Full text of "Memoirs of the Queensland Museum"

See other formats

of the Queensland Museum | Nature

ISSN 1440-4788
brisbane 1 15 march 2010

55 ( 1 )

Memoirs of the Queensland Museum | Nature

55 ( 1 )

Minister: The Hon. Anna Bligh, MP, Premier of Queensland and Minister for the Arts

CEO: I.D. Galloway, PhD

Editor in Chief: J.N.A. Hooper, PhD

Subject Editors: CJ. Burwell, PhD; S.Van Dyck, PhD; J. Healy, PhD DSc; A.G. Cook,
PhD; PJ.F. Davie, PhD.

Managing Editor: S.M. Verschoore BA(VA)

PUBLISHED BY ORDER OF THE BOARD

15 MARCH 2010

© The State of Queensland, the Queensland Museum
PO Box 3300, South Brisbane 4101, Australia
Phone 61 3840 7555
Fax 61 7 3846 1226
www.qm.qld.gov.au

National Library of Australia card number
ISSN 0079-8835

COVER: Photograph of Mixophyes coggeri egg eight days after laying.

NOTE

Papers published in this volume and in all previous volumes of the Memoirs of the Queensland Museum
may be reproduced for scientific research, individual study or other educational purposes. Properly
acknowledged ouotations mav be made but queries regarding the republication of any papers should be
addressed to the Editor in Chief. Copies of the journal can be purcnased from the Queensland Museum Shop.

A Guide to Authors is displayed at the Queensland Museum web site www.qm.qld.gov.au/organisation/
publications/ memoirs/ guioetoauthors.pdf

A Queensland Government Project

Typeset at the Queensland Museum
Printed by Watson Ferguson & Co.,
655 Toohey Road, Salisbury Qld 4107

Queensland ' queensland museum

Government www.qm.qld.gov.au

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Breeding behaviour of the Barred
Frog Mixophyes coggeri

Conrad J. HOSKIN

School of Earth and Environmental Sciences, James Cook University, Cairns, Qld 4878, Australia;
New address: School of Botany and Zoology, The Australian National University, Canberra, ACT
0200, Australia. Email: conrad.hoskin@gmail.com

Citation: Hoskin, C.J. 2010 03 15. Breeding behaviour of the Barred Frog Mixophyes coggeri.
Memoirs of the Queensland Museum — Nature 55(1): 1-7. Brisbane. ISSN 0079-8835. Accepted:
26 March 2007.

ABSTRACT

A recent analysis of specimens assigned to the Northern Barred Frog Mixophyes schevilli
Loveridge, 1933 of the Wet Tropics region of north-east Queensland revealed three
genetically and morphologically distinct species (Mahony et al. 2006). Mixophyes
schevilli was retained as a species distributed in the northern and central Wet Tropics,
M. coggeri Mahony, Donnellan, Richards & McDonald. 2006 was described as a species
distributed throughout the Wet Tropics, and M. carbinensis Mahony, Donnellan, Richards
& McDonald, 2006 was described as a species restricted to the Carbine and Windsor
Tablelands (Mahony et al. 2006). All three species are large, terrestrial, rainforest-
restricted stream breeders. Importantly, the calls and breeding biology of the three
species have not been described, and differences in habitat preferences and other
ecological aspects have not been resolved. Here I present the advertisement (mating)
call of M. coggeri and an observation of the breeding behaviour of this species. The
call is a deep, reverberating ‘worg’ and the primary call parameters are presented. The
breeding behaviour is described in detail and is characterised by the female flicking
fertilised eggs up onto rocks and the bank overhanging a side-pool in the stream. The
calls and breeding biology of the other two Wet Tropics Mixophyes species remain unresolved.
□ advertisement call, oviposition, Myobatrachidae, Australia, north-east Queensland,
Wet Tropics.

Until recently Mixophyes schevilli was the sole
described Mixophyes species from the rainforest
of the Wet Tropics region, between Townsville
and Cooktown in north-east Queensland. A
recent analyses, however, split M. schevilli
into three genetically and morphologically
distinct species in the Wet Tropics: two widely
distributed species, M. schevilli and M. coggeri,
and a species restricted to the Carbine and
Windsor Tablelands, M. carbinensis (Mahony et
al. 2006). Little information has been published
on the breeding biology of Mixophyes schevilli,
the call is generally described as a deep 'wark'
(e.g. Barker et al. 1995; McDonald 2000) and
oviposition has been reported as eggs 'laid on soil

under banks above water' (McDonald 2000). The
recent revision of M. schevilli makes it unclear
which of the three species this information referes
to, and descriptions of the breeding behaviour of
each species are required. The breeding behaviour
of the Wet Tropics species is of particular interest
given the diversity in call structure and egg laying
behaviour among the four Mixophyes species
in south-east Australia (discussed below). Here
1 describe the call and a breeding observation
of M. coggeri from the Kuranda region in the
central Wet Tropics. Mixophyes coggeri occurs in
lowland and upland rainforest through much
of the Wet Tropics region, from Paluma (near
Townsville) to Big Tableland (near Cooktown)

Memoirs of the Queensland Museum | Nature • 2010 • 55(1) • www.qm.qld.gov.au

Hoskin

(Mahony et al. 2006). Across much of this distri-
bution the species co-occurs with M. schevilli,
while on the Carbine and Windsor Tableland
it is sympatric with M. carbinensis (Mahony et
al., 2006).

The calls, breeding observation and habitat
notes presented herein come from the Kuranda
region (16°45'-16°51'S, 145°33'-145°40'E, altitude
300-450m), west of Cairns. In this region I have
observed M. coggeri across the full spectrum
of rainforest types, from well-developed wet
rainforest to thin riparian rainforest strips in
otherwise open, sclerophyll-dominated forest.
Mixophyes coggeri calls from the vicinity of pools
on slow-moving streams that range in substrate
from sandy and boggy through to rocky. Males
and females are regularly observed foraging
at night, generally along stream banks and
at times some distance from streams. I have
not observed Mixophyes schevilli at these sites,
although it is known to occur in the Kuranda
region (Mahony et al., 2006). All individuals
herein were identified asM. coggeri based on the
diagnostic morphological and pattern characters
presented in Mahony et al. (2006): dorsal pattern
consisting of a series of irregular blotches (e.g.
Fig. 1), large size (male SVL > 80 mm) (Table 1),
broad head shape (HW/SVL 0.45-0.47) (Table 1),
and aspects of the patterning of the posterior
surface of the thigh.

CALL CHARACTERISTICS

The calls of three male M. coggeri were recorded
on Streets Ck (16°49'34"S 145°39'22"E) between
21 :00 and 23:30 hrs on 1 March 2007. These were
the only M. coggeri calling along a 350 m stream
transect that night and no females were observed.
Males 1 and 2 were calling on opposite sides of
the stream where a deep stream pool flowed into
riffles, and male 3 was calling 150 m upstream
near a stream pool between cascades (Fig. 2A).
The weather was overcast, warm, humid and
still. All three males were calling from elevated
earthy stream banks approximately 3 m from the
water and were partly covered by leaf-litter but
with the front half of the body exposed. Calls
were recorded with a Marantz DAT recorder
and a Sennheiser directional microphone, and
air temperature was taken (Table 1). Following
recording, each frog was placed in a clear plastic
bag and identified and measured (Table 1), after
which it was released at the exact point of capture
and photographed. Male 3 was identified as that
involved in the breeding observation the previous
night (described below), based on a comparison
of the dorsal pattern in photographs and the
fact that male 3 was calling within 5 m of where
the breeding pair had been found the previous
night. The software Soundruler 0.9.6.0 was used
to measure the following call parameters:
call interval (time from the end of one call
to the beginning of the next), call duration

TABLE 1 . Call characteristics of Mixophyes coegeri. Table presents the average and range (in brackets) of each

/u»*R a ^I?, eter ^w, r three males ' aIon S with ^eir snout to vent length (SVL), tibia leneth (TI ) head width
S the\hree^males'° (HW/SVL) ' and the air *™perature (T°cf tAb lasi'row is thf avlra^ of all^S

Individual

SVl (mm)

TL (mm)

1 l\Y (mm)

HW S\ I

Call int.
(sec.)

Call duration
(sec.)

Pulses
per call

Pulses
per sec.

Dominant
Freq. (Hz)

T°C

Male 1

87.2

52.6

39.6

0.45

(8-67)

0.218

(0.216-0.219)

(13-13)

59.7

(59.4-60.1)

536

(520-550)

23.5

Male 2

82.4

54.4

38.6

0.47

(29-78)

0.220

(0.200-0.237)

(12-14)

60.0

(59.2-60.4)

576

(560-580)

23.5

Male 3

85.8

51.9

40.1

0.47

(16-48)

0.259

(0.257-0.260)

(15-15)

58.0

(57.6-58.5)

574

(560-580)

23.0

Average

85.1

53.0

39.4

0.46

0.232

59.2

562

23.3

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Breeding behaviour of the Barred Frog

FIG. 1. Mixophye s coggeri, pair in amplexus.

(time from the beginning of the first
pulse to the end of the last pulse), pulses
per call, pulse rate (number of pulses
divided by call duration), and dominant
frequency (frequency at which the call
is of greatest intensity). Five successive

calls were analysed for each male to
give the average and range for each call
parameter for each male (Table 1).

The typical advertisement call of M.
coggeri in the Kuranda region is a deep,

FIG. 2. Stream habitat on Streets Ck, with the ovinosition site being the rock overhang marked by
an arrow in photo A and in the close-up photo (B).

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Hoskin

FIG. 3. Mixophi/es coggeri eggs adhered to the under surface of the rock overhang 12 hours after laying.

reverberating 'worg' repeated infrequently.
Call parameters of the three males recorded
at Streets Ck are presented in Table 1. Of the
approximately 50 calls recorded all were a
single 'worg', except one which was a double
'worg worg'. More recordings are required
to assess call variation in M. coggeri across the
extent of its range.

1 have heard a distinctly different, higher
pitched and less reverberating 'wark' or 'wark
wark' call at other sites in the northern and
central Wet Tropics, similar to that presented
for 'M. schevilli' by Stewart (1998a). This is likely
to be the mating call of M. schevilli but will
remain unresolved until the call of this species
is characterised and matched with genetic
or morphology/ pattern data. 1 have also
recorded a similar 'wark' call from Mixophyes
on the Windsor Tableland (where M. schevilli
has not been recorded (Mahony et al. 2006))
from males that match the morphology of M.

carbincnsis but not M. coggeri. This suggests
the call of M. carbincnsis is similar to that of M.
schevilli and differs from the deep, reverberating
call of M. coggeri. However, once again this
requires recordings matched with genetic or
morphology/ pattern data.

BREEDING OBSERVATION

An observation of M. coggeri breeding was
made on Streets Ck on the night of 28 Febuary/1
March 2007. Weather conditions were overcast,
warm (25°C), humid and still on the night. Heavy
rain had fallen over the previous fortnight but
little had fallen over the previous few days and
the stream level had dropped back to a 'normal'
level. A pair of M. coggeri were found in amplexus
(Fig. 1) at 22:30 hrs, halfway up a steep earthy
bank and about lm from the edge of a shallow
side-pool connected to the stream (top right of
Fig. 2A). Amplexus at this stage was axillary.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Breeding behaviour of the Barred Frog

with the male clasping the female just behind the
pectoral region. The pair remained amplexed for
the remainder of the observations but movements
described below were those of the female as
she was carrying the male. Movement was
characterised by short bursts of activity by the
female separated by long periods where the pair
remained perfectly still. At 23:45 hrs the female
moved to the edge of the side pool (mid right
of Fig. 2A) and the pair sat there until 00:30
hrs when the female worked her way through
the middle of a rock pile in the creek, ending
up in a shallow pool beneath a rock overhang
(marked on Figs 2A & B). At 00:45 hrs the pair
emerged from the overhang and sat on the edge
of the main stream pool (mid right of Fig. 2B)
until 01:30 hrs when the female moved around
the side-pool back to the point where they first
approached the stream (mid right of Fig. 2A).
The pair nestled into muddy leaf-litter on the
edge of the side-pool for 10 minutes before the
female jumped into the side-pool, swam across
it and then carried the male back through the
rock pile to the small pool below the overhang
(01:45 hrs). At this point no eggs had been laid
in the overhang or in the side-pool. The position
of the male had changed such that at this stage
amplexus appeared inguinal (axillary earlier).
The pair moved around in the small shallow pool
below the overhang (marked in Figs 2A & B) and
then at about 02:00 hrs started flicking eggs and
water up onto the ceiling and walls of the rocky
overhang. Sitting or floating in shallow water the
female would pause, lean forward and flick a
spray of eggs and water upwards with her back
legs (presumably immediately after the eggs had
emerged from her cloaca and been fertilised by
the male). Then the amplexed pair would move
around in the pool, pause, and flick again.
The egg flicking behaviour was similar to that
photographed in M. fasciolntus Giinther, 1864
(Anstis 2002, p. 218). This behaviour was still
continuing at 02:30 hrs when observations were
ceased, and at this point eggs and dripping
water were evident on the roof of the overhang

FIG. 4. Mixophyes coggeri egg 8 days after laying.
The tadpole is clearly discernable on the left hand
side of the egg, with its head facing forwards and the
tail curving away to the right. The tadpole hatched
out during rainfall shortly after the photograph
was taken.

and to a lesser degree on the adjacent sloping
rock walls and earth banks.

Observations the next day (1 March 2007)
revealed the pool beneath the overhang to be
about 80 cm long, 40 cm wide and from 5-15 cm in
depth. The rock above the pool made a roof about
20-30 cm above the water surface and the pool
was almost completely surrounded by sloping
rock and earth, with just two small 1 cm deep
channels linking it to the stream (Fig. 2B). The
rock pile was positioned between a slowly flowing
pool (3 m wide, 50 cm deep) in the main stream
channel and a linked side-pool (2 m wide, 20 cm
deep) (Fig. 2A). Most eggs (approximately 300)
were stuck to the rock roof of the overhang (Fig.
3) and were clumped (but generally in a single
layer) directly above tine water in a 60 cm by 30
cm strip, about 20 cm above the water surface.
Others (approximately 150 eggs) were scattered
on the sloping earth bank and sloping rocks
around the pool, generally within 10 cm (but up

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Hoskin

to 30 cm) from the water, and a small number of
eggs (about 40) were visible in the water. This
gives a clutch size estimate of about 500 eggs.
The eggs were pigmented (creamy brown)
and with visible animal and vegetal poles (Fig.
3). The eggs stuck to the overhang hung with
their darker, animal pole upwards. The egg
diameter averaged about 3.8 mm (about 4.2 mm
including the firm, clear egg capsule).

When I returned seven days later (8 March
2007) no live eggs were observed on the roof of
tine overhang but a small number of dry, shrivelled
eggs (approximately 50) remained. I assume the
remainder had hatched out and dropped into the
pool. Approximately 50 eggs were still present
on an earthy bank above the waters edge and
these were now large and a well-developed larva
was clearly visible in each (Fig. 4). The larva in
the eggs resembled the few hatchlings that were
visible in the pool below the overhang. It had
rained very little since egg laying and the water
level in the pool had dropped and it was now
isolated from the stream. Heavy rain began falling
as the observations were being completed and
an egg on an earth bank about 5cm from the
water was observed to rupture as it was wet by
water dripping off rocks into the overhang. The
hatchling wriggled vigorously until it entered
the shallow pool. The hatchling was patterned
with dark and golden markings and a dark bar
across the base of the tail. Observations ceased
due to heavy rain and it was assumed that other
eggs would have hatched as water ran off rocks
into the overhang, and that the pool below the
overhang would have re-connected with the
rising stream. Overall, the egg laying behaviour
and development of the eggs is similar to that
described for M. fasciolatus and M. iteratus
Straughan, 1968 (Anstis 2002; Harry Hines,
pers. comm.).

DISCUSSION

The mating call and breeding biology of the
Wet Tropics Mixophyes species is of particular

interest given the differences observed between
the four south-east Australian species, which can
be broken into two groups based on call structure
and breeding behaviour/habitat: 1. M. fasciolatus
and M. iteratus give a 'wark' or 'woh' style call
and their breeding behaviour is characterised
by the female (in amplexus) flicking fertilised
eggs up onto rocks or banks overhanging stream
pools, whereas, 2. M. fleayi Corben & Ingram,
1987 and M. balbus Straughan, 1968 give a
stuttered 'ok-ok-ok-ok-ok' style call and lay
their eggs in a circular depression (constructed
by the female during amplexus) in gravel or
leaf-litter in shallow stream riffles (Barker et
al. 1995; Stewart 1998b; Lewis 2000; Meyer
et al., 2001; Anstis, 2002). The call and breeding
biology of M. coggeri conforms to group 1
(M. fasciolatus and M. iteratus). In particular,
M. coggeri shows similarities to M. iteratus in
being of very large body size and having a
deep, reverberating call. Further research is
required to assess call variation across the
range of M. coggeri and to resolve the mating
call and breeding biology of M. schevilli and M.
carbinensis. This will determine whether the Wet
Tropics Mixophyes display the variation in call
structure and breeding behaviour seen in the
south-east Australian species. Characterising
differences in the advertisement call and habitat
requirements between the three Wet Tropics
species will also allow an assessment of the
mechanisms of reproductive isolation between
the sympatric species pairs.

ACKNOWLEDGEMENTS

I thank Greg Dawe and Eleanor Hoskin for
assistance, and the reviewers for constructive
comments. The research was conducted during
a Tropical Landscapes Joint Venture Project
between James Cook University (Miriam Goosem)
and the Queensland Department of Main Roads,
assessing the impact of highways on frogs.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Breeding behaviour of the Barred Frog

LITERATURE CITED

Anstis, M. 2002. Tadpoles of south-eastern Australia: a
guide with keys. (New Holland: Sydney).

Barker , }., Grigg, G. & Tyler, M. 1995. A field guide to
Australian frogs. (Surrey Beatty & Sons: Sydney).

Cogger, H.G. 2000. Reptiles and amphibians of Australia
(6th ed.). (Reed New Holland: Sydney).

Hero, J.-M. & Fickling, S. 1997. A guide to stream-
dwelling frogs of the Wet Tropics rainforests. (James
Cook University: Queensland).

Lewis, B. 2000. A breeding observation of the
Stuttering Frog ( Mixophyes balbus) in northern
New South WaTfes. Herpetofauna 30: 30-33.

Mahony, M., Donnellan, S.C., Richards, S.J.
& McDonald, K. 2006. Species boundaries

among barred river frogs, Mixophyes (Anura:
Myobatrachidae) in north-eastern Australia, with
descriptions of two new species. Zootaxa 1228:
35-60.

McDonald, K.R. 2000. Frogs. In: Wildlife of Tropical
North Queensland, eds. M. Ryan & C. Burwell.
(Queensland Museum: Brisbane).

Meyer, E„ Hines, H. & Hero, J.-M. 2001. Wet forest
frogs of south-east Queensland. (Griffith University:
Queensland).

Stewart, D. 1998a. Audio CD: Australian frog calls:
tropical north-east. (Nature Sound: Mullumbimby,
NSW).

1998b. Audio CD: Australian frog calls: subtropical
east. ( Nature Sound: Mullumbimby, NSW).

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Bishop

A Triassic conchostracan from near Murgon, SEQ.

Eastern Australian Early- to Mid-Triassic conchostracans
have been described by Webb (1978) from the Sydney
Basin and Tasch (1979)" from the Bowen Basin. A major
summary of Gondwanan occurrences, including Australia,
was provided by Tasch (1987). 1 here report the first Mid to
Late Triassic conchostracan from the Aranbanga Volcanics
Group (Donchak et al. 1999), 14 km north of Murgon, southeast
Queensland. The genus reported here is cosmopolitan in
distribution (Tasch 1969) with many species known from
the southern continents (Tasch 1987).

The Aranbanga Volcanics are a diverse suite of volcanics
and volcanogenic sedimentary rocks occupying a wide band
throughout the south Burnett region of south’ east Queensland.
It contains a diverse flora and a fauna including insects and
freshwater bivalves. The Aranbanga Volcanics have been
ascribed a Mid to Late Triassic age based on the flora and a
K-Ar date (Day et al. 1983).

Systematic Palaeontology
Order CONCHOSTRACA Sars, 1867
Suborder SPIN1CAUDATA Linder, 1945
Superfamily CYZICOIDEA Stebbing, 1910
Family EUESTHERIIDAE Defretin, 1965
Euestheria Deperet & Mazeran, 1912

Type Species. Posidcmia minuta Von Zeiten, 1833 by sub-
sequent designation of Raymond (1946), from the Upper
Triassic of Great Britain.

Euestheria acampestria sp. nov. (Fig. 1)

Material. Holotype: QMF54807, right valve. Paratypes
QMF54794-54806, QMF54808 (11 left, 3 right valves); all
from QML1331, Aranbanga Volcanics Group mid- to late
Triassic, near Murgon; 151° 53’ 20” E, 26° 08' 10" S. Note that
these specimens are selected from a slab bearing several
hundred individuals.

Diagnosis. Valves medium sized (up to 8.8 mm long); hinge
short; micro-ornament fine and papillate.

Etymology. Latin; campestria meaning flat plain, prefix 'a-'
meaning without; acampestria referring to an apparent lack
of a flat elongate hingeline.

Description. Valves 6.1-8.8 mm long, elliptical with width
approximately two-thirds of length; umbo small, subterminal,
approximately one-sixth of the length of the valve in from
point of maximum anterior bulge. Dorsal margin completely
rounded, with hingleline very short (to the extent of appearing
absent). Valve is fairly symmetrical about the midline
transverse axis (perpendicular to dorsal margin). Maximum
ventral bulge approximately at mid-length.

Growth bands comarginal, distinct, rugose, 10-16 per valve
(in specimens under discussion, with larger valves having
more bands), more pronounced toward ventral margin where
they are closely spaced; elsewhere bands are more widely
spaced. Growth bands nonexistent near umbo.

FIG. 1. Euestheria acampestria sp. nov.; A, Holotype QMF54807,
right valve, lateral view, scale bar = 1 mm; B, Paratype QMF54805,
right valve, lateral view, scale bar = 1 mm; C, Detail of ornament on
paratype QMF54801, scale bar « 0.5 mm. D. Detail of ornament,
paratype QMF54804, scale bar = 0.5 mm.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A Triassic Conchostracan

Micro-ornament uniformly spaced, small, round papillae,
which are slightly smaller than the distance between growth
lines; papillae more visible toward the margin.

Remarks. The material is assigned to Euestheria on the basis of
prominent growth lines, the papillate ornament, resembling
the 'polygons' in the type species and the relatively short
hinge line. The material is similar to Cyzicus (Euestheria)
dualis Tasch 1987 from the Lower Triassic Panchet Formation
of India and Cyzicus (Lioestheria) disgregaris Tasch 1987 from
the Jurassic Blizzard Heights and Storm Peak localities of
Antarctica, with approximately symmetrical and elliptical
valves and the small subterminal umbo lacking growth
bands near it. It is distinguishable from C. dualis however
by the lack of an elongate ningeline along the dorsal margin
and from C. disgreganous by its fewer, more widely spaced
growth bands, an umbo more anteriorly located and lack of
barred ornament. The papillate micro-ornament is similar
to Cyzicus (Euestheria) castaneus Tasch 1987 from the Jurassic
at Blizzard Heights and Storm Peak in Antarctica, but in C.
castaneus the microstructure is finer. C. castaneus also differs
by having a more elongate hingeline. This is also the case for
Endolimnadiopsis eicnwakh (Netshajev) (Shen, 1985) from
the Upper Permian Kazan Formation of Russia. The species
Endolimnadiopsis rusconii Gallego, 2004 from the Upper
Triassic Cacheuta Formation of Argentina differs from E.
acampestria in having an elongate posterior hingeline and
finer, variably-sized papillate micro-ornament.

Acknoweldgements

Paul Tierney, Scott Hocknull and Mark Saul are thanked
for supplying the specimens. Drs Alex Cook and Peter Jell
are thanked for their input in the writing and editing of the
manuscript.

Literature Cited

Day, R.W., Whitaker, W.G, Murray, C.G., Wilson, I.H. & Grimes,
K.G. 1983. Queensland Geology, A companion volume to the
1:2 500 000 scale geological map (1975). Publication of the
Geological Survey of Queensland, v 383.

Defretin-Lefranc, S. 1965, Etude et revision de phyllopodes
conchostrac£s en provenance d'U.R.S.S. Annual Geological
Society Norway. T. 85.

Dep6ret, C. & Mazeran, P. 1912. Les Estheria du Permian d'Autun,
Soriet6 d'Histoire Naturalle d'Autun Bulletin 25:165-173.

Donchak, P.J.T., Cranfield, L.C, & Pascoe, GS. 1999. Murgon 1:100
000 Geological Map. (Queensland Government, Brisbane).

Gallego, O.F. 2004. First record of the family Palaeolimnadiopseidae
Defretin-Le Franc, 1965 (Crustacea-Conchostraca) in the
Triassic of Argentina. Journal of South American Earth Sciences
18:223-231.

Raymond, P.E. 1946. The genera of fossil Conchostraca- an order
of bivalved Crustacea, Bulletin Harvard University Museum of
Comparative Zoology. 96: 218-307.

Shen, Y.B. 1985. Classification and evolution of the Family
Palaeolimnadiopseidae (Conchostraca). Scientia Sinica (Series
B) 28: 888-894.

Tasch, P. 1969. Branchiopoda. Pp. R129-R191. In Moore, R.C. (ed.)
Treatise on Invertebrate Paleontology, Part R. Arthropoda
4. (Geological Society of America and University of Kansas
Press, Lawrence, Kansas)

1979. Permian and Triassic Conchostraca from the Bowen Basin
(with a note on a Carboniferous leaiid from the Drummond
Basin), Queensland. Bureau of Mineral Resources, Geology and
Geophysics Bulletin 185: 31-43.

1987. Fossil Conchostraca of the Southern Hemisphere and
Continental Drift: Geological Society of America, Memoir
165: 1-290.

Von Zeiten, G.H. 1833. Die versteinerungen Wtlrtembergs: Stuttgart
(Zusammengestelt von G.H. Zeiten) p. 72, pi 54 not seen.

Webb, J.A. 1978. A new Triassic Palaeolimnadiopsis (Crustacea,
Conchostraca) from the Sydney Basin, New South Wales.
Alclieringa 2: 261-267.

Peter J. Bishop, Queensland Museum, Geosciences, 122 Gerler
Rd, Hendra Qld. 4011, Australia. 16th of January 2009.

Memoirs of the Queensland Museum

Nature • 2010 • 55(1)

A new subgenus and five new species of Australian
Glow-worms (Diptera: Keroplatidae: Arachnocampa)

Claire H. BAKER

School of Integrative Biology, The University of Queensland, Brisbane Qld, 4072, Australia. Email:
bugsyclaire@yahoo.com.au

Baker, C. H. 2010 03 15. A new subgenus and five new species of Australian glow-worms (Diptera:
Keroplatidae: Arachnocampa). Memoirs of the Queensland Museum — Nature 55(1): 11-43. Brisbane.
ISSN 0079-8835. Accepted: 25 February 2008.

ABSTRACT

A new subgenus, Arachnocampa (Lucifera) subgen. nov., is described to include the
Tasmanian species, Arachnocampa tasmaniensis Ferguson and the Mount Buffalo glow-
worm, Arachnocampa buffaloensis sp. nov. The new subgenus is separated from the
subgenera Arachnocampa ( Arachnocampa ) and Arachnocampa ( Campara ) by differences in
wing venation. The subgenus Arachnocampa now includes only the New Zealand species
A. luminosa (Skuse) which differs from species of Lucifera and Campara by its unusual
method of vertical pupal suspension. The Australian species A. tasmaniensis, A. flava
Harrison and A. richardsae Harrison are redescribed. Five new species are described:
A. buffaloensis sp. nov. (Victoria) within the subgenus Lucifera; and A. tropica sp.
nov. (north Queensland), A. gippslandensis sp. nov. (eastern Victoria), A. otwayensis sp.
nov. (western Victoria) and A. girraweenensis sp. nov. (southeast Queensland/northern
New South Wales) within the subgenus Campara. □ troglophile, rainforest, threatened
species, glow-worm, Keroplatidae, Arachnocampa, Lucifera, Campara.

The genus Arachnocampa (Edwards 1924)
contains species with bioluminescent larvae
that are commonly known as glow-worms.
Historically Arachnocampa contained four
described species, three of which are endemic
to Australia: A. flava Harrison from southeast
Queensland (Perkins 1935; Harrison 1966);
A. richardsae Harrison from New South Wales
(Harrison 1966) and A. tasmaniensis Ferguson
from Tasmania (Ferguson 1925). The fourth
species, A. luminosa (Skuse 1890), is endemic
to New Zealand. Arachnocampa are recognised
as commercially valuable organisms due to the
importance of glow-worms to tourism (Baker
2002). However, little was known about the
insects around which this industry is based
(Baker 2004). In Australia, Arachnocampa have
been documented from several locations well
separated from the known distributions of the
three described species (Table 1). Until this

study, these populations had been tentatively
assigned to either A. flava or A. richardsae (e.g.
Crosby 1978). The most recent reviews of the
Arachnocampa taxonomic literature are by Pugsley
(1983), Meyer-Rochow (2007) and Baker (2009).

The first description of an Australian glow-
worm was of A. tasmaniensis from the Ida
Bay Caves, Tasmania (Ferguson 1925). The
description was based on the adult form, with
a brief mention of larval biology. Other large
colonies of glow-worms known from Numinbah
in the Nerang Valley, Queensland and from New
South Wales (Perkins 1935) were subsequently
described as A. flava and A. richardsae respec-
tively, and both species were placed in a
new subgenus, Campara Harrison, separated
from the Tasmanian and New Zealand
species based on wing venation (Harrison
1966). McKeown (1935) noted the presence

Memoirs of the Queensland Museum | Nature • 2010 • 55(1) • www.qm.qld.gov.au

Baker

FIG. 1. Map of eastern Australia showing locations
of published Arachnocampa sites based on literature
surveys. Table 1 lists the names of each site. Three
species of Arachnocampa have been described from
these sites with many other sites documented but with
no follow-up morphological identification until the
present study.

of Arachnocampa in the Otway Forest, Victoria.

. Crosby (1978) later tentatively identified this
population as A. richardsae. However, this colony
of Arachnocampa represents a new species, and is
described here. Traditionally glow-worms have
been included within the Mycetophilidae but
following a review of the higher taxonomy of
the Mycetophiloidea, Arachnocampa and other
bioluminescent genera were placed in the family
Keroplatidae (Matile 1981).

Arachnocampa are highly susceptible to desi-
ccation and therefore require high humidity or
direct contact with water within their habitat for

FIG. 2 Map of eastern Australia showing the collection
localities of specimens used for morphological
examination. Regional groups are encompassed
within circled areas. Site names are found in Table 2.

survival (Richards 1960; Baker & Merritt 2003;
Baker 2004). All published Arachnocampa sites
in Australia are in rainforest, wet sclerophyll
forest or limestone or granite boulder caves
(Fig. 1, Table 1). Many of these records simply
noted the presence of larvae and provided
no additional information. Furthermore, the
taxonomy of Australian Arachnocampa species
in relation to their apparently widespread
but patchy distribution is poorly understood,
with the last taxonomic study by Harrison
(1966). Arachnocampa display limited mobility,
a shortened adult dispersal stage, and are
reliant on specific high humidity environments
for survival (Richards 1960; Baker 2004).
Their primary reliance on habitats of ancient

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Taxonomy of Australian glow-worms

TABLE 1. Published records of known Arachnocampa spp. locations in Australia illustrated in FIG. 1.
Holotype locations from published descriptions indicated as follows: A. flava holotype *, A. richardsae
holotype **, A. tasmaniensis holotype ***.

1 Aust.

1 State

Location of Arachnocampa spp. colony

Reference

QLD

1. Natural Bridge, Springbrook National Park *

Perkins 1935, Harrison 1966

2. River Cave, Girraween National Park

Finlayson 1982

NSW

3. Gloucester Cave, Gloucester

Harrison 1966, Eberhard & Spate 1995

4. Grose Valley, Blue Mountains

Currey 1966, Harrison 1966

5. Hazelbrook, Blue Mountains

Harrison 1966

6. Newnes railway tunnel **

Harrison 1966

7. Bundanoon

McKeown 1935

VIC

8. Underground River Cave, Mt Buffalo National Park

Crosby 1978

9. Walhalla mine shafts, Walhalla

Harrison 1966

10. Madsen's Track, Melba Gully State Park

Department of Conservation 1994

! TAS

11. Loongana (Old tourist cave)

Geode 1967

12. Gunns Plains Caves (Guns Plains Tourist Cave)

McKeown 1935, Geode 1967

13. Mole Creek Caves (Marakoopa, Wet,

Lynds and Westmoreland caves)

Geode 1967, Ferguson 1925

14. Kelley Basin (small unnamed cave)

Geode 1967

15. Florentine Valley and Junee Area (Cashion

Creek, Growling Swallet and Junee Caves)

Geode 1967

16. Ida Bay CavesJEntrance and exit caves) ***

Geode 1967, Ferguson 1925

Gondwanan origin, means that many popu-
lations are likely to have been fragmented in
refugial habitat' pockets for millions of years
through natural contractions and expansions,
with little to no chance of present day dispersal
across large geographic boundaries (Baker 2004;
Baker et al. 2008).

Australian glow-worm populations are under
increasing pressure due to their restricted
habitat range, critical habitat requirements, and
increasing tourism popularity in these regions
(Baker 2002, 2004). Therefore there is an urgent
need for careful management to ensure the long-

term sustainability of these species in Australia.
Correct management guidelines can only be
procured through a better understanding of
the species present and, as such, this study was
long overdue.

The taxonomic descriptions appearing in this
paper were intended to be published prior to
a molecular phylogenetic study of Australian
glow-worms (Baker et al. 2008). However, the
molecular study was published before the
present descriptive paper, and the scientific names
of the new taxa described here were inadvertently
used in that paper. Baker et al. (2008) did not

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Baker

TABLE 2. Map locations for Figure 2 with site names and codes used in descriptions. GPS coordinates fot-
each site are listed in the text. All sites are epigean unless specified as a cave or man-made tunnel.

Map location
Figure 2

Site Name

Site Code
used in text

Mossman Gorge

NQLD1

Mt Lewis

NQLD2

Lamb Range National Park

NQLD3

Mt Hypipamee

National Park

NQLD4

Bartle Frere Cave,
Wooroonooran

National Park

NQLD5

Bartle Frere stream

NQLD6

Mungalli Falls

NQLD7

Charmilllan walking trail,
Tully Gorge State Park

NQLD8

Birthday Creek Falls,

Paluma National Park

NQLD9

Tamborine Mountain

TAM1

i 12

Natural Bridge,

Springbrook National

Park (large overhang)

CALD1

Springbrook Plateau

CALD2

Springbrook National Park

CALD3

Lamington National Park

CALD4

South Bald Rock Cave,
Girraween National Park

G1RRA1

i, li

Ramsey Creek Cave,
Girraween National Park

GIRRA2

Mt Warning National Park

CALD5

Protestors Falls, Nightcap
National Park

CALD6

Washpool National Park

NNSWl

New England National Park

NNSW2

Dorrigo National Park

NNSW3

Barrington Tops

National Park

NNSW4

Gloucester Cave

NNSW5

Newnes Railway tunnel
(man-made tunnel)

SNSW1

Map location
Figure 2

Site Name

Site Code
used in text

Waterfall Springs
Conservation Park

SNSW2

Upper Kurrajong

SNSW3

Grand Canyon walk. Blue
Mountains National Park

SNSW4

Bundanoon

SNSW5

"The Grotto" Fitzroy

Falls National Park
(large overhang)

SNSW6

Underground River Cave,

Mt Buffalo National Park

BUFFI

Melba Gully State Park

WVIC1

Beauchamp Falls,

Otway National Park

WVIC2

Hopetoun Falls, Otway
National Park

WVIC3

Beauty Spot Reserve

WVIC4

Grey River Picnic area,
Angahook-Lorne State Forest

WVIC5

She-oak Picnic area,
Angahook-Lorne State Forest

WVIC6

Upper Yarra Valley
mine tunnel

EVIC1

O'Shannassy Weir
(man-made tunnel)

EVIC2

Britannia Creek

Cave, State Forest

EVIC3

Shining Star Gold Mine,
Warburton (tunnel)

EVIC4

Shiprock Falls Cave,

Kilnkurth State Forest

EVIC5

Labertouche Cave

EVIC6

Walhalla Mine tunnel

EVIC7

Gunns Plains Caves

TAS1

Marakoopa Cave

TAS2

Sassafras Cave

TAS3

Derby Mine tunnel

TAS4

Francis Gully, Ida Bay

TAS5

- «_ _ __

Entrance ('Mystery

Creek') Cave, Ida Bay

TAS6

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Taxonomy of Australian glow-worms

FIG. 3. Wing venation of Arachnocantpa (dorsal view of
right wing). Dots indicate the location of campaniform
sensillae. C, costa; CuAl, anterior branch of cubitus;
CuA2, posterior branch of cubitus; m-cu, posterior
section of medial-cubital cross vein; M, media; r-m,
radial-medial cross vein; R, radial vein; 1A, anal
vein. Wing length measurements were taken from
basal end to the wing apex.

fix holotypes for the new species, so the species
names used there are not available (International
Code of Zoological Nomenclature, 1999, Article
16.4). The subgenus name Lucifera as used by
Baker et al. (2008) is also not available. It is a
nomen nudum because the authors did not fix a
type species (International Code of Zoological
Nomenclature, 1999, Article 13.3). However, in
order to preserve the link with the taxa referred
to in the molecular paper (Baker et al. 2008),
the same names are used here for the formal
descriptions of the new subgenus and species.

MATERIALS AND METHODS

Taxon Sampling, Rearing and Labelling.

Arachnocampa larvae were collected from 49
sites including many new locations (Fig. 2,
Table 2). A GPS (eTrex®) was used to record
site coordinates, however, due to dense

FIG. 4. Morphology and terminology of Arachnocampa
thorax (revised from Matile 1990). ab tg 1, abdominal
tergite 1; anepm, anepimeron; anepst, anepistemum;
a spr, anterior spiracle; cx, coxa; hit, haltere; keptst,
katepisternum; ftgt, laterotergite; mtg, mediotergite;
mtkepst, metakatepisternum; mtn, metanotum; pal,
post alar region; patg, paratergite; p spr, posterior
spiracle; preset sut, prescutal suture; set, scutum;
sctl, scutellum; spal, supra-alar region.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Baker

FIG. 5. Slide mounted Aradmocampa flava larval head
capsule (dorsal view), ant, antennal stubb; 1, labrum; s,
stemmatal eye.

FIG. 6. Scanning electron micrograph (SEM) of
Aradmocampa flava head capsule (ventral view),
ant, antennal stubb; 1, labrum; man, mandible; max,
maxilla with maxillary teeth; s, stemmatal eye.

1 mm

FIG. 7. SEM of setae on the distal end of Arachnocampa
flava larva. These setae are used by larvae for
movement.

FIG. 8. 'Tiger stripe' colouration of the Aradmocampa
abdomen (each segment is lighter basally and
darker apically).

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Taxonomy of Australian glow-worms

FIG. 9. Adult Araclwocampa antenna showing the
progressive shortening of antennal segments from
segments 1-13. Segment 14 is longer than 13, and is
swollen, with a narrowed, apex.

0.5 mm

FIG. 10. Adult Araclwocampa head showing the
placement of the ocelli on a raised ridge. The median
ocellus faces anteriorly while the lateral ocelli are
directed laterally.

rainforest coverage the Geoscience Australia
website was used to obtain GPS coordinates
for some localities (http://www.ga.gov.
au/map/names/). Specimens were sent live
on ice to the University of Queensland where
they were placed in individual containers for
rearing to adulthood following methods
outlined in Baker & Merritt (2003). Each examined
specimen was labelled with the following data:
state, locality, GPS coordinates, habitat, date and
collectors. Holotype specimen are lodged in their

corresponding state museums with paratypes
distributed between the other state museums.

Terminology and Measurements. Morphological
terminology follows McAlpine (1981) with minor
modifications listed in the text. Terminology and
measurements of wings and thoracic segments
are illustrated in Figs 3-4. Measurements and
terminology of the larval head capsule are
indicated on Figs 5-6. Measurements listed in
the text are taken primarily from the holotype,
with paratype variation listed in parentheses or
as ranges. In some instances, variation is listed
from specific populations within a regional
group, but mostly ranges are given separately for

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Baker

males and females from entire regional groups.
Measurements were taken from at least six to
ten specimens where possible. If the number
of available specimens of a species was low, all
specimens were measured. Counts of the number
of maxillary teeth were taken from the 5 lh instar
exuviae after the onset of pupation or adult
emergence. Larvae moult their exoskeleton and
push it to the posterior end of the pupal case
(Baker & Merritt 2003). This method was used
to avoid potential differences in the numbers of
maxillary teeth among different instars.

Preparation and Examination of Slide-Mounted
Material. Specimens stored in 70% ethanol were
progressively dehydrated (increasing by 10%
concentration at each change) to 100% ethanol,
cleared in 10% KOH on a heating block for
15-30 minutes, and washed in distilled water
before being slide-mounted in glycerin or
glycerin gel. Characters were examined using a
Zeiss™ Stemi SV6 binocular dissecting micro-
scope or Zeiss™ Axioscop binocular compound
microscope. Pictures were taken using a colour
3CCD digital camera mounted onto the micro-
scope. Measurements and digital images were
taken using Scion Image V. 1.62. Scale bars and
annotations were added in Adobe Photoshop
5.5. Scanning electron micrographs (SEM)
were taken by Anthony O'Toole and Frederic
Beaulieu, School of Life Sciences, University
of Queensland. Images were taken on a Phillips
XL20 at the centre for Microscopy and Micro-
analysis, University of Queensland.

Type Material. Holotypes were chosen as the
specimen that best represented the species based
on morphological characters. Type specimens
of previously described species proved to be
brittle and shrivelled, and of limited use for
morphological examination. Therefore fresh
specimens obtained from the type localities were
used to prepare redescriptions.

Permits For Specimen Collection. Specimens
included in this study were collected under

the following scientific permits: Queensland,
permit number W4/002525/00/SAA, New
South Wales permit number B2111, Tasmania
permit number 00093, Victoria permit number
10000827. Abbreviations for field collectors listed
in the material examined sections are as follows:
AM, Andrew McDowell; CB, Claire Baker; AR,
Andrew Ridley; DM, David Merritt; ASH,
Andrew Shek; JF, Josh Fartch; AS, Amanda
Smith; SB, Stuart Baird.

Type Deposition. Type specimens of described
species were borrowed from the Australian
National Insect Collection (ANIC), Canberra
and The South Australian Museum (SAM),
Adelaide. Types have been deposited as per
permit requirements, with holotypes deposited
in their respective state insect collections and
paratvpes deposited across various collections
as follows: Australian Museum (AM), Sydney;
Museum Victoria (MV), Melbourne; Queensland
Museum, (QM), Brisbane; Tasmanian Museum
and Art Gallery (TM), Hobart; University of
Queensland Insect Collection (UQIC), Brisbane.

SYSTEMATICS

CHECKLIST OF THE GENUS
ARACHNOCAMPA.

The following taxonomic checklist is modified
from Matile (1981). A new subgenus, Lucifera,
is proposed to include A. tasmaniensis and A.
buffaloensis. Four new species are allocated to
the subgenus Campara, with the type species, A.
luminosa in subgenus Arachnocampa.

FAMILY KEROPLATIDAE
SUBFAMILY ARACHNOCAMPINAE

Genus Arachnocampa (Edwards, 1924)

Arachnocampa Edwards, 1924: 177. Type species: Bolitophila
luminosa Skuse, 1890 by original designation.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Taxonomy of Australian glow-worms

FIG. 12A-H. Arachnocampa spp. wings. A, A. tasmaniensis; B, A. buffalocnsis; C, A. tropica ; D, A.flava ; E, A.
girraweenensis ; F, A. richardsae ; G, A. gippslandensis ; H, A. otwayensis. cams, campaniform sensillae.

Subgenus Arachnocampa Edwards

Arachnocampa Edwards, 1924: 177 (as genus). Type species:
Bolitophila luminosa Skuse, 1890 by original designation.

Subgenus Lucifera subgen. nov.

tasmaniensis Ferguson, 1925: 487. Australia (TAS). Type
species

buffalocnsis sp. nov. Australia (VIC)

Subgenus Campara Harrison

Campara Harrison, 1966: 880. Type species: Arachnocampa
richardsae Harrison 1966 by original designation.
flava Harrison, 1966: 880. Australia (Queensland)
richardsae Harrison, 1966: 881 . Australia (New South Wales)
tropica sp. nov. Australia (Queensland)

girraweenensis sp. nov. Australia (Queensland/ New South
Wales)

otwayensis sp. nov. Australia (Victoria)
gippslandensis sp. nov. Australia (Victoria)

REDESCRIPTION OF
ARACHNOCAMPA EDWARDS

DIAGNOSIS. Empodia and pulvilli absent. Radial
sector originating far before middle of wing and far
before tip of costa. R 4 absent. Larval head capsule as
long as broad. Labrum hood-like. Larval mandibles
longer than broad with prominent teeth. Larval
antennae vestigial. Posterior abdominal segment of
larva with papillae. Pupation taking place in larval
web. Pupa not actively mobile, but upon disturbance

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Baker

FIG. 13. Antennae of Arachnocampa spp. A, A. tasmaniensis antennal base indicating position at which
diameter of 1 st segment was measured; B, A. tasmaniensis male antennal apex. The apex of this species is
elongate; C, A. buffaloensis female antennal apex; D, A. richardsae female antennal apex.

is capable of small movements within remains of
larval snare.

Description. BIOLOGY: Larvae produce
bioluminescence from posterior end. Larvae
construct snares (webs) from silk and sticky
mucous droplets in which they lie within a
mucous tube. Snares consist of a number of silk
'fishing lines' with sticky mucous droplets for
prey capture.

Morphology. Larval antenna vestigial and
in the form of low hemispherical pale yellow
protuberance (Figs 5-6). Larva with rows of fine
setae at distal end (Fig. 7) and pair of large anal
papillae. Larval mandibles with four apical
teeth and one median tooth on inner surface
(Fig. 6). Adult antennae with fourteen flagellar
segments, longest segment at base, segments
progressively shortening to 13 th segment (Fig.

120

9); 14 th segment longer than 13 th , narrowing to
a short, rounded apex. Front dark brown. Ocelli
on prominent raised ridge facing anteriorly and
laterally (Fig. 10); strong dark setae covering
ridge. Face medium brown with thick cluster
of setae over facial carina. Eyes covered with
mat of fine, dark interommatidial hairs. Row
of setae along lateral margins of scutum. Leg
segments uniformly brown except lighter basal
section of coxa. Hind legs darker. Legs covered
in fine setae. Two apical tibial spurs on ventral
side of mid and hind leg. One apical tibial spur
on fore leg. Strong setae (quarter of length
of ventral spurs) along tibiae and tarsi and
clustered around dorsal apex of tibiae. Wing
veins brown. Darker on costal (C), subcostal
(Scl) and radial veins (R). Strong dark setae
on C, (setae on Scl in A. buffaloensis only. Fig.
11A) and R 1 and R 5 veins (Fig. 11B). Abdominal

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Taxonomy of Australian glow-worms

FIG. 14. Arachnocampa spp. larval sclerotised head plates indicating placement of campaniform sensillae
(with arrows). This arrangement of the campaniform sensillae differentiates the subgenus Campara (C-H)
from the subgenus Araclmocmnpa (A-B). A, A. tasmaniensis; B, A. buffaloensis; C, A. tropica ; D, A.flava ; E, A.
girraweenensis ; F, A. richardsae; G, A. gippslandensis; H, A. otwayensis.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Baker

segments 2-7 with tergites and stemites covered in
black hairs. Abdominal segment 1 with tergite
hairy, sternite with hairs on distal edge only. 2
with more setae on basal abdominal segments
than <J. Eggs clearly visible through adult
$ abdomen. All adult abdominal segments
yellow to dark brown, darker at distal end of
each segment, giving distinctive 'tiger stripe'
colouration (Fig. 8).

KEY TO THE SUBGENERA OF
ARACHNOCAMPA (MODIFIED
FROM EDWARDS 1924)

1 . Wing with m-cu cross vein basal to r-m cross
vein; fore basitarsus equal to, or up to 1.3
times length of fore tibia; pupae suspended
vertically; endemic to New Zealand

. . Araclmocampa ( Araclmocampa ) Edwards

2. Wing with m-cu cross vein basal to r-m cross
vein (Fig. 12A-B); fore basitarsus 1.55-2
times length of fore tibia; pupae suspended
horizontally; endemic to Tasmania
and Mt Buffalo, Victoria, Australia
Araclmocampa (Lucifera) subgen. nov.

3. Wing with m-cu cross vein distal to r-m

cross vein (Fig. 12C-H); basal segment
of fore tarsus 1. 3-1.5 times length of fore
tibia; pupae suspended horizontally; found
in eastern Australia Araclmocampa

(Campara) Harrison

REDESCRIPTION OF SUBGENUS
ARACHNOCAMPA BASED
ON EDWARDS (1924)

Diagnosis. Adult wing with m-cu basal to r-m
(Fig. 12A-B). bm-cu 0.72-1.6 times length of
m-cu (Fig. 12A-B). bm-cu straight (Fig. 12A-
B). Scape with full ring of setae around distal
end. Larval head capsule with campaniform
sensillae in block formation with setae (Fig.
14A-B). Larval malpighian tubules white (Fig.
15A-D). Pupae suspended vertically with one
silk line connecting to thoracic region.

122

ARACHNOCAMPA (LUCIFERA)
SUBGEN. NOV.

Araclmocampa (Lucifera) Baker et al., 2008 nomen nudum

Type species. Araclmocampa tasmaniensis Ferguson,
1925.

This study has supported Edward's (1924)
division of Araclmocampa into two subgenera,
Araclmocampa and Campara, based on the position
of the m-cu vein. However, I have identified
additional characters that further subdivide
Araclmocampa. Consequently, I propose a
new subgenus, Lucifera, for A. tasmaniensis
Ferguson and A. buffaloensis sp. nov. Lucifera
is distinguished from Campara by wing
venation and sensillum pattern. The subgenus
Araclmocampa is separated from Lucifera by the
unique method of vertical pupal suspension of
its now single constituent species, Araclmocampa
luminosa (Skuse).

Diagnosis. Adult wing with m-cu basal to r-m (Fig.
12A-B). bm-cu 0.72-1.6 times length of m-cu (Fig. 12A-
B). bm-cu straight (Fig. 12A-B). Scape with full ring
of setae around distal end. Larval head capsule with
campaniform sensillae in block formation with setae
(Fig. 14A-B). Larval malpighian tubules white (Fig.
15A-D). Pupae suspended horizontally with one silk
line connecting to thoracic region and another to the
abdominal region.

Etymology. Lucifera, literally meaning 'light
bringing', refers to the unique ability of Araclmocampa
to produce light. It also has reference to the enzyme
luciferase and substrate luciferin that produce the
light in conjunction with adenosine 5'-triphosphate
(ATP) and oxygen.

KEY TO SPECIES OF THE
SUBGENUS LUCIFERA

1. Fore basitarsus 1.75-2 times length of fore

tibia, endemic to Tasmania tasmaniensis

Ferguson

2. Fore basitarsus 1.55 times length of fore
tibia, endemic to Mt Buffalo, Victoria

buffaloensis sp. nov.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Taxonomy of Australian glow-worms

FIG. 15. Arachnocampa spp. larvae (live whole mounts). A, A. tasmaniensis (epigean); B, A. tasmaniensis (cave); C,
A. tasmaniensis (cave); D, A. buffaloensis ; E, A. tropica ; F, A.flaxm ; G, A. girrazveenensis (cave2); H, A. girraweenensis
(cavel); I, A. richardsae (cave);’J, A. ricliardsae (epigean); R, A. gippslandensis (cave); L, A. gippslandensis (cave),
mt, malphigian tubules.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Baker

REDESCRIPTION OF SUBGENUS
CAMPARA BASED ON EDWARDS (1924)

Diagnosis. Adult wing with m-cu distal to
r-m (Fig. 12C-H). bm-cu 1. 4-4.5 times length of
m-cu (Fig. 12C-H). bm-cu curved. Scape with
half ring of setae on ventral side of distal end.
Larval head capsule with campaniform sensillae
in line with setae rather than block formation
(Fig. 14C-H). Larval malpighian tubules white
(Fig. 15G, K) to light brown (Fig. 15H, I, J, L) to
black (Fig. 15E-F). Pupae suspended horizontally
with one silk line connecting to thoracic region
and another to abdominal region.

KEY TO SPECIES OF THE
SUBGENUS CAMPARA

To accurately distinguish species in this
subgenus, all life stages are needed. Genetic
data and mating trials clearly indicate them to
be distinct species. Large geographic boundaries
between species and no evidence of sympatric
speciation mean geographic information is often
the easiest way to distinguish these species.

1. 5 th instar larva with 10 maxillary teeth on
one side and 11 on the other (Fig. 16B, D) . . . 2

— 5 th instar larva with 9 maxillary teeth on
one side and 10 on the other (Fig. 16A, Q. . . 3

2. Larval malpighian tubules black (Fig. 15E).

Adult laterotergite with 2-30 setae (Fig. 17C);
bm-cu 1.7-2. 5 times length of m-cu (Fig.
12C). Found in wet tropical rainforests of
north Queensland tropica sp. nov.

— Larval malpighian tubules white to light

brown (Fig. 15G-H). Adult laterotergite
with 0-5 setae (17E); bm-cu 1.8-4. 5 times
length of m-cu (Fig. 12E). Found in caves
in Girraween National Park, south-east
Queensland and rainforest in northern
New South Wales (Washpool, Dorrigo
and New England National Parks)
girraweenensis sp. nov.

124

3. Larval malpighian tubules black (Fig. 15F).

Adult laterotergite with 1-4 setae; bm-cu 1.9-
2.7 times length of m-cu (Fig. 12D). Found
in isolated rainforest gullies in Kroombit
Tops National Park and rainforest of the
Mt Warning Caldera including northern
New South Wales .flava

— Not with above combination of characters
4

4. Larval malpighian tubules pale to medium

brown (Fig. 151, J). Adult laterotergite with
0-21 setae; bm-cu 1.5-3 times length of m-cu
(Fig. 12F). Found in rainforested regions of
central to southern New South Wales (Blue
Mountains, Bundanoon, Fiztroy Falls)
richardsae Harrison

— Not with above combination of characters.

Adult with never more than 3 small setae
on laterotergite. Found in rainforest areas
in southern Victoria 5

5. Adult body length 10-12.2 mm; wing length

6.5- 8 mm; 9 antennal length 3. 2-3. 3 mm;
bm-cu 1.75-2.6 times length of m-cu (Fig.
12G). Found in Gippsland rainforest
and isolated cave systems of Victoria
gippslandensis sp. nov.

— Adult body length 6.5-8 mm, wing length

4.5- 5.5 mm, 9 antennal length 2-2.2mm; bm-
cu 2.4-3.3 times length of m-cu (Fig. 12H).
Found in rainforest west of Melbourne
(including Otway National Park and Melba
Gully and Angahook-Lorne State Parks)
otwayensis sp. nov.

Arachnocampa ( Lucifera )
tasmattiensis Ferguson, 1925

Araclmocampa tasmaniensis Ferguson, 1925: 487; Harrison,
1966: 879 (redescription).

Material Examined. HOLOTYPE (in copula with
paratype ?), bearing the label: 'Ida Bay caves,
Arthur M. Lea, December, 1909. In total darkness
fully 1/4 mile from entrance', condition: poor. Card
mounted, both holotype and paratype r . missing
antennae, legs and wings glued to card (SAM).

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Taxonomy of Australian glow-worms

Other Material. TASMANIA, 3dd, 5??, Mole Creek
Karst National Park, Marakoopa cave, 41°35'S,
146°17'E, 1712001, CB/DM, collected as larvae, reared
to adults in incubator bv CB, TAS2#1,2,3,7,9,11,16;
299, Sassafras cave, 41°34'S, 146°21'E, l.x.2001, CB/
AS, collected as larvae, reared to adults in incubator
by CB, TAS3#3-4; 3(Jd, Francistown, property of
Arthur Clarke, rainforest gully along stream, 43°18.5'S,
146°59.4'E, 1612001, CB, collected as larvae, reared to
adults in incubator by CB, TAS5#l-2,5; 5Jd» 799, Ida
Bay caves. Mystery Creek cave (Entrance cave), 43°2 7.7'
S, 146°50.0'E, 161.2001, CB/DM, collected as larvae,
reared to adults in incubator bv CB, TAS6#l-2, 6-8, 12-
14, 17-18, 22-23.

The following redescription and measurements
are based primarily on J TAS6#22, with other
specimens measured for ranges and variation.

Diagnosis. Wing with rn-cu located basal to
r-m (Fig. 12A); bm-cu straight (not curved),
angled backwards towards wing base and 0.72-
1.6 times length of m-cu; bm-cu parallel to r-m
(Fig. 12A). Laterotergite with 0-10 setae. Fore
basitarsus 1.75-2 times length of fore tibia.
Larvae with rows of 9 and 10 maxillary teeth
plus lobe (Fig. 16A).

Redescription. Adult. $ holotype length 11.2
mm (range 10-13.5 mm (cave), 8-10.5 mm
(epigean)). Wing length J 6.5 mm (5.5-8 mm
(cave (?d)/ 5-6mm (epigean ^d), 6-9 mm (cave
9$), 5 mm (epigean ?$)). Antennal length 5mm
(d'd), 3mm ($?). Head: Antennae brown, bases
of first few flagellar segments lighter. Scape
and pedicel brown. Flagellum with dense mat
of brown hairs, except basal quarter of first
segment. Diameter of scape 2.4 times diameter
of basal flagellar segment (dd 2.3-2.9x, 99 2.6-
3x). First flagellar segment 1.2 times length
of second flagellar segment (JS l.l-1.5x, 9$
1.3-1. 6x). Palpi yellow/brown; apical segment
lengthened and slightly dilated (also darker);
light covering of dark setae on palps. Proboscis
yellow/brown (99) to pale yellow Thorax:
Cluster of long, dark setae on supra-alar region.
Shorter, finer setae on post alar area. Laterotergite

with 0/2 small setae (TAS2 1-10 setae, TAS5 1-4,
TAS6 0-3); sometimes with unequal numbers
of setae on each side. Scutum with medium
to dark brown shading. Pleura entirely dark
brown (encompassing laterotergite, anepimeron,
anepistemum, katepsisternum) (Fig. 17A). Legs:
Coxae white to pale yellow, darkening to dark
brown at apices, more extensively so on hind
coxa (Fig. 17A). Long dark hairs covering about
one third of each coxa, predominantly at apices
and on ventral surfaces. Mid basitarsus 1.1 times
length of mid tibia; hind basitarsus 0.71 times
length of hind tibia (99 0.77x). kVmgs: Entire
wing shaded except for fainter patch basally.
Campaniform sensilla located on d wings as
follows: 3 on Rj, 4 on Rs, and 1 on proximal end of
R 5 (Fig. 12A). TAS2 <?cJ (Rl; Rs: R5) 2-3: 4-6: 1-3,
99 2-3: 4-6: 2-4, TAS5 1-2: 2-4: 2-3, 99 2: 2: 2,
TAS6 <$<$ 1-4: 4-6: 1-2, 9 3/7: 5-6: 1-3). Halteres
pale yellow/ white basally, brown apically with
fine hairs. Abdomen : External genitalia light
brown. Eggs. Slightly ovoid. 0.55 x 0.5 mm,
red/brown. Larva. Larval body colouration
with sclerotised brown head capsule (TAS6-
cave), body with faint brown/red pigmentation
beginning at anterior end, extending down
75% of lateral edges of body (Fig. 15C) (range
TAS2-cave: pigmentation green, extending
down 60% of lateral edges (Fig.l5B); TAS5-
epigean: brown/black pigmentation extending
entire length of lateral edges (Fig. 15A). Larve
up to 4 cm long in fifth instar. Snares ~ 30 cm
long in caves, <3 cm in epigean settings. Pupa.
Suspended from distal tip of abdomen and
anterior part of abdomen.

Distribution. Arachnocampa tasmaniensis colonies
are relatively widespread throughout Tasmania
(Goede 1967; Eberhard 1999, 2000). Larvae are
primarily found in large numbers in limestone
caves. Populations are also widespread in rain-
forest and along creek banks and walking trail
cuttings (Baker 2004; Baker et al. 2008).

Memoirs of the Queensland Museum 1 Nature • 2010 • 55(1)

Baker

nn < Aradmocampci spp maxillary teeth. A. A. gippslandensis (10 maxillary teeth plus lobe); B, A. tropica
oirrmnpwn^ryn i^ 6th P?,Y S Iobe ) ; c < A. flava (9 and 10 teeth plus lobe (numbers 1-9! + lobe); D, A.
fhat on the oth ° ^ max illary teeth). In most species the number of teeth on one maxilla is different to

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Taxonomy of Australian glow-worms

FIG. 17. Arachnocampa spp. thoracic segments. A, A. tasmaniensis ; B, A. buffaloetisis; C, A. tropica ; D, A.flava ;
E, A. girraweenensis; F, A. richardsae ; G, A. gippslandensis; H, A. otwayensis. anepm, anepimeron; anepst,
anepisternum; kepst; ltgt, laterotergite; pal, post alar region; spal, supra-alar region.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Baker

Comments. Tasmanian populations from which
adults were successfully reared for morphological
examination were collected from four distinct
areas and habitats. Three populations were
sampled from limestone caves in both northern
(TAS2, 3) and southern (TAS6) areas of the
island, and one population was sampled from
rainforest in the south (TAS5). Other populations
identified during surveys did not produce
adults for morphological examination (TAS4),
or larvae were not collected from the site for
identification (TAS1).

Obvious morphological variation was evident
between cave and rainforest (epigean) popula-
tions (e.g. in larval colouration, adult body and
wing lengths), suggesting that environmental
differences experienced by cave and rainforest-
dwelling populations have led to phenotypic
changes within this species. The close proximity
of caves to rainforest populations enables genetic
exchange between these habitat types (Baker
2004, Baker et al. 2008). Some morphological
variation is evident between northern and
southern cave populations. For example, the
number of setae on the laterotergite varies from
0-3 for southern cave populations, and 1-10 for
northern cave populations. However, this may be
due to differing degrees of rainforest population
migration into caves or general plasticity within
this taxon rather than evidence of speciation.
Morphological characters used in this study
support Ferguson's (1925) description of
this species and its placement within the new
subgenus Lucifera.

Araclmocampa ( Lucifera )
buffaloensis sp. nov.

Araclmocampa buffaloensis Baker et al. 2008 (unavailable name).

Material. HOLOTYPE $, Australia, Victoria, Mount
Buffalo National Park, Underground River cave, glow-
worm grotto, 36°43'S, 151°17'E, 13.iii.2000, CB, collected
as adult, killed and stored in 70% ethanol, BUFF#2
(deposited in MV). PARATYPES. VICTORIA, 4 adult
; v ' Mount Buffalo National Park, Underground River

128

cave, glow-worm grotto, 36°43'S, 151°17'E, 131.2001,
CB/JF, collected as larva, reared to adult in incubator
by CB, BUFF#1,3,4,5. (paratypes deposited in MV
and QM)

The following description and measurements
are based primarily on the holotype, with other
specimens measured for ranges and variation.
No males were reared from larvae of this colony.

Diagnosis. Adult body length 10-12 mm. Wing
length 6.5-7 mm. m-cu close to r-m (Fig. 12B);
bm-cu running parallel to r-m and 1-1.6 times
length of m-cu (Fig. 12B). Fore basitarsus 1.55
times length of fore tibia. Laterotergite bare (1
seta on some paratypes). Larvae with row of 10
maxillary teeth plus lobe (Fig. 16A). Larval body
colouration: head capsule sclerotised brown,
body red, green, blue and white in order from
anterior to posterior end. Pigmentation limited
to anterior end (Fig. 15D).

Description. Adults ($$). Body length 11 mm;
wing length 7 mm; antennal length 2.8 mm.

Head. Antenna light brown, scape and pedicel
light brown, flagellum with dense mat of
brown hairs except basal sixth of first flagellar
segment. Diameter of scape 2.5 times diameter
of basal flagellar segment (range 2.3-2.5x). First
flagellar segment 1.4 times length of second
segment (1.3-1. 45x). Front brown, ridge at
dorsal corner of eye ending before centre of
front. Longitudinal, deep sulcus above base of
antennae. Palpi pale yellow. Apical segment
lengthened and slightly dilated (also slightly
darker). Proboscis pale yellow. Thorax: Large
cluster of long, dark setae on supra-alar region,
smaller much shorter hairs on post alar area
(Fig. 17B). Thoracic dorsum with medium
brown shading. Pleura darker brown, lighter in
crevice between laterotergite and anepimeron
(Fig. 17B). Legs: Coxae white to pale yellow.
Faint traces of darkening at apices, more
extensively so on hind coxa. Dark hairs covering
about one third of each coxa. Mid basitarsus
equal to length of mid tibia; hind basitarsus

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Taxonomy of Australian glow-worms

0.81 times length of hind tibia. Wing: Shading
covering entire wing, m-cu basal to r-m, but
closely approximated; r-m sloped forwards
towards apex of wing (Fig. 12B); brn-cu 1-1.6
times length of m-cu; m-cu sloping forwards
towards apex of wing. Halteres pale yellow
basally, brown apically. Campaniform sensillae
located on R veins as follows: 1 on 6 on Rs,
and 2 on R 5 (1-3: 6-7: 1-3) (Fig. 12B). Abdomen:
External genitalia light brown. All abdominal
segments dark brown (paratypes: light-medium
brown; darker at distal end of each abdominal
segment). Eggs. Not seen. Larva. Up to 4 cm in
length. Larval snares up to 40 cm in length.
Pupa. Suspended from apex of abdomen and
anterior part of abdomen. Bracing thread from
thorax to main anterior abdominal suspension
thread. Head area black, thorax red/brown,
abdomen pale yellow (other Mt Buffalo pupae
suspended as for A.flaim).

Etymology. The specific name refers to the type
locality, a single cave in Mt Buffalo National Park,
Victoria.

Distribution. At present this species is known
only from the Underground River Cave (granite
boulder infill cave, 300 m in length with constant
water flow) in Mt Buffalo National Park, Victoria.
Other caves on private land within the region
have been reported to contain glow-worms.
However, as yet entry to these other sites has
not been granted to the author.

Comments. Data from molecular and morph-
ological studies support A. buffaloensis as a
sister species to the geographically isolated
Tasmanian species A. tasmaniensis (Baker 2004;
Baker et al. 2008). Theories explaining the current
geographic isolation of these sister species are
presented in Baker et al. (2008). The cave in
which A. buffaloensis is found is sub-alpine, where
snowfall and subsequent spring melt flooding are
an annual event. "The extreme seasonal variations
and low humidity outside the cave habitat greatly
decrease the chances of survival for colonies

not within the safety of the cave environment.
This species was nominated by the author
and has since been listed as threatened under
the Victorian Government's Flora and Fauna
Guarantee Act 1988 on the basis of its extremely
restricted habitat of one cave.

Arachnocampa (Campara) tropica sp. nov.

Arachnocampa tropica Baker et al. 2008 (unavailable name)
Arachnocampa tropicus Baker et al. 2008 (unavailable name)

Material. HOLOTYPE <T, Australia, north Queensland,
Mt Lewis Rd, Rainforest road cuttings near stream,
16°34.9’S, 145°17'E, 10.vii.2000, CB/AM, collected
as larva, reared to adult in incubator by CB,
NQLD2#2 (deposited in QM T152171). PARATYPES.
QUEENSLAND, same data as holotype. NQLD2#6;
2?$, same data as holotype NQLD2#1, 5; 2-LS, 52$,
Daintree National Park, Mossman Gorge section,
council road off Mossman Gorge carpark, road cuttings,
16°25'S, 145°20'E, 14.vii.2000, CB/AM, collected as
larvae, reared to adults in incubator by CB, NQLD1 #1-6,
8; 1$, Paluma National Park, Birthday Creek Falls, road
cuttings near car park, 18°59'S, 146°10'E, 8.vii.2000,
CB/AM, collected as larva, reared to adult in incubator
by CB, NQLD9#1; IT, Wooroonoran National Park:
Josephine Falls section, Bartle Frere trail, western
approach, Bobbin Bobbin Falls, stream banks, 17°22'S,
145°47E, 9.vii.2000, CB/ AM, collected as larva, reared to
adult in incubator by CB, NQLD6#1; lv, Wooroonoran
National Park: Josephine Falls section, Bartle Frere trail,
western approach, granite boulder cave off trail
near summit, 17°22'S, 145°49'E, 9.vii.2000, CB/AM,
collected as larva, reared to adult in incubator by
CB, NQLD5#3; 5oT, 1$, Mt Hypipamee National
Park, Dinner Falls walking trail banks near stream,
17°26'S, 145°28'E, 9.vii.2000, CB/AM, collected as
larva, reared to adult in incubator by CB, NQLD4#1-
6; 1$, Mt Hypipamee National Park, Dinner Falls
walking trail banks near stream, 17°26'S, 145°28'E,
11. vi. 2002, CB/AM, collected as larva, reared to
adult in incubator by CB, NQLD4Ri#l; 2TT, 2$$,
Tully Falls State Park, Charmillan walking trail banks
near stream, 17°43'S, 145°31'E, 9.vii.2000, CB/AM,
collected as larva, reared to adult in incubator by CB,
NQLD8#1-4. (paratypes deposited in AM, MV, QM,
TM and UQ1C).

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Baker

The following description and measurements
are based primarily on the holotype, with other
specimens measured for ranges and variation.

Diagnosis. Adult body length 7-8 mm (S3),
7-9 mm (?$). Wing length 5-6 mm 5.5-

6.5 mm (??). Antennal length 3-3.5 mm (c?c?)/
2mm ($?). bm-cu curved backwards and 1.7-

2.5 x length of m-cu. m-cu sloping forwards (Fig.
12C). Laterotergite with 2-12 (<JcJ), 3-30 (;?) setae.
Larvae with 10 and 11 maxillary teeth (Fig. 16B).
Larval head capsule sclerotised brown, tx)dy with
faint brown/red pigmentation beginning at
anterior end, extending down entire length of
lateral edges of body. Malpighian tubules black
(Fig. 15E).

Description. Adult. $ holotype body length 8
mm; wing length 5 mm; antennal length 3.2 mm.

Head. Antennae yellow to brown, lighter at basal
end of first flagellar segment; scape and pedicel
yellow. Flagellum with dense mat of brown
hairs, except basal quarter of first flagellar
segment. Diameter of scape 2.25 times diameter
of basal flagellar segment in holotype <$ ( jj
2.25-2.7x, : 2 2.6-3.4x). First flagellar segment 0.88
times length of second flagellar segment (o c? 0.88-
1.4x, $2 0.93-1. 2x). Front medium to dark brown.
Dark setae on head. Face medium brown to
dark brown with thick cluster of setae over
facial carina. Palpi pale yellow; apical segment
lengthened and slightly dilated; light covering
of dark setae on palps. Proboscis pale yellow
to brown. Thorax: Cluster of long, dark setae
on supra-alar region and on post alar area (Fig.
17C). Laterotergite with 6 small setae (cJo 2-12,
?? 2-30); sometimes with unequal numbers of
setae on each side. All thoracic dorsum with
medium brown shading. Pleura (encompassing
laterotergite, anepisternum, anepimeron and
katepisternunr) slightly darker medium brown
and more heavily shaded in centre of sclerites
(Fig. 17C). Legs: Coxae pale yellow, darkening to
medium/ dark brown at apices, more extensively
so on hind coxa. Long, dark hairs covering

130

about one quarter of each coxa, predominantly
at apices and ventral surfaces. Other segments
of legs yellow/brown. Fore basitarsus 1.5 tim^s
length of fore tibia (1.2-1.5x); Mid basitarsus
equal to length of mid tibia (22 0.83x); hind
basitarsus 0.83 times length of hind tibia (2$
0.66-0.83x). Wings: Shading covering entire
wing except for a fainter patch at wing base.
Halteres pale yellow basally, brown apically
with fine hairs. Campaniform sensilla located
on cJd wings as follows: 2 on Rj, 4 on Rs, and 3
on R s (Fig. 12C) (£3 range 1-3: 3-6: 1-3, $$ 1-3:
3-5: 1-4). Abdomen: External genitalia medium
brown. Eggs. 0.48 x 0.43 mm, yellow/brown
with slightly red tinge (virgin eggs). LarVa.
Intestines green. Larval snares <3 cm. Pupa.
Suspended from apex of abdomen and anterior
part of abdomen.

Etymology. The specific name tropica refers to the
type locality's distribution within the world heritage-
listed Wet Tropics region of north Queensland,
Australia.

Distribution. This species has been recorded
in rainforest and small granite boulder caves
throughout the Wet Tropics rainforests of far
north Queensland. It is likely that this species
is more widespread in the rainforests of this
region, however a detailed survey is yet to be
conducted. Populations have been observed
to expand rapidly during the northern wet
season (November - March), only to drop back in
numbers through the remainder of the year due
to drier conditions (personal observations).

Comments. Morphological and molecular data
distinguish A. tropica from all other species in the
genus. A large geographic distance separates
this northern clade from other Australian
populations of Arachnocampa. Genetic data
suggest two species may be present in far north
Queensland following two phases of expansion
into the tropics (Baker 2004, Baker et al. 2008).
Verification of this distinction requires genetic
analysis at the population level. Individual

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Taxonomy of Australian glow-worms

colonies of this species were usually low in
numbers of individuals and density, with no
spectacular displays noted. Larvae collected from
the Wet Tropics were heavily parasitised by an
undescribed species of ichneumonid wasp (Baker
2004; Chris Burwell, pers. comm.). This wasp has
not been recorded from colonies further south
but could represent a significant threat to the
tourism industry associated with glow-worms
in southeast Queensland if it dispersed or was
translocated to the region.

Arachnocampa ( Campara )
flava Harrison, 1966
Arachnocampa flava Harrison 1966: 882.

Material Examined. HOLOTYPE $ (QM Reg. No.
T6430) Queensland, Numinbah 21.iv.1935 F.A. Perkins
(QM). OTHER MATERIAL. NEW SOUTH WALES,
3do, 4?$, Mt Warning National Park, Korrumbyn
picnic area, vertical creek banks, 28°24'S, 153°16'E,
27.iii.2000, CB/JF, collected as larvae, reared to adults
in incubator by CB, CALD5#l-7; 2$?, Mt Warning
National Park, Korrumbyn picnic area, vertical creek
banks, 28°24'S, 153°16'E, 5.vii.2002, AR/CB, collected
as larvae, reared to adults in incubator by CB,
CALD5#Ri1-2; 4 4.$, Nightcap National Park,
Protesters Falls walk, vertical creek banks, 28°34'S,
153 1 7 E, 27.iii.2000, CB/JF, collected as larvae, reared to
adults in incubator by CB, CALD6# 1-8. QUEENSLAND:
6>a, 10, , Springbrook-Mudgeeraba Road, Spring-
brook Gully, vertical creek banks, 28°12.5'S, 153°1 7.5'E,
23. v. 2001, CB, collected as larvae, reared to adults in
incubator by CB, CALD3#4, 12, 14-17, 21, 23, 31, 33, 34,
39, 41-43, Rix.

Diagnosis. Adult body length 6.5-8.5 mm (3o),
7-10 mm (99); wing length 4.5-5 mm 4.6-
6.5 mm ( 99 ); antennal length 3.5-4 mm (tfrf), 2
mm ( 99 ). bm-cu curved backwards and 1. 9-2.7
times length of m-cu (Fig. 12D). Laterotergite
with 0-7 setae. Larval colouration: head capsule
sclerotised brown, body with dark brown/red
pigmentation extending along entire lateral
edges of body (Fig. 15F). Intestines green.
Malpighian tubules black. Larval maxillae with
rows of 9 and 10 teeth (Fig. 16C).

Description. Adult. $ holotype body length 8
mm; wing length 5mm; antennal length 4 mm.

Head. Antennae light to medium brown, light
yellow at basal end of first few flagellar segments.
Scape and pedicel pale yellow. Flagellum
with dense mat of brown hairs, except basal
fifth of first segment. Diameter of scape 2.7
times diameter of basal flagellar segment in
holotype $ (JJ 2.6-3.1x) and 3-3.6 times length
of second flagellar segment in 9?- Scape with
very sparse setae around ventral apex. First
flagellar segment 1.3 times length of second
flagellar segment (<J<J and 99 1.1-1 ,3x). Front
medium to dark brown. Strong dark setae
covering ridge. Face medium brown with
thick cluster of setae over facial carina. Palpi
pale yellow; apical segment lengthened and
slightly dilated; light covering of dark setae
on palps. Proboscis yellow/brown ( 99 ) to pale
yellow (jo). Thorax : Cluster of long, dark setae
on supra-alar region. Shorter, finer setae on
post alar area. Laterotergite with 2/1 small
setae (0-7); sometimes with unequal numbers
of setae on each side. Scutum with medium
yellow/brown shading. Pleura (encompassing
laterotergite, anepisternum, anepimeron and
katepistemum) medium to dark brown, crevices
between sclerites slightly lighter. Legs: Coxae
white to pale yellow, darkening to dark brown
at apices, more extensively so on hind coxa.
Dark hairs covering about one quarter of each
coxa, predominantly at apices and on ventral
surfaces. Other segments of legs yellow to brown.
Fore basitarsus 1.5 times length of fore tibia
(99 l.lx); mid basitarsus 1.1 times length of
mid tibia (99 0.77x); hind basitarsus 0.71 times
length of hind tibia (99 0.77x). Wings : Shading
covering entire wing with only a fainter basal
patch. Halteres pale yellow basally, brown
apically with fine hairs. Campaniform sensillae
located on 9$ wings as follows: 2 on Rj, 2-4
on Rs, and 1-3 on R g (beyond r-m) (99 1-3: 2-5:
1-3). Abdomen: External genitalia yellow/brown.
Eggs. 0.42 x 0.36 mm, orange/cream coloured

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Baker

when deposited; darkening over time to dark
red/ brown. Larva. 3-4 cm in length in 5 th instar.
Larval snares <5 cm. Pupa. Suspended from
apex of abdomen and anterior part of abdomen.
Abdomen orange.

Distribution. Populations are found in rain-
forest gullies and along rainforest stream banks
throughout the Mt Warning caldera. Colonies
reach their highest numbers in one congregation
at a large overhang. Natural Bridge, Springbrook
National Park. Other notable colonies are in the
Sprinbrook Plateau, the Border Ranges, Main
Range and Tamborine Mountain (David Newell,
herpetologist, Southern Cross University, Conrad
Hoskin, herpetologist, Australian National
University, pers. comm.). These additional pop-
ulations were not surveyed for this study.

Comments. Elevated numbers of A.flava within
rainforest colonies occur during periods of warm,
wet weather (Baker 2002). The distribution of A.
flava now includes colonies from Mt Warning
(CALD5) and Nightcap National Park (CALD6).
Colonies at Tamborine Mountain (TAM1),
O'Reilly's, Lamington National Park (CALD4),
and Kroombit Tops (samples sent from Harry
Hines, Queensland National Parks) are tentatively
placed in this species based on molecular evidence
from two mitochondrial genes (Baker et al. 2008).
The author would like to conduct a morphological
examination of adults from the Kroombit Tops
population to ascertain their correct placement,
as a large barrier of unsuitable habitat exists
between these populations. However as yet,
only preserved larvae have been obtained.

Arachnocampa ( Campara )
girraiveenensis sp. nov.

Arachnocampa girraiveenensis Baker et al. 2008 (unavailable
name).

Material. HOLOTYPE <J, Australia, Queensland,
Girraween National Park, South Bald Rock cave,
28°54’S, 152°01'E, 24.iii.2001, ASH/CB/DM, coll-
ected as larva, reared to adult in incubator by CB,
G1RRA2#12 (deposited in QM T152149). PARATYPES.

QUEENSLAND, ?, Girraween National Park, South
Bald Rock cave, 28°54'S, 152°01'E, 6.vii.2002, AR/
CB, collected as larva, reared to adult in incubator
bv CB, GIRRA2Ri#A; 233, 52?, Girraween National
Park, South Bald Rock cave, 28°54'S, 152°0l'E

24.111.2001, ASH/CB/DM, collected as larva e '
reared to adults in incubator by CB, G1RRA2#1, 4-(s,

8 10-11; 233, 1?, Girraween National Park, Ramsey
Creek, Underground River cave, 28°51'38.9's
151°55'47.3E, 6.vii.2002, AR/CB, collected as lar V a e "
reared to adults in incubator by CB, GlRRAlRi#i- 3 ;
lcJ, 3?$, Girraween National Park, Ramsey Creek,
Underground River cave, 28°51'38.9'S, 151°55'47.3E,

23.111.2001, ASH/CB/DM, collected as larvae, reared
to adults in incubator by CB, GlRRAlRi#4,8,9,li .
NEW SOUTH WALES, 1$, Bellingen Shire, Cleaver's
Bridge, 30°27'S, 152°54'E, 25.iii.2000, CB/JF, collected
as larva, reared to adult in incubator by CB, NNSW2#1 ;
3J j, 1?, Barrington Tops National Park, Gloucester
Tops, Sharpe's Creek Walk, trail banks near creek,
32°03'S, 151°40'E, 24.iii.2000, CB/JF, collected as
larva, reared to adult in incubator by CB, NNSW4#1,
3-5; 333/ 2;V, Washpool National Park, Washpool walk,
trail banks near stream, 29°16'S, 152°22'E, 26.iii.2000,
CB/JF, collected as larvae, reared to adults in incubator
by CB, NNSWl#l-5. (paratypes deposited in AM, MV,
QM, TM and UQIC).

The following description and measurements are
based primarily on the holotype, with ranges taken
from other specimens.

Diagnosis. Adult body length 8.5-11 mm (cave
33 )> 7-8.5 mm (epigean 33 )/ 8-11 mm (cave
99), 7-9 mm (epigean 2$); wing length 5-6.2 mm
(cave 33), 6-7.2 mm (cave 99), 5-6 mm (epigean
33 and 99); antennae 4-5 mm (cave cJcJ), 2-3
mm (cave 99), 3-3.5 mm (epigean 33 ), 2 mm
(epigean 29). Laterotergite with 0-5 setae, bm-
cu curved and angled backwards, 2-4.5 times
length of m-cu (cave) (1.8-3.2x epigean) (Fig.
12E). Larvae with 10 and 11 maxillary teeth (Fig.
16D). Larval head capsule sclerotised brown,
body with brown/red pigmentation beginning
at anterior end, extending down 30% of lateral
edges of body (Fig. 15G, H). Malpighian tubules
white to pale brown. Larval snares <5 cm.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Taxonomy of Australian glow-worms

Description. Adult. 3 holotype (cave) body length
11 mm; wing length cave 6.2 mm; antennal length
5 mm. Head: Antennae medium brown, light
yellow at bases of first few flagellar segments.
Scape and pedicel pale yellow. Flagellum with
dense mat of brown hairs, except basal sixth of
first segment. Diameter of scape 2.6 (cave 3 3
range 2.3-2.6, epigean 33 2.2-2.8) times diameter
of basal flagellar segment, in 33 and 2.4-3 .3 times
length of second flagellar segment in cave 2?
(epigean 22 3-3.7x). First flagellar segment 1.4
times length of second flagellar segment (range
l.l-1.4x). Front dark brown. Face medium to
dark brown with thick cluster of setae over
facial carina. Palpi yellow/ brown; apical
segment lengthened and slightly dilated (also
darker); light covering of dark setae on palps.
Proboscis yellow. Thorax: Cluster of long, dark
setae on supra-alar region. Shorter, finer setae
on post alar area. Laterotergite with 3/3 dark
setae (range cave 33 3, epigean 33 0-3, cave 2?

1- 5, epigean $2 0-4); sometimes with unequal
numbers of setae on each side. Row of small
setae along lateral margins of scutum. Scutum
with medium brown shading. Pleura light to
medium brown (encompassing laterotergite,
anepisternum, anepimeron and katepisternum);
anepimeron lighter (Fig. 17E). Legs: Coxae pale
yellow, darkening to light to medium brown
at apices, more extensively so on hind coxa.
Long dark hairs covering about one third of
each coxa, predominantly at apices and ventral
surfaces. Fore basitarsus 1 .4 times length of fore
tibia (?$ 1.4x); mid tarsus 1.1 times length of
mid tibia (22 0.83x); hind basitarsus 0.71 times
length of hind tibia (2$ 0.65x). Wings: Halteres
pale yellow/ white basally, yellow/brown
apically with fine hairs. Campaniform sensillae
located on 3 wings as follows: 3 on R t , 4 on Rs,
and 3 on R 5 (Fig. 12E) (33 cave range 2-4: 3-4:

2- 3, 33 epigean 2-3: 2-3: 1-3). Abdomen: External
genitalia medium brown. Eggs. Slightly ovoid.
0.56 mm x 0.48 mm, yellow/brown (virgin $2
eggs). Larval snares <4 cm. Pupa. Suspended
from apex of abdomen and anterior part of

abdomen. Abdomen yellow/brown with eggs
clearly visible.

Etymology. The specific name refers to the type
locality of Girraween National Park, Queensland, in
which the two isolated cave populations are located.

Distribution. This species is restricted to two
continuously wet granite boulder caves in
Girraween National Park, Queensland, extending
into nearby rainforest within and around Wash-
pool National Park (NNSW1), New England
National Park (NNSW2), Dorrigo National
Park (NNSW3), Barrington Tops National Park
(NNSW4) and one small cave in Gloucester,
New South Wales.

Comments. Morphological differences are evi-
dent between the cave and rainforest (epigean)
colonies described here (including body size,
colour and the number of setae on the later-
otergite). However, these differences are likely
to result from the environmental conditions
in the habitat in which they live rather than
reflect species differences. Larval snares are
short despite being in a cave environment.
Molecular data support A. girratveenensis as
the sister species to A.flava, and link the cave
populations of A. girraweenensis in Queensland
to those in nearby rainforest, now separated by
unsuitable habitat (Baker et al. 2008).

Arachnocampa (Campara)
richardsae Harrison, 1966

Arachnocampa richardsae Harrison, 1966: 881.

Material Examined. HOLOTYPE 3 bearing the label:
Australia, Newnes Railway Tunnel, New South Wales.
Collected as pupa 15.x. 1961, adult 20.x. 1961, coll.
A.M. Richards (ANIC).

Other Material. NEW SOUTH WALES, 3 33 , 4??,
Wollemi National Park, Newnes Railway Tunnel,
33°11'S, 150°14'E, 14.vii.2000, CB/JF, collected as
larva ( 1 3 and IS collected as adults), reared to adult
in incubator by CB, SNSWl#l-4, 6-8; 2^ . , Wollemi
National Park, Newnes Railway Tunnel, 33°11'S,
150°14'E, 28. vi. 2002, CB/AR, collected as larvae.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Baker

reared to adults in incubator by CB, SNSWlRi#l,
4; 3cJcJ, Blue Mountains National Park, Grand
Canyon Walk, rainforest canyon near stream, 33°
39'S, 150°19'E, 23.iii.2000, CB/JF, collected as larvae,
reared to adults in incubator by CB, SNSW4#l-3; 15 ,
Blue Mountains National Park, Grand Canyon Walk,
rainforest canyon near stream, 33°39'S, 150°19'E,
28.vi.2002, CB/ AR, collected as larva, reared to adult
in incubator bv CB, SNSW4Ri#4; 2J $, 452 , Morton
National Park, Glow-worm Grotto, overhang over
waterfall in rainforest, 34°39’S, 150°29'E, 23.iii.2000,
CB/JF, collected as larvae, reared to adults in incubator
by CB, SNSW6#l-6.

The following redescription and measure-
ments are based primarily on a male, SNSW1#1,
from Wollemi National Park, Newnes Railway
Tunnel, with other specimens measured for
ranges and variation.

Diagnosis. Body length 10-11 mm (S3 cave),
7-7.5 mm (S3 epigean), 9-11.5 mm ($2 cave), 9
mm (22 epigean); wing length 6.5-7.5 mm (S3
cave), 5.2-6 mm (epigean 33), 6-8.5 mm (22 cave),
5-6 mm ( 2 $ epigean); 22 cave antennae 3 mm,
epigean 22 2-2.2 mm. m-cu well distal to r-m.
r-m cross vein straight, bm-cu curved and 1.5-3
times (SNSW1 1.8-2.8x, SNSW6 2.3x, SNSW4 1.5-
3x) length of m-cu, sloping backwards, towards
wing base (Fig. 12F). m-cu sloping forwards.
Laterotergite with 0-21 setae. Larval malpighian
tubules light to medium brown (Fig. 151, J). Larval
maxillary teeth 9 and 10 plus lobe (cf Fig. 16A,
C). Larval colouration: head capsule sclerotised
brown, body with green pigmentation extending
along entire length of lateral edges of body. Some
brown/ red pigmentation at posterior end (Fig.
151, J).

Redescription. Adults. 3 holotype body length
11 mm; wing length 7.5 mm; antennal length 5
mm. Head: Antennae brown, lighter at bases of
first few flagellar segments. Scape and pedicel
yellow to light brown. Flagellum with dense
mat of brown hairs, except basal sixth of first
segment. Diameter of scape 2.4 times diameter
of basal flagellar segment in 3 (2.2-3.4x). First

flagellar segment 1.04 times length of second
flagellar segment (1.04-1.4x). Front dark brown.
Dark setae behind ridge. Face medium to dark
brown with cluster of setae over facial carina.
Palpi yellow to light brown; apical segment
lengthened and slightly dilated; light covering
of dark setae on palps. Proboscis yellow/brown
( 22 ) to pale yellow (So)- Thorax: Cluster of very
long, dark setae on supra-alar region and post
alar area (Fig. 17F). Laterotergite with 20/21
small setae on 3 (33 6-21, 8-17 on pi) (SNSYV4
33 1-7, SNSW4/SNSW6 22 0-3); sometimes
with unequal numbers of setae on each side.
Scutum with medium brown shading. Pleura
medium brown (encompassing laterotergite,
anepisternum, anepimeron and katepistemum).
Legs: Coxae pale yellow, darkening to medium
brown at apices, slightly more extensively so on
hind coxa. Very long dark hairs covering about
one third of each coxa, predominantly at apices
and ventral surfaces. Fore basitarsus 1 .3 times
length of fore tibia (22 1.2x); Mid basitarsus 1.1
times length of mid tibia; hind basitarsus 0.66
times length of hind tibia ( ; . 0.65x). Wings:
Shading covering entire wing, only with a
fainter patch at base. Halteres pale yellow
basally ( 22 ), white (33), brown apically with
fine hairs. Campaniform sensillae located on
3 wing as follows: 4 on R v 4 on Rs, and 4 on R-
(Fig. 12F) (SNSW1 So 3-4: 3-5: 4-6, SNSW1 22
3-4: 4-5: 4/7, SNSW4 So 2-3: 2-4: 3-5, SNSW6 22
2:3:2, SS 2: 3-4: 2). Abdomen: External genitalia
medium to dark brown. Eggs. 0.55 x 0.48 mm,
cream brown - red/brown (virgin 22 e ggs).
Larva. 5 th instar 3-4 cm in length. Larval snares
<5cm. Pupa. Suspended from apex of abdomen
and anterior part of abdomen.

Distribution. This species is found in fragmented
rainforest areas from Gosford (SNSW2), the Blue
Mountains (SNSW1, 3, 4) to Bundanoon (SNSW5)
and Fitzroy Falls (SNSW6), New South Wales.

Comments. Initially described by Harrison (1966),
the largest known colony is found in the man-

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Taxonomy of Australian glow-worms

made Newnes railway tunnel. This is probably
due to the water seepage within the tunnel, the
abundance of prey items and protection from fire.
Morphologically, cave and epigean populations
exhibit differences in body size and colouration.
Again these differences are likely to be due
to morphological plasticity rather than reflect
species differences. Harrison's use of colour for
distinguishing A. flava from A. richardsae can
be problematic given the large colour variation
within these species. The darkened malpighian
tubules in the larvae of A. flava is a more reliable
character to separate the species.

Aracltnocampa (Cam para)
gippslandensis sp. nov.

Araclmocampn gippstaiideusis Baker et at. 2008 (unavailable

name).

Material Examined. HOLOTYPE ..Australia, Victoria,
Yarra Valley, Old Warburton Road, Shining Star mine
tunnel, 37°46'S, 145°38'E, 12.i.2001, CB, collected as
larva, reared to adult in incubator by CB, EVIC4#3
(deposited in MV). PARA TYPES: VICTORIA, 2yi,
Yarra Valley, Old Warburton Road, Shining Star mine
tunnel, 37°46'S, 145°38'E, 16.iii.2000, CB/JF, collected
as larva, reared to adult in incubator by CB, EV1C4#1-
2; 1$, Yarra Valley, Old Warburton Road, Shining Star
mine tunnel, 37°46'S, 145°38'E, 12.L20O1, CB, collected as
larva, reared to adult in incubator by CB, EV1C4#4; Id,

I . , Yarra State Forest, Britannia Creek Rd, Britannia
Creek cave, 37°48'S, 145°40'E, 14.vi.2002, CB/AM,
collected as larvae, reared to adults in incubator by
CB, EVlC3#3-4; 1 ., Yarra Ranges National Park,
Peninsula Road, Goldfields Walk, Upper Yarra
goldmine tunnel, 37°39'S, 145°53'E, 14.vi.2002, CB/
AM, collected as larvae, reared to adults in incubator
by CB, EVIC1#2; 3dd, 1 , Walhalla, Long Tunnel
Extended, horizontal mine shaft, 37°57'S, 146°27'E,
14. vi. 2002, CB/AM, collected as larvae, reared to
adults in incubator by CB, EVIC7Ri #1-4. (paratypes
deposited in AM, MV, QM, TM and UQIC).

The following description and measurements
are based primarily on the holotype, with other
specimens measured for ranges and variation.

Diagnosis. Adult body size 10-12 mm (Jd),

II. 5-12.2 mm (99); wing length 6. 5-7.5 mm

{<$<$)' 7-8 mm (99); antennal length 4.5-5 mm
3.2-3.3 mm ($9)- 0-1 setae on laterotergite.
m-cu well distal to r-m. r-m at 90 degrees to R
(EVIC3) or anteriorly sloped slightly backwards
towards the base of wing (EV1C1, 4, 7) (Fig. 12G).
bm-cu 1.75-2.6 times length of m-cu, curved,
and angled backwards towards base of wing
(EVIC1, 3, 7) (EVIC4 bm-cu 1.3-1 .4 times length
of m-cu, straight, angled forwards towards apex
of wing). Larval maxillae with rows of 9 and
10 teeth (Fig. 16A). Larval colouration: Head
capsule sclerotised brown, body with faint
brown/ red pigmentation beginning at anterior
end, extending down 25% of lateral edges of
body (Fig. 15K, L). Intestines brown to green.
Malpighian tubules white to pale brown (Fig.
15K, L).

Description. Adult. 3 holotype body length 11.5
mm; wing length 7 mm; antennal length 5 mm.
Head: Antennae brown, lighter at bases of first
few flagellar segments. Scape and pedicel light
yellow to brown. Flagellum with dense mat of
brown hairs, except basal fifth of first segment.
Diameter of scape 2.8 times diameter of basal
flagellar segment in J (j'J 2-2.9x, 9$ 2.7-3.1x).
First flagellar segment 1.5 times length of second
flagellar segment ($<$ 1.3-1. 5x, $9 1.09-1.4x).
Front dark brown. Face medium brown with
thick cluster of setae over facial carina. Palpi
yellow/ brown; apical segment lengthened and
slightly dilated (also darker); light covering of
dark setae on palps. Proboscis yellow/ brown.
Thorax: Cluster of long, dark setae on supra-
alar region (Fig. 17G). Shorter, finer setae on
post alar area. Scutum medium brown shading.
Pleura sclerotised dark brown (Fig. 17G)
(encompassing laterotergite, anepisternum,
anepimeron and katepisternum). Legs: Coxae
yellow, darkening to medium brown at apices,
more extensively so on hind coxa. Long dark
hairs covering about one third of each coxa,
predominantly at apices and on ventral surfaces.
Other segments of legs brown. Fore basitarsus
1.4 times length of fore tibia (cJcJ and 9$); mid

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Baker

basitarsus 1.1 times length of mid tibia (So and
??): hind basitarsus 0.71 times length of hind
tibia (?$ 0.66x). Wings: Halteres pale yellow/
white basally, brown apically with fine hairs,
m-cu distal to r-m but position variable: EVIC1
with m-cu closer to r-m, EVIC3 and EVIC7 with
m-cu closer to M vein split, EVIC4 with m-cu
in middle. Campaniform sensillae located on 3
holotype wing as follows: 4 on Rl, 5 on Rs, and
5 on R5 (Fig. 12G) (SS 3-4: 4-5: 3-5, 99 3-4: 3-5:
4-5). Abdomen: External genitalia light brown
with black tips. Eggs. Slightly ovoid. 0.5 mm x
0.6 mm, brown (virgin 29 eggs), larva. Larval
snares <3 cm. Pupa. Suspended from apex
of abdomen and anterior part of abdomen.
Threads turn golden after pupation (although
may be an artefact of the environment).

Etymology. The specific name refers to the type
locality of Gippsland, Victoria, Australia.

Distribution. Arachnocampa gippslandensis colonies
are found in rainforest and man-made mine
tunnels in the Gippsland region of Victoria.

Comments. There is strong morphological and
molecular support for the separation of the
eastern Victorian Arachnocampa populations as
a distinct species (Baker et al. 2008). Specimens
from this species were first noted in the literature
by Harrison (1966) from samples collected at
Walhalla by Elery Hamilton-Smith in 1965. At that
time, this population was tentatively grouped
with A. richardsae, but without morphological
examination. Recent fires (February 2009) through
this region may have greatly decreased the
available habitat of this species and, as such,
field surveys are recommended.

Arachnocampa (Campara)
otwayensis sp. nov.

Arachnocampa otwayensis Baker et al., 2008 (unavailable name).

Material Examined. HOLOTYPE _J, Australia, Victoria,
Melba Gully State Park, Madsen Track, stream and
trail banks, 38°41.8'S, 143°22.2'E, 10.vi.2002, CB/
AM, collected as larva, reared to adult in incubator

136

by CB, WVICIRil (deposited in MV). PARATYPES.
VICTORIA, Id, Hopetoun Falls, stream banks,
38°40'S, 143°34'E, 9.iii.2000, CB/JF, collected as larva,
reared to adult in incubator by CB, WVIC3#1; 1 J, 19,
Angahook-Lorne State Park, Kennett River picnic
area, stream banks, 38°40'S, 143°49'E, 9.iii.2000, CB/
JF, collected as larvae, reared to adults in incubator
by CB, VVVIC5#2-3; 1 j, 1 2, Angahook-Lorne State
Park, Sheoak picnic area, stream banks, 38°33'S,
143°56'E 12.iii.2000, CB/JF, collected as larvae,
reared to adults in incubator by CB, WVIC6#l-2;
IS, Angahook-Lorne State Park, Sheoak picnic area,
stream banks, 38°33'S, 143°56'E, 10.vi.2002, CB/ AM,
collected as larvae, reared to adults in incubator by
CB, WVIC6Ri#l ; 2 SS, 1 - Beauchamp Falls, base of
waterfall and surrounding stream banks, collected as
larvae, reared to adults in incubator by CB, 38°39'S,
143°37'E, 12.iii.2000, CB/JF, WVIC2#l-3; Id, 12,
Beauchamp Falls, base of waterfall and surrounding
stream banks, SS^S, 143°37'E, 10.vi.2002, CB/ AM,
collected as larvae, reared to adults in incubator by
CB, WVIC2Ri#l-2. (paratypes deposited in AM, MV,
QM, TM and UQIC).

The following description and measurements
are based primarily on holotype, with other
specimens measured for ranges and variation.

Diagnosis. Adult body size 6.5-8.5 mm (SS), 8
mm ( 99 ); wing length 4.5-5.5 mm S3, 5-6 mm
29 - Antennae 4-5 mm SS, 2-2.2 mm ’ 9 - 0-5 setae
on laterotergite. m-cu located halfway between
r-m and fork of M (well beyond r-m) (Fig. 12H).
r-m straight, bm-cu curved, 2.4-3.3 times length
of m-cu, sloping slightly backwards, towards
base of wing (Fig. 12H). Larvae with 9 and 10
maxillary teeth plus lobe (cf Fig. 16A, C).

Description. Adult. S holotype body length
7 mm; wing length 5 mm; antennal length 4.5
mm. Head: Antennae brown, pale yellow at
basal end of first few flagellar segments. Scape
and pedicel yellow. Flagellum with dense mat
of brown hairs, except basal seventh of first
segment. Diameter of scape 2.6 times (2.6-
3.2x) diameter of basal flagellar segment in SS
(99 2.8-3.6x). First flagellar segment 1.3 times
length of second flagellar segment (<J(J 1.2-1.3x,

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Taxonomy of Australian glow-worms

$?1.2-1.4x). Front medium to dark brown. Face
medium brown with cluster of setae over facial
carina. Palpi very pale yellow; apical segment
lengthened and slightly dilated; light covering
of dark setae on palps. Proboscis pale yellow.
Thorax: Cluster of long, dark setae on supra-alar
and post alar regions (Fig. 17H). Laterotergite
with 0-3 small setae (0-5); sometimes with
unequal numbers of setae on each side. Medium
brown shading on scutum. Pleura dark brown
(encompassing laterotergite, anepisternum,
anepimeron and katepisternum). Entire thorax
moderately shaded (light to medium brown).
Legs: Coxae yellow to brown, darkening to dark
brown at apices, more extensively so on hind
coxa. Long dark hairs covering about one third
of each coxa, predominantly at apices and on
ventral surfaces. Other segments of legs brown.
Fore basitarsus 1.3 times length of fore tibia
(9$ l.lx): Mid basitarsus 0.83 times length of
mid tibia (?9 0.74x): Hind basitarsus 0.62 times
length of hind tibia ( >y 0.65x). Whigs: Halteres
pale yellow basally, brown apically with small
hairs. Campaniform sensillae located on rJo
wings as follows: 2 on Rl, 3 on Rs, and 1 on R5
(Fig. 12H) (cJJ 1-2: 2-4: 1-3, total average = 7,
range 6-8), ($$ 2-3: 4-5: range 1-3, total average
= 9, range 8-10). Abdomen: External genitalia
medium brown. Eggs. Slightly ovoid. 0.48 x 0.45
mm, cream/brown. Larva. Larval snares <3 cm.
Pupa. Suspended from apex of abdomen and
anterior part of abdomen.

Etymology. Tine specific name refers to Otway National
Park, Victoria, which contains many colonies of
this species.

Distribution. This species is restricted to rainforest
gullies and stream banks in Otway National
Park, Melba Gully State Park and Angahook-
Lorne State Park and adjoining private land
containing rainforest, southwest Victoria.

Comments. McKeown (1935) noted the
presence of Arachrwcampa in the Otway Forest,
Victoria. This population of Arachnocampa species

remained undescribed until the current study,
although previously it had been tentatively
identified as A. richardsae (Crosby 1978). Molec-
ular data indicate A. otwayensis is the sister
species to A. gippslandcnsis + A. richardsae (Baker
et al. 2008). The colouration of fresh, live larvae of
this species was not recorded before they were
placed in ethanol or reared to the adults.

DISCUSSION

At the time of Harrison's (1966) revision of the
Australian glow-worms, the genus Arachnocampa
was included in the Mycetophilidae. Matile
(1981) subsequently separated the monophyletic
Arachnocampa from Mycetophilidae based on
seven larval characters, many taken from Edwards
(1924), placing them in the family Keroplatidae.

Morphologically, Arachnocampa is unusual
in that populations of the same species exhibit
morphological plasticity depending on their
habitat type. Major morphological differences
are evident between populations occurring
in caves and in nearby epigean habitats (e.g.
A. richardsae: SNSW1 vs SNSW2) thus making
morphological comparisons between these
groups detailed and lengthy. An example
of this morphological variation is that cave
populations display limited pigmentation, if
any, along the lateral edges of the larval body
(Figs 15B-D, G, H, K, L), whereas the larvae
of nearby rainforest populations are generally
heavily pigmented down the entire length of
the lateral edges of the body (Figs 15 A, E, F,
J). The darker pigmentation may simply be
the result of living in an epigean environment
(due to contact with sunlight), and is possibly
an adaptation to escape predation as the darker
colouration makes them more difficult to see
on dark earthy banks and vegetation. Cave-
dwelling Arachnocampa larvae are lighter and
creamier in colouration as they have no contact
with pigment altering ultra-violet light (Meyer-
Rochow pers. comm.), and pigmentation is

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Baker

presumably not a necessary adaptation for
camouflage in a dark environment. Variation
in larval pigmentation between geographically
close populations has made morphological
identification in the larval stage difficult. Larval
descriptions include both cave and epigean
colony colouration differences to illustrate this
plasticity. The relatively recent availability of
some of the man-made 'cave' environments (e.g.
mine shafts in Victoria and the Newnes railway
tunnel in New South Wales) and the obvious
degree of morphological variation between these
colonies and nearby epigean colonies strengthens
the idea that such characters are plastic within the
genus. Cave populations, such as those in Girra-
ween National Park caves (A. girraweenensis) and
Mt Buffalo cave (A. buffaloensis), that now appear
entirely restricted to the cave environment may
begin to exhibit other troglobitic attributes, as
gene-flow from epigean colonies is extremely
unlikely. Arachnocampa buffaloensis larvae
show characteristics (e.g. larval colouration,
arrangement of campaniform sensillae on the
larval head plate and wings, wing venation) that
clearly differentiate them from other main-
land Australian Arachnocampa. Whether this is
a result of their radically different habitat type,
-or simply chance phenetic mutation remains to
be tested.

Other characters that may be important in
separating populations are largely variable
between cave and rainforest populations. One
of these characters, larger body size (mirrored
by larger wing and antennal lengths), is evident
in all cave dwelling populations and has been
suggested to result from a more suitable habitat
(Pugsley 1980, 1984) and greater availability of
prey (Richards 1960). In southern New South
Wales, an artificial cave (a railway tunnel; built
between 1906 and 1907 and abandoned in 1912)
(NSW National Parks and Wildlife Service
1996) has provided a relatively new habitat
for one A. richardsae colony (SNSW1). The

larvae inhabiting this tunnel exhibit a much
larger body size range than nearby rainforest
populations (SNSW2-6), again indicating the
limited usefulness of size as a character when
dealing with organisms adapted to different
habitats. A newly constructed limestone replica
cave on Tamborine Mountain, Queensland,
now contains thousands of A. flava originally
bred from nearby rainforest gullies. This pop-
ulation now exhibits markedly larger body size
after only four years of breeding within the cave
system (personal observations).

A correlation has been made between latitude
and body size in a number of insect groups (e.g.
Muscidae: Alves & Belo 2002; Curculionidae:
Chown & Klok 2003; Culicidae: Sota 1994).
Epigean glow-worm colonies share very similar
body length ranges from north Queensland to
New South Wales and as do some in Victoria
(see species descriptions for body ranges).
However, an increase in body size is evident
in cave females, and to a smaller degree, cave
males (there are fewer data available for cave
males) with increasing latitude, with the largest
specimens recorded from caves in Tasmania,
Epigean populations near to caves also show
larger body sizes, potentially due to breeding
with large-bodied cave glow-worms within the
same species (e.g. A. gippslandensis). This increase
in body size may be attributed to decreased
temperatures in caves rather than a direct link
with latitude. Decreased temperatures increase
larval development times, thereby making
food resources available for longer periods.
This hypothesis has been tested in weevils and
correlations were found between relatively
aseasonally situated colonies and increased
body size (Chown & Klok 2003). As larger
caves have a relatively constant temperature
due to thermal inertia (de Freitas & Schmekal
2003), this theory may also apply to the
large limestone Tasmanian caves occupied
by A. tasmaniensis. Further studies utilising

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Taxonomy of Australian glow-worms

morphometric analysis are recommended to
identify possible factors driving the observed
morphological variations.

All the newly described species share the
same basic biological attributes in relation to
snare building and maintenance (refer to Baker
& Merritt 2003 for description of this behaviour).
Differences occur in the length of the snare,
but again, this is more likely an artefact of
the environment in which they live. Cave
environments inhabited by Arachnocampa are
less exposed to wind turbulence and therefore
snares can reach lengths of 20-40 cm without
tangling. In contrast, epigean Arachnocampa
are usually exposed to many weather related
factors which combine to ensure snare lengths
are short.

The cohabitation of this group in both cave and
epigean environments presents an interesting
model system for studying the transition of
a troglophilic species to a troglobitic species.
There have been many theories as to how this
transformation occurs (see Holsinger 2000 for
a review). For example, some authors believe
that founder epigean individuals happen
upon subterranean habitat and adapt to this
environment. Other theories suggest species
exhibit pre-existing characteristics that enable
them to flourish in this new environment. If
any, the second scenario appears more likely
for Arachnocampa. For instance, their reliance
on a constantly humid environment, their
need for darkness to attract prey (with their
bioluminescence) and their need for slow air
movement for long snare building are examples
of how a subterranean environment could
provide a particularly suitable habitat. The
species are equally adapted to rainforest or cave
habitats. However, it is the cave habitat that
often provides the prerequisites for colonies to
reach very large numbers (i.e. the increase in
overhang space for larval snare building).

Living in darkened environments, it is likely
that species recognition is based on adult
pheromone attraction and therefore future
species identification for this genus may involve
cuticular hydrocarbon analysis. Population level
molecular analysis is recommended to ascertain
historical gene flow between populations. In
this area, microsatellite analysis may provide
useful answers to further species identification
in this genus.

ACKNOWLEDGEMENTS

This research is an outcome of a PhD project
funded by the Sustainable Cooperative Research
Centre, established by the Australian Govern-
ment. Additional funding for the project came
from Australian Geographic, Xenogen Coporation
Pty Ltd, ABRS, JPT Tour Group (National Tour
Company), Aries Tours Pty Ltd, Forest of Dreams
and Queensland National Parks Service.

I would like to thank Greg Daniels, Jeff
Skevington, Patrice Bouchard, Christine Lambkin
and David Yeates for consultations regarding
dipteran taxonomy; Susan Wright and Chris
Burwell from the Queensland Museum for
assistance with specimen deposition and Anthony
O'Toole for providing the SEM photographs. Josh
Fartch assisted with the figures. Mark Gahan,
Kirsten Scott, Dan Pedersen, David Merritt and
David Yeates provided helpful comments on
the manuscript. Thanks must also go to those
who assisted with surveys, collections or sent
samples to me. These people include: David
Chitty, Arthur Clarke, Neil Collinson, Michael
Driessen, Lee Etherington, Bill Goebels, Harry
Hines, Ian Houshold, Deb Hunter, Bevan Jenkins,
Alison Marion, Lloyd Oldfield, Ian Roche, Andy
Spate, Dave Smith and Gary Summers.

LITERATURE CITED

Alves, S. M. & Belo, M. 2002. Morphometric variations
in the housefly, Musca aomestica, (L.) with
latitude. Genetica 115: 243-251.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1) 39

Baker

Baker, C.H. 2002. A biological basis for management
of glow-worm (Diptera: Keroplatidae) popu-
lations of ecotourism significance. Wildlife
Tourism Report Series 21. Pp. 76. Gold Coast:
Sustainable Tourism CRC.

2004. Australian Glow-worms (Diptera: Keroplatidae:
Arachnocampa spp.): Distribution, diversity,
identity and management. Unpubl. Ppl82. PhD
thesis, University of Queensland, Brisbane.

2009. Australian Glow-worms (Diptera; Keroplatidae;
Arachnocampa): An overview of their distribution,
taxonomy and phylogenetic relationships. In
Bioluminescence in Focus - A collection of
illuminating essays Ed. V.B. Meyer-Rochow.
Pp 385. Research Signpost, India,

Baker, C.H., Graham, G.C., Scott, K.D., Yeates
& D.K., Merritt, D.J. 2008. Distribution and
phylogenetic relationships of Australian glow-
worms Arachnocampa (Diptera, Keroplatidae).
Molecular Phylogenetics and Evolution 48: 506-
514.

Baker, C.H. & Merritt, D.J. 2003. Life cycle of an
Australian glow-worm A rachnocampa flam Harrison
(Diptera: Keroplatidae: Arachnocampinae).
Australian Entomologist 30: 45-55.

Chown, S.L. & Klok, C.J. 2003. Altitudinal body size
dines: latitudinal effects associated with changing
seasonality. Ecography 26: 445-455.

Crosby, D.F. 1978. Glowworms. Victorian Entomologist
8(6): 56-59. 6

Currey, J.E.B. 1966. Reflections on the Colony of
New South Wales: (Landsdowne Press).

De Freitas, C.R. & Schmekal, A. 2003. Condensation
as a microclimate process: Measurement,
numerical simulation and prediction in the glow-
worm cave. New Zealand. International Journal of
Climatology 23: 557-575.

Department of Conservation and Natural Resources.
1994. Melba Gully State Park draft management
plan. Pp. 21. Melbourne: National Parks Service-
South West Area.

Eberhard, S. 1999. Cave fauna management and
monitoring at Ida Bay, Tasmania. Nature
Conservation Report 99/ 1, pp. 37: Tasmania Parks
and Wildlife Service, Tasmania.

Eberhard, S. 2000. Reconnaissance survey of cave
fauna management issues in the Mole Creek Karst
National Park, 1 asmania. Nature Conservation
Report 2000/1, pp. 38: Department of Primary
Industry, Water and Environment.

Eberhard, S.M. & Spate, A.P. 1995. Cave Invertebrate
Survey: Towards an atlas of New South Wales
cave fauna. Pp. 112. National Parks and Wildlife
Service of New South Wales.

Edwards, F.W. 1924. A note on the 'New Zealand
Glow-worm' (Diptera, Mycetophilidae). Annual
Magazine ofNahiral History series 9 (14): 175-179.

Ferguson, E.W. 1925. Description of a new species
of Mycetophilidae (Diptera) with luminous larvae.
Proceedings of the Linnean Society of New South
Wales 50: 487-488.

Finlayson, B. 1982. Granite caves in Girraween National
Park, southeast Queensland. Helictite 20: 53-59.

Goede, A. 1967. Tasmanian cave fauna: character
and distribution. Helictite 6: 71-85.

Harrison, R.A. 1966. Australian glow-worms of the

f en us Arachnocampa Edwards. Pacific Insects 8:
77-883.

Holsinger, J.R. 2000. Ecological Derivation, Colonisation,
and Speciation. In. Wilkins H., Culver D. C. and
Humphreys W. F. (eds) Ecosystems of the World:
Subterranean ecosystems, Amsterdam: Elsevier
Science B.V.

Matile, L. 1981. Description d'un Keroplatidae du
Cretace moyen et donnees morphologiques
et taxinomiques sur les Mycetophiloidea
(Diptera). Annals de la Societe entomolgique de
France (New Series) 17: 99-123.

1990. Reserches sur la systematic et 1'evolution
des Keroplatidae (Diptera, Mycetophiloidea).
Paris: Memoires du Museum National D'Histoire
Naturelle. 148: 1-654.

Mcalpine, J.F. 1981. Morphology and Terminology-
Adults. Pp. 674. In McAlpine J. F., Peterson
B. V., Shewell G. E. , Tesky H. J., Vockeroth
J. R. and Wood D. M. (ed.) Manual of Neartic
Diptera, Vol. 1, Quebec: Canadian Government
Publishing Centre.

Mckeown, K.C. 1935. Insect wonders of Australia. Pp
252. Sydney: Angus and Robertson.
Meyer-Rochow, V.B. 2007. Glow-worms: a review
of Arachnocampa spp. and kin. Luminescence
22(13): 251-265.

New South Wales National Parks and Wildlife Service.
(1996). Wollemi National Park walking track
guide. Newnes and the glow worm tunnel,
pp. 1 - 19: NSW National Parks and Wildlife
service, Blackheath.

Perkins, F.A. 1935. 'Glow-worms' and other insects
collected at Numinbah, Nerang Valley, Easter

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Taxonomy of Australian glow-worms

Camp, Queensland Naturalists' Club, 1935. Vie
Queensland Naturalist 9: 84-85.

Pugsley, C.W. 1980. Ecology of the New Zealand
glowworm in caves at Waitomo. Pp. 107. PhD
Thesis, Auckland: Auckland University.

1983. Literature review of the New Zealand
glowworm Arachnocampa luminosa (Diptera:
Keroplatidae) and related cave-dwelling Diptera.
Nezv Zealand Entomologist 7: 419-424.

1984. Ecology of the New Zealand glowworm,
Arachnocampa luminosa (Diptera: Keroplatidae),
in the glowworm cave, Waitomo. Journal of the
Royal Society of New Zealand 14: 387-407.

Richards, A.M. 1960. Observations on the New Zealand
glow-worm Arachnocampa luminosa (Skuse) 1890.
Transactions of the Royal Society of New Zealand
88:559-574.

Skuse, F.A.A. 1890. Description of a luminous
dipterous insect (Fam. Mycetophilidae) from
New Zealand. Proceedings of the Linnean Society
of New South Wales. Second series 5: 677-679.

Sota, T. 1994. Larval diapause, size and autogeny in
the mosquito Aedes togoi (Diptera, Culicidae)
from tropical to sub-arctic zones. Canadian
Journal of Zoology-Revue Canadienne de Zoology
72: 1462-1468.

Memoirs of the Queensland Museum

Nature • 2010 • 55(1)

Revision of Flindersichthys denmeadi Longman
1932, a marine teleost from the Lower Cretaceous
of the Great Artesian Basin, Queensland

Alan BARTHOLOMAI

Director Emeritus, Queensland Museum, PO Box 3300, South Brisbane, Qld 4101, Australia.

Citation: Bartholomai, A. 2010 03 15. Revision of Flindersichthys denmeadi Longman 1932,
a marine teleost from the Lower Cretaceous of the Great Artesian Basin, Queensland. Memoirs of the
Queensland Museum — Nature 55(1): 43-68. Brisbane. ISSN 0079-8835. Accepted: 20 January 2009.

ABSTRACT

Prepared and naturally exposed specimens of the large, fossil teleost, Flindersichthys
denmeadi Longman, 1932, in the collections of the Queensland Museum, form the
basis of a revision of this relatively common, carnivorous, Lower Cretaceous (latest
mid to late Albian) marine fish. Originally described from a single specimen, this larger
sample has enabled the skeletal morphology of the taxon to be amplified and, where
necessary, clarified. Flindersichthys has been referred to the Order Elopiformes within
the Superorder Elopomorpha but is considered incertae sedis at familial level. Nearly all
referred specimens have been derived from marine sediments of the Toolebuc and Allaru
Formations in the northern part of the Eromanga Basin, part of the Great Artesian Basin
but rare occurrences are recorded from the Normanton Formation in the Carpentaria
Basin. The morphological description remains incomplete, especially in the post-cranial
area, with the material studied being almost exclusively cranial or from the very anterior
of the body. □ Teleostei, Elopomorpha, Elopiformes, Flindersichthys denmeadi; Lower
Cretaceous (latest mid to late Albian); Great Artesian, Eromanga and Carpentaria Basins;
Toolebuc, Allaru and Normanton Formations; Euroka Arch.

Lower Cretaceous (latest mid to late Albian)
marine sediments of the Great Artesian Basin in
central Queensland, Australia, have long been
the source of a rich suite of fossil vertebrates and
invertebrates. The earliest described vertebrate
taxa were identified in Etheridge (1872) and the
fauna is now known to include a range of marine
fishes comprising both chondrichthyans and
osteichthyan actinopterygians, together with
numerous marine reptiles and even terrestrial
vertebrates that were washed or carried out into
the epeiric sea or which died and were preserved
in littoral or deeper water situations. Among the
actinopterygians already recorded, Bartholomai
(1969) has revised the pachyrhizodontid teleost,
Pachyrhizodus marathonensis (Etheridge Jnr.
1905) and later (Bartholomai 2004) he revised

the aspidorhynchid, Richmondichthys sweeti
(Etheridge Jnr. & Smith Woodward 1891).
The ichthyodectiform, Cooyoo australis (Smith
Woodward 1894) was revised by Lees &
Bartholomai (1987), while a probable neoteleost,
Dugaldia emmelta, was described by Lees
(1990). Kear (2007) added the pachycormid,
Australopachycormus hurleyi Kear from the
Toolebuc Formation near Boulia in the north-
west of the Eromanga Basin portion of the
Great Artesian Basin.

The current study represents a continuation
of work on this latest mid to late Albian fish
fauna. It is almost exclusively based on a review
of exposed and acetic acid prepared specimens
in the collections of the Queensland Museum,
mostly those added over the last few decades

Memoirs of the Queensland Museum | Nature • 2010 • 55(1) • www.qm.qld.gov.au

Bartholomai

through field work by the author, his research
assistant Ms Tempe Lees (retired), other staff
of the institution, especially Mr Terry Tebble,
Senior Preparator (retired) and through donations
by members of the public,.

Almost all the fossil fish referable to
Flindersichthys denmeadi have been derived
from the marine Toolebuc Formation (with
less from the Allaru Formation) in the northern
Eromanga Basin, the largest part of the
Artesian Basin. Indeed, no material referable
to Flindersichthys has yet been discovered from
the Eromanga Basin away from the inflow area
across and below the basement Euroka Arch
south of the current Gulf of Carpentaria and
in the nearby Hughenden- Richmond area of
north-central Queensland. Rare occurrences do
occur further north of the Euroka Arch, within
the Carpentaria Basin, from near Weipa, Cape
York and at the Little Bynoe River crossing, in
marine sediments interpreted as the Normanton
Formation. Dating of the Toolebuc Formation
as latest mid to late Albian is based on its
correlation with the Pseudoceratium ludbrookiae
dinoflagellate zone and the upper Coptospora
paradoxa- Pliimopollenites pannosus zone (Moore
. et al. 1986; McMinn & Burger 1986). A more
refined age of early Late Albian for the
Toolebuc has been suggested by Henderson
(2004), based on ammonite and nannofossil
biostratigraphy, with the Formation being no
older than the upper Prediscosphyraera columnata
zone and no younger than the Mortoniceras
inflatum zone. However, the less precise age
has been applied in the current work. The
Toolebuc contributes only some 5-35 m of the
total ca.2 km thickness of Lower Cretaceous
marine sediment in the Eromanga Basin. The
Allaru Formation deposited conformably on
the Toolebuc is much thicker than the Toolebuc
but is also considered to be of Late Albian age,
again within the P. pannosus zone (Burger
1986). The Normanton Formation is generally

1 44

regarded as having been deposited during the
early Cretaceous.

Marine incursions into the Great Artesian
Basin during the Lower Cretaceous were rel-
atively short lived. Surface expression of
the sedimentary deposits is extensive but
poor because of extremely shallow dips and
low topographic elevation. These physical
characteristics have resulted in deep weathering
and limited exposures through erosion of the
predominant Toolebuc facies, viz. coquinite
and organic-rich shale (Henderson 2004).
Fortunately, many of the fossils found at the
surface exist as the nucleus around which
calcium carbonate was deposited within the
sediments, creating hardened calcareous
concretions. These appear to be related to the
coquinite units and are often exposed on
the surface by deflation or are concentrated
in such gullies and streams as are rarely
encountered. More complete teleost skeletons,
not covered by concretionary structures, have
been discovered in the coquinite units during
recent excavations by the Kronosaurus Korner
Museum, Richmond (Stumkat, pers. comm.).

The fossil fish remains in the current study are
therefore predominately of cranial specimens
preserved in concretions, occasionally including
partial skeletons that are mostly of the anterior
of the body. Scattered, disarticulated material
is present and there are very rare instances
where more complete skeletons have been
encountered. Most species, including F. denmeadi,
are represented by large to very large indi-
viduals. The occasional presence of more
complete skeletons and identification of new
taxa, regardless of the completeness of the
material, encourages the continuation of further
exploration, as does the presence of masses of
disarticulated bones of very small individuals
that are occasionally found in the coquinites at
the surface. Small actinopterygians have also
been encountered rarely in cores recovered from
depth in the sequence (e.g. Parfrey, 1990).

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Revision of Flindersichthys denmeadi

The fossil fish fauna in general has come
from the arc of sediments stretching from near
Hughenden, in the northeast of the Eromanga
Basin westwards through and above Richmond
and Julia Creek and then down to beyond
Boulia, in the northwest of the Eromanga Basin.
The age attributed to the fauna suggests it has
the potential to add to a better understanding
of early teleost radiation in the oceanic and near-
oceanic waters off the dispersing Gondwanaland
tectonic plates. Unfortunately, Lower Cretaceous
(Aptian) sediments from within the Great
Artesian Basin have limited fossil fish records.
Specimens from above and close to the Euroka
Arch duringthe later Toolebuc-Allaru limes must
have had ready access to tidal, oceanic waters
from the north, with the majority of fish finds
continuing around the north and northwest of
the Eromanga Basin, in keeping with a suggested
anti-clockwise current flow within the epeiric
sea. Glikson & Taylor (1986) concluded that
the Euroka Arch intermittently restricted the
area south of the Gulf of Carpentaria and that
facies changes show that the eastern side of
the entry strait across the Arch, from which a
concentration of fossil fishes has been located,
was shallow for a greater distance from land
than usual. The entry was partially blocked by
the N-S St. Elmo Structure towards the west.
Depositional conditions in both Toolebuc and
Allaru times are discussed in Wade (1993), as
well as in Henderson (2004).

ABBREVIATIONS USED IN TEXT FIGURES

ace anterior ceratohyal

ang angular

ao antorbital

apal autopalatine

art articular

asp autosphenotic

bh basihyal

boc basioccipital

brr branchiostegal ray

bsp basisphenoid

cl cleithrum

de dermethmoid

den dentary

df dilatator fossa

dhh dorsal hypohyal

dpal dermopalatine

d.pl dental plate

ecp ectopterygoid

enp endopterygoid

epo epiotic

exo exoccipital

fahm hyomandibular facet

fica foramen for internal carotid artery

fm foramen magnum

foa foramen for orbital artery

fr frontal

fsp foramen for occipital nerve

fuv fused vertebral centrum

hm hyomandibular

ic intercalary

io infraorbital (1-5)

iop interoperculum

l. e lateral ethmoid

mes mesethmoid

mpt metapterygoid

m. s.c mandibular sensory canal

mx maxilla

op operculum

ors orbitosphenoid

ot.s.c otic sensory canal

pa parietal

par parasphenoid

part.p .... postarticular process

pee posterior ceratohyal

pci postcleithrum

pec f pectoral fin

pmx premaxilla

pop preoperculum

pro prootic

Memoirs of the Queensland Museum 1 Nature • 2010 • 55(1)

Bartholomai

pro.ic .... prootic-intercalar bridge

psp pterosphenoid

ptf post-temporal fossa

pto pterotic

ptfc posterior opening of pars jugularis

ptt post-temporal

qu quadrate

rart retroarticular

scl supracleithrum

sc pi sclerotic plate

smx supramaxilla (1-2)

so supraorbital

soc supraoccipital

soc sp . . . . supraoccipital spine

sop suboperculum

so.s.c supraorbital sensory canal

stt supratemporal

sy symplectic

tpbh basihyal tooth plate

vhh ventral hypohyal

vo ....... vomer

I foramen for olfactory tract

II foramen for optic tract

Vllhm .... foramen for hyomandibular trunk
:of facial

VHot foramen for otic branch of facial

IX foramen for glossopharangeal

X foramen for vagus

SYSTEMATIC DESCRIPTIONS
AND DISCUSSIONS

Division Teleostei
Superorder Elopomorpha
Order Elopiformes
Family incertae sedis

Discussion. Reference of fossil taxa within an
acceptable Linnean classification is problem-

146

atical in most instances for those that, in the
fauna are shown to possess generalised
elopomorph characters. Forey et al. (1996)
indicate that the elopomorph fishes, including
those referable to the Order Elopiformes, have
a rich but patchy fossil record extending back
to the Lower Cretaceous (Valanginian) and
even to the Upper Jurassic (Kimmeridgian)
where Anaethalion is concerned (see Forey 1973),
considerably earlier than the marine Lower
Cretaceous (Albian) sediments from which
Flindersichthys has been recorded. However, these
authors acknowledge the generalised features
of many elopomorph fossils that result in their
relegation as incertae sedis at various ranks
and conclude that a number of older taxa
traditionally referred to the Elopiformes should
be regarded as Elopomorpha incertae sedis.
Nonetheless, it has been considered reasonable
to adopt a less conservative position in regard
to the present taxon (see below).

It has been felt prudent to maintain an incertae
sedis status for Flindersichthys, at the familial
level, rather than to refer it to the Family
Megalopidae, action taken by Taverne (1999)
in regard to the somewhat similar elopiform,
Arratiaelops, from the Lower Cretaceous Wealdon
of England and Belgium.

Flindersichthys Longman, 1932

Flindersichthys Longman, 1932: 89; Taverne, 1999: 91-3.

Generic Diagnosis (emended). A very large
elopiform with prognathous lower jaw, reaching
in excess of 1.25 metres in total estimated body
length. Neurocranial roof moderately broad,
slightly depressed posteromedially. Maximum
depth of neurocranium at occipital region. Orbit
relatively small. Dermethmoid laterally with
prominent, elongated, posteroventral processes
and broad, medial, dorsal ridge, terminating in
elevated opening above anterior of elongate,
interfrontal fontenelle. Rostral absent. Outer
margin of olfactory capsule defined anteriorly
by dermethmoid process and posteriorly by

Memoirs of the Queensland Museum I Nature • 2010 • 55(1)

Revision of Flindersichthys denmeadi

lateral ethmoid below and by rarely preserved,
relatively small supraorbital above. Antorbital
present. Nasal plate-like. Parietal subovate
but sometimes squarer anteriorly, longer than
broad. Supraoccipital with strong posterior
spine. Dilatator fossa elongate, anteriorly deep,
shallow, partially roofed along its entire length by
the pterotic, reduced posteriorly by prominent,
curved ridge forming roof of hyomandibular facet.
Strong, subparallel ridges occur longitudinally on
the pterotic, parietal and the back of the frontal.
Exoccipital excluded from margin of post-
temporal fossa. Intercalar contributes significant-
ly to the prootic-intercalar bridge. Autosphenotic
spine extended posteroventrally into stout, lateral
ridge. Descending lamina of frontal 'V' shaped,
large, dished and deeply plicated longitudinally
meeting anterodorsal margins of autosphenotic
and pterosphenoid and posterodorsal margin
of orbitosphenoid. Orbitosphenoid very large,
extended dorsoventrally and anteriorly by ossified
interorbital septum reaching to parasphenoid,
with large, prominently walled foramen
posteroventrally. Parasphenoid near planar,
deepening into inverted 'Y' shape anteriorly,
ventrolaterally supported to below orbit by
posteriorly tapering vomerine processes.
Vomer with small, multiserial, villiform
teeth. Dermopalatine with larger teeth. Endo-
pterygoid and metapterygoid very large.
Maxilla extends posteriorly well beyond level
of orbit. Sclerotic plates present. Circumorbital
series incomplete above orbit. Jaws massive,
with dentary shallower anteriorly and with
relatively low coronoid process. Retroarticular
not fused. A relatively extensive external part
of the angular anteriorly meets the dentary
and posteriorly joins with the retroarticular
to produce prominent postarticular process.
Mandibular sensory canal opens medially. Ang-
ular and articular contribute to articulatory cup
internally, separated by prominent fissure. Gular
elongate but narrow. Dentition on premaxilla,
maxilla and dentary of very small, multiserial,
villiform, conical teeth represented mainly by

hollow, circular bases, set on expanded oral
plates that often extend beyond the buccal
margins. Occasional slightly larger, slightly
recurved teeth occur internally. Hyomandibular
with elongate process strengthened by strong,
medial, angular ridge. Anterior ceratohyal very
large, posteriorly deep, fenestrated towards
dorsal margin. At least 14 pairs of branchiostegal
rays present. Operculum and suboperculum
large. Supratemporal extremely large. Ganoine
present. Pectoral splint and at least 15 pectoral fin
rays present. Vertebrae large, cylindrical, much
shorter than high, laterally with numerous, fine,
longitudinal striae.

Type and only species. Flindersichthys denmeadi
Longman, 1932.

Flindersichthys denmeadi Longman, 1932
(Figs 1-8).

Flindersichthys denmeadi Longman, 1932: 69-97; Taverne,

1999: 91-3!

Specific diagnosis (emended). As for the genus.

Holotype. QMF2210, nearly complete skull. Flinders
River, one mile east of Richmond, NCQ., from sedi-
ments now interpreted as Toolebuc Formation.

Material examined. QMF2388, partial posterior of
skull and anterior of body, Hughenden district, NCQ.
QMF5780, partial skull, crushed dorsoventrally,
Stewart Creek, Hughenden, NCQ. QMF5798, partial
skull, 'Boree Park' Station, west of Richmond, NCQ.
QMF11042, posterior of neurocraium. Little Bynoe
River crossing, NCQ. QMF12878, incomplete skull
and body, 'loronto Park' Station, 3 km. east of
homestead in gully on ridge above creek, Toolebuc
Fm„ NCQ. QMF13735, QMF13736, QMF13769,
QMF13797, QMF13800, QMF13804, QMF13809,
QMF13811, QMF13848, QMF13888, partial skeletons,
'Dunraven' Station, nr. Hughenden, banks of
unnamed branch of Stewart Creek, NCQ. QMF13720,
nearly complete skull, 'Dunraven' Station, nr.
Hughenden, unnamed tributary of Stewart Creek,
NCQ. QMF13714, partial skull, 'Dunraven' Station,
nr. Hughenden, Stewart Creek, west of Pelican
Bore, NCQ. QMF13743, QMF13755, QMF13813,
partial skeletons, 'Dunraven' Station, nr. Hughenden,
Stewart Creek, downstream from Pelican Bore,
NCQ. QMF13707, QMF13715, QMF52273, partial
skull, 'Dunraven' Station, nr. Hughenden, cobble

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Bartholomai

bed nr. Pelican Bore, NCQ. QMF13719, QMF13737,
QMF13807, partial skeleton, 'Dunraven' Station, nr.
Hughenden, Stewart Creek, upstream from junction
with Soda Creek, NCQ. QMF13808, partial skeleton,
'Dunraven' Station, nr. Hughenden, Stewart Creek,
upstream from Pelican Bore, NCQ. QMF15215,
partial skull, 'Dunraven' Station, nr. Hughenden,
NCQ. QMF1 3601, posterior of skull and anterior of
body, 'Dunraven' Station, nr. Hughenden, at Pelican
Bore, NCQ. QMF15986, partial skeleton, 'Alderley'
Station, via. Hughenden, NCQ. QMF18916, skull and
partial body, upper tributary of Mvall Creek, 'York
Downs' Station, nr. Weipa, Cape York, at 12° 40'S
and 142° 22'E, mapped as undifferentiated Rolling
Downs Group but most probably Normanton Fm.
based on BMR Weipa 1 drill core, of Albian age,
Carpentaria Basin (see Smart et al., 1980). QMF52274,
almost complete skull, 'Dunraven' Station, steep gully
of Stewart Creek, at 20° 29.98' S, 148° 55.5' E, NCQ.

Formations and Age. Marine Toolebuc and Allaru
Formations of Lower Cretaceous (latest mid to
late Albian) age and Normanton Formation of
Early Cretaceous age.

Description. A very large species with an esti-
mated length exceeding 1.25metres and with an
estimated body depth of more than 30 cm.

Neurocranium. The neurocranium is slightly
less than three times as long as deep. Width
of the neurocranium is often difficult to assess
accurately because of distortion that is usual
during preservation. However, it appears that
maximum width across the occipital region, is
only slightly more than across the area of the
autosphenotic ridges. Maximum depth is in the
occipital area. The orbit is relatively small and
comprises about 17 % of the total length of the
neurocranium.

The dermethmoid is a thin element forming
the anterior end of the neurocranial roof.
The bone is elongate, anteriorly narrow and
broadest posteriorly and is constricted in dorsal
view about one-third its length from the front.
It is broadly 'V'-shaped anteriorly in dorsal
view and anteriorly possesses a low, broad,
median swelling becoming a well-defined ridge
running longitudinally; this opens posteriorly

1 48

through a large, slightly elevated and anteriorly
directed pit at the anterior of an interfrontal
fontenelle. A poorly developed groove parallels
the lateral margin of the anterior point. No
ethmoid commissure appears present. The front
of the ethmoid area slopes posteroventrally and is
sometimes separated by an unossified area from
the mesethmoid below. The mesethmoid bears a
maxillary facet on each side, behind the broadly
pointed tip. It is possible that a further facet
exists along each of the converging surfaces of
the anterior point to accommodate the dorsal
wing of the premaxilla. Ventral to the maxillary
facet, the mesethmoid carries an articulating
surface for contact with the autopalatine head.
Internally, the ethmoid area appears largely
cartilaginous posteriorly in some specimens.
The anterior of the mesethmoid is narrowly
dumbbell-shaped above the anterior of the
vomer and appears to then expand posteriorly
to broadly support the upper surface of that
bone. A prominent, elongate, lateral process
of the dermethmoid curves posteroventrally
from near the middle of the dorsal margin,
delimiting the anterior and anteroventral parts
of the olfactory capsule. The lateral process is,
in turn, sutured to the supraorbital and lateral
ethmoid, towards the back and also below the
lateral and posterior margins of the olfactory
capsule. The dorsal surface of the dermethmoid
in front of the anterodorsal margin of the
olfactory capsule is shallowly indented to
accommodate the front of the nasal.

The frontals make up the major part of the
skull roof. Each frontal meets that of the other
side along a medial, longitudinal suture that
becomes interdigitated posteriorly. The lateral
margin progressively widens above the lateral
ethmoid. Anterior to this, the dorsal surface of
the frontal is pointed to above the front of the
olfactory capsule where it meets the back of the
dermethmoid. It also provides the margin of the
unossified, interfrontal fontenelle that variably
separates the frontals anteriorly. Above the

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Revision of Flindersichthys denmeadi

FIG. 1. Flindersichthys denmeadi Longman. Composite reconstruction of cranium in right lateral view; length
ca. 31.8 cm.

orbit the lateral margin again expands before
being gently concave for a relatively short
distance, reflecting the presence of a relatively
small eye. Greatest width of the frontal occurs
above the autosphenotic. The frontal then
narrows posteriorly to its junction with the
parietal. The dorsal surface of the cranium
along the medial suture is shallowly depressed
in the posterior moiety but this is often accen-
tuated by lateral or dorsoventral crushing
during preservation. The supraorbital sensory
canal is roofed and lies within a longitudinal
ridge that extends from near the anterolateral
margin behind the nasal, continuing back to the
surface of the parietal. Strong subparallel, near
longitudinal, subsidiary ridges emanate from
above the middle of the autosphenotic and
the main ridge onto the parietal while others

pass posterolaterally onto the pterotic. Pores
associated with the supraorbital sensory canal
system occur irregularly towards the inner ends
of some of the shallow grooves between the
subsidiary ridges, especially posterolaterally
and on the parietal. Ventrally/ below the orbit,
the frontal has a prominent, dished, V'-shaped
descending lamina that is characterised by
numerous, deep, longitudinal placations that
often separate variable small to very small
foramina. The lamina meets the dorsal part of
the autosphenotic within the orbit, as well as
the pterosphenoid and orbitosphenoid along
its raised rims.

The nasal is rarely preserved and is only
known from its anterior moiety. It is a thin, plate-
like bone and carries the supraorbital sensory

Memoirs of the Queensland Museum 1 Nature • 2010 • 55(1)

Bartholomai

FIG. 2. Flindcrsichtyys denmeadi Longman. QMF13720, view of right side of incomplete skull (prepared by
acetic acid treatment), unnamed tributary of Stewart Creek, 'Dunraven' Station, nr. Hughenden, NCQ.,
length 48 cm.

canal and would have covered much of the
dorsomedial margin of the olfactory capsule.

No rostral element has been identified.

The parietal is a relatively small, irregularly
ovate bone that is sometimes more squared
anteriorly in some specimens. It is longer than
broad and meets its counterpart along the
midline except for a short distance posteriorly
where the supraoccipital intrudes onto the
dorsal surface for a short distance. The dorsal
surface is ornamented by the extension of the
main ridge of the supraorbital sensory canal
and subsidiary ridges that are subparallel with
the midline.

The posterolateral part of the skull roof is
formed by the dorsal portion of the prominent
pterotic. This forms the bulk of the roof of
the post-temporal fossa. The posteromedial
margin meets the dorsal part of the epiotic
along a sinuous junction. The dilatator fossa
is moderately elongate, anteriorly deep and
shallow and is contributed to and partially

1 50

roofed along its length by the lateral margin
of the pterotic, which extends anteriorly to
above the middle of the autosphenotic ridge.
The dilatator fossa is limited posteriorly by a
strong, obliquely anteroventrally curving ridge
that excludes the fossa from the posterolateral
comer of the neurocranium and forms the upper
rim of the hyomandibular facet. The lower rim
of the hyomandibular facet provides the roof
of the relatively shallow sub-temporal fossa.
Medially, the pterotic contributes to the lateral
wall of the post-temporal fossa. Posteriorly, it
contributes only relatively minimally to the
dorsolateral margin of the post-temporal fossa
external to its junction with the epiotic and
above the dorsally extended intercalar. The
dorsal surface of the pterotic is ornamented
by continuations of the relatively strong,
subparallel, subsidiary ridges, slightly angled
away from the main supraorbital sensory canal
ridge and pit lines from the supraorbital sensory
canal are present posterolaterally.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Revision of Flindersichthys denmeadi

FIG. 3. Flindersichthys denmeadi Longman. Composite reconstructions of neurocranium. A, dorsal view; B,
left lateral view; C, ventral view.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Bartholomai

The epiotic is visible posteriorly and dorsally,
the latter comprising a minor expansion along
the posterodorsal margin of the neurocranial
roof, together with a knob-like epiotic process
development. The epiotic also forms almost all
the medial margin of the opening of the post-
temporal fossa and contributes to its medial
wall but joins with the intercalar in excluding
the exoccipital from the ventromesial margin
of the post-temporal fossa. It is slightly dished
across the posterior surface but not to the extent
of forming a subepiotic fossa.

The supraoccipital is a relatively small bone
which, posteriorly lies between the epiotics
and posterodorsally extends onto the dorsal
surface between the back of the parietals. It is
shield-shaped in posterior view and possesses
a stout, dorsomedial spine, directed posteriorly
from the upper part of the posterior surface (2.7
cm long in the holotype). The bone is slightly
dished across the posterior surface.

The exoccipital is excluded from contributing
to the margin of the opening of the post-temp-
oral fossa by the epiotic and intercalar. It meets
its counterpart in a vertical junction above
the foramen magnum but does not appear to
meet below this foramen. It is dished inwards
in posterior view but not developed into a
definable subepiotic fossa. A small foramen
is present for the occipital nerve, close to the
foramen magnum. The exoccipital has a short
junction with the intercalar posterolaterally
and a much longer junction with the epiotic
posterodorsally. The back of the lateral face of
the exoccipital is not well preserved. However,
a large foramen for the vagus nerve is present.
Anteroventrally to this is a slightly smaller
foramen for the glossopharangeal nerve.
Ventrally, the exoccipital curves deeply over
the dorsal surface of the basioccipital and then
curves sharply anterodorsally towards and
then below the anterior base of the intercalar-
prootic bridge.

1 52

The intercalar forms a large cap over the
posteroventral corner of the base of the opening
of the post-temporal fossa. It is particularly
well developed between its junctions with
the exoccipital, epiotic and pterotic within the
fossa. The posteroventral corner of the post-
temporal fossa margin is extended backwards
to form an angled knob on the intercalar to
support the ventral limb of the post-temporal
bone. Although the full extent of the anterior
extension of the intercalar on the lateral face
of the neurocranium has not been observed, a
strong base for an anterior ridge is present as a
significant contribution to the prootic-intercalar
bridge standing clear of the neurocranial wall
below the subtemporal fossa.

The basioccipital is broadly 'W'-shaped in
posterior view with the central part contributing
the posterior myodome. The posterior portion
of the basioccipital extends dorsally to meet
the exoccipital and is firmly united with a
thin vertebral centrum that is often difficult
to recognise as discrete in adult specimens.
In younger individuals, the vertebra tapers
dorsally in lateral view. Dorsally there is a short
autogenous neural arch and spine but there
are no pits developed to accommodate these
on the dorsal surface of the attached vertebra.
In lateral view, the basioccipital extends to the
prootic along a near 45°, posterodorsally angled
junction that reaches the anterior base of the
exoccipital. The longitudinal junction with the
exoccipital is broadly convex ventrally.

The prootic is imperfectly preserved anteriorly
in all specimens and this may reflect poor
ossification of this element. It is a relatively large
bone. The lateral face is extended posteriorly
into the anterior moiety' of the prootic-intercalar
bridge and contributes the bulk of the anterior
of the relatively deep, subtemporal fossa. Fora-
mina for the hyomandibular trunk of the
facial nerve and for the orbital artery are well
developed towards the anterior of the lateral
face of the prootic and many small foramina

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Revision of Flindersichthys denmeadi

are present immediately anterior to the base of
the prootic-intercalar bridge. Above the base
of the bridge, the junction of the prootic and
exoccipital is penetrated by a foramen, believed
to be for the jugular vein.

The autosphenotic is large and has an
expanded anteroventral face that descends
medially from a longitudinally elongated, 'ridge-
like' autosphenotic spine, curving broadly
then flattening below the frontal and then
recurving to meet the posterolateral part of the
pterosphenoid. The opening for the otic branch
of the facial nerve is close to this contact. The
autosphenotic also has an extensive junction
anterodorsally with the back of the prominent,
'V'-shaped descending lamina of the frontal. The
anterior of the hyomandibular facet is partially
accommodated on the posterior of the bone.

The pterosphenoid anterodorsally has a sig-
nificant, raised contact with the descending

FIG. 4. Flindersichthys denmeadi Longman. Composite
reconstruction of posterior of neurocranium.

10cm

FIG. 5. Flindersichthys denmeadi Longman. Composite reconstruction of left hyopalatine bones in lateral view.

Memoirs of the Queensland Museum 1 Nature • 2010 • 55(1)

Bartholomai

FIG. 6. Flindersichthys denmeadi Longman. Medial view of partial right hyoid bar and remains of anterior
branchiostegal rays in QMF5798.

lamina of the frontal and a short contact with the
orbitosphenoid, behind the raised wall of the
large foramen at the back of that element. The
bone also appears to provide a dorsal margin
for the large opening for the optic tract.

The orbitosphenoid is a very large element,
ventrally extending well anterior to the ptero-
sphenoid. Posterodorsally, the bone unites with
the anterior of the large, 'V'-shaped, descend-
ing lamina of the frontal along a raised rim. The
dorsal part of the orbitosphenoid is laterally
bulbous. An extensive, ossified, interorbital
septum is present and its contact with the
orbitosphenoid is sometimes deeply and
coarsely interdigitated but the junction is often
difficult to see. The contact is 'Y' -shaped in
section and the anteroventrally angled septum
meets the upper surface of the parasphenoid
over an elongated contact. Posteriorly, the orbit-

osphenoid bears a large, foramen in front of
the short junction with the pterosphenoid,
surrounded by high walls around its base.
The anterior margin of the orbitosphenoid is
occasionally overlain by irregular flanges of
bone from the lateral ethmoid.

The basisphenoid has not been identified in
any specimen.

The lateral ethmoid is large and is curved
anteroventrally to contact the posterior of
the lateral dermethmoid process, below the
anterior of the supraorbital. It also underlies
the supraorbital below the back of the olfactory
capsule, where it contacts the mesethnioid
medially. It provides smooth anterior and dorsal
upper surfaces for the front of the orbit and
appears to meet its counterpart from the other
side at the midline. It is sutured to the frontal
posteriorly and is sometimes incompletely

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Revision of Flindersichthys denmeadi

ossified where it meets and sometimes overlaps
the anterior of the orbitosphenoid. The dorsal
surface in some specimens appears to be exposed
minimally between the back of the olfactory
capsule, the possible lateral margin of the
nasal and the supraorbital. Anteroventrally,
it is terminated medial to this area by a large
cartilaginous meniscus but does not otherwise
meet the parasphenoid.

The parasphenoid is elongate and near planar,
extending posteriorly from its contact with
the vomer but is not well exposed in most
specimens. It extends back to a position near
the posterior rim of the basioccipital. Towards
the junction with the vomer, the bone deepens
dorsally assuming an inverted 'Y'-shape and
is sandwiched laterally between the posterior
arms of that bone. It is relatively narrow below
the orbit, broadening posteriorly and forming
the anterior floor of the posterior myodome.
The central area is ventrally produced into a
sharp, strongly developed median ridge, below
the orbital area. This is reduced to a more
rounded ridge anterior to the ascending wing,
immediately anterior to an elevated medial
foramen, presumably for the buccohypophysial
canal. The bone then subdivides into two
irregular and bulbous areas before reducing
rapidly to reach the attached vertebra of the
basioccipital. The notch between the posteriorly
arms is penetrated by a large foramen leading
into the myodome. The anterior of the base of
each of the posterior arms is penetrated by a
well defined foramen for the internal carotid
artery. The posterior of the broad, ascending
wing is slightly longer than the anterior and
is produced dorsally. A small area of villiform
teeth exists on the ventral surface of the
parasphenoid below the back of the orbit in
some individuals.

The vomer is incompletely exposed in all
specimens. It is broadly attached to the base of
the mesethmoid. It appears to have a broadened,
heart-shaped dentigenous head covered with

villiform teeth that are somewhat smaller
than those on the dermopalatine. The middle
of the toothed surface appears to have been
longitudinally subdivided by an area of even
smaller teeth. It extends and tapers posteriorly
below the anterior of the parasphenoid and has
elevated wings that sandwich the sides of the
anterior of the parasphenoid.

Very large anterior sclerotic plates are present
but posterior plates have been preserved only
rarely.

Hi/opalatine bones. The series is deeper than it
is long in adult specimens, with the quadrate-
mandibular articulation lying well behind the
back of the orbit.

The hyomandibular is angled slightly post-
erior to the vertical. The head of the bone
has anteriorly inclined, anterior and posterior
articulating surfaces that fit into the relevant
parts of the hyomandibular facet. Externally,
the surface is marked by a strong ridge
descending from below the posterior part of
the head of the bone. Posterior to this ridge is
a deep groove separating it from a short, thin
posterodorsal margin above the opercular
process. The opercular process is elongate and is
strengthened by a sharply crested central ridge
that merges with that from the posterior head of
the bone. A sharp ridge curves anterodorsally
from the main body of the bone, broadening to
the anterior head of the bone. The anterior of
the hyomandibular is expanded transversely
and longitudinally into a cup-shaped surface.
The area between the anterior ridge and the
dorsal margin was thin, as was that ventral to
the anterior ridge. The ventral limit of the main
shaft of the bone appears to broadly abut the
symplectic.

The symplectic is rarely seen but is elongate,
curved and inclined anteriorly at approximately
45° to the vertical. It fits into a deep groove in
the back of the quadrate and is partially overlain
by the preoperculum over much of its length.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Bartholomai

The quadrate is a very large, large fan-shaped
bone. Posterodorsally, the back of the bone behind
the deep notch for the symplectic, continues
the scooped surface from the hyomandibular
to accommodate the anterior base of the
preoperculum. The external ridge formed by
this, extends to just above the expanded condylar
surface. Posterior to the condyle is a notch that
probably limited the extent of the gape. The dorsal
margin of the quadrate has a cartilage connection
with the shallow groove along the ventral
edge of the metapterygoid. Anterodorsally, the
quadrate is laterally shelved to accommodate the
uppermost supramaxilla.

The metapterygoid is a significant element.
The thin posterior of the bone is broadly curved
to fit into the cup-shaped anteroventral surface
of the hyomandibular and preoperculum. It
is flexed along a line approximately 30° to the
horizontal that separates the bulk of the bone
from a thin, anterodorsal wing that overlaps
the back of the endopterygoid and part of the
anteroventral margin of the hyomandibular.
A partially separated, well-defined ridge is
occasionally present along the flexure anteriorly.

The endopterygoid extends posteriorly to
be overlapped laterally by part of the meta-
pterygoid. The bone is thin and has a broadly
convex dorsal margin. It is also shallowly curved
medially along a longitudinal line. The bone is
relatively deep, meeting theectopterygoid along
an elongated junction and apparently overlies
part of the dermopalatine. Teeth appear to have
been lacking on the endopterygoid.

The ectopterygoid is reasonably well-exposed
in the holotype. It is shallow anteriorly, with a
dorsolaterally developed process, presumably
to support infraorbital bones. The posterior
moiety of the bone is curved ventrally to meet the
quadrate along the upper two-thirds of its anterior
margin. The ventral margin of the ectopterygoid
overlies the upper margin of the dermopalatine.
The bone appears to be edentulous.

The autopalatine is ossified but appears to
be edentate. It interdigitates deeply with the
ectopterygoid. It is irregular dorsally and carries
a large articulating surface for cartilaginous
contact with the lateral ethmoid. Anteriorly,
it has a strong, elongate process for additional
articulations with the mesethmoid and maxilla.

The dermopalatine is plate-like and very elong-
ate and carries multiserial, recurved teeth
that were at least as well developed or even
larger than the largest of the teeth on the jaws
and are developed in continuity with those
of the vomer. It extends anteriorly below the
autopalatine to meet the back of the vomer.

Dermal upper jcnv. The upper jaw extends from
the tip of the snout, somewhat posterior to the
anterior of the symphysis of the lower jaws,
back to below the posterior of the orbit. The
premaxilla is small, comprising slightly less
than 14% of the total length of the upper jaw.
The upper jaw is gently convex laterally over
its posterior moiety, with the maxilla becoming
more convex anteriorly towards the snout.
Convexity increases even more around the
anterior of the premaxilla. The oral border is
gently sinuous in lateral view.

The premaxilla is relatively deep anteriorly,
with its maximum depth equal to about 80%
of its length. The bone tapers posteriorly,
meets its counterpart medially and overlies the
anterolateral base of the maxilla. The oral border
is gently transversely curved into a plate-like
structure that extends outwards beyond the
line of the bulk of the external surface. The oral
plate bears numerous, very small, multiserial,
villiform teeth (ca. 400 per square centimeter).
Nearly all have been lost n preserved specimens
and are represented only by their shallow,
round sockets.

The maxilla is elongate and relatively robust.
Measurements for its length vary from 17.5
cm. to 25.2 cm. while maximum depth ranges
from 3.1 cm to 4.0 cm. Anteriorly, the maxilla

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Revision of Flindersichthys denmeadi

is ventrolaterally grooved to accommodate the
overlying end of the premaxilla. It also extends
and curves dorsomedially and is raised into a
simple rounded head that would have abutted
the anterolateral surface of the mesethmoid.
Posterior to this articulating surface and
internally, there is an emargination of the
inner surface of the maxilla that together with
a strong, elongate but low dorsal process for
the cartilagenous attachment of the front of
the palatine. The maxilla is deepest below the
orbit (ca. 20% of its length) becoming slightly
shallower posteriorly. However, its minimal
depth occurs immediately posterior to the
palatine attachment. The dorsal margin in lateral
view is shallowly curved below the orbital area
and is grooved dorsolaterally to accommodate
the supramaxillae over the posterior moiety.
The grooving does not reach the thin posterior
margin, which is ornamented with a number
of fine, near-longitudinal ridges. As with the
premaxilla, the oral border is plate-like and
is expanded laterally along its entire length
as a slightly transversely convex platform that
provides a base for numerous, small, multi-
serial, villiform teeth, largely represented by
their sockets. These are circular and shallow
and are present in similar numbers to those on
the premaxilla. Teeth are rarely preserved but,
where present, are small to very small, conical or
slightly recurved and are enameled. Posteriorly,
the tooth plate extends slightly beyond the
lateral surface of the bone.

Two supramaxillae are present, partially
overlying each other. The anterior supramaxilla
is extended as a narrow spine to below the
middle of the orbit. Its anterior limit is slightly
overlain by a weakly developed and thin dorsal
flange from the maxilla and extends only
slightly more anteriorly than the overlying
anterior spine of the posterior supramaxilla.
The lateral face of the anterior supramaxilla
bears an irregular, longitudinal, shallow groove
that disappears ventrally. Posterodorsally, the

FIG. 7. Flindersichthys denmeadi Longman. Composite
reconstruction of medial view of posterior of left
mandible, mostly from QMF52273, approximately
natural size.

bone is extended into a point set at about 60° to
the horizontal that is ornamented dorsally for
a short distance by an area of minor serrations.
The posterior supramaxilla is a hook-shaped
element whose anterior and anteroventral
margins are slightly overlain by the anterior
supramaxilla. The bone extends anteriorly in
a narrow spine and is posterodorsally flanged
and irregularly serrated at the margin to overlie
the quadrate. The posteroventral expansion
loosely overlies the upper margin of the maxilla
and is also irregularly serrated at its margin.
The dorsal part of the bone is thin and steps up
abruptly to the much thicker ventral part. The
longitudinal base of the step is slightly grooved
and bears shallow pockets mesially.

Mandible. Regardless of the fact that the mandi-
ble protrudes in front of the premaxilla, giving
the mouth an upwardly directed, prognathous
gape, the total length of the lower jaw is only
slightly longer than that of the upper jaw. The
symphysis is relatively shallow and slopes

571

Memoirs of the Queensland Museum 1 Nature • 2010 • 55(1)

Bartholomai

posteriorly, giving the skull a rounded anterior
outline in lateral view.

The dentary forms more than two-thirds
of the mandible. The thin ventral margin is
considerably inflected. The anterior of the
dentary is markedly curved inwardly towards
the symphysis. In lateral view, the oral margin
is slightly concave and the dentary increases
gradually in depth below this margin to a
maximum depth anterior to a relatively weakly
developed coronoid process. Teeth are borne on
the surface of a transversely flattened to slightly
convex oral plate. The teeth are rarely preserved
but are represented by circular sockets of
numerous, small, multiserial, villiform teeth.
Where present, these are tiny, enamel-capped
and often slightly recurved. The oral plate
expands posteriorly and is generally extended
beyond the lateral surface of the mandible.
The plate is sometimes free of the extreme
posterodorsal surface of the dentary, indicating
presence of cartilage in young individuals.
The longitudinal base of the lateral surface of
the mandible is broadly rounded above the
well defined groove that separates this surface
from the expanded and inflected ventral flange.
The ventral flange extends backwards over the
entire length of the dentary and is deepest about
one-third the distance back from the symphysis.
At least nine large foramina for the mandi-
bular sensory canal are present within the
longitudinal groove in the dentary.

The posterior of the dentary laterally overlies
a moderately large angular. Near the poster-
oventral corner of the mandible, this has a short,
angled, sutured junction with the retroarticular.
A well-defined groove is present immediately
above the ventral margin of the angular, over-
lying the posterior of the mandibular canal. The
angular is ornamented with relatively coarse
ridges that are slightly inclined from the hori-
zontal. Dorsally, it provides the lateral rim of
the articulatory cup and extends around and
higher than the posterior of the cup, being

produced into a strong, extended, postarticular
process. Internally, the angular and articular
bones are distinct, with each contributing to
the articulatory facet. The posterior, angular
part of the facet is separated from the articular
part by a transverse, presumably cartilage-filled
fissure. The mandibular sensory canal opens
medially above the retroarticular.

The retroarticular caps the posteroventral
comer of the mandible. It extends anterolaterally
to unite with the angular and posteriorly forms
a truncated, near vertical back to the lower one-
third of the postarticular process. Internally, it
remains distinct and does not contribute to the
articulatory facet. A small nugget of bone is
present internally within the posterior surface
of the mandible on some specimens and this is
interpreted as an endosteal articular element.
However, there does not appear to be any
connection between this and the symplectic.

Circumorbital series. The circumorbital ring is
poorly preserved or absent in all specimens.
Dorsal components are rarely preserved and
individual elements are often difficult to
distinguish from underlying bones. Most,
especially those in the cheek, were very thin and
were generally distorted and fragmented during
preservation where they were preserved.

The antorbital is present meeting the anterior
base of the supraorbital. It barely reaches the
anterior margin of the orbit below the lateral
and posterior margins of the olfactory capsule.
It is sigmoidal in shape and expands ventrally,
while broadly meeting the anterior of the first
infraorbital.

The supraorbital meets the antorbital towards
the middle of the lateral margin of olfactory
capsule and extends backwards, tapering
above the anterodorsal margin of the orbit. It
is a relatively small bone that also lies along
the outer margin of the anteroventral process
of the lateral ethmoid and the posterior and
posterolateral margins of the olfactory capsule.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Revision of Flindersichthys denmeadi

FIG. 8. Flindersichthys denmeadi Longman. QMF52273, view of left side of anterior of skull showing
dermethmoid and olfactory capsule, anterior hyopalatine bones and remains of displaced gill arches,
'Dunraven' Station, nr. Hughenden, cobble bed nr. Pelican Bore, NCQ.

The anteroventral margin is ornamented with
short digitations. Posteriorly, it does not
reach the anterior of the dermosphenotic but
terminates bluntly against the frontal, leaving
the circumorbital ring incomplete.

The first infraorbital appears relatively well
developed as an elongated sub-triangular element
contributing the anterior and anteroventral
margins of the orbit. It tapers posteriorly along
its base and is ornamented by a number fine
ridges that give the surface the appearance of
a series of small, lozenge-shaped components.

The second infraorbital is not well represented
but appears sigmoidal in shape. It is a small
bone, tapered anteriorly and posteriorly, contri-
buting the middle of the ventral orbital margin
above the anterior spines of the supramaxillae.
The third infraorbital appears to have been the
largest element, elongate and deep. It has a
thickened anterodorsal margin that carries the
infraorbital sensory canal. The fourth infraorbital
is also large and provides the posterior part
of the orbital margin. The fifth infraorbital is
elongate and large. In lateral view it expands

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Bartholomai

anterodorsally, in front of and below the auto-
sphenotic ridge and provides the posterodorsal
margin of the orbit. Foramina for the infraorbital
sensory canal vary greatly in size. Those in the
first infraorbital appear very small, while
that in the third infraorbital is quite large with
the opening supported by marginal ridges.
Ornamentation of infraorbitals 3-5 is only
present over part of their surfaces and comprises
numerous, short ridges, subparallel to the upper
and lower margins, attenuated anteriorly and
thickened posteriorly. A thin covering of ganoine
is present.

The dermosphenotic is a truncated wedge-
shaped element and is inserted anterior to and
around the front of the autosphenotic ridge
and above the back of the orbit. It is penetrated
by several large pores from the infraorbital
sensory canal and has a rugose dorsal surface.
Ventrally, it has numerous, very short, pointed
projections. The anterior of the dermosphenotic
is rough but did not meet the posterior of the
supraorbital.

Hyoid arch, gill arches and gular plate. The
ceratohyal is composed of both anterior and
posterior parts, united by cartilage. The
posterior ceratohyal has not been completely
exposed in any of the prepared specimens and
evidence for the presence of an interhyal is also
lacking. The anterior ceratohyal is a very large
bone, much deeper posteriorly than anteriorly.
Posterior depth of the anterior ceratohyal varies
from 6.5 cm to 9.6 cm. It is thickened at its anterior
margin but is relatively thin posteriorly. It is
strongly fenestrated close to its dorsal margin
about half- way along its length.

The upper and lower hypohyals are separated
from one another and from the front of the
anterior ceratohyal by short gaps that would
have been filled with cartilage in life. The upper
hypohyal is much larger than the lower and its
anterior curves medially. Its anterior margin
appears thicker than its posterior margin.

1 60

Several small knobs are present anterolaterally
on its dorsal surface, presumably to lie against
the basihyal. A foramen exists on the inner
side of the upper hypohyal for the passage
of the afferent hyoidean artery, although this
foramen would normally be expected on the
dorsal surface of the bone. The lateral surface
of the upper hypohyal bears a significant
foramen close to its posterior margin. A broken
end of an oval-sectioned bone lies between the
upper parts of the two upper hypohyals. This is
probably an ossified part of the urohyal.

A plate covered with very fine rounded tooth
sockets, similar to those on the mandibular and
maxillary tooth plates, is present anterior to
the upper hypohyals. This is interpreted as a
basihyal toothplate and was possibly associated
with underlying bone, interpreted as part of the
basihyal. This has a posteroventral articu-
lation to meet the processes of the hypohyals.
The element appears to be about as long as
the hypohyals.

Gill arches are generally incomplete and
usually displaced. Those in QMF52273 are
the best yet observed and are illustrated in
Figure 8. It is likely that those exposed mostly
represent anterior elements. They include parts
of what are believed to be hypobranchials
(possibly 1 and 2). Part of an anterior end of a
ceratobranchial has a greatly deepened ventral
flange immediately behind its articulatory end,
separated from the medial body of the element
by a sharply angular ridge that strengthens
posteriorly up the face of the bone, broadening
the dorsomedial surface. The most obvious
component comprises a very large, robust epi-
branchial, probably epibranchial 1. Distally,
this presents a large, oval socket, set at a right
angle to the proximal head of the bone. It is
12 cm. long and 4.1cm. wide at its proximal
end. The uncinate process is robust, short and
only slightly separated by a shallow groove.
A strong flange defines a deep medial groove,
broadening the width of the body of the bone

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Revision of Flindersichthys denmeadi

towards its centre. Remains of a further wide
epibranchial (possibly epibranchial 2) with its
uncinate process lost and lacking its distal
moiety is preserved immediately posterior to
the possible epibranchial 1. This has an angular
anterolateral margin and is deeply 'spoon'-
shaped along the medial shaft of the bone, at
which point it is 2.2 cm. wide. A partial, stout,
short infrapharangobranchial (possibly the
first) is present above the possible epibranchial
1 and the back of the neurocranium. Numerous
disassociated tooth plates bearing sockets for
minute, multiserial, villiform teeth are widely
distributed among the gill arch elements. No
attempt has been made to reconstruct the gill
arch because of the uncertainty relating to identi-
fication of the component elements.

There are at least 14 branchiostegal rays
present on each side. The most anterior of these
are subdivided towards their tips while the
posterior rays are increasingly spathiform. At
least the first five rays of the series are attached
to the ventral margin of the anterior ceratohyal,
which is scalloped to receive them.

The gular plate is elongate, extending poster-
iorly to below the articulatory cup. It is relatively
narrow and bears broad, concentric, posteriorly
convex ridges. Its posterior margin tapers and
is marked by a central emargination.

Opercular series. The preoperculum is a large
bone, with a relatively elongate, curved lower
margin to below the middle of the quadrate and
a more gently curved posterior margin from
above the anteroventral limit of the operculum.
It has a raised anterior rim. This provides a
smoothly concave anterior face that is more
deeply developed from about the middle of
the metapterygoid to the mid-quadrate. The
raised rim fits behind the raised posterior of
the quadrate. The dorsal moiety of the anterior
rim of the preoperculum fits and supplements
the main ridge of the hyomandibular, ventral
to the level of the opercular process. The preo-

percular ridge is ornamented by short, irregular
ridges and grooves that parallel the anterior
face of the bone in the upper two-thirds of
the structure. The dorsal margin of the bone
is extended into an elongated dorsal process
lateral to the opercular process. The anterior
moiety of the bone is thickened, especially
ventrally and is often separated abruptly from
the thin posterior of the element. At least nine
foramina are present along the posterior base of
this thickened anterior rim, associated with the
preopercular sensory canal. The posterior of the
rim is gently curved back onto the dorsolateral
surface of the bone, becoming more angular
ventrally. A series of minor, irregular, ridges
radiate across the surface from the back of the
rim, opposite the level of the dorsal limit of the
quadrate.

The operculum is a large element, apparently
rounded ventrally but with a prominent post-
eroventrally inclined demarcation above the
ventral edge, set at about 40° to the horizontal.
The bone articulates with the opercular process
of the hyomandibular and this is achieved through
a short but broad, anterolaterally and slightly
dorsally directed process and facet. Below this
process, the anterior margin is stepped medially
to allow the back of the preoperculum to slightly
overlie it. A slightly raised ridge closely parallels
this step and is occasionally better developed.
This appears to carry a sensory canal that opens
through foramina. The surface of the operculum
is ornamented by a radiating series of shallow
grooves emanating from behind the articulating
process and the surface is sometimes dimpled.

The suboperculum is also a large, elongate,
deep element, extending slightly beyond the
front of the operculum. It has a curved lower
margin that is ornamented by fine ridges para-
lleling much broader and stronger, flattened
ridges that radiate from a centre about one-third
the distance from the anterior of the element.
The upper one-half of the bone is occasionally

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Bartholomai

somewhat rugose and is overlain to a reasonable
degree by the ventral edge of the operculum.

The interoperculum appears to have been a
small element.

Pectoral girdle and fin. The supratemporal is
considered here, although it is not strictly a
part of the pectoral girdle. It is an extremely
large, thin element, extending from the back of
the braincase to near the dorsal end of the supra-
cleithrum, covering much of the post-temporal.
It curves ventrolaterally to a rounded lower
margin and has a bilobate posterior margin
with the smaller upper lobe separated from the
lower by a cleft. The supratemporal commissure
is at least partially open, with several deep,
longitudinal grooves running posteriorly from
near the front of the upper moiety. Fine ridges
paralleling the lower margins are supplemented
by similar ridges radiating across the lower
surface.

The post-temporal is a smaller, flat plate of
bone that is slightly convex laterally. Its dorsal
margin is slightly convex and posteriorly curls
into a broad groove on the supracleithrum. The
middle of its anteroventral margin is extended
into a blunt, obtuse, triangular point separating
tw<5 shallow indentations. Much of the bone
underlies the supratemporal. The middle
projection of the anteroventral margin has a
large foramen ventrally associated with the
lateral line. This opens from a roofed canal that
itself opens into a groove along the margin,
towards the supracleithrum. The ventral inter-
calary limb is acutely angled ventrally and is
slightly curved ventrally but is slender, linking
the bone to the back corner of the post-temporal
fossa. A process to the epiotic process has not
been observed. Additional foramina penetrate the
lateral surface in the ventral moiety of the bone.

The supracleithrum is a very large bone that is
dorsoventrally deep and slightly convex laterally,
more so in the area of an extended posterodorsal
process that underlies the post-temporal. In

this area the bone is thin but the anterior of
the bone is much thicker. The external surface
anteroposteriorly is moderately convex and is
extended along its front margin into a broad,
convex plate that underlies the operculum and
is separated by a significantly concave upper
margin from the posterodorsal process. The
lateral line canal is externally roofed, running
obliquely across the upper part of the bone to a
posteriorly directed foramen and groove at the
back of the medial surface. The supracleithrum
is ornamented by weak, dorsoventral ridges and
a deep medial groove from near the middle of
the upper part of the lateral surface.

The cleithrum is a large element that curves
anteroventrally and that has a wavy external
surface. A large, near vertical, roofed canal
extends within a well-defined plication from a
large foramen at the ventral margin. The dorsal
margin of the cleithrum extends beneath the
ventral part of the supracleithrum.

A moderately large, thin, postcleithrum is
present behind the junction of the cleithrum
and supracleithrum. This is ornamented with
fine radiating ridges. Other post-cleithral elements
have not been observed.

The coracoid is elongate and has a slightly
concave ventral margin. It forms a sharp keel
ventrally where it meets its counterpart from
the other side. Between the coracoid and the
cleithrum there is a large interosseous foramen.
Posteriorly, the coracoid contacts the scapula
laterally. Its contact with the mesocoracoid has
not been observed.

The scapula spreads over the inner side of the
cleithrum and encloses the scapular foramen.
Support for the first of the fin rays is directly
to the endochondral girdle but radial elements
have not been preserved.

At least 15 pectoral fin rays are present but
are only known proximally. A pectoral splint
is present.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Revision of Flindersichthys denmeadi

Vertebral column. No specimen in the available
collections possesses a complete vertebral column.
It is thus impossible to assess the number of
vertebrae present. Vertebral centra are large,
near cylindrical, with anterior centra circular
in section and much shorter than they are
deep. All are amphicoelous and pierced by the
notochord. Laterally, all centra are marked by
very fine longitudinal ridges, many of which
subdivide or anastomose. Neural arches are
autogenous and each centrum bears two deep
pits dorsally to accommodate them. Anterior
neural arches are expanded and bear strong,
slender, straight neural spines that are separate
from their counterparts and that are marked
by a central groove. The centrum present as
part of the neurocranium lacks dorsal pits
and the first neural arch is not as expanded
as those following. Parapophyses are present
on all anterior vertebrae except that attached
as part of the neurocranium. These are more
ventral on the most anterior centra and are
positioned progressively more dorsally along
the abdominal part of the column.

Other skeletal elements. These are not present in
the available sample.

DISCUSSION

Taverne (1999), in describing the genus,
Arratiaeiops from the Early Cretaceous Wealden
of Europe, has analysed characters that could
be considered to determine the superordinal
position of that taxon. Six were regarded
as supportive of assigning the genus to the
Elopomorpha, these being :

(1) oral dentition borne on dental plates;

(2) retroarticular fused to angular (a principal
apomorphy of Elopomorpha, see Nelson,
1973);

(3) premaxillae articulate with the mesethmoid;

(4) parasphenoid has lost its basipterygoid
process;

(5) hyomandibuiar has a long, narrow shaft; and

(6) pectoral fin has a pectoral splint.

Of these, Flindersichthys clearly possesses
characters 1, 4, 5 and 6 and is believed to possess
character 3. To these, Forey et al. (1996) addition-
ally regarded the presence of rostral ossicles
a putative character of elopomorphs but this
character was not emphasised by Taverne
(1999). No rostral ossicle has yet been observed
in Flindersichthys .

Regarding fusion of the retroarticular and
angular. Nelson (1973) suggested that the poster-
ior of the primitive teleost mandible is comprised
of three separate bones, the angular, the articular
and the retroarticular. The Elopomorpha were
regarded as derived because of fusion of the
angular and retroarticular, with both elements
contributing to the articularatory facet.
Flindersichthys usually has the retroarticular
separate from the angular, with the former not
contributing to the jaw articulation. Only the
British Museum (Natural History) specimen
numbered P. 59694, illustrated by Taverne
(1999) and undoubtedly referable to F. denmeadi,
appears to have the retroarticular fused to the
angular without an obvious suture, but the area
of possible separation of the bones could have
been obscured because of outward rotation of
the oral margin of the mandible. Taverne (1999),
however, states in his description that the retro-
articular is fused to the angular. If this is so,
then the character may be variable, reducing
reliability of any suggestion that Flindersichthys
is not an elopomorph. It is concluded that
Flindersichthys should be included within the
Superorder Elopomorpha and that early radiation
within the Elopomorpha included taxa with a
non-fused retroarticular.

Forey et al. (1996) concluded that the Elopo-
morpha contained three orders, the Elopiformes,
the Albuliformes and the Notacanthiformes,
with the Order Elopiformes embracing two
families, the Elopidae and the Megalopidae.
The Albuliformes were believed to include
albulids, pterothrissids and possible stem-

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Bartholomai

group Cretaceous albulids ( Osmeroides and
Brannerion). Again, Taveme (1999) has provided
apomorphic character states for Arratiaelops
that serve to support referral of that genus to
the Elopiformes. Flindersichthys has generalised
morphology suggesting it also is referable to
the Order Elopiformes. Applying Taverne's
listed character states for the elopiformes,
Flindersichthys has a very prognathous mandible,
with a deep dentary; its post-temporal fossa is
large and deep; it has a large, rounded, olfactory
capsule, surrounded anteriorly by a lateral
dermethmoid process and posteriorly by the
lateral ethmoid (at least in part); however, the
mesethmoid appears to be incompletely ossified
in some individuals; parasphenoid dentition is
reduced; the supratemporal is very large; and
the parietal is penetrated by the supraorbital
sensory canal. Referral of Flindersichthys to the
Order Elopiformes is herein proposed, but with
less certainty than at the superordinal level.
Although Flindersichthys appears to super-
ficially resemble the megalopid elopids, it is
difficult to reconcile a number of its character
states with diagnoses at the familial level, as
outlined by Forey (1973). For this reason, the
genus is considered incertae sedis below the
ordinal level.

Longman (1932) observed that the holotype
of Flindersichthys denmeadi, QMF2210, had been
subjected to 'considerable distortion' with
the roof of the skull crushed down obliquely,
displacing several of the elements. The marked,
longitudinal depression in the posterior of the
neurocranial roof of the holotype, regarded
by him as one of the morphological characters
of diagnostic importance, is re-interpreted as being
emphasised by the same distortion. The sample
now available shows that nearly all specimens
referred to F. denmeadi have been distorted to
some extent, reflecting the position of the body at
the time of burial and preservation and the relative
low strength of many of the neurocranial ele-
ments. In some, e.g. QMF13714 and QMF13719,

1 64

lateral compression has resulted in the collapse
of the neurocranial roof along the mid-line,
giving the impression that the species in life
was much narrower relative to its depth
than it must have been. Other specimens (e.g.
QMF12707, QMF13715 and QMF13888) were
crushed dorsoventrally, artificially spreading the
neurocranial elements, reducing the convexity
of the skull roof and resulting in the impression
of a more fusiform body shape than the animal
had, in reality, during life. No specimen appears
completely undistorted and the reconstruction
of the skull in Figure 1 is based upon aspects
of those individuals in which distortion is
believed to have had minimal effect (e.g. the
lateral view of the holotype, together with those
of QMF13720 and QMF52274). QMF13720 is
illustrated as prepared in Figure 2.

The additional material available has per-
mitted a reassessment of the morphological
characters considered by Longman (1932) to
be of diagnostic significance in defining the
taxon. A detailed composite reconstruction of
the neurocranium in Flindersichthys denmeadi,
based upon the holotype and referred material,
is presented in Figures 3 and 4. Longman (1932)
had suggested that the parietals were not
separated posteriorly on the neurocranial roof.
However, these elements are partially separated
posteriorly by a short, anterior extension of the
supraoccipital, similar to the situation observed
by Forey (1973) in living Elops hawaiensis and
Tarpon atlanticus. However, the parietals in F.
denmeadi are usually relatively smaller bones.
The 'oblique furrows' on the dorsal surface of
the pterotic of the holotype, noted by Longman
(1932), are present on all specimens studied
but are variable in both strength and number
from specimen to specimen and extend onto
the frontal as well. The lateral spine of the
autosphenotic, with its 'oblique trough' in its
median surface (the anterior base of the dilatator
fossa), also noted by Longman (1932), is present
as a significant feature in all specimens where it

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Revision of Flindersichthys denmeadi

is preserved but again, it is somewhat variable.
Longman (1932) regarded the autosphenotic
spine in F. denmeadi as a remarkable feature
in the supraorbital region and considered it
'perhaps as distinctive as that in ... Xiplwctinus'.
Morphologically, the autosphenotic spines
in these two genera are very different, with
that in Flindersichthys being longitudinally and
obliquely elongated in lateral extent and more
appropriately described as an autosphenotic
ridge, while that in Xiphactinus is distinctly
peg-like (as depicted in Bardak, 1965).

Width of the neurocranial roof is quite variable,
reflecting the extent and direction of the crushing
that occurred during preservation, as well as the
maturity of the individual at the time of death,
together with intraspecific variation. Certainly,
Longman's (1932) estimated width of at least 75
mm across the occipital area for the holotype,
appears in error. The measure was at least 105
mm, in keeping with the autosphenotic width
of 101 mm. Longman (1932) correctly noted that
the autosphenotic width approximated that at
the occipital margin.

The bone regarded by Longman (1932) as the
suborbital plate is a part of the very expanded
metapterygoid. A similar portion of a bone ident-
ified by Longman as a post-orbital plate is also
part of the expanded metapterygoid, while
the anterior of the bone considered by him
to represent a preorbital part of the 'massive
plate of bone' exposed in the cheek area is here
shown to represent part of the endopterygoid.
A composite reconstruction of the hyopalatine
bones in F. denmeadi is provided in Figure 5.

The operculum in F. denmeadi is not sutured
to the suboperculum as suggested by Longman
(1932) but, as shown in Figures I and 2, their
relationship includes ventral overlap (below
an angled flexure on the operculum), with the
overlap covering about 50% of the surface of
the suboperculum.

Longman (1932) correctly suggested the
presence of a gular plate. However, he mistook
the flanged lower margins of the dentaries as
part of a pair of gular plates. The single plate is
narrow and elongate but is largely masked in
the holotype. The holotype also exhibits parts of
the anterior and posterior ceratohyals (the latter
identified by Longman as an epihyal). Figure 6
is a composite reconstruction. Longman further
suggested that there was a minimum of eight
branchiostegal rays on each side, whereas the
number is here shown to be at least 14. There
is evidence on vertebral centra of the presence
of a small, central perforation for a persistent
notochord, a character that was considered
absent by Longman (1932).

Circumorbital bones are very rarely pre-
served and, where present, are fragmented and
distorted and difficult to separate from under-
lying elements. Apart from those positioned
anteriorly and dorsally, they were generally
very thin bones that were not strongly attached,
although all were often lost before they were able
to be preserved in situ. Those for QMF52274 are
the best preserved but the reconstruction shown
in Figure 1 is composite. The poor state of
preservation of the base of the neurocranium,
especially of the anterior of the prootic, is possibly
attributable to thin bones as much as to crushing
during fossilisation.

The massive expansion of the orbitosphenoid
dorsoventrally and anteriorly by the ossification of
the interorbital septum produces a development
that parallels that of the ossified orbitosphenoid
septum in some albuliformes (especially osmer-
oidids and albulids). This represents a significant
morphological difference from described mega-
lopids. However, F. denmeadi is not believed
to be closely related to albuloids, differing in
many characters including possession of covered
sensory canals and large, sub-parallel post-tempo-
ral fossae; lack of a sub-epiotic fossa; presence of
a larger intercalary and prootic-intercalar bridge;

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Bartholomai

and with the quadrate/ mandibular articulation
positioned well behind the orbit.

As noted above, the posterior of the lower
jaw in Flindersichthys (Fig. 7) provides an
interesting organisation of component elements.
Nelson (1973) has reported on different patterns
of bone fusion in this area of the lower jaws in
a number of Late Mesozoic and Early Tertiary
teleosts as a guide to clarifying relationships
within the main groups. A prominent postarti-
cular process of the angular bone is shown to
be especially well developed in the Cretaceous
forms. The articulatory facet in F. denmeadi
is divided between two of the bones of the
posterior of the mandible (unlike that in the
pachyrhizodontids which Nelson records as
being single and large, reflecting the fusion
of the angular and articular, leaving the
retroarticular separate). Nelson (1973) indicates
that this arrangement in pachyrhizodontids is
reflected in a medial opening of the mandibular
sensory canal on the angular, more like that in
the Ichthyodectiformes. An enclosed mandi-
bular canal with the posterior opening posi-
tioned medially is a condition that Maisey &
Blum (1991) conclude supports the view of
Patterson & Rosen (1977) that this primitively
characterises teleosts. In modern elopiform
genera like Flops and Megalops, Nelson (1973)
notes that the posterior part of the articulatory
facet is partly separated from the articular
portion by a transverse, cartilage-filled fissure
and partly by the separation between the angular
component and the articular bone. This is
similar to the situation in F. denmeadi but the
modem elopiforms mentioned lack the unfused
retroarticular observed but possibly variable in
Flindersichthys. It is possible that the presence
of a usually separate retroarticular, medial
opening of the mandibular sensory canal on
the angular, separation of the angular and
articular parts of the articulatory facet and the
high postarticular process, as well as exclusion
of the retroarticular from the articulatory facet.

define presumably more primitive character
states retained in some lineages within the early
elopiform radiation, including that containing
Flindersichthys.

Again as noted above, Taverne (1999) in
redescribing the European Lower Cretaceous
elopiform Oligopleurus vectensis Smith Woodward,
1890 and assigning it to the new genus, Arratiae-
lops, compared it with an incomplete skull of F.
denmeadi, numbered P. 59694 from the collections
of the British Museum (Natural History). The
specimen as illustrated (Taverne, 1999, fig. 11) is
somewhat distorted, with a number of elements
redistributed. With access to a larger sample
in the current study, it is probable that some of
Taveme's morphological interpretations may not
be correct. The question of a fused retroarticular
and angular has been addressed above. The
relatively low postarticular process, as illustrated,
may also reflect outwards rotation of the oral
surface of the mandible. The identity of the
two fragments identified as infraorbitals 1 and
2 are difficult to determine from the drawing
but differ greatly from infraorbitals 1 and 2
in the Queensland Museum specimens. The
dermal upper jaw has been illustrated below
the main part of the skull, while the suggested
basibranchial is comparatively too elongate
and is most likely part of the endopterygoid.
The suggested fibrous membrane of the eye is
probably from the ossified interorbital septum.
The dermethmoid does not appear to present
a medial ridge and a pit at the raised anterior
margin to the interfrontal fontanelle. Sufficient
morphological distinctions exist, especially in
the circumorbital series, the hyopalatine series
and the internal bones of the orbital area, having
regard to Taverne's (1999) description and
illustrations for Arratiaelops, to conclude that
Flindersichthys is distinct.

The dentition and other morphological
features in F. denmeadi are not dissimilar to
those in a number of other Cretaceous teleosts,
both elopoid and albuloid. For example, Maisey

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Revision of Flindersichthys denmeadi

& Blum (1991) have shown almost identical
dentition in the large albuloid, Paraelops Silva
Santos, 1971, from the Lower Cretaceous
Santana Formation, Lagoa de Dentro, Chapado
do Araripe, Pernambuco, Brazil, South America,
described as numerous, small and conical teeth,
attached to inflated bony pads extending laterally
onto labial surfaces of maxilla, premaxilla and
dentary. The lateral expansion onto the dentary
appears to have been even greater than in F.
denmeadi. The similarity between the dentition
in such genera and the nature of their support
is considered to represent synapomorphy.

The presence of small, isolated teleost vertebrae
within the mouth in QMF13720 and the nature
of the dentition, supports the conclusion that F.
denmeadi was a predator of as yet unidentified
smaller, shallow water marine fishes. Almost all
specimens here referred have been collected from
sediments deposited close to the suggested inflow
/ outflow point for the marine transgressions
and subsequent regressions to and from the
epeiric sea of the Great Artesian Basin and the
Carpentaria Basin. This suggests that F. denmeadi
was probably not able to exist comfortably at
any great distance from the open oceans and,
as such was probably not greatly different or
was even conspecific with contemporaneous,
extracontinental elopomorphs.

LITERATURE CITED

Bardack, D. 1965. Anatomy and evolution of chirocen-
trid fishes. Paleont. Contr. Uni. Kansas, Vertebrata.
10 : 1 - 88 .

Bartholomai, A. 1969. The Lower Cretaceous elopoid
fish Pachyrhizodus marathonensis (Etheridge
Jnr.). Pp. 249-263 In Campbell, K.S.W. (ed.)
Stratigraphy and Palaeontology. Essays in Honour
of Dorothy Hill. (Australian National University
Press, Canberra).

2004. The large aspidorhynchid fish, Rkhmondichthys
sioeeti (Etheridge Jnr. and Smith Woodward, 1891)
from Albian marine deposits of Queensland.
Memoirs of the Queensland Museum 49(2): 521-36.

Burger, D. 1986. Palynology, cyclic sedimentation and
palaeoenvironments in the Late Mesozoic of the
Eromanga Basin. Pp 53-70. In Gravestock, D.
I., Moore, P.S. & Pitt, G.M., (eds). Contributions
to the Geology and Hydrocarbon Potential of the
Eromanga Basin. Geoogical Society of Australia
Special Publication, 12.

Etheridge, R. 1872. Description of Palaeozoic and
Mesozoic fossils of Queensland. Quarterly Journal
of the Geological Society, 28: 317-360.

Etheridge, R. Jnr. 1905. Description of the mutilated
cranium of a large fish from the Lower Cretaceous
of Queensland. Records of the Australian Museum
6:5-8.

Forey, P.L. 1973. A revision of the elopiform fishes,
fossil and recent. Bulletin British Museum (Natural
History), Geology, Supplement 10: 1-222.

Forey, P.L., Littlewood, D.T.J., Ritchie, P. & Mayer,
A. 19%. Interrelationships of elopomorph fishes.
Pp 175-191 In Striassny, M.J., Parenti, L. &
Johnson, G.D. (eds) Interrelationships of Fishes.
(Academic Press, San Diego)

Greenwood, P.H., Rosen, D.E., Weitzman, S.H. &
Myers, G.S. 1966. Phyletic studies of teleostean
fishes, with a provisional classification of
living forms. Bulletin American Museum Natural
History 131: 339-456.

Henderson, R.A. 2004. A mid-Cretaceous association
of shell beds and organic-rich shale: bivalve
exploitation of a nutrient-rich, anoxic sea-floor
environment. Palaios 19: 156-69.

Rear, B.P. 2007. First record of a pachycormid fish
(Actinopterygii: Pachycormiformes) from the
Lower Cretaceous of Australia. Journal. Vertebrate
Paleontology 27(4): 1034-8.

Lees, T. 1990. A probable neoteleost, Dugaldia emmelta
gen. and sp. nov. from the Lower Cretaceous of
Queensland, Australia. Memoirs of the Queensland
Museum 28(1): 79-88.

Lees, T. & Bartholomai, A. 1987. Study of a Lower
Cretaceous actinopterygian (Class Pisces) Cooyoo
australis from Queensland, Australia. Memoirs of
the Queensland Museum 25(1): 177-92.

Longman, H.A. 1932. A new Cretaceous fish. Memoirs
of the Queensland Museum 10(2): 89-97.

Maisey, J.G. & Blum, S. 1991. Pareolops Silva Santos,
1971. Pp. 238-247 In Maisey, J.G. (ed.) Santana
Fossils : An Illustrated Atlas. Contributions to IUGS
Project No. 242, the Cretaceous of South America
(T.F.H. Publications, Inc., Neptune City)

McMinn, A. & Burger, D. 1986. Palynology and
palaeoenvironment of the Toolebuc Formation.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Bartholomai

Pp 139-154. In Gravestock, D.I., Moore, P.S. &
Pitt, G.M., (eds). Contributions to the Geology
and ' Hydrocarbon Potential of the Eromanga Basin.
Geological Society of Australia Special Publication, 12.

Moore, P.S., Pitt, G.M. & Dettmann, M.E. 1986. The
Early Cretaceous Coorikiana Sandstone and
Toolebuc Formation: their relationship in the
southwestern Eromanga Basin. Pp 97-114. In
Gravestock, D.I., Moore, P.S. & Pitt, G.M., (eds).
Contributions to the Geology and Hydrocarbon
Potential of the Eromanga Basin. Geoogical Society
of Australia Special Publication, 12.

Nelson, G. 1973. Notes on the structure and relation-
ships of certain Cretaceous and Eocene teleostean
fishes. American Museum Novitates, 2524: 1-31.

Parfrey, S. 1990. Catch of the week. Queensland
Government Mining journal, June, 1990: 276.

Patterson, C. & Rosen, D.E. 1977. Review of
ichthyodectiform and other Mesozoic teleost
fishes and the theory and practice of classifying
fossils. Bulletin, American Museum of Natural
History 158(2): 81-172.

Silva Santos, R. da. 1971. Nouveau genre et epece d'
Elopidae du bassin s^d i men tai rede la Chapada
do Araripa. Amis da Academia Brasileira de Ciencias
43(2): 439-442.

Wade, M. 1993. New Kelaenida and Va m pyromorpha :
Cretaceous squid from Queensland. Memoirs
Association of Australasian Palaeontologists 15:
353-374.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A new Albian Teleost, Euroka dunravenensis
gen. et sp. nov. and a new family, Eurokidae,
from the Eromanga Basin of Queensland

Alan BARTHOLOMAI

Director Emeritus, Queensland Museum, PO Box 3300, South Brisbane Qld, 4101, Australia.

Citation: Bartholomai, A. 2010 02 15. A new Albian teleost, Euroka dunravenensis gen. et sp. nov.
and a new family, Eurokidae, from the Eromanga Basin of Queensland. Memoirs of the Queensland
Museum - Nature 55(1): 69-85. Brisbane. ISSN 0079-8835. Accepted: 16 January 2009.

ABSTRACT

The Lower Cretaceous (latest middle to late Albian) marine sediments of the Eromanga
Basin portion of the Great Artesian Basin in Queensland, Australia continue to yield
additional species of teleosts that expand knowledge of the early radiation of the
Elopomorpha. The new genus and species, Euroka dunravenensis, is described from the
Toolebuc Formation and a new family, the Eurokidae, has been created to accommodate
this very specialised, large predator that has highly adapted morphological features
especially those associated with the roof of its mouth. The parasphenoid and vomer are
very broad to compensate for its solid, bar-like anterior hyopalatine elements. Vomerine
and palatine teeth are extremely large. The lower jaw is also very strongly developed,
with its lower margin recumbent and with a solid, broad, internal shelf present to further
strengthen and accommodate the extent of the ventral penetration of the large teeth
from the roof of the mouth. It appears that circumorbital bones may have been much
reduced in number but heavy while the preoperculum was sigmoidally shaped and also
robust. The neurocranium was very broadened and flattened. The taxon shows that early
radiation of the elopomorphs was more complex than previously recognised. □ Teleost,
Albian, Eromanga Basin, Toolebuc Formation, Euroka dunravenensis, Elopomorpha,
?Albuloidei, Eurokidae.

Among collections of teleost fishes made by
the author and staff of the Queensland Museum
from the Toolebuc and Allaru Formations of
Lower Cretaceous (latest middle to late Albian)
age in the northern part of the Eromanga Basin
(a part of the Great Artesian Basin) are a number
of early representatives of the Elopomorpha.
These are being progressively prepared and are
adding significant detail to the morphology of
already identified species or are proving to repre-
sent taxa that are new to the described fauna.

Entry of oceanic waters into the epeiric sea that
covered much of inland Queensland during the
deposition of the Albian sediments was over the

basement structure known as the Euroka Arch.
This partially separated the Eromanga Basin from
the more northerly Carpentaria Basin, both parts
of the expansive Great Artesian Basin. The
Toolebuc Formation is a thin body of sediment
(much thinner than the Allaru Formation) that
is especially rich in remains of marine fossils and,
being widespread but poorly exposed, shows
evidence for some variation in its vertebrate faunal
composition from place to place. Most discoveries
occur within residual calcilutite concretions
located at the surface, as at localities above the
Euroka Arch. However, there is increasing interest
in rare exposures of the sedimentary rock itself
and in quarries being worked by local authorities

Memoirs of the Queensland Museum | Nature • 2010 • 55(1) • www.qm.qld.gov.au

Bartholomai

for road work. Discoveries of more complete
specimens of several species of fish have
been reported from the Kronosaurus Korner
Museum, Richmond, working such quarries
in its immediate vicinity in central northern
Queensland.

The crushed and somewhat distorted neuro-
cranium that is the holotype of Euroka
dunrcwenensis, was prepared by acetic acid treat-
ment from an isolated concretion, from close
to the area underlain by the Euroka Arch and
the referred posterior of a skull is from near
Richmond. Unfortunately, no post-cranial remains
have been found, other than those immediately
behind the neurocranium of the holotype and
it is to be hoped that future field work by the
Queensland Museum or by local museums in
the area will provide additional detail for this
enigmatic fish.

The author wishes to acknowledge the prep-
aration work undertaken on the holotype by
his Assistant, Ms Tempe Lees and the critical
review and input into the manuscript by Dr Alex
Cook of the Queensland Museum.

ABBREVIATIONS USED IN TEXT FIGURES

ace anterior ceratohyal

ang angular

apal autopalatine

asp autosphenotic

bo basioccipital

brr branchiostegal ray

de dermethmoid

den dentary

df dilitator fossa

dpal dermopalatine

e.com ethmoid commissure

ec P ectopterygoid

en P endopterygoid

e P° epiotic

ex ° exoccipital

exo.ic exoccipital-intercalar ridge

fahm hyomandibular facet

fm foramen magnum

fr frontal

hm hyomandibular

ic intercalar

io infraorbital

iop interoperculum

l. e lateral ethmoid

mes mesethmoid

mpt metapterygoid

m. s.c mandibular sensory canal

mx maxilla

op operculum

ors orbitosphenoid

ot.s.c otic sensory canal

pa parietal

par parasphenoid

part.p postarticular process

pop preoperculum

pro prootic

psp pterosphenoid

ptf post-temporal fossa

pto pterotic

qu quadrate

rart retroarticular

smx supramaxilla

soc supraoccipital

soc sp supraoccipital spine

sop suboperculum

so. s.c supraorbital sensory canal

suf subtemporal fossa

vhh ventral hypohyal

vo vomer

VII hm foramen for hyomandibular trunk

of facial

VII ot foramen for otic branch of facial

IX glossopharangeal foramen

X foramen for vagus

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Euroka dunravenensis gen. et sp. nov. and Eurokidae

SYSTEMATIC PALAEONTOLOGY

Order Elopiformes

Suborder ?Albuloidei

Family EUROKIDAE fam. nov.

Family Diagnosis. Elopiform fishes in which
the skull is posteriorly very broad and very
shallow. Subtemporal fossa is very weakly
developed. Snout is moderately elongated and
the mouth is terminal. Ethmoid commissure is
incomplete. Supraorbital sensory canal does not
extend onto parietal and its ridge is anteriorly
shortened. Parietals are relatively small, diamond-
shaped and separated posteriorly by supra-
occipital. Occipital condyle is formed by
basioccipital. Otic bullae insignificant. Ossified
interorbital septum to the parasphenoid is
elongated and very thickened posteriorly. Lateral
ethmoid is relatively small but complex, with its
posteromedial body firmly sutured ventrally to
lateral margin of parasphenoid and posterolateral
base abutting endopterygoid. Parasphenoid is
edentulous and very broad. Vomer is solidly
attached to the mesethmoid and locked in
laterally by vertical, reciprocal, interlocking
processes; it bears very large, near-conical teeth.
Ectopterygoid, endopterygoid, dermopalatine
and autopalatine are all very robust, united
into solid bar; both palatine elements have very
large, slightly incurved, conical teeth. Maxilla
posteriorly carries relatively small, conical teeth,
ankylosed to internal shelf and supported
laterally by lateral flange. Two supramaxillae
are present. Mandible is robust, strengthened
by large internal shelf. Dentary deep anteriorly
with ventral margin recumbent. Low coronoid
process well anterior to articulatory facet.
Mandibular teeth are larger than maxillary
but are similarly attached; two, much larger,
anterior, tusk-like teeth present. A small, robust,
retroarticular is present. Gular is well developed
and at least 20 branchiostegal rays exist on each

side, with posterior rays very spathiform and
elongated. Circumorbital series much reduced.
The preoperculum is thickened anteriorly
and sigmoidal in shape. Interoperculum and
suboperculum are elongated and deep.

Euroka gen. nov.

Generic Diagnosis. As for the species until further
taxa are defined.

Type Species. Euroka dunravenensis sp. nov.

Etymology. Named for the subsurface Euroka Arch
above and to the south of which the holotype of the
type species was collected.

Euroka dunravenensis sp. nov.

(Figs 1-8 )

Holotype. QMF13766, almost complete skull, crushed
dorsoventrally and posterolaterally with a degree
of rolling giving it an exaggerated fusiform
appearance, lacking the premaxilla, most of the
maxilla and possibly some cheek bones and part of the
posterior of the neurocranium, from bank of unnamed
tributary of Stewart Creek, "Dunraven" Station, N of
Hughenden, NCQ.

Referred Specimen. QMF12759, posterior of incomp-
lete neurocranium, showing some dorsoventral
crushing with slight anterior rotation of ventral
elements and with some dorsal shedding of surface
bone, from "Redcliffe" Station, ENE of Richmond,
NCQ.

Formation. Toolebuc Formation of latest middle to
late Albian (Lower Cretaceous) age.

Etymology. The species was named for "Dunraven"
Station, N of Hughenden, NCQ, the property from
which many fossil fish species have been collected.

Diagnosis. Large elopiform teleost with skull
in holotype exceeding 30 cm in length and over
13 cm broad posteriorly. Neurocranial roof
longitudinally flat, posteriorly broad and shallow
(5.7 cm deep) and only slightly arched across
occipital region, lacking strong ornamentation.
Dilitator fossa shallow but very broadly roofed.
Anterior of hyomandibular facet deeply pocketed
and shared by both autosphenotic and prootic.
Dermethmoid much broadened posteriorly

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Bartholomai

FIG. 1. Euroka dunravenensis gen. et sp. nov., Holotype, QMF13766, neurocranium in dorsal view.

and firmly sutured to the frontals; overlying
medial body of mesethmoid and hooked over
its anterodorsal surface. Mesethmoid complex,
supporting robust vomer. Two large vomerine
teeth present, slightly recurved and slightly
longitudinally oval in section, with anterior
exceeding 12 mm in length and posterior about
8 mm in length. Parasphenoid anteriorly very
broad. Epiotic process weak. Supraoccipital spine
relatively well developed. Angular, horizontal
flexure present across exoccipitals, reducing
sub-epiotic fossa. Subtemporal fossa almost
non-existant, with bones on side of brain
case almost horizontal. Post-temporal fossa
angled anteromedially, with reduced external
opening. Intercalar contributes minimally to
exoccipital-intercalar ridge that terminates
bluntly at exoccipital/prootic suture. Auto-
sphenotic with weak spine and with ventral

base broadly rounded; posterior face near
vertical. Pterosphenoid lobate, relatively large.
Orbitosphenoid with very thick interorbital
septum. Infraorbitals reduced in number, with
posteroventral infraorbital strong, elongated,
lozenge-shaped. Quadrate / mandibular artic-
ulation below back of orbit. Two large auto-
palatine teeth present, with anterior 8 mm
long and posterior 14 mm long, separated
from from four, large, dermopalatine teeth
by diastema. Anterior dermopalatine tooth 15
mm long and 7 mm in diameter. Palatine teeth
round in section and slightly incurved at tips.
Anterior ceratohyal elongate. Lower jaws very
large, deep, basally recumbent, strengthened by
broad internal shelf. Coronoid process low, well
separated from jaw articulation. Teeth pointed,
circular in section with tips slightly incurved
and slightly larger than those on maxilla with

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Euroka dunravenensis gen. et sp. nov. and Eurokidae

latter up to 2.8 mm long. Mandibular teeth up
to 9 mm long above rim of dentary (excluding
two, larger, anterior, tusk-like teeth). Gular
elongate, narrow anteriorly, broader and
scalloped posteriorly between numerous longi-
tudinal ridges.

Description. A large teleost probably exceeding
a metre in total length. Neurocranium of
holotype crushed dorsoventrally and postero-
laterally. Sample lacks premaxilla, most of maxilla
and, if ever present, most of the infraorbitals. The
postcranial skeleton is largely unknown. The post-
erior neurocranial surface is shallowly concave
transversely in dorsal view. The surface slopes
anteriorly above and below a broadly angled,
transverse flexure at the level of the foramen
magnum at the junction of the exoccipitals
with the epiotics and supraoccipital. The
neurocranium is shallow in the occipital region
but this still represents the deepest part of the
skull. Occipital breadth in the holotype is ca.
13 cm but depth is not able to be measured.
Occipital breadth in QMF12759 is 12.5 cm, while
its occipital depth is 5.87 cm. Postorbital length
in this specimen is 9.96 cm.

Neurocranium. The frontals and parietals make
up the bulk of the longitudinally flat neuro-
cranial roof. Transversely, the roof is only
slightly convex in the occipital region. The
frontals have strong, broad junctions with the
dermethmoid. Each frontal carries a strong,
curved, longitudinal ridge for the supraorbital
sensory canal, positioned about one-half the
distance from the medial frontal suture to the
lateral border above the orbit. Posteriorly, the
neurocranial roof is slightly depressed medially
but this is probably accentuated by crushing.
Above the orbit, the frontal is flattened across
its sutured junction with its counterpart. The
dorsal surface is only very slightly ornamented
with variable ridges that originate from centres
above the posterior of the orbit. Those more
directly linking the centres of ossification are
somewhat stronger. The anterolateral surface

bears much broader, low, radiating folds from the
same centres. The suture that unites each frontal
and parietal is difficult to locate in the holotype but
is more readily identifiable in QMF12759.

The dermethmoid in dorsal view is a large,
spool-shaped, relatively simple element, much
broader posteriorly than anteriorly, with a
waisted section above the olfactory capsule. It is
thin and anteriorly is bluntly pointed in dorsal
view. Its posterior surface is slightly depressed
medially, within low, anteriorly converging
ridges that unite above the waisted area to
form a slightly raised anteriorly surface. More
sharply defined ridges are present immediately
behind and above the olfactory capsule near
the lateral margins of the bone, separated by
shallow grooves from the inner ridges. These
disappear posteriorly. The dermethmoid is
bluntly pointed and overlies the body of the
mesethmoid, wrapping around its anterodorsal
border and contributing ventrolaterally to the
upper border of a single premaxillary facet. This
"hook" meets the vomer and is separated from
the medial dorsal convexity of the dermethmoid
by a shallow, well-defined groove that opens
posteriorly on each side, separating the two
longitudinal ridges on the bone. This groove is
partially capped by remains of thin bone and
appears to represent an incomplete, anteriorly
curved, ethmoid commissure.

The mesethmoid is a complex element
that appears to be fully ossified. Anteriorly, it
continues ventrally from its junction with the
dermethmoid, providing a shallow articulation
dorsolaterally, the premaxillary facet. Ventrally,
it meets and completes the anterior support
for the vomer. Anterolaterally, it extends
dorsally as an interlocking process, then extends
posterodorsally, to underlie the back of the
dermethmoid, meeting the upper surface of
the vomer and anterior of the parasphenoid,. It
contributes to the olfactory capsule posterventral
to an interlocking, vertical arm of the vomer.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Bartholomai

apal

brr 10 P

FIG. 2. Euroka dunravenensis gen. et sp. nov., Holotype, QMF13766, neurocranium in lateral view.

FIG. 3. Euroka dunravenensis gen. et sp. nov., Holotype, QMF13766, neurocranium in ventral view.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Euroka dunravenensis gen. et sp. nov. and Eurokidae

FIG. 4. Euroka dunravenensis gen. et sp. nov., Holotype,
QMF13766, anterior hyopalatine bones. A, lateral
view; B, dorsal view; C, medial view.

The parietal is a small bone, broader than it is
long but its sutures with surrounding bones are
difficult to define. It lies behind the posterolateral
back of a depressed posterior of the neurocrania 1
surface. A dorsal wedge of the supraoccipital
separates the parietals. A continuation of the
ricige on the frontal, above the supraorbital
sensory canal, extends towards the parietal but
does not appear to reach it. The parietal dorsal
surface is relatively unomamented.

The pterotic makes up the bulk of the post-
erolateral part of the neurocranial roof and
contributes the lateral and dorsolateral margins
of the small, post-temporal fossa. Medially, the
junction of the pterotic with the parietal is short.
The cranium is widest at the posteroventral

FIG. 5. Euroka dunravenensis gen. et sp. nov., Holotype,
QMF13766, sketch of anterior of neurocranium, lateral
view (premaxilla not preserved).

corner of the pterotics. The dilatator fossa is
shallow and very broad and is almost fully
roofed by the pterotic. The anterior of the
dilatator fossa is provided by the autosphenotic.
A large fenestra is present at the anterior of
the dilatator fossa, penetrating medially. The
pterotic also contributes the bulk of the roof of
the hyomandibular facet that is separated from
the base of the dilatator fossa by a strong, near-
horizontal ridge. The otic sensory canal is not
well defined but appears to have been covered.
Dorsomedially the junction of the pterotic and
parietal is short. The post-temporal fossa is well
developed and angled anteromedially. Within
the post-temporal fossa, the pterotic makes a
broad, transverse contact with the intercalar
but only just extends ventrally to form the
extreme posteroventral margin of the fossa. Its
longitudinal contact with the epiotic is in the
base of the fossa. Posteroventral ly, the pterotic
is expanded and has long, sutured contacts with
the intercalar and prootic.

The autosphenotic is subtriangular with a
low, anterolateral shelf. Its body is tapered
laterally and expanded at the tip into a reduced
autosphenotic spine that does not extend beyond

Memoirs of the Queensland Museum 1 Nature • 2010 • 55(1)

Bartholomai

the margin of the neurocranial roof. The base of
the spine is broadly convex anteroposteriorly
and is ornamented with low ridges. Anteriorly,
the face is penetrated by a foramen for the otic
component of nerve VII. The posteromedial
face is near vertical and forms the anterior of
the dilatator fossa.

The supraoccipital is a relatively small element
that is slightly transversely concave, positioned
medially in the dished posterior neurocranial
surface. It extends about one-half the distance
from the midline to the inner margin of the
post-temporal fossa. It is pocketed on each side
of a moderate, posteriorly directed spine, with
the pocket separated from the exoccipital by a
dorsolateral ridge. The pocket extends across
the back of the neurocranium onto the epiotic
as a weak subepiotic fossa. The supraoccipital
extends relatively broadly onto the dorsal surface
of the neurocranium to separate the posterior
of the parietals. It is strongly united with the
epiotic laterally.

The epiotic extends laterally from its junction
with the supraoccipital to near the middle of the
dorsal rim of the post-temporal fossa beyond a
muted epiotic process. It has greater expression
on the dorsal surface than the supraoccipital. Its
slightly pocketed posterior surface contributes
to the poorly developed subepiotic fossa. It
provides much of the dorsomedial margin and
wall of the post-temporal fossa.

The exoccipital is relatively large but appears
to be excluded from below the ventromedial
margin of the post-temporal fossa by the epiotic
and intercalar. It lies below the transverse flexure
of the posterior neurocranial surface and provides
the margins of the foramen magnum. It is
penetrated by the foramen for the occipital nerve.
It does not appear to form the floor of the foramen
magnum. It curves anteroventrallyclose to the
basioccipital and extends anteriorly to just
behind the back of the parasphenoid. Laterally,
it bears a low but major contribution to an

exoccipital-intercalar ridge to the prootic. The
foramen for the vagus nerve is small, beside the
posterodorsal margin of the basioccipital and its
junction with the exoccipital.

The intercalar is a relatively large element
that caps the posteroventral, inner comer of the
post-temporal fossa and extends across and
dorsally to and slightly above the ventromedial
part of the margin. It provides most of the
posteroventral neurocranial surface, meeting the
exoccipital along an elongated posterolaterally
angled ridge, broadening anteromedially and
contributing minimally to a low, exoccipital-
intercalar ridge that terminates bluntly above
the prootic-exoccipital suture. A foramen near
this appears to be the opening for the pars jugu-
laris. Laterally, the subtemporal fossa is almost
non-existant.

The basioccipital forms the occipital condyle.
Dorsally, it bears two deep pits for attachment
of a neural arch and ventrally is deeply grooved
towards the occipital condyle, at the back of
the myodome, posterior to the prongs of the
parasphenoid.

The prootic is large and complex. It is the
largest element of the lateral neurocranial wall.
Posterolaterally, it lies at only a slight angle
from the horizontal and links laterally with the
autosphenotic and pterotic, posteriorly with the
intercalar, the exoccipital and the basioccipital,
ventromedially with the exoccipital, the basio-
ccipital and parasphenoid and anteriorly with
the autosphenotic and pterosphenoid. Ante-
rolaterally, it contributes the facet for the
anterior of the hyomandibular. The shape of
the hyomandibular facet indicates that the
hyomandibular possessed a very broad head.
A large foramen close to the ascending wing
of the parasphenoid is for the hyomandibular
trunk of the facial nerve. The foramen for the
orbital artery is present anteromedially and
close to this. The anterior face of the prootic
forms much of the posterior margin of the orbit.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Euroka dunravenensis gen. et sp. nov. and Eurokidae

3cm

FIG. 6. Euroka dunravenensis gen. et sp. nov., QMF12759,
ventral view of posterior of partial neurocranium.

Crushing has masked the foramina for the
oculomotor nerve and the anterior opening of
the pars jugularis.

The pterosphenoid is moderately large,
near horizontal, being only slightly inclined
anteriorly. Its anterolateral margin is rounded.
Two small foramina are present towards the
medial margin. The surface of the bone is plicated
radially towards the lateral margin.

The basisphenoid is represented only by part
of the upper arm of its broadly Y-shaped shape.
This is angled posteriorly and separated from
much of the medial margin of the petrosphenoid
by an elongate, slit-like foramen for the optic
nerve, but this has probably been narrowed by
crushing.

The orbitosphenoid is represented only by
its posterior margin. The dorsal attachment to
the frontal is very solid and the bone has only
minor penetration on each side, just above
the level of the front of the pterosphenoid,

FIG. 7. Euroka dunravenensis gen. et sp. nov., QMF12759,
posterior view back of partial neurocranium.

possibly for passage of the olfactory tract. The
orbitosphenoid is attached to the upper surface
of the parasphenoid by an ossified interorbital
septum that is massively solid, posteriorly almost
as wide as the parasphenoid itself.

The parasphenoid is only partially known.
It is very robust, short posteriorly, with the back
produced into forked, short, thickened prongs.
The ascending wings are almost non-existant
and the foramina for the internal carotid
arteries are present near their bases. Anterior to
the ascending wings, the bone is transversely
slightly convex with an edentulous but rugose
surface. A medial foramen is present below
the back of the orbitosphenoid, possibly for
the efferent pseudobranchial artery. The
lateral margin of the parasphenoid is greatly
expanded, especially towards the front,
greatly increasing its contribution to the roof of
the mouth, compensating for the blockiness of
the anterior hyopalatine elements. The extreme
anterolateral margin is slightly deepened and
appears to have contributed to the support for
the anterior process of the dermopalatine.

The lateral ethmoid is a relatively complex
element with its laterally visible body well post-
erior to the mesethmoid, firmly uniting the base
of the frontal to the thin, upturned, lateral margin
of the parasphenoid. This mesial, posteroventral
projection together with the back of the body

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Bartholomai

of the bone constitutes the posterior of the
minor, anterior orbital margin. Anterodorsally,
the body of the bone laterally provides the
back of the olfactory capsule, while at the
posteroventral external comer, a prominent but
short process provides the articulatory surface
for the cartilaginous connection with the dorsal
process of the ectopterygoid. Separation of these
two articulatory surfaces in the holotype shows
that the backwards displacement of the dorsal
neurocranial elements relative to the lower
neurocranial elements, was of the order of 3.5
cm. Dorsally, the bone extends and underlies
the concave, lateral margin of the frontal, pro-
gressively thinning to the anterolateral edge
of the olfactory capsule. Anteromedially, it
thins towards the anterior of the ossified inter-
orbital septum and probably does not meet
its counterpart from the other side. It projects
anteriorly to the posterior of the mesethmoid,
providing additional support for the upper,
medial part of the parasphenoid.

The vomer is a very solid, unpaired bone
broadly supported anterodorsally by the base
of the mesethmoid below the hooked anterior of
the dermethmoid. Posterodorsally, it thickens
to meet the body of the mesethmoid and
laterally meets the expanded anterior margin
of the parasphenoid. It is produced into a
vertical, interlocking process into the side of the
mesethmoid, midway along the lateral margin.
Ventrally, the vomer is broadly concave along
its midline and anterolaterally bears a very
large, pointed, 12 mm long tooth on each side
in the holotype. The tooth is longitudinally
ovate in section and its tip is recurved. A some-
what smaller but otherwise similar tooth is
present on each side at the edge of the vomer
and is separated from that at the front by a long
diastema. This tooth is some 8 mm long and lies
inside the front of the parasphenoid. Again, the
left side counterpart of this tooth is not visible
in the holotype. Posteroventral ly, the vomer is
pocketed to contribute, with the base of the

front of the parasphenoid, to the facet for the
anterior process of the dermopalatine.

Hyopalatine bones. Hyopalatine bones are, in part,
incompletely preserved or are masked by overly-
ing elements. The hyomandibular is covered
by other elements but the hyomandibular facet
is elongate and subparallel to the neurocranial
roof, suggesting that the head of the hyopalatine
was broad and the bone was somewhat less
inclined anteriorly than the distortion noted in
the holotype suggests.

The quadrate is expanded and widely fan-
shaped. The articulation for the lower jaw is
transversely broad and inclined anteriorly but
this is again possibly somewhat accentuated
by posterior movement of the neurocranium
relative to the lower jaw. A deep groove that
probably accommodated the symplectic, separates
an expanded posterior margin. The extent of its
contact with the metapterygoid is masked and
the symplectic is similarly obscured. Contact
with the descending arm of the ectopterygoid
was limited.

The ectopterygoid is a deep, robust bone, with
its lateral surface slightly convex dorsoventrally
and with its dorsal surface expanding medially
and separated by an abrupt drop to the endop-
terygoid. Towards the back, it carries a sharp
ridge that crosses towards the posterior
margin. It is firmly united anteriorly with the
dermopalatine along a deeply sinuous junction.
The posterior of the ectopterygoid is broadly
attenuated and slightly flexed ventrally and
this bears a shallow, facet to accommodate
the anterodorsal edge of the quadrate. Ante-
rolaterally, the ectopterygoid extends dorsally
above the level of the dermopalatine as a strong
articulatory lateral process to contact the lateral
ethmoid.

The endopterygoid is somewhat shallower but
longer bone than the ectopterygoid. It is strongly
attached posterolaterally to the ectopterygoid
and extends dorsally across as the slightly

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Euroka dunravenensis gen. et sp. nov. and Eurokidae

FIG. 8. Euroka dunravenensis gen. et sp. nov., reconstruction of QMF12759, posterior of partial neurocranium.
A, ventral view; B, posterior view.

concave surface of the united elements to meet
the autopalatine, behind the anterior process.
This surface bears fine, irregular, longitudinal
ridges. Its medial suture with the dermopalatine
is angled posteroventrally. The bone meets the
lateral arm of the dermopalatine along a dorsal
rim that is deeper and sharper than that at the
dorsolateral contact with the ectopterygoid, with

the latter extending only to slightly below the level
of the strong process on the ectopterygoid.

The autopalatine is a short but robust nugget
of bone, forming the knob-like anterior process
as well as the anteromedial and part of the
anterior of the hyopalatine series. It is firmly
united posteriorly with the anterior surfaces of
the dermopalatine. The dorsal surface is sloped

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Bartholomai

anterodorsally and somewhat dorsomedially
as the anterior process and, while generally
flattened transversely and gently curved
longitudinally, it is marked by several small
knobs and ridges at the outer edges.The ventral
surface carries two large, pointed teeth that
are near circular in section and that have their
tips incurved and directed slightly backwards.
The anterior tooth is smaller being 8 mm long,
while the second tooth is 14 mm long.

The dermopalatine is a very large, deep,
element that forms the majority of the ante-
romedial and anterolateral parts of the solid
hyopalatine series. Laterally, it is deep and
robust and extends posteriorly to its deeply digi-
tated suture with the ectopterygoid. The
lateral surface is gently dorsoventrally convex.
Internally, the dermopalatine is attenuated from
above its dentition along its junction with the
endopterygoid. It bears at least four, extremely
large, pointed and slightly incurved teeth along
the anteroventral border, with the anterior one
ca. 15mm long and 7 mm in diameter. These
are closely placed in a single row but are
separated by a short diastema from those on
the autopalatine.

The metapterygoid is represented only by
fragments of thin bone, largely masked by
external bones of the skull.

Dermal upper jaw. The maxilla is known only
from a posterior portion of the bone on the
right side of the skull. It is relatively shallow
and extends posteriorly to just in front of the
articulation with the lower jaw. It has a narrow
and thin upper flange that thickens markedly
towards the tooth row, giving it a strongly
convex profile across its ventral moiety. It is
gently convex along its length but it appears to
have been sinuous anteriorly. It is ornamented
externally behind the tooth row with numerous,
fine ridges and grooves and, more dorsally by
numerous, very small tubercles. The back of the
maxilla is convex, without a posterior notch.

Internally, several small conical processes are
directed medially. Maxillary dentition comprises
a limited number of small, conical teeth in a
single series, ankylosed to an internal shelf and
supported laterally by a low flange. Teeth are
up to 2.8 mm long.

A small but robust part of supramaxilla 2 is
firmly attached to the posterodorsal corner of
the maxilla, over the thin upper flange. The
anterior of the bone is notched and then expands
markedly in its anterodorsal moiety to a stout,
rounded process that is directed posteromedially.
This may have provided support for the back of
the robust infraorbital positioned above. The
head of the process bears a convex articulating
surface. The anterior continuation of the flange
of the maxilla and the notch in the anterior face
of supramaxilla 2 supports the presence of a
supramaxilla 1, lost during fossilisation in the
holotype.

Mandible. The mandible is incomplete anter-
iorly but is only slightly reduced by this. It
is elongate, very robust, deep and strongly
developed with a shallow symphysis. The two
sides of the lower jaw are only moderately
divergent and the anterior is shallowly U-shaped
across the symphysis. The base of the lower jaw
is strongly inflected, especially in the area of a
lower, well-developed flange.

The dentary comprises most of the mandible.
The ventral margin is near horizontal in lateral
view but the bone itself is anteriorly more
expansive dorsoventrally than it is posteriorly.
The oral margin curves gently ventrally, anterior
to the low coronoid process, except where it
descends more abruptly at the extreme anterior
to the symphysis. It is tooth bearing over much
of its length. The dentary teeth are moderately
large, generally being much larger than those
on the maxilla. Each is round in section, with
slightly recurved tips. Remains of 12 teeth
are present, each ankylosed at its base to an
extensive, internal shelf of bone, slightly below

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Euroka dunravenensis gen. et sp. nov. and Eurokidae

a thin lateral flange of bone forming the oral
border that supports the lateral bases of the
teeth. A shallow groove exists internal to the
bases of the teeth. Teeth in the back of the row
increase gradually in size to the fourth last tooth,
with smaller teeth then present to remains of
the base of the third tooth from the front of
the row preserved in the holotype. This tooth
base and that in front of it suggest they were
much larger than all other teeth in the dentary
row and were tusk-like. In an undistorted
specimen, they would have lain outside the
large vomerine teeth. A very well developed,
internal shelf is present medial to the bases of
the teeth, curving abruptly to the lateral extent
of the interior of the inflected basal flange of the
bone. No coronoid teeth are present although
the size of the inner shelf is similar to that where
such teeth are present. The external surface of
the dentary is externally marked by a deep,
longitudinal groove that separates the upper,
swollen body of the bone from the broad, thin
flange running from the symphysis to below
the articulation. The groove is penetrated
by a number of prominent foramina for the
mandibular sensory canal. The upper one-half
of the main body of the dentary towards the
rear of the tooth row is ornamented by very fine
tubercles, becoming slightly larger posteriorly.

The angular is relatively small and deeply
sutured to the dentary. It forms the outer
part of the articulatory cup and posteriorly is
elevated into a high, strong, prominent post-
articular process. The interior of the posterior
of the mandible is not exposed.

The retroarticular is present, represented by
a small but robust knob of bone attached to
the posterior end of the mandible, with contact
with both the angular and the back of the basal,
inflected flange of the dentary.

Circumorbital series. Only one infraorbital has
been well preserved, positioned above the
posterior of the supraorbital in the holotype.

This is a robust, elongated, lozenge-shaped bone,
with the anterior more attenuated. It is generally
thicker than are most infraorbitals, ornamented
externally with low, broad ridges and has the
posterior moiety of its dorsal margin bevelled.
Below the anterodorsal margin, the bone is
thickened above an irregular, longitudinal
groove and ventral flange. This is penetrated by
small, posteroventrally directed canals and pores
associated with the enclosed, infraorbital sensory
canal. A prominent groove is present around
the posteroventral edge before being directed
posterodorsally across the lateral surface towards
the anterior of the preoperculum.

Fragmentary remains of what appears to
be other Infraorbitals and part of a possible
dermosphenotic are present, displaced to above
and behind this well preserved element.

Hyoid arch, gill arches and gular plate. The hyoid
arch is only partly exposed. Only the posterior
of a robust hypohyal element is exposed in the
holotype. Articulating with this is much of the
body of an extremely large ceratohyal. This is
thickened anteriorly and becomes thinner and
broader posteriorly, extending beyond the end
of the mandible.

Branchiostegal rays are well exposed in the
holotype, especially on the right side which
has at least 20 rays present. The anterior 14
are relatively thin and distally pointed, with
a thickened lateral surface, separated from a
thinner surface by a longitudinal groove. The
grooves diminish and almost disappear in the
more posterior of these anterior rays; and these
rays have a thickened posterior margin and a
thin flange that broadens proximally. Within the
series of branchiostegal rays, the more posterior
6 or possibly more become increasingly spathi-
form. Because of their thinning, they have been
subjected to fretting of their edges, making
it difficult to separate them. The last ray is
extremely spathiform and has ornamentation
of fine tubercles along its dorsal margin.

811

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Bartholomai

Gill arches are only represented by fragments
that are not well enough preserved to warrant
description.

The gular is present, lying between and
dorsal to the lower flanges of the dentaries in
the holotype. It is elongate, narrowly U-shaped
ventrally towards the front, broadening and
clearly scalloped posteriorly, with short, sub-
parallel, longitudinal ridges extending into
points and, with the intervening grooves,
producing the scalloped margin.

Opercular series. The preoperculum is a solid,
robust, sigmoidally-shaped bone with its sensory'
canal opening posteroventrally. It has a thickened
anterior moiety that is slightly poteriorly curved at
its dorsal tip and that is strongly curved anteriorly
into a deep, ventral point that lies along the
posterior of the quadrate. Its dorsal end lies
close to the back of the pterotic but this may
have resulted from displacement. The thickened
anterior moiety is externally roughened and
further ornamented by' three broad grooves
running subparallel to the posterior border. The
posterior moiety of the bone is thinner, curves
gently over the operculum and is more strongly
curved posteroventrally to disappear near the
anterior base of the element. Some shedding
may have occurred at the posteroventral margin.

The operculum is known only from a small,
poorly preserved portion of the bone, present
over only one-half of the posterior body depth
from above.

The interoperculum is an elongated bone
underlying the posteroventral part of the
preoperculum and extending to the back of the
lower jaw. It is thickened along its anterodorsal
margin and thins posteroventrally. Fine tubercles
are present below the anterodorsal margin.

The suboperculum is very elongate and deep,
bearing irregular and rare, shallow grooves
curving obliquely from near its anterodorsal
margin. Its depth may be exaggerated by loss

of the ventral portion of the operculum that
would have partially overlain it.

Vertebral column. Only a single anterior vertebral
centrum has been preserved. This is a simple
bone, lacking any laterally developed fossae. A
notochordal pit is present. Remains of several
neural arches are present, including one which
was probably associated with the pits in the
basioccipital, behind the foramen magnum.

DISCUSSION

The genus Euroka has superficial morpho-
logical character states that suggest it could have
affinities with one of the existing families of the
Suborder Albuloidei, either the Pterothrissidae
Gill, 1893 or the Albulidae Bleeker, 1859,
both of which have living as well as fossil
representatives. Of these, closest similarities
lie with the albulids, a family known from
the Upper Cretaceous from Europe and
North America (Forey 1973). A third family,
the Osmeroididae, was established by Forey
(1973) and was considered to possess few
of the specialisations of the albulids or ptero-
thrissids but was none the less regarded as a
link between the Elopidae at the base of the
elopiform radiation and the more specialised
albuloids. It could be argued that, because of
the limited number of specimens involved in
the establishment of E. dunravenensis, it would
be prudent to refer Euroka to the Albuloidei
incertae sedis, rather than to erect a further
family, the Eurokidae. However, the specialised
morphological character states in the new
taxon, especially those associated with the roof
of the mouth, support formal differentiation at
this time.

The Australian material is derived from the
marine Toolebuc Formation within the Eromanga
Basin part of the Great Artesian Basin, reliably
dated as Lower Cretaceous (latest mid to late
Albian) on the basis of dinoflagellate and spore-
pollen zonation (Moore et al. 1986; McMinn &

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Euroka dunravenensis gen. et sp. nov. and Eurokidae

Burger 1986). Later but less convincing dating
of the Toolebuc by Henderson (2004) based on
ammonite work suggested slight refinement
of the dating but still retained a Late Albian
age for the Formation. This indicates that early
elopomorph radiation in the Australian region
included local lineages closely contemporaneous
with the earliest of the recorded osmeroidids from
the northern hemisphere. The material presented
specialised morphology differing significantly
from the later occurring albulid taxa.

Marine transgressions into the Great Artesian
Basin were short-lived with limited access
to open oceanic waters across the basement
structure of the Euroka Arch, south of the
present Gulf of Carpentaria. This suggests
that access to new, niche opportunities in the
epeiric sea could have resulted in relatively rapid
evolutionary responses from taxa entering the
Basin, leading to endemic forms with specialised
features, departing from trends developed within
the extracontinental gene pool. On the other hand.

TABLE 1. Differences between Euroka and Albula.

Character

Euroka

Albula

Anterior cranial roof

Relatively shorter

Elongated

Dermethmoid

Posteriorly broad

Posteriorly narrow

iii

Anterior ethmoid area

Complex/blocky

Complex/ attenuated

Vomer

With very large teeth

Needle-like teeth

Parasphenoid

Very broad esp. anteriorly

Less broad anteriorly

Parasphenoid teeth

Edentulous

Many rounded teeth

vii

Cranial depth

Posteriorly very shallow

Relatively less shallow

viii

Posterior cranial width

Very wide

Relatively narrower

Post-temporal fossa

Relatively small

Relatively larger

Subepiotic fossa

Insignificant

Large

Subtemporal fossa

Almost lacking

Very large

xii

Otic bullae

Insignificant

Very inflated

xiii

Interorbital septum

Very thick

Thin

xiv

Hyopalatine series

Anterior block-like

Anterior long, thin

Palatine teeth

Extremely large

Small teeth if present

xvi

Maxillary teeth

Few posteriorly, small

Edentulous

xvii

Supramaxilla

Two

Single

xviii

Premaxilla articulation

To single facet

Double articulation

xix

Infraorbitals

Reduced, robust

Thinner, expansive

Mandible

Solid, deep, elongated

Short, tapering

xxi

Dentary base

Inflected

Not inflected

xxii

Dentary teeth

Large, anterior fangs

Fine needle-like band

xxiii

Retroarticular

Present

Absent

xxiv

Postarticular process

Prominent

Absent

XXV

Coronoid process

Low

High

xxvi

Quadrate articulation

Below orbit back

Below front of orbit

xxvii

Gular

Relatively large

Small if present

xxviii

Preoperculum

Sigmoidal, thickened

Boomerang-shaped

xxix

Interoperculum

Elongate, deep

Small

XXX

Suboperculum

Elongate, deep

Large, hooked

xxxi

Branchiostegal rays

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Bartholomai

the localities from which E. dunravenensis have
come are both close to the inflow/ outflow over
the Euroka Arch, in possibly shallower waters
to the east and southeast. For this reason,
establishment of the Family Eurokidae draws
attention to the possibly unique evolutionary
responses within this less interactive, possibly
early albuloid lineage. Certainly, pre-existing
familial and even ordinal diagnoses such as those
presented by Forey (1973), with classifications
modified as in Forey et al. (1996), would
need to be altered considerably and probably
unnecessarily to accommodate Euroka.

E. dunravenensis differs significantly when
compared with albulids, based on the living
Albula vulpes (Linnaeus 1758). These differences
are summarised in Table 1. The Lower-Middle
Eocene remains of A. oweni (Owen 1840) from
Europe and North Africa, illustrated and
discussed by Forey (1973) show even fewer
similarities with specific character states in E.
dunravenensis than are present in A. vulpes.

The enlargement of the orbitosphenoid in E.
dunravenensis by addition of a massively ossified
septum is similar to but more solidly developed
than that in the Albulidae and in Flindersichthys
Longman, 1932 and other taxa (in press) also from
the Albian marine sediments of Queensland.
However, strengthening of the orbitosphenoid
area by ossification of an interorbital septum
to contact with the parasphenoid, has been
stated to occur sporadically in other groups by
Forey (1973), who considers this character to
lack strong phylogenetic value. Strengthening
of the linkages between the parasphenoid and
the roof of the neurocranium would increase
support for the roof of the mouth. This would
assist in overcoming the stresses associated
with the bites employing the solid structure
and massive teeth of the anterior hyopalatine
bones. It is interesting to note the lack of teeth
on the parasphenoid at the back of the roof of
the mouth in this regard. This is at variance with

the general situation in albuloids where parasphe-
noid tooth development is usually considerable.

It is unfortunate that the preserved remains
of E. dunravenensis lack almost all of the post-
cranial skeleton. Certainly, the fusiform shape
of the neurocranium, including the strongly
inflected base of the lower jaw and the apparent
loss of supraorbital and most of the infraorbital
elements suggests the possibility of an eel-like
body form. The increased solidity of many of
the neurocranial bones would argue against
rapid swimming and could be associated with
a less mobile, more sedentary existence, with
the animals being secretive, awaiting passing
prey rather than actively hunting for food.
Large, gripping teeth would be advantageous
in ensuring that any prey captured from
concealment would be unlikely to escape as
they were dragged back into the possible area of
concealment. The eels of the elopioform Order
Anguilliformes are represented in the fossil
record from the Upper Cretaceous (Middle
Cenomanian) but the morphology displayed
by E. dunravenensis is generally more in keeping
with interpretation of the taxon as being more
albuliform. The relatively poor knowledge of
the fossil history of the eels, noted by Forey et
al. (1996), at this time precludes consideration of
Euroka as a possible stem group genus associated
with anguilliform evolutionary radiation within
early elopomorph radiation.

Regardless, the presence of E. dunravenensis
in the Albian fossil fish fauna of the Great
Artesian Basin indicates that early elopomorph
evolution was none the less more complex than
was believed previously.

LITERATURE CITED

Forey, P.L. 1973. A revision of the elopiform fishes,
fossil and recent. Bulletin, British Museum (Natural
History), Geology, Supplement. 10: 1-222.

Forey, P.L., Littlewood, D.T.L., Ritchie, P. & Mayers,
A. 1996. Interrelationships of elopomorph fishes
Pp. 171-195. Striassny, M.J., Parenti, L. &

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Euroka dunravenensis gen. et sp. nov. and Eurokidae

Johnson, G.D.,(eds). Interrelationships of Fishes.
(Academic Press, San Diego).

Henderson, R.A. 2004. A mid-Cretaceous association
of shell beds and organic-rich shale: bivalve
exploitation of a nutnent-rich, anoxic sea-floor
environment. Palaois. 19: 156-69.

McMiinn, A. & Burger, D. i986. Palynology and
palaeoenvironment of the Toolebuc Formation.

Moore, P.S., Pitt, G.M. & Dettmann, M. E. 1986. The
Early Cretaceous Coorikiana Sandstone and
Toolebuc Formation: their relationship in the
southwestern Eromanga Basin. Pp 97-114. In
Gravestock, D.I., Moore, P.S. & Pitt, G.M. (eds)
Contributions to the Geology and Hydrocarbon
Potential of the Eromanga Basin. Geological Society
of Australia Special Publication, 12.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Bishop

A Conchostracan from the Carboniferous Ducabrook
Formation

The Lower Carboniferous Ducabrook Formation in the
Drummond Basin, central Queensland contains a diverse
fauna including fish and tetrapod remains (Thulborn et al,
19%, Warren & Turner, 2004). The formation was deposited
in paralic and estuarine environments (Parker & Webb 2008).
Although invertebrate remains are less common than those
of vertebrates, molluscs, ostracodes and conchostracans are
known. This note describes the conchostracans from near the
Ducabrook Station, south of Bogantungan, approximately
350 km west of Rockhampton. Biostratigraphic control on the
sequence was provided by Playford (1977, 1978, 1985) who
indicated a late Visdan age for the Ducabrook Formation.

Systematic Palaeontology
Order CONCHOSTRACA Sars, 1867
Suborder SP1N1CAUDATA Linder, 1945
Superfamily CYZICOIDEA Stebbing, 1910
Family LIOESTHER1IDAE Raymond, 1946
Lioestheria Deperet & Mazeran, 1912
Lioestheria sp. (Fig. 1)

Material. QMF54809, left valve; QMF54810, right valve;
QMF54811, conjoined valves; all from QML1117, Ducabrook
Formation, Carboniferous (Visean) near Ducabrook Station,
S. of Bogantungan, central Queensland.

Description. Valves 7-8 mm long, elliptical with width
approximately two-thirds of length; umbo small, subterminal,
approximately one-fifth of the length of the valve in from point
of maximum anterior bulge. Hingeline long, occupying
the middle two thirds of the dorsal margin; valve
slightly asymmetrical about the midline transverse axis
(perpendicular to dorsal margin), with a slightly narrower
tapering to the point of maximum posterior bulge giving
the valve an 'arrowhead' appearance. Maximum ventral
bulge located approximately one-third of the length
of the valve from the anterior. Growth bands comarginal,
distinct, rugose and number 13-14 per valve (in specimens
under discussion, with larger valves having more bands),
more pronounced and closely spaced at anterior margin.
Region near umbo with one or two growth bands. Micro-
ornament not preserved.

Remarks. The material is similar to Palaeolimnadia
(Crandilimnadia) arcoensis Tasch & Jones 1979 from the Lower
Triassic Mount Goodwin Formation in the Bonaparte
Basin, northwestern Australia (see Mory 1991). It can be
distinguished by significantly fewer growth bands, and
an extended hingeline, both features absent in P. arcoensis.
Palaeolimnadia ( Crandilimnadia ) profunda Tasch & Jones
1979, also from the Mount Goodwin Formation lacks an
extended hingeline, has growth lines near parallel to
the dorsal margin for most of the valve length, unlike the
Ducabrook specimens. The Ducabrook specimens are close
to Cyzicus (Lioestheria) sp. undet. 1 Tasch 1979 from the
Carboniferous Anderson Formation in the Canning Basin
but are distinguished by being more asymmetrical about
the midline transverse axis.

FIG. 1. Cyzicus (Lioestheria) sp., A. QMF54809, left valve, scale bar = 1
mm. B. QMF54810, right valve, scale bar = 1 mm. C. QMF54811,
left and right valve, scale bar = 1 mm.

ACKNOWLEDGEMENTS

Anne Warren is thanked for providing the specimens.
Drs Alex Cook :<nd Peter Jell are thanked for their input in
the writing and editing of the manuscript.

LITERATURE CITED

Audouin, V. 1837. Communications, In Societe Entomologique de
France, Annates 5:5-516. Entomologie Bulletin, 1 ere trimester,
Session 1: ix-xi.

Deperet, C & Mazeran, P. 1912. Les Estheria du Permian d'Autun.

Societe 1 1‘Histoirc Naturalle d'Autun Bulletin 25:165-173.

Mory, A.J. 1991. Geologv of the offshore Bonaparte Basin,
northwestern Australia. Geological Survey of Western Australia
Report 29:1-47.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Conchostracan

Parker, K.E. & Webb, J.A, 2008. Estuarine deposition of a mid-Visean
tetrapod unit, Ducabrook Formation, central Queensland:
Implications for tetrapod dispersal. Australian Journal of Earth
Sciences 55: 509-530.

Playford, G. 1977. A Lower Carboniferous palynoflora from the
Drummond Basin, East-Central Queensland. Proceedings of
the Royal Society of Queensland 88. 75-81 .

1978. Lower Carboniferous spores from the Ducabrook Formation,
Drummond Basin, Queensland. Palaeontographica. Abteilung
B: 167: 105-160.

1985. Pal vnology of the Australian Lower Carboniferous; a review.
Compte Rendu - Congres International de Stratigraphie et
de Geologic du Carbonifere - International Congress on
Carboniferous Stratigraphy and Geology: 10: 247-265.

Raymond, P. 1946. The Genera of Fossil Conchostraca-an order of
bivalved Crustacea: Journal of Paleontology 30: 1248-1257.

Tasch, P. 1979. Permian and Triassic Conchostraca from the Bowen
Basin (with a note on a Carboniferous leaiid from the Drummond
Basin), Queensland. Bureau of Mineral Resources, Geology and
Geophysics Bulletin 185: 31-43.

Tasch, P. & Jones, P.J. 1979. Permian and Triassic Conchostraca
from the Bowen Basin (with a note on a Carboniferous leaiid
fro the Drummond Basin), Queensland. Bureau of Mineral
Resources, Geology and Geophysics Bulletin 185: 31-43.

Thulborn, T., Warren, A., Turner, S. & Hamley, T. 1996. Early
Carboniferous tetrapods in Australia. Nature 381. 777-780.

Warren, A. & Turner, S. 2004. The first stem tetrapod from the Lower
Carboniferous of Gondwana. Palaeontology 47:151-184.

Peter J. Bishop, Queensland Museum, Geosciences, 122
Gerler Rd, Hendra Q. 4011, Australia. 3 December 2009.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Distribution of the estuary stingray
(Dasyatis fluviorum ) in Australia

Simon J. PIERCE
Michael B. BENNETT

School of Biomedical Sciences, The University of Queensland, St Lucia, Qld 4072, Australia. Email:
simon@giantfish.org

Citation: Pierce, S.J. & Bennett, M.B. 2010 02 15. Distribution of the estuary stingray (Dasyatis
fluviorum ) in Australia. Memoirs of the Queensland Museum — Nature 55(1): 89-97. Brisbane. ISSN
0079-8835. Accepted: 19 August 2009.

ABSTRACT

The extant distribution of Dasyatis fluviorum Ogilby, 1908 is confined to rivers, estuaries
and near-shore marine waters along approximately 1700 km of the eastern Australian
coastline from central New South Wales (34° 4 ' S) northwards to central Queensland
(20° 37 ' S). Records of D. fluviorum from northern Australia, Papua New Guinea and
eastern Indonesia appear to be misidentifications of Dasyatis sp. 1 (as described by
White et al. 2006). An analysis of historical records shows no evidence of a contraction
in distribution in south-eastern Australia as had previously been reported. However, the
preferred habitats of D. fluviorum are heavily impacted by human activities throughout
much of its range. □ Biogeography, Chondrichthyes, Myliobatiformes, elasmobranch,
distribution, conservation.

Six of the seven freshwater and estuarine-
specialist elasmobranch species in Australian
waters (Last 2002) are listed as globally threatened
on the 2007 IUCN Red List of Threatened Species
(1UCN 2007). Little is known about the elasmo-
branch species inhabiting estuarine and mangrove
ecosystems, though these habitats have suffered
extensive modification and degradation from
anthropogenic processes worldwide (Halpern et
al. 2007; Lotze et al. 2006), and the sharks and rays
that prefer to inhabit these environments may
have suffered concomitant population declines
(Compagno 2002; Last 2002). This situation is
exemplified by the estuary stingray ( Dasyatis
fluviorum) Ogiiby, 1908.

Dasyatis fluviorum is known from a number
of rivers, estuaries and adjacent coastal waters
on the highly-urbanised eastern coast of
Australia (Last 2002). These areas are subject to
a variety of anthropogenic pressures including
urbanisation, habitat degradation, aquaculture

and major fisheries, all of which are likely
to impact on this species (Kyne et al. 2003;
Pogonoski et al. 2002). The estuary stingray
was listed as Vulnerable on the 2003 Red List
of Threatened Species (Kyne et al. 2003) based
on an apparent range contraction along the
New South Wales (NSW) coast and anecdotal
reports of population declines over the species'
range (Last 2002; Last & Stevens 1994; Kyne et
al. 2003; Pogonoski et al. 2002). The extant
distribution of this poorly-known species has
remained unclear (Last 2002; Pogonoski et al.
2002). Here we review all published literature,
collate museum collection records and summarise
unpublished sighting records to delineate the
historical and extant distribution of the estuary
stingray in Australasia.

MATERIALS AND METHODS

All available references and sighting records
for D. fluviorum were located in the peer-

Memoirs of the Queensland Museum | Nature • 2009 • 55(1) • www.qm.qld.gov.au

Pierce & Bennett

FIG. 1. Nominal Dasyatis sp. 1 (White et al., 2006)
specimens from Indonesia and northern Australia.
A, ANFC specimen H 5286-01 caught in the Kamora
River estuary, Papua; B, NTM specimen S.14424-011
caught at Pocock's Beach in the Kakadu National
Park; C, NTM S.15930-008 caught in the Keep River
estuary, western Northern Territory. Scale bars
represent 50 mm.

reviewed and grey literature through searches
of published references, CrossSearch (ISI Web
of Knowledge) databases and World Wide Web
content using internet search engines. Details
of museum and other major collection records
were obtained through the Online Zoological
Collections of Australian Museums (http://
www.ozcam.gov.au) which included sightings
and records from the Australian Museum
(AM), the Australian National Fish Collection
(ANFC) and the Museum and Art Gallery of
the Northern Territory (NTM). Records from
the Queensland Museum (QM) were obtained
directly from that institution. Unpublished
observations and sighting records were added
from interviews with experienced observers.
Although data collation focused on positive
reports of D. fluviorum presence, other major
surveys of potential or nominal D. fluviorum
habitat that did not record the species from
certain regions were also noted.

RESULTS

MELANESIA AND NORTHERN AUSTRALIA

Pre-2000 records of D. fluviorum from the
coasts of eastern Indonesia (Fig. 1 A) and Papua
New Guinea (PNG) (Kailola 1975, 1987) and
more recent observations from the Northern
Territory in Australia are likely to have been
misidentifications of an undescribed species
currently known as the Merauke stingray,
Dasyatis sp. 1 (White et al. 2006). This species is
similarly-coloured to D. fluviorum and occupies
a similar near-shore and estuarine habitat
(White et al. 2006). Photographs of a specimen
that was initially identified as D. fluviorum
(H. Larson pers. comm. Fig. IB) caught from
Pocock's Beach, West Alligator Head, Kakadu
National Park, show a long tail and well-
developed row of medial thorns. These features
are characteristic of Dasyatis sp. 1 (White et
al. 2006), but are not present in similarly-
sized D. fluviorum from Moreton Bay (S. Pierce
unpub. data). Another, larger specimen from
the Keep River estuary in the western Northern

Memoirs of the Queensland Museum | Nature • 2009 • 55(1)

Distribution of Dasyatis fluviorum

Territory (H. Larson, pers. comm.. Fig. 1C), has
a similar morphology and these two specin ens
are provisionally assigned to Dasyatis sp. 1

A number of pre-1995 records from the Nort lem
Territory have lx;en shown to be misidentificat ons
of the freshwater whipray Himantura dalyt nsis
(Last & Manjaji-Matsumoto 2008) basec on
descriptions and photographs (Last 2 )02;
Thorburn et al., 2003). Prior to this species b -ing
identified from the region. Stead (1963) st ited
that D. fluviorum had been recorded, ur der
synonyms, at Port Darwin in 1878 and 1881 and
from the Arafura Sea in 1880. Pollard (1174)
caught a single specimen, listed as "Dasyatis
sp. (fluviorum?)" , from the East Alligator River
and noted that further captures had been made
from an adjacent freshwater lagoon. Dasyatis
fluviorum were not recorded during freshwater
elasmobranch surveys in 2002 (Thorburn et al.,
2003) or among bycatch of the Northern Prawn
Fishery (Stobutzki et al. 2002).

Records of D. fluviorum from Darnley Island in
the Torres Strait proved to be misidentifications of
Neohygon kuhlii (Muller and Henle, 1841). Dasyatis
fluviorum has not been recorded from fisheries
catches in the Gulf of Carpentaria (Blaber et al.
1994; Stobutzki et al. 2002) or from fisheries-
independent studies in northern Queensland
(Blaber 1990; Blaber et al. 1995). Herbert et
al. (1995) conducted a major freshwater fish
survey of the Cape York Peninsula region and
recorded stingrays from several rivers, listing
species as either "Dasyatis sp." or “Dasyatis
sp. (fluviorum?)" . A photograph in a resulting
publication (Herbert and Peeters, 1995) listed
as D. fluviorum is actually H. dalyensis. A more
recent survey of freshwater and estuarine
elasmobranchs on the peninsula did not record
D. fluviorum, though H. dalyensis was present in
the region (Thorburn et al. 2003).

Material. Kamora River estuary, Panua, ANFC H 5286-
01 (photograph); Pocock's Beacn, West Alligator
I lead NTM S.14424-011 (photograph); NTM S.15930-
008 (photograph); Darnley Island, Torres Strait, QM I
1519, 1 1520 (both examined).

CENTRAL AND SOUTHERN QUEENSLAND

The confirmed distribution of D. fluviorum
extends south from Repulse Bay in central
Queensland (20° 37' S; Fig. 2). There are records
of the species from South Beach near Mackay,
Warginburra Peninsula north of Yeppoon, Port
Curtis near Gladstone (Connolly et al. 2006),
Hervev Bay (A. Gutteridge, pers. comm.), Mary
River (McPhee & Skilleter 2005), Noosa River
(J. Johnson pers. comm.) and the Maroochy
and Mooloolah rivers (Schlacher et al. 2005)
on the Sunshine Coast. The species is common
within Moreton Bay (Johnson, 1999; Pillans et
al. 2007) and has been reported to occur in the
Brisbane River "above tidal influence" (Ogilby
1908; Stead 1963). Dasyatis fluviorum has also
been recorded from the Gold Coast Seaway
(J. Johnson pers. comm.), Coombabah Creek
(Ross 1999), Nerang River (Morton 1989) and
Tallebudgera Creek (Morton 1992) on the Gold
Coast in southern Queensland.

Material. Repulse Bay, ANFC H 4421-01 (site record);
South Beach, Mackay, ANFC H 4595-01 (site record);
Warginburra Peninsula, AM 134333008 (site record);
Moreton Bay, QM 1 26914, 1 30175, I 627, ANFC H
5769-01 (site records); Brisbane River, QM 1 1530, I
900, 1 2101, 1 11928, 1 235, 1 7720 (site records).

NEW SOUTH WALES

The majority of NSW records of D. fluviorum
were from rivers and harbours. In northern
NSW, D. fluviorum has been reported from
Belongil Creek near Byron Bay (Parker 1999),
from between 20 and 28 m depth off the coast
east of Iluka, from the Clarence River (Kroon
& Ansell 2006) where it has been recorded "at
least as far inland as Grafton" (Stead 1963), and
the Macleay (Gibbs et al. 1999; Macbeth et al.
2002), Manning (Gibbs et al. 1999), Wallamba
(Gibbs et al., 1999), Hunter (Gibbs et al. 1999;
Ruello 1976) and Hawkesbury Rivers (Gray
1990). There were records of D. fluviorum from
Port Jackson between 1881 and 1914, and the
species was caught from the Parramatta River
and Sydney Harbour between 1975 and 1984
(P. Gibbs pers. comm.). Nine D. fluviorum were

Memoirs of the Queensland Museum | Nature • 2009 • 55(1)

911

Pierce & Bennett

caught in Botany Bay fish surveys in 1978 and
1979 (SPCC 1981) and one was caught in April
2000 (Williams et al. 2004). Dasyatis fluviorum
was listed on a checklist of fishes recorded from
the Hacking River (34° 4' S; Pease & Herbert
2002), and this was the southern-most record
of the species.

Material. East of Iluka, ANFC H 4170-01 (site
record); Clarence River, ANFC H 5964-01, FI 5964-
02 (site records); Hawkesbury River, AM 114625,
119951001 (site records); Port Jackson, AM B8395,
113456 (site records).

HABITAT PREFERENCES

Dasyatis fluviorum has been most commonly
recorded from shallow coastal waters, particu-
larly over mangrove-fringed sand/ mud intertidal
flats in sheltered bays and estuarine areas. The
maximum depth record was from between
20 and 28 m depth offshore. The species may
penetrate rivers to pure freshwater (Stead,
1963), though salinity preferences of 14 - 35
were noted in the Hunter River (Ruello, 1976).
Mean monthly surface temperatures at the
north of D. fluviorum' s distribution (Repulse
Bay) range from an estimated 24°C in July to
29°C in January, while the temperature ranges
from approximately 17°C in August to 23°C in
February near the southern extent of distribution
in Botany Bay (Department of Defence, 2007).

DISCUSSION

The results of this study indicate that Dasyatis
fluviorum is an eastern Australian endemic,
restricted to approximately 1700 km of the
eastern coast from the Hacking River in NSW (34°
4' S) in the south to Repulse Bay in Queensland
(20° 37' S) in the north. Considerable sampling
effort in tropical north Queensland has failed
to record the species (J. Johnson, pers. comm.).
The centre of abundance for D. fluviorum appears
likely to be the sheltered bays and estuaries of
southern Queensland based on habitat prefer-
ences, distribution data and sighting records.

Moreton Bay and Hervey Bay represent two
likely population centres (Kyne et al. 2003).

The provisional identification of Dasyatis sp.

1 from the Northern Territory is the first record
of this species from Australia. Dasyatis sp. 1 is
not readily distinguishable from D. fluviorum
(W. White pers. comm.) and is currently
known solely from several juvenile specimens
from Papua (White et al. 2006). The distinct
taxonomic status of Dasyatis sp. 1 and the
specific identity of Northern Territory Dasyatis
cf. sp. 1 requires confirmation, as the results of the
present study indicate that the extant distribution
of D. fluviorum is substantially smaller (>50%)
than was previously recognised (Last & Stevens
1994).

Recent (post-2000) observations of D. fluviorum
from the central NSW coast, estuaries and rivers
indicate that the southern extent of this species
has not contracted northwards in recent decades.
This hypothesised range contraction (Last &
Stevens 1994; Pogonoski et al. 2002) was used
as partial justification for listing D. fluviorum as
Vulnerable on the IUCN Red List of Threatened
Species (Kyne et al. 2003). The revised distri-
bution for the species presented here does
not alter the current listing of the species as
Vulnerable. In fact, the relatively small range
of D. fluviorum along the east coast may also
add criteria Blab as an additional risk factor,
based on (1) an extent of occurrence of <20000
km 2 , (2) possible population fragmentation
because of specialised habitat preferences and
(3) continued decline in habitat quality in many
of the locations where the species is known to
occur (IUCN 2001).

Records of D. fluviorum are predominantly
from tidal rivers, estuaries, sheltered harbours,
and only occasionally the adjacent coastline.
The extent of this species' interchange
between these habitats is unknown. Few data
are available on the movement patterns of
demersal rays. Short-term acoustic tracking
studies on D. lata (Garman 1880) (Cartamil

Memoirs of the Queensland Museum | Nature • 2009 • 55(1)

Distribution of Dasyatis fluviorum

et al, 2003) and Urobatis halleri (Cooper 1863)
(Vaudo and Lowe, 2006) have shown low rates
of movement, while longer-term (months to years)
tracking and tagging studies on D. brevicaudata
(Hutton 1875) (Le Port et al. 2008), Neotrygon
kuhlii (Pierce et al. 2009) and U. halleri (Vaudo
& Lowe 2006) have found most individuals to
be relatively site-resident. Studies of D. sabina
(Lesueur 1824) in coastal lagoons found most
individuals to be present year-round (Snelson
et al. 1988), though two tagged specimens
Were recaptured 80 and 97 km from the point
of tagging along the eastern coast of the
United States (Schwartz & Dahlberg 1978).
It is unknown how the relatively specialised
habitat preferences of D. fluviorum affect its
dispersal ability, although the single capture of
a specimen in 20-28 m depth near the Clarence
River mouth may imply that the species is
capable of coastal movements. As habitat
isolation can increase the overall extinction
risk for a species (Reed 2004), determining the
extent of interchange between what may be
relatively isolated D. fluviorum populations is
an important topic for future research.

The results of the current survey indicate that
U. fluviorum faces several contemporary extrinsic
threatening processes. Several publications list
stingrays as bycatch of near-shore commercial
fisheries along the northern and central NSW
coasts. These records are ambiguous, recording
species as "Dasyatididae/ Urolophidae (mixed
S PP')” (Liggins et al. 1996; Liggins and Kennedy,
19 96), "Dasyatis sp." (Gray 2002), "Dasyatis
S P- (estuary stingray)" (Gray et al. 2003) or "D.
U'etidis (estuary stingray)" (Gray et al. 2001).
However, these areas lie within the revised
distribution of D. fluviorum and the species is
rikely to have been a component of these catches.
Commercial prawn trawling (Gray et al. 1990;
M cPhee & Skilleter 2005; Ruello 1976) and
estuarine gill net fisheries (Gray 2002), as well
* s re creational angling (Steffe et al. 2007; West
^ Gordon 1994) occurs in most known D.
/ l <viorum habitats throughout its distribution.

Habitat modification and degradation are also
likely to affect D. fluviorum. Many large rivers
have been modified through the construction of
flood mitigation barriers, which can periodically
lead to influxes of water with a low dissolved
oxygen content and low pH (Gibbs et al.
1999; Kroon & Ansell 2006) causing fish kills
(Macbeth et al. 2002). Occasional fish kills, from
a variety of causes, have been noted in many

93|

Memoirs of the Queensland Museum | Nature • 2009 • 55(1)

Pierce & Bennett

known D. fluviorum habitats (Ruello 1976; Steffe
et al. 2007). Many areas are also affected by
pollutants (Ruello 1976; Williams et al. 2004),
such as excess nitrogen loads (Schlacher et al.
2005) or organochlorines (Birch & Taylor
2000). Potential habitats have also been modified
for residential developments such as canal estates
(Morton 1992) or commercial industries such
as ports (Connolly et al. 2006). Limited data on
the use of modified habitats by elasmobranchs
in southern Queensland (Morton 1989, 1992;
Ross 1999) have not shown D. fluviorum to utilise
residential canal estate habitats, although the
species was present in adjacent unmodified areas.
The majority of known D. fluviorum habitats,
and particularly large rivers and harbour areas,
are subject to multiple potential threatening
processes (i.e. Birch & Taylor 2000; Johnson
1999; Ruello 1976).

Although anecdotal evidence suggests signi-
ficant population decline overall (Kyne et al.,
2003), D. fluviorum has been shown to persist
in some highly-modified habitats for long time
periods (> 100 years in some cases). It is difficult
to ascertain the contemporary abundance of D.
fluviorum in these habitats. Poor water quality
or pollutants has been linked with endocrine
dysfunction, health disturbances and
decreased reproductive health in elasmobranch
populations (Gelsleichter et al. 2005, 2006),
which can potentially lead to reduced rates of
population growth (Cortes & Parsons 1996).
Habitat. degradation has also been linked to
increased mortality in juvenile sharks and
a possible long-term decrease in carrying
capacity (Jennings et al. 2008). While these factors
have not been examined in D. fluviorum, it
is possible that reductions in the quality
and quantity of habitat have contributed to the
species' population decline. Large stingrays are
relatively difficult fish to sample because their
size, benthic habit and general anatomy makes
them difficult to enmesh in nets. Evaluation
of their current status and abundance would
be assisted by species-specific data collection

during trawl surveys in deeper riverine waters
and seine-netting in intertidal areas.

The revised distribution of D. fluviorum
along the eastern coast overlaps with a large
proportion of the most urbanised and modified
coastal areas in Australia. The reduction, mod-
ification and degradation of inshore, and
particularly estuarine habitats are likely to
directly or indirectly impact upon D. fluviorum
populations throughout much of this range.
The species is also impacted by a variety of
threats, such as commercial and recreational
fisheries, that are known to directly reduce
ray populations. This poorly-known species is
worthy of increased attention from conservation
managers.

ACKNOWLEDGEMENTS

Thanks to Helen Larson for information on
NTM specimens and editorial comments. Phillip
Gibbs provided extremely useful information
on NSW records. Thanks also to Patricia Kailola,
Jeffrey Johnson, Adrian Gutteridge, Stephen
Taylor, Peter Kyne, William White, Colin
Simpfendorfer and Charlie Huveneers for
contributing additional records and comments.
This research was supported by The University
of Queensland, Project AW ARE Asia-Pacific
and the Tangalooma Marine Education and
Research Foundation. Thanks also to one
anonymous reviewer for helpful comments on
this manuscript.

LITERATURE CITED

Birch, G.F. & Taylor, S.E. 2000. Distribution and
possible sources of organochlorine residues in
sediments of a large urban estuary. Port Jackson,
Sydney. Australian ]ournal of Earth Sciences 47:
749-756.

Blaber, S.J.M., Brewer, D.T. & Harris, A.N. 1994.
Distribution, biomass and community structure
of demersal fishes of the Gulf of Carpentaria,
Australia. Australian Journal of Marine and
Freshwater Research 45: 375-396.

Blaber, S.J. M„ Brewer, D.T. & Sal ini, J.P. 1995. Fish
communities and the nursery role of the shallow

Memoirs of the Queensland Museum | Nature • 2009 • 55(1)

Distribution of Dasyatis fluviorum

inshore waters of a tropical bay in the Gulf of
Carpentaria, Australia. Estuarine, Coastal and
Shelf Science 40: 177-193.

Blaber, S.J.M., Brewer, D.T., Salini, J.P. & Kerr, J.
1990. Biomasses, catch rates and abundances of
demersal fishes, particularly predators of prawns,
in a tropical bay in the Gulf of Carpentaria,
Australia. Marine Biology 107: 397-408.

Cartamil, D.P., Vaudo, J.J., Lowe, C.G., Wetherbee,
B.M. & Holland, K.N. 2003. Diel movement
patterns of the Hawaiian stingray, Dasyatis lata:
implications for ecological interactions between
sympatric elasmobranch species. Marine Biology
142: 841-847.

Compagno, L.J.V. 2002. Freshwater and estuarine
elasmobranch surveys in the Indo-Pacific region:
threats, distribution and speciation. Pp. 168-
180. In Fowler, S. L., Reed, T. M. & Dipper, F.
A. (eds) Elasmobranch biodiversity, conservation
and management. Proceedings of the international
seminar and workshop, Sabah, Malaysia, July 1997.
Vol. 25. Occasional Papers of the IUCN Species
Survival Commission. (IUCN: Gland and
Cambridge).

Connolly, R.M., Currie, D.R., Danaher, K.F., Dunning,
M., Melzer, A., Platten, J. R., Shearer, D., Stratford,
P. j., Teasdale, P. R. & Vandergragt, M. 2006.
Intertidal wetlands of Port Curtis: ecological patterns
and processes, and their implications. Technical
Report No. 43. (CRC for Coastal Zone, Estuary
and Waterway Management: Brisbane).

Cortes, E. & Parsons, G.R. 1996. Comparative
demography of two populations of the bonnet-
head shark (Sphyrna tiburo). Canadian Journal of
Fisheries and Aquatic Sciences 53: 709-718.

department of Defence. 2007. Surface temperature
search tool. (Database, Directorate of
Oceanography and Meteorology, Department
of Defence, Commonwealth of Australia).
Available from: http:/ / www.metoc.gov.au.

Gelsleichter, J„ Manire, C.A., Szabo, N.J., Cortes, E.,
Carlson, J. & Lombardi-Carlson, L. 2005.
Organochlorine concentrations in bonnethead
sharks ( Sphyrna tiburo) from four Florida estuaries.
Archives of Environmental Contamination and
toxicology 48: 474-483.

Gelsleichter, J., Walsh, C.J., Szabo, N.J. & Rasmussen,
L. E. L. 2006. Organochlorine concentrations,
reproductive physiology, and immune function
in unique populations of freshwater Atlantic
stingrays ( Dasyatis sabina) from Florida's St.
Johns River. Chemosphere 63: 1506-1522.

Gibbs, P., McVea, T. & Louden, B. 1999. Utilisation of
restored wetlands by fish and invertebrates. NSW
Fisheries Final Report Series No. 16. (NSW
Fisheries Office of Conservation: Cronulla).

Gray, C.A. 2002. Management implications of
discarding in an estuarine multi-species gill net
fishery. Fisheries Research 56: 177-192.

Gray, C.A., Kennelly, S.J. & Hodgson, K. E. 2003.
Low levels of bycatch from estuarine prawn
seining in New South Wales, Australia. Fisheries
Research 64: 37-54.

Gray, C.A., Kennelly, S.J., Hodgson, K.E., Ashby, C.J.
T. & Beatson, M. L. 2001. Retained and discarded
catches from commercial beach-seining in Botany
Bay, Australia. Fisheries Research 50: 205-219.

Gray, C.A., McDonall, V.C., & Reid, D.D. 1990. By-
catch from prawn trawling in the Hawkesbury
River, New South Wales: species composition,
distribution and abundance. Australian Journal
of Marine and Freshivater Research 41: 13-26.

Halpem, B.S., Selkoe, K. A., Micheli, F. & Kappel, C.V.
2007. Evaluating and ranking the vulnerability
of global marine ecosystems to anthropogenic
threats. Conservation Biology 21: 1301-1315.

Herbert, B. & Peeters, J. 1995. Freshwater fishes of far
north Queensland. (Queensland Department of
Primary Industries: Brisbane).

Herbert, B.W., Peeters, J.A., Graham, P.A. & Hogan,
A.E. 1995. Freshwater fish and aquatic habitat survey
of Cape York Peninsula. Cape York Peninsula Land
Use Strategy. (Office of the Co-ordinator General
of Queensland: Brisbane, Department of the
Environment, Sport and Territories: Canberra,
Queensland Department of Primary Industries:
Brisbane).

IUCN. 2001. IUCN Red List Categories and Criteria:
Version 3.1. (IUCN: Gland and Cambridge).

IUCN. 2007. 2007 IUCN Red list of threatened species.
(Database, IUCN). Available from: http:/ /
www.i ucnredlist.org.

Jennings, D.E., Gruber, S.H., Franks, B.R., Kessel, S.
T. & Robertson, A.L. 2008. Effects of large-scale
anthropogenic development on juvenile lemon
shark ( Negaprion brevirostris) populations of
Bimini, Bahamas. Environmental Biology of Fishes
83: 369-377.

Johnson, J.W. 1999. Annotated checklist of the fishes of
Moreton Bay, Queensland, Australia. Memoirs
of the Queensland Museum 43: 709-762.

Kailola, P.J. 1975. A catalogue of the fish reference
collection at the Kanudi fisheries research laboratory,
Port Moresby. Research Bulletin No. 16.

Memoirs of the Queensland Museum | Nature • 2009 • 55(1) 95

Pierce & Bennett

(Department of Agriculture, Stock & Fisheries:
Port Moresby).

1987. The fishes of Papua New Guinea. A revised
and annotated checklist. Vol. 1. Myxinidae to
Synbranchidae. Research Bulletin No. 41.
(Department of Fisheries and Marine Resources:
Port Moresby).

Kroon, F.J. & Ansell, D.H. 2006. A comparison of
species assemblages between drainage systems
with and without floodgates: implications for
coastal floodplain management. Canadian Journal
of Fisheries and Aquatic Sciences 63: 2400-2417.

Kyne, P.M., Pollard, D.A. & Bennett, M.B. 2003.
Dasyatis fluviorum. (Database, 2007 1UCN Red
list of threatened species). Available from:
http:/ / www.iucnredlist.org/search/details.
php/ 41797/summ.

Larson, H.K. 1999. Report to Parks Australia North,
on the estuarine fish inventory of Kakadu National
Park, Northern Territory, Australia. (Museums and
Art Galleries of the Northern Territory: Darwin).

Last, P.R. 2002. Freshwater and estuarine elasmo-
branchs of Australia. Pp. 185-193. In Fowler, S. L.,
Reed, T. M. & Dipper, F. A. (eds) Elasmobranch
biodiversity, conservation and management.
Proceedings of the international seminar and
workshop, Sabah, Malaysia, July 1997. Vol. 25.
Occasional Papers of the IUCN Species Survival
Commission. (IUCN: Gland and Cambridge).

Last, P.R. & Manjaji-Matsumoto, B.M. 2008. Himcmtura
dalyensis sp. nov., a new estuarine whipray
(Myliobatoidei: Dasyatidae) from northern
Australia. Pp. 283-291. In Last, P. R„ White, W. T. &
Pogonoski, J. J. (eds) Descriptions of neio Australian
chondrichthyans. CSIRO Marine and Atmospheric
Research Paper No. 022. (CSIRO Marine and
Atmospheric Research: Hobart).

Last, P.R. & Stevens, J.D. 1994. Sharks and rays of
Auslralia. (CSIRO: Australia).

Le Port, A., Sippel, T. & Montgomery, J.C. 2008.
Observations of mesoscale movements in the
short-tailed stingray, Dasyatis brevicaudata from
New Zealand using a novel PSAT tag attachment
method. Journal of Experimental Marine Biology and
Ecology 359: 110-117.

Liggins, G.W. & Kennelly, S.J. 1996. By-catch from
prawn trawling in the Clarence River estuary,
New South Wales, Australia. Fisheries Research
25: 347-367.

Liggins, G.W., Kennelly, S.J. & Broadhurst, M.K. 1996.
Observer-based survey of by-catch from prawn
trawling in Botany Bay and Port Jackson, New

South Wales. Marine and Freshwater Research 47:
877-888.

Lotze, H.K., Lenihan, H.S., Bourque, B. J., Bradbury,

R. H., Cooke, R. G., Kay, M. C., Kidwell, S. M.,
Kirby, M. X., Peterson, C. H. & Jackson, J. B.
C. 2006. Depletion, degradation, and recovery
potential of estuaries and coastal seas. Science
312: 1806-1809.

Macbeth, W.G., Pollard, D.A., Steffe, A.S., Morris,

S. & Miller, M. 2002. Relative abundances of
fish and crustaceans and water quality following
the fish kill of early March 2001 in the Macleay
River, northern New South Wales. Pp. 61-100. In
Kennelly, S. & McVea, T. (eds) Scientific reports
on the recovery of the Richmond and Macleay
Rivers following fish kills in February and March
2001. NSW Fisheries Final Report Series No. 39.
(NSW Fisheries, Cronulla).

McPhee, D.P. & Skilleter, G.A. 2005. The set pocket
(stow) net prawn fishery of the Mary River
(Queensland, Australia) and its by-catch.
Proceedings of the Royal Society of Queensland 112:
39-46.

Morton, R.M, 1989. Hydrology and fish fauna of
canal developments in an intensively modified
Australian estuary. Estuarine, Coastal and Shelf
Science 28: 43-58.

1992. Fish assemblages in residential canal
developments near the mouth of a subtropical
Queensland estuary. Australian Journal of Marine
and Freshwater Research 43: 1359-1371.

Muller, J. & Henle, F.G.J. 1841. Systematische
beschreibung dcr plagiostomen. (Veit & Co.: Berlin).

Ogilby, J.D. 1908. On new genera and species of fishes.
Proceedings of the Royal Society of Queensland 21:
1 - 26 .

Parker, P.G. 1999. Fish assemblages at Julian Rocks
and the adjacent waters of northern New South
Wales, Australia. Australian Zoologist 31: 134-
160.

Pease, B.C. & Herbert, S. 2002. Checklist of fishes from
the Hacking River catchment, including the
Port Hacking estuary, in central New South
Wales. Wetlands (Australia) 20: 66-79.

Pierce, S.J., Pardo, S.A. & Bennett, M.B. 2009.
Reproduction of the blue-spotted maskray
Neotrygon kuldii (Myliobatoidei: Dasyatidae) in
southeast Queensland, Australia. Journal of Fish
Biology 74: 1291-1308.

Pillans, S., Ortiz, J.-C., Pillans, R.D. & Possingham,
H.P. 2007. The impact of marine reserves on
nekton diversity and community composition

Memoirs of the Queensland Museum | Nature • 2009 • 55(1)

Distribution of Dasyatis fluviorum

in subtropical eastern Australia. Biological
Conservation 136: 455-469.

Pogonoski, J.J., Pollard, D.A. & Paxton, J.R. 2002.
Conservation overview and action plan for
Australian threatened and potentially threatened
marine and estuarine fishes. (Environment Australia:
Canberra).

Pollard, D.A. 1974. Report 1. The freshwater fishes
of the Alligator Rivers 'Uranium Province' area
(Top End, Northern Territory), with particular
reference to the Magela Creek Catchment (East
Alligator River System). In Conway, N. R., Davy,
D. R., Giles, M. S„ Newton, P. J. F. & Pollard, D.
A. (eds). The Alligator Rivers Fact Finding Study
AAEC/E305. (Environmental and Public Health
Division, AAEC: Lucas Heights).

Reed, D.H. 2004. Extinction risk in fragmented
habitats. Animal Conservation 7: 181-191.

Ross, J. 1999. The influence of residential canal estates on
estuarine fish assemblages in the Gold Coast region,
Australia. Unpubl. MSc thesis, The University
of Queensland, Brisbane.

Ruello, N.V. 1976. Environmental and biological studies
of the Hunter River. Operculum June: 76-84.

Schlacher, T.A., Liddell, B., Gaston, T.F. & Schlacher-
Hoenlinger, M. 2005. Fish track wastewater
pollution to estuaries. Oecologia 144: 570-584.

Schwartz, F.J. & Dahlberg, M.D. 1978. Biology and
ecology of the Atlantic stingray, Dasyatis sabina
(Pisces: Dasyatidae), in North Carolina and
Georgia. Northeast Gulf Science 2: 1-23.

Snelson, F.F., Williams-Hooper, S.E. & Schmid, T.H.
1988. Reproduction ana ecology of the Atlantic
stingray, Dasyatis sabina, in Florida coastal lagoons.
Copeia 1988: 729-739.

SPCC. 1981. The ecology of fish in Botany Bay.
Environmental control study of Botany Bay. BBS
23. NSW State Pollution Control Commission.
Sydney, Australia.

Stead, D.G. 1963. Sharks and rays of Australian seas.

(Angus and Robertson: Sydney).

Steffe, A.S., Macbeth, W.G. & Murphy, J.J. 2007.
Status of the recreational fisheries in two
Australian coastal estuaries following large fish-
kill events. Fisheries Research 85: 258-269.

Stobutzki, I.C., Miller, M.J., Heales, D.S. & Brewer,
D.T. 2002. Sustainability of elasmobranchs caught
as by catch in a tropical prawn (shrimp) trawl
fishery. Fishery Bulletin 100: 800-821.

Thorburn, D.C., Peverell, S„ Stevens, J.D., Last, P.R.
& Rowland, A.J. 2003. Status offreshioater and
estuarine elasmobranchs in northern Australia.
(Report to the Natural Heritage Trust).

Vaudo, J.J. & Lowe, C.G. 2006. Movement patterns
of the round stingray Urobatis halleri (Cooper)
near a thermal outfall. Journal of Fish Biology 68:
1756-1766.

West, R. & Gordon, G. 1994. Commercial and
recreational harvest of fish from two Australian
coastal rivers. Australian Journal of Marine and
Freshwater Research 45: 1259-1279.

White, W.T., Last, P.R., Stevens, J.D., Yearsley,
G. K., Fahmi & Dharmadi. 2006. Economically
important sharks and rays of Indonesia. (Australian
Centre for International Agricultural Research:
Canberra).

Williams, R.J., Bryant, A. & Ledlin, D.E. 2004.
Biodiversity of the Georges River catchment: aquatic
biodiversity. (NSW Department of Infrastructure,
Planning and Natural Resources: Sydney).

Memoirs of the Queensland Museum | Nature • 2009 • 55(1)

Van Dyck, Janetzki & Sheridan

Structure of a burrow of the fawn hopping-mouse
Notomys cervinus (Rodentia: Muridae).

At approximately 0030 hrs on 4 May 2009, an adult hopping-
mouse, subsequently identified as Notomys cervinus, was
chased over a straight line distance of 100 m (although the
actual path followed was zigzagged) until it disappeared into
a vertical burrow (23° 46' 08.9 ,f S, 141° 01'13.6"E, GDA94) on
a treeless, mostly grassless, daypan between Lake Constance
and Hunters Gorge, Diamantina National Park, south-
western Queensland. The 3 cm-diameter hole into which
it retreated was flush with the daypan and not associated witli
anv spoil from the burrow, however, another similar-sized
hotelrSOcm away had an obvious plume of granulated spoil
(silt) flaring north-west from the hole but not surrounding
the hole to form a conical depression.

Before beginning the excavation (under torch light) at
the spoil-associated hole, a butterfly net was spread over
the second hole 150 cm away. Dry silt was then shovelled
from the inner side of the vertical exit shaft (i.e. the area
between the two holes) until, at a depth or 43 cm, the
burrow took a right-angle bend to lead horizontally in an
arc toward the other exit. After 20 cm the burrow swung
out to take on a semi-elliptical orientation. As successive
biscuits of silt were shovelled off while following the 3 cm
wide burrow, a hopping-mouse (adult female) fled the other
exit and was trapped in, and immediately extracted from,
the butterfly net. At approx. 90 cm along the burrow, four-
nestling mice were found in a chamber that expanded to
12 cm diameter. The young, whose eyes were unopened,
were probably about 14 days old (see Watts 1983). They
were found on a scant mattress of thin, chewed grass stems, the
only vegetable matter to be found in the burrow complex.
The chamber was palpably warm and humid.

Two burrows led from tine nesting chamber, a 13 cm blind
tunnel that was a perpendicular offshoot to the main burrow
ellipse and another that constituted the natural extension
of tne main burrow. This led to the second vertical exit/
entrance tunnel approximately 40 cm further on. The second
vertical pipe (the one into which the chased mouse retreated
from the clay pan) was 37 cm long and 3 cm in diameter.
From the base of this pipe, the burrow continued on, but
swung around to form the other arc of an ellipse that would
eventually reconnected it with the entrance where the
excavation began (Figs 1-3).

During excavation of the second half of the burrow
ellipse, another hopping-mouse (adult male) erupted from
a tunnel and began leaping around the open depression of the
excavation. This individual was eventually hand-captured
when it jumped up the inside of one of our shirts. Using
a small portable plastic mouse cage furnished with a few
calico collection bags and with the addition of the original
nesting vegetation, the two adult mice were reintroduced
to the four young, which on capture had been provided
with a hot water bottle. Within minutes, the female was
suckling the young.

Mid-way along the returning sweep of the burrow ellipse
were two more blind tunnels: one, a short (14 cm) horizontal
drive into the centre of the ellipse, and the other, a much
longer (at 54 cm) upward-inclined, tangential spur, that
began as a wide, smooth-walled gallery for approximately

FIG. 1. Plan view of burrow complex with flagging tape highlighting
tunnels. Scale divisions in 1 cm and 10 cm intervals (photo: HJ).

FIG. 2. Diagrammatic representation of burrow complex.

FIG. 3. Excavation (photo: JS).

Memoirs of the Queensland Museum

Nature • 2010 • 55(1)

Van Dyck, Janetzki & Sheridan

18 cm, but then continued as a very narrow (2 cm) pass-
ageway. From this spur a live sub-adult male hopping-mouse
was caught while trying to dig its way out.

Later, after releasing this subadult and driving away
in our vehicle, another subadult was spotlit running from
the opposite direction toward and eventually into the
excavation pit from which it then quickly emerged and ran
off. Within the 15 minutes it took us to drive approx. 2 km
through claypan and gibber away from the excavation, we
recorded by spotlighting an additional adult and a juvenile
N. ceivinu s together, then a subadult, and finally two more
juveniles*.

Some physical features of this burrow complex would
appear to reduce the risk of snake predation. Firstly, it
was positioned in stark 'lunar' habitat mostly devoid of
vegetation or surface cracks, and so perhaps less likely to
be visited by diurnal snakes due to the degree of exposure.
Secondly, the horizontal elliptical nature of the tunnelling
with the provision of pop holes at either 'end' of the loop
could ensure escape from almost anywhere within the
system. Thirdly, the nesting 'chamber' was not positioned
a f the end of a blind tunnel but in an expanded section of
the main burrow. Even if nestlings were too young to
escape, the reproducing adult(s) could. Blind galleries
may be more important in providing points from which
rapid-response exits can be dug when established exits
are suddenly blocked. In this regard, the very narrow,
upward rising tangential spur from which a subadult was
collected toward the final sages of the excavation, probably
represented the ongoing frantic execution of such a tunnel.
Some features of the construction are more puzzling. For
example, how, in the digging process 40 cm underground,
?o the mice connect the ends of the horizontal loop? How
ls burrow spoil (Fig. 4) moved up the vertical entrance/
escape shaft? And why construct an all-horizontal system
when stepping-up of burrow elements could provide early
Warning of incoming water or, escape from flooding?

Parts of Diamantina National Park were inundated by
uoodwaters between January and April 2009. This site
was only 1.5 km from the main Diamantina River channel
an d at the same elevation (Fig. 6). Clearly, the superficial,
contorting crust on the claypan on which the burrow
e xcavation occurred indicated its relatively recent receipt
of substantial rain and/or inundating floodwaters (Fig.
o)- The ease of excavation ( ipso facto burrow construction)
through the homogeneous silt substrate may have resulted
tom relatively recent inundation, and the generational
fccumulation of burrow spoil may contribute to the
establishment of the small raised hummocks that dot the
clay pa ns . i n time these hummocks may deflect surface
Water from entrance tunnels, but their capacity to trap
wind-blown seeds among grass tufts and scattered rubble,
‘hay be more significant.

While not labyrinthine in construction this one burrow
complex builds on earlier reports that N. cervittus con-
ducts 'very simple' burrows free of side galleries and
,, evo jd of radiating warrens (Finlayson 1939). It confirms
in a f individuals live in small family groups (Finlayson
tl u Watts & Aslin 1981). We acknowledge however, that

1e burrow system we describe here may have been still

FIG. 4. Spoil plume from an adjacent N. cervinus burrow (photo JS).

FIG. 5. Northern view from excavation site across claypan (photo: HJ).

FIG. 6. Eastern view from excavation site toward Diamantina River
channel (photo: HJ).

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Van Dyck, Janetzki & Sheridan

TABLE 1 Mammals trapped (T), hand-caught (H), or spotlit (S) in the vicinity of the excavation (all coordinates are GDA94) between 28
April and 4 May 2009 (975 trap-nights; 27 hrs spotlighting).

Dasyuroides bymei

23 -.wiw rs, urnw i n. «] u u m> 2/5/09

23*45'48.2"S, 141*08'26.5"E (T, subadult F) 3/5/09

Planigale tenuirostris

23*34'35.9"S, 141*06'17.0"E (T) 1/5/09

1 Sminthopsis crassicaudata

23‘33'37.0"S, 141*0745.7 W E (H) 30/4/09

23‘43'02.2"S, 141*10'565"E (T. juv) 4/5/09

Sminthopsis macroura

23*36*41 .6"S, 14r03T5.2''E (T) 30/4/09

\ Notomys cervinus

23* 39* 44.2"S 141 * 10' 38.6"E (H) (28/4/09)

23* 41' 393"S 141* 04' 58.6"E (S) (29/4/09)

23* 46' 35.4-S 141* 09' 08.5"E (H) (2/5/09)

23* 45' 34.4"S 141* 08' 25.2~E (H) (2/5/09)

23' 46' 08.9"S141* 01' 13.6"E (H, 2 adull, 4 juv, 1 subadult) (S, 1 subadult) 4/5/09

*23*45'50.0"S, 141*0V20.0"E (S, 1 adult, 1 juv) 4/5/09

•23*45'31.6"S, 141*01'29.6"E (H, subadult) 4/5/09

•23*45T0.0"S, 141 '0T50.0"E (S, 2 juv) 4/5/09

Leggndina fbrresti

23*33'37.0"S, 141*07' 45.7"E (H, juv) 30/4/09

23*44'22.2”S, 141*10'30.6"E(H) 2/5/09

under construction, or may have been modified by reptiles
or other mammals. Desert burrows are generally celebrated
for ensuring cool humid conditions in hot climates but,
inasmuch as this excavation occurred on a very cold night,
it highlighted a burrow's capacity to also provide warm,
humid conditions for adults and nestlings during periods
of low ambient temperature.

Literature Cited

Finlayson, H.H. (1939). On mammals from the Lake Eyre Basin,
■part IV The Monodelphia. Transactions of the Royal Society of
South Australia 63: 88-117 (as Notomys aistoni).

Watts, C.H.S. (1983). Fawn Hopping-mouse Notomys cervinus. Pp
432-433. In R. Strahan (ed.) The Complete Book of Australian
Mammals. Angus and Robertson: Sydney.

Watts, C.H.S. & H.J. Aslin (1981). The Rodents of Australia. Angus
and Robertson: Sydney.

Steve Van Dyck, Heather Janetzki and John Sheridan, Queensland
Museum, PO Box 3300 South Brisbane, 4101, Qld, Australia.
steve.vandyck@qm.qld.gov.au; 5 August 2009.

100

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Distribution of the porcupine ray Urogymnus
asperrimus (Bloch & Schneider, 1801) in Australian
waters, with new records from Queensland

Susan M. THEISS

Sensory Neurobiology Group, School of Biomedical Sciences, The University of Queensland, St.
Lucia, Qld 4072, Australia. Email: s.theiss@uq.edu.au

Peter M. KYNE

School of Biomedical Sciences, The University of Queensland, St. Lucia, Qld 4072; present address:
Tropical Rivers and Coastal Knowledge, Charles Darwin University, Darwin, NT 0909.

Leslie A. CHISHOLM

Marine Parasitology Laboratory, School of Earth and Environmental Sciences (DX 650 418), The
University of Adelaide, North Terrace, Adelaide, SA 5005.

Citation: Theiss, S.M., Kyne, P.M. & Chisholm, L.A. 2010 03 15. Distribution of the porcupine ray
Urogymnus asperrimus (Bloch & Schneider, 1801) in Australian waters, with new records from
Queensland. Memoirs of the Queensland Museum -Nature 55(1): 101-101. Brisbane. ISSN 0079-8835.
Accepted: 20 May 2008.

ABSTRACT

Three specimens of the porcupine ray Urogymnus asperrimus (Bloch & Schneider,
1801) are reported from Heron Island on the Great Barrier Reef, Qld. These are the
first records from the southern Great Barrier Reef and represent the southernmost
records for this species on the east coast of Australia. An immature male with a disc
width (DW) of 650 mm and two females measuring 620 mm DW and 545 mm DW were
caught on the eastern side of the island using hand or seine nets. The two females were
released alive after examination. Some morphometric data from two of the individuals
are provided. The distribution, biology and ecology of this species are poorly-known, with
only five catalogued Australian specimens held in Australian museums. The majority of
these are not whole specimens and are in poor condition. There is further scattered
information from photographs and live sightings. All known Australian records of U.
asperrimus are summarised here. There are records of the species across tropical
northern Australia, from Ningaloo Reef, WA (22°43’S) to Heron Island, Qld (23°26’S).
□ Myliobatoidei, Dasyatidae, Urogymnus, southern Great Barrier Reef, Heron Island.

The porcupine ray Urogymnus asperrimus
(Bloch & Schneider 1801) is a large dark brown
to greyish batoid, identified by an oval-shaped
disc covered in plate-like denticles and
sharp thorns (Last & Stevens 1994). The tail
acks stinging spines and skin folds (Last &
levens 1994), the latter separating the genus
r °m the closely-related Dasyatis Rafinesque,

1810. Urogymnus asperrimus occurs in tropical
waters throughout the Indo-West Pacific from
east Africa to Fiji, and in the tropical Eastern
Central Atlantic off west Africa (Last & Stevens
1994). It attains a maximum disc width (DW) of
at least 1470 mm and is found inshore on sand
and coral rubble substrate near reefs (Last &
Stevens 1994; White et al. 2006). Very little

^ e moirs of the Queensland Museum | Nature

2010 • 55(1) • www.qm.qld.

101

Theiss, Kyne & Chisholm

is known about the biology of U. asperrimus,
and detailed information on distribution and
habitat is lacking.

Despite the wide distribution of U. asperrimus,
it has not regularly been reported (Compagno
2000). Previous Australian records are limited
to only a few specimens with little to no accom-
panying data, along with anecdotal sightings,
photographs and live sightings from Western
Australia (WA). The first Australian record of
the species appears to be an individual caught
in 1770 by the crew of the H.M. Bark Endeavour
on James Cook's first voyage to the Pacific.
This individual was taken off the mouth
of the Endeavour River, with a description
provided by botanist Daniel Solander, fitting
that of U. asperrimus (description reprinted in
Whitley 1939). There is only one previously
catalogued specimen from the east coast of
Australia (Queensland Museum, QM 1.1112-
4), which is registered as the holotype of the
subspecies U. asperrimus solanderi Whitley,
1939. Whitley (1939) used Solander's description
of the Endeavour River specimen, together with
limited material (teeth, velum maxillare, buccal
processes and skin) to 'give a new subspecific
name to distinguish it from the Bombay type
of the species', but did not provide an actual
description of the subspecies, nor distinguishing
features between it and the nominate form.
Urogymnus africanus (Bloch & Schneider 1801)
and Raja africana Bloch & Schneider 1801, are
both synonyms of U. asperrimus, with the
only other valid species in the genus being the
pincushion ray U. ukpam (Smith 1863) of west
African freshwater basins.

This paper reports on the location and
habitat of three new U. asperrimus individuals
collected from eastern Australia and summarises
the known Australian records of the species.

MATERIALS AND METHODS

The first new specimen was examined at
the Heron Island Research Station (HIRS),
The University of Queensland, before being

transported to and lodged at the Queensland
Museum (QM), Brisbane, where it is preserved
in ethanol. The second and third specimens
were kept alive overnight in a large sea-
water tank at HIRS before being measured,
photographed and released. Location, date, DW,
disc length (DL), total length (TL), weight (for tire
live individuals only), sex and maturity, where
possible, as well as some selected morphometries
for two individuals were recorded from the
specimens following Last & Stevens (1994).

Information on Australian records of U.
asperrimus in museum collections, as well as
confirmed live sightings, were gathered in
order to provide a summary on the distribution
and biology of this species in Australian waters,
and to compare this information with the new-
ly recorded individuals presented here. The
following abbreviations are used for insti-
tutions: AMS, Australian Museum, Sydney;
NTM, Museums and Art Galleries of the Northern
Territory, Darwin; QM, Queensland Museum,
Brisbane; WAM, Western Australia Museum,
Perth.

RESULTS

All three U. asperrimus specimens were collected
from Heron Island (23°26'S, 151°54'E), part of
the Capricorn Bunker Group of islands at the
southern end of the Great Barrier Reef, 72 km
off the coast of Gladstone, Qld.

The first specimen (QM 1.31178, immature
o, 650 mm DW, 665 mm DL, 1375 mm TL) was
collected on 9 July 1998 at Shark Bay on the
eastern end of Heron Island by seine net at
dusk Other measurements: 120 mm snout
ength, 290 mm head length, 250 mm trunk
ength and 835 mm tail length. The second
specimen (?, 620 mm DW, 650 mm DL, 1270 mm

o' Tuf kg ) was coll ected on 9 July 2006 from
e fallows of Shark Bay after sunset by hand
net. The third specimen (y, 545 mm DW, 570
mm DL 1130 mm TL, 11.70 kg) (Fig. 1 A) was
collected on 1 February 2008, also from Shark

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Urogymnus asperrimus in Australian waters

FIG. 1 . Photographic records of the porcupine ray, Urogymnus asperrimus. A, Dorsal view of a 545 mm DW $ U.
asperrimus collected from Heron Island, Qld and released alive. Scale bar= 10 cm; B, Underwater lateral view of
U. asperrimus from Fitzroy Reef, Qld. (Photo: Brett Vercoe).

Bay, using a seine net on the incoming tide in
the afternoon. Other measurements: 100 mm
snout length, 230 mm head length, 235 mm trunk
length and 665 mm tail length. This specimen
was recaptured two days later with a hand net
at high tide, approaching dusk, in nearly the
same location within Shark Bay. Photographs
of white spots found along the margin of the
disc were used to identify the animal as the
previously caught specimen. Maturity was not
determined as the animals were released alive.
But as females mature by 1000 mm DW (White
e t al., 2006) it is assumed that both specimens
Were immature. Minor abrasions were noted
°n the dorsal surface of the disc of the second
specimen, but these were unrelated to the method
°f capture.

provided here. Of the 25 known records, 10 are
live sightings recorded from visual surveys, six
are from photographs (Fig. 1), three are dried skins
(or parts of), one is small parts of an individual,
two are whole specimens preserved in ethanol,
one was used for research purposes, and one
specimen could not be located. An additional
historical account appears to represent the first
Australian record from 1770. With the exception
of the two ethanol preserved specimens, the
catalogued material is generally in a poor
state. Accompanying data is limited for many
records. Some records have a detailed note
of location and most records include a date of
capture, but actual specimen data, including
size, sex and maturity, are lacking in most
instances (see Table 1).

There are 25 known records of U. asperrimus
from Australian waters (Table 1). Including
the three new records reported here, there are
nine records from Qld, 15 from WA and one
from the Northern Territory (NT). A further
two registered records held in Australian
collections (AMS and QM) are from the Gilbert
stands, part of the Republic of Kiribati in the
Western Central Pacific. As these records were
not obtained from Australian waters, they
ar e not included in the summary of records

DISCUSSION

Australian records of the porcupine ray are
scattered across tropical northern Australia,
with the southernmost point of occurrence on
the west coast, Ningaloo Reef, WA (22°43'S) and
on the east coast. Heron Island, Qld (23°26'S).
The new specimens are the only confirmed
records of the species from Heron Island and
the southern Great Barrier Reef. They also

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

103

Theiss, Kyne & Chisholm

Remarks

Taken on Cook's first voyage
to the Pacific, Description
provided by D. Solander
(sec Whitley, 1939).

0
z

i§

1 «

bo
Q U

"O C

$/ r, os

1 &1S

li'
If =|

fill

§ % JB J
ca .2. 2 £

No collection data. Specimen
location unknown

Caught in net, released alive

Found dead on beach,
not collected

Based on incidental sighting

during underwater visual

surveys of reef fish

Whole specimen in ethanol

Healed umbilical scar

Released alive. Mass 16.04 kg

Underwater photograph

taken at 10 m depth

Two underwater photographs

taken at 5 m depth

Nine confirmed sightings from

two locations, with photograph

Released alive. Recaptured

in same location 2 days

later. Mass 1 1.70 kg

Beach-washed specimen,

dried skin in poor condition

Whole specimen in ethanol, tail

tip missing. No collection data.

Dried piece of skin only

Sex and
i maturity

Female,

immature

Female

Male,

immature

Female,

immature

Female,

probably

immature

Female,

probably

immature

Female,

probably

immature

Total

i Length

(mm)

■

•

■

1375

1000

1270

•

■

•

1130

■

1221

■

Disc

Length

(mm)

■

665

490

650

•

■

570

■

642

•

Disc Width

(mm)

*■

■

•

650

460

620

■

545

■

617

■

Collector

Crew of H.M.
Bark Endeavour

J. Brazier

J.R. Tosh

G.F. Mees

| J.B. Hutchins

J.B. Hutchins

H. Malcolm

L.A. Chisholm

L. Squire Jr.

S.M. Theiss

B. Vercoe

| B. Carter

J.D. Stevens

S.M. Theiss

Date

1770

1888

08/04/1913

1959

07/1991

1998

09/07/1998

11/06/2003

09/07/2006

16/2/2 007

04/2007

01/02/2008

Longitude

-t

■3*

141°39'E

-r

grp

CO cn

Latitude

c n

ui cn

Location caught

Endeavour River

mouth, Qld

Kingsmill

Islands, WA

Damey Island, Qld

Dampier, WA

Exmouth Gulf, W A

Wellesley

Islands, Gulf of
Carpentaria, Qld

Heron Island, Qld

Off Weipa, Gulf of

Carpentaria, Qld

Heron Island, Qld

Fitzroy Reef, Qld

I Northern end of

Flynn Reef, Qld

>. 8 5

III

c o i
c a, c

S <4 iz

Heron Island, Qld

*fc

§ -S
5 c

n Registration

number

Historical

account

AMS 1.1721

QM 1.1112-4

WAM P.4631-

0001

Photograph

Live sighting

QM 1.31178

(present

manuscript)

Research

specimen

Photographs

(present

manuscript)

Photograph

Cl,

Cl.

Live

sightings and

photograph

Photographs

(present

manuscript)

Unregistered <

5 £

VAM E

'.29573-001

Museun

WAM

! WAM

—

■

•

■

WAM V

Memoirs of the Queensland

104

Museum | Nature • 2010 • 55(1)

TABLE 1. Summary of all known Australian records of Urogymnus asperrimus (Bloch & Schneider, 1801). AMS, Australian Museum,
Sydney; NTM, Museums and Art Galleries of the Northern Territory, Darwin; QM, Queensland Museum, Brisbane; WAM, Western Australia
Museum, Perth.

Urogymnus asperrimus in Australian waters

represent the known southern distribution limit
for the species on the east coast of Australia.

The eastern side of Heron Island is primarily
a sandy bottom habitat of the reef lagoon with
some coral rubble leading out to the inner reef.
Large aggregations of batoids can be found in
the lagoon, most commonly the pink whipray
Himantumfai Jordan & Seale, 1906, the cowtail
ray Pastinachus atrus (Macleay, 1883) and
the giant shovelnose ray Glaucostegus typus
(Bennett, 1830). All three U. asperrimus
specimens, however, appeared to be solitary.
Heron Island and the southern Great Barrier
Reef are well surveyed regions and the low
recorded abundance of the porcupine ray is
probably an accurate indicator of the rarity
of this species locally. Indeed, throughout its
Australian range, this species is not regularly
recorded and there is a lack of reliable records
to confirm distribution, along with basic
biological information. Although U. asperrimus
is rarely encountered, it may be more widely
distributed across northern Australia than
current records indicate. In this area relatively
turbid conditions markedly reduce the effective-
ness of underwater visual survey techniques,
compared to in clearer Great Barrier Reef and
west coast waters. Overall, little biological
information can be gathered from the existing
Australian records.

Behavioural observations made from
live sightings at Ningaloo Marine Park
Provide some insight into the ecology of this
species (J.D. Stevens, pers. comm.). Urogymnus
a sperrimus appears to plough strongly through
the substrate when feeding, which is different
to the feeding strategies of other large rays,
a nd there seems to be some suggestion that
larger individuals are found at greater depths
(J D. Stevens, pers. comm.).

Urogymnus asperrimus is listed as Vulnerable
°n the IUCN Red List of Threatened Species due
to the lack of regular records and an appar-
ent decrease in abundance (Compagno, 2000).
Although handling of U. asperrimus is difficult

due to its rough ciorsal surface, this species
is commonly caught in South East Asia and
used for its meat, cartilage and skin, which is
of particularly high value (White et al., 2006).
Basic biological information such as age, growth
and reproduction are lacking, making an
assessment of the species' resilience to fishing
pressure difficult. Its rarity, inshore habitat and
large size, however, may imply that it is unlikely
to sustain prolonged targeted fishing.

ACKNOWLEDGEMENTS

Thanks go to Jeff Johnson for providing
details on the specimen lodged at QM as
well as other Australian records; Sue Morrison
and Claire Bartron for information on lodged
specimens at WAM; John Stevens for providing
information gathered from visual surveys in
Western Australia; Brett Vercoe for supplying a
photograph of the Fitzroy Reef specimen; Janine
Caira and Kirsten Jensen for details on the
Weipa specimen; Nathan Hart, Shaun Collin,
Blake Harahush, and Carla Atkinson for help
with fieldwork; and the staff of the Heron Island
Research Station, The University of Queensland.

LITERATURE CITED

Compagno, L.J.V. 2000. Urogymnus asperrimus. In
IUCN 2007. 2007 IUCN Red List of Threatened
Species, www.iucnredlist.org

LAST, P.R. & STEVENS, J.D. 1994. Sharks and rays of
Australia. (CSIRO Division of Fisheries: Hobart).

White, W.T., Last P.R., Stevens, J.D., Yearsley,
G.K., Fahmi & Dharmadi. 2006. Economically
important sharks and rays of Indonesia.
(Australian Centre for International Agricultural
Research: Canberra).

Whitley, G.P. 1939. Taxonomic notes on sharks
and rays. Australian Zoologist 9(3): 227-262.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

105

The Julia Creek dunnart and other prey
of the barn owl in Mitchell grass downs
of north-western Queensland

Patricia A. WOOLLEY

Department of Zoology, La Trobe University, Melbourne, Vic 3086. Email: p.woolley@latrobe.edu.au

Citation: Woolley, P.A. 2009 03 15. The Julia Creek dunnart and other prey of the barn owl in
Mitchell grass downs of north-western Queensland. Memoirs of the Queensland Museum — Nature
55(1): 107-117. Brisbane. ISSN 0079-8835. Accepted: 18 February 2009.

ABSTRACT

Analysis of the contents of pellets produced by barn owls, Tyto alba, has provided
information on the distribution of the Julia Creek dunnart, Sminthopsis douglasi, and
five other species of small terrestrial mammals. The pellets were collected from 28
localities in Mitchell Grass downs country around the town of Julia Creek in north-western
Queensland. Mammals formed the largest component of the diet of the owls, followed by
birds. Reptiles, frogs and arthropods were also preyed upon. □ Prey of Tyto alba, Sminthopsis
douglasi, dasyurid marsupials, rodents, birds, Queensland, Mitchell grass downs.

Barn owls, Tyto alba, feed mainly on small
terrestrial mammals but they are known to take
other vertebrates and insects (Taylor 1994).
Studies on the diet of barn owls in Australia
(see Morton 1975 which includes a summary
of earlier studies; Morton et al. 1977; Morton &
Martin 1979; Valente 1981; Smith & Cole 1989;
Debus et al. 1999; Hey wood & Pavey 2002)
have shown that a variety of mammals, birds,
reptiles, frogs and arthropods may be eaten, and
that rodents such as Rattus villosissimus and Mns
ihusculus, species that undergo large fluctuations
M population size, often form the bulk of the
diet. Barn owls usually swallow their prey
w hole and the undigested remains, including
hones, hair, feathers, scales and hard parts of
insects that are all generally readily identifiable,
are regurgitated in compact pellets. Each pellet
ls thought to contain most of the remains from
a single foraging expedition (Taylor 1994).
Freshly ejected pellets are soft and covered with
mucous which, as it dries, gives them a smooth,
dark, glazed appearance. In dry conditions the
glaze is lost after about 10 days but the pellets
remain firm and darkly coloured for eight or

more months (from Barn Owls On Site http://
www.barnowltrust.org.uk/content_images/
pdf [July 2008]). Barn owls roost in caves, old
buildings, tree hollows and, occasionally, in
trees, and pellets accumulate at the roost site.

Analysis of the contents of pellets provides
not only information on the diet of the owl but
also information on the composition of the
fauna in the foraging area. This may lead to the
detection of uncommon species, as happened
in the case of a then little known species of
dasyurid marsupial, the Julia Creek Dunnart,
Sminthopsis douglasi (Woolley 1992). This
species, at the time of its description by Archer
(1979), was known from only four specimens in
museum collections. These had been collected
from three localities between the towns of Julia
Creek and Richmond, Queensland. As a result of
survey work commenced in mid-1990 (Woolley,
1992) the skeletal remains of Julia Creek
dunnarts were found among a deposit of bones
that were considered to have formed from the
disintegration of owl pellets in a hollow tree.
Trapping in an area close to where the bones

1071

Memoirs of the Queensland Museum | Nature • 2010 • 55(1) • www.qm.qld.gov.au

Woolley

TABLE 1. The localities and collecting sites together with the dates on which owl pellets were collected, the
number and condition (glazed - G, unglazed - U) of the pellets, and whether or not bam owls were present
at the site. Localities (abbreviation in parentheses) ordered by date of earliest collection.

Locality and
collecting sites

Date of
collection

No. of
pellets

Locality

total

Bam

owls

Eureka (EUR) - 20TWS.
141°48'E

abandoned homestead

30.4.1992

72(C)

ves

27.5.1992

43(G)

yes

29.6.1992

40(G)

ves

3.8.1992

33(G)

ves j

5.10.1992

17(G)

ves

6.7.1994

33 (15U,
18G)

yes

24.7.19*1

15(G)

ves

2.94991

18(C)

271

Ves

Crendon (CRE) - 21°26'S,

142°07'E

shearing shed

-.11.1993

31 (U)

Nelia (N) - 21T39'S,

1423 3'E

a) derelict house, b)
racecourse buildings

19.6.1994 a

28(G)

yes

10.6.1995 b

50(G)

ves )

26.7.1995 b

19(G)

yes

9.10.1995 b

9(G)

27.4.1996 b

36(C)

142

no j

Nelia West (NW) -
20°42'S, 142°04'E derelict
shearer's quarters

19.6.1994

19 (U)

21.3.1997

66 (U)

Osbert (O) - 20‘30'S,
141°46'E outbuildings on
unoccupied property

20.6.1994

215 (92U,
123G)

yes j

17.6.1995

54 (26U,
28G)

269

Lvrian (L) - 19°27S,

14 132^ abandoned
homestead, outbuildings,
nearbv tree hollow

21.6.1994

30(G)

ves

6.10.1995

56 (35U,

21 G)

Auckland Downs
(AD) - 2(n5'S, 141 U 46'E
derelict outbuilding

3.7.1994

22 (U)

Huddersfield (H) -
20”57S, UF53'E shearing
shed, outbuildings

4.7.1994

76 (U)

29.4.1995

33(G)

ves

6.6.1998

82(G)

ves

303.1999

12(G)

ves I

3.6.1999

81(G)

ves

10.9.2000

6(U)

290

Eulolo (EUL) - 2134'S,
141°33'E outbuildings on
unoccupied property

5.7.1994

264 (U)

264

Penola Downs (PD) -
J 21*31'$, 141°27&dferelict
shearer's quarters

5.7.1994

115 (U)

115

Carrum (CAR) -
20°53'S, 141°43'E
unoccupied house

6.7.1994

9(G)

yes 1

Armidale (ARM)

- 20°52'S, 141°41'E
shearing shed

6.7.1994

18(G)

Waterloo Plains (WP)

- 20°57S, 141°43'E
unoccupied house,
shearing shed

7.7.1994

258 (191U,
67G)

yes

.8.1994

4(G)

yes

.9.1994

6(G)

268

yes

Kelloshiel (K) - 21°01'S,
141°42'E shearing shed

7.7.19*1

5(U)

Locality' and
collecting sites

Date of
collection

No. of
pellets

Locality

total

Bam

owls

Eastern Creek (EC) -

20 , '54'S, 141*4715 tree
hollow at Waterloo

Rd crossing

7.7.1994

20 (9U,

11G)

yes

Longford Plains (LP)

- 20"46'S, 141"46'E
abandoned homestead

8.7.1994

57(G)

yes

Toorak (T) - 21”02'S,
141’48'E a) tree hollow
in paddock no. 8, b)
yards between paddocks
, 1 and 2, c) bam

8.7.1994 a

9(G)

ves

2654998 b

45(G)

yes

24.5.1999 c

40(G)

ves

-11.1999c

11(G)

ves

20.8.2000 c

85(G)

yes

1211.2001c

22(G)

210

ves

Eddington (ED) - 21139*5,
141*33'E shearing shed

9.74994

16 (U)

Ardbrin (ARD) - 2 LIB’S,

14237E shearing shed,
derelict shearer's quarters

15.64995

260 (99U,

161C)

yes

168.1995

38(G)

ves

19.9.1995

>**(G)

18.6.19%

25(C)

13.3.1997

16 <U)

28.10.1999

15(C)

376

Consentes (CON)

- 2036'S, 141°36'E
derelict outbuilding

17.61995

98 (55U,

43G)

Broad lands (B) - 21"28'S,

14137E disused
shearing shed

22.6.1995

32 (U)

Minamere (M) - 2D°52'S,

14238'E a) abandoned
homestead, b) meat house
at shearer's quarters

11.7.1995 a

23(G)

ves

15.8.1995 b

277 (170U,

107C)

yes

19.9.1995 b

21(G)

ves

1540.1995 b

40(G)

ves

27.11.1995 b

26(G)

yes

18.1.19% b

33(G)

20.2.19% b

2(G)

193.19% b

19 (G)

24.4 4 996 b

6(G)

3.6.1998 b

9(G)

456

Canobie (CAN) -

m6'S, 140°58'E (hut
at Hawkes Nest bore)

7.10.1995

72 (45U,
27G)

Shilmalier (S) -

I 2030'S, 141°58'E

disused shearing shed

17.4.19%

24 (U)

Quambetook (Q) -

21°12'S, 1423 0'E
shearing shed

17.104996

293 (159U.
134G)

293

yes

Proa (P) - 20‘34'S.

142WE

shearer's shower block

30.5.1998

46(G)

ves

28.5.1999

32(G)

ves

2931999

15(G)

ves

1414999

5(G)

ves

Yorkshire Downs (YD) -

20°52'S, 141°58'E
under trees

3.114999

6(G)

ves

31.5.2000

3(C)

“ Rosevale (R) -

1 21°06'S, 14200'E
abandoned homestead

9.9.2000

113(G)

113

yes

108

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Prey of barn owl in Queensland

were found led to the collection of live Julia
Creek dunnarts in 1992.

The present study of the prey of the Bam Owl
in north-western Queensland was undertaken
primarily to obtain information on the dis-
tribution of the Julia Creek Dunnart which, in
turn, might lead to the finding of other trappable
populations for study of tine biology of the species,
currently listed as Endangered (EPBC Act 1999).

METHODS

Collection of pellets. Searches for pellets of the
Barn Owl were made between April 1992 and
November 2001 in Mitchell Grass downs country
around the town of Julia Creek in north-western
Queensland, an area encompassing part of the
presently known range of S. douglasi (Kutt 2003).
Pellets were found in abandoned homesteads
and outbuildings, in tree hollows and under trees
in 28 localities (Table 1). At some localities pellets
were collected from more than one site, and
sometimes one or two barn owls were present
at the site where the pellets were found. Most
collecting localities were revisited, or checked
by local residents, after the first collection was
made. At some, no further accumulations of
pellets were found and barn owls, if previously
present, were no longer roosting at the site. Tire
collecting sites at Eureka, Longford Plains and
Armidale were demolished early in the collecting
period. Pellets of the Southern Boobook, Ninox
novaeseelandiae, which were also found at some
sites but not collected, were easily distinguished
from those of the Barn Owl by their generally
smaller size and friable structure.

Intact pellets were picked up, classed as either
unglazed or glazed, and packed singly for tran-
sport to the laboratory.

Preparation of pellets. Pellets were soaked
individually in water, to which a few drops of
detergent had been added, until soft (usually 30-
60 minutes). The pellets were teased apart with
forceps in a white enamel tray. By a process of

differential flotation followed by sieving using
a 1.2 mm mesh sieve the lighter hair, feathers
and debris were separated from the heavier
bones. Hair from some pellets, and distinctively
patterned or coloured feathers, were saved. Once
the water was relatively clear the bones and other
hard parts were collected from the tray and
placed in a petri dish to air dry. Before it was
discarded the material in the sieve was felt for
bones that may have been trapped in fur.

Identification of contents. Identification of the
remains of vertebrate prey items was based
primarily on characteristics of the skulls, teeth
and lower jaws as seen in reference specimens.
For rodents these included differences in the
size of the skull, length of the molar tooth row
and width of the molars. Notched incisors and
a difference in the shape of the anterior edge
of the zygomatic plate helped to distinguish
M. musculus from Leggadina forresti. Among the
dasyurids, the larger size (length of dentaries,
width of third upper molar, canine teeth,
pelvic girdle) was used to distinguish adult
specimens of S. douglasi from adult Sminthopsis
macrourn. Dentaries of juvenile specimens
of Sminthopsis that lacked teeth upon which
identification could be made were assigned, if
an adult was present in the same pellet, to the
same species. The skeletal remains of Planigale
were assumed to be those of P. ingrami based
largely on the observation that this was the only
species of Planigale either trapped or found in
the study area (Woolley & Mifsud unpub.
observations). Differences in the pelvic girdle
(shape of obturator foramen and anterior edge of
pubic bones) were found useful in distinguishing
dasyurid marsupials from small rodents, and the
rodents M. musculus and LJbrresti from each other,
and provided additional confirmation of identity.
Experts (see acknowledgments) assisted with
the identification of some taxa. The identity of
one mammal was confirmed by examination
of hair structure, and of some birds by distinctive
feathers. The minimum number of individuals
of a vertebrate prey item in a pellet was calculated

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

109

Woolley

TABLE 2. The number of pellets obtained at each locality, and the number containing each taxon. For
each taxon the number of pellets as a percentage of the total is shown in square parentheses, followed by
the number of individuals (arthropods were not counted). Key to localities in Table 1.

Locality

No. of

No. of pellets containing remains of each taxon, [% of total) . (no. of individuals)

pellets

Mammals

Birds

Reptiles

Frogs

Arthropods

EUR

271

227 [83.71 (295)

36 [13.2| (82)

14 [5.11(15)

1 [05] (1)

13 [4.7]

CRE

10 [32.2) (10)

28 [90.31 (73)

5 116.1 1(8)

19 161.21

1 N

142

114 [80.21(185)

25 [17.61 (36)

22|15.41(55)

23 [16.2| (82)

60 [42.2]

64 |75.21 (85)

9 [10.5[ (17)

7 [8.2| (7)

6 [7.0] (24)

22 [25511

1 O

269

149 [553| (215)

96 [35.6| (187)

46 [17.1 [(176)

8 [29[ (18)

50 [183J

1 L

B4 [97.6] (131)

1 |H 10)

i [i.ii (i)

1 [1 11 (1)

3 [341

I AD

9 [40 9| (18)

9 1 40.9](1 5)

8 [36.3| (35)

6 [27.2[

290

214 [73.71 (277)

77 [26.5[ (132)

19 [6.51 (33)

7 [2.4] (13)

69 [23.7)

I EUL

264

255 (96.5) (635)

29 [10.9) (39)

9 [3 41 (11)

3 [1.11 (6)

22 [8.3|

115

111 |96.5] (271)

1 [0 «l (2)

1 [0.8] (2)

1 |08)

| CAR

9 [100J 08)

1(111] (2)

2 [22.2| (5)

2 [22.2)

i ARM

16 [88.8] (33)

5 [27.71 (14)

1 [5.5] (1)

3 [16.6]

268

170 [63.41 (295)

145 [54.11 (-168)

9 [3.31 (11)

10 [3.7] (16)

101 [37.6|

3 [60.0] (8)

3 [60.0] (13)

1 [20.010)

2 [40.0|

16 [80.01 (20)

3 [15.0| (13)

1 15.0]

54 [94.7J (80)

2(351(2)

1 [1.7]

210

184 [87.6] (252)

60 [28.5] (108)

13 [6.11 (26)

10 [4.71 (14)

38 [18,0|

9 [56.2| po)

3 [18.71 (3)

11 [68.7] (48)

6 [37.5| (18)

12 [75.0[

ARD

376

329 [87.51 (664)

51 [13.5] (75)

65 [17.2] (159)

4 [1.0] (7)

89 [23.6]

CON

93 [94.8J (141)

1 110)0)

1 [1-0] (2)

2 [2.0[ (22)

16 [16.31

30 [93.7J (47)

11311(1)

456

351 [76 9] (613)

137 130.0| (310)

53 [11.6| (135)

10 12.1)(21)

141 [30.9]

|| CAN

64 [88.8] (75)

1 [13] (1)

3 H 1[(3)

9 [12.5| (30)

14 |19.41

| s

8 I33.3J 03)

1 [4 1 1(1)

14 [58.3] (55)

18 |75.0] (58)

18 [75.0)

293

161 [54.9| (232)

194 [66.2) (493)

41 [13 9) (140)

88 [30.0]

66 [67.31 (90)

31 [31.6J (37)

4 [4.01 (4)

1 [1.0| 0)

23 [23.41

9 [1001 (9)

113

59 [52.2| (71)

42 [37.11 (42)

15 [132]

| All localities

3749

2868 [75.5]

983 [26.21 (2053)

348 [9.3]

130 [3.4]

829 [22.11

(4793)

(907)

(373)

from the number of dentaries (mammals), lower
mandibles (birds), lower jaws (reptiles) and ilia
(frogs). The number determined in this way was
often supported by counts of other elements e.g.
for mammals the number of upper jaws and pelvic
bones matched the number of dentaries, and for
some birds the number of sacra / keels/ gizzards
matched the number of lower mandibles.
Mammals and birds were identified to species,
reptiles to family and frogs to order.

Identification of the remains of invertebrate
prey items, which included several orders of
arthropods, was based on wing fragments,
head capsules, mandibles and legs. The number
of individuals of each invertebrate prey
item in a pellet was generally not determined.
The bulk of the material extracted from the
pellets for identification has been lodged in the
Queensland Museum.

110

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Prey of barn owl in Queensland

Number of pellets

FIG. 1. Number of pellets containing the countable
remains of prey items for each taxon. In the case of
arthropods (mostly grasshoppers) the actual number of
individuals was not determined. The number of pellets
expressed as a percentage of all pellets examined
(3,/49) is shown at the end of each oar, together with
the number of individuals in parentheses.

RESULTS

The contents of 3,749 pellets collected between
April 1992 and November 2001 have been
examined. They were obtained from 28 localities
(Table 1), and the number of pellets per locality
ranged from 5 to 456. At 16 localities, 15 of which
were revisited, no more pellets were obtained
after the initial collection. The pellets from 8
(CRE, AD, EUL, PD, K, ED, B and S) of these
16 localities were probably not of recent origin;
they had lost their glaze and no barn owls were
present at the sites at the time the pellets were
collected. At some localities (e.g. WP, A, and
M) where sequential collections were made the
first, usually large, collection contained both
unglazed and glazed pellets. However it was
not known if the barn owls, present at the time
the pellets were collected, had been roosting
at the site continuously over a long period,
or if the older (unglazed) pellets represented
an earlier accumulation. The pellets varied
greatly in size, ranging from balls about 2 cm in
diameter to cylinders about 8 cm in length. The
largest pellets were usually found to contain
the remains of a single large prey item e.g. either a
large rodent or bird.

The prey items identified included mammals,
birds, reptiles, frogs and arthropods (mostly

orthopterans). The contribution of each taxon
to the diet (Table 2, Fig. 1) was based, in the
case of vertebrate prey items, on the number
of pellets that contained countable remains,
i.e. remains upon which a count of the number
of individuals eaten could be made and, for
arthropods, the number of pellets in which they
were found. Mammals were present in 75.5% of
all pellets, birds in 26.2%, reptiles in 9.3%, frogs in
3.4% and arthropods in 22.1%. The contribution
of each taxon to the diet varied between localities
(Table 2). Mammals were represented in the
pellet collection from every locality, and all taxa
in the collections from 16 localities. Birds were not
found in the pellets from two localities, reptiles
from seven, frogs from nine and arthropods
from two.

Mammals. The species identified included
three rodents ( R . villosissimus, L. forresti and the
introduced house mouse M. musculus), three
dasyurids (Planigale ingrami, S. macroura and S.
douglasi), two bats (a single Saccolnimus flaviventris
from Eulolo and one other, tentatively identified
as Mormopterus beccarii, from Huddersfield) and
one cat (Felis cntus) from Toorak. Identification of
the cat (a kitten) based on teeth was confirmed
by examination of the structure of hairs from
the same pellet. The last three species, of which
only a single individual of each was found, are not
further considered. No Sminthopsis crassicaudnta,
a species known to the author to occur to the
east of the study area at Leslew Downs (20°59'S,
142°55'E) and to the south-west at McKinlay
(21°16'S, 14T17'E), were found among the remains
of mammals in the pellets.

Rattus villosissimus (mass up to 280 g) is the larg-
est of the mammals frequently preyed upon and
up to four individuals, but never more than
two with adult dentition, were found in a single
pellet. Many pellets contained recognisable hair
and post-cranial bones of this species but not the
countable remains (dentaries). Of the smaller
rodents as many as six L. forresti (mass up to
20 g), or five M. musculus (mass up to 25 g),

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

111

Woolley

TABLE 3. The number of mammals in the pellets from each locality, and the number of pellets containing the
remains of each of the six species of mammals in each locality. The number of individuals of each species is
shown in parentheses and this number, as a percentage of the "mammals in the locality, in square parentheses.
For all localities the number of pellets, as a percentage of the total number collected (3,749), and the number
of each species as a percentage of the total number of mammals (4,793) is shown in bold figures. Key to
localities in Table 1.

['Locality (no.

No. of pellets containing remains of each species, |% of mammals], (no. of individuals)

mammals)

R. villosissimus

L forresti

M. musculus

P. ingrami

S. macroura

S. douglasi

EUR (295)

201 (81.0) (239)

14 [7.11 (21)

20 [8.5] (25)

1 [03] (1)

5 [3.1] (9)

CRH (10)

2 120.0] (2)

1 [10.0] (1)

8 [70.0] (7)

] N (185)

63 [36.2] (67)

7 [3.8] (7)

21111(2)

52 [50.8] (94)

7 [43] (8)

7 [3.8] (7)

NW (85)

56 (85.9] (73)

1 [1.2] (1)

5 18.2[ (7)

1 [1*2] (1)

3 [3.5] P)

j 0(215)

96 [51.6| (111)

17 [9.8| (21)

31 [25.11(54)

13 [8.8] (19)

7 [4.7] (10)

L (131)

83 [91 .6] (120)

9 [7.6] (10)

1 [0.8] (1)

j AD (18)

3 |16.7] (3)

8 [77.8] (14)

1 15.51 (1)

|| H (277)

152 157.8] (160)

12 [4.71 (U)

38 [19.1| (53)

6 [15] (7)

39 [15.9] (44)

i EUL (635)

47 [8.7] (55)

194 [52.0] (330)

26 [6.8] (43)

43 [10.2| (65)

98 [223] (141)

1 [0.1| (1)

PD (271)

38 (15.5] (42)

85 [72.7] (197)

3 [l-H (3)

7 [3.3] (9)

12 [7.0] (19)

1 [0.4] (1)

| CAR (18)

4 |33.3] (6)

3 [16.7] (3)

1 [11.1| (2)

4 [38.9] (7)

ARM (33)

10 139.4] (13)

7 [45.5] (15)

4 [15.1] (5)

I WP(295)

88 |353] (104)

1 [0.3] (1)

3 [1.0] (3)

81 [60.7] (179)

3(101(3)

5 117] (5)

k<8)

2 [87.5| (7)

1 [12.5] (1)

EC (20)

15 ]80.0] (16)

3 (15.0| (3)

1 [5.01(1)

[ LP (80)

53 [90.0] (72)

1 [1.251 (1)

3 [5.0|(4)

1 [1.251 (1)

2 [2.5] (2)

| T (252)

125 [54.3] (137)

35 [18.6] (47)

6 [2.4] (6)

26 [15.1] (38)

8 14.41(11)

13 [5.2] (13)

ED (10)

3 [30.0] (3)

5 [50.0] (5)

1 [10.01 (1)

1 [10.0] (1)

ARD (664)

237 [39.4] (262)

6 [0.9] (6)

140 [55.0] (365)

18 [35] (23)

8 [1.2] (8)

CON (141)

87 172.3| (102)

2 [4.3] (6)

8 [220] pi)

1 1 1.4) (2)

B (47)

22 149.0( (23)

13 [44.7] (21)

1 [2.11(1)

2 [4-2] (2)

M (613)

205 [40.81 (250)

31 [83] (51)

33 [10.9] (67)

115 [32-5| (199)

8 [1.5] (9)

29 [6.0] (37)

CAN (75)

57 [77.4] (58)

1 [1.3] (1)

9 [20.0] (15)

1 P-3] (1)

S (13)

7 [ 7 6.9] (10)

3 [231] (3)

j Q (232)

79 [37.1] (86)

5 [2.61 (6)

79 [47.4] (110)

19 [8.2] (19)

10 [4.7] (11)

| P (90)

21 126.7] (24)

4 [53] (5)

11 [16.7] (15)

7 18-91 (8)

31 (42.2) (38)

YD (9)

8 [88.9] (8)

i inn a)

R (71)

58 |84.6] (60)

3 [7.0[ (5)

1 [5.6] (4)

1 [1 41 (1)

1 [1.4] (1)

J All localities

1810 (2093)

437 (745)

75 (126)

720 (1341)

215 (286)

172 (202)

| % pellets

48.3

11.6

2.0

19.2

5.7

4.6

% mammals

43.7

15.5

2.6

28.0

6.0

4.2

were found in a pellet. Up to 14 individuals of
the smallest dasyurid preyed upon, P. ingrami
(mass up to 5 g), but usually only one or two
of the larger dasyurids, S. macroura (mass up
to 25 g) and S. douglasi (mass up to 70 g), were
found in a pellet. If larger numbers of either S.
macroura or S. douglasi were found they were
always individuals assessed as juveniles. It was

not uncommon to find the remains of more than
one species of mammal in a single pellet. R.
mllosissimus was found in 48.3% of all pellets, L.
forresti in 11.6%, M. musculus in 2.0%, P. ingrami
in 19.2%, S. macroura in 5.7% and S. douglasi in
4.6% (Table 3, Figure 2). The localities at which
the remains of each of these species were found
in the pellets can be seen in Figure 3.

112

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Prey of barn owl in Queensland

Birds. Seventeen species were identified. Of
the 2053 individuals found in the pellets 30
(1.4%) could not be identified. The species most
commonly preyed upon was the Zebra Finch,
Taetticrpygia guttata (75.9% of the total number of
birds), followed by tine Budgerigar, Melopsittacus
undulatus (7.8%); Black-faced Woodswallow,
Artamus cinereus (4.2%); White-winged Triller,
Lalage sueurii (2.5%); Red-capped Robin, Petroica
goodenovii (2.3%); White-winged Fairy-wren,
Malurus leucopterus (1.6%); Singing Bushlark,
Mirafra javanica (1.6%) as well as, each at less
than 1.0%, swallows, Hirundo sp.; Golden-
headed Cisticola, Cisticola exilis ; Little Button-
quail, Turnix velox ; Red-chested Button-
quail, T. pyrrlwtlwrax ; Baillon's Crake, Porzana
pusilla ; Brown Songlark, Cincloramphus cruralis;
Rufous Songlark, C. mathewsi; Spiny-cheeked
Honeyeater, Acanthagenys rufogularis ; pigeons,
Phaps sp. and Australian Pratincole, Stiltia
isabclla. Up to seven zebra finches (mass 12-13
g) were found in a single pellet, but seldom
more than two of any other species.

Reptiles. Representatives of three families, the
Gekkonidae, Agamidae and Scincidae, were
found in the pellets. Most of the jaw bones,
upon which the count of individuals was made,
were very delicate and some may have been
overlooked in the preparation of the pellets.
Thus the total number of individuals found
(907) may be an underestimate of the extent
to which reptiles are preyed upon. Some of
the 907 specimens could not be identified, and
some were not placed to family with certainty
but the majority were considered to be geckoes
(762 individuals), followed by agamids (85)
and skinks (11). Frequently large numbers
of geckoes (up to 18) were found in a single
pellet. Geckoes were often seen in numbers
at roosting sites in old buildings and so may
have been very readily accessible to resident
owls. Confirmation of identity for some of
the agamids was provided by the presence of
portions of tail, still covered with rough skin.

FIG. 2. Number of pellets containing each species of
rodent (R. villosissimus, L. forresti and M. musculus)
and dasyurid (S. douglasi, S. macroura and P. ingrami)
preyed upon oy the owls. The number of pellets,
expressed as a percentage of all pellets examined
(3,749), is shown at the end of each bar together with
the number of individuals in parentheses.

One agamid was further identified as a juvenile
Pogona brevis (G. J. Witten pers. comm.).

Frogs. The remains of 373 frogs were found in
the pellets. The ilium, upon which the estimate
of numbers was based, ranged in length from 11
to 33 mm. Up to 20 individuals with short (13-
14 mm) ilia were found in some pellets. Some
indication of the size of the frogs preyed upon
by the owls can be gained from measurements
taken from the dry remains of an unidentified
frog found in the study area. The body length
of this specimen was about 75 mm, and the
length of the ilium, 26.5 mm.

Arthropods. The remains of a few large spiders
(O. Araneae) and one centipede (not further
identified) were found in the pellets but the
majority of the arthropods were insects,
including beetles (O. Coleoptera), bugs (O.
Hemiptera), mantids (O. Mantodea), and crickets
and grasshoppers (O. Orthoptera). Grasshoppers
were by far the most numerous of the arthropod
prey items, and their eggs were frequently seen
in the pellets. Some of the beetles and bugs were
very small and they may have been ingested
along with other prey such as frogs.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

113

Woolley

Other items. Seeds considered to have been
ingested along with a prey item were found in
many pellets, especially those containing bird
remains. The large oval seeds of the paddy
melon, Cucumis melo, were often found in pellets
that contained the remains of R. villosissimus.

DISCUSSION

Consistent with the findings of studies on
the diet of barn owls referred to above the prey
of the owls in the study area in north-western
Queensland included a variety of mammals,
birds, reptiles, frogs and arthropods, with
mammals forming the major component of the
diet. The remains of individual mammals were
found in three quarters of all pellets examined,
and mammals were taken in greater numbers
than individuals of other vertebrate taxa. All six
species of small terrestrial mammals (3 rodents, 3
dasyurids) preyed upon were previously known
to occur in the study area.

The most frequently taken species was R.
villosissimus (43.7% of all mammals) and it was
found in pellets from all but three localities.
The area around Julia Creek is thought to be
a 'refuge' area for R. villosissimus when it is
not in plague proportions (Carstairs 1974).
This species has been found to form a large
component of the diet of the Barn Owl in some
other studies e.g. Morton et al., 1977 (41.6%);
Valente, 1981 (60.9%); Debus et al„ 1999 (65%).

Smaller contributions to the diet were made
by the other two species of rodents. In the case
of L. forresti (15.5% of all mammals) the majority
(527 of the 745 individuals) were found in
pellets collected in July 1994 from just two of the
nineteen localities in which it was detected. In
one of the two, Eulolo, they constituted 52% of
all mammals from that locality and at the other,
Penola Downs, 73%. Although this species is
seldom encountered in large numbers (Watts
& Aslin 1981; Dickman et al. 2000) it seems
that it was plentiful in these localities at the
time the pellets accumulated. Morton & Martin

(1979) also found it to form a large component
of the diet of barn owls in 2 of their 11 study
sites (59% at Coober Pedy and 20% at Warrina).
The introduced house mouse, M. musculus, the
least frequently preyed upon species (2.6% of
all mammals) was found in pellets in only eight
localities and the majority (110 of 126 individuals)
were taken in just two; at Eulolo it constituted
6.8% of the mammals and at Minamere, 10.9%.
When numbers of house mice were high at
Fowler's Gap, N.S.W., they formed the bulk of
the diet (up to 96.4%) of the owls (Morton &
Martin 1979).

Planigale ingrami, the smallest of the three
dasyurids, was the second most frequently
preyed upon species (28% of all mammals) and
it was found in pellets from all localities except
one. Smmthopsis macroura (6% of all mammals)
was preyed upon a little more frequently than
S. douglasi (4.2%), and both species were found
throughout the study area in a majority (21,
22 respectively) of the localities from which
pellets were obtained. Sminthopsis douglasi has
not previously been recorded as prey of the
barn owl but S. macroura, which has a much
larger distribution, has been found in owl
pellets from other areas. Heywood & Pavey
(2002) found it to be the major prey item (81.8%
of identified items) in a sample of pellets from
Connell's Lagoon on the Barkly Tableland at a
time when numbers of R. villosissimus, a species
known to occur in the same locality, were not
found in the pellets. In the present study S.
macroura never formed more than 38.9 % of the
mammalian prey items at any one locality.

Among birds, the vertebrate taxon making
the second largest contribution to the diet of the
owls, the species most frequently preyed upon
was the Zebra Finch, T. guttata. It is not known
if this species was generally more abundant in
the study area than others, or if some aspect
of its behaviour makes it more susceptible
to predation. It is known to nest in colonies
throughout the year (Zann, 1996) and this may

114

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Prey of barn owl in Queensland

(a)

140 141 142 143

(b) Longitude (°E)

(d) Sm

(e) Pi

(f) Rv

(9) Lf

(h) Mm

FIG. 3. Distribution of the species of dasyurids and rodents based on remains found in barn owl pellets
collected in north-western Queensland, a) location of the study area; bl the collecting localities (see Table 1
for full names) around the town of Julia Creek (JC); c) Sd = S. aouglasi; a) Sm = S. macroura; e) Pi = P. ingrami ;
f) Rv = R. villosissimus; g) L/= L. forresti and f) Mm = M. museums. A filled symbol indicates a record of a
species; an open symbol, no record.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

115

Woolley

lead to several (up to seven) being taken at a
time: barn owls have been observed to disturb
communally roosting birds by beating their wings
against the bushes in which the birds are roosting
(Bunn et al. 1982).

No assessment of either seasonal or long term
changes in the availability of prey species at
the various localities within the study area can
be made because the pellets were collected
opportunistically over a period of 8-9 years, and
in many cases the period over which the pellets
had accumulated was not known. Differences
between localities in the relative importance
of the various prey taxa (see Tables 2 and 3)
probably reflect what was readily available
to the owls at the time they were present in
the area. Frogs may form a larger component
of the diet during wet periods, and insects
such as grasshoppers when local infestations
occur. The absence of a particular taxon from
some localities may be due in part to the small
number of pellets found at the locality.

From the information obtained on the distri-
bution of the three species of rodents (R.
villosissimus, L.forresti and M. musculus) and the
three dasyurids (S. douglasi, S. macroura and P.
ingrami) it seems likely that this suite of small
mammals, with the possible exception of M.
musculus, may be found throughout the study
area. The absence of a species from the pellets
from any one locality does not necessarily mean
that the species does not occur there. S. douglasi,
the species of particular interest, was not found
in pellets collected from Crendon in 1993, but
a live individual had been obtained from
that locality in 1992. To date, live animals or
carcasses of S. douglasi have been collected
from 6 of the 28 localities, including Lyrian,
Crendon, Nelia, Toorak, Proa and Yorkshire
Downs (Woolley 1992; Woolley & Mifsud
unpub. observations). Specimens have also been
collected from other localities within the study
area, including Euraba, Julia Creek, Edith Downs
and Euroka (Woolley 1992), and one was found

alive near Penola Downs (at 21°28'S, 141°10'E)
in the wet season of 1997-98 (S. Malone pers.
comm.). These localities all lie within the known
range of the Julia Creek Dunnart (Kutt 2003).

ACKNOWLEDGMENT

My thanks are due to many people including
all the property owners who allowed access to
collecting sites; to B. Spreadborough, S. Malone,

A. Alloway, S. O'Connor and especially G.
Mifsud for assistance with the collection of owl
pellets; to assistants, volunteers, students of
Marsupial Biology at La Trobe University and
the late Natalie Smith (who studied a subset of
the collection for her Honours project in 1995)
for help with the preparation of the pellets, and
to G. Richards (bats), W. Boles (birds), P. Couper
(reptiles), T. New and M. Cairns (arthropods), and

B. Triggs (hair) for assistance with identification
of material. As well as confirming the identity
of a large number of the specimens of L. forresti
A. Baynes provided valued assistance in the
compilation of data and ideas for the presentation
of the results. P. Green kindly assisted with the
preparation of Figure 3. Financial assistance
was received from La Trobe University and the
Queensland Department of Environment and
Heritage.

LITERATURE CITED

Archer, M. 1979. Two new species of Sminthopsis
Thomas (Dasyuridae: Marsupialia) from northern
Australia, S. butleri and S. douglasi. Australian
Zoologist 20: 327-345.

Bunn, D.S., Warburton, A.B. & Wilson, R.D.S. 1982.
The Bam Owl. (The Pitman Press: Bath.)

Carstairs, J.L. 1974. The distribution of Rattus
villosissimus (Waite) during plague and non-
plague years. Australian Wildlife Research 1
95-106.

Debus, S. J. S., Rose, A. B. & Harris, J. 1999. Diet of
the Bam Owl Tyto alba at the Diamantina Lakes,
western Queensland. Sunbird 29: 26-27.

Dickman, C.R., Leung, L.K.-P. & Van Dvck, S.M. 2000.
Status, ecological attributes and conservation

116

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Prey of barn owl in Queensland

of native rodents in Queensland. Wildlife Research
27: 333-346.

Heywood, M. R. & Pavey, C. R. 2002. Relative
importance of plague rodents and dasyurids as
prey of bam owls in central Australia. Wildlife
Research 29: 203-207.

Kutt, A.S. 2003. New records of the Julia Creek
Dunnart Sminthopsis douglasi in central-north
Queensland. Australian Zoologist 32: 257-260.

Morton, S.R. 1975. The diet of the Barn Owl Tyto alba
in southern Victoria. Emu 75: 31-34.

Morton, S.R., Happold, M., Lee, A.K. & MacMillan,
R.E. 1977. The diet of the Bam Owl, Tyto alba, in
south-western Queensland. Australian Wildlife
Research 4: 91-97.

Morton, S.R. & Martin, A. A. 1979. Feeding ecology
of the Barn Owl, Tyto alba, in arid southern
Australia. Australian Wildlife Research 6: 191-204.

Smith, J.D.B. & Cole, J. 1989. Diet of the Bam Owl, Tyto
alba, in the Tanami Desert, Northern Territory.
Australian Wildlife Research 16: 611-624.

Taylor, L 1994. Bam Owb: Predator-Prey Relationships and
Conservation. (University Press: Cambridge).
Valente, A. 1981. Vertebrate remains in pellets of the
Bam Owl, Tyto alba, from Planet Downs Station,
south-western Queensland. Australian Wildlife
Research 8: 181-185.

Watts CHS. & Aslin, H.J. 1981 . The Rodents of Australia.
(Angus and Robertson Publishers: Sydney).

Woolley, P. A. 1992. New records of the Julia Creek
Dunnart, Sminthopsis douglasi (Marsupialia:
Dasyuridae). Wildlife Research 19:779-783.

Zann, R. A. 1996. The Zebra Finch. A Synthesis of Field
and Laboratory Shidies. (Oxford University Press:
Melbourne).

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

117

Cook & McHenry

A homonym of Leptocleidus Andrews. 1922
(Sarcopterygia, Plesiosauria. Leptocleidoidea,

Leptocleididae) and a replacement name for Leptocleidus
Mueller 1936 (Platyhelminthes. Monogenea).

Leptocleidus Andrews 1922 is an iconic plesiosaurian
which is known from the Early Cretaceous of the United
Kingdom, South Africa and Australia and is the root genus
of the monophyletic family-level clade Leptocleididae (White
1940), as well as the superfamily-level clade Leptocleidoidea
(Druckenmiller & Russell 2008) and the rank free higher clade
Leptocleidia (Ketchum & Benson 2009). Andrews erected
the genus on L. superstes, from the Wealden Clay of the
United Kingdom

Independently Mueller (1936) erected Leptocleidus upon
the L. mcgalonchus for a taxon of platyhelminth flatworm.
Shortly after its designation Leptocleidus Muller was
synonvmised with Cleidodiscus Muller 1934 by Mizelle &
Hughes (1938) and later with Urocleidus Mueller 1936 by
Price (1968). This nomenclatural arrangement was utilised
until Leptocleidus Mueller was resurrected by Sullivan et al.
(1978), who asserted its place as an independent genus.

Leptocleidus Mueller is invalid as the name is preoccupied
and a new name must be chosen for the genus. We here
designate Muellerocleidus gen. nov. as a replacement name
for Leptocleidus Mueller 1936. The name honours Justus F.
Mueller.

This new name eliminates the homonym and any confusion
between the nomenclature of a Cretaceous plesiosaur and a
modern platyhelminth.

Literature cited

Andrews, C.W. 1922. Descriptions of a new plesiosaur from
the Weald Clay of Berwick (Sussex). Quarterly lournal of the
Geological Society of Loudon 78: 285-298.

Druckenmiller, P.S. & Russell, A.P. 2008. A phytogeny of Plesiosauria
(Sauropterygia) and its bearing on the systematic status of
Leptocleidus Andrews, 1922. Zootnxa 1863 1-120.

Ketchum, H.F. & Benson. R.B.J. 2009. Global interrelationships of
Plesiosauria (Reptilia, Sauropterygia) and the pivotal role of
taxon sampling in determining the outcome of phylogenetic
analyses. Biological Reviews 2009: doi 10.1111/jl469-
185X.2009.001 07.x.

Mizelle.. J.D., & Hughes, R.C. 1938. The North American freshwater
Tetraonchinae. Ameriam Midland Naturalist 20: 341-353.

Mueller, J.F. 1934. Parasites of Oneida Lake fishes, part IV. Roosevelt
Wild Life Annals 3(4):336-358. (not seen)

1936. Studies on North American Gyrcdactyloidea. Transactions of
the American Microscopical Society 55(1): 55-72.

Price, C.E. 1968. Notes on the trematode genera Cleidodiscus and
Urocleidus. Quarterly lournal of the Florida Academy of Science
30: 61-67.

Sullivan, J.R., Mayes, M.A., Rogers, W.A. & Becker D.A. 1978.
Resurrection and redescription of the genus Leptocleidus
Mueller 1936 (Monogenoidea) with notes on the habitat and
distribution of L, megalonchus. lournal of Parasitology 64(5)-
810-812.

White, T.E 1940. Holotype of Plesiosaurus longirostris Blake and the
classification of plesiosaurs, lournal of Paleontology 14:451-467.

Alex G. Cook, Queensland Museum, Geosciences, 122 Gerler

Rd, Hendra, Qld 4011 Australia & Colin R. McHenry, School of

Engineering, University of Newcastle, University Drive, Callaghan,

NSW, 2308, Australia. 4 December 2009.

118

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

The Tribe Dufouriini (Diptera: Tachinidae:
Dexiinae) recorded from Australia with
the description of two new species

Bryan K. CANTRELL

Biodiversity Program, Queensland Museum, PO Box 3300, South Brisbane, Qld 4101, Australia.
Email: bjlcantrell@ozemail.com.au

Chris J. BURWELL

Biodiversity Program, Queensland Museum, PO Box 3300, South Brisbane, Qld 4101, Australia.
Email: bjlcantrell@ozemail.com.au

Environmental Futures Centre and Griffith School of Environment, Griffith University, Nathan, Qld,
4111, Australia.

Citation: Cantrell, B.K. & Burwell, CJ. 2010 03 15. The tribe Dufouriini (Diptera: Tachinidae: Dexiinae)
recorded from Australia with the description of two new species. Memoirs of the Queensland
Museum — Nature 55(1): 119-131. Brisbane. ISSN 0079-8835. Accepted: 8 August 2009.

ABSTRACT

Rondania albipilosa sp. nov. and R. cinerea sp. nov. are described and Chetoptilia
angustifrons Mesnil is recorded from Australia, confirming the presence of the tribe
Dufouriini in the Australasian Region. Other Australian species of Rondania are known,
but remain undescribed because insufficient specimens are available to adequately
resolve species limits. The distributions of Chetoptilia Rondani and Rondania Robineau-
Desvoidy are extended to Australia. Two host records are noted, that of Chetoptilia
angustifrons emerging from prepupae of a chrysomelid beetle and Rondania cinerea
from an adult weevil. The presence of fully developed first instar larvae in the oviducts of
specimens of Rondania albipilosa and R. cinerea indicate ovolarviparity, consistent with
other Dexiinae. □ Tachinidae , Dexiinae, Dufouriini, Chetoptilia, Rondania, new species,
new record, Australia.

The Tachinidae form a conspicuous element of
the Australian dipteran fauna and are primarily
parasitoids of other insects. The subfamily
Dexiinae (= Proseninae of earlier authors) is
a relatively distinct group of tachinids whose
hosts are mainly Coleoptera (beetles). The
Dufouriini are currently regarded as a tribe
within the Dexiinae (see O'Hara & Wood 2004),
but the included genera are poorly understood
phylogenetically and have an unusual mix
of specialised characters, indicating that the
tribe is probably not monophyletic as presently
constituted.

Cantrell (1988) and Cantrell & Crosskey (1989)
noted the existence of specimens of Dufouriini (as

Dufouriinae) from Australia in collections, but
did not formally record the taxon from Australia
because no species were described. Even today,
the tribe is represented in Australian collections
by relatively small numbers of specimens (often
only of one sex) and the hosts remain unknown
for many species.

This study was prompted by examination of a
series of eleven adults of Chetoptilia angustifrons
Mesnil from Queensland that were reared by
CJB. Initially we thought that they represented a
new species, but Dr D.M. Wood (CNQ recognised
their similarity to C. angustifrons and loaned a
male and female of the latter from the Philippines

Memoirs of the Queensland Museum | Nature • 2010 • 55(1) • www.qm.qld.gov.au

119

Cantrell & Burwell

FIG. 1. Locality records for Australian Dufouriini. ■, Chetoptilia angustifrons Mesnil; ■ , Rondania albipilosa
sp. nov.; • , Rondania cinerea sp. nov.; A, undescribed species of Rondania.

for comparison. It was immediately clear that
all specimens were conspecific although the
two older specimens from the Philippines were
rather faded. This allowed the redescription
of C. angustifrons based largely on Australian
specimens.

The second part of this study was based on
examination of ca 80 specimens of Rondania spp.
from Australian insect collections (Fig. 1), enabling
the description of two new species from southern
Australia. However, other Australian species of
Rondania remain undescribed because we were
unable to resolve doubts about species limits from
study of the small number of available specimens.

While our aim was to record the presence of the
Dufouriini in Australia, we needed to consolidate

our knowledge of the tribe by studying
representatives of non-Australian genera and
comparing them against the Australian taxa. This
was achieved largely by referring to published
literature, but we also examined specimens of a
number of exotic taxa, including C. angustifrons
from the Philippines, Neotropical Comyops van
der Wulp and Ebenia Macquart and the European
species Dufouria clialybeata (Meigen), Eugynmopeza
braueri Townsend, Rondania cucullata Robineau-
Desvoidy and R.fasciata (Macquart).

MATERIAL AND METHODS

Morphological terminology and abbreviations
follow Crosskey (1973, 1976), except that ac not
acr is used for the acrostichial setae and the

120

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Australian Dufouriini (Diptera: Tachinidae: Dexiinae)

TABLE 1. Regional distribution of world genera of Dufouriini. Abbreviations: Neo, Neotropical; Nea,
Nearctic; Palae, Palaearctic; Afro, Afrotropical; Aust, Australasian; Or, Oriental.

Genus

Neo

Nea

Palae

Afro

Aust

Comyops

✓

Ebenia

✓

Euoestrophasia

✓

Jamacaria

✓

Oeslropliasia

✓

Chetoptilia

✓

Dufburia

✓

Eugyrmiopeza

✓

1 Microsoma

✓

Pandelleia

✓

Plesina

✓

Romlania

✓

1 Mesnilana

✓

! Khinophoroidcs

✓

word seta or setae is omitted in combination
with the relevant abbreviation e.g. pra means
pre-alar seta.

Morphological abbreviations: ac, acrostichial
setae; ad, anterodorsal; A.s., antennal segment;
av, anteroventral; dc, dorsocentral setae; fr, frontal
setae; ia, intra-alar setae; if, interfrontal area; iv,
inner vertical setae; mmp, median marginal pair
of setae on abdominal tergites; oc, ocellar setae;
orb, orbital setae; ov, outer vertical setae; pd,
posterodorsal; pf, parafacials; pfr, parafrontals;
pra, pre-alar setae; post-, postsutural; pv,
posteroventral; sa, supra-alar setae; stpl, stem-
opleural setae; T, abdominal tergite; v, ventral.

All measurements are in millimetres. Measure-
ments of body length and V:HW (ratio of width
of vertex at level of posterior ocelli to maximum
width of head across eyes, both viewed dorsally)
are expressed as means with ranges given in
brackets; number of specimens measured is
also shown.

Institutional abbreviations: ANIC, Australian
National Insect Collection, Canberra; CNC,

Canadian National Collection of Insects, Ottawa;
QM, Queensland Museum, Brisbane; UQIC,
University of Queensland Insect Collection, Bris-
bane; WAM, Western Australian Museum, Perth.

SYSTEMATIC^

As stated above, an in-depth study of the
Dufouriini was never our intention. However,
our literature- and limited specimen-based review
of the tribe allowed us to place the Australian
species within Chetoptilia and Rondania with a
degree of confidence.

The Dufouriini are known from all major
zoogeographic regions except Oceania
(Barraclough 2005; Cantrell & Crosskey 1989;
Crosskey 1976, 1984; Guimaraes 1977 (as
Oestrophasiini); Herting 1984; O'Hara & Wood
2004) with the greatest generic diversity in
the Palaearctic, Afrotropical and Neotropical
regions (Table 1). The genera are diverse in
appearance and the tribe is almost certainly
not monophyletic as presently constituted.
However, all genera for which the hosts are

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

121

Cantrell & Burwell

known parasitise Coleoptera, particularly
Chrysomelidae and Curculionidae.

The composition of Dufouriini remains open
to debate as several genera remain poorly
studied. Based upon our brief overview, we
include fourteen genera in the tribe, distributed
as shown in Table 1. Anthomyiopsis Townsend,
Freraea Robineau-Desvoidy and Rossimyiops
Mesnil are no longer regarded as Dufouriines
(Herting 1984; O'Hara and Wood 2004; Cerretti
et al. 2009) and we exclude the problematic
genus Kambailimyia Mesnil. The doubtful posi-
tion of the latter was discussed by Crosskey
(1976) when he provisionally included it in the
Oriental Dufouriini.

DIAGNOSIS OF AUSTRALIAN DUFOURIINI

The Australian Dexiinae comprise three tribes,
Dexiini, Dufouriini and Rutiliini. As inferred
above, it is difficult to succinctly characterise
the Dufouriini to facilitate their easy recognition
and the Australian fauna is no exception. Their
muscoid facies and enlarged subscutellum
clearly identify them as Tachinidae, but
assigning specimens to a subfamily or tribe is
often problematic, particularly if host data are
lacking. Small size (length 3-5 mm), bare eyes
and a long petiolate cell R5 will help to identify
Rondania as belonging to the Dexiinae, as will
the small size (4-5 mm), bare eyes (with enlarged
upper facets in the male), metallic blue/ green
colour and cell R5 open or just closed at the
wing margin in Clietoptilia. Knowledge that
the flies were reared from beetle hosts would
help confirm placement in the Dexiinae.

The Dufouriini may be distinguished from the
Australian Dexiini and Rutiliini by characters
including small size, bare parafacials, lack of a
facial carina, bare propleuron, scutellum with
2-3 pairs of marginal setae and their generally
non-bristly facies.

Diagnosis. Head: eye bare, strongly approx-
imated or holoptic (with upper eye facets
sometimes enlarged) in but widely separated
in $ (facets of uniform size); ocellar triangle
prominent in 5 because of eye approximation;
facial carina absent; antennal axis at or below
level of eye middle; epistome not prominent,
but sometimes extending anterior to profrons
in profile; facial ridge bare; pf bare; antenna
short, usually falling well short of epistome;
arista micropubescent to plumose; oc present;
orb present in ?, absent in cJ; iv usually present
in both sexes, but often weak and hairlike in J;
ov variable, often absent or indistinguishable
from postocular setulae; fr present in both sexes,
usually in irregular rows; vibrissa present, but
often not clearly distinguished from other setae
on genal margin; palp present, well-developed;
mouthparts (mentum plus labellum) normally
developed, shorter than head height.

Thorax: prosternum and propleuron bare;
chaetotaxy often irregular, with variation in
both numbers and strength of setae, particularly
dorsally; humeral callus usually with 2 setae; pra
absent or weakly present, smaller than first post-
ia; usually 2 post-ia; 2(3)+3 dc; l(2)+2(3) ac; 2
stpl, anterior seta usually weaker; scutellum
with 2 or 3 pairs of marginal setae, usually
with strong apicals and basals, weaker laterals
present or absent; wing with cell R 5 narrowly
open, closed at the margin, or petiolate; leg
setae often reduced.

Abdomen: Tl+2 variable, from weakly excavate
to excavate virtually to hind margin; chaetotaxy
often irregular, with variation in both numbers
and strength of setae; ovipositor a simple
eversible tube, typically concealed in Clietoptilia,
but usually at least partly protruding from
preabdomen in Rondania.

122

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Australian Dufouriini (Diptera: Tachinidae: Dexiinae)

KEY TO AUSTRALIAN GENERA
AND SPECIES OF DUFOURIINI

Users of this key should be aware that there

are several undescribed Australian species of

Rondania (see discussion under Rondania below).

1 . Arista plumose; cell R 5 narrowly open or just

closed at wing margin; bend of M rounded
and vein approaching wing margin at an
acute angle (Fig. 2A); Tl+2 excavate almost
to hind margin; thorax, including legs, and
abdomen dark, conspicuously shining black
or green-black; thorax with thin dusting
of silver pollinosity under some angles of
light, particularly in 3 eyes holoptic and
upper eye facets conspicuously enlarged
(Fig. 2B); ovipositor not usually visible
externally Chetoptilia angustifrons Mesnil.

— Arista bare or pubescent; cell R 5 closed

and distinctly petiolate; bend of M evenly
rounded and vein approaching R 4+5 at almost
a right angle (Fig. 4A, B); T1 +2 excavate
at most in basal half; body colour grey-
toned often with pale or yellow areas on
head, legs or abdomen; 3 eyes holoptic or
dichoptic, with or without enlarged upper
eye facets; ovipositor normally visible as a
simple tubular structure extending beneath
abdomen ( Rondania spp.) 2

2. Abdomen grey (with darker rings around

setal bases) in $, in 3 similar medially and
distally but lateral areas of Tl+2 to T4
yellow or pale; [thoracic pleura dark-haired;
scutellum with 3 pairs of marginal setae,
weak laterals present; 3 eyes narrowly
separated by distance approximately equal
to that between lateral ocelli; ? vertex broader
(V:HW > 0.35)] R. cinerea sp. nov.

— Abdomen predominantly pale or yellow 3

3. Abdomen yellow, without pattern of dark
spots or bands; thoracic pleura pale-haired;
scutellum with 2 pairs of marginal setae,
laterals absent; [cj eyes holoptic; ? vertex

relatively narrow (V:HW < 0.25)] R .

albipilosa sp. nov.

— Abdomen yellow, often shining, with pattern
of dark markings, typically small median and
lateral spots or transverse bands on T3 to
T5, but sometimes limited to median spots;
if dark abdominal markings absent, then
either scutellum with 3 pairs of marginal

setae or pleura dark-haired undescribed

species of Rondania.

Chetoptilia Rondani

Chetoptilia Rondani, 1862: 166; Crosskey, 1976: 176; Herting,
1984: 159.

Chetoptilia is an Old World genus comprising
six species: C. puella (Rondani) from Europe (type
species); C. plumicornis Villeneuve from Africa;
C. cyanea Mesnil and C. metallica Mesnil from
Madagascar; C. bunnanica (Baranov) from Asia;
and C. angustifrons from Asia and Australia.
Chetoptilia species are characterised by dark
metallic colouration, a plumose or pubescent
arista, a bare prosternum, three pairs of
marginal scutellar setae, cell R 5 open or just
closed at the wing margin and Tl+2 excavate
almost to the hind margin. The eyes are bare
(sparsely short-haired at high magnification),
holoptic or closely approximated in the male,
with the upper facets enlarged.

Chetoptilia angustifrons Mesnil
(Figs 1, 2, 3A-B)

Chetoptilia angustifrons Mesnil, 1953: 164; Crosskey, 1976: 176.

Material. Philippines, 13, Luzon, Manila, xi.1914
(CNC); 1 ., Luzon, Limay, 21. iv. 1913, G. Boettcher
(? type, abdomen missing) (CNC). [Both specimens
from the L. P. Mesnil collection labelled Paraptilops
angustifrons Mesn.]. AUSTRALIA, QUEENSLAND, 4 3,
19, Ellis Beach, 16°44'S, 145°WE, 28.iv.1998, C.J. Burwell
and C.M. Rodriguez, ex prepupa of Aspidimorpha
deusta on Ipomaea pes-caprae (Convolvulaceae); 33,
29, Bramston Beach, 17°21'S, 146°01'E, 3. v. 1998,
C.J. Burwell and C.M. Rodriguez, ex prepupa
of Aspidimorpha deusta on Ipomaea pes-caprae
(Convolvulaceae); 1 3 , 16 km N of Boonah, 27°54'S
152°41'E, 14-15.xii.1996, C.J. Burwell, ex prepupa of
Cassida sp. on Polymeria calycina (Convolvulaceae).
All QM except as indicated.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

123

Cantrell & Burwell

FIG. 2. Chetoptilia angustifrons Mesnil. A, 9, dorsal view; B, S head, lateral view; C, J terminalia, lateral view.

Diagnosis.' Small flies, shining green-black,
but prescutum and scutum with fine dusting
of silver pollinosity in female; tibiae and tarsi
somewhat duller; antenna and palp dull yellow/
orange; arista plumose; face dull silver
pollinose; eyes bare (sparsely short-haired at high
magnification); c J eyes holoptic, eyes widely
separated; occiput pale-haired except for some
dark setulae adjacent to postocular row; Tl+2
to T5 lacking discal setae; ovipositor fully retracted
in all available specimens.

Re-description. Male. Body length 4.62 (4.1 -
4.8) (n=9). Head: eyes holoptic, facets enlarged
in slightly more than upper half, with clear
line of demarcation between upper and lower
facets (Fig. 2B); if area triangular, reddish-
brown; pf, pfr and face faintly silver pollinose;
ocellar triangle prominent; oc weak, hairlike,
proclinate; 4-5 pairs of irregular fr between lunula
and eye convergence; orb absent; iv convergent,
often crossed at tips; ov not much stronger than
postocular setulae; antenna yellow, inserted at or
below level of middle of eye, short, not reaching

124

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Australian Dufouriini (Diptera: Tachinidae: Dexiinae)

epistome; arista yellow, plumose; epistome not
prominent; vibrissa inserted at level of epistome,
a few fine hairs above; facial ridge bare; pf bare;
gena black-haired; mentum relatively short, not
as long as palp, latter yellow with darker bases;
occiput dark, pale-haired except for some dark
setulae adjacent to postocular row.

Thorax shining black; small pra present, much
weaker than first post-ia; 1+2 ia; 2(?3)+3 dc;
l(?2)+2 ac; propleuron and prosternum bare;
2 stpl, anterior seta weaker; humeral callus
with 2 setae; scutellum with 3 pairs marginal
setae (strong crossed apicals and slightly
weaker basals and laterals); fore tibia with 1
pv; mid tibia without submedian v but with
1 submedian ad and 2 weaker pd; hind tibia
with irregular ad and pd rows; wing with cell
R 3 open or just closed at wing margin; m-cu
weakly sinuate, shorter than length of M from
insertion of m-cu to bend; 2 ntl costal sector
haired ventrally; basal node of R 4+5 with a few
fine dark setulae on both wing surfaces; veins
yellow; wing membrane clear, but upper and
lower calypter opaque, smoky-coloured.

Abdomen shining black; Tl+2 excavate almost
to hind margin, without mmp; T3 with mmp;
T4, 5 with marginal row of setae; T3-5 subequal
in length dorsally. Terminalia as in Fig. 2C.

Female. (Fig. 2A) Body length 4.3 (4.2 - 4.4) (n=3);
V:HW 0.25 (0.24 - 0.26) (n=4). Similar to
male except as noted. Head: eyes broadly
separated, all facets of uniform size, if area matt
reddish-brown, pf, pfr and face faintly silver
pollinose; gena and area between eye margin
and postocular row of fine setulae lightly silver
pollinose; 4-6 irregular fr; 2 pro-, 1 reclinate
orb; oc strong, proclinate; iv strong, crossed;
ov weaker, diverging; A.s.3 relatively longer, its
tip almost reaching epistome. Thorax: prescutum
and scutum with fine dusting of silver pollinosity;
mid tibia with submedian v; upper and lower

FIG. 3. Puparia. A, Chetoptilia angustifrons Mesnil,
dorsal view showing puparium in remains of beetle
host prepupa; B, C, ventral view of posterior part of
puparium showing non-functional posterior larval
spiracles: B, Chetoptilia angustifrons Mesnil C, Rondatiia
cinerea sp. nov. as, anterior (larval) spiracles; he, head
capsule of beetle prepupa; p, puparium; ps, posterior
(larval) spiracles.

calypter translucent white. Abdomen: ovipositor
not visible externally in specimens examined.

Puparium. Oblong, brown, with black non-
functional posterior larval spiracles raised above
puparial cuticle (Fig. 3B), in end view larval
spiracles semicircular, narrowly separated, each
comprising about 12 elongate slits arranged in a
crescent; non-functional anterior larval spiracles
projecting anteriorly on diverging cuticular
protuberances, each spiracle with a narrow

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

125

Cantrell & Burwell

black base and bearing about 10 terminal minute
openings. Puparial respiratory horns absent.

The puparium lies within remains of host
prepupa (Fig. 3A), with both the anterior and
posterior ends of the puparium erupting through
the ventral cuticle of the host prepupa. The
posterior end projects from between the bases of
the fore and mid legs of the host prepupa so that
the posterior (larval) spiracles on the puparium
are located just behind the head capsule of
the beetle. The anterior end of the puparium
projects between the 5 th and 6 th or the 6*’ 1 and
7 th abdominal segments of the host prepupa,
and the anterior (larval) spiracles are clearly
exposed.

Remarks. Chetoptilia angustifrons possesses all
the generic characters noted above and was
reared from a chrysomelid leaf beetle larva; this
host association is also typical of the genus. Based
on literature (Baranov 1938; Mesnil 1953, 1968;
Tschnorsnig and Herting 1994; Villeneuve 1942)
the plumose arista separates C. angustifrons from
C. cyanea and C. metallica from Madagascar and
C. burmanica from Myanmar, all of which have
a pubescent arista. The holoptic eyes in the
male and lack of discal setae on T5 distinguish
C. angustifrons from the European C. puella and
C. plumicornis from Africa.

As noted above, the north Queensland
specimens of C. angustifrons were reared from
Aspidimorpha deusta, a hispine chrysomelid beetle
which occurs in northern Australia, New Guinea,
Timor, Indonesia and Malaysia (Borowiec 1992;
Boroweic & Swiytojariska 2008), suggesting the
fly may have a wider distribution than presently
known. Like C. angustifrons, A. deusta has been
recorded from the Philippines (Leyte Island;
Medvedev 1995), although Medvedev suggested
the beetle had been introduced. Consequently,
the identity of the host of C. angustifrons in the
Philippines requires confirmation.

Rondania Robineau-Desvoidy

Rondania Robineau-Desvoidy, 1850: 192; Herting, 1984: 160.

Rondania is primarily an Old World genus,
comprising nine species including R. albipilosa sp.
nov. and R. cinerea sp. nov. described below from
Australia; R. cucullata Robineau-Desvoidy (type
species), R. ditnidiata (Meigen), R. dispar (Dufour),
R.fasciata (Macquart) and R. rubcns Herting from
mainland Europe; plus R. insularis (Bigot) (Canary
Islands). The sole New World representative is
R. dorsalis (Coquillett) from North America. We
examined specimens of R. cucidlata and R.fasciata
in addition to the Australian species. Several
undescribed Australian species of Rondania are
also known, but cannot be named at present
because insufficient specimens are available to
accurately resolve species limits.

Rondania is characterised by grey and yellow/
orange colouration; a pubescent or bare arista;
cell R 5 closed at the wing margin or petiolate
and Tl+2 excavate only basally.

All Australian specimens of Rondania examined
were collected below the Tropic of Capricorn
(23°26.5'S), mostly from mainland Australia
but with a few records from Tasmania (Fig. 1).
Rondania cinerea sp. nov. is a distinctive species
with a grey poilinose abdomen, described from
the type series collected in southern Western
Australia, supplemented by specimens collected
in Canberra and a few intervening localities.
Rondania albipilosa sp. nov. is a smaller species
with pale-haired thoracic pleura; it is known
only from Western Australia, where the
distributions of both species overlap to some
extent in the Esperance to Madura region.

Confidently defining the species limits of
the two Rondania species described below is
complicated by a number of specimens which
closely resemble those of the type series but
differ in abdominal colour pattern, having
more extensive dark markings. Consequently
we have taken a conservative approach and
excluded these specimens from the type

126

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Australian Dufouriini (Diptera: Tachinidae: Dexiinae)

FIG. 4. Rondania spp. nov. A, C, R. albipilosa sp. nov.: A, $ lateral view; C, $ head, dorsal view; B, D, E, R
cinerea sp. nov.: B, $ lateral view; D, $ head, dorsal view; E, cj terminalia, lateral view.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

127

Cantrell & Burwell

series, but briefly discuss them in the remarks
sections of the descriptions. Similar variations
in abdominal colour pattern are also found in a
number of the undescribed Australian species,
as well as variation in the length of the petiole
of cell Rg. Whether colour patterns and petiole
length are highly variable intraspecifically,
or these specimens are specifically distinct is
uncertain and may only be resolved by the
acquisition of additional material.

Rondania albipilosa sp. nov.

(Figs 1, 4 A, C)

Etymology. Latin albus, white and pilus, hair; referring
to the pale-haired pleura in this species.

Material. Holotype > Western Australia, The Humps,
19 km N Hyden, 32°19'S, 118°57E, 1011986, G. and A.
Daniels, mv lamp (QM). PARATYPES: WESTERN
AUSTRALIA, 49, same data as holotype (UQ1C);
19, 57 km S of Norseman, 32°38'S, 121°32'E,
31.xii.1985, G. and A. Daniels, mv lamp (UQIC);
Id, 2$, 8 miles (13 km) SW Mt Ragged, 33°27'S,
123°28'E, 22.iii.1968, l.F.B. Common and M.S.
Upton (ANIC). OTHER MATERIAL: WESTERN
AUSTRALIA, 19, Madura, 31°56'S, 126°58'E,
20.iii.1968, l.F.B Common & M.S. Upton (abdomen
in glycerine) (ANIC).

Diagnosis. Small flies; head, legs and abdomen
yellow, thorax grey pollinose on prescutum and
scutum, but scutellum dark shining, without
pollinosity; pleura grey pollinose, pale-haired;
chaetotaxy difficult to discern on prescutum
and scutum of $, except for posterior dc and ac;
scutellum with 2 pairs of marginal setae; eyes
holoptic, $ eyes narrowly separated.

Description. Female. (Fig. 4A, C) Body length 2.8
(2.4 - 3.0) (n=4); V:HW 0.215 (0.20 - 0.23)
(n=9). Head: generally pale; if area relatively
narrow, matt yellow; pfr and pf yellow with
thin silver pollinosity; ocellar triangle dark;
upper occiput dark except centrally; eye bare;
oc proclinate; 2 proclinate orb; 5-6 pairs fr; small
iv present, convergent but not crossed; ov not
discernible from postocular setulae; antenna
yellow-orange, arista somewhat darker; setae

and hairs of head generally pale, including weak
vibrissa, but upper fr, orb, oc and iv darker.

Thorax densely grey pollinose on prescutum
and scutum, but scutellum dark shining, without
pollinosity; pair of thin medial dark vittae on
prescutum; chaetotaxy difficult to discern on
prescutum and scutum except for posterior dc
and ac; scutellum with 2 pairs of marginal setae
(apicals and basals); stpl 0+1; dorsal setae and
hairs dark, those on pleura pale; legs yellow,
mostly pale-haired; fore femur with pd and
pv setal rows; mid tibia without submedian
v seta; hind femur with ad setal row; cell Rg
closed, petiole at least as long as length of M
from bend to its junction with R 4+5 ; row of
approximately 5 dark setulae on R 4+5 between
node and r-m crossvein; m-cu joining M much
closer to r-m than bend.

Abdomen yellow-brown, without pollinosity,
dark-haired, without discernible setae; basal
half of Tl+2 excavate; narrow dark bands
on posterior margins of T4 and T5 in some
specimens; pale tubular ovipositor exserted
and recurved beneath body in many specimens,
sometimes extending beyond head; T5 longer
than T4.

Male. Body length 4.0 (n=l). Similar to 9 except
as noted. Head: eyes holoptic, facets enlarged
in slightly more than upper halves; if area trian-
gular due to eye convergence; ocellar triangle and
vertex dark; ocellar triangle prominent, with weak
proclinate oc; 7-8 pairs of weak fr between level of
lunula and eye convergence; antenna yellow, but
arista and dorsal edge of A.s.3 somewhat darker;
A.s.l and 2 with black setulae dorsally; poorly
differentiated vibrissa and setae along margin
of oral cavity black; epistome and profrons
coincident in profile; gena with black hairs
anteriorly, otherwise pale-haired; palp and
mentum yellow.

Thorax finely silver pollinose dorsally on
prescutum and scutum; 4 narrow dark vittae
on prescutum and scutum; dorsal setae dark;

128

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Australian Dufouriini (Diptera: Tachinidae: Dexiinae)

pleura grey pollinose, pale-haired; thoracic setae
stronger than in $: humeral callus with 2 setae; pra
absent; 1 sa; 1 weak post-ia near suture separating
postalar calli; 2+4 dc, posterior strongest; 2+?3 ac;
legs yellow with black hairing; hind tibia with
sparse ad and pd setal rows.

Remarks. The following females in ANIC are
similar to those of R. albipilosa but have more
pronounced dark markings on the abdomen
as indicated below. The taxonomic significance
of these variant abdominal colour patterns
is undetermined. MATERIAL. WESTERN
AUSTRALIA, 1?, 1 km NNW Eucla Pass, 31°05'S,
128°52'E, 10. iv. 1983, E.S. Nielsen and E.D.
Edwards - thin dark marginal bands on Tl+2-
5; 19, 7 km S by E of Albany, 35°05'S, 117°54'E,
18. iv. 1983, E.S. Nielsen and E.D. Edwards -
posterior portion on T3 and all T4 and T5
dark. SOUTH AUSTRALIA, 12, 14 km WNW
Renmark, 34°07'S, 140°37'E, 7.xi-13.xii.l995, K. R.
Pullen, flight intercept/ pitfall trap - small median
and lateral dark spots on T3, T4 and scutellum
lightly dusted with silver pollinosity.

The female of R. albipilosa has a distinctive
facies, characterised by its relatively small size
(<3.0mm), extensive pale coloration and reduced
thoracic chaetotaxy. Although the male has more
normally developed thoracic chaetotaxy, it
otherwise generally resembles the female. This
combination of characters sets it apart from R.
cucuUata which is larger, robustly setose and
has extensive grey colouration. Other exotic
species of Rondania also more closely resemble
R. cucuUata as discussed below under R. cinerea
sp. nov.

Rondania cinerea sp. nov.

(Figs 1, 3C, 4B, D, E)

Etymology. Latin, cinereus, ash-coloured, grey;
referring to the grey body colour.

Material. Holotype 9: Western Australia, 19 miles (30
km) N by E of Mundrabilla HS, 31°51'S, 127°51'E,
16.X.1968, Britton, Upton, Balderson (ANIC).

PARATYPES: WESTERN AUSTRALIA, 3$, same
data as holotvpe; 12, Mt Boyatup, 74 miles (118
km) E of Esperance, 33°44'S, 123 d 02'E, 23.iii.1968,
I.F.B. Common and M.S. Upton; 1$, 28 miles (45
km) W of Madura, 31°56'S, 126°58'E, 30.iv.1968,
I.F.B. Common and M.S. Upton; 1$, Drummond's
Cove, Geraldton, 28°46'S, 114°16'E, 26. ix. 1972,

N. McFarland, at light; 2$, 5 km ENE Caiguna,
32°15'S, 125°32'E, ll.iv.1983, E.S. Nielsen and E.D.
Edwards; 39, 30 km SE by S Carnarvon, 25°07'S,
113°50'E, 29.iv.1971, Upton and Mitchell; all in
ANIC; 5c?, 20 km N Eneabba, 29°49'S, 115°16'E,
21-23. ix. 2002, G. and M. Wood, in WAM. OTHER
MATERIAL: WESTERN AUSTRALIA, 62, 30 km SE
by S Carnarvon, 25°07'S, 1 13°50'E, 29.iv.1971, Upton
and Mitchell; 2^?, Sandstone, 27°59'S, 119°18'E,
emerged 22.vii.1976 ex adult weevil Acantholophus
niveovittatus collected 30. vi. 1976, A.M. and M.J.
Douglas. NORTHERN TERRITORY, 1?, Tempe
Downs, 24°23'S, 132°25'E, 11.ix.1963, P. Ranford.
SOUTH AUSTRALIA, 11, Blanchetown, 34°21'S,
139°37'E, 26.xii.1966, Z. Liepa. AUSTRALIAN
CAPITAL TERRITORY, 1$, Canberra, 35°17'S,
149°13'E, 17.xi.1929, l.M. Mackerras; 11c?, Canberra,
12.xi.1936, W.J. Rafferty; 59, Canberra (Black Mtn),
collected 7.U.1957, 20.ii.f960, 12.xii.1960, 7.xi.l961 and
3.V.1967, I.F.B. Common, light trap. All in ANIC.

Diagnosis. Small flies; thorax and abdomen
grey (latter pale laterally in <?); head and legs
yellowish in 9 , dark in pleura dark-haired;
setae on prescutum and scutum of 9 normally
developed (not reduced as in R. albipilosa) and
scutellum with 3 pairs of marginal setae; o eyes
narrowly separated, 9 eyes broadly separated.

Description. Female. (Fig. 4B, D) Body length
3.6 (3.0 - 4.5) (n=21); V:HW 0.39 ' (0.37 -

O. 40) (n=21). Head yellow; eyes bare, broadly
separated, pfr not much narrowed dorsally;
if area dull yellowish; pfr, pf and gena with
yellow ground colour and silver pollinosity;
vertex silver pollinose, upper part with dark
ground colour and hairing, lower part yellow
with pale hairing; A.s.l and 2 yellow, A.s.3
darker, especially on outer surface; arista dark,
bare; antenna short, falling short of epistome
bv a distance approximately equal to length of
A.s.3; epistome slightly more pronounced
than profrons in profile; palp yellow; mentum

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

129

Cantrell & Burwell

somewhat darker; oc prominent, proclinate;
4-5 pairs irregular fr; 2 pro-, 1 reclinate and
laterally divergent orb; iv convergent but not
crossed; ov smaller than iv, divergent; vibrissa
relatively strong, row of weaker bristles below
(along genal margin) and 1 above vibrissal
insertions; gena dark-haired.

Thorax grey, with darker rings around setal
bases, black-haired; humeral callus with 2 setae;
pra absent; l(?2)+2 ia; 2+3 dc; 2(3)+3 ac; scutellum
with 3 pairs of marginal setae, strong crossed
apicals, strong basals and weak laterals; legs
yellow, tarsi somewhat darker; fore femur with
pd and pv setal rows; mid tibia with 2 ad and
1 submedian v setae; hind femur with ad setal
row; hind tibia with 2-3 ad, 1 av and 2 pd setae;
wing veins yellow; cell R 5 closed, petiole at
least as long as the length of M from the bend
to its junction with R 4+5 ; node of R 4+5 with 4-5
small black setulae; m-cu joining M closer to
bend than r-m.

Abdomen grey, with darker rings around
setal bases and shifting silver/bronze polli-
nosity; Tl+2 excavate in basal half, with mmp;
T3 and T4 with marginal row of setae; T5 with
ill-defined rows of discal and marginal setae;
pale tubular ovipositor partly protruding beyond
tip of T5 in some specimens.

Male. Body length 3.7 (3.2 - 4.0) (n=15). Similar to
V except as noted: Head: ground colour mainly
dark, with silver pollinosity; antenna pale
yellow, although A.s.3 and arista darker; frons
very constricted, narrowest part approximately
as wide as distance between lateral ocelli;
eye facets enlarged dorsally, with clear line of
demarcation visible in well-preserved specimens;
weak iv and ov, converging but not crossed;
row of fine postocular setulae. Thorax: legs
dark. Abdomen pale laterally on Tl+2-4, pale
area just extending to T5. Terminalia as in Fig. 4E.

Puparium. Oblong, brown, tapering posteriorly
towards non-functional larval posterior spiracles,
these set on diverging shining black cylindrical

bases, narrowly separated medially (Fig. 3C); each
spiracle comprises about 12 small slits, arranged
in a crescent. Anterior larval spiracles missing
from 2 available puparia. Puparial respiratory
horns absent.

Remarks. The following males in WAM are
similar to R. cinerea, but have yellow abdomens
with median dark markings on T3-5 and a
more rounded bend in Ml. The terminalia of
one male were briefly examined and showed
differences in the shape of the cerci and surstyli
compared with R. cinerea, suggesting they are
a distinct species. MATERIAL. WESTERN
AUSTRALIA, 2$, 20 km N Eneabba, 21-23.ix.2002,
G. and M. Wood.

Rondattia cinerea has a facies most closely
resembling R. cucullata, but may be distinguished
by the petiolate cell R 5 (open or just closed at
wing margin in R. cucullata). Rondania dimidiata
(Meigen) and R. dispar (Dufour), however, have
a petiolate cell R5, but the petiole is relatively
short in both species, at most a third the length
of M from the bend to its junction with R 4+s
The petiole in R. cinerea is at least as long as
the length of M from the bend to its junction
with R 4 + cj. Rondania fasciata (Macquart) has a
long petiole like R. cinerea, but is distinguished
by its extensive yellow abdominal markings.
[R. cinerea was not compared against literature
descriptions of R. insularis (Bigot) (Canary
Islands), R. rubens Herting (Spain), or R. dorsalis
(North America).]

BIOLOGY

The Tachinidae exhibit several reproductive
strategies, mainly oviparity or ovolarviparity
where the egg develops in utero to a first-
instar larva contained within the chorion. As
soon as these are laid, the larvae break out
and begin to search for a host. In some cases,
females actually deposit their eggs on the
host. The Dexiinae appear to be exclusively
ovolarviparous (Cantrell 1988). In the course

130

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Australian Dufouriini (Diptera: Tachinidae: Dexiinae)

of this study, we noted the presence of fully
developed first instar larvae in the oviducts of
specimens of Rondania albipilosa and R. cinerea.
Because of their small size, these larvae were
not mounted for closer examination, but they
confirm that the Australian Dufouriini are
ovolarviparous like other Dexiinae.

Published records indicate that Dufouriini
are parasitoids of beetles, primarily leaf beetles
(Chrysomelidae) and weevils (Curculionidae)
(see Baranov 1938; Cox 1994; Guimaraes
1977, Kovaric & Reitz 2005, Parker et al. 1950;
Tschorsnig & Herting 1994). Most records are
of flies emerging from adult beetles. However,
Cerretti & Mei (2001) recorded Eugymnopeza
braueri Townsend parasitising adult Blaps gibba
Laporte (Tenebrionidae) and give detailed
descriptions of the life history of E. braueri
in Italy. The two known host records for the
Australian Dufouriini are discussed below.

Chetoptilia angustifrons is a solitary endo-
parasitoid of tortoise beetles (cassidoid Hispinae:
Chrysomelidae) and has been reared from two
species, Aspidimorpha deusta (Fabricius) and
an undescribed species of Cassida L. Tortoise
beetle larvae are external leaf feeders and
most, including the above species, retain their
cast larval exuviae upon a pair of supra-anal
processes, forming a dorsal shield. The 5 th and
final instar has a dorsal shield comprising the
exuviae of the four previous instars. In addition
to the exuviae, many tortoise beetles, including
both the known hosts of C. angustifrons, incorp-
orate their own faecal material into the dorsal
shield. Prior to pupation, 5 th instar tortoise
beetle larvae usually glue the ventral surfaces
of their anterior abdominal segments to the
substrate and undergo a prepupal stage of one
to a few days.

Mature larvae of C. angustifrons pupariate
within the dead remains of their host (Fig.
3A). The specimens of C. angustifrons from north
Queensland emerged from prepupae of A. deusta

attached to the leaves of its host plant Ipomoea
pes-caprae (L.) (Convolvulaceae), a common
trailing perennial growing above the strandline
of sandy beaches. The male reared from the
Cassida species in southeast Queensland also
emerged from a prepupa, attached to leaf
of the beetle's host plant. Polymeria calycina
R. Br. (Convolvulaceae). The position of the
fly puparium within the host prepupa was
described above. The reverse orientation of the
puparium facilitates the eclosion of the adult
fly as the posterior abdominal segments of the
beetle prepupa are slightly elevated above the
substrate to enable the reflection of the dorsal
shield over the dorsum of the prepupa.

No information is available on the oviposi-
tional behaviour of C. angustifrons, nor on
which life stage of the beetle host is initially
attacked. However, it is interesting to note that
the larvae of both known hosts incorporate
faecal material into their dorsal shields in all
instars. In contrast, the larvae of several species
of Australian Cassida (often placed within the
subgenus Taiwania Spaeth) have dorsal shields
that are largely devoid of faecal material; only
the first instars deposit a small blob of faeces
on the supra-anal processes (Hawkeswood et
al. 1997; CJB unpub. data). Two of these species,
Cassida diomma Boisduval and C. compuncta
(Boheman) are common in south-eastern Queens-
land, and despite CJB having reared a large
number of field collected larvae, prepupae
and pupae of both species, C. angustifrons has
not been found attacking either (CJB unpub.
data). Perhaps adult females of C. angustifrons
use chemical odours emanating from the faeces
incorporated into the dorsal shields as host
finding cues.

Most extralimital species of Chetoptilia for which
host records are known also attack tortoise beetles
(cassidoid Hispinae). However, the life stage
from which the adult fly emerges varies,
with species recorded from larvae, pupae and
adult beetles: C. plumicomis from an unidentified

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

131

Cantrell & Burwell

cassidoid larva in Uganda (Villeneuve 1942); C.
cyanea from a pupa of Aspidimorpha apicalis (Klug)
in Madagascar (Mesnil 1968) and C. burmanica
from overwintering adults of Craspedcmta leayana
(Latrielle) in Myanmar (Garthwaite 1939). The
only known exception to this pattern of cassidoid
hosts is found in the type species of the genus, C.
puella, a parasitoid of adult weevils ( Bytiscus behdae
L., Curculionidae) (Tschorsnig & Herting 1994).

Rondania cinerea attacks adults of the weevil
Acantholophus niveovittatus (Curculionidae;
Amycterinae). Two males in ANIC emerged
from adult A. niveovittatus collected beneath
Spinifex in Western Australia by Athol Douglas
(see Zimmerman 1993: 214). However, no indi-
cation is given as to whether the specimens
emerged from the same or separate adult
weevils. The fly puparia are devoid of host
remains and it is assumed that the mature fly
larvae leave the host to pupariate in a similar
manner to R. cucidlata (see below). Acantholophus
is a widespread and diverse genus known from
all Australian states and territories and with more
than 50 described species (Zimmerman 1993).

Extralimital species of Rondania for which
host records are known, are also parasitoids
of adult Curculionidae (Tschorsnig & Herting
1994). Female Rondania use their reflexed tub-
ular ovipositor to lay eggs directly onto the
adult host. Rondania dimidiata lays its eggs in the
mouthparts of adults of Brachyderes incanus L.
(De Fluiter & Blijdorp 1935). Rondania cucullata
lays its eggs near the anal opening of adults
of Cleonus mendicus Gyll., the mature larvae
leaving the host to pupariate in soil (Menozzi
1939). Eugymnopeza braueri Townsend, a para-
sitoid of the ground beetle Blaps gibba
Laporte (Coleoptera; Tenebrionidae), also lays
its eggs in the mouthparts of adult beetles, but
pupariates within the dead host beetle (Cerretti
& Mei 2001).

ACKNOWLEDGEMENTS

We thank the following collaborators for helpful
discussions on the taxonomy of Dufouriini,
access to literature and loan of non-Australian
specimens for study: Dr David Barraclough
(School of Biological and Conservation Sciences,
University of KwaZulu-Natal, Durban, South
Africa), Dr Pierfilippo Cerretti (Centro Nazionale
Biodiversita Forestale, Verona, Italy), Dr James
O'Hara (CNC) and Dr Monty Wood (CNC).
Geoff Thompson (QM) took the photographs of
adult flies, Karin Koch (QM) prepared the dis-
tribution map and Susan Wright (QM) provided
technical assistance. We also thank Greg Daniels
(UQIC) and Dr David Yeates (ANIC) for loan
of specimens and Dr Barbara Baehr (QM) for
assistance with German translation.

LITERATURE CITED

Baranov, N. 1938. Neue indo-australischeTachinidae.
Bulletin of Entomological Research 29: 405-414.

Barraclough, D.A. 2005. Rhinophoroides minutus, a new
genus and species of rare nocturnal Dufouriini
(Diptera: Tachinidae: Dexiinae) from South
Africa. African Entomology 13: 380-384.

Borowiec, L. 1992. A review of the tribe Aspidimorphini
of the Australian Region and Papuan Subregion
(Coleoptera; Chrysomelidae: Cassidinae). Genus
3: 121-184.

Borowiec, L. & Swiytojahska, J. 2008. Cassidinae of
the World - an interactive manual (Coleoptera:
Chrysomelidae). http://www.biol.uni.wroc.pl/
cassidae/katalog%20intemetowy/ index.htm.
Cantrell, B.K. 1988. The comparative morphology of the
male and female postabdomen of the Australian
Tachinidae (Diptera) with descriptions of
some first-instar larvae and pupae. Invertebrate
Taxonomy 2: 81-221.

Cantrell, B.K. & Crosskey, R.W. 1989. Tachinidae. Pp.
733-784. In, Evenhuis, N.L. (Ed), Catalog of the
Diptera of the Australasian and Oceanian Regions.
Bishop Museum Special Publication 86. (Bishop
Museum Press: Honolulu and E. J. Brill: Leiden).

Cerretti, P. & Mei, M. 2001. Eugymnopeza braueri
(Diptera, Tachinidae) as parasitoid of Blaps gibba
(Coleoptera, Tenebrionidae), with description

132

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Australian Dufouriini (Diptera: Tachinidae: Dexiinae)

of the preimaginal instars. Italian Journal of
Zoology 68: 215-222.

Cerretti, P., De Biase, A. & Freidberg, A. 2009.
Systematic study of the genus Rossimyiops Mesnil
(Diptera: Tachinidae). Zootaxa 1984: 31-56.

Cox, M.L. 1994. The Hymenoptera and Diptera
parasitoids of Chrysomelidae. Pp. 419-468. In,
Jolivet, P.H., Cox, M.L. & Petitpierre, E. (Eds),
Novel aspects of the biology ofOtrysotnelidae. (Klu ver
Academic Publishers).

Crosskey, R.W. 1973. A taxonomic conspectus of the
Tachinidae (Diptera) of Australia, including keys
to the supraspecific taxa and taxonomic and host
catalogues. Bulletin of the British Museum (Natural
History), Entomology, Supplement 21: 1-221.

1976. A taxonomic conspectus of the Tachinidae
(Diptera) of the Oriental Region. Bulletin of the
British Museum (Natural History), Entomology,
Supplement 26: 1-357.

1984. Annotated keys to the genera of Tachinidae
(Diptera) found in tropical and southern Africa.
Annals of the Natal Museum 26: 189-337.

De Fluiter, H.J. & Blijdorp, P.A. 1935. De grauwe
Dennensnuitkever Brachyderes incanus L.
Tijdschrift Plantenziekten 41: 143-211.

Garthwaite, P.F. 1939. Biology of Calopepla leayana
Latr. (Chrysomelidae, Col.) and the possibilities
of control. Indian Forest Records (New Series)
Entomology 5: 237-277.

Guimaraes, J.H. 1977. A review of the tribe
Oestrophasiini Brauer & Bergenstamm (Diptera,
Tachinidae). Papeis Avulsos de Zoologica, Sao
Paulo 30: 215-238.

Hawkeswood, T.J., Takizawa, H. & Jolivet, P.H. 1997.
Observations on the biology and host plants of
the Australian tortoise beetle, Cassida compuncta
(Boheman), with a description of the larva, pupa
and adult (Insecta: Coleoptera: Chrysomelidae).
Mauritiana 16: 333-339.

Herting, B. 1984. Catalogue of Palearctic Tachinidae
(Diptera). Stuttgarter Beitrdge zur Naturkunde,
Serie A (Biologie) 369: 1-228.

Kovarik, P. W. & Reitz, S.R. 2005. Oestrophasia (Cenosoma)
sabroskyi (Diptera: Tachinidae), a parasitoid of
Artipus floridanus (Coleoptera: Curculionidae):
taxonomy and bionomics. Annals of the
Entomological Society of America 98: 245-251.

Medvedev, L.N. 1995. Chrysomelidae (Coleoptera)
from Leyte Island, Philippines. Stuttgarter Beitrdge
zur Naturkunde Serie A (Biologie) 526:1-22.

Menozzi, C. 1939. Parassiti e predatori del Conorrhynchus
mendicus Gyll. (Coleoptera-Curculionidae),
dannoso alia bietola da zucchero in Italia, e
loro importanza nella lotta biologica contro
questo fitofago. Verhandlungen VII International
Kongress of Entomology, vol. IV. Friendlander &
Sohn, Berlin, pp. 2561-2575.

Mesnil, L.P. 1953. Nouveaux tachinaires d'Orient (2 e
partie). Bulletin and Annals of the Entomological
Society of Belgium 89: 146-178.

1968. Quelques remarquables Tachinaires de
Madagascar (Dipt. Tachinidae). Verhandlungen
der Natutjbrschenden Gesellschaft in Basel 79: 44-55.

O'Hara, J.E. & Wood, D.M. 2004. Catalogue of the
Tachinidae (Diptera) of America north of
Mexico. Memoirs on Entomology, International 18.
(Associated Publishers: Gainesville, Florida).
410 pp.

Parker, H.L., Berry, P.A. & Silveria, A. 1950. Vegetable
weevils and their natural enemies in Argentina
and Uruguay. United States Department of
Agriculture Technical Bulletin No. 1016: 1-28.

Robineau-Desvoidy, J.B. 1850. Essai Myiodaires.
Myiodaires des environs de Paris (Suite). Annales
de la Societe Entomologique de France 2: 183- 209.

Rondani, C. 1862. Species italicae ordinis Dipterorum
in genera characteribus definita, ordinatum
collectae, methodo analytica distinctae, et novis
vel minus cognitis descriptis. (Part 4 Muscidae:
Phasiinae-Dexiinae-Muscinae-Stomoxidinae.)
Dipterologiae italicae prodromus 5, 239pp.

Tschorsnig, H.-P. & Herting, B. 1994. The tachinids
(Diptera: Tachinidae) of Central Europe:
identification keys for the species and data on
distribution and ecology. Stuttgarter Beitrdge
zur Naturkunde Serie A (Biology) 506: 1-170.

Villeneuve, J. 1942. Descriptions de Myodaires
sup6rieurs nouveaux (Dipt. Tachinidae. Bulletin
de la Societe entomologique de France 47: 50-55.

Zimmerman, E.C. 1993. Australian weevils. Volume
III. (CSIRO Publications, Melbourne). 854 pp.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

133

Dinosaur footprints from the lower Jurassic
of Mount Morgan, Queensland

Alex G. COOK

Queensland Museum Geosciences, 122 Gerler Rd, Hendra, Qld 4011, Australia. Email: alex.cook@
qm.qld.gov.au

Nupur SAINI

School of Earth and Environmental Sciences, University of Wollongong, NSW 2522, Australia.

Scott A. HOCKNULL

Queensland Museum Geosciences, 122 Gerler Rd, Hendra, Qld 4011, Australia.

Citation: Cook, A.G., Saini, N. & Hocknull, S.A. 2010 03 15. Dinosaur footprints from the Lower
Jurassic of Mount Morgan, Queensland. Memoirs of the Queensland Museum — Nature 55(1):
135-146. Brisbane. ISSN 0079-8835. Accepted: 5 May 2009.

ABSTRACT

Over 190 partial and complete dinosaur prints which include six trackways are preserved
in the ceilings of a disused clay mine near Mount Morgan central eastern Queensland,
Australia. These represent the best record of Early Jurassic dinosaur footprints thus far
discovered within Australia. Anomoepus dominates with other morphologies present
including, Grallator, cf. Eubrontes, and Skartopus and several indeterminate prints. Only
one possible manus print was observed. All preserved tracks are short walking tracks.
□ Lower Jurassic, dinosaur trackways, ornithopod, theropod, dinosaur footprints,
Razorback Beds, Mount Morgan.

Early Jurassic records of dinosaurs in Australia
are scant with no Early Jurassic dinosaur skeletal
material known, and only isolated occurrences
of dinosaur prints known in Queensland
(Bartholomai 1966; Bartholomai in Hill, Playford
& Woods, 1966; Thulborn 1994). Dinosaur prints
and trackways are known in Australia from the
mid-Triassic onwards into the Late Cretaceous
(Thulborn & Wade 1984; Thulborn 1998) and
have proved essential in delimiting the diversity
and evolutionary importance of otherwise
unknown elements of the dinosaurian faunas
of the continent. Dinosaur footprints have been
recognised near Mount Morgan for over fifty
years, but there has not been any documentation
of their occurrence, composition and inferred
palaeobiological information. Reconnaissance
reports of this ichnofauna were made by Staines
(1954), Bartholomai (1966) and Molnar (1991),

but only isolated prints and no trackways
were figured. This work concerns the dinosaur
footprints prints exposed in the 'Fireclay Caverns'
mine at Mount Morgan and the few trackways
preserved in the ichnological assemblage.

Mount Morgan gold and copper deposit was
exploited between 1882 and 1990. In support of
mining operations, were a series of clay mines
supplying raw material with which to make
bricks for the smelting furnaces. These 'fireclay
caverns' operated spasmodically between 1886
and 1925. Activity in the mine and subsequent
chemical weathering exposed the lowermost
surfaces of at least three layers of dinosaur
footprints. A prime reason for the delay in
investigating these occurrences of trackways
was that they are exposed on the ceiling of
the mine, 10-12 m above the mine floor. This
rendered replication, photography or removal

Memoirs of the Queensland Museum | Nature • 2010 • 55(1) • www.qm.qld.gov.au

135

Cook, Saini & Hocknull

difficult. One small area of footprints was
exposed in a suspended drive and was within
4-5 metres of the mine floor. It was these prints
that Staines (1954) photographed.

STRATIGRAPHIC AND
SEDIMENTOLOGICAL SETTING

Mount Morgan is located 38 km SW of the
regional centre of Rockhampton, central eastern
Queensland (Fig. 1). A ~65 m thick succession
non-marine Jurassic sediments unconformably
overlies Devonian basement (Fig. 2) and has been

commonly called the Razorback Beds (Playford
& Cornelius 1967, Day et al., 1983). Studies
by Playford & Cornelius (1967) of these units
indicate a lowermost Jurassic age and correlated
the Razorback Beds with the Evergreen Formation
of the Precipice Sandstone within the Great
Artesian Basin to the west and south west.

Nearly 65 m of Jurassic sediments are exposed
near the Mount Morgan Mine. Detailed logging
of this section has been interpreted to represent
three distinct facies groups; (1) Basal pebbly
sandstone; (2) siltstone and (3) cross-bedded
coarse sandstone.

FIG. 1. Locality and general geology of Mt Morgan with map of Fireclay caverns after Staines (1954). Main
footprint sites marked A-E.

136

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Dinosaur footprints from Queensland

Basal Pebbly sandstone facies. This facies lies
unconformably atop Devonian porphyry and is
approximately 20 m thick. Coarse massive quartz
and sublithic sandstones, minor polymict con-
glomerates medium grained quartz sandstone
units 0.6-2. 2 m thick are intercalated with
sporadic mudstone and siltstone units. Minor
trough cross bedding and planar laminations
are present in the coarser units. Scour bases and
lithic breccia lenses are present in the lower parts
of coarse units. Generally the facies fines upwards
and is directly overlain by the thick siltstone facies
described below. The sequence is interpreted
as an alluvial fan deposit with sporadic channel
and limited floodplain deposition.

Siltstone facies. Two metres of fine-medium silt-
stone is overlain by monotonous fine-grained
laminated siltstone, over 16 m thick which domi-
nates the middle of the sequence. Sedimentary
structures include planar and parallel ripple cross
laminations, isolated outsize quartz pebbles
and subangular clay clasts. Significant pyrite
crystals are distributed throughout which are
considered secondary given the nature of the
proximal orebody. Dinosaur prints are preserved
in the uppermost one metre of this facies. The unit
is interpreted as a lacustrine system which in
uppermost parts shallows to form a regressive
lacustrine system onto which footprints were
impressed.

10m

Coarse quartzose planar
cross-bedded sandstones.
Minor conglomeratic lenses,
minor sandy siltstones.
Alluvial fan setting.

Laminated siltstone and fine
sandy siltstones.

Lacustrine setting.

Basal pebbly sandstones and
conglomerates, sandy siltstones
and coarse sandstones.

Riverine and alluvial fan setting.

Cross bedded sandstone facies. In the study area
this forms the uppermost 25 m of exposed
section and is dominated by massive and planar
and trough cross bedded coarse quartz and
sublithic sandstones. Bedding sets are up to 2 m
thick and have common basal pebble lags. The
facies is interpreted as part of a small braided
river system which built out over a restricted
lacustrine environment.

PRESERVATION

Footprints are preserved as hyporelief and
transmitted hyporelief prints throughout most

FIG. 2. Stratigraphic column for Razorback Beds in
the vicinity or the Fireclay Caverns, Mount Morgan.

of the clay caverns mine (Fig. 1). They are
preserved in several layers of the uppermost
1 m of siltstone facies within the sequence.
Their preservation mostly >10 m above the
mine floor was a major impediment to earlier
study. In addition the small area in which
the prints are less than 5 m above the mine
floor could not be cast as the prints were too
fragile. This was exacerbated by pyrite growth
throughout the siltstone producing large areas
of friable and extremely fragile rock. Finally in

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

137

Cook, Saini & Hocknull

2007, using major scaffolding, a small section of
footprints was moulded. The mould is lodged
with the Queensland Museum (QMF54079).

METHODS

Initial work which formed part of reconnai-
ssance studies (Cook et a]., 2002) simply photo-
graphed the major footprint-bearing panels.
These were then assembled as a photo mosaic
for analysis. Two laser devices 10 cm apart were
used to provide a scale for further work with
photo mosaics. Individual images were either
solarised or embossed in Adobe ™Photoshop
7.0. Rectified images were imported into
ArcMAP™ 9.0 for further measurement.
Because of the large distance between the floor of

the caverns and the ceiling, approximately 10
m, all measurements are approximate.

Calculations of hip height follow Alexander
(1976) and Thulborn (1990).

DISTRIBUTION

The prints are present in many parts of the
mine ceiling. Areas with more concentrated prints
sets were given informal names for convenience.
Trackways are preserved in 'Bat Cave', 'Main
Entrance', 'Mezzanine Hall', 'North Bat Cave'
and 'Ladder hall' (Fig. 3).

No preferred overall orientation of prints
was observed within the caverns (Fig. 3). Two
areas with slight preference for trends of prints
directed to the WSW and WNW for 'Bat Cave'
and the 'Main Entrance' where there was an
EW-NW bipolar orientation.

PRINTS

Only one possible manus print was observed
in the caverns despite the observations of
Bartholomai (1966) reported in Molnar (1991).
Most prints were tridactyl pes prints. Largest
prints were that of cf. Eubrontes had Foot Length
(FL)=40 cm and the smallest, that of Attomoepus
FL=5 cm. Seven print morphologies were
identified, but only three can be confidently
assigned to ichnotaxa. Many prints are eroded,
with many of their features distorted by
erosion. An equally large number of prints are
transmitted prints or partial transmitted prints
with digit 'breakthrough' from overlying layers.
All prints are preserved in hyporelief.

Attomoepus (Fig. 4)

This is the dominant print present within
the caverns represented by 69 full prints and
61 partial prints which can be confidently
assigned. FL varies from 5 cm to 27 cm with most
prints ranging in FL from 10-15 cm. Divaricance

138

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Dinosaur footprints from Queensland

FIG. 4. A-C, Anomoepus from the Fireclay Caverns, Mount Morgan; D, Isolated partial pes and possible
manus print, Anomoepus.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

139

Cook, Saini & Hocknull

FIG. 5. A, Large transmitted theropod print; B, Stubby-
toed omithopod print Note tracks are transmitted with
breakthroughs on deeper distal digits, especially III.

is between 85-105°. Digits are relatively broad
and a semiclaw is present in some specimens.
Digits are long and tapered with digit (D) III
longer and more tapered than DII or DIV. Slight
medial swelling is present on Dill, however there
is a pronounced swelling distally on DIV. Distal
ends tapered more sharply on Dill, but are
subrounded on the other two digits. Interdigital
angle is greater on DII-III than DIII-IV.

One print is associated with an additional
manus impression (Fig. 4D), consisting of three
short stubby digits, approximately 7 cm long,
with a divaricance of approximately 85°. These
lie just anterior to a distinct partial Anomoepus
pes print.

Four short trackways are attributable to
Anomoepus. Anomoepus has been recorded in
the Precipice Sandstone of Queensland by
Thulborn (1994), but the specimens lodged
in the Queensland Museum by Thulborn are
generally smaller and have slightly narrower
digits. Moyenisauropus has been used by many
authors for prints similar in morphology to
Anomoepus, but larger than 20 cm FL. A number
of authors however assert that the forms are
indistinguishable (Lockley & Meyer 2000;
Olsen & Rainforth 2003; Olsen & Gatton 1984;
Thulborn 1994). For the purposes of this work
we regard them as size variants.

cf. Eubrontes (Fig. 6)

Fourteen individual prints and one trackway
are assigned to this ichnotaxon. FL is between
16-40 cm. The prints assigned to cf. Eubrontes
differ from Grallator in both size and that Dili
is reduced in comparison to DII and DIV. In the
Mount Morgan prints, the DII-IV divarification
is between 50° and 70°. Claw marks are present
in three of the 14 prints observed. These represent
the largest of the prints at Mount Morgan.

Skartopus (Fig. 8C)

This is represented by a single small tridactyl
print with FL of 5 cm, interdigital angles <40°.
Digit III slightly longer than others. The single
print is distinctive and unlike any other in the
Fireclay Caverns. Skartopus is known from the
medial Cretaceous of western Queensland
(Thulborn & Wade 1984).

140

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Dinosaur footprints from Queensland

FIG. 6. cf. Eubrontes prints forming a short trackway near the entrance to Fireclay Caverns

Grallator (Fig. 5)

Tridactyl print, digit III is longer than DII
and DIV, which are nearly equal in length.
Divarification is between 85-95°. No hallux
impression is present. Most prints are between
4-18 cm in FL but a single, much larger, print
28 cm long is present. This morphotype is
represented by 16 complete and 8 partial prints.
One short trackway is attributable to Grallator.
The divarification is high in comparison to types
of Grallator (Olsen, Smith & McDonald 1998)
and other described footprints, but the general
foot morphology is consistent. Lack of detail in
the phalangial pad marks makes discrimination
of individual phalangial swellings difficult.

The earliest record in Australia of grallatoroid
tracks is that of Thulborn (1998) who described

five prints from the mid-Triassic (Carnian)
of Queensland. Grallator and its synonyms
(Thulborn 1998) are common prints within Early
Jurassic assemblages worldwide (Ellenberger
1974; Olsen & Galton 1984; Rainforth 2001).

INDETERMINATE PRINT 1 (Fig 7A-C)

Large tridactyl prints, up to 24 cm FL, with a
divaricance of 70-100°. Digits are long and taper
to a point. Rear of the print is crescent-shaped.
DII-III angle is less than DIII-IV. Dill deep has
a medial swelling. Two examples of this print
have a probable hallux impression which extends
approximately one quarter the distance behind
the proximal part of the print. The sharp distal
ends of the digit and the size suggest that these
prints are attributable to a theropod.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

141

Cook, Saini & Hocknull

FIG. 7. Large indeterminate, weathered theropod prints.

142

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Dinosaur footprints from Queensland

INDETERMINATE PRINT 2 (FIG. 8A)

FL approximately 8 cm, consisting of three
widely divaricant digits with an overall
divaricance of 95°. Digits are short and blunt
v/ith rounded distal ends. Dill is slightly longer
than D1V and has a slight swelling approaching
the distal margin. DIV deep. DII shallower with
less pronounced swelling at the distal end.

This print differs from Indeterminate print 3
by its wider divaricance and the connectedness
of the digits. The blunt nature of the prints suggest
an ornithopod origin.

INDETERMINATE PRINT 3 (Fig. 8B)

Foot length is approximately 12 cm. Print
consists of three blunt digits with a divaricance
of 50°. Dill longer and deeper than DII and DIV,
with DIV longer than DII. Slight swelling at the
proximal end of Dill. The three digits are almost
unconnected on the print. The blunt toed nature
of this print suggests an ornithopod origin

SIZES OF PRINTMAKERS

Thulbom (1990) suggested that the approximate
relationship of FL to hip height and his
formulae and conversion factors are followed
here. Anoinoepus prints suggest most animals
with a hip height of 25-131 cm. One example
suggested an animal with a hip height of 161 cm,
but this single print is aberrant. Prints ascribed to
cf. Eubrontes suggest animals in the range 70-
180 cm at the hip and those assigned to Grallator
indicate hip heights of between 20-120 cm.

ASSEMBLAGE

The variety of the prints within the assemblage
is skewed by our decision to lump ichnotaxa.
Nevertheless what is striking about this fauna
is the high percentage of anomoepid prints
in relation to theropod prints (Fig. 9). Prints
attributable to ornithopods constitute just
over two-thirds of the assemblage. This is in
contrast to other Triassic-Jurassic assemblages
worldwide which show greater percentages of

FIG. 8. A,B. small, blunt-toed probable ornithipod
prints. A, digits show wide divaricance; B, Digit
impressions only with moderate divaricance; C, Isolated
Skartopus print"

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

143

Cook, Saini & Hocknull

mAnomeopus b Grallator a cf Eubrontes

h indet theropod □ indet omitho pod

FIG. 9. Pie chart of proportions of prints found
within the Fireclay Caverns. n=192. Single Skartopus
print included in Indeterminate theropod prints.

grallatoroid and other theropod prints (Lockley
& Hunt, 1995; Lockley & Meyer, 2000, Gierlinski
& Sawicki, 1998; Gierlinski & Niedzwiedzki,
2002; Niedzwiedzki & Pienkowski, 2004). A
possibility is that the ecological setting for the
site, essentially at lakes edge could account for
the greater numbers of ornithopods, utilising
the lake as a regular watering hole.

TRACKWAYS

Six trackways (Fig. 10) are present in the clay
caverns, the longest of which is represented
by 6 successive prints. Three are attributed to
Anomeopus and of the remaining one to ?Eubrontes
and two to Grallator. Pace lengths and other
data are summarised in Table 1. Stride length
to hip height ratios (Alexander 1976, Thulborn
1990) are all less than 1.3 indicating cursorial
gaits for all the short trackways present. Data is
too incomplete to make any other meaningful
comment on speeds, however it is clear all the
trackways represent walking tracks.

144

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Dinosaur footprints from Queensland

TABLE 1. Morphometric data for trackways within the Mt Morgan Fireclay Caverns.

[ Track and
Print field
number

Full/

partial

Length

(cm)

Width

(cm)

Right/ left

Pace

Length

Stride

length

Pace

angulation

Direction/

azimuth

taxon

urn

Grallator r

LHFP26

7.1

198

LHFP27

4.4+

4.5+

28.3

163

194

\ LHFP28

9.9

5.7+

28.5

56.4

211

LHT2

Anomeopus

LHFP33

9.5

10.4

43.4

171

LHFP34

8.8

9.3

24.6

168

170

LHFP35

5.6+

8.2

12.6

37.7

173

165

LHFP36

10.4

11.4

165

—

MET1

Anomeopus

MEFFP15

19.2

20.9

322

MEFPP14

15.5+

14.3+

31.1

321

MEFPP17

16.9

20.2

337

MET2

Grallator

MEBFB30

17.5

15.7

102

191

MEBFB19

18.5

18.4

98.5

145

182

MEBFB18

18.7

21.7

138

209

MEBFB11

16.4

19.6

213

BCT1

?Eubrontcs

BCFP5

21.9

22.6

26.9

262

BCFP4

6.2+

7.3+

35.7

173

263

BCFP3

20.6

20.7

24.8

180

276

BCFP2

5.6+

285

EHT1

Anomeopus 1

EHFP4

101

227

EHFF5

90.7

135

229

EHFP6

9.0+

2.2+

143

227

EHFP7

142

230

EHFP8

127

226

EHFP9

11.4+

1.9

___

228

ACKNOWLEDGEMENTS

We thank the staff of the Rockhampton office of
the Department of Mines and Energy (formerly
Natural Resources and Water) for access to the
site, and Tourism Queensland for invaluable
funding assistance. The Australian Age of
Dinosaurs is thanked for assistance in moulding
the prints. Kristen Spring and Matthew Ng are

thanked for field assistance. Alan Bartholomai
and Matt White are thanked for helpful reviews
of this manuscript

LITERATURE CITED

Alexander, R.McN. 1976. Estimates of the speeds of
dinosaurs. Nature 261: 129.

Bartholomai, A. 1966. Fossil footprints in Queensland.
Australian Natural History 15: 147-150.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

145

Cook, Saini & Hocknull

Cook, A.G., Hocknull, S.A.& Spring. K.A. 2002. Report
on the Mount Morgan Clay Caverns dino-
saur footprint and trackways site. Queensland
Museum, Brisbane. (Unpublished).

Day, R.W., Whitaker, W.G., Murray, C.G. Wilson,
I.H. & Grimes, K.G. 1983. Queensland Geology-
a companion guide to the 1:2 500 000 scale
geological map (1975). Geological Survey of
Queensland Publication 383.

Ellenberger, P. Contribution a la classification
des pistes de vertebres du Trias; les types du
Stormberg d'Afrique du Sud (I). Palaeovertebrata
Memoire Extraordinaire. Montpellier. Pp 1-134.

Gierlinski, G. 1991. New dinosaur ichnotaxa from the
Early Jurassic of the Holy Cross Mountains.
Palaeogeograyhy, Palaeocliniatology, Palaeoecology
85: 137-148.

Gierlinski, G & Niedzwiedzki, G. 2002. Enigmatic
dinosaur footprints from the Lower Jurassic of
Poland. Geological Quarterly 46(4):467-472.

Gierlinski, G. & Sawacki, G. 1998. New sauropod tracks
from the Lower Jurassic of Poland. Geological
Quarterly 42(4):477-480.

Hill, D„ Plavford, G„ & Woods, J.T. 1966.
Jurassic fossils of Queensland. (Queensland
Palaeontographical Society, Brisbane) 32pp.

Lockley, M. & Hunt, A.P. 1995. Dinosaur tracks
and other fossil footprints of the western United
States. (Columbia University Press, New York).
338 pp.

Lockley, M. & Meyer, C. 2000. Dinosaur tracks
and other fossil footprints of Europe. (Columbia
University Press, New York). 323pp.

Molnar, R.E. 1991. Fossil reptiles in Australia. Pp
605-702 in Rich, P.V., Baird, R.F., Thompson,
E. &. Monaghan, J. eds. Fossil Vertebrates of
Australasia. Pioneer Design Studio, Monash
University, Melbourne.

Niedzwiecki, G. & Pienkowski, G. 2004. A
dinosaur track association from the Early
Jurassic deltaic deposits of Podole near Opatow,
Poland. Geological Quarterly 48(4): 333-338.

Olsen, P.E. & Galton, P.M. 1984. A review of the
reptile and amphibian assemblages from the
Stormberg of southern Africa, with special
emphasis on the footprints and the age of the
Stormberg. Palaeontologia Africana. Haughton
Memorial Volume 92-1 16.

Olsen, P.E. & Rainforth, E.C. 2003. The early Jurassic
omithischian dinosaurian ichnogenus Anomoepus.
Pp 314-368 In LeTomeau, P.M. & Olsen, P.E. (eds).
The Great Rift Valleys of Pangea in Eastern North
America, Sedimentology and Paleontology Volume 2.
Columbia University Press.

Olsen, P.E., Smith. J.B., & Mcdonald, N.G. 1998.
Type material of the type species of the
classic theropod footprint genera Eubrontes,
Anchisauripus, and Grallator (Early Jurassic,
Hartford and Deerfield Basins, Connecticut and
Massachusetts, USA). Journal of Vertebrate
Paleontology 18(3): 586-601.

Playford, G & Cornelius, K.D. 1967. Palvnological
and lithostratigraphic features of the Razorback
Beds, Mount Morgan district, Queensland.
Papers of the Department of Geology,
University of Queensland 6: 81-96.

Rainforth, E.C. 2001. Late Triassic - Early Jurassic
dinosaur ichnofaunas, eastern North America
and southern Africa. Journal of Vertebrate
Paleontology 21 (supplement): 91 A.

Staines, H.R.E. 1954. Dinosaur footprints at Mount
Morgan. Queensland Government Mining Journal 55-
483-485.

Thulborn, R.A. 1990. Dinosaur Tracks. Chapman &
Hall, London. 410pp.

1994. Omithopod dinosaur tracks from the Lower
Jurassic of Queensland. Alcheringa 18: 247-258.

1998. Australia's earliest theropods: footprint
evidence in the Ipswich Coal Measures (Upper
Triassic ) of Queensland. Gaia 15: 301-311.

Thulborn, R. A. & Wade, M. 1984. Dinosaur trackways
in the Winton Formation (mid-Cretaceous) of
Queensland. Memoirs of the Queensland Museum
21: 413-518

146

Memoirs of the Queensland Museum

Nature • 2010 • 55(1)

Fossil fruit of the Macadamieae (Proteaceae) in
the Tertiary of eastern Australia: Eureka gen. nov.

Mary E. DETTMANN
H. Trevor CLIFFORD

Queensland Museum, Geosciences, Hendra Facility, 122 Gerler Rd, Hendra, Qld 4011, Australia.
Email: mary.dettmann@qm.qld.gov.au

Citation: Dettmann, M.E. & Clifford, H.T. 2010 03 15. Fossil fruit of the Macadamieae (Proteaceae)
in the Tertiary of eastern Australia: Eureka gen. nov.. Memoirs of the Queensland Museum —
Nature 55(1): 147-166. Brisbane. ISSN 0079-8835. Accepted: 13 October 2009.

ABSTRACT

Eureka gen. nov. is proposed to accommodate fossil fruits recovered from several mid-
Tertiary (early Oligocene-Miocene) sites in eastern Australia. The type (E. welcomensis
sp. nov.) and other described species (E.(a/. Conchotheca) turgida (F. Muell.) comb,
nov., E. spechtii sp. nov., E. harslettiae sp. nov.) have bilaterally symmetrical, ellipsoidal
pericarps that are uniloculate, 1-seeded, and tardily dehiscent along the ventral suture
and dorsal hinge line. The pericarp wall is radially structured and has a branched
vasculature system capped by fibres embedded in a thick mesocarp; the radial fibres
extend into the exocarp, and the endocarp is thin or lacking. The seed is orthotropous,
attached apically and with a 2-layered seed coat that has an endotestal cuticle bearing
impressions of transfer cells. Eureka is morphologically and anatomically consistent with
fruits of extant members of Bleasdalea F. Muell., Hicksbeachia F. Muell., and Gevuina
Molina of subtribe Gevuininae (tribe Macadamieae, family Proteaceae). Accordingly, the
fossil fruit evidence implies that the Gevuininae had a former distribution range spanning
some 20° of latitude in eastern Australia and taken together with fossil cuticle evidence
a trans-Tasman distribution no longer mirrored in the present vegetation. Eureka also
displays some resemblance to the tardily dehiscent fruits of extant Panopsis (subtribe
Macadamiinae, tribe Macadamieae); however, morphological/anatomical details of
extant members are imprecisely known. □ Australia, Bleasdalea, Eureka, fossil fruit,
Gevuina, Hicksbeachia, Macadamieae, Tertiary.

Among recently recovered woody fossils
from Tertiary sediments at several localities
in Queensland are several different types of
1-seeded, tardily dehiscent, uniloculate fruits
with vascularised pericarps similar to those of
fruits borne by several extant genera included
within tribe Macadamieae (Proteaceae) as
recently redefined (Weston & Barker 2006; Mast
et al. 2008). The Queensland fossil fruits are
spheroidal to ellipsoidal, near-symmetrical about
the vertical axis, and their thick pericarp has a
distinctive radial structure reflecting a branched
vasculature system embedded in a thick, woody

mesocarp. Similarly structured pericarps charac-
terise fruits borne by extant genera of tribe
Macadamieae. The distinctive vasculature
pattern with 'third-order radial vascular bundles'
(Johnson & Briggs 1975, p.134; Douglas 1995,
p.18) comprises prolific radial branching of
the main vertical and lateral (tangentially-
aligned) bundles in a woody ground mass of
predominantly radially oriented fibre sheaths
that surround the bundles as detailed for fruits
of several extant members of the tribe (Filla
1926; Strohschen 1986a,b). The Queensland
fossil fruits with their near vertical symmetry,

Memoirs of the Queensland Museum | Nature • 2010 • 55(1) • www.qm.qld.gov.au

147

Dettmann & Clifford

near-smooth outer surface of pericarp, and
one apically attached unwinged seed having
a 2-layered, non-woody seed coat are more
similar to fruits of several genera ( Gevuina
Molina, Hicksbeachia F. Muell., and Bleasdalea
F. Muell.) of subtribe Gevuininae than to other
members of tribe Macadamieae. Previously
described fossil fruits that possess a thick
pericarp with third order vasculature were
assigned to Conchotheca turgida F. Muell., a taxon
reported from mid-Tertiary deep lead sediments
in Victoria (Mueller 1874a, Deane 1925) and
Tasmania (Johnston 1880). Mueller (1874a) did
not comment on the botanical affinity of C. turgida,
but examination of his protocol material confirms
these fossils, too, are morphologically consonant
with fruits of subtribe Gevuininae, being similar
to those of Gevuina, Bleasdalea, and Hicksbeachia.
Evaluation of Mueller's protocol material and of
the Queensland fossil fruits form the basis of the
present study. Moreover, comparisons have been
undertaken of the fossil material with fruits of
extant Gevuina, Hicksbeachia, and Bleasdalea.

Extant Gevuina is monotypic and endemic
to southern South America; Bleasdalea has two
species, one each in New Guinea and eastern
Australia, and Hicksbeachia, with two species, is
endemic to eastern Australia (Fig 1 A; Weston &
Barker 2006). Fossil cuticles that are consistent
with those of subtribe Gevuininae have been
reported from Eocene sediments of Western
Australia (Carpenter & Pole 1995) and from
the Miocene of New Zealand (Pole 1998).
Fossil wood similar to, but not identical with,
Gevuina has been reported from the Oligocene
of Patagonia (Pujana 2007).

The fossil pollen taxa, Propylipollis reticulo-
scabratus (Harris) Harris and Martin, and P.
tripartus Harris, which occur in Campanian-
Tertiary sediments of Australia, New Zealand
and Antarctica, are morphologically similar to
pollen of Gevuina, Bleasdalea, and Hicksbeachia
(Martin 1982; Dettmann & Jarzen 1990, 1996,
1998). P. reticuloscabratus however, has been

shown to be insufficiently discriminatory of
the extant taxa in parsimony analyses (Sauquet
et al. 2009); thus far, P. tripartus has not been
tested in any comparable cladistic analysis.

MATERIAL

Fruits reported upon here include woody and
charcoalified and permineralised specimens.
The charcoalified specimens were recovered
from subsurface sediments near Blackwater
and Moranbah, central Queensland and
Bundaberg, southeastern Queensland and the
permineralised specimens are from an outcrop at
Glencoe, central Queensland; all are held in the
Queensland Museum (QMF). Also examined are
woody and charcoalified fruits collected in the late
1800' s and early 1900's from several goldfield sites
in Victoria and New South Wales and housed
in the Queensland Museum, the Australian
Museum (AMF) and the Museum of Victoria
(NMVP). The last-mentioned collection contains
the type and other specimens of the protocol
material of Conchotheca turgida F. Muell.,1874,
which was originally housed in the Museum
of the Geological Survey of Victoria (GSVF).
Locality details of occurrences of Eureka gen.
nov. are as follows (see also Fig. IB).

Queensland, a) Picardy Station, near Moranbah
(21°5'17.6'S 147°50'34.3'E) Rio Tinto Exploration
Hole RDPD98MA21, sands, silts and lignites
at 123-133 m; Early Oligocene (Dettmann &
Clifford 2001).

b) Near Blackwater (24°1'1.3'S 148°48'50'E) South

Blackwater Coal Pty Ltd Hole R8736, sands
and lignites at 82m; Early-Late Oligocene
(Dettmann & Clifford 2001).

c) Glencoe Station (23° 36 SI 48° 06' E), near Capella,

silcrete outcrop; 01igocene-?Early Miocene
(Rozefelds 1990; Rozefelds & Christophel 1996;
Dettmann & Clifford 2001).

d) N of Bundaberg (24° 46'49'S 152°18'17'E),
Herberts on Welcome Creek Drain, subsurface

148

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Fossil Proteaceous fruits

sands and clays; Early Miocene (Dettmann
& Clifford, 2003).

e) Near Clifton (27° 55'S 151°55'E), Queensland
Water Resources borehole, sands and lignites
at 45-46 m; Tertiary (?Miocene).

frlexv South Wales. Near Orange (33°17'S 149°
06'E, Forest Reefs Mine, deep lead sediments;
late Middle-early Late Miocene (Johnson 1989;
Dettmann & Clifford 2001)

Victoria, a) Nintingabool (Crucible Co. Shaft),
near Haddon (37° 18'S 146°32'E), SW of
Ballarat, deep lead sediments; Oligocene-
Miocene (Greenwood et al. 2000; Dettmann &
Clifford 2001; Holdgate et al. 2006).

b) Foster (37°10'S 146°14'E), deep lead sediments;
Oligocene-Miocene (Rozefelds & Christophel
1996; Holdgate et al. 2008).

Tasmania. Brandy Creek, Beaconsfield (41°12'S
148°49'E), deep lead sediments; Oligocene
(Forsyth 1989)

METHODS

Fruits studied include carbonaceous specimens
retaining well preserved morphological and
anatomical characters of their woody tissues to
severely vitrinised specimens retaining external
and internal morphology but with few anatomical
characters preserved. Some of the carbonaceous
fruits are pyritised and, as a result of oxidation
since collection, several have fragmented. Per-
mineralised specimens are composed of
amorphous and/or cryptocrystalline quartz
that has preserved external and internal
morphology of the fruits, but no anatomical
detail of the original tissues is preserved.

Charcoalified specimens were photographed
either after whitening with ammonium chloride
to highlight surface features or without whitening
to illustrate internal characters including those
of the locules, seeds, and internal anatomy.
Permineralised specimens were photographed

without whitening. Thin sections of charcoalified
specimens were cut using a slide microtome
and mounted in glycerine jelly on glass slides
for transmitted light microscope analysis.
Fragments of several specimens were mounted
on stubs and gold plated for scanning electron
microscope analysis. Anatomical features of
seed coats were examined in transmitted light
after clearing in a weak solution of sodium
hypochlorite, followed by thorough washing
in distilled water.

Photographs were captured on either T Max
film or digitally using an Olympus Cameida
C-5050 camera. Images were processed using
Photoshop 5LE.

SYSTEMATIC DESCRIPTION
FAMILY - PROTEACEAE JUSS.

SUBFAMILY - GREVILLEOIDEAE ENGL.
TRIBE - MACADAMIEAE C. VENKATA RAO

SUBTRIBE - GEVUININAE L.A.S.

JOHNSON & B.G. BRIGGS

Genus -Eureka gen.nov.

Derivation of name. With reference to the
association of 'Eureka' (Eureka Stockade, Eureka
Flag, Eureka Deep Lead) with early gold mining
activities in and around Ballarat, Victoria.

Generic diagnosis. Fruit unilocular, indehiscent
to tardily dehiscent woody follicle; ellipsoidal
or spheroidal, symmetrical about the vertical
axis. Style base near-opposite stalk in the
vertical plane; ventral suture and dorsal hinge
line approximately equal in length; splitting
along ventral suture and dorsal hinge into 2
near equal portions. Surface smooth or near-
smooth. Pericarp largely composed of meso-
carp, the outer thicker part with a branched
vascular system composed of vertically aligned

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

149

Dettmann & Clifford

FIG. 1. Polar stereographic projections, based on Lawver & Gahagan (2003). A, Present Day showing
distribution-range of extant members of Subtribe Gevuininae (excluding Cardivellia); B, Early Miocene (20
Mya) showing localities of fossil Eureka (*), fossil cuticles (C) of Subtribe Gevuininae, and fossil wood (W)
questionably allied to Subtribe Gevuininae.

(tangential) primary bundles from which branch
radially aligned secondary bundles that are
surrounded by tangentially and radially
aligned fibre bundles; fibres interspersed with
parenchyma. Exocarp and endocarp thin. Seed
solitary, unwinged, orthotropous, attached
apically, filling, or almost so, locule. Seed
coat 2-layered, the outer layer incompletely
enveloping the inner; outer layer a cuticle
with impressions of transfer cells; inner layer a
cuticle with anticlinal walls of rectilinear cells.

Type species. Eureka welcomensis sp. nov.

Remarks and comparison. Eureka gen. nov.
accommodates uniloculate fruits having a
branched vascular system surrounded by fibre
bundles in the mesocarp, and a single apically
attached, orthotropous seed. In these respects
Eureka differs from Conchotheca F. Muell., emend.
Dettmann & Clifford 2005, which includes fruits

having vertically aligned, unbranched vascular
bundles in the mesocarp and laterally attached
seeds. One taxon, C. turgida, allocated to
Conchotheca by Mueller (1874a) is characterised
by a pericarp with a branched vascular system
and an apically attached, orthotropous seed;
this species is transferred herein to Eureka.

Eureka is distinct from Conchocaryon F. Muell.
1879, which has asymmetric fruits and laterally
attached winged seeds (Mueller 1879; Dettmann
& Clifford 2005). Both Plesiocapparis F. Muell.1871
and Celyphina F. Muell. 1871 have indehiscent,
unilocular, near-smooth fruits with branched
vasculature in the mesocarp. They differ from
Eureka in possessing large stone cell complexes
(not radially oriented fibres) that cap and
surround the vascular bundles in the fruit wall
(Dettmann & Clifford, in prep.).

150

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Fossil Proteaceous fruits

FIGS 2-16. Eureka welcomensis so. nov.; 2-4, Lateral, apical and basal views of holotype with seed (arrowed),
QMF51143; scale bar = 5 mm; 5-6, Pericarp, lateral view showing surface of locule and external lateral view
showing V-shaped scar (arrowed) of style base, AMF11099; scale bar = 5 mm; 7, 8, Pericarp, surfaces
of locule, lateral view, QMF51144; scale bar =5mm; 9, Detail of style scar (arrowed) on external surface
of fruit, QMF51145; scale bar = 2.5 mm; 10, Pericarp wall, transverse section at right angles to the dorso-
ventral plane showing vasculature associated with style base (arrow), QMF51145; scale bar = 1 mm; 11,
12, Pericarp wall, section in plane of dorsal hinge line and ventral suture showing vasculature comprising
eroded vertical (v) bundles is cross section and radially aligned secondary bundles (r) surrounded by fibre
complexes embedded in a ground mass of parenchyma, QMF51145; scale'bar = 1 mm; 13, 14, Pericarp wall,
structure in transverse sections showing wall fibre bundle complexes surrounding vasculature, QMF51149;
scale bar = 100 pm; 15, 16, Pericarp wall in transverse section showing fibre bundles surrounding branched
vasculature and detail of fibres, QMF51150; scale bar = 100 gm and 10 gm respectively.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1) 151|

Dettmann & Clifford

Species - Eureka welcomensis sp. nov.

(Figs 2-24)

Holotype. QMF51143 (Figs 2-4). Prolate ellipsoidal,
unilocular, 2-valved fruit, partially dehisced into
2 subequal valves along ventral suture and dorsal
hinge line; stalk scar indistinct, scar of style base
oriented transverse to dorso-ventral plane extending 4-5
mm from apex towards base on mid face of each valve.
Pericarp 4-5 mm thick in lateral regions, thinner (3 mm)
at apex. Seed solitary, apically attached, orthotropous.
Vertical axis 19mm; lateral axes 19 mm, 16 mm.

Other material. QMF51144-QMF51151 inclusive,
AMF11099.

Type locality. Herberts on Welcome Creek Drain, N
of Bundaberg, Queensland; Early Miocene.

Diagnosis. Fruit woody, prolate ellipsoidal,
near symmetrical about the vertical axis;
indehiscent or tardily dehiscent, unilocular,
with 1 apically attached seed. Stalk attachment
basal, inconspicuous; style base represented
by a linear groove oriented transverse to the
dorso-ventral plane and extending 4-5 mm from
apex on lateral surface of each valve; groove
underlain by vascular strands that extend into
the pericarp and terminate near the seed cavity.
Pericarp with near smooth outer surface,
thickest (4-6 mm) in basal and mid regions of
lateral surfaces, tapering to 3-4 mm at apex and
along both ventral suture and dorsal hinge line;
wall comprises thin exocarp overlying a thick
vascularised mesocarp and a thin cuticle-like
endoearp. Inner mesocarp predominantly of
vertically aligned vasculature surrounded by
fibre sheaths in a ground mass of parenchyma;
at or near inner/outer interface of mesocarp
the vascular bundles branch radially; outer
mesocarp composed of the radially directed
vascular branches and surrounding fibre sheaths.
Preserved seed coat comprises exotesta of a thin,
structureless cuticle, an endotestal cuticle bearing
impressions of polygonal-shaped transfer cells
(40-60 pm diameter) having internally directed
fimbriae, and a tegmic cuticle with impressions
of anticlinal walls of rectilinear cells (55-70 pm
long, 20-25 pm wide).

Dimensions. Pericarp (8 specimens); vertical
axis 12.5 (18.5) 23 mm, lateral axes 12.5 (16.8) 22
mm (in plane of dehiscence) x 10 (15.6) 20 mm
(at right angles to dorso-ventral plane). Seed (4
specimens); vertical axis 12-16 mm, lateral axis
8-10 mm.

Derivation of name. With reference to the type
locality and in allusion to The Welcome' gold
nugget, recovered during 1858, from a Deep
Lead at Bakery Hill near Ballarat, Victoria.

Remarks. The majority of specimens were
collected in an undehisced state (Figs 2-4), but
on drying after collection, the valves of several
of them separated either partially or wholly
from the apex and along the ventral suture
thereby to reveal their seed coats attached to
the pericarp wall (Fig. 17).

Distribution. Herberts on Welcome Creek
Drain, N of Bundaberg; near Clifton, Darling
Downs; Picardv Station near Moranbah Hole
RDPD98MA21, 123-133 m, Qld; Forest Reef,
near Orange, NSW.

Age range. Early Oligocene-early Late Miocene.

Species - Eureka turgida (F. Muell.)
comb. nov. (Figs 25-30)

1874a Conchotheca turgida Mueller, p.42; Pl.X, 8, 12 (non figs
5-7, 9-11)

Lectotype. NMVP53987; Mueller, 1874a, PI. X, fig. 8;
Figs 25-30 herein. One valve of globose unilocular
fruit, 12 mm long, 11 mm wide (dorsiventrally),
stalk inserted into basal depression. Pericarp wall
2-3 mm thick; seed attached apically, with remnants
of seed coat loosely adherent to pericarp in basal
half of locule.

Type locality. Nintingbool, (Crucible Co.Shaft, -23.2
m) near Haddon, SW of Ballarat, Victoria; basal
sediments of a sequence beneath basalt: Oligocene-
Miocene.

Diagnosis. Fruit woody, indehiscent or tardily
dehiscent, spheroidal, unilocular, with 1 seed.
Stalk 2-2.5 mm in diameter, inserted into basal
depression 4-5 mm in diameter; style base

152

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Fossil Proteaceous fruits

FIGS 17-30 Eureka welcomensis sp. nov. and Eureka turgida (F. Muell.) comb, nov.; 17-24, Eureka welcomensis so.
nov 17, Interior surfaces of opened fruit with presented seed (arrowed), QMF51 148 ; scale bar = 5 mm; 18,
19 Internal and external surface, showing style base (arrowed), of one valve of opened fruitQMh51 14/; scale,
b a ' r = 5 m m; 20, Cuticle of funicle and seed coat at base of seed, QMF51 150; scale bar = 100 pm; 21, 22, Inner
laver of seed coat as viewed under light and scanning electron microscopes, cuticle showing impressions ot
polygonal-shaped transfer cells, QN&5U50; scale bar = 100 pm; 23, 24, Outer layer of seed coat as viewed
under light and scanning electron microscopes, cuticle showing elongate cells, QMF51150; scale bar 100
unr 25-26, Eureka turgida (F. Muell.) comb, nov., lectotype, interior surfaces of opened fruit showing traces to
stalk (s) and stigma (st) and remnants of seed coat (arrow), NMVP53987; scale bar -2.5 mm, 27, Lee otypeas
illustrated by Mueller, NMVP53987; scale bar =2.5 mm; 28, Detail of pericarp wall NMVP53987; scale bar - 1
mm; 29, 30, Seed coat tissue, outer and inner layers respectively NMVP53987; scale bar - 10 pm.

Memoirs of the Queensland Museum 1 Nature • 2010 • 55(1)

153

Dettmann & Clifford

represented by an ellipsoidal groove, the
long axis 2-3 mm and oriented transverse to
the dorso-ventral plane on lateral surface of
valves; groove underlain by vascular strands
that extend into the pericarp and terminate
near the seed cavity. Pericarp with a near-
smooth outer surface; wall 2.5-3 mm thick in
mid regions of lateral surfaces, tapering to 1.5-2
mm at base of style; composed of thin exocarp
overlying a thick vascularised mesocarp and a
thin cuticle-like endocarp. Mesocarp a dense
groundmass of fibre sheaths that surround the
vasculature and are predominantly aligned in a
radial direction. Preserved seed coat comprises
exotesta of a thin, structureless cuticle, an
endotestal cuticle bearing faint impressions of
polygonal cells (40-50 urn in diameter) having
occasional internally directed fimbriae, and a
tegmic cuticle with impressions of anticlinal
walls of rectilinear cells (80-120 pm long, 20-25
pm wide).

Dimensions. Pericarp (2 specimens); vertical
axis 11, 12 mm, dorsiventral axis 10, 11 mm.

Comparison. E. turgida is smaller and the
pericarp wall is thinner than that of £. welcomensis.
Moreover, the latter species lacks a basal
depression as occurs in E. turgida.

Distribution and age. Known only from the
type lbcality: Oligocene-Miocene.

Species - Eureka spechtii sp. nov. (Figs 31-51)

1874a Conchotheca turgida Mueller, p.42; Pl.X, 5-7, 9-11 (non
figs 8,12)

Holotype. (here designated) NMVP53958 (Mueller,
1874a, PI. X , fig. 7), Figs 31-34 herein. Whole fruit,
unilocular, ellipsoidal, vertical axis 13.5 mm, lateral
axes each 10.5 mm. Stalk scar 3 mm.

Other material. NMVP52980, NMVP53096,

NMVP53098, NMVP53960, QMF13208.

Type locality. Nintingbool, (Crucible CoShaft, -23.2 m)
near Haddon, SW of Ballarat, Victoria; basal sediments
of a sequence beneath basalt: Oligocene-Miocene.

Diagnosis. Fruit woody, asymmetrically prolate
ellipsoidal, near-smooth externally, unilocular,
1-seeded; indehiscent or tardily dehiscent,
dehiscing from the apex along the sinuous, ridged
ventral suture. Stalk scar indistinct or visible, 0.5-
1.5 mm in diameter, inserted at base; style base
indistinct, apical, underlain by vascular strands
that extend into the pericarp and terminate near
the seed cavity. Pericarp wall 2-2.5 mm thick
in mid regions of lateral surfaces, thickening to
2.5-3.5 mm at base of style; composed of thin
exocarp overlying a thick radially vascularised
mesocarp and a thin cuticle-like endocarp.
Mesocarp 1. 8-2.2 mm in thickness in mid
regions of lateral surfaces, composed of a dense
groundmass of fibre sheaths that surround the
vasculature and arranged predominantly in a
radial direction. Preserved seed coat comprises
exotesta of a thin, structureless cuticle, an
endotestal cuticle bearing faint impressions of
polygonal cells (40-50 pm in diameter) having
occasional internally directed fimbriae, and a
tegmic cuticle with impressions of anticlinal
walls of rectilinear cells (80-120 pm long, 20-25
pm wide).

Dimensions. Pericarp (8 specimens); vertical
axis 10 (11.8) 14 mm; lateral axes 8 (9.8) 11 mm
(in plane of dehiscence) x 7.5 (8.4) 10.5 mm (at
right angles to dorsal-ventral plane).

Derivation of name. Named in honour
of Raymond Louis Specht, distinguished
Australian ecologist.

Comparison. E. spechtii differs from E. turgida
in shape (prolate ellipsoidal vs. spheroidal),
wall thickness (thickest vs thinnest in apical
regions) and the ventral suture line (sinuously
ridged vs. straight).

Distribution. Nintingbool (Mueller, 1874a, 1874b);
Foster (Deane, 1925), Victoria; Brandy Creek
(Johnston 1880), Tasmania; Darling Downs (see
Johnston 1880, p.27), Queensland.

154

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Fossil Proteaceous fruits

Age range. Early Oligocene-early Late Miocene.

Species - Eureka harslettiae sp. nov.

(Figs 52-64)

Holotype (here designated).QMF51 153, Figs 52-
53. Whole fruit, unilocular, ellipsoidal, vertical axis 19
mm, lateral axes each 15.5 mm. Stigma scar 2.5 mm
wide at apex , tapering along its length (7.5 mm) on
both lateral surfaces.

Other material. QMF 51154-QMF 51157 inclusive.

Type locality. South Blackwater Coal Pty Ltd Hole
R8736, 82 m, Queensland: Early-Late Oligocene.

Diagnosis. Fruit woody, indehiscent or tardily
dehiscent, with a near-smooth surface; prolate
ellipsoidal to subspheroidal with a prominent
ridge (2-3 mm high) that encircles the stone in
the longitudinal plane and situated along the
ventral suture and dorsal hinge line; unilocular,
1-seeded. Stalk attachment basal, inconspicuous;
style base represented by a V-shaped groove
oriented perpendicular to the dorso-ventral
plane and extending 7-9 mm from apex on
lateral surface of each valve; groove widest (1.5-
2.5 mm) at apex; underlain by vascular strands
that extend into the pericarp and terminate near
the seed cavity. Pericarp wall 4-6 mm thick, but
thinning slightly around base of style; mostly
composed of vascularised mesocarp external to
a thin, cuticle-like layer (?endocarp). Mesocarp
3.S-5.8 mm in thickness, composed of a dense
groundmass of fibre sheaths that surround the
vascular mesh and are predominantly aligned
in a radial direction. Seed coat comprises exotesta
of a thin unstructured cuticle, and an endotestal
cuticle bearing faint impressions of polygonal
cells (40-50 pm in diameter) having sparsely
distributed, internally directed fimbriae.

Derivation of name. Named in honour of
Morwenna Jean Harslett, Queensland naturalist.

Dimensions. Pericarp (12 specimens); vertical
axis 15 (16.7) 21 mm; lateral axes 13 (15.5) 21
mm (in plane of dehiscence) x 8 (11.8) 15 mm
(at right angles to dorsal-ventral plane).

Remarks and Comparison. The holotype and
all but one of the other specimens examined are
charcoalified and their internal morphological
and anatomical features are preserved. The
sole permineralised example studied is known
only from its external morphology. The species
is similar in size to E. welcomensis, but is distinct
in possessing a prominent ridge that encircles
the fruit along the ventral suture and dorsal
hinge line. Moreover, the V-shaped groove of the
stigma scar is longer than that in E. welcomensis.

Distribution. South Blackwater Coal Pty
Ltd Hole R8736, 82 m; Glencoe Station, near
Capella, Queensland.

Age range. Early-Late Oligocene.

AFFINITIES OF EUREKA

Pericarp vasculature of Eureka is consistent
with that of fruits borne by several genera of the
Proteaceae included within tribe Macadamieae,
and in particular by members of subtribes
Macadamiinae, Gevuininae, and Virotiinae
(sensu Weston and Barker, 2006; Mast et al.,
2008). The pericarps of genera included within
these subtribes possess a complexly branched
vascular system surrounded by sclerenchyma
and/or fibre sheaths and/or stone cell complexes.
The main vascular bundles are predominantly
vertically aligned, occurring within the middle
region of the mesocarp, and external to a narrow
zone of parenchyma and/or sclereids of the
innermost mesocarp. The vertical bundles
branch radially, the radial branches extending
into the outer region of the mesocarp where
they are sheathed by sclerenchyma; the tissues
between them are composed of fibres, stone cell
complexes and/ or thin-walled (and sometimes
succulent) parenchyma. Thus, the middle region
of the pericarp with its vertically aligned,
sheathed vasculature is densely structured
and may be woody and endocarp-like,
whereas the outer region of the pericarp is
predominantly radially structured (Figs 65-68,

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

155

Dettmann & Clifford

FIGS 31-41. Eureka spechtii sp. nov. 31-34, Holotype in lateral (31, 32), apical (33), and basal (34) views showing
dorsal suture (ds) and stalk scar (s), \'MVP53598; scale bar = 2.5 mm; 35-37, External (35) and internal (36,3/)
views of pericarp showing style base (arrowed) and apically attached seed (arrow), NMVP53096; scale bar
= 2.5 mm; 38-39, Seed illustrated in Fig. 37 before (38) and’ after (39) treatment with sodium hypochlorite
showing positions of chalaza (ch) and micropyle (m), NMVP53096; scale bar = 1 mm; 40, Transfer cells of
seed coat of holotype, NMVP53598; scale bar = 10 pm.

156

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Fossil Proteaceous fruits

FIGS 41-51. Eureka speclitii sp. nov. 41-45. Lateral (41-43), apical (44) and basal (45) views of pericarp,
NMVP53960; scale bar = 2.5mm. 46-47. Internal views of pericarp, NMVP53098; scale bars = 2.5 mm and
0.5mm respectively. 48. Seed morphology showing chalaza (ch) and micropvle (m), NMVP53098; scale bar =
1 mm. 49. Locule tissue, NMVP53598; scale bar = 40 Jim. 50, 51. Anatomy of seed coat showing transfer cells
(50) and inner cuticle of elongate cells (51), NMVP53598; scale bars = 40pm.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

157

Dettmann & Clifford

FIGS 52-64. Eureka harslettiae sp. nov.; 52, 53, Holotype, lateral (52) and apical (531 views, QMF51153; scale bar
= 5 mm; 54, 55, Internal views of fruit cut transversely, showing seed (S) in locule and cavities (arrows) from
which vascular bundles have been eroded from pericarp, QMF51154; scale bar = 5 mm; 56, Internal view
of pericarp cut longitudinally, QMF51155; scale oar = 5 mm; 57, 58, Lateral and apical views, QMF51156;
scale bar = 5 mm; 59, 60, Lateral and apical views of permineralised specimen, QMF51157; scale bar = 5 mm;
61-63, Longitudinal sections of pericarp showing vascular tissue (v) surrounded by fibre bundle complexes,
QMF51155; scale bars = (100 pm, 0.5 mm, 1 mm respectively); 64, Longitudinal section of pericarp showing
radial orientation of fibre complexes surrounding partially eroded vascular bundles of the inner mesocarn
region, QMF51155; scale bar = 1 mm.

158

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Fossil Proteaceous fruits

TABLE 1. Fruit and seed wall characters of members of Subtribes Macadamiinae and Gevuininae (sensu Weston
& Barker 2006), and fossil! c r ^eka .\ Information sources, additional to present observations, on development
and/or anatomy are from Filla (1926), Francis (1928), Hartung & Storey (1939), Hohman (1978), and Strohschen
(1986a,b).

Subtribe/
fossil taxon

Fruit type/
dehiscence

Peri

:arp

Seed coat

Exocarp

Outer mesocarp

Middle

mesocarp

Inner

mesocarp

oe., outer
integument

ie., outer
integument

Seed sheath

Macadamiinae

follicle or

drupaceous

follicle;

?delayed

dehiscence

epidermal
cells with
stomata and
trichomes

leathery, radial
vase, and fibres

leathery vert,
vase.; outer
surface smooth

parenchyma '

woody
or ?thin
cuticle
( Panopsis )

crystal layer,

polygonal

cells

thin, fragile

1 -layered
(derived
from inner
integument)

Gevuininae

(excluding

Cardwellia)

drupaceous/

follicle;

delayed

dehiscence

epidermal
cells with
stomata and
trichomes

succulent-
fibrous, radial
vase, and fibres

woody, vert,
vase.; outer
surface smooth

parenchyma

and sdereids
adherent
to testa

thin cuticle

crystal layer,

polygonal

cells

1 -layered
with elongate
cells (derived
from cuticle
between outer
and inner
integuments)

Eureka

drupaceous/

follicle;

delayed

dehiscence

not known

?succulent,

radial vase,
and fibres

woody, vert.

vase.; outer
surface smooth

parenchyma

and sdereids
adherent
to testa

thin cuticle

crystal layer,

polygonal

cells

1-layered with
elongate cells

73-77, 80A-B). This pattern of vasculature was
detailed for Macadamia, Brabejutn, Hicksbeachia,
and Gevuina by Filla (1926) and designated as
fruit Type VI among the pericarps of Proteaceae
known to him. Later, the vasculature pattern
of pericarps was identified as a characteristic
attribute (Johnson & Briggs 1973; Douglas 1995)
of genera now included within subtribes
Macadamiinae, Gevuininae, and Virotiinae (sensu
Weston & Barker 2006).

Within these subtribes, fruits with a vascu-
larised, radially structured pericarp and bearing
1-2, vertically attached, wingless seeds as occurs in
Eureka, are borne by Macadamia F. Muell., Panopsis
Salisb., Brabejum L. and Lasjia P.H. Weston &
A.R. Mast (subtribe Macadamiinae); Heliciopsis
Sleumer, Athertonia L. A.S. Johnson & B.G. Briggs,
and Vi ro tin L.A.S. Johnson & B.G. Briggs (subtribe
Virotiinae); and Sleurmerodendon Virot, Euplassa
Salisb., Gevuina Molina, Bleasdalea F. Muell.,
Hicksbeachia F. Muell., Kerrnadecia Brongn. & Gris.,
and Turrillia A.C.Sm. (subtribe Gevuininae)

(Sleumer 1 955; Virot 1 968; Johnson & Briggs 1975;
Smith & Haas 1975; Smith 1985; Steyermark
2004; Weston & Barker 2006; Mast et al. 2008).
Tire vascularised tissues of fruits of the extant
genera are of mesocarp origin as confirmed from
detailed developmental studies (Francis 1928;
Hartung & Storey 1939; Strohschen 1986a,
b) of fruits of Macadamia temifolia F. Muell.,
M. in tegrifolia Maiden & Betche (Subtribe
Macadamiinae), and Hicksbeachia pinnatifolia
F. Muell. (Subtribe Gevuiinae). Furthermore, an
endocarp s.str. (ie. a proliferation of woody tissue
derived from the inner epidermis of the ovary
wall) is not fomred in mature fruit of the studied
taxa. It seems likely that the woody or sometimes
leathery 'endocarp' referred to by some authors
(e.g. Sleumer 1955; Virot 1968; Smith & Haas
1975) in morphological descriptions of fruits
of the extant taxa within subtribes Virotiinae,
Gevuininae, and Macadamiinae instead repre-
sents tissues of the mesocarp.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

159

Dettmann & Clifford

FIGS 65-79. Fruits and seeds of Hicksbeachia pinnatifolia F. Muell. (65-72) and Gevuina avellana Molina (73-78),
and fruit of Athertonia diversifolia (C.T. White) L.A.S. Johnson & B.G. Briggs (79); 65-68, Pericarp Hicksbeachia
pinnatifolia, longitudinal (65, 67) and transverse (66, 68) sections of pericarp showing vasculature comprising
vertically aligned primary bundles (pb) of inner mesocarp and radially aligned secondary bundles (sfcp
extending into the exocarp; scale bars 65, 66 = 10 mm, 67 = 1 mm; 69, Seed H. pinnatifolia, showing chalaza
(ch) and micropyle (m); scale bar = 10 mm. 70-72. Tissues associated with seeds of H. pinnatifolia; 70, Inner
epidermis of testa showing polygonal cells with internally directed fimbriae (crystal layer); scale bar = 50
pm; 71, Endocarp sclereids detached from pericarp and closely adherent to testa; scale bar = 100 pm; 72,
Chalazal tissue showing vasculature; scale Dar = 0.5 mm; 73-77, Pericarp of Gevuina avellana, lateral view
(73), longitudinal (74, 76), and transverse (75, 77) sections; scale bar = 73-75 = 5 mm, 76, 77 = 1 mm; 78,Seed
tissue G. avellana, inner epidermis of testa showing polygonal cells with internally directed fimbriae (crystal
cells); scale bar = 50 pm; 79, Athertonia diversifolia, pericarp with exocarp removed, lateral view showing
surface reticulation of mesocarp; scale bar = 5 mm.

160

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Fossil Proteaceous fruits

LS TS

FIG. 80. Diagram illustrating pericarps of A,
Hicksbeachia pinnatifolia and; B, Eureka welcomensis
in longitudinal (LS) and transverse (TS) sections,
ip, inner pericarp; mp, middle pericarp; op, outer
pericarp; st, stigma scar; vb, vascular bundles.

Fruits of several genera within the Maca-
damiinae (Macadamia, Nothorites, and Lasji) are
characterised as follicular (Willis 2007; Mast et
al. 2008) as demonstrated from developmental
studies of Macadamia (Francis 1928; Hartung
& Storey 1939; Strohschen 1986a). The fruits
dehisce along the entire or partial length of the
ventral suture of the mature pericarp. Fruits
of the other extant Macadamiinae ( Brabejum
and Panopsis) are characterised as indehiscent
to tardily indehiscent (Willis 2007; Mast et
al. 2008). However, by contrast to the woody
pericarp of Eureka, the pericarp of Brabejum lacks
significant lignified tissue (Filla 1926), whereas
detailed anatomical and developmental studies
have yet to be undertaken on fruits of Panopsis
(Edwards & Prance 1993).

Developmental studies confirm that the indehi-
scent to tardily dehiscent pericarps of Hicksbeachia

pinnatifolia are transitional between a follicle and
a nut (Strohschen, 1986b). Fruits of this and other
taxa included in subtribe Gevuininae (except
Cardioellia) and those of subtribe Virotiinae
are described as 'drupaceous', in possessing a
woody, vascularised middle mesocarp and
an outer, somewhat fleshy mesocarp with
radially orientated vascular and fibre bundles
(Weston & Barker 2006, p. 338; Mast et al. 2008).
In fruits of the Virotiinae, the woody middle
mesocarp possesses a reticulate or pitted
outer surface (Sleumer 1955; Rozefelds 1990,
1992; Fig. 79) whereas, except for Sleumerodendron
austrocaledonicum, the middle mesocarp of
indehiscent fruits of the Gevuininae has a smooth
surface (Figs. 73, 80A). The surface of the
woody middle mesocarp of S. austrocaledonicum
is sculpted into a series of irregular branched
longitudinally-aligned, ribbon-like strips that
are separated by transversely disposed fibres
(Association Endemia 2001).

Thus the smooth-surfaced, vascularised,
radially structured pericarps of Eureka are more
similar in morphology to the indehiscent/ tardily
dehiscent fruits of Panopsis (Macadamiinae)
and the Gevuininae than to those of the Virotiinae
and Malagasiinae, each of which has a well
defined synapomorphy. In the Virotiinae the inner
mesocarp surface is pitted or reticulate, and in the
Malagasiinae the outer mesocarp lacks prominent
radiating vascular bundles (Weston & Barker
2006). Moreover, in Macadamiinae seeds possess
a structureless, fragile seed sheath derived from
the inner integument, whereas in studied seeds of
the Gevuininae the seed sheath is a cuticle with
elongate cells derived from a cuticle between the
inner and outer integument (Table 1). Seeds of
Eureka have an inner seed layer morphologically
comparable to that of Bleasdalea, Gevuina and
Hicksbeachia As the phylogeny of the character
states displayed by Eureka has not been explored,
it is uncertain whether they are synapomorphic
of the Gevuininae; this matter is the subject of
ongoing research.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

161

Dettmann & Clifford

Fruits of Eureka are prolate ellipsoidal, near
symmetrical about the vertical axis, with the
stalk and style in vertical alignment and the
dorsal hinge line and ventral suture near equal
in length. Extant Gevuininae having fruits
possessing these features are Gevuina, Hicksbeachia,
Bleasdalea, Euplassa, and Turrillia. Indehiscent or
tardily dehiscent fruits of other genera comprising
Subtribe Gevuininae differ in being asymmetric
with respect to the vertical axis ( Sleumerodendron
- see Smith & Haas 1975) or in possessing an
angled inner mesocarp ( Kermadeda - see Smith
& Haas 1975). Eureka possesses a single,
wingless, orthotropous seed with a two layered
seed coat, characters that are in accord with those
detailed for Gevuina and Hicksbeachia (Hohmann
1978; Strohschen 1986b) and observed by us
in these two genera and in Bleasdalea (Figs 70-
72, 78). In each of the fossil species, the outer
layer of the seed coat is closely adherent to
the inner surface of the pericarp (Figs 65, 74)
and envelops a cuticle bearing impressions
of polygonal cells having internally directed
fimbriae; these layers are interpreted to represent
the outer and inner epidermis of the testa
as has been precisely detailed for Gevuina and
Hicksbeachia (Hohmann 1978; Strohschen 1986b).
Strohschen's developmental studies confirm that
in Hicksbeachia pinnatifolia F. Muell. tissues of the
ovary wall and the outer layer (outer epidermis
of outer integument) of the seed coat may be
fused as is evident in Gevuina avellana Moliner
(Hohmann 1978) and in Eureka welcomensis.
Moreover, in both extant taxa (and in most
grevilleoid genera) the inner epidermis of the
testa is composed of polygonal-shaped, calcium
oxalate-bearing, transfer cells with internally
directed fimbriae (Hohman 1978; Strohschen
1986b). A further cuticle with impressions of
elongate cells occurs beneath the testa and
develops from the cuticles between the inner and
outer integuments and the cuticle surrounding
the nucellus (Strohschen 1986b); a morpho-
logically comparable cuticle occurs beneath the
testal cuticles in Eureka welcomensis, E. turgida,

and E. spechtii (Figs 23, 24, 29, 51). Anatomical
and morphological features of fruits of extant
species of Turrillia and Euplassa (Gevuininae)
and Panopsis (Macadamiinae) are unavailable
and so detailed comparisons between these
extant fruits and fossil Eureka are precluded.

In summation, fossil fruits included within
Eureka have pericarps that possess a mosaic of
morphological, anatomical, and organisational
attributes consistent with pericarps of the
Macadamiinae and Gevuinineae. However,
seed coat structure of Eureka is more similar to
that of the Gevuininae than the Macadamiinae
(Table 1) and so the fossil taxon is provisionally
placed in the Gevuininae.

IMPLICATIONS OF EUREKA

The fossils herein allocated to Eureka collectively
imply a wide distribution of their parental
plants in eastern Australia during Oligocene
and Miocene times. Oldest occurrences are in
sediments beneath basalts dated as 27-34 Mya
(Oligocene) in central Queensland (Picardy,
Blackwater) and youngest well-dated occurrences
are also in sediments beneath basalts of 11.2-13
Mya (late Middle-early Late Miocene) at Forest
Reefs Mine, New South Wales. Other localities
from which Eureka has been reported are less
precisely dated, but are within an Oligocene-
Miocene age frame (34-5 Mya). If Eureka is a
member of Gevuininae, this age range has
implications for the history and past distribution
of subtribe and for age calibrations relevant to
divergence of those taxa of the subtribe having
tardily dehiscent fruits with wingless seeds.
Extant members of the subtribe that bear tardily
dehiscent fruits with 1-2 wingless seeds are
Hicksbeachia and Bleasdalea in northeastern
Australia-New Guinea, Kermadeda and
Sleurmerodendron in New Caledonia, Turrillia
in Fiji and Vanuatu, and Euplassa and Gevuina in
South America (Weston & Barker 2006). Euplassa
with 20 species has the widest distribution

162

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Fossil Proteaceous fruits

range (mid-low latitudes of South America)
whilst each of the other six genera has a narrow
distribution range and few species (Fig. 1). Initially
allocated to three separate subtribes within tribe
Macadamieae (Johnson & Briggs 1975) molecular
and morphological analyses (Hoot & Douglas
1998; Weston & Barker 2006) demonstrated the
above-named genera together with Cardwellia
(follicular fruit with winged seeds) form a
monophyletic clade. The biogeography of the
clade was investigated subsequently to test
whether the present trans-southern Pacific
distribution of the subtribe is better explained
by vicariance or by transoceanic dispersal (Barker
et al. 2007; Mast et al. 2008; Sauquet et al. 2009).

Age estimates based on Bayesian phylo-
genetic and molecular dating together with
fossil calibration points returned an age of
51.4 (±10.5) Mya for divergence of subtribe
Gevuininae (Barker et al. 2007, Fig.2, Node
F), an age consistent with a vicariance model
for its present distribution. By contrast, in
their analyses of a more refined phylogeny
of tribe Macadamieae (Mast et al. 2008) the
minimum age for the most recent common
ancestor (MRCA) of subtribe Gevuininae was
set at 34 Mya based on the age of fossil cuticles
assigned to the Gevuininae and reported from
Middle Eocene sediments (Carpenter & Pole
1995). Even so. Mast et al. (2008) considered
the analyses failed to discriminate between
disjunction of the subtribe at 34 Mya or 13.6 Mya,
and hence between a vicariance or dispersal
model to account for the extant distribution of the
genera in the Gevuininae. However, they propose
an inferred age of 11.5 Mya for divergence
between Bleasdalea (New Guinea and Australia)
and its sister Hickslvachia (Australia), and 8.1 Mya
for disjunction between the sister taxa Kermadecia
(New Caledonia) and Turillia (Fiji and Vanuatu).
Accordingly, they argue that long distance
dispersal accounts for the distribution of the latter
sister taxa since the inferred age of disjunction
postdates the tectonic events that account for

the present day geography of this part of the
southwest Pacific (Schellart et al. 2006). They
propose that tardily dehiscent fruits with a
spongy pericarp may be sufficiently buoyant
for long distance water dispersal and suggest
the Antarctic Circumpolar Current (ACC) may
serve as a possible mechanism. This current is
now believed to have been initiated in the Late
Oligocene at 25-23 Mya (Lyle et al. 2009), and is
confined to latitudes of approximately 40°S.

A further analysis based entirely on molecular
data (Sauquet et al. 2009) yielded a cladogram
whose topology for the subtribe differed
from that presented by Mast et al. (2008), and
different crown group age estimates for the
constituent extant members of the clade. The age
constraint of 70.7 Mya selected for the MRCA
of crown Macadamiaeae (Sauquet et al. 2009;
Fig. 1, Node I) is based on the Late Cretaceous
occurrence of the fossil pollen taxon Propylipollis
crotonioides Dettmann & Jarzen, which in
parsimony analyses was considered suitable
for calibration (Sauquet et al. 2009, Supporting
Information, Fig. S2, Table SI). On this basis, the
crown group age of Gevuininae was estimated
as 37.4 Mya and 19.2 Mya for the crown group
age of the clade sister to Cardwellia (comprising
Bleasdalea and its sister taxa, all of which bear
tardily dehiscent fruits). As discussed above
the fruits of Eureka closely resemble those of
Bleasdalea, Gevuina and Hicksbeacliia and so
the occurrence of Eureka in sediments as old
as, or older than, 34 Mya (ie Early Oligocene)
suggests that the crown group age of the sister
clade to Cardwellia, proposed by Sauquet et al.
(2009), may need to be reconsidered.

Based on occurrences of Eureka reported
herein, the Oligocene-Miocene distribution of
members of the subtribe having tardily dehiscent
fruits spanned some 20° of latitude (~60°- 40°S
palaeolatitude) in eastern Australia (Fig IB).
Other fossils reported from within this latitudinal
band and accepted as those of the Gevuininae
(Mast et al. 2008) are cuticles from the Middle

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

163

Dettmann & Clifford

Eocene of southern Western Australia (Carpenter
& Pole 1995) and the Early Miocene of New
Zealand (Pole 1998). Thus, the fossil evidence
demonstrates a former trans-Tasman distribution
of the Gevuininae no longer mirrored in the
present vegetation.

Questionably allied to the subtribe is fossil
wood reported as similar to, but not identical
with, that of Gevuina from the Oligocene of
southern South America (Pujana, 2007). The
pollen taxon, Propylipollis reticuloscabratus (Harris)
Martin & Harris, which is widely distributed in
Australian Campanian-Pleistocene sediments
and known also from the Campanian-Eocene
of the Antarctic Peninsula (Dettmann & Jarzen
1996), may not represent evidence of the subtribe
in South America, as has been determined by
Sauquet et aL (2009, Supporting Information).
Though these authors undertook a cladistical
analysis embracing pollen of all extant genera
of Proteaceae and 25 fossil proteaceous-like
pollen taxa, the results must be treated with
caution because of the 22 pollen characters
scored for pollen of extant species relatively
few were available for most fossil taxa.

ACKNOWLEDGEMENTS

The authors thank E.Thompson, Museum of
Victoria and R. Jones, Australian Museum for
loan of fossil material from Victoria and New
South Wales; to T. Spencer, South Blackwater
Coal, G. Muggeridge, S. Mepes, and H.
Rewald, CRA, and to P. Baker, Department of
Natural Resources, Bundaberg for collection of
fossil fruits. D. Lee, Department of Geology,
University of Canterbury, New Zealand, P.
Weston, Royal Botanic Gardens, Sydney, A.
Specht, School of Environmental Science
and Management, Southern Cross University,
Lismore, G. Guymer, Queensland Herbarium,
R. McKinnon, Brisbane Botanic Gardens, and
W. Cooper, Atherton, Queensland are thanked
for securing fruits of extant Gevuina avellana,

Bleasdalea bleasdalei, Hicksbeachia pinnatifolia
and H. pilosa. G. Guymer and P. Weston kindly
arranged the loan of fruits of several species of
Virotia. We are especially grateful to P. Weston
and R. Hill for helpful reviewer's comments.

LITERATURE CITED

Association Endemia Nouvelle-Caledonie. 2001.
La Flore, Fiche Descriptive, Sleurmerodendron
austrocaledonicum. http:/ /www.endemia.
nc/plante/fichephp?code=1163 (accessed 25
August, 2009).

Barker, N.G., P.H. Weston, F. Rutschmann & H. Sauquet.
2007. Molecular dating of the 'Gondwanan' plant
family Proteaceae is only partially congruent
with the timing of the break-up of Gondwana.
Journal of Biogeography 34: 2012-2027.

Carpenter, R . )., & M.S. Pole. 1995. Eocene plant fossils
from the Lefroy and Cowan Paleodrainages,
Western Australia. Australian Systematic Botanu
8: 1107-1154. y

Deane, H. 1925. Tertiary fossil fruits from deep lead,
Foster, south Gippsland. Records of the Geological
Survey of Victoria 4: 489-492.

Dettmann, M.E., & H.T. Clifford. 2001. Fossil record
of Elaeocarpus fruits. Memoirs of the Queensland
Museum 46: 461-497.

2003. Miocene palynofloras from subsurface sedi-
ments near Bundaberg, southeastern Queensland.
Memoirs of the Queensland Museum 49: 261-267.

2005. Fossil fruit of the Grevilleeae (Proteaceae) in
the Tertiary of eastern Australia. Memoirs of the
Queensland Museum 51: 423-438.

Dettmann, M.E., & D.M. Jarzen. 1990. The Antarctic/
Australian rift valley: Late Cretaceous cradle
of northeastern Australasian relicts? Review of
Palaeobotany and Palynology 65: 131-144.

1996. Pollen of proteaceous-type from latest
Cretaceous sediments, southeastern Australia.
Alcheringa 20: 103-160.

1998. The early history of the Proteaceae: the pollen
record. Australian Systematic Botany 11: 401-438.

Douglas, A.W., 1995. Morphological Features. In
P. McCarthy (ed.), Flora of Australia Volume
16, Eleagna’ceae, Proteaceae I, 14-20. CSIRO,
Melbourne, Australia.

Edwards, K.S., & Prance, G.T. 1993. New Species
of Panopsis (Proteaceae) from South America.
Kew Bulletin 48: 637-662.

164

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Fossil Proteaceous fruits

Filla, F. 1926. Das pericarp der Proteaceae. Flora 120:
99-142.

Forsyth, S.M. 1989. The Tamar Graben. In C.F.
Burrett & E.L. Martin (eds). Geology and Mineral
Resources of Tasmania. Special Publication of the
Geological Society of Australia 15: 358-361.

Francis, W.D. 1928. The anatomy of the Australian
bush nut (Macadamia ternifolia). Proceedings of
the Royal Society of Queensland 39: 43-53.

Greenwood, D.R., A.J.V.Valdala., & J.G. Douglas. 2000.
Victorian Paleogene and Neogene macrofloras:
a conspectus. Proceedings of the Royal Society of
Victoria 112: 65-92.

Hartung, M.E., & W.B. Storey. 1939. The development
of the fruit of Macadamia ternifolia. Journal of
Agricultural Research 59: 397-406.

Hohman, B. 1978. Zur mikroskopischen Identifizierung
von chilenischen Haselnussen (Gevuina avellana
Molina). Zeitsclirift fiir Lebensmittel-Untersuchung
utid Forschung 166: 304-307.

Holdgate, G.R., M.W. Wallace, S.J. Gallagher, B.E.
Wagstaff & D. Moore. 2008. No mountains to
snow on: major post-Eocene uplift of the East
Victoria Highlands; evidence from Cenozoic
deposits. Australian Journal of Earth Sciences 55:
211-234.

Holdgate, G.R., M.W. Wallace, S.J. Gallagher, R.B.
Witten, B. Stats & B.E. Wagstaff. 2006. Cenozoic
fault control on 'deep lead palaeoriver systems,
Central Highlands, Victoria. Australian Journal
of Earth Sciences 53: 445-468.

Hoot, S.B. & Douglas, A.R. 1988. Phylogeny of
the Proteaceae based on aptB and aptB-rcL
intergeneric spacer region sequences. Australian
Systematic Botany 11: 301-320.

Johnson, R.W. 1989. Intraplate volcanism in eastern
Australia and New Zealand. (Cambridge University
Press and Australian Academy of Science:
Cambridge).

Johnson, L.A.S., & B.G. Briggs. 1975. On the Proteaceae
- evaluation and classification of a southern
Family. Botanical Journal of the Linneati Socieh/ 70:
83-182.

Johnston, R.M. 1880. Note on the discovery of
Spondylostrobus smythii Muell., and other
fossil fruits in the deep lead drift at Brandy
Creek Goldfield. Report of the Royal Society of
Tasmania for 1879: 81-90.

Lawver, L.A. & Gahagan, L.M. 2003. Evolution
of Cenozoic seaways in the circum-Antarctic

region. Palaeogeography, Palaeoclimatology,
Palaeoecology 198: 11-37.

Lyle, M., S. Gibbs, T.C. Moore & D.K. Rea. 2009. Late
Oligocene initiation of the Antarctic Circumpolar
Current: evidence from the South Pacific. Geolosv
35:691-694.

Martin, A.H.R. 1982. Proteaceae and the early
differentiation of the central Australian flora. In
W. R. Barker & P. J. Greenslade (eds.). Evolution
of the Flora and Fauna of Arid Australia, 77-83.
Peacock Publications, Adelaide.

Mast, A.R., C.L. Willis, E.H.Jones, K.M. Downs And
P.H. Weston. 2008. A smaller Macadamia from
a more vagile tribe: inference of phylogenetic
relationships, divergence times, and diaspore
evolution in Macadamia and relatives (Tribe
Macadamieae; Proteaceae). American Journal of
Botany 95: 843-870.

Mueller, F. 1871. New vegetable fossils of Victoria.
Reports of the Mining Surveyors and Registrars,
Victoria, Quarter ending 30th September, 1871:
39-41.

1873. New vegetable fossils of Victoria. Reports of
the Mining Surveyors and Registrars, Victoria,
Quarter ending 30th September, 1873: 41-42.

1874a. New vegetable fossils of Victoria. Reports of
the Mining Surveyors and Registrars, victoria.
Quarter ending 30th September, 1874: 41-42.

1874b. Observations on new vegetable fossils of the
auriferous drifts. Geological Survey of Victoria,
John Ferres, Government Printer, Melbourne,
Australia.

1879. Observations on new vegetable fossils of
the auriferous drifts. Reports of the Mining
Surveyors and Registrars, Victoria, Quarter
ending 31st December, 1878: 37-39.

1882. New vegetable fossils of Victoria. Reports of
the Mining Surveyors and Registrars, Victoria,
Quarter ending 31st March, 1882: 43-44.

1883. Observations on new vegetable fossils of the
auriferous drifts. Second Decade, Geological
Survey of Victoria, John Ferres, Government
Printer, Melbourne, Australia.

Pole, M.S. 1998. The Proteaceae record in New Zealand.
Australian Systematic Botany 11: 343-372.

Pujana, R. 2007. New fossil woods of Proteaceae from
the Oligocene of southern Patagonia. Australian
Systematic Botany 20: 119-125.

Rozefelds, A.C. 1990. A mid Tertiary rainforest flora
from Capella, central Queensland. In J.G. Douglas,
and D.C. Christophel (eds). Proceedings, Third

Memoirs of the Queensland Museum 1 Nature • 2010 • 55(1)

165

Dettmann & Clifford

International Organization of Palaeobotany
Conference, Melbourne, 1988, 123-136. A-Z
Publishers, Melbourne, Australia.

1992. The subtribe Hicksbeachiinae (Proteaceae) in the
Australian Tertiary. Memoirs of the Queensland
Museum 32: 195-202.

1995. Miocene Wilkinsonia fruits (Hicksbeachiinae,
Proteaceae) from the base of the Yallourn
Formation, Latrobe Valley, Victoria. Papers and
Proceedings of the Royal Society of Tasmania 129:
59-62.

Rozefelds, A.C., & D.C. Christophel. 19%. Elaeocarpus
(Elaeocarpaceae) fruits from theOligo-Miocene
of eastern Australia. Papers and Proceedings of
the Royal Society of Tasmania 130: 41-48.

Sauquet, H„ P.L. Weston, C.L. Anderson, N.P. Barker,
D.C. Cantrill, S.R. Mast, & V. Savolainen. 2009.
Contrasted patterns of hyperdiversification in
Mediterranean hotspots. PNAS, 106, 221-225 and
Supporting Information http://www.pnas.
org/ cgi/ content/ short/ 08056071 06

Schellart, W.P., G.S. Lister, & V.G. Troy. 2006. A late
Cretaceous and Cenozoic reconstruction of the
southwest Pacific region: Tectonics controlled
by subduction and slab rollback processes.
Earth-Science Reviews 76: 191-233.

Sleumer, H. 1955. Proteaceae. Flora Malesiana 5: 147-206.

Smith, A.C. 1985. Flora vitiensis nova, A new flora
of Fiji (Spermatophvtes only). Volume 3. Pacific
Tropical Botanical Garden, Honolulu, Hawaii,
USA.

Smith, A.C., & J. Haas. 1975. Studies of Pacific Island
plants. XXIX. Bleasdalea and related genera of

the Proteaceae. American Journal of Botany 62
133-147.

Steyermark, J.A. 2004. Proteaceae. In Steyermark
J.A., Berry, P.E., Yatskievych, K. & Holst, B.K
(eds). Flora of the Venezuelan Guayana, 8: 384-393
Missouri Botanical Garden Press, St Louis,
Missouri, USA.

Strohschen, B. 1986a. Contributions to the biology
of useful plants. 4. Anatomical studies of fruit
development and fruit classification of the
Macadamia Nut (Macadamia integrifolia Maiden
& Betche). Angewandte Botanik 60: 239-247.

1986b. Contributions to the biology of useful plants
5. Anatomical studies of fruit development
and fruit classification of the Monkey Nut
(Hicksbeachia pinnatifolia F. Muell. Angewandte
Botanik 60: 249-256.

Virot, R. 1968. Proteac6es. In A. Aubreville (ed.), Flore
de las Nouvelle-Catedonie et Dependances 2: 5-254.
Museum National d'Histoire Naturelle, Paris,
France.

Weston, P.H., & N.P. Barker. 2006. A new
suprageneric classification of the Proteaceae
with an annotated checklist of genera. Telopea
11: 314-344.

Willis, CL. 2007. Inference of phylogenetic relationships
in Macadamia and relatives (Tribe Macadamieae;
Proteaceae) using three chloroplast and three
nuclear DNA regions. Honors Thesis, Florida
State University D-Scholarship Repository.
Website http:/ / www.dscholoarship.lib.fsu.edu/
uncergrad/323 (accessed 30 October, 2007).

166

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A review of the Australian Polyrhachis ants of the
subgenera Myrmhopla Forel and Hirtomyrma subgen.
Nov. (Hymenoptera: Formicidae: Formicinae)

Rudolf J. KOHOUT

Queensland Museum, PO Box 3300, South Brisbane, Qld 4101, Australia. Email: rudolf.kohout@qm.qld.
gov.au

Citation: Kohout, R.J. 2010 03 15: A review of the Australian Polyrhachis ants of the subgenera
Myrmhopla Forel and Hirtomyrma subgen. nov. (Hymenoptera: Formicidae: Formicinae). Memoirs
of the Queensland Museum - Nature 55(1): 167-204. Brisbane. ISSN 0079-8835. Accepted: 13
April 2009.

ABSTRACT

The Australian species of the Polyrhachis subgenus Myrmhopla are revised. A total of eight
Australian species are recognised in four species-groups; four species in the sexspinosa-
group. two species in the bicolor- group and single species in each of the dives- and
mucronata- groups. A new subgenus Hirtomyrma is proposed to house ten species formerly
included within the P. viehmeyeri- group of the subgenus Myrmhopla. Polyrhachis dispar
sp. nov. is described and placed in the sexsp/'nosa-group. Polyrhachis bicolor nigripes is
raised to specific status and reported from Australia for the first time. The extralimital
species P. cyrtomyrmoides Donisthorpe is considered synonymous with P. mucronata
Fr. Smith. A neotype for P. sexspinosa (Latreillei) and lectotype for P. reclinata Emery
are designated. All species are illustrated and their distribution and nesting habits are
summarised. Keys to the subgenera of Australian Polyrhachis and to species of Myrmhopla
and Hirtomyrma are included. □ Polyrhachis , Myrmhopla, Hirtomyrma, bicolor-group,
dives-group, mucronata-group, sexspinosa-group, Australia, distribution.

Myrmhopla was established by Forel (1915) as
a subgenus of Polyrhachis Fr. Smith, 1857, with
Formica armata (Le Guillou, 1842) as the type
species. Forel did not define his new subgenus
but Emery (1925) later delimited Myrmhopla
as follows (translation): "Worker. - Dorsum of
thorax rounded, that is to say not marginate,
except in some species (groups cryptoceroides
and viehmeyeri); pronotal spines shorter than
propodeal, sometimes absent; metanotal groove
variable. Petiole variable amongst the groups and
species; body of petiole in form of an elongate
node, angled anterodorsally in profile or, to the
contrary, forming a thick scale, higher than
long, angular or rounded in front; generally
bearing single pair of spines very variable in
form, size and direction; rarely the spines are

hook-like; in many species where they form a
gaster embracing arc, there is between spines
also a pair of teeth or small vertical spines. First
gastral segment large. Female. - Very similar to
the worker, with spines usually stronger and
shorter."

When Emery published his diagnosis of
Myrvtlwpla, the subgenus already included some
140 species and subspecific forms. In an attempt
to partition the high degree of diversity within
such a large subgenus, he subdivided Myrmhopla
into six species-groups. Dorow (1995) divided
the subgenus further, recognising 16 species-
groups, the six proposed by Emery and ten that
he established as new. Five of these groups are

Memoirs of the Queensland Museum | Nature • 2010 • 55(1) • www.qm.qld.gov.au

167

Kohout

relevant to the Australian fauna; the bicolor,
dives, mucronata, sexspinosa and mehmeyeri-groups.

However, as mentioned by previous authors
(Bolton, 1975; Dorow, 1995), the large degree
of morphological diversity within Myrmhopla
presents problems with maintaining the sub-
genus as it was originally perceived. Virtually
none of the characters originally used by Emery
(1925) to define Myrmhopla consistently apply to
the species currently placed within the subgenus
and some characters vary within a single species-
group. The concept of the subgenus has widely
been criticised (Hung, 1967) and the formation
of numerous species-groups within Myrmhopla
has only partially alleviated the problem.

Considerable morphological differences
between various species-groups are evident
throughout the subgenus Myrmhopla, but
nowhere as markedly as in relation to the P.
viehmeyeri-group. For example, a marginate
mesosoma is a particularly significant character
separating species of the viehmeyeri-group from
the rest of Myrmhopla, except perhaps some
species of the extralimital P. crypitoceroides-group
(e.g. P. cryptoceroides Emery) (Kohout, 2006a).
In some respects, viehmeyeri-group species
resemble members of the subgenus Hedontyrma
Forel as they share a spinose and marginate
mesosoma and a petiole featuring a more-or-
less flat dorsum. However the characteristic
vermiculate sculpturation, bristle-like pilosity
and distinct reddish-brown colouration of
species of the viehmeyeri-group clearly separate
them from Hedomyrma species. The most
remarkable feature of viehmeyeri-group species
is their subterranean nesting habit combined
with a sophisticated parasitic relationship with
certain groups of ectatommine and poneroid
ants (Maschwitz et al., 2003). The morphological
and behavioural distinctness of the viehmeyeri-
group is further supported by a preliminary
molecular phylogeny of Polyrhacliis (S.K.A.
Robson, pers. comm.) that places the viehmeyeri-
group (i.e. P. loiueryi) closest to species of the

subgenus Chariomyrma Forel (P. lata Emery
and P. sokolova Forel) and rather distant from
representatives of other Myrmhopla species-
groups. Considering these facts, I believe that
the viehmeyeri-group should be removed front
the subgenus Myrmhopla and a new subgenus
Hirtomyrma is proposed below to incorporate
its constituent species.

METHODS

Publication dates and the spelling of species
epithets and authors' names follow Bolton et
al. (2007). This study is principally based on
the worker caste but notes are provided on
associated queens. Associated males of only a
few species are known and present in the ANIC
and/or QM spirit collections. Their diagnosis
is beyond the intended scope of this paper and
has not been attempted.

The localities at which ants were collected by
the Bishop Museum's collectors, were checked
against that institution's list of New Guinean
localities (BPBM, 1966, unpublished). In some
cases the latitude and longitude co-ordinates,
or altitude, are only roughly approximate. The
use of the terms 'New Guinea' or 'Bismarck
Archipelago' alone indicate the delimitation
of these regions in a biogeographic sense
regardless of current political boundaries.

Illustrations. Photographs of specimens were
taken with an Olympus SZX1 2 Stereomicroscope
and DP70 digital camera. Images were processed
using Helicon Focus (Mac OSX version) and
Photoshop CS2 (Adobe Inc., USA) software.
The holotypes of P. dispar sp. nov., P. bamaga
Kohout, P. eremita Kohout and P. rustica Kohout,
the paratype of P. loioeryi Kohout and type-
compared specimens from Australian localities
of other species are illustrated. The illustrations
of P. sexspinosa (Latreille) are of the neotype
designated below.

Standard measurements and indices. Measure-
ments and indices follow those of Kohout (2006):

168

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A review of the Australian Polyrhachis Ants

TL = Total length (the necessarily compc site
measurement of the outstretched lengt i of
the entire ant measured in profile); HL = F ead
length (the maximum measurable length 01 the
head in perfect full face view, measured f om
the anterior-most point of the clypeal bordi r or
teeth, to the posterior-most point of the occi lital
margin); HW = Head width (width of the 1 ead
in perfect full face view, measured immedic tely
in front of the eyes); Cl = Cephalic index ( TW
x 100/ HL); SL = Scape length (excluding the
condyle); SI = Scape index (SL x 100/HW); PW
= Pronotal width (greatest width of the pror otal
dorsum); MTL = Metathoracic tibial length (max-
imum measurable length of the tibia of the hind
leg). All measurements were taken using a
Zeiss SR stereomicroscope with an eyepiece
graticule calibrated against a stage micrometer.
All measurements are expressed in millimetres
(mm).

Abbreviations. Names of the most frequently
listed collectors are abbreviated as follows:
ANA - Alan N. Andersen; CJB - C.J. Burwell;
DJC - D.J. Cook; GBM - G.B. Monteith; RJK -

R. J. Kohout; RWT - R.W. Taylor; SKR - S.K.
Robson. Other abbreviations used in specimen
data are: NP - National Park; Pen. - Peninsula;
PNG - Papua New Guinea; R. - River; Ra. - Range;
Rd - Road; rf. - rainforest; Stn - Station; w -
worker/s.

Institutions and depositories, (with the
names of cooperating curators) AMNH -
American Museum of Natural History, New York,
NY, USA (Dr J.M. Carpenter); AMSA - Australian
Museum, Sydney, NSW, Australia (Drs D. Britton,
D. Smith); ANIC - Australian National Insect
Collection, CSIRO, Canberra, Australia (Dr

S. O. Shattuck); BMNH - The Natural History
Museum, London, UK (B. Bolton); BPBM -
Bernice P. Bishop Museum, Honolulu, HI,
USA (K.T. Arakaki); HNHM - Hungarian
Natural History Museum, Budapest, Hungary
(Dr J. Papp); IZAS - Institute of Zoology,
Ukrainian Academy of Sciences, Kiev, Ukraine

(Dr A.G. Radchenko); JCUT - James Cook
University, Townsville, Queensland, Australia (Dr
S.K.A. Robson); JWGU - Johan Wolfgang Goethe-
Universitat, Frankfurt am Main, Germany
(Prof. Dr U. Maschwitz); MCZC - Museum
of Comparative Zoology, Harvard University,
Cambridge, MA, USA (Dr S. Cover); MNHA -
Museum of Nature and Human Activities, Hyogo
Pref. University, Hyogo, Japan (Dr Y. Hashimoto);
MNHN - Museum National d'Histoire Naturelle,
Paris, France (Dr J. Casevitz Weulersse); MSNG -
Civic Museum of Natural History "G. Doria",
Genova, Italy (Drs R. Poggi, V. Raineri); MVMA

- Museum of Victoria, Melbourne, Vic., Australia
(Dr K. Walker); NMNH - National Museum of
Natural History, Washington, DC, USA (Dr T.R.
Schultz); OXUM - Hope Entomological Collec-
tions, University Museum, Oxford, UK (Drs C.
O'Toole, D.J. Mann); QM - Queensland Museum,
Brisbane, Qld, Australia (Dr C.J. Burwell); TERC

- Tropical Ecosysytems Research Centre, CSIRO
Sustainable Ecosystems, Darwin, NT, Australia
(Dr A.N. Andersen).

SYSTEMATICS

Genus Polyrhachis Fr. Smith, 1857

Polyrhachis Fr. Smith, 1857: 58. Type species: Formica bihamata
Drury, 1773 by original designation.

KEY TO AUSTRALIAN SUBGENERA OF
POLYRHACHIS
(based on worker caste)

1. Mesonotum armed with a pair of upwards
and backwards curved spines; petiole
distinctly higher than full height of mesosoma,
terminating in a pair of hook-like spines
(Fig. IK) (arboreal) (Cape York Pen.)
. . . .Polyrhachis (only P. bellicosa Fr. Smith)

- Mesonotum without spines; petiole lower

than full height of mesosoma, dorsum armed
with spines or teeth of various configurations
or virtually unarmed 2.

2. Mesosoma completely laterally immarginate

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

169

Kohout

— Mesosoma at least partly laterally marginate
5.

3. Small species (HL 1 .25-1 .65); dorsum of meso-

soma relatively short, strongly longitudinally
and transversely convex; pronotal spines
reduced to acute teeth or absent; body
uniformly black, highly polished 4.

— Small to large species (HL 1.40-3.60); dorsum
of mesosoma elongate, only weakly to
moderately convex; pronotal spines relatively
long, acute; colour of body variable, mostly
black, but also reddish-brown or bicoloured
(Figs 1A-F, 4A-F) (arboreal or lignicolous)
(tropical Qld and NT) . . . Myrmhopla (part)

4. Sides of head with longitudinal carina

separating gena from ventral parts of head;
propodeal spines, if present, very short;
petiole scale-like, armed with four spines or
teeth of various lengths and configurations,
(Fig. 1C) only rarely reduced to mere
denticles (P. brevinoda Kohout) (arboreal) (NT,
Qld and coastal NSW) Cyrtomyrma

— Sides of head without longitudinal carina;
propodeal spines long; petiole columnar,
armed with a pair of horizontal spines that
conform to shape of first gastral segment and
a pair of distinct intercalary teeth (Fig. 3E,
F) (arboreal) (Cape York Pen. and north Qld)

Mvrmhopla (part) (only P. mucronata

Fr. Smith)

5. Pronotal humeri simply rounded or, at
most, bluntly angular (Figs 1A, D) 6.

— Pronotal humeri armed with spines or
acute teeth (Figs 1 E, G) 9.

6. Dorsum of petiole usually narrowly rounded,
rarely with a distinct platform (P. thusnelda
Forel), armed with a pair of more-or-less
horizontal, backwards directed or diverging,
acute spines; propodeal spines rarely hook-
like (P. ammonoeides Roger); propodeal spines
always present (Fig. ID), mostly horizontal
or weakly elevated (subterranean, rarely
lignicolous or lithocolous) (Australia-wide,
except south, south-west and Tasmania)

Hagiomyrma

— Dorsum of petiole variable, with two, three

or four upward directed spines or teeth of
various lengths and configurations, or petiole
virtually unarmed; propodeal spines present
or reduced to mere teeth 7.

7. Small species (HL 1 .15-1 .65); petiole columnar

with two or three spines; body light coloured,
mostly yellowish- or reddish-brown (arboreal)
(tropical north Qld and NT only) 8.

— Small to large species (HL 1.10-2.80);

petiole scale-like, usually with four teeth
or short spines, but rarely also with one
(P. pseudothrinax Hung) or two (e.g. P.
prometheus Santschi) elongated spines or
virtually unarmed with only shallow median
emargination (Fig. 1A); body mostly black
or rarely reddish-brown (P. incerta Kohout)
(subterranean or lignicolous, nocturnal
and crepuscular foragers) (Australia-wide,
incl. Tasmania). Campomyrma

8. Petiole armed with three spines, middle spine

distinctly longer than lateral spines; propodeal
spines acute, distinctly elevated or virtually
vertical (Fig. 1J) (arboreal) (NT and north
Qld) Myrmothrinax

— Petiole armed with two short spines;
propodeal spines reduced to short, up-
turned teeth (Fig. 1H) (arboreal) (north Qld)

Myrmatopa (part) (only

P. lombokensis Emery)

9. Pronotal humeri produced into broad-based

short teeth with lateral margins distinctly
expanded, virtually laminate; mesonotal
and propodeal margins often elaborate,
variously incised or with laterally dilated
laminate lobes (e.g. P. schoopae Forel); body
mostly broad and stocky (Fig. IB) with
short appendages, generally with abundant
pilosity and pubescence often masking
underlying sculpturation (subterranean or
rarely lignicolous) (Australia-wide, except
Tasmania) Chariotnyrtna

— Pronotal humeri produced into spines or
acute teeth with lateral margins distinct.

170

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A review of the Australian Polyrhachis Ants

but not laminate; mesonotal and propodeal
margins simple, never elaborate; body
elongate with relatively long appendages and
only rarely with abundant pilosity and/
or pubescence (e.g. P. (Hedomyrma) clotho
Forel) 10.

10. Pronotal humeri produced into long,

horizontal, anteriorly directed spines;
propodeal spines reduced to short, upturned
teeth; petiole scale-like, armed with a pair of
acute, upward directed spines and a tooth
or blunt angle situated below their bases
(Fig. 1G) (lignicolous or terrestrial) (Cape
York Pen.) Myrma

— Pronotal humeri produced into horizontal

or variously elevated, mostly anterolaterally
directed spines or acute teeth; propodeal
spines present or rarely reduced to acute
teeth (P. (Myrmatopa) alphea Fr. Smith);
petiole columnar with a pair of horizontal
or elevated spines 11.

11. Small species (HL < 1.75); propodeal spines
reduced to acute teeth; petiole armed
with two, rather long, upward directed
spines (arboreal) (Cape York Pen.)

. . . Myrmatopa (part) (only P. alphea Fr. Smith)

— Larger species (HL > 1.75); propodeal spines

always present; petiole with more-or-less flat
dorsum, armed with a pair of horizontal or
variously elevated spines 12.

12. Pronotal dorsum flat; humeri armed with

acute, somewhat elevated, slender spines;
propodeal spines acute, usually longer than
pronotal pair, variously elevated; dorsum
of petiole with rather vague, posteriorly
sloping platform; dorsa of head, mesosoma
and petiole with characteristic vermiculate
sculpturation and short, bristle-like, brown
hairs (Figs IF, 6A-F, 7A-F) (subterranean)
(Qld) Hirtomyrma

— Pronotal dorsum more-or-less longitudinally
and transversely convex; humeri armed
with mostly horizontal, dorsally flattened,
laterally or anterolaterally directed acute
spines or teeth (Fig. IE); propodeal spines

mostly horizontal, but also elevated; dorsum
of petiole with rather conspicuous platform
(except in P. clio Forel); body sculpturation
never vermiculate, ranging from rugose to
rather smooth; silvery or rich golden pilosity
and pubescence in various densities over
most body surfaces (lignicolous or rarely
lithocolous) (tropical and coastal temperate
Australia, except central and southern
regions) Hedomyrma

Subgenus Myrmhopla Forel, 1915

Myrmhopla Forel, 1915: 107. Type species: Formica armata Le
Guillou, 1842 by original designation.

KEY TO AUSTRALIAN SPECIES OF THE
SUBGENUS MYRMHOPLA
(based on worker caste)

1. Larger species (HL >2.00); all body surfaces,
including appendages, with relatively long,
erect hairs; appressed or suberect pubescence
present in various densities but usually not
completely hiding underlying sculpturation;
gastral pubescence organised in midline
pattern (Fig. 5C, E) (sexspinosa-group) 2.

— Smaller species (HL <2.00); body pilosity and
pubescence highly variable between species-
groups, from rather abundant to virtually
lacking; gastral pubescence, if present, not
forming midline pattern (Figs 2C, E, 3C, E)

2. Smaller species (HL <2.70); head and anterior
portion of mesosoma black with propodeum
and petiole mostly reddish-brown; vertex of
head coarsely rugose; pronotal dorsum rather
smooth and polished or weakly and shallowly
sculptured; short, sparse pubescence mostly
greyish or white (Fig. 4B, E-F) glabriwta Clark.

— Larger species (HL >3.00); head and body
mostly black; dorsum of head and pronotum
reticulate-punctate or rugose beneath rather
long, golden or silvery pubescence .... 3.

3. Head strongly tapered behind eyes; occipital
margin narrow, forming lateral lobes that are
more-or-less visible with head in full face

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

171

Kohout

view (Fig. 5A, B); antennal scapes relatively
long (SI >190) 4.

— Head not strongly tapered behind eyes;

occipital margin broad, without distinct
lateral lobes (Fig. 4A); antennal scapes
relatively short (SI <160) dispar sp. nov.

4. Propodeal spines generally vertical to main
axis of body or even inclined anteriorly;
dorsum of head and mesosoma rather
coarsely rugose (Fig. 5B, E-F) . . sexspiiwsa
(Latreille)

— Propodeal spines oblique to main axis of

body, directed posteriorly; dorsum of head
and mesosoma shallowly and finely sculp-
tured beneath dense pubescence (Fig. 5A,
C-D) reclinata Emery

5. Body distinctly bicoloured; head and meso-
soma black with gaster orange or light
reddish-brown; appendages black, brown
or orange (Fig. 2A-F) (bicolor-group) ... .6.

— Body unicoloured, black, with appendages

black or reddish-brown (Fig. 3A-F) 7.

6. Mandibles, apical antennal segments and
gaster orange or light reddish-brown; anten-
nal scapes and legs mostly black or very
dark reddish-brown; mesosoma in lateral
view with evenly convex outline (Fig. 2B, E-F)
nigripes Emery

— Mandibles, antennae, legs and gaster orange

or light reddish-brown; mesosoma in lateral
view with mesonotum virtually flat (Fig.
2A, C-D) bicolor Fr. Smith

7. Pronotal spines relatively long and slender;
body black, closely and uniformly reticulate-
punctate, opaque (Fig. 3A, C-D) (dives- group)
dives Fr. Smith

— Pronotal spines reduced to minute teeth;
body jet-black with legs and apical antennal
segments black to medium reddish-brown;
sculpturation consisting very fine, super-
ficial reticulation with scattered shallow
pits (Fig. 3B, E-F) (mucronata-group)
mucronata Fr. Smith

POLYRHACHIS BICOLOR SPECIES-GROUP

The Polyrhachis bicolor species-group was
established by Dorow (1995) who subdivided
the former P. dmes-group (as delimited by
Emery, 1925) and transferred many of its
original constituents into three, earlier established
groups (P. armata-, sexspiiwsa- and viehmeyeri-
groups), or into five groups he newly proposed (P.
arachne-, bicolor-, cephalotes-, hector- and mucronata-
groups). As presently defined, the bicolor- group
includes only four species. However, about
11 infraspecific forms are currently associated
with the name-bearing species, P. bicolor.
Many of these forms apparently represent valid
species and, in addition, at least twice as many
closely related new species are in collections
awaiting description. This relatively small, but
widespread and complicated group is in great
need of revision but, with only two species
relevant to the Australian fauna, this is beyond
the scope of the present paper.

Diagnosis, (modified from Dorow, 1995)
Worker: Mostly small to medium-sized ants (HL
1 .15-1 .85) with general characteristics of the genus.
Mandibles smooth or very finely, longitudinally
striate, rather polished with small piliferous pits
towards bases. Anterior clypeal margin with
shallow, central, medially emarginate flange,
laterally flanked by teeth or acute angles. Head
semicircular in side view, oval in frontal view;
genae immarginate. Eyes moderately to strongly
convex, clearly exceeding lateral cephalic outline
in full face view. Mesosoma totally immarginate,
armed with rather slender spines. Petiole nodi
form with a pair of lateral spines usually em-
bracing first gastral segment, without intercalary
spines or teeth. Antennal scapes and tibiae slender
and long, spider-like. Sculpturation of head,
mesosoma and petiole mostly a fine punctation,
usually obscured by rich pubescence, producing
a matt appearance. Gaster shagreened or finely
reticulate-punctate, opaque. All body surfaces
with abundant, relatively long, erect hairs and
silvery to golden, appressed or suberect pub-

172

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A review of the Australian Polyrhachis Ants

FIG. 1. Australian Polyrhachis subgenera, dorsal view of mesosoma and petiole in outline (pilosity omitted):
A - P. (Campomyrma) creusa Emery; B - P. (Chariomyrma) schoopae Forel; C - P. (Cyrtomyrma) australis Mayr;
D - P. (Hagtomyrma) penclope Forel; E - P. (Hedomyrma) cupreata Emery; F - P. (Hirtomyrma) loweryi Kohout;
G - P. (Myrma) foreli kohout; H - P. (Mynmtopa) lombokensis Emery; f- P. (Myrmhopla) dives Fr. Smith; J - P.
(Mymiotlirinax) queenslandica Emery; K - P. (Polyrhachis) bellicosa Fr. Smith (not to scale).

escence. Body bicoloured, mostly black with
gaster and appendages light reddish-brown or
amber-coloured (as in P. bicolor), or virtually
unicoloured with body black and gaster, including
appendages, black or very dark reddish-brown
(as in P. longipes Fr. Smith, 1859).

Queen. Apart from sexual characters, very similar
to worker. Armament of pronotum, propodeum
and petiole distinctly reduced with spines
shorter and stronger. Sculpturation, pilosity and
colour virtually identical to worker.

Distribution and biology. Polyrhachis bicolor-
group species are distributed throughout

south-east Asia, including Myanmar, India,
Thailand, Malaysia, Singapore, Vietnam and the
Philippines, extending south to Indonesia, New
Guinea and northern Australia. Members of
the bicolor - group are arboreal nesters, building
polydomous nests of silk and vegetation debris
among the leaves of mostly rainforest trees and
shrubs (Robson & Kohout, 2007).

Polyrhachis bicolor Fr. Smith, 1958
(Figs 2A, C-D, 9 A, 10D)

Polyrhachis bicolor Fr. Smith, 1858: 65. Holotype queen. Type
locality: BURMA (= MYANMAR), BMNH (examined).

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

173

Kohout

Polyrhachis bicolor var. concolor Forel, 1910: 129. Syntype
workers, queen, male. Tvpe locality: PHILIPPINES, Luzon,
Manila (C.S. Banks), BSMP, MHNG QM (examined).
Synonymy by Kohout, 1998: 515.

Other material. INDIA, Bengal Baigachi, vi-
viii.1943 (L.H. Weatherill) (w); S. Andamans, Port Blair,
4.xii.l906 (B. Osmaston) (w, 2, 3 )- S.E. THAILAND,
Songkhla Prov., TonNga Chang Wildlife Sanctuary,
06°56'N, 100"14'E, 12.i.2002, lowland rf. (Surachai
Tongierm) (w). SINGAPORE, Mandai, 01°27'N,
103°46'E, <5 m, 21. xi. 1988 (P.S. Ward, #9581-4)
(w). VIETNAM, Saigon, 8.U.1925 (F. Silvestri) (w).
PHILIPPINES, Los Banos (F.X. Williams) (w); Luzon,
Manila, 20. iv. 1918 (McGregor) (w); Camarines
Sur, Panicuason Villiage, 18 km E of Naga City,
13°40'N, 123°19'E, 500-550m, 12.iii.2003 (D. General
& G.D. Alpert) (w); Palawan, Honda Bay, ii.1988
(J.H. Martin) (w); Negros Or., Dumaguete, 1942
(J.W. Chapman) (w); Mindanao Or., Gingoog,
Anakan Lbr. Co. (A. Reyes) (w). WEST MALAYSIA,
Kulala Lumpur, i-ii.1989 (M. Edmunds) (w). EAST
MALAYSIA, SABAH (as British Nth Borneo), W
Coast Residency, Ranau, 500 m, 22-25. i. 1959 (T.C.
Maa) (w, 2 ); Kinabalu Park, 19. v. 1995 (Shanmuga
Sundram) (w); Maliau Basin, Ginseng Camp,
04°44'N, 116°55'E, 700 m, 27.ii-ll.iii.2005 (RJK &
Effazilla Waty acc. 05.27) (w); ditto, Agathis Camp,
04°41'N, 116 6 54'E, c. 500 m, 16-19.iii.2005 (RJK &
Lina Thomas acc. 05.70) (w). SARAWAK, Kuching
(J. Hewitt) (w); Nanga Pelagus, nr Kapit, 180-585m,
7-14.viii.1958 (T.C. Maa) (w). BRUNEI, Brunei-
Muara Distr., Tanjung Semesta, Brunei R., 5.vii.l994
(RJK et al. acc. 94.83) (w); Belait Distr., 1-2 km SE of
Melilas Longhouse, 16.vii.1994 (RJK acc. 94.124) (w).
INDONESIA, JAVA, Batavia (= Jakarta), iii.1921 (no
further data) (w); Buitenzorg (= Bogor), 21 .xii.1912
(V. Karavaiev #2382) (w); ditto, 4.ix.l909 (Bryant
& Palmer) (w, 2 ); SUMATRA, Pematang, Siantar,
1937 (W:M. Mann, NGS SI) (w, ,)• PAPUA NEW
GUINEA, West Sepik Prov., Torricelli Mts, Lumi,
400-550 m, 03°28'S, 142°02'E, 4-13.viii.1984 (RJK acc.
84.283) (w). Central Prov., Thaira Boat Harbour, c. 15
km ESE of Port Moresby, 09°3TS, 147H 7'E, 5.ix.l984,
mangroves (RJK acc. 84.436) (w, , ). AUSTRALIA,
NORTHERN TERRITORY, Holmes Jungle, c. 15
km NE of Darwin, 12°25'S, 130°58'E, 16.xi.1993,
monsoon rf. clearing (RJK acc.93.35) (w, , ); Darwin,
Nightcliff, 2.ix.l960 (J.L. Gressitt) (w); Berry Springs
NP, 12°42'S, 130°59'E, 10.ii.1994 (RJK acc. 94.4) (w);
ditto, 21.vii.1981 (BBL) (w).

Worker. Dimensions: TL c. 6.00-7.06; HL 1.43-
1.68; HW 1.12-1.31; Cl 77-79; SL 1.96-2.34; SI 172-
182; PW 0.87-1.03; MTL 2.34-2.68 (10 measured)

Mandibles with 5 teeth, progressively reducing
in length towards base. Anterior clypeal margin
with shallow median flange, laterally flanked
by acute angles. Clypeus with poorly defined,
posteriorly weakly elevated, median carina;
clypeus virtually straight in profile with rather
shallow basal margin. Frontal carinae sinuate
with well raised margins; frontal furrow indistinct.
Sides of head in front of eyes converging
anteriorly tow'ards mandibular bases in virtually
straight line; behind eyes sides rounding into
convex occipital margin. Eyes strongly convex, in
full face view clearly breaking lateral cephalic
outline. Ocelli indistinct. Mesosoma laterally
immarginate. Pronotal dorsum weakly convex
in profile; humeri armed with slender, relatively
long, acute, anterolaterally directed spines with
tips slightly turned upwards. Promesonotal
suture distinct; mesonotum straight in profile
with metanotal groove marked by slight step in
outline. Propodeal dorsum rather short with a
pair of slender, obliquely elevated, subparallel,
acute spines. Petiole nodiform with medially
weakly elevated dorsum and pair of relatively
long and slender, laterally and posteriorly curved,
acute spines. Anterior face of first gastral tergite
rounding in evenly convex line onto dorsum of
segment.

Mandibles smooth and polished with shallow
piliferous pits. Head, mesosoma and gaster
closely reticulate-punctate with sculpture almost
completely hidden by overlying pubescence.
Spines weakly rugose at bases, smooth and
polished towards tips. Gaster finelly shagreened.

Mandibular masticatory borders with curved,
golden hairs. Anterior clypeal margin medially
with several, medium length, anteriorly directed,
golden setae. Head, including clypeus, mesosoma
and gaster with numerous, mostly erect and
variously curved, somewhat untidy, long silvery
hairs, some longer than greatest diameter of eyes;
hairs on gaster mostly posteriorly directed. Very
distinct, relatively long, silvery pubescence almost
completely hiding underlying sculpturation
on head, mesosoma and petiole, except spines.

174

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A review of the Australian Polyrhachis Ants

FIG. 2. Polyrhachis (Myrmlwpla) species from Australia. Head in full face view (top); Dorsal view (left); Lateral
view (right). A,C-D, P. bicolor Fr. Smith; B,E-F, P. nigripes Emery (not to scale).

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

175

Kohout

Gastral pubescence more appressed and some-
what diluted, not obscuring fine sculpturation.

Head, mesosoma and petiole black; mandibles,
median portion of anterior clypeal margin,
antennae, legs, coxae, tips of spines, subpetiolar
process and gaster, orange or light reddish-
brown.

Queen. Dimensions: TL c. 8.97; HL 1.87; HW
1.40; Cl 75; SL 2.62; SI 187; PW 1.78; MTL 3.06
(1 measured).

Very similar to worker and apart from sexual
characters, including three ocelli, fully developed
mesosoma and wings, differing as follows:
Pronotal spines reduced to more-or-less
triangular, acute teeth, barely longer than basal
widths. Mesoscutum with anterior margin widely
and evenly rounded in dorsal view; median
line bifurcate anteriorly; parapsides rather flat
anteriorly, raised posteriorly. Mesoscutellum
weakly convex, slightly elevated above
dorsal plane of mesosoma; metanotal groove
distinct. Propodeal spines very short, obliquely
elevated, somewhat dorsoventrally flattened,
tips rounded. Petiolar spines similar to those
in worker but stronger at base and distinctly
shorter; dorsum of petiole with rather distinct,
blunt intercalary tooth.

Males and immature stages (eggs, larvae and
pupae) deposited in the QM spirit collection.

Remarks. Polyrhachis bicolor is a widespread
species recorded from south-east Asia south
to Indonesia, New Guinea and the Northern
Territory in Australia. As noted by Kohout
(2008: 295), across its distribution, P. bi color forms
a large number of overlapping populations that
differ to some extent from the holotype. About
eleven infraspecific forms are presently associated
with P. bicolor, however, after examination of a
large amount of material from across the entire
range I consider the Australian population to
represent the nominal form.

Polyrhachis bicolor is a relatively common species
at suitable localities around Darwin, where it

builds nests of silk and vegetation debris among
the leaves of trees and shrubs notably along the
margins of monsoon rainforests. In spite of being
a rather common species in mangroves and
lowland forests along the Gulf of Papua, it has
not yet been recorded from Cape York Peninsula
or elsewhere in north Queensland.

Polyrhachis niqripes Emery, 1897 stat. nov.

(Figs 2B, E-F, 9A)

Polyrhachis bicolor var. nigripcs Emery, 1897: 592. Holotype

worker. Type locality: NEW GUINEA: Paumomu River

(= Angabanga R.) (L. Loria), MSNG. Stat. nov.

Other material. INDONESIA, IRIAN JAYA,
Eramboe, 80 km ex Merauke, l.ii.1960 (T.C. Maa)
(w); Hollandia area, Cyclops Mts, W. Sentani, 2 00-
1000m, 22-25. vi. 1959 (T.C. Maa) (w). PAPUA NEW
GUINEA, West Sepik Prov., Torricelli Mts, Lumi,

x. 1984 (D. Waisi) (w); Wum, Upper Jimmi Valley,
840m, 16.vii.1955 (J.L. Gressitt) (w); Morobe Prov.,
Huon Pen., lower Busu R., 27.iv-12.v.l955, lowland rf.
(E.O. Wilson) (w - compared with holotype); Central
Prov., Karema, Brown R., 8-1 1 .iii.1955, lowland rf. (E.O.
Wilson #601) (w); Bisianumu, E of Port Moresby, 500
m, 3.ix.l959 (T.C. Maa) (w); Catalina Estate, 48 km N of
Port Moresby, 500 m, 3.iv.l959 (T.C. Maa) (w); Laloki, nr
Port Moresby, 30.vii-2.ix.1959 (T.C. Maa) (w); Rouna,

xi. 1968 (N.L.H. Krauss) (w); Aroa Pltn, 16.V.1957 (J.H
Barrett) (w). AUSTRALIA, QUEENSLAND, Cape York
Pen., Iron Ra. NP, Claudie R., 12°46'S, 143°16'E, <20m,
2.X.2000, monsoon rf. (R]K acc. 2000.136, 137) (w, 9 ).

Worker. Dimensions: TL c. 5.74-6.25; HL 1.40-
1.59; HW 1.15-1.31; Cl 79-82; SL 1.87-2.06; SI 157-
169; PW 0.87-1.03; MTL 2.18-2.46 (4 measured).

Mandibles with 5 teeth. Anterior clypeal margin
with shallow, median flange, laterally flanked
by rather acute teeth. Clypeus with poorly
defined, posteriorly elevated median carina,
weakly convex in profile with rather shallow
basal margin. Frontal carinae sinuate with well
raised margins; frontal furrow indistinct. Sides
of head in front of eyes converging anteriorly
towards mandibular bases in straight line;
behind eyes sides rounding into convex occipital
margin. Eyes strongly convex, relatively
large, in full face view clearly breaking lateral
occipital margin. Ocelli indistinct. Mesosoma
laterally immarginate. Pronotal and mesonotal
dorsa forming even, weakly convex line in profile;

176

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A review of the Australian Polyrhachis Ants

pronotal humeri armed with fine, slender, rela-
tively long, acute, anterolaterally directed spines,
tips slightly turned downwards. Promesonotal
suture distinct; metanotal groove feebly marked
by rather flat, bowed line. Propodeal dorsum
rather short with a pair of obliquely elevated,
fine, slender, subparallel, acute spines. Petiole
nodiform, dorsum medially weakly elevated,
armed with a pair of relatively long, slender,
laterally and posteriorly curved, acute spines.
Anterior face of first gastral segment widely
rounding onto dorsum of segment.

Mandibles rather smooth or very finely,
longitudinally striate with shallow piliferous pits.
Head, mesosoma and gaster closely reticulate
punctate with sculpture almost completely hidden
by overlying pubescence. Spines weakly rugose,
tips rather smooth and polished. Gaster finely
shagreened.

Mandibular masticatory borders with a few
curved, golden hairs. Anterior clypeal margin
medially with several, medium length, anteriorly
directed, golden setae. Head, including
clypeus, mesosoma and gaster with numerous,
mostly erect and variously curved, somewhat
untidy, long silvery hairs, some longer
than greatest diameter of eyes; hairs on gaster
somewhat shorter and mostly posteriorly
directed. Very distinct, relatively long, silvery
pubescence almost completely hiding underlying
sculpturation on head, mesosoma and petiole,
except apical portions of spines. Gastral pub-
escence more appressed and diluted, not
obscuring fine sculpturation.

Head, including anterior clypeal margin,
antennal scapes, mesosoma, including fore
coxae, spines, and petiole, including subpetiolar
process, black; legs, including mid and hind
coxae and basal antennal segments dark to very
dark reddish-brown or black; mandibles, apical
antennal segments and gaster orange or light
reddish-brown. Dorsum of first gastral tergite
with darker, irregular, somewhat diffused,
reddish-brown patch medially.

Queen. Dimensions. TL c. 8.52; HL 1.84; HW
1.50; Cl 81; SL 2.34; SI 156; PW 1.75; MTL 3.81
(1 measured)

Essentially like worker and apart from sexual
characters, including three ocelli and complete
thoracic structure, differing as follows: pronotal
spines distinctly reduced in length, about twice
as long as their basal widths. Mesoscutum
with anterior margin evenly rounded in dorsal
view; median line bifurcate anteriorly and
dorsally; parapsides weakly raised posteriorly.
Mesoscutellum slightly elevated above dorsal
plane of mesosoma; metanotal groove distinct.
Propodeal spines short, obliquely elevated,
their inner margins continued inwards but not
meeting medially. Petiolar spines similar to those
in worker but stronger and distinctly shorter;
dorsum of petiole bluntly raised medially.
Sculpturation, pilosity and colour virtually
identical to worker.

Males unknown. Immature stages (eggs and
larvae) deposited in the QM spirit collection.

Remarks. Polyrhachis bicolor and P. nigripes
are certainly very closely related. The main
differences separating them are their colour
patterns which appear to be constant in all
material examined. Another key difference is
the profile of the mesosoma which features a
virtually flat mesonotum in P. bicolor (Fig. 2D),
while in P. nigripes the pronotal and mesonotal
dorsa form a weak, but evenly convex line
(Fig. 2F). In addition, P. bicolor has distinctly
more slender spines, notably those on the
propodeum, which, although they are closer
together, are clearly separated at their bases
and are parallel for their full length. In contrast
the spines in P. nigripes are distinctly stronger,
with the inner bases of the propodeal spines
somewhat continuous across the propodeal
dorsum, forming a medially incomplete 'U' shape
in dorsal view. The propodeal spines are also
slightly divergent and weakly curved along their
length. New Guinean and Australian specimens
of P. nigripes are very similar in appearance.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

177

Kohout

The distribution of P. nigripes in New Guinea
appears to be concentrated mainly along the
coast of the Gulf of Papua, with only a few
doubtful records from Irian Jaya and the north
of the island. In Australia, P. nigripes has been
recorded only once from Iron Range National
Park on Cape York Peninsula where a nest of
silk and vegetation debris was collected on
the ground in monsoon rainforest along the
Claudie River. The twig and leaves upon which
the nest was built were completely dry and
had apparently fallen very recently from its
original position higher on the tree. It contained
a dealate queen, 19 workers and brood (many
eggs and larvae).

POLYRHACHIS DIVES SPECIES-GROUP

The Polyrhachis dives species-group was
originally delimited by Emery (1925) and has
previously contained as many as 77 species
and subspecies. Dorow (1995) redefined the
group and transferred a number of species into
the earlier established P. armata-, sexspinosa- and
viehtneyeri-groups (all Emery, 1925), or into his
newly proposed P. arachne-, bicolor-, cephaloles-,
hector- and mucronata-groups. The P. dives-group,
as presently defined, includes about 14 species
and subspecies with only one Australian species,
P. dives.

Diagnosis, (modified from Dorow, 1995)
Worker. Mostly medium-sized ants (HL 1.40-2.00),
some species exhibiting slight polymorphism.
Mandibles rather densely longitudinally striate
or rugose with numerous piliferous pits. Anterior
clypeal margin with central, medially emarginate
flange, laterally flanked by acute teeth. Head
semicircular in side view, almost circular in
frontal view. Genae immarginate or with a short
carina running about half way from occipital
margin towards mandibular bases (as in some
extralimital species, e.g. P. lacteipennis Fr. Smith,
1858). Eyes rather flat or only moderately convex,
in full face view not or only marginally exceeding
lateral cephalic outline. Mesosoma totally

immarginate. Pronotum armed with rather short
or only moderately long spines (except in P.
dives belli Forel, 1912, where pronotal spines are
slender and relatively long); propodeal spines
slender and elevated with their tips curved
outwards. Petiole with lateral spines, that in
most species conform to shape of gaster, and
a pair of distinct intercalary teeth. Body rather
distinctly, more-or-less regularly, reticulate-
punctate (as in P. dives), moderately rugose
(as in P. lacteipennis) or coarsely foveolate (as
in P. menelas Forel, 1904). Gaster shagreened or
closely punctate. Body with only a few, short,
erect hairs; closely appressed, mostly silvery or
pale golden pubescence rather sparse over head
and body (as in P. dives) or virtually lacking
(as in P. lacteipennis ). Gaster with somewhat
longer, silvery or golden pubescence, that is
virtually lacking in several extralimital species.
Body and appendages mostly black with gaster
black or very dark reddish-brown.

Queen. Queen in several species (e.g. P. dives)
distinctly larger than worker with usual
characters identifying full sexuality, including
three ocelli, complete thoracic structure and
wings. Spines distinctly shorter with pronotal
spines reduced to acute angles. Propodeal spines
modified into blunt, horizontal, posteriorly
directed and somewhat dorso-ventrally
compressed stubs; petiolar spines very short,
only weakly curved, almost straight. Body
sculpturation, pilosity and colour identical to
that in worker.

Distribution and biology. The Polyrhachis dives
species-group is the most widespread species-
group within Myrnihopla. It stretches from
Guam Island in the Pacific, throughout east
and south-east Asia (e.g. China, Japan, Taiwan,
Philippines, Malaysia, Thailand, Myanmar,
Nepal, India, Sri Lanka), the Middle East and
Arabian Peninsula (e.g. Pakistan, Iran, Iraq,
Israel, Saudi Arabia, Oman, Yemen) and reaches
as far west as Morocco in northern Africa. From
southern Asia it extends southwards to Indonesia,
New Guinea and northern Australia. Members

178

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A review of the Australian Polyrhachis Ants

of this group are mostly arboreal, building
nests upon the leaves and branches of trees and
shrubs, preferably in open habitats, such as
grassy woodlands, open forests and swamps
(Fig IOC). However, some extralimital species
(e.g. P. lacteipennis Fr. Smith) were observed
to be lignicolous or subterranean nesters.
The incorporation of silk and occasionally carton
occurs in all arboreal nests. These nests can be
either mono- or polydomous. Single and multiple
queen colonies have been documented in P. dives
(see Robson & Kohout, 2007).

Polyrhachis dives Fr. Smith, 1857
(Figs II, 3 A, C-D, 8A, 10C,)

Polyrhachis dives Fr. Smith, 1857: 64. Holotvpe worker.
Type locality: SINGAPORE (A.R. Wallace), BMNH

(examined).

Polyrhachis (Mymthopla) exulans Clark, 1941: 91, pi. 13, fig.
24. (w.) Type locality: AUSTRALIA, Northern Territory,
Koolpinyan (C.L. Barrett). Synonymy by Kohout, 1988:
433.

(For full synonymic citations see Bolton, 1995: 347)

Other material. CHINA, Canton (C.W. Howard)
(w); Mokanshan (N. Gist Gee) (w); Amoy (S. Ling)
(w); Triet Giang Prov., xii.2001 (Bui Tuan Viet #3)
(w). HONG KONG (J. Fellowes #1) (w) FORMOSA,
Abato (Silvestri) (w); Kusukusu, 12.iv.1932 (L. Gressitt)
(w); Rokki, 17. vi. 1932 (L. Gressitt); T'ai Chung,
18. iv. 1962 (A.C.F. Hung) (w). GUAM I., Andersen
Air Force Base, iv.& xii.2003 (L. Hansen) (w, $, J).
MYANMAR (as BURMA), (no further data) (w); Carin
Cheba, 900-1 10m, v-xii.1888 (L. Fea); Bhamo, vii.1885
(Fea) (w). ASSAM, Cherrapoonii (Smythies) (w);
Misamari, iv.1944 (A.C. Cole) (w). THAILAND,
Chiang Mai Prov., Doi Ithanon NP, 16.viii.1992
(D.G. Furth) (w); Khao Yai N.P., 800-1000 m,
19.viii.1992 (D.G. Furth) (w); Payao, 12.ix.1951 (D. &
E. Thurman); Petchburi, Kowyoi Nongchumphon
Nonpoe, 28.vi.1952 (R.E. Elbel). VIETNAM, Thua
Thien Hue Prov., Phong Dien Distr., iv.2001 (Bui
Tuan Viet #7, 8) (w); Lang Son Prov., Cao Loc
Distr., i.2001 (Bui Tuan Viet #4) (w). SINGAPORE,
no further data) (Bryant & Palmer) (w); ditto (no
urther data (Baker) (w). PHILIPPINES, Luzon,
Manila, 19.ii.1918 (Me Gregor) (w); Luzon, Luneta
Hill, Baguio, 1450 nt, l.v.1981 (W.L. Brown) (w,
j); Negros Or., Dumaguete, 1923 (].W. Chapman)
(w); ditto. Horns of Negros, 3600', 1942-43 (J.W.
Chapman) (w); Negros, Victorias, 20.xii.1927 (W.D.
Pierce) (w); Baguio, (J.W. Chapman) (w); Mindanao,
Misamis, Gingoog, Anakan Lor. Co. (A. Reyes) (w);
Bukidnon, Musuan Maramag, 3. i. 1985 (C.K. Starr

& Pinto) (w). WEST MALAYSIA, Selangor, Kulala
Lumpur (Army Scrub Typhus Unit); Selangor,
UKM campus, 30.viii.1992, D.G. Furth) (w). EAST
MALAYSIA, SARAWAK, Nanga Pelagus, nr Kapit,
180-585 m, 7-14. viii. 1958 (T.C. Maa). SABAH, Tawau
Distr., Kalabakan, 8-15. xi. 1958 (T.C. Maa); Tawau,
Quoin Hill, Cocoa Res. Stn, 4.xii.l962 (Y. Hirashima);
13km N of Marak Parak, 12.V.1985 (C.K. Starr) (w,
.). BRUNEI, Belait Distr., 1-2 km SE of Melilas
Longhouse, 16.vii.1994 (RJK accs 93.10, 94.127) (w, ?);
Bandar Seri Begawan, Gadong University grounds,
11. iv. 1993 (RJK acc.93.2) (9); Tutong Distr., Bukit
Sulang nr Lamunin, 20-23.vii.1994 (RJK acc. 94.129)
(w, ?[. INDONESIA, KALIMANTAN BARAT, 3
km N of Putussibau, 00°51'N, 112°55'E, 19-20.
vi.1996 (C. Reid) (w, 9 ); KALIMANTAN TENGAH,
Banjarmasin, 'Suaka Insan' Hospital, 6.vi.l983 (M.
Bordsen). SUMATRA, Pematang, Siantar, 1937
(W.M. Mann, NGS SI Exp.); Brastagi, 1937 (Mann,
NGS SI Exp.) (w); Medan, 21.x. 1993 (Y.v. Nierop)

S ; Langkat, ii.1925 (N. Dengas) (w). JAVA, Kebun
a, Bogor Botan. Gdns, OB^S, 106°47'E, 290 m,
7-12.ix.1999 (S.K.A. Robson #816) (w); ditto, 31.1993
(Yayuk R. Suhardiono) (9); Semarang, 13.X.1927 (L.G.
Kafshoven) (w); Buitenzorg (= Bogor) (J. Barbour)
(w). SULAYVESI, Lore-Lindu NP, nr Dongi-Dongi
shelter, 975-1020 m, 01°15'S, 120°20'E, 4-<Lxii.l985
(Mai. traps) (C.v. Achterberg) (w); Latimodjong Mts
Clagg) (w); Malino, viii.1937 (C.T.& B.B. Brues)
(w). SERAM, Piroe, 1937 (W.M. Mann, NGS SI Exp.)
(w, 9 , ,T); ditto, above Haruru, nr Masohi, 50-150 m,
18.iii.1981 (W.L. Brown) (w, ?, cj). SULU, Mangole, 2
km W Mandafuhi Camp, c. 70 m, 7-24. iii. 1995 (Mai.
traps) (C.v. Achterberg & Y. Yasir) (w). FLORES,
Nangagete, c. 60 km E Maumere, 19. vii.T 972 (W.L.
Brown) (w). BALI, 1937 (Springer) (w). AMBON,
Amboina DEI, 1938 (F.I. Buitenbos) (w); ditto, 1937
(W.M. Mann, NGS SI Exp.) (w, $, <J). IRIAN JAY A,
Wamena, 1700 m, 10-25. ii. 1960 (T.C. Maa) (w);
Hollandia, 13.iii.1960 (T.C. Maa) (w); Waris, S of
Hollandia, 450-500 m, 16-23.viii.1959 (T.C. Maa) (w).
PAPUA NEW GUINEA, West Sepik Prov., Torricelli
Mts, Lumi, 03°28'S, 142°02'E, 400-500 m, 4-13.viii.1984
(RJK acc. 84.249) (w); ditto, x.1984 (D. Waisi) (w, );
nr Aitape, 03°09'S, 142°21'E, 3.viii.l984 (RJK acc.
84.218) (w, V); nr Yapsiei, ii. 1984 (S. van Dyck) (w);
Subitana, Sogeri, 26. v. 1957 (J.H. Barrett) (w); New
Britain Prov., Linga Linga Pltn, W of Willaumez
Pen., 9.iv.l956 (J.L. Gressitt) (w); ditto, Sio N Coast,
600 m, 24.vii.1956 (E.J. Ford Jn.) (w); Morobe Prov.,
Huon Pen., Mongi-Mape Watersheds, Sattelberg
vie., 660 m, 4.iv.l955 (E.O. Wilson #722) (w); ditto,
Mongi Watershed, Gemeheng, 1200-1300 m, 11-13.
iv.1955 (E.O. Wilson #773) (w); Mt Missim (Stevens)
(w); Wau, Mendi, 1600-1700 m, xi.1971 (N.L.H.
Krauss) (w); Lae, 19.ix.1949 (N.L.H. Krauss) (w); Port

Memoirs of the Queensland Museum 1 Nature • 2010 • 55(1)

179

Kohout

Moresby, 7.vi.l955 (E.O. Wilson #520) (w); 7km S of
Maprik, 03.42'S, 143.03'E, 120 m, ll.ii.1989 (P.S. Ward
#10175) (w); West New Britain Prov., Dami Res.
Stn, 12 km SW of Hoskins, 27.viii.1984 (E.J. Brough)
(w); New Ireland Prov., Lelet Plateau, 03°20'S,
151°56'E, 800-1000 m, 19-24.vii.1984 (RJK acc. 84.95,
99) (w). AUSTRALIA, NORTHERN TERRITORY,
Koolpinyah, 1933 (C.L. Barrett) (w); Holmes Jungle,
15 km NE of Darwin, 12°25'S, 130°58'E, 16.xi.1993
(RJK acc. 93.33) (w). QUEENSLAND, Yarrabah, c.
11 km E of Cairns, 16°56'S, 145°52'E, 22-24.vii.1980
(RJK accs 80.113, 130, 141) (w, 2 , o); South Mission
Bch, 2 km NbyW, 17°55'S, 146°05'E, 13.vi.1980
(RJK accs 80.14, 15, 17, 18) (w, y) ; Tully, ii.1958
(Darlingtons) (w); 2 km NbyW of Sth Mission Beach,
17°55'S, 146°05'E, 13.vi.1980 (RJK acc. 80.14); Halifax,
8.vi.l919 (F.X. Williams) (w).

Worker. Dimensions. TL c. 5.39-7.71; HL 1.40-1.93;
HW 1.18-1.65; Cl 84-91; SL 1.56-1.96; SI 119-132; PW
0.87-1.15; MTL 1.87-2.56 (34 measured).

Mandibles with 5 teeth. Anterior clypeal margin
with emarginate, shallow, median flange, laterally
flanked by acute teeth. Clypeus virtually straight
in profile, posteriorly rounding into weakly
impressed basal margin. Frontal triangle distict.
Frontal carinae sinuate with moderately raised
margins. Sides of head in front of eyes weakly
convex towards mandibular bases; behind eyes
sides widely rounding into rather flat occipital
margin. Eyes only moderately convex, in full
face view not or only marginally exceeding
lateral cephalic outline. Ocelli lacking, relative
positions marked by shallow pits in cephalic
structure. Mesosoma immarginate. Pronotum
rather weakly convex in profile; humeri armed
with straight, horizontal, anterolaterallv directed,
acute spines. Promesonotal suture distinct;
mesonotum convex in profile. Metanotal groove
poorly indicated; propodeal dorsum armed with
slender, obliquely elevated, divergent spines,
tips curved upwards and outwards. Petiole
with dorsolaterally projecting acute spines that
conform to shape of anterior gaster; dorsum
medially with pair of distinct intercalary teeth.
Anterior face of first gastral tergite higher than
full height of petiole, widely rounding onto
dorsum of segment.

Mandibles densely and closely longitudinally
striate with piliferous pits. Head, mesosoma
and petiole distinctly, more-or-less regularly,
reticulate-punctate. Sides of mesosoma somewhat
more deeply sculptured with numerous pits
on meso- and metapleurae. Spines sculptured at
bases, smooth and polished towards tips. Gaster
finely shagreened.

Mandibles at masticatory borders with
numerous, curved, golden hairs and short,
appressed hairs towards bases. Anterior clypeal
margin with several anteriorly projecting
setae medially and a few shorter setae fringing
margin laterally. Clypeus with a few, paired,
medium length, erect hairs. Apical segments on
dorsum and venter of gaster with a number of
erect, relatively long, mostly posteriorly directed,
golden hairs. Closely appressed, medium length,
pale golden pubescence in various densities
over most of head and body. Pubescence
more silvery on sides of head, mesosoma and
petiole. Gastral pubescence somewhat longer,
rich golden and more abundant dorsally
where it almost completely hides underlying
sculpturation; pubescence paler and less dense
on gastral venter.

Black, with only condylae and tip of apical
antennal segments, light yellowish-brown;
mandibular teeth dark reddish-brown.

Queen. Dimensions: TL c. 8.72-9.32; HL 2,03-2.18;
HW 1.78-1.90; Cl 85-91; SL 1 .93-2.09; SI 108-112; PW
1.81-1.90; MTL 2.62-2.74 (14 measured).

Very similar to worker, apart from sexual
characters, shorter spines and the following:
anterior clypeal margin with median flange more
deeply emarginate; clypeus with basal margin
virtually flat in lateral view. Eyes more convex,
always exceeding lateral cephalic outline. Pronotal
spines reduced to minute teeth. Mesoscutum in
lateral view relatively high, widely rounding
onto flat dorsum with anterior margin evenly
rounded in dorsal view; median line bifurcate
anteriorly and posteriorly; parapsides flat.
Mesoscutellum flat, not elevated above dorsal

180

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A review of the Australian Polyrhachis Ants

plane of mesosoma. Metanotal groove distinctly
impressed. Propodeal spines very short, directed
posteriorly, weakly elevated dorsally. Petiolar
spines short, projecting laterally, very weakly
curved. Sculpturation, pilosity, pubescence and
colour identical to those in worker.

Males and immature stages (eggs, larvae and
pupae) deposited in QM spirit collection.

Remarks. Polyrhachis dives is a very widespread
species ranging from south-east Asia south to
northern Australia, with recent reports of its
occurrence as far east as Guam Island in the
Pacific (L. Hansen pers. comm.). Throughout its
distribution, P. dives is a morphologically very
stable species with only a few, rather insignificant
differences between individuals, even those from
widely separated localities. However, south-east
Asian specimens generally have a more deeply
emarginate anterior clypeal margin and eyes that
only rarely exceed the lateral cephalic outline.
Most also have the tips of the propodeal spines
more distinctly curved outwards and longer
petiolar spines that are somewhat curved
downwards from their midlength. In contrast,
the anterior clypeal margin in Australian and
New Guinean specimens is only shallowly
emarginate and the eyes clearly exceed the lateral
cephalic outline. The propodeal spines are only
weakly curved outwards and the petiolar spines
somewhat shorter and less curved.

In Australia P. dives is known from two
isolated populations, one in the Northern
Territory and the other in northern Queensland.
It prefers mostly open savannah woodlands and
swampy coastal plains, where it builds its silk
and/or carton-based, relatively large, usually
polydomous nests between the branches and
leaves of small trees and shrubs (Fig. IOC).

POLYRHACHIS MUCRONATA
SPECIES-GROUP

The Polyrhachis mucronata species-group of the
subgenus Myrmhopla was delimited by Dorow

(1995) who subdivided the earlier established
P. d/ues-group (Emery, 1925). Dorow listed
36 species and subspecies within the group,
however, the status of several subspecific forms
still remains unresolved. Two new species were
recently described from Sulawesi (Kohout, 2008)
and one species is considered a junior synonym
(see below). Only a single species of the group, P.
mucronata, is relevant to the Australian fauna.

Diagnosis, (modified from Dorow, 1995) Worker:
Small to medium-sized ants (HL 1.25-2.10)
with general characteristics of the genus.
Mandibles mostly longitudinally striate or finely
rugose with numerous piliferous pits. Anterior
clypeal margin with shallow, median flange (as
in P. mucronata), or shallowly truncate (as in P.
retrorsa Emery, 1900). Head usually semicircular in
side view, oval in frontal view; genae immargin-
ate. Eyes moderately to strongly convex, clearly
exceeding lateral cephalic outline in full face view.
Mesosoma totally immarginate, usually highly
convex and relatively short (as in P. mucronata),
but also somewhat elongated and distinctly less
convex (as in P. tristis Mayr, 1867). Pronotum
armed with acute teeth (as in P. mucronata), or
rarely with long slender spines (as in P. amana
Fr. Smith, 1861), or simply rounded (as in P.
moeschi Forel, 1912). Propodeal spines relatively
long and strong in most species, however,
also short (as in P. orpheus Forel, 1911). Petiole
columnar with a pair of lateral spines usually
embracing first gastral segment; spines mostly
slender, but also remarkably massive (as in P.
lucidula Emery, 1893 and P. ridleyi Forel, 1912).
Dorsum of petiole with a pair of more-or-less
distinct intercalary teeth, except in some species
(e.g. P. amana and P. orpheus). Sculpturation
of head, mesosoma and petiole ranging from
rather smooth and highly polished (as in P.
emmae Santschi, 1920) to closely punctate (as
in P. oedacantha Wheeler, 1919). Gaster usually
more finely sculptured, shagreened and polish-
ed, only rarely closely punctate, opaque (as
in P. tristis). Body pilosity and pubescence
virtually lacking in most species, however, in

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

181

Kohout

P. mitrata Menozzi, 1932 and P. retrorsa whole
body covered with rather diluted, whitish pub-
escence. Body mostly black, rarely with purple
metallic reflections (as in P. oedacantha and P.
phalerata Menozzi, 1926). Gaster black or reddish-
brown with appendages ranging from orange
or light reddish-brown to black.

Queen. Queen very similar to worker with
usual differences indicating caste, including
three ocelli, complete thoracic structure and
wings. Body armature, notably propodeal and
petiolar spines distinctly shorter and stronger.
Sculpturation, pilosity and colour essentially as
in worker.

Distribution and biology. The Polyrhachis
mucrcmata species-group is distributed throughout
east and south-east Asia (China, Philippines, Laos,
Malaysia, Thailand, Myanmar, India, Sri Lanka),
extending south to Indonesia (Sumatra, Java,
Sulawesi) and New Guinea (including Bismarck
Archipelago), reaching the southern limit of its
distribution in northern Queensland. The known
members of this group are arboreal nesters,
building nests of silk and vegetation debris
upon the leaves of rainforest trees and shrubs,
mostly in the lower arboreal zone (Fig. 10E).

Polyrhachis mucronata Fr. Smith, 1859
(Figs 3B, E-F, 9B, 10E)

Polyrhachis mucronatus Fr. Smith, 1859: 140. Holotype worker.
Type locality: INDONESIA: Aru Is. (A.R. Wallace), OXUM
(examined).

PolMachis (Myrmhopla) cyrtomyrmoides Donisthorjpe, 1947:
195. Syntype worker, queen.'Type locality: INDONESIA,
IRIAN JAYA (as Dutch New Guinea), Marlin Bay, viii.1944
(E.S. Ross), CASC, BMNH (examined). Syn. n'ov.

Other material. INDONESIA, Aru I. (no further data)
(w). IRIAN JAYA (as Dutch New Guinea), Maffin Bay,
17.vi.1944 (E.S. Ross) (w); Hollandia area, W. Sentani,
Cyclop Mts, 150-1000 m, 16-19.vi.1959 (T.C. Maa) (w);
Waris, S of Hollandia, 450-500 m, 1 6-23.viii.1959 (T.C.
Maa) (w). PAPUA NEW GUINEA, West Sepik Prov.,
Torricelli Mts, Lumi, 400-550 m, 03°28'S, 142°02'E,
4-13. viii.l 984 (RJK acc. 84.228, 249) (w); ditto, x.1984
(D. Waisi) (w); Morobe Prov., nr Wampit, c. 35km W
of Lae, 06°45'S, 146°40'E, c. 50 m, 24& 27. viii.l 984 (RJK
acc. 84.365) (w); Naru, Gogol R., 20 km SW of Madang,
05°21'S, 145°41'E, 22.viii.1984 (RJK acc. 84.336) (w);

Central Prov., Eilogo Rd., 4 km ESE of Sogeri,
09°25'S, 147°27'E, c. 500 m, 4.ix.l984 (RJK acc. 84.427)
(w); Varirata NP, 550-760 m, 5-9.ii.1981 (W.L. Brown)
(w). AUSTRALIA, QUEENSLAND, Cape York Pen.,
Iron Range, 12°43'S, 143°18'E, 26-31.vii.1981 (RJK accs
81.146, 147); ditto, l-17.viii.1978 (S.van Dvck); Mt
Hedlev, 1-2 km N of Home Rule, 15°45'S, l45°17'E,
200-3(50vm, 11. vi. 1996 (RJK & CJB acc. 96.44) (w);
Cedar Bay NP, 15°48'S, 145°19'E, 16.vi.1997 (SKR
#589) (w); Pilgrim Sands, c. 1 km NW of Cape
Tribulation, 16°04'S, 145°28'E, 12-15.vi.1996 (RJK at al.
acc. 96.47) (w); Cape Tribulation NP, 16°04'S, 145°27'E,
6.xii.l985 (RJK acc. 85.5) (w, -); Canopy Crane site.
Cape Tribulation, 16°06'S, 145°27'E, 20-21. ii.2000 (RJK
accs 2000.17, 21) (w, ?); Oliver Ck, c. 8 km SW of Cape
Tribulation, 16°08'S, 146°26'E, 14.vi.1998, lowland
rf. (RJK acc. 98.53 (w); Me Lean Ck, 19 km Sby W of
Cape Tribulation, 16°15'S, 145°26'E, 15. vi. 1996 (RJK
acc. 96.54) (w); Bellenden Ker, Cableway Base Stn,
17°16'S, 145°54'E, 17-24.X.1981 (GBM & Earth watch
Exp.) (w); Russel R., Bellenden Ker Landing, 17°16'S,
145°56'E, l-9.xi.1981 (GBM & Earthwatch Exp.) (w);
Josephine Falls, 17°26'S, 145°51'E, 12.ii.1996 (GBM)
(w); Garradunga, Seymour Ra„ c. 7 km N of Innisfail,
17°28'S, 146°01 V E, <l'00 m, 5-6.vi.1996 (RJK at al. 96.31)
(w, 1 ); Mission Bch, 17°45'S, 146°00'E, 10 m, 20. i. 1996
(SKR #101) (w); Dunk I., 17°57'S, 146°09'E, viii.1927
(H. Hacker) (2); Hinchinbrook I., Gayundah Ck,
18°21'S, 146°14'E, 100-500 m, 8-18.xi.1984 (GBM) (w).

Worker. Dimensions: (holotvpe cited first) TL c. 5.55,
5.34-6.50; HL 1.43, 1.40-1.65; HW 1.17, 1.12-1.31; Cl 82,
76-82; SL 1.84, 1.65-2.12; SI 157, 146-164; PW 1.00, 0.94-
1.15; MTL 1.93, 1.81-2.31 (31 measured).

Anterior clypeal margin medially with shallow
flange, flanked laterally by acute angles. Clypeus
with poorly defined median carina; clypeus in
profile straight anteriorly, posteriorly rounding
into moderately impressed basal margin.
Frontal carinae sinuate with moderately raised
margins. Sides of head in front of eyes weakly
convex converging towards mandibular bases;
behind eyes sides rounding into convex occipital
margin. Eyes relatively large and convex, in
full face view clearly exceeding lateral cephalic
outline. Ocelli lacking. Mesosoma immarginate.
Pronotal humeri armed with short, acute teeth.
Promesonotal suture distinct; mesonotum vir-
tually flat in profile. Mesopleural process a blunt
lobe; metanotal groove indicated as a thin, bowed
line. Propodeal dorsum very short, armed with
rather strong, obliquely directed, diverging.

182

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A review of the Australian Polyrhachis Ants

acute spines; inner borders of spines continued
medially for a short distance, but failing to meet,
propodeal dorsum descending into relatively
high, oblique declivity in uninterrupted line.
Petiole columnar, dorsum armed with two
strong lateral spines that conform to shape of
anterior gaster and two short, acute intercalary
teeth medially. Anterior face of first gastral
tergite straight, widely rounding onto dorsum
of segment.

Mandibles very finely, longitudinally striate
with numerous piliferous pits. Head and dorsum
of mesosoma and petiole finely reticulate-
punctate with sculpturation distinctly coarser
on sides of mesosoma; meso- and metapleurae
weakly rugose and somewhat wrinkled. Gaster
finely shagreened.

Mandibles at masticatory and outer borders
with a few, curved, golden hairs. Anterior
clypeal margin with several anteriorly directed
setae medially and a pair of longer, suberect hairs
arising from just behind margin. Tuft of medium
length, erect hairs in front of subpetiolar process.
Apical segments on dorsum and venter of gaster
with a number of semierect, posteriorly directed,
golden hairs. Very short, closely appressed pub-
escence arising from minute punctures and pits,
distributed over most body surfaces, more distinct
on gastral dorsum.

Body black, polished; mandibular bases,
antennal scapes, fore coxae and tarsi black or very
dark reddish-brown; mandibular masticatory
borders, condylae, apical antennal segments
and most of legs, except tarsi, light to medium
reddish-brown.

Queen . Dimensions: TL c. 6.95-7.81; HL 1.65-1.75;
HW 1.25-1.34; Cl 75-77; SL 1.96-2.12; SI 153-160;
PW 1.56-1.68; MTL 2.18-2.37 (10 measured).

Very similar to worker with usur* differences
indicating caste, including three ocelli, complete
thoracic structure and wings. Median clypeal
carina somewhat more distinct. Pronotal teeth
dorsoventrally flattened; mesoscutum in lateral

view with anterior margin widely rounding
onto flat dorsum; median line bifurcate ant-
eriorly; parapsides rather flat, weakly raised
posteriorly; mesoscutellum convex, elevated
above dorsal plane of mesosoma; metanotal
groove strongly impressed. Propodeal spines
shorter than in worker, almost straight; petiolar
spines similar to those in worker but shorter.
Sculpturation, pilosity, pubescence and colour
virtually identical to worker.

Males and immature stages (eggs, larvae and
pupae) deposited in QM spirit collection.

Remarks. Among the Australian species of
Myrmhopla, Polyrlmchis mucronata is very distinct,
however, it superficially resembles some species
of the subgenus Cyrtomyrma Forel. Despite their
general similarity and identical nesting habits,
P. mucronata can be easily distinguished from
Cyrtomyrma species by the form of its petiole.
The petiolar node in P. mucronata is columnar
and armed with two, more-or-less horizontal,
posteriorly directed spines that conform to
the shape of the anterior gaster. The petiole
also bears a pair of short, intercalary teeth. In
contrast, in Cyrtomyrma species, the petiole is
scale-like and armed with four teeth or spines
of various configurations. Additionally, all
Australian Cyrtomyrma species have a distinct
posterolateral carina separating the gena from
the ventral parts of the head. This carina is
lacking in P. mucronata-group species. The two
groups also differ in their immature stages with
the pupae of Myrmhopla species being enclosed
in cocoons, while in Cyrtomyrma species the
pupae are naked (Kohout, 2006b).

Polyrhachis mucronata ranges from the islands
of eastern Indonesia and New Guinea south
to northern Queensland where it reaches the
southern limit of its distribution. It is a relatively
common species in the monsoon rainforests of
Cape York Peninsula and the lowland rainforests
of the Wet Tropics. Colonies of P. mucronata build
nests the leaves of rainforest trees and shrubs.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1) 183

Kohout

using mostly larval silk combined with leaf
fragments and other vegetation debris (Fig. 10E).

When describing Polyrhachis cyrtomyrmoides,
Donisthorpe (1947: 195) emphasised its close
resemblance to ants of tine subgenus Cyrtomyrma
but failed to note its remarkable similarity to P.
mucronata. I had the opportunity to examine the
syntypes of P. cyrtomyrmoides, supplemented
by additional specimens in the CASC collection
(Maffin Bay, 17.vi.1944, E.S. Ross). Their com-
parison with the holotype of P. mucronata
(OXUM) and abundant material of that species
from New Guinea and Australia (ANIC, BMNH,
BPBM, MCZC and QM) revealed that the two
species were very similar. As noted earlier
by Kohout (2000: 206), the Australian and
New Guinean populations of P. mucronata
differ somewhat from the holotype, with some
specimens from Australia being remarkably close
to those of P. cyrtomyrmoides, sharing the rather
highly polished appearance which is typical of
the latter species. However, the density of the
generally fine reticulate-punctate sculpturation
varies not only between different populations,
but also to some extent between specimens of a
single colony. The colour of the legs also varies,
generally being a lighter red in Australian
specimens and a darker, reddish-brown in
most of New Guinean specimens. In addition,
the eyes vary in the degree of their convexity
with specimens from the Wet Tropics having
the eyes simply convex, similar to those in P.
cyrtomyrmoides, while the eyes in specimens
from the mid and northern Cape York Peninsula
are virtually protuberant. When considered
individually, specimens of some populations
appear quite distinct, however, when all the
available material from the whole distribution
range is compared, no consistent differences
between P. cyrtomyrmoides and P. mucronata are
evident. I therefore believe that the separate
specific status of the former is not justified and
consider them conspecific.

POLYRHACHIS SEXSPINOSA
SPECIES-GROUP

The Polyrhachis sexspinosa species-group
of the subgenus Myrmhopla Forel, 1915 was
established by Emery (1925), who included 12
species and subspecies from New Guinea and
south-east Asia. Bolton (1975) revised the world
fauna of the group and recognised 12 valid
species with all of the included infraspecific
taxa considered synonyms. Three new species
from the Philippines were later added, one
subspecies (P. sexspinosa reclinata Emery, 1887)
raised to specific status (Kohout, 1987) and one
species (P. barnardi Clark, 1928) synonymised
(Kohout & Taylor (1990), raising the number
of valid species of the group to 15. Dorow
(1995) recognised 1 7 species as constituents of the
group, including P. melpomene Emery, 1897 and
P. olybria Forel, 1912. However, these two species
were later transferred to different subgenera (P.
melpomene and its junior synonym P. dolichocepluila
Viehmeyer, 1914 to subgenus Hedomyrma and P.
olybria to the nominal subgenus Polyrhachis), and
two former subspecies (P. arcuspinosa waigeuensis
Donisthorpe, 1943 and P. sexspinosa esuriens
Emery, 1897) were raised to specific status by
Kohout (1998). At present the P. sexspinosa species-
group comprises 19 valid species, including one
(P. spinosa Mayr, 1867) recently elevated to its
original specific status (Kohout, 2008) and one
described below as new (P. dispar). Four species
of the group are relevant to the Australian fauna
with two (P. dispar and P. glabrinota Clark,
1930) considered endemic. They appear to be
derived from common ancestral stock, such as
the New Guinea-based species P. aureovestita
Donisthorpe, 1937 and P. bubastes Fr. Smith,
1863 (Bolton, 1975) and speciated after loosing
connection with the maternal New Guinean
populations following the sinking of the con-
tinental bridge between New Guinea and
Australia. On the other hand, the more robust
populations of the relatively common P.
sexspinosa (Latreille, 1802) and P. reclinata
Emery, 1887 maintained most of their original

184

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A review of the Australian Polyrhachis Ants

characteristics, with specimens from Cape York
Peninsula indistinguishable from their New
Guinean counterparts.

Diagnosis. Worker: Relatively large ants (HL
>2.0), except for the rather small, extralimital P.
nofra Bolton, 1975 (HL 1.52-1.53), with general
characteristics of the genus. Mandibles smooth
and polished apart from small piliferous pits
towards bases. Anterior clypeal margin arcuate,
often obtusely truncate medially or with shallow
median emargination. Head contracted post-
eriorly, distinctly narrower behind than in front of
eyes. Eyes with short, erect hairs; strongly convex
or protuberant, clearly exceeding lateral cephalic
outline in full face view. Occipital margin with
more-or-less developed lateral angular prom-
inences which are usually visible with the head in
full face view (except in P. dispar and P. glabrinota).
Mesosoma immarginate; pronotum strongly
convex, humeri armed with rather strong, mostly
forward curved spines. Promesonotal suture
strongly impressed. Mesonotum flat or weakly
convex in profile; mesopleural process present as a
simple lobe (as in P. dispar and P. sexspinosa) or a
dentiform structure, that can be acute or obtuse
(as in P. glabrinota and P. reclinata); metanotal
groove usually replaced by a minutely raised
ridge. Propodeum armed with a pair of spines
that are either vertical or inclined forwards
(as in P. sexspinosa), or posteriorly reclined
and often curved (as in P. dispar, P. glabrinota
and P. reclinata). Petiole nodiform with a pair
of lateral spines, without intercalary spines or
teeth. Gaster, when contracted, broadly ovate.
Sculpturation of head, mesosoma and petiole
ranging from weak to rather heavy, gaster
usually finely shagreened. All body surfaces,
including appendages, with relatively long, erect
hairs. Pubescence mostly appressed or suberect,
somewhat radiating, present in various densities
but usually not completely hiding underlying
sculpturation. Pubescence on gastral dorsum
organised in a characteristic midline pattern in
virtually all species (except P. nofra) or almost
completely absent (P. exotica Kohout, 1987).

Body mostly black, or partly reddish-brown
(as in P. glabrinota), with appendages usually
lighter, reddish-brown.

Queen. Very similar to worker with usual chara-
cters of full sexuality, including three ocelli,
complete thoracic structure and wings. Armament
of pronotum, propodeum and petiole somewhat
reduced with spines shorter, less curved and
usually more stubby. Sculpturation, pilosity,
pubescence and colour virtually as in worker.

Distribution and biology. Polyrhachis sexspinosa-
group species are distributed throughout the
Indo-Australian region, with only a few species
known from the Oriental region and one from the
Solomon Islands, hi Australia, the group is limited
to the northern part of Cape York Peninsula in
Queensland, north of the 14°S parallel. Members
of the sexspinosa-group are mostly arboreal
nesters, building nests of silk and vegetation
debris between the foliage of rainforest trees and
shrubs. However, some species deviate from
this behaviour and have their own characteristic
methods of nesting. Colonies of P. sexspinosa
invariably build pocket-like nests of silk,
vegetation debris and bark fragments against the
trunks of rainforest trees (Fig. 10A-B). In contrast,
colonies of P. dispar, P. glabrinota and P. reclinata
at Iron Range National Park were frequently
found nesting within the hollow internodes of
a bamboo, Bambusa forbesii. Only one colony
of P. reclinata was located under the bark of
a living tree, while most nests of P. glabrinota
were constructed between leaves, usually in
the lower arboreal zone, about 2-3 m above the
ground.

Polyrhachis dispar sp. nov.

(Figs 4A, C-D, 8A)

Polyrhachis (Myrmhopla) barnardi Clark, 1928: 39, pi. 1, figs
37-38 (in part); Kohout & Taylor, 1990: 519.

Polyrhachis barnardi Clark; Bolton, 1975: 6 (in part),

Material. HOLOTYPE: QUEENSLAND, Cape York
Pen., Iron Ra„ 12°43'S, 143°18'E, 26-31.vii.1981, rf„
ex nest in dry bamboo internode, R.J. Kohout acc.
81.202 (worker). PARATYPES: data (and nest) as

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

185

Kohout

FIG. 3. Polyrlmchis (Myntihopla) species from Australia. Head in full face view (topi; Dorsal view (left); Lateral
view (right). A,C-D, P. dives Fr. Smith; B,E-F, P. mucronata Fr. Smith (not to scale).

186

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A review of the Australian Polyrhachis Ants

for holotype (64 workers, alate queen); data as for
holotype, except RJK accs 81.148, 195 (8 paratype
workers); Cape York, W.B. Barnard (2 workers
from original series of P. banmrdi Clark in ANIC &
MCZC). Type deposition: Holotype (QM T152321),
most paratype workers and paratype queen (from
holotype nest) in QM; 4 paratype Workers (2 from
holotype nest) each in ANIC, BMNH and MCZC; 2
paratype workers each in AMNH, AMSA, CASC,
MHNG, MVMA, NMNH and NMHU.

Other material. QUEENSLAND, Cape York Pen.,
Lockerbie Scrub, 10°46'S, 142°29'E, 19-24.iii.1987,
rf. (RJK acc. 87.61) (w, $); ditto, 25.i-12.ii.1984 (J.H.
Sedldfek) (Y); Iron Ra., 21.vi.1948 (D.P. Vernon) (w);
ditto, viii.1949 (N.L.H. Krauss) (w).

Worker. Dimensions: (holotype cited first) TL c. 13.10,
11.39-14.10; HL 3.12, 2.87-5.40; HW 2.12, 2.02-2.44;
Cl 68, 68-74; SL 3.43, 3.17-3.72; SI 162, 152-165; PW
1.75, 1.51-1.96; MTL 4.33, 4.08-4.69 (25 measured).

Anterior clypeal margin arcuate, with shallow
emargination medially; clypeus with distinct,
blunt, median carina; clypeus virtually straight
in profile, posteriorly rounding into moderately
impressed basal margin. Frontal triangle distinct.
Frontal carinae sinuate with acute, almost ver-
tically raised margins; central area relatively
narrow with distinct posterior carina. Sides of
head in front of eyes weakly convex; behind
eyes sides converging into relatively wide
occipital margin. Eyes strongly convex, in full
face view exceeding lateral cephalic outline.
Median ocellus poorly developed, lateral ocelli
lacking; positions indicated by distinctly raised
cephalic sculpture. Pronotal dorsum weakly
convex in profile; humeri armed with strong,
acute, more-or-less horizontal, anterolaterally
curved spines; promesonotal suture distinct.
Mesonotum rather flat in profile; mesopleural
process a blunt, simple lobe. Propodeal spines
relatively long, obliquely elevated from their
bases, curved posteriorly and moderately
divergent. Dorsum of petiole with a pair of
laterally and posteriorly directed, acute spines,
with their tips turned slightly downwards.

Clypeus finely reticulate-punctate; sides of
head in front of eyes smooth and polished;
sculpturation on head increasing in density and

intensity posteriorly, vertex and sides towards
occipital margin coarsely sculptured. Dorsum
and sides of mesosoma rugose; sculptural
intensity decreasing posteriorly with propodeal
dorsum and declivity relatively smooth. Gaster
very finely shagreened.

Mostly whitish or pale yellow hairs present
over all body and appendages, including antennal
scapes and spines, with longest hairs on pronotal
and mesonotal dorsa exceeding greatest diameter
of eye in length. Rather diluted, mostly greyish,
appressed or suberect pubescence variously
developed, most dense and more appressed
on clypeus, dorsal mesosoma and petiole,
somewhat suberect and variously radiating
on sides of mesosoma. Sides of head and vertex
with distinctly less dense, suberect pubescence
partly revealing underlying sculpturation. Gaster
with relatively long, mostly golden appressed
pubescence, ranging from somewhat diluted
on venter to dense on dorsum, where it is
arranged in a characteristic midline pattern.

Body generally black; mesosoma and petiole
in some specimens very dark reddish-brown;
masticatory borders of mandibles, apical antennal
segments and legs, except tarsi, medium reddish-
brown.

Queen. Dimensions: (paratvpe queen cited first)
TL c. 14.08, 13.18-14.21; HL 3.17, 3.07-3.38; HW 2.17,
2,17-2.47; Cl 68, 68-73; SL 3.43, 3.28-3.73; SI 158,
148-159; PW 2.37, 2.42-2.57; MTL 4.33, 4.23-4.74 (5
measured).

Very similar to worker with usual differences
indicating full sexuality. Mesoscutum virtually
as wide as long with lateral margins converging
anteriorly into narrowly rounded margin;
median line slightly raised; dorsum rather low
and flat in lateral view with parapsides only
weakly raised posteriorly. Mesoscutellum flat,
not elevated above dorsal plane of mesosoma;
metanotal groove distinct. Spines similar to
those in worker, but shorter with propodeal pair
distinctly less elevated. Sculpturation, pilosity,
pubescence and colour virtually identical to
worker.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

187

Kohout

Male unknown. Immature stages (larvae of
various stages of development and pupae) in
QM spirit collection.

Remarks. With its relatively broad occipital
margin, Polyrhachis dispar is easily recognised
within the sexsphnosa-group. It is most similar
to P. reclinata and, in addition to the lack of
lateral occipital lobes, it can be distinguished by
its less convex pronotal dorsum and distinctly
rugose dorsum of the head and mesosoma.
The pronotal dorsum in P. reclinata is distinctly
higher and the sculpturation of the vertex,
sides of head and pronotal dorsum consists of
very fine reticulation beneath a rich pubescence
that is much more diluted in P. dispar. Also, the
mesopleural process in P. dispar is present as a
blunt, simple lobe, while in P. reclinata the lobe
is ventrally dentiform.

Kohout & Taylor (1990: 519) examined all
the available specimens of the P. barnardi
Clark, 1928 type series and concluded that it
comprised two distinct species. The lectotype
(originally furnished with a red tag inscribed
'Type') and two paralectotypes (all MVMA)
were identified as P. sexspinosa (Latreille) and,
consequently, P. barnardi has been considered
a synonym of that species. Two remaining
paralectotypes (that match Clark's original
description and illustrations of P. barnardi)
were deposited in other institutions (ANIC,
MCZC) and are included in the type series of
P. -dispar.

Polyrhachis glabrinota Clark, 1930
(Figs 4B, E-F, 8B)

Polyrhachis (Myrmhopla)glabrinotum Clark, 1930: 13, fig. 1 nos
11, 11a. Syntype workers. Type locality: QUEENSLAND,
Cape York (W.B. Barnard), ANIC, BMNH, MCZC, MVMA,
(examined).

Polyrhachis glabrinota Clark; Bolton, 1975: 8.

Other material. QUEENSLAND, Cape York Pen.,
Lockerbie Scrub, 10°46'S, 142 l ’29'E, 19-24.iii.1987,
rf. (RJK acc. 87.38, 60, 61) (w, ?); Barnaga, 10°53'S,
14223'E, 18-24.iii.1987, rf. (RJK acc 87.11) (w); Jardine
R., 11°08'S, 142°35'E, 14.X.1979 (M.S.& B.J. Moulds) (w);
Hann Ck, W of Moreton Telegraph Stn, 28.vi.1988 (G.

Kenning) (w); 14 km WSW of Captain Billy Landing
11 U 41'S, 142°42'E, 9.ix.l993, rf. (S.O. Shattuck #3802-
16) (w); Iron Ra., West Claudie R„ 3-10.xii.1985, rf 50
m (JBM & DJC) (w); Iron Ra., 12'43'S, 143 l, 18'E, 1-17.
viii.1978 (S.van Dyck) (w); ditto, 26-31. vii. 1981 rf
(RJK accs 81.133, 134, 167, 180) (w, .); ditto, Gordon
Ck x-ing, 6.X.2000 (RJK acc. 00.174) (w); 9 km ENE of
Mt Tozer, Iron Ra. NP, 12"43'S, 143 l T7'E, 10.vii.1986
(D.C.F. Rentz) (w); 11 km ENE of Mt Tozer, 11-16.
vii. 1986 (T.Weir & A. Calder) (w).

Worker. Dimensions: (syntypes cited first) TL
c. 10.18-11.00, 9.78-11.95; HL 2.50-2.64, 2.46-
2.96; HW 1.62-1.69, 1.53-1.93; Cl 63-66, 59-66-
SL 3.12-3.32, 2.93-3.35; SI 190-197, 181-197; PW
1.44-1.48, 1.40-1.78; MTL 3.73-3.88, 3.58-4.18 (22
measured).

Anterior clypeal margin arcuate, entire. Clypeus
with blunt median carina, straight or weakly
convex in profile, posteriorly rounding into
well impressed basal margin. Frontal triangle
distinct. Frontal carinae sinuate with only
moderately raised margins; central area with
median longitudinal carina. Sides of head in front
of eyes weakly convex; behind eyes sides distinct-
ly tapering into relatively narrow occipital margin
with lateral lobes weakly developed, not visible
with head in full face view. Eyes protuberant, in
full face view exceeding lateral cephalic outline.
Ocelli lacking. Pronotal dorsum convex in
profile; summit with more-or-less distinct,
shallow, longitudinal depression; humeri armed
with strong, acute, more-or-less horizontal, ante-
rolaterally curved spines; promesonotal
suture distinct. Mesonotum virtually flat in
profile; mesopleural process a well-developed
lobe. Propodeal spines relatively long, obli-
quely elevated from their bases, curved
and moderately divergent. Dorsum of petiole
with a pair of laterally and weakly posteriorly
directed, acute spines, with their tips turned
slightly downwards.

Clypeus and sides of head in front of eyes
finely reticulate-punctate, rather smooth and
polished; density and intensity of sculpturation
increasing posteriorly with central area, vertex
and sides of head towards occipital margin

188

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A review of the Australian Polyrhachis Ants

deeply and coarsely sculptured. Pronotal
dorsum rather smooth and polished or with
shallow weak rugosity. Mesonotum, sides of
mesosoma and petiole rugose with propodeal
dorsum relatively smooth. Gaster very finely
shagreened.

Numerous, mostly white or pale yellow,
relatively short hairs, not exceeding greatest
diameter of eye in length, present over entire
body and appendages, including antennal
scapes and spines. Rather diluted, white or grey,
appressed or suberect pubescence developed
in various densities over all body surfaces, more
dense and closely appressed on clypeus, sides
of head and fore coxae; somewhat suberect
and variously radiating pubescence on sides of
mesosoma. Gaster with mostly white, appressed
pubescence arranged in characteristic midline
pattern.

Generally bicoloured; head black, mesosoma
black or dark reddish-brown with propodeum
and petiole distinctly lighter. Colour variable;
black in most recently collected specimens
but distinctly lighter reddish-brown in some
syntypes. Mandibular masticatory borders, tips
of apical antennal segments, and legs, including
coxae, light to medium reddish-brown.

Queen. Dimensions: TL c. 11.34-13.10; HL
2.65-2.96; HW 1.62-1.81; Cl 59-64; SL 3.17-3.53;
SI 183-199; PW 1.90-2.21; MTL 3.93-4.33 (9
measured).

Apart from sexual characters, very similar to
worker. Mesoscutum virtually as wide as long
with lateral margins converging into narrowly
rounded anterior margin; median line slightly
raised; dorsum flat in profile with parapsides
only weakly raised posteriorly. Mesoscutellum
flat, marginally elevated above dorsal plane of
mesosoma, posteriorly rounding into metanotal
groove. Spines similar to worker but shorter
with propodeal pair horizontal. Sculpturation,
pilosity, pubescence and colour identical to
worker.

Males and immature stages (eggs, larvae in
various stages of development and pupae) in
QM spirit collection.

Remarks. With its smaller size, rather smooth
and polished pronotal dorsum and short, sparse
pubescence, P. glabrinota is easily distinguished
from all other Australian members of the
group. However, it closely resembles P.
bubastes Fr. Smith, 1863 from New Guinea,
with both species similar in size, general body
form and shape of the petiole. They differ in
the intensity of the pronotal sculpturation that,
in P. glabrinota, is only weakly sculptured,
rather smooth and polished, while it is very
coarsely sculptured in P. bubastes. The body in
P. glabrinota is also generally bicoloured with
the parts of mesosoma and petiole more-or-less
reddish-brown, while the body in P. bubastes is
mostly uniformly black.

Recent examination of a considerable amount of
sexsvinosa-group material from eastern Indonesia
ana New Guinea has revealed a number of
specimens closely comparable to P. glabrinota.
Like that species, they have smooth pronotal
dorsums and are bicoloured, but more distinctly
than most of the recently collected Australian
specimens. The reddish-brown colour of the
specimens from Australia is somewhat darker
and more-or-less restricted to the propodeum,
petiole and appendages, while the specimens
from the Aru Islands and New Guinean
mainland are distinctly bicoloured with the
head black and all of the mesosoma, petiole and
appendages light to medium reddish-brown.
Tnese specimens also bear a rather prominent,
median, longitudinal depression on the pro-
notal dorsum, which is only shallowly
indicated in some of the Australian specimens.
Specimens from Wammar, Aru Is (V.
Karavaiev #2595, IZAS) were misidentified
as P. rugifrons Fr. Smith (Karavaiev, 1927: 25),
while a series from Wanuma, Madang District
(N.L.H. Kraus, BPBM) were furnished with an
identification tag of unknown origin inscribed
'cf. bubastes new'. In addition, two nest series
of closely comparable specimens, complete with

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

189

Kohout

FIG. 4. Polyrhachis (Myrtuhopla) species from Australia. Head in full face view (top); Dorsal view (left); Lateral
view (right). A,C-D, P. dispar sp. nov.; B,E-F, P. glabritwta Clark (not to scale).

190

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A review of the Australian Polyrhachis Ants

FIG. 5. Polyrhachis (Myrmhopla) species from Australia. Head in full face view (top); Dorsal view (left); Lateral
view (right). A,C-D, P. rechnata Emery; B,E-F, P. sexspinosa (Latreille) (not to scale).

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

191

Kohout

sexuals, were more recently collected at Lumi,
Torricelli Mts (RJK accs 84.271, 283). Specimens
of all these populations vary in several characters,
including the direction and length of the propo-
deal spines and the relative height of the petiole,
but overall the specimens are very similar.
However, when directly compared with P.
glabrinota, these specimens appear rather distinct
and in spite of some similarity I am hesitant to
consider them conspecific. Like Bolton (1975), I
consider P. glabrinota to be an Australian endemic
and believe that the Indonesian and New Guinean
material belong to a closely allied, but separate
biological species.

Polyrhachis glabrinota is a predominantly arbore-
al species that builds nests between leaves in the
lower arboreal zone. However, several colonies
were located nesting in bamboo internodes. In
contrast, the nests of the closely allied, possibly
polydomous. New Guinean species noted above,
were collected from the crown of a recently
felled rainforest tree.

Polyrhachis reclinata Emery, 1887
(Figs 5A, C-D, 8C)

Polyrhachis sexspinosa var. reclinata Emery, 1887: 236. Lectotype
worker (by present designation) and paralectotype
workers. Type locality: NEW GUINEA, Fly River, xii.75
(L.M. D' Albertis), MSNG (examined). Polyrhachis sexspinosa
(Latreille, 1802); Bolton, 1975: 12 (in part).

Polyrhachis reclinata Emery; Kohout & Taylor, 1990: 518.
Raised to species.

LECTOTYPE SELECTION

I have examined five syntypes of the original
series of P. sexspinosa reclinata, three of which
were lodged in Emery's collection and two in the
separately housed main collection of the MSNG.
The specimens bear the following labels: 'Nuova
Guinea, Fly River, xii.75, L.M. D' Albertis', 'P.
sexspinosa Latr. var. reclinata Em.' and 'Collezione
Emery'. Besides the syntypes, I had the oppor-
tunity to examine additional specimens iden-
tified as P. sexspinosa reclinata and lodged in
various other collections, however, in spite of
some being furnished with a tag inscribed 'Co-
type', their locality labels appeared to be of

doubtful origin and not identical to those on
the syntypes.

To establish the nomenclatural stability of
the species and to prevent any future selection
of a lectotype from outside of Emery's original
series, I have designated, in accordance with
Article 74.7.3 of the International Code of Zoological
Nomenclature (Fourth Edition), a syntype worker
specimen as the lectotype of Polyrhachis reclinata
Emery. The specimen is lodged in the Emery
collection (MSNG) and, in addition to the labels
listed above, it bears a red original tag inscribed
'Typus'. The remaining specimens of the original
series consequently become paralectotypes and
are labelled accordingly.

Other material. PAPUA NEW GUINEA, Morobe
Prov., Huon Pen., lower Busu R., 8-10.V.1955, lowland
rf. (E.O. Wilson #923, 984, 1003) (w, 2). Western
Prov., Oriomo Govt. Stn, 26-28.X.1960 (J.L. Gressitt)
(w); Daru I. (no further data) (w); Central Prov.,
Eigolo Rd., 4 km ESE of Sogeri, 09°25'S, 147°27'E,
4.ix.l984, c. 500 m, rf. (RJK accs 84.427, 423, 424) (w,
2 ). AUSTRALIA, QUEENSLAND, Cape York Pen.,
Lockerbie Scrub, 10°46'S, 142°29'E, 19-23.iii.1987, rf.
(RJK accs 87.21, 30, 31, 32, 33, 34, 39, 48, 64, 69, 71, 72)
(w, .); ditto, i.1958 (Darlingtons) (w); Bamaga, 10°53'S,
142°23'E, 18-23.iii.1987, rf. (RJK acc. 87.9) (w); Bamaga-
Lockerbie, 25.i-12.ii.1984 (J.H. SedUCek) (w); Iron Ra„
12’43'S, 143‘18'E, 1-17. viii.l 978 (S. Van Dyck) (w); ditto,
l-3.vii.1976 (P. Filewood) (w); ditto, 26-31 .vii. 1981, rf.
(RJK accs 81.145, 148, 163, 168, 185, 195, 1%, 201, 209,
210, 212) (w, 2 ); ditto, i.1958 (Darlingtons) (w); ditto.
West Claudie R., 3-1 0.xii. 1985, rf. (GBM & DJC) (w);
ditto, Lockhard R. Ranger Stn., 12°45'S, 143°17'E,
26.viii.2004, rf edge (G.D. Alpert) (w); 11 km ENE
of Mt Tozer, 12°43 V S, 143°18'E (T. Weir & A. Calder)
(w).

Worker. Dimensions, (lectotype cited first): TL c.
14.01,11.44-14.31; HL 3.48, 2.92-3.48; HW 2.37,1.89-2.37;
Cl 68, 62-68; SL 4.13, 3.53-4.13; SI 174, 174-195; PW
1.86, 1.56-1.86; MTL 5.09, 4.28-5.11 (23 measured).

Anterior clypeal margin arcuate, narrowly
and shallowly emarginate medially; clypeus
with blunt median carina, clypeus weakly
convex in profile, posteriorly curving into
shallow basal margin. Frontal carinae sinuate
with acute, highly raised margins; central

192

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A review of the Australian Polyrhachis Ants

area with median longitudinal carina. Sides of
head in front of eyes convex; behind eyes sides
strongly tapering into narrow occipital margin
with rather weakly developed occipital lobes.
Eyes strongly convex, in full face view clearly
exceeding lateral cephalic outline. Median ocellus
usually present, lateral ocelli lacking; positions
indicated by distinctly raised cephalic sculpture.
Pronotal dorsum weakly convex in profile with
anterior slope almost flat; humeri armed with
acute, more-or-less horizontal, anterolaterally
directed spines; promesonotal suture distinct.
Mesopleural process a simple dentiform lobe.
Mesonotal dorsum flat or weakly convex in
profile. Propodeal spines long, divergent,
obliquely elevated from their bases and more-or-
less curved posteriorly. Dorsum of petiole with a
pair of slender, widely diverging, acute spines.

Clypeus and sides of head finely reticulate-
punctate, distinctly smooth; intensity of
sculpturation increasing posteriorly with a few
shallow rugae on vertex and towards occipital
margin. Mesosoma rugose dorsally and lat-
erally, except for relatively smooth propodeal
dorsum. Gaster very finely shagreened.

Mostly yellow or pale golden hairs, of var-
iable length and density over entire body
and appendages, including antennal scapes.
Relatively long, suberect, mostly golden or brassy
pubescence variously developed, most dense
and more appressed on dorsum of head and
mesosoma where it almost completely hides
underlying sculpturation. Gastral pubescence
ranging from somewhat diluted on venter to
dense on dorsum, where it is arranged in a
characteristic midline pattern.

Body generally black; propodeum and petiole
often lighter, reddish-brown; masticatory
borders of mandibles, antennae and legs, in-
cluding coxae, medium reddish-brown with
fore coxae and tarsi distinctly darker.

Queen. Dimensions: TL c. 13.55-14.46; HL 3.12-3.28;
HW 1.96-2.07; Cl 61-64; SL 3.83-3.88; SI 185-195; PW
2.02-2.12; MTL 4.64-4.79 (4 measured).

Differing from worker in usual characters
identifying full sexuality, including three
ocelli, complete thoracic structure and wings.
Mesoscutum marginally wider than long
with narrowly rounded anterior margin in
dorsal view; median line raised; dorsum flat
in lateral view with parapsides only weakly
raised posteriorly. Mesoscutellum’ flat, not
elevated above dorsal plane of mesosoma;
metanotal groove distinct. Spines similar to
those in worker but shorter, propodeal pair
distinctly less elevated. Sculpturation, pilosity
and colour virtually identical to worker.

Males and immature stages (eggs, larvae in
various stages of development and pupae) in
QM spirit collection.

Remarks. Polyrhachis reclinata is a rather
common species closely resembling P. sexspinosa.
However, they are easily separated by the
following characters. The head in P. sexspinosa is
strongly tapered behind tire eyes and the occipital
margin forms prominent, well-developed lateral
lobes that are clearly visible in full face view.
In addition, the vertex and sides of the head
behind the eyes are rather coarsely rugose and
covered with more-or-less diluted, mostly
white or silvery pubescence. The propodeal
spines in P. sexspinosa are virtually vertical to
the main axis of the body, or are even inclined
forwards when viewed laterally. In contrast, the
head behind the eyes in P. reclinata is generally
shorter and distinctly less strongly tapered
with the lateral occipital lobes relatively weakly
developed and less prominent in full face view.
The sculpturation on the clypeus and sides of
the head is rather smooth, with a few shallow
rugae on the vertex, and is usually masked
by rather dense, golden or brassy, appressed
pubescence. The propodeal spines are oblique
to the main axis of the body and usually curved
posteriorly in side view.

The type locality of P. reclinata is the Fly River
delta, which is only about 150 km from Cape
York Peninsula on the opposite side of Torres

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

193

Kohout

FIG. 6. Polyrhachis (Hirtomyrma) species from Australia. Head in full face view (top); Dorsal view (left);
Lateral view (right). A,C-D, P. bamaga Kohout; B,E-F, P. eremita Kohout (not to scale).

194

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A review of the Australian Polyrhachis Ants

FIG. 7. Polyrhachis (Hirtomyrma) species from Australia. Head in full face view (top); Dorsal view (left)
Lateral view (right). A,C-D, P. loweryi Kohout; B,E-F, P. rustica Kohout (not to scale).

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

195

Kohout

Strait. Specimens from either side of Torres
Strait are closely comparable, differing only in
the shape and direction of spines. The pronotal
spines of the specimens from the Fly River are
more slender, longer and less strongly curved
forwards compared to Australian specimens.
The propodeal spines are almost straight
while they are shorter and more strongly
curved posteriorly in Australian specimens.
The mesopleural process is a simple, ventrally
dentiform lobe in both New Guinean and
Australian specimens.

Polyrhachis reclinata is known from lowland
rainforest along the Gulf of Papua, including the
Fly River delta (type locality), the lower Busu
River, the Port Moresby area and Daru Island.
In Australia P. reclinata has been collected

from Lockerbie Scrub south to Iron Range.
Most colonies collected at the latter locality
were nesting within the hollow internodes of
Batnbusa forbesii growing as several relatively
large stands within monsoon rainforest. Only a
single colony of this species was located nesting
under the bark of a living tree.

Polyrhachis sexspinosa (Latreille, 1802)
(Figs 5B, E-F, 8D, 10A-B,)

Formica sexspinosa Latreille, 1802: 126, pi. 4, fig. 21. Neotype
worker (by present designation). Type locality: PAPUA
NEW GUINEA, West Sepik Prov., Pes Mission, 12 km
SW of Aitape, 03°11'S, 142°15'E, <50 m, rf„ 3.viii 1984
(RJK acc. 84.207), AN1C (examined).

Polyrhachis barnardi Clark 1928; 39, pi. 1, figs 37-38 (in part).
Synonymy, and lectotype designated by Kohout & Taylor
1990: 519 (see under remarks section of P. dispar above).
Polyrhachis sexspinosa (Latreille). Fr. Smith, 1858: 59; Bolton,
1975: 12.

196

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A review of the Australian Polyrhachis Ants

NEOTYPE DESIGNATION

The holotype queen of this species should
be lodged in the MNHN, but an extensive
search by the author failed locate it there or in
any collections examined and it is considered
lost. To establish the nomenclatural stability of
the species, 1 have designated, in accordance
with Article 75.3.5. of the International Code
of Zoological Nomenclature (Fourth Edition), a
worker specimen as the neotype of Polyrhachis
sexspinosa (Latreille). The specimen was selected
from a colony of 50+ workers, a dealate queen
and several males collected from a pocket nest
on the trunk of a rainforest tree. The neotype
has been deposited in ANIC.

Other material. INDONESIA, Aru Is, Wokam, 1908
(Merton) (w); ditto, Kobror, 1925 (Karavaiev

#2534) (w). PAPUA NEW GUINEA, West Sepik
Prov., Pes Mission, c.12 km WSW of Aitape, 03°11'S,
142°15'E, <50 m, 31.vii-3.viii.1984 (RJK accs 84.148,
171, 207) (w, 2, (J); Northern Prov., Owen Stanley
Ra., Mamba Pltn, c. 7 km WNW of Kokoda, 08°51'S,
147°41'E, 500 m, 31.viii-l.ix.1984 (RJK accs 84.398, 399,
403) (w, V). AUSTRALIA, QUEENSLAND, Cape York,
xi.1927 (W.B. Barnard); Cape York Pen., Lockerbie
Scrub, 10°46'S, 142°29'E, 19-23.iii.1987 (RJK accs 87.40,
41, 42, 43, 48, 57, 70) (w, 2); Iron Ra., 12°43'S, 143°18'E,
26-31.vii.1981 (RJK accs 81.162, 172, 173) (w, ); ditto,
l-3.vii.1976 (P. Filewood) (w); ditto. West Claudie R.,
3-10.xii.1985 (GBM & DJC) (w).

Worker. Dimensions. TL c. 12.60-14.76; HL 3.12-
3.58; HW 1.84-2.17; Cl 57-62; SL 3.78-4.48; SI 195-
216; PW 1.81-2.03; MTL 4.79-5.74 (27 measured).

Anterior clypeal margin arcuate, often with
very shallow median emargination. Clypeus
with median carina, virtually straight in profile.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

197

Kohout

posteriorly curving into moderately impressed
basal margin. Frontal carinae sinuate with acute,
highly raised margins. Sides of head in front of
eyes almost straight, weakly diverging towards
mandibular bases; behind eyes sides strongly
tapered into narrow occipital margin with
strongly developed occipital lobes. Eyes strongly
convex, in full face view clearly exceeding
lateral cephalic outline. Ocelli lacking, positions
indicated by distinctly raised cephalic sculpture.
Pronotal dorsum convex in profile, distinctly
higher than mesonotum; humeri armed
with strong, acute, more-or-less horizontal,
anterolaterally directed spines; promesonotal
suture distinct. Mesopleural process a distinct,
dentiform lobe. Mesonotum and propodeum
weakly convex in profile. Propodeal spines
relatively long, virtually straight, in lateral view
usually vertical to main axis of body or inclined
forwards, tips sometimes curved posteriorly
or inwards. Dorsum of petiole with a pair of
lateral, dorsoposteriorly directed, acute spines.

Mandibles with numerous piliferous pits.
Clypeus and sides of head finely reticulate-
punctate with intensity and density of sculp-
turation distinctly increasing posteriorly,
vertex and occiput rather coarsely sculptured.
Mesosomal dorsum rugose with intensity
decreasing posteriorly with propodeal dorsum
finely reticulate-punctate. Gaster very finely
shagreened.

Mostly silvery or grey, erect or variously
curved hairs of variable lengths and densities
over the entire body and appendages, including
antennal scapes. Longest hairs on mesosomal
dorsum, clearly longer than greatest diameter
of eye. Head and mesosoma with somewhat
untidy, relatively long, mostly silvery, suberect
pubescence that never completely hides under-
lying sculpturation; most dense and more
appressed on dorsum of head and mesosoma,
somewhat radiating on sides. Gastral pubescence
ranging from diluted on venter to dense on
dorsum, where it is arranged in characteristic
midline pattern.

Body black; mandibles, except bases, apical
antennal segments and legs, including coxae,
mostly light to medium reddish-brown.

Queen. Dimensions: TL c. 13.46-16.28; HL 3.17-
3.78; HW 1.86-2.37; Cl 57-63; SL 3.73-4.59; SI 189-
211; PW 2.02-2.82; MTL 4.74-5.64 (11 measured).

Queen very similar to worker; apart from usual
characters identifying full sexuality, including
three ocelli, complete thoracic structure and
wings differing mainly in configuration of
spines; pronotal spines shorter, projecting
more anteriorly, tips directed forwards; pro-
podeal spines distinctly shorter, oblique to
main axis of body, weakly curved posteriorly;
petiolar spines straight, distinctly shorter than
in worker; sculpturation, pilosity and colour
pattern identical to worker.

Males and immature stages (eggs, larvae in
various stages of development and pupae) in
QM spirit collection.

Remarks. Polyrhachis sexspinosa was originally
described from a queen collected in the 'East
Indies', however like Bolton (1975), I have
been unable to locate the type in any of the
collections examined and consider it lost. I
therefore support the earlier decision of Kohout
& Taylor (1990: 518-520) who recognised the
specimens of a nest series from Papua New
Guinea (see above) as the voucher specimens of
P. sexspinosa by designating one of these workers
as the neotype for this name (ANIC, QM).

Polyrhachis sexspinosa resembles P. reclinata,
with their main distinguishing characters given
in remarks section under the latter species.
Polyrhachis sexspinosa is a relatively common and
widespread species reported from most of the
New Guinean mainland and islands of eastern
Indonesia, including several doubtful records
from the southern Philippines (Kohout, 1989).
In Australia it ranges from Lockerbie Scrub south
to Rocky River, northeast of Coen. Polyrhachis
sexspinosa is somewhat singular within the
sexspinosa-group in that it builds pocket-like nests

198

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A review of the Australian Polyrhachis Ants

of silk, vegetation debris and bark fragments
against the trunks of rainforest trees (Figs 10A-B).

Subgenus Hirtomyrtna subgen. nov.

Myrmhopla Forel, 1915: 107 as subgenus of Polyrhachis Fr.
Smith (in part).

Myrmhopla Forel; Emery, 1925 (in part - as P. viehmeyeri-
group).

Myrmhopla Forel; Kohout, 1990 (in part - as P. viehmeyeri-
‘ group).

Myrmhopla Forel; Dorow, 1995 (in part - as P. vielnneyeri-
' group).

Type species. Polyrhachis hirla Viehmeyer, 1914.

The subgenus Hirtomyrma as conceived here
effectively replaces the Polyrhachis viehmeyeri-
group that was originally established by Emery
(1925), within the subgenus Myrmhopla Forel, for
two, rather unusual species from New Guinea (P.
hirta Viehmeyer, 1913 and P. viehmeyeri Emery,
1921). A third species (P. davydovi Karavaiev) from
the Aru Is, Indonesia, was described by Karavaiev
in 1927. The former group was reviewed by
Kohout (1990) who described four new species
from northern Australia (P. bamaga Kohout,
P. eremita Kohout, P. loweryi Kohout and P.
rustica Kohout) and two extralimital species (P.
grcensladei Kohout from the Solomons and P.
stigmatifera Kohout from Seram I., Indonesia).
An additional species from South East Asia (P.
lama Kohout) was described by Kohout in 1994,
thus raising the number of known species of
the group to ten.

Diagnosis. Worker. Medium sized-ants (HL
1.75-2.20) with general characteristics of the
genus. Mandibles with 4 or 5 teeth, very finely
longitudinally striate. Anterior clypeal margin
truncate medially; posterior margin usually
deeply impressed. Frontal carinae rather
flat, widely separated. Eyes with numerous,
short, erect hairs, strongly convex, almost
hemispherical, clearly exceeding lateral cephalic
outline in full face view. Median ocellus distinct
(as in P. eremita and P. loweryi), vestigial (as in P.
rustica ) or lacking (as in P. bamaga). Pronotal and
propodeal dorsa laterally marginate, virtually
flat; mesonotal dorsum transversely convex

Memoirs of the Queensland Museum | Nature • 2010

with less distinct, rather blunt, lateral margins.
Pronotum armed with somewhat dorsally
flattened, acute spines (except in extralimital
P. lama); their length, direction and degree
of elevation usually highly variable, even
asymmetrical, within a single species (as in P.
rustica). Propodeal spines acute, usually longer
than pronotal pair, variously elevated. Dorsum
of petiole with poorly defined, more-or-less
posteriorly sloping platform, bearing a pair of
widely separated, diverging spines, without
intercalary spines or teeth. Head, mesosoma
and petiole with characteristic vermiculate-
rugose sculpturation and bristle-like hairs,
distinctly shorter than maximum eye diameter.
Gaster shagreened, or with base of first tergite
finely micro-reticulate and more-or-less shiny
(as in most Australian species). Body mostly
light to dark reddish-brown, with mandibles,
clypeus, frontal carinae, spines and posterior
margins of gastral tergites usually narrowly
bordered very dark brown.

Queen. Differing from worker in usual
characters identifying full sexuality, including
three ocelli, complete thoracic structure and
wings. Spines shorter and more stubby than
in worker. Sculpturation, pilosity and colour
essentially as in worker.

Distribution and biology. The main distribution
of the subgenus Hirtomyrma ranges from the
Moluccas, New Guinea and Bismarck Archi-
pelago to the Solomons and extends south to
northern Australia. However, the range of one
species (P. lama Kohout, 1994) extends from
Java and Hong Kong to the Tibetan Plateau
in the Himalayas. All known species of this
group are apparently social parasites of other
ants, notably Ectatomminae and Ponerinae.
The association of P. loweryi with the genus
Rhytidoponera Mayr and the extralimital P. lama,
with the genus Diacamma Mayr, were studied
by Prof. U. Maschwitz (formerly of Johan
Wolfgang Goethe-Universitat in Frankfurt am
Main, Germany) and his team (Maschwitz et al.
2000, 2003).

55(1) 199 1

Kohout

KEY TO AUSTRALIAN SPECIES OF THE
SUBGENUS HIRTOMYRMA
(based on worker caste)

1. Base of first gastral tergite finely shagreened,

rather polished (Fig. 6E); mandibles with 4
teeth 2.

— Base of first gastral tergite closely,

transversely striate, opaque (Fig. 6C);
mandibles with 5 teeth, but with basal tooth
often vestigal bamaga Kohout

2. Larger species (HL >1.85); median ocellus

well developed, distinct 3.

— Smaller species (HL <1.80); median ocellus

rather small, indistinct rustica Kohout

3. Body bicoloured, medium reddish-brown

with most of head, pronotal collar and
anteromedian patch on mesosomal dorsum
light yellowish-brown; antennal scapes
shorter (SI <144) eremita Kohout

— Body more-or-less unicoloured, dark reddish-

brown, with only mandibles, spines and
subpetiolar process lighter; antennal scapes
longer (SI >146) loweryi Kohout

Polyrhachis bamaga Kohout, 1990
(Figs 6A, C-D, 9C)

Polyrhachis bamaga Kohout, 1990: 500, fig. 4. Holotype and
paratype workers. Type locality: QUEENSLAND, Cape
York Pen., Bamaga (R.J. Kohout), QM (examined).

- Remarks. In spite of revisiting the type
locality of P. bamaga specifically to collect more
specimens (see Maschwitz et al., 2003), the types
remain the only specimens of this species known.
They were originally collected along the edge of
a lowland riverine rainforest, foraging over low
vegetation in association with Polyrltachis paxilla
Fr. Smith, 1863 (see Kohout, 1990: 501).

Poli/rliachis eremita Kohout, 1990
(Figs 6B, E-F, 9C)

Polyrhachis eremita Kohout, 1990: 502. Holotype and paratype
workers. Type locality: QUEENSLAND, 4-10 km N of
Marlborough (R.J. Kohout), QM (examined).

Remarks. A thorough search conducted at
the type locality and in brigalow forest north
of Marlborough (see Maschwitz et al., 2003),
produced no nests of the large Rhytidoponera
species with which P. eremita was thought to
be associated. I also visited the area in 2006
and found no Rhytidoponera nests, as well as
a general decline in the number of other ant
species previously found there in 1981. This
could be due to considerable degradation of the
habitat, including substantial deforestation.

Poli/rhachis loiveryi Kohout, 1990
(Figs IF, 7A, C-D, 9C)

Polyrhachis Imveryi Kohout, 1990: 505. Holotype and paratvpe

workers. Type locality: QUEENSLAND, Miles (6.B.

Lowery), ANiC (examined).

Other material. QUEENSLAND, c. 5.5 km NNW of
Miles, 26°36'S, 150°10T, 6-8.iii.2000, dry (U. Maschwitz
et al.) (w, $, j 1 - associated with Rhytidoponera sp.).

Queen, (not previously described) Dimensions:
TL c. 9.27; HL 2.00; HW 1.56; Cl 78; SL 2.34; SI 150;
PW 1.81; MTL 3.22 (1 measured).

Closely resembling worker and apart from
sexual characters, including three ocelli,
complete thoracic structure and wings, differing
as follows: pronotal spines distinctly reduced to
more-or-less triangular, acute teeth. Mesoscutum
with anterior margin evenly rounded in dorsal
view; in profile anterior face distinctly swollen,
widely rounding onto flat dorsum; median line
bifurcate dorsally; parapsides flat anteriorly,
weakly raised posteriorly. Mesoscutellum in
lateral view elevated above dorsal plane of
mesosoma, relatively flat, rounding posteriorly
into distinct metanotal groove. Propodeal
spines short, obliquely elevated; petiolar spines
similar to those in worker but distinctly shorter,
their inner margins continued medially and
posteriorly, forming rather blunt, V-shaped
posterior margin of petiolar dorsum. Sculturation,
pilosity and colour virtually identical to worker.

Males and immature stages (eggs, larvae
and pupae) originally deposited in the
Forschungsinstitute Senckenberg, Frankfurt am

200

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A review of the Australian Polyrhachis Ants

FIG. 10. Nests of Australian Polyrhachis (Myrmhopla) species. A-B. Pocket nest of Polyrhachis sexspinosa
(Latreille) attached to a buttress of a caulifloral rainforest tree at Lockerbie Scrub, Cape York Peninsula
(Photo R.J. Kohout); C. Polydomous nest of Polyrhachis dives Fr. Smith on small sapling in north Queensland
(Photo S.K.A. Robson); D. Nest of Polyrhachis bicolor Fr. Smith in curled leaf (Photo A. N. Andersen); E. Nest
of Polyrhachis mucronata Fr. Smith (Photo ]. Wright).

Main, Germany (see Maschwitz et ai, 2003) were
accidently destroyed (Dorow, pers. comm.).

Remarks. A successful visit to the vicinity of
Miles (see above) yielded numerous nests of
Rhytidoponera spp. ( aciculata - and co/tnc.vn-groups).
Following examination of ten nests of the aciculata-
group sp. and five nests of the coiwexa-group sp.,
six colonies of P. lowcryi were located within the

nests of the former. However, on a subsequent
visit to the same locality, several worker spe-
cimens of P. loweryi were also located within a
nest of a Rln/tidoponera species belonging to the
convexa- group.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

201

Kohout

Polyrhachis rustica Kohout, 1990
(Figs 7B, E-F, 9C)

Polyrhachis rustica Kohout, 1990: 505. Holotype and paratype
workers, queen. Type locality: QUEENSLAND, 4km N
of Collinsville (B.B. Lowery), ANIC (examined).

Other material. QUEENSLAND, Chewko Rd. nr
Mareeba, 3.viii.l975 (B.B. Lowery) (w - associated
with Rhytidoponcra laurata Roger).

Remarks. Both the original localities at Collins-
ville and Mareeba were visited in February, 2000
by Maschwitz et al. (2003) in search of P. rustica.
However, in spite of locating and examining
numerous nests of two unidentified Rhytidoponcra
species, no associated Polyrhachis specimens
were found.

ACKNOWLEDGEMENTS

I am very grateful to the Australian Biological
Resources Study for a research grant supporting
my work on the systematics of Australian
Polyrhachis ants. This work was also supported
by three Ernst Mayr Grants that allowed me
to travel and study specimens in the Museum
of Comparative Zoology, Harvard University
and several other museums and institutions
in the USA and UK. I am also indebted to
the Smithsonian Institution for a Short-Term
Visitor Grant to study the Polyrhachis ants in the
collection of the National Museum of Natural
History, Washington. My sincere thanks go
to Drs Steve Shattuck (ANIC), Stefan Cover
.(MCZC) and Barry Bolton (BMNH) for unlimited
access to the collections in their care. 1 thank
Prof. U. Maschwitz (formerly of JWGU) for the
opportunity to join the expedition to northern
and central Queensland, financed by the
Deutsche Forschungsgemeinschaft. My thanks
are due to Dr Yoshiaki Hashimoto (MNHA) for
his patience and care in the preparation of the
digital images used for illustrations, to Geoff
Thompson (QM) for producing the distribution
maps and to Ms Natalie Barnett (ANIC) for
computer-based enhancement of the line draw-
ings. I also thank Dr Gary Alpert and Mary
Corrigan (both Harvard University) for their

hospitality during my visits to the MCZC. To D r
Premek Hamr (Upper Canada College, Toronto
Canada) I owe my thanks for assistance with the
translation of the French diagnosis of Myrmhopi a
I thank the curators and staff of the museums
and institutions listed earlier for their help and
kindness during my visits and for providing
loans of types and other specimens. My sincere
thanks go to my colleagues, Drs Chris Burwell
Geoff Monteith, Ms Susan Wright (all QM), f 0 ^
their valuable support during the course of this
study. I would also like to extend my gratitude
to the Environmental Protection Agency and the
Department of Natural Resources for issuing
the permits to allow collecting in Queensland's
National Parks and State Forests. My thanks
are also due to Parks Australia, Department
of Environment and Heritage and to the Parks
and Wildlife Commission of the Northern
Territory for permits to collect in Kakadu and
other National Parks in the Northern Territory
Finally a special thankyou to Chris Burwell
(QM) for reading and commenting on a draft of
the manuscript.

LITERATURE CITED

Bolton, B. 1975. The sexspinosa-group of the ant genus
Polyrhachis F. Smith (Hym. Formicidae). Journal
of Entomology (Series B) 44: 1-14, figs,

1995. A new general Catalogue of llw Ants of the World.
Harvard University Press: Cambridge, Mass.,
504 pp.

Bolton, B„ Alpert, G.D., Ward, P.S. & Naskrecki, P.
2007. Bolton's Catalogue of Ants of the World: 1758-
2005. Harvard University Press, Cambridge
Mass., CD-ROM.

Clark, J. 1928. Australian Formicidae. Journal of the Royal
Society of Western Australia 14: 29-41 .

1930. New Formicidae, with notes on some little-
known species. Proceedings of the Royal Society of
Victoria 43: 2-25.

1941. Australian Formicidae. Notes and new species.
Memoirs of the National Museum of Victoria 12:
71-94.

Donisthorpe, H. 1947. Some new ants from New
Guinea. Annals and Magazine of Natural History
(11)14: 183-197.

202

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

A review of the Australian Polyrhachis Ants

Dorow, W.H.0. 1995. Revision of the ant genus Poly-
rhachis Smith, 1857 (Hymenoptera: Formicidae:
Formicinae) on subgenus level with keys,
checklist of species and bibliography. Courier
Forschungsinstitut Senckenberg 185: 1-113.

Drury, D. 1773. Illustrations of Natural History.
Wherein are exhibited upwards of two hundred
and twenty figures of exotic insects 2. 90 pp.
London.

Emery, E. 1887. Catalogo delle formiche esistenti
nelle collezioni del Museo Civico di Genova.
Parte terza. Formiche della regione Indo-Malese
e dell' Australia. Annali del Museo Civico di Storia
Naturale di Genova 4(2): 209-258.

1897. Viaggio di Lamberto Loria nella Papuasia
orientale. 18. Formiche raccolte nella Nuova
Guinea dal Dott. Lamberto Loria. Annali
del Museo Civico di Storia Naturale di Genova
(2)18(38]: 546-594.

1925. Hymenoptera, Fam. Formicidae, subfam.
Formicinae. In: Genera Insectorum. (Wytsman ed.)
Fasc. 183. Bruxelles. 302 pp.

Forel, A. 1910. Fourmis des Philippines. Philippine
Journal of Science 5 (sect. D): 121-130.

1915. Results of Dr. E. Mjoberg's Swedish scientific
expeditions to Australia, 1910-1913. 2. Ameisen.
Arkiv for Zoologi 9(16): 1-119.

Hung, A.C.F. 1967. A revision of the ant genus
Polyrhachis at the subgeneric level. Transactions
of the American Entomological Society 93: 395-422.

International Commission On Zoological Nomen-
clature, 1999. International Code of Zoological
Nomenclature (Fourth Edition). London. 306pp.

Kohout, R.J. 1987. Three new Polyrhachis sexspinosa-
group species from the Philippines (Hymenoptera:
Formicidae: Formicinae). Memoirs of the
Queensland Museum 25(1): 169-176.

1988. Nomenclatural changes and new Australian
records in the genus Polyrhachis (Hymenoptera:
Formicidae: Formicinae). Memoirs of the
Queensland Museum 25(2): 429-438.

1990. A review of the Polyrhachis viehmeyeri species-
group (Hymenoptera: Formicidae: Formicinae).
Memoirs of the Queensland Museum 28(2): 499-508.

1994. Polyrhachis lama, a new ant from the Tibetan
plateau (Formicidae: Formicinae). Memoirs of the
Queensland Museum 35(1): 137-138.

1998. New synonyms and nomenclatural changes in
the ant genus Polyrhachis Fr. Smith (Hymenoptera:

Formicidae: Formicinae). Memoirs of the
Queensland Museum 42(2): 505-531.

2000. A review of the distribution of the Polyrhachis
and E chinopla ants of the Queensland Wet
Tropics (Hymenoptera: Formicidae: Formicinae).
Memoirs of the Queensland Museum 46(1): 183-
209.

2006a. A review of the Polyrhachis cryptoceroides
species-group with description of a new species
from Thailand (Hymenoptera: Formicidae).
Myrmecologische Nachrichten 8: 145-150.

2006b. Review of Polyrhachis (Cyrtomyrma) Forel
(Hymenoptera: Formicidae: Formicinae) of
Australia, Borneo, New Guinea and the Solomon
Islands with descriptions of new species. Memoirs
of the Queensland Museum 52(1): 87-146.

2008. A review of the Polyrhachis ants of Sulawesi
with keys and descriptions of new species
(Hymenoptera: Formicidae: Formicinae). Memoirs
of the Queensland Museum 52(2): 255-317.

Kohout, R.J. & Taylor, R.W. 1990. Notes on Australian
ants of the genus Polyrhachis Fr. Smith, with
synonymic list of the species (Hymenoptera:
Formicidae: Formicinae). Memoirs of the Queensland
Museum 28(2): 509-522.

Latreille, P.A. 1802. Histoire Naturelle des Fourmis, et
recueil de memoires et d'obsewations sur les abeilles,
les araignees, les faucheurs, et autres insectes. Paris.

Le Guillou, E.J.F. 1842. Catalogue raisonn£ des
insectes hymenopteres recueillis dans le voyage
de circumnavigation des corvettes 1' Astrolabe et
La Zelde. Annates de la Societe Entomologique de
France 10(1841): 311-324.

Maschwitz, U., Dorow, W.H.O., Buschinger, A. &
Kalytta, G. 2000. Social parasitism involving
ants of different subfamilies: Polyrhachis lama
(Formicinae) an obligator)' inquiline of Diacamma
sp. (Poneriane) in Java. Insectes Sociaux 47 (2000):
27-35.

Maschwitz, U., Go, C., Dorow, W.H.O., Buschinger,
A. & Kohout, R.J. 2003. Polyrhachis loweryi
(Formicinae): A guest ant paraziting Rhytidoponera
sp. (Ponerinae) in Queensland, Australia. Insectes
Sociaux 50 (2003): 69-76.

Robson, S.K.A & Kohout, R.J. 2005. Evolution of nest-
weaving behaviour in arboreal nesting ants of
the genus Polyrhachis Fr. Smith (Hymenoptera:
Formicidae). Australian Journal of Entomology
(2005) 44(2): 164-169.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

203

Kohout

2007. A review of the nesting habits and socioecology
of the ant genus Polyrhachis Fr. Smith. Asian
Myrmecology 1: 81-99.

Smith, Fr. 1857. Catalogue of the hymenopterous insects
collected at Sarawak, Borneo; Mount Ophir,
Malacca; and at Singapore, by A.R. Wallace.
Journal of the Proceedings of the Linnean Society
of London, Zoology 2: 42-88.

1858. Catalogue of Hymenopterous Insects in the collection
of the British Museum 6. Formicidae. London.

1859. Catalogue of hymenopterous insects collected
by Mr. A.R. Waflace at the Islands of Aru and
Key. Journal of the Proceedings of the Linnean Society
of London, Zoology 3: 132-158.

1863. Catalogue of Hymenopterous insects collected
by Mr A.R. Wallace in the Islands of Mysol,
Ceram, Waigiou, Bouru and Timor. Journal of the
Proceedings of the Linnean Society, Zoology 7: 6-48.

204

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Rabaulichthys squirei, a new species of Sailfin
Anthias (Serranidae: Anthiinae) from the Coral Sea

John E. RANDALL

Bishop Museum, 1525 Bernice St, Honolulu, HI 96817 USA. Email: jackr@hawaii.rr.com

Fenton WALSH

Northern Barrier Reef Marine Life, Miles St. Cairns, Qld 4870 Australia.

Citation: Randall, J.E. & F. Walsh 2010 03 15 Rabaulichthys squirei, a new species of Sailfin Anthias
(Serranidae: Anthiinae) from the Coral Sea. Memoirs of the Queensland Museum - Nature 55(1V
205-211. Brisbane. ISSN 0079-8835. Accepted: 23 September 2009.

ABSTRACT

Rabaulichthys squirei is described as a new species of the serranid subfamily Anthiinae
from 12 specimens collected at Flinders Reef and Flora Reef in the Coral Sea from 6
to 54 m. A huge population was also observed at Holmes Reef in 55 m. The species is
distinct from three others of the genus by the combination of a high count of 23 or 24
lower-limb gill rakers, relatively long snout, a very high spinous portion of the dorsal fin
of the male, and subtle differences in colouration. It forms small elusive aggregations
that feed on zooplankton, mainly over open substrata of coral rubble. □ Coral Sea,
Serranidae, Anthiinae, Rabaulichthys, new species.

The Sailfin Anthias, Rabaulichthys altipintiis,
was described as a new genus and species of
serranid fish of the subfamily Anthiinae by
Allen (1984) from specimens collected near
Rabaul, New Britain. Six specimens, 34.8-45.5
mm SL, were collected from small aggregations
about a meter above a sloping coral-rubble
bottom in 30-40 m. This species typically forms
aggregations of about three to eight individuals.
The new genus was characterised by a slender
body, the dorsal fin of 10 weak flexible spines,
and 15 or 16 rays, the spinous part greatly
elevated in the male, 19-21 pectoral rays, over
50 lateral-line scales, no teeth on the vomer and
palatines, and fleshy papillae on the edge of the
posterior half of the orbit.

A second species of the genus, R. stigmaticus,
was described by Randall & Pyle (1989) from four
specimens collected in 35 m over a rubble bottom
at Ari Atoll, Maidive Islands. Its occurrence in

Sri Lanka was confirmed by a photograph of an
adult male sent by Rohan Pethiyagoda.

A third species, Rabaulichthys suzukii, was
described by Masuda & Randall (2001) from three
specimens taken in 5-15 m at Izu Peninsula,
Shizuoka Prefecture, Japan.

As mentioned by Randall & Pyle (1989) and
Masuda & Randall (2001), the Bishop Museum
in Honolulu has 40 specimens of a small species
of Rabaulichthys that were collected by dredging
in 1972 by the US National Marine Fisheries
Service at Condor Reef in the Caroline Islands
from a depth range of 37-46 m. This species is
distinct in its very slender body and having
a filament on the tip of each caudal-fin lobe.
Because of the poor condition of the specimens
and no information on life color, this species
has not been described.

Two specimens of a fifth species of the genus
were collected in October, 2007 by aquarium fish

Memoirs of the Queensland Museum | Nature • 2010 • 55(1) • www.qm.qld.gov.au

205

Randall & Walsh

collector Tim Bennett from 54 m at Flinders Reef
in the Coral Sea, 220 km east of Townsville,
Queensland. One fish is a male with the
expected very high spinous portion of the
dorsal fin, and the other a female with a dorsal
fin of near-uniform height. The male has a lateral-
line scale count of 57 on one side and 59 on the
other, whereas the female has counts of 51 and
52 lateral-line scales. This is more variation than
expected for a single species of anthiine fish, so
we waited for additional material. Ten months
later, 10 specimens were collected by Cadel Squire
at Flora Reef in 6-9 m. Their lateral-line scale
counts of both sides filled the gap between 52
and 57, with seven counts of 54 scales and five
of 55 scales.

We present here the description of the new
species of Rabaulichthys from the Coral Sea, the
first of the genus for Australian waters.

METHODS AND MATERIALS

Lengths of specimens are given as standard
length (SL), measured from the front of the
upper lip to the base of the caudal fin (posterior
end of the hypural plate); head length (HL) is
measured from the same anterior point to the
posterior end of the opercular flap; body depth
is taken vertically from the base of first dorsal
spine; body width is the maximum width just
.posterior to the gill opening; orbit diameter is
the greatest fleshy diameter, and interorbital
width the least fleshy width; upper-jaw length is
taken from the front of the upper lip to the mid-
posterior end of the maxilla; caudal-peduncle
depth is the least depth, and caudal-peduncle
length the horizontal distance between verticals
at the rear base of the anal fin and the caudal-
fin base; lengths of fin spines and rays of the
dorsal and anal fins are measured to their
extreme base; caudal-fin length is the horizontal
distance from the base of the fin to a vertical at
the tip of the longest ray; caudal concavity is
the horizontal distance between verticals at the

tips of the longest and shortest rays; pectoral-fin
length is the length of the longest ray; pelvic-fin
length is measured from the origin of the pelvic
spine to the tip of the longest soft ray. Pectoral-
fin ray and lateral-line scale counts were made
on both sides. Gill-raker counts were made on
the first gill arch of the right side and include
rudiments; the raker at the angle is contained
in the count of the lower-limb.

In the description of the new species, data in
parentheses refer to paratypes. Measurements
in Table 1 and the diagnosis are given as percent
of the standard length, those in the description
as proportions of the standard length or head
length, rounded to the nearest .05.

Specimens for this study have been deposited
in the fish collections of the Bernice P. Bishop
Museum, Honolulu (BPBM); the Queensland
Museum, Brisbane (QM), the National Museum
of Natural History, Washington, D.C. (USNM);
and the Western Australian Museum, Perth
(WAM).

KEY TO THE SPECIES OF RABAULICHTHYS

1. Snout short, 5.5-6.4% SL; lower-limb gill

rakers 21-22; length of longest dorsal spine
of mature male 25.3-26.2% SL; body depth
26.6-29.7% SL (New Britain) altipinnis

— Snout not short, 6.8-7.4% SL; lower-limb

gill rakers 23-24; length of longest dorsal
spine of mature male 28.7-39.0% SL; body
depth 23.8-27.7% SL 2

2. A narrow, elliptical, black band parallel to

back near middle of soft portion of dorsal
fin; a large, pale-edged, nearly square, dark
reddish blotch on body below soft portion
of dorsal fin (Maidive Islands and Sri Lanka;
only males available) stigmaticus

— No black band in soft portion of dorsal fin;

no large, nearly square, pale-edged, dark
reddish bar on body below soft portion of
dorsal fin 3

3. Body depth 23.8-25.3% SL; caudal-fin length

206

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Rabaulichthys squirei, a new species of Sailfin Anthias

of males 31.8-35.8% SL (Coral Sea)
squirei, sp. nov.

— Body depth 26.0-27.7% SL; caudal-fin length
of males 25.5-31.2% SL (only males available,
Shizuoka Prefecture, Japan) suzukii

Rabaulichthys squirei Randall & Walsh sp. nov.
(Figs. 1 A-D)

Etymology. We are pleased to name this species for
Cadel Squire who collected most of the type specimens.

Material. HOLOTYPE: QM 1.38237, 62.2 mm
SL, Coral Sea, Flinders Reef, north end, 17°24'1"S,

148°25'3"E, 54 m, barrier net, T. Bennett, 1/10/2007.
PARATYPES. QM 1.38441, 53.0 mm, same data as
holotype; QM 1.38420, 4: 36.2-47.6 mm SL, Coral Sea,
Flora Reef, 16°45'S, 147°46'E, 6-9 m, 8/2008; BPBM
40978, 2: 35.0-46.6 mm; USNM 395918, 2: 37.2-43.1
mm; and WAM P.33113-001, 2: 37.8-42.5 mm, all with
same data as QM 1.38420.

Diagnosis. Dorsal rays X,16; anal rays 111,7;
pectoral rays 19—21 (rarely 21); lateral-line scales
51-59; gill rakers 9 + 23-24;' body depth 23.8-25.3%
SL; head length 28.5-30.6% HL; snout length 6.8-
7.4% SL; longest dorsal spine of mature males
28.7-31.3% SL; caudal fin of mature males 31.8-

TABLE 1. Proportional measurements of type specimens of Rabaulichthys squirei as percentages of the
standard length

Holotype

Para types

QM 1.38237

QM 1.38441

QM 1.38420

WAM P.33113

QM 1.38420

Standard length (mm)

62.2

53.0

47.6

47.2

42.5

39.5

36.2

Sex

male

female

male

female

Body depth

24.2

23.8

25.2

25.0

24.0

24.5

25.3

Body width

14.1

14.7

13.2

13.7

14.8

14.6

13.8

Head length

28.7

28.5

29.2

30.3

30.6

30.4

30.5

Snout length

7.4

7.1

7.0

6.8

7.0

6.8

7.3

Orbit diameter

7.6

7.9

8.5

8.7

9.0

9.6

Interorbital width

8.8

8.6

8.9

9.2

9.5

9.7

Upper-jaw length

12.3

12.2

13.0

13.4

13.5

12.8

13.5

Caudal-peduncle depth

10.7

11.5

11.6

11.8

11.0

10.9

Caudal-peduncle length

24.2

24.5

25.2

25.0

25.1

24.0

Predorsal length

25.8

26.3

25.2

26.0

26.7

26.6

27.8

Preanal length

57.8

59.8

59.5

59.7

57.4

59.9

58.7

Prepelvic length

29.0

27.9

28.5

28.7

27.8

27.7

27.6

Dorsal-fin base

60.5

60.2

60.5

60.3

60.5

60.8

58.7

First dorsal spine

9.0

7.5

9.2

8.7

7.7

7.9

Second dorsal spine

15.2

9.7

16.3

15.0

14.9

10.1

Longest dorsal spine

30.4

13.2

31.0

31.3

27.4

13.7

13.0

Longest dorsal ray

19.3

12.2

18.8

19.1

12.9

13.1

Anal-fin base

17.7

17.3

16.8

16.5

17.8

16.8

17.7

First anal spine

4.7

4.5

4.4

4.6

4.7

5.1

5.2

Third anal spine

9.5

8.2

9.3

9.5

9.6

8.8

9.1

Longest anal ray

14.4

13.0

12.6

13.7

14.4

15.2

13.3

Caudal-fin length

31.8

27.0

34.7

35.2

35.8

27.7

28.3

) Caudal concavity

21.4

13.7

17.6

22.5

24.0

12.8

13.9

Pectoral-fin length

25.7

25.2

25.4

25.9

26.2

25.7

Pelvic-spine length

12.9

13.3

12.7

12.1

13.5

13.4

13.8

Pelvic-fin length

22.9

20.8

23.0

23.4

23.5

20.8

22.1

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

207

Randall & Walsh

35.8% SL; no black band in soft portion of dorsal
fin; a series of indistinct dark bars of unequal
width in middle third of standard length, none
with pale margins; colour in life light red, the
scale centres pale, the dark bars reddish brown;
an oblique, pale-edged, brownish orange band
from snout through eye and across cheek; pelvic
fins black with red rays; remaining fins pink to
yellowish. Largest specimen, 62.2 mm SL.

Description. Dorsal-fin rays X,16, the first 3
(3-6) soft rays unbranched, the last branched to
base; anal-fin rays III, 7, all soft rays branched;
pectoral-fin rays 19 (19-21, only one of 11
paratypes with 21), the rays branched except
upper two and lowermost; pelvic rays 1,5, all soft
rays branched; principal caudal-fin rays 15, the
middle 13 branched, upper and lower procurrent
caudal-fin rays 12, the posterior three segmented;
lateral-line scales 57 (51-59); scales above lateral
line to middle of spinous portion of dorsal fin
3.5; scales below lateral line to origin of anal fin
15; lowest count of circumpeduncular scales
28; gill rakers 9 + 24 (9 + 23-24, five of 11 with
24); pseudobranchial filaments of holotype 15;
vertebrae 10 + 16; supraneural (predorsal) bones
2, projecting toward first neural spine; first two
dorsal spines close together, associated with
first dorsal pterygiophore, in space between
first two neural spines; pterygiophores of third
and fourth dorsal spines in space between
. second and third neural spines; remaining
pterygiophores one on one for each vertebra.

Body elongate, the depth 4.15 (3.95-4.2) in SL;
body compressed, the width a little more than
one-half body depth, 1.95 (1.95-2.0) in HL; snout
length 3.9 (4.0-4.5) in HL; eye diameter 3.8 (3.2-
3.6) in HL; posterior half of orbit with 14 small

fleshy papillae; interorbital space convex, the
least fleshy width 3.25 (3.25-3.4) in HL; caudal-
peduncle depth 2.7 (2.6-2.85) in HL; caudal-
peduncle length 1.2 (1.15-1.25) in HL.

Mouth moderately large, the maxilla reaching
slightly posterior to middle of eye, the upper-
jaw length 2.35 (2.25-2.4) in HL mouth terminal
and moderately oblique, forming an angle of
about 40° to horizontal axis of body; posterior
end of maxilla truncate, the upper corner
rounded; front of upper jaw with a pair of stout,
blunt, forward-projecting, conical teeth (double
on one side of holotype), about one-fourth pupil
diameter in length, separated by an indented
symphyseal gap of one-half pupil diameter; side
of upper jaw with a row of 16 small, slender,
conical teeth, progressively shorter posteriorly,
the tips not reaching edge of upper lip; lower jaw
with a pair of stout, very blunt, canine teeth
that project laterally as much as anteriorly,
and fit medially to upper canine teeth when
jaws closed; side of lower jaw with a row of 16
teeth, more slender than upper teeth; anterior
teeth of lower jaw retrorse and progressively
larger to eighth, which is recurved; remaining
teeth more slender, near-vertical, and unevenly-
spaced; no teeth on vomer or palatines; tongue
thin and leaf-like with sharply pointed tip, the
upper surface with small papillae. Gill rakers long
and slender, the longest at angle three-fourths
orbit diameter.

Anterior nostril a short fleshy tubule, the rim
higher posteriorly, about three-fourths pupil
diameter before centre of eye; posterior nostril
ovate, without a rim, about 1.5 anterior nostril
diameters dorsoposterior to anterior nostril.

FIG. 1. A, Rabnulichthys squirci, holotype, QM 1.38237, male, 62.2 mm SL, Flinders Reef, Coral Sea (F. Walsh).
B, para type, immature female, 35.0’mm SL, Flora Reef, Coral Sea (F. Walsh). C, paratype, WAM P.33113,
male, 42.5 mm SL, Flora Reef, stress colour pattern (F. Walsh); D, paratype, QM 1.38420, female, 36.2 mm SL,
Flora Reef, stress colour pattern (F. Walsh); E, R. altminnis, holotype, WAM P.28179, male, 45.5 mm SL, New
Britain (G.R. Allen); F, R. altipinnis, paratype, WAM P.28280, female, 38.6 mm SL, New Britain (G.R. Allen).
G, R. stigmaticus, male, Sri Lanka (R. PetHiyagoda); H, R. suzukii, male, Suruga Bay, Japan (A. Mishiku).

208

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Rabaulichthys squirei, a new species of Sailfin Anthias

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

209

Randall & Walsh

Opercle with three small flat spines, the middle
one largest and most posterior; opercular mem-
brane extending a spine's length beyond tip
of middle spine; posterior edge of preopercle
of holotype with 45 serrae, progressively larger,
in general, ventrally, with only five small serrae
continuing on ventral edge (largest paratype
with 34 serrae posteriorly on preopercle, and
smallest paratype with 29).

Scales coarsely ctenoid, less so ventrally, and
cycloid on abdomen before anus; head fully
scaled, including maxilla, except a narrow zone
around orbit that is broader anteriorly to include
nostrils; scales dorsally on snout progressively
smaller and more embedded anteriorly, the
extreme front naked; no scales on dorsal and
anal fins; scales on caudal fin extending nearly to
posterior margin; scales present on about basal
one-third to two-fifths of pectoral fins; pelvic fins
with a pointed, midventral, scaly process that
extends two-thirds length of pelvic spine.

Lateral-line ascending in a straight line to 3.5
scale rows below base of fifth dorsal spine, then
following contour of back to base of caudal
fin. Pores of cephalic lateral-line system small.
Those apparent are: dorsally at front of snout,
above posterior nostril, one on each side in inter-
orbital, and as a series of very small pores around
orbit, ending in pore below anterior nostril; 11
pores in preopercular-mandibular series.

Origin of dorsal fin slightly anterior to upper
end of gill opening, the predorsal length 3.9 (3.6-
3.85) in SL. Dorsal-fin spines very slender and
flexible; first spine 3.2 (3.2-3.95) in HL; second
spine 1.9 (1.8-2.05) in HL in males, (2.95-3.0)
in females; fifth spine longest, longer than HL
in males, 3.3 (3.2-3.5) in SL, and (7.3 -7.7) in
females; first soft ray longest, 1.5 (1.55-1.6) in
HL in males, (2.15-2.3) in females. First anal-fin
spine 6.1 (5.9-6.6) in HL; second spine notably
thicker than third spine, but slightly shorter;
third spine 3.0 (3.1 -3.5) in HL; second soft ray
longest, 2.0 (2.0-2.3) in HL. Caudal fin lunate.

its length in males 3.15 (2.8-2.9) in SL, in females
(3.5-3.7) in SL; caudal concavity in males 1.35
(1.25-1.65) in HL, in females (1.15-1.2) in HL.
Pectoral fins pointed, the middle rays longest,
1.1 (1.15-1.2) in HL. Pelvic spine 2.2 (2.15-2.5)
in HL; third soft ray longest, 1.15 (1.25-1.3) in
HL in males, (1.35-1.45) in females.

Colour of holotype in alcohol pale yellowish
with four dusky bars on side of body, about
one-fourth of their length above lateral line;
first dark bar a double bar, the third bar about
twice as broad as others and darker anteriorly;
occipital region dusky; pelvic fins with pale grey
rays and blackish membranes; remaining fins
with pale lavender-grey rays and translucent
pale yellowish membranes.

Colour of holotype in life as in Fig. 1A. The
colour of paratypes is shown in Fig. 1B-D. The
darker red and more strongly barred pattern of
the male and female of Figs. C and D are stress
colouration.

Remarks. The genus Rabaulichthys is currently
represented by four described species from widely
separated localities of the Indo-Pacific region:
Maidive Islands and Sri Lanka, New Britain,
Japan, and with the description of R. squirei, reefs
of the Coral Sea off Queensland.

All four species share the same fin-ray counts,
and any difference in the number of lateral-line
scales will be modal at best. Among the meristic
characters, only the lower-limb gill-raker count
provides a slight separation of the type species,
R. altipinnis, from the remaining three species.
The salient morphological difference is in the
more slender body of R. squirei, 23.8-25.3% SL,
compared to 25.5-29.7% SL for the others. Al-
though there is a sharing of salient features of
colour pattern, red to pink overall, with an
oblique, pale-edged, brownish orange band from
the eye across the cheek, and predominantly
black pelvic fins in the male, some colour
differences are evident from a comparison of
the images of Fig. 1.

210

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Rabaulichthys squirei, a new species of Sailfin Anthias

The species of Rabaulichthys form small aggre-
gations that feed on zooplankton above the
substratum. They are generally found over open
stretches of coral rubble, and therefore avoid
the many predators of coral reefs. Because of
their small size, they readily take shelter in the
rubble. The paucity of specimens of this genus
can be partly explained by this habitat, which is
not often visited by scubadivers, whether sport
divers or aquarium fish collectors.

Our first two specimens of R. squirei, the 62.2-
mm male holotype and the 53.0-mm female
paratype, were collected from Flinders Reef in 54
m. Our next 10 specimens, which include males
that measured 35.0-47.6 mm SL, were collected
in 6-9 m at Flora Reef. In view of the difference
in size of the specimens, the different depth of
the two collections, and the different localities,
we wondered if the fish might represent two
different species. The collectors assured us
they are the same, explaining that the species
is generally larger in deeper water. We could
find no differences in colour or morphology in
specimens from the two localities.

In addition to Flinders Reef and Flora Reef,
this species has been seen at Holmes Reef (1 6°43'S,
147°55'E) in about 55 m, where thousands occur
over a very large expanse of nearly flat rubble
bottom. They swim together in small groups, often
as only two or three individuals, and feed from
1-3 m above the substratum. In shallower water,
this species sometimes mixes with aggregations
of other anthiine fishes, especially Pseudanthias
dispar (Herre), and is easily overlooked.

Mature males of all the species of Rabaulichthys
have the spinous portion of the dorsal fin greatly
elevated, and as expected, display it fully erect

during courtship. The first species of the genus
was named R. alhpinnis for this feature. Ironically,
it has the lowest spinous dorsal fin of the four
known species. Mature males also have longer
and more lunate caudal fins and longer pelvic
fins than females (see Table 1).

Immature males and females of Rabaulichthys
squirei are morphologically indistinguishable, but
black pigment develops in the pelvic fins of the
males before the dorsal spines become longer.

ACKNOWLEDGMENTS

We thank foremost Cadel and Lyle Squire
and Tim Bennett for their special effort to collect
this elusive species for us. We are also very
grateful to Gerald R. Allen, Rohan Pethiyagoda,
and Akihiko Mishiku (via Hiroshi Senou) for
providing photographs of species of Rabaulichthys.
The photograph of the male of R. suzukii is filed as
KPM-NR 34642 in the image Database of Fishes
of the Kanagawa Prefectural Museum of Natural
History. Thanks are also due Loreen R. O'Hara of
the Bishop Museum for x-rays, and Helen A.
Randall for review of the manuscript.

LITERATURE CITED

Allen, G.R. 1984. A new genus and species of anthiid
fish from Papua New Guinea. Revue frangaise
d'Aquariologie 11: 47-50.

Masuda, H., & Randall, J.E. 2001. Rabaulichthys suzukii,
a new anthiine fish from Japan. Ichthyological
Research 48: 77-81 .

Randall, J.E. & Pyle, R.M. 1989. A new species
of anthiine fish of the genus Rabaulichthys
(Perciformes: Serranidae) from the Maidive
Islands. Special Publication of the J.L.B.
Smith Institute of Ichthyology 47: 1-6.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

211

A putative hybrid of the Murray Crayfish, Euastacus
armatus (Crustacea: Decapoda: Parastacidae)

Diana STREET
Geoffrey EDNEY
Darryl ROWE
Susan H. LAWLER

Department of Environmental Management and Ecology, La Trobe University, Wodonga, VIC
3690, Australia. Email: s.lawler@latrobe.edu.au

Citation: Street, D., Edney, G., Rowe, D. & Lawler, S.H. 2010 03 15. A putative hybrid of the Murray
Crayfish, Euastacus armatus, (Crustacea: Decapoda: Parastacidae). Memoirs of the Queensland
Museum - Nature 55(1): 213-224. Brisbane. ISSN 0079-8835. Accepted: June 2009

ABSTRACT

An unusual population of freshwater crayfish of the genus Euastacus found in the
East Buffalo River, Victoria, is morphologically distinguishable from other populations,
and in particular has a marbled or camouflage pattern on the carapace. The cheliped
dactylar spine counts and the number of zygocardiac teeth anterior to the ventral ear of
the zygocardiac ossicle (TAP) are somewhat reduced but these values overlap with the
range of the most similar species (E. armatus). Genetic sequences from the C01 and
16S mitochondrial DNA regions were also undertaken, and the results indicated that
these animals may represent an aberrant population of E. armatus (the Murray Crayfish)
or perhaps a hybrid between E. armatus and E. woiwuru, two species that are found
downstream and upstream, respectively, of the population investigated. The conservation
status of the population remains unclear. □ Decapoda, Parastacidae, mitochondrial
DNA, Buffalo River, conservation, Murray Cray, hybrid.

Australian freshwater crayfish are an ancient
and diverse group that are increasingly in need
of conservation (Horwitz 1990a; Merrick 1997;
O'Brien 2007). Spiny freshwater crayfishes
of the genus Euastacus are found on the east
coast of Australia, from the Great Dividing
Range in Victoria to isolated mountains in
northern Queensland (Clark 1941; Morgan
1986, 1988, 1997). New species have recently
been described from north-eastern New South
Wales (Coughran 2002, 2005). Most species in
the genus are considered short-range endemics
(Harvey 2002), and looking at a map of their
distributions (eg. Shull et al. 2005) makes it
easy to see why; Euastacus species are usually lim-
ited to a single catchment or mountain top. They
are most commonly found in cold, clear mountain

streams and rivers, and their preferred habitat
may be at risk due to climate change.

The Murray Cray, Euastacus armatus (von
Martens 1866), has the widest distribution in
the genus, being found in both the Murray
and Murrumbidgee Rivers and their tributaries
(McCarthy 2005; Gilligan et al. 2007). The
species is considered to be remarkably invariant
morphologically, even across its broad range
(Morgan 1986).

Euastacus woiwuru (Morgan 1986) is a small
species which occurs on both sides of the Great
Dividing Range in Victoria. This species is found
in the Dandenong ranges near Melbourne, as well
as in central and northern Victoria. Its most closely
related species is Euastacus kershawi (Smith 1912),

Memoirs of the Queensland Museum | Nature • 2010 • 55(1) • www.qm.qld.gov.au

213

Street, Edney, Rowe & Lawler

FIG. 1. A, the marbled Euastacus and B, E. arnmtus, both
found at Schultz Track on the East Buffalo River.

the Gippsland spiny crayfish (Shull et al. 2005).
Both E. armatus and E. woiwuru are found in the
Buffalo River in northeast Victoria.

In 2002 we visited Dandongadale, on the
Buffalo River near the junction of the Rose
River, in order to recollect Euastacus woiwuru
that had first been collected at that locality by
P..Horwitz in 1982 (Morgan 1986). When we
searched the area this time we could only find
E. armatus. We continued to search upstream
and found an unusual population of crayfish at
Schultz Track in 2002 with a marbled carapace
(Fig. 1). Further searches of the Buffalo River
were interrupted by the bushfires of 2003,
which closed some roads for 18 months. When
we were able to return to the site, it had been
altered by bulldozers and the water was
affected by heavy erosion and ash from the
fire. We were unable to find any crayfish at that
time. In 2006 the fires came again, with heavy
fire-fighting taking place near Schultz Track.
When the roads opened again in 2007, we were

finally able to collect a number of crayfish at
Schultz Track and upstream of the site. We were
also thus able to determine that the marbled
population of Euastacus at the Schultz Track
site actually occurred at the boundary between
E. armatus and E. woiwuru populations.

The marbled Euastacus at Schultz Track have
a distinctive colour pattern on the carapace (Fig.
1), and a few other characters that differ from
the other spiny crayfish species in the river.
We originally misidentified these crayfish as
Euastacus crassus (Riek 1969) and included
the DNA in a large phylogeny of Euastacus
(specimen number KC2654 in Shull et al. 2005),
however, its position on the phylogeny showed
clearly that it was not E. crassus (Shull et al.
2005). The specimen was then sent to Dr John
Short of the Queensland Museum, who said he
believed it to be a new species (pers. com.). As
such it was listed as an undescribed species in
a recent review of the conservation status of
Victorian freshwater crayfish (O'Brien 2007).

For measures to be invoked toward the
management of crayfish populations, species
have to be recognised as threatened and in need
of protection (Merrick 1997). The conservation
status of the unknown crayfish at Schultz
Track could not be clarified until its taxonomic
status was resolved. Further collections were
interrupted by road closures due to wild fires
in 2003 and 2006. We were finally able to collect
more individuals in 2007, and discovered that
the site where we found the unusual population
occurred on the species boundary between
E. armatus and E. woiwuru. In this study we
compare the unusual crayfish with other local
Euastacus species using both morphological and
genetic characters.

METHODS

Description of the study site. The Buffalo River
flows north from the Barry Mountains and
joins the Ovens River near Myrtleford, Victoria.

214

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Murray Crayfish, Euastacus armatus

destructor.

South of Lake Buffalo (a man-made reservoir),
the Buffalo River is adjacent to pine plantations
and cattle stations as well as State Forest and
the Alpine National Park. Schultz Track, where
we found the marbled Euastacus, is in the
East Branch of the upper Buffalo River, about
a kilometre above where the East and West
Buffalo join (Fig. 2).

Collection of specimens. Specimens were coll-
ected using drop nets, bait nets, dip nets and
by hand. Concerns for the conservation of this
population meant that only two individuals
could be retained from each site, so some
animals were released after a portion of a leg

was removed for DNA analysis. Because they
can grow their legs back, this allowed us to get
genetic samples without killing the animals. It
did, however, limit the number of specimens
available for taxonomic work. Some specimens
were collected and released live without remov-
ing any legs. They were used to record the
distribution of crayfish in the river (Fig. 2) but
are not part of the material examined.

Material collected but not retained. Oierax destructor:
Buffalo River, Manna Gum Campsite, VIC, (36°50'S,
146°39'E), 5 Mar. 2002, G. Edney, 2 .

Euastacus armatus. Buffalo River, 1.5 km upstream of
Schultz Track, VIC (37°00'S, 146°49'E), 1 May 2007,
G. Edney, 2 , 9.

Taxonomic and morphometric examinations.

A total of 45 crayfish were examined for 38
characteristics and 15 measurements that were
turned into ratios following to Morgan (1986, 1987
1997). Twenty eight of these were E. armatus,
seven were E. woizouru, three were £. crassus,
one was E. reiki and six were marbled Euastacus
from the East Buffalo River. A Categorical
Principal Component Analysis (CATPCA) was
performed using SPSS version 15 to determine
which characters best distinguished species.

Material examined for morphological characters.

Currently in the research collection in the Department
of Environmental Management and Ecology, but
will be deposited in the Museum of Victoria after
current ongoing research is completed.

Euastacus armatus. Tallangatta Creek, VIC, (36°17'S,
147°33'E), 20 Apr. 1995, S. Lawler, 5 2 ; Koetong

Creek, VIC, (36°06'S, 147°27'E), 15 May 1995, G.
Closs, ,_J, 3 . . ; Hinces Creek, Burrowa Pines N.P.,
NSW, (36°05'S, 147°46'E), 26 April 1995, G. Closs
& M. Shirley, 5 (J, 4 2; Murray River, Bamawartha
VIC, (36°02'S, 146°45'E), 11 July 1995, J. Sloan, , ; King
River, Oxley, VIC, (36°27'S, 146°22'E), 24 June 1995, M.
Versteegen, J; Tumbarumba Creek, Tumbarumba,
NSW, (35°51'S, 148°02'E), 9 July 1995, M. Versteegen, 2
cJ, 9; Nug Nug, Buffalo River, VIC, (36°40'S, 146°41'E),
6 Dec. 1996, P. Suter, V; Ovens River, VIC, (36°02'S,
146°11'E), Sep. 1999, B. Holloway, Manna Gum
Campsite, Buffalo River, VIC, (36°5()'S, 146°39'E), 5 Mar.
2002, G. Ednev, a; Schultz Track, East Buffalo River,
VIC, (36°59'S,T46°48'E), 10 Mar. 2002, G. Edney, <J.

Memoirs of the Queensland Museum 1 Nature • 2010 • 55(1)

215

Street, Edney, Rowe & Lawler

Euastacus woiwuru. Rollason's Falls, Mt. Buffalo NP,
VIC, (36°42'S, 146°47'E), 15 Feb. 2000, M. Chapman,
cJ; Dobson's Creek, Fern Tree Gully, VIC, (37°52'S,
145°19'E), 23 Mar. 2002, K. Sewell, J; West Buffalo
River, VIC, (37°02'S, 146°46'E), 23 Nov. 2006, G.
Edney, 2 rj; East Buffalo River, 1.5 km upstream of
Schultz Track, VIC, (37°00'S, 146°49'E), 30 Apr. 2007,
G. Edney, East Buffalo River, 4.5 km upstream
of Schultz Track, VIC, (37°01'S, 146°49'E), 15 Jul.
2007, G. Edney, £; Dandongadale River, near Lake
Cobbler, VIC, (37°01'S, 146 S 37'E), 30 Aug. 2007, G.
Edney, 9-

Euastacus rieki. Tumbarumba Creek, Tumbarumba,
NSW, (35°51'S, 148°02'E), 9 July 1995, M. Versteegen,

Euastacus crassus. Basalt Hill, Falls Creek, VIC, 17
Jan. 2006, D. Heinze, Native Dog Flat, Buchan
River, VIC, (36°90'S, 148°09'E), 26 Feb. 2000, G.
Edney, <J; Tributary of Big River, Dartmouth Dam,
VIC, (36°39'S, 147°18'E), 26 Jan. 2007, G. Edney, V.

Marbled Euastacus. Schultz Track, East Buffalo River,
VIC, (36°59'S, 146°48'E), 10 Mar. 2002, G. Edney,
QMW 26596, KC2654, J; Schultz Track, East Buffalo
River, VIC, (36°59'S, 146°48'E), 10 Mar. 2002, G. Edney,
9; Schultz Track, East Buffalo River, VIC, (36°59'S,
146°48'E), 10 Jul. 2002, G. Ednev, 2; Schultz Track,
East Buffalo River, VIC, (36°59'S, 146°48'E), 21 Apr.
2007, G. Edney & S. Lawler, 9; Schultz Track, East
Buffalo River, 'VIC, (36°59'S, 146°48'E), 22 Apr. 2007,
G. Edney & S. Lawler, 2 2-

Genetic analysis. The mitochondrial genes COI
and 16S were used because they have been used
extensively to clarify taxonomy and examine
evolutionary processes in freshwater crayfish
(Crandall et al. 1995, 1999; Versteegen & Lawler,
1997; Lawler & Crandall 1998; Hughes & Hillyer
2003; Austin et al. 2003; Munasinghe et al. 2003;
Shull et al. 2005; Gouws et al. 2006; Ponniah
& Hughes 2004, 2006). These gene regions have
been used to find cryptic species in other fresh-
water macroinvertebrates (Chenoweth & Hughes
2003; Baker et al. 2004).

DNA was extracted from tissue (usually gill or
a bit of a leg) using guanidium iso-thiocyanate
(GIT) buffer and a phenol-chloroform extraction
as in Crandall et al. (1995). The DNA was resus-
pended in lOOpI, two microlitres of which was

used as the template for a polymerase chain
reaction (PCR).

Two different PCR products were amplified
from the mitochondrial genome: 720 base pairs
of the cytochrome c oxidase subunit 1 (COI)
and 503 base pairs of the 16S rRNA. COI was
amplified using the LCO1490 and HC02198
primers from Folmer et al. 1994. The 16S rRNA
fragment was amplified using the 16sL and
1472 primers from Shull et al. 2005. A BioRad
PTC-0200 DNA Engine Peltier Thermal Cycler
was used to amplify the DNA, with details of
the reaction mixtures and temperature profiles
available from Street 2007. PCR products were
sent to Macrogen Inc. (Seoul, Korea) for single
extension DNA sequencing.

Material used for Mitochondrial DNA Amplification.

Euastacus armatus. Arml= Manna Gum Campsite,
Buffalo River, VIC, (36°50'S, 146°39'E), 5 Mar. 2002, G.
Edney, <J; Arm2 = Ovens River, VIC, (36°02'S, 146°1TE),
Sep. 1999, B. Holloway, , ; Arm3 = Brad Betts Memorial,
Yackandandah Creek, VIC, (36°20'S, 146°48'E), 29 Jun.
2007, G. Ednev; Arm4 = Buffalo River, VIC, 10 Mar.
2002, G. Edney, QMW26582, KC2653.

Euastacus woiwuru. Woil = East Buffalo River, 1.5
km upstream of SchultzTrack, VIC, (37°00'S, 1 46°49' E),
30 Apr. 2007, G. Ednev, <}; Woi2 = West Buffalo River,
VIC, (37°02'S, 146°46'E), 23 Nov. 2006, G. Edney,
Woi3 = East Buffalo River, 4.5 km upstream of Schultz
Track, VIC, (37°01'S, 146°49'E), 15 Jul. 2007, G. Ednev, 2;
Woi4 = Dobson's Creek, Fern Tree Gully, VIC, (37°52'S,
145°19'E), 23 Mar. 2002, K. Sewell, <J; Dandongadale
River, near Lake Cobbler, VIC, (37°01 'S, 146°37’E), 30
Aug. 2007, G. Edney, Woi5 = West Buffalo River,
VIC, (37°02'S, 146°46'E), 23 Nov. 2006, G. Edney, J.

Euastacus crassus. Cral = Native Dog Flat, Buchan River,
VIC, (36°90'S, 148°09'E), 26 Feb. 2000, G. Edney, <J;
Cra2 = Native Dog Flat, Buchan River, VIC, (36°90'S,
148°09'E), 19 Mar. 2002, G. Edney, KC2649, J; Cra3
= Native Dog Flat, Buchan River, VIC, (36°90'S,
148°09'E), 20 Mar. 2002, G. Edney, KC2720. Euastacus
yarraensis. Yar 1 = Love Creek, ViC, (38°48'S, 143°58'E
), 1 Jan. 2004, K. Sewell & G. Edney, KC2831; Yar 2
= Cockatoo, VIC (37°94'S, 145°49'E) 21 Mar. 2002,
KC2651.

Marbled Euastacus. Unkl = Schultz Track, East
Buffalo River, VIC, 10 Jul. 2002, G. Ednev, ?; Unk2
= Schultz Track, East Buffalo River, ViC, 10 Mar.

216

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Murray Crayfish, Euastacus armatus

0.01

FIG. 3. Neighbour-joining consensus tree using 16S and COI mitochondrial DNA sequences. Interior
branch test probabilities are shown on nodes with bootstrap values shown in parentheses, both using 100,000
replications. Arm = £. armatus, Unk = Marbled Euastacus, Woi = £. woiwuru, Yar = £. yarraensis, Cra = E. crassus
and the outgroup is Euastacus australasieusis. For information on collection sites see methods section.

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

217

Street, Edney, Rowe & Lawler

TABLE 1. Some morphological characters of the Euastacus of the Buffalo River, Victoria. The characters for
E. armatus and E. woiwuru agree with Morgan (1986), who examined many more individuals.

Marbled Euastacus

E. armatus

E. woiwuru

Number examined

TAP

4-5

5-7

7-9

Urocardiac Ridge

9-10

9-11

Marginal mesal
dactvlar spines

0-1

0-3

0-1

Dorsal mesal
dactylar spines

0-5

0-3

2-5

i *■

Mesal carpal spines

Male cuticle partition

yes

Telsonic spines

yes

2002, G. Edney, QMW 26596, KC2654, <$; Unk3 =
Schultz Track, East Buffalo River, VIC, 22 Apr. 2007,
G. Ednev & S. Lawler, 2; Unk4 = East Buffalo River,
1.5 km upstream of Schultz Track, VIC, (37°00'S,
146°49'E), 30 Apr. 2007, G. Edney & D. Street (leg
only); Unk5 = East Buffalo River, 1.5 km upstream of
Schultz Track, VIC, 1 Mav 2007, G. Edney & D. Street
(leg only); Unk6 = East Buffalo River, Schultz Track,
VIC, 22' Apr. 2007, G. Edney & S. Lawler; Unk7 =
East Buffalo River, Schultz Track, VIC, 22 Apr. 2007,
G. Edney & D. Street (leg only); Unk8 = Schultz
Track, East Buffalo River, VIC, (36°59'S, 146°48'E), 21
Apr. 2007, G. Edney & S. Lawler, 2; Unk9 = Schultz
Track, East Buffalo River, VIC, 3 July 2007, G. Edney
& D. Street, (leg only).

Phylogeny construction. DNA sequences were
aligned using Cluster W in the computer
program MEGA Version 4.0 (Kumar et al., 2004;
Tamura et al., 2007). Only specimens that were
successfully sequenced for both mitochondrial
gene regions were used for phylogeny con-
struction. Euastacus australasiensis (KC2637) was
used as an outgroup, and other sequences were
included for comparison, including the marbled
Euastacus KC2654, E. armatus KC2653, E. crassus
KC2720 and E. yarraetisis KC281 and KC 2651
(Shull et al. 2005).

A consensus tree for the two gene regions
was constructed using the neighbour joining
method in MEGA. Two different probabilities,
the bootstrap and interior branch test, were mea-
sured for each node on the phylogeny (Fig. 3).

RESULTS

Distribution. We found a total of 10 crayfish
in the Buffalo River that, while clearly belonging
to the genus Euastacus, did not resemble any
known species. Three of these 'marbled'
Euastacus were collected in 2002 and seven in
2007. Four of the animals caught in 2007 were
released on site after removing a portion of
a leg for genetic analysis. All of these animals
were found in a stretch of river only a few
kilometres long in the East Buffalo River (Fig. 2).

Euastacus armatus are widely distributed in the
Buffalo River below the lake, and in the Ovens
River into which the Buffalo River empties.
They were also found near Dandongadale (the
locality), at Schultz Track, and 1.5 km upstream
of Schultz Track, where they occurred in sym-
patry with E. woiwuru.

Euastacus woiwuru were found on top of
Mount Buffalo, where they had not been
previously recorded, but were not found near
Dandongadale where they had been recorded
in 1982 (Morgan 1986). They were also found
at Schultz Track and 1.5 km, 4.5 km, and 6.8
km upstream from Schultz Track in the East
Buffalo River. They were also found in the West
Buffalo River and in the upper Dandongadale
River, so their distribution is in the headwaters

218

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Murray Crayfish, Euastacus armatus

of the catchment (See Fig. 2). Both E. armatus
and E. woiwuru are found in sympatry with the
marbled Euastacus 1.5 km upstream of Schultz
Track.

The common yabby, Cherax destructor, was
found in the Buffalo River, 6 km south of
Dandongadale (Fig. 2), and Engaeus burrows are
common throughout the catchment. We have
not dug up any burrows or actively sampled the
burrowing crayfish, but the species is probably
Engaeus cymus (Clark 1936) (Horwitz 1990b).

During the taxonomic examinations of old
collections, we found another case of Euastacus
armatus in sympatry with another species
of Euastacus. A collection from a single site
(Tumbarumba Creek, NSW in 1995) was found
to contain both E. armatus and Euastacus reiki
(Morgan 1997). These species both have white
claws and similar spination, highlighting the
degree to which different species can appear
similar in this genus. One of the characters
used to distinguish E. reiki from E. armatus is
the TAP count.

Morphological examinations. A total of 45
Euastacus specimens were scored for 38 character-
istics and 15 measurements that were turned
into ratios according to Morgan (1986, 1987,
1997). Our ratios fell well within the range of
Morgan's data for all the species examined, and
hierarchical cluster analyses in SPSS grouped
the animals into species clusters (Street 2007).
The CATPCA test in SPSS produced a list of
morphological characters that were most useful
in differentiating these species of Euastacus.
A shorter list emphasising the morphological
differences in the Schultz Track population is
shown in Table 1.

The characters used to distinguish the three
Euastacus species were the male cuticle partition,
telsonic spines, mesal carpal spines and the TAP
(Morgan, 1986, 1997). The number of teeth anterior
to the posterior margin of the zygocardiac ossicle
ear, or TAP, is a morphological character that

was described by Francois (1962), is considered
to be a useful character for crayfish taxonomy
(Growns & Richardson 1990) and has been used
in the description of Euastacus species (Morgan
1986; 1988; 1997). The TAP of E. armatus is
between 5 and 7, while the range of TAP in E.
woiwuru is between 7 and 9 (Morgan 1986).

The Schultz Track population, or the marbled
Euastacus, had TAPs at or below the range of
E. armatus, with at least half of them having
five teeth in one ear and four in the other.
These assymetrical individuals were given a
TAP score of 4.5, while somel individuals had
four teeth in both ears, which is outside of the
published range of E. armatus (Table 1).

All of the six marbled specimens examined
for morphology were small, with occipital cara-
pace lengths (OCLs) less than 44 (28, 28, 28, 29,
37 43). The larger ones had white on the tips
of their claws. It is possible that they develop
white claws as they grow larger. E. armatus also
change from green or brown claws to white
as they grow, but the size where this change
occurs seems to vary between populations (pers.
obs., SL).

Genetic analysis and phylogeny. Phylogenies
constructed for the two mitochondrial DNA
regions (16S and COI) produced the same
topology. The consensus tree is shown in Fig. 3.
The phylogeny shows all the marbled Euastacus
in the same clade with E. armatus. According
to the mitochondrial DNA phylogeny, the
unusual Euastacus is not a separate species from
E. armatus, however, DNA analysis using nuclear
genes (Ji et al. 2003) may yet show that this
population is different.

DISCUSSION

An unusual population of crayfish of the
genus Euastacus was found in a very short
stretch of river. This population is designated
the marbled Euastacus in this paper, and its

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

219

Street, Edney, Rowe & Lawler

taxonomic status remains unclear. The location
of this population between two different species
(Fig. 2) suggests that it could be a hybrid.

Genetic sequences mitochondrial DNA genes
16S and COI were used to test the hypothesis
of a hybrid zone. These genes were chosen
because they have been used for many studies of
speciation and population structure of Australian
freshwater decapods (Crandall et aL, 1995; Hughes
and Hillyer, 2003; Chenoweth & Hughes, 2003;
Shull et aL, 2005; Schultz et al., 2007).

The use of the mitochondrial gene cytochrome
oxidase (COI) to identify species is also known
as barcoding (Mitchell 2008). Although it has
been used to identify crvptic species in Crustacea
(Witt et al. 2006) and in insect groups (Rubinoff &
Sperling 2004; Hebert et al. 2004; Smith et al. 2006),
there are also instances in which COI has been
unable to distinguish between species (Meier
et al. 2006; Hickerson et al. 2006; Whitworth et
al. 2007). Nevertheless, some authors claim it
is effective for measuring hybridisation events,
defining hybrid zones and discovering cryptic
species (Rubinoff & Holland 2005).

Failures to resolve species boundaries using
mitochondrial genes are more likely when the
species have recently diverged, or in cases of
inter-specific hybridisation (Shaw 2002). Nelson
et al. (2007) were able to identify nine species of
blowflies using DNA barcoding, but misidentified
the one hybrid specimen using this techinique.

Our data clearly separated known species
of the genus Euastacus, with £. armatus and E.
woiwuru forming well defined monophyletic
groups (Fig. 3). The marbled Euastacu s fell within
the E. armatus clade using consensus sequences
of the mitochondrial genes COI and 16S. This
pattern is suggestive rather than conclusive,
however, because mitochondrial DNA is inherited
directly from the mother without undergoing
recombination. If the mothers of the hybrids
consistently belong to the species E. armatus,
this pattern would still occur. Behavioural

constraints could result in a bias during inter-
specific matings, particularly if the animals
involved have a consistent size difference. For
example, we know that small males can mate
with larger females, and the species E. armatus
usually grow larger than E. woiwuru (Morgan
1986).

Some authors say that E. armatus is not found
in sympatry with other members of its genus
(Gilligan et al. 2007). This perception may be
due to the lack of sampling in areas of potential
overlap, because we have found E. armatus
and E. woiwuru at the same site in the Buffalo
River, and we also identified at least one site
in New South Wales where E. armatus is found
in sympatry with E. rieki. We did not recognise
the sympatry at the time of collection (in 1995)
because £. rieki had not yet been described
(Morgan 1997). Euastacus reiki and E. armatus
are so similar morphologically that we have to
dissect the gastric mill to tell them apart, which
makes them very hard to distinguish in the field.

Almost 30 years ago, Euastacus woiwuru were
found near Dandongadale, but only Murray
Crays (E. armatus) are found there now. Euastacus
woiivuru are now found about 20 kms south of
Dandongadale in the East and West branches of
the Buffalo River. The apparent movement of the
Murray Cray upstream could be explained by
the alteration of habitats by human disturbance
or by the warming of the river due to climate
change or many years of drought. Either way,
the change in distribution may indicate that
the species boundary between E. armatus and
£. woiwuru has been moving for decades. The
fact that an unusual morphological variant is
found at this boundary is highly suggestive of
a hybrid zone. However, if they are hybrids,
they do not appear to form a self sustaining
breeding population, because both parent
species are present at the localities where the
marbled form is found.

220

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Murray Crayfish, Euastacus armatus

Marbled Euastacus are morphologically distinct
in their colouration and the fact that some of
their TAPs (the number of teeth anterior to the
posterior margin of the zygocardiac ossicle ear)
are below the published range of E. armatus and
E. woiwuru. Interestingly, Morgan states that
the TAPs of E. woiwuru are generally lower in
Murray River tributaries like the Buffalo River,
and that a specimen collected at Dandongadale
was unusual (Morgan 1986, p. 49). Perhaps this
location, which is at the edge of the distribution of
both species, contains some unique morpho-
logical variants.

The marbled Euastacus were all very small,
with OCLs (occipital carapace length) below 44
mm. However, since the larger of tine marbled
Euastacus had white tips on their claws, it is
possible that the marbled crays are a unique
juvenile variant, and that these animals grow
into adult Murray Crays with white claws and
a uniformly coloured carapace. Normally,
juvenile Murray Crays do not have the white
claws that are a distinctive feature of the adults,
but neither do they have a marbled pattern.
The substrate at this site did not obviously
differ from other stretches of the river where
E. armatus juveniles do not show this marbled
pattern, but it is possible that there is a yet to be
identified environmental factor affecting their
colour. Because our permit did not allow us to
retain and raise these animals, we were unable
to confirm the adult colouration of the marbled
specimens.

Many threatening processes are controllable,
but for measures to be invoked toward the
management of crayfish populations, species
have to be recognised as threatened and in need
of protection (Merrick 1997). The sedentary
nature of crayfish and limited gene flow
between catchments (Fetzner & Crandall 2001;
Gouin et al. 2006) makes them susceptible to
over-fishing and habitat alteration (Merrick 1997;
O'Brien 2007). The first step toward listing a
population is to determine if it is different enough

to be considered a conservation unit (Fraser &
Bernatchez 2001).

Unique fauna are often found in unregulated
headwater streams (Baker et al. 2004), and
this study provides an example of an unusual
population of spiny crayfish. The conservation
status of this population needs to be clarified.
The location is subject to several threatening
processes. Cows are allowed to roam freely in
and out of the river, and fires and fire control
measures combined to significantly disturb the
site in 2003 and 2006. Given the fire history of
the area, we can expect reduced water flows in
the catchment for decades.

Speciation is a slow process in this genus
(Ponniah & Hughes 2006). If we have found an
unusual population of E. armatus, it could still be
different enough to be deserving of conservation.
On the other hand, finding this crayfish at the
boundary of two Euastacus species means that
the possibility of hybridisation must be taken
seriously, particularly since introgression is a
possibility (Ballard & Whitlock 2004; Funk &
Omland 2003; Shaw 2002).

We are planning to sequence nuclear genes
using ITS primers (Ji et al. 2003) and will continue
to make field observations. Investigation of fresh-
water crayfish species boundaries in the Buffalo
River may enlighten us about the ecology and
evolution of Australian spiny freshwater crayfish.

ACKNOWLEDGEMENTS

This research was supported by a series
of permits from Victorian Fisheries (RP906,
RP562, RP751), New South Wales Fisheries
(FSP/CW/135), Victorian Department of Sus-
tainability and Environment (10002702 and
10004059) and by animal ethics permits from
La Trobe University (AEC Invert/03-4(W) and
AEC07/21(W)).

We would like to thank Christine Street, Susan
Street, Robbie Schaeffer, Kim Sewell, Ryan

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

221

Street, Edney, Rowe & Lawler

Sewell, David Blair, Phil Suter, Ben Holloway,
Mick Chapman, Gerry Closs, Michael Shirley,
Mardi Versteegen, Dean Heinze and Dale McNeil
for help with field work and collections. We would
also like to thank Warren Paul for assistance with
statistical analyses, Leo McGuire for the use of
his cabin, and Peter Davies for comments on an
earlier version of this manuscript.

LITERATURE CITED

Austin, C.M., Nguven, T.T.T., Meewan, M.M. &
Jerry, D.R. 2003. The taxonomy and phylogeny
of the ' Cherax destructor’ complex (Decapoda:
Parastacidae) examined using mitochondrial 16S
sequences. Australian Journal of Zoology 51: 99-110.

Baker, A.M., Hughes, J.M., Dean, J.C. & Bunn, S.E.
2004. Mitochondrial DNA reveals phylogenetic
structuring and cryptic diversity in Australian
freshwater macroinvertebrate assemblages.
Marine and Freshwater Research 55: 1126-1130.
Ballard, J.W. & Whitlock, M.C. 2004. The incomplete
natural history of mitochondria. Molecular
Ecology 13: 729-744.

Chenoweth, S.F. & Hughes, J.M. 2003. Speciation
and phylogeograpny in Caridina indistincta, a
complex of freshwater shrimps from Australian
heathland streams. Marine and Freshwater Research
54: 807-812.

Clark, E. 1936. The freshwater and land crayfishes
of Australia. Memoirs of the National Museum of
Victoria 10: 5-58.

' 1941. Revision of the genus Euastacus (Crayfishes,
Family Parastacidae), with notes on the
distribution of certain species. Memoirs of the
Natural History Museum of Victoria 12: 7-30.
Crandall, K.A., Lawler, S.H. & Austin, C.M. 1995.
A preliminary examination of the molecular
phylogenetic relationship of the crayfish genera of
Australia (Decapoda: Parastacidae). Freshwater
Crayfish 10: 18-30.

Crandall, K.A., Fetzner, Jr., J.F., Lawler, S.H.,
Kinnerslev, M. & Austin, C. M. 1999. Phylogenetic
relationships among the Australian and
New Zealand genera of freshwater crayfishes
(Decapoda: Parastacidae). Australian Journal of
Zoology 47: 199-214.

Coughran, J. 2002. A new spec : es of the freshwater
crayfish genus Euastacus (Decapoda: Parastacidae)

from north-eastern New South Wales, Australia.
Records of the Australian Museum 54: 25-30.

2005. New crayfishes (Decapoda: Parastacidae)
from north-eastern New South Wales, Australia.
Records of the Australian Museum 57: 361-374.

Edney, G.N., Mcneil, D.G. & Lawler, S.H. 2002. The
Swamp Yabby ( Cherax sp.) of the Murray River
Catchment. The Victorian Naturalist 119(4)-
200-204.

Fetzner, J.W., Jr. & Crandall, K.A. 2001. Genetic
variation. In Chapter 10 Biology of Freshwater
Crayfish. (D.M. Holdich, Ed.), Pp. 291-326.
Blackwell Science, Oxford.

Fetzner, J. W., Jr. & Crandall, K.A. 2003. Linear habitats
and the nested clade analysis: an empirical
evaluation of geographic vs. river distances using
an Ozark crayfish (Decapoda: Cambaridae).
Evolution 57: 2102-2118.

Folmer, O., Black, M„ Hoeh, W„ Lutz, R. & Vrijenhoek,
R. 1994. DNA primers for the amplification of
mitochondrial cytochrome c oxidase subunit I
from diverse metazoan invertebrates. Molecular
Marine Biology and Biotechnology 3: 294-299.

Francois, D. D. 1962. A revision of the Australian
genus Euastacus (Decapoda: Parastacidae). Ph.D.
Thesis, Cornell University, Ithaca, New York.

Fraser, D.J. & Bernatchez, L. 2001. Adaptive
evolutionary conservation: towards a unified
concept for defining conservation units. Molecular
Ecology 10: 2741-2752.

Funk, D.J. & Omland, K.E. 2003. Species - level
paraphyly and polvphyly: frequency, causes, and
consequences, with insights from mitochondrial
DNA. Annual Review of Ecology and Evolution
Systematics 34: 397-423.

Gilligan, D., Rolls, R., Merrick, J., Lintermans, M„
Duncan, P. & Kohen, J. 2007. Scoping the knowledge
requirements for Murray crayfish ( Euastacus armatus).
(NSW Department of Primary Industries:
Narrandera).

Gouin, N„ Grandjean, F. & Souty-Grosset, C. 2006.
Population genetic structure of the endangered
crayfish Austropotamobius pallipes in France based
on microsatellite variation: biogeographical
inferences and conservation implications.
Freshwater Biology 51: 1369-1387.

Gouws, G., Stewart, B.A. & Daniels, S.R. 2006.
Phylogeographic structure of a freshwater
crayfish (Decapoda: Parastacidae: Cherax
pretssii) in south-western Australia. Marine and
Freshwater Research 57: 837-848.

222

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Murray Crayfish, Euastacus armatus

Growns, I. & Marsden, T. 1998. Altitude separation
and pollution tolerance in the freshwater
crayfish Euastacus spinifer and E. autraliasiensis
(Decapoda: Parastacidae) in coastal flowing
streams of the Blue Mountains, New Soutn
Wales. Proceedings of the Linnean Society of
Neic South Wales 120: 139-145.

Growns, 1. & Richardson, A. 1990. A comparison of
the gastric mills of nine species of Parastacid
crayfish from a range of habitats, using multi-
variate morphometries (Decapoda, Astacoidae).
Crustaceana 58: 33-44.

Harvey, M. 2002. Short range endemism in the
Australian fauna: some examples from non-
marine environments. Invertebrate Systematics
16: 555-570.

Hazlett, B.A., Lawler, S.H. & Edney, G. 2007. Agnostic
behaviour of the crayfish Euastacus armatus and
Cherax destructor. Journal of Marine and Fresh-
water Behaviour and Physiology 40(4): 257-266.

Hebert, P.D.N., Penton, E.H., Bums, J.M, Janzen,
D.H. & Hallwachs, W. 2004. Ten species in
one: DNA barcoding reveals cryptic species
in the neotropical skipper butterfly Astraptes
fulgerator. Proceedings of the National Academy
of Science of the United States of America 101(41):
14812-14817.

Hickerson, M.J., Meyer, C.P. & Mortiz, C. 2006. DNA
barcoding will often fail to discover new animal
species over broad parameter space. Systematic
Biology 55: 729-739.

Heller, C. 1865. Crustacea. Reise 'Novara' 2: 100-104.

Horwitz, P. 1990a. The conservation status of Australian
freshwater Crustacea, with a provisional list of
threatened species, habitats and potentially
threatening processes. Australian National Parks
and Wildlife Service, Canberra, Report Series
No. 14.

1990b. A taxonomic revision of species in the
freshwater crayfish genus Engaeus Erichson
(Decapoda: Parastacidae). Intertebrate Taxonomy 4:
427-614.

Hughes, J.M. & Hillyer, M.J. 2003. Patterns of
connectivity' among populations of Cherax
destructor (Decapoda: Parastacidae) in western
Queensland, Australia. Marine and Freshwater
Research 54: 587-596.

Ji, Y.J., Zhang, D.X. & He, L.J. 2003. Evolutionary
conservation and versatility of a new set of
primers for amplifying the ribosomal internal
transcribed spacer region in insects and other
invertebrates. Molecular Ecology 3: 581-585.

Kumar, S., Tamura, K. & Nei, M. 2004. MEGA 3:
integrated software for molecular evolutionary

§ enetics analysis and sequence alignment.
riefings in Bioinformatics 5: 150-163.

Lawler, S.H. & Crandall, K.A. 1998. The relationship
of the Australian freshwater crayfish genera
Euastacus and Aslacopsis. Proceedings of the Linnean
Society ofNeu> South Wales 119: 1-8.

McCarthy, B. 2005. Distribution of Murray crayfish
(Euastacus armatus) in the Mallee region 2004.
(Murray-Darling Freshwater Research Centre,
Lower Basin Laboratory, Mildura).

Meier, R„ Shiyang, K„ Vaidya, G. & Ng, P. K. L. 2006.
DNA barcoding and taxonomy in Diptera: a
tale of high intraspecific variability and low
identification success. Systematic Biology 55:
715-728.

Merrick, J.R. 1997. Conservation and field management
of the freshwater crayfish, Euastacus spinifer
(Decapoda: Parastacidae), in the Sydney Region,
Australia. Proceedings of the Linnean Society of New
South Wales 118: 217-225.

Milne Edwards, H. 1837. Histoire naturelle des
crustaces, comprenant l'anatomie, la physiologie
et la classification de ces animaux. Librairrie
Encyclopediquc de Roret. 2: 1-57.

Mitchell, A. 2008. DNA barcoding demystified.
Australian Journal of Entomology. 47 (3) 169-173.

Morey, J. & Hollis, G. 1997. Australia's most diverse
crayfish habitat? Memoirs of the Museum of
Victoria. 56(2): 667-669.

Morgan, G. J. 1986. Freshwater crayfish of the genus
Euastacus Clark (Decapoda: Parastacidae) from
Victoria. Memoirs of the Museum of Victoria.
47: 1-57.

1988. Freshwater crayfish of the genus Euastacus
Clark (Decapoda: Parastacidae) from Queensland.
Memoirs of the Museum of Victoria. 49: 1-49.

1997. Freshwater crayfish of the genus Euastacus
Clark (Decapoda: Parastacidae) from New South
Wales, with a key to all species in the genus.
Records of the Australian Museum 23: 1-110.

Munasinghe, D., Murphy, N. & Austin, C.M. 2003.
Utility of mitochondrial DNA sequences from
four gene regions for systematic studies of
Australian freshwater crayfish of the genus Qierax
(Decapoda: Parastacidae). Journal of Crustacean
Biology 23(2): 402- 417.

Nelson, L.A., Wallman, J.F. & Dowton, M. 2007. Using
COI barcodes to identify forensically and

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

223

Street, Edney, Rowe & Lawler

medically important blowflies. Medical and
Veterinary Entomology 21: 44 - 52.

Nguyen, T.T.T., Austin, C.M., Meewan, M.M., Schultz,
M.B.A. & Jerry, D.R. 2004. Phylogeographv of
the freshwater crayfish Cherax destructor Clark
(Parastacidae) in inland Australia: Historical
fragmentation and recent range expansion.
Biological Journal of the Linnean Society 83: 539-550.

O'brien, M. 2007. Freshwater and terrestrial crayfish
(Decapoda, Parstacidae) of Victoria: status,
conservation, threatening processes and
bibliography. The Victorian Naturalist. 124(4):
210-229.

Ponniah, M. & Hughes, J.M. 2004. The evolution of
Queensland spiny mountain crayfish of tire genus
Euastacus. I. Testing vicariance and dispersal with
interspecific mitochondrial DNA. Evolution 58(5):
1073-1085.

Ponniah, M. & Hughes, J. M. 2006. The evolution
of Queensland spiny mountain crayfish of the
genus Euastacus. II. Investigating simultaneous
vicariance with intraspecific genetic data.
Marine and Freshwater Research 57: 349-362.

Riek, E. F. 1969. The Australian freshwater crayfish
(Crustacea: Decapoda: Parastacidae), with
descriptions of new species. Australian Journal
of Zoology 17: 855-918.

Rubinoff, D. & Holland, B.S. 2005. Between two
extremes: mitochondrial DNA is neither the
panacea nor the nemesis of phylogenetic and
taxonomic influence. Systematic Biology 54(6):
952-961.

Rubinoff, D. & Sperling, F.A.H. 2004. Mitochondrial
DNA sequence, morphology and ecology yield
contrasting conservation implications for two
threatened buckmoths (Hemileuca: Saturniidae).
Biological Conservation 118: 341-351.

Schultz, M.B., Smith, S.A., Richardson, A.M.M.,
Horwitz, P., Crandall, K.A. & AUSTIN,
C.M. 2007. Cryptic diversity in Etigaeus
Erichson, Geocharax Clark and Gramastacus
Riek (Decapoda: Parastacidae) revealed by
mitochondrial 16S rDNA sequences. Invertebrate
Systematics 21: 569-587.

Shaw, K.L. 2002. Conflict between nuclear and
mitochondrial DNA phylogenies of a recent

species radiation: what mtDNA reveals and
conceals about modes of speciation in Hawaiian
crickets. Proceedings of the National Academy of
Science 99(25): 16122-16127.

Shull, H.C., Perez-Losada, M., Blair, D., Sewell, K.,
Sinclair, E.A., Lawler, S., Ponniah, M. & Crandall,
K.A. 2005. Phylogeny and biogeography of
the freshwater crayfish Euastacus (Decapoda:
Parastacidae) based bn nuclear and mitochondria]
DNA. Molecular Phylogenetics and Evolution 37:
249-263.

Smith, G. 1912. Tire freshwater crayfishes of Australia.
Proceedings of the Zoological Society of London.
1912: 144-171.

Smith, M.A., Woodley, N.E., Janzen, D.H., Hallwachs,
W. & Hebert, P.D.N. 2006. DNA barcodes reveal
cryptic host-specificity within the presumed
polyphagous members of a genus of parasitoid
flies (Diptera: Tachinidae). Proceedings of the
National Academy of Science of the United'' States of
America 103(10): 3657 - 3662.

Street, D.C. 2007. A review of the status of an unusual
opulation of freshwater crayfish of the genus
uastacus found in the East Buffalo River,
Victoria. Honours Thesis, La Trobe University,
Wodonga, Victoria.

Tamura, K., Dudley, J., Nei, M. & Kumar, S. 2007.
MEGA 4: Molecular evolutionary genetics analysis
(MEGA) software version 4.0. Molecular Biology
and Evolution 10.1093/ molbev/msm092.

Versteegen, M. & Lawler, S.H. 1997. Population
genetics of the Murray River crayfish, Euastacus
armatus. Freshwater Crayfish 11: 146 -157.

Von Martens, E. 1866. On a new species of Astacus.
Annals and Magazine of Natural History 3(17V
359-360.

Whitworth, T.L., Dawson, R.D., Magalon, H. & Baudry,
E. 2007. DNA barcoding cannot reliably identify
species of the blowfly genus Protocalliphora
(Diptera: Calliphoridae). Proceedings of the Royal
Society B 274: 1731 -1739.

Witt, J.D.S., Threloff, D.L. & Hebert, P.D.N. 2006.
DNA barcoding reveals extraordinary cryptic
diversity in an amphipod genus: implications
for desert spring conservation. Molecular Ecology
15: 3073-3082.

224

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

New records of the Japanese devilray Mobula
japanica (Muller & Henle 1841) for Australian waters

Kathy A. TOWNSEND

Moreton Bay Research Station, Centre for Marine Studies, The University of Queensland, 25
Flinders Ave, Dunwich, Qld 4183, Australia. Email: kathy.townsend@uq.edu.au

Peter M. KYNE

School of Biomedical Sciences, The University of Queensland, St Lucia, Qld 4072, Australia: present
address: Tropical Rivers and Coastal Knowledge, Charles Darwin University, Darwin, NT 0909,
Australia.

Citation: Townsend, K.A. & Kyne, P.M. 2010 03 15. New records of the Japanese devilray Mobula
japanica (Muller & Henle 1841) for Australian waters. Memoirs of the Queensland Museum -
Nature 55(1): 225-230. Brisbane. ISSN 0079-8835. Accepted: 2 February 2009.

ABSTRACT

New Australian records of the Japanese devilray Mobula japanica (Muller & Henle 1841)
are documented from photographs taken at Ribbon Reefs, Fraser Island, off Noosa and
off Southport, Qld, and from a specimen beach-washed on North Stradbroke Island,
Qld. This specimen, measuring 2224 mm disc width, was located on 02/09/2007 and
is the first sexually mature male recorded in Australian waters. Morphometries for this
specimen are generally consistent with previously published accounts. Where differences
exist, these are marginal. Within Australian waters, M. japanica has thus far only been
recorded from the east coast, within the latitudinal range 14°57'-32°59'S. The recent
increase in records suggests the species is more common in Australian waters than
previously thought. □ Myliobatoidei, Mobulidae, North Stradbroke Island.

The Japanese devilray Mobula japanica (Muller
& Henle 1841) (Rajiformes: Myliobatoidei:
Mobulidae) is thought to be circumglobal in warm
temperate and tropical waters of the Atlantic,
Pacific and Indian Oceans (Last & Stevens 1994).
However, its detailed distribution, along with its
biology is poorly-known, and it was previously
thought to be rare, or at least rarely documented
in Australian waters (Kyne et al. 2005).

The family Mobulidae is comprised of two
genera: Manta Bancroft, 1829 and Mobula
Rafinesque, 1810, separated by the position of the
mouth; at end of snout tip in Manta, and ventral
on the head and posterior of the snout tip in
Mobula. There are nine recognised valid species
of Mobula worldwide (Notarbartolo-di-Sciara
1987). Distinguishing characters of Mobula japanica
include: elliptical spiracles situated dorsal to the

level of the pectoral fins, white tip to dorsal
fin, very long tail (when unbroken) and the
presence of a caudal spine (Notarbartolo-di-
Sciara 1987; Last & Stevens 1994). This latter
character is shared only with Mobula mobular
(Bonnaterre 1788), which is possibly endemic
to the Mediterranean, although the systematics
concerning the relationship between M. japanica
and M. mobular remain unresolved (Notarbartolo-
di-Sciara 1987).

Four species of mobulids have been confirmed
for Australian waters. Manta birostris (Donndorff
1798) and Mobula eregoodootenkee Garman 1913
are reasonably common in tropical waters, while
Mobula thurstoni (Lloyd 1908) is known from
a single record off Mackay, Qld, and Mobula
japanica from two previous specimens from the
east coast (Last & Stevens 1994; Kyne et al. 2005).

Memoirs of the Queensland Museum | Nature • 2010 • 55(1) • www.qm.qld.gov.au

225

Townsend & Kyne

Mobula japanica is a medium to large devilray,
attaining a disc width (DW) of at least 3100 mm
(Paulin et al. 1982).

The first recorded Australian specimen of M.
japanica was a 1880 mm DW individual collected
by net fisherman, inshore Lake Macquarie, NSW
(32°59’S, 151°35'E) on 04/04/1968 (AMSIB.8021).
The sex is unknown, as only the head was
lodged in the AMS collection, however, from
accompanying original photographs it appears
to be female (Kyne et al. 2005). The second
specimen was a 1088 mm DW immature male
found beach-washed on Eurong Beach, Fraser
Island, Qld (25°31'S, 153°08'E) on 17/08/2000
(QM1.33855) (Kyne et al. 2005). This paper presents
information on a number of new records for
Australian waters, in particular a beach-washed
specimen from North Stradbroke Island, Qld.

METHODS

All records were identified using Notarbartolo-
di-Sciara (1987) and Last & Stevens (1994). The
North Stradbroke Island specimen was reportedly
alive when first located by a member of the
public. Upon collection it was dead but very fresh.
Morphometries and photographs were taken
of the fresh specimen on arrival at the Morelun
Bay Research Station, Dunwich. Thirty-four
• morphometries were recorded, including 29
following the methodology of Notarbartolo-di-
Sciara (1987), and additional measurements of
the claspers and caudal spine. Morphometries
were compared to previously published results
from the E Pacific (Notarbartolo-di-Sciara 1987)
and the Fraser Island specimen (Kyne et al.
2005). The following abbreviations are used for
institutions: AMS, Australian Museum, Sydney;
QM, Queensland Museum, Brisbane.

Following fresh examination, the specimen
was transported to the QM. Space restrictions
precluded fixation and storage of the whole
animal, however, a fibreglass cast of the fresh
whole specimen was made for public display at

the QM. The head (including cephalic fins) and
region posterior from the dorsal fin (including
dorsal fin, pelvic fins, claspers, tail and caudal
spine) were dissected from the whole specimen
and lodged in theQM Ichthyological Collection.
Some sections of gill arches were also preserved,
as well as liver and muscle tissue samples in
alcohol for possible future DNA analysis.

Photographs were obtained of individuals
from a number of locations along the eat coast
of Qld: Fraser Island (beach-washed specimen;
individual not retained), off Southport (live
sightings), Ribbon Reefs (live sighting) and off
Noosa (fisheries bycatch; specimens not retained).

RESULTS AND DISCUSSION

A 2224 mm DW mature male M. japanica
(QMI. 38137) (Fig. 1A-D) was beach-washed
on Flinders Beach, North Stradbroke Island,
Qld, Australia (27°24'59"S, 153°29'12”E) on
02/09/2007. Table 1 presents morphometries
for the specimen. For comparison, those from
the second Australian specimen (Fraser Island,
Qld; QMI.33855; Kyne et al. 2005) and the
range of values for the 29 measurements given
in Notarbartolo-di-Sciara (1987) for specimens
from the E Pacific are also included.

The measured size of the specimen (2224 mm
DW) is within the reported maximum for the
species, which is at least 3100 mm DW (Paulin
et al. 1982; Last & Stevens 1994). White et al.
(2006) estimated size at maturity for male M.
japanica from eastern Indonesian waters (DW w
with confidence intervals) as 2016 mm (1984-
2050 mm) DW. The North Stradbroke specimen
represents the only known mature male of the
Australian records.

Over one third of the 29 morphometric
measurements were outside of the range reported
by Notarbartolo-di-Sciara (1987) for specimens
from the E Pacific (Table 1). Nine of the 11 of these
were above the previously published range, while

226

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

New records of Mobula japanica

TABLE 1. Proportional dimensions as thousandths of disc width for Mobula japanica from North Stradbroke
(QMI.38137) and Fraser Islands (QMI.33855; Kyne et al. 2005), Qld, Australia, and those given by Notarbartolo-
di-Sciara (1987). * denotes proportional measurements which fall outside the previously reported range.

QMI.38137

QMI.33855

Notarbartolo-di-Sciara (1987) Range

Location

North Sradbroke Is.,
Qld

Fraser Is., Qld

Eastern Pacific

1 . Disc width (mm)

2224

1088

1316-2259

| 2. Disc length

544*

445

448-531

3. Anterior projection

354

310

321-379

4. Rostrum to pelvic fin

560

453

457-561

5. Predorsal length

461*

373

373-441

6. Dorsal fin base

39-55

7. Dorsal fin height

35-47

1 8. Precloaca! distance

441*

388

367-430

' 9. Tail length

517*

1245

641-1075

10. 1st gill opening length

47-57

11. 2nd gill opening length

51-62

12. 3rd gill opening length

49-61

13. 4th gill opening length

44-58

14. 5th gill opening length

34-44

15. 1st interbranchial distance

122*

108-119

16. 5th interbranchial distance

46-69

17. Rostrum to 1st gill openings

122*

105

103-121

18. Rostrum to 5th gill openings

225*

190

193-221

19. Pelvic fin length

123

108-131

20. Cephalic fin length

128*

107

117-127

21. Cephalic fin width

34*

52-68

22. Eyeball diameter

21-26

23. Cranial width

188*

161

148-169

24. Preoral length

58*

38-43

25. Head length

153

138

132-163

26. Mouth width

124

103

110-126

27. Intemarial distance

108

99-110

28. Upper toothband length

83-94

29. Lower toothband length

81-96

Clasper length inner

110

Clasper length outer

Clasper width

Dorsal spine length

Ventral spine length

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

227

Townsend & Kyne

two were below (tail length and cephalic fin
width). The tail was damaged, explaining the
lower range for that value. Kyne et al. (2005)
noted the unreliability of this as a diagnostic
feature due to its susceptibility to damage. The
following measurements associated with the
head were all above the previously published
range: cranial width, preoral length and cephalic
fin length, as were disc length and the predorsal
length (Table 1). However, with the exception
of tail length, all differences were marginal (0.1
to 2.0% of DW), and may be related to either
measuring technique (although the present
study followed Notarbartolo-di-Sciara (1987)
very closely), or a limited number of replicates
from the original study (n=19; Notarbartolo-di-
Sciara, 1987). Alternatively, some differences may
reflect regional or population-level variability.
Genetic analysis may be able to shed some light
on the latter possibility.

Table 2 summaries the 13 known records of M.
japcmica for Australia. In addition to the North
Stradbroke Island specimen detailed above,
a number of photographs of M. japmiica were
obtained from Qld waters. Ian Banks filmed an

individual at 6 m depth off Main Beach, Gojj
Coast (27°58'S, 153°25’E) on 13/02/2006 and l ater
filmed two individuals nearby (27"56'S, 153°25'g\
on 11/05/2008 (still photographs were provide
from the video footage). On 03/10/2007, Juli e
Meles photographed a beach-washed individual
1 km north of McLaughlan Rocks on Fraser
Island (25°19'S, 153‘T3’E). The sex of the specimen
was not noted, but it was estimated to measure
3100 mm DW (information provided by J e ff
Johnson, QM and Ivan Thrash, QNPWS). Chri s
Witty photographed an individual on a re e f
between Ribbon Reef #9 and #10 on the Great
Barrier Reef (14°57’S, 145°40'E) at -10-15 m
depth on 01/01/2008 (a total of four devilray s
were sighted, but only one individual could b e
identified as M. japanica) (information provided
by Qamar Schuyler, Undersea Explorer), a
professional fisher caught five specimens from
670 to 1060 mm DW by gill net at 7-12 m depth
on -20/05/2008 just north of the Noosa River
mouth (26°23'S, 153°05’E). All were released and
sex was not recorded (information provided by
Jeff Johnson, QM and S. McCulloch, QDPI & F).

TABLE 2. Summary of Australian records of Mobula japanica.

Record &
specimen no.

(if applicable)

Date of record

Location

Co-ordinates

Collection method

Reference/ source

1. AMSIB.8021

04 Apr 1968

Lake Macquarie, NSW

32°59’S, 151°35'E

Estuarine net
bycatch

Kyne et al. (2005)

2. QMI.33855

17 Aug 2000

Fraser Island, QLD

25°31'S, 153°08'E

Beach-washed

Kyne et al. (2005)

13 Feb 2006

Off Main Beach,

Gold Coast, QLD

27°58’S, 153°25'E

Live sighting
(photographs)

I. Banks

4. QMI.38137

02 Sep 2007

North Stradbroke

Island, QLD

27 <> 25 , S, 153°29’E

Beach-washed

Present study

03 Oct 2007

Fraser Island, QLD

25“19’S, 153°13'E

Beach-washed

(photographs)

J. Meles & I. Thrash

01 Jan 2008

Ribbon Reefs, Great

Barrier Reef, QLD

14°57S, 145"40’E

Live sighting
(photographs)

C. Witty & Q.

Schuyler (Undersea
Explorer)

7-8.

11 May 2008

Gold Coast Seaway, QLD

27°56'S, 153°25'E

Live sighting
(photographs)

I. Banks

9-13.

-20 May 2008

Off Noosa, Sunshine

Coast, QLD

26°23'S, 153°05'E

Gillnet bycatch
(photographs)

S. McCulloch
(QDPI&F)

228

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

New records of Mobula japanica

FIG. 1. Mobula japonica, QMI.38137, 2224 mm DW. A, Whole animal, dorsal view. B, Whole animal, ventral
view. C, Head, lateral view. D, Dorsal fin, caudal spine and tail base, lateral view. (Photos: P. Fugelli).

Mobula japanica is the only devilray species
found within the Indo-Pacific which possesses
a caudal spine. However, Notarbartolo-di-Sciara
(1987) noted that the caudal spine can occasionally
be missing in the species (2.5% of specimens
examined). Furthermore, for M. japanica speci-
mens from New Zealand waters, Paulin et al.
(1982) noted that in many of the individuals
examined, the caudal spine was very small or
broken. As such, when identifying specimens
in the field, the absence of a caudal spine does
not necessarily discount M. japanica. Diagnostic
characters specified in Notarbartolo-di-Sciara
(1987) and Last & Stevens (1994) should be used in
combination to ensure the correct identification
of Indo-Pacific mobulids.

Very little is known about the stock structure
of M. japanica, despite its wide distribution.
Although thought to be circumtropical, the
presently known distribution is disjunct.
The connectivity of regional populations or
stocks, for example, where the species occurs
in the E Pacific and the W Pacific, is unknown.
Even locally within Australia and SE Asia it is
unknown whether there is one interbreeding
stock, or several distinct populations. The latter
would have implications for mobulid fisheries
where they occur regionally in Indonesia and
the Philippines, with the risk of stock depletion
greatly increased.

In Australia, M. japanica has only been recorded
along the east coast, with present records within

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

229

Townsend & Kyne

the latitudinal range 14°57'-32°59'S. The closest
records appear to be from New Zealand where
the species is relatively common in waters off
the North Island during the summer months
to at least 38°22'S (Paulin et al. 1982; Duffy &
Abbott 2003). Since the Japanese devilray is
commonly caught in Indonesian waters (White
et al. 2006), it is expected that it will be recorded
more widely from northern and western
Australian waters in the future. The number
of recent beach-washed records, together with
live sightings and fisheries bycatch, suggests a
more common occurrence off eastern Australia
than previously thought.

ACKNOWLEDGMENTS

We thank Jack Jackson and the Straddie Wild-
care group for notification of tine North Stradbroke
Island specimen, Fabrice Jaine for assistance with
specimen examination and Petter Fugelli for
photography of that specimen, Jeff Johnson
(Ichthyology, QM) for curatorial assistance and
providing information on additional records,
Ian Banks, Julie Meles, Ivan Thrash (QNPWS),
Chris Witty and Qamar Schuyler (Undersea
Explorer), and S. McCulloch (QDPI&F) for

access to photographs and information on addi-
tional records.

LITERATURE CITED

Duffy, C.A.J. & Abbott, D. 2003. Sightings of
mobulid rays from northern New Zealand, with
confirmation of the occurrence of Manta birostris
in New Zealand waters. New Zealand journal of
Marine and Freshwater Research 37: 715-721.

Kyne, P.M., Johnson, J.W., Courtney, A.J. & Bennett
M.B. 2005. New biogeograpHical information
on Queensland chondrichthyans. Memoirs of
the Queensland Museum 50: 321-327.

Last P.R. & Stevens J.D. 1994. Sharks and rays of
Australia. (CSIRO Division of Fisheries: Hobart).

Notarbartolo-di-Sciara, G. 1987. A revisionary
study of the genus Mobula Rafinesque, 1810
(Chondrichthyes: Mobulidae) with the description
of a new species. Zoological journal of the Linnean
Society 91: 1-91.

Paulin, C.D., Habib, G„ Carey, C.L., Swanson, P.M
& Vos, G.J. 1982. New records of Mobula
japanica and Masturus lanceolatus, and further
records of Luvaris imperialis (Pisces: Mobulidae,
Louvaridae) from New Zealand. New Zealand
journal of Marine and Freshwater Research 16: 11-17.

White, W.T., Giles, J., Dharmadi & Potter, I.C. 2006.
Data on the bycatch fishery and reproductive
biology of mobulid ravs (Myliobatiformes)
in Indonesia. Fisheries Research 82: 65-73.

230

Memoirs of the Queensland Museum

Nature • 2010 • 55(1)

A technique for examination of diagnostic
characters of penicillate millipedes

Megan SHORT
Cuong HUYNH

School of Life and Environmental Sciences, Deakin University, 221 Burwood Hwy, Burwood, VIC
3125, Australia. Email: mshort@deakin.edu.au

Citation: Short, M. & Huynh, C. 2010 03 15. A technique for examination of diagnostic characteristics
of penicillate millipedes. Memoirs of the Queensland Museum — Nature 55(1): 231-234. Brisbane.
ISSN 0079-8835. Accepted: 5 August 2009.

ABSTRACT

We describe a simple method for examination of penicillate millipedes. The internal tissues
are dissolved and the stain Fast Green used to give excellent contrast for dissection and
to provide clear visualisation of characters necessary for identification of both freshly
collected and museum specimens. □ Histology, Fast Green stain, Diplopoda, Penicillata,
Polyxenida.

Millipedes in the subclass Penicillata, order
Polyxenida have been overlooked in the study
of terrestrial invertebrate fauna in Australia.
Penicillates are commonly known as bristly,
dwarf or pincushion millipedes. They differ
from other millipedes in that adults are less than
5 mm in length, the cuticle is unmineralised and
the animals are covered in bristles or trichomes
(Hopkin & Read 1992). Due to their small size
and soft structure, penicillates need to be
specially prepared for high-power microscopic
examination in order to clearly view their
diagnostic characters. Diagnostic features used
include insertion patterns of the cephalic,
tergal and caudal trichomes; number and
nature of antennal and leg sensilla; details of
the mouthparts (labrum and gnathochilarium);
and number of ocelli (Nguyen Duy-Jacquemin
2006; Short & Huynh 2006). No information is
available on histological techniques used to
examine penicillates other than preparation of
whole mounts. This paper describes the technique
we have developed to make permanent mounts
using the stain Fast Green, a stain more commonly
used in botanical microscopy (Ruzin 1999) and
illustrates a selection of the characters used in
identification of genera and species. The method

has been modified from that described by
Upton (1991) for mounting small arthropods for
microscopic examination.

MATERIALS AND METHOD

Due to their small size penicillate millipedes
are most commonly collected from bark and litter
samples using Tullgren funnel extractions and
stored in 70-80% ethanol. Selected specimens at
least 2 mm in size are then prepared for micro-
scopic examination as described below:

1. Removal and mounting of trichomes: This
allows the clearest view of the pattern of
trichome insertion points. Place a specimen
in a drop of 100% isopropanol on a glass
slide and remove trichomes with fine
forceps and a microprobe. Gently disperse
to separate them and after the isopropanol
has evaporated cover the trichomes with
a drop of DPX mounting medium and
cover-slip.

2. Maceration and clearing: Body contents are
digested while retaining details of the exo-
skeleton including sensory hairs and colxal
gland openings. Place previously preserved

Memoirs of the Queensland Museum 1 Nature • 2010 • 55(1) • www.qm.qld.gov.au

231

Short & Huynh

FIG. 1. Technique for opening up exoskeleton of
penicillate millipedes in order to remove contents
and prepare exoskeleton for staining.

specimens in an Eppendorf tube with 15%
potassium hydroxide and heat gently in a
water-bath for 2 minutes at 80°C followed
by rinsing in distilled water. Damaged
specimens or those preserved for many
decades need just 1 minute in potassium
hydroxide, while fresh specimens require a
longer period of up to 12 hours immersion.

3. Dehydration and staining: Rinse the cleared
specimen in water and then place into
20% acetic acid for 2 minutes to neutralise,
followed by dehydration through a series
of ethanol solutions from 70%, 80%, 90% to
100% (2 minutes per solution), followed by
2 minutes in 1% Fast Green in 100% ethanol.
Return the specimen to 100% ethanol to
remove excess stain.

FIG. 2. Exoskeleton of body (A) and head (B) after slide
preparation showing both dorsal and ventral features.
Scale bars = 500pm.

4. Cleaning and dissection: The cleared and stain-

ed millipede is opened up using microprobes
and micro-scissors in a series of steps as
illustrated in Fig. 1, followed by removal of
body contents. The Head is detached from the
body. Removal of contents and opening out
of the body and head are required to facilitate
viewing of morphological characters of the
exoskeleton.

5. Re-staining and mounting: Return the head
and body to 1 % Fast Green solution in 100%
ethanol for 2 minutes followed by 100%
isopropanol for 1 minute and finally into
xylene for at least 1 minute. Specimens can
be left in xylene until ready to mount. Mount

232

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

Technique to examine penicillate millipedes

FIG. 3. Examples of diagnostic features visible after
preparation and staining exoskeleton of Unixenus
mjoebergi (Verhoeff) with Fast Green. A, eye showing
ocelli; B, gnathochilarium showing palpi sensilla; C,
pattern of trichome insertion points on right half of
third tergite; D, sensilla on 6 th antennal article. Scale
bars = 50 gm.

the head and body separately in a drop of
xylene on a glass slide followed by a drop of
mounting medium DPX. DPX is preferred as
it dries faster than Canada Balsam or Euparal.
When mounting the head, tease apart the
antennae and mouthparts with microprobes
so that all relevant diagnostic features are
clearly visible. When mounting the body,
arrange the split body so that both dorsal
and ventral surfaces are in the same plane
for ease of viewing.

Images were taken using a Nikon Coolpix
4500 digital camera with an Olympus Vanox
compound microscope.

RESULTS AND DISCUSSION

The body and head after preparation and
staining are illustrated in Fig. 2. Having a
permanent preparation of both dorsal and ventral
features in the one field of view, enables efficient
examination and documentation of diagnostic

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

233

Short & Huynh

features. Fig. 3 illustrates a selection of these
features. The method provides clearer views of
features in comparison with a whole mount in
which the gut contents often obscure important
features such as the pattern of tergal trichome
insertions. A further advantage is that staining
prior to dissection enables the specimen to
be more easily visualised for dissection after
maceration renders penicillates transparent
and hard to manipulate. As well only one speci-
men is required to view all features. This is an
important factor when only a small number of
specimens is available. A final advantage of the
technique is reduction of the depth of field of the
preparation with the body opened out and the
contents removed.

We have used the method successfully with
specimens from the penicillate families Synx-
enidae, Polyxenidae and Lophoproctidae includ-
ing museum specimens lacking any colouration
after decades in preservative.

ACKNOWLEDGEMENTS

Comments from R. Mesibov and an anony-
mous reviewer improved an earlier version of
this paper.

LITERATURE CITED

Hopkin, S.P. & Read, H.J. 1992. The Biology of
Millipedes. (Oxford University Press: Oxford).

Nguyen Duy-Jacquemin, M. 2006. Condexenus, a
new genus of the millipede faily Synxenidae
(Diplopoda, Polyxenida) from Namibia,
Norwegian Journal of Entomology 53: 237-248.

Ruzin, S. E. 1999. Plant microtechnique and microscopy.

(Oxford University Press: New York).

Short, M. & Huynh, C. 2006. Redescription of
Phryssonotus novehollandiae (Silvestri, 1923)
with details of post-embryonic stadia. Noru>egian
Jounial of Entomology 53: 211-222.

Upton, M. S. 1991. Methods for Collecting, Presewing,
and Studying Insects and Allied Forms (4 th Edition).
(Australian Entomological Society: Brisbane).

Memoirs of the Queensland Museum | Nature • 2010 • 55(1)

^ IV53

234

CONTENTS CONTINUED

NOTES

BISHOP, PJ.

A triassic conchostracan from near Murgon, SEQ 8

BISHOP, PJ.

A Conchostracan from the Carboniferous Ducabrook
formation 86

VAN DYCK, S„ JANETZKI, H. & SHERIDAN, J.

Structure of a burrow of the fawn hopping-mouse Notomys

cervinus (Rodentia: Muridae) 98

COOK, A.G. & MCHENRY, C.R.

A homonym of Leptocleidus Andrews, 1 922 (Sarcopterygia,
Plesiosauria, Leptocleidoidea, Leptodeididae) and a replacement
name for Leptocleidus Mueller 1 936 (Platyhelminthes,
Monogenea) 118

CONTENTS

HOSKIN, CJ.

Breeding behaviour of the Barred Frog Mixophyes coggeri 1

BAKER, C.H.

A new subgenus and five new Australian glow-worm species

(Diptera: Keroplatidae: Arachnocampa spp.) 11

BARTHOLOMAI, A.

Revision of Flindersichthys denmeadi Longman 1932, a marine
teleost from the Lower Cretaceous of the Great Artesian Basin,

Queensland 43

BARTHOLOMAI, A.

A new Albian Teleost, Euroka dunravenensis gen. et sp. nov.
and a new family, Eurokidae, from the Eromanga Basin of

Queensland 69

PIERCE, SJ.& BENNETT, M.B.

Distribution of the estuary stingray ( Dasyatis fluviorum ) in
Australia 89

THEISS, S.M., KYNE, P.M. & CHISHOLM, LA.

Distribution of the porcupine ray Urogymnus asperrimus (Bloch
& Schneider, 1801) in Australian waters, with new records from

Queensland 101

WOOLLEY, P.A.

The Julia Creek dunnart and other prey of the barn owl in
Mitchell grass downs of north-western Queensland 107

CANTRELL, B.K. & BURWELL, CJ.

The tribe Dufouriini (Diptera: Tachinidae: Dexiinae) recorded
from Australia with the description of two new species 119

COOK, A.G., SAINI, N. & HOCKNULL, S.A.

Dinosaur footprints from the Lower Jurassic of Mount Morgan,

Queensland.. 135

DETTMANN, M.E. & CLIFFORD, H.T.

Fossil fruit of the Macadamieae (Proteaceae) in the Tertiary of
eastern Australia: Eureka gen. nov 147

KOHOUT, RJ.

A review of the Australian Polyrhachis ants of the subgenera
Myrmhopla Forel and Hirtomyrma subgen. nov. (Hymenoptera:

Formicidae: Formicinae) 167

RANDALL. J.E.&F. WALSH

Rabaulichthys squirei, a new species of Sailfin Anthias (Serranidae:

Anthiinae) from the Coral Sea 205

STREET, D„ EDNEY G„ ROWE, D. & LAWLER, S.H.

A putative hybrid of the Murray Crayfish, Euastacus armatus,

(Crustacea: Decapoda: Parastacidae) 213

TOWNSEND, K.A. & KYNE, P.M.

New records of the Japanese devilray Mobula japanica (Muller &

Henle 1841) for Australian waters 225

SHORT, M.& HUYNH, C.

A technique for examination of diagnostic characteristics of
penicillate millipedes 231

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Memoirs of the Queensland Museum"

See other formats