Content uploaded by Neville Stewart Pledge
Author content
All content in this area was uploaded by Neville Stewart Pledge on Sep 20, 2018
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=ujvp20
Download by: [203.1.252.5] Date: 10 April 2017, At: 17:55
Journal of Vertebrate Paleontology
ISSN: 0272-4634 (Print) 1937-2809 (Online) Journal homepage: http://www.tandfonline.com/loi/ujvp20
A Pliocene mekosuchine (Eusuchia: Crocodilia)
from the Lake Eyre Basin of South Australia
Adam M. Yates & Neville S. Pledge
To cite this article: Adam M. Yates & Neville S. Pledge (2017) A Pliocene mekosuchine (Eusuchia:
Crocodilia) from the Lake Eyre Basin of South Australia, Journal of Vertebrate Paleontology, 37:1,
e1244540, DOI: 10.1080/02724634.2017.1244540
To link to this article: http://dx.doi.org/10.1080/02724634.2017.1244540
View supplementary material
Published online: 18 Nov 2016.
Submit your article to this journal
Article views: 105
View related articles
View Crossmark data
ARTICLE
A PLIOCENE MEKOSUCHINE (EUSUCHIA: CROCODILIA) FROM THE LAKE EYRE
BASIN OF SOUTH AUSTRALIA
ADAM M. YATES*
,1
and NEVILLE S. PLEDGE
2
1
Museum of Central Australia, Museum and Art Gallery of the Northern Territory, 4 Memorial Avenue, Alice Springs, Northern
Territory 0870, Australia, adamm.yates@nt.gov.au;
2
South Australian Museum, North Terrace, Adelaide, South Australia 5000, Australia
ABSTRACT—A new genus and species of crocodilian, Kalthifrons aurivellensis, is described from a channel sand deposit incised
into the Oligo-Miocene Etadunna Formation on the western shore of Lake Palankarinna in the Lake Eyre Basin, South Australia.
The channel sand is interpreted as an outcrop of the Mampuwordu Sand Member of the Tirari Formation, which has been assigned
an early Pliocene age. The taxon can be diagnosed by the extremely elongate and narrow anterior process of the frontal pair and a
distinctively beveled and laterally expanded prefrontal contribution to the orbit margin. The broadly triangular rostrum, at least
partially interlocking dentition, smooth dental carinae, and weak labiolingual compression resemble unspecialized crocodilian
species, and it was probably a generalist aquatic predator. It possessed character states that indicate that it was not a member of
Crocodylus, and it is referred to the endemic Australasian clade Mekosuchinae. The first overlying unit in the Lake Eyre Basin
sequence to produce diagnostic crocodilian remains is the mid- to late Pliocene Pompapillina Member of the Tirari Formation. It
contains a true species of Crocodylus but no indication of Kalthifrons aurivellensis or indeed any other mekosuchine, indicating that
the latter may have been driven to extinction via competitive replacement.
http://zoobank.org/urn:lsid:zoobank.org:pub:C4E93358-D80F-4E4D-B7FF-BFFB93EDCD54
SUPPLEMENTAL DATA—Supplemental materials are available for this article for free at www.tandfonline.com/UJVP
Citation for this article: Yates, A. M., and N. S. Pledge. 2016. A Pliocene mekosuchine (Eusuchia: Crocodilia) from the Lake
Eyre Basin of South Australia. Journal of Vertebrate Paleontology. DOI: 10.1080/02724634.2017.1244540.
INTRODUCTION
The origin of ‘true’ crocodiles of the genus Crocodylus is sur-
prisingly recent given the popular view that they are ‘living fos-
sils.’ They are even more recent immigrants to the continent of
Australia. Molecular clock estimates place the origin of the
crown clade in the late Miocene (Oaks, 2011), whereas fossil evi-
dence suggests that the genus first entered Australia during the
Pliocene, after dispersing across marine barriers from Southeast
Asia via the Malay Archipelago (Molnar, 1979; Willis, 1997a).
Before this time, Australia had been host to an endemic clade of
crocodilians known as the Mekosuchinae (Willis, 1997a; Salisbury
and Willis, 1996). The origin of Mekosuchinae is obscure, but phy-
logenetic analyses consistently find them to be more closely related
to Crocodylus than to Alligator (Salisbury and Willis, 1996; Brochu,
1999; Brochu and Storrs, 2012). Mekosuchines underwent a drop in
diversity in the latter part of the Neogene, with the loss of the
large-prey specialist Baru, the aquatic generalist Australosuchus,
and a range of dwarf taxa such as Trilophosuchus and Mekosuchus
from continental Australia (Willis, 1997a). How many of these
extinctions, if any, can berelated to competitive exclusion by invad-
ing Crocodylus is unknown. It is known that crocodilian diversity
worldwide underwent a reduction during the late Miocene to early
Pliocene, as a response to cooling global climate and increased ari-
dification (Markwick, 1998; Mannion et al., 2015). It seems quite
likely that at least some of these mekosuchine taxa (e.g., Australo-
suchus, Baru, Trilophosuchus,andMekosuchus) were extinct
before the close of the Miocene (Willis and Molnar, 1991; Willis,
1993, 1997a, 2001), at least on the continent of Australia (Mekosu-
chus did survive on offshore islands, Balouet and Buffetaut, 1987;
Balouet, 1991; Mead et al., 2002). If this wereso, there was noover-
lap between these genera and Crocodylus in Australia, and the lat-
ter may have been simply moving into empty niches when it
colonized the continent. A similar pattern is observed in South
America where coincident climatic and hydrographic changes in
the late Miocene led to local extinctions that paved the way for col-
onization of immigrant Crocodylus (Moreno-Bernal et al., 2016)
Apparently the very broad-snouted mekosuchine, Pallimn-
archus, from the Plio-Pleistocene was able to coexist with Croco-
dylus in Australia, although rarely at the same site (Molnar,
1982; Willis and Molnar, 1997). Through the long temporal over-
lap between Crocodylus and Pallimnarchus, there are only four
recorded sites that contain both taxa and only two that contain
Pallimnarchus together with a broad-snouted, generalist species
of Crocodylus (Willis and Molnar, 1997), suggesting that some
ecological segregation kept the two taxa from occupying the
same habitat. The long-lived mekosuchine taxon Quinkana was
able to survive the advent of Crocodylus in Australia up to the
late Pleistocene, when it too went extinct (Molnar, 1981). In this
case, competition with Crocodylus was probably avoided
through the planocraniid-like specializations of Quinkana for
terrestrial predation (Molnar, 1981; Megirian, 1994).
Here we report on a probable mekosuchine from the Pliocene that
was living in the Lake Eyre Basin, immediately before the arrival of
*Corresponding author.
Color versions of one or more of the figures in this article can be found
online at www.tandfonline.com/ujvp.
Journal of Vertebrate Paleontology e1244540 (15 pages)
Óby the Society of Vertebrate Paleontology
DOI: 10.1080/02724634.2017.1244540
true Crocodylus. Its skull form appears to be that of a generalist,
broad-snouted crocodile, and it is a good candidate for a taxon that
went extinct due to competition with the newly arrived Crocodylus.
Institutional Abbreviations—NMV, Museum Victoria,
Melbourne, Australia; NTM, Museum and Art Gallery of the
Northern Territory, Darwin, Australia; QM, Queensland
Museum, Brisbane, Australia; SAM, South Australian Museum,
Adelaide, Australia; UCMP, University of California Museum
of Paleontology, Berkeley, California, U.S.A.
GEOLOGICAL SETTING
The specimen derives from the Golden Fleece locality on the
western shore of Lake Palankarinna, South Australia, in the
Lake Eyre Basin (Fig. 1). This vertebrate fossil assemblage
occurs in pale gray fluviatile channel sand that is incised into the
gray-green Oligo-Miocene claystones of the Etadunna Forma-
tion. The specimen consists of multiple associated parts (a
cranium and two mandibular rami) and bears no trace of adher-
ent claystone, ruling out the possibility that it was reworked
from the Etadunna Formation. Underlying the sand, and sepa-
rating it from the Etadunna Formation, is a thin sheet of selenite.
This is interpreted as an evaporite deposit that is penecontem-
poraneous with the channel fill and may record a total drying out
of a waterhole on this part of the channel. The channel sand is
capped by gypcrete. The vertebrate assemblage consists largely
of disassociated turtle shell elements (tentatively referred to
Elseya sp. cf. E. lavarackorum), with some crocodilian
osteoderms, shed teeth, and vertebral fragments. The skull and
jaws described in this paper are unique for this site, and no others
have been found. The vertebrate assemblage is interpreted as a
mass death assemblage that formed when a waterhole dried out,
killing aquatic taxa such as turtles and crocodilians.
The age of the locality cannot be determined through bio-
chronology because no chronologically useful vertebrate fossils
have been found at the site. The crocodilian described herein is
FIGURE 1. Locality maps showing the loca-
tion of the Tirari Desert, Lake Palankarinna,
and the Golden Fleece locality. A, South Aus-
tralia; B, Tirari Desert; C, northwestern shore of
Lake Palankarinna showing various fossil locali-
ties, redrawn from Woodburne et al. (1994:
fig. 6). Contour interval in Cis 4 m; contours
are relative and taken from the floor of the lake,
sea level is close to 16 m.
Yates and Pledge—A mekosuchine crocodilian from Australia (e1244540-2)
quite distinct from Australosuchus clarkae, the only crocodilian
known from the Etadunna Formation, suggesting that the age
of the Golden Fleece locality might differ from that of the Eta-
dunna Formation. The channel sand of the Golden Fleece local-
ity could potentially represent a channel facies of the Etadunna
Formation, but no similar channels of Etadunna age are known
at Lake Palankarinna. The Golden Fleece locality occurs near
outcrops of the Mampuwordu Sand Member of the Tirari For-
mation (Fig. 2). The Mampuwordu Sand Member (Callen and
Tedford, 1976) fills a channel that is incised into the Etadunna
Formation (Stirton et al., 1961; Tedford et al., 1992). Some
stratigraphers have classified the Mampuwordu Sand itself as a
channel facies within the Oligo-Miocene Etadunna Formation
(Alley, 1998). However, the unit contains derived macropodid
and diprotodontid fossils that clearly indicate a Pliocene age
(Tedford et al., 1992; Megirian et al., 2010). Furthermore, the
base of the Mampuwordu Sand Member is disconformable with
the underlying Etadunna Formation and conformable with the
overlying main body of the Tirari Formation (Tedford et al.,
1992). Utilizing these biostratigraphic and lithostratigraphic
data and combining them with paleomagnetic data, an age of
3.9 Ma was estimated for the unit (Tedford et al., 1992). We
regard the Golden Fleece locality as a probable outcrop of the
Mampuwordu Sand Member of the Tirari Formation with an
early Pliocene age.
SYSTEMATIC PALEONTOLOGY
CROCODILIA Gmelin, 1789
?LONGIROSTRES Harshman et al., 2003
MEKOSUCHINAE (Balouet and Buffetaut, 1987)
KALTHIFRONS AURIVELLENSIS, gen. et sp. nov.
(Figs. 3–11)
Etymology—Genus name from the Dieri word ‘kalthi,’ mean-
ing spear and the Latin word ‘frons,’ meaning forehead or brow.
A reference to the long, narrow, pointed anterior process of the
frontal bone. Gender is masculine. Species name from the latin
word ‘aurum,’ meaning gold, ‘vellus’ meaning fleece, and the
adjectival suffix ‘-ensis,’ meaning to originate from. Refers to the
type locality.
Diagnosis—A crocodilian with the following autapomorphies:
narrow, pointed anterior process of frontal extending between
prefrontals and nasals for 64% of frontal length; and prefrontal
contribution to orbital margin with lateral flange bearing a dorso-
laterally facing fossa. Differing from all other mekosuchines
(except Australosuchus clarkae) by: small, anteroposteriorly
narrow supraoccipital contribution to dorsal skull roof. Differs
from A. clarkae in: relatively narrower premaxillae; occlusion of
fourth dentary tooth in widely open notch; more anteriorly posi-
tioned orbits (level with 11th maxillary alveolus vs. behind 14th);
relatively greater width of frontal interorbital region (43% of
orbital length vs. 15%); and smaller, elongated oval supratempo-
ral fenestra with correspondingly wider interfenestral bar. Differs
from Crocodylus in: quadratojugal forms posteroventral corner
of infratemporal fenestra (quadratojugal process extending onto
lower temporal bar); more anteriorly positioned orbits; pterygoid
with pair of ridges extending anteriorly from lateral margins of
choana; and labiolingually compressed tooth crowns.
Holotype—SAM P35062, a poorly preserved skull with incom-
plete associated lower jaws (Figs. 3–5, 8).
Locality and Stratigraphy—Golden Fleece locality, Lake Pal-
ankarinna, South Australia. Probably early Pliocene Mampu-
wordu Sand Member of the Tirari Formation.
Tentatively Referred Material—Various isolated teeth, osteo-
derms, and vertebral fragments, registered as SAM P35062, from
the type locality (Figs. 9, 10).
FIGURE 2. Composite stratigraphic col-
umn of the Etadunna and Tirari formations
in the northern third of the western side of
Lake Palankarinna, showing the occurrence
of fossil crocodilians. Black skull silhouettes
represent occurrences from the northwestern
shore of Lake Palankarinna, gray silhouettes
represent occurrences from the same lithos-
tratigraphic units further afield. Stratigraphic
thickness (left) is in meters. Based on strati-
graphic profiles from Woodburne et al.
(1994:fig. 11) and Tedford et al. (1992:fig. 3).
Yates and Pledge—A mekosuchine crocodilian from Australia (e1244540-3)
DESCRIPTION
Preservation of the Holotype
The skull was found lying upside down, and the poor pres-
ervation of the ventral features of the skull is probably due
to their deterioration prior to burial. The braincase is largely
absent, and all that remains is a formless mass of small bone
fragments mixed with the enclosing, sandy matrix (Fig. 4B).
Most of the teeth have also disintegrated, with some frag-
mentary shards remaining in some of the alveoli and one
unerupted crown remaining intact in a posterior alveolus of
the right maxilla. The individual alveoli are hard to discern
due to disintegration of the bone and infilling of the alveoli
with bone fragments and matrix. The palate is extensively
damaged. The anterior ends of the palatines are missing, as
are the anterior margins of the suborbital fenestrae. Conse-
quently, it is not possible to ascertain the size, shape, and
position of the suborbital fenestrae. A piece consisting of
the pterygoid plate from in front of the choana and the pos-
terior ends of the palatines has become displaced and lies
toward the left side of the skull (thus appearing on the right
in ventral view). The left ectopterygoid is missing, whereas
the right is slightly displaced and missing the posterior
pterygoid process (Fig. 4B). Only the bones of the dorsal
skull are moderately well preserved, but even here the ante-
rior ends of the premaxillae, the left posterior temporal
region, and the condyle of the right quadrate are missing
due to erosion.
General Skull Form
The skull is small for a crocodilian, but larger than dwarf
mekosuchin taxa such as Mekosuchus spp., Volia athollander-
soni, and Trilophosuchus rackhami (Fig. 5). It is short, broad,
and triangular in dorsal view (Fig. 3). Its proportions are typical
of a generalist, broad-snouted crocodilian (Brochu, 2001). The
skull is dorsoventrally flattened, with a platyrostral snout
(Fig. 4A). This has probably been accentuated by post-burial
compaction, and it is difficult to assess how tall the snout was in
life. Crocodilians with flattened to moderately deep snouts fre-
quently develop dorsal bosses over the roots of the fifth maxil-
lary teeth, presumably because the maxilla roofs over the root of
this large tooth, which causes an upward bulge such as in Brochu-
chus pigotti (Conrad et al., 2013:fig. 6a, b) and Crocodylus thorb-
jarnarsoni (Brochu and Storrs, 2012:fig. 1a, f). In contrast, the
maxilla of highly altirostral taxa, such as Quinkana, is more
FIGURE 3. Holotype skull of Kalthifrons aurivellensis, gen. et sp. nov. (SAM P35062), in dorsal view. A, photograph; B, photograph with interpreta-
tion of sutural relationships and morphology. Note that the green areas are due to mould release agent having soaked into the porous matrix. Abbrevi-
ations:f, frontal; itf, infratemporal fenestra; j, jugal; l, lacrimal; mx, maxilla; n, nasal; na, naris; o, orbit; p, parietal; pmx, premaxilla; po, postorbital; prf,
prefrontal; q, quadrate; qj, quadratojugal; so, supraoccipital; sq, squamosal; stf, supratemporal fenestra. Scale bar equals 200 mm.
Yates and Pledge—A mekosuchine crocodilian from Australia (e1244540-4)
steeply inclined in the region of the fifth maxillary tooth and con-
sequently does not develop a strong dorsal boss (pers. observ. of
Quinkana spp.; NTM P9464-167, QM F31057, cast of AM
F57844). The snout of K. aurivellensis has well-developed bosses
on the maxilla, suggesting that the snout was at best only moder-
ately deep, perhaps reaching a similar proportion to that of Baru
darrowi (Willis et al., 1990:fig. 6a). The snout (measured from
the rostral tip to the level of the anterior ends of the orbits) occu-
pies 56% of the total skull length as preserved and is 1.46 times
longer than it is broad at its base. There is a deep constriction
between the premaxilla and maxilla for receiving the fourth den-
tary tooth. The notch is broadly open in dorsal and lateral views
as in most crocodilians, unlike the semi-enclosed condition of
Australosuchus clarkae (Willis and Molnar, 1991). The snout
bears a mild lateral expansion at the level of the enlarged fifth
maxillary tooth. The orbits are large, with a maximum dorsoven-
tral diameter that exceeds the interorbital distance (Fig. 5). The
anterior ends are pointed and draw level with what is interpreted
as the 11th maxillary tooth position, indicating that the orbits
were anteriorly positioned, similar to those of Baru darrowi (Wil-
lis et al., 1990). The skull table has a planar dorsal surface and is
trapezoidal in dorsal view. The transverse width at its posterior
end is 1.61 times greater than the anteroposterior length of the
table. The lateral margins diverge posteriorly from the sagittal
plane at an angle of 12(measured from the undeformed right
side). The small, widely spaced supratemporal fenestrae are ellip-
tical, with their long axes oriented anteroposteriorly as in Trilo-
phosuchus rackhami (Willis, 1993:fig. 1b). Despite their relatively
small size, their opening is largely unrestricted by overhanging
bones of the skull roof, although the posterior part of the parietal
rim is slightly overhanging. At its minimum width, the interfenes-
tral bar is as wide as the supratemporal fenestrae are long. The
posterior margin of the posterior skull table is gently and evenly
concave along its width. The right infratemporal fenestra appears
as a relatively large subrectangular opening, but its size and
shape have probably been altered by the loss of bone from its
posterior and dorsal margins.
Cranial Bones
Premaxilla—The premaxillae are relatively narrower than
in most other mekosuchines, including Australosuchus clar-
kae. The maximum width of the pair is half the width of the
base of the snout. The external naris is missing its anterior
margin, but even the preserved portion is longer than it is
wide, indicating an elongate narial opening in contrast to the
broad openings seen in other mekosuchines. Short, narrow
triangular posterior processes of the premaxilla insert
between the maxilla and the nasal. Both left and right pro-
cesses are incomplete, but the articulation facet for the left
process on the lateral side of the left nasal indicates that it
extended no further posteriorly than the level of the third or
fourth maxillary tooth. Ventrally, there is a pit developed at
each of the anteromedial ends of the preserved portion of
the premaxillae for the reception of the first pair of dentary
teeth. These do not appear to have merged with the incisive
foramen as they do in Baru darrowi (Willis et al., 1990:
fig. 1c). The left premaxilla shows the remains of three poste-
rior premaxillary teeth, but it is not known whether the miss-
ing anterior portion held one or two teeth. The posterior-
most alveolus lies directly posterior to the penultimate alveo-
lus. Damage around the edges of the incisive foramen makes
it impossible to discern its size and shape accurately, but
enough premaxillary palate is preserved to show that it was
small and anteriorly placed.
Maxilla—The maxilla bears a low swelling on its dorsal surface
above the root of the enlarged fifth maxillary tooth. Such swel-
lings are typical of many crocodylians, including mekosuchines
and crocodylids. At the posterior end of the maxilla, the dorso-
medial surface that is adjacent to the lacrimal is depressed, creat-
ing a broad, shallow fossa that faces dorsolaterally. On the
palatal surface, poor preservation makes it impossible to accu-
rately count alveoli, but there appears to be 10 preserved tooth
positions with space for more on the left maxilla and six tooth
positions preserved posterior to the enlarged fifth tooth on the
FIGURE 4. Kalthifrons aurivellensis, gen. et sp.
nov., SAM P35062, holotypic skull. A, lateral and
B, ventral views. Scale bar equals 200 mm.
Yates and Pledge—A mekosuchine crocodilian from Australia (e1244540-5)
right maxilla, for a total of at least 11 maxillary teeth on each
side. There were probably more posterior teeth that have been
lost to erosion. The anterior maxillary teeth appear to have been
close set, and there may not have been enough room for recep-
tion pits between them, although the lack of preserved interal-
veolar bone makes it impossible to be sure. A large reception pit
is developed between the sixth and seventh alveoli, indicating
that at least a partially interlocking dentition was present. A low
alveolar wall forms the medial margin of at least the third,
fourth, and fifth alveoli. As in most, but not all, crocodilians, the
maxillae do not participate in the margin of the orbit, neither are
there any posterior processes of the maxilla inserting into the lac-
rimal or between the lacrimal and the nasal.
Nasal—The dorsal surface of the nasal pair is flat and extends
anteriorly to the external naris, preventing a midline contact of
the premaxillae behind the naris. The pair is roughly spindle-
shaped, narrowing at both the anterior and posterior ends. The
posterior ends are separated for a distance of 25 mm by the nar-
row anterior process of the frontal pair that inserts between
them. The pointed posterior ends of the nasals insert between
the anterior process of the frontals and the prefrontals.
Lacrimal—The lacrimals are elongate, irregular triangular bones
that form the anterior ventrolateral margin of the orbit and its
pointed anterior corner. They extend further anteriorly than the pre-
frontals so that the medial margin of each passes along the lateral
margins of the prefrontal and the nasal. Each lacrimal bears a thick
preorbital ridge that extends from the anterior corner of the orbit to
the anterior tip of the lacrimal. This preorbital ridge differs from that
of Crocodylus porosus in that it is not strongly raised dorsally above
the level of the snout but rather results from a distinct change in
slope between the horizontal median roof of the snout and the dorso-
laterally facing lateral sides of the snout. Slight swelling of the bone
along this angle in slope accentuates its thick, ridge-like appearance.
The same form of preorbital ridge is present in Baru wickeni (Willis,
1997b:fig. 12).
Prefrontal—The prefrontals (Fig. 6) form the anterodorsal mar-
gin of the orbits. In dorsal view, the prefrontal bears a low convex lat-
eral projection along this margin that produces a distinct
emargination in the rim of the orbit. This is best seen on the right
side because the apex of the projection on the left side is damaged.
Instead of forming a sharp flat or raised edge, as in most crocodyli-
ans, the orbital margin of the prefrontal forms a dorsolaterally bev-
eled margin. This bevel bears a shallow depression at the level of the
lateral projection. A second depression is developed at the posterior
end of the dorsal surface of the prefrontal, between the beveled lat-
eral margin and the anterior process of the conjoined frontals. A
low, thin dorsal crest separates the posterior depression from the
beveled orbital margin. The prefrontal-frontal suture follows an ‘L’-
shaped path, with the posterior end of the prefrontal terminating
bluntly to form the horizontal branch of the ‘L.’
Frontal—This bone is divisible into two distinct regions: a broad
posterior, interorbital plate and a long, narrow, spike-like anterior
process. The anterior process has a pointed tip that inserts between
the posterior ends of the nasal pair. Most of its dorsal surface is
smooth, but the posterior-most centimeter of its length bears rugose
ornament that is present on the rest of the frontal. The length of the
anterior process is exceptional, occupying just over 60% of the total
length of the frontal (Fig. 7A). The blunt posterior ends of the inset
prefrontals create an abrupt constriction of the frontal that marks
the boundary between the posterior plate and the anterior process,
which is level with the minimum interorbital distance. The interor-
bital distance is 80% of the maximum dorsoventral diameter of the
orbit. The dorsal surface of the trapezoidal posterior plate is
FIGURE 6. Kalthifrons aurivellensis, gen. et sp. nov., SAM P35062, holo-
typic skull, close-up of right prefrontal and adjacent bones in dorsal view
showing autapomorphic features of the prefrontal orbital margin. Abbrevia-
tions:bev, bevel; foss,fossa;lat exp, lateral expansion. Scale bar equals
20 mm.
FIGURE 5. Kalthifrons aurivellensis, gen. et sp. nov., SAM P35062,
measurements of holotypic skull. Units are in mm.
Yates and Pledge—A mekosuchine crocodilian from Australia (e1244540-6)
transversely concave, with gently raised orbital margins. A low mid-
line keel is present at the anterior end and terminates about halfway
along the length of the posterior plate. The frontoparietal suture is
gently anteriorly concave and lies on the dorsal skull roof anterior to
the supratemporal fossae.
Jugal—The right jugal is the better preserved of the pair and
forms the basis of this description. It is a slender bone that is similar
in proportions to Australosuchus clarkae (Willis and Molnar, 1991:
fig. 6b, d). It reaches a maximum depth immediately anterior to the
base of the postorbital bar. The degree of anterior expansion is mod-
est and reaches a maximum depth of 1.6 times greater than the mini-
mum depth of the lower temporal bar, which is in contrast to the
broadly flared jugals of altirostral taxa such as Baru sp., where the
anterior depth is 2.25 times greater than the minimum depth of the
lower temporal bar (e.g., NTM P5263). Like Baru sp., the suborbital
region of jugal bears an irregular depression on its outer, or dorsolat-
eral, surface. The base of the postorbital bar is inset, and the raised
dorsal rim of the jugal body creates a shallow trough between the
postorbital bar and the dorsolateral surface. Like in Australosuchus
clarkae, the ventral margin is only very weakly concave, which is in
contrast to the more strongly arched ventral margins seen in Baru
spp. and mature Crocodylus porosus. The dorsal margin is convex,
with the highest point located at the level of maximum depth, imme-
diately anterior to the postorbital bar.
Postorbital—The postorbitals form the anterolateral corners of
the skull table. The dorsal surface is flat and heavily ornamented
with deep pits. The slender postorbital bar descends from the antero-
lateral corner. The dorsal end of the bar is not inset, with its antero-
lateral surface flush with that of the skull table. The medial side of
the postorbital forms the anterolateral margin of the supratemporal
fenestra. The postorbital wall of the fenestra is oriented vertically,
with no supratemporal fossa extending onto the dorsal surface of the
postorbital. Anteromedially, the postorbital sutures with the frontal
and the parietal. The postorbital-parietal suture extends anteriorly
for a distance of about 5 mm from the anteromedial margin of the
supratemporal fenestra onto the dorsal surface of the skull table, sep-
arating the fenestra from the posterior margin of the frontal. The
postorbital-squamosal suture extends laterally from the lateral mar-
gin of the supratemporal fenestra to the lateral margin of the skull
table. Poor preservation and adherent matrix prevent us from tracing
the course of this suture across the lateral and ventral surfaces of the
skull table.
Parietal—The width of the interfenestral bar between the supra-
temporal fenestrae is 51% of the total length of the parietal, an
unusually broad proportion in comparison with most other mekosu-
chines, except an unnamed species of Baru from Alcoota (pers.
observ. of NTM P5335). The parietal wall of the supratemporal
fenestra plunges vertically at its anterior end, whereas the ventral rim
flares laterally to meet the quadrate at the posterior end. The lateral
flare of the posterior ventral rim produces a supratemporal fossa that
is visible inside the dorsal rim of the fenestra when viewed dorsally.
Preservation is too poor to determine if the parietal also makes con-
tact with the squamosal under the anterior opening of the cranioqua-
drate canal. The posterior end of the lateral rim of the interfenestral
bar forms a weakly raised lip that slightly overhangs the supratempo-
ral fenestra. The dorsal surface of the interfenestral bar between
these two slightly raised rims is gently depressed. Behind the supra-
temporal fenestrae, the parietals form a flat region that extends
FIGURE 7. Comparison of Kalthifrons aurivellensis, gen. et sp. nov., with Quinkana spp. A, frontal in dorsal view of K. aurivellensis (SAM P35062);
B, frontal in dorsal view of Q. sp. cf. fortirostrum (NTM P6282); C, posterior skull deck in right dorsolateral view of K. aurivellensis (SAM P35062);
D, posterior skull deck in left dorsolateral view (reversed for comparison) of Q. timara (cast of NMV P179632). Abbreviations:ap, anterior process of
the frontal; ls, lateral sulcus of the squamosal; nas, articular surface for contact with the nasals; om, orbital margin; plp, posterolateral process of the
squamosal; stf, supratemporal fenestra. Scale bars equal 20 mm.
Yates and Pledge—A mekosuchine crocodilian from Australia (e1244540-7)
posteriorly to the rear margin of the skull table. The posterior margin
is gently concave, except at the midlinewherethereisasmall,
roughly semilunate notch that accepts the dorsal exposure of the
supraoccipital. The dorsal surface of the parietal immediately ante-
rior to this notch is raised into a small, low, irregular bump. Laterally
the parietal-squamosal suture extends from the posterior rim of the
supratemporsal fenestra to the posterior margin of the skull table.
Squamosal—The squamosal forms the posterolateral corner of
the table and the posterolateral margin of the supratemporal fenes-
tra. The dorsolateral margin is rounded and is not raised into a ridge.
In dorsal view, the posterior lateral margin meets the occipital mar-
ginatanangleof70
ontheleftsideand80
on the right, with the
asymmetry caused by post-burial distortion. On the right squamosal,
the posterolateral corner continues into the base of the subdermal
posterolateral prong of the squamosal before being terminated by
breakage. In lateral view, this prong descends posteroventrally at a
shallow angle from the skull table, unlike the steeply descending
margins seen in some derived mekosuchines (e.g., Volia athollander-
soni; Molnar et al., 2002:fig. 2o) but not others (e.g., Quinkana
timara; pers. observ. of NMV P179632). Anterior to this prong, the
lateral surface of the squamosals is mostly obscured by adherent
matrix and coats of glue, but enough is exposed to make out that
there are parallel margins of a smooth, narrow, lateral sulcus for ear
flap valve musculature (Fig. 7C).
Anteriorly, the squamosal meets the posterior end of the postor-
bital to form a broad, dorsally flat bar, lateral to the supratemporal
fenestra. The postorbital-squamosal suture crosses this bar trans-
versely. As it crosses over the dorsolateral margin of the table and
onto the lateral surface, it extends anteriorly, defining the dorsal bor-
der of an anterior prong of the squamosal that extends along the ven-
trolateral margin of the postorbital, a condition typical of most
crocodilians, including several mekosuchines (e.g., Baru wickeni;
pers. observ. of NTM P91164). It is not possible to trace the suture
further, so it cannot be determined if the anterior end of the squamo-
sal underplates the postorbital.
Quadratojugal—The left quadratojugal is missing, and not much
of the right one can be seen. Its contact with the squamosal is
obscured under supporting matrix and is likely crushed and broken,
given the state of the bones under the skull table. The posterior
region is missing. Only the ventral surface adjacent to the posterior
angle of the lower temporal fenestra is tolerably well preserved
(Fig. 8). This shows that the quadratojugal formed the margin of the
posterior angle with a pointed anterior process extending along the
dorsomedial margin of the lower temporal bar.
Supraoccipital—Only the dorsally exposed portion of the supra-
occipital can be seen; the occipital surface is crushed and obscured
by matrix. The dorsal contribution to the skull table was restricted
to a small, roughly semilunate sliver that is just over 19 mm wide
FIGURE 8. Kalthifrons aurivellensis, gen. et
sp. nov., SAM P35062, holotypic skull, ven-
tral and slightly medial oblique views of the
right lower temporal fenestra. A, photograph;
B, photograph with interpretation of sutural
relationships and morphology. Stippled areas
represent matrix, loose shards of bone, and
bones unrelated to the area of interest.
Hatched area represents exposed articular
surface. Abbreviations:itf, infratemporal
fenestra; j, jugal; qj, quadratojugal; qjas,
articular surface for reception of the quadra-
tojugal; qj-js, quadratojugal-jugal suture.
Scale bar equals 50 mm.
Yates and Pledge—A mekosuchine crocodilian from Australia (e1244540-8)
and 7.5 mm long. This crescent fits into a corresponding notch in
the posterior margin of the parietals. In contrast, the supraoccipital
of most other mekosuchines, including Kambara taraina
(Buchanan, 2009:fig. 3a), Trilophosuchus rackhami (Willis, 1993:
fig. 2b), Baru wickeni (pers. observ. of NTM P902-5), Volia athol-
landersoni (Molnar et al., 2002:fig. 2g), and Mekosuchus sanderi
(Willis, 2001:fig. 2), forms a triangular wedge that inserts deeply
into the dorsal skull roof.
Palatine—Only the posterior ends of the palatines are pre-
served (Fig. 9). Little can be said other than they lack the poste-
rior transverse flaring seen in some Crocodylus species (e.g.,
Crocodylus porosus; Grigg and Gans, 1993:fig. 40.3b).
Ectopterygoid—The preserved portions of the right ectoptery-
goid seem unremarkable (Fig. 9). The maxillary process is elongate
and tapering with a subtriangular cross-section. The anterior tip is
too poorly preserved to determine if it was forked, or if the tip
inserted into the maxilla away from the margin of the suborbital
fenestra. The anteromedial margin from the maxillary process to
the pterygoid suture forms an evenly concave curve without any
bulge that would have constricted the suborbital fenestra.
Pterygoid—The anterior region of the pterygoid plate is pres-
ent, but it is in a poor state of preservation (Fig. 9). Nevertheless,
it shows a number of distinctive characters, several of which are
significant for determining the systematic position of Kalthifrons
aurivellensis. A short section of pterygoid-palatine suture on the
right side indicates that the palatines extend posterior to the level
of the posterior margin of the suborbital fenestra so that there
was no contribution from the pterygoids to the interfenestral
narial canal. The pterygoid component of the posterior margin of
the suborbital fenestra was at least as broad as the posterior width
of the palatines, a derived condition shared with some derived
mekosuchines (e.g., Baru darrowi, NTM P87110-31). Most nota-
bly, the pterygoid plate is arched immediately posterior to the
pterygoid-palatine suture so that the plate faces slightly postero-
ventrally if the palatines are oriented horizontally. This arch is
correlated with a deep fossa on the pterygoid plate, anterior to
the choana in more complete mekosuchines (e.g., Kambara
implexidens, QM F29662). Each pterygoid bears a small sharp-
crested prominence either side of the midline, along the broken
posterior margin. These prominences represent the anterior ends
of a pair of ridges that extend anteriorly from the lateral margins
of the choana in mekosuchines (e.g., Kambara taraina, Buchanan,
2009:fig. 6; Baru darrowi, pers. observ. of NTM P87110-31).
Mandible—Mandibular rami were found associated with the
holotype skull (Fig. 10). Unfortunately, they are in such a poor state
of preservation that very few details worth describing can be
observed. The ventral part of the lateral surface of the angular was
heavily ornamented with deep pits, whereas there is no trace of orna-
mentation preserved on the lateral surface of the dentary. This does
not mean that the dentaries were smooth in life, because the outer
layers of bone of this specimen have been shed, but it does indicate
that the posterior end of the dentary was far more lightly orna-
mented than the angular. The ventral splenial-dentary suture can be
traced along the ventromedial margin. Like most crocodilians,
except some mekosuchines (e.g., Baru darrowi, NTM P87115-15),
the splenial contributes only a thin edge to the ventral surface of the
mandible and the ornament is apparently restricted to the dentary,
although poor surface preservation may have erased some weak
splenial ornament.
Perhaps the most remarkable feature of the mandible is the
broad, flat ventral surface of the dentaries that is offset from the lat-
eral surface by a right-angled bend along the ventrolateral margin
of the dentary (Fig. 10D). In this respect, the mandible resembles
those referred to Quinkana (e.g., NTM P5334 and QM F7816).
Dentition—Only three tooth crowns of the holotype are par-
tially complete, with one in the right maxilla and two in the right
dentary. All occur towards the posterior ends of their respective
tooth rows and have the short, broad, and bluntly rounded shape
typical of the posterior teeth of many crocodilians (Fig. 11D, E).
The crowns bear smooth mesial and distal carinae and are labio-
lingually compressed, with labiolingual widths that range from
63% to 74% of the mesiodistal length.
Better preserved isolated crocodilian tooth crowns have been
found at the same locality (Fig. 11A–C). Given that Golden Fleece
appears to be a small, low-diversity assemblage formed when a local
pool dried up, it seems reasonable to assume that these teeth belong
to Kalthifrons aurivellensis, although the referral must remain tenta-
tive. These teeth include anterior teeth that are taller, more acutely
pointed, and slightly less labiolingually compressed than the poste-
rior teeth. The largest complete specimen is 21.8 mm high and has a
labiolingual width that is 76% of the mesiodistal length (SAM,
unregistered material). The crown has a straight longitudinal axis in
labial and lingual views, but is gently curved to the lingual side in
mesial and distal views.
Vertebrae—There is a fragmentary procoelous vertebral cen-
trum from the Golden Fleece locality (SAM P, unregistered
FIGURE 9. Kalthifrons aurivellensis, gen. et sp. nov., SAM P35062, holotypic skull, ventral view of the posterior palatal region. A, photograph;
B, photograph with interpretation of the preserved palatal elements. Abbreviations:ec, ectopterygoid; pal, palatine; pt, pterygoid; vptr, ventral ptery-
goid ridge. Scale bar equals 100 mm.
Yates and Pledge—A mekosuchine crocodilian from Australia (e1244540-9)
material). It does not display any characters of diagnostic value
within Crocodylia.
Osteoderms—Several crocodilian osteoderms were recovered
from the Golden Fleece locality that, like the isolated tooth crowns,
can be tentatively referred to Kalthifrons aurivellensis (Fig. 12).
They range from square to rectangular, with low to tall midline keels.
The dorsal surface is ornamented with a series of round pits, except
for a narrow articulation facet that extends along the anterior mar-
gin. The anterior edge of this facet is straight, without an anterior
projection seen on some basal crocodilian osteoderms. The largest
preserved osteoderm reaches a mediolateral width of 62.3 mm.
CLADISTIC ANALYSIS
Kalthifrons aurivellensis was scored for 178 character traits
(Supplementary Data 1) and added to a matrix of 37 other taxa
(Supplementary Data 2). The matrix is a reduced version of an
in-progress work designed to examine the inter- and intrarela-
tionships of Mekosuchinae; consequently, the taxon sampling is
skewed towards Mekosuchinae and their likely longirostran
outgroups.
Mekosuchinae is represented by a high proportion of named
species-level taxa. Those that were omitted (Kambara molnari,
Quinkana babarra, and Baru huberi) could not be scored for
many of the characters and behave like wildcard taxa. The scores
for Quinkana fortirostrum were supplemented by material from
the Mio-Pliocene Ongeva Local Fauna of the Waite Formation
because although this material may not be Quinkana fortirostrum
sensu stricto, it appears to be more closely related to it than other
named Quinkana species and adds valuable morphological char-
acter states to this poorly known species (Yates, unpubl. data).
Pallimnarchus is in need of taxonomic revision. It is based on sev-
eral fragmentary nonoverlapping specimens that are currently
placed in two species. It is not at all certain that the many isolated
FIGURE 10. Kalthifrons aurivellensis, gen.
et sp. nov., SAM P35062, right mandibular
ramus. A, photograph in lateral view; B, pho-
tograph in medial view; C, photograph in
ventral view; D, photograph in lateral view
with interpretive overlay and transverse sec-
tions through dentigerous region of the man-
dible; E, photograph in medial view with
interpretive overlay; F, photograph in ventral
view with interpretive overlay. Abbrevia-
tions:an, angular; asd, articular surface for
reception of the dentary; d, dentary; emf,
external mandibular fenestra; fic, foramen
intermandibularis caudalis; sa, surangular; sp,
splenial; vlm, ventrolateral margin of the
dentary. Scale bar equals 100 mm.
Yates and Pledge—A mekosuchine crocodilian from Australia (e1244540-10)
fragments have been correctly apportioned to the two species, or
indeed if parts of Crocodylus sp. have not been included. Here we
combine all Pallimnarchus specimens into a single terminal taxon
and consider only those specimens that can be positively excluded
from Crocodylus. Mekosuchus kalpokasi is treated as a synonym
of Mekosuchus inexpectatus following Holt et al. (2007).
Crocodyloidea is represented by Crocodylus porosus, Crocodylus
moreletii, Crocodylus thorbjarnarsoni, Osteolaemus tetraspis, Voay
robustus, Mecistops cataphractus, Brochuchus pigotti,and
‘Crocodylus’ megarhinus. Gavialoidea is represented solely by the
tomistomines Maroccosuchus zennaroi, Dollosuchoides densmorei,
Toyotamaphimeia machikanensis,andTomistoma schlegelii. Gavia-
lis gangeticus was excluded because morphological atavism tends to
draw the taxon to the base of Crocodilia (Gatesy et al., 2003),
whereas several independent molecular data sets strongly place the
taxon as the extant sister taxon of Tomistoma schlegelii (Harshman
et al., 2003; Oaks, 2011). Putative extinct gavialines were also
excluded because it is currently impossible for simple parsimony
analyses of morphological data to disentangle those taxa that are
true gavialines with atavistic morphology from those that are basal
crocodilian taxa with convergent longirostrine morphology. This is
an interesting research question that needs to be tackled with more
sophisticated techniques, but is ignored here, where we are con-
cerned with the relationships of Kalthifrons aurivellensis either inside
or outside of Mekosuchinae.
Stem longirostrans are represented in the analysis by Prodi-
plocynodon langi,‘Crocodylus’ affinis,‘Crocodylus’ depressi-
frons, and Brachyuranochampsa eversolei. Alligatoroidea are
represented by Leidyosuchus canadensis, Brachychampsa mon-
tana, and Alligator mississippiensis. Lastly, two outgroup taxa,
Borealosuchus formidibalis and Bernissartia fagesii, were
included.
A heuristic search of the matrix was run using PAUP version 4.0
(Swofford, 2003) with a random addition sequence and 1000 repli-
cates. The search returned 566 most parsimonious trees, each with a
length of 588 steps (Fig. 13). The strict consensus of these trees
resolves many clades but also includes several significant polytomies.
Most significantly, Kalthifrons aurivellensis is recovered as a part of a
monophyletic Mekosuchinae. In this analysis, Mekosuchinae is
unambiguously supported by a penultimate premaxillary tooth as
large, or larger, than the fifth maxillary tooth (character 9, state 1);
concavities on the ventral surface of the pterygoids anterior and lat-
eral to the choana (character 86, state 1); largest maxillary tooth
greater than twice the diameter of the smallest interfestoonal maxil-
lary tooth (character 105, state 1); and second dentary tooth less than
60% of the diameter of the first (character 107, state 1). The follow-
ing character states may also diagnose Mekosuchinae but are ambig-
uous because the state is unknown in the basal A. clarkae:a
FIGURE 11. Kalthifrons aurivellensis, gen. et sp. nov., teeth. A–C, iso-
lated shed anterior tooth crown, tentatively referred to K. aurivellensis,
gen. et sp. nov. (SAM unregistered material, from type locality), in A,
labial, B, anterior, and C, basal views. D–E, posterior right dentary tooth
of the holotype (SAM P35062) in D, lingual and E, occlusal views. Scale
bar equals 10 mm.
FIGURE 12. Osteoderms tentatively
referred to Kalthifrons aurivellensis, gen. et
sp. nov. A–C, osteoderm, probably from right
side of nuchal shield (SAM P35062), in A,
dorsal, B, ventral, and C, posterior views. D–
F, paravertebral osteoderm from dorsal
shield (SAM P uncataloged, from type local-
ity) in D, dorsal and E, anterior views. F,
accessory osteoderm from dorsal shield in
dorsal view. G, unattached osteoderm, possi-
bly from flank, in dorsal view (SAM P unreg-
istered material, from type locality). Scale
bar equals 20 mm.
Yates and Pledge—A mekosuchine crocodilian from Australia (e1244540-11)
FIGURE 13. Results of the cladistic analysis of crocodilian relationships. A, strict consensus tree of 566 most parsimonious trees (tree length D588)
with select higher taxa named. B, strict consensus tree with support values (bootstrap support percentages followed by decay index) for each clade.
Only bootstrap support values over 50% are represented on the cladogram.
Yates and Pledge—A mekosuchine crocodilian from Australia (e1244540-12)
dorsoventrally expanded medial process of the prefrontal (character
22, state 0); occipitally exposed face of the quadrate ventrolateral to
the exoccipital (character 68, state 1); absence of a notch in the poste-
rior margin of the suborbital fenestra (character 71, state 0); ventral
ridges on the pterygoids extending anteriorly from lateral margins of
choana (character 87, state 1); anteroventrally oriented choanal
opening (character 89, state 1); lateral exposure of the prootic on the
braincase wall highly limited, largely covered by quadrate and later-
osphenoid (character 93, state 1); ornamentation along the ventrolat-
eral margin of the splenial (character 125, state 1); and
anteroposterior length of the neural spine of the third cervical verte-
bra is less than half of its centrum length (character 159, state 1). The
content of Mekosuchinae is uncontroversial, consisting only of
Kalthifrons aurivellensis and taxa that previously had been consid-
ered to be mekosuchines. The position of Mekosuchinae within
Crocodilia is less resolved, however. A large polytomy includes
Mekosuchinae together with Gavialoidea, Crocodyloidea,
‘Crocodylus’ depressifrons,‘Crocodylus’ megarhinus,andBrachyur-
anochampsa eversolei. Inspection of a sample of the most parsimoni-
ous trees indicates that Mekosuchinae can occupy a position both
inside Longirostres (as part of Crocodyloidea) or on the longirostran
stem.
Within Mekosuchinae, Australosuchus clarkae is resolved as the
sister taxon of all other mekosuchines. Within the clade of all other
mekosuchines, the base is represented by a large polytomy that
includes Kalthifrons aurivellensis. Inspection of a sample of most par-
simonious trees reveals that it is the variable position of Kalthifrons
aurivellensis itself that is largely responsible for this loss of resolution.
In some trees, Kalthifrons aurivellensis branches early as a sister
taxontothespeciesofKambara Call remaining mekosuchines,
whereas in others it is a derived member of Mekosuchini (the clade
of moderately to highly altirostral mekosuchines containing Baru,
Quinkana, and the dwarf mekosuchines). The lability of Kalthifrons
aurivellensis stems both from its poor preservation and from the con-
sequently low number of characters that can be scored, as well as the
incongruent nature of those character states that can be determined.
The relationships found in this analysis are, for the most part, not
very robust. In particular, the clades of interest, the Mekosuchinae
and all mekosuchines exclusive of Australosuchus clarkae (which
include Kalthifrons aurivellensis), are extremely weak, with boot-
strap frequencies of less than 5% (1000 bootstrap replicates, fast
stepwise addition) and a decay index of just 1 step (Fig. 13).
DISCUSSION
Among crocodilian taxa known from the Neogene of Aus-
tralia, Crocodylus can be ruled out as a potential systematic posi-
tion for K. aurivellensis. No diagnostic apomorphies of
Crocodylus described by Brochu (2000) can be observed on the
specimen, and some characters display traits that are inconsistent
with Crocodylus, such as the quadratojugal forming the ventro-
lateral corner of the infratemporal fenestra, absence of postero-
dorsal process of the jugal that extends up the posterior margin
of the infratemporal fenestra, and the presence of labiolingually
compressed teeth. Most significantly, the admittedly poorly pre-
served pterygoid plate shows that the area in front of the choana
is depressed relative to the base of the palatines and that there is
a pair of ridges extending anteriorly from the lateral margins of
the choana. These are found to be synapomorphies of Mekosu-
chinae in our analysis and indicate that Kalthifrons aurivellensis
is probably a part of the Australian mekosuchine radiation.
Although the evidence placing Kalthifrons aurivellensis in
Mekosuchinae is weak, there is even less evidence placing it in
any other group. As discussed above, K. aurivellensis does not
belong in Crocodylus. The only Neogene crocodilian from Aus-
tralia that is neither Crocodylus nor clearly a member of Meko-
suchinae is Harpacochampsa camfieldensis (Megirian et al.,
1991). The systematic position of this taxon has remained
unclear, with some authors indicating that it too may be a mem-
ber of the mekosuchine radiation (Willis, 1997a; Brochu, 2001).
As biogeographically tidy as that position may be, there is little
evidence to support it and other character traits, such as an ante-
riorly flaring lateral squamosal groove, an anterior squamosal
projection that does not underplate the postorbital, and a rela-
tively robust postorbital bar with a rather prominent anterolat-
eral protuberance, suggest that it may be gavialid (Yates,
unpubl. data). The presence of a gavialid in the Neogene of Aus-
tralia is not entirely unsurprising, because gavialids (including
both tomistomines and gavialines) appear to have been excellent
dispersers across marine barriers and were widespread from the
neotropics, through the Old World, across to Southeast Asia
(Hua and Jouve, 2004; Piras et al., 2007; V
elez-Juarbe et al.,
2007; Jouve et al., 2015). Regardless of its affinities, the narrow,
elongate rostrum, broad supratemporal fenestrae that dominate
the skull table, robust postorbital bar, and anteriorly flaring lat-
eral squamosal groove of Harpacochampsa camfieldensis differ
sharply from those of Kalthifrons aurivellensis and indicate that
these two taxa are not close relatives within Crocodilia.
Accepting that Kalthifrons aurivellensis is a mekosuchine, it is
not possible to determine its systematic position within this
clade. This is in part because poor preservation prevents deter-
mination of many character states, but also because the character
states that can be observed present conflicting signals. For exam-
ple, Kalthifrons aurivellensis displays a small, anteroposteriorly
narrow dorsal exposure of the supraoccipital, a plesiomorphic
trait otherwise restricted to the basal Australosuchus clarkae
amongst mekosuchines, whereas the anteriorly positioned orbits
are a derived trait that is diagnostic of derived mekosuchines in
the tribe Mekosuchini (Salisbury and Willis, 1996).
Currently, Kalthifrons aurivellensis is known only from its type
locality. Other crocodilian fragments from other localities in the Pli-
ocene Mampuwordu Sand Member have been mentioned before.
These include some fragments referred to Australosuchus clarkae
(Willis and Molnar, 1991). The authors noted that the apparent lon-
gevity of Australosuchus clarkae that is implied by these Pliocene
occurrences is remarkable and does not seem wholly plausible. They
suggested that reworking of older specimens, cataloging errors, or a
misidentification of stratigraphic position during collection explains
these anomalous records (Willis and Molnar, 1991). One of these
fragments, UCMP 100027, which consists of a frontal with adjoining
parts of the prefrontals, can be compared with Kalthifrons aurivellen-
sis. This specimen differs from Kalthifrons aurivellensis in lacking the
highly elongate anterior process and matches Australosuchus clarkae
in having an extremely narrow interorbital bridge. Thus the referral
of UCMP 100027 to Australosuchus clarkae stands. Reworking or
mistaken stratigraphic attribution during collection remain likely
explanations for this anomalous occurrence. Given that Kalthifrons
aurivellensis is a crocodilian generalist like Australosuchus clarkae
and that most sympatric crocodilian faunas support only one gener-
alist taxon, or have ecologically segregated generalist crocodilians
(Brochu, 2001), we consider it unlikely that Australosuchus clarkae
and Kalthifrons aurivellensis were sympatric.
Other crocodilian specimens known from the Mampuwordu Sand
Member include teeth that were referred to Crocodylus without
comment (Tedford et al., 1992). There are no diagnostic characters
of Crocodylus present in isolated teeth, and it is possible that these
teeth belong to Kalthifrons aurivellensis (or are reworked Australo-
suchus clarkae teeth). Lastly, Hecht and Archer (1977) reported a
maxillary fragment from the Woodard Quarry in the Mampuwordu
Sand Member that they identified as sebecosuchian on the basis of a
vertical maxillary wall (i.e., a highly altirostral snout) and strongly
labiolingually compressed ziphodont teeth. These characters are
also present in the mekosuchine genus Quinkana, and the specimen
cannot be distinguished from that genus. Although the specimen
(UCMP 113956) has been reported as lost (Willis, 1997a), it was
recently relocated (P. Holroyd, pers. comm., 2016) and can be clearly
Yates and Pledge—A mekosuchine crocodilian from Australia (e1244540-13)
distinguished from K. aurivellensis by its strongly mediolaterally
compressed teeth with recurved apices and the vertical maxillary
wall. The ziphodont teeth of Quinkana differ from those of K. auri-
vellensis in having much stronger labiolingual compression, posteri-
orly recurved apices, and denticulate carinae. Further differences
between K. aurivellensis and Quinkana can be observed when other
Quinkana specimens from other stratigraphic units are considered.
For example, the frontal of Q.sp.cf.fortirostrum (NTM P6282;
Fig. 7B) lacks the extremely elongate anterior process of K. aurivel-
lensis, and the squamosals of Q. timara and Q.sp.cf.fortirostrum
have rugose ornamentation that extends from the dorsal surface ven-
trally into lateral sulcus, unlike K. aurivellensis (Fig. 7C, D). Thus,
Quinkana sp. and K. aurivellensis indicate that there are two ecolog-
ically distinct crocodilian taxa present in the Mampuwordu Sand
Member.
Higher up in the Tirari Formation, the Pompapillina Member con-
tains an undescribed, broad-snouted, generalist species of Crocody-
lus (Yates, pers. observ. and unpubl. data) and shows no evidence
for the continued survival of Kalthifrons aurivellensis (Fig. 2). Based
on palaeomagnetic and biochronological data, the maximum and
minimum age constraints of the Pompallina Member are 3.9 and
3.4 Ma, respectively, with the age for the unit probably lying close to
the younger end of this range (Tedford et al., 1992). Thus, Crocody-
lus appears to have arrived in the Lake Eyre Basin in the mid-Plio-
cene between 3.9 and 3.4 Ma, at a time when Kalthifrons
aurivellensis may have still been extant, given its presence at 3.9 Ma.
This is the first example of immediate turnover from an endemic
mekosuchine to Crocodylus in the Australian fossil record. Although
the record from the Lake Eyre Basin is insufficient at this stage to
demonstrate that the extinction was caused by the arrival of Croco-
dylus, it is consistent with such an interpretation. Although a number
of diverse crocodilian faunas are know from around the world (e.g.,
Kuhn, 1938; Willis, 1997b; Scheyer et al., 2013; Salas-Gismondi
et al., 2015), it is usual for only a single generalist species to be pres-
ent (Brochu, 2001). The platyrostral, or mildly altirostral, broad-
snouted and partially interlocking dentition of Kalthifrons aurivel-
lensis indicate that it was an ecological generalist with a lifestyle simi-
lar to broad-snouted Crocodylus. This would have placed
Kalthifrons aurivellensis in direct competition with Crocodylus and
suggests competitive exclusion as a driver for the extinction of Kalth-
ifrons aurivellensis.
ACKNOWLEDGMENTS
We thank B. McHenry, J. McNamara, and J. Thurmer for their
assistance on the field expedition that uncovered the holotype speci-
men. We also thank P. Holroyd and E. Holt (UCMP) for their dili-
gent and successful search for the mislaid ziphodont maxilla from
the Mampuwordu Sand Member and images of that specimen. A.M.
Y. wishes to thank M. S. Y. Lee and B. McHenry (SAM) and S.
Hocknull and K. Spring (QM) for their assistance while visiting their
respective institutions. A.M.Y. also thanks M. Binnie (SAM) for
arranging a loan of the holotype specimen.
LITERATURE CITED
Alley, N. 1998. Cainozoic stratigraphy, palaeoenvironments and geologi-
cal evolution of the Lake Eyre Basin. Palaeogeography Palaeocli-
matology Palaeoecology 144:239–263.
Balouet, J.-C. 1991. The fossil vertebrate record of New Caledonia; pp.
1383–1403 in P. Vickers-Rich, J. M. Monaghan, R. F. Baird, and T.
H. Rich (eds.), Vertebrate Palaeontology of Australasia. Pioneer
Design Studio, Melbourne, Victoria.
Balouet, J.-C., and E. Buffetaut. 1987. Mekosuchus inexpectatus, n. g., n.
sp., crocodilien nouveau de l’Holoc
ene de Nouvelle Caledonie.
Comptes Rendus de l’Acad
emie des Sciences, Paris 304:853–857.
Brochu, C. A. 1999. Phylogeny, systematics, and historical biogeography of
Alligatoroidea. Society of Vertebrate Paleontology Memoir 6:9–100.
Brochu, C. A. 2000. Phylogenetic relationships and divergence timing of
Crocodylus based on morphology and the fossil record. Copeia
2000:657–673.
Brochu, C. A. 2001. Crocodylian snouts in space and time: phylogenetic
approaches toward adaptive radiation. American Zoologist 41:564–585.
Brochu, C. A., and G. W. Storrs. 2012. A giant crocodile from the Plio-
Pleistocene of Kenya, the phylogenetic relationships of Neogene
African crocodylines, and the antiquity of Crocodylus in Africa.
Journal of Vertebrate Paleontology 32:587–602.
Buchanan, L. A. 2009. Kambara taraina sp. nov. (Crocodylia, Crocodyloidea),
a new Eocene Mekosuchine from Queensland, Australia, and a revision
of the genus. Journal of Vertebrate Paleontology 29:473–486.
Callen, R. J., and R. H. Tedford. 1976. New late Cainozoic rock units and
depositional environments, Lake Frome area, South Australia.
Transactions of the Royal Society of South Australia 100:125–168.
Conrad, J. L., K. Jenkins, T. Lehmann, F. K. Manthi, D. J. Peppe, S.
Nightingale, A. Cossette, H. M. Dunsworth, W. E. H. Harcourt-
Smith, and K. P. McNulty. 2013. New specimens of ‘Crocodylus’
pigotti (Crocodylidae) from Rusinga Island, Kenya, and generic
reallocation of the species. Journal of Vertebrate Paleontology
33:629–646.
Gatesy, J., G. Amato, M. Norell, R. DeSalle, and C. Hayashi. 2003. Com-
bined support for wholesale taxic atavism in gavialine crocodylians.
Systematic Biology 52:403–422.
Gmelin, J. F. 1789. Caroli a Lin
ne, Systema Naturae, Tomus I. Pars III.
G. E. Beer, Leipzig.
Grigg, G., and C. Gans. 1993. Morphology and physiology of the Croco-
dylia; pp. 326–336 in C. G. Glasby, G. J. B. Ross, and P. L. Beesley
(eds.), Fauna of Australia, Volume 2A: Amphibia and Reptilia.
Australian Government Publishing Service, Canberra, Australian
Capital Territory.
Harshman, J., C. J. Huddleston, J. P. Bollback, T. J. Parsons, and M. J.
Braun. 2003. True and false gharials: a nuclear gene phylogeny of
Crocodylia. Systematic Biology 52:386–402.
Hecht, M. K., and M. Archer. 1977. Presence of xiphodont crocodilians in
the Tertiary and Pleistocene of Australia. Alcheringa 1:383–385.
Holt, T. R., S. W. Salisbury, T. Worthy, C. Sand, and A. Anderson. 2007. New
material of Mekosuchus inexpectatus (Crocodylia: Mekosuchinae) from
the Quaternary of New Caledonia. Unpublished conference paper.
Available at www.researchgate.net/publication/43494531. Accessed
January 18, 2016.
Hua, S., and S. Jouve. 2004. A primitive marine gavialoid from the Paleo-
cene of Morocco. Journal of Vertebrate Paleontology 24:341–350.
Jouve, S., B. Bouyac, M. Amaghzazc, and S. Meslouhd. 2015. Maroccosu-
chus zennaroi (Crocodylia: Tomistominae) from the Eocene of
Morocco: phylogenetic and palaeobiogeographical implications of
the basalmost tomistomine. Journal of Systematic Palaeontology
13:421–445.
Kuhn, O. 1938. Die Crocodilier aus dem mittleren Eoz€
an des Geiseltales
bei Halle. Nova Acta Lepoldina, N.F. 39:313–328.
Mannion, P. D., R. B. J. Benson, M. T. Carrano, J. P. Tennant, J. Judd,
and R. J. Butler. 2015. Climate constrains the evolutionary history
and biodiversity of crocodylians. Nature Communications 6:8438.
doi: 10.1038/ncomms9438.
Markwick, P. J. 1998. Fossil crocodilians as indicators of Late Cretaceous
and Cenozoic climates: implications for using palaeontological data
in reconstructing palaeoclimate. Palaeogeography, Palaeoclimatol-
ogy and Palaeoecology 137:205–271.
Mead, J. I., D. W. Steadman, S. H. Bedford, C. J. Bell, and M. Spriggs.
2002. New extinct mekosuchine crocodile from Vanuatu, South
Pacific. Copeia 2002:632–641.
Megirian, D. 1994. A new species of Quinkana Molnar (Eusuchia: Croco-
dylidae) from the Miocene Camfield Beds of northern Australia.
The Beagle, Records of the Museums and Art Galleries of the
Northern Territory 11:145–166.
Megirian, D., P. F. Murray, and P. M. A. Willis. 1991. A new crocodile of
the gavial ecomorph morphology from the Miocene of Northern
Australia. The Beagle, Records of the Museums and Art Galleries
of the Northern Territory 8:135–157.
Megirian, D., G. J. Prideaux, P. F. Murray, and N. Smit. 2010. An Austra-
lian land mammal age biochronological scheme. Paleobiology
36:658–671.
Molnar,R.E.1979.Crocodylus porosus from the Pliocene Allingham Forma-
tion of north Queensland. Results of the Ray E. Lemley Expeditions,
Part 5. Memoirs of the Queensland Museum 19:357–365.
Yates and Pledge—A mekosuchine crocodilian from Australia (e1244540-14)
Molnar, R. E. 1981. Pleistocene ziphodont crocodilians of Queensland.
Records of the Australian Museum 33:808–834.
Molnar, R. E. 1982. Pallimnarchus and other Cenozoic crocodiles of
Queensland. Memoirs of the Queensland Museum 20:657–673.
Molnar, R. E., T. Worthy, and P. M. A. Willis. 2002. An extinct Pleisto-
cene endemic mekosuchine crocodylian from Fiji. Journal of Verte-
brate Paleontology 22:612–628.
Moreno-Bernal, J. W., J. Head, and C. A. Jaramillo. 2016. Fossil crocodi-
lians from the High Guajira Peninsula of Colombia: Neogene faunal
change in northernmost South America. Journal of Vertebrate
Paleontology. doi: 10.1080/02724634.2016.1110586.
Oaks, J. R. 2011. A time-calibrated species tree of Crocodylia reveals a
recent radiation of the true crocodiles. Evolution 65:3285–3297.
Piras, P., M. Delfino, L. Del Favero, and T. Kotsakis. 2007. Phylogenetic
position of the crocodylian Megadontosuchus arduini and
tomistomine palaeobiogeography. Acta Palaeontologica Polonica
52:315–328.
Salas-Gismondi,R.,J.J.Flynn,P.Baby,J.V.Tejada-Lara,F.P.Wesselingh,
and P.-O. Antoine. 2015. A Miocene hyperdiverse crocodylian commu-
nity reveals peculiar trophic dynamics in proto-Amazonian mega-wet-
lands. Proceedings of the Royal Society of London B, Biological
Sciences 282:20142490. doi: 10.1098/rspb.2014.2490.
Salisbury, S. W., and P. M. A. Willis. 1996. A new crocodylian from the
early Eocene of south-eastern Queensland and a preliminary inves-
tigation of the phylogenetic relationships of crocodyloids. Alcher-
inga 20:179–226.
Scheyer, T. M., O. A. Aguilera, M. Delfino, D. C. Fortier, A. A. Carlini,
R. S
anchez, J. D. Carrillo-Brice~
no, L. Quiroz and M. R. S
anchez-
Villagra. 2013. Crocodylian diversity peak and extinction in the late
Cenozoic of the northern Neotropics. Nature Communications
4:1907. doi: 10.1038/ncomms2940.
Stirton, R. A., R. H. Tedford, and A. H. Miller. 1961. Cenozoic stratigra-
phy and vertebrate paleontology of the Tirari Desert, South Aus-
tralia. Records of the South Australian Museum 14:19–61.
Swofford, D. A. 2003. PAUP*: Phylogenetic Analysis Using Parsimony
(*And Other Methods), version 4.0. Sinauer Associates, Sunder-
land, Massachusetts.
Tedford, R. H., R. T. Wells, and S. F. Baghoorn. 1992. Tirari Formation
and contained faunas, Pliocene of the Lake Eyre Basin, South
Australia. The Beagle, Records of the Museums and Art Galleries
of the Northern Territory 9:173–194.
V
elez-Juarbe, J., C. A. Brochu, and H. Santos. 2007. A gharial from the
Oligocene of Puerto Rico: transoceanic dispersal in the history of a
non-marine reptile. Proceedings of the Royal Society of London B,
Biological Sciences 274:1245–1254.
Willis, P. M. A. 1993. Trilophosuchus rackhami gen. et sp. nov., a new
crocodilian from the early Miocene limestones of Riversleigh,
northwestern Queensland. Journal of Vertebrate Paleontology
13:90–98.
Willis, P. M. A. 1997a. Review of fossil crocodilians from Australasia.
Australian Zoologist 30:287–298.
Willis, P. M. A. 1997b. New crocodilians from the Late Oligocene White
Hunter site, Riversleigh, northwestern Queensland. Memoirs of the
Queensland Museum 41:423–438.
Willis, P. M. A. 2001. New crocodilian material from the Miocene of Riv-
ersleigh (northwestern Queensland, Australia); pp. 64–74 in G. C.
Grigg, F. Seebacher, and C. E. Franklin (eds.), Crocodilian Biology
and Evolution. Surrey Beatty and Sons, Chipping Norton, New
South Wales.
Willis, P. M. A., and R. E. Molnar. 1991. A new middle Tertiary crocodile
from Lake Palankarinna, South Australia. Records of the South
Australian Museum 25:39–55.
Willis, P. M. A., and R. E. Molnar. 1997. A review of the Plio-Pleistocene
crocodilian genus Pallimnarchus. Proceedings of the Linnean Soci-
ety of New South Wales 117:224–242.
Willis, P. M. A., P. Murray, and D. Megirian. 1990. Baru darrowi gen. et
sp. nov., a large, broad-snouted crocodyline (Eusuchia: Crocodyli-
dae) from mid-Tertiary freshwater limestones in Northern Aus-
tralia. Memoirs of the Queensland Museum 29:521–540.
Woodburne, M. O., B. J. MacFadden, J. A. Case, M. S. Springer, N. S.
Pledge, J. D. Power, J. M. Woodburne, and K. B. Springer. 1994.
Land mammal biostratigraphy and magnetostratigraphy of the Eta-
dunna Formation (Late Oligocene) of South Australia. Journal of
Vertebrate Paleontology 13:483–515.
Submitted February 29, 2016; revisions received July 21, 2016;
accepted September 5, 2016.
Handling editor: R. Butler.
Yates and Pledge—A mekosuchine crocodilian from Australia (e1244540-15)
