Oecologia (1988) 76:399-407
9 Springer-Verlag1988
Temporal spacing of breeding activity
in the natterjack toad, Bufo cMamita
Ulrich Sinseh
Zoologisches Institut der Universitfit Bonn, Poppelsdorfer SchloB, D-5300 Bonn 1, Federal Republic of Germany
Summary. The reproductive behaviour of the toad Bufo
calamita was studied in a large population in the northern
Rhineland, Germany. Toads bred at four spatially sepa-
rated but neighbouring areas including 34 breeding ponds.
Three methods were applied to monitor quantitatively the
intensity of breeding and migratory activity : capture-mark-
recapture, a mechanical tracking device, and radio-teleme-
try. In 1987 a twofold temporal spacing of breeding activity
was observed: (1) division of the reproductive period into
early, main, and late breeding periods; (2) subdivision of
each breeding period into several short calling periods.
Since weather conditions were favourable during most of
the reproductive period, environmental factors contributed
little (23.4%41.0%) to the temporal variation in breeding
intensity as shown by nmltivariate statistics, The period
of vocalization (April to August) included the activity of
three distinct mate groups. The first group called from April
to mid-May (2 calling periods), the second from May to
early July (5 calling periods) and the third from mid-July
to August (2 calling periods). Within a calling period most
males stayed near their calling sites and migratory activity
was low. Extensive movements were detected when a new
group of males immigrated from the feeding regions to the
breeding areas and males that had been calling previously
emigrated, at least partly. However, exchange of males
among the neighbouring breeding areas did not exceed 5%,
though neighbouring breeding choruses were clearly aud-
ible. It is suggested that the energetic costs of vocalization
impeded continuous calling activity in individuals and ac-
counted for the periodical variations in chorus size. The
short-time variations of breeding activity therefore probab-
ly reflect the physiological limits of toads in a situation
where the weather permits continuous reproductive activity.
Key words: Bufo calamita -Reproduction - Temporal spac-
ing - Vocalization - Migration
Two main categories of reproductive behaviour are found
in azuran amphibians: explosive breeding and prolonged
breeding (Wells 1977; Arak 1983a). In Central Europe the
first category is exemplified by the common toad Bufo bufo
(Eibl-Eibesfeldt 1950; Heusser 1967, 1968), and the second
by the natterjack toad B. calamita (Beebee 1979). The repro-
ductive strategy of B. bufo includes a complex pattern of
annual migrations which serves to gather the reproductive
individuals of a population at the breeding pond within
Oecologia
1 or 2 weeks (Heusser 1968; Sinsch 1987, I988a). B. cala-
rnita, however, breeds continuously over about 4 months
from April to August, in Germany (Bregulla 1986; Sinsch
1988b). Reproductive males are found in the breeding
ponds during the whole reproductive period and (orm chor-
uses which attract both sexes (Mathias 1971; Flindt and
Hemmer 1972; L6rcher and Schneider 1973; Arak 1983a).
Females approach the ponds exclusively for spawning and
leave immediately afterwards (Niekisch 1981). The acousti-
cal communication of B. caIamita during the reproductive
period has been analysed thoroughly in terms of structure
(Flindt and Hemmer 1968a; Heusser 1969; L6reher and
Schneider 1973; Weber 1978; Tejedo 1986) and mate choice
(Flindt and Hemmer 1972; Arak 1983a, b; H6gtmld and
Robertson 1986). Vocalization in breeding choruses, how-
ever, exhausts anurans considerably (Bucher el: al. 1981)
and Hyta crucifer has been shown to employ anaerobic
metabolism during calling (Pough and Gatten 1984). Be-
cause of the high energetic costs of vocalization it is rather
unlikely that the same males are engaged in chorusing
throughout the whole reproductive period. In fact, there
are indications that some males arrive at the breeding ponds
several weeks earlier than others (Flindt and Hemmer
1968 b; Bridson 1976). Nevertheless, the migratory behav-
iour of natterjack toads has never before been surveyed
over a complete reproductive period at both individual and
population levels.
Moreover, there is disagreement about the influence of
environmental factors on the breeding activity of this spe-
cies. Mathias (1971) failed to observe any relationship be-
tween activity and weather conditions, whereas Tejedo
(1986) atributed the temporal variation in numbers of
breeding toads almost entirely to the action of meteorologi-
cal variables. Since the first study concerned a population
in Britain but the second a population in Spain, it remains
unclear whether geographical aspects are also relevant.
The present study investigates the migratory behaviour
as wetl as the environmental influence on the breeding activ-
ity of the natterjack toad. A population size of more than
3000 adult toads (Sinsch 1988b) made it possible to quan-
tify both aspects. Three methods were used to monitor the
movements of the toads quantitatively: capture-mark-re-
capture calculations at the population level, and mechanical
tracking devices and radio telemetry at the individual level.
Parallel records of the microclimate within the habitat made
it possible to evaluate the relative influence of environmen-
tal factors by multivariate methods.
400
Material and methods
Study area
The study was conducted in the northern Rhineland near
Siegburg. West Germany, at altitudes of 52-58 m. The B.
calamita population studied ( > 300 adults) inhabited an
area of about 8 km 2 composed of sandy gravel pits, bare
ground, pasture, and maize and wheat fields. Four spatially
separated areas included a total of 34 breeding ponds
(Fig. 1). Area I included 6 ponds of 1303 m 2 maximum
total water surface, area II 4 ponds of 516 m 2, area III
18 ponds of 3310 m 2, and area IV 6 ponds of 1880 m 2.
These areas are referred to as breeding areas. Of the 34
ponds, 29 were shallow and dried up during the study peri-
od (further details in Sinsch 1988b).
General procedure
In this paper I am concerned with the "reproductive peri-
od" defined as the period from the first to the last observa-
tion in a year of calling males. The term "calling period"
is used to subdivide the reproductive period and means
the sequence of consecutive days from one local minimum
of chorus size to the next. A calling period includes also
the arrival of females and spawning activity at the local
maximum of chorus size. Since the number of spawn strings
per calling period differed considerably (see results, Fig. 2),
the term "breeding period" is used to include consecutive
calling periods from one local minimum of oviposition to
the next.
Preliminary observations on the breeding behaviour of
the study population were conducted in 1986, from 1 June
to 31 July. During the 1987 reproductive period (5 April-
2 August) the study area was visited at least four times
a week and daily during periods of spawning. The nocturnal
observations started 30min after sunset (April: 9 p.m.,
July: 11 p.m.). Males and females present at the breeding
sites were captured and counted and their snout-vent length
(_+ i mm) and body mass (_+ 0.5 g) were measured. All cap-
tures and measurements were performed by two people in
a regular way: in area III for i h, then area II for 0.5 h
and area I for 0.5 h. Thus, the counts of toads do not
represent the total number of reproductive toads per night
but the number of toads captured in 4 men-hours. Spawn
1500
N II I sion. After closing the wound with sutures the toads were
III
[m] ~iiiiiiiiiiiiii
iv
0 [m] 1500
Fig. 1. Spatial relations and extent (hatched) of the breeding areas
(I-IV) of the toad population studied
strings were counted during daytime (including area IV)
and the actual surface area of all breeding ponds was esti-
mated to the nearest square meter. Air temperature, relative
air humidity, atmospheric pressure, and precipitation were
recorded continuously at 10 cm above ground using a Lain-
brecht meteorgraph in area III.
Monitoring methods
Estimates of the numbers of breeding toads were obtained
by capture-mark-recapture calculations based on the algo-
rithms in Riist (1969). During April 1987 all toads staying
within areas I and III were captured, marked site-specifi-
cally by toe-clipping, and released again in situ. In area
I a total of 133 males was marked, and in area III 89 males.
During May 1987 another 24 toads were toe-clipped in area
II.
The individual paths of t4 males were monitored using
a mechanical tracking device. The device consisted of a
sewing machine bobbin a holder weighing ca. 6.5 g. The
holder was fastened to the toad's back with an elastic tape
around the waist. The bobbin contained about 60 m of
thread (for further details see Dole 1965). Each toad was
fitted with a tracking device at the capture site and released
in situ. On the morning of the following day the thread
unwound during nocturnal activity marked the path of the
toad and was plotted exactly on a map of the locality.
The thread supply was changed daily during the tracking
period, if neccessary. In 1986 eight calling males were
tracked at area I on 5-12 June, and another six males at
area II on 5-14 July.
The migratory behaviour of five males and one female
between two breeding periods was monitored by radio te-
lemetry. Single-stage transmitters (Custom Electronics, Ur-
bana Illinois, mass with battery: 2.5 g) were implanted into
the abdominal cavity of the toad. Toads were narcotized
in 0.05% solutions of MS 222 (ethyl-m-aminobenzoate
methanosulfonate). The transmitter was introduced
through a 1-cm lateral incision which was closed afterwards
by 3-4 sutures. The toads usually recovered within 2 h and
displayed normal behaviour. After another 2 h the toads
were released at the capture site and tracked with a CE 12
receiver (Custom Electronics). The range of the transmitters
varied from 40 m in burrowing toads to 200 m in toads
migrating on bare ground. The location of each toad was
recorded hourly from sunset to 1 p.m. and at 10 a.m. At
the end of the tracking period all toads were captured again
and the transmitters explanted through a contralateral inci-
released within their habitat. Wound infection was never
observed; on the contrary, the implantation wounds had
always healed by the end of the experiment. These toads
were tracked in area II from 10 April to 5 May 1987.
Statistical analysis
The Kruskal-Wallis H-test was used to test for differences
among the snout-vent length data. All multivariate methods
were applied to data sets of 18 variables with 102 observa-
tions each and refer exclusively to measurements at area
III. Variables are: MALES/FEMALES = numbers of indi-
viduals per 2 men-hours; S P A W N = n u m b e r s of newly de-
posited strings; W A T E R l = a c t u a l surface area of the
breeding ponds; WATER2 = surface area of ponds without
tadpoles; TEMPDMAX, TEMPSS, TEMPNMIN, TEMP-

N A V = air temperature: daily maximum, at sunset, night
minimum, night mean; T E M P R A T E = m e a n hourly de-
crease of temperature during night; H U M I D M I N , H U M -
ISS, H U M I N M A X , H U M I N A V = r e l a t i v e air humidity:
daily minimum, at sunset, night maximum, night mean;
PRESSR, PRESS=absolute atmospheric pressure at sun-
rise and sunset; P R E S - 2 4 , P R E S - 1 2 = c h a n g e of atmo-
spheric pressure 24 or 12 h before sunset; SOIL = moisture
of soil, as hours since last rainfall; R A I N D , R A I N N =
hours of rainfall during day and night. First a principal
component analysis was performed on the matrix of the
linear correlation coefficients of all variables. Principal
components with eigenvalues > 1 were extracted and sub-
mitted to an orthogonal VARIMAX-rotation. This kind
of rotation attempts to simplify the columns of the factor
matrix by making all values close to either 0 or 1. Those
variables which loaded the same principal component as
MALES or FEMALES were considered associated. A sec-
ond stepwise multiple regression (procedure: forward selec-
tion) was used to establish statistical models to predict the
numbers of either males or females (dependent variable)
from the corresponding environmental factors (independent
variables). Selection of variables stopped if the increase of
explained variance R 2 was not significant (F-statistics, P >
0.05). Finally discriminant functions were computed to pre-
dict presence or absence of males and females and spawn
deposition. All calculations were performed by F O R T R A N
77 programs based on algorithms summarized in Flury and
Riedwyl (1983). Circular correlation was tested using the
methods described in Batschelet (1981).
Results
Time course of breeding activity
The toads of the study population did not breed contin-
uously throughout the reproductive period, but showed a
regular pattern with respect to individual numbers and de-
position of spawn strings in the breeding ponds (Fig. 2).
The reproductive period in 1987 was subdivided into an
early (April to mid-May), a main (late May to early July),
and a late breeding period (mid-July to August). The whole
reproductive period comprised nine calling periods. In 1986
tadpoles of different sizes in the local ponds demonstrated
that there had been breeding activity before June. Neverthe-
less, the late breeding period (July) was clearly separated
from the previous one. The pattern of periodical fluctuation
was parallel in all breeding areas, but drying up of the
ponds impeded breeding activity in area I in May and July
and in area II in April (Figs. 3-5).
Principal component analysis revealed little covariance
of the measured environmental variables with the numbers
of either males or females (Table 1). Numbers of individuals
were associated with the available water surface area
(WATER1, WATER2) and to a far lower degree with the
ambient temperature (TEMPDMAX, TEMPSS, TEMPN-
MIN, TEMPNAV). Tests for circular correlation with
moon phases gave negative results (r = 0.084, P > 0.05). Sta-
tistical models were constructed that predicted the numbers
of males or females at the breeding sites based on both
a water surface variable (WATER1) and a night tempera-
ture variable (TEMPSS or TEMPNAV, Table 2). However,
the contribution of these factors to the variance in toad
numbers was low: 41.0% in males and 23.4% in females.
401
1986 1987
~) 100
0 , d, ,
-
!~ 0 ~5 i8' i ?-III
Fig. 2. Time course of breeding activity represented as the number
of males captured in areas I-III (upper graph) and the number
of spawn strings in areas I-IV (lower graph). The total number
of spawn strings deposited during each calling period is also given
Discriminant functions based on all environmental vari-
ables predicted more successfully presence or absence of
males and females and spawn deposition (Fig. 6). Neverthe-
less, the rates of wrong classification ranged from 12.2%
to 25.0% (Table 3).
Migratory behaviour of males during the reproductive period
In 1987 the toe-clipped individuals demonstrated a remark-
able fidelity to the breeding area where they had originally
been captured. Only 4 out of 133 males marked in area
I were recaptured in areas II and III, whereas none of the
24 toads marked in area II and 89 marked in area III were
ever observed at other sites. However, every breeding chor-
us of at least 20 males in any breeding area was clearly
audible in each of the neighbouring ones.
The absolute and relative recapture rates of marked
males decreased during the course of the reproductive peri-
od (Fig. 7). Yet not even during the first maximum of call-
ing activity (April in areas I and III, May in area II), were
all previously marked toads recaptured simultaneously: at
most 60% were. This low rate was probably due to the
inactivity of some toads and also to the limited period of
capture. Since the estimates of individuals within a given
area depend on the total number Arm of marked individuals,
it was assumed that e.g. a recapture rate of 53 toads in
area III meant that all 89 marked males were present. Dur-
ing the following calling periods Arm for every breeding area
was corrected by multiplication by the quotient: maximum
number of males recaptured during one night of a calling
period/maximum number of toads recaptured during the
first calling period.
Estimates of the numbers of males present during the
nine calling periods revealed that at all sites many toads
402

Table 1 A, B. Principal component analysis of data sets each containing 18 variables with 102 observations in area III. Variable names
are explained in methods section. Six principal components (PC) with eigenvalues > 1 are presented, after orthogonal VARIMAX-rotation
A Association of male numbers and environmental conditions

PCI PC2 PC3 PC4 PC5 PC6

MALES 0.329 0.051 0.168 0.726 --0.154 0.070

WATER/ 0.226 0.201 0.193 0.802 - 0.032 0.026
WATER2 - 0.175 0.171 - 0.022 0.836 - 0.045 - 0.014
TEMPDMAX 0.738 --0,539 --0.063 --0.100 0.105 --0.177
TEMPSS 0.865 -0.413 -0.002 0.054 0.018 -0.117
TEMPNMIN 0.961 0.166 - 0.011 0.150 0.008 --0,066
TEMPNAV 0.972 -0.062 -0.028 0.107 0.006 - 0.121
TEMPRATE --0.023 0,877 --0.095 0,109 -0.024 0.091
HUMIDMIN -0.082 0.665 0.355 0.333 -0.134 0.122
HUMISS 0.084 0.654 0.588 0.250 0.025 0.118
HUMINMAX -- 0.080 0.030 0.927 0,076 -- 0.079 0.054
HUMINAV 0.044 0.328 0.858 0.172 0.046 0.165
PRESS 0.058 - 0.191 - 0.054 - 0.104 0.930 - 0.251
PRESSR - 0.024 - 0.153 0.020 - 0.082 0.924 0.281
PRES-24 -0.122 -0.004 0.055 0.019 0.152 0.923
P R E S - - 12 --0.174 0.161 0.147 0.044 - 0.J45 0.877
SOIL 0.178 -0.586 --0.206 - 0.256 0.318 -0.229
RAIND -0.169 0.647 0.241 0.132 -0.334 -0.093
RAINN -0.302 0.510 0.327 -0.105 -0.135 -0.175
B Association of female numbers and environmental conditions

PC1 PC2 PC3 PC4 PC5 PC6

FEMALES 0.249 --0.085 0.093 -- 0.156 0.098

WATER1 0.236 0.26t 0.204 -0.016 0.755 0.015
WATER2 - 0.162 0.226 - 0.012 - 0.027 0.825 - 0.024
TEMPDMAX 0.744 --0.536 --0.075 --0.109 --0.072 -0.171
TEMPSS 0.868 - 0.405 - 0.009 0.021 0.076 - 0.114
TEMPNMIN 0.961 0.186 -0.002 0.004 0.128 -0.071
TEMPNAV 0.973 -0.047 - 0.025 0.005 0,105 --0.122
TEMPRATE -0.027 0.892 -0.072 -0.036 0.051 0.080
HUMIDMIN -0.091 0.667 0.376 -0.140 0.298 0.111
HUMISS 0.079 0.655 0.607 0.017 0.210 0.110
HUMINMAX 0.078 0.016 0.926 - 0.077 0.059 0.056
HUMINAV 0.042 0.321 0.866 0.042 0.142 0.162
PRESS 0.061 -0.190 -0.056 -0.928 -0.101 -0.249
PRESSR -- 0.023 - 0.151 0.018 -- 0.922 -- 0.069 0.282
PRES 24 - - 0.120 0.015 0.053 0.157 0.008 0.921
PRES-- 12 --0.179 0.153 0.150 --0.147 0.072 0.875
SOIL 0.179 -0.600 -0.220 0.320 -0.218 - 0.221
RAIND -0.189 0.606 0.263 -0.350 0.157 -0.101
RAINN -0.320 0.465 0.342 -0.150 -0.099 -0.181

m i g r a t e d i n t o the b r e e d i n g areas in late M a y a n d m i d - J u l y , the c o u r s e o f the r e p r o d u c t i v e period. In area I, h o w e v e r ,

while others e m i g r a t e d in late J u n e (Figs. 3-5). D u r i n g
A p r i l the m i g r a t o r y b e h a v i o u r o f m a l e s differed a m o n g the
three areas: there was s t r o n g i m m i g r a t i o n into a r e a I, no
t o a d s p r e s e n t at a r e a II, a n d an a l m o s t u n c h a n g e d n u m b e r
o f m a l e s in a r e a III. T h e analysis o f size frequencies in
b r e e d i n g males p r o v i d e d f u r t h e r evidence t h a t in a r e a I I I
three distinct g r o u p s o f males b r e d at different times
(Fig. 8). T h e size distributions o f the calling p e r i o d s A p r i l - /
a n d M a y - I were n o t significantly different (H-test, P >
0.05), b u t were significantly different f r o m the succeeding
ones (H-test, P < 0.001). A g a i n , the d i s t r i b u t i o n s f r o m M a y -
2 to July-1 did n o t differ significantly (H-test, P > 0 . 0 5 ) ,
while a significant c h a n g e in size-class f r e q u e n c y o c c u r r e d
in the July-2 calling period. (H-test, P < 0 . 0 1 ) . T h e initially
p r e d o m i n a n t small males were r e p l a c e d by larger ones o v e r
the size frequencies o f males r e m a i n e d u n c h a n g e d f r o m
A p r i l to A u g u s t (H-test, P > 0 . 0 5 , Fig. 9). In a r e a II
n u m b e r s o f males w e r e t o o small for analysis.
T h e i n d i v i d u a l p a t h s o f m a l e s d u r i n g 3 days o f extensive
calling activity w e r e m o n i t o r e d in a r e a I d u r i n g J u n e 1986
(Fig. 10). T h e m i g r a t o r y b e h a v i o u r o f all individuals was
essentially the same. D u r i n g d a y l i g h t they r e m a i n e d h i d d e n
u n d e r plants n e a r the b r e e d i n g ponds. I m m e d i a t e l y after
sunset they m o v e d to a calling site within the w a t e r b o d y
o r at the edges o f the p o n d . D u r i n g the n i g h t the m o n i t o r e d
individuals stayed in this small a r e a a n d called. M o s t m o v e -
m e n t s within the calling sites were r e a c t i o n s to o t h e r ap-
p r o a c h i n g toads. A limited site fidelity was f o u n d in s o m e
individuals which were r e c a p t u r e d up to eight times within
the s a m e p o n d . A t the end o f the first July calling period,
 403

Table 2 A, B. Statistical models to predict the number of individuals Breeding Activity

MALES
in breeding area III on the basis of the environmental variables9
The best variable subsets as computed by stepwise multiple regres-
sion are presented9 Variable names are explained in methods sec-
tion. Rz is the measure of determination, that is, the portion of
~176
i I
9
8oo
3oo-I
II
I~
ESTIMATE
At Area I

75~ FEMALES
total variance explained by the action of the independent variables
.c~PAWN PER
A Dependent variable: number of males '4 CALLING
PERIOD

Variable Coefficient P Contr. to R z Total R 2 2~ / l' so~

Intercept
WATER1
TEMPNAV
--/5.306
0.028
2.282
0.1365
<0.0001
0.0154
0.386
0.024 0.410
o00t I / IF ,o
B Dependent variable: number of females

Variable Coefficient P Contr. to R 2 Total R 2

0tA,,,Wli
April May June July 1987
ol A p r i l May June' July ]987

Fig. 3. Breeding activity in area I. (Left: Capture rate of males

as individuals per 1 man-hour (hatched columns) and total number
Intercept -2.212 0.0507 (NTOT) of reproductive males estimated from the recapture rates9
WATER/ 0.002 <0.0001 0.208 Squares represent a single estimate, otherwise weighted means and
TEMPSS 0.183 0.0226 0.026 0.234 the corresponding 99.9% confidence interval (vertical bar) are giv-
en. The temporal range of estimates is indicated by horizontal bars.
Right: Reproductive activity of females is represented by the
number of spawn strings counted during each calling period
Table 3A, B. Statistical significance of the discriminant functions
(Fig. 6) and predictability of presence or absence of males, females,
and oviposition (spawn) at the breeding ponds in area III 40 B r e e d i n g Activity
MALES
A Significance levels (]8 degrees of freedom) ESTIMATE At Area II
NTOT + 3 SD
ac
Variable Eigen- Canonical Wilks Chi- P
value Correlation Lambda Squared
9
2c FEMALES
Male 0.741 0.652 0.574 50.47 0.00006 o
Female 0.664 0.632 0.601 46.34 0.00026 g ,PA ZE"
Spawn 0.442 0.554 0.693 33.34 0.0]520 o lC 10 ~RIOD
C~
(D
B Success rates (%) of discriminant classification
;z; C: 0.
April' May ' June' July ]987 April May June July 1987
Actual Predicted
Fig. 4. Breeding activity in area II. Left: Capture rate of males
Present Absent as individuals per 1 man-hour (hatched columns) and total number
(NTOT) of reproductive males estimated from the recapture rates.
Males Females Spawn Males Females Spawn Weighted means and the corresponding 99.9% confidence interval
(vertical bar) are given. The temporal range of estimates is indicated
Present 85.5 87.8 80.0 14.5 12.2 20.0 by horizontal bars. Right: Reproductive females are represented
Absent 17.5 19.7 25.0 82.5 80.3 75.0 by the capture rate (hatched columns) and the number of spawn
strings counted during each calling period

400
Breeding Activity
1986, in area II migratory activity increased and toads left
At Area III
the ponds (Fig. 11). Of 8 monitored males, 7 stayed within
m~TE -~ ~L I I
a radius of 100 m, while 1 departed to a maize field 340 m "~ aOO
away. Following a foraging period of a b o u t 2 weeks, 6
of these toads called again in late July 1986. The tracked
o FEMALES
by radio-telemetry in area III showed a similar attachment
~200 50-
to their breeding area between the April-1 and May-1 o%
" ~ SPAWN PER l
breeding periods in 1987. r ,c o ,

190

Migratory behaviour of females during the breeding season

,JL MJ!
,m
At the beginning of a calling period females were never
0
found within a breeding area. They always arrived at the z April May June July 1987 April May June J u l y 1987
ponds after several days of extensive vocalization of local Fig. 5. Breeding activity at area III. Presentation of data as in
males (Figs. 4, 5). They stayed at most 3 days but usually Fig. 4, except for the capture rate, here given as toads per 2 men-
only 1 day, spawned, and left. N o n e of 27 toe-clipped fe- hours
404

i0 j MALES ,' .

i,o::1
g o. I ~=~A~=s

2
- ,~ "~-~
& IIIIf~
7///////4,,
~ E 3 I 1 I_1 I l i l t

~//////~../~, ~--/// / / / // / / / / /~"

6o~////~ /~Y 1 - - . / . A ~ / / / / ~ / /
I //
i ,#
,

2 tO SPAWN pr,~,~at
,.~ ^ ..,ill I I
~~ =~ ,:',=~ g , 2 : '= t,=l
I01~ /' i I I '1 I I t' 'i' '1 i
absent

I' 'f'"'i"'T 'l '"

Fig, 8. Distribution of size classes (4 mm width) of toads captured
in area III. Left: distributions of each calling period; Right: histo-
-4 -3 -2 -1 0 1 2 3 4 grams summarizing those distributions which and did not differ
Canonical Variable among each other significantly (H-Test, P>0.05)
Fig. 6. Discriminant grouping patterns of observations in area IIi.
Values on the abscissa are of discriminant scores, and observations
are grouped at class widths of 0.25 units. Days with males, females, _ i

or new spawn strings at the breeding ponds are grouped above ~ ~ - ~ L ~ ~

the score line (shaded) and those without are below (unshadedhisto- E~ ~NS
~.=s) rria.r representgroup centroid= ~//////2ZmJ ,r = -t / / J / ~ / / ' ~
IW---

~o_
It
i=0~ ill ~ P
ol. lui.,, nmlln.a. . u l. r
~
~176 //~" 7 / - / - ~ / - - ~
. -

i ~,,o:i ll\! ~io 5~ 72 40 5e 7~ 40 5e, 71~ [ram]

~ii!iJlll~
S n o u t - V e n t - L e n g t h h~ 4 r a m Classe,
Fig. 9. Distribution of size classes (4 mm width) of toads captured
~o0 in area I. Presentation of data as in Fig. 8

40 the reproductive period into early, main and late breeding

2 periods; (2) subdivision of each breeding period into several
April May J u n e ' July 1987
Fig, 7. Absolute and relative recapture rates of,toe-clipped toads
in area III throughout the reproductive period of"1987
males was recaptured after spawning within any breeding
area. The size frequencies of females varied significantly
among the breeding areas and also among the calling peri-
ods (H-test, P<0.01, Figs. 8, 9).
One female was tracked by radio-telemetry after ovipo-
sition in April 1987 in area III. This individual left area
III directly and migrated with a speed of 200-450 mtnight
towards the south. Her final position was 985 m away from
the spawning site at the edge of a wheat field.
Discussion
The observed population of Bufo calamita showed a two-
fold temporal spacing of breeding activity: (1) division of
short calling periods. Different intensities of spawning ac-
tivity have also been observed in other populations
throughout the whole geographical range of this toad: in
Sweden two maxima of spawn deposition in May and July
(Andrdn and Nilson 1985; HSglund and Robertson 1986),
in England and Poland two maxima in April and June
(Mathias 1971; Kowalewski 1974; Arak 1983b), in Ger-
many three maxima in April, June, and July/August (Bre-
gulla 1986), and finally in Spain three maxima in January,
late February, and March (Tejedo 1986). Thus a division
of the reproductive period into two or three breeding peri-
ods is a common feature ofB. ealamita populations whatev-
er their geographical location.
Nonetheless little is known about the reasons for these
variations of breeding intensity within a population. Since
toads were ectothermal organisms, exogenous factors such
as weather conditions might account for the observed fluc-
tuations. In fact meteorological conditions influence consid-
erably the onset and duration of breeding activity at the
northern and southern edges of their range: low tempera-
tures impede breeding before May in Sweden (Andrdn and
Niison 1985), while in Spain long periods of drought post-

5O[
[m]
0 0
0 [m] 30
Fig. t0. Paths of four males in area I during 5-8 June 21986.Capture
sites in pasture are indicated by S, the surface area of the breeding
ponds is hatched
pone or even prevent breeding (Diaz-Paniagua 1986; Te-
jedo 1986). Thus, at least near the limits of distribution
a great part of variation of breeding intensity is due to
environmental factors.
In the central region of geographical distribution weath-
er conditions are rarely outside the tolerance range for
breeding activity (Mathias 1971). By multivariate analysis,
the present study confirms that meteorological factors ac-
count for only a small portion of temporal fluctuations,
with maxima of oviposition in April, June, and late July
1987. It was not even possible to distinguish clearly days
with oviposition from those without on the basis of climatic
data (discriminant analysis). The mark recapture calcula-
tions and differences in size frequency rather suggest that
the three breeding periods are due to the immigration and
emigration of three distinct groups of males.
The first indication of distinct immigration waves of
males was found by Flindt and Hemmer (1968b). They
found that larger ("older") toads tended to arrive earlier
at the breeding pond than smaller ("younger") ones and
concluded that different temperature sensitivity might be
responsible. However, recent studies have demonstrated
that there exists only a slight, if any, correlation between
body size and age of a toad (Hemelaar 1981, 1985; Gittins
et al. 1982; Gibbons and McCarthy 1983; Acker et al. 1986;
Jorgenson et al. 1986). In addition the succession from large
to small individuals seems to be a local phenomenon, since
405
100
wheat
field
Ira]
0 [m] 60
Fig. 11. Paths of three males in area II during 7-14 July 1986.
All paths start at the edges of ponds (horizontally hatched). Dots
indicate where toads rested during daylight. The area between the
roads was bare ground including several compost heaps covered
with herbs (cross-hatched)
in area III the order was reversed and in area I size frequen-
cies did not change at all. Thus, a more reliable conclusion
is that the size of the toads reflects their feeding condition
rather than age.
The origin of the three male groups remains to be dis-
cussed. In area III all single estimates of active toads were
about 100 individuals without change until late May indi-
cating a local group of males. In area I, however, estimates
showed steadily increasing toad numbers suggesting immi-
gration from the adjacent areas. Therefore, the first maxi-
mum of calling activity was probably due to toads which
had hibernated within or at least near the later breeding
area. Since in area I the size-class frequencies of reproduc-
tive males did not differ among the three breeding periods,
all reproductive activity apparently relied on the same pool
of males. In area III the analysis of size frequencies indi-
cated immigration of males from different locations during
the second and third maxima of calling activity. Since the
speed of migration of B. calamita was similar to that of
B. bufo (van Gelder et al. 1986), this species is probably
also capable of moving over large distances. By chance,
one marked male was recaptured 2075 m away from its
breeding area. The migratory range of B. calamita is cur-
rently being studied.
The physiological mechanism which causes an individ-
ual natterjack toad to breed earlier or later than other
members of the same population is unknown. However,
406
multimodal variations of spawning intensity have also been
observed in other Bufo species with prolonged breeding:
B. melanostictus (Jorgenson et al. 1986) and B. spinulosus
(Sinsch, in prep). Studies on the ovarian cycte of B. melanos-
tictus confirmed that all reproductive stages can be observed
during most parts of the year, but with different relative
frequencies (Jorgenson et al. 1986). In India this species
breeds during the wet season with two distinct activity max-
ima. Heavy rainfalls seem to trigger oviposition in the main
breeding periods. However, the asynchrony of individuals
is not yet understood in this species either. In conclusion,
the physiological basis of the three groups of individuals
which breed at different times remains subject to specula-
tion. Long-term studies are needed to decide whether e.g.
the date of first sexual maturity determines the annual onset
of breeding activity in every individual during its life span.
The duration of breeding activity in individuals varied
considerably. Since the duration of the ovarian cycle in
temperate-zone anurans is about 6 months (Jorgenson
1986), it is most likely that females breed only once a year.
In fact female natterjack toads have never been observed
to spawn more than once a year in the field (Beebee 1979),
and more recent studies, including this one, confirm this
opinion. Males, however, were certainly involved in several
calling periods. Three individuals distinguished by toe-clip-
ping and special colouration and/or size even participated
in all calling periods. Since there were frequently breaks
of several days between the recaptures, it is likely that even
these individuals did not breed continuously.
The periodical variations of calling intensity in males
within the reproductive period have never been described
before. The numbers of calling males at the breeding areas
fluctuated sinusoidally. In 1987 a total of nine temporally
spaced calling periods was observed. There are at least three
possible reasons why previous studies failed to detect these
short-term variations: (1) small population size; (2) low
frequency of observations; and (3) unfavourable meteoro-
logical conditions.
In small populations temporal variations of breeding
intensity are difficult to interpret because of the low number
of both individuals and spawn strings. During two repro-
ductive periods Niekisch (1981) observed daily a population
of 47 toads which deposited 10 strings in 1978 and another
13 in 1979. Although the number of toads present at the
breeding ponds (at most 12) did not show regular patterns
of variation, apparent breaks of about 10 days between
periods of spawn deposition indicate short-term spacing:
e.g. in 1979, 6-16 May: 8 strings; 27 May: 2 strings; and
7-11 June: 3 strings (cited from Fig. 7 and Table 3 in Niek-
isch 1981).
Daily recording of toad activity or oviposition is neces-
sary to detect short-term variations. This condition was
not met in the studies of Andr~n and Nilson (1985) in Swe-
den or of Bregulla (1986) in Germany.
Meteorological variables such as rainfall and tempera-
ture affect breeding activity. Various authors state that nat-
terjack toads prefer rainy days for spawning (Kowaleski
1974; Niekisch 1981; Tejedo 1986) and require night tem-
peratures of at least 8 ~ C for chorusing (Flindt and Hemmer
1968b; Mathias 1971; L6rcher and Schneider 1973). Dur-
ing dry and cold reproductive periods in southern Spain
weather variables accounted almost entirely (94.6%,
91.6%) for variation in numbers of breeding males and
females (Tejedo 1986). During the 1987 study reported here,
rain fell almost daily and temperature rarely dropped below
the critical limit, and multivariate analysis demonstrated
that the measured rainfall and temperature variables did
not limit the toads' activity. Therefore, the recorded short-
term variations in breeding activity were apparently due
to endogenous changes in the toads. Since vocalization is
expensive in terms of energy (Bucher et al. 1981; Pough
and Gatten 1984), it is likely that resting periods are re-
quired to recover, even if environmental conditions do not
impede calling. In conclusion, short-term variations in
breeding activity probably reflect the physiological limits
of toads during weather conditions which permit continu-
ous reproductive activity.
Acknowledgements. I thank the following institutions for issuing
the necessary permissions: Regierungsprfisident Kgln, Untere Na-
tur- und Landschaftsschutzbeh6rde Siegburg, Ordnungsamt St.
Augusfin, Umweltamt St. Augustin. I am grateful to H. Schneider
for helpful comments on this paper. Especially I am indebted to
R Canales for his indefatigable assistance in the field throughout
the study period.
References
Acker PM, Kruse KC, Krehbiel EB (1986) Aging Bufo marinus
by skeletochronology. J Herpet 20:570-575
Andr6n C, Nilson G (1985) Breeding pool characteristics and re-
procution in an island population of natterjack toads, Bufo
calamita Laur., at the Swedish west coast. Amphibia-Reptilia
6:137-142
Arak A (1983 a) Male-male competition and mate choice in anuran
amphibians. In: Mate choice (ed Bateson): 181-210
Arak A (1983b) Sexual selection by male-male competition in nat-
terjack toad choruses. Nature 306:261-262
Batschelet E (1981) Circular statitistics in biology. Academic Press
(London New york Toronto Sydney San Francisco) 372 p
Beebee TJC (1979) A review of scientific information pertaining
to the natterjack toad Bufo calamita throughout its geographi-
cal range. Biol Conserv 16:107-134
Bregulla D (1986) Untersuchungen zur Wasserchemie von Kreuz-
kr6ten-Laichgew/issern. Salamandra 22:173 179
Bridson RH (1976) The distribution of the natterjack toad in south
west Scotland 1976. South Scotland, internal report. Nature
Conservancy Council, cited from Beebee (1979)
Bucher TL, Ryan M J, Bartholomew GA (1981) The cost of croak-
ing in a frog Physalaemus pustulosus (Leptodactylidae). Am
Zool 20: 909
Diaz-Paniagua C (1986) Reproductive period of amphibians in the
biological reserve of Donana (SW Spain). In: Studies in herpe-
tology (ed Rocek Z):429-432
Dole JW (1965) Summer movements of adult leopard frogs Rana
pipiens Schreber in northern Michigan. Ecology 46:236-255
Eibl-Eibesfeldt I (1950) Ein Beitrag zur Paarungsbiologie der Erd-
kr6te (Bufo bufo L.). Behaviour 2:217-233
Flindt R, Hemmer H (1968 a) Analyse des akustischen Geschlecht-
serkennungsmechanismus (Befreiungsrufe) bei Kr6ten (Genus
Bufo). Experientia 24:285 286
Flindt R, Hemmer H (1968b) Beobachtungen zur Dynamik einer
Population von Bufo bufo und Bufo calamita. Zool Jb Syst
95 : 16~186
Flindt R, Hemmer H (1972) Untersuchungen zur Reaktion yon
Bufo calamita und Bufo viridis auf arteigene Rufe. Biol Zbl
91 : 597-600
Flury B, Riedwyl H (1983) Angewandte multivariate Statistik. Gus-
tar Fischer Verlag (Stuttgart New York) 187 p
Gelder JJ van, Aarts ttMJ, Staal HJWM (1986) Routes and speed
of migrating toads (Bufo bufo L.): a telemetric study. Herp
J1:111-114
Gibbons MM, McCarthy TK (1983) Age-determination of frogs

and toads (Amphibia, Anura) from northwestern Europe. Zool
Scr 12:145-151
Gittins SP, Steeds JE, Williams R (1982) Population age-structure
of the common toad (Bufo bufo) at a lake in Mid-Wales deter-
mined from annual growth rings in the phalanges. Brit J Herpet
6:249-252
Hemelaar ASM (1981) Age determination of male Bufo bufo (Am-
phibia, Anura) from the Netherlands, based on year rings in
phalanges. Amphibia-Reptilia 1 : 223-233
Hemelaar A (1985) An improved method to estimate the number
of year rings resorbed in phalanges of Bufo bufo (L.) and its
application to populations from different latitudes and alti-
tudes. Amphibia-Reptilia 6 : 323-342
Heusser H (1967) Wanderungen und Sommerquartiere der Erd-
kr6te (Bufo bufo L.). Unpubl Ph D, Univ Ziirich
Heusser H (1968) Die Lebensweise der Erdkr6te (Bufo bufo L.):
Wanderungen und Sommerquartiere. Rev Suis Zool
75 : 928-982
Heusser H (1969) Unkreaktion mit Befreiungsruf beim Weibchen
der Kreuzkr6te Bufo calamita. Experientia 25:121-123
H6glund J, Robertson J (1986) Species isolating mechanisms in
a population of sympatric toads. In: Studies in herpetology
(ed Rocek Z) : 553-556
Jorgensen CB (1986) External and internal control of patterns of
feeding, growth and gonadal function in a temperate zone an-
uran, the toad Bufo bufo. J Zool 210:211-241
Jorgensen CB, Shakuntala K, Vijayakumar S (1986) Body size,
reproduction and growth in a tropical toad, Bufo melanostictus,
with a comparison of ovarian cycles in tropical and temperate
zone anurans. Oikos 46:379-389
Kowalewski L (1974) Observations on the phenology and ecology
of amphibia in the region of Czestochowa. Acta Zool Cracov
19:3191-458
407
L6rcher K, Schneider H (1973) Vergleichende bio-akustische Un-
tersuchungen an der Kreuzkr6te, Bufo calamita (LAUR.), und
der Wechselkr6te, Bufo v. viridis (LAUR.). Z Tierpsychol
32:506-521
Mathias JH (1971) The comparative ecologies of two species of
amphibia (Bufo bufo and Bufo calamita) on the Ainsdale sand
dunes national nature reserve, unpubI Ph D thesis, Univ Man-
chester
Niekisch M (1981) Beitrag zu Biologie und Schutz der Kreuzkr6te
(Bufo ealamita LAUR.). Decheniana 135:88-103
Pough FH, Gatten RE jr (1984) The use of anaerobic metabolism
by frogs Hyla crucifer in a breeding chorus. J Comp Physiol
A 78:337-340
Rfist H (1969) Sch/itzung der Amphibienbest~inde in einem Teich
durch ein Wiederfangverfahren. Viertelj Nat-forsch Ges Zfirich
114:279-291
Sinsch U (1987) Orientation behaviour of toads (Bufo bufo) dis-
placed from the breeding site. J Comp Physiol A 161:715-727
Sinsch U (1988a) Seasonal changes in the migratory behaviour
of the toad Bufo bufo : direction and magnitude of movements.
Oecologia 76 : 390-398
Sinsch U (1988 b) Auskiesungen als Sekund~irhabitate ffir bedrohte
Amphibien und Reptilien. Salamandra (in press)
Tejedo M (1986) Comunicacion acustica y seleccion sexual en Bufo
calamita (Laurenti 1768). Unpubl thesis, Universidad de Cor-
doba
Weber E (1978) Distress calls of Bufo calamita and Bufo viridis
(Amphibia, Anura). Copeia 1978:354-356
Wells KD (1977) The social behaviour of anuran amphibians.
Anim Behav 25 : 666-693
Received January 16, 1988