UNIVERSITY OF

ILLINOIS LIBRARY

AT URBANA-CHAMPAIGN

BIOLOGY

JUL 1 1 y^^

FIELDIANA
Zoology

Published by Field Museum of Natural History

Volume 58, No. 6 August 31, 1971

Mating Calls of Some Frogs From Thailand

W. Ronald Heyeri

Biology Department, Pacific Lutheran University

During a year's ecological study of amphibians and reptiles at
the Sakaerat Experimental Station, located 250 km. N.E. of Bang-
kok, I had the opportunity to observe and record vocal activity of
22 species (11 genera and 4 families) of frogs. The mating calls of
some of these species have been described in word form previously,
but to my knowledge none has been analyzed from recordings.

Dr. Joe T. Marshall has also recorded mating calls of some Thai
frogs. Those species recorded by him with which I have had no
field experience are not included herein. His material is on deposit
at the American Museum of Natural History.

t

METHODS AND MATERIALS

All calls were recorded on Scotch Brand No. 150-9 magnetic
recording tape at 73^ i.p.s. using a Uher 4000 Report-L portable
tape recorder. Analysis was by a Kay Electronics Sonagraph
Model 6061B, using the 80-8000Hz frequency scale with narrow
band filter (45 Hz) unless otherwise stipulated. The amplitude
display unit was used to resolve temporal intensity patterns, par-
ticularly pulse repetition frequency and pulses per note. Playback
was by a Crown Series 8000 tape recorder. The system, record to
playback, is essentially flat ±3db over the frequency range of the
analyzer.

Calling individuals were captured and preserved for positive
identification in most cases; at least one calling specimen for each
species was captured with the exception of Kaloula pulchra. The

' Formerly Field Research Associate, Division of Amphibians and Reptiles,
Field Museum of Natural History.

Library of Congress Catalog Card Number: 72-165190

Publication 1132 61 jj^g LIBRARY 0?, THE

62 FIELDIANA: ZOOLOGY, VOLUME 58

tapes and specimens are deposited at Field Museum of Natural
History.

The terminology that has accumulated in association with call
analyses is difficult to apply consistently to different types of frog
calls. The following definitions are those used by current workers
(Straughan, personal communication).

Since the darkness of the sonagraph trace is proportional to the
source intensity, dominant frequency was taken as the darkest
portion of the frequency traces on the sonagram. Where no discrete
dominant was present, I took the entire frequency range of the
darkest portion as the dominant. A fundamental frequency was
recognized where a distinct frequency band lay between the dominant
frequency and the baseline on the sonagram. Harmonic content
is described when there is a regular pattern of frequency bands on
the sonagram. A change of frequency within a note (fig. 14C) is
called frequency modulation. Intensity modulation refers to a
noticeable change in intensity within a note (fig. 5).

A call group consists of a series of repeated calls. A call consists
of a series of repeated notes (fig. 9), a single note (fig. 3), or a series
of pulses (fig. 7). A note may consist of a single pulse (no pulse),
may be partially pulsed (fig. 3), or may consist of a series of pulses
(fig. 2A).

Temperature is known to affect pulse rate. Unfortunately,
temperatures were not recorded in every instance, but the records
for which temperatures are available are presented in the species
accounts.

ACKNOWLEDGMENTS

Noel Kobayashi, Prasert Lohavanijaya, Frank Nicholls, Sukhum
Pongsapipatana, all of the Applied Scientific Research Corporation
of Thailand, my wife Miriam and daughter Laura facilitated the
field work in Thailand, especially at Sakaerat. Jay M. Savage and
Ian Straughan, University of Southern California, allowed me the
use of the tape recorder and sonagraph. Dr. Straughan gave
needed assistance with the sonagraph. Richard G. Zweifel, Ameri-
can Museum of Natural History, loaned Marshall's tapes and
specimens for comparison. Robert F. Inger, Field Museum of
Natural History, and Straughan carefully read the manuscript.
Field work in Thailand was supported by National Science Founda-
tion grant GB 7845X.

HEYER: MATING CALLS OF THAILAND FROGS

63

8-1

N

■

■ T.

^

»

6-

>
O

5-

z

4-

3

O

,^-

UJ

u.

2-

1-

LidhLJi/LiJLJu^JiJl/iJy>.A/iJuJJ

TIME

(SECONDS)

Fig. 1.
display.

The mating call of Bufo melanostictus, WRH 69-4, with amplitude

SPECIES ACCOUNTS
Bufonidae. Figure 1.

Eight calls from three specimens of Bufo melanostictus were
analyzed (fig. 1). The calls are long trills of from 4-30 seconds.
The pulse rate for a call recorded at an air temperature of 25 °C is
13.2 pulses per second, counting all pulses. Maximum sound
energy is spread over the frequency range 1000-1700 hz. The call
does not have harmonics.

8-|

N

■

X

^

X

b-

>

o

5-

2
UJ

4-

13

O

>^-

UJ

u.

2-
1-

A

okM

I n i ( »

lillliM

m%_

ill

[til

.^f*L:^lli>Ub--

TIME

(SECONDS)

Fig. 2. A, The mating call of Calluella guttulata, WRH 69-3, with amplitude
display. Bufo melanostictus in background. B, The mating call of Glyphoglossus
molossus, WRH 69-1, with amplitude display.

64

FIELDIANA: ZOOLOGY, VOLUME 58
Microhylidae. Table 1, Figures 2-9.

The call of Calluella guttulata (fig. 2A) consists of a series of
pulsed notes lasting from .14-.37 seconds. Each note is composed
of from 4-10 pulses; the pulse rate ranges from 27-33 per second.

_ 8

N

I 7-

UJ

mm/itiiimimmmtmmtiiiMm

hMP'HK

TME

(SECONDS)

*-4

2

Fig.
display.

3. The mating call of Kaloula pulchra, WRH 69-31, with amplitude

Maximum sound energy of the poorly tuned note is spread over the
frequency range 700-3200 hz. There is no distinct fundamental
frequency; poorly differentiated harmonics may or may not be
present. Frequency and intensity modulations are absent.

The call of Glyphoglossus molossus (fig. 2B) consists of a single
partially pulsed note lasting from .30-. 32 seconds. Each note is
composed of from 12-19 pulses; the pulse rate ranges from 40-59
pulses per second. Maximum sound energy of the moderately
well-tuned note is spread over the frequency range 300-1100 hz.
The dominant frequency is the same as the fundamental. Two
moderately differentiated harmonics are present at approximately
2000 and 3500 hz. The lower harmonic has the more energy. In-
tensity modulation is present.

The call of Kaloula pulchra (fig. 3) consists of a single partially
pulsed note lasting .56-.60 seconds. Each note is composed of from
18-21 pulses; the pulse rate ranges from 32-35 pulses per second.
The dominant frequency of the well tuned note is 250 hz. Well-
defined harmonics are absent; the dominant frequency is equal to
the fundamental. Frequency and intensity modulations are absent.

The call of Microhyla herdmorei (fig. 4) consists of a series of
pulsed notes lasting .09-.26 seconds. Each note is composed of 3-9

TIME

(SECONDS)

Fig. 4. The mating call of Microhyla berdmorei, WRH 69-39, with amplitude
display.

8-1

n

"

I

7-

>t

6-

o

5-

z

4-

D

■

o

6-

Ui

U.

2-

1-

TIME

(SECONDS)

Fig. 5. The mating call of Microhyla butleri, WRH 69-11, with amplitude
display.

^ 8

I 7-:

6-
> 5-
g 4-
O 3-

Ui

liiiiiiifii

U* ' I I i I i * I r -

t'ftf r?f ? f I

TIME

(SECONDS)

Fig. 6. A, The mating call of Microhyla heymonsi, WRH 69-9. B, The
mating call of Microhyla pulchra, WRH 69-17.

65

66

FIELDIANA: ZOOLOGY, VOLUME 58

_ 8-1

N

^ 6-

5 4-
§3-
^2-

■-t-i-^-i- *ij, ^i.^t'^jux^^g

I

TIC (SECONDS)

Fig. 7. The mating call of Microhyla inornata, WRH 69-19.

pulses; the pulse rate ranges from 33-35 per second. The dominant
frequency of the moderately well-tuned note is spread over the
1500-1800 hz frequency range. Well-defined harmonics are absent;
the fundamental frequency equals the dominant. Frequency and
intensity modulations are absent.

The call of Microhyla hutleri (fig. 5) consists of a series of partially
pulsed notes, each note lasting .16 -.21 seconds. Each note is
composed of 6-7 pulses per note; the pulse rate ranges from 28-44
pulses per second. Maximum sound energy of the poorly tuned
note is spread over the frequency range 1200-4500 hz. Harmonics
are absent; the fundamental frequency equals the dominant, A

TME

(SECONDS)

Fig. 8. The mating call of Microhyla inornata, WRH 69-19, played at half
speed, analyzed with wide band filter, with amplitude display. The frequency
scale is doubled and the time scale is half of the scales figured.

HEYER: MATING CALLS OF THAILAND FROGS

67

iUmmmmmmi

mMami

TIME (SECONDS)

Fig. 9.
display.

The mating call of Microhyla ornata, WRH 69-2, with amplitude

strong intensity modulation is present, giving a distinctive, rising
sound to the note when heard by the human ear.

The call of Microhyla heymonsi (fig. 6A) consists of a series of
pulses, a call lasting .48 seconds. The single call analyzed had 11
pulses at a pulse rate of 23 pulses per second recorded at an air
temperature of 28*0. Maximum sound energy of the poorly tuned
call is spread over the frequency range 1700-3000 hz. Harmonics
are absent; the fundamental frequency equals the dominant. The
call lacks frequency and intensity modulation.

The call of Microhyla inornata (figs. 7, 8) consists of a series of
pulses. The call lasts from .79-2.02 seconds. Each call is composed
of 52-83 pulses; the pulse rate is approximately 66 pulses per
second. Maximum sound energy of the poorly tuned call is spread
over the frequency range 4400-6500 hz. Harmonics are absent; the
fundamental frequency equals the dominant. The call lacks fre-
quency and intensity modulation. The call sounds like a cricket
rather than a frog.

The call of Microhyla ornata (fig. 9) consists of a series of notes,
each note lasting .23-.31 seconds. Each note is composed of 10-18
pulses; the pulse rate ranges from 53-60 pulses per second recorded
at air temperatures of 25-28°C. Maximum sound energy of the
poorly tuned note is spread over the frequency range 1200-3500 hz.
The fundamental frequency ranges from 80 500 hz. A poorly
differentiated harmonic lies in the 4500-5000 hz frequency range.
An intensity modulation is distinct in most of the tracings, giving a
rising sound to the note when heard by the human ear.

c

c -^

o

0) t. cS

e3

05 O 01

W

o>

>» ?

5 Q>

c

Si

J<!

C

c«

<D

O

05

^

c

c

^

g

c

(U

Ol

2? ^^

fii

tn

CO

0) o

on

^

^

^

S>

CIS

c8

a

c3

0) ^

n

^

«*-

s

o

3

rn

C

11

«

■5

bfi

0)

C

03

■!->

o

rt

o

05 C
0) o

J2

03

Si

^

J

Si

rt

C3

rt

rt

ca

cU

3

b

o

O

o

o

o

o

O

o

o

o

(M

I-H

c»

lO

o

C4

J3

1

o

1
O

o

1

o

1

©

1

o

o

o

o

o

©

t-

ec

in

IM

t-

3 0)

©

©

©

o

©

©

iC

\ei

lO

CD

1

1

1

CO
I

i

a

So;

©

"o

3

PUl

o

§

g

•i

:3

c

05

05

^

a

o

o

0) c

CO

i

-o

*5

§1

r

<J

a;
Xi

3

Q

"o

I— 1

i-H

3

w

:z;

^

II

z

.J, n

00 rH

U3 rH ,-H

^ Oi

CO '

SS ci £i

i-( rH -<1<

U
O)

a

5

(J

O O

1^

5

s

68

HEYER: MATING CALLS OF THAILAND FROGS

69

I

The call of Microhyla pulchra (fig. 6B) consists of a series of
pulses. The call lasts 1.42-1.92 seconds. Each call is composed
of 10-20 pulses; the pulse rate ranges from 7-10 per second recorded
at an air temperature of 25 °C. Maximum sound energy of the
moderately poorly tuned call is spread over the frequency range
1000-2500 hz. Harmonics are absent; the fundamental frequency
equals the dominant. The call lacks frequency and intensity
modulation.

Each species call is quite distinct when heard in the field; these
distinct differences are tabulated (Table 1).

Ranidae. Table 2, Figures 10-15

At the research site, the vast majority of Ooeidozyga laevis
uttered two types of calls (A and C of Table 2 and fig. 10). Call
type A consists of a single note lasting .03- .08 seconds. Each note
consists of a single pulse. Maximum sound energy of the well-
tuned call centers about two frequency ranges, 2000 hz and/or
3600 hz. The dominant frequency may be associated with the
second and/or fourth harmonic frequencies. Three to seven har-
monic frequencies are present at approximately 2000, 2800, 3600,
4200, 5000, and 5800 hz. The call lacks frequency and intensity
modulation.

Call type C consists of a series of notes, each note lasting .03-. 05
seconds. Each note is composed of a single pulse. The pulse rate
ranges from 8-9 per second recorded at an air temperature of 30°C.

I 7-

A

5-

4-

2-

^

**

i V i ; J i t V t t i i

-e <* -.

^^r---. «c --T*^ ■c* -"^ -f «**

TME

(SECONDS)

Fig. 10. Mating calls of Ooeidozyga laevis.
Call type C, WRH 69-13.

A, Call type A, WRH 69-5. C,

70

FIELDIANA: ZOOLOGY, VOLUME 58

^ o-

w

I 7-

X

^ 6-

^5-

A

B

5 4-

-

3

4Bte

O 3-

'^"

^2-

—

1-

■

TME

(SECONDS)

Fig. 11. Mating calls of Ooeidozyga laevis calling from the same pond.
Call type A, WRH 69-16. B, Call type B, WRH 69-15.

A,

Maximum sound energy of the moderately poorly tuned call is spread
over the frequency range 1700-1900 hz. The dominant frequency
is equal to the fundamental. A single rather indistinct harmonic is
found in the 3500-4000 hz frequency range. A slight increasing or
decreasing frequency modulation is found in the harmonic fre-
quencies at the end of the call.

The call group consists of irregularly spaced call type A's ending
with call type C.

Dr. Marshall recorded individuals of Ooeidozyga which had a dif-
ferent call. On April 3, 1969 I recorded both call types at the same
locality (fig. 11). Our first reaction was that two sympatric species

I 7-

X

^ 6-

g4H
S3

I

^HiiM^V#MAMlMMMMMi«l

I
TME (SECONDS)

Fig. 12. Mating call of Ooeidozyga laevis illustrating call types A and B from
a single individual, WRH 69-36, with amplitude display.

HEYER: MATING CALLS OF THAILAND FROGS 71

were involved. The unusual call, note type B (Table 2, fig. 11), was
the only call heard from individuals of the "second" species. That
is, the call of these individuals lacked the typical type C end of an
Ooeidozyga laevis call group. Comparison of the sonagrams of call
types A and B recorded in sympatry (fig. 11) indicate they differ

M

X 7

>-
o

I

'^ ' ^w«^

5- ^^|»|WW^#|»i»iiiii|»»l»|<<»t<iiiMii|

.?

1 ' I

I 2

TME (SECONDS)

Fig. 13. The mating call of Ooeidozyga lima, WRH 69-34.

basically only in note duration. Calling males of both call types
were captured and proved morphologically indistinguishable. On
August 2, 1969, I recorded one male Ooeidozyga that produced all
three note types (types A and B only shown on fig. 12) . This male
was from the same pond at which recordings were made on April 3.
I interpret the type A and B notes to be individual variation within
a single species.

Ooeidozyga lima has two call types similar to 0. laevis; a call group
consisting of irregularly spaced single note calls ending with a call
consisting of many rapid notes. I recorded only the first part of the
call (Table 2, fig. 13) . The call consists of a single note lasting .05 .06
seconds. Each note consists of two pulses. Maximum sound energy
of the poorly tuned call is spread over the frequency range 2800-3200
hz. The call lacks harmonics; the fundamental frequency is equal
to the dominant. The call lacks frequency and intensity modulation.
The call of 0. lima sounds very distinct from any of the call types of
0. laevis.

The call of Rana limnocharis (fig. 14A) consists of a series of
partially pulsed notes, each note lasting .11-. 16 seconds. Each
note is composed of 5-7 pulses. Maximum sound energy of the
poorly tuned call is spread over the frequency range 2300-2800 hz.

72

FIELDIANA: ZOOLOGY, VOLUME 58

^ 6-

Z /I

, JkJK

1

0 — LL

UJ 4-
§3-

XL. ^

1-

mim~ '«^

t

d^

"I

1

2

TME (SECONDS)

Fig, 14. A, The mating call of Rana limnocharis, no number, with amplitude
display. B, The mating call of Rana livida, WRH 69-40 with amplitude display.
C, The mating call of Rana nigromttata, WRH 69-27, with amplitude display.

A fundamental frequency lies at approximately 1250 hz. No har-
monics above the dominant are present. The call lacks frequency
and intensity modulation.

The call of Rana livida (fig. 14B) consists of a partially pulsed
note. The only note analyzed is composed of two pulses, the note
lasts .04 seconds. The dominant frequency of the well-tuned call
modulates from 900-1800 hz. A single harmonic modulates from
1500-2200 hz; the fundamental frequency equals the dominant.

The call of Rana nigrovittata (fig. 14C) consists of a partially
pulsed note. The duration of the call ranges from .16 -.21 seconds.
Each note is composed of 3-4 pulses. The dominant frequency of
the well-tuned note modulates in a complex manner (see fig. 14C)
between 900 and 1500 hz. A fundamental frequency modulates
between 500 and 1000 hz. Two to three harmonics above the domi-
nant modulate between approximately 1200-1800, 1400-2000, and
1700-2000 hz respectively.

Rana pileata is the only species of Rana recorded that exhibited
two call types. Call type A (fig. 15, first two tracings) sounds like
a cross between WONK! and a stone dropping in water. The call
consists of a pulsed note lasting .12-. 14 seconds. Each note is
composed of 2-4 pulses. Maximum sound energy of the poorly
tuned call is spread over the frequency range 200-1000 hz. There
are no harmonic frequencies; the fundamental frequency equals the
dominant. A frequency modulation occurs between 500 and 1000
hz in one component of the tracing.

'E jq

m

o

< m

-rt

(U

0)

cS

■n

T1

>-i

d

o

rt

a>

D^

^ o

c

0)

<1> ^ CQ
03 O (D

C3 ^

o

-D

_m

s

OS

c

S

1) Ui S

c c g c c

m CO o 5K to M

^ ^ gi JO ^

Co CO p^ 03 CS

>>

o

o

O C

o

o

c —

(M

(M

ajTS N

(M

C^

tr

1

O

^

o

o

S-Q

lO

00

y-* CO

<M i-( i-H

u

e«

03

a

w

4?

O

^3

C

>.

o

e«

m

S3

(h

ci3

crt

a

S

cS

^= o
3 c« C
Oh "

Co y Z O

O

3 «<-i

00

o

r

CO

o

i-H O

tH IM rH

(M ■<d< rH

•s

-§

73

74

FIELDIANA: ZOOLOGY, VOLUME 58

TME

(SECONDS)

Fig. 15. The mating call of Rana pileata WRH 69-33, with amplitude display.

Call type B (fig. 15, last three tracings) sounds like snoring. The
call consists of a series of pulsed and partially pulsed notes, each note
lasting from .11-. 16 seconds. Each note is composed of 5-8 pulses.
The dominant frequency of the moderately well-tuned call lies
between 250-500 hz. No harmonic frequencies are present. The
notes of the call may have slight frequency modulations.

A call group of Rana pileata consists of call A given at spaced
intervals; call B is given at intervals only when the frogs are most
actively calling.

Rhacophoridae. Table 3, Figures 16-20

The call of Chirixalus nongkhorensis (fig. 16A) is composed of
from 1-2 nonpulsed notes; i.e., each note consists of a single pulse.
Each note lasts .02 .03 seconds. Maximum sound energy of the
poorly tuned call is spread over the frequency range 3300-4000 hz.
Harmonic frequencies are absent; the fundamental frequency equals
the dominant. The call lacks both frequency and intensity modula-
tion.

The call of Chirixalus vittatus (fig. 16B) consists of a series of
from 1-4 nonpulsed notes, i.e., each note consists of a single pulse.
Only one note was recorded and analyzed. The note lasted .05
seconds. Maximum sound energy of the moderately poorly tuned
note is spread over the frequency range 1750-2000 hz. Harmonic
frequencies are lacking; the fundamental frequency equals the
dominant. The call lacks both frequency and intensity modulation.
The first few times Chirixalus vittatus called, there was but one note

HEYER: MATING CALLS OF THAILAND FROGS

75

- 81

X7-

^

V

i!

^ 6-

•J'

^5-

A

B

54-

3 ,

0 3-

UJ

1-

1 — -— ^ -i

V

TME

(SECONDS)

Fig. 16. A, The mating call of Chirixalus nongkhorensis, WRH 69-32.
The mating call of Chirixalus vittatus, WRH 69-37, with ampHtude display.

B,

per call. Later in the season, the species commonly had up to four
notes per call.

The call of Philautus parvulus (fig. 17) consists of a series of 1-7
nonpulsed notes, i.e., each note consists of a single pulse. Each
note lasts from .02-.04 seconds. Maximum sound energy of the
poorly tuned call is spread over the frequency range 2250-3250 hz.
Harmonic frequencies are lacking; the fundamental frequency equals
the dominant. The call lacks both frequency and intensity modula-
tion. Philautus parvulus starts calling with one note per call, then
increases by one note each succeeding call until the maximum
number of notes per call is reached. Chirixalus vittatus and Microhyla
pulchra also called in this manner, but not with the observed absolute

8-
7-

J

6-

5-
4-

3-
2-
1-

r

r

f

r =

? •

I

1

2

TME

(SECONDS)

Fig, 17. The mating call of Philautus parvulus, WRH 69-30.

76

FIELDIANA: ZOOLOGY, VOLUME 58

regularity of adding one note per call as observed in Philautus
parvulus.

The call group of Polypedates leucomystax consists of two call
types (fig. 18) . Call type A (fig. 18A) consists of a pulsed note lasting
.23-. 38 seconds. Each note is composed of 4-5 pulses. Maximum

_ 8

>- 5-

5 41

-_, ; --. --- ♦-- ,-».»■ '

mm

i j! !
Mil

I ft

TME (SECONDS)

Fig. 18. Mating calls of Polypedates leucomystax. A, Call type A, WRH
69-14. B, Call type B, WRH 69-23.

sound energy of the poorly tuned call is spread over the frequency
range 300-2600 hz. Harmonic frequencies are lacking; the funda-
mental frequency equals the dominant. The call lacks frequency
modulation, but may have a slight increase in intensity at the end.
Call type B (fig. 18B) consists of a pulsed note lasting from .12-. 25
seconds. Each note is composed of 2-4 pulses. Maximum sound
energy of the poorly tuned call is spread over the frequency range
1700-3100 hz. Harmonic frequencies are lacking; the fundamental
frequency equals the dominant. The call lacks both frequency and
intensity modulation. There did not appear to be any pattern of how
the call types were organized into a call group.

The call group of Rhacophorus appendiculatus is composed of two
call types (fig. 19). Call type A (fig. 19A) consists of a single note.
The only note analyzed is composed of a single pulse lasting .04
seconds. Maximum sound energy of the moderately well-tuned note
lies between 2500-2600 hz. Harmonic frequencies are lacking. The
fundamental frequency appears to equal the dominant. The call
lacks both frequency and intensity modulation. Call type B (fig.
19B) consists of a pulsed note. The only note analyzed is composed
of seven pulses, lasting .48 seconds. Maximum sound energy of the

HEYER: MATING CALLS OF THAILAND FROGS

77

8-1

M

'

I

r-

U

^

6-

>

O

5-

z

4-

■D

'

a

6-

1^

2-

1-

B

*- - -=*v*if<c-

I 2

TME (SECONDS)

Fig. 19. Mating calls of Rhacophorus appendiculatus. A, Call type A, WRH
69-26. B, Call type B, WRH 69-26.

poorly tuned call is spread over the frequency range 1600-2500 hz.
The recording and subsequent sonagram are of too poor a quality to
determine if a distinct fundamental frequency is present or frequency
or intensity modulations are present. A call group started with type
A given at various intervals, ended with a series of call type B calls.
The call group of Rhacophorus bimaculatus is composed of two
call types (fig. 20). Call type A (fig. 20A) consists of a single pulsed
note lasting .04 .09 seconds. Maximum sound energy of the poorly
tuned call is spread over the frequency range 1900-2800 hz. Har-
monic frequencies are lacking; the fundamental frequency equals the
dominant. The call lacks both frequency and intensity modulation.
Call type B (fig. 20B) consists of one or two pulsed notes, each lasting

_ 8-1

~ 6

5 4H
§3^

i^

B

TME (SECONDS)

Fig. 20. Mating calls of Rhacophorus bimaculatus.
69-22. B, Call type B, WRH 69-35.

A, Call type A, WRH

>

a

o
25

< m < m

M

0)

0)

"S

'§

TJ

T3

J3

a
o

o

8

^

«

u

cU
43

CO

P?

S

*S

g

o

'i

03

C

.s

"o

0)

1

Q

3

CJ3

'w

42

S

"o

3

_w

C

0!

£8

J3

o

-M

M

<D

3 C

C
'•♦3

CO

OS

"o

rt

cm

ri

c

a

C

C

a

« -^

hi

0)

a;

0)

a>

a>

oj t, rt

CO

CO

CO

n

^

X!

-Q

^

J3

-£ >

cU

cd

cs

c«

c4

eS

ea "^

o

o

o

o

o

o

o

o

o

o

«o

I— (

to

IC

00

(M

05

(M

(M

N

O

A

o

o

o

o

o

o

o

o

CO

t-

\a

«o

Oi

y-t Oi 1-i

CO

-4J

1^

b

0)

0)

C

-u

03

^

a

t-i

3^

^

M

1

0"=

u

N

t;

>.

o

"rt

c
o

C
CIS

to

w

^^

cS

cS

a

«

i

•o

O

fc

^

00

s

3

W

iz;

in

+^ S
O O

a

SS ci- d-

iM rH eo

o

eo i-H

00 i-H

■I I

1

I

O '-H

j-\ <M

78

HEYER: MATING CALLS OF THAILAND FROGS 79

.10-.34 seconds. Each note is composed of 3 8 pulses. Maximum
sound intensity of the poorly tuned call is spread over the frequency
range 1800 3300 hz. The recordings and subsequent sonagrams are
of poor quality, but it appears that a weak fundamental frequency
is spread over 80-500 hz, and that no pronounced frequency or
intensity modulations are present. Call group composition is identi-
cal to that of Rhacophorus appendiculatus.

The calls of Chirixalus nongkhorensis and vittatus sound similar
to the human ear as do those of Rhacophorus appendiculatus and
bimaculatus. Chirixalus nongkhorensis and vittatus occur in sympatry
at the research site; with practice I was able to discriminate the calls
of the two species at a given pond. As shown on Table 3, there is a
difference in frequencies. Rhacophorus appendiculatus and bimacula-
tus were never taken in sympatry. The recordings of each of these
two species are of poor quality and do not allow for detailed compari-
sons between the species.

DISCUSSION

The usefulness of mating calls in determining relationships is
well established, particularly when two morphologically similar
species exhibit divergent call types. The usefulness of mating calls
in determining relationships above the species level is not as well
documented. The data presented herein demonstrate that mating
calls can aid in determining higher relationships, but the method has
limitations.

One way of approaching the data is to look at the calls of the
members of genera and families for common uniting characteristics.
An extension of this approach is to look for a call prototype from
which all calls of the group may be derived. The microhylid and
rhacophorid calls analyzed appear as though the calls of each of the
families could have been derived from a single call prototype. The
prototype for the microhylid call would be composed of several
pulses per note over a rather broad frequency range. If this proto-
type is assumed, the call has been channeled in many directions
through selection. In some cases, the frequency response has be-
come narrowed as in Kaloula pulchra, or broadened as in Microhyla
butleri; the pulses have been jammed together as in Microhyla
inornata or spread apart as in Microhyla pulchra. As shown in Table
1, the frequencies between 250 and 6500 hz have been partitioned
among the species. Certain species, as Microhyla butleri and M.
ornata, have been able to modulate the intensity of the call.

80 FIELDIANA: ZOOLOGY, VOLUME 58

The rhacophorid prototype would consist of a component with
a wide frequency range, lacking harmonics. The calls of Chirixalus
and Philautus recorded are similar in having a single type of call
which may consist of from one to several notes. Each of the species
of Polypedates and Rhacophorus recorded has two call types. In
Polypedates both call types are composed of several pulses per note,
whereas the species of Rhacophorus have one call type that has a
single pulse. In all forms with two call types, the difference be-
tween call types within a call group of a given species are in the
number of pulses and frequency range.

Another way of organizing the data is to sort the calls by what
appear to be different methods of sound production; in other words,
sorting for fundamental types of calls. The calls reported appear to
fit into three fundamental types: 1) Bufonidae, 2) Rajia (Ranidae),
and 3) Microhylidae, Rhacophoridae, and Ooeidozyga (Ranidae).
If these groupings do indicate different mechanisms of sound produc-
tion, two conclusions may be made. In some instances, calls may
be useful in differentiating groupings, such as Ooeidozyga from Rana,
and bufonids from the other families. In other instances, mating
call production may give no indication of relationships, such as
indicated by members of three families sharing common sound re-
production mechanisms (type 3 call). This, in turn, may indicate
that methods of sound production are limited in frogs.

Mating calls may correlate with phenomena other than those
associated with relationships. Konishi (1970) indicates two con-
tradicting evolutionary pressures involving the physical properties of
sound: 1) lower frequencies carry farther than higher frequencies, yet
2) "higher frequencies and wider frequency ranges . . . are more
suitable for localization by binaural comparisons of intensity or time
differences . . ." (p. 68). In more complex habitats, it would appear
logical to assume that the exact location of the calling male by the
female would be the most important factor; that is, frog calls in
complex habitats would be predicted to have higher frequencies than
those in simple habitats. To test this hypothesis, the data were
organized in the following way. For the period of February 23, 1969
through December 31, 1969 records were kept on calling activity at
ponds in four different ecological settings: 1) a pond next to a high-
way in an open situation; 2) three small ponds resulting from pits
dug for road fill in a deciduous dipterocarp forest; 3) a series of ponds
resulting from a small spring in partially cleared dry evergreen forest;

HEYER: MATING CALLS OF THAILAND FROGS

81

N

X

6n

5-

4-

>-
o

z

UJ

ID
O
UJ

tr.

3-

2-

4

B

3

4

Fig. 21. Average frequency of dominant of each species in four vegetation
tjrpes. Vegetation type 1 is the most open, type 4 the densest. A, All species.
B, Microhylids only.

and 4) a naturally occurring temporary pond in the dry evergreen
forest. From 1 to 4 represents a grade of increasing complexity and
density of the vegetation. Results of calling activity by species were
pooled for the year for each pond situation. An average dominant
frequency was determined for each species by taking the average of
the ranges reported. In the cases where a frog call contained two or
more note types an average value of all the averages of the different
note types was used. The frequency values were then plotted by
ecological situation (fig. 21). There does appear to be a trend of
higher dominant frequencies associated with denser vegetation (fig.
21A). It is obvious, however, that this trend is accounted for solely
by members of the family Microhylidae (fig. 21B).

The chorus structure of members of the Microhylidae give further
insight into the ecological meaning of physical parameters of sound
coding systems. Male Glyphoglossus and Kaloula call while floating
in the water of just-filled rain pools. Both are large species, locally
abundant while calling, and at Sakaerat were found in the more open
habitats. The two species have the lowest dominant frequencies of
the microhylids recorded. The males can attract females over long
distances; the females need only to locate the ponds, not individual
males. All other species of microhylids analyzed call from the banks
or from vegetation on the water surface and are smaller species.

m

82 FIELDIANA: ZOOLOGY, VOLUME 58

The call of Microhyla herdmorei is characterized by a narrow fre-
quency range of the dominant and a lower maximum dominant
frequency than any of the other species of Calluella or Microhyla
analyzed. Males were calling from isolated pools next to a flowing
river (a habitat of rapid sound attenuation) and were the only species
calling from such a habitat. Around each suitable pool, numerous
males were calling. Apparently, the critical factor is similar to the
situation as observed in Glyphoglossus and Kaloula; attracting
females to a suitable microhabitat rather than to an individual male.
The remaining species are characterized by calling males dispersed
around a suitable pond situation. Here the apparent critical factor
is the point location of the calling male by the female. The calls of
this group of species are higher and have wider frequency ranges,
satisfying the physical requirements for more accurate binaural
point location of the sound source (Konishi, 1970).

Using the microhylids as an example, mating calls give the follow-
ing information with respect to relationships. The calls of members
of the family show a unity at the family level (remember the small
sample size, however!). At the other extreme, the calls are most
useful in distinguishing the species. The call data aid in distinguish-
ing certain categories, but yield no information on the phylogenies
of the categories involved. The correlation of call characteristics
with habitat differences and types of mating call chorus indicate that
other factors besides those associated with relationships are involved
in the evolution of mating calls.

LITERATURE CITED

Konishi, M.

1970. Evolution of design features in the coding of species-specificity. Amer.
Zool., 10 (1), pp. 67-72.

I

UNIVERSITY OF ILLINOIS-URBANA

590. 5FI C001

FIELDIANA, ZOOLOGYSCHGO
55-58.61 1968-72

3 0112 009379840

1