Content uploaded by Jan Willem Arntzen
Author content
All content in this area was uploaded by Jan Willem Arntzen on Oct 07, 2018
Content may be subject to copyright.
Reproductive biology of the parsley frog, Pelodytes punctatus, at
the northernmost part of its range
A.G. Toxopeus, M. Ohm, J.W. Arntzen1
Institute of Taxonomic Zoology,
University
of Amsterdam, Mauritskade 57, P.O. Box 4766, 1009 AT
Amsterdam,
The Netherlands
1 address for correspondence and reprint requests: Department of Zoology, University of Leicester,
University
Road, Leicester LE1 7RH, United Kingdom
Abstract. The reproductive biology
and population dynamics
of Pelodytes
punctatus
were studied at the
breeding season over a three year period in a coastal dune system
located at the extreme northwestern
border of the species'
range. Adult population size estimates
ranged from about 100 in the first year to 60
in the third year. Males were remarkably sedentary near the pond under artificially provided shelters.
Many were observed
during the most of the breeding season
which lasted from mid-March or the end of
March to the end of April or mid-May.
Most spawning
took place in the second half of March or early
April. In two years out of three a second
period of spawning
involving
fewer animals was observed
in the
first half of May. Both
periods of spawning
coincided
with, or shortly
followed,
periods
of rising
median air
temperature. Egg-clutches
were deposited
in the deepest parts of pond, mainly on submerged vegetation
not reaching the surface.
An average
sized clutch contained approximately
360 eggs. Development
of the
embryos
until hatching took from 4 to 14 days, depending on the ambient temperature. Larval develop-
ment and growth were fast. Recently
metamorphosed froglets
at a size of around 18 mm were found from
the end of May onwards.
Juveniles may reach adult size in the autumn of the year that they were born.
Adult frogs did not show a strong fidelity to the breeding pond between years. In the study area the
population structure of Pelodytes punctatus
seems to be best described by a source - sink model in which
flourishing populations in the dunes give rise to short lived satellite
populations outside the dunes.
Introduction
The parsley frog, Pelodytes punctatus (Daudin, 1803), is distributed over a major part of
western and south-western Europe. The French Channel coast is at the extreme north-
western part of its range (fig. 1). The species was discovered by Giard (1890) to occur
north of Boulogne-sur-Mer. In 1951 the species was found again by the late D. Hillenius
(pers. comm., 1976), while Arntzen and Gerats (1976) documented the existence of some
flourishing populations in the area.
We studied the reproductive biology and population dynamics of the parsley frog with
the aim of gaining a better understanding of its ecological requirements, especially those
that may govern its existence at the northern edge of its range.
132
Figure 1. Distribution of Pelodytes punctatus.
The study area in north-western France is: 1)
UTM grid NF27,
(Crespo
and Oliviera, 1989);
indicated
by an arrow. Reliable records documenting
the outer range of the
species
are numbered as follows.
2)
Portalegre,
(Boscá, 1880); 3)
Los Pedroches
(Lope7-jurado
et al., 1980); 4)
San Pablo
de los Montes,
(March 1984,
I. Esteban and A. Mugica, pers. comm.); 5) Uruefia,
(June 1990,
M.
Garcia-Paris,
pers. comm.);
6) Cisneros,
(March 1980, J.W.
Arntzen
and A. Zuiderwijk, unpubl.);
7) UTM
grid WN14, (A. Bea, 1985); 8) Saint
Jean de Pied-de-Port,
(Martinez Rica, 1983); 9) Bellevue,
Sangatte
and Peuplingues
(Kroese
en van Leeuwen,
1979),
confirmed
by M. Ohm and D. van der Kroef in
1983; 10-14) Institute Géographique National maps 1/50,000 nr. XXXI-10 (Montmedy),
XXXIII-13
(Chambley),
XXXIII-30
(Seyssel),
XXXIV-38
(Chorges),
XXXVII-42
(Menton),
all in Le Garff (1989);
15)
Castino, (Bruno, 1977); 16)
Rapallo, Genova, (Bruno, 1977;
Sindaco and Andreone, 1988).
Numbers are
placed at the outside of the indicated distribution. The occurrence of the species
in Belgium
(Schreitmüller
and Wolterstorff,
1923;
see Parent 1984 for further references), Luxemburg (Hoffmann, 1958), Germany
(Fournel, 1836 in: von Bedriaga, 1891)
and northwestern
Spain and Portugal (Regulez
Fernández, 1988;
Vieira, 1887;
Crespo, 1971)
is unconfirmed
(Holandre, 1851 in: von Bedriaga, 1891; Parent, 1970, 1984;
Arntzen, 1981;
Crespo and Oliveira, 1989;
M. Garcia-Paris.
pers. comm., 1991).
The map comprises
two as
yet unpublished
records in southeastern
Spain
(Alcoy
and Vélez
Rubio)
made
in 1990 and 1991
by M. Garcia-
Paris, completing
recent maps such as in Crespo and Oliveira
(1989).
133
Material and methods
Research was carried out in a coastal dune system ('Dunes de Slack') south of Amble-
teuse in the département Pas-de-Calais, France during the breeding season from 1981 to
1983. In the inner dune slacks many natural freshwater ponds can be found, while some
ponds have been enlarged and others newly created to serve for duck hunting. Three
ponds in an accessible part of the dunes were selected for the study on the parsley frog.
Two small and shallow ponds (maximum size of 90 m2, up to 35 cm deep) were located
close to one another in an abandoned sand quarry. (These ponds are very similar to the
dune slack depicted by Arntzen, 1981: fig. 31, and by Arntzen and Teunis, 1993). A
bigger and slightly deeper (up to 50 cm) horseshoe shaped pond with a surface of
approximately 1900 m2 was also selected. The ponds are separated by 300 m in direct
line. Due to an unacceptably high level of vandalism in the quarry, Horseshoe Pond only
was studied in 1982 (by Den Uijl, 1982) and in 1983. Each summer, the ponds gradually
decreased in size to dry out completely (in the quarry) or to reduce to about one third of
its original size (Horseshoe Pond).
The margins of Horseshoe Pond were covered with a dense vegetation of Leptodictyum
riparium, Claytonia perfoliata, Veronica scutella and Poa annua while higher up a dense Salix
repens, Salix purpurea and Hippophae rhamnoides vegetation and planted Pinus nigra was
found. In all ponds the aquatic vegetation was scarce, mainly consisting of Mentha
aquatica, Potamogeton perfoliatus,
Chara sp. and some Phragmites australis and Cirsium arvense.
In marshy areas of Horseshoe Pond Hydrocotyle vulgaris, Rumex hydrolapathum, Iris pseu-
dacorus and Juncus inflexus
was found. The margins of the two quarry ponds were almost
barren. All three ponds were virtually unshaded.
The parsley frog is a nocturnal creature which hides during the day under vegetation
cover, trash or in self-dug holes. To trace the animals more easily than could otherwise
be achieved, artificial hiding places were created. These were pieces of wood, rags,
metal, carpet etc., varying in size from 0.4 to I m2 that were collected from the local
refuse dump. These were distributed around the pond at more less regular intervals
averaging from 1 1 m in 1981 to 8 m in 1983. The artificial shelters were placed in early
spring at a short distance from the pond margin (50 cm to 100 cm in spring), were they
remained for the whole study period, after which they were removed. The shelters were
checked for frogs on a daily basis. All animals were measured dorsally for snout - vent
length (SVL, from the tip of the snout to the posterior side of the urostyl). In 1981 and
1983 all captured animals were sexed (males are characterized by breeding pads on their
breast, arms and fingers). All metamorphosed animals with a SVL of mm could be
sexed unequivocally. They were released on the spot after being toe-clipped to allow
individual recognition. The adult population size (NJ was estimated by census, for
males and females (Nf) separately. Males were counted directly and the estimated
population (N £ equalled the total number of different males observed. For the females
Nf was estimated indirectly as equalling the total number of egg-clutches found in the
pond. To gain an impression about the frequency with which frogs go unnoticed, we
134
calculated a 'census oversight index', by dividing the number of days any individual frog
was observed in the time span between its first and last appearance around the pond by
the total number of opportunities (days) for observation and we then averaged this
measure for all males and all females.
The clear ponds were searched daily for newly deposited egg-strings. For each finding
the following parameters were recorded: location (relative to a grid system along the
shore of the pond), length of the egg-string in a straight line (as a multiple of 5 mm along
the unmanipulated egg-string), depth of the pond at egg-string location (in cm), the
depth of the water column from the centre of the egg-string to the surface (in cm) and
the number of days it took half of the eggs in the clutch to hatch. In 1983, embryonic
development was followed in more detail for seven clutches including some of the
earliest and some of the latest spawnings of the year. The water temperature was
measured daily on the spot by a minimum-maximum thermometer. Since it proved
difficult to find larvae in the pond after they had hatched, and to avoid the mixing of
larvae of different ages, growth was studied for a sample of larvae in netting cage (50 cm
high, diameter 70 cm) that was placed around an egg-string on a representative site in
the pond (see fig. 6). Growth was also studied for a single larval cohort in a pond at close
proximity in 1975. Measurements taken on larvae were total length (TL) from the tip of
the snout to the tip of the tail and SVL from the tip of the snout up to and including the
posterior side of the body.
Daily minimum and maximum air temperature were obtained from the meteorologi-
cal station Boulogne-semaphore (altitude 0 m) in Boulogne-sur-Mer, situated 10 km
from the study area.
Results
Breeding and spawning
The number of male frogs found under the artificial hiding sites during the course of
1981 is shown in fig. 2. On 12 March the first male was found in the quarry. In
Horseshoe Pond the first males were observed on March 25. Their numbers were
highest by the middle of April (in Horseshoe Pond) and early May (in the quarry). From
those points in time onwards the numbers decreased gradually until by earlyJune all the
males had gone. Fifty-one different males were altogether observed in Horseshoe Pond
and 33 in the quarry ponds. Thirty-four males (67%) stayed at the shore of Horseshoe
Pond for a longer time and were observed four times or more. Some of them were found
throughout most of the study period. Others apparently were around for much shorter
periods and were found only once, or a few times (table 1 Towards the end of the study
period, males not observed before occasionally arrived at the pool while others left,
creating a certain turnover in the population. Fewer females than males were observed
both at Horseshoe Pond (n=10) and in the quarry (n=9). In contrast to the pattern
observed for the males, relatively few females at Horseshoe Pond were seen four times or
135
Figure 2. Daily
observed number
of males,
females and newly deposited egg-clutches of Pelodytes punctatus
in
northwestern France in 1981 in a) the Quarry Ponds and b) Horseshoe Pond. Shading
indicates
periods in
which no observations were made. Periods
of rising
air temperature
(trend
of increase of median values
per 24
hours)
are shown
in the lower
figure by hatched bars.
136
Table 1. Number of times
individual
Pelodytes
punctatus
were observed at Horseshoe
Pond. For males,
the
total number observed is shown to approximate
population size (details
see text).
more. Compared to the males, females stayed at the pond shore line for a short time
being observed fewer times (table 1). ).
In 1981 spawning took place in two periods, the first period being from the end of
March to early April and the second period being in early May (fig. 2). A total of 46
clutches was found in Horseshoe Pond and 15 clutches were found in the quarry. In
making up 46 clutches the first period of egg-deposition was more important than the
second one, when only 12 clutches were deposited. Three more clutches were found
mid-April, in between the first and the second periods of spawning.
Similar temporal patterns of presence and abundance were found in 1982 and 1983,
with the highest numbers of adults present in the second half of April (in 1982) or the
first half of May (in 1983) (fig. 3). In 1982 most egg clutches (28 out of 35, 80%) were
deposited in the first half of April and in 1983 in the second half of March (23 out of 34,
68%). Five egg clutches were observed in May 1982 and none in May 1983. In each of
the study years egg deposition was preceded by a period of rising air temperatures (figs.
2 and 3).
Egg deposition
Apart from a few egg-strings that were not attached to any substratum, or that were torn
loose by the action of wind and waves, all the clutches were found deposited against
plant material in a zig-zag manner typical for Pelodytes punctatus (Toxopeus, 1985).
Submerged, short stiff branches from Phragmites australis and Cirsium arvense were most
often used for this purpose. Approximately nine out of ten clutches were deposited on
submerged vegetation that did not reach the water surface and about three out of four
clutches were laid closer to the bottom of the pool than to the surface. In 1981, 30 out of
44 (68%) of the clutches that were observed on the spot were deposited in the deep part
of the pond, 27% were deposited in areas of intermediate depth and 5% were deposited
137
Figure 3. Daily
observed
numbers of males,
females and newly deposited egg-dutches
of Pelodytes punctatus
in
north-western
France in Horseshoe Pond in a) 1982,
and b) 1983.
Legend
as in fig. 2.
138
Table 2. Depth of Horseshoe Pond at the spots where egg clutches of Pelodytes punctatus
were deposited.
V as measured in 1982
in the shallow part of the pond. A similar pattern was found in 1982 (table 2). In 1983
more clutches were deposited at places of intermediate depth than at deep sites, but it
should be noted that in that year the pond was less deep than in the years before and
deep sites were not available.
The yearly average size of the egg strings ranged from 70 mm (SD=26.2 mm, n=36) in
1982 to 81 mm (SD=27.0, n=33) in 1983. These values are not significantly different
from one another. The number of eggs in an average sized string is estimated at 360, but
clutches existed with as few as 100 and as many as 700 eggs. Interestingly, in 1981 the
mean length of the clutches from the first period of spawning was significantly larger
than those from the second period (83 mm versus 67 mm; t-test, p<0.05). No correlation
was found between the pond depth at the spawning site and the size of the clutches (in all
three years the product-momènt correlation coefficient r was <0.2 and the correspond-
ing p value >0.05).
Embryonic development,
larval growth, metamorphosis
and adult size
With the exception of a single moulding egg clutch in 1982, all clutches had developing
eggs and, despite the presence of potential predators like the newt species Triturus
cristatus and T. vulgaris in Horseshoe Pond, the disappearance of eggs or entire clutches
was not observed. (See Isle de Dreneuf, 1862, for a report of predation on a Pelodytes egg-
clutch). The development of the eggs until hatching ranged from a maximum of 19 days
at the start of the season (in 1983) to a minimum of 3 days (observed in May 1981). A
strong negative relationship was observed between water temperature and age at
hatching. At a median temperature of 90 C embryonic development till hatching took
10-14 days; at 170 C this was reduced to 4 days (fig. 4). The first free-swimming larvae
were observed the earliest on April 6th (in 1983) and at the latest on April 16 (in 1981). ).
Spawning was not observed in 1975 but is likely to have peaked in early April when
temperature rose sharply after a three week long cold spell with average median air
temperatures below 50 C (Arntzen and Gerats, 1976). By the end of May the larvae from
these clutches had reached an average SVL of 16.1 mm. The small variance in size of
this cohort (SD=1.2 mm, n=50) indicates that the tadpoles were of similar age. A further
two weeks of rapid growth made the sample reach maximum mean body size of 24.6 mm
139
Figure 4. Duration of embryonic development
as a function of water temperature (median
value per 24
hours)
for seven selected
egg-strings
of Pelodytes punctatus
in Horseshoe Pond. The curve is fitted by a third
order polynomial.
(fig. 5). By then hind limbs had appeared in all tadpoles. Fore-limbs were visible behind
the transparent skin by June 19 and in the sample taken nine days later fore-limbs had
appeared in all but one of the specimens. We assume that most larvae had meta-
morphosed at July 8 because only a single (metamorphosing) larvae could be obtained at
that date.
In the cage that was placed around an egg-string, mean SVL increased from 3 mm for
23 hatching embryo's on April 20, 1982 to 18 mm for tadpoles at June 14. By then hind
limbs had appeared, followed six days later by the fore limbs. Eight larvae (35%) had
completed metamorphosis at July 14 with a SVL of 17 mm (fig. 5). Metamorphosed
froglets in close proximity to the ponds were observed in 1981 from 6 to 27 July (n=5), in
1982 from 15-21 May (n=5) and in 1983 on May 26 (n=14). The number of days passed
since the peak in spawning till the day the first metamorphs were observed ranged from
67 days in 1983, approximately 90 days in 1975 and in 1982, to 102 days in 1981.
Average size of the metamorphs in the biggest sample, that of 1983 (n=14) was in the
range of 15-21 mm. By the end of August 1983 a sample from this cohort (n=8) had
obtained a size range of 21-28 mm.
Adult males as measured in 1983 were on average significantly smaller than females
(SVL of 37.5±3.02 mm, n=24 versus 40.6±3.56 mm, n=11; p<0.05 in t-test).
Adult population size and spatial distribution
Altogether 33 different males, 9 females and 15 egg clutches were observed in and
around the quarry ponds in 1981. The number of males observed around Horseshoe
140
Figure 5. Growth
of tadpoles of Pelodytes punctatus
in north-western France.
Average
snout - vent
length (SVL)
is indicated
by squares,
total length by dots and standard deviation
by vertical
bars. Numbers
refer
to sample
size. Data points connected
by interrupted lines are approximated;
details see text. The graph connecting
solid
symbols represents
larval
growth
in a cage
(data from
den Uijl, 1982;
details
see text and fig.
6).
To the
right average
SVL and standard deviation for adult males
(m)
and females
(g, as measured in 1983.
Pond in that year was 51 and 24 were observed in 1983. The vast majority were
recorded more than once, with a maximum of 56 records over 78 observation days and
low values were obtained for the index of daily census oversight (0.35 in 1981 and 0.24
in 1983). It is therefore unlikely that many male frogs will have been overlooked. Even
frogs using a single shelter for a single day are not likely to have escaped our attention,
and we conclude that Nm closely reflects reality. Comparing the number of egg-strings
in Horseshoe Pond with the observed number of females under the shelters it is
concluded that many females have gone unnoticed, i.e. they do not seem to use the
shelters to the same extent as the males. The female population size, estimated from a
count of egg-clutches, was 46 in 1981 and went down to 35 in 1982 and to 34 in 1983.
Assuming that each female deposits only once a year, the female population size in 1981
1
was 46 individuals. The estimated sex-ratio expressed as 1Vm over l%r (5 1 /46
in 1981 and
24/34 in 1983), in neither year is significantly different from unity (G-test for goodness
141
of fit). The total estimated population size for the adults declined from 97 in 1981 to 58
in 1983. On the assumption that sex-ratio was at unity in 1982 Na can be estimated from
Ñf to give a value of 70 for that year.
Out of 51 males marked around Horseshoe Pond in 1981, nine ( 18%) were recaptured
in later years. Interestingly, six more recaptures were found that were originally marked
in the quarry at 300 m distance (6 out of 33, 18%). Only one female was ever recaptured
between years. The parsley frog seems to swap breeding site easily from year to year
between the quarry and Horseshoe Ponds. Unfortunately, this behaviour with the
adults moving away from their original breeding site to take up residency at another site
precludes the estimation of yearly survival rates.
A different pattern is observed during the breeding season when the males are
remarkably sedentary. They were found under the same shelter up to 38 times in a row,
with no gaps in the recording. Other males occasionally moved to another shelter, but
the pattern as a whole is that of sedentariness. In 1981 most males had taken up
residency on the spit of land (fig. 6 spot C), with another aggregation at the opposite
corner of the pond (fig. 6 spot E), spanning that part of the pond where most egg-strings
were deposited. Similar patterns of clumped male frog distribution were found in 1982
and 1983
(fig. 7). Visual inspection of figs 6 and 7 reveals a tendency for the females not
choosing the shelters where male frogs were residing. In 1981 and 1983 most females
were found close to, but outside the areas were males aggregated. Choruses of males
displaying their mating call in the evenings (Hotz, 1971; Paillette et al., 1992), were
observed near the centre of the pond. Most males were calling from above the water
surface, although some called from a submerged position.
Discussion
The parsley frog, Pelodytes punctatus, is among the least known species of the Palaearctic
herpetofauna. For just over a century now it has been known to occur in Pas-de-Calais,
close to the northern edge of its range. However, the existence of a relatively large
(Ña=60-100) and a successfully reproducing population north of Boulogne-sur-Mer
is only documented for Horseshoe Pond. In estimating Na several assumptions are
made. For the males it is assumed that none of them go unnoticed, or - in other words -
that all, at some time or other, use the shelters provided for a period of time long enough
(approximately 24 hours) to be observed at at least one of the daily checking routines
during the breeding season.
As judged from the comparison of the number of females and the number of egg-
strings, females, in contrast to the males, do not seem to be using the shelters to an
extent that N fwould be reliably estimated. However, the clear pond does allow a correct
count of all egg-strings. Some reports indicate that females deposit several clutches in a
single mating. Studying the breeding of the parsley frog in captivity, Raehmel (1983),
Toxopeus (1985) and Hartley (1990) all observed 2 or 3 clutches laid by single females,
with a time span between deposits of 15 minutes in an interrupted amplexus with spatial
142
Figure 6. Map of Horseshoe Pond with the egg clutches of Pelodytes
punctatus
indicated as found in 1981
(round symbols, n=44), 1982
(square symbols, n=35)
and 1983
(triangle
symbols, n=33). Position
of cage
in
which larvae were raised is indicated
by an asterisk. Isobaths are given
for pond depths
of 15 and 25 cm, as
measured in 1982.
Areas
along
the shore line are schematically
indicated where
many
males
(dotted
shading),
and females
(grey shading)
were found
in 1981. For the spatial
distribution of the adults in 1982
and 1983 see
fig. 7.
Figure 7. Distribution of adult
Pelodytes punctatus
around Horseshoe Pond in 1981, 1982,
and 1983. Number
of male observed at shelters is indicated above the horizontal
axis;
females shown below the horizontal axis
(with
the exception
of 1982
data when data for males and females were pooled).
Note that the total surface
covered
by the graphs
varies over the years not only
as a function of the number of animals
found,
but also
with the number of artificial
hiding
sides
(28
in 1981,
36 in 1982
and 41 in 1983).
Letters A to E indicate the
corners of the pond as shown
in fig. 6.
143
relocation (Hartley, 1990) to two hours in a continued amplexus (Toxopeus, 1985). In
another study (van den Elzen, 1976) it is implicitly stated that generally more than 10 0
clutches are deposited per female! If in the wild indeed more than one clutch is
deposited per female, as seems to be the case in captivity, this would mean that Nfis
overestimated. This could be by as much as a factor 1.7 in the quarry and a factor of 3.0
- 4.4 in Horseshoe Pond (cf. tables I and 2). Circumstantial evidence counting against
this interpretation is that the clutches are spaced out all over the deeper parts of the
pond (fig. 6) and that no indications were found for two or more of the shorter egg-
strings to be deposited in close proximity to each other. We suggest that several short
clutches instead of a single larger one are normally laid only in shallow water where not
enough vertical space is available for the female to climb the vegetation during egg-
deposition. (See the description of the breeding behaviour of the frog by Toxopeus,
1985). Observations supporting this view were made in two ponds in very close prox-
imity to each other but differing in depth, in a sand-quarry near Montelimar in
southeastern France (J. W. Arntzen and A. Zuiderwijk, 1983 unpubl.). It was found that,
the freshly deposited egg-strings found in the shallower of the two ponds (8 cm versus 18 8
cm water depth) were significantly shorter than those found in the deeper pond (52.2 ±
9.6 mm, n=ll, versus 74.7±24.5 mm, n=15; p<0.01, t-test). For the relatively deep
Horseshoe Pond we therefore favour the interpretation in which each egg-clutch repre-
sents an adult female. A corollary of the alternative explanation in which females each
year deposit several clutches, is a male biased sex-ratio that at Horseshoe Pond would
significantly differ from unity (p<0.05; G-test for goodness of fit).
The artificial shelters are intensively used by the males. The fact that during the
breeding season the males congregate in groups in and around the pond seems to
indicate the existence of a lek breeding system (cf. Bourne, 1992). Interestingly, indica-
tions are found that females assemble at places that are satellite to these male groups
(figs 6 and 7), and we speculate that they enter the lek when ready for breeding, acquire
a mate and enter the pond in copula. It may be that the observed preferences for the
deepest parts of the pond to mate, and for non-emerging aquatic vegetation to use as a
substratum for egg-deposition (cf. H6ron-Royer, 1879) are interdependent. The sugges-
tion that the parsley frog prefers the shallowest parts of the ponds to spawn (von
Bedriaga, 1891), where vegetation is densest (H6ron-Royer, 1879), cannot be
confirmed.
Compared to most other Palaearctic anurans the embryonic development of the
parsley frog is rapid and larval growth is fast (Diaz-Paniagua, 1988). Embryo's are
reported to hatch in 8.5 to 10 days (von Bedriaga, 1891; Raehmel, 1983; Diaz-Paniagua
and Arizabalaga, 1987) and to complete metamorphosis in a further 73 to 97 days
(Balcells, 1955; Raehmel, 1983; cf. Girard, 1989), all in line with our observations. In
hibernating tadpoles it may take up to 8 months to reach metamorphosis (Lataste, 1876;
van den Elzen, 1976). In southern France tadpoles are reported to metamorphose at a
SVL of apto 25 mm in 2-3 months after hatching (von Bedriaga, 1891). In a study of 5
species of anurans in southern Spain, the fast growth of Pelodytes punctatus
tadpoles is only
144
matched by that of Pelobates cultripes (Diaz-Paniagua, 1988), reaching a large size at
metamorphosis (up to 23 mm) in approximately 70 days.
The growth curve obtained for free-swimming tadpoles of the parsley frog at the
northern edge of its range is surprisingly similar to that at the very south of its range
(data from Diaz-Paniagua and Arrizabalaga, 1987; Diaz-Paniagua, 1988, 1989 and
pers. comm., 1992). In this context, two observations are relevant. Firstly, in the south
of Spain, the frog breeds in November and the larvae develop during the coldest months
of the year (Diaz-Paniagua, 1988), while in the north the larval period falls in spring and
early summer. Secondly, in the north most successfully reproducing populations of the
parsley frog are found in the warmest environments. Applying a method in which the
(temperature dependant) inversion of sucrose into glucose was used to determine
additively the water temperature of ponds, Kleyn and Smit (1986) demonstrated that in
spring and early summer the dune ponds in the same study area (some of them with
parley frogs) were significantly warmer than similar sized ponds outside the dunes. The
temperature difference averaged at 3.50 C. We have shown that ambient temperature
strongly influences developmental rates in the parsley frog. It may well be that effec-
tively the temperature regimes in ponds at opposite sites of the N-S distribution during
larval development are altogether not markedly different.
Algae and detritus constitute the bulk of the diet of tadpoles of the parsley frog (Diaz-
Paniagua, 1989). They possess an efficient filter apparatus and a long digestive tract that
enables them to successfully exploit the algal blooms (Sokol, 1981; Diaz-Paniagua,
1989) that develop in early spring in the otherwise oligotrophic dune slacks, acquiring a
large size at metamorphosis in a relatively short larval period.
With several months available for terrestrial feeding, metamorphs may be able to
reach adult size (28 mm) before they go into hibernation. Indeed, some specimens were
recorded reaching adult size as early as the end of August in the year that they were
born. That small animals may be sexually mature was shown by Hartley's (1990)
observation of a successfully reproducing female of only 25 mm. Considering the fast
larval and juvenile growth as documented in the present study, we assume that in
northern France the parsley frog normally breeds for the first time at the beginning of its
second year. Compared to other Palaearctic anurans the juvenile phase of the parsley
frog is unusually short.
Several hypothesis can be put forward that would explain the existence of a second
peak of breeding activity as observed in 1981 and in 1982. The second peak of breeding
could be caused by: i) some particularly successful females producing not one but two
clutches a year (cf. Wells, 1976); ii) new recruits to the breeding population that had not
yet maturated earlier that year (cf. T J.C. Beebee, in: Silverin and Andrén, 1992); or iii)
by females that postponed oviposition, in response to environmental conditions such as
drought (cf. Silverin and Andrén, 1992). Our observation that clutches laid by Pelodytes
punctatus in the second period are smaller than those laid in the first period is not in line
with the latter scenario; it does, however, not allow to discriminate between the other
two explanations. Smaller clutches may be expected to be produced by new breeders
145
because recruits will be comparatively small and female fecncity may be related to body
size. Alternatively, resources might be insufficient to produce a second clutch that is
equally large to the first one, considering the short time span available (cf. Howard,
1978). To complicate matters further, some authors report autumnal breeding of
Pelodytes punctatus in Italy and southern France (Von Bedriaga, 1891; Sindaco and
Andreone, 1988), but such has not been observed in the d6partement Pas-de-Calais.
Despite extensive inventories in Pas-de-Calais, the Parsley Frog is rarely found
outside the dunes (Zuiderwijk and Hooghiemstra, 1975; Arntzen and Gerats, 1976;
Kroese and Van Leeuwen, 1979), and a similar distribution pattern is found in the
d6partement Somme, to the south of Pas-de-Calais (Godet, 1990). Outside the dune
areas the species has been found in an abandoned marble quarry and in two cattle
drinking troughs in pasture land. Outside the dunes the presence / absence of Pelodytes
punctatus in Pas-de-Calais is not clearly associated to any environmental variable, as
seems to be the case for western Europe on the whole; generally, its presence is localized
and most difficult to predict (A. Zuiderwijk, pers. comm., 1983; Paillette et al., 1992).
We have shown the adult frogs to be quite migratory and we assume that in this way new
populations are occasionally founded. On the whole however opportunities for success-
ful reproduction will be rare and many populations may not survive. Such a 'source -
sink' pattern of population structure fits the distributional observations very well. In
Pas-de-Calais the main sources of dispersal are found in the dunes. These sandy habitats
provide easy shelter for the adults as well as relatively warm and oligotrophic aquatic
conditions for the tadpoles. The ultimate factor governing successful reproduction in the
dunes may well be annual rainfall. Occasionally ponds were observed to dry out well
before metamorphosis while in other years, with less than average levels of precipitation,
many ponds ceased to exist altogether. In such conditions a flexible mode of life with the
potential to quickly invade new ponds when they come into existence is a requirement to
survive.
Acknowledgements. We thank Dirk van der Kroef for assistance in the field, Dr. Mario Garcia-Paris
and Dr. P. Galin for their help in the construction
of fig.
1,
Carlos Abrahams for polishing
the english
text,
Mr. P. Sauvage for support and the organization 'Espace Naturel Régional, Littoral Pas-de-Calais' for
access to the study sites.
References
Arntzen, J.W. (1981):
Kikkers en padden (Anura).
In: De amfibieën en reptielen
van Nederland, Belgie
en
Luxemburg, p. 57-107. Sparreboom,
M., Ed., Rotterdam, Balkema.
Arntzen, J.W., Gerats, A.G.M. (1976): Een ecologisch
onderzoek aan de amfibieën van Noord-West
Frankrijk met de nadruk op biotooppreferenties
en predatierelaties. M.Sc. Thesis, Institute of
Taxonomic
Zoologie, University
of Amsterdam,
Amsterdam.
Arntzen, J.W., Teunis, S.F.M. (1993):
A six year study on the population dynamics
of the crested newt
(Triturus
cristatus) following
the colonization
of a newly
created pond. Herpetol.
J. (in press).
Balcells,
E.R. (1955):
Contributions to the study of the life
cycle
of Spanish amphibians. Br. J. Herpetol. 2:
1-6.
146
Bedriaga, J. von (1891):
Die Lurchfauna Europa's. I. Anura. Froschlurche. Universitäts
Buchdruckerei,
Moskau.
Bea,
A. (1985):
Atlas de los anfibios
y reptiles
de Alava,
Vizcaya
y Guipuzcoa.
In: Atlas de los Vertebrados
Continentales de Alava,
Vizcaya y Guipuzcoa, p. 56-99.
Alvarez, J.,
Castien, E., Mendiola, I., Eds.,
Sociedad de Ciencias
Aranzadi Zientzi Elkartea.
Bosca,
E. (1880):
Catalogue
des Reptiles
et Amphibiens
de la Peninsule
Iberique et des Iles Baleares.
Bull.
Soc. Zool. Fr. 5: 240-287.
Bourne, G.R. (1992):
Lekking
behavior
in the Neotropical frog Ololygon
rubra.
Behav. Ecol. Sociobiol. 31:
173-180.
Bruno, S. (1977):
Anfibi
d'Italia: Salientia: II. Pelodytes punctatus
(Studi
sulla fauna erpetologica italiana).
Natura (Milano)
68: 175-180.
Crespo, E.G. (1971):
Anfíbios
de Portugal continental des colecçöes
do Museo
Bocage.
Archos
Mus. Boc.
(2e serie) 3: 203-303.
Creapo, E. G., Oliveira,
M. E. (1989):
Atlas de distribuição
dos anfibios e répteis
de Portugal continental.
Serviço
Nacional de Parques Reservas e Conservação
de Naturaleza, Lisboa.
Diaz-Paniagua,
C. (1988):
Temporal segregation
in larval amphibian
communities in temporary
ponds at
a locality
in SW Spain. Amphibia-Reptilia
9: 15-26.
Diaz-Paniagua,
C. (1989):
Larval diets of two anuran species, Pelodytes punctatus
and Bufo bufo,
in SW
Spain.
Amphibia-Reptilia
10: 71-75.
Diaz-Paniagua,
C., Arrizabalaga,
B. (1977):
Development
and growth
rates of coexisting
Pelobates
cultripes
and Pelodytes punctatus.
In: Proc. 4th OGM Societas Europaea Herpetologica, p. 115-118. Van
Gelder,
J.J., Strijbosch,
H., Bergers, P J.M., Eds., Nijmegen, Faculty of Sciences.
Elzen, P. van der, (1976):
Remarques sur la biologie de Pelodytes punctatus
Daudin 1802 (Amphibia,
Pel-
obatidae)
en Camargue. Rev. Fr. Aquariol. 3: 99-104.
Garff, B., Le (1989): Pelodytes punctatus
(Daudin, 1802)
Pélododyte ponctué, ou "Grenouille" persillée.
In:
Atlas de Repartition des Amphibiens
et Reptiles de France, p. 70-71.
Castanet, J., Guyetant, R.,
Eds., Paris, Société Herpétologique
de France.
Giard, A. (1890):
La laboratoire de Wimereux en 1889,
(recherches
fauniques).
Bull. Scientifique
Fr. Belg.
22: 60-87.
Girard, F. (1989): Observations sur l'activité de divers batraciens dans une dune littorale de Loire
Atlantique. Bull. Soc. Herpétol. Fr. 51: 27-32.
Godet, D. (1990):
Données écologiques,
légendes
et traditions populaires
relatives a l'herpétofaune de la
Somme. Bull. Soc. Herpétol. Fr. 56: 27-37.
Hartley, S. (1990):
Spawning
and rearing the parsly frog (Pelodytes punctatus)
in captivity. Br. Herp. Soc.
Bull. 31: 19-25.
Héron-Royer (1879):
Note sur l'oeuf et la première période embryonnaire
du Pélodyte ponctué Pelodytes
punctatus
Dugès. Bull. Soc. Zool. Fr. 4: 229-239.
Hoffmann, J.
(1958):
Sur la présence
de Pelodytes punctatus
Daudin au Grand-Duché
de Luxembourg.
Arch.
Inst. Grand-Ducal de Luxembourg
N.S. 25: 235-237.
Hotz, H. (1971):
Analyse
des Paarungsrufes
von Pelodytes punctatus (Daudin 1803)
(Amphibia, Pelobatidae).
Experientia 27: 851-852.
Howard, R.D. (1978):
The influence of male-defended
oviposition
sites on early embryo mortality in
bullfrogs.
Ecology
59: 789-798.
Isle du Dréneuf, A. de 1' (1862):
Un nouveau batracien urodèle de France (Triton blasii).
Ann. Sci. Nat.
(Paris)
17: 363-375.
Kleijn, J.,
Smit, E. (1986):
Oecologisch
onderzoek naar Triturus helveticus
en Triturus
vulgaris
in Noord-west
Frankrijk. M.Sc. thesis, Institute of Taxonomic Zoologie,
University
of Amsterdam, Amsterdam.
Kroese J., Leeuwen,
F. van (1979):
Amfibieën
rond Ambleteuse.
Verslagen
en Technische Gegevens
19,
Institute of Taxonomic Zoologie,
University
of Amsterdam,
Amsterdam.
Lataste, F. (1876): Essai d'une faune herpétologique de la Gironde. Actes Soc. Linn. Bordeaux 30:
193-544.
Lopez-Jurado,
L.F., Ruiz, M., Santaella, R. (1980):
Características de la distribución de los anfiguos
anuros y urodelos en la Provincia
de Cordoba (España).
Actas. IIa Reunión Iberoamer. Cons. Zool.
Vert. 1: 15-25.
Martinez Rica, J.P. (1983):
Atlas herpetológico
del Pirineo. Munibe 35: 51-80.
147
Paillette, M., Oliveira, M.E. Rosa, H.D. Crespo, E.G. (1992):
Is there a dialect in Pelodytes punctatus
from
southern Portugal? Amphibia-Reptilia
13: 97-108.
Parent, G.H. (1970):
Le Pélodyte
ponctué, Pelodytes punctatus
(Daudin)
exist et-il en Belgique
et au Grand-
Duché de Luxembourg?
Note préliminaire. Naturalistes
Belges
51: 333-337.
Parent, G.H. (1984):
Atlas des batraciens et reptiles de Belgique.
Cahiers d'Ethologie Appliquée
4: i-iv,
1-198.
Raehmel,
C.-A. (1983):
Pelodytes punctatus (Daudin),
der Westliche Schlammtaucher -
seine
Pflege
und seine
Zucht. Herpetofauna 27: 22-24.
Regulez Fernández, O. (1988): Primeras citas de Pelodytes punctatus (Daudin, 1802) y Podarcis muralis
(Laurenti, 1768)
en Galicia. Rev. Esp. Herpetologia
3: 145-147.
Schreitmüller, W., Wolterstorff,
W. (1923):
Beiträge
zur Fauna Nord- und Nordost-Frankreichs und der
angrenzenden Gebiete Belgiens.
Arch. Naturg. (Berlin)
89A: 119-168.
Silverin, B., Andrén, C. (1992):
The ovarian cycle
in the natterjack toad, Bufo
calamita,
and its relation to
breeding behaviour. Amphibia-Reptilia
13: 177-192.
Sindaco, R., Andreone, F. (1988):
Considerazioni sulla distribuzione di "Pelodytes
punctatus" (Daudin,
1802)
in territorio italiano (Amphibia,
Anura,
Pelodytidae).
Atti Mus. Civ. Stor. nat. Trieste 41: 161-167.
Sokol,
O.M. (1981):
The filter apparatus of larval
Pelodytes punctatus
(Amphibia:
Anura).
Amphibia-Reptilia
2: 195-208.
Toxopeus, A.G. (1985):
Die Eiablage
bei Pelodytes punctatus (Daudin, 1802) (Salientia: Pelobatidae).
Sala-
mandra 21: 222-224.
Uijl, H. den (1982): Pelodytes punctatus,
een zwerver,
aan de rand van zijn bestaan.
M.Sc.
Thesis,
Institute of
Taxonomic
Zoology,
University
of Amsterdam,
Amsterdam.
Vieira, L. (1887):
Catálogo dos amphibios e répteis de Portugal existentes actualmente no Museu Zoo-
lógico
da Universidade de Coimbra. Relatório do Prof. de Zool. 1885-1886,
Coimbra.
Wells,
K.D. (1976):
Multiple egg clutches in the green frog (Rana clamitans)
Herpetologica
32: 85-87.
Zuiderwijk,
A., Hooghiemstra,
H. (1975): Enige oecologische waarnemingen
aan de amphibieen in de
Boulonnais.
M.Sc. Thesis, Institute of Taxonomic
Zoology, University
of Amsterdam,
Amsterdam.
