A new species of frog of the Eleutherodactylus lacrimosus assemblage (Leptodactylidae) from the western Amazon Basin, with comments on the utility of canopy surveys in lowland rainforest

Abstract

We describe a new species of Eleutherodactylus from the lowlands of the western Amazon Basin. The new species is referred to the Eleutherodactylus unistrigatus group, lacrimosus assemblage. It differs from other members of the group by having a dorsal olive-green coloration with an interorbital creamy yellow stripe that extends posterolaterally and reaches the level of the sacrum, and low ulnar and tarsal tubercles. The new species inhabits western Amazon tropical rainforests and has been found in arboreal bromeliads by day and on vegetation by night. We discuss the effect of lack of sampling in the forest canopy in our understanding of tropical amphibian communities. Based on work conducted at two localities in Ecuadorian Amazonia, we find that even limited sampling effort in the canopy can greatly improve efficiency of biological inventories.

Full text

Herpetologica, 62(2), 2006, 191–202
Ó 2006 by The Herpetologists’ League, Inc.
A NEW SPECIES OF FROG OF THE ELEUTHERODACTYLUS
LACRIMOSUS ASSEMBLAGE (LEPTODACTYLIDAE) FROM THE
WESTERN AMAZON BASIN, WITH COMMENTS ON THE UTILITY
OF CANOPY SURVEYS IN LOWLAND RAINFOREST
JUAN M. GUAYASAMIN
1,6,8
,SANTIAGO R. RON
2,6
,DIEGO F. CISNEROS-HEREDIA
3
,
W
ILLIAM LAMAR
4
, AND SHAWN F. MCCRACKEN
5,7
1
Natural History Museum & Biodiversity Research Center, Department of Ecology and Evolutionary Biology,
The University of Kansas, Lawrence, KS 66045-7561, USA
2
Section of Integrative Biology and Texas Memorial Museum, The University of Texas, Austin, TX 78712, USA
3
Colegio de Ciencias Biolo
´
gicas & Ambientales, Universidad San Francisco de Quito, Ave. Interocea
´
nica,
Campus Cumbaya
´
, Edif. Maxwell, Decanato. Casilla Postal 17-12-841, Quito, Ecuador
4
University of Texas at Tyler, Department of Biology, 3900 University Blvd., Tyler, TX 75799, USA
5
Texas State University, Department of Biology, 601 University Drive, San Marcos, TX 78666, USA
6
Museo de Zoologı
´
a, Centro de Biodiversidad y Ambiente, Escuela de Biologı
´
a, Pontificia Universidad Cato
´
lica
del Ecuador, Av. 12 de Octubre y Roca, Aptdo. 17-01-2184, Quito, Ecuador
7
Tadpole Organization, 2214 South First Street, Austin, TX 78704, USA
A
BSTRACT: We describe a new species of Eleutherodactylus from the lowlands of the western Amazon
Basin. The new species is referred to the Eleutherodactylus unistrigatus group, lacrimosus assemblage. It
differs from other members of the group by having a dorsal olive-green coloration with an interorbital creamy
yellow stripe that extends posterolaterally and reaches the level of the sacrum, and low ulnar and tarsal
tubercles. The new species inhabits western Amazon tropical rainforests and has been found in arboreal
bromeliads by day and on vegetation by night. We discuss the effect of lack of sampling in the forest canopy
in our understanding of tropical amphibian communities. Based on work conducted at two localities in
Ecuadorian Amazonia, we find that even limited sampling effort in the canopy can greatly improve efficiency
of biological inventories.
Key words: Anura; Canopy sampling; Ecuador; Eleutherodactylus aureolineatus; E. lacrimosus;
E. unistrigatus; Leptodactylidae; New species; Peru
THE GENUS Eleutherodactylus comprises 621
recognized species (Amphibiaweb, 2005) dis-
tributed in Central and South America, as well
as the West Indies. The Eleutherodactylus
lacrimosus assemblage is a phenetic subgroup
of the larger E. unistrigatus group (sensu
Lynch and Duellman, 1997) and currently
contains 14 species: Eleutherodactylus apicu-
latus Lynch and Burrowes 1990 (but see
Heyer and Hardy, 1991:444); E. boulengeri
Lynch 1981; E. brevifrons Lynch 1981; E.
bromeliaceus Lynch 1979; E. dorsopictus
Rivero and Serna 1988 ‘‘1987’’; E. eremitus
Lynch 1980; E. lacrimosus (Jime
´
nez de la
Espada 1875); E. mendax Duellman 1978a;
E. olivaceus Ko
¨
hler, Morales, Lo
¨
tters, Reichle,
and Aparicio 1998; E. petersorum Lynch 1991;
E. prolixodiscus Lynch 1978; E. schultei
Duellman 1990; E. tayrona Lynch and Ruı
´
z-
Carranza 1985; and E. zimmermanae Heyer
and Hardy 1991. Most members of this
assemblage likely are bromeliad-dwelling and
have a small body size (SVL in males, 14.5–
26.6 mm; in females, 20.3–33.8 mm) and
broad, flat, pointy heads.
The morphological resemblance among the
species in the Eleutherodactylus lacrimosus
assemblage has been explained as the result
of convergent evolution in response to their
microhabitat use (bromeliads; Lynch and Ruı
´
z-
Carranza, 1985), rather than as a result of
common ancestry, although no formal test has
been performed to distinguish between these
two hypotheses. Until robust phylogenetic
hypotheses are available, the E. lacrimosus
assemblage is recognized only as a matter of
convenience (Lynch and Ruı
´
z-Carranza, 1985).
Recent collection efforts in the canopy of
Amazon rainforest and examination of speci-
mens previously identified as Eleutherodacty-
lus lacrimosus revealed the presence of an
8
CORRESPONDENCE: e-mail, juanm@ku.edu
191

undescribed species. Herein, we describe this
new species and discuss the effect of sampling
the forest canopy in our understanding of
tropical amphibian communities.
M
ATERIALS AND METHODS
The diagnosis and description of the new
species follow those of Lynch and Duellman
(1997). We examined alcohol-preserved speci-
mens from the herpetological collections at
Museo de Zoologı
´
a of the Universidad Cato
´
lica
del Ecuador, Quito (QCAZ); Universidad San
Francisco de Quito, Quito (DFCH-USFQ);
Natural History Museum of The University of
Kansas, Lawrence (KU); and The Natural
History Collection at Texas A&M University –
Texas Cooperative Wildlife Collection, Col-
lege Station (TCWC). In addition to the type
series of the new species, specimens examined
are listed in Appendix I; one of us (WL)
examined an additional specimen (CRG
931; adult female from Reserva Allpahuayo,
25 km S Iquitos) of the new species that
currently forms part of a collection by Carlos
Rivera G.
Morphological measurements were taken
as described in Guayasamin (2004) and are:
(1) snout–vent length (SVL); (2) tibia length;
(3) foot length; (4) head length; (5) head width;
(6) interorbital distance; (7) upper eyelid
width; (8) internarial distance; (9) eye-to-nostril
distance; (10) snout–eye distance; (11) eye
diameter; (12) tympanum diameter; (13) eye-
to-tympanum distance; (14) radioulna length;
(15) hand length; (16) Finger-I length. Sexual
maturity was determined by the presence of
eggs or convoluted oviducts in females or by
the presence of vocal slits in males. Color pat-
terns in life weretaken from field notes and color
photographs.
At Yasunı
´
Scientific Research Station of the
Universidad Cato
´
lica del Ecuador (YSRS),
visual encounter surveys (mainly ground-level
VES) have been carried out regularly since
1995, and sporadic canopy searches (especially
in tank bromeliads) were carried out during
one week in May 2002.
At Tiputini Biodiversity Station of the
Universidad San Francisco de Quito (TBS),
surveys (ground-level transect VES and quad-
rats) have been carried out since 1998; at this
locality, canopy surveys were conducted within
a 100 m-long system of canopy bridges, and
annual bromeliad inspections were performed
(surveys of all bromeliads on three trees once
per year). During the VES, most of the
amphibians found were collected; thus, the
number of specimens recorded was used as a
proxy for inventory sampling effort.
SFM conducted bromeliad patch sampling
using single-rope climbing technique to
access canopy bromeliads in July and August
of 2004. Two trees in each of four quadrat
plots were surveyed; five bromeliads were
removed from each tree at varying vertical
heights and placed in large plastic bags to be
lowered to the forest floor. Then, the
obtained bromeliads (n 5 40) were searched
by removal of individual leaves to facilitate
collection of all anurans.
Eleutherodactylus aureolineatus sp. nov.
Fig. 1A
Eleutherodactyluss lacrimosus — Duellman,
1978.
Eleutherodactylus sp. nov.—Ron, 2001; Read,
2000; Cisneros-Heredia, 2003.
Holotype.—QCAZ 20712 (field no. SC
10141), an adult male collected by Italo Tapia
and SRR at the Yasunı
´
Scientific Research
Station, Universidad Cato
´
lica del Ecuador
(768 249 190 W, 008 409 320 S; altitude
230 m), Provincia de Orellana, Ecuador, on
7 May 2002.
Paratopotypes.—QCAZ 19534, adult female
collected by Samael Padilla on December 3,
2001; QCAZ 20713 (field no. 10142), subadult
male collected by Italo Tapia and SSR on
7 May 2002.
Paratypes.—Ecuador: Provincia de Orellana:
Tiputini Biodiversity Station, Universidad
San Francisco de Quito (768 109
190 W, 008
379 050 S, 190–250 m): DFCH-USFQ 0443,
adult male collected by DFCH and Andre
´
s
Leo
´
n on 15 February 2001; DFCH-USFQ
0559, juvenile collected by DFCH on 31 July
2001; DFCH-USFQ 0734, subadult male, and
DFCH-USFQ 0736, adult female collected by
DFCH and Andre
´
sLeo
´
n on 15 August 2002;
TCWC 90334, juvenile collected by SFM on
18 May 2004; TCWC 90335, 90338, 90339,
adult males, TCWC 90336), subadult female,
TCWC 90337, 90340, adult females, and
TCWC 90341, juvenile male collected by
192 HERPETOLOGICA [Vol. 62, No. 2

SFM on 31 July 2004; and TCWC 90342, adult
female collected by SFM on 08 Aug 2004;
Provincia de Sucumbı
´
os: Limoncocha (008 249
S, 768 379 W; altitude 220 m): KU 106967
(adult male) and 104623 (juvenile) collected by
W. E. Duellman on 22 November 1966 and 16
August 1966, respectively; KU 123402, juve-
nile collected by M. L. Crump on 14 July 1968;
Santa Cecilia (008 039 N, 768 589 W; altitude
340 m): KU 148902, 148906, adult males
collected by M. L. Crump on 05 October
1971 and 25 February 1972, respectively.
Peru: Departamento de Loreto: Explornapo
Lodge, junction between Rı
´
o Sucusari and Rı
´
o
Napo (038 169 S, 728 549 W): KU 220426, adult
female collected by WL on 17 June 1992.
Diagnosis.—A member of the Eleutherodac-
tylus unistrigatus group, lacrimosus assemblage
FIG. 1.—Species in the Eleutherodactylus lacrimosus assemblage. (A) Eleutherodactylus aureolineatus, QCAZ 19534,
female, SVL 27.8 mm (photo by SRR); (B) E. boulengeri KU 169055, female, SVL 30.6 mm (William E. Duellman); (C) E.
eremitus, KU 165884, male, SVL 19.7 mm (WED); (D) E. lacrimosus, KU 119513, female, SVL 24.5 mm (John D.
Lynch); (E) E. mendax, KU 162290, SVL 19.6 mm (WED); (F) Eleutherodactylus petersorum, KU 165993, juvenile
female, SVL 20.7 mm (WED). Color photographs are available at AmphibiaWeb (http://elib.cs.berkeley.
edu/aw/).
June 2006] HERPETOLOGICA 193

(as defined by Lynch and Duellman, 1980,
1997; Lynch and Ruı
´
z-Carranza, 1985) having
(1) skin texture of dorsum finely shagreen, that
on the venter areolate (sensu Lynch and
Duellman, 1997); discoidal fold low; dorsolat-
eral folds absent; (2) tympanic membrane not
differentiated in males and slightly evident in
females; tympanic annulus evident, with
supratympanic fold obscuring upper and
posterodorsal edges, horizontal diameter of
tympanum 31–39% of eye diameter; (3) snout
subacuminate in dorsal view, protruding in
profile; (4) upper eyelid lacking tubercles or
with one small tubercle; cranial crests absent;
(5) dentigerous process of the vomer tri-
angular, each bearing 0–6 teeth; (6) males
with vocal slits and median subgular vocal sac;
white nuptial pads present; (7) first finger
shorter than the second; Fingers III–IV
bearing rounded discs about twice as wide as
digits; (8) fingers with narrow lateral fringes;
(9) antebrachial tubercle evident, ulnar
tubercles low or absent; (10) tarsal tubercles
low or absent; (11) inner metatarsal tubercle
oval, two-to-three times as long as round outer
metatarsal tubercle; supernumerary plantar
tubercles low, at the base of Toes III and IV;
(12) toes with narrow lateral fringes; webbing
absent; fifth toe much longer than third; (13) in
life, males with dorsum olive green with an
interorbital golden yellow stripe that extends
above the eyes and dorsolaterally to the
sacrum, and females with dorsum olive brown
to dark brown with an interorbital creamy
yellow stripe that extends above the eyes and
dorsolaterally to the sacrum; venter bright
yellow to greenish cream; posterior surfaces of
thighs olive green; (14) adults small, SVL in
males 19.7–28.8 mm (

x 5 23.6 6 2.77, n 5 9),
in females 26.3–30.5 mm (

x 5 27.8 6 1.86,
n 5 7); males with white testes (mesorchium).
Comparison with similar species.—
Eleutherodactylus aureolineatus differs from
all other species of the E. lacrimosus assem-
blage (i.e., E. apiculatus, E. boulengeri, E.
brevifrons, E. bromeliaceus, E. dorsopictus, E.
eremitus, E. lacrimosus, E. mendax, E. peter-
sorum, E. prolixodiscus, E. schultei, E. tayr-
FIG. 2.—Species in the Eleutherodactylus lacrimosus assemblage. (A) Eleutherodactylus prolixodiscus, KU 132733,
male, SVL 25.1 mm (photo by Stephan R. Edwards); (B) E. schultei, KU 212222, male, SVL 27.4 mm (William E.
Duellman); (C) E. zimmermanae, amplectant pair, NMW 32112:2–3 (Walter Ho
¨
dl); (D) E. bromeliaceus, KU 212214,
male, SVL 22.5 mm (WED). Color photographs are available at AmphibiaWeb (http://elib.cs.berkeley.edu/aw/).
194 HERPETOLOGICA [Vol. 62, No. 2

ona, E. zimmermanae) by its olive-green dorsal
coloration with a creamy-yellow interorbital
stripe that extends above the eyes and
dorsolaterally to the sacrum (Figs. 1, 2).
Additionally, Eleutherodactylus aureolineatus
differs from the sympatric E. lacrimosus by
lacking tubercles on dorsum (few tubercles
present in E. lacrimosus) and by being larger
(see Table 1). Differences among species in
the E. lacrimosus assemblage are summarized
in Table 1.
Description of holotype.—Adult male
(QCAZ 20712) with head as wide as long;
snout subacuminate in dorsal view and pro-
truding in lateral view, relatively short (snout–
eye distance 16.9% SVL), with small papilla at
tip (Fig. 3); in lateral view, canthus rostralis
distinct; loreal region slightly concave; nostrils
protuberant, directed dorsolaterally; interor-
bital area flat, broader than upper eyelid
(upper eyelid width 61.8% interorbital dis-
tance); cranial crests absent; upper eyelid
lacking tubercles; tympanic membrane not dif-
ferentiated from surrounding skin; tympanic
annulus distinct, round, with supratympanic
fold obscuring upper and posterodorsal edges
of annulus (Fig. 3); tympanum diameter 34.4%
of eye diameter; postrictal tubercles absent.
Choanae small, nearly elliptical, not concealed
by palatal shelf of maxillary; dentigerous pro-
cess of the vomer triangular, widely separated,
posteromedial to choanae, each bearing two
or five small teeth (bilateral variation within
individual); shortest distance between dentig-
erous processes about 36% distance between
choanae; tongue longer than wide, granular,
with small notch in posterior border.
Texture of skin of dorsum and flanks
shagreen, dorsolateral folds absent; venter
areolate; discoidal fold low; cloacal sheath
absent. Forearm slender; radio–ulna length
20.6% SVL; ulnar tubercles and ulnar fold
absent; hand length longer than radio–ulna
length (hand length 26.7% SVL); fingers with
narrow lateral fringes; relative lengths of
fingers I , II , IV , III; palmar tubercle
incompletely bifid, thenar tubercle oval
(Fig. 4); subarticular tubercles round, prom-
inent; supernumerary palmar tubercles pres-
ent at the base of all fingers (Fig. 4); disc cover
of Finger I slightly expanded, those of Fingers
II–IV extensively expanded (Fig. 4); outer
discs of fingers as wide as those of toes; all disc
covers with elliptical ventral pads defined by
circummarginal grooves.
Hind limbs relatively slender; tibia length
53.1% SVL; foot length 82.9% of tibia length;
tarsal fold and tarsal tubercles absent; heel
(tibiotarsal articulation) lacking tubercles; toes
with narrow lateral fringes (Fig. 4); subartic-
ular tubercles round, prominent; inner meta-
tarsal tubercle oval, about 2.4 3 the size of
subconical outer tubercle; supernumerary
plantar tubercles low, at the base of toes
(Fig. 4); disc covers slightly expanded; toes
with defined pads; disc pads nearly elliptical;
relative lengths of toes I , II , III , V , IV
(Fig. 4); tip of Toe V reaching proximal border
of distal subarticular tubercle of Toe IV; tip of
Toe III reaching distal border of medial
subarticular tubercle of Toe IV.
Coloration of holotype in life.—Dorsum
light olive green with an interorbital creamy
yellow stripe that extends above the eyes
and dorsolaterally to the sacrum, dorso-lateral
portion of stripe faint; surfaces of hind limbs
and forearm olive green; dorsal surface of arm
greenish yellow; flanks olive green with lighter
flecks. Venter bright yellow; throat greenish
cream, becoming pale gray towards the jaw;
ventral surfaces of hind limbs greenish salmon;
tibiofibula externally visible, white; sides of
head olive green, becoming greenish cream
towards the lips. Iris reddish bronze with
reddish brown median horizontal streak (S. R.
Ron, field notes, 8 May 2002).
Coloration of holotype in ethanol.—Dorsum
brown with an interorbital whitish stripe that
extends above the eyes and dorsolaterally to
the sacrum, posterior portion of stripe faint;
surfaces of limbs brown; flanks light brown;
groin cream. Throat, venter, and underside of
limbs cream; palms and soles cream with some
brown pigmentation on external fingers and
toes.
Measurements of holotype (mm).—SVL 5
24.3; tibia length 5
12.9; foot length 5 10.7;
head length 5 9.4; head width 5 9.3; upper
eyelid width 5 2.1; interorbital distance 5 3.4;
eye diameter 5 3.2; eye-to-nostril distance 5
2.9; snout-to-eye distance 5 4.1; tympanum
diameter 5 1.1; eye-to-tympanum distance 5
1.1; internarial distance 5 1.8; radioulna legth 5
5.0; hand length 5 6.5; finger I length 5 4.3;
finger II length 5 4.4. For morphometric
variation, see Table 2.
June 2006] HERPETOLOGICA 195

TABLE 1.—Diagnostic morphological characters in species of the Eleutherodactylus lacrimosus assemblage. For E. lacrimosus, SVL is given for lowland (,800 m) and highland
(900–1100 m) populations (adults only).
SVL males SVL females Tubercles on upper eyelid Ulnar tubercles Tarsal tubercles Dorsal coloration in life Source
E. apiculatus 17.8–21.8 21.6–26.3 One small
nonconical
tubercle
Absent Present,
but indistinct
Tan or yellowish tan with or
without irregular shaped
blotches; some with
middorsal yellow stripe or
fine yellow stripes
Lynch and
Duellman,
1997;
this work
E. aureolineatus 19.7–28.8 26.3–30.5 Absent or one
low tubercle
present
Low or
absent
Low or absent Olive green with creamy
yellow interorbital bar and
dorsolateral stripes that
extend posterolateraly
to the level of sacrum
This work
E. boulengeri 18.6–25.6 27.3–33.8 Present,
numerous and
nonconical
Present Low or absent Orange-tan to brown with
darker spots or marks;
some with dark
dorsolateral stripes;
interorbital bar usually
present
Lynch, 1981;
this work
E. brevifrons 15.1–19.7 21.2–25.0 Low or absent Absent or
present,
but indistinct
Low or absent Pale yellow to brown with
indistinct to bold brown
markings washed with green;
some with dorsalateral stripes
Lynch, 1981;
this work
E. bromeliaceus 16.7–23.2 22.9–28.1 Two or three
small tubercles
Absent Present, low White with or without spots
and/or reticulation
Lynch, 1979
E. dorsopictus 19.0–22.0 21.6–31.6 ? Present, small Low or absent ? Rivero and
Serna, 1988
‘‘1987’’
E. eremitus 17.2–21.8 27.1–27.6 One conical
tubercle or
several small
tubercles
Absent or
present,
but indistinct
Present, small Green, usually with darker
markings; some with black
spots or redish brown
dorsaleteral stripes
Lynch and
Duellman,
1997; this work
E. lacrimosus Low: 16.1–20.0
High: 19.6–23.7
Low: 20.6–24.4
High: 24.4–32.5
Absent Absent Absent Pale golden brown with or
without brown spotting
Lynch and
Duellman, 1980
E. mendax 19.4–21.7 22.6–28.0 Low or absent Present Present Green with various markings;
some with orange-tan or
yellow-tan dorsolateral
stripes
Duellman, 1978a;
this work
E. olivaceus 17.8–20.7 — Two or three
small tubercles
Present, small Present, small Olive green with black spots
and markings; cream
interorbital bar present
Ko
¨
hler et al., 1998
196 HERPETOLOGICA [Vol. 62, No. 2

Variation.—In some specimens, the snout
is round in dorsal view and lacks the papilla
at tip (QCAZ 20713, KU 123402, DFCH-
USFQ 0559). In one female (DFCH-USFQ
0736), the papilla is poorly defined. On
the inner margin of the tarsus of most
individuals (KU 104623, 106967, 123402,
148902, 148906, DFCH-USFQ 0734, 0443,
TCWC 90335–42), three low tubercles are
evident under magnification. The absence of
tarsal and ulnar tubercles in some of the
specimens of Eleutherodactylus aureolineatus,
including the holotype, may be a preservation
artifact. We have observed that low tubercles
are distinguishable in living individuals, but
seem to be absent when examining preserved
material.
The number of teeth on the dentigerous
processes of the vomers is variable; some
specimens (KU 148906, QCAZ 19534, 20713,
TABLE 1.—Continued
SVL males SVL females Tubercles on upper eyelid Ulnar tubercles Tarsal tubercles Dorsal coloration in life Source
E. petersorum 14.5–19.9 20.3–23.1 One conical
tubercle
Absent Absent Usually uniform pale green to dull
green; some tan or brown with
or without brown markings
Lynch and
Duellman, 1980
E. prolixodiscus 19.3–25.4 23.0–26.7 Low or absent Present, low Absent Dorsum with distinct green patch;
canthal and postocular stripe
(ending about sacrum) present
Lynch, 1978;
this work
E. schultei 23.5–26.6 28.4–34.0 Numerous, low Present, low Numerous,
low
Tan, reddish brown, or green,
bordered or not by narrow dark
brown line from snout to point
above vent
Duellman, 1990
E. tayrona 15.3–25.1 22.6–30.2 Low or absent Low or absent Absent Reddish-brown; some with cream
to dull orange vertebral stripe
Lynch and
Ruı
´
z-Carranza,
1985; this work
E. zimmermanae 19.1–21.2 22.4–25.8 Several tubercles Present Present Light to medium brown with
darker markings; yellow
interorbital bar present or absent
Heyer and
Hardy, 1991
FIG. 3.—Dorsal (A) and lateral (B) views of head of
holotype of Eleutherodactylus aureolineatus, QCAZ
20712, SVL 5 24.3 mm.
June 2006] HERPETOLOGICA 197

DFCH-USFQ 0734, TCWC 90335, 90337,
90338, 90340, 90342) have 5 or 6 teeth, one
specimen (TCWC 90339) has 4 teeth, whereas
others (KU 104623, 106967, 123402, DFCH-
USFQ 0443, 0736, TCWC 90336, 90341) have
only 1 to 3 or lack them altogether (KU
148902, TCWC 90334). Two to three small
ulnar tubercles are present in CRG 931,
TCWC 90335, 90337–40, 90342. In life, males
have a pale (DFCH-USFQ 0443, TCWC
90339) or dark (QCAZ 20713, DFCH-USFQ
0734, TCWC 90335, 90338, 90341) olive-
green dorsum; the interorbital stripe is bright
golden to yellow during night and pale golden
during day, and it extends above the eyes and
dorsolaterally down to the sacrum or to mid-
flanks; coloration of ventral surfaces of arms
and legs is similar to that on venter. Females
have a darker coloration, with olive brown
to dark brown dorsal coloration and the
FIG. 4.—Ventral view of hand and foot of Eleuther-
odactylus aureolineatus. (A) Hand, QCAZ 19534, female;
(B) foot, QCAZ 19534, female; (C) hand and forearm,
arrows indicate low ulnar tubercles, TCWC 90338, male;
(D) foot and tarsus, arrows indicate low tarsal tubercles,
TCWC 90338, male.
TABLE 2.—Measurements (in mm) and morphological
proportions (in percentages) of adult males and females
of Eleutherodactylus aureolineatus (range over mean 6
standard deviation).
Males n 5 9 Females n 5 7
SVL 19.7–28.8 26.3–30.5
23.6 6 2.77 27.5 6 1.58
Tibia length 11.5–15.3 13.5–15.4
12.8 6 1.33 14.5 6 0.68
Foot length 9.2–12.6 11.3–12.5
10.4 6 1.24 11.9 6 0.40
Head length 8.6–11.1 10.1–10.8
9.1 6 0.87 10.4 6 0.22
Head width 8.7–12.0 10.2–12.2
9.56 1.19 11.1 6 0.70
Upper-eyelid width 2.1–2.9 2.3–2.9
2.3 6 0.29 2.5 6 0.21
Interorbital distance 2.7–4.3 3.3–4.3
3.2 6 0.56 3.7 6 0.33
Eye diameter 2.7–4.1 3.3–4.0
3.25 6 0.42 3.6 6 0.27
Eye-to-nostril distance 2.6–3.6 3.0–3.4
2.9 6 0.34 3.3 6 0.15
Snout-to-eye distance 3.9–5.2 4.2–5.0
4.2 6 0.45 4.7 6 0.29
Tympanum diameter 1.0–1.5 1.2–1.5
1.2 6 0.18 1.3 6 0.15
Eye-to-tympanum distance 0.7–1.6 0.9–1.6
1.1 6 0.34 1.2 6 0.27
Internarial distance 1.7–2.5 2.0–2.7
1.9 6 0.29 2.2 6 0.25
Radioulna length 4.9–6.7 5.7–6.4
5.35 6 0.62 6.0 6 0.27
Hand length 5.8–7.9 7.4–8.0
6.5 6 0.74 7.6 6 0.25
Finger I length 2.8–4.85 3.9–5.3
4.0 6 0.66 4.9 6 0.46
Finger II length 4.0–5.5 5.1–5.7
4.8 6 0.76 5.3 6
0.30
Tibia length/SVL 50.4–58.4 48.8–58.2
Foot length/SVL 40.9–46.7 41.0–46.2
Head width/SVL 38.3–44.2 38.6–42.5
Head length/SVL 36.8–44.7 33.8–39.5
Radioulna length/SVL 20.6–25.9 18.7–24.2
Upper eyelid width/
Interorbital distance 60.0–86.2 60.5–81.8
Tympanum diameter/
Eye diameter 34.4–43.3 34.3–42.1
Finger I length/Finger
II length 88.7–97.7 90.1–96.6
198 HERPETOLOGICA [Vol. 62, No. 2

interorbital stripe is creamy yellow (DFCH-
USFQ 0736, TCWC 90337, 90340, 90342),
coloration of ventral surfaces of arms and legs
are dark brown, unlike venter coloration.
Venter coloration varies from bright yellow
(QCAZ 20712, DFCH-USFQ 0736, 443,
TCWC 90335) and pale cream (DFCH-USFQ
0734, TCWC 90336, 90339, 90341), to green-
ish creamy (QCAZ 20713, TCWC 90337,
90338, 90340, 90342). Posterior surfaces of
thighs vary from olive green to dark brown.
Juvenile (DFCH-USFQ 0559) with dark
brown dorsum, ventral surfaces white bluish,
posterior surfaces brown with indistinct darker
transversal bands and flecks, cream interorbit-
al stripe, iris dark brown. In preservative, the
soles and palms have a pale brown coloration
(CRG 931; DFCH-USFQ 0736); the dorsum is
creamish brown (KU 148906), reddish
brown (DFCH-USFQ 0443) or dark brown
(DFCH-USFQ 0736) with cream dorsolateral
stripes.
Distribution and natural history.—
Eleutherodactylus aureolineatus has been
found at localities below 350 m in elevation,
in the upper Amazon Basin of eastern Ecuador
and Peru (Fig. 5). The vegetation type at the
Ecuadorian localities is Amazonian Evergreen
Lowland Forest (Palacios et al., 1999). The
YSRS and the TBS are located on the bank of
the Rı
´
o Tiputini at elevations between 190–
250 m. Limoncocha is a oxbow lake on the
north bank of the Rı
´
o Napo at an elevation of
220 m. At these localities, the vegetation is
composed of terra firme forest (nonflooded
forest), varzea (flooded forest), and seasonally
flooded forest. The vegetation type and climate
at Santa Cecilia (Provincia de Sucumbı
´
os, 340 m)
have been described in detail by Duellman
(1978b). On 7 May 2002, the holotype and one
subadult male (QCAZ 20713) were collected
by day in a tank bromeliad, at a height of 25 m,
on a tree in terra firme forest. In the same
bromeliad, an adult male and at least 10
tadpoles of Osteocephalus planiceps (e.g.,
QCAZ 20873) were found. On 7 May 2002,
in a tree hole 7 m below the same bromeliad,
an adult male and female and several tadpoles
(QCAZ 20875) of Phrynohyas resinifictrix
were found; the next day, an additional adult
female of P. resinifictrix was collected in the
same hole (S. R. Ron, fieldnotes). An adult
female E. aureolineatus (QCAZ 19534) was
collected by night over a window on a building
of the YSRS on 3 December 2001. An adult
male E. aureolineatus (DFCH-USFQ 0443)
was collected while calling by night over a leaf
ca. 45 m height on a Ceiba tree in terra firme
forest. A subadult male (DFCH-USFQ 0734)
and an adult female (DFCH-USFQ 0736)
were collected by day inside a bromeliad ca.
40 m height on a tree in terra firme forest
(D. F. Cisneros-Heredia, fieldnotes). A female
E. aureolineatus (not collected) was observed
during day inside a bromeliad ca. 40 m height,
together with an adult Dendrobates ventrima-
culatus (D. F. Cisneros-Heredia, fieldnotes). A
juvenile E. aureolineatus (TCWC 90334) was
collected by night ca. 30 m height on a handrail
of a wooden canopy tower surrounding a Ceiba
tree in terra firme forest at TBS on 18 May 04;
two adult males (TCWC 90335, 90338), one
adult female (TCWC 90337) and one juvenile
female (TCWC 90336) were collected by day
inside a bromeliad ca. 38 m height on a tree in
terra firme forest during bromeliad patch
sampling on 31 July 2004; on the same tree,
an adult male (TCWC 90339), an adult female
(TCWC 90340) and a juvenile male (TCWC
90341) were collected by day inside a brome-
liad ca. 35.5 m height on 31 July 2004 (S. F.
McCracken, fieldnotes). An adult female
(TCWC 90342) was collected by day inside
a bromeliad ca. 23.5 m height on a tree in terra
firme forest during bromeliad patch sampling
on 8 Aug. 2004; on the same tree, an E.
aureolineatus (not collected) was observed by
day inside a bromeliad ca. 23.5 m height on
FIG. 5.—Distribution of Eleutherodactylus aureolinea-
tus (circles). Map modified from Torres-Carvajal (2005).
June 2006] HERPETOLOGICA 199

8 Aug. 2004 (S. F. McCracken fieldnotes). Two
males E. aureolineatus (KU 148902, 148906)
were found at night on low vegetation (W. E.
Duellman, fieldnotes); one juvenile (KU
123402) was found on forest floor by day (M.
L. Crump, fieldnotes), while another juvenile
(DFCH-USFQ 0559) was found in low
vegetation (50 cm) by night (D. F. Cisneros-
Heredia, fieldnotes). Twelve out of fifteen
adults with ecological data have been found at
heights .20 m suggesting that E. aureolinea-
tus is predominantly a canopy dweller.
Etymology.—The specific name aureolinea-
tus comes from the Latin words aureus
meaning golden, and linea meaning line.
The name is used in reference to the dorso-
lateral golden stripe characteristic of the new
species.
Remarks.—Most of the diversity of the
Eleutherodactylus lacrimosus assemblage is
found in the cloud forest of the Andes. Only
five species (E. aureolineatus, E. bromeliaceus,
E. lacrimosus, E. mendax, and E. zimmerma-
nae) are known from the Amazonian lowlands
(,1000 m; see IUCN et al., 2004).
D
ISCUSSION
Neglect of Canopy in
Amphibian Inventories?
Most amphibian inventories in tropical rain-
forests are carried out at ground level, typically
along shallow strata (only ;2 m), characterized
by higher humidity and lower ambient tem-
perature than other microhabitats. Therefore,
accounts of the diversity, community compo-
sition, and abundance of amphibians in these
structurally complex forests are biased toward
a relatively localized environmental space.
Eleutherodactylus aureolineatus and other
members of the lacrimosus assemblage are
good examples of species that can be over-
looked as a consequence of microhabitat
sampling bias.
This bias could be influential in inventories
on which visual encounter surveys (VES),
transect surveys, or quadrats are the pre-
dominant sampling methods. An example is
given by inventories carried out in the YSRS
and at the TBS. At the YSRS, between 1995
and 2003, a total of 1540 specimens, belong-
ing to 72 amphibian species, have been
collected (QCAZ database). Despite the large
sampling effort, only one E. aureolineatus
(QCAZ 19534) has ever been collected
below the canopy. The only additional E.
aureolineatus known from this locality (QCAZ
20712–13) were collected during sporadic
canopy searches (especially in tank bro-
meliads) carried out during one week in May
2002.
At the TBS, between 1998 and 2001, almost
600 specimens, belonging to 87 species of
amphibians were collected (Cisneros-Heredia
2003, D. F. Cisneros-Heredia, unpublished
data). All paratypes (except for one juvenile
found near the floor, DFCH-USFQ 0559)
were collected in the canopy. Aside from the
type series, nine additional specimens were
also collected during the yearly canopy surveys
(these specimens were not preserved). Also, E.
aureolineatus seems to be an abundant species
in the canopy; bromeliad patch sampling of 40
bromeliads among eight trees resulted in the
collection of eight specimens.
The specimens of Eleutherodactylus aureo-
lineatus from the YSRS represent 13% of the
total number found in the canopy searches,
and the specimens from the TBS represent
27%. These results suggest that E. aureolinea-
tus is a proportionally more abundant anuran
within higher strata of the forest. In addition,
at YSRS, the canopy searches yielded 13
specimens belonging to 6 species: Dendro-
bates aff. ventrimaculatus, Nyctimantis rugi-
ceps, Osteocephalus deridens, O. fuscifacies,
O. planiceps, and Phrynohyas resinfictrix.At
the TBS, the canopy searches yielded 32
records belonging to 13 species: Hyalinoba-
trachium munozorum, Cochranella sp., Den-
drobates ventrimaculatus, D. duellmani,
Eleutherodactylus lacrimosus, Gastrotheca
longipes, Nyctimantis rugiceps, Osteocephalus
deridens, O. planiceps, O. taurinus, Phryno-
hyas coriacea, P. resinifictrix
, and Sphaeno-
rhynchus carneus. Bromeliad patch sampling
yielded 11 specimens belonging to two addi-
tional species: Dendrobates ventrimaculatus
and Eleutherodactylus sp. Similarly to E.
aureolineatus, all these species found in
canopy situation have been extremely rare
during VES, and three (Osteocephalus deri-
dens, O. fuscifacies, Dendrobates duellmani)
were undescribed until recently (Jungfer et al.,
2000; Schulte, 1999); the scarcity of these
200 HERPETOLOGICA [Vol. 62, No. 2

species in herpetological collections is likely
an artifact of survey methodology.
Although our sampling effort was limited,
the results are consistent with the suggestion
that canopy surveys can improve species
richness assessments and inventory efficiency
in tropical rainforests (e.g., Longino and
Colwel, 1997; Kalko and Handley, 2001).
Sampling methodologies that distribute cap-
ture effort more evenly among vertical strata of
rainforests have the potential to enhance
significantly our understanding of tropical
amphibian communities.
Acknowledgments.—For comments on this manuscript,
we thank R. Brown. Fieldwork at Yasunı
´
was assisted by
I. G. Tapia. Access to KU and TCWC specimens was
provided by L. Trueb and L. Fitzgerald, respectively.
Loans from QCAZ were arranged by L. A. Coloma.
Research by JMG is partially supported by The University
of Kansas, Fundacio
´
n Numashir para la Conservacio
´
nde
Ecosistemas Amenazados, and a fellowship from the
Fundacio
´
n para la Ciencia y Tecnologı
´
a del Ecuador
(FUNDACYT) under the sponsorship of the Departa-
mento de Ciencias Biolo
´
gicas of the Pontificia Uni-
versidad Cato
´
lica del Ecuador. Research by DFCH
was partially supported by Universidad San Francisco
de Quito, Tiputini Biodiversity Station, the 2002 Re-
search Training Program, National Museum of Natural
History, Smithsonian Institution, Smithsonian Women’s
Committee, and Ma. Elena and L. Heredia. Research
by SFM is partially supported by Texas State University –
Department of Biology and the TADPOLE Organization.
LITERATURE CITED
AMPHIBIAWEB. 2005. AmphibiaWeb: Information on
amphibian biology and conservation. Berkeley, Califor-
nia. Available at: ,http://amphibiaweb.org/.. Accessed
on 25 November 2005.
C
ISNEROS-HEREDIA, D. F. 2003. Herpetofauna de la
Estacio
´
n de Biodiversidad Tiputini, Amazonı
´
a Ecua-
toriana. En: De la Torre, S. y G. Reck. (Eds.). Ecologı
´
ay
Ambiente en el Ecuador: Memorias del I Congreso de
Ecologı
´
a y Ambiente. CD. Universidad San Francisco
de Quito. Quito, Ecuador.
D
UELLMAN, W. E. 1978a. Three new species of Eleuther-
odactylus from Amazonian Peru
´
(Amphibia: Anura:
Leptodactylidae). Herpetologica 34:264–270.
———. 1978b. The biology of an Equatorial herpetofauna
in Amazonian Ecuador. The University of Kansas
Museum of Natural History, Miscellaneous Publications
65:1–352.
———. 1990. A new species of Eleutherodactylus from
the Andes of northern Peru (Anura: Leptodactylidae).
Journal of Herpetology 24:348–350.
G
UAYASAMIN, J. M. 2004. A new species of Eleuthero-
dactylus (Anura: Leptodactylidae) from the north-
western lowlands of Ecuador. Herpetologica 60:
103–116.
H
EYER, W. R., AND L. M. HARDY. 1991. A new species of
frog of the Eleutherodactylus lacrimosus assembly from
Amazonia, South America (Amphibia: Anura: Lepto-
dactylidae) from the northwestern lowlands of Ecuador.
Proceedings of the Biological Society of Washington
104:436–447.
IUCN, Conservation International, and NatureServe.
2004. Global Amphibian Assessment. Available at:
http://www.globalamphibians.org. Accessed on 21 Oc-
tober 2005.
J
UNGFER, K. S., R. RON,R.SEIPP,A.ALMENDA
´
RIZ, AND
G. KO
¨
HLER. 2000. Two new species of hylid frogs, genus
Osteocephalus, from Amazonian Ecuador. Amphibia &
Reptilia 21:327–340.
K
ALKO, E. K. V., AND C. O. HANDLEY. 2001. Neotropical
bats in the canopy: diversity, community structure, and
implications for conservation. Plant Ecology 153:
319–333.
K
O
¨
HLER, J., V. R. MORALES,S.LO
¨
TTERS,S.REICHLE, AND
J. APARICIO. 1998. A new green frog, genus Eleuther-
odactylus, from Bolivia and Peru
´
(Amphibia, Anura,
Leptodactylidae). Studies on Neotropical Fauna &
Environment 33:93–99.
L
ONGINO, J. T., AND R. K. COLWELL. 1997. Biodiversity
assessment using structured inventory: capturing the
ant fauna of a tropical rainforest. Ecological Applica-
tions 7:1263–1277.
L
YNCH, J. D. 1978. A new Eleutherodactylus frog from the
Andes of northern Colombia (Leptodactylidae). Copeia
1978:17–21.
———. 1979. Leptodactylid frogs of the genus Eleuther-
odactylus from the Andes of southern Ecuador. The
University of Kansas Museum of Natural History,
Miscellaneous Publications 66:1–62.
———. 1981. Two new species of Eleutherodactylus from
western Colombia (Amphibia: Anura: Leptodactylidae).
University of Michigan, Occasional Papers of the
Museum of Zoology 697:1–12.
L
YNCH, J. D., AND W. E. DUELLMAN. 1980. The Eleuther-
odactylus of the Amazonian slopes of the Ecuadorian
Andes (Anura: Leptodactylidae). The University of
Kansas Museum of Natural History, Miscellaneous
Publications 69:1–86.
———. 1997. Frogs of the genus Eleutherodactylus
(Anura: Leptodactylidae) in western Ecuador: system-
atics, ecology, and biogeography. The University of
Kansas Natural History Museum Special Publication 23:
1–236.
L
YNCH, J. D., AND P. M. RUI
´
Z-CARRANZA. 1985. A synopsis
of the frogs of the genus Eleutherodactylus from the
Sierra Nevada de Santa Marta, Colombia. Occasional
Papers of the Museum of Zoology, University of
Michigan 711:1–59.
P
ALACIOS, W., C. CERO
´
N,R.VALENCIA, AND R. SIERRA.
1999. Las formaciones naturales de la Costa del
Ecuador. Pp. 109–119. In R. Sierra (Ed.), Propuesta
Preliminar de Clasificacio
´
n de Vegetacio
´
n para el
Ecuador Continental. Proyecto INEFAN/GEF-BIRF
and EcoCiencia, Quito, Ecuador.
R
EAD, M. 2000. Frogs of the Ecuadorian Amazon: a guide
to their calls. Compact Disc. Morley Read Productions,
Fowey, Cornwall, U. K.
R
IVERO,J.A.,AND M. A. SERNA. 1988 ‘‘1987’’. Tres nuevas es-
pecies de Eleutherodactylus (Amphibia: Leptodactylidae)
June 2006] HERPETOLOGICA 201

de Antioquia, Colombia. Caribbean Journal of Sciences
23:386–399.
R
ON, S. R. 2001. Anfibios del Parque Nacional Yasunı
´
,
Amazonı
´
a ecuatoriana. [on line]. Ver. 1.3 (2 Marzo
2001). Museo de Zoologı
´
a, Pontificia Universidad
Cato
´
lica del Ecuador. Quito, Ecuador. Available at:
http://www.bio.utexas.edu/grad/ecuador/web/yasuni/esp/
anfyas.htm. Accessed on 10 October 2005.
S
CHULTE, R. 1999. Pfeilgiftfro
¨
sche, Artenteil—Peru.
Nikola. Stuttgart, Germany.
T
ORRES-CARVAJAL, O. 2005. A new species of iguanian lizard
(Stenocercus) from the western lowlands of southern
Ecuador and northern Peru. Herpetologica 61:75–85.
Accepted: 19 December 2005
Associate Editor: Joseph Mendelson
APPENDIX I
Specimens Examined
Eleutherodactylus apiculatus (KU 212534–38, para-
types). Eleutherodactylus boulengeri (KU 169056–61,
paratypes). Eleutherodactylus brevifrons (KU 169006–13,
paratypes). Eleutherodactylus bromeliaceus (KU 174524–
25, paratypes). Eleutherodactylus eremitus (KU 140878,
179085–86; paratypes). Eleutherodactylus lacrimosus (KU
110782, neotype; 148889–908, 123391–98, DFCH-USFQ
D50). Eleutherodactylus mendax (KU 173234, holotype;
171897, paratype). Eleutherodactylus petersorum (KU
143508, holotype). Eleutherodactylus prolixodiscus (KU
132727–33, paratypes). Eleutherodactylus schultei
(KU 209498–504, paratypes).
Herpetologica, 62(2), 2006, 202–220
Ó 2006 by The Herpetologists’ League, Inc.
THREE NEW SPECIES OF COPHIXALUS (ANURA:
MICROHYLIDAE) FROM SOUTHEASTERN NEW GUINEA
FRED KRAUS
1,2
AND ALLEN ALLISON
1
1
Bernice P. Bishop Museum, 1525 Bernice Street, Honolulu, HI 96817, USA
A
BSTRACT: We describe three new species of a cryptic species complex of Cophixalus from southeastern
New Guinea and adjacent islands. They are readily distinguished from all other Papuan species by their small
size, presence of vocal slits in males, absence of toe webbing, relatively long legs, short snouts, and vertical
lores. Their calls consist of a series of high-pitched peeps. Intraspecific color pattern is highly variable and
variant patterns are shared among species. Members of this complex can be distinguished from each other on
the basis of adult body size, extent of development of the first finger and disc, snout width, tympanum size, and
advertisement calls. The new species are among the most common frogs where found. Two of the species are
currently known only from their type localities, but the third ranges more broadly in the Southeast Peninsula
of New Guinea and D’Entrecasteaux Islands.
Key words: Anura; Cophixalus; Microhylidae; New Guinea; New species
IN THE COURSE of conducting herpetological
surveys in Milne Bay Province, Papua New
Guinea, during 2002, we found an undescribed
species of Cophixalus to be among the most
common components of the anuran commu-
nity at a number of locations. It has been known
for approximately 30 years and was featured
in Menzies (1975:pl. 12C) but has not re-
ceived a name or formal description. Menzies
(1975:60) further noted that a second, similar
species might replace this species at higher
elevations. In expanding our initial surveys to
additional localities, apparent differences in
call parameters lent support to Menzies’ sup-
position that more than one species was
involved, although morphologically they are
virtually impossible to tell apart in the field.
Additional collections and recordings of similar
animals from Central Province made it clear
that the situation was more complicated than
suggested by Menzies. It is now apparent that
the frog referred to as ‘‘Cophixalus species’’ by
Menzies (1975:59–60) is actually a complex of
cryptic species. Here we describe three of
these species but note that additional members
may remain to be discovered. This brings the
2
CORRESPONDENCE: e-mail, fkraus@hawaii.edu
202 HERPETOLOGICA [Vol. 62, No. 2

total number of Papuan members of the genus
to 20 and the total for the genus (ranging from
Halmahera to Australia) to 34 (Gu¨ nther, 2003;
Kraus and Allison, 2000).
M
ATERIALS AND METHODS
Specimens were fixed in 10% buffered
formalin and transferred to 70% ethanol for
storage. All measurements were made with
digital calipers or an optical micrometer to the
nearest 0.1 mm, with the exception that disc
widths were measured to the nearest 0.01 mm.
Measurements, terminology, and abbrevia-
tions follow Zweifel (1972, 1985) and Kraus
and Allison (2001): body length from snout–
vent (SVL); tibia length from heel to skin fold
of knee (TL
fold
); tibia length from heel to outer
surface of flexed knee (TL
knee
); diameter of
eye (EY); distance from anterior corner of eye
to center of naris (EN); internarial distance,
between centers of external nares (IN); dis-
tance from anterior corner of eye to tip of
snout (SN); head width at widest point, typi-
cally at the level of the tympani (HW); head
length, from tip of snout to posterior margin
of tympanum (HL); horizontal tympanum di-
ameter (TY); hand length from proximal edge
of palm to tip of 3
rd
finger (HandL); foot length
from proximal edge of sole to tip of 4
th
toe
(FootL); width of the third finger disc (3
rd
F);
width of the fourth toe disc (4
th
T). We also
introduce a new measure, length of the first
finger, from the most proximal margin of web-
bing between the first and second fingers to
the tip of the disc (1
st
F). We provide two
measures of TL to facilitate comparison with
all earlier studies of the genus, which have
generally evolved from using TL
fold
(e.g.,
Zweifel, 1956) to TL
knee
(e.g., Zweifel, 1985).
We recorded calls in the field using a
Sennheiser K6 microphone and either a Sony
Professional Walkman WM-D6C cassette
recorder or a Sony MDSJE480 minidisc re-
corder. Call structure was analyzed using the
computer program Avisoft-SASLab Pro(v4.34).
We confirmed by dissection generic assign-
ment of the frogs using the presence of an
eleutherognathine jaw, absence of clavicles
and procoracoids, presence of expanded finger
discs larger than toe discs, fifth toe shorter
than the third, M. depressor mandibulae aris-
ing mostly from the dorsal musculature, and
alary process of premaxilla projecting dorsally
(Burton and Zweifel, 1995; Menzies and Tyler,
1977; Parker, 1934; Zweifel and Parker, 1989).
Type specimens are deposited in the Bernice
P. Bishop Museum, Honolulu (BPBM) and
Papua New Guinea National Museum and Art
Gallery, Port Moresby (PNGNM).
Pairwise statistical comparisons of SVL
means utilized Student’s t-test in SigmaStat,
version 2.0. All other statistical comparisons
involved general linear model ANOVA to
determine if two-parameter regressions (most
commonly one morphometric parameter
against SVL) varied among samples. Signifi-
cant pairwise differences were detected using
Tukey’s test. These tests were performed using
Minitab 14. To save space, we summarize
morphometric differences among samples or
species with simple bivariate ratios. In the
accompanying tables we highlight those ratios
whose constituent variables have significantly
different relationships among strata (sexes,
geographic samples, species) in the ANOVA
tests. A posteriori canonical variates and
discriminant function tests were performed
using PAST, version 1.32, to evaluate how well
the presumptive species could be discrimi-
nated from each other in multivariate space.
Unless otherwise noted, all latitude and
longitude coordinates use the Australian Geo-
detic Datum, 1966 (AGD 66).
Cophixalus variabilis sp. nov.
Holotype.—BPBM 20165 (field no. FK
10633), adult male (Figs. 1, 2), collected by
F. Kraus on SE slope Mt Pekopekowana,
10.28518 S, 150.18228 E (WGS 84), 330 m,
Milne Bay Province, Papua New Guinea, 24
May 2004.
Paratypes.—Papua New Guinea: Milne Bay
Province: near Upaelisafupi Stream, Cloudy
Mts, 10.49718 S, 150.23308 E (WGS 84), 715 m,
10–16 April 2002 (BPBM 15772–803); same
locality, 11–20 April 2002 (PNGNM 23747–
51, 23838–47); Cloudy Mts, 10.50798 S,
150.22278 E (WGS 84), 1010 m, 19 April 2002
(BPBM 15804); Cloudy Mts, 10.50958 S,
150.22248 E, 980–1000 m, 23 April 2002
(BPBM 15805–08); same data as holotype
but 7 May 2002 (BPBM 15809–10); along
Wailahabahaba Creek, SE slope Mt Pekope-
kowana, 10.28268 S, 150.15488 E (WGS 84),
June 2006] HERPETOLOGICA 203

590–615 m, 9–10 May 2002 (BPBM 15811–14);
same locality, 14–17 May 2002 (PNGNM
23850–51); along Sagabada River, SE
slope Mt Pekopekowana, 10.28208 S, 150.14938
E, 1600 m, 12 May 2002 (BPBM 15815); SE
slope Mt Pekopekowana between 10.28158 S,
150.16168 E, 620–680 m, 13 May 2002 (BPBM
15816–21, PNGNM 23848–49); Central Prov-
ince: Moroka, 9.42908 S, 147.59328 E, 302 m, 21
March 2005 (BPBM 21886–89).
Referred specimens.—Papua New Guinea:
Milne Bay Province: E slope Oya Tabu (5 Mt
Kilkerran), Fergusson Island, 9.45768 S,
150.78888 E (WGS 84), 960 m, 23 August
2002 (BPBM 16273–74, PNGNM 23852); NE
slope Oya Tabu, Fergusson Island, 9.45888 S,
150.77878 E, 1500 m, 28 August 2002 (BPBM
16275); NE slope Oya Tabu, Fergusson Island,
9.45488 S, 150.80438 E (WGS 84), 320–360 m,
31 August 2002 (BPBM 16276–78, PNGNM
23853–54); S slope Oya Waka, Fergusson
Island, 9.45628 S, 150.55968 E (WGS 84),
980 m, 9–13 September 2002 (BPBM 16279–
85, PNGNM 23855–56); near Saidowai, Nor-
manby Island, 9.95608 S, 150.95088 E (WGS
84), sea level, 1 October 2002 (BPBM 16286).
Diagnosis.—A small (adult males 13.6–18.6
mm, adult females 14.5–19.7 mm) species of
Cophixalus having a tuberculate dorsum; well-
developed first finger (1
st
F/HandL 5 0.15–
0.28); well-developed, expanded, flattened
discs on all fingers and toes; toes free of
FIG. 1.—(A) Cophixalus variabilis, paratype (BPBM 15772), male, Cloudy Mountains, Milne Bay Province; (B) C.
timidus, paratype (BPBM 18092), female, Mt Simpson, Milne Bay Province; (C) C. timidus, paratype (BPBM 18096),
female, Mt Simpson, Milne Bay Province; (D) C. timidus, paratype (BPBM 18104), female, Mt Simpson, Milne Bay
Province; (E) C. sisyphus, paratype (BPBM 19302), male, Mt Obree, Central Province; and (F) C. sisyphus, paratype
(BPBM 19303), male, Mt Obree, Central Province.
204 HERPETOLOGICA [Vol. 62, No. 2

webbing; vocal slits in males; sexual mono-
morphism in tympanum size (TY/SV 5 0.035–
0.075 in males, 0.045–0.071 in females);
moderately long legs (TL
knee
/SVL 5 0.46–
0.58, TL
fold
/SVL 5 0.41–0.53); a relatively
short, broad snout (EN/IN 5 0.72–1.00);
vertical lores; ventral pectoral spots usually
obscure or absent; shanks typically banded or
mottled with dark brown or black; heavily
melanized ventral surfaces in most individuals,
and a call consisting of a series of 1–31 peeps.
Cophixalus variabilis is distinguished from
C. ateles, C. bewaniensis, C. pipilans, C.
pulchellus, C. shellyi, and C. sphagnicola by
its tuberculate dorsum, variegated or striped
color pattern, and in having the first finger of
normal length, not reduced in size, with a disc;
from C. verrecundus in having discs on the
fingers and toes; from C. tagulensis in having
the toes free of webbing; from C. daymani in
having vocal slits in males, vertical lores, and
longer legs (TL
fold
/SVL 5 0.31–0.38 in C.
daymani); from C. parkeri and C. kaindiensis
in having the third toe distinctly longer that the
fifth (subequal in C. parkeri and C. kaindien-
sis); and from C. nubicola in its longer legs
(TL
fold
/SVL 5 0.35–0.38 in C. nubicola). It is
distinguished from all remaining Papuan
Cophixalus by its much smaller size and
heavily melanized ventral surfaces but also
differs from C. cheesmanae and C. verrucosus
in its short, broad (vs. pointed) snout; from
C. cryptotympanum in having vocal slits in
males and a tuberculate dorsum; and from
C. biroi in its vertical lores and more variable
color pattern.
Description of holotype.—An adult male.
Head moderately wide (HW/SVL 5 0.39),
with vertical loreal region; canthus rostralis
rounded, slightly concave when viewed from
above; nostrils directed laterally, much closer
to tip of snout than to eyes; internarial distance
considerably larger than distance from naris
to eye (EN/IN 5 0.80, IN/SVL 5 0.096, EN/
SVL 5 0.077); snout rounded and projecting
when viewed from the side, angulate when
viewed from above; eyes moderately large (EY/
SVL 5 0.12); eyelid approximately 3/4 the
width of the interorbital distance; tympanum
indistinct and small (TY/SVL 5 0.064), with-
out a clear annulus. Dorsal and lateral surfaces
smooth (probably an artifact of preservation).
Supratympanic fold absent. Ventral surfaces
generally smooth but weakly granular on
abdomen. Fingers unwebbed, bearing discs
with terminal grooves; relative lengths 3 . 4 .
2 . 1; first finger and disc well-developed.
Finger discs approximately 2.5 times widths of
penultimate phalanges, except for the first
finger disc, which is only approximately 1.5
times width of penultimate phalanx (Fig. 2A).
Articular tubercles low and indistinct; meta-
carpal tubercles poorly developed. Toes
unwebbed, bearing discs with terminal
grooves; relative lengths 4 . 3 . 5 . 2 . 1.
Toe discs smaller than those of fingers; disc of
fourth toe approximately twice width of
penultimate phalanx; disc of first toe approx-
imately same width as penultimate phalanx.
Subarticular tubercles lacking but having
thicker, lighter skin in same areas; inner
metatarsal tubercle narrow and elongate, outer
lacking. Hind legs moderately long (TL
knee
/
SVL 5 0.49, TL
fold
/SVL 5 0.44).
Dorsum with variegated color pattern.
Dorsum posterior to scapular region light
brown with orange cast and with few small
brown blotches; anterior to scapular region
chocolate brown, including top of head and
sides of face. Sides with same antero-posterior
division of colors but with band of dense dark
brown spotting posterior to forearm; anteri-
orly, from behind eye to behind forearm,
ground color light gray. Tops of limbs tan,
mottled and blotched with black or dark brown;
backs of thighs tan heavily mottled with dark
brown; ill-defined triangular dark brown patch
surrounds the anus. Short, dark brown post-
ocular line ending approximately halfway to
forearm insertion; followed by a dark brown
dash above forearm and then the band of dark
brown lateral spots. Entire dorsum covered
FIG. 2.—Palmar views of left hands of (A) Cophixalus
variabilis paratype (BPBM 15806), (B) C. timidus para-
type (BPBM 18101), and C. sisyphus paratype (BPBM
19318).
June 2006] HERPETOLOGICA 205

with minute black punctations, invisible except
under magnification. Iris dark brown. Venter
light pearl gray, densely dusted with black
punctations and flecks, densest from chin to
chest and decreasing in density posteriorly;
undersides of limbs same.
Measurements (in mm).—SVL 5 15.6,
TL
fold
5 6.9, TL
knee
5 7.6, HW 5 6.1, HL 5
5.5, IN 5 1.5, EN 5 1.2, SN 5 2.1, EY 5 1.9,
TY 5 1.0, HandL 5 3.9, FootL 5 7.6, 3
rd
F 5
1.08, 4
th
T 5 0.96.
Variation.—The range of mensural variation
in the samples is considerable (Tables 1–3).
However, consistent morphological differen-
ces between the sexes are virtually nonexistent,
judging from comparisons made within the
large sample from the Cloudy Mts. For this
sample, females are larger than males (t 5
3.76, P 5 0.001), but a significant interaction
does not occur between sex and any of the
parameter pairs used in the morphometric
ratios. Given this absence of sexual morpho-
metric differences, we pooled males and
females for our comparisons of geographic
variation among our three largest samples
(Table 2).
Body size is significantly larger in specimens
from Fergusson Island than in specimens from
the adjacent mainland (F 5 27.629, df 5 2, P ,
0.001; Fergusson vs. Cloudy Mts p 5 3, q 5
9.569, P , 0.001; Fergusson vs. Owen Stanley
Mts q 5 9.168, P , 0.001); specimens from
the two mainland localities cannot be distin-
guished (q 5 0.461, P 5 0.943). Morphometric
relationships that differ geographically are IN
vs. SVL, TY vs. SVL, HW vs. SVL, 3
rd
F vs.
SVL, 4
th
T vs. SVL, EN vs. IN, EY vs. SN, HL
vs. HW, and 1
st
F vs. HandL (Table 2). The
frogs from the Cloudy Mts have significantly
wider IN at any given SV than do those from
Fergusson Island (T 5 2.719, P 5 0.0216) and
are almost significantly different from the
sample from Owen Stanley Mts (T 5 2.254,
P 5 0.0683) but the slopes remain identical.
The sample from the Cloudy Mts has wider TY
than does that from Fergusson Island for any
given SV (T 5 4.037, P 5 0.0004) but the
slope is the same and the remaining pairwise
comparisons show no differences. The sample
from the Owen Stanley Mts has larger HW for
any given SV than does that from Fergusson
Island (T 5 3.328, P 5 0.038) but there is no
difference in slopes and all other sample
comparisons are identical. The sample from
Fergusson Island has smaller 3
rd
F at any given
SVL than do those from the Cloudy Mts (T 5
2.996, P 5 0.0101) or Owen Stanley Mts (T 5
3.683, P 5 0.0012), but there is no difference
in slopes. The sample from the Owen Stanley
Mts has significantly larger 4
th
T at any given
SVL than do those from the Cloudy Mts (T 5
3.9824, P 5 0.0004) or Fergusson Island (T 5
3.927, P 5 0.0005), but there is no difference
in slope. The sample from Fergusson Island
increases more rapidly in EN with increasing
IN than do the samples from the Cloudy Mts
or Owen Stanley Mts (F 5 3.39, df 5 2, P 5
0.039). The samples differ in the rate at which
EY increases with respect to SN (F 5 3.56,
df 5 2, P 5 0.033), with that from Fergusson
Island increasing at the most rapid rate, that
from the Owen Stanley Mts at the slowest rate,
and that from the Cloudy Mts intermediate
between these extremes. The sample from
Fergusson Island has a relatively longer HL
for any given HW compared to that from the
Owen Stanley Mts (T 5 2.7459, P 5 0.0195)
but neither differs from the sample from the
Cloudy Mts and the slopes do not differ, and it
has a relatively longer 1
st
F for any given
HandL than do the samples from the Cloudy
Mts (T 5 2.594, P 5 0.0318) or Owen Stanley
TABLE 1.—Sexual variation in mensural characters for
the Cloudy Mountains sample of Cophixalus variabilis.
Numbers in parentheses indicate sample sizes.
Males (20) Females (21)
Character Mean Range Mean Range
SVL (mm) 15.5 13.6–16.9 16.8 14.0–19.4
TLfold/SVL 0.47 0.42–0.51 0.48 0.41–0.51
TLknee/SVL 0.52 0.47–0.56 0.52 0.47–0.56
EN/SVL 0.085 0.073–0.094 0.082 0.071–0.090
IN/SVL 0.099 0.089–0.110 0.096 0.085–0.107
SN/SVL 0.13 0.12–0.15 0.13 0.11–0.15
TY/SVL 0.054 0.040–0.072 0.060 0.045–0.071
EY/SVL 0.13 0.12–0.13 0.12 0.11–0.13
HW/SVL 0.39 0.36–0.43 0.37 0.35–0.40
HL/SVL 0.33 0.31–0.34 0.32 0.26–0.34
HandL/SVL 0.28 0.25–0.32 0.28 0.25–0.32
FootL/SVL 0.47 0.43–0.53 0.48 0.43–0.53
3
rd
F/SVL 0.059 0.049–0.067 0.062 0.047–0.078
4
th
T/SVL 0.048 0.040–0.054 0.049 0.039–0.058
EN/IN 0.86 0.79–1.00 0.86 0.75–1.00
EN/SN 0.65 0.57–0.74 0.63 0.54–0.70
EY/SN 0.98 0.87–1.11 0.94 0.81–1.05
3
rd
F/4
th
T 1.2 1.0–1.5 1.3 1.1–1.4
HL/HW 0.84 0.76–0.93 0.87 0.79–0.95
1
st
F/HandL 0.21 0.18–0.26 0.22 0.19–0.25
206 HERPETOLOGICA [Vol. 62, No. 2

Mts (T 5 3.140, P 5 0.0074) but there is no
difference among the slopes.
A number of mensural characters show
apparent ontogenetic variation, with adults
and juveniles having different covariation
slopes for a number of character pairs (Table
3). Juveniles have relatively longer legs (for
TL
fold
vs. SVL, F 5 4.94, P 5 0.031; for TL
knee
vs. SVL, F 5 6.13, P 5 0.017), longer snouts
(for EN vs. IN, F 5 7.83, P 5 0.007; for EN vs.
SN, F 5 8.07, P 5 0.007), wider heads (HW
vs. SVL, F 5 9.78, P 5 0.003), and longer feet
(for FootL vs. SVL, F 5 4.65, P 5 0.0036).
Color variation is considerable. In preser-
vative, most specimens are variegated with
black and/or dark brown flecks and spots on
a light tan to dark gray ground color. Many
specimens have longitudinal striping of two
types: of 88 specimens, 24 have a light tan mid-
dorsal stripe and 16 have a lighter, tan to
russet, mid-dorsal band; a few specimens have
both, that is, a narrow mid-dorsal stripe
contained within a wider mid-dorsal band. Of
the specimens having the narrow tan stripe, 11
also have narrow tan stripes on the backs of the
thighs and shanks (n 5 8) or only the shanks (n
5 3). Thirty-nine specimens have the heels
and proximal upper surfaces of the thighs
conspicuously lighter than the surrounding
regions, producing a light, triangular posterior
patch when the animals sit at rest. Clear
lumbar ocelli are present in 26 specimens,
obscure ocelli are present in 16, and ocelli are
absent in 22. The shanks of most individuals
are mottled or banded with darker brown or
black. Venters have a straw ground color
flecked with dark gray, with the amount of
gray flecking varying considerably. In speci-
mens with the least flecking, overall ventral
coloration is a dirty light gray; most specimens
have a two-toned ventral coloration, with a dark
chin and throat followed by a lighter chest and
abdomen. One or two (usually one) cream or
dirty white pectoral spots are present on each
side of the pectoral region in many specimens.
In 15 specimens examined, these spots are
distinct, in 29 specimens obscure, and in 20
specimens absent.
Color in life.—In life, BPBM 15773 was
noted to have a tan dorsal ground color with
darker brown wash and some scattered small
russet blotches, a light tan mid-dorsal stripe
and inguinal ocelli, black blotches in the scap-
ular region, and a gray venter. Others (e.g.,
BPBM 15798) were noted to have a tan face
and heel marks on an otherwise dark brown
ground. These are the two most commonly
observed color patterns, although the tan
TABLE 2.—Geographic variation in mensural characters for three samples of Cophixalus variabilis comparing only adults.
Numbers in parentheses indicate sample sizes; asterisks indicate character pairs having significantly different covariant
relationships among locations using Tukey’s test (see text).
Cloudy Mts (41) Owen Stanley Mts (18) Fergusson Island (16)
Character Mean Range Mean Range Mean Range
SVL (mm) 15.5 13.6–19.4 15.7 14.1–17.9 17.3 15.3–19.7
TLfold/SVL 0.47 0.41–0.51 0.49 0.43–0.53 0.48 0.46–0.52
TLknee/SVL 0.52 0.46–0.56 0.53 0.46–0.58 0.53 0.50–0.57
EN/SVL 0.083 0.071–0.094 0.082 0.073–0.092 0.084 0.075–0.092
IN/SVL* 0.097 0.085–0.110 0.095 0.078–0.104 0.091 0.082–0.098
SN/SVL 0.13 0.11–0.15 0.13 0.12–0.15 0.13 0.12–0.14
TY/SVL* 0.057 0.040–0.072 0.051 0.039–0.075 0.048 0.035–0.066
EY/SVL 0.13 0.11–0.14 0.13 0.12–0.14 0.12 0.12–0.13
HW/SVL* 0.38 0.33–0.43 0.35 0.36–0.43 0.37 0.32–0.39
HL/SVL 0.32 0.26–0.35 0.32 0.28–0.35 0.32 0.30–0.33
HandL/SVL 0.28 0.25–0.32 0.27 0.19–0.34 0.27 0.25–0.30
FootL/SVL 0.48 0.43–0.53 0.48 0.42–0.52 0.47 0.44–0.51
3
rd
F/SVL* 0.061 0.047–0.078 0.063 0.052–0.070 0.057 0.048–0.065
4
th
T/SVL* 0.048 0.039–0.058 0.053 0.041–0.062 0.048 0.042–0.055
EN/IN* 0.86 0.79–1.00 0.87 0.74–1.00 0.92 0.82–1.00
EN/SN 0.64 0.54–0.74 0.62 0.55–0.70 0.64 0.58–0.74
EY/SN* 0.96 0.78–1.19 0.97 0.83–1.11 0.95 0.88–1.05
3
rd
F/4
th
T 1.3 1.0–1.5 1.2 1.1–1.4 1.2 1.1–1.3
HL/HW* 0.86 0.76–0.95 0.83 0.77–0.90 0.87 0.82–0.98
1
st
F/HandL* 0.22 0.18–0.26 0.21 0.15–0.25 0.24 0.21–0.28
June 2006] HERPETOLOGICA 207

ground color can also be predominantly russet
and some animals are entirely tan with a few
black flecks (e.g., BPBM 15772; Fig. 1A).
Occasional specimens have a dark ground
color with a narrow, light tan mid-dorsal stripe
and others are dark brown with a broad, russet
mid-dorsal band. Venters are almost always so
heavily flecked with melanin as to appear dark
gray, but occasional specimens are light gray.
Call.—We recorded 24 complete calls from
nine individuals (Table 4; Fig. 3). The average
call consists of 12–13 peeping notes (range 1–
31) with a mean repetition rate of 2.36 notes/s
(range 0.84–3.32 notes/s). The notes averaged
114 ms in duration (range 78–196 ms) and the
intervals between notes averaged 414 ms
(range 196–1070 ms). The mean dominant
frequency was 4175 Hz.
The holotype (BPBM 20165) produced two
different types of calls and was observed to
alternate between them (Table 4; Figs. 3A–F).
The first two complete calls that we recorded
had a mean repetition rate of 1.85 notes/s
while the next three calls had a repetition rate
of 3.27 notes/s. There were also differences in
the number of notes per call. The mean num-
ber of notes in the first two calls was 17.5
(range 10–25). This decreased to 9.7 (range 5–
14) for the next three calls. BPBM 20165 was
the only individual that we directly observed
alternating between the two different calling
rates, but this behavior could account for the
rather large spread that we observed in call
repetition rate (1.54–3.32 notes/s for mainland
populations of Cophixalus variabilis). BPBM
20165 also had a shorter mean call note (84
ms) than did the other individuals recorded
from the New Guinean mainland (mean 5 119
ms; n 5 15 calls from six individuals).
The two individuals from Fergusson Island
(BPBM 16273–74) had a much slower note
repetition rate (Table 4; Fig. 3G–I) than did
those from the New Guinean mainland. The
mean repetition rate for the two specimens
from Fergusson Island was 0.85 notes/s (n 5 4;
range 0.84–0.86 notes/s) compared to 2.67
notes/s for the seven specimens from the
mainland (n 5 20; range 1.53–3.33 notes/s).
These differences mainly reflect differences
in the duration of the internotes of these two
populations. The notes in the population from
Fergusson Island, which averaged 136 ms in
duration (range 102–196 ms), were only slightly
longer than those for mainland populations,
which averaged 110 ms (range 78–133 ms).
However, the mean internote duration for the
population from Fergusson Island was 1042 ms
(range 990–1070 ms) compared to a mean of
281 ms (range 196–532 ms) for individuals from
the New Guinean mainland.
During drought conditions in August and
September, 2002, males on Fergusson Island
were only heard to call on the four nights (out
of five weeks of field activity) having rain.
The two individuals from Fergusson Island
were recorded on one of these four nights and
the generally dry conditions may have affected
their calling rate. We heard only four individ-
uals calling irregularly during the first few
hours of darkness. In our experience, other
species of New Guinean frogs tend to call in
small numbers at irregular intervals and at
slower rates under drought conditions than
they do during wet weather. Unfortunately, we
do not have data for the populations from
Fergusson Island during wet weather, but
our overall impression of the call was that
TABLE 3.—Ontogenetic variation in mensural characters
for the Cloudy Mts sample of Cophixalus variabilis. Adults
were determined by sexual maturity (presence of enlarged
ova in females and vocal slits in males) and juveniles are
those animals smaller than the smallest adult. Numbers
in parentheses indicate sample sizes; asterisks indicate
character pairs having significantly different covariant
relationships between the samples using Tukey’s test (see
text). Differences in SVL were not compared.
Adults (41) Juveniles (7)
Character Mean Range Mean Range
SVL (mm) 16.1 13.6–19.4 11.8 10.1–15.4
TL
fold
/SVL* 0.47 0.41–0.51 0.48 0.41–0.52
TL
knee
/SVL* 0.52 0.46–0.56 0.53 0.46–0.58
EN/SVL 0.083 0.071–0.094 0.093 0.081–0.109
IN/SVL 0.097 0.085–0.110 0.109 0.099–0.119
SN/SVL 0.13 0.11–0.15 0.15 0.13–0.17
TY/SVL 0.057 0.040–0.072 0.051 0.044–0.065
EY/SVL 0.13 0.11–0.14 0.14 0.13–0.16
HW/SVL* 0.38 0.33–0.43 0.39 0.36–0.45
HL/SVL 0.32 0.26–0.35 0.35 0.31–0.38
HandL/SVL 0.28 0.25–0.32 0.27 0.26–0.30
FootL/SVL* 0.48 0.43–0.53 0.46 0.41–0.52
3
rd
F/SVL 0.061 0.047–0.078 0.056 0.050–0.062
4
th
T/SVL 0.048 0.039–0.058 0.048 0.036–0.059
EN/IN* 0.86 0.75–1.00 0.86 0.75–0.93
EN/SN* 0.64 0.54–0.74 0.64 0.60–0.67
EY/SN 0.96 0.78–1.19 0.96 0.85–1.13
3
rd
F/4
th
T 1.3 1.0–1.5 1.2 0.9–1.4
HL/HW 0.86 0.76–0.95 0.89 0.83–0.95
1
st
F/HandL 0.22 0.18–0.26 0.18 0.17–0.20
208 HERPETOLOGICA [Vol. 62, No. 2

it seemed relatively tepid compared to
other populations of Cophixalus variabilis
and its close relatives observed under wet
conditions.
Individuals were observed to call both day
and night but predominantly from dusk
through the first several hours of darkness.
Calling during daytime was more frequent
under cloudy conditions or immediately fol-
lowing a rain but never was observed to involve
as many animals as nocturnal calling. Males
most often called from exposed sites on
vegetation 1–2 m above ground, but daytime
calling was conducted from hidden locations
within hollow trees, under bark, or even in
underground burrows.
Two individuals from the New Guinean
mainland were recorded during the afternoon.
Their calls (n 5 10) averaged 5.1 notes (range
1–29) with a repetition rate of 2.96 notes/s.
The other five individuals from the mainland
produced calls (n 5 10) that averaged 16.9
(range 5–29) notes and had a mean repetition
rate of 2.41 notes/s. The duration of the notes
was similar in the two groups (116 ms during
the day and 105 ms at night). The internote
duration in the diurnal calls was slightly
shorter than for that for calls recorded at night
(215 ms vs. 341 ms). Dominant frequencies
were similar (4190 Hz for the diurnal calls;
4307 Hz for nocturnal calls).
In 13 of the 23 calls with two or more notes,
the first note is the longest note in the call (in
some cases by only a few ms). In the holotype
(BPBM 20165), the first note is the longest in
only one of five complete calls.
Ecological notes.—The species ranges from
sea level (Normanby Island) to 1500 m
(Fergusson Island), but most of our records
occurred from 300–1000 m. The species was
always found in closed-canopy lowland and
mid-elevation rainforest, except for the eleva-
tional outliers. The sea-level specimen was
found in a sago swamp; the 1500 m specimen
in a more open montane forest dominated
by Metrosideros (Myrtaceae). Where found,
the species appears to be widely distributed
throughout the forest, showing no obvious
tendency to cluster near streams, unlike
some other members of the genus (Menzies,
1975: 59).
Animals of both sexes make no apparent
attempt to hide themselves at night and are
easily encountered fully exposed while perch-
ing on vegetation. In this choice of perch
sites this species closely resembles species
of Albericus. During the above-mentioned
drought, several specimens were captured
hiding in moist soil beneath rocks alongside
a small forested stream at 320–360 m. Except
for two males calling during the daytime, these
were the only animals to be seen on the ground.
Judging from observations made incidental
to dissections, ants appear to form a significant
portion of the diet for this species.
Syntopic congeners include Cophixalus
verrucosus and an undescribed Cophixalus
related to C. ateles. The former is a larger
species with similar perching habits to C.
variabilis but shows some tendency to cluster
around stream courses; the latter is a slightly
smaller species that typically inhabits the
ground or vegetation within ca. 20 cm of the
ground.
Etymology.—The masculine Latin adjective
‘‘variabilis’’ means ‘‘changeable’’ and refers to
the great color-pattern and call-rate variation
seen in this species.
Distribution.—We have collected Cophix-
alus variabilis at several localities in south-
eastern Milne Bay Province (Fig. 4), including
the Cloudy Mts (700–1000 m), Owen Stanley
Mts (300–600 m), Fergusson Island (300–
1500 m), and Normanby Island (sea level).
Additionally, we have collected the species
near Aieme Creek (300 m), east of Port
Moresby, in Central Province (Fig. 4). We
presume the species to be restricted to the
Southeast Peninsula and adjacent islands.
Remarks.—The values in calling rate of
animals from Fergusson Island fall outside
the already wide range of variation observed in
mainland specimens of Cophixalus variabilis
(Table 4) and differ as greatly as that seen in
the closely related species described next.
Nonetheless, we tentatively include our mate-
rial from the D’Entrecasteaux Islands in C.
variabilis for two reasons. First, except for
a tendency toward larger body size, we can
find no morphological features that serve to
distinguish D’Entrecasteaux material from
mainland material. Second, the only calls we
could record of animals from Fergusson Island
were obtained during a prolonged drought. It
was our impression that these frogs, and other
syntopic species, were calling in a sluggish
June 2006] HERPETOLOGICA 209

TABLE 4.—Call characteristics for Cophixalus variabilis, C. timidus, and C. sisyphus.
Species/Specimen Call no. Locality Time (h)
Elevation
(m)
Temperature
(C)
Total
notes
in call
Mean
note
duration
(s)
Mean
internote
duration (s)
Repetition
rate
(notes/s)
Dominant
frequency (Hz)
Cophixalus variabilis
Uncaptured No. 1 A Cloudy Mts ca. 1800–2000 715 NR 29 0.112 0.477 1.696 4050
Uncaptured No. 2 A ’’ 1230 ’’ 27.5 2 0.111 0.222 2.843 4080
’’ B ’’ ’’ ’’ ’’ 5 0.117 0.236 2.801 4270
’’ C ’’ ’’ ’’ ’’ 1 0.123 — — 4030
’’ D ’’ ’’ ’’ ’’ 2 0.121 0.236 2.659 4240
’’ E ’’ ’’ ’’ ’’ 2 0.121 0.232 2.658 4350
’’ F ’’ ’’ ’’ ’’ 2 0.131 0.217 2.687 4410
BPBM 15772 A ’’ 1840 ’’ 23.0 27 0.122 0.532 1.530 4350
Uncaptured No. 3 A ’’ 1920 ’’ 23.4 20 0.132 0.282 2.406 4290
’’ C ’’ ’’ ’’ ’’ 19 0.133 0.269 2.481 4350
Uncaptured No. 4 A ’’ ’’ ’’ ’’ 10 0.125 0.279 2.469 4210
BPBM 15782 A ’’ 1340 ’’ 24.5 10 0.110 0.197 3.240 4230
’’ B ’’ ’’ ’’ ’’ 11 0.108 0.197 3.264 4150
’’ C ’’ ’’ ’’ ’’ 4 0.107 0.199 3.211 4030
’’ D ’’ ’’ ’’ ’’ 12 0.108 0.196 3.276 4110
BPBM 20165 A
1
Owen Stanley Mts 1830 330 23.6 1 0.080 — — 4380
’’ B ’’ ’’ ’’ ’’ 10 0.078 0.463 1.851 4260
’’ C ’’ ’’ ’’ ’’ 25 0.082 0.458 1.851 4220
’’ D ’’ ’’ ’’ ’’ 5 0.078 0.226 3.257 4630
’’ E ’’ ’’ ’’ ’’ 10 0.087 0.217 3.325 4450
’’ F ’’ ’’ ’’ ’’ 14 0.096 0.208 3.263 4260
BPBM 21886 A Moroka 1830 302 25.5 14 0.098 0.547 1.551 4260
’’ B ’’ ’’ ’’ ’’ 12 0.102 0.550 1.532 4160
’’ C ’’ ’’ ’’ ’’ 16 0.102 0.535 1.569 4180
BPBM 16273 A Fergusson Island 2040 960 18.8 11 0.102 1.062 0.858 3920
’’ B ’’ ’’ ’’ ’’ 16 0.121 1.070 0.840 3930
BPBM 16274 A ’’ 2050 ’’ ’’ 28 0.196 0.990 0.842 3730
’’ B ’’ ’’ ’’ ’’ 31 0.125 1.047 0.853 3640
Cophixalus timidus
BPBM 18098 A Mt Simpson 1100 2170 22.3 9 0.229 0.882 0.898 4000
’’ B ’’ ’’ ’’ ’’ 13 0.228 0.859 0.918 4080
’’ C ’’ ’’ ’’ ’’ 12 0.220 0.877 0.910 4000
’’ D ’’ ’’ ’’ ’’ 12 0.239 0.873 0.898 3940
’’ E ’’ ’’ ’’ ’’ 11 0.243 0.835 0.926 3950
BPBM 18107 A ’’ 1920 1490 18.0 14 0.203 0.967 0.854 4500
’’ B ’’ ’’ ’’ ’’ 10 0.213 0.956 0.855 4660
’’ C ’’ ’’ ’’ ’’ 9 0.194 0.948 0.874 4710
BPBM 18108 A ’’ 1930 ’’ ’’ 12 0.264 1.127 0.718 4840
’’ B ’’ ’’ ’’ ’’ 13 0.256 1.147 0.713 4710
’’ C ’’ ’’ ’’ ’’ 13 0.185 0.896 0.922 4600
’’ D ’’ ’’ ’’ ’’ 13 0.192 0.887 0.924 4550
’’ E ’’ ’’ ’’ ’’ 14 0.265 1.043 0.764 4600
Uncaptured No. 5 A
2
’’ ’’ ’’ ’’ 13 0.193 0.888 0.989 4490
Cophixalus sisyphus
BPBM 18316 A
3
Mt Obree 2115 600 22.0 9 0.145 0.309 2.196 4580
’’ B ’’ ’’ ’’ ’’ 10 0.159 0.304 2.148 4460
’’ C ’’ ’’ ’’ ’’ 10 0.160 0.305 2.142 4600
BPBM 19299 A ’’ 1910 1610 20.2 7 0.183 0.452 1.569 5000
’’ B ’’ ’’ ’’ ’’ 7 0.221 0.513 1.353 4860
’’ C ’’ ’’ ’’ ’’ 6 0.223 0.504 1.365 4620
’’ D ’’ ’’ ’’ ’’ 6 0.219 0.489 1.404 4890
BPBM 19300 A ’’ 1925 ’’ ’’ 7 0.183 0.452 1.569 4940
’’ B ’’ ’’ ’’ ’’ 7 0.191 0.459 1.535 4890
’’ C ’’ ’’ ’’ ’’ 7 0.194 0.476 1.484 4710
BPBM 19301 A ’’ 2030 ’’ ’’ 7 0.179 0.447 1.597 5100
’’ B ’’ ’’ ’’ ’’ 5 0.196 0.445 1.548 5190
210 HERPETOLOGICA [Vol. 62, No. 2

fashion during this drought (when calling at
all, which was extremely infrequently) and we
are not convinced that the slow call rates
obtained by us are not an artifact of the
adverse weather conditions. It remains to
obtain calls from specimens from the D’Entre-
casteaux Islands during wet weather condi-
tions to determine whether the apparent call
differences seen in our material are consistent,
and hence of potential taxonomic importance,
or merely vary with weather conditions. Until
that time, it seems most prudent to include
these samples within C. variabilis. Because
of their questionable status, calls from the
animals from Fergusson Island will not be
used in comparisons with the new species that
follow.
Cophixalus timidus sp. nov.
Holotype.—BPBM 18098 (field no. FK
7472), adult male (Figs. 1, 2), collected by F.
Kraus on NE slope Mt Simpson, 10.03168 S,
149.57678 E, 2170 m, Milne Bay Province,
Papua New Guinea, 21 February 2003.
Paratypes.—Papua New Guinea: Milne Bay
Province: NE slope Mt Simpson, 10.03808 S,
149.59848 E, 1800 m, 17 February 2003
(BPBM 18092); same data as holotype but
19 February 2003 (BPBM 18093–97); same
data as holotype (BPBM 18099–105, PNGNM
24002); N slope Mt Simpson, 10.03648 S,
149.57498 E, 2480 m, 21 February, 2003
(BPBM 18106); 0.5 km S Bunisi, 10.02098 S,
149.59478 E, 1490 m, 23 February 2003
(BPBM 18107–12); same locality but 24
February 2003 (BPBM 18113); NE Bunisi,
10.01408 S, 149.61298 E, 1520 m, 26 February
2003 (BPBM 18114); ca. 1 km S Bunisi,
10.02458 S, 149.59478 E, 1490–1540 m, 27
February 2003 (BPBM 18115–17); Bunisi,
10.01718 S, 149.60028 E, 1420 m, 25 February
2003 (PNGNM 24003–24006).
TABLE 4.—Continued.
Species/Specimen Call no. Locality
Time (h) Elevation
(m)
Temperature
(C)
Total
notes
in call
Mean
note
duration
(s)
Mean
internote
duration (s)
Repetition
rate
(notes/s)
Dominant
frequency
BPBM 19306 A ’’ 2105 ’’ 17.8 8 0.195 0.534 1.367 4990
’’ B ’’ ’’ ’’ ’’ 8 0.202 0.530 1.361 4840
’’ C ’’ ’’ ’’ ’’ 8 0.202 0.523 1.374 4830
BPBM 19308 A ’’ 1950 ’’ 19.2 8 0.158 0.451 1.636 5230
BPBM 19316 A ’’ 2030 1795 16.5 8 0.166 0.487 1.524 5570
’’ B ’’ ’’ ’’ ’’ 8 0.159 0.488 1.540 5600
’’ C ’’ ’’ ’’ ’’ 8 0.167 0.466 1.571 5630
1
Partial call – presumed last note of series.
2
Two notes cannot be accurately measured because of interference from another frog; call statistics based on 11 notes and 8 internotes; repetition rate based
on entire call.
3
Partial call – call statistics based on 8 full notes included in recording.
FIG. 3.—(A) Waveform, (B) power spectrum, and (C)
spectrogram of a slow-call sequence of Cophixalus
variabilis, BPBM 20165, holotype, from southern Owen
Stanley Mts; recorded 1830 h, air temperature 23.6 C.
(D)–(F) same data for a fast-call sequence from the same
animal. (G)–(I) same data for a call from BPBM 16273,
Oya Tabu, Fergusson Island; recorded 2040 h, air
temperature 18.8 C.
June 2006] HERPETOLOGICA 211

Diagnosis.—A small (adult males 13.5–17.5
mm, adult females 14.4–21.1 mm) species of
Cophixalus having a tuberculate dorsum; well-
developed first finger (1
st
F/HandL 5 0.09–
0.26); discs with well-defined terminal grooves
on all fingers and toes except the first finger,
which has a narrower, less flattened disc that
either lacks a terminal groove or has it weakly
developed (present typically but not exclusively
in large females); toes free of webbing; vocal
slits in males; sexual dimorphism in tympa-
num size (TY/SVL 5 0.031–0.052 in males,
0.048–0.084 in females); relatively short legs
(TL
knee
/SVL 5 0.42–0.52, TL
fold
/SVL 5 0.37–
0.46); a relatively short, broad snout (EN/IN 5
0.75–1.00); vertical lores; light yellow blotch in
the groin; 1–4 (usually 1) small yellow spots on
each side of the pectoral region; shanks
typically uniform dark gray; and a call consist-
ing of a series of 9–14 peeps.
Cophixalus timidus differs from all pre-
viously described Papuan Cophixalus in
the same suite of characters as exhibited by
C. variabilis, with which it is very similar, and
Cophixalus timidus differs from C. variabilis
in having (1) sexual dimorphism in tympanum
size; (2) somewhat shorter legs (TL
knee
/SVL 5
0.42–0.52 vs. 0.46–0.58 in C. variabilis); (3)
disc on first finger less well developed, usually
barely wider than penultimate phalanx, and
not as flattened as in C. variabilis; (4) usually
two (sometimes more) small yellow pectoral
spots surrounded by black, which are more
obvious than in C. variabilis; (5) typically
a light yellow blotch in the groin; (6) generally
no dark banding on the shanks; (7) longer call
notes (mean 0.185–0.265 ms vs. 0.078–0.133
ms in C. variabilis); and (8) slower call rate
(mean 0.713–0.989 notes/s vs. 1.530–3.325
notes/s in mainland C. variabilis).
Description of holotype.—An adult male.
Head moderately wide (HW/SVL 5 0.42),
with vertical loreal region; canthus rostralis
rounded, straight when viewed from above;
nostrils directed laterally, much closer to tip
of snout than to eyes; internarial distance con-
siderably larger than distance from external
naris to eye (EN/IN 5 0.88, IN/SVL 5 0.103,
EN/SVL 5 0.090); snout slightly projecting
when viewed from the side, shallowly angulate
when viewed from above; eyes moderately
large (EY/SVL 5 0.12); eyelid approximately
2/3 the width of the interorbital distance;
tympanum indistinct and small (TY/SVL 5
0.038). Dorsal and lateral surfaces weakly
tuberculate. Supratympanic fold weakly de-
veloped. Ventral surfaces generally smooth
but weakly granular on abdomen. Fingers
unwebbed, bearing discs with terminal
grooves; relative lengths 3 . 4 . 2 . 1; first
finger and disc moderately well-developed.
Finger discs approximately twice widths of
penultimate phalanges, except for first finger
disc, which is just barely wider than penulti-
mate phalanx (Fig. 2B). Articular tubercles low
and indistinct; metacarpal tubercles poorly
developed. Toes unwebbed, bearing discs with
terminal grooves; relative lengths 4 . 3 . 5 .
2 . 1. Toe discs smaller than those of fingers,
that of fourth toe approximately 1.5 times
width of penultimate phalanx, that of first
approximately same width as penultimate
phalanx. Subarticular tubercles lacking but
with thicker, lighter skin in some areas; inner
metatarsal tubercle narrow and elongate, outer
lacking. Hind legs relatively short (TL
knee
/
SVL 5 0.46, TL
fold
/SVL 5 0.42).
Dorsum with brown ground color heavily
and uniformly stippled with tiny black and
silver punctations, imparting an overall uni-
form dark gray appearance. Vestigial black
postocular bar present on each side; a few small
black spots present above forearm insertions.
Narrow cream cresent on each side of lumbar
region; remnants of lumbar ocelli but lacking
central dark blotches. Dark interrupted band
on forearm; legs unbanded. Iris dark brown
with numerous tiny silver specks arrayed in
a largely radial pattern. Venter light straw
FIG. 4.—Map of southeastern New Guinea and adjacent
islands, showing the ranges of Cophixalus variabilis (filled
circles; star is type locality), C. timidus (square), and C.
sisyphus (triangle).
212 HERPETOLOGICA [Vol. 62, No. 2

heavily stippled with tiny black punctations,
densest on chin and throat, slightly sparser on
posterior of abdomen and thighs. Pair of small
cream spots on the pectoral region.
Measurements (in mm).—SVL 5 15.6,
TL
fold
5 6.5, TL
knee
5 7.2, HW 5 6.5, HL 5
5.2, IN 5 1.6, EN 5 1.4, SN 5 2.0, EY 5 1.9,
TY 5 0.6, HandL 5 4.2, FootL 5 6.8, 3
rd
F 5
0.94, 4
th
T 5 0.83.
Variation.—Mensural variation in the sam-
ples is not great, although sexual differences
exist (Table 5). Females are generally larger
than males (t 5 5.536, P , 0.001). The
tympana of females are noticeably larger than
those of males and their relative size increases
with increasing SVL, whereas that of males
decreases slightly (F 5 4.30, P , 0.001).
Females have larger toe discs that increase
with increasing SV relative to those of males
(for 4
th
T vs. SVL, F 5 2.22, P 5 0.035).
Only two juveniles were available for
comparison and these have relatively wider
and longer snouts (IN/SVL 5 0.115–0.118,
SN/SVL 5 0.16) and larger heads (HW/
SVL 5 0.42, HL/SVL 5 0.37–0.40) than do
adults (Table 5).
Color pattern of Cophixalus timidus is not
as variable as in C. variabilis. The majority of
specimens are dorsally dark gray in alcohol,
many uniformly so but most with a few
scattered black flecks or small blotches. A
few specimens are light gray mottled with
black and dark brown, one is light brown with
similar mottling. A narrow, light mid-dorsal
stripe is seen in only six of the 31 specimens,
but is distinct in only one; a single specimen
has a wide, light-gray mid-dorsal band. One of
the individuals having a light mid-dorsal stripe
also has light stripes on the proximal portion of
the back of each thigh but these extend only
approximately one-third the length of the
thigh. Tan heels are seen in eight specimens
and, as in C. variabilis, are correlated with
presence of mid-dorsal stripe (4 specimens) or
band (1 specimen). Shanks typically lack any
trace of dark banding, but irregular bands
are present on the shanks of three, two of
which also have tan heels. Clear lumbar ocelli
are present in 22 of 31 specimens; of those
without, six (of nine) have the dark central
blotch of the expected ocellus but lack its light
anterior and lateral margins. Venters have
a straw ground color heavily flecked with dark
gray in all specimens, imparting a dark gray
appearance to the venter overall. In a few
specimens the gray flecking is somewhat less
dense, imparting a slightly lighter color to the
abdomen and proximal undersides of the
thighs. In those specimens, overall ventral
coloration appears two-toned to the naked eye,
with a dark chin and throat followed by
a lighter chest and abdomen. One to four
(usually one) distinct, cream spots are present
on each side of the pectoral regions in all
except one specimen.
Color in life.—In life, BPBM 18092 was
noted to have ‘‘Dorsum brown with yellow
lumbar ocelli and a few yellow flecks on snout.
Legs somewhat lighter; heels tan, margined in
black. Occasional black flecks on sides and
limbs. Tops of thighs and forearms burnt
orange. Venter light yellow heavily dusted with
gray; under legs same. Three cream spots on
chest. Posterior belly and groin lemon yellow
dusted with gray. Iris bronze with black
horizontal streaks.’’ BPBM 18095–96 had
similarly dark brown dorsa but BPBM 18097
had an ochre dorsum and yellow on the backs
of the thighs as well as the groin. BPBM 18110
was noted to have the dorsum dark brown
TABLE 5.—Sexual variation in mensural characters for
the type series of Cophixalus timidus from Mt Simpson.
Numbers in parentheses indicate sample sizes; asterisks
indicate character pairs having significantly different
covariant relationships between the sexes using Tukey’s
test (see text).
Males (14) Females (15)
Character Mean Range Mean Range
SVL (mm) 14.7 13.5–17.5 17.5 14.4–21.1
TLfold/SVL 0.43 0.37–0.46 0.41 0.38–0.45
TLknee/SVL 0.48 0.42–0.52 0.46 0.42–0.49
EN/SVL 0.088 0.080–0.096 0.087 0.081–0.093
IN/SVL 0.097 0.087–0.103 0.096 0.085–0.110
SN/SVL 0.14 0.12–0.14 0.14 0.12–0.15
TY/SVL* 0.043 0.031–0.052 0.070 0.058–0.084
EY/SVL 0.13 0.12–0.14 0.13 0.12–0.14
HW/SVL 0.39 0.35–0.42 0.38 0.34–0.42
HL/SVL 0.33 0.30–0.36 0.34 0.33–0.37
HandL/SVL 0.26 0.22–0.28 0.26 0.24–0.27
FootL/SVL 0.44 0.37–0.49 0.43 0.40–0.46
3
rd
F/SVL 0.059 0.045–0.075 0.056 0.046–0.068
4
th
T/SVL* 0.050 0.043–0.055 0.048 0.045–0.055
EN/IN 0.91 0.86–1.00 0.91 0.82–1.00
EN/SN 0.64 0.56–0.70 0.64 0.60–0.69
EY/SN 0.96 0.84–1.05 0.95 0.84–1.05
3
rd
F/4
th
T 1.2 1.0–1.6 1.2 1.0–1.4
HL/HW 0.86 0.80–0.92 0.91 0.82–0.96
1
st
F/HandL 0.16 0.13–0.26 0.17 0.12–0.21
June 2006] HERPETOLOGICA 213

mottled with tan, tan heels, and tan mid-dorsal
stripe and tan stripes on the backs of the
thighs. Its venter was dark gray with minute
light gray flecks, its iris was tan, and in contrast
to most other specimens it lacked the yellow
patches in the groin. The dorsum of PNGNM
24005 was dark burnt orange.
Call.—We recorded 14 complete calls from
four individuals (Table 4; Fig. 5A–C). The
average call consists of 12 peeping notes (range
9–14) with a mean repetition rate of 0.94
notes/s (range 0.71–0.99 notes/s). The notes
averaged 223 ms in duration (range in call
means 185–265 ms) and the intervals between
notes averaged 937 ms (range in call means
823–1147 ms). The mean dominant frequency
was 4402 Hz (range 3940–4880 Hz). In all calls
the first note was longer than the succeeding
notes, averaging 325 ms (range 235–367 ms)
compared to 215 ms (range 142–291 ms) for
subsequent notes.
As for Cophixalus variabilis, animals were
heard to call both day and night but the large
majority called during the first several hours of
darkness. Daytime calling was only observed
when animals were safely sequestered in
difficult-to-reach refugia. Males called from
0–2 m above the ground on available vegeta-
tion. Unlike C. variabilis and the species to
follow, males of C. timidus could be very wary
and difficult to capture, especially at upper
elevation sites.
Ecological notes.—This species was found
to occur in closed-canopy mid-elevation rain-
forest, montane cloud forest, and open mon-
tane mixed shrub-grassland (with Styphelia
and Deschampsia as common elements).
Streams were virtually absent from these
areas, so any tendency to cluster near streams
could not be assessed.
Specimens collected in mid-elevation
(;1500 m) rainforest behaved similarly to
Cophixalus variabilis in calling behavior, call-
ing while perched on stems or leaves of shrubs
or from within Pandanus leaf axils, and not
being unduly difficult to obtain. Individuals
observed in the upper-elevation (;2500 m)
open grassland/shrubland called most fre-
quently from low grass tussocks and were
virtually impossible to collect while calling.
Most specimens obtained from this general
area were obtained in daytime by removing
standing cover of a Dicranopteris linearis
thicket and raking through the wet leaf litter
underneath. This site was a forest-fernland
ecotone on a former landslip at 2170 m
elevation.
Syntopic congeners include Cophixalus
verrucosus and two undescribed Cophixalus
of unclear relationships. Cophixalus verruco-
sus is a larger species also scandent on low
vegetation but which shows some tendency to
cluster around stream courses; one of the
undescribed forms is a dwarf species also
found in the Dicranopteris leaf litter from
which several C. timidus were obtained; the
second undescribed species was only found
on the ground near a mid-elevation (1500 m)
stream.
Etymology.—The masculine Latin adjective
‘‘timidus’’ means ‘‘fearful’’ and refers to the
wary calling behavior of this species, which
contrasts with that of its close relatives de-
scribed herein.
Distribution.—We have collected or ob-
served C. timidus at several localities on the
northern slope of Mt Simpson, ranging in
elevation from 1400–2500 m. It is currently
known only from this area.
Cophixalus sisyphus sp. nov.
Holotype.—BPBM 19300 (field no. FK
8910), adult male (Figs. 1, 2), collected by F.
Kraus on W slope Mt Obree, 9.44568
S,
FIG. 5.—(A) Waveform, (B) spectrogram, and (C) power
spectrum of a call sequence of Cophixalus timidus, BPBM
18098, holotype, from Mt Simpson; recorded 1100 h, air
temperature 22.3 C. (D)–(F) same parameters for call
sequence of Cophixalus sisyphus, BPBM 19300, holotype,
from Mt Obree; recorded 1920 h, air temperature 20.2 C.
214 HERPETOLOGICA [Vol. 62, No. 2

148.00648 E, 1580–1640 m, Central Province,
Papua New Guinea, 30 January 2004.
Paratypes.—Papua New Guinea: Central
Province: 1.5–2 km E Dorobisoro, 9.45928 S,
147.93728 E, 600 m, 7 October 2003 (BPBM
18316); same data as holotype (BPBM 19299,
19301–05); same data as holotype but 31
January 2004 (BPBM 19306); W slope Mt
Obree, 9.44478 S, 148.00928 E, 1640–1680 m,
31 January 2004 (BPBM 19307); same data as
holotype but 2 February 2004 (BPBM 19308–
11); W slope Mt Obree, 9.45838 S, 148.02728
E, 1870 m, 4 February 2004 (BPBM 19312–
14); W slope Mt Obree, 9.45748 S, 148.02778
E, 1760–1870 m, 5 February 2004 (BPBM
19315–19, PNGNM 24007–11).
Diagnosis.—A small (adult males 12.0–14.1
mm, adult females 12.7–13.6 mm) species of
Cophixalus having a tuberculate dorsum;
poorly developed first finger (1
st
F/HandL 5
0.09–0.14); discs with well-defined terminal
grooves on all fingers and toes except the first
finger, which has a poorly developed, unflat-
tened disc usually lacking a terminal groove
(but is weakly developed in one large female);
toes free of webbing; vocal slits in males; sexual
dimorphism in tympanum size (TY/SVL 5
0.039–0.068 in males, 0.071–0.081 in females);
relatively short legs (TL
knee
/SVL 5 0.43–0.52,
TL
fold
/SVL 5 0.37–0.46); a relatively short,
broad snout (EN/IN 5 0.73–0.87); vertical
lores; 1–4 (usually 1) small yellow spots on
each side of the pectoral region; and a call
consisting of a series of 5–10 peeps.
Cophixalus sisyphus differs from all pre-
viously described Papuan Cophixalus except
C. ateles, C. bewaniensis, C. pipilans, C.
pulchellus, C. shellyi, and C. sphagnicola in
the same suite of characters as exhibited by C.
variabilis and C. timidus, except that it shares
with C. ateles, C. bewaniensis, C. pipilans, C.
pulchellus, C. shellyi,andC. sphagnicola a
shortened first finger lacking a well-developed
disc. It differs from all six of these in its
tuberculate dorsum and variegated or striped
color pattern.
Cophixalus sisyphus differs from C. varia-
bilis in having (1) smaller body size (SVL 5
12.0–14.1 mm vs. 13.6–19.7 in C. variabilis);
(2) sexual dimorphism in tympanum size; (3)
a shorter first finger (1
st
F/HandL 5 0.09–0.14
vs. 0.15–0.28 in C. variabilis) with a poorly
developed, unflattened disc almost always
lacking a terminal groove (vs. a wide, flattened
disc with a well-developed terminal groove
in C. variabilis); (4) somewhat shorter legs
(TL
knee
/SVL 5 0.43–0.52 vs. 0.46–0.58 in C.
variabilis); (5) usually two (sometimes more)
small yellow pectoral spots surrounded by
black, which are more obvious than in C.
variabilis; and (6) longer call notes (0.158–
0.223 ms vs. 0.107–0.159 ms in C. variabilis).
Cophixalus sisyphus differs from C. timidus
in having (1) smaller body size (SVL 5 12.0–
14.1 mm vs. 13.6–19.7 in C. timidus); (2)
broader snout (IN/SVL 5 0.102–0.117, EN/
IN 5 0.73–0.87 vs. IN/SVL 5 0.085–0.110,
EN/IN 5 0.75–1.00 in C. timidus); (3)
a shorter first finger (1
st
F/HandL 5 0.09–
0.14 vs. 0.12–0.26 in C. timidus); (4) absence
of a light yellow blotch in groin; and (5) faster
call rate (1.353–2.196 notes/s vs. 0.713–0.989
notes/s in C. timidus).
Description of holotype.—An adult male.
Head moderately wide (HW/SVL 5 0.39),
with slightly oblique loreal region; canthus
rostralis rounded, straight when viewed from
above; nostrils directed laterally, much closer
to tip of snout than to eyes; internarial distance
considerably larger than distance from ex-
ternal naris to eye (EN/IN 5 0.80, IN/SVL 5
0.109, EN/SVL 5 0.088); snout slightly
projecting when viewed from the side, shal-
lowly angulate when viewed from above; eyes
moderately large (EY/SVL 5 0.13); eyelid
approximately 2/3 the width of the interorbital
distance; tympanum indistinct and small (TY/
SVL 5 0.058). Dorsal and lateral surfaces ru-
gosely tuberculate. Supratympanic fold weakly
developed. Ventral surfaces granular. Fin-
gers unwebbed, bearing discs with terminal
grooves except on first finger; relative lengths 3
. 4 . 2 . 1; first finger very short (1
st
F/
HandL 5 0.10), poorly developed, disc ab-
sent. Finger discs approximately 1.5 times
widths of penultimate phalanges (Fig. 2C).
Articular tubercles low and indistinct; meta-
carpal tubercles absent. Toes unwebbed,
bearing discs with terminal grooves; relative
lengths 4 . 3 . 5 . 2 . 1. Toe discs
approximately same size as those of fingers,
that of fourth toe approximately 1.5 times
width of penultimate phalanx, that of first little
wider than penultimate phalanx. Subarticular
tubercles lacking; inner metatarsal tubercle
narrow, elongate, and poorly developed; outer
June 2006] HERPETOLOGICA 215

lacking. Hind legs relatively short (TL
knee
/
SVL 5 0.45, TL
fold
/SVL 5 0.41).
Dorsum with dark gray central hourglass
pattern extending from between eyes to above
vent, border slightly darker than center,
anterior end brown, with narrow light gray
vertebral stripe on posterior half; vestige of
light gray crescent on each side of lumbar
region of hourglass. Sides lateral to hourglass
uniform light gray. Light gray patch on top of
snout anterior to eyes, this mottled with dark
gray blotches. Black bar extends from behind
eye to corner of jaw. Dark band on center of
each shank; dark patch on each knee; posterior
of tarsus black. Iris dark brown heavily stippled
with silver. Venter gray densely stippled with
minute black and silver punctations and with an
irregular patchwork of small brown spots, best
developed under thighs and on margin of chin.
Measurements (in mm).—SVL 5 13.7,
TL
fold
5 5.6, TL
knee
5 6.2, HW 5 6.5, HL 5
5.2, IN 5 1.6, EN 5 1.4, SN 5 2.0, EY 5 1.9,
TY 5 0.6, HandL 5 4.2, FootL 5 6.8, 3
rd
F 5
0.94, 4
th
T 5 0.83.
Variation.—Mensural variation in the sam-
ples is slight (Table 6). Few females were
collected, making comparison of the sexes
difficult, but in contrast with Cophixalus
variabilis and C. timidus, there is no observed
sexual size dimorphism in our sample of C.
sisyphus (Table 6). However, as for C. timidus,
sexual dimorphism in tympanum size is
obvious, with the tympana of females notice-
ably larger than those of males. Female
tympana increase in relative size with in-
creasing SVL, whereas those of males decrease
slightly (F 5 2.22, P , 0.036). Juveniles are
also not available in our sample, making on-
togenetic comparisons impossible.
Color pattern variation in Cophixalus sisy-
phus is considerable, resembling C. variabilis
in this respect and differing from C. timidus.
In preservative, the dorsa may be more or less
uniformly gray or with a broad, mid-dorsal
hourglass band. Of the (largely) uniformly
colored specimens, six are uniform light gray,
six are light gray with small black flecks or
blotches, four are uniform dark gray with
a light gray triangular patch on the top of the
snout, and two are dark gray with scattered
small black flecks. Of the specimens with the
hourglass pattern, three have a dark gray hour-
glass with light gray sides, three have the
hourglass light gray with dark gray sides, and
three have the hourglass light gray with light
gray sides. In all cases, the lateral margins
of the hourglass are rimmed with black. Five
specimens have a narrow light gray mid-dorsal
stripe, which may occur in combination with
either the uniform or hourglass-patterned
dorsa, and one of these has similar narrow
stripes on the backs of each thigh and shank.
Half of the specimens have a single, central
dark band on each shank and half have no
banding. Seven specimens have tan heels.
Fifteen specimens have lumbar ocelli, but in
all but two or three of these the ocelli are
obscure; eleven specimens lack lumbar ocelli
entirely. Venters have a light gray ground color
heavily and uniformly stippled/suffused with
dark gray, giving a uniformly dark gray appear-
ance to the entire venter, with no contrast
apparent to the naked eye between the
intensity of coloration on the chin/throat vs.
chest/abdomen. One to four (usually one)
distinct, cream pectoral spots are present on
each side of the pectoral regions in all except
one specimen.
Color in life.—In life, BPBM 18316 was
noted to have ‘‘Dorsum variegated dark and
TABLE 6.—Sexual variation in mensural characters for the
type series of Cophixalus sisyphus from Mt Obree.
Numbers in parentheses indicate sample sizes; asterisk
indicates character pair having significantly different
covariant relationship between the sexes using Tukey’s
test (see text).
Males (24) Females (3)
Character Mean Range Mean Range
SVL (mm) 13.3 12.0–14.1 13.3 12.7–13.6
TLfold/SVL 0.43 0.39–0.46 0.44 0.43–0.44
TLknee/SVL 0.48 0.43–0.52 0.49 0.49–0.50
EN/SVL 0.087 0.079–0.096 0.091 0.088–0.094
IN/SVL 0.110 0.100–0.117 0.108 0.104–0.110
SN/SVL 0.14 0.12–0.15 0.15 0.15–0.16
TY/SVL* 0.052 0.039–0.063 0.078 0.071–0.081
EY/SVL 0.13 0.12–0.15 0.14 0.13–0.14
HW/SVL 0.39 0.36–0.43 0.37 0.36–0.40
HL/SVL 0.33 0.30–0.35 0.34 0.33–0.35
HandL/SVL 0.25 0.23–0.27 0.26 0.24–0.27
FootL/SVL 0.41 0.38–0.44 0.41 0.40–0.43
3
rd
F/SVL 0.054 0.045–0.063 0.053 0.050–0.057
4
th
T/SVL 0.055 0.049–0.060 0.052 0.047–0.056
EN/IN 0.79 0.73–0.87 0.84 0.80–0.86
EN/SN 0.62 0.58–0.72 0.60 0.57–0.63
EY/SN 0.93 0.81–1.11 0.90 0.86–0.95
3
rd
F/4
th
T 1.0 0.9–1.1 1.0 0.9–1.1
HL/HW 0.84 0.77–0.90 0.90 0.85–0.96
1
st
F/HandL 0.10 0.09–0.14 0.11 0.09–0.14
216 HERPETOLOGICA [Vol. 62, No. 2

light mud brown, with two brown-yellow
lumbar hemi-ocelli. Venter dark gray, flecked
with light gray, being more heavily flecked
posteriorly. Rear of thighs gray-brown with
pale brown flecks. Iris light mud brown.’’
BPBM 19299, from 1000 m higher elevation,
had ‘‘Dorsum reddish brown with scattered
small black blotches (mostly laterally), an
ochre vertebral stripe, and a ridge running
behind eye to suprascapular region. Iris
reddish brown. Black chevron above anus.
Face gray to black. Hidden surfaces of thighs
brown stippled with gray. Chin black, re-
mainder of venter light gray heavily stippled
with dark gray, so overall appearance is almost
black.’’ In contrast, BPBM 19300 (the holo-
type) was noted to have a tan ground color with
a dorsal brown hourglass blotch and a tan iris;
BPBM 19301 to have a tan mid-dorsum,
brown sides, and a narrow black stripe in-
tervening between the two; BPBM 19304 to
have a uniformly russet-brown dorsum;
PNGNM 24008 a brown dorsum with a tan
vertebral stripe and dark brown dorsolateral
stripes bordered below by tan and then yellow
brown; and PNGNM 24009 to be uniformly
russet.
Call.—We recorded 18 complete calls from
seven individuals (Table 4; Fig. 5D–F). The
average call consists of 8 peeping notes (range
5–10) with a mean repetition rate of 1.60 notes/s
(range 1.35–2.26 notes/s). The notes averaged
187 ms in duration (range in call means 158–
223 ms) and the intervals between notes
averaged 454 ms (range in call means 304–
534 ms). The mean dominant frequency was
4975 Hz (range 4460–5630 Hz). In all the calls
the first note was longer than the succeeding
notes, averaging 265 ms (range 213–318 ms;
n 5 18) compared to 172 ms (range 131–221;
n 5 117) for subsequent notes.
Animals called perched on leaves 0.15–1.5
m above ground and calling began after dark in
all instances observed by us. A train of calls
would be delivered every few minutes. Calling
animals were not energetic but were not
especially shy either.
Ecological notes.—This species was found
to occur in somewhat disturbed, closed-
canopy, low-elevation rainforest and in pri-
mary, closed-canopy, mid-elevation rainforest.
Animals largely occupied rather steep slopes
but a few were obtained on an adjacent ridge-
top as well; none was obtained in the im-
mediate vicinity of streams.
The only syntopic congener found with
Cophixalus sisyphus was a small terrestrial
species that is either referable to C. ateles or is
a new close relative of that species. An un-
described species of Cophixalus also inhabited
small stream margins in the nearby vicinity
but appeared not to occur syntopically with
C. sisyphus.
Etymology.—The name is a proper noun in
apposition and refers to the founder of
Corinth, who was condemned by the gods to
rolling a boulder to the top of a mountain only
to have it immediately roll down upon attain-
ing the summit, providing him an eternity of
futile labor. The name alludes to the seeming
task involved in sorting out the taxonomy of the
species described herein.
Distribution.—We have collected C. sisy-
phus only on the western slope of Mt Obree,
Central Province, at elevations ranging from
600–1800 m.
D
ISCUSSION
The three members of this species complex
are superficially very similar and difficult to
distinguish without detailed morphological
examination. Nonetheless, each can be differ-
entiated from the others by a combination of
subtle differences. Canonical variates analysis
separates all three species into cohesive groups
and confirms our placement of the specimens
from the D’Entrecasteaux Islands within
Cophixalus variabilis (Fig. 6). If pairwise dis-
criminant functions are run there is clear
separation of taxa in all cases, with no mis-
assigned cases for any comparison, and with
Hotelling’s T
2
values ranging from 1.263 10
14
to 2.23310
35
. These results hold true whether
the C. variabilis sample is restricted to only
mainland specimens or includes the material
from the D’Entrecasteaux as well. These
results are confirmed by general linear model
ANOVA analyses of paired variables using
Tukey’s tests. Although morphological differ-
ences in simple summary ratios can be subtle
(Table 7), covariate differences are frequently
taxonomically discriminating (Table 8).
The most striking feature of the members of
this species complex is the great variation in
color pattern features, which includes mottling
June 2006] HERPETOLOGICA 217

and both narrow and broad dorsal striping, tan
anal and heel patches, and leg striping, in
addition to simply unicolored dorsa. The
genetic basis for this variation remains un-
known but the phenomenon is not unprece-
dented. The inclusion of both unicolored/
mottled and striped/banded phenotypes with-
in single populations is similar to that seen in
Cophixalus ornatus from Australia (Zweifel,
1985) and in some Caribbean species of the
unrelated Eleutherodactylus (Schwartz and
Henderson, 1991). Within our complex,
Cophixalus timidus is slightly anomalous in
showing a more limited range of color pattern
features and a lower frequency of the more
striking features such as dorsal striping, band-
ing, and tan heels.
The call rate variation in Cophixalus varia-
bilis is another unusual feature requiring
comment inasmuch as that parameter is often
useful in distinguishing closely related species
of Papuan microhylids (e.g., Kraus and Allison,
2002; Zweifel, 2000). Indeed, we use that
character to distinguish C. timidus from the
other two members of the complex. Specimens
of C. variabilis typically call at two different
rates within any given chorus, the rates dif-
fering by a factor of approximately two. Given
this variation, it might be wondered if we have
inadvertently confounded two syntopic species
in our description of C. variabilis. But this
interpretation is obviated by our directly ob-
serving the holotype to alternate between the
slow and fast call rates within the course of a
few minutes. No other calling specimens were
in the area at this time. Furthermore: (1) C.
variabilis is morphologically cohesive (Fig. 6)
and we could discern no distinguishing mor-
phological features in either univariate or
multivariate space either within or among
geographic samples; (2) other relevant call
FIG. 6.—Scatter plot of canonical variate axes 1 and 2
with convex hulls delimiting the multivariate ranges for
mainland Cophixalus variabilis (filled circles), Fergusson
Island C. variabilis (open circles), C. timidus (squares),
and C. sisyphus (triangles).
TABLE 7.—Variation in mensural characters among the three species of Cophixalus comparing only adult males. Numbers
in parentheses indicate sample sizes.
Cophixalus variabilis (49) Cophixalus timidus (14) Cophixalus sisyphus (24)
Character Mean Range Mean Range Mean Range
SVL (mm) 15.9 13.6–18.6 14.7 13.5–17.5 13.3 12.0–14.1
TLfold/SVL 0.48 0.42–0.53 0.43 0.37–0.46 0.43 0.39–0.47
TLknee/SVL 0.53 0.46–0.58 0.48 0.42–0.52 0.48 0.43–0.52
EN/SVL 0.084 0.073–0.094 0.088 0.080–0.096 0.087 0.079–0.096
IN/SVL 0.096 0.078–0.110 0.097 0.087–0.103 0.110 0.100–0.117
SN/SVL 0.13 0.12–0.15 0.14 0.12–0.14 0.14 0.12–0.15
TY/SVL 0.050 0.035–0.075 0.043 0.031–0.052 0.052 0.039–0.068
EY/SVL 0.13 0.12–0.14 0.13 0.12–0.14 0.13 0.12–0.15
HW/SVL 0.37 0.35–0.43 0.39 0.35–0.42 0.39 0.36–0.43
HL/SVL 0.32 0.28–0.35 0.33 0.30–0.36 0.33 0.30–0.35
HandL/SVL 0.27 0.19–0.32 0.26 0.22–0.28 0.25 0.23–0.27
FootL/SVL 0.48 0.42–0.53 0.44 0.37–0.49 0.41 0.38–0.44
3
rd
F/SVL 0.060 0.049–0.073 0.059 0.045–0.075 0.054 0.039–0.063
4
th
T/SVL 0.050 0.040–0.062 0.050 0.043–0.055 0.055 0.044–0.060
EN/IN 0.87 0.79–1.00 0.91 0.86–1.00 0.79 0.73–0.87
EN/SN 0.64 0.55–0.74 0.64 0.56–0.70 0.62 0.58–0.72
EY/SN 0.97 0.86–1.11 0.96 0.84–1.05 0.93 0.81–1.11
3
rd
F/4
th
T 1.2 1.0–1.5 1.2 1.0–1.6 1.0 0.9–1.2
HL/HW 0.84 0.76–0.93 0.86 0.80–0.92 0.84 0.77–0.90
1
st
F/HandL 0.21 0.15–0.26 0.16 0.13–0.26 0.10 0.09–0.14
218 HERPETOLOGICA [Vol. 62, No. 2

parameters, such as dominant frequency, and
note length are the same within populations for
specimens exhibiting the two call rates; (3)
though not frequently investigated, variation in
notes/s for other Papuan microhylids can be of
a similar magnitude as seen in our samples (e.g.,
Gu¨ nther, 2000). Given these considerations,
we are confident that our mainland samples of
syntopic frogs with the fast and slow call rates
represent the same species.
The taxonomic status of the frogs from the
D’Entrecasteaux Islands is not certain. Mor-
phologically they are indistinguishable from
mainland samples of C. variabilis (Fig. 6), but
our recorded individuals were calling at a rate
slower than seen in mainland C. variabilis.We
suspect the calls of the frogs from Fergusson
Island to be atypical and think it inadvisable
to draw taxonomic conclusions without hav-
ing this issue clarified by further recordings.
If future research shows the slow calling
rate observed in our limited sample is consis-
tent among frogs of this complex from the
D’Entrecasteaux Islands, recognition of an
additional species may be warranted.
The nearest relatives of the Cophixalus
variabilis complex are difficult to discern. In
quantitative morphometrics, the new species
seem to differ least from the considerably
larger C. biroi, although the species could not
easily be confused with each other for size and
color-pattern differences alone. The new
species are also similar to C. daymani in
having a short snout and to C. cryptotympa-
num in having vertical lores—both characters
of limited occurrence in Cophixalus—but both
of these species lack vocal slits in males
(Zweifel, 1956) and differ in other morpho-
logical and color-pattern features. The remain-
ing Papuan Cophixalus species appear more
TABLE 8.—Pairwise covariate differences among the three species of Cophixalus for the character pairs in the left-hand
column, comparing only adult males. Sample sizes as for Table 7. C. variabilis 5 var, C. timidus 5 tim, C. sisyphus 5 sis.
Different in slope Different in Y intercept
Characters var vs. tim var vs. sis tim vs. sis var vs. tim var vs. sis tim vs. sis
TLfold vs. SVL T 5 7.987 T 5 5.151
P , 0.0001 P , 0.0001
TLknee vs. SVL T 5 7.005 T 5 5.110
P , 0.0001 P , 0.0001
EN vs. SVL
IN vs. SVL
SN vs. SVL T 5 2.867
TY vs. SVL T 5 4.506 P 5 0.0182
P 5 0.0001
EY vs. SVL
HW vs. SVL
HL vs. SVL T 5 2.774 T 5 2.656
P 5 0.0187 P 5 0.0255
HandL vs. SVL T 5 2.738 T 5 3.663
P 5 0.0205 P 5 0.0013
FootL vs. SVL T 5 7.117 T 5 5.639
P , 0.0001 P , 0.0001
3
rd
F vs. SVL T 5 3.920 T 5 3.041
P 5 0.0005 P 5 0.0088
4
th
T vs. SVL
EN vs. IN T 5 6.763 T 5 4.803
P , 0.0001 P , 0.0001
EN vs. SN T 5 3.711 T 5 3.086
P 5 0.0011 P 5 0.0078
EY vs. SN T 5 5.938 T 5 3.871
P , 0.0001 P 5 0.0006
3
rd
F vs. 4
th
T T 5 8.590 T 5 4.869
P , 0.0001 P , 0.0001
HL vs. HW T 5 2.569 T 5 2.373
P 5 0.0320 P 5 0.0520
1
st
F vs. HandL T 5 5.406 T 5 4.434 T 5 2.488
P , 0.0001 P 5 0.0001 P 5 0.0400
June 2006] HERPETOLOGICA 219

divergent and seem even less likely candidates
for any near relationship to the new form. It is
conceivable that the nearest relatives of the C.
variabilis complex may be found in Queens-
land. In superficial appearance, size, color-
pattern variation, and some morphological
details, the new species resemble C. ornatus;
nonetheless, various combinations of snout
shape, tibia length, and call differences clearly
distinguish the Papuan species from all Aus-
tralian forms (cf. Zweifel, 1985). Should the
Australian species prove to be closer relatives
of the C. variabilis complex than the currently
known Papuan species, it remains uncertain
what process would allow sister lineages to
occupy the Southeast Peninsula of New
Guinea and the Queensland mountains 600–
800 km distant, apparently without occupy-
ing intervening areas.
Acknowledgments.—We thank C. Kishinami and
K. Igeta for specimen processing and documentation;
C. Kishinami for helpful comments on the manuscript;
A. Taylor for statistical advice; B. Evans for help producing
the figures; J. Anamiato, P. Banige, I. Bigilale, N. Billy, B.
Iova, D. Iova, S. Kidabu, F. Malesa, W. Mamana, B. Sagara,
I. Sagara, D. Salepuna, J. Torena, B. Uruwa, B. Yuaru,
Allen, Dage, Don, Genta Sr., Genta Jr., Gule, James,
Jeffrey, Lawasi, Munda, Pipi, Suki, and Tanunu for field
assistance; and S. Andrew, Aussiya, H. Korage, D. Mitchell,
B. Salepuna, and B. Yawi for logistical field assistance. We
thank the many landowners who kindly granted us
permission to survey on their lands; the PNG National
Museum and Art Gallery for providing in-country collab-
orative assistance; and the Department of Environment
and Conservation, National Research Institute, and Milne
Bay Provincial Government for permission to conduct this
research. This research was supported by NSF grant DEB-
0103794. Contribution No. 2005-018 of the Pacific Bi-
ological Survey.
LITERATURE CITED
BURTON, T. C., AND R. G. ZWEIFEL. 1995. A new genus of
genyophrynine microhylid frogs from New Guinea.
American Museum Novitates 3129:1–7.
G
U
¨
NTHER, R. 2000. Albericus laurini species nova, the first
record of the genus Albericus (Anura, Microhylidae)
from the west of New Guinea. Mitteilungen der
Museum fu¨ r Naturkunde Berlin, Zoologische Reihe
76:167–174.
———. 2003. First record of the microhylid frog
genus Cophixalus from western Papua, Indonesia,
with descriptions of two new species. Herpetozoa 16:
3–21.
K
RAUS, F., AND A. ALLISON. 2000. Two new species of
Cophixalus from New Guinea. Journal of Herpetology
34:535–541.
———. 2001. A review of the endemic New Guinea
microhylid frog genus Choerophryne. Herpetologica
57:214–232.
———. 2002. A new species of Xenobatrachus (Anura:
Microhylidae) from northern Papua New Guinea.
Herpetologica 58:56–66.
M
ENZIES, J. I. 1975. Handbook of Common New Guinea
Frogs. Wau Ecology Institue, Wau, Papua New Guinea.
M
ENZIES, J. I., AND M. J. TYLER. 1977. The systematics
and adaptations of some Papuan microhylid frogs which
live underground. Journal of Zoology, London 183:
431–464.
P
ARKER, H. W. 1934. A Monograph of the Frogs of the
Family Microhylidae. British Museum (Natural History),
London, U.K.
S
CHWARTZ, A., AND R. W. HENDERSON. 1991. Amphibians
and Reptiles of the West Indies: Descriptions, Distri-
butions, and Natural history. University of Florida Press,
Gainesville, Florida, U.S.A.
Z
WEIFEL, R. G. 1956. Results of the Archbold Expeditions.
No. 72. Microhylid frogs from New Guinea, with
descriptions of new species. American Museum Nov-
itates 1766:1–49.
———. 1972. Results of the Archbold Expeditions. No.
97. A revision of the frog subfamily Asterophryinae
family Microhylidae. Bulletin of the American Museum
of Natural History 148:411–546.
———. 1985. Australian frogs of the family Microhylidae.
Bulletin of the American Museum of Natural History
182:265–388.
———. 2000. Partition of the Austropapuan microhylid
frog genus Sphenophryne with descriptions of new
species. Bulletin of the American Museum of Natural
History 253:1–130.
Z
WEIFEL, R. G., AND F. PARKER. 1989. New species of micro-
hylid frogs from the Owen Stanley Mountains of Papua
New Guinea and resurrection of the genus Aphanto-
phryne. American Museum Novitates 2954:1–20.
Accepted: 19 December 2005
Associate Editor: Joseph Mendelson
220 HERPETOLOGICA [Vol. 62, No. 2