Postilla

SOS SL
P37
y Museu
al History
iversity
ven, CT 06511

LEE

Postilla umber 197

14 April 1986

(Received 25 October 1984)
Abstract

A nearly complete vertebral column and
portions of an associated pelvis and hind limb
from Sind Province, southern Pakistan, are
described and identified as those of a marine
crocodile belonging to the poorly known
family, Dyrosauridae. The fossil is the most
complete of its kind yet known from Asia and
adds to our knowledge of the postcranial
anatomy of the dyrosaurs. Its occurrence in
the late Paleocene (Thanetian) Lakhra
Formation places it among the earliest of
Asian dyrosaurid crocodiles and strengthens
the view that the Dyrosauridae spread from
Africa to Asia along the shores of Tethys. The
source rocks are of marine origin. Remains of
dyrosaurid crocodiles are generally restricted
to marine sediments of Africa, North America,
and South America but have been only rarely
found in marine rocks of Asia.

Key Words

Crocodylia, Dyrosauridae, Mesosuchia,

Pakistan, Paleocene, Sind Province, Thanetian.

© Copyright 1986 by the Peabody Museum of
Natural History, Yale University. All rights reserved.
No part of this publication, except brief quotations
for scholarly purposes, may be reproduced without
the written permission of the Director, Peabody
Museum of Natural History.

A Dyrosaurid Crocodile
(Crocodylia: Mesosuchia) from the

Paleocene of P 4 HSON ay

APR 295 1986

\

Glenn W. Storrs

Introduction

During the period of 23 December 1981 to 1
January 1982 a joint field expedition from the
Yale Peabody Museum of Natural History and
the Geological Survey of Pakistan (GSP) under
the direction of J. D. Archibald and

H. Shaheed conducted a geological and
paleontological survey of the Paleocene rocks
of the Ranikot Group in the northern Lakhi
Range of the Lower Indus Basin, Pakistan.
This work was carried out under the auspices
of the University of Michigan Museum of
Paleontology and the Geological Survey of
Pakistan as part of a continuing survey of
Paleocene and Eocene continental sediments
in Pakistan. The Yale-GSP field area
comprised three ephemeral stream valleys,
known locally from north to south as Rahman
Doro (which dissects Bara Dome or the
Lakhra anticline), Bara Nala (Bara Nai), and
Lohige Nala (including, south of the main
drainage, its tributary Barhi Nala). These
valleys lie in the foothills at the northern end
of the Lakhi Range, a group of north-south
trending mountains in the Dadu District of
Sind Province at the southern end of Pakistan
(Fig. 1). During the course of field
investigations, the articulated vertebral column
of a large crocodilian was discovered and
collected from the field area (Fig. 2)
approximately 10 kilometers due east from the
British colonial rest house settlement of Amri
and the west bank of the Indus River, and
approximately 170 kilometers from and 27°
northeast of Karachi. Approximate coordinates
of the field area are 26°10'N latitude; 67°53’E
longitude.

2 A Dyrosaurid Crocodile from
the Paleocene of Pakistan

Postilla 197

34° eS
, Peawacoi st aibagl
AFGHANISTAN d : ae

30°

wal

IRAN

26°

ss

64°

ARABIAN SEA

68°

Fig. 1
Outline map of Pakistan. Northern Lakhi Range is
located within circle and detailed in Fig. 2.

Geologic Setting

Several stratigraphic units are exposed in the
field area. The disposition, description, and
nomenclatural history of these formations has
been discussed in detail by numerous authors
(e.g., Gingerich et al. 1979; Hunting Survey
Corporation 1961; Khan 1956; Shah 1977;
Vredenburg 1909, 1928; Williams 1959; etc.).
Lowermost in the section lies the Pab
Sandstone of latest Cretaceous
(Maastrichtian) age. Above this are found the
three formations of the Ranikot Group, oldest
of which is the Khadro Formation of Williams

JACOBABAD

ei a
es

fi
ona

Zi

7 |.

INDIA

(1959) (‘‘Cardita beaumonti beds" of Blanford
1876; “‘Daphro Beds’’ of Gingerich et al.
1979). On the basis of abundant foraminifera,
the Khadro has been labeled Late Cretaceous
to (primarily) early Paleocene or ‘‘Danian”’ in
age (Shah 1977). The top of the Khadro
Formation is marked by a basaltic trap
possibly related to the Peninsular India
Deccan Vent volcanism of the early Tertiary.
The specific flow found in the Khadro has not,
to my knowledge, been dated. Above the
“Deccan trap”’ lie the Paleocene Bara and
Lakhra formations, respectively, which are
unconformably overlain by the Eocene Laki

Sr nn nn nnn nn eat Et ttttEIIy II nISSSSSISIS SSS

3 A Dyrosaurid Crocodile from
the Paleocene of Pakistan

Postilla 197

Fig. 2

Detail of Fig. 1 with approximate topographic
contours in meters above sea level. Approximate
locality (W1) of dyrosaur fossil, GSP No. 1020, is
indicated by triangle.

Formation (Shah 1977). The Deccan trap
basalt and the resistant, cliff-forming limestone
of the Laki Formation present convenient and
easily recognizable markers that delimit the
local extent of the Bara and Lakhra
sediments.

While the primary research interest of the
Yale-GSP field party centered around the
Paleocene Bara Formation (Lower Ranikot
Formation of Hunting Survey Corporation

/ Ranikot©

i

20km

1961), the overlying Lakhra Formation (Upper
Ranikot Formation) was also surveyed. The
Lakhra Formation is approximately 242 m thick
at the type section on the southern flank of
the Lakhra anticline (Rhaman Doro) and
invertebrate fossils from its richly fossiliferous
sediments indicate a late Paleocene
(Thanetian) age (Shah 1977). The Lakhra
Formation consists of alternating calcareous
sands, silts, shales, and dominant foraminiferal

4 A Dyrosaurid Crocodile from
the Paleocene of Pakistan

Postilla 197

Fig. 3

Barhi Nala, Sind Province, Pakistan. Khadro
Formation shales and dyrosaur locality (W1) lie in
foreground. Cliff-forming Laki limestone rises in
distance.

limestones of varying thicknesses. All
gradations between limestone and shale can
be found. The sands are varicolored and
poorly sorted, fine to coarse grained, thin to
thick bedded, with cross-stratification and
occasional ferruginous nodules. The shales,
are clayey, gypsiferous, and poorly indurated.
The characteristic limestones are richly
fossiliferous, thin to thick bedded, nodular,
brecciated, and arenaceous. They are
generally brown weathering (Shah 1977).
Although of marine origin at the base, the
Bara Formation has generally been regarded
as predominantly representing fluvial
environments, owing in part to the paucity of

invertebrate fossils (Gingerich et al. 1979). The

Lakhra Formation is unquestionably marine,
however, and suggests frequent, short-lived
regressions of intertidal mudflat and estuarine
environments between episodes of
transgression and offshore marine deposition.

On 30 December 1981, while prospecting
in Barhi Nala within Lohige Nala, J. D.
Archibald discovered the skeleton of a fossil
crocodile in marine sediments of Paleocene
age (Yale-GSP locality W1) which are
unquestionably part of the Lakhra Formation.
These were greensand rocks situated some
distance (although unmeasured) below the
obvious cliffs of the Laki limestone (Figs. 3, 4)
and presumably above the Deccan trap. The
Deccan trap is believed to lie subsurface in
this area. The locality consisted of nearly
horizontal beds of soft, brown and
greenish-gray, gypsiferous shales. A
NNW-to-SSE-trending fault separated these
rocks from steeply dipping (about 60°) beds
to the east. Approximately 150 m to the west,
limestones were discovered approximately 15
m below the shales. The poorly exposed
strata above the locality were also dominated
by limestones (Yale-GSP field notes). Fossil

5 A Dyrosaurid Crocodile from
the Paleocene of Pakistan

Postilla 197

Fig. 4
Detail of dyrosaur locality (W1) in Barhi Nala. Laki
Formation cliffs beyond.

invertebrates recovered from the matrix and
surrounding area match those collected by
the Yale team in the Lakhra Formation at
Rhaman Doro and include Ostrea sp.,
Venericardia sp., Crassatella sp., ?Conus sp.,
?Turritella sp., numerous indeterminate
bivalves and gastropods, solitary
scleractinians, worm tubes, and fragments of
crab shell. Shark teeth, fish vertebrae, and
rare carbonized plant debris were also found
at both localities.

Preservation of the Fossil

The skeleton was discovered weathering from
the side of a small hill. A large section of the
articulated cervical and thoracic regions of the
vertebral column was separated from the tail
section by a break in the column. These
sections lay in a coil which extended into, and
was excavated from, the hillside (Figs. 5, 6).
The sacral vertebrae and pelvis had been

6 A Dyrosaurid Crocodile from
the Paleocene of Pakistan

Postilla 197

Fig. 5

Dyrosaur vertebral column, GSP No. 1020, in situ,
after partial excavation. Left lateral surface
exposed. Caudal vertebrae to left. Break in column
at top.

destroyed by erosion, but parts of some of
these elements, ribs, and portions of a hind
leg were collected below the site as float. No
part of the skull was found.

The condition of the bones is only fair and
although harder than the crumbly siltstone
matrix, they are delicate, easily broken, and
badly fractured. Occasionally, the common
veins of secondarily deposited gypsum have
entered and distorted individual bones. The
skeleion is for the most part encrusted with
an extremely hard, ferruginous, concretionary
layer averaging approximately 5 mm in
thickness. Although the concretion has been
mechanically removed from the bone in some
places, it has often made detailed study of
specific features difficult. However, the
concretion takes the general form of the
bones and so with the use of composite
reconstructions incorporating fully prepared
areas, little information about the external
appearance of the bones has been lost. The

configuration of the vertebrae indicate the
skeleton is that of a dyrosaurid crocodile.
Most strikingly, the vertebrae are platycoelous,
indicating that the animal was a mesosuchian
in the classical sense (although the
Mesosuchia is likely to be an artificial
grouping). Additionally, the fossil was
discovered, presumably untransported, in
marine sediments of Paleocene age. The
Dyrosauridae represent the last radiation of
marine mesosuchians (Buffetaut 1981) and the
only family of mesosuchians (possibly
excluding the problematic terrestrial
Sebecidae of South America) to survive past
the Cretaceous—Tertiary boundary. While it is
as yet virtually impossible to identify isolated
postcrania of the dyrosaurs to genus, as the
postcranial skeleton exhibited very little
variation within the family (Buffetaut, personal
communication, 1984; Moody and Buffetaut
1981), the completeness of this fossil warrants
its description. The fossil is number 1020 in

A Dyrosaurid Crocodile from Postilla 197
the Paleocene of Pakistan

Fig. 6

lllustration of dyrosaur vertebral column, GSP
No. 1020, after preparation. Right lateral surface
shown. Caudals at top. Bar = 10 cm.

8 A Dyrosaurid Crocodile from
the Paleocene of Pakistan

the collection of the Geological Survey of
Pakistan.

Description

ORDER Crocodylia GMELIN, 1788
SUBORDER Mesosuchia HUXLEY, 1875
FAMILY Dyrosauridae DE STEFANO, 1903

The total length of the anterior series of
articulated vertebrae is approximately 135 cm.
This series begins with the fused atlas
centrum and axis intercentrum. The neural
arch and the intercentrum of the atlas are
missing. This is followed by the axis, seven
additional cervical vertebrae (totaling nine as
is typical for most crocodilians), and fourteen
dorsals. Typical cervical ribs remain articulated
with the third through seventh cervicals.
Seventeen thoracic ribs lie across the
vertebral column, some articulated with their
transverse processes. A few isolated rib
fragments were also collected. The second
group of articulated vertebrae consists entirely
of caudals and is approximately 60 cm long.
Eleven proximal caudals are present. The first
ten are complete with their respective
chevrons. All of the preserved caudal
vertebrae have transverse processes. From
between the two vertebral series an isolated
dorsal vertebra and a single sacral vertebra
are preserved. The second sacral is missing,
along with perhaps one posterior dorsal, one
or more anteriormost caudals, and the greater
part of the tail’s extremity.

The fact that most of the spines and all of
the arches are intact throughout the column is
noteworthy. Dyrosaurid vertebrae are usually
found as isolated centra and it is very rare for
the neural arch to remain attached. This is
due to the frequently weak neurocentral
sutures of dyrosaurs (Buffetaut 1978b).

Part of the pelvic girdle and right hind leg
are present. These are composed of the
articulated right ilium and proximal part of the
right ischium, three fragments of the right
pubis, the acetabular portion of the left ilium,
the proximal end of the right tibia, and the
right astragalus. A single crocodilian tooth

Postilla 197

fragment was discovered in the matrix of the
skeleton. However, as other crocodile teeth
were found in the vicinity, this is possibly from
a separate individual.

Atlas-axis Complex

The atlantal centrum forms a small,
wedge-shaped lozenge approximately 3 cm
wide and 3 cm high. It is fused with the axial
intercentrum. This bone is thickest dorsally. It
is articulated with the anterior face of the
much larger axis centrum. The centrum of the
axis is approximately 5 cm long, 3 cm wide,
and carries a neural arch with a long (7.5 cm),
low neural spine, which projects past the
posterior end of the centrum. In this regard,
the vertebra resembles the axis of the living
Gavialis gangeticus. Indeed, the atlas-axis
complex of Gavialis is closely similar to that of
the fossil, except for the platycoelous nature
of the latter.

Third through Ninth Cervical Vertebrae

The centra of these vertebrae are cylindrical
and platycoelous. The eighth cervical centrum
iS approximately 4 cm high, 4 cm wide, and
5.5 cm long. The capitular facets of the
vertebrae are situated low on the centrum and
the tubercular facets lie at the tip of the
transverse processes of the neural arches.
The ribs of vertebrae 3 through 7 are short
and bladelike and resemble those of Gavialis.
Each is supported by a V-shaped brace
formed by the capitulum and tuberculum. The
‘blade’ projects both craniad and caudad
parallel to the axis of the column.

The posterior two cervical vertebrae are
transitional to the dorsals and bear
parapophyses that are located progressively
higher on the centra than those of the anterior
cervicals. The diapophyses are long and
stout. The ribs of these transitional vertebrae
resemble those of the dorsals as they are
long and robust. The posterior two cervical
vertebrae also possess strong, keellike
hypapophyses much as in other dyrosaurs, for
example, Hyposaurus from the Cretaceous of
New Jersey (Owen 1849) and Brazil (Cope

9 A Dyrosaurid Crocodile from
the Paleocene of Pakistan

Postilla 197

1886) and the Paleocene of Mali (Buffetaut
1980; Dollo 1914; Swinton 1930). Those of the
Pakistani crocodile are swept back,
subtriangular in lateral outline, and
approximately 2.5 cm long from base to tip. In
contrast, the small hypapophyses of the
modern Alligator mississippiensis are present
on all cervicals and point craniad. Similarly, all
parapophyses of Alligator are located near the
anterior end of each centrum, whereas in the
fossil each is found at the approximate
midpoint.

It is in the neural spines that differences
from other crocodilians are most evident. The
spines of vertebrae 3 through 7 are short and
narrow with rounded tops, whereas the last
two cervicals have very tall (maximum 12 cm),
rectangular spines. The spines of Hyposaurus
are more rounded. The tall spines of the
Pakistani crocodile begin abruptly and then,
together with the anterior dorsal spines,
gradually decrease in size caudad.
Dyrosaurids are distinguished from other
crocodilians by such tall cervical spines and
Buffetaut (1979a) has suggested that they
provided surfaces of attachment for powerful
neck muscles from the occiput. Dyrosaurids
had large, longirostrine skulls and a strong
nuchal ligament at this point probably
supported the heavy head.

Dorsal Vertebrae

The dorsal vertebrae continue the pattern
observed in the cervicals in that long,
rectangular, anterior neural spines become
progressively shorter toward the middle of the
column where they become nearly equal
(approximately 6 cm) in height. Throughout
the column each spine is uniformly thin with
little or no lateral thickening at its top. The last
few preserved dorsals have neural spines that
are directed somewhat craniad. These spines
are also rectangular in outline. These last
dorsals represent the lumbar region because
they have stout transverse processes but no
associated ribs. The long transverse
processes of the anterior dorsals are swept
backwards at an angle of approximately 45°
from the perpendicular. This angle decreases

posteriorly. The dorsal diapophyses of
Alligator are, on the other hand, all set nearly
perpendicular to their centra.

The anterior dorsal centra of the fossil are
only slightly taller than they are wide. The
centrum of the fourth dorsal is 3.5 cm wide,
4.5 cm high and 5 cm long. Caudad the
centra become more nearly cylindrical. The
thirteenth dorsal is 4 cm wide, 4 cm high, and
5 cm long. Well-developed hypapophyses are
characteristic of the anterior dorsals of
dyrosaurs, and they are present on the first
three dorsals of the fossil. These are similar to
those of the cervicals. Vertebrae 13 and 14
each have a small hypapophysis. The
parapophyses of the first two dorsals lie
along, and are divided by, faint neurocentral
sutures. There are no uncinate processes on
any of the associated dorsal ribs.

Sacral Vertebra

This isolated bone, though poorly preserved,
is large and robust. The centrum is
approximately 5.5 cm long and 4.5 cm wide,
but is only 3.5 cm high. It is thus semielliptical
in cross section. The articular faces of the
centrum are expanded. A single fused sacral
rib, (7 cm) long and stout, is intact.

Caudal Vertebrae

These vertebrae are characteristically
dyrosaurian and strikingly similar to those of
Hyposaurus. They indicate a powerful, laterally
compressed tail. The centra are deeper than
they are broad (3.5 cm by 3 cm, respectively,
in the third preserved vertebra); preserved
centra average approximately 4.5 cm in
length. The articular surfaces are
subrectangular and the ventral surfaces are
deeply concave. Prominent chevron facets are
present at both ends of the centra. The
chevrons are long and rectangular as are the
neural spines. The complete chevron of the
third preserved caudal is 13 cm in length
whereas that of the tenth is only about 6 cm
long. The spine of the sixth preserved caudal
is approximately 7 cm tall.

$n

10 A Dyrosaurid Crocodile from Postilla 197
the Paleocene of Pakistan

Fig. 7
Right ilium and ischium, GSP No. 1020. A, medial
aspect; B, lateral aspect. ac, acetabulum; //, ilium; is,
ischium; sa, sacral articulation. Shaded portions are
reconstructed.

11 A Dyrosaurid Crocodile from
the Paleocene of Pakistan

Postilla 197

Siem

Fig. 8

Right pubis, GSP No. 1020, dorsal aspect. Proximal
end at top. Shaded portions reconstructed after
Hyposaurus.

Pelvic Girdle

The ilium and ischium (Fig. 7), except for their
larger sizes, are little different from those of
Alligator. Slight differences are found in the
shape of the acetabulum and the high
anterodorsal spur, or costalis tubercle, on the
ilium of the fossil. This spur, and indeed the
entire pelvis, is virtually identical to that of the
Hyposaurus specimen described by Troxell
(1925). As in Hyposaurus, the acetabular

perforation is deeply incised into the ischium,
and although filled with concretionary material,
the anterior opening of the acetabulum
appears to be more nearly closed by the ilium
and ischium than in Alligator. The right pubis
is represented by three fragments which
indicate a very long (possibly 20 cm),
spatulate blade as in Hyposaurus (Fig. 8), and
not the short, broad type found in Alligator.
The longest dimension of the right ilium is
approximately 15 cm. The posterior tip of the

12 A Dyrosaurid Crocodile from Postilla 197
the Paleocene of Pakistan

Fig. 9

Right tibia, GSP No. 1020. A, extensor aspect; B,
flexor aspect; C, proximal aspect, anterior to top; D,
posterior aspect; E, anterior aspect. Bar = 5 cm.

13 A Dyrosaurid Crocodile from
the Paleocene of Pakistan

bone is missing. The breadth of the right
ischium at the iliac articulation and across the
acetabulum is approximately 8 cm.

Tibia

The proximal end of the right tibia (Fig. 9) has
a maximum diameter of approximately 4.5 cm.
The articular region is well formed, indicating
an adult animal. The head is robust and the
two areas of contact with the femoral
condyles are individually distinct. The posterior
articulation is much larger than that of
Alligator, creating a more nearly triangular
proximal surface. This surface is much higher
anteriorly than it is posteriorly and laterally. A
well-defined trough is present between the
condylar articulations on the flexor surface of
the bone. On the extensor surface the
cnemial crest is prominent.

Astragalus

The astragalus is not unlike that of Alligator.
The ball joint for articulation with the first
metatarsal is well defined, as are the peg for
the calcaneum, the facet for the tibia, and the
contact with the fibula. The bone is
approximately 4.5 cm long.

Discussion

The late Mesozoic and Paleogene
Dyrosauridae is a family of poorly known,
longirostrine mesosuchian crocodilians. Most
were large (up to 9 m in length), and they
were probably powerful swimmers primarily
adapted to life in the littoral marine
environment. They are so far known from the
Maastrichtian through the Eocene, most
notably from northern Africa. During the
Paleogene, dyrosaurid crocodiles diversified,
exhibiting several combinations of jaw
structure and dentition (Buffetaut 1979a).
Thevenin (1911) first proved the crocodilian
affinities of the familial type, Dyrosaurus
Pomel, 1894, from the early Tertiary
phosphate beds of southern Tunisia (see also
De Stefano 1903; Nopcsa 1905; Sauvage

Postilla 197

1904; Thomas 1893). Additional dyrosaur
fossils representing several taxa (e.g.,
Atlantosuchus, Dyrosaurus, Hyposaurus,
Phosphatosaurus, Rhabdognathus,
Sokotosuchus, Tilemsisuchus) have since
been collected from marine deposits in
Algeria, Angola, Egypt, lvory Coast, Mali,
Morocco, Niger, Nigeria, Saudi Arabia,
Senegal, and Togo (Buffetaut 1979a).
Bergounioux (1955, 1956), Buffetaut (1976a,
1976b, 1978c, 1979a, 1979b, 1980), Buffetaut
and Wouters (1979), Buffetaut et al. (1982),
Halstead (1975), Madden et al. (1979), Moody
and Buffetaut (1981), and others have studied
and discussed many of these specimens.
Buffetaut has also shown that Hyposaurus
from New Jersey, Brazil, and Mali

(“ Congosaurus,”’ ‘‘Sokotosaurus,”’
‘‘Wurnosaurus’’) is a dyrosaurid (1976a, 1980),
and that indeterminate dyrosaurs occur in
Pakistan (1977, 1978b) and possibly also in
Burma (1977, 1978a).

The occurrence of Asian dyrosaurid
crocodile remains has been reviewed by
Buffetaut (1978a, 1978b). One caudal vertebra
of a possible dyrosaur is known from the late
Eocene of Burma. It is from a freshwater
deposit and is perhaps the latest and the
easternmost dyrosaur occurrence known
(Buffetaut 1978a). However, the caudals of
certain terrestrial mesosuchians are quite
similar (Buffetaut, personal communication,
1984). Asian marine dyrosaurids are
surprisingly rare. With the exceptions of a
single vertebra from an Eocene oyster bed
limestone of Punjab (northern Pakistan) and
one dorsal and one caudal centrum from the
‘Danian’ Khadro Formation of the Lakhi
Range (described by Lydekker 1879), all
previously known Pakistani dyrosaur fossils
have come from freshwater deposits
(Buffetaut 1978b). Virtually every specimen
from Africa, North America, and South
America is known to be from marine rocks.

Dyrosaur fossils from freshwater deposits
in Pakistan are numerous, but usually consist
of isolated vertebral centra, occassional skull
and jaw fragments, scutes, and teeth,
suggesting considerable postmortem
transport. All of them are Eocene in age

14 A Dyrosaurid Crocodile from
the Paleocene of Pakistan

(Buffetaut 1978b). Buffetaut (1978b) suggests
that they may represent immature individuals
that were hatched from inland broods and
had not yet journeyed seaward to live along
the coast as adults. Most of the vertebrae are
small (about 4 cm in length). The abraded
nature of the bones is consistent with riverine
or estuarine transport.

The present specimen represents by far
the most complete dyrosaurid yet discovered
in Asia. The preserved portion of the column
and the average vertebral length of
approximately 5 cm suggests an adult
individual of about 4 m length. The fossil is
one of the oldest Asian finds, excepting the
two above-mentioned vertebrae from the
Khadro Formation. Apparently the individual
(from its deep caudals, presumably an active
swimmer) was part of the littoral fauna.
Shallow-water marine invertebrates (e.g.,
Ostrea sp., corals, decapods, etc.) and the
ubiquitous teeth of the sand shark Odontaspis
sp. and the nurse shark Ginglymostoma sp.
confirm the near-shore marine nature of the
matrix. The glauconitic content of the siltstone
(approximately 80%) and its calcite cement
also support this interpretation. The articulated
condition of the column suggests little
transport of the carcass and only moderate
disruption after deposition. The animal, though
deposited in shallow water, was presumably
lying below wave base where some of its
exposed bones provided a temporary
substrate for small epifaunal oysters and
solitary scleractinians. There is also,
incidentally, no direct evidence that any of the
sharks were scavenging upon the carcass.
Their teeth are common throughout the
sediments of the locality.

The dyrosaurs of Pakistan, owing to their
fragmentary record, have not yet been
identified generically. At least some of them
may be congeneric with African forms where
the dyrosaurid record is most complete
(Maastrichtian to middle Eocene) and where
their greatest known diversity was achieved in

Postiila 197

the Paleocene (Buffetaut 1978b). Northern
Africa and Pakistan are geographically related
as components of the Tethyan maritime
province, and it is not unlikely that
crocodilians, especially such strong swimmers
as dyrosaurs, could have spread from one
area to the other. The present specimen,
while similar to Hyposaurus in most respects,
cannot be positively linked with any of the
African forms. Part of a slender, curved tooth
associated with, and probably belonging to it,
may rule out the large Phosphatosaurus
whose teeth were generally massive and
stout. However, the other and smaller Tertiary
dyrosaurs had long slender teeth. Only when
good skull material of Pakistani dyrosaurids is
found can more definitive identifications be
made.

Acknowledgments

This work was supported in part by a
Smithsonian Institution Foreign Currency
Exchange Grant to P. D. Gingerich, by the
Yale Peabody Museum of Natural History
Division of Vertebrate Paleontology, and by
the Geological Survey of Pakistan. | thank J.
D. Archibald, J. U. McClammer, Jr., R. M.
Schoch, and H. Shaheed for field work in
Pakistan and for subsequent discussions. S.
M. |. Shah kindly authorized collection and
study of the material. | am grateful to both J.
D. Archibald and P. D. Gingerich for allowing
me to study the specimen. K. A. Waldron
spent many hours assisting in the preparation
and study of the fossil and offered much
useful advice. | also thank J. U. McClammer,
Jr. for providing photographs of the field area
and W. K. Sacco for technical assistance. E.
Buffetaut and Wann Langston, Jr. provided
helpful information and also read and
criticized the manuscript, as did J. D.
Archibald, P. D. Gingerich, J. H. Ostrom, B. H.
Tiffney, K. M. Waage. | am grateful for their
comments.

15 A Dyrosaurid Crocodile from Postilla 197
the Paleocene of Pakistan

Literature Cited

Bergounioux, F. M. 1955. Les Crocodiliens fossiles des dépdts phosphatés du Sud-Tunisien. C. R. Acad.
Sci. Paris 240:1917-18.

1956. Les Reptiles fossiles des dépdts phosphatés sud tunisiens. Ann. Min. Geol. Tunis:1—105.
Blanford, W. T. 1876. On the geology of Sind. Rec. India Geol. Surv. 9:8-22.

Buffetaut, E. 1976a. Une nouvelle définition de la famille des Dyrosauridae de Stefano, 1903 (Crocodylia,
Mesosuchia) et ses consequences: inclusion des genres Hyposaurus et Sokotosuchus dans les
Dyrosauridae. Geobios 9, 3:333-36.

1976b. Sur la répartition geographique hors d'Afrique des Dyrosauridae. Crocodiliens
mésosuchiens du Crétacé terminal et du Paléogéne. C. R. Acad. Sci. Paris 283, D:487-90.

1977. Données nouvelles sur les Crocodiliens paleogénes du Pakistan et de Birmanie. C. R. Acad.
Sci. Paris 285, D:869-72.

1978a. A dyrosaurid (Crocodylia, Mesosuchia) from the Upper Eocene of Burma. Neues Jb. Geol.
Palaeontol. Monatsh. 5:273-81.

1978b. Crocodilian remains from the Eocene of Pakistan. Neues Jb. Geol. Palaeontol. Abh. 156, 2:

262-83.

1978c. Les Dyrosauridae (Crocodylia, Mesosuchia) des phosphates de |’Eocene inférieur de
Tunisie: Dyrosaurus, Rhabdognathus, Phosphatosaurus. Geol. Mediterr. 5, 2:237-56.

1979a. Sokotosuchus ianwilsoni and the evolution of the dyrosaurid crocodilians. Niger. Field,
Monogr. 1 (Suppl. to Vol. 44):31-41.

1979b. Présence du crocodilien Phosophatosaurus (Mesosuchia, Dyrosauridae) dans le Paleocene
du Niger et Mali. Palaeontol. Z. 53, 3/4:323-33.

1980. Les crocodiliens paléogeénes du Tilemsi (Mali): un apergu systematique. Paleovertebrata
(Montpellier), Mem. Jubil. R. Lavocat:15-35.

1981. Die biogeographische Geschichte der Krokodilier, mit Beschreibung einer neuen Art,
Araripesuchus wegeneri. Geol. Rundschau 70, 2:611-24.

Buffetaut, E., V. de Buffréenil, A. de Ricqles and Z. V. Spinar. 1982. Remarques anatomiques et
paléohistologiques sur Dyrosaurus phosphaticus, crocodilien mesosuchien des phosphates Yprésiens de
Tunisie. Ann. Paléont. 68, 4:327-41.

Buffetaut, E. and G. Wouters. 1979. Atlantosuchus coupateéZi, n.g., n.sp., un nouveau dyrosauride
(Crocodylia, Mesosuchia) des phosphates Montiens du Maroc. Bull. Trimest. Soc. Géol. Normandie et Amis
Muséum 66, 4:85-90.

Cope, E. D. 1886. A contribution to the vertebrate paleontology of Brazil. Proc. Am. Philos. Soc. 23:1-21.
De Stefano, G. 1903. Nuovi Rettili degli stratia fosfato della Tunisia. Boll. Soc. Geol. Ital. 22, 1:51-80.
Dollo, L. 1914. Sur la découverte de téléosauriens tertiaires au Congo. Acad. R. Belg., Bull. Cl. Sci.:
288-98.

Gingerich, P. D., D. E. Russell, D. Sigogneau-Russell, J.-L. Harbenberger, S. M. |. Shah, M.
Hassan, K. D. Rose and R. H. Ardrey. 1979. Reconnaissance survey and vertebrate paleontology of
some Paleocene and Eocene formations in Pakistan. Contrib. Mus. Paleont. Univ. Michigan 25, 5:105-16.
Halstead, L. B. 1975. Sokotosuchus ianwilsonin.g., n.sp., a new teleosaur crocodile from the Upper
Cretaceous of Nigeria. J. Niger. Min. Geol. Metal. Soc. 11, 1-2:101-3.

Hunting Survey Corporation. 1961. Reconnaissance geology of part of West Pakistan (Colombo Plan
Cooperative Project). Toronto: Canada Gov., p. 1-550.

Huxley, T. H. 1875. On Stagonolepis robertsoni, and on the evolution of the Crocodilia. Q. J. Geol. Soc.
London 31:423-38.

Khan, M.H. 1956. Ranikot Series. /n Lexique Stratigraphique International, ed. by L. Dubertret. Proc. 20th
Cong. Internat. Geol., Mexico, 3 Asia 8A:223-25.

Lydekker, R. 1879. Fossil Reptilia and Batrachia. Mem. Geol. Surv. India, Palaeontol. Indica, 4, |, 3:1-36.
Madden, C. T., |. M. Naqvi, F. C. Whitmore, Jr., D. L. Schmidt, W. Langston, Jr. and R. C. Wood.
1979. Paleocene vertebrates from coastal deposits in the Harrat Hadan area, At Taif region, Kingdom of
Saudi Arabia. US Geol. Surv., Saudi Arabia Mission, Proj. Rep. 269:1-29.

Moody, R. T. J. and E. Buffetaut. 1981. Notes on the systematics and palaeoecology of the crocodiles
and turtles of the Metlaoui Phosphates (Eocene) of southern Tunisia. Tertiary Res. 3, 3:125-40.

Nopcsa, F. 1905. Le Dyrosaurus des Phosphates de Gafsa. Bull. Soc. Géol. Fr. 4:138.

NN

SMITHSONIAN INST!

TT
06 5173

3 9088 012

16 A Dyrosaurid Crocodile from Postilla 197
the Paleocene of Pakistan

Owen, R. 1849. Notes on remains of fossil reptiles discovered by Prof. Henry Rogers of Pennsylvania,
U.S., in Greensand formations of New Jersey. Q. J. Geol. Soc. London 5:380-83.

Pomel, A. 1894. Sur le Dyrosaurus thevetensis. C. R. Acad. Sci. Paris 118:1396.

Sauvage, H.E. 1904. Critique of ‘Nuovi Rettili degli stratia fosfato della Tunisia, par Giuseppe de
Stefano.’’ Rev. Crit. Paléozool. 8:156-57.

Shah, S. M. 1. 1977. Stratigraphy of Pakistan. Mem. Geol. Surv. Pakistan 12:1-138.

Swinton, W. E. 1930. On fossil Reptilia from Sokoto Province. Bull. Geol. Surv. Niger. 13:9-56.
Thevenin, A. 1911. Le Dyrosaurus des phosphates de Tunisie. Ann. Paléont. 6:95-108.

Thomas, P. 1893. Description de quelques fossiles nouveaux ou critiques des terrains tertiaires et
secondaires de la Tunisie receuillis en 1885 et 1886 par M. Philippe Thomas. Explor. Sci. Tunis 3:1-46.
Troxell, E. L. 1925. Hyposaurus, a marine crocodilian. Am. J. Sci., 5th ser., 9, 54:489-514.
Vredenburg, E. W. 1909. Mollusca of the Ranikot Series, introductory note on the stratigraphy of the
Ranikot Series. Mem. Geol. Surv. India, Palaeont. Indica, n.s. 3, 1:5-19.

1928. A supplement to the Mollusca of the Ranikot Series. Mem. Geol. Surv. India, Palaeont.
Indica, n.s. 10, 4:1-75.

Williams, M. D. 1959. Stratigraphy of the Lower Indus Basin, West Pakistan. Proc. 5th Congr. World
Petroleum, New York, 1, 19:377-90.

The Author

Glenn W. Storrs. Department of Geology

and Geophysics and Peabody Museum of

Natural History Division of Vertebrate

Paleontology, Yale University, 170 Whitney

Avenue, P.O. Box 6666, New Haven, CT

06511. ISSN No. 0-912532-02