most of the fall-out breccia is
reworked back into the crater cavity. If so, the
lowest unit of the Panna Formation, if recovered
in future, should be investigated for an iridium
anomaly and shock metamorphic minerals (Fig.
6B). The crater floor is composed of younger
Deccan flows from the adjoining Western Ghat
section that took place around ~65Ma (Couirtil-
lot, 1990; Duncan & Pyle, 1988). The Deccan

floor at the annular trough and the overlying
Panna Formation narrow down its age to close to
the KT boundary. Recent isotopic dating of the
Shiva melt rock, as discussed in the following
section, has yielded an ^0 Ar/? Ar date of 65Ma
(Basu et al., 1994). Finally, the Carlsberg rifting
within the crater basin was formed during the
magnetic chron 29R (Naini & Talwani, 1982).

Although it is generally believed that the Bom-
bay High was formed during the breakup of India-
Seychelles at the KT boundary (Basu et al., 1982),
no tectonic mechanism has been offered to ex-
plain this spectacular uplift (~ 10km high) of the
Precambrian granite with a thin veneer of Deccan
Traps at the passive margin of the Bombay shelf
(Fig. 6B). We propose that this structural high
represents one of the central peaks of the Shiva
impact that originally underlay the transient
crater. The other two central peaks, Praslin and
Mahe Islands in the Seychelles (Fig. 5A), also
rebounded upward at the same time as they con-
tain similar Precambrian granites and younger
KT intrusive rocks (Devey & Stephens, 1992),
These central peaks are composed of deformed
and fractured rocks that have been stratigraphi-
cally uplifted distances comparable to the crater
depth (Melosh, 1989). The fractured nature of the
central peak of the Bombay High can be seen in
Figure 6B. The crystalline rocks beneath the
Shiva Crater are shattered, as in the Ries Crater
of Germany, as inferred from the low seismic
velocity beneath the H4 horizon (Rao & Talukdar,
1980). Similarly, Baker (1967) reported a
‘megablock zone’ — a chaotic assemblage of
gigantic blocks of granite, each up to 13m high,
often marked with surface fluting at the Mahe
uplift in the Seychelles. Evidence from drill
holes, geochronology and seismicity suggest that
the magmatism at the Seychelles Bank occurred
at the KT boundary, producing both Deccan lavas
and alkaline igneous complexes (Devey &
Stephens, 1992).

IMPACT MELT ROCKS. The Shiva impact must
have produced enormous volumes of impact
melt, breccias and shocked materials between the
central structure and the rim. Yet, these impact
signatures are difficult to interpret because of
mobilization and mixing of thick lava flows from
Carlsberg Ridge, subsequent burial by the Dec-
can lava, as well as lack of drilling and seismic
data below the Deccan floor, However, there is
some indirect evidence that indicates a lava-like
impact melt was emplaced radially within and
outside the crater.

502 MEMOIRS OF THE QUEENSLAND MUSEUM

SHTRA SSS

;

ANNULAR TROUGH ':;
(Surat Basin) 9

BOMBAY

WHO |,
V

55: RATNAGIRI
CUSPREZZFAULTTE
Oil/Gas fie

Deccan Traps

ld ja^ueg) WI gar

l nx
WW

Precambrian Granite

Hot Springs

= APPARENT RADIUS OF THE SHIVA CRATER -

-240 ki
xm m) CRATER RIM 2km
(Panvel Flexure)

BOMBAY HIGH PANNA DEPRESSION -

E Ns

A f Neogene/Quaternary Deccan Traps (~ KT boundary) R
m | Neogané FEE] Precambrian Granite
= g! EEH
B

Paleogene 20 km

pu
Pose Panna Formation (Paleocene)

FIG. 6. A, Location and tectonic framework of Bombay High in relation to India representing the eastern-half of
the Shiva Crater (simplified from Rao & Talukdar, 1980). B, seismic cross-section along the line AB to show
the central uplift (Bombay High), annular trough (Panna Depression) and the crater rim (Panvel Flexure)
(simplified from Rao & Talukdar, 1980).

One of the most intriguing features following along the radii of the Shiva Crater (Figs 3, 7).
the Deccan flood basalt volcanism is the occur- They are manifested in plug-like bodies and
rence of several post-tholeiitic alkali igneous minor intrusions in the western and northwestern
complexes of nepheline-carbonatite affinities, part of the Deccan volcanic province and are

KT EVENTS IN INDIA

limited in space and volume compared to the vast
expanse of tholeiitic lavas (Bose, 1980; De,
1981). They are clearly defined by zones of
gravity highs (Biswas, 1988). Extrusive rocks are
relatively rare except in the Kutch rift zone, in-
dicative of fissure eruption. Devey & Stephens
(1992) described contemporary alkaline in-
trusives in several islands in the Seychelles
microcontinent, especially in Mahe, North Island
and Silhouette Island within the Deccan lavas.
Geochemically they are very similar to the Murud
alkaline dikes of Bombay. Recent Aar/? Ar
dating of these alkaline complexes indicates
65Ma, precisely coinciding with the KT bound-
ary (Basu et al, 1993; Pande et al, 1988; Devey &
Stephens, 1992). In the Anjar section of the ex-
trusive alkaline complexes, cosmic iridium and
osmium anomalies have been reported from the
interbed (Bhandari et al., 1994). Their distribu-
tions are shown in Figures 3 & 7A.

Two spectacular volcanic plugs within the
Shiva Crater need discussion. One is Fortune
Bank igneous center south of the Seychelles, the
other is a newly discovered buried structure near
Bombay. The Fortune Bank is an unusual sub-
marine volcanic center, about 15kms high and
50km across, with anomalous magnetic and
gravity signatures within the Deccan volcanics
(Girling 1992). Devey & Stephens (1992) inter-
preted this structure as a large alkaline intrusion
contemporary with similar volcanics around the
Seychelles. Negi et al. (1993) identified a large
volcanic cone-like intrusive, about 12km high
and 35km in diameter, buried 6km below the
Deccan Trap near Bombay, east of the Panvel
Flexure. It is defined by a high gravity and ther-
mal anomaly. They found that the crustal thick-
ness in this site is half of the normal Moho depth,
and the granitic basement is almost missing in this
region. They interpreted the structure as a large
fossil conduit, formed at the point of collision of
a large KT meteorite. For ease of description, we
designate this igneous complex the Napsi struc-
ture after the last names of the discoverers (Negi,
Agrawal, Pandey, Singh). Since alkaline in-
trusives are known from this area, and they show
high gravity anomalies relative to the surround-
ing Deccan Trap, we interpret the Napsi structure
as a massive alkaline complex similar to the
Fortune Bank (Fig. 7A).

The origin of these alkaline igneous complexes
within the Deccan volcanics has been debated for
many years. They have been linked to fractional
crystallization of parent tholeiitic magma (West,
1958), an early stage of Carlsberg rifting

503

(Thompson & Nelson, 1972; Devey & Stephens,
1992) and a late stage of Reunion mantle plume
(Bose, 1980; De, 1981; Basu et al. 1993). How is
it possible to derive from the same mantle source
large volumes of tholeiite extrusives followed by
limited occurrences of alkaline intrusives? Why
are the alkaline rocks so heavy relative to the
Deccan Traps and so easily demarcated by
gravity anomalies? How can we explain the abun-
dance of these alkaline rocks in and around the
Shiva Crater in the form of volcanic cones? Some
of these buried volcanic cones, such as Napsi
structure and Fortune Bank even dwarf Mount
Everest. Alkaline igneous complexes are com-
monly found as melt sheets associated with
several Canadian craters (Grieve, 1987). The spa-
tial and temporal correlations of these alkaline
igneous complexes with the Shiva Crater are
intriguing, making a causal relationship likely.
All these alkaline igneous complexes show the
following features indicative of impact origin:

1, The centers of alkaline magmatism are
clustered around the Shiva Crater in radial
fashion but conspicuously absent in other parts of
the Deccan province (Figs 3 & 7A);

2, Their age matches exactly with the KT im-
pact event;

3, They are all defined by positive gravity
anomalies;

4, They have restricted distributions and occur
within the Deccan volcanics as post-tholeiitic
intrusives or plugs;

5,Their parent melt composition was
homogeneous, but later differentiation within the
plug has produced varieties and compositional
layering as in the case of impact melts in several
Canadian craters such as Brent and Manicougan;

6, They show higher alkali content than the
country rock;

7, They show evidence of crustal contamina-
tion.

This evidence suggests that the alkaline ig-
neous rocks around the Shiva Crater were formed
by crystallization from impact melted country
rocks. We hypothesize that these alkaline rocks
represent melt ejecta produced by impact-in-
duced mixing and melting of the target rocks, and
were emplaced radially on downrange side of the
trajectory (Fig. 7A). The close isotopic and age
relationship of the impact melt with the younger
Deccan volcanics such as the Ambenali Forma-
tion (Pande et al., 1986; Devey & Stephens, 1992)
indicates remelting of these rocks. We believe
three groups of rocks, younger Deccan volcanics,
platform carbonates and evaporites, and

504

Carlsberg Ridge

NN
aN

CRATER WALL

SHIVA
CRATER

=

TRAJECTORY OF PROJECTILE

CENTRAL UPLIFT ES;

ARTIFICIAL CRATER

MEMOIRS OF THE QUEENSLAND MUSEUM

* N
Tt A Anjar : N
Z p vA N
i a D NC
Lr DISTRIBUTION OF MELT EJECTA
CQ. Gimar (Alkaline Igneous Complex)
IS E Kadi

wane A Phenai Mata
- ^ [i Amba Dongar
ES

~ A Barwaha

ae 100km | —-
"NOn Deccan-Reunion

EA Hotspot

\\

ME P k
OF EJECTA-

FIG. 7. A, Schematic three-dimensional view of the Shiva Crater showing the radial distribution of the impact
melt rocks in the form of volcanic plugs; younger Deccan volcanics are removed to show the morphology of
the crater; oblong, teardrop shape of the crater and the asymmetric distribution of the melt rocks consistent with
an oblique impact event along the NE downrange direction. B, Artificial crater produced by low-angle (~15°)
oblique impact in the laboratory mimics the shape and ejecta distribution of the Shiva Crater (simplified from

Schultz & Gault, 1990).

Precambrian basement were involved in the
genesis of the melt rocks of the Shiva Crater. The
younger Deccan volcanics covering the
Precambrian basement of the Bombay and
Seychelles region at the KT boundary were the
main target rocks. However, the Precambrian
granite itself was also involved in the impact, as
indicated by the unusually thin crust in the Bom-
bay area with missing granitic layer (Negi et al.,
1993) as well as by evidence of crustal con-
tamination in the alkaline suites (Paul et al, 1977;
Basu et al., 1993). The Shiva Crater was also

inferred to have been excavated on a shallow sea
platform containing thick carbonate and
evaporitic rocks such as we see today around the
Bombay High and Rann of Kutch. Thus the melt
rock can be interpreted as the mixing product of
Deccan Traps, Precambrian granite and the thick
Late Cretaceous sequences of carbonate/evaporite-
rich sediments. These Cretaceous sediments are
still preserved in the adjacent Kutch, Saurashtra
and Rajasthan areas, covered by the Deccan
Traps, and are encountered by subsurface drilling
(Biswas, 1988). However, the alkaline rocks are

KT EVENTS IN INDIA

TABLE 1. Large Earth-crossing and Earth-approach-
ing asteroids (diameter > 10km; after Wetherill &
Shoemaker, 1982).

[|] «ass
Apo ae

— | DIAM. (Km)
114

E
| 1866 |

generally undersaturated with a high K20/NaO
ratio relative to the Deccan tholeiites (Bose,
1980). Such an ‘anomaly’ is not unusual with an
impact event. Alkaline volcanic rocks are com-
monly present as impact melt in most of the
Canadian craters. Grieve (1987) discussed the
compositional variation of these alkaline melt
rocks, where melt sheets have a higher K20/NaO
ratio than the target rock. The reason for this
alkali enrichment is not clear; he believed that
either it is caused by selective elemental
vaporization and condensation during the melt
and vapor formation, or hydrothermal alteration.
Since the impact was largely oceanic, the latter
explanation of enrichment of alkali from sea
water is likely. The gravity high of the alkaline
igneous complexes most likely reflects mass con-
centration associated with dense impact melt and
uplift of the silicate basement rock along the
crater rim (Figs. 3, 7A).

The radial distribution of melt sheets in the
down-range direction of the Shiva Crater is in-
triguing (Fig. 7A). They might have developed
either during the passage of the shock waves
when fluid ejecta was emplaced downrange, or
during the collapse of the transient crater when
radial fractures were formed. Melosh (1989) dis-
cussed the mechanism by which radial fracture
patterns develop during the collapse of a large,
multi-ring crater. Radiating fractures from the
central uplift to the rim are known from the
Manicougan Crater of Canada (Grieve et al.,
1988) and the Wells Creek Structure of Tennessee
(Stern, 1968).

SIZE AND TRAJECTORY OF THE IMPACTOR.
Throughout its history the Earth has been im-
pacted by countless meteorites. There are two
broad categories of meteorites with orbits that
bring them close to the Earth: comets and
asteroids. Comets are composed in large part of
water, ice and other volatiles and therefore are
more easily fragmented than rocky or metallic
asteroids. Although smaller meteorites pose little
threat, impacts by large objects (diameter > Ikm)
constitute the greatest hazard, with their potential

505

for global environmental damage and mass mor-
tality. The most famous large impact on planet
Earth is the KT event that killed the dinosaurs and
other contemporary biota. Wetherill &
Shoemaker (1982) have summarized the current
knowledge of Earth-crossing and Earth-orbiting
asteroids and their probability of impacting the
Earth. They estimate that out of 1,000 near-Earth
asteroids (NEA) that have diameters greater than
lkm, three exceed the 10km range. These are
1866 Sisyphus (11.4km), 433 Eros (19.6km) and
1036 Ganymede (39.6km) (see Table 1). Both
Eros and Ganymede can be perturbed into Earth-
crossing orbits by close encounters with Mars.
They speculate that asteroids as large as 20km in
diameter probably have struck the Earth in the last
few billion years, and 10km diameter bodies ap-
parently may impact every 40 million years.
Recently Paola Farienella of the Universty of Pisa
in Italy and colleagues simulated eight computer
models to calulate the probability of colliding the
20km-wide asteroid Eros with the Earth. One
forecast of this catastrophic impact is alarming; it
could happen in the next 1.14 million years
(Desonie, 1996). Because of this continued threat
from space, NASA has organized an International
Spacegurad Survey Network to detect and
monitor near-Earth-objects. In February 1996
NASA sent its Near Earth Asteroid Rendezvous
(NEAR) spacecraft to examine Eros from a close
distance. The spacecraft is scheduled to arrive in
close proximity to this asteroid in 1999 to
scrutinize its surface in great detail for the under-
standing of its origin (Bell, 1996).

Although hypervelocity impacts normally
create circular craters, impacts at a low angle
(S15? from the horizontal) often generate elongate
craters such as the Messier and Schiller Craters
of the Moon (Wilhelms, 1987) and the Rio Cuarto
Crater in Argentina (Schultz & Lianza, 1992).
Craters formed by artificial oblique impact are
generally oblong (Moore, 1976; Gault &
Wedekind, 1978). The shape of an artificial crater
formed by oblique impact at 15? (Schultz &
Gault, 1990) is like a teardrop, where the pointed
end indicates the downrange direction (Fig. 7B).
In an oblique impact, the crater and its ejecta are
bilaterally symmetrical about the plane of the
trajectory, but the distribution of ejecta is con-
centrated asymmetrically on the downrange side.
The shape of the Shiva Crater and the distribution
of melt ejecta are almost identical to those of the
artitificial crater (Fig. 7A). If so, the impact that
produced the Shiva Crater was probably oblique
along a SW-NE trajectory as evident from the

506

direction of the longer diameter of the oblong
crater; the tip of teardrop indicates that the
downrange direction was NE. Howard & Wil-
shire (1975) described flows of impact melt of
large lunar craters both outside on crater rims and
inside on crater walls, where asymmetric dis-
tribution of this melt sheet can be used to deter-
mine impact trajectory. The rim pools tend be
concentrated on the inferred downrange side. The
asymmetric concentrations of sheet melts on the
NE side of the Shiva Crater indicate the
downrange direction (Fig. 7A).

We estimate that a 40km diameter asteroid,
about the size of the Amor object Ganymed, could
have created the Shiva Crater, initiated the
Carlsberg rift and excavated the crustal materials
into mantle reservoirs resulting in basaltic vol-
canism (Fig. 5A). An impact of this magnitude
may be a rare event, but seems possible, because
of the presence of even larger craters on the Moon
and Mars. Currently there are at least three known
craters which are close to 200km across:
Vredefort in South Africa (2,000Ma), Sudbury in
Canada (1,849Ma) and Chicxulub (65Ma) in
Mexico. The Shiva Crater is not unique, but near-
ly so. The only older similar structure in its size
range is the eastern shore (Nastapoka Arc) of
Hudson Bay in Canada: this 600km wide depres-
sion is believed to be an impact scar of Archean
age (Dietz, 1993). If properly understood, the
Shiva Crater is the largest impact structure
produced in the Phanerozoic and is consistent
with the environmental havoc wreaked at the KT
boundary.

A new model of lithosphere thinning by
meteoritic impact is proposed here to explain the
origin of the Carlsberg ridge, rift volcanism and
the Shiva Crater. In this model the lithosphere
could be excavated and shattered by a projectile
of considerable size (~40km) to initiate a
midoceanic ridge. Asteroids strike the Earth at an
average speed of 25km/sec and transfer consider-
able kinetic energy to the target rocks
(Shoemaker, 1983). The pressures exerted on the
meteorite and target rock can exceed 100GPa;
temperatures can reach several thousand degrees
Celsius; and impacting energy would generate a
100-million megaton blast (Grieve, 1990). Such
a hypervelocity impacting body penetrates the
target rocks to two or three times its radius
(Grieve, 1987). An asteroid of 40km diameter
would produce cratering and associated tectonic
rebound-collapse effects sufficient to shatter the
80km thick lithosphere that could form plate
boundaries and continental rifts. This concept

MEMOIRS OF THE QUEENSLAND MUSEUM

opens up a new field of research to investigate
whether plate tectonics may be influenced by
impacts of large bodies.

NEW LINKS BETWEEN THE CHICXULUB
AND THE SHIVA CRATERS.

A Multiple-impact Model. If both the Chicxulub
and Shiva Craters are real and were formed at KT
boundary time, is there any genetic link between
them? Hartnady (1989) suggested that if a low-
angle, oblique primary impact occured in the
Southern Hemisphere near India, then bolide
ricochet may have resulted in secondary impacts
a few minutes later in the Northern Hemisphere.
However, since the trajectory of the Shiva Crater
is from SW to NE, Hartnady's model fails to
explain the origin of both the Shiva Crater and the
Chicxulub Crater by oblique impact ricochet
process. However, the recent crash of 21 frag-
ments of comet Shoemaker-Levy 9 (S-L 9) on
Jupiter inspired scientists searching for similar
multiple impacts on other bodies in the solar
system. Crater chains were soon discovered on
Jupiter's satellites and on the Moon (Levy et al.,
1995). The cometary fragments of S-L 9 did not
collide simultaneously on Jupiter but spread in
time over 5 days. If the original KT meteorite
broke into several fragments, as in the case of
Shoemaker-Levy, and the larger one formed the
Shiva Crater, the second impact, almost after 12
hours (or odd multiple thereof), could create the
Chicxulub as the Earth rotated anticlockwise
around its axis. If the collisions of two fragments
were spread over 12 hours, two antipodal craters
could be formed by meteorite fragments around
a great circle. If so, one can predict that additional
KT impact scars, if discovered in future, should
lie on this great circle joining the Shiva and the
Chicxulub structures. This great circle is named
here the Alvarez Impact Belt in honor of Luis and
Walter Alvarez for their pioneering work (Fig.
8A).

There may be signs of this 'string of pearls'
effect along the Alvarez Impact Belt in the form
of a crater chain. Various authors have predicted
a third KT impact site on the Pacific plate which
surprsingly lay along the Alvarez Impact Belt.
Frank Kyte (pers. comm.) made a dramatic dis-
covery of a tiny fragment (- 3mm) of the KT
bolide in a drill core from DSDP site 576 in the
western North Pacific. The bolide chip held
micrometer-size metallic grains that are up to
8796 nickel and rich in iridium. From the
geochemical signature of the chip, Kyte specu-
lates that the KT projectile is probably an

KT EVENTS IN INDIA

CRATERS, (7 -
U^ M
N

[45 "
NS Mp,
b ACT a LL
aos '& SOUTH » |

WS ste |
à |o

DSDP 576».
ua M

/
T
i

507

SEISMIC WAVES —

e i

SHOCK WAVES, X = Ede

NS 77 MANTLE _
IMPACT — egt — — ^
SOURCES. SHIVA

‘CRATER
i ENS".

FIG. 8. Possible links between the Shiva and Chicxulub Craters. A, locations of the Chicxulub and the Shiva
Craters along the ‘Alvarez Impact Belt’ at KT boundary time; note, impact debris and extinction events are
concentrated along this belt. Both craters may have originated when two fragments from a larger meteorite
crashed on a rotating earth over the course of 12 hours. B, near-antipodal positions of the Chicxulub and the
Shiva Craters may also indicate alternative scenario that a large impact on one side of the Earth (near India)
might have produced a similar signature on the far side (near Mexico) by axial focusing of seismic waves

(modelled after Boslough et al., in press).

asteroid, not a comet, that slammed into Earth at
a shallow angle. This is the first direct evidence
regarding the carrier for iridium. The chip may
have broken off of the asteroid before it crashed
on the Pacific plate. At 65Ma, the Pacific impact
site was located midway between the Chicxulub
and the Shiva Craters right on the Alvarez Impact
Belt (Fig. 8A). Kyte et al. (1994) also identified
additional five KT boundary sites on the Pacific
Plate around 576, characterized by an iridium
anomaly, spherules and shocked quartz; they all
cluster along this DSDP site 576. The point of
collision of the KT projectile on the Pacific plate
is further supported by other evidence. Robin et
al. (1994) concluded from spinel compositions at
the KT boundary sites of the Pacific plate that
multiple impacts might have occurred from a
single disrupted bolide, where the largest objects
would have impacted in the Pacific and Indian
Ocean. Their spinel distributions and proposed
impact site at the Pacific fit nicely with the Al-
varez Impact Belt. Finally, all the known
localities with an iridium anomaly and shocked
quartz related to the KT impact event (see Alvarez

& Asaro, 1990) cluster symmetrically on either
side of the Alvarez Impact Belt. Distribution of
impact-related minerals and trace elements along
this belt coincides nicely with biogeographic
selectivity of extinction at low latitudes (Keller,
1994). The high latitudes probably served as a
refuge for many organisms, especially for plants.
If so, the distribution of several impact sites and
variation of composition of impact-generated
minerals along the Alvarez Impact Belt could be
explained by multiple impact events on a rotating
Earth, around the equator, rather than by a single
large impact. Such a latitudinal gradient, with
most extinctions occurring in the tropics, is exact-
ly what is to be expected by a chain of three
impact sites in India, Mexico and Pacific plate
respectively. These sites were formed when a
large parent body broke apart into a string of
smaller asteroidal fragments in the inner solar
system and crashed into the planet.

Recently, Alvarez et al. (1995) raised three
paradoxes in KT boundary sediments which are
apparently difficult to reconcile in a single impact
hypothesis, 1) In North America the KT boundary

508

reveals the double layer of ejecta: the lower layer
consists mainly of iridium spike and altered
spherules, whereas the upper contains shocked
quartz. 2) Shocked quartz is more abundant and
coarser grained at longitudes west than east of the
Chicxulub Crater. Various KT boundary sites in
Europe, Africa and Asia show little or no
evidence of shocked quartz and may represent a
‘forbidden zone’ for shocked quartz distribution.
3) The proposed impact energy (>100GPa)
released by collision of a 10km diameter asteroid
at Chicxulub Crater would produce impact melt,
not the moderate-pressure shock lamellae. Al-
varez et al. linked all these three anomalies to
differential timing of emplacement of target
rocks. However, we speculate that the double
layer ejecta at KT boundary sediments reflects
two different sources of impact events: the lower
iridium spike and impact spherules layer might
have come from the earlier Shiva impact, whereas
shocked quartz layer was emplaced during the
Chicxulub event. This travel time sorting at two
different impact sites may explain stratigraphic
superposition of the double layer ejecta in North
America and lack of upper layer in the distant
areas of Europe, Asia and Africa. We propose that
the energy released at the Shiva impact site was
hundreds of gigapascals through the collision of
a larger bolide fragment that produced mainly
impact melt, emplacing it around the crater
vicinity. From this impact site, spherules and
iridium spikes were emplaced distally and global-
ly. The high impact pressure would explain why
we could not detect any shocked quartz grains at
the KT boundary sections in India after repeated
searches, as well as a general lack of shocked
quartz in adjacent landmasses in Asia, Europe and
Africa. On the other hand, the impact energy at
the Chicxulub site was moderate to the tune of a
few tens of gigapascals, producing mainly shock-
ed quartz ejecta that were emplaced on the
western side of the crater as the Earth rotated
anticlockwise. Occurrence of multiple iridium
spikes at some KT boundary sections (Officer et
al., 1987) may be attributed to different timing of
arrival from separate sources of impact sites.

Antipodal Crater Pairs Model. The antipodal
locations of the Chicxulub and Shiva craters at
the KT boundary are intriguing and suggest an
alternative scenario (Fig. 8B). A hypervelocity
impact is now known to have important geomor-
phic effects at its antipode. For example, Watts et
al. (1991) documented unusual surface features
such as 'disrupted terrains' antipodal to crater

MEMOIRS OF THE QUEENSLAND MUSEUM

basins on the Moon, Mercury and icy satellites.
Similarly, Rampino & Caldeira (1992) proposed
that antipodal focusing of impact energy may lead
to Deccan volcanism and hotspot activity. Bos-
lough et al. (1994) proposed a new model to
explain how energy from a large impact on the
Earth's surface would couple to its interior and
focus axially at its antipode. However, their
model is based on a vertical impact, whereas most
impacts occur obliquely (Schultz & Gault, 1990).
Because the Shiva Crater is much larger than the
Chicxulub, it is likely that the Shiva Crater is the
primary impact event at the KT boundary,
whereas the Chicxulub may be its antipodal effect
(Fig. 8B). The oblique impact may explain the
departure of a few hundred kilometers from the
true antipodal position. However, the presence of
an iridium anomaly at the Chicxulub antipode is
difficult to explain by the axial focusing
mechanism. Similarly, the Pacific site of collision
is anomalous in this model. The multiple-impact
model is preferred here because of an unusual
concentration of impact deposits and
biogeographic selectivity of extinction at low
latitudes along the Alvarez Impact Belt.

DID THE SHIVA IMPACT TRIGGER
DECCAN VOLCANISM?

DISTRIBUTION AND EXTENT OF DECCAN
VOLCANISM. The close of Cretaceous time was
marked by the outpouring of the enormous Dec-
can lava flows, spreading over vast areas of
western, central and southern India (Fig. 9). The
Deccan Traps cover 800,000km" of west-central
India and extend seaward along more than 500km
of Arabian sea coastline, reaching as far as the
continental shelf and beyond (Devey &
Lightfoot, 1986). Deccan lava flows also spread
across the Seychelles-Saya de Malha Bank, im-
plying that their original extent may be more than
1.5 million km? (Krishnan, 1982; Devey &
Stephens, 1992). They are extremely flat, with
most dips less than 1, and rest mainly on the
Precambrian granitic basement. Significant
departures from horizontal occur, in particular, in
the Western Ghats, west of the Panvel Flexure.
Deep-seismic sounding studies reveal that the
thickness of the Deccan Traps varies from about
100m in the northeastern part to about 2km along
the west coast (Kaila, 1988). Deccan volcanism
is considered to be one of the largest continental
flood basalt deposits in the Phanerozoic (Courtil-
lot et al., 1986).

KT EVENTS IN INDIA

STRUCTURAL FEATURES. Four important
structural lineaments are associated within the
Deccan volcanic province: the east-west trending
Narmada Rift (Choubey, 1971); north-south
trending Cambay rift (Biswas, 1982); northwest-
southeast trending Godavari Rift (Qureshy et al.,
1988); and the arcuate Panvel Flexure (Auden,
1949) (Fig. 9). The Narmada, Cambay and
Godavari Rifts are older geofractures (Biswas,
1987), and may have influenced the distribution
of the Deccan lava basalts. The Panvel Flexure,
on the other hand, is syntectonic with the younger
flows (Wai subgroup) of the Western Ghats Dec-
can volcanism (Auden, 1949) and is interpreted
here as the faulted outer rim of the Shiva Crater
(Chatterjee, 1992), Three volcanic subprovinces
are identified within the Deccan associated with
these rift systems: the Narmada subprovince,
north and south of the Narmada Rift including the
outlier of Rajahmundry in the Godavari Basin;
the Saurashtra subprovince, covering the Deccan
exposures in Saurashtra, west of the Cambay Rift;
the Western Ghats subprovince, including the
thick volcanic sequence east and west of the
Panvel Flexure. The boundary between the Nar-
mada and Western Ghats subprovinces is some-
what diffuse and can be demarcated in the field
by mineralogy and the nature of the lava flow. A
combination of field mapping with petrochemical
and isotopic studies permits division of the thick
lava pile of the Western Ghats section into 3
subgroups (Kalsubai, Lonavala & Wai) and 12
formations, with a progressive decrease in age
from north to south (Beane et al., 1986; Hooper
et al., 1988). Similarly, Subbarao et al. (1988)
recognized 3 new formations in the Narmada
region (Narmada, Manpur & Mhow), but their
relationships with the Western Ghats section is
not clear.

TIMING AND DURATION OF DECCAN VOL-
CANISM. The temporal coincidence of the main
pulse of the Deccan volcanism with the KT
boundary led many authors to believe that a major
asteroid impact might have initiated this massive
volcanism (Alvarez et al., 1982; Rampino, 1987;
Alt et al, 1988; Negi et al., 1993; Chatterjee &
Rudra, 1993). However, critics have pointed out
that the Deccan volcanism started at least a mil-
lion years before the impact event, making the
causal relationship less likely (Courtillot, 1990;
Sutherland, 1994; Alvarez et al., 1994; Bhandari
et al., 1995). The possible link between the KT
extinction and the Deccan flood basalts has
spurred detailed analyses to determine the timing

509

and duration of Deccan volcanism from
radiometric, palaeomagnetic and palaeontologic
constraints. We have also sampled various Dec-
can stratigraphic sections to determine the KT
boundary event layer. Here we synthesize all
available data to estimate the absolute age and
age span of the Deccan volcanism in the context
of its origin and subsequent influence on the
biotic crisis at the KT boundary.

Geochronology. Because of the altered nature of
the Deccan basalt, previous attempts to determine
its age by K/Ar method were unsatisfactory, with
results ranging from 102 to 30Ma (Alexander,
1981). However, recent “Ar/*?Ar dates of the
stratigraphically controlled thick sequence of the
Western Ghats section cluster around a narrow
span of age from 69 to 64Ma, with a major
eruptive phase around 65Ma (Duncan & Pyle,
1988; Courtillot et al, 1988; Venkatesan et al.,
1993; Baksi, 1994). Although a close temporal
correspondence between the main pulse of Dec-
can volcanism and the KT boundary is indicated
by isotope dating, early phases of eruption may
have started at least 1Ma before the KT event.

Palaeomagnetism. Recent palaeomagnetic
studies indicate that only 3 magnetic chrons
(30N, 29R, & 29N) are represented in the thick
Deccan lava pile, where the main eruptive phase
in the Western Ghats section corresponds with
chron 29R (Gallet et al., 1989; Courtillot, 1990;
Vandamme & Courtillot, 1992). This normal-
reversed-normal (NRN) magnetostratigraphy
appears to be a powerful tool in correlating wide-
ly separated Deccan basalt provinces.
Palaeomagnetic results support a shorter span of
Deccan volcanism, from 67 to 64Ma, centered
around chron 29R (65Ma). In this case, the whole
eruptive history would cover only 1-2Ma.

Palaeontology. Palaeontologic evidence comes
from thick fossiliferous sedimentary beds as-
sociated with the basaltic flows. These interbeds
are traditionally named according to their physi-
cal position relative to the basal Deccan flow;
they may underlie the flows (infratraps, such as
the Lameta Beds) or they may be intercalated
within the flows (intertraps). Stratigraphically
infratraps are believed to be older than the inter-
traps (Krishnan, 1982, p. 415). This distinction is
not clear-cut in regional biostratigraphic analysis.
Local infratrappean beds may appear intertrap-
pean in large-scale mapping. Sahni & Khoshla
(1994) proposed a neutral term, ‘Deccan basalt
volcano-sedimentary sequence' (DBVSS) for
these trappean beds. These sedimentary beds
occur marginal to the Deccan outcrops in the

510

Narmada and Saurashtra subprovinces, indicat-
ing that these vast thicknesses of lava flows were
not extruded all at once; volcanic activity was
punctuated periodically. In between the flows are
fluvial or lacustrine deposits of trappean beds that
contain abundant remains of plants, inver-
tebrates, dinosaurs and their eggs, We made ex-
tensive sampling of these trappean beds in search
of the iridium anomaly, shocked quartz and tek-
tites to detect the KT boundary layer, but the
results were negative, reinforcing the palaeon-
tologic observations that most of these trappean
sediments are older than the KT boundary.
Palaeontologic evidence drawn from palynoflora
Aquilapollenites, charophytes, non-marine
ostracods, the selachian /gdabatis, pelobatid
frogs, anguid lizards, booid snakes, pelomedusid
turtles, abelisaurid, titanosaurid and ankylosaurid
dinosaurs and palaeoryctid mammals indicate a
Maastrichtian age for these trappean beds, but the
upper limit is unknown (Sahni & Bajpai, 1988;
Chatterjee, 1992). More precise data comes from
oil-exploration wells in the Godavari Basin,
where the Narsapur well has encountered KT
boundary flows (Govindan, 1981). Here the
lower marine intertrap has yielded the planktonic
zone fossil Abathomphalus mayaroensis of Late
Maastrichtian age, whereas the upper post-trap-
pean bed contains Globorotalia praecursoria of
the P2 planktonic foraminifera zone of Early
Palaeocene age. Palaeontological data define
broad limits of Deccan volcanism from 67.5 to
60.5Ma in the Godavari Basin.

Stratigraphic Calibration. A schematic correla-
tion of Deccan Trap sequences between the
western Ghats and the Narmada/Saurasthra sub-
provinces is shown in Fig. 10, combining
radiometric, palaeomagnetic and palaeontologic
data. The following observations can be made:

1) The volcanism north of the Narmada Rift
started somewhat earlier than the southern part.
The Lameta beds are restricted to the north of this
lineament and provide palaeontological control
on the onset of volcanism. So far, only one mag-
netic chron, 30N, has been identified in this
region associated with the infratrap (Sahni &
Bajpai, 1988). In the Bara Simla section of Jabal-
pur, north of the rift zone, the Deccan flow over-
lying the dinosaur-bearing Lameta Group shows
normal polarity (30N). This flow has not yet been
dated by the Ar-Ar technique. Dinosaur bones,
nesting sites, microvertebrates, palynoflora
Aquilapollenites and the selachian /gdabatis sug-
gest a Maastrichtian age for the Lameta Beds
(Chatterjee, 1992; Jaeger et al., 1989; Sahni &

MEMOIRS OF THE QUEENSLAND MUSEUM

Bajpai, 1988; Sahni et al., 1994). Sauropod nest-
ing sites occur in a specific lithotype, a pedogeni-
cally modified sandy carbonate that forms a
distinct marker bed. This egg-bearing bed can be
traced for almost 1,000km, from Balasinor,
Dohad and Hathni to Jabalpur in the northern part
of the Narmada Rift (Mohabey, 1984, 1987;
Srivastava et al., 1986; Sahni et al., 1994).

2) The 670m thick volcanic sequence at Kal-
ghat-Mhow region, north of the Narmada linea-
ment, lacks infratraps but shows three trap
formations (Narmada, Manpur and Mhow) with
normal (30N) and reverse (29R) chrons (Sub-
barao et al., 1988; Vandamme & Courtillot, 1992)
and can be equated with the Western Ghats sec-
tion.

3) The Saurashtra subprovince is not well con-
strained in age by palaeontological and
palaeomagnetic analysis. Here the Deccan Traps
overlie the 50m marine sequence of Late
Cretaceous Wardhan Member and can be corre-
lated with the volcanics of the northern part of the
Narmada Rift (Biswas, 1988).

4) South of the Narmada lineament, in the
Nagpur-Umrer-Dongargaon intertrappean beds,
a thick sequence of reverse polarity (29R) is
followed by a normal sequence (29N) (Sahni &
Bajpai, 1988). In these intertraps dinosaur bones,
fragmentary eggshells and microvertebrates of
Late Maastrichtian are common. In Dongargaon
the basal flow has yielded a precise Ar-Ar plateau
age of 66.4+1.9Ma (Duncan & Pyle, 1988).

5) Vadamme & Courtillot (1992) reported a
29R-29N reversal chron from the Rajahmundry
outcrop. Geochronological studies suggest an age
of ~64Ma coeval with 29N chron (Baksi et al.,
1994). In the Narsapur well, trappean beds can be
tied to two foraminiferal zone boundaries in the
type KT boundary sections of Italy (Alvarez et
al., 1987; Govindan, 1981): the lower, infra-trap
bed correlates with the Abathomphalus mayoren-
sis zone (Maastrichtian), while the upper, post-
trap bed corresponds with the Globorotalia
praecursoria of P2 zone (Early Palaeocene).

6) The thick flows of Deccan volcanism in the
Western Ghats sections show three subgroups
(Kalsubai, Lonavla, and Wai) and are well-con-
strained by radiometric and palaeomagnetic data;
volcanic activity was short-lived and reached its
major peak of activity during chron 29R
(~65Ma), followed by a short interval of chron
29N. Lack of trappean beds makes it impossible
to estimate the age of Western Ghat sections by
palaeontological methods.

KT EVENTS IN INDIA

7) Adefinite KT boundary section with iridium
anomaly has been identified recently in the inter-
trappean beds of Anjar in Kutch, where
radiometric ages of the traps cluster around 65Ma
(Bhandari et al., 1994). As discussed earlier, we
believe this section of alkaline basalt flows is
directly related to the impact event.

8) The duration of Deccan volcanism may
range from 67.5 to 60.5Ma (palaeontologic con-
straints), from 69 to 64Ma (geochronologic con-
straints), or from 67 to 64Ma (palaeomagnetic
constraints). Thus the minimum age range of
eruption may be around 67Ma to 64Ma, about
3My in duration.

9) Because of the presence of foraminiferal
zones and magnetic reversal chrons, the Deccan
stratigraphic sequence can be equated with the
magnetostratigraphic type section of the KT
boundary at Gubbio, Italy (Alvarez et al., 1987).

ORIGIN OF DECCAN VOLCANISM. A cause-
and-effect connection between impact and Dec-
can volcanism has been the subject of extensive
discussion and speculation. If the Deccan vol-
canism started 1Ma before the KT boundary
event, and extended over 3Ma, as combined
evidence of isotopic dating, magnetic anomalies
and palaeontology suggests, then the Shiva im-
pact did not initiate the Deccan volcanism. Dec-
can volcanism predated the impact event
(Courtillot, 1990; Sutherland, 1994; Bhandari et
al., 1995). Impact and Deccan volcanism are in-
dependent, having occurred by chance at about
the same time. Deccan volcanism was already
active when when the KT mpact occured near the
Bombay coast. However, the impact might have
shaken the Earth’s mantle violently to enhance
the spectacular Deccan outburst precisely at the
time of the KT boundary.

The spatial and temporal coincidence of Dec-
can volcanism with the Carlsberg Rift and the
Reunion hotspot activity at the KT boundary
sheds critical insights into its origin. The enor-
mous thickness of the lava pile in the Western
Ghats sections associated with compound flows
and ash beds indicate that the major eruptive
source for Deccan volcanism must be located
near the Bombay area, where evidence of both
hotspot and rift magmatism are present. There has
always been controversy as to whether the plume
or rifting was the initiating factor for the Deccan
volcanism. This conflict can be resolved if we can
determine accurately the timing of initial eruption
and duration of Deccan volcanism. Morgan
(1981) proposed that the Deccan flood basalts

were the first manifestation of the Reunion
hotspot that subsequently produced the hotspot
trails underlying the Laccadive, Maldive and
Chagos islands; the Mascarene Plateau; and the
youngest volcanic islands of Mauritius and
Reunion, Recent DSDP data confirm that the age
of the volcanism decreases from north to south,
from the Deccan to the Reunion hotspot (Back-
man et al., 1988). Thus the geometry and the age
range of these volcanic provinces, islands and
submarine ridges are consistent with the rapid
northward motion of the Indian plate over a fixed
hotspot (Morgan, 1981; Duncan & Pyle, 1988).

Although the hotspot model is very attractive
in explaining the Deccan flood basalt volcanism
and linear volcanic chains of the western Indian
Ocean, there are distinctions in both trace element
and isotope geochemistry between present-day
Reunion eruptives and those of the Deccan
province; the likely source of the Deccan vol-
canism is similar to rift volcanism rather than the
Reunion hotspot (Mahoney, 1988). Further
geochemical and geothermal evidence suggests
that Deccan magmas were generated at relatively
shallow (35-45km) depth in Mid-Ocean Ridge
Basalt (MORB) mantle and rules out the pos-
sibility of its origin by a deep mantle plume (Sen,
1988). However geophysical evidence indicates
that the continental crust was extremely thin in
the Western Ghats section under the plume (Negi
et al., 1993). Morever, Ellam (1992) showed con-
vincingly that the thinned lithosphere of Western
Ghats is the reason for this trace element dis-
crepancy between the Deccan volcanism and the
Reunion hotspot.

Was rifting triggered by doming above the
Reunion hotspot? Some workers (White & Mc-
Kenzie, 1989; Hooper, 1990) argued that the
Reunion hotspot actually created the Carlsberg
rift along which Deccan volcanism erupted.
However, as discussed earlier, the Carlsberg rift-
ing did not start before chron 29R, whereas Dec-
can volcanism started somewhat earlier around
30N. If the Sarnu-Dandali and Mundawara vol-
canics of Rajasthan are regarded as the earliest
manifestation of the Deccan volcanism in Penin-
sular India and the initial location of the Deccan-
Reunion hotspot (Basu et al., 1993), then the
Deccan volcanism must have started 3.5 million
years earlier than the timing of the Carlsberg Rift,
making the causal link unlikely. Moreover, if the
Carlsberg Rift was triggered by the Deccan-
Reunion hotspot, its geographic location would
be at the center of the Shiva Crater, offshore of
the Bombay coast. However, the hotspot track

512

indicates that the Reunion
hotspot always lay farther
east within the Indian con-
tinent, probably near Igat-
puri at the KT boundary
(Fig. 9). If we consider the
Rajasthan volcanics as the
earliest and northernmost
activity of the Reunion
hotspot, it would be at least
500km northeast from the
Carlsberg Rift at the time of
eruption. Thus the timing of
the eruption and the loca-
tion of the Reunion hotspot
do not suggest any close
link between plume genera-
tion and rifting. On the
other hand, at the KT
boundary time, the impact
had been coincidentally
close enough to the
Reunion hotspot to activate
the major phase of the vol-
canic outbursts (Figs. 7A,
9). Reviewing all the
evidence, the Deccan-
Reunion hotspot remains
the best model for the origin
of Deccan volcanism.

How did the Deccan lava
cover such an enormous
area of India? The intercon-
nected rift basins may be
implicated in the distribu-
tion of Deccan lavas. Prior
to the onset of Deccan vol-
canism the palaeodrainage
of the Narmada and
Godavari Rivers was
directed toward the Bay of
Bengal (Krishnan, 1982, p. 17). The Cambay Rift
basin was tilted northward and westward at that
time (Biswas, 1987). This centripetal pattern of
drainage system was further accentuated by
doming of the Western Ghats section around Igat-
puri by the uprising plume (Fig. 9). Lava
generated from the Deccan-Reunion hotspot
flooded the Narmada, Cambay and Godavari rift
basins. These lava rivers traveled many hundreds
of kilometers in all directions with occasional
flooding in the overbank areas. The main reason
these flows could travel such great distances is
their unusually large volume and rapid rate of
eruption, coupled with their low viscosity. The

INDIAN OCEAN

Deccan Traps.

^ CAMBAY RIF
fe} O|

SEYCHELLES
j -Bombay

RR |

MEMOIRS OF THE QUEENSLAND MUSEUM

4
`
`
`
‘
68.5 Je Sarnu-Dandall
\

Sa
^
he

5 so Narayanpur )

? C Jabalpur tm

KITE T NARMADA RIFT NM b-
wo Agir eM -

>i WM o Siolondi
* Igatpuri

E Pisdura
Dongargaon
BAY OF BENGAL.

<y] Deccan Traps

t| Precambrian Granite

o Dinosaur localities

200 km

FIG. 9. Sketch map showing the localities of Maastrichtian dinosaurs around the
Deccan volcanic province before the impact event; microcontinent Seychelles
was adjacent to India; asterisks indicate the Deccan-Reunion hotspot track
during the rapid northward drift of India. During Maastrichtian, the Deccan-
Reunion hotspot was located near Igatpuri forming a domal structure with
centripetal distribution of lavas; bold arrows indicate the possible directions of
intercanyon flows of Deccan lavas along the drainage of the Narmada, Godavari
and Cambay basins. Intercanyon flows may explain large areal distribution of

final distribution of the Deccan Traps may reflect
the topography at the time of distribution and the
palaeosplope of these interconnected rift basins
(Fig. 10). Similar long-distance intercanyon
flows are known from the Columbia Plateau
basalt of western United States, where the
Pomona Member of the Saddle Mountain Basalt
flowed down the ancestral Columbia River from
north-central Idaho to the Pacific Coast (Hooper,
1992).

MAASTRICHTIAN DINOSAURS OF INDIA

Although dinosaur (implying nonavian
dinosaur throughout the text) extinction was a

KT EVENTS IN INDIA

THANETIAN

ul
z
iu
o
o
u
-
x
m

KT BOUNDARY

MAASTRICHTIAN

SENONIAN

DECCAN TRAPS

Upper Range ol Volcanism

--«- Lower Range of Volcanism

C40 ar38Ar

EN Normal
Reverse

TRAPPEAN SEDIMENTS,

FIG. 10, Stratigraphic correlation of Deccan volcanic provinces in relation to KT boundary by geochronologic,
palaeomagnetic and palaeontologic constraints. The minimum age span of Deccan volcanism may range from
67Ma to 64Ma; (data from Vandamme & Courtillot, 1992; Harland et al., 1989; Jaeger et al., 1989; Subbarao
etal., 1988; Sahni & Bajpai, 1988; Chatterjee, 1992; and personal observations).

global phenomenon, there are very few places in
the world where dinosaur-bearing sediments
document the crucial KT boundary events; most
of these exposures are known only from western
North America (Wyoming, Montana, Alberta,
Saskatchewan, Colorado, New Mexico and
Texas), Mongolia and India. As a result, the
tempo of dinosaur extinction, whether sudden or
gradual, has been debated intensely (Clemens,
1982; Sloan et al., 1986; Sheehan et al., 1991). In
this context, Indian biostratigraphic evidence
may play acrucial role, since most of the dinosaur
bones are known almost exclusively from the
Deccan volcano sedimentary sequence such as
Lameta Group which was deposited close to the
KT boundary time. In the KT boundary section of
Anjar, Gujrat, the last dinosaur bones occur
precisely at the iridium layer. This is probaly a
unique biostratigraphic evidence for a sharp trun-
cation of dinosaurs supporting the sudden extinc-
tion event.

Maastrichtian dinosaurs from India are ex-
tremely rare. Some fragmentary remains of
dinosaurs are also known from the Kallamedu
Formation of Tamilnadu, South India (Fig. 9).
Most of the Lameta dinosaurs came from a single
quarry on the western slope of Bara Simla Hill of
Jabalpur. This quarry in 1917-19 produced
various dinosaurs including titanosaurids,
coelurids, ornithomimids, allosaurids and
ankylosaurids (Huene & Matley, 1933). Although
the fauna from the Lameta Group indicates a
moderate diversity of dinosaur community that

inhabited India as it collided with Asia near the
very end of the Cretaceous, most of these bones
are fragmentary limb, girdle and vertebral ele-
ments which have few diagnostic characteristics.
Furthermore, all of the previous Lameta dinosaur
collection, housed at the Geological Survey of
India, Calcutta and the Natural History Museum,
London, has been missing for the last 50 years.
Thus the affinity of Indian Maastrichtian
dinosaurs has never been thoroughly investigated
in recent time because of lack of material. The
only valuable specimen remaining from the early
collection is a partial skull of a juvenile theropod
Indosuchus, now at the American Museum of
Natural History (AMNH 1955), collected by Bar-
num Brown in 1922 and subsequently described
by Chatterjee (1978).

During the past thirty years, many attempts had
been made to locate this important quarry at Bara
Simla Hill without any success. The site was
covered with thick vegetation, and no landmarks
or quarry maps of the original excavation survive.
In 1988, we found this classical site and began
exploring. A preliminary excavation has un-
covered the original bone beds with sauropod and
theropod remains, but most of the new material is
yet to be prepared. We collected a partial skeleton
of a large titanosaurid associated with limbs,
girdles, vertebrae and a beautiful braincase. The
large size (femur more than a meter long), un-
divided cervical neural spine, and strongly
procoelous caudal vertebrae establish its
titanosaurid affinity. In addition, several cranial

514

fragments of Indosuchus were also recovered. In
1995, we discovered a rich graveyard of dinosaur
bones in the mudstone facies of the Lameta Group
near Raiholi village of Gujrat, western India. The
most remarkable discovery from this quarry rep-
resents a nearly complete skeleton of a large
theropod, about the size of Allosaurus. We have
also collected various bones of titanosaurs and
ankylosaurs from this quarry. Abundant sauropod
eggs were also found from adjacent Dohad
locality (Fig. 9). Finally, we have traced the miss-
ing dinosaur bones from Jabalpur described by
Huene & Matley (1933) at the Indian Museum,
Calcutta, last year. With this new information, we
hope to get a first good look at the last dinosaurs
in India, Here we provide a preliminary descrip-
tion of the new finds to evaluate the systematics
of Indian Maastrichtian dinosaurs.

SAURISCHIAN DINOSAURS.

Sauropoda: Family Titanosauridae Lydekker,
1885. Titanosaurids were first recognized from
the Late Cretaceous of India by Richard Lydekker
in 1877 on the basis of their peculiar procoelous
caudal vertebrae, and were later found in contem-
poraneous beds in South America, North
America, Africa, Madagascar, Europe and Asia.
They were among the largest dinosaurs that ever
lived, but their origin and relationships are poorly
understood, in part because no good skulls are
known. They were the dominant herbivores of
Gondwana during the Cretaceous, but radiated
into Laurasian continents. The titanosaurids are
characterized by vertebral attributes, especially
by short cervicals, undivided neural spines in
anterior dorsals, 6 sacrals and procoelous caudals
(McIntosh, 1990). One of the most intriguing
features of titanosaurids is the presence of body
armor in the form of bony scutes and knobby
dermal bones, known from Saltasaurus from Ar-
gentina (Bonaparte & Powell, 1980). There is no
doubt that the dermal ossicle collected by Bar-
num Brown (AMNH 1359) from the Bara Simla
locality, which was attributed to stegosaurs
(Huene & Matley, 1933) and ankylosaurs
(Coombs, 1978), actually belongs to titanosaurids.

Huene & Matley (1933) described two
titanosaurid genera from the Lameta Group,
based on size differences: the gracile-
Titanosaurus indicus, Lydekker and a new robust
species, Antarctosaurus septentrionalis. They
figured a partial braincase of Antarctosaurus
which compares well with A. wichmannianus of
South America. However, the braincase of Indian
Titanosaurus was unknown at that time. Later,

MEMOIRS OF THE QUEENSLAND MUSEUM

Berman and Jain (1982) described the braincase
of a small titanosur from the Lameta Group at
Dongargaon and concluded that new braincase is
fundamentally different from Antarctosaurus.
The associated postcranial material collected
from the site, especially vertebrae, resembles that
of Titanosaurus indicus (Jain, 1989). It is likely
that the Dongargaon braincase belongs to T. in-
dicus (Jain, pers. comm.). Thus, from the brain-
case morphology and postcranial characteristics,
the coexistence of at least two titanosaur genera,
Antarctosaurus and Titanosaurus, in the Lameta
Group can be established. This conclusion is fur-
ther strengthened with the discovery of two addi-
tional braincases, one belonging to Titanosaurus,
the other to Antarctosaurus, which are described
below. Braincases are conservative which makes
them excellent tools for identifying taxa; they are
less susceptible to convergences that characterise
skeletal features associated with feeding and
locomotion. Using braincase morphology as a
guide, the distinction between Titanosaurus and
Antarctosaurus appears very clear-cut. In
Titanosaurus the basipterygoid processes are ex-
tremely short, reduced and lie almost at the level
of basal tubera; moreover, the parocicipital
processes are wide and moderately curved
ventrally. In Antarctosaurus, the basipterygoid
processes are very long, slender and directed
considerably ventrally below the level of the
basal tubera. The paroccipital processes in this
genus are narrow and highly curved downward.
In sauropods, the nature of the basipteryoid
process is intimately linked to the attitude of the
quadrate; the short basipterygoid process is as-
sociated with the vertical quadrate, whereas the
long basipterygoid process indicates highly slant-
ing quadrate.

Antarctosaurus septentrionalis Huene & Mat-
ley, 1933. The new braincase (ISI R 162)
recovered from Bara Simla site is beautifully
preserved and provides a wealth of anatomical
information. All the bones are tightly sutured so
demarcation of individual elements is difficult.
Although the braincase described by Huene and
Matley (1993, Fig. 5) as Antarctosaurus sep-
tentrionalis is fragmentary, the preserved part
shows close resemblance to the new material,
especially in the construction of basipterygoid
process and paroccipital process. The new
specimen probably belonged to a young in-
dividual, as it lacks the laterosphenoid, orbitos-
phenoid and skull roof. In this respect, this
specimen is very similar to that of a French

KT EVENTS IN INDIA

515

FIG. 11. Braincase of Antarctosaurus septentrionalis, a sauropod from the Lameta Group of Jabalpur, based on
ISI R 162; A, caudal view; B, lateral view; C, rostral view; D, ventral view; E, dorsal view of basioccipital
showing pons varioli. bo=basioccipital, bpt=basipterygoid process, bs=basisphenoid, bt=basal tubera, eo=ex-
occipital, fo=fenestra ovalis, ic=internal carotid artery, mf=metotic foramen, op=opisthotic, pif=pituitary fossa,
popr=paroccipital process, pro=prootic, soc=supraoccipital; foramina for cranial nerves in Roman numerals.

titanosaurid (Loeuff et al., 1989), which also
lacks the anterior and dorsal parts,

The braincase is short and extremely deep so
that the basioccipital and the basisphenoid are
telescoped (Fig. 11). The most outstanding fea-
ture is the highly enlarged basipterygoid proces-
ses, which are slender and directed downward as
in several other titanosaurid genera such as
Amargosaurus, Nemegtosaurus and Saltasaurus,
(McIntosh, 1990). On the lateral wall of the basip-
terygoid process, at the level of the basal tubera,
the openings for the canals of the internal carotid
artery are visible. These tunnel through the bone
and emerge at the base of the pituitary fossa
through separate openings. The pituitary fossa is
deep on the dorsal surface of the basisphenoid.
The posterior wall of the fossa, the dorsum sellae,
is pierced laterally by a pair of openings for the
abducens (VI) nerves. On the lateral wall of the
braincase, there are two prominent foramina at
the middle ear region: the caudal one, the metotic
foramen is very large and transmitted nerves IX-

XI and the caudal branch of the jugular vein; the
rostral one, the fenestra ovalis, is relatively small
to receive the stapes. Rostral to the fenestra ovalis
is a small aperture indicating the outlet for the
facialis (VII) nerve. Farther rostrally each prootic
is notched by the trigeminal foramen (V) which
would be enclosed by the laterosphenoid. Both
laterosphenoid and orbitosphenoid bones are
missing in our specimen. In the occiput, the
strong nuchal crest on supraoccipital and
obliteration of sutures between the exoccipitals,
the supraoccipital and the basioccipital, can be
seen. The paroccipital processes are narrow,
wing-like structures extending outward and con-
siderably downward. The occipital condyle is
large and spherical. Each exoccipital bone near
the lower rim of the foramen magnum is pierced
by a foramen for the hypoglossal (XII). At the
floor of the braincase, the basioccipital shows a
median cavity for the pons variolii (Fig. 11).

Titanosaurus indicus Lydekker, 1877. Berman
& Jain (1982) described a sauropod braincase (ISI

516

MEMOIRS OF THE QUEENSLAND MUSEUM

bpt

FIG. 12. Braincase of Titanosaurus indicus, a sauropod from the Lameta Group of Raiholi, based on ISI R 467.
A, caudal view; B, lateral view; C, rostral view. Abbreviations as in Fig. 11; f=frontal, Is-laterosphenoid,
os-orbitosphenoid, p=parietal, ps=parasphenoid rostrum.

R 199) from the Dongargon which was later
assigned to Titanosaurus indicus (Jain, pers.
comm.). The specimen is not well-preserved
which makes the detailed interpretation some-
what difficult. This deficiency is remedied by the
discovery of a better specimen from Raiholi site
(ISI R 467). The new specimen is very similar to
the Dongargaon specimen in size and morphol-
ogy (Fig. 12), but fundamentally different from
that of Antarctosaurus.

Asin all sauropods, the braincase is completely
ossified. The occipital condyle is hemispherical
except for being somewhat concave dorsally to
give a kidney-shaped appearance. Ventrally, the
basal tubera are subdued and completely fused
with the basipterygoid processes. The basip-
terygoid processes are very small and closely
appressed. The foramen magnum is somewhat
oval and smallerthan the occipital condyle. At the
base of the foramen magnum, each exoccipital is
pierced by a single canal for cranial nerve XII. In
Dongargaon specimen, there is a prominent

protuberance on the occipital bone representing
the articular facet of proatlas. This feature is
lacking in our specimen. The opisthotic is in-
timately fused with the exoccipital; it is directed
laterally and somewhat ventraly as a robust wing
of paroccipital process. The supraoccipital is fair-
ly massive and forms the dorsal roof of the
foramen magnum. Laterally it is overlapped by a
ventral flange of the parietal.

In lateral aspect, the cranial foramina are
beautifully preserved. In the middle ear region
two foramina are visible; the rostral, smaller one
is the fenestra ovalis; the caudal, large one is the
metotic foramen for nerves IX-XI. Farther
rostrally the large foramen, shared between the
prootic and laterosphenoid, is the trigeminal
opening for nerve V. Ventral to it, the entrance of
the internal carotid artery to the pituitary fossa
can be seen at the base of the basipterygoid
process. Rostral to trigeminal foramen, three
foramina are visible in a vertical row in the latero-
sphenoid-orbitosphenoid-basisphenoid complex.

KT EVENTS IN INDIA

The dorsal foramen indicates the exit for nerve
IV, the middle one for nerve III and the ventral
one for nerve VI. At the base of the parasphenoid
rostrum a large opening within the orbitos-
phenoid indicates the lateral aperture for nerve II.
In rostral aspect the fused orbitosphenoid-basis-
phenoid complex narrows considerably to form a
median vertical ridge as seen in a Romanian
titanosaurid specimen (Weishampel et al., Fig.
15). Dorsal to the parasphenoid rostrum, the large
cavity indicates the pituitary fossa.

Theropoda. Huene & Matley (1933) described
two large predatory dinosaurs from the Lameta
Group of Jabalpur. /ndosuchus raptorius and In-
dosaurus matleyi as late survivors of allosaurs.
Both taxa are known from partial skull fragments
as well as several postcranial elements. The dis-
tinction between Indosuchus and Indosaurus was
based on the structure of the fronto-parietal
region. In Indosuchus, the fronto-parietal region
has a narrow crest as seen in tyrannosaurs. The
skull roof is flat between the orbits and the
postfrontal is smooth. In Indosaurus, on the other
hand, the parietal is broad, the lower surface of
the frontal is wide and the transverse crest lies
above and below the orbit. The frontals are con-
cave and decline in front. The supratemporal
fossa is short and broad as in allosaurs. Huene &
Matley also described a maxilla of Indosuchus
which is identical to a similar element collected
by Barnum Brown in 1922 from the Lameta
Group (AMNH, 1955). Later, Chatterjee (1978)
described partial jaws of AMNH specimens of
Indosuchus as a juvenile tyrannosaurid, partly on
the basis of narrow fronto-parietal crest, elon-
gated supratemporal fossae, incisiform premaxil-
lary teeth and similar dental formula. He accepted
Indosaurus as an allosaurid.

The classification of Indousuchus and In-
dosaurus has been difficult in the past because of
fragmentary material. Despite marked
similarities to tyrannosaurs, Indosuchus is char-
acterized by the absence of preantorbital
fenestrae in the maxilla. Bonaparte & Novas
(1985) described a large skull of a new theropod,
Abelisaurus comahuensis, from the Maastrich-
tian Allen Formation of Rio Negro, Argentina.
Later, Bonapart et al. (1990) described a virtually
complete, articulated skeleton of another
theropod, Carnotaurus sastrei, from the Middle
Cretaceous Gorro Frigio Formation of Argentina.
Carnotaurus shows unusual frontal horns. Its for-
limbs are highly reduced, with extremely short
radius and ulna. Bonaparte and associates placed

517

Abelisaurus, Carnotaurus and Indosuchus in the
same family, Abelisauridae. This family is
defined by large infratemporal fenestra, elon-
gated quadrate, posteriorly directed squamosal
with a ventral, rod-like process and a small max-
illay fenestra near the preorbital opening. Com-
parison of Abelisaurus and Indosuchus clearly
indicates close similarity between these two
genera. Indosuchus is not a tyrannosaurid as was
supposed earlier, but possibly an abelisaurid. Al-
though abelisaurs were the dominant predators
throughout Gondwana during the Cretaceous
Period, our knowledge of these enigmatic
theropods is still limited. The occurrence of a
supposed abelisaur in the Late Cretaceous of
France has been recognized recently (Buffetaut et
al, 1988). Although abelisaurids superficially
resemble large predatory dinosaurs such as al-
losaurs and tyrannosaurs, they are more primi-
tive, probably a separate lineage evolving from
ceratosaurs.

Various small coelurosaurs described from the
Bara Simla quarry by Huene & Matley (1933),
such as Jubbulpuria tenuis and Laevisuchus in-
dicus are intriguing. We need additional material
to assess their relationships.

Family Abelisauridae Bonaparte & Novas, 1985.
Indosuchus raptorius Huene, 1933. The type
specimen of Indosuchus raptorius, now housed
at the Indian Museum, Calcutta is allocated to this
species. Some of our recent finds may also belong
to this species. For example, we have collected
additional cranial bones of Indosuchus, such as
the lacrimal, jugal and posterior part of the jaw
(ISI R 163) from the Bara Simla Hill of Jabalpur.
Furthermore, we have unearthed a nearly com-
plete skeleton of Jndosuchus from Raiholi site of
Gujrat. The teeth are compressed laterally and
serrated; the tooth crown is extremely low; the
ratio of the crown height to rostro-caudal width
is 1.5. The vertebrae are amphicoelous but lack
pleurocoels. The scapulocoaracoid is narrow with
a prominent acromion process. The forelimb is
short relative to the hindlimb; the ratio to femoral
to humeral length is about 2. Unlike Carnotaurus
and Tyrannosaurus, the forelimbs of Indosuchus
were as along as in allosaurs. Unfortunately, the
postcranial elements of Abelisaurus are un-
known. The iliac blade of Indosuchus has a deep
preacetabular and long postacetabulr process; the
pubis is expanded distally to form a foot. The
hindlimb bones are hollow, thin-walled, stout and
resemble the corresponding elements of Car-
notaurus. The femur has a spherical, inturned

518

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 13. Skeletal restoration of Indosuchus raptorius, a theropod from Raiholi site, showing preserved bones;
skull modified from Chatterjee (1978). Restoration is based on disarticulated elements (ISI R 401 - 454).

head with a distinct neck. The lesser trochanter is
fairly well-developed but lower than the greater
trochanter. The tibia is robust with a distinct
cnemial crest; distally it remains unfused with the
astragalus. The astragalus is hemicylindrical in
form, but its ascending process is relatively short.
A preliminary skeletal reconstruction of In-
dosuchus is shown in Fig. 13 and restoration of
the skull in Fig. 14.

Indosaurus matleyi Huene 1933. Another
theropod partial skull from the Bara Simla quarry,
Indosuchus matleyi, is generally allied to Al-
losauridae (Huene & Matley, 1933; Chatterjee,
1978), but Molnar (1990) pointed out some at-
tributes in Indosaurus similar to those of Car-
notaurus of Argentina and allocated it to
Abelisauridae. Recently we have collected a Car-
notaurus-like atlas-axis complex from the
Lameta Group of Raiholi area, which strengthens
Molnar's assessment.

ORNITHISCHIAN DINOSAURS. Various or-
nithischian fragments have occasionally been
reported from the Late Cretaceous of India, but
their identity is dubious except for the
ankylosaurian Lametasaurus.

Ankylosauria. Lametasaurus indicus Matley
1923. Lametasaurus is known from the sacrum
ilia, tibia, spine and armor (Matley, 1923; Huene
& Matley 1933). It is a fossil chimera of mistaken
association containing parts of two different
dinosaurs: armour fragments often attributed to
stegosaurs, nodosaurs (Huene & Matley, 1933),
or ankylosaurs (Coombs & Maryanska, 1990),
whereas sacrum, ilia and tibia probably belong to

theropods (Chakravarti, 1935), such as
abelisaurids (Molnar, 1990). Yet Huene & Matley
mentioned, but did not figure, two additional
types of dermal scutes from the same bed that
could be ankylosaurian (Galton, 1981). Recently
we have recoverd definite remains of an
ankylosaur from the Raiholi site. The new
material represents isolated vertebrae,
scapulocoracoid, humerus, femur and several
pieces of armor such as hollow lateral spikes and
solid dorsal scutes, typical of ankylosaurs, which
will be described in a separate paper.

Stegosauria nomen dubium. Dravidosaurus
blanfordi Yadagiri & Ayyasami (1979). A small,
late surviving stegosaur was described from the
marine Cretaceous Kossmaticeras theobal-
dianum Zone Trichinopoly Group by Yadagiri &
Ayyasami (1979). In 1991, we visited the site and
found only fragmentary remains of plesiosaurs.
We also examined the holotype and could not see
anything related to the stegosaurian plates and
skull claimed by these authors. Instead, the bones
are highly weathered limb and girdle elements
and may belong to plesiosaurs.

DISTRIBUTION OF DINOSAURS. It appears
from the above discussion that Indian Late
Cretaceous dinosaurs remain rather poorly
known. Fossil records indicate that they lived
around the fringe of the Deccan volcanic
province and adapted to the harsh environments
(Fig. 9). Yet this distribution reflects an artifact of
preservation because of unusual protection of
intertrappean beds by the Deccan basalts. Most
of the nonmarine Cretaceous deposits in penin-

KT EVENTS IN INDIA

sular India were removed by
erosion except for the little
patch in the Kallamedu area
near Madras. As a result, the
extent of dinosaur distribution,
other than in the Deccan vol-
canic province, is unknown.
Although Masstrichtian
dinosaurs have been reported
from all Gondwana continents
including Antarctica, reasonably
complete skeletons have been
recovered only from South
America. The disjunct distribu-
tion of the titanosaurid-
abelisaurid assemblage in
South America, India and
Europe is interesting in the
context of drifting continents in
Late Cretaceous time. Palaeon-
tologic evidence suggests that
the India/Eurasia collision took
place during this time, facilitat-
ing the migration of northern
fauna to India (Chatterjee,
1992; Jaeger et al., 1989;
Prasad et al., 1994). Isolation
between the northern and
southem continents produced
dramatically different distribu-
tions among dinosaurs. In contrast to the nothern
hemisphere, the dominant herbivores in the Late
Cretacous of India and South America were
titanosaurids rather than ornithischians, whereas
the large predators were abelisaurids instead of
tyrannosaurs (Fig. 15).

SAUROPOD NESTING SITES. Sauropod nest-
ing sites from the Lameta Group underlying the
Deccan volcanics represent one of the most ex-
tensive fossil hatcheries in the world (Mohavey
& Mathur, 1989; Srivastava et al., 1986; Sahni et
al., 1994), They can be traced almost continuous-
ly along the north of Narmada Rift from Anjar to
Jabalpur for more than 1,000km, wherever the
Lameta Group is exposed (Fig. 9). The eggs are
particularly restricted to a thin (3-12m), sandy
carbonate unit, interpreted as calcretized
palaeosoil (Sahni & Khosla, 1994). Several
hundred nests with 3-13 eggs and abundant frag-
mentary eggshells are found in this unit. Most of
these eggs belong to titanosaurids. A partial
skeleton of a baby titanosaur has been discovered
among the nesting grounds in Gujrat (Mohabey,
1987), but Jain (1989) questioned its identity and

519

FIG. 14. Composite restoration of the skull of Indosuchus raptorius, an
abelisaurid theropod from the Lameta Group; left lateral view; based on
AMNH 1753, 1955, 1960 and ISI R 163. a=angular, ar=articular, d=den-
tary, j=jugal, I-lacrimal, m=maxilla, n=nasal, p=parietal, pm-premaxilla,
po =postorbital, q=quadrate, qj=quadratojugal, sa=sarangular (modified
from Chatterjee, 1978).

assigned it to a booid snake. We have collected
excellent specimens of titanosaurid eggs from
Jabalpur and Dohad areas. The eggs are spherical
with an average diameter of 15cm and were laid
in linear or circular fashion. Similar titanosaurid
eggs are known from France and Spain.

BIOTIC EFFECTS OF THE KT BOUNDARY
CATASTROPHIC EVENTS

The Cretaceous Period was rich in biodiversity.
Yet by the end of the period, there was an un-
precedented ecological crisis leading to extinc-
tions of many groups of organisms, from the giant
dinosaurs to microscopic plankton. The abrupt-
ness and importance of the KT boundary has been
recognized from the birth of stratigraphy, as it
was the basis for the demarcation of two geologic
eras. Although both impacts and Deccan vol-
canism coincided with the KT boundary, their
relative importance for global climatic instability
and disruption of life is still hotly debated. There
are three possible ways in which extinctions
could have been brought about: 1, an abrupt
cataclysm triggered by the impact; 2, a gradual
extinction process induced by the prolonged Dec-

520

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 15. Life restoration of the Masstrichtian Indian dinosaur community, In the foreground are two abelisaurid
theropods, Jndosuchus raptorius confronting a herd of titanosaurid sauropods, Titanosaurus indicus.

can volcanism that polluted the atmosphere and
changed the world’s climate; or 3, the combined
effects of both impact and volcanism might have
contributed to environmental crisis leading to
mass extinction. The fossil record of victims and
survivors across the KT boundary have been used
by the proponents of both impact and volcanic
models to champion their views.

BIOTIC EFFECTS IN INDIA. Since India was
ground zero for both impact and Deccan vol-
canism, this would be an ideal place to search for
the evidence of crisis on local biota. Unfortunately,
the terrestrial sequence coinciding exactly with
the KT boundary is limited in India as in other
parts of the world. It is very unlikely that such a
singular event would leave dense, multispecific
dinosaur bone beds in the area of greatest mass
mortality, unless the catastrophic bed is quickly
buried and protected. Surprisingly, there is one
dramatic event of mass mortality preserved in
India which may be tied to the Shiva impact. In
the Anjar KT boundary section of Kutch, the
dinosaurs died out suddenly at the KT boundary
which is defined on the basis of the iridium
anomaly and isotopic dating (Bhandari et al.,
1994). Here, the KT boundary section is a lcm
thick, pinkish clay layer above a cross-bedded
limestone unit which contains associated
titanosaurid bones. No dinosaur bones occur
above the impact layer. We collected some KT
boundary samples and dinosaur bones from this
section and sent them to Dr Moses Attrep Jr, of
Los Alamos National Laboratory for geochemical

analysis. Dr Attrep found a signicant enrichment
of iridium (348 parts per trillion) from the clay
sample, but could not detect any anomaly in the
dinosaur bone (82 ppt) or limestone unit. This is
the first unequivocal evidence that the dinosaur
extinction occurred precisely at the time of
meteoritic impact.

The influence of Deccan volcanism on contem-
porary biotas can be inferred from the fossil
record in the trappean beds. The most obvious
local catastrophe posed by Deccan volcanism
was the habitat destruction of 800,000km^ of
tropical forest by extensive lava flows — an area
of about the size of Texas (Fig. 9). The advancing
flow and ash fallout must have destroyed the
complex ecosystem for a large number of plant
and animal species. The largest land vertebrates,
the dinosaurs, were most affected by loss and
fragmentation of habitats. The surviving popula-
tions of flora and fauna were forced to move out
to the restricted space available in peninsular
India, where they had to compete for food and
resources. However, Deccan volcanism consisted
largely of nonexplosive, tholeiitic eruptions
similarto Kilauea and Reunion emissions, though
on a grander scale. Recurrent eruptions do not
disrupt much of the rich biota on the islands of
Hawaii and Reunion. Although basaltic flows of
relatively low viscosity commonly destroy
plants, they move slowly enough that they seldom
threaten the lives of the animals. Luxuriant
vegetation thrived in the Deccan province as
indicated by palaeobotanical evidence. Between
the flows are the trappean beds that contain plant

KT EVENTS IN INDIA

remains of fungi, bryophytes, ferns, conifers and
angiosperms. This flora, a very extensive one, is
decidedly unusual in that it consists of seeds,
fruits, leaves, trunks, branches and the like —
thousands upon thousands of them. In 1988, we
visited the little-known ‘National Fossil Park’ in
Shahpura, about 80km from Jabalpur. This is
probably one of the best petrified forests in the
world, preserving this magnificient Deccan flora.
We believe that these vast floral accumulations
may signify recurrent deforestation events induced
by Deccan ashfalls. In spite of floral deteriora-
tion, no major extinction event has been recorded
among the Deccan flora (Lakhanpal, 1970).
Within the intertrappeans beds, magnificent
fossil accumulations indicate that life was
resilient in this hostile environment. Deccan vol-
canism cannot be the proximate cause of the mass
extinction.

The long-term hazard of Deccan volcanism
would be several trillion tons of toxic gases
pumped into the upper atmosphere, which would
result in global climatic perturbations disrupting
the ecosystem. Within the trappean beds, espe-
cially in the Lameta Group of rocks, the green-
house effect can be identified, by the climate
becoming more severe, with marked seasonal
changes and aridity. There were large alkaline
playa lakes around the periphery of the Deccan
province where dinosaur bones and eggs were
preserved in hot, arid conditions (Sahni et al.,
1994). These lakes were subjected to alternate
wet and dry seasons. Although pollution from the
Deccan volcanism might have had some adverse
effect on vertebrate populations, Prasad et al.
(1994) could not detect any change in the com-
position or population of biota from infratraps to
intertraps, during episodic volcanic activity. They
concluded that the Deccan volcanism was not
detrimental to life. There is another line of
evidence which supports this theory. Midoceanic
ridges spewing tholeiitic lava, toxic gases and
sulfides, support a rich oasis of life, including
tubeworms, polychaetes, crustaceans and mol-
lusks. In fact, submarine vents provide a refuge
for many relict forms that might have escaped the
KT extinction (Tunnicliffe, 1991), Extrusion of
the Deccan flows certainly affected the local flora
and fauna by habitat destruction and pollution,
but had little direct regarding the major extinc-
tions.

Eggs are generally more sensitive indicators of
environmental stress. As mentioned earlier, the
world’s largest fossil nesting sites have been
discovered recently in the northern part of the

Narmada lineament around the fringe of the Dec-
can province stretching for almost 1,000km.
These eggs were apparently laid on wet, sandy-
limey soils and were later subjected to prolonged
intervals of aridity that resulted in considerable
dehydration (Sahni et al., 1994). The abundant
distribution of eggs suggests that sauropod
population was thriving there when Deccan vol-
cano was erupting. Sarkar et al. (1992)
reconstructed the palaeoecology of the Lameta
titanosaurids from stable isotope (80 and C)
analysis of their eggs from the Kheda district of
Gujrat. They found oxygen isotope data indicat-
ing that titanosaurids drank water from rivers and
evaporative pools, whereas carbon isotope data
suggesting they consumed C3-type plants (small
palms, conifers, dicot shrubs, etc.) in a semiarid
environment. It is well-known that pathological
conditons in living reptiles and birds is caused by
hormonal imbalance which can be induced by
psychic stress due to overcrowding in a restricted
space, or sudden a change in the environment
(Erben et al., 1979). Surprisingly, in various nest-
ing sites around the Deccan Province, embryos,
hatchlings and juvenile skeletons are almost
absent. Did these dinosaur eggs fail to hatch
because of halogen poisoning from the Deccan
flows? This seems unlikely because dinosaur
eggs have been found at the KT boundary section
in Anjar (Ghevaria & Srikarni, 1990). Unlike the
condition in French and Spanish nesting sites
(Erben et al., 1979), we could not detect any
pathological abnormality in shell thickness and
growth.

Was there any other mechanism of gradual
reproductive failure among Indian dinosaurs?
Heat from Deccan volcanism could have altered
the sex ratio of the local dinosaur population.
Paladino et al. (1989) proposed that sex in
dinosaurs may be controlled by incubation
temperature in the nests as seen among living
reptiles such as turtles and crocodiles. They
argued that during the Late Maastrichtian,
climatic perturbations may have led to unisexual
dinosaur populations and eventual extinction.
This is an interesting concept but difficult to
evaluate from the fossil record. Moreover, if
dinosaur physiology was similar to that of birds
(Bakker, 1975), temperature during incubation
may not have had any role in sex determination.
The occurrence of dinosaurs and eggs in the
sediments intercalated with with Deccan lava
flows indicates that volcanism had little direct
effect on their demise. However, the pollution

522

from Deccan volcanism may have some global
impact in conjunction with impact.

IMPACT-EXTINCTION MECHANISMS. One
of the most important questions about the KT
extinctions concerns its timing and duration. The
extinction of six groups of marine microfossils —
planktic and benthic forams, coccoliths,
radiolarians, dinoflagellates and diatoms — took
place precisely at the impact horizon and are
compatible with this scenario (Smit, 1982;
Thierstein, 1982; Alvarez et al., 1984).
Brachiopods (Surlyk & Johnson, 1984) and am-
monites (Ward et al., 1991) also show a similar
abrupt pattern of extinction at the KT boundary.
Similarly, pterosaurs, plesiosaurs and mosasaurs
in the marine realm became extinct suddenly at
the end of the Maastrichtian (Buffetaut, 1990).
There is no question that the oceanic food chain
collapsed by environmental changes around the
KT boundary. In terrestrial communities, the
sharp floral event exactly at this boundary strong-
ly suggests a causal link with the impact scenario
(Tschudy & Tschudy, 1986). Careful sampling
indicates that dinosaurs in Montana died out
suddenly at the KT boundary (Sheehan et al.,
1991). We see the similar evidence of abrupt
disappearance of dinosaurs precisely at the
iridium KT boundary layer in Anjar section of
Gujrat. Thus, the palaeontological record favors
a sudden and simultaneous extinction pattern at
the KT boundary among certain groups of or-
ganisms.

If, as seems likely, one or more meteorite im-
pacts did play a role in the mass extinction event,
what mechanism was involved? How devastating
was the collision and how did it affect life?
Various models have been proposed in recent
times to explain the killing mechanism. A mas-
sive impact would generate earthquakes of high
magnitude, inject a large volume of dust into the
stratosphere, darken the skies, halt photosyn-
thesis, ignite global fires, initiate tsunamis,
produce acid rains, decrease ocean surface
alkalinity and devastate the biosphere. Millions
of organisms would die instantly from the direct
effect of the impact — shock heating of the at-
mosphere by the expanding fireball (Alvarez
1986, 1987; Emiliani et al., 1981; 1987; Melosh
et al, 1990). Comet Shoemaker-Levy 9 provided
proof that the KT impact ignited global fires. As
the ejecta fell back into the atmosphere after 20
minutes of each mpact, they reentered with a
release of titanic energy; the heat from this
reentry was so intense it was easily detected from

MEMOIRS OF THE QUEENSLAND MUSEUM

Earth (Weissman, 1995). Similarly, when the KT
ejecta reentered into Earth’s atmosphere, they
ignited terrestrial forests and plant covers. The
wildfires consumed oxygen and poisoned the
atmosphere with CO. This massive impact would
generate a 100-million megaton blast — 1000
times more powerful than the explosion of all
nuclear arsenals of the world. The impact force
would have vaporized not only the meteorite but
also melted the target rock. This melt ejecta
would spread in large waves, until it had formed
a large crater. Huge tsunamis produced by the
oceanic impact could destroy coastal habitats
across the globe (Sharpton & Ward, 1990).

Available data suggest that the extinction of
carbonate-secreting plankton was especially
severe at the KT boundary (Smit, 1982; Keller,
1989). Acid trauma is likely a consequence of
many biotic crisis in the oceans. Various impact
mechanisms have been proposed to produce large
volumes of HNO3 and H2SOs for acidification of
surface marine waters. Shock heating of the at-
mosphere would cause nitrogen and oxygen to
combine with steam to form nitric acid (Prinn &
Fegley, 1987). The sheer size of the projectile
cannot by itself account for the devastation. The
unique composition of the crash sites provided
additional arsenal. Because the target rock at the
Yucatán Peninsula (and western coast of India)
has a thick-cover of CaSOs, an impact there could
have released as much as a trillion tons of SO?
(D'Hondt et al., 1994). Release of devolatilized
SO» would form sulfuric acid aerosols which
would contribute to a rapid decline in global
surface temperature and halt photosynthesis. The
subsequent showers of acid rain must have been
agents for decreased ocean alkalinity and destruc-
tion of life. All of these events had cascading
effects through the terrestrial and marine ecosys-
tems to collapse the food chain.

All the extinction mechanisms cited above are
based on a 10km-sized meteorite. If we consider
the impact of a 40km meteorite, the devastation
would be much more traumatic on the biosphere.
One obvious consequence of a larger oceanic
impact would be partial vaporization of the photic
zone, raising the atmosphere's temperature to
hundreds of degrees, and destoying all life in the
ocean surface. Thus, a large body impact may
explain the marine regression by vaporization of
water which is generally associated with the KT
event. The impact-induced thermal radiation
would ignite global wildfires on land and destroy
ecological niches (Melosh et al., 1990).

KT EVENTS IN INDIA

MECHANISM OF VOLCANIC-EXTINC-
TION. Many palaeontologists are skeptical about
the scenario of impact holocaust. They argue that
the KT extinction was neither global, nor instan-
taneous, but occurred over an extended period of
time, because different organisms disappear at
different levels at or near the KT boundary. They
look for extinction agents that could explain a
gradual extinction event and favor terrestrial
causes, such as prolonged Deccan volcanism and
regression of sea level, as main contributing factors
for the biotic crisis (Clemens, 1982; Officer et al.,
1987; Hallam, 1987; Stanley, 1987, p. 167;
Keller, 1989; Zinsmeister et al., 1989; Courtillot,
1990).

However, the gradual extinction pattern seen
among some organisms may be an artifact of
preservation and declining sampling quality. Sig-
nor & Lipps (1982) showed theoretically how a
sudden catastrophic extinction would appear to
have been gradual in the fossil record, if the
record was not dense. The Signor-Lipps effect
weakens the case for a gradual extinction. Fur-
thermore, the instantaneous extinction in both
continental and marine organisms is not a neces-
sary corollary of the impact theory (Alvarez et al.,
1984). A large impact could trigger environmen-
tal crisis, and the latter could, in turn, cause
extinctions spread over 10 to 10° years (Hsu et
al., 1982). The most ecologically-sensitive
organisms were probably the first victims, with
progressively more tolerant groups succumbing
in the later stages. The short term physical event
may lead to long term biological events.

There are many examples of pre-boundary and
post-boundary extinction events (Keller, 1979)
which led many palaeontologists to advocate a
close relationship between the biotic crisis and
Deccan volcanism. There is no doubt that such a
massive volcanic outburst over an extended
period would have deleterious environmental
consequences. Based on data known from the
Laki eruption in 1783, the Deccan eruption must
have pumped large volumes of SO», HCl and ash
into the atmosphere to cause global cooling,
immense amounts of acid rain, reduction in
alkalinity and pH of the surface ocean and ozone
layer depletion that might be harmful to both
terrestrial and marine organisms (Hallam, 1987).
Deccan volcanism must have produced large-
scale aerosol clouds of sulphur and CO2. The
reduction in temperatures caused by sulphur
aerosols in the stratosphere act in the opposite
direction to the greenhouse gases such as carbon
dioxide. Emissions of CO» from the Deccan vol-

523

canism would directly increase the atmospheric
and oceanic CO», leading to sustained global
warming.

SYNTHESIS. Some of the consequences of an
asteroid impact and of massive volcanism would
be quite similar, such as pollution of the atmos-
phere, darkness resulting from dust (either ejecta
or ash), suppression of photosynthesis, acid rain,
global cooling, carbonate crisis in the ocean
waters, environmental stress, devastation of
ecosystems and the collapse of the food chain.
The scenario of extinction by food chain disrup-
tion is compatible with both impact and volcanic
hypotheses. It is generally believed that the biotic
crisis was most severe and sometimes limited to
tropical-subtropical regions while high latitude
faunas and floras were little affected (Keller,
1994). The geographic locations of two or three
impact sites, Deccan volcanism and the distribu-
tion of impact deposits along the Alvarez Impact
Belt may explain this geographic selection of KT
extinction (Fig. 6B). The sudden cooling by
H2S04 aerosol and warming by CO2 would result
in chaotic climatic perturbations which would be
especially traumatic for global organisms. How-
ever massive volcanism would lack some of the
titanic lethal features that are generally associated
with a large body impact, such as gigantic shock
effects leading to vapor plume, global fire, ther-
mal pulse, evaporation of the photic zone and
concomitant sea regression, chondritic metal
toxicity and volatilization of target rocks. Impacts
probably made more dramatic changes to KT
environments globally and a more traumatic
crisis to the ecosystem than volcanic emissions.
Impacts also perturbed the environment so sud-
denly and catastrophically that most organisms
failed to adapt to these changes in such a short
time and perished. In contrast, prolonged vol-
canism allowed enough time for some organisms
to adapt, or for others to disappear gradually.
Global fire must have decreased the amounts of
atmospheric oxygen which would further exacer-
bate the hostile conditions already developing
among terrestrial and marine life. Selectivity and
a step-wise extinction pattern can be better ex-
plained by the volcanic model. It seems that im-
pact was the proximate cause to the biotic crisis
at the KT boundary, whereas long-term vol-
canism produced harmful changes that enhanced
climatic stress and finished off the extinction
process.

524

CONCLUSION

We have outlined two catastrophic events at the
KT boundary in India: a giant meteorite impact at
the India-Seychelles rift margin and the spec-
tacular volcanic outbursts of Deccan volcanism.
These two events were so closely intertwined in
space and time that it is difficult to untangle them
in proper chronologic order. The subsurface
stratigraphy and geophysical data suggest that the
Shiva Crater is a potential candidate for the long-
sought KT impact scar, Biostratigraphic,
geochronologic and palaeomagnetic evidence
indicates a KT boundary age.

We speculate that an oblique impact by a40km
asteroid produced the oblong, complex, Shiva
Crater on the continental shelf of western India,
600km across and more than 12km deep. Part of
the crater rim survives in the Panvel Flexure and
the Narmada Fault near western India and in the
Amirante Arc south of the Seychelles. The central
uplifts are represented by the elevated peaks of
the Bombay High, Praslin and Mahe. Between
the central peaks and the crater rim lay an annular
trough, now preserved as the Surat Basin, the
Dahanu-Panna Depression and the Amirante
Basin, each containing thick Tertiary sediments
(Fig. SA). The impact melts were emplaced
radially downrange of the trajectory and are
clustered around the Shiva Crater in the form
plugs of alkaline igneous complexes. Radio-
metric, palaeontologic and palaeomagnetic data
suggest that the Shiva Crater was formed at the
KT boundary.

We are aware that one serious shortcoming of
the Shiva Impact hypothesis is the documentation
of shock metamorphism in crater proximity.
There are several possible reasons for this
deficiency. First, we could not acquire any KT
boundary core samples from Oil and Natural Gas
Commission of India regarding Bomay High
area. These samples are crucial to detect impact
signatures such as iridium anomaly, shocked
quartz, melt rocks, breccias, suevites and shatter
cones. Second, the impact took place in a shal-
low-marine setting. The tsunami generated by
this impact would have been efficient at removing
ejecta material. Third, some of the shocked sig-
natures may be preserved in the deeper part of the
basin, yet to be penetrated by drilling. Fourth,
contemporary Deccan volcanism and impact melt
which had filled part of the crater and crater
exterior must have consumed or erased other
evidence of shocked features such as shatter
cones, breccia and ejecta. The thick blanket of

MEMOIRS OF THE QUEENSLAND MUSEUM

Deccan Traps on the western part of India must
have concealed most of the crucial evidence of
impact signatures. As in the case of Chicxulub,
we need more subsurface data to confirm the
impact origin of the Shiva structure. An interna-
tional effort is needed to unravel the morphology
of the Shiva Crater.

It is highly coincidental that critical evidence
for both the Chicxulub and Shiva Craters came
from petroleum prospecting, because the sedi-
ments overlying the central peak of a complex
crater form a domal structural trap. The Shiva
crater now preserves a complicated geological
history, as well as valuable economic resources
— the largest oil field in India. Thus the Shiva
impact had created not only an immense crater
and continental rifting, but also affected global
climate, the chemistry of the world’s ocean, a
biotic crisis and perhaps even the production of
oil traps.

The near-antipodal positions of the Chicxulub
and Shiva structures suggest two possibilities:
Either, two large meteorite fragments crashed on
a rotating Earth at a short interval, creating two
large scars along the Alvarez Impact Belt. Or, a
large impact on one side of the Earth produced a
similar signature on the far side. The first model
finds support from the unusual concentration of
impact deposits and biogeographic selectivity of
extinction at low latitudes (Keller, 1994) along
this Alvarez Impact Belt.

The impact-induced model of Deccan volcanism,
though very appealing, is rejected because of con-
flict of timing. The revised calibration of Deccan
volcanism indicates that the eruption began over
1Ma before the KT event and extended for 3Ma.
If so, then the Shiva impact did not initiate the
Deccan plume activity.

Although both impact and volcanism might
have contributed to inhospitable environments
and biotic crisis at the KT boundary, the impact
must have played a major role in the killing
mechanism. Both catastrophes contributed
heavily to the breakdown of stable ecological
communities and disrupted the biosphere. In the
aftermath of the KT extinction, some forms of
life, such as bird and mammals rebounded. They
formerly played minor roles, but assumed
prominence after the extinction. Today, they are
abundant in both numbers of species and popula-
tions and can be found on every continent oc-
cupying virtually all available niches. The
selectivity of KT extinction is still puzzling. For
example, terrestrial dinosaurs became the vic-
tims, but their flying counterparts — birds —

KT EVENTS IN INDIA

survived the catastrophe. Birds rose like the
Phoenix from the funeral pyre of terrestrial
dinosaurs with renewed youth and beauty to
carry their ancestral heritage into the Cenozoic
cycle.

ACKNOWLEDGMENTS

We thank Fernando E. Novas for inviting us to
contribute this paper; Soumya Chatterjee and
Ralph E. Molnar for reviewing the manuscript;
Sibani Chatterjee, P.K. Mazumdar, Dhuiya Padhi,
Shyamal Roy and Lakshman Mahankur for field
assistance; and Michael W. Nickell for illustra-
tions. We are specially grateful to authorities of
the Gun and Carriage Factory of Jabalpur for
access to their land, and Sevalia Cement Corpora-
tion of Gujrat for accommodation and hospitality.
We thank Frank Kyte and M.B. Boslough for
sharing their unpublished works, and Moses
Attrep Jr for iridium analysis. The research was
supported by the National Geographic Society,
Dinosaur Society, Texas Tech University and In-
dian Statistical Institute.

LITERATURE CITED

ADATTE, T., STINNESBECK, W. & KELLER, G.
1994. Mineralogical correlations of near KT
boundary deposits in northeastern Mexico:
evidence for long-term deposition and vol-
canoclastic influence. Lunar and Planetary In-
stitutre Contribution 825: 1-3.

ALEXANDER, P.O. 1981. Age and duration of Deccan
volcanism: K-Ar evidence. Pp. 244-278. In K.V.
Subbarao & R.N. Sukeshwala (eds) ‘Deccan
Volcanism’. Geological Society of India Memoir

ALT, D., SEARS, J.M. & HYNDMAN, D.W. 1988.
Terrestrial maria: the origins of large basalt
plateaus, hotspot tracks and spreading ridges.
Journal of Geology 96; 647-662.

ALVAREZ, L.W. 1987 Mass extinctions caused by
large bolide impacts. Physics Today 40(7): 24-33.

ALVAREZ, L.W., ALVAREZ, W., ASARO, F. &
MICHEL, H.V. 1980. Extraterrestrial cause for the
Cretaceous-Tertiary extinction. Science 208:
1095-1108.

ALVAREZ, W. 1986. Toward a theory of impact crises.
EOS, Transactions, American Geophysical Union
67(35): 649-658.

ALVAREZ, W., ALVAREZ, L.W., ASARO, F. &
MICHEL, H.V. 1982. Current status of the impact
theory for the terminal Cretaceous extinction.
Geological Society of America Special Paper 190:
305-315.

1984. The end of the Cretaceous: sharp boundary or
gradual transition? Science 223: 1183-1186

525

ALVAREZ, W. & ASARO. F. 1990. An extraterrestrial
impact. Scientific American 263(4): 78-84.
ALVAREZ, W., ARTHUR, M.A., FISCHER, A.G.,
LOWRIE, W., NAPOLEINE, G., SILVA, I.P. &
ROGGENTHEN, W.M. 1987. Upper Cretaceous-
Palaeocene magnetic stratigraphy at Gubbio, Italy.
V. Type section for the Late Cretaceous-
Palaeocene geomagnetic reversal time scale.
Geological Society of America Bulletin 88: 383-

389.

ALVAREZ, W., ASARO. F., CLAEYS, P.,
GRAJALES-N., J.M., MONTNARI, A. & SMIT,
J. 1994. Developments in the KT impact Theory
since Snowbird II. Lunar and Planetary Contribu-
tions 825: 3-4.

ALVAREZ, W., CLAEYS, P. & KIEFFER, S.W. 1985.
Emplacement of Cretaceous-Tertiary boundary
shocked quartz from Chicxulub Crater. Science
269: 930-935.

ALVAREZ, W., SMIT, J., LOWRIE, W., ASARO, F.,
MARGOLIS, S.V., CLAEYS, P., KASTNER, M.
& HILDEBRAND, A.R. 1992. Proximal impact
deposits at the Cretaceous-Tertiary boundary in
the Gulf of Mexico: a iestudy of DSDP Leg 77
Sites 536 and 540. Geology 20: 697-700.

ANDERSON, R.R. & HARTUNG, J.B. 1988. The
Mason structure in north-central Iowa: the largest
known impact site in the United States and a K-T
boundary astroblem. EOS, Transactions
American Geophysical Union 69(44): 1291.

AUDEN, J.B. 1949. Dykes of western India: a discus-
sion of their relationship with the Deccan Traps.
Transactions of the National Institute of Science
India. 3:123-157.

1974. Afganistan-West Pakistan. Pp. 235-253. In
A.M. Spencer (ed.) Geological Society Special
Publication 4, London.

BACKMAN, J., DUNCAN, R.A., PETERSON, L.C.,
BAKER, P.A., BAXTER, A.N., BOERSMA, A.,
CULLEN, J.L., DROXLER, A.W., FISK, M.R.,
GREENOUGH. J.D., HARGRAVES, R.B.,
HEMPEL, P., HOBART, M.A., HURLEY, M.T.,
JOHNSON, D.A., MACDONALD, A.M., MIK-
KELSEN, N., OKADA, H., RIO, D., ROBIN-
SON, S.G., SCHNEIDER, D., SWART, P.K.,
TATSUMI, Y., VANDAMME, D., VILKS, G. &
VINCENT, E. 1988. Mascarene Plateau. Proceed-
ings of the Ocean Drilling Program 5: 5-15.

BAKER, B.H. 1967. The Precambrian of the Seychelles
Archipellago. Pp. 122-132. In K. Rankama (ed.)
*Precambrian Research'. (Interscience: New
York).

BAKER, B.H. & MILLER, J.A. 1963. Geology and
chronology of the Seychelles Islands and structure
of the floor of the Arabian Sea. Nature 199: 346-
348.

BAKKER, R.T. 1975. Dinosaur renaissance. Scientific
American 232: 58-78.

BAKSI, A.K. 1994. Geochronological studies on
whole-rock basalts, Deccan Traps, India: evalua-
tion of the timing of volcanism relative to the K-T

526

boundary. Earth and Planetary Science Letters
121: 43-56.

BASU, A.R., RENNE, P.R., DASGUPTA, D.K.,
TEICHMANN, F. & POREDA, RJ. 1993. Early
and late alkali igneous pulses and high- He plume
origin for the Deccan Flood Basalts. Science 261:
902-906.

BASU, D.N., BANERJEE, A. & TAMHANE, D.M.
1982. Facies distribution and petroleum geology
of the Bombay offshore basin, India. Journal of
Petroleum Geology 5(1): 51-75.

BEANNE, J.E., TURNER, C.A., HOOPER, P.R., SUB-
BARAO, K.V. & WALSH, J.N. 1986. Stratig-
raphy, composition and form of the Deccan
basalts, Western Ghats, India. Bulletin of Vol-
canology 48: 61-83.

BEESON, D., GARTNER, S., KELLER, G.,
MacLEOD, N., MEDUS, J., ROCCHIA, R. &
ROBIN, E. 1994. The KT boundary along the
Brazos River, Falls County, Texas: multidiscipli-
nary stratigraphy and depositional environment.
Lunar and Planetary Institute Contributions 825:
9-10.

BELL, J. 1996. Far journey to a near asteroid.
Astronomy 24(3): 42-47.

BEKOV, G.I., LETOKHOV. V.S., RADAEV, V.N.,
BADYUKOV, D.D. & NAZAROV, M.A. 1988.
Rhodium distribution at the Cretaceous/Tertiary
boundary analyzed by ultrasensitive laser
photoionization. Nature 332: 146-148.

BERMAN, D.S. & JAIN, S.K. 1982. The braincase of
a small sauropod dinosaur (Reptilia: Saurischia)
from the Upper Cretaceous Lameta Group,
Central India, with review of Lameta Group
localities. Annals of Carnegie Museum 51: 405-
422.

BHANDARI, L.L. & JAIN, S.K. 1984. Reservoir geol-
ogy and its role in the development of the L-III
reservoir, Bombay High Field, India. Journal of
Petroleum Geology 7(1): 27-46.

BHANDARI, L.L. & SHUKLA, P.N. 1995. Impact did
not trigger Deccan volcanism: evidence from
Anjar K/T boundary intertrappean sediments.
Geophysical Research Letters 22: 433-436.

BHANDARI, L.L., SHUKLA, P.N., GHEVARIA, Z.G.
& SUNDARAM, S.M. 1994. KT boundary in
Deccan intertrappeans; chemical anomalies and
their implications. Lunar and Planetary Institute
Contributions 825: 10-11.

BISWAS, S.K. 1982. Western rift basins of India and
hydrocarbon prospects. Oil and Gas Journal 80:
224-232.

1987. Regional tectonic framework, structure and
evolution of the western marginal basins of India.
Tectonophysics 135: 307-327.

1988. Structure of the western continental margin of
India and related igneous activity. Pp. 371-393.
In Subbarao, K.V. (ed.) ‘Deccan Flood Basalts’.
Geological Society of India Memoir 10.

BLUM, J.D., CHAMBERLIN, C.P., HANGSTON,
M.P., KOEBERL, C., MAREN, L.E.,

MEMOIRS OF THE QUEENSLAND MUSEUM

SCHURAYTZ, C. & SHARPTON, V.L. 1993.
Isotopic composition of the K/T boundary impact
glass with melt rock from the Chicxulub and
Manson impact structures. Nature 364: 325-327.

BOHOR, B.F., FORD, E.E., MODRESKI, PJ. &
TRIPLEHORN, D.M. 1984. Mineralogic
evidence for an impact event at the Cretaceous-
Tertiary boundary. Science 224: 867-869.

BOHOR, B.F., MODRESKI, PJ. & FORD, E.E. 1987.
Shocked quartz in the Cretaceous-Tertiary bound-
ary clays: evidence for a global distribution.
Science 230: 705-709.

BOHOR, B.F., & SEITZ, R. 1990, Cuban K/T
catastrophe. Nature 344: 593.

BONAPARTE, J.F. & NOVAS, F.E. 1985. Abelisaurus
comahuensis, n.g., n.sp., Carnosauria del Cretacio
Tardio de la Patagonia. Ameghiana 21: 259-265.

BONAPARTE, J.F, POWELL, J.F. & POWELL, J.E.
1980. A continental assemblage of tetrapods from
the Upper Cretaceous beds of El Brete,
northwestern Argentina (Sauropoda-
Coelurosauria-Carnosauria-Aves). Memoires de
la Société Géologique de France. 139: 19-28.

BONAPARTE, J.F., POWELL, J.F. & CORIA, R.A.
1990. Carnotausus sastrei Bonaparte, the horned,
lightly built carnosaur from the Middle
Cretaceous of Patagonia. Contributions In
Science, Natural History Museum of Los Angeles
County 416: 1-42.

BOSE, M.K. 1980. Alkaline magmatism in the Deccan
volcanic province. Journal of the Geological
Society of India 21: 317-329.

BOSLOUGH, M.B., CHAEL, E.P., TRUCANO, T.G.
& CRAWFORD, D.A. 1996. Axial focusing of
energy from a hypervelocity impact on Earth.
International Journal of Impact Engineering, in
press.

BOURGEOIS, J., HANSEN, T.A., WILBERG, P.L., &
KAUFFMAN, E.G. 1988. Atsunami depositat the
Cretaceous-Tertiary boundary in Texas. Science
241: 567-570.

BUFFETAUT, E. 1989, Archosaurian reptiles with
Gondwanan affinities in the Upper Cretaceous of
Europe. Terra Nova 1: 69-74.

1990. Vertebrate extinctions and survival across the
Cretaceous-Tertiary boundary. Tectonophysics
171: 337-345.

CARLISLE, D.B. 1993. Extraterrestrial amino acids at
the Cretaceous-Tertiary boundary of Alberta. Na-
ture 361: 551-553.

CARLISLE, D.B. & BRAMAN, D.R. 1991.
Nanometric-size diamonds in the Cretaceous/Ter-
tiary boundary clay of Alberta. Nature 352: 708-
709.

CHAKRAVARTI, D.K. 1935. Is Lametasaurus indicus
an armored dinosaur? American Journal of
Science 230: 138-141.

CHATTERJEE, S. 1978. Indosuchus and Indosaurus,
Cretaceous carnosaurs of India. Journal of
Palaeontology 52: 570-580.

KT EVENTS IN INDIA

1990. A possible K-T impact site at the India-
Seychelles boundary. Abstract, Lunar and
Planetary Science Conference XXI, pp. 182-183.

1992. A kinematic model for the evolution of the
Indian plate since the Late Jurassic. Pp. 33-62. In
Chatterjee, S. & Hotton, N. (eds) ‘New Concepts
in Global Tectonics’, (Texas Tech University
Press: Lubbock).

CHATTERJEE, S. & RUDRA, D.K. 1993. Shiva
Crater: a possible K/T boundary impact stucture
at the India-Seychelles plate margin. P. 38. In 74th
Annual Meeting of the Pacific Division, AAAS,
Program with Abstracts. (abstract)

CHOUBEY, V.D. 1971. Narmada-Son lineament, India.
Nature 232: 38-40.

CLEMENS, W.A. 1982. Pattern of extinction and sur-
vival of the terrestrial biota during the
Cretaceous/Tertiary transition. Geological
Society of America Special Paper 190: 407-413.

CLUBE, S.V.M. & NAPIER, W.M. 1982. The role of
episodic bombardment in geophysics. Earth and
Planetary Science Letters 57: 251-262.

COOMBS, W.P. Jr 1978. The families of the ornithis-
chian dinosaur order Ankylosauria. Palaeontology
21: 143-170.

COOMBS, W.P. Jr & MARYANSKA. T. 1990.
Ankylosaur. Pp. 456-483. In Weishampel, D.B.,
Dodson, P. & Osmólska, H. (eds) 'The
Dinosauria'. (University of California Press:
Berkeley).

COURTILLOT, V. 1990. A volcanic eruption. Scientific
American 263(4): 85-92.

COURTILLOT, V. , FERAUD, G., MALUSKI, H.,
VANDAMME, D., MOREAU, M.G. & BESSE,
J. 1988. Deccan flood basalts and the
Cretaceous/Tertiary boundary. Nature 333: 843-
846.

COURTILLOT, V. , BESSE, J., VANDAMME, D.,
MONTIGNY, R., JAEGER, J.J. & CAPETTA, H.
1986. Deccan flood basalts at the Cretaceous/Ter-
tiary boundary? Earth and Planetary Science Let-
ters 80: 361-374,

DE, A. 1981. Late Mesozoic-Lower Tertiary magma
types of Kutch and Saurashtra. Pp. 327-339. In
Subbarao, K.V. & Sukeshwala, R. N. (eds)
Geological Society of India Memoir 3, Bangalore.

DESONIE, D. 1996. The threat from space. Earth 5(4):
24-3].

DEVEY, C.W. & LIGHTFOOT, P.C. 1986. Vol-
canological and tectonic control of stratigraphy
and structure in the western Deccan Traps. Bul-
letin of Volcanology 48: 195-207.

DEVEY, C.W., & STEPHENS, W.E. 1992, Deccan-re-
lated magmatism west of the Seychelles-India rift.
Pp. 271-291. In Storey, B.C., Alabaster, T.A. &
Pankhurst, R.J. (eds) ‘Magmatism and the Causes
of Continental Break-up’. Geological Society
Special Publications 68, London.

D'HONDT, S., SIGURDSSON, H., HANSON, A.,
CAREY, S. & PILSON, M. 1994, Sulfate
volatilization, surface-water acidification and ex-

527

tinction at the KT boundary. Lunar and Planetary
Institute Contribution 825: 29-30.

DIETZ, R.S. 1961. Vredefort ring structure: meteorite

impact scar? Journal of Geology 69: 499-516.

1993, Review of terrestrial evidence for major cos-
mic impacts. P. 40. In 74th Annual Meeting of the
Pacific Division, AAAS, Program with Abstracts.
(abstract)

DIETZ, R.S. & McHONE, J. 1990. Isle of Pines (Cuba),
apparently not K/T boundary impact site.
Geological Society of America Abstracts with
Programs 22(7): A79. (abstract)

DUNCAN, R.A. & PYLE, D.G. 1988, Rapid eruption
of the Deccan flood basalts at the Cretaceous/Ter-
tiary boundary. Nature 333: 841-843.

ELLAM, R.M, 1992. Lithospheric thickness as a con-
trol on basalt geochemistry. Geology 20: 153-156.

ENGELHARDT, W. von 1984. Melt products from
terrestrial impact structures. 27th International
Geological Congress Proceedings 19: 149-163.

ERBEN, H.K. & WEDEPHOL, K.H. 1979.
Palaeobiological and isotopic studies of eggshells
from a declining dinosaur species. Palaeobiology
5(4): 380-414.

FISHER, R.L., ENGEL, C.G. & HILDE, T.W.C. 1968.
Basalts dredged from the Amirante Ridge, western
Indian Ocean. Deep Sea Research 15: 521-534.

FRENCH, B.M. 1984. Impact event at the Cretaceous-
Tertiary boundary: a possible site. Science 226:
353.

FRENCH, B.M. & SHORT, N.M. (eds) 1968, ‘Shock
Metamorphism of Natural Materials'. (Mono
Book Corporation: Baltimore). 644 pp.

GALLET, Y., WEEKS, R., VANDAMME, D. &
COURTILLOT, V. 1989. Duration of Deccan Trap
volcanism: a statistical approach. Earth and
Planetary Science Letters 93: 273-282.

GALTON, P.M. 1981. Craterosaurus pottonensis Sel-
ley, a stegosaurian dinosaur from the Lower Juras-
sic of England, and a review of Cretaceous
stegosaurs. Neues Jarrbuch fiir Geologie und
Palántologie, Abhhandlungen 160: 73-95.

GAULT, D.E., & WEDEKIND, J.A. 1978. Experimen-
tal studies of oblique impact. Pp. 3843-3875. In
‘Proceedings, 9th Lunar and Planetary Science
Conference’. (Pergamon Press: New York).

GHEVARIYA, Z.G. & SRIKARNI, C. 1990. Anjar
Formation, its fossils and their bearing on the
extinction of dinosaurs. Pp. 106-109. In A. Sahni
& A. Jolly (eds) *Cretaceous Event Stratigraphy
and the Correlation of the Indian Nonmarine
Strata' (Laser Print: Chandigarh).

GILMORE, J.S., KNIGHT, J.D., ORTH, C.J., PIL-
MORE, C.L. & TSCHUDY, R.H. 1984. Trace
element patterns at a non-marine Cretaceous-Ter-
tiary boundary. Nature 307: 224-228.

GIRLING, C.M. 1992. Hydrocarbon habitat of the
Seychelles microcontinent related to plate tec-
tonics and palaeogeography. Pp. 223-231. In
Plummer, P.S. (ed.) ‘Proceedings of the Indian

528

Ocean Petroleum Symposium’, (United Nations:
Victoria, Seychelles).
GOVINDAN, A. 1981. Foraminifera from the Infra and
Intertrappean surface sediments of Narsapur Well
1 and age of the Deccan Trap flows. Pp. 81-93. In
‘Indian Colloquium on Micropalaeontology and
Stratigraphy, 9th. Proceedings’. (Dehra Dun).
GRIEVE, R.A.F. 1982. The record of impact on Earth:
implications for a major Cretaceous/Tertiary im-
pact event. Geological Society of America Special
Paper 190: 25-37.
1987. Terrestrial impact structures. Annual Review
of Earth and Planetary Sciences 15: 245-270.

1990. Impact cratering on the Earth. Scientific

American 261(4): 66-73.

GRIEVE, R.A.F., WOOD, C.A, GARVIA, J.B.,
McLAUGHLIN, G. & McHONE, J.F. Jr 1988.
* Astronaut's Guide to Terrestrial Impact Craters’,
(Lunar and Planetary Institute: Houston).

HARLAND, W.B., ARMSTRONG, R.L., COX, A.V.,
CRAIG, L.E., SMITH, A.G. & SMITH, D.G.
1990. ‘A Geologic Time Scale 1989'. (Cambridge
University Press: Cambridge). 263 pp.

HARTNADY, C.J.H. 1986. Amirante Basin, western
Indian Ocean: possible-impact site at the
Cretaceous-Tertiary extinction bolide? Geology
14: 423-426.

1989. Oblique impact hypothesis relates Deccan
volcanism to multiple northern hemisphere
craters to Cretaceous-Tertiary boundary.
Abstracts, 28th International Geological Con-
gress, 2: 33-34. (abtracts)

HILDEBRAND, A.R. & BOYNTON, W.V. 1990.
Proximal Cretaceous-Tertiary boundary impact
deposits in the Caribbean. Science 248: 843-847.

HILDEBRAND, A.R., PENFIELD, G.T., KING, D.A.,
PILKINGTON, M., CAMARO, A.Z., JACOB-
SON, S.B. & BOYNTON, W.B. 1991. Chicxulub
Crater: a possible Cretaceous/Tertiary boundary
impact crater on the Yucatan Peninsula, Mexico.
Geology 19: 867-871.

HILDEBRAND, A.R., ASARO, F, ATTREP, M. Jr.,
BERMUDEZ-SANTANA, J.C., BONIS, S.,
CEDILLO-PARDO, E., CLAEYS, P., GON-
ZALEZ-CASILDO, V., GRAJALES-CASILDO,
J.M,, GRAJALES-NISHIMURA, J.M.,
GREGOIRE, D.C., ORITZ-ALEMAN, C.,
PILKINGTON, M., SACCHEZ-RIOS, M.A.,
SMIT, J. & STANSBERRY, J.A. 1994, The Chic-
xulub Crater and its relation to the KT boundary
ejecta and impact-wave deposits. Lunar and
Planetary Institutre Contribution 825: 49-50.

HILDEBRAND, A.R., PILKINGTON, M., CON-
NORS, M., ORTIZ-ALEMAN, C. & CHAVEZ,
R.E. 1995. Size and structure of the Chicxulub
Crater revealed by horizontal gravity gradients
and cenotes. Nature 376: 415-417.

HOOPER, P.R. 1982. The Columbia River Basalts.
Science 215: 1463-1468.

MEMOIRS OF THE QUEENSLAND MUSEUM

1990. The timing of crustal extension and the erup-
tion of continental flood basalt. Nature 345: 246-
249.

HOOPER, P.R., SUBBARAO, K.V. & BEANE, J.E.
1988. The giant plagioclase basalts (GPBs) of the
Western Ghats, Deccan Traps. Pp. 135-144. In
Subbarao, K.V. (ed.) ‘Deccan Flood Basalts’.
Geological Society of India Memoir 10.

HOWARD, K.A. & WILSHIRE, H.G. 1975. Flows of
impact melt at lunar craters, Journal of Research,
U.S. Geological Survey 3: 237-251.

HSU, KJ., McKENZIE, J.A. & HE Q.X. 1982. Ter-
minal Cretaceous environmental and evolutionary
changes. Geological Society of America Special
Paper 190: 317-328.

HUENE, F. von & MATLEY, C.A. 1933. The
Cretaceous Saurischia and Omithischia of the
Central Provinces of India. Palaeontologica Indica
21: 1-74.

IZETT, G. 1990. The Cretaceous/Tertiary boundary in-
terval, Raton Basin, Colorado and New Mexico.
Geological Society of America Special Paper 249:
1-100.

I G., DALRYMPLE, G.B. & SNEE, L.W. 1991.

Ar’ Ar Age of the Cretaceous-Tertiary bound-
ary tektites from Haiti. Science 252: 1539-1541.
IZETT, G., COBBAN, W.A., OBRADOVICH, J.D. &
RUNK. M.J. 1993. The Manson impact structure:
Arf" Ar Age and its distal impact ejecta in the
Pierre Shale in southwestern South Dakota.
Science 262: 729-732.

JAEGER, J.J., COURTILLOT, V. & TAPPONNIER, P.
1989. Palaeontological view of the ages of the
Deccan Traps, the Cretaceous/Tertiary boundary,
and the India-Asia collision. Geology 17: 316-
319.

JAIN, S.L. 1989, Recent dinosaur discoveries in India,
including eggshells, nests and coprolites. Pp. 99-
108. In Gillette, D. & Lockley, M.G. (eds)
‘Dinosaur Tracks and Traces’. (Cambridge
University Press: Cambridge).

JANSON, L.F. 1993. Cometary impacts into ocean:
their recognition and threshold constraints for
biological extinctions. Paleogeography,
Paleoclimatology. Paleoecology 104: 271-286.

JOHNSON, D.A., BERGGREN, W.A. & DAMUTH,
J.E. 1982. Cretaceous ocean floor in the Amirante
Passage: tectonic and oceanographic implications.
Marine Geology 47: 31-343.

KAILA, K.L., MURTY, P.R.K., RAO, V.K. &
KHARETCHKO, G.E. 1981. Crustal structure
from deep-seismic surroundings along the Koyna
II (Kelsi-Loni) profile in the Deccan Trap area,
India. Tectonophysics 73: 365-384.

KAMEN-KAYE, M. 1985. Mesozoic columns below
the Seychelles Bank, western Indian Ocean. Jour-
nal of Petroleum Geology 8(3): 323-329.

KELLER, G. 1989. Extended period of extinctions
across the Cretaceous/Tertiary boundary in
planktonic foraminifera of continental-shelf sec-
tions; implications for impact and volcanism

KT EVENTS IN INDIA

theories. Geological Society of America Bulletin
101: 1408-1419.

1994. Global biotic effects of the KT boundary
event: mass extinction restricted to low latitudes?
Lunar and Planetary Institute Contributions 825:
57-58.

KELLER, G., MacLEOD, N., LYONS, J.B. & OF-
FICER, C.B. 1993. No evidence for Cretaceous-
Tertiary boundary-age deep water deposits in the
Caribbean and Gulf of Mexico. Geology 21: 776-
780.

KOEBERL, C. 1993. Chicxulub Crater, Yucatan: tek-
tites, impact glasses and the geochemistry of target
rocks and breccias. Geology 21: 211-214.

KOEBERL, C., MURALI, A.V., SHARPTON, V.L. &
BURKE, K. 1988. Geology and geoschemistry of
the Kara and Ust-Kara impact structures (USSR).
EOS, Transactions American Geophysical Union
69: 1263.

KOEBERL, C., SHARPTON, V.L., MURALI, A.V. &
BURKE, K. 1990. Kara and Ust-Kara impact
structures (USSR) and their relevance to the K/T
boundary event. Geology 18: 50-53,

KRING, D.A. & BOYNTON, W.V. 1992. Petrogenesis
of an augite-bearing melt rock in the Chicxulub
structure and its relationships to the K/T impact
spherules in Haiti. Nature 358: 141-144.

KRISHNAN, M.S. 1982. ‘Geology of India and
Burma’. (CBS Publishers: Delhi). 536 pp.

KYTE, F.T., BOSTWICK, J.A. & ZHOU L. 1994. The
KT boundary on the Pacific plate. Lunar and
Planetary Institute Contribution 825: 64-65.

LAKHANPAL, R.N. 1970. Tertiary floras of India and
their bearing on the historical geology of the
region. Taxon 19: 615-695.

LEVY, D.K., SHOEMAKER, E.M. & SHOEMAKER,
C.S. 1995. Comet Shoemaker-Levy 9 meets
Jupiter. Scientific American 273(2): 84-91.

LOEUFF, J. le, BUFFETAUT, E., MECHIN, P. &
MECHIN-SALESSY, A. 1989. A titanosaurid
dinosaur braincase (Saurischia, Sauropoda) from
the Upper Cretaceous of Var (Provence, France).
Comptes Rendus de I’ Academie des Sciences de
Paris 309: 851-857.

LOPEZ-RAMOS, E. 1975. Geological summary of the
Yucatan peninsula. Pp. 257-282. In Nairn, A.E.M.
& Stehli, FG. (eds) ‘The Ocean Basins and
Margins', vol. 3. (Plenum Press: New York).

LYONS, J.B. & OFFICER, C.B. 1992. Mineralogy and
petrology of the Haiti Cretaceous/Tertiary section.
Earth and Planetary Science Letters 109: 205-224.

MacINTYRE, R.M., DICKIN, A.P., FALLICK, A.E. &
HALLIDAY, A.N. 1985. An isotopic and
geochronological study of the younger igneous
rocks of the Seychelles. EOS, Transactions
American Geophysical Union 66: 1137.

MAHONEY, J.J. 1988. Deccan Traps. Pp. 151-194. In
Macdougall, J.D. (ed.) ‘Continental Flood
Basalts’. (Kluwer Academic Publishers:
Dordrecht).

529

MATLEY, C.A. 1923. Note on an armored dinosaur
from the Lameta beds of Jubbulpore. Records
Geological Survey of India 55: 105-109,

MAURASSE, FJ. & MAURASSE, M.R. 1990. The
Cretaceous-Tertiary boundary impact site in the
Caribbean. Geological Society of America
Abstracts with Programs 22(7): A77. (abstract)

MAURASSE, FJ. & SEN, G. 1991. Impacts, tsunamis,
and the Haitian Cretaceous/Tertiary boundary
layer. Science 252: 1690-1693.

McHONE, J.F., NIEMAN, R.A., LEWIS, C.F. &
YATES, A.M. 1989. Stishovite at the Cretaceous-
Tertiary boundary, Raton, New Mexico. Science
243: 1182-1184.

McINTOSH, J.S. 1990. Sauropoda. Pp. 345-401. In
Weishampel, D.B., Dodson, P. & Osmólska, H.
(eds) ‘The Dinosauria’. (University of California
Press: Berkeley).

McLEAN, D.M. 1985. Deccan Traps mantle degassing
in the terminal Cretaceous mass extinction.
Cretaceous Research 6: 235-259.

MELOSH, H.J. 1989, ‘Impact Cratering: A Geologic
Process'. (Oxford University Press: New York).
245 pp.

MELOSH, H.J., SCHENIDER, N.M., ZAHNLE, KJ.
& LATHAM, D. 1990. Ignition of global wildfires
at the Cretaceous/Tertiary boundary. Nature 343:
251-254.

MEYERHOFF, A.A., LYONS, J.B. & OFFICER, C.B.
1994. Chicxulub structure: a volcanic sequence of
Late Cretaceous age. Geology 22: 3-4.

MEYERHOFF, A.A. & KAMEN-KAYE, M. 1981.
Petroleum prospects of Saya de Malha and
Nazareth Banks, Indian Ocean. American As-
sociation of Petroleum Geologists Bulletin 65:
1344-1347.

MOHABEY, D.M. 1984. The study of dinosaurian eggs
from Infratrappean limestone in Kheda District,
Gujrat. Journal of Geological Society of India 23:
329-337.

1987. Juvenile sauropod dinosaur from Upper
Cretaceous Lameta of Panchmahals District,
Gujrat, India. Journal of Geological Society of
India 30(3): 210-216.

MOLNAR, R.E. 1990. Problematic theropods:
"Carnosauria'. Pp. 306-317. In Weishampel, D.B.,
Dodson, P. & Osmólska, H. (eds) "The Dinosauria’.
(University of California Press: Berkeley).

MOORE, H.H. 1976. Missile impact centers (White
Sands Missile Range, New Mexico) and applica-
tion to lunar research. Geological Survey Profes-
sional Papers 812B: B1-B47.

MORGAN, W.J. 1981. Hotspot tracks and the opening
of the the Atlantic and Indian Oceans. Pp. 203-
213. In Emiliani, C. (ed.) "The Sea’, Vol. 7. (Wiley:
New York).

NAINI, B.R. & TALWANI, M. 1982, Structural
framework and evolutionary history of the con-
tinental margin of western India. American As-
n of Petroleum Geologists Memoir 34:
167-193.

530

NEGI, J.G., AGARWAL, P.K., SINGTH, A.P. & PAN-
DEY, O.P. 1992. Bombay gravity high and erup-
tion of Deccan flood basalts (India) from a shallow
secondary plume. Tectonophysics 206: 341-350.

NEGI, J.G., AGARWAL, P.K., PANDEY, O.P. &
SINGH, A.P. 1993. A possible K-T boundary
bolide impact site offshore near Bombay and trig-
gering of rapid Deccan volcanism. Physics of the
Earth and Planetary Interiors 76: 189-197.

NORTON, I.O. & SCLATER, J.G. 1979. A model for
the evolution of the Indian Ocean and the breakup
of Gondwanaland. Journal of Geophysical Re-
search 84: 6803-6830.

OCAMPO, A.C. & POPE, K.O. A KT boundary section
from northern Belize. Lunar and Planetary In-
stitute Contribution 825: 86-87.

OFFICER, C.B., HALLAM, A., DRAKE, C.L.,
DEVINE, J.D. & MEYERHOFF, A.A. 1987. Late
Cretaceous paroxysmal Cretaceous/Tertiary ex-
tinctions. Nature 326: 143-149.

OFFICER, C.B., DRAKE, C.L. & PINDELL, J.L.
1992. Cretaceous-Tertiary events and the Carib-
bean caper. GSA Today 2(4): 69-75.

OLMEZI, L, FINNEGAN, D.L., & ZOLLER, W.H.,
1986. Iridium emissions from Kilauea volcano.
Journal of Geophysical Research 91: 653-664.

ORTH, CJ. 1989. Geochemistry of the bio-event
horizons. Pp. 37-72. In Donovan, S.K. (ed.) ‘Mass
Extinctions: Processes and Evidence'. (Columbia
University: New York).

OWEN, M.R. & ANDERS, M.H. 1988. Evidence from
cathodluminiscence for non-volcanic origin of
shocked quartz at the Cretaceous/Tertiary bound-
ary. Nature 334: 145-147.

PALADINO, F.V., DODSON, P., HAMMOND, J.K. &
SPOTILA, J.R. 1989. Temperature-dependent sex
determination in dinosaurs? implications for
population dynamics and extinction. Geological
Society of America Bulletin Special Paper 238:
63-70.

PANDE, K., VENKATESAN, T.R., GOPALAN, K.,
KRISHNAMURJ[HY, P. & MACDOUGALL,
J.D. 1988. " Ar/” Ar ages of alkali basalts from
Kutch, Deccan volcanic province, India. Pp. 145-
150. In Subbarao, K.V. (ed.) ‘Deccan Flood
Basalts'. Geological Society of India Memoir 10.

PAUL, D.K., POTTS, PJ., REX, D.C. & BECKIN-
SALE, R.D. 1977. Geochemical and petrogenetic
study of the Girnar igneous complex, Deccan
volcanic province, India. Geological Society of
America Bulletin 88: 227-234.

PENFIELD, G.T. & CAMAGRO, Z.A. 1982. Defini-
tion of a major igneous zone in the ventral Yucatán
platform with aeromagnetics and gravity.
Geophysics 47: 448-449,

POWAR, K.B. 1981. Lineament fabric and dyke pattern
in the western part of the Deccan volcanic
province. Pp. 45-57. In Subbarao, K.V. & Sukesh-
wala, R.N. (eds) ‘Deccan Volcanism’. Geological
Society of India Memoir 3.

MEMOIRS OF THE QUEENSLAND MUSEUM

PRASAD, G.V.R., JAEGER, J.J., SAHNI, A.,
GHEERBRANT, E. & KHAJURIA, C.K. 1994.
Eutherian mammals from the Upper Cretaceous
(Maastrichtian) intertrappean beds of Naskal,
Andhra Pradesh, India. Journal of Vertebrate
Palaeontology 14: 260-277.

PRINN, R.G. & FEGLEY, B.F. Jr 1987. Bolide impacts,
acid rain, and biospheric traumas at the
Cretaceous-Tertiary boundary. Earth and
Planetary Science Letters 83: 1-15.

QUERESHY, M.N., BRAHMAN, N., GARDE, S.C. &
MATHUR, B.K. 1968. Gravity anomalies and the
Godavari rift, India. Geological Society of
America Bulletin 79: 1221-1230.

RAMPINO, M.R. 1987. Impact cratering and flood
basalt volcanism. Nature 327: 468.

RAMPINO, M.R. & CALDEIRA, K. 1992. Antipodal
hotspot pairs on the earth. Geophysical Research
Letters 19: 2011-2014.

RAO, R.P. & TALUKDAR, S.N. 1980. Petroleum geol-
ogy of Bombay High Field, India. American As-
sociation of Petroleum Geologists Memoir 30:
487-506.

ROBIN, E., GAYRAUD, J., FROGET, L. & ROC-
CHIA, R. 1994. On the origin of the regional
variation in spinel compositions at the KT bound-

. Lunar and Planetary Institute Contribution
825: 96-97.

SAHNI, A. & BAJPAI, S. 1988. Cretaceous-Tertiary
boundary events: the fossil vertebrate, palacomag-
netic and radiometric evidence from peninsular
India. Journal of Geological Society of India 32:
382-396.

SAHNI, A. & KHOSHLA, A. 1994. Palaeobiological,
taphonomical and palaeoenvironmental aspects of
Indian Cretaceous sauropod nesting sites. Gaia 10:
215-223.

SAHNI, A., TANDON, S.K., JOLLY, A.S., BAJPAI, S.,
SOOD, A. & SRINIVASAN, S. 1994. Upper
Cretaceous dinosaur eggs and nesting sites from
the Deccan volcano-sedimentary province of
peninsular India. Pp. 204-226. In Carpenter, K.,
Hirsch, K.F. & Horner, J.R. (eds) ‘Dinosaur Eggs
and Babies’. (Cambridge University Press:
Cambridge).

SARKAR, A., BHATTACHARYA, S.K. &
MOHABEY, D.M. 1991. Stable-isotope analyses
of dinosaur eggshells: palacoenvironmental im-
plications. Geology 19: 1068-1071.

SCHULTZ, P.H. & GAULT, D.E. 1990. Prolonged
global catastrophes from oblique impacts.
Geological Society of America Special Paper 247:
239- 261.

SCHULTZ, P.H. & LIANZA, R.E. 1992. Recent graz-
ing impacts on the Earth recorded in the Rio
Cuarto Crater field, Argentina. Nature: 234-237,

SEN, G. 1988. Possible depth of origin of primary
Deccan tholeiitic magma. Pp. 25-51. In Subbarao,
K.V. (ed.) ‘Deccan Flood Basalts'. Geological
Society of India Bulletin 10.

KT EVENTS IN INDIA

SHARPTON, V.L. & WARD, P.D. (eds) 1990. Global
catastrophes in earth history. Geological Society
of America Special Paper 247: 1-631.

SHARPTON, V.L., DALRYMPLE, G.B., MARTIN,
L.E., RYDER, G., SCHUARAYTZ, B.C. & UR-
RUTIA-FUCUGACHI, J. 1992. New links be-
tween the Chicxulub impact structure and the
Cretaceous/Tertiary boundary. Nature 359: 819-
821.

SHEEHAN, P.M., FASTOVSKY, D.E., HOFFMAN,
R.G., BERGHAUS, C.B. & GABRIEL, D.I.
1991. Sudden extinction of the dinosaurs: latest
Cretaceous, Upper Great Plains, U.S.A. Science
254: 835-839.

SHOEMAKER, E.M. 1983. Asteroid and comet bom-
bardment of the Earth. Annual Review of Earth
Sciences 11: 461-494,

SHOR, G.G. & POLLARD, D.D. 1963. Seismic in-
vetigation of Seychelles and Saya de Malha
Banks, north-west Indian Ocean. Science 142:
48-49.

SIGNOR, P.W. III & LIPPS, J.M. 1982. Sampling bias,
gradual extinction patterns, and catastrophes in the
fossil record. Geological Society of America Spe-
cial Paper 190: 291-296.

SLOAN, R.E., RIGBY, J.K, VAN VALEN, L.M. &
GABRIEL, D. 1986. Gradual extinction and
simultaneous ungulate radiation in the Hell Creek
Formation, Science 232: 623-633.

SMIT, J. 1982. Extinction and evolution of planktonic
foraminifera at the Cretaceous/Tertiary Boundary.
Geological Society of America Special Paper 190:
329-352.

SMIT, J., MONTARI, A., SWINBURNE, N.H.M.,
HILDEBRAND, A.P., MARGOLIS, S.V.,
CLAEYS, P., LOWRIE, W. & ASARO, F. 1992.
Tektite-bearing, deepwater clastic unit at the
Cretaceous-Tertiary boundary in northeastern
Mexico. Geology 20: 99-103.

SMIT, J., ROEP, T.B., ALVAREZ, W., MONTANARI,
S. & CLAEYS, P. 1994. Stratigraphy and
sedimentology of KT clastic beds in the Moscow
Landing (Alabama) outcrop: evidence for impact-
related earthquakes and tsunamis. Lunar and
Planetary Institute Contributiom 825: 119-120.

SRIVASTAVA, S., MOHABEY, D.M., SAHNI, A. &
PANT, S.C. 1986. Upper Cretaceous dinosaur egg
clutches from Kheda District (Gujrat, India).
Palaeontographica A 193; 219-233.

STANLEY, S.M. 1987. ‘Extinction’. (W.H. Freeman:
San Francisco). 242 pp.

STERN, R.G. 1968. The Wells Creek structure, Ten-
nessee. Pp. 323-337. In French, B.M. & Short,
N.M. (eds) ‘Shock Metamorphism in Natural
Minerals’. (Mono Book Corporation: Baltimore).

STINNESBECK, W., BARBARIN, J.M., KELLER,
G., LOPEZ-OLIVA, D.G., PIVNIK, D.A.,
LYONS, J.B., OFFICER, C.B., ADATTE, T.,
GRAUP, G., ROCCHIA, R. & ROBIN, E. 1993.
Depositional of channel deposits near the
Cretaceous-Tertiary boundary in northeastern

531

Mexico: catastrophic or “normal” sedimentary
deposits? Geology 21: 797-800.

SUBBARAO, K.V., SREENIVASA RAO, M.,
RAMASUBBA REDDY, N., PRASAD, C. V.R.K.
& HARIHARAN, M. 1988. Pp. 199-212. In Sub-
barao, K.V. (ed.) ‘Deccan Flood Basalts’.
Geological Society of India Memoir 10, Ban-
galore.

SURLYK, F. & JOHNSON, M.B. 1984. End-
Cretaceous brachiopod extinctions in the chalk of
Denmark. Science 23: 1174-1177.

SUTHERLAND, FL. 1994, Volcanism around K/T
boundary time — its role in an impact scenario for
the K/T extinction events. Earth Science Reviews
36: 1-26.

SWISHER, C.C., GRAJALES-NISHIMURA, J.M.,
MONTANARI, A., MARGOLIS, S.V., CLAEYS,
P., ALVAREZ, W., RENNE, P., CEDILLO-
PARDO, E., MAURRASSE, FJ.-M.R., CURTIS,
G.H., SMIT, , . & McWILLIAMS, M.O. 1992.
Coeval ^ Ar/"" Ar ages of 65.0 million years ago
from Chicxulub crater melt rock and Cretaceous-
Tertiary boundary tektites. Science 257: 954-958.

THIERSTEIN, H.R. 1982. Terminal Cretaceous
plankton extinctions: a critical assessment.
Geological Society of America Special Paper 190:
385-405.

THOMPSON, G. & NELSON, W.G. 1972. The petrol-
ogy of oceanic crust and fracture zones in the
Atlantic Ocean: evidence of a new kind of sea
floor spreading. Journal of Geology 80: 526-538.

TSCHUDY, R.M. & TSCHUDY, B.D. 1986. Extinction
and survival of plant life following the
Cretaceous/Tertiary boundary event, western inte-
rior, North America. Geology 14: 667-670.

TUNNICLIFFE, V. 1991. The biology of hydrothermal
vents: ecology and evolution. Marine Biology
and Oceanography: an Annual Review. 29: 319-
407.

TUREKIAN, K.K. 1982. Potential of '*’0s/'*°Os as a
cosmic versus terrestrial indicator in high iridium
layers of sedimentary strata. Geological Society of
America Special Paper 190: 243-249.

VANDAMME, D. & COURTILLOT, V. 1992.
Palaeomagnetic constraints on the structure of the
Deccan Traps. Physics of the Earth and Planetary
Interior 74: 241-261.

VENKATESAN, T.R., PANDE, K. & GOPALAN, K.
1993, Did Deccan volcanism pre-date the
Cretaceous/Tertiary transition? Earth and
Planetary Science Letters 119: 181-189.

WARD, P.D., KENNEDY, W.J., MacLEOD. K.G. &
MOUNT, J.F. 1991. Ammonite and inoceratid
bivalve extinction patterns in Cretaceous/Tertiary
boundary section of the Biscay region (south-
western France, northern Spain). Geology 19:
1181-1184.

WATTS, A.W., GREELEY, R. & MELOSH, H.J. 1991.
The formation of terrains antipodal to major im-
pacts, Icarus 93: 159-168.

532

WEISHAMPEL, D.B, GRIGORESCU, D. & NOR-
MAN, D.B. 1991. Dinosaurs of Transylvania: Is-
land biogeography in the Late Cretaceous.
National Geographic Research and Exploration
7(2): 196-215.

WEST, W.D. 1958. The petrography and petrogenesis
of forty-eight flows of Deccan Trap penetrated by
boring in western India. Transactions National
Institute of Science, India. 4(1): 1-56.

WETHERILL, G.W. & SHOEMAKER, G.M. 1982.
Collision of astronomically observable bodies
with the Earth. Geological Society of America
Special Paper 190: 1-13.

WHITE, R.S. & McKENZIE, D.P. 1989. Volcanism at
rifts. Scientific American 261(1): 62-71.

MEMOIRS OF THE QUEENSLAND MUSEUM

WILHELMS, D.E. 1987. The geologic history of the
moon. U.S. Geological Survey Professional Paper
M1348: 1-293.

WILSON, E.O. 1992. ‘The Diversity of Life’. (W.W.
Norton: New York). 424 pp.

WOLBACH, W.S., GILMOUR, I., ANDERS, E., ORTH,
CJ. & BOOK, R.R. 1988. Global fire at the
Cretaceous-Tertiary boundary. Nature 324: 665-669.

YADAGIRI, P. & AYYASAMI, K. 1979. A new
stegosaurian dinosaur from the Upper Cretaceous
sediments of South India. Journal Geological
Society of India 20: 521-530.

ZINSMEISTER, W.J., FELDMAN, R.M., WOOD-
BURNE, M.O. & ELLIOT, D.H. 1989. Latest
Cretaceous/earliest Tertiary transition on
Seymour Island, Antarctica. Journal of Palaeon-
tology 63: 731-738.

EARLY AVIAN EVOLUTION IN THE SOUTHERN HEMISPHERE: THE FOSSIL
RECORD OF BIRDS IN THE MESOZOIC OF GONDWANA

LUIS M. CHIAPPE

Chiappe, L.M. 1996 12 20: Early Avian evolution in the Southern Hemisphere: The fossil
record of birds in the Mesozoic of Gondwana, Memoirs of the Queensland Museum 39(3):
533-554. ISSN 0079-8835.

The record of Gondwanan Mesozoic birds, including osteological specimens, feathers and
traces, is critically reviewed. Data regarding the paleoenvironment and associated biota of
each record is provided. Several occurrences of controversial status in the Late Triassic-Early
Jurassic of Africa and the Cretaceous of Australia and Africa, and misguided reports from
the Cretaceous of South America are also discussed. The Mesozoic record of Gondwanan
birds is limited, although it has provided relevant insights about the early evolution of birds.
Undisputable records are known from the Cretaceous of South America, Australia, Africa
and Antarctica. This material indicates that during the Cretaceous, Gondwanan birds were
widely distributed, inhabiting a broad range of environments and developing various modes
of life. [ ] Birds, Mesozoic, Gondwana, evolution.

Se revisa criticamente el registro Mesozoico de aves de Gondwana, incluyendo restos óseos,
plumas y trazas. Para cada registro, se provee información con respecto al paleoambiente y
la fauna asociada. Se discuten varios registros de status controvertido en el Triásico
Tardío-Jurásico Temprano de Africa, y el Cretacico de Australia y Africa, como también
reportes erroneos del Cretácico de América del Sur. El registro Mesozoico de aves de
Gondwanaes limitado, si bien ha provisto información relevante sobre la temprana evolución
de las aves. Registros confiables son conocidos del Cretácico de América del Sur, Australia,
Africa y Antártida. Este material indica que durante el Cretácico las aves de Gondwana
estuvieron extensamente distribuidas, habitando un amplio espectro de ambientes y desar-
rollando diversos estilos de vida.

Department of Ornithology, American Museum of Natural History, Central Park West at
79th Street, New York, NY 10024, USA; 5 April 1996.

The early evolutionary history of birds has been
reconstructed mostly on the basis of fossil
evidence provided by Mesozoic deposits of the
northern hemisphere. It is not until recently that
significant discoveries of Mesozoic birds from
Gondwana have been made, and that this new
evidence has strongly influenced our previous
conceptions of avian evolution (Chiappe, 19952).

The indisputable record of Mesozoic birds in
Gondwana is restricted to the Cretaceous, al-
though bird-like footprints have been found in
pre-Cretaceous deposits. Among today's
Gondwanan continental masses, South America,
Antartica, Australia and Africa have provided
Cretaceous evidence of birds. The South
American (Chiappe, 1991) and Australian (Vick-
ers-Rich, 1991) records have been previously
discussed elsewhere, but additional new findings
in these landmasses, along with the discovery of
Cretaceous birds in Antarctica and Africa, and the
recent interest in the pre-Cretaceous bird-like
footprints from the latter continent, call for an
update of the Gondwanan record. In this paper the
Gondwanan record (Fig. 1, Table 1) is reviewed,

including data on paleoenvironment and as-
sociated biota, and its contribution to the under-
standing of the early history of birds is
emphasised.

MATERIALS AND METHODS

The taxonomical nomenclature employed here
follows the traditional use of the taxon-name
Aves. Hence, the term Aves names a clade,
defined as the common ancestor of Archaeop-
teryx lithographica and Neornithes (modern
birds) and all its descendants (Fig. 2). Gauthier
(1986) and others (Perle et al., 1993, 1994; Norell
et al., 1993) proposed restricting the taxon-name
Aves to the crown-group of Theropoda, that is the
common ancestor of all extant forms plus all its
descendants. These authors (including myself in
some cases) have used the term Avialae to name
the clade traditionally regarded as Aves, while the
latter term was used to replace the taxon-name
Neornithes of the classical ornithological
taxonomy. Perhaps the most attractive advantage
of Gauthier's (1986) proposal is the idea of

534

MEMOIRS OF THE QUEENSLAND MUSEUM

2b 2a
AUSTRALIA

15

FIG. 1. Maps of A, South America and Argentina-Chile; B, Australia; C, Antarctica; and D, Africa, showing the
occurrences of Mesozoic birds. W = Early Cretaceous occurrences. @ = Late Cretaceous occurrences. Numbers
pointing each locality correspond to those used in Table 1. 1, Koonwarra, Victoria; 2a, "Warra', Queensland;
2b, ‘Canary’, Queensland; 3, Santana do Cariri, Ceará; 4, area of Plaza Huincul, Neuquén; 5a, city of Neuquén,
Neuquén; 5b, Puesto Tripailao, Río Negro; 6, Estancia Los Alamitos, Río Negro; 7, Salitral Moreno, Río Negro;
8, Bahía San Vicente, Concepción; 9, area of Ingeniero Jacobacci and ‘Monton-I10’, Río Negro; 10, El Brete,
Salta; 11, Quebrada del Tapon, Salta; 12, Cape Lamb, Vega Island; 13, Marambio (Seymour) Island; 14, area
of Agadir (Morocco); 15, Berivotra, Mahajanga (Madagascar).

having a widely used term, such as Aves, as-
sociated with a clade that can be diagnosed by
numerous synapomorphies (many soft-tissue
characters in addition to the osteological ones). It
should be noted, however, that many of these
characters (e.g., soft-tissue characters) have an
ambiguous optimization when fossils are in-
cluded in the analysis. The greatest argument
against this proposal appears to be simply the
universal way in which the name Aves has been
associated to the clade composed of all birds (note
that a phylogenetic definition is given above)
during the last 130 years. Names are terms
applied to a definition and ideally they should be
conventional. For this reason I would keep using
the term Aves in the way it is defined above. The
casual terms ‘birds’ and ‘modern birds’ refer to
Aves and Neornithes, respectively. The collective
expression 'non-avian theropods' is used without
any systematic meaning.

Several of the occurrences here regarded as
Mesozoic birds are represented by footprints. It
is well known that the identification of a footprint
with a particular trackmaker always involves
some degree of uncertainty. Several factors such
as the stance, march, speed and the composition
of the substrate may affect the morphology of the
footprint and consequently its taxonomic inter-
pretation (Padian & Olsen, 1984; Unwin, 1989).
In this paper, the recognition of a footprint as
avian is based primarily on the criteria sum-
marised by Lockley et al. (1992). Most significant
criteria used by Lockley et al (1992) are: 1) small
size; 2) slender digit impressions with undifferen-
tiated pad impressions; 3) wide divarication angle
between digits II and IV; 4) plantarily directed
hallux; 5) slender claws; and 6) distal curvature
of lateral claws away from the axis of the pes.

The treatment of the avian occurrences of a
particular continent are organised in the follow-

FOSSIL RECORD OF BIRDS IN THE MESOZOIC GONDWANA 535

TABLE 1. Avian occurrences in the Mesozoic of Gondwana. Alb=Albian, Apt=Aptian, Bar=Barremian,
Cam=Campanian, Con=Coniacian, Mae=Maastrichtian, Tur=Turonian, San=Santonian.

DISTRIBUTION

1 Aves indet. Koonwarra, Victoria

HORIZON/AGE | EVIDENCE REFERENCE

AUSTRALASIA
Strzelecki Group,

Bar-Apt Talent et al., 1966; Rogers, 1987

2ab Enantiornithes: Warra Station (2a), Canary Toolebuc Fm., Alb
Nanantlus eos 2b), Queensland

Molnar 1986; Molnar pers. comm.

3 Aves indet. Brazil

SOUTH AMERICA
Santana de Cariri, Ceará, Santana Fm., Apt- Martins-Neto & Kellner, 1988;

Alb Kellner, 1990

Alvarezsauridae: Area of Plaza Huincul, Río Neuquén Fm., Novas & Coria,1990; Novas, this
" F bone
Patagonykus puertai Neuquén, Argentina Tur-Con volume
Alvarezsauridae: City of Neuquén, Neuquén, | Río Colorado Fm., Bonaparte, 1991; Novas, this
h à bone
Alvarezsaurus calvoi Argentina Con-San volume

Argentina

P . Estancia Los Alamitos, Río | Los Alamitos Fm. $
nm Omithothoraces indet. Negro, Argentina Cam-Mae Chiappe, 1992a

Patagopteryxdeferrariisi City of Neuquén, Neuquén, | Rio Colorado Fm., bone Alvarenga & Bonaparte, 1992;
ede Argentina Con-San Chiappe, 1992a
Enantiornithes: RM Sete Rio Colorado Fm., bene Chiappe, 1992a; Chiappe &

Neuquenornis volans p 3 c Con-San Calvo, 1994

: , Salitral Moreno, Río Negro, Allen Fm., VES
7 Ornithurae indet. Arena Cam-Mae bone Powell, 1987; Chiappe, 1992a
Ornithurae: Gaviidae: Bahía San Vicente Quiriquina Fm., Bone Lambrecht, 1929; Olson, 1992
Neogaeornis wetzeli Concepción, Chile Cam-Mae
Aves det , Area of Ln Jacobaccl, Pa Casamiqueta, 1987; Leonardi,
9 “Patagonichornis "Monton-110", Río Negro, | Unknown Fm., Mae ichnite z
a is à 1987; Chiappe, 1991
venetiorum Argentina
10 | Enantiorithesindet. | ElBrete,Salta,Argentina | LechoFm.Mae | bone | Walker, 1981; Chiappe, 1991
10 Hune omite; t El Brete, Salta, Argentina Lecho Fm., Mae bone Walker, 1981; Chiappe, 1991
Enantiornis leali
10 Biointióniiter: . El Brete, Salta, Argentina Lecho Fm., Mae bone Walker, 1981; Chiappe, 1993
Soroavisaurus australis
jo | Enantiomithes: Lectavis | s e Salts, Argentina | Lecho Fm., Mae bone Walker, 1981; Chiappe, 1993
bretincola
Enantiornithes:
10 Yungavolucris El Brete, Salta, Argentina Lecho Fm., Mae bone Walker, 1981; Chiappe, 1993
brevipedalis
Aves indet.: Quebrada del Tapón, Salta, " "TOR Alonso & Marquillas, 1986;
at Yacoraitichnus avis Argentina Yacoraite Fm., Mae schnite Chiappe, 1991
ANTARCTICA 3
Omithurae: López de Bertodano : "
12 Presbyomithidae indet. Cape Lamb, Vega Island Fm., Cam-Mae bone Noriega & Tambussi, 1995
Omithurae: Gaviidae i López de Bertodano 3 .
indet. Marambio (Seymour) Island Fm, “Mae Chatterjee, 1989; Olson, 1992
AFRICA
14 Aves indet. Area of Agadir, Morocco | Unknown Fm., Mae ichnite Amborg is Lapparent 8954:
Lockley et al., 1992
Aves indéi: Berivotra, Mahajanga, Maevarano Fm., bone Kitiise, pers: comi,
NS Madagascar Cam

ing hierarchical arrangement: Aves, Metornithes,
Ornithothoraces, Omithurae, Neornithes. In-
determinate occurrences within a particular clade
are listed first.

When anatomical nomenclature is applied, this
follows Baumel & Witmer (1993), using the
English equivalents of the Latin terminology.

536

* Taxa recorded in Gondwana

Ornithurae

Ornithothoraces
Metornithes

Aves (=Avialae)

MEMOIRS OF THE QUEENSLAND MUSEUM

7 *
Neornithes

Ichthyornithiformes
Hesperornithiformes

*
Patagopteryx

TRA *
Enantiornithes
Iberomesornis

A *
Alvarezsauridae

Archaeopteryx

FIG. 2. Cladogram illustrating the relationships between Alvarezauridae, Enantiornithes, Patagopteryx defer-
rariisi and Neornithes (see Chiappe et al., this volume).

Institutional abbreviations. DNPM, Depar-
tamento Nacional da Produção Mineral (Rio de
Janeiro, Brazil); GP, Instituto de Geociéncias,
Universidade de São Paulo (Sao Paulo, Brazil);
GPMK, Geologisch-Paláontologisches Institut
und Museum, Kiel; LEIUG, Department of Geol-
ogy, Leicester University (Leicester, United
Kingdom); MACN, Museo Argentino de Cien-
cias Naturales (Buenos Aires, Argentina);
MUCPy, Museo de Ciencias Naturales, Univer-
sidad Nacional del Comahue (Neuquén, Argen-
tina); NMV, National Museum of Victoria
(Melbourne, Australia); PVL, Fundación-
Instituto Miguel Lillo, Universidad Nacional de
Tucumán (San Miguel de Tucumán, Argentina);
PVPH, Museo Cármen Funes (Plaza Huincul,
Argentina); QM, Queensland Museum (South
Brisbane, Australia; SAMK, South African
Museum (Cape Town, South Africa); SMNK,
Staatliches Museum für Naturkunde Karlsruhe
(Karlsruhe, Germany).

CONTROVERSIAL OCCURRENCES

Ellenberger (1972) described a series of bird-
like footprints from the Lower Stormberg beds
(Late Triassic) of South Africa, recognising
several tridactyl and tetradactyl small ichno-
species (Trisauropodiscus galliforrna, T.
phasianiforma, T. levis, T. aviforma) and other
medium size tetradactyls (T. popompoi, T. super-
aviforma). Later, Ellenberger (1974) reported
several other footprints from the upper Storm-
berg beds (Early Jurassic) recognising the
ichnogroups carnavians (Masitisisauropus

palmipes, M. angustus, M. exiguus, Masitisisauro-
pezus minimus, M. minutus, M. perdiciforma) and
lacunavians (Masitisisauropodiscus turdus, M.
fringilla). Interestingly, among the carnavians
(M. palmipes), he claimed the presence of
feather-like traces associated to impressions of
the thoracic limb. Unfortunately, Ellenberger's
descriptions appear to have been made directly in
the field, and no holotype or referred specimen
has ever been collected and deposited in an in-
stitution (Lockley, pers. comm.). This situation
limits any comparison to the poorly figured
illustrations of his papers (Ellenberger, 1972,
1974). In the case of the feather-like structures
(M. palmipes), as pointed out by Molnar (1985),
there is no compelling evidence to regard these
traces as feathers and the footprints themselves
are probably not of an avian theropod. The dif-
ferent footprints of Trisauropodiscus, however,
have the small size and large divarication angle
between digits II and IV characteristic of avian
trackmakers (Lockley et al., 1992). Lockley et al.
(1992) remarked upon the similarity of
Trisauropodiscus with footprints reported by
Ishigaki (1985, 1986) from the Early Jurassic of
the Atlas Mountains in Morocco, pointing out that
they would have been considered as being made
by an avian trackmaker if they were of late
Mesozoic age. However, the existence of pre-
Late Jurassic birds has yet to be documented. The
recent allocation of Protoavis texensis (Chatter-
jee. 1987, 1991) within Aves is not supported by
the available evidence. Several authors have ex-
pressed their scepticism of Chatterjee's claims
(see Ostrom, 1987, 1991; Anderson, 1991;

FOSSIL RECORD OF BIRDS IN THE MESOZOIC GONDWANA

Monastersky, 1991), and after examination of the
material, I also disagree with his hypothesis
(Chiappe, 1995a). Therefore, given the fact that
indisputable pre-Late Jurassic birds are yet to be
documented in addition to the problems as-
sociated with the nature of footprints (see
Materials and Methods), I prefer to leave these
tracks out of the discussion of true birds.

Dettmann et al. (1992: note added in proof)
mentioned the discovery of new avian fossils
from the Early Cretaceous (Albian) Griman
Creek Formation at Lightning Ridge (north-
central New South Wales, Australia). Thanks to
R.E. Molnar, I had the opportunity to examine
casts of this material which comprise the distal
ends of two small bones, possibly tibiotarsi. The
specimens are fragmentary and I am not com-
pletely confident of their identification as
tibiotarsi. However, if they are tibiotarsi, then the
proximal tarsals are completely fused to the tibia
and the fibula does not reaches the tarsus. These
two characteristics are unique of avians among
dinosaurs (Gauthier, 1986; Cracraft, 1986;
Chiappe & Calvo, 1994; Chiappe, 1995b, 1996),
and support Dettmann's et al. (1992) allocation.
In light of the fragmentary condition of these
specimens, however, I would wait for further
evidence before regarding them as truly avian.

The occurrence of a putative bird bone was
mentioned by Smith (1988) in his analysis of the
paleoenvironment of a Late Cretaceous crater-
lake in Stompoor, central South Africa. Unfor-
tunately, this specimen (SAMK-6486) is a
fragmentary left femur and is missing the distal
end and most of the proximal one. Although
hollow, small and bird-like in aspect, SAMK-
6486 does not show any character that can be
regarded as avian.

Mussi & Alonso (1989) reported an isolated
bird footprint from the Late Cretaceous
(Maastrichtian) deposits of the Yacoraite Forma-
tion at Dique Cabra Corral (Province of Salta,
Argentina). In my opinion this ‘footprint’ is ac-
tually a peculiar disposition of small concretions
embedded in the limestone slab and not a track
made by a living organism.

Finally, in 1991 Icommented on the occurrence
of a fragmentary carpometacarpus from the Late
Cretaceous (Maastrichtian) Bauru Formation of
southern Brazil (Chiappe, 1991). Alvarenga
(pers. comm.), however, has recently identified
this bone as the extant Yellow-billed cuckoo
(Coccyzus americanus), being the result of con-
tamination of the Cretaceous material.

537

THE AVIAN RECORD
AUSTRALASIA

AVES Linnaeus, 1758

MATERIAL EXAMINED. At least five feathers (Fig.
3A), including covert and down ones (Talent et al.,
1966; Waldman, 1970; Rogers, 1987; Vickers-Rich,
1991), Waldman (1970) and Rogers (1987) published
on NMV P.26059A and P.26059B, slab and counter-
slab, respectively.

LOCALITY AND HORIZON. Koonwarra, southeast
Victoria, Australia. Strzelecki Group, Early Cretaceous
(Aptian; Vickers-Rich, 1991; Dettmann et al., 1992).

REMARKS. These feathers (Fig. 3À) represent
the oldest evidence of birds from this continent.
Rogers (1987) studied NMV P.26059A and
P.26059B in detail. This feather, regarded as
either a primary covert or an alular, was found to
be different from any modem feather (Rogers,
1987). This is interesting considering that the
only avian material from the Cretaceous of
Australia that allows a more precise taxonomic
identification belongs to the Enantiornithes (see
below), a group not closely related to modern
birds (Fig. 2).

The lacustrine mudstones of Koonwarra have
provided one of the best Mesozoic continental
biota from Australia, including a diversified flora
(e.g., bryophytes, ferns, conifers and an-
giosperms; see Douglas & Williams, 1982), and
a large variety of insects, crustaceans, spiders,
earthworms, bryozoans, bivalves and fishes (Det-
tmann et al., 1992). This biotic association be-
comes of particular interest considering that
southeastern Victoria was situated in polar
latitudes (~75°-85°S) during the early Cretaceous
(Vickers-Rich & Rich, 1989). Douglas & Wil-
liams (1982) interpreted the paleoenvironment at
Koonwarraas forested (primarily evergreen forests)
in a warm to cool-temperate climate with moderate
seasonality. However, Vickers-Rich & Rich
(1989) emphasised the possibility of annual near-
freezing conditions and a prolonged polar night.

METORNITHES Perle et al., 1993
ORNITHOTHORACES Chiappe, 1995
ENANTIORNITHES Walker, 1981

Nanantius eos Molnar, 1986

MATERIAL EXAMINED. HOLOTYPE: QMF12992
(Fig. 3C), complete left tibiotarsus (Molnar, 1986).
REFERRED SPECIMENS: QMF31813, proximal end
of a tibiotarsus (Molnar & Kurochkin, in press).

538

FIG. 3. Early Cretaceous birds from South America and Australasia. A,
isolated feather (GP/2T-136) from the Late Aptian Crato Member of the
Santana Formation near Santana do Cariri, northeastern Brazil (after Mar-
tins Neto & Kellner, 1987). B, Aptian feathers from the Strzelecki Group
at Koonwarra, southeastern Victoria, Australia (after Vickers-Rich, 1976).
C, medial (left) and cranial (right) views of the tibiotarsus (QMF12992) of
Nanantius eos from the Albian Toolebuc Formation at Warra Station,
western Queensland, Australia (after Molnar, 1986). Arrow points to the
broad medial condyle characteristic of enantiornithine birds. In C scale 2 1cm.

QMF12991, an anterior dorsal vertebra that may also
be Nanantius eos (Molnar & Kurochkin, in press).

LOCALITY AND HORIZON. QMF12992 and
QMF12991 were found at Warra Station, east side of
Hamilton River near Boulia, west Queensland,
Australia. QMF31813 is from Canary Station in west
Queensland, Australia. Toolebuc Formation, Early
Cretaceous (Albian; Molnar, 1986; Dettmann et al.,
1992).

REMARKS. Molnar (1986) correctly placed
Nanantius eos into the Enantiornithes. The
tibiotarsus of N. eos shows a round proximal
articular surface, a smooth craniolateral cnemial
crest, and a wide and bulbous medial condyle
(Fig. 3C), which are enantiornithine synapomor-

MEMOIRS OF THE QUEENSLAND MUSEUM

phies (Chiappe, 1992b, 1993).
N. eos is a small enantiornith-
ine, approximately the size of
the American robin (Turdus
/ migratorius) (Molnar, 1986).
Nevertheless, the complete
fusion of the proximal tarsals to
the tibia in the holotype of N.
eos suggests that this repre-
sents an adult individual. The
size of N. eos is significantly
smaller than-any of the South
American enantiornithines,
and comparable to the size of
the enantiornithines Alexomis
antecedens (Brodkorb, 1976)
and Concomis lacustris (Sanz
et al., 1995) from the Cretaceous
of Mexico and Spain, respec-
tively.

Molnar & Kurochkin (in
press) have identified QMF-
31813 as Nanantius sp. remark-
ing on a few subtle differences
with respect to Nanantius eos.
In my opinion, this specimen
should be placed within N. eos.
QMF1299], found at less than
5cm from the holotype of N.
eos, may also belong to this
species (Molnar & Kurochkin,
in press), but these authors
have pointed out that it belongs
to an individual larger than the
holotype of this species.

The Toolebuc Formation is
composed of offshore marine
limestones deposited during
the middle-late Albian (Molnar
& Thulborn, 1980; Dettmann et
al., 1992). The holotype specimen was associated
with fish remains, turtles, ichthyosaurs and
pterosaurs (Molnar, 1986). The common occur-
rence of fossil logs within these deposits might
indicate that the shoreline was forested near the
type locality of Nanantius eos (Dettmann et al.,
1992). N. eos is the only occurrence of a
Gondwanan enantiornithine in marine deposits.

SOUTH AMERICA
AVES Linnaeus, 1758
MATERIAL EXAMINED. SEVERAL ISOLATED

FEATHERS: GP/2T-136 (Fig. 3B), a flight feather
(Martins-Neto & Kellner, 1988); DNPM MCT 1493-R,

FOSSIL RECORD OF BIRDS IN THE MESOZOIC GONDWANA

a down feather (Kellner et al. 1994); LEIUG 114369,
a semiplume (Martill & Filgueira, 1994); LEIUG
115562, SMNK 1247 PAL (counterpart), a contour,
body feather (Martill & Frey, 1995).

LOCALITY AND HORIZON. All feathers are from
the Araripe Basin in the State of Ceará, Brazil. GP/2T-
136 and DNPM MCT 1493-R come from an indeter-
minate area around Nova Olinda (Kellner, pers.
comm.); LEIUG 115562, SMNK 1247 PAL and
LEIUG 114369 are from the Mina de Antone Phillipe,
near Tatajuba. Santana Formation, Crato Member,
Early Cretaceous (Late Aptian; Pons et al, 1990).

REMARKS. GP/2T-136 is a fairly small feather
(approximately 64mm long), which might have
belonged to a bird of the size of the neotropical
brush finches (Atlapetes) (Kellner et al., 1991). It
has a strong rachis and a very narrow outer vane
(Kellner et al., 1991). The fact that this feather has
asymmetric vanes (Fig. 3B) suggests it belonged
to a flying bird (Feduccia & Tordoff, 1979), al-
though no taxonomic assignment is possible with
only this evidence. In LEIUG 115562 (and its
counterpart, SMNK 1247 PAL), a small body
feather, the color pattering is preserved. This con-
sists of a series of transverse dark and light bands
(Martill & Frey, 1995). Barbules have been
preserved in both DNPM MCT 1493-R and
LEIUG 114369 (Kellner et al., 1994; Martill &
Filgueira, 1994). The presence of semiplumes,
flight, body and down feathers in the Crato Mem-
ber highlights the differentiation of feather types
as early as the Early Cretaceous.

The paleoenvironment of the Crato Member
has been generally regarded as a calm, fresh water
lake existing under arid climatic conditions
(Maisey, 1991), although restricted connections
to marine waters may have existed (Martill &
Filgueira, 1994; Martill & Frey, 1995). Sediments
consist of laminated limestones with intervening
beds of sands and shales (Maisey, 1991; Martill
& Filgueira, 1994). The feathers were found in
the laminated limestones, regarded as the mar-
ginal lake facies (Kellner et al., 1991; Martill &
Filgueira, 1995). The fossil assemblage recorded
from the Crato Member includes an abundant
variety of insects (Grimaldi, 1990), along with
scorpions, spiders, fishes, frogs and several floral
elements (Maisey, 1991).

Yacoraitichnus avis Alonso & Marquillas,
1986

MATERIAL EXAMINED. Slab (without number
when published) with several footprints (Fig. 4A)
(Alonso & Marquillas, 1986; Alonso, 1989).

539

LOCALITY AND HORIZON. Quebrada del Tapón,
Valle del Tonco, Department of San Carlos, Province
of Salta, Argentina. Yacoraite Formation, Late
Cretaceous (Maastrichtian; Alonso & Marquillas,
1986).

REMARKS. Yacoraitichnus avis consists of
medium size footprints (approximately 80mm
long), lacking the hallux impression (Alonso &
Marquillas, 1986) (Fig. 4A). This ichno-species
is one of the few avian tracks that shows digital
pad impressions (Lockley et al., 1992). Alonso &
Marquillas (1986) suggested that Y. avis could
either belong to a galliform-like neognathine or
to an enantiornithine bird, although they finally
regarded it as indeterminate among avians. The
hypothesis of a galliform affinity becomes un-
likely considering that the earliest known gal-
liform comes from the Early Eocene of North
America. Likewise, despite the fact that diverse
ensemble of enantiornithine birds is known from
the Lecho Formation (see below) — a unit con-
sidered to be part of the same depositional event
of the Yacoraite Formation (Gomez Omil et al.,
1989) — no ichnites have been ever found as-
sociated with these and other enantiornithine
remains making the latter hypothesis untestable
(Chiappe, 1991).

The footprints of Yacoraitichnus avis occur in
a narrow band of green claystone one meter
above deposits with a variety of non-avian
dinosaur footprints (Alonso & Marquillas, 1986;
Alonso, 1989). Remains of fishes and crocodiles
are also known to occur in the Yacoraite Forma-
tion at Quebrada del Tapón (Alonso & Marquil-
las, 1986). The footprints at Quebrada del Tapón
occur in coastal plain beds deposited along the
western margin of an extensive body of water
(Alonso, 1989), This coastal plain was peri-
odically submerged because of changes in the
level of the water. Alonso (1989) concluded that
the footprints of Y. avis were produced on an
exposed layer of mud, left after a period of inun-
dation,

‘Patagonichnornis venetiorum’ Casamiquela,

MATERIAL EXAMINED. Footprints (no repository
has been published; see Casamiquela, 1987).

LOCALITY AND HORIZON. Area of Ingeniero
Jacobacci and 'Monton-110', Province of Rio Negro,
Argentina. The footprints from around Ingeniero
Jacobacci are referred to the Late Cretaceous without
any further stratigraphical information (see Casami-

540

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 4. Late Cretaceous footprints from South Americaand Africa. A, Yacoraitichnus avis from the Maastrichtian

Yacoraite Formation at Quebrada del Tapón, Province of Salta, Argentina (after Alonso & Marquillas, 1986).
B, C, D, bird tracks from Maastrichtian deposits near Agadir, southern Morocco (after Ambroggi & Lapparent,

1954).

quela, 1987); the ones from 'Monton-T10' are listed as
Late Maastrichtian (see Leonardi, 1987: Plate XVIIT).

REMARKS. Casamiquela (1987) reported the
discovery of bird footprints associated with
tracks of hadrosaurs and, although they were
neither described nor illustrated, he named them
*Patagonichnornis venetiorum'. On the other
hand, a cast with several bird tracks from the Late
Maastrichtian of the Argentine Río Negro
Province (‘Monton-I10’) was illustrated by
Leonardi (1987), although no further geographi-
cal reference was given. These latter tracks have
long, slender digits and large divaricate angles
between digits II and IV (Leonardi, 1987). Al-
though not verified, it is likely that Leonardi
(1987) was illustrating what Casamiquela (1987)
named ‘P. venetiorum’.

Casamiquela (1987) allocated ‘Patagonichnor-
nis venetiorum’ in the extinct Cimolopterygidae
within the Charadriiformes, the group that in-
cludes most extant shorebirds. As I have pre-
viously pointed out (Chiappe, 1991), there is no
evidence to support such taxonomic allocation.
The Cimolopterygidae are only known from ele-
ments of the forelimb and thoracic girdle (Brod-
korb, 1963a; Olson & Parris, 1987; Tokaryk &
James, 1989), and they are not the only
Cretaceous charadriiforms (Olson, 1985; Olson
& Parris, 1987). Furthermore, there are other
shorebirds in the Cretaceous, such as Gansus
yumenensis from the Early Cretaceous of China
(Hou & Liu, 1984) and a new Presbyornithidae
from the Late Cretaceous of Antarctica (Noriega
& Tambussi, 1995; see below). Moreover, the
morphology of the hind limb of the enantior-
nithine Lectavis bretincola from the Late Cre-
taceous Argentine Province of Salta (Chiappe,
1993; see below) suggest wading habits (Walker,
1981), making it equally likely that the footprints
from Ingeniero Jacobacci belonged to an enan-
tiornithine bird. Due to the lack of information, I

prefer to consider the identity of these footprints
as indeterminate among avians.

METORNITHES Perle et al ., 1993
ALVAREZSAURIDAE Bonaparte, 1991

Patagonykus puertai Novas, this volume

MATERIAL EXAMINED. HOLOTYPE: PVPH-37,
isolated dorsal, synsacral and caudal vertebrae, along
with portions of thoracic limb and coracoids, and pelvic
girdle and limb (see Novas, this volume; in press).

LOCALITY AND HORIZON. Sierra del Portezuelo,
Province of Neuquén, Argentina. Rio Neuquén Forma-
tion, Late Cretaceous (Turonian-Coniacian; Legarreta
& Gulisano, 1989).

REMARKS. Patagonykus puertai is unques-
tionably a new taxon, although the actual associa-
tion of the different elements of PVPH-37 into a
single individual or even a single species may be
questionable (Chiappe et al., this volume). As
emphasised by Novas (this volume; in press; see
also Chiappe et al., this volume), several charac-
ters in the synsacrum, pelvic girdle and thoracic
limb indicate a close relationship between P.
puertai, the Patagonian Alvarezsaurus calvoi
(Bonaparte, 1991; see below) and Mononykus
olecranus from the Late Cretaceous of central
Asia (Perle et al., 1993, 1994). P. puertai had the
bizarre morphology of the thoracic limb of M.
olecranus (Perle et al., 1993, 1994), but this bird
was significantly larger than its closest relatives,
being about the size of the non-avian theropod
Deinonychus antirrhopus (Novas & Coria,
1990).

The Río Neuquén Formation is composed of
fluvial deposits (Cazau & Uliana, 1973). In addi-
tion to Patagonykus puertai, this formation has
provided remains of fishes, amphibians, turtles
and titanosaurid dinosaurs (Novas & Coria, 1990;
Bonaparte, 1992; Novas, pers. comm.).

FOSSIL RECORD OF BIRDS IN THE MESOZOIC GONDWANA

Alvarezsaurus calvoi Bonaparte, 1991

MATERIAL EXAMINED. HOLOTYPE: MUCPv-
54, articulated specimen including several cervical and
dorsal vertebral remains, three synsacral and 13 caudal
vertebrae, a scapula and part of a coracoid, portions of
both ilia, femora, tibiae and proximal tarsals, and
metatarsals and pedal phalanges (Bonaparte, 1991).

LOCALITY AND HORIZON. City of Neuquén,
Province of Neuquén, Argentina. Rio Colorado Forma-
tion, Bajo de la Carpa Member, Late Cretaceous (Con-
iacian-Santonian; Chiappe & Calvo, 1994).

REMARKS. Novas (this volume) included
Patagonykus puertai and Mononykus olecranus
along with Alvarezaurus calvoi in the Al-
varezauridae, a taxon recognised by Bonaparte
(1991) to include the latter species. Bonaparte
(1991) remarked on the differences of alvarez-
saurids to all theropods then known. In fact, this
distinction was emphasised by his creation of the
Alvarezsauria, a taxon that included both AI-
varezsauridae and A, calvoi. Alvarezsauria, how-
ever, should be disregarded in that it only adds
redundancy to the taxonomic system. Chiappe et
al. (this volume) have presented evidence for the
avian affinity of A. calvoi, which along with all
alvarezsaurids are the sister-group of all birds
other than Archaeopteryx lithographica.

The fluvial sandstones of the Bajo de la Carpa
Member in the city of Neuquén have provided
abundant remains of continental tetrapods such
as the flightless bird Patagopteryx deferrariisi
(Alvarenga & Bonaparte, 1992; see below), the
enantiornithine bird Neuquenomis volans
(Chiappe & Calvo, 1994; see below), noto-
suchian and baurusuchid crocodiles, dynilisid
ophidians, small non-avian theropods, and
titanosaurid sauropods (Bonaparte, 1991;
Chiappe & Calvo, 1994). The beds of the Bajo de
la Carpa Member have been regarded as high
energy channels, deposited in a paleoenviron-
ment of braided streams (Cazau & Uliana, 1973).
In the area of the city of Neuquen, however, the
characteristics of the fossil fauna and its preser-
vation (i.e., small, articulated tetrapods, absent in
other localities) suggest a lower energy deposi-
tional environment than the one envisioned pre-
viously (Chiappe & Calvo, 1994).

ORNITHOTHORACES Chiappe, 1995

MATERIAL EXAMINED. MACN-RN-976 (Fig. 5),
an isolated cervical vertebra.

LOCALITY AND HORIZON. Estancia Los Alamitos,
Cerro Cuadrado, Province of Río Negro, Argentina

541

(Bonaparte et al., 1984; Bonaparte, 1986. Los Alamitos
Formation, Late Cretaceous (Campanian-Maastrich-
tian; Bonaparte, 1987).

REMARKS. This is the only avian specimen so
far recorded from the rich fossil fauna of the Los
Alamitos Formation (Bonaparte et al., 1984;
Bonaparte, 1987, 1990). The presence of a well-
developed ventral process (Figs 5F, G) indicates
that MACN-RN-976 was a caudal element within
the cervical series. MACN-RN-976 has a fully
heterocoelic centrum (Figs 5F, G) and in this
respect differs from both Patagopteryx defer-
rariisi and Enantiornithes in which the posterior
cervicals have an incipient degree of heterocoely
(Chiappe, 1992a, 1996). The presence of
heterocoelic cervical vertebrae (including an in-
cipient degree of development) diagnoses the
clade formed by the common ancestor of Enan-
tiornithes and modem birds plus all its descen-
dants (unnamed node; see Chiappe, 1992a,
1995b, 1996; Chiappe & Calvo, 1994). The frag-
mentary nature of MACN-RN-976 makes it dif-
ficult to be more specific in its allocation.

The mudstone to sandstone deposits of Los
Alamitos Formation at Cerro Cuadrado have
yielded abundant remains of invertebrates, fishes,
frogs, snakes, turtles, mammals and non-avian
dinosaurs (Bonaparte et al., 1984; Bonaparte,
1987, 1990). Paleoenvironmentally, this forma-
tion has been interpreted as a shallow, permanent
brackish body of water (Andreis, 1987;
Bonaparte, 1990).

Patagopteryx deferrariisi
Alvarenga & Bonaparte, 1992

MATERIAL EXAMINED. HOLOTYPE: MACN-N-
03, a partially complete specimen including several
cervical, thoracic, synsacral and caudal vertebrae, por-
tions of the wing and shoulder, part of the ilium and
hind limb elements (Alvarenga & Bonaparte, 1992;
Chiappe, 1992). REFERRED SPECIMENS: MACN-
N-11 (Fig. 6), an almost complete skeleton; MUCPv-
48 (Fig. 6), a complete hind limb associated with
several fragmentary bones; MACN-N-10, MACN-N-
14 and MUCP-207, fragmentary specimens (Chiappe,
19922).

LOCALITY AND HORIZON. City of Neuquén,
Province of Neuquén, Argentina. Río Colorado Forma-
tion, Bajo de la Carpa Member, Late Cretaceous (Con-
iacian-Santonian; Chiappe & Calvo, 1994).

REMARKS. Patagopteryx deferrariisi is a hen-
sized, flightless bird known from the same
deposits as Alvarezsaurus calvoi (see above). Al-
varenga & Bonaparte (1992) and Alvarenga

542 MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 5. Ornithurine specimens from the Maastrichtian Allen Formation at Salitral Moreno (A-E), and the
Campanian-Maastrichtian Los Alamitos Formation at Estancia Los Alamitos (F, G), Province of Rio Negro,
Argentina. A, B, proximal half of left tibiotarsus (PVL-4730) in cranial (A) and lateral (B) views; C, distal end
of the right humerus (PVL-4731) in cranial view (the radius is still articulated); D, proximal end of right ulna
(PVL-4731) in ventral view; E, distal and proximal ends of right ulna and carpometacarpus (PVL-4731) in
caudal and dorsal views, respectively. F, G, heterocoelic vertebra (MACN-RN-976) in cranial (F) and left lateral
(G) views. dc, dorsal condyle of humerus; ep, extensor process of carpometacarpus; r, radius; u, ulna; vc, ventral
condyle of humerus. Scales = 1cm.

FOSSIL RECORD OF BIRDS IN THE MESOZOIC GONDWANA 543

FIG. 6. Patagopteryx deferrariisi from the Coniacian-Santonian Rio Colorado Formation at the city of Neuquén,
Province of Neuquén, Argentina. A, right lateral view of the pelvis (MACN-N-11); B, ventral view of the right
humerus, ulna and radius (MACN-N-11); C, right lateral view of the skull (MACN-N-11); D, cranial view of
the left femur (MUCPv-48); E, cranial view of the left tibiotarsus and fibula (MUCPv-48); F, medial view of
the left tarsometatarsus and phalanges MUCPv-48. Scale = lcm.

544

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 7. Transverse sections of the enanthiornathine femora of PVL-4273 (A) and MACN-S-01 (B) from the
Maastrichtian Lecho Formation at Estancia El Brete, Province of Salta, Argentina. Arrows show different lines
of arrested growth (LAGs) in the compacta of these bones. Scales = 100jum.

(1993) related this species to the ratites (ostriches
and its allies). A more parsimonious interpreta-
tion, however, supports the sister-group relation-
ship between P. deferrariisi and Ornithurae
(Chiappe, 1992a, 1995a, b; Chiappe & Calvo,
1994; Chiappe et al., this volume) (Fig. 2). Fur-
thermore, the hypothesis of ratite affinities has
been seriously challenged by recent histological
studies documenting the presence of growth rings
or lines of arrested growth (LAGs) in the compacta
of the femora of both P. deferrariisi and the
Enantiornithes (Chinsamy et al., 1994, 1995).
This pattern of bone microstructure is absent in
ornithurine birds, including ratites. The presence
of cyclical pauses in bone deposition indicates
that P. deferrariisi, and the Enantiornithes, had
a rate of growth and associated physiology
different from that of any modern bird (Chinsamy
etal., 1994, 1995). In consequence, compelling
osteological and histological evidence supports
the hypothesis that P. deferrariisi acquired flight-
lessness independently from ratites and any other
known flightless avian group (Chiappe, 19954, b,
1996).

Curiously, Nessov (1993) recently related the
Coniacian Kuszholia mengi (Nessov, 1992)
from the Bissekty Formation at Dzhyrakuduk
(Uzbekistan) to Patagopteryx deferrariisi.
Nessov (1992) briefly described and illustrated
two fragmentary synsacra, which appear to be all
the available material, and concluded that K.
mengi could have been either a bird or a non-
avian theropod. Comparisons based on such
limited evidence are very hard to make. The
synsacrum of K. mengi, however, differs from that

of P. deferrariisi in the presence of pleurocoels
and a concave caudal articular surface.

ENANTIORNITHES Walker, 1981

MATERIAL EXAMINED. Large assortment of
postcranial elements and a jaw housed at the
Fundación-Instituto Miguel Lillo (Tucumán) and the
Museo Argentino de Ciencias Naturales (Buenos
Aires) (Walker, 1981; Chiappe, 1993, 1996).

LOCALITY AND HORIZON. Estancia El Brete,
Department of Candelaria, Province of Salta, Argen-
tina. Lecho Formation, Late Cretaceous (Maastrich-
tian; Bonaparte et al, 1977; Gómez Omil et al.,
1989).

REMARKS. The Maastrichtian deposits of the
Lecho Formation (Gómez Omil et al., 1989) have
provided the most diverse and abundant collec-
tion of Gondwanan Cretaceous avians
(Bonaparte et al, 1977; Bonaparte & Powell,
1980; Walker, 1981; Chiappe, 1991, 1993). Most
of these avian remains were preliminarily studied
by Walker (1981), who recognised them as part
of a basal monophyletic group, the Enantior-
nithes. The phylogenetic relationships of the
Enantiornithines have been a matter of debate
(Chiappe, 1995a, b), although most authors
regard them as basal birds. Recent cladistic
analyses (Chiappe, 1995a, b, 1996; Chiappe &
Calvo, 1994; Chiappe et al., this volume) support
the hypothesis that the Enantiornithes is the
sister-group of the clade formed by Patagopteryx
deferrariisi and Ornithurae (Fig. 2). The number
of species represented in the enantiornithine as-
semblage of El Brete is not yet known. It is likely,

FOSSIL RECORD OF BIRDS IN THE MESOZOIC GONDWANA

however, that more than the A
four named species (see below)
are present in this site. The fact
that most bones are preserved
disarticulated prevents a more
precise evaluation of the actual
specific diversity.

As mentioned in the discus-
sion of Patagopteryx defer-
rariisi, recent histological
studies of enantiornithine
femora from this collection
have documented the presence
of LAGs in the compacta of
these bones (Chinsamy et al.,
1994, 1995) (Fig. 7). As
pointed out by Chinsamy et al.
(1994, 1995), cyclical bone
deposition in Enantiornithes
and other basal birds suggests
important physiological dif-
ferences with respect to their
modem relatives.

The avian assemblage from
El Brete (including the four
species discussed below) is in
association with remains of
sauropod and non-avian
theropod dinosaurs (Bonaparte
et al., 1977; Bonaparte &
Powell, 1980). The fine-
grained sandstone entombing this assemblage has
been regarded as deposited in a fluvial-lacustrine
coastal plain, with abundant vegetation and
ponds (Bonaparte et al., 1977).

Enantiornis leali Walker, 1981

MATERIAL EXAMINED. PVL-4035 (Fig. 8), a
specimen including the proximal half of the humerus,
the proximal portion of the scapula, and a coracoid
(Walker, 1981). REFERRED SPECIMENS: PVL-
4020, several thoracic limb and girdle elements; PVL-
4039, PVL-4055, two isolated scapulae; PVL-4023,
PVL-4181, two isolated ulnae.

LOCALITY AND HORIZON. Estancia El Brete,
Department of Candelaria, Province of Salta, Argen-
tina. Lecho Formation, Late Cretaceous (Maastrich-
tian; Bonaparte et al., 1977; Gómez Omil et al.,
1989).

REMARKS. Unfortunately, Walker (1981) never
described this species but just listed its name in
the figure's caption. Below is a brief description
of the holotype specimen. A more detailed
description of all the available material would not

545

FIG. 8. Enantiomis leali (PVL-4035) from the Maastrichtian Lecho Forma-
tion at Estancia El Brete, Province of Salta, Argentina. A, proximal half of
left humerus in cranial view; B, left coracoid in dorsal view; C, left scapula
in medial view. ac=acromion, bc-bicipital crest, fn=foramen for
supracoracoid nerve, fs-dorsal fossa of coracoid. Scale = lcm.

be appropriate for this paper, but it will be
provided elsewhere (Chiappe & Walker, in
preparation).

The humerus of Enantiornis leali is very robust
(Fig. 8A). The cranial side of the head is deeply
concave. Distal to the head there is a deep, sub-
circular fossa. As in other enantiornithines, the
bicipital crest is inflated and projected
cranioventrally (Fig. 8A). The ventral tubercle is
projected caudally and, most peculiarly, it is
proximodistally perforated by a foramen. The
pneumotricipital foramen is not developed. The
coracoid is elongated (Fig. 8B). Its shoulder end
is lateromedially compressed and it slopes dor-
sally. Medially, between the articular facet for
the humerus and the acrocoracoid process, there
is a subcircular tubercle. As typical of all enan-
tiornithines, the dorsal surface of the coracoid is
deeply excavated by a triangular fossa (Fig. 8B).
Slightly above the vertex of this fossa is the dorsal
opening of the foramen for the supracoracoid
nerve. The position of this foramen differs from
that in some other enantiomithines, where it
opens inside the fossa. The scapula has a well-
developed acromion (Fig. 8C). Its most charac-

546

teristic feature is the presence of a circular pit,
medially situated between the acromion and the
articular facet for the humerus.

Enantiornis leali was a fairly large bird, ranging
between the size of a skua (Catharacta skua) and
a turkey vulture (Cathartes aura). This species is
represented by elements of the thoracic limb and
girdle only, while all other named enantiornithine
species from El Brete are known from elements
of the pelvic limb (Chiappe, 1993). Thus, there is
the possibility that E. leali might be a synonym
of one of the three other enantiomithine species
from E] Brete.

Lectavis bretincola Chiappe, 1993

MATERIAL EXAMINED. HOLOTYPE: PVL-4021-1
(Fig. 9A, B), a tibiotarsus and tarsometatarsus missing
the distal portion (Chiappe, 1993).

LOCALITY AND HORIZON. Estancia El Brete,
Department of Candelaria, Province of Salta, Argen-
tina. Lecho Formation, Late Cretaceous (Maastrich-
tian; Bonaparte et al., 1977; Gómez Omil et al.,
1989).

REMARKS. The pedal morphology of Lectavis
bretincola together with the inflation of the
medial condyle of the tibiotarsus (Chiappe, 1993;
Chiappe & Calvo, 1994) supports the allocation
of this taxon in the Enantiornithes. The
phylogenetic relationships of L. bretincola to
other enantiornithines are not clearly understood,
although a recent cladistic analysis indicates that
this taxon lies outside the Avisauridae a group of
small to medium size arboreal enantiornithines
(Chiappe, 1993; see below). The very elongate
tibiotarsus and tarsometatarsus of L. bretincola
(Fig. 9A, B) suggest that it had wading habits
(Walker, 1981).

Yungavolucris brevipedalis Chiappe, 1993

MATERIAL EXAMINED. HOLOTYPE: PVL-4053
(Fig. 9E, F), a nearly complete tarsometatarsus.
REFERRED SPECIMENS: Several tarsometatarsi
(PVL-4040, PVL-4052, PVL-4268, PVL-4692)
(Chiappe, 1993).

LOCALITY AND HORIZON. Estancia El Brete,
Department of Candelaria, Province of Salta, Argen-
tina. Lecho Formation, Late Cretaceous (Maastrich-
tian; Bonaparte et al., 1977; Gómez Omil et al.,
1989).

REMARKS. The presence of a metatarsal IV far
smaller than metatarsals II and III, a tubercle on
the dorsal face of metatarsal II, and a trochlea of
metatarsal II broader than that of metatarsals II

MEMOIRS OF THE QUEENSLAND MUSEUM

and IV (Fig. 9E, F), identify this taxon as an
enantiornithine (Chiappe, 1993). The
phylogenetic relationships of Yungavolucris
brevipedalis to other enantiomithines are not
resolved, however. Like Lectavis bretincola, a
cladistic analysis within the Enantiornithes indi-
cates that it is not an avisaurid enantiornithine
(Chiappe, 1993). The remarkable asymmetry of
the tarsometatarsus of Y. brevipedalis (Fig. 9E, F)
suggests that this species was probably aquatic
(Walker, 1981).

AVISAURIDAE Brett-Surman & Paul, 1985

Soroavisaurus australis Chiappe, 1993

MATERIAL EXAMINED. HOLOTYPE: PVL-4690
(Fig. 9C, D), a tarsometatarsus (Chiappe, 1993).
REFERRED SPECIMENS: PVL-4048, a tarso-
metatarsus articulated to several phalanges (Chiappe,
1993).

LOCALITY AND HORIZON. Estancia El Brete,
Department of Candelaria, Province of Salta, Argen-
tina. Lecho Formation, Late Cretaceous (Maastrich-
tian; Bonaparte et al, 1977; Gómez Omil et al.,
1989).

REMARKS. Soroavisaurus australis can be
regarded as an enantiornithine on the basis of the
pedal characteristics mentioned for Yungavolucris
brevipedalis, in addition to derived features
shared with Neuquenornis volans (e.g., strong
plantar projection of the medial rim of the
trochlea of metatarsal III, transverse convexity of
dorsal surface of the mid-shaft of metatarsal III;
see Chiappe, 1993; Chiappe & Calvo, 1994). The
avian affinity of S. australis was questioned by
Brett-Surman & Paul (1985), who regarded it as
non-avian dinosaur, together with Avisaurus
archibaldi from the Late Cretaceous Hell Creek
Formation of North America. The avian affinity
of the Avisauridae, however, has been definitive-
ly established with the discovery of fairly com-
plete specimens showing undisputable
enantiornithine features (Chiappe, 1992b, 1993;
Huchinson, 1993; Chiappe & Calvo, 1994).

A recent cladistic analysis (Chiappe, 1993; Var-
ricchio & Chiappe, 1995) indicates that
Soroavisaurus australis is the sister-group of the
North American avisaurids (Avisaurus archibaldi
and A. gloriae). This study demonstrates the non-
monophyletic status of the El Brete enantior-
nithine assemblage, which is composed of taxa
sharing a most recent common ancestor with
North American taxa. The pedal morphology of

FOSSIL RECORD OF BIRDS IN THE MESOZOIC GONDWANA

547

IV

Hl Il IV

FIG. 9. Enantiornithine tarsometatarsi (dorsal and plantar views) from the Maastrichtian Lecho Formation at
Estancia El Brete, Province of Salta, Argentina (after Chiappe, 1993). A, B, Lectavis bretincola (PVL-4021-1);
C, D, Soroavisaurus australis (PVL-4690); E, F, Yungavolucris brevipedalis (PVL-4053). dp=dorsomedial
projection of the distal end of metatarsal II, fe=fenestra between metatarsals III and IV, rm=elongate proximal
ridge for muscle attachment, tu=tubercle for the attachment of the M. tibialis cranialis, II-IV, metatarsals II-IV.

Scale = lcm.

S. australis indicates that this species was capable
of perching.

Neuquenornis volans
Chiappe & Calvo, 1994

MATERIAL EXAMINED. HOLOTYPE: MUCPv-
142 (Fig. 10), fairly complete skeleton preserving
the caudal portion of the skull, elements of the wing

and shoulder, portions of the hind limb and a few
thoracic vertebrae (Chiappe & Calvo, 1994). REFERRED
SPECIMEN: MACN-RN-977, distal end of humerus.

LOCALITY AND HORIZON. MUCPv-142 comes
from the city of Neuquén, Province of Neuquén. MACN-
RN-977 was found at Puesto Tripailao, approximately
30km southwest of General Roca, Province of Río
Negro, Argentina. Río Colorado Formation, Bajo de la

548

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 10. Neuquenornis volans (MUCPv-142) from the Coniacian-Santonian Río Colorado Formation at the city
of Neuquén, Province of Neuquén, Argentina (after Chiappe & Calvo, 1994). FUR-furcula, LCO=left coracoid,
LHU-left humerus, LSC=left scapula, LTI-left tibiotarsus, LTM=left tarsometatarsus, PEL?=pelvis,
RCM-right carpometacarpus, RCO=right coracoid, RFE=right femur, RHU-right humerus, RTI=right tibiotar-

sus, RTM=right tarsometatarsus.

Carpa Member, Late Cretaceous (Coniacian-San-
tonian; Chiappe & Calvo, 1994).

REMARKS. Neuquenornis volans, was a falcon-
sized, active flyer whose enantiomithine affinity
is supported by numerous synapomorphies in the
thoracic limb (e.g., distal end of humerus cranio-
caudally compressed) and girdle (e.g., coracoid
with a convex lateral margin anda wide triangular
fossa on its dorsal surface), pelvic limb (e.g.,

well-developed posterior trochanter of femur)
and vertebral column (e.g., lateral grooves on
thoracic centra) [see Chiappe & Calvo (1994) for
a complete list of characters] (Fig. 10). The pedal
anatomy indicates that N. volans is a member of
the Avisauridae (Chiappe, 1993) and like
Soroavisaurus australis, was capable of perching
(Chiappe & Calvo, 1994). Specimen MACN-
RN-977, although fragmentary, is identical in

FOSSIL RECORD OF BIRDS IN THE MESOZOIC GONDWANA

size and morphology to the holotype of Neu-
quenornis volans.

ORNITHURAE Haeckel, 1866

MATERIAL EXAMINED. PVL-4730, a proximal
half of a tibiotarsus (Powell, 1987), PVL-4731, por-
tions of humerus, ulna, radius, carpometacarpus, and
the proximal phalanx of the major digit (Fig. 5).

LOCALITY AND HORIZON, Salitral Moreno, 20km
south of General Roca, Province of Rio Negro, Argen-
tina. Allen Formation, Late Cretaceous (Maastrichtian;
Powell, 1987).

REMARKS. These two specimens were found in
the same quarry (Powell, pers. comm.). PVL-
4730 exhibits two well-developed cnemial crests,
typical of ornithurine birds (Figs 5A, B). PVL-
4731 is significantly larger than PVL-4730 and,
although these specimens have not been studied
in detail, it probably represents a different species
(Fig. 5C-E). The presence of a well-projected,
pointed metacarpal process of the carpometacar-
pus (Fig. 5E) and a cranio-caudally expanded
proximal phalanx of the major digit both support
the allocation of PVL-4731 to the Ornithurae.
The fossiliferous beds of the Allen Formation
at Salitral Moreno have provided a variety of
dinosaurian remains including hadrosaurids
(Powell, 1987), titanosaurid sauropods (Salgado
& Coria, 1993), non-avian theropods (Coria, pers.
comm.) and turtles (Salgado & Coria, 1993).

NEORNITHES Gadow, 1893
GAVIIDAE Allen, 1897

Neogaeornis wetzeli Lambrecht, 1929

MATERIAL EXAMINED. HOLOTYPE: GPMK-
123, a right tarsometatarsus (Lambrecht, 1929).

LOCALITY AND HORIZON. West end of Bahía San
Vicente, Province of Concepción, Chile. Quiriquina
Formation, Late Cretaceous (Campanian-Maastrich-
tian; Biró Bagóczky, 1982).

REMARKS. Lambrecht (1929) considered
Neogaeornis wetzeli as a hesperornithiform, a
group of Cretaceous marine foot-propelled divers
mostly known from North American beds (Olson,
1985; Chiappe, 1995a). Brodkorb (1963b)
remarked on the affinity of this species to the
hesperornithiform Baptornis advenus, including
both in the Baptornithidae, in which he was fol-
lowed by Martin & Tate (1976). The hesperor-
nithiform relationship of N. wetzeli, however,
was recently challenged by Olson (1992). After

549

repreparing the specimen, Olson (1992) regarded
itas a loon and placed it within the foot-propelled,
modern Gaviidae.

The Quiriquina Formation has afforded a great
variety of invertebrates, fishes and reptiles (Biró-
Bagóczky, 1982). The depositional environment
of the Quiriquina Formation appears to have been
a tidal plain (Biró-Bagóczky, 1982).

ANTARCTICA

AVES Linnaeus, 1758
METORNITHES Perle et al., 1993
ORNITHOTHORACES Chiappe, 1995
ORNITHURAE Haeckel, 1866
NEORNITHES Gadow, 1893
GAVIIDAE Allen, 1897

MATERIAL EXAMINED. A single specimen
preserving parts of the skull, limbs, and a few cervical
vertebrae (Chatterjee, 1989).

LOCALITY AND HORIZON. Marambio (Seymour)
Island, Antarctic Peninsula. López de Bertodano For-
mation, Late Cretaceous (Campanian-Maastrichtian;
Medina et al., 1989).

REMARKS. The first report of an Antarctic
Cretaceous bird was given by Chatterjee (1989),
who reported a loon-like bird from the Cam-
panian-Maastrichtian López de Bertodano For-
mation of Marambio (Seymour) Island.
Chatterjee's (1989) allocation of this fossil within
the modern gaviiforms was corroborated by
Olson (1992). This specimen represents the most
southern ocurrence of gaviiforms and along with
Neogaeomis wetzeli (Lambrecht, 1929; see
above) the oldest known loon.

The López de Bertodano Formation is primari-
ly composed of marine medium-grained to silty
sandstones (Zinsmeister, 1982; Medina et al.,
1989), enclosing a diverse assemblage of micror-
ganisms (Martinez Machiavello, 1987), plants
(Askin, 1990) and invertebrates (Zinsmeister,
1982), along with fishes (Grande & Chatterjee,
1987), plesiosaurs (Gasparini et al., 1984) and
mosasaurs (Gasparini & del Valle, 1984) among
vertebrates. Medina et al. (1989) have sum-
marised the paleoenvironment of this unit as
ranging from an offshore marine shelf to near
shore facies, including intertidal plains and del-
taic environments.

PRESBYORNITHIDAE Wetmore, 1926

MATERIAL EXAMINED. MLP 93-1-3-1 and MLP
93-1-3-2, partial skeletons, apparently conspecific
(Noriega & Tambussi, 1995).

550

LOCALITY AND HORIZON. Cape Lamb, Vega Is-
land, Antarctic Peninsula, Cape Lamb strata, correlated
to the López de Bertodano Formation (Unit B), Late
Cretaceous (Campanian-Maastrichtian; Medina et al.,
1989).

REMARKS. MLP 93-1-3-1 was briefly
described by Noriega & Tambussi (1995). These
authors regarded it as a member of the
anseriforms Presbyornithidae (Ericson, 1992), an
extinct group of Paleogene wading birds (Olson,
1985) previously represented in the southern con-
tinents by the Eocene Telmabates from Patagonia
(Howard, 1955). If this interpretation is correct,
these new findings expand the record of pres-
byornithids to the Cretaceous.

The marine deposits from where these fossils
were recovered belong to the Maastrichtian sec-
tion of the Late Cretaceous Cape Lamb strata
(Zinsmeister, pers. comm.). These beds have
been correlated to those of the López de Ber-
todano Formation at Marambio (Seymour) Is-
land, although regarded as more marginal
deposits within the same basin (Pirrie et al.,
1991). The associated fauna primarily includes
amonites, gastropods, fishes and plesiosaurs
(Zinsmeister, pers. comm.). The fact that these
birds were found in marine deposits does not
necessarily mean that they were oceanic birds,
although it suggests that at least they inhabited
marine littoral environments.

AFRICA

AVES Linnaeus, 1758

MATERIAL EXAMINED. Several footprints (Figs
4B, C, D) of unknown repository (Lockley etal., 1992).

LOCALITY AND HORIZON. 16km east of Agadir,
Morocco (Ambroggi & Lapparent, 1954). Calcareous
Maastrichtian beds (no formational name is given),
Late Cretaceous.

REMARKS. Ambroggi & Lapparent (1954)
regarded as avian a series of small tridactyl
footprints (Figs 4B, C, D), about 3cm long, as-
sociated with other footprints of non-avian
dinosaurs and lizards. The bird tracks were brief-
ly described and poorly illustrated. One of them
(Fig. 4D) was interpreted as having been made by
a semi-palmate bird.

METORNITHES Perle et al., 1993

MATERIAL EXAMINED. Hind limb elements of two
con-specific specimens (Krause, pers. comm.).

MEMOIRS OF THE QUEENSLAND MUSEUM

LOCALITY AND HORIZON. Near the village of
Berivotra, Mahajanga Basin, northwestern Madagas-
car. Maevarano Formation, Late Cretaceous (Cam-
panian; Krause et al., 1994).

REMARKS. The specimens belong to a new,
undescribed taxon. Retention of several
plesiomorphic characters (e.g., incompletely
fused tibiotarsus and tarsometatarsus) suggest
that this is a very primitive bird, although its
phylogenetic placement remains to be analysed.
This finding is extremely important because these
specimens not only represent the first Mesozoic
African bird known by skeletal evidence but also
the first pre-Late Pleistocene avian from
Madagascar.

The fossiliferous section of the Maevarano For-
mation consists of fluvial white sandstones. The
poorly known Late Cretaceous fossil record of
Madagascar has become substantially improved
with the recent discovery of an ensemble of ver-
tebrates from the Maevarano Formation that, in
addition to the above mentioned birds, includes
fishes, frogs, turtles, lizards, snakes, crocodiles,
dinosaurs and mammals (Krause et al., 1994, in
press).

DISCUSSION

In spite of a recent plethora of new findings, the
Gondwanan record of Mesozoic birds is still very
limited and restricted to the Cretaceous. This
record is most diverse and extensive in South
America, where birds are known from the Aptian
to the Maastrichtian and from many localities.
Within Australasia it is limited to the Early
Cretaceous (Aptian-Albian) of Australia and in
Antartica it is restricted to the latest Cretaceous
(Maastrichtian). Likewise, in Africa only a few
footprints and a recently discovered new taxon
from the Late Cretaceous of Morocco
(Maastrichtian) and Madagascar (Campanian),
respectively, stand as evidence of presence of
Mesozoic birds in this continent.

Although the record is limited, the Mesozoic
birds from Gondwana have contributed sig-
nificantly to the growth of knowledge on early
avian evolution. Taxa such as the Enantiornithes
and Patagopteryx deferrariisi were among the
firstto provide evidence about the series of trans-
formations which occurred between the primitive
morphology of Archaeopteryx lithographica and
that of more derived birds such as Hesperor-
nithiformes and Ichthyornithiformes (Walker,
1981; Chiappe, 19922, 1995a, b, 1996; Chiappe
& Calvo, 1994). More recently, the histological

FOSSIL RECORD OF BIRDS IN THE MESOZOIC GONDWANA

studies carried out on the bones of Enantiomithes
and P. deferrariisi led to the first inferences on the
rate of growth and related physiology of basal
birds derived from reliable evidence (Chinsamy
et al., 1994, 1995).

Using the Gondwanan record of Mesozoic
birds for large scale paleobiogeographic analysis
(e.g., Bonaparte, 1986, 1991; Chiappe, 1991;
Noriega & Tambussi, 1995) is considered unreli-
able on the basis of the still insufficient evidence.
For example, a Gondwanan origin of the Enan-
tiornithes was claimed by Bonaparte (1986,
1991) on the basis of assumptions such as their
‘oldest’ record and their ‘high’ taxonomic diversity
in South America [see Humphries & Parenti
(1986) for a discussion of these theoretical
issues]. This idea, however, has been shown to be
an erroneous generalization based on an inade-
quate record. At present an abundant diversity of
enantiornithine birds is known worldwide during
the Cretaceous, and their oldest record to date is
not in Gondwanan continents but in Europe and
Asia (Chiappe & Calvo, 1994; Chiappe, 1995a;
Sanz et al. 1995).

A general look at the Mesozoic record of birds
from Gondwana reflects a broad, although
punctuated, spectrum of geographic distribu-
tions, habitats and modes of life. During the
Cretaceous, Gondwanan birds occupied a large
latitudinal range, from near the equator to close
to the Southern pole. These fossils have been
recovered primarily from inland environments,
although birds probably inhabited marine realms
and seashores as well. The paleoenvironmental
conditions of the localities in which Mesozoic
Gondwanan birds were found range from warm
and arid to near-freezing, at least during winter.
A variety of modes of life have been inferred from
their anatomy, such as flightless cursorials to
foot-propelled divers, waders and perching active
fliers.

ACKNOWLEDGEMENTS

I am especially grateful to F. Novas (Museo
Argentino de Ciencias Naturales, Buenos Aires)
and R. Molnar (Queensland Museum, South Bris-
bane) for their invitation to participate in this
Symposium and their valuable help in various
aspects of this paper. I am grateful to J. Bonaparte
(MACN, Buenos Aires), S. Chatterjee (Texas
Tech. University, Lubbok), R. Coria (Museo Car-
men Funes, Plaza Huincul), A. Kellner
(American Museum of Natural History, New
York), R.E. Molnar, J. Noriega (Museo de La

Plata, La Plata), F Novas, S. Olson (National
Museum of Natural History, Washington), J.
Powell (Universidad Nacional de Tucumán,
Tucumán), L. Salgado (Universidad Nacional del
Comahue, Neuquén), R. Smith (South African
Museum, Cape Town) and C. Tambussi (MLP, La
Plata) who allowed me to consult material under
their care or provided casts for this study. I thank
H. Alvarenga, R. Coria, P. Ericson (Swedish
Museum of Natural History, Stockholm), M.
Lockley (University of Colorado, Denver), V.
Ovtsharenko (AMNH, New York), M. Reguero
(MLP, La Plata), L. Salgado and W. Zinsmeister
(Purdue University, West Lafayette) for provid-
ing general data (often unpublished) and discus-
sions on some aspects of the paper. F. Novas and
P. Vickers-Rich (Monash University, Clayton)
made useful reviews to the original manuscript.
P. Dandonoli (AMNH, New York) and S. Zetkus
(AMNH, New York) provided grammatical help
for the final English manuscript. F. Ippolito
(AMNH) prepared part of the illustrations. F.
Novas provided financial assistance for attending
to the Symposium. This research has been sup-
ported by a Frick Research Fellowship of the
American Museum of Natural History.

REFERENCES

ALONSO, R.N. 1989. Late Cretaceous dinosaur track-
ways in northern Argentina. Pp. 223-228. In Gil-
lette, D.D. & Lockley, M.G. (eds) 'Dinosaur tracks
and traces’. (Cambridge University Press).

ALONSO, R.N. & MARQUILLAS, R. A. 1986. Nueva
localidad con huellas de dinosaurios y primer
hallazgo de huellas de aves en la Formación
Yacoraite (Maastrichtiano) del norte Argentino.
IV Congreso Argentino de Paleontología y
Bioestratigrafia, Mendoza. Actas 2: 33-41

ALVARENGA, H.M.F. & BONAPARTE, J.F. 1992. A
new flightless land bird from the Cretaceous of
Patagonia. Proceedings 11 Intemational Sym-
posium of Avian Paleontology, Los Angeles, pp.
51-64.

AMBROGGI, R. & DE LAPPARENT, A.F. 1954. Les
empreintes de pas fossiles du Maestrichtien d'-
Agadir. Notes du Service Geologique du Maroc
10(122): 43-57.

ANDERSON, A, 199]. Early bird threatens Ar-
chaeopteryx's perch. Research News (1991): 35.

ANDREIS, R.R. 1987. Stratigraphy and Paleoenviron-
ment. In Bonaparte, J.F. (coord.) ‘The Late
Cretaceous fauna of Los Alamitos, Patagonia,
Argentina’. Rev. Mus. Arg. Ciencias Nat. “Bernar-
dino Rivadavia" 3(3): 103-110.

ASKIN, R.A. 1990. Campanian to Paleocene spore and
pollen assemblages of Seymour Island, An-

552

tarctica. Review of Palaobotany and Palinology
65: 105-113.

BAUMEL, J.J. & WITMER, L.M. 1993. Osteologia. In
Baumel, J.J., King, Breazile, A.S., Evans, J.E.,
Vanden Berge, H.E. & J.C. (eds) ‘Handbook of
Avian Anatomy: Nomina Anatomica Avium’,
second edition. Publications of the Nuffal
Omithological Club 23: 45-132.

BIRÓ-BAGÓCZK Y, L. 1982. Revision y redefinición
de los “Estratos de Quiriquina", Campaniano-
Maastrichtiano, en su localidad tipo, en la isla
Quiriquina, 36°37’ Lat. Sur, Chile, Sudamérica,
con un perfil complementario en Cocholque. III
Congreso Geológico Chileno, Actas 1: A29-A64.

BONAPARTE, J.F. 1986. History of the terrestrial
Cretaceous vertebrates of Gondwana. IV Con-
greso Argentino de Paleontología y
Bioestratigrafía, Actas 2: 63-95.

1987. The Late Cretaceous fauna of Los Alamitos,
Patagonia, Argentina. Parts I to IX. Revista del
Museo Argentino de Ciencias Naturales “Bernar-
dino Rivadavia", Paleontologia 3(3): 103-178.

1990. New Late Cretaceous mammals from the Los
Alamitos Formation, northern Patagonia. Nation-
al Geographic Research 6(1): 63-93.

1991. Los vertebrados fósiles de la Formación Río
Colorado de Neuquén y cercanias, Cretácico Su-
perior, Argentina. Revista del Museo Argentino
de Ciencias Naturales "Bernardino Rivadavia",
Paleontologia 4(3): 17-123.

1992. Una nueva especie de Triconodonta (Mam-
malia), de la Formación Los Alamitos, provincia
de Río Negro y comentarios sobre su fauna de
mamíferos. Ameghiniana 29(2): 99-110.

BONAPARTE, J.F., SALFITTY, J.A., BOSSI, G. &
POWELL, J.E. 1977. Hallazgos de dinosaurios y
aves cretácicas en la Formación Lecho de El Brete
(Salta), próximo al limite con Tucumán. Acta
Geológica Lilloana 14: 5-17.

BONAPARTE, J.F. & POWELL, J.E. 1980. A continen-
tal assemblage of tetrapods from the Upper
Cretaceous beds of El Brete, northwestern Argen-
tina (Sauropoda-Coelurosauria-Carnosauria-
Aves). Mémoirs de la Société Géologique de
France (Nouvelle Série) 139: 19-28.

BONAPARTE, J.F., FRANCHI, M.R., POWELL, J.E.,
& SEPULVEDA, E.G. 1984. La Formación Los
Alamitos (Campaniano-Maastrichtiano) del
Sudeste de Río Negro, con descripción de
Kritosaurus australis n. sp. (Hadrosauridae). Sig-
nificado paleogeográfico de los vertebrados.
Revista de la Asociación Geológica Argentina
(3-4): 284-299.

BRODKORB, P. 1963a. Birds from the Upper
Cretaceous of Wyoming. Proceedings of the XIII
Intemational Omithological Congress, pp. 55-70.

1963b. Catalogue of fossil birds. Bulletin of the
Florida State Mus., Biol. Sci. 7: 179-293.

1976. Discovery of a Cretaceous bird, apparently
ancestral to the Orders Coraciformes and Picifor-
mes (Aves: Carinatae). In Olson, S.L. (ed.) Col-

MEMOIRS OF THE QUEENSLAND MUSEUM

lected papers in avian paleontology honoring the
9Oth birthday of Alexander Wetmore. Smith-
sonian Contnbutions to Paleobiology 27: 67-73.

CASAMIQUELA, R. 1987. Novedades en icnologia de
vertebrados en la Argentina. X Congreso
Brasileiro de Paleontologia, Rio de Janeiro, Anais
1: 445-456.

CAZAU, L.B. & ULIANA, M.A. 1973. El Cretácico
superior continental de la Cuenca Neuquina. V
Congreso Geológico Argentino, Actas 3: 131-163.

CHATTERJEE, S. 1987. Skull of Protoavis and early
evolution of birds. Joumal of Vertebrate Paleon-
tology 7(3) (Supp.): 14A.

1989. The oldest Antarctic bird. Joumal of Ver-
tebrate Paleontology 9(3) (Supp.): 16A.

1991. Cranial anatomy and relationships of a new
Triassic bird from Texas. Philosophical Transac-
tions of the Royal Society (Biological Sciences)
332 (1265): 271-346.

CHIAPPE, L.M. 1991. Cretaceous birds of Latin
America. Cretaceous Research 12: 55-63.

1992a. Osteología y sistemática de Patagopteryx
deferrariisi Alvarenga y Bonaparte (Aves) del
Cretácico de Patagonia. Filogenia e historia
biogeográfica de las aves Cretácicas de América
del Sur. Tesis Doctoral, Facultad de Ciencias
Exactas y Naturales, Universidad de Buenos
Aires.

1992b. Enantiornithine tarsometatarsi and the avian
affinity of the Late Cretaceous Avisauridae.
Se of Vertebrate Paleontology 12 (3): 344-

50.

1993, Enantiornithine (Aves) tarsometatarsi from
the Cretaceous Lecho Formation of northwestern
Argentina. American Museum Novitates 3083:
1-27.

1995a, The first 85 million years of avian evolution.
Nature 378: 349-355.

1995b. The phylogenetic position of the Cretaceous
birds of Argentina: Enantiornithes and Patagop-
teryx deferrariisi. Courier Forschungsinstitut
Senckenberg 181: 55-63.

1996. Late Cretaceous birds of southern South
America: anatomy and systematics of Patagop-
teryx and Enantiornithes. In Arratia, G. (ed.) Con-
tnbutions of Southern South America to
Vertebrate Paleontology. Münchner Geowis-
senschaftliche Abhandlungen (A)30:203-244.
(Verlag Dr Pfeil, München).

CHIAPPE, L.M. & CALVO, J.O. 1994. Neuguenornis
volans, a new Enantiornithes (Aves) from the
Upper Cretaceous of Patagonia (Argentina). Jour-
nal of Vertebrate Paleontology 14(2): 23-246.

CHIAPPE, L.M., NORELL, M.A., & CLARK, J.M.
1996, this volume. Phylogenetic position of
Mononykus from the Late Cretaceous of the Gobi
desert. Memoirs of the Queensland Museum
39(3);: 557-582.

CHINSAMY, A., CHIAPPE, L.M., & DODSON, P.
1994. Growth rings in Mesozoic birds. Nature
368: 196-197.

FOSSIL RECORD OF BIRDS IN THE MESOZOIC GONDWANA

1995. Mesozoic avian bone microstructure:

physiological implications. Paleobiology 21 (4).

CRACRAFT, J. 1986. The origin and early diversifica-
tion of birds. Paleobiology 12(4): 383-399.

DETTMANN, M.E., MOLNAR, R.E, DOUGLAS,
J.G., BURGER, D., FIELDING, C., CLIFFORD,
H.T., FRANCIS, J., JELL, J., RICH, T., WADE,
M., VICKERS-RICH, P., PLEDGE, N., KEMP,
A. & ROZEFELDS, A. 1992. Australian
Cretaceous terrestrial faunas and floras:
biostratigraphic and biogeographic implications.
Cretaceous Research 13: 207-262.

DOUGLAS J.G. & WILLIAMS, G.E. 1982. Southern
polar forests: the early Cretaceous floras of Vic-
toria and their palaeoclimatic significance.
Palaeogeography, Palaeoclimatology,
Palaeoecology 39: 171-185.

ELLENBERGER, P. 1972. Contribution a la classifica-
tion des pistes de vertébrés du Trias: Les types du
Stormberg d' Afrique du Sud (I). Palaeovertebrata
(Memoire Extraordinaire): 1-104.

1974. Contribution a la classification des pistes de
vertébrés du Trias: Les types du Stormberg d'-
Afrique du Sud (IIeme Partie: Le Stormberg Su-
perieur — I Le biome de la zone b/1 ou niveau de
Moyeni: ses biocénoses). Palaeovertebrata
(Memoire Extraordinaire): 1-141.

ERICSON, P. 1992. Evolution and systematic elements
of the Paleogene family Presbyornithidae. Abs. III
Symposium Society of Avian Paleontology and
Evolution 1: 16 (Frankfurt).

FEDUCCIA, A. & Tordoff, H.B. 1979. Feathers of
Archaeopteryx: asymetric vanes indicate
aerodynamic function. Science 203: 1021-1022.

GASPARINI, Z. & DEL VALLE, R. 1984. Mosasaurios
(Reptilia, Sauria) cretácicos en el continente
Antártico. IX Congreso Argentino de Geología,
S.C. de Bariloche. Actas: 423-431. —

GASPARINI, Z., DEL VALLE, R. & GONI, R. 1984.
Un elasmosaürido (Reptilia, Plesiosauria) del
Cretácico superior de la Antártida. Contribuciones
Científicas del instituto Antártico 305: 1-24.

GAUTHIER, J. 1986. Saurischian monophyly and the
origin of birds, In Padian, K. (ed.) The origin of
birds and the evolution of flight. Memoirs of the

, California Academy of Sciences 8: 1-55.

GOMEZ OMIL, R.J., BOLL, A. & HEMANDEZ, R.M.
1989. Cuenca Cretácico-Terciaria del Noroeste
Argentino (Grupo Salta). In Chebli, G.A. & Spal-
letti, L.A. (eds) Cuencas Sedimentarias Argen-
tinas. Serie Correlaciones Geológicas 6: 43-64.

GRANDE, L. & CHATTERJEE, S. 1987. New
Cretaceous fish fossils from Seymour Island, An-
tarctic Peninsula, Paleontology 30(4): 829-837.

GRIMALDI, D.A. (ed.) 1990. Insects from the Santana
Formation, Lower Cretaceous, of Brazil. Bulletin
of the American Museum of Natural History 195:
1-191.

HOU L. & LIU Z. 1984. A new fossil bird from Lower
Cretaceous of Gansu and early evolution of birds.
Scientia Sinica (Series B) 27(12): 1296-1302.

553

HOWARD, H. 1955. A new wading bird from the
Eocene of Patagonia. Amencan Museum
Novitates 1710: 1-25.

HUTCHISON, J.H. 1993. Avisaurus: a ‘dinosaur’
grows wings. Joumal of Veffebrate Paleontology
13(3) (Supp): 43A.

HUMPHRIES, C.J. & PARENTI, L.R. 1986. Cladistic
biogeography. Oxford Monographs on Biogeog-
raphy 2: 1-98.

ISHIGAKI, S. 1985. Dinosaur footprints of the Atlas
Mountains I. Nature Study, (Osaka) 31(1): 5-7.

1986. Dinosaur footprints of the Atlas Mountains II.
Nature Study (Osaka) 32(10): 6-9.

KELLNER, A.W.A., MARTINS-NETO, R.G. &
MAISEY, J.G. 1991. Undetermined feather. Pp.
376-377. In Maisey, J.G. (ed.) ‘Santana fossils: an
illustrated atlas". (TFH Publications: Neptune,
N.J.).

KELLNER, A.W.A., MAISEY, J.G. & CAMPOS, D.A.
1994. Fossil down feather from the Lower
Cretaceous of Brazil. Palaeontology 37(3): 489-
492.

KRAUSE, D.W., HARTMAN, J.H., WELLS, N.A.,
BUCKLEY, G.A., LOCKWOOD, C.A., WALL,
C.E., WUNDERLICH, R.E., RABARISON, J.A.
& RANDRIAMIARAMANANA, L.L. 1994.
Late Cretaceous mammals. Nature 368: 298.

KRAUSE, D.W., HARTMAN, J.H. & WELLS, N.A. in
press. Late Cretaceous vertebrates from Madagas-
car: implications for biotic change in deep time. In
Patterson, B. & Goodman, S. (eds) ‘Natural and
Human-Induced Change in Madagascar’.
(University of Chicago Press: Chicago).

LAMBRECHT, K. 1929. Neogaeomis wetzeli n.g., n.
sp., der erste Kreidevógel der südlichen Hemis-
phaere. Palaeontologische Zeitschrift 11: 121-
128.

LEGARRETA, L. & GULISANO, C.A. 1989. Análisis
estratigráfico de la cuenca Neuquina (Triásico
superior-Terciario inferior). In Chebli, G. Spalletti,
L. (eds) Cuencas Sedimentarias Angentinas. Série
Correlación Geológica 6: 221-243.

LEONARDI, G. (ed.) 1987. Glossary and manual of
tetrapod footprint palaeoichnology. Departamen-
to Nacional da Producao Minaral (DNPM), Brasil.

LOCKLEY, M.G., YANG, S.Y., MATSUKAWA, M.,
FLEMING, F. & LIM, S.K. 1992. The track record
of Mesozoic birds: evidence and implications.
Philosophical Transactions of the Royal Society
of London B 336: 113-134.

MAISEY, J.G. (ed.) 1991. 'Santana fossils: an il-
lustrated atlas'. (TFH Publications: Neptune,
N.J.).

MARTIN, L.D. & TATE Jr, J. 1976, The skeleton of
Baptornis advenus (Aves: Hesperornithiformes).
In Olson, S.L. (ed.) Collected papers in avian
paleontology honoring the 90th birthday of
Alexander Wetmore. Smithsonian Contributions
to Paleobiology 27: 35-66.

MARTINEZ MACHIAVELLO, J.C. 1987.
Bioestratigrafía diatómica de un perfil del

554

Cretácico superior de la Isla Vicecomodoro
Marambio, Antártida. Ameghiniana 24(3-4): 277-
288.

MARTINS-NETO, M.R. & KELLNER, A.W.A. 1988.
Primeiro registro de pena na Formacao Santana
(Cretáceo inferior), Bacia do Araripe, nordeste do
Brasil. Anais da Academia Brasileira de Ciencias
60: 61-68.

MARTILL, D.M. & FILGUEIRA, J.B.M. 1994. A new
feather from the Lower Cretaceous of Brazil.
Palaeontology 37(3): 483-487.

MARTILL, D.M. & FREY, E. 1995, Colour pattering
preserved in Lower Cretaceous birds and insects:
The Crato Formation of N.E. Brazil. N. Jb. Geol.
Paláont. Mb. H. 2: 118-128.

MEDINA, FA., SCASSO, R.A., DEL VALLE, R.A.,
OLIVERO, E.B., MALAGNINO, E.C. &
RINALDI, C.A. 1989. Cuenca Mesozoica del
márgen nororiental de la Península Antártida. In
Chebli, G.A, & Spalletti, L.A. (eds) Cuencas
Sedimentanas Argentinas. Serie Correlaciones
Geológicas 6: 443-465.

MOLNAR, R.E. 1985. Alternatives to Archaeopteryx:
a survey of proposed early or ancestral birds. Pp.
209-218. In Hecht, M.K., Ostrom, J.H., Viohl, G.
& Wellnhofer, P. (eds) The beqinnings of Birds.
Proceedinqs of the Intemational Archaeopteryx
Conference Eichstatt,

1986. An enantiornithine bird from the Lower
Cretaceous of Queensland, Australia. Nature
322: 736-738.

MOLNAR, R.E. & THULBORN, R.A. 1980. First
pterosaur from Australia. Nature 288: 361-363.

MOLNAR, R.E. & KUROCHKIN, E. In press. New
material of enantiornithine birds from the
Australian Early Cretaceous. Alcheringa.

MONASTERSKY, R. 1991. The lonely bird. Science
News 140: 104-105.

MUSSI, L.F. & ALONSO, R.N. 1989. Hallazgo de una
icnita ‘aviforme’ en la Formacion Yacoraite
(Cretácico tardio), Salta. VI Jornadas Argentinas
de Paleontología de Vertebrados, San Juan.
Resümenes: 86-89.

NESSOV, L.A. 1992, Review of localities and remains
of Mesozoic and Paleogene birds of the USSR and
the description of new findings. Russian Joumal
of Ornithology 1(1): 7-50.

1993. News from the members. In Society of Avian
Paleontology and Evolution, Information Letter:
14,

NORELL, M.A., CLARK, J.M. & CHIAPPE, L.M.
1993. Naming names. Nature 366: 518.

NORIEGA, J.I. & TAMBUSSI, C.P. 1995. A Late
Cretaceous Presbyornithidae (Aves: Anserifor-
mes) from Vega Island, Antarctic Peninsula:
Paleobiogeographic implications. Ameghiniana
32(1): 57-61.

NOVAS, F.E. 1996, this volume. Alvarezsauridae,
Cretaceous maniraptorans from Patagonia and
Mongolia. Memoirs of the Queensland Museum
39(3):675-702.

MEMOIRS OF THE QUEENSLAND MUSEUM

In press. Anatomy of Patagonykus puertai
(Theropoda, Maniraptora, Alvarezsauridae) from
the Late Cretaceous of Patagonia. Journal of Ver-
tebrate Paleontology.

NOVAS, FE. & CORIA, R. 1990. Unusual Cretaceous
theropods from Patagonia. Joumal of Vertebrate
Paleontology 9(3) (Supp.): 37A.

OLSON, S.L. 1985. The fossil record of birds. In Farner,
D.S., King, J.R. & Parkes, K.C. (eds) ‘Avian
Biology’ 8: 79-238. (Academic Press: New York).

1992. Neogaeornis wetzeli Lambrecht, a Cretaceous
loon from Chile (Aves: Gaviidae). Joumal of
Vertebrate Paleontology 12(1): 122-124.

OLSON, S.L. & PARRIS, D.C. 1987. The Cretaceous
birds of New Jersey. Smithsonian Contributions
to Paleobiology 63: 1-22.

OSTROM, J.H. 1987. Protoavis, a Triassic bird? Ar-
chaeopteryx 5: 113-114.

1991. The bird in the bush. Nature 353: 212.

PADIAN, K. & OLSEN, P.E. 1984, The fossil trackway
Pteraichnus: not pterosaurian, but crocodilian.
Journal of Paleontology 58: 178-184.

PERLE, A., NORELL, M.A., CHIAPPE, L.M. &
CLARK, J.M. 1993, Flightless bird from the
Cretaceous of Mongolia. Nature 362: 623-626.

PERLE, A., CHIAPPE, L.M., BARSBOLD, R.,
CLARK, J.M. & NORELL, M.A. 1994, Skeletal
morphology of Mononykus olecranus
(Theropoda: Aves) from the Late Cretaceous of
Mongolia. American Museum Novitates 3105:
1-29,

PIRRIE, D., CRAME, J.A. & RIDING, J.B. 1991. Late
Cretaceous stratigraphy and sedimentology of
Cape Lamb, Vega Island, Antarctica. Cretaceous
Research 12: 227-258.

PONS, D., BERTHOU, P. Y. & CAMPOS, D.A. 1990.
Quelques observations sur la palynologie de 1-
Aptien Supérieur et de lAlbien du Bassin d'-
Araripe (N.E. du Brésil). I Simpósio sobre a Bacia
do Aranpe e Bacias /ntenores do Nordeste, Crato.
Atas: 241-25

POWELL, J.E. 1987. Hallazgo de un Dinosaurio
Hadrosáurido (Ornithischia, Ornithopoda) en la
Formación Allen (Cretácico Superior) de Salitral
Moreno,Provincia de Río Negro, Argentina. X
Congreso Geológico Argentino, San Miguel de
Tucumán. Actas 3: 149-152.

ROGERS, D. 1987. Description and interpretation of a
Lower Cretaceous fossil feather from Koonwarra,
South Gippsland, Victoria. 3rd. Yr. Thesis.
Monash University, Clayton, pp. 1-70.

SALGADO, L. & CORIA, R. 1993. El género
Aeolosaurus (Sauropoda, Titanosauridae) en la
Formación Allen (Campaniano-Maastrichtiano)
de la Provincia de Río Negro, Argentina.
Ameghiniana 30(2): 119-128.

SANZ, J.L., CHIAPPE, L.M. & BUSCALIONI, A.
1995. The osteology of Concornis lacustris (Aves:
Enantiornithes) from the Lower Cretaceous of
Spain and a re-examination of its phylogenetic

FOSSIL RECORD OF BIRDS IN THE MESOZOIC GONDWANA

relationships. Amencan Museum Novitates 3 133:
1-23.

SMITH, R.M.H. 1988. Palaeoenvironmental
reconstruction of a Cretaceous crater-lake deposit
in Bushmanland, South Africa. Pp. 27-41. In
Heine, K. (ed.) Palaeoecology of Africa and the
surrounding islands. Proceedings of the VIIIth
biennal conference held at the University of
Bloemfontein.

TALENT, J.A., DUNCAN, P.M. & HANDLEY, P.L.
1966. Early Cretaceous feathers from Victoria.
Emu 66: 81-86.

TOKARYK, T.T. & JAMES, P.C. 1989. Cimolopteryx
sp. (Aves, Charadriiformes) from the Frenchman
Formation (Maastrichtian), Saskatchewan.
Canadian Journal of Earth Sciences 26(12): 2729-
2730.

UNWIN, D.M. 1989. A predictive method for the iden-
tification of vertebrate ichnites and its application
to pterosaur tracks. Pp. 259-274. In Gillette, D.D.
& Lockley, M.G. (eds) ‘Dinosaur tracks and
traces’. (Cambridge University Press).

VARRICCHIO, D.J. & CHIAPPE, L.M. 1995. A new
bird from the Cretaceous Two Medicine Forma-
tion of Montana. Journal of Vertebrate Paleontol-
ogy 15(1): 201-204.

555

VICKERS-RICH, P. 1976. The history of birds on the
island continent Australia. Proceedings 16th In-
temational Ornithological Congress, Canberra,
pp. 53-65.

1991. The Mesozoic and Tertiary history of birds on
the Australian plate. Pp. 721-808. In Vickers-
Rich, P., Monaghan, J.M., Baird, R.F. & Rich,
T.H. (eds) Vertebrate Palaeontology of
Australasia. (Pioneer Design Studio and Monash
University Publications: Clayton).

VICKERS-RICH, P. & RICH, T.H.V. 1989. Polar
dinosaurs and biotas of the early Cretaceous of
southeastern Australia. National Geographic Re-
search 5(1): 15-53.

WALDMAN, M. 1970. A third specimen of a lower
Cretaceous feather from Victoria, Australia. Con-
dor 72: 377.

WALKER, C.A. 1981. New subclass of birds from the

Cretaceous of South America. Nature 292: 51-
53.

ZINSMEISTER, W. 1982. Review of the Upper
Cretaceous-Lower Tertiary sequence on Seymour
Island, Antarctica. Joumal of the Geological
Society of London 139: 779-785.

PHYLOGENETIC POSITION OF MONONYKUS (AVES: ALVAREZSAURIDAE) FROM

THE LATE CRETACEOUS OF THE GOBI DESERT
LUIS M. CHIAPPE, MARK A. NORELL & JAMES M. CLARK

Chiappe, L.M., Norell, M.A. & Clark, J.M. 1996 12 20: Phylogenetic position of Mononykus
(Aves: Alvarezsauridae) from the Late Cretaceous of the Gobi Desert. Memoirs of the
Queensland Museum 39(3) 557-582. Brisbane. ISSN 0079-8835,

In order to assess the phylogenetic relationships of Mononykus, a cladistic analysis was
performed. Using velociraptorine theropods as outgroups, the analysis resulted in a single
most parsimonious cladogram. In this cladogram the monophyletic Alvarezsauridae (includ-
ing Mononykus and the Argentine Alvarezsaurus and Patagonykus) is the sistergroup of all
other birds except Archaeopteryx. The monophyly of Aves (= Avialae sensu Gauthier) is
supported by seven unambiguous synapomorphies, four of which are present in Mononykus.
These characters include fewer than 26 caudal vertebrae, caudal vertebrae with short distal
prezygapophyses, teeth with unserrated crowns and a caudal tympanic recess opening only
inside the tympanic cavity. The monophyly of Metornithes (Aves exclusive of Archaeop-
teryx) is supported by six unambiguous synapomorphies all of which occur in Mononykus.
Among these characters are the presence of prominent ventral processes on the cervicodorsal
vertebrae, a carpometacarpus, a prominent antitrochanter in the pelvis and a rectangular,
carinate sternum. Furthermore, six synapomorphies (all present in Mononykus) ambiguously
diagnose both Aves and Metomithes. The distribution among avian and nonavian taxa of all
these characters is discussed. Several authors have criticised the hypothesis of avian relation-
ships for Mononykus. In this paper we address those criticisms. We also discuss the rationale
of testing phylogenetic hypotheses within a cladistic framework and establish that our critics
have not furnished much beyond a priori speculation. [ ] Birds, Cretaceous, phylogeny,
homology.

L.M. Chiappe, Department of Ornithology, American Museum of Natural History, Central
Park West at 79th Street, New York, NY 10024, U.S.A.; M.A. Norell, Department of
Vertebrate Paleontology. American Museum of Natural History, Central Park West at 79th
Street, New York, NY 10024, U.S.A.; J.M. Clark, Department of Biological Sciences, George

Washington University, Washington, DC, 20052, U.S.A.; 5 April 1996.

One of the central problems in palaeontology is
the origin of the major groups of terrestrial am-
niotes. One of these, birds, has received intense
scrutiny ever since, and even before, the first
specimens of Archaeopteryx lithographica were
collected. Since the publication of the London
specimen in 1861 (Meyer, 1861), this and sub-
sequent specimens of Archaeopteryx have been
key elements in discussions of bird origins.

Surprisingly, except for specimens of
Hesperornis and Ichthyornis, it took over a
hundred years for significant new specimens of
basal birds to be recognised (Chiappe, 1995a, b).
The last 15 years, however, has seen surprising
progress on this front and many important new
specimens of basal birds have been described
(e.g., Walker, 1981; Kurochkin, 1985; Sanz et al.,
1988, 1995; Chiappe, 1991, 1993, 1995a, b; AI-
varenga & Bonaparte, 1992; Sanz & Buscalioni,
1992; Sereno & Rao, 1992; Wellnhofer, 1992,
1993; Zhou et al., 1992; Chiappe & Calvo, 1994;
Hou et al., 1995).

Mononykus olecranus (Perle et al., 1993) from
the Late Cretaceous of Mongolia, is one of the
most unusual of these (Fig. 1). Mononykus is
larger than most basal birds and instead of well
developed wings it possessed stout arms that are
ridiculously short, terminating in a robust hand
with a hypertrophied digit (Perle et al., 1993,
1994; Norell et al., 19932). Many other aspects of
its morphology are a peculiar melange of primi-
tive, derived and just plain weird (Perle et el.,
1994). Apparently, Mononykus was a common
element of the Late Cretaceous fauna of Central
Asia (Norell et al., 19932). In the last few years
many specimens have been collected at several
Mongolian localities by the Mongolian American
Museum Paleontological Project (Novacek et al.,
1994; Dashzeveg et al. 1995). Furthermore,
specimens collected during the 1920s from the
Djadokhta Formation at Bayn Dzak (Norell et al.,
1993a) and the Iren Dabasu Formation in north-
ern China have been recently found in the collec-
tions of the AMNH. In earlier papers, we
predicted that such a highly recognisable mor-

558

phology may foster discovery of
members of this clade in other
faunas, where material is usually
more poorly preserved. Recently,
close relatives of Mononykus have
been identified in the Late
Cretaceous of Argentine Patagonia
(Novas, this volume). Because of
this discovery a discussion of the
phylogenetic relationships of
Mononykus fits the scope of a sym-
posium on Gondwanan dinosaurs.
In our 1993 paper we proposed a
hypothesis based on shared
derived characters placing
Mononykus in a sister-group
relationship to all birds except Ar-
chaeopteryx (Perle et al., 1993,
1994). The discovery of
Patagonykus and the reinterpreta-
tion of Alvarezsaurus as another
relative of Mononykus (see Novas,
this volume) documented that the Alvarez-
sauridae (e.g., Alvarezsaurus, Mononykus, and
Patagonykus) comprise a diverse, but monophyletic
group of primitive birds not only present in the Late
Cretaceous of central Asia but also in southern
South America and probably in western North
America (Holtz, 1994a). In this paper, we sum-
marise the evidence supporting the sistergroup
relationship of Alvarezsauridae to all other birds
except Archaeopteryx, making emphasis on Mono-
nykus [see Novas (this volume) for information on
Alvarezsaurus and Patagonykus]. Phylogenetic
relationships among Alvarezsauridae are dis-
cussed elsewhere (see Novas, this volume).

al., 1994).

MATERIALS AND METHODS

ANATOMICAL NOMENCLATURE. Anatomi-
cal terms mostly follow Baumel & Witmer
(1993), using the English equivalents of the Latin
terminology. The extrapolation of modern avian
nomenclature to successive sister groups and
even non-avian theropods is based on acceptance
of the theropodan hypothesis of avian origins
(Ostrom, 1976a; Gauthier, 1986). For most fea-
tures of modern birds it is possible to trace
homologous structures in more basal birds and
non-avian theropods.

TAXONOMIC NOMENCLATURE. In recent
years there has been disagreement as to what taxa
comprise Aves. Traditionally it is used to name a
group including all species descended from the
last common ancestor of Archaeopteryx

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 1. Skeletal reconstruction of Mononykus (modified from Perle et

lithographica and modern birds (Neornithes), al-
though Hennig's (1969, 1981) concepts of total
group (e.g., Ax, 1987; Patterson, 1993) and crown
group (e.g., Gauthier, 1986; Norell et al., 1993b;
Perle et al., 1993, 1994) have been also applied
to this clade. Using the latter concept, Gauthier
(1986) recognised the term Avialae to name the
clade traditionally named with the term Aves (see
also Rowe & Gauthier,1992; De Queiroz &
Gauthier,1992). In doing this, Gauthier (1986)
replaced the term Neornithes — traditionally used
to name this group — by redefining the term
Aves. In this paper — based on the preferences of
the senior author — we follow the traditional
nomenclature and therefore we use the term Aves
to name a clade composed of the common ances-
tor of Archaeopteryx lithographica and modern
birds (Neornithes) plus all its descendants. Note
that this definition uses a node-based phylo-
genetic definition (see De Queiroz & Gauthier,
1992) and not a character-based — usually feathers
— definition of the taxon Aves. Therefore, the
inclusion of a particular taxon within Aves is only
based on its genealogical history. The collective
terms ‘modern birds’ and ‘birds’ refer to all mem-
bers of the monophyletic taxa Neornithes and
Aves, respectively, while ‘nonavian theropods’
refer to all theropod outgroups of birds, used
without any implication of monophyly.

PHYLOGENETIC ANALYSIS. The character
analysis includes 99 characters (Appendix 1). Ten
of these are multistate characters. All characters
were treated as additive (any two states are

MONONYKUS FROM THE GOBI DESERT

separated by a number of steps that equals their
absolute arithmetic difference) except for four
multistate characters (34, 40, 64 & 92) that were
treated as non-additive (any two states are
separated by a single step). The data set was
processed using the computer program ‘Hennig
86’ (Farris, 1988). Using the implicit enumeration
(ie) command — which generates trees that are
certain to be of minimal length — a single most
parsimonious tree was obtained (Fig. 2). In order
to address the character optimisation in alterna-
tive topologies, the resultant tree was re-rooted
by using the dos equis (xx) command of ‘Hennig
86’ (see Discussion),

Because our objective is to assess relationships
rather than diagnose all groups, only characters
in which the derived state (or at least one of the
derived states of a multistate character) is present
in two or more different terminal taxa were
analyzed. In order to avoid influence on the con-
sistency index, derived characters exclusive to a
single terminal taxon were not included (Car-
penter, 1988; Wiley et al., 1991). As the result of
the phylogenetic analysis, however, some charac-
ters are autapomorphies of different taxa (e.g.,
characters 64a, 98; Fig. 2).

Polarisation of character states used in the char-
acter analysis was established by using velocirap-
torine theropods as the outgroup. We are aware
that the use of a different outgroup (e.g., Troodon-
tidae, Oviraptoridae) might result in a different
tree topology. This decision, however, was based
on a combination of previous phylogenetic re-
search with the anatomical information currently
at hand: Several recent phylogenetic hypotheses
on maniraptoran dinosaurs have considered
velociraptorine theropods (e.g., Deinonychus,
Velociraptor, Adasaurus) as closely related to
birds (Gauthier, 1986; Novas, 1992; Holtz,
1994b) and the available information on other
nonavian maniraptoran dinosaurs (e.g., troodon-
tids, oviraptorids, segnosaurids) is still limited.

The analysed ingroup included taxa
hypothesised to be closely related to Mononykus
such as the Argentine Alvarezsaurus (Bonaparte,
1991) and Patagonykus (Novas, this volume),
along with the best represented avian taxa: Ar-
chaeopteryx, Iberomesornis, Enantiornithes,
Patagopteryx, Hesperornithiformes, Ichthyor-
nithiformes and Neornithes.

MATERIAL EXAMINED. The anatomical inforrna-
tion on Velociraptorinae was based mostly on the
AMNH and YPM's specimens of Deinonychus antir-
rhopus (Ostrom, 1969) and the holotype, and several
new unpublished specimens (see Norell et al., 1992) of

559

Velociraptor mongoliensis (Osborn, 1924). Additional
information was taken from Ostrom (1969, 1974,
1976b, 1990) and Barsbold (1983).

Five specimens of Mononykus olecranus were used in
this study: the holotype (MGI 107/6) and specimens
MGI N100/99, MGI 100/975, MGI 100/977 and IGM
100/1001. The holotype specimens of Alvarezsaurus
calvoi (Bonaparte, 1991) and Patagonykus puertai
(Novas, this volume) from the Late Cretaceous of
Patagonia (Argentina) were also studied.

The osteological data on Archaeopteryx lithographica
was taken from De Beer (1954), Wellnhofer (1974,
1992, 1993), Ostrom (19762), Martin (1983),
Whetstone (1983), Walker (1985), Bühler (1985), Wit-
mer (1990) and Elzanowski & Wellnhofer (1995) as
well as examination of the Eichstátt, Solnhofen and
London specimens, and a cast of the Berlin specimen.
The holotype specimen of /beromesornis romerali
(Sanz et al., 1988; Sanz & Bonaparte, 1992) was also
studied. Almost all the available material of Enantior-
nithes was examined, including undescribed
specimens from the Late Cretaceous of Argentina (El
Brete; see Chiappe, 1993, 1996) and specimens of
Enantiornis leali (Walker, 1981; Chiappe, this
volume), Lectavis bretincola (Chiappe, 1993), Yun-
gavolucris brevipedalis (Chiappe, 1993),
Soroavisaurus australis (Chiappe, 1993), Neuquenor-
nis volans (Chiappe & Calvo, 1994), Concornis
lacustris (Sanz & Buscalioni, 1992; Sanz et al., 1995)
and Cathayornis yandica (Zhou et al., 1992). Casts of
the enantiomithines Sinornis santensis [Sereno & Rao,
1992; arguments supporting its inclusion within Enan-
tiornithes are presented elsewhere (Chiappe, 1995b)],
Avisaurus gloriae (Varricchio & Chiappe, 1995),
Avisaurus archibaldi (Brett-Surman & Paul, 1985;
Chiappe, 1992b, 1993) and Nanantius eos (Molnar,
1986) were also studied. All the available material of
Patagopteryx deferrariisi (Alvarenga & Bonaparte,
1992; Chiappe, 1992a, 1996) was also examined. The
anatomical data on the Hesperornithiformes mostly
derives from the seminal monograph of Marsh (1880),
and papers of Martin & Tate (1976), Martin (1980,
1983, 1984), Biihler et al. (1988), Witmer (1990) and
Elzanowski (1991). Specimens at the AMNH, FMNH,
UK and YPM were also examined. Information on
Ichthyornithiformes was derived mainly from Marsh’s
(1880) description of Jchthyornis dispar and Ichthyor-
nis victor and the study of specimens labeled as Ich-
thyornis sp. at the YPM. The skeletal material of
different representatives of several groups of modern
paleognathes and neognathes (Aves) was surveyed.
This information was supplemented by such general
osteological papers such as Jollie (1957), Webb (1957),
Bellairs & Jenkin (1960), King & McLelland (1984)
and Baumel & Witmer (1993).

Institutional Abbreviations. AMNH, American
Museum of Natural History (New York); FMNH,
Field Museum of Natural History (Chicago);
MGI, Mongolian Geological Institute (Ulan

560 MEMOIRS OF THE QUEENSLAND MUSEUM

Carinàtag Neornithes

Ornithurae Ichthyornithiformes

Hesperornithiformes

Patagopteryx

Ornithothoraces 2 M
Enantiornithes

Metornithes Iberomesornis

Mononykus

Alvarezsauridae Patagonykus

Aves

Alvarezsaurus
Archaeopteryx

FIG. 2. Analysis cladogram. Coding: asterisked characters (*) have equivocal optimization; the symbol (-)
indicates a character reversal; ‘a’ and ‘b’ refer to states 1 and 2, respectively, of a multistate character (see
Appendix 1). NODE 1 (Aves = Avialae sensu Gauthier, 1986): 11*, 15*, 19a, 23a, 46*, 59a*, 68*, 71, 72*,
74*, 81, 83*, 84*, 85a, 86*, 87, 88, 90*, 95*, 96*. NODE 2 (Metornithes): 6*, 8*, 11*, 13, 14a*, 15*, 29, 30,
31*, 34a*, 34b*, 36*, 39, 41*, 44, 45*, 46*, 47*, 51*, 55*, 59a*, 68*, 69, 72*, 74*, 83*, 84*, 86*, 90*, 92a*,
95*, 96*. NODE 3 (Ornithothoraces): 1*, 3*, 5*, 6*, 7*, 8*, 10*, 11*, 14a*, 18, 21*, 23b, 24*, 25, 26*, 27,
31*, 32*, 34a*, 34b*, 36*, 37*, 38, 40a*, 40b*, 41*, 45*, 47*, 50*, 51*, 53*, 54a*, 55*, 60*, 63*, 67*. 70*,
72*, 83*, 84*, 86*, 90*, 92a*, 93*, 96*, 97*. NODE 4: 1*, 3*, 5*, 6*, 7*, 10*, 11*, 16, 17, 19b*, 21*, 24*,
26*, 31*, 32*, 34b*, 35, 36*, 37*, 40a*, 40b*, 45*, 47*, 50*, 51*, 53*, 54a*, 55*, 56, 60*, 62, 63*, 67*, 70*,
72*, 73, 76*, 83*, 84*, 86*, 90*, 92a*, 93*, 96*, 97*. NODE 5: 2*, 4*, 6*, 7*, 10*, 11*, 19b*, 28, 31*, 32*,
40a*, 40b*, 45*, 47*, 40, 54b, 55*, 61, 66, 67*, 70*, 72*, 76*, 82, 85b, 87*, 90*, 92a*, 93*, 04*, 97*. NODE
6 (Ornithurae): 2*, 4*, 6*, 7*, 9, 11*, 12, 14b, 19b*, 20, 22, 32*, 40a*, 40b*, 42, 43, 48, 52, 57, 58, 59b, 64b,
65, 67*, 70*, 72*, 87*, 91*, 92a*, 93*, 94*, 97*, 99*, NODE 7 (Carinatae): 11*, 32?, 33, 40b*, 70*, 87*, 91*,
92a*, 92b*, 99*, NODE 8 (Alvarezsauridae, see Novas, this volume): 6*, 8*, 11*, 14a*, 23*, 31*, 34a*, 34b*,
36*, 41*, 45*, 47*, 48*, 51*, 55*, 75*, 78*, 79*, 80*, 83*, 84*, 89, 90*, 92a*, 96*. NODE 9: 6*, 8*, 11*,
14a*, 23*, 31*, 34a*, 36*, 41*, 47*, 48*, 51*, 75*, 76, 77, 78*, 79*, 80*, 83*, 84*, 90*, 92a*, 96*, Resultant
apomorphies: Mononykus (64a), Enantiornithes (98), Neornithes (98). In nodes 4 and 5, if character 76 is
synapomorphic, it becomes a reversal in node 6 (Ornithurae).

Bator); UK, Museum of Natural History, Univer-
sity of Kansas (Lawrence); YPM, Yale Peabody
Museum (New Haven).

CHARACTER ANALYSIS

This analysis resulted in a single most par-
simonious cladogram with low homoplasy
(length, 143; rescaled consistency index, 0.76;
retention index, 0.81). In this cladogram (Fig. 2)
the monophyly of Alvarezsauridae (Mononykus,
Patagonykus and Alvarezsaurus; see Novas, this
volume) is supported, and this group is the sister-
group of all birds other than Archaeopteryx. In an
earlier paper we coined the term ‘Metornithes’ to
name this monophyletic group (Perle et al.,
1993).

Below we describe those characters
synapomorphic of both Aves and Metornithes and

which are known to be present in Mononykus.
Reference is made to the condition in the ingroup
and outgroup taxa, along with that found in other
nonavian theropods. Missing entries are in most
cases not mentioned (see data matrix in Appendix
1 for character scoring).

CHARACTERS SUPPORTING THE MONO-
PHYLY OF AVES. The monophyly of Aves (=
Avialae sensu Gauthier, 1986) is supported by
seven unambiguous synapomorphies (Fig. 2).
Four of these (character states 19a, 71, 81 & 85a)
are present in Mononykus. The available material
of Mononykus and the remaining Alvarezsauridae
does not allow determination of the condition in
two of these characters (87 & 88), while
Mononykus shows the primitive condition for the
remaining character (23a). The four avian
synapomorphies present in Mononykus are:

MONONYKUS FROM THE GOBI DESERT

Archaeopteryx Enantiornithes

FIG. 3. Avian dental morphology of Archaeopteryx, the enantiornithine

Cathayornis and Hesperornis. Drawings not to scale.

1) Caudal vertebral count smaller than 25-26
elements (character 19). In moder birds, the tail
is composed of a series of free caudal vertebrae
and the pygostyle. The free vertebral count of
modern birds ranges between four and eight,
being typically between five and seven (Ver-
heyen, 1960). The pygostyle is usually composed
of five or six vertebrae (Baumel & Witmer, 1993).
Hence, the total caudal count of modern birds
(free caudals + pygostyle) includes fewer than 15
elements. Among hesperornithiforms, Marsh
(1880) estimated 12 elements in the tail of
Hesperomis and Martin & Tate (1976) illustrated
14 caudal vertebrae in Baptornis. Iberomesornis
has eight free caudals and a large pygostyle.
Some of the elements forming the pygostyle of
Iberomesornis are clearly distinguishable. Sanz
& Bonaparte (1992) correctly estimated that 10
to 15 elements form the pygostyle. Therefore,
Iberomesornis has a caudal count of no more than
23 vertebrae. In Archaeopteryx, the number of
caudal elements ranges from 20 to 23 in the
different specimens (Ostrom, 19762).

Velociraptorines and other nonavian theropods,
on the contrary, have much longer tails with at
least 36 vertebrae (Osborn, 1916; Lambe, 1917;
Osmolska et al., 1972; Madsen, 1976). There are
approximately 36 to 40 caudal vertebrae in
Deinonychus (Ostrom, 1969). Curiously, the
troodontid Sinornithoides (Russell & Dong,
1993a) has 27 preserved caudals and, as es-
timated by these authors, a total caudal count of
no more than 30 elements.

The number of caudal vertebrae in Mononykus
is significantly lower than in the outgroup and
similar to the number found in Archaeopteryx. In
specimens MGI N100/99 and MGI 100/975, 19
caudal elements are preserved. Some distal ele-
ments are missing but based on the size and
morphology of the last preserved elements we are

Hesperornis

561

confident in our estimate that the
number of caudal elements of
Mononykus was not larger than
25-26.

2) Teeth with unserrated
crowns (character 71). Neor-
nithine birds lack teeth, but a
variety of basal birds bear both
cranial and mandibular dentition.
In the teeth of Hesperornithifor-
mes, Ichthyornithiformes and Ar-
chaeopteryx the enamel of the
crowns is smooth, lacking serra-
tions (Martin et al., 1980; Martin,
1985) (Fig. 3). The same condi-
tion is present in the Early
Cretaceous enantiornithine Cathayornis (Zhou et
al., 1992).

In contrast, adult velociraptorine theropods
have serrated crowns (Osborn, 1924; Ostrom,
1969; Currie et al., 1990) (Fig. 4). This is the case
for most other non-avian theropods in which the
enamel is serrated in at least some areas of the
crown tooth (Currie et al., 1990; Fiorillo & Cur-
rie, 1994; see Ostrom, 1991 for a few exceptions).
In the description of Archaeornithoides, El-
zanowski & Wellnhofer (1992) considered the
absence of dental serrations as a synapomorphy
of the clade formed by the latter taxon and birds.
The only, and very fragmentary, specimen of
Archaeornithoides is clearly a juvenile and there-
fore not an adequate specimen for phylogenetic
inferences. As we have recently shown,
dromaeosaurid neonates lack serrations as well
(Norell et al., 1994), suggesting that in
dromaeosaurids, teeth became serrated during
postnatal ontogeny. A similar ontogenetic
modification is known to occur in extant non-
avian archosaurs (i.e., crocodiles). The absence
of dental serrations in the juveniles of theropods
closely related to birds might indicate that the
avian tooth morphology arose through
heterochrony (Norell et al., 1994),

In our preliminary description of Mononykus
(Perle et al., 1993) and in a later paper (Perle et
al., 1994), we described a tooth that was found
isolated inside the fragmentary skull. The crown
of this tooth possesses rostral and caudal carinae
and lacks serrations (Fig. 5). Confirmation of this
dental morphology has come from a recently
discovered articulated specimen (MGI 100/977),
including the skull, from the Djadokhta-like red
beds of Ukhaa Tolgod (Dashzeveg et al., 1995),
in the southwestem Mongol Gobi. In specimen
MGI 100/977 both cranial and mandibular teeth

562

FIG. 4. Teeth of Velociraptor (AMNH65 18). Note the
serrated margins.

are preserved in their natural position, and they
lack serrations.

3) Caudal tympanic recess opens inside the
collumelar recess and not in the paroccipital
process (character 81). The caudal tympanic
recess of modern neornithine birds, the recess
formed by the caudal evagination of the tympanic
air sac (Witmer, 1990), consistently opens inside
the collumelar recess (Witmer, 1990; Baumel &
Witmer, 1993). A similar configuration of the
tympanic region occurs in Hesperomithiformes
and Archaeopteryx (Witmer, 1990) (Fig. 6).

The caudal tympanic recess is well-preserved
in arecently discovered braincase of Velociraptor
(Norell et al., 1992). CAT scan imaging has
shown that, as in modern birds, it extends inside
the paroccipital process. An important difference
with birds (Witmer, pers. comm.), however, is
that the caudal tympanic recess opens on the
rostral surface of the paroccipital process, outside
the collumelar recess. This braincase configura-
tion is known to occur in several non-avian
theropods such as Dromaeosaurus and Itemirus
(Currie, 1995), Struthiomimus and a new
maniraptoran from the St Mary River Formation
(Witmer & Weishampel, 1993; Witmer, pers.
comm.) and Protoavis (Chatterjee, 1991), which
we do not regard as a bird (see Chiappe, 1995b).
Interestingly, Currie & Zhao (1993b) have men-
tioned that this external paroccipital opening is
absent in troodontids.

The tympanic region of Mononykus resembles
that of Archaeopteryx, Hesperornis and neor-

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 5. Electron micrograph of a tooth of Mononykus
(MGI 107/6). Note the complete absence of serra-

Archaeopteryx

nykus
aara Monony
a
J RTR AOF
POP AT

MX

2 C

FIG. 6. Braincase morphology of Archaeopteryx
(redrawn from Walker, 1985) and Mononykus (MGI
107/6). Note that the caudal tympanic recess does not
open into the paroccipital process. AFQ=articular
facet for the quadrate, AOF=antorbital fossa,
MXzmaxilla, POP=paroccipital process, Q=quad-
rate, RTR=rostral tympanic recess. Drawings not to
scale.

9
a

MONONYKUS FROM THE GOBI DESERT

ones Gallimimus

Deinonychus

563

FIG. 7. Caudal vertebrae of Deinonychus (after Ostrom, 1969) and Gallimimus (twenty second caudal; after
Barsbold & Osmólska, 1990) in dorsal and lateral views. Caudal vertebrae of Archaeopteryx (tenth to fifteenth
vertebrae; after Wellnhofer, 1974). PRZ-prezygapophysis. Drawings not to scale.

nithine birds. In the braincase preserved in the
holotype specimen (Perle et al., 1994), the paroc-
cipital process is not perforated by any foramen
(Fig. 6) and the caudal tympanic recess opens in
the collumelar recess (as in Archaeopteryx; con-
tra Currie, 1995).

4) Short or reduced prezygapophyses in distal
caudal vertebrae (character 85). The caudal ver-
tebrae of modern birds typically have small, or
even absent, prezygapophyses. Caudal
prezygapophyses are absent in Hesperomithifor-
mes (Marsh, 1880; Martin & Tate, 1976) and
Patagopteryx (Chiappe, 1992a). In Ichthyor-
nithiformes (Marsh, 1880), distinct but short
prezygapophyses are present in the proximal
caudals. Within the Enantiornithes, this character
is not determinable in the caudal vertebrae of
either Concornis (Sanz et al., 1995) or Cathayor-
nis (Zhou et al., 1992). Nevertheless, a caudal
vertebra preserved in an as yet undescribed enan-
tiornithine specimen from Alabama (Lamb et al.,
1993) has short prezygapophyses. The free
caudal vertebrae of Iberomesornis (Sanz &
Bonaparte, 1992) also bear short or reduced
prezygapophyses as well. In Archaeopteryx, the
caudal prezygapophyses appear to be fairly short,
extending only slightly over the preceding ver-
tebra (Fig. 7). This is the condition present in the
Eichstátt specimen (Wellnhofer, 1974). In the
London specimen, the prezygapophyses seem to
be longer, however it is hard to provide an ac-
curate estimate of their cranial projection because
the articulations between the centra are not ex-

posed. In any case, it is clear that the
prezygapophyses of the London specimen are far
shorter than those of several non-avian theropods
(see below), projecting less than 25% the length
of the preceding vertebra.

The presence of remarkably long, rod-like
prezygapophyses in the caudal series of
velociraptorine theropods is well known
(Ostrom, 1969, 1990) (Fig. 7). In Deinonychus,
the elongate prezygapophyses are present in all
caudals distal to the eighth or ninth element
(Ostrom, 1969). Elongated prezygapophyses —
yet not to extent of the extremely apomorphic
condition seen in velociraptorines — are known
to occur in the distal caudals of several non-avian
theropods, in particular in the middle and distal
portions of the tail (see Lambe, 1917; Ostrom,
1969; Barsbold, 1974). The omithomimid Gal-
limimus, for example (Fig. 7), has distal caudal
prezygapophyses extending up to two-thirds the
length of the preceding vertebra (Barsbold &
Osmólska, 1990) and in Allosaurus they extend
for at least half the length of the preceding ele-
ment (Madsen, 1976).

In Mononykus, the caudal prezygapophyses,
and in particular those of the distal portion of the
tail, are short and do not extend to the preceding
vertebra. Interestingly, in contrast to those non-
avian theropods with long caudal prezygapophyses,
the proximal caudal vertebrae of Mononykus
have prezygapophyses that are longer — though
still relatively short — than those of the distal
vertebrae (Fig. 8).

564

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 8. Caudal vertebrae of Mononykus (MGI 100/975). A, B, first? to seventh? vertebrae in dorsal and lateral
view. C, 13th? to 19th? vertebrae in lateral view.Note the short prezygapophyses (PRZ). Scale = 1cm.

CHARACTERS SUPPORTING THE MONO-
PHYLY OF METORNITHES. The monophyly
of Metornithes (Perle et al., 1993), the clade
composed of the closest common ancestor of
Mononykus and Neornithes plus all its descen-
dants, is supported by six unambiguous
synapomorphies. All these synapomorphies are
present in Mononykus. These synapomorphies
are:

1) Prominent ventral processes on cervicodor-
sal vertebrae (character 13,). The cervicodorsal
vertebrae of ornithurine birds (Neornithes,

Hesperornithiformes and Ichthyornithiformes)
bear prominent ventral processes for the origin of
M. longus colli ventralis (Chiappe, 1992a, 1996),
a primary depressor of the neck (Zusi, 1962) (Fig.
9). These processes are also well developed in the
cervicodorsal vertebrae of Iberomesornis, Enan-
tiornithes and Patagopteryx deferrariisi, but they
appear to be absent in Archaeopteryx (Chiappe,
1996).

In contrast, ventral processes are only slightly
developed in velociraptorine theropods (Ostrom,
1969; Gauthier, 1986) (Fig. 9) and are usually

MONONYKUS FROM THE GOBI DESERT

Deinonychus Sinraptor

eornithes

N
— Vep—

FIG. 9. Dorsal vertebrae of Deinonychus (fourth? dor-
sal; after Ostrom, 1969) and Sinraptor (first dorsal;
after Currie & Zhao, 1993a) in caudal and lateral
views. Dorsal vertebrae of a neornithine bird in
cranial and lateral views. Note the large neural canal
and the absence of hypantrum in neornithine birds.
Hy-hyposphene, Nc=neural canal, Vep=ventral
process. Drawings not to scale,

absent (Osborn, 1916; Lambe, 1917; Madsen,
1976; Ostrom, 1978; Bonaparte et al., 1990) or
barely developed (Russell & Dong, 1993a; Currie
& Zhao, 1993a) in other non-avian theropods
(Fig. 9).

In Mononykus ventral processes are present in
the cervicodorsal region (Fig. 10). In the holotype
specimen (Perle et al., 1994), the last two transi-
tional vertebrae between cervicals and dorsals
have small ventral processes but the first
preserved dorsal bears a much more prominent,
well-developed process (Fig. 10). The develop-
ment of the ventral process of the first(?) dorsal
of Mononykus resembles that of more advanced
birds and not the blunt process of velociraptorine
theropods.

2) Sternum of longitudinal rectangular shape
(character 29). In the Ornithurae the sternum is
typically large and rectangular, with its
craniocaudal axis longer than the transverse axis.
In these birds, the ratio between maximum length
and maximum width (including lateral
trabeculae) (ML/MW) is usually larger than 1.4
(Chiappe, 1996). (In some birds with broad ster-
na, such as Cuculidae, Caprimulgidae, Picidae
and Trogonidae, this ratio is slightly smaller).
This derived condition is also present in the Enan-

565

tiornithes. The sterna of the enantiornithines Neu-
quenornis (Chiappe & Calvo, 1994) and Concor-
nis (Sanz et al., 1995) are large and rectangular,
with a ML/MW ratio of at least 1.66 in the former
taxon. In Archaeopteryx, in contrast, the sternum
is transversely broader than long as has been
recently described by Wellnhofer (1993) in the
*Solenhofer Aktien-Verein' specimen (Fig. 11).

In velociraptorines and other non-avian
theropods in which stemal ossifications are
known, these are formed by two quadrangular
plates, which sometimes fuse to each other
(Lambe, 1917; Barsbold, 1983; Bonaparte et al.,
1990; Currie & Zhao, 1993a). Currie & Zhao
(1993a) considered the absence of fusion be-
tween sternal plates as related to immaturity. In
Velociraptor (Fig. 11), the ML/MW ratio is about
1.04 (Barsbold, 1983) and approximately 0.95
and 0.77 in the oviraptorids Oviraptor and In-
genia, respectively (Barsbold, 1983).

Mononykus has a longitudinally rectangular
sternum (Perle et al., 1993, 1994), as it has been
described in the holotype specimen and cor-
roborated by specimen MGI 100/977 in which the
sternum is preserved in natural position. The
ML/MW ratio of the stemum of the holotype
specimen of Mononykus is at least 1.93 (Fig. 12).

3) Ossified sternal keel (character 30). In neor-
nithine birds generally (except in ratites and some
other flightless birds), the sternum has a large
ventral keel from which the main flight muscles
arise. A sternal carina is also present in ichthyor-
nithiforms (Marsh, 1880), in the enantiomithines
Neuquenornis (Chiappe & Calvo, 1994), Concor-
nis (Sanz et al., 1995), and Cathayornis (Zhou et
al., 1992), but it is absent in hesperornithiforms
(Marsh, 1880) and Archaeopteryx (Wellnhofer,
1993) (Fig. 11).

In velociraptorines (Fig. 11) and most non-
avian theropods in which sternal ossifications are
known, the carina is completely absent (Lambe,
1917; Barsbold, 1983; Bonaparte et al., 1990). An
exception has been recently reported by Currie &
Zhao (19932) who described a low, blunt ventral
keel in the sternum of Sinraptor, a taxon related
to Allosaurus.

In Mononykus the sternum has a well-
developed ventral keel (Fig. 12). Although
carinate, this sternum differs from all other
carinate sterna in that it is subtriangular in cross-
section and not T-shaped. This latter condition
has been considered an apomorphy of
Mononykus (Perle et al., 1994).

4) Distal carpals fused to metacarpals forming
a carpometacarpus (character 39). Neornithine

566

FIG. 10. Dorsal vertebrae of Mononykus (MGI 107/6).
A, D, anterior first preserved vertebra in lateral and
caudal view; B, C, anterior second preserved vertebra
in lateral and caudal view. COF=costal fovea
(parapophysis), DOP=dorsal process, IPF=in-
frapostzygapophysial fossa, WC=vertebral canal,
VEP=ventral process. Scale = Icm.

birds have a carpometacarpus formed by the
fusion of several central and distal carpals with
the metacarpals of the alular, major and minor
digits (digits I, IT, III) (Fig. 13). Fusion of carpal
and metacarpal bones to form a carpometacarpus
also occurs in ichthyornithiforms, enantior-
nithines and Patagopteryx (Chiappe, 1996). In
contrast a carpometacarpus does not occur in
Archaeopteryx (Fig. 13), in which the metacar-
pals fuse neither with each other nor with the
distal carpals (Ostrom, 19762).

The carpals and metacarpals of velociraptorine
theropods are not fused to each other (Ostrom,
1969, 1990) (Fig. 13), a condition common to all
non-avian theropods (Madsen, 1976; Barsbold,

MEMOIRS OF THE QUEENSLAND MUSEUM

1983; Weishampel et al., 1990; Currie & Zhao,
1993a; Russell & Dong, 1993a, b).

In Mononykus, metacarpals and at least one
distal carpal are fused into a massive, quadran-
gular carpometacarpus (Perle et al., 1994) (Fig.
14). Sutures and lines of contact — best seen on
the dorsal surface of the holotype specimen — are
present between the alular, minor and major
metacarpals (metacarpals I, II and IIT). Earlier we
considered that at least one carpal homologous to
the ‘semilunate’ bone of non-avian theropods and

Archaeopteryx (see Ostrom, 1976a) was fused to

the proximal end of the alular metacarpal (Perle
et al., 1994). This interpretation was based on the
convex, pulley-like morphology of the proximal
end of the carpometacarpus (Fig. 14) which
resembles the condition in Archaeopteryx and
non-avian maniraptoran dinosaurs. Were the
'semilunate' carpal not fused to the alular
metacarpal, the proximal end of the latter would
be nearly flat to slightly concave, as in non-avian
theropods.

Velociraptor Archaeopteryx

FIG. 11. Sterna of Velociraptor (after Barsbold, 1983)
and Archaeopteryx (after Wellnhofer, 1993) in
ventral view. Drawings not to scale.

MONONYKUS FROM THE GOBI DESERT

567

FIG. 12. Sternum of Mononykus (MGI 107/6). A, stereopair in ventral view; B, cranial view; C, lateral view.

5) Pelvis with prominent antitrochanter (char-
acter 44, Appendix 1). In the caudodorsal angle
of the acetabulum of ornithurine birds there is a
prominent articular facet, the antitrochanter, typi-
cally formed by contributions of the ischium and
ilium (Fig. 15). A prominent antitrochanter is
found in both Enantiornithes and Patagopteryx
but it is absent in Archaeopteryx (Chiappe, 1996)
(Fig. 15).

In contrast, a prominent antitrochanter is not
developed in velociraptorine theropods (Ostrom,
1969, 1976b) (Fig. 15) norin any other non-avian
theropod (see Osborn, 1916; Osmólska et al.,
1972; Madsen, 1976; Weishampel et al., 1990;
Currie & Zhao, 1993a; Zhao & Currie, 1993).
Russell & Dong (1993a) reported the presence of
an antitrochanter in the troodontid Sinor-
nithoides, but the absence of detailed illustrations
along with the fact that we have not seen the
specimen prevents comparisons with those of
birds.

In Mononykus, the pelvis has a very robust and
well-developed antitrochanter (Perle et al., 1994)
(Fig. 16). The degree in which both ischium and
ilium contribute to the formation of the an-
titrochanter of Mononykus is obscured by the fact
that in adult specimens such as the holotype, the
ischium and ilium are fused. The fact that the

robust antitrochanter of Patagonykus (Novas, this
volume) is formed by equal contributions of both
ilium and ischium suggests that this was probably
the case for Mononykus. In contrast to most or-
nithurine and enantiornithine birds the an-
titrochanter is developed below the dorsal margin
of the acetabulum, and its main axis is not
oriented dorsocaudally but caudoventrally (Figs
15, 16). A similar position and orientation of the
antitrochanter, however, is present in Patagop-
teryx (Chiappe, 1992a, 1996).

6) Ischium more than two-thirds of pubic length
(character 69). In neornithine birds the pubis is
longer than the ischium but this difference is
typically much less than one-third the length of
the ischium (Fig. 15). The pubis is only slightly
longer than the ischium in hesperornithiforms,
Patagopteryx (Fig. 15) and the enantiornithines
Concornis (Sanz et al., 1995) and Sinornis
(Sereno & Rao, 1992). In contrast, in Archaeo-
pteryx the length of the ischium is between 44 to
48% that of the pubis (Wellnhofer, 1985, 1992)
(Fig. 15). Interestingly, the ischium appears to be
proportionally shorter in the *Solenhofer Aktien-
Verein’ specimen, recognised as a different
species — Archaeopteryx bavarica instead of
Archaeopteryx lithographica — by Wellnhofer
(1993).

568 MEMOIRS OF THE QUEENSLAND MUSEUM

MJ

FIG. 13. Distal carpals and metacarpals of Deinonychus, Archaeopteryx, Enantiornithes, Ichthyornis and
Neornithes. Note the fusion of these elements in the three latter taxa.

FIG. 14. Carpometacarpus of Mononykus (MGI 106/7). A, distal view; B, ventral view; C, dorsal view; D,
proximal view. ALM=alular metacarpal, AMA=articular facet of major metacarpal, AMlI-articular facet of
minor metacarpal, FCE=central proximal articular facet, FLA=lateral proximal articular facet, FMI-proximal
articular facet of minor metacarpal, MAM=major metacarpal, MIM=minor metacarpal,S AM-suture between
alular and major metacarpals, SMM=suture between major and minor metacarpals.

MONONYKUS FROM THE GOBI DESERT

Deinonychus Archaeopteryx

Enantiornithes | Hesperornithiformes

Ant=antitrochanter. Drawings not to scale.

Velociraptorine theropods, as in other non-
avian maniraptoran theropods (e.g., Russell &
Dong, 1993a), have ischia that are two-thirds or
less the length of the pubis (Gauthier, 1986). In
Deinonychus (Ostrom, 1976b) and Adasaurus
(Barsbold, 1983), the ischium is nearly 50% the
size of the pubis (Fig. 15). In contrast, in
Mononykus, both pubis and ischium — oriented
some 45? caudoventrally — have delicate, rod-
like shafts of subequal length (Fig. 17).

Gauthier (1986) has hypothesised that the more .

similar length of the pubis and ischium in or-
nithurine birds was acquired through reduction of
the former bone. Nevertheless, the fact that the
obturator process of the ischium of these birds is
located proximally and not mediodistally as in
velociraptorines or Archaeopteryx suggests the
opposite. It is probably the elongation of the
ischiadic blade that accounts for the proportion
seen in ornithurine birds.

AMBIGUOUS SYNAPOMORPHIES EX-

CLUSIVE OF BOTH AVES AND METORN-
ITHES. Additional support for the avian affinity
of Mononykus and the Alvarezsauridae comes
from six other characters in which optimisation is
ambiguous or equivocal for the present character
distribution. A closer examination of the data
indicates that the ambiguity for this optimisation
is mostly derived from the fact that these charac-
ter states are uncertain in Archaeopteryx. It is

Patagopteryx

Neornithes

C
Ant
Ant Ant

FIG. 15. Lateral view of the pelvis of Deinonychus, Archaeopteryx,
Patagopteryx, Enantiornithes, Hesperornithiformes and Neornithes.

569

important to note that given a
known morphology for Archaeo-
pteryx, these characters would be
unambiguously synapomorphic of
either Aves or Metornithes. The
fact that these character states are
clearly not present in the outgroup,
and that they represent
symplesiomorphies for clades
more derived than Alvarez-
sauridae, provides further support
for our hypothesis. These
synapomorphies are:

1) Wide vertebral foramen in
dorsal vertebrae, vertebral
foramen/cranial articular facet
ratio greater than 0.40 (character
15). The dorsal vertebrae of
modern, neornithine birds possess
a large vertebral foramen
(Chiappe, 1996) (Fig. 9). Despite
wide variation, in the sample of
neornithine birds taken for this
study, the ratio between the verti-
cal diameters of the vertebral
foramen and the cranial articular surface ranges
approximately from 0.55-2.75. Typically, the
anterior dorsals give larger ratios than the
posterior ones. In hesperornithiforms and ich-
thyornithiforms this value is at least 0.70 (Marsh,
1880). In Enantiomithes and Patagopteryx, al-
though the size of the vertebral foramen falls
among the lower values observed in neornithine
birds, the ratio is clearly greater than 0.40.

A quite contrasting condition occurs in
velociraptorines and other non-avian theropods
in which the vertebral foramen of the dorsal ver-
tebrae is very small (Fig. 9), with the above ratio
being much lower than 0.40 (see for example
Ostrom, 1969; Madsen, 1976; Currie & Zhao,
19932). In the holotype specimen of Mononykus,
however, the ratio between the vertebral
foramen/articular cranial facet is approximately
0.75 in the anterior-most dorsals (Fig. 10) and
about 0.58 in the more posterior biconvex ver-
tebra. This ratio is approximately 0.45 in the only
dorsal vertebra of Patagonykus that preserves the
cranial portion (Novas, this volume, in press).

2) Lack of contact between ischial terminal
processes (lack of ischial symphysis) (character
46). In all omithurine birds, excepting only the
Rheidae (i.e., rheas), the terminal processes of the
ischia do not contact with each other. The is-
chiadic terminal processes do not contact each
other in Enantiornithes (based on Concornis; see

570

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 16. Pelvis of Mononykus (MGI 106/7). A, stereopair of lateral view of the ilium and pubis; B, ventral view
of the ilium; C, proximal view of the pubis. ANT=antitrochanter, PUP=pubic peduncle, SUC=supraacetabular
crest.

Sanz et al., 1995) and Patagopteryx (Chiappe,
1992a, 1996).

In velociraptorines and the remaining non-
avian theropods (Romer, 1956; Osmólska et al.,
1972; Madsen, 1976; Currie & Zhao, 1993a) the
ischia form a terminal symphysis. In contrast, in
Mononykus the distal ends of the ischia do not
contact each other (Perle et al., 1993, 1994), as is
clearly visible in specimens MGI N 100/99 and
MGI 100/975 in which the pelvic elements are in
articulation.

3) Fibular tubercle for M. iliofibularis laterally
projecting (character 59, Appendix 1). In the fibula
of ornithurine birds, just proximal to the spine,
there is a caudolaterally, or caudally, projecting
tubercle for the insertion of M. iliofibularis, an
important flexor of the tibiotarsus. The fibula is
not well known for most non-ornithurine birds. In
Patagopteryx, however, it is well-preserved and
exhibits a robust tubercle for the M. iliofibularis,
which projects directly laterally instead of
caudolaterally or caudally (Chiappe, 19922).

In velociraptorine theropods and other non-
avian theropods, this tubercle (also known as the
anterolateral process) typically projects
craniolaterally (Osborn, 1916; Lambe, 1917;

Ostrom, 1969; Osmólska et al., 1972; Welles,
1984; Currie & Zhao, 19932).

Mononykus shares the condition of Patago-
pteryx, with a robust and laterally oriented M.
iliofibularis tubercle (Fig. 18). Assuming that the
anterolateral process of non-avian theropods and
the tubercle for M. iliofibularis of ornithurines are
homologous, this tubercle must have migrated
caudally during the early evolution of birds. The
lateral position of this tubercle in both
Mononykus and Patagopteryx represents an
‘intermediate’ stage in this transformation.

4) Quadratojugal not contacting the squamosal
(character 68). Neornithine birds are charac-
teristic among theropods (and most archosaurs)
in that the quadratojugal is a rod-like bone lacking
a dorsal process for its articulation with the
squamosal, which forms a free, ventrally
projected zygomatic process. This derived condi-
tion is clearly present in Hesperornithiformes
(Marsh, 1880; Bühler et al., 1988; Elzanowski,
1991). In Patagopteryx and Ichthyornithiformes
the quadratojugal is not known (at least for the
published specimens), but the presence of a
zygomatic process indicates that the squamosal
and the quadratojugal do not contact each other.

MONONYKUS FROM THE GOBI DESERT

FIG. 17. Pelvis of Mononykus (MGI 100/975, left; MGI N 100/99, right) in lateral view. Note the subequal length
of the pubis (PUB) and ischium (ISC). Scale = 1cm.

In Archaeopteryx, the quadratojugal has a dor-
sal process comparable to that of non-avian
theropods, as it can be seen in the counter-slab of
the recently found 'Solenhofer Aktien-Verein'
specimen (Wellnhofer, 1993). The morphology
of the squamosal of Archaeopteryx has been a
matter of strong debate (cf. Wellnhofer, 1974;
Whetstone, 1983; Bühler, 1985; Walker, 1985;
Martin, 1991). In the ‘Solenhofer Aktien-Verein'
specimen, however, the squamosal appears to be
well preserved and it shows a prominent ventral
process (Elzanowski & Wellnhofer, 1995).
Regardless the presence of these processes, it is
not clear whether the quadratojugal and
squamosal contacted each other, hence this char-
acter has been scored as uncertain for the
Solnhofen bird.

In contrast to the condition present in neor-
nithines, the squamosal of velociraptorines and
all other non-avian theropods has an extensive
contact with the quadratojugal (see Osborn, 1912;
Colbert & Russell, 1969; Osmólska et al., 1972;
Madsen, 1976; Weishampel et al., 1990; Currie &
Zhao, 1993a; Clark et al., 1994). This relationship
of the squamosal to the quadratojugal is in fact
the primitive amniote condition (Romer, 1956).

In Mononykus, the quadratojugal forms a rod-
like ossification, identical to the condition in
neornithine birds, as it seen in the recently col-
lected IGM 100/1001 from Ukhaa Tolgod (Dash-
zeveg et al., 1995). Furthermore, the squamosal

lacks any ventral projection. Clearly, these two
bones do not contact each other.

5) Absence of medial fossa on the proximal end
of the fibula (character 75). The medial surface
of the fibula of neornithine birds is generally flat.
This 1s the case in Hesperornithiformes and
Patagopteryx (Chiappe, 1992a). As remarked
above, the fibula is missing or poorly preserved
in most non-ornithurine birds, and the present
character is uncertain for Enantiornithes,
Iberomesornis and Archaeopteryx.

In contrast, in Deinonychus the medial face of
the proximal end of the fibula is excavated by a
shallow fossa. This medial fossa is much more
prominent in other non-avian theropods such as
ornithomimids, tyrannosaurids (Lambe, 1917),
Sinraptor (Currie & Zhao, 19932) and Allosaurus
(Madsen, 1976).

The proximal end of the fibula of Mononykus
is flat in its medial surface, with no excavation
(Fig. 18). A fragment of the fibula of Patagonykus
shows that this derived morphology was also
presentinthe Argentine taxon (Novas, this volume).

6) Absence of postorbital-jugal contact (char-
acter 95). The bird skull is characterised by
having the orbit confluent with the archosaurian
infratemporal fenestra (Zusi, 1993), a derived
feature achieved by the reduction of the postorbi-
tal-jugal bar. This derived morphology is known
to occur in Neornithes [although secondarily
modified in some lineages (e.g., Psittaciformes,

Galliformes)], Ichthyornithifor-
mes (Marsh, 1880) and Hesper-
ornithiformes (Witmer & Martin,
1987; Biihler et al., 1988) in which
the postorbital bone is absent. As
with other cranial features, the
presence or absence of a postorbi-
tal bone in Archaeopteryx is con-
troversial. Wellnhofer (1974)
regarded as a postorbital bone an
impression on the counter-slab of
the Eichstütt specimen, an iden-
tification followed by Walker
(1985) who considered several
fragments between the quadrate
and the fronto-parietal suture as
portions of the postorbital. Both
Wellnhofer (1974) and Walker
(1985) regarded the postorbital
bone to contact the dorsocaudally
projected caudal portion of the
jugal, a feature well-preserved in
the 'Solenhofer Aktien-Verein'
specimen (Wellnhofer, 1993).
Whetstone (1983), Bühler (1985)
and Martin (1991), however, con-
sidered that a postorbital bone was
absent in Archaeopteryx, and that this bone had
no contact with the jugal bar caudally confining
the orbit.

In velociraptorine theropods, as in all other
nonavian theropods (Osborn, 1912; Colbert &
Russell, 1969; Osmólska et al., 1972; Madsen,
1976; Weishampel et al., 1990), the jugal has a
robust dorsal process that contacts a ventral process
of the postorbital closing the orbit caudally.

Mononykus presents an intermediate condition
between the morphology of non-avian theropods
and that of more advanced birds such as Hesper-
ornithiformes or Neomithes. In Mononykus (MGI
100/977) the postorbital has a long, slender
ventral process, but this process does not reach
the jugal. In fact, opposite the postorbital's ventral
process, the jugal has a smooth, convex surface
and no trace of a dorsal process is present. The
orbit of Mononykus was clearly not closed caudally
but conected with the infratemporal fenestra.

DISCUSSION

This cladistic analysis supports the allocation
of Mononykus, along with Patagonykus and Al-
varezsaurus, within Aves (i.e., Avialae sensu
Gauthier, 1986). This hypothesis is supported by
four unequivocal synapomorphies of Aves

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 18. Fibula of Mononykus (A, B, MGI N 100/99; C, D, AMNH-
6570). A, D, lateral view; B, caudal view; C, medial view. IFT-tubercle
for the M. iliofibularis. A, B, scale = 0.5cm; C, D, scale = 1.0cm.

present in Mononykus (character states 19a, 71,
81 & 85a) and six unequivocal synapomorphies
(character states 13, 29, 30, 39, 44 & 69) support-
ing the monophyly of Metornithes (Perle et al.,
1993), the clade composed of the common ances-
tor of Mononykus and Neornithes, plus all its
descendants. The allocation of Mononykus within
Aves is further supported by its possession of
several other derived characters which diagnose,
although equivocally, both Aves and Metornithes
(character states 15, 46, 59a, 68, 74 & 95). The
equivocal status of these synapomorphies means
that pending determination of the condition in
Archaeopteryx, these characters might become
synapomorphies of either Aves or Metornithes.
Although the precise status of these characters is
yet unclear, they are both absent in the outgroup
and symplesiomorphic of less inclusive clades
(e.g., Ornithothoraces, Ornithurae), and ex-
amination of the data indicates that there is no
optimisation dependance for these characters.
Our initial description of Mononykus (Perle et
al., 1993), although brief, identified many of the
characters diagnosing Metornithes presented in
this paper. The avian affinity of Mononykus has
been corroborated by Novas (this volume) on the
basis of a cladistic analysis of a different data set,
but criticised by others (Patterson, 1993; Feduc-

MONONYKUS FROM THE GOBI DESERT

cia, 1994; Martin & Rinaldi, 1994; Ostrom, 1994:
Wellnhofer, 1994).

As with any hypothesis, our hypothesis of
relationship can be, and should be, tested by
others. The rationale for cladistic analysis
(described in detail by Farris, 1983 and Schoch,
1986) dictates that phylogenetic hypotheses are
tested by the distribution of characters among
taxa. Thus, to falsify our hypothesis: 1, characters
supporting an alternative relationship for
Mononykus must be identified; and 2, the weight
of the evidence must support this alternative.
Because the published criticisms do not furnish
such evidence, our disagreements with these
critics primarily concerns methodological issues
and our criterion for evidence and testability in
phylogenetic reconstruction.

Patterson’s (1993) criticism was not focused on
the phylogenetic position of Mononykus. Instead
it reflects nomenclatural issues, which Norell et
al. (1993b) subsequently addressed. Others
centered their criticisms on the peculiar forelimb
specialisations of Mononykus. For example,
Wellnhofer (1994: 306) states that ‘it would be
very difficult to imagine how a primitive bird
wing, such as that of Archaeopteryx, could have
evolved into a forelimb like that of Mononykus'.
Such an assumption lacks rigor (Chiappe et al.,
1995). If one agrees with Wellnhofer’s argument
it would be ‘very difficult to imagine’ how the
flippers of a seal evolved from the forelimb of an
ancestral carnivore. Likewise, Ostrom (1994)
uses the flightless condition of Mononykus to
claim that its keeled sternum must have evolved
convergently, as a burrowing adaptation. Indeed,
Ostrom misleadingly cites the keeled sternum as
the only evidence we provide in support of our
hypothesis, a claim belied by the figure from our
original paper reproduced in his article, with five
characters highlighted. The logic behind this ar-
gument seems to be that structures never change
function, so that if similar structures have dif-
ferent functions they cannot be homologous.
Thus, the explanation of this structure as an adap-
tation for burrowing takes precedence over the
explanation of this structure as evidence for a
close relationship between Mononykus and birds.
The fallacy of such arguments has been pointed
out many times (e.g., Gould & Vrba, 1982;
Lauder, 1994, 1995). Function and structure are
not always phylogenetically correlated (Lauder,
1995) and one wonders whether Ostrom would
consider the forelimb of Deinonychus to be non-
homologous with that of Archaeopteryx because
they have different functions.

573

In their zeal to refute our phylogenetic
hypothesis, some critics claim that the similarities
we pointed out do not exist. For example, Feduc-
cia (1994: 32) states 'the keeled breastbone
doesn't resemble that of birds, but it is very much
like that of a mole’ . While we do not deny that the
stemum of Mononykus is similar to that of some
moles, this similarity is irrelevant to comparisons
between Mononykus and other maniraptoran ar-
chosaurs unless a close relationship between
these archosaurs and this group of placental
mammals is being seriously entertained. More to
the point, the undeniable resemblance between
the sternum of Mononykus and that of other birds
cannot be ignored simply because it is at odds
with a favorite scenario of bird evolution.
Feduccia's other arguments against the avian af-
finities of Mononykus are mistaken or mislead-
ing. Within the context of his argument that birds
are unrelated to dinosaurs, he claims that
Mononykus ‘has many typical theropod dinosaur
features, including a large 'dinosaur 'tail, a small
head, and no collarbones’ (Feduccia, 1994: 32).
The presence of many 'dinosaur' features in birds
has been broadly documented (e.g., Ostrom,
19762; Gauthier, 1986; Weishampel et al., 1990).
Incidentally, none of these features are those
pointed out by Feduccia. The presence of a ‘large
‘dinosaur’ tai in Mononykus is erroneous,
whereas the significance of the size of the head
and the absence of clavicles (if not due to lack of
preservation) is unclear. As noted above (charac-
ter 19; see Characters Supporting the Monophyly
of Aves, Character Analysis), the tail of
Mononykus is like that of Archaeopteryx in being
shorter than that of non-avian theropod dinosaurs.
As for the lack of furcula, the presence of either
unfused clavicles or furculae in several groups of
theropod dinosaurs (see Barsbold, 1983; Bryant
& Russell, 1993; Russell & Dong, 1993a) indi-
cates that if they are indeed absent in Mononykus
then this absence is a derived feature of this taxon.
With regard to the skull, at the time our earlier
paper was published the only evidence from the
skull was a portion of a braincase and a bone
tentatively identified as part of the maxilla. For
these reasons our reconstruction of the skull was
clearly indicated as such. We now know, from the
recently discovered skull of specimen MGI
100/977, that the skull of Mononykus is actually
larger than what we thought (Fig. 1). However,
the suggestion that non-avian theropod dinosaurs
have smaller skulls than birds is incorrect.

Martin & Rinaldi (1994) claimed that
Mononykus is not avian but related to or-

574

nithomimid theropods (see also Martin, 1995).
Unlike most of our critics, they at least proposed
a specific hypothesis of relationships for
Mononykus. Our differences with Martin &
Rinaldi are both methodological and empirical.
These authors both misunderstand and mis-
represent cladistic techniques. In their view
‘clearly almost any outcome is possible in that
sort of an analysis [cladistic analysis]. Martin &
Rinaldi (1994) prefer to enumerate a series of
characters that supposedly differentiates
Mononykus from Ichthyornis, and which are al-
leged to be similarities between Mononykus and
ornithomimids. Methodological differences
aside, the problem with their procedure is that the
majority of the characters listed for Mononykus
are problematic or absent in this taxon. For ex-
ample, Martin & Rinaldi claimed that Mononykus
‘lacks or has few teeth’, ‘has a small head’, ‘no
furcula', ‘no free carpals’ , ‘unusually short ribs’ ,
‘a long tail’, ‘astragalus enlarged to the point
that the calcaneum is reduced to a nub or lost’,
and ‘no antitrochanter on the pelvis’. As men-
tioned earlier, Mononykus has numerous tiny
teeth and a skull which is larger than that il-
lustrated in our original reconstruction, a tail
shorter than any non-avian theropod, and a well-
developed antitrochanter. The furcula (see above)
and free carpals are not necessarily absent but not
preserved in the known specimens. In fact, facets
on the proximal articular surface of the car-
pometacarpus indicate that free carpals are miss-
ing (Perle et al., 1994).

Likewise, the astragalus and calcanum of
Mononykus are completely fused (Perle et al.,
1994), preventing estimation of their relative
sizes. The new specimen MGI 100/977 shows
that the ribs are not particularly short. Obviously
Martin & Rinaldi have based their comments on
our original reconstruction (Perle et al., 1993:
Fig. 2) despite the fact that the ribs were indicated
(along with most of the skull) as unpreserved
elements. The remaining characters used by Mar-
tin & Rinaldi (1994) are either primitive (e.g.,
elongated haemal arches, preacetabular ilium
short), autapomorphic (e.g., ischium slender,
reduced forelimbs, enlarged metacarpal 1), or so
highly variable as to be of uncertain generality
(e.g., postacetabular ilium elongate, long neck).
Of all the characters listed by Martin & Rinaldi,
only the ‘enlarged metacarpal I (alular metacar-
pal in this paper) is derived and shared by or-
nithomimids. Nevertheless, the morphology of
this metacarpal in Mononykus is distinctly dif-
ferent than that of ornithomimids. In or-

MEMOIRS OF THE QUEENSLAND MUSEUM

nithomimids, metacarpal I is more than half the
length of metacarpals II and III (Barsbold &
Osmólska, 1990); in advanced ornithomimids all
the metacarpals are nearly equivalent in length
(e.g., Gallimimus, Ornithomimus). Aside from
length, however, all known ornithomimids have
metacarpals that are proportionally similar. This
is not true in Mononykus where the alular
metacarpal (equivalent to metacarpal I) is sig-
nificantly larger than the major and minor
metacarpals (metacarpals II and III, respectively)
(Fig. 14). This condition (an alular metacarpal
that is larger than the remaining metacarpals) is
synapomorphic of a group containing Mononykus
and Patagonykus, and it is unrelated to the con-
dition seen in ornithomimids. By rejecting our
phylogenetic hypothesis, our critics have im-
plicitly indicated that the evidence used to in-
clude Mononykus within Aves (in addition to the
evidence presented here) is nonhomologous (i.e.,
homoplastic). These authors, however, have
failed in providing logical support for such a
claim. As extensively discussed elsewhere (e.g.,
Patterson, 1982; Rieppel, 1992, 1994; Pinna,
1991; Hall, 1994), homology is a two-statement
concept ultimately based on the congruence of
characters. Primary homology refers to state-
ments about the similarity of characters prior to
phylogenetic constructions, while secondary
homology refers to interpretations of common
origins through character congruence on a par-
ticular tree (Pinna, 1991); the largest number of
characters which congruently support a specific
hypothesis of relationships are considered as
homologies. Therefore, a statement of nonhomol-
Ogy — as used by our critics — cannot be derived
from a ‘priori’ observation, but it should be ul-
timately revealed by the mismatch between
primary and secondary homology. In order to say,
for example, that the carinate sternum or the
absence of a jugal-postorbital contact shared by
Mononykus and more advanced birds is not
homologous, it is necessary to have an alternative
hypothesis in which the ‘congruence’ between
the distribution of characters supporting that
hypothesis is maximised.

In using a cladogram to determine evolutionary
relationships, one need only assume that the
hierarchical distribution of characters reflects the
evolutionary relationships of taxa. The possible
function of a particular structure, or a scenario
about how a particular structure or function arose
are not relevant. Instead, the relationships iden-
tified by cladistic analysis are the ones that pro-
vide the framework for testing functional or

MONONYKUS FROM THE GOBI DESERT

adaptational hypotheses (Coddington, 1988;
Gatesy, 1995; Witmer, 1995) and to use these
scenarios as evidence for or against a phylo-
genetic hypothesis confuses the phenomenon to
be explained with the explanation for the
phenomenon (Brady, 1985).

The avian affinity of Mononykus is supported
by the present study and corroborated inde-
pendently (Novas, this volume). Re-rooting this
cladogram so that Alvarezsauridae falls outside
Aves requires a number of additional steps. Four
and six additional steps are required for the plac-
ing Alvarezsauridae either as the sister-group of
Aves or the clade formed by Aves plus Velocirap-
torinae, respectively. As for any phylogenetic
hypothesis, the addition of characters, taxa, or
both to the data set may change the topology of
the resultant cladogram. It is hoped that the dis-
covery of more fossil taxa and better specimens
(in particular representatives of the lineage lead-
ing to Mononykus, Patagonykus and Alvarez-
saurus) will provided additional character
information to further test this cladogram. Fur-
thermore, an alternative outgroup hypothesis
may modify our result. Several derived characters
shared by Mononykus and other birds appear in
some troodontid taxa but are absent in velocirap-
torines (e.g., characters 44 & 81). These
similarities may suggest that troodontids are
closer to birds than velociraptorines are; however
this needs to be evaluated in light ofa larger, more
inclusive analysis. Our choice for velocirap-
torines instead of troodontids is based on the fact
that the latter are poorly known and their selection
as an outgroup would have added numerous miss-
ing entries to the data set.

CONCLUSIONS

The cladistic analysis presented here supports
the inclusion of Mononykus and the Alvarez-
sauridae, within Aves. The Alvarezauridae is the
sister-group of Ornithothoraces and is closer to
Neomithes than is Archaeopteryx.

The alternative, that Mononykus is not a mem-
ber of Aves, is based on misleading evolutionary
assumptions and ‘a priori’ speculation. Falsifica-
tion of our hypothesis requires that a new
hypothesis better summarising the pattern of
similarities shared by Mononykus and birds be
proposed. Or, with the addition of new characters
or taxa, our evidence for relationship (synapo-
morphy) is contradicted. What this means is that
it is insufficient to authoritatively proclaim that
Mononykus is not avian (or not a bird), without

575

proposing an alternative hypothesis of what it is
related to (supported by character evidence).

Several years ago evolutionary biologists ar-
rived at a consensus that hypotheses of
phylogenetic relationships should be based on the
simplest explanation for the distribution of char-
acters shared among organisms. It is unfortunate
that our critics support their claims with argu-
ments that lie outside the lines of modern sys-
tematics and comparative biology.

ACKNOWLEDGEMENTS

We are especially grateful to F. Novas (Museo
Argentino de Ciencias Naturales) and R.E. Mol-
nar (Queensland Museum) for their invitation to
participate in this Symposium. F. Novas kindly
provided data on unpublished specimens and
financial assistance to L. Chiappe for attending
the Symposium. We also thank F. Novas and L.
Witmer for reviewing the original manuscript.
The illustrations were prepared by M. Ellison
(AMNH). This research has been supported by
the Frick Fund of the AMNH and the NSF (DEB-
9407999).

REFERENCES

ALVARENGA, H.M.F. & BONAPARTE, J.F. 1992. A
new flightless land bird from the Cretaceous of
Patagonia. Pp. 5164. Proceedings of the II Inter-
national Symposium of Avialian Paleontology,
Los Angeles.

AX, P. 1987. ‘The Phylogenetic System’. (Wiley &
Sons: Chichester). 340pp.

BARSBOLD, R. 1974. Saurornithoididae, a new family
of small theropod dinosaurs from central Asia and
North America. Results of the Polish-Mongolian
paleo ntological expedition, Part V. Paleontologia
Polonica 30: 522.

1983. Carnivorous dinosaurs form the Late
Cretaceous of Mongolia. Trudy Paleon-
tologicheskogo Instituta Akademii Nauk SSSR
19: 1117. (In Russian).

BARSBOLD, R. & OSMOLSKA, H. 1990. Ornitho-
mimosauria. Pp. 225-244. In Weishampel, D.B.,
Dodson, P. & Osmólska, H. (eds.) 'The
Dinosauria'. (University of California Press:
Berkeley).

BAUMEL, J.J. & WITMER, L.M. 1993. Osteologia. In
Baumel, J.J., King, A.S., Breazile, J.E., Evans,
H.E.& Vanden Berge, J.C. (eds.) ‘Handbook of
Avian Anatomy: Nomina Anatomica Avium’,
second edition. Publications of the Nuttal Or-
nithological Club 23: 45-132.

BELLAIRS A. D'A. & JENKIN, C.R. 1960. The
skeleton of birds. Pp. 241-300. In Marshall, A.J.
(ed.) ‘Biology and comparative physiology of
birds’. (Wiley and Sons: New York).

576

BONAPARTE, J.F. 1991. Los vertebrados fósiles de la
Formación Río Colorado de Neuquén y cercanias,
Cretácico Superior, Argentina. Revista del Museo
Argentino de Ciencias Naturales *Bernardino
Rivadavia' (Paleontología) 4(3): 17-123.

BONAPARTE, J.F, NOVAS, FE. & CORIA, R.A.
1990. Carnotaurus sastrei Bonaparte, the horned,
lightly built Carnosaur from the Middle
Cretaceous of Patagonia, Contributions in Science
416: 142. Natural History Museum of Los Angeles
County.

BRADY, R.H. 1985. On the independence of cladistics.
Cladistics 1: 113-126.

BRETT-SURMAN, M.K. & PAUL, G.S. 1985. A new
family of bird-like dinosaurs linking Laurasia and
Gondwanaland. Journal of Vertebrate Paleontol-
ogy 5(2): 133-138.

BRYANT, H.N. & RUSSELL, A.P. 1993. The occur-
rence of clavicles within Dinosauria: implications
for the homology of the avian furcula and the
utility of negative evidence, Journal of Vertebrate

Paleontology 13(2): 171-184.

BUHLER, P. 1985. On the morphology of the skull of
Archaeopteryx. Pp.135-140. In Hecht, M.K.,
Ostrom, J.H., Viohl, G. & Wellnhofer, P. (eds.)
"The beginnings of Birds’. Proceedings of the
International Archaeopteryx Conference,

. Eichstatt.

BUHLER, P., MARTIN, L.D. & WITMER, L.M. 1988.
Cranial kinesis in the Late Cretaceous birds
Hesperornis and Parahesperornis. Auk 105: 111-
122.

CARPENTER, J.M. 1988. Choosing among multiple
equally parsimonious cladograms. Cladistics 4:
291-296.

CHATTERJEE, S. 1991. Cranial anatomy and relation-
ships of a new Triassic bird from Texas.
Philosophical Transactions of the Royal Society
of London, Series B, 332(1265): 277-346.

CHIAPPE, L.M. 1991. Cretaceous birds of Latin
America. Cretaceous Research 12: 55.

1992a. Osteología y sistemática de Patagopteryx
deferrariisi Alvarenga y Bonaparte (Aves) del
Cretácico de Patagonia. Filogenia e historia
biogeográfica de las aves Cretácicas de América
del Sur. Tesis Doctoral, Facultad de Ciencias
Exactas y Naturales, Universidad de Buenos
Aires, 429pp.

1992b. Enantiornithine tarsometatarsi and the avian
affinity of the Late Cretaceous Avisauridac. Jour-
nal of Vertebrate Paleontology 12(3): 344-350.

1993. Enantiornithine (Aves) tarsometatarsi from
the Cretaceous Lecho Formation of northwestern
Argentina. American Museum Novitates 3083:
1-27.

1995a. A diversity of early birds. Natural History
104(6): 52-55.

1995b. The first 85 million years of avian evolution.
Nature 378: 349-355.

1996. Late Cretaceous birds of southern South
America: anatomy and systematics of Patagop-

MEMOIRS OF THE QUEENSLAND MUSEUM

teryx and Enantiornithes. In Arratia, G. (ed.) Con-
tributions of southern South America to ver-
tebrate paleontology. Münchner
Geowissenschaftliche Abhandlungen (A)30:
203-244. (Verlag Dr Pfeil, München).

1996, this volume. Early avian evolution in the
southern hemisphere: the fossil record of birds in
the Mesozoic of Gondwana. Memoirs of the
Queensland Museum 39(3): 533-554.

CHIAPPE, L.M. & CALVO, J.O. 1994. Neuquenornis
volans, a new Enantiornithes (Aves) from the
Upper Cretaceous of Patagonia (Argentina). Jour-
nal of Vertebrate Paleontology 14(2): 230-246.

CHIAPPE, L.M., NORELL, M.A. & CLARK, J.M.
1995. Comment on Wellnhofer's new data on the
origin and early evolution of birds. C. R. Acad.
Sci. Paris 320(11): 1031-1032.

CLARK, J.M., PERLE, A. & NORELL, M.A. 1994,
The skull of Erlicosaurus andrewsi, a Late
Cretaceous ‘Segnosaur’ (Theropoda:
Therizinosauridae) from Mongolia. American
Museum Novitates 3115; 1-39.

CODDINGTON, J.A. 1988, Cladistic tests of adapta-
tional hypotheses. Cladistics 4: 3-22.

COLBERT, E.H. & RUSSELL, D.A. 1969. The small
dinosaur Dromaeosaurus. American Museum
Novitates 2380: 1-49.

CURRIE, P.J. 1995. New information on the anatomy
and relationships of Dromaeosaurus albertensis
(Dinosauria, Theropoda). Journal of Vertebrate
Paleontology 15(3): 576-591.

CURRIE, P.J., RIGBY, J.K. Jr & SLOAN, R.E. 1990.
Theropod teeth from the Judith River Formation
of southem Alberta, Canada. Pp.107-125. In Car-
penter, K. & Currie, PJ. (eds.) “Dinosaur Sys-
tematics: Approaches and Perspectives'.
(Cambridge Univ. Press).

CURRIE PJ. & ZHAO, X-J. 1993a. A new carnosaur
(Dinosauria, Theropoda) from the Jurassic of Xin-
jiang, People's Republic of China, Canadian Jour-
nal of Earth Sciences 30(10-11): 2037-2081.

1993b. A new troodontid (Dinosauria, Theropoda)
braincase from Dinosaur Park Formation (Cam-
panian) of Alberta. Canadian Journal of Earth
Sciences 30(10-11): 2231-2247.

DASHZEVEG D., NOVACEK, M.J., NORELL, M.A.,
CLARK, J.M., CHIAPPE, L.M., DAVIDSON, A.,
MCKENNA, M.C., DINGUS, L., SWISHER, C.
& PERLE A. 1995. Extraordinary preservation in
a new vertebrate assemblage from the Late
Cretaceous of Mongolia. Nature 374: 446-449,

De BEER, G. 1954. Archaeopteryx lithographica. A
study based upon the British Museum specimen.
Publication of the British Museum (Natural His-
tory), No 224: 168. London.

ELZANOWSKI, A. 1991. New observations on the
skull of Hesperornis with reconstructions of the
bony palate and otic region. Postilla 207: 120.

ELZANOWSKI, A. & WELLNHOFER, P. 1992. Anew
link between theropods and birds from the
Cretaceous of Mongolia. Nature 359: 821-823.

MONONYKUS FROM THE GOBI DESERT

1995. The skull of Archaeopteryx and the origin of
birds. Archaeopteryx 13: 41-46.

FARRIS, J. 1983. The logical basis of phylogenetic
analysis. Pp. 136, Vol. 2. In Platnick, N.I. & Funk,
V.A. (eds.) ‘Advances in cladistics'. (Columbia
Univ. Press: New York).

1988. Hennig 86 references. Documentation for ver-
sion 1.5. Privately published.

FEDUCCIA, A. 1994, The Great Dinosaur Debate.
Living Bird 13(4): 28-33.

FIORILLO, A. & CURRIE, P. 1994, Theropod teeth
from the Judith River Formation (Upper
Cretaceous) of south-central Montana. Journal of
Vertebrate Paleontology 14(1): 74-80.

GATESY, S.M. 1995. Functional evolution of the
hindlimb and tail from basal theropods to birds.
Pp. 219-234. In Thomason, J. (ed.) ‘Functional
morphology in vertebrate paleontology’.
(Cambridge Univ. Press).

GAUTHIER, J. 1986. Saurischian monophyly and the
origin of birds. In Padian, K. (ed.) ‘The origin of
birds and the evolution of flight’. Memoirs of the
California Academy of Sciences 8: 1-55.

GOULD, S.J. & VRBA, E.S. 1982. Exaptation — a
missing term in the science of form. Paleobiology
8(1): 4-15.

HALL, B.K. 1994, ‘Homology, structure, and function’.
(Academic Press: San Diego). 483pp.

HENNIG, W. 1969. ‘Die stammesgeschichte der
Insekten’. (E. Kramer: Frankfurt am Main). 436 pp.

1981. ‘Insect Phylogeny’. (John Wiley & Sons: New
York).

HOLTZ, T. 1994a. The arctometatarsalian pes, an un-
usual structure of the metatarsus of Cretaceous
Theropoda (Dinosauria: Saurischia). Journal of
Vertebrate Paleontology 14(4); 480-519.

1994b. The phylogenetic position of the Tyran-
nosauridae: implications for theropod sys-
tematics. Journal of Paleontology 68(5):
1100-1117.

HOU L., ZHOU Z., MARTIN, L.D. & FEDUCCIA, A.
1995. A beaked bird from the Jurassic of China.
Nature 377: 616-618.

JOLLIE, M.T. 1957. The head skeleton of the chicken
and remarks on the anatomy of this region in other
birds. Journal of Morphology 100(3): 389-436.

KING, A.S. & McLELLAND, J. 1984. Pp. 334. ‘Birds,
their Structure and Function’ (2nd ed.). (Bailliere
Tindall: London).

KUROCHKIN, E.N. 1985. A true carinate bird from
Lower Cretaceous deposits in Mongolia and other
evidence of early Cretaceous birds in Asia.
Cretaceous Research 6: 271-278.

LAMB, J.P., Jr, CHIAPPE, L.M. & ERICSON, P.G.P.
1993. A marine enantiornithine from the
Cretaceous of Alabama. Journal of Vertebrate
Paleontology 13(3), Sup.: 45A.

LAMBE, L.M. 1917. The Cretaceous theropodous
dinosaur Gorgosaurus. Memoirs of the Geologi-
cal Survey of Canada 100: 1-84.

577

LAUDER, G.V. 1994. Homology, structure, and func-
tion. Pp. 151-196, In Hall, B. (ed.) ‘Homology: the
hierarchical basis of comparative biology’.
(Academic Press: San Diego).

1995. On the influence of function from structure.
Pp.118. In Thomason, J. (ed.) ‘Functional mor-
phology in vertebrate paleontology’. (Cambridge
Univ. Press).

MADSEN, J.A. 1976, Allosaurus fragilis. A revised
osteology. Bulletin of the Utah Geological and
Mineral Survey 109: 1-163.

MARSH, O.C. 1880. Odontornithes: a monograph on
the extinct toothed birds of North America. United
States Geological Exploration of the 40th Parallel.
(Government Printing Office: Washington). 201

PP.

MARTIN, L.D.1980. Foot-propelled diving birds of the
Mesozoic. 17 Congressus Internationalis Or-
nithologici, Actas 2: 237-242. Berlin.

1983. The origin and early radiation of birds. Pp.
291-338. In Bush, A.H. & Clark, G.A. (eds.)
“Perspectives in ornithology’. (Cambridge
University Press).

1984. A new Hesperomithid and the relationships of
the Mesozoic birds. Transactions of the Kansas
Academy of Science 87(34): 141-150.

1985, The relationships of Archaeopteryx to other
birds. Pp.177-183. In Hecht, M.K., Ostrom, J.H.,
Viohl, G. & Wellnhofer, P. (eds.) "The beginnings
of birds’. Proceedings of the International Ar-
chaeopteryx Conference, Eichstátt.

1991. Mesozoic birds and the origin of birds, Pp.
485-539. In Schultze, H.P. & Trueb, L. (eds.),
Origins of the higher groups of tetrapods: con-
troversy and consensus. (Comstock Publ. As-
sociates: Ithaca).

1995, The relationship of Mononykus to or-
nithomimid dinosaurs. Journal of Vertebrate
Paleontology 15(3) (Supp.): 43A.

MARTIN, L.D. & TATE, J. Jr 1976. The skeleton of
Baptornis advenus (Aves: Hesperornithiformes).
In Olson, S.L. (ed.) Collected papers in avian
paleontology honoring the 90th birthday of
Alexander Wetmore. Smithsonian Contributions
to Paleobiology 27: 35-66.

MARTIN, L.D. & RINALDI, C. 1994, How to tell a bird
from a dinosaur. Maps Digest 17(4): 190-196.

MARTIN, L.D., STEWART, J.D. & WHETSTONE,
K.N. 1980. The origin of birds: structure of the
tarsus and teeth. Auk 97: 86-93.

MEYER, H. von 1861. Archaeopteryx lithographica
(Vogel Feder) und Prercdactylus von Solenhofen.
N. Jb. Min. Geol. Paláont.: 678-679. Stuttgart.

MOLNAR, R.E. 1986, An enantiornithine bird from the
Lower Cretaceous of Queensland, Australia. Na-
ture 322: 736-738.

NORELL, M.A., CLARK, J.M. & PERLE, A. 1992.
New dromaeosaur material from the Late
Cretaceous of Mongolia. Journal of Vertebrate
Paleontology 12(3), Sup.: 45A.

578

NORELL, M.A., CHIAPPE, L.M. & CLARK, J. 1993a.
New limb on the avian family tree. Natural History
102(9): 38-43.

NORELL, M.A., CLARK, J.M. & CHIAPPE, L.M.
1993b. Naming names. Nature 366: 518.

NORELL, M.A., CLARK, J.M., DASHZEVEG D.,
BARSBOLD R., CHIAPPE, L.M., DAVIDSON,
A.R., McKENNA, M.C. & NOVACEK, MJ.
1994. A Theropod Dinosaur Embryo, and the Af-
finities of the Flaming Cliffs Dinosaur Eggs.
Science 266: 779-782.

NOVACEK, M., NORELL, M.A., McKENNA, M. &
CLARK, J.M. 1994. Fossils from the Flaming
Cliffs. Scientifc American 271 (6): 36-43.

NOVAS, FE. 1992. La evolución de los dinosaurios
carnívoros. Segundo Curso de Paleontologia en
Cuenca, Actas, pp. 125-163. Cuenca: Instituto
‘Juan de Valdes’.

1996, this volume. Alvarezsauridae, Cretaceous
maniraptorans from Patagonia and Mongolia.
Memoirs of the Queensland Museum 39(3):675-
702.

In press. Anatomy of Patagonykus puertai
(Theropoda, Maniraptora, Alvarezsauridae) from
the Late Cretaceous of Patagonia. Journal of Ver-
tebrate Paleontology.

OSBORN, H.F. 1912. Crania of Tyrannosaurus and
Allosaurus. Memoirs of the American Museum of
Natural History, Part 1: 130.

1916. Skeletal adaptations of Ornitholestes,
Struthiomimus, Tyrannosaurus. Bulletin of the
American Museum of Natural History 35: 733-771.

1924. Three new Theropoda, Protoceratops zone,
central Mongolia. American Museum Novitates
144: 1-12.

OSMOLSKA, H., RONIEWICZ, E. & BARSBOLD,
R. 1972. A new dinosaur, Gallimimus bullatus n.
gen. n. sp. (Ornithomimidae) from the Upper
Cretaceous of Mongolia. Palaeontologia Polonica
27: 103-143.

OSTROM, J.H. 1969. Osteology of Deinonychus antir-
rhopus, an unusual theropod dinosaur from the
Lower Cretaceous of Montana. Bulletin of the
Peabody Museum of Natural History 30: 1-165.

1974. The pectoral girdle and forelimb function of
Deinonychus (Reptilia: Saurischia): A correction.
Postilla 165: 1-11.

1976a. Archaeopteryx and the origin of birds. Bio-
logical Journal of the Linnean Society 8: 91-182.

1976b. On a new specimen of the lower Cretaceous
theropod dinosaur Deinonychus antirrhopus.
Breviora 439: 1-21.

1978. The osteology of Compsognathus longipes
Wagner. Zitteliana 4: 73-118.

1990. Dromaeosauridae. Pp. 269-279. In Weisham-
pel, D.B., Dodson, P. & Osmólska, H. (eds.) ‘The
Dinosauria'. (University of California Press:
Berkeley).

1991. The question of the origin of birds. Pp.467-
484. In Schultze, H.P. & Trueb, L. (eds.) 'Origins

MEMOIRS OF THE QUEENSLAND MUSEUM

of the higher groups of tetrapods: controversy and
consensus’. (Comstock Publ. Associates: Ithaca).

1994. On the origin of birds and of avian flight. In
Prothero, D.R. & Schoch, R.M. (conv.) Major
Features of Vertebrate Evolution. Short Courses
in Paleontology 7: 160-177.

PATTERSON, C. 1982. Morphological characters and
homology. In Joysey, K.A. & Friday, A.E. (eds.)
Problems of phylogenetic reconstruction. Sys-
tematics Association Special Volume 21: 21-74.
(Academic Press: London).

1993. Bird or dinosaur? Nature 365: 21-22.

PERLE, A., NORELL, M.A., CHIAPPE, L.M. &
CLARK, J.M. 1993. Flightless bird from the
Cretaceous of Mongolia. Nature 362; 623-626.

PERLE A., CHIAPPE, L.M., BARSBOLD, R.,
CLARK, J.M. & NORELL, M.A. 1994. Skeletal
morphology of Mononykus olecranus
(Theropoda: Avialae) from the Late Cretaceous of
Mongolia. American Museum Novitates 3105:
1-29.

PINNA, M. 1991. Concepts and tests of homology in
the cladistic paradigm. Cladistics 7: 367-394.

QUEIROZ, K. De & GAUTHIER, J. 1992.
Phylogenetic taxonomy. Annual Review of Ecol-
ogy and Systematics 23: 449-480.

RIEPPEL, O. 1992. Homology and logical fallacy. Jour-
nal of evolutionary biology 5: 701-715.

1994. The role of paleontological data in testing
homology by congruence. Acta Paleontologica
Polonica 38(314): 295-302.

ROMER, A.S. 1956. ‘Osteology of the Reptiles’.
(University of Chicago Press: Chicago). 772 pp.

ROWE, T. & GAUTHIER, J. 1992, Ancestry, paleon-
tology, and definition of the name Mammalia,
Systematic Biology 41(3): 372-378.

RUSSELL, D.A. & DONG, Z-M. 1993a. A nearly
complete skeleton of a new troodontid dinosaur
from the Early Cretaceous of the Ordos Basin,
Inner Mongolia, People’s Republic of China.
Canadian Journal of Earth Sciences 30(10-11):
2163-2173.

1993b. The affinities of a new theropod from the
Alxa Desert, Inner Mongolia, People’s Republic
of China. Canadian Journal of Earth Sciences
30( 10-11): 2107-2127.

SANZ, J.L., BONAPARTE, J.F. & LACASA RUIZ, A.
1988. Unusual Early Cretaceous birds from Spain,
Nature 331: 433-435.

SANZ, J.L. & BONAPARTE, J.F. 1992, Iberomesornis
romerali, a fossil small bird articulated skeleton
from the early Cretaceous of Spain. Proceedings
of the II International Symposium of Avialian
Paleontology, Los Angeles. Pp. 39-49.

SANZ, J.L. & BUSCALIONI, A.D. 1992. A new bird
from the Early Cretaceous of Las Hoyas, Spain,
and the early radiation of birds. Paleontology
35(4): 829-845.

SANZ, J.L., CHIAPPE, L.M. & BUSCALIONI, A.D.
1995. The osteology of Concornis lacustris (Aves:
Enantiomithes) from the Lower Cretaceous of

MONONYKUS FROM THE GOBI DESERT

Spain, and a re-examination of its phylogenetic
relationships. American Museum Novitates 3 133:
1-23

SCHOCH, R.M. 1986. ‘Phylogeny reconstruction in
paleontology’. (Van Nostrand Reinhold Co.: New
York). 353 pp.

SERENO, P. & RAO C.1992. Early evolution of avian
flight and perching: new evidence from Lower
Cretaceous of China. Science 255: 845-848.

VARRICCHIO, D.J. & CHIAPPE, L.M. 1995. A new
bird from the Cretaceous Two Medicine Forma-
tion of Montana. Journal of Vertebrate Paleontol-
ogy 15(1): 201-204.

VERHEYEN, R. 1960. Considerations sur la colonne
vertebrale des oiseaux (non-passeres). Bulletin
Institut Royal des Sciences Naturelles de Belgique
36(42): 1-24.

WALKER, A.D. 1985. The braincase of Archaeopteryx.
In Hecht, M.K., Ostrom, J.H., Viohl, G. &
Wellnhofer, P. (eds.) The Beginnings of Birds.
Proceedings of the International Archaeopteryx
Conference, Eichstatt. Pp.123-134.

WALKER, C.A. 1981. New subclass of birds from the
Cretaceous of South America. Nature 292: 51-53.

WEBB, M. 1957. The ontogeny of the cranial bones,
cranial peripheral and cranial parasympathetic
nerves, together with a study of the visceral
muscles of Struthio. Acta Zoologica 38: 81-203.

WEISHAMPEL, D.B., DODSON, P. & OSMOLSKA,
H. 1990. ‘The Dinosauria'. (Univ. of California
Press: Berkeley). 733 pp.

WELLES, S.P. 1984. Dilophosaurus wetherilli
(Dinosauria, Theropoda) osteology and com-
parisons. Palaeontographica A, 185: 85-180.

WELLNHOFER, P. 1974. Das fünfte Skelettexemplar
von Archaeopteryx. Palaeontographica A, 147:
169-216.

1985. Remarks on the digit and pubis problems of
Archaeopteryx. In Hecht, M.K., Ostrom, J.H.,
Viohl, G. & Wellnhofer, P. (eds.) ‘The Beginnings
of Birds’. Proceedings of the International Ar-
chaeopteryx Conference, Eichstátt. Pp.113-122.

1992. A new specimen of Archaeopteryx from the
Solnhofen limestone. Proceedings of the II Inter-
national Symposium of Avialian Paleontology,
Los Angeles. Pp. 3-23.

1993. Das siebte Exemplar von Archaeopteryx aus
den Solnhofener Schichten. Archaeopteryx 11:
1-48.

1994. New data on the origin and early evolution of
birds. C. R. Acad. Sci. Paris 319(11): 299-308.

WHETSTONE, K.N. 1983. Braincase of Mesozoic
birds: I. New preparation of the "London" Ar-
chaeopteryx. Journal of Vertebrate Paleontology
2(4): 439-452.

WILEY, E.O., SIEGEL-CAUSEY, D., BROOKS, D.R.
& FUNK, V.A. 1991. The compleat cladist. The
Univ. Kansas, Special Publ. 19: 1-158.

WITMER, L.M. 1990. The craniofacial air sac system
of Mesozoic birds (Aves). Zoological Journal of
the Linnean Society 100: 327-378.

579

1995, The extant phylogentic bracket and the impor-
tance of reconstructing soft tissues in fossils. In
Thomason, J. (ed.) ‘Functional morphology in
vertebrate paleontology'. (Cambridge Univ.
Press: Cambridge). Pp. 1933.

WITMER, L.M. & MARTIN, L.D. 1987. The primitive
features of the avian palate, with special reference
to Mesozoic birds. In Mourer-Chauviré, C.
(coord.) L'evolution des oiseaux d'apres le
temoignage des fossiles. Documents des
Laboratoires de Geologie Lyon 99: 21-40.

WITMER, L.M. & WEISHAMPEL, D.B.1993.
Remains of theropod dinosaurs from the Upper
Cretaceous St Mary River Formation of
northwestern Montana, with special reference to a
new maniraptoran braincase. Journal of Vertebrate
Paleontology 13(3), Suppl.: 63A.

ZHAO X-J. & CURRIE, P.J. 1993. A large crested
theropod from the Jurassic of Xinjiang, People’s
Republic of China. Canadian Journal of Earth
Sciences 30(10-11): 2027-2036.

ZHOU Z., JIN F. & ZHANG J-Y. 1992. Preliminary
report on a Mesozoic bird from Liaoning, China.
Chinese Science Bulletin 37(16): 1365-1368.

ZUSI, R.L.1962. Structural adaptations of the head and
neck in the black skimmer. Publications of the
Nuttal Ornithological Club 3: 1-101.

1993, Patterns of diversity in the avian skull. In
Hanken, J. & Hall, B.K. (eds) “The Skull’ 2: 391-
437, (University of Chicago Press: Chicago).

APPENDIX 1

Skeletal characters and data-matrix used in the
cladistic analysis. Multistate characters have
been considered as additive except for characters
34, 40, 64, and 92 in which any two distinct states
are separated by a single step. In the text and Fig.
2, ‘a’ and ‘b’ refer to states 1 and 2, respectively,
of a multistate character. Scoring: 0 = primitive;
1,2 = derived; ? = missing or uncertain.

1. Premaxillae in adults. 0. unfused. Z. fused.

2. Maxillary process of premaxillary. 0.
restricted to the rostral portion. 7. forming most
of the facial margin.

3. Frontal process of premaxilla. 0. short. 7.
long, extending caudally to the level of lacrimals.

4. Premaxillary teeth. 0. present. 7. absent.

5. Osseous external naris. 0. smaller, or 7. con-
siderably larger than antorbital fenestra.

6. Maxillary fenestrae. 0. present. J. absent.

7. Rostral jugal border. 0. away. 1. very close to
the caudal margin of the osseous external naris.

8. Ectopterygoid. 0. present. J. absent.

9. Quadrate orbital process (pterygoid ramus).
0. broad. J. sharp and point-shaped.

580

10. Quadratojugal. 0. sutured to the quadrate.
1. articulating in a cotyle in the lateral face of the
quadrate mandibular process.

11. Quadrate pneumaticity. 0. absent. J.
present.

12. Articular pneumaticity. 0. absent. 1. present.

13. Prominent ventral processes on cervico-
dorsal vertebrae. 0. absent. 7. present.

14. Dorsal vertebral count. 0. 13-14. 1. fewer
than 13. 2. fewer than 11.

15. Wide vertebral foramen in thoracic ver-
tebrae, vertebral foramen/cranial articular facet
ratio (vertical diameter) larger than 0.40. 0. ab-
sent. J. present.

16. Synsacrum formed by 0. less. 7. more than
8 vertebrae.

17. Heterocoelous cervical vertebrae. 0. absent

I. present.

18. Pygostyle. 0. absent Z. present.

19. Caudal vertebral count. 0. more than 35. 1.
fewer than 25-26. 2. fewer than 15.

20. Ossified uncinate processes. 0. absent 1.
present.

21. Scapula and coracoid articulation. 0.
through a wide, sutured articulation. 7. through
more localised facets.

22. Procoracoid process. 0. absent. 7. present.

23. Coracoid shape. 0. short. 7. elongated with
subrectangular profile. 2. strut-like.

24. Scapulocoracoid articulation. 0. at the
shoulder (proximal) end of coracoid. /. well
below to it.

25. Supracoracoid nerve foramen. 0. centrally
located. J. situated (often as an incision) in the
medial margin of coracoid.

26. Coracoid and scapula. 0. placed in the same
plane. J. forming a sharp angle at the level of the
glenoid cavity.

27. Scapular caudal end. 0. blunt. 1. sharp.

28. Scapular shaft. 0. straight. Z. sagittally
curved.

29. Stemum. 0. subquadrangular to transversal-
ly rectangular. 1. longitudinally rectangular.

30. Ossified sternal keel. 0. absent. 7. present.

31. Proximal and distal humeral ends. 0. twisted

1. expanded nearly in the same plane.

32. Ulna. 0. shorter, I. longer or nearly
equivalent to humerus.

33. Humeral head. 0. concave cranially and
convex caudally. 7. globe shaped, craniocaudally
convex.

34. Ventral tubercle of humerus. 0. projected
ventrally. 7. proximally 2. or caudally, separated
from the humeral head by a deep capital incision.

MEMOIRS OF THE QUEENSLAND MUSEUM

35. Humerus with well developed transverse
ligamental groove. 0. absent. J. present.

36. Humeral distal condyle location. 0. mainly
on distal aspect. 1. cranial aspect.

37. Semilunate ridge on ulnar dorsal condyle.
0. absent. 7. present.

38. Ulnar shaft considerably thicker than radial
shaft; radial shaft/ulnar shaft ratio. 0. larger 1.
smaller than 0.70.

39. Distal carpals and metacarpals. 0. unfused.
I. fused forming a carpometacarpus.

40. Extensor process on carpometacarpus. 0.
absent. 7. round shaped. 2. with a sharp point.

41. Pelvic elements. 0. unfused. 7. fused or
partially fused.

42. Small acetabulum; acetabulum/ilium ratio
equal or smaller than 0.11. 0. absent. 1. present.

43. Pubis more or less parallel to ilium and
ischium. 0. absent. J. present.

44. Prominent antitrochanter. 0. absent. 1.
present.

45. Iliac fossa for M. cuppedicus (= M.
iliofemoralis internus). 0. present. 7. absent.

46. Ischiadic terminal processes. 0. in contact.
1. lacking contact.

47. Pubic apex. 0. in contact. 1. lacking contact.

48. Pubis shaft laterally compressed throughout
its length. 0. absent. 7. present.

49. Pubic foot. 0. present. J. absent.

50. Femur with distinct fossa for capital liga-
ment. 0. absent. 7. present.

51. Femoral anterior trochanter. 0. nearly con-
fluent with the greater trochanter. 7. or fused to it
forming the trochanteric crest.

52. Femur with prominent patellar groove. 0.
absent. J. present.

53. Femoral popliteal fossa distally bounded by
a complete transverse ridge. 0. absent. 1. present.

54. Tibiofibular crest in the lateral condyle of
femur. 0. absent. J. poorly developed. 2.
prominent.

55. Femoral posterior trochanter. 0. present. J.
absent.

56. Tibia, calcaneum and astragalus. 0. unfused
or poorly coossified (sutures still visible). 7. com-
plete calcaneo-astragalar-tibial fusion.

57. Cranial cnemial crest on tibiotarsus. 0. ab-
sent. 7. present.

58. Extensor canal on tibiotarsus. 0. absent. 7.
present.

59. Fibula with tubercle directed, for M. iliof-
bularis. 0. anterolaterally. 7. laterally. 2.
caudolaterally or caudally.

MONONYKUS FROM THE GOBI DESERT 581

[ Taxalcharacters [1/2/3/4|5/6|7[8|9[10 11[12 13 14[15 16 17 [18 19] 20] 21[22 23 24 25 | 26 27 | 28 29 | 30|31 32 | 33 [34 /35[ 36]
Velociraptorinae 0,00]00|00/00/?/0/0/0]0]0,0]0]0]1/]0]0|1/0/0]0/0/0/0/0|0|/0/0/0|0|0
Archaeopteryx 0,0000000]0/0/?/0)0/|?]0]0|0|1/0/0]0/|1/0|0/0]0/0|/0/0]/0/0/?/0|0|O
Mononykus 0/000/?0|?2/00]?|?|1]?/1]0]0]0] 1? 0|0/0]0|0]|0/|0|0/1/1/1|0|0|1[0|1
Alvarezsaurus |?|?|?|?|?|?,?2|?|? ?|? |? ? |? |?|0]0|?|? ?|0|0|?|0|0|0|?|0|?|?|?| 2|? |?2|?|?
Patagonikus |?|?|?|?|2,2(2/2|2| 2| 2|? 2|? | 1]? |?2|?|?|?|0|0]0]0|0|? |? ?2|?|?|?|?2|0|1]0| 1
Iberomesornis ?|*|?|?|?|2|2|2] 2|? | 2| 1|1]?]0]0| 1| 1]0]|?2|?2|2/?| 1/?| 1/0|?|?|?|1/]0| ?| O0] O
Enantiornithes 01]01/2?|10]?2|? 0]1|?|1]1/1/|1/]?/0|1|0]2|1/1/1/|1|0/1|1/]0]1/]0]2|1| 1
Patagopteryx ?|*»|2|2?]2/2|0 1|] 1/0] 1 1] 1| 1] 12]? ]0|1|]0/2|1|1|1]|1|1/2|2|1]0/0]2 ?| 1
Hesperomithiformes|1/1/1/4/ 1/4/41] 1/1) 1/0| 1| 1[2/ 11 / 1/1]2/1/1|1|2/1]1|1/|1/1/1]0|1[0[0]2 |? | 1
Ichthyornithiformes |[1,2|1/2|222/2,1 1| 1 1 12/1] 1/,0|1|]?]?2|1|1]2|]1|1]1]1|1/1/1]1|1|1|2|1 | 1
Neomithes 114|[1]1) 113/13] 11 1/1]1]12]11]1]1/1]12]1]1]11/2/3]|1]12[11]1/1]3]1]1]|1/2]1| 1
38|39|40|41| 42 43| 44| 45| 46 47 | 48 49| 50/51/52] 53] 54/55] 56| 57| 58| 59|60| 61/62/63] 64/65) 66| 6768 | 69
Velociraptorinae [0,0/0|0/0/0/0|0/0/0/0]0|0/0/0|]0/0|0|0|0|0|0|0|0/0/0/0|0|0|0|0 0
Archaeopteryx |0 0|0|0|0/0/0/0/0|?/0|0|0/?|0|0|?/0|0|0]|00/?|0/|0|0/0/0]0|0/? 0
Mononykus 0/0/1]0|1/]0]0|1/|1|1|]1|1|1/0/1|0|0|0|1/|0/|0|0/|1|1]0/?]0/1|0|0|O 1
Alvarezsaurus |? | ? ?|?|?|0|?|?|1|?|?|?|?|?|?|?|?|?|1|0|?|0|?|0]0|?|?] 0| ?] 0|] O ?
Patagonikus 0/01|0]0|?]01|1]?|1|1|0/|?|0|0/0]0|1|0]0|0|? 0|0|?|?/|0|0]?2|2|?1|?
Iberomesornis |?*|1|?|? 1,0)0/?|?|?|?]0]|?|?|? 0??? 0,0? ?/?/|0|0/?]0]0]0|?|?|?
Enantiornithes |1/1/1,1/1/0|?|1/0]1/0|0]0/1/|1]0/1|1]0/1/0/0|?|1|]0|1|1/0|0[|0|1 1
Patagopteryx ?|1]1]?|1]0]0]|1/ 1/1] 1|[0|1/2]|?]0/ 1|2|1|1|0]0|1]|1|1|1|1]0]0|1]0 1
Hesperomitifomes| ? | 1|? |? | 11,1 1/1/1]1|1)1|?]?2/1]12|1]1,1/1)2|1/1|1/1]2|1|1[|1 1
Ichthyornithiformes | 1/1/12 1 11 1/|1/1/1]1/1/1]1]1/1]2/1|1|1|1]2]1/1|1|1]|2]|1]|1| 1 1
Neornithes 11|1120/1|[1]1/1|101]1]1]1]1/1/1]1/2]1/1/1]1]2]1/1|1/1|2|1|1]1[|1| 1
| Velodraptornae |0|0/0/0/0|0/0/0|0/0]0/0/0|0|0|0/0|0/0/0/0|0/0/0/0|0|0|0|0.| O
Archaeopteryx |0|1|1/0/?/|0|0]?|0.0]/0/1]0/?2|?|1/1|1|1/|0|?/]0/0]0/?2|?|?]0|0| O
Momnyus :(0|1,0|0/1|1|1|1|1/1|1/ 1? ?|1/ 1|0]?2|?|1|1/?2|1|0]0|1|?|0 ? | ?
. Alvarezsaurus ?2|?|0/|0|?2|2|0)0)? ? ?,?7|?|?|?|1|]0]?7 ? 1,?|]2|?2/|?2|2]?)?]? | ?| ?
Patagonikus ?]?]0]|?|?|1]1|1|1|1|1|]?]|?]?]0]|?[|?|?|?|1|?|?|?|?|0|? |?|?| 2 | 2
Iberomesomis |?|?,?|0|?|0/0]0]0|0]?|?/?|?]|?|1|?]1/?2|0]?2|?2]|?2|? ?|?|?2|2|? | ?
Enantiornithes 1|1|[1/1]1]0/1[0]0]0]0|7|0|1|1]1|1[1|?|0|0|0 ?|?2|0|2|1|?2| 1 | O
Patagopteryx ?|?*|]0|1]1|0 1/0|1]0]0]?|1/|1|1]2/|1|]0|?|1|1|0|?|?|?|?|1]?7| ? | O
Hesperomithiformes| O| 1| 1|1|1/0/0/0[0/0/? 1|,1|]1|1[2|[0|0/1/|0]1]?]1|1|1]1]0]1/0]| ?
Ichthyornithiformes | * 1 1|1/?]0/?,0/0/0/0|?|?|?|1|1|1|2]?2]0|1|1|]0|1]2|1|1/]2|0 | 1
Neomithes 1|j?|1[1]1 0,0,0,0/0|]1|/1|1|1/|2|1|1|1]0|1[1|2]|1|1]1/|1]1]| 1 | 1

Character data matrix.

60. Fibular articulation. 0. with the cal- 71. Teeth (adult). 0. with serrated crowns. 1.
caneum. 1. greatly reduced distally, without unserrated crowns.

articulation with the calcaneum. 72. Supracetabular lip. 0. present. 7, absent.
61. Metatarsals II-IV completely fused one to 73. Cervical ribs. 0. articulated with vertebrae.
each other. 0. absent. 1. present. 1. fused to vertebrae forming the costal processes.
62. Distal tarsals. 0. free. 1. completely fusedto 74. Proximal end of fibula. 0. excavated by a
the metatarsals. medial fossa. J. nearly flat.
63. Metatarsal V. 0. present. 1. absent. 75. Hypertrophied olecranon process. 0. absent.
64. Proximal end of metatarsal IIT. 0. in the 7- Present.
same plane as metatarsals II & IV. Z. reduced, not ae procoelous. 0. absent. 1.

reaching the tarsals (arctometatarsalian condi-

tion). 2. plantarily displaced with respect metatar- 77. Caudal portion of the synsacrum forming a
sals II & IV. prominent ventral keel. 0. absent. 1. present.

78. Caudal articular surface of synsacrum con-
vex. 0. absent. J. present.

79. Humerus. 0. with two distal condyles. J.
single condyle.

80. Prominent ventral projection of the

65. Well developed tarsometatarsal intercon-
dylar eminence. 0. absent. 7. present.

66. Tarsometatarsal distal vascular foramen. 0.
absent. J. present.

67. Iliac brevis fossa. 0. present. 7. absent. lateroproximal margin of the proximal phalanx of
68. Quadratojugal-squamosal contact. 0. digit I. 0. absent. 1. present.
present. 7. absent. 81. Caudal tympanic recess. 0. opens on the
69. Ischium. 0. less than two-thirds. 7. two- rostral margin of the paraoccipital process. J.
thirds or more of pubis length. opens into the collumelar recess.
70. Lateral processes on the sternum. 0. absent. 82. Quadrate. 0. with two distal condyles. J.

I. present. with three condyles forming a triangle.

582

83. Basicranial fontanelle on the ventral surface
of the basisphenoid. 0. present. 7. absent.

84. Hyposphene-hypantrum accesory interver-
tebral articulations in trunk vertebrae. 0. present.
1. absent.

85. Distal caudal prezygapophyses. 0. elongate.
1. short 2. absent.

86. Prominent acromion in the scapula. 0.
absent. J. present.

87. Completely reverted hallux (arch of ungual
phalanx of digit I oposing the arch of the unguals
of digits II-IV). 0. absent. 7. present.

88. Caudal maxillary sinus. 0. absent. J.
present.

89. Procoelous caudals. 0. absent. 1. present.

90. Carotid processes in intermediate cervicals.
0. absent. 1. present.

MEMOIRS OF THE QUEENSLAND MUSEUM

91. Ungual phalanx on major digit (digit ID. 0.
present. J. absent.

92. Dentary teeth. 0. set in sockets. 7. set in a
groove 2. absent.

93. Postorbital. 0. present J. absent.

94. Fossa for the femoral origin of M. tibialis
cranialis. 0. absent. Z. present.

95. Postorbital-jugal contact. 0. present. 7. ab-
sent.

96. Subequal cotyla of ulna. 0. present. 7. ab-
sent.

97. Costal facets in sternum. 0. absent. J.
present.

98. Bony mandibular symphysis. 0. absent. 1.
present.

99. Proximal phalanx of manal major digit
(digit II). 0. of normal shape. J. flat and
craniocaudally expanded.

NEW DATA ON THE ANKYLOSAURIAN DINOSAUR FROM THE LATE
CRETACEOUS OF THE ANTARCTIC PENINSULA

ZULMA GASPARINI, XABIER PEREDA-SUBERBIOLA & RALPH E. MOLNAR

Gasparini, Z., Pereda-Suberbiola, X. & Molnar, R.E. 1996 12 20: New data on the ankylosaurian
dinosaur from the Late Cretaceous of the Antarctic Peninsula. Memoirs of the Queensland
Museum 39(3) 583-594. Brisbane. ISSN 0079-8835.

Ankylosaurian remains have been found in the Late Cretaceous (Campanian) of the Antarctic
Peninsula (James Ross Island). The material includes a lower jaw, teeth, cervical, dorsal,
sacral? and caudal vertebrae, ribs, parts of the scapula and ilium, autopodial bones, and
armour from a single small individual. The Antarctic ankylosaur is probably a nodosaurid,
based on the tooth form. The fragmentary material doesn't permit accurate identification at
the generic or specific levels, so the specimens are referred to Nodosauridae indet. Current
data on ankylosaurian distribution supports the hypothesis of a late dispersal of nodosaurids
from the northern hemisphere to the Antarctic Peninsula via South America, although an
early migration during Late Jurassic or Early Cretaceous time is possible. The Antarctic
ankylosaur was small and lived in a high-latitude, although rather mild climate. [J
Dinosauria, Ankylosauria, Late Cretaceous, Antarctic Peninsula, Gondwana,
palaeobiogeography.

Z. Gasparini, Museo de La Plata, Departamento de Paleontología de Vertebrados, Paseo
del Bosque, 1900 La Plata, Argentina; X. Pereda-Suberbiola, Universidad del País
Vasco/Euskal Herriko Unibertsitatea, Facultad de Ciencias, Departamento de Estratigrafía
y Paleontología, Aparto 644, 48040 Bilbao, Spain; R.E. Molnar, Queensland Museum, P.O.

Box 3300, South Brisbane, Queensland 4101, Australia; 16 June 1996.

In January 1986, ankylosaurian remains were
recovered during fieldwork by the Instituto
Antártico Argentino in Late Cretaceous (Cam-
panian) sediments of James Ross Island, north-
east of the Antarctic Peninsula (Gasparini et al.,
1987). This was the first dinosaur to be dis-
covered on the Antarctic continent (Olivero et al.,
1991). Subsequently, other dinosaur remains
have been collected from the Late Cretaceous of
the Antarctic Peninsula, including a hyp-
silophodont ornithopod from the Campanian or
Maastrichtian of Seymour Island (Hooker et al.,
1991; Milner et al., 1992) and a theropod from the
Coniacian or Santonian of James Ross Island
(Molnar et al., this volume). Other occurrences of
dinosaurs in Antarctica include a large crested
theropod, Cryolophosaurus ellioti, and a
Plateosaurus-like prosauropod from the Lower
Jurassic of Mount Kirkpatrick, nearer the South
Pole (Hammer & Hickerson, 1994).

A preliminary description of the ankylosaurian
material and a discussion of its palaeo-
biogeographical implications were given by
Gasparini et al. (1987) and Olivero et al. (1991).
As a result of the preparation of the matrix in
which the fossils were imbedded, two additional
teeth were found. A latex cast was made from a
mould of three articulated cervical vertebrae.
Other material not described in the first account

includes parts of the scapula and ilium, vertebral
pieces, metapodials and phalanges. In this paper,
we discuss the affinities of the Antarctic
ankylosaur in the light of the known material.
New palaeobiogeographical data allow us to
comment further on ankylosaurian distribution
during the Late Cretaceous.

GEOGRAPHICAL AND GEOLOGICAL
FRAMEWORK

The ankylosaurian material was discovered by
the Argentinian geologists E. Olivero and R.
Scasso about 2km inside Santa Marta Cove, in the
north of James Ross Island (Olivero et al., 1991).
The remains come from the Gamma Member of
the Santa Marta Formation, in the lowermost part
of the Marambio Group (Gasparini et al., 1987).
They were recovered from locality D6-1, about
90m above the base of the Gamma Member. The
lithology consists of massive green silty
sandstones with abundant trace fossils. The
ankylosaurian bones were associated with fish
vertebrae and marine invertebrates, including
bivalves, gastropods and nautiloid cephalopods
(e.g., Cymatoceras). The unusual association of
nautiloid phragmocones with the vertebrate
bones could be explained as the result of strand-
ing of shells along a beach (Woodburne &
Zinsmeister, 1984). The ammonite assemblages

584

found above and below the dinosaur beds suggest
a probably Late Campanian age (Olivero et al.,
1991; Olivero, 1992).

The ankylosaurian remains could belong to a
single specimen, as previously reported, although
size differences between the bones suggest the
possibility of two distinct individuals. The fossils
were recovered from a small area about 6 meters
square (Olivero et al., 1991). Most bones are
fragmentary and have been much shattered by
frost action. Preparation of the remains was car-
ried out in the laboratory of the Museo de La Plata
using mechanical techniques.

Collection designations. MLP, Museo de La
Plata, La Plata; QM, Queensland Museum, Bris-
bane; ROM, Royal Ontario Museum, Toronto.

SYSTEMATIC PALAEONTOLOGY
Subclass DINOSAURIA Owen, 1842
Order ORNITHISCHIA Seeley, 1888
Suborder ANK YLOSAURIA Osbom, 1923
Family NODOSAURIDAE Marsh, 1890
Genus and species indet.

MATERIAL EXAMINED. A partial lower jaw with a
tooth in situ, two isolated teeth, a cervical vertebra and
a latex cast prepared from a natural mould of three
articulated cervical vertebrae, two dorsal centra from
the presacral rod, parts of the ?sacrum, three fragmen-
lary caudal vertebrae, rib fragments, a fragmentary
scapula, a fragment of ilium, four metapodials and two
phalanges, and a collection of dermal elements, repre-
senting five different kinds of armour. All the remains
are kept at the Departamento de Paleontología de Ver-
tebrados, Museo de La Plata with the registration num-
ber MLP 86-X-28-1.

PROVENANCE AND AGE, Santa Marta Cove, north-
ern James Ross Island, northeast of the Antarctic
Peninsula; locality D6-1, lower section of the Gamma
Member, Santa Marta Formation, Marambio Group;
Late Cretaceous, Upper Campanian (Gasparini et al.,
1987; Olivero et al., 1991; Olivero, 1992).

DESCRIPTION

A preliminary description of the specimens was
given by Gasparini et al. (1987) and Olivero et al.
(1991). Here we describe for the first time addi-
tional bones, including teeth, vertebrae, parts of
the pectoral and pelvic girdles, and autopodial
elements. Because the only other reasonably well
known ankylosaur from the southern hemisphere
isthe specimen of Minmi described in this volume
(QM F18101), detailed comparison will be made
where possible with that specimen.

Lower jaw (Fig. 11-J). The only preserved portion
of the lower jaw is a section of the left dentary,

MEMOIRS OF THE QUEENSLAND MUSEUM

including the sides and upper margin, but lacking
the ventral margin (Gasparini et al., 1987). Nine
alveoli are visible, including the foramina for tooth
emplacement on the lingual side. The anterior and
posterior parts of the tooth row are missing. The
specimen is about 15mm thick and about 60mm
long, as preserved. In occlusal view, the tooth row
is slightly curved. One replacement tooth remains
in place in the fifth preserved alveolus. The crown
is broken but it looks like those of two isolated
teeth described below. The dentary is a thick
bone, as usual in ankylosaurs (Galton, 1983). It
bears no armour. Four foramina are present on the
labial side of the dentary. Medially, the splenial
is missing and so the Meckelian canal is open.
The lack of fusion of the splenial suggests that the
lower jaw does not belong to a fully grown
ankylosaur.

Teeth (Fig. 1A-H). In addition to the tooth
preserved in situ in the lower jaw, there are two
isolated teeth. Both teeth retain the crown and
part of the root. The maximum crown width is
about 7mm and the crown height about 7.5mm.
As preserved, the more complete tooth is 20mm
high. The teeth are typically ankylosaurian, with
a leaf-shaped, laterally compressed crown
(Coombs, 1990). They bear seven or eight den-
ticles on the anterior margin and five on the
posterior, each row of denticles separated from
the other by a small apical cusp. The teeth are of
about the same size as those of Minmi (QM
F18101), but in Minmi the crowns are relatively
higher (7.5 X 6mm rather than 7.5 X 7mm) and
only have five or six anterior and three posterior
denticles. A prominent, rugose basal cingulum is
visible on each side of the crown. The cingulum
is better developed and more basally placed on
the labial side than on the other. Moreover, the
lingual cingulum is arched apically and thicker
posteriorly. The presence of a two-faced asym-
metrical cingulum is a common feature among
nodosaurids, e.g., Sauropelta (Ostrom, 1970) and
Edmontonia (Carpenter, 1990). The cingula are
more nearly symmetrical in Minmi, with one only
slightly more basal than the other, and the lingual
cingulum lacks the arch seen here. On the lingual
side of the crown there are a set of small ridges
(or sulci) that do not correspond with the den-
ticles, but extend from the bases of the denticles
to the cingulum, and may be continuous with the
vertical wrinkles of the cingulum. There seem to
be slight ridges on the labial side as well that are
continuous with the denticles, but these extend
only a short distance below them, and do not
extend across the entire crown. The pattern of

ANKYLOSAURIAN DINOSAUR FROM THE ANTARCTIC PENINSULA 585

FIG. 1. MLP 86-X-28-1, ankylosaur cranial remains from the Late Cretaceous (Campanian) of James Ross Island,
Antarctic Peninsula. A, C, two teeth in labial view; and B, D, lingual view; scale = 5mm (4x). E, tooth in lingual
view; F, distal; G, labial; and H, mesial views; scale = 5mm. I, Left dentary with a tooth in situ in medial view;
scale 10mm; and J, in occlusal view.

586 MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 2. MLP 86-X-28-1, ankylosaur vertebrae from the Late Cretaceous (Campanian) of James Ross Island,
Antarctic Peninsula. A, D, anterior cervical vertebra in anterior view; and B, E, left lateral view. C, Latex cast
from a mould of three articulated vertebrae in right lateral view (0.67x). F, two fused dorsal vertebrae from
the ?presacral rod in dorsal view; G, ventral; and I, J, anterior views. H, Caudal vertebra in anterior view.

Scale = 20mm.

vertical ridges on the crown is different from that
of Edmontonia. In the latter, the ridges are con-
tinuous with the denticles, i.e., at the margin of
the crown the ridges terminate in denticles

(Coombs, 1990; Coombs & Maryanska, 1990).
The root is slightly incurved and lingually convex
in anterior and posterior views. It is separated
from the crown by a slight constriction. If the

ANKYLOSAURIAN DINOSAUR FROM THE ANTARCTIC PENINSULA

587

FIG. 3. MLP 86-X-28-1, ankylosaur limb girdle remains from the Late Cretaceous (Campanian) of the James
Ross Island, Antarctic Peninsula. A, Scapula in articular (ventral) view. The inset indicates the humeral (g) and
coracoid (c) articular surfaces. Medial is to the top. B, Ilium in lateral view. Scale = 20mm.

teeth belong to the lower jaw, they come from the
left dentary (it is likely that they belong to the
same individual).

Vertebrae and ribs (Fig. 2). About twelve more or
less fragmentary vertebrae have been recognised.
Most are quite damaged and shattered. A latex
cast was made from a mould of three articulated
cervical vertebrae (Fig. 2C). These vertebrae are
small, with a total length of the series of about
123mm. The fusion of the neural archs to the
centra suggests that they do not belong to a
juvenile individual. The best preserved vertebra
consists of part of the centrum, neural arch
pedicles, transverse processes, prezygapophyses
and the base of the neural spine. The centra are
slightly amphicoelous, wider than long, and high
as is common in ankylosaurs (Coombs &
Maryanska, 1990). The parapophyses are placed
near the anterior border, at mid-height of the
centrum in the first preserved vertebra and be-
come moderately higher in those further back.
The neural canal is circular to slightly oval, and
its diameter reaches a third of the width of the
centrum. The transverse processes are short and
oriented almost horizontally. The prezygapo-
physes are well-developed and form an angle
with each other of about 60°. On the basis of these
characters, the vertebrae are interpreted as com-
ing from the middle of the cervical series (see
Eaton, 1960).

In addition, an isolated cervical vertebra is
known (Fig. 2F-G, I-J). This vertebra is about the
same size or slightly smaller than the articulated
cervicals. The centrum is also amphicoelous and
wider than long. The neural canal is broader than
high anteriorly, but higher than broad posteriorly.
The parapophyses appear to be on the lower half
of the lateral side. This suggests that the vertebra
comes from the anterior half of the neck and was
possibly situated anterior to the articulated series.
The cervicals are similar to those of Minmi, but

differ in detail. The pedicles of the neural arches
seem anteroposteriorly thinner in Minmi, and the
centra are basically amphiplatyan with only
slightly concave (or convex) central articular
faces.

Two co-ossified centra (Fig. 2F-G) which
preserve parts of the neural arches were originally
described as sacral vertebrae (Gasparini et al.,
1987: pl. 2, fig. 3). However, the neural canal of
the vertebrae, as exposed dorsally, seems to be
very narrow for the sacral region. In addition, the
centra are only slightly compressed dorsoventral-
ly and there is no fusion of the diapophyses and
parapophyses lateroventrally, as occurs in the
sacrals. Accordingly, these vertebrae probably
come from the presacral rod, not the sacrum.
They are tentatively regarded as the first and
second vertebrae from the rod because the ar-
ticular face of the best preserved centrum shows
no trace of fusion with another vertebra. Ver-
tebrae from the front of the presacral rod of Minmi
paravertebra (QM F10329) differ in having a
neural canal that is less laterally compressed, but
centra that more compressed.

On the other hand, the other fragmentary fused
elements may belong to the sacral region. The
more complete specimen preserves parts of two
possible co-ossified vertebrae, with broken
lateral and dorsal attachments. The ventral sur-
face is flat and narrower in the articular region
between the elements. The specimen is too frag-
mentary for accurate identification, but it may be
part of the sacral neural arches.

Abouteight fragments of dorsal ribs are known.
The ribs are T-shaped to L-shaped in cross section
proximally, a common morphology among
ankylosaurs (Coombs, 1978), although found in
some other dinosaurs as well.

At least three fragmentary caudal vertebrae are
present. One of them contains a centrum with a
fragmentary transverse process projecting

588

ventrolaterally (Fig. 2H), in anterior view this
caudal superficially resembles that of a mosasaur.
However it is amphicoelous, not procoelous. Two
other caudals retain only the lower face of the
centrum. A shallow, median groove and articular
surfaces for chevrons are visible ventrally. The
centra are very short and probably come from the
anterior part of the tail. The caudal vertebrae are
all similar in size and could belong to the same
individual.

Scapula (Fig. 3A). A fragmentary left scapula is
the only recognisable part of the pectoral girdle.
It includes the glenoid region but the scapular
spine is not preserved. The scapula and coracoid
were not co-ossified. This, together with small
size of the scapula, suggests that it could belong
to an immature individual.

Ilium (Fig. 3B). The pelvic girdle is represented
by a fragment of ilium. The specimen shows a
slightly sigmoid lateral border. The dorsal surface
is convex but the ventral surface is flat. The piece
probably comes from the central portion of the
preacetabular process of a right ilium which in
ankylosaurs — differently to other dinosaurs —
is unusually twisted laterally to lie in a horizontal
plane (Coombs, 1978; Coombs & Maryanska,
1990).

Metapodials and phalanges (Fig. 4). Four frag-
mentary metapodials and two phalanges are
known. The metapodials are broken and lack
proximal articular regions. They are relatively
broad and bear massive distal articulations, which
are approximately perpendicular to the long axis
of the shaft (Fig. 4A-B). The moderate robustness
of the metapodials suggests that they could
belong to the pes rather than to the manus; al-
though in fact, ankylosaurian metatarsals are
more slender than metacarpals (Coombs &
Maryanska, 1990). Three of the metapodials
seem roughly similar in shape but comparatively
smaller than metatarsals II, III and IV of the
nodosaurid Sauropelta (see Ostrom, 1970: pl.
26). The phalanges are very different in form.
One is cube-shaped, slightly asymmetrical in
ventral view, and longer than wide (Fig. 4C). It
probably comes from digit I. The other is a very
short, disc-like bone (Fig. 4D). It is tentatively
identified as the second phalanx of digit II or III.
The disc-like phalanx is about 1.5 times as wide
as the other. This difference in size can be inter-
preted in two ways: either one phalanx was from
the manus and the other from the pes, or they are
from two different individuals.

Armour (Fig. 5). Dermal elements are repre-
sented by five different types (Gasparini et al.,

MEMOIRS OF THE QUEENSLAND MUSEUM

B D

FIG. 4. MLP 86-X-28-l, ankylosaur manual or pedal
remains from the Late Cretaceous (Campanian) of the
James Ross Island, Antarctic Peninsula. A, B, two
metapodials; and C, D, two phalanges in anterior
view. Scale = 20mm.

1987): a) keeled, hollow-based scutes; b) massive
bulging plates; c) co-ossified flat scutes, overlap-
ping each other; d) oval, low-keeled scutes; and
e) tiny button-like ossicles.

Several fragmentary plates of the first type (a)
are known. These are massive, with an irregular
dorsal surface and prominent keel. Ventrally, the
surface is deeply hollowed in the largest
preserved specimen (Olivero et al. 1991, fig. 2c).
These plates were originally regarded as cranial
ossifications and so tentatively referred to the
Ankylosauridae (Gasparini et al., 1987). The
specimens are too fragmentary to confirm this.
The occurrence of prominent lateral horny plates
on the skull roof is a conspicuous character of
ankylosaurids (Maryanska, 1977; Coombs &
Maryanska, 1990) but similar structures could be
sporadically present in nodosaurids (Kirkland,
pers. comm., 1995).

The second type (b) of armour consists of large,
rough-surfaced plates (Gasparini et al., 1987, pl.
I, fig. 6; Olivero et al., 1991, fig. 2d). The dorsal
surface is sculptured with numerous small pits
and rugosities. The ventral face is irregular and
hollowed. These plates were originally inter-
preted as a possible tail-club but there is no direct
evidence (i.e. fusion to distal caudal vertebrae)
confirming this.

The third kind (c) of armour consists of co-os-
sified flat scutes overlapping each other and
enclosed by small polygonal ossicles (Fig. 5C-D:
Gasparini et al., 1987, pl. II, fig. 4). The scutes are
subcircular to oval in form, irregular in outline
and devoid of a dorsal keel. This pattern resembles
that of the sacral armour of Polacanthus-like
nodosaurids, which are characterised by the
fusion of a mosaic of plates into a rigid shield

ANKYLOSAURIAN DINOSAUR FROM THE ANTARCTIC PENINSULA

FIG. 5. MLP 86-X-28-], ankylosaur dermal plates from the Late Cretaceous
(Campanian) of James Ross Island, Antarctic Peninsula. Osteoderms, in-
cluding A-D, overlapped flat plates; and E, F, an oval low-keeled scute in
dorsal view. G, skull roof ossification? Scale = 20mm (0.5x).

(Hulke, 1888; Kirkland & Carpenter, 1994; Pereda-
Suberbiola, 1994). As far as known, a synsacral
shield could be present in the Antarctic ankylo-
saur but further material is need to confirm it.

The small to medium-sized scutes (d) are oval
in shape and bear a low dorsal keel (Fig. 5E-F:
Gasparini et al., 1987, pl. II, fig. 5). They were
probably situated in longitudinal rows on the
body, as diagnostic for thyreophorans (Sereno,
1986). The ventral surfaces of the scutes are
generally flat as in nodosaurids (Coombs, 1978).
Together with small ossicles, one of these scutes
has been preserved associated with a dorsal rib
and ossified tendons (Gasparini et al., 1987, pl. I,
figs 3,4). This suggests that the small scutes were
arranged in an intercostal position along the
trunk, as in Minmi and generally among
ankylosaurs.

Finally, there are numerous polygonal ossicles
(e) of very small size, less than 4mm in diameter

589

(Gasparini et al., 1987, pl. II, fig.
6). These ossicles probably
floated in the skin and formed
a continuous pavement between
the large dermal elements per-
mitting supple movements of
the body (Carpenter, 1984). A
histological study of these os-
sicles is currently in progress
(de Ricglés and collaborators,
in preparation).

The new specimen of Minmi
(QM F18101) also shows five
types of dermal armor, but only
two of them (types d and e
given here) clearly match those
of the James Ross Island
ankylosaur. These kinds seem
to be found in many
ankylosaurian taxa. Minmi
clearly lacked Polacanthus-
like sacral armor, and had
larger (5-6mm diameter) dermal
ossicles.

DISCUSSION

The general morphology of
the lower jaw, teeth, vertebrae,
ilium and armour allow confi-
dent attribution of the dinosaur
from James Ross Island to the
Ankylosauria. At least three
derived characters of
ankylosaurs (Sereno 1986;
Coombs & Maryanska 1990)
are present: posterior dorsal vertebrae fused to
form a presacral rod, ilium rotated into the
horizontal plane, and body covered by a mosaic
of armour plates of several shapes.

In a preliminary account, Gasparini et al.
(1987) tentatively assigned the material to the
family Ankylosauridae on the following charac-
ters: skull co-ossifications with lateral projec-
tions, teeth with cingula, and tail-club. However,
our redescription of the material advises caution
in regard to these features. In fact, on the basis of
tooth form the ankylosaurian remains from the
Antarctic Peninsula look more likely to be
nodosaurid than ankylosaurid. Coombs &
Maryanska (1990) pointed out that the two
families can usually be distinguished on dental
features. These authors cited several characters
that are useful in this regard, e.g., conspicuous
basal cingulum, cingulum typically higher on the

590

TABLE 1. Gondwanan ankylosaurs.

Nodosauridae indet. (Molnar & Wiffen, 1994)

TAXON AGE PROVENANCE
Nodosauridae indet. (Coria, 1994) Campanian-Maastrichtian | Patagonia, Argentina

MEMOIRS OF THE QUEENSLAND MUSEUM

North Island, New Zealand

Campanian-Maastrichtian

Nodosauridae indet. (this paper; Gasparini et al., 1987;
Olivero et al., 1991

Minmi paravertebra Molnar, 1980 (Molnar, 1980, 1991, this
volume: Molnar & Frey, 1987

Ankylosauria indet. (Rich & Rich, 1989
Ankylosauria (Chatterjee & Rudra, this volume)

James Ross I., Antarctic

Campanian a
P Peninsula

Aptian-Albian Queensland, Australia

Victoria, Australia

Gujarat, India

Aptian-Albian
Maastrichtian

lingual face, and grooves on crown generally
better developed (Coombs & Maryanska, 1990).
All of these features are present in the teeth from
James Ross Island and this suggests that they
belong to a nodosaurid ankylosaur. The differen-
ces in tooth form and the absence of a sacral
*buckler' in Minmi are the clearest distinctions
between the Antarctic beast and Minmi. Thus they
seem not to be closely related — insofar as we
can tell when many corresponding elements are
either missing (in the Antarctic specimen) or not
yet exposed (in Minmi). The currently-known
material from James Ross Island is too fragmen-
tary for generic or specific identification and is
here provisionally referred to Nodosauridae
indet. Further material is needed to accurately
define the affinities of the Antarctic ankylosaur.

The geologists who discovered the material
thought it derived from a single partial skeleton
of small size (Olivero, pers. comm.). Only some
of the bone fragments were recovered because the
ice was very thick around the fossils and the
bones were broken by frost action. However,
revision of the material at La Plata raised the
possibility that two individuals were represented.
Thus size differences between some of the bones,
e.g., phalanges, would be accounted for. Also, on
the one hand the neural arches are fused to the
centra, which supports the idea that the vertebrae
do not belong to a juvenile. But, on the other, the
scapula and coracoid (not found) were not fused,
in contrast to the usual condition of adult
ankylosaurs (Coombs & Maryanska, 1990).
Nonetheless these conditions are both seen in the
skeleton of Minmi (QM F18101) which clearly
derives from a single (articulated) individual. The
differences in size between the phalanges may be
due to one being manual and the other pedal. So
there is no strong evidence that two individuals
are represented.

The absence of fusion of the scapula to the
coracoid suggests immaturity, although fusion
does not take place in Hylaeosaurus (Pereda-

Suberbiola, 1993) nor Edmontonia (Carpenter,
1990). So the absence of fusion of the splenial and
dentary is perhaps more convincing. Fusion of the
neural arches to the centra indicates that the
animal was not very young. So most of the skeletal
remains from James Ross Island are tentatively
regarded as those of an immature, small in-
dividual. The estimated body length of the An-
tarctic ankylosaur is 3 or 4m.

PALAEOBIOGEOGRAPHICAL
SIGNIFICANCE

Ankylosaurs were quadrupedal, armoured
dinosaurs that lived from the Middle Jurassic to
the Late Cretaceous (Coombs & Maryanska,
1990). The distribution of the two families is
probably vicariant. The more conservative
Nodosauridae are known in Europe from the Cal-
lovian to the Maastrichtian (Galton, 1983;
Pereda-Suberbiola, 1992), and in North America
from the Tithonian to the Maastrichtian (Car-
penter & Breithaupt, 1986; Kirkland & Car-
penter, 1994). The club-tailed Ankylosauridae are
present in Asia from the Bathonian-Callovian to
the Maastrichtian (Maryanska, 1977; Dong,
1993), and in North America during Campanian-
Maastrichtian times (Coombs, 1986).

Unquestionable ankylosaurs from Gondwana-
land include discoveries from the Lower
Cretaceous of Australia and from the Late
Cretaceous of New Zealand, South America and
Antarctica (Table 1). Chatterjee & Rudra (this
volume) also report an ankylosaur from India. In
Australia, there are five Aptian-Albian occurren-
ces in Queensland, including the holotype of
Minmi paravertebra (Molnar, 1980; Molnar &
Frey 1987) and a nearly complete, articulated
skeleton referable to Minmi (Molnar, 1991, this
volume), and two occurrences in Victoria (Rich
& Rich, 1989). Ankylosaurs are also present in
the Campanian-Maastrichtian of North Island,
New Zealand (Molnar & Wiffen, 1994) and that

ANKYLOSAURIAN DINOSAUR FROM THE ANTARCTIC PENINSULA

591

TABLE 2. Specimens from Gondwanaland incorrectly referred to the Ankylosauria.

Loricosaurus scutatus Huene 1929 Maastrichtian

Argentina

TAXON PROVENANCE STATUS

Titanosaurid sauropod (Powell, 1980;
Bonaparte & Powell, 1980)

Lametasaurus indicus Matley 1923
(Matley, 1923; Huene & Matley, 1933)

Maastrichtian

Brachypodosaurus gravis Chakravarti
1934 (Chakravrati, 934)

[f ce e ascariensis Piveteau
isi 1923; Russell et al.,

Campanian

Madagascar

Composiite theropod, EHE and crocodilan
(Chakravarti, 1935; Walker, 1964, Molnar &
Frey, 1987)

India

| Maman | da | Stegosaur? (Galton, 1981)

unknown

Albian

Acanthopholidae indet (Lapparent,

Tv iata M c
of Río Negro Province, Patagonia, Argentina
(Coria, 1994; Salgado & Coria, in press). The list
is completed with these remains from the Cam-

panian of the Antarctic Peninsula (Gasparini et
al., 1987; Olivero et al., 1991).

Ankylosaurs have also previously been
reported from the Early Cretaceous of North
Africa (Lapparent, 1960) and the Late Cretaceous
of India (Matley, 1923; Huene & Matley, 1933;
Chakravarti, 1934; Coombs, 1978) and Madagas-
car (Piveteau, 1923; Russell et al., 1976). Never-
theless, these remains are probably not
ankylosaurian and can be assigned to other
groups of dinosaurs or archosaurs (Table 2; see
Molnar & Frey, 1987, for a discussion). As far as
known, no conclusive evidence confirms the oc-
currence of ankylosaurs in Africa (Molnar &
Frey, 1987; X. P.-S., pers. obs.), India (Chak-
ravarti, 1935; Walker, 1964; Galton, 1981;
Molnar & Frey, 1987) other than that reported by
Chatterjee & Rudra, nor Madagascar (Sues,
1980). Dermal scutes from Argentina
(Loricosaurus scutatus Huene, 1929) have been
removed from the ankylosaurs to the titanosaurid
sauropods (Powell, 1980; Bonaparte & Powell,
1980).

The presence of ankylosaurs in Antarctica
during the Late Cretaceous can be explained as
the result of dispersal from the northern con-
tinents to southern Gondwana (see discussion in
Gasparini et al., 1987; Olivero et al., 1991). The
time of this passage is not well known but two
hypotheses may be considered:

First, an early migration during the Late Juras-
sic via Africa or South America. At this time,
nodosaurids were present in Europe and North
America and palaeobiogeographical evidence
shows that passage to Gondwanaland was pos-
sible (Galton, 1977). If so, ankylosaurs were
present in southern continents before the breakup

Niger

Stegosaur? (Molnar & Frey, 1980; X. Pereda-S.,
pers. obs.) 1 - -

of Gondwana (Bonaparte 1986). According to
Molnar (1980, 1991, this volume), the mixture of
primitive and specialised characters in Minmi
could be the result of endemic evolution (because
of geographical isolation) of this lineage of
ankylosaurs in Australia during the Early
Cretaceous. There is no evidence of Jurassic
ankylosaurs in Gondwanaland and there seems
no close affinity between the Antarctic
ankylosaur and Minmi. On the other hand, the
Late Cretaceous polar dinosaur fauna from New
Zealand, that includes an ankylosaur, has been
interpreted as probably representative of the
terrestrial assemblages that inhabited Late
Cretaceous Antarctica (Molnar & Wiffen, 1994).
Unfortunately, a comparison between the
Antarctic ankylosaur and that from New Zealand
is not possible with the available material.

Second, a late dispersal, during the Late
Cretaceous, via South America. The vertebrate
interchange between North America and South
America in the Campanian is well documented
(Bonaparte, 1986; Gayet et al., 1992).
Ankylosaurs could have traversed South America
and reach Antarctica. This hypothesis is supported
by the presence of a nodosaurid in Patagonia at
this time (Coria, 1994; Salgado & Coria, in
press). A continuous or intermittant land connec-
tion between Antarctica and South America is
likely for some period during the Late Cretaceous
(Olivero et al., 1991) and the Early Paleogene
(Woodburne & Zinsmeister, 1984; Gasparini et
al.,1986; Marenssi et al., 1994).

The occurrence of an armoured dinosaur in the
Antarctic Peninsula is also interesting because it
is one of the rare occurrences of ankylosaurs in
high latitudes. During the Late Cretaceous, the
Antarctic Peninsula was situated near its present
position, at about 64°S (Zinsmeister, 1987).
Other reports of ankylosaurs in high palaeo-

592

latitudes are of the nodosaurid Edmontonia from
the Late Cretaceous of Alaska (Gangloff, 1995)
at about 70°N, Minmi from the uppermost Early
Cretaceous of Australia (Molnar, 1980, 1991, this
volume) between 70° and 80°S (Veevers et al.,
1991), and an indeterminate ankylosaur from the
Late Cretaceous of New Zealand (Molnar & Wif-
fen, 1994) at least 66°S (see Barron, 1986 and
Veevers et al., 1991, for palaeogeographical
maps). In addition to the ocurrence of the An-
tarctic ankylosaur in high latitudes, there is inde-
pendent evidence of a diversified flora suggesting
a rather mild climate and a relatively high
humidity, at least in the area where the dinosaur
remains were found (Baldoni, 1992).

As noted by Molnar & Wiffen (1994), although
large ankylosaurs (up to 9m) are known from both
Asia and North America, there is no indication of
any from the Gondwanan continents more than
4m long, including Australia, New Zealand, An-
tarctica, and South America. With the exception
of the latter find, all the others probably represent
components of insular, near-polar faunas.

Finally, the ankylosaur remains from James
Ross Island were recovered from shallow marine
deposits associated with marine invertebrates
(Olivero et al., 1991). It should be noted that
among ankylosaurs, ankylosaurid specimens are
known only from continental deposits (Table 22.1
of Coombs & Maryanska, 1990) although articu-
lated nodosaurid bones have (occasionally) been
reported from marine formations (Horner, 1979;
Carpenter et al., 1995; Coombs & Deméré, 1996).
In addition there have been several recent dis-
coveries of disarticulated nodosaurid specimens
in marine formations as well (Coombs, 1995;
Gangloff, 1995; Holden, 1996; Lee, 1996). This
taphonomic difference suggests that by the Late
Cretaceous nodosaurids preferred nearshore en-
vironments, whereas ankylosaurids probably
lived in more inland habitats. A similar ecological
segregation has been suggested among
hadrosaurid dinosaurs (Horner, 1979; Carpenter
et al., 1995). Walter Coombs (pers. comm., 1996)
pointed out that recent discoveries of specimens
from southern Califoria (Coombs & Deméré,
1996), Alaska (Gangloff, 1995) and, reportedly,
Hokkaido (Holden, 1996) suggest a circum-
Pacific distribution of nodosaurids, reminiscent
of that of desmostylians. He also noted that the
Alaskan specimen (referred to Edmontonia sp.) is
a skull very like that of ROM 1215 from the
continental Campanian of Alberta. This suggests
that nodosaurids were not restricted to marginal
marine habitats — and nodosaurids show no am-

MEMOIRS OF THE QUEENSLAND MUSEUM

phibious features in their morphology — but per-
haps had a broader tolerance of habitats than
ankylosaurids. To sum up, the Antarctic
ankylosaur remains were found in shallow
marine sediments and in high palaeolatitudes, a
typical condition among nodosaurids but un-
known in ankylosaurids. This indirectly supports
the nodosaurid affinities of the Antarctic
dinosaur.

ACKNOWLEDGEMENTS

We thank J. Moly and O. Molína for the
preparation of the material (Museo de La Plata),
F. Castets (Museo de La Plata) and C. Abrial
(CNRS, Paris) for the photographs, M. Lezcano
for the drawings and K. Carpenter and W.P.
Coombs, Jr, for their helpful comments. The
senior author appreciates the assistance of A.C.
Milner (BMNH, London) and M. Norell
(AMNH, New York) during the study of collec-
tions in their care.

LITERATURE CITED

BALDONI, A. 1992, Palinología de la Formación Santa
Marta, Cretácico superior de la Isla James Ross,
Antártida. Pp. 359-374. In Rinaldi, C. (ed.)
*Geología de la Isla James Ross’. (Instituto
Antártico Argentino, Buenos Aires).

BARRON, E.J. 1986. Global Cretaceous Paleogeog-
raphy — International Geologic Correlation Pro-
gram Project 191. Palaeogeography,
Palaeoclimatology, Palaeoecology 59: 207-214.

BONAPARTE, J.F. 1986. History of the terrestrial
Cretaceous vertebrates of Gondwana. Actas IV
Congreso Argentino de Paleontologia y
Bioestratigrafia, Mendoza 2: 63-95,

BONAPARTE, J.F. & POWELL, J.E. 1980, A continen-
tal assemblage of tetrapods from the Upper
Cretaceous beds of El Brete, northwestern Argen-
tina (Sauropoda-Coelurosauria-Carnosauria-
Aves) Mémoires de la Société géologique de
France 139: 19-28.

CARPENTER, K. 1984. Skeletal reconstruction and life
restoration of Sauropelta (Ankylosauria:
Nodosauridae) from the Cretaceous of North
America. Canadian Journal of Earth Sciences 21:
1491-1498.

1990. Ankylosaurs systematics: example using
Panoplosaurus and Edmontonia (Ankylosauria:
Nodosauridae). Pp. 281-298. In Carpenter, K. &
Currie, P.J. (eds) 'Dinosaur Systematics:
Perspectives and Approaches'. (Cambridge
University Press: Cambridge).

CARPENTER, K. & BREITHAUPT, B. 1986. Latest
Cretaceous occurrences of nodosaurid
ankylosaurs (Dinosauria: Ornithischia) in western
North America and the gradual extinction of the

ANKYLOSAURIAN DINOSAUR FROM THE ANTARCTIC PENINSULA

dinosaurs. Journal of Vertebrate Paleontology 6:
251-257.

CARPENTER, K., DILKES, D. & WEISHAMPEL,
D.B. 1995. The dinosaurs of the Niobrara Chalk
Formation (Upper Cretaceous, Kansas). Journal
of Vertebrate Paleontology 15: 275-297.

CHAKRAVARTI, D.K. 1934. On a stegosaurian
humerus from the Lameta beds of Jubbulpore.
Journal of the Geolological, Mineralogical and
Metallurgical Society of India 6: 75-79.

1935. Is Lametasaurus an armoured dinosaur?
American Journal of Science 30: 138-141.
CHATTERJEE, S. & RUDRA, D.K. 1996, this volume.
KT events in India: impact, rifting, volcanism and
dinosaur extinction. Memoirs of the Queensland

Museum 39(3); 489-532.

COOMBS, W.P., Jr 1978. The families of the ornithis-
chian dinosaur order Ankylosauria. Palaeontology
21: 143-170.

1986. A juvenile ankylosaur referable to the genus
Euoplocephalus (Reptilia, Ornithischia). Journal
of Vertebrate Paleontology 6: 162-173.

1990. Teeth and taxonomy in ankylosaurs. Pp. 269-
279. In Carpenter, K. & Currie, P.J. (eds)
‘Dinosaur Systematics: Perspectives and
Approaches’. (Cambridge University Press:
Cambridge).

1995. A nodosaurid ankylosaur (Dinosauria: Or-
nithischia) from the Lower Cretaceous of Texas.
Journal of Vertebrate Paleontology 15: 298-312.

COOMBS, W.P., Jr & DEMERE, T.A. 1996, A Late
Cretaceous nodosaurid ankylosaur (Dinosauria:
Ornithischia) from marine sediments of coastal
California. Journal of Paleontology, 70: 311-326.

COOMBS, W.P., Jr & MARYANSKA, T. 1990.
Ankylosauria. Pp. 456-483. In Weishampel, D.B.,
Dodson, P. & Osmólska, H. (eds) 'The
Dinosauria'. (University California Press:
Berkeley).

CORIA, R.A. 1994. Sobre la presencia de dinosaurios
ornitisquios acorazados en Sudamérica.
Ameghiniana 31: 398,

DONG, Z. 1993. An Ankylosaur (ornithischian
dinosaur) from the Middle Jurassic of the Junggar
Basin, China. Vertebrata PalAsiatica 20: 83-87.

EATON, T. 1960. A new armored dinosaur from the
Cretaceous of Kansas. University of Kansas,
Paleontological Contributions 8: 1-24. CHECK

GALTON, P.M. 1977. The ornithopod Dryosaurus and
a Laurasia — Gondwanaland connection in the
Upper Jurassic. Nature 268: 230-232.

1981. Craterosaurus pattonensis Seeley, a
stegosaurian dinosaur from the Lower
Cretaceous of England, and a review of
Cretaceous stegosaurs. Neues Jahrbuch für
Geologie und Paläontologie, Abhandlungen 161:
28-46.

1983. Armored dinosaurs (Ornithischia:
Ankylosauria) from the Middle and Upper Juras-
sic of England. Palaeontographica A 182: 1-25.

593

GANGLOFF, R.A. 1995. Edmontonia sp., the first
record of an ankylosaur from Alaska. Journal of
Vertebrate Paleontology 15: 195-200.

GASPARINI, Z., FUENTE, M. de la & DONADIO, O.
1986. Los reptiles cenozoicos de la Argentina.
Implicancias paleoambientales y evolución
biogeográfica. Actas IV Congress Argentino de
Paleontología y Bioestratigrafia, Mendoza 2: 119-
130.

GASPARINI, Z., OLIVERO, E., SCASSO, R. &
RINALDI, C. 1987. Un ankylosaurio (Reptilia,
Ornithischia) campaniano en el continente
antártico. Anais X Congr. Brasi. Paleont. Rio de
Janeiro 1987: 131-141. CHECK

GAYET, M., RAGE, J.-C., SEMPERE, T. & GAG-
NIER, P.-Y. 1992. Modalités des échanges de
vertébrés continentaux entre l' Amérique du Nord
et l'Amérique du Sud au Crétacé supérieur et au
Paléocéne. Bulletin de la Société géologique de
France, 163: 781-791.

HAMMER, W.R. & HICKERSON, W.J. 1994. A
Crested Theropod Dinosaur from Antarctica.
Science 264: 828-830.

HOLDEN, C. 1996. On the dinosaur trail. Science, 271:
1061,1063.

HOOKER, J.J., MILNER, A.C. & SEQUEIRA, S.
1991. An omithopod dinosaur from the Late
Cretaceous of West Antarctica. Antarctic Science
3: 331-332.

HORNER, J.R. 1979. Upper Cretaceous dinosaurs from
the Bearpaw Shale (marine) of south-central Mon-
tana with a checklist of Upper Cretaceous
dinosaur remains from marine sediments in North
America. Journal of Paleontology, 53: 566-577.

HUENE, F. von 1929, Los saurisquios y ornitisquios del
Cretáceo argentino. Anales de la Museo de La
Plata, ser. Paleont. 3: 1-196.

HUENE, F. von & MATLEY, C.A. 1933. The
Cretaceous Saurischia and Ornithischia of the
Central Provinces of India. Memoirs of the
Geological Survey of India, Paleontologia Indica
21: 1-74.

HULKE, J.W. 1888. Supplemental note on Polacanthus
foxii, describing the dorsal shield and some parts
of the endoskeleton, imperfectly known in 1881.
Philosophical Transactions of the Royal Society
of London 178: 169-172.

KIRKLAND, J.I. & CARPENTER, K. 1994. North
America's First Pre-Cretaceous Ankylosaur
(Dinosauria) from the Upper Jurassic Morrison
Formation of Western Colorado. Brigham Young
University Geological Studies 40: 25-42.

LAPPARENT, A.F. de 1960. Les Dinosauriens du ‘Con-
tinental intercalaire' du Sahara central. Mémoires
de la Société géologique de France 88A: 1-57.

LEE, Y.-N. 1996. A new nodosaurid ankylosaur
(Dinosauria: Ornithischia) from the Paw Paw For-
mation (Late Albian) of Texas. Journal of Ver-
tebrate Paleontology 16: 232-245.

MARENSSI, S., REGUERO, M., SANTILLANA, S.
& VIZCAINO, S. 1994, Eocene land mammals

594

from Seymour Island, Antarctica:
palaeobiogeographical implications. Antarctic
Science 6: 3-15.

MARYANSKA, T. 1977. Ankylosauridae (Dinosauria)
from Mongolia. Paleontologia Polonica 37: 85-
151.

MATLEY, C.A. 1923. Note on an armoured dinosaur
from the Lameta Beds of Jubbulpore. Records of
the Geological Survey of India 55: 105-109.

MILNER, A.C., HOOKER, JJ. & SEQUEIRA, S.
1992. An ornithopod dinosaur from the Upper
Cretaceous of the Antarctic peninsula. Journal of
Vertebrate Paleontology 12 (suppl. to 3): 44A.
(abstract)

MOLNAR, R.E. 1980. An ankylosaur (Ornithischia:
Reptilia) from the Lower Cretaceous of southern
Queensland. Memoirs of the Queensland Museum
20: 77-87.

1991. A nearly complete articulated ankylosaur
from Queensland, Australia. Journal of Ver-
tebrate Paleontology 11 (suppl. to 3): 47A.
(abstract)

1996, this volume. Preliminary report on a new
ankylosaur from the Early Creataceous of
Queensland, Australia. Memoirs of the
Queensland Museum 39(3): 653-668.

MOLNAR, R.E. & FREY, E. 1987. The paravertebral
elements of the Australian Ankylosaur Minmi
(Reptilia: Ornithischia, Cretaceous), Neues
Jahrbuch für Geologie und Paläontologie, Ab-
handlungen 175: 19-37.

MOLNAR, R.E., LOPEZ ANGRIMAN, A. &
GASPARINI, Z. 1996, this volume. An Antarctic
Cretaceous theropod. Memoirs of the Queensland
Museum 39(3):669-674.

MOLNAR, R.E. & WIFFEN, J. 1994. A Late
Cretaceous polar dinosaur fauna from New
Zealand. Cretaceous Research 15: 689-706.

OLIVERO, E. 1992. Asociaciones de amonites de la
Formación Santa Marta (Cretácico tardío), Isla
James Ross. Pp. 47-75. In Rinaldi, C. (ed.)
*Geología de la Isla James Ross'. (Instituto
Antártico Argentino: Buenos Aires).

OLIVERO, E., GASPARINI, Z., RINALDI, C. &
SCASSO, R. 1990. First record of dinosaurs in
Antarctica (Upper Cretaceous, James Ross Is-
land): palaeogeographical implications. Pp. 617-
622. In Thomson, M.R.A., Crame, J.A. &
Thomson, J.W. (eds) *Geological Evolution of
Antarctica'. (Cambridge University Press:
Cambridge).

OSTROM, J.H. 1970. Stratigraphy and paleontology of
the Cloverly Formation (Lower Cretaceous) of the
Bighorn Basin area, Wyoming and Montana. Bul-
Een Peabody Museum of Natural History 35:

MEMOIRS OF THE QUEENSLAND MUSEUM

PEREDA-SUBERBIOLA, X. 1992. A revised census
of European Late Cretaceous nodosaurids (Or-
nithischia: Ankylosauria): last occurrence and
possible extinction scenarios. Terra Nova 4: 641-
648.

1993. Hylaeosaurus, Polacanthus, and the sys-
tematics and stratigraphy of Wealden armoured
dinosaurs. Geological Magazine 130: 767-781.

1994. Polacanthus (Ornithischia, Ankylosauria), a
transatlantic armoured dinosaur from the Early
Cretaceous of Europe and North America.
Palaeontographica A 232: 133-159.

PIVETEAU, J. 1923. Contribution à l'étude des forma-
tions lagunaires du Nord-Ouest de Madagascar.
Bulletin de la Société géologique de France 26:
33-38.

POWELL, J.E. 1980. Sobre la presencia de una ar-
madura dérmica en algunos dinosaurios
titanosáuridos. Acta Geologica Lilloana 15: 41-
47.

RICH, T.H.V. & RICH, P.V. 1989. Polar dinosaur and
biotas of the Early Cretaceous of southeastern
Australia. National Geographic Research 5: 15-
53.

RUSSELL, D., RUSSELL, D., TAQUET, P. &
THOMAS, H. 1976. Nouvelles récoltes de
Vertébrés dans les terrains continentaux de la
région de Majunga (Madagascar). Comptes Ren-
dus sommaire de la Société géologique de France
5: 204-208.

SALGADO, L. & CORIA, R.A. in press. First evidence
of an armoured ornithischian dinosaur in the Late
Cretaceous of North Patagonia, Argentina.
Ameghiniana.

SERENO, P.C. 1986. Phylogeny of the bird-hipped
dinosaurs (Order Ornithischia). National
Geographic Research 2: 234-256.

SUES, H.-D. 1980. A pachycephalosaurid dinosaur
from the Upper Cretaceous of Madagascar and its
paleobiogeographical implications. Journal of
Paleontology 54: 954-962.

VEEVERS, J.J., POWELL, McA., C. & ROOTS, S.R.
1991, Review of seafloor spreading around
Australia. I. Synthesis of the patterns and spread-
ing. Australian Jorunal of Earth Sciences 38:373-
389,

WALKER, A.D. 1964. Triassic reptiles from the Elgin
area: Ornithosuchus and the origin of carnosaurs.
Philosophical Transactions of the Royal Society
London B 248: 53-134.

WOODBURNE, M.O. & ZINSMEISTER, W.J. 1984.
The first land mammal from Antarctica and its
biogeographic implications. Journal of Paleontol-
ogy 58: 913-948.

ZINSMEISTER, W.J. 1987. Cretaceous paleogeog-
raphy of Antarctica, Palaeogeography,
Palaeoclimatology, Palaeoecology 59: 197-206.

CRETACEOUS DINOSAURS OF AFRICA: EXAMPLES FROM CAMEROON
AND MALAWI

LOUIS L. JACOBS, DALE A. WINKLER & ELIZABETH M. GOMANI

Jacobs, L.L., Winkler, D.A. & Gomani, E.M. 1996 12 20: Cretaceous dinosaurs of Africa:
Examples from Cameroon and Malawi. Memoirs of the Queensland Museum 39(3): 595-610.
Brisbane. ISSN 0079-8835.

Africa became progressively isolated during the middle portion of the Cretaceous Period as
it rifted apart from other Gondwana continents and the southern Atlantic Ocean was
completed. Within this geological context, which controls the occurrence and distribution of
Cretaceous fossil localities over most of Africa, there is considerable variety in composition
and occurrence of fossils. In the Koum Basin of Cameroon in western Africa, abundant
footprints of five morphotypes, but scrappy skeletal remains and teeth of the ornithopod
Ouranosaurus and other taxa are found. In the Karonga area of northern Malawi, in
southeastern Africa, more complete skeletal remains are known, including the sauropod
Malawisaurus, which shows affinity to South American titanosaurids, and an undescribed
sauropod taxon. These may demonstrate a range of morphological diversity in sub-Saharan
Cretaceous sauropods comparable to that found in South America. [ ] Cretaceous, Africa,
Sauropoda, palaeobiogeography, crocodilia, stratigraphy.

Louis L. Jacobs, Dale A. Winkler & Elizabeth M. Gomani, Department of Geological
Sciences and Shuler Museum of Paleontology, Southern Methodist University, Dallas, Texas

75275 USA; 15 August 1996.

Africa is a vast continent with a sporadic fossil
record, remarkably good for some time intervals
in some regions, yet dismally bad in others, yield-
ing a disjointed and sketchy picture of past life on
the continent. Our research into the African
Cretaceous is focused on sub-Saharan Africa, a
largely tropical to subtropical Precambrian shield
area, heavily vegetated for the most part, and
deeply weathered, often into fossil-destroying
lateritic soils. In this paper we are primarily
concerned with two areas of Lower Cretaceous
outcrop: one in Cameroon and the other in
Malawi. Both are associated with extensional
tectonics along reactivated Precambrian mobile belts
(Mateer et al., 1992). Our primary objective in this
paper is to provide a general comparison between
the two; that is, between eastern and western
African dinosaur faunas of the Early Cretaceous
using the Koum Basin, Cameroon, and the
Dinosaur Beds of Karonga, Malawi, as examples
(Fig. 1). We will also take this opportunity to
examine briefly African Cretaceous dinosaurs
and their relationship to those of South America.

Africa became isolated from Madagascar,
India, Australia and Antarctica during the Late
Jurassic (Pittman et al., 1993). One of the most
profound events affecting the physical geography
of the Cretaceous world was the opening of the
southern Atlantic Ocean by the rifting of South
America and Africa. Beginning in the earliest
Cretaceous, tectonic rifting that culminated in

the isolation of Africa began inthe south between
the continents. By the Late Albian to Cenoman-
jan the African terrestrial fauna may have been
isolated from that of South America by the Atlan-
tic (Pittman et al., 1993). These geographic chan-
ges would have precipitated climatic change by
modifying oceanic and atmospheric currents dis-
tributing heat across the Cretaceous world.

In the Early Cretaceous, volcanism in southern
Africa and the reactivation of the Precambrian
East African mobile belt resulted in Cretaceous
fossil deposits in Malawi, and along the Zambezi
River and its tributaries (Eby et al., 1995). Also
in the Early Cretaceous, as noted above, on the
west side of the continent, South America and
Africa were rifting apart along a fault-rift-rift
triple junction (Popoff, 1988), completing the
Atlantic Ocean by Cenomanian or earlier time.

Thus, the Cretaceous fossil record is largely
controlled by the tectonic framework of the con-
tinent, which provided the geologic setting in
which fossil deposits formed. The chronology of
African Cretaceous localities has yet to be docu-
mented in detail. Many identifications as
presented in the literature are preliminary or
based on incomplete material. Table 1 is a com-
pilation of published records of Cretaceous
dinosaurs from Africa. This is provided as a
guide to published literature only, in recognition
of the rather tentative nature of most of the
entries.

596

KOUM BASIN, CAMEROON

The Koum Basin is an isolated basin associated
with the Benue Trough, a major structural feature
of West Africa (Figs 1, 2). The Benue Trough is
an aulacogen, the failed arm of the triple junction
that led to the completion of the Atlantic Ocean,
extending from the Gulf of Guinea through
Nigeria, then dividing into northern and southern
branches. The southern branch is called the Yola
Arm, which passes into Cameroon. While the
Koum Basin is not the only fossiliferous basin
associated with the Yola Arm (see also Brunet et
al., 1988; Dejax et al., 1989; Michard et al., 1990),
it is of interest here because of the dinosaur fauna
it contains.

The most complete study of the Koum Basin
(also referred to as Mayo Rey Basin in Flynn et
al., 1987) and its fauna to date is that of Congleton
(1990). The Koum Basin (Figs 2, 3) is an 80km
long half-graben oriented east-west, bounded to
the north by a fault, and filled with up to 3,000m
of fluviatile, overbank, and lacustrine Cretaceous
sediments (Tillement, 1971). It is surrounded by
Precambrian metamorphic rocks. Cretaceous
sediments are exposed primarily along the Mayo
Rey, an east to west flowing tributary to the
Benue River.

Congleton (1990) referred to the Cretaceous
sediments as the Koum Formation, in which he
recognised two areally restricted and well-
defined members. The Mbissirri Member is com-
posed of reddish fine-grained silty mudstones,
clay shales, thin limestones and sandstones.
There appear to be cyclical repetitions of cross-
bedded sandstones and thicker mudstones within
a sequence generally coarsening upwards. Con-
gleton (1990) suggested that the Mbissirri Mem-
ber represented lacustrine and aggraded
meandering stream sediments within a fine-
grained meander belt. Carbonised plant frag-
ments, conchostracans, turtle carapace fragments
and dinosaur footprint localities (designated KB-
3, KB-17, KB-18, KB-23 in Fig. 3) have been
found in the Mbissirri Member.

The Mbissirri Member is overlain along the
northern margin of the basin by the coarse-
grained Grés de Gaba Member. The contact is
gradational. The sediments of the Grés de Gaba
Member are medium to coarse-grained cross-
bedded arkosic sandstones, conglomerates, and
interbedded mudstones and palaeosols. Con-
gleton (1990) interpreted the Grés de Gaba Mem-
ber as representing a coarsegrained, braided
fluvial system. Bone-bearing localities (KB-6,

MEMOIRS OF THE QUEENSLAND MUSEUM

Karonga 4

FIG. 1. Location map of Koum Basin, Cameroon and
the Dinosaur Beds, Mwakasyunguti area, Karonga,
Malawi. Lakes Chad and Malawi indicated in black.

KB-8, KB-13), but no footprints, are known from
the Grés de Gaba Member. Congleton et al.
(1992) present a section measured at locality
KB-6, the most productive bone locality. The
base of the locality is coarse sand containing
fragmentary scattered bones and teeth repre-
senting predominantly, in order of abundance,
iguanodontian, theropod and sauropodomorph
dinosaurs. One tooth was identified as
thyreophoran(?); we consider that identification
dubious and the specimen possibly confused
with a crocodilian. The iguanodontian teeth
resemble those of Ouranosaurus. The section
fines upwards and small bones, representing par-
ticularly anurans and the crocodyliform
Araripesuchus, but also mammals (Brunet et al.,
1990, Jacobs et al., 1988), are concentrated in the
lower portion of a mudstone, at the top of which
a calcareous palaeosol is developed. Insect trace
fossils resembling hymenopteran larval cases are
found in the palaeosol. The sequence is suc-
ceeded by a coarse sandstone representing the
initiation of the next fluvial cycle. A neural arch
and neural spine attributed to the region of the
sixteenth dorsal vertebra of Ouranosaurus (based
on height of spine and position of the transverse
processes) was found in conglomeratic sandstone
at locality KB-13, overlying KB-6 by ap-
proximately 5 meters.

CRETACEOUS DINOSAURS OF AFRICA

200 km

NIGERIA

ESE RRR E ws

597
GONGOLA ARM at
e 1 i + CHAD
x^ r i
xn “A LeQaroua.
YOLA ARM LA T,
n Q9 1
: / KOUM BASIN
z ( s en
CAMEROON i CENTRAL
* AFRICAN
* REPUBLIC
e Yaoundé `

FIG. 2. Distribution of Cretaceous and younger sediments (stippled pattern) in the Benue trough (redrawn from

Allix, 1983).

Congleton et al. (1992) suggested that faunas
of the Koum Basin were correlative with the
locality of Gadoufaoua, Niger, based on the com-
mon occurrence of Araripesuchus and
Ouranosaurus. The age of Gadoufaoua is usually
considered Aptian (Taquet, 1976). Colin et al.
(1992) report the ostracode Cypridea minuscula
from the Koum Basin (locality KB-24, Mbissirri
Member). This genus was originally described
from the Candeias Formation, Reconcavo Basin,
Brazil. The Candeias Formation is placed within
the Rio da Serra local stage (Moura et al., 1994),
which ranges from the lower Berriasian to the
lower Hauterivian stages (Ponte, 1994). Brunet et
al. (1988) estimated the age of the nearby Mayo
Oulo Léré Basin to be approximately around the
Hauterivian-Barremian transition based on
plants, particularly the palynomorph
Dicheiropollis etruscus. A similar age is sug-
gested for the Babouri-Figil Basin, which also
contains dinosaur tracks, for the same reason
(Dejax et al., 1989). It appears likely that the
Koum Basin is Early Cretaceous in age, no
younger than Aptian, but perhaps older.

Dinosaur footprints are relatively abundant in
the Koum Basin. Congleton (1990) recognized
five morphotypes of tracks corresponding to
theropod, ornithopod and sauropod, consistent
with incomplete skeletal remains from elsewhere
in the basin. The most prolific trackway site is
KB-17 in the Mbissirri Member, where tracks
occur in four successive sandstone strata. Small
theropod prints are the most abundant mor-
photype. Taken together, the trackways are
directed north-northeast or south-southwest con-
sistently, suggesting that the pathways remained
constant over the flooding events that deposited
the track-bearing strata and that a consistent bar-
rier guided the movements of the dinosaurs.

A considerable amount of variation in stride
length and other indicators of relative speed or
inferred behavior is reported by Congleton
(1990). Interpreted gaits include walking, trot-
ting, sprinting, acceleration, deceleration and
hobbling. The majority of the trackmakers appear
to have been walking, however, based on relative
stride length and pace angle. Smaller trackmakers
appear relatively more energetic than larger

598

+ +
+e + « ToGaroua" *
K.

Tcholliré

Fault

MEMOIRS OF THE QUEENSLAND MUSEUM

+ T + + id
Dur a cpu T Me Onde Rr
"E `
' To Touboro | N
A A
Lower Grés de Gaba Member | koum SY
retaceous
[|] Mbissirri Member Fm.
Precambrian |$ $+ Metamorphic basement
KB fossil locality 0 10 20 km

FIG. 3. Geological map of the Koum Basin, northern Cameroon (redrawn from Congleton, 1990).

trackmakers, as interpreted from inferred stride
rates.

KARONGA, MALAWI

The Dinosaur Beds of northern Malawi are
located along the Sitwe River in the Mwakasyun-
guti area of Karonga District. Malawi is a long,
narrow country located in the southern portion of
the East African Rift. Karonga is the north-
ernmost province and Mwakasyunguti is located
approximately 70km southwest of Karonga Dis-
trict Headquarters. Jacobs et al. (1992) provided
an overview of previous palaeontological re-
search in Malawi.

The Sitwe River near Mwakasyunguti flows
through a north-south trending half graben, paral-
leling the Malawi Rift System (Kaufulu, 1989;
Tiercelin et al., 1988) and bordered by
Precambrian metamorphic basement. Three
sedimentary formations are exposed in the half
graben (Fig. 4): the Dinosaur Beds are deposited
on the basement and are unconformably overlain
by the Plio-Pleistocene Chiwondo Beds, which
are in turn unconformably overlain by the Pleis-
tocene Chitemwe Beds.

The Dinosaur Beds are commonly tilted to the
northeast with dips ranging from 14-25? north-
east. They are cut by two obvious faults (Fig. 4).

One fault trends north-south, cutting across al-
most the entire study area. It juxtaposes sedi-
ments of the upper Dinosaur Beds against the
lower Dinosaur Beds and the Precambrian base-
ment. Drag on this fault results in southwesterly
dips for the upper part of the Dinosaur Beds. The
second clearly visible fault cuts across Cretaceous
and Cenozoic sediments, juxtaposing the entire
package against Precambrian basement.

Dixey (1928; see also Gomani, 1993) divided
the Dinosaur Beds into an upper and a lower
member based mainly on color. The lower mem-
ber is unfossiliferous red and purple mud and
calcite-cemented sandstone, locally con-
glomeratic with some mottling and dessication
cracks. The upper member is composed primarily
of fossiliferous alternating red and white
crossbedded fluvial sandstones deposited along
braided streams. At least two distinct modes of
bone preservation are found in the Dinosaur Beds.
Nearly complete and articulated notosuchid
crocodyliforms (Clark et al., 1989; Gomani,
1993, in press) are found within burrows in aban-
doned channel sediments. A more diverse suite of
dinosaurs and other taxa occur as isolated bones,
associated clusters, or articulated elements in
course poorly-sorted sandstone (Gomani, 1993).
Plant macrofossils are rare with only two diaspores

CRETACEOUS DINOSAURS OF AFRICA

z Po
CD-9, 10, 11, 15, 17
* CD-8
€* CD-1,2,3,6,7

599

skeletons, which apparently
died in burrows (Gomani, in
press). It has a long and flat jaw
articulation surface and a
heterodont dentition with mul-
ticusped teeth (Clark et al.,
1989; Gomani, in press) that
appear to be similar to Can-
didodon from Brazil (Carvalho
& Campos, 1988; Carvalho,
1994) and Chimaerasuchus

from China (Wu et al., 1995).
Although distinct, a close
phylogenetic relationship of
the Malawi notosuchid to Can-
didodon and Chimaerasuchus
appears likely, but cannot be
tested until better material of
these genera is found (Gomani,

Alluvium
Chitimwe Beds
Chiwondo Beds

Upper member | Dinosaur Beds
Lower member

Holocene

Plio-
Pleistocene

[73
E]
L]
Cretaceous} E54
Precambrian Metamorphic basement "

Lower

€ CD-12 Dinosaur Beds fossil locality

Fault

FIG. 4. Geological map of Mwakasyunguti area, Karonga District, Malawi

(redrawn from Gomani, 1993).

having been recovered (Jacobs, 1990), and pollen
is not preserved. The vertebrate fauna indicates
an Early Cretaceous age (Jacobs et al., 1990).

The vertebrate fauna, including fish (but not
lungfish), anurans, turtles, crocodyliforms and
dinosaurs (Jacobs et al., 1990, 1992; Gomani,
1993), has been collected from localities desig-
nated as CD-l to CD-17 (Fig. 4). Anurans are
represented by two skull fragments: one has
rugose dermal bones and the other one has no
sculpturing (Jacobs et al., 1990). The
crocodyliforms include an interesting notosuchid
(Gomani, in press), aff. Araripesuchus sp., and
teeth of an unidentified taxon. Aff. Araripesuchus
is represented by isolated ziphodont teeth which
are indistinguishable from specimens from
Cameroon (Congleton, 1990). The latter
resemble the type of Araripesuchus wegeneri
from Niger (Buffetaut, 1981), which Kellner
(1994) argues may be generically distinct from
the genotypic South American species. The
notosuchid species is represented by complete

Air photo: No. 0890 Run 9 Mw 4 78

in press). The crus and tarsal
bone articulations of the
Malawi notosuchid are vertical
and the distal condyles of the
femur are posteriorly directed.
These indicate an erect posture.
The ventral position of the oc-
cipital condyle and foramen
magnum indicates that the head
was held perpendicular to the
neck (Gomani, in press).

Based on field identifica-
tions, it was suggested that the
Malawi dinosaur fauna con-
sists of at least five taxa includ-
ing a titanosaurid, a diplodocid, two morphs of
theropods and a stegosaur. However, examina-
tion of Argentinean dinosaur specimens and fur-
ther preparation of the Malawian material
necessitate re-evaluation. There are at least two
species of sauropods, one of which is
Malawisaurus dixeyi (Haughton, 1928; Jacobs et
al., 1990, 1992, 1993; Gomani, 1993), the other
of which is undescribed and has ‘pencil-like’
teeth. At least two species of theropods are
present, based on teeth (Jacobs et al., 1990, 1992).
The most abundant elements belong to Malawi-
saurus, a titanosaurid.

Malawisaurus is a titanosaurid because it has a
transversely expanded ischium and strongly
procoelous anterior caudal vertebrae, derived
characters that are considered diagnostic for the
Titanosauridae, or possibly a more inclusive
monophyletic group containing Malawisaurus.
Titanosaurids are known to have dermal armor
(Depéret, 1896; Bonaparte & Powell, 1980).

600

Malawisaurus lacks direct evidence of dermal
armor, but calcite pseudomorphs shaped like der-
mal armor were found associated with the bones
in the same quarry (Jacobs et al., 1993).

One cervical vertebra referred to Malawisaurus
(Mal-180, Fig. 5A) has a low neural spine that is
not bifurcated. The prezygapophyses extend
10cm beyond the anterior end of the centrum but
the postzygapophyses do not extend posteriorly
beyond the centrum. The centrum lacks
pleurocentral cavities. The cervical ribs are
coosified to the centrum and have cranial proces-
ses that terminate at the anterior limit of the
centrum. The shafts of the ribs extend beyond the
centrum posteriorly. In Saltasaurus loricatus
(PVL 4017-139) of Argentina and in cf.
Titanosaurus from Brazil (Powell, 1986), the
prezygapophyses do not extend beyond the
centrum while the postzygapophyses do extend
beyond the centrum. In addition, Saltasaurus and
cf. Titanosaurus cervical vertebrae have
pleurocentral cavities on the centra. In anterior
cervicals of Saltasaurus neural spines do not rise
above the neural arches while in posterior cervi-
cals low neural spines rise above the neural ar-
ches. Based on comparison with Saltasaurus, this
vertebra of Malawisaurus (Mal-180) is probably
posterior to the fourth cervical in position.

Dorsal vertebrae attributed to Malawisaurus
(Mal-181, and Mal-182, Figs 5B-D) indicate no
evidence of hyposphene-hypantrum articular
surfaces. The centra are opisthocoelous and
pleurocoels are anteriorly restricted. The
transverse processes of Mal-181 are wide and
slightly inclined dorsally. The broad-based, un-
divided neural spine has elongate anterolateral
depressions that are at the same level as the post-
zygapophyses. A well-developed thin prespinal
lamina occurs at the anterior base. The postspinal
lamina is thicker than the prespinal lamina.

The base of the undivided neural spine in Mal-
182 is narrower than in Mal-181. The prespinal
lamina is well-developed. In anterior dorsals
(dorsals 1-4) of sauropods the parapophyses are
located on the centra while on the posterior
dorsals, the parapophyses are high on the centro-
diapophysial lamina that extends from the
diapophysis (McIntosh, 1990). Mal-181 and Mal-
182 are not anterior dorsals because the centra
lack parapophyses. In Epachthosaurus, the
parapophyses become progressively higher
posteriorly on dorsals. The transverse processes
of Mal-182 are broken, thus the height of the
parapophyses and relative vertebral position can-
not be determined. The prezygapophyses in Mal-

MEMOIRS OF THE QUEENSLAND MUSEUM

182 are closer together than in Mal-181. The
undivided neural spine is high and inclined
caudally in Mal-181, but it is vertical in Mal-182.
Supradiapophysial laminae in Mal-181 and Mal-
182 are well-developed. Deep lateral excavations
and high prespinal laminae also occur in Salta-
saurus loricatus, Andesaurus delgadoi and
Argyrosaurus superbus from Argentina
(Bonaparte & Powell, 1980; Calvo & Bonaparte,
1991). In Saltasaurus, caudally inclined neural
spines occur in more anterior dorsal vertebrae and
vertical neural spines occur in posterior dorsal
vertebrae (Powell, 1986). Based on comparison
with these taxa, Mal-182 probably is more
posterior in the vertebral column of
Malawisaurus than Mal-181.

Caudal vertebrae of Malawisaurus are
described by Jacobs et al. (1993). The vertebrae
progress from procoelous to gently amphicoelous
posteriorly. The anterior caudals are strongly
procoelous, the prezygapophyses extend slightly
beyond the centrum as in Titanosaurus sp. from
Argentina (Powell, 1986) and Alamosaurus from
North America (Gilmore, 1946). However,
Alamosaurus and Titanosaurus are more derived
because all their posterior caudal vertebrae are
procoelous. Malawisaurus also shares low caudal
neural spines with Andesaurus and derived
titanosaurids. Malawisaurus is distinct from
Andesaurus and Epachthosaurus because these
genera have hyposphene-hypantrum articular
surfaces on dorsal vertebrae (Calvo & Bonaparte,
1991) that are absent in Malawisaurus. All caudal
vertebrae of Epachthosaurus are procoelous.

Isolated teeth similar to those associated with
titanosaurids from Argentina (Huene, 1929 PI. 1.
Figs 12-13; Powell, 1986), Brazil (Kellner, this
volume) and to those associated with
Alamosaurus from North America (Kues et al.,
1980; Lucas & Hunt, 1989) occur in the Dinosaur
Beds. These teeth are slender and conical.
Presence of relatively long slender and conical
(peglike) teeth is one of the diagnostic features of
Titanosauridae and Diplodocidae (McIntosh,
1990; Titanosauroidea and Diplodocoidea of Up-
church, 1995). Thus, the slender teeth in the
Dinosaur Beds may be titanosaurid or diplodocid.
The presence of diplodocids in the Dinosaur Beds
is suggested by the derived features of an un-
described mandible. It has a short tooth row and
the ramus turns at à sharp angle towards the
symphysis where the dorsal margin of the man-
dible flares outward. These characters are also
present in South American Antarctosaurus
(Huene, 1929; Huene & Matley, 1833; Powell,

CRETACEOUS DINOSAURS OF AFRICA

601

Diplodocoidea of Upchurch,
1995), it is clear that the mor-
phological diversity of both
South American and African
saurpods may be greater than
we previously anticipated.

A pelvis from CD-10 is a
titanosaurid as suggested by
the pelvis of Epachthosaurus
and Saltasaurus. The anterior
blade of the ilium of these
titanosaurids is curved
anterolaterally, a feature that
may be derived for this group
of sauropods. Further descrip-
tion of the pelvis from Malawi
awaits complete preparation.

DISCUSSION AND
CONCLUSION

Dinosaurs of Cretaceous age
in Africa are widely distributed
geographically (Table 1). The
data used in the compilation of
Table 1 were derived from
Broom (1904), Haughton
(1928), Dixey & Smith (1929),
Dixey (1937, 1939), Greigert et
al. (1954), Lavocat (1954),
Lapparent (1960), Said (1962),
Broin et al. (1971,1974), Ken-
nedy & Klinger (1972), Russell
et al. (1976), McLachlan &
McMillan (1976), Taquet
(1976,1982,1984), El-Khashab
(1977), Monbaron (1978),
Klitzsch et al. (1979),
Pentel kov & Voronovsky
(1979), Sues & Taquet (1979),
Dutuit & Ouazzou (1980),
Sues (1980), Galton & Coombs
(1981), Rich et al. (1983),
Cooper (1985), Kent &
Gradstein (1985, 1986), Wil-

FIG. 5. Vertebrae of Malawisaurus. A, middle cervical, Mal-180, left lateral
view; scale = 10cm. B, C, D, dorsals of Malawisaurus (B, C, Mal-181, right
lateral and posterior view, respectively). D, Mal-182, left lateral view. B-D

liam & Savage (1986), Flynn et
al. (1987), Handford (1987),

scales = 3cm.

1986). Those authors consider Antarctosaurus a
titanosaurid, but others have suggested
diplodocid or diplodocoid relationships (Jacobs
et al., 1993; Hunt et al., 1994; Upchurch, 1994,
1995). While these are given as characters of
Diplodocidae (Berman & McIntosh, 1978;

Kennedy et al. (1987), Mateer
(1987), Shrank (1987),
Bouaziz et al. (1988), Brunet et al. (1988), Carroll
(1988), Buffetaut (1989a, 1989b), Dejax et al.
(1989), Jacobs et al. (1989, 1990, 1992, 1993),
Colin & Jacobs (1990), Congleton (1990), Jacobs
(1990), Klitzsch & Squyres (1990), Lefranc &
Guiraud (1990), Weishampel (1990), Wycisk

602

(1990), Moody & Sutcliffe (1991), Colin et al.
(1992), Congleton et al. (1992), Mateer et al.
(1992), Werner (1993a, 1993b), Wescott et al.
(1993), Krause & Dodson, (1994); Sereno et al.
(1994, 1996), Rauhut & Werner (1995), Forster
(1996), Sampson et al. (1996), Krause et al. (in
press), and the references therein. The most
widely distributed sauropods in Africa are
titanosaurids. This is important for palaeo-
biogeographical comparison between Africa and
South America because titanosaurids are also the
most widely distributed geographically and most
numerous sauropods in South America (Huene,
1929; Huene & Matley, 1933; Bonaparte &
Powell, 1980; Powell, 1986; Weishampel, 1990).
In Africa, Cretaceous titanosaurids are reported
from the Campanian of Madagascar (Russell et
al., 1976; Sues & Taquet, 1979; Sues, 1980;
Forster, 1996; Sampson etal., 1996; Krause et al.,
in press), the Senonian of Sudan (Werner, 1993a,
1993b), the Turonian-Santonian of Kenya
(Arambourg & Wolff, 1969; Weishampel, 1990;
Westcott et al., 1993), the Cenomanian of Egypt
(Stromer, 1932), the Albian of Niger (Lapparent,
1960), the Aptian of Malawi (Jacobs et al., 1992,
1993).

Titanosaurid sauropods are clearly important
elements in the Cretaceous fauna of both South
America and Africa. Titanosaurids are consider-
ed to include Aegyptosaurus, Aeolosaurus,
Alamosaurus, Andesaurus, Antarctosaurus,
Argentinosaurus, Argyrosaurus, Epachthosaurus,
Hypselosaurus, | Janenschia, Laplatasaurus,
Macrurosaurus, Magyarosaurus, Malawisaurus,
Saltasaurus, Titanosaurus and questionably
Campylodoniscus. The most diagnostic and most
commonly used synapomorphy of titanosaurids
is strongly procoelous caudal vertebrae. Having
all caudal vertebrae procoelous is derived relative
to having posterior caudals amphicoelous.

Malawisaurus has gently amphicoelous
posterior caudals. This feature is also seen in
Janenschia from Tanzania (Janensch, 1922;
1961), Aeolosaurus rionegrinus from Argentina
(Powell, 1986; 1987) and Macrurosaurus from
England (Seeley, 1876). No anterior caudal ver-
tebrae are known for Andesaurus (Calvo &
Bonaparte, 1991). The posterior caudal (Calvo &
Bonaparte, 1991, Figs 4A-C) of the latter is
described as amphiplatyan. Huene & Matley
(1933) described an amphiplatyan posterior
caudal and associated it with 7itanosaurus
indicus. The use of the terms gently amphicoelous
as compared to amphiplatyan in this case may be
a distinction without a difference. Having gently

MEMOIRS OF THE QUEENSLAND MUSEUM

amphicoelous posterior caudal vertebrae may
be fairly common among titanosaurids and taxa
cannot be excluded from the group because of
amphicoelous posterior caudals.

Diplodocids in the Cretaceous of Africa are
reported from the Cenomanian of Morocco and
the Albian of Niger, Algeria and Morocco
(Lavocat, 1954; Lapparent, 1960; Taquet, 1976).
These are represented by dorsal vertebrae,
humerus, sacrum and teeth of Rebbachisaurus
and isolated pencil-like teeth (not figured) of an
unnamed taxon from Niger (Taquet, 1976). The
identification of the unnamed taxon is based on
comparison with Dicraeosaurus teeth from the
Kimmeridgian of Tanzania. Because some pen-
cil-like teeth are referred to titanosaurids, the
teeth from Niger require further study for confi-
dent assignment to taxon. In the Cretaceous of
South America, diplodocids (diplodocoids of Up-
church, 1995) are represented by Rebbachisaurus
(recently reported by Calvo & Salgado, 1995, in
a paper we have not yet had the opportunity to
evaluate) and Amargasaurus (Salgado &
Bonaparte, 1991). Amargasaurus is most closely
related to the Late Jurassic Dicraeosaurus from
Tendaguru, Tanzania (Upchurch, 1995). The
African titanosaurid Malawisaurus is more
derived than Janenschia from Tendaguru and
more similar to South American Andesaurus. The
undescribed jaw from the Dinosaur Beds of
Malawi may be diplodocid but the phylogenetic
relationships to other sauropods will be presented
elsewhere.

Early Cretaceous sediments and fauna from
Cameroon and Malawi are depositionally dif-
ferent and taxonomically distinct. In Cameroon,
fossils occur in lacustrine, meanderbelt channel,
overbank and braided stream deposits, while in
Malawi dinosaur fossils occur predominantly in
course-grained, braided fluvial sediments lacking
finegrained overbank deposits. Footprints occur
in Cameroon but not Malawi. Bones are more
complete in Malawi. Cameroon appears to have
a more diverse total fauna than Malawi. In
Cameroon, the ornithopod Ouranosaurus is
present while Malawi has no ornithopods. In
Malawi, sauropods are the best represented ele-
ments, while they are rare in Cameroon. In
neither area do we have a reasonably complete
understanding of the fauna because of inadequate
sampling, but both contribute to the palaeon-
tological baseline for Africa. Compared to other
African localities, the Koum Basin appears most
similar to Gadoufaoua, Niger, which is close in
proximity and probably in age; but neither Koum

CRETACEOUS DINOSAURS OF AFRICA 603

Tapparent, 1960; Said, 1962; El-Khashab,
19

Elaphrosaurus |Cenomanian | Marsa Matruh, Egypt (Baharija Fm)

Kasr-es-Souk, Morocco (Tegana Fm) Lapparent, 1960; Monbaron, 1978

mer rq Southern Sahara, Niger Sereno et al., 1994
Majungasaurus npa Marsa Matruth, Egypt (Phosphatic Beds) Lapparent: 1960; Said, 1962; El-Khashab,
: NE Russell et al., 1976; Sues & Taquet, 1979;
Campanian tunga District, Madagascar Sues, 1980; Foster, 1996; Sampson et al.,
1996; Krause et al., in press
Bahariasaurus Turonian- Agadez, Niger (In Beceten Fm) Greigert et al., 1954; Broin et al., 1974
Santonian 8 7 MB B "i t p
, Marsa Matruh, Egypt (Baharija Fm); Greigert et al., 1954; Lapparent 1960;
Cenomanian Tahoua, Niger (Farak Fm): orthern Said, 1962; Taquet, 1976; El-Khashab,
Province, Sudan (Wadi Milk Fm) 1977; Werner, 1993a
Albian Agadez, Niger (Continental Intercalaire) — | Lapparent, 1960
1 Marsa Matruh, Egypt (Baharija Fm); North- | Lavocat, 1954; Lapparent, 1960; Said,
ocho a i Cenomanian ern Province, Sudan, (Wadi Milk Fm); Kem | 1962; Ei-Khashab' 977; Werner, 1993a;
g Kem, Morocco, (Kem Kem Beds) Sereno et al., 1996
Aiba cr Pammada du Guir, Morocco (Continental Buffetaut, 1989a, 1989b

Tamourhest, Vaga, & aera Ageia,
paces, Niger (Continental Intercaleue)i Lapparent, 1960; Monbaron, 1978;
Kasr-es-Souk, Morocco (Tegana Fm); ppareni, à , ,
Gharyan, Libya (Continental Intercalaire) | Bouaziz et al., 1988
Medenine, Tunisia (Chenini Fm)

Albian

———:
Medenine, Tunisia (Chenini Fm) Bouaziz et al., 1988
Deltadromeus | |Cenomanian Kem Kem, Morocco, (Kem Kem Beds) Sereno et al., 1996
Spinosaurus |Turonian- nd Province, Kenya (Turkana | Weishampel, 1990; Wescott et al., 1993

Cenomanian Marsa Matruh, Egypt (Baharija Fm) Lapparent, 1960; Said, 1962; El-Khashab, 1977

Albian- Hammada du Guir, Morocco (Continental
Cenomanian Red Beds) Buffetaut, 1989a, 1989b

Albian Medenine, Tunisia (Chenini Fm) Bouaziz et al., 1988; Weishampel, 1990
Aptian Gadoufaoa, Niger (Elrhaz Fm) Taquet, 1976; 1982; 1984

Mee Mayo Djarendi, Cameroon (Koum Basin) Corp eton d Congleton et al., 1992;
28 pnosaund Aptian Agadez, Niger (Elrhaz Fm) Lapparent, 1960
Barremian-

“Aptian Mayo Djarendi, Cameroon (Koum Basin) |Congleton 1990; Colin et al., 1992

Dromaeosaurid |Cenomanian Northern Province, Sudan (Wadi Milk Fm) | Rauhut & Werner, 1995

Theropoda Marsa Matnuh, Egypt (Nubian sandstone); ; WATT aie 4
undescribed Campanian Majunga District, Madagascar Sati, 1902; Mir Rer 1 p; Fester,
and indet. (Maevarano Fm) 2 P Li

Turonian- Rift S ley Province, Kenya (Turkana
rits

Santonian Weishampel, 1990; Wescott et al., 1993

Wagh and Adran Algeria (Continental i
: ntercalaire); aryan, Libya ontinenta e :

Albian Intercalaire):-A. ades, Niger (Continental Lapparent, 1960; Broin et al., 1971
Intercalaire); Medenini, Tunisia (Chenini Fm)

Aptian Agadez, Niger (Elrhaz Fm) Lapparent, 1960; Broin et al., 1974
Tumveipide Mayo Djarendi, Cameroon (Koum Basin) |Congleton, 1990; Colin et al., 1992
pApuan |

Barremian | GPC Prince South Africa (Sundays | Rich et al., 1983; Weishampel, 1990

604 MEMOIRS OF THE QUEENSLAND MUSEUM

TABLE 1 (cont.)

| Taxon | STAGE | __—sSCLOCALITY(ROCKUNIT). | REFERENE è |
Tithonian- Cape Province, South Africa (Upper ; .
Kirkwood; Enon Fm Rich et al., 1983; Mateer, 1987

Theropod Maastrichtian | Agadir, Morocco (Unnamed unit) Weishampel, 1990

footprints
|Cenomanian | Laghout, Algeria (Unnamed unit Weishampel, 1990

Barremian- . a A Flynn et al., 1989; Congleton, 1990;
Mayo Djarendi, Cameroon (Koum Basin) | Coo eieton et al. 1992; Colin etal., 1992

Sauropoda - Diplodocidae

. " t, 1960; Said, ; El-Khashab,
cf craeosaurus |Cenomanian | Marsa Matruh, Egypt (Baharija Fm) Minera » 1960; Said, 1962; El-Khasha

Rebbachisaurus | Cenomanian Kem Kem, Morocco, (Kem Kem Beds) Lavocat, 1954; Sereno et al., 1996

Adrar, Wargla and Tamenghest, Algeria "
Continental Intercalaire); Agadez, Niger |Greigert et al., 1954; Lavocat, 1954;

Albian Continental Intercalaire); Tahoua, Niger |Lapparent, 1960; Taquet, 1976;
arak Fm); Kasr-es-Souk, Morocco (Tegana | Monbaron, 1978

m); Medenini, Tunisia (Chenini Fm)

| Tithonian- Cape Province, South Africa (Upper Broom, 1904; McLachlan & McMillan,
Algoasaurus | Hauterivian — |Kirkwood Fm) "P 1976

A Aptian Agadez, Niger (Elrhaz Fm) Lapparent, 1960; Taquet, 1976, 1982, 1984

Titanosauridae

Sues & Taquet, 1979; Sues, 1980; Forster,

Majunga District, Madagascar ` ,
Guise varus Fm) 8 1926; Sampson et al., 1996; Krause et al.,

Turonian-

Santonian Weishampel, 1990; Wescott et al., 1993

Rift Valley Province, Kenya (Turkana
Grits)

gypt (Baharip Fm); Greigert et al., 1954; Lapparent, 1960;
ak Fm Said, 1962; El-Khashab, 1977

Agadez, Niger (Continental Intercalaire) Greigert et al., 1954; Lapparent, 1960;
" Haughton, 1928; Jacobs et al., 1990, 1992,
199
Agadir, Morocco (Unnamed unit) Weishampel, 1990
: Majunga District, Madagascar Suez & Taquet, 1979; Sues, 1980; Foister,
Mbeveraho Fm) p 1996 1 ‘

Santonian Natal, South Africa (Unnamed unit Kennedy et al., 1987

Agadez, Niger (Unnamed unit); Broin et al., 1974

: 1 a gs Schrank, 1987; Werner, 1993a; 1993b;
| Northern Province, Sudan (Wadi Milk Fm) Rauhut & Werner, 1995
| Aptian Agadez, Niger (In Beceten Fm); Greigert et al., 1954; Broin et al., 1974;
| P Gadoufaoa, Niger (Elrhaz Fm) Taquet, 1976, 1982
Tithonian- Cape Province, South Africa (Upper
| een Pret [Forster 199
| Brachiosauridae

; : Wargla and Adrar, Algeria (Continental

| ?Pleurocoelus Agadez, Niger (Continental Intercalaire Lapparent, 1960
Tithonian-
Hauterivian

Aegyptosaurus

=
S
&
s.
S
E
z
€
z
3
5
E
S
[uz]
R |
=
E
m
E
g
z
8
$
5
io]
a

| Cape Province, South Africa (Upper McLachlan & McMillan, 1976; Rich et
| Kirkwood Fm) ^"

| Brachiosaurid |Tithonian- Cape Province, South Africa T

E mE 75 [ope Poris (Upper [Forier 1996

i

| Camarasauridae

g
-
©
06
o

| Camarasaurid |Tithonian- Cape Province, South Africa r A
erent i [eame (Upper [Rich eral, 1983

CRETACEOUS DINOSAURS OF AFRICA 605

TABLE 1 (cont.)

TAXON STAGE LOCALITY (ROCK UNIT) REFERENCE
Sauropoda indet. | Turonian- Rift Valley Province, Kenya : .
and uridescribed (TurkanaGrits) Weishampel, 1990; Wescott et al., 1993

Marsa Matruh, Egypt (Baharija Fm); Lapparent, 1960; Said, 1962; El-Khashab,
Madagascar (Ankarafantsika Fm) 1977; Mateer et al., 1992

Cenomanian

Adrar, Algeria (Continental Intercalaire
and Unnamed unit); Medenine, Tunisia
(Cheni Pio); Gao, Mali (Continental
ntercalaire)

Karonga, Malawi (Dinosaur Beds); Dixey & Smith, 1929; Dixey, 1937, 1939;
Chirunda Hill, Mozambique (Lupata Pentel'kov &Voronovsky, 1979; Colin &
Series); PUE Valley, Zambia Jacobs, 1990; Jacobs et al., 1990, 1992,
Dinosaur Beds 1993

Lapparent, 1960; Broin et al., 1971;

Albian Bouaziz et al., 1988

EM Mayo Djarendi, (Koum Basin) Brunet et al., 1990
[d sarei Southern Sahara, Niger Sereno et al., 1994

Barremian-
Aptian

Sauro

Flynn et al., 1989; Congleton, 1990;
footprints C 1992

Mayo Djarendi, Cameroon (Koum Basin) lefon etal.

Tithonian: Marrakech, Morocco (Unnamed unit) Dutuit & Ouazzou, 1980

guanodontidae

Ornithopoda - I

Mayo Djarendi, Cameroon (Koum Basin); Ip p Seded oe 390. Cien *

Agadez, Niger (Elrház Fm) 1990; Congleton et al., 1992

Northem Province, Sudan (Wadi Milk Fm); | Rauhut & Werner, 1995; Sereno et al.,
Kem Kem, Morocco, (Kem Kem Beds) 1996

Medenini, Tunisia (Continental "
Intercalaire); Agadez, Niger (Elrhaz Fm) _|Lapparent, 1960; Taquet, 1976

Car a Provinoe, South Africa (Upper Rich et al., 1983; Foster, 1996

Barremian-
Aptian

Cenomanian

Aptian

Ouranosaurus

Iguanodontian
indet.

Tithonian-
Hauterivian

Hypsilophodontidae

| Valdosaurus | Aptian Agadez, Niger (Elrhaz Fm Lapparent, 1960; Taquet, 1976, 1982, 1984
Tithonian- Cape Province, South Africa (Kalahari
Kangnasaurus Hauterivian D pe it Cooper, 1985

Digadsptedonn Northern Province, Sudan (Wadi Milk Fm) | Rauhut & Werner, 1995

Ornithopoda Turonian- Rift Valley Province, Kenya (Turkana ; A

inde Pa Santonian Gnt) S Weishampel, 1990; Wescott et al., 1993
Cenomanian |Northern Province, Sudan (Wadi Milk Fm) | Werner, 1993a; Rauhut & Werner, 1995
Tithonian- Cape Province, South Africa (Upper McLachlan & McMillan, 1976

Hauterivian Kirkwood Fm
Cenomanian Kem Kem, Morocco, (Kem Kem Beds) Sereno et al., 1996

Mayo Djarendi (Koum Basin) Pn poer AA Seen, 1290;

Omithopod -
footprints Barremian-
Aptian

? Pachycephalosauria - ?Pachycephalosauridae

Sues & Taquet, 1979; Sues, 1980;

Majunga District, Madagascar Sampson et al., 1996a,b; Krause et al., in

Majungatholus
?= (Maevarano Fm) press

Campanian

Majungasaurus)

Thyreophora - Stegosauridae

Paranthodon |Tithonian- | Cay MM pat South Africa (Upper Galton & Coombs, 1981; Rich et al., 1983

Nodosauridae

Nodosaurid " Gao, Mali; Agadez, Niger (Continental : ,

indet. Albian Tniercalzire E E Lapparent, 1960; Weishampel, 1990
indet. Mm Eee mi f Mayo Djarendi, Camminont Basin): | Congleton, 3990; Congleton et aly 15922

606

nor the Dinosaur Beds of Malawi can be shown
to be particularly similar to other African
localities. This is interesting with respect to the
differentiation and development of endemism in
the African dinosaur fauna in that it highlights
both the chronological and geographic shortcom-
ings of the record as we now know it.

Ongoing work by groups of researchers in the
Sahara, southeast Africa and Madagascar is
elucidating further the history of dinosaurs on this
great continent. For example, Sereno et al. (1994)
described the theropod Afrovenator, a tetanuran
more primitive than Spinosaurus, and mentions
sauropod remains with unclear affinities, but not
diplodocid or titanosaurid, from the Early
Cretaceous of Niger. The bones were found in
channel fills and show evidence of fluvial
transport. They interpret the area to have had a
uniform climatic regime because of the absence
of caliche and fossil wood without growth rings.
Work in the Cenomanian of Morocco (Sereno et
al., 1996) led to the discovery of the coelurosaur
Deltadromeus and the description of the skull of
the large allosauroid Carcharodontosaurus.
Sereno et al. (1996) and Sereno (1996) concluded
that dinosaurian endemism began abruptly in the
Late Cretaceous.

Late Cretaceous work in Madagascar recently
by field parties led by David Krause have resulted
in significant new discoveries, particularly of
titanosaurids, the theropod Majungasaurus and
birds. These discoveries, and the ones that are sure
to follow, may prove to be particularily enlightening
because Madagascar remained joined to the
Gondwana landmass including India, Australia,
Antarctica and South America into the Early
Cretaceous, long after it had rifted from the west
coast of Africa (Krause & Dodson, 1994; Forster,
1996; Sampson et al., 1996; Krause et al., in press).

ACKNOWLEDGMENTS

Our field work in Africa has been a cooperative
effort among colleagues from Malawi,
Cameroon, France, Britain and the United States.
We wish to thank Yusuf Juwayeyi, the Malawi
Department of Antiquities, and our Malawian
colleagues in the field and laboratory, especially
Fidelis Morocco, John Chilachila, Serino Mithi,
the late James Khomu and the villagers around
Mwakasyunguti and Ngara. We also thank Kent
Newman, Alisa Winkler, Zefe Kaufulu, Laura
MacLatchy, William Downs, Louis Taylor and
John Congleton. For the Cameroon project we
especially thank the Ministére de l'Enseignement

MEMOIRS OF THE QUEENSLAND MUSEUM

Supérieur et de la Recherche Scientifique, Col-
lege de France and Ministere des Relations
Exterieures, Philipe Taquet, M.M. Soba, G.
Ekodeck, Joseph Hell, and all the members of
Project International de Recherché dans le
Cénozoique/Crétacé d'Afrique de l'Ouest
(Cameroun). We thank David Baker for Fig. 5.
The following people allowed access to collec-
tions in their charge: the late Hermann Jaeger and
Wolf-Dieter Heinrich (Museum für Naturkunde,
Berlin); Jose Bonaparte, Guillermo Rougier and
Fernando Novas (Museo Argentino de Ciencias
Naturales, Buenos Aires); Jaime Powell (Museo
de Ciencias Naturales, Tucumán); Angela C.
Milner (The Natural History Museum, London);
Jennifer Clack (University Museum of Zoology,
Cambridge); and Gillian King (South African
Museum). Support for this project has been
provided by The Dinosaur Society, the National
Geographic Society, the U.S. National Science
Foundation (BSR-8700539 to LLJ), the Institute
forthe Study of Earth and Man, Caltex Oil (Malawi)
Limited, the Huntingfield Corporation, American
Airlines, Johnson & Johnson Orthopaedics and
the Carl B. and Florence E. King Foundation.

LITERATURE CITED

ALLIX, P. 1983. Environments mésozoiques de la par-
tie nord-orientale du fossé de la Bénoué (Nigeria):
stratigraphie — sédimentologie — evolution
géodynamique. Travaux des Laboratoires des
Science de la Terre, St. Jéróme Marseille
(Laboratoire associé au Centre National de al
Recherche Scientifique No. 132, 'Etudes
géologiques Ouestafricaines’) Ser. B No. 21, 200 pp.

ARAMBOURG, C. & WOLFF, R.G. 1969. Nouvelles
données paléontologique sur l'age des ‘Grés du
Lubur' (Turkana Grits) á l'ouest du lac Rodolphe.
Comptes Rendus Sommaire Société géologique
de France 1969: 190-192.

BERMAN, D.S. & McINTOSH, J.S. 1978. Skull and
relationships of the Upper Jurassic sauropod
Apatosaurus (Reptilia, Saurischia). Bulletin of
Carnegie Museum of Natural History 8: 1-35.

BONAPARTE, J.F. & POWELL, J.P. 1980. A continen-
tal assemblage of tetrapods from the Upper
Cretaceous beds of El-Brete, northwestern Argen-
tina (Sauropoda-Coelurosauria-Carnosauria-
Aves). Mémoires de la Société géologique de
France 139: 19-28.

BOUAZIZ, S., BUFFETAUT, E., GHANMI, M.,
JAEGER, J.J., MARTIN, M., MAZIN, J.M. &
TONG, H. 1988. Nouvelles découverte de vertébrés
fossiles dans l' Albien du Sud Tunisien. Bulletin
de la Société géologique de France 8: 335-339.

BROIN, F. de, GRENOT, C. & VERNET, R. 1971. Sur
la découverte d'un nouveau gisement de vertébrés
dans le Continental Intercalaire saharien: la Gara

CRETACEOUS DINOSAURS OF AFRICA

Samani (Algérie). Comptes Rendus de
l'Académie des Sciences de Paris, Ser. D 272:
1219-1221.

BROIN, F. de, BUFFETAUT, E., KOENIGUE, J.C.,
RAGE, J.C., RUSSELL, D., TAQUET, P.,
VERGNAUD-GRAZZIN, C. & WENZ, S. 1974.
La faune de vertébrés continentaux du gisement
d'In Beceten (Sénonien du Niger). Comptes Ren-
dus de l'Académie des Sciences de Paris 279:
469-472.

BROOM, R. 1904. On the occurrence of an opis-
thocoelian dinosaur (Algoasaurus bauri) in the
Cretaceous beds of South Africa. Geological
Magazine 1: 445-447.

BRUNET, M., DEJAX, J., BRILLANCEAU, A., CON-
GLETON, J., DOWNS, W., OUPERON-LAV-
DOVENEIX, M., EISENMANN, V., FLANAGAN,
K., FLYNN, L., HEINTZ, E., HELL, J., JACOBS,
L., JEHENNE, Y., NDIENG, E., MOUCHELIN, G.
& PILBEAM, D. 1988. Mise en évidence d'une
sédimentation précoce d'âge Barrémien dans le
fossé de Bénoué en Afrique occidantale (Bassin
du Mayo Oulo Léré, Cameroun), en relation avec
l'ouverture de I’ Atlantique sud. Comptes Rendus
de l' Académie des Sciences de Paris, Ser. II 306:
1125-1133.

BRUNET, M., COPPENS, Y., DEJAX, J., FLYNN, L,.
HEINTZ, E., HELL, J., JACOBS, L., JEHENNE,
Y., MOUCHELIN, G., PILBEAM, D. & SUDRE,
J. 1990. Nouveaux mammiféres du Crétacé
inférieur de Cameroun, Afrique de l' Ouest. Com-
ptes Rendus de l’ Académie des Sciences de Paris,
Ser. II 310: 1139-1146.

BUFFETAUT, E. 1981. Die biogeographische Ges-
chichte der Krokodilier, mit Beschreibung einer
neuen Art, Araripesuchus wegeneri. Geologische
Rundschau 70: 611-624.

1989a. New Remains of Spinosaurus from the
Cretaceous of Morocco. Archosaurian Articula-
tions 1; 65.

1989b. New remains of the enigmatic dinosaur
Spinosaurus from the Cretaceous of Morocco
and the affinities between Spinosaurus and
Baryonyx. Neues Jahrbuch für Geologie und
Paläontologie, Abhandlungen 2: 79-87.

CALVO, J.O. & BONAPARTE, J.F. 1991. Andesaurus
delgadoi gen. et sp. nov. (Saurischia-Sauropoda),
dinosaurio Titanosauridae de la Formación Río
Limay (Albiano-Cenomaniano), Neuquen, Ar-
gentina, Ameghiniana 28: 303-310.

CALVO, J.O. & SALGADO, L. 1995. Rebbachisaurus
tessonei sp. nov. a new Sauropoda from the Al-
bian-Cenomanian of Argentina; New evidence on
the origin of the Diplodocidae. Gaia 11: 13-33.

CARROLL, R.L. 1988. ‘Vertebrate paleontology and
evolution'. (W.H. Freeman and Company: New
York). 698 pp.

CARVALHO, LS. 1994. Candidodon: um crocodilo
com heterodontia (Notosuchia, Cretáceo inferior-
Brasil). Anais da Academia Brasileira de Ciéncias
66; 331-345.

607

CARVALHO, I.S. & CAMPOS, D. d'A. 1988. Um
mamífero triconodonte do Cretáceo Inferior do
Maranhao, Brasil. Anais da Academia Brasileira
de Ciéncias 60: 437-446.

CLARK, J.M., JACOBS, L.L. & DOWNS, W.R. 1989.
Mammal-like dentition in a Mesozoic crocodilian.
Science 244: 1064-1066.

COLIN, J.K. & JACOBS. L.L. 1990. On the age of the
Malawi Dinosaur Beds: evidence from
ostracodes. Comptes Rendus de |’ Académie des
Sciences de Paris Ser. II 331: 1025-1029.

COLIN, J.K., BRUNET, M., CONGLETON, J.D.,
DEJAX, J., FLYNN, L.J., HELL, J. & JACOBS,
L.L. 1992. Ostracodes lacustres des bassins d'áge
Crétacé inférieur du Nord Cameroun Hama-Kous-
sou, Koum, et Babouri-Figuil. Revue de
Paléobiologie 11: 357-372.

CONGLETON, J. 1990. Vertebrate paleontology of the
Koum Basin northern Cameroon, and ar-
chosaurian paleobiogeography in the Early
Cretaceous. MS thesis, Southern Methodist
University, Dallas, Texas, 236 pp.

CONGLETON, J., FLYNN, L.J., JACOBS, L.L.,
BRUNET, M., DEJAX, J., HELL, J. & PIL-
BEAM, D. 1992. Preliminary correlation of con-
tinental sediments of the Koum Basin, northern
Cameroon. Pp. 213-219. In N.J. Mateer & Chen,
P.J. (eds) ‘Aspect of nonmarine Cretaceous
geology’. (China Ocean Press: Beijing).

COOPER, M.R. 1985. A revision of the ornithischian
dinosaur Kangnasaurus coetzeei Haughton, with
classification of the Ornithischia. Annals of South
African Museum 95; 281-317.

DEJAX, J., MICHARD, J., BRUNET, M, & HELL, J.
1989, Emprientes de pas de dinosauriens datées
du Crétacé inférieur dans le Bassin de Babouri-
Figuil (Fossé de la Bénoué, Cameroun). Neues
Jahrbuch für Geologie und Paläontologie, Ab-
handlungen 178: 85-108.

DEPERET, C. 1896. Note sur les dinosauriens
sauropodes et théropodes du Crétacé supérieur de
Madagascar. Bulletin de la Société géologique de
France, Ser. 3 24: 176-196.

DIXEY, F. 1928. The Dinosaur Beds of Lake Nyasa.
Transactions of the Royal Society of South Africa
16: 55-66.

1937. The geology of part of the upper Luangwa
Valley, northeastern Rhodesia. Journal of the
Geological Society of London 93: 52-74.

1939. The Early Cretaceous valley-floor peneplain
of the Lake Nyasa region and its relation to
Tertiary rift structures. Quarterly Journal of the
Geological Society of London 95: 75-108.

DIXEY, F. & SMITH, W.C. 1929. The rock of the
Lupata Gorge and north side of the Lower Zam-
bezi. The Geological Magazine 66: 245-259.

DUTUIT, J.M. & OUAZZOU, A. 1980. Découverte
d'une piste de dinosaure sauropode sur le site
d'empreintes de Demnat (Haut-Atlas Marocain).
Mémoires de la Société géologique de France 59:
95-102.

608

EBY, G. N., RODEN-TICE, M., KRUEGER, H.L.,
EWING, W., FAXON, E.H. & WOOLLEY, A.R.
1995. Geochronology and cooling history of the
northern part of the Chilwa Alkaline Province,
Malawi. Journal of African Earth Sciences 20:
275-288.

EL-KHASHAB, B. 1977. A review of the fossil reptiles
of Egypt. Geological Survey, Egypt Mining
Authority Paper 62: 1-9.

FLYNN, L.J., BRILLANCEAU, A., BRUNET, M.,
COPPENS, Y., DEJAX, J., DUPERON-
LAUDOUENEIX, M., EKODECK, G., FLAN-
NAGAN, K.M., HEINTZ, E., HELL, J.,
JACOBS, L.L., PILBEAM, D.R., SEN, S. &
DJALLO, S. 1987. Vertebrate fossils from
Cameroon, West Africa. Journal of Vertebrate
Paleontology 7: 469-471.

FORSTER, C.A. 1996. The fragmentation of
Gondwana: Using dinosaurs to test biogeographic
hypotheses. Sixth North American Paleontologi-
cal Convention, Abstracts of Papers. Paleon-
tological Society Special Publication 8: 127.

GALTON, P.M. & COOMBS, W.P. 1981. Paranthodon
africanus (Broom) a stegosaurian dinosaur from
the Lower Cretaceous of South Africa, Geobios 3:
299-309.

GILMORE, C.W. 1946. Reptilian fauna of the North
Horn Formation of central Utah. U.S. Geological
Survey, Professional Paper 210c: 1-52.

GOMANI, E.M. 1993. A crocodilian from the Early
Cretaceous Dinosaur Beds from Mwakasyunguti
area, Karonga District, northem Malawi. MS Thesis,
Southern Methodist University, Dallas, 87 pp.

In press, A crocodyliform from the Early Cretaceous
Dinosaur Beds, Northern Malawi. Journal of Ver-
tebrate Paleontology.

GREIGERT, J., JOULIA, F. & LAPPARENT, A.F. de
1954. Repartition stratigraphique des gisements
de vertebres dans le Crétacé du Niger. Comptes
Rendus de l' Académie des Sciences de Paris 239:
433-435.

HANDFORD, C.R. 1987. Turkana Grits — a Cretaceous
braided alluvial system in northern Kenya.
Abstracts, The American Association of
Petroleum Geologists 71: 654.

HAUGHTON, S.H. 1928. On some remains from the
Dinosaur Beds of Nyasaland. Transactions of the
Royal Society of South Africa 16: 69-83.

HUENE, F. von 1929. Los Saurisquios y Ornithisquios
de Cretaceo Argentino. Anales del Museo de La
Plata, 3: 1-196.

HUENE, F. von & MATLEY, C.A. 1933. The
Cretaceous Saurischia and Omithischia of the
central Province of India. Memoirs of the Geologi-
cal Survey of India, Palaeontologia Indica 26: 1-74.

HUNT, A.P., LOCKLEY, M.G., LUCAS, S.G. &
MEYER, C.A. 1994. The global sauropod fossil
record. Gaia 10: 261-279.

JACOBS, B.F. 1990, Lower Cretaceous diaspores from
Md Africa. National Geographic Research 6:

16-518.

MEMOIRS OF THE QUEENSLAND MUSEUM

JACOBS, L.L., CONGLETON, J.D., BRUNET, M.,
DEJAX, J., FLYNN, L.J., HELL, J.V. &
MOUCHELIN, G. 1988. Mammal teeth from the
Cretaceous of Africa. Nature 336: 158-160.

JACOBS, L.L., FLANAGAN, K.M., BRUNET, M.,
FLYNN, L.J., DEJAX, J. & HELL, V. 1989.
Dinosaur footprints from the Lower Cretaceous of
Cameroon, West Africa. Pp. 349-351. In D.D.
Gillette & Lockley, M.G. (eds) 'Dinosaur tracks
and traces'. (Cambridge University Press:
Cambridge).

JACOBS, L.L., WINKLER, D.A. & DOWNS, W.R.
1992. Malawi's paleontological heritage. Oc-
casional Papers of the Malawi Department of An-
tiquities 1: 5-22.

JACOBS, L.L., WINKLER. D.A., DOWNS, W.R. &
GOMANI, E.M. 1993. New material of an Early
Cretaceous titanosaurid sauropod dinosaur from
Malawi. Palaeontology 36: 523-534.

JACOBS, L.L., WINKLER, D.A., KAUFULU, Z.M. &
DOWNS, W.R. 1990. The Dinosaur Beds of
northern Malawi, Africa. National Geographic
Research 6: 196-204.

JANENSCH, W. 1922. Das Handskelett von Gigan-
tosaurus robustus und Brachiosaurus brancai aus
den Tendaguru-Schichten Deutsch-Ostafrikas.
Centralblatt. Mineralogie Geologie Palaeon-
tologie 1922: 464-480.

1961. Die Gliedmaszen und Gliedmaszengürtel der
Sauropoden der Tendaguruschichten. Palaeon-
tographica (suppl. 7) 3: 177-235.

KAUFULU, Z.M. 1989, Sedimentary conditions within
a Plio-Pleistocene graben at Karonga, northern
Malawi and their implications for the prehistoric
record of the area. Palaeoecology of Africa 20:
99-108.

KELLNER, A.W.A. 1994. Comments on the
paleobiogeography of Cretaceous archosaurs
during the opening of the South Atlantic Ocean.
Acta Geologica Leopoldensia 17: 615-625.

1996. Remarks on Brazilian dinosaurs. This volume.

KENNEDY, W. J. & KLINGER, H.C. 1972. A
Texanites-Spinaptychus association from the
Upper Cretaceous of Zululand. Palaeontology 15:
394-399.

KENNEDY, W. J., KLINGER, H.C. & MATEER, NJ.
1987. First record of an Upper Cretaceous
sauropod dinosaur from Zululand, South Africa,
South African Journal of Science 83: 173-174.

KENT, V.D. & GRADSTEIN, E.M. 1985. A Cretaceous
and Jurassic geochronology. Geological Society
of America Bulletin 96: 1419-1427.

1986. A Jurassic to recent chronology. Pp. 45-50. In
Vogt, P.R. & Tucholke, B.E. (eds) ‘The geology
of North America Vol. M. The western North
Atlantic region'. (Geological Society of America.)

KLITZSCH, E. & SQUYRES, C.H. 1990. Paleozoic
and Mesozoic geological history of northeast
Africa based upon new interpretation of Nubian
strata. The American Association of Petroleum
Geologists 74: 1203-1211.

CRETACEOUS DINOSAURS OF AFRICA

KLITZSCH, E., HARMS, L.C., LEJAL-NICOL, A. &
LIST, F.K. 1979. Major subdivisions and deposi-
tional environments of Nubia strata, southwest
Egypt. The American Association of Petroleum
Geologists 63: 967-974.

KRAUSE, D.W. & DODSON, P. 1994. The premaxilla
of Majungasaurus (Theropoda), Late Cretaceous,
Madagascar: Implications for relationships. Jour-
nal of Vertebrate Paleontology 14: 32A. (Abstract)

KRAUSE, D.W., HARTMAN, J.W. & WELLS, N.A.
In press. Late Cretaceous vertebrates from
Madagascar: Implications for biotic change in
deep time; In Patterson, B. & Goodman, S. (eds)
‘Natural and Human-induced Change in
Madagascar’ (Smithsonian Institution Press:
Washington, D.C.).

KUES, B.S., LEHMAN, T.M. & RIGBY, J.K., Jr 1980.
The teeth of Alamosaurus sanjuanensis, a Late
Cretaceous sauropod. Journal of Paleontology 54:
864-869.

LAPPARENT, F.A. de 1960. Les dinosauriens du ‘Con-
tinental Intercalaire' du Sahara central. Mémoire
de la Société géologique de France 88a: 1-57.

LAVOCAT, R. 1954, Sur les dinosauriens du Continen-
tal Intercalaire des Kem-Kem de la Daoura. Cong.
Geol. Int. Alger, Quest. Divers., 3^ Part., pp. 65-68.

LEFRANC, J.C. & GUIRAUD, R. 1990. The Continen-
tal Intercalaire of northwestern Sahara and its
equivalents in the neighboring regions. Journal of
African Earth Sciences 10: 27-77.

LUCAS, S.P. & HUNT, A.P. 1989. Alamosaurus and the
sauropod hiatus in the Cretaceous of the North
American Western Interior. Geological Society of
America Special Paper 238: 75-85

MATEER, N.J. 1987. A new report of a theropod
dinosaur from South Africa. Palaeontology 30:
141-145.

MATEER, N.J., WYCISK, P., JACOBS, L.L.,
BRUNET, M., LUGER, P., ARUSH, M.A.,
HENDRICKS, F., WEISSBROD, T.,
GVIRTMAN, G., MBEDE, E., DINA, A.,
MOODY, R.T.J., WEIGELT, G., EL-NAKHAL,
H.A., HELL, J. & STETS, J. 1992. Correlation of
nonmarine Cretaceous strata of Africa and the
Middle East. Cretaceous Research, 13: 273-318.

McINTOSH, J.S. 1990. Sauropoda. Pp. 345-401. In
Weishampel, D.B., Dodson, P. & Osmolska, H.
(eds) ‘The Dinosauria’ (University of California
Press: Berkeley).

McLACHLAN, I.R. & McMILLAN, LK. 1976.
Review and stratigraphic significance of southern
Cape Mesozoic palaeontology. Transactions of
the Geological Society of South Africa 79: 197-
212.

MICHARD, J., BROIN, F. de, BRUNET, M. & HELL,
J. 1990. The oldest specialized neosuchian
crocodile from Africa (Early Cretaceous,
Cameroon), with ‘eusuchian’ characters. Comptes
Rendus de I’ Académie des Sciences de Paris, Ser.
IL 311: 365-371.

609

MONBARON, M. 1978. Nouveaux ossement de
dinosauriens de grande taille dans le bassin Juras-
sico-Crétacé de Taguelft (Atlas de Beni Mellal,
Maroc). Comptes Rendus de l'Académie des
Sciences de Paris 287: 1277-1279.

MOODY, R.T.J. & SUTCLIFFE, J.C. 1991. The
Cretaceous deposits of the Iullemmeden Basin of
Niger, central west Africa. Cretaceous Research
12: 137-157.

MOURA, J.A., ARAI, M., BRITO, LA.M., CAMPOS,
D.d’A., CARVALHO, M.S.S., CASSAB, R.C.T.,
COIMBRA, J.C., DUARTE, L., HASHIMOTO,
A.T., LIMA, M.R. & SANTOS, E.M. 1994. The
Recóncavo-Tucano-Jatobá Basin. Pp. 125-160. In
Beurlen, G., Campos, D. d'A. & Viviers, M.C.
(eds) 'Stratigraphic range of Cretaceous Mega-
and Microfossils of Brazil' (Universidade Federal
do Rio de Janeiro, Centro de Ciéncias
Matematicas e da Natureza, Instituto de
Geociéncias: Rio de Janeiro).

PENTEL'KOV, V.G. & VORONOVSKY, S.N. 1979,
Radiometric age of the Mbalizi carbonatite, Tan-
zania, and correlation with other carbonatites of
the Rukwa-Malawi Rift zone. Doklady
Akademiia Nauk SSSR 235: 92-94.

PITTMAN, W.C. III, CANDE, S., La BRECQUE, J. &
PINDELL, J. 1993, Fragmentation of Gondwana:
The separation of Africa from South America. Pp.
15-34. In Goldblatt, P. (ed.) ‘Biological relation-
ships between Africa and South America' (Yale
University Press: New Haven).

PONTE, E.C. 1994. The geology of the Brazilian
Cretaceous sedimentary basins. Pp. 1-21. In Beur-
len, G., Campos, D. d'A. & Viviers, M.C. (eds)
‘Stratigraphic range of Cretaceous Mega- and
Microfossils of Brazil' (Universidade Federal do
Rio de Janeiro, Centro de Ciéncias Matematicas e
da Natureza, Instituto de Geociéncias: Rio de
Janeiro).

POPOFF, M. 1988. Du Gondwana à l'Atlantique sud;
les connexions du fossé de la Bénoué avec les
bassins du Nord-Est Brésilien jusqu'à l'ouverture
du golfe de Guinée au Crétacé inférieur. Journal
of African Earth Sciences 7: 409-431.

POWELL, J.E. 1986. Revision de los Titanosauridos de
America del Sur. Ph.D. Dissertation, Universidad
Nacional de Tucumán, Argentina. 340 pp.

1987. The Late Cretaceous fauna of the Los
Alamitos, Patagonia, Argentina. Part IV The
titanosaurids. Paleontologia 3: 148-153.

RAUHUT, O.W.M. & WERNER, C. 1995. First record
of the family Dromaeosauridae (Dinosauria:
Theropoda) in the Cretaceous of Gondwana (Wadi
Milk Formation, northern Sudan).
Paláontologische Zeitschrift 69: 475-489.

RICH, T.H., MOLNAR, R.E. & RICH, P. V. 1983. Fossil
vertebrates from the Late Jurassic or Early
Cretaceous Kirkwood Formation, Algoa Basin,
Southern Africa. Transactions of the Geological
Society of South Africa 86: 281-291.

610

RUSSELL, D., RUSSELL, D., TAQUET, P. & THOMAS,
H. 1976. Nouvelles récoltes de vertebres dans les
terrains continentaux du Crétacé Supérior de la
region de Majunga (Madagascar). Comptes Ren-
dus Sommaire Société géologique de France 5:
204-208.

SAID, R. 1962. ‘The geology of Egypt’. (Elsevier
Publishing Company: Amsterdam) 377 pp.
SALGADO, L. & BONAPARTE, J.F. 1991. Un nuevo
saurópodo Dicraeosauridae, Armagasaurus
cazaui gen. et sp. nov., de la Formación La Amar-
ga, Neocomian de la Provincia del Neuquen, Ar-

gentina. Ameghiniana 28: 333-346.

SAMPSON, S.D., FORSTER, C.A., KRAUSE, D.W.
& DODSON, P. 1996a. New dinosaur discoveries
from the Late Cretaceous of Madagascar: Implica-
tions for Gondwanan biogeography. Sixth North
American Paleontological Convention, Abstracts of
POE Paleontological Society Special Publication

SAMPSON, S.D., KRAUSE, D.W., FORSTER, C.A.,
DODSON, P. & RAVOAVY, F. 1996b. Non-avian
theropod dinosaurs from the Late Cretaceous of
Madagascar and their paleobiogeographic im-
plications. Journal of Vertebrate Paleontology
16(Sup): 62A.

SCHRANK, E. 1987. Biostratigraphic importance of
microfloras from the Late Cretaceous clastic series
of northwestern Sudan. Cretaceous Research 8:
29-42,

SEELEY, H.G. 1876. On Macrurosaurus semnus
(Seeley), a long-tailed animal with procoelous
vertebrae from the Cambridge Upper Greensand,
preserved in the Woodwardian Museum of the
University of Cambridge. Quarterly Journal of the
Geological Society, London 32: 440-444.

SERENO, P.C. 1996. Lineage sorting and dispersal
among Cretaceous dinosaurs. Sixth North
American Paleontological Convention, Abstracts
of Papers. Paleontological Society Special Publi-
cation, 8: 353.

SERENO, P.C., DUTHEIL, D.B., IAROCHENE, M.,
LARSSON, H.C.E., LYON, G.H., MAGWENE,
P.M., SIDOR, C.A., VARRICCHIO, D.J. & WIL-
SON, J.A. 1996. Predatory dinosaurs from the
Sahara and Late Cretaceous faunal differentiation.
Science 272: 986-991.

SERENO, P.C., WILSON, J.A., LARSSON, H.C.E.,
DUTHEIL, D.B. & SUES, H.D. 1994. Early
Cretaceous dinosaurs from the Sahara. Science
266: 267-271.

STROMER, E. 1932. Ergebnisse der Forschungsreisen
Prof. E. Stromers in den Wüsten Agyptens. II.
Wirbeltierreste der Baharije-Stufe (unterstes
Cenoman). 11. Sauropoda. Abhandlungen der
Bayerischen Akademie der Wissenschaften Math-
ematisch-naturwissenschaftliche Abteilung, NF
10: 1-21.

SUES, H.D. 1980. A pachycephalosaurid dinosaur from
the Upper Cretaceous of Madagascar and its im-
plications, Journal of Paleontology 54: 954-962.

MEMOIRS OF THE QUEENSLAND MUSEUM

SUES, H.D. & TAQUET, P. 1979. A pachycephalo-
saurid dinosaur from Madagascar and Laurasia-
Gondwanaland connection in Cretaceous. Nature
279: 633-634.

TAQUET, P. 1976. Géologie et paléontologie du gise-
ment du Gadoufaoua (Aptien du Niger). Cahiers
de Paléontologie x: 1-191.

1982. The Aptian fossil locality of Gadoufaoua.
National Geographic Society Research Reports
14: 649-653.

1984, Une curieuse spécialisation du crâne de cer-
tains dinosaures carnivores du Crétacé: le museau
long et étroit de Spinosauridés. Comptes Rendus
de l'Académie des Sciences de Paris 299: 217-
222.

TIERCELIN, J.J., CHOROWICZ, J., BELLON, H., et
al. 1988. East African Rift System: offset, age, and
tectonic significance of the Tanganyika-Rukwa-
Malawi intracontinental transcurrent fault zone.
Tectonophysics 148: 241-252.

TILLEMENT, B. 1971. Hydrogeologie du Nord-
Cameroun. Republique Federale du Cameroun,
Direction des Mines et de la Geologie, Bulletin 6:
1-294.

UPCHURCH, P. 1994. Sauropod phylogeny and
palaeoecology. Gaia 10: 249-260.

1995, The evolutionary history of sauropod
dinosaurs. Philosophical Transactions of the
Royal Society of London 349: 365-390.

WEISHAMPEL, D.B. 1990. Dinosaurian distribution.
Pp. 63-139. In Weishampel, D.B., Dodson, P. &
Osmolska, H. (eds) ‘The Dinosauria’ (University
of California Press: Berkeley).

WERNER, C. 1993a. Aspects on terrestrial Upper
Cretaceous ecosystems of Egypt and northern
Sudan. Contribution from the Paleontological
Museum, University of Oslo 364: 71-72.

1993b. Late Cretaceous continental vertebrate
faunas of Niger and northern Sudan. Pp. 401-405.
In Thorweihe, U. & Schandelmeier, H. (eds)
*Geoscientific Research in Northeast Africa’
(Balkema: Rotterdam).

WESCOTT, W.A., MORLEY, C.K. & KARANJA,
E.M. 1993. Geology of the "Turkana Grits' in the
Lariu Range and Mt. Porr areas, southern Lake
Turkana, northwestern Kenya. Journal of African
Earth Sciences 16: 425-435.

WILLIAM, P.G. & SAVAGE, R.J.G. 1986. Early rift
sedimentation in the Turkana Basin, northern
Kenya. Pp. 265-283. In Frostick, L.E., Renaut,
R.W., Reid, I. & Tiercelin, J.J, (eds) 'Sedimenta-
tion in the African Rifts’ (Blackwell Scientific
Publications: Oxford).

WU X, C., SUES, H.D. & SUN, A. 1995. A plant-eating
crocodyliform reptile from the Cretaceous of
China. Nature 376: 378-680.

WYCISK, P. 1990. Stratigraphic update of the non-
marine Cretaceous from SW Egypt and NW
Sudan. Cretaceous Research 12: 185-200.

REMARKS ON BRAZILIAN DINOSAURS
ALEXANDER W.A. KELLNER

Kellner, A.W.A. 1996 12 20: Remarks on Brazilian Dinosaurs. Memoirs of the Queensland
Museum 39(3): 611-626. Brisbane. ISSN 0079-8835.

To date the dinosaur record in Brazil is limited to saurischians and no confirmed evidence
(except probably footprints) of ornithischians is presently known. Two dinosaur taxa were
described from the Triassic Santa Maria Formation: Spondylosoma absconditum and
Staurikosaurus pricei. The dinosaurian affinities of Spondylosoma are questioned in the
literature; Staurikosaurus is regarded by some authors as closely related to Herrerasaurus
from Argentina. No Jurassic dinosaurs are known from Brazil, but there are several
Cretaceous occurrences. Titanosaurid sauropods (Antarctosaurus brasiliensis, several un-
described postcranials and teeth) and theropods (teeth) are present in the Late Cretaceous
strata of the Bauru Group. Spinosaurid theropods have been found in the Albian strata of
northeast Brazil and are apparently closely related to some African forms. This distribution
pattern may be a product of an Early Cretaceous vicariant event. Spinosaurids as regarded
here are not an exclusive Gondwanan group and are also present in Europe (Baryonyx). A
literal interpretation of the stratigraphic record, therefore, suggests that dinosaur faunal
interchanges between Europe and Gondwana may have occurred not only from South to
North as previously supposed, but in the opposite direction as well. This hypothesis is very
preliminary and can be tested by new and more complete dinosaur material from Brazil and
Africa. [ ] Brasil, dinosaurs, Triassic, Cretaceous, spinosaurid, soft tissue.

Alexander W.A. Kelner, Museu de Ciéncias da Terra/Setor de Paleontologia, Departamento
Nacional da Produgáo Mineral, Rio de Janeiro/ Fellow Conselho Nacional de Desenvol-

vimento Científico e Tecnólogico- CNPq (Brazil); 5 April 1996.

Dinosaur remains are recorded from several
localities in Brazil (Fig. 1). To date only remains
of saurischian dinosaurs are known in this
country and no confirmed evidence (except some
possibly ichnofossils) of ornithischian taxa have
been found (see Campos & Kellner, 1991, for a
review). Many specimens have been collected,
but most are still undescribed. Apart from the
considerable work done on footprints and track-
ways, (e.g., Leonardi, 1984), little is actually
known from the Brazilian dinosaur fauna.

This paper reviews the most important dinosaur
occurrences in Brazil, particularly those from the
Santa Maria Formation (Triassic), Santana For-
mation (Early Cretaceous) and Bauru Group
(Late Cretaceous). The preliminary description of
new specimens (mostly teeth) and the
palaeobiogeographic information that can be in-
ferred based on the present knowledge of the
Brazilian dinosaurs is also discussed.

SANTA MARIA FORMATION

Triassic dinosaurs in Brazil are only found in
the Santa Maria Formation (Parana Basin, south
Brazil) which is the most fossiliferous unit of that
period in this country. Rhynchosaurs, cynodonts
and dicynodonts are frequently found preserved
in red sandstones and siltstones. The age of this

unit is regarded as ranging from Anisian to Car-
nian based essentially on the fossil vertebrate
assemblages (Barberena et al., 1985).

Two supposed dinosaurs have been reported
from the Santa Maria Formation. The first one is
Spondylosoma absconditum named by Huene
(1942) based on some postcranial elements. Col-
bert (1970) mentioned that two small and com-
pressed teeth with serrated edges may also belong
to the same specimen. Romer (1956) listed Spon-
dylosoma within Prosauropoda, which was fol-
lowed by others (e.g., Colbert, 1970).
Unfortunately no recent detailed description of
this material is available which could provide
more evidence about the systematic position of
this taxon. According to Sues (1990), the
dinosaurian affinity of Spondylosoma has not
been conclusively presented so far.

A second dinosaur from this unit, Stauriko-
saurus pricei, is only known by the type
specimen. This taxon was named by Colbert
(1970) on the basis of an incomplete skeleton.
Another partial postcranial skeleton found in the
Late Triassic Ischigualasto Formation from Ar-
gentina was attributed to cf. Staurikosaurus sp.
(Brinkman & Sues, 1987), but is now referred to
Herrerasaurus (Novas, 1994).

612

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 1. Map showing the most important Brazilian dinosaur localities known so far (after Campos & Kellner,
1991). In Rio Grande Do Sul: 1, Santa Maria. In São Paulo: 2, Presidente Bernardes; 3, Adamantina; 4, Pacaembu
Paulista; 5, Guararapes; 6, Sáo José do Rio Preto; 7, Ibirá; 8, Colina; 9, Monte Alto. In Minas Gerais: 10,
Uberaba/Peirópolis. In Mato Grosso: 11, Morro do Cambambe. In Ceará: 12, Sobradinho; 13, Chapada do
Araripe. In Paraíba: 14, Rio do Peixe. In Maranháo: 15, Bacia de Sáo Marcos. In Amazonas: 16, Nova Olinda

do Norte.

The phylogenetic position of Staurikosaurus
has been controversial. In the original descrip-
tion, Colbert (1970) placed this taxon in the
Saurischia, regarding it as possibly a carnivorous
prosauropod (see Galton,1977). Benedetto
(1973) considered Staurikosaurus closely related
to Herrerasaurus from the Ischigualasto Forma-
tion (Argentina) and classified both in the Her-

rerasauridae, which he excluded from
Theropoda. Galton (1973) suggested that
Staurikosaurus was probably a primitive
theropod, but classified this taxon as Saurischia
incertae sedis. Later, Galton (1977) placed
Staurikosaurus in its own higher taxonomic unit
(Staurikosauridae), separating the Brazilian
species from Herrerasaurus, and considered both

REMARKS ON BRAZILIAN DINOSAURS

613

FIG. 2. Brazilian spinosaurid Angaturama limai; right lateral view of the anterior portion of the skull (Universidade
de São Paulo - USP, São Paulo: GP/2T-5). m=maxilla, pm-premaxilla, sc=sagittal crest; scale = 50mm.

as Saurischia incertae sedis. Brinkman & Sues
(1987) also disagreed with the hypothesis of close
relationship between Staurikosaurus and Her-
rerasaurus, tegarding both taxa as successive
outgroups for Saurischia + Ornithischia. More
recently Novas (1992, 1994) and Sereno & Novas
(1992), based on the study of new material of
Herrerasaurus, suggested that the Argentinean
form was closely related to Staurikosaurus, and
considered both as basal theropods of the
monophyletic group Herrerasauridae.

SANTANA FORMATION

The Santana Formation (Araripe Basin, north-
east Brazil) is one of the most fossiliferous units
in the world. This formation is divided into three
members, named from bottom to top: Crato, Ipubi
and Romualdo (Beurlen, 1971). According to
palynological data (Pons et al., 1990), the age of
those strata varies from Aptian-Albian (Crato
Member) to middle Albian (Romualdo Member).

The palaeontological content of the Santana
Formation is very diverse. The Crato Member is
very rich in plants, insects and fishes, but includes
pterosaurs (Frey & Martill, 1994; Campos &

Kellner, 1996) and birds (so far only feathers).
The Romualdo Member is well known for the
variety and quantity of fishes, but plants,
pterosaurs and crocodilians are also found (see
Maisey, 1991). Dinosaur material is rare and
presently restricted to the calcareous nodules of
the Romualdo Member.

The first dinosaur reported from the Santana
Formation is an isolated bone, tentatively iden-
tified as an ischium of an ornithischian dinosaur
(Leonardi & Borgomanero, 1981). Although
probably dinosaurian, the fragmentary nature of
this specimen casts doubt about its anatomical
identification. This material needs further
preparation and at this point should be regarded
as Dinosauria incertae sedis.

Another dinosaur specimen reported from the
Romualdo Member is the anterior portion of a
skull (incomplete premaxilla-maxilla; Figs 2, 3)
and represents a new theropod (Kellner, 19942),
Angaturama limai (Kellner & Campos, 1996).
The rostrum of this dinosaur is very compressed
(more than in any other theropod) and the distal
portion is expanded (‘spoon-shaped’). A sagittal
crest formed by the premaxillae is present on the
anterior part of the snout. This specimen repre-

614 MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 3. Spinosaurid Angaturama limai (USP GP/2T-5). A, Stereopair of the palatal view. pm=premaxilla; scale
= 50mm. B, Right 6th premaxillary tooth, stereopair of the transverse section; scale = 10mm. C, Lingual view;
scale = 10mm.

sents the first occurrence of a spinosaurid in found in the Cenomanian rocks of Egypt
Brazil. (Stromer, 1915) and the spinosaurids found in

Spinosauridae is considered a group of unusual Early Cretaceous strata of Gadoufaoua, Niger
theropods that includes Spinosaurus aegyptiacus (Taquet, 1984). Baryonyx found in the Barremian

REMARKS ON BRAZILIAN DINOSAURS

615

FIG. 4. Isolated spinosaurid tooth from Cretaceous strata of Morocco, housed in the Departamento Nacional da
Produção Mineral (DNPM), Rio de Janeiro. A, Stereopair of the lingual view; scale = 10mm. B, Posterior view;
note the wear surface on the basal portion of the tooth; scale = 10mm.

of England (Charig & Milner, 1986) is considered
a spinosaurid too (Paul, 1988; Buffetaut, 1989a;
Kellner, 1994a; Kellner & Campos, 1996; but see
Charig & Milner, 1990). Buffetaut (1989a) also
attributed to Spinosauridae (Spinosaurus cf. S.
aegyptiacus) a fragmentary maxilla with almost

circular alveoli (no teeth were preserved) found
in continental red beds of Albian-Cenomanian
age in the Taouz region, southern Morocco. More
materials from Morocco referable to
Spinosauridae are isolated teeth, preserved in red-
dish sandstone (Fig. 4) that were collected near

616

the region of Ksar-es Souk (Henry Galiano, pers.
comm. 1994) and come possibly from the Tegana
Formation (Albian-Cenomanian?). Those teeth
are very similar to the ones reported from Albian
strata of Southern Tunisia, which are attributed to
Spinosaurus sp. (Bouaziz et al., 1988).

A potential synapomorphy of Spinosauridae
that can be observed in the Brazilian specimen is
the ‘crocodilian-like’ teeth, with subcircular
transverse sections and striated enamel. The
anterior and posterior keels of the teeth in the
Brazilian spinosaurid are unserrated, which is the
same condition as in Spinosaurus aegyptiacus
and the teeth found in Tunisia and Morocco (Fig.
4), but different from the finely serrated teeth in
Baryonyx (Charig & Milner, 1986; 1990). Other
derived characters shared by spinosaurids are: the
particular shape of the rostrum (concave ventral
margin of the upper jaw, which corresponds to a
convex dorsal margin of the lower jaw); the
presence of seven premaxillary teeth; and, the
extemal naris displaced backwards (observed in
Baryonyx and indicated in the specimens found
in Gadoufaoua and in Brazil). The last two fea-
tures cannot be observed in Spinosaurus aegyp-
tiacus since the upper jaw of this species remains
unknown.

Recently Martill et al. (1996) described another
theropod from the Romualdo Member, /rritator
challengeri, which they regarded as representing
a new clade (Irritatoridae) of maniraptoran
dinosaurs. The materia] includes an almost com-
plete skull and mandible which lacks the rostral
portion.

Judging from the published picture, this
specimen was still unprepared at the time of their
study (Martill et al., 1996: 6). It is also not clear
if the matrix from several cranial openings has
been completely removed or if those openings are
only indicated on the picture of the specimen
(compare Martill et al., 1996, Figs 2 & 3).

Despite listing several characters that Z. chal-
lengeri apparently shares with different
theropods, the authors have not made clear why
they regard it as a maniraptoran. The only cranial
character used by Gauthier (1986) to diagnose the
Maniraptora is the absence or reduction of the
prefrontal. The presence or absence of this bone
cannot be verified in 7. challengeri since Martill
et al. (1996, Fig. 3) listed this bone but failed to
indicate its presence (and consequently its shape
and proportions) in the skull. Holtz (1994: 1107),
reviewing the phylogeny of theropods, listed as a
cranial synapomorphy of maniraptorans the
‘jugal expressed on the rim of the antorbital

MEMOIRS OF THE QUEENSLAND MUSEUM

fenestra’ . If the cranial sutures indicated by Mar-
till et al. (1996) are correct, the jugal of 7. chal-
lengeri does not participate in, the antorbital
fenestra. All other synapomorphies of the
Maniraptora are based on postcranial bones
(Gauthier, 1986; Holtz, 1994).

Curiously, Martill et al. (1996) did not compare
the skull of 7. challengeri with the spinosaurid,
Angaturama limai, from the same deposit
(Kellner, 1994a; Kellner & Campos, 1996). Nor
did Martill et al. (1996) compare their specimen
with Baryonyx described by Charig & Milner
(1986; 1990). They did, however, notice
similarities between the dentitions of /. chal-
lengeri and Spinosaurus from Egypt (Stromer,
1915), but pointed out that the ‘mandible of
Spinosaurus would certainly not fit with the den-
tal margin of the maxilla and premaxilla' of their
material (Martill et al., 1996: 8). It is intriguing
how this conclusion was reached, since their
specimen lacks the rostral part of the skull.

Preliminary comparison of /. challengeri with
other spinosaurid taxa indicates that they share at
least two synapomorphies: transverse section of
the teeth sub-circular and external nares dis-
placed backwards. Those features suggest that Z.
challengeri is a spinosaurid and so there seems
no justifiation for the Irritatoridae, here con-
sidered a junior synonym of the Spinosauridae.

The comparison of /. challengeri with An-
gaturama limai is difficult since they are based
on different parts of the skull. Both, however,
have unserrated teeth. This feature, associated
with the fact that they come from the same
deposit, raises the possibility that they represent
the same taxon. The preserved posterior portion
of the skull of A. limai, however, is apparently
higher and more laterally compressed than the
preserved anterior portion of the skull of Z. chal-
lengeri. Although there is a good possibility that
the specimens belong to the same taxon, regard-
ing them as synonymous is premature until more
complete material is found.

Another dinosaur specimen from the Santana
Formation mentioned in the literature is an in-
complete sacrum (Frey & Martill, 1995). The
material, only partially prepared, was attributed
to a possible oviraptorosaurid on the basis of the
presence of pleurocoels. Since oviraptorosaurids
are a highly specialised group of theropods whose
presence has only definitely been recorded from
North America and Asia (Barsbold et al., 1993),
the assignment of this incomplete material to this
group should be viewed with caution, as indeed
was pointed out by Frey & Martill (1995).

REMARKS ON BRAZILIAN DINOSAURS

617

FIG. 5. Small theropod from the Santana Formation, Araripe Basin (MCT 1502-R). A, Complete nodule; scale
= 50mm. B, Detail of soft tissue between femur, tibia and fibula; scale = 10mm. C, Detail of soft tissue near
one pedal phalanx; scale = 10mm. cd=caudal vertebrae, d=digits of pes, fe=femur, fi=fibula, mt=metatarsals,
pe=remains of pelvis, ph=phalanx of pes, s=soft tissue, ti=tibia.

Further dinosaur material from the Santana
Formation includes limb bones, one series of nine
vertebrae (three sacrals and six caudals), and a
complete pelvis with articulated posterior limb
elements and several vertebrae (dorsal, sacral and
caudal). Recently a calcareous nodule containing
both hindlimbs, partial pelvis and several ver-
tebrae of a small theropod (length of femur ap-
proximately 175mm) was found (Figs 5, 6). This
specimen is particularly striking because of the
soft tissue associated with many bones. Since
epidermis and muscle fibers are preserved in
three dimensions, this might be the best fossilised
soft tissue of a dinosaur known so far (Kellner,
1996). Furthermore, longitudinal and transverse

sections through a fragment of one femur showed
the presence of rod-like structures filling the
channels for blood vessels. These structures might
be either mineralisations filling those channels or
they could represent the replacement of the actual
blood vessels. All this material is currently being
prepared and will be described elsewhere.

BAURU GROUP

The Bauru Group comprises one of the most
extensive continental sedimentary sequences in
South America. Outcrops of this unit are found in
many Brazilian States (Parana, Sáo Paulo, Minas
Gerais, Goiás and Mato Grosso do Sul) and ex-

618

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 6. Scanning electron micrographs of soft tissue preserved in the small theropod (MCT 1502-R) from the
Santana Formation, A, Muscle fibres; scale = 50j.m. B, Longitudinal section through a fragment of the femur
showing rod-like structures preserved in the channels for the capillary blood vessels of the bone; scale = 100j.m;
C, Transverse section through the same bone showing the rod-like structures; scale = 501m; D, Detail of one
of the rod-like structures; scale = 5pm.

tend into Paraguay. The most widely accepted
stratigraphic subdivision for those strata was pro-
posed by Soares et al. (1980), who recognised
four formations: Caiuá, Santo Anastácio,
Adamantina and Marília. To these, Souza (1984)
included the Uberaba Formation, which is inter-
digitated with the Adamantina Formation. Essen-
tially all these stratigraphical units are composed
by sandstones and siltstones, deposited in a semi-
arid environment by a meandering river system,
and all are of Late Cretaceous age. Among the

fossil vertebrates reported are crocodilians,
dinosaurs, turtles and fishes, most of which came
from the upper layers (Adamantina/Uberaba,
Marília formations).

Although dinosaur remains from the Bauru
strata were reported at the beginning of this cen-
tury (Ihering, 1911) and more specimens were
collected since, very little is known about the
dinosaurian fauna of this unit. A few incomplete
specimens were assigned to European and North
American taxa, which may be erroneous (Cam-

REMARKS ON BRAZILIAN DINOSAURS

pos & Kellner, 1991). The only new species
described is the sauropod Antarctosaurus
brasiliensis, based on very fragmentary materia]
(incomplete humerus, femur and dorsal vertebra)
collected in Sáo Paulo (Arid & Vizotto, 1971).
Antarctosaurus was first reported in Cretaceous
sediments of Argentina and regarded as a
titanosaurid (Huene, 1929; McIntosh, 1990a).
More recently, based on the jaw morphology,
Jacobs et al. (1993) questioned the titanosaurid
affinities of the Argentinean genus, and regarded
Antarctosaurus as a diplodocid. Due to the frag-
mentary nature of the material referred to An-
tarctosaurus brasiliensis, there is no particular
feature that would support any particular generic
affinity of the Brazilian species.

Recently Bertin et al. (1993) reported the
presence of an incomplete upper jaw and one
tooth referred to Abelisauridae, and an incom-
plete dentary of a coelurosaur. These potentially
important specimens, however, were not
described or figured and no diagnostic feature
was presented that could support their assign-
ment. The presence of an abelisaur in the Bauru
group would be of considerable interest for the
palaeobiogeography, since Abelisauridae are
considered to have a Gondwanan distribution
(Bonaparte & Novas, 1985).

Most of the dinosaur collecting in the Bauru
strata was done by Llewellyn Ivor Price near
Uberaba and Peirópolis in Minas Gerais where an
ongoing excavation is currently being carried out
(Campos, 1993). The work resulted in an exten-
sive collection of postcranial dinosaur material
which has yet to be described.

The most common dinosaur remains found near
Peiropolis are sauropod and theropod teeth
(Kellner, 1994b). The sauropod teeth are typical-
ly pencil-like: long, slender and subcircular in
cross-section. The labial side is slightly more
inflated than the lingual side. The enamel has a
brownish colour and a rough surface (Fig. 7).
Small, unserrated carinae can be observed in well
preserved specimens, although in most they are
absent (possibly worn down). In some teeth the
enamel is almost completely lost, exposing the
dentine that has a light color and a smooth sur-
face. Wear surfaces are present and vary possibly
according to the tooth’s position in the jaw and its
age. Based on wear surfaces, the following tooth
classes have been recognised:

1) teeth with wear surfaces present on the lin-
gual side and either absent or very limited on the
labial side (Figs 7D-F);

619

2) teeth with wear surfaces present on the lin-
gual and labial side (Figs 7A-C);

3) teeth with wear surfaces present on different
parts of the tooth.

Most teeth are of class two, with the lingual
wear surface always more developed than the
labial one. Some teeth exhibit a combination of
wear surfaces (class three), most on the labial and
lateral side of the tooth.

The curvature of the teeth changes from almost
straight to slightly curved. This variation may
also be related to the tooth’s position in the jaw.
The roots of the teeth can be distinguished from
the enamel by their smooth texture and white
color. All roots are very small, suggesting that
most of the preserved teeth were replaced during
the animal’s life.

The pencil-like shaped teeth are typical of two
sauropod groups: Titanosauridae and Diplo-
docidae (McIntosh, 1990a, 1990b). Since all
sauropod postcranial bones so far reported from
the Bauru strata (no skull material is known from
this unit) are typical of titanosaurids (e.g., Powell,
1987; Kellner & Borgomanero, 1989), I assume
that the teeth described here belong to this group.

There is, however, some morphological varia-
tion regarding tooth morphology in titanosaurids.
The teeth from the Bauru strata are very similar
to those reported by Kues et al. (1980) in Alamo-
saurus, but both differ from the ones described in
Malawisaurus which, according to Jacobs et al.
(1993), are more flattened labial-lingually.

As far as tooth morphology can suggest, there
is no evidence of more than one titanosaurid
sauropod taxa in the Bauru Group. There are two
almost complete pelves, however, of different
morphology that indicate the presence of at least
two sauropod taxa. The same applies to
postcranial elements, all collected in Minas
Gerais (Campos, 1993).

Theropod teeth are found in greater abundance
than sauropod ones. There are several teeth with
distinct morphologies among the material col-
lected near Peirópolis, which range from large
blade-like, symmetrical teeth to small, inflated
and asymmetrical teeth (Figs 8, 9). The carinae
are always serrated, although their configuration
might change according to the position of the
tooth in the jaw and its age. In one tooth, the
middle portion of the posterior carina follows a
sigmoid curve (Figs 9A-B), which might have
been pathological. Another tooth has a split
anterior carina (Figs 9C-D). Split carinae were
recently observed in tyrannosaurids (Erickson,
1995) and in one theropod tooth from the Fruit-

620

land Formation of the San Juan Basin, housed in
the New Mexico Museum of Natural History,
Albuquerque (NMMNH P-25068; pers. observ.).

The potential for identifying theropods at dif-
ferent taxonomic levels using the morphology of
the teeth was already demonstrated elsewhere
(e.g., Currie et al., 1990), and is certainly worth
investigating in the Bauru material. In a prelimi-
nary study of the theropod teeth found in the
region of Peirópolis, Kellner (1995) identified six
categories:

1) teeth very compressed laterally (blade-like);
anterior margin curved distally and posterior mar-
gin straight; anterior and posterior carinae with 3
(base) and 2.5 (tip) denticles per mm; denticles
straight, longer than wide, and chisel-like (Figs
8C-D);

2) teeth similar to those of category 1 but less
compressed laterally; anterior and posterior carinae
with 3 (base) and 2 (tip) denticles per mm; den-
ticles straight, longer than wide, and slightly hooked
on the basal part of the crown (Figs 9A-B);

3) teeth curved labial-lingually with oval cross-
section at the base; anterior carina with 2-3 den-
ticles per mm and posterior with 2.5 (base) to 1.5
(tip) denticles per mm; denticles on posterior
carina larger than on anterior; denticles straight,
tend to be wider towards the tip of the crown, and
are proportionately longer than in categories 1
and 2, but smaller than in categories 5 and 6 (not
figured);

4) very similar to those of category 3, but with
split anterior carina (only one specimen, Figs
9C-D);

5) recurved and laterally compressed teeth
smaller than in category 1; anterior carina with 3
denticles per mm and posterior with 3-4 denticles
per mm; denticles on posterior carina larger than
those on anterior; denticles pointed, inclined and
hooked; blood grooves extend from between the
bases of adjacent denticles onto the surface of the
crown, particularly on the posterior basal portion
(not figured);

6) slightly recurved teeth with labial portion
inflated and lingual portion flattened; basal cross-
section oval to sub-circular; carinae with 2 den-
ticles per mm; denticles on posterior carina larger
than on anterior; denticles inclined, hooked, and
wider than those of category 5 (Figs 8A-B).

MEMOIRS OF THE QUEENSLAND MUSEUM

Based on their morphologies Kellner (1995)
suggested that the teeth described in categories 1,
3, 5 and probably 6, represent different theropod
taxa, unless the theropods in this unit present
accentuated heterodonty. The finding of complete
theropod jaws from those layers is needed to
confirm this hypothesis.

Other dinosaur remains reported from the
Bauru strata include eggs. The first was described
by Price (1951) and regarded as belonging to a
sauropod dinosaur. The material is not very well
preserved and only the impression of the inner
portion of the egg shell can be identified in some
parts of the specimen.

Further dinosaur eggs were reported by Cam-
pos & Bertini (1985) and referred to Ceratopsia
(Ornithischia) because of their general similar
shape with those found in the Cretaceous rocks of
the Gobi Desert, Mongolia. Recently, Norell et al.
(1994) have found theropod remains associated
with eggs very similar to ones previously at-
tributed to Ceratopsia in a new locality of the
Gobi Desert. This raises the possibility that the
eggs reported by Campos & Bertini (1985) also
belong to theropods.

Until detailed descriptions and comparisons of
the fossil eggs found in the Bauru strata are avail-
able, including an analysis of the microstructure
of the egg shell — which may or may not turn out
to be of systematic value — those specimens
should only be considered Dinosauria incertae
sedis.

DISCUSSION

Mesozoic sedimentary strata, especially of
Cretaceous age, are well represented in Brazil and
provide a high potential for the preservation of
dinosaur remains. Despite this fact very few
specimens have been recovered, reflecting the
lack of collected fossils in this country. This
limited information on the Brazilian dinosaur
faunas seriously restricts our knowledge of
dinosaur diversity and distribution in this part of
South America. Nevertheless, the few studies done
so far allow some preliminary considerations.

The dinosaurian affinities of the Triassic Spon-
dylosoma are not clear (Sues, 1990). Stauriko-
saurus, the other Brazilian dinosaur of that
period, is now regarded as a primitive theropod

FIG. 7. A-C, Titanosaurid tooth from the upper portion of the Late Cretaceous Bauru group. A, Stereopair of the
labial view; note wear surface. B, Lingual view; note wear surface. C, Lateral view. D-F, Smaller titanosaurid
tooth from the upper portion of the Late Cretaceous Bauru group. D, Stereopair of the labial yiew; almost no
wear surface. E, Lingual view; note wear surface. F, Lateral view. Scale = 10 mm.

Uu
[2
=]
<
Uu
o
Z
a
z
x
d
N
<
[^4
m
z
o
v
x
[4
<
=
ea]
[74

622

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 8. A-B, Theropod tooth from the upper portion of the Late Cretaceous Bauru group. A, Stereopair of the
labial view. B, Posterior view; scale = 10mm. C-D, Large theropod tooth from the upper portion of the Late
Cretaceous Bauru group. C, Stereopair of the lingual view. D, Posterior view; scale = 10mm.

and has been referred to the Herrerasauridae
(Novas, 1992; 1994). So far no palaeo-
biogeographical pattern is apparent from these
occurrences other than a possible faunistic link
between the Southern part of Brazil and Argen-
tina during the Triassic.

The Late Cretaceous Antarctosaurus brasilien-
sis is regarded as a titanosaurid sauropod (Arid &
Vizotto, 1971), although the fragmentary nature
of the type specimen makes it difficult to establish
its phylogenetic position within Titanosauridae.
Titanosaurid sauropods in Late Cretaceous strata
of Brazil are not unexpected, since they occur
widely in other parts of South America (Huene,
1929; Bonaparte, 1986; Powell, 1987; McIntosh,
1990a). Although titanosaurids have been regarded
as ‘typical’ Gondwanan sauropods (e.g.,

Bonaparte, 1984), they are also reported from
Europe and North America (e.g., McIntosh 1990a).
This has been used to suggest that faunal interchan-
ges between northern continents and Gondwana
happened during the Late Cretaceous (e.g.,
Bonaparte, 1984; Buffetaut, 1989b). A better un-
derstanding of the ingroup relationships among
titanosaurid taxa 1s needed to test this hypothesis.

The occurrence of spinosaurids from Albian
strata of Brazil (Kellner, 1994a; Martill et al.,
1996; Kellner & Campos, 1996) is very interest-
ing from the palaeobiogeographical point of
view. Due to the fragmentary nature of most
specimens attributed to Spinosauridae, their in-
group relationships are very difficult to establish.
Nonetheless, the Brazilian and some African
spinosaurids (Spinosaurus aegyptiacus and iso-

REMARKS ON BRAZILIAN DINOSAURS 623

FIG. 9. A-B, Theropod tooth from the upper portion of the Late Cretaceous Bauru group. A, Stereopair of the
labial view. B, Posterior view; note the irregular carina; scale = 10mm. C-D, Theropod tooth from the upper
portion of the Late Cretaceous Bauru group. C, Stereopair of the lingual view. D, Anterior view; note the split
carinae; scale = 10mm.

624

lated teeth found in Tunisia and Morocco) have
developed teeth with unserrated carinae. This
synapomorphic character suggests a close
relationship between the Brazilian and those
African spinosaurids.

Baryonyx, known only from Europe, lacks this
derived dental feature. Despite the discussion
whether this European taxon should be classified
in its own higher taxonomic unit (i.e.,
Baryonychidae), there is a consensus view that
Baryonyx and spinosaurids are closely related
(Charig & Milner, 1986, 1990; Buffetaut, 1989a;
Paul, 1988). The morphological features of the
anterior portion of the rostrum and the particular
structure of the teeth shared by Baryonyx and
spinosaurids supports this hypothesis. I suggest
that Baryonyx is a primitive member of the
spinosaurid clade, lacking the dental character
which seems to unite at least some of the
Gondwanan forms.

The above phylogenetic hypothesis supports
two biogeographic points:

a) that Brazilian and African spinosaurid dis-
tribution may be the result of an Early Cretaceous
vicariant event (i.e., opening of the South Atlantic
Ocean);

b) that the spinosaurid clade is not an exclusive-
ly Gondwanan group as previously proposed
(e.g., Bonaparte, 1986; Russell, 1993), but was
originally more widespread. In the middle part of
the Cretaceous an apparently monophyletic
group of spinosaurids occupied western
Gondwana, where they diversified and eventual-
ly gave rise to the African and Brazilian taxa.

In addition, a literal interpretation of
stratigraphic data suggests that dinosaur dispersal
events in the Cretaceous were not only directed
from Gondwana to Europe as proposed by Rus-
sell (1993), but may also have occurred in the
reverse direction. All spinosaurids from Brazil
and Africa are Albian or younger, while
Baryonyx, here regarded as a primitive member
of this clade, is Barremian. The limited inform-
ation about the taxa mentioned, particularly
regarding the African and Brazilian forms, makes
the previous suggestion only a very preliminary
hypothesis. Note that it is still possible that
spinosaurids arose in Gondwana at an even ear-
lier time, and that Baryonyx may represent a
plesiomorphic lineage which migrated into
Europe prior to the break-up of western
Gondwana. More complete dinosaur material
from Brazil and from Africa are needed to test this
hypothesis.

MEMOIRS OF THE QUEENSLAND MUSEUM

ACKNOWLEDGEMENTS

I thank John G. Maisey (American Museum of
Natural History - AMNH, New York) and
Diógenes de Almeida Campos (Setor de Paleon-
tologia, Departamento Nacional da Produção
Mineral - DNPM, Rio de Janeiro) for joining me
in the 1993 field season in the Late Cretaceous
strata of the Bauru Group; Lídia Prata Ciabotti
(Fundacáo Cultural de Überaba, Uberaba) and
Luiz Carlos Borges Ribeiro (Centro de Pesquisas
Paleontológicas Llewellyn Ivor Price — CPPLIP,
Peirópolis) for providing accommodation at the
CPPLIP and for the access to specimens;
Donizetti Fontes Calçado, João Ismael da Silva,
Vanderlei Messias da Silva and Vanderlon Mes-
sias da Silva (CPPLIP) for help during the field
work and preliminary preparation of specimens;
Henry Galiano (New York) for material and in-
formation regarding the spinosaurid teeth from
Morocco; George Olshevsky for calling my at-
tention to the paper by Martill et al. (1996);
Robert L. Evander (AMNH) for help during the
preparation of the spinosaurid teeth from Moroc-
co; Diógenes de Almeida Campos (DNPM) and
Spencer Lucas (New Mexico Museum of Natural
History, Albuquerque, New Mexico) for access to
specimens; Max Hecht (City University of New
York, New York) and Fernando Novas (Museo
Argentino de Ciencias Naturales Bernardino
Rivadavia, Buenos Aires) for discussions, sug-
gestions and reviewing earlier drafts of the
manuscript. I also thank the Society of Vertebrate
Paleontology (Bryan Patterson Award), Geology
Department of Columbia University (Summer
Field Grant) and the Department of Vertebrate
Paleontology of the AMNH (Axelrod Fund) for
supporting my field work in Brazil.

LIETRATURE CITED

ARID, E.M. & VIZOTTO, L.D. 1971. Antarctosaurus
brasiliensis, um novo saurópode do Cretáceo Su-
perior do sul do Brasil. Anais 25° Congresso
Brasileiro de Geologia 2: 297-305, Sao Paulo.

BARBERENA, M.C., ARAUJO, D.C., & LAVINA,
E.L. 1985. Late Permian and Triassic tetrapods of
southern Brazil. National Geographic Research
1(1):5-20.

BARSBOLD, R., MARYANSKA, T. & OSMÓLSKA, H.
1993. Oviraptorosauria. Pp. 249-258. In
Weishampel, D.B., Dodson, P. & Osmolska, H.
(eds) ‘The Dinosauria’. (University of California
Press: Berkeley).

BENEDETTO, J.L. 1973. Herrerasauridae, nueva
familia de saurisquios Tridsicos. Ameghiniana
19: 89-102.

REMARKS ON BRAZILIAN DINOSAURS

BERTINI, R.J., MARSHALL, L.G., GAYET, M. &
BRITO, P. 1993. Vertebrate faunas from the
Adamantina and Marilia formations (Upper Baurá
Group, late Cretaceous, Brazil) in their strati-
graphic and paleobiogeographic context. Neues
Jahrbuch Geol, Paláont. Abh. 188(1): 71-101.

BEURLEN, K. 1971. As condigées ecológicas e
faciológicas da formação Santana na Chapada do
Araripe (Nordeste do Brasil). Anais Academia
Brasileira de Ciéncias 43(supl.): 411-415.

BONAPARTE, J.F. 1984. Late Cretaceous faunal inter-
changes of terrestrial vertebrates between the
Americas. Pp. 19-24. In Reif, W.E. & Westphal, F.
(eds) "Third Symposium on Mesozoic Terrestrial
Ecosystems, Short Papers’. (ATTEMPTO Verlag:
Stuttgart).

1986. Hystory of the terrestrial Cretaceous Ver-
tebrates of Gondwana, Actas IV Congreso Argen-
tino di Paleontologia y Biostratigrafia 2: 63-95.

BONAPARTE, J.F. & NOVAS, FE. 1985. Abelisaurus
comahuensis, n.g., n.sp., Carnosauria del Cretacico
tardio de Patagonia. Ameghiniana 21(2-4): 259-
265.

BOUAZIZ, S., BUFFETAUT, E., GHANMI, M,,
JAEGER, J.J., MARTIN, M., MAZIN, J.M. &
TONG, H. 1988, Nouvelles découvertes de
vertébrés fossiles dans l'Albien du Sud tunisien.
Bull. Soc. géol. France 4(2): 335-339.

BRINKMAN, D.B. & SUES, H.D. 1987. A
staurikosaurid dinosaur from the Upper Tríassic
Ischigualasto Formation of Argentina and the
relationships of the Staurikosauridae. Palaeontol-
ogy 30: 493-503.

BUFFETAUT, E. 19892. New remains of the enigmatic
dinosaur Spinosaurus from the Cretaceous of
Morocco and the affinities between Spinosaurus
and Baryonyx. Neues Jahrbuch für Geologie und
Paläontologie, Monatshefte 1989(2): 79-87.

1989b. Archosaurian reptiles with Gondwanan af-
finities in the Upper Cretaceous of Europe. Terra
Nova 1: 69-74.

CAMPOS, D.A. 1993, The vertebrates of Uberaba,
Brazil. Journal of Vertebrate Paleontology 13
(supl. to 3): 28A.

CAMPOS, D.A. & BERTINI, R. 1985. Ovos de dinos-
sauro da Formação Uberaba, Cretáceo Superior do
Estado de Minas Gerais. Bol. Resumos 9° Con-
gresso Brasileiro de Paleontologia: 19, Fortaleza.

CAMPOS, D.A. & KELLNER, A.W.A. 1991,
Dinosaurs of the Santana Formation with com-
ments on other Brazilian occurrences. Pp. 372-
375. In Maisey, J. (ed.) ‘Santana Fossils, an
illustrated atlas’. (TFH Publishing: Neptune, NJ).

1996. An unusual crested pterosaur fom the Early
Cretaceous of Brazil. Journal of Paleontology
16(Sup. 3): 25A.

CHARIG, A.J. & MILNER, A. 1986. Baryonyx, a remark-
able new theropod dinosaur. Nature 324: 359-361.

1990, The systematic position of Baryonyx walkeri,
in the light of Gauthier’s reclassification of the
Theropoda. Pp, 127-140. In Carpenter, K & Cur-

625

rie, P.J. (eds) ‘Dinosaur systematics: perspectives
and approaches’. (Cambridge University Press:
Cambridge).

COLBERT, E.H. 1970. A saurischian dinosaur from the
Triassic of Brazil. American Museum Novitates
2405: 1-39.

CURRIE, P., RIGBY, J.K., Jr, & SLOAN, R.E. 1990.
Theropod teeth from the Judith River Formation
of southern Alberta, Canada. Pp. 107-125. In Car-
penter, K. & Currie, P.J. (eds) 'Dinosaur
sytematics: perspectives and approaches’.
(Cambridge University Press: Cambridge).

ERICKSON, G.M. 1995. The evolution of split carinas
in tyrannosaur teeth from the Late Cretaceous
western interior. Journal of Vertebrate Paleontol-
ogy 15 (2): 268-274.

FREY, E. & MARTILL, D. 1994. A new pterosaur from
the Crato Formation (Lower Cretaceous, Aptian)
of Brazil. Neues Jahrbuch für Geologie und
Palaontologie, Abhandlungen, 194 (2/3): 379-412.

FREY, E. & MARTILL, D. 1995. A possible ovirap-
torosaurid theropod from the Santana Formation
(Lower Cretaceous, ?Albian) of Brazil. Neues
Jahrbuch fur Geologie und Palaontologie, Monat-
shefte, 1995 (7): 397-412.

GALTON, P.M. 1973. On the anatomy and relationships
of Efraasia diagnostica (HUENE) n. gen., a
prosauropod dinosaur (Reptilia: Saurischia) from
the Upper Triassic of Germany. Paláont. Z.
47(3/4): 229-255.

GALTON, P. M. 1977. On Staurikosaurus pricei an
early saurischian dinosaur from the Triassic of
Brazil, with notes on the Herrerasauridae and
Poposauridae. Paláontologische Zeitschrift
51(3/4): 234-245.

GAUTHIER, J. 1986. Saurischian monophyly and the
origin of birds. In Padian, K. (ed.) "The Origin of
Birds and the Evolution of Flight". Memoirs of the
California Academy of Sciences, 8: 1-55.

HOLTZ Jr, T. R. 1994, The phylogenetic position of the
Tyrannosauridae: implications for theropod sys-
tematics. Journal of Paleontology, 68 (5): 1100-
1117.

HUENE, F. von, 1929. Los Saurisquios y Ornithisquios
de Cretáceo Argentino. An. Mus. La Plata (ser. 2)
3: 1-196.

1942. Die fossilen Reptilien des südamerikanischen
Gondwanalandes. Ergebnisse der Sauriergrabun-
gen in Südbrasilien 1928/29. Beck'sche Ver-
lagsbuchhandlung, Munich. 332 pp.

IHERING, R, 1911, Fósseis de São José do Rio Preto,
Revista Mus. Paulista, 8: 141-146.

JACOBS, L.L., WINKLER, D.A., DOWNS, W.R. &
GOMANI, E.M. 1993, New material of an Early
Cretaceous titanosaurid sauropod dinosaur from
Malawi. Paleontology 36(3): 523-534.

KELLNER, A.W.A. 1994a. Comments on the
paleobiogeography of Cretaceous archosaurs
during the opening of the South Atlantic Ocean.
Acta Geologica Leopoldensia, 17; 615-625,

626

1994b. Preliminary report on the microvertebrate
fauna from the Late Cretaceous Bauru strata near
Peirópolis, Minas Gerais, Brazil. Journal of Ver-
tebrate Paleontology 14 (supl. to 3): 31A.

1995. Theropod teeth from the Late Cretaceous
Bauru Group near Peirópolis, Minas Gerais,
Brazil. Pp. 66-67. In Kellner, A.W.A. & Viana
C.F. (eds) Atas do XIV Congresso Brasileiro de
Paleontologia, Rio do Janeiro.

1996. Fossilized theropod soft tissue. Nature, 379: 32.

KELLNER, A.W.A. & BORGOMANERO, G. 1989. A
coleção paleontológica Borgomanero, Parte 1:
Vertebrados. Anais 11° Congresso Brasileiro de
Paleontologia 2: 925-939.

KELLNER, A.W.A. & CAMPOS, D.A. 1996. First
Early Cretaceous theropod dinosaur form Brazil
with comments on Spinosauridae. Neues
Jahrbuch fur Geologie und Palaontologie, Ab-
handlungen, 199 (2): 151-166.

KUES, B.S., LEHMAN, T.H. & RIGBY, J.K., Jr 1980.
The teeth of Alamosaurus sanjuanensis, a Late
Cretaceous sauropod. Journal Pal, 54: 864-869.

LEONARDI, G. 1984. Le impronte fossili di dinosauri.
Pp. 165-186. In Bonaparte, J.E, Colbert, E.H.,
Currie, P.J., Ricqles, A. de., Kielan-Jaworowska,
Z., Leonardi, G., Morello, N. & Taquet, P. (eds),
*Sulle orme di dinosauri'. (Erizzo: Venezia).

LEONARDI, G. & BORGOMANERO, G. 1981. Sobre
uma possível ocorréncia de Ornithischia na
formação Santana, chapada do Araripe (Ceará).
Revista Brasileira de Geociências, 11: 1-4.

MAISEY, J.G. (ed.) 1991. ‘Santana Fossils, An il-
lustrated Atlas’, (TFH Publishing: Neptune, NJ).
459 pp.

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. Journal
of the Geological Society of London, 153: 5-8.

McINTOSH, J.S. 1990a. Sauropoda. Pp. 345-401. In
Weishampel, D.B., Dodson, P. & Osmolska, H.
(eds) ‘The Dinosauria'. (University of California
Press: Berkeley).

1990b. Species determination in sauropod dinosaurs
with tentative suggestions for their classification.
Pp. 53-69. In Carpenter, K. & Currie, P.J. (eds)
‘Dinosaur systematics: perspectives and
approaches'. (Cambridge University Press:
Cambridge).

NORELL, M.A., CLARK, J.M., BARSBOLD, R., &
CHIAPPE, L. 1994. Oviraptorid and dromaeo-
saurids and their association with presumed
theropod nests. Journal of Vertebrate Paleontol-
ogy 14 (supl. to 3): 40A.

NOVAS, FE. 1992. Phylogenetic relationships of the
basal dinosaurs, the Herrerasauridae. Palaeontol-
ogy 35(1): 51-62.

1994, New information on the systematics and
postcranial skeleton of Herrerasaurus is-
chigualastensis (Theropoda: Herrerasauridae)
from the Ischigualasto Formation (Upper Trias-

MEMOIRS OF THE QUEENSLAND MUSEUM

sic) of Argentina. Journal of Vertebrate Paleon-
tology 13(4): 400-423.

PAUL, G.S. 1988 . ‘Predatory dinosaurs of the World’.
(Simon & Schuster: New York). 464 pp.

PONS, D., BERTHOU, P.Y. & CAMPOS, D.A., 1990,
Quelques observations sur la palynologie de T'-
Aptien Supérieur et de l'Albien du bassin d'-
Araripe (N.E. du Brésil). Pp. 241-252. In Campos,
D.A., Viana, M.S.S., Brito, P.M. & Beurlen, G.
(eds) ‘Atas 1? Simposio sobre a Bacia do Araripe
e Bacias Interiores do Nordeste'.

POWELL, J.E. 1987. Morfologia del esqueleto axial de
los dinosaurios titanosauridos (Saurischia,
Sauropoda) del Estado de Minas Gerais, Brasil.
An. 10? Congresso Brasileiro de Paleontologia 1:
155-171.

PRICE, L.I. 1951. Um ovo de dinosaurio na Formação
Baurá do Cretáceo do Estado de Minas Gerais.
Div. Geol. Min., Notas prelim. e estudos 53: 1-7.

ROMER, A.S. 1956. ‘The osteology of the reptiles’.
(University Chicago Press: Chicago) 772 pp.

RUSSEL, D.A. 1993. The role of Central Asia in
dinosaurian biogeography. Canadian Journal of
Earth Sciences 30: 2002-2012.

SERENO, P.C. 1994. The pectoral girdle and forelimb
of the basal theropod Herrerasaurus ischigualas-
tensis. Journal of Vertebrate Paleontology 13(4):
425-450.

SERENO, P.C. & NOVAS, KE. 1992. Complete skull
and skeleton of an early dinosaur. Science 258:
1137-1140.

1994. The skull and neck of the basal theropod
Herrerasaurus ischigualastensis. Journal of Ver-
tebrate Paleontology 13(4): 451-476.

SOARES, P.C., LANDIM, P.M.B., FULFARO, V.J. &
SOBREIRO NETO, A.F. 1980. Ensaio de
caracterização estratigráfica do Cretáceo no Es-
tado de São Paulo: Grupo Baurú. Revista
Brasileira de Geociências, 10(3): 177-185.

SOUZA Jr, J.J. 1984. O Grupo Bauri na porção mais
setentrional da Bacia Sedimentar do Paraná. An.
33° Congresso Brasileiro de Geologia 2: 944-957,
Rio de Janeiro.

STROMER, E., 1915, Ergebnisse der Forschungsreisen
Prof. E. Stromers in den Wüsten Agyptens, II.
Wirbeltier-Reste der Baharije-Stufe (unterstes
Cenoman) 3. Das Original des Theropoden
Spinosaurus aegyptiacus nov. gen., nov. spec.,
Abhandlungen der Bayerischen Akademie der
Wesenschaften, mathematiche - naturwissenscha
- ftenliche Abteilung 28(3): 1-32, 2 pls.

SUES, H.D. 1990. Staurikosaurus and Herrerasauridae.
Pp. 143-147. In Carpenter, K. & Currie, P.J. (eds)
‘Dinosaur systematics: perspectives and
approaches’. (Cambridge University Press:
Cambridge).

TAQUET, P. 1984. Une curieuse spécialisation du crane
de certains Dinosaures carnivores du Crétacé: le
museau long et étroit des Spinosauridés, Comptes
Rendus de l' Academie des Sciences, (Paris), ser.
II, 299(5): 217-222.

GONDWANAN DINOSAURS OF INDIA: AFFINITIES AND PALAEOBIOGEOGRAPHY

RAMINDER S. LOYAL, ASHU KHOSLA & ASHOK SAHNI

Loyal, R.S., Khosla, A. & Sahni, A. 1996 12 20: Gondwanan dinosaurs of India: affinities
and palaeobiogeography. Memoirs of the Queensland Museum 39(3) 627-638. Brisbane.
ISSN 0079-8835.

The record of Indian dinosaurs is now well known and extends from Late Triassic to the
terminal Cretaceous. The Indian dinosaurs are based on fragmentary cranial, skeletal and
egg material which is a cause of some uncertainty in the analysis of their taxonomic affinities,
age and palaeobiogeography. The Indian dinosaurian record starts with the Triassic Maleri
Formation, which yields Alwalkeria maleriensis, a coelurosaur. The overlying Dharmaram
Formation contains phytosaurs, aetosaurs, plateosaurids and a sphenosuchid. The early
Jurassic Kota Formation has yielded numerous bones (cranial and postcranial) of
Barapasaurus tagorei, besides semionotid fishes, coelacanths, pleurosaurs, crocodiles and
mammals. The Cretaceous dinosaurian record is from sedimentary sequences associated with
the Deccan volcanics and is dominated by skeletal remains of titanosaurid sauropods and
theropods, such as Indosuchus raptorius and Indosaurus matleyi. The Kallamedu Bone Bed
(Ariyalur Formation) has yielded bones and a tooth of Megalosaurus. The cranial and
postcranial skeleton of a stegosaur Dravidosaurus blanfordi is known from the Trichinopoly
Group. Khosla & Sahni (1995) have classified Indian dinosaur eggs and eggshell fragments
into eight new oospecies (in the oofamilies Megaloolithidae and Subtiliolithidae):
Megaloolithus cylindricus, M. jabalpurensis, M. mohabeyi, M. baghensis, M. dholiyaensis,
M. walpurensis, M. padiyalensis and Subtiliolithus kachchhensis. Besides these Late
Cretaceous eggshells, footprints are known from the Bhuj Formation of Early Cretaceous
age (Ghevariya & Srikarni, 1990). Coprolites of dinosaurs and (probable) chelonians are
known from the Late Cretaceous of Pisdura, and include four categories of ribbed and
non-ribbed forms. The palaeobiogeographic analysis of the Indian dinosaurs implies they
are part of a cosmopolitan distribution. However, migrations through an earlier Madagascan
connection and through island arcs in the Cretaceous (represented by Dras volcanics) and
earlier collision arcs and microplates in Late Permian to Cretaceous (represented by north-
ernmost Gondwana fragments of Tibet, Iran, Iraq and Afghanistan) led to an influx of some
taxa. These paths of migration are significant in explaining the cosmopolitan character of
the Indian biotas and their similarities to Laurasian faunas during the drift of the Indian Plate,
[O India, dinosaurs, Mesozoic, tracks, dinosaur nests, eggs.

R.S. Loyal, A. Khosla & A. Sahni, Centre of Advanced Study in Geology, Panjab University,
Chandigarh, 160 014, India; 1 July 1996.

Dinosaur remains in India are found in the
Triassic-Cretaceous sequence of peninsular India
(Table 1), however Middle Jurassic-Middle
Cretaceous gaps are quite significant. In this con-
text, the Pranhita-Godavari valley, where a thick
continental vertebrate-rich sequence of Late Per-
mian-Cretaceous age is exposed, has assumed
considerable significance as a site for
Gondwanan dinosaurs.

While the Pranhita-Godavari Valley (Fig. 1)
represents a repository of dinosaurs of Triassic-
Jurassic age, the younger terminal Cretaceous
dinosaur fauna is most diverse. It occurs along the
Narmada Valley of central India, associated with
the Lameta Formation and is spread over more
than 10,000 sq. km., below the Deccan vol-
canosedimentary sequence (Sahni et al., 1994,
Fig. 2). In addition, Cretaceous dinosaurs also

occur towards the south, associated with the
Ariyalur Formation.

TRIASSIC

The oldest record of Indian dinosaurs comes
from the Late Triassic Maleri Formation, which
is exposed near the village Maleri (near Tandur)
in the district Adilabad, Andhra Pradesh.
Lithologically, red clays are abundant, along with
lime-pellet rocks (calcirudites, calcarenites) and
quartzose sandstones (Kutty & Sengupta, 1989).

The biotic elements of Maleri Formation in-
clude vertebrates, coprolites, invertebrates and
plants. Dominant in the Lower Maleri, is Al-
walkeria maleriensis, which constitutes the ear-
liest Indian dinosaur (Chatterjee, 1987) and is
based on imperfectly known fragmentary
material, a partial skull, vertebrae, a femur and

628

astragali. It is referred to the basal
theropods, belonging to a group of
slightly built predaceous
dinosaurs.

The overlying Dharmaram For-
mation is composed of a basal
thick sandstone followed by more
sandstones and clay beds. The
sandstones resemble the Maleri
sandstones, but are more coarse-
grained and gritty (Kutty et al.,
1987). The lower part of the Dhar-
maram Formation — besides the
occurrence of the fish Xenacanthus
and Ceratodus — is characterised
by the presence of a phytosaur
(Nicrosaurus), dominant aetosaurs
and a small prosauropod (Kutty &
Sengupta, 1989). The Upper Dhar-
maram Formation, on the other
hand, is far richer and contains a
large plateosaurid, an ornithischian
and a sphenosuchid (Kutty, 1969;
Kutty et al., 1987). Although as the
fauna remains undescribed as yet,
a Late Norian-Rhaetian age is
preferred by most workers for this
fauna.

MEMOIRS OF THE QUEENSLAND MUSEUM

Chitur
e

o
18
Ankis apr Eat

FIG. 1. Map of Pranhita-Godavari Valley with fossil- bearing localities

JURASSIC

The continental Kota Formation, of Early
Jurassic age, lies conformably above the Dhar-
maram Formation and in turn is overlain by Gan-
gapur Formation of Early Jurassic age (Kutty et
al, 1987). The stratotype of this formation is
named after Kota village, on the eastern bank of
the Pranhita River. This formation has provided
a vast amount of information about its exceed-
ingly rich continental fauna, comprising fishes,
reptiles, mammals, freshwater ostracodes, con-
chostrachans and land insects (Kutty et al., 1987).

The lithic units of the Kota Formation are char-
acterised by sandstones (ferruginous arkoses),
with cross-bedding, and limestones with desicca-
tion polygons and worm-bored layers (Robinson,
1970; Jain, 1980; Yadagiri & Rao, 1987).

Field work conducted by palaeontologists of
the Indian Statistical Institute in the Kota Forma-
tion during 1960-61 led to the recovery of about
300 sauropod bones from a bone bed occurring
between a sandstone-clay lens. This vertebrate
material was assiduously studied by Dr S.L. Jain
and his colleagues, who established a new

(after Jain et al., 1975).

sauropod, Barapasaurus tagorei, one of the best
known Early Jurassic dinosaurs in the world
(Jain, 1975, 1979). A complete mounted skeleton
of this dinosaur is displayed at the Geology
Museum, Indian Statistical Institute, Calcutta.

Barapasaurus tagorei was a large sauropod
with slender limbs, spoon-shaped teeth with
coarse denticles, opisthocoelous cervical and anterior
dorsal centra, with the other centra platycoelous.
Neural spines are not bifurcate and the centra are
not cavernous, but have oval depressions in the
lateral surface. The ilium possesses a well-
developed anterior process, the ischium is rela-
tively slender and rod-like distally and the pubis
has a well-developed terminal expansion.

Jain et al. (1979) commented on the level of
development of Barapasaurus, stating that it is
intermediate between the prosauropod stage and
that of known sauropods (e.g., the limbs, though
graviportal, are slender). The anterior caudals
have not developed procoelous centra. The
sacrum is narrow, with the pelvic depression
small in comparison to length of the pubis. Mc-
Intosh (1990) tentatively assigned Barapasaurus

GONDWANAN DINOSAURS OF INDIA

629

TABLE 1. Gondwana stratigraphy of Pranhita-Godavari Valley (modified after Kutty et al., 1987).

Deccan volcano-sedimentary sequences

Late Cretaceous frogs, fishes, charop

mammals, dinosaurs (titanosaurids),
hytes, ostracodes

sandstones, shales, marls, associated
with Deccan basalts

Gangapur
Early Cretaceous

Kota

Early Jurassic early mammals

Dharmaram .
late Late Triassic

prosauropods

Gleichenia, Pagiophyllum Ptilophyllum

holostean fishes, sauropods, pterosaurs,

ferruginous sandstones conglomerates

ritty sandstones, pink mudstones,
erruginous sandstones

sandstones, siltstones, clays and a
limestone band

sandstones, red clays

UT » ON
red clays, sandstones, lime-pellet rocks

late Middle Triassic

Late Permian-Early Triassic

to the Vulcanodontidae (Cooper, 1984) on the
basis of the narrow sacrum, deeply furrowed
caudals, long forelimbs, apron-like pubis, elon-
gate metatarsals and teeth with coarse denticles.
The associated fauna of the Kota dinosaurs
includes semionotid fishes (Tetragonolepis,
Lepidotes, Parapedium), a coelacanth (Indo-
coelacanthus robustus), a pholidophorid fish
(Pholidophorus), a pterosaur (Campylognathoides)
and teleosaurid crocodiles (Kutty et al., 1987;
Jain & Roy Chowdhury, 1987). Early Jurassic
mammals from the Kota Formation are repre-
sented by Kotatherium, Trishulotherium and In-
dotherium (Datta, 1981; Yadagiri, 1984; 1985).

CRETACEOUS

SKELETAL MATERIAL. The Indian Cretaceous
dinosaur record is associated with the non-marine
post-Gondwana sequence of central and western
India. This sequence is represented by Lameta
Formation, which has received considerable at-
tention for the last 70 years for its fairly extensive
dinosaurian fauna, Stratigraphically, the
dinosaur-rich vertebrate beds underlie the Dec-
can Traps, with an aerial extent of more than
10,000 sq. kms. As such, this vertebrate sequence
is crucial to the dating and duration of the traps,
more so as the Late Cretaceous extinction
mechanisms might be related to the eruptions
(Courtillot et al., 1986; Buffetaut, 1987).

The stratotype of the Lameta Formation is
situated at Lametaghat, near Jabalpur on the
banks of Narmada river. The classic dinosaur
areas here are Bara Simla Hill and Chui Hill, in

Maleri a", metoposaurs, aetosaurs, phytosaurs,
early Late Triassic thynchosaurs
Bhimaram Sandstone

labyrinthodont, dicynodont

Yerrapalli stahleckeriid, kannemeyeriid
early Middle Triassic dicynodonts
Kamthi

dicynodonts from basal beds

sandstones (ferruginous in lower part),
red clays

red violet clays, sandstones, pellet rocks

ferruginous, non-felspathic sandstones,
urple siltstones

addition to Pisdura in Maharashtra. The first
detailed account of the geology of the Jabapur
area was given by Matley (1921), who prepared
an excellent map of the area ona scale measuring
8 inches to a mile, along with a description of
various lithic units.

Huene & Matley (1933) described postcranial
elements of the titanosaurid sauropods, viz.
Titanosaurus indicus and Antarctosaurus sep-
tentrionalis from Bara Simla Hill, and a third
sauropod Laplatasaurus madgascariensis from
Pisdura. From the same locality carnosaurs were
also described: Indosuchus raptorius and In-
dosaurus matleyi, along with coelurosaurs, such
as Compsosuchus solus, Laevisuchus indicus,
Jubbulpuria tenuis, Coeluroides largus, Dryp-
tosauroides grandis, Ornithomimoides mobilis
and O.(?) barasimlensis.

Of the aforementioned forms Antarctosaurus
septentrionalis has been given a new generic
name Jainosaurus by Hunt et al. (1994) because
of its distinctness from the Argentinian form, the
genotypic A. wichmannianus. The main differen-
ces are in braincase of the Indian and Argentinian
forms. Earlier Indian braincase studies on
‘Antarctosaurus’ septentrionalis were by Berman
& Jain (1982) however the specimens were
recovered from Dongargaon near Pisdura. Fur-
ther, Indosaurus and Indosuchus regarded by
Chatterjee (1978) as a megalosaur and tyran-
nosaurid (respectively), are now considered as
abelisaurids (see Chatterjee, this volume).

As compared to the saurischian fauna, the in-

formation and taxonomic assignment of ornithis-
chians have remained largely in doubt.

630

Rahioli Batasinor
. we

(pode

Takli
Nande

MEMOIRS OF THE QUEENSLAND MUSEUM

o

80

Jabalpur

R; Chui Hill section

Bara Simla section
Chota Simla section
Lametaghat section

* Pisdura

Pavna *

*Asifabad

FIG. 2. Localities of dinosaur eggshells and nests in India.

Lametasaurus indicus described from Bara Simla
Hill by Matley (1933) was at that time considered
to be a stegosaurian. Later Chakarvarti (1934)
and Walker (1964) opined that this genus could
not be an ankylosaurian. Recently, works by Ber-
man & Jain (1982), Buffetaut (1987) and Molnar
& Frey (1987) opined that Lametasaurus could
well have been a titanosaurid. Molnar (pers.
comm., 1996) believes it to be a chimaera (based
on bones and skeletons of saurischians and
crocodilians). However a new ankylosaurian
specimen is reported by Chattergee & Rudra (this
volume).

From South India, the dinosaur studies were
first conducted by Blanford (1862), who reported
bones and a tooth of Megalosaurus from the same
horizon in the Kallamedu Bone Bed (Ariyalur
Formation, Late Cretaceous). Matley (1929)
reported fragmentary bones of Titanosaurus
along with teeth of a megalosaur and bones of a
stegosaurian. Much later, Yadagiri & Ayyasami
(1979) described cranial and postcranial skeleton
of stegosaur Dravidosaurus blanfordi from the
Trichinopoly Group (Turonian-Santonian). In
this work, a brief mention is also made of the
presence of bones of a sauropod, theropod and
stegosaur from the Kallamedu Bone Bed.

EGGSHELLS. The Late Cretaceous eggshell
material and nesting sites have been most abun-
dant and have assumed considerable significance

over the last decade, providing evidence of a large
dinosaur hatchery (Figs 2, 3). The eggshells and
nesting sites underlie the Deccan Trap flows and
are uniformly found in a hard sandy carbonate
with strong development of a calcretised
palaeosol towards the top (Tandon et al., 1990;
Jolly et al., 1990; Sahni et al., 1994, Fig. 3). In
addition, Late Cretaceous eggshells are also
known from the sandy limestones of the Kal-
lankurichi Formation of the Ariyalur sequence
(Kohring et al., 1996).

Up to now, considerable work has been done to
categorise main eggshell morphotypes (discussed
later) based on megascopic and microscopic
characters (Mohabey, 1982; 1984; Srivastava et
al., 1986; Vianey-Liaud et al., 1987; Jain 1989;
Sahni, 1989; Mohabey & Mathur, 1989; Jolly et
al., 1990; Sahni, 1993; Sahni et al., 1994 and
many others). This research and the present work,
have established five eggshell morphotypes,
three of which have been assigned to titanosaurid
sauropods (parataxonomic oofamily Megalo-
olithidae, Zhao, 1979) and the fourth to the or-
nithoid type. In a recent work, Khosla & Sahni
(1995) established eight oospecies including
seven of the oogenus Megaloolithus referable to
sauropods and one, Subtiliolithus, an ornithoid.
These oospecies are:

Megaloolithus cylindricus Khosla & Sahni
(1995)(PI. 1, Fig. A): spherical eggs 12-20cm in
diameter; eggshell thickess varying from 1.7-

GONDWANAN DINOSAURS OF INDIA

INDEX

Em Sandy Carbonate
Sandy mori
E

4] Sandy clay

b^

Balasinor

631

A
2m

o p
< p
< dk
é o

Eggshell .
Sandstone Fragments An Jar
Conglomerate Nests
hert i
[e er Dinosaur
Basait banes

rare
aT.

Precambrian basement

TA

FIG. 3. Stratigraphic sections with associated eggshells and nests.

3.5mm; spheroliths cylindrical; pores subcir-
cular; pore canals long and narrow; tightly pack-
ed subcircular basal caps, 0.2-0.5mm in diameter.

Megaloolithus jabalpurensis Khosla & Sahni
(1995)(PI1. 1, Figs B, C): spherical eggs 14-16cm
in diameter; eggshell thickness 1-2.3mm; nodose
external surface; spheroliths with varying width
and shapes; elongate to circular pores; tightly
packed basal caps (more so than in M.
cylindricus) 0.1-0.5mm in diameter.

Megaloolithus mohabeyi Khosla & Sahni
(1995): spherical eggs 16-19cm in diameter; egg-
shell thickness 1.8-1.9mm; nodose external sur-
face; spheroliths long and fused to adjacent ones;
growth lines convex to undulating; pores ellipti-
cal; broad and semicircular basal caps, 0.14-
0.21mm in diameter.

Megaloolithus baghensis Khosla & Sahni
(1995)(PI. 1, Fig. D): spherical eggs 14-20cm in
diameter; eggshell thickness 1.0-1.7mm; nodes
discrete to fused; spheroliths fan-shaped, distinct
or partially fused and look more compressed
when ending in multiple nodes; pores subcircular
to elliptical; basal caps swollen-ended, 0.2-
0.3mm in diameter.

Megaloolithus dholiyaensis Khosla & Sahni
(1995): eggshell thickness 1.47-1.75mm; nodose
outer surface; spheroliths cylindrical and fan-
shaped; pore canals straight; basal caps subcir-

cular to conical, isolated to fused, 0.15-0.30mm
in diameter. So far found only as fragments.

Megaloolithus walpurensis Khosla & Sahni
(1995): eggshell 3.5-3.6mm thick; spheroliths ir-
regularly fan-shaped and showing fusion towards
the outer surface; pore canals short, thin, slender,
with slit-like openings; basal caps conical to sub-
circular, 0.25-0.30mm in diameter. So far found
only as fragments.

Megaloolithus padiyalensis Khosla & Sahni
(1995): eggshells 1.12-1.68mm thick; nodose ex-
ternal surface; small irregular, slender
spheroliths; small and large pore canals; basal
caps circular to semicircular, 0.07-0.21mm in
diameter. So far found only as fragments.

Subtiliolithus kachchhensis Khosla & Sahni
(1995)(Pl. 1, Fig. E): eggshells extremely thin,
0.35-0.45mm; external surface exhibits irregular-
ly spaced microtubercles; double layered; spongy
layer poorly defined; mammillary layer 1/2-1/3
of the total eggshell thickness; mammillae tightly
packed, 0.03-0.05mm in diameter. So far found
only as fragments.

The eggshells referred to a new morphotype
(Elongatoolithidae) were recovered from near
Rahioli (district Kheda, Gujarat) and work on
them is in progress. This morphotype is currently
being described in detail with Dr D.M. Mohabey,
GSI, who has also recovered the same type of

632 MEMOIRS OF THE QUEENSLAND MUSEUM

PLATE 1. A, Radial view of Megaloolithus cylindricus (VPL/AK 220), Pat Baba Mandir, Jabalpur (Madhya
Pradesh). Note cylindrically shaped spheroliths. B, Radial view of Megaloolithus jabapurensis (VPL/AK 275),
Bara Simla Hill, Jabalpur (Madhya Pradesh), showing small and large fan-shaped spheroliths. C, Radial view
of Megaloolithus jabalpurensis (VPL/AK 276), Bara Simla Hill, Jabalpur (Madhya Pradesh). Note sweeping
extinction pattern of spheroliths. D, Radial view of Megaloolithus baghensis (VPL/AK 556), Pisdura,
Chandrapur district, Maharashtra. Note small and large fused spheroliths with moderately arched growth lines.
E, Radial view of Subtiliolithus kachchhensis (VPL/AK 580), Anjar, District Kachchh, Gujarat. Note two-
layered eggshell. Mammillary layer is well defined while spongy layer is faintly developed. ML=mammillary
layer, Pc=pore canal; scale = 500m. (VPL/AK=Vertebrate Palaeontology Lab./Ashu Khosla).

GONDWANAN DINOSAURS OF INDIA

eggs. These eggshells are ellipsoid in shape, with
a thickness of 1.2-1.6mm, and ornamentation
varying from sagenotuberculate (at poles) to
lineartuberculate in the equatorial region. The
microstructure shows a double layer with radiat-
ing structures in the mammillary zone —the latter
being about half of the total thickness. Gently
curving growth striations are observed in the
outer spongy layer. The spheroliths are slender
and fused towards the upper part.

TRACKS. As well as the skeletal and eggshell
material, dinosaur pedal tracks are also known
from the Early Cretaceous Bhuj Formation ex-
posed in the Pakhera and Fatehgarh areas of
Kachchh (Ghevariya & Srikarni, 1990). From the
Pakhera area, tracks belonging to a juvenile and
an adult individual were reported. These types
were interpreted as ornithopod, as claws were
absent. From the Fatehgarh area, diversely
shaped (rectangular, oval and diamond shaped)
tracks were described from hard sandstone; the
associated grey shale bears an Upper Gondwanan
plant, Ptilophyllum.

PALAEOENVIRONMENTS

The Triassic Maleri Formation, has long been
thought to have been deposited in a river valley
system, the sandstones representing the channel
deposits and clays suggesting an interchannel
floodplain (Jain, 1990). The cross-bedded and
lenticular sandstones represent cut-off meanders
of the mainstream, with clays deposited on water-
logged floodplains (Sengupta, 1970). Jain (1990)
described the general environment of the Maleris
as indicative of well-aerated country, inhabited
by aquatic vertebrates, in a forest habitat. The
waterholes and their environs were dominated by
amphibians and arboreal vertebrates. The
eosuchian Malerisaurus probably lived on
floodplains or marshy land close to lakes.

The environment for the overlying Dharmaram
Formation, from its base to the top of the lower
Kota Formation is similar to that of an upwardly-
fining sequence deposited under a laterally shift-
ing meandering stream (Rudra, 1982).

As mentioned earlier, the overlying Kota For-
mation is a typical continental deposit. There
were two kinds of depositional environment:
sandstones and clays representing fluviatile con-
ditions (river deltas), and thin limestones suggest-
ing deposition in inland lakes (Robinson, 1970).
Further, two different climatic conditions also
prevailed; one with moderately high relief and
rainfall and the other low relief and rainfall. Tasch

633

et al. (1975), on the basis of conchostracans, and
Govindan (1975), on the presence of ostracodes,
supported a fresh water environment for the
deposition of the limestone within small basinal
ponds. According to Rudra (1982) the Kota Lime-
stone suggests periodic precipitation of impure
lime muds in freshwater playa-type lakes. The
presence of inclined and vertical burrows indi-
cates quiet sedimentary episodes, and pelletoid-
bearing laminations indicate a very slow rate of
sedimentation (Maulik & Rudra, 1986). Periods
of drought are also suggested by the presence of
semionotid fish in cracked limestone — this cir-
cumstance has been explained by Jain (1980;
1983) as owing to the mortality of semionotids in
the oxygen-deficient evaporating water of the
Kota lakes. The large sauropods, such as
Barapasaurus, inhabited land close to the Kota
lake and fed upon the aquatic vegetation, while
the crocodiles fed on fishes and lived on the banks
(Jain & Roychowdhury, 1987). Broadly, the Kota
lake environment is pictured as a shallow water
body, with pholidophorids and coelacanths in-
habiting its deeper parts.

The Late Cretaceous Lameta Formation has
provided a fairly extensive record of terrestrial
dinosaurs, ostracodes, charophytes and inver-
tebrates; the environmental interpretation has
therefore been of much fascination.

The marine interpretation of the Lameta For-
mation is based on the presence of algal-like
structures, glauconite beds and thalassinoid bur-
rows (Chanda, 1963; Chanda & Bhattacharya,
1966). Singh (1981) opined that both the Bagh
and Lameta Formations are the product of a
Turonian-Senonian marine transgression;
wherein an estuarine tidal flat complex was indi-
cated by crab burrows, bioturbated carbonate and
marl facies.

The marine interpretation was first challenged
by Brookfield & Sahni (1987) who believed that
the evidence favouring a marine environment
may have been misinterpreted, for instance the
algal structures in the limestones are pedogenic
carbonate structures of palaeosols. Further, on the
basis of the dinosaur eggs, bones and associated
freshwater biota (charophytes, ostracodes and
other invertebrates), it was concluded that the
entire Lameta sequence consists of pedogenical-
ly-modified piedmont calcrete deposited in a
semi-arid environment dominated by a large
river.

Subsequently, on the basis of Matley's (1921)
map of the Jabalpur area, Tandon et al. (1990)
studied the sedimentology and facies of sections

634

SUBDUCTION ZONE

1 KARAKORAM
2 SHYOK BACK ARC BASIN
3 KOHISTAN DRAS

FIG. 4. Late Cretaceous palaeoposition of the Indian
plate and Kohistan-Dras Island Arc (after Sharma,
1987; Sahni & Bajpai, 1991).

at Lametaghat, Chui Hill and Bara Simla Hill in
detail. At Bara Simla Hill they identified several
undulations associated with 'relief highs' and
‘relief lows’ — these were interpreted (through
computer studies) to represent shoulders and
ramps on an ancient geomorphic surface. The
relief high characterises dinosaur nest-bearing
nodular carbonates, while the relief lows are as-
sociated with a marly dinosaur bone-bearing
horizon. Both these palaeo-relief surfaces form
part of fanpalustrine flat system with repeated
sheetwash activity, wherein the upper sandy car-
bonate represents emergent areas with high
pedogenic modification, while the lower marly
facies has low pedogenic modification (Tandon
et al., 1990; Sahni et al., 1994).

Further, the nesting sites, in as much as they are
associated with pedogenic calcrete, fringed small
lakes. Desiccation and subsequent sheetwash ac-
tivity occurred leading to rapid matrix cementa-
tion and preservation of large dinosaur nests
(Jolly et al., 1990). In this context extensive field
work at Bara Simla Hill, Chui Hill and
Lametaghat suggest a single stratigraphic level of
sauropod nests associated with pedogenic

MEMOIRS OF THE QUEENSLAND MUSEUM

calcretes in the Lower Limestone layer (Sahni &
Khosla, 1994). Evidence of multiple layers of
nesting is absent; the different topographic levels
can be explained by the undulatory character of
the palaeogeomorphic surface (Tandon et al.,
1990). As such there is a strong evidence for ‘site
selectivity’ on the part of the dinosaurs, rather
than site fidelity. Moreover, the pedomorphic sur-
face along the entire stretch of terrain from Jabal-
pur in the east towards Gujarat in west was
selected by the sauropods for nesting sites (Jolly
et al., 1990).

An alluvial-limnic environment under semi-
arid conditions was suggested for various suben-
vironments of the Lametas exposed in
Dongargaon Hill and Dhamni-Pavna sections by
Mohabey et al. (1993). The eggshell-bearing
horizons, especially near Pavna, are associated
with overbank facies. Pink and red clays suggest
the leaching action of reducing waters with oc-
casional ponding of water, as shown by fresh-
water charophytes and ostracodes. However, at
Pisdura and Dongargaon, a channel facies with
immature pebbly and gritty sandstones with
calcrete layers was delineated, with cross-
stratification, convoluted bedding and channel
fill structures. The lacustrine facies is charac-
terised by finely laminated silty clays and con-
tains Lepisosteus, Lepidotes and ostracodes
(Candionella and Mongolianella).

The palaeophysiologic conditions of the Indian
Cretaceous eggshells have been characterised
mainly on shell thickness, pore distribution, den-
sity, pathologic thickening or thinning and
resorption of mammillae (Sahni et al., 1994). The
porosity values, which reflect nest humidity, for
the eggshells are substantially less; these values
are measured by water vapour conductance
which vary from 2.65mg/(day.tor) to
3.49mg/(day.tor); such low values are probably
due to the semi-arid environment.

The OP and C” values of the Lameta eggshells
indicate that dinosaurs drank water from rivers
and evaporating pools and consumed C-3 type
plants, such as palms, dicot shrubs and conifers
(Sarker et al., 1991).

PALAEOBIOGEOGRAPHY

Until recently, the main arguments supporting
the hypothesis of Gondwanaland were derived
from glaciation and distinctive floras. In recent
years geophysics and oceanography have con-
tributed suggestive new data, but it is the infor-
mation from Gondwanan vertebrates that has

GONDWANAN DINOSAURS OF INDIA

proved to be of inestimable value in providing a
clearer palaeobiogeographic analysis in the
scenario of drifting plates.

A look into the affinities of the Indian
Gondwana tetrapods clearly provides sufficient
evidence of Laurasian influence in these
palaeocommunities, with no endemism. The Tri-
assic Maleri dinosaur Alwalkeria is more or less
similar to Coelophysis of North America and
Procompsognathus of Germany. The lower Juras-
sic Kota dinosaur, Barapasaurus tagorei bears
affinities to the European Ohmdenosaurus
liasicus, Patagonian Patagosaurus fariasi and
Chinese Shunosaurus lii.

Significant evidence, however, is the discovery
of Maastrichtian mammal Deccanolestes hislopi
in the Intertrappeans of Naskal, Andhra Pradesh
(Prasad & Sahni, 1988). This genus shows mor-
phological similarity to the North American
Cimolestes and Mongolian Kennalestes.

Palaeorycted mammals are also known from
the Paleocene of Morocco (Cappetta et al., 1978).
In addition to palaeoryctids, molar teeth belong-
ing to Docodontidae are known from the Late
Jurassic of India (Prasad & Khajuria, 1990). Be-
sides, pelobatid frogs, lizards and turtles, known
from the Intertrappean beds of Nagpur, are also
known from North American and Russian
localities.

Amongst the invertebrates, the Intertrappean of
Nagpur has yielded Cypris, Candona, Al-
tanicypris, which are similar to Mongolian-
Chinese form. Charophytes, such as Platychara
perlata known from the Intertrappeans, are close-
ly similar to P. compressa of North America. The
aforementioned picture of lack of endemism and
the cosmopolitan distribution of Indian
dinosaurs, has been utilised and interpreted by
some stabilists, who have invoked static models
of Indian-Eurasian plates and their contiguity in
the past (Chatterjee, 1984; 1987; Chatterjee &
Hotton, 1986). These workers are of the view that
India was never an island continent during its
convergence towards Eurasia, and believed in a
narrower Tethys with India occupying a position
between Somalia and Eurasia in its pre-drift posi-
tion (Chatterjee & Hotton, 1986). Also, the
presence of Laurasian biotas in India and migra-
tions, may be a consequence of land corridors
between India and Eurasia that were maintained
during Cretaceous (Chatterjee, 1992).

However the work of ardent mobilists, notably
Sahni (1984), Sahni & Bajpai (1991) and Loyal
(1984, 1985), have seriously questioned the
validity of Chatterjee’s (1984, 1987) conclusions.

635

Sahni (1984), while analyzing and admitting the
Eurasiatic faunal presence in the Indian
Gondwanas, concluded that migration of land
faunas took place during the Cretaceous-
Palaeocene from the eastern part of Africa, as
India was then close to Madagascar. Later, Sahni
et al. (1987) and Sahni & Bajpai (1991) suggested
that the necessary land passage might have been
provided by the island arcs of Dras Volcanics
(Fig. 4); this activity might have been initiated
earlier, in Late Jurassic (Sharma, 1987). In this
context, an early Cretaceous India-Eurasia col-
lision was suggested by Jaeger et al. (1989).

Further, to account for the cosmopolitan dis-
tribution, one of us (RSL) is of the opinion that
two dispersal corridor routes existed for
Gondwanaland-Laurasia intermigrations, viz.
India-South Tibet-Afghanistan-Iran-Turkey and
India-Africa-Siberia (Loyal, 1985). The frag-
mented portion of northernmost Gondwanaland
(Tibet, Iran, Iraq, Afghanistan) formed a migrat-
ing festoon of island arcs, establishing links with
Eurasia throughout Late Permian-Cretaceous.
Therefore, these paths of migration (in Laurasia
and Gondwanaland) were extremely significant
in as much as they help explain the lack of faunal
endemism of the Indian biotas during its time of
geographic isolation and rapid northward jour-
ney.

Finally, though there exists now (as aforemen-
tioned) a growing body of vertebrate and inver-
tebrate faunal data from the Indian Plate, pointing
to strong India-Laurasia links similar palaeo-
biogeographic data from other Gondwanic areas
(Africa, S. America, Antarctica, Australia) is al-
most meagre. Therefore, in this context,
Laurasia-Gondwanaland faunal comparisons
remain poorly understood: hence urgent and ex-
tensive fieldwork in the Mesozoic terrain of Gon-
dwanic regions is necessary.

ACKNOWLEDGEMENTS

The authors are grateful to Dr R.E. Molnar for
review and valuable comments on various
aspects of Gondwanan dinosaurs. One of us (AK)
is thankful to the Council of Scientific and In-
dustrial Research (CSIR, New Delhi) for finan-
cial assistance in the form of a Senior Research
Fellowship.

LITERATURE CITED

BERMAN, D.S. & JAIN, S.L., 1982. The braincase of
a small sauropod dinosaur (Reptilia: Saurischia)
from upper Cretaceous Lameta Group, Central

636

India, with review of Lameta Group localities,
Annals of Carnegie Museum 51: 405-422.

BROOKFIELD, M.E. & SAHNI, A., 1987. Palaeoen-
vironments of the Lameta Beds (Late Cretaceous)
at Jabalpur, Madhya Pradesh, India: Soils and
biotas of a semi-arid alluvial plain. Cretaceous
Research 8: 1-14.

BUFFETAUT, E., 1987. On the age of dinosaur fauna
from the Lameta Formation (Upper Cretaceous of
Central India). Newsletters in Stratigraphy 18:
1-6.

CAPPETTA, H., JAEGER, J.J., SEBATIER, M.,
SUDRE, J. & VIANEY-LIAUD, M. 1978.
Decouverté dans le Paleocene du Maroc des plus
anciens mammiferes eutheriens d'Afrique.
Geobios 11: 257-263.

CHAKRAVARTI, D.J. 1934. On a stegosaurian
humerus from the Lameta Beds of Jubbulpore.
Quarterly Journal of the Geological, Mining and
Metallurgical Society of India 6(3): 75-79.

CHANDA, S.K., 1967. Petrogenesis of the calcareous
constituents ofthe Lameta Group around Jabalpur,
M.P., India. Journal of Sedimentary Petrology 37:
425-437.

CHANDA, S.K. & BHATTACHARYA, A, 1966. A
re-evaluation of the stratigraphy of the Lameta —
Jabalpur contact around Jabalpur, M.P. Journal of
the Geological Society of India 7: 91-99.

CHATTERJEE, S. 1978. A primitive parasuchid
(phytosaur) reptile from the Upper Triassic Maleri
Formation of India. Palaeontology 21: 83-127.

1984. The drift of India: A conflict in plate tectonics.
Mémoires de la Société géologique de France
147: 43-48.

1987. The new theropod dinosaur from India with
remarks on the Gondwana-Lauriasia connection
in the Late Triassic. Pp. 183-189. In McKenzie,
G.D. (ed.) 'Gondwana six: Stratigraphy,
Sedimentology and Paleontology'. Geophysical
Monographs 41.

CHATTERJEE, S. & HOTTON, N. 1986. The paleo-
position of India. Journal of Southeast Asian Earth
Science 1: 145-189.

CHATTERJEE, S. & RUDRA, D.K. 1996. KT events
in India: impact, rifting, volcanism and dinosaur
exticntion. Memoirs of the Queensland Museum,
this volume 489-532.

COOMBS, W.P. & MARYANSKA, T. 1992,
Ankylosauria. Pp. 456-483. In Weishampel, D.B.,
Dodson, P. & Osmólska, H. (eds) 'The
Dinosauria'. (University of California Press:
Berkeley).

COURTILLOT, V, BESSE, J. VANDAMME, D.,
MONTIGN Y, R., JAEGER, J.J. & CAPPETTA,
H. 1986. Deccan flood basalts at the
Cretaceous/Tertiary Boundary. Earth and
Planetary Science Letters 80: 361-374.

DATTA, P.M. 1981. The first Jurassic mammal from
India. Zoological Journal of Linnean Society,
London 73: 307-312.

MEMOIRS OF THE QUEENSLAND MUSEUM

GALTON, P.M. 1981. Craterosaurus pottonensis
Seeley, a stegosaurian dinosaur from the Lower
Cretaceous of England anda review of Cretaceous
stegosaurs, Neues Jahrbuch fiir Geologie und
Paläontologie, Abhandlungen 161: 28-46

GHEVARIYA, Z.G. & SRIKARNI, C. 1990. Anjar
Formation, its fossils and their bearing on the
extinction of dinosaurs. Pp. 106-109. In Sahni, A.
& Jolly, A. (eds) ‘Cretaceous Event Stratigraphy
and the Correlation of Indian Nonmarine Strata’.
(Panjab University: Chandigarh).

HUENE, F. von & MATLEY, C.A. 1933. The
Cretaceous Saurischia and Ornithischia of the
Central Provinces of India. Memoir of Geological
Survey of India, Palaeontologia Indica ns 21: 1-
72.

JAEGER, J.J., COURTILLOT, V. & TAPPONIER, P.
1989. Paleontological view of the ages of the
Deccan Traps, the Cretaceous Tertiary Boundary
and the India-Asia Collisions. Geology 17: 316-
319.

JAIN, S.L., 1980. The continental Lower Jurassic Fauna
from the Kota Formation, India. Pp. 99-123, In
Jacobs, L.L. (ed.) ‘Aspects of Vertebrate History’.
(Museum of Arizona Press: Flagstaff).

1983. Spirally coiled coprolites from the Upper
Triassic Maleri Formation, India. Palaeontology
26: 813-829,

1990. An Upper Triassic vertebrate assemblage
from Central India. Bulletin of the Indian
Geologists Association, 67-84.

JAIN, S.L., KUTTY, T.S., ROY CHOWDHURY, T. &
CHATTERJEE, S., 1975. The sauropod dinosaur
from the Lower Jurassic Kota Formation of India.
Proceedings of the Royal Society, London Ser. A.
188: 221-228.

1979. Some characteristics of Barapasaurus
lagorei, a sauropod dinosaur from the Lower
Jurassic of Deccan, India, Pp. 204-261, In Laskar,
B. & Raja Rao, C.S. (eds) ‘Fourth International
Gondwana Symposium’. (Geological Survey of
India: Calcutta).

JAIN, S.L. & ROY CHOWDHURY, T. 1987. Fossil
vertebrates from the Pranhita-Godavari valley
(India) and their stratigraphic correlation. Pp. 219-
228, In McKenzie, G.D. (ed.) “Gondwana Six:
Stratigraphy, Sedimentology and Palaeontology’,
Geophysical monograph 41.

JOLLY, A., BAJPAI, S. & SRINIVASAN, S. 1990.
Indian sauropod nesting sites (Maastrichtian,
Lameta Formation): a preliminary assessment of
the taphonomic factors at Jabapur, India. Pp. 78-
81. In Sahni, A. & Jolly, A. (eds) ‘Cretaceous
Event Stratigraphy and the Correlation of the In-
dian Nonmarine Strata'. (Panjab University:
Chandigarh).

KHOSLA, A. & SAHNI, A. 1995, Parataxonomic clas-
sification of Late Cretaceous dinosaur eggshells
from India. Journal of the Palaeontological
Society of India. 40: 87-102.

GONDWANAN DINOSAURS OF INDIA

KOHRING, R., BANDEL, K., KORTUM, D. & PAR-
THASARTHY, S. 1996. Shell structure of a
dinosaur egg from the Maastrichtian of Ariyalur
(Southern India). Neues Jahrbuch fiir Geologie
und Paláontologie, Monatshefte 1996: 48-64.

KUTTY, T.S. 1969. Some contribution to the stratig-
raphy of the Upper Gondwana Formation of the
Pranhita Godavari Valley, Central India. Journal
Geological Society of India 10: 33-48.

KUTTY, T.S., JAIN, S.L. & ROY CHOWDHURY, T.
1987. Gondwana sequences of the northern Pran-
hita-Godavari Valley: its stratigraphy and ver-
tebrate faunas. Palaeobotanist 36: 214-229.

KUTTY, T.S. & SENGUPTA, D.P. 1989. The Late
Triassic Formations of the Pranhita-Godavari Val-
ley and their vertebrate succession — a reappraisal.
Indian Journal of Earth Science 16: 189-206.

LOYAL, R.S., 1984. Bearing of paleontological data on
the collision models of the Indian plate. Pp. 284-
285. 27th International Geological Congress',
Moscow.

1985. Migration routes of Gondwana vertebrates
and drift models of the Indian Plate. 6th Interna-
tional Gondwana Symposium, Ohio, 64
(abstract).

MATLEY, C.A. 1921. Thecoprolites of Pisdura, Central
Provinces. Records of the Geological Survey of
India 74: 535-547.

1929, The Cretaceous dinosaurs of the Trichinopoly
district and the associated rocks with them.
Records of the Geological Survery of India 61:
337-349.

MAULIK, P.K. & RUDRA, D.K. 1986. Trace fossils
from the freshwater Kota Limestone of the Pran-
hita — Godavari Valley, South Central India. Bul-
letin of Geological, Mining and Metallurgical
Society of India 54: 114-123.

McINTOSH, J.S. 1990. Sauropoda. Pp. 345-401. In
Weishampel, D.B., Dodson, P. & Osmólska, H
(eds) ‘The Dinosauria'. (University of California
Press: Berkeley).

MOHABEY, D.M. 1982. On the occurrence of
dinosaurian fossi] eggs from Infratrappean Lime-
stones, Kheda, Dist. Gujarat. Current Science 52:
1194,

1984. The study of dinosaurian eggs from Infratrap-
pean limestones in Kheda district, Gujarat. Jour-
nal of Geological Society of India 25: 329-337.

MOHABEY, D.M. & MATHUR, U.B., 1989, Upper
Cretaceous dinosaur eggs from new localties of
Gujarat. Journal of Geological Society of India 33:
32-37.

MOHABEY, D.M., UDHOJI, S.J. & VERMA, K.K.
1993. Palaeontological and sedimentological ob-
servations on non-marine Lameta Formation
(Upper Cretaceous) of Maharashtra, India: their
palaeoecological and palaeoenvironmental sig-
nificance, Palaeogeography, Palaeoclimatology,
Palaeoecology, 105: 83-94.

637

MOLNAR, R.E. 1980. Australian Late Mesozoic ter-
restrial tetrapods: some implications. Mémoire de
la Société géologique de France 139: 131-143.

MOLNAR, R.E. & FREY, E., 1987. The paravertebral
elements of the Australian ankylosaur Minmi
(Reptilia: Ornithischia, Cretaceous). Neues
Jahrbuch für Geologie und Paläontologie, Ab-
handlungen 175: 19-37.

PRASAD, G.V.R. & KHAJURIA, C.K. 1990. Late
Mesozoic mammals from India and their sig-
nificance. Pp. 50-51. In Sahni, A. & Jolly, A. (eds)
“Cretaceous Event Stratigraphy and the Correla-
tion of the Indian Nonmarine Strata’, (Panjab
University: Chandigarh).

PRASAD, G.V.R. & SAHNI, A. 1988. First Cretaceous
mammal from India. Nature 332: 638-640.

SAHNI, A. 1984. Cretaceous-Paleocene terrestrial
faunas of India: Lack of endemism during drifting
of the Indian Plate. Science 226: 441-443.

1989. Paleoecology of the Late Cretaceous dinosaur
eggshell sites from peninsular India. Pp. 179-185.
In Gillette, D.D. & Lockley, M.G. (eds)
‘Dinosaur Tracks and Traces’ (Cambridge
University Press: Cambridge).

1993, Eggshell ultrastructure of Late Cretaceous
Indian dinosaurs. Pp. 187-193 In Kobayashi, J.,
Mutevei, H. & Sahni, A. (eds) ‘Structures, forma-
tion and evolution of fossil hard-tissues’. (Tokai
University Press: Tokyo).

SAHNI, A. & BAJPAI, S. 1991. Eurasiatic elements in
the Upper Cretaceous non marine biotas of penin-
sular India. Cretaceous Research 12: 177-183.

SAHNI, A. & KHOSLA, A. 1994. Paleobiological,
taphonomical and paleoenvironmental aspects of
Indian Cretaceous nesting sites. Pp. 215-224. In
Lockley, M.G., dos Santos, V.F., Meyer, C.A. &
Hunt, A. (eds) ‘Aspects of Sauropod
Paleobiology’, Gaia 10: 215-224.

SAHNI, A., RANA, R.S. & PRASAD, G.V.R., 1987.
New evidence for paleobiogeographic intercon-
tinental Gondwana relationships based on late
Cretaceous-earliest Paleocene coastal faunas from
Peninsular India. Pp. 207-218. In McKenzie, G.D.
(ed.) ‘Gondwana Six: Stratigraphy, Sedimentology
and Palaeontology’. Geophysical monograph 41.

SAHNI, A., TANDON, S.K., JOLLY, A., BAJPAI, S.,
SOOD, A. & SRINIVASAN, S. 1994. Upper
Cretaceous dinosaur eggs and the nesting sites
from the Deccan volcano-sedimentary province of
peninsular India. Pp. 204-226. In Carpenter, K.,
Hirsch, K. & Horner, J. (eds) 'Dinosaur eggs and
babies'. (Cambridge University Press:
Cambridge).

SARKAR, A. BHATTACHARYA, S.K. &
MOHABEY, D.M. 1991. Stable isotope analysis
of dinosaur eggshells, paleoenvironmental im-
plications. Geology 19: 1068-1071.

SENGUPTA, S. 1970. Gondwanan sedimentation
around Bheemaram (Bhimaram), Pranhita-
Godavari Valley, India. Journal of Sedimentary
Petrology 40: 140-170.

638

SHARMA, K.K. 1987. Crustal growth and two stage
India-Eurasia collision in Ladakh. Tectonophysics
134: 17-28.

SINGH, I.B. 1981. Palaeoenvironment and palaeogeog-
raphy of the Lameta Group sediments (Late
Cretaceous) in Jabalpur area, India. Journal of the
Palaeontological Society of India 26: 38-53.

SRIVASTAVA, S., MOHABEY, D.M., SAHNI, A. &
PANT, S.C. 1986. Upper Cretaceous dinosaur egg
clutches from Kheda district Gujarat, India: their
distribution, shell ultrastructure and palaeoecol-
ogy. Palaeontographica Ser. A 193: 219-233.

TANDON, S.K., VERMA, V.K., JHINGRAN, V.,
SOOD, A., KUMAR, S., KOHLI, R.P. & MIT-
TAL, S. 1990. The Lameta beds of Jabalpur,
Central India: Deposits of fluvial and pedogeni-
cally modified semi-arid fan-palustrine flat sys-
tem. Pp. 27-30 In Shani, A. & Jolly, A. (eds)
“Cretaceous Event Stratigraphy and Correlation of
Indian Nonmarine Strata’, (Panjab University:
Chandigarh).

TASCH, P., SASTRY, M.V.A., SHAH, S.C., RAO,
B.RJ., RAO, C.N. & GHOSH, S.C., 1973. Es-
thiriids of the Indian Gondwanas: Significance for
continental rift. Pp. 445-452. In Campbell, K.S.W.

MEMOIRS OF THE QUEENSLAND MUSEUM

(ed.) “Gondwana Geology.’ (Australian National
University Press: Canberra).

VIANEY-LIAUD, M., JAIN, S.L. & SAHNI, A. 1987.
Dinosaur eggshells (Saurischia) from the Late
Cretaceous Intertrappean and Lameta Formation
(Deccan, India). Journal of Vertebrate Paleontol-
ogy 7: 408-424.

WALKER, A.D. 1964. Triassic reptiles from the Elgin
area: Ornithosuchus and origin of carnosaurs.
Philosophical Transactions of Royal Society of
London Ser. B 248: 53-134.

YADAGIRI, P. 1984. New symmetrodonts from the
Kota Formation (Early Jurassic), India. Journal of
Geological Society of India 25: 514-521.

1985. An amphidontid symmetrodont from the
Early Jurassic Kota Formation, India. Zoological
Journal of Linnean Society 85: 411-417.

YADAGIRI, P. & AYYASAMI, K., 1979. A new
stegosaurian dinosaur from Upper Cretaceous
sediments of South India. Journal of Geological
Society of India 20: 521-530.

YADAGIRI, P. & RAO, B.R.J., 1987. Contribution to
the stratigraphy and vertebrate fauna of Lower
Jurassic Kota Formation, Pranhita-Godavari Val-
ley, India. Palaeobotanist 36: 230-244.

OBSERVATIONS ON THE AUSTRALIAN ORNITHOPOD DINOSAUR,
MUTTABURRASAURUS

RALPH E. MOLNAR

Molnar, R.E. 1996 12 20: Observations on the Australian ornithopod dinosaur, Muttabur-
rasaurus. Memoirs of the Queensland Museum 39(3) 639-652. Brisbane. ISSN 0079-8835.

New material and a restudy of old suggests that Muttaburrasaurus is not an iguanodontid
ornithopod, but probably diverged from the iguanodontian-hadrosaur lineage prior to the
divergence of taxa such as Dryosaurus and Tenontosaurus. The type skull differs from a
second, older skull from the Allaru Mudstone in ways that suggest a trend toward the
evolution of a continuous more-or-less planar sheet of enamel along the labial face of the
maxillary dentition. Teeth from Lightning Ridge (Griman Creek Fm.) may pertain to a more
plesiomorphic species of Muttaburrasaurus. [_] Muttaburrasaurus, Ornithopoda, Australia.

Ralph E. Molnar, Queensland Museum, P.O. Box 3300, South Brisbane, Queensland, 4101,

Australia; 15 August 1996.

The discovery of a new skull of Muttaburra-
saurus (from ‘Dunluce’, north-central
Queensland) and further preparation of the
holotype skull has yielded interesting new infor-
mation bearing on its evolution and systematic
position. Because preparation of both skulls
proved more difficult and hence slower than ex-
pected, it is worthwhile presenting interim results
that substantially alter the phylogenetic inter-
pretation of Bartholomai & Molnar (1981) and
Molnar (1984). New information on cranial fea-
tures, especially those relevant to understanding
the phylogenetic position of Muttaburrasaurus in
relation to the classification presented by Sereno
(1986), are given here. Some remarks on
postcranial features are also included.

Collection designations. AM, Australian Museum,
Sydney; NMV, Museum of Victoria, Melbourne;
QM, Queensland Museum, Brisbane.

DESCRIPTION

HOLOTYPE SKULL (QM F6140). The original
description of the skull of Muttaburrasaurus
langdoni included no description of palatal or
braincase structure and these are described here,
insofar as feasible with the still incomplete state
of preparation. Comments on the maxilla, the post-
orbital bar and the general form of the skull
follow.

Quadrate. The quadrate is a robust columnar
element, similar to those of other advanced
ornithopods. The articular end remains in articu-
lation with the mandible, and so cannot be ex-
amined in detail. The body is straight in lateral
view as in Iguanodon bernissartensis, not curved
as in 7. atherfieldensis. The posterior surface,
inclined to face slightly medially, tapers upwards

to become a sharp ridge formed by the meeting
of the lateral and medial surfaces of the element.
A broad pterygoid wing projects anteromedially
to meet the pterygoid in a smooth curve (Fig. 1).
Posteriorly this process bears a marked concavity,
adjacent to a vertical sulcus on the upper part of
the body of the bone. The lateral face (jugal wing)
projects anteriorly beneath the quadratojugal and
is embayed by the round margin of the quadrate
foramen.

Exoccipital. Presumably the exoccipitals and
opisthotics are fused, but only the exoccipital
portion of the compound element is easily visible,
to best advantage on the left side. A dorsoventrally
deep paroccipital process extends posterolateral-
ly from the condyar region. Although the distal
end is lacking, there is no indication that the
process was declined distally as in /guanodon and
hadrosaurids (Fig. 2).

Medially a stout pillar descends from the par-
occipital process to abut on the basioccipital and
form the dorsolateral part of the occipital con-
dyle. Part of the condyle, separated by a shallow
groove from the ventral moiety, may be made up
by the exoccpital. Because the contacts are fused
it isn't clear that this is actually part of the exoc-
cipital, as there is another possible contact — a
discontinuity in the surface grain of the bone —
at the junction of the pillar and the base of the
paroccipital process. Just anterior to the
presumed junction, there is an anteroposteriorly-
directed concavity. This has yet to be fully
cleaned but is likely to accommodate foramina
for cranial nerves X, XI and XII.

Basioccipital. Only two points may be added to
the original description. The articular surface of
the condyle extends forward to form a lip

640

FIG. 1. Cutaway interpretation of braincase and
posterior palatal structures of Murtaburrasaurus
langdoni, QM F6140. Bars indicate cuts through the
jugal arch, prefrontal region and quadrate wing of
pterygoid. Dots indicate broken bone surfaces.
Bt=basal tubera (basisphenoid), Ep=epipterygoid,
F=frontal, Oc=occipital condyle, P=parietal,
Pal=palatine, Po=paroccipital process, Pt=pterygoid
(quadrate wing), Q=quadrate, Sq=squamosal,
V=presumed trigeminal foramen.

‘overhanging’ the neck of the condyle, so the
surface describes an arc in excess of 180° as in 7.
atherfieldensis (Norman, 1986, fig. 18). This lip
connects with the basal tubera by a low, blunt keel
on the ventral face of the condylar neck.
Basisphenoid. Only the inferior, and some of the
lateral, surface of the basisphenoid is clearly
visible and it carries large basal tubera, confluent
medially to form a rugose transverse ridge. The
left tuber is deflected sharply posteriorly and the
right is incomplete. A large foramen on each side
penetrates the tubera at the junction with the keel
connecting to the occipital condyle. On the mid-
line the ventral face of the confluent tubera has a
larger foramen (Fig. 3). This foramen is in the
same relative position as that of the median eus-
tachian foramen (f. intertympanicum) of modern
crocodilians, but whether or not it represents that
structure cannot be determined at present.
About 2cm in front of the tubera, two robust
4cm-long basipterygoid processes extend
ventrolaterally and somewhat anteriorly to con-
tact the pterygoids. Just anterior to the tuber on
the left side, a sheet of bone curls out from the
basisphenoid (or prootic?) joining it to the quad-
rate process of the pterygoid. Although so far only
partly prepared this seems to be very similar to a
character state used to diagnose the
Pachycephalosauria: Prootic-basisphenoid plate

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 2. Skulls of A, Iguanodon bernissartensis; and B,
Muttaburrasaurus langdoni in occipital view, to
show the difference in forms of the paroccipital
processes. To scale. (A redrawn from Norman, 1980,
and B from Bartholomai & Molnar, 1980.)

present which extends laterally from the brain-
case, contacting the pterygoquadrate wing and
effectively separating subtemporal and occipital
regions (Sereno, 1986). In Muttaburrasaurus this
contact also separates the subtemporal (adductor)
chamber from the occipital region.

Pterygoid. Both right and left pterygoids are
preserved, but that of the right is visible only in
the basipterygoid-quadrate region, and that of the
left is fragmentary. The quadrate process is a
vertical sheet, with an acutely angled posterior
margin, projecting posteriorly from the basip-
terygoid contact and laying ventral to the
pterygoid wing of the quadrate (‘Pt’ in Fig. 1,
where the entire extent is not shown). It bears a
broad, blunt median longitudinal ridge.
Anteroventrally it contacts another sheet laying
in a plane inclined at about 45? to the horizontal.
The extent of this portion cannot be determined
without further preparation. The basipterygoid
joint, which seems to have been immobile, is just
anterior to the contact between these two sheets.
In front of the joint a tapered process of the
pterygoid extends medially and curves through
180° to form a (more or less) horizontal hook-like
structure (Fig. 3, ‘mp’).

Epipterygoid. A few fragmentary sheets of much
abraded bone between the laterosphenoid and the
pterygoid on the right side probably represent the
epipterygoid. They show little other than that one
was present in the usual position (Fig. 1, ‘Ep’).
Palatine. The right palatine is visible and some-
what worn. It seems to be a thick, triangular plate
rising sharply from the maxilla, with its dorsal
margin just below or contacting the prefrontal
(Fig. 1).

Maxilla. In QM F6140 the maxillae are orientated
so that their toothrows have been brought

OBSERVATIONS ON MUTTABURRASAURUS

641

FIG. 3. The skull of Muttaburrasaurus langdoni, QM F6140, in ventral view. Bj=basipterygoid joint, Bp=basip-
terygoid process (basisphenoid), D=dentary, H=hyoid piece, M=maxilla, mf=median ventral foramen of
basisphenoid, mp=medial hooked process of pterygoid, Pb-pterygoid-braincase contact. Other abbreviations

as in Fig. 1. Scale = 50mm.

together, with the crowns only 5cm apart,
directed toward one another with the labial faces
of the teeth facing ventrally. The maxillary
toothrows of the ‘Dunluce’ skull (QM F14921)
have also been brought together, about 3cm apart,
by crushing. The orientation of the maxillae in
this specimen suggests that those of the holotype
have been folded underneath the skull, so that
their lateral surfaces have come to face ventrally.
In fact this has also happened to a lesser extent
in the ‘Dunluce’ skull, as shown by fracturing of
the posterior part of the right maxilla. Such frac-
tures are not apparent in the type skull, which has
either suffered plastic deformation or rotation of

the maxillae about their dorsal margins. Since
such rotation has been proposed as a masticatory
device in the iguanodontian-hadrosaurian lineage
(e.g., Weishampel, 1984; Norman & Weisham-
pel, 1985), it is interesting to see if it was present
in Muttaburrasaurus. 'The lateral surface of the
posterior part of the left maxilla of the type skull
curves through a 90? arc (in the frontal plane)
beneath the orbit, from the nearly horizontal sur-
face adjoining the jugal to the nearly vertical
surface at the alveoli. But just forward of this the
alveolar region has been flexed so that the teeth
become directed medially, as described above.
This suggests that this maxilla at least is

642

deformed, and provides no
evidence for (or against)
flexure at the maxillary-nasal
junction. The reconstruction of
the type skull (Fig. 4) corrects
for this distortion.

Postorbital bar. Unlike the
condition of most other or-
nithopods the postorbital bar of

MEMOIRS OF THE QUEENSLAND MUSEUM

M. langdoni is more extensive
transversely than antero-
posteriorly (Fig. 5). In other
words, itis broad. The descend-
ing process of the postorbital is
4cm wide (transversely) but
only 1.75cm in the longitudinal B
dimension. The ascending
process of the jugal is similarly
formed, so the postorbital bar
partly divides the orbital cavity
from the adductor chamber. A
similar state, postorbital bar
transversely broadened with an
interdigitating postorbital-
jugal suture, is used by Sereno
(1986) to partly define the
Goyocephalia, the node just
distal to the Pachycephalo-
sauria. The contact between the
jugal and postorbital in M.
langdoniis tight, but only actual-
ly interdigitates at the tip of the
ascending process of the jugal.
Skull form. Molnar (1995)
presented an argument, based
on the type skull, that the pos-
torbital region was enlarged relative to those of
other ornithopods. This may be seen in Fig. 6, and
has directed the orbits very slightly to the front.
Although dorsoventrally deep, the snout was
seemingly relatively narrow. The ‘Dunluce’ skull,
in spite of its crushing, confirms that the postor-
bital region was approximately twice as broad as
the snout.

'DUNLUCE' SKULL (QM F14921). The second
skull was collected by Dr Mary Wade in 1987
from the Allaru Mudstone (not the Toolebuc Fm.,
as reported by Molnar, 1995) on Dunluce station,
about halfway between Hughenden and Rich-
mond in north-central Queensland. The Allaru
Mudstone, a gray calcareous marine argillite, un-
derlays the Mackunda Fm. from which the
holotype derives. Thus this second specimen
derives from an older population (although still

FIG. 4. Reconstruction of skulls of A, Muttaburrasaurus langdoni (QM
F6140); and B, Muttaburrasaurus sp. (QM F14921). Restored portions in
dashed lines. The region of the jugal and of the junction between the nasal
bulla and -body of the maxilla in QM F14921 is worn and hence the
structures shown (form of jugal-maxillary contact and clefts beneath bulla)
may represent structures just below the surface of the skull. The last
maxillary tooth is very poorly preserved, so it is not certain if the last
alveolus as drawn represents a single alveolus or two.

Albian) than the type. The skull is probably more
complete that the holotype skull but has been
laterally sheared and crushed, so that the the nasal
bulla has been flattened and the skull roof has
come to lie almost coplanar with the right side
(Fig. 7). The postorbital region is severely
damaged on the left, so that the quadrate now
rests in the position ofthe posterior orbital margin
(Fig. 7A). The postorbital is missing entirely on
this side, and the (empty) impression for its recep-
tion may be seen on the ascending process of the
jugal. Because of the damage, and because
preparation is still in a preliminary stage, a full
description of this skull will not be given here.
Instead selected relevant features are reported.
This skull is identified as Muttaburrasaurus on
the basis of its general form, as well as the
apomorphic nasal bulla and tooth form. Because
of the damage and incomplete preparation, as

OBSERVATIONS ON MUTTABURRASAURUS

FIG. 5. The skull of Muttaburrasaurus langdoni (QM F6140) in anterolateral
view, showing the breadth of the postorbital bar (Pob). Dots indicate broken
bone surface; open dots indicate matrix. m=mandible, J=jugal, o=orbit,
a=antorbital fossa, r=root of maxillary tooth, L=lachrymal, Other abbrevia-

tions as in Fig. 3.

well as the possession of partly erupted maxillary
crowns anteriorly in the toothrow (those of M.
langdoni are completely erupted along the entire
visible maxillary toothrow) and differences in
form, it is premature to assign it to M. langdoni,
and it is referred to Muttaburrasaurus sp.

Nasal bulla. The anterior termination of nasal
bulla is not preserved on the type skull, but on
OM F14921 it seems to be present. Although
much of the bony surface of this specimen has
been lost surface bone is present at the front
margin of the nasal bulla. Here the surface forms
an anteriorly-directed vertical face, which then
presumably bends through a right angle — the
flexure itself is still embedded in carbonate — to
give an adjacent horizontal bone surface extend-
ing anteriorly to the break at the front of the
specimen. Although the nasal bulla has been flat-
tened during preservation, it seems unlikely such
transverse compression could create an illusory
anterior termination to the bulla. This observation
has been incorporated into the new reconstruction
of the skull in Fig. 4B.

Although the nasal bulla of QM F4921 is badly
crushed and incompletely prepared, some fea-
tures can be better seen than in the holotype. On
the right side, a crescentic element makes up the
anterodorsal part of the lateral wall of the bulla
(Fig. 7B). The left is worn and incompletely

643

prepared and so difficult to in-
terpret, but gives no reason to
doubt or modify the interpreta-
tion here given. This element
contacts that, presumably the
nasal, making up the dorsal
roof of the bulla posteriorly,
and what seem to be slit-like
nares anteriorly. The concave
inferior margin of the bone is
unbroken, so that the
posteroventral part of the
lateral wall must have been
formed by another element, not
preserved. Thus the cresecentic
element may be, or be part of,
the premaxilla. This element
cannot be seen on the bulla of
the type skull, where instead
the lateral wall seems to be
made up of a single element.
There is no clear indication of
narial openings on such of the
maxillae as is preserved
anterior to the bulla.

Ventrally the bulla seems to
have been separated from the
underlying maxilla by two clefts, separated from
each other by a thin sheet of bone (Fig. 7A). This
sheet is continuous with the maxilla anteriorly, so
the lower of the clefts seems to be part of the
maxilla. These structures are also apparent on the
left side, but cannot be seen in the type skull,
presumably because they are internal structures
exposed by wear on the ‘Dunluce’ skull. A thin
median septum is apparent within the bulla, as in
the type skull (which also has a second vertical
sheet, described in Molnar [1984] more laterally
placed at the left edge of the left narial opening).
The bulla of the type skull also shows on its badly
worn left side one, or perhaps two, thin, seemingly
horizontal internal sheets of bone.

Antorbital fossa. The apparent antorbital fenestra
on the left side of the 'Dunluce' skull (Fig. 7) is
entirely surrounded by abraded bone that had
been exposed prior to discovery. This has ex-
posed a triangular antorbital fossa of respectable
size, approximately 5 x 7cm.

Tooth replacement. The holotype skull is unusual
among ornithopods, as well as other reptiles, in
that there is no indication of the sequence of tooth
replacement in the maxillary dentition. In fact,
replacement teeth can be seen only where the
skull has been broken transversely behind the
nasal bulla. (The dentary crowns now visible

644

FIG. 6. Reconstructed skull of Muttaburrasaurus
langdoni (QM F6140) in dorsal view. (From Molnar,
1995.)

seem to show an alternating pattern of replace-
ment.) The ‘Dunluce’ skull has a complete, or
almost complete but bady weathered, right max-
illary toothrow. There are 22 tooth positions, with
what seem to be replacement crowns in positions
4 and 6 (from the front). Only 15 positions are
visible in the type skull, probably 6 to 21 assuming
that it, too, had 22 maxillary teeth, and the
anterior-most teeth are not well preserved. Thus
some replacement in the type skull cannot com-
pletely be ruled out, but there is no indication of it.

TEETH

Maxillary teeth. Only the maxillary teeth of M.
langdoni were described by Bartholomai & Mol-
nar (1981). Although 17 maxillary crowns are
exposed on the left side of the type skull, only
three preserve substantial portions of the lateral
face. These crowns differ from those of all other
ornithischians in that they show an almost flat
(slightly convex) lateral face marked by a series
of parallel ridges and grooves, all of which are
equally developed (Fig. 8). The ridges extend to
the base of the crown anteriorly (mesially) but
become progressively shorter posteriorly. They
— with one possible exception — do not con-
verge. None of the crowns shows clear indication
of a primary ridge. On two, however, a primary
ridge adjacent to the posterior (distal) margin of
the crown is evident. This presumed ridge is only
marked as such by an adjacent sulcus along the
posterior margin (this sulcus is the ‘deep groove’
of Bartholomai & Molnar, 1981). Each crown
slightly overlaps that posterior to it.

Maxillary crowns are also known in the
*Dunluce' skull, QM F14921, and probably — as
isolated teeth — from Lightning Ridge. The latter
were held in the private collection of Ms
Elizabeth Smith but have been donated to the
Australian Museum, and represented by casts in
the Queensland Museum. As mentioned pre-

MEMOIRS OF THE QUEENSLAND MUSEUM

viously, most of the exposed maxillary crowns of
the ‘Dunluce’ skull are poorly preserved, with
their lateral faces badly weathered, but a for-
tuitous break on the left side exposed a well-
preserved crown near the posterior end of the
toothrow (Fig. 8B, E), probably number 19 or 20.
The omament of this crown is like that of the the
type skull, but differs in some notable ways. The
lateral face is almost flat, slightly concave, and
bears 11 low ridges. These ridges are relatively
larger than those of the holotype skull and broader,
with flat — in some cases slightly grooved —
tops. Most apparent, however, is the clear primary
ridge near the distal margin with 8 of the low
ridges mesial to it and 3 distal. In some ways this
crown more closely resembles those of Atlascopco-
saurus loadsi (NMV P157390) than those of the
holotype skull. However, the maxillary crowns of
QM F14921 do resemble those of the holotype in:
outline of the crown; almost flat lateral face; large
number of low, parallel grooves and ridges; and
a low primary ridge near the distal margin.

The teeth from Lightning Ridge (New South
Wales) from the Smith collection include one
worn specimen preserving the root and base of
the crown (QM F14420 [cast]), and a second that
is almost complete (OM F14421 [cast]). (A third
tooth, AM F81865, may also pertain to Muttabur-
rasaurus). They derive from the Griman Creek
Fm. at McNamara's Three-Mile Field and were
collected in, or just before, 1986. These were
previously noted, but not described, by Molnar &
Galton (1986). The first is too incomplete for
consideration here — and may derive from the
dentary — but the second shows a crown basical-
ly similar to that of the 'Dunluce' skull (Fig. 8F).
It has a series of 6 low, parallel ridges mesial to a
primary ridge, fewer than in QM F14921, and 1-2
distal to it in a shallow sulcus. It is also smaller
than the ‘Dunluce’ tooth, about 69% as wide. The
crown is referred to Muttaburrasaurus, rather
than Arlascopcosaurus, on two character states: a
curved (not angulate) proximal margin to the
crown; and the extension of the primary ridge to
the base of the lateral face. In A. loadsi the
primary ridge extends to the distal margin of the
crown rather than the proximal, and there is no
real proximal margin, rather the mesial and distal
margins converge (Fig. 8G).

Unlike any of the crowns in either the type or
*Dunluce' skulls, this crown (QM F14421) shows
the lingual surface, which bears a series of at least
four parallel, shallow grooves. Unfortunately,
due to the silicification of the specimen it is not
possible to determine if the surface bore enamel

OBSERVATIONS ON MUTTABURRASAURUS 645

FIG. 7. The partially prepared skull of Muttaburrasaurus sp. (QM F14921), from the left A, and right B. The
skull has been crushed, and the apparent antorbital fenestra and opening in the lower part of the nasal bulla are
artefacts of erosion on the left side. The left quadrate has been displaced anteriorly to the region of the anterior
margin of the laterotemporal fenestra.B=nasal bulla; R=skull roof.

and, if so, how much. The maxillary crowns of A. plete crown referred to Muttaburrasaurus sp.
loadsi also have a series of shallow, parallel from near Hughenden (QM F12541) (Fig. 8C).
grooves on their lingual faces as does an incom-

646

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 8. Teeth of Muttaburrasaurus. A, Anterior maxillary teeth of M. langdoni (QM F6140). B, Posterior
maxillary crown of M. sp. (QM F14921). C, Dentary crowns of M. sp. (QM F12541). D,, Right 11th or 12th
maxillary crown of M. langdoni (QM F6140). Section through crown at bottom, showing position of presumed
primary ridge. E, Left posterior maxillary crowh of M. sp. (QM F14921). Section through crown, as in E, at
bottom. F, Maxillary crown of M. sp. (QM F14421). G, Two maxillary crowns of Atlascopcosaurus loadsi
(NMV 157390).H, Reconstructed dentary crown of Muttaburrasaurus, based on QM F12541 and QM F6140,

Dentary teeth. Dentary crowns have been ex-
posed on the type skull, and also occur in a very
fragmentary specimen (QM F12541) from the
Allaru Mudstone from ‘Iona’, southeast of
Hughenden. This was tentatively referred by
Molnar (1984) to 'an unusually large
hypsilophodont' on the basis of what are now
recognised as plesiomorphic ornithopod or even
euornithopod character states. This specimen is
part of an associated partial skeleton, which has
remained in Hughenden. It consists of parts of
two dentary crowns found in a small block of
carbonate, together with natural molds of parts of
four other teeth. One of these represents the lin-
gual face of the maxillary tooth mentioned above.

No dentary crown is completely exposed in
either specimen, so only the upper half (or a little
more) toward the tip may be described. The
crown seems to be rounded, perhaps leaf-shaped,
in form. À prominent carina occupies the middle,
with a series of lower, parallel ridges, varying in
size, occupying the face mesial and distal to it
(Fig. 8C, H). These ridges are inclined to the
carina and terminate at the margin in small, taper-
ing denticles. The denticles are featureless, ex-
cept that some divide distally, so that one ridge

may bear two denticles. In at least one specimen
(QM F12541) the carina itself bears four or five
low, parallel ridges basally. In all discernible fea-
tures the dentary crowns of the type skull match
that of QM F12541. A single displaced dentary
tooth of the *Dunluce' skull is also indistinguish-
able from these.

POSTCRANIUM (HOLOTYPE)

Manus. Initially Muttaburrasaurus was believed
to be an iguanodontian, closely related to
Camptosaurus (Bartholomai & Molnar, 1981):
this work was carried out before the influence of
phylogenetic systematics was felt in ornithis-
chian taxonomy. At that time an incomplete flat-
tened, tapering element was thought perhaps to
represent a thumb-spike. If Muttaburrasaurus is
indeed an ankylopollexian, this is a plausible
interpretation. However, the bone is poorly
preserved, abraded and broken. Therefore — in
the absence of any assurance that Muttabur-
rasaurus is in this lineage and is as or more
advanced than Camptosaurus — the piece cannot
be reliably identified as a thumb-spike and so
does not constitute evidence that Muttabur-
rasaurus is an ankylopollexian.

OBSERVATIONS ON MUTIABURRASAURUS

FIG. 9. Reconstruction of the skeleton of Muttabur-
rasaurus langdoni showing what parts are actually
preserved. Scale = 1m.

Pes. During the reconstruction (Fig. 9) of M.
langdoni, the pedal elements were re-examined.
Further cleaning of the pieces revealed a contact
between the proximal and distal parts of metatar-
sal III. This shows that the metatarsus is shorter
than originally believed (Fig. 10), and reduces the
anomalous disparity in length between metatarsal
III and the others.

SYSTEMATIC POSITION

Since the original description of Muttaburra-
saurus much new work on advanced ornithopods
has appeared. The original interpretation of its
systematic position was based significantly on
Dodson (1980), but the re-interpretation of or-
nithischian phylogeny by Sereno (1986) suggests
that a review of the position of Muttaburrasaurus
is in order. Consideration of the original and new
character states suggests that Muttaburrasaurus
is not an iguanodontid (or camptosaurid), but a
basal ornithopod whose lineage diverged early
from the iguanodont-hadrosaur lineage.

Because of the incompleteness of the Muttaburra-
saurus material (Fig. 9) many of the character
states used by Sereno (1986) are not observable.
The remaining states are assessed for Muttaburra-
saurus, based on the type specimen but with some
observations from the ‘Dunluce’ specimen. Three
(mandibular dentition offset medially; moderate
coronoid process; pubic peduncle of ilium less
robust than ischial peduncle) of the seven states
used by Sereno to diagnose the Genasauria are
present in Muttaburrasaurus, so its inclusion in
that group is reasonable.

States are considered in order of the nodes they
define.

ORNITHOPODA

647

External opening of the antorbital fossa of
moderate size or smaller: Present. The antorbital
opening is present but small in the type skull.
Unfortunately shedding of the surficial bone
makes it impossible to tell just how small the
opening was. The ‘Dunluce’ skull seems to have
a relatively large antorbital opening for an or-
nithischian, however as mentioned previously
this is an artefact arising from the loss of surficial
bone. The type skull is fortuitously broken at the
antorbital fenestra and shows that a relatively thin
plate of bone laterally walls the antorbital fossa
around the small antorbital opening.

The remaining three character states are not
presently determinable, but the single state
known gives no reason to doubt that Muttabur-
rasaurus is an ornithopod.

HYPSILOPHODONTIA
Prepubic process rod-shaped; wider transversely
than tall dorsoventrally: Absent. The prepubic
process in Muttaburrasaurus is mediolaterally
compressed, taller than wide and not rod-shaped.
The other three character states are not deter-
minable for Muttaburrasaurus, but the state here
assessed gives no evidence for its being a hyp-
silophodontian.

IGUANODONTIA
Leaf-shaped denticles: Absent. Although den-
ticles are present on the dentary teeth they are
tapered in form, not leaf-shaped (Fig. 8C, D). The
maxillary crowns do not (observably) have den-
ticles.
Strong primary ridge on medial surface of den-
tary crowns: Present (Fig. 8C, D). Strong primary
ridges are present on the dentary crowns in the
type and 'Dunluce' skulls, and QM F12541.
External opening of the antorbital fossa is rela-
tively small or entirely absent: Present, as men-
tioned above, in the type skull.
Quadratojugal reduced in size relative to the
quadrate: Absent. As may be seen from Fig. 44°
(and fig. 2, Bartholomai & Molnar, 1981) the
quadratojugal is about as large (in its lateral ex-
tent) as the quadrate, maybe larger. It is not
reduced relative to the quadrate.
Femur with weak anterior intercondylar groove
and deep posterior intercondylar groove: Absent.
The femur does indeed have a deep posterior
intercondylar groove (Fig. 11A), but the anterior
intercondylar groove is deep as well (Fig. 11B).
The other nine character states are not presently
determinable, but as only two of the five deter-
mined agree with Sereno's states for iguanodon-

648

FIG. 10. Pes of M. langdoni (QM F6140) in dorsal
(right) and plantal (left) views. Scale = 150mm.

tians, Muttaburrasaurus is probably not an
iguanodontian.

DRYOMORPHA

Maxillary crown narrower anteroposteriorly
than the opposing dentary crown: Present. The
preservation of maxillary and dentary teeth in
juxtaposition in the snout region of the type skull,
shows that the maxillary crowns are indeed nar-
rower than the dentary crowns, which overlap in
their medial exposure. Overlap is less in the max-
illary toothrow (Fig. 8A).

Lateral maxillary primary ridge stronger than the
medial dentary primary ridge: Absent. Although
there is a primary ridge on the maxillary crowns
(not mentioned in Bartholomai & Molnar, 1981),
this ridge is placed adjacent to the anterior
(mesial) edge of the lateral face of the crown, and
is less prominent than the corresponding ridge of
the dentary crowns (Fig. 8A, B, E, F vs. C, D).
Diamond-shaped maxillary and dentary crowns;
crowns with rounded anterior and posterior
corners: Absent. As may be seen from Fig. 8,
neither maxillary nor dentary crowns have
diamond-shaped lateral faces, although in section
they do exhibit rounded anterior and posterior
margins (which is what I believe to be meant by
‘anterior and posterior corners’).

Space separating ventral margin of quad-
ratojugal from jaw articulation: Present, in the
type skull (Fig. 4A).

Ischial shaft round in cross section; transversely
compressed distally: Absent. Only the proximal
portions of both ischia are preserved so the
presence of distal compression cannot be
verified. However, the proximal parts of the shaft
are compressed not round in section.

MEMOIRS OF THE QUEENSLAND MUSEUM

Two (or two and a half) of the five determinable
character states (out of a total of eight) for
dryomorphs can be found in Muttaburrasaurus.
In view of the lack of match with the character
states for more basal nodes it seems likely that
Muttaburrasaurus is not a dryomorph, but the
states that can be assessed are equivocal.

ANKYLOPOLLEXIA
Close packing along the toothrow and in the
replacement series eliminating spaces between
the bases of the crowns of adjacent functional
teeth: Absent. The maxillary teeth are closely
packed, but present an almost rectangular form in
lateral view which eliminates the spaces between
the crowns of adjacent functional teeth without
leaving any space left over for the replacement
crowns (Fig. 8A). No replacement crowns are
visible in the maxillary series in either specimen,
although only a small part of the series is visible
in the ‘Dunluce’ skull. The dentary toothrows do
show replacement crowns filling in the gaps be-
tween the bases of adjacent crowns, and have
much more the standard appearance of or-
nithopod toothrows.
Prominent primary ridge on the lateral side of the
maxillary crown: Absent. Although, as mentioned
above, there are primary ridges on the lateral faces
of the maxillary crowns, they are not prominent.
And, due to their placement, they are almost
impossible to discern in lateral view in the type,
and can most easily be made out in cross section
where the maxillary teeth have been broken.
Ornamentation of apical margin of individual
denticles: Absent, no denticles show any indica-
tion of ornament.
Cervical neural spines very weak or absent: Ab-
sent. The cervical series has yet to be completely
prepared, but the neural spines seem to be well-
developed.
Robust, arching cervical postzygapophyses
posterior to axis: Present. Such postzy-
gapophyses are present at least in the anterior
postaxial cervical series.
Moderate opisthocoely in cervical vertebrae 4-9;
slight opisthocoely in dorsal vertebrae 1-2: Ab-
sent. The dorsals, having been found disarticu-
lated and rather broken, have so far resisted
efforts at determining their position. The anterior
cervicals show moderate opisthocoely, but a
posterior centrum is only slightly opis-
thocoelous.

Only one of these character states (or one and
a half, being liberal about the first state) of six
determinable (out of a total of nine) match those

OBSERVATIONS ON MUTTABURRASAURUS

FIG. 11. Left femur of M. langdoni (QM F6140) in A,
posterior; and B, anterior views. A block of matrix
(bar) still adheres to the proximal surface of the neck.
Scale = 200mm.

of ankylopellecians, thus Muttaburrasaurus
seems not to be an ankylopellecian, contrary to
the view of Bartholomai & Molnar (1981).

STYRACOSTERNA

At least 25 vertical columns in maxillary and
dentary tooth rows: Absent. Eighteen columns
are visible in place in the type skull. Ifthe portion
ofthe maxilla anterior to the anteriormost preserv-
ed tooth also held teeth there is space to accomodate
at most only three additional columns. The back of
the maxilla could accommodate one, or at most two,
teeth, making the maximum estimate of maxillary
columns 23. There are 22 on the right side in the
*Dunluce' skull.

Lanceolate-shaped maxillary crowns: Absent
(Fig. 8A, B, E, F).

Strong opisthocoely in cervical vertebrae, begin-
ning with the third cervical: Absent. The cervicals
are opisthocoelous, but only weakly.

649

Humerus with proximally and posteriorly
prominent head: Present. The right humerus of
the type has a prominent, posteriorly-projected head.
Distal end of prepubic process moderately ex-
panded dorsoventrally: Present? This was believed
to have been present by Bartholomai & Molnar
(1981), and cannot presently be re-assessed.
Pubis with distinct, stout iliac peduncle: Absent.
The proximal piece of pubis is abraded, but the
form and orientation of the acetabular region
suggests that a distinct iliac peduncle was not
present (Fig. 12).

Femur with deep anterior intercondylar groove:
Present, as discussed above (Fig. 11B).

Two or three of the determinable seven charac-
ter states (out of 12 total) are present, which
suggests that Muttaburrasaurus is not a
styracosternan.

IGUANODONTOIDEA

External nares enlarged: Absent. Although the
rostrum is missing from both skulls, the portions
of the narial regions preserved give no indication
that the nares were enlarged. The “Dunluce’ skull,
although crushed in this region, suggests that the
nares were slit-like and restricted to the anterior
and dorsal faces of the nasal bulla. The type skull
is less forthcoming, but shows no evidence con-
tradicting this. In Jguanodon and hadrosaurs the
naris is noticeably enlarged. There is no indica-
tion of similar enlargement in Muttaburrasaurus.
At least slight transverse narrowing of cranium
from postorbital region, dorsal view: Absent.
Molnar (1995) showed this is clearly not the case.
Paroccipital process relatively broader
proximally and narrower distally: Absent. This is
also not the case in Muttaburrasaurus, with the
paroccipital process being broad (dorsoventrally)
distally. The distal extremity of the process is
missing, but the posterior face of the quadrate is
nearly complete. It indicates that the distal end of
the paroccipital process was not declined as in
Iguanodon.

Manual digit V with at least three phalanges:
Present. If the manual digits were correctly inter-
preted by Bartholomai & Molnar (1981), then
manual digit V has at least three phalanges.

The one out of four determinable states (from a
total of 12) indicates that Muttaburrasaurus is not
an iguanodontoid.

On the basis of what is known about Muttabur-
rasaurus it cannot be accommodated in any node
distal to that of Ornithopoda, and even that is not
as secure as might be wished. Presumably it is a
genasaur, bui may represent a lineage that

650

——t

FIG. 12. Left (incomplete) pubis of M. langdoni (QM
F6140). Arrow indicates pubic notch. Ac=
acetabulum, PUR ta ata process, Prp=prepubic
process. Scale = 50m

diverged from the iguanodont-hadrosaur line
prior to the divergence of Tenontosaurus.

PALAEOBIOLOGICAL SPECULATIONS

The type skull and that from ‘Dunluce’ differ in
unexpected ways, specifically in maxillary tooth
form and, apparently, the structure of the nasal
bulla. There may be other differences, regarding
the form of the jugal-maxillary contact (Fig. 7)
and absence of replacement crowns in the maxi-
llary series, but the state of preservation and
preparation makes these uncertain. Therefore the
question naturally arises as to the significance of
the differences. This is also linked to the differen-
ces of the isolated maxillary teeth from Lightning
Ridge. The differences may be interpreted in
several fashions, as: stages in evolution; samples
of population variation; examples of sexual
dimorphism; stages of growth. I see no reason to
regard them as pathological variants.

The two skulls may simply represent variant
individuals, but if so the differences in the nasal
bulla imply considerable variation in this struc-
ture. Because of this, skulls discovered in the
future would be expected to also vary in structure
from those now known. Although admittedly not
calculable, the likelihood of finding by chance
two extreme variants in bullar, and dental, struc-
ture would seem small. The uniformity of maxi-
llary tooth form in the type skull argues against
the difference from the ‘Dunluce’ specimen being
due to variation in form along the toothrow.
Growth seems an inappropriate explanation for
differences between two skulls of almost equal
size. It may, nonetheless, be relevant to the teeth
from Lightning Ridge, which are smaller than
those in the Queensland skulls. However, those
teeth also differ in form from those of both skulls.

MEMOIRS OF THE QUEENSLAND MUSEUM

Changes in tooth form with ontogeny have
never been systematically described in ornithis-
chians, despite having good ontogenetic series of
ceratopians (Brown & Schlaikjer, 1940; Dodson,
1976) and hadrosaurs (Dodson, 1975). Some of
the recently discovered embryonic or hatchling
dinosaur material have no teeth suitable for com-
parison with mature teeth (Carpenter, 1994;
Chure et al., 1994) but some do. In the absence of
systematic studies, the differences will be sum-
marised here. Jacobs et al. (1994) figure a
juvenile ankylosaur tooth that seems not to differ
significantly from adult teeth. Since the specimen
was not referred to a known taxon appropriate
adult teeth for comparison are not available, so
this is a tentative conclusion. Carpenter (1982)
and Hatcher et al. (1907) reported that teeth of
juvenile ceratopians lack the bifurcated root of
adult teeth. Horner and Currie (1994) describe
and figure teeth of embryonic and neonatal
specimens of Hypacrosaurus stebingeri. Unless
the adult teeth, so far undescribed, are very different
from all other adult hadrosaur teeth the neonatal
teeth are more plesiomorphic in appearance
(Horner & Currie, 1994, figs 21.4A, 21.22, 21.23).
The crowns are broader, leaf-shaped, and in the
maxillary teeth have a more prominent primary
ridge and marginal denticles. Superficially they
have more resemble iguanodontian than
hadrosaurian teeth. The tooth from Lightning
Ridge does not show these kinds of differences
from those of the Queensland Muttaburrasaurus,
which suggests that it is not from a juvenile.

The nasal bulla is a large, prominent, cranial
structure. Such structures in hadrosaurs may well
have been associated with sexual selection and
sexual dimorphism (Molnar, 1977; Dodson, 1975),
so this is not unlikely in Muttaburrasaurus as
well. Even the difference in tooth form may be
related. Shine (1989) points out that females,
particularly during the reproductive season, may
have different nutritional requirements than males.
In some ducks, for example, this is reflected in the
lamellar form of the bills (Shine, 1989). Although
the specific differences in Muttaburrasaurus
food requirements obviously remain unknown —
as does its food in general — this is a possible
explanation for the difference in tooth form.

That the differences in form may represent
evolutionary changes is suggested by the dif-
ferences in age of the material. The type specimen
derives from the Mackunda Fm. of Late Albian
age, the ‘Dunluce’ specimen from the underlying
Allaru Mudstone also Late Albian, and the
Lightning Ridge teeth from the Griman Creek

OBSERVATIONS ON MUTTABURRASAURUS

Fm., from the Middle Albian (Exon & Senior,
1976; Burger, 1986). Thus the differences may
define a temporal sequence. This notion fits the
changes in form of the maxillary teeth, with a
progressive reduction in the prominence of the
primary ridge and its progressive shift to distal
edge of the lateral face. The Lightning Ridge
teeth suggest a third trend, toward an increased
number of secondary ridges. The presence of two
possible replacement teeth in the ‘Dunluce’ skull
and none in the type skull is consistent with a
trend toward the replacement of the maxillary
toothrow en masse. The adaptive — or otherwise
— nature of these changes cannot now be as-
sessed. However, the result of the trends provides
an almost unbroken strip of enamel along the
maxillary toothrow for the dentary dentition to
operate against. (This remains true even if there
is some individual replacement of the anterior
maxillary teeth.) This strip would presumably be
only minimally impaired during the en masse
replacement of the maxillary teeth. This is con-
sistent with an adaptive change in the toothrow to
give, or improve, a shearing function.

Thus of the five potential explanations for the
differences considered here two, pathology and
ontogentic development, are rejected. The
remaining three, variation, sexual dimorphism
and phylogenetic trends, remain viable alterna-
tives. If the evolutionary trends are real — and
they are based on only a few specimens — it
should be possible to find an ancestral form that
is consistent with them. As I (Molnar, 1991, 1992)
have argued that the Australian dinosaur fauna
was partially isolated from those elsewhere, such
a form should be available in Australia. The only
other Australian ornithopods are those from
Lightning Ridge (New South Wales) described by
Molnar & Galton (1986) and those from southern
Victoria described by Rich & Rich (1989). A
tooth from Lightning Ridge (QM F9505, a cast),
mentioned by Molnar (1984), shows no particular
resemblance to those of Muttaburrasaurus, nor
do the maxillary teeth of Leaellynasaura
amicagraphica (Rich & Rich, 1989, fig. 3A).
Those of Atlascopcosaurus loadsi (NMV
P157390, referred) from the Eumeralla Fm.
(deposited near the Aptian-Albian boundary) do
show a prominent primary ridge, arising from the
base of distal margin of the crown. A literature
survey of the teeth of other small or plesiomor-
phic ornithopods (Camptosaurus, Dryosaurus,
fabrosaurs, Hypsilophodon, Kangnasaurus,
Parksosaurus, Tenontosaurus, Thescelosaurus,
Yandusaurus, Zephyrosaurus) shows nothing

matching the maxillary teeth of Atlascop-
cosaurus, althougsome of those of Yandusaurus
hongheensis (He, 1979) show a primary ridge
displaced toward the distal margin.

The maxillary teeth of Atlascopcosaurus pro-
vide a possible plesiomorphic form of the teeth of
the Muttaburrasaurus lineage, and a link to the
form of teeth in overseas ornithopods, such as
Yandusaurus. This is consistent with the greater
age of Atlascopcosaurus, and provides some
evidence that the differences do represent evolu-
tionary trends rather than sexual dimorphism or
simple variation. Nonetheless the sample size of
specimens is so small that these other potential
explanations should not be completely ruled out.

CONCLUSIONS

Examination of the holotype (QM F6140) and
a second specimen (QM F14921) of Muttabur-
rasaurus indicates that it cannot be referred to the
Hypsilophodontia, Iguanodontia, Ankylopol-
lexia or more distal nodes of the iguanodontian
lineage as defined by Sereno (1986). It may rep-
resent a lineage that diverged from this line before
the divergence of Tenontosaurus.

Differences in bullar structure and dental form
suggest that QM F14921, from the Allaru
Mudstone, is slightly more primitive than the type
specimen of Muttaburrasaurus langdoni, from
the younger, Late Albian Mackunda Fm. Varia-
tion and sexual dimorphism are deemed possible,
but less likely, explanations. This material, together
with a tooth from the Griman Creek Fm. at Lightn-
ing Ridge, suggests that the teeth in this lineage
evolved to progressively reduce the size of the
primary ridge on the maxillary teeth, progressive-
ly displace it towards the posterior (distal) edge
of the lateral face and increase the number of
secondary ridges. This provided a continuous,
lightly corrugated strip of enamel along the lateral
side of the upper dentition. This condition is consis-
tent with the earlier (Bartholomai & Molnar, 1980)
suggestion of a shearing dentition in Muttaburra-
saurus. The maxillary teeth of Atlascopcosaurus
(NMV P157390) are close to the expected an-
cestral state of those of Muttaburrasaurus.

ACKNOWLEDGEMENTS

Robert Walker discovered the specimen at
"Dunluce' and he, Ninian Stewart-Moore and Mary
Wade organised its donation to the Queensland
Museum. New preparation of the type skull was
carried out by Joanne Wilkinson and Angela
Maree Hatch, and of the ‘Dunluce’ skull by Rudy

652

Kohout and Ms Hatch. Peter Arnold provided
welcome assistance during the study. Tom Rich
and Laune Bierne contributed helpful comments
and the latter provided Figs 9 and 10. The Lightn-
ing Ridge material was discovered by Lars and
Bea Tape, and they, together with Ms Elizabeth
Smith, arranged for its donation to the Australian
Museum. The photos are by Geoff Wright and
Bruce Cowell, QM photography section.

LITERATURE CITED

BARTHOLOMAI, A. & MOLNAR, R.E. 1981. Mut-
taburrasaurus, a new iguanodontid (Ornithischia:
Ornithopoda) dinosaur from the Lower
Cretaceous of Queensland. Memoirs of the
Queensland Museum 20: 319-349,

BROWN, B. & SCHLAKJER, E.M. 1940. The struc-
ture and relationships of Protoceratops. Annals of
the New York Academy of Sciences 60: 133-266.

BURGER, D. 1986. Palynology, cyclic sedimentation,
and palaeoenvironments in the Late Mesozoic of
the Eromanga Basin. Geological Society of
Australia, Special Publication 12: 53-70.

CARPENTER, K. 1982. Baby dinosaurs from the Late
Cretaceous Lance and Hell Creek formations and
a description of a new species of theropod. Con-
btributions to Geology, University of Wyoming,
20:123-134.

1994. Baby Dryosaurus from the Upper Jurassic
Morrison Formation of Dinosaur National Monu-
ment. Pp. 288-297. In, Carpenter, K., Hirsch, K.F.
& Horner, J.R. (eds) ‘Dinosaur Eggs and Babies’
(Cambridge University Press: Cambridge).

CHURE, D., TURNER, C. & PETERSON, F. 1994, An
embryo of Camptosaurus from the Morrison For-
mation (Jurassic, Middle Tithonian) in Dinosaur
National Monument, Utah. Pp. 298-311. In, Car-
penter, K., Hirsch, K.F. & Horner, J.R. (eds)
*Dinosaur Eggs and Babies' (Cambridge Univer-
sity Press: Cambridge).

DODSON, P. 1975. Taxonomic implications of relative
growth in lambeosaurine hadrosaurs. Systematic
Zoology 24: 37-54.

1976. Quantitative aspects of relative growth and
sexual dimorphism in Protoceratops. Journal of
Paleontology 50: 929-940.

1980. Comparative osteology of the American or-
nithopods Camptosaurus and Tenontosaurus.
Mémoirs de la Société géologique de France (ns)
59: 81-85.

EXON, N. F. & SENIOR, B.R. 1976. The Cretaceous
of the Eromanga and Surat Basins. BMR Journal
of Australian Geology & Geophysics 1: 33-50.

HATCHER, J.B., MARSH, O.C. & LULL, R.S. 1907.
The Ceratopsia. United States Geological Survey,
Monograph 49.

HE X. 1979. A newly discovered ornithopod dinosaur
— Yandusaurus from Zigong, Sichuan. Pp. 116-
123. In, *Contribution to International Exchange

MEMOIRS OF THE QUEENSLAND MUSEUM

of Geology, Pt. 2, Stratigraphy and Paleontology.'
(Sichuan Scientific and Technical Publishing
House: Chengdu). (In Chinese.)

HORNER, J.R. & CURRIE, P.J. 1994. Embryonic and
neonatal morphology and ontogeny of a new
species of Hypacrosaurus (Ornithischia, Lam-
beosauridae) from Montana and Alberta. Pp. 312-
336. In, K. Carpenter, K.F. Hirsch & J.R. Horner
(eds), ‘Dinosaur Eggs and Babies’ (Cambridge
University Press: Cambridge).

JACOBS, L.L., WINKLER, D. A., MURRY, P.A. &
MAURICE, J.M. 1994. A nodosaurid scuteling
from the Texas shore of the Western Interior
Seaway. Pp. 337-346. In, Carpenter, K., Hirsch,
K.F. & Horner, J.R. (eds) ‘Dinosaur Eggs and
Babies' (Cambridge University Press: Cambridge).

MOLNAR, R.E. 1977. Analogies in the evolution of
display structures in ornithopods and ungulates.
Evolutionary Theory 3: 165-190.

1984. Ornithischian dinosaurs in Australia. Pp. 151-
156. In, Reif, W.-E. & Westphal, F. (eds) ‘Third
Symposium on Mesozoic Terrestrial Ecosystems,
Short Papers.' (ATTEMPTO Verlag: Tübingen).

1991. Fossil reptiles in Australia. Pp. 605-702. In,
Rich, P. V., Monaghan, J. M., Baird, R. F., Rich,
T.H., Thompson, E. M. & Williams, C. (eds)
“Vertebrate Palaeontology in Australasia.’
(Pioneer Design Studio: Melbourne).

1992. Paleozoogeographic relationships of
Australian Mesozoic tetrapods. Pp. 259-266. In,
Chatterjee, S. & Hotton, N. III, (eds) ‘New Con-
cepts in Global Tectonics.’ (Texas Tech Univer-
sity Press: Lubbock).

1995, Possible convergence in the jaw mechanisms
of ceratopians and Muttaburrasaurus. Pp. 115-
117. In, Sun A. & Wang Y., (eds) 'Sixth Sym-
posium on Mesozoic Terrestrial Ecosystems and
Biota, Short Papers.' (China Ocean Press: Beijing).

MOLNAR, R.E. & GALTON, P.M. 1986, Hypsilopho-
dontid dinosaurs from Lightning Ridge, New
South Wales, Australia. Geobios 19; 231-239.

NORMAN, D.B. 1986. On the anatomy of Iguanodon
atherfieldensis (Ornithischia: Ornithopoda). In-
stitut Royal des Sciences Naturelles de Belgique,
Mémoires 56; 281-372.

NORMAN, D.B., & WEISHAMPEL, D.B. 1985. Or-
nithopod feeding mechanisms: their bearing on
the evolution of herbivory. The American
Naturalist 126: 151-164.

RICH, T.H.V. & RICH, P.V. 1989. Polar dinosaurs and
biotas of the Early Cretaceous of southeastern
Australia. National Geographic Research 5: 15-53.

SERENO, P.C. 1986. Phylogeny of the bird-hipped
dinosaur (Order Ornithischia). National
Geographic Research 2: 234-256.

SHINE, R. 1989, Ecological causes for the evolution of
sexual dimorphism: a review of the evidence. The
Quarterly Review of Biology 64: 419-461.

WEISHAMPEL, D.B., 1984. Evolution of jaw
mechanisms in ornithopod dinosaurs. Advances in
Anatomy, Embryology and Cell Biology 87: 1-110.

PRELIMINARY REPORT ON ANEW ANKYLOSAUR FROM THE EARLY
CRETACEOUS OF QUEENSLAND, AUSTRALIA

RALPH E. MOLNAR

Molnar, R.E. 1996 12 20: Preliminary report on a new ankylosaur from the Early Cretaceous
of Queensland, Australia. Memoirs of the Queensland Museum 39(3): 653-668. Brisbane.

ISSN 0079-8835.

The skeleton of a small ankylosaur, referred to the genus Minmi (Molnar 1980), was
recovered from the Albian Allaru Mudstone of north-central Queensland. The specimen
shows most of the contacts between the cranial bones and preserves most of the dorsal armor
in place. Unusual features include an apparently very large inferior process of the premaxilla,
thin ventral sheets of the nasals, an apparently rod-like, vertical lachrymal, and a pronounced
coronoid process on the mandible. The ilium is joined to two sacrals by a broad ‘bridge’ of
bone. The skeleton is thought to derive from a mature or almost mature individual, and may
have been mummified before burial. [ ] Early Cretaceous, ankylosaur, Australia, taphonomy.

Ralph E. Molnar, Queensland Museum, P.O. Box 3300, South Brisbane, Queensland,

Australia; 15 August 1996.

The most complete dinosaurian specimen yet
found in Australia was discovered by Mr Ian
levers in north-central Queensland in November
1989. While resting on an outcrop of limestone
concretions, he noticed fossil bone in them. He
then contacted the Queensland Museum and the
specimen was collected in January 1990. Mr
Ievers subsequently located further blocks, with
parts of the neck and shoulder girdle, which he
kindly forwarded to the museum.

This specimen represents the most complete
dinosaurian material yet found in eastern
Gondwanaland (Australia and Antarctica) and
one of the most complete Early Cretaceous
ankylosaurs. In addition it is one of the few
ankylosaurs to clearly show the contacts between
the cranial bones and the disposition of the dorsal
dermal armor, almost all of which seemingly
remains in place.

Because both preparation and study have yet to
be completed, this account will necessarily em-
phasise the features that have been studied, espe-
cially the dermal armor and cranial morphology.
More cursory comments on postcranial elements
are also included. A complete description,
together with an analysis of the phylogenetic
relationships and description of another recently-
discovered thyreophore skeleton (probably con-
specific), is contemplated for the near future.
Popular discussions of the specimen have been
published in Japan (Molnar, 1994) and Australia
(Molnar, 1991a, 1991b), and color photos of the
reassembled specimen have been published in
popular books (Lambert, 1993; Lambert & Bunt-
ing, 1995; Tomida & Sato, 1995).

Collection designations. AM, Australian
Museum, Sydney; QM, Queensland Museum,
Brisbane.

Abbreviations. b, sheet of bone linking scarum
and ilia; Bo, basioccipital; D, ossification of der-
mal armor; Dy, dentary; Eo, exoccipital (paroc-
cipital process); F, frontals; fm, foramen
magnum; J, jugal; L, (presumed) lachrymal; If,
left femur; lh, left humerus; li, left ilium; lu, left
antebrachium; m, unremoved matrix; Mx, maxi-
Ila; N, nasals; n, neural spines; o, orbit; P, parie-
tals; Pf, prefrontal; Pm, premaxilla; Po,
postorbital; Pof, postfrontal; Pt, pterygoid; Q,
quadrate; Qj, quadratojugal; R, nasal passage; ri,
right ilium; S, presumed sinus chamber; s, skull;
So, supraorbital; Soc, (presumed) suraoccipital;
Sq, squamosal; ‘Sq’, squamosal, postfrontal &
postorbital; sr, sacral ribs; stf, supratemporal
fenestra; t, transverse processes.

OCCURRENCE AND STRATIGRAPHY

The specimen (QM F18101) was found on a
low ridge in open country south of the Flinders
River on Marathon Station, east of Richmond,
north-central Queensland. The bones were par-
tially enclosed in silty buff-grey micrite concre-
tions. These concretions derive from weathering
of the Albian Allaru Mudstone, which consists of
mudstones and siltstones, often calcareous,
deposited in the inland sea.

The carbonate was deposited immediately
around the skeleton, with no large masses free
from bone. This suggests that the nodule may
originally (prior to its weathering in the soil) have
roughly preserved the body outline of the carcass.

654

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 1. Reassembled skeleton of the ankylosaur, Minmi sp. (QM F18101), from north-central Queensland. This
photo was taken prior to beginning preparation. Scale = 250mm.

The ribs, dorsal armor and tips of the neural
spines were exposed on the lower surfaces of the
blocks almost as though artificially prepared, but
the ilia were almost completely covered by car-
bonate. The remaining elements, especially the
distal limb pieces, were found either in small
blocks of carbonate or as isolated bones in the
soil.

The skeleton was intimately associated with
abundant plates of Inoceramus shell, a few
Echinorhinus teeth and part of one small, incom-
plete — and unfortunately indeterminate — am-
monite. The Allaru on ‘Marathon’ has also
yielded large teleosts, ichthyosaurs (Etheridge,
1888) and a nearly complete small sauro-
pterygian, as well as marine invertebrates, but no
other terrestrial forms. Nonmarine tetrapods from
the Allaru elsewhere in central Queensland in-
clude Muttaburrasaurus sp. (Molnar, this
volume) and a pterosaur. Six other specimens
represent armored dinosaurs, probably
ankylosaurs, all but one of which are fragmen-
tary: they will not be described here.

IDENTIFICATION

Only a single armored dinosaur, Minmi
paravertebra (Molnar, 1980), is known from
Austrtalia. This specimen matches the holotype
M. paravertebra in the possession of paraver-
tebrae, ossified aponeuroses and tendons along
the dorsal vertebral column. However the type of
M. paravertebra derives from the Aptian Bungil
Fm. whereas this specimen is from the Albian.
Since study of the new material is not yet com-

plete, and there may be differences of form in the
ribs, the question of its specific allocation
remains open. Furthermore the type material was
too incomplete for a phylogenetic analysis, al-
though the possession of paravertebrae is an
autapomorphy for the genus (cf. Molnar, 1980).
Hence the ‘Marathon’ specimen is herein desig-
nated Minmi sp. Pending the planned
phylogenetic analysis the genus Minmi is
presumed to belong to the Ankylosauria, but is
not allocated to a family.

Only five Cretaceous thyreophores are known
from the southern hemisphere outside Australia
(cf. Molnar & Frey, 1987). Ankylosaurian
material has recently been described from Upper
Cretaceous beds in Antarctica (Gasparini et al.,
1987; Olivero et al., 1991; Gasparini et al., this
volume), New Zealand (Molnar & Wiffen, 1994),
Argentina (Salgado & Coria, in press) and India
(Chatterjee & Rudra, this volume). A stegosaur,
Dravidosaurus blanfordi, has been reported from
the Coniacian of India (Yadagiri & Ayyasami,
1979), the only stegosaur described from the Late
Cretaceous. All of these are represented by sub-
stantially less complete specimens than that from
‘Marathon’. Isolated ankylosaur elements have
also been found in the Lower Cretaceous sedi-
ments of Victoria (Rich & V.- Rich, 1994).

DISPOSITION OF SKELETON

Since this is only a preliminary description of
the skeleton, the completeness and condition of
the specimen when collected are presented here
separately from the description. The specimen

A NEW ANKYLOSAUR FROM THE EARLY CRETACEOUS OF QUEENSLAND

f,
CO. ener
Ls
S3
Ih
lu

By
C Ee

655

Sy
bi T

-LIu:elllcloo =>

FIG. 2. The pattern of dorsal armor of Minmi sp. (QM F18101). Elements of the armor are shown in heavy outline,
and endoskeletal elements (skull roof, tips of neural spines & transverse processes, dorsal & sacral ribs and ilia)

are in light outline. Scale = 100mm.

includes most of the skull and the articulated axial
skeleton back to the proximal tail, to probably the
ninth caudal (Figs 1, 2). The arches of some
vertebrae have been crushed. The neural arch of
the axis was found isolated in the soil about a
metre away from the block enclosing the skull
and anterior cervicals. The more distal part of the
tail was not found and the distal parts of both left
limbs are also missing. The left shoulder girdle
seems complete. The coracoids are not fused to
the scapulae, and were displaced. The left
humerus is almost complete and the proximal
parts of both left radius and ulna are present. The
left ilium is nearly complete, and the proximal
halves of both ischium and pubis are also present
on that side. Much of the right pelvis, however,
has been lost. Both femora are essentially com-
plete, but weathered, as are the crural elements.
Only a few disarticulated manual and pedal ele-
ments were found, although the right manus may
be embedded beneath the rib cage.

The specimen was upside down when ex-
cavated, except perhaps for the block containing
the skull which had been uncovered before our
arrival. The neural canal of one vertebra
(enclosed in a small block of carbonate) contains
sediment (with embedded dermal ossicles) at the
top and white calcite crystals below clearly in-
dicating the inverted position of the trunk in the
ground.

When reassembled, the vertebral column is
straight, except for the cervical column near the
skull which inclines to the right at about 20° (Fig.
1). The base of the tail extends straight out
posteriorly from the trunk and the pieces of the
more distal part of the tail fit together to also give
a straight column. The ribs on the left side have
been flattened and so have lost their natural cur-
vature, but the ribs on the right — except for those
just in front of the pelvis — have been rotated
backwards. The right ilium has apparently be-
come separated from the sacrum and displaced
forward and laterally. Both left fore and hind
limbs extend laterally horizontally from the
trunk, but both right limbs are folded across the
belly. Both forelimbs are flexed at the elbow and
both hindlimbs at the knee. The left femur is in
situ in the acetabulum but projects horizontally,
parallel to the top of the ilium. The trunk gives
the impression of having collapsed to lay flat on
the substrate.

MATURITY

Recently there has been much emphasis (e.g.,
Rowe & Gauthier, 1990) on using adult charac-
ters (in the sense of skeletal maturity) in
phylogenetic classification and the unusual fea-
tures of this specimen raise the question of
whether or not it was mature, or nearly so. This
skeleton was probably 2.5-3.5m long when com-

656

plete. This is a small ankylosaur (cf. Coombs &
Maryanska, 1990): does it represent a small adult
or an immature individual? All known early
ankylosaurs (Dracopelta, Priodontognathus,
Sarcolestes and Tianchiasaurus) are small, as are
later, insular ankylosaurs such as Struthiosaurus.
Minmi is a relatively early ankylosaur and, in
addition, may have evolved on the east
Queensland island of the Aptian (Dettmann et
al., 1992). The holotype specimen of Minmi
paravertebra was found in Aptian marine sedi-
ments (Bungil Fm.) near this island (Molnar,
1980) and probably derived from there (Fig. 3).
If the lineage evolved there a small adult body
size is plausible. Late Cretaceous European in-
sular ankylosaurs (Struthiosaurus) are of about
the same size as the holotype and the ‘Marathon’
animal, and notably smaller than those from
western North America and central Asia.

Some features — absence of fusion of the
cranial elements, absence of fusion of the
scapulae and coracoids, absence of fusion of the
pelvic elements — in addition to its small size,
suggest that it was immature at death. But there
are also reasons to suggest that was mature or
nearly so. There are five other specimens from
Queensland certainly or plausibly referred to the
genus Minmi: QM F33565 and F33566 (probably
deriving from a single individual and so con-
sidered here), AM F35259, QM F33286, QM
F10329 (the holotype) and an unregistered
specimen that will probably be housed at the
Australian Museum. All of these, except the first
which is also the most different morphologically
and hence possibly represents a distinct taxon, are
within 10% of the linear dimensions of the
‘Marathon’ skeleton. Ignoring the first specimen
would suggest that all known specimens of
Queensland ankylosaurs are mature but small, but
including it indicates that five of six are imma-
ture, if the ‘Marathon’ skeleton is from an imma-
ture individual. It seems prima facie unlikely that
of six specimens that were, judging from their
geographical distribution, probably drawn from
different populations, all but one would represent
immature individuals at the same stage of growth.
Nonetheless this is not impossible, especially as
Horner (1979) pointed out that most occurrences
of dinosaurs in marine beds are of immature
individuals.

Although the junctions between the cranial
bones generally are not fused both frontals and
parietals are fused along the midline, and the
contacts in braincase are fused. Fusion of the
cranial elements is thought to be a marker of

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 3. Shorelines of Australia during the ages when
Minmi lived, A, Aptian, B, Albian. Heavy lines mark
contemporaneous shorelines, light lines modern
shorelines. The triangles indicate the localities of
Minmi paravertebra (A) and Minmi sp. (B). Both
specimens probably drifted from the eastern
Queensland landmass, which in the Aptian was a set
of islands and in the Albian a peninsula. (Redrawn
from Dettmann et al., 1992)

adulthood in ankylosaurs (Coombs &
Maryanska, 1990), but lack of fusion is also the
plesiomorphic state and cranial fusion may have
developed only in later ankylosaurs. The maxilla
of Priodontognathus, probably early Late Juras-
sic in age, shows no indication of having been
fused to the other cranial bones (Galton, 1980).
In the ‘Marathon’ specimen the coracoids are not
fused to the scapulae. This lack of fusion is not
considered a juvenile character by Coombs &
Maryanska (1990) and some ankylosaurs lack
fused scapulocoracoids as adults (Carpenter,
1990; Pereda-S., 1994).

Currie (pers. comm., 1991) observed that clear-
ly juvenile Pinacosaurus have only ossicles
making up the armor of the trunk. In the
‘Marathon’ Minmi the trunk armor comprises

A NEW ANKYLOSAUR FROM THE EARLY CRETACEOUS OF QUEENSLAND

657

FIG. 4. Skull, in dorsal view, of Minmi sp. (QM F18101). Anterior is to the left. Elements of the dermal armor
remain in the narial region and behind the quadrate. Scale = 50mm.

both ossicles and scutes, suggesting it was adult,
or nearly so.

Thus although there are some grounds (contra
Molnar, 1994) to think this specimen is a juvenile,
these are not entirely compelling and it may simp-
ly be plesiomorphic. Tentatively the ‘Marathon’
specimen is best regarded as representing an al-
most mature, or newly mature, individual.

DISPOSITION OF ARMOR

Armor is present across the dorsum of the neck
and trunk, as well as on the tail and limbs (Fig.
2). Several kinds of dermal elements are present:
large nuchal, scapular, pelvic, appendicular and
caudal scutes; small dorsal scutes; ossicles; and
(probably) triangular caudal plates.

The broadly oval nuchal scutes are larger (to
13cm wide x 9cm long) than any of the dorsal
scutes, and comparable in size to the pelvic scutes
and caudal plates. Four (three of which are
preserved) form an incomplete transverse band
just behind the skull. The pair on either side are
in contact (preserved on the left), but the medial
of each pair are separated by a gap of 5.5cm.
Behind this is a suite of one large (9.5 x 75cm)
three medium (c. 5.5 x 3cm) and two small scutes
(c. 2 x 1.5cm) arranged with the large one
medially placed and the others spread laterally
over the shoulder region. These shoulder scutes,
like those of the back, are roughly oval with a low
longitudinal keel and aligned parasagittally.
They are preserved only on the left side.

The smaller scutes of the back are arranged
linearly, at least towards the midline. A row of at

658

FIG. 5. Skull, in dorsal view, of Minmi sp. (QM
F18101). Anterior is to the left. Occipital face not
shown. The element on the left side of the nasals is a
large piece of dermal armor placed in what is thought
to be its proper position. Abbreviations given in text.
Scale = 50mm.

least ten lay to either side of the vertebral column,
extending from the base of the neck at least to the
pelvic region. Further laterally, the arrangement
is more haphazard, although this may arise from
post-mortem disturbance. In all cases the scutes
are placed between the ribs. The arrangement at
the shoulder may be an extension of that of the
back, but this is not clear. Unlike the shoulder
scutes, on the back the lateral scutes are slightly
larger than the medial.

Along the lateral side of the left ilium is a row
of three large flat plates (to 10.5 x 7cm). Basically
oval in form, they come to a sharp point posterior-
ly, and so look like tear-drops. Three of these are
preserved, arranged sequentially along the
posterior portion of the lateral margin of the ilium
and curving around behind it. All except the last
are displaced from the ilium, with but a few

A

MEMOIRS OF THE QUEENSLAND MUSEUM

medium or small scutes in the intervening region.
Unfortunately, because this part of the specimen
is almost completely flattened it is not clear
whether these plates lay flush with the skin, or if
the posterior points projected outwards above the
hind limb.

The neck, back and limbs were probably com-
pletely covered by small ossicles embedded in
the skin. Layers of ossicles are also found ventral
to the vertebrae and ribs, in what was the body
cavity. This suggests that like the holotype of M.
paravertebra, the belly was also covered by a
chain-mail of these ossicles. Some of the scutes,
particularly those of the shoulder region, have the
appearance of having been formed from — or at
least augmented by — the accretion of ossicles.

The ossicles are rectangular to trapezoidal in
outline and seem have two forms. One is convex
with faint concentric ridges or striae, and the
other an almost stellate form of sharp intercon-
necting ridges surrounding distinct pockets. Ex-
amination of ossicles freed during preparation,
however suggests that there is only a single form.
Presumably the ‘stellate’ face was directed exter-
nally and the convex, pillow-like face internally.
The appearance of both forms in the armor
preserved in place suggests that some have be-
come inverted without having been displaced.
Unbroken ossicles measure about 4 x 6mm.

The tail bore at least two kinds of armor. A row
of two or three large, smooth triangular plates are
preserved along one side of the tail. Their orien-
tation is unclear. As preserved, they project
ventrolaterally, but all are broken at their bases,
as if they had been forced downwards into this
position during preservation. In addition, there
are keeled, pitted, roughly rectangular scutes,
each slightly less than half as long as the trian-
gular plates. Again their arrangement is not

FIG. 6. Pattern of cranial elements in dorsal view for A, Minmi sp. B, Pinacosaurus grangeri; and C, Scelidosaurus
harrisonii. Not to scale (Pinacosaurus after Maryanska (1977), Scelidosaurus after Coombs, et al. (1990)).

ANEW ANKYLOSAUR FROM THE EARLY CRETACEOUS OF QUEENSLAND

659

FIG, 7. Skull, in lateral view, of Minmi sp. (QM F18101). Anterior is the left. Part of the integumentary dermal
armor remains intact behind the skull at right. Inset shows matrix (open dots), broken bone (dots) and still

adherent nuchal ossicles (hatching). Scale = 50mm.

known, as all seem to be out of place, some
having come to rest inserted between the bases of
the large plates and the caudal vertebrae. They
may have formed a row along the bottom of the
tail, two to every triangular plate.

The appendicular armor is not as completely
preserved, however both sets of limbs carried
large scutes, as well as small ossicles. The
forelimb had a subrounded, keeled scute midway
along the humerus and a low, round, pitted scute
at the lateral side of the elbow, along the ulna. The
hind limb bore a large keeled scute along the back
of the calf.

CRANIAL MORPHOLOGY

Although the form of the ankylosaurian skull as
a whole is well known, the form of the individual
cranial bones is not. Only a single skull (of
Pinacosaurus) is known showing the contacts
between the cranial bones, and those only for the
skull roof (Maryanska, 1977). For the rest of the
taxa only an isolated maxilla has been described
in Galton’s (1980) study of Priodontognathus.
The ‘Marathon’ specimen, however, shows all of
these contacts, except in the braincase where they
are fused.

The skull looks like a pentagonal box, because
all of the sides (and occipital face) meet the skull

roof at approximately right angles. It is slightly
longer (240mm as preserved) than wide
(195mm). The skull roof is almost flat, but the
nasals arch slightly (c. 3mm) above the level of
the skull roof. Unfortunately, the skull has been
broken and slightly crushed in places, fortunately,
little seems to be missing and the crushing has not
done much damage.

FIG. 8. Reconstruction of skull and left mandible, in
lateral view, of Minmi sp. (QM F18101). Anterior is
to the left. Hatching represents missing parts or broken
surfaces. Abbeviations given in text. Scale = 50mm.

660

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 9. Skull, in ventral view, of Minmi sp. (QM F18101). Anterior is to the left. Inset shows areas still covered
by matrix (open dots) and broken bone surfaces (dots). Scale = 50mm.

On the top of the skull both the grooves
presumably demarcating the boundaries of the
dermal plates and the sutures are clearly
preserved (Figs 4, 5). There are also a few dermal
ossicles. The skull roof elements are fewer and
simpler than in Pinacosaurus, basically similar in
pattern to those of Scelidosaurus (Coombs,
Weishampel & Witmer, 1990, fig. 20.1) (Fig. 6).
The nasals, prefrontals, frontals, parietals and
*squamosals' and supraorbitals contribute to the
dorsal face. The supratemporal fenestrae are en-
tirely obscured by the ‘squamosals’, parietals and
maybe the frontals.

The side of the skull is made up by the
premaxilla, maxilla, lachrymal, a little of the
supraorbital, and the *squamosal' and jugal (Figs

7, 8). The orbit is large, laterally directed, and
roughly hexagonal in shape. Just in front of it, the
sides of the skull converge towards the narial
region, forming the abbreviate snout. In the
nodosaurid Edmontonia the snout (from tip of
premaxilla to center of the orbit) comprises about
70% of the total cranial length, but in the
ankylosaurid Euoplocephalus, and the
‘Marathon’ skull (as preserved), it is about 60%.
The antorbital fenestra has been closed, possibly
by a backwards extension of the maxilla. Nor is
there any laterotemporal fenestra, which was
seemingly covered by the jugal and, maybe, the
'squamosal'. The palate is still largely un-
prepared, but there is so far no indication of the
median vomerine septum, and if it were present

A NEW ANKYLOSAUR FROM THE EARLY CRETACEOUS OF QUEENSLAND

FIG. 10. Diagram of skull, in anterior view, of Minmi
sp. (QM F18101) showing the the ventral sheets of
the nasals. Abbreviations given in text.

and in place, it would expected to have been
exposed by now.

Atransverse fracture, most obvious on the pala-
tal surface (Fig. 9), crosses the skull through the
front of the subtemporal fenestrae. On the left the
back of the maxilla and the overlying anterior
process of the jugal, have been displaced into the
orbital cavity: on the right the subtemporal
fenestra seems undamaged, but is occupied by a
piece of bone, perhaps part of the coronoid
process. Both pterygoids have been broken from
their contacts with the basipterygoid processes
and displaced dorsally, more so on the left side.
The postorbital region on the left may have been
slightly crushed and the right maxilla has been
broken and pushed into the palatal vacuity. The
left maxillary teeth have all either fallen from
their sockets, or been snapped off at the neck, so
that a few (four) crowns lie adjacent and lateral
to their respective alveoli. The snout is incom-
plete, but the inferior processes of the premaxillae
still associated with the maxillae, suggest that
only a little is actually missing, and this is cor-
roborated by the left mandible which is too short
to have fitted a much longer skull.

Premaxilla. The premaxilla is represented by a
large, smooth plate forming at least 5096 of the
length of the snout (Fig. 8). This plate seems to
be a greatly expanded inferior process of the
premaxilla. It overlaps the vertical plate of the
maxilla behind, and seemingly rests in a groove
in the dorsal margin of the anterior part of the
body of the maxilla. The plate is basically flat, but
is slightly convex posteriorly and slightly con-
cave dorsally. Its oblique anterodorsal edge
presumably forms the posterior margin of the
external naris.

Maxilla. The maxilla is a low, broad, elongate
element, that forms the ventral margin of the
snout and underlies the orbital cavity (Fig. 8). The

661

body is massive and broad anteriorly becoming
thinner but broader posteriorly. A vertical plate
rises behind and adjacent to the inferior process
of the premaxilla. This plate ascends to the dorsal
face of the skull to contact the prefrontal above.
In palatal view, the maxillary toothrow is curved,
slightly concave laterally. At least 21 alveoli are
present, increasing in size posteriorly. The
posterior alveoli are transversely broad, but con-
stricted centrally to look like the figure ‘8’. The
medial lobe of the alveolus was (and still is in
some) occupied by the root of the replacement
tooth, almost as large as that of the functional
tooth.

Nasal. The nasals are elongate elements, lon-
gitudinally arched (and incomplete) anteriorly
but concave transversely, forming a central lon-
gitudinal groove. Posteriorly they expand
markedly in the transverse plane, giving them a
roughly ‘T’ shape in dorsal view. They meet the
frontals in an almost transverse very slightly in-
terdigitating contact, and laterally they are
smoothly overlapped by the prefrontals. The
anterior moiety medially and posteriorly bounds
the external nares. Medial to the nares thin sheets
of bone project ventrally at least 35mm from the
margins of the nasals, giving the bone an M-
shaped form in section (Fig. 10). It is not known
how, or if, the passage situated between these
sheets contacted the laterally placed chambers.
Prefrontal. The prefrontals are elongate elements,
extending anteriorly from the posterior sutural
contacts with the frontal and ‘squamosal’ to the
edge of the skull in front of the orbits (Fig. 5).
Each joins the supraorbital posterolaterally in a
suture, and the vertical plate of the maxilla in
front of that in a smooth contact. They are ex-
posed only on the dorsum of the skull.

Lachrymal. The anterior margin of the orbit is
composed of a set of unfamiliar, rodlike elements
(Fig. 8). The presumed lachrymal is a vertical
rodlike element, that extends along this margin.
It gradually tapers ventrally, to abut on the max-
illa. In addition there are two other rod-like ele-
ments, or processes, extending ventrally along
the front edge of the orbit, one in anterior to and
the other behind the presumed lachrymal. The
anterior may be a descending process of the
prefrontal or supraorbital, but the possible junc-
tion is obscured.

Supraorbital. The subtriangular supraorbital
roofs the orbital cavity between the prefrontals
and 'squamosals' (Fig. 5). Beneath it posteriorly
(on the right side) is an elongate lightly orna-
mented element (Fig. 8), that appears to be part

662

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 11. Skull, in posterior view, of Minmi sp. (QM F18101). The upper part of the foramen magnum is occluded,
especially on the left, by matrix containing parts of the atlas-axis complex. Also on the left is part of the
integumentary dermal armor of the neck, just lateral to the distal end of the left paroccipital process. Scale =

25mm.

of the dermal armor, or perhaps a second supra-
orbital.

Frontal. The frontals are fused and basically
hexagonal in form, joining the nasals anteriorly,
the parietals posteriorly, the prefrontals antero-
laterally, and the 'squamosals' posterolaterally
(Fig. 5). The dorsal sulcus of the nasals continues
posteriorly across the frontals to terminate just
anterior to the parietal contact.

Jugal. Behind the maxilla and posteroventral to
the orbit is the large, flat, smooth, triangular jugal
(Fig. 8). It has the form of two plates, the vertical
triangular one behind forming the side of the
cheek region of the skull, and a horizontal one in
front that floors the orbital cavity. It overlaps the
quadratojugal posteriorly — which is thereby
obscured from lateral view — and in turn is
overlapped by a small, triangular dermal element
at its posteroventral extremity.

Squamosal, Postfrontal and Postorbital. The
bone here called the ‘squamosal’ is probably the
fused squamosal, postfrontal and postorbital. It is
a hexagonal bone, situated at the posterolateral
corner of the skull, that forms the upper part of
the cheek region (Figs 5, 8). It is overlapped by
the jugal below, and sutures to the supraorbital in
front and the frontal medially. Unlike the smooth
jugal, its lateral face is ornamented by subparallel
curved grooves directed posterodorsally, becom-
ing horizontal posteriorly.

Quadratojugal. The quadratojugal is medial to
the posterior part of the jugal and lateral to the
lower part of the quadrate. It contributed little if

anything to the lateral side of the skull, but is
visible along the ventral margin of the jugal arch.
In posterior aspect it is revealed as a thin element
rising vertically probably to contact the
*squamosal'. A large foramen separates it from
the quadrate just above their ventral contact.
Pterygoid. Only the broad, thin horizontal quad-
rate processes of the pterygoids are so far
revealed (Fig. 9). They project laterally from the
basisphenoid region where the junctions are
broken, to elongate contacts with the quadrates.
Quadrate. Both quadrates are present, but that on
the right is obscured. The articular condyles are
missing from both. The left is broken at midshaft,
but not displaced. The quadrate seems to form a
vertical pillar. The lateral inclination of almost
45? of the ventral portion is due to the crushing
of the back part of the skull on the left side. The
strong, medial vertical plate-like process projects
anteromedially, to reach the pterygoids.

FIG. 12. Skull in posterior view. Abbreviations given
in text. Open dots indicates matrix. Scale = 25mm.

A NEW ANKYLOSAUR FROM THE EARLY CRETACEOUS OF QUEENSLAND

FIG, 13. Mandible (dentary), in lateral view, of Minmi
sp. (QM F18101). Anterior is to the left. Only the
dorsal margin of the body and anterodorsal margin of
the coronoid process are preserved, the rest of the
edges are breaks. Scale = 25mm.

Parietal. The parietals are transversely elongate
elements, fused at the midline, and located along
the posterior margin of the skull (Figs 5, 12).
Their posterior face seems to extend ventrally for
3cm to contact the paroccipital processes, and
overlap the occipital faces of the “squamosals’.
Exoccipital. The paroccipital processes extend
directly laterally and flare at their distal ends, to
make up the ventral half of the occipital face of
the skull (Fig. 12). From each side a stout pedicle
projects ventromedially and a little posteriorly to
abut on the basioccipal below the foramen mag-
num. These exoccipital processes and the basioc-
cipital each contribute about 1/3 to the ventral
margin of the foramen magnum.

Supraoccipital. A large supraoccipital in the form
of a capital lambda, or inverted ‘V’, may be
present above the foramen magnum (Fig. 12) but
further cleaning is necessary to confirm this.
Basioccipital. The occipital condyle, on the
basioccipital, sits at the end of a stout, tapering
*neck' that is broader than the condyle itself. In
posterior aspect the condyle is reniform and it is
directed posteriorly (Fig. 12), rather than
posteroventrally.

Basisphenoid. Only the ventral surface of the
basisphenoid is yet exposed. Between the basip-
terygoid processes the ventral face of the basis-
phenoid is nearly flat, with a small median
foramen or depression (Fig. 9). Posteriorly the
basisphenoid tubera lie close together near the
midline. There is no excavation between and
behind them as in certain taxa (Struthiosaurus,
Pereda-S. & Galton, 1994), but only a smooth
surface, facing almost directly backwards, ex-
tending to the base of the neck of the occipital
condyle. The basioccipital-basisphenoid junction
is fused.

Cranial dermal armor. On each side of the skull,
roof, three small dermal ossicles sit at the ‘apex’
of the supraorbital-prefrontal-‘squamosal’ con-

663

tact, and a large one sits just lateral to them (Fig.
5). An elongate ossicle sits atop the posterior rim
of each ‘squamosal’. Dermal ossicles are also
found at several places on the sides of the skull,
along the posterior margin to the jugal, and the
postreior margin of the orbit. A large, elongate
ossicle just below the lateral margin of supraor-
bital, forms the dorsal margin of the orbit. A
smaller, triangular ossicle overlies the junction of
the lachrymal with the supraorbital. These two
may be the second and third supraorbitals, char-
acteristic of thyreophores (Sereno, 1986), al-
though if so they differ from those of the others
in that neither is exposed on the top of the skull.
A thin, flat, triangular ossicle overlying the
posteroventral ‘corner’ of the jugal (and quad-
ratojugal) is presumably the quadratojugal plate,
also found in ankylosaurids.

Dentary. The left mandible is very incomplete as
everything medial to the lateral sides of the al-
veoli has been lost. Only the dentary remains
(Fig. 13). But much of its length is preserved,
from the down-curved anterior region to well
back into the coronoid process. This strong
coronoid process arises abruptly lateral to the
back of the toothrow, and is clearly more exten-

FIG. 14. Three teeth of Minmi sp. (QM 18101) as
preserved after having fallen from their alveoli. A
partially prepared tooth is at the left, two completely
prepared teeth at center, and a block of matrix and
consolidant at right. Scale = 5mm.

664

sive than in any other described
thyreophore. The distance
from the declined anterior part
of the toothrow near the front
end of the mandible to the
coronoid suggests that the
mandible was relatively short.
Teeth. Two sets of isolated teeth
are available, one a crown from
the anterior part of the skull
(presumably the anterior maxi-
lla), and the other a set of three
adherent teeth (Fig. 14) from
the mid-maxillary (or den-
tary?) region. The anterior
crown has a cingulum 2-3mm
deep, with the fluted part of the
crown 2-3mm high. The cin-
gulum is broad and bulbous,
3mm wide, and the crown
above it only Imm wide. Seven
denticles, and the furrows be-
tween them, are present on one
side of the crown, whilst the
other is almost smooth, with
only the lateralmost furrows
present. The mid-maxillary
teeth are 18-26mm long. The
roots are straight and cylindri-
cal, constricted at the neck and
open, although perhaps broken,
at the tip. The best crown is
7mm high, including a 3mm
cingulum. One crown has
seven denticles and the remain-
ing two have nine. The enamel is covered with
very fine protuberances, giving a finely pebbled
appearance under a magnifier. The crowns are
6mm long and 3mm broad at their cingula and
2mm broad just above the cingula. The upper
margin of the cingulum is distinct on both
faces. Furrows extend all the way from edge to
cingulum, unlike the anterior tooth, where
they extend only a short distance from the den-
ticles.

VERTEBRAE

So far only the sacrals and middle caudals have
been prepared. The sacrals are partially obscured
by ossicles that presumably fell onto them from
the ventral body wall, so little can be said of them.
However it is clear that although the last centrum
of the presacral rod was ventrally flattened the

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 15. A, Left humerus, in anterior view, of Minmi sp. (OM F18101). Some
ossicles adhere to it distally, and the matrix visible along both medial and
lateral margins holds a sheet of ossicles in place on its reverse (cf. Fig. 1).
B, Proximal segments of left radius and ulna of Minmi sp. (QM F18101) in
medial view. The block of matrix joining both bones proximally holds
dermal armor in place on its reverse (cf. Fig. 1). Scale = 50mm.

synsacral centra lack a ventral sulcus, at least
posteriorly.

The more distal of these have a centrum that is
hexagonal in section, with a ventral sulcus. The
most distal preserved have a broad, flat dorsal
surface, with the neural arch restricted to the
middle third. Proximally the centra remain
hexagonal, but become deeper, and then become
quadrangular in section. The ventral sulcus be-
comes progressively shallower anteriorly, until
the ventral face is simply flat. The caudals are
slightly amphicoelous.

Although the prezygapophyses are relatively
long, they are not as elongate as in ankylosaurids.
There is no indication of a tail club, or that the
distal part of the tail was modified into a rigid rod.
All parts of the tail preserved suggest that some
flexibility was possible.

Several series of ossified tendons extend along
the tail, mostly above the transverse processes. So
they probably were part of the epaxial muscula-

A NEW ANKYLOSAUR FROM THE EARLY CRETACEOUS OF QUEENSLAND

FIG. 16. Reconstructed pelvis of Minmi sp.; A, in
dorsal view; and B, lateral. In A the dashed line
indicates the approximate margin of the left ilium, and
the posterior portion of the ilium has been completed
from a second specimen, QM F33286. The right side
of the reconstructed pelvis is a reflection of the left of
the specimen. The broad ‘bridge’(b) linking sacrum
and ilium may be seen. Scale = 200mm.

ture, although there are one or two that were in
the muscles below the transverse processes.

FORELIMB

Scapula. The left scapula is preserved just below
the dermal armor. Although preparation is not
complete, it shows to have a (mediolaterally) thin
blade c. 27cm long, and a well-developed
pseudacromial process 3.3cm high. (A complete,
but fractured, left coracoid is also present but not
yet studied.)

Humerus. The left humerus (26.5cm long) is
nearly complete with only the proximolateral
‘corner’ of the deltopectoral crest missing (Fig.
15A). The shaft is constricted near its midlength
and the ends enlarged, but not nearly as strongly
as in other thyreophores. The head is
anteroposteriorly broadened and set about a third

665

of the way out from the proximal trochanter. It is
supported behind by a buttress extending about a
third of the way down the shaft. There is also a
ridge along the back of the deltopectoral crest, so
forming a shallow fossa between this and the
buttress supporting the head. The deltopectoral
crest extends about a third of the way down the
shaft and lacks the distal expansion described in
ankylosaurids (Coombs & Maryanska, 1990).

The distal articular surface, at the elbow, sub-
tends an angle of almost 180? anteroposteriorly.
The elbow was probably a hinge joints as in most
tetrapods.

Antebrachial elements. The proximal end of the
ulna (Fig. 15B) is still incompletely prepared, but
it seemingly lacks an olecranal process, so
prominent in other thryeophores. Instead there is
a flattened, expanded, mushroom-like proximal
articular surface that makes an angle of about 45?
with the long axis of the shaft. The radial sulcus
is well-developed. The shaft is slender and dis-
tinctly flattened, with shallow grooves in the
broad faces giving an almost figure-8 form to the
cross-section.

Little is preserved of the radius (Fig. 15B),
although the shaft is slender and oval in cross-sec-
tion. Both radius and ulna are noticably more
slender than in other thyreophores.

HINDLIMB

Ilium. The ilium is well-preserved on the left side.
It is a flat, moderately broad element, with the
antacetabular processes inclined laterally at about
35° (Fig. 16A). It is unique among ankylosaurs in
that the postacetabular process is elongate. In
Euoplocephalus the postacetabular process
makes up about 27% of the iliac length, while in
Stegosaurus this is 33%. In the ‘Marathon’
specimen it is at least 38%. The ilium is connected
to the sacrum by a broad ‘bridge’ of bone. This
structure joins the medial margin of the ilium to
the dorsal portions of the neural spines of at least
three, maybe five, sacrals. These are probably in
the middle of the series, although precisely which
sacrals cannot be determined until further
preparation. The sheet-like 'bridge' seems to
overlie the sacral ribs and no trace of it can be
seen ventrally although the sacral ribs are
prominent. It is well preserved on the right side
in OM F18101, but indistinct on the left. It is also
present in QM F33286. To my knowledge noth-
ing like this has been reported in other
ankylosaurs, but it seems very similar to the sheet

666

FIG. 17. A, Pubis in lateral; and B, anterior view. Scale
= 25mm.

of bone that covers the sacral ribs in Stegosaurus
(Gilmore, 1914).

Pubis. The left pubis is nicely preserved, although
incomplete distally (Figs 16, 17). It has a block-
like body, of ‘standard’ form and a long slender
postpubis, that is mildly blade-like in form
proximally.

Ischium. The ischium (Figs 16, 18) looks much
like those of other ankylosaurs, broad subtrian-
gular proximally and rodlike distally.

Femur. The femur is a robust, straight element,
still incompletely prepared. The head is set off
medially, as in nodosaurids and the lesser
trochanter is not fused to the greater along its
length, but is set off by a shallow groove.
Crural & pedal elements. The tibia and fibula are
present, the latter at least in part, but have yet to
be studied, and no fibulae have been found. Two
isolated claw-like unguals resemble those of
Acanthopholis.

TAPHONOMY

The skeleton was found upside down in a
marine deposit. There being no reason to believe
this was an amphibious dinosaur, presumably the
body washed out to sea. The skeleton was com-
pletely articulated — except perhaps for the feet,
the distal end of the tail and the axis — and the
small dermal ossicles of the back are still largely
(or entirely) in place. This suggests that the skin
was present and intact on the back right up to the
time of burial. No evidence, such as tooth marks,
that the carcass was scavenged before it came to
lie on the sea floor has been seen, although a few

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG, 18. Lower face of left pelvic block of QM F18101
showing left(?) ischium (Isc) at center. Inset shows
matrix indicated by open dots, and blocks of ossciles
by dots. Scale = 50mm.

teeth of the small shark Echinorhinus were found
in association with the skeleton and may indicate
that a little scavenging occurred after it sank.
Since the skin of the back seems to have been
present, the skeleton may have been held together
by it, at least in part. If the skin remainined intact,
some of the ligaments and capsular tissue might
have done so, too. However, the preservation of
the left femur still in contact with the acetabulum
but directed horizontally outwards, suggests that
the ligaments and capsular tissue had relaxed or
decayed enough to permit what was presumably
a very un-life-like attitude. Furthermore, the dis-
covery of the neural arch of the axis some dis-
tance from the skull and neck suggests that this
element had become freed of its ligamentous (and
bony) attachments. The remaining cervical ver-
tebrae were all found associated or articulated
and enclosed within the calacreous nodules,
which in turn suggests that the axial neural arch
became freed before the nodule started to form.
Either the carcass was more or less complete
when it washed out to sea, or it had dried out and
mummified. As it was found some distance from
the reconstructed shorelines, it had presumably
floated for a while before sinking to the bottom.
If intact the carcass may have decayed, formed
gas bubbles (which would help it float) and even-
tually ruptured the belly wall, allowing it to sink.
Or it may have floated as a dry mummy until it
became waterlogged. The ventral sheet of os-

A NEW ANKYLOSAUR FROM THE EARLY CRETACEOUS OF QUEENSLAND

sicles seems to have been disrupted and fallen
into the body cavity, consistent with rupture by
decay gasses. But the dorsal sheet of ossicles has
‘sunk’ in between the ribs, which is consistent
with mummification, and difficult to explain if
that did not occur. Furthermore, the anomalous
position of the neural arch of the axis is consistent
with this explanation: the carcass had dried out to
the extent of allowing it to come free of its attach-
ment to soft tissues and the rest of the skeleton.
While the carcass was sinking the arch fell away
to come to rest on the seafloor some distance from
the remainder of the skeleton. This is not, how-
ever, the only possible explanation, it may have
been carried off and then dropped by one of the
small scavenging sharks.

The evidence seems in favor of the interpreta-
tion of mummification prior to being swept out to
sea by one point, the ‘sinking in’ of the dorsal
armor and hence the dorsal hide between the ribs.
Perhaps this is also consistent with the interpreta-
tion of having bloated and burst, but it isn’t ob-
vious how. Further work on this aspect is planned
as, if true, the mummification interpretation has
some interesting implications. First the carcass
would have had to have been exposed long
enough to dry out, without being noticibly
scavenged, and then washed into a river or the
sea. This implies an extended, perhaps seasonal,
dry period followed by rains heavy enough to
remove the carcass, and a surprising lack of
scavengers. The seasonality is not inconsistent
with what is known about Albian palaeoclimates
in Queensland (Dettmann et al., 1992), but per-
haps suggests that the seasons were more marked
than generally thought.

ACKNOWLEDGEMENTS

The importance of the specimen was recog-
nised by Mr Ian levers, who kindly donated it to
the Queensland Musuem. Ken Carpenter, Xabier
Pereda-Suberbiola, W.P. Coombs Jr, Tatiana
Tumanova, Laurie Bierne and Peter Trusler
provided much helpful discussion and comment.
The specimen was skillfully prepared by Angela
Maree Hatch and Joanne Wilkinson. I am much
indebted to all of these for their assistance during
the study of this specimen.

LITERATURE CITED

CARPENTER, K. 1990. Ankylosaur systematics: ex-
ample using Panoplosaurus and Edmontonia
(Ankylosauria: Nodosauridac). Pp. 281-298. In,
Carpenter, K. & Currie, P.J. (eds) ‘Dinosaur

667

Systematics’, (Cambridge University Press:
Cambridge).

CHATTERJEE, S. & RUDRA, D.K. 1996. KT events
in India: impact, rifting, volcanism and dinosaur
exticntion. Memoirs of the Queensland Museum,
this volume 489-532.

COOMBS, W.P. Jr & MARYANSKA, T. 1990.
Ankylosauria. Pp. 456-483. In, Weishampel, D.B.,
Dodson P. & Osmólska, H. (eds) ‘The
Dinosauria'. (University of California Press:
Berkeley).

COOMBS, W.P. Jr, VEISHAMPEL, D.B. & WITMER,
L.M. 1990. Basal Thyreophora. Pp. 427-434. In,
Weishampel, D.B., Dodson P. & Osmólska, H.
(eds) ‘The Dinosauria'. (University of California
Press: Berkeley).

DETTMANN, M.E., MOLNAR, R.E., DOUGLAS,
J.G., BURGER, D., FIELDING, C., CLIFFORD,
H.T., FRANCIS, J., JELL, P., RICH, T., WADE,
M., RICH, P.V., PLEDGE, N., KEMP, A. &
ROZEFELDS, A. 1992. Australian Cretaceous
terrestrial faunas and floras: biostratigraphic and
biogeographic implications. Cretaceous Research
13: 207-262.

ETHERIDGE, R. Jr 1888. On additional evidence of the
genus Ichthyosaurus in the Mesozoic rocks
(‘Rolling Downs Formation’) of northeastern
Australia. Proceedings of the Linnean Society of
New South Wales 2: 405-409.

GALTON, P.M. 1980. Priodontognathus phillipsii
(Seeley), an ankylosaurian dinosaur from the
Upper Jurassic (or possibly Lower Cretaceous) of
England. Neues Jahrbuch für Geologie und
Paláontologie, Monatshefte 1980: 477-489.

GASPARINI, Z., OLIVERO, E., SCASSO, R. &
RINALDI, C. 1987. Un ankylosaurio (Reptilia,
Omithischia) campaniano en el continente antar-
tico. Anais IV Congreso Brasileiro de Paleon-
tologia, Rio de Janeiro 1: 131-141.

GASPARINI, Z., PEREDA-SUBERBIOLA, X. &
MOLNAR, R.E. 1996. New data on the
ankylosaurian dinosaur from the Late Cretaceous
of the Antarctic Peninsula. This volume.

GILMORE, C.W. 1914, Osteology of the armored
Dinosauria in the United States National Museum,
with special reference to the genus Stegosaurus.
United States National Museum, Bulletin 89: 1-
136.

HORNER, J.R., 1979. Upper Cretaceous dinosaurs
from the Bearpaw Shale (marine) of south-central
Montana with a checklist of Upper Cretaceous
dinosaur remains from marine sediments in North
America. Journal of Paleontology 53: 566-577.

LAMBERT, D. 1993. P. 127. "The Ultimate Dinosaur
Book’. (Dorling Kindersley Ltd.: London).

LAMBERT, D. & BUNTING, E. 1995. P. 43. ‘The
Visual Dictionary of Prehistoric Life’. (Dorling
Kindersley Ltd: London).

MARYANSKA, T. 1977, Ankylosauridae (Dinosauria)
from Mongolia. Palaeontologia Polonica 37: 85-
151.

668

MOLNAR, R.E. 1980. An ankylosaur (Ornithischia:
Reptilia) from the lower Cretaceous of southern
Queensland. Memoirs of the Queensland Museum
20: 77-87.

1991a. Minmi paravertebra. The Fossil Collector
(Melbourne) 34: 5-10.

1991b. The case of the mysterious dinosaurs of the
outback. Wildlife Australia (Brisbane) 28:26-27.

1994. A dragon of the south. Kyoryugaku Saizensen
8: 102-111. (in Japanese)

1996, this volume. Observations of the Australian
omithopod dinosaur Muttaburrasaurus.Memoirs
of the Queensland Museum, this volume: 39 (3):
639-652.

MOLNAR, R.E. & FREY, E. 1987. The paravertebral
elements of the Australian ankylosaur Minmi
(Reptilia: Ornithischia, Cretaceous). Neues
Jahrbuch für Geologie und Paläontologie, Ab-
handlungen 175: 19-37.

MOLNAR, R.E. & WIFFEN, J. 1994. A late Cretaceous
polar dinosaur fauna from New Zealand.
Cretaceous Research 15: 689-706.

OLIVERO, E.B., GASPARINI, Z., RINALDI, C.A.,
SCASSO, R. 1991. First record of dinosaurs in
Antarctica (Upper Cretaceous, James Ross Is-
land): palaeogeographical implications. Pp. 617-
622. In, Thomson, M.R.A., Crame, J.A. &
Thomson, J.W. (eds) ‘Geological Evolution of
Antarctica’. (Cambridge University Press:
Cambridge).

PEREDA-SUBERBIOLA, J. 1994. Polacanthus (Or-
nithischia, Ankylosauria), a Transatlantic ar-

MEMOIRS OF THE QUEENSLAND MUSEUM

moured dinosaur from the Early Cretaceous of
Europe and North America. Palaeontographica A
232: 133-159.

PEREDA-SUBERBIOLA, J. & GALTON, P.M. 1994.
A revision of the cranial features of the dinosaur
Struthiosaurus austriacus Bunzel (Ornithischia:
Ankylosauria) from the Late Cretaceous of
Europe. Neues Jahrbuch fiir Geologie und
Paläontologie, Abhandlungen 191: 173-200.

RICH, T.H.V. & VICKERS-RICH, P. 1994.
Neoceratopsians and ornithomimosaurs:
dinosaurs of Gondwana origin? Research & Ex-
ploration 5: 129-131.

ROWE, T., & GAUTHIER, J., 1990. Ceratosauria. Pp.
151-168. In, Weishampel, D.B., Dodson P. &
Osmólska, H. (eds) ‘The Dinosauria’. (University
of California Press: Berkeley).

SALGADO, L. & CORIA, R.A. in press. First evidence
of an armoured ornithischian dinosaur in the Late
Cretaceous of north Patagonia, Argentina.
Ameghiniana.

SERENO, P.C. 1986. Phylogeny of the bird-hipped
dinosaurs (order Ornithischia). National
Geographic Research, 2:234-256.

TOMIDA Y. & SATO T. 1995. P. 13. ‘Stegosaurus’.
(Nature Editors, Kaisei-sha Publishing Co.) (in
Japanese)

YADAGIRI, P. & AYYASAMI, K. 1979. A new
stegosaurian dinosaur from Upper Cretaceous
sediments of south India. Journal of the Geologi-
cal Society of India 20: 521-530.

AN ANTARCTIC CRETACEOUS THEROPOD
R.E. MOLNAR, ALEJANDRO LOPEZ ANGRIMAN & ZULMA GASPARINI

Molnar, R.E., López Angriman, A. & Gasparini, G. 1996 12 20: An Antarctic Cretaceous
theropod. Memoirs of the Queensland Museum 39(3): 669-674. Brisbane. ISSN 0079-8835.

The distal part of a theropod tibia has been recovered fromthe Coniacian-Santonian Hidden
Lake Fm. near Cape Lachmann, James Ross Island. The piece closely resembles the
corresponding region in Megalosaurus and, more closely, in Piatnitzkysaurus. This suggests
that a persistently plesiomorphic tetanuran lineage inhabited the Antarctic. The relatively
small size of the animal argues against a mean annual temperature below 15°C in its
environment. [ ] Antarctica, theropod, Late Cretaceous, biogeography, tetanuran.

R.E. Molnar, Queensland Museum, P.O. Box 3300, South Brisbane, Queensland, 4101,
Australia; Alejandro López Angríman, BRIDAS, Av. L. N. Alem 1180, (1001) Buenos Aires,
Argentina; Zulma Gasparini, Depto. Paleontología de Vertebrados, Museo de La Plata,

(1900) La Plata, Argentina; 5 July 1996.

During the summer of 1988 one of the authors
(A.L.A.) discovered a small bone in the
Cretaceous deposits 4km north of Col Crame, in
the Cape Lachman region, northwestern James
Ross Island (Fig. 1). Ina basin in the northwestern
sector of the Weddell Sea, the origin of the island
is related to the opening of that sea during the
breakup of Gondwanaland (Medina et al., 1992).
The specimen derives from the middle section of
the Hidden Lake Fm., Gustav Gr. (Ineson et al.,
1986), referred to the Coniacian-Santonian
(Buatois & López Angriman, 1992a). It is the
first Cretaceous Antarctic theropod and the oldest
Cretaceous tetrapod from Antarcica.

Prior to Angríman's discovery, ornithischians
were the only Cretaceous dinosaurs known from
Antarctica (Gasparini et al., 1987; Olivero et al.,
1991; Hooker et al., 1991). The Late Cretaceous
theropods from New Zealand (Molnar & Wiffen,
1994) imply that theropods also lived in An-
tarctica at that time (cf. Molnar, 1989) but this
discovery verifies their occurrence and helps fill
out our knowledge of Antarctic faunal evolution.
A theropod is known from the Early Jurassic of
Antarctica (Hammer & Hickerson, 1994).
Collection designations. ANSP, Academy of
Natural Sciences, Philadelphia; MLP, Depar-
tamento de Paleontología Vertebrados, Museo de
La Plata, La Plata.

DESCRIPTION

The specimen (MLP: 89-XII-1-1) is the distal
end — representing probably 10-15% of the total
length — of a left tibia. The piece is complete
save for very small portions of the extremities of
the medial malleolus and the flange that backs the
fibula. In general form it is (Fig. 2) is similar to

the corresponding part of the tibia in
Megalosaurus bucklandi, Piatnitzkysaurus
floresi and Poekilopleuron bucklandii. Seen from
the front, the shaft is expanded laterally to form
a broadly rounded post-fibular flange, and
medially into an angularly truncate medial mal-
leolus. A broad, flat-surfaced prominence
proximomedially bounds the ascending process
and dorsal face of the astragalus, extending
proximolaterally across the anterior face from the
medial malleolus. Its edges are marked by sharp
angulations. This is termed the medial buttress.
The posterior face is almost flat but with a slight
longitudinal concavity laterally. Above the post-
fibular flange the surface of this concavity
abruptly slopes anteriorly to meet the anterior
surface of the shaft at a sharp edge.

The broken end reveals a central cavity sur-
rounded by relatively thin-walled bone (Fig. 2D).
The shaft is tear-shaped in section at the break.
Fragments of bone in the calcite fill of the central
cavity suggest that the shaft was crushed just
above the broken end before or shortly after burial.
The distal end is slightly concave from the front
and triangular when viewed distally, with the
anterior surface forming the longest edge and the
inclined lateral edge the shortest. The distal sur-
face is shallowly concave in the middle, becom-
ing mildly convex both medially and laterally.

The form of the anterior surface of the distal
tibia (Fig. 2C) indicates that the ascending
process of the astragalus was moderately low,
relatively narrow and restricted to the lateral half
of the tibial shaft. Thus the astragalus would have
been unlike those of ornithomimosaurs or tyran-
nosaurs in form, but probably much like that of
Poekilopleuron bucklandii.

670

ANTARCTIC

© C. LACHMAN
PENINSULA od

« Me
(evano l.
58°W he

FIG. 1. Map of James Ross Island, showing where the
theropod tibia was found (dot). C=Cape, I=Island.

IDENTIFICATION. The general form and rela-
tively thin-walled hollow shaft indicate that this
tibia derives from a theropod. Dryosaurs have
also been reported with thin-walled limb elements
(Chinsamy, 1995) and dryosaurs or similar forms
are known from Late Cretaceous New Zealand
(Wiffen & Molnar, 1989), hence probably in-
habited Antarctica at this time. However no other
dinosaurs have the astragalar ascending process
set into a depression of the anterior face of the
distal tibia, dorsomedially bounded by an abrupt
step or offset in the anterior face. This feature is
characteristic of neotheropods (Sereno et al.,
1994), and isn’t found in herrerasaurs (Fig. 3).
The truncate medial malleolus also occurs in a
restricted group of theropods (Table 1), but this
probably represents a stage in the evolution of
distal tibial form (Fig. 3).

The work on theropod astragali of Welles &
Long, 1974, (with the correction of Carpenter,
1992) accords well with recent phylogenetic
analyses of theropods (Holtz, 1994; Russell &
Dong, 1993; Sereno et al., 1994). A general ten-
dency toward increased height and breadth of the
ascending process can be seen in the lineage
leading to arctometatarsalians (Fig. 3). Further-
more, a survey of figured tibiae indicates that
distal tibial form is reasonably distinctive for the
recognised theropod groups (Table 1).

The distal tibia of ceratosaurs is variable be-
cause of the evolution of the ankle in this group.
In distal view the tibia is not anteroposteriorly com-
pressed in the Triassic and Early Jurassic forms,
as it is in later ceratosaurs and tetanurans. The
ascending process is absent or low and narrow,
the medial buttress is absent in Coelophysis and
Syntarsus but present in Dilophosaurus, and the
medial malleolus is quite subdued or absent in

MEMOIRS OF THE QUEENSLAND MUSEUM

Coelophysis and Dilophosaurus but present and
angular in Syntarsus (Raath, 1969; Welles, 1984;
Colbert, 1989). Early ceratosaurs, e.g.,
Coelophysis, (and herrerasaurs) lack an ascend-
ing process and the astragalus interlocks with an
offset distal surface of the tibia, the offset being
visible anteriorly (Padian, 1986, fig. 5.5).

The neoceratosaur distal tibia is known only in
Ceratosaurus and Xenotarsosaurus, where the
distal expansion is approximately symmetrical
and the fossa for the ascending process low. In
Ceratosaurus the medial buttress is broad with
the ‘step’ extending nearly horizontally at least
2/3 of the way across the shaft (about halfway in
the Hidden Lake specimen) and the medial mal-
leolus does not extend as far proximally along the
shaft, so is more pointed in anterior view (Gil-
more, 1920); the distal tibia of Xenotarsosaurus is
similar but with a broadly rounded malleolus
(Martinez et al., 1987). Dromaeosaurs (i.e.,
Deinonychus antirrhopus) have a broad, high,
ascending process, lack the medial buttress —
probably correlated with the broadening of the
ascending process — and have a broadly rounded
medial malleolus (Ostrom, 1969). Arctometatar-
salian theropods also have a broad, high ascend-
ing process and lack the medial buttress, and have
a nearly symmetrical] distal end (in anterior or
posterior view) with no distinctive form to the
medial malleolus (Welles & Long, 1974). Other
coelurosaurs (e.g., Calamosaurus) similarly have
a high, broad ascending process and no medial
buttress but the medial malleolus is similar in
silhouette, from in front, to that of the Hidden
Lake tibia (Lydekker, 1891). There is little data
available on oviraptorosaurs, but judging from
Chirostenotes and Microvenator, the form is
similar to that in arctometatarsalians, with the
exception that the fibular flange seems truncate
in the former genus (Ostrom, 1970; Currie &
Russell, 1988). The segnosaur distal tibia remains
unknown or undescribed. That of Yangchuano-
saurus shows a low ascending process, a medial
buttress different in form and an angular medial
malleolus (Dong et al., 1983). In Sinraptor the
ascending process is low and narrow, the medial
buttress essentially similar to that of
Ceratosaurus, and with an angular medial mal-
leolus (Currie & Zhao, 1993). In distal view the
tibia is more anteroposteriorly compressed. Al-
losaurus has a broader, higher ascending process,
a narrower medial buttress, and a generally
similar medial malleolus that is, however, more
medially projecting and more rounded in outline
(Gilmore, 1920; Madsen, 1976).

AN ANTARCTIC CRETACEOUS THEROPOD

671

FIG. 2. The theropod distal tibia (MLP: 89-XII-1-1) from the Hidden Lake Fm., James Ross Island, in: A,
posterior; B, medial; C, anterior; D, proximal; E, distal; and F, lateral views. mb=medial buttress, mm-medial

malleolus, pf=flange behind fibula, scale = lcm.

The distal tibiae of Megalosaurus bucklandi
(Huxley, 1870; Hulke, 1879) and Poekilopleuron
bucklandii (Eudes-Deslongchamps, 1837) —
possibly related forms — Erectopus superbus
(Sauvage, 1882) — of uncertain affinities — and
Piatnitzkysaurus floresi (Bonaparte, 1986) — a
plesiomorphic allosauroid — are the most
similar. However, the tibiae of torvosaurids
(specifically Afrovenator abakensis, Eustrepto-
spondylus oxoniensis and Torvosaurus tanneri),
animals similar in other ways to megalosaurs,
differ in having an angular medial malleolus (e.g.,
Britt, 1991). Of those with similar distal tibiae,
the most similar is that of P. floresi, which
matches that of the Hidden Lake theropod in the
forms of the medial malleolus and buttress —
particularly the curve of the step adjacent to the
facet for the astragalar ascending process — and
the fibular flange. The distal tibiae of the pre-
viously noted megalosaurs, and of Erectopus,
differ in that the fibular flange projects distally,
forming the distalmost part of the tibia. This does

not occur in Piatnitzkysaurus or the Hidden Lake
tibia. There are slight differences: the Hidden Lake
tibia is less compressed distally, its fibular flange
projects less and its medial buttress is slightly
broader than in P. floresi. Nonetheless, had the
Hidden Lake tibia been found instead in the
Cañadon Asfalto Fm. (Callovian-Oxfordian) of
Argentina, it would likely have been attributed to
an immature Piatnitzkysaurus.

The similarity of the Hidden Lake tibia to those
of megalosaurs and Piatnitzkysaurus, all Middle
Jurassic forms, and Erectopus, an Albian form,
implies that the Hidden Lake theropod represents
a lineage probably derived from relatively
plesiomorphic megalosaur or primitive al-
losauroid (but not sinraptorid) stock, unrelated to
the lineages culminating in the more or less con-
temporaneous arctometatarsalians. Interestingly,
the only other known Antarctic theropod, the
Early Jurassic Cryolophosaurus ellioti (Hammer
& Hickerson, 1994), is also now thought to be a
plesiomorphic allosauroid (Hammer, pers.

672

comm., 1996). Generic identifica-
tion of the Hidden Lake specimen
is not possible, unless more is
recovered.

DISCUSSION

The Hidden Lake tibia is the
same size as the distal end of the
tibia of the ornithomimosaur
Coelosaurus antiquus (ANSP
9222) which is about 40cm long
overall. If the proportions were the
same, then the Hidden Lake
theropod would have been about
3-3.5m long. If, on the other hand
it was more robust, with the tibia
proportioned like those of Piat-
nitzkysaurus floresi, then it would
have been about 22-23cm long and
the whole animal about 2.5-3m
long. Assuming that it is from a
mature animal, and we have no
reason to think otherwise, in either
case it represents a moderately
small theropod about the size of a
large Coelophysis. This theropod
would have been too small to have
been ectothermic and to have lived
under a climate with a mean annual
temperature of less than 15?C (cf.

THEROPODA-
Medial
buttress

Symmetrical

~
NEOTHEROPODA

Anteroposterior!
compressed distal tibia

N

TETANURAE ~

Truncate medial malleolus ~

Ascending process broad & high

development of malleolus & fibular flange ^ ——
(loss of truncate medial malleolus) /

MEMOIRS OF THE QUEENSLAND MUSEUM

Herrerasauridae —

l-

Ccelophyside D
Dilophosaurus N
Elaphrosaurus ) \

CERATOSAURIA

Xenotarsosaurus /
\ | C3
Ceratosaurus L\

Torvosaurus L\

Jd Megalosaurus b) E» p \

Truncate
medial malleolus Piatnitzkysaurus 7 JN
Sinraptor A HA

Allosaurus AN

Acrocanthosaurus JI

Deinonychus IK

Oviraptoridae

Troodontidze LA

CARNOSAURIA
—

AVETHEROPODA

idæ \
COELUROSAURIA Tyrannosauridae / E

Loss of medial buttress ?

=
Ornithomimosauria l \
(sc)

Molnar & Wiffen, 1994; Spotila et FIG. 3. Phylogenetic relationships among theropods, with astragalar and

al., 1991). This suggests that James
Ross Island enjoyed a rather mild
climate, at least in some places,
during the Coniacian-Santonian.
The specimen was found in the
middle section of the Hidden Lake
Fm., in the level of the calcareous
sand with abundant carbonised
material. Associated trace fossils
include Planolites sp. and
Palaeophycus sp., and remains of logs without
the borings of Teredolites that are often found in
the upper and lower levels of this formation
(Buatois & López Angriman, 1992b). The en-
vironment of deposition was a developing dis-
tributary plain in the central fan of a fan delta
depositional system (Buatois & López Angriman,
19922) into which, we infer, the specimen had
been transported after death.

This tibia suggests a more primitive theropod,
or at least one with less advanced distal tibial
structure, than was common in the Late
Cretaceous Asiamerica or South America. Thus it

distal tibial characters discussed in text assigned to nodes. A and B
indicate presumed reversals of character state: A, to uncompressed
distal tibia; B, to angular (rather than truncate) medial malleolus. Along
the right margin are diagrams of the distal tibiae of the genera included,
with that for Poekilopleuron bucklandii (P.b.) added next to that for
Megalosaurus, and that for the Hidden Lake theropod (HLt) next to
Piatnitzkysaurus. Cladogram after Holtz (1994), with the positions of
Sinraptor, following Sereno, et al. (1994), and Piatnitzkysaurus, fol-
lowing Molnar, et al. (1990), added. (Tibial diagrams redrawn from the
literature, except for the Hidden Lake theropod and Piatnitzkysaurus).

suggests that lineages deriving from Middle
Jurassic forms persisted in the Antarctic.

ACKNOWLEDGEMENTS

The specimen was collected with the kind
logistical assistance of the Instituto Antártico Ar-
gentino, and the participation of Drs Christian
Cherniglasov, Roberto Llorente and especially
Luis Buatois. We also appreciate the help of Fer-
nando Novas (Museo Nacional de Ciencias
Naturales, Buenos Aires), Jaime Powell (Instituto
Miguel Lillo, Tucumán), Alec Ritchie (Australian
Museum, Sydney) and H. Phillip Powell (Oxford

AN ANTARCTIC CRETACEOUS THEROPOD

673

TABLE 1. Theropod distal tibial form. In the last column, ‘lateral’ indicates that the lateral flange is more strongly
developed than the medial malleolus; ‘symmetric’ that flange and malleolus are about equally developed and
symmetrical in anterior view; ‘both’ that flange and malleolus are equally developed but not symmetrical .

Medial malleolar
Taxa
development
Herrerasauridae slight

Coelophysidae
Dilophosaurus

absent

Elaphrosaurus

Medial buttress Astragalar ascending process
width height

Distal tibial
development

moderate symmetric

Xenotarsosaurus |

Ceratosaurus

Torvosaurus present

both

lateral

moderate

moderate low

Megalosaurus truncate present

truncate present

moderate low | lateral

2 ? l lateral

Piatnitzkysaurus

Sinraptor angular present moderate LL low lateral
Allosaurus truncate present moderate low lateral
Acrocanthosaurus truncate present ? ?

Deinonychus rounded absent broad symmetric
Oviraptoridae slight low or absent broad | both — |
Avimimus i | absent
| Tyrannosauridae rounded | absent broad
Troodontidae i absent symmetric

Ornithomimidae

Janensch (1925

University Museum, Oxford) who made avail-
able theropod material in their care, and Angela
Milner (British Museum) for hypsilophodontian
material. Don Baird (Princeton University)
provided a cast of the tibia of Coelosaurus which
washelpful and Jim Farlow (University of
Indiana/Purdue University), William Hammer
(Augustana College) and Tom Holtz, Jr (Univer-
sity of Maryland) all provided helpful comments.
We much appreciate all their assistance.

LITERATURE CITED

BONAPARTE, J.F. 1986. Les Dinosaures (Car-
nosaures, Allosauridés, Sauropodes,
Cétiosauridés) du Jurassique moyen de Cerro
Cóndor (Chubut, Argentine). (part 1). Annales de
Paléontologie (Vert.-Invert.), 72: 247-289.

BUATOIS, L.A., & LOPEZ ANGRIMAN, A. 1992a.
Evolución de sistemas deposicionales en el Grupo
Gustav, Cretácico de la Isla James Ross, Antártida.
Pp. 263-297. In, Rinaldi, C. (ed.) ‘Geología de la
Isla James Ross’. (Instituto Antártico Argentino;
Buenos Aires).

BUATOIS, L.A. & LOPEZ ANGRIMAN, A. 1992b.
Trazas fósiles y sistemas deposicionales del Grupo
Gustav, Cretácico de la Isla James Ross, Antártida.

! Welles & [5 (1974) describe the astragalar ascending process of Elaphrosaurus bambergi as high and broad, citing

inference. Hence the proportions are here considered to be unknown.

symmetric

Pp. 239-262. In, Rinaldi, C. (ed.) ‘Geologia de la
Isla James Ross’. (Instituto Antártico Argentino:
Buenos Aires).

BRITT, B.B. 1991. Theropods of Dry Mesa Quarry
(Morrison Formation, Late Jurassic), Colorado,
with emphasis on the osteology of Torvosaurus
tanneri. Brigham Young University Geology
Studies, 37: 1-72.

CARPENTER, K, 1992. Tyrannosaurids (Dinosauria)
of Asia and North America. Pp. 250-268. In,
Mateer, N. & Chen PJ. (eds) ‘Aspects of Non-
marine Cretaceous Geology’. (China Ocean Press:
Beijing).

COLBERT, E.H. 1989. The Triassic dinosaur
Coelophysis. Bulletin of the Museum of Northern
Arizona. 57: 1-174.

CURRIE, P. J., & RUSSELL, D.A. 1988. Osteology and
relationships of Chirostenotes pergracilis (Sauris-
chia, Theropoda) from the Judith River (Oldman)
Formation of Alberta, Canada. Canadian Journal
of Earth Sciences, 25: 972-986.

CURRIE, P. J. & ZHAO, X. 1993. A new carnosaur
(Dinosauria, Theropoda) from the Jurassic of Xin-
jiang, People’s Republic of China. Canadian Jour-
nal of Earth Sciences, 30: 2037-2081.

DONG, Z., ZHANG, S. & ZHOU, Y. 1983. The
dinosaurian remains from Sichuan Basin, China.
Palaeontologia Sinica, 162: 1-145.

674

EUDES-DESLONGCHAMBPS, J.-A. 1837. Mémoire
sur le Poikilopleuron bucklandii, grand saurien
fossile, intermédiare entre les crocodiles et les
lézards. Mémoire de la société Linnéenne de Nor-
mandie, 6: 37-146.

GASPARINI, Z., OLIVERO, E., SCASSO, R. &
RINALDI, C. 1987. Un ankylosaurio (Reptilia,
Ornithischia) Campaniano en el continente Antar-
tico. Anais do X Congresso Brasileiro de Paleon-
tologia, 131-141.

GILMORE, C.W. 1920. Osteology of the carnivorous
Dinosauria in the United States National Museum,
with special reference to the genera Antrodemus
(Allosaurus) and Ceratosaurus. United States Na-
tional Museum, Bulletin, 110: 1-159.

HAMMER, W.R. & HICKERSON, W.J. 1994. A
crested theropod dinosaur from Antarctica.
Science, 264: 828-830.

HOLTZ, T.R., Jr 1994, The phylogenetic position of the
Tyrannosauridae: implications for theropod sys-
tematics. Journal of Paleontology, 68: 1100-1117.

HOOKER, J.J., MILNER, A.C. & SEQUEIRA, S.E.K.
1991. An ornithopod dinosaur from the late
Cretaceous of West Antarctica. Antarctic Science,
3: 331-332.

HULKE, J.W. 1879. Note on Poikilopleuron bucklandi
of Eudes-Deslongchamps (pere) identifying it
with Megalosaurus bucklandi. Quarterly Journal
of the Geological Society of London, 35; 233-238,

HUXLEY, T.H. 1870. On the classification of the
Dinosauria with observations of the Dinosauria of
the Trias. Quarterly Journal of the Geological
Society of London, 26: 31-50.

INESON, J., CRAME, J. & THOMSON, M. 1986.
Lithostratigraphy of the Cretaceous strata of west
James Ross Island, Antarctica. Cretaceous Re-
search, 7: 141-159.

JANENSCH, W. 1925. Die Coelurosaurier und
Theropoden der Tendaguru-Schichten Deutsch-
Ostafrikas. Palaeontographica (Suppl. 7) 1: 1-99,

LYDEKKER, R. 1891. On certain ornithosaurian and
dinosaurian remains. Quarterly Jounral of the
Geological Society of London, 47: 41-44.

MADSEN, J. H., Jr 1976. Allosaurus fragilis: a revised
osteology. Utah Geological and Mineral Survey,
Bulletin, 1091: 1- 163.

MARTINEZ, R., GIMENEZ, O., RODRIGUEZ, J. &
BOCHATEY, G. 1987. Xenotarsosaurus bonapartei
nov. gen. et sp. (Carnosauria, Abelisauridae), un
nuevo Theropoda de la Formacion Bajo Barreal,
Chubut, Argentina. Actas IV Congreso Argentino
de Paleontología y Bioestratigrafía, 23-31.

MEDINA, F., BUATOIS, L., & LOPEZ ANGRIMAN,
A. 1992, Estratigrafía del Grupo Gustav en la Isla
James Ross, Antártida. Pp. 167-192. In, Rinaldi,
C. (ed.) 'Geología dela Isla James Ross'. (Instituto
Antártico Argentino: Buenos Aires).

MOLNAR, R.E. 1989. Terrestrial tetrapods in
Cretaceous Antarctica. In, Crame, J.A. (ed.)
‘Origins and Evolution of the Antarctic Biota’,

MEMOIRS OF THE QUEENSLAND MUSEUM

Geological Society Special Publication, 47: 131-
140.

MOLNAR, R.E., KURZANOV, S.M. & DONG, Z.
1990. Carnosauria. Pp. 169-209. In, Weishampel,
D.B., Dodson, P. & Osmólska, H. (eds) “The
Dinosauria'. (University of California Press:
Berkeley).

MOLNAR, R.E. & WIFFEN, J. 1994. A Late
Cretaceous polar dinosaur fauna from New
Zealand. Cretaceous Research, 15: 689-706.

OLIVERO, E., GASPARINI, Z., RINALDI, C. &
SCASSO, R. 1991. First record of dinosaurs in
Antarctica (Upper Cretaceous, James Ross Is-
land): palaeogeographical implications. Pp. 617-
622. In, Thomson, M., Crame, J. & Thomson, J.,
(eds) ‘Geological Evolution of Antarctica’.
(Cambridge University Press: Cambridge).

OSTROM, J.H., 1969. Osteology of Deinonychus an-
tirrhopus, an unusual theropod from the Lower
Cretaceous of Montana. Bulletin, Peabody
Museum of Natural History, 30: 1-165.

1970. Stratigraphy and paleontology of the Cloverly
Formation (Lower Cretaceous) of the Bighorn
Basin area, Wyoming and Montana. Bulletin,
Peabody Museum of Natural History, 35: 1-234.

PADIAN, K. 1986. On the type material of Coelophysis
Cope (Saurischia: Theropoda) and a new
specimen from the Petrified Forest of Arizona
(Late Triassic: Chinle Formation). Pp. 45-60. In,
Padian, K. (ed.) ‘The Beginning of the Age of
Dinosaurs’. (Cambridge University Press: New
York).

RAATH, M.A., 1969. A new coelurosaurian dinosaur
from the Forest Sandstone of Rhodesia. Arnoldia,
4: 1-25.

RUSSELL, D.A. & DONG, Z. 1993, The affinities of a
new theropod from the Alxa Desert, Inner Mon-
golia, People's Republic of China. Canadian Jour-
nal of Earth Sciences, 30: 2107-2127.

SAUVAGE, H.E. 1882. Recherches sur les reptiles
trouvés dans le Gault de l'est du Basin de Paris.
Mémoires de la Société géologique de France, (3),
4: 439-442,

SERENO, P.C., WILSON, J.A., LARSSON, H.C.E.,
DUTHEIL, D.B. & SUES, H.-D. 1994. Early
Cretaceous dinosaurs from the Sahara. Science,
266: 267-271.

SPOTILA, J.R., O'CONNOR, M.P., DODSON, P. &
PALADINO, F.V. 1991. Hot and cold running
dinosaurs: body size, metabolism and migration.
Modern Geology 16: 203-227.

WELLES, S.P. 1984. Dilophosaurus wetherilli
(Dinosauria, Theropoda). Osteology and com-
parisons. Palacontographica, A, 185: 85-180.

WELLES, S.P. & LONG, R.A. 1974. The tarsus of
theropod dinosaurs. Annals of the South African
Museum, 64: 191-218.

WIFFEN, J. & MOLNAR, R.E. 1989. An ornithopod
dinosaur from New Zealand. Geobios, 22:531-
536.

ALVAREZSAURIDAE, CRETACEOUS BASAL BIRDS FROM PATAGONIA
AND MONGOLIA

FERNANDO E. NOVAS

Novas, F.E. 1996 12 20: Alvarezsauridae, Cretaceous basal birds from Patagonia and
Mongolia. Memoirs of the Queensland Museum 39(3): 675-702. Brisbane. ISSN 0079-8835.

Alvarezsauridae represents a clade of bizarre birds with extremely reduced but powerful
forelimbs. Twenty synapomorphic features shared by Patagonykus, Alvarezsaurus and
Mononykus supports Alvarezsauridae as a monophyletic group of avialan theropods. Diag-
nostic characters, mainly referred to vertebral, forelimb, pelvic and hindlimb anatomy,
emerge from acladistic analysis of 74 derived features depicting Alvarezsauridae as the sister
taxon of the avialian clade Ornithothoraces. Since the origin and early diversification of the
Alvarezsauridae probably took place during, or prior to, the Early Cretaceous, their common
presence in Patagonia and Mongolia reflects a wider geographical distribution over the world,
prior to the development of major geographical barriers between Laurasia and Gondwana
during Aptian to Cenomanian times. [ ] Alvarezsauridae, Patagonykus, Mononykus, birds.

Los Alvarezsauridae constituyen un clado de extrañas aves basales, caracterizados por sus
miembros anteriores extremadamente reducidos, aunque proporcionalmente robustos.
Veinte sinapomorfías compartidas por Patagonykus, Alvarezsaurus y Mononykus sustentan
la hipótesis que Alvarezsauridae conforma un grupo monofilético de terópodos avialanos.
Los caracteres diagnósticos de Alvarezsauridae se refieren principalmente a la columna
vertebral, miembros anteriores y posteriores, y pélvis, y emergen de un análisis cladístico de
74 rasgos derivados muestran a Alvarezsauridae como el grupo hermano del clado avialiano
Ornithothoraces. En base a las hipótesis filogenéticas propuestas, se estima que el origen y
temprana radiación de los alvarezsáuridos habría ocurrido, al menos, durante el Cretácico
temprano. Este dato permite suponer que estos terópodos se habrían dispersado en varios
continentes (p.ej.; América del Sur, América del Norte, y Asia) antes que se instalaran
barreras geográficas de importancia entre Laurasia y Gondwana durante el Cretácico
‘Medio’.

Fernando E. Novas, Museo Argentino de Ciencias Naturales, Av. Angel Gallardo 470, 1405
Buenos Aires, Argentina; 1 February 1995.

Alvarezsauridae (Bonaparte, 1991) is a clade of
bizarre avialan theropods from Upper Cretaceous
rocks of Mongolia and Patagonia. At present they
are known by three different species: the
Patagonian Alvarezsaurus calvoi (Bonaparte,
1991), Patagonykus puertai gen. et sp. nov.
(Novas, in press a) and the Mongolian
Mononykus olecranus (Perle et al., 1993, 1994;
Chiappe et al., this volume). By far, the latter
species is the best represented one, being known
from complete cranial and postcranial skeletons
(Perle et al, 1993; 1994; Chiappe et al., this
volume). On the contrary, the South American
taxa Patagonykus and Alvarezsaurus are not
completely known and the most serious lack of
information refers to the skull. Nevertheless, the
available osteological material pertaining to the
Patagonian forms is informative, allowing recog-
nition of autapomorphic features diagnostic of
each of the Patagonian species. Although incom-
pletely represented, Patagonykus and Alvarez-
saurus are significant phylogenetically since they

retained the plesiomorphic state for several fea-
tures that Mononykus shares with birds, more
derived than Archaeopteryx.

Better documented now than they were five
years ago (Bonaparte, 1991), some aspects of the
phylogenetic relationships of the alvarezsaurids
are now better understood. For example, it is now
clear that Alvarezsauridae does not constitute a
theropod branch of uncertain relationships, as
originally interpreted by Bonaparte (1991); on
the contrary, they are deeply internested within
Tetanurae, Coelurosauria and Maniraptora be-
cause they exhibit hypapophyses on vertebrae
from the cervicothoracic region, semilunate car-
pal, retroverted pubis, posterodorsal margin of
ilium ventrally curved in lateral view and pubic
foot cranially reduced (Gauthier, 1986).

There are, however, several features that make
these theropods particularly interesting: first,
they exhibit peculiar adaptations in the forelimbs
and vertebral column, the functional significance
of which is controversial (e.g., Perle et al., 1993,

676

1994; Ostrom, 1994); second, alvarezsaurids
share with birds several apomorphic resemblan-
ces, more derived than Archaeopteryx, raising

new questions about the early evolution of birds '

(Perle et al., 1993, 1994; Chiappe et al., this
volume); third, alvarezsaurids are known from
distant Upper Cretaceous localities of the world
(e.g., Patagonia and Mongolia) and thus are inter-
esting from a palaeobiogeographical point of
view.

ABBREVIATIONS. AMNH, American Museum
of Natural History, New York; BSP, Bayerische
Staatssammlung für Paläontologie, Munich; CM,
Carnegie Museum of Natural History, Pittsburgh;
HMN MB, Humboldt Museum für Naturkunde,
Berlin; GI, Geological Institute, Mongolian
Academy of Sciences, Ulan Bator; MACN,
Museo Argentino de Ciencias Naturales 'B.
Rivadavia', Buenos Aires; MCZ, Museum of
Comparative Zoology, Cambridge; MLP, Museo
de La Plata, La Plata; MUCPV, Museo de Cien-
cias Naturales, Universidad Nacional del Coma-
hue, Neuquén; PVL, Paleontología de
Vertebrados, Fundación ‘Miguel Lillo’, San
Miguel de Tucumán; PVPH, Paleontología Ver-
tebrados, Museo Municipal 'Carmen Funes',
Plaza Huincul, Neuquén; PVSJ, Museo de Cien-
cias Naturales, Universidad Nacional de San
Juan, San Juan; USNM, United States National
Museum, Washington, DC; YPM, Yale Peabody
Museum, New Haven.

MATERIALS AND METHODS

MATERIAL EXAMINED. A comparative study of the
holotypes of Patagonykus puertai (PV PH 37), Alvarez-
saurus calvoi (MUCPV 54) and Mononykus olecranus
(GI N107/6 cast) was conducted. The following
specimens were also studied: Albertosaurus libratus
(AMNH 5468), Alectrosaurus olseni (AMNH 6554),
Allosaurus fragilis (AMNH 5767), Archaeopteryx
lithographica (HMN MB 1880/81 and casts of London
and Eichstátt specimens), Archaeornithomimus
asiaticus (AMNH 6566, 6567, 6570), Caiman
latirostris (pers. collection), Compsognathus longipes
(BSP AS I 536), Deinonychus antirrhopus (AMNH
3015, MCZ 4371, YPM 5205, 5206, 5236), Her-
rerasaurus ischigualastensis (PVSJ 373), Iberomesor-
nis romerali (MACN unnumbered cast), Meleagris
gallopavo (pers. collection), Mussaurus patagonicus
(MLP-68-III-27-1), Ornitholestes hermani (AMNH
619), Ornithomimus velox (AMNH 5355), Or-
nithomimus sedens (USNM 2164) and Piat-
nitzkysaurus floresi (MACN-CH 895).

ANATOMICAL TERMINOLOGY. I follow the
terminology of Clark, 1993. ‘Cranial’ and

MEMOIRS OF THE QUEENSLAND MUSEUM

*caudal' are used here in place of 'anterior' and
‘posterior’, respectively.

SYSTEMATIC NOMENCLATURE. [ascribe to
the notion of phylogenetic (node-based or stem-
based) definitions for all taxa (de Queiroz &
Gauthier, 1994). Aves is defined to encompass all
the descendants of the most recent common an-
cestor of Ratitae, Tinami and Neognathae
(Gauthier, 1986); Avialae includes Archaeop-
teryx lithographica, Aves and their most recent
common ancestor; Maniraptora includes all those
theropods more closely related to Aves than to the
Ornithomimidae (Gauthier, 1986). With respect
to Maniraptora, I am not following the
synapomorphy-based definition given by Holtz
(1994), who has also included ornithomimids and
tyrannosaurids within Maniraptora on the as-
sumption that the ancestors of these two taxa also
possessed the diagnostic features of Maniraptora
(e.g., raptorial manus, etc.). The node-based
definition originally given by Gauthier (1986:30)
perfectly fits to the clade formed by
Dromaeosauridae plus Avialae, even accepting
the monophyly of Arctometatarsalia (= El-
misauridae + Avimimus + Tyrannosauridae +
(Troodontidae + Ornithomimosauria); sensu
Holtz, 1994).

SYSTEMATIC PALAEONTOLOGY

Basic information on the new taxon
Patagonykus puertai is provided here. A detailed
anatomical description and discussion of the
autamorphies diagnosing this species are given
elsewhere (Novas, in press a).

COELUROSAURIA Huene, 1920
MANIRAPTORA Gauthier, 1986
AVIALAE Gauthier, 1986
METORNITHES Perle et al., 1993
ALVARESAURIDAE Bonaparte, 1991

Patagonykus gen. nov.

Patagonykus puertai gen. et sp. nov.

MATERIAL EXAMINED. HOLOTYPE PVPH 37,
two incomplete dorsal vertabrae, incomplete sacrum,
two proximal and two distal caudal vertabrae; incom-
plete left and right coracoids, proximal and distal ends
of both humeri, right proximal portions of ulna and
radius, and distal portion of left ulna, articulated car-
pometacarpus and first phalanx of digit I of the right
manus; incomplete ungual phalanx probably cor-
responding to digit I; portions of ilia, proximal ends of
ischia, and portions of pubes; proximal and distal por-

CRETACEOUS MANIRAPTORAN THEROPODS

677

Ornithomimidae Tyrannosauridae Deinonychus Archaeopteryx Alvarezsaurus Mononykus Patagonykus Ornithothoraces

COELUROSAURIA

MANIRAPTORA

ALVAREZSAURIDAE

FIG. 1. Cladogram depicting the phylogenetic relationships among Patagonykus, Alvarezsaurus and Mononykus,

and five immediate outgroups.

tions of right femur, and distal end of the left; proximal
and distal ends of both tibiae, fused with proximal
tarsals; metatarsals II and III fused to distal tarsals III;
several pedal phalanges.

HORIZON AND AGE. Portezuelo Member of
the Río Neuquén Formation (possibly Turonian,
Late Cretaceous; Cruz et al., 1989), Sierra del
Portezuelo, 22km west of Plaza Huincul City,
Neuquén Province, NW Patagonia, Argentina.
The quarry is situated 500m NW of National
Route 22.

DIAGNOSIS. Patagonykus puertai is an alvarez-
saurid avialian theropod diagnosed by the follow-
ing: postzygapophyses in dorsal vertebrae with
ventrally curved, tongue-shaped lateral margin;
dorsal, sacral and caudal vertabrae with a bulge
on the caudal base of the neural arch; humeral
articular facet of coracoid transversely narrow;
internal tuberosity of humerus subcylindrical,
wider at its extremity rather than in its base;
humeral entepicondyle conical and strongly
projected medially; first phalanx of manual digit
I with proximomedial hook-like processes; ecto-
condylar tuber of femur rectangular in distal
view.

PHYLOGENETIC RESULTS

Seventy four derived features were coded as
binary and their distribution examined in three
terminal taxa and five outgroups (see Appendix).
The data matrix was subjected to parsimony
analysis using the implicit enumeration (ie) com-

mand in HENNIG 86 (version 1.5) by J.S. Farris
(1988). A single most parsimonious tree was ob-
tained (Fig. 1), with a length of 102 steps, a
consistency index of 0.72 and a retention index
of 0.74. This tree depicts Alvarezsauridae as the
sister group of Ornithothoraces, but it must be
emphasised that the tree supporting Alvarez-
sauridae outside Avialae (namely as the sister
taxon of Deinonychus plus Avialae) differs in five
evolutionary steps (characters listed in the Ap-
pendix). Until more evidence becomes available,
I will consider alvarezsaurids as birds (e.g.,
avialians more derived than Archaeopteryx), in
agreement with Perle et al. (1993, 1994) and
Chiappe et al. (this volume).

OUTGROUP RELATIONSHIPS. The following
taxa have been chosen for outgroup comparisons:
Ornithothoraces, Archaeopteryx, Deinonychus,
Tyrannosauridae and Ornithomimidae. Although
there is a diversity of opinion about the
phylogenetic arrangement of the terminal taxa,
there is agreement among authors (e.g., Bakker et
al., 1988; Novas, 1991; 1992; Holtz, 1994) that
the Tyrannosauridae are more closely related to
Ornithomimidae, Dromaeosauridae, Aves and
other coelurosaurs, than to Allosaurus (contra
Gauthier, 1986; Molnar et al., 1990). Another
point of consensus is that Dromaeosauridae and
Avialae form a clade (e.g., Maniraptora;
Gauthier, 1986; Novas, 1991; 1992; Holtz, 1994).

In reference to the Avialae, the available data
supports Alvarezsauridae as the sister taxon of
Omithothoraces, with Archaeopteryx as the out-

678

group (Perle et al., 1993, 1994; Chiappe et al., this
volume).

TERMINAL TAXA. Patagonykus puertai, Al-
varezsaurus calvoi and Mononykus olecranus
were chosen as operational taxonomic units
(OTU’s) for parsimony analysis. The latter two
taxa are diagnosed below on the basis of apomor-
phic characters, which are numbered and
preceded by a letter identifying their presence in
the corresponding taxon (A, Alvarezsaurus; M,
Mononykus).

Alvarezsaurus calvoi Bonaparte, 1991. This
taxon was originally diagnosed by Bonaparte
(1991) on the basis of several features, some of
which are problematic. For example, presence of
‘cervical pleurocoels ... 5 or 6 sacrals ... ilium low
and long ... unfused metatarsals and tarsals ...
astragalus with wide condyles ... metatarsal III
narrower in caudal view respect to the remaining
metatarsals ... metatarsal IV greater proximally
than the other metatarsals’ are all characters
widely distributed within Tetanurae (Gauthier,
1986), and clearly none of these features is
autapomorphic for Alvarezsaurus calvoi. Other
features (e.g., cranial sacrals with a slight axial
depression on ventral surface; ilium with pos-
tacetabular blade greater than the preacetabular
one) are widely — and unevenly — distributed
among several non-avialian and avialian taxa and
their status is difficult to verify. Other characters
originally included in the diagnosis of this taxon
(Bonaparte, 1991), such as ‘neural spines ves-
tigial in cervical and cranial dorsal vertebrae’
and ‘caudal sacrals with narrow ventral margin’
are features also present in Mononykus and
Patagonykus and are better interpreted as diag-
nostic of the Alvarezsauridae. Bonaparte has also
listed the small size of the specimen as a diagnos-
tic feature of Alvarezsaurus. However, the lack of
fusion of the centra and respective neural arches
of the cervical vertebrae, as well as the unfused
centra of the sacrals, reveal that the holotype
specimen of Alvarezsaurus calvoi was an imma-
ture individual that probably did not reach its
maximum body size.

Restudy of the partial skeleton of Alvarez-
saurus calvoi (MUCPV 54) allowed recognition
of the following autapomorphies:

Al) Cervical centra amphicoelous (Fig. 2).
Bonaparte (1991) originally recognised this char-
acter as diagnostic of Alvarezsaurus. This condi-
tion sharply contrasts with that present in the
remaining Tetanurae (including Mononykus), in
which the cranial articular surfaces of cervical

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 2. Cervical vertebrae of Alvarezsauridae. A-D,
F, Alvarezsaurus calvoi (MUCPC 54). A, B, F, cervi-
cal ?5th. D, 7th. C, 8th. E, G, Mononykus olecranus
(GI N107/6 cast). E, ?6th to ?8th. G, 9th. (A, lateral;
B, cranial; C, D, E, dorsal; F,G, ventral views). ep-
epipophysis, n=neural spine, p=parapophysis, pc=
processus caroticus, pl=pleurocoel; scale = 10mm.

centra are flat or convex. Procoelous cervical
centra are also present in Ornithomimidae (e.g.,
Archaeornithomimus AMNH 6566, 6567, 6570).

A2) Cervical postzygapophyses dorsoventrally
flattened, paddle-shaped in dorsal view, and with
a pair of strong craniocaudal ridges (Fig. 2). Al-
varezsaurus exhibits paddle-shaped, cranio-
caudally elongate postzygapophyses on cervical
vertebrae (Bonaparte, 1991). This condition con-
trasts with other theropods (e.g., Piatnitzky-
saurus, Archaeornithomimus, Ornitholestes,
Deinonychus) in which the postzygapophyses are
rectangular, not constricted at their bases, and
with a convex dorsal surface bearing a prominent

CRETACEOUS MANIRAPTORAN THEROPODS

epipophysis. In the caudal cer-
vicals of Alvarezsaurus the
postzygapophyses exhibit a
strong, craniocaudally-
oriented buttress running along
the medial margin. This condi-
tion is seen in cervicals 7 to 9,
but in cervicals 7 and 8 a lateral
crest is also present, bounding
a shallow basin over the dorsal
surface of the postzygapo-
physes. This condition of the
dorsal surface of the
postzygapophyses contrasts
with that present in other
theropods, including Mononykus
(GI N 107/6) in which the
dorsal surface of the cervical
postzygapophyses is smooth
and transversely convex.

A3) Length of distal caudals
more than 200% of the length
of proximal caudals. In Al-
varezsaurus (MUCPV 54) the
centrum of the distalmost
preserved caudal (presumably
corresponding to the region of
caudals 15 through 18), is
213% of the length of the most
proximally preserved caudal.
This condition resembles that
of Archaeopteryx (Wellnhofer,
1974; 1988; 1993), in which
the longest tail vertebrae
(caudals 12 and 13) represents
185 to 287% of the length of the
proximal caudal vertebrae
(caudals 1 through 3). The
elongation of the distal caudal
segments in Alvarezsaurus
sharply contrast with the proportions seen in non-
avialian maniraptorans (e.g., Ornitholestes,
Sinornithoides, Deinonychus; Osborn, 1917;
Ostrom, 1969; Russell & Dong, 1993b) in which
the length of the distal caudals represents no more
than 175% the length of the proximal caudals.
Alvarezsaurus also differs from other alvarez-
saurids, in which the length of caudal vertebrae
remains more or less similar along the tail (e.g.,
Mononykus; Perle et al., 1994), or they are con-
siderably smaller than the proximal ones, as in
Patagonykus (PVPH 37). As for the Or-
nithothoraces, the distal caudal vertebrae are
uniformly short (e.g., Baptornis, Hesperornis,
Ichthyornis, Patagopteryx; Marsh, 1880; Martin

679

p

E
j

FIG. 3. Pectoral girdle of alvarezsaurids. A, Alvarezsaurus calvoi (modified
from Bonaparte, 1991). B, E, Mononykus olecranus (GI N107/6 cast). C,
D, Patagonykus puertai (composite reconstruction based on left and right
coracoids of PVPH 37). (A, B, C, left lateral view of scapula and coracoid;
D, E, caudal view of left coracoid). r=craniocaudal ridge, scale = 20mm.

& Tate, 1976; Alvarenga & Bonaparte, 1992),
but this condition can not be easily considered
ancestral for the clade, since in basal ornitho-
thoracines (e.g., Jberomesornis, Sinornis; Sanz et.
al., 1988; Sereno & Rao, 1992) the distal caudals
are strongly fused forming a pygostyle, prevent-
ing measurement of the length of each vertebral
segment.

A4) Scapular blade slender and reduced (Fig.
3). Bonaparte (1991) included this feature in his
original diagnosis. Although the distal end of the
scapula of Alvarezsaurusis broken off, the cranial
and caudal margins of the blade tend to converge
distally, suggesting the lack of a distal expansion
as present in other theropods (e.g., Mononykus,
Archaeopteryx, Tyrannosaurus, Allosaurus;

680

FIG. 4. Ungual phalanx of first manual digit of Alvarezsauridae. A, B, C,
Alvarezsaurus calvoi (MUCPC 54). D, E, F, Mononykus olecranus (GI
N107/6). (A, D, lateral; B, E, ventral; C, F, proximal views). Scale = 10mm.

Osborn, 1905; Madsen, 1976; Ostrom, 1976a;
Perle et al., 1994). The scapula of Alvarezsaurus
seems to be more slender than that of Mononykus
(Perle et al., 1994), a conclusion that emerges
when the scapula is compared with other skeletal
elements: for example, in Alvarezsaurus
(Bonaparte, 1991) the scapula represents 47% of
the craniocaudal length of the iliac blade, instead
in Mononykus (Perle et al., 1994) the scapula
represents 86% of the craniocaudal length of the
ilium.

A5) Ungual phalanx of digit I ventrally keeled
(Fig. 4). Revision of the holotype specimen of
Alvarezsaurus calvoi (MUCPV 54) allowed iden-
tification of a manual ungual phalanx, originally

MEMOIRS OF THE QUEENSLAND MUSEUM

undescribed by Bonaparte
(1991). This phalanx is
remarkably similar to that of
Mononykus (see character 22),
although in Alvarezsaurus the
first manual ungual exhibits a
strongly developed ventral
keel on its proximoventral sur-
face. By contrast, manual
claws of most theropods (e.g.,
Allosaurus, Ornithomimus,
Deinonychus, Patagonykus,
Mononykus; Marsh, 1896;
Ostrom, 1969; Madsen, 1976;
Perle et al., 1994) this keel is
absent and the ventral ungual
surface is transversely rounded
or flattened.

Mononykus olecranus Perle
et al., 1993. The list of charac-
ters offered below differs from
that originally given by Perle et
al. (1993, 1994), not only be-
cause some features exhibit a
wider distribution than pre-
viously thought, but also since
new features have been recog-
nised. Mononykus differs from
other alvarezsaurids in the fol-
lowing autapomorphies:

M1) Absence of pleurocoels
in cervical vertebrae (Fig. 2).
Presence of pleurocoels in cer-
vical vertebrae is a common
feature among Theropoda
(Gauthier, 1986). This condi-
tion seems to be ancestral for
the Alvarezsauridae, because
pleurocoel openings are
present in neck vertebrae of
Alvarezsaurus (Bonaparte,
1991). Contrarily, cervical vertebrae of
Mononykus lack pleurocoels (Perle et al., 1994),
a character convergently acquired in ornithurine
birds (Chiappe, in press).

M2) Presence of sulcus caroticus in cervical
vertebrae (Fig. 2). In Mononykus the cranio-
ventral margin of the cervical centra is complex,
due to the presence of a craniocaudal groove
laterally bounded by a strongly developed ventral
processes. This ventral process of Mononykus
resembles the processus caroticus of modern
birds, in which the major muscle mass of M.
longus colli ventralis is attached (Baumel & Wit-
mer, 1993). This character is present in other

CRETACEOUS MANIRAPTORAN THEROPODS

Mesozoic birds (e.g., Ichthyor-
nis, Hesperornis and
presumably Patagopteryx;
Marsh, 1880; Alvarenga &
Bonaparte, 1992), although it is
unknown in other avialans
(e.g., Iberomesornis, Neu-
quenornis, Sanz et al.,1988;
Chiappe & Calvo, 1994). The
phylogenetic status of this
character is uncertain (i.e.,
synapomorphic of Metornithes
or autapomorphic of Mono-
nykus), mainly because the sul-
cus and lateroventral processes
are absent in the cervicals of
Alvarezsaurus calvoi (MUCPV
54).

M3) Presacral vertebrae with
diapophyses and parapophyses
occupying the same level (Fig.
5). Perle et al. (1994) noted this
peculiar condition for
Mononykus which is unique
among Theropoda. The
preserved dorsal vertebrae of
Patagonykus (PVPH 37) show
the ancestral archosaur condi-
tion in which the parapophyses
are cranioventrally placed with
respect to the diapophyses. The
few dorsal vertebrae known in
Alvarezsaurus correspond to
the cranial region (Bonaparte,
1991).

M4) Dorsal vertebrae lacking hyposphene-
hypantrum, and postzygapophyses lateroventral-
ly oriented (Fig. 5). As Perle et al. (1994)
described, all of the presacral vertebrae of
Mononykus lack hyposphene-hypantrum ar-
ticulations. However, these authors did not in-
clude this feature in the diagnosis of Mononykus,
but interpreted it as an equivocal synapomorphy
of Metornithes (Chiappe et al., this volume).
However, I do not agree with this interpretation.
Although it is true that Mononykus lacks hypos-
phenes (viz., the postzygapophyses are elongate
and separated from each other by a deep cleft),
the same is not true for Patagonykus, because the
postzygapophyses are ventrally confluent in a
block-like hyposphene, the proportions of which
do not significantly differ from those of other
theropods such as Deinonychus and Allosaurus
(Ostrom, 1969; Madsen, 1976).

—

681

—

FIG. 5. Dorsal vertebrae of theropod taxa showing the morphology of the
postzygapophiseal region. A, B, Deinonychus antirrhopus (‘11th? dorsal
vertebra’; from Ostrom, 1969). C, D, Mononykus olecranus (‘middorsal’;
modified from Perle et al., 1994). E, F, Patagonykus puertai (middorsal;
PVPH 37). (A, C, E, left lateral; B, D, F caudal view). d=diapophysis,
h=hyposphene, l-lateroventral margin of postzygapophysis, p=
parapophysis; scale = 10mm.

Another curious aspect of Mononykus is that
the articular surface of the postzygapophyses is flat
and faces lateroventrally, a condition uniformly
present along the dorsal series (Perle et al., 1994).
On the contrary, the articular facet of the post-
zygapophyses in other theropods (e.g., Al-
losaurus, Deinonychus, Patagonykus) is
ventrally concave and faces more ventrally than
laterally.

M5) Cranial dorsal vertebrae transversally
compressed. In Mononykus (Perle et al., 1994) the
centra of the cranial dorsal vertebrae are strongly
compressed transversely. As a result, a
pronounced ventral keel is present in the cranial
dorsal vertebrae. This condition is absent in the
available dorsals of Patagonykus (PVPH 37), in
which the centra are transversely wider and
ventrally flat. Preserved cranial dorsal vertebrae
of Alvarezsaurus (MUCPV 54) are transversely
compressed and a slight ventral keel is present,
but it is not so marked as in Mononykus.

682

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 6. Middorsal vertebrae of alvarezsaurids showing
the morphology of central articular facets. A, B,
caudal portion of dorsal vertebra of Patagonykus
puertai (PVPH 37). C, D, Mononykus olecranus
(from Perle et al., 1994). (A, C, right lateral view; B,
D, ventral; broken bone surface indicated by dashed
lines). prz=prezygapophysis of another dorsal ver-
tebra, adjacent caudally; scale = 10mm.

M6) Caudal dorsal vertebrae strongly
procoelous. Mononykus is one of the few
theropods in which caudal dorsal vertebrae are
strongly procoelous. Patagonykus resembles
Mononykus in that caudal dorsal vertebrae exhibit
the procoelous condition, but they differ in that
the convexity of the caudal articular surface of
dorsal vertebrae is considerably more
pronounced in Mononykus than in the Patagonian
taxon (Fig. 6). In Alvarezsaurus the condition of
the dorsal vertebrae is unknown. As Perle et al.
(1993, 1994) pointed out, a strong procoelous
condition for caudal dorsals is uncommon among
theropods and the other case in which it was
reported is the ornithothoracine bird Patagop-
teryx (Alvarenga & Bonaparte, 1992; Chiappe,
1992).

M7) Extreme transverse compression ofthe last
sacral vertebra (Fig.7). The last sacral of
Mononykus exhibits extreme transverse compres-
sion. This condition sharply differs from that
present in Patagonykus and Alvarezsaurus, in
which the sacral centra are considerably less
compressed transversely. The transverse com-
pression described for Mononykus is accom-
panied by a ventral projection of the centrum
below the level of the caudal articular surface.

FIG. 7. Last sacral vertebrae of alvarezsaurids. A, B,
Patagonykus puertai (PVPH 37). C, D, Mononykus
olecranus (GI N107/6 cast). (A, C, right lateral view;
B, D, caudal). Scale = 10mm.

This modification is evident when the last sacral
of both Mononykus and Patagonykus is compared
in caudal aspect (Fig.7). In the first taxon the
ventral keel nearly equals the dorsoventral depth
of the caudal articular surface of the centrum,
while in Patagonykus the ventral keel is consid-
erably less developed with respect to the caudal
articular surface. Alvarezsaurus shows the same
condition as Patagonykus.

M8) Coracoidal shaft elliptical in lateral view
(Fig. 3). The coracoid of Mononykus is elliptical,
as seen in lateral aspect, being craniocaudally
long and dorsoventrally low. This morphology is
in sharp contrast to that of other maniraptorans
(e.g., Deinonychus, Sinornithoides, Archaeop-
teryx, Ornithothoraces; Ostrom, 1969, 1974,
1976a; Russell & Dong, 1993; Walker 1981;
Chiappe, 1996) in which the coracoid is dor-
soventrally deeper than craniocaudally long with
a rectangular to strut-like shape. The coracoid of
Mononykus resembles the ancestral theropod
condition (Gauthier, 1986), and is better inter-
preted as an evolutionary reversal that is diagnos-
tic for this taxon. The actual shape of the
coracoids of Patagonykus and Alvarezsaurus is
not known (the reconstruction given in Fig. 3 is
approximate). Hence, the distribution of this
character may be wider than thought and its
phylogenetic status different.

CRETACEOUS MANIRAPTORAN THEROPODS

M9) Coracoid transversely
flat (Fig.3). In Mononykus (GI
N107/6) the lateral surface of
the coracoid is slightly convex
craniocaudally, but it is
lateromedially flat. Instead, in
Patagonykus, as well as other
theropods (e.g., Allosaurus,
Deinonychus, Archaeopteryx),
the coracoid is strongly in-
flected medially, with the
lateral surface proximodistally
and craniocaudally convex.
Also, Patagonykus has a sharp
craniocaudal ridge along the
lateral surface of the coracoid.
Unfortunately, the coracoid of
Alvarezsaurus is too poorly
preserved to discern the condi-
tion of this feature.

M10) Sternum with thick
carina. The presence of an os-
sified sternal keel is interpreted
as a synapomorphy of Metor-
nithes (Perle et al., 1993, 1994;
Chiappe et al., this volume).
Mononykus, however, is
peculiar among avialians in
that the sternal carina is
transversely thick and V-
shaped in cranial view, instead
of being transversely narrow
and T-shaped as in or-
nithothoracine birds (Perle et
al, 1994, Chiappe et al., this
volume). Unfortunately, the
sternum is unknown in both Al-
varezsaurus and Patagonykus
and for this reason the condi-
tion described above for
Mononykus constitutes an am-
biguous autapomorphy of the
later taxon.

M11) Radius with extensive
articular surface for the ulna
(Fig. 8). In Mononykus (Perle et al., 1994) the
proximocaudal portion of the radius forms a
single, extensive surface for articulation with the
ulna. In Patagonykus, instead, two
proximocaudal surfaces for the ulnar articulation
are present. However, these surfaces are consid-
erably smallerthan those of the Mongolian taxon.

M12) Radius with carpal articular facet hyper-
trophied (Fig. 8). Mononykus is unique among
theropods in the unusual development of the

20mm.

683

FIG. 8. Forelimbs of alvarezsaurids. A, B, composite reconstruction based
on left and right forelimb bones of Patagonykus puertai (PVPH 37). C, D,
Mononykus olecranus (from Perle et al., 1994). (A, C, lateral view; B, D,
caudal view of first phalanx and ungual phalanx of digit I). c=carpometacar-
pus, elp=extensor ligamentary pit, h=humerus, ph=first phalanx, pr=
proximomedial ridge, r=radius, ul=ulna, ug=ungual phalanx; scale =

radiocarpal articular facet (Perle et al., 1994).
This morphology is almost certainly absent in
Patagonykus, because the shaft of the radius is
rod-like and triangular in cross section, lacking
indications of the presence of an hypertrophied
radiocarpal articular facet.

M13) First phalanx of digit I with a very
prominent proximocaudal process (Fig. 8). In
Mononykus the proximocaudal comer of the first
phalanx of digit I develops a prominence that
wraps over the ginglymus of metacarpal I. This

684

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 9. Pelves of maniraptorans in lateral aspect. A, Adasaurus mongoliensis (from Barsbold, 1983). B,
Patagonykus puertai (composite reconstruction based on left and right bones of the pelvis of PVPH 37). C,
Mononykus olecranus (from Perle et al., 1994). Not to scale.

process, probably related to the insertion of
strong extensor muscles, is absent in other
theropods, including Patagonykus.

M14) Pubis caudoventrally oriented (Fig. 9). In
Mononykus (Perle et al., 1994; Chiappe et al., this
volume) the main axis ofthe pubic shaft describes
an angle of nearly 70? with the proximal iliac
surface of the pubis. In Patagonykus, instead, the
proximal portion of the pubic shaft is oriented
almost perpendicularly with respect to this sur-
face of the pubis. The differences in pubic
retroversion documented within Alvarezsauridae
supports the interpretation that a strong
caudoventral orientation of the pubis, becoming
parallel to the ischium, evolved more than once
within Metornithes: once in Mononykus and
again in birds more derived than Archaeopteryx
(Wellnhofer, 1974; 1988; 1993).

M15) Pubic foot absent (Fig. 9). Mononykus
lacks a distal pubic foot (Perle et al., 1993, 1994;
Chiappe et al., this volume). This feature was
originally thought to be an ambiguous
synapomorphy of Metornithes, because a distal
foot was present in the basal ornithothoracine
Sinornis (Sereno & Rao, 1992, Perle et al., 1993;
Chiappe, 1995c). However, the absence of a dis-
tal foot in the pubis of Mononykus is interpreted
as autapomorphic of this taxon, because a well
developed distal expansion is documented in the
pubis of Patagonykus and basal ornithothoracine
birds. The pubes are not preserved in Alvarez-

saurus nor in the basal ornithothoracine bird
Iberomesornis (Sanz et al., 1989). In the context
of all the evidence, I interpret the lack of a pubic
foot as independently evolved in Mononykus and
ornithothoracines more derived than Patagop-
teryx.

M16) Ischium extremely reduced (Fig. 9). In
Mononykus (Perle et al., 1993, 1994; Chiappe et
al., this volume) the ischium is markedly reduced.
In Patagonykus the ischium is also reduced with
respect to other theropods (e.g., Ornitho-
mimidae), but its proximal end, at least, is more
massive than that of Mononykus. Patagonykus
retained a well defined ischiac pedicle on the
ilium, with the antitrochanter extending over both
ilium and ischium. The articulation between
pubis and ischium is dorsoventrally deep in the
Patagonian taxon. In Mononykus the ilium and
ischium are strongly fused, and the antitrochanter
is formed by both pelvic bones (Perle et al., 1994).

M17) Femoral distal condyles transversely ex-
panded, nearly confluent below popliteal fossa
(Fig. 10). In Mononykus (Perle et al., 1994) the
medial condyle of the femur is extremely ex-
panded transversely, its transverse axis being
nearly 75% of its craniocaudal extension. Perle et
al. (1993, 1994) noted the distal enclosure of the
popliteal fossa resembling the condition present
in more derived birds (Chiappe & Calvo, 1994;
Chiappe, 1992, 19952). This enclosure of the
popliteal fossa results from the transverse expan-

CRETACEOUS MANIRAPTORAN THEROPODS

sion of both the medial distal condyle and the
ectocondylar tuber, which are almost in contact
with each other, distal to the popliteal fossa. On
the contrary, in Patagonykus, as in other non-
avialian theropods (e.g., Deinonychus, Ornitho-
mimidae, Tyrannosauridae) this fossa is entirely
open distally, because both the medial condyle
and the ectocondylar tuber are less expanded
transversely. The modification described above
for the femur of Mononykus correlates with that
present in the proximal tibia, in which the outer
condyle greatly expands transversely, contacting
with the inner condyle of the same bone. In regard
to the tibia, Patagonykus exhibits the ancestral
condition, with the outer condyle less expanded
transversely.

M18) Tibia with accessory (medial) cnemial
crest. As interpreted by Perle et al. (1993, 1994),
the presence of a smooth crest on the medial face
of the proximal tibia is convergent with Or-
nithurae, because a medial crest is absent in
Maniraptora ancestrally (e.g., Deinonychus;
Ostrom, 1969), as well as in Patagonykus.

M19) Outer malleolus of distal tibia
craniocaudally thick (Fig. 11). In Alvarezsaurus
and Patagonykus the outer malleolus of the distal
tibia is craniocaudally narrow with respect to that
portion of the calcaneum that is in front of it,
representing nearly 36% of the craniocaudal
diameter of the calcaneum. In Mononykus, in-
stead, the outer malleolus is craniocaudally thick-
er, representing nearly 143% of the craniocaudal
dimension of that portion of the calcaneum that
is in front of it.

M20) Astragalocalcaneum with deep intercon-
dylar groove (Fig. 11). In Mononykus the fusion
between astragalus and calcaneum is stronger,
and the intercondylar groove deeper, than in both
Patagonykus and Alvarezsaurus. In the latter two
taxa the astragalocalcaneal suture is still visible.

M21) Ascending process of astragalus laterally
displaced (Fig. 11). In Mononykus the ascending
astragalar process is deeply notched along its
medial margin, resulting in an ascending process
transversely narrow in cranial view (Perle et al.,
1994). Also, the ascending process of the
astragalus appears to be more laterally placed
than in other maniraptorans (e.g., Patagonykus,
Alvarezsaurus, Deinonychus), as suggested by
the overlap of the lateral margin of this process
with the lateral margin of the tibia. This condition
contrasts with that present in other Coelurosauria
(e.g., Tyrannosauridae, Ornithomimidae,
Dromaeosauridae, Patagonykus, Alvarezsaurus)
in which the ascending process of the astragalus

685

A

mc

et

et mc
p

mc
Em

FIG. 10. Left distal femora of alvarezsaurids. A, B, C,
Patagonykus puertai (PV PH 37). D, E, F, Mononykus
olecranus (from Perle et al., 1994). (A, D, caudal
view; B, E, medial; C, F, distal). et=ectocondylar
tuber, Ic=lateral condyle, mc=medial condyle, pf=
popliteal fossa; scale = 20mm.

is transversely wide. Particularly the medial mar-
gin of this process is not notched as in
Mononykus, but extends in a nearly straight line
from the medial condyle of the astragalus to the
proximal tip of the ascending process. The lateral
margin of the ascending process does not reach
the lateral border of the tibia. Interestingly, the
condition of Mononykus resembles that of
modern avians, in which the ‘pretibial bone’
(homologous to the ascending process of the
astragalus; McGowan, 1985) is transversely nar-
row and laterally displaced.

M22) Femoral trochanteric crest present. In
Coelurosauria ancestrally (e.g., Ornithomimidae,
Tyrannosauridae), as well as in Deinonychus
(MCZ 4371), Patagonykus, Alvarezsaurus and

686

FIG. 11. Right distal tibia and astragalocalcaneum of Alvarezsauridae. A, B,
C, Alvarezsaurus calvoi (MUCPV 54). D, E, F, Patagonykus puertai
(PVPH 37). G, H, I, Mononykus olecranus (GI N107/6). (A, D, G, cranial
view; B, E, H, lateral; C, F, I, distal; reconstruction indicated by dashed
lines). as=astragalus, apa=ascending process of astragalus, c=calcaneum,

f=fibula, om=out tibial malleolus; scale = 10mm.

Archaeopteryx, both anterior and greater femoral
trochanters are separated by a cleft. In contrast
Mononykus exhibits a femoral trochanteric crest
(e.g., anterior trochanter undivided from the
greater trochanter), a derived condition that is
shared with more derived birds.

M23) Fibula does not articulate with the tarsus
(Fig. 11). This feature was originally considered
by Perle et al. (1993) as synapomorphic of Metor-
nithes, because it is shared by Mononykus and the
Ornithothoraces. However, other alvarezsaurids
exhibit the plesiomorphic condition: in
Patagonykus the fibula is incomplete, but the
presence of a deep socket on the proximal cal-
caneal surface suggests that in this taxon the
fibula articulated with the tarsus, a condition
that also found in Alvarezsaurus (Bonaparte,
1991).

MEMOIRS OF THE QUEENSLAND MUSEUM

INGROUP RELATION-
SHIPS. Patagonykus and
Mononykus share seventeen
characters which are absent or
unknown in Alvarezsaurus.
The monophyly of Alvarez-
sauridae is supported here by
eleven derived characters.

Patagonykus + Mononykus
clade. Several features support
the conclusion that
Patagonykus and Mononykus
are more closely related than
either is to Alvarezsaurus.
Some of these features are ab-
sent in Alvarezsaurus and are
readily interpreted as synapo-
morphic of the Patagonykus +
Mononykus clade (characters 1
& 2), others are contingent
upon the results of the cladistic
analysis (characters 3 & 4).
However, the condition of most
of the apomorphies uniting
Patagonykus and Mononykus
is unknown for Alvarezsaurus,
because of the fragmentary na-
ture of the available material.
Consequently, a large set of
characters (5 through 17) used
to unite Patagonykus +
Mononykus may become
synapomorphic of a more in-
clusive group (e.g., Alvarez-
sauridae), pending additional
information on Alvarezsaurus:

1) Caudal articular surface of the centra of the
last sacral and proximal caudal vertebrae strongly
spherical. Mononykus, Patagonykus and Alvarez-
saurus Clearly differ from other theropods since
they share last sacral and most of the caudal
vertebrae with a ball-shaped caudal surface (char-
acter 19). However, Alvarezsaurus seems to be
less derived than other alvarezsaurids because the
caudal surfaces of the last sacral vertebra and
caudal vertebrae are not as spherical as in
Patagonykus and Mononykus. Moreover, the
procoelous condition in Alvarezsaurus appears to
be confined to the last sacral and to the caudal
vertebrae, since the cranial articular surface of the
presumed sacral 2 of Alvarezsaurus is almost
planar.

2) Sacral vertebrae ventrally keeled (Fig. 7). In
Allosaurus (Gilmore, 1920), Gallimimus

CRETACEOUS MANIRAPTORAN THEROPODS 687

(Osmólska et al., 1972) and
Archaeornithomimus (AMNH
6567) the ventral surface of
sacral centra is smooth and
convex in cross-section. This
condition also applies to
Ornithomimus (USNM 2164;
see Gilmore, 1920), although
in this taxon the sacral centra
are ventrally grooved. The
sacral centra of Deinonychus
(MCZ 4371), Ornitholestes
(AMNH 619) and Archaeo-
pteryx (Wellnhofer, 1974,
1993) are not keeled ventrally.
Sacral vertebrae are not keeled
in Jberomesornis and Sinornis
(Sanz et al., 1988; Chiappe,
pers. comm.) and the same is
true for more derived birds
(e.g., Ichthyornis, Hesperor-
nis; Marsh, 1880): other
avialans, instead, bear a ventral
groove (e.g., Baptornis,
Patagopteryx; Martin & Tate,
1976; Chiappe, 1992; Perle et
al, 1994). Alvarezsaurids dif-
fer from the remaining
Coelurosauria in the presence
of a ventral keel in the caudal
sacral vertebrae. However, the
degree of transverse compres-
sion of this keel varies among
alvarezsaurids: in Alvarez-
saurus (MUCPV 54) the
presumed sacral 1 is transver-
sely wide and ventrally convex
in cross-section, while the
sacrals 2-3 are ventrally
grooved, resembling the condi-
tion present in Ornithomimus
(Gilmore, 1920). The penul-
timate sacral vertebra of
Alvarezsaurus (presumably
sacral 4) is more compressed
transversely than more cranial
sacrals; this vertebra is slightly
keeled, but a rudimentary
ventral canal is still present.
The last sacral (presumably

FIG. 12. Humeri of theropod taxa, A, E, Mononykus olecranus (modified
from Perle et al., 1994). B, F, Patagonykus puertai (composite reconstruc-
tion based on left and right humerus of PVPH 37). C, D, G, Deinonychus
antirrhopus; C, proximal end (from Ostrom, 1969), D, G, distal end of
humerus (reconstruction based on AMNH 3015 and MCZ 4371). (A, B, C,
D, cranial view; E, F, G, distal). dc=distal condyle, dp=deltopectoral crest,
h=humeral head, ec=ectepicondyle, en=entepicondyle, it=internal
tuberosity, re=radial condyle, uc=ulnar condyle; scale = 10mm.

ing the maximum compression in the last sacral

sacral 5) of Alvarezsaurus bears a ventral keel, (presumably sacral 5). In Mononykus the first
but it is not so prominent and transversely com- sacral is transversely more compressed than the
pressed as in Patagonykus and Mononykus. In dorsal vertebrae (Perle et al., 1994) and exhibits
Patagonykus (PVPH 37) the transverse compres- a slight longitudinal ridge along its ventral sur-
sion of the ventral keel increases caudally, show- face. I interpret Mononykus as more derived than

688

B

pm

—

FIG. 13. First phalanx of digit I of maniraptoran
theropods. A, B, Deinonychus antirrhopus phalanx of
the left manus (A, taken from Ostrom, 1969; B, from
YPM 5206). C, D, Mononykus olecranus phalanx of
the right manus (C, from Perle et al., 1994; D, from
GIN107/6). E, F, Patagonykus puertai phalanx of the
right manus (PVPH 37). (A, C, E, proximal view; B,
D, F, distal). pm=proximomedial ridge; scale =
10mm.

other alvarezsaurids in this respect, since the
sacrals are extremely compressed laterally (Perle
et al., 1994; see character M7).

3) Femoral fourth trochanter present. A femoral
fourth trochanter is present in basal coelurosaurs
(e.g., Ornithomimidae, Tyrannosauridae). This
structure has been retained, albeit reduced, in
dromaeosaurids as it is seen in Deinonychus
(MCZ 4371; contra Ostrom, 1976b). Archaeop-
teryx, Alvarezsaurus, and Ornithothoraces lack
the fourth trochanter, a condition interpreted as
synapomorphic of Avialae. In the context of the
evidence, the shared presence of a fourth
trochanter in Mononykus and Patagonykus is
considered as an apomorphic reversal.

4) Supracetabular crest present. The absence of
a supracetabular crest is hypothesised as

MEMOIRS OF THE QUEENSLAND MUSEUM

synapomorphic of Maniraptora, because such a
crest is lacking in Deinonychus, Archaeopteryx,
Alvarezsaurus and the Ornithothoraces. Follow-
ing that, the presence of such a crest in the ilium
of Patagonykus and Mononykus is interpreted as
a secondary reversal.

5) Posterior dorsal vertebrae procoelous (Fig.
6). Patagonykus and Mononykus share
procoelous caudal dorsal vertebrae. However, in
the first taxon the convexity of the caudal ar-
ticular surface is considerably less marked than
in Mononykus, in which this surface is ball-
shaped. Interestingly, the development of a ball-
shaped, caudal articular surface in sacral
vertebrae is almost the same in Patagonykus and
Mononykus (Fig. 7) suggesting that the
procoelous condition evolved from caudal to dor-
sal vertebrae. Mid and caudal dorsal vertebrae are
unknown in Alvarezsaurus (Bonaparte, 1991),
but it is possible that they may have been am-
phiplatyan or amphicoelous, since the disarticu-
lated second sacral vertebra of the holotype
(MUCPY 54) has a flat caudal surface. Alvarez-
saurus, however, 1s coded as a question mark for
this character.

6) Bicipital tubercle of coracoid absent (Fig. 3).
A bicipital tubercle is absent in the coracoids of
Mononykus, Patagonykus and presumably also in
Alvarezsaurus (unfortunately, most of the caudal
and distal portions of the coracoid are missing in
the holotype). This condition sharply contrasts
with that present in Theropoda ancestrally. For
example, in Allosaurus the bicipital tubercle is
slightly marked, but it forms a distinct
prominence on the lateral surface of the coracoid,
as is seen in caudal view (Madsen, 1976). In
ornithomimids (e.g., Ornithomimus AMNH
5355; Archaeornithomimus AMNH 6567, 6566)
the biceps tubercle is also prominent, but itis even
more developed in maniraptorans. Such is the
case for Deinonychus (YPM 5236), which ex-
hibits a biceps tubercle more developed than in
most theropods (Ostrom, 1974) and Archaeo-
pteryx in which the tubercle is proportionally
larger than that of Deinonychus (Ostrom, 1974;
Wellnhofer, 1988, 1993). In avialians more
derived than Archaeopteryx (e.g., Enantiornithes;
Walker, 1981) the acrocoracoid (hypothesised as
the homologue of the bicipital tubercle; Ostrom,
1976a), consists of a robust structure, that in the
Ornithothoraces is proximally placed with
respect to the glenoid facet (Chiappe, 1992).

7) Forelimbs less than 20% of hindlimb length.

In Theropoda ancestrally the forelimbs represent
40-53% of hindlimb length, as it occurs in

CRETACEOUS MANIRAPTORAN THEROPODS

Herrerasaurus, Coelophysis, Syntarsus,
Dilophosaurus, Piatnitzkysaurus and Compsog-
nathus (Raath, 1969, Colbert, 1989; Ostrom,
1978; Welles, 1984; Novas, 1994; Sereno, 1994).
Possession of extremely short forelimbs (with
respect to hindlimb length) seems to have
evolved more than once in theropod evolution
(Novas, 1992; Perle et al., 1994). Examples of
such shortening are seen in the neoceratosaurian
theropods (e.g., Ceratosaurus, Carnotaurus;
Bonaparte et al., 1990; Novas, 1992), the basal
tetanurine Torvosaurus (Galton & Jensen, 1979;
Holtz, 1994) and the coelurosaurian Tyran-
nosauridae (Novas, 1991, 1992; Holtz, 1994). In
the latter taxon the forelimbs represent 22-26%
of the hindlimb length (Lambe, 1917), but in
Mononykus the forelimbs are even shorter than in
tyrannosaurids, since they represent nearly 18%
of the length of the hindlimbs (Perle et al., 1993,
Fig. 2). In Patagonykus (PVPH 37) the forelimbs
are known from portions of humerii, ulnae and
the almost complete left manus (carpometacarpus
plus digit I). Although it is not possible to obtain
a ratio of forelimb versus hindlimb length for
Patagonykus, the proportions of the currently
available bones of this taxon suggest that the
forelimbs also were very short. Bonaparte (1991)
arrived at the conclusion that Alvarezsaurus pos-
sibly had reduced forelimbs on the basis of the
proportionally reduced scapular girdle of this
theropod. This suspicion is supported by the stout
aspect and large size of the ungual phalanx of
digit I (Fig. 4). The morphology of this ungual
allows the prediction that the proximal elements
of the forelimb of Alvarezsaurus were mor-
phologically similar to those of Mononykus and
Patagonykus (Perle et al., 1993, 1994).

8) Humeral head with major transverse axis
ventrolaterally inclined with respect to the lon-
gitudinal axis of the humerus, and internal
tuberosity proximally projected (Fig. 12). In
Mononykus the humeral head is lateroventrally
inclined with respect to the longitudinal axis of
the humerus. Additionally, the internal tuberosity
is proximally projected, reaching nearly the same
level as the humeral head. By contrast, in other
theropods (e.g., Deinonychus; Ostrom, 1969) the
major transverse axis of the humeral head is
horizontally held with respect to the longitudinal
axis of the humerus, or it is inclined
ventromedially (e.g., /chthyornis; Marsh, 1880)
with respect to the longitudinalaxis. Also, in most
theropods the internal tuberosity is a cone-shaped
structure (e.g., Ceratosauria, Tyrannosauridae,
Ornitholestes AMNH 619; Rowe & Gauthier,

689

A

FIG. 14. Ilium of maniraptoran theropods in ventral
aspect. A, right ilium of Alvarezsaurus calvoi
(MUCPY 54). B, left ilium (reversed) of Deinonychus
antirrhopus (AMNH 3015). Sandstone matrix indi-
cated by stippling. bs=brevis shelf, cp=fossa for
origin of M. cuppedicus, pp=pubic peduncle; scale =
20mm.

1990; Osborn, 1917), or it forms a longitudinally
expanded prominence that is medially and slight-
ly caudally projected, and distally placed with
respect to the humeral head (e.g., Deinonychus,
Ichthyornis; Marsh, 1880; Ostrom, 1969). The
known humerus of Patagonykus has unconnected
proximal and distal portions, such that determina-
tion of the orientation of the humeral head and
internal tuberosity with respect to the humeral
shaft is difficult to assert. However, the proximal
end of the humerus of Patagonykus closely
resembles that of Mononykus, and the features
described above for the later taxon seem to apply
also to the Patagonian species. Since the humerus
is unknown in Alvarezsaurus (Bonaparte, 1991)
the peculiar morphology of the proximal end of
the humerus shared by Patagonykus and

690

Mononykus is considered as an ambiguous
synapomorphy of the Alvarezsauridae.

9) Humerus with a single distal condyle; ulna
and radius tightly appressed proximally, forming
a cup-like articular surface for the humerus (Fig.
12). In Theropoda, ancestrally (for example,
Piatnitzkysaurus MACN-CH 895; Deinonychus
AMNH 3015, MCZ 4371; Ornitholestes AMNH
619; Ornithomimidae; Barsbold & Osmdlska,
1990), the distal end of the humerus exhibits two
distal condyles, the radial one being larger than
the ulnar condyle. Both ulnar and radial condyles
are separated by an intercondylar groove, a con-
dition that 1s retained in recent birds (Baumel &
Witmer, 1993). Also, the humeral cotylus of the
ulna is subtriangular in proximal view, with an
acute cranial projection and a craniolateral con-
cavity for the reception of the radius. In
Theropoda, ancestrally, the radius and ulna lack
proximal articular facets between them. Also, the
ulna and radius bear independent proximal ar-
ticular surfaces for the ulnar and radial condyles
of the humerus, respectively. Alvarezsaurids, in-
stead, are unique among archosaurs in the
peculiar propodial-epipodial articulation, con-
sisting in the presence of a single, and well
developed, condyle in the distal humerus which
articulates with a continuous, cup-like articular
surface formed by both radius and ulna. In
Patagonykus and Mononykus the humeral cotylus
of the ulna is subelliptically shaped and lacks the
lateral concavity to accommodate the radius. The
latter bone has moved entirely over the cranial
aspect of the ulna and both bones are tightly
articulated proximally.

Some minor differences exist in the propodial-
epipodial articulation of alvarezsaurids (Fig. 12):
while in Patagonykus the humeral distal condyle
is ball-shaped and elliptical in contour, in
Mononykus it is trochlear and trapezoidal in distal
aspect. The complex morphology of the distal
humeral condyle of Mononykus is due to the
strong craniocaudal expansion of its medial side,
equalling its transverse axis (Fig. 12A, E). Fur-
thermore, the proximal articulation between the
radius and ulna is extremely extended in
Mononykus (Perle et al., 1994) and is readily
interpreted as autapomorphic for the Mongolian
taxon (see character M11).

10) Olecranal process of ulna strongly
developed (Fig. 8). The development of the
olecranal process is variable within Dinosauria,
although it is feebly developed in most theropods
(e.g., Syntarsus, Piatnitzkysaurus, Allosaurus,
Ornithomimidae, Deinonychus, Archaeopteryx,

MEMOIRS OF THE QUEENSLAND MUSEUM

Omnithothoraces; Ostrom, 1969, 1976a; Raath,
1969; Madsen, 1976; Bonaparte, 1986a; Baumel
& Witmer, 1993; Barsbold & Osmólska, 1990).
Patagonykus and Mononykus differ from the
remaining dinosaurs because they share a strong-
ly developed ulnar olecranon process: that in
Mononykus represents 40% of the whole length
of the bone (Perle et al., 1994). The relative size
and the stout morphology of the ungual phalanx
of manual digit I of Alvarezsaurus (see characters
7 & A5) suggests that the presence of a well
developed olecranal process probably had a dis-
tribution wider than thought.

11) Ulnar caudal margin straight (Fig. 8).
Gauthier (1986) considered ‘ulna bowed
posteriorly' as a diagnostic character of Manirap-
tora. Avialae and Deinonychus, and apparently
also in Troodon and Sinornithoides (Russell,
1969; Russell & Dong, 1993b) the caudal border
of the ulna is uniformly convex, excepting the
distal extremity that, as in other tetanurines, is
posteriorly projected (the ulna is slightly bowed
in the Ornithomimidae, although it is not convex
caudally but cranially — that is towards the
radius; Nicholls & Russell, 1985; Barbold &
Osmólska, 1990; Osmólska et al., 1972). In
Mononykus (Perle et al., 1994) the ulna is straight
in lateral and cranial aspects. Preserved portions
of the ulna of Patagonykus show that the caudal
margin is straight. The straight shaft of the ulna
in alvarezsaurids is interpreted as an evolutionary
reversal that might be related to the strong reduc-
tion of the forelimbs.

12) Carpometacarpus massive, short and quad-
rangular (Fig. 8). Patagonykus and Mononykus
share a carpometacarpus (= semilunate carpal +
first metacarpal) that is dorsoventrally com-
pressed and quadrangular in dorsal view, a con-
dition quite unusual for a theropod (Perle et al.
1994; Novas, in press a). In Patagonykus and
Mononykus the distal condyle of metacarpal I is
transversely wide and dorsoventrally com-
pressed, with a shallow dorsoventral groove. In
contrast to other theropods such as Her-
rerasaurus (PVSJ 373; Sereno, 1994),
Coelophysis (Raath, 1969), Torvosaurus (Galton
& Jensen, 1979), Allosaurus (Madsen, 1976) and
Deinonychus (Ostrom, 1969), metacarpal I is
proximodistally long and transversely narrow
and the distal end forms a ginglymoid articula-
tion. Mononykus, and presumably also
Patagonykus, differ from other coelurosaurs
(e.g., Oviraptor, Deinonychus, Velociraptor,
Archaeopteryx; Ostrom, 1976a; Barsbold et al.,
1990) in that the semilunate carpal articulates

CRETACEOUS MANIRAPTORAN THEROPODS

distally only with metacarpal I, instead of with
both metacarpal I and II as in Maniraptora an-
cestrally (Gauthier, 1986). Size disparity between
metacarpal I and the semilunate carpal is readily
apparent in alvarezsaurids: the transverse width
of metacarpal I nearly matches that of the semi-
lunate carpal. This disparity is probably due to
hypertrophy of the metacarpal I.

The main difference between Mononykus and
Patagonykus is that in the first taxon the semi-
lunate articulation of the carpometacarpus is con-
siderably more extended cranially and distally
than in Patagonykus, resulting in a very close
approximation between the semilunate articula-
tion and the distal condyle of the metacarpal I. In
Deinonychus and Archaeopteryx (Ostrom, 1969,
19762) the medial (cranial, if rotated) margin of
the metacarpal I forms a narrow ridge, resembling
the condition described for Mononykus (Perle et
al., 1994).

13) Digit I larger than the remaining digits of
the hands (Fig. 8). The hand of most coelurosaurs
is characterised by being gracile and elongate,
with the first digit smaller than the second digit,
both in length and in transverse diameter
(Ostrom, 1969; Gauthier, 1986; Barsbold et al.,
1990). In Mononykus, instead, digit I is much
larger (in both transverse width and proximodis-
tal length) than digits II and III. As noted by Perle
et al. (1994), metacarpals II and III are not only
strongly reduced, but they are fused to each other
and with metacarpal I, without the delimitation of
any intermetacarpal space. Interestingly, the
oviraptorosaur Ingenia (Barsbold et al., 1990)
constitutes the only known non-alvarezsaurid
theropod in which digit I is proportionally larger
than the outer digits, although the degree of the
reduction of the external digits is not so marked
as in Mononykus. Close resemblances of the
available manual bones of Patagonykus and
Alvarezsaurus suggest that the South American
taxa also shared the condition described above for
Mononykus.

14) Phalanx I of manual digit I showing B-
shaped proximal articular surface, hook-like
proximomedial processes, symmetrical distal
ginglymus and deep extensor ligamentary pit
(Fig. 13). Patagonykus and Mononykus are uni-
que among Archosauria in the morphology of the
first phalanx of digit I. This phalanx is
craniocaudally wide and lateromedially com-
pressed, resulting in a curious proximal articula-
tion, which describes a horizontal ‘B’ in proximal
aspect. This morphology sharply contrasts with
that seen in other theropods, in which the

691

proximal contour of the first phalanx of digit I is
triangular (e.g., Deinonychus; Ostrom, 1969), or
describes a vertical rectangle (as in Allosaurus;
Madsen, 1976). Another peculiarity is the
presence of a pair of strongly developed
proximomedial ridges bearing muscle scars. Dis-
tally, the ginglymus of the first phalanx forms a
craniocaudally wide, symmetrical pulley,
proximally preceded by a deep extensor ligamen-
tary pit (Figs 8 & 13). By contrast, in other
dinosaurs (e.g., Mussaurus MLP 68-II-27-1;
Deinonychus YPM 5206; Allosaurus; Madsen,
1976), the ginglymus is transversely more com-
pressed and the extensor pit is absent. Further, the
flexor ligamentary pit is more marked in
Mononykus and Patagonykus than in other
theropods.

15) Medial condyle of femur transversely wide
and distally flat (Fig. 10). The medial condyle of
the distal femur is transversely narrow and distal-
ly convex in non-avialian theropods (e.g., Al-
losaurus CM 21726; Tyrannosaurus CM 9380;
Deinonychus MCZ 4371), Archaeopteryx
(Eichstatt specimen, cast) and early Ornitho-
thoraces (e.g., Enantiornithes, MACN unnum-
bered cast). In Patagonykus and Mononykus,
instead, the medial condyle of the distal femur is
rectangular and distally flat. In the Patagonian
taxon the transverse axis of this condyle repre-
sents nearly 66% of its craniocaudal extension,
while in Mononykus it is 7596. The transverse
extension of the medial condyle of the distal
femur resembles that present in recent birds, such
as Rhea (pers. collection).

16) Ectocondylar tuber caudally projected, well
behind the medial distal condyle (Fig. 10). In
Allosaurus (Gilmore, 1920), Ornitholestes
(AMNH 619), Gallimimus (Osmólska et al.,
1972), Tyrannosaurus (USNM 10.754) and
Deinonychus (MCZ 4371), the ectocondylar crest
of the femur does not caudally surpass the level
of the medial femoral condyle. In Patagonykus
and Mononykus, instead, the ectocondylar crest
strongly projects caudally, surpassing the medial
femoral condyle. This condition resembles that
present in modern birds (e.g., Rhea) in which the
crista tibiofibularis (the caudal portion of which
is considered to be homologous with the ectocon-
dylar tuber; Chiappe, 1996) is more caudally
projected than the medial condyle of the femur.

. 17) Fibular condyle of femur conical and

projected distally (Fig. 10). In Mononykus and
Patagonykus the fibular condyle of the femur
constitutes a well defined, cone-shaped structure,
which is strongly projected distally. This condi-

692

tion contrasts with that seen in other theropods
(e.g., Allosaurus, Archaeornithomimus, AMNH
6570; Deinonychus) in which the lateral condyle
of distal femur is only slightly more projected
distally with respect to the medial condyle. The
lateral condyle of the femur is conical in some
ornithomimids (e.g., Gallimimus, Osmólska et
al., 1972, pl. 46), as well as in the tyrannosaurid
Alectrosaurus (AMNH 6554), but in neither case
is it so prominent as in alvarezsaurids. The con-
dition of this femoral condyle in Archaeopteryx,
Iberomesornis and the enantiornithine Sinornis is
unknown (Ostrom, 1976a; Wellnhofer, 1974;
Sanz et al., 1988; Sereno & Rao, 1991, Chiappe,
1995c), but in more derived avialians (e.g.,
Hesperornis, Ichthyornis, Patagopteryx, Enan-
tiornithes; Marsh, 1880; Walker, 1981; Alvarenga
& Bonaparte, 1992) the fibular condyle of femur
is not conical but smoothly convex craniocaudal-
ly and transversely.

Alvarezsauridae Bonaparte, 1991. Alvarez-
sauridae is here defined to encompass
Patagonykus puertai, Alvarezsaurus calvoi,
Mononykus olecranus and all the descendants of
their most recent common ancestor. Alvarez-
sauridae is diagnosed on the basis of nine unam-
biguous synapomorphies, plus other two
ambiguous traits, which are listed and analyzed
below:

18) Cervical vertebrae with craniocaudally
short and dorsoventrally low neural spines (Fig.
2). Alvarezsaurus and Mononykus exhibit strong-
ly reduced neural spines on the cervical vertebrae
(Bonaparte, 1991; Perle et al., 1994), This condi-
tion strongly differs from those in other
coelurosaurs. As an example, in Ornitholestes
(AMNH 619) the neural spine of cervical ?4 is
dorsoventrally deep and craniocaudally short,
similar to Deinonychus (Ostrom, 1969). The
neural spine of a mid-caudal cervical of Ornitho-
lestes, instead, is dorsoventrally low but axially
long. In Archaeopteryx, cervicals 4 and 5
(Wellnhofer, 1974) exhibit neural spines propor-
tionally taller than Mononykus (Perle et al., 1994)
and Alvarezsaurus (Bonaparte, 1991). Unfor-
tunately, the neck vertebrae of basal avialians is
almost unknown (e.g., /beromesornis, Sinornis;
Sereno & Rao, 1992) and the condition is un-
known in Patagonykus. As originally noted by
Bonaparte (1991), gracile cervical vertebrae with
reduced neural spines are present in or-
nithomimids, a condition that is here interpreted
as independently evolved.

19) Sacral and caudal vertebrae procoelous
(Fig. 7). In Theropoda ancestrally the sacral and

MEMOIRS OF THE QUEENSLAND MUSEUM

caudal vertebrae are amphiplatyan or slightly am-
phicoelous (e.g., Piatnitzkysaurus MACN-CH
895; Archaeornithomimus AMNH 6567; Alber-
tosaurus AMNH 5458; Deinonychus MCZ
4371). In the recently described seventh
specimen of Archaeopteryx (Wellnhofer, 1993),
the disarticulated and well preserved proximal
caudals appear to be amphiplatyan, as suggested
by the straight cranial and caudal margins of the
caudal centra, parallel to each other in lateral
view. The condition is unknown in basal Or-
nithothoraces (e.g., Jberomesornis; Sanz et
31.1988) due to the firm articulation among
caudal segments forming a pygostile. A
procoelous condition of caudal dorsals, sacrals
and caudals has been documented only in
Patagopteryx among ornithothoracines
(Chiappe, 1992, 19952), a condition that was
previously interpreted as convergently acquired
with Alvarezsauridae (Perle et al., 1993, 1994).
In Hesperornithiformes, ancestrally (e.g.,
Baptornis; Martin & Tate, 1976), the caudal
centra are amphicoelous or amphiplatyan. In
Ichthyornis (Marsh, 1880) the cranial surface of
the first sacral centrum and the caudal surface of
the last sacral vertebra are concave and the caudal
vertebrae are amphicoelous.

Mononykus, Patagonykus and Alvarezsaurus
exhibit the last sacral and most of the caudal
vertebrae with a ball-shaped caudal surface. As
commented above (character 1) some variation
exists in the convexity of the caudal surface
within Alvarezsauridae.

The resemblances between the proximal
caudals of Patagonykus and eusuchian crocodiles
(e.g., Caiman; pers. collection) are noteworthy,
mainly because they share ball-shaped caudal
articular surfaces, robust, craniocaudally short
neural spines, transversely thin ligamentary scars
at the base of the neural spines, and the base of
the neural spines with a deep excavation between
pre- and postzygapophyses. It is not possible to
assert whether these osteological resemblances
between Patagonykus and extant crocodiles cor-
respond with similarities in distribution and
development of the epaxial musculature, but the
existence of a strong procoelous condition sug-
gests a high degree of movement all along the tail,
in contrast with most tetanurine theropods, in-
cluding birds (Gauthier, 1986).

20) Caudal sacral centra transversely com-
pressed. As noted above (character 2) the sacral
vertebrae of alvarezsaurids are transversely com-
pressed, in contrast with other theropods in which
the sacrum does not exhibit such a condition.

CRETACEOUS MANIRAPTORAN THEROPODS

21) Haemal arches of proximal caudals dor-
soventrally elongate. This feature was interpreted
by Martin & Rinaldi (1994) in support of non-
avialian affinities of Mononykus. However, in the
context of all the evidence, the presence of elon-
gate haemal arches in Mononykus and Alvarez-
Saurus is interpreted here as an apomophic
reversal, and consequently as diagnostic of the
Alvarezsauridae.

22) Ungual phalanx of manual digit I stout and
robust (Figs 4 & 8). Perle et al. (1993, 1994) have
noted that the ungual phalanx of digit I of
Mononykus is robust and less arched than in other
theropods (e.g., Deinonychus, Archaeopteryx)
and with the proximoventral area forming a flat
surface lacking a flexor tubercle. These authors
have indicated also, that the ungual of the first
digit represents 226% of the proximodistal length
of the carpometacarpus (Perle et al., 1994). As
commented above, the manual ungual of Alvarez-
saurus is remarkably similar to that of
Mononykus, in being quadrangular in proximal
aspect, with the proximoventral surface flat and
the flexor tubercle absent. This manual ungual,
outstandingly larger than any ungual of the foot,
exhibits deep proximal concavities for articula-
tion with the first phalanx of digit 1. These con-
cavities are separated by a prominent ridge. The
ungual phalanx of digit I of Alvarezsaurus lacks
the foramina that pierce the proximoventral sur-
face of that of Mononykus (Perle et al., 1994).

23) Pubic pedicle of ilium slender (Fig. 9). Like
other dinosaurs the pubic pedicle of the ilium of
Patagonykus and Mononykus is subtriangular in
cross-section and lateromedially compressed.
However, in alvarezsaurids the pubic pedicle is
elongate and craniocaudally narrow, in contrast
with other tetanurines (e.g., Allosaurus, Omitho-
mimidae, Ornitholestes, Deinonychus, Archaeo-
pteryx, Enantiornithes; Madsen, 1976; Osborn,
1917; Ostrom, 1969, 1976b; Walker, 1981;
Barsbold & Osmólska, 1990) in which the pubic
peduncle of the ilium is craniocaudally thick.

24) Pubic peduncle cranioventrally projected.
In Theropoda, ancestrally (e.g., Compsognathus,
Allosaurus, Ornitholestes, Tyrannosauridae,
Omithomimidae, /ngenia, Chirostenotes), the
pubic pedicle is anteroventrally oriented. This
condition is also seen in Oviraptor (AMNH
6517). In Deinonychus and Archaeopteryx the
pedicle clearly surpasses ventrally the level of the
ischiac pedicle. Also, the cranial margin of the
pubic pedicle is straight and slopes caudoventral-
ly. The caudal margin of the pedicle is more
curved caudoventrally increasing the participa-

693

tion of the ilium in the acetabular surface. In
Deinonychus (Ostrom, 1969; 1976b) and
Archaeopteryx (Martin, 1983) the pubic pedicle
inclines caudoventrally 115-130° with respect to
the longitudinal axis of the iliac blade (where the
dorsal margin is assumed horizontal). In Enan-
tiornithes (Martin, 1983) the angle is 140° ap-
proximately. In contrast, in Gallimimus
(Osmólska et al., 1972, pl. 50) and Ornitholestes
(AMNH 619) the pubic pedicle inclines 50-60?
with respect to the longitudinal axis of the blade
and in Tyrannosaurus (Osborn, 1917) the angle is
nearly 70?. Alvarezsaurus (Bonaparte, 1991),
Mononykus (Perle et al., 1993) and Patagonykus
retained the ancestral condition, with the pubic
pedicle cranioventrally oriented. The angulation
1s not possible to calculate in Patagonykus.

The ilio-pubic articulation is too highly
modified in Hesperornis to discern the inclination
of the pubic peduncle. Unfortunately, most of the
ilium is lost in Jberomesornis (Sanz et al., 1988).
In Apatornis, and presumably also in /chthyornis
(Marsh, 1880), the cranial margin of the pubic
pedicle slopes caudoventrally. In Patagopteryx
neither specimen preserves the pubic peduncle
complete, but judging from the available material
(Chiappe, 1992) the pubic peduncle seems to be
vertical, but not caudoventrally oriented.
Caudoventral orientation of the pubic peduncle is
also seen in neornithine birds (e.g., Apteryx).

25) Fossa for M. cuppedicus absent (Fig. 14).
A strongly developed fossa for the femoral
protractor M. cuppedicus (Rowe, 1986) is present
in the cranioventral corner of the ilium of Alber-
tosaurus (AMNH 5458), Ornithomimus (USNM
2164), Deinonychus (AMNH 3015, MCZ 4371),
Archaeopteryx (HMN MB 1880/81) and Enan-
tiornithes (Walker, 1981). On the contrary, the
lateral surface of the preacetabular blade of the
ilium of Alvarezsaurus (Bonaparte, 1991),
Patagonykus (PVPH 37) and Mononykus (Perle
et al., 1994) is strongly convex dorsoventrally,
lacking a fossa for the M. cuppedicus on the
ventral margin of the iliac blade. The absence of
a fossa for the M. cuppedicus has been interpreted
as synapomorphic of the avialian clade formed by
Patagopteryx and the Ornithurae (Chiappe,
1996) and more recently (Chiappe et al., this
volume) as an ambiguous synapomorphy of
Metornithes. This uncertainty rises from the un-
known condition in /beromesornis (Sanz et al.,
1988) and the presence of such an iliac fossa in
enanthiornithine birds (Walker, 1981). I prefer to
interpret the presence of an iliac fossa as primitive
for Ornithothoraces (see character matrix in the

694

Appendix). Consequently, the lack of an iliac
fossa for the M. cuppedicus is interpreted, with
reservations, as an apomorphic character conver-
gently acquired by alvarezsaurids and the clade
formed by Patagopteryx plus Ornithurae.

26) Supraacetabular crest ending cranially
above the pubic pedicle. In Patagonykus (PVPH
37) and Mononykus (GI N107/6) the supra-
acetabular crest is almost restricted to the dorsal
portion of the acetabular aperture, ending abrupt-
ly above the pubic pedicle. A similar condition is
present in Alvarezsaurus (MUCPY 54) in which
the feebly developed supraacetabular crest does
not extend over the pubic pedicle of the ilium.
This condition contrasts with that present in other
theropods (e.g., Piatnitzkysaurus MACN-CH 895;
Allosaurus, AMNH 813; Archaeornithomimus
AMNH 6576; Ornithomimus AMNH 5421;
Albertosaurus AMNH 5458, 5664; Ornitholestes
AMNH 619; Deinonychus MCZ 4371, AMNH
3015; Archaeopteryx HMN MB 1880/81) in
which the supraacetabular crest (even in the
reminiscent condition present in derived
theropods) extends cranially in continuity with
the lateral border of the pubic pedicle.

27) Postacetabular blade of ilium with brevis
shelf caudolaterally oriented and medial flange
ventrally curved (Fig. 9). In Alvarezsaurus and
Mononykus the brevis fossa is present (e.g., a
caudoventral basin bounded by a well developed
brevis shelf and the medial flange of the ilium;
Novas, in press b). The fragmentary nature of the
ilium of Patagonykus prevents knowledge of this
character. The loss of a discrete brevis shelf and
fossa apparently constitutes a synapomorphy of
the Maniraptora, and its re-acquisition is con-
sidered an apomorphic reversal diagnostic of
Alvarezsauridae. Interestingly, the or-
nithothoracine Patagopteryx also exhibits a
brevis shelf and fossa and, along with alvarez-
saurids, constitutes one of the few avialians in
which this feature is present (Chiappe, 1996).

28) ‘Posterior’ trochanter on proximal femur
absent. This feature was originally described by
Ostrom (1976a, b) for Deinonychus and Ar-
chaeopteryx. After that, the presence of such a
prominence was recognised also in Enantior-
nithes (Chiappe & Calvo, 1994; Chiappe, 1996)
and Sinornithoides (Russell & Dong, 1993b). The
absence of a ‘posterior’ trochanter in the femur of
Patagonykus, Mononykus and Alvarezsaurus is
interpreted as an apomorphic reversal.

MEMOIRS OF THE QUEENSLAND MUSEUM

DISCUSSION

Alvarezsaurids have been recorded at present
in Late Cretaceous formations in Patagonia and
Mongolia. The recorded taxa are: Patagonykus
puertai from the Río Neuquén Formation
(Turonian; Cruz et al., 1989), Alvarezsaurus
calvoi from the overlying Río Colorado Forma-
tion (Coniacian-Santonian; Bonaparte, 1991;
Cruz et al., 1989; Chiappe & Calvo, 1994) and
Mononykus olecranus documented in the Nemegt
Formation (Maastrichtian; Perle et al., 1993,
1994). Chiappe et al. (this volume) have iden-
tified bones of species related to Mononykus in
palaeontological collections previously made by
Mongolian and American palaeontologists from
the Chinese Iren Dabasu Formation, and the
Mongolian Tugrugeen Shireh, Bayn Dzak,
Ukhaa Tolgod and Barun Goyot (Khermeen
Tsav) formations, thought to be Campanian in age
(Jerzykiewicz & Russell, 1991; Currie & Eberth,
1993). More recently Holtz (1995: 511) con-
sidered ‘Ornithomimus’ minutus (Laramie For-
mation, Late Maastrichtian; Marsh, 1892;
Weishampel, 1990) as a possible member of the
*Mononykus lineage'. However the description
given by Marsh (1892) is insufficient to support
such an assignment, since the features described
for the metatarsals of ‘O’. minutus are not unique
to Mononykus, but are also shared by other
theropods with the arctometatarsalian condition,
as for example Troodontidae, Ornithomimidae,
Avimimus, Tyrannosauridae (Holtz, 1994). In
sum, although presence of alvarezsaurids may be
expected in other continents (e.g., North
America), reliable records are only known from
South America and Asia.

On the basis of the analysis presented above,
the common alvarezsaurid ancestor evolved
eleven evolutionary novelties (e.g., sacral ver-
tebrae procoelous; caudal sacral centra transver-
sely compressed and sharply keeled ventrally;
ungual phalanx of manual digit I stout and robust;
pubic pedicle of ilium slender; fossa for M. cup-
pedicus absent; supraacetabular crest ending
cranially above the pubic pedicle; haemal arches
dorsoventrally elongate; brevis fossa present;
‘posterior’ trochanter on femur absent; etc.).
However, this list of diagnostic traits for the Al-
varezsauridae may be larger with the inclusion of
several characters, the condition of which is un-
known for Alvarezsaurus. For example, the stout
morphology of the first digit ungual of the manus
of Alvarezsaurus (see characters A5 & 22) sug-
gests that the first manual digit of this taxon was

CRETACEOUS MANIRAPTORAN THEROPODS

powerfully constructed and on the basis of this
evidence it is also expected that Alvarezsaurus
possessed extremely short forelimbs plus all the
synapomorphic traits described for Mononykus
and Patagonykus (characters 7-14).

The Patagonian alvarezsaurids Patagonykus
and Alvarezsaurus are more primitive than those
from Asia, in accordance with their greater age.
However, despite their close geographic and
stratigraphic provenance, Patagonykus and
Alvarezsaurus do not exhibit derived characters
in common suggesting closer affinities within
Alvarezsauridae. On the contrary, there are four
features that Patagonykus shares with
Mononykus, exclusive of Alvarezsaurus: caudal
articular surface of the centra of the last sacral and
first caudal vertebrae strongly spherical, sacral
vertebrae ventrally keeled, supraacetabular crest
present and femoral fourth trochanter present.
These features suggest that Patagonykus and
Mononykus are descendants from a common an-
cestor not shared with Alvarezsaurus (Fig. 1).

Discovery of alvarezsaurids less derived than
Mononykus suggests that the surprisingly mcdern
avialian features present in the later taxon are the
result of convergent evolution. Such derived
avialian (e.g., ornithothoracine) characters of
Mononykus are: lack of hyposphene-hypantrum
articulation, pubis caudoventrally directed,
femoral popliteal fossa distally closed, accessory
cnemial crest, fibula not contacting tarsus and
astragalar ascending process transversely narrow
and laterally displaced. Mononykus makes a good
case for evolutionary convergence, showing that
at least some derived features were acquired more
than once in the early evolution of birds. In a more
general context, the arctometatarsalian condition
of the metatarsals (Holtz, 1994) was acquired at
least twice in theropod evolution: once in the
Arctometatarsalia clade of theropods (Holtz,
1994) and independently in the derived alvarez-
saurid Mononykus.

A phylogenetic diagram (Fig. 15) depicting the
phylogenetic relationships of the alvarezsaurids,
complemented with biochronological informa-
tion of terminal taxa (e.g., Patagonykus, Alvarez-
saurus and Mononykus) and immediate
outgroups (e.g., Ornithothoraces), suggests that
the origin and diversification of the Alvarez-
sauridae occurred before the Turonian, probably
during Valanginian times (131-138my; Haq &
Van Eysinga, 1994). This suspicion about the
time of origin of the Alvarezsauridae is in agree-
ment with the currently known biochronology of
the Ornithothoraces, the oldest representative of

695

MAASTRICHTIAN

CAMPANIAN

Mononykus

li

Alvarezsaurus
Patagonykus

[— CONIACIAN I
TURONIAN

CENOMANIAN

ALBIAN

Ornithothoraces

Archaeopteryx

FIG. 15. Phylogenetic diagram depicting phylogenetic
relationships of the Alvarezsauridae, complemented
with biochronological information.

which (Sinornis santensis; Sereno & Rao, 1992)
is known from Valanginian rocks. This informa-
tion indicates that the Alvarezsauridae has a long
evolutionary history, the recorded portion of
which approximately spans 24.5my; i.e., from
Turonian (91my) to Maastrichtian times
(66.5my; Haq & Van Eysinga, 1994).

Presence of alvarezsaurids in the Late
Cretaceous rocks in Mongolia and Patagonia is
puzzling, mainly beeause alvarezsaurids are one
of only two taxa (the other is the Titanosauridae,
represented in Mongolia by Opisthocoelicaudia
skarzynskii; Giménez, 1993; Salgado & Coria,
1993) which are shared by the sharply different
Late Cretaceous faunas of South Ameriea and
Asia (e.g., Bonaparte,1986b; Bonaparte &
Kielan-Jawarowska, 1987; Russell, 1993). There
are numerous examples of Cretaceous
Gondwanan terrestrial vertebrates which appear
to lack close phylogenetic affinities with those
from Laurasia. Bonaparte (1986b) and Bonaparte
& Kielan-Jawarowska (1987) pointed out the
remarkable differences in faunal composition
among Laurasian and Gondwanan continents,
considering such dissimilarities as a direct conse-
quence of the physical separation of both super-

696

continental landmasses, which lasted almost the
entire Cretaceous period, a time span of 70 to
80my. Russell (1993) has also agreed that several
dinosaur taxa evolved separately in
Gondwanaland, although he considered that the
grouping of terrestrial vertebrates into
'Laurasian' and “Gondwanan’ assemblages was
established by Aptian-Albian times.

The shared presence of alvarezsaurids in South
America and Asia admits two possible alterna-
tives. One explanation considers alvarezsaurid
species from South America and Asia as vicariant
taxa descended from an ancestral species widely
distributed over northern and southern landmas-
ses; this wide distribution would have occurred
before major barriers to overland dispersal
among Laurasia and Gondwana were emplaced
during Aptian to Cenomanian times (Lillegraven
et al., 1979). Following this interpretation, the
origin of the Alvarezsauridae must be traced back
to Valanginian times, during which Gondwanan
and Laurasian continents occupied closer posi-
tions than in later times (Scotese et al., 1992). The
alternative explanation considers Alvarezsaurus
and Patagonykus as endemic taxa from Gond-
wana (e.g., South America; Novas & Coria, 1990a;
Bonaparte, 1991), which evolved in isolation during
Cenomanian to Santonian times; in this context,
alvarezsaurids may be interpreted as later emigrants
to Asia (via North America?) when continental
connections occurred during the Campanian
(Bonaparte, 1986b). This later hypothesis agrees
with the available fossil record of alvarezsaurids
(Fig. 15) and also with palaeogeographic
reconstructions (e.g., Lillegraven et al., 1979;
Scotese et al., 1992) and palaeobiogeographic
interpretations of the evolution of the vertebrate
faunas of Gondwana as a whole (e.g., Bonaparte,
1986b; Bonaparte & Kielan-Jawarowska, 1987).

Current geographic documentation of the AI-
varezsauridae in South America and Asia sug-
gests that this clade successfully occupied a wide
range of environmental conditions. For example,
the Neuquén Group, which includes among
others the Río Neuquén and Río Colorado For-
mations, constitutes a succession of sandstones
and mudstones deposited under fluvial and
lacustrine conditions (Digregorio & Uliana,
1980; Legarreta & Gulisano, 1989). The fauna
recorded at the Portezuelo Member of Río
Neuquén Formation is made up of gastropods
(Cazau & Uliana, 1973), fishes, amphibians,
turtles, crocodiles, small ornithischians, small to
large theropods and large sauropods (Novas &
Coria, 1990b). That of the Río Colorado Forma-

MEMOIRS OF THE QUEENSLAND MUSEUM

tion is a very rich fauna, including the following
taxa (Bonaparte, 1991; Alvarenga & Bonaparte,
1992; Chiappe & Calvo, 1994): notosuchid and
sebecosuchid crocodiles, dinilysid booids,
sauropods, basal ornithothoracine birds (e.g.,
Patagopteryx, Neuquenornis) and ceratosaurian
theropods (Velocisaurus). In the other extreme of
sedimentological conditions in which alvarez-
saurids were recorded, are the Asiatic Djadokhta
(Tugrugeen Shireh) and Barun Goyot (Khermeen
Tsav) Formations, deposited under subaerial con-
ditions of sand dunes, small lakes and streams, in
hot and semi-arid climate in areas lacking a per-
manent fluvial system (Gradzinsky et al., 1977;
Osmólska, 1980). It seems to be clear that al-
varezsaurids inhabited a wide range of palaeo-
environments (from desertic environments, as those
indicated by the sedimentology of the Djadokhta
Formation, to more humid conditions as sug-
gested by the fluvial deposits of the Nemegt, Río
Neuquén and Río Colorado Formations).

As currently known,"Alvarezsauridae includes
small forms the size of a turkey (nearly 1m long)
such as Alvarezsaurus and Mononykus, and larger
animals, up to 2m, such as Patagonykus. They seem
to have had a role of predators of small animals,
presumably insects (Perle et al., 1994), a
suspicion based on the small size of the head and
tooth reduction. However, other food items can
not be dismissed. Chiappe (1995b) has recently
speculated that alvarezsaurids may have used the
hand claws to strip bark or perhaps stems from
low-growing vegetation. In fact, alvarezsaurids
repeat the case of almost uncertain feeding habits
as with the ornithomimosaurians, for which car-
nivorous, herbivorous, or omnivorous habits were
variously proposed (e.g., Sanz & Perez-Moreno,
1995). The unusual morphology of the forelimbs
is not readily interpreted in reference to behavior
and does not help in the elucidation of feeding
habits. As commented by Perle et al. (1993, 1994)
the sternum and forelimbs of Mononykus
resemble those of moles mainly due to the
presence of a keeled sternum, humerus short and
expanded, ulna with elongate olecranon process
and stout, slightly curved unguals. These res-
emblances have been interpreted by Ostrom (1994)
as indicative of fossorial habits for Mononykus.
However, Chiappe (1995b) has dissmissed this
interpretation. Mononykus differs from a mole
because in the latter the manus is proportionally
large, not only because the number of digits is
unreduced, but each digit is robust and a lunate
sesamoid is added on the medial side of the
manus, enlarging its palmar surface. Also, the

CRETACEOUS MANIRAPTORAN THEROPODS

body proportions of Mononykus sharply contrast
with that of digging mammals (e.g., moles and
edentates; Hildebrand, 1975), in which the body
is compact and the neck, forelimbs and hindlimbs
are shortened.

ACKNOWLEDGMENTS

I thank L.M. Chiappe, L. Witmer and J.F.
Bonaparte for their valuable criticisms and sug-
gestions on this and other papers on alvarez-
saurids; M. Norell, J. Clark and A. Perle for
sharing unpublished information on Mononykus.
lam grateful to C. Gulisano (Petrolera Argentina
San Jorge), L. Legarreta and M. Uliana (Astra Oil
Company) for their help and insightful comments
on the geology of the Neuquén Basin. For access
to specimens in their care, I thank M. Norell
(AMNH), P. Wellnhofer (BSP), D. Berman (CM),
V.-D. Heindrich (HMN), J.F. Bonaparte
(MACN), C. Shaff (MCZ), R. Pascual (MLP), O.
De Ferrariis and J. Calvo (MUCPV), J. Powell
(PVL), R.A. Coria (PVPH), A. Monetta (PVSJ),
M. Brett-Surman (USNM), J. Ostrom (YPM).
This study was supported by The Dinosaur
Society, Collection Study Grant (AMNH) and by
Petrolera Argentina San Jorge (Buenos Aires).
Figures were skillfully prepared by M. Lezcano.

LITERATURE CITED

ALVARENGA, H.M.F. & BONAPARTE, J.F. 1992. A
new flightless land bird from the Cretaceous of
Patagonia. Pp. 51-64. In Campbell, K.E., Jr (ed.)
‘Papers in Avian Paleontology Honoring Pierce
Brodkorb’. (Los Angeles County Museum of
Natural History: Los Angeles).

BAKKER, R.T., WILLIAMS, M. & CURRIE, P.J.
1988. Nanotyrannus, a new genus of pygmy
tyrannosaur, from the latest Cretaceous of Mon-
tana. Hunteria 1(53): 1-30.

BARSBOLD, R. 1983. Carnivorous dinosaurs from the
Cretaceous of Mongolia. Transactions, Joint
Soviet-Mongolian Paleontological Expedition 19:
5-120. (In Russian).

BARSBOLD, R. , MARYANSKA, T. & OSMOLSKA,
H. 1990. Oviraptorosauria. Pp. 249-258. In
Weishampel, D.B., Dodson, P. & Osmólska, H.
(eds) "The Dinosauria'. (University of California
Press: Berkeley).

BARSBOLD, R., & OSMOLSKA, H. 1990. Or-
nithomimosauria, Pp. 225-244. In Weishampel,
D.B., Dodson, P. & Osmólska, H. (eds) ‘The
Dinosauria'. (University of California Press:
Berkeley).

BAUMEL, J.J. & WITMER, L.M. 1993. Osteologia.
Pp. 45-132. In Baumel, J.J., King, A.S., Breazile,
J.E., Evans, H.E. & Vanden Berge, J.C. (eds)
‘Handbook of Avian Anatomy: Nomina

697

Anatomica Avium’ (Second Editiion) Publica-
tions of the Nuttall Ornithological Club 23.
BONAPARTE, J.F. 1986a. Les dinosaures (car-
nosaures, allosauridés, sauropodes, cetiosauridés)
du Jurassique Moyen de Cerro Céndor (Chubut,
Argentine). Annales de Paléontologie (Vertébrés-

Invertébrés) 72: 247-386.

1986b. History of the terrestrial Cretaceous ver-
tebrates of Gondwana. IV Congreso Argentino de
Paleontología y Bioestratigrafía Actas 2: 63-95.

1991. Los vertebrados fósiles de la Formación Río
Colorado, de la ciudad de Neuquén y cercanías,
Cretácico Superior, Argentina. Revista del
Museo Argentino de Ciencias Naturales ‘B.
Rivadavia', Paleontología 4: 17-123.

BONAPARTE, J.F. & KIELAN-JAWAROWSKA, Z.
1987. Late Cretaceous dinosaur and mammal
faunas of Laurasia and Gondwana. Pp. 24-29. In
Currie, P.M. & Koster, E.H. (eds) ‘Fourth Sym-
posium on Mesozoic Terrestrial Ecosystems,
Short Papers’. Tyrrell Museum of Palaeontology
Occasional Papers 3.

BONAPARTE, J.F., NOVAS, FE. & CORIA, R.A. 1990.
Carnotaurus sastrei Bonaparte, the horned, light-
ly built carnosaur from the middle Cretaceous of
Patagonia. Contributions in Science, Natural His-
tory Museum of Los Angeles County 416: 1-42.

CAZAU, L.B. & ULIANA, M.A. 1973. El Cretácico
Superior continental de Ja Cuenca Neuquina.
Quinto Congreso Geológico Argentino, Buenos
Aires, Actas 3: 131-163.

CHIAPPE, L.M. 1992. Osteología y sistemática de
Patagopteryx deferrariisi Alvarenga y Bonaparte
(Aves) del Cretácico de Patagonia. Filogenia e
historia biogeográfica de las aves Cretácicas de
América del Sur. Tesis Doctoral, Facultad de Cien-
cias Exactas y Naturales, Universidad de Buenos
Aires, 429 pp.

1995a. The phylogenetic position of the Cretaceous
birds of Argentina: Enantiornithes and Patagop-
teryx deferrariisi. Courier Forschungsintitut
Seckenberg 181: 55-63.

1995b. A diversity of early birds. Natural History 6:
52-55.

1995c. The first 85 million years of avian evolution.
Nature, 378; 349-355.

1996. Late Cretaceous birds of southem South
America: anatomy and systematics of Enanthior-
nithes and Patagopteryx deferrariisi. Pp. 203-
244. In Arratia, G. (ed.) 'Contributions of
Southern South America to Vertebrate
Paleontology'. Münchener Geowis-
senschaftliche Abhandlungen 30. (Verlag Dr
Pfeil: München).

CHIAPPE, L.M. & CALVO, J. 1994. Neuquenornis
volans, a new Enanthiornithes (Aves) from the
Upper Cretaceous of Patagonia (Argentina). Jour-
nal of Vertebrate Paleontology 14: 230-246.

CHIAPPE, L.M., NORRELL, M. & CLARK, J. 1996,
this volume. Phylogenetic position of Mononykus

698

from the Upper Cretaceous of the Gobi desert.
Memoirs of the Queensland Museum39(3): 557-582.

CLARK, G, 1993. Termini situm et directionem partium
corporis indicates. Pp. 1-6. In Baumel, J., King,
A., Breazile, J., Evans, H. & Vanden Berge, J.
(eds) Handbook of avian anatomy: Nomina
anatomica avium. Publications of the Nuttall Or-
nithological Club 23.

COLBERT, E.H. 1989, The Triassic dinosaur
Coelophysis. Museum of Northern Arizona Bul-
letin 57: 1-160.

CRUZ, C.E., CONDAT, P., KOZLOWSKI, E. & MAN-
CEDA, R. 1989. Anilisis estratigrafico secuencial
del Grupo Neuquén (Cretácico superior) en el
valle del Río Grande, Provincia de Mendoza. I
Congreso Argentino de Hidrocarburos, Actas 2:
689-714.

CURRIE, P.J. & EBERTH, D.A., 1993. Palaeontology,
sedimentology and palaeoecology of the Iren
Dabasu Formation (Upper Cretaceous), Inner
Mongolia. People’s Republic of China.
Cretaceous Research 14: 127-144.

DIGREGORIO, J.H. & ULIANA, M.A. 1980. Cuenca
Neuquina. Academia Nacional de Ciencias de
Córdoba, Argentina. Geología Regional Argen-
tina (2): 985-1032.

FARRIS, J.S. 1988. Hennig86, version 1.5. (Computer
program distributed by its author,41 Admiral St,
Port Jefferson Station, NY 11776).

GALTON, P.M. & JENSEN, J.A. 1979. A new large
theropod dinosaur from the Upper Jurassic of
Colorado. Brigham Young University, Geological
Studies 26: 1-12.

GAUTHIER, J.A. 1986. Saurischian monophyly and
the origin of birds. Pp. 1-55. In Padian, K. (ed.)
‘The Origin of Birds and the Evolution of Flight’.
California Academy of Sciences, Memoir 8.

GILMORE, C.W. 1920. Osteology of the carnivorous
Dinosauria in the United States National Museum,
with special reference to the genera Antrodemus
(Allosaurus) and Ceratosaurus. United States Na-
tional Museum, Bulletin 110: 1-159.

GIMÉNEZ, O. 1993. Estudio preliminar del miembro
anterior de los saurópodos titanosáuridos.
Ameghiniana 30: 154.

GRADZINSKY, R., KIELAN-JAWAROWSKA, Z. &
MARYANSKA, T. 1977. Upper Cretaceous
Djadokhta, Barun Goyot and Nemegt formations
of Mongolia, including remarks on previous sub-
divisions. Acta Geologica Polonica 27: 281-318.

HAQ, B.U. & VAN EYSINGA, E.W.B. 1994. ‘Geologi-
cal time table'. Fourth edition. (Elsevier: Amster-
dam).

HILDEBRAND, M. 1975. ‘Analysis of vertebrate
structure’. (John Wiley & Sons: New York) 710 pp.

HOLTZ, T. Jr 1994, The phylogenetic position of the
Tyrannosauridae: implications for theropod sys-
tematics, Journal of Paleontology, 68: 1100-1117.

1995. The arctometatarsalian pes, an unusual struc-
ture of the metatarsus of Cretaceous Theropoda

MEMOIRS OF THE QUEENSLAND MUSEUM

(Dinosauria: Saurischia). Journal of Vertebrate
Paleontology, 14: 480-519.

JERZYKIEWICZ, T. & RUSSELL. D.A. 1991. Late
Mesozoic stratigraphy and vertebrates of the Gobi
Basin. Cretaceous Research 12: 345-377.

LAMBE, L.M. 1917. The Cretaceous theropodous
dinosaur Gorgosaurus. Memoirs of the Geologi-
cal Survey of Canada 100: 1-84.

LEGARRETA, L. & GULISANO, C. 1989. Análisis
estratigráfico secuencial de la Cuenca Neuquina
(Triásico Superior-Terciario Inferior). Pp. 221-
243. In Chebli, G. & Spalletti, L. (eds) 7 Cuencas
Sedimentarias Argentinas. Serie de Correlación
Geológica, 6. (Universidad Nacional de Tucumán:
Tucumán).

LILLEGRAVEN, J.A., KRAUS, MJ. & BOWN, T.M.
1979. Paleogeography of the world of the
Mesozoic. Pp. 277-308. In Lillegraven, J.A.,
Kielan-Jawarowska, Z. & Clemens, W.A. (eds)
‘Mesozoic Mammals. The first two-thirds of
mammalian history'. (University of California
Press: Berkeley).

MADSEN, J.H. 1976. Allosaurus fragilis: a revised
osteology. Utah Geological and Mineral Survey
Bulletin 109: 1-163.

MARSH, O.C. 1880. ‘Odontornithes: a monograph on
the extinct toothed birds of North America’. (Govt
Printing Office: Washington) 201 pp.

1892, Notice of new reptiles from the Laramie For-
mation. American Journal of Science (ser. 3) 43:
449-453.

1896. "The dinosaurs of North America'. Sixteenth
Annual Report of the U.S. Geological Survey,
Washington, 414 pp.

MARTIN, L.D. 1983. The origin and early radiation of
birds. Pp. 291-338. In Bush, A.H. & Clark, G.A.
Jr (eds) ‘Perspectives in ornithology’. (Cambridge
University Press: New York).

MARTIN, L.D. & TATE, J., Jr 1976. The skeleton of
Baptornis advenus (Aves: Hesperornithiformes).
In Olson, S.L. (ed.) 'Collected papers in avian
paleontology honoring the 90th birthday of
Alexander Wetmore'. Smithsonian Contributions
to Paleobiology 27: 35-66.

MARTIN, L.D. & RINALDI, C. 1994, How to tell a
bird from a dinosaur. Maps Digest 17(4): 190-196.

McGOWAN, C. 1985. Tarsal development in birds:
evidence for homology with the theropod condi-
tion. Journal of Zoology (A) 206: 53-67.

MOLNAR, R.E., KURZANOV, S.M. & DONG Z.
1990. Carnosauria. Pp. 169-209. In Weishampel,
D.B., Dodson, P. & Osmólska, H. (eds) "The
Dinosauria'. (University of California Press:
Berkeley).

NICHOLS, E.L. & RUSSELL, A.P. 1985. Structure and
function of the pectoral girdle and forelimb of
Struthiomimus altus (Theropoda: Or-
nithomimidae). Palaeontology 28: 638-667.

NOVAS, FE. 1991. Los Tyrannosauridae, gigantescos
dinosaurios celurosaurios del Cretácico Tardío de
Laurasia. Ameghiniana 28; 410 (Abstract).

CRETACEOUS MANIRAPTORAN THEROPODS

1992. La evolución de los dinosaurios carnívoros.
Pp. 125-163. In Sanz, J.L. & Buscalioni, A. (eds)
‘Los dinosaurios y su entorno biótico'. Actas II
Curso de Paleontologia en Cuenca, Instituto ‘Juan
de Valdés’, Ayuntamiento de Cuenca, España.

1994. New information on the systematics and
postcranial skeleton of Herrerasaurus is-
chigualastensis (Theropoda: Herrerasauridae)
from the Ischigualasto Formation (Upper Trias-
sic) of Argentina. Journal of Vertebrate Paleon-
tology, 13: 400-423.

In press a. Anatomy of Patagonykus puertai
(Theropoda, Avialae, Alvarezsauridae), from the
Late Cretaceous of Patagonia. Journal of Ver-
tebrate Paleontology.

In press b. Dinosaur monophyly. Journal of Ver-
tebrate Paleontology.

NOVAS, F.E. & CORIA, R.A. 1990a. Unusual
Cretaceous theropods from Patagonia. Journal of
Vertebrate Paleontology 10 (suppl. 3): 143A.

NOVAS, F.E. & CORIA, R.A. 1990b. Nuevos
dinosaurios terópodos del Cretácico Superior de
Patagonia. Ameghiniana 26: 247 (Abstract).

OSBORN, H.F. 1905. Tyrannosaurus, Upper
Cretaceous carnivorous dinosaur (second com-
munication). Bulletin of the American Museum of
Natural History 22: 281-296.

1917. Skeletal adaptations of Ornitholestes, Struthio-
mimus, Tyrannosaurus. Bulletin of the American
Museum of Natural History 35: 733-771.

OSMOLSKA, H. 1980. The Late Cretaceous vertebrate
assemblages of the Gobi Desert, Mongolia.
Mémoires de la Société géologique de France 59:
145-150.

OSMOLSKA, H., RONIEWICZ, E. & BARSBOLD,
R. 1972. A new dinosaur, Gallimimus bullatus n.
gen., n. sp. (Ornithomimidae) from the upper
Cretaceous of Mongolia. Palaeontologia Polonica
27: 103-143.

OSTROM, J.H. 1969. Osteology of Deinonychus antir-
rhopus, an unusual theropod from the Lower
Cretaceous of Montana. Peabody Museum of
Natural History, Bulletin 30: 1-165.

1974. The pectoral girdle and forelimb function of
Deinonychus antirrhopus (Reptilia: Saurischia):
a correction. Postilla 165: 1-11.

1976a. Archaeopteryx and the origin of birds. Bio-
logical Journal of the Linnean Society 8: 91-182.

1976b. On a new specimen of the lower Cretaceous
theropod dinosaur Deinonychus antirrhopus.
Breviora 439; 1-21.

1978. The osteology of Compsognathus longipes
Wagner. Zitteliana 4: 73-118.

1994. On the origin of birds and of avian flight. In
Prothero, D.R. & Schoch, R.M. (Conv.), ‘Major
Features of Vertebrate Evolution. Short Courses
in Paleontology' 7: 160- 177.

PERLE, A., CHIAPPE L.M., BARSBOLD, R., CLARK,
J.M. & NORRELL, M.A. 1994. Skeletal morphol-
ogy of Mononykus olecranus (Theropoda:

699

Avialae) from the Late Cretaceous of Mongolia.
American Museum Novitates 3105: 1-29.

PERLE, A., NORRELL, M.A., CHIAPPE, L.M. &
CLARK, J.M. 1993. Flightless bird from the
Cretaceous of Mongolia. Nature 362: 623-626.

QUEIRCZ, K. De & GAUTHIER, J.A. 1994. Toward
a phylogenetic system of biological nomenclature.
Trends in Ecology and Evolution, 9: 27-31.

RAATH, M. 1969. A new coelurosaurian dinosaur from
the Forest Sandstone of Rhodesia. Amoldia 4: 1-25.

ROWE, T. 1986, Homology and evolution of the deep
dorsal thigh muscles in birds and other Reptilia.
Journal of Morphology 189: 327-346.

ROWE, T. & GAUTHIER, J.A. 1990. Ceratosauria. Pp.
151-168. In Weishampel, D.B., Dodson, P. &
Osmólska, H. (eds) “The Dinosauria’. (University
of California Press: Berkeley).

RUSSELL, D.A. 1969, A new specimen of
Stenonychosaurus from the Oldman Formation
(Cretaceous) of Alberta. Canadian Journal of
Earth Sciences 6: 595-612.

1993, The role of Central Asia in dinosaurian
biogeography. Canadian Journal of Earth Scien-
ces 30: 2002-2012.

RUSSELL, D.A. & DONG, Z. 1993a. The affinities of
anew theropod from the Alxa Desert, Inner Mon-
golia, People's Republic of China, Canadian Jour-
nal of Earth Sciences 30: 2107-2127.

1993b. A nearly complete skeleton of a new troodon-
tid dinosaur from the Early Cretaceous of the
Ordos Basin, Inner Mongolia, People's Republic
of China. Canadian Journal of Earth Sciences 30:
2163-2173.

SALGADO, L. & CORIA, R. 1993. Consideraciones
sobre las relaciones filogenéticas de Opistho-
coelicaudia skarzynskii (Sauropoda) del Cretácico
superior de Mongolia. Ameghiniana 30: 339,

SANZ, J.L., BONAPARTE, J.F. & LACASA RUIZ, A.
1988. Unusual Early Cretaceous birds from Spain.
Nature 331: 433-435,

SANZ, J.L. & PEREZ-MORENO, B. 1995, Or-
nitomimosaurio hallado en Cuenca. Investigación
y Ciencia 2: 31-33.

SCOTESE, C.R. & GOLONKA, J. 1992. PALEOMAP
Paleogeographic Ailas, PALEOMAP Progress
Report #20, Department of Geology, University
of Texas at Arlington.

SERENO, P. 1994. The pectoral girdle and forelimb of
the basal theropod Herrerasaurus ischigualasten-
sis. Journal of Vertebrate Paleontology 13: 425-
450.

SERENO, P. & RAO, C. 1992. Early evolution of avian
flight and perching: new evidence from Lower
Cretaceous of China. Science 255: 845-848.

WALKER, C.A. 1981. New subclass of birds from the
Cretaceous of South America, Nature 292: 51-53.

WELLES, S. 1984. Dilophosaurus wetherilli
(Dinosauria, Theropoda): osteology and com-
parisons. Palaeontographica 185: 85-180.

WEISHAMPEL, D.B. 1990. Dinosaurian distribution.
Pp. 63-139. In Weishampel, D.B., Dodson, P. &

700

Osmélska, H. (eds) ‘The Dinosauria’. (University
of California Press: Berkeley).

WELLNHOFER, P. 1974. Das fiinfte Skelettexemplar
von Archaeopteryx. Palaeontographica (A) 147:
169-216.

1988. Ein neues Exemplar von Archaeopteryx. Ar-
chaeopteryx, 6: 1-30.

1993. Das siebte Exemplar von Archaeopteryx aus
den Solnhofener Schichten. Archaeopteryx, 11:
1-48.

APPENDIX

TERMINAL TAXA. Character list and taxon-
character-state matrix.

The coding of 34 characters used to diagnose
three of the six maniraptoran terminal taxa are
given below. Some features (e.g., characters M8,
M22, M23) have been incorporated in the list
below, but have emerged as autapomorphies of
some terminal taxa after a cladistic numerical
analysis of ingroup relationships was carried
out. Data from Bonaparte (1991), Perle et al.
(1994), Chiappe et al. (this volume) and Novas
(in press a). Scoring: 0, primitive; 1, derived; ?,
missing or uncertain.

Patagonykus puertai

P1. Dorsal vertebral postzygapophyses. 0. with
lateral margin describing a continuous convex curve in
ventral aspect. J, ventrally sinuous, with a tongue-
shaped lateral projection.

P2. Base of the neural arch in postcervical vertebrae.
0. caudally smooth and flat. 7. with a bulge on both
sides of the neural canal.

P3. Humeral articular facet of coracoid. 0. transver-
sely wide, being nearly as wide transversely as it is
anteroposteriorly. 7. transversely narrow, being as wide
transversely as the surface for the articulation for the
scapula.

P4. Internal tuberosity of humerus. 0. conical,
craniocaudally flattened, or pyramid-shaped. 7. sub-
cylindrical, wider at its extremity rather than in its base.

PS. Humeral entepicondyle. 0. longitudinal
prominence on the anteromedial margin of distal
humerus, and it is almost anteriorly projected. J. coni-
cal-shaped and strongly projected medially.

P6. First phalanx of manual digit I. 0. ventral surface
bounded at both sides by slightly marked ridges on its
proximal portion. 7. with proximomedial hook-like
processes.

P7. Ectocondylar tuber of femur. 0. robust and ellip-
tical-shaped in distal aspect. 7. transversely com-
pressed, rectangular in distal view.

Alvarezsaurus calvoi

Al. Cervical centra. 0. amphiplatyan or opis-
thocoelous. 7. amphicoelous.

A2. Cervical postzygapophyses. 0. rectangular in
dorsal view, not constricted at their bases, and with

MEMOIRS OF THE QUEENSLAND MUSEUM

convex dorsal surface. 7. paddle-shaped in dorsal view,
dorsoventrally flattened, and with a pair of strong
craniocaudal ridges.

A3. Length of distal caudals with respect to proximal
caudals. 0. less than 175%. I. more than 200%.

A4. Scapular blade. 0. craniocaudally wide and dis-
tally expanded. J. slender and reduced.

A5. Ungual phalanx of digit I. 0. with ventral surface
transversely rounded or flattened. 7. ventrally keeled.

Mononykus olecranus

M1. Pleurocoels in cervical vertebrae. 0. present. 1.
absent.

M2. Sulcus caroticus in cervical vertebrae. 0. absent.
I. present.

M3. Relative position of parapophyses in presacral
vertebrae. 0. below the level of the diapophyses. J.
occupying the same level as the diapophyses.

MA. Hyposphene-hypantrum in dorsal vertebrae, 0.
present. 7. absent.

MS. Centra of cranial dorsal vertebrae. 0. transver-
sally rounded. 7. strongly compressed transversally,
forming a pronounced ventral keel.

M6. Centra of caudal dorsal vertebrae. 0. slightly
concave or convex. 1. strongly prococlous, with caudal
articular surface hemispherical.

M7. Last sacral centrum. 0. elliptical or transversely
compressed in cross-section. J. extremely compressed
transversally, forming a pronounced ventral ‘keel’.

M8. Coracoidal shaft. 0. dorsoventral depth not ex-
ceeding its anteroposterior width. 7. expanded ventral-
ly, subrectangular in profile, dorsoventral depth more
than 13056 of anteroposterior width.

M9. Coracoidal shaft. 0. with distal half strongly
inflected medially. 7. transversely flat and
craniocaudally concave.

M10. Sternal carina.0. slender. 1. thick.

M11, Ulnar articular surface of the radius.0. small
and restricted to the proximal portion of the radius, J.
extensive.

M12. Carpal articular facet of radius. 0. restricted to
the distal portion of the bone. 7. hypertrophied, ex-
tended proximally over the caudal and cranial surfaces
of the radius.

M13, Proximocaudal process on the first phalanx of
digit I. 0. absent. 7. very prominent.

MIA. Pubic shaft orientation. 0. almost perpen-
dicular to the iliac surface of pubis. 7. caudoventrally
oriented, describing an angle of 70? with the iliac
surface of pubis.

MIS. Pubic foot. 0. present. J. absent.

M16. Ischiac articular surfaces for ilium and pubis.
0. well developed. 7. extremely reduced.

M17. Femoral distal condyles. 0. separated below
popliteal fossa. 7. transversely expanded, nearly con-
fluent below popliteal fossa.

M18. Accesory (medial) cnemial crest on tibia. 0.
absent. 7. present.

M19. Outer malleolus of distal tibia. 0. craniocaudal-
ly narrow in respect to the calcaneum. 7. craniocaudally
thick.

CRETACEOUS MANIRAPTORAN THEROPODS

M20. Astragalocalcaneal intercondylar groove. 0.
shallow. 7. deep.

M21. Astragalar ascending process. 0. transversally
wide and not displaced laterally, without overlap onto
the lateral margin of distal tibia. 7. transversally narrow
and laterally displaced, with overlap onto the lateral
margin of distal tibia.

M22. Femoral trochanteric crest. 0. anterior
trochanter separated from greater trochanter by a cleft.
1. undivided.

M23. Fibula. 0. articulates with the tarsus. J. does
not articulates with the tarsus.

INGROUP CLADES. Character list and taxon-
character-state matrix.

The coding and distribution are shown below
for 74 characters in six coelurosaur taxa and in
two proximate outgroups. All characters are bi-
nary. Characters gathered from Gauthier (1986),
Russell & Dong (1993a,b), Chiappe (1995, 1996),
Chiappe et al. (this volume) and Novas (in prep.).

1. Caudal articular surface of the centra of the last
sacral and first caudal vertebrae. 0. slightly convex. J.
strongly spherical.

2. Sacral vertebral centra. 0. transversely rounded or
craniocaudally grooved. 7. strongly keeled ventrally.

3. Femoral fourth trochanter. 0. present. J. absent.

4. Supracetabular crest. 0. present. J. absent.

5. Caudal dorsal vertebrae. 0. amphiplatyan or am-
phicoelous. J. procoelous.

6. Bicipital tubercle of coracoid. 0. present. J. absent.

7. Forelimbs to hindlimb length. 0. 40-53%. 1. less
than 26%.

8. Proximal humerus. 0. major transverse axis of
humeral head horizontally oriented with respect to
longitudinal axis of the humerus, and internal
tuberosity distally placed with respect to humeral head.
1. major transverse axis of humeral head ventrolateral-
ly inclined with respect to longitudinal axis of the
humerus, and internal tuberosity proximally projected.

9. Radial and ulnar condyles of distal humerus. 0.
Separated by an intercondylar groove, and radius and
ulna loosely articulated proximally, retaining inde-
pendent articular surfaces for both radial and ulnar
condyles of humerus, respectively. J. a single condyle
on distal humerus for articulation with radius and ulna,
which are tightly appressed proximally, and provided
with a cup-shaped proximal articular surface.

10. Olecranal process of ulna. 0. feebly developed.
1. strongly developed.

11. Ulnar posterior margin. 0. sigmoid. 7. uniformly
convex.

12. Carpometacarpus. 0 slender, elongate, with in-
termetacarpal space. J. massive, short, quadrangular
with no intermetacarpal space.

13. Digit I proportions. 0. transverse dimension sub-
equal to digit II, and longitudinally shorter than digit
II. 7. digit I larger than the remaining digits of the
hands.

14. Phalanx 1 of manual digit I. 0. with triangular-
shaped proximal articular surface, proximoventral sur-

701

face almost flat, distal gynglymus transversely com-
pressed, and the extensor pit absent. 7. with B-shaped
proximal articular surface, hook-like proximomedial
processes, symmetrical distal ginglymus, and deep ex-
tensor ligamentary pit.

15. Medial condyle of femur. 0. transversely narrow
and distally convex. J. transversely wide and distally
flat.

16. Ectocondylar tuber of distal femur caudally
projected, well behind the medial distal condyle. 0.
absent. 7. present.

17. Fibular condyle of femur. 0. convex or forming
a depressed low cone, and slightly surpassing distally
the medial condyle. 7. sharply conical and distally
projected respect to the medial condyle.

18. Cervical vertebrae neural spines. 0. dorsoventral-
ly deep and craniocaudally short. 7. craniocaudally
short and dorsoventrally low.

19. Sacral and caudal vertebrae. 0. amphiplatyan or
amphicoelous. 7. procoelous.

20. Last sacral centra. 0. ventrally convex in cross-
section, sometimes bearing a longitudinal groove. 7.
transversely compressed and keeled ventrally.

21. Haemal arches of proximal caudals. 0. dor-
soventrally elongate. 7. dorsoventrally depressed.

22. Ungual phalanx of manual digit I. 0. dor-
soventrally deep, with proximal articular surface ellip-
tical shaped. 7. stout and robust, dorsoventrally depressed,
with proximal articular surface quadrangular.

23. Pubic pedicle of ilium. 0. craniocaudally wide. J.
craniocaudally narrow.

24. Pubic pedicle of ilium. 0. cranioventrally
projected. 7. caudoventraly projected.

25. Fossa for M. cuppedicus on ilium. 0. transversely
wide, with sharp bounding margins. 7. absent.

26. Supraacetabular crest. 0. extended cranially in
continuity with the lateral border of the pubic pedicle.
1. ending cranially above the pubic pedicle.

27. Postacetabular blade. 0. brevis shelf
caudolaterally oriented, and medial flange ventrally
curved (viz., brevis fossa present). 7. postacetabular
blade vertical, and medial flange strongly reduced,
perpendicular to iliac blade (viz., brevis fossa absent).

28. Posterior trochanter on proximal femur. 0. ab-
sent. 7. present.

29. Vertebral foramen. 0. small. 7. wide.

30, Number of caudals. 0. 35 or more. 7. less than 25.

31. Neural spines of caudal vertebrae. 0. present on
caudals 1-23. Z. confined to caudals 1-12.

32. Transverse processes in caudal vertebrae. 0.
reduction begins in caudal 25-16. 7. in caudal 12 at least.

33. Haemal arches. 0. become longer than deep
behind caudal 17. 7. behind caudal 10.

34. Mid-caudals prezygapophyses. 0. elongate. 7. short.

35. Length of distal caudals. 0. as long as the
proximal caudals. 7. distal caudals more than 130% of
the length of proximals.

36. Ossified sternal carina. 0. absent. 7. present.

37(=M8). Coracoidal shaft. 0. dorsoventral depth not
exceeding its anteroposterior width. 7. expanded

702

ventrally, subrectangular in profile, dorsoventral depth
more than 130% of anteroposterior width.

38. Coracoidal shaft with respect to the proximal
articular surface for the scapula. 0. shaft transversely
flattened or slightly medially inflected in posterior
view. 1. strongly inflected posteromedially.

39. Dorsal fossa on caudal process of coracoid. 0.
present. 7. absent.

40. Bicipital tubercle of coracoid. 0. slightly marked
or absent. J. strongly developed.

41. Forelimbs. 0. no more than 53% of hindlimb
length. 7. exceeding 7596 of hindlimb length.

42. Forelimbs. 0. not exceeding 75% of hindlimb
length. 7. 86% or more of hindlimb length.

43. Shape of the internal tuberosity (— bicipital
tubercle) of humerus. 0. conical. 7. craniocaudally
compressed and longitudinally elongated.

44. Radius/humerus length ratio. 0. no more than
0.71. 1. 0.76 or more.

45. Carpometacarpus. 0. carpals and metacarpals
unfused. J. carpometacarpus present.

46. Distal carpal. 0. proximodistally flattened. J.
semilunate.

47. Digit I proportions. 0. digit I ends at level of
mid-length of phalanx 2, digit II. 7. digit I ends at level
of mid-length of phalanx 1, digit II.

48. Posterodorsal margin of ilium. 0. straight. J.
curves ventrally in lateral view.

49. Posterior end of ilium. 0. dorsoventrally deep,
squared or rounded. J. dorsoventrally low and sharply
pointed.

50. Antitrochanter on ilium. 0. slightly marked. J.
prominent.

51. Pubic apron. 0. transversely wide and
proximodistally long. 7. strongly reduced transversely
and restricted to the distal 1/3 of the pubic length.

52. Pubis. 0. cranioventrally oriented. 7. caudo-
ventrally oriented respect to the pubic pedicle of ilium.

53. Pubic apices. 0. in contact. J. not in contact.

54. Obturator process on ischium. 0. present. 1.
absent (the cranioventral margin of ischium is almost
straight).

55. Ischial to pubic length. 0. elongate ischium,
slightly shorter than pubis or femur. J. nearly half of
pubis or femoral length.

56. Medial condyle of femur. 0. dorsoventrally deep.
1. dorsoventrally depressed.

57(=M22). Femoral trochanteric crest. 0. anterior
trochanter separated from greater trochanter by a cleft.
1. undivided.

58. Adductor fossa and associated craniodistal crista
of distal femur. 0. present, prominent. 7. reduced or
absent.

59. Tibia-femur proportions. 0. tibia no more than
15% longer than the femur. 1. tibia 25% longer than the
femur.

60(=M23). Fibula. 0. articulates with the tarsus. 7.
does not articulate with the tarsus.

61. Articulations of quadrate and squamosal. 0. quad-
rate articulates only with squamosal, the latter contact-
ing both the quadratojugal and the postorbital. 7.

MEMOIRS OF THE QUEENSLAND MUSEUM

quadrate articulates with both prootic and squamosal,
and the latter contacting neither the quadratojugal nor
the postorbital.

62. Serration of teeth. 0. present. 1. absent.

63. Teeth crown-root constriction. 0. absent. 1.
present.

64. Ulnar distal condyle. 0. transversely compressed
and craniocaudally extended approximately in the
same plane as humero-ulnar flexion-extension. J. sub-
triangular in distal view, with a dorsomedial condyle,
and twisted more than 54? with respect to proximal end.

65. Calcaneum. 0. posteroventral projection present.
1. strongly reduced.

66. Pubic distal foot. 0. cranial projection present. 7.
absent.

67. Prominent ventral processes of cervico-dorsal
vertebrae. 0. absent. I. present.

68. Large, rectangular ossified sternum. 0. absent. 1.
present.

69. Terminal processes of ischia. 0. in contact. J. not
in contact.

70. Fibular tubercle for M. iliofibularis. 0.
craniolaterally projected. 7. laterally, caudolaterally, or
caudally directed.

71. Proximal and distal humeral ends. 0. twisted. 1.
expanded in nearly the same plane.

72. Pelvic elements. 0. unfused. 7. fused or partially
fused.

73. Proximal end of fibula. 0. excavated by a medial
fossa. I. nearly flat.

74(=M4). Hyposphene-hypantrum in dorsal ver-
tebrae. 0. present. 1. absent.

TAXON CHARACTER STATES.
Ornithomimidae

00000 00000 00000 00100 00000 00000 00000 00000
00000 00000 00000 00000 00000 00000 0000
Tyrannosauridae

00000 00000 00000 00000 00000 00000 00000 00070
00000 00000 00000 00000 00000 00000 0000
Deinonychus

00010 00000 10000 00000 10010 01100 11101 01000
10110 10100 01001 10100 00001 00000 0000
Archaeopteryx

00110 00070 10000 07000 10010 01171 11111 01111
11110 11110 01001 20110 01101 100?? 00??
Ornithothoraces

00110 00000 10000 00700 10010 01111 11111 11111
11111 11111 11311001111 1111? 11111 1111
Mononykus

11001 11111 01111 11111 01101 10011 11110 10010
00001 10101 11110 11111 11110 11111 1111
Patagonykus

11001 11111 01111 11?11 ?1101 1901? ???10 ??010
20071 1?2?1 110?? 10120 ???10 07??? 7010
Alvarezsaurus

0011? 22272 27277 ??111 011?1 100?? 10111 7202?

EARLIEST EVIDENCE OF DINOSAURS FROM CENTRAL GONDWANA
M.A. RAATH

Raath, M.A. 1996 12 20: The earliest evidence of dinosaurs from central Gondwana.
Memoirs of the Queensland Museum 39(3): 703-709. Brisbane. ISSN 0079-8835.

Rocks of the Elliot Formation and its regional equivalents in southern Africa, which straddle
the Triassic-Jurassic boundary, have traditionally been regarded as the earliest dinosaur-bear-
ing sediments in this region. The identification of tridactyl Grallator-like tracks at a site in
the Eastern Cape Province, South Africa, and the discovery of an isolated fragment of a
prosauropod femur in the central Zambezi Valley of Zimbabwe, both in rocks assigned to
equivalents of the underlying Molteno Formation on stratigraphic and palaeobotanical
grounds, indicate the presence of an earlier dinosaurian fauna. These traces constitute the
earliest known evidence of dinosaurs in this region of Central Gondwana. [ ] Triassic,
dinosaur tracks, rhynchosaur, prosauropod, South Africa, Zimbabwe, Molteno.

M.A. Raath, Port Elizabeth Museum, Port Elizabeth, South Africa. (Current address:
Bernard Price Institute for Palaeontological Research, University of the Witwatersrand,

Johannesburg, South Africa); 5 April 1996.

Two fossil discoveries during the last decade in
two far-flung localities in southern Africa —
Maclear in the north-eastern Eastern Cape
Province of South Africa, and the Cabora Bassa
Basin of the Central Zambezi Valley of Zim-
babwe (Fig. 1) — have yielded evidence of a
dinosaur fauna predating that of the Elliot Forma-
tion of the Stormberg Group in South Africa
(Raath, Kitching, Shone & Rossouw, 1990;
Raath, Oesterlen & Kitching, 1992).

The Elliot Formation and its regional
equivalents have generally been regarded as the
earliest dinosaur-bearing formations in this part
of Africa. According to Olsen & Galton (1984)
the Elliot spans the Triassic-Jurassic boundary
(Carnian to Sinemurian). On the basis of its fossil
content in South Africa, Kitching & Raath (1984)
recognised two distinct biozones in the Elliot
Formation: a lower Euskelosaurus Range Zone
characterised by the presence of the relatively
large prosauropod dinosaur Euskelosaurus, ac-
companied by capitosaurid amphibians, uniden-
tified rauisuchid thecodontians and rare cynodont
therapsids. This zone is followed by the Mas-
sospondylus Range Zone, characterised by the
abundance of the smaller prosauropod dinosaur
Massospondylus and a diverse fauna which in-
cludes amphibia, a chelonian, crocodilians,
dinosaurs other than prosauropods, therapsids
and even rare primitive mammals (Kitching &
Raath, 1984).

The discoveries which are the subject of this
contribution have only recently come to the atten-
tion of scientists, although one of them has been
known to the local populace for a long time. From
the early years of this century, trout anglers fish-

ing along the Pot River in the vicinity of the town
of Maclear (Fig. 1) in the Eastern Cape Province
of South Africa have known of the presence of
unusual markings in the rocks lining the river's
banks, which have long been known locally as
*petrified spoor' (animal tracks). Although these
tracks have featured in local newspapers from
time to time, they escaped the notice of geologists
and palaeontologists. As far as I can determine,
ours was the first detailed investigation of the site
(Raath et al., 1990). It confirmed that the marks
included vertebrate tracks and that animals of
probably more than one kind were involved, at
least one of them a dinosaur. It was further con-
cluded that the rocks in which the tracks are
preserved belong to the Molteno Formation
(Raath et al., 1990), which underlies the Elliot
Formation.

Until now, no unequivocal evidence of tetrapod
vertebrates has been found in Molteno rocks —
other than reports by P. Ellenberger of vertebrate
tracks of many kinds in Lesotho in beds which he
assigned to the Molteno, including tracks of a
variety of dinosaurs (see synthesis in Ellenberger,
1970). Persistent doubts about the reliability of
these stratigraphic assignments has meant that
this work has been largely overlooked. The only
unquestionable animal fossils previously
recovered from the Molteno have been some
palaeoniscid fish (Semionotus sp.) and a variety
of invertebrates, mainly insects. Anderson &
Anderson (1984: 40) state in their summary of
Molteno fossils: 'Tetrapods — Bone is not
preserved in the Molteno Formation’ .

The early Mesozoic deposits of Zimbabwe are
palaeontologically very similar to their South

704

African counterparts. In
that country, the main
dinosaur-bearing unit, the
Forest Sandstone Forma-
tion, contains a typical but
rather depauperate ‘ter-
minal Karoo fauna’ which
is dominated by prosauro-
pod and small theropod
dinosaurs, like the South
African middle and upper
Elliot Formation fauna.
However, the Zimbabwean
fauna lacks many of the
other elements represented
in South Africa, particularly
the ornithischian dinosaurs,
advanced therapsids and
early mammals.

As in South Africa, the
Zimbabwean beds which
underlie the traditional
dinosaur-bearing beds have
until now also been con-
sidered barren of tetrapods
(see e.g., Bond, 1973). The
fossils usually associated with these beds in Zim-
babwe include silicified logs (cf. Dadoxylon spp),
particularly from the Pebbly Arkose Formation
and plant leaf impressions (mainly the
pteridosperm genus Dicroidium) and rare inver-
tebrates (freshwater bivalve molluscs) (see sum-
mary in Bond, 1973). On the basis of the
dominance of the characteristic Triassic fossil
plant genus Dicroidium, Bond (1973) concluded
that these beds in Zimbabwe (including the Pebb-
ly Arkose, the Ripple Marked Flags and the Fine
Red Marly Sandstone) were equivalents of the
Molteno Formation of the Karoo Supergroup in
South Africa (late Triassic: Carnian, Anderson &
Anderson, 1984). Some lungfish remains were
recovered from beds of this age in Lake Kariba in
1973, and until the recent tetrapod discoveries in
the Central Zambezi Valley (Raath, Oesterlen &
Kitching, 1992), they were the only vertebrate
remains known from these strata in Zimbabwe.

1975: fig. 5).

SUMMARY OF THE TWO OCCURRENCES

MACLEAR DINOSAUR TRACKS, SOUTH
AFRICA. The most prominent trace in the track-
site at Maclear, Eastem Cape Province, South
Africa, is a straight groove ploughed into the mud
(Fig. 2a), with other less obvious similar traces
nearby. Raath et al. (1990) tentatively interpreted

AFRICA

Dande rhynchosaur
and dinosaur site

MEMOIRS OF THE QUEENSLAND MUSEUM

ANTARCTICA

Maclear dinosaur
tracksile

WoW Lsnv

FIG. 1. Reconstruction of Gondwana prior to separation, showing approximate
positions of the two localities referred to in the text (modified from Colbert,

these grooves as the tail drags of quadrupedal
dinosaurs, but no clear unequivocal footprints can
be associated with them, so this suggestion is
doubtful at best. It has been suggested that a more
likely interpretation is that they are drag marks
from an object floating in the river, although the
remarkably detailed preservation of the rolled
edges of the groove indicate very low velocity of
the current responsible.

On the other hand, the majority of the tracks are
clear and well preserved, consisting of plentiful
tridactyl footprints of a small bipedal dinosaur
(Fig. 2b, c). Similar tracks are well known from
anumber of other places in southern Africa, espe-
cially in Lesotho (Ellenberger, 1970; Olsen &
Galton, 1984) and from one locality in Zimbabwe
(Raath, 1972). Since publication of the report by
Raath et al. (1990), several more reports of
similar well-preserved tridactyl dinosaur
footprints have been received from the farming
districts south and west of Maclear.

Raath et al. (1990) concluded that the tridactyl
tracks are assignable either to the ichnogenus
Grallator (see Olsen & Galton, 1984: fig. 3H) or
Atreipus (Olsen & Baird, 1986: figs 6.3, 6.4). But
Atreipus is defined as ‘habitually quadrupedal
ichnites’ (Olsen & Baird, 1986: 62), and since no
manus impressions are associated with any of the
tridactyl pes impressions at Maclear, it appears
that this ichnogenus can be eliminated. Accord-

EARLIEST EVIDENCE OF DINOSAURS FROM CENTRAL GONDWANA

705

FIG. 2. Maclear dinosaur tracksite. A, the straight ploughed trace — possibly a tail-drag. B, tridactyl prints
arranged in several trackways (note the infilled desiccation cracks on the surface). C, single isolated tridactyl
print, Grallator sp., from a ripple-marked surface. D, frond impression of the pteridosperm genus Dicroidium

from near the tracksite.

ing to Olsen & Galton (1984: 97), Grallator-like
tracks were 'almost certainly made by small
theropod dinosaurs’, which therefore indicates
the presence of small theropods at Maclear when
these sediments were laid down.

RHYNCHOSAURS AND DINOSAUR, ZIM-
BABWE. The rhynchosaur fossils from the
Dande Communal Lands in Zimbabwe consist
mainly of scattered, isolated, often fragmentary
bones and teeth, including the unmistakable
tooth-studded maxillary dental plates of these
highly specialised herbivorous archosauro-
morphs (Fig. 3a). They evidently belong to a

single rhynchosaur taxon, showing a consider-
able range in size.

Rhynchosaur taxa are well-known to be useful
for dating the beds from which they come (Chat-
terjee, 1980; Benton, 1983) and the state of the
diagnostic dental and other cranial characters in
the Zimbabwean rhynchosaurs identifies them as
Hyperodapedon, of the family Rhynchosauridae,
sub-family Hyperodapedontinae (Raath et al.,
1992). This has significance for dating the strata
from which they come.

The single fragmentary non-rhynchosaur bone
found with rhynchosaur material at one of the
Dande sites bears the characteristic crest-like
fourth trochanter of a prosauropod dinosaur

706

femur (Fig. 3b, c). The fragment is broken off just
below the trochanter and most of the head is
missing, but enough remains to be reasonably
confident of its identification as a prosauropod.
Although it is small compared to most of the
known southern African prosauropods, such as
Euskelosaurus and Massospondylus, the
specimen is taken to be from an adult based on
the degree of ossification of scars of minor sur-
face features like the vascular groove running
along the anterior surface to the nutritive foramen
opposite the fourth trochanter (Fig. 3c).

If this identification is correct, it demonstrates
the presence of prosauropod dinosaurs when the
Dande beds were laid down.

GEOLOGY AND AGE OF THE TWO
OCCURRENCES

MACLEAR TRACKSITE. The Maclear tracks
are preserved in a medium-to-finegrained cross-
bedded silty sandstone belonging to the
Stormberg Group, from somewhere near its base.
The Stormberg succession consists of the thick,
coarse fluvial sandstones and finer shales of the
Molteno Formation at its base, overlain by the
brick-red mudrocks and sandstones of the Elliot
Formation, which in turn are overlain by the pale,
finer-grained, mainly aeolian sandstones of the
Clarens Formation. The thick Drakensberg lavas
cap the sequence (Fig. 4). This entire sequence is
well exposed in the Barkly Pass — a road pass
through the Drakensberg mountains just 50km
west of the Maclear tracksite. Visser (1984) notes
that in this area, the Molteno Formation makes up
the valley floor at the base of the Drakensberg
mountains, The mountains themselves are made
up of Elliot Formation mudrocks at the base, with
the pale sandstones of the Clarens Formation
above, and capping the sequence on the mountain
peaks are the Drakensberg lavas. Visser (cited in
Raath et al., 1990) locates the Molteno-Elliot
contact ‘right at the foot of the (Barkly) pass in
the Tsomo River and this position agrees with the
boundary defined by (A.L.) du Toit on his map
(?1910) of the area.’.

The Pot River has cut its deep valley through
the flat plains at the foot of the mountains — i.e.,
through Molteno deposits. The beds through
which it has cut might be described as ‘uppermost
Molteno’, and one coarse sandstone ledge from
this unit overlooks the fossil tracksite about 100m
below. This sandstone is lithologically reminis-
cent of the Indwe Sandstone Member, a persistent
and highly characteristic marker horizon in the

MEMOIRS OF THE QUEENSLAND MUSEUM

FIG. 3. Zimbabwean rhynchosaur locality. A, snout of
a juvenile rhynchosaur, Hyperodapedon, in palatal
view, showing the characteristic maxillary tooth-
plates. B, fragmentary prosauropod left femur in
lateral view; note crest-like fourth trochanter (ar-
rowed). C, anterior surface of the femur in (B) show-
ing the vascular groove (arrowed) leading to the
nutritive foramen. Scales = lcm.

Molteno Formation, all of which suggests that the
fossil tracks are in Molteno beds, substantially
below the contact with the Elliot Formation.
Moreover, within a kilometre of the tracksite,
and on a stratigraphic level either the same as or

EARLIEST EVIDENCE OF DINOSAURS FROM CENTRAL GONDWANA 707

slightly above it, a finegrained
shale outcrop contains abun-

FORMATION DIVISION BIOZONE

dant plant fossils dominated by Drakensberg [Lavas]
the characteristic Molteno Formation

pteridosperm genus,
Dicroidium (Fig. 2d).

No bones of dinosaurs are
known from Molteno beds in
South Africa, but they are cer-
tainly known from the higher
levels of the Stormberg Group
and its equivalents in many
southern African localities (see
Raath, 1980; Cooper, 1981;
Kitching & Raath, 1984 and
references therein); in fact,

STORMBERG

GROUP

Fu | —— ——]
Clarens Formation Middle ENDE

Lower

Massospondylus
U -
per Range-zone

Middle

Elliot Formation
Euskelosaurus
Lower Range-zone
ée caedi
Molteno Formation Upper New "pre-Elliot"
dinosaur fauna

there are several good bone-
producing sites in the Barkly
Pass, within 50km of the
Maclear tracksite.

BEAUFORT
GROUP

The apparently consistent FIG. 4. Summary of Stormberg stratigraphy in South Africa.

stratigraphic distribution of

dinosaurs in the Stormberg led Kitching & Raath
(1984) to propose the two biozones mentioned
earlier — the Euskelosaurus Range Zone in the
lower parts of the Elliot Formation, and the Mas-
sospondylus Range Zone in the middle and upper
parts, extending into the base of the overlying
Clarens Formation.

A complicating factor in the Maclear case is
that the only dinosaurs so far known from the
Euskelosaurus Range Zone are prosauropods —
no theropods have been discovered to date. How-
ever, at least one theropod taxon, Syntarsus, is
relatively common in the overlying Massospon-
dylus Range Zone (Raath, 1980; Kitching &
Raath, 1984). Tracks attributed to Syntarsus are
known from Zimbabwe in beds which correlate
with the upper Elliot Formation (Raath, 1972). In
terms of general morphology, the Zimbabwe
?Syntarsus tracks are similar to the Maclear
tridactyl prints, but Syntarsus is too small to have
made the Maclear tracks. The identity of the
track-maker therefore remains unresolved.

The conclusion that these tracks are in Molteno
sediments also demands a reassessment of the
validity of Ellenberger's (1970) stratigraphic as-
signments.

ZIMBABWEAN RHYNCHOSAUR AND
DINOSAUR SITE. The rhynchosaur specimens
and the prosauropod from Zimbabwe were dis-
covered during mapping by P.M. Oesterlen of the
Zimbabwean Geological Survey, in the Dande
Communal Lands in the Lower Zambezi Valley

during the late 1980s (Raath et al., 1992). The
rocks exposed here range in age from late
Palaeozoic to early Mesozoic (Broderick, 1984;
1989; Oesterlen, 1989) and the rhynchosaur
material came from beds which Oesterlen
mapped as Pebbly Arkose Formation (Fig. 5).
This formation is of late Triassic age and is con-
sidered an equivalent of the South African Mol-
teno Formation (Bond, 1973).

Chatterjee (1969; 1980) and Benton (1983)
point out that rhynchosaurs fall into three family
or subfamily groups whose stratigraphic distribu-
tion neatly coincides with the division of the
Triassic into early, middle and late divisions. The
subfamily Hyperodapedontinae is characteristic
of, and apparently confined to, the late Triassic.
Benton (1983) includes in this subfamily the East
African form 'Supradapedon', which comes
from the late Triassic beds of Tunduru, Tanzania,
and is very similar to the Zimbabwean form. On
these grounds, the Zimbabwean rhynchosaurs are
assigned to the late Triassic, and, by association,
so is the dinosaur found with them.

A late Triassic (Molteno-equivalent) age for the
Zimbabwean occurrences is further supported by
the fact that, as in the South African case, the
typical late Triassic plant genus Dicroidium is
found in the vicinity of the Dande rhynchosaur
localities, although so far not in direct
stratigraphic association with the vertebrates.

It is therefore concluded that on stratigraphic
and palaeontological grounds, the Maclear
(South Africa) and Dande (Zimbabwe) sites show

708

MEMOIRS OF THE QUEENSLAND MUSEUM

Formations in the

Period Group

| Cretaceous |

Jurassic

Upper
Karoo

Mesozoic

Triassic

Cabora Bassa Basin,

Dande Sandstone

Forest Sandstone

Pebbly Arkose
(from which the rhynchosaurs
and the dinosaur bone come)

South African
Equivalents

[Post-Karoo]

Zimbabwe

Stormberg Group

Angwa Sandstone

FIG. 5. Stratigraphy of Stormberg equivalents in Zimbabwe (adapted from Raath, Oesterlen & Kitching, 1992).

the presence of dinosaurs during Molteno times
(?lower Carnian, Anderson & Anderson, 1984:
40) and, subject to a reassessment of Ellen-
berger’s (1970) stratigraphic assignments, these
are therefore the earliest indications of dinosaurs
in this part of Central Gondwana.

ACKNOWLEDGEMENTS

I am grateful to the organisers of the Sym-
posium on Gondwanan Dinosaurs, especially
Fernando Novas and Ruben Cuneo for their in-
vitation to take part and for their assistance which
made it possible. Similarly I thank the Board of
Trustees of the Port Elizabeth Museum, especial-
ly the Chairman, John Grieve, for releasing me to
attend and for generous support of my trip. Final-
ly I thank my colleague Nielen Schaefer and my
friends and collaborators on the two projects dis-
cussed here, notably James Kitching, Phil Oester-
len, John Orpen and Gerald Spilkin for the many
ways in which they helped.

REFERENCES

ANDERSON, J.M. & ANDERSON, H.M. 1984. The
fossil content of the Upper Triassic Molteno For-
mation, South Africa. Palaeontologia africana, 25:
39-59.

BENTON, M.J. 1983. The Triassic reptile Hyper-
odapedon from Elgin: functional morphology and
relationships. Philosophical Transactions of the
Royal Society, London, B 302: 605-717.

BOND, G. 1973. The Palaeontology of Rhodesia. Bul-
letin of the Geological Survey of Rhodesia, 70:
1-121.

BRODERICK, T.J. 1984. A geological interpretation
across a portion of the mid-Zambezi Valley lying
between the Mkanga and Hunyani rivers, Guruve

District. Annals of the Zimbabwe Geological Sur-
vey, 9: 59-79.

1989. An interpretation of the geology of the Cabora
Bassa Basin, mid-Zambezi Valley. Annals of the
Zimbabwe Geological Survey, 14: 1-11.

CHATTERJEE, S. 1969. Rhynchosaurs in time and
space. Proceedings of the Geological Society,
London, 1658: 203-208.

1980. The evolution of the rhynchosaurs. Memoires
de la Societe geologique de France, N.S. 139:
57-65.

COLBERT, E.H. 1975. Early Triassic tetrapods and
Gondwanaland. Memoires du Museum National
d’Histoire Naturelle, N.S. A 88: 202-215.

COOPER, M.R. 1981. The prosauropod dinosaur Mas-
sospondylus carinatus from Zimbabwe: its biol-
ogy, mode of life and phylogenetic significance.
Occasional Papers of the National Museums of
Rhodesia, B., Natural Sciences 6(10): 689-840.

ELLENBERGER, P. 1970. Les niveaux paleontologi-
ques de premiere apparition des Mammiferes
primordiaux en Afrique du Sud et leur ichnologie:
Etablissement de zones stratigraphiques detaillees
dans le Stormberg du Lesotho (Afrique du Sud)
(Trias superieur a Jurassique). Pp. 343-370. In
Haughton, S.H. (ed.) 'IUGS, 2nd Symposium on
Gondwana Stratigraphy and Palaeontology’.
(Council for Scientific and Industrial Research:
Pretoria).

KITCHING, J.W. & RAATH, M.A. 1984. Fossils from
the Elliot and Clarens formations (Karoo Se-
quence) of the northeastern Cape, Orange Free
State and Lesotho, and a suggested biozonation
based on tetrapods. Palaeontologia africana 25:
111-125.

OESTERLEN, P.M. 1989. The geology of the Dande
West area (Western Cabora Bassa Basin) — a
preliminary report. Annals of the Zimbabwe
Geological Survey 14: 12-20.

OLSEN, P.E. & BAIRD, D. 1986. The ichnogenus
Atreipus and its significance for Triassic

EARLIEST EVIDENCE OF DINOSAURS FROM CENTRAL GONDWANA

biostratigraphy. Pp. 61-87. In Padian, K. (ed.) “The
Beginning of the Age of Dinosaurs’. (Cambridge
University Press: Cambridge).

OLSEN, P.E. & GALTON, P.M. 1984. A review of the
reptile and amphibian assemblages of the
Stormberg of Southern Africa, with special em-
phasis on the footprints and the age of the
Stormberg. Palaeontologia africana 25: 87-110.

RAATH, M.A. 1972. First record of dinosaur footprints
from Rhodesia. Arnoldia (Rhodesia) 5(27): 1-5.

1980. The theropod dinosaur Syntarsus (Saurischia:
Podokesauridae) discovered in South Africa.
South African Journal of Science 76(8): 375-376.

709

RAATH, M.A., KITCHING, J.W., SHONE, R.W. &
ROSSOUW, G.J. 1990. Dinosaur tracks in Trias-
sic Molteno sediments: the earliest evidence of
dinosaurs in South Africa? Palaeontologia
africana 27: 89-95.

RAATH, M.A., OESTERLEN, P.M. & KITCHING,
J.W. 1992. First record of Triassic Rhynchosauria
(Reptilia: Diapsida) from the Lower Zambezi Val-
ley, Zimbabwe. Palaeontologia africana 29: 1-10.

VISSER, J.N.J. 1984. A review of the Stormberg Group
and Drakensberg Volcanics in southern Africa.
Palaeontologia africana 25: 5-27.

SIGNIFICANCE OF POLAR DINOSAURS IN GONDWANA
THOMAS H. RICH

Rich, T.H. 1996 12 20: Significance of polar dinosaurs in Gondwana. Memoirs of the
Queensland Museum 39(3) 711-717. Brisbane. ISSN 0079-8835.

Polar dinosaurs have been found at four localities in the southern hemisphere and eight in
the northern. Three groups of dinosaurs, neoceratopsians, ornithomimosaurs and ovirap-
torosaurs, previously known only from the Late Cretaceous of the northern hemisphere were
present in the Early Cretaceous polar dinosaur fauna of southeastern Australia. Labyrin-
thodont amphibians were also represent, but do not occur in the same deposits as crocodilians.
Enlarged optic lobes of the hypsilophodontid Leaellynasaura, suggesting enhanced ability
to see under low light conditions, is the only adaptation to a polar environment yet recognised
in a dinosaur. Polar dinosaurian taxa are not unique at the familial level, but belong to more
widespread families. Polar dinosaurs probably lived under cold climates quite unlike those
tolerated by modern reptiles. [ ] Early Cretaceous, southeastern Australia, dinosaurs,
labyrinthodonts, palaeoclimate.

Thomas H. Rich, Museum of Victoria, P.O. Box 666E, Melbourne, Victoria, Australia; 5

April 1996.

Only one area within the palaeo-Antarctic
Circle has produced dinosaurs — southeastern
Australia (Aptian-Albian, early Cretaceous). Four
more are known to have been close to the palaeo-
Antarctic Circle — southeastern Queensland,
Australia (early or middle Jurassic), Beardmore
Glacier, Antarctica (early Jurassic), James Ross
and Vega Islands, Antarctica (Campanian and
Maastrichtian, late Cretaceous) and North Island,
New Zealand (probably Maastrichtian, late
Cretaceous). In the Northern Hemisphere, eight
areas (late Jurassic? to late Cretaceous) are
known to have produced dinosaurs within the
palaeo-Arctic Circle (Fig. 1). Most of the polar
dinosaur material from both hemispheres was
acquired during the past twenty years and much
of it remains to be described (Table 1).

Polar Gondwana may have served as both a
refuge and birthplace for some dinosaur groups.

From the early or middle Jurassic of
southeastern Queensland has come a partial
skeleton of one of the earliest sauropods,
Rhoetosaurus brownii Longman, 1927.

A single astragalus suggests that the well-known
form Allosaurus may have persisted into the Ap-
tian of southeastern Australia after having be-
come extinct elsewhere at the end of the Jurassic.

On the basis of one or a few bones, the presence
of three groups previously known in the late
Cretaceous of the Northern Hemisphere has been
suggested in the early Cretaceous of southeastern
Australia. An ulna from the Aptian there bears an
uncanny resemblance to that of the Maastrichtian
protoceratopsian Leptoceratops gracilis from Al-
berta, Canada. This suggests that Polar

Gondwana may have been the place of origin for
the neoceratopsians, for they are known nowhere
else prior to the Late Cretaceous.

Ornithomimosaurs are represented in the Al-
bian of southeastern Australia by femora distinct
enough to base a new genus and species, Timimus
hermani Rich & Vickers-Rich, 1994. As yet un-
described ornithosaur vertebrae are known from
both the same Albian site that produced femora
from southeastern Australia as well as Aptian
sites a few hundred kilometres to the east. This
material suggests a presence for this group on the
Gondwana continents prior to when or-
nithomimosaurs are best known in the northern
hemisphere, the late Cretaceous. However, the
recent publication of Shuvosaurus inexpectatus
Chatterjee, 1993 from the late Triassic of Texas,
implies that ornithomimosaurs have had a much
longer history in the northern hemisphere than
previously suspected. The late Jurassic Elaphro-
saurus bambergi from Tendaguru, Tanzania,
formerly regarded as an ornithomimosaur, has
recently been allocated to the abelisaurs (Holtz,
1994).

Oviraptorosaurs, previously represented ex-
clusively in the late Cretaceous of the northern
hemisphere, appear to have been present in the
Albian of southeastern Australia based on a par-
tial surangular and a vertebra (Currie, Vickers-
Rich & Rich, 1996).

Associated with these dinosaurs is the most
unexpected and best documented temporal range
extension of a major tetrapod group. Two decades
ago, labyrinthodont amphibians were thought to
have become extinct at the end of the Triassic.

712 MEMOIRS OF THE QUEENSLAND MUSEUM

Taloona Station, Queensland, A\

Victoria, Australia

Mangahouanga Stream, New Zealand

Put
T Ue
Nm. m

James Ross & Vega Islands, Antarctica Beardmore Glacier, Antarctica

B PS eee

we YOR Y —— =<

Ankylosauria Aves Carnosaur Carnosaur footprint Small Carnosaur or Theropod Ceratopsidae

— 2 4.

Hypsilophodontidae, 5 different genera — lguanodontidae footprint Labyrinthodont amphibian Marsupialia

f 2 , €

Plesiosaur Prosauropod Protoceratopsian Pterosaur Sauropod

FIG. 1. Distribution of polar dinosaurs, Southern Hemisphere (this page) and Northern Hemisphere (facing page).
Dinosaur sites in polar latitudes are not common, most having been discovered in the last 20 years. None are
fully studied. The higher diversity in southeastern Australia is probably due to greater collecting effort and
easier accessibility. When some of the other sites are more fully studied, they may provide similar diversity.
Base map from Smith, Hurley & Briden (1981).

SIGNIFICANCE OF POLAR DINOSAURS IN GONDWANA 713

Spitzbergen
eK

Sangarskaya Svita, Siberia e

IJ -eK
Kakanaut, Siberia

Bylot Island,
Canada
Colville River, Alaska

Northwest Territories,
Canada

Talkeetna Mountains,
Alaska

Yukon Territory, Canada

EN a 4m Geological Age of Site or Sites

Crocodilia Dromaeosauridae Hadrosauridae Hypsilophontidae J Jurrassic, 145 - 208 million years

Early Jurassic, 178 - 208 million years

IJ - Late Jurassic, 145 - 157 million years
- Early Cretaceous, 97 - 145 million years
Mosasaur Multituberculata ^ Ornithomimosauridae ^ Oviraptosauridae Placentalia

Late Cretaceous, 65 - 97 million years

Quantity of Fossil Material

- Greater than 1,000 bones
"o
e E 100 - 1,000 bones
ee =
Stegosauridae Tritylodontid Troodontidae Turtle e Less than 100 bones
FIG. 1 (cont.).
Subsequent to that, they have been described (Dong, 1985) and late Jurassic of Mongolia

from the early Jurassic of Australia (Warren & (Shishkin, 1991). Nearly complete lower jaws
Hutchinson, 1983), middle Jurassic of China with teeth, pectoral girdle elements, and ver-

714

tebrae form the basis for recognition of their
presence in the Aptian of southeastern Australia
(Warren, Kool, Cleeland, Rich & Rich, 1991;
Warren, Rich & Vickers-Rich, in press).

Crocodilian-like in their overall morphology,
the latest survivors of this once dominant group
of amphibians may have found a final refuge in
Aptian polar southeastern Australia because of a
greater tolerance of cold, as is true of modern
frogs and salamanders, than the crocodilians
which appeared there subsequently in the Albian
when conditions became warmer. There is no
trace of an Albian labyrinthodont in the south-
eastern Australian deposits where crocodilians
are represented by a few dermal scutes.

Among polar dinosaurs from either hemi-
sphere, only one feature of one individual has yet
been interpreted as an adaptation to life in a high
latitude environment. This may reflect more the
fact that very little is known about these animals
and less published, than that they rarely displayed
marked differences between themselves and their
lower latitude contemporaries. The feature in
question is the enlarged optic lobes of the brain
that occurred on the holotype of the hyp-
silophodontid Leaellynasaura amicagraphica
Rich & Rich, 1989 from the Aptian of south-
eastern Australia. Hypertrophy of this structure
formed the basis for the suggestion that this
animal had enhanced ability to see under the low
light conditions that would have prevailed during
the prolonged periods of continuous darkness
each Winter.

Hypsilophodontid dinosaurs are generally a
rare element in most dinosaur assemblages. Even
where specimens are relatively common, as on
the Isle of Wight, their taxonomic diversity is not
great. Southeastern Australia is a marked excep-
tion to that generalisation. At least six species in
five or six genera occur there; just over half the
total dinosaurs recognised to date. Currie (pers.
comm.) has suggested that hypsilophodontids
may have been primarily an upland group at
lower latitudes, hence their general rarity there,
and are better represented in southeastern
Australia because of its cooler conditions.

In this regard it may be noteworthy that the
dinosaur known from the late Cretaceous of Vega
Island is a hypsilophodontid or dryosaurid (Mil-
ner & Hooker, 1992) and that of the four found at
Mangahouanga Stream, New Zealand, one is a
probable dryosaurid, a family closely related to
hypsilophodontids (Wiffen & Molnar, 1989).
This explanation for the apparent preponderance
of these groups at high latitude may well be true.

MEMOIRS OF THE QUEENSLAND MUSEUM

However, in the much more fossiliferous Lis-
comb Bonebed locality on the Colville River,
Alaska, hypsilophodontids are all but unknown
despite the extensive sieving programme carried
out there (Clemens & Nelms, 1993; pers. obs.).
The age of the site is Maastrichtian and by that
time there may not have been the variety of
hypsilophodontids there was earlier in the
Cretaceous. But that does not explain why the one
known to occur there is not more abundant.

Today, there are no avian or terrestrial mam-
malian families restricted to the polar regions. As
far as the available record has been analysed, the
same can be said of the polar dinosaurs as pre-
viously independently noted by Molnar & Wiffen
(1994). Although new genera and species have
been recognised among them, they all belong to
families also known at lower latitudes.

However, given the fragmentary nature of
much of the evidence for the existance of various
groups, this could be an artefact. If one had a
single tooth of a litoptem and that group was
otherwise unknown, depending on the species,
the most parsimonious familial identification
might be Equidae. In that case one would be
parsimonious but one would be wrong, The
recent identification of protoceratopsians in the
Aptian of Australia (Rich & Vickers-Rich, 1994)
could be a parallel case. On the basis of a single
bone, rather than propose an entirely new group
of vertebrates which would share an uncanny
resemblance in the form of the ulna to
protoceratopsians but would in some other way
be distinct from them, the specimen was allocated
to this known taxon which extended its record not
only to another continent but also backwards in
time by at least 15 million years. At our present
state of knowledge that is an identification which
is quite plausible but could well eventually prove
to be fundamentally in error.

The Gondwanan polar dinosaur localities that
form groups of the same age are quite meagre in
their total faunal lists, making comparisons quite
preliminary. It is fortuituous, however, that in the
broadest terms, they share a common facies. Both
early Jurassic sites are flood plain deposits and
the three late Cretaceous ones, nearshore marine.
It may be of some note that the New Zealand late
Cretaceous locality shares ankylosaurs and
dryosaurids/hypsilophodontids with the two sites
in the Antarctic Peninsula, the only two dinosaurs
known to occur in the latter region.

There is no evidence such as tillites to suggest
that continental ice sheets existed during the
Mesozoic at high latitudes (Frakes, Francis &

SIGNIFICANCE OF POLAR DINOSAURS IN GONDWANA

TABLE 1. Gondwana Polar Dinosaur Localities.

1, Mt Kirkpatrick, Beardmore Glacier area,
Transantarctic Mountains, Antarctica, 165°E,
84°S

Rock unit. Falla Formation (fluviatile)

Age. Jurassic on stage-of-evolution of dinosaurs
sean 61°S (Smith, Hurley & Briden
1981)

Palaeoenvironment. Foreland basin flood plain
Recent Reference. Hammer & Hickerson (1994)
Fauna. Cryolophosaurus ellioti, plesiomorphic
allosaurid (Hammer, pers. comm. to Molnar
1995); Small theropod; Prosauropod; Tritylodon-
tid; Pterosaur

2, Australia, Southeastern Queensland, Taloona
Station near Roma, 149°03+3’E, 26°05+3’S.
Rock unit. Walloon Coal Measures (Injune Creek
Group)

Age. Early Jurass ic [?Bajocian]

Palaeolatitude. 65°S (Douglas & Williams
1982), 56?S ( Smith, Hurley & Briden 1981)
Palaeoenvironment. Intracratonic flood plain
Reference. Longman (1927)

Fauna. Sauropoda: Rhoetosaurus brownei

3, Victoria, Australia, numerous sites centred on
38°45’S, 143°30’E. (Otway Group) and 38?40' S,
145°40’E (Strzelecki Group)
Rock units. Strzelecki and Otway Groups
(fluviatile)
Age. Strzelecki Group Aptian except Koonwarra
locality which is Albian; Otway Group Albian
Palaeolatitude. Strzelecki Group, 77.8°S; Otway
Group, 66.8°S. Minimum estimates because of
thermal overprinting between 75 & 100my BP
(Whitelaw 1993).
Palaeoenvironment. Rift valley flood plain
Recent Reference. Rich & Vickers-Rich (1994)
Fauna.Hypsilophodontidae: Fulgurotherium
australe, Leaellynasaura amicagraphica, Atlas-
copcosaurus loadsi, Victorian hypsilophodontid
Type 1, Victorian hypsilophodontid Type 2, Vic-
torian hypsilophodontid Type 3

Omithomimidae: Timimus hermani

Ankylosauria, cf. Minmi

Protoceratopsidae: Aff. Leptoceratops

Oviraptorosaur?

Carnosauria: Allosaurus sp.

Testudines: Chelycarapookus arcuatus (tes-
tudines)

Testudines: Cryptodira

715

Plesiosaur

Pterosaur

Aves

Labyrinthodont: New genus and species (War-
ren et al., in press)

Ceratodontidae: Ceratodus avus, Ceratodus
nargun, Ceratodus sp., Coccolepis woodwardi,
Wadeicthys oxyops, Koonwarria manifrons,
Leptolepis koonwarri

4, Antarctica, Antarctic Peninsula, James Ross
Island, 57.9°W, 63.9°S

Rock unit. Santa Marta Formation (Marambio
Group)

Age. Campanian

Palaeolatitude. 65°S (Scotese et al. 1988)
Palaeoenvironment. Nearshore marine

Recent Reference. Gasparini, Pereda-S. & Molnar
(1996, this volume)

Fauna. Nodosauridae

5, Antarctica, Antarctic Peninsula, James Ross
Island, 57.9°W, 63.9°S

Rock unit. Hidden Lake Formation

Age. Coni acian-Santonian

Palaeolatitude. 65°S (Scotese et al. 1988)
Palaeoenvironment. Nearshore marine

Recent Reference. Molnar, Angriman &
Gasparini (1996, this volume)

Fauna. Theropod

6, Antarctica, Antarctic Peninsula, Vega Island,
57.5°W, 63.7°S

Rock unit. Cape Lamb Member of the Lopez de
Bertodano Formation

Age. Late Campanian - early Maastrichtian
Palaeolatitude. 65°S (Scotese et al. 1988)
Palaeoenvironment. Near shore marine

Recent Reference. Milner & Hooker (1992 ).
Fauna. Hypsilophodontidae

7, New Zealand, North Island, Mangahouanga
Stream, 176°45’E, 39°S

Rock unit. Maungataniwha Sandstone

Age. Late Cretaceous, Campanian - Maastrichtian
Palaeolatitude. 55°S (Scotese et al. 1988)
Palaeoenvironment . Nearshore marine

Recent Reference, Molnar & Wiffen (1994)
Fauna. Dryosauridae, Theropod, Ankylosauria,
Sauropoda, Testudines, Plesiosauroidea,
Mosasauridae, Pterosauria

716

Syktus, 1992). However, it has been inferred on
the basis of the occurrence of dropstones as large
as 3m across in finegrained marine sediments
deposited during the late Jurassic and early
Cretaceous that winter seasonal ice did form at
high palaeolatitudes in both hemispheres (Frakes,
Francis & Syktus, 1992).

In southeastern Australia, studies assessing the
mean annual palaeotemperature have been car-
ried out on the sediments producing the
dinosaurs. Palaeobotanical evidence based on
leaf margin and stomatal structure together with
the overall composition of the flora have been
taken to suggest a mean annual temperature of
+10°C (Parrish et al., 1991) while an oxygen
isotope estimate is -6°C (Gregory et al. 1989), the
difference between Chicago and Point Barrow,
Alaska, today. While the biological implications
of these two estimates are quite different, they are
concordant in that the palaeoclimate was far from
tropical.

No matter what the palaeotemperature was,
polar dinosaurs would had to have adapted to
prolonged periods of annual darkness each
Winter. There have been suggestions that in the
geological past the earth's rotational axis might
have been significantly closer to being oriented
perpendicular to the plane of the ecliptic (e.g.,
Douglas & Williams, 1982). If this had occurred,
the length of continuous darkness each Winter at
high latitudes would have been reduced. How-
ever, the criticism of Laplace (1829) against the
earth's obliquity having shifted more than a few
degrees from its present orientation as it does over
a period of about 41,000 years (one of the com-
ponents of the Milankovitch cycle), has never
been refuted.

Of all the polar dinosaur sites, that on the Col-
ville River, Alaska (Fig. 1), hasthe greatest poten-
tial to yield a detailed picture about these animals.
For more than 200km on the left bank of that river
are beautifully exposed outcrops that range in age
from Albian to Maastrichtian. Fossil bones have
been found at many places along that river. By
contrast, the potential of sites which are shown as
having less than 100 bones is much more
problematical.

The work on the polar dinosaurs of Gondwana
is just beginning. Just how distinctive they were
from their lower latitude contemporaries is all but
unknown. The Falla Formation of Antarctica
seems the most promising area at the moment in
terms of future potential. Southeastern Australian
outcrops are limited to just 4 square kilometres
which have been thoroughly prospected. In

MEMOIRS OF THE QUEENSLAND MUSEUM

another four years, the major sites there may be
worked out. Outcrops of the Wallon Coal
Measures are rare in southeastern Queensland. In
the Mangahouanga Stream locality of New
Zealand the fossil vertebrates are recovered from
loose boulders in the stream itself and dinosaur
remains are quite rare. The Ross Island occurren-
ces, like the New Zealand ones, are marine but
unlike there, partial skeletons rather than isolated
bones have been found. Although Ross Island is
unlikely to produce major concentrations of
dinosaur remains, with persistent prospecting
over the large available area, it may be expected
to continue to contribute significantly to our
knowledge of Gondwana polar dinosaurs. Clear-
ly, to find significantly more information about
Gondwana polar dinosaurs outside the area of the
Falla Formation, new areas need to be inves-
tigated such as the Cretaceous coalfields of New
Zealand (Rich, 1975).

LITERATURE CITED

CHATTERJEE, S. 1993, Shuvosaurus, a new theropod.
Research & Exploration 9: 274-285.

CLEMENS, W.A., & NELMS, L.G. 1993. Paleo-
ecological implications of Alaskan terrestrial ver-
tebrate fauna in latest Cretaceous time at high
paleolatitudes. Geology 21: 503-506.

CURRIE, PJ., VICKERS-RICH, P, & RICH, T.H.
1996. Possible oviraptorosaur (Theropoda,
Dinosauria) specimens from the Early Cretaceous
Otway Group of Dinosaur Cove, Australia. Al-
cheringa 20: 73-79.

DONG Z. 1985. A Middle Jurassic labyrinthodont
(Sinobrachyops placenticephalus gen. et sp. nov.)
from Dashaupu, Zigong, Sichuan Province.
[Chinese with English abstract.] Vertebrata
Palasiatica 23: 301-307.

DOUGLAS, J.G. & WILLIAMS, G.E. 1982. Southern
polar forests: the Early Cretaceous floras of Vic-
toria and their palaeoclimatic significance.
Palaeogeography, ^ Palaeoclimatology,
Palaeoecology 39: 171-185.

FRAKES, L.A., FRANCIS, J.E. & SYKTUS, J.I. 1992.
‘Climate Modes of the Phanerozoic’. (Cambridge
University Press: Cambridge). xi, 274 pp.

GASPARINI, Z., PEREDA-SUBERBIOLA, X. &
MOLNAR, R.E. 1996, this volume. New data on
the ankylosaurian dinosaur from the Late
Cretaceous of the Antarctic Peninsula. Memoirs
of the Queensland Museum 39(3): 583-594,

GREGORY, R.T., DOUTHITT, C.B., DUDDY, LR.,
RICH, P.V. & RICH, T.H. 1989. Oxygen isotopic
composition of carbonate concretions from the
lower Cretaceous of Victoria, Australia: implica-
tions for the evolution of meteoric waters on the
Australian continent in a paleopolar environment.
Earth and Planetary Science Letters 92: 27-42.

SIGNIFICANCE OF POLAR DINOSAURS IN GONDWANA

HAMMER, W.R. & HICKERSON, W.J. 1994. A
crested theropod dinosaur from Antarctica.
Science 264: 828-830.

HOLTZ, T.R. Jr 1994. The phylogenetic position of the
Tyrannosauridae: implications for theropod sys-
tematics. Journal of Paleontology 68: 1100-1117.

LAPLACE, P.S. 1829. ‘Mecanique Celeste’, vol. II.

LONGMAN, H.A. 1927. The giant dinosaur
Rhoetosaurus brownei. Memoirs of the
Queensland Museum 9: 1-18

MILNER, A.C. & HOOKER, J.J. 1992. An ornithopod
dinosaur from the upper Cretaceous of the An-
tarctic Peninsula. Journal of Vertebrate Paleontol-
ogy 12: 44A.

MOLNAR, R.E., LOPEZ ANGRIMAN, A. &
GASPARINI, Z. 1996, this volume. An Antarctic
Cretaceous theropod. Memoirs of the Queensland
Museum 39(3):669-674.

MOLNAR, R.E. & WIFFEN, J. 1994. A late Cretaceous
polar dinosaur fauna from New Zealand.
Cretaceous Research 15: 689-706.

PARRISH, J.T., SPICER, R.A., DOUGLAS, J.G.,
RICH, T.H. & VICKERS-RICH, P. 1991. Con-
tinental climate near the Albian South Pole and
comparison with climate near the North Pole.
Geological Society of America Abstracts with
Programs 23: A302.

RICH, T.H. 1975. Potential pre-Pleistocene fossil
tetrapod sitesin New Zealand. Mauri Ora 3: 45-54.

RICH, T.H. & RICH, P.V. 1989. Polar dinosaurs and
biotas of the early Cretaceous of southeastern
Australia. National Geographic Research 5: 15-53.

RICH, T.H. & VICKERS-RICH, P. 1994. Neoceratop-
sians & Ornithomimosaurs: Dinosaurs of

717

Gondwana Origin? Research & Exploration 10:
129-131.

SCOTESE, C.R., GAHAGAN, L.M. & LARSON, R.L.
1988. Plate tectonic reconstructions of the
Cretaceous and Cenozoic ocean basins. Tec-
tonophysics 155: 27-48

SHISHKIN, M.A. 1991. Labirintodont iz pozdney yury
Mongolii. Palaeontological Zhurnal, no. 1, pp.
81-95. [A labyrinthodont from the Late Jurassic of
Mongolia. Paleontological Journal 1: 78-91.]

SMITH, A.B., HURLEY, A.M. & BRIDEN, J.C. 1981.
‘Phanerozoic Paleocontinental World Maps’.
(Cambridge University Press: Cambridge). 102 pp.

WARREN, A.A. & HUTCHINSON, M.N. 1983. The
last labyrinthodont? A new brachyopoid (Am-
phibia, Temnospondyli) from the Early Jurassic
Evergreen Formation of Queensland, Australia.
Philosophical Transactions of the Royal Society,
Series B, 303: 1-62.

WARREN, A.A., KOOL, L., CLEELAND, M., RICH,
T.H. & VICKERS-RICH, P. 1991. An Early
Cretaceous labyrinthodont. Alcheringa 15: 327-
332.

WARREN, A.A., RICH, T.H. & VICKERS-RICH, P. in
press. The last, last labyrinthodont? Palaon-
tographica, Abt. A.

WHITELAW, M.J. 1993. Paleomagnetic Paleolatitude
determinations for the Cretaceous vertebrate
localities of southeastern Australia-high latitude
dinosaur faunas. Journal of Vertebrate Paleontol-
ogy 13, supplement to Number 3: 62A.

WIFFEN, J. & MOLNAR, R.E. 1989. An Upper
Cretaceous ornithopod from New Zealand.
Geobios — Paleontologie, Stratigraphie,
Paleoecologie 22: 531-536.

EARLY CRETACEOUS POLAR TETRAPODS FROM THE GREAT SOUTHERN RIFT

VALLEY, SOUHTEASTERN AUSTRALIA
PATRICIA VICKERS-RICH

Vickers-Rich, P. 1996 12 20: Early Cretaceous polar tetrapods from the Great Southern Rift
Valley, southeastern Australia. Memoirs of the Queensland Museum 39(3): 719-723. Bris-
bane. ISSN 0079-8835.

Early Cretaceous deposits from southeastern Australia record a cold, extensively forested
environment. Tetrapod fossils in channel fills, gravity flows, lag and point bar deposits in
the Aptian Wonthaggi and Albian Middle Eumeralla formations. The fossils occur mainly
in horizontally-stratified, clast-supported conglomerates and massive, matrix-supported
conglomerates, Leaf mats indicate that several species of deciduous plants shed their leaves
together, presumably in winter. Taphonomy of the lake beds at Koonwarra indicate seasonal
freezing. A variety of dinosaurs — including hypsilophodonts, ankylosaurs, neoceratopsians,
allosaurs, dromaeosaurs, oviraptorosaurs and ornithomimosaurs — were present as well as
pterosaurs, plesiosaurs, temnospondyls and crocodilians. The latter two groups did not occur
together and the temnospondyls lived under either cooler or higher-energy conditions than
the crocodilians. [] Cretaceous, Australia, dinosaur, environment of deposition,
palaeoclimate, temnospondyl.

P. Vickers-Rich, Dept of Earth Sciences, Monash University, Clayton, Victoria, Australia,

3168; 5 April 1996.

Breakup of Gondwana began in the Late Trias-
sic. Australia and Antarctica were the last two
continents of the supercontinent to separate, com-
mencing in the Late Jurassic. During the Early
Cretaceous, between 125 and 105 million years
ago, what is now the southern coast of Victoria,
Australia, was part of a rift valley formed be-
tween the two continents during the initial phase
of that separation. As separation continued and
sped up, the floor of the rift valley sank. As a
consequence of that event, volcanoes that
probably lay near the Lord Howe Rise, poured
quantities of ash (estimated at 50,000 cubic
kilometres) into the rift valley where it was then
reworked by the rivers and streams, the precur-
sors to the green sandstones and mudstones that
now form prominent cliffs for about 200km of the
Victorian coastline (Fig. 1).

Since that rift valley was formed, Australia has
drifted far to the north, while Antarctica remained
close to its Early Cretaceous position straddling
the South Pole. In the Early Cretaceous
southeastern Australia lay well within the An-
tarctic Circle. The dinosaurs, other fauna and
plants that lived in this region thus contended
with prolonged periods of continuous darkness
each year, just as musk ox and reindeer do today.
However, the geochemical and botanical climatic
indicators suggest that the environment in the
Early Cretaceous of Victoria was not frigid as are
similar latitudes today, and extensive forests

clothed many parts of the rift valley and its flanks
at times.

ENVIRONMENTAL SETTING OF THE
TETRAPOD FAUNA

A small, but growing, collection of fossil tetrapods
are known from several locales in southeastern
Australia (Currie, Vickers-Rich & Rich, 1996;
Gross, Rich & Vickers-Rich, 1993; Molnar, Flan-
nery & Rich, 1981; 1985; Rich & Rich, 1989;
Rich & Vickers-Rich, 1994; Rich et al., 1988;
1992; Vickers-Rich & Rich, 1993; Warren et al.,
1991). These fossils have been collected from
three geologic units: 1, the younger Middle
Eumeralla Formation of Early Albian age in the
Otway Group; 2, the older undifferentiated Won-
thaggi Formation of Aptian age in the Strzelecki
Group; and 3, the Aptian San Remo Member near
the base of the Wonthaggi Formation.

Fossils are concentrated in a few facies within
these units: i.e., 1, mainly those preserved as a
consequence of rapid deposition in a new channel
cut as a consequence of a single major flood; 2,
as gravity flows of sediment, either mud or rock
debris; and to a lesser extent, 3, lag and point bar
deposits within major river systems (A. Constan-
tine, pers. comm.).

The bones are most commonly preserved in
horizontally stratified, clast-supported con-
glomerates and massive, matrix-supported con-
glomerates. A 'clast-supported conglomerate'

720

refers to one in which the individual pebbles or
boulders contact one another, whereas a ‘matrix-
supported conglomerate’ is one in which the
quantity of finegrained rock between the pebbles
or boulders is so great that they do not contact one
another. If the matrix could be removed from a
clast-supported conglomerate, the volume oc-
cupied by the conglomerate would not decrease,
whereas in the case of a matrix-supported con-
glomerate, it would.

The horizontally stratified, clast-supported
conglomerates are typically of the order of 5-
20cm thick and are characterised by pebble to
cobble-sized clasts formed of clay and mudstone.
Such deposits are stratified and contain little car-
bonaceous material. This type of bone-bearing
conglomerate typically occurs at the base of
channel complexes where significant erosion into
underlying floodplain claystones and mudstones
is evident. Such deposits are interpreted as having
formed by streams or rivers breaking their banks
during major floods, with the resultant flood
waters flowing out over the surrounding flood-
plain, picking up bones and plant debris and
concentrating them in erosion scours.

The massive, matrix-supported conglomerates,
on the other hand, differ in that the pebble to
boulder-sized clasts are not touching each other.
Instead, they appear to be suspended in a finer-
grained matrix composed of fine to medium-
grained sandstone. Mudstone is again the
dominant clast type, and plant remains are not
as common. Conglomerates of this type are
interpreted as debris flows, which form as the
result of bank collapses within the confines of
channels.

The regional setting for the two geological units
that have produced significant fossil collections
shows distinct change through time, from a
predominantly ‘meandering’ to a ‘braided’ sys-
tem upsequence. The older Wonthaggi Formation
is characterised by thick floodplain deposits and
channels with moderate to high sinuosity. There
are thick accumulations of horizontally stratified
to low angle crossbedded sediments, which indi-
cate rapid aggradation under transitional to high
flow regimes — suggestive of occasional flash
flooding. The formation consists of about 60%
sandstone and 40% mudstone and reflects vary-
ing flow regimes, perhaps due to discharge levels
controlled by snow melt. Vertebrate fossils have
been recovered from a variety of lithofacies
within the formation, but mainly from horizontal-
ly stratified, clast-supported conglomerates and
massive, matrix-supported conglomerates.

MEMOIRS OF THE QUEENSLAND MUSEUM

The younger Middle Eumeralla Formation con-
sists primarily of sandstone (more than 70%) and
is characterised by a classic braided river
lithofacies and architecture. Sinuosity of channel
deposits is low, and the channels were wide and
shallow. There are thick sequences of both
floodplain and lacustrine sediments with lenses
of sand-sized particles representing individual
channels stacked one on top of another. Ver-
tebrates occur in only a few environmental set-
tings, somewhat in contrast to the situation in the
Wonthaggi Formation — in the Middle Eumeral-
la bones occur predominantly in sediments
formed when a channel broke its bank and flowed
out over the surrounding floodplain, or in those
formed by loading and collapse of non-vegetated
sand bars in the main channel during a peak flood
stage that caused liquefaction of sand and then
down-slope mass flow.

THE BIOTA: PALAEOFLORA AND
INVERTEBRATE FAUNA AS
PALAEOCLIMATIC INDICATORS

The palaeoflora of southeastern Australia
during the Early Cretaceous was dominated by
conifers, ferns, cycads, ginkgoes and lower-
growing horsetails and bryophytes. Angiosperms
were present, but only as prostrate or small her-
baceous forms. The structure and diversity of this
flora suggest a mean annual temperature of 8-
10?C (Douglas, 1969; 1971; Drinnan & Chambers,
1986; Parrish et al., 1991). Some plants in this
flora were evergreens, while others were clearly
deciduous, the best evidence being fossilised leaf
mats suggestive of simultaneous shedding of
leaves by several species. The evergreen plants
possessed leaves with thick cuticle and micro-
phyllus (small) leaves. Such leaf morphology is
consistent with a climate characterised by sig-
nificant variation in temperature throughout the
year or a fluctuating water supply — conditions
that would be expected within a continental land
mass distant from the ocean. Leaf mats themselves
are indicative of leaves having fallen in a short
period of time — which can be brought on by
pronounced seasonal changes in light, temperature
or water availability or a combination of these
factors. Pronounced seasonal contrasts are ex-
pected of such inland environments at such
latitudes as was the case in southern Victoria in
the Early Cretaceous (Parrish et al., 1991).

One unusual locality to have produced fossil

plants, invertebrates, fishes and birds, is Koon-
warra. Koonwarra is an inland site in the Won-

CRETACEOUS POLAR TETRAPODS FROM SOUTHEASTERN AUSTRALIA

ANTARCTICA

721

AUSTRALIA

FIG. 1. Sketch of junction between SE Australia and E Antarctica in the Early Cretaceous. Separation began in
the Late Jurassic and resulted in a rift valley being formed on the interplate boundary. Into this rift valley poured
a vast quantity of volanogenic sediments derived from volcanoes perhaps laying to the east in the vicinity of
the Lord Howe Rise. Large rivers flowed across the floor of this rift valley, fed perhaps in part by meltwater of
snow presumably located at high altitude on mountains on the margins of the rift valley or the volcanoes which
produced the volcanogenic sediments. The Early Cretaceous tetrapod and plant fossils occur for the most part
in sediments that were laid down in small streams feeding into the larger rivers on the floor of the rift valley.
Subsequent to the Early Cretaceous, the sediments deposited on the floor of the rift valley were first lithified as
they were buried under additional sediment. Then late in the Cainozoic, these sediments were uplifted to form

the Strzlecki and Otway Ranges.

thaggi Formation of the Strzelecki Group and
represents the remains of an ancient lake. Most
other vertebrate sites in southeastern Victoria are
coastal exposures, and fossils were deposited not
in the quiet waters of large lakes, but in the more
energetic riverine and floodplain environments.
Insects and other invertebrates, primarily larval
forms, recovered from the Koonwarra locality are
most closely related to forms typical of cool Tas-
manian mountain lakes today (Jell & Duncan,
1986), relationships that are indicative of
temperatures similar to those reflected by the
palaeoflora. Waldman (1971), who studied the
Koonwarra fish concentration, suggested that the
entire fossil accumulation may have been due to
winterkill — when the lake froze over and oxygen
supply was greatly reduced.

Further evidence that temperatures were cool
and that ice may have formed at times during the
year occurs in contemporaneous sequences in
central Australia. Boulders up to 3m in diameter
have been found in otherwise finegrained marine
sediments of the Bulldog Shale near Andamooka,
South Australia (Frakes & Francis, 1988). Frakes
& Francis have suggested that these boulders
dropped to the bottom of the shallow sea as
icebergs in which they floated, melted away. Al-
though there is no preserved evidence — such as
tillites — for glacial activity in Australia at this
time, unlike for the earlier Permian time, Frakes
has suggested that montane glaciation could have
been active and, in places, these glaciers might
have reached the sea at the base of drainage
systems.

722

In summary, the suite of animals, plants and
sedimentological data suggest that the climate of
southeastern Australia during the late Early
Cretaceous when dinosaurs are known to have
lived there, was somewhat cooler than at present,
but temperatures were by no means frigid as
similar high polar latitudes are today. O!5/0!6
evidence, however, suggests that mean annual
temperatures approached 0°C at times during
deposition of the dinosaur-bearing sediments.

VERTEBRATE ASSEMBLAGES OF THE
EARLY CRETACEOUS OF POLAR
AUSTRALIA

The Early Cretaceous terrestrial vertebrate as-
semblages of southeastern Australia are
dominated by dinosaurs, in particular hyp-
silophodontids. There are at least five genera and
six species of this family known from Victoria,
half as yet unnamed. The diversity of hyp-
silophodontids in these south polar latitudes is
unmatched anywhere else in the world, including
localites with hundreds of thousands of bones and
high diversity in the total dinosaur assemblage.
Currie (pers. comm.) has suggested that at lower
palaeolatitudes, hypsilophodontids may have
been upland forms and thus not frequently repre-
sented at the lower elevations where most
dinosaur fossils accumulated. Hypsilophodon-
tids, then, may well have been preadapted for the
conditions of southeastern Australia and conse-
quently thrived there. Hyspilophodonts are
known from both the Otway and Strzelecki
Groups and thus have an age range in these se-
quences from Aptian to Albian, with some sites
perhaps being as old as Valanginian in the
Strzelecki Group.

On the basis of femoral morphology, five genra
of hypsilophodontids have been recognised. To
date, formal scientific names have been assigned
to only three of these: Fulgurotherium, Leael-
lynasaura and Atlascopcosaurus. Prominent
optic lobes preserved on an endocast of Leael-
lynasaura suggest that this dinosaur had unusual-
ly enhanced ability to process visual signals. In a
polar setting, the most plausible explanation for
this acuity would be that it improved visual ability
under the low light conditions, which would have
prevailed during the months of continuous dark-
ness of the polar Winter. Since both palaeobotani-
cal and geochemical studies suggest that Winter
temperatures would have probably dipped below
freezing, and since this is a prohibitive tempera-
ture for activity of modern reptiles, it is tantalis-

MEMOIRS OF THE QUEENSLAND MUSEUM

ing to think that Leaellynasaura may have been
homeothermic — thus allowing the increased
visual acuity to have an adaptive advantage.

Other ornithischians in the fauna include an
ankylosaur (based on the cross-sectional outline
of a rib, a scute and a few teeth) and an Aptian
neoceratopsian (based on an ulna with a remark-
able resemblance to that of Leptoceratops
gracilis from the latest Cretaceous of Alberta).

Allosaurids, oviraptorosaurs and ornithomimid
theropods are known. An astragalus resembling
that of Allosaurus has been recognised in the
Aptian Wonthaggi Formation of the Strzelecki
Group (Molnar, Flannery & Rich, 1981; 1985).
Oviraptorosaurs (Currie, Vickers-Rich & Rich,
1996) and ornithomimids (Rich & Vickers-Rich,
1993) are both known from the Middle Eumeralla
Formation in the Otway Group, and ornitho-
mimids have also been recovered from the older
Strzelecki Group (Wonthaggi Formation).
Footprints of small theropods have been recorded
in the Otway Group.

The neoceratopsian, dromaeosaur and ovirap-
torosaur fossils from Australia are among the
oldest records of these groups anywhere in the
world. Other components of the fauna include the
young temnospondy] amphibians, which are rep-
resented by more than twenty bones, including a
pair of mandibles with the teeth in situ. Two bones
of pterosaurs and half a dozen plesiosaur teeth
have also been recovered. Evidently the
plesiosaurs were freshwater animals, as all
sedimentological and palaeontological data point
to a fluviatile source for the containing sediments.

A few remains of crocodiles have been
recovered in the Middle Eumeralla Formation in
the Otway Group, but never have they been found
together with temnospondyl remains, restricted
as they are to the older Strzelecki Group. A warm-
ing trend up section in the Aptian-Albian se-
quence may explain this apparent faunal change
— the replacement of temnospondyls by
crocodiles.

The Victorian Aptian record of temnospondyls
is the most recent for the group anywhere in the
world. These amphibians were very crocodile-
like functionally — in body form and in tooth
morphology. Temnospondyls occur in the older
Aptian sediments where temperatures were lower
than in the younger Albian sediments that bear a
few dermal scutes of crocodiles. Perhaps the
rising temperatures allowed the invasion of
crocodilians into an area from which they had
been excluded by the cold waters, thus bringing
them into direct competition with the temnospon-

CRETACEOUS POLAR TETRAPODS FROM SOUTHEASTERN AUSTRALIA

dyls, followed by extinction of the latter. Today
amphibians, such as frogs and the Giant Japanese
Salamanders are able to cope with temperatures
well below those tolerated by living reptiles.
The evidence is suggestive, but not definitive,
however, for there is one other possible explana-
tion that cannot be ruled out. Temnospondyl fos-
sils are known from the Strzelecki Group only in
the high energy sediments that represent
fanglomerates pouring off the margins of the rift
valley. This coarse facies occurs widely in the
western exposures of the Aptian Strzelecki
Group, but is less common in the younger Aptian
Otway Group. Perhaps temnospondyls were
facies controlled, and thus their absence is owing
to sparsity ofthe coarse fanglomerate facies in the
younger sediments of the Otway Group, rather
than their extinction by the Albian owing to
temperature increase or some other factor.

CONCLUSION

Although genera endemic to southeastern
Australia occur in these Early Cretaceous as-
semblages, there is nothing yet recognised as
unique as the modern day koala or kangaroo. All
of the tetrapods found to date can be readily
accommodated in families known from other
continents. But what is clear is that in the Early
Cretaceous, southern Australia served as a refuge
allowing some groups to live well beyond their
time elsewhere in the world (e.g., Allosaurus,
temnospondyls and some fish and plant groups).
This area also nurtured novelty — it may have
been the cradle for such groups as the neoceratop-
sians, dromaeosaurs and oviraptorosaurs — a
cradle from which they dispersed later,
northwards, to meet with great success in North
America and Asia.

LITERATURE CITED

CURRIE, P.J., VICKERS-RICH, P. & RICH, T.H.
1996. Possible oviraptorosaur (Theropoda,
Dinosauria) specimens from the Early Cretaceous
Otway Group of Dinosaur Cove, Australia. Al-
cheringa 20: 73-79.

DOUGLAS, J.G. 1969. The Mesozoic floras of Victoria,
Parts 1 & 2. Memoir of the Geological Survey of
Victoria 28.

723

1972, The Mesozoic floras of Victoria, Part 3.
Memoirs of the Geological Survey of Victoria 29.
DRINNAN, A.N. & CHAMBERS, T.C., 1986. Flora of
the lower Koonwarra fossil bed (Korumburra
Group), South Gippsland, Victoria. Association of
Australasian Palaeontologists Memoir 3: 1-77.
FRAKES, L.A. & FRANCIS, J.E. 1988. A guide to
Phanerozoic cold polar climates from high-
latitude ice-rafting in the Cretaceous. Nature 333:

547-549,
GROSS, J.D., RICH, T.H. & VICKERS-RICH, P. 1993.
Dinosaur Bone Infection. National

Geogeographic Research and Exploration 9: 286-
293

MOLNAR, R.E., FLANNERY, T.F. & RICH, T.H.
1981. An allosaurid dinosaur from the Cretaceous
of Victoria, Australia. Alcheringa 5: 141-146.

1985. Aussie Allosaurus after all. Journal of Paleon-
tology 59: 1511-1513.

PARRISH, J.T., SPICER, R.A., DOUGLAS, J.G.,
RICH, T.H. & VICKERS-RICH, P. 1991. Con-
tinental climate near the Albian South Pole and
comparison with climate near the North Pole.
Geological Society of America, Abstracts with
Programs 23: A302. (abstract)

RICH, P.V., RICH, T.H., WAGSTAFF, B., MCEWEN-
MASON, J., DOUTHITT, R.T., GREGORY, R.T.
& FELTON, A. 1988. Evidence for low tempera-
tures and biologic diversity in Cretaceous high
latitudes of Australia. Science 242: 1403-1406.

RICH, T.H. & RICH, P.V. 1989. Polar dinosaurs and
biotas of the Early Cretaceous of southeastern
Australia. National Geographic Society Research
Reports 5: 15-53.

RICH, T.H., RICH, P.V., WAGSTAFF, B.E., McEWEN-
MASON, J.R.C., FLANNERY, T.E, ARCHER,
M., MOLNAR, R.E. & LONG, J.A. 1992. Two
possible chronological anomalies in the Early
Cretaceous tetrapod assemblage of southeastern
Australia. Pp. 165-176. In Chen, P. & Mateer, N.
(eds) ‘Proceedings First International Symposium
on Nonmarine Cretaceous Correlations'. (China
Ocean Press: Beijing).

RICH, T.H. & VICKERS-RICH, P. 1994. Neoceratop-
sians & ornithomimosaurs: dinosaurs of
Gondwana origin? National Geographic Research
and Exploration 10: 129-131.

WALDMAN, M. 1971. Fish from the freshwater Lower
Cretaceous of Victoria, Australia, with comments
on the palaeoenvironment. Special Papers in
Palaeontology 9: 1-124.

WARREN, A.A., KOOL, L., CLEELAND, M., RICH,
T.H. & VICKERS-RICH, P. 1991. An Early
Cretaceous labyrinthodont. Alcheringa 15: 327-
332.

DINOSAURIAN PALAEOBIOLOGY: A NEW ZEALAND PERSPECTIVE

J. WIFFEN

Wiffen, J. 1996 12 20: Dinosaurian palaeobiology: a New Zealand perspective. Memoirs of
the Queensland Museum 39(3): 725-731. Brisbane. ISSN 0079-8835.

The first known dinosaur bone from New Zealand was identified from the Late Cretaceous
marine sandstones at the Mangahouanga Stream site in 1980 (Molnar, 1981). Since then
isolated bones indicate that a variety of herbivores and carnivores were present after the
separation from Gondwana 80-85 million years ago until their extinction. Though geographi-
cally polar in origin, survival for a long period on an island landmass suggests a temperate
climate prevailed as New Zealand drifted north. [ ] New Zealand, Cretaceous, dinosaur,

palaeoclimate, palaeobotany, taphonomy.

J. Wiffen, 138 Beach Road, Haumoana, Hawke's Bay, New Zealand, 5 July 1996.

The only dinosaur fossils so far found in New
Zealand come from the Late Cretaceous
Maastrichtian-Campanian marine sediments in
the Mangahouanga Stream fossil site, North Is-
land, New Zealand. These bones represent a
diverse dinosaur fauna and the only record of
terrestrial life from Mesozoic New Zealand. They
include vertebrae and phalanges from 2-3 species
of theropod, vertebrae (Fig. 1A) and a rib from an
ankylosaur, an ilium (Fig. 1B) from an or-
nithopod and a rib fragment (Fig. 1C) from a
sauropod (Molnar, 1981; Wiffen & Molnar, 1989;
Molnar & Wiffen, 1994).

The bones are found in hard calcareous concre-
tions, locally derived from the upper layers of the
Maungataniwha Sandstone. This was laid down
on the eastern coastline of ancient New Zealand
in Campanian-Maastrichtian (Haumurian-
Piripauan) times and is now exposed at the Man-
gahouanga Stream (Fig. 2). Isolated bones were
carried from the adjacent landmass by rivers
probably in seasonal floods — the larger bones
dropped as the flow slowed through lagoonal or
estuarine areas — prior to burial in riverborne
debris in nearshore wave base deposits
(Crampton & Moore, 1990; Moore & Joass,
1991; Wilson & Moore, 1988). They occur along
with fossils of marine origin.

The described terrestrial bones are easily recog-
nised (Fig. 3) and well preserved, with diagnostic
features. So it seems likely that the bones weren’t
transported far from land to the site. When ex-
tracted from the rock, there is little apparent
abrasive damage on the bone surface that oc-
curred prior to burial. Most of the damage seen is
recent, from exposure after the concretions have
eroded out of the enclosing rock and split due to
temperature extremes (i.e., winter frosts and hot

dry summers), with resulting surfaces worn by
running water.

SIGNIFICANCE

By and large the record of terrestrial life from
ancient New Zealand isn’t as good as those from
elsewhere. Acid conditions produced by the high
rainfall and extensive forest cover are thought to
account for the generally poor fossil record
(Fleming, 1962).

Consequently although the number of dinosaur
bones identified to date is small, they are sig-
nificant for several reasons:

1) They show that dinosaurs inhabited this part
of Gondwana prior to its separation from the
Marie Byrd Land area of western Antarctica
(Stevens, 1985).

2) They are the first evidence that terrestrial
vertebrates survived in New Zealand after its
separation from Gondwana, as opposed to having
arrived there by dispersal.

3) This is the only known Southern Hemisphere
region where dinosaurs lived on a small island,
for up to 15-20 million years, unti] their extinc-
tion.

What evolutionary changes, if any, occurred in
these forms in the presumably static island en-
vironment, in the absence of migrations to New
Zealand and with (presumably) decreased com-
petition, are not known. To date no unique fea-
tures have been seen to suggest morphological
changes or geographic or climatic adaptations:
however, considerably more complete fossil
material is required to determine this.

Although similar taxa have been found from
this period in Antarctica (Hooker et al., 1991;
Hammer & Hickerson, 1994) no closely-related
forms have been detected, in spite of the variety
of taxa — ornithopod, sauropod and ankylosaur,

726

FIG. 1. A, Articulated nodosaurian caudal vertebrae (CD 546) from the
Maungataniwha Sandstone, Mangahouanga Stream, in posterolateral obli-
que view. The specimen unfortunately exceeded the depth of focus of the
camera. CD, New Zealand Geological Survey Collection, Lower Hutt.
Scale = lcm. B, Posterior part of the right ilium (CD 529) of a dryosaur-like
ornithopod from the Maungataniwha Sandstone, Mangahouanga Stream.
Scale = lcm. C, Fragment of sauropod rib (CD 542) from the Maun-
gataniwha Sandstone, Mangahouanga Stream. Scale = 5cm.

probably nodosaurian, dinosaurs — that have been
identified from the Mangahouanga site (Molnar
& Wiffen, 1994): the ankylosaurid bones found
by the British Museum are as yet undescribed.
Recent discoveries, which include a neoceratop-
sian, Timimus hermani and a possible ovirap-
torosaur (Rich & Rich, 1994; Currie et al., 1996)
from Victoria, new ankylosaur material (Molnar,

MEMOIRS OF THE QUEENSLAND MUSEUM

this volume) and a second skull
of the large ornithopod Mut-
taburrasaurus (Molnar, this
volume) from Queensland and
a range of new Jurassic
material from Antarctica
(Hammer & Hickerson, 1994)
suggest the early and
widespread distribution of
dinosaurs on the southern con-
tinent. So a greater under-
standing of the distributions of
these polar dinosaurs, and their
relationships to those from
New Zealand, should result
when more becomes known
about these new discoveries.

The difficulties of close com-
parisons with New Zealand
dinosaur bones are due to the
paucity of fossil material col-
lected and identified to date,
hence the difficulty in finding
homologous elements for com-
parison, as well as the dis-
crepancy in geological age
from most other Australasian
/Antarctic sites.

FLORA AND FAUNA

The New Zealand marine
fossil record is relatively good,
and the Mangahouanga Stream
site has contributed fossils to
both marine (Crampton &
Moore, 1990; Glaessner, 1980;
Feldmann, 1993; Wiffen,
1981) and terrestrial records —
including (possibly fresh-
water) turtles (Gaffney, pers.
comm., 1989), flying reptiles
(Fig. 4)(Wiffen & Molnar,
1988), a coleopteran (Craw &
Watt, 1987) and a cockroach,
which is still under study (Fig. 5).

A considerable quantity of
fossil wood and plant material
has been collected as well and this is currently
being described by J.I. Raine. The available fossil
plant material (wood, leaves, seeds and cone
scales) is probably biased due to its preservation
in shallow wave-base marine sediments — with
only the tougher plant material surviving
transportation by river to the region of deposition.

DINOSAURIAN PALAEOBIOLOGY: A NEW ZEALAND PERSPECTIVE 727

FIG, 2. The Mangahouanga Stream fossil site, showing
a concretion eroding from the Maunataniwha
Sandstone. Dinosaurian, and other, bones have been
recovered from some of these concretions.

However eight types of angiosperm leaves and
several seed capsules have been recognised and
a considerable amount of podocarp-like foliage,
similar to present-day material — though
araucarian wood, leaves and cone scales
dominate the collections (Raine, 1990;
Crampton, 1990). It is hoped that ultimately the
study of this material, combined with pollen and
spore samples, will give a clearer indication of the
botanical environment in which these dinosaurs
survived.

POLAR DINOSAURS?

While it is evident that New Zealand’s dinosaur
population was subpolar in the geographic sense
(Fig. 6) (Molnar & Wiffen 1994), it is not readily
apparent that the climate and conditions in which
they survived during the subsequent 15-20 mil-
lion years were polar in the climatic sense. Water
temperatures of 14.3°C are suggested from
oxygen isotope studies by Stevens & Clayton
(1971) and the marine record shows that
plesiosaurs, mosasaurs and turtles were common
in offshore waters in the Late Cretaceous. On
land, there is evidence of continuous forest and
plant growth — e.g., the Jurassic fossil forest at
Curio Bay, Southland, the plant beds at Port
Waikato, the lignite and plant fossils in both
South and North Islands, material from Shag
Point and Cretaceous material from Putarau,
northwestern Nelson, Kaipara Harbour and Man-
gahouanga Stream (Bose, 1975; Edwards, 1926;
Ettingshausen, 1891; Johnson, 1993; Kennedy,
1993; Mildenhall, 1970; Stevens, 1985) — which
would have been essential to maintain a her-

bivorous dinosaur population, which in turn
provided the food for the carnivorous dinosaurs.

However, palaeobotanical evidence appears to
be accumulating to support the view of a cold
Australasian polar climate, i.e., dominant an-
giosperms that were deciduous, dormant in
periods of cold and darkness and dropping all
their leaves over a short period to form leaf mats
(Vickers-Rich & Rich, 1993; Johnson, 1993),
though Pole (1993) suggests that this is still based
on scant evidence. If true, presumably the
deciduous plants would have, to some extent,
replaced the conifers. From a practical point of
view, such a deciduous vegetation would seem to
have provided a meagre winter diet for the active,
medium-sized dryosaur-like herbivorous
dinosaur, while having the larger herbivores
(sauropods) dependent on foraging for and
browsing on fallen leaf mats to sustain life over
prolonged periods of darkness and low tempera-
ture seems an unlikely scenario. How much nutri-
ment could be derived from cold, semi-frozen,
fallen leaves alone does not appear to have been
calculated, but it was probably substantially
lower than in still-attached leaves, because most
nutrients are withdrawn from leaves before they
are dropped. The feasibility of sauropods of even
modest proportions ‘grazing’ on such a leaf mat
is unclear, and even the question of how
sauropods supported themselves in more
temperate climates is unresolved (Ostrom, 1985).
There is no evidence that herbivorous dinosaurs
roamed New Zealand in large herds, as occurred
in North America at this period, and the balance
of both herbivores and carnivores would have
been ultimately controlled on this island
landmass by the quality, quantity, reproductive
rate and availablity of the vegetation that formed
the food supply. The deciduous vegetation does
not appear to have been adequate to support the
known dinosaurs through a cold polar winter.

Migration was not an option on this relatively
small island landmass (Molnar & Wiffen, 1994).
There is nothing in the dinosaur bones so far
identified from the Mangahouanga Stream site to
indicate dwarfism, as suggested for Victorian
dinosaurs by Vickers-Rich & Rich (1993): a use-
ful strategy if food resources or climatic condi-
tions made hibernation necessary for survival.
But hibernation for the relatively large dinosaurs
(Allosaurus-size theropods and medium-size
sauropods) found in New Zealand seems unlike-
ly. Even remaining stationary for long periods in
groups to conserve warmth would be unlikely for
dinosaurs of this large size, while laying down

728 MEMOIRS OF THE QUEENSLAND MUSEUM

DINOSAURIAN PALAEOBIOLOGY: A NEW ZEALAND PERSPECTIVE

FIG. 4. Pterosaur distal ulna (CD 467) from the Maun-
gatanwha Sandstone, Mangahouanga Stream. Scale
= Icm.

and curling up to keep the extremities warm
would be improbable and finding sufficient
handy caves to accommodate a number of
animals of this size unbelievable. So there are no
obvious behavioral strategies for the dinosaurs to
have coped with a cold polar winter, either.

The difficulty of comparing the environment
and dinosaurs from New Zealand and Australia
arises from the difference in geological ages of
these dinosaur faunas. The Australian dinosaurs
range from Jurassic to Early Cretaceous, while
Australia was still attached to Gondwana,
whereas the only known bones from New
Zealand are Late Cretaceous (Maastrichtian, 65-
70 million years) when New Zealand had been
adrift — and an isolated island — for ap-
proximately 15-20 million years after separation
from Gondwana. It would appear that dinosaurs
either adapted to whatever climatic changes they
encountered over their 15-20 million years of
occupation or the climate on the New Zealand
landmass remained relatively stable and
temperate. Doubtless even with a reasonably
stable and temperate maritime climate, as probab-
ly existed, there were times when temperatures
dropped or rose and conditions were unfavorable
for dinosaurian life — with subsequent decreases
in population size and reproduction at that time,
such as occurs in the wild (and also with domes-
ticated) animals today, but so far as is known of
insufficient magnitude to wipe out any species
before the end of the Cretaceous.

A living example of the resiliency that favours
long-term survival is the Tuatara (Sphenodon
punctatus): to quote Vickers-Rich & Rich (1993),
‘this reptile can live in conditions down to 5°C,
as long as it can sun itself . It must be remem-

729

FIG. 5. Fossil cockroach, about 43mm long, from the
Maungataniwha Sandstone, Mangahouanga Stream.
The head was not found, but the right antenna (a) is
present.

bered that Sphenodon punctatus was a New
Zealand resident prior to and since the separation
from Gondwana and that it apparently survived
everything, including the Pleistocene ice ages
from around 2 million to 14,000 years ago: an
example of endurance over a long period of fluc-
tuating temperature. Admittedly, the relatively
small size and lower metabolic rate of the Tuatara
allowed it to hibernate using holes and tunnels
made by birds, or tree roots, for shelter. On the
other hand the Moas, flightless ratite birds of
large proportions whose ancestors are also
believed to have lived on New Zealand since it
separated from Gondwana, would have had dif-
ficulty in hibernating during the Ice Age, but
survived until shortly after the arrival of man —
around 1,000 years ago. This in spite of their
higher metabolic rates and hence increased sus-
ceptibility to low temperatures and winter food
shortages.

These are just two of a number of the forms

which are known to have survived, regardless of
climatic changes and turbulent geological events,

FIG. 3. A, Plesiosaurian vertebrae exposed in a concretion at the Mangahouanga Stream locality. B, Pedal phalanx
of a large theropod exposed in a concretion at this locality. Both specimens show the characteristic structure of
fossil bone as it is seen in the field and show how the bone can be easily recognised in situ.

730

OX

60S

ay
NS
NZ

FIG. 6. Map of Late Cretaceous Southern Hemisphere
in south polar projection showing the position of New
Zealand (NZ). SP=South Pole.

since New Zealand’s isolation 80-85 million
years ago — though, as yet, there is little fossil
evidence of their long period of residence. So it
is not unexpected that the dinosaurs whether ec-
tothermic or endothermic — considered by some
to be the most successful forms of life on earth —
were capable of surviving whatever changes oc-
curred within their period of occupation of New
Zealand. Especially as these changes seem to
have been substantially less great than those un-
dergone during the Caenozoic.

SUMMARY

Although the New Zealand contribution of
dinosaur fossils is small at this stage, they do
contribute to the overall knowledge of dinosaurs
from the southern supercontinent, their dispersal
and their survival until their worldwide extinction
at the end of the Cretaceous. The fossils show that
dinosaurs did inhabit New Zealand after it
separated from Gondwana early in the Cam-
panian. They are the oldest evidence for ter-
restrial vertebrates in New Zealand, and are the
only near-polar dinosaurs from the Southern
Hemisphere that lived on an island landmass.

The climate of New Zealand at that time seems
to have been temperate, judging from the
dinosaurian and plant fossils. Some plant material
suggests that leaves were dropped during the
winter forming leaf mats although conifers, not
these deciduous forms, still appear to have

MEMOIRS OF THE QUEENSLAND MUSEUM

dominated the forests in this region. The
recovered dinosaur material shows no special
adaptations to cold climates. However, the sur-
vival of Tuataras and Moas through the climatic
variations that occurred in New Zealand since its
separation from Gondwana shows that both en-
dothermic and ectothermic forms were capable of
surviving here for long periods. There is no
reason to think that dinosaurs were not able to do
So, too.

It is hoped that other New Zealand sites yield-
ing Mesozoic terrestrial fossils will be found, and
that more material will add to what we already
know and expand our understanding of
dinosaurian life and survival here on New
Zealand and our place in the evolution of
Gondwanan dinosaurs.

ACKNOWLEDGEMENTS

I would like to thank Mike Pole, I.L. Daniel and
J.P. Lovis for information regarding Cretaceous
plant life in New Zealand, J.I. Raine for his con-
tinuing study of the Mangahouanga Stream site
plant fossils, and R.E. Molnar for his editorial
assistance.

LITERATURE CITED

BOSE, M.N. 1975. Araucaria Haastii Ettingshausen
from Shag Point, New Zealand. The Paleobotanist
22: 76-80.

CRAMPTON, J.S. & MOORE, P.R. 1989, Environ-
ment of deposition of the Maungataniwha
Sandstone (Late Cretaceous), Te Hoe River area,
western Hawke's Bay, New Zealand. New
Zealand Journal of Geology and Geophysics 33:
333-348.

CRAW, R.D. & WATTS, J.C. 1987. An Upper
Cretaceous beetle (Coleoptera) from Hawke's
Bay. Journal of the Royal Society of New Zealand
17: 395-398.

CURRIE, PJ., VICKERS-RICH, P. & RICH, T.H.
1996. Possible oviraptorosaur (Theropoda,
Dinosauria) specimens from the Early Cretaceous
Otway Group of Dinosaur Cove, Australia. Al-
cheringa 20: 73-79.

DANIEL, I.L., LOVIS, J.D. & REAY, M.B. 1989. A
brief introduction on the Mid-Cretaceous
megaflora of the Clarence Valley, New Zealand.
Pp. 27-29. In Douglas, J.G. & Christophel,
M.D.C. (eds) ‘Proceedings of the Third Interna-
tional Organisation of Palaeobotany Conference’.
(Melbourne).

EDWARDS, W.N. 1926. Cretaceous plants from
Kaipara, New Zealand. Transactions of the New
Zealand Institute 56: 121-128.

FELDMANN, R.M. 1993. Additions to the fossil
decapod crustacean fauna of New Zealand. New

DINOSAURIAN PALAEOBIOLOGY: A NEW ZEALAND PERSPECTIVE

Zealand Journal of Geology and Geophysics 36:
201-211.

FLEMING, C.A. 1962. New Zealand biogeography, A
paleontologist’s approach. Tuatara 10: 53-108.

GLAESSNER, M.F. 1980. New Cretaceous and Ter-
tiary crabs (Crustacea: Brachyura) from Australia
and New Zealand. Transactions of the Royal
Society of South Australia 104: 171-192.

HAMMER, W.R. & HICKERSON, W.J. 1994. A
crested dinosaur from Antarctica. Science 264:
828-830.

HOOKER, J.J., MILNER, A.C. & SEQUERIA, S.E.K.
1991. An ornithopod dinosaur from the Late
Cretaceous of West Antarctica. Antarctic Science
3: 331-332.

ISAAC, M.J., MOORE, P.R. & JOASS, YJ. 1971.
Tahora Formation: the basal facies of a Late
Cretaceous transgressive sequence, northeastern
New Zealand. New Zealand Journal of Geology
and Geophysics 34: 227-236.

JOHNSON, K. 1993. Ancient leaves write history of
forests. National Geographic 148, 2
(‘Geographica’ column: not paginated)

KENNEDY, L.M. 1993. Palaeoenvironment of an
Huamaurian plant fossil locality, Pakawau Group,
N.W. Nelson. Thesis.

MILDENHALL, D.C. 1976. Checklist of valid and
invalid plant macrofossils from New Zealand.
Transactions of the Royal Society of New
Zealand, Earth Scienes 8: 77-89.

MOLNAR, R.E. 1981. A dinosaur from New Zealand.
Pp. 91-96. In Vella, P. & Creswell, M. (eds)
*Gondwana Five'. (A.A. Balkema: Rotterdam).

1996, this volume.. Observations of the Australian
ornithopod dinosaur, Muttaburrasaurus.
Memoirs of the Queensland Museum 39(3): 639-
652.

731

MOLNAR, R.E. & WIFFEN, J. 1994. A Late
Cretaceous polar dinosaur fauna from New
Zealand. Cretaceous Research 15: 689-706.

OSTROM, J.H. 1987. Romancing the dinosaurs. The
Sciences 27: 56-63.

POLE, M. 1992. Cretaceous macrofloras of Eastern
Otago, New Zealand. Australian Journal of
Botany 40: 169-206.

RAINE, J.I. 1990. Late Cretaceous plant macrofossils
from Mangahouanga Stream. P. 111. In
Campbell, H.J., Hayward, B.W. & Mildenhall,
D.C. (eds) ‘Geological Society of New Zealand
Annual Conference, Napier 1990, Programme and
Abstracts'. (Geological Society of New Zealand:
Lower Hutt).

SCARLETT, R.J. & MOLNAR, R.E. 1984. Terrestrial
bird or dinosaur phalanx from the New Zealand
Cretaceous. New Zealand Journal of Zoology 11:
271-275.

STEVENS, G. 1980. ‘New Zealand adrift.” (A.H. &
A.W. Reed: Wellington), 442 pp. (Pp. 226-231)

VICKERS-RICH, P. & RICH, T.H. 1993. Australia’s
polar dinosaurs. Scientific American 269(1): 50-
55.

WIFFEN, J. 1981. The first Late Cretaceous turtles from
New Zealand. New Zealand Journal of Geology
& Geophysics 24: 293-299,

WIFFEN, J. & MOLNAR, R.E. 1988. First pterosaur
from New Zealand. Alcheringa 12: 53-59.

1989. An Upper Cretaceous ornithopod from New
Zealand. Geobios 22: 531-536.

WILSON, G.J. & MOORE, P.R. 1988. Cretaceous-Ter-
tiary boundary in the Te Hoe River area, western
Hawke’s Bay. New Zealand Geological Survey
Records 33: 34-37.

CONTENTS
PART 1 (Issued 1 May, 1996)

BENNETT, T.

The museum and the citizen.: ..-.i-..--1 Mrd er NAN ciae uela ar rtr ee i o ae eT pa !
TROTTER, R.

Women in museums: the shift from second-class to full citizenship. . DERART LT

McALEAR, D.

First peoples, museums and citizenship. 22.02... isses sonne atgyti iiis 79
TROTTER, R.

From monoculture to multiculture: the response of Australian museums 1o multiculturtism , |, 47
TROTTER, R. m

Museums: access and participation 6... 2... ce eee cc mane 115

PART 2 (Issued 20 July, 1996)

ARNOLD, P.W. & HEINSOHN, G.E.
Phylogenetic status of the In'awaddy Dolphin Orcaella brevirostris (Owen in Gray);
acladistic analysis bryan iaa sees eee be ebb sede es tensed baa ees 141
BRAILOVSKY, H. & MONTEITH, G.B.
A new species of Pomponatius Distant from Australia pe Het. higine
reje itan thocorini)...... TET II PME S
BULL, K.H. & WHITTIER, J
Annual pattern of ani of the Brown Tree Snake (Boiga irregularis) in southeastern
Queensland teamed eet eee cps bk tte AER Ur e d nme oda 483
BURWELL, CJ.
_ Revision of the Australian genus Microrropesa Macquart (Diptera: Tachinidae: Tachinini).... 211
CLIFFORD, H.T.
Geometrical study of a case of Leptophloeium australe (McCoy) Walton from Queensland . . ., 227
COUPER, eh poche: a J.A., MARSTERSON, S.P, & SHEA, G.M,
ria naufragus inm et sp.nov., a sand-swimming skink from Fraser Island, Queensland ., 233
COVACEVICH Ni A. & COUPER.
Aspidites ramsayi (Boidae); inthe Brigalow Biogeographic Region of Queenslands

occurrence, conservation status and ie bibly associations. .......s een 243
COVACEVICH, J.A., COUPER, P.J. & MCDONALD, K.R
Lerista allenae (Scincidae: Lygosominae): 60 years from exhibition to extinction? _........- 247

"Two new species of ‘false spider crabs (Crustacea: Brachyura: Hymenosomatidac)
from New Caledonia ~. o.. se aea e mte c cece eee Rm mh 257
NG, P.K.L. & DE FORGES, R.
The Hymenosomatidae (Crustacea: Decapoda: Brachyura) of New Caledonia, with
descriptions of two new genera and two new species... sse ee eee eee eee 263
DAVIE, P.J.F. & GUINOT, D.
Two new freshwater crabs in Australocarcinus Davie, with remarks on the Trogtoplacinae
Guinot and Goneplacidae MacLeay (Crustacea: Decapoda: Brachyura)............... 277
JAMIESON, B.G.M. & GUINOT, D.
Ultrastructure of the spermatozoon of Australocarcinus riparius (Crustacea: Brachyura:
Gonoplacidae: Trogloplacinae) |... eer e rn 289
GARDZINSKA, J.
New specics and records of Astieae (Arancae: Salticidae) from Australia and Papua
New Guinea siasi ironi Ld ad nies Ti IE TEST potatoe T 297
GULLAN, PJ. & STEWART, A.C.
A new genus and species of ant-associated coccid aime ise Coccidae: = Myzdecsoiitun]
from Canthium Lam. (Rubiaceue)., .. .. Setty ca raia 307
HEALY, J.M. & LAMPRELL, K.L.
The Atlantic-Mediierranean Mah. Syria gibba {Oliv Toortalidecs Myon in

Port Phillip Bay, Victoria. , vo (2a. 315
MANNING, B. & KOFRON, C.P.
Evolution and. zoogeography of Australian freshwater turtles... 0.0.0... esee ese, 319
PATERSON, R.A. & VAN DYCK, S.
Perinatal skeletal injuries in em balaenopterid whales... eee nn 333

SEEMENTS. K.D. & RANDALL, J.B.
our new records of surgeonfishes (Perciformes: Acanthuridae) from the Great Barner Reef .. 339
STANISIC., J. J

New land snails from Boggomoss environments in the Dawson SM socastpm
Queensland (Eupulmonata: Charopidae and Camaenidae) , . ; ei ales A

STANISIC, J.
A new camaenid land snail from the Wet Tropics Biogeographic Region, northeastem
Queensland (Eupulmonata: Camaenidae) ....- 2.2... ee eee eee eee eee eee 355
STOREY, RI, & HOWDEN, H.F.
Revision of Australoxeriella Howden & Storey in Australia (Coleoptera: Scarabaeidac:

Aphodiinae)
TAFE, DJ. a GREENWOOD, J.G:
A. new species of Schizotrema (Cumacea: Nannastacidae) from Moreton Bay, Queensland .... 381
TAFE, D.J. & GREENWOOD, J.G,

The Bodotriidae (Crustacea: Cumacea) of Moreton Bay, Queensland -.....-..-...-..--.-- 391
NOTES x
LEE, M.S.Y. i

Possible affinities between Varanus ane and 1 Megalanie prisca |. sess 232
COUPER, P.J., COVACEVICH, J.A. & MCDONALD

A Bandy Bandy with a difference 2.0... ccc cece k beet ete t eme eme 242
COUPER, PJ., COVACEVICH, J.A., MONTEITH, G.B., JAGO, K. JANETZKI, H. & ROBERTS, L,

Feeding habits of the rin p-lailed gecko, Cyrtodactylus louisiadensis .. ........ cessisse 288
HERO, J.M. & FICKLING, S.
IRWIN A: Reproductive characteristics of female frogs from mesic habitats in Queensland ..........., 306

Survival of a large Crocodylus proosus despite significant lower jaw loss |... ... iiio. 338
BURNETT, S. & NOLEN, J.

Fruit eating by the gecko Gehyra dubia in Townsville 2... 20.00.60. o eee ee ee e 364
DEER, R.

Reptile diversity in a Callitris forest in central Queensland's brigalow belt ...............- 390
COUPER, P.J.

Nephrurus asper (Squamata; Gekkonidae): sperm storage and other reproductive dala ....... 487
PART 3 (Issued 20 December, 1996)

PROCEEDINGS OF THE GONDWANAN DINOSAUR SYMPOSIUM

CHATTERJEE, S. & RUDRA, D.K.
KT events in India: impact, rifting, volcanism and dinosaur extinction .. 2... .......2..-24- 489
CHIAPPE, L.M.
Early avian evolution in the Southem Hemisphere; the fossil record of birds 3n the Mesozoic
Of Gondwana . oo... cece ee nee phele-pelepicp sihes etehi oriol o $33
CHIAPPE, L.M., NORELL, M.A. & CLARK, J.M.
Phylogenetic position of C Mene (Aves: Alvarezsauridae) from the Late Cretaceous
of the Gobi Desert qM" 557
GASPARINI, Z., PEREDA SUBERBIOLA. X. & MOLNAR, R.E.
New data on the ankylosaurian dinosaur from the Late Cretaceous of the Antarctic Peninsula. . 583
JACOBS, L.L., WINKLER, D.A. & GOMANI, E.M.

Cretaceous dinosaurs of Africa: examples from Cameroon and Malawi ,.., . - CIC Eds f 595
KELLNER, A.W.A.

Remarks on Brazilian dinosaurs... sees mr IRR IRI 611
LOYAL, R.S., KHOSLA, A. & SAHNI, A.

Gondwanan dinosaurs of India: affinities and palacobiogeography -.......-.. 2.422.222 eee 627
MOLNAR, R

Observations on the Australian ornithopod dinosaur, Muttaburrasauras , . TIBUS 639
MOLNAR, te Ben

ary report on a new ankylosaur from the Early Cretaceous of Queensland, Australia .. 653
MOLNAR, "e E. LOPEZ ANGRÍMAN, A. & GASPARINI, Z.

An Antarctic Cretaceous theropod ,.......2.., essen Penne eee thins de thea sate cee be oe 669
NOVAS, F.E.

Alvarezsauridae,Cretaceous basal birds from Patagonia and Mongolia .....-.4...02.2..22, 615
RAATH, M.A.
— Earliest evidence of dinosaurs from Central Gondwana .-....-.--.-..-+ vary treme ned 703

` Significance of polar dinosaurs in Gondwana . . eei atl rre aar E a Sr ET

VICKERS-RICH, P.
Early Cretaceous polar tetrapods from the Great Southem Rif Valley, sem

Austral. eed atdala ai n e ma ries ible ea da ueri Da. a. o. 719
WIFFEN, J,
Dinosaurian palacobiology: a New Zealand perspective... esee the deea nee 725

CONTENTS

CHATTERJEE, S. & RUDRA, D.K.

KT events in India: impact, rifting, volcanism and dinosaur extinction ..............200065
CHIAPPE, L.M.

Early avian evolution in the Southern Hemisphere: the fossil record of birds in the

Mesozoic of Gondwana). .... 5 ccc cc cece ease oen S ATE A EA hera bend
CHIAPPE, L.M., NORELL, M.A. & CLARK, J.M.

Phylogenetic position of Mononykus (Aves: Alvarezsauridae) from the Late Cretaceous

Ge Ihe GOW DEI oie, canes cena sees ee Leas eka ton eh SR ee er la

GASPARINI, Z., PEREDA-SUBERBIOLA, X. & MOLNAR, R.E.

New data on the ankylosaurian dinosaur from the Late Cretaceous of the Antarctic Peninsula. .
JACOBS, L.L., WINKLER, D.A. & GOMANI, E.M.

Cretaceous dinosaurs of Africa: examples from Cameroon and Malawi ............... suu.
KELLNER, A.W.A.

Remarks on Brazilian dinosaurs... .......... 0000s cece cece ehm hh hen
LOYAL, R.S., KHOSLA, A. & SAHNI, A.

Gondwanan dinosaurs of India: affinities and palaeobiogeography ................200000-
MOLNAR, R.E.

Observations on the Australian ornithopod dinosaur, Muttaburrasaurus.......... lese.
MOLNAR, R.E

Preliminary report on a new ankylosaur from the Early Cretaceous of Queensland, Australia . .
MOLNAR, R.E., LOPEZ ANGRIMAN, A. & GASPARINI, Z.

‘An Antarctic Cretaceous theropod POTUM IC aR eae vtr
NOVAS, F.E.

Alvarezsauridae,Cretaceous basal birds from Patagonia and Mongolia ................ LL.
RAATH, M.A.

Earliest evidence of dinosaurs from Central Gondwana .........sssessssssseseoesossess
RICH, T. |

Significance of polar dinosaurs in Gondwana ...........sessssssesssesosussssesessese
VICKERS-RICH, P

Early ied polar tetrapods from the Great Southern Rift Valley, southeastern

PAGINA ies i. ob croce ca EEE ede ea WU OLI ITE SEP RU TRIER
WIFFEN, J.
Dinosaurian palaeobiology: a New Zealand perspective... 2... 2 oe eee ee eee