Bioacoustics of Hylarana celebensis (Peters, 1872) (Anura: Ranidae) From
Sulawesi
[Bioakustik Kodok Hylarana celebensis (Peters, 1872) (Anura: Ranidae) Asal
Sulawesi]
Hellen Kurniati
Zoology Division of Research Center for Biology, Indonesian Institute of Sciences (LIPI),
Widyasatwaloka Building-LIPI, Jalan Raya Cibinong Km 46, Cibinong 16911, West Jawa.
E-mail: hkurniati@yahoo.com
Memasukkan: Januari 2015, Diterima: Juni 2015
ABSTRACT
Hylarana celebensis (Peters, 1872) is an endemic frog to Sulawesi, the species being a member of family
Ranidae. The presence of the frog in its habitat is easily detected from its advertisement call; males usually call
in a chorus, they call to each other in a large group. Bioacoustic of calls of a typical individual male of H.
celebensis have not been described in detail, although it is very easy to find this species in freshwater swamps,
permanent ponds, or slow-flowing waters in the lowland areas. The purpose of the bioacoustic analysis on H.
celebensis’s calls that were recorded at Bahodopi area is to build a reference collection to be compared with H.
celebensis’s bioacoustics from other regions in Sulawesi. Because of the wide distribution of this frog in
Sulawesi; genetic structure of H. celebensis population may also follow the genetic structure of Ingerophrynus
celebensispopulation. Calls of H. celebensis have two types of calls, i.e. pure tone and pulse; however, pure
tones have three variation, namely pure tone type 1, pure tone type 2 and pure tone type 3; however pulsed call
has only one type.
Keywords: Anura, Hylarana celebensis, bioacoustics, Sulawesi.
ABSTRAK
Hylarana celebensis ( Peters, 1872) adalah kodok endemik Sulawesi, yang mana jenis ini adalah anggota dari
suku Ranidae. Keberadaan kodok H. celebensis di habitatnya sangat mudah terdeteksi melalui suara
panggilannya; jantan biasanya bersuara dalam kelompok yang ramai, mereka bersahut-sahutan satu sama lain
dalam kelompok besar. Bioakustik suara panggilan individu jantan H. celebensis yang khas belum
diinformasikan secara detail, meskipun sangat mudah untuk menemukan jenis kodok ini di rawa air tawar,
kolam permanen atau perairan berarus lambat di daerah dataran rendah Sulawesi. Tujuan dari analisis
bioakoustik kodok H. celebensis yang direkam di wilayah Bahodopi adalah sebagai acuan dalam pengungkapan
bioakoustik kodok H. celebensis dari wilayah lain di Sulawesi, karena distribusi kodok ini meluas ada di
Sulawesi, kemungkinan besar struktur genetik dari populasi H. celebensis juga mengikuti struktur genetik
populasi kodok Ingerophrynus celebensis. Suara panggilan dari jantan H. celebensis memiliki dua tipe
gelombang suara, yaitu nada murni (pure tone) dan nada pulsa (pulse). Nada murni memiliki tiga variasi, yaitu
nada murni tipe 1, nada murni tipe 2 dan nada murni tipe 3, sedangkan gelombang suara nada pulsa hanya
memiliki satu tipe.
Kata Kunci: Anura, Hylarana celebensis, bioakustik, Sulawesi
INTRODUCTION
Hylarana celebensis (Peters, 1872) is a
frog species endemic to Sulawesi, and is a
member of the family Ranidae. Adult male and
female of H. celebensis have slim body, which
is the length of the body (SVL) of both sex are
not much different; the average SVL of adult
males that were collected from Bahodopi area,
Central Sulawesi (Figure 1) was 42.5 mm
(number of individuals 16), while the average
SVL of adult females was 48.6 mm (number of
individuals 5).
According to Iskandar & Mumpuni (2004),
the distribution of H. celebensis in Sulawesi covers
an area of low-lying west coast of Central Sulawesi,
lowland areas west and east coasts of North Sulawesi
and Gorontalo to Sorowako region in South
Sulawesi. The vertical distribution of this frog covers
lowlands below 700 m elevation above sea level
(asl). With more additional locations that have been
surveyed in Sulawesi, information on distribution of
H. celebensis has become more widespread; the frog
was not just found in the region of Central Sulawesi
2
Hellen Kurniati
and North Sulawesi, but also found in the east
coast of Central Sulawesi (including Bahodopi
area) and in the lowland area of Southeast Sulawesi.
Wanger et al. (2011) found H. celebensis in Lore
Lindu National Park area; Gillespie et al. (2005) also
found the frog on Buton Island and other islands
around the island.
The presence of H. celebensis in its natural
habitat is easily detected from its advertisement call;
males usually call in a chorus, they respond to the
advertisement call in a large male group.
Bioacoustics of calls that a typical of adult male of H.
celebensis have not been described in detail, although
it is very easy to find this frog in its natural habitats
such as freshwater swamps, permanent ponds or slow
flowing waters in low land areas, which are typically
found in Bahodopi area in Central Sulawesi Province
(Figure 2). The main objective to describe
bioacoustics of H. celebensis that was originated
from Bahodopi area use as a basic reference in
comparison with bioacoustics of H. celebensis from
the other regions in Sulawesi; because the frog has
wide distribution in Sulawesi, and most likely the
population genetic structure of H. celebensis may
also follow genetic structure of the population
Ingerophrynus celebensis, as revealed by Evans et al.
(2008). Possible relationships between bioacoustics
of I. celebensis and genetics have not been proven.
However that the genetic structure of the population
does affect the pattern of bioacoustics was proven in
Duttaphrynus melanostictus (Wei et al. 2012) and
also for Polypedates leucomystax (Sheridan et al.
2010).
METHODOLOGY
Calls of four adult males H. celebensis
were recorded in the Bahodopi area (S
02048'37.4 "; E 122004'58.5"; 258 m asl),
Central Sulawesi. Date of recording on 9-10
April, 2013 by using a Sony PCM-M10 recorder
using a sampling frequency of 94 kilohertz (kHz) and
a bit rate of 24 bits, with the air temperature of 27.4°
C. In the lab, recorded calls were first normalized and
then converted to 48000 Hz (= Hertz) and 16 bits by
using Adobe Audition software version 3.0 to ensure
highest possible recording quality for frequencies up
to 24 kHz. Adobe Audition 3.0 software was also
used to describe the oscillogram and audiospec-
trogram of each call type of H. celebensis.
Visualization process of sound waves in the form of
oscillograms, audiospectrograms and the dominant
frequency were taken by using FFT (Fast Fourier
Transformation; 1024 points) at using a Hanning
window. The terms were used in the description of
the calls followed to Pettitt et al. (2012). Ratio of
Coefficient Variant (CV) was calculated to determine
the "static" and "dynamic" of vocalizations (Gerhardt
1991); CV calculations followed Krebs (1989).
RESULTS
Calls of H. celebensis have two types, i.e.
pure tone and pulses (Figure 3 and 4); calls of
two males were recorded for the type of pure
tone and also two males for the type of pulse.
Pure tone calls have three variations, namely
pure tone type 1, pure tone type 2 and pure tone
type 3 (Figure 3); these three types of pure tone
were released randomly in five minutes long
recording and the tones are frequency modulated in
various ways (see Figure 3); however pulsed
call has only one type (Figure 4). Three types of
pure tones were released by the males in
irregular tempo, while the pulses were released
with a relatively in regular tempo; therefore
Figure 1. Adult male of Hylarana celebensis from
Bahodopi area, Central Sulawesi.
Figure 2. Bahodopi area (black circle) which is located at
east coast of Central Sulawesi Province.
3
Bioacoustics of Hylarana celebensis (Peters, 1872) (Anura: Ranidae)
pulse rate and call rate of pulsed call type can be
calculated by following Pettitt et al. (2012).
Both types of sound waves are described in
detail below:
I. Pure Tone Calls
I.1. Pure tone type 1
Visualization results of the audio spectrogram
show the amplitude of this type is modulated with
harmonic number is between 1-2; first harmonic (or
fundamental frequency) is strong whereas the second
harmonic is weak (Figure 3). The average value of
the first harmonic is 1563.66 ± 248.61 Hz (CV =
15.90%), with a frequency range is between 937-
2015 Hz; while the average value of the second
harmonic is 2777.95 ± 521.11 Hz, with a frequency
range is between 1593-4078 Hz (see Table 1). Up to
5 harmonics may be present in the signal. The
frequency range of the first and second harmonic is
very wide (> 1000 Hz), this value indicates that the
filtering of the vocal apparatus seems to be rather
broad, most signal energy was between 937 Hz and
2015 Hz. The average duration of the pure tone type
1 is 40.71 ± 11.06 ms (CV = 27.17%), with a range
of durations between 23-91 ms. The duration of the
pure tone type 1 is shorter than the duration of the
two other types of pure tones (see Table 1).
I.2. Pure tone type 2
Results obtained from the audio spectrogram
show that frequency of this type is an ascending
modulation at an early stage and then descending
modulation at the end; the number of harmonics is
between 2-3, with the second harmonic being strong
while the two others are weak (Figure 3). The
average value of the fundamental frequency (or
first harmonic) is 1366.32 ± 248.25 Hz (CV =
18.17%), with a frequency range is between 483
-1734 Hz; the average value of the second
harmonic is 2139.74 ± 384.67 Hz, with a
frequency range is between 1593 to 2695 Hz;
while the average value of the second harmonic
is 2904.58 ± 401.92 Hz, with a frequency range
is between 1828-3304 Hz (see Table 1). The
frequency range of the fundamental frequency
and the first harmonic and the second is very
wide (> 1000 Hz), this value also indicates that
the filtering of the vocal apparatus seems to be
rather broad, most signal energy was between
483 Hz and 1734 Hz . The average duration of
the pure tone type 2 is 80.63 ± 30.35 ms (CV =
37.64%), with a range duration is between 37-
151 ms.
I.3. Pure tone type 3
Visualization results of audiospectrogram
show that frequency of this type is slightly
modulated and has 4-7 harmonics, which is the
first and the second harmonic are strong (Figure 3).
The average value of the fundamental frequency is
1120.67 ± 196.98 Hz (CV = 17.58%), with a
frequency range is between 937-1453 Hz; the
average value of the first harmonic is 1839.50 ±
216.56 Hz, with a frequency range is between 1453
to 2015 Hz (see Table 1); while the average value of
the second harmonic is 2640.17 ± 213.18 Hz, with a
frequency range is between 2390 to 2953 Hz (see
Table 1). Frequency range of the fundamental
frequency and the first harmonic and the second is
not wide (approximately 500 Hz), this value indicates
that the filtering of the vocal apparatus seems to be
not broad, most signal energy was between 1453
Hz and 2015 Hz. The average duration of the
pure tone type 3 is 104.17 ± 34.08 ms (CV =
Figure 3. Oscillogram and audiospectrogram of pure tone calls of Hylarana celebensis. From left to right: pure
tone type 1, pure tone type 2 and pure tone type 3.
4
Hellen Kurniati
32.71%), with a range duration is between 62-
157 ms. The duration of a pure tone type 3 is the
longest compared to the other two types of pure
tones (see Table 1).
Three types of pure tone calls have
irregular tempo. This character indicates that the
pure tone calls are not spontaneous or clicking
vocalizations, the calls can be controlled by the
males. The average of fundamental frequency
decreases successively from pure tone type 1 to
pure tone type 3, but the number of harmonics
and duration of the pure tone increases from
pure tone type 1 to pure tone type 3 (see Table
1). Two males released three types of pure tone
might be caused by pulse background noise in
which a large group of male H. celebensis were
calling in chorus, because the two males were
about one meter outside the swamp habitat. This
phenomenon occurs on Hylarana (= Rana)
taipehensis that live near airports in Thailand, in
which individual males will release calls of
different frequencies in a noisy environment due
to human activity (Sun & Narins 2005);
however pure tone variations that were released
by H. celebensis was natural, because there was
no anthropogenic noise (resulting from human
activities) in the environment around H.
celebensis habitat during calls recording.
Visualization of the three types of pure tone
looks similar to visualization of pure tones that
were released by H. taipehensis (Sun & Narins
2005). Pure tone type 1, type 2 and type 3 of H.
celebensis are similar to pure tone type (g), type
(e) and type (f) of H. taipehensis (see Figure 3
and Figure 6), however both frogs species have
different fundamental frequencies; fundamental
frequency of H. celebensis lower than the
fundamental frequency of H. taipehensis.
The three types of pure tones of H.
celebensis are dynamic calls; it is indicated by
the CV values more than 12% at the
fundamental frequency; as well as the duration
of each type of pure tones. These CV results
indicate that the intensity and duration of pure
tones calls can be controlled by male of H.
celebensis. This phenomenon also belongs to
Hylarana chalconota (Marquez & Eekhout
2006) and Phrynoidis aspera (Kurniati 2013).
II. Pulsed Calls
There is only one type of pulsed call of H.
celebensis (Figure 4a). The call type is a regular
call that is often advertised in chorus by males
H. celebensis. The pulse type of H. celebensis
has two sets of pulses, which is one call consists
of 3-5 pulses and a pulse is also composed of
many sub-pulses (> 10 sub-pulses) (see Figure
4b). As a result of this rapid and varying
amplitude modulation the spectrogram shows
merely ‘noise’ (Figure 4) in which no harmonic
structure can be discerned. To the ear such
sounds appear scratch-like, whereas pure tone
sounds as described above have a pleasant
sound quality to the human ear.
The average number of pulses in every call
is 4 ± 0.31, with a range of pulse is between 3-5.
The average duration for the pulse is 131.87 ±
10.57 ms (CV = 8.01%), with a range duration
is between 100-164 ms; while the average
duration of a call is 788.23 ± 70.09 ms (CV =
8.89%), with a range call duration is between
548-963 ms (see Table 1). The average duration
between pulses is 210.17 ± 16.77 ms (CV =
7.98%), with a range duration between pulse is
182-250 ms; pulse rate that were taken from the
average duration is 4.76/sec. The average
duration between call is 9699.05 ± 2034.10 ms
(CV = 20.97%), while the range of durations is
between call is 7358-16390 ms; call rate that
was taken from the average duration is 0.10/sec.
Audio spectrogram visualization of pulse
shows that there was no dominant frequency,
but it is a span with a wide spectrum (Figure
4a). Average frequency of the lower spectrum is
1636.08 ± 151.72 Hz (CV=9.27%), with a range
frequency is between 1125 to 1968 Hz, while
the average frequency of the upper spectrum is
3326.15 ± 322, 01 Hz (CV=9.68%), with a
range frequency is between 2906-4500 Hz;
average width spectrum of the pulses is 1690.21
± 384.62 Hz (CV=22.75%), with a range
frequency is between 1032 to 2860 Hz (Table
1).
DISCUSSION
The pulsed call type is a common call
which is released regularly by male H.
celebensis ; this call is usually used as a
5
Bioacoustics of Hylarana celebensis (Peters, 1872) (Anura: Ranidae)
reciprocal call among males in a large group in
its natural habitat, include grassy swamp,
permanent pool and slow-moving stream.
Pulsed calls type have relatively regular tempo
that is shown by the value of CV<12%, the
duration one pulse (CV=8.01%), the duration
between pulse (CV=7.98%) and the duration of
one call (CV=8.89%). CV value is more than
12% in the duration between calls (CV=
20.97%) indicates that there is an irregularity
tempo on calls; this condition is common in frog’s
bioacoustics, because the releasing time of the call on
frog depends on many environmental factors (Goutte
et al. 2013).
Visualization results of H. celebensis’ pulsed
call shows that there is no dominant frequency on the
audio spectrogram; therefore the analysis was carried
out in another way by looking lower frequency and
upper frequency of the spectrum. The width of the
spectrum is the reduction result of the upper
frequency to the lower frequencies (Figure 5). Lower
and upper frequencies of the pulse are static with
CV<12%, but the width of the spectrum is dynamic
(CV = 22.75%).
The pulse type call of H. celebensis is similar
with the pulse type call of H. nicobariensis as
described by Jehle and Arak (1998) and Malkmus et
al. (2002); the difference between the two calls type
is the number of pulses per call; call of H. celebensis
consists of 3-5 pulses, while in H. nicobariensis
consists of 6-10 pulses; the duration of the pulse
on H. celebensis is between 100-164 ms,
whereas on H. nicobariensis is between 50-60
ms; the average pulse rate on H. celebensis is
Figure 4a. Oscillogram and audiospectrogram of pulsed calls of Hylarana celebensis.
Figure 4b. Enlarged oscillogram to show the sub-pulses of pulsed call of Hylarana celebensis.
Figure 5. Wide spectrum (black color) of pulsed calls of Hylarana celebensis.
6
Hellen Kurniati
4.76/sec, whereas 7.00/sec on H. nicobariensis;
duration between pulses on H. celebensis is 182
-250 ms, whereas in H. nicobariensis is 70-80
ms. The lowest pulse frequency and the highest
pulse frequency on H. celebensis is between
1124-4500 Hz (see Table 1), whereas on H.
nicobariensis is between 1000-4000 Hz
(Malkmus et al. 2002).
A
B
Figure 6. Oscillogram, audiospectrogram and intensity (power) of Hylarana taipehensis (A) from Thailand
that looks similar to the three types of pure tone calls of Hylarana celebensis (B) from Sulawesi (image
source of A: Sun & Narins 2005). (g) a pure tone that similar to pure tone type 1; (e) a pure tone similar
to the pure tone type 2; (f) a pure tone similar to pure tone type 3.
Table 1. The average value, standard deviation (SD), minimum and maximum values [(x ± SD) (minimum-
maximum)] of the frequency and duration of four types of calls of Hylarana celebensis. n = total num-
ber of the analyzed calls.
Pure tone type 1
Pure tone type 2
Pure tone type 3
Pulse type
(n=78)
(n=18)
(n=6)
(n=87)
1563,66±248,61
1366,32±248,25
1120,67±196,98
(937-2015)
(483-1734)
(937-1453)
2777,95±521,11
2139,74±384,67
1839,50±216,56
(1593-4078)
(1593-2695)
(1453-2015)
2904,58±401,92
2640,17±213,18
(1828-3304)
(2390-2953)
40,71±11,06
80,63±30,35
104,17±34,08
(23-91)
(37-151)
(62-157)
4±0,31
(3-5)
1636,08±151,72
(1125-1968)
3326,15±322,01
(2906-4500)
1690,21±384,62
(1032-2860)
131,87±10,57
(100-164)
210,17±16,77
(182-250)
788,23±70,09
(548-963)
9699,05±2034,10
(7358-16390)
Number of pulse/call
-
-
-
Call characters
Call type
Fundamental frequency-Hz
-
Harmonic 1-Hz
-
Harmonic 2-Hz
-
-
Duration of one pure tone-ms
-
Lower frequency-Hertz
-
-
-
Upper Frequency-Hz
-
-
-
Width of spectrum-Hz
-
-
-
Duration of one pulse-ms
-
-
-
Duration between calls-ms
-
-
-
Duration between pulses-ms
-
-
-
Duration of one call-ms
-
-
-
7
Bioacoustics of Hylarana celebensis (Peters, 1872) (Anura: Ranidae)
CONCLUSIONS
The bioacoustics of H. celebensis has two
types of calls, i.e. pure tone and pulse. Pure tone
calls consists of three types: (a) pure tone type
1; (b) pure tone type 2; and (c) pure tone type 3.
Pure tone type 1 is a modulated call which has 1
-2 harmonics and the fundamental frequency
range is between 937-2015 Hz. Pure tone type 2
is a modulated call which has 2-3 harmonics
and the fundamental frequency range is between
483-1734 Hz. Pure tone type 3 is a slightly
modulated call which has 4-7 harmonics and the
fundamental frequency range is between 937-
1453 Hz. The pulse type call has only one type,
which has no dominant frequency on its
audiospectrogram, but it is a ban with a wide
spectrum which range of the spectrum is
between 1032-2860 Hz.
ACKNOWLEDGEMENTS
The author is addressed to PT. Hatfield
Indonesia and PT Vale Indonesia, which have
sponsored the exploration of the herpetofauna
group activities in the Bahodopi area, Central
Sulawesi. Many thank also goes to Mr. Saiful,
Mr. Wardi and Mr. Iliyas who have helped the
author during the fieldwork. Finally, the authors
would like to thanks Dr. Arjan Boonman for his
comments and English editing to this article.
REFERENCES
Evans, BJ., JA. McGuire, RM. Brown, N.
Andayani & J. Supriatna. 2008. A coalescent
framework for comparing alternative models
of population structure with genetic data:
evolution of Celebes toads. Biology Letters
4:430–433.
Gerhardt, HC. 1991. Female mate choice in
treefrogs: static and dynamic criteria. Animal
Behaviour 42: 615-635.
Gillespie, G., S. Howard, D. Lockie, M. Scroggie &
Boeadi. 2005. Herpetofaunal richness and
community structure of off shore islands of
Sulawesi, Indonesia. Biotropica 37 (2): 279-
290.
Goutte, S., A. Dubois & F. Legendre. 2013.
The Importance of Ambient Sound Level
to Characterise Anuran Habitat. PLoS
ONE 8(10): e78020. doi: 10.1371/journal.
pone. 0078020.
Iskandar, D & Mumpuni. 2004. Hylarana
celebensis. In: IUCN 2013. IUCN Red List of
Threatened Species. Version 2013.2.
iucnredlist.org>. Down loaded on 06 March
2014.
Jehle, R & A. Arak. 1998. Graded call variation in
the Asian cricket frog Rana nicobariensis.
Bioacoustics 9(1): 35-48.
Krebs, CJ. 1989. Ecological methodology. Harper &
Row Publisher. New York.
Kurniati, H. 2013. Keragaman suara kodok Puru
Besar, Phrynoidis aspera (Gravenhorst, 1829)
asal Jawa Barat. Berita Biologi 2 (1): 47-60.
Malkmus, R., U. Manthey, G. Vogel, P. Hoffmann
& J. Kosuch. 2002. Amphibians & reptiles of
Mount Kinabalu (North Borneo). Fuldaer
Verlagsanstalt, Germany.
Marquez, R & XR. Eekhout. 2006. Advertisement
calls of six species of anurans from Bali,
Republic of Indonesia. Journal of Natural
History 40 (9–10): 571–588.
Pettitt, BA., GR. Bourne & MA. Bee. 2012.
Quantitative acoustic analysis of the
vocal repertoire of the golden rocket frog
(Anomaloglossus beebei). Journal Acoustical
Society of America 131 (6): 4811–4820.
Sheridan, JA., D. Bickford & KF-Y. Su. 2010.
An examination of call and genetic
variation in three wide-ranging southeast
asian anuran species. The Raffles Bulletin
of Zoology 58(2): 369–379.
Sun, JWC & PM. Narins. 2005. Anthropogenic
sounds differentially affect amphibian
call rate. Biological Conservation 121:
419-427.
Wanger, TC., I. Motzke, S. Saleh & DT. Iskandar.
2011. The amphibians and reptiles of the Lore
Lindu National Park area, Central Sulawesi,
Indonesia. Salamandra 47 (1): 17-29.
Wei, L., L. Zhao, X. Ma, X. Fan, X. Ma & Z. Lin.
2012. Advertisement Call Variability in the
Black-spined Toad Bufo melanostictus
(Anura: Bufonidae) during the Breeding
Season in Lishui, Zhejiang, China. Asian
Herpetological Research 3(2): 157–162.
Document Outline - Blank Page
- Blank Page
- Blank Page
- Blank Page
- Blank Page
- Blank Page
- Blank Page
- Blank Page
- Blank Page
- Acover_JBI2016(1)_E_ISSN.pdf
- ABcover_JBI2016(1)_E_ISSN.pdf
Dostları ilə paylaş: |