Records of the Western Australian Museum

ISSN 0312 3162

Records

of the

Western Australian
Museum

Volume 23 Part 3 2007

Records

of the

Western Australian Museum

The Records of the Western Australian Museum publishes the results of research into all branches of natural
sciences, and social and cultural history, primarily ba,sed on the collections of the Western Australian
Museum and on research carried out by its staff members.

Collections and research at the Western Australian Museum are centred on Earth and Planetary Sciences.
Zoology, Anthropology, Archaeology and History. In particular the following areas are covered: systematics,
ecology, biogeography and evolution of living and fossil organisms; palaeontology; mineralogy; meteoritics;
anthropology and archaeology; history; maritime hi.story and maritime archaeology; and conservation.

Western Australian Museum

Perth Cultural Centre, James Street, Perth, Western Australia 6000

Mail: Locked Bag 49, Welshpool DC, Western Australia 6986

Tel. (08) 9212 3700

Fax. (08) 9427 2881

E-mail reception @mu,seum. wa.gov.au

Minister for Culture

and the Arts The Hon. Sheila M'Hale MLA

Chair of Trustees Mr Tim Ungar BEc, MAICD, FAIM
Executive Director Dr Dawn Casey PSM FAHS

Editor Dr Kenneth J. McNamara BSc(Hons), PhD, FAAAS

Editorial Board Dr Alex Baynes MA, PhD

Dr Alex Bevan BSc(Hons), PhD
Ms Ann Delroy B At Hons), MPhil
Dr Paul Doughty BSc(Hons). PhD
Dr Bill Humphreys BSc(Hons), PhD
Dr Barry Hutchins BSc(Hons), PhD
Dr Moya Smith BA(Hons) DipEd. PhD

The Records of the Western Australian Museum is published approximately three times per year. A series of
Supplements is also produced. The Records are available for sale and exchange, the current price being $11 plus
postage per part. Each volume comprises four parts. Sub.scriptions can be taken out for whole volumes at a rate of $33
plus postage. Supplements can be purchased from the W.A. Museum Bookshop. Prices on request. Prices include GST.

Printed and published by the Western Australian Mu.seum
© Western Au,stralian Museum, Febmary 2007
ISSN 0312 3162

Cover: Sample developmental stages for the direct developing frogs

Arenophrs'ne rotunda (row 1) and Myohatrachus gouldii (row 2).

Records of the Western Australian Museum 23: 213-217 (2007).

^ .

A new species of Gnathoxys (Coleoptera: Carabidae: Carabinae)
from an urban bushland remnant in Western Australia

Nadine A. Guthrie

Department of Environment and Conservation, Wildlife Research Centre,
PO Box 51, Wanneroo, Western Australia 6946, Australia
Email: Nadine.Guthrie@dec.wa.gov.au

Abstract - Gnathoxys pannuceus sp. nov. is described and illustrated from a
specimen collected from Woodman Point Reserve, Western Australia. It is
distinguished from other Gnathoxys species by the highly distinctive
wrinkled pattern on the elytral surface, a feature that immediately
distinguishes it from all other members of the genus.

INTRODUCTION

The ground beetle genus Gnathoxys Westwood
1842 is endemic to Australia and 16 species are
currently recognised. The majority of these species
occur in the southwestern region of Western
Australia and seven occur along the Murray-
Darling River system (Moore etal. 1987). Gnathoxys
punctipennis Macleay 1873, also occurs along the
southern coastal regions between southwest
Western Australia into the South Australian Gulf
region. Two species have not been seen since initial
collection: G. irregularis Westwood 1842, reported
from Port Essington in the Northern Territory by
Westwood (1842) and G. sulcicolhs Sloane 1910,
from central Australia (no type locality was
reported for this species; Moore et al. 1987).

Since initial collection, little has been determined
about the ecology, taxonomic relationships or
distributions of the various species within
Gnathoxys. Generic relationships are not clear;
however, Macleay (1864) suggested tentative
associations between C. tesselatus Macleay 1864
and Promecoderus based upon the dilation of the
male fore tarsus. Roig-Juhent (2000) however,
determined that within the Broscitae, Gnathoxys
forms a natural grouping with the other Australian
endemic genera Cerotalis, Adotela and
Brithysternuin and is only distantly related to other
genera (including Promecoderus, Creobius and
Cascellius) within the subtribe Creobiina of the
Broscitae. Gnathoxys is unique among this group
and is defined by possessing fore tibia with two
medial teeth, fore and middle tibia markedly
palmate, and maxillary and labial palpal apical
segments of males securiform (Roig-Juhent 2000).

The broscitines are diverse and medium sized
ground-dwelling beetles distributed in temperate,
subarctic and subantarctic regions of the world, and
are generally absent from the tropics (Davidson and

Ball 1998). Within Australia, the group is an
important element in the beetle fauna of arid areas
(Matthews 1980). There are 11 recognised
Australian genera, with all species endemic, but
two genera are represented by other species outside
of Australia (Moore et. al. 1987).

A recent survey of urban bushland remnants in
Perth, Western Australia surveyed a number of sites
in the metropolitan area and a large number of
carabid specimens, including several species of
Gnathoxys, were collected in pitfall traps (How et.
al. 1996; Guthrie 2001). Amongst these samples was
a single representative of an unusual new species of
Gnathoxys collected from Woodman Point, south of
Perth. The distinctive appearance of the elytra!
sculpturing of this specimen is sufficient to suggest
that this represents a new species, and the species is
named and described in this paper.

MATERIALS AND METHODS

The specimen was collected using wet pitfall
traps along a 100 metre transect set with an ethylene
glycol mix (400 ml of 70% ethylene glycol, 30%
water). The specimen was stored in 75% ethyl
alcohol until identification and removal of the
genitalia, and then pinned. The specimen is lodged
in the Western Australian Museum, Perth (WAM).

Measurements were taken using a stereo
microscope with vernier callipers and expressed in
millimetres. Body length was measured from the
apical margin of the labrum to the apex of the elytra
(T-L). Length of pronotum was taken along the
midline (P-L). Fore tibia length was measured from
the femur joint to tip of 1“ tibial tooth (FT-L).

The gross genital morphology was examined by
relaxing the specimen in a mixture of soapy
distilled water and 2% acetic acid. The genitalia
were then dissected out and cleared overnight in

214

N.A. Guthrie

cold 10% potassium hydroxide. Once cleared, the
pH of dissected parts was neutralised in dilute
acetic acid. The dissected male genitalia were
placed in glycerine for examination (Liebherr 1990).

SYSTEMATICS
Family Carabidae
Subfamily Carabinae
Tribe Broscitae

Genus Gnathoxys Westwood, 1842
Type species

None designated by Westwood, but originally
included nominotypical species; Gnathoxys
granulans Westwood, 1842; Gnathoxys irregularis
Westwood, 1842. Gnathoxys granularis Westwood,
1842 by subsequent designation of Roig-Juhent
( 2000 ).

Gnathoxys pannuceus, sp. nov.

Figures 1-7

Material examined

Holotype

Male, Woodman Point Reserve, Western
Australia, site WP2 [32°07'50"S 115°45'28"E], wet
pitfall trap, 4 November 1994-19 January 1995,

Figure 1 Dorsal habitus of Gnathoxys pannuceus
holotype; male total length 13.3 mm

collected by J. M. Waldock and M. S. Harvey (WAM
#38293).

Diagnosis

Gnathoxys pannuceus is similar in overall
appearance and size to G. crassipes Sloane but is
distinguished from all other Gnathoxys species by
being heavy in appearance with a large head
relative to overall size. The pronotum is strongly
globular in shape with a distinct medial sulcus and
faint wrinkles on the otherwise smooth dorsal
surface. The pronotum and elytra margins have fine
long setae in greater abundance than other similarly
sized Gnathoxys. The most obvious character that
separates this species from all others in the genus is
the striking elytral pattern. Whereas G. granularis
has distinct granulated areas on the elytra, and
other Gnathoxys species possess elytral patterns
consisting of foveae, punctures or similar
depressions, this species has a highly distinctive
wrinkled pattern.

Description

Male (holoti/pe) (Figures 1, 2)

Total length = 13.3 mm; elytral length/width = 7.6/
5.9 mm; pronotum length/width = 4.3/5. 1 mm; head
length = 3.0 mm; fore tibia length = 2.9 mm. Colour
entirely black without bronze or olive sheen, with
dark orange eyes.

A new species of Gnathoxys

215

Head. Very long, heavy mandibles approximately
2/3 of head length, slightly curved downward. Inner
mandible edge straight and toothless, curved
toward apex with deep overlap ot mandible apices.
Mandibular groove wide and shallow,
approximately half mandible length, mandibular
ridge very narrow. Single seta at groove apex and
non-setiferous puncture on outer curve of
mandibles near apex. Single seta on medial surface
of 2""' segment of palp and on ventral surface of
basal segment of maxilla palp. Maxillae and labial
apical palp segments securiform. Two fine setae on
anterior mentum medial margin and one on either
side of extremities of basal maxilla. Labrum slightly
broader than long, bifid and rounded. Medial
sulcus extremely faint. A fringe of setae on under
side of labial anterior margin and three pairs of
setae on labial dorsal anterior margin. Outer labrum
edges yellow with remainder reddish brown. Eyes
round, convex and not prominent or overly large.
Antenna short, moniliform with single seta on scape
and segments 4-11 covered dorsally and ventrally
with thick short setae. Supraorbital seta posterior to
eye, with supraorbital sulcus running forward,
terminated posterior to mandibular ridge. Deep
latero-medial sulcus on either side of head, initiated
in line with anterior half of eye, extended directly
forward to lateral extremities of clypeus. Clypeus
medially and anteriorly depressed, with one mid
and two lateral creases medially aligned.

Prothorax. Pronotum very rounded, sub-spherical
with very weak extensions at cervTcal collar
insertion point. Narrow pronotal margin with setae
in anterior and posterior third of margin. Medial
sulcus extended forward to anterior margin. Lateral
wrinkles traverse pronotum surface, strongest near
medial sulcus, lateral margins and towards
thickened and blunt basal margin. Prosternum with
wrinkles around sparse cluster of setae in front of
each leg (widest anteriorly), wrinkles continue onto
proepimeron, tubercles reduced to slight swollen
areas between anterior coxa.

Pterothorax. Elytra are sub-quadrate, slightly
longer than broad with rounded sides and apex.
Peduncle thick and short with heavy shoulders
projecting. Elytral margin very thin, with five setae
evenly spaced along anterior two thirds of margin.
Apical declivity finely granulated, extending over
posterior one sixth of elytra. Granulations extending
along lateral margins, diminished anteriorly. Four
setae evenly spaced along dorsal edge of apical
declivity on each elytron. Dorsal surface of elytra
finely creased and wrinkled with extremely
irregular sulci, reminiscent of "crumpled
aluminium foil re-flattened" (Figure 3).

Abdomen. Ventrites bipunctate medially, with
final seta pair positioned on medial portion of
apical margin.

Legs. Foreleg: Trochanter ventral surface with one

Figure 3 Gnathoxys pannuceus sp. nov. Detail of
apical elytral surface showing extremely
irregular sulci.

punctate seta. Femur with one cluster of setae on
anterior ventral edge, two setae on posterior ventral
edge, three setae on centre of posterior dorsal edge
and a cluster centrally positioned on dorsal surface.
Two teeth present on outer fore tibial edge, medial
one smaller, both with a seta positioned on
posterior distal margin. Linear arrangement of three
setae along midline in line with antennal cleaning
organ. A row of fine setae along inner edge of tibia
terminated at cleaning organ. Rounded, flattened
apical tooth directed distally. Tarsomeres triangular
with outer lateral edge extended, narrowed distally
towards 2"'^ tarsomere. Tlrree or four stiffened setae
on both tarsomere edges. Apical tarsomere filiform
with 2 setae on lateral edges, tipped with
symmetrical short curved claws.

Midleg; Coxae with a cluster of setae on anterior
surface, one seta on ventral surface, and one
ventrally on trochanter. Clusters of setae present on
anterior, dorsal and posterior femoral surfaces.
Femur widened dorso-ventrally. Tibia with linear
rows of stiff setae orientated distally on anterior
and posterior surfaces. A triangular apical tooth
with stiffened setae fonning a fringe around distal
surface of tibia at tarsus insertion point. Two similar
sized apical teeth inserted below tarsus. Tarsal
arrangement identical to foretarsus.

Flindleg: Coxae with two setae on apical and basal
margins. Cluster of setae on posterior and dorsal
surfaces of trochanter. Long setae in curved linear
clusters on posterior and ventral surfaces of femur.
Long setae sparsely distributed on distal ventral
and dorsal third of femur. Tibia elongate, flattened
with widened distal end. Rounded apical tooth on
tibia broad and short. Tibia edge serrated weakly,
serrations with rounded points. Stiffened short
setae in linear rows thickly cover tibial surfaces.

216

N.A. Guthrie

Figures 4-7 Gnathoxys pannuceus sp. nov. 4, genital ring; 5a, right (dorsal) view median lobe; 5b, left (ventral) view
median lobe; 6, left paramere; 7, right paramere; scale bar = 1 mm

Shortened apical teeth, equal in length set below
tarsal insertion point. Tarsal arrangement identical
to anterior tarsus.

Male Genitalia. Genital ring ovoid in shape, with
slight concavity towards basal third, and thin edges
and no extensions (Figure 4). Median lobe (Figure
5a,b) thick, with no curvature and a small hook at
apex. Orifice dorsally placed behind apex. Left and
right sides of median lobe not symmetrical, with
left (or ventral view) extended on upper surface
near orifice. Parameres dissimilar (Figures 6, 7), left
with extension on inner edge, extended to apical
third of paramere. Right paramere larger and
thicker, with thick setal brush extended from apex
to mid-length, almost equal to adeagus in length.

Female: Unknown.

Etymology

The specific epithet is derived from the Latin
adjectival pannuceus (lesser ragged, wrinkled,
shrivelled) pertaining to the characteristic dorsal
surface of the elytra.

Remarks

Gnathoxys pannuceus sp. nov. was collected from
the type locality at Woodman Point and, although
the pitfall traps were left open for twelve months,
only a single specimen was collected. Searches at
the type locality during the same season over
several years have failed to locate any further
specimens, suggesting that this species is locally
uncommon or inhabits a microhabitat that is not
effectively sampled through pitfall traps.

Numerous unidentified forms of Gnathoxys exist
in collections (Western Australian Museum,
Agriculture W.A. and the Australian National
Insect Collection; author's unpublished
observations). Sloane (1898) listed several
unidentified Gnathoxys specimens but his
descriptions and comments are too brief to
satisfactorily ally any of the descriptions with these
unidentified forms. It is also highly likely that more
species of Gnathoxys will be collected in poorly
surveyed areas of southwestern Australia.
Therefore, a comprehensive revision of the genus

A new species of Gnathoxys

217

incorporating all available material, including
currently undescribed forms and the old types is
required immediately.

ACKNOWLEDGEMENTS

I thank M.S. Harvey and J.M. Waldock for access
to the carabid specimens collected during the
Ground Fauna of Urban Bushland Remnants in
Perth Survey. Thanks also to T. Houston, A. Szito
and T. Weir for access to carabid beetle collections
at the Western Australian Museum, Department of
Agriculture, WA and Australian National Insect
Collection, Canberra. I also thank B.P. Moore for
bringing to my attention the significance of this
unusual specimen. Jane McRae kindly
photographed the holotype for me. Finally, I thank
my supervisors, Pierre Horwitz (Edith Cowan
University) and Mark Harvey for their unending
advice and support throughout this project.

REEERENCES

Davidson, R.L. and Ball, G.E. (1998). The tribe Broscini in
Mexico: Rawlinsius papillatus, new genus and new
species (Insecta: Coleoptera: Carabidae) with notes on
natural history and evolution. Annals of the Carnegie
Museum 67 (4): 349-378.

Guthrie, N.A. (2001). Aspects of the Taxonomy and
Ecology of Ground Beetle (Carabidae) Assemblage on
the Swan Coastal Plain (with particular reference to
habitat fragmentation on the Quindalup Dune
System). Masters Thesis, Edith Cowan University.
How, R.A., Harvey, M.S., Dell, ]. and Waldock, FM.
(1996). Ground Fauna of Urban Bushland Remnants
in Perth. Unpublished Report to the Australian
Heritage Commission.

Liebherr, J.K. (1990). A new tribal placement for the
Australasian genera Homethes and Aeolodermus
(Coleoptera; Carabidae: Odacathini). Pan-Pacific
Entomologist 66 (4):312-321.

MacLeay, W.J. (1864). On the scaritidae of New Holland.
2"*^ paper. Transactions of the Entomological Society
of New South Wales!: 134-154.

MacLeay, W.J. (1873). Miscellanea entomologica.
Transactions of the Entomological Society of New
South Wales 2: 319-370.

Matthews, E.G. (1980). A guide to the genera of Beetles
of South Australia. Part 1 Archostemata and
Adephaga. Special Educational Bulletin Series South
Australian Museum, Adelaide.

Moore, B.P., Weir, T.A. and Pyke, J.E. (1987). Rhysodidae
and Carabidae. Pp20-320. In D.W. Walton (ed).
Zoological Catalogue ot Australia. Volume 4.
Coleoptera: Archostemata, Myxophaga and
Adephaga, 444pp. (Canberra: Australian Government
Publication Service).

Roig-Juhent, S. (2000). The subtribes and genera of the
tribe Broscini (Coleoptera: Carabidae): cladistic
analysis, taxonomic treatment, and biogeographical
considerations. Bulletin of the American Museum of
Natural History 255: 1-90.

Sloane, T.G. (1898). On Carabidae from West Australia
sent by Mr A. M. Lea (with descriptions of new
genera and species, synoptic tables, etc). Proceedings
of the Linnean Society of New South Wales 23: 444—
520.

Sloane, T.G. (1910). Studies in Australian Entomology.
No. XVI New species of Carabidae. Proceedings of
the Linnean Society of New South Wales 35: 378M:06.

Westwood, J.O. (1842). On the Scaritideous beetles of
New Holland. Pp 81-90 pi. 24 In J.O. Westwood.
Arcana Entomologica or illustrations of new rare and
interesting insects. London : Smith Vol 1 iv 192 pp.

Manuscript received 29 April 2005; accepted 10 July 2006

Records of the Western Australian Museum 23: 219-234 (2007).

Observations of the biology and immature stages of the sandgroper
Cylindraustralia kochii (Saussure), with notes on some
congeners (Orthoptera: Cylindrachetidae)

Terry F. Houston

Western Australian Museum, Locked Bag 49, Welshpool DC, Western Australia 6986, Australia
Email: terry.houston@museum.wa.gov.au

Abstract - Field and laboratory observations of Cylindraustralia kochii are
presented with notes on some congeners. Nymphs and adults create galleries
in moist soil by compression of the soil with their powerful fore legs,
burrowing to depths of up to 1.9 m. During the cooler months and 1-2 days
after rain, sandgropers commonly burrow long distances close to the soil
surface producing conspicuous raised trails. Adults and nymphs of various
sizes were found throughout the year. Eggs and early immatures of the genus
(and family) are described for the first time. Pedicellate eggs of C. kochii were
suspended singly in closed chambers 40-190 cm deep in moist soil. A 'larval'
stage hatches from the egg and moults to a first instar nymph while still in
the egg chamber. Five nymphal instars are indicated by morphometric and
morphological data. Eggs are laid from autumn to spring but hatching was
only observ'ed in mid summer. A duration of at least 12 months is indicated
for first instar nymphs, so the complete life cycle may extend over several
years. Examination of gut contents revealed that sandgropers are omnivorous,
consuming a wide array of plant, fungal and arthropod material. Plant food
included root, stem, leaf, flower and seed tissue. Cannibalism occurred in one
very dense population of C. kochii. Otherwise, no insect predators or
parasitoids were encountered. Associated organisms included gregarines and
Amoeba (Protista) in the intestines, rhabditid nematodes in the genital
chambers of adults, and six species of mesostigmatid and astigmatid mites
which adhered externally to the body. Nymphs and adults produce an
odorous, probably defensive secretion from a pair of abdominal glands.

Key words; subterranean insects, ethology, ecology, parasites

INTRODUCTION

Sandgropers, once regarded as degenerate mole
crickets (e.g., Tindale 1928), are now classified with
the short-horned grasshoppers (suborder Caelifera)
and form the family Cylindrachetidae within the
superfamily Tridactyloidea (Rentz 1996). Included
with them in this superfamily are the Tridactylidae
('pygmy mole crickets') and Ripipterygidae ('mud
crickets') (Gunther 1994; Flook et al. 1999). All
cylindrachetids are burrowing insects, highly
modified for a subterranean existence. The body
shape is cylindrical, the fore legs are highly
modified for digging, the reduced mid and hind
legs recess into the sides of the abdomen, simple
eyes replace the compound eyes, antennae and cerci
are reduced, and wings are entirely absent (Figures
1-3). Of all the orthopteroid insects, they are
considered to be the most strongly modified
morphologically for a subterranean life (Kevan
1989).

In the most recent revision of the family (Gunther
1992), three genera and 16 species were recognized.

Fourteen species are Australian, one is Argentinean
and one putatively occurs in New Guinea. Gunther
erected a new genus, Cylindraustralia, to contain 13
of the Australian species. Prior to his revision, all
known Australian species were placed in
Cylindracheta Kirby, a genus Gunther restricted to
one species from the 'Top End' of the Northern
Territory. Cylindraustralia species occur widely
across the Australian continent but are absent from
the south-eastern portion.

Although the taxonomy of cylindrachetids has
been reasonably well studied, their biology has
received scant attention (Barrett 1928; Tindale 1928;
Richards 1980; Gunther 1992; Rentz 1996). Some
published information is misleading or incorrect
and nothing has been recorded hitherto of the eggs
and early immature stages. Of course, living almost
wholly subterranean lives, the insects are rarely
observed and make difficult subjects for study.

In Western Australia, sandgropers have gained a
reputation as agricultural pests, being reported to
damage wheat, barley, oats, sweet lupins and

220

T.F. Houston

Figures 1-5 Cylindraustralia kocbii. (1-2) Adult female and male, respectively (note bands of pigmentation around
abdomen, in male interrupted dorsally on segments 7-9). (3) Last stage nymph (note absence of abdominal
pigmentafion; dark marks' along dorsal median line are gaps in underlying faf body visible through
transparent integument). (4-5) Surface trails produced by adults burrowing just beneath surface of
ground. (4) simple trails in natural bushland; (5) branched trails on compacted sand surface of farm road.

tagasaste between Perth and Geraldton (Richards
1980; Rentz 1996; Wiley 2000). Only anecdotal and
circumstantial evidence, though, was produced by
these authors to show that sandgropers were the
cause of the observed plant damage.

While the insects themselves are rarely
encountered, their characteristic trails (Figure 4) are
a common sight on bare sandy ground in Western
Australia. Two species (C. kochii (Saussure) (syn.
psammophila (Tindale)) and C tindalei Gunther)
are known to be extant in and around Perth.

The present study was undertaken in an attempt
to elucidate the life histories, behaviour and ecology
of sandgropers.

MATERIALS AND METHODS

Over 900 spirit-preserved specimens of
Cylindraustralia in the collection of the Western
Australian Museum were examined in this study.
Most were collected by the author from 2002-2005,
the remainder being donated by members of the

Biology of sandgropers

221

farming community and the general public in
response to a media appeal. By far the bulk of the
material studied was comprised of C. kochii while
most of the remainder consisted of C. tindalei.

Although sandgropers have occasionally been
found in pitfall traps, the author's deployment of
gutter traps and pitfall traps combined with drift
fences at a number of sites failed to yield specimens.
Following on foot close behind farm ploughs
turning over soil under pasture yielded many
specimens. Others were obtained from near-surface
galleries: by driving back and forth along sandy
roads and firebreaks on the margins of bushland
shortly after rain, it was possible to recognize fresh
trails where they crossed the vehicle's tyre tracks
(Figure 5). Most specimens obtained for this study,
however, were turned up by digging with a spade
beneath pastures and weeds on farms.

Study sites where significant work was
undertaken are as follows (short-hand names used
in this paper appear in quotation marks):
"Dandaragan site"- Annamullah Farm, 6 km NNE
of Dandaragan, 30°38'S, 115°45'E; "Fforrocks site" -
Willi Gulli North Farm, 18 km W of Northampton
(2 km E of Horrocks), 28°22'S, 114°27'E; "Eurardy
site" - Eurardy Station, 89 km N of Northampton,
27°34'S, n4°40'E; M. and D. Webb's farm, 23 km E
of Northampton, 28°18'52"S, 114°51'58"E; and
"Great Sandy Desert", various sites approximately
220-280 km SE of Broome, between 19°04'13"S,
f23°44'05"E, and 19°17'52"S, 124°26'27"E.

Various methods of killing and preserving
specimens were trialled. For the purposes of later
dissection, best results were obtained by freezing
specimens. Where this was impractical, freshly
killed specimens were injected with and stored in
10% formalin (although injection caused the
abdomen to inflate and extend). Several specimens
were killed by spraying the head and thorax with
electrician's freezer and were then immediately
dissected in saline to check for living parasites or
commensals in the gut, abdominal cavity and
genital tracts.

Live specimens were maintained in containers of
moist sand or sandy loam with various plants: Cape
Weed {Arctotheca calendula), Wild Oats {Avena
tatua), and seedlings germinated from commercial
'mixed budgie seed'. Glass-bottomed and clear
plastic containers permitted observations of
burrowing activity. Eggs were reared on tissue
wads in glass vials in humid boxes. The boxes were
kept at room temperature (18-30°C) and open vials
of saturated salt solution provided moderate
humidity.

Specimens were identified by comparison with
specimens in the Western Australian Museum
determined by Dr Kurt Gunther and by means of
Dr Gunther's 1992 revision of the family
Cylindrachetidae. Some specimens from the

Horrocks and Great Sandy Desert sites could not be
matched to any of Gunther's taxa and appear to
represent undescribed species referred to below as
'Species A' and 'Species B', respectively.

The pronotal width of all specimens was
measured to determine the number of instars. The
pronotum is a rigid structure that is easily and
reliably measured across its greatest width.

Population sampling at the Dandaragan site was
undertaken approximately every second month
although the October sample was not in sequence
with the rest. The method used was to excavate a
large pit at least 1 x 2 m in area and 1-2 m deep
using a spade and trowel and to collect every
specimen encountered as the soil was turned over.
Excavation required 2-4 days.

OBSERVATIONS AND DISCUSSION
Life Stages and Morphology

Adults

Apart from having completely developed
genitalia, adults are distinguishable from nymphs
in having the abdominal integument wholly or
largely tan-coloured (Figures 1, 2) (in males of C
kochii the tan pigmentation is usually broken by
narrow, colourless, intersegmental bands). The
abdominal integument of all nymphal stages, by
contrast, is completely colourless and, being
transparent, the abdomen appears white or cream
because of the underlying fat body (Figures 3, 11).

Males and females are similar in size. Among a
sample of C. kochii adults from the Dandaragan
site, pronotal widths of males ranged from 6.75-
8.20 mm (mean 7.35 ± 0.33, n = 20) and of females
from 6.80-8.30 mm (mean 7.4 ± 0.44, n = 14). The
sexes are also very similar morphologically but can
be distinguished by the external genitalia. As
Gunther (1992) noted, males possess a pair of short,
stout spines on the paraprocts near the insertions of
the cerci (Figure 13). Females lack these spines and,
instead, possess a pair of rudimentary gonovalves,
the tips of which sometimes protrude slightly
beyond the apical margin of the 8**^ abdominal
sternite (S8, Figure 12). In C. kochii (but not other
species), adult males are further distinguished by a
large unpigmented patch on the dorsal side of
abdominal segments 7-9 (Figure 2).

Eggs were first observed in the oviducts of
dissected females of C kochii. These oviducal eggs
were elongate-ovoidal, c. 7.0 mm long and 3.3 mm
in diameter, flat to slightly concave on one side
(thus being bilaterally symmetrical), and had a tiny
appendage c. 0.5 mm long anteriorly (Figure 6). The
chorion was smooth, unsculptured and translucent

222

T.F. Houston

Figures 6-11 Cylindraustralia kochii: (6) mature eggs dissected from oviduct (for detail of apical appendages, see
Figures 21-23); (7) laid eggs showing attachment pedicels and adhesive disks with adherent sand grains
(collected in October, chorions dull and opaque); (8) vertical section of earth showing two freshly laid
eggs suspended in their chambers (upper egg has a drop of ground water on right side and a fungus
grows on chamber floor); (9) freshly laid egg with red chorion; (10) larva shortly after eclosion (for more
details see Figures 24-26); (11) newly emerged first instar nymph with its eggshell.

Biology of sandgropers

223

off-white. The appendage consisted of a doughnut-
shaped mass of gelatinous material about 0.7 mm in
diameter attached to a central disc which was in
turn connected axially to the egg by a short flexible
stalk or pedicel (Figures 21-23). At 400x
magnification the gelatinous mass was observed to
consist of tightly packed bundles of fibrils with
their free ends outermost. This appendage later
proved to be a device for attachment of the egg to
the substrate.

Laid eggs of C kochii were first observed in situ
at the Dandaragan site in May 2003 when over 40
were uncovered, each enclosed in a small chamber
(Figure 8). The eggs were suspended from the
ceilings of their chambers on short flexible pedicels,
the upper ends of which expanded into rounded
discs (Figures 7, 24). The discs were firmly
cemented to the soil by some substance that proved
to be water-insoluble. Otherwise, the eggs were free
of contact with the soil. Most eggs in this lot were
translucent white (like oviducal eggs) and
presumably freshly laid. A few were wholly dark
red (Figure 9) while others were white variously
mottled with pink. The red/pink pigmentation was
confined to the chorion and, in the wholly red
individuals, to the stalk and disc as well but never
extended to the yolk which was completely
colourless. Many eggs, too, bore a drop of clear
liquid on one side (Figure 8) - evidently ground
water that had trickled down from the chamber
ceilings. At the same site in October 2004, 32 eggs
were excavated. The majority were wholly or partly
pink and only six were pure white but, in all cases,
the chorion was dull and opaque.

No laid or oviducal eggs were found for C.
tindalei but one near-mature egg (4.3 mm long) in
an ovariole had a gelationous appendage much like
that of oviducal eggs of C. kochii. Oviducal eggs of
Species B, however, lacked a pedicel and
attachment disc. Instead, each egg had a flat apical
cap of gelatinous material (ca. 0.8 mm diameter)
directly and broadly attached to the chorion.

The glueing of eggs to the substrate, and
particularly their suspension on pedicels, is
something not reported for other tridactyloid
families. Eggs of Tridactylidae and Ripipterygidae,
lack any sort of appendage as far as currently
known. Eggs of one tridactylid have been reported
to be laid in batches of 10-20 in the ends of galleries
(Urquhart 1973, cited by Gunther 1994) while those
of ripipterygids are laid singly in excavations made
with the gonovalves much as in the manner of
acridids (Schremmer 1972; Gambardella 1971; both
cited by Gunther 1994).

Larva

In the laboratory, eggs eclosed to a pre-nymphal
stage or 'larva' (Figures 10, 25-27), the equivalent of
the 'vermiform larva' of the Acrididae (Uvarov

1966). The larva was a setose individual of
distinctive form enveloped in a thin, transparent
membrane (the 'provisional cuticle' of Uvarov).
This membrane, unlike that of acridids, lacked setae
and spicules but on the median line of the frons had
a thin, brownish, sclerotized and slightly serrated
Carina (Figures 26, 27), presumably an egg-burster.
Other characteristics were: fore legs reflexed
backwards against body; prothorax much wider
than long and slightly biconvex (weakly depressed
medianly); and mesothorax not encapsulating hind
part of prothorax. This stage is short-lived, the
provisional cuticle being shed almost immediately
after eclosion from the egg, or at least within a
couple of hours, giving rise to the first nymphal
instar.

Nymphs

An individual of typical sandgroper form with
the fore legs directed anteriorly emerged from the
larval skin (Figure 11). In keeping with convention
(David Ragge, pers. comm.), this stage should be
regarded as the first nymphal instar. It is at first
wholly white with pink eyes but gradually (over a
period of days) develops tan colouration in the
head and thorax as the cuticle hardens and the eyes
turn black. These changes occur before the nymph
leaves the egg chamber.

Nymphs are much like adults and are
comparatively uniform morphologically. However,
the development of the external genitalia provides
some characters enabling determination of the sex
of an individual and (in females) the particular
instar to which it belongs. Tentative determination
of the number of nymphal instars in C. kochii was
made possible by measurement of a large number
of nymphs of various sizes and the hatching of early
stages from eggs in the laboratory.

The size-frequency distribution for all C. kochii
collected from the Dandaragan site (Figure 28)
reveals four peaks suggesting the existence of four
nymphal instars. However, as the larger size classes
were poorly represented, the histogram may not
present an accurate picture. If the relative increase
in pronotal width from instar to instar was constant
in keeping with Dyar's 'law' (CSIRO 1991), one
would expect another peak to occur around the 5.0
mm mark.

Anatomical evidence for the existence of five
nymphal instars was found on the eighth abdominal
sternite (S8) of females: the vaginal opening is
evident from the first instar, and increases in size
and shifts rearward with each moult; in later instars,
the gonovalves form from the hind margin of S8
(Figures 14-18).

Male nymphs can be recognized by the absence of
the vaginal groove and/or developing gonovalves.
Additionally, from about the instar, they possess
developing paraproct spines. These are at first

224

T.F. Houston

Figures 12-20 Sketches of genital areas of Cylindraustralid kochii (not to same scale): (12) underside of apex of
abdomen of adult female, somewhat inflated to show various sclerites and apices of gonovalves
(normally hidden behind 8th sternite); (13) same of adult male, showing copulatory spines (solid black)
on paraprocts (pp); (14-18) eighth sternite of lst-5th female nymphal instars, respectively, showing
development of vaginal opening and gonovalves (abdominal sclerites of early instars are unsclerotized
and ill-defined, thus approximate boundaries of S8 are indicated by broken lines); (19) eighth sternite of
adult female showing outline of gonovalves; (20) presumed juxtaposition of hind ends of male and
female during copulation (only with this arrangement could copulatory spines of male engage hind
edge of S8 of female, pulling it down and permitting intromission of genital armature into vagina).
Abbreviations; gv, gonovalves; pp, paraprocts; S8, eighth sternite; vo, vaginal opening.

almost imperceptible, colourless tubercles but, in
later instars, they become more pronounced and
more acute and, in the final nymphal instar, acquire
pigmentation and are strongly sclerotized (c/.
Figure 13).

Putative stridjilatory organ
A putative stridulatory apparatus on the mouth-
parts of cylindrachetids was described and figured
by Gunther (1992) and Rentz (1996). It consists of a
field of microscopic tubercles arranged in rows on

the ventral surface of each mandible and a single
row of about seven short ridges on the opposing
dorsal surface of the basal segment of each
maxillary palpus. Gunther noted this apparatus in
both sexes. It is now clear that it occurs in all
nymphal instars as well. Thus, it is unlikely that the
apparatus plays a part in mate-attraction, if in fact it
produces sound at all. I detected no stridulatory
sounds from sandgropers, even when holding them
close to my ear.

Lawrence and Britton (1994, pi. 2) described and

Biology of sandgropers

225

27 !

Figures 21-27 Cylindraustralia kochii. (21-24) Sketches
of apical appendage of egg: (21) prior to
laying, top (axial) view; (22) same, lateral
view; (23) same, sectional view; (24) after
oviposition (everted adhesive disk is
cemented to sand grains in ceiling of egg
chamber, its outer edges being poorly
defined). Abbreviations: ad, adhesive
disk; m, mucilaginous ring; p, flexible
pedicel. (25-27) Sketches of larva: (25)
lateral view (note reflexed fore log); (26)
anterior view of head and prothorax
showing location of frontal carina (= egg-
burster, arrowed); (27) frontal carina in
left lateral view, not to scale. Scale lines,
1 mm.

January 2004

l-l-i

i ^ ^

rl

ra , t73 , ea ,

May 2003

CL

i ra , IT. , B , ^ ^ , 1 , B , —

June-Julv 2003

October 2004

J

I

December 2003

1

Combined data

h

I m

- ^ .

g ^ ^ i ^ ^ 1 i _

6 i i 6 6

Pronotal width in r

Figure 28

Frequencies of various size classes (based on
pronotal width) of nymphs (hatched) and
adults (solid black) of Cylindraustralia kochii
collected in different months at the
Dandaragan study site. The combined data
set is based on specimens from the six
seasonal samples plus some additional ones.
Note that above a pronotal width of 5.00 mm,
size class intervals increase from 0.2 to 0.5
mm.

226

T.F. Houston

figured similar patches of minute tubercles (termed
'asperities') on the dorsal surfaces of the mandibles
of certain pyrochroid and cucujid beetle larvae but
did not attribute any function to them.

Odour glands

Live specimens of C kochii (and other species)
often emitted a strong, slightly pungent odour
when handled. Gunther (1992) noted a number of
integumental glands and gland openings in
cylindrachetids although he did not discuss their
functions. They included a gland opening on the
inside of each fore femur, glandular tissue in each
mid and hind tibia, a gland opening in each
laterosternite of the 3'*^ abdominal segment and, in
males of C. kochii, an area of glandular tissue
beneath abdominal tergites 7-9. In order to
determine the origin of the odour, each of the gland
areas of a freshly killed adult of C. kochii was
excised in turn, crushed between the fingers and
the residue sniffed to check for odour. Only the 3'“^
abdominal segment produced a very strong and
lasting odour identical to that noticed in handling
live specimens.

Dissection revealed a gland sac attached to each of
the two gland openings on the 3'^‘* abdominal
segment. These sacs are evidently reservoirs for the
gland secretion. An apodeme adjacent to each gland
opening provides attachment for a muscle (possibly
serving to open or close a valve). The gland openings
and sacs were found in adults of both sexes and all
nymphal instars. Consequently, the gland secretion
is unlikely to play a role in mate attraction and a
defensive function seems more likely.

Ecology and Behaviour

Habitat

Field observations, reinforced by museum
collection data, reveal that sandgropers inhabit a
wide variety of sandy soils including calcareous
and siliceous sands and sandy loams. C. kochii
inhabits diverse habitat types including coastal
dunes, sand plains with heath ('kwongan') or
shrubland vegetation, red desert dunes with
tussock grasses (principally Triodia spp.), red sandy
loams with open eucalypt woodland and
comparatively hard sandy loams with Acacia
shrubland in the Gascoyne of WA.

Hundreds of specimens of C kochii were
collected from agricultural land beneath pastures of
mixed weeds including Cape Weed, lupins, clover
and exotic grasses or beneath young cereal crops
(wheat and barley). Some of this land had been
cleared for several decades and the nearest
remnants of native vegetation were tens of
kilometres away. A smaller number of specimens of
C. kochii and C. tindalei were collected from
suburban gardens beneath exotic plants or patches

of weeds. Clearly these sandgropers are not
dependent on native flora.

Burrowing

Observation of specimens in moist sand in glass-
bottomed and clear-sided containers revealed that
they create galleries by parting the soil ahead of
them with synchronous lateral motions of their fore
legs, compressing it to the sides. They do not loosen
soil and shift it behind them the way many other
burrowing insects do. After each stroke of the fore
legs, the insects shuffle forwards on the mid and
hind legs. Upward motions of the head, observed in
hand-held specimens, may also help compact the
walls of the galleries. By twisting the fore body as
they progress, the insects are able to compact the
soil up and down as well as sideways. The galleries
so-formed are smooth-walled, cylindrical and only
marginally wider than the insects creating them.
Sandgropers move easily and quickly both forwards
and backwards within their galleries. Only the mid
and hind legs are invmh'ed in walking, the fore legs
being held stiffly forwards off the substrate.

At the Dandaragan site, adults and large nymphs
of C kochii were frequently excavated from depths
of 1. 0-1.8 m (and given the presence of eggs at 1.9
m, adult females must at times have burrowed to
this depth). They could not have gone much deeper
because of a gravel layer at 2 m. Smaller nymphs
were also found in numbers at depths of 1.5 m or
more, although many (if not all) of them would
hav^e hatched there, line soil at depths of 40 cm and
deeper was very compact and could be cut in
blocks. It is testimony to the strength of
sandgropers' fore legs that they are able to force a
passage through such a compact medium.

Specimens were usually found in horizontal to
somewhat inclined burrows, rarely in vertical
galleries. In some cases they had found their way
into large earthworm shafts which abounded at the
Dandaragan site. The galleries of several nymphs
and adults that were traced carefully wound
erratically downwards, having horizontal, inclined
and vertical sections. Several adults and late stage
nymphs were encountered at the ends of galleries
facing away from the blind ends. How these
individuals could have executed turns, allowing
them to reverse into these galleries, remains
unexplained.

During summer excavations at the Dandaragan
site, no live sandgropers were found in the top 20
cm of the soil (the A horizon) which was dry and
hard. All specimens occurred in the moist subsoil.
The A horizon, however, was almost honeycombed
in places with large galleries created on previous
occasions. In winter, too, no sandgropers were
found in surface soil that had become dry. They
ventured into the surface zone only when it was
damp following recent rain.

Biology of sandgropers

227

A common habit of sandgropers is to burrow long
distances just beneath the surface of the soil
producing raised ridges or 'trails' on the surface
(Figures 4, 5). Adults of C kochii burrow 1-2 cm
beneath the surface, smaller nymphs at
comparatively shallower depths. Beneath each
raised trail (and scores were examined) was a
gallery. The longest continuous section of trail
observed was 10 m but the insects may travel much
further than this. Trails can persist for weeks or
even months and often criss-cross the ground.

In southern Western Australia which has a
Mediterranean climate, several sandgroper species
produce trails only during the cooler, wet months
of the year from about April or May to September
or October and only for 1-2 days after soaking rain
while the surface soil is moist. Fresh trails appeared
throughout the day but not at night. Heavy showers
also elicited trail-forming by Species B at the Great
Sandy Desert sites in July 2005. Typically, this
tropical area has dry winters, receiving its rainfall
during the summer monsoon season.

It was sometimes found that sandgropers had
backed up one or more metres from the blind
(leading) ends of their near-surface galleries. Also,
many trails and their underlying galleries branched,
especially those occurring on compacted surfaces
such as dirt roads (Figure 5). Evidently, when the
insects encounter an obstacle, such as soil that is too
hard to penetrate, they back up and strike off in a
different direction.

Counts of the stages and sexes of sandgropers
collected while trail-forming are shown in Table 1.
For C. kochii, the behaviour seems to involve
mainly adult males (92% of specimens), suggesting
it could be associated with mate-seeking. A similar
but less pronounced trend is noted for C. tindalei
(72% of specimens). By contrast, both sexes were
almost equally represented for Species B. Larger
samples will be required to determine if there are
persistent species differences here. Given that
nymphs as well as adults engage in trail-forming,
this behaviour may represent a general dispersal

Table 1 Numbers of specimens of sandgropers
collected while trail-forming (i.e., burrowing
just beneath the surface of the soil causing a
raised ridge). The species C. arenivaga
(Tindale) was observed by the author in the
Gibson Desert in 1982.

Species

adult

males

adult

females

nymphs

C. arenivaga

2

C. kochii

22

1

1

C. tindalei

13

4

1

C. tindaleil

1

C. tindalei x kochii (?)

1

Species B

8

7

4

mechanism. By burrowing close to the soil surface
which yields, sandgropers would be able to
progress faster and with less effort compared with
burrowing at greater depth and still maintain cover.

Soil moisture is clearly important to the
burrowing activities of sandgropers. First, it softens
the soil (sandy loams often become mortar hard
when dry). Second, it binds sand grains ensuring
that galleries remain open behind the insects,
providing them with a ready means of retreat.

Egg chambers

Egg chambers (Figure 8) measured c. 20 mm in
length, were smooth and evenly concave at one end
and rough at the other. They appeared to have been
formed from the blind ends of horizontal or slightly
inclined galleries through back-filling of the access
burrows following oviposition.

While egg chambers were clearly separate, they
were often loosely aggregated. For example, at the
Dandaragan site, one group of 19 chambers
occurred within a block of soil measuring c. 30 x 20
X 20 cm. Within this group there were tighter
clusters of 2-5 chambers, the chambers being
separated by as little as 1-2 cm. Egg chambers were
found at depths of from 40-190 cm.

The process of egg chamber formation and
oviposition was not observed but must involve the
female in some special manoeuvres including at
least two reversals of direction. As a female creates
a blind horizontal gallery, the end of which will
become the egg chamber, she would face into the
blind end. To oviposit in this blind end, she would
need to reverse direction and, to attach her egg to
the ceiling, must lie on her back. As the egg is
extruded from the vagina, the adhesive disc on its
anterior (leading) end would contact the ceiling and
cement the egg in place. The female must then
withdraw and reverse direction again in order to
attend to closure of the brood chamber (females
having no strongly sclerotized structures at their
hind ends that could serve to scrape or push soil). It
would be impossible for a female to reverse
direction in the narrow confines of a typical gallery,
yet I observed nothing that could have served as a
'turning chamber'. However, I did encounter some
widened sections of gallery (about twice as wide as
usual) which could have been the source of soil
used for back-filling access burrows.

Population density and distribution

Other than finding specimens in near-surface
galleries following rain, attempts to find
sandgropers in bushland areas by means of digging
were unsuccessful, even though many holes were
dug in areas where trails were common. This would
suggest that either the insects were sparsely
distributed or they were deeper than my
excavations (usually not deeper than 50 cm).

228

T.F. Houston

At the Dandaragan site, however, a very different
situation prev^ailed. In mid May 2003 when initial
observations were made by the author, a 1 x 1 m
hole dug almost anywhere in a paddock carrying
only pasture produced one or more specimens. For
example, one exploratory excavation about Im x Im
X 80 cm deep yielded 11 small nymphs. About 50 m
distant, an excavation 1 m x 1 m x 30 cm deep
yielded five adult males but no nymphs.

The greatest density was recorded at the same site
during excavation in late March/early April 2004
when the largest and deepest pit was dug (3 m x 1
m X 1.8 m [in part]). Calculations produced a figure
of c. 100 specimens for each square metre of surface.
Specimens were absent from the dry A horizon (c.
20 cm deep) but were numerous at all depths of the
moist B horizon down to 180 cm. The greatest
density occurred in the 60-90 cm deep zone (101
specimens/m-’) and the 160-180 cm deep zone (100
specimens/m-'’>. The size/frequency distribution of
this sample is represented in Figure 28.

At the Horrocks site in August 2003, the author
excavated seventeen Im x Im pits to a depth of at
least 40 cm at various locations around the farm to
check for the presence of sandgropers. All were in
deep yellow sand under pasture. In one paddock,
only C. kochii was found. In an adjoining paddock,
mainly Species A was found with an occasional C
kochii. There was very little observable difference
between these two paddocks in terms of soil and
pasture cover. Several excavations in a paddock
situated in a vale produced no specimens at all.
Clearly, the distribution of sandgropers is patchy in
seemingly suitable habitat, but what factors
determine the presence or absence of these insects
has yet to be determined.

Food and feeding

Examination of the gut contents of 62 winter-
collected and 100 summer-collected specimens of
C. kochii and C. tindalei revealed that they had
consumed a diversity of materials, most of it being
of vascular plant origin although insect and
arachnid remains were also identified in many
specimens. Fungal tissues, including hyphae,
sporangia and spores, were present more often than
not, but mostly in small quantities. Sand grains, too,
were almost always present throughout the
intestine but, comprising only a minor component
of gut contents, were probably accidentally ingested
with the food. The food was well masticated and
finely divided, so identification sometimes required
comparison of tissues at the cellular level under a
compound microscope.

Sloughed peritrophic membranes were always
present in the gut and enclosed the food, regardless
of the quantity of the latter.

Ingested plant material consisted mainly of
underground parts (roots and stolons) but also

comprised aerial parts such as stems and leaves of
grasses (including cooch, wheat and barley),
dicotyledonous leaves (e.g.. Cape Weed), floral
bracteoles of Asteraceae, and seeds of several kinds.
Most seed tissue was not identified but several
specimens of C. kochii from the ITorrocks and
Dandaragan sites had eaten seeds of 'double-gee'
(Emex australis), a pest weed in these areas.
Double-gee seeds are contained in hard, spined
fruits which the insects evidently chew open.

Plant material varied from fresh (e.g., white
rootlet or chlorophyll-containing leaf tissue) to old
and partly decomposed (brownish tissue
containing lots of fungal hyphae and spores). The
presence of chlorophyll-containing leaf tissue
matching that of wheat leaves in the intestines of
sandgropers collected from wheat fields could be
taken as convincing evidence that the insects
damage wheat as reported by Richards (1980).
There is some possibility, though, that sandgropers
may simply be availing themselves of stems and
leaves pulled into the soil by cutworms (noctuid
moth larvae) rather than being primary pests. On a
farm east of Northampton, the author examined a
patch of barley crop purportedly thinned by
sandgropers. Numerous young barley plants had
turned yellow and many loose stems and leaves
were found partly pulled into the soil. Excavation
around these damaged plants yielded not
sandgropers but numerous pink cutworm larvae
[Agrotis munda Walker). These cutworms are
reported to cause the kind of damage observed
(Common 1990).

Fungal tissue in gut samples consisted mostly of
rusts, saprophytic and mycorrizal fungi probably
ingested with root, stem and leaf tissue. In a few
samples, though, significant amounts of fungal
tissue suggested direct browsing, one such sample
containing VAM (v'esicular-arbuscular mycorrhiza)
spores (Dr Neale Bougher, pers. comm.).

A variety of invertebrates were identified among
gut contents (Table 2). In most cases, the remains of
only one or two insects were present. However, two
adults of C. tindalei had consumed numerous
worker termites, clearly demonstrating purposeful
predation rather than accidental ingestion. Most of
the listed invertebrates are likely soil inhabitants,
even the native bee. Six or more insect egg chorions
about 4 mm long and possibly from acridid
grasshopper eggs were found in the gut of one
adult male of C kochii.

Cannibalism was encountered in the very dense
population of C. kochii at the Dandaragan site
among summer-collected specimens (see Table 2).
Second and older instar nymphs and adults had
consumed first instar nymphs which formed the
bulk of the population at the time. In several
individuals, the gut contents included fragments of
both the front and hind ends of the prey, providing

229

Biology of sandgropers

Table 2 List of arthropod food items identified among the gut contents of Cylindraustralia kochii and C. tindalei and
the numbers of specimens in which they were found.

Food item

C. kochii

C tindalei

Dermaptera

1

Isoptera - workers

2

4

?Orthoptera; ?Acrididae - eggs

1

Orthoptera: Cylindrachetidae - nymphs

17

Hemiptera: Fulgoroidea

1

Diptera: Mycetophilidae - larva

1

Diptera: Sciaridae - adult

1

Diptera; Sciaridae - larva

1

Diptera: Cylcorrapha - larva (T' instar?)

1

Lepidoptera; Noctuidae, Agrotis - larva

1

Lepidoptera: unidentified larva

3

Coleoptera: Scarabaeidae, Melolonthinae - adult

3

1

Hymenoptera; Formicidae - worker

10

2

Hymenoptera: Colletidae, Dermatohesma - adult

1

Araneae

1

Acarina

2

3

unidentified chitinous remains

7

convincing evidence of predation as opposed to
accidental ingestion.

Evidence that first instar nymphs consume some
or all of their eggshells was found in 11 specimens
collected at Dandaragan in January and March/
April. Among the gut contents were fragments of
fenestrated membrane (consistent with the outer
layer of sandgroper egg chorion) plus large
numbers of colourless, refractive, spherules
(diameter ca. 0.01 mm). Similar spherules occur in
clusters in the inner layer of the egg chorion.
Further evidence was obtained when four newly
hatched first instar nymphs were maintained in
glass vials with their eggshells. Torn edges of the
chorions, at first entire, became distinctly serrated

and eroded due to the feeding activity of the
nymphs.

When first instar nymphs leave their egg
chambers deep in the soil, their most likely food
source would be the very fine roots found to lace
the soil there.

In terms of gut contents, there were some notable
differences between specimens collected in 'winter'
(May to September; see Table 3) and those collected
in 'summer' (December to April; see Table 4): 93%
of winter specimens had eaten plant material
compared with only 13% of summer specimens.
Only 44% of summer specimens showed evidence
of recent feeding and 57% of those had eaten an
insect (in 18 of 25 cases, another sandgroper). The

Table 3 Summary of gut contents of sandgroper specimens collected during 'winter' months (i.e. late April to
September) from various localities in south-western Australia.

Species

No. of specimens examined

Numbers of specimens that had eaten certain items

Plant material

Seed material

insect/mite

C kochii

45

42

24

12-^

C. tindalei

17

11

3

10

Table 4 Summary of gut contents of specimens of Cylindraustralia kochii collected in 'summer' months (i.e., December
to early April) from the Dandaragan site. Cannibalism is represented in the column headed 'Sandgroper'. The
column headed 'Egg chorion' refers to first instars that appear to have consumed their own egg chorion after
hatching. For an explanation of the right-hand column, see under Predators, Parasites and Associated
Organisms - Amoebae.

Sample n Numbers of specimens that had eaten certain items

Any Plant Sand Other Egg Number with amoebae

food matter groper insects chorion and/or rectal convolutions

Early Dec.

40

6

6

1

2

3?

28

Late January

30

20

5

13

2

2

22

March/Apri!

30

18

2

4

3

9

10

Combined

100

44

13

18

7

14

60

230

T.F. Houston

low incidence of plant-feeding during summer
might suggest that the insects avoid the dry surface
layers of the soil where most of the grass and herb
roots occur. In summer, too, cooch grass was the
only live plant at the study site. Yet, among the
plant material consumed by some summer-
collected specimens were grass-leaf and seed
tissues. Their presence in specimens collected at
depths of 60-95 cm suggests that those individuals
had recently ventured to or near the surface to feed.

The occurrence of amoebae in the gut of summer-
collected specimens and the seemingly associated
condition referred to here as 'rectal convolutions' is
discussed in detail below under Predators, Parasites
and Associated Organisms.

Faecal pellets observed in the rectum, were
usually solid, roughly cylindrical, and enclosed in
peritrophic membrane.

The gut contents of two adult females of Species
B consisted mostly of various plant tissues along
with small amounts of arthropod chitin. Gunther
(1992) and Tindale (1928) recorded plant tissue and
insect chitin in the alimentary tracts of a further
three species of Cylindraustralia, so omnivory is
clearly widespread in the genus.

Annual Life Cycle and Development

Adults and nymphs of a broad range of sizes were
present in population samples of C. kochii collected
throughout the year (Figure 28). From the
histograms it will be seen that the first and second
nymphal instars were by far the most numerous
stages present in each sample. The third and fourth
instars, by contrast, were very poorly represented,
being scarcer even than the fifth instar. As revealed
by Table 5, laid eggs were found in the soil at
intervals throughout most of the year. The
occurrence of eggs and early instars through most
months of the year initially suggested the
possibility of year-round breeding in C. kochii. This
possibility, however, is not supported by other
observations.

Dissection of adult females collected from May to
August revealed that most carried eggs ready to lay
in the oviducts or at least had near-mature ova in
the ovaries. For example, five adult females

ploughed up on 10 May 2003 all carried eggs ready
to lay. By contrast, the ovarioles of the only four
females collected in summer (late January and late
March) had no ova near egg-size. Instead, each
ovariole contained only a series of very small to
minute ova. Additionally, the spermathecae of all
four females were devoid of sperm. Thus it is likely
that these females were very young, pre-
reproductive individuals.

At the Dandaragan site, freshly laid (translucent)
eggs were found only during the late May and
June-July visits. All eggs found later than July
through to December were opaque and showed no
signs of embryological development. Developed
and hatching eggs were only found in January.

Several apparently freshly laid eggs collected in
June/July were maintained in the laboratory for
several weeks during which time they turned
opaque and some succumbed to mould but none
hatched. A few were opened to check for signs of
embryological development but none was found.

In January 2004, a number of opaque eggs were
excavated, some showing signs of development (eye
spots and legs vaguely visible through the chorion).
On this same occasion, empty egg-shells were
found along with tiny, clearly newly emerged
nymphs in several chambers. A number of eggs
hatched over subsequent days. During a March-
April excavation at the same site, only one (opaque)
egg was found.

In October 2004, 32 opaque eggs were excavated
at Dandaragan and returned to the laboratory.
Although a few succumbed to mould attack, turned
black and/or shrivelled, most eggs remained
outwardly unchanged until late February 2005.
Four eggs hatched between 24 February and 1
March 2005 and several more probably would have
hatched had they not been dissected to check for
embryological development. The first such
dissections were on 17 January: two eggs contained
small embryos and another a live, almost fully
developed larva. On 3 February, a number of eggs
were wet with distilled water to varying degrees
and over varying periods from one day to two
weeks to see if this might induce eclosion.
However, these treatments were ineffectual. Eleven

Table 5 Dates when eggs of Cylindraustralia kochii were excavated from soil.

Month (days), year

Location

Comments

January (28-31), 2004

Dandaragan

Many, opaque, with embryos or hatching

March (29)-April (1), 2004

Dandaragan

One, opaque

May (28-30), 2003

Dandaragan

Many, translucent

June (30) -July (2), 2003

Dandaragan

Many, translucent

August (20-26), 2003

Horrocks

Two

October (27-29), 2004

Dandaragan

Many, opaque

November (16), 2002

Mullaloo (Perth)

Two, opaque

December (3-6) 2003

Dandaragan

A few, opaque

Biology of sandgropers

231

eggs remaining unhatched on 14 March were
dissected and, while life had expired in all of them,
embryological development had proceeded to
varying stages in several and two contained fully
formed larvae.

Four first instar nymphs reared from eggs in late
February/early March 2005 were maintained alive
in moist soil with germinating mixed budgie seed.
They thrived (as evidenced by their increasingly
large abdomens) but succumbed to disease one bv
one, the last surviving for seven months. None
moulted to the second instar.

Taking account of the above data, it seems likely
that oviposition occurs from May to August; the
egg chorion is shiny and translucent at first but
gradually turns dull and opaque; eggs remain
dormant until mid-summer when they develop and
hatch. If it is the norm that hatching is restricted to
mid-summer, then the year round presence of first
instar nymphs suggests that this stadium endures
for at least twelve months. If each instar were to be
equally long-lived, the whole life cycle of C. kochii
would extend over at least five years.

The scarcity of third and fourth instars in most
population samples is difficult to explain. Only in
the January and March-April 2004 samples were
significant numbers of third instar nymphs present
(Figure 28). If, as it seems, the life cycle occupies
several years, then the absence or scarcity of a
particular stage in the population could simply
reflect a past year in which fewer eggs were laid or
in which mortality of early stages was heavy. In
order to gain a clearer and more reliable picture of
population structure and change through the year,
it will be necessary to gather larger samples. In this
study, excavation by spade greatly restricted the
area of soil that could be turned over, especially at
greater depths. Additionally, it is possible that
vibrations caused by digging might have caused
some larger specimens to flee the excavation sites
via existing galleries. Rearing specimens in captivity
will also be necessary to determine longevity in the
various instars and reliable data on longevity is
necessary to interpret population structure.

Fecundity

Females have ten ovarioles per oviduct. Although
a maximum of 14 eggs ready to lay were found in
one individual (kept captive in a small container of
soil for several weeks and therefore prevented from
ovipositing) no more than seven were found in
several other adult females. As each egg is laid
singly in its own chamber, the rate of egg
production must be comparatively low. What is not
known is how long females go on ovipositing and
how many eggs they would lay in their lifetime.

Mating

No observations of mating were made. Attempts

to induce copulation by placing pairs of adults
together in small containers proved unsuccessful.
Flowever, examination of the copulatory organs of
freshly killed adults strongly suggests that mating
individuals must come together 'tail' to 'tail' and
venter to venter (somewhat as in Figure 20). The
phallus cannot be exserted very far and has little
flexibility. By making contact as in Figure 20, the
hooks on the paraprocts of the male could engage
the hind margin of sternite 8 of the female, pulling
it down to open the vagina and the phallus would
be orientated at just the right angle to permit
intromission.

Copulation could hardly occur within the
confines of normal galleries but it might occur in
the widened sections of galleries noted under
Burrowing above. Alternatively, copulation might
occur on the surface of the ground. To check this
possibility, nocturnal searches by torch-light were
undertaken where sandgropers were known to be
present in dense populations. Searches were made
in both wet and dry weather conditions but no
surface activity was encountered.

Predators, Parasites and Associated Organisms

According to several farmers, 'crows' (actually
ravens) gather in flocks to predate on sandgropers
turned out of the soil during ploughing of pastures.
Johnstone and Storr (2004) recorded sandgropers
from the guts of the Australian Raven. Farmers also
report that foxes dig sandgropers from their surface
trails and one observer noted the remains of
sandgropers in fox scats.

This study found no evidence that sandgropers
(either adults or immatures) are subject to attack by
insect predators or parasitoids. If truly free of such
attacks they would be a rarity among the insects.
Evidently, their wholly subterranean existence,
perhaps combined with their very hard
integuments (anteriorly) and their chemical
defences, serve to shield them from such enemies.

Grcgarines

The mid guts of many specimens of C kochii and
C. tindalei were found to contain white bodies up
to 2 mm long which superficially resembled insect
ova or maggot-like larvae. These proved to be
'gamonts' of protistan parasites of the genus
Gregarina, class Apicomplexa (formerly Sporozoa).
They were present in varying numbers, rarely just
one or two, frequently dozens and occasionally
hundreds when they packed the lumen of the
midgut. Another stage in the life cycle of these
organisms, the spherical 'gamontocyst', was
observed frequently in faecal pellets in the rectum.
Gregarines were found in both adults and nymphs
of various sizes from all study sites. Their incidence
was comparatively low among dissected specimens
collected from May to August, 16 of 50 C. kochii

232

T.F. Houston

and 3 of 18 C. tindalei being infested. Their
incidence was very much higher in the early
December sample of C kochii from the Dandaragan
site, 32 of 38 dissected specimens being infested.
However, at the same site, only 3 of 30 dissected
specimens from late January and none of 30 from
late March/early April were infested. This dramatic
reduction could be correlated perhaps with the
apparent cessation or significant reduction of
feeding observed in summer populations (see under
Food and Feeding).

Amoebae

Specimens of C. kochii collected from the
Dandaragan site in summer months exhibited
another protistan occupant of the midgut: gold-
coloured, single-celled organisms tentatively
identified as amoebae. These occurred in varying
numbers from just a few up to hundreds in the
peritrophic membranes of the mid gut (made more
visible by the absence of food material). It appeared
that the amoebae did not survive their passage
through the gut. Instead, they broke down in the
posterior part of the mid gut or in the hind gut
where they became concentrated in mucus-like
material in narrow peritrophic membrane tubules.
In the rectum, the tubules became convoluted and
compacted into soft, translucent, honey-coloured
pellets. Specimens whose rectal contents consisted
only of convolufed tubules almost invariably had
amoebae in the mid gut and their intestines were
either devoid of food material or contained only
minor quantities. Convoluted tubules were noted in
60% of summer-collected specimens (see Table 4 for
details). These observations suggest that infestation
of the gut by amoebae is associated with (perhaps
even causes) a cessation of feeding. Given the
absence of feeding, the amoebae are possibly
ingested through the imbibition of ground-water
(made possible by heavy summer rains). Because
the amoebae appear not to survive their passage
through the sandgropers, they cannot be considered
to be parasites.

Nematodes

Nematodes identified as 'dauers' (non-feeding,
resting or dispersal stage larvae) of the family
Rhabditidae and possibly the genus Rhabditis (Dr
Kerrie Davies, pers. comm.) proved to be common
occupants of the genital chambers of C kochii and
C. tindalei in the northern parts of their ranges
(north of the latitude of Geraldton). No such genital
occupants were found in specimens south of
Geraldton. In one specimen of C. kochii, dauers
occurred also in a depression of the fore femur.

Only about 0.5 mm in length, dauers frequently
formed tightly packed masses comprised of dozens
or even hundreds of individuals beneath the
phallus of male hosts. Dauers were also found in

the vaginas of three adult females. When a freshly
killed male sandgroper was dissected in saline
solution, the nematodes were at first still but, on
being disturbed with a needle, quickly became
active, flexing their bodies strongly back and forth
and dispersed in the saline. In some cases, however,
a few to many of the nematodes were dead, brown
and stiff.

Only occasional nematodes were encountered
among gut contents and were possibly accidentally
ingested with the food. None were encountered
elsewhere among the internal organs of the insects.
At the Dandaragan site, two egg chambers
contained dead detached eggs with clusters of
nematodes of various sizes on, in and around the
latter. These nematodes were identified as bacterial-
feeding cephalobids (common soil inhabitants) and
an unidentified species, possibly Mesorhabditis (Dr
Kerrie Davies, pers. comm.). Neither kind
represented the same species as the dauers in the
sandgropers' genital tracts.

As sandgropers carried dauers only in part of
their range and no other part of this particular
nematode's life cycle was found to be closely
associated with the insects, the nematodes mav
simply be using them as dispersal agents. Questions
remaining unanswered are - how do so many
dauers find their way into the genital chambers,
where do they come from and are the dauers
transferred between the sexes during copulation?
Sexual transmission of nematodes has been
reported to occur in certain other orthopterans (e.g.,
Luong et ai. 2000).

Mites

Phoretic deutonymphs (non-feeding, dispersal
stage nymphs, also known as hypopi) of six species
of mites were found externally on a number of
individuals of C. kochii and C. tindalei. Thev
occurred, sometimes singly, sometimes clustered,
on various sheltered parts of the body: inner sides
of fore legs, flanks of abdomen beneath mid and
hind femora, and in folds of abdominal segments.
These mites were identified bv Dr Barry O'Connor
(pers. comm.) and their names and host associations
are listed in Table 6. Dr O'Connor noted that some
members of unnamed genus 1 are associated with
termites in the USA and central America while
unnamed genus 2 is similar to taxa (e.g., Forcellinia)
associated with ants and termites.

Fungi

Many dead eggs were found in chambers during
excavation at the Dandaragan site in July and most
of these were heavily coated with various kinds of
fungi. Even seemingly fresh, suspended eggs often
had fungal hyphae (bright yellow, black or
colourless) growing over their surfaces and some
were dotted with fungal sporangia.

Biology of sandgropers 233

Table 6 Mites recorded from the bodies of sandgropers in the present study.

Mite taxa

C. kochii

C. tindalei

Order Acariformes: Suborder Astigmata

Acaridae - unnamed genus 1

+

Acaridae - unnamed genus 2, species 1

+

Acaridae - unnamed genus 2, species 2

+

Acaridae - Sancassania sp.

+

Histiostomatidae, Histiosoma sp.

+

Order Parasitiformes: Suborder Mesostigmata

Ascidae? (Lasioseius?)

+

Defences

No biting or other defensive behaviours were
observed while handling specimens except that,
when restrained, the insects sought to 'burrow'
their way to freedom with their powerful fore legs.
When exposed during excavation, the insects
always attempted to burrow back into the soil or
withdrew into their galleries.

The characteristic odour produced by
sandgropers (see above under Odour glands)
probably serves a defensive function.

CONCLUSION

Many Tridactyloidea are heavily dependent on
fresh-water bodies for their survival. Some 'pygmy
mole crickets and mud crickets' (Tridactylidae and
Ripipterygidae, resp.) inhabit the margins of lakes,
streams and rivers, often in humid tropical
environments, where they burrow and feed in the
damp surface layers of mud or sand (Gunther 1994).
The Argentinian cylindrachetid, Cylindroryctes
spegazzinii (Giglio-Tos), lives in the gritty shores of
lakes and associated rivers (Gunther 1992).
Cylindraustralia species, however, live well away
from free water and many inhabit semiarid to arid
habitats. Nevertheless, the present study has
indicated that they are still dependent on soil
moisture and no specimens were ever found in
truly dry soil.

Despite the gains from the present study, many
basic questions concerning cylindrachetid biology
remain to be answered, even for the principal
subject C. kochii. For example, how long is the
complete life cycle? How long do adults survive?
Where, when and how do they mate? How many
eggs does a female produce in her lifetime? Are
there any insect predators or parasitoids not found
in this study? At what rates do sandgropers burrow
near the surface and at depth? Do they continually
burrow into fresh soil or do they (at least at times)
return to home burrows? Do they exhibit daily
patterns of activity?

ACKNOWLEDGEMENTS

This study was made possible by the generous

assistance of many individuals. While I cannot
name all of those who provided me with
specimens 1 am deeply grateful to the following
for their interest, support, and hospitality during
work on their properties: Paul Bloomer of
Mullaloo (Perth); Ian and Denise Edgar of
Annamullah farm, Dandaragan; Ross and Nola
Johnson of Willi Gulli farm, Horrocks; Margaret,
Bruce, and Keith Quicke of Eurardy Station; and
Maurice and Dezi Webb, Chapman Valley east of
Northampton. Voluntary assistance was
generously provided by Otto Mueller (fieldwork)
and Nihara Gunawardene (data entry and
specimen mensuration). For the identification of
organisms found in association with sandgropers
and for information about them I am indebted to
the following: Kerrie Davies, Waite Campus,
University of Adelaide (nematodes); Barry
O'Connor, University of Ann Arbor, Michigan,
(mites); Neale Rougher, CSIRO Forestry, Perth,
and Elaine Davison, Curtin University of
Technology (both fungi). For encouragement,
information and suggestions of useful references I
wish to thank Kurt Gunther, David Rentz, and
Marlene Zuk. David Ragge provided guidance on
numbering the nymphal stages. David Rentz and
Winston Bailey kindly read an early draft of this
paper and made useful suggestions for its
improvement.

REFERENCES

Barrett, C. (1928). Notes on Cylindracheta. Victorian
Naturalist 44: 266-267.

Common, I.F.B. (1990). Moths of Australia, Melbourne
University Press, Carlton, Victoria.

CSIRO (1991). The Insects of Australia. A textbook for
students and research workers (Second edition),
Melbourne University Press, Carlton, Victoria.

Flook, P.K., Klee, S. and Rowell, C.H.F. (1999). Combined
molecular phylogenetic analysis of the Orthoptera
(Arthropoda, Insecta) and implications for their
higher systematics. Systematic Biology 48: 233-53.
Gambardella, L.A. de (1971). Oviposicion y eclosion de
Rhipipteryx notata Burm. 1838. Revista Peruana de
Entomologia 14: 282-285 (cited by Gunther (1994); not
seen).

Gunther, K.K. (1992). Revision der Familie

234

T.F. Houston

Cylindrachetidae Giglio-Tos, 1914 (Orthoptera,
Tridactyloidea). Deutsche Entomologische Zeitschrift
39 (4/5): 233-291.

Gunther, K.K. (1994). Die Tridactyloidea-Fauna
Kolumbiens (Orthoptera, Caelifera). Deutsche
Entomologische Zeitschrift 41(1): 45-56.

Johnstone, R.E. and Storr, G.M. (2004). Handbook of
Western Australian birds. Volume II. Passerines
(blue-winged pitta to goldfinch). Western Australian
Museum, Perth.

Kevan, D.K.M. (1989). Grigs that dig and grasshoppers
that grovel. Revue d'Ecologie et de Biologie du Sol 26:
267-289.

Lawrence, J.F. and Britton, E.B. (1994). Australian
Beetles, Melbourne University Press, Carlton,
Victoria.

Luong, L.T., Platzer, E.G., Zuk, M. and Giblin-Davis,
R.M. (2000). Venereal worms: sexually transmitted
nematodes in the decorated cricket. Journal of
Parasitology 86: 471^77.

Rentz, D.C.F. (1996). Grasshopper Country: the abundant
orthopteroid insects of Australia. University of New
South Wales Press, Sydney.

Richards, K.T. (1980). The sandgroper - a sometimes not-
so-friendly Western Australian. Journal of
Agriculture - Western Australia 21: 52-53.

Schremmer, F. (1972). Ein Massenvorkommen von
bodenbewohnenden Rhipipteryx forceps
(Tridactyloidea) in den kolumbianischen Anden.
Pedobiologia 12: 317-322 (cited by Gunther (1994);
not seen).

Tindale, N.B. (1928). Australian mole crickets of the
family Gryllotalpidae (Orthoptera). Records of the
South Australian Museum 4: 1^2.

Urquhart, F.A. (1973). Some notes on the sand cricket
[Tridactylus apicalis Say). Canadian Field-Naturalist
51: 28-29 (cited by Gunther (1994); not seen).

Uvarov, B. (1966). Grasshoppers and Locusts, a
Handbook of General Acridology, vol. 1, Cambridge
University Press, Cambridge

Wiley, T. (2000). Insect posts of the perennial fodder
shrub tagasaste. Farmnote 48/2000, Western
Australian Department of Agriculture, Perth.

Zuk, M. (1987). The effects of gregarine parasites on
longevity, weight loss, fecundity and developmental
time in the field crickets Gryllus veletis and G.
pennsylvanicus. Ecological Entomology 12: 349-354.

Manuscript received 7 December 2005; accepted 10 July 2006

Records of the Western Australian Museum 23: 235-240 (2007).

Temporal variation in ground-dwelling invertebrate
biomass in the Goldfields of Western Australia

Scott A. Thompson^'^ and Graham G. Thompson^'^

ATA Environmental, Dilhorn House, 2 Bulwer Street, Perth,

Western Australia 6000, Australia (present address)

Centre for Ecosystem Management, Edith Cowan University, Joondalup Drive,
Joondalup, Western Australia 6027, Australia
^Corresponding author. Email: g.thompson@ecu.edu.au.

Abstract - We examined temporal variation in invertebrate biomass based on
pit-trapping data from the semi-arid goldfields region of Western Australia
(W.A.). Invertebrate dry biomass varied significantly among taxa, seasons
and from year-to-year. There was a peak in dry biomass for all taxonomic
groups from September to January that was followed by a significant decline
for most families by April, and invertebrate biomass was lowest in mid-
winter. For Araneae, Blattodea, Scorpionida and Chilopoda there was a
significant rapid decline by April, whereas for Coleoptera, Orthoptera and
Isopoda the rate of decline was slower. Other than during the winter survey,
the dry biomass of Formicidae was unchanged. Chilopoda and Blattodea
constituted the highest proportion of the biomass captured and the dry mass
of individuals from these taxa was generally higher than that for the other
invertebrate taxa. There was a positive correlation between invertebrate
biomass and the number of reptiles caught but not with the number of
mammals caught.

INTRODUCTION

Invertebrates are important in the functioning of
nearly all natural environments and a change in
their diversity and abundance can potentially affect
the whole ecosystem. One potential impact that
stems from their importance is as a prey source for
reptiles, some small mammals (e.g., dasyurids) and
many birds. Invertebrates are also increasingly
being used as indicators of rehabilitation success in
a variety of situations (Andersen 1994; Bisevac and
Majer 1998, 1999a, 1999b; Lobry de Bruyn 1999;
Madden and Fox 1997; Majer and Brown 1997;
McGeoch 1998). All too often, one-off surveys are
undertaken for terrestrial fauna at a particular site,
and the data are used to characterise species
richness and abundance or to provide benchmarks
against which future impacts on the faunal
assemblages are assessed. Such surveys pay little
attention to temporal variation (Cowan and How
2004; Thompson and Thompson 2005a).

Our objective here was to: a) characterise seasonal
patterns in invertebrate biomass; b) compare
seasonal patterns in the Goldfields with those that
have been described elsewhere in W.A.; and c)
relate invertebrate biomass to the activity of
insectivorous vertebrates.

METHODS AND SITES

We sampled the invertebrates on six occasions
(Dec 2000, Jan 2001, April 2001, June 2001, Sept
2001 and Dec 2001) in nine undisturbed sites near
Ora Banda (30°27'S, 121°4'E; approximately 50 km
north of Kalgoorlie, W.A.) to establish the annual
cycle of variation in biomass.

Ora Banda lies on Archaen granites that underlie
lateritic gravel soils. The vegetation was
heterogenous, ranging from Eucalypt-Casuarina-
Mulga woodlands interspersed with Acacia, to
sparsely distributed spinifex {Triodia spp.) and
shrubs (Acacia spp.) to dense shrubs (Acacia spp.,
Atriplex spp., Allocasuarina spp.). The nine
undisturbed areas were located in different habitats
based on major vegetation types identified for the
area by Mattiske Consulting (1995).

Other researchers sampling invertebrates have
used small pit-traps filled with a preservative
(Andersen et al. 2003; Bisevac and Majer 1999a,
1999b; Brennan et al. 1999). However, based on
results of a pilot trial in September 2000, larger
invertebrates (e.g., beetles, centipedes, spiders)
were not easily caught in small diameter (~
40mm) vials filled with a preservative. Therefore,
20 L pit-trap buckets without a preservative were

236

S.A. Thompson, G.G. Thompson

used for surveying the ground-dwelling
invertebrates.

Eight pit-trapping lines, each containing three 20
L PVC buckets and three 150mm PVC pipes
(600mm deep) that were used as pit-traps, were
alternated and evenly spaced along 30m long fly-
wire drift fences (250mm high) at each of the nine
study sites. Each trapping line was approximately
20m apart. Invertebrates were collected daily using
forceps from each of the 20 L pit-trap buckets, for
six days for each of the six survey periods.
Invertebrates were not collected from the PVC pipe
pit-traps. These same pit-traps were used for
surveying reptiles and small mammals (see
Thompson et al. 2003; Thompson and Thompson
2005b). We appreciate that captured vertebrates or
other invertebrates may have eaten some of the
invertebrates caught in these pit-traps, but we
believe that the number destroyed would be low
compared to the total abundance, as pit-traps were
cleared each morning. Small invertebrates, and in
particular ants that died and dehydrated in the pit-
traps, were very difficult to collect at the bottom of
the buckets. As a consequence, we would have
under-sampled the very small invertebrates;
however, this will have had little consequence on
our estimate of temporal variation in terrestrial
invertebrate biomass given the magnitude of these
variations.

All invertebrates collected were initially
preserved in 70% ethanol. The preserved
invertebrates were later sorted into the following
groups: Formicidae (ants); Coleoptera (beetles);
Chilopoda (centipedes); Blattodea (cockroaches);
Orthoptera (grasshopper and crickets); Isopoda
(slaters); Scorpionida (scorpions); Araneae (spiders)
and others. These particular invertebrate groups
were chosen because they are easily identified in
reptile and mammal stomach contents and were
consistent with how other authors report reptile
stomach content data (e.g., Pianka 1986).

The invertebrate catch from all 24 PVC buckets at
each site on each day were grouped. All
invertebrates were removed from alcohol, placed
into vials and dried for four days at 35°C in a
controlled temperature room under fan-forced
airflow. Four days was sufficient to remove the
moisture and reach a constant mass. The time taken
to dry invertebrates was tested after the first survey
period in December 2000. There was no change in
invertebrate mass between the third and fourth
days of drying. The number of individual
invertebrates in each group was counted and all
samples were weighed to four decimal places using
electronic scales.

Data analysis

The dry biomass of invertebrates was used for all
calculations. Variation in invertebrate biomass

among the six survey periods was examined using
a full factorial ANOVA [Biomass (g/24 pit-trap
nights) = site + taxonomic group -i- survey period +
site’^group -i- site^survey period + taxonomic
group*survey period + site*taxonomic
group^survey period] using statistiXL (http://
www.statistiXL.com/; V.I.5). Variance in the
ANOVA model came from differences among the
six days of data for each site for each survey period.

Correlation coefficients were calculated to
demonstrate relationships between the total
invertebrate dry biomass for invertebrate group and
the number of reptiles and mammals caught during
each survey period. In addition, the number of
small dasyurids caught was correlated with
invertebrate dry biomass, as they are almost
exclusively insectivorous. Significance level at a =
0.05 was used for all analyses.

RESULTS

During six surveys (7776 pit-trap nights) from
December 2000 to December 2001, 15069 individual
invertebrates were captured with a total drv body
mass of 1402.4 g. Chilopoda, Blattodea and Araneae
constituted more than two thirds of the inv'ertebrate
biomass captured (Table 1). The 'other' group,
which accounted for 9.8% of the total biomass,
included gastropods, mantids, earwigs, stick-
insects, moths and larvae. Coleoptera had a higher
dry mass than Orthoptera, Isopoda and
Scorpionida. The average dry mass for individuals
was highest for Chilopoda and Blattodea (Table I).
Counting many of the small dehydrated ants
proved to be an impossible task and their
individual body mass was not able to be accurately
assessed, but it was probably the lowest of the
groups assessed.

Among seasons vaiiation in invertebrate biomass

There was a significant difference among factors
(e.g., seasons, taxonomic groups and sites), and a

Table 1 Average mass of individuals and proportion
of the biomass represented by each group of
invertebrates

Group Mean individual Percentage

dry mass (g) of biomass

Formicidae 3 27

Coleoptera 0.121 11.87

Chilopoda 0.392 25.67

Blattodea 0.323 21.42

Orthoptera 0.076 4.42

Isopoda 0.020 1.78

Scorpionida 0.093 2.50

Araneae 0.090 19.30

Other Q nn

(g/24 pit-trap nights)

Temporal variation in invertebrate biomass

237

« E
® 0)
2 H

Figure 1 Mean dry biomass for each invertebrate
group for each of the six survey periods
showing means ± 1 se.

significant interaction among factors for
invertebrate biomass (Table 2; Figure 1).

A post-hoc Tukey test after the full factorial
ANOVA on all survey periods showed that
December 2000 and January 2001 had a higher
invertebrate biomass than all other survey periods,
including December 2001 (Table 3). June 2001 had
the lowest invertebrate biomass. December 2001
biomass was significantly less than December 2000
but the same as for April 2001 and September 2001
(Table 3). A post-hoc Tukey test indicated that
Coleoptera and Araneae were the only two families
that were significantly different between the two
December surveys.

There was no significant difference between the
biomass for any taxonomic group for the December
2000 and January 2001 survey periods. With the
June catch removed from the dataset, the biomass
of Formicidae was not significantly different among
the seasons (T, „5 " ^.80, P = 0.53), whereas for all
other taxonomic groups there were significant
differences in the dry biomass among seasons
(Figure 1). Chilopoda, Blattodea, Scorpionida and
Araneae all showed a significant reduction in dry
biomass from the January to the April survey
period. The decline from January to April was not
significant for the other invertebrate groups, but it
was significant from January to June. The variation
in biomass for Orthoptera and Isopoda, and
Chilopoda and Blattodea across the six survey
periods were similar (Figure 1). Araneae dry
biomass was high for the December 2000 and
January 2001 survey periods, then remained low for
the subsequent periods.

Abundance of reptiles and mammals relative to
invertebrate biomass

There was a positive correlation between the
biomass of Coleoptera, Chilopoda, Blattodea,
Scorpionida and Araneae and the number of
reptiles caught during each survey period, but the
total number of mammals and the number of
dasyurids caught were not significantly correlated
to the dry invertebrate biomass for any of the
families (Table 4).

Table 2 ANOVA results for comparison of variation in invertebrate biomass among sites, seasons and taxonomic
groups.

Source

SS

Df

MSq

f-value

P value

Site

41.235

8

5.154

4.403

< 0.0001

Season

258.906

5

51.781

44.234

< 0.0001

Group

400.062

7

57.152

48.821

< 0.0001

Site*Season

90.821

40

2.271

1.940

< 0.0001

Site* Group

258.412

56

4.614

3.942

< 0.0001

Season* Group

257.958

35

7.370

6.296

< 0.0001

Site*Season*Group

399.909

280

1.428

1.220

0.011

S.A. Thompson, G.G. Thompson

238

Table 3 Variation in invertebrate mass (g/24 pit-trap nights) among survey periods. P values are from a post-hoc
Tukey test after an ANOVA. Bold F values represent a significant difference.

Survey period

Mean ± se

December

January

April

June

September

2000

2001

2001

2001

2001

December 2000

0.803 +0.129

January 2001

0.968 ± 0.139

0.22

April 2001

0.273 ± 0.040

<0.01

<0.01

June 2001

0.020 ± 0.006

<0.01

<0.01

<0.01

September 2001

0.443 ± 0.080

<0.01

<0.01

0.19

<0.01

December 2001

0.436 + 0.059

<0.01

<0.01

0.23

<0.01

-1.0

Table 4 Correlation between invertebrate family biomass and the number of reptiles and mammals caught during the
six survey periods. Bold F values represent a significant correlation.

Formicidae

Coleoptera

Chilopoda

Blattodea

Orthoptera

Isopoda

Scorpionida

Araneae

Reptiles

0.558

0.963

0.984

0.913

0.703

0.739

0.873

0.975

Mammals

0.414

0.510

0.203

0.214

0.041

-0.006

0.663

0.414

Small dasyurids

0.474

-0.152

-0.274

-0.188

-0.305

-0.165

0.111

-0.275

F values

Reptiles

0.250

0.002

<0.001

0.011

0.119

0.093

0.023

0.001

Mammals

0.415

0.301

0.699

0.684

0.939

0.992

0.152

0.415

Small dasyurids

0.342

0.774

0.599

0.721

0.557

0.755

0.834

0.599

DISCUSSION

In the south-west of W.A. there have been a few
phenological investigations of invertebrate activity.
Majer ancd Nichols (1998) reported that the number
of ants showed appreciable intra- and inter-specific
variation over a 14 year period in the forested areas
of south-western Australia, with detectable patterns
not clearly evident. Postle (1985) reported soil and
litter invertebrate numbers around Dwellingup in
the south-west of Australia being highest in autumn
and progressively declining to a low in December
before beginning to increase in February. In
contrast, Majer (1985b) and Majer and Koch (1982)
reported herbivorous invertebrate numbers were
negatively correlated with rainfall at sites at Perth,
Dwellingup and Manjimup in the south-west of
Australia with lowest numbers in winter, and
higher levels of activity in spring, summer and early
autumn. Predator insects at the Perth site were most
active from late autumn to early spring (Majer and
Koch 1982) and low in summer, whereas
invertebrate numbers were lowest at Dwellingup in
May and June, and at Manjimup in June and July
(Koch and Majer 1980). The invertebrate
decomposers were most active in winter and spring
at the two most northerly sites (Perth and
Dwellingup), but at Manjimup, they were most
active during summer (Koch and Majer 1980). At
Katanning in the wheatbelt to the east of these three
sites (e.g., Perth, Dwellingup and Manjimup), ants
were most active during the December to March
period (Majer 1985a). In the arid Tanami desert,
Paltridge and Southgate (2001) reported significant
fluctuation in invertebrate biomass between survey

periods, with the lowest catch rates being recorded
in winter.

Ora Banda is in the semi-arid Goldfields region of
W.A. and receives regular winter rain (May to July),
and thunderstorms and irregular heavy rain
resulting from decaying cyclones and low pressure
systems that cross the W.A. coast in the Pilbara
during late summer (Figure 1). Summer rain can
cause local flooding and leave ephemeral ponds for
weeks. Mean monthly maximum summer
temperatures are in the low 30s and drop to the low
20s in winter (Figure 1).

The most obvious general feature of invertebrate
biomass around Ora Banda was the higher biomass
for all families during the summer of 2000/01 and
the steady decline into winter and an increase in the
following spring. There was no difference in the
dry biomass for any taxonomic groups between
December 2000 and January 2001. For Coleoptera,
Chilopoda, Blattodea, Scorpionida and Araneae, the
very obvious peak (Figure 1) in dry biomass during
December-January was followed by a significant
decline by April and a further drop to June. For
Orthoptera and Isopoda, the rate of decline was
slower, but the dry biomass for these species was
very low in June. For Formicidae, there was no
difference among the five survey periods when the
June data were excluded. There was no difference
in dry biomass for any taxonomic group between
September and December 2001, but the overall
biomass was higher in December 2000 than in
December 2001. These data suggest that the biomass
of invertebrates increases rapidly at the end of
winter. It then remains the same from September to

Temporal variation in invertebrate biomass

239

January, and then declines to a low value in mid
winter. This is similar to that reported by Majer
(1985a) for the semi-arid wheatbelt and Paltridge
and Southgate (2001) for the arid Tanami Desert. At
other sites in the more mesic south-west of W.A.
the pattern seems more variable and perhaps linked
to foraging strategy and diet.

Reptiles were most active when the invertebrate
biomass was high. This might be expected as a
majority of the reptiles around Ora Banda eat
invertebrates and, for many, invertebrates are their
primary prey. However, many of the small
mammals caught (e.g., Cercartetus concinnus, Mus
musculus, Pseudomys bolami, P.
hermannsburgensis) either do not eat invertebrates
or they constitute only a small proportion of their
diet, and the activity patterns for these species is
probably not linked to invertebrate abundance. In
contrast, most of the small dasyurids are almost
exclusively insectivorous, and it might be expected
that their behaviour and activity patterns are linked
to invertebrate abundance. However, there was no
correlation between the number of small dasyurids
caught during each survey period and dry
invertebrate biomass. It would therefore be
expected that body condition of dasyurids around
Ora Banda would be lower in winter when
invertebrates were scarce, and they would put on
weight in summer because of the increased food
supply, and this would be when they are likely to
be reproductively active.

Chilopoda and Blattodea constitute the highest
proportion of the biomass captured, and the dry
mass of individuals was higher than for other
invertebrate taxa. Centipedes and cockroaches are
generally nocturnal and are therefore probably an
important prey source for many of the small
mammals in the area. Spiders are also relatively
plentiful and vary in dry body mass, providing a
range of prey sizes for reptiles, amphibians and
small mammals that prey upon them.

Given varying seasonal and year-to-year
fluctuations for different invertebrate taxa, the use
of the abundance of invertebrates as a bio-indicator
of ecosystem restoration should be undertaken with
considerable caution. In most circumstances where
a faunal assemblage is used as a bio-indicator, there
is a presumption that most of the variance in
abundance and species richness is directly related
to ecosystem development and not environmental
or variables unrelated to the restoration success
(Thompson and Thompson 2005b). A single
terrestrial survey of invertebrates is only able to
describe the assemblage for a particular period in
time, as relative abundance varies both seasonally
and from year-to-year. Therefore, in circumstances
where invertebrate monitoring data are used to
measure the success of a restoration area compared
with an adjacent undisturbed area, the two areas

must also be surveyed simultaneously. In most
circumstances, our current level of knowledge is
such that we cannot separate natural year-to-year
variation in invertebrate assemblages or biomass
from variations attributable to stochastic events
such as fire, grazing, drought or unseasonally heavy
or no rainfall.

ACKNOWLEDGEMENTS
This research was undertaken with ethics
approval granted by Edith Cowan University and
licences issued by the Department of Conservation
and Land Management. This research was
financially supported by OMG Cawse Nickel and
Barrick Kanowna, for which we are very
appreciative.

REFERENCES

Andersen, A. N. (1994). Ants as Indicators of Restoration
Success Following Mining in Northern Australia. In
1994 AusIMM Annual Conference. Australian Mining
Looks North - The Challenges and Choices. 5-9
August 1994. Australian Institute of Mining and
Metallurgy, Darwin.

Andersen, A. N., Hoffmann, B. D. and Somes, J. (2003).
Ants as indicators of minesite restoration: community
recovery at one of eight rehabilitation sites in central
Queensland. Ecological Management and Restoration
4 Sup: S12-S19.

Bisevac, L. and Majer, ]. D. (1998). Invertebrates as
success indicators for mine site rehabilitation. In C. J.
Asher and L. C. Bell (eds), Proceedings of the
Workshop on Indicators of Ecosystem Rehabilitation
Success 23-24 October. Australian Centre for Mining
Environmental Research, Melbourne.

Bisevac, L. and Majer, ]. D. (1999a). Comparative study
of ant communities of rehabilitated mineral sand
mines and heathland. Western Australia. Restoration
Ecology 7'. 117-126.

Bisevac, L. and Majer, J. D. (1999b). An evaluation of
invertebrates for use as success indicators for minesite
rehabilitation. In W. F. Ponder and D. Lunney (eds).
The other 99%: The conservation and biodiversity of
invertebrates: 46-49. Transactions of the Royal
Zoological Society of New South Wales: Sydney.
Brennan, K. E. C., Majer, J. D. and Reygaert, N. (1999).
Determination of an optimal pitfall trap size for
sampling spiders in a Western Australian Jarrah
forest. Journal of Insect Conservation 3: 297-307.
Cowan, M. A. and How, R. A. (2004). Comparisons of
ground vertebrate assemblages in arid Western
Australia in different seasons and decades. Records
of the Western Australian Museum 22: 91-100.

Koch, L. E. and Majer, J. D. (1980). A phenological
investigation of various invertebrates in forest and
woodland areas in south-west of Western Australia.
Journal of the Royal Society of Western Australia 63:
21-28.

Lobry de Bruyn, L. A. (1999). Ants as bioindicators of soil

240

S.A. Thompson, G.G. Thompson

function in rural environments. Agriculture,
Ecosystems and Environment 74: 425^41 .

Madden, K. E. and Fox, B. J. (1997). Arthropods as
indicators of the effects of fluoride pollution on the
succession following sand mining. Journal of Applied
Ecology 34: 1239-1256.

Majer, J. D. (1985a). Seasonality of ants (Formicidae) in
south-western Australia. In P. Gieenslade and J. D.
Majer (eds), Soil and litter invertebrates of some
Australian Mediterranean-type ecosystems, School of
Biology, Bulletin number 12, Western Australian
Institute of Teclmology, Perth.

Majer, J. D. (1985b). Seasonality of epigaeic invertebrates
at Perth, Dwellingup and Manjimup. In P. Greenslade
and J. D. Majer (eds). Soil and litter invertebrates of
some Australian Mediterranean-type ecosystems.
School of Biology, Bulletin Number 12, Western
Australian Institute of Technology, Perth.

Majer, J. D. and Brown, E. (1997). The role of
invertebrates in ecological functioning. In C. J. Asher
and L. C. Bell (eds). Fauna habitat reconstruction after
mining, 10-11 October 1997, Australian Centre for
Mining Environmental Research, Adelaide.

Majer, J. D. and Koch, L. E. (1982). Seasonal activity of
hexapods in woodlands and forest leaf litter in the
south-west of Western Australia. Journal of the Royal
Society of Western Australia 65: 37-45.

Majer, J. D. and Nichols, O. G. (1998). Long-term
recolonisation patterns of ants in Western Australian
rehabilitated bauxite mines with reference to their use
as indicators of restoration success. Journal of
Applied Ecology 35: 161-182.

Mattiske Consulting Pty Ltd (1995). Flora and vegetation
of the Cawse find area, Woodward-Clyde, Perth.

McGeoch, M. A. (1998). The selection, testing and
application of terrestrial insects as bioindicators.
Biological Reviews of the Cambridge Philosophical
Society 73: 183-201.

Paltridge, R. and Southgate, R. (2001). The effect of
habitat type and seasonal conditions on fauna in two
areas of the Tanami Desert. Wildlife Research 28: 247-
260.

Pianka, E. R. (1986). Ecology and natural history' of desert
lizards: Analyses of the ecological niche and
community structure, Princeton University Press,
Princeton.

Postle, A. C. (1985). Density and seasonality of soil and
litter invertebrates at Dwellingup. In P. Greenslade
and J. D. Majer (eds). Soil and litter invertebrates of
some Australian Mediterranean-type ecosystems,
School of Biology, Bulletin Number 12, Western
Australian Institute of Technology, Perth.

Thompson, G. G., Thompson, S. A., Withers, P. C. and
Pianka, E. R. (2003). Diversity and abundance of pit-
trapped reptiles of arid and mesic habitats in
Australia: Biodiversity for environmental impact
assessments. Pacific Conservation Biology' 9: 120-35.

Thompson, S. A. and Thompson, G. G. (2005a). Temporal
variation in reptile assemblages in the Goldfields of
Western Australia. Journal of the Roy'al Society of
Western Australia 88: 25-36.

Thompson, G. G. and Thompson, S. A. (2005b). Mammals
or reptiles, as surveyed by pit-traps, as bio-indicators
or rehabilitation success for mine sites in the
goldfields region of Western Australia? Pacific
Conservation Biology 11: 268-286.

Manuscript received 17 August 2005; accepted 24 August
2006

Records of the Western Australian Museum 23: 241-257 (2007).

A new species of rock-dwelling hylid frog (AnuraiHylidae)
from the eastern Kimberley region of Western Australia

Paul Doughty and Marion Anstis^

’Department of Terrestrial Vertebrates, Western Australian Museum,
49 Kew Street, Welshpool WA 6106, Australia
^corresponding author - e-mail: Paul.Doughty@museum.wa.gov.au
^26 Wideview Road, Berowra Heights NSW 2082, Australia

Abstract - Australia's documented frog diversity slowly continues to grow
owing to genetic tests for cryptic species and ongoing exploration of remote
regions. Recent collecting trips in Western Australia's east Kimberley region
resulted in the discover}' of a new' rock-dwelling hylid frog, Litoria staccato
sp. nov. The new species is closely related to the much more widely
distributed L. coplandi, W'hich also breeds in the same rocky creeks. Litoria
staccato sp. nov. is a small to moderate-sized frog characterised from co-
occuring species by a combination of a moderately pointed snout, expanded
terminal discs, half-W'ebbed toes and a mottled appearance wdth variable
colouration (reddish browm, grey or beige). The advertisement call consists of
a rapid burst of irregularly-spaced notes, followed by groups of softer calls
comprised of single or complex notes. Compared to L. coplandi, L. staccato
sp. nov. is slightly smaller, has reduced webbing between the toes, different
colouration and pattern (including diffuse vertebral and dorsolateral stripes),
reduced glandular tissue at the angle of the jaw and a highly divergent call.
Tadpoles show some adaptations to stream-living but also have body shape
affinities associated with ground hylid pond-dwelling types such as L.
inermis. The now species has only been found near Wyndham in the far north
of Western Australia, and no specimens have been detected in existing
museum collections indicating a restricted distribution. Owing to its
remoteness and complex geology, the Kimberley region may hold other
undiscovered rock-dwelling species with small natural ranges.

Key words: frog, Kimberley, Litoria, rock-dw’elling, tadpole

INTRODUCTION

Frogs of the genus Litoria are prominent among
northern Australian vertebrate fauna. Here they
have radiated into a diversity of forms specialized
for different lifestyles, including species that are
strongly associated with rocky streams and pools
along escarpments. There are currently three small
rock-dwelling hylids from the humid Kimberley to
Arnhem Land region of northern Australia; L.
coplandi, L. personata and L. meiriana. All three
species have expanded terminal discs on their
fingers and toes and are encountered along rocky
creeks, water holes and escarpments. Tyler and
Davies (1978) initially placed L. coplandi in its own
monotypic species group. Barker et al. (1995) placed
the rock-dwelling forms either directly in a "L.
iatopalmata" group (L. personata) or in "other
Litoria” (L. coplandi and L. meiriana). Before being
formally described, specimens of L. coplandi were
placed in "L. Iatopalmata watjulumensis" but later
described as a separate taxon by Tyler (1968a). Tyler
et al. (1978) compared the new taxon L. personata

to various L. Iatopalmata group members, but not
to L. coplandi. Recent molecular work indicates that
all three rock-adapted hylids may be only distantly
related (S. Donnellan personal communication),
suggesting that they evolved an association with
flowing water and pools on rocks independently.
Litoria meiriana is likely to be only distantly related
to the other two species based on morphological
(adults and tadpoles), behavioural and genetic
differences (Tyler and Davies 1978; Tyler et al. 1983;
S. Donnellan personal communication).

Potential threats to the native frogs of the tropical
Kimberley region in Western Australia from
introduced species such as the cane toad {Bufo
marinus: Bufonidae) and chytrid fungus have
generated concern about the future status of frogs
there. As a result, new surveys are being conducted
to estimate the true diversity of the region. Initial
surveys conducted in the wet season of 2005-2006
in the east Kimberley have revealed a previously
unknown taxon closely allied to, and syntopic with,
the rock frog, L. coplandi. Here we describe this

242

P. Doughty, M. Anstis

taxon as a new species and present information on
the male advertisement call, embryonic and tadpole
development and the breeding habitat.

MATERIALS AND METHODS

We examined 12 adult specimens of the new
taxon and compared them with its suspected close
allies L. coplandi and L. personata. Morphological
measurements generally follow Tyler (1968b) with
some modifications (see Table 1). Measurements
that could be made on either side of the body (e.g.,
tarsus length) were measured on the right side of
the animal, unless this was damaged or misshapen.
Measurements were made under a Leica MZ6
dissecting scope with digital vernier callipers to the
nearest 0.01 mm. We also calculated the following

ratios (see Table lA for abbreviations); HL/HW, IN/
lO, EN/IN, TL/SVL, TarL/SVL and TarL/TL.

We compared the calls of two males of the new
species with the call of one L. coplandi and one L.
meiriana. Calls were recorded on a Marantz
PMD670 digital recorder with a BeyerDynamic
M88N microphone. Sound analysis was carried out
on Cool Edit Pro and Raven 1.3b (Charif etal. 2004).

We collected a sample of embryos just prior to
hatching close to where calling males and a gravid
female had been collected the previous night. Six
hatchlings and a small sample of capsules from the
same clutch were also preserved. A sample of live
hatchlings was collected and reared to
metamorphosis to confirm identity. In addition,
another sample of small tadpoles at stages 26-27
(Gosner 1960) found in the same pool and

Table 1 Characters measured with abbreviations and explanations.
Character Abbrev. Explanation of Measurement

A. Adults

Snout-vent length

SVL

Inter-limb length

ILL

Head length

HL

Head width

HW

Eye-naris distance

EN

Interorbital span

to

Internarial span

IN

Naris-mouth distance

NM

Eye diameter

EL

Tympanum length

TymL

Forearm length

FL

Hand length

HandL

Third finger disc width

3^‘^FDW

Tibia length

TL

Tarsus length

TarL

Foot length

FootL

Fourth toe disc width

4'hTDW

B. Tadpoles

Total length

TL

Body length

BL

Body depth

BD

Body width

BW

Body width at eyes

EBW

Tail muscle depth

BTM

Tail muscle width

BTMW

Tail depth

TD

Dorsal fin depth

DF

Tail muscle depth

TM

Ventral fin depth

VF

Inter-orbital span

lO

Inter-narial span

IN

Eye to naris

EN

Narial diameter

N

Snout to spiracle

SS

Snout to naris

SN

Snout to eye

SE

Eye diameter

ED

Oral disc width

ODW

From tip of snout to posterior tip of urostyle
From axilla to groin

From tip of snout to posterior edge of tympanum

Width of head at centre of tympani

From anterior corner of eye to posterior edge of naris

Distance between anterior corners of eyes

Distance between inner edges of nares

Posterior edge of naris to upper edge of jaw

Anterior to posterior corners

Anterior to posterior edges

Elbow to proximal edge of palmar tubercle

Tip of 3"** finger to proximal edge of palmar tubercle

Maximum transverse width of 3 ”^ finger disc

Measured with leg in natural resting position, from knee to tarsus

Measured with leg in natural resting position, from proximal end of tarsus to

proximal edge of inner metatarsal tubercle

From tip of 4'*’ toe to proximal end of inner metatarsal tubercle

Maximum transverse width of 4"^ finger disc

From tip of snout to tail tip

From tip of snout to end of body

Maximum height of body

Widest point of body in dorsal view

Body width at level of eyes in dorsal view

Depth of tail muscle at base

Width across tail muscle at base in dorsal view

Measured at midpoint of tail

Measured at tail depth

Measured at tail depth

Measured at tail depth

Measured in dorsal view

Measured in dorsal view

Measured in dorsal view

Measured in dorsal view

Measured at maximum in ventral view

Rock-dwelling hylid frog

considered likely to be this species, was collected
and reared to metamorphosis. Tadpoles were
reared in 50 cm diameter containers of stream water
to a depth of 14 cm, rocks and leaf litter from the
stream where they were collected. Water was
aerated and temperatures ranged from about 16-
36°C during development.

Tadpole descriptions follow Anstis (2002).
Abbreviations for tadpole morphometric characters
follow Anstis and Tyler (2005) and are given in
Table IB. Measurements were made with an ocular
micrometer attached to a microscope and vernier
callipers. Embryos and tadpoles were drawn with
the aid of a camera lucida, and photographs of live
tadpoles taken using a Nikon D70 and 60 mm
macro lens.

SYSTEMATICS

Family HYLIDAE Rafinesque 1815
Genus Litoria Tschudi 1838
Litoria staccato sp. nov.

243

Chattering Rock Frog
Figures 1-5

FFolotype

WAM R162611. Adult male collected near "The
Grotto", 30 km south of Wyndham, Western
Australia (15.72540°S, 128.27953°E), by P. Doughty
and C. Mills on 30 January 2006. Liver sample
stored at -75°C at the Western Australian Museum,
Welshpool.

Paratypes

WAM R162512, R162514 (males) and WAM
R162513 (female) collected 8 January 2006 by P.
Doughty, J. Francis and M. Anstis (15.71466°S,
128.27288°E); WAM R162537-8 (males) collected on
15 January 2006 by P. Doughty, J. Francis and C.
Mills (15.72506°S, 128.27951°E); WAM R162612-6
(males) and WAM R162620 (female) collected on 30
January 2006 by P. Doughty and C. Mills
(15.72540°S, 128.27953°E). Liver samples stored at -
75°C at the Western Australian Museum,
Welshpool.

A

B

C

Figure 1 Head (A), chin and hand (B) and foot (C) of the naale holotype of Litoria staccato (WAM R162611).

244

P. Doughty, M. Anstis

Embryos and Tadpoles

WAM R162946-7 (embryos), WAM R162948-57
(tadpoles) collected 9 January 2006 by M. Anstis
and P. Doughty (15.71 466°S, 128.27288°E).

Diagnosis

A small to moderate-sized rock-dwelling hylid
with moderately pointed snout, medium build and
slender limbs. Tips of fingers widely expanded and
toes half-webbed. Dark lateral head stripe present
but not clearly defined; pale triangular patch
usually discernible on snout. Lateral head stripe
continues beyond tympanum and fades posteriorly
into broader mottled lateral stripe that demarcates
lateral and ventral zones. Dorsal colour of males
variable, ranges from reddish brown to slate grey to
beige; females reddish brown. There are variably
expressed diffuse darker vertebral, dorsolateral and
lateral stripes.

Distinguishable from similar-sized ground hylid
frogs of the Kimberley-Arnhem Land region by
possession of broadly expanded discs on tips of
fingers and toes (not L. inermis, L. latopalmata, L.
nasuta or L. pallida which lack expanded terminal
discs), toes half-webbed (not L. Copland!, L.
meiriana or L. wotjulumensis which have fully
webbed toes) and mottled dorsal colouration with
diffuse lateral head stripe, vertebral and
dorsolateral stripes (not L. personata which has
strong lateral head stripe and uniform-coloured
dorsum) (see also Comparison with other species,
below). The male call consists of a series of rapid,
high-pitched irregularly spaced notes, interspersed
with short and complex softer calls (Figure 3C).

Description of holotype

Head narrow and triangular with moderately
pointed snout and prominent eyes (Figure lA). In
profile, snout gradually narrows to oblique tip.
Nares positioned on tip of snout under canthus
rostralis, slightly oval, opening dorsolaterally and
slightly forwards. Canthus rostralis straight with
moderately sharp edge; loreal region steep-sided
and concave. Tympanum prominent and circular,
distinct annulus present except for dorsal edge.
Small cluster of 5-6 glandular nodules between
lower posterior edge of tympanum and insertion of
forearm. Vomerine teeth a pair of smooth ridges
anterior to medium-large oblique choanae. Tongue
oval, tapers posteriorly, free edge blunt and
unnotched.

Arms short and slender. Fingers long, slender and
unwebbed but with weak lateral fringes (Figure IB).
Palmar tubercles at base of outer portion of wrist
prominent and paisley-shaped (narrow end
pointing towards fingers). Large tubercles present
on finger joints with smaller tubercles on palm.
Nuptial pad comprised of fine layer of small dark
rugose tubercles on inner margin of finger.

Fingers in order of length: 3>4>1>2. Tips of fingers
with broad discs: P‘ and 2"'^ fingers approximately
2x wider, and 3^'^ and 4* fingers approximately f .5x
wider than distal phalanx in life (in preservative,
discs L5x and lx wider, respectively).

Legs long and slender. Distinct fold of skin above
knee. A fringe runs along inner tarsus and connects
to inner metatarsal tubercle. Moderate sized inner
metatarsal tubercle narrow, projects distally (Figure
1C). Outer metatarsal tubercle small and oval,
projects towards toes. Feet narrow. Toes in order of
length: 4>5>3>2>1. Webbing between P' and 2"'^ and
between 2""* and 3"'^ toes to proximal end of distal
phalanges on each toe. Webbing between 3’’'^ and 4*''
toes to just beyond proximal joint of distal phalange
on 3''^ toe, and to base of proximal end of
penultimate phalanx on 4* toe. Webbing between
4‘'’ and 5‘*' toes to base of proximal end of
penultimate phalanx on 4'^ toe and to just above
proximal end of distal phalanx on 5* toe. Lateral
fringes on all toes beyond webbing. Toe discs only
slightly wider than penultimate phalanx in life (in
preservative, approximately the same width).
Medium conical subarticular tubercles on joints of
toes with minute tubercles on plantar surface.

Skin on dorsum and limbs smooth. Belly granular
with slight transverse crease between arms, towards
anterior edge of arm insertion. Underside of
posterior edge of thighs with larger flattened
granulation. Coccvx forms prominent ridge that
protrudes slightly beyond end of body. Cloaca
positioned just below coccyx, projects dorso-
posteriorly.

Dimensions of holotype (mm)

SVL 30.5; ILL 13.15; HL 11.66; HW 10.83; EN 2.76;
lO 5.3; IN 3.21; NM 1.64; EL 3.21; TymL 2.30; FL
6.15; HandL 7.60; O^-'FDW 0.72; TL 15.72; TarL 8.40;
FootL 11.69; 4*TDW 0.56; HL/SVL 0.38; HL/HW
1.08; EN/IN 0.86; EN/IE 0.52; TL/SVL 0.52; TarL/
SVL 0.28; TarL/TL 0.53.

Colour in life

Dorsum light reddish brown (Figure 2A). Faint,
darker, narrow vertebral and wider dorsolateral
stripes present, the latter forming a diffuse border
between dorsal and lateral zones. Lateral head
stripes dark grey, not sharply defined along snout,
with diffuse dorsal and ventral edges. Lateral head
stripe begins narrowly at rostrum passing through
nostril and lower half of eye; continues posteriorly
from eye through tympanum, extending just above
dorsal edge of tympanum; angles downwards
towards ventral surface, fading diffusely just over
half-way between insertion of arms and legs;
continues as diffuse mottled border between lateral
and ventral zones. A subtle, yet distinct, paler
triangular patch on snout is defined dorsally by
border of lateral head stripes and posteriorly by

Rock-dwelling hylid frog 245

D

Figure 2 Adult frogs in life of Litoria staccato showing colour variation. A) holotype male (WAM R162611) with
reddish brown colour; B) beige male (WAM R162514); C) calling slate grey male (uncollected); D) reddish
brown female (WAM R162513).

246

P. Doughty, M. Anstis

diffuse darker bar between eyes. As triangular
patch narrows towards tip of snout, it broadens
slightly and contacts nares before terminating just
anterior to nares. Lateral head stripes continue
forward to join at rostrum tip. Upper lip mottled
with diffuse black. Lower lip pale with dark
mottling not extending to chin. Chin darkly
stippled anterior to vocal sac, and less stippled
towards margin of jaw. Lower two-thirds of iris
brown, upper third bright copper gold, pale gold
border above pupil, less distinct below pupil.

Tympanum unpigmented except for darker patch
extending from dorsal edge to centre. Annulus of
tympanum pale. Pale lemon yellow wash over
upper lip (below stripe), sides and posterior surface
of thighs. Bright lemon yellow wash over groin,
fades anteriorly. Flanks and posterior surfaces of
thighs diffusely mottled with reddish brown colour
of dorsum. Dorsal surface of limbs reddish brown
(same as dorsum) with diffuse darker mottling.
Dorsal surfaces of arms mottled, fingers paler,
especially 2"'^ and S''*. Outer edge of forearms with
darker mottling. Dorsal surface of legs dark with
some mottling, especially on posterior edge of
thighs where blotches form an uninterrupted line.
Belly and ventral surface of limbs pale white,
undersurfaces of feet dark brown.

Colour in preservative

Dorsal surfaces much darker than in life - dark
slate to chocolate brown - with vertebral and
dorsolateral stripes much less apparent. Dark lateral
head stripe poorly defined with ground colour
discernible beneath; continues past tympanum and
fades on side near arms. Yellow wash in groin
barely discernible. Undersurfaces pale yellow,
hands and feet dark.

Variation

Male body sizes varied only slightly - the smallest
was 29 mm and the largest 33 mm (Table 2). The
build, proportions and general appearance of male
specimens generally agreed with the holotype
except for the following (WAM prefixes excluded
below). Shape of rostrum varied from sharp and
angular (R162614, R162616) to more broadly
rounded (R162612). Glandular tissue at angle of jaw
similar to holotype for most males, but in R162616,
nodules were higher and more prominent, and in
R162512, skin was nearly smooth. Nuptial pads
ranged from less developed (lighter and less
extensive; R162512) to very heavy and extensive
(R1 62537). R162515 possessed slightly rougher pads
than other males.

The two female specimens had lengths of 35.5 and
36.5 mm - larger than any of the 10 males. Both
females were collected near calling males and were
heavily gravid. Other than overall body size there
were no obvious differences between males and

Table 2 Summaries of characters and ratios measured
for Litoria staccato, L. coplandi and L.
personata. MeanlS.D. (range). Sample sizes
are for species unless noted. See Table lA for
abbreviations.

Character

L. staccato

N = 12

L. coplandi

N = 56

L. personata
N = 12

SVL

31.4±2.4

33,4+4.0

30.8+2.4

(29-36.5)

(24.5-43.0)

(26.5-33.5)

ILL

13.0±2.1

13.8+2.1

12.3+1.9

(10.4-18.4)

(8.8-18.0)

N = 54

(9.5-15.3)

HL

11.3+0.7

(12.4+1.3)

11.2+0.8

(10.4-13.1)

(8.6-15.2)

(9.7-12.4)

HW

10.7+0.8

11.8+1.4

9.9+0.7

(9.8-12.6)

(8.9-14.6)

(8.6-11.0)

EN

2.8+0.2

3.2+0.4

3.0+0.3

(2.6-3.1)

(2.6-4.1)

(2.4-3.2)

lO

5.4+0.3

6.6+0.7

5.9+0.5

{4.8-6.0)

(5.0-8.2)

(5.0-6.5)

IN

3.2+0.2

3, 2+0.4

3.2+0.2

(2.8-3.6)

(2.4^.1)

(2.9-3.4)

NM

1. 8+0.2

2.1+0.3

1. 7+0.2

(1. 5-2.1)

N = ll

(1. 6-2.7)

N = 55

(1.5-1. 9)

EL

3.2+0.3

3.7+0.4

3.3+0.4

(2.8-4.0)

(2,9-4.5)

(2.6-1.0)

TymL

2.3+0. 1

2.6+0.3

2.5+0.6

(2.0-2.5)

(2.1-3.8)

(1.9-3.3)

FL

6.4+0.7

6.9+0.7

6. 6+0. 6

(5.7-7.9)

(5.5-8.6)

(5.7-8.0)

HandL

7.6+1 .2

8.4+1 .3

7.7+0.6

(6.0-9.6)

(5.8-10.3)

(6.6-8.5)

N = 11

S'TDW

0.98+0.15

1.14+0.23

0.93+0.22

(0.77-1.23)

(0.63-1.81)

(0,60-1.13)

N = 8

z

II

CD

N = 8

TL

16.0+1.3

18.7+2.2

17.4+1.6

(14.3-18.4)

(14.8-24.1)

(14.7-20.0)

TarL

8.4+0.6

9.1+1.1

9.2+1 .0

(7.4-9.7)

(7.4-11,4)

(7.2-10.6)

FootL

11.9+1.2

13.3+1.8

12.1+1.3

(10.8-14.4)

(10.4-18.2)

N = 55

(10.1-13.9)

4'hTDW

0.73+0.10

0.98+0.25

0.72+0.21

(0.63-0.94)

(0.58-1.68)

(0.48-1,04)

N = 8

N = 45

N = 8

HW/SVL

0.34+0.01

0.36+0.01

0.32+0.01

(0.31-0.36)

(0.32-0.36)

(0.32-0.36)

HL/HW

1.05+0.03

1.04+0.05

1.13+0.04

(1.02-1.10)

(0.82-1.16)

(1.07-1.21)

EN/IN

0.89+0.04

0.98+0.08

0.93+0.06

(0.81-0.95)

(0.81-1.16)

(0.82-1.04)

EN/IO

0.52+0.02

0.48+0.04

0.50+0.04

(0.47-0.56)

(0.41-0.57)

(0.45-0.57)

TL/SVL

0.51+0.02

0.56+0.04

0.57+0.03

(0.46-0.55)

(0.45-0.65)

(0.49-0.60)

TarL/SVL

0.27+0.01

0.27+0.02

0.30+0.02

(0.25-0.29)

(0.22-0.33)

(0.24-0.33)

TarL/TL

0.52+0.02

0.49+0.02

0.53+0.03

(0.49-0.55)

(0.44-0.53)

(0.48-0.58)

Rock-dwelling hylid frog

247

females in morphological characters, but the small
number of females prevented further evaluation.

Colouration of males was variable. In addition to
the reddish brown of the holotype and paratypes
R162537-8, R162612 and R162616, other individuals
were bright beige while active in life (R162514 and
other uncollected males - Figure 2B). Still others
were slate grey (R162512, R162613-5 - Figure 2C).
Mottling on the dorsum was also variable - some
individuals had darker mottling (e.g., R162612)
while others had only faint variegations (e.g.,
R162514). Collected individuals changed colour
from generally vivid while active to more dull and/
or mottled the following day, obscuring the diffuse
vertebral, dorsolateral and lateral streaks.

The lateral head stripe ranged from relatively
demarcated (e.g., R162613) to diffuse grey (e.g.,
R162514) with borders never sharply defined. In
most males, the lateral head stripes did not meet at
the tip of the snout, but in one other individual
(R162537) they joined, as in the holotype. The paler
snout patch outlined by the lateral head stripes and
the diffuse posterior bar between the eyes varied in
definition from very clear (e.g., R162537) to poorly
defined (R1 62612). Presence of the thin vertebral
and wider dorsolateral streaks was highly variable.
In some specimens, stripes were relatively solid and
dark (e.g., R1 62613, R162615), in others there was
only a slight stripe (R162514), or heavy mottling
that obscured stripes (R162612). The border
between lateral and ventral regions varied from a
smooth transition with little marking (R162514), to
a mottled transition zone (R162612, R162614), to a
darker stripe (R162613, R162614). Mottling on
posterior edge of thighs ranged from diffuse (e.g.,
R162614-5; as for holotype), to faint uniform
stippling (R162612-3), or very faint stippling
(R162514).

The two female specimens were similar in
colouration - both had the dull reddish brown
background colouration seen in several males, with
moderate to heavy dark mottling on dorsal surface.
Snout patches were less prominent and vertebral,
dorsolateral and transverse bars weakly defined.
Female R162513 was lighter overall, including paler
sides, no stippling on chin and only faint stippling
on back of thighs (Figure 2D). Female R162620 was
darker, with mottled sides, light stippling on chin
and mottling on back of thighs similar to some
males.

Advertisement call

The calls of the holotype male (R162611) and a
paratype (R162612) were recorded on 30 January
2006 between 7 and 9 pm. The air temperature 1 cm
above the males was 28.7°C (R162611) and 26.6°C
(R162612), and the temperature of the flowing
water about 5 cm below the surface was ~ 29°C for
both.

The call of the holotype of L. staccato is presented
in Figure 3C. It consists of a sequence of rapid,
high-pitched, irregularly spaced, short (staccato)
notes, followed by a series of softer and more
widely spaced notes with occasionally more
complex notes (Figure 3C-F). The holotype male
called 3.8 times per minute with call duration
averaging 6.5 s (maximum - 15 s). Notes in the main
call are irregularly spaced, sounding similar to a
Morse code signal. There were an average of 25
notes/call and 4.3 notes/s. The notes increased
slightly in amplitude during the call (Figure 3C).
Each note consisted of a series of 14-19 pulses that
increased in amplitude gradually with a sharper
decrease, and with dominant frequencies of 2-3 and
4-6 kHz (Figure 3D). Between the main calls, the
much less frequent softer calls were delivered in
small clusters of typically 3-4 notes (up to six).
These notes were made up of 5-8 pulses with
dominant frequencies at 2, 3.5 and 5 kHz (Figure
3E). A third type of call was occasionally given
among the softer calls that consisted of a rapid, trill-
like series of modulated pulses with several peaks
(Figure 3F). During the 10 minute recording, the
male only began to give these more complex calls in
the middle third of the calling sequence. These
complex notes were made up of 34-^8 pulses, had 4
or 5 peaks in amplitude during the brief (0.15 s) call
and had dominant frequencies at 1.5 and 3.5 kHz.

The paratype male (R162612) had very similar call
characteristics for the main call, but did not give the
soft or complex calls between the main calls. These
two individuals were calling on either side of a
stream > 5 m apart. In both recordings, other males
called simultaneously in response to each other. A
gravid female (R162120) was captured within 2 m
of R162612.

Breeding choruses

Litoria staccato males called in choruses of 2-6
males in slow-flowing sections of a rocky creek at
one site, and around shallow water in crevices or
under boulders at another site located on an
escarpment. Calling sites included exposed rocks,
within crevices and under overhanging vegetation.
One male (R1 62612) was observed calling ~ 10 cm
above the water (head facing down and towards the
stream) while clinging vertically to a ~ 50 cm
boulder at the creek's edge. No males were
observed to be within 5 m of each other and males
often called from positions on opposite sides of the
water body. Calls of males in breeding choruses
occurred synchronously.

Embryos

A single clutch of embryos was collected that
were either just prior to hatching or just hatched at
stages 20-21 with capsules partly decomposed. The
clutch was collected from a very small and shallow

248

P. Doughty, M. Anstis

A

B

D

Figure 3 Oscillograms (upper) and sonograms (lower) of male advertisement calls. A) Litoria meiriana (WAM
R162521); B) L. coplandi (uncollected); C) L. staccato holotype (WAM R162611); D) L. staccato main call; E) L.
staccato soft call; F) L. staccato complex call.

Rock-dwelling hylid frog

249

rock pool (70 x 30 cm and 2-3 cm deep) segregated
by about two metres from the main creek, most of
which was flowing at a reduced water level beneath
large boulders. The pool contained leaf litter and
tannin-stained water and was on a rock shelf where
several calling males and a gravid female rvere
found the previous night. The sample of small
tadpoles collected at stages 26-27 was taken from
the same pool. The remaining jelly capsules were
covered with silt and most were decomposing, but
those of six embryos which haci died earlier at
about stages 13-14 were still intact, and these had a
mean external capsule diameter of 3.83 mm (3.54-
3.86 mm).

Measurements of embryos are shown in Table 3.
Two embryos at stage 20 had shorter gills, darker
fins and a less arched dorsal fin than those at stage
21.

Stage 21 (Figure 4A). - Dorsum and tail muscle
appear black macroscopically; area above head
(lateral view) translucent grey; tail fins dusky grey;
snout angular in lateral view; abdomen broad in
dorsal view, yolk white; optic bulge discernable but
barely pigmented; two pairs well developed
external gills with 3-4 upper and 5-6 lower
filaments; adhesive organs black and prominent;
deep triangular stomodaeum bordered by labial
ridges; narial pits visible.

Stage 23 (Figure 4B). - Reached on 10 January;
dorsum very dark brown with scattered iridophores
over snout, brain, eyes and tail muscle; yolk whitish
with network of melanophores dorsolaterally;
lateral line organs faintly visible; tail fins dusky
grey, melanophores anteriorly across dorsal and

partly lateral surface of muscle. Snout broad in
dorsal view and rounded in lateral view; eyes well
developed, cornea clear; external gills slightly
reduced, upper and lower branches of similar
length, 4—5 upper and 6—7 lower filaments; adhesive
organs broad and flattened; nares perforated,
opening anteriorly, quite widely spaced and
situated right on tip of snout; labial ridges broader,
upper ridge divided; jaw sheaths visible, keratin
just visible on edge of upper sheath; operculum
open on both sides, short tubular projection on edge
of left side - juts outwards (probable early
development of spiracle); tail fins well arched, tip
broadly rounded; myotomes visible along muscle.

Tadpoles

The largest tadpole grew to a maximum total
length of 52.0 mm and body length 17.5 mm (stage
38). Table 3 presents measurements of tadpoles.
Tadpoles in captivity were predominantly bottom
dwellers and mostly grazed on live algae on rocks
and on sediments. Initially water was not aerated
and while most tadpoles appeared to grow
normally, some died. Aeration was then introduced
and the remainder survived, became more agile and
grew more steadily. Tadpoles frequently remained
in the vicinity of the source of aeration, holding
onto rocks with the oral disc. If disturbed, they
rapidly darted under rocks or leaves.

Table 4 describes pigmentation development in
life. In preservative, all golden, silver and copper
iridophores are lost, together with lighter brown
pigment, leaving only the darker melanophore
patterns visible on the dorsum and tail. The venter

Table 3

Morphometric measurements of tadpoles of Litoria staccato, in mm (see Table IB for abbreviations). Number
of specimens: stages 20-21 = 7, 25 = 2, stages 26-29, 32, 38-40 = 1, stage 36 = 3, stage 46 = 2.

Stage

20-21 25 26 27 28

29

32

36

36

36 38

39

40

46

TL

5.72 12.39, 12.23 20.0 21.5 26.0

29.0

35.0

42.6

40.5

44.0 52.0

49.0

48.3

17.5, 19.0

(5.24-5.98)

BL

4.99, 4.99 7.72 8.05 11.1

12.07

13.36

16.74

15.77

16.3 17.5

17.7

17.7

BD

6.15

6.76

7.24

9.01

8.69

9.98

9.66

BW

6.72

7.08

7.24

9.98

9.33

10.94

10.78

EBW

6.64

7.08

7.08

8.05

9.17

10.46

9.82

BTM

2.29

2.73

3.22

4.18

3.54

5.0

4.83

BTMW

2.09

2.41

2.57

3.54

3.7

5.15

5.15

TD

5.49

6.27

6.6

8.05

8.13

9.17

8.69

DF

1.88

2.09

2.25

2.57

2.65

2.98

3.05

TM

1.88

2.25

2.25

3.22

3.13

3.7

2.57

VF

1.72

1.93

2.09

2.25

2.33

2.57

3.05

lO

3.44

3.78

3.86

4.34

4.34

4.5

4.34

IN

1.88

2.09

2.09

2.57

2.57

2.57

2.57

EN

1.93

2.09

2.09

2.57

2.57

2.73

2.73

N

0.28

0.28

0.32

0.3

0.32

0.32

0.32

SS

6.44

7.08

8.05

9.41

9.01

10.3

9.98

SN

1.28

1.28

1.61

2.01

1.61

2.25

1.61

SE

3.38

3.38

3.7

4.83

4.34

5.15

4.34

ED

1.36

1.45

1.93

2.09

2.25

2.57

2.57

ODW

3.19

3.28

3.36

4.18

3.93

4.51

4.67

250

P. Doughty, M. Anstis

Figure 4 Embryos, tadpole and oral disc of Litoria staccato. A) hatchling at stage 20, bar = 1 mm; B) stage 23, bar = 1
mm; C) tadpole at stage 36, bar = 5 mm; D) oral disc, specimen at stage 36, bar = 1 mm.

appears dark grey-blue and the paler snout colour
is not visible. Description of the morphological
changes during development are presented below.

Stage 25. - Reached by 14 January; body shape
cylindrical, similar to Type 2 hylids (Anstis 2002);
eyes near lateral; tail fins well arched, tail tip
rounded.

Stages 26 and 27 (Figure 4A). - Mostly similar to
later stages described below, apart from size and
pigmentation changes (see below), in body features,
mouthparts and tail features, but the distance from
fhe eyes to the tip of the snout is shorter and the
snout is a little narrower in dorsal view.

Stages 32-39. - Medium body size when full
grown, as wide as deep across abdomen to about
stage 32, slightly wider than deep across abdomen
from about stage 36 onwards; snout rounded in
dorsal view, gradually becomes broader and
slightly more streamlined anterior to eyes from
about stage 34 onwards; eyes near lateral, slightly
dorsolateral in later stages; nares small, quite
widely spaced, open anterolaterally, slightly closer
to tip of snout than to eyes; spiracle fairly short,
broad, opens dorsoposteriorly below horizontal
body axis posterior to midpoint of body; vent tube
dextral (type a; Anstis 2002), narrow,\nd opens

Rock-dwelling hylid frog

Table 4 Pigmentation of Litoria staccato tadpoles at different larval stages (Gosner 1960).

251

Stage Dorsum and Eyes

Sides

25 Melanophores over dorsum;
gold iridophores over most
of dorsum (except over
darker base of body); some
small dark patches over
vertebral region.

26-27 Dorsum mostly uniform
(Fig. 5A) golden; areas above brain,

around nares, over abdomen
and base of body a little
darker; iris golden above and
below pupil, black at each
side and across top.

32 Dull golden brown or darker
(Fig. 5B) brown with layer of fine

copper-gold iridophores over
most of head and body, dark
longitudinal stripe down
each side of vertebral region
and dark patch over base of
body, indisfinct darker mask
bridges eyes.

36-39 Diffuse melanophore clumps
(Fig. over dorsal fin and muscle of
5C,D) tail; a few diffuse gold clusters
and flecks over muscle and
both fins; darker pigmented
veins over muscle and fins
(some outlined with gold);
copper stripe extends from
middle of base of body just
onto dorsal surface of muscl;
dense copper-gold covers
most of iris.

Gold patch beginning on
each side of abdomen at
base of body, denser
iridophores posterior
to gill region.

Gold clusters cover upper
half of abdomen, merging
down sides to orange-
gold, dark background
beneath; lower half of
abdomen orange-gold,
opaque white beneath;
orange-gold from gills to
eyes, clearer below;
distinct pale gold
longitudinal patch midway
down body along each side
of abdomen, just anterior
to base of body; another
similar but narrower
vertical patch just posterior
to gill region; pigment
lighter anteriorly.

Distinct lateral gold bar
present at base of body
during at least stages
26-28 now mostly
obscured.

Mottling covers upper
two-thirds of body,
denser by stage 38
onwards.

42 Pale triangle on snout anterior
(Fig. 5E) to eyes visible, demarcated
posteriorly by diffuse darker
bar bridging eyes.

Venter

Mostly transparent,
bordered by dense
melanophores and
gold stippling.

Brilliant orange-gold
over abdomen,
sparser over gills and
clear over buccal
region.

Opaque silver-white
with copper sheen,
clearer below mouth.

Opaque silver right
up to mouth from
stage 36.

Tail

Fins clear, dorsal surface of
muscle dark, capped with
gold patches spaced along
length, lateral surface
stippled with melanophores,
a few gold iridophores
anteriorly.

fins mostly clear with some
dark veins; few gold specks
and melanophores on dorsal
fin; fine melanophores over
muscle anteriorly, gold
stippling dorsolaterally over
anterior third; lateral surface
of muscle mostly
unpigmented posteriorly;
some gold clusters anteriorly
over lower half. Anterior
edge of ventral fin bordered
with pale gold, gold clusters
over vent tube.

Diffuse melanophore clumps
over dorsal fin and muscle of
tail; few diffuse gold clusters
and flecks over muscle and
both fins; darker pigmented
veins over muscle and fins
(some outlined with gold).

Darker mottling covers most
of tail; numerous pigmented
veins, some outlined with
gold; gold clusters anteriorly
on dorsal fin, copper-gold
along anterior edge of
ventral fin and over vent
tube; darker mottling denser
and covers entire tail by
stage 38.

posteriorly, dorsal edge partly unattached behind.
Fins moderately arched and taper to somewhat
elongate, narrowly rounded tip; dorsal fin begins
just onto base of body, initially low then rises more
distinctly to highest point anterior to midpoint of
tail before tapering; ventral fin less arched.

Oral disc (Figure 4D). - Near ventral in direction
in life (anterior medial margin tilts slightly
upwards); ventral in preservative. Marginal
papillae surround entire disc; anterior marginal
papillae mostly in a single row medially to partway
down lateral margins, increasing to two offset rows

252

P. Doughty, M. Anstis

rlU

-rC*

.-

Figure 5 Live tadpoles and metamorph of Litoria staccato. A) stage 26 (lateral view); B) stage 32 (dorsal view); C, D)
stage 36 (dorsal and lateral views); E) stage 42; F) stage 46. Bar in each photo = 5 mm.

Rock-dwelling hylid frog

253

beyond this down each side of anterior half; some
have as few as 10-25 medial papillae in a single
row across top of disc before two rows begin on
each side. Four to six rows of mostly small
submarginal papillae at each side of disc; two row's
offset slightly longer papillae around posterior
margin; may be only one row' initially at each side
of margin, to up to three row's medially in some.
Two anterior and three posterior tooth row's. A'
continuous, usually with medial pleat (Figure 5B),
A- has a narrow' medial gap, rows continuous,
P^very slightly shorter. Jaw sheaths medium, quite
distinctly serrated and fairly narrowly arched, with
long flared lateral processes.

Metamorphosis. - Tadpoles collected at stage 26
on 9 January began to metamorphose on 11
February (33 d later), and hatchlings collected on 9
January first metamorphosed on 20 February (42 d
later). Assuming that early development from egg
to hatching is likely to take about 3 d in the shallow
warm w'ater of the initial pool, minimum larval life
span in captivity for the hatchling group was about
45 d. New'ly metamorphosed froglets had
colouration similar to adults (Figure 5F). Head not
quite as proportionately long yet as in adults.
Terminal discs and webbing as for adults. Two
newly metamorphosed froglets measured 17.5 and
19.0 mm SVL.

Distribution

Currently known from only tw'O locations near
"The Grotto", approximately 30 km south of
Wyndham, Western Australia (Figure 6). Both
locations occur in the rocky southern portion of
Parry's Lagoon Nature Reserve east of the Great
Northern Highway. The entire collections of L.
coplandi at the WA Museum (529 specimens), SA
Museum (98 specimens). Museum and Art Gallery
of the Northern Territory (190 specimens),
Queensland Museum (77 specimens) and
Australian Museum (151 specimens) were checked
for the diagnostic characteristics of L. staccato. No
specimens of L. staccato were detected. This
indicates L. staccato's distribution is apparently
restricted to the small area where the type series
was collected. How'ever, ow'ing to the inaccessibility
of the Kimberley region due to the rugged terrain
and large areas with no vehicular access, it is likely
that the new species will be found elsew'here in the
eastern Kimberley, possibly to the northwest of the
two known sites and to the east in the Northern
Territory where similar habitats occur.

Habitat

Individuals of L. staccato were found in tw'o areas
with flowing water. The first was a steep rocky
ridge with a slow trickle of water running under
large boulders where males were calling, and where
the eggs and tadpoles were collected (see above).

The second area (where the holotype was collected)
was a creek that ran down a rocky ridge, about 2-3
km long (Figure 7). Both sources of water came
from underground streams that flowed from near
the top of ridges.

The vegetation at the rocky ridge sites where L.
staccato occurs is sparse but dominated by Triodia
wiseana with Cochlospermum fraseri, Calytrix
exstipulata and stunted Erythrophlem
chlorostachys. Along the watercourses where L.
staccato was calling were Triodia pungens,
Terminalia volucris, Ficus sp. and occasionally the
boab tree Adansonia gregorii.

Etymology

Specific name 'staccato' is from the Italian musical
term, and refers to the short detached sound of the
individual repeated notes of the male advertisement
call. It is to be treated as a noun in apposition.

Comparison with other species

1. Adults

In the eastern Kimberley, L. staccato may be
potentially confused with several species of
ground-dwelling Litoria which have pointed
snouts, such as L. nasuta, L. pallida, L. inermis and
L. tornieri. All of these species have narrow
terminal discs on the fingers, whereas L. staccato
has wider, expanded discs. Litoria nasuta has an
elongate head with a strongly pointed snout and
prominent longitudinal stripes. Litoria tornieri has
a smooth dorsum, uniform pale body colour and a
strongly contrasting dark lateral head stripe that
breaks up posterior to the tympanum. Litoria
inermis has a poorly defined lateral head stripe
similar to L. staccato, but possesses raised tubercles
over the dorsal surface, unlike the smooth skin of L.
staccato. Although some L. pallida also possess a
poorly defined lateral head stripe, they can be
distinguished by very narrow' terminal discs on the
fingers, slightly raised tubercles on dorsum,
distinctive penetrating call W'ith much longer notes
and selection of mostly still water breeding sites.
Litoria wotjulumensis often breeds along rocky
streams, has moderately expanded discs on the
fingers and toes and also has a complex call with
elements similar to L. staccato. However, L.
wotjulumensis is a much larger species (almost
double the length of L. staccato), has a more
elongate head, possesses a strong, broad lateral
head stripe and has fully webbed toes. Litoria
meiriana also occurs along rocky creeks and rock
holes and occurs in the Kimberley and Northern
Territory. However, its most obvious difference
from L. staccato is its much smaller size (~ 20 mm).
In addition, L. meiriana is dorsoventrally
compressed, has tubercular skin and fully w'ebbed
toes.

254

P. Doughty, M. Anstis

■

Figure 6 Distribution of Litoria coplandi, L. personata and L. staccato in northern Australia.

The two other rock-dwelling species with similar
habits to L. staccato and thus most likely to be
confused with it, are compared in more detail.
Table 2 presents summaries of morphological
measurements of L. staccato, L. coplandi and L.
personata. Litoria coplandi reaches a larger body
size, and females of both L. coplandi and L. staccato
are larger than males. The relative head width of L.
coplandi was wider than the other two species
(Table 2). Hind limb proportions of L. coplandi
and L. staccato were similar, but L. personata had
longer hindlimbs. Thus, L. staccato is characterised
by a narrower head relative to L. coplandi and
shorter tibia and tarsus lengths compared to L.
personata.

A number of other characters further distinguish
these three rock-dwelling forms. The most reliable
morphological character to distinguish the syntopic
L. staccato and L. coplandi is the extent of webbing
between the toes. In L. staccato the webbing is
reduced, for example the distal two phalanges on
the 4^'’ toe are free of webbing and the distal
phalanges of the other toes are also free of webbing.
In L. coplandi the webbing extends to the last
phalanx on the 4‘'" toe and to the terminal discs on
the remaining toes. The hands and feet of L. staccato
are more gracile than the more heavily built L.
coplandi. The webbing between the toes of L.
personata is only slightly more extensive than L.
staccato and much reduced relative to L. coplandi.
Another consistent character among the three

species is the glandular tissue at the angle of the
jaw. This tissue is pronounced and raised into
several discrete nodules in L. coplandi, much
reduced in L. staccato (fewer and lower in profile)
and absent in L. personata.

All three rock-dwelling hylids possess differences
in dorsal colour and patterns that can be used to
distinguish them, but these are individually
variable and some are not retained or less evident
in preservative. Ground colour of L. coplandi and
L. personata ranges from light to medium brown,
whereas L, staccato ranges from beige to slate grey
to reddish brown (the majority of individuals).
Litoria coplandi and L. personata have a relatively
uniform dorsal colour. In contrast, many L. staccato
individuals have more extensive mottling and
possess variably expressed vertebral, dorsolateral
and lateral stripes. The presence and prominence of
a lateral head stripe is another way to separate
them. Litoria personata has a strong, clearly defined
lateral head stripe, L. staccato has a less prominent
stripe with diffuse borders and L. coplandi lacks a
lateral head stripe (unique in the L. lesueuri
complex; Tyler 1968a; Barker et al. 1995).

2. Advertisement call

For the purposes of comparison, we present the
calls of two sympatric rock-dwelling hylids for
which no sonograms have been published, L.
coplandi and L. meiriana (Figure 3A,B). Both males
called within 5 cm of the edge of exposed rock

Rock-dwelling hylid frog

255

Figure 7 Type locality of Litoria staccato near
Wyndham, Western Australia.

pools. The temperature 1 cm above the calling L.
meiriana was 28.2°C with a water temperature of
32.1°C; the L. coplandi male was recorded shortly
after, and wdthin 50 m of the L. meiriana male.

The call of L. meiriana is most similar to L.
staccato owing to the notes of the main call being
irregularly spaced (similar to Morse code).
However, L. meiriana has a much higher-pitched
call (dominant frequencies of 2-3 and 4-6 kHz) and
the softer notes between main calls occur singly
(Figure 3A), unlike in L. staccato (Figure 3C). The
call of L. coplandi (Figure 3B) is easily
distinguished by a combination of regularly-spaced
notes, longer duration call (> 20 s), main call begins
very softly then increases steadily in amplitude,
dominant frequencies of 1-2 and 3—4 kHz and
single softer notes between main calls (as in L.
meiriana).

The call of L. wotjulumensis (not shown or
analyzed) is highly distinctive and very complex
{personal observations). The call contains loud,
sustained sequences of calls that abruptly double in
rate. The sustained calls can last for over 30 s and
are usually followed by a series of complex trills,
similar to the complex trill-like notes of L. staccato,
but given more frequently. Owing to the few males

that were recorded, our comparative results are best
considered preliminary. More detailed sound
analyses of more individuals and species is likely to
yield additional differences between species and
also provide estimates of variation within species.
The calls of all species discussed above may be
heard on www.museum.wa.gov.au/frogwatch.

3. Eggs and tadpoles

The remains of the egg mass collected indicate
that eggs may be laid in fairly loose clumps
attached to substrate rock. The eggs of L. coplandi
also have been found laid on the floor of shallow
rock pools singly or in small clumps (Tyler et al.
1983).

The tadpoles of L. staccato are distinguishable
from L. coplandi as early as at stage 25, when the
mouthparts are complete, as L. coplandi tadpoles
have two rows of continuous anterior papillae and
L. staccato have only a single continuous row across
part or all of the anterior margin. In addition, fully
grown L. coplandi tadpoles have a more distinctly
streamlined body form and a wider oral disc that
appears to be slightly more suctorial than that of L.
staccato. Of the other species of hylid tadpoles
which are found in stream pools in the escarpment
areas of the region, L. staccato have a generally
similar body size and shape to those of L. inermis
and L. wotjulumensis tadpoles, although they
become slightly more streamlined anteriorly than L.
inermis in later stages. Both L. inermis, L.
wotjulumensis and all other known ground hylid
species (with the exception of L. coplandi) in the
Kimberley region of northern Australia have a
narrow medial gap in the anterior papillae.

Remarks

New species of frogs are still being described in
Australia, especially in the northern tropics and the
eastern margin of the continent, where they are
most diverse. Recent descriptions include the
discovery of a highly distinctive stream-dwelling
tree frog in north Queensland, L. andirrmalin
(McDonald 1997), and a cryptic species of Uperoleia
near Darwin, U. daviesae (Young et al. 2005).
Genetic techniques and analysis of calls are
resulting in further cryptic species being uncovered
in frogs previously considered to be one species
(e.g., L. lesueuri, which has now been split into
three species, Donnellan and Mahoney 2004; see
also Donnellan et al 1983; Floskin 2004).

The recent discovery of L. staccato highlights the
possibility that more undescribed species of frogs
may occur in the Kimberley region. Other than the
sealed Great Northern Highway and the unsealed
Gibb River Road, only the Mitchell Plateau has been
reasonably sampled for frogs. Many surveys to
other regions (e.g., Kendrick and Rolfe 1991) were
designed to collect surface-active terrestrial

256

vertebrates but did not specifically target frogs and
did not involve night searches when breeding males
are easily located by their calls. Future wet season
frog surveys involving night work, recording of
male calls and taking tissue samples for molecular
analysis are likely to yield more undescribed
species in the Kimberley Region.

Little is known of L. staccato. Breeding choruses
occurred along rocky creeks up ridges or beside
seeps running down rock faces. In the area near
The Grotto where the type series was collected,
several other species were calling. Calling from
ponds on the flats between the ridges were the
myobatrachids Crinia bilingua, Opisthodon
ornatus, Notaden melanoscaphus and Uperoleia
lithomoda, and the hylids Cyclorana australis, C.
longipes, L. bicolor, L. pallida and L. nasuta.
Calling along large flowing rocky creeks at the
base of the ridges were U. borealis and L.
wotjulumensis. Further up the ridge, calling males
of U. borealis, Limnodynastes lignarius, Litoria
coplandi and L. staccato occurred along small
flowing rocky creeks. Near the top of the ridge,
only Limnodynastes lignarius and Litoria staccato
occurred. The reduced webbing on the feet of L.
staccato (compared to L. coplandi) may indicate
they are somewhat less aquatic, consistent with
their distribution further up the two ridges than L.
coplandi, where there is less water in creeks. Much
more work is needed to gain a better
understanding of the habits and distribution of L.
staccato in the Kimberley region and possibly in
adjacent parts of the Northern Territory.

ACKNOWLEDGEMENTS

We thank CALM "toadbusters" J. Francis and C.
Mills for assistance in the field during the discovery
and collection of the type series. G. Graham, G.
McKae and the staff of CALM-Kununurra provided
accommodation and logistical support for our
surveys during the 2005-2006 wet season. We thank
C. Stevenson (Western Australian Museum) for the
drawings of the adults and the distribution map
and J. Francis for the photograph of a calling male
L. staccato (Figure 2C). Technical advice and loan of
equipment for audio recordings and analysis were
provided by J. D. Roberts (University of Western
Australia) and C. Gerhardt (University of Missouri).
We thank P. TIorner (Museum and Art Gallery of
fhe Northern Territory), P. Couper (Queensland
Museum) and M. Flutchinson (South Australian
Museum) for loan of specimens, and S. Reynolds
(Charles Darwin Universify) and P. Oliver
(Adelaide University) for searching fhrough the L.
coplandi collections for specimens of the new
species. This work was made possible through the
Western Australian Museum's Alcoa Frog Watch
programme funded by Alcoa of Australia. This

P. Doughty, M. Anstis

paper is dedicated to the volunteer "toadbusters" of

the Kununurra community.

REFERENCES

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

Anstis, M. and Tyler, M. J. (2005). Breeding biology of
Litoria microbelos (Cogger) (Anura: Hylidae).
Transactions of the Royal Society of South Austalia
129: 43M8.

Barker, J., Grigg, G. C. and Tyler, M. J. (1995). A Field
Guide to Australian Frogs. Surrey Beatty & Sons,
Chipping Norton, NSW.

Charif, R. A., Clark, C. W. and Fristrup, K. M. (2004).
Raven 1.2 User's Manual. Cornell Laboratory of
Ornithology, Ithaca, NY, USA.

Donnellan, S., Adams, M., Hutchinson, M. and
Baverstock, P. R. (1993). The identification of cryptic
species in the Australian herpetofauna; a high
research priority (pp. 121-125). In Lunney, D. and
Ayers, D. (eds.) Herpetology in Australia: a diverse
discipline. Surrey Beatt}' & Sons, Chipping Norton,
NSW, Australia.

Donnellan, S. C. and Mahoney, M. J. (2004). Allozyme,
chromosomal and morphological variability in the
Litoria lesueuri species group (Anura: Hylidae),
including a description of a new species. Australian
Journal of Zoology 52:1-28.

Gosner, K. L. (1960). A simplified table for staging
anuran embryos and larvae with notes on
identification. Herpetologica 16:183-190.

Hoskin, C. J. (2004). Australian microhylid frogs
(Cophixalus and Austrochaperina): phylogeny,
taxonomy, calls, distributions and breeding biology.
Australian Journal of Zoology 52:237-269.

Kendrick, P. G. and Rolfo, J. K. (1991). The reptiles and
amphibians of Kimberley rainforests (pp. 347-359). In
McKenzie, N. L., Johnston, R. B. and Kendrick, P. G.
(eds.) Kimberly Rainforests of Australia. Surrey
Beatty & Sons, Chipping Norton, NSW.

McDonald, K. R. (1997). A new stream-dwelling Litoria
from the Melville Range, Queensland, Australia.
Memoirs of the Queensland Museum 42:307-309.

Tyler, M. J. (1968a). A taxonomic study of hylid frogs of
the Hyla lesueuri complex occurring in north-western
Australia. Records of the South Australian Museum
15:711-727.

Tyler, M. J. (1968b). Papuan hylid frogs of the genus
Hyla. Zoologishce Verhhandelingen 96:1-203.

Tyler, M. ]. and Davies, M. (1978). Species groups within
the Australopapuan hylid frog genus Litoria Tschudi.
Australian Journal of Zoology 63 (supplement):l-47.

Tyler, M. J., Davies, M. and Martin, A. A. (1978). A new
species of hylid frog from the Northern Territory.
Transactions of the Royal Society of South Australia
102:151-157.

Tyler, M. J., Crook, G. A. and Davies, M. (1983).
Reproductive biology of the frogs of the Magela
Creek system. Northern Territory. Records of the
South Australian Museum 18: 415^0.

Rock-dwelling hylid frog

257

Young, J. E., Tyler, M. ]. and Kent, S. A. (2005).
Diminutive new species of Uperoleia Grey (Anura:
Myobatrachidae) from the vicinity of Darwin,
Northern Territory, Australia. Journal of Herpetology
39: 603-609.

Manuscript received 25 July 2006; accepted 18 October 2006

APPENDIX

Comparative material examined.

Abbreviations: SAM - South Australian Museum;
NT - Museum and Art Gallery of the Northern
Territory; QM - Queensland Museum; note
specimen numbers without one of these prefixes are
from the Western Australian Museum.

Litoria coplandi

Males - WAM R103060, R108792, R110746,
R114039, R114090, R119091, Rn4092, R129193,

R137838, R137384, R137385, R140357, R140362,

R152951, R162520, R162523, R162524, R162535,

R162536, R162539, R162547, R162548, R162549,

R162950, R162581, R162596, R162597, R162602,

R162603, R1 62609, R162610, QM J54933, QM J56592,
QM J56588, QM J56595, QM J56580.

Females - R97942, R114088, R127332, R137382,
R137389, R138879, R138883, R138894, R140351,
R140352, R140361, R140369, QM J53809, QM J56584,
QM J56596.

Juveniles (sex unknown) - R95599, R129194,
R95509, R87922.

Litoria personata

Males - NT R16886, NT R18794, NT R18795, NT
R19807, NT R19809, NT R20466, SAM R16773, SAM
R16774.

Females - NT R20467, NT R20468, SAM R16831,
SAM R16832.

Juvenile — SAM R16829.

Note added in proof.

Field trips in 2006-2007 have recorded L. staccato
from the Mitchell Plateau and Prince Regent Nature
Reserve, greatly extending its distribution.

Records of the Western Australian Museum 23; 259-271 (2007).

Direct development in two Myobatrachid Frogs, Arenophryne rotunda
Tyler and Myobatrachus gouldii Gray, from Western Australia

Marion Anstis^ J. Dale Roberts^ and Ronald Altig^

' 26 Wideview Rd., Berowra Heights, NSW 2082, Australia. Email: frogpole@tpg.com.au
^School of Animal Biology (M092), University of Western Australia,

35 Stirling Highway, Crawley, Western Australia 6009 Australia
^Department of Biological Sciences, Mississippi State University, Mississippi State, MS 39762 USA

Abstract - The closely related Western Australian myobatrachid frogs
Arenophryne rotunda and Myobatrachus gouldii deposit eggs in burrows
that are dug by the adults in moist sand. Embryonic dev'elopment requires up
to two months and is completed entirely within the jelly capsule. The
developmental stages of these two taxa are described and compared with
those of the South American direct developing leptodactylid frog
Eleutherodactylus coqui.

Key words: Australia, direct development, embryo, endotrophic,
myobatrachid

INTRODUCTION

The frogs Arenophryne rotunda and
Myobatrachus gouldii (Myobatrachidae) are widely
distributed in semi-arid and arid regions of
southwestern Australia (Tyler et al. 2000). Both
species are forwards borrowers that oviposit deep
underground in moist sand where embryos
undergo direct development, an endotrophic
breeding mode in which all embryonic
development through to a froglet takes place within
the jelly layers of the egg (Altig and Johnson 1989).
Arenophryne rotunda calls from July-November
(austral winter to spring). Pairs of males and gravid
females not in amplexus have been found together
in November at a mean depth of 45 cm, and in
February and April (late summer to autumn) at
mean depths of 75-78 cm, but eggs were only found
in April (Roberts 1984). Myobatrachus gouldii calls
from September-February (spring to late summer);
a male and female burrow together, not in
amplexus, into deep, moist sand where they appear
to remain together until autumn when they deposit
eggs at depths of 80-115 cm (Roberts 1981, 1984).
Tyler's (1976a) suggestion of a close relationship
between these two species and with Metacrinia was
supported by Maxson and Roberts (1985), Read et
al. (2001) and the recent analysis by Frost et al.
(2006).

Direct development in amphibians has evolved in
at least seventeen genera from nine families of
anurans worldwide (Thibaudeau and Altig 1999).
Although the life histories of a number of these
species have been described, especially for the
genus Eleutherodactylus (e.g. Gitlin 1944; Jameson

1950; Wake 1978; Townsend and Stewart 1985),
there are no available descriptions of the Australian
species which include the myobatrachid genera
Arenophryne, Myobatrachus and Metacrinia and
the microhylid genera Austrochaperina and
Cophixalus.

The South American leptodactylid genus
Eleutherodactylus consists of several direct
developing species and the field staging system
developed for E. coqui by Townsend and Stewart
(1985) is the most comprehensive system available
for this breeding mode. We describe some
preserved embryonic material in the Western
Australian Museum of A. rotunda and M. gouldii
and compare them to E. coqui (see Table 3 and
Discussion). Brief comparisons to Australian direct
developing microhylids and also to species from
other Australian endotrophic guilds including the
nidicolous, paraviviparous and exoviviparous
species are made where relevant. These are not
direct developers because they have a hatched
tadpole stage {sensu Altig and Johnston 1989), but
have some similar characteristics to A. rotunda and
M. gouldii in early stages.

MATERIALS AND METHODS

Fifteen embryos of A. rotunda from four clutches
collected near Shark Bay, WA and reared in the
laboratory in April 1981 by J. D. Roberts, were
preserved at irregular intervals in Tyler's fixative
(Tyler 1962) and transferred to 70% ethanol when
accessioned into the West Australian Museum:
WAM R97047-50, 97053, 97057, R97059-60 (see

260

M. Anstis, T.D. Roberts, R. Altig

Figure 1 Stages 3, 4 and 6 (Townsend and Stewart, 1985) of Arenophryne rotunda. A and B = stage 3, anterior and
lateral view; C = stage 4, dorsal view; D, E and F = stage 6, anterior, dorsal and posterior views. Scale bar
represents 1 mm. Arrows indicate features highlighted in bold in Table 1.

Appendix 1). Nine embryos up to stage 13 of
Townsend and Stewart (T&S; 1985) from one clutch
of M. gouldii were collected 15 km north-east of
Perfh, WA, then reared and preserved at irregular
intervals: WAM R97036-40. Six individuals just
prior to hatching and recently hatched from four
marked nests in the field were preserved affer being
excavafed in April 1982: WAM R97041-42, 97044-45
(see Appendix 2). All embryos were reared in total
darkness at ambient room temperatures in the
laboratory which were lowered slightly
(approximately 17-20°C) to better simulate cooling
conditions at the nest sites in the field.

Measurements were taken with an ocular
micrometer attached to a Wild M5 stereoscopic
microscope and drawings were prepared with the
aid of a camera lucida. The photograph (Figure 4F)
was taken with a Nikon D70 digital SLR camera
and 60 mm micro lens. Embryos were staged using
the system of Townsend and Stewart (1985) which
was devised for the direct developing leptodactylid
E. coqui, with additional references to toe
development based on the staging table for aquatic
larvae of Gosner (1960). For the sake of
completeness, descriptive observations on egg
clutches provided for A. rotunda and M. gouldii by

261

Development in Arenophryne and Myobatrachus

Figure 2 Stages 6, 7, 9, and 15 (Townsend and Stewart, 1985) of Arenophryne rotunda. A = stage 6, dorsolateral view
B and C - stage 7, lateral and anterior views; D and E = stage 9, lateral and ventral views; and F = stage Id'

hatching, ventral view. Scale bar represents 1 mm. Arrows indicate features highlighted in bold
in labie 1. o o

Roberts (1984 and 1981, respectively) are
summarised prior to the descriptions for each
species, with additional notes on development
(Roberts, unpubl. data). Embryos in stages 1, 2, 3-7,
9-11, 13 and 15 are described and most stages are
illustrated (Figures 1-4). Brief observations were
made on live embryos during early cell division.
The partial deterioration of the youngest preserved
embryos of A. rotunda (stages 1 and 2), and
specimens of M. gouldii at stages 11 and 13, limited
their descriptions.

Results

The two species have various morphological

characteristics in common. Both have a generally
similar parallel progression through the
developmental stages described in Tables 1 and 2.
Measurements of embryos for each species are
given in the Appendices and Table 3 summarises
key differences between the Australian species and
E. coqui.

Development of Arenophryne rotunda
Clutch sizes of fertilised eggs ranged from 4-11
(mean 7, n = 5). Ovarian development commences
in spring (late August), but ovum maturation is not
completed until late summer. Three females
collected in February 1981 contained 8, 8 and 4 pale

262

M. Anstis, J.D. Roberts, R. Altig

1,
c —
o

2 -
O "O

C OC

i-ia

o to

£

fcC

e o

U

■n -c

^ c/3

C U

-- a lu

a cj ӣ
3 'a C ra '2

dj .t; D O

IX -D t C

-t: °

o g .a ■£■ aj

£ t; > S _y

tn u u

£ £

c/i r3

S c

oj XI _ .is

> « T3 .S-'P

O g ^ "a. -£ ra

g G. 3 :a .£ a

ra -a g o c o

d ra CO >, D."0

^ c

2 ra

J- (/2
03

C 5-

(D

1 ^

^ j:! P
£ - ,
V3

3 3 «
C O

C

O

’cb

GJ

■' ^ V

‘ WD:^

5 O

i -D ^

' r- C

n:

cc

b .!£

O -D

S"H ^
Krs «

c

o

= “O

’5:.£ ^
~ c

U3 c/3 O

£ <

C -O
CJ u-

c ^

3 Y x:

;£ 2 -Y
c/3 ^ —

-c >

C cr>

S 3 ^
3

3 3

5-1 U

rj ’ ■ ^
Cu

tc

Q- D-.«-
D O O-

*73
"O C

c ^

>H ^

3

- 3^2
§ ■“ o

? rs ^

0/ Q>

Excess

c/3 a> .U c/3 Mh J3

uj ;2

m -a £

g Xj o ^
^ X2 :g §
.ECO

-a c
S "a c ^ ^

- - ^ C ,!2

>,x: >

c :3

■o

CO

O

XI

« a

C CO

.2 "o —

bO Cl

2 $ g

T3 2 2

03 Cl b
OJ "O

c^ C

O

CL
(30 ^

S c>

F ^

O o
O- >

2

c §

QJ Ji
c/3 > Xi

F — ^

M g £

o i g

»-i w

c > a 73

Cl -c c ^ C .X
■S oV, > "= 0

c — ' c >,

c c ^-E -2 0

£ 'o .9- E fcC ■p :
cn £ an Xi 2 O .

O -Y i

•2 > ■— 73
6 C.^ S C-

.£ ra S =
c ^ — 9^

'E,

ir £

U ^ 03

^3= s ^
O .gc-2

£ c/3 •£ G;

. g .2P

•33 U

■ .2 73

■g "? -K s.

c o

iF C/3 ^ C

£ E c c

-=: CO a

■3> ”

> >^ a

jC' — ^

^ to CO
C -- 03

^ CO >

^ O

••/ • “ K.

>

O ^ L.

-^X3 o

p -p c o X<i

O §

SP I

E Q

x a, >
S. s ■£

I cu _

, 73 C

1

~ u

to

Development in Arenophryne and Myobatiachus

263

C 0.1 ■

“ > "u
Cu O O

R !r >

n *T 3

" S 3

'r

r 3 U r- —

3 -fi

m

j= '

■C sc
n 3
Cu o
o ^ ^
SC e *

2 .5 «
“ ’a; 4<

slO ii E.

j:' w u

a>

— CD H

i

i-o 2

TO ra o n

^ o

X ^

c

CJD ^ '

.^r M- I

1 g '

s

I 2 ^

Sj ^ ,

>. ^ '
T3 V> :
o a>

Xi >

^ cL

' ^ CD ^

c ^ ^

«j T3
' a» TO o

: -O Q> O

1 'g g 3

L-- 3

° O-’ra .=

1 ~

' rf; > ‘■^1

’ *pS • "• '*

I TO 3 > y

: c «jC _

' 3 1 ;; =S 2

1 -Ji 3 O r

; .5 >.•;

-a ^. -

>

o E

H c
i ° o

-2 o ■- o

T3 73 C >-

2 “ ^ i

■x: T3 — -C
C C O W
OJ TO >^ —

e

bC

a

c

a;

£ o I bO

_ .S “S 'o

TO bC - ■

U O ^

o. c

^ a»
Q. >

O

o ^

u o

I ™ ^

9 3 s >- s

5 -C c/i 1/5 Q- .

2 -D

U- TO

^ _TO ^

a

O O-
^ C D

E E

LT) ^

u

O) OJ
3 TO bC C

£ d C .-3 «

a> ^ c ^ > 3

^ ^ X C CD d

(1> O -D O CD

5 wdI ^

0-3

g 2 c .-g .E 00 '25

3 £ 5 E 5 . 2 .^ TO

42

o ^ -a
S £ ^

2 5S $

£ .E

2 J=
a £

"O

o

Ui

o

E

£ G TO

£ -E E

^*C

-u "g

TO

c

2 E 2 t:

i o o

^ o

‘ >■

OJ £ g i.'

c O

■ ,r, ^ it

TO O ,2 t^ ^

- C TO ^
J 3 CD u- >

V

Q

> c
o -x

t3 O

3 g

u ^
p Cl.

TO rzJ

c

>s-^

' ' u-

0, Q o

> (N XI

C TJ ^ bC

- ^ 3

bo c y ^
x p X c

CD

E

3 2

CD ^

U X

P J _

— ■ T3

O OC C ^

^ E|

£| g S

c y -5 -

1/

1 / g .E

C ra

iu£0-

- g P= ^ -

T3 ^ -P ^ D.

— aj u ^ c g

S g; g 9 ■“ £

+-! u £ E x: T 3

^ "O g 0 /
— g XI xc
g Oh O P

o ’-' .5

X 1 2 ^

E 3 g .E

X -13 C X

-73

O)

C

V

X X
3 .P

, P

n1

TO .5

> E c °

>> S £ I

r -p £i £

in .S^uT

UU CNI

264

p > Xi
C O 01

C // ,bC

D- o s' ra —

, w :r

f/) >

S ^ £

.2 c ^
—

i's

R- E T3

rz 733
j- O

C

D.;

c/2 H

i 3 .£P^ J!5 s .2 00

C'-ri O-uci: C'oo

■*' ^
£ £

' -a c

:3c ^

' S E -7

« Q X5 ^ ^

^ £ C 3 0^

{. 01 tc

P GO a> Jp g c

: i: £ o c

-a 73
c O s

^ 57 c

' E 2 ^

— sy G> -C .5

2 :p 3 oc 60 ^

^ O “ 4» C

-§ >. m JS ;=

t: o £ :^ - O

;J f- fj O o

> .£ c >^ d >

bi) w

E Q

M. Anstis, J.D. Roberts, R. Altig

C/2H-.S rrj::

5 CjOpiji r-TJ „ bo C

o2oi2oCu-E -SP^

T3 JO > -2 >,T3 O 0= O- 5

^ j-<

*7 : Cl u -

Xt ^

'x a.

: £

33 u C C

73

O > CTo c>

-i: _2 >

-S 'c P -2

Cl

-c 'g= D. g P ■§

^ X 3 U C/2 C

E m -

>'£
n uG

g O-C -ii M
p "o o o (fl 22
— ^ c > >-, c «

I 3

' ^ C OJ o»

" O 2 7b is

c 2; x: :S £

n W 01 .2 n

n IT) "2 rS ^ j- 5"

£ g i c Ob

Sd s ^

:! '53 E

s If If

C Xi _D S 70

SP 3

■5 I

5 ■§
-a

^ s s

' S 7

p u
o Xi

' 2 2^
c ^

■? ^
y >^

c

s

X >- o
J. 0 Ct3
C P 73

f = «

1- rc ^

"3 "C
f3 !/2 • —

X (72 TJ X» >

<
rrs ^

X G;
X
£ .t:

GJ rz

,ob .E -§ 8

1- 'o p=
>- "0 r

2- 2 3 ■“
3 ir, 2
•-2x3
.52 c 2 2
> to !X-5

« _s n ■p:

3 ra op E
E > — -p
° ° c «
IN 3

B c.s'-s

Development in Arenophryne and Myobatrachus

265

E ^

5 >, c — ^

Xi -p o

> .SI-

c c:

X 7

n 01 5^

^J■X t; 0.

^ 4 = 2 S

u a>

~zi n ^

"i .§^ >

5 >

•ii ^

C ^

O 3 5

- O

x.

fr: JT'X 3
G tC ii ^ Xi

D: c o a> o

X ^ %i: — at

c ^ Cl*-

^ ^ I E p -S

o 5 ^ 2 o c ^

X .A u OJ

C ^ '"4—

C Ci

■G u Vj c!

U X r -- ■-U.

Cl. C 2 O 2

X t U -C

s Ci

‘IC 2 — {j
X ^

23 ^

O ^ ^

r^ '[Z aS

L- £

Ct 2Z .£ Tj

■p u- — rz O

.£P > c/) S E I,

C Ct C iC

— — Tj O C

I

_D i £ E

- oo ?P ?P

X CO 5 .=

^ c; - ^

5 I

£ X

X 2*1

CC X
2= -Q

^ £

>^X X
^ it:

^ C ^

' 4 - r 3 T 5

I

X Xi u >^ >

X U

: ^

' “O .21

■ -p X

_ ^ X

u- O ^
Xt E

ZZ

•— X

p X c

S CL ^

E ~
P c o c

X X X Cu

2 •- Q

E

X

C -1-1 _

X P E

^ L-

E C c
— o '
~o ^

C £

a;

s; CL 1- o

o 3 3 01 X

>. 0) £ ;: Cl. 5

cj "p u C o) ^

-0 'x ^ X ^ Q
S > S o u

^ c g -g I

r o
,S X a. E

&c*

2*1

60 ^

C ^

.-3

V fcH

”0 ra X c
• 1-1 'P e ^
-* *.C C X
QJ u at ^

^ C E B.

X

E S

X X

r e

X X
2 73

CL

"O 70

X ■" -O £

c .£

^ X uS

-c: JH

^ r3 ^ —

^ V4- 2Z 23 n

0 ° ^ bC £

X c = c o

5 § 55 C c

I, g

^ O X

1 I ^ £ I

' ^ X X E 33

^ O 0 i on U

= .SP 53 o S

■ c -a c d o

266

M. Anstis, J.D. Roberts, R. Altig

Table 3 Differences between available preserved stages of A. rotunda and M gouldii and those of similar live stages for
E. coqui. As no observations of behaviour or ECD (endolymphatic calcium deposits, visible in life) were avail-
able for the Australian genera, these are not included here for E. coqui. Stages prior to stage 4 and features
which are the same for each are excluded. T&S = Townsend and Stewart stages (1985).

T&S Eleutherodactylus

Arenophryne and Myobatraebus

4 • eye bulges distinct

• gill arches present, but no gills present

• tail bud first apparent

• eye bulges discernable

• slight gill arches

• tail bud elongates enough to bend around yolk to one side

5 • forelimbs round to ovoid, external

• forelimb buds beneath operculum

• eyes prominent, unpigmented • eyes partly pigmented

• gill buds first appear from gill arches, gill circulation • indistinct gill arches, no gills

• tail bud elongates enough to bend, small thin fin • tail long (especially Myobatraebus), wraps around

yolk, fins well developed and vascular (Myobatraebus),

6 • forelimbs develop externally

• eye distinct from re.st of head, pupil clear

• gills well developed

• tail over one-half final length, small,
membranous fin

• forelimbs develop beneath operculum

• eye pigment well developed, slight choroid fissure

• no external gills develop

• fins low, poorly developed, slightly vascular
(Arenopbryne), or well dewloped and more

• widely scattered melanophores over dorsum

c'ascular (Myobatraebus)

• some fine melanophore stippling over brain
and vertebral region

7 • hind limbs with obvious knee joints, foot paddles

first evident

• elbow visible on forelimbs

• foot develops up to 3 early toe nubs

• one forelimb begins to break through operculum

• tail Vj final length, fin almost full size, vascular

(Arenopbryne)

• tail at full length, fins remain low (Arenophry'ne)

• beginning of first gut wall within yolk

9 • limbs elongate, digits on hands and feet

• forelimbs fully erupted in Arenopbn'ne; still beneath

• tail V3 full length with full size fin

• pigmentation expands to about midway down yolk

operculum alongside head in Myobatraebus; toe digits
similar to Gosner stages 35-36

• tail full length in both

• pigmentation expands around sides and partly
over venter

• beginning of small, conical projection on inside centre
of lower lip, small notch in centre of upper lip

10 • toes to V, of hatching length

• forelimbs now erupted in Myobatraebus; toes all

• pigmentation dense on dorsum, less on head

individually separate and a bit longer, similar to

Gosner stage 37

• pigmentation uniform over entire dorsum and sides,
denser over head and vertebral region

11-12 • tail full length with full fin

• egg tooth first develops on upper lip by late stage 12

• tail at full length from stage 9

• nonkeratinised conical projection on inside lower lip

13 • toes full length, toe pads first evident

• limbs well advanced, subarticular tubercles develop on
hands and feet, no toe pads

• egg tooth develops keratin

• rio egg tooth, nonkeratinised conical projection fits into
deeper notch in upper lip when mouth closes

• yolk reserve still large

• intestinal development now obscured by pigment

15 • tail remnant half or less of full length at hatching

• no tail remnant prior to hatching

yellow, mature ovarian eggs with a mean diameter
of 4 mm. An adult male found on 1 April 1981 was
sitting on a clutch of seven eggs buried in sand at a
depth of 80 cm and soil temperature in the nest site
at 0845 hr was 25°C. It was not possible to
determine if there was a burrow leading to the frog
and eggs, or a chamber around them, as the sand

caved in during excavation. The eggs in stage 1 had
a mean capsule diameter of 5.5 mm (5.0-6.0 mm)
and initially adhered together in a cluster bv means
of the sticky outer surface of each capsule. As they
became covered in sand, individual capsules
separated. Another clutch at the same depth was
unattended by an adult (Roberts 1984).

Development in Arenophryne and Myobatrachus

267

Figure 3 Stages 5, 6, and 9 (Townsend and Stewart, 1985) of Myobatrachus gouldii. A = stage 5, dorsal view; B and C =
stage 5, dorsal and partial posterior views (line indicates forelimb beneath operculum in B and vent in C); D
and E = stage 6, anterior and posterior views; and F = stage 9, lateral view. Scale bar represents 1 mm.
Arrows indicate features highlighted in bold in Table 2.

Embryos at stages 1, 2-4, 6, 7, 9 and 15 are
described in Table 1, and illustrated in Figures 1
and 2. Embryos are unpigmented during stages 1-4.
Live embryos were not easily studied due to fine
sand over the capsules. A pair of frogs collected on
1 April 1981, laid 11 eggs (clutch 4) some time
between 1-3 April and cleavage furrows were
observed during early to mid-cleavage on 4 April.
Gastrulation and blastopore formation seemed
typical of those described for aquatic tadpoles
(Gosner 1960) and the dorsal lip was a distinct
indentation. Estimating 2 April as the approximate

date eggs were laid, late gastrula was reached after
about 5 days and the neural plate began to form
(stage 2) after about 10 days. Stage 5 was reached
after about 16 days (none preserved) and stage 6
after 24 days.

Hatching and embryonic life span
Well formed froglets from clutch 4 were observed
twitching within the capsules from about 45 days
after the eggs were laid, and after about 50 days
some were unhatched and adpressed tightly against
the capsule wall with no yolk remaining. The last

268

M. Anstis, J.D. Roberts, R. Altig

Figure 4 Stages 9, 10, 13 and 15 (Townsend and Stewart, 1985) of Myobatrachus gouldii. A = stage 9, anterior view,
forelimb beneath operculum; B and C = stage 10, lateral and \’entral views; D = stage 13, lateral view; and E
and F = stage 15, ventral view and photograph of lateral view, just prior to hatching. Scale bar represents 1
mm. Arrows indicate features highlighted in bold in Table 2.

froglet hatched (yolk present in intestinal loops) on
12 June 1981, about 64 days after the eggs were laid
(SV = 10.4 mm, weight 0.22 g). On 5 June, one
froglet from another clutch which had been
collected at stage 1 (dorsal lip to mid-gastrula) on 1
April, was found beginning to hatch with one hind
limb extended through the capsule wall. When
removed and washed to remove sand, the
remaining jelly layers came free and the froglet
became active and soon began to burrow (SV = 9.9
mm, weight 0.24 g). There was a distinct middorsal
stripe and the remnant yolk mass was quite large.

Minimum embryonic life span for this individual
was about 65 days (estimating four days from
fertilisation to mid-gastrula). Three recent
hatchlings 1-2 weeks old measured 11.2-11.4 mm
(mean 11.3 mm) and all weighed 0.25 g.

Development of Myobatrachus gouldii
Clutch sizes of fertilised and ovarian eggs ranged
from 9-38 (mean 25, n = 5; Roberts 1981). The mean
diameter of 28 ovarian eggs from a female caught in
February 1979 is 5.1 mm, and mean diameters of 26,
38 and 9 ovarian eggs from three females collected

Development in Arenophryne and Myobatrachus

269

in November and December 1979 are 5.0, 5.3 and
4.9 mm, respectively (Roberts 1981). The mean
capsule diameter of 23 live embryos from one clutch
at early cell division collected in February 1979 is
7.4 mm (± 0.5 SD). The live embryos are creamy
white, and the surface of the capsules are sticky to
the touch, fairly tough and covered with fine sand.
They soon become like firm, round balls to touch.
Early cell division in live embryos at stage 1 is
similar to that described for Heleioporus eyrei at
Gosner stage 4 (Packer 1966) with four, dorsal
micromeres and two perpendicular, incomplete
cleavage furrows at the vegetal pole. Embryos at
the earliest preserved stages available at stage 5
(T&S) have small limb buds, but unlike E. coqui at
this stage, the eyes are partially pigmented (see
Discussion).

Stages 5, 6, 9, 10, 11, 13 and 15 are described in
Table 2, and illustrated in Figures 3 and 4.

Hatchlings

Hatching was not observed. Five fully formed
froglets just prior to and just after hatching
measured 8.9-10.8 mm (mean 10.1), were exact
miniatures of the adult in form and pigmentation
and began to burrow soon after hatching.

In summary, A. rotunda and M. gouldii share the
following characteristics; large, unpigmented ova
encapsulated with a thin outer layer (that becomes
fairly tough) and an inner jelly layer, no external gills
or adhesive organs, early limb bud development
prior to any optic pigmentation, no spiracle and
forelimbs covered by the operculum until at least
stage 7. Eyes develop pigment gradually from stage
5 and increase noticeably in diameter during stages
6-8. The neural tube is initially raised above the large
yolk (stages 4-5), gradually flattens and broadens
from about stage 6 onward, then as vertebrae
develop, the vertebral column appears as a broad,
thickened ridge. The vent tube begins to develop
from about stage 4, the gut gradually develops from
initial divisions in the yolk at stage 6, into a thick
intestinal coil by stage 9, and a small internal flap
develops inside each naris from about stage 9.

The mouth begins as a small stomodaeal pit at
stage 4, then becomes a simple slit that gradually
widens with jaw development and never develops
the oral mouthparts of a tadpole. During stage 9,
a small flexible conical structure (visible when the
mouth is opened) begins to project upwards
from the inside centre of the lower lip and inserts
into a corresponding notch centred in the upper lip;
this projection becomes more defined in subsequent
stages and the inner margin of the notch deepens
posteriorly. These structures remain in the adult and
are also present in the Australian microhylids (Anstis
unpublished observations). Frog-like features of the
head develop from as early as stage 9.

DISCUSSION

Comparative development

Although not all stages were available for the two
species, an adequate comparative understanding of
their development can be gained from the existing
material, because in those stages where direct
comparison was possible, similarities were quite
evident and differences were minor. No pairs of
adults for either species were observed in amplexus
and the mode of fertilisation could not be
determined, but as eggs are laid in sand, internal
fertilisation could be advantageous. The relatively
large size of the ova and the small clutch sizes are
also characteristic of direct developing species (e.g.,
2.0-10.0 mm and clutch sizes of 1-94; Thibaudeau
and Altig 1999). Based on the similarity in early cell
division noted here between H. eyrei and M.
gouldii, it is likely that cleavage is holoblastic, but
more live material needs to be studied to verify this.
The tough external capsules may protect
developing embryos but do not prevent desiccation
in M. gouldii (Roberts 1981). Death by desiccation
may be a result for embryos of both species if
normal winter rains are delayed.

Arenophryne rotunda has a shorter tail with low
fins and a narrower muscle and much narrower tail
tip than M. gouldii. Myobatrachus gouldii has a
long tail with more prominent fins that provide a
greater degree of vascularisation, a broad muscle,
broadly rounded tip and the tail is well advanced
by stage 5 (Figure 3A, Table 3). Pigmentation is
generally less dense in A. rotunda during stages 4—
9. The forelimbs emerge through the epidermis
during stage 7 for A. rotunda and about stage 10 for
M. gouldii.

Arenophryne rotunda and M. gouldii have
forward burrowing behaviour (Tyler et al. 1980;
Main et al. 1959; Lindgren and Main 1961), and the
minute flap in the narial canal which persists in
adults, possibly prevents sand particles being
lodged in the nostrils during burrowing.

Hatching

Hatching in these species has not been fully
observed, but in A. rotunda, one embryo pushed a
hind limb through the capsule wall at the onset of
hatching. In a description of the hatching process of
the microhylid Cophixalus darlingtoni from Papua
New Guinea, Tyler (1976b) observed that prior to
hatching, the embryo used only abrupt,
outstretched movements of the arms and legs to
split the capsule. From the one observation of the
A. rotunda hatchling, it appears that the hatching
process in A. rotunda, and probably M. gouldii, is
similarly precipitated by abrupt movements of the
limbs, since the outer layer of the jelly capsule is
dry and tough and the embryos already have quite
robust forelimbs.

270

M. Anstis, J.D. Roberts, R. Altig

The turgidity of the egg jellies of direct
developers would seem to require the use of an egg
tooth during hatching, and embryos of species of
Eleutherodactylus are known to poke at the inside
of the egg capsule with the keratinised egg tooth on
the upper lip (Townsend and Stewart, 1985;
Duellman and Trueb, 1986). In A. rotunda and M.
gouldii, however, there is no egg tooth, only the
small, nonkeratinised conical projection described.

Comparisons with Eleutherodactylus coqui

The two Australian myobatrachids differ from E.
coqui in that they deposit eggs in subterranean sites,
they do not develop an egg tooth, the initial
development of the forelimbs is internal prior to
stage 7 or 10 (exposed from stage 4 in E. coqui), the
tail is more advanced in development by stage 5
(M. gouldii) and there are no external gills. Apart
from the differences noted above and those in Table
3, they have a generally similar developmental life
history to E. coqui, but it has not been possible to
adequately compare aspects of gut, mouth and eye
development (choroid fissure), vitelline circulation
and behaviour.

Comparisons with other Australian myobatrachids and
microhylids

Arenophryne rotunda and Myobatrachus gouldii
share key features typical of direct development as
defined by Altig and Johnston (1989) including the
lack of mouthparts and a spiracle. The absence of a
spiracle and mouthparts are also typical of other
Australian endotrophic guilds including Assa
darlingtoni and Bryobatrachus nimbus (Anstis
2002). The paraviviparous genus Rheobatrachus
and the nidicolous species of Geocrinia, however,
have a vestigial spiracle and much reduced
mouthparts, including a few very small lateral
marginal papillae and nonkeratinized jaw ridges
(Anstis unpublished observations; Watson and
Martin 1973; Tyler and Davies 1983). Adhesive
glands are absent in A. rotunda and M. gouldii
and in Spicospina flammocaerulea from
southwestern Australia, a species with aquatic
development in which the hatchlings are fully
supported within thick algae mats (Dziminski and
Anstis 2004), negating the need for adhesive
glands.

Exposed forelimb bud development throughout
embryonic stages is found in the microhylid genus
Cophixalus (Tyler 1976b; Anstis unpublished
observations) and also in the earlier stages of
Philoria, which has terrestrial, nidicolous larvae.
In at least three species of Philoria (P.
sphagnicolus, P. kundagungan and P. loveridgei),
all four limb buds are initially exposed from about
Gosner stage 20, but the forelimbs are soon
covered by the operculum and continue
development internally during larval stages.

breaking through the operculum at Gosner stage
42 (Anstis 1981; De Bavay 1993; Ingram and
Corben 1975; Anstis 2002).

Further studies on the Australian direct
developing genera are required to improve our
understanding of their morphology, physiology
and general biology, including mode of fertilisation,
embryonic behaviour, life span and the hatching
process, so that adequate future comparisons can be
made with other direct developing genera.

ACKNOWLEDGEMENTS
The Western Australian Museum is gratefully
acknowledged for the loan of the specimens
studied. Field work and research was supported by
the University of Western Australia. We thank the
staff at the University of Western Australia for their
assistance. K. Thumm and several reviewers offered
helpful suggestions on the manuscript.

REFERENCES

Altig, R. and Johnston, G.F. (1989). Guilds of anuran
larvae: relationships among developmental modes,
morphologies, and habitats. Herpetological
Monographs 3: 81-109.

Anstis, M. (1981). Breeding biolog}' and range extension
for the New South Wales frog Kyarranus
sphagnicolus (Anura; Leptodactylidae). Australian
Journal of Herpetology 1: 1-9.

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

De Bavay, J. M. (1993). The developmental stages of the
sphagnum frog, Kyarranus sphagnicolus Moore
(Anura: Myobatrachidae). Australian Journal of
Zoology 41: 275-93.

Duellman, W.E. and Trueb, L. (1986). Biology of
Amphibians. McGraw-Hill Book Company, New
York.

Dziminski, M. A. and Anstis, M. (2004). Embryonic and
larval development of the sunset frog, Spicospina
flammocaerulea (Anura: Myobatrachidae), from
southwestern Australia. Copeia 2004: 893-899.

Frost, D.R., Grant, T., Faivovich, J., Bain, R., Haas, A.,
Haddad, C.F.B., de Sa, R.O., Donnellan, S.C.,
Raxworthy, C.J., Wilkinson, M., Channing, A.,
Campbell, J.A., Blotto, B.L., Moler, P., Drewes, R.C.,
Nussbaum, R.A., Lynch, J.D., Green, D. and Wheeler,
W.C. (2006). The amphibian tree of life. Bulletin of
American Mirseum of Natural History. 297: 1-370.
Gitlin, D. 1944. The development of Eleutherodactylus
portoricensis. Copeia 1944: 91-98.

Gosner, K. L. (1960). A simplified table for staging
anuran embryos and larvae with notes on
identification. Herpetologica 16: 183-190.

Ingram, G. J. and Corben, C.J. 1975. A new species of
Kyarranus (Anura: Leptodactylidae) from
Queensland, Australia. Memoirs of the Queensland
Museum. 17: 335-339.

Development in Arenophryne and Myobatrachus

271

Jameson, D. L. 1950. The development of
Eleutherodactylus latrans. Copeia 1950: 44-46.

Lindgren, E., and Main, A.R. (1961). Natural history
notes from Jigalong, IV: Frogs. West Australian
Naturalist?: 193-195.

Main, A. R., Littlejohn, M.J. and Lee, A.K. (1959). Ecology
of Australian frogs, p. 396-411. In Biogeography and
Ecoiogy in Austraiia, eds A. Keast, R. L. Crocker and
C. S. Christian. Junk: The Hague.

Maxson, L. R. and Roberts, J.D. (1985). An immunological
analysis of the phylogenetic relationships between
two enigmatic frogs, M gouidii and A. rotunda.
Journai of Zooiogy (London) 207: 289-300.

Packer, W. C. (1966). Embryonic and larval development
of Eieieioporus eyrei (Amphibia: Leptodactylidae).
Copeia 1966: 92-97.

Read, K., Keogh, J.S., Scott, l.A.W., Roberts, J.D. and
Doughty, P. (2001). Molecular phylogeny of the
Australian frog genera Crinia and Geocrinia and
allied taxa (Anura: Myobatrachidae). Moiecuiar
Phyiogenetics and Evoiution 21: 294-308.

Roberts, J. D. (1981). Terrestrial breeding in the
Australian leptodactylid frog Myobatrachus gouidii
(Gray). Austraiian Wiidiife Research 8: 451-62.

Roberts, J. D. (1984). Terrestrial egg deposition and direct
development in Arenophry'ne rotunda Tyler, a
myobatrachid frog from coastal sand dunes at Shark
Bay, W.A. Austraiian Wiidiife Research 11: 191-200.

Thibaudeau, G., and Altig, R. (1999). Endotrophic
Anurans, p. 170-188. In Tadpoies: the Bioiogy of
Anuran Larvae, eds R. W. McDiarmid and R. Altig.
University of Chicago Press, Chicago.

Townsend, D. S. and Stewart, M.M. (1985). Direct
development in Eieutherodactyius coqui (Anura:
Leptodactylidae): a staging table. Copeia 1985: 423-
436.

Tyler, M. J. (1962). On the preservation of anuran
tadpoles. Australian Journal of Science 25: 222.

Tyler, M. J. (1976a). A new genus and two new species of
leptodactylid frogs from Western Australia. Records
of the Western Austraiian Museum 4: 45-52.

Tyler, M. J. (1976b). Frogs. Australian Naturalist Library.
Collins, Australia.

Tyler, M. J., Roberts, J.D. and Davies, M. (1980). Field
observations on Arenophryne rotunda Tyler, a
leptodactylid frog of coastal sandhills. Australian
Wildlife Research 7: 295-304.

Tyler, M. J., Smith, L.A. and Johnstone, R.E. (2000). Frogs
of Western Austraiia. Western Australian Museum,
Perth.

Wake, M. H. 1978. The reproductive biology of
Eieutherodactyius jasperi (Amphibia, Anura,
Leptodactylidae), with comments on the evolution of
live-bearing systems. Journai of Herpetoiogy 12(2):
121-133.

Watson, G. F., and Martin, A. A. (1973). Life history,
larval morphology and relationships of Australian
leptodactylid frogs. Transactions of the Royai Society
of South Austraiia 97: 33^5.

APPENDIX 1

Collection and preservation dates (day/month/
1981), stage (Townsend and Stewart 1985), and
embryo dimensions (mm, diameters to stage 9,
snout-vent length for stage 15) of Arenophryne
rotunda. WAM = West Australian Museum. N = 1

in each case, see footnote.

Clutch WAM Coll. Pres.

Stage

Dimensions

1

R97056

1/4

1/4

1

5.0

2

R97054

1/4

1/4

1

4.5x4.4

4

R97046

3/4

10/4

2

5.0

4

R97047

3/4

16/4

3

5.0x4.7

1

R97057

1/4

10/4

4

5.2x5.0

4

R97048

3/4

24/4

6’

5.0x4.2

3

R97049

1/4

16/4

6

5.1x4.5

3

R97050

1/4

24/4

7

4.8x4.6

4

R97052

1/4

15/5

7

5.7x4.7

1

R97059

1/4

24/4

9

5. 6x5.5

3

R97053

1/4

22/5

15

6.6 SVL

1

R97060

1/4

1/5

15

4.8 SVL

' external egg diameter of one individual = 6.1 mm.

APPENDIX 2

Collection and preservation dates, stage
(Townsend and Stewart 1985), and embryo
dimensions (mm, diameters to stage 10, snout-vent
length for stages 13-15) of Myobatrachus gouidii.
WAM = West Australian Museum. N = 1 in each
case, except R97037 = 4, range in parenthesis.

Clutch

WAM

Coll.

Pres.

Stage

Dimensions

1

R97036

4/3/81

4/3/81

5

5.7x5.6

1

R97036

4/3/81

4/3/81

5

5.7x5.7

1

R97037

4/3/81

13/3/81

6 5. 1x5.5

(4.8-5.4x5.2-5.8)

1

R97038

4/3/81

27/3/81

9

6.4x6.2

1

R97040

4/3/81

16/4/81

9

6.6x6.4

1

R97039

4/3/81

3/4/81

10

6.1x5.8

2

R97041

14/4/82

14/4/82

13

8.5 SVL

3

R97042

14/4/82 14/4/82

15

10.8 SVL’

3

R97043

23/4/82 23/4/82

15

10.5 SVL

5

R97044

7/4/81

7/4/81

15

8.9 SVL

5

R97044

7/4/81

7/4/81

15

10.0 SVL^

4

R97945

14/4/81

14/4/81

15

10.1 SVL

’ Egg dimensions = 11.3 x 10.8 mm
^ Egg dimensions = 11.3 x 10.5 mm

Records of the Western Australian Museum 23: 273-305 (2007).

Two new species of the Delma tincta group (Squamata: Pygopodidae)

from northwestern Australia

Brad Maryan’*, Ken P. Aplin^ and Mark Adams’

’ Department of Terrestrial Vertebrates, Western Australian Museum,

Locked Bag 49 Welshpool DC, Perth, WA, 6986, Australia
’Australian National Wildlife Collection, CSIRO Division of Sustainable Ecosystems,

PO Box 284 Canberra, ACT, 2601, Australia

’Evolutionary Biology Unit, South Australian Museum, North Terrace, Adelaide, SA, 5000, Australia

* Corresponding author

Abstract - Analysis of allozyme and morphological variation has revealed
that two pygopodid lizard species are presently confused under Delma pax
Kluge, 1974. Delma pax is redescribed and shown to be confined to the Pilbara
region, while a closely related, new species is described from the arid deserts
of western and central Australia. A second new species, endemic to the Cape
Range Peninsula, is also described. Among Western Australian specimens
previously referred to D. borea Kluge, 1974, those from the Pilbara islands are
confirmed; however, all specimens from the Pilbara mainland and arid desert
localities are reallocated to other taxa. Both of the newly described species
belong to an expanded Delma tincta group which displays a complex
biogeographic pattern in northwestern Australia. An updated key to the
Delma spp. of Western Australia is provided.

INTRODUCTION

There are currently 25 described species of
pygopodid lizard known from Western Australia
(Wilson and Swan 2003). Of these, Delma is the
most speciose genus with ten species, five of which
were described in a comprehensive taxonomic
revision by Kluge (1974a). Since this revision,
several additional species and subspecies have been
recognized in Western Australia, including D.
butleri Storr, 1987, D. haroldi Storr, 1987 and D.
fraseri petersoni Shea, 1991. Shea (1991) advocated
synonymy of D. haroldi and D. butleri and this
view is supported in a recent phylogenetic study of
pygopodid lizards by Jennings et al. (2003). These
authors also elevated D. f. petersoni to full species
status and advocated transfer of Aclys concinna
Kluge, 1974 to Delma [as suggested also by Kluge
(1976)]. Aplin and Smith (2001) highlighted further
taxonomic complexity in the widespread D.
australis Kluge, 1974 and in D. butleri, with
preliminary investigations suggesting that both are
composites.

Shea (1991) proposed a Delma tincta group to
include D. tincta, D. borea and D. pax, based on
their similar patterning, the usual presence of a
single large temporal scale bordering each parietal
scale, and their largely allopatric pattern of
geographic distributions. The integrity of this
grouping was strongly supported by Jennings et al.

(2003) analysis of DNA sequence variation among
pygopodid lizards and by their combined
morphological and molecular analysis. However,
neither dataset was able to resolve the relationships
among the three species.

Problems with the taxonomy of the Delma tincta
group were noted by field herpetologists working
in the Pilbara region in Western Australia.
Application of a published key (Storr et al. 1990)
resulted in some Pilbara specimens being identified
as D. borea Kluge, 1974, which is otherwise known
from the Kimberley region. Northern Territory,
western Queensland (Kluge 1974; Shea 1987; Shea
1991) and northwestern South Australia (Ehmann
2005). Similar identification problems were
apparent on the Cape Range Peninsula, where
specimens initially identified as D. pax from the
Exmouth region (Storr and Hanlon 1980) were
subsequently transferred to D. tincta De Vis, 1888
(Storr etal. 1990).

This study presents molecular and morphological
evidence for the recognition of two new species of
Delma in northwestern Australia, both of which
were previously confused with D. pax and/or D.
tincta. The results further reinforce the opinion of
Shea (1987) and others (e.g. Aplin and Smith 2001)
that our knowledge of the taxonomy of the
morphologically conservative genus Delma is far
from complete.

274

B. Maryan, K.P. Aplin, M. Adams

METHODS

Morphological analysis

This study is based on the examination of material
held in the Western Australian Museum (WAM),
Northern Territory Museum (NTM), Australian
Museum (AM) and South Australian Museum
(SAM). The "R" prefix has been omitted for all
WAM, NTM, AM and SAM specimens and, unless
otherwise indicated, specimen registration numbers
refer to the herpetological collection of the Western
Australian Museum. Sex of individuals was
determined by dissection and inspection of gonads;
some immature or poorly preserved individuals
were left unsexed. Head scale terminology,
methods of scale counting and morphometric
measurements follow those used by Shea (1987),
except that all scales between postnasal and
circumocular granules are counted as loreals (after
Storr et al. 1990). Bilateral loreal counts were
averaged if different.

For the purpose of this study the following
morphometries data were taken with digital vernier
calipers and plastic ruler: snout-vent length (SVL),
tail length (Tail L), head depth immediately behind
eye (Head D), head length from tip of snout to
posterior margin of ear (Head L), head width
between ear (Head W), hindlimb length from
junction of limb flap with body to distal tip of flap
(Hindlimb L), mouth length from tip of snout to
oral rictus (Mouth L), rostral depth between dorsal
and ventral extremes of scale (Rostral D), rostral
width between lateral extremes of scale (Rostral W),
snout length from tip of snout to anterior margin of
eye (Snouf L) and eye width between anterior and
posterior extremes of transparent cornea (Eye W).
All measurements are reported in millimeters (mm)
and characters recorded from the right side only.
Specimens preserved in a circular or twisted
position were straightened on a flat surface when
measured for snout-vent and tail length. Tails were
not measured if they were recently broken or
obviously regenerated, as suggested by a clear
break in colouration or patterning. However, x-rays
are necessary to reliably distinguishable original
and fully regenerated tails and these were not taken
during this study. Accordingly, tail measurements
are not used in any taxonomic sense and statistical
data are provided for descriptive purposes only.

For each species, the possibility of sexual
dimorphism in body measurements and scale
counts were explored by Analysis of Variance
(ANOVA), following tests for normality and
homogeneity of variance. Pairwise interspecific
statistical comparisons were similarly conducted,
using pooled-sex or single sex samples as
appropriate. Contrasts are regarded as statistically
significant if p values were less than 0.05.

Because significant sexual dimorphism was

observed in SVL for most species, variation in the
size and proportions of the head was further
examined by Analysis of Covariance (ANCOVA).
For each taxon. Head L was first regressed against
SVL for each sex and ANCOVA used to test for
equality of the slopes and intercepts. All other
dimensions were then regressed against Head L to
test for differences in head proportions between the
sexes.

Interspecific differences were explored by first
preparing bivariate plots of all dimensions against
SVL, separately for each sex. The head dimensions
were then combined using Principal Component
Analyses (PCA; based on covariance matrices) to
produce a simplified representation of the
morphometric variation. All analyses were
performed on untransformed data after biv'ariate
plots showed essentially linear patterns of relative
growth among the various measures and no
significant growth-related increase in variance in
any dimension. Statistical analyses were
performed with MINITAB Release 14.20 or
GenStat Release 6.1.

All WAM specimens of D. borea and D. tincta
collected subsequent to Kluge (1974) were assessed
for three characters; supranasal scale division,
midbody scale row count and the identity of the
supralabial scale positioned beneath the eye to
quantify the intraspecific variation and determine
the effectiveness of these characters for
identification. Due to their geographic proximity to
the new species described herein, all speciniens
from the Western Australian Pilbara islands plus all
available D. borea from the southern sector of the
Northern Territory were also examined (see
Appendix 1).

Allozymc analysis

Frozen liver or heart tissues for allozyme
electrophoresis were obtained from the frozen
tissue collections of the Western Australian and
South Australian Museums for 12 specimens of
typical D. pax, six specimens of the 'desert'
morphotype, four specimens of the 'Cape Range'
morphotype, and three specimens of each of D.
tincta and D. borea from localities in northwestern
Australia. We also included samples identified as
D. butleri (n = 9) and D. haroldi (n = 4), drawn from
across the geographic range of this sibling pair (or
geographically variable taxon; Shea 1991). A total of
41 specimens were represented in the study (see
Appendix 2 for voucher details). We used allozyme
electrophoresis to test the hypothesis that each of
the identified morphotypes within the Delma tincta
group represents a distinct evolutionary species.
The samples of D. butleri and D. haroldi were
included as members of a second species group to
provide a perspective on genetic diversity within
the D. tincta group.

Pygopdids from NW Australia

275

Allozyme electrophoresis was carried out on
cellulose acetate gels (Cellogel©) using the
principles and procedures detailed in Richardson et
al. (1986). The following enzymes or non-enzymatic
proteins displayed sufficient activity and resolution
to allow allozymic interpretation:- aconitase
hydratase (ACON, EC 4.2. 1.3), acid phosphatase
(ACP, EC 3. 1.3. 2), aminoacylase (ACYC, EC
3.5.1.14), adenosine deaminase (ADA, EC 3.5.4.4),
alcohol dehydrogenase (ADH, EC 1.1. 1.1),
carbonate dehydratase (CA, EC 4.2. 1.1), diaphorase
(DIA, EC 1.6.99), enolase (ENOL, EC 4.2.1.11),
esterase (EST, EC 3.1.1), fructose-bisphosphatase
(FDP, EC 3.1.3.11), fumarate hydratase (FUM, EC

4.2.1 .2) , glyceraldehyde-3-phospate dehydrogenase
(GAPD, EC 1.2.1.12), guanine deaminase (GDA, EC

3.5.4.3) , lactoylglutathione lyase (GLO, EC 4.4.1.5),
aspartate aminotransferase (GOT, EC 2. 6. 1.1),
glycerol-3-phosphate dehydrogenase (GPD, EC

1.1. 1.8) , glucose-6-phosphate isomerase (GPI, EC

5. 3. 1.9) , guanylate kinase (GUK, EC 2. 7. 4. 8),
isocitrate dehydrogenase (IDH, EC 1.1.1.42), cytosol
aminopeptidase (LAP, EC 3.4.11.1), L-lactate
dehydrogenase (LDFl, EC 1.1.1.27), malate
dehydrogenase (MDH, EC 1.1.1.37), "malic"
enzyme (ME, EC 1.1.1.40), mannose-6-phosphate
isomerase (MPI, EC 5. 3. 1.8), nucleoside-
diphosphate kinase (NDPK, EC 2. 7.4. 6), dipeptidase
(PEPA, EC 3.4.13), tripeptide aminopeptidase (PEP-
B, EC 3.4.11), proline dipeptidase (PEPD, EC 3.4.13),
phosphogluconate dehydrogenase (6PGD, EC
1.1.1.44), phosphoglucomutase (PGM, EC 5.4.2. 2),
pyruvate kinase (PK, EC 2.7.1.40), superoxide
dismutase (SOD, EC 1.15.1.1), L-iditol
dehydrogenase (SRDH, EC 1.1.1.14) and triose-
phosphate isomerase (TPI, EC 5. 3. 1.1). The
nomenclature used to refer to loci and allozymes
follows Adams et al. (1987).

The allozyme data were analysed in several ways.
In the first instance. Principal Co-ordinates Analysis
(PCoA) was employed to assess the genetic
affinities of individuals, independently of any a
priori grouping based on morphology. Following
an initial PCoA on all 41 specimens analysed,
subsequent PCoAs were then undertaken on each
of the three subsets of specimens which clustered
together and comprised more than one
morphotypic form. The rationale underlying this
'stepwise' use of multiple PCoAs to identify genetic
groups from first principles, plus the
methodological details involved, are presented in
Smith and Adams (2006).

Having defined the major genetic groupings
using PCoA, the phylogenetic relationships among
these groups were explored by constructing a
Neighbor joining tree from pairwise Nei's genetic
distances. This analysis was undertaken using the
NEIGHBOR computer program contained with
PHYLIP 3.5c (Felsenstein 1993), and the resultant

tree drawn using TREEVIEW (Page 1996). A
measure of the robustness of clades was obtained
by bootstrapping the allele frequency data for 100
pseudoreplicates, using a BASIC program written
by M. Adams.

A second measure of genetic divergence was
obtained by calculating the percentage fixed
differences (% FDs) among groups. As argued by
Richardson et al. (1986), the number of "fixed" or
diagnostic differences between populations is more
biologically relevant when determining species
boundaries than are Nei D values, which may be
quite large even in the absence of any genuinely
diagnostic loci.

RESULTS

Morphological analysis

Initial recognition of the potential new species
emerged during a careful examination of all D. pax
specimens and of D. borea specimens from the
southern Kimberley in Western Australia and from
southern Northern Territory. During this
morphological survey, special attention was paid to
the identity of the supralabial scale positioned
beneath the eye and the detail and intensity of head
patterning at various stages of maturity. Using these
characters in combination, it was possible to detect
subtle but consistent differences between three
morphologically diagnosable geographic entities.
These were: (i) true D. pax from the Pilbara with
strong juvenile head pattern that fades early in
ontogeny, (ii) a distinctive, inornate 'Cape Range'
morphotype with similarities to each of D. pax and
D. borea and (iii) a widespread 'desert' morphotype
with a persistent well-developed head pattern. Each
of these taxa appeared to be quite distinct from each
of D. borea and D. tincta.

Before undertaking any morphometric
comparisons, we examined the linear
measurements and scale counts from each of the
putative taxa and geographically proximate
samples of D. borea and D. tincta for evidence of
sexual dimorphism. Statistically significant sexual
dimorphism was observed in each species but with
contrasting expression in each (Tables 1 and 2).

In all putative taxa, females are significantly
longer bodied (SVL) than males. In typical D. pax
and the 'desert' morphotype the mean SVL of
females is 109% and 112% larger than that of
conspecific males (Table 1). This value is slightly
lower in D. borea (107%). The small sample of the
'Cape Range' morphotype gives an estimate of
dimorphism of 112%. These observations are
consistent with Kluge's (1974: 34) observation for
pygopodids that "the female of a given species
almost always attains a larger size than the male."

In each of D. pax, the 'desert' morphotype and D.

276

B. Maryan, K.P. Aplin, M. Adams

Table 1 Summary' of mensural and mcristic data gathered in this study, presented separately for each sex. The
'desert' and 'Cape Range' morphotypes are listed in this and all subsequent tables as D. desmosa an .
tealei, respectively, reflecting the ultimate taxonomic arrangement. Also shown are data for the redefined D.
pax and for geographically proximate samples of D. borea and D. tincta. Values are mean ± one standard

deviation, range and sample size (n).

D. tealei

D. desmosa

D. pax

D. borea

D. tincta

SVL

d

73.7 ±1.49
70-77
(4)

70.0 ±1.47
60-80
(24)

74.1 ± 1.42
55-93
(42)

70.0 ± 1.10
54-88
(42)

66.6 ±1.72
58-72
(9)

2

82.2 ± 2.56
77-88
(4)

78.4 ± 2.53
56-90
(15)

81.0 ±1.46
58-98
(39)

74.8 ± 1.18
54-95
(50)

79,4 ± 4.62
66-92

(5)

Tail L

d

146.3 + 33.0
107-212
(3)

198.2 ± 10.7
85-275
(24)

184.4 ± 7.67
109-271
(32)

172.9 ± 6.86
57-240
(35)

189.3 ± 17.61
102-263
(9)

2

142.2 ± 29.7
87-210
(4)

187.5 ± 13.0
80-257
(14)

198.5 ± 6.31
117-257
(33)

176.5 ± 9.21
53-259
(42)

228.6 ±11.87
200-260
(5)

Ventrals

d

50.5 ± 0.5
50-52
(4)

51.8 ±0.30
48-56
(24)

54.2 ± 0.30
50-58
(43)

53.7 ± 0.46
47-62
(42)

48.7 ± 0.78
44-52
(9)

9

51.5 ±0.5
50-52
(4)

52.9 ± 0.45
50-58
(15)

56.0 ± 0.32
52-60
(39)

54.3 ± 0.34
50-58
(50)

52.4 ± 0.68
50-58
(5)

Head L

d

8.50 ± 0.09
8.36-8.77
(4)

8.04 ± 0.13
7.08-9.28
(24)

8.51 ±0.10
6.96-9.88
(43)

8.13 ± 0.09
6.79-9.39
(42)

7.58 ± 0.16
6.96-8.22
(9)

9

8.81 ± 0.27
8.18-9.39
(4)

8.52 ±0.16
7.21-9.65
(15)

8.72 ±0.10
7.45-9.78
(39)

8.34 ± 0.10
6.68-9.89
(50)

8.20 ± 0,35
7.16-9.02
(5)

Head W

d

5.55 ± 0.09
5.35-5.75
(4)

4.76 ± 0.08
3.91-5.54
(24)

5.11 ±0.08
3.96-6.48
(43)

4.82 ± 0.08
3.91-5.99
(42)

4.52 ± 0.16
3.88-5.24
(9)

9

5.53 ± 0.23
5.11-6.16
(4)

5.11 ±0.15
3.85-6.06
(15)

5.34 ± 0.10
3.71-6.73
(39)

5.01 ± 0.07
3.69-6.06
(50)

4.77 ± 0.32
3.97-5.59
(5)

Head D

d

4.49 ± 0.27
3.85-5.04
(4)

4.39 ± 0.08
3.60-5.13
(24)

4.52 ± 0.07
3.51-5.97
(43)

3.91 ± 0.07
3.16-5.04
(42)

3.79 ±0.10
3.35-4.39
(9)

9

4.44 ± 0.13
4.11-4.73
(4)

4.70 ± 0.15
3.61-5.65
(14)

4.68 ±0.12
3.33-6.25
(39)

4.07 ± 0.07
3.06-5.45
(49)

3.94 ± 0.22
3.44^.52
(4)

Mouth L

d

5.89 ± 0.14
5.51-6.18
(4)

5.83 ± 0.12
4.88-6.84
(24)

6.29 ± 0.12
5.24-8.40
(43)

5.65 ± 0.08
4.63-6.79
(42)

6.23 ± 0.21
4.87-6.99
(9)

9

5.85 ±0.11
5.59-6.13
(4)

6.19 ±0.14
4.99-6.93
(15)

6.49 ± 0.10
5.02-7.99
(39)

5.80 ± 0.07
4.51-6.82
(50)

6.71 ± 0.34
5.76-7.89
(5)

tincta the number of enlarged ventral scales is
significantly higher in females than males (Table 2),
with mean ventral scale counts in females being 1.8,
1.1 and 3.7 (scales) greater than the conspecific male

values, respectively (Table 1). In contrast, D. borea
and the 'Cape Range' morphotype do not show
significant sexual dimorphism in this feature,
although in each case the mean value for females is

Pygopdids from NW Australia
Table 1 (cont.)

Ill

D. tealei

D. desmosa

D. pax

D. borea

D. tincta

Snout L

d

3.53+0.11

3.32-3.78

(4)

3.19 ± 0.06
2.33-3.86
(24)

3.46 ± 0.05
2.74-4.76
(43)

3.28 ± 0.04
2.69^.07
(42)

3.06 ± 0.10
2.76-3.76
(9)

?

3.51 ±0.11
3.35-3.84
(4)

3.39 ± 0.09
2.61-3.93
(15)

3.52 ± 0.05
2.96-4.29
(39)

3.40 ± 0.04
2.63-4.09
(50)

3.39 ± 0.21
2.76^.00
(5)

Rostral W

d

1.88 + 0.05
1.77-2.01
(4)

1.62 ±0.04
0.93-1.88
(24)

1.66 ±0.02
1.26-2.06
(43)

1.54 ±0.02
1.23-1.84
(42)

1.44 ±0.03
1.28-1.66
(9)

2

1.85 ±0.08
1.63-2.06
(4)

1.79 ±0.04
1.51-2.27
(15)

1.78 ±0.03
1.40-2.19
(38)

1.58 ±0.02
1.17-1.98
(50)

1.61 ±0.06
1.48-1.87
(5)

Rostral D

d

0.92 ± 0.07
0.76-1.10
(4)

0.95 ± 0.02
0.52-1.24
(24)

1.03 ±0.02
0.81-1.25
(43)

0.87 ± 0.01
0.71-1.16
(42)

1.00 ± 0.03
0,81-1.14
(9)

2

1.09 ±0.06
0.93-1.25
(4)

1.00 ±0.03
0.67-1.18
(15)

1.06 ±0.01
0.84-1.33
(38)

0.89 ± 0.01
0.64-1.21
(50)

0.98 ± 0.06
0.73-1.06
(5)

Eye W

d

1.73 ±0.13
1.48-2.04
(4)

1.46 ±0.04
0.83-1.74
(24)

1.56 ±0.02
1.23-1.84
(42)

1.50 ±0.02
1.34-1.79
(42)

1.55 ±0.07
1.16-1.88
(9)

2

1.97 ±0.14
1.56-2.13
(4)

1.54 ±0.04
1.16-1.79
(15)

1.56 ±0.02
1.28-1.93
(39)

1.49 ±0.02
1.19-1.81
(50)

1.60 ±0.08
1.39-1.79
(5)

Hindlimb L

d

3.20 ± 0.21
2.91-3.81
(4)

3.63 ± 0.10
2.59-4.42
(24)

3.60 ± 0.09
1.70-4.92
(43)

2.61 ± 0.05
1.73-3.43
(41)

2.63 ± 0.14
1.73-3.16
(9)

2

2.89 ± 0.27
2.08-3.31
(4)

2.89 ±0.13
1,95-3.53
(15)

2.62 ± 0.06
1.95-3.80
(39)

2.10 ± 0.04
1.46-2.95
(49)

2.38 ±0.16
1.99-2.95
(5)

Loreals

d

7.25 ± 0.66
6-9
(4)

6.85 ± 0.22

5-9

(24)

7.37 ±0.18
5-10
(43)

7.50 ± 0.19

4-9

(40)

4.96 + 0.20

4-6

(9)

2

6.25 ± 0.48
5-7
(4)

6.87 ± 0.29

5-8

(15)

7.54 ± 0.19
5-10
(39)

7.24 ± 0.19
4-11
(50)

5.30 ± 0.46

4.5-7

(5)

Hindlimb

scales

d

8 ± 0.00

8

(4)

8 ± 0.00

8

(24)

8.51 ± 0.88
8-10
(43)

8.00 ± 1.25
5-10
(41)

5 ± 0.00

5

(9)

2

8 ± 0.00

8

(4)

8 + 0.00

8

(15)

8.56 ± 0.91
8-10
(39)

7.66 ± 1.17

5-9

(50)

5 ± 0.00

5

(5)

higher than that for males. Kluge (1974) reported
significant sexual dimorphism in mean ventral
counts (always greater in females than males) in
four species of Delwa [D. australis, D. impar
(Fischer, 1882), D. nasuta Kluge, 1974, and D.
tincta], with means differing by 3-5 scales in each
case. Species that Kluge (1974) found to be non-
dimorphic in this attribute include D. borea, D.

fraseri Gray, 1831, D, grayii Smith, 1849, D. inornata
Kluge, 1974, D. molleri Liitken, 1863 and D. plebeia
De Vis, 1888. Kluge's (1974) samples of D. nasuta
and D. inornata were both composites as they both
included specimens subsequently referred to D.
butleri (Storr 1987; Shea 1991).

Sexual dimorphism in hindlimb length is
expressed in each of D. pax, the 'desert'

278

B. Maryan, K.P. Aplin, M. Adams

Table 2 Statistical analysis (ANOVA) of intraspecific sexual dimorphism in selected mensural and menstic characters
for each of D. tealei, D. desmosa, D. pax, D. borea, and D. tincta.

D. tealei

D. desmosa

D. pax

D. borea

D. tincta

SVL

F = 8.218
d.f. = 1,7

P = 0.029

F = 9.600
d.f. = 1,38

P = 0.004

F = 11.380
d.f. = 1,80

P = 0.001

F = 8.311
d.f. = 1,91

P = 0.005

F = 9.753
d.f. = 1,13

P = 0.009

Tail L

F = 0.008
d.f. = 1,6

P = 0.931

F = 0.387
d.f. = 1,37

P = 0.538

F = 2.027
d.f. = 1,64

P = 0.160

F = 0.093
d.f. = 1,76

P = 0.761

F = 2.364
d.f. = 1,13

P = 0.150

Vcntrals

F = 2.000
d.f. = 1,7

P = 0.207

F = 5.231
d.f. = 1,38

P = 0.028

F = 17.595
d.f. = 1,81

P <0.001

F = 1.175
d.f. = 1,91

P = 0.281

F = 10.105
d.f. = 1,13

P = 0.008

Head L

F = 1.108
d.f. = 1,7

P = 0.330

F = 4.933
d.f. = 1,38

P = 0.033

F = 1.995
d.f. = 1,81

P = 0.162

F = 2.274
d.f. = 1,91

P = 0.135

F = 3.407
d.f. = 1,13

P = 0.090

Head W

F = 0.008
d.f. = 1,7

P = 0.931

F = 4.464
d.f. = 1,38

P = 0.041

F = 2.967
d.f. = 1,81

P = 0.089

F = 2.839
d.f. = 1,91

P = 0.096

F = 0.576
d.f. = 1,13

P = 0.463

Head D

F = 0.0296
d.f. = 1,7

P = 0.869

F = 3.741
d.f. = 1,37

P = 0.061

F = 1.246
d.f. = 1,81

P = 0.268

F = 2.591
d.f. = 1,90

P = 0.111

F = 0.498
d.f. = 1,12

P = 0.495

Mouth L

F = 0.047
d.f. = 1,7

P = 0.835

F = 3.296
d.f. = 1,38

P = 0.078

F = 1.508
d.f. = 1,81

P = 0.223

F = 1.706
d.f. = 1,91

P = 0.195

F = 1.543
d.f. = 1,13

P = 0.238

Snout L

F = 0.016
d.f. = 1,7

P = 0.904

F = 3.044
d.f. = 1,38

P = 0.089

F = 0.536
d.f. = 1,81

P = 0.466

F = 2.955
d.f. = 1,91

P = 0.089

F = 2.513
d.f. = 1,13

P = 0.139

Rostral W

F = 0.084
d.f. = 1,7

P = 0.781

F = 5.659
d.f. = 1,38

P = 0.023

F = 7.498
d.f. = 1,80

P = 0.008

F = 1.582
d.f. = 1,91

P = 0.212

F = 6.231
d.f. = 1,13

P = 0.028

Rostral D

F = 2.894
d.f. = 1,7

P = 0.140

F = 1.354
d.f. = 1,38

P = 0.252

F = 1.320
d.f. = 1,80

P = 0.254

F = 0.837
d.f. = 1,91

P = 0.363

F = 0.101
d.f. = 1,13

P = 0.756

EyeW

F = 1.555
d.f. = 1,7

P = 0.259

F = 1.622
d.f. = 1,38

P = 0.211

F = 0.001
d.f. = 1,80

P = 0.979

F = 0.149
d.f. = 1,91

P = 0.700

F = 0.208
d.f. = 1,13

P = 0.657

Hindlimb L

F = 0.839
d.f. = 1,7

P = 0.395

F = 18.702
d.f. = 1,38

P <0.001

F = 65.496
d.f. = 1,81

P <0.001

F = 49.009
d.f. = 1,89

P <0.001

F = 1.103
d.f. = 1,13

P = 0.314

Loreals

F = 1.500
d.f. = 1,7

P = 0.267

F = 0.001
d.f. = 1,38

P = 0.973

F = 0.405
d.f. = 1,81

P = 0.526

F = 0.887
d.f. = 1,89

P = 0.349

F = 0.641
d.f. = 1,13

P = 0.439

morphotype and D. borea, with males having
longer hindlimb flaps in each taxon (Table 1).
Bivariate plots of this measurement against SVL for
each of these taxa (Figure lA-C) show that variance
in hindlimb length is low at early growth stages
(low SVL) and that sexual dimorphism emerges
through life as a result of more rapid growth of the
hindlimb, relative to SVL, in males than females.
The different relative growth trajectory of each sex
is confirmed by results of ANCOVA for each of D.

pax and D. borea (Table 3). Results for the 'desert'
morphotype are not statistically significant but this
may be due to the lack of smaller females in the
sample. Too few individuals of the 'Cape Range'
morphotype were available and too few specimens
of D. tincta were examined to determine the extent
of hindlimb sexual dimorphism in each of these
taxa. Somewhat surprisingly, the number of
hindlimb scales is not sexually dimorphic in any of
the studied species (Table 1). Kluge (1974) did not

hind limb length (mm) ® hind limb length (mm) ^ hind limb length (mm)

Pygopdids from NW Australia

279

4 . 5 -

4 . 0 -

3 . 5 -

3 . 0 -

2 . 5 -

2 . 0 -

1.5

1.0-L

50 60 70 80 90 100

4.5

4.0

3 . 5 -

3.0

2 . 5 -

2.0

1.5
1.0

• I

« I
t*t

C

5 . 0 -

4 . 5 - '

4 . 0 -

3 . 5 -

3 . 0 -

2 . 5 -

2 . 0 -

1 . 5 -
1.0 •

-r 1 r

30 40 50 60 70 80 90

snout-vent length (mm)

Figure 1 Bivariate plots of Hindlimb L against SVL for
each of D. pax (A); D. borea (B) and the
'desert' morphotype (C). In each plot, males
(diamonds) are distinguished from females
(squares) and unsexed individuals (circles).
The plots demonstrate that male-biased
sexual dimorphism in hindlimb length in
each species arises through more rapid
growth of this appendage in males than in
females.

present hindlimb lengths but did report a lack of
sexual dimorphism in hindlimb scale counts in all
Delma species.

For head dimensions, only a few statistically
significant or near significant contrasts (Table 2) are
observed between the sexes in each of D. pax (Head
W, Rostral W), D. borea (Head W, Snout L), with
mean values for females exceeding those of males
in all cases. In contrast, the 'desert' morphotype
shows female-biased sexual dimorphism in most
head dimensions. Examination of bivariate plots of
each head dimension against SVL (Figure 2A-C)
indicate a clear lack of sexual dimorphism in D.
borea, with no differences in the slopes or intercepts
of regression lines, but a more complex situation in
each of D. pax and the 'desert' morphotype. In these
taxa, regression slopes are slightly higher in males
than females, indicating a more rapid growth of the
head relative to SVL in males than in females.
However, for both taxa ANCOVA-s were not
significant for any head dimension against SVL
(Table 3). Bivariate plots of all other head
dimensions against Head L for each taxon failed to
reveal any sexual dimorphism in head proportions
(Figure 2D-F for Head W against Head L) and this
was also confirmed by non-significant results from
ANCOVA (not shown). Males in each of these
species of Delma thus develop a slightly larger head
than females through life, but without any obvious
proportional changes.

No sexual dimorphism was observed in loreal
counts. This finding is consistent with that of Kluge
(1974) for other Delma species and for pygopodids
generally. Uniquely among pygopodids, Lialis
burtonis Gray, 1835 is sexually dimorphic in the
number of supralabial scales (Kluge 1974; 132).

Table 4 gives a summary of pairwise statistical
comparisons among D. pax, D. desmosa and D.
borea for various measurements and scale counts,
with separate comparisons for each sex. Comments
on statistically significant contrasts are provided
under the individual species accounts.

For head dimensions, interspecific contrasts were
examined separately for each sex by bivariate plots
and then by PCA (results not shown). No clear
interspecific differences were found. Instead, the
head appears to be remarkably conservative in
proportions among all of the species examined.

Allozymc analysis

We were able to score a total of 43 presumptive
allozyme loci. Nine loci (Estl, Gapd, Idhl, Lap,
Ldhl, Ldh2, Mdb, Pk, and Tpi) were invariant and
hence uninformative for assessing genetic
relationships among individuals. Appendix 2
presents the allozyme profiles of the 41 specimens
examined at the 34 variable loci.

The initial PCoA on all specimens revealed the
presence of four discrete clusters, labeled A-D on

280

B. Maryan, K.P. Aplin, M. Adams

“D

10.0
? 9.5

g .

sz 9.0
S 8.5
8.0

7.5
7.0

6.5

__^ 10.0

E

^ 9.5
_c

c 9.0
_a>

TJ

m 8.5

JZ

8.0

7.5

7.0

• « •
*•

if-;- •

si

* * •

E 9.5

g

S 9-0

cn

^ 8.5

"O

TO

g 8.0

7.5

7.0

60 70

90 100

% :«!•

70 80

B

60 70 80

snout-vent length (mm)

I

£ 6.0
■g

I-

TO

OJ

5.0

4.5

**« •
•1 •

,06 • •

6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0

D

—I 1 1 1 1 r

7.5 8.0 8.5 9.0 9.5 10.0

E

E

E 6.0

8.0 8.5 9.0 9.5

head length (mm)

Figure 2 Bivariate plots of Head L against SVL and Head W against Head L for each of D. pax (A, D); D. borea (B, E)
and the 'desert' morphotype (C, F). In each plot, males (squares) are distinguished from females (circles). The
plots demonstrate slightly more rapid growth of the head relative to SVL in each of D. pax and the 'desert'
morphotype, and a lack of differentiation between the sexes in head proportions.

Figure 3. As shown, only group D comprised
specimens of a single a priori taxon (i.e. D. tincta).
All other groups were composites; group A
comprised specimens displaying either the pax or
'desert' morphotype, group B contained specimens
referable to the 'Cape Range' morphotype or D.
borea, and group C was a mix of both D. butleri
and D. haroldi.

In order to determine whether all taxa were

independently diagnosable by their allozyme
profiles, a second round of PCoAs was undertaken
on individuals within each of the composite groups
A, B, and C (Figure 4). Unequivocal discrimination
was indeed obtained between D. pax and the
'desert' morphotype (Figure 4A) and between D.
borea and the 'Cape Range' morphotype (Figure
4B). The outcome was more complex for group C,
since while haroldi was distinguishable from

Pygopdids from NW Australia

281

Table 3 Statistical analysis (ANCOVA) of intraspecific sexual dimorphism for hindlimb and selected head dimensions
for each of D. borea, D. desmosa and D. pax. Regression values are slope (a) ± s.e. and intercept (i). All
regressions are highly significant and all contrasts passed tests of homogeneity of variance.

Comparison

SEX

Hindlimb L
vs SVL

Head L
vs SVL

Head L vs
Head W

Loreals

D. borea

S

a = 0.052 ± 0.007
i = 0.590

F = 5.67
d.f. = 1,92

P = 0.019

a = 0.073 ± 0.006
i = 2.99

F = 2.06
d.f. = 1,94

P = 0.155

a = 0.565 ± 0.065
i = 0.230

F = 0.01
d.f. = 1,94

P = 0.921

9

a = 0.025 ± 0.009
i = 0.30

a = 0.061 ± 0.007
i = 3.88

a = 0.554 ± 0.086
i = 0.310

D. pax

<S

a = 0.065 ± 0.010
i = -1.15

F = 35.93
d.f. = 1,79

P < 0.001

a = 0.082 ± 0.009
i = 2.41

F = 2.44
d.f. = 1,79

P = 0.123

a = 0.547 ± 0.103
i = 0.443

F = 2.54
d.f. = 1,80

P = 0.115

9

a = 0.027 ± 0.005
i = 0.27

a = 0.066 ± 0.005
i = 3.33

a = 0.752 + 0.077
i = -1.297

D. desmosa

3

a = 0.072 ± 0.012
i = -1.83

F = 0.24
d.f. = 1,36

P = 0.624

a = 0.085 ± 0.008
i = 2.33

F = 0.52
d.f. = 1,36

P = 0.477

a = 0.676 ±0.111
i = -0.699

F = 0.19
d.f. = 1,36

P = 0.662

5

a = 0.059 + 0.023
i = -2.20

a = 0.073 ± 0.015
i = 2.87

a = 0.572 ± 0.208
i = 0.450

Table 4 Statistical analysis of pairwise interspecific differences between each of D. desmosa, D. pax, and D. borea for
selected mensural and meristic characters. The available sample of D. tealei is too small to yield significant
results.

Comparison

SEX

SVL

Ventrals

HeadL

Eye W

Hindlimb L

Loreals

pax vs desmosa

3

F = 3.582
d.f. = 1,65

P = 0.063

F = 27.707
d.f. = 1,66

P <0.001

F = 7.002
d.f. = 1,66

P = 0.010

F = 5.452
d.f. = 1,65

P = 0.023

F = 0.037
d.f. = 1,66

P = 0.847

F = 3.149
d.f. = 1,66

P = 0.081

?

F =0.817
d.f. = 1,53

P = 0.370

F = 27.178
d.f. = 1,53

P <0.001

F = 1.113
d.f. = 1,53

P = 0.296

F = 0.318
d.f. = 1,53

P = 0.575

F = 3.936
d.f. = 1,53

P = 0.053

F =3.560
d.f. = 1,53

P = 0.065

pax vs borea

3

F = 5.118
d.f. = 1,83

P = 0.026

F = 0.612
d.f. = 1,84

P = 0.436

F = 7.084
d.f. = 1,84

P = 0.009

F = 4.618
d.f. = 1,83

P = 0.035

F = 73.475
d.f. = 1,83

P <0.001

F = 0.237
d.f. = 1,82

P = 0.628

2

F = 11.196
d.f. = 1,88

P = 0.001

F = 12.295
d.f. = 1,88

P = 0.001

F = 6.860
d,f. = l,88

P = 0.010

F = 6.391
d.f. = 1,88

P = 0.013

F = 45.916
d.f. = 1,87

P <0.001

F = 1.173
d.f. = 1,88

P = 0.282

desmosa vs borea

3

F = 1.232
d.f. = 1,107

P = 0.270

F = 16.981
d.f. = 1,108

P <0.001

F = 3.264
d.f. = 1,108

P = 0.074

F = 1.015
d.f. = 1,65

P = 0.318

F = 10.466
d.f. = 1,107

P = 0.002

F = 4.551
d.f. = 1,63

P = 0.037

2

F = 0.131
d.f. = 1,103

P = 0.718

F = 11.240
d.f. = 1,103

P = 0.001

F = 0.004
d.f. = 1,103

P = 0.952

F = 1.385
d.f. = 1,64

P = 0.244

F = 19.146
d.f. = 1,102

P <0.001

F = 0.931
d.f. = 1,64

P = 0.338

butleri, the latter also displayed considerable
heterogeneity which broadly manifested itself as
three geographically-based clusters herein referred
to as 'western', 'central', and 'eastern' (Figure 4C,
Appendix 1). Thus the final outcome of the four
PCoAs was the recognition of nine Operational
Taxonomic Units (OTUs) among the 41 specimens
examined, each diagnosable from all others using
stepwise PCoA of the allozyme data. Table 5
compares allele frequencies for each OTU at the 34
informative loci, while Table 6 presents pairwise

genetic distance (Nei D and % fixed difference)
values.

In general, each of the OTUs is well-differentiated
genetically from all others, with only five of the 36
pairwise comparisons involving fewer than six
fixed differences (equivalent to 12%FD). Regarding
the five exceptions, all but one occurred among the
four OTUs identified within group C {butleri/
haroldi); indeed, in the case of D. haroldi versus
'central' D. butleri the two OTUs shared alleles at
all loci (0 %FD, Table 6).

282

B. Maryan, K.P. Aplin, M. Adams

Figures Principal Co-ordinates
Analysis of the 41
specimens included in
the allozyme study. The
'desert' and 'Cape Range'
morphotypos are listed
in this and all subsequent
figures as D. desmosa
and D. tealei, respect-
ively, reflecting the ulti-
mate taxonomic arrange-
ment. The relative PCoA
scores have been plotted
for the first (X-axis) and
second (Y-axis) dimen-
sions, w^hich individually
explained 43% and 14%
respectively of the total
multivariate variation.

• D. pax

'Aa^

0 D. desmosa

U 1 B

A D. tealei

A D. borea

+ D. tincta

\ + + \

□ D. butleri

■ D. haroldi

-f \

D

a \

■

□

A vC ■ ■

.,/■■■ C

^ /CD

O iil \

• • • •

The only other pairwise comparison not
characterized by multiple fixed differences is that
between typical pax and the 'desert' morphotype.
These OTUs displayed a single fixed difference (=
2%FD) and a modest Nei D of 0.08 (Table 6). In
contrast, the 'Cape Range' morphotype shows fixed
differences at 21% of loci to each of D. pax and the
'desert' morphotype (Nei D = 0.25-0.27) and a
closer association with D. borea (12%FD and Nei D
= 0.16). Pairwise contrasts within the D. butleri/
haroldi group range from 0-14% for fixed
differences and 0.04 to 0.21 for Nei D, with a closer
affinity between D. haroldi and eastern D. butleri
on the one hand, and between 'western' and
'central' populations of D. butleri on the other.

The Neighbour-Joining tree constructed from
pairwise Nei D values (Figure 5) shows a deep
division of the OTUs into two groups, one
containing D. butleri and D. haroldi, and the other
containing D. tincta, D. pax and D. borea and both
the 'desert' and 'Cape Range' morphotypes. Within
this latter group, D. tincta appears to be the most
divergent, with the remaining four OTUs forming a
common group made up of two pairs of OTUs: pax
-t- 'desert' and borea + 'Cape Range'.

Figure 4 Principal Co-ordinates Analyses for each of
the three groups identified in the initial PCoA
(Figure 3). A) PCoA of group A specimens;
the first and second dimensions individually
explained 34% and 13% respectively of the
total variance. B) PCoA of group B specimens;
the first and second dimensions individually
explained 75% and 16% respectively of the
total variance. C) PCoA of group C
specimens; the first and second dimensions
individually explained 26% and 15%
respectively of the total variance. Codes,
legends, and general layout as per Figure 3.

• D. pax
O D. desmosa

A

B

A D. tealei
A D. borea

western” □

“eastern”

“central”

O D. butleri
■ D. haroldi

Pygopdids from NW Australia

283

Table 5 Allele frequencies at 34 variable loci for the nine OTUs identified in the allozyme study. For polymorphic loci,
the frequencies of all but the rarer/rarest alleles are expressed as percentages and shown as superscripts
(allowing the frequency of each rare allele to be calculated by subtraction from 100%). A dash indicates no
genotypes assignable at this locus.

Locus

pax

desmosa

tealei

bore a

tincta

butleri

"western"

butleri

"central"

butleri

"eastern"

haroldi

Aconl

a

a

a

a

a

b»»,c

a“b

b

Acon2

d^b

c

c»^d

b“(P^a

a««,b

d

d“e

a^*,d^^,c

Acpl

a

a

a

b

a

a

a

a

a

Acp2

b

b

b

b

a

a

a

a

a

A eye

b“a

b

a

a

a'’^b

a

a

a

a^^c

Ada

b

b®c

b

b

b'^^d

b

b

b

b^a

Adhl

b

c^’^b

b

b

b

b

b*^a

b

b

Adh2

d

a^“,c

c

c

a

c

b“ c

c

b

Ca

a

a

a

a

a

b

b

b

b

Dia

b^’,e

b™,a

h««,j

h66 gl7 d

g“h

c“P^g'^i’^h'^j

g

-

Enol

c

c

b

c

c

c“,b^®,a

c

c

c^^b

Est2

d“e

d

b

b'>^,c

b^5,a

b

b

b

b*®,c

Fdp

c

c’^b

a

a

a

a

a

a

a

Fum

b

b

b

b

b

b»«,d

b^d

b* c

b"^a'^d

Gda

c

c

e

d

d

c“^,a

c

c

c^b

GIo

c*b

c“,a

c

c

c

d

d

d

d

Gotl

b*,a

b

b

b

b

b

b

c

b

Got2

b“a

a

b

b

b

c

c

c

c^^d

Gpd

b

b

b

b

b^a

b

b

b

b

Gpi

b

b«a

b

b

b

c

c

c

c

Guk

b"=,a

b

b

a“,b

a

a

a

a

a

Idh2

c

c

c

c

c

b^a

b

b

b^^a>^d

Mel

c

c%a

c

c

c

b"^c

b

b

c^b

Mpi

b

b

b

b

b

a

a

a

a

Ndpkl

a

a

a

a

a

b

b

b

b

Ndpk2

b

b“,c

b

b«^a

b

b

b®d

b

b

PepA

c“,b

c

c

c

c

c*^,a

c

d

c

PepB

e

e

e

e^^b

e“,b

a

b“,e^^a'^c

PepD

c^,d

d5^C

c*^,a

d“c“f

d“b

d“c“ f

d

c*,d

d“f",e’lg

6Pgd

b®,c

b

h”,a

b

b<’^c

d

d

e

d

Pgml

c

c™,b

e

g

d

c‘’^,a

e««,f

f

e“f

Pgm2

a

a

a

a“b

a

a

a

a

a

Sod

d*g

d

d

c

d

d

d^^a

d

a“,b^^e'^f

Srdh

b’^a

b

b

b

d%b

b“ c

c

c

c

Table 6 Genetic distance matrices for the nine OTUs of Delma identified by the Principal Co-ordinates Analyses.
Lower triangle = %FDs; upper triangle = Nei Ds

OTU

pax

desmosa

tealei

borea

tincta

butleri

"western"

butleri

"central"

butleri

"eastern"

haroldi

pax

-

0.08

0.25

0.27

0.27

0.55

0.55

0.68

0.53

desmosa

2

-

0.27

0.30

0.31

0.58

0.61

0.74

0.59

tealei

21

21

-

0.16

0.27

0.48

0.53

0.68

0.49

borea

21

23

12

-

0.17

0.50

0.54

0.67

0.49

tincta

21

21

16

12

-

0.40

0.43

0.56

0.40

butleri 'western'

37

40

35

35

28

-

0.07

0.13

0.10

butleri 'central'

40

42

37

35

33

5

-

0.15

0.04

butleri 'eastern'

47

49

49

44

42

14

9

_

0.21

haroldi

38

43

36

36

31

7

0

12

_

284

B. Maryan, K.P. Aplin, M. Adams

—D. pax

D. desmosa

■D. tealei

■D. borea

D. tincta

rD. butleri - western

D. butleri - eastern

D. butleri - central

for treating these morphologically distinct
populations as discrete evolutionary lineages. First,
in addition to their single fixed difference at the
Adh2 locus, they also displayed major differences
in allele frequency at a further three loci (Acon2^
DP = 63%, AdhP DP = 75%, Gof2’ DP = 83%; Table
5). Second, the spatial distribution of variation in
each of the 'near fixed' loci within each taxon is not
clustered in specific localities around the periphery
of the Pilbara, as might be expected if regular gene
flow was occurring between the two morphotypes,
nor is it arranged in any geographic pattern that
might be identified as a genetic dine. Last, each of
the D. pax and the 'desert' morphotypes have quite
large geographic distributions (see below), which
nevertheless appear to abut around the perimeter of
the Pilbara uplands, involving a total distance of
many hundreds of kilometres. Such a geographic
arrangement ought to facilitate gene flow between
the two forms, yet they appear to maintain their
morphological distinctiveness across their ranges.

In the following section we diagnose two new
species of Delma, redefine D. pax as a taxon
restricted to the Pilbara uplands, and comment on
the distribution and morphology of D. borea
populations in Western Australia.

'—D. haroldi

Figures Neighbor-joining tree depicting the
phylogenetic affinities of nine OTUs of
Delma, based on Nei distances and rooted at
the midpoint of the longest branch. Bootstrap
proportions of 50% or greater from 100
pseudo-replications are indicated for all
nodes. Scale represents a Nei D of 0.1.

The case for recognition of the 'Cape Range'
morphotype as a distinct species is strongly
supported by the genetic evidence. Although this
population was historically associated first with D.
pax and then with D. tincta, its genetic affinities
clearly lie with D. borea. Nevertheless, the 'Cape
Range' morphotype and D. borea are well-
differentiated genetically, with a total of six fixed
differences and a Nei D of 0.16 between them. This
is equivalent to the observed genetic differentiation
between D. borea and D. tincta (12%FD and Nei D
of 0.17), two species that are broadly sympatric (but
rarely syntopic; Shea 1991) across northern
Australia. Furthermore, the 'Cape Range'
morphotype is readily distinguished on several
morphological criteria from D. borea (see below).

In contrast, the 'desert' morphotype is weakly
differentiated from typical D. pax, with only a
single observed fixed difference in their allozyme
profiles. Despite this, there remains a strong case

SYSTEMATICS

Delma tealei sp. nov.

Figures 6-7

Material examined

Holotype

153811 in the Western Australian Museum, an
adult female collected on 12 September 2003 by B.
Maryan and D. Algaba on Charles Knife Road, Cape
Range, Western Australia (22°07'08"S 114°03'44"E).
Liver sample preserved in -75°C ultrafreeze at W.A.
Museum.

Paratypes

Sex indicated in brackets.

Western Australia: 52934-35 (both F) Shothole
Canyon (22°03'S 114°02'E); 82532 (M) 6 km W
Exmouth (21°56'S 114°04'E); 88548 (F) 2 km E Yardie
Creek mouth (22°20’S 113°49'E); 102837 (M) Cape
Range National Park (22°09'01"S 113°59'52"E);
153813 (M) 2 km S Yardie Homestead Caravan Park
(21°53'37"S 114°00’34"E); 153819 (M) Shothole
Canyon (22°03'49"S 114°00'42"E).

Diagnosis

A moderately small species of Delma (SVL up to
88mm) with modally 14 midbody scales, two pairs
of supranasals and relatively plain colouration apart
from variegated ventrolateral scales on forebody.

Pygopdids from NW Australia

285

Figure 6 Holotype (153811) of Delma tealei, photographed in life (B. Maryan).

Adults lack any trace of dark markings on head
or neck. Differs from the otherwise similar D. borea
in lower modal midbody scale count, typically the
third supralabial positioned below the eye, absence
of pattern on head and neck in adults and longer
hindlimb flaps in both sexes.

Description

Rostral with obtuse apex, penetrating between
rostral supranasals; two pairs of supranasals, caudal
pair much larger; rostral supranasals in moderate
contact with first supralabial; caudal supranasals in
point to moderate contact with nostril; postnasal
single; loreals 5-9, subequal; suboculars 3-4;
supraciliaries 5, fifth much larger; supraoculars 2,
second wider than first; supralabials 5, third
elongate and positioned below eye, fifth much
smaller; infralabials 4, third elongate; occipital scale
present; upper temporals 2. General form of head
and details of scalation illustrated in Figure 7.
Midbody scale rows 14; transversely enlarged
ventral scales 50-52; hindlimb scales 8.

Morphological Variation; 82532 has a small scale
partly wedged between second and third
supralabial on left side; 52934 has upper temporals
divided on both sides.

Colouration and patterning

In preservative, upper and lateral surfaces light
grey or light to dark brown, head slightly darker.
Supralabials pale to dark brown and infralabials
pale with brownish vertical streaks or blotches
mostly centered on first and third sutures along
series. Lateral scales on forebody typically
variegated, bases greyish white to white, centres
blackish (mostly a dark smudge) and apices greyish

Figure 7 Head scalation of Delma tealei holotype
(153811) in lateral (top) and dorsal (bottom)
views.

to brownish grey. Variation includes individuals
(e.g., 102837) with barely discernible variegation;
and others (e.g., 52934) with distinct white-centred
scales bordered by black smudging and with light
brown apices. Lower surface greyish white or white
with diffuse dark smudging on posterior edges of
some scales.

In life, a subtle pinkish flush is noticeable on the
dorsal and lateral scales immediately forward of
and behind vent (e.g., 153811, 153813, 153819); this
colour is lost in preservative.

No immature specimens are available for this

286

B. Maryan, K.P. Aplin, M. Adams

species. Accordingly, it is not known whether or
iiot it displays the ontogenetic fading of head and
neck patterning displayed by D. pax and some D.
borea (see below for details).

Details of Holotype

Snout-vent length (mm) 79; tail 210; loreals 8;
midbody scale rows 14; ventrals 52; hindlimb scales
8. Light grey upper and lateral surface, supralabials
smudged grey aligned with dark vertical streaks on
infralabial sutures 1-3, variegated lateral scales on
forebody bases greyish white, apices greyish and
some scales with blackish centres. Lower surface
white and unpatterned.

Etymology

Named for zoologist Mr Roy Teale, in recognition
of his contribution to Western Australian natural
history and the collections of the Western
Australian Museum, and his active support of
numerous taxonomic research projects.

Distribution and sympatry

Apparently restricted to the Cape Range
Peninsula of North West Cape in Western Australia

(Figure 8), a heavily dissected limestone plateau,
sparsely vegetated with Triodia, shrubs and low
eucalypts; gorges within the range are more heavily
vegetated (Storr and Hanlon 1980).

Three other species of Delma are recorded on the
Cape Range Peninsula. Delma nasuta Kluge, 1974
and D. tincta De Vis, 1888 are known from multiple
localities and the regional sample is consistent with
other populations of these widespread taxon. A
third taxon, currently associated with D. australis
Kluge, 1974 of southern Australia, is known from a
single specimen (132470) collected at Shothole
Canyon. Specimens of D. tincta were collected on
the same occasion as D. tealei at four localities
(Shothole Canyon, 52933; Cape Range National
Park, 102838; 2 km S Yardie Homestead Caravan
Park, 153814; Charles Knife Road, 153820).

Comparison with other species

Delma tealei will be compared first with D. borea
and D. tincta, the two species with which it is most
similar to, then with each of the regionally
sympatric D. nasuta and D. australis, and finally
with geographically distant congeners with which
it shares important characters.

Pygopdids from NW Australia

287

Delma tealei is morphologically most similar to
populations of D. borea on the western Pilbara
islands (e.g., 28656, 37371, 37406, 48559). These taxa
are similar in body size and share two pairs of
supranasals and some indication of variegated
ventrolateral scales on the forebody. However, all
populations of D. borea have higher midbody scale
row counts (modally 16 versus 14), some indication
in adults of pale brown bands on the head and neck,
and typically the fourth supralabial positioned
below the eye (typically the third in D. tealei).

Delma tealei and D. tincta share modally 14
midbody scales and indication of variegated
ventrolateral scales on the forebody. Delma tincta
has one pair of supranasals (two pairs in D. tealei)
and dark dorsal head markings that are especially
distinctive on immature specimens but remain
visible on most adult specimens (Storr et al. 1990)
including individuals (e.g., 52933, 102838, 153814)
from some of the same localities as the new species.
Delma tincta also has smaller hindlimb flaps than
D. tealei (Table 1). The hindlimb scale counts are
correspondingly lower in D. tincta (5 versus 8).

Delma nasuta from Cape Range Peninsula and
elsewhere grows to larger size (SVL up to 112mm
versus 88mm) and has a more elongate snout,
higher midbody scale row counts (modally 16 or 18
versus 14), more loreals (6-23 versus 5-9) and a
reticulated or spotted body pattern formed by a
dark spot or emargination on numerous body
scales.

The Cape Range Peninsula specimen of D.

'australis' is smaller than D. tealei (SVL 57mm
versus up to 88mm) and further differs in having
one pair of supranasals (versus two), more midbody
scale rows (modally 18 versus 14), and very
different patterning that includes fine black lateral
bars on the neck and throat. It also shares an
unusual arrangement of the loreal scales with
typical D. australis (loreal row is broken by
prefrontal-supralabial contact versus continuous in
D. tealei).

Delma pax and D. dcsmosa are both allopatric to
D. tealei (Figure 8). They are distinguished by
higher midbody scale row counts (modally 16
versus 14) and uniformly coloured ventrolateral
scales on the forebody. Adult D. desmosa differ
further by tbe ontogenetic retention of dark dorsal
head markings.

Habitat

The holotype was raked (using a 3-prong
cultivator) from dead Triodia clumps on a low hill
vegetated with Triodia and sparse shrubs on brown
stony loam (Figure 9). The paratypes were collected
in the same manner except for 88548 that was found
beneath an exfoliated limestone slab on heavy
Eucalyptus leaf litter, and 102837 that was pit-
trapped in a valley floor surrounded by low
limestone breakaways (P. Kendrick, pers. comm.).
All collection sites for this species combine
hummock grass and limestone, an association that
is overwhelmingly dominant on the Cape Range
Peninsula.

Figure 9 Low stony hills covered with dense Triodia at the Charles Knife Road, Cape Range WA, the type locality for
Delma tealei (B. Maryan).

288

B. Maryan, K.P. Aplin, M. Adams

Remarks

Delma tealei was originally thought to represent
a southern outlier population of D. pax when first
collected during herpetofaunal surveys (Storr and
Hanlon 1980). The combination of two pairs of
supranasals and third upper labial in subocular
position probably influenced this decision.
However, in preparation for publication of a
handbook to the gekkonoid lizards of Western
Australia (Storr et al. 1990), fresh examination
resulted in the transfer of this population to D.
tincta. This action probably reflected the shared
condition of 14 midbody scale rows in each of D.
tealei and D. tincta.

In Storr et al. (1990) the species account for D.
tincta (incorporating D. tealei) included the
statement 'usually one (occasionally two) pairs
of supranasals' but without reference to a
specific population. We consulted Kluge (1974)
and also examined all Western Australian
Museum holdings of D. tincta (see Appendix 1)
to ascertain whether this statement holds true
following exclusion of D. tealei. Kluge (1974)
examined 168 specimens and encountered a
single individual (22323) with unilateral division
(left side) of the supranasals. Similarly, in a total
of 163 specimens examined by us, we found only
one example (104426) with bilateral division into
two pairs of asymmetrically shaped supranasals,
and another (85190) with unilateral division
(right side). Both of these examinations suggest
that any individual variation away from the
conditional state of undivided supranasals in D.
tincta is extremely rare. Accordingly, we believe
that the reference by Storr et al. (1990) to
supranasal multiplication in D. tincta was in
specific reference to specimens from the Cape
Range Peninsula referred herein to D. tealei.
Delma inornata from eastern Australia appears
to be the only' species of Delma that exhibits
regular intraspecific variation (around 10%) in
having either 1 or 2 pairs of supranasals (Kluge
1974: 103). Among Western Australian Delma
the combination of 14 midbody scale rows and 2
pairs of supranasals is unique to D. tealei.

Delma tealei would probably receive an lUCN
conservation rating of 'Least Concern' on account
of the lack of evidence for any population decline
and most of its geographic range being protected
within the Cape Range National Park. However, in
many areas on the Cape Range Peninsula,
introduced Buffel Grass (Cenchnis ciliaris), has
virtually replaced the original ground cover (Aplin
1998) and there is an identified priority to monitor
and manage its spread (Keighery and Gibson 1993).
Particular attention should be given to the impact
on species such as Delma tealei that are probably
dependent on Triodia and other hummock grasses
for their survival.

Delma desmosa sp. nov.

Figures 10-12

Material examined

Holotypc

102657 in the Western Australian Museum, an
adult female collected on 10 October 1996 by S. van
Leeuwen at Site Cooma 4, Little Sandy Desert,
Western Australia (24°06T7"S 120°19'30"E). Liver
sample preserved in -75°C ultrafreeze at W.A.
Museum.

Para types

Sex indicated in brackets.

Western Australia: NTM 17987 (M) Sandfire
Flat (19°47'S 121°09'E); 45809-10 (both M) Wallal
Downs Homestead (19°47’S 120°38'E); 63313 (M)
Djaluwon Creek (20°20'S 127°26'E); 64001 (M)
Anketell Ridge (20°24'S 122°07'E); 64097 (F)
Staffords Bore (20°21'S 127°24’E); 64143 (F)
Breaden Pool (20°15'S 126°34'E); 64186 (F) 1 km S
Waddawalla Well (21°41'S 125°46'E); 75798 (F)
Dragon Tree Soak (19°39'S 123°23'E); 75830 (M)
Anna Plains Homestead (19°15'S 12r29’E); 87007
(F) Sandfire Roadhouse (19°46'S 121°06'E); 87353
(M) 3 km SE Wallal Downs Homestead (19°47'S
120°40"E); 88535-41 (M, F, F, F, M, F, F) 55 km S
Anna Plains Homestead (19°44'S 121°28'E); 94757,
94776-77 (F, M) 80 km S Telfer Mine (22°20T2''S
122'’02'26’'E); AM 100853, 101548 (both M) 6.6 km
N Sandfire Roadhouse (19°19'S 121°16’E); 102650
(M) Cooma 5, Little Sandy Desert (24°06'41"S
120°19T0"E); 108477 (M) is'km S Lake Hancock
(24°27'S 124°50'E); 114555 (F) Sandfire Roadhouse
(]9°46’S 121°06'E); 126496 (M), 126498 (M) Gibson
Desert Nature Reserve (24°43'S 124°52'E); 132802
Warri Airstrip (24°15'S 124°24'E); 139089 (M)
Mandora Station (19°45T6"S 12r26'59"E); 140442
(M) Yanneri Lake (24°27'08"S 120°29'02"E); 145073
(M) Officer Basin area (26°55'58"S 125°16'44'’E);
151252 (F) Townsend Ridges (26°20'25"S
126°56'26"E). Northern Territory: NTM 14901 (F)
12 km SW Sangsters Bore (20°52'S 130°16'E); NTM
15038 (M) Uluru National Park (25°2TS 131°0TE);
NTM R15138 (M), NTM 15144 (M), NTM 15146
(F), NTM 15151 (M) 12 km SW Sangsters Bore
(20°52'S 130°16’E); NTM 15230 (M) 17 km W
Sangsters Bore (20°48’S 130°14'E); NTM 15501 (M)
Uluru National Park (25°21'S 131°01'E); NTM
20250 (M) Sangsters Bore (20°51'09"S
130°23'09"E); NTM 26789 Henbury (24°34'S
133°30'E); NTM 32301 (M) 10 km WSW Sangsters
Bore (20°44’S 130°16'E); NTM 34489 (F) Ayers
Rock (25°20'S 131°0rE). South Australia; SAM
48671 (M) 9.3 km NNW Cheeseman Peak
(27°19'46"S 130°17'36"E); SAM 59561 (F) 3.3 km W
Mount Holder, Birksgate Range (27°08'43"S
129°39'51"E).

Pygopdids from NW Australia

289

Figure 10 Holotype (102657) of Delma desmosa, photographed in life (B. Maryan).

Diagnosis

A moderately small, stout species of Delma (SVL
up to 90mm) with modally 16 midbody scales, two
pairs of supranasals and distinctive dark dorsal
head markings present throughout life (any
ontogenetic fading is restricted to markings forward
of the eyes).

Description

Rostral with obtuse apex, penetrating between
rostral supranasals; two pairs of supranasals, caudal
pair much larger; rostral supranasals in moderate
contact with first supralabial and caudal
supranasals in point contact or only narrowly
separated from the nostril; postnasal single; loreals
3-9, subequal; suboculars 3-4; supraciliaries 5, fifth
much larger; supraoculars 2, second wider than
first; supralabials 5-6, third typically elongate and
positioned below eye (rarely, fourth is below eye);
posteriormost supralabial much smaller; infralabials
4 (rarely 5), third elongate; occipital present; upper
temporals 2.

General form of head and details of scalation
illustrated in Figure 11. Midbody scale rows 16;
transversely enlarged ventral scales 48—59; hindlimb
scales 8-10.

Morphological Variation: 75798 and 132802 have
third supralabial divided on left side; 94757 has this
scale divided on both sides. The location of
accessory supralabial suture (e.g., anterior, centre

or posterior) determines whether it is the third or
fourth that is positioned below the eye.

64186 has third infralabial divided on right side;
1 14555 has this scale divided on both sides.

64097 has second supraocular fused with fourth
supraciliary on both sides.

63313 has an upper loreal that is interposed
between first and second loreals and contacts
second supralabial on both sides.

88539 has supraoculars fused into one scale on
right side.

NTM 15146 has a small scale interposed between
caudal pair of supranasals and rostral supranasals
divided into two scales on right side.

AM 100583 has first supraciliary fused with an
upper loreal on right side.

AM 101548 has three small scales interposed
between third and fourth supraciliary on right side.

Colouration and patterning

In preservative, upper and lateral surface grey to
greyish brown merging with light brown on tail
(particularly regenerated portion). Irregular black
smudging on dorsal scales in some individuals (e.g.,
64186, 88537, 15038, 15151, 15501). Lateral scales on
forebody are plain. Lower surface immaculate
white.

Head of juveniles and adults typically with three
to four dark brown to black dorsal to lateral bands,
that narrow as they descend and terminate obtusely

290

B. Maryan, K.P. Aplin, M. Adams

Figure 11 Head scalation of Delma desmosa holotvpe
(102657) in lateral (top) and dorsal (bottom)
views.

on mental scale, on infralabials and behind the ear.
Dark head bands are most intense on immature
specimens (Figure 12A) but remain well-defined in
most adults (Figure 12C; ontogenetic fading is a
common occurrence in other Delma spp.).
Interspaces between dark bands light brown,
greyish to white and usually widest on back of head
through ear and neck. Supralabials and infralabials
whitish in between dark bands. Below the mouth,
bands are typically centred on suture between
mental and first infralabial, and on suture between
second and third infralabials; suture betw'een first
and second infralabials is invariably clear;
occasionally, the first dark head band completely
covers the first mfralabial but not its anterior suture
(e.g., 140442).

In adults that show signs of ontogenetic fading of
the head pattern (Figure 12B), the first band (on
snout and lores) and second band (over eyes)
typically become diffuse and merge to form a cream
to light browm colour forward of eyes. The broader
third and fourth dark bands (forward of and behind
ears, respectively) are persistent in adults and are
rarely diffuse or broken. In some individuals (e.g.,
15038, 34489, 88537) coalescent dark smudges
positioned transversely across forebody are
suggestive of a fifth dark band. Adult specimens of
D. desmosa with pronounced ontogenetic fading
come from localities spread across the western half
of the range of the species, including localities in
the Great Sandy Desert (e.g., 64097, 64143, 75830)
that are remote from the range of D. pax in the
Pilbara region. The wide geographic distribution of

these individuals make it unlikely that they are the
product of introgression between D. pax and D.
desmosa.

In life, some adults have pale orange-brown
interspaces between the dark bands; this pigment is
lost in preservative.

Details of Holotype

Snout-vent length (mm) 87; tail 163; loreals 8 on
left side, 9 on right; midbody scale rows 16; ventrals
59; hindlimb scales 9. Greyish brown upper and
lateral surface.

Indication of four dark brown head bands as
follows: first on lores is diffuse and terminates as
smudge on first infralabial; second terminates on
suture between second and third infralabial, leaving
preceding suture clear white; third and fourth
bands are dark brown and well-defined.

Etymology

From the Greek desmos, a chain, tie, or band, in
specific reference to the distinctive and persistent
dorsal head bands of this species.

Distribution and sympatry

Widespread in arid desert regions of western and
central Australia (Figure 8) extending west to the
vicinity of the 80 mile beach (Anna Plains and
Wallal Downs Stations), south to the Little Sandy
Desert and Officer Basin area and east through the
Great Sandy, Tanami and Great Victoria Deserts
into central Northern Territory (Sangsters Bore,
Uluru National Park and Henbury) and
northwestern South Australia (Cheeseman Peak).

The geographic distributions of D. desmosa and
D. pax appear to be allopatric (Figure 8). Currently
the two species are known to occur within 90 km of
each other (e.g., 102650, 102657 from Little Sandy
Desert versus 125452 from 30 km E Newman,
respectively). Specimens from these proximate
localities do not show' anv admixture of characters
as might be expected if significant levels of gene
flow' w'ere occurring across a contact zone or step
dine.

Five Delma species have geographic distributions
that overlap that of D. desmosa: D. borea, D. butleri,
D. haroldi, D. nasuta and D. tincta. Among these
species, the greatest morphological similarity occurs
between D. desmosa and D. borea (see below). The
distributions of D. borea and D. desmosa are
broadly overlapping in the south Kimberley,
southern Northern Territory and northwestern
South Australia (Figures 8, 15) but there are few
known instances of actual sympatry. Recent
collections by P. Kendrick of the Department of
Environment and Conservation have extended the
mainland W.A. range of D. borea south to the
vicinity of Mandora (e.g., 112725-26, 139058,

Pygopdids from NW Australia

291

(A) Immature
D. desmosa

(B) Adult
D. desmosa

(C) Adult
I), desmosa

(D) Immature
I), pax

Clear

Broad

Clear

\

Smudge

Smudge

Narrow

(G) Adult D. horea

Smudge

Figure 12 Head patterning of A, D. desmosa, immature from Townsend Ridges (151252), B, D. desmosa adult with
weak bands forward of eyes from Staffords Bore (64097), C, D. desmosa adult with strong bands from Ayers
Rock, Northern Territory (NTM 34489), D, D. pax immature from 82km E Port Hedland (140396), E, D. pax
sub-adult with fading bands from DeGrey River Station (132549) and arrows indicating clear or pigmented
suture and broad auricular band in D. desmosa, all shown in dorsal and lateral views and F, D. pax adult
from Potter Island (139353) showing complete lack of head patterning, G, D. borea adult with strong bands
from Mitchell Plateau (77201) and arrows indicating pigmented suture and narrow auricular band^ shown
in lateral views.

292

B. Maryan, K.P. Aplin, M. Adams

139062-63) where D. desmosa (e.g., 139089) is also
recorded. Currently this represents the only known
instance of sympatry for these two species in
Western Australia. However, the single specimen of
D. desmosa was obtained on the crest of a Triodia-
covered sandridge, while specimens of D. borea
came from the edge of a spring with Melaleuca
leucadendra on clayey soil. Records of D. desmosa
and D. borea from the arid southern Northern
Territory and northwestern South Australia also
tend to come from different localities. Systematic
faunal surveys have only recently recorded D. borea
from northwestern South Australia (Robinson ef al
2003) and the Uluru National Park (McAlpin 2005).
Generally, D. borea prefei's areas of stony or heavy
soils with Triodia and tussock grasses or savanna
woodland with grass/leaf litter (B. Maryan, personal
observation) and appears to not occupy Triodia on
sandplains (Kluge 1974; 82), the preferred habitat of
D. desmosa. The available information thus
suggests a degree of habitat partitioning between D.
desmosa and D. borea in areas where their ranges
interdigitate.

In Western Australia, other recorded instances of
sympatry involving D. desmosa include D. baroldi
and D. nasuta in the Little Sandy Desert and Central
Ranges (B. Maryan and P. Doughty, personal
obserx'ation) and D. butleri has been collected from
the Officer Basin area, where D. desmosa (e.g.,
145073) is also recorded. In the Northern Territory,
Reid et al. (1993) records both D. butleri and D.
nasuta from the same survey site as 'D. pax' (= D.
desmosa) in Uluru National Park.

Comparison with other species

Delma desmosa will be compared first with D.
pax and D. borea, two species with which it has
previously been confused, and then with other
regionally sympatric Delma spp.

Delma pax and D. desmosa are similar in body
proportions (Table 1) and agree in most details of
head and body scalation. The head pattern of
juveniles is also similar, consisting of four
'opalescent' black bands and pale interspaces
(Figure 12A, D), but the fate of this pattern is very
different. In D. desmosa the bands are retained
through to adult life (Figures 10, 12C) while in D.
pax they undergo pronounced ontogenetic fading
(Figure 12E) such that adults typically lack any head
pattern (Figures 12F, 14). Another difference in
head pattern concerns the ventral extent of the
anterior dark head bands; in D. pax the first (on
snout) and second (through eyes) bands typically
terminate on the infralabials, while in D. desmosa
they extend onto the mental scale and below the
infralabials, and are visible in ventral view. In
addition, the first dark head band on the snout in D.
pax is regularly weak or absent and usually distinct
in D. desmosa. Immature specimens of D. pax

typically have three dark brown smudges on the
lower lip, situated over the sutures between the
mental and anterior two infralabial scales (Figure
12D-E). In D. desmosa the suture between the first
and second infralabial typically is unpigmented
(Figure 12A-C).

On meristic and mensural data, D. desmosa
averages slightly smaller than D. pax for most
measurements (Table 1). This contrast is
particularly striking for males, reflecting more
pronounced sexual dimorphism in D. pax than D.
desmosa (Table 1). In both sexes the mean ventral
scale count is significantly lower in D. desmosa than
in D. pax (Table 4).

Delma borea is similar to D. desmosa in overall
body size but has shorter hindlimb flaps and a
significantly higher average number of ventral
scales in both sexes (Tables 1, 4). The two species
are readily distinguished by head prattern. Delma
borea has a pale inter-band on the back of head
which narrows dorsally but broadens or 'forks' at
the ear aperture (Figure 12G). In D. desmosa the
auricular inter-band typiically is of even width or
only slightly broader laterally (Figure 12A-C). Most
D. borea also have some variegated ventrolateral
scales OTT the forebody. These scales are unpatterned
in D. desmosa. Delma borea typically have the
fourth supralabial prositioned below the eye
(typically the third in D. desmosa).

Delma borea and D. pax both have three dark
brown smudges situated betw'een the nostril and
the eye, and positioned over the mental and
infralabial scale sutures (Figure 12D-E, G). These
smudges are absent only in some adults devoid of
any head pattern (Figure 12F). In contrast, the
suture between the first and second infralabial
typically is clear in D. desmosa; in occasional
specimens (e.g., 140442) the first head band
completely covers the upper portion of fhe first
infralabial scale.

Apart from D. borea, the only Delma species with
distributions overlapping that of D. desmosa are D.
butleri, D. baroldi, D. nasuta and D. tincta. D.
desmosa is easily distinguished from the first three
by having dark dorsal head markings and from D.
tincta by the presence of two pairs of supranasal
scales (one pair in D. tincta), a higher midbody scale
row count (modally 16 versus 14) and unpatterned
ventrolateral scales on the forebody.

Habitat

The holotype was raked (using a 3-prong
cultivator) from Eucalyptus chippendalei leaf litter
in dune swale on red sand with groundcover of
Thryptomene and Triodia (Figure 13). Notes
accompanying some Western Australian paratypes
include "raked from dead Triodia clumps and
shrubs on crest of sandridges" (e.g., 88535-41); "pit-
traprped on claypan with Acacia over mixed grasses

Pygopdids from NW Australia

293

Figure 13 Triodia sandplain in swale with red sand dune covered with Eucalyptus chippendalei. Acacia and Triodia
at the Little Sandy Desert WA, the habitat for Delma desmosa (B. Maryan).

and samphire" (e.g., 75798); "along minor drainage
lines with fringing Eucalyptus^' (e.g., 63313, 94776-
77) and "active at night on road in sandplain with
sparse Acacia over Triodia" (e.g., 114555). Habitat
details for paratypes from Northern Territory and
northwestern South Australia are "pit-trapped in
Triodia grassland" (e.g., 14901, 15038, 15138, 15144,
15146, 15151, 15230, 15501, 20250); "mulga
woodland over bluebush and tussock grasses on
sandplain" (e.g., 48671) and "under loose stones on
rocky hillside" (e.g., 59561).

Remarks

The specific characteristics of D. desmosa have
created prior confusion between D. borea and D.
pax. Shea (1991) drew attention to this problem by
mentioning a specimen from Ayers Rock (NTM
1319; renumbered 34489) that .shares D. borea and
D. pax scalation characters. This specimen is herein
referred to D. desmosa (Figure 12C). Systematic
faunal surveys by Reid et at. (1993) and McAlpin
(2005) also mentioned problems with identification
and variously assigning a Delma sp. to D. borea or
D. tincta in previous reports, but identified by them
as 'D. pax'. Their accounts of 'D. pax' from Uluru
National Park are most likely based on individuals
of D. desmosa.

Ehmann (1992: 94) and Reid et al. (1993: 49)
illustrate D. desmosa from the Great Sandy Desert,
Western Australia and from Uluru National Park,
Northern Territory respectively; in both
publications the specimens are identified as D. pax.

Delma pax Kluge, 1974

Figure 14

Delma pax Kluge (1974: 113-117). 14804 in the
Western Australian Museum, an adult female
collected on 21 May 1961 by G.M. Storr at Jones
River, Western Australia (20°58'S, 117°23'E).

Revised diagnosis

A moderately small, stout species of Delma (SVL
up to 98mm) with modally 16 midbody scales, two
pairs of supranasals, and a plain adult colouration
due to pronounced ontogenetic fading of dorsal
head markings (markings prominent in juveniles).

Description

Rostral with obtuse apex, penetrating between
rostral supranasals; two pairs of supranasals, caudal
pair much larger; rostral supranasals in moderate
contact with first supralabial and caudal
supranasals in point contact or only narrowly
separated from the nostril; postnasal single; loreals
4-10, subequal; suboculars 2-4; supraciliaries 5
(rarely 4), fifth much larger; supraoculars 2, second
wider than first; supralabials 5-6, third typically
elongate and positioned below eye (occasionally,
fourth is below eye), posteriormost supralabial
much smaller; infralabials 4, third elongate;
occipital present; upper temporals 2. Midbody scale
rows usually 16 [14 in 12.4 % of specimens; 18 in
one specimen (119045; from Port Hedland)];
specimens with 14 midbody scale rows are Bohemia

294

B. Maryan, K.P. Aplin, M. Adams

Figure 14 Adult Delma pax from Meentheena, photographed in life (B. Maryan).

scattered throughout the range of D. pax and often
come from the same localities as specimens with 16
midbody scale rows. Storr et al. (1990) state that D.
pax has 'rarely 17' midbody scale rows but without
citing specimen details. Transversely enlarged
ventral scales 50-60; hindlimb scales 8-10.

Morphological Variation: Kluge (1974) examined
16 D. pax and recorded only a single specimen with
the fourth supralabial below the eye on one side
only (SAM 3445 from Pilgangoora Well). Out-
examination of a further 97 specimens found this
condition in unilateral or bilateral states in 20% of
individuals, as follows: fourth supralabial below
eye on both sides in 73146, 102137, 113387, 119045-
046, 129930, 132657, 135919, 139294, 140396, 145680
and 146649; fourth supralabial below eye on right
side only in 127829, 135320, 139353 and^ 146591; on
left side only in 9943, 73841 and 129658.

Other variants are;

9946 has fusion of second and third supraciliaries
(total 4) on both sides; 132606 has same condition
on right side only.

81390 has rostral contacting caudal supranasals
and thus separating rostral supranasals; 129930 has
small scale interposed between rostral supranasals.

1 19045 has two upper loreal scales on right side.

145748 has rostral and caudal supranasals fused
on right side.

Colouration and patterning

In preservative, upper and lateral surface brown
to reddish brown merging into pale grey on lower

lateral surfaces. Lateral scales on forebody are plain.
Lower surface white and unpatterned.

Head of juveniles typically with strong pattern of
transverse bands (Figure 12D). Intensity of bands
diminishes with increasing body size (= age) such
that head and neck of large adults are typically light
to reddish brown and unpatterned (Figures 12F,
14). Where traces of pattern are retained (e.g.,
129658, 132548, 139170, 140021), this consists of
very diffuse pale brown spaces ('ghosting' of bands)
between the ears and behind the eyes.

Head pattern of immature specimens consists of
three to four brown to blackish brown bands that
narrow as they descend and terminate obtusely on
the infralabials and behind the ear. The first band
(on snout) is v'ariably developed and may be absent,
even on juveniles. The bands crossing the back of
the head and the neck are broader and more distinct
and the pale interbands are typically of even width,
without any lateral widening. The interband spaces
are pale reddish brown in life but this typically
fades to a lighter brown or a greyish-white (e.g.,
140396) in preservative. A narrow and faint pale
band is usually present on the neck behind the
posteriormost dark band; this is, occasionally
followed on the side of the neck by a narrow dark
band (e.g., 140396).

Supralabials and infralabials of juveniles whitish
in between dark bands. Dark bands extend below
mouth and terminate on suture lines between
mental and first infralabial scales, between second
and third infralabials, and between third and fourth

Pygopdids from NW Australia

295

infralabials. These suture line smudges undergo
ontogenetic fading in concert with the general head
pattern.

Distribution and sympatry

Widespread throughout Pilbara region of Western
Australia (Figure 8) with southerly extension to
northern Gascoyne at Turee Creek, extending north
to DeGrey River Station, east to Carawine Gorge
and 30km east of Newman and southwest to Mount
Minnie and Cane River Stations. Also occurs on
Potter Island off Pilbara coast. Endemic to Western
Australia.

As noted earlier, D. pax appears to be allopatric
with respect to the closely related D. desmosa
(Figure 8). However, D. pax is regionally sympatric
in the Pilbara region with D. butleri, D. haroldi, D.
elegans Kluge, 1974, D. nasuta and D. tincta.
Recorded instances of sympatry involving D. pax
include D. haroldi, D. nasuta and D. tincta from
multiple localities throughout the Pilbara region.
For instance, in the vicinity of Port Hedland, D. pax,
D. haroldi and D. tincta have all been observed
crossing the same section of sealed road at night.
Delma pax and D. nasuta were observed together in
the same context in the vicinity of Newman (B.
Maryan and B. Bush, personal observation). Delma
pax also has been recorded with D. elegans from
several localities including Meentheena and
Pannawonica.

Comparison with other species

Delma pax will be compared first with D. borea,
and then with other regionally sympatric Delma
spp. For comparison with D. desmosa see the
preceding account.

Both sexes of D. pax average larger than D. borea
in almost all linear measurements (Tables 1, 4),
while average ventral scale counts are significantly
higher in female D. pax than female D. borea but
not in males. Body pattern in D. pax features plain
ventrolateral scales on the forebody, whereas D.
borea typically has variegated scales in this area.
Specimens of D. pax that retain traces of the
immature head pattern (Figure 12E) are
distinguished from D. borea by having a band on
back of head that is slightly broader and of more
even width (in D. borea the auricular band is
narrower mid-dorsally but broadens laterally, often
'forking' at the ear aperture; Figure 12G). Delma
pax typically have the third supralabial positioned
below the eye (typically the fourth in D. borea).

As indicated above, five other Delma species
occur in regional sympatry with D. pax (D. butleri,
D. haroldi, D. elegans, D. nasuta and D. tincta). The
relatively inornate D. butleri bears a superficial
resemblance to adult D. pax but differs in having
more complex patterning on the lips, side of head
and neck (variably marked with brown and white

spots, blotches or v'ertical streaks). Delma haroldi is
more distinct with narrow wavy pale bands (but no
dark bands) across the head and neck. Delma
elegans has five or six dark head bands that descend
obliquely forward and also has higher midbody
scale row counts (modally 18 versus 16). Delma
nasuta has a longer, sharper snout with a spotted or
reticulated pattern on the dorsal and ventral
surfaces. Finally, D. tincta has one pair of
supranasals (versus two in D. pax), lower midbody
scale row counts (modally 14 versus 16), fewer
loreal scales on average (Table 1) and variegated
scales on the lateral forebody (versus plain scales in
D. pax). Delma tincta generally averages smaller
than D. pax in linear dimensions (Table 1).

Habitat

Delma pax occupies a variety of habitats in the
Pilbara region including sandy riverside flats and
stony slopes with heavy soils. It is most frequently
obtained from Triodia clumps but also shelters in
flood debris along dry watercourses. The species is
often observed at night active on sealed roads (B.
Maryan, personal observation).

Remarks

Kluge (1974: 116-117) illustrates the head region
of both immature and adult D. pax. Adult D. pax is
illustrated by Wilson and Knowles (1988: 249), Storr
et al. (1990: Plate 17.3), Cogger (2000: 290) and
Wilson and Swan (2003: 117). Wilson and Swan
(2003: 111) illustrate an immature D. pax (140396)
from the Port Hedland district, mislabeled as D.
borea.

Remarks on the distribution of Delma borea

The taxonomic changes proposed above also help
clarify the species limits and geographic
distribution of D. borea. As delimited here, D. borea
in Western Australia is a moderately small, stout
species of Delma (SVL up to 98mm) with modally
16 (rarely 17) midbody scales, two pairs of
supranasals, the fourth supralabial scale positioned
below the eye (unilaterally third supralabial below
eye in two out of 114 specimens), and variegated
ventrolateral scales on forebody. Juveniles possess
well-defined dark head bands. Adults undergo
ontogenetic fading to varying degree and usually
possess indistinct pale brown bands on the head
and neck. These morphological characteristics are
consistent with previous accounts of D. borea in the
more easterly parts of its range (Kluge 1974).

Figure 15 shows the distribution of D. borea in
Australia, based on our reassessment of specimens
in the collection of the Western Australian,
Northern Territory and South Australian Museums!

In Western Australia, this species ranges from the
Kimberley southwest to Mandora inland of the 80
mile beach, and south to the Edgar Ranges, 25 km E

296

B. Maryan, K.P. Aplin, M. Adams

Downs and Denison Range, it is present on
numerous islands off the Kimberley coast
(Troughton, Naturalist, Coronation, Heywood,
Sunday, Augustus, King Hall, Cockatoo and
Koolan) and at least three islands off the Pilbara
coast (Barrow, Hermite and Rosemary). It also
occurs in the Northern Territory, western
Queensland and northwestern South Australia
(Kluge 1974; Shea 1987, 1991; Ingram and Raven
1991; Ehmann 2005). In the Northern Territory, D.
borea is most common in the Top End, with the
southernmost records at Wave Hill, Helen Springs
and 50 km S MacArthur River camp (Shea 1991); it
appears to be sparsely distributed south of the 20"
parallel that includes its occurrence in northwestern
South Australia on Aboriginal Lands (Ehmann
2005). Kluge's (1974: Figure 47) map for this species
shows a record in northwestern South Australia but
without an equivalent specimen listed under
Paratypes; we assume that the map is in error.

Three specimens included by Kluge (1974: 192)
within D. borea warrant special mention. Specimen
25201 from 32 km E Jiggalong was also mapped as
D. borea by Storr et al. (1990) and further cited by

Shea (1991) as representing this taxon. This
specimen is confirmed here as a member of the D.
tincta group due to the presence of a enlarged
upper temporal scale bordering each parietal. It
differs from each of D. borea and from D. desmosa
in being more slender bodied, and differs from D.
desmosa in having more pronounced variegation of
the ventrolateral scales on the forebody. In both of
these respects, the specimen resembles D. tincta.
However, it differs from D. tincta in having paired
supranasals and 16 midbody scale rows, and in
these respects, more closely resembles D. borea and
D. desmosa. This specimen might represent an
outlier of D. borea, a somewhat aberrant D.
desmosa, or another, as yet unrecognized taxon.

Specimen SAM 5058 referred to D. borea also by
Shea (1991) from the Warburton Range is similar in
most respects to specimen 25201. The colour pattern
is similar to both D. borea and D. tincta, most
notably in the presence of variegated ventrolateral
scales on the forebody. The specimen has two pairs
of supranasal scales and the fourth supralabial
positioned bilaterally below the eye, both features
shared with typical D. borea. However, it is slender

Figure 15 Map of Australia showing distribution of D. borea (shading) and location of specimens examined in this
study (triangles). Arrows indicate outlier populations at Mandora, western Pilbara islands and northwestern
South Australia and the two specimens (circles with cross) from 32 km E Jiggalong and Warburton Range
that are regarded as D. sp. incertae sedis piending further field survey and analysis.

Pygopdids from NW Australia

297

bodied and has 14 midbody scales, more typical of
D. tincta. The presence of variegated ventrolateral
scales on the forebody and the absence of strong
head patterning distinguish the specimen from
regionally sympatric D. desmosa. The desert
regions are poorly sampled for herpetofauna (How
and Cowan 2006) and until further material is
available, we recommend that the Warburton Range
and Jiggalong populations be treated as Delma sp.
incertae sedis.

Specimen SAM 4475 from Tambrey is listed twice
by Kluge (1974; 192), once as a paratype of D. borea
and again as a paratype of D. elegans. Advice from
the South Australian Museum indicates that there
have been no changes made in relation to this
registration number which is currently attached to a
paratype of D. elegans (M. Hutchinson, persona]
communication); the duplicate listing under D.
borea is assumed to be an editorial error.

DISCUSSION

Taxonomic diversity of the Delma tincta group
The present study has identified two additional
members of the Delma tincta group as defined by
Shea (1991). Delma desmosa from the arid sand
deserts of western and central Australia is a close
relative of D. pax which is now recognized as
endemic to the Pilbara region. The ranges of D. pax
and D. desmosa appear to be allopatric or
parapatric (current records suggest a gap of no
more than 90 km between populations). These
species are weakly differentiated genetically and
they differ morphologically mainly in features of
colouration (most notably the degree of ontogenetic
fading of the juvenile head markings), and in the
degree to which sexual dimorphism is expressed.
However, the fact that an abrupt boundary is
maintained between the two taxa over a very large
distance around the periphery of the Pilbara
uplands indicates that they represent discrete
evolutionary lineages, each with its own set of
ecophysiological requirements, and thus warrant
specific recognition. Gene flow between the
populations, if it occurs at all, is clearly limited and
insufficient to influence the genetic or
morphological characteristics of the spatially
adjacent populations. Nevertheless, contact zones
between D. pax and D. desmosa should be sought
in which to investigate the nature and extent of
genetic interactions between these taxa.
Furthermore, ecological and behavioural
comparisons of these closely related species might
yield valuable insights into the adaptive
significance of head patterning in Delma.

Delma tealei, an endemic of the Cape Range
Peninsula, is most similar morphologically to D.
borea and to a lesser extent, to D. tincta, all three

taxa sharing the unusual characteristic of variegated
scales on the ventrolateral forebody. The allozyme
data suggest a possible sibling relationship between
D. borea and D. tealei, and a more remote
association with D. tincta. However, the level of
divergence is much greater than that observed
between D. pax and D. desmosa, and identifies D.
tealei as a well established lineage within the D.
tincta group.

Biogeography of the Delma tincta group

The D. tincta group as a whole has a 'Torresian'
distribution (sensu Cogger and Heatwole 1981).
This appears to be unique within the genus Delma,
since other species groups recognized on
morphological (Kluge 1974; Shea 1991) or molecular
criteria (Jennings et al. 2003) hav'e geographic
distributions that either range across southern
Australia (D. australis + D. torquata Kluge, 1974; D.
fraseri + D. grayii; D. petersoni + D. inomata), are
confined to eastern Australia (D. impar + D. molleri
+ D. mitella Shea, 1987), or are centred on the arid
inland region (D. butleri/D. haroldi + D. nasuta).
With a total of five species, the D. tincta group is
also the most speciose of the major intrageneric
lineages identified to date. However, as noted
above, this may reflect a lack of complete taxonomic
discrimination in some other groups, most notably
in the D. butleri/D. haroldi complex and in the D.
australis group (Aplin and Smith 2001).

Within the Delma tincta group, the widely
distributed D. tincta appears to be broadly
sympatric with each of D. borea, D. pax and the two
additional members described in this paper.
Whether this entails instances of true syntopy is not
known. However, in our view the likelihood of
syntopy is enhanced by the unusual morphological
characteristics of D. tincta within this group, in
particular its relatively small adult size and slender
build (as reflected by a reduced number of midbody
scale rows). Delma tealei, which is at least
regionally sympatric with D. tincta on the Cape
Range Peninsula, shares several of these
characteristics and it would be of great interest to
know more about the ecological interaction between
these species.

Delma pax and D. desmosa have allopatric or
parapatric distributions, the former confined to the
Pilbara region, and the latter found in the
surrounding sandy deserts and extending east into
the central Australian deserts. Along its northern
margin, the range of D. desmosa appears to
interdigitate with that of D. borea but with no
records of syntopy. These taxa differ in relatively
subtle aspects of body colouration, meristics and in
relative hindlimb flap length, and they may be
weakly differentiated ecologically and subjected to
mutual competitive exclusion. Somewhat
surprisingly, given the high degree of

298

B. Maryan, K.P. Aplin, M. Adams

morphological similarity, D. desmosa and D. borea
are well-differentiated genetically.

The ranges of D. borea and D. pax approach
regional sympatry in northwestern Australia.
Delma pax is restricted to the mainland Pilbara
region where it occupies a variety of local habitat
types being only recently recorded on Potter
Island off the Pilbara coast. Delma borea, in
contrast, is present on several of the western
Pilbara islands. On Barrow Island it is found
together with D. nasuta and D. tincta and on
Hermite Island it is recorded with D. nasuta
(Burbidge et al. 2000).

The disjunct occurrence of D. borea in
northwestern Australia begs explanation. The flora
and fauna of Barrow and Hermite Islands off the
Pilbara coast are closely allied with those of the
Cape Range Peninsula (Keighery and Gibson 1993;
Baynes and Jones 1993; Kendrick 1993), reflecting
not only the geological similarity between the two
areas (both are anticlinal structures comprised of
Miocene limestones) but also that during periods of
lowered sea level through the late Pliocene and
Pleistocene, both formed rocky plateaux on a sandy,
emergent continental shelf. Despite this overall
similarity, the Pilbara islands host a number of
'northern monsoonal' faunal elements that are
absent from the Cape Range Peninsula. One of these
is D. borea, perhaps replaced on the peninsula by
the morphologically similar D. tealei. Others
include a murid rodent Pseudomys nanus (Gould,
1858) and a skink Ctenotus angusticeps Storr, 1988.
Another 'northern monsoonal' mammal species, the
Northern Nailtail Wallaby Onychogalea unguifera
(Gould, 1841), is represented in an early Holocene
subfossil assemblage from the Montebello islands
(Veth 1993). These taxa are typically associated
either with grassland communities on coastal plains
(F. nanus and O. unguifera) or with coastal
samphire communities (C. angusticeps), vegetation
types that were most likely broadly continuous
along the emergent northwestern continental shelf
during periods of lower sea level. The present
distribution of D. borea suggests that it, too, was a
member of this now emergent continental shelf
community that survives in relictual form only on
the Pilbara islands.

Molecular clock estimates suggest a staggered
origin of the major species group lineages within
Delma during the early to mid-Miocene, around 20-
28 million years ago (Jennings et al. 2003). The
Delma tincta group probably arose during the latter
part of this radiation. Divergence of the modern
species lineages (D. pax, D. tincta, D. borea)
probably occurred during the late Miocene (ca. 8-9
million years ago). Other speciation events in
Delma typically are of similar or even greater
antiquity, if the molecular dock estimates are
accepted as valid (Jennings et al. 2003).

Delma tealei shows a similar level of genetic
differentiation to the other previously recognized
members of the D. tincta group and thus may also
have originated during the late Miocene. This
corresponds closely to the time of emergence of the
Cape Range (Wyrwoll et al. 1993) and it is tempting
to link the two events via a dispersal or vicariance
event. However, the probable great antiquity of the
D. tealei lineage (as indicated by its genetic
distinctiveness) leaves open the possibility that D.
tealei was formerly more widespread under the
very different bioclimatic regime of the late Tertiary
period and is relictual on Cape Range Peninsula.
Delma desmosa, in contrast, is genetically close to
its sibling D. pa.vand these taxa probably diverged
during the late Pliocene or Pleistocene (i.e., within
the last 2-3 million years). In broad terms, this
corresponds to the period of progressive
desertification of Australia (White 1994; Fujioka et
al. 2005) and it seems likely that this later period of
diversification within the D. tincta group occurred
in response to the emergence of new kinds of arid
zone habitats.

Species identification in Delma and a new
dichotomous key

As noted in the Introduction, this study was
initiated in response to seemingly anomalous
identifications of Pilbara Delma specimens obtained
through application of the key published in Storr et
al. (1990: 114). In most cases, this confusion
involved individuals of D. pax or D. desmosa in
which the third supralabial scale is divided
unilaterally or bilaterally, thereby leading to
ambiguity at the second key couplet in which the
relationship of supralabial scales to the eye is used.
Similar difficulties were also encountered as a
consequence of intraspecific variation in other
'diagnostic' characters within the genus Delma,
including the condition of the supranasal scales, the
number of midbody scale rows and aspects of head
patterning. Fundamentally, these difficulties reflect
the fact that the genus Delma is speciose yet
morphologically conservative. Moreover, problems
of identification are compounded by marked
ontogenetic transformations in head pattern that
occur in some species of Delma but not in others.

In conclusion, we offer a revised dichotomous key
to the Delma species of Western Australia. Given the
difficulties in accurate identification of this group,
we suggest that the key be used only as a first step
towards taxonomic identification of adult specimens,
which should then be confirmed by direct
comparison with voucher material or through
genetic analysis. Moreover, if possible, we
recommend that the following conditions be met
prior to application of the key: 1) that the stage of
sexual maturity of individual animals is determined;
2) that 'typical' scalation characters are determined

Pygopdids from NW Australia

299

through examination of locally obtained series rather
than individual specimens; and 3) that close attention
is paid not only to the small number of standard
diagnostic features employed in the key but also to
subtleties of head and body scale patterning within
regional Delma assemblages.

Key to Western Australian Delma

1 . Typically one pair of supranasals 2

Typically two pairs of supranasals 3

2. Smaller (SVL up to 57mm); typically 18

midbody scale rows; side of neck and
forebody usually finely barred with black; no

broad dark bands across head and neck

australis

Longer (SVL up to 92mm); typically 14
midbody scale rows; side of forebody usually
with variably coloured scales; broad dark
bands across head and neck (often fade with
age) tincta

3. Typically third supralabial below the eye;

typically 14 midbody scale rows; Cape Range

Peninsula tealei

Either third or fourth supralabial below the
eye; between 16-20 midbody scale rows 4

4. Typically third (occasionally fourth)

supralabial below the eye 5

Typically fourth supralabial below the eye .... 6

5. Dark bands across head and neck persistent at

all ages; two dark smudges on infralabials

below lores; deserts desmosa

Dark bands across head and neck absent in
adults; three dark smudges on infralabials
below lores; Pilbara pax

6. Typically 16 midbody scale rows 7

Typically 18 midbody scale rows 8

7. Throat white; Kimberley and some Pilbara

islands borea

Throat with fine dark variegations; southwest
of WA fraseri

8. Dark bands across head and neck descend

obliquely forwards, terminating acutely;

Pilbara elegans

Dark bands run straight across head and neck
and meet to form black bands across the chin
and throat; southern Great Victoria Desert ...
petersoni

9. Narrow, wavy pale bands across head and neck

haroldi

No pale bands across head and neck 10

10. Side of forebody with numerous pale vertical

streaks or bars; lower surface yellow ... grayii

Side of forebody without numerous pale
vertical streaks or bars; lower surface white
with or without dark markings 11

11. Snout long; dorsal scales spotted and flecked
with dark brown; ventral scales usually dark-

edged nasuta

Snout short; dorsal scales finely dark-edged;
ventral scales without dark edges butleri

ACKNOWLEDGEMENTS

This study would not have been possible without
financial support for the genetic examinations from
Methanex Australia Pty. Ltd. R. How, P. Doughty
and C. Stevenson of the Western Australian
Museum provided support and assistance during
the study. We are grateful to R. O'Shea formerly of
the Western Australian Museum for doing the head
drawings and taking photographs of preserved
specimens. L. Schmitt of the University of Western
Australia kindly provided statistical advice. P.
Horner and G. Dally of the Northern Territory
Museum, M. Hutchinson and C. Secombe of the
South Australian Museum and R. Sadlier of the
Australian Museum kindly allowed access to
specimens. G. Shea, M. Hutchinson and P. Doughty
offered insightful comments of the manuscript.

Thanks also to the many field herpetologists in
Western Australia who have made substantial
contributions to the Delma collection of the Western
Australian Museum, most notably D. Algaba, B.
Bush, R. Browne-Cooper, G. Harold, P. Kendrick,
D. Pearson, D. Robinson and R. Teale.

REFERENCES

Adams, M., Baverstock, P.R. and Reardon, T. (1987).
Electrophoretic resolution of species boundaries in
Australian Microchiroptera. I. Eptesicus (Chiroptera,
Vespertilionidae). Australian Journal of Biological
Sciences 40: 143-162.

Aplin, K.P. (1998). Three new blindsnakes (Squamata:
Typhlopidae) from northwestern Australia. Records
of the Western Australian Museum 19; 1-12.

Aplin, K.P. and Smith, L.A. (2001). Checklist of the frogs
and reptiles of Western Australia. Records of the
Western Australian Museum Supplement No. 63: 51-
74.

Baynes, A. and Jones, B. (1993). The mammals of Cape
Range peninsula, north-western Australia. Pp. 207-
225 in W.F. Humphreys (ed.) The Biogeography of
Cape Range, Western Australia. Records of the
Western Australian Museum Supplement No. 45.
Burbidge, A.A., Blyth, J.D., Fuller, P.J., Kendrick, P.G.,
Stanley, F.J. and Smith, L.A. (2000). The terrestrial
vertebrate fauna of the Montebello Islands, Western
Australia. CALM Science 3: 95-107.

Cogger, H.G. and Heatwole, H. (1981). The Australian
reptiles: Origins, biogeography, distribution patterns

300

B. Maryan, K.P. Aplin, M. Adams

and island evolution, Pp 1331-1373 in Ecological
biogeography of Australia (A. Keast ed.) Dr W. Junk,
The Hague.

Cogger, H.G. (2000). Reptiles and Amphibians of
Australia. Reed New Holland, Frenchs Forest, NSW.

Ehmann, H. (1992). Encyclopedia of Australian Animals.
Reptiles. Angus & Robertson, Sydney.

Ehmann, H. (2005). South Australian Rangelands and
Aboriginal Lands Wildlife Management Manual.
Department of Water, Land and Biodiversity
Conservation, South Australia.

Felsenstein, J. (1993). PHYLIP (Phylogeny Inference
Package. Version 3.5c). Distributed by the author.
University of Washington, Seattle.

Fujioka, T., Chappell, J. Honda, M., Yatsevich, I., Fifield,
K. and FAbel, D. (2005). Global cooling initiated stony
deserts in central Australia 2-4 Ma, dated by
cosmogenic 21Ne-10Be. Geology 33: 993-996.

How, R.A. and Cowan, M.A. (2006). Collections in space
and time: geographical patterning of native frogs,
mammals and reptiles through a continental gradient.
Pacific Conservation Biology 12: 111-133.

Ingram, G.J. and Raven, R.J. (1991) eds. An atlas of
Queensland's Frogs, Reptiles, Birds and Mammals,
Queensland Museum.

Jennings, W.B., Pianka, E.R. and Donnellan, S. (2003).
Systematics of the Lizard Family Pygopodidae with
Implications for the Diversification of Australian
Temperate Biotas. Systematic Biology 52: 757-780.

Keighery, G. and Gibson, N. (1993). Biogeography and
composition of the flora of the Cape Range peninsula.
Western Australia. Pp. 51-85 in W.F. Humphreys
(ed.) The Biogeography of Cape Range, Western
Australia, Records of the Western Australian
Museum Supplement No. 45.

Kendrick, P.G. (1993). Biogeography of the vertebrates of
the Cape Range peninsula. Western Australia. Pp.
193-206 in W.F. Humphreys (ed.) The Biogeography
of Cape Range, Western Australia. Records of the
Western Australian Museum Supplement No. 45.

Kluge, A.G. (1974). A taxonomic revision of the lizard
family Pygopodidae. Miscellaneous Publications of
the Museum of Zoology, University of Michigan No.
147: 1-221.

Kluge, A.G. (1976). Phylogenetic relationships in the
lizard family Pygopodidae: An evaluation of theory,
methods and data. Miscellaneous Publications of the
Museum of Zoology, University of Michigan No. 152:
1-72.

McAlpin, S. (2005). Uluru - Tjuta National Park
Vertebrate Fauna Resurvey; 2004. Consultancy Report
to Parks Australia North. The author, Lismore.

Page, R.D.M. (1996). TREEVIEW: An application to
display phylogenetic trees on personal computers.
Computer Applications in the Biosciences 12: 357-
358.

Reid, J.R.W., Kerle, J.A. and Morton, S.R. (1993). Uluru
Fauna. The Distribution and Abundance of Vertebrate
Fauna of Uluru (Ayers Rock-Mount Olga) National
Park, N.T. Australian National Parks and Wildlife
Service series - Kowari, Canberra.

Richardson, B.J., Baverstock, P.R. and Adams, M. (1986).
Allozyme Electrophoresis: A Handbook for Animal
Systematics and Population Studios. Academic Press,
Sydney.

Robinson, A.C., Copley, P.B., Canty, P.D., Baker, L.M.
and Nesbitt, B.J. (2003). A biological survey of the
Anangu Pittjantjatjara Lands South Australia. South
Australian Department for Environment and
Heritage, Adelaide.

Shea, G.M. (1987). Two new Species of Delma (Lacertilia:
Pygopodidae) from northeastern Queensland and a
Note on the Status of the Genus Aclys. Proceedings of
the Linnean Society of New South Wales 109: 203-
212 .

Shea, G.M. (1991). Revisionary notes on the genus Delma
(Squamata: Pygopodidae) in South Australia and the
Northern Territory. Records of the South Australian
Museum 25: 71-90.

Smith, L.A. and Adams, M. (2006). Revision of the Lerista
muelleri species-group (Lacertilia: Scincidae) in
Western Australia, with a redescription of Lerista
muelleri (Fischer, 1881) and the description of nine
new species. Records of the Western Australian
Museum (in press).

Storr, G.M. (1987). Three new legless lizards
(Pygopodidae) from Western Australia. Records of
the Western Australian Museum 13: 345-355.

Storr, G.M. and Hanlon, T.M.S. (1980). Herpetofauna of
the Exmouth Region, Western Australia. Records of
the Western Australian Museum 8; 423-439.

Storr, G.M., Smith, L.A. and Johnstone, R.E. (1990).
Lizards of Western Australia. Ill Geckos & Pygopods.
Western Australian Museum.

Storr, G.M,, Smith, L.A. and Johnstone, R.E. (2002).
Snakes of Western Australia. Revised Edition.
Western Australian Museum.

Veth, P. (1993). The Aboriginal occupation of the
Montebello Islands, north-west Australia. Australian
Aboriginal Studies 24: 81-89.

White, M.E. (1994). After the greening, the browning of
Australia. Kangaroo Press, Kenthurst, Australia.

Wilson, S.K. and Knowles, D.G. (1988). Australia's
Reptiles. A photographic reference to the Terrestrial
Reptiles of Australia. Collins, Sydney.

Wilson, S.K. and Swan G. (2003). A Complete Guide to
Reptiles of Australia. Reed New Holland, Frenchs
Forest, NSW.

Wyrwoll, K.-H., Kendrick, G.W. and Long, J.A. (1993).
The geomorphology and Late Cenozoic
geomorphological evolution of the Cape Range -
Exmouth Gulf region. In W.F. Humphreys (ed.) The
Biogeography of Cape Range, Western Australia.
Records of the Western Australian Museum
Supplement No. 45: 1-23.

Manuscript received 9 November 2005; accepted 10 November
2006

Pygopdids from NW Australia

301

APPENDIX 1

Lists of Specimens examined

Legend for Museum registration numbers: WAM = Western Australian Museum; NTM = Northern Territory Museum;
SAM = South Australian Museum.

Delma borea Kluge 1974

SAM; 42018-20, 42046, 42189 2 km W Hanging Knoll
(26°19'23"S 130‘’23'36"E).

NTM: 1317 Barrow Creek (21°31'S 133°53'E); 1574
Tanami Bore (19°58'S 129°40'E); 5371 Barrow Creek
(21°31'S 133“53'E); 5824-25, 6516-6517, 6610 Wave Hill
(17°29'S 130°57'E); 6594 70 km N Top Springs (16°00'S
131°56'E); 9133 Keep River National Park (15°45'S
129°05'E); 12727 George Gill Ranges (24°19’S 131°35’E);
13047 Ord River (16°10'S 128°44'E); 15414 Chewings
Range (23°40'S 132°54'E); 16473 Sambo Bore, Wave Hill
Station (18°53’S 130°40'E); 18042 Alyawarre Desert Area
(20°44'S 135°50'E); 20644 Finke Gorge National Park
(24°04'30"S 132°40’39"E); 20660 Finke Gorge National
Park (24“04'02"S 132°37'36"E); 21128 Carpentaria
Highway (16°44'S 135°02'E); 22702 Macdonncll Ranges
(24°27T3"S 134°24'59"E); 23764 Gregory National Park
(16°49'43''S 130°11'01''E); 23765 Gregory National Park
(16°50'20"S 130°10'58"E); 23807 Gregory National Park
(16°47'50"S 130°09T5"E); 25495 Jasper Gorge (16°05'50"S
130°45T8''E); 34524-25 Arltunga Ruins (23°26'S 134°42'E).
WAM: 13496 Yirrkala (12°15'S 136°53’E); 21852 8 km N
Kalumburu (14°14'S 126“37’E); 21980 Darwin (12°27'S
130°50'E); 23480 Nightcliff (12°23’S 130°52'E); 24001 11
km N Adelaide River (13°08'S 131°08’E); 24198 Helen
Springs (18°26'S 133“52'E); 26224 Parap (12°25'S
130°52'E); 28656 Barrow Island (20'’50'S 115°25'E); 34331-
32 Yirrkala (12°15'S 136°53'E); 37371 Rosemary Island
(20°29'S 116°35'E); 37406 Hermite Island (20°29'S
115°31'E); 40296, 40835 Darwin (12“27'S 130°50'E); 41271
Augustus Island (15°19'S 124°32'E); 41370 Heywood
Island (15°20'S 124°20'E); 41409 Coronation Island
(15°00'S 124°56'E); 43071-75 Crystal Creek (14°31'S
125°48'E); 43119 Port Warrender {14°34'S nS^hl’E);
43151, 43185, 43204, 43211, 43341-42 Mitchell Plateau
(14°52'S 125°50'E); 44278 Gcikie Gorge (18°05'S 125‘’43'E);
44566-71 Behn River Mouth, Lake Argyle (16°15'S
128°45'E); 44572-75 Ord River {16°17'S 128°47'E); 45066
Napier Range (17°13'S 124°38'E); 48559 Shark Point,
Barrow Island (20°52'S 115‘’25'E); 51277 East Palm Spring,
Denison Range (19°20'S 128°20'E); 52670 Lake Argyle
(16°07'S 128°44'E); 54141 Edgar Ranges (18°50'S
123°15'E); 56208-09 Crystal Creek (M^SO’S 125M7'E);
57039 Doongan Homestead (15°23'S 126°18'E); 60352 3
km E Nicholson (18°03'S 128°55’E); 69845 Koolan Island
(16°09'S 123°45'E); 70365 12.5 km 309“New Lissadell
Homestead (16°36'12"S 128°27'34"E); 70385 10.5 km
249°New Lissadell Homestead (16“43'S 128°27'E); 70564
5.2 km 202°Mount Percy' (17°39'37"S 124°54’24"E); 70582
6 3 km 172°Mount Percy (17°40'23"S 124°56'03"E); 70625
5.2 km 202° Mount Perc^ (17°39'37"S 124°54'24"E); 75376,
75398 12 km NW New Lissadell Homestead (16°37'S
128°28'E); 75533 11 km WNW New Lissadell Homestead
(16°39'S 'l28°28'E); 77201 Mitchell Plateau (14°44'30"S
125°47'00"E); 77472 Camp Creek (14°53'30"S

125°45'00"E); 79063 Brooking Springs Station (17°54'S
125°16'E); 80028 Sunday Island (16°25'S 123°11’E); 81286
Koolan Island (16°09'S 123°45'E); 94881 Lake Argyle
(16°07'S 128°42'E); 96784 Sale River (16°05'S 124°40'E);
96828 Camden Head (15°30'S 124°37'E); 96944 The
Dromedaries (16°34'20"S 124°56 40 E); 97091 3.7 km NW
Mount Daglish (16°15'05"S 124°56'00"E); 99774, 99776 10

km SW Silent Grove (17°06'55’’S 125°10'30"E); 101335 30
km ENE Calwynyardah Homestead (17°57'S 125°02'E);
103120, 103129 Purnululu National Park (17°26'S
128°24'E); 103151 Purnululu National Park {17°33'S
128°15'E); 103207 Purnululu National Park (17°10'S
128°44'E); 103384, 103395 Purnululu National Park
(17°15'S 128°18’E); 103483 Purnululu National Park
{17°32'S 128°21'E); 103489-90 Purnululu National Park
(17°29'S 128°22'E); 103733 Koolan Island (16°09'S
123°45’E); 106284 Augustus Island (15°26'S 124°36'E);
108737 10 km N Gordon Downs Homestead (18°40'S
128°35'E); 108815 30 km SE Gordon Downs Homestead
(18°56’S 128°47’E); 112725-26 Mandora Station (19°47’52"S
121°26'53’'E); 114462 King Hall Island {16°05'S 123°25'E);
138149 Napier Range (17°13'S 124°38'E); 139058, 139062-
63 Mandora Station (19°47'52"S 121°26'53"E); 141530
Quanbun Downs Station (18°21'27”S 125°13’10''E).

Delma pax Kluge 1974

NTM; 9939 24 km N Port Hedland (20°23'S 118°48'E);
9943 13 km N Port Hedland (20°25'S 118°42'E); 9945 15
km N Port Hedland (20°25'S 118°42"E); 9946 22 km N
Port Hedland (20°23'S 118°47'E).

WAM: 14803 19 km W Mundabullangana (20°30'S
117°53’E); 51620 10 km NE Mount Newman (23°17'S
119°45'E); 58965 Marble Bar Pool (21°16’S 119°42'E);
64701 Nullagine (21°54'S 120°06’E); 64986 Dampier
(20°40'S 116°42'E); 68370-71 Between Nullagine and Roy
Hill (22°15'S 120°00'E); 70103 Between Dampier and
Karratha (20°45’S 116°45'E); 73144, 73146 24.2 km
234°Marillana Homestead (22°46'00"S 119°13'08"E);
73542-43 Dampier (20°40'S 116°42‘E); 73604 24.2 km
234°Marillana Homestead (22°46'00"S 119°13'08"E); 73841
22 km S Roebourne (21°06'S 117°05'E); 76469 10 km SSW
Cooya Pooya Homestead {21°07'S 117°07'E); 80379
Carrawella Well (21°43’S 115°3TE); 80599 8 km SE
Peedamulla Homestead (21°55'S 115°40'E); 80995-96
South Hedland (20°24'S 118°36'E); 81387-91 Miaree Pool
(20°5rS 116°37'E); 82605 Carawine Gorge (21°29'S
121°01'E); 83153 Karratha (20°44'S 116°51'E); 84982
Dampier Archipelago (20°33'S 116°42’E); 87854 Wickham
(20°40'S 117°07'E); 90886 Woodstock {21°37'01"S
118°57T3"E); 94660-61 Crossing Pool (21°35'S 117°06'E);
94882 Mardie Station (21°15'S 115°50'E); 102046 2 km N
Crossing Pool (20°53'S 116°40'E); 102066 Karratha
{20°53'S 116°40'E); 102091 Dampier (20°40'S 116°42'E);
102115 7 km NE Mount Windell (22°37'28"S 118°36'26"E);
102137 5 km NNE Mount Windell (22°36'16"S
118°34'0r'E); 102149 10 km ENE Mount Windell
(22°35'58"S 118°38'20"E); 104176 Woodstock {21°36'34"S
119°01T7"E); 104297 Woodstock (21°36'25'’S 119°02'23"E);
106257, 106278-79, 108711, 108791, 113387 South Hedland
(20°24'S 118°36’E); 114437 Wittenoom (22°14'S 118°20'E);
116300 King Bay (20°38'S 116°45'E); 119045-46 South
Hedland (20°24'S 118°36'E); 120021 3.5 km NE Mount
Brockman (22°28'S 117°18'E); 120030 Hope Downs
(22°58'00"S 119°09'45"E); 120735 Boodarie Hill (20°24'S
118°3rE); 125023 Yandicoogina (22°43'S 119°01'E);
125452 30 km E Newman (23°19'S 120°02'E); 127829
Mount Brockman (22°25'05"S 117°18'03’’E); 129658 120
km NW Newman (22°59'45"S 119°18'30"E); 129930 West
Angelas (23°15'S 118°40'E); 132548 DeGrey River Station

302

B. Maryan, K.P. Aplin, M. Adams

{20'’13'14"S 119°09'58"E); 132549 DcGrey River Station
(20°17’16"S 119°12'36"E); 132593, 132596 Burrup
Peninsula (20°36'45"S 116‘’47'37"E); 132606 Burrup
Peninsula (20°40'49"S 116“44'37"E); 132657 Burrup
Peninsula (20“31'40"S 116'’49'11"E); 135320, 135336-37
Cape Lambert (20°48'36"S 116°56’31"E); 135632 Bea Bea
Creek (2r43'S 118°44'E); 135919-20 32 km SW South
Hedland (20°36'S 118°22'E); 137857 Munjina Roadhouse
(21°59'S 119°45'E); 139170-71 Cane River Homestead
(22°05’19’'S 115°37'31"E); 139294 Meentheena (21°1713"S
120'’27'34"E); 139352-53 Potter Island (20°57'S 116°08'E);
139369 Meentheena (21°13'56"S 120°19'40"E); 139457
Mount Minnie Homestead (21°58’23"S 115°25'51"E);
140021 Millstream-Chichester National Park {21°10'53"S
117°03'28''E); 140396 82 km E Port Hedland {20°18'53"S
119“24'4r’E); 141269-70 24 km ESE Port Hedland (20°23'S
118°48'E); 141311 Cape Preston area (20°50'00"S
116°09'47"E); 145512 98 km S Port Hedland (21°09'36"S
118°48'36”E); 145544 80 km S Port Hedland (21°00'36"S
118°42'00''E); 145569 34 km S Port Hedland {20°36'36"S
118°36'36"E); 145614 18 km S Port Hedland (20°28'12"S
118=’35'24’'E); 145680 Abydos Station (2r25'S 118°55'E);
145748 Chichester Range' (22°04'44"S n8“58'40"E); 145753
Chichester Range (22°01'01"S 118°58'55"E); 146591 124
km S Port Hedland (21“26'53"S 118°55'24"E); 146649 80
km S Port Hedland (21°00'36"S 118°42'00"E); 151161 Tom
Price area (22°37T3"S 117°44'37"E).

Delma tincta De Vis 1888

WAM: 3440 La Grange (18°40’S 122°01'E); 4511 East
Chapman (28°40'S 114°50'E); 8109 Wandagee Station
(23°49’S 114°27'E); 9782-84 Carnarvon (24°53’S 113°40'E);
10615 Minilya (23°51'S 113°58'E); 11494 Learmonth
district (22°15'S 114°05'E); 12114 Kimberley Research
Station (15°39'S 128°42’E); 13653 VVyndham (15°29'S
128°07'E); 13838 Kalumburu (14°18'S 126°38'E); 13933
Mount Pleasant * (32°02'S n5'’5rE); 14791-95
Northampton (28°21'S 114°38'E); 14801

Mundabullangana (20°31'S 118°03'E); 17683 Turee Creek
Station (23°37'S 118°39'E); 22323 Nabawa (28°30'S
n4°47'E); 22366 Kimberley Research Station (15°39'S
128'’42'E); 24812 Binnu (28°02’S 114°40'E); 25221
Murchison House (27°39'S 114“14'E); 28370

Coordevvandy (25°36'S 115°58'E); 28391 Murchison
House (27°39'S 114°14'E); 28454 Barrow Island (20°48’S
115°24'E); 30259 Carnarvon (24°53’S 113°40'E); 31397 35
km NE Mingenew (29°03'S 115“37'E); 31487 Eradu
(28°42'S 115“02'E); 44555-65 Lake Argyle (16°10'S
128°44'E); 47854 Barrow Island (20°52'S 115°22'E); 48560-
62 Barrow Island (20°52'S 115°25'E); 50091 Yalgoo
(28°21’S 116°41'E); 51003-04 Exmouth (21°56'S 114°07’E);
51641 Marandoo (22°38'S 118°08'E); 52933 Shothole
Canyon (22°03'S 114°02'E); 53791-93 Gascoyne Junction
area (25°06’S 115°13'E); 54606 Wooramel Homestead
(25°44’S 114°17'E); 55019 Hamelin Homestead (26°26'S
114°12’E); 55094 Wooramel Homestead (25°44'S
114°17'E); 55406-07, 55440 71 km W Barrv' Caves (19°52'S
136°03'E); 58413 5 km N Coulomb 'Point (17°19'S
122'’10'E); 59687, 59689 Meeberrie Homestead (26°58'S
115°58'E); 62208 Mingenew (29°12'S 115°26'E); 62416 5
km W Williambury Homestead (23°54'S 115°10'E); 63678
25 km NNW Winning Homestead (22V56'S 114°27'E);
66313 36 km 137°Mount Meharry (23°12'30"S
118°49'30"E); 66314 34 km 136°Mount Meharry
(23°11'40'’S 118°49’30"E); 66327 36 km 137°Mount
Meharry (23°12'30"S 118°49'30"E); 67606-09 Marble Bar
(21°10’S 119°44'E); 67806 Hamelin Pool (26°24'S 114°10'E);

67905 36 km 137°Mount Meharry (23M2'30"S
118°4930"E); 69779 Mount Bruce (22°35’S 118°10'E);
70757, 70761, 70764 30.2 km 238°Marillana Homestead
(22°46'55"S 119°09'35"E); 71059 Hamelin Homestead
(26“26'S li4°12'E); 73138 30.2 km 238°Marillana
Homestead (22°46'58"S 119°09'35"E); 73897 Pender Bay
area (16°45'S 122°49'E); 78239 70 km W Barry Caves
{19°51'S 136°02’E); 81330 57 km NNE Nanutarra
Roadhouse (22°01'S 115°36'E); 83152 Karratha {20°44’S
116°5rE); 83210 Carnarvon (24°53'S 113°40'E); 84150-52
Yalgoo (28°21’S 116°41'E); 85190 8 km ESE Kununurra
(15°49'S 128°48'E); 86429 Hamelin Homestead (26°26’S
114°12'E); 88547 Carnarvon (24'’53'S 113°40'E); 91132 10
km NE Paynes Find (29°11'S 117°42'E); 92727 Hamelin
Homestead (26°26'S 114°12'E); 93701 53 km NNE

Broome (17°32'S 122°25'E); 95291-93 Walga Rock (27°24'S
117°28'E); 99180 Woodstock Station (21°36'35"S
118''57'44"E); 101246 Galena (27°50'S 114°41'E); 101278
Barrow Island (20°48'S 115“24'E); 102154 10 km ENE
Mount Windell (22'’35’58"S 118“38'20"E); 102401 Barlee
Range Nature Reserve (23“04'47"S 115°47T4"E); 102815
Burrup Peninsula (20°4039"S 116M5T1"E); 102838 Cape
Range National Park (22°09'01"S 1 13°59'52"E); 102852
Meentheena (21°14T6"S 120°23'31"E); 139140 Meentheena
(21°25'18"S 120°25’36"E); 104426, 105987 Carnarvon
(24°53'S 113“40'E); 112511 Urala Station (21°47'04"S
114°52'07"E); 112689 10 km SSW Onslow (21°43'51''S
115°05'49"E); 112690 5.5 km SE Onslow (21°40'33"S
115°08’42'’E); 112691 11 km S Onslow (21°44'27"S
115°06'46"E); 112715 5.5 km SE Onslow (21°40'33"S
1]5°08'42"E); 112716 12 km SE Onslow (21°42'39"S
115°11'49"E); 112747 Bibawarra Crossing (24°53'S
113°42'E); 113012, 113030 Lesley Salt Works (20°14’50"S
118°50'50"E); 114101-02 Carnarvon Airport (24°54'S
113°39'E); 114391-92 9 km NE Broome (17°55'S 122“15'E);
114490 Wicherina Dam (28°44'S 115°00'E); 115018
Spalding Park (28°39'S 114°38'E); 116439 15 km NNW
Carlton Hill Homestead (15°23'39"S 128°28T3"E); 116545
Depot Hill (29°08'S 115°21'E); 117215 Narngulu (28°49'S
114°41'E); 117342 Hope Downs (22°56'45"S 119°07'30"E);
120020 3.5 km NE Mount Brockman (22°28'S 117H8’E);
125032 Yandicoogina (22°43T4"S 118°59’26"E); 127718,
127768, 127792 5 km S Mount Tom Price Mine (22°47'55"S
117°46T0"E); 129587, 129623 120 km NW Newman
(22°55'S 118°54'E); 131752 Mount Robinson (22°57T9"S
118°46'14"E); 132209 Urala Station (21°47'09"S

114°31'58"E); 135322 Cape Lambert (20°45'16"S

117°04'52"E); 135422 Mount Brockman (22°18-38"S

117M5'08"E); 135487 Urala Station (2r46'58"S

114°52'11"E); 137953 35 km NNE Kununurra (15°35'20"S
128°59'00"E); 138222 Karijini National Park (22°37'S
118°17'E); 138226 Karijini National Park (23°01'S
118°43'E); 138243 Karijini National Park (22°59'S
118°44'E): 139140 Meentheena (21°25T8''S 120°25'36”E);
139282 Meentheena (21M7'07"S 120°24'55''E); 139308
Meentheena (2ri4'41"S 120'’19'20''E); 139314 Meentheena
(21°13'04"S 120°27'20"E); 139321 Meentheena (21°15'20"S
120°27'18"E); 139328 Meentheena (21°16'54"S

120°27'58"E); 141273 22 km ESE Port Hedland (20°23'S
118°47'E); 141584 1 km N Quobba Homestead (24°22'24"S
113°24T9"E); 141585-86 Quobba Station (24°27'42"S
113°24'28"E); 145250 5 km S Mount Tom Price Mine
(22°48'31"S 117“47'09'’E); 145650 235 km SSW Port
Hedland (22°23'24"S 118°58'48’'E); 146589, 146645 228 km
SSW Port Hedland (22°20'24"S 119°00'00"E); 146890
Mirima National Park (15°47'S 128°44'E); 146957
Kalumburu (14°18'S 126°38'E); 151059-60 10 km E

Pygopdids from NW Australia

303

Carnarvon (24°53’S 113°46'E); 153814 2 km S Yardie
Homestead Caravan Park (21°53'37'’S 1 14°00'34"E);
153820 Charles Knife Road (22°07'08"S 114°03'44"E);
153821 Bullara Station (22°48'33"S 113°56'39"E).

*As noted by Kluge (1974), this locality record is
considered erroneous.

Delma sp. incertae sedis

SAM: 5058 Warburton Range (26°06'S 126°39'E); WAM:
25201 32 km E Jiggalong (23°22'S 121°05'E).

Specimens examined in allozyme analysis

Delma pax

WAM: 104297 Woodstock Station; 120021 3.5 km NE
Mount Brockman; 120030 Hope Downs; 125452 30 km E
Newman; 132548 De Grey River Station; 132596, 132606
Burrup Peninsula; 135920 South Hedland; 139171 Cane
River Homestead; 139294 Meentheena; 140021 Millsteam-
Chichester National Park; 141270 24 km ESE Port
Hedland.

Delma desmosa sp nov.

WAM: 102650, 102657 Little Sandy Desert; 114555
Sandfire Roadhouse; 132802 Warri Airstrip; 139089
Mandora; 145073 Officer Basin area.

Delma tealei sp nov.

WAM: 102837 Cape Range National Park; 153811 Charles
Knife Road; 153813 2 km S Yardie Homestead Caravan
Park; 153819 Shothole Canyon.

Delma bore a

WAM: 139058, 139063 Mandora; 141530 Quanbun Downs
Station.

Delma tincta

WAM: 102401 Barlee Range Nature Reserve; 102838 Cape
Range National Park; 114391 9 km NE Broome.

Delma butler!

"western"

WAM: 120819 Peron Peninsula (26° OO'S 113° 30'E);
141590 Boolathana Station (24° 39’S 113°42'E); 127461
East Yuna Nature Reserve (28° 20'S 115°00'E); 144711
Bungalbin Hill (30°24'S 119°38’E).

"central"

WAM: 106163 Mount Frazer (25°38'S 118°23'E); 135249
Wiluna (26°35’S 120°14'E); 145072 Officer Basin (29°58’S
123°46'E).

SAM: 35027 Bloodweed Bore (26°57'S 140°57'E).

"eastern"

SAM: 45210 Peebinga Conservation Park (34° 58'S 140°
50’E).

Delma haroldi

WAM: 102123 Mount Windell (22° 39’S 118°33'E); 135924
Sandfire Roadhouse (19° 46'S 121° 05'E); 145653 Port
Hedland (20° 18'S 118° 35'E); NTM: 16484 Wave Hill
Station (17°29'S 130°57'E).

304

B. Maryan, K.P. Aplin, M. Adams

« 2 .S

S 5

H t;
2

^ -a
c H

^ E
° " ^
0) ^ S

,Q 5

.2 ^ ■£

g i2 i

g-s

C

O-

^ *Sd>

t— ' ^ u

T!; " I

^ E w

C/5 a>

V c/5 '

S ^ "

2^2

Ci- t

O v'
O) V

E "a S

>> P E!

N c/5

o 00 a>

I/PJS

POS

liuSj

pSj9

adsj

gdaj

yddj

Z>ldpN

ndpN

idpv

im

ZHPl

WD

fdo

pdD

Zioo

lioo

oiD

epD

ainj

dpd

zm

PUH

P'Q

ej

ZHPV

mpv

epv

dAdy

zdDV

idoy

ZUODy

luooy

C/5 r,

a» £
u 3

< z

■o *0

fC5

r 3 r 3 Q

Xi^^-QX5-Q_QXi I^XiXi
'O'OTJ-n’TS'O'O’O’O'O’U-o
rsrarsacococcrSGfOC

UUUUUUUUUUUUXj

u

CJ U CJ

CJ JO

^ x»

Xi-DXiXiX»X<XiX5XJ^X5

2 Ti 7! ^ *0

Jij Xi

Jli Xi

XI Xi -D X)

XC-^UCJCJUCJCJCJCJ

XXXXXXXXX CJ

TO c^

TO TO TO TO

X X

U CJ

U U

X X
X X
X X
X X

X X

U CJ

u u

-TO
X X
X X

xxxxxxxxxx

CJUUCJCJCJUCJCJU

X X X X

UUCJCJUCJCJCJUCJCJCJUCJ

X X
X X
X X
X X
X X

X X
X X
X X
X X

X X
X X
X X
X X

X TO

X X
X X
X X

X X

X X
X X

XXX

XXX
I X X

TO TO TO

CJUUCJCJCJUUCJCJCJ

X X

CJ u

CJCJCJCJCJCJCJUCJUCJCJCJCJCJCJCJCJ

XXXXXXXXXXXXXX

CJCJCJCJCJUCJCJUCJCJCJCJCJ

X X

CJ CJ

^ -TO

CJ u

X
TO TO

X X
X X
-S XI

TO

CJ -Q
CJ u

X
TO TO

-O X3
X X
X X

^ TJ

CJ u

T3

u u

X X X X

-TO "TO
X X
X X

TO

TO

X X X X X X

TO TO TO TO TO

”0 "O "O ”0 T3

TO TO

-u 'O
X X
X X
X X

TO TO

X X

TO TO

-O
u u

X o

TO TO

TJ -O
X X
X X
X X
X X

TO TO

UCJCJUCJUCJCJ

TOTOTOTOTOTOTOTO

73 ~0
X X
X X

TO

X X
X X
X X

X
X X
X X
X X

TO TO

"O "O 73 73 "C "C CJ

X X
X X

TO TO

•p u

TOTOTOTOTOTOTOTOTOTOTOTOTOTOTOTOTOTO

C3n CsI
CN O
rt O
O <N

§ §

< <

O IN
Lo
O NI

o m

(N fN

§ s
< <
5 ^

00 vO
"If o
m LT,
(N (N

ro tn

1 §

< <

^£5 O
O (N
M3 ON
CM in
m ro

S. i
< <
^ §

< <
^ 5

S 2
< <
^ 5

O IN

in in

NO VO
CM CM

o o

§ §
< <
5 ^

S E

XXXXXXXXXXXXSS

TOTOTOTOTOTOTOTOTOTOTOTOQj’li

n,a,a.a,a,a,a.a,a,Q,a,a.t3-o

desmosa WAM114555

desmosa WAM132802

desmosa WAM139089

desmosa WAM145073

Appendix 2 (cont.)

Pygopdids from NW Australia

HP^S

b

b

b

b

b

b

b

d

d

b

c

b

c

c

b

c

c

c

c

c

c

c

c

POS

l3-Ci-l3-T3uuu-t3-oxiTJ lT3^'aTJ’^-t3'®ran:S^

a

a

a

a

a

ab

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

cja;ocjfcOC>CfcCT3"a"aii 1 raoiuooi'-^'ojajaicu't;

pScI9

.Qxixi-^xijixixi uxi-aT3-aT3TJt3T3

adsj

gdad

e

e

Q

e

e

0

c

e

e

e

be

e

be

e

e

be

be

a

be

ac

b

be

be

yddd

c

c

c

c

c

c

c

c

c

c

c

ac

c

c

c

c

c

d

c

c

c

c

c

ZndpM

b

b

b

b

b

b

ab

b

b

b

b

b

b

b

b

b

bd

b

b

b

b

b

b

ndpN

a

a

a

a

a

a

a

a

a

a

b

b

b

b

b

b

b

b

b

b

b

b

b

idiAi

b

b

b

b

b

b

b

b

b

b

a

a

a

a

a

a

a

a

a

a

a

a

a

lan

c

c

c

c

c

c

c

c

c

c

b

b

c

b

b

b

b

b

b

c

c

be

c

ZPPI

c

c

c

c

c

c

c

c

c

c

b

ab

b

b

b

b

b

b

b

ab

b

b

bd

WD

b

b

b

b

a

a

ab

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

jdo

b

b

b

b

b

b

b

b

b

b

c

c

c

c

c

c

c

c

c

c

c

c

c

pdo

XjXiXiXiXjXi_QXia:Xi4=i IXi IXi IXiXiXiXi_DX)Xi

Zioo

XiXl-QXiXt-DXiXl-QX) U U U U U U U U U U U 0*0

UOQ

oiO

c

c

c

c

c

c

c

c

c

c

d

d

d

d

d

d

d

d

d

d

d

d

d

epD

cj(:joajT3'a-a-0-0-auouuJ^ouuuuuu_^

Lunj

b

b

b

b

b

b

b

b

b

b

bd

b

bd

b

b

b

b

be

b

bd

ab

b

b

dpd

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

ZiSd

b

b

b

b

be

be

b

ab

b

b

b

b

b

b

b

b

b

b

b

b

be

IOU3

b

b

b

b

c

c

c

c

c

c

r

ac

c

c

b

c

c

c

c

b

c

c

c

B;a

i bc£ I'Sb’fccM)^ i i i i

ej

a

a

a

a

a

a

a

a

a

a

b

b

b

b

b

b

b

b

b

b

b

b

b

ZPPV

c

c

c

c

c

c

c

a

a

a

be

c

c

b

c

c

c

c

b

b

b

b

b

mpv

i ixixi ixixi-gxijaxixi

epv

dAdy

a

a

a

a

a

a

a

a

a

b

a

a

a

a

a

a

a

a

a

a

a

ac

ac

Zday

b

b

b

b

b

b

b

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

idoy

a

a

a

a

b

b

b

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

Zuojy

c

c

c

c

c

c

cd

ac

be

b

d

a

a

d

a

ab

d

de

d

a

d

c

ad

luoDy

a

a

a

a

a

a

a

a

a

a

b

b

be

b

b

b

a

b

a

a

ab

a

ab

Accession

Number

WAM102837

W AMI 53811

WAM153813

WAM153819

WAM139058

WAM139063

WAM141530

W AMI 02401

W AMI 02838

WAM114391

WAM106163

WAM120819

WAM127461

WAM135249

WAM141590

WAM144711

WAM145072

SAM45210

SAM35027

WAM102123

WAM135924

WAM145653

NTM16484

'u' ^

Species

tealei

tea lei

tealei

tealei

borea

borea

borea

tincta

tincta

tincta

butleri (

biitieri (

butieri 1

butieri 1
butieri (

butieri {

butieri I
butieri |
butieri |
haroldi

haroldi

haroldi

haroldi

Records of the Western Australian Museum 23: 307-308 (2007).

Short communication

First record of the freshwater sawfish, Pristis microdon,
from southwestern Australian waters

Justin A. Chidlow

Department of Fisheries, Western Australian Marine Research Laboratories,
PO Box 20, North Beach, Western Australia 6920
Email: jchidlow@fish.wa.gov.au

Sawfishes (family Pristidae) are large (up to 7m)
modified batoids with a blade-like snout edged
with pairs of rostral teeth. They occur worldwide in
sub-tropical and tropical shallow coastal sea,
estuaries and freshwater systems (Last and Stevens
1994; Compagno and Last 1998). There are between
five and seven recognised species worldwide, with
five species represented in Australian waters (Last
and Stevens 1994). Sawfish populations have been
extirpated from many parts of their original global
range by gillnetting and trawling and are easily
entangled in nets by their toothed rostra
(Simpfendorfer 2000). The little that is known about
the biology of sawfish suggests they have low rates
of reproduction (Tanaka 1991; Compagno and Last
1998; Wilson 1999; Simpfendorfer 2000; Thorburn
et al. 2004). This combined with their susceptibility
to fishing gear, make sawfish a high risk species
and all have subsequently been listed globally as
critically endangered under the lUCN Red List
Assessment 2006 (Compagno etal. 2006).

Pristis microdon Latham, 1794

Pristis microdon is a medium to large sawfish that
in Australia grows to at least 361cm TL (Tanaka
1991), but is reported to reach up to 700cm TL in
other locations (Last and Stevens 1994). They are
born at around 50cm in length after a five month
gestation period, with litter sizes ranging between 1
and 12 (Wilson 1999). In the western Atlantic P.
microdon matures at between 240cm and 300cm TL
(Compagno and Last 1998). Tanaka (1991) reported
two male specimens from New Guinea, one
measuring 247cm that was immature and a 361cm
specimen that was mature. In Australian waters, P.
microdon feeds on fish such as catfish, small
crustaceans and molluscs (Allen 1982; Cliff and
Wilson 1994; Pogonoski et al. 2002; Thorburn et al.
2004).

Pristis microdon occurs inshore and in intertidal
areas and is usually found in freshwater drainages,
lakes and estuaries where it can penetrate as far as
400km from the coast (Morgan et al. 2004). In the
Indo-West Pacific it ranges from New Guinea, SE
Asia, northern Australia and west to South Africa

(Last and Stevens 1994; Compagno and Last 1998).
Pristis microdon may also occur in the Atlantic and
eastern Pacific if P. perotteti Muller & Henle, 1841
and P. zephyreus Jordan & Starks in Jordan, 1895
are synonymised with this species (Compagno and
Last 1998). In Australia, the freshwater sawfish is
known to occur in the Ord, Durack and Fitzroy
Rivers (Western Australia), the Adelaide, Victoria
and Daly Rivers (Northern Territory), and the
Gilbert, Mitchell, Norman and Leichhardt Rivers
(Queensland) (Last and Stevens 1994; Pogonoski et
al. 2002; Thorburn et al. 2004). Only recently has P.
microdon been reported from marine waters
(Thorburn etal. 2004).

Southwestern Australian P. microdon

A female P. microdon was captured by a
commercial shark fisher operating demersal gillnets
in southwestern Australian waters on the 5* of
February 2003. The capture location was
approximately six miles east of Cape Naturaliste
(33°31'S, 115°07'E) in 32m of water. The sawfish was
estimated to be 3.5m in length TL when landed and
appeared to be healthy. The specimen was
processed and the fisher retained the remaining
trunk, fins and saw. I positively identified the
processed sawfish as P. microdon using an
identification kev provided by Last and Stevens
(1994).

The partial length (origin of the first dorsal fin to
the insertion point of the second dorsal fin) was
95cm (approximate as the trunk had been cut in
half). The rostral saw length was 79cm with 19 pairs
of teeth that extended to the basal quarter of the
saw (Figure 1). The interspace between rostral teeth
at the base of the saw was 4cm, and 3cm between
the teeth at the tip of the saw (Figure 1). A groove
was present along the posterior margin of all rostral
teeth. The origin of the first dorsal-fin was located
anterior to the pelvic-fin origin and the height of
the first dorsal-fin was 32cm. The second dorsal-fin
height was 31cm. The ventral lobe of the caudal-fin
was small, but distinct. The upper and lower
postventral caudal-fin margins measured 44.5cm
and 11.5cm respectively.

308

J.A. Chidlow

Figure 1 Rostral saw from a female Pristis microdon, measuring approximately 350cm in total length, captured off
Cape Naturaliste, Western Australia. See text for description of measurements.

This record of P. microdon from southwestern
Australia extends the range of the species
approximately 1600 km south of its previously
known southern limit, Cape Keraudrcn, Western
Australia (Thorburn ef al. 2004) and provides
further confirmation that P. microdon utilizes
marine waters.

ACKNOWLEDGEMENTS
I wish to thank J. Nelson who thoughtfully
informed the Shark Research Section, Department
of Fisheries WA of the capture, and provided
assistance in identifying and collecting data from
the specimen. 1 would also like to thank P. Last
from CSIRO Marine Research, Hobart, R. McAuley
from Department of Fisheries WA, Perth and C.
Simpfendorfer from the Mote Marine Laboratory,
Florida for their assistance in positively identifying
the specimen.

REFERENCES

Allen, G.R. (1982). A field guide to inland fishes of
Western Australia. Western Australian Museum,
Perth, Western Australia. 86pp.

Cliff, G. and Wilson, G. (1994). Natal sharks board's
guide to sharks and other marine animals. Natal
Sharks Board, p33.

Compagno, I..J.V. and Last, J.D. (1999). Pristiformes:
Pristidae. In K.E. Carpenter and V.H. Niem (eds).
FAO species identification guide for fishery purposes.
The living marine resources of the IV'esfern Central
Pacific. Volume 3. Batoid fishes, chimaeras and bony
fishes part 1 (Elopidae to Linophrvnidae). Pp. 1410-
1417. FAO, Rome.

Compagno, L.J.V., Cook, S.F. and Fowler, S.L. (2006).
Pristis microdon. In lUCN 2006 lUCN Red List of
Threatened Species.

Jordan, D.S. and Starks, E.C. (1895). In The fishes of
Sinaloa. D.S. Jordan. Proceedings of the Califfornia
Academy of Sciences (Ser. 2) 377-314.

Last, P.R. and Stevens, J.D. (1994). Sharks and rays of
Australia. CSIRO, Melbourne, 513 pp.

Latham, J. (1794). .An essay on the various species of
sawfish. Transactions of the Linnean Society of
London 2 (23); 273-282.

Morgan, D.L., Allen, M.G., Bedford, P. and Horstman,
M. (2004). Fish fauna of the Fitzroy River in the
Kimberley region of Western Australia - including
Bunuba,Gooniyandi, Ngarinyin, Nyikina and
Walmajarri Aboriginal names. Records of the Western
Australian Museum 22:147-161.

.Muller, J and Henle, F.G.J. (1841). Systematische
Beschreibung der Plagiostomen Berlin. Plagiostomen
i-xxii + 1-200.

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

Simpfendorfer, C.A. (2000). Predicting population
recovery rates for endangered western Atlantic
sawfishes using demographic analysis.
Environmental Biology of Fishes 58: 371-377.

Tanaka, S. (1991). Age estimation of freshwater sawfish
and sharks in northern Australia and Papua New
Guinea. The University Museum, University of
Tokyo. Nature and Culture 3: 71-82.

rhorburn, D., Morgan, D., Gill, H., Johnson, M., Wallace-
Smith, FI., Vigilante, T., Gorring, A., Croft, I. and
Fenton, J. (2004). Biology and cultural significance of
the freshwater sawfish (Pristis microdon) in the
Fitzroy River, Kimberley, Western Australia. Report
to the Threatened Species Network 2004. 57 pp.

Thorson, T.B. (1982). The impact of commercial
exploitation on sawfish and shark populations in
Lake Nicaragua. Fisheries 7(2): 2-10.

Wilson, D. (1999). Freshwater sawfish Pristis microdon.
Australia New Guinea Fishes Associations' A-Z
notebook of native fre.shwater fish. ANGFA Bulletin
41.

Mnuiiscript received 23 August 2004; accepted 3 August 2006

Guide to Authors

Subject Matter:

Reviews, observations and results of research into
all branches of natural science and human studies
will be considered for publication. However,
emphasis is placed on studies pertaining to We.stem
Australia. Longer papers will be considered for
publication as a Supplement to the Records of the
Western Australian Museum. Short communications
should not normally exceed three typed pages and
this category of paper is intended to accommodate
observations, results or new records of significance,
that otherwise might not get into the literature, or
for which there is a particular urgency for
publication. All material mu.st be original and not
have been published elsewhere.

Presentation:

Authors are advised to follow the layout and style
in the most recent issue of the Records of the
Western Australian Museum including headings,
tables, illustrations and references.

The title should be concise, informative and
contain key words necessary for retrieval by
modern searching techniques. An abridged title (not
exceeding 50 letter spaces) should be included for
use as a running head.

An ab.stract mu.st be given in full length papers
but not short communications, summarizing the
scope of the work and principal findings. It should
normally not exceed 2% of the paper and should be
suitable for reprinting in reference periodicals.

The International System of units should be used.

Numbers should be spelled out from one to nine
in descriptive text; figures used for 10 or more. For
associated groups, figures should be used
consistently, e.g., 5 to 10, not five to 10.

Spelling should follow the Concise 0.xford
Dictionary.

Systematic papers must conform with the
International Codes of Botanical and Zoological
Nomenclature and, as far as possible, with their
recommendations.

Synonymies should be given in the short form
(taxon, author, date, page) and the full reference
cited at the end of the paper. All citations, including
those associated with scientific names, must be
included in the references.

Manuscripts:

The original and two copies of manuscripts and
figures should be submitted to the Editor, c/-
Publications Department, Western Australian
Museum, Locked Bag 49, Welshpool DC, Western
Australia 6986. They must be in double-spaced
typescript on A4 sheets. All margins should be at
lea.st 30 mm wide. Tables plus heading and legends
to illustrations should be typed on separate pages.
The desired position for insertion of tables and
illustrations in the text should be indicated in
pencil. Tables should be numbered consecutively,
have headings which make them understandable
without reference to the text, and be referred to in
the text.

High quality illustrations are required to size
(16.8 cm X 25.2 cm) or no larger than 32 cm x 40
cm with sans serif lettering suitable for reduction to
size. Photographs must be good quality black and
white prints, not exceeding 16.8 cm x 25.2 cm.
Scale mu.st be indicated on illustrations. All maps,
line drawings, photographs and graphs, should be
numbered in sequence and referred to as Figure/s in
the text and captions. Each must have a brief, fully
explanatory caption. On acceptance a computer disk
containing all corrections should be sent with
amended manuscript. The disk should be marked
with program (e.g. Word, WordPerfect, etc).

In papers dealing with historical subjects
references may be cited as footnotes. In all other
papers references must be cited in the text by author
and date and all must be listed alphabetically at the
end of the paper. The names of journals are to be
given in full.

Processing:

Papers and short communications are reviewed
by at least two referees and acceptance or rejection
is then decided by the editor.

The senior author is sent one set of page proofs
which must be returned promptly.

The senior author will receive fifty free offprints
of the paper. Additional offprints can be ordered at
page proof stage.

Records of the Western Australian Museum
Volume 23 Part 3 2007

CONTENTS

Nadine A. Guthrie

A new species of Gnathoxys (Coleoptera: Carabidae: Carabinae) from an

urban bushland remnant in Western Australia 213

Terry F. Houston

Observations of the biology and immature stages of the sandgroper
Cylindraustralia kochii (Saussure), with notes on some congeners

(Orthoptera: Cylindrachetidae) 219

Scott A. Thompson and Graham G. Thompson

Temporal variation in ground-dwelling invertebrate biomass in the

Goldfields of Western Australia 235

Paul Doughty and Marion Anstis

A new species of rock-dwelling hylid frog (AnuraiHylidae)

from the eastern Kimberley region of Western Australia 241

Marion Anstis, J. Dale Roberts and Ronald Altig

Direct development in two Myobatrachid Frogs, Arenophryme

rotunda Tyler and Myobatrachus gouldii Gray, from Western Australia 259

Brad Maryan, Ken P. Aplin and Mark Adams

Two new species of the Delma tincta group (Squamata; Pygopodidae)

from northwestern Australia 273

SHORT COMMUNICATION

Justin A. Chidlow

First record of the freshwater sawfish, Pristis microdon, from
southwestern Australian waters

307