ArticlePDF Available

Reproduction and terrestrial direct development in Sri Lankan shrub frogs (Ranidae: Rhacophorinae: Philautus)


Abstract and Figures

Two distinct nesting behaviours are reported from 98 clutches (80 in situ and 18 ex situ) of 17 species of direct-developing Philautus from Sri Lanka. The species examined, belong to three communities spanning 400–2,200 m elevation. Sixteen species are 'ground nesters', depositing 6–155 cream or white eggs in 5–35 mm deep nests excavated by the female in the rainforest floor. This is the first record of terrestrial egg burying in soil by the direct-developing anurans, a behaviour commonly employed by tetrapod reptiles. A single species (Philautus femoralis) however, is an arboreal nester, depositing 7–22 green, adhesive eggs in a disc-like mass on the underside of leaves. Parental care is not provided by any of the species. Larval development in Philautus differs from the Neotropical direct-developing leptodactylid, Eleutherodactylus coqui: the former lack external gills and have a rudimentary cement gland, a coiled gut and a larger, spatulate, heavily vascularized tail. They also lack an egg tooth. The first two of these apparently primitive characters have so far been observed only in aquatic anuran tadpoles and not in direct-developing larvae. Furthermore, the Philautus larvae studied possess atrophied internal gills, respiration apparently being facilitated primarily via the tail.
Content may be subject to copyright.
THE RAFFLES BULLETIN OF ZOOLOGY 2005 Supplement No. 12: 339–350
© National University of Singapore
Mohomed M. Bahir
Wildlife Heritage Trust, 95 Cotta Road, Colombo 8, Sri Lanka
Email: bahir@wht .org
Madhava Meegaskumbura
Wildlife Heritage Trust, 95 Cotta Road, Colombo 8, Sri Lanka
Department of Biology, Boston University, 5 Cummington Street, Boston, MA, 02215, USA
Kelum Manamendra-Arachchi
Wildlife Heritage Trust, 95 Cotta Road, Colombo 8, Sri Lanka
Email: kelum@
Christopher J. Schneider
Department of Biology, Boston University, 5 Cummington Street, Boston, MA, 02215, USA
Rohan Pethiyagoda
Wildlife Heritage Trust, 95 Cotta Road, Colombo 8, Sri Lanka (author for correspondence)
Email: rohan@
ABSTRACT. Two distinct nesting behaviours are reported from 98 clutches (80 in situ and 18 ex situ) of 17
species of direct-developing Philautus from Sri Lanka. The species examined, belong to three communities
spanning 400–2,200 m elevation. Sixteen species are ‘ground nesters’, depositing 6–155 cream or white eggs in
5–35 mm deep nests excavated by the female in the rainforest floor. This is the first record of terrestrial egg
burying in soil by the direct-developing anurans, a behaviour commonly employed by tetrapod reptiles. A
single species (Philautus femoralis) however, is an arboreal nester, depositing 7–22 green, adhesive eggs in a
disc-like mass on the underside of leaves. Parental care is not provided by any of the species.
Larval development in Philautus differs from the Neotropical direct-developing leptodactylid, Eleutherodactylus
coqui: the former lack external gills and have a rudimentary cement gland, a coiled gut and a larger, spatulate,
heavily vascularized tail. They also lack an egg tooth. The first two of these apparently primitive characters
have so far been observed only in aquatic anuran tadpoles and not in direct-developing larvae. Furthermore, the
Philautus larvae studied possess atrophied internal gills, respiration apparently being facilitated primarily via
the tail.
KEY WORDS. direct development, Philautus, Rhacophorinae, staging table, ground nesting, conservation
Sri Lanka is considered a global amphibian hotspot
(Meegaskumbura et al., 2002; Manamendra-Arachchi &
Pethiyagoda, 2005). A survey of anurans in about 350 locations
in Sri Lanka during the period 1993–2003 served to uncover a
large number of new species of frogs, the vast majority of
them tree frogs of the genus Philautus (Meegaskumbura et
al., 2002). Recently, a similar diversity of anurans has been
announced also in the Western Ghats of India (Biju, 2001).
Based essentially on indirect evidence—the chance discovery
clutches of large terrestrial eggs containing embryos—rather
than observations of reproductive ethology, it has long been
known that some Sri Lankan rhacophorines now assigned to
Philautus have terrestrial direct-developing eggs (Günther, 1876;
Ferguson, 1876; Kirtisinghe, 1946, 1957; Dutta & Manamendra-
Arachchi, 1996). Dring (1979) showed that the type species of
Philautus,P. aurifasciatus too, is a direct developer (see also
Alcala & Brown, 1982), while Yong & Ramly (1987) described
direct development of Philautus from Malaysia.
Bahir et al.: Reproduction in Sri Lankan Philautus
Eleven genera of Ranidae (including Rhacophorinae and
Mantellinae) are known to exhibit direct development of which
however, only Philautus and Taylorana are distributed in
Asia (Thibaudeau & Altig, 1999). Marmayou et al. (2000)
showed that endotrophy had arisen independently in these
two genera, while Bossuyt & Milinkovitch (2000) showed
that direct development, among other remarkable similarities,
between Mantidactylus (Mantellinae, until then considered
a subfamily within Rhacophoridae) and Philautus are in fact
the result of remarkable convergence.
Although about 110 nominal species of Philautus are
recognized as valid (Bossuyt & Dubois, 2001; Manamendra-
Arachchi & Pethiyagoda, 2005), terrestrial direct development
in this genus has been observed in relatively few species.
Dring (1987) observed reproduction in Sunda Shelf species,
while Kanamadi et al. (1996) and Patil & Kanamadi (1997)
observed direct development in an Indian species of
Philautus. Arboreal direct development has also been
observed in the Indian species P. bombayensis (see Bossuyt
et al., 2001), P. tinniens (see Bossuyt & Dubois, 2001) and P.
glandulosus (see Biju, 2003). None of these authors however,
provided data beyond the breeding event itself.
The discoveries of significant hitherto unsuspected species
richness within this genus in Sri Lanka and the Western Ghats,
both of which are generally considered herpetologically ‘well
studied’, suggests that the radiation of Philautus in tropical
Asia could parallel that of the direct-developing
Eleutherodactylus (Leptodactylidae), a remarkably speciose
genus which includes over 600 Neotropical species. While
development in a handful of Eleutherodactylus species,
especially E. coqui, has been well studied (Townsend &
Stewart, 1985, 1994; Ovaska & Rand, 2001; Rogowitz et al.,
2001; see also Callery et al., 2001 and references therein),
hardly any data are available on reproductive ethology and
development in Philautus.
It now seems likely that Thibaudeau & Altig’s (1999) reckoning
that direct developers comprise about 1,000 of the world’s
approximately 5,000 amphibian species (Hanken, 1999; Stuart
et al., 2004) was in fact an underestimate. Despite the frequency
of this developmental mode however, the terrestrial and
arboreal eggs of endotrophic anurans are usually well-
concealed and difficult to find, making direct observation
problematical. Indeed, Thibaudeau & Altig (1999) noted with
some justification that “developmental data on endotrophic
anurans generally are sparse or absent.” Here we attempt to
address this deficiency by presenting data on reproduction
and development in 17 species of Sri Lankan Philautus, which
includes a cloud-forest community of nine species in the
central highlands of the island.
Our study has also been motivated by the need to develop
captive breeding techniques as a tool of last resort for the ex
situ conservation of Critically Endangered species. Given that
a large number of Sri Lankan Philautus species have already
become extinct (Stuart et al., 2004; Manamendra-Arachchi &
Pethiyagoda, 2005), and given population declines or extreme
rarity in others, it is imperative that methods for large scale ex
situ conservation programmes be developed and made
Breeding behaviour (both in captivity and in the wild: see
Table 1) was observed in 17 species of Philautus at the four
sites listed below, principally at Site 1, incidentally at other
sites. Three of the 17 species are as yet undescribed, and are
referred to here by comparison with species they resemble:
Philautus cf. alto, P. cf. silus and P. cf. sordidus.
Site 1. – WHT Field Station (06º50’N, 80º40’E, 1,650 m above
sea level) adjacent to the 61.9 km2 Agra-Bopath Forest Reserve
in the central highlands of Sri Lanka. We selected twelve
species for captive breeding (Philautus alto,P. asankai, P.
caeruleus, P. femoralis,P. frankenbergi, P. microtympanum,
P. schmarda,P. sarasinorum,P. silus,P. cf. silus, P. cf. sordidus
and P. viridis).Nine of these inhabit secondary forest and
scrub in the 25 ha Wildlife Heritage Trust field station site
(06º50’36” N, 80º40’37” E; alt. 1,650 m) and adjacent forest.
The other three species were from Horton Plains (2,135 m alt.,
06º46’ N, 80º47’ E, Philautus frankenbergi), Peak Wilderness
(on the Bogawantalawa-Balangoda road, 1,300 m alt., 06º45’
N, 80º42’ E, P. caeruleus), and Check Polleat Gap (900 m alt.,
06º56’ N, 80º30’ E, Philautus cf. silus).
Temperatures at Site 1 vary from daily minima in the range
12.3–22.1ºC, to maxima in the range 18.9–28.1ºC. Annual rainfall
at the site is approximately 2,000 mm yr-1 (the nearest
meteorological station, at Nuwara Eliya, 06º57’N, 80º47’E, at
an altitude of 1,710 m records an annual average rainfall of
approximately 1,900 mm). Relative humidity is high (>70%)
throughout much of the year, reaching 100% during periods
of mist cover. Wind speeds during the southwest monsoon
(May–September) are typically high (12–20 km hr-1), reaching
a maximum of 86 km hr-1 during the study period.
Site 2. – Peradeniya, near Hantana (07º16’36” N, 80º37’E,
600–1,000 m) and Gannoruwa (07º17’ N, 80º35’ E, altitude 600
m) Forest Reserves, for Philautus hallidayi, P. rus,P.
sarasinorum,P. cf. sordidus and P. zorro.
Site 3. – Eastern Sinharaja (06º24’ N, 80º38’ E, altitude 1,060
m) for P. decoris.
Site 4. – Knuckles mountains (07º24’ N, 80º47’ E, altitude
1,100 m) for Philautus cf. alto and P. femoralis.
Captive breeding. Mature imagos of the 17 species of
Philautus studied here were maintained in terraria for up to
30 months, fed on wild-collected grasshoppers, crickets and
cultured fruit flies, coated with Sera®Reptimineral C vitamin
powder. The frogs were reared in all-glass terraria, 60×30×30–
75×75×45 cm (l×w×h) in dim light comparable with that of the
forest understorey. No artificial lighting was used. The terraria
were provided with a substratum of humus-rich forest topsoil
about 5–10 cm deep, about 50% of the floor area being covered
with an approximately 3 cm deep layer of leaf litter from the
forest floor. Branches from native local shrubs and potted
plants were provided as perches. In-terraria relative humidity
was maintained at > 90% using a hand-operated horticultural
sprayer at least three times daily at approximately 8 h intervals.
A few frogs developed eye infections, which we suspect was
contracted from bacteria in the stream water used for spraying:
given their wholly terrestrial life cycle, these frogs would not
normally come into extended contact with surface water, which
may contain pathogens not found in rain/mist water. Bacterial
cultures showed these infections to be the result of coliform
and beta haemolytic streptococci (Group C) bacteria; these
were treated successfully with ophthalmic chloramphenicol
Climate data (accuracy: rainfall to ±0.2 mm; temperature to
±0.5ºC; wind speed to ±1 km hr-1; and relative humidity to
±5%) at Site 1 were measured and logged on a Davis Weather
Monitor II system using Weatherlink® software. Additionally,
in-terraria temperature and humidity were measured on an
Oregon Scientific® EMR963HG cable-free thermo-hygrometer
(accuracy ±0.1º C). Metric measurements were made using
dial vernier callipers (accuracy ±0.05 mm). Specimens and ova
were weighed on a Acculab® PP-2060D electronic balance
(accuracy ±1 mg).
Abbreviations of larval measurements are as follows: BD
(body depth), maximum depth of head and body; BW (body
width), maximum width of body; HBL (head-body length),
distance from tip of snout to posterior extremity of body; TD
(tail depth), maximum depth of the tail including fin; TL (total
length), distance from tip of snout to tip of tail.
Eggs and embryos were preserved in a solution of equal parts
(by volume) of 10% buffered formalin and 70% ethanol. All
the specimens preserved in this study, mostly embryos, are
in the collection of the Wildlife Heritage Trust of Sri Lanka,
registration numbers WHT 5473–5784. After study, adult frogs
and most froglets were released at the same locations from
which they had originally been collected.
A simple regression analysis was performed for data on the 8
species of ground nesting Philautus,P. alto,P. asankai, P.
microtympanum,P. sarasinorum,P. schmarda,P.silus,P. cf.
sordidus and P. viridis, from the Agra-Bopath community to
investigate the relationships between mean female snout–
vent length and the means of the following variables: male
snout–vent length; number of eggs per clutch; mass of eggs;
egg diameter; egg-mixing duration; nest-covering duration;
and incubation time.
Breeding behaviour. – All 17 species of Philautus bred in
captivity. Observations of breeding in 10 species were also
made in the wild (see Table 1). Despite in-terrarium relative
humidity being maintained > 90% through frequent spraying,
mating always coincided with episodes of heavy rainfall and/
or mist cover, usually accompanied by high (80–100%) ambient
relative humidity (Fig. 1).
The 17 species studied exhibited two distinct reproductive
behaviours: (a) depositing eggs in soil on the forest floor (16
species); and (b) depositing adhesive eggs on the leaves of
understorey shrubs (one species, Philautus femoralis).
Ground nesting. – (see Figs. 2 and 3). Excluding P. femoralis,
Sri Lankan Philautus males perch and vocalize from the
branches of trees and shrubs 0.3–5 m above ground, or in
terrarium, from the leaves of plants (15–40 cm above floor
level). In the forest, males are entirely arboreal, descending to
the forest floor only to nest. Males advertise throughout the
year and during daytime, but more intensively at night,
especially during periods of rainfall and episodes of high
(>80%) relative humidity (Fig. 1). Females usually occupy the
lower branches of understorey shrubs.
In terrarium, courtship commences with the male calling from
his perch (Fig. 2a) and a female advancing towards him (Fig.
2b). Ripe ova are externally visible in females in mating
condition (Fig. 2b) except in P. schmarda, the only species
with heavily tuberculated skin among the species studied;
large eggs are visible in all ripe females however, in ventral
view. Males were never observed to move from their perches
until visual contact was made with the female prior to amplexus
(Fig. 2b). The male then moves towards the female and engages
in axillary amplexus (sensu Duellman & Trueb, 1986) while
still perched on a branch or, in some species, on the ground
(Fig. 2c). The male ceases to call once amplexus commences,
whereupon the pair descends to the forest floor (Figs. 2d;
3a). By this time their dorsal and lateral coloration changes to
match that of leaf litter or soil (Figs. 2j, k; 3a).
Table 1. The number of successful matings observed in captivity
and in the wild for 17 species of Sri Lankan Philautus, including
three undescribed species.
species number of matings
in captivity in the wild
Philautus alto 20 2
Philautus cf. alto 31
Philautus asankai 31
Philautus caeruleus 1–
Philautus decoris 1–
Philautus femoralis 11 2
Philautus frankenbergi 11
Philautus hallidayi 21
Philautus microtympanum 42
Philautus rus 2–
Philautus sarasinorum 31
Philautus schmarda 1–
Philautus silus 5–
Philautus cf. silus 1–
Philautus cf. sordidus 2–
Philautus viridis 19 6
Philautus zorro 11
total 80 18
Bahir et al.: Reproduction in Sri Lankan Philautus
Total Rain Average Humidity
Rainfall (mm day-1)
Relative humidity (%)
Fig. 1. Relative humidity (percent) and rainfall (mm day-1) from April to July 2002 at the WHT field station, indicating in green the dates on
which Philautus mating occurred (weather data for few days are unavailable), usually at the onset of periods of sustained rainfall and when
relative humidity was 80–100 %.
Fig. 2. Mating behaviour of ground-nesting Philautus. a, a male calls from his perch; b, ripe female approaches male; c, amplexus occurs on
a branch; d, the pair descends to the forest floor, the female locates a suitable nest-site and commences excavation; e, female uses snout and
rear limbs to widen the cavity; f, oviposition occurs within the nest cavity; g, after oviposition, the male departs and the female uses her
hands to mix the eggs with soil; h, after mixing, she uses her hands to cover the nest with soil and then abandons it; i, a pair of Philautus viridis
excavate a nest in soil (a few leaves have been removed to facilitate the photograph); j, a pair of P. viridis excavate a nest in bare soil.
The female then locates a suitable patch of ground on which
to nest, usually under leaf litter in moist soil (Fig. 3b). There,
with the male still in axillary amplexus, she excavates a cavity
1.5–50 cm3 (Figs. 2, 3c) using primarily her forelimbs and to a
lesser extent her hind limbs and snout (Fig. 2f). Moist soil is
scooped to the sides using both hands sequentially,
completing 3–9 strokes with one hand before changing to the
other. As excavation proceeds, the female slowly revolves
her body clockwise and anti-clockwise, using her hind limbs
and snout to shape and enlarge the excavation. Body rotation
occurs in steps of about 60–180º at a time. From the inception
of nest excavation to oviposition the female keeps her head
depressed against the nest wall. A single pair each of P. alto
and P. cf. sordidus used only their hind limbs and snouts to
excavate the cavity, rotating their bodies as in other species;
but this may have been because the soil was saturated with
water—these species may use burrowing methods similar to
the other species when in unsaturated soil.
Amplexus lasts 1.5–29.5 hr, during which 6–155 eggs, 3.7–5.7
mm in diameter, are laid (Figs. 2f, 3d). After oviposition, the
female begins mixing the eggs with soil and the male
dismounts and returns to a perch. Egg mixing is performed by
the female alone (Figs. 2g, 3e) using her hands and rotating
her body as in nest excavation. Mixing results in the eggs
being mixed with sperm and covered with grains of sand and
decaying leaf matter, while absorbing water from the moist
soil. This action serves also to separate the eggs from one
another, possibly facilitating increased respiration. After mixing
the eggs, the female depresses the clutch into the cavity by
swivelling her body 1–4 rotations in either sense, and then
loosely covers the nest with soil and debris using her hands
(Fig. 2i) and sometimes her snout, employing the same actions
as in nest cavity excavation, following which she abandons
the clutch. We noted also that the eggs absorb water during
the mixing process, resulting in a volume and weight increment
of 165–185% and 90–95% respectively in the case of P. viridis
eggs, the only species for which this was quantified.
Regression analysis shows that for mating pairs, female SVL
is strongly correlated to average number of eggs (R2 = 0.929)
and male snout–vent length (R2 = 0.822) (Fig. 7) but not egg
size or other variables (R2= <0.29). Incubation takes from 24–
68 days, with imagos emerging between stages 12–14 (see
staging table, Fig. 5, Table 2), with or without the tail fully
absorbed but with the yolk sac still visible. All ground-nesting
Philautus froglets show adult colouration.
Leaf nesting.(see Fig. 4). The breeding sequence of the
arboreal, leaf-nesting Philautus femoralis commences as for
the ground-nesting species, with the male advertising while
perched on a leaf about 0.3–3 m above ground. A gravid female
with ripe eggs (Fig. 4a) approaches the male, whereupon
axillary amplexus occurs (Fig. 4b). The pair in amplexus do
not descend to the ground as in ground-nesting species, but
the female selects a leaf suitable for deposition of eggs, about
0.3–2 m above ground, on which she deposits 7–22 (n = 13)
adhesive, green eggs while the male fertilizes them (Fig. 4c).
Unlike the ground-nesting species, mating P. femoralis do
not appear to alter colour at any stage of the reproductive
process. During egg laying the male places his vent on the
leaf, slightly below the female’s vent, and thrusts each egg
under the female’s body as he fertilizes the clutch. The male
departs soon after the last egg is laid, while the female sits on
the adhesive eggs from 1–3 h, occasionally pressing an egg
down more firmly using her feet (Fig. 4d). When compared
with ground nesters, the females of P. femoralis do not revolve
their bodies on the egg clutch or attempt to camouflage the
a b c
Fig. 3. Egg-laying sequence of Philautus alto from Agarapatana. a, the pair in amplexus descend to the ground; b, the female selects a
nesting site and starts excavating a nest pit; c, a further stage of digging the nest pit; d, laying eggs inside the pit, while the male is still
in amplexus; e, the female mixes the eggs and soil together after the male’s departure; and f, the female covers and camouflages the nest
pit, and abandons it.
Bahir et al.: Reproduction in Sri Lankan Philautus
nest. The female then abandons the clutch (Fig. 4e). Brown-
black froglets emerge (Fig. 4f) 37–49 (n = 5) days later.
The eggs of both ground- and leaf-nesting species of Philautus
are intolerant of desiccation. Three clutches of P. femoralis were
rendered unviable following a three-day period during which
relative humidity was reduced to < 75%. Similarly, egg clutches
(n = 6) of ground nesting species became unviable following
exposure to air or by failure to moisten the soil for 4–5 days
during periods of low (< 75%) relative humidity.
Juveniles of all ground nesting species were observed on
leaf litter throughout the year, but were more abundant during
periods of havy rainfall. Philautus femoralis juveniles and
clutches, however, were seen only during periods of sustained
rainfall, especially during the May–September southwest
Direct development in Sri Lankan Philautus. – Fourteen
stages (Fig. 5, Table 2) are distinguishable in the development
of the Philautus species reported on here. Post-larval growth
in the first 12 months was recorded for a single species,
Philautus cf. sordidus, at which point the juveniles had
reached, on average, 63% and 68% of adult male and female
SVL, respectively. Males among these began vocalizing at 15
Description of embryos. Philautus silus. Stages 6–13.
General description: eyes laterally orientated; mouth
subterminal, open, lower and upper lips prominent.
Stage 6 (WHT 5625). BH 3.0 mm, BW 3.4 mm, HBL, 4.5 mm,
TD, 2.0 mm, TL, 6.6 mm. Snout rounded in all aspects; lower
eyelid not visible; endolymphatic calcium deposits visible;
mouth width 0.8 mm; forelimbs concealed under opercular
Fig. 5 (opposite page). Developmental stages 1–14 of P. viridis. Stage 1 (dorsolateral aspect), egg unpigmented, creamy white, no limb-buds
visible, neural groove closed, head and tail ends differentiable; Stage 2, a (lateral aspect), front limb-buds feebly defined; cement gland visible;
b (dorsal aspect), hindlimb-buds distinct, tail bud visible; Stage 3, a, b (lateral aspect), unpigmented subdermal eyes visible; mouth opens; tail
about twice its length at Stage 2, laterally curved left or right; endolymphatic calcium deposits visible; cement gland visible; Stage 4 (lateral
aspect),dorsal and lateral surface of yolk sac pigmented; eyes pigmented, not subdermal; nostrils appear; tail about ¼ diameter of yolk-sac,
almost its maximum length; hind-limb buds elongate; cement gland indistinct; Stage 5 a (posterior aspect); b, hind limb bud of tadpole; notch
separating toes 4 and 5 distinct; Stage 6 a (posterior aspect), toe 3 demarcated; coiled gut forms; b, hind limb, toe 3 demarcated; Stage 7 a
(posterior aspect), toes 1 and 2 demarcated; toes 3, 4 and 5 clearly separated; b, hind limb, toes 3, 4 and 5 clearly separate; c, hind limb, toes
1 and 2 demarcated; toes 3–5 clearly separate; Stage 8 a (lateral aspect), tips of toes enlarged, ‘elbow slit’ appears in forelimbs, still concealed
under opercular fold; lower eyelid appears; b, hind limb, toes 1–5 distinctly separated; Stage 9 (lateral aspect), forelimb elbow emerges; hand
subdermal; Stage 10 (lateral aspect), forelimbs fully emerged; nictitating membrane visible; metatarsal tubercle distinct; subarticular tubercles
visible on toes; tail fin spatulate, fully developed; Stage 11 a (lateral aspect), digits enlarged, rounded; palmar tubercles visible; b (ventral
aspect of hand), enlarged digits rounded; palmar tubercles visible; c (ventral aspect of foot), enlarged digits rounded; subarticular tubercles
visible; Stage 12 (lateral aspect of head), upper eyelid visible; Stage 13 (lateral aspect), tail reduced; gape reaches mid-length of eye; palm and
foot pigmented, more densely dorsally and laterally; Stage 14 a (dorsal aspect at emergence of imago), tail reduced to a small protuberance;
terminal groove visible on discs; yolk sac still present; b (lateral aspect of head), gape reaches beyond mid-length of eye. Scale bar = 1 mm.
Fig. 4. Egg laying sequence of Philatus femoralis from Agarapatana. a, the female in breeding condition approaches a male; b, the pair engage
in amplexus on a leaf; c, the female lays eggs on the underside of the leaf while the male fertilizes the eggs; d, the female sits on the clutch after
the male departs; e, the developing clutch; f, froglets emerge from the nest.
Fig. 5. (For caption, see opposite page.)
14a 14b
Bahir et al.: Reproduction in Sri Lankan Philautus
Table 2. Principal characters used to stage embryonic development in P. viridis (see Fig. 5).
Incubation Stage Developmental characters
period (days)
6 1 Egg unpigmented, creamy white; no limb-buds visible; neural groove closed; head and tail ends
8 2 Hind-limb buds clearly visible; tail bud visible; forelimb buds feebly defined.
10 3 Unpigmented subdermal eyes visible; mouth opens; tail about twice its length at Stage 2, laterally
curved left or right; endolymphatic calcium deposits visible.
13 4 Dorsal and lateral surface of yolk sac pigmented; eyes pigmented, not subdermal; nostrils appear; tail
about ¼ diameter of yolk-sac, almost its maximum length; hind-limb buds elongate.
16 5 Notch separating toes 4 and 5 distinct.
19 6 Toe 3 demarcated.
20 7 Toes 1 and 2 demarcated; toes 3, 4 and 5 clearly separated.
23 8 Tips of toes enlarged; ‘elbow slit’ appears for forelimbs, concealed under opercular fold; lower eyelid
appears; Toes 1–5 distinctly separated.
27 9 Forelimb elbow emerges; hand subdermal.
29 10 Forelimbs fully emerged; nictitating membrane visible; metatarsal tubercle distinct; subarticular tubercles
visible on toes; tail fin fully developed.
31 11 Enlarged digits rounded; palmar tubercles visible.
35 12 Upper eyelid visible.
39 13 Tail reduced; gape reaches mid-length of eye; palm and foot pigmented, more densely dorsally and
46 14 Tail reduced to a small protuberance; gape reaches beyond mid-length of eye; terminal groove visible
on discs; yolk sac still present.
iv v
Fig. 6. Micrographs of developmental stages of P. sarasinorum and P. hallidayi. a, Stage 3 larvae of P. hallidayi (i, eye, before deposition of
pigments; ii, four internal gill chambers; iii, forelimb bud); b, P. sarasinorum, anterior lateral view of Stage 4 larvae (iv,cement gland; v, eye,
after deposition of pigments); c, P. sarasinorum stage 7 larvae (vi, forelimbs concealed under opercular fold; vii, coiled gut; viii, hind limbs
before disks develop); d, P. hallidayi stage 11 larvae (ix, large, spatulate, heavily-vascularized tail).
fold, elbow slit not differentiable; tips of toes not enlarged,
toes 3–5 distinct; tail large, spatulate, curved left, reaching
forelimb, concealed under opercular fold; dorsal and anal fins
deep, vascularized, their tips rounded; head, mid-dorsal and
anterior side of body heavily pigmented; a pigmented mid-
ventral band present; hind limbs unpigmented; coiled gut
visible below forelimbs.
Stage 8 (WHT 5631). BH 3.1 mm, BW 3.5 mm, HBL, 4.6 mm,
TD, 2.6 mm, TL, 9.7 mm. Snout rounded in all aspects; lower
eyelid visible; endolymphatic calcium deposits visible; mouth
width 1.0 mm; elbow slit (see Fig. 5, stage 8a) prominent on
forelimb, concealed under opercular fold; tips of toes enlarged,
toes 1–5 distinctly separated; tail large, spatulate, curved right,
reaching posterior margin of eye; dorsal and anal fins deep,
vascularized, their tips rounded; head, mid-dorsal and anterior
side of body heavily pigmented; hind limbs slightly
pigmented; tail and ventral foot lack pigmentation.
Stage 12 (WHT 5635). BH 3.2 mm, BW 3.9 mm, HBL, 6 mm, TD,
2.4 mm, TL, 11.7 mm. Snout rounded in all aspects; lower and
upper eyelids prominent; endolymphatic calcium deposits
indistinct; mouth width 1.5 mm; forelimbs well developed;
vent slit prominent; tips of toes enlarged; toes 1–5 distinctly
separated; subarticular tubercles visible; tail curved left,
reaching up to forelimb; dorsal and anal fins deep,
vascularized, their tips rounded; body heavily pigmented
except for posterior side of abdomen, and ventral surfaces of
hand and foot; tail lacks pigmentation.
Stage 13 (WHT 5641). BH 3.2 mm, BW 4.2 mm, HBL 5.1 mm,
TD 1.9 mm, TL, 9.1 mm. Snout obtusely pointed in all aspects;
lower and upper eyelids prominent; endolymphatic calcium
deposits indistinct; mouth width 1.8 mm; gape reaches mid-
length of eye; tail reduced, curved right, reaching forelimbs;
dorsal and anal fins not deep, vascularized, their tips rounded;
body heavily pigmented except posterior surface of abdomen,
and ventral surfaces of hand and foot; tail lacks pigmentation.
Reproductive behaviour. Kanamadi et al. (1996) and Patil &
Kanamadi (1997) reported on ethology and development
respectively in an Indian species of Philautus they identified
as P. variabilis. This species, originally described from Sri
Lanka by Günther (1859), was reported extinct by
Manamendra-Arachchi & Pethiyagoda (2005). It is also likely
that it was, like all other Philautus species occurring on the
island, endemic to Sri Lanka. We suspect therefore that Patil
& Kanamadi (1997) misidentified this species, but cannot
verify this inference as they do not list the preserved material
on which their study was based. Biju (2001) reported the
discovery of more than 100 species of putatively new anuran
species, most of them Philautus, in south-western India—
we presume the species studied by Kanamadi et al. (1996)
and Patil & Kanamadi (1997) was an undescribed one or one
described recently by Kuramoto & Joshy (2004).
Patil & Kanamadi’s (1997) observations are nevertheless of
value in comparing Sri Lankan and south Indian Philautus,
which form a monophyletic group (Meegaskumbura et al.,
2002). Although Patil & Kanamadi report no detailed
ethological observations, larval development in the Indian
species studied by them appears to differ in several respects
from Sri Lankan Philautus reported here—a development
period of 12 days, vs. 24–68 days for Sri Lankan Philautus;
the appearance of external gills (Patil & Kanamadi, 1997: fig.
2i) vs. absence of external gills; and the early, simultaneous
appearance of hind and forelimbs (Patil & Kanamadi, 1997:
Fig. 2d, comparable to Stage 3 in Sri Lankan Philautus) vs.
forelimbs concealed under opercular fold until Stage 10 (Fig.
5.10). We speculate that the much shorter incubation period
recorded by Patil & Kanamadi (1997) is a consequence of the
higher ambient temperature at their study site (Dharwat) than
at the Sri Lankan sites.
Kanamadi et al.’s (1996) behavioural observations too, differ
from ours in several respects: their females attacked the males
after copulation, vs. voluntary departure of the male in our
study; and the female guarded the eggs for 3 h after
oviposition, vs. female abandoned the nest immediately after
she concealed it. We observed neither aggression between
the sexes nor nest guarding by either parent in the 85 clutches
of ground-nesting species we studied. We suspect that
Kanamadi et al.’s (1996) results may have been influenced by
their apparent failure to provide their terrarium with an
adequately deep soil and leaf-litter substrate.
Dring (1987) described the deposition of 6–8 eggs by
Philautus kerangae in the pitcher of Nepenthes bicalcarata
in Sarawak, and deduced the occurrence of male parental care
from the presence of a calling male beside the clutch. He
noted also that embryos of P. kerangae had subdermal
forelimbs (as in the Sri Lankan species studied by us, but not
the Indian one described by Patil & Kanamadi, 1997). In
common with the Sri Lankan embryos reported on here,
Dring’s specimens lacked beaks, labial teeth, expanded lips
and oral suckers. Dring (1987) also noted a snout-tip tubercle
in females of P. petersi, a high-altitude (1,300–1,500 m) Bornean
and Malaysian species, and speculated that this may facilitate
nest excavation. No such character has been observed in any
Sri Lankan Philautus.
Alcala & Brown (1982) recorded arboreal clutches of Philautus
lissobrachius (from leaf axils of Pandanus trees and
Asplenium ferns) and P. schmackeri (from leaf axils of
Asplenium ferns in the Philippines). Brown & Alcala (1994)
also recorded clutches of P. surdus (5–19 eggs) from the leaf
axils of Pandanus and from the “base” of arboreal Asplenium
ferns, together with clutches of a species thought to be P.
acutirostris from leaf axils of wild banana plants and aerial
ferns. We have hitherto failed to find evidence that Sri Lankan
Philautus use leaf axils as nest sites. Further, in their Fig. 5,
Alcala & Brown (1982: 224) note and depict the presence of a
prominent operculum in their embryos, a structure we have
not observed in the embryos of Sri Lankan Philautus.
With more than 600 species, the Neotropical leptodactylid
genus Eleutherodactylus comprises the largest and arguably
the hitherto best studied group of endotrophic anurans
Bahir et al.: Reproduction in Sri Lankan Philautus
(Hanken, 1999). Breeding and nesting behaviour of four
species are well studied (Townsend & Stewart, 1994; Bourne,
1997; Rogowitz et. al., 2001; Ovaska & Rand, 2001); parental
care occurs in all but one of these. Oviposition sites are known
for 16 species of Eleutherodactylus (Schwartz & Henderson,
1991; Thibaudeau & Altig, 1999). Fourteen of these deposit
eggs under various objects while E. coqui deposits eggs on
elevated sites (Thibaudeau & Altig, 1999) and E. diastema
deposits its clutch between leaves, plant stems or inside
bromeliads (Ovaska & Rand, 2001). No Eleutherodactylus is
known to excavate a nest in soil as Sri Lankan Philautus do.
All but one of the 44 described extant species of Sri Lankan
Philautus occur in the island’s south-western wet zone,
annual rainfall > 2,000 mm (Manamendra-Arachchi &
Pethiyagoda, 2005; Meegaskumbura & Manamendra-
Arachchi, 2005), a distribution probably determined by the
dependence of these species on loose, moist, shaded soil in
which to deposit their eggs. We note that egg separation has
not been observed in other direct-developing species
(Thibaudeau & Altig, 1999) and conclude that this behaviour
serves to space eggs and increase the effective area available
for aeration, while also facilitating better distribution of sperm.
This is especially relevant given the much larger clutch sizes
observed in Sri Lankan Philautus (6–155 ova) compared to
other South and Southeast Asian Philautus (5–62 ova: Alcala
& Brown, 1994; Patil & Kanamadi, 1997).
Terrestrial nest excavation has evidently not hitherto been
recorded in other direct-developing anurans (Thibaudeau &
Altig, 1999), though it is a behaviour commonplace among
reptiles. While terrestrial nest excavation in Philautus shows
marked similarities to tetrapod reptiles, this behaviour is
different in these frogs in that the pair remain in amplexus
while the female excavates the nest and mixes the eggs, sperm
and soil after oviposition.
Although two Indian Philautus species are arboreal leaf
nesters (Bossuyt et al., 2001; Biju, 2003), the only known Sri
Lankan leaf nester, Philautus femoralis, differs from them by
laying green eggs only on the underside of leaves, whereas
the Indian species lay white or cream-coloured eggs on the
upper surface of leaves (Bossuyt et al., 2001; Biju, 2003). For
P. femoralis, nesting on the underside of leaves, rather than
on upper surfaces, may offer several advantages, such as
preventing predators from locating the clutch; reducing risk
of the eggs being dislodged by rain or other mechanical
damage; and reducing exposure to solar radiation.
The colour change that ground-nesting Philautus undergo
(Fig. 2 and 3) may have an adaptive function through reducing
the risk of predation, especially given that these frogs often
nest in daytime. On the other hand, the bright-green Philautus
femoralis, which deposits its eggs on the underside of leaves,
does not change colour.
The critical humidity dependence of Philautus eggs makes
them vulnerable to even short periods of desiccation. For
example, P. femoralis eggs at Agrapatana incubate for 37–49
days, subjecting the clutch to risk from stochastic climate
events. This danger is underlined by the fact that juveniles of
ground nesters were observed throughout the year and those
of Philautus femoralis, only during periods of sustained
rainfall. Climate warming could therefore place this fauna at
risk, though arboreal nesters would appear to be more
immediately threatened than ground nesters.
The strong correlation between female body size and number
of eggs (though not egg diameter) has been noted also in
other species of frogs (Duellman & Trueb, 1986). The lack of
correlation between female size and egg diameter and mass
however, could be the result of inconsistencies in our
measurements: eggs almost double in volume immediately
after release (through absorption of water from the soil),
rendering repeatable measurement difficult.
Comparative development. Embryos of Sri Lankan
Philautus show a degree of external morphological
convergence with Eleutherodactylus coqui. Embryos of
Philautus differ from those of E. coqui however, in having a
Fig. 7. a,female snout–vent length (FMSVL) shows a strong positive correlation (R2 = 0.822, n = 8) with male snout–vent length (MSVL)
for mating pairs; b, FMSVL shows a strong positive correlation with number of eggs (R2 = 0.920). Other reproductive variables (egg mass;
egg diameter; egg-mixing duration; nest covering duration; and incubation time) show only weak correlations (R2 = <0.296, n = 8).
D E
      
 
        
female snout-vent length
male snout-vent length
female snout-vent length average number of eggs
larger, spatulate, vascularized tail; absence of external gills;
presence of internal gill chambers; presence of a cement gland;
presence of a coiled gut; forelimbs concealed under opercular
fold up to an advanced stage of development (Fig. 5.10); and
the absence of an egg tooth (Fig. 5.13, 5.14) (see Townsend &
Stewart, 1985 for development in Eleutherodactylus coqui).
Philautus species also differ in the tempo and sequence of
development of some characters, such as early enlargement
of the tail (vs. relatively late enlargement in
The relatively large (compared to SVL) vascularized tail in
Philautus is evidently important for respiration since the
embryo lacks external gills or well-developed internal gills.
Kirtisinghe (1946) noted the presence of four pairs of atrophied
internal gills and concluded that it was the heavily vascularized
tail and not the gills that facilitated respiration in Philautus
embryos (see Fig. 6d). In Eleutherodactylus, the tail expands
only after the external gills are absorbed (see Fig. 1 in
Townsend & Stewart, 1985), reinforcing the importance of
the respiratory role of the tail in terrestrial direct developers.
The consistent orientation of the tail to the left in
Eleutherodactylus coqui has been remarked on by
Thibaudeau & Altig (1999: 178–179) and Wassersug (2000).
Although the tail in the larval specimens of P. silus described
above happen to be orientated to the left, we note that in
general the larvae of Sri Lankan Philautus have the tail
orientated both left and right, and occasionally even folded
under the body. Mature larvae were also seen to re-orientate
themselves within the egg, moving the tail from left to right
and vice versa. We note also Thibaudeau & Altig’s (1999:
179) questioning of the respiratory function of the vascularized
tail in E. coqui: “Whether or not there is sufficient increase in
vascularity to suggest a specific respiratory function for the
tail is debatable.” We suspect however, that in Philautus the
tail does indeed play an important role in respiration given
that the gills are internal and atrophied in all the Sri Lankan
larvae hitherto studied.
We suspect that the poorly developed cement gland in
Philautus is a primitive character. This ‘gland’, which is a
feature common in the early stages of the exotrophic larvae of
aquatic tadpoles, facilitates attachment to aquatic objects
(Duellman & Trueb,1986; Callery et al., 2001); such a function
is, of course, irrelevant to endotrophic larvae.
A coiled gut has not been observed in other lineages with
direct-developing embryos. In Sri Lankan Philautus, this
persists without degeneration up to stage 14. A coiled gut is
usually associated with aquatic tadpoles that rely on a
herbivorous diet, the elongated gut facilitating digestion and
absorption (Callery et al., 2001). The function of such a gut in
early stages of Philautus is difficult to explain: we suspect it
has a role in the absorption of the yolk associated with rapid
development, although it could simply be a primitive character.
Philautus have, however, lost many other larval characters
associated with aquatic tadpoles, such as keratinized denticles
around the mouth and a lateral-line sensory organ.
Conservation. – Little attention has been paid up to now to
the conservation of Sri Lanka’s Amphibia, despite the island’s
‘hotspot’ status (Myers et al., 2000; Meegaskumbura et al.,
2002) and the listing of some 44 species, including 27
Philautus, in the various ‘threatened’ categories of the IUCN
Red List (see More than 95% Sri
Lanka’s original rainforest extent of about 15,000 km2 has
disappeared during the past two centuries: only about 750
km2 now survives, mostly in the form of < 10 km2 fragments
surrounded by tea and rubber plantations, and human
settlements. The attrition caused by island effects (Ferraz et
al., 2003), taken together with edge effects, invasive alien
species, poor land management, industrial emissions,
irresponsible pesticide use, rainwater acidification and global
and local climatic change are likely to result in continued
species loss. This is particularly relevant because several
species are restricted to individual forest fragments. While
Sri Lanka’s human population density (300 km-2) is the highest
among the world’s 25 biodiversity hotspots (Cincotta et al.,
2000), human population density in the biodiversity-rich
south-western wet zone (to which all but one species of
Philautus are restricted) is more than twice the national
average. The pressures on forest fragments, in terms of
deforestation and habitat alteration, are therefore immense.
Given Sri Lanka’s extraordinary amphibian species richness, it is
imperative that every effort be made to conserve this remarkable
fauna, especially in view of its still being so poorly known: of the
17 species treated here, eight have been described only recently
(Manamendra-Arachchi & Pethiyagoda, 2005), while a further
five remain to be described. Philautus shows a high degree of
localized endemism within the Western Ghats-Sri Lanka
biodiversity hotspot. This fauna is gravely at risk, and its security
will depend on the conservation of habitats, a fuller understanding
of the ecology of the species involved and intensive measures,
both in and ex situ, to restore populations of threatened species.
Considering the difficulties associated with addressing many of
these threats, ex situ conservation may, in the short term, provide
the only realistic prospect for many Critically Endangered
species. We hope that the data in this paper will contribute to the
body of knowledge necessary for such conservation efforts,
which are now a matter of overwhelming priority.
We thank the Forest Department and the Department of Wildlife
Conservation for various permissions associated with this study;
Indraneil Das (University Malaysia, Sarawak) and an anonymous
reviewer for providing much valuable comment that helped
improve the manuscript; D. Kandamby (National Museum, Sri
Lanka), S. Nanayakkara (WHT) and Kelvin Lim (National
University of Singapore) for access to material; Tzi Ming Leong
and Donna O’Connell for discussions; Indraneil Das for
literature; S. Kankanam-Gamage, K. Wewelwala, S. Batuwita,
Kalana Maduwage and Anjana Silva for their generous help
caring for frogs at the WHT field station; James Hanken and
Ryan Kerney (Museum of Comparative Zoology, Harvard) for
access to and training in lab techniques; and S. D. Atukorale
(National Hospital, Colombo) for identifying pathogens. M.M.
Bahir et al.: Reproduction in Sri Lankan Philautus
acknowledges the Society of Systematic Biologists’ Graduate
Student Award (2002) that facilitated part of this work; and M.M.B.
acknowledges the Durrell Wildlife Conservation Trust for a
fellowship on endangered species management in 2000, and the
Wildlife Trust, USA, for its support of this research; and R. P.
acknowledges the Declining Amphibian Populations Task Force
for a grant that supported this work.
Alcala, A. C. & W. Brown, 1982. Reproductive biology of some
species of Philautus (Rhacophoridae) and other Philipine anurans.
Philippines Journal Biology, 11: 203–226.
Biju. S. D., 2001. A synopsis to the frog fauna of the Western Ghats,
India. Occasional Papers of the Indian Society for Conservation
Biology, 1: 1–24.
Biju. S. D., 2003. Reproductive mode in the shrub frog Philautus
glandulosus (Jerdon, 1853) (Anura: Rhacophoridae). Current
Science, 84: 283–284.
Bossuyt. F. & A. Dubois, 2001. A review of the frog genus Philautus
Gistel, 1848 (Amphibia, Anura, Ranidae, Rhacophorinae).
Zeylanica,6: 1–112.
Bossuyt, F. & M. C. Milinkovitch, 2000. Convergent adaptive
radiations in Madagascan and Asian ranid frogs reveal covariation
between larval and adult traits. Proceedings of the National
Academy of Sciences, 97: 6585–6590.
Bossuyt. F., K. Roelants, L. Spithoven & M. H. Daro, 2001. Philautus
bombayensis (Bombay Oriental Shrub-frog). Reproduction.
Herpetological Review,32: 34–35.
Bourne. R. B., 1997. Reproductive behaviour of terrestrial breeding
frogs Eleutherodactylus johnstonei in Guyana. Journal of
Herpetology, 31: 221–229.
Brown. W. & A. C. Alcala, 1994. Philippine frogs of the family
Rhacophoridae. Proceedings of California Academy of Sciences,
48: 185–220.
Callery, E. M., H. Fang & R. P. Elinson, 2001. Frogs without
polliwogs: evolution of direct development. BioEssays,23:233–
Cincotta, R. P., J. Wisnewski & R. Engelman, 2000. Human
populations in the biodiversity hotspots. Nature,404: 990–992.
Dring, J. C. M., 1979. Amphibia and reptiles from Northern
Trengganu, Malaysia with description of two new geckos:
Cnemaspis and Crytodactylus.Bulletin of British Museum (Natural
History), Zoology, 34: 181–241.
Dring, J., 1987. Bornean tree frogs of the genus Philautus
(Rhacophoridae). Amphibia-Reptilia,8: 19–47.
Duellman, W. E. & L. Trueb, 1986. Biology of Amphibians. McGraw-
Hill Book Co., New York. xix+670 pp.
Dutta, S. K. & K. Manamendra-Arachchi, 1996. The amphibian
fauna of Sri Lanka. Wildlife Heritage Trust of Sri Lanka, Colombo.
232 pp.
Ferguson, W., 1876. Singular Ceylonese frogs. Annals Magazine of
Natural History, (4), 18: 356–357.
Ferraz, G., G. J. Russell, P. C. Stouffer, R. O. Bierregaard, S. L. Pim
& T.E. Lovejoy, 2003. Rates of species loss from Amazon forest
fragments. Proceedings of National Academy of Sciences, 100:
Günther. A., 1859. Catalogue of the Batrachia Salientia in the
collection of the British Museum. Taylor & Francis, London.
xvi+160 pp, xii pl.
Günther. A., 1876. Note on the mode of propagation of some
Ceylonese tree-frogs, with description of two new species.
Annals Magazine of Natural History, (4), 17: 377–380, pl. 20.
Hanken. J., 1999. Larvae in Amphibian development and evolution.
In: Hall, B. K. & M. H. Wake (eds.), The origin and evolution of
larval forms. Academic Press, New York. Pp. 61–108.
Kanamadi, R. D., H. N. Nandihal, S. K. Saidapur & N. S Patil, 1996.
Parental care in the frog, Philautus variabilis (Günther) Journal
of Advanced Zoology, 17: 68–70.
Kirtisinghe, P., 1946. The presence in Ceylon of a frog with direct
development on land. Ceylon Journal of Science, (B), 23: 109–
Kirtisinghe, P., 1957. The Amphibia of Ceylon. Published by the
author, Colombo. xiii+112 pp, 1 pl.
Kuramoto, M. & S. H. Joshy, 2004. Two new species of the genus
Philautus (Anura: Rhacophoridae) from the Western Ghats,
southwestern India. Current Herpetology,22: 51–60.
Manamendra-Arachchi, K. & R. Pethiyagoda, 2005. The Sri Lankan
shrub-frogs of the genus Philautus Gistel, 1848 (Ranidae:
Rhacophorinae), with description of 27 new species. In: Yeo, D.
C. J., P. K. L. Ng & R. Pethiyagoda (eds.), Contributions to
biodiversity exploration and research in Sri Lanka. The Raffles
Bulletin of Zoology, Supplement No. 12: 163–303.
Marmayou, J., A. Dubois, A. Ohler, E. Pasquet & A. Tillier, 2000.
Phylogenetic relationships in the Ranidae. Independent origin of
direct development in the genera Philautus and Taylorana.
Comptes Rendus de l’Académie des Sciences, Serie III: Sciences
de la Vie —Life Sciences, 323(3): 287–297.
Meegaskumbura, M., F. Bossuyt, R. Pethiyagoda, K. Manamendra-
Arachchi, M. Bahir, M. Milinkovitch & C. J. Schneider, 2002.
Sri Lanka: an amphibian hot spot. Science, 298: 379.
Meegaskumbura, M. & K. Manamendra-Arachchi, 2005. Description
of eight new species of shrub frogs (Ranidae: Rhacophorinae:
Philautus) from Sri Lanka. In: Yeo, D. C. J., P. K. L. Ng &
R.Pethiyagoda (eds.), Contributions to biodiversity exploration
and research in Sri Lanka. The Raffles Bulletin of Zoology,
Supplement No. 12: 305–338.
Myers, N., R. A. Mittermeier, C. G. Mittermeier, G. A. B. da Fonseca
& J. Kent, 2000. Biodiversity hotspots for conservation priorities.
Nature,403: 853–858.
Ovaska, K. & A. S. Rand, 2001. Courtship and reproductive behavior
of the frog Eleutherodactylus diastema (Anura: Lepdodactylidae)
in Gamboa, Panama. Journal of Herpetology, 35: 44–50.
Patil, N. S. & R. D. Kanamadi, 1997. Direct development in the
rhacophorid frog, Philautus variabilis (Günther). Current Science,
73: 697–701.
Rogowitz, G. L., C. L. Candelaria, L. E. Denizard & L. J. Melendez,
2001. Seasonal Reproduction of a Neotropical frog, the cave
coqui (Eleutherodactylus cooki). Copeia,2001: 542–547.
Schwartz, J. J. & R. W. Henderson, 1991. Amphibians and reptiles of
the West Indies: descriptions, distribution and natural history.
University of Florida Press, Gainesville. 720 pp.
Stuart, S., J. S. Chanson, N. A. Cox, B. E. Young, A. S. L. Rodrigues,
D. L. Fischman & R. W. Waller, 2004. Status and trends of
amphibian declines and extinctions worldwide. Science, 306:
Thibaudeau, G. & R. Altig, 1999. Endotrophic anurans: development
and evolution. In: Mcdiarmid R.W. & R. Altig (eds.), Tadpoles:
the biology of anuran larvae. University of Chicago Press,
Chicago. Pp.170–188.
Townsend, D. S. & M. M. Stewart, 1985. Direct development in
Eleutherodactylus coqui (Anura: Leptodactylidae): a staging table.
Copeia, 1985: 423–436.
Townsend, D., S. & M. M. Stewart, 1994. Reproductive ecology of
the Puerto Rican frog Eleutherodactylus coqui.Journal of
Herpetology, 28: 34–40.
Wassersug, R. J., 2000. Tadpoles: the biology of anuran larvae. Copeia,
2000: 1125–1134.
Yong, H. S. & I. Ramly, 1987. Conquest of the land: direct
development in a Malaysian Philautus tree frog. Nature
Malaysiana 13: 4–7.
... Furthermore, during dry conditions, the frogs can retreat into moist litter to avoid desiccation. A thick litter layer also helps retain moisture in the sub-surface environment, which is essential for direct developing frogs such as P. zorro that lay eggs in moist soil (Bahir et al. 2005). The type and quantity of litter may also affect arthropod availability (Heinen 1992), which is the main food source of amphibians (Solé and Rödder 2010). ...
... Higher temperatures can increase evaporation from leaf litter (Kreye et al. 2018) and influence the distribution of P. zorro, which depends on the leaf litter moisture for survival and reproduction. In general, reduced moisture levels can reduce the survival of adult terrestrial amphibians (Rohr and Madison 2003) and desiccate the direct developing eggs (Bahir et al. 2005). This assertion is further strengthened by the observation that during the comparatively dry months of the study period, P. zorro was rarely found at sites in which they had previously been recorded. ...
Full-text available
Basic ecological information is of critical importance for the development of conservation strategies for endangered amphibian species. The Gannoruwa Shrub Frog (Pseudophilautus zorro) is an endangered endemic and a geographically restricted frog found in mid elevations of the central highlands of Sri Lanka whose ecology is little known. We assessed: (1) aspects of microhabitat associations; (2) the realized niche to identify potential areas of occupancy using Ecological Niche Modeling (ENM); and (3) estimated the Extent of Occurrence (EOO) and Area of Occupancy (AOO) of P. zorro with new locality data and predicted data. The results indicate that there is a significant positive relationship between leaf litter thickness and number of individuals of P. zorro in quadrats and a significant relationship between percentage of canopy cover and number of individuals of P. zorro in quadrats. Pseudophilautus zorro is more abundant in broad-leaf forests than pine forests. ENM predicted highly suitable areas (suitability > 80%) in the Kandy, Matale, and Kegalle districts in Sri Lanka. The EOO and AOO of P. zorro with new occurrence data was 558 km 2 and 9 km 2 , and with predicted data, 2,374 km 2 and 612 km 2 respectively, indicating an increase in EOO and AOO. The data on microhabitat preferences and predicted habitats will be invaluable to formulate in-situ and ex-situ conservation strategies for this species.
... Forelimbs then erupt through the operculum and continue growing exposed until hatching. This is the case in direct-developing arthroleptids, ceratobatrachids, myobatrachids, rhacophorids, and the Neotropical microhylid Myersiella (interestingly, not Australopapuan asterophryines) (e.g., Alcal a, 1962;Anstis, 2008;Anstis et al., 2011;Bahir et al., 2005;Izecksohn et al., 1971;Schweiger et al., 2017). The full forelimb concealment in these groups may last 1-6 developmental stages (e.g., TS9 to TS10 in Arthroleptis spp., TS5 to TS11 in Metacrinia nichollsi; Anstis, 2008;Schweiger et al., 2017). ...
In frogs, the loss of the free‐living tadpole stage implies deep transformations in morphology and development, and the result is a combination of traits exclusive to direct development with the reduction or absence of typically larval features. Two traits stand out in terraranans, a mostly Neotropical clade with ca. 1200 species known or suspected to have direct development: (1) the development of forelimbs, which unlike in tadpoles and in several unrelated direct‐developing lineages, in this clade is described to occur completely exposed during ontogeny, and (2) the reduction of gills and their physiological replacement by highly vascularized tails for gas exchange. In this contribution, we report unusual aspects of the embryonic ontogeny in the Confused Robber frog Oreobates berdemenos, which complicate the interpretation of morphological diversity in Terrarana and in direct‐developing frogs in general. First, forelimbs develop initially entirely concealed by a larval operculum. Second, tail fins develop dorsal and ventral to the tail muscular core but their arrangement changes to lateral during ontogeny. Third, the expansion of the tail fins appears to be a plastic character whose variation likely responds to individual conditions of humidity and/or oxygenation of each egg within the clutch. Finally, we report cells with the morphology of hatching gland cells in the rostral region of embryos at stages about to hatch. If not vestigial, these cells could assist the egg tooth, which is characteristic of terraranans, with chemical softening of the egg capsules during hatching. We report morphological aspects of the embryonic ontogeny in the Confused Robber frog Oreobates berdemenos, which stand out from what is known in other related direct‐developing frogs. For instance, the tail fins in terraranan embryos are well‐developed and vascularized, and likely compensate the reduction or absence of gills. In O. berdemenos, tail development is also polymorphic, with some tail fins enveloping the embryo completely and some only its posterior half. Our preliminary experiments suggest that fin expansion could respond to individual conditions of humidity and/or oxygenation of eggs within the clutch: a fully enveloping tail would provide a larger and more efficient surface for gas exchange under dry conditions. Further unusual features in this species are a complete operculum concealing part of the forelimb development, and putative hatching gland cells present at hatching. These features, not described to date in terraranan frogs, represent embryonic/larval aspects retained from ancestral biphasic ontogenies.
... In particular, among Caudata, most species of the family Plethodontidae experience it (Wake and Hanken, 1996;Marks and Collazo, 1998;Kerney, 2011). Among Anura it is found in many families, including Brevicipitidae, Eleutherodactylidae, Hemiphractidae, Leiopelmatidae, Megophryidae, Microhylidae, Rhacophoridae, etc. (Bahir et al., 2005;Callery et al., 2001;Anstis et al., 2011;Gomez-Mestre et al., 2012). It is also characteristic of some Gymnophiona (Wake, 1992). ...
Full-text available
The relative independence of ontogenetic processes, which is characteristic of amphibians, determines the possibility of shifts in the rate and timing of the ontogenetic events, or heterochronies. Heteroch-ronies give rise to ontogenetic and morphological diversity in amphibians, which can occur without significant genetic changes. Heterochrony-related neoteny, miniaturization and paedomorphic underdevelopment, as well as loss of ancestral traits due to their transition into a latent capacities state with the possibility of secondary recapitulation are phenomena typical of both recent and fossil amphibians.
... Even though the taxonomy and phylogeny of amphibians (Manamendra-Arachchi & Meegaskumbura, 2012;Wickramasinghe et al., 2013) are well studied in Sri Lanka, relatively little attention has been paid to ecological and behavioral characterization. Other than for vocalization (Samarasinghe, 2011) and some information on breeding (Bahir et al., 2005;Karunarathna & Amarasinghe, 2007, 2010, literature on the biology and ecology of P. popularis is not available. The Common shrub-frog Pseudophilautus popularis (Manamendra-Arachchi & was described in 2005 and information on taxonomic features and distribution has been published. ...
This study reports activity budget and perch characteristics of the Sri Lankan endemic shrub frog Pseudophilautus popularis in a wetland-home garden setting in an urban area. Thirty-two frogs were studied from 18:30–06:00h and 45% of the time they were found to be stationary without any activity. The highest percentage of time was utilised for calling and courtship (52%). This study reveals that the frogs occupy different heights during their nightly activity period, starting from the ground level and gradually moving up among the vegetation. They were reported to reach the maximum heights around 23:00h, remaining there for nearly two hours and then retreating downwards towards the dawn. The results emphasize the importance of maintaining diversity of vegetation especially in terms of stratification in ensuring habitat quality to conserve this species.
... However, within the order Anura it is estimated that there are more than 1000 species of directdeveloping frogs, in which the whole development, from the embryo to the adult-like froglet, takes place inside the egg-grouped into the Neotropical superfamily Brachycephaloidea (see Frost, 2021). In addition, direct development, it seems, has evolved independently several times in Anura (Duellman & Trueb, 1986) and is observed among diverse families in addition to those contained in Brachycephaloidea (Drewry & Jones, 1976;Wake, 1978;Townsend & Del Pino, 2019); Mantellidae (Vences & Glaw, 2001); Microhylidae (Zweifel, 1972); Myobatrachidae (Roberts 1984;Anstis, 2008); and Rhacophoridae (Bahir et al., 2005). ...
Anurans have served as models in evolutionary studies on reproduction and development. Several generalizations have emerged but it is unclear whether they represent the diversity of the entire group. We studied the development of two species of the Neotropical family Cycloramphidae. Frogs in this family have peculiar spawning sites and developmental biologies. Whereas some cycloramphid tadpoles are exotrophic, quasi-terrestrial and thrive in films of water that flow over rocky outcrops, others are endotrophic, terrestrial and complete their development outside the eggs laid on the forest floor, under logs or in crevices. Cycloramphid tadpoles have aspects of their morphology distinct from those of pond and stream dwellers; they hatch in more advanced stages of development and have typical tadpole structures, either in a modified form or absent. Herein we present a study of the tadpoles of C. lithomimeticus and T. miliaris, in order to document and discuss alterations in the ontogeny of several larval structures. These changes seem to be associated with adaptations to life in the peculiar habitats where the tadpoles thrive. We infer that the observations we present are explained by development being associated with aspects of terrestrialization and endotrophism.
... Daily observations were made to note the progression of the developmental stages of the eggs. The stages of the eggs were categorised by comparing the eggs to photographs of thr various stages provided by Bahir et al. (2005). ...
Full-text available
Many frogs undergo direct development, in which the embryo develops within the egg and lacks a tadpole stage, being hatched morphologically similar to the adults. Such reproductive behaviour ensures the evolutionary success of their lineages without the limitation of needing a moist substrate for egg deposition. The species in the genus Raorchestes are characterised as small-sized frogs that undergo terrestrial direct development. However, observations of the reproductive behaviour of many species in Raorchestes are lacking. Here, we report the first examination of bush frogs, Raorchestes longchuanensis, from southern China undergoing direct development. Under laboratory conditions, we found that R. longchuanensis can lay two clutches in a single month (as two were laid in the month of June). Only 10–12 eggs were laid per clutch. The hatching period of R. longchuanensis was 25–30 days.
... Direct development evolved numerous times during amphibian phylogeny (e.g., Schweiger et al. 2017), and in this context, it is reasonable to expect different sets of clade-specific features. This is already reported in several character systems exhibiting high variation in the presence, morphology, and developmental timing across anuran groups, such as gasexchange structures (e.g., gills, opercules, tails, abdominal sacs), hatching structures (egg tooth, hatching cells), adhesive glands, vent tube, and limbs (e.g., Hanken et al. 1992;Bahir et al. 2005;Anstis et al. 2007;Narayan et al. 2011;Goldberg and Vera Candioti 2015;Schweiger et al. 2017). Skeletal ontogeny seems to follow the same pattern, with heterochronic shifts likely different among groups. ...
Direct development implies transformations with respect to the anuran biphasic life cycle, including changes in embryonic anatomy. In the clade Brachycephaloidea, skeletal ontogeny is known in Eleutherodactylus coqui of the basal family Eleutherodactylidae. In this work, we study it in four species representing the two other families in the group. We worked with developmental series of Oreobates barituensis, Haddadus binotatus, Ischnocnema henselii, and Brachycephalus ephippium. Specimens were prepared following protocols of clearing and staining and histology. In the cranium, results show an overall shared pattern that, as summarized for E. coqui, combines a partial recapitulation of aspects of the ancestral biphasic ontogeny with a profound remodeling that includes lost/novel structures and heterochronic shifts of developmental events. Among these transformations are the absence of suprarostrals and trabecular horns and the precocious ossification of jaw and suspensorium. In addition, each lineage shows particular features such that skull ontogeny varies interspecifically. In turn, the morphogenesis of the axial and appendicular skeleton is highly conserved, with main variations including the extent of ossification at hatching. Along with some external features such as the egg tooth and the enveloping tail with transversely arranged fins, an ossification sequence with extremely accelerated ossification of jaws and suspensorium could be distinctive of Brachycephaloidea.
en Bush frogs of the genus Philautus are a species‐rich group of the Asian tree frogs Rhacophoridae, which are known for their diverse reproductive biology. Within Philautus, reproduction has been described via endotrophic tadpoles and by direct terrestrial development. Here, we provide results of phylogenetic analyses based on the most comprehensive sampling of Bornean Philautus to date. As a result of an integrative taxonomic study using mitochondrial and nuclear markers, along with morphological and bioacoustic data, we describe a spectacular new species of this genus from the island of Borneo. The ecology of the new species of Philautus is closely associated with the carnivorous pitcher plant, Nepenthes mollis. The unusually large eggs are laid in the fluid of the pitcher and the endotrophic tadpoles, characterized by reduced mouthparts, small oral orifice and large intestinal yolk mass, complete their development in this environment. Molecular data and synapomorphic larval characters support the sister group relationship of the new species to P. macroscelis: both belong to the early diverged lineages in the Philautus tree, whose phylogenetic relationships could not be fully resolved. The new record of endotrophic tadpoles challenges again the hypothesis that terrestrial direct development is the plesiomorphic mode in this genus. Further, we discuss the nature of the frog‐plant interaction that could represent a new case of mutualism. The frog provides the plant with a source of nitrogen by depositing yolk‐rich eggs in the liquid of the pitcher. The plant, on the other hand, offers an exclusively protected space for the development of tadpoles in a habitat that otherwise has few permanent bodies of water and many competing frog species. Abstrakt de Die artenreiche Gattung Philautus gehört zu den Ruderfröschen (Rhacophoridae), die für ihre vielfältige Fortpflanzungsbiologie bekannt sind. Innerhalb von Philautus wurde bisher die Fortpflanzung mit endotrophen Kaulquappen und durch Direktentwicklung an Landnachgewiesen. In der vorliegenden Arbeit stellen wir die Ergebnisse phylogenetischer Analysen vor, die auf dem bisher umfangreichsten Sample von Philautus von der Insel Borneo basieren. Als Ergebnis dieser integrativen taxonomischen Studie unter Verwendung mitochondrialer und nukleärer Marker, in Kombination mit morphologischen und bioakustischen Daten, beschreiben wir eine spektakuläre neue Art dieser Gattung. Die Ökologie dieser Art ist eng mit der fleischfressenden Kannenpflanze Nepenthes mollis verbunden. Die ungewöhnlich großen Eier werden in der Flüssigkeit der Kannenpflanze abgelegt. Die endotrophen Kaulquappen, die durch ein reduziertesMundfeld, eine kleine Mundöffnung und eine große Dottermasse im Darm gekennzeichnet sind, vollenden ihre Entwicklung in den Kannen. Molekulare Daten und synapomorphe larvale Merkmale unterstützen die Schwestergruppenbeziehung der neuen Art zu P. macroscelis: beide gehören zu den basalen Linien in der Gattung Philautus, deren phylogenetische Beziehungen allerdings nicht vollständig aufgelöst werden konnten. Der neue Nachweis von endotrophen Kaulquappen bei Philautus stellt erneut die Hypothese in Frage, dass Direktentwicklung der plesiomorphe Fortpflanzungsmodus in dieser Gattung ist. Weiterhin diskutieren wir die Art der Frosch‐Pflanze‐Interaktion, die einen neuen Fall von Mutualismus darstellen könnte. Der Frosch bietet der Pflanze eine Stickstoffquelle, indem er dotterreiche Eier in der Flüssigkeit der Kanne ablegt. Die Pflanze wiederum liefert einen exklusiven, geschützten Raum für die Entwicklung der Kaulquappen in einem Habitat, das ansonsten wenige permanente Gewässer und viele konkurrierende Froscharten aufweist.
Amphibians are among the world's most endangered category of vertebrates. 67% of amphibians in Sri Lanka are threatened with extinction due largely to anthropogenic activity. A basic knowledge of habitat requirements and a fine-scale distribution of species is important to the survival of amphibians. Pseudophilautus zorro is commonly referred as the Gannoruwa shrub frog.
The normal development of the brown frog Rana sakuraii is observed from oocyte to completion of metamorphosis in the laboratory at 15±2°C. Tadpoles of this species can develop completely without any food. Their development is divided into 40 stages and this paper provides descriptions and illustrations of the external features at each stage. After hatching, some developmental events occur simultaneously and elongation of hindlimb buds and toe formation are relatively rapid. On the average, the tadpoles reach the stage 36 with their maximum total length 22.8 mm after 43.9 days. They complete metamorphosis after 56.5 days, reaching the final stage 40 with 8.0 mm in snout to vent length. The labial teeth development of this non-feeding tadpole is generally weak and the formula at a given stage is greatly variable among individuals.
Full-text available
Literature survey indicates that more than 200 species of anurans are known from India. However, no serious attempts have been made to comprehensively study the amphibian fauna of this region. The Western Ghats of India, a global hotspot, constitutes one of the amphibian rich areas in tropical Asia. Of the 206 species of anurans reported from India, 104 are known to occur in the Western Ghats. A seven-year field survey and taxonomic study of anurans of this region resulted in collecting more than 200 species, about twice the above number. Published works have reported only 52 species of the Old World treefrogs from India (Chirixalus: 4, Nyctixalus: 1, Philautus: 25, Polypedates: 5, Rhacophorus: 16 and Theloderma:1) but the present study reveals that more than 100 ‘species’ of this group are present in the Western Ghats alone. Four new genera and about 115 new species from the Western Ghats are still awaiting scientific identity Most of them belong to the shrubfrog genus, Philautus (Biju, unpublished data). Field survey was conducted giving special attention to the breeding behaviour and reproductive biology The life cycle of 31 species of anurans was studied: they include one species each of Euphlyctis, Hoplobatrachus, Micrixalus, Microhyla, Minervarya, two species of Indirana, three species of Nyctibatrachus, and 21 species of rhacophorids/nes. Rhacophorids/ nes studied include Rhacophorus calcadensis, R. lateralis, R. malabaricus, Polypedates maculatus, P. ‘pleurostictus’, P. pseudocruciger, four undescribed species of Philautus, two undescribed species of Polypedates and nine other undescribed species belonging to different genera in Rhacophoridae/nae.
During breeding season gravid females Philautus variabilis were attracted by calling males. They bave moved towards the males. Male suddenly grasped the female resulting in an axillary amplexus. After 8-10 hours females began to ovulate. 60-62 large sized unpigmented eggs were deposited in single clutch. Soon after the spawning the females began to collect the spreadout eggs beneath the abdomen using fore and hind limbs. Meanwhile, male moved towards the female. Females aggressively attacked the male by their forelimbs to chase away the intruder. Under the laboratory conditions females guarded the eggs for three hours. P. variabilis exhibits parental care by egg attendance.
I studied courtship, amplexus, and oviposition behavior of leaf litter and cavity breeding frogs (Eleutherodactylus johnstonei) in Georgetown, Guyana, using modified plastic planters equipped with observation windows. Both sexes used pots with and without plants as diurnal retreats, but there was a preference for pots with plants. Calling males showed no preference for pots with or without plants. Females preferred pots with plants for oviposition. A gravid female initiated courtship by touching a calling male, who then led the female to one or more potential oviposition sites. The female either accepted or rejected cavities as nest sites, and sometimes abandoned the male if the nest cavities were unacceptable. Unacceptable cavities had direct contact with soil. Only 45% of the mating pairs employed axillary amplexus; males usually just sat on the females' backs. Twenty-five percent of the amplectic frogs then employed the reverse hind leg clasp after 0.50-1.75 h in the nest cavity. Oviposition lasted for 0.75-1 h, and approximately 14 externally fertilized, unpigmented eggs were laid in a pyramid. Fecundity was positively correlated with female size. The time from initial contact of the male by the female to egg laying was 4-6 h. Eggs hatched in approximately 13 d, and newly hatched froglets had snout-to-vent lengths of about 3.5 mm. Clutches were found during all months of the year; however, most clutches were deposited during the wettest months of the year. On average, males sired 3.3 clutches/year, while females produced 4.3 clutches/year. The most striking feature of E. johnstonei's reproductive behavior is that a few pairs used a reverse hind leg clasp during amplexus; a behavior only known for E. coqui that has internal fertilization.
The Puerto Rican frog Eleutherodactylus coqui has terrestrial reproduction and direct development, as is the case for all known species of this large, neotropical group. To facilitate field measurements of embryonic age in E. coqui, a staging table was developed based on changes in the external morphology of embryos. Pre-hatching development was divided into 15 stages. Principal diagnostic features include formation of external gills, progressive pigmentation of the eyes, growth and differentiation of the limbs, development of endolymphatic calcium deposits, expansion of a pigmented body wall, growth of a large, vascularized tail and changes in embryonic behavior. We have used the table to stage the eggs of E. antillensis and E. portoricensis, related Puerto Rican species. A literature review suggests that our table may be useful as a general staging scheme for the genus.