Download the full book of proceedings - SEH

Biota
Vol. 3, No. 1-2, 2002
Revija za biologijo in ekologijo
Journal of biology and ecology
Proceedings of the
11th Ordinary General Meeting of Societas
Euro pea Herpetologica (SEH),
Zalec, Slovenija, July 13-17, 2001
Drustvo za proucevanje ptic in varstvo narave
Society of Bird Research and Nature Protection
Drustvo varuhov okolja Radoziv
Environmental Society Radoziv

Biota 3 1-2,2002
11th Ordinary General Meeting of Societas
Europea Herpetologica (SEN),
Zalec, Slovenija, July 13-17, 2001
Organised by:
Environmental Society Radoziv, Zalec
Organising committee:
Klavdija Kac
Miran Orozim
Stasa Tome
Milan Vogrin
Gregor Vovk Petrovski
Tanja Vovk Petrovski
Edited by:
Milan Vogrin
Linguistic collaboration:
Victor Kennedy
Scientific committee:
Dr. Wolfgang Bohme (Germany)
Dr. Zoltan Korsos (Hungary)
Dr. Michael R.K. Lambert (United Kingdom)
Dr. Claud Miaud (France)
Msc. Stasa Tome (Slovenia)
Dr. Marco Zuffi (Italy)
The recommended citation for particular paper is e.g.:
Crnobrnja-lsailovic, J. 2003: Notes of diurnal activity in Vipera ammodytesof the central
Balkans. In: Vogrin, M. (ed.). Proceedings of the 11th Ordinary General Meeting of
Societas Europea Herpetologica (SEH), Zalec, Slovenia, July 13-17, 2001. Biota 3 (1-2):
The papers in this issue are ordered according they alphabetical order of authors.

Vsebina/Contents
3 1-2, 2OO2
Milan VOGRIN
Preface 7
Clanki/Articles
Jelka CRNOBRNJA-ISAILOVIC
Notes of diurnal activity in Vipera ammodytes of the Central Balkans 9
Andris CEIRANS
Reptiles and amphibians of the Gauja National Park, Latvia 17
Augusto GENTILLI, Stefano SCALI, Francesco BARBIERI & Franco BERNINI
A three-year project for the management and the conservation of amphibians
in Northern Italy 27
Giinter GOLLMANN, Birgit GOLLMANN, Christian BAUMGARTNER & Andrea
WARINGER-LOSCHENKOHL
Spawning site shifts by Rana dalmatina and Rana temporaria in response
to habitat change 35
Kurt GROSSENBACHER
First results of a 20-year-study on Common Toad Bufo bufo in the Swiss Alps 43
Daniela GUICKING, Ulrich JOGER, Michael WINK
Molecular Phylogeography of the Viperine Snake Natrix maura and
the Dice Snake Natrix tessellata: first results 49
Adrian HAILEY & Michael R.K. LAMBERT
Comparative growth patterns in Afrotropical giant tortoises (genus Geochelone) 61
Miodrag JOVANOVIC, Dragana BURIC & Zoran MARKOVIC
Tertiary reptiles of the central part of the Balkan peninsula 67
Zoltan KORSOS & Balazs TROCSANYI
Herpetofauna of Round Island, Mauritius 77
Bilal KUTRUP & Nurhayat YILMAZ
Preliminary data on some new specimens of Vipera barani collected from Trabzon
(Northeastern Turkey) 85
A.C.AA. MEESKE, N. SCHNEEWEISS & K.J. RYBCZYNSKI
Reproduction of the European Pond Turtle Emys orbicularis
in the northern limit of the species range 91
Michele MENEGON & Sebastiano SALVIDIO
Notes on habitat, egg-laying and first record of the advertising
call of Hyperolius kihangensis 703
Zoltan Tamas NAGY, Ulrich JOGER, Daniela GUICKING & Michael WINK
Phylogeography of the European Whip Snake Coluber (Hierophis)
viridiflavus as inferred from nucleotide sequences of the mitochondrial
cytochrome b gene and ISSR genomic fingerprinting 109
Christian PASTORELLI, Paolo LAGHI & Dino SCARAVELLI
Seasonal activity and spatial distribution of a Speleomantes italicus
population in a natural cave 119
Christian PASTORELLI, Paolo LAGHI & Dino SCARAVELLI
Speleomantes antipredator strategies: a review and new observations 127
Galina V. POLYNOVA & Olga E. POLYNOVA
Tail autotomy as an index of human influence on the
Alsophylax pipiens population in the Bogdino-Baskunchak state reserve 133

Biota 3/i-a, 2002
Laura RACCA
The conservation of the Agile Frog Rana dalmatina in Jersey (Channel Islands) 141
Sebastiano SALVIDIO, Gabriele ALARIO, Maura Valeric PASTORINO & Mirko FERRETTI
Seasonal activity and abundance of Speleomantes ambrosii in cave habitats 149
Stefano SCALI & Augusto GENTILLI
A comparison of main heathlands in northern Italy and their importance
for amphibian populations 155
Stefano SCALI, Claudia CORTI, Augusto GENTILLI, Luca LUISELLI,
Edoardo RAZZETTI & Marco A.L. ZUFFI
Continental versus Mediterranean European Whip Snake
Hierophis viridiflavus: a morphometric approach 161
Galina. S. SUROVA
The role of frog egg aggregations as a control of abiotic factors 167
Judit VOROS, Zoltan KORS6S & Ferenc SZALAY
A comparative morphological study of the two Hungarian
discoglossid toad species Bombina spp 173
Vit ZAVADIL & Arnost L. SIZLING
Morphological variability in the newts of the Cristatus group 181
Marco A.L. ZUFFI, Augusto GENTILLI, Edoardo RAZZETTI & Stefano SCALI
Transition-hybridization areas in parapatric species
of Vipera aspis group from northern Italy 191
Nove knjige/Book reviews
Milan VOGRIN
CABELA, A., GRILLITSCH, H. & TIEDEMANN (eds.) 2001:
Atlas zur Verbreitung und Okologie der Amphibien und Reptilien in Osterreich 197
Editor acknowledgements 198

Preface
)ta 3/1-2, 2OO2
The third volume of Biota is entirely devoted to herpetology. This volume includes 24 contributions
presented during the 11th Ordinary General Meeting (OGM) of SEH - Societas
Europaea Herpetologica.
The OGM in Zalec (Slovenia) was attended by 86 participants from 24 countries who contributed
37 oral presentations and 56 posters. Some of them, as full papers, are reproduced
in this volume.
This volume goes to press much later than we expected. This is largely because of the sometimes
extensive discussions we had (the referees and I) with the authors on how to make
their work more comprehensible. Despite the delay, I hope that our readers (and authors as
well) will find these efforts worthwhile.
I am also grateful to all who helped in different ways, both at the congress and during the
preparation of the proceedings. Here I wish to thank the past president of SEH, prof. dr.
Wolfgang Bohme, and the past and former secretary of SEH, dr. Michael R.K. Lambert, for
their encouragement and work during the congress.
Finally, I wish to mention that for the first time at the OGM of SEH, we started with a competition
for "best student talk" and "best student poster". Congratulations Daniela and
Nathalie! Daniela's paper can be found in these proceedings.
Thanks also to all referees (see the list of referees) for their time and help and to all participants
with whom I have (almost with all) very pleasant discussions.
-»»
I hope that we will see you some day in Slovenia again!
Milan Vogrin

CRNOBRNJA - ISAILOVIC BJOta 3/i-a. 2002 9
Notes of diurnal activity in Vipera
ammodytes of the Central Balkans
Jelka CRNOBRNJA-ISAILOVIC
Department of evolutionary biology, Institute for biological research,
29. Novembra 142, 11000 Belgrade, FR Yugoslavia
E-mail: jelka@ibiss.bg.ac.yu
Abstract
Diurnal activity in Vipera ammodytes, the most widespread viper in the Balkans, was
analysed using a sample of records collected from 1981 to 2001 in Serbia and
Montenegro. The inspiration for this study arose from several questions summarized as
one: is it possible to successfully define activity of specimens in space and time by a
combination of states of several environmental variables? A restricted number of measured
environmental variables, as well as spatial/temporal heterogeneity of records,
influenced the course of the study. The results pointed to significantly different diurnal
activity patterns of the sexes in two seasons (spring and summer; p < 0.05). Distribution
of male and female records in two seasons and three main parts of the day was also
non-random. The observed differences between male and female diurnal activity patterns
would be the consequence of their reproductive activity schedules during two seasons.
Differences in diurnal activity sensu sTricto seem to be related to insolation regimes
of two seasons.
Key words: Vipera ammodytes, Central Balkans, diurnal activity
Received 1 October 2001; accepted 4 November 2001

10 Biota 3 i-:>., ?.oo:>. CRNOBRNJA - ISAILOVIC
INTRODUCTION
The Balkan Peninsula comprises most of
the European part of the sand viper's
species area (see in: Crnobrnja-lsailovic &
Haxhiu 1997). In Serbia, this viper was
not recorded north of the Sava and the
Danube river flows. Vipera ammodytes
(Figure 1) is a common inhabitant of
canyons and gorges, as well as of areas
belonging to the vegetation zone of
open, mostly oak xerophilous forests in
Serbia and Montenegro. Syntopic reptile
species in the continental part of the
Central Balkans frequently include:
Podarcis mural/s, Lacerta viridis,
Corone/la austriaca, Testudo hermanni,
and in particular habitats also Lacerta
pratico/a pontica, Ab/epharus kitaibelli
and Coluber caspius.
Figure 1. Vipera ammodytes, adult female,
from Southeastern Serbia (photo: J.
Crnobrnja-lsailovic).
Knowledge of the sand viper population
dynamic and activity patterns in Serbia
and Montenegro still does not go beyond
the level of anecdotal observations. My
intention in this study was to arrange diffuse
data into statistically useful information
and to provoke more detailed studies
in the future, comparable to those
already done for related species (Money
1994, Zuffi 1999, Ujvari & Korsos 2000).
MATERIAL AND METHODS
Vipera ammodytes was recorded during
the author's inventory of amphibian and
reptile species in Serbia and Montenegro,
from 1981 to today (Figure 2). The analyzed
area broadly covers the geographic
space between 42° and 44°40' North latitude.
The specimens were detected within
a range of altitudes 10 - 1600m above sea
level. The quality of environmental data
accompanying the records varied from
case to case, depending on the accessibility
of equipment. The terrain was checked
visually during the day time by slowly
walking through various habitats along the
planned transect route. Specimens were
captured and sexed where possible. Some
sand viper habitats were checked from
midnight to 02h A.M., but nocturnal activity
of the species was not detected.
In total, 67 records were accompanied
with data concerning at least one environmental
variable (Table 1). Individuals were
recorded during spring (1 in March, 4 in
April, 14 in May and 2 in the first half of
June) and summer (1 in the third week of
June, 43 in July and 2 in August). Among
them, only mature individuals (41) were
chosen for analysis. Juveniles were omitted
from the study because overall sample size
was too small for such a structured analysis.
The sex of captured animals was recognized
by presence/absence of
hemipenises. It is worth mentioning that
15 adults were measured before being
released. Their overall size varied from 404
to 670 mm. The records were categorized
according to accompanying variables and
processed through simple (two variables)
and multivariate (more than two variables)
correspondence analyses (Statistica 5.0).
RESULTS
The results of ordinary correspondence
analysis (Table 2.) point to different activity
patterns of the two sexes: males were
significantly more "visible" during the

CRNOBRNJA - ISAILOVIC Biota 3 1-2,2002 11
Figure 2. A. European part of V. ammodytes species area. B. Location of records used in this
analysis.

12 Biota 3/i-a, 2002 CRNOBRNJA - ISAILOVIC
Table 1. Variables processed through various correspondence analyses.
Variable
Sex
Season
Part of the day
Exposure
Behaviour
spring, while females were more
"exposed" during the summer (Table 3).
Differences in activity patterns between
sexes in relation to other environmental
variables were not detected. Also, the
results suggest that the activity time of
adult sand vipers generally differs
between two seasons (p < 0.05): during
the spring, the sand vipers were usually
seen in the middle of the day, and during
the summer, mostly in the morning hours
(Table 3). The results of multiple correspondence
analysis pointed to significant
association of the different sexes with
two seasons (spring and summer) and
three main parts of the day (Table 2):
males were more often noticed in the
spring middays and females in summer
mornings (Figure 2). Analysis, which
included sets of four as well as all five
Category
Male
Female
Spring
(21stMarch-23rdJune)
Summer
(23rd June-20111 September)
Morning
(07A.M. - 1 1 A.M.)
Midday
(11A.M.-04P.M.)
Afternoon
(04PM. - 08 P.M.)
South
Southeast
East
Northeast
North
West
Taking res tin shade
Basking
Moving
Hiding
Escaping
Mating
N° of cases
16
19
21
46
14
20
7
10
8
12
1
4
3
9
8
4
2
11
2
recorded variables, also showed significant
non-random associations (Table 2).
However, it is important to mention that
no statistical significance tests are customarily
applied to the results of a correspondence
analysis; the primary purpose
of the technique is to produce a simplified
(low-dimensional) representation of the
information in a large frequency table.
Increasing the number of variables diminishes
discrimination power when the
sample is relatively small (Table 2). For
that reason I avoided interpretation of
results where more than three variables
were tested for associations.
DISCUSSION
Seasonal differences in activity patterns
among sexes were observed in European
populations of several viperide species

CRNOBRNJA - ISAILOVIC BJOta 3/1-2, 2002 13
Table 2. A summary of simple correspondence analysis. Df = degrees of freedom; P = probability;
n.s.=non-significant; *=p < 0.05; ***=p < 0.001
Variables
Sex X Season
Sex X part of the day
Sex X exposition
Sex X behavior
Season X part of the day
Season X exposition
Season X behavior
part of the day X exposition
Part of the day X behavior
Exposition X behavior
SIMPLE CORRESPONDENCE ANALYSIS
X2 test value
5.57
5.35
5.89
6.68
7.98
9.31
5.82
16.05
11.70
26.33
df
1
2
5
5
2
6
5
12
10
20
P
0.0182
0.0689
0.2452
0.2452
0.0185
0.1570
0.3246
0.1890
0.3059
0.1553
significance
A
MULTIPLE CORRESPONDENCE ANALYSIS
Sex X season X part of the day
Total inertia
1.333
Dimensions
4
Total X2
135.345 36
df
P
0.000
significance
***
No. Dim. Singular
value
Eigenvalue %Inertia Cumulative% X2
1
0.751 0.564 42.26 42.26 57.20
2
0.613 0.376 28.21 70.48 38.18
3
0.462 0.214 .^ 16.03 86.50 21.69
Sex X season X part of the day X exposition
Total inertia
2.500
Dimensions
10
Total X2
317.228
df
169
P
0.000
significance
***
No. Dim. Singular
value
Eigenvalue %Inertia Cumulative% X2
1
0.734 0.538 21.53 21.53 68.31
2
0.683 0.467 18.67 40.20 59.22
3
0.574 0.329 13.16 53.36 41.76
Sex x season X part of the day X exposition x behavior
Total inertia
2.600
Dimensions
13
Total X2
401.404 289
Df
P
0.000
significance
***
No. Dim. Singular
value
Eigenvalue %Inertia Cumulative% X2
1
0.739 0.546 21.01 21.01 84.35
2
0.638 0.407 15.67 36.68 62.90
3
0.604 0.365 14.04 50.73 56.38
(recently: Biella & Volkl 1993 for Vipera
berus; Bonnet & Naulleau, 1993, Saint
Girons 1994, Monney 1994 and Zuffi
1999 for Vipera aspis; Baron 1997 for
Vipera ursinii ursinii). Generally, the earlier
emergence of male vipers from hiber-
n.s.
n.s.
n.s.
*
n.s.
n.s.
n.s.
n.s.
n.s.
nation was explained by their need for
completion of spermatogenesis (Nilson
1980). It was also noted that pregnant
females are more easily seen during the
summer than non-reproductive females
or adult males (Andren, 1985). Having in

14 Biota 3 1-2, aooa CRNOBRNJA - ISAILOVIC
Table 3. Percent of records within categories sex and time in two seasons.
Sex
Males
Females
Total
part of the day
Morning
Midday
Afternoon
Total
Spring
75%
25%
100%
8%
64%
28%
100%
Summer
33%
67%
100%
52%
38%
10%
100%
Figure 3. Associations of records having different categories of variables SEX, SEASON and
PART OF THE DAY in the space described by the first two correspondent axes.
E
0
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-1.5
2D Plot of Column Coordinates. Dimension: 1 x 2
Input Table (Rows x Columns). 7x7 (Burt Table)
MALES
MIDDAY
AFTERNOON
-1.0 -0.5 0.0 0.5
mind poor quality of input data in this
study (the absence of consistent longterm
monitoring in the same habitat), I
can only suppose that sand viper males in
the Central Balkans are more active in the
spring searching for territory and for
mates, like their relatives. For V.
ammodytes, such behavior has been sporadically
noticed in the field but still is not
statistically confirmed or rejected by
Dimension 1, Eigenvalue: .56349 (42.26% of Inertia)
1.0 1.5
appropriate experiments. Ritual male
combats are also observed in this species
(Steward 1971).
The fact that most sand viper females
were recorded in the summer logically
points to their reproductive status: breeding
in V. ammodytes in the continental
part of the Central Balkans takes place in
the spring (I have records of courtship at

CRNOBRNJA - ISAILOVIC Biota 3/i-a, 2002 15
the end of April in 1984 in the northern
part of the area studied, but I have also
seen courtship at the end of May 1996 in
the southern part); also, the female
caught at the end of August 1997 at
1200m above sea level had signs of
recent parturition. Accordingly, pregnancy
of sand viper females in the study area
lasts through the summer months.
Parturition probably happens earlier at
lower altitudes - from the beginning of
August in lowland areas to the end of
August/beginning of September in habitats
near the upper altitudinal limit.
Pregnant females need more energy
input than males or non-reproductive
females, so they spend more time basking
(Bozhanskii & Kudryavcev 1986, Andren
1985 for V. berus). I collected most of the
summer data during July (the presumed
pregnancy period) and females predominated
among the specimens observed.
Consequently, one can suppose that most
of the sand viper females observed during
the summer months in the Central
Balkans were pregnant.
Significant differences in time of activity
among males and females of V.
ammodytes in Serbia and Montenegro
seem to be a direct consequence of thermal
differences between the two seasons
in the temperate climatic region. A large
part of the snakes' motion in space is
connected with choosing the optimum
insolation regime (Bozhanskii &
Kudryavcev 1986). The early morning
hours during the spring in this region are
still too cold for heliothermic reptile
species, while summer mornings are more
favorable for daily activities. On the other
hand, midday is the most comfortable
part of the day in the spring, while in the
summer many reptile species usually
spend the warmest part of the day in
their dens.
Acknowledgements
My colleagues Dr. Marco Zuffi (Pisa) and Dr Jens B. Rasmussen (Copenhagen) helped
greatly with their suggestions about literature. I would also like to thank the anonymous
reviewers for constructive comments, and Oliver Isailovic for careful preparation of illustrations.
REFERENCES
ANDREN, C. 1985: Risk of predation in male and female adders, Vipera berus (Linne).
Amphibia-Reptilia 6: 203-206.
BARON, J. - P. 1997: Demographie et dynamique d' une population francaise de Vipera
ursinii ursinii (Bonaparte, 1835). These de Doctorat, Ecole Pratique des Hautes
Etudes, 201 p.
BIELLA, H-J. & VOLKL, W. 1993: Die Biologie der Kreuzotter (Vipera berus, L. 1758) in
Mitteleuropa - ein kurzer Uberblick. Mertensiella 3: 311-318.
BONNET, X. & NAULLEAU, G. 1993. Relations entre la glycemie et I' activite saisonniere
chez Vipera aspis L. Amphibia-Reptilia 14: 295-306.
BOZHANSKII, AT. & KUDRYAVCEV, S.V. 1986: Ecological Observations of the Rare Vipers
of the Caucasus. In: Rocek, Z. (ed.) Studies in Herpetology: 495-498.
CRNOBRNJA-ISAILOVIC, J. & HAXHIU, I. 1997: Vipera ammodytes. In: Case, J..-P,
CABELA, A. , CRNOBRNJA-ISAILOVIC, J., DOLMEN, D., GROSSENBACHER, K.,
HAFFNER, P., LESCURE, J., MARTENS, H., MARTINEZ-RICA, J.P., MAURIN, H.,
OLIVEIRA, M.E., SOFIANIDOU, T.S., VEITH, M. & ZUIDERWIJK, A. (eds.) Atlas
of amphibians and reptiles in Europe. SEH &MNHN (IEGP/SPN), Paris.
MONNEY, J.-C. 1994: Comparaison des cycles annuels d'activite de Vipera aspis et Vipera
berus (Ophidia, Viperidae) dans une station des Prealpes Bernoises (ouest de la

16 Biota 3 1-2,2002 CRNOBRNJA - ISAILOVIC
Suisse). Bull.Soc.Herp. Fr. 71-72: 49-61.
NILSOIM, G. 1980. Male Reproductive Cycle of European Adder, Vipera berus, and its
Relation to Annual Activity Periods. Copeia 4: 729-737.
SAINT-GIRONS, H. 1994: Les risques de predation lies a la reproduction chez un Viperidae
ovovivipare, Vipera aspis L, d' apres les observations visuelles. Amphibia-Reptilia
15:413-416.
STEWARD, J.W. 1971: The snakes of Europe. Fairleigh Dickinson Univ. Press, Rutherford-
Madison-Teancek.
UJVARI, B., KORSOS, Z. & PECHY, T. 2000: Life history, population characteristics and conservation
of the Hungarian meadow viper (Vipera ursinii rakosiensis). Amphibia-
Reptilia 21: 267-278.
ZUFFI, M. A. L. 1999: Activity patterns in a viviparous snake, Vipera aspis (L.), from
Mediterranean central Italy. Amphibia-Reptilia 20: 313-318.

CEIRANS Biota 3/1-2.2002 17
Reptiles and amphibians of the
Gauja National Park, Latvia
Andris CEIRANS
Department of Zoology and Animal Ecology, Faculty of Biology, University of Latvia
Kronvalda bulv. 4, Riga LV-1586, Latvia
E-mail: andrisc@lanet.lv
Abstract
An inventory of the herpetofauna of the Gauja National Park, located in the north-central
part of Latvia, was carried out in 1999-2000. Its objectives were to determine
species composition, status, and habitat preferences. The main attention was focussed
on reptiles. Data were collected along transects located throughout the territory of the
Park. The total length of transects was 166.2 km, and numerous separate observations
of various species were also recorded. Common and widespread species were Lacerta
vivipara, Bufo bufo, Rana temporaria, and Rana synklepton esculenta. Anguis fragilis
was found mostly in a dry pine, pine-spruce forest on the terrace of the ancient valley
of Gauja River. A large population of Matrix natrix was found in the southern part of the
Park in deciduous and coniferous forests. A few populations of Lacerta agiliswere found
in dry pine forests, and on the banks of the Gauja River. Rana atvalis was a rare species,
more frequently found in high moors. There were also several records of Triturus cristatus
and T. vulgaris in the Gauja National Park. The required conservation activities are
discussed.
Key words: reptiles, amphibians, habitats, Gauja National Park, Latvia
Received 6 September 2001; accepted 28 October 2001

ta 3/1-2, 2OO2 CEIRANS
INTRODUCTION
Gauja National Park is located in the
north-central part of Latvia, about 35 km
north-west of RTga. The Park was established
in 1973, and it was then the second
National Park in the territory of the
former USSR. The area of the Park is
91,745 ha; forest occupies 48,592 ha
(Pilats 2000). The area is dissected by the
ancient valley of the Gauja River, together
with valleys of its numerous tributaries
and side ravines. Gauja National Park is
one of the florally richest regions in
Latvia, with a high diversity of forests.
The largest proportion of old broadleaved
forests in Latvia is found near the
town of Sigulda (Pilats 2000). However,
boreal coniferous forests dominate the
Park. The largest high moor is Sudas-
Zviedru mire (2575 ha), found in the
southern part of Gauja National Park.
There are also several smaller mires in the
central and northern parts of the Park.
Agricultural landscapes, although occupying
a large part of the Park, are fragmented
by many forest stands, shrubs,
and fallow lands that serve as refuges for
wild animals. Gauja National Park also
includes many historically significant
objects and popular tourist sites.
The present survey is part of a large-scale
inventory of flora, fauna and habitats of
Gauja National Park carried out in 1999
and 2000. The aim of the inventory of
herpetofauna was to determine the
species composition and status of reptile
and amphibian species, and to describe
their habitats in the Park. The main
attention was focussed on reptiles,
because studies of these animals in
Latvia have been very few and since
information on their preferred habitats is
limited.
METHODS
Data was collected mainly on transects in
the field seasons of 1999 (from 06.08. to
08.09.) and 2000 (from 17.04. to
14.07). The total length of transects was
166.2 km. Censuses were carried out in
all of the main habitat groups of the Park
(Table 1). The type of habitat was determined
from short descriptions made in
the field. This information was supplemented
by data from forest plans of the
State Forest Service. The syntaxonomical
classification of Latvian forests has not
yet been fully developed (Prieditis 1999),
and is therefore supplemented (Table 1)
by the forest stand classification applied
in the forestry industry (Buss 1997). The
transects were evenly distributed
throughout the area. Therefore, the
transect length for a particular habitat is
roughly proportional to the area that the
habitat occupies in the Park. An exception
was agricultural landscapes, which
were considerably less represented in
transects (31.0 % of total transect length
in comparison to 45 % coverage in the
Park).
The routes were carried out in warm and
dry weather, as the main attention was
focussed on reptiles. Each observation of
'a reptile or amphibian was mapped at a
1:50.000 scale. For every reptile specimen
observed, a brief description of the
site was made, and in most cases it was
supplemented later with information
from the data base of the State Forest
Service. Transects were located mostly
along sites with potentially highest reptile
density (roadsides, fringes, cuttings,
clearings etc.), and extrapolation of this
data to the whole habitat could provide
misleading results. Densities were used
only for Anguis fragilis and Lacerta vivipara
for comparing various habitats. To
describe the occurrence of the Common
Lizard Lacerta vivipara in habitats, two
parameters were used: the number of
the individuals and the number of
records. As the movement range for
Lacerta vivipara is up to 80 m
(Zarnolodchikov & Avilova 1989), two
records were considered to be separate if

CEIRANS 3/1-2, 2OO2 19
Table 1. Transect length of different habitats of the Gauja National Park
Plant communities after Prieditis 1999* and Kabucis 2000**, forest types after Buds 1997.
Habitat
Dry pine forests on poor sandy
soil
Dry mesotrophic pine and pinespruce
forests
Dry mesotrophic spruce forests
Moist eutrophic broad-leaved
forests
High moors with pine and pine
forests on wet peat
Degraded high moors and pine
forests on drained peat
Other wet forest types
Edges of dry pine, pine-spruce
forests
Edges of dry spruce and leaf
tree forests
Agricultural landscapes,
meadows and shrubs
Total
Plant community
Cladonio-Pinetum,Vaccinio
vitis-idaeae-Pinetum*
Vaccinio myrtilli-Pinetum,
several types not classified
yet*
Oxalido-Piceetum exeelsae*
Querco-Tilietum*
Sphagnion
magellanici**,Vaccinio
uliginosi-Pinetum*
not classified yet*
the distance between individuals was
more than 100 meters. This served to
reduce the effect of occasional observa- ~~*l
tions of a large number of individuals at
the same site due to higher activity in
optimal weather conditions or better visibility
for the observer. Juveniles were
excluded from these data. In the analysis
of dominant tree species in habitats, data
acquired from transect censuses were
supplemented with descriptions of
twelve locations where the reptile
species were observed outside transects.
For amphibians, the type of habitat was
determined later from State Forest
Service forest plans and its data base.
Amphibians along transects were not
counted, and densities were not calculated
in cases when numerous individuals
were observed. The transect method was
not applied for the recording of newts,
and information on these animals was
collected occasionally.
Species distribution maps were prepared
using the Baltic Co-ordinate System,
Forest type
Cladinoso-callunosa,
Vaccinosa
Myrtillosa,
Hylocomiosa
Oxalidosa
Aegopodiosa
moor, Sphagnosa,
Caricoso-phragmitosa
Callunosa turf, me).,
Vaccinosa turf. mel.
Myrtilloso-sphagnosa,
Dryopterioso-caricosa
Myrtillosa,
Hylocomiosa
Oxalidosa,
Aegopodiosa
Transect length,
km (%)
2.8(1.7)
78.1 (47.0)
9.3 (5.6)
2.3(1.4)
8.4(5.1)
1.7(1.0)
1.1 (0.7)
7.5 (4.5)
3.4(2.0)
51.6(31.0)
166.2 (100.0)
Transverse Mercator Projection (TM-
1993). The number of 1x1 km squares of
this Co-ordinate System, in which
species were observed, was used to estimate
the occurrence of various species in
the Gauja National Park. A total of 269
or 27.5 % of all 1x1 km squares of the
Park were visited in the survey. The
number of the squares is not necessary
identical to the number of locations indicated
for species in the results, as several
locations in the same square are possible.
RESULTS
Six reptile and five amphibian species
were found during the survey. At least
one species was observed in 201 or 74.7
% of the visited 1x1 km squares.
Observation frequency for the various
species is shown in Table 2.
Reptiles
The Sand Lizard Lacerta agilis was found
in 2 areas: dry pine Pinus sylvestris forest

20 3/1-2, 2OO2 CEIRANS
Table 2. Occurrence of reptiles and amphibians in 1x1 km squares of the Baltic Co-ordinate
System that were crossed by transects in the Gauja National Park
Species
Reptiles
Lacerta agilis
Lacerta vivipara
Anguis fragilis
Natrix natrix
Vipera bents
Amphibians
Triturus cristatus
Triturus vulgaris
Bufo bufo
Rana arvalis
Rana temporaria
Rana synklepton esculenta
in the south-western part of the Park (2
locations), and banks and terraces of the
river Gauja and smaller tributaries in the
central part of the Park (5 locations).
The habitats of Lacerta agilis can be
grouped in two types:
1. low and sparse Pinus sylvestris growing
(sometimes mixed with birch Betula
spp. and spruce Picea abies) on dry sand,
with a tree height of 2-7 m and canopy
cover of 10-30 %; herb layer dominated
by grasses (Poa, Festuca, Calamagrostis)
and, in some cases, with horsetail
(Equisetum); herb cover of 10-50%;
habitat found on the banks of the river
and in some disturbed habitats (such as
old sand quarries) on terrace (5 locations);
2. fringes of dry and tall pine forest
(Vaccinio vitis-idaeae - Pinetum,
Vacdnio myrtilli - Pinetum associations)
on sandy soil on terrace (2 locations).
The Common Lizard Lacerta vivipara was
more or less evenly distributed throughout
the territory of the Gauja National
No. of
squares
6
59
3
9
1
1
1
91
8
133
40
% of visited
squares
2.2
21.9
3.0
3.3
0.4
0.4
0.4
33.8
3.0
49.4
14.9
Park. This species was observed in most
of the habitats, except some types of
forest. The highest density was in high
""moors and wet pine forests on peat,
especially in drained sites, and along the
fringes of various dry forests (Table 3).
The dominant tree species in Lacerta
vivipara wet forest habitats usually was
pine Pinus sylvestris, while deciduous
trees (Betula spp., Populus tremula,
Quercus robur, Tilia cordata) were characteristic
of open habitats with separate
trees or small groups of trees, such as in
open agricultural landscapes.
Lacerta vivipara preferred forest habitats,
especially mature dry forest, where it
was observed exclusively on relatively
open sites such as forest clearcuts, grassy
roadsides, banks, fringes and sites with
large gaps in the forest canopy.
The Slow Worm Anguis fragilis was
found rarely, in upland pine dominated
forest areas in the whole territory, and
only in dry forest types. Specimens were
usually observed on or near paths. Two

CEIRANS 3/1-2, 2OO2 21
Table 3. Occurrence of Lacerta vivipara on transects in various habitats.
Only the habitats where species were found are included in the list. The first value indicates
the number of observations, the second - density of findings. For the transect
length in the habitat see Table 1. Separate observations that were made outside the
transects are not included.
Habitat
Dry mesotrophic pine and pine-spruce forests
High moors with pine and pine forests on wet peat
Degraded high moors and pine forests on drained peat
Fringes of dry pine, pine-spruce forests
Fringes of dry spruce and leaf tree forests
Agricultural landscapes, meadows and shrubs
locations (or 0.71 records/km) were in
pine forest on poor sandy soil (association
Vaccinia vitis-idaeae - Pinetum),
and seven locations (0.09 records/km) in
pine and pine-spruce forest on
mesotrophic soil. Pinus sylvestris was the
dominant tree species in all locations.
The Grass Snake Natrix natrix was regularly
observed in the south-western part
of the Park, particularly near Sigulda.
The species inhabited the following habitats:
young and patchy Alnus incana and
Betula stands and meadows with pools
located on the bed of the ancient Gauja
river valley (5 locations), old broadleaved
forests with Fraxinus excelsior,
Quercus robur, and Ulmus glabra on the
slopes of the valley (association Querco-
Tilietum, 2 locations), and dry
mesotrophic pine and pine-spruce
forests on terraces of the valley (4 locations).
In forest habitats, it was usually
found in habitat edges and roadsides.
There were also some records of Natrix
natrix in 1985-1987 from other parts of
the Park, mostly along the Gauja River
(Z. Bruneniece, unpublished Bachelor's
thesis).
The Adder Vipera berus was found in
only two locations during the survey. The
first was on the edge of a clay quarry
with young and sparse Betula and Picea
abies, with Calluna vulgaris in the
ground cover. This site was located on a
Records
12/0.15
5/0.60
3/1.79
5/0.67
2/0.59
15/0.29
Individuals
15/0.19
5/0.60
8/4.71
5/0.67
2/0.59
17/0.33
slope with western exposure, within dry
mesotrophic pine and spruce forests in
the northern part of the Park. The other
observation of Vipera berus was made
about 500 m outside the eastern border
of the Park, on the grassy railway
embankment near a young Betula-Salix-
Alnus incana stand in a hilly landscape.
Vipera berus is also present in the northeastern
part of Sudas-Zviedru mire
(observed by M. Deicmane); however,
the species was not detected there during
the transect counts.
Amphibians
Two species were common on transects,
and were found in various habitats
throughout the entire territory of the
Park - the Common Toad Bufo bufo and
the Common Frog Rana temporaria.
Rana temporaria was the dominant
species in all natural habitats, except for
the Sudas-Zviedru high moor and sometimes
also the broad-leaved forests on
the slopes of the Gauja River valley. Bufo
bufo were fewer in number than Rana
temporaria in all dry forest habitats, with
the exception of a few cases in broadleaved
forest on valley slopes (association
Querco-Tilietum). This species was
infrequent on high moors and in wet
pine forests on peat. Both Bufo bufo and
Rana temporaria were relatively common
on agricultural landscapes, especial-

22 Biota 3/1-2,2002 CEIRANS
ly Bufo bufo, which was a usual inhabitant
of the local villages. Both species
were absent in the driest fragments of
pine forests on sandy soil (association
Cladinio-Pinetum).
Green frogs Rana synklepton esculenta
were common in aquatic and semi-aquatic
habitats in forest and agricultural landscapes,
especially wet habitats with many
pools. Green frogs were usual inhabitants
of ponds on agricultural farms. Most of
the examined individuals belonged to the
Pool Frog Rana lessonae; a few specimens
of the Green Frog Rana esculenta
were found in the largest lake of the Park
(Ungura Lake, 394 ha).
The Moor Frog Rana arvalis was a rare
species; a few individuals were occasionally
observed in the entire territory of the
Park. It was more common in the central
parts of the Sudas-Zviedru mire, the
largest high moor of the Gauja National
Park, and in the surrounding wet forest
where it was the dominant amphibian
species.
Amphibian roadkills are fairly common in
the Gauja National Park, especially of
Bufo bufo. The number of killed individuals
of this species on country roads during
spring migration in one case
(26.04.2000) reached 43 toads on a
0.025 km long road span, and in two
other cases (17.04.2000 and
17.05.2000), there were 11 and 12 killed
toads per 0.035 and 1.5 km, respectively.
Separate killed individuals of this species
were a common sight in or near human
settlements. Brown frogs were seldom
killed on roads, but on one day
(17.05.2000), 11 dead frogs per km were
recorded.
Two newt species were found at one
location each. A dead specimen of the
Great Crested Newt Triturus cristatus was
found on a road on the outskirts of the
town of Cesis. The road is located
between a wet deciduous tree forest area
and a garden area. There have been also
reports from zoologists (V. Pilats, A.
Minde) about findings in four other locations
in ponds near or in human settlements.
Larvae of the Smooth Newt
Triturus vulgaris were found in the central
part of the Park in a small forest lake with
swampy banks surrounded by wet pine
forest. There are two reports (from M.
Deicmane, M. Kalnins) about findings in
human settlements.
Comments on some other species
In 1988, a few individuals of the Fire-bellied
Toad Bombina bombina were
observed (by S. Inberga) near Sigulda in a
popular tourist area, but the species has
not survived in this location. Apparently
the observed individuals were released by
man. Bombina bombina in Latvia is on
the periphery of the distribution, and the
species inhabits only the southern part of
Latvia - Bauska and Daugavpils Districts
(unpublished data).
The Natterjack Toad Bufo calamita was
found (by M. Kalnins) in the 1990s in two
locations within a few kilometres of the
"Southern border of the Gauja National
Park, and the presence of this species in
the Park is highly possible.
There are records of the European Pond
Turtle Emys orbicularis in 1914 and 1925
in the territory of the present-day Gauja
National Park (Silins, Lamsters 1934).
Since then, there have been no new
records of this species, in spite of well
developed tourism in the Park. In Latvia,
there have been rare but regular findings
of Emys orbicularis in various regions. In
most cases, these animals were possibly
released by man, but in a few locations in
southern Latvia (Daugavpils and Dobele
. districts) this species is observed regularly
(personal communications by E. Tone, M.
Pupins) and small populations of this
species are expected there.

CEIRANS Biota 3/i-2, 2OO2 23
DISCUSSION
A survey using methods similar to those
of the present study was carried out in
1994-1997 in the Kemeri National Park
(Ceirans, in press). That National Park is
located in the central part of Latvia, but
only 15 % of its territory is covered by
agricultural land and human settlements
(compared to 45 % in the Gauja National
Park). The frequency of records (% of visited
1x1 km squares where species was
found) for Lacerta vivipara, L. agilis and
Vipera berus was about the same in both
Parks. The frequency for Anguis fragilis
and Matrix natrix records was respectively
about 3 and 5 times higher in the
Kemeri National Park. For the latter
species, this difference is probably associated
with climatic differences.
The large proportion of agricultural landscapes
in the Gauja National Park affects
to a great extent the status of species.
Four species of reptiles and amphibians
(Lacerta vivipara, Bufo bufo, Rana ternporaria,
and Rana synklepton esculenta)
are common and widespread in the
whole territory. All of these species are*»
typical of agricultural landscapes of the
Park, where farming methods are not as
intensive as in Western Europe or in some
other Latvian regions. The remaining
species are not adapted to these landscapes
(Anguis fragilis, Vipera berus,
Triturus vulgaris, Triturus cristatus) or
(and) are near the limits of their climatic
tolerance (Lacerta agilis, Natrix natrix),
which are the reasons for their rarity.
Rana arvalis was found in the same habitats
as Rana temporaria, but usually in
considerably lower numbers (excepting
for some high moor areas). Among frogs
killed on roads in spring 1999-2000 in the
southern part of the Gauja National Park,
not far from the Sudas-Zviedru high moor
(I. Valikova, unpublished Bachelor's thesis),
Rana arvalis and R. temporaria individuals
were found in proportions from
1:5 to 1:10.
The Adder Vipera berus population is
probably declining in the Gauja National
Park. In a survey carried out in 1985-
1987, and based on verified reports from
local residents, (Z. Bruneniece, unpublished
Bachelor's thesis), records of
Anguis fragilis, Natrix natrix, and Vipera
berus were present in a proportion
1:0.9:0.6, compared to a proportion of
0.6:1:0.05 in the present survey (1999-
2000) based on transect routes. As the
total number of findings of these three
species was considerably higher in the
former survey, a decrease of all three
species is possible. Populations of Anguis
fragilis, Natrix natrix, and Vipera berus
show long-term declining trends in
Finland (Terhivuo 1993), and decline of
Vipera berus is recorded for North
European Russia (Orlov & Ananjeva
1995). Long-term declining trends for
these species in Latvia is also possible.
Lacerta vivipara inhabits diverse habitats
in both forest and open landscapes of the
Park. For this species, a change of dominant
tree species depending on humidity
and canopy closure in the habitat was
observed. There was a correlation
between the occurrence of Lacerta vivipara
and that of pine in wet and closed
habitats, and that of deciduous trees in
dry open habitats. The latter possibly
reflects the dominance of deciduous trees
in open landscapes of the Park, and not
that deciduous trees were preferred by
Lacerta vivipara. However, these shifts
can reflect the microclimatic preferences
of the species. Pine stands are usually well
lit, and in the case of wet stands (such as
in high moor or forest on wet peat), fairly
open. Deciduous tree stands in wet
sites are usually shaded and may be too
cool for Lacerta vivipara. In the open
landscapes, summer temperatures can be
considerably higher than in the forest
sites. Sparse, dry and relatively open pine
stands may be too dry and hot for Lacerta
vivipara, and suitable rather to Lacerta
agilis which inhabits such sites on the

24 Biota 3/1-2,2002 CEIRANS
n high moors of the Moscow region of
Russia, the most favourable habitat for
Lacerta vivipara was observed to be wet
sites with sparse low pine stands and
Myrica gale, Ledum palustre,
Eriophorum vaginatum, and Sphagnum
in the herb and moss layer
(Zamolodchikov & Avilova 1989). In
high moors of the Gauja National Park,
Lacerta vivipara preferred sites with a
stable water regime, such as sparse low
pine-birch stands near old harvested
peat areas or in partially drained moor
areas, with characteristic swards of
Calluna vulgaris and grasses (Molinia
caerulea etc.) in the herb layer.
During the survey, large numbers of
killed amphibians on roads were
observed in relatively few cases, mostly
regarding Bufo bufo. There are data
from the southern part of the Park, near
Ligatne (I.Valikova, unpublished
Bachelors thesis), where amphibians
killed on roads were counted during the
spawning period (April - first 10 days of
May) for two seasons (1999-2000). The
maximum numbers of kills were 212
Rana temporaria, 35 Bufo bufo, and 14
Rana an/alls in one spring per 1 km of
road. In one case, about 15 km outside
the border of the Gauja National Park,
382 killed Rana temporaria and 70 Bufo
bufo per km were counted. This indicates
that the number of killed brown
frogs recorded in the Park in the present
survey - maximum 11 frogs/km - is an
REFERENCES
underestimate, probably due to the timing
of field work after the end of the
amphibian spawning season.
The following conservation and research
activities are recommended for the Gauja
National Park:
- Creating a data base of records at least
for rare species (Lacerta agilis, Anguis
fragilis, Natrix natrix, Vipera berus,
Triturus vulgaris, T. cristatus, Rana
an/alls) using the 1999-2000 inventory
project as a base;
- More research on ecology of all species
is needed to encourage the development
of a management plan for reptiles and
amphibians in the Gauja National Park;
- Development of a network of protected
microreserves for Triturus cristatus.
The species is included in Appendices II
of both the Bern Convention
(Anonymous 1992a) and ED Directive
92/43EEC (Anonymous 1992b), and in
the 2nd category (vulnerable species) of
the Red Data Book of Latvia (Ingelog et.
al. 1993);
- Measures should be made to avoid
"•amphibian road kills. In cases of mass
road kills, conservation activities such as
construction of roadside fences and
underground passages should be
planned;
- Increasing public awareness of reptile
and amphibian conservation in the Park.
Further tourism development could have
a negative effect on some species, particularly
snakes.
ANONYMOUS 1992a: Appendix II, Strictly Protected Fauna Species to the Convention on
the Conservation of European Wildlife and Natural Habitats, Bern, 1979.
Directorate of Environment and Local Authorities, Strasbourg.
ANONYMOUS 1992b: EU Directive 92/43/EEC on the Conservation of Natural Habitats and
Wild Fauna and Flora, Brussels.
BUSS, K. 1997: Forest ecosystem classification in Latvia. Proceedings of the Latvian Academy
of Sciences. Section B 51: 204-218.
CEIRANS, A. in press: Reptiles and anurans of the Kemeri National Park, Latvia. In: Heikkila,

CEIRANS Biota 3/i-a, 2002 25
R. & Lindholm, T. (eds.). Biodiversity and conservation of the boreal nature.
Proceedings of the 10 years anniversary symposium of the Nature Reserve
Friendship. The Finnish Environment 485.
INGELOG, T., ANDERSSON, R. & TJERNBERG, M. (eds.) 1993: Red Data Book of the Baltic
Region. Part 1. Lists of Threatened Vascular Plants and Vertebrates. Swedish
Threatened Species Unit, Uppsala.
KABUCIS, I. (ed.) 2000: Biotopu rokasgramata. Eiropas Savienibas aizsargajamie biotopi
Latvija [Handbook of habitats. Protected habitats by European Union in Latvia].
Latvijas dabas fonds, Riga (in Latvian).
ORLOV, N.L. & ANANJEVA, N.B. 1995: Distribution of amphibians and reptiles and their
relict populations in the Gulf of Finland and Lake Ladoga. Memoranda Societatis
pro Fauna et Flora Fennica 71: 109-112.
PILATS, V. 2000: Emeralds of Latvia. Opportunities for nature tourism. Ministry of
Environmental Protection and Regional Development, Riga.
PRIEDITIS, N. 1999: Latvian forest: nature and diversity. WWF, Riga.
SILINS, J. & LAMSTERS, V. 1934: Latvijas rapuli un abinieki [Reptiles and amphibians of
Latvia]. Valters un Rapa, Riga (in Latvian).
TERHIVUO, J. 1993: Provisional atlas and status of populations for the herpetofauna of
Finland in 1980-92. Annales Zoologici Fennici 30: 55-69.
ZAMOLODCHIKOV, D.G. & AVILOVA, K.V. 1989: Materials on biology of the Common
Lizard, Lacerta vivipara, in a high moor in Western Moscow region. In:
Amphibians and reptiles of Moscow Region. Nauka, Moscow: 147-152 (in
Russian).

GENTILLI, SCALI, BARBIERI & BERNINI Biota 3/i-a, 2002 27
A three-year project for the
management and the conservation
of amphibians in Northern Italy
Augusto GENTILLI1, Stefano SCALI2r
Francesco BARBIERI1 & Franco BERNINI1
1 Dip. Biologia Animale, University of Pavia, p.za Botta 9, 27100 Pavia, Italy
2 Museo Civico di Storia Naturale, c.so Venezia 55, 20121 Milano-ltaly
Abstract
In 1998 the Lombardy District (Regione Lombardia, Northern Italy), in co-operation
with the Italian Ministry of Environment, started an integrated three-year project for
conservation of amphibians in twelve regional parks, in both mountain and plain areas.
The aims of the project were translocation of some threatened amphibian species and
habitat management of some endangered environments. The target species were
Salamandra salamandra, Triturus camifex, Pelobates fuscus insubricus, Bombina variegata,
Hyla intermedia, Rana latastei, R. dalmatina and R. temporaria. The target habitats
are natural and artificial ponds and springs located in areas where these environments
are disappearing.
The translocations involved P. fuscus insubricus and R. latastei; eggs of these species
were collected in areas in Northern Italy in accordance with taxonomic and conservation
interests. Eggs and tadpoles were bred in semi-natural conditions by the biologists
of the University of Pavia. Tadpoles were released immediately before metamorphosis.
One half of bred tadpoles were released in the areas where the eggs were collected.
The new and restored ponds in most cases have been spontaneously colonized by seven
amphibian species: Salamandra salamandra, Bufo bufo, Hyla intermedia, Rana dalmatina,
R. latastei, R. temporaria and R. synklepton esculenta. The successful metamorphosis
and survival after the wintering period of the translocated specimens were also
observed. These preliminary results underline the importance and the validity of these
methods of translocation and habitat management for amphibian conservation projects.
Key words: Amphibians, conservation, translocation, habitat management, Northern
Italy
Received 24 September; accepted 12 December 2001

28 Biota 3/1-2,2002 GENTILLI, SCALI, BARBIERI & BERNINI
INTRODUCTION
The use of habitat management and
translocations for amphibian conservation
has been proposed by many authors
(Fog 1988, Burke 1991, Dodd & Seigel
1991, Reinert 1991, Bray 1994, Langton
et al. 1994, Amtkjaer 1995, Andren &
Nilson 1995a, b, Cooke & Oldham 1995,
Gubbels 1995, Hels & Fog 1995, Juul
1995, Zvirgzds et al. 1995, Griffiths,
1996, Agger 1997, Briggs 1997, Fog
1997, Jensen 1997, Lacoste & Durrer
1998). A three year project based on
these methods was started in 1998 to
preserve amphibian species and their
habitats in Lombardy (Northern Italy).
The project was made possible thanks to
the appropriation of € 250.000 by the
Regione Lombardia and the Italian
Ministry of the Environment. The
authors describe the methods used during
the project and its preliminary results.
MATERIALS AND METHODS
The project involved 12 natural parks,
seven of which are located in the Po
Plain (Table 1) and five in the Alps (Table
2). The present or historical presence of
amphibian species in the parks was
assessed by bibliographic and personal
data (Andreone et al. 1993, Societas
Herpetologica Italica 1996, Bonini et al.
2000).
Using these data, we identified target
species based on two main criteria:
endemic or rare taxa considered important
by European laws (Habitats
Directive 92/43/European Community:
T. carnifex, P. fuscus insubricus, B. variegata,
H. intermedia, R. dalmatina, R.
latastei) and species that, although not
rare in Northern Italy, are in decline in
some areas (i.e., S. salamandra, H. intermedia,
R. dalmatina and R. temporaria).
The literature on these species was
examined to gather data on their ecological
needs. Only Pelobates fuscus insubricus
and Rana latastei were chosen for
translocations, because they are quite
rare in many areas of Lombardy and they
are endemic to the Po Plain.
Forty suitable areas for habitat management
and translocations were identified
following many field surveys with park
staff. The causes of amphibian decline in
these sites were analysed and some
Actions for habitat restoration were proposed,
in accordance with the guidelines
of the I.U.C.N., of the National Wildlife
Institute (I.N.F.S.), and of the Societas
Table 1. Target species and conservation measures chosen for lowland parks; HM
Habitat Management, Rl = Reproduction, PR = Population reinforcement.
Park
Parco Pineta di Appiano
Gentile e Tradate
Parco Ticino
Parco Agricolo Sud
Parco Adda Sud
Parco Serio
Parco Oglio Sud
Parco Mincio
PLAIN PARKS
Target species
Salamandra salamandra
Triturus carnifex
Rana dalmatina
Pelobates fuscus
Pelobates fuscus
Rana latastei
Pelobates fuscus
Rana latastei
Rana latastei
Rana latastei
Rana latastei
HM
Conservation
measures
HM,RI
HM,RI
HM,RI
HM,RI
HM,PR
HM,RI
HM
HM

GENTILLI, SCALI, BARBIERI & BERNINI Biota 3/1-2,2002 29
Table 2. Target species and conservation measures chosen for alpine parks; HM =
Habitat Management, Rl = Reintroduction, PR = Population reinforcement.
Park
Parco Monte Barro
Parco Orobie Valtel lines!
Parco Colli di Bergamo
Parco Adamello
Parco Alto Garda Bresciano
Herpetologica Italica (I.N.F.S. 1995,
Stanley Price & Fairclough 1997, Societas
Herpetologica Italica 1997). In particular,
we chose public areas that guaranteed
major protection status over private
ones; the sites where habitat restoration
was not possible were discarded. The
choice of sites was also based on ecological,
habitat structural and human factors
(Scali et al. 2002). In addition, the chosen
sites were contiguous with other
suitable damp areas, in order to re-create
metapopulations.
Many techniques were used for habitat
restoration:
• creation of new ponds: some pools
were excavated in suitable areas to
increase reproductive sites;
• restoration of old ponds: some desiccated
or almost desiccated existing
ponds were excavated to increase water
levels during the breeding period;
• waterproofing of ponds: new and old
ponds were waterproofed with clay or
PVC, if necessary, to guarantee water
permanence;
• increase of laying points: some deadwood
was introduced to facilitate egg
anchorage;
• excavation of tributary canals: when
necessary, they were dug to allow water
permanence;
• positioning of metallic grids: these were
used to stop predator fish entrance;
• removal of fish: these vertebrates were
removed, if already present, using an
ALPINE PARKS
Target species
Salamandra salamandra
Rana temporaria
Bombina variegata
Rana temporaria
Salamandra salamandra
Hyla intermedia
Rana temporaria
Conservation measures
HM
HM
HM
HM,PR
HM
electrostunner, and they were released in
contiguous damp areas.
Translocations were carried on only after
habitat management and only if we considered
reintroductions or population
reinforcements necessary: i) when a population
was extinct and re-colonization
was impossible or unlikely, ii) when a
population had strongly declined and its
full recovery was unlikely or it required a
very long time.
The eggs of Pelobates fuscus insubricus
and Rana latastei were collected in sites
where these species are quite abundant
and located near the translocation areas,
to guarantee genetic homogeneity. In
fact, no geographic barriers are present
in areas involved in the plan, and there
are no borders to the biogeographic
regions. Furthermore, no information
about the genetics of Rana latastei and
Pelobates fuscus insubricus are available
for these areas, and no funds were
appropriated for this kind of research.
Eggs were transported to Bosco Negri
Natural Reserve, which is managed by
the Italian League for Bird Protection
(LIPU), and they were hatched in seminatural
conditions by the authors. Eggs
coming from different clumps were
mixed to increase genetic diversity.
Tadpoles were released only when their
hind legs were developing; one half of
them were released in the donor sites,
because we do not want to impoverish
those populations.

30 Biota 3/i-a, 2002 GENTILLI, SCALI, BARBIERI & BERNINI
RESULTS
To date, 56 habitat management projects
have been carried out in 34 different
sites. During the year 2000 a total of
2000 tadpoles of P. fuscus and 12,000 of
R. latastei were bred; in 2001, the number
of R. latastei increased to 28,000. In
the same year P. fuscus did not breed,
due to unfavourable climatic conditions.
In the former year we made six reintroductions
of P. fuscus insubricus, and two
reintroductions and three population
reinforcements of R. latastei. During the
second year we made seven reintroductions
and six population reinforcements
of R. latastei.
Metamorphosis occurred without any
particular problem in both years and
some juveniles of the spadefoot toad
were found in spring 2001. The habitat
management works led to an improvement
of environmental conditions for
amphibians in general. In fact, the populations
of different species already living
in those areas often colonized the new
ponds and the number of egg clumps
increased in the second year (Table 3 and
DISCUSSION
Preliminary data suggest that the project
is quite successful. This result is apparent
from the increase of breeding specimens
of the species already existing in managed
areas and by the successful metamorphosis
of many tadpoles of the
translocated species. No information is
available at the present time about
future reproductive success of the managed
populations. To achieve these data,
a monitoring phase is required in future
years. The breeding operations, the
translocations, and the habitat management
works will continue through 2002
to improve the preliminary results.
This kind of project guarantees a fast and
easy way to reconstitute wild populations;
in fact, amphibians can be bred at
a low cost and a large number of future
breeders can be achieved in a short time;
moreover, the breeding and the release
do not involve behavioural problems, as
is often reported for birds and mammals
(Bloxam & Tonge 1995).
It is important to underline that many of
The increased species are protected by
European laws (Habitat Directive). Our
observations highlight the importance of
a co-ordinate project for an homogeneous
line of conduct and for an optimization
of economic resources.

GENTILLI, SCAU, BARBIERI & BERNINI Biota 3/i-a, 2002 31
Table 3. Increase of already present species and colonization by new species in the lowland
parks.
Plain parks
Parco Pineta di Appiano Gentile
e Tradate
Parco Ticino
Parco Agricolo Sud
Parco Adda Sud
Parco Serio
Parco Oglio Sud
Parco Mincio
Increase
T. vulgaris
T. camifex
R dalmatina
R. synklepton esculenta
R. latastei
R. latastei
Colonization
S. salamandra
T. carnifex
R. dalmatina
T. vulgaris
B. bufo
H. intermedia
R. latastei
R. dalmatina
R. synklepton esculenta
H. intermedia
R. synklepton esculenta
H. intermedia
R. synklepton esculenta
R. synklepton esculenta
R. synklepton esculenta
R. synklepton esculenta
Table 4. Increase of already present species and colonization by new species in the
alpine parks.
Plain parks Increase Colonization
Parco Monte Barro
Parco Orobie Valtellinesi
Parco Colli di Bergamo
Parco Adamello
Parco Alto Garda Bresciano
"*& salamandra
R. temporaria
S. salamandra
B. variegata
R. temporaria
S. salamandra R, temporaria
REFERENCES
AGGER, P. 1997: The background for the habitat management, monitoring and nature
restoration in Denmark. Memoranda pro Societatis Fauna et Flora Fennica 73: 87-
90.
AMTKJAER, J. 1995: Increasing populations of the Green Toad (Bufo viridis) due to a pond
project on the island of Samso. Memoranda pro Societatis Fauna et Flora Fennica
71:77-81.
ANDREN, C. & NILSON, G. 1995a: Re-introduction of the Fire-bellied Toad Bombina
bombina in southern Sweden. Memoranda pro Societatis Fauna et Flora Fennica
71:82-83.
ANDREN, C. & NILSON, G. 1995b: Translocation of amphibians and reptiles. Memoranda
pro Societatis Fauna et Flora Fennica 71: 84-87.
ANDREONE, F., FORTINA, R. & CHIMINELLO, A. 1993: Natural history, ecology and conservation
of the Italian Spadefoot toad, Pelobates fuscus insubricus. [Storia nat-

32 )ta 3/1-2, 2OO2 GENTILLI, SCALI, BARBIERI & BERNINI
urale, ecologia e conservazione del Pelobate Insubrico, Pelobates fuscus insubricus].
Scientific reports, Zoological Society "La Torbiera", Agrate Conturbia
(NO), 2: 1-92.
BLOXAM, Q.M.C. & TONGE, S.J. 1995: Amphibians: suitable candidates for breedingrelease
programmes. Biodiversity and Conservation 4: 636-644.
BONINI, L, GENTILLI, A., RAZZETTI, E. & SCALI, S. 2000: Progetto atlante erpetologico
lombardo: cartine definitive (27 maggio 2000). Societas Herpetologica Italica
Sezione Lombardia, Centra Stampa dell'Universita di Pavia: 1-24.
BRAY, R. 1994: Case study: A programme of habitat creation and great crested newt introduction
to restored opencast land for British Coast Opencast. In: Gent, T. & Bray,
R. (eds.). Conservation and management of the great crested newts Triturus
cristatus, 11 January 1994 Kew Gardens, Richmond, Surrey. English Nature,
Surrey: 113-126.
BRIGGS, L. 1997: Recovery of Bombina bombina in Funen County, Denmark. Memoranda
pro Societatis Fauna et Flora Fennica 73: 101-104.
BURKE, R.L. 1991: Relocations, repatriations, and translocations of amphibians and rep tiles:
taking a broader view. Herpetologica 47: 350-357.
COOKE, A.S. & OLDHAM, R.S. 1995: Establishment of populations of the common frog,
Rana tetnporaria and common toad, Bufo bufo, in a newly created reserve following
translocation. Herpetological Journal 5: 173-180.
DODD, C.K. Jr. & SEIGEL, R.A. 1991: Relocation, repatriation, and translocation of amphibians
and reptiles: are they conservation strategies that work? Herpetologica 47:
336-350.
FOG, K. 1988: Pond restoration on Bornholm. Memoranda pro Societatis Fauna Flora
Fennica 64: 143-145.
FOG, K. 1997: A survey of the results of pond projects for rare amphibians in Denmark.
Memoranda pro Societatis Fauna et Flora Fennica 73: 91-100.
GRIFFITHS, R.A. 1996: Newts and Salamanders"of Europe. Poyser Ltd., London: 1-188.
GUBBELS, R. 1995: The restoration of habitats for Amphibians by the Roer and Overmaas
Waterboard with special attention for the restoration of the most important
breeding biotope of the tree frog (Hyla arborea L.) in the Dutch province of
Limburg. In: Llorente; A., Santos, X. & Carretero, M.A. (eds.). Scientia
Herpetologica, Asociacion Herpetologica Espanola, Barcelona: 361 - 364.
HELS, T. & FOG, K. 1995: Does it help to restore ponds? A case of the Tree Frog (Hyla
arborea). Memoranda pro Societatis Fauna et Flora Fennica 71: 93-95.
I.N.F.S. 1995: Documento sulle immissioni faunistiche. Linee guida per le introduzioni, reintroduzioni
e ripopolamenti di Uccelli e Mammiferi. Ill Convegno Nazionale Biologi
della Selvaggina, Bologna 9-11 Febbraio 1995.
JENSEN, B.H. 1997: Relocation of a garlic toad (Pelobates fuscus) population. Memoranda
pro Societatis Fauna et Flora Fennica 73: 111-113.
JUUL, K. 1995: New ponds in Aarhus (Denmark). Memoranda pro Societatis Fauna et Flora
Fennica 71: 146-148.
LACOSTE, V. & DURRER, H. 1998: Past distribution and current status of the common
spadefoot (Pelobates fuscus) in the Plain of the Upper Rhine and strategies of
reintroduction. In: Miaud, C. & Guyetant, R. (eds.). Current studies in her- petology,
SEH, Le Bourget du Lac, France: 239-247.
LANGTON, T.E.S., BECKETT, K., MORGAN, K. & DRYDEN, R.C. 1994: Translocation of a
crested newt Triturus cristatus population from a site in Crewe, Cheshire, to a
nearby receptor site. In: Gent, T. & Bray, R. (eds.). Conservation and man agement
of the great crested newts Triturus cristatus, 11 January 1994 Kew Gardens,

GENTILLI, SCALI, BARBIERI & BERNINI BJOta 3/1-2,2002 33
Richmond, Surrey. English Nature, Richmond: 92-103.
REINERT, H.K. 1991: Translocation as a conservation strategy for Amphibians and Reptiles:
some comments, concerns, and observations. Herpetologica 47: 357-363.
SCALI, S., GENTILLI, A., BARBIERI, R, BERNINI, F. & VERCESI, A. 2002: An integrated project
for conservation of amphibians in some protected areas in Northern Italy. Atti
3° Convegno "Salvaguardia Anfibi", Lugano 23-24 Giugno 2000,
Cugecstre Edizioni, Penne: 151-158
Societas Herpetologica Italica 1996: Atlante prowisorio degli Anfibi e dei Rettili italiani.
Annali del Museo Civico di Storia Naturale "Giacomo Doria", Geneva 91: 95-
178.
Societas Herpetologica Italica 1997: Linee guida per le reintroduzioni, i ripopo lament! e le
introduzioni di Anfibi e Rettili. In: Commissione Conservazione S.H.I.,
Herp.Lex.98: 6-11.
STANLEY PRICE, M.R. & FAIRCLOUGH, A. 1997: Translocation of wildlife: the ILJCN position
statement and general considerations on behavioural constraints to release.
Supplement! Ricerche di Biologia della Selvaggina 27: 25-38.
ZVIRGZDS, J., STASULS, A/I. & VILNITIS, V. 1995: Reproduction of the European Tree Frog
(Hyla arborea) in Latvia. Memoranda pro Societatis Fauna et Flora Fennica 71:
139-142.

COILMANN, COLLMANN, BAUMGARTNER & WARINGER BJQta 3/1-2, 2002 35
Spawning site shifts by Rana
dalmatina and Rana temporaria in
response to habitat change
Gunter GOLLMANN1, Birgit GOLLMANN1,
Christian BAUMGARTNER2 & Andrea
WARINGER-LOSCHENKOHL2
'Institut fur Zoologie, Universitat Wien, Althanstr. 14, A-1090 Wien, Austria
E-mail: guenter.gollmann@univie.ac.at
2lnstitut fur Okologie und Naturschutz, Universitat Wien, Althanstr. 14, A-1090 Wien,
Austria
Abstract
Between 1995 and 2001 we investigated spatial patterns of spawn deposition by Rana
dalmatina and R. temporaria in the flood retention basin of the Mauerbach stream in
western Vienna (Austria). This basin was constructed in the late 19th century; until
1996 it was bypassed by the stream flowing in a straight channel and contained several
temporary pools. In the course of reconstruction works the basin was flooded for prolonged
periods in 1997 and 1998 and a meadow close to the stream, outside the retention
basin, was inundated. Since 1998 the stream has been running partly through the
basin, connecting all the pools that existed previously. Owing to the rise of the water
table, a few new shallow pools formed in the upper part of the basin. Since 2000,
beaver dams have stabilized the water level in the lower part of the basin. Clutches of
R. dalmatina were usually laid separately, attached to vegetation such as stalks of reed,
whereas egg masses of R. temporaria were often aggregated in clusters. Both species
readily adopted new water bodies and ceased to use some areas frequented in the early
years of the study. A shift to shallower spawning sites, which was more distinct in R.
temporaria than in R. dalmatina, may be interpreted as predator avoidance or a preference
for warm microhabitats under generally colder conditions.
Key words: amphibians, ecology, oviposition, stream restoration, Ranidae
Received 10 November; accepted 15 December 2001

36 Biota 3/i-a, 2002 GOLLMANN, GOLLMANN, BAUMGARTNER & WARINGER
INTRODUCTION
Spawn is an important stage in the lifehistory
of most amphibians. Selection of
oviposition sites by spawning frogs has
consequences for the risks of desiccation
and predation on embryos and larvae as
well as for growth opportunities of the
tadpoles.
Two species of brown frogs, Rana dalmatina
and Rana temporaria, have widely
overlapping distributions in Europe
(Gasc et al. 1997), but syntopic occurrence
is rare in many regions. Their lifehistories
are similar; in Central Europe
both species breed in early spring. The
question whether their distributions are
influenced by interspecific competition
has remained unresolved (e.g. Riis 1988,
Rohrbach & Kuhn 1997). Competition
among amphibians is expected to occur
mainly during the larval stage (Wilbur
1996). Hence, spatial and temporal patterns
of spawn deposition determine the
potential for competition among tadpoles,
and may also influence the outcome
of this competition.
To assess possibilities for competition
and conditions of co-occurrence, we
have investigated oviposition, embryonic
and larval development in syntopic populations
of R. dalmatina and R. temporaria
at the western outskirts of Vienna
Figure 1. Water bodies in the study area 1995-1997. Outlines show maximal size of
temporary pools that mostly dried during summer.

GOLLMANN, GOLLMANN, BAUMGARTNER & WARINGER Biota 3/1-2, 2002 37
since 1995 (Baumgartner et al. 1996,
Gollmann et al. 1999). Reconstruction
works affecting the study area, including
a stream restoration project, caused profound
habitat change. Here, we describe
the spatial patterns of spawn deposition
over seven study years.
STUDY AREA AND HABITAT CHANGE
Observations were made in the flood
retention basin of the Mauerbach stream.
This basin was constructed in the late
19th century to retard water during flood
peaks. It measures approximately 350 x
70 m and until 1997 was bypassed by the
stream flowing in a straight channel.
Rgure 2. Water bodies in the study area since 1998.
Pool dry in 2001
Beaver dam
Temporary pools in the central and lower
parts of the basin usually dried in late
spring or early summer (Figure 1). In
1995, water from the stream never
flowed into the basin. On 8 April 1996
melting of snow in upstream areas caused
a brief flooding of the basin, which led to
the formation of a new ephemeral pool at
the upper end of the basin.
In autumn 1996 reconstruction works
started, with the aims of improving flood
protection (of downstream districts of
Vienna) and redirecting part of the
stream through the retention basin. In
spring of 1997, the basin was flooded by
the stream for prolonged periods, and a

38 Biota 3/1-2, 2002 GOLLWIANN, GOLLMANN, BAUAAGARTNER & WARINGER
meadow close to the stream, directly
above the retention basin, was inundated
(Figure 1). Heavy rainfalls in July
1997 caused deep submersion of the
retention basin, which lasted until
February 1998, when debris from the
outlet was removed. Since then, the
stream has been running partly through
the basin, connecting all ponds that had
existed previously. Owing to the rise of
the water table, a few new shallow
ponds formed in the upper parts of the
basin. Since winter 1999/2000, beavers
have inhabited the basin; they have
gradually been building dams, which
have increased and stabilized the water
level in the retention basin (Figure 2).
MATERIAL AND METHODS
During the brown frog breeding season
the study area was usually visited daily,
rarely at two day intervals. Minimummaximum
thermometers and watergauges
were placed at several sites in the
study area.
Each egg mass was individually marked
Figure 3. Partition of the study area
into four sections (see text and
Figure 4).
with a floating disk of cork tied to the
jelly with a string (1995-1999) or with a
straw pushed through the center of the
spawn clump (2000-2001). Position of
the clutch, water depth at the spawning
site and depth of the water column
above the spawn clump were recorded;
one egg was removed for species determination
by enzyme electrophoresis
(Baumgartner et al. 1996), which was
performed for most clutches.
RESULTS
To summarize spatial patterns of spawn
deposition, the study area is here divided
into four sections (Figures 3, 4): partitions
of the basin correspond to slight
rises in the ground, whereas pools outside
the basin are summarized as area D.
In the first study year, only pools in the
lower and central parts of the basin (A,
B) were available as breeding habitats.
Newly created water bodies, such as the
meadow inundated in 1997 (area D),
were readily accepted for spawning by
both species.
Figure 4. Number of clutches of R. dalmatina
and R. temporaria per year in the four parts of
the study area (see Figure 3).
1995 1996 1997 1998 1999 2000 2001
Rana dalmatina H Rana temporaria

GOLLMANN, GOLLMANN, BAUMGARTNER & WARINCER 3/1-2, 2002 39
Figure 5. Water depth at the spawning sites: Lower symbols indicate depth of the water
body at the spawning site (mean and standard deviation) for each species and year,
upper symbols show mean values for depth of the water column above the spawn
clump.
A •*•»
ffl 20-
30~
40 -1
A Sana dalmatina
4» Sana temporaria
\s of
1995 1996 1997 1998 1399 2000 2001
always deposited singly, often attached
to a central axis provided by a stalk of
reed or a branch. In the first four study
years, the earliest clutches were deposited
at the same site (the easternmost pool
in area B, Figure 3). When the water
level in the basin increased due to flooding
(since 1997), some clutches were laid
in deep water, whereas others were
deposited in shallow marginal areas,
leading to enhanced variation in water
depth at the spawning sites (Figure 5). In
2001, most egg masses were laid close
to the water surface.
Spawn of R. temporaria was often
aggregated in clusters, to which new egg
masses were added over several days.
When the pools in the basin were connected
to running water, spawning of
this species became restricted to marginal
areas; deep pools were no longer
used, which resulted in an overall shift to
lower mean values for water depth at
the spawning site (Figure 5).
In later years, both species ceased to
spawn in water bodies which had
become much deeper (areas A, B). Both
species laid eggs in a bay of the stream
outside the basin (area D) and in shallow
muddy pools in the uppermost part of
the basin, which were not connected to
the stream (area C; these pools were
mostly dry in 2001).
DISCUSSION
The stream restoration project took some
unexpected turns, such as a catastrophic
flood in summer 1997 during the construction
phase, and the immigration of
beavers. Overall, water bodies available
as frog spawning sites in "the study area
became much larger, more stable, and
cooler. Numbers and diversity of preda-

40 Biota 3/i-a, 2002 GOLLMANN, GOLLMANN, BAUMGARTNER & WARINGER
tors such as fish (including pike, Esox
lucius), birds (mallards, rails, herons) and
aquatic insects have increased greatly.
The effects of habitat change on population
dynamics are hard to predict. In the
short run, connection of spawning pools
to the stream was certainly advantageous.
Otherwise, no tadpoles could
have survived to metamorphosis in the
very dry years 2000 and 2001. For R.
dalmatina, the expansion of the water
bodies means loss of terrestrial habitat;
hence, a reduction in the number of
breeding frogs in the study area can be
expected. In the long run, decreased
desiccation risk is counteracted by
increased predation risk, but the outcome
of these opposing trends is uncertain,
and may differ between species. In
some riverine floodplains of southern
Germany, habitat management that
increased flooded areas or the number of
permanent pools led to growth of R.
temporaria populations (Kuhn 2001,
Laufer 2001, Utschick 2001).
Shifting spawning sites to the shallow
margins may have two reasons: predator
avoidance or a preference for warm
microhabitats under the generally cooler
conditions. Predation on frog eggs
seemed to be negligible in our study area
(unpublished data). Laurila & Aho
(1997) showed that R. temporaria did
not avoid spawning in pools containing
fish which prey on tadpoles. If predation
risk did at all influence spawning site
selection at our study site, adult frogs
may have evaded deep water because of
the presence of predatory fish such as
pike.
There was some opportunity for interspecific
competition among brown frog
tadpoles, especially in early study years
with drying pools, but its importance for
larval growth and survival was probably
subordinate to abiotic factors (hydroperiod,
water temperature, water velocity)
and predation.
Acknowledgements
We thank the students of the University
of Vienna field course in amphibian ecology
for their help in collecting and
analysing data. GG acknowledges support
by the Austrian Academy of
Sciences, in the framework of ecological
monitoring of river restructuring commissioned
by the Municipality of Vienna
(Magistratsabteilung 45, Wasserbau).
REFERENCES
BAUMGARTNER, C., BITSCHI, N., ELLINGER, N., GOLLMANN, B., GOLLMANN, G., KOCK,
M., LEBETH, E. & WARINGER-LOSCHENKOHL, A. 1996: Laichablage und
Embryonalentwicklung von Springfrosch (Rana dalmatina BONAPARTE, 1840)
und Grasfrosch (Rana temporaria LINNAEUS, 1758) in einem syntopen
Vorkommen (Anura: Ranidae). Herpetozoa9: 133-150.
GASC, J.-P., CABELA, A., CRNOBRNJA-ISAILOVIC, J., DOLMEN, D., GROSSENBACHER, K.,
HAFFNER, P., LESCURE, J., MARTENS, H., MARTINEZ RICA, J.P., MAURIN, H.,
OLIVEIRA, M.E. , SOFIANIDOU, T. S. , VEITH, M. & ZUIDERWIJK, A. (eds.)
1997: Atlas of Amphibians and Reptiles in Europe. Societas Europaea
Herpetologica and Museum National d'Histoire Naturelle (IEGP/SPN), Paris.
GOLLMANN, G., BAUMGARTNER, C., GOLLMANN, B. & WARINGER-LOSCHENKOHL, A.
1999: Breeding phenology of syntopic frog populations, Rana dalmatina and
Rana temporaria, in suburban Vienna. Verhandlungen der Gesellschaft fur
Okologie 29: 357-361.
KUHN, J. 2001: Amphibien in der Wildflusslandschaft der oberen Isar (Bayern):
Auswirkungen der "Teilruckleitung" seit 1990 und des Spitzenhochwassers

GOLLMANN, GOLLMANN, BAUMGARTNER & WARINGER Biota 3/1-2, 2002 41
1999. Zeitschrift fur Feldherpetologie 8: 43-56.
LAUFER, H. 2001: Amphibien in den Poldern Altenheim (Oberrhein, Baden-Wurttemberg):
Bestandsentwicklung und Auswirkung von Hochwassern. Zeitschrift fur
Feldherpetologie 8: 203-214.
LAURILA, A. & AHO, T. 1997: Do female common frogs choose their breeding habitat to
avoid predation on tadpoles? Oikos 78: 585-591.
RIIS, N. 1988: The present distribution of Rana dalmatina and Rana temporaria in southern
Scandinavia explained by a theory of competitive exclusion. Mem. Soc. Fauna
Flora Fennica 64: 104-106.
ROHRBACH, T. & KUHN, J. 1997: Der Springfrosch (Rana dalmatina) im westlichen
Bodenseeraum 1994-1996: Verbreitung, Bestande, Laichgewasser. In: Krone, A.,
Kiihnel.K.-D. & Berger.H. (eds:). Der Springfrosch (Rana dalmatina)- Okologie
und Bestandssituation. Rana / Sonderheft 2, Rangsdorf: 251-261.
UTSCHICK, H. 2001: Auswirkungen der Staustufe Perach auf die Amphibienbestande der
Aue (Unterer Inn, Bayern) in der Wildflusslandschaft der oberen Isar (Bayern).
Zeitschrift fur Feldherpetologie 8: 119-129.
WILBUR, H.M. 1996: Multistage live cycles. In: Rhodes, O.E., Jr., Chesser, R.K. & Smith,
M.H. (eds.) Population dynamics in ecological space and time. The University of
Chicago Press. Chicago & London: 75-108.

GROSSENBACHER 3/1-2, 2OO2 43
First results of a 20-year-study on
Common Toad Bufo bufo in the
Swiss Alps
Kurt GROSSENBACHER
Museum of Natural History, Bernastrasse 15, CH-3005 Bern
E-mail: groba@nmbe.unibe.ch
Abstract
From 1982 until 2001 a population of 150-200 Common Toad Bufo bufo on the Grosse
Scheidegg near Grindelwald in the Bernese Alps (altitude 1850m) was registered every
spring at the breeding site. The toads were marked in the first years with toe clipping,
in the last 9 years with transponders. The number of first time breeders showed a deep
depression in the late 80s and recovered in the nineteen-nineties. One third of all males
and two thirds of all females reproduce only once. The maximum reproductive rate of
a female is seven times within 11 years. Most females reproduced every second year,
but reproduction in two successive years is not rare. The maximum age observed is 25
years, of a male marked in 1983 which was present at the breeding place during 16
years. The maximum age of a female is 20 years. Reproduction success was very low in
the nineteen-nineties, but an effect on the adult population has not yet appeared.
Keywords: Bufo bufo, reproduction, age, longevity, population dynamics, high altitude,
Swiss Alps.
Received 29 October; accepted 18 December 2001

44 Biota 3/1-2,2002 GROSSENBACHER
INTRODUCTION
Hemelaar (1988) reported a 3-year study
of the population structure of Common
Toad, comparing several populations
throughout Europe. All but one population
were situated in lowlands, the only
exception being a population above
Grindelwald in the Swiss Alps. The Dutch
herpetologists marked the animals with
toe-clipping and analysed age by skeletochronology
(Hemelaar et al. 1987).
Our study picked up where Hemelaar
left off in 1985, focused on the same
mountain population, and has continued
through the 2001 breeding season. Here
we examine aspects of 20 years of data
coming from this population.
STUDY AREA
The site of this study of an alpine
Common Toad population is situated
above Grindelwald in the Bernese
Oberland, below the pass Grosse
Scheidegg. The altitude of the pass is
1960 m and the pond, at 1850 m, is only
a few meters from the road and easily
accessible. The pond measures about 30
x 10 m and the maximum depth is
approximately one metre. The vegetation
consists of Potamogeton alpinus,
Sparganium angustifolium, Carex rostrata
and Carex fusca and has clearly
increased in cover over the last 20 years
(pers. obs.). Other resident amphibian
populations include Triturus alpestris,
Common Toad and a few Rana temporaria,
the eggs and larvae of which are
usually eliminated by the newts.
METHODS
As a basis (1982-1984) we had all the
individually marked toads of Hemelaar
(1988). The initial objective was to follow
the life histories of the marked toads
to the end.
We initially avoided individual marking
with toe clips because of the number of
toes (4) that needed to be removed.
Instead, we marked newly captured individuals
by removing one toe upon first
date of arrival, allowing the definition of
year classes (1985-1992). Later (1993-
2001) we used transponders to identify
individuals.
We visited the pond during the evening
every 4-5 days, starting each year at the
beginning of snow-melt. It was impossible
to find all toads during daylight
because most were well-hidden in the
Carex-vegetation. At dusk (9:30 to
10:00 p.m.) we started catching all animals
we observed. The length and the
weight of the animals were measured,
the toe-code or the current transponder
number recorded and first-capture animals
were fitted with a transponder. This
process generally lasted 2-3 hours.
RESULTS
Only several first results can be presented
here; the amount of results is huge,
and only a small part has been analyzed.
The recapture rate of the toads from
1985 on, marked individually in 1982-84
(143 males, 121 females), was generally
low in the females, never exceeding 9
females per year. The last female of this
group was present in 1993 (female no.
411, which came to breed 7 times within
11 years). The decrease of the males is
slower and more regular, starting with 53
recaptured males in 1985. The longest
living male (No. 418) was regularly present
at the breeding place for 16 years
(1983-1998)! In contrast to the females,
most of the long-lived males migrated
every year to the site of reproduction.
Table 1 presents the years of breeding of
those females of the group 1982-84
who were present at least four times. In
most cases a 2-year-cycle can be recognised.
In some cases the females arrived
in 2, and in rare cases 3, consecutive
years. In the cases of the females no.
195, 424 and 428 it is likely that they
were present in at least one year

GROSSENBACHER Biota 3/i-a, ac 45
Table 1. »Life story« of individually marked females which were present at least four
times at the breeding place. X = present at the breeding place, but not weighed; 74:
weight of a female before spawning; (55): weight of a female after having partially
spawned.
Female
No.
411
151
225
82
195
424
428
1982
X
X
1983
X
X
X
X
1984
X
X
X
1985
72
1986
68
61
50
between 1983 and 1987 or 1988 respectively,
but they escaped our control.
In 1982-84 the age of 108 males and 66
females was determined by skeletochronology
by Hemelaar et al. (1987).
The combination with our recapture
results enables us to present in Figure 1
the final age they reached. Most females
don't get older than 10-12 years. This is
the age of sexual maturity and they
come only once to breed; few of them
get older up to a maximum of 20 years.
The males' ages are much more scat-
1987
75
56
48
60
66
1988
76
63
59
70
64
1989
74
66
72
1990
74
64
70
1991
74
73
1992
67
(55)
1993
82 7x
6x
5x
4x
4x
4x
4x
tered; some of them reach sexual maturity
at 5-7 years, normally at 8-10 years,
but many get much older, with a maximum
of 25 years.
A very high proportion of the females
(62%) were present only once at the
breeding place (Figure 2); one quarter
came twice, almost 10% 3 times, a.s.o.
up to 7 times. One third of the males
were present only once at the breeding
place. The number of multiple breeders
in males is irregularly scattered over the
scale up to 16 times. 25 males (17% of
Figure 1. Final age of 108 males and 66 females for which the age was determined in
1982-84.
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Age in years

46 Biota 3/i-a, 2002 GROSSENBACHER
Figure 2. Percentage of males and females present once, twice, three times a.s.o. at the
breeding place.
50
40
20
PI n „
5x 7x 8x 9x 10x
all individually followed males) were
present in 8 to16 consecutive years. The
mean value of presence for females is
1.5 times, for males 3 times.
DISCUSSION
Most of the males with determined age
originated from the two years 1973 and
1974, and most of the females with
known age from 1972 (Figure 3).
Hemelaar et al. (1987) considered the
age structure of the spawning populations
as unstable. This means a pulsation
of the age classes occurs along the time
axis, depending on the time span since
the last year with high offspring production.
It seems that in several years before
1972 very few or no offspring survived
(the age classes after 1974 were too
young to be sexually mature in 82-84).
When Hemelaar et al. (1987) analysed
the population in 1982-84, they found a
relatively young population with no
males older than 13 and no females
older than 18. Several of these animals
got much older than these maximum
presence at breeding place
ages in the following 10-14 years. An
analysis of the age structure of the same
population in the early 90s would have
come to a very different result, with a set
of old animals, both females and males,
around 20 years old. Thus an age analysis
of a toad population always depends
on the space of time since the last year
with a high recruitment. The reproductive
success of such populations seems to
vary from one year to the next to a very
high extent. In many years the production
is nearly zero, as in our case in the
late 60s until 1971. Otherwise, at least a
few old animals should have survived
until 1982-84. We observed a similar
phenomenon during the 90s, with no
year evidencing high recruitment. It is
not yet clear if this is a normal fluctuation
or the beginning of a population decline.
Because of the long delay between birth
and maturity (males 8-10, females 11-12
years) a decline would only be observable
in adults more than a decade later.
Another interesting point is the fact that
the years of high recruitment are differ-

GROSSENBACHER Biota 3/i-a, 2002 47
Figure 3. Year of birth for males and females for which the age was determined in 1982-
84.
40
35
30
25
20
15
10
5
0
d males
• females
1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980
ent in males and females. 1972 (best
year for females) was a cold summer
with temperatures below the average
from April to October and a very cold
September. In 1973 (best year for males)
the summer was much warmer, with a
very warm August and September. 1974
(good year for males) was again rather
cold with the exception of a very warm
August. August is normally the time of
metamorphosis; in cold years it is
September.
According to Piquet (1930), Common
Toad produces a higher proportion of
females under cold water conditions, in
warmer water a higher proportion of
males. One possible explanation for our
observations is a temperature-dependent
sex determination. But this can only
be considered as an indication in this
direction. Other explanations, such as a
different survival rate, are possible too.
The result that 2/3 of all females breed
only once is not surprising. Kuhn (1994)
observed similar or higher rates of mortality.
In the lowlands of England the
annual mortality was much higher with
70-85% (Gittins et al. 1985, Reading
1989). But in contrast to these studies,
we never found dead females in the
pond. We presume that overwintering
stress is responsible for this high mortality.
The cause of the much higher life
expectancy (20 years) in the mountains
than in lowlands (Kuhn 1994: 9 years)
can be found in the much shorter activity
time per season which prevails at the
altitude of 1850 m a.s.l., between 3.5
and 5 months. In accordance with this
higher life expectancy it is not surprising
that some females breed more often (6,
even 7 times) than ever found in the
lowlands (Kuhn 1994: 5 times). The difference
is rather small because a 2-yearcycle
is more frequent in the mountains
than in the plain. A detailed analysis of
the relation between 1-year- and 2-yearcycles
and a possible change of it will be
presented in a later paper; more data
must be collected. What is really new
and astonishing about these current
observations is the extremely old age
and long presence at the breeding place
of several males. In most other studies
males did not reach an age of 10 years,
and a presence of 8 times or more at the

48 3/1-2, 2OO2 GROSSENBACHER
breeding place is consequently impossible.
Because of extended longevity of toads
in higher altitude populations, an analysis
of possible changes in the population
structure cannot yet be made. A/lore
research is necessary until the age class-
es marked individually by transponders
between 1993 and 95 have reached the
end of their lifespan. Even after 20 years
of population marking, many questions
are still open, and this long-term study
must go on.
Acknowledgements
Thanks to A. Hemelaar for beginning the study and sharing data, to Peter Pearman,
who helped me improving this manuscript, and to many colleagues and friends who
helped me in the field over this long period, especially Silvia Zumbach and Beatrice
Luscher.
REFERENCES
GITTINS, S.P., KENNEDY, R.I., WILLIAMS, R. 1985: Aspects of a population age-structure of
the common toad (Bufo bufo) at Llandrindod Wells Lake, mid-Wales. Brit. J.
Herpetol., London 6: 447-449.
HEMELAAR, A. 1988: Age, growth and other population characteristics of Bufo bufo from
different latitudes and altitudes. Journal of Herpetology 22/4: 369-388.
HEMELAAR, A., CLAESSEN, V., WIJNANDS, H. 1987: Enkele karakteristieken van een
voortplantingspopulatie van de gewone pad (Bufo bufo) uit het gebergte van
Zwitserland. Lacerta45:130-139.
KUHN, J. 1994: Lebensgeschichte und Demographie von Erdkrotenweibchen Bufo bufo
bufo (L.). Zeitschrift fur Feldherpetologie 1: 3-87.
PIQUET, J. 1930: Determination du sexe chez les batraciens en fonction de la temperature.
Rev. Suisse Zool. 37:173-281.
READING, CJ. 1989: Annual fluctuations in common toad (Bufo bufo): breeding numbers
at a pond in southern England (1980-1889). Abstracts First World Congress of
Herpetology, 11-19 September 1989, University of Canterbury, United Kingdom.

GUICKING, JOCER & WINK Biota 3/1-2,2002 49
Molecular Phylogeography of the
Viperine Snake Natrix maura and
the Dice Snake Natrix tessellata:
first results
Daniela GUICKING1, Ulrich JOGER2,
Michael WINK1
1lnstitut fur Pharmazeutische Biologic, Im Neuenheimer Feld 364, 69120 Heidelberg,
Germany
!Hessisches Landesmuseum, Zoologische Abteilung, Friedensplatz 1, 64283 Darmstadt,
Germany
E-mail: daniela.guicking@urz.uni-heidelberg.de
Abstract
First results on phylogeographic patterns, based on complete sequences of the mitochondrial
cytochrome b gene, are presented for Viperine Snake Natrix maura and Dice
Snake Natrix tessellata. Three major phylogeographic clades are identified in N. maura.
The phylogenetically oldest groups of haplotypes are found in Tunisia/Sardinia, and in
Morocco; a third clade comprises all European specimens. Genetic distances (uncorrected
p-distances) are 3.9% to 4.6% between major clades, but do not exceed 1.4%
between different European populations. Due to higher genetic diversity in southern
parts of the European range, it is suggested that the Iberian peninsula served as a glacial
refugium, from where France and Northwestern Italy were colonized during postglacial
times.
In N. tessellata, genetic differences between distinct populations reach 8.2%. Major
clades correspond to an apparently relict population in central Greece, to populations in
Egypt/Jordan, Kasakhstan, Eastern Turkey, Crete and most of Europe. Within Europe
the Balkan region is suggested to have served as a glacial refugium during the
Pleistocene. Apparently, a few lineages colonized central Europe in postglacial times and
gave rise to the populations found today from Romania to Germany. Low but clear
genetic differences between Italian and central European populations suggest that Italy
was colonized by a distinct genetic lineage during the late Pleistocene and remained isolated
from central European populations. These first results are consistent with general
hypotheses on phylogeographic patterns and postglacial colonization routes in
European animal and plant species.
Key words: phylogeography, intraspecific diversity, Natrix maura, Natrix tessellata,
cytochrome b
Received 28 September; accepted 26 October 2001

50 Biota 3/i-a, 2002 GUICKING, JOGER & WINK
INTRODUCTION
Intraspecific differentiation of a species is
a complex outcome of geographic,
demographic and ecological factors that
have operated throughout its evolutionary
history (Walker & Avise 1998). It
should be particularly apparent in taxa
that show only limited mobility. As this is
generally true for reptiles, most studies
of intraspecific variability in these vertebrates
provide evidence for the existence
of distinct lineages or morphotypes that
can be well correlated to geographic
regions (e.g. Emys orbicularis: Fritz 1996,
Lenk et al. 1999, Natrix natrix: Thorpe
1984a, Thorpe 1984b, reviewed in
Kabisch 1999; Elaphe obsoleta: Burbrink
et al. 2000). Designation of subspecies is
the traditional way to account taxonomically
for high intraspecific diversity and
to distinguish lineages that differ by morphology,
coloration, genetics and geographic
distribution.
The genus Natrix has recently been confined
to the water snakes of the western
palaearctic. It comprises four species:
Viperine Snake Natrix maura, Dice Snake
N. tessellata, Grass Snake N. natrix, and
Big-head European Grass Snake N.
megalocephala. However, the species
status of N. megalocephala, which is
restricted to a small area east and northeast
of the Black Sea (Orlow & Tunijew,
1999), is controversial, due to its similarity
to N. natrix (Bohme 1999). Molecular
phylogenetic data of the genus Natrix
suggest that N. maura is the most ancestral
of the three main species. N. tessellata
and N. natrix are closely related and
form the more derived sister group to N.
maura. A detailed study on this topic will
soon be published by R. Lawson et al.
Of the three main species, N. natrix
shows the broadest ecological tolerance
and the largest distribution area. It
occurs throughout southern and central
Europe into England and Scandinavia,
throughout great parts of Siberia east-
wards to Lake Baikal, in the Middle East
and in the most northerly parts of Africa
(Kabisch 1999). N. maura and N. tessellata
are ecologically more restricted and
occupy similar ecological niches. Both
species are highly dependent on aquatic
habitats and need warmer climates than
N. natrix. Accordingly, their distribution
areas do not extend as far north (see Fig.
2). N. maura is found in northwestern
Africa, on the Iberian peninsula, in
southern France and in northwestern
Italy (Schatti 1999). The range of N. tessellata
extends from Italy throughout the
Balkan, Middle East, southern Asia into
China, and includes northern Egypt
(Gruschwitz et al. 1999). Isolated small
populations of Dice Snakes are found in
western Germany and in the Czech
Republic. The only sympatric populations
of Dice Snake and Viperine Snake are
found in northwestern Italy and at Lake
Geneva in Switzerland. However, at Lake
Geneva only N. maura is autochthonous,
whereas N. tessellata was introduced by
man (Mebert 1993).
Intraspecific variability has been studied
to some extent at least in all three
species, but only in N. natrix have different
subspecies been described to
account for the high diversity (Thorpe
1984a, Thorpe 1984b, reviewed in
Kabisch 1999). In N. maura no subspecies
are currently recognized
although morphological studies have
found some differences in pholidosis and
colouration between geographically distant
populations. However, these differences
appear too small and too inconsistent
to justify designation of subspecies
(Schatti 1999). In N. tessellata two subspecies
are distinguished. The small population
of the Romanian Black Sea island
of Serpilor is separated as N. t heinrothi
(Hecht 1930) from the nominate form N.
t. tessellata, which includes all other
populations of the Dice snake. According
to Gruschwitz et al. (1999), the sub-

GUICKING, JOGER & WINK Biota 3/i-a, 2002 51
species concept of N. tessellata needs to
be revised.
We have chosen a phylogeographic
approach based on molecular data to
investigate the intraspecific differentiation
of Viperine Snake and Dice Snake:
1) to find out whether intraspecific
genetic differentiation exists, 2) whether
genetic clades correlate to geographic
regions, and 3) whether the results can
be used to reconstruct a possible
microevolutionary history for the species
in Europe. In the present contribution we
describe first results based on mitochondria!
cytochrome b DMA sequences.
AAATERIAL AND METHODS
132 Dice Snakes and 80 Viperine Snakes
were sampled covering great parts of
their distribution ranges. As a source of
total genomic DNA, blood samples taken
from the caudal vein of living snakes, tissue
samples collected from dead animals
or museum material, and shed skins
were used. DNA was isolated from small
aliquots of the samples following standard
proteinase k and phenol chloroform
protocols (Sambrook et al. 1989).
Polymerase chain reaction (PCR) was
performed to amplify the target
sequence using the primers L14724NAT
(5'-GAC CTG CGG TCC GAA AAA CCA-
3') and H16064 (5'-CTT TGG TTT ACA
AGA ACA ATG CTT TA-3'; Burbrink et al.
2000) situated in the two RNA-genes
flanking the cytochrome b gene in reptiles.
PCR was performed in 50 ul volume
containing 0.75 units of Amersham
Pharmacia Biotech Taq polymerase, 0.2
mM of each dNTR 50 mM KCI, 1.5 mM
MgCI2, 0.5% Triton x-100, 10 mM Tris-
HCI (pH 8.5) and 0.01% BSA. 10 pmol
of each primer and 50-100 ng of the
template DNA were used. After an initial
denaturing step for 4 minutes at 94° C,
31 cycles were performed with denaturing
50 seconds at 94° C, annealing 50
seconds at 52° C, and primer extension
90 seconds at 72° C. A final extension
step of 5 minutes at 72° C followed.
PCR products were sequenced directly
with the dideoxy chain termination
method (Sanger et al. 1977) using the
Cycle Sequencing Kit (Amersham
Pharmacia Biotech, RPN 2438/RPN
2538) in combination with fluorescently
labelled primers. For cycle sequencing
primers mt-b2 (5'-GCC CAG AAm GAT
ATT TGT CCT CA, modified from Kocher
et al. 1989), NATf (5'-ACT CAG ATA
TyG ATA AAA TCC C-3'), eNAT (5'-TAG
GCA AAT AGr AAG TAT CAT TCT GG-
3'), and mt-le (5'-TCA AAC CCG AAT
GAT ACT TCC TAT T-3') were used. An
initial denaturing step at 94° C for 3 minutes
was followed by 26 cycles at 94° C
(30 sec) and at 55° C (60 sec).
Fluorescently labelled fragments were
analysed on an automated Sequencer
(Amersham Pharmacia Biotech, ALF-
Express II). The sequences were aligned
manually. Complete cytochrome b gene
sequences (1117 nt) were used for this
study.
Phylogenetic analyses were performed
using the program packages MEGA
(Kumar et al. 2001) and PAUP version
4.0b8 (Swofford 2001). For outgroup
rooting in N. maura the cytochrome b
sequence of Nerodia fasdata (kindly provided
by R. Lawson) and in N. tessellata
sequences of N. maura were used.
Maximum Parsimony analyses were conducted
with the heuristic search
approach of PAUP using the tree-bisection-and
reconnection swapping algorithm.
Bootstrap analyses were performed
with the heuristic search
approach. For comparison, Neighbor
Joining and preliminary Maximum
Likelihood trees with estimated parameters
were calculated.

52 Biota 3/i-a, 2002 GUICKING, JOGER & WINK
RESULTS AND DISCUSSION
Natrix maura
Phylogenetic analyses
Neighbor Joining, Maximum Parsimony
and preliminary Maximum Likelihood
analyses yielded similar phylogenetic
trees. The parsimony approach found
eight shortest trees for N. maura phy-
logeny. The strict consensus tree is
shown in Figure 1. The intraspecific comparison
of different cytochrome b gene
haplotypes for N. maura yielded 103
variable sites (9.22% of all sites), of
which 82 (7.34%) are parsimony informative.
99% or 100% bootstrap support was
Figure 1. Natrix maura phylogeny: Strict consensus cladogram of 8 most parsimonious
trees; lengths 295 steps (Cl = 0.8407, HI = 0.1593, Rl = 0.9226). Nerodia fasciata was
used for outgroup rooting. Bootstrap values (500 replicates) of more than 50% are indicated
at bifurcations for major lineages. Numbers following the localities indicate more
than one identical sequence from the same location.
Natrix
maura
99
100
100
58
97
78
87
r Italy Ponte Organasco
L Italy Voghera 2
Switzerland Lake Geneva 6
Spain Olot 2
Spain Olot 2
Spain Olot 2
Spain Olot
France Cote d'Azur
Switzerland L'Allondon
France Lac Bourget
Italy Varzi
Italy Sta. Maigheriia
Italy Sta. Margherita 2
Spain Amposta
France Camargue
France Niort 3
Spain Pyrenees 2
Spain Galicla 2
France Dpt Herault
France Belcastell
France Lac Salagon 2
Spain Delia de L'Ebre S
p Spain Extremadura
1 *- Spain Extremadura
Spain Extremadura
Spain Extremadura 3
Spain Extremadura
I Spain Extremadura
Portugal Higuera
r- Spain Extremadura
f^- Spain Extremadura
"1— Spain Benissa
'— Spain Extremadura
_r France Lac Salagon 2
Spain Galicia
Spain Extremadura
Spain Benissa
Tunisia Tamerza
Sardinia 2
Tunisia An Nafidah
Tunisia Nabul
Morocco Tabounaht 4
Morocco Haut Atlas
Morocco Haul Atlas 2
Morocco Haut Atlas
Morocco Tabounaht
Morocco Tabounaht 2
Morocco Moyen Atlas
Nerodia fasdaia
"European
Clade"
"Tunisian
Clade"
"Morrocan
Clade"

GUICKING, JOGER & WINK Biota 3/i-a, 2002 53
Figure 2. Location of sample sites and major mtDNA clades in Natrix maura (open symbols)
and Natrix tessellata (black symbols). The dashed line indicates the distribution
range of N. maura in the west and N. tessellata in the east. Different symbols represent
different genetic clades, in N. maura: square: Moroccan clade, triangle: Tunisian clade,
circle: European clade; in N. tessellata: rhomb: Egyptian clade, star: loannina clade, triangle:
Turkish clade, square: Kazakhstan clade, cross: Crete clade, circle: European
clade. Fat lines suggest approximate boundaries between mtDNA clades.
obtained for three strictly monophyletic
mtDNA clades of Viperine Snakes (Figure
1). The three genetic clades correspond
well to geographic locations of the samples
(Figure 2). One clade is comprised
of Viperine Snakes from Tunisia and
Sardinia, one clade is located in
Morocco, and the third clade comprises
all European specimens.
The two clades from northern Africa are
apparently phylogenetically older than
the European clade. However, current
data are not unambiguous in the positioning
of the two African clades.
Depending on the algorithm and optimality
criterion used, either the Tunisian
clade or the Moroccan clade appears to
be more ancestral. Bootstrap values of a
little less than 50% for either of the two
possibilities further indicate that the positioning
of these clades relative to each
other cannot be reliably inferred from
the cytochrome b data set.
Close genetic relationship between
Sardinian and Tunisian Viperine Snakes
suggest that the Viperine Snake was
introduced to Sardinia by man. This has
already been suggested by Schatti
(1999) due to morphological similarities
of individuals from the two localities.
Haplotypes and pairwise genetic comparisons
Pairwise genetic distances are given as
interclade and intraclade comparisons in
Table 1. The three major clades are separated
from each other by genetic distances
of 3.9-4.6%. Intraclade distances
reach 1.34% between different
European samples. In all instances, the
interclade differences are greater than
the intraclade distances, indicating that
the three major clades represent independent
evolutionary lineages.
According to the concept of the molecular
clock, genetic distances can provide
information on the divergence time of
two distinct lineages. With an assumed
evolutionary rate for snake mtDNA of
1.3% per one million years (de Queiroz

54 Biota 3/i-a, 2002 GUICKING, JOGER & WINK
Table 1. Interclade and maximum intraclade genetic distances (p-distances) in a) Natrix
maura and b) Natrix tessellata, based on complete sequences of the mitochondrial
cytochrome b gene.
a) Natrix maura
Glade "Tunisian" "Moroccan" "European"
"Tunisian" 0.45%
"Moroccan" 3.9-4.5% 0.90%
"European" 3.9-4.6% 3.9-4.6% 1.34%
b) Nafrix tessellata
Glade "loannina"
"loannina" 0.27%
"Egyptian" 8.0-8.1%
"Kazakhstan" 7.8-8.2%
"Turkish" 7.4-8.1%
"Crete" 7.9-8.1%
"European" 7.3-8.1%
et al. in press, adopted from Macey et al.
1998, who obtained sequence divergence
data for some agamid lizards), we
can cautiously estimate the divergence
time of the three major clades in N.
maura to some 3 to 3.5 million years
ago, i.e., in the middle/late Pliocene.
Larger sample sizes within the European
clade allow some preliminary statements
on the microevolutionary history of the
Viperine Snake in Europe. Intraclade
genetic distances of 1.34% or less suggest
that radiation of the European
ancestor into present haplotypes started
about one million years ago and corresponds
well to the major Pleistocene
glaciation cycles. Especially the last 700
000 years were dominated by major
glacials with a roughly 100 000 year
cycle that alternated with relatively short
interglacials (reviewed in Hewitt 1996).
The climatic fluctuations of the
Pleistocene are known to have greatly
influenced the distribution of intraspecific
polymorphism in cold-sensitive species
"Egyptian" "Kazakhstan" "Turkish"
0.27%
7.3-7.7%
7.2-7.8%
7.3%
6.9-7.3%
2.42%
3.7^.9%
6.6-6.8%
6.4-7.0%
3.13%
5.6-6.8%
5.6-7.0%
"Crete" "European"
0%
2.96% 0.80%
of European flora and fauna (Hewitt
1996, Taberlet et al. 1998). Pleistocene
climatic conditions caused extinction of
northern populations and contraction of
range to southern refugia during cold
periods as well as northward expansion
from refugia during subsequent warmings
(Hewitt 1996, Taberlet et al. 1998).
Such processes imply successive bottlenecks
and a loss of genetic diversity in
northern populations, whereas highest
genetic diversity is expected in southern
refugia (Taberlet et al. 1998). The main
Pleistocene refugia in Europe were located
on the Iberian peninsula, in Italy and
in the Balkans (e.g. Emys orbicularis:
Lenk et al. 1999). Postglacial colonization
of central Europe originated from
the Balkan refugium or from the Iberian
peninsula, whereas northward expansion
of Italian populations in most species
was prevented by the barrier of the Alps
(Taberlet et al. 1998).
In N. maura, highest genetic diversity -
indicated by distinct haplotypes occur-

GUICKING, JOGER & WINK
ring in the same area (e.g. in the
Extremadura) or identical haplotypes
occurring at geographically distant localities
- is found on the Iberian peninsula,
suggesting this as the main Pleistocene
refugium. Samples from the northeastern
part of the distribution range, that is
northwestern Italy, Switzerland, the Cote
d'Azur and Olot in Spain, on the other
hand, share closely related haplotypes
(Figure 1). These regions were apparently
colonized by a single lineage after the
last glaciation.
Matrix tessellata
Phylogenetic analyses
In N. tessellata, six major phylogenetic
clades can be distinguished, independent
of the calculation method. The strict
consensus tree of 96 most parsimonous
trees is shown in Figure 3. Within the
ingroup, 218 variable sites (19.5% of all
sites) were found in 1117 nucleotides of
the cytochrome b gene. 183 (16.4%) of
these are parsimony-informative.
The two most ancestral clades correspond
to samples from Egypt/Jordan and
to an apparently isolated population
from central Greece, Lake loannina
(Figures 2, 3). Most samples from
Turkey, together with all samples from
Armenia, Azerbaijan, Georgia and
Kazakhstan, form another clade, which is
further subdivided and therefore is treated
here as two different clades, the
Kazakhstan clade and the Turkish clade.
The best represented group corresponds
to samples from Europe, except for those
from Lake loannina. Because of high
genetic distances between European
mainland samples and samples from the
island of Crete (Table 1), we separate the
Crete samples as a clade of their own.
Approximate geographic boundaries
between different clades are indicated in
Figure 2.
The existence of a population with
ancestral haplotypes at Lake loannina is
55
one of the most surprising results of this
study. Several scenarios can be drawn to
explain the parapatric occurrence of different
haplotypes in Greece. Both lineages
might have descended from a common
ancestor that inhabited Europe several
million years ago and was then biogeographically
separated into different
populations which evolved independently
afterwards. The existence of a phylogenetically
ancestral haplotype within
the distribution area of the more derived
European haplotypes might further suggest
that Europe was inhabited by a different
genetic lineage of Dice Snakes in
the far past. This scenario would assume
that the early European Dice Snakes
became almost entirely extinct, with the
only descendants of that lineage today
being left at Lake loannina, whereas the
present European haplotype has
descended from a lineage that colonized
Europe much later. A third possible
explanation is that the population at
Lake loannina derives from an anthropogenetically
introduced population.
However, this possibility seems less likely,
because our data do not suggest any
relationship of the loannina haplotype to
haplotypes found elsewhere, and the
region of Lake loannina is known to
inhabit endemic lineages of several other
organisms as well (P. Kyriakopoulou-
Sklavounou, pers. cornm.).
Haplotypes and pairwise genetic comparisons
Genetic distances between the six clades
are 3.0-8.2% (Table 1), suggesting independent
evolution of the most distant
lineages since the early Pliocene.
Intraclade genetic distances are smaller,
reaching a maximum of 2.42% and
3.13% in the two deeply subdivided
clades of Kazakhstan and Turkish samples
(Tab. 1). Samples from Crete are
separated from mainland European samples
by genetic distances of up to

56 3/1-2, 2OO2 GUICKING, JOGER & WINK
Figure 3. Matrix tessellata phylogeny: Strict consensus cladogram of 96 most parsimonious
trees; lengths 517 steps (Cl = 0.6716, HI = 0.3284, Rl = 0.9197). N. maura
sequences were used for outgroup rooting. Bootstrap values (100 replicates) greater
than 50% are indicated at bifurcations for major lineages. Numbers after the location
indicate more than one identical sequence from the same location.
71
81
|— 99
100
p Italy Lazio 4
*- Italy Toscana
p Switzerland Lago di Lugano
*- Switzerland Lago di Lugano
85 (~ Hungary Balaton
*— Hungary Balaton
S4 ,f Romania Tulcea
— P- Romania Tulcea 3
*— Romania Tulcea
_ . . _
r Turkey Sarkale 5
J""1- Georgia Batumi
lOpJLr Turkey Malatya
j I *- Armenia Eranus
n\ Azerbaijan Kubatly
92
nn r-T TurKey EsMK>9a
LZZ-P" Turkey Yenioaga 5
|_r Turkey Sinop
^ Turkey Catalzeytin
Turkey Biracik
innrT Kazakhstan Alma-Ata
98 f^P- Kazakhstan Alma-Ata 2
tesseflata 10°
' Georgia Agara
r Eavut
*- Jordan Jarash 2
* «« r Greece Lake loannina
1UU P- Greece Lake loannina
|_r Greece Lake loannina
"- Greece Lake loSnnina
r~~ Nattix maura France Lac Bourget
fj— Natrix maura Italy Ponte Organasco
a
' — • Matrix maura Spain Extremadura
2.96%; the divergence time of the two
lineages could therefore be estimated at
a little more than two million years ago,
i.e., at the transition of Pliocene and
Pleistocene.
Within the European Clade, genetic distances
of 0.80% suggest that radiation
"European
Clade"
"Crete Clade"
"Turkish
Clade"
"Kazakhstan
Clade"
"Egyptian Clade
"loannina
Clade"
into present haplotypes occurred during
the middle and late Pleistocene.
Phylogenetic relationships are, however,
only weakly resolved. Three apparently
independent genetic lineages can be distinguished:
one comprises samples from
Lake Balaton in Hungary, one comprises

GUICKING, JOGER & WINK
samples from Tulcea at the mouth of the
Danube river in Romania, and one
includes the Italian and Swiss samples. A
fourth lineage that is resolved by
Neighbor Joining and Maximum
Likelihood analyses, but does not appear
in the Maximum Parsimony approach,
includes all samples from Germany, the
Czech Republic and some samples from
Bulgaria and Romania. The Balkan region
most likely served as a Pleistocene
refugium, from where the different lineages
postglacially invaded the northern
parts of the present distribution area.
A close relationship between Italian and
Balkan specimens indicate that Italy did
not play a major role as a glacial
refugium in the Dice Snake, but was
rather colonized only during the late
Pleistocene. Genetic distinction of Italian
from central European Dice Snakes
apparently reflects the barrier function of
the Alps.
Conclusions
Our data show a remarkable intraspecific
genetic subdivision of Viperine Snake
and Dice Snake. Both species are composed
of several monophyletic mtDNA
clades that can be well correlated to geographic
regions. Lack of co-existence of
major haplotypes and large genealogical
gaps between most clades suggest long-
Biota 3/i-a, Z, 2OO2 57
term extrinsic barriers to gene flow
(Walker & Avise 1998). Phylogeographic
data from European specimens provide
the first information on the microevolutionary
history of the two species in
Europe. To complement the existing
data, we would like to include further
sample sites, especially from underrepresented
regions and possible suture zones
of different lineages. We are also planning
to include nuclear markers in our
study (ISSR-PCR genetic fingerprints, see
Wink et al. 2001) to study possible introgression
between adjacent lineages.
Finally, the data presented here provide
evidence for the existence of distinct
genetic lineages that have evolved independently
over several million years,
both in Viperine Snake and Dice Snake.
Genetic distances between different
clades are comparatively high and lie
within the range of distances between
closely related species and subspecies in
other reptiles (Johns & Avise 1998, de
Queiroz et al. in press). Therefore, our
data provide a first basis for a taxonomic
revision of the two species. Distinction
of different lineages, at least at the subspecies
level, seems desirable to account
for the high intraspecific variability.
However, morphological studies need to
be performed to accomplish such a taxonomic
revision.
Acknowledgements
We are very grateful to all those who have supported and still support our study by collecting
or providing sample material. These are: T. Amann, P. Lenk, H.-P. Eckstein, P.
Godlay, A. Pieh, X. Bonnet, S. Ursenbacher, E. Razzetti, A. Hille, M. Gruschwitz, S. Lenz,
A. Herzberg, W. Fiedler, A. Sproll, A. K. Smole-Wiener, B. Gautschi, D. Modry, M. A. L.
Zuffi, N. Godetsch, G. Mantziou, N. Rastegar-Pouyani, R. Griffiths, D. Ristow, N. Orlov,
S. Tome, A. Kapla, M. Vogrin, U. Utiger and many others. Samples from the mouth of
the Ebro river in Spain were provided by X. Santos and collected with permission from
the "Direccio General del Medi Natural, Generalitat de Catalunya" no. 6539
(26.10.1990). The "Struktur- und Genehmigungsdirektion Nord" in Koblenz, Germany,
gave permission to collect samples of German Dice Snakes. We thank R. Lawson for his
constant help with emerging problems and for providing unpublished sequences. A.
Rasmussen and M. Ivanov will probably note that we tried to keep in mind some of
their helpful critiques given at the SEH meeting in Slovenia. Our study is financially sup-

58 Biota 3/1-2, 2002 GUICKING, JOGER & WINK
ported by the Deutsche Forschungsgemeinschaft (JO-134/7 and WI-719/18).
REFERENCES
BOHME, W. 1999: Handbuch der Reptilien und Amphibien Europas, Bd. 3/IIA (Schlangen
II). Aula-Verlag, Wiesbaden: 482.
BURBRINK, F. T., LAWSON, R. & SLOWINSKI, J. B. 2000: Mitochondrial DMA phylo- geography
of the polytypic North American Rat Snake (Elaphe obsoleta): A cri-tique
of the subspecies concept. Evolution 54: 2107-2118.
De QUEIROZ, A., LAWSON, R. & LEMOS ESPINAL, J. A. 2002: Phylogenetic relation- ships
of North American Garter Snakes (Thamnophis) based on four mito chondrial
genes: How much DNA sequence is enough. A/lolekular Phylogenetics and
Evolution 22: 315-329.
FRITZ, U. 1996: Zur innerartlichen Variabilitat von Emys orbicularis (Linnaeus, 1758). 5b.
Intraspezifische Hierarchie und Zoogeographie (Reptilia: Testudines: Emydidae).
Zoologische Abhandlungen Staatliches Museum fur Tierkunde Dresden 49: 31-
71.
GRUSCHWITZ, M., LENZ, S., MEBERT, K. & LANKA, V. 1999: Matrix tessellata (Laurenti,
1768) - Wurfelnatter. In: W. Bohme. Handbuch der Reptilien und Amphibien
Europas, Bd. 3/IIA (Schlangen II). Aula-Verlag, Wiesbaden: 581- 644.
HECHT, G. 1930: Systematik, Ausbreitungsgeschichte und Oekologie der europaischen
Arten der Gattung Tropidonotus (Kuhl) H. Boie. Inaugural-Dissertation, Univ.
Berlin.
HEWITT, G. M. 1996: Some genetic consequences of ice ages, and their role in diver gence
and speciation. Biological Journal of the Linnean Society 58: 247-276.
JOHNS, G. C. & AVISE, J. C. 1998: A comparative summary of genetic distances in the vertebrates
from the mitochondrial cytochrome b gene. Molecular Biology and
Evolution 15: 1481-1490.
KABISCH, K. 1999: Natrix natrix (Linnaeus, 1758) - Ringelnatter. In: W. Bohme. Handbuch
der Reptilien und Amphibien Europas, Bd. 3/IIA (Schlangen II). Aula-Verlag,
Wiesbaden: 513-580.
KOCHER, T. D., THOMAS, W. K., MEYER, A. EDWARDS, S. V., PAABO, S., VILLABLANCA,
F. X. & WILSON, A. C. 1989: Dynamics of mitochondrial DNA evolution in animals:
Amplification and sequencing with conserved primers. Proceedings of the
National Academy of Sciences, USA 86: 6196-6200.
KUMAR, S.r TAMURA, K., JAKOBSEN, I. B. & NEI, M. (2001): MEGA: Molecular
Evolutionary Genetics Analysis, Version 2.0. Pennsylvania State University, PA,
USA.
LENK, P., FRITZ, U., JOGER, U. & WINK, M. 1999: Mitochondrial phylogeography of the
European pond turtle, Emys orbicularis (Linnaeus 1758). Molecular Ecology 8:
1911-1922.
MACEY, J. R., SCHULTE, J. A., ANANJEVA, N. B., LARSON, A., RASTEGAR-POUYANI, N.,
SHAMMAKOV, S. M. & PAPENFUSS, T. J. 1998: Phylogenetic relationships
among agamid lizards of the Laudakia caucasia species group: testing hypotheses
of biogeographic fragmentation and an area cladogram for the Iranian
Plateau. Molecular Phylogenetics and Evolution 10: 118-131.
MEBERT, K. 1993: Untersuchung zur Morphologie und Taxonomie der Wurfelnatter Natrix
tessellata (Laurenti 1768) in der Schweiz und im sudlichen Alpenraum.
Diplomarbeit, Univ. Zurich.
ORLOW, N. L. & TUNIJEW, B. S. 1999: Natrix megalocephala Orlow und Tunijew, 1987 -
GroBkopf-Ringelnatter. In: W. Bohme. Handbuch der Reptilien und Amphibien
Europas, Bd. 3/IIA (Schlangen II). Aula-Verlag, Wiesbaden: 505- 512.

GUICKING, JOCER & WINK >ta 3/1-2, 2002 59
SAMBROOK, J., FRITSCH, E. F. & MANIATIS, T. 1989: Molecular Cloning: A laboratory manual.
Cold Spring Harbor Laboratory, Cold Spring Harbor, New York.
SANGER, F., NICKLEN, S. & COULSON, A. R. 1977: DNA sequencing with chain-ter- minating
inhibitors. Proceedings of the National Academy of Sciences, USA 74: 5463-
5467.
SCHATTI, B. 1999: Natrix maura (Linnaeus, 1758) - Vipernatter. In: W. Bohme. Handbuch
der Reptilien und Amphibien Europas, Bd. 3/IIA (Schlangen II). Aula-Verlag,
Wiesbaden: 483-503.
SWOFFORD, D. L. 2001: PAUP*. Phylogenetic Analysis Using Parsimony (*and Other
Methods), Version 4.0b8. Sinauer Associates, Sunderland.
TABERLET, P., FUMAGALLI, L., WUST-SAUCY, A.-G. & COSSON, J.-F. 1998: Comparative
phylogeography and postglacial colonization routes in Europe. Molecular Ecology
7: 453-464.
THORPE, R. S. 1984a: Geographic variation in the Western grass snake (Natrix natrix helvetica)
in relation to hypothesized phylogeny and conventional subspecies.
Journal of Zoology, London 203: 345-355.
THORPE, R. S. 1984b: Multivariate patterns of geographic variation between the island and
mainland populations of the eastern grass snake (Natrix natrix natrix). Journal of
Zoology, London 204: 551 -561.
WALKER, D. & AVISE, J. C. 1998: Principles of phylogeography as illustrated by fresh water
and terrestrial turtles in the southeastern United States. Annual Reviews in
Ecology and Systematics 29: 23-58.
WINK, M., GUICKING, D. & FRITZ, U. 2001: Molecular evidence for hybrid origin of
Mauremys iversoni Pritchard et McCord, 1991, and Mauremys pritchardi
McCord, 1997 (Reptilia: Testudines: Bataguridae). Zoologische Abhandlungen
Staatliches Museum fur Tierkunde Dresden 51: 41-49

HAILEY & LAMBERT Biota 3/1-2,2002 61
Comparative growth patterns in
Afrotropical giant tortoises
(genus Geochelone)
Adrian HAILEY1 & Michael R.K. LAMBERT2*
'Department of Biological Sciences, University of Zimbabwe, PO Box MP167, Mount
Pleasant, Harare, Zimbabwe
Present address: School of Biological Sciences, University of Bristol, Woodland Road,
Bristol BS8 1UG, United Kingdom
E-mail: adrian.hailey@bristol.ac.uk
"Corresponding author: Natural Resources Institute, University of Greenwich at
Medway, Central Avenue, Chatham Maritime, Kent ME4 4TB, United Kingdom
Present address: Environmental Initiatives, Lydbrook House, Upper Lydbrook,
Gloucestershire GL17 9LP, United Kingdom
E-mail: lambertmrk@aol.com
Abstract
Growth to huge sizes is characteristic of giant tortoises. Growth data for leopard tortoises
Geochelone pardalis in eastern and southern Africa, the continental Sahelian
giant tortoise Geochelone sulcata, and the insular Aldabran giant tortoise Geochelone
gigantea are reanalysed, and their growth patterns compared. Geochelone pardalis
shows great body size variation, with size in northern populations (Republic of
Somaliland) comparable to that of insular giant tortoises. All species and populations fitted
a two-phase growth curve: an initial asymptotic growth period followed by slow
indeterminate linear growth. Asymptotic growth was non-Bertalanffy in all groups,
most individuals fitting Gompertz or logistic-by-mass models best Juvenile growth (for
ages 3-8) was also approximately linear. Variation in populations and species was in
juvenile growth rate, asymptotic size, rate of subsequent indeterminate linear growth,
and old age survival rate. Juvenile growth rates and asymptotic sizes of G. pardalis from
Somaliland resembled other populations, with large mean size due to high survival rates
and continued linear growth after the asymptote. Juvenile growth rates and indeterminate
linear growth in Geochelone sulcata resembled those of G. pardalis, but larger
asymptotic sizes were reached. Geochelone gigantea had much higher juvenile growth
rates than the continental species, and asymptotic size and rate of subsequent indeterminate
linear growth were lower in a high-density (Grande Terre) population - with size
achieved by G. sulcata similar - than a low-density (lie Malabar) one. Juvenile growth
rate was unaffected by population density, and relative to larger tortoises was higher in
the Grande Terre population, the only group to approach Bertalanffy-type growth.
Key words: Africa, asymptotic size, Geochelone, growth curve, growth rate, tortoise
Received 1 September 2001; accepted 23 February 2002

62 Biota 3/1-2,2002 HAILEY & LAMBERT
INTRODUCTION
Growth to huge sizes is perhaps the most
striking feature of giant tortoises. There
are five Afrotropical species; two are on
the mainland continent of Africa. The
leopard tortoise Geochelone pardalis has
an extensive north-south range in eastern
and southern E. & S. Africa, while the
African spurred or Sahelian giant tortoise
Geochelone sulcata has an east-west
range across the Sahelian region (Iverson
1992).
Geochelone pardalis in populations of
Somalia's North-West Zone (Republic of
Somaliland) (hereafter referred to as
Somaliland) have been shown to be substantially
larger than in those further
south, especially between the Equator
and Tropic of Capricorn. Size frequencies
indicate that on average, Somaliland tortoises
even exceed those in the eastern
Cape Province, South Africa (Lambert et
al. 1998), where the largest individuals
of this species have been recorded
(Branch et al. 1990).
Geochelone pardalis in Somaliland experience
wet summers, with copious green
vegetation available as food, and cool
dry winters when refuge is sought and
activity is minimal. A not dissimilar climatic
regime prevails also in the Sahelian
region of Africa, from Mauritania and
Senegal in the west to Eritrea and
Ethiopia in the east, coincident with the
range of C. sulcata, the world's largest
mainland tortoise. The size of this
Sahelian tortoise was approached by
large individual G. pardalis in Somaliland.
Large size could thus be due to the similarity
of climatic conditions, which may
be particularly favourable for growth in
both the Sahelian region and northernmost
Somaliland.
The main aim of this work, therefore,
was to compare growth of G. pardalis
tortoises in Somaliland with that of G.
pardalis from other regions in E. & S.
Africa, and also with Sahelian G. sulcata.
In particular, this was to show whether
large size in Somaliland was due only to
a high growth rate, which could be purely
environmental, or to a different
growth pattern, which is more likely to
reflect genetic differences associated
with speciation.
This alludes to an important question of
whether growth differences in tortoises
are phenotypic or genetic. Growth of
tortoises is not suited to experimental
analysis of this problem, because many
differences are based on the timing of
reduced growth at the asymptote, and
would require study over several decades
to obtain the answer. Any interspecific
differences, such as between G. pardalis
and G. sulcata, must have a genetic
basis. The possible importance of phenotypic
effects can be judged by comparison
of closely-related populations in different
ecological conditions.
Neighbouring island populations of the
Aldabra giant tortoise Geochelone
gigantea on Aldabra Atoll, Indian Ocean,
differ markedly in population density,
and so growth in this species was examined
in two such populations based on
data in Bourn & Coe (1978). Geochelone
gigantea is one of three extant giant tortoise
species inhabiting islands in the
Indian Ocean, but the only one with
substantial populations remaining.
Studies of growth in G. pardalis have
been made by several investigators in
Somaliland and elsewhere in E. & S.
Africa (Lambert 1995); Serengeti
National Park, N.E. Tanzania (Lambert et
al. 1998), and in the Sengwa Wildlife
Research Area, W. Zimbabwe (Hailey &
Coulson 1999).
An outline of main results is given in this
account; data in more detail have been
published elsewhere (Hailey & Lambert
2002).
METHODS
Growth measurements, based on growth

HAILEY & LAMBERT 1-2, 2O02 63
rings, were made on G. pardalis in four
regions of eastern and southern Africa,
on G. sulcata in the western Sahel, and
on C. gigantea on two islets of Aldabra
Atoll, Indian Ocean.
The straight-line size measurement (in
mm) used for all three tortoise species,
either from direct measurement or calculated
(Bourn & Coe 1978, Lambert 1993,
Lambert 1995, Lambert et al. 1998,
Hailey & Coulson 1999), was the midline
straight carapace length (intermarginal
notch to supracaudal scute). Tortoises
were also weighed to within 0.5 kg up to
50 kg, and 1 kg above 50 kg. For consistency,
a single relationship between mass
(in g) and length was used for all populations
of C. pardalis.
The growth pattern in chelonians has
been described by a variety of mathematical
models. As adopted by Hailey &
Coulson (1999), a number of FORD-
WALFORD plots may be used to fit
BERTALANFFY, GOMPERTZ, logistic by
length or logistic by mass growth curves.
This provides a simple method for distinguishing
between major types of growth
model, and thus the growth pattern in
animals. FORD-WALFORD plots are
especially useful in chelonians, with
growth histories represented by yearly
deposition of keratinous rings, and
depend on the availability of a large
number of growth increments to give a
significant negative regression for one or
more growth curves. Excluding the possibly
incomplete outer growth ring of all
tortoises, only individuals with 10+ increments
(i.e. with 11+ usable growth rings
and 12+ in total) were used.
Following Hailey & Coulson (1999), and
using a common size measure, tortoise
populations were compared by means of
FORD-WALFORD graphical plots to
determine the growth patterns. These
methods were applied to G. sulcata to
show whether the large G. pardalis from
Somaliland had growth more similar to
this species (with which it has a slightly
overlapping geographical range in S. E.
Ethiopia; Lambert 1999) or to other populations
of G. pardalis.
Asymptotic size was estimated from the
intercept on the x axis of the best-fitting
significant negative regression of a
FORD-WALFORD plot on BERTA-
LANFFY, GOMPERTZ or logistic by
length or mass axes.
The BERTALANFFY plot may in some
cases give a significant positive slope
with few data points, where only the initial
increasing phase of growth is represented.
This is not a fit to the BERTA-
LANFFY curve as such, since this requires
decreasing growth rate to an asymptotic
size. A positive slope only would give the
nonsensical result of continuously accelerating
growth to an infinite size.
RESULTS
The Leopard Tortoise Geochelone
pardalis
For G. pardalis, the BERTALANFFY
model did not give the best fit for any
tortoise in any of the four regions.
Tortoises from Somaliland and Serengeti
mostly fitted GOMPERTZ (17 individuals)
or logistic by mass (15 individuals)
models best, with the remaining seven
individuals fitting the logistic by length
model best. Tortoises from E. & S. Africa
mostly fitted the GOMPERTZ model best
(15/19 individuals), while most of those
from Sengwa fitted the logistic by mass
model best (20/23 individuals).
For all regions in which G. pardalis were
measured in sub-Saharan Africa, growth
increments showed an early growth rate
of 10-15 mm/yr in years 1-2. Compared
to Somaliland tortoises, the growth rate
of museum E. & S. Africa tortoises
increased only slightly, to about 15
mm/yr, although the variation in growth
rate with age was significant (ANOVA,
F1W56 = 1.89, P < 0.02) over the years 1 -
16. The relatively flat growth pattern of

64 Biota 3/1-2,2002 HAILEY & LAMBERT
E. & S. Africa tortoises explains why the
GOMPERTZ model fits these animals
best. In other regions, after an increasing
rate of growth, the rate declined in older
animals where there were data available.
For Serengeti and Sengwa tortoises,
asymptotic size in males and females was
significantly different, and so growth
increments were examined separately for
the sexes in these two regions.
From growth curves constructed for G.
pardalis from different regions in sub-
Saharan Africa, growth in museum specimens
from E. & S. Africa appeared to be
much slower. However, differences in
growth in the other three regions were
small. After 12 years, the divergent
growth data among these regions is due
to small and changing sample sizes.
Asymptotic sizes in G. pardalis from,
respectively, Sengwa, Serengeti, E. & S.
Africa and Somaliland were fairly consistent,
with ascending means of 295 ±
S.D. 33, 324 ± S.D. 63, 337 ± S.D. 117
and 344 ± S.D. 113 mm, and there was
no significant difference between them.
Initial sizes at growth ring 0 (hatchling
size), with ascending means of 31 ± S.D.
6, 45 ± S.D. 7, 47 ± S.D. 8 and 55 ± S.D.
4 mm in, respectively, E. & S. Africa,
Somaliland, Serengeti and Sengwa tortoises,
did, however, differ significantly
in the four regions (F3,?9 = 52.3,
P

HAILEY & LAMBERT Biota 3/i-a, 2002 65
taken up to 27 years for the lie Malabar
population and 17 years for the population
on Grande Terre. Growth curves for
the two C. gigantea populations, one of
low density (lie Malabar) and the other
of high density (Grande Terre), were
compared with that of G. sulcata.
Juvenile growth rate of G. gigantea over
years 3-6 did not differ between low
density lie Malabar and high density
Grande Terre populations. Juvenile
growth rate at 37.6 mm/yr was almost
twice the mean of 21.6 ± S.D. 7.9
mm/yr for G. pardalis.
Growth ring increments for G. gigantea
over the first 17 years were also used to
fit FORD-WALFORD plots, lie Malabar
mean data fitted the GOMPERTZ curve
best, and the BERTALANFFY curve least
well; asymptotic size was 613 mm.
Grande Terre data also fitted the GOM-
PERTZ curve best, and the BERTA-
LANFFY curve least well, but asymptotic
size at 470 mm was much smaller.
Because of reduced adult growth, but
normal juvenile growth in the dense
population, the Grande Terre tortoises
are indeed the only ones to which the
BERTALANFFY curve provides a reasonable
fit, although the GOMPERTZ curve
is the best fit.
DISCUSSION AND CONCLUSIONS
All species and populations of the
Afrotropical tortoises under discussion
fitted a two-phase growth curve, with an
initial period of asymptotic growth up to
about 20 years followed by indeterminate
linear growth. Asymptotic growth
was non-BERTALANFFY in all populations,
and juvenile growth was approximately
linear over years 3-8.
Adult size of G. gigantea from the dense
Grande Terre population was similar to
that of G. sulcata, and to old Somaliland
G. pardalis. Thus all can be grouped as
»giant tortoises*, but they achieve huge
sizes in different ways. The differences lie
in the rate of juvenile growth, the
asymptotic size reached, the rate of linear
growth after this, and the duration of
this period of slow growth.
Changes in these variables accounted for
all the growth curves observed:-
* Slow juvenile growth - G. pardalis in E.
& N. Africa
* Fast juvenile growth - G. gigantea on
Aldabra.
* Higher asymptote than G. pardalis - G.
sulcata in the western Sahel, and G.
gigantea on Aldabra, especially on lie
Malabar
* Faster overall growth than other tortoises
- G. gigantea on lie Malabar
* Longer-continued late linear growth of
old animals - G. pardalis in Somaliland.
The high rate of growth of G. gigantea
may be due to a longer activity season
with tropical maritime conditions on
Aldabra compared to mainland Africa.
Geochelone gigantea on lie Malabar
were active throughout the year, while
adults on Grande Terre, due to the lack
of feeding opportunities in shade, were
inactive for no more than two months
during the dry season. In contrast, G.
pardalis at Sengwa and G. sulcata in the
western Sahel were inactive for five
months of the dry season.
The reduced body size of tortoises in the
dense Grande Terre population, compared
to those on lie Malabar without
the same ecological constraints, is a
proven case of environmental effects.
With possible access to a different range
of herbaceous food species from adults,
juveniles on Grand Terre appear to be
shielded from competition with adults.
The high growth rates in early years give
a closer approximation to a BERTA-
LANFFY-type growth curve than in any
other species or population.

66 Biota 3/i-a, 2002 HAILEY & LAMBERT
Acknowledgements
Cost of travel and subsistence for M.R.K.L to attend the 11th Ordinary General Meeting
of Societas Europaea Herpetologica in 2alec (near Celje), Slovenia, 13-17 July 2001,
was covered by the Higher Education Funding Council of England (HEFCE) through
approval of the Natural Resources Institute, University of Greenwich at Medway
(Chatham), who also provided time for meeting participation, preparation of an oral
presentation and manuscript writing for the proceedings volume.
REFERENCES
BOURN, D. & COE, M. 1978. The size, structure and distribution of the giant tortoise population
of Aldabra. Philosophical Transactions of the Royal Society London (B)
282:139-175.
BRANCH, W. R., BAARD, E. & DE VILLIERS, A. 1990. Some exceptionally large south- ern
African chelonians. Journal of the Herpetological Association of Africa 37: 53-54.
HAILEY, A. & COULSON, I. M. 1999. The growth pattern of the African tortoise
Geochelone pardalis and other chelonians. Canadian Journal of Zoology 77:
181-193.
HAILEY, A. & LAMBERT, M.R.K. 2002. Comparative growth patterns in Afrotropical giant
tortoises (Reptilia Testudinidae). Tropical Zoology 15: 121-139.
IVERSON, J. B. 1992. A revised checklist with distribution maps of the turtles of the world.
Privately printed, Richmond, Indiana, USA.
LAMBERT, M. R. K. 1993. On growth, sexual dimorphism, and the general ecology of the
African spurred tortoise, Geochelone sulcata, in Mali. Chelonian Conservation
and Biology 1: 37-46.
LAMBERT, M. R. K. 1995. On geographical size variation, growth and sexual dimor- phism
of the leopard tortoise, Geochelone pardalis, in Somaliland. Chelonian
Conservation and Biology 1: 269-278.
LAMBERT, M.R.K. 1999. On conservation of the Sahelian giant tortoise, Geochelone sulcata,
pp. 255-261. In: Miaud, C. & Guyetant, R. (eds.). Current studies in herpetology,
Le Bourget de Lac (SEH) [Proceedings of the 9th Ordinary General
Meeting of the Societas Europaea Herpetologica, 25-29 August 1998, Le Bourget
du Lac, France]. SEH, Le Bourget du Lac, France: 255-261.
LAMBERT, M. R. K., CAMPBELL, K. L. I. & KABIGUMILA ,J. D. 1998. On growth and morphometrics
of leopard tortoises, Geochelone pardalis, in Serengeti National Park,
Tanzania, with observations on effects of bushfires and latitudinal varia- tion in
populations of eastern Africa. Chelonian Conservation and Biology 3: 46-57.

JOVANOVIC, 9URIC & AAARKOVIC BJOta 3/1-2, 2002 67
Tertiary reptiles of the central
part of the Balkan peninsula
Miodrag JOVANOVIC, Dragana DURIC &
Zoran AAARKOVIC
Natural History Museum, Njegoseva 51, 11 000 Beograd, Yugoslavia
E-mail: geopal@ptt.yu
Abstract
This paper includes a list, with short remarks, of all fossils of Tertiary reptiles discovered
in the period 1896-2001 in the central part of the Balkan Peninsula, composed of
Serbia, Republic of Srpska and FRY Macedonia. The total number of discovered Tertiary
reptile fossils in the central part of the Balkan Peninsula is about 55 taxons in 22 localities.
First recordings are noted for chelonians of family Chelidrydae, lizards of genus
Ophisaurus and family Gerhonotidae, Mediterranean snakes of family Colubridae, large
subtropic snakes of genus Vipera, and certain crocodilians.
Key words: Tertiary, chelonians, lizards, snakes, crocodilians, Serbia, Republic of Srpska,
Macedonia.
Received 9 January 2002; accepted 15 July 2002

68 3/1-2, 2002 JOVANOVIC, BURIC & MARKOVIC
INTRODUCTION
A first, but incomplete, list of Tertiary
reptiles discovered on the territory of former
SFR Yugoslavia (whose eastern half
we here present under the name of the
central part of the Balkan Peninsula),
was published by Paunovic (1983). From
1983 to the present (2001), our knowledge
of Tertiary reptiles in this part of
Europe has become both significantly
different and larger. Within the period
1985-2001, Jovanovic, the curator of
the Natural History Museum and a biologist,
has surveyed all older accessible
paleoherpetological material from the
collections of the Natural History
Museum in Belgrade and other institutions
in Serbia. These results were published
in several papers (Jovanovic 1989,
1990, 1995a, b). From 1990-1999,
Markovic, a geologist and curator of the
Natural History Museum, discovered by
fieldwork a whole array of new localities
with remains of Tertiary reptiles. Since
1999, Jovanovic and Markovic have
been joined in paleoherpetological
research by Novakovic, a biologist and
curator of the Natural History Museum,
who initiated revision of all former findings.
This work has resulted in this, the
most complete list of all fossil finds of
Tertiary reptiles in the central part of the
Balkan Peninsula.
METHODS
The authors of this paper have revised
older available fossil material and preliminarily
identified new fossil material.
They also provide comments on previous
identifications and reasons behind new
determinations, wherever necessary. In
cases when cited fossil specimens are
known only from literature data, the
authors did no revision but gave their
opinion in comments. This is explicitly
stated for all such fossil remains and
those which are known only from credible
oral reports.
Many fossils cited in this paper could not
be determined to species level but only
to genus, family or some higher taxonomy
units. However, regardless of this
fact, all these remains were treated as
real species, in both a biological and a
paleontological sense. For the specimens
that are presently kept in the
Paleontological collection of the natural
history Museum in Belgrade, that was
not explicitly stated since it was understood.
All material was found together with
mammal fossils, and therefore is dated
according to MN zones (Steininger
1999).
LIST OF LOCALITIES AND RECORDED
SPECIES
EOCENE
Stip (FRY Macedonia):
In this locality, an almost complete shell
of a turtle (Temnoclemmys mazedonica
Pasic & Klincarski 1959) was found. This
fossil may be part of the palentological
collection in the Skopje Museum of
Natural History. This is not only the oldest
known record of turtles in
Macedonia, but also the oldest known
record of any Tertiary reptile on the territory
of the central Balkans.
EARLY MIOCENE
Djacki potok. Aleksinac (Serbia) MN1:
Several almost complete turtle shells
were found in vicinity of the Aleksinac
coal mine (Stevanovic 1969). It is not
known where these fossils are kept
today.
Kraljevo village. Aleksinac (Serbia) MN1:
Three vertebrae, approximately five cm
long (Crocodilia gen. et sp. indet), were
collected by R. Stevanovic.
Dubrava coal mine. Aleksinac (Serbia)
MN1:
According to Stevanovic (1969), at
Djacki Potok and in the Dubrava coal

JOVANOVIC, BURl£ & MARKOVIC Biota 3/1-2,2002 69
Table 1. List of Tertiary fossils with their localities and ages in MN zones (for explanation
see text)
fossil
Testudines gen. et sp. indet
Temnoclemmys mazedoniaca
Trionyx sp.
Emys sp.
Mauremmvs serbica
Testudo sp.
Testudo kalksburgensis
Geomyda sp.
Testudinidae gen. et sp. indet
Chelidridae gen. et sp. indet.
Reptilia gen.et sp. indet.
Crocodylia gen.et sp. inded.
Crocodylus mormiensis
Aligarotinae gen.et sp. indet
Tomistoma egsenburgensis
Sauna geaet sp. indet.
Squamata gen.et sp. indet
Vipera sp.
Vijjeridae gen.et sp. indet.
Colubridae gen.et sp. indet
Alethinophidia gen.et sp. indet
Natrix cf. natrix
Natrix sp.
Coluber cf. viridiflaviis
Coluber cfgemonensis
Elaphe cf.aitatorlineata
Elaphe cf. longissima
Elaphe sp.
Chameleontidae gen.et sp. indet
Anguidae gen.et sp. indet
Lacertidae gen.et sp. iadet.
Lacerta sp.
Anguis sp.
Pseudapits sp.
Ophisaums sp.
?Gerrhonotinac gen.et sp. indet
E
1
1
1
MW
2,3
4
i
1
1 3,19
1
1
1
1 5
1
MM4
List of localities:
1 - Stip, (FRY Macedonia)
2 - Djacki potok, Aleksinac (Serbia)
3 - Okno Dubrava, Aleksinac (Serbia)
4 - Bogovina (Serbia)
5 - Ugljevik (Republik of Srpska)
6 - Jankova Klisura (Serbia)
7 - Sibnica (Serbia)
8 - Milicevo brdo, Krusevica (Serbia)
9 - Jama AAorava (Serbia)
10 - Popovac (Serbia)
11 - Prebreza, Blace (Serbia)
12 - Lazarevac, Trstenik (Serbia)
mine several well preserved turtle
(Testudines) shells were found, with horn
plates and visible markings. Besides the
shells, specimens also included preserved
limbs. Shell lengths were about 20 cm. It
6
7
7
7
MN5
9
3
9
20
mammal zones
MN
MN6
7+8
21
11,13
13
10
13
10
11
11,12
10,21
10
10 1
10
1
22
14
14
14
14
14
14
MN
mn
15
IS
IS
15
15
MN13 MN14
16
1
17
17
17
17
17
17
17
17
17
17
17
MN16
13 - Lestane (Serbia)
14 - Vracevic, Monastery Bogovadje
(Serbia)
15 - Beluska and Prevalac, Veles
(FRY Macedonia)
16 - Grocka (Serbia)
17 - Kamenjak, Bukulja (Serbia)
18 - Beocin (Serbia)
19 - Village Kraljevo, Aleksinac
(Serbia)
20 - Mala Miliva, Despotovac (Serbia)
21 - Zitni potok, Prokuplje (Serbia)
22 - Village Crvca, Levac (Serbia)
is not known where the specimens are
kept today. At the same locality limb
bones of crocodilians (Crocodylia gen. et
sp. indet) as well as parts of joints and
one vertebra were found (Stevanovic P.
IS

70 Biota 3/i-a, 2002 JOVANOVIC, BURIC & MARKOVIC
viva voce, Stevanovic 1969).
Bogovina (Serbia) MN1:
Parts of a turtle (Trionyx sp.) shell were
found in the Bogovina coal mine
(Laskarev 1949). It is not known where
the specimens are kept today.
Ugljevik (Republic of Srpska) MN1:
Lignite layers in this Neogene lake in
Dinarides have originated in equivalents
of ottnangian (Vujnovic et al 2000). The
bottom of Ugljevik lake was gently
descending during the accumulation of
coal, and its water was, at least in one
part of the period, free-flowing (Krstic
2000). Fauna of the lake and its shores
was, according to our findings, abundant
in amphibians, birds and mammals that
were subtropical or tropical in character.
Of paleoherpetofauna, remains of the
lower jaw of a lizard were found. The
teeth resemble those of the genus
Anguis or Ophisaurus.
Jankova Klisura (Serbia) MN4:
Osteoderms of crocodilians (Crocodylia
gen. et sp. indet) were found in a coal
mine (Pavlovic & Djurkovic 1962). It is
not known where the specimens are kept
today.
Sibnica (Serbia) MN4:
This locality has been known since 1967
(Petronijevic 1967). Of the fossils found,
several were identified: fragments of
intermaxillar bones and vertebras of
some lizards (Sauria gen. et sp. indet.), a
fragment of the lower jaw of a
chameleon (Chamaeleontidae gen. et sp.
indet), and a thoracic vertebra of a snake
(Vipera sp.).
MIDDLE MIOCENE
Milicevo brdo. Krusevica (Serbia) MN5:
In this locality, part of a tortoise (Testudo
sp.) shell was found (Laskarev, 1936). It
is not known where the specimen is kept
today. This fragment of a Testudo is not
only the first find of this tortoise in our
region but also the first registered find of
any Tertiary reptile in Serbia.
Jama Morava. Despotovac (Serbia)
MN5:
There are literature data from this locality
(Petronijevic 1967), recorded as part
of a turtle shell (Testudines) and a tooth
of a small reptile. These fossil remains are
presently kept at the Geology faculty. As
there is neither picture not description,
and the place where the specimens are
kept is unknown, any kind of revision is
impossible.
Mala Miliva. Despotovac (Serbia) MN5:
Fossil osteoderms were found belonging
to certain Anguidae.
Zitni Potok. Macina. Prokuplje (Serbia)
MN6:
In this locality plastron fragments
(Testudines gen. et sp. indet) and a tooth
of a crocodilian (Crocodilia) were found
at a depth of 11m in the sandy-gravel
sediments.
Popovac (Serbia) MN6:
Several fossils of various reptiles were
found in a cement marl mine in Popovac.
Among them, turtle remains are the
most important, and include plastron
fragments of Mauremmys serbica
Jovanovic, 1995b and carapace fragments
of a tortoise (family Testudinidae).
Overall, the most important findings in
reptile paleofauna in the territory of the
central Balkans are crocodilian remains:
Tomistoma eggenburgensis Toula & Kail
1885, part of a jaw with teeth (Pejovic
1951). It is not known where the specimen
is kept today. The fossil fragment of
crocodilian jaw that was identified by
Pejovic through odontological analysis
could not be located by the authors of
this paper, so revision could not be done.
There is a possibility that identification
done by Pejovic is inadequate, as according
to Antunes (1987) and Antunes &
Ginsburg (1989), fossil species of genus
Tomistoma can hardly be determined
just by teeth.
Crocodylus moraviensis Jovanovic 1995a
part of a skull with jaws. The holotype is
kept at the local Museum in Paracin.

I
JOVANOVld, DURIC & MARKOVIC 3/1-2, 2OO2 71
Crocodylidae gen. et sp. indet, several
fragments of skull bones, teeth, osteoderms,
etc. (Jovanovic 1989),
Crocodylus sp indet, fragment of premaxilla
and nasal bones with preserved
nasal opening. This fossil is presently
kept at the Museum in Paracin. As this
fragment shows a very convex rostrum,
it may be assumed that it belonged to
some crocodilian with long jaws, perhaps
a gavial.
Aligatorinae indet, fragments of upper
and lower jaw with teeth (Jovanovic,
1995a). The fossil is presently kept in the
Museum in Paracin. The size of the angle
between the left and right branches of
the jaws enables us to conclude that this
fossil remain belonged to a crocodilian
with wide and short jaws, such as an alligator.
Prebreza. vicinity of Blace (Serbia) MN6:
Vegetation of the locality where reptiles
lived was of savannah type. Precipitation
was likely to be rare and sporadic. Also
sporadic were savannah bushfires that
drove animals into lakes where they
drowned and later fossilized (Milosevic
1967).
Turtle remains from genus Trionyx sp
were identified. These were carapax
fragments belonging to one small and
one large turtle. They differ from each
other not only in size but also by morphology
of costalia (bone shape and
sculptured look of their upper surfaces).
Besides, these two bones were found in
two differently coloured and differently
granulated sandstone fragments. This
points either to possible paleoecological
differences in habitats of these turtles (if
the places where both specimens were
found were primary ones) or to the existence
of some greater time distance
between them. Because of morphological
and most probably paleoecological
differences, we have decided to treat
these remains as two different species.
Also in this locality, remains were found
of two plastrons of certain larger, most
probably aquatic turtles that are presently
kept at the paleontological collection
of the Natural History Museum in
Belgrade. The size, shape and morphology
of these plastrons are similar to those
of turtles in the Chelidridae family.
Turtle remains (plastron fragments, hand
and fossilized eggs) described in the
paper by Milosevic (1967) were determined
to the order level (Testudines);
however, there is some doubt regarding
Milosevic's taxonomic determination.
The plastron fragments were lost and
revision is impossible, while morphoanatomic
analysis of the fossil hand
remains (number of phalangae in the fingers
and size) show that it probably does
not belong to a chelonian. The »turtle
eggs« as they are called by Milosevic
(1967) are actually only their moulds,
almost without any traces of shell.
During the revision of all fossil material
brought earlier from Blace and additional
sieving of sediment, bird remains were
discovered. Egg size and distribution
within the sediment have repeatedly
reminded the authors of bird eggs within
a nest. As bird remains were found in
the same layer at the same locality, it is
probable that the »turtle eggs« from
Blace are actually a nest of eggs of some
Miocene water bird. The theory that fossil
eggs from Blace belong to a turtle was
also doubted by Mikhailov id his paper
on classification of fossil amniote eggs
(Mikhailov 1991).
Lazarevac. Trstenik (Serbia) MN6:
Horn plates remains of a reptile and few
osteoderms of undetermined Anguidae
were found in the village of Lazarevac in
2001. They are now kept in the Natural
History Museum in Belgrade.
Lestane (Serbia) MN6:
In this locality fragments of three chelonian
genera were found (Trionyx sp.,
Testudo sp., Emys sp.). The fossil
remains of turtles were found in the layer

72 Biota 3/i-a, 2002 JOVANOVIC, BURIC & MARKOVIC
of Lower Sarmate, but there is a possibility
that they were moved there from
some older, for example Badenian, sediments
(Jovanovic 1998).
Vracevici near Bogovadje Monastery
(Serbia) MN7+8:
This locality is situated at the place
reached by Paratethys during the mid
Miocene (Prisjazhnjuk et al 2000). The
locality was either a small lake or a
waterhole. According to the mammal
fossils also discovered here, it may be
concluded that the climate was moist
and warm, tropical or subtropical.
Several jaw fragments from the
Lacertidae family were found, as well as
osteoderms and skull parts of the
Anguinae subfamily or the
Gerrhonotinae subfamily and the vertebrae
of snakes (Colubridae and
Viperidae).
Village Crnca. Levac (Serbia) MN7+8:
Fragments of upper and lower jaws of
unknown reptiles (Sauria gen. et sp.
indet.) were collected in 1975 by the
geologist Dolic.
LATE MIOCENE
Beluska and Prevalec. vicinity of Veles
(Macedonia) MN11-J-12:
The area where the reptiles found here
lived was similar to the present-day
African savannah. Around the lake in the
valley there was a lot of grass and bushes.
Vegetation was present mostly
around water basins, lakes and rivers
(Pavlovic & Markovic 1990).
Among the fossils found were tortoise
(Testudo sp, unknown where the specimen
is kept), lizard Lacerta sp. (Ciiric
1957 also not known where it is kept), as
well as the right mandible, left maxilla
and osteoderms of Pseudopus sp. Fossil
remains of this lizard from Veles were
described for the first time in a paper by
Pavlovic & Markovic (1990). In that
paper, of all the material found, only the
osteoderms were determined to be the
remains of a lizard Ophisaurus panonni-
cus, or as accepted here, Pseudopus
panonnicus.
The right mandible of a lizard from Veles
seems to have been discovered by
Pavlovic and Laskarev after WWI; however,
they neither identified it nor mentioned
it in reports. During later repeated
sieving of sediments, Markovic &
Pavlovic discovered the left maxilla and
several osteoderms. From an odontological
viewpoint, the mandible and maxilla
of the fossil Pseudopus sp from Veles are
very similar to each other and probably
belonged to two individuals of the same
species. On the other hand, the jaws and
teeth of Pseudopus sp from Veles differ
morphologically from those of Middle
European fossil Pseudopus pannonicus
as well as from the recent Mediterranean
Pseudopus apodus.
At the same locality, fragments of one
larger and one smaller snake vertebra
were found. The osteological characters
point out that these are vertebra of a
viper, most probably two different individuals,
and trunk vertebra of an
Alethinophidia. Based on collected material
and reports, during the excavations
in Veles, Pavlovic and Laskarev paid most
attention to larger fossil remains while
neglecting smaller ones. Therefore, their
finding and mentioning of reptile vertebra
(Reptilia gen. et sp. indet, PavlovL
1922) shows that it was large, that is,
belonging to a very large specimen of
lizard or snake. Also, two vertebra fragments
were found that were unusual in
shape (elongated and large, with broken
ventral parts) and determined to be
Reptilia gen.et sp. indet. As there was no
other material to compare them with,
they could not be defined more precise-
ly'
Grocka (Serbia) MN13: *
Part of an epiplastron was found
(Geoemyda sp.) This fragment is insufficient
to precisely define the turtle
species.

JOVANOVIC, BURld & MARKOVIC Biota 3/1-2,2002 73
EARLY PLIOCENE
Kamenjak (Serbia) MM 14:
This locality is very rich in various fossil
remains of herpetofauna. The following
items were identified: fragments of parietal
bones, jaws and vertebrae of Lacerta
sp., many vertebra of Anguis sp, osteoderms
of Ophisaurus sp, fragments of
right maxilla of Anguis or Ophisaurus as
well as many torsal vertebra of snakes
(Natrix sp., Coluber sp., Elaphe sp.,
Vipera sp.).
MIDDLE PLIOCENE
Beocin (Serbia) MN16:
In this locality, part of a turtle shell were
found and identified as Testudo kalksburgensisTou\a,
1896, (Mlynarski 1966).
The fossil remains of this tortoise are
kept at the Geological Institute in
Hungary. Testudo kalksburgensis from
Beocin is the first fossil find of a turtle
registered in Vojvodina.
DISCUSSION AND CONCLUSION
At the end of this paper, we must mention
that some of the fossil remains
determined here to the genus level will
be further described in later papers and
determined to the species level. Some of
them represent new species for science.
In some of the earlier papers it is mentioned
that almost all fossil remains of
reptiles in this part of Europe were found
accidentally, for example, while excavating
mammal fauna, and that they were
never considered very important.
However, reptiles and amphibians are
very good ecological indicators, that is,
they are susceptible to various climatic
changes (especially temperature and
humidity). Better knowledge of fossil
reptiles in the central Balkans would
surely enable us to solve a whole array of
paleoecological and paleogeographic
problems.
In the paper "Miocene Crocodilians of
Serbia," read at the third symposium on
fauna of Serbia (Jovanovic 1989), the
author shared the hypothesis that
remains of Tertiary reptiles should be
looked for at the same places where
remains of Tertiary mammals were
found, especially on the south exposures
of the localities. During the last several
years, Markovic has found remains of
small mammals together with fossils of
chelonians, lizards and snakes at almost
all visited Tertiary and Quaternary localities
in Serbia. This preliminary research
has already partially confirmed the
above-mentioned assumptions of the
authors.
This list includes various species of
Tertiary reptiles found in various Tertiary
localities in the central part of the
Balkans. However, the realistic number
of species of Tertiary reptiles that inhabited
this part of Europe was certainly tens
of times larger than those we have mentioned
(Jovanovic 1996 a, b, 1998).
Unfortunately, here we must repeat and
stress that only future, systematic paleoherpetological
research in the central
Balkans can give a realistic picture of the
Tertiary reptiles of this region.

74 Biota 3/1-2, 2002 JOVANOVld, 0UR|£ & MARKOVlC
Acknowledgements
The authors of this paper are especially grateful to our colleagues, mineralogist Voislav
Simic and geologist Gordana Jovanovic, for shared data, suggestions, and advice.
REFERENCES
ANTUNES, M.T. 1987: Affinities and Taxinomical Status of Miocene Longirostrine
Crocodilians from Western Europe with Remarks on Phylogeny, Paleoecology and
Distribution. Comun. Serv. Geol. Portugal 73: 49-58.
ANTUNES, M.T. & GINSBURG, L. 1989: Les Crocodiliens des faluns miocenes de I'Anjou.
Bull.Mus.natn.Hist.nat. 4 C: 2: 79-99.
ClRlC, A. 1957: Pikermijska fauna iz okoline Titovog Velesa. Glas. prir. muz. srp. zeml je, A
8: 1-82. In Serbian.
JOVANOVIC, G. 1998: Prilog poznavanju sarmata Lestana kod Beograda. Vesnik geozavoda,
serija A,B, Geologija, 48: 69-73rln Serbian.
JOVANOVIC, M. 1989: Krokodili miocena Srbije. Treci simpozijum o fauni SR.Srbije. Uvodni
referati i rezimei, 58. In Serbian.
JOVANOVIC, M. 1990: Tercijarni gmizavci Srbije: sadasnje stanje i buduca istrazivan ja. XII
kongres na geolozite na Jugoslavia. 1: 348-358. In Serbian.
JOVANOVIC, M. 1995 a: Crocodylus moraviensis, nova vrsta tercijarnog krokodila iz laporaca
Popovca kod Paracina (Serbia, Jugoslavia). Zapisnici SGD za 1990. i 1991.
37-38. In Serbian.
JOVANOVIc, M. 1995 b: Mauremys serbica nova vrsta slatkovodne kornjace iz tercijara
Popovca kod Paracina (Serbia, Jugoslavija). Zapisnici SGD za 1990 i 1991. 39- 43.
In Serbian.
JOVANOVIC, M. 1996 a: Fossil reptilia from the Popovac quare. Neogene of Central Serbia.
Neogene of Paratethys Yugoslav Working Group. JGCP, Project 329: 47.
JOVANOVIC, M. 1996 b: O diverzitetu i sudbini tercijarnih vodozemaca i gmizavaca centralnog
dela Balkanskog kopna. Desbilten, 2. 26-28. In Serbian.
JOVANOVIC, M. 1998: Majdan cementnih laporaca Popovca (Paracin, Serbia) - jedno od
najvaznijih nalazista miocenskih krokodila u Evropi. Zastita prirode 48-49, za
1995-1997: 351-353. In Serbian.
KRSTlC, N. 2000: Some ostracodes from the Dinaric Alps neogene and its paleoecolo gy.
Zbornik radova Medunarodnog simpozijuma geologija i metalogenija Dinarida i
Vardarske zone. ANU Republike Srpske. 1: 207-214. Banja Luka, Srpsko Sarajevo.
In Serbian.
LASKAREV, V. 1936: Miocenska fauna kicmenjaka iz okoline sela Krusevice (Bukulja). Geol.
anali Balk.pol. 13: 14-27. In Serbian.
LASKAREV, V. 1949: O novim nalascima fosilnih sisara kod Despotovca sa kratkim pre gledom
tercijarnih fauna sisara nadenih u Srbiji. Glas.SANU, 192: 45-62. In Serbian.
MIKHAILOV, K. E. 1991: Classification of fossil eggshells of amniotic vertebrata. Acta
Paleontologica Polonica 36: 193-238.
MILOSEVIC, V. 1967: O nalasku fosilnih jaja u miocenskim sedimentima okoline Blaca u
Toplici. Glas.Prir.Muzeja A, 22: 17-28. In Serbian.
MLYNARSKI, M. 1966: Die fosilen Schildkroten in den ungarishen Samlungen. Acta Zool.
Cracov. 11:223-288
PAUNOVlC, M. 1983: Prilog poznavanju rasprostranjenosti mezozojskih i kenozojskih
Amphibia i Reptilia u Jugoslaviji. Geol.vjesnik, 36: 79-89. In Croatian.
PAVLOVlC, G. & MARKOVlC, Z. 1990: Dopuna poznavanju pikermiske faune iz oko line
Titovog Velesa na osnovu novih nalazaka sitnih kicmenjaka. XII Kongres geologa
Jugoslavije. 1: 401-405. Ohrid. In Serbian.
PAVLOVlC, M. & BURKOVlC, R. 1962: Miocenski sisari iz rudnika "Jankova Klisura" u

JOVANOVld, 9URIC & MARKOVIC BJOta 3/i-a, 2002 75
Toplickom basenu (Serbia). Geol.anali Balk.pol., 29. In Serbian.
PAVLOVlC P 1922: Muzej srpske zemlje u 1921. g., Godisnjak SKA, 30: 1-23. In Serbian.
PASlC, M. & KLINCARSKI, V. 1959: Nalazak fosilne kornjace u sedimentima gornjeg eocena
okoline Stipa (Makedonija). Vesnik 27: 61-72. In Serbian.
PEJOVlC D. 1951: O nalasku vilice fosilnog krokodila u cementnim laporcima Popovca kod
Paracina. Zbornik radova SAND, 16: 103-107. In Serbian.
PETRONIJEVIC Z. 1967: Srednjomiocenska i donjosarmatska ("Stajerska") fauna sisara
Srbije. Paleontologica jugoslavica, 7: 1-157. In Serbian.
PRYSJAZHNJUK, V, KOVALENKO, V. & KRSTlC N. 2000: On the terrestrial and fresh- water
molluscs from neogene of western Serbia. Zbornik radova Medunaro dnog simpozijuma
geologija i metalogenija Dinarida i Vardarske zone. AND Republike
Srpske 1: 219-224. Banja Luka, Srpsko Sarajevo. In Serbian.
STEININGER, F. F. 1999: Chronostratigraphy, Geochronology and Biochronology of the
Miocene "European Land Mammal Mega-Zones" (ELMMZ)and the Miocene
"Mammal-Zones (MN-Zones)". In: Roessner G. & Heissig K. (eds.): The Miocene
Mammals of Europe. 9-24. F. Pfeil, Munchen.
STEVANOVIC, R. 1969: Stratigrafski prikaz miocenske ugljonosne serije aleksinackog basena.
Zapisnici SGD, (1964., 1965., 1966. i 1967.). 239-247. In Serbian.
VUJNOVIC, L. et al. 2000: Lacustrine neogene of the'-Dinarides. Zbornik radova Medunarodnog
simpozijuma geologija i metalogenija Dinarida i Vardarske zone. AND
Republike Srpske, 1: 197-206. Banja Luka, Srpsko Sarajevo. In Serbian.

Biota 3/1-2,2002 77
Herpetofauna of Round Island,
Mauritius
Zoltan KORSOS1 & Balazs TROCSANYI2
'Department of Zoology, Hungarian Natural History Museum, Baross u. 13, H-1088
Budapest, Hungary
E-mail: korsos@zoo.zoo.nhmus.hu
department of Conservation, Danube-Drava National Park Directorate, Tettye ter 9,
H-7625 Pecs, Hungary
E-mail:balazs.trocsanyi@ktm.x400gw.itb.hu
Abstract
Round Island, north of Mauritius in the Indian Ocean, is inhabited by eight species of
reptiles (no amphibians), five of which are endemic. The original vegetation of the
island was clear-cut in the 18th century. Introduced goats and rabbits were eliminated
relatively recently to protect the remnants of the palm savannah, then the whole island
was designated as a nature reserve. A Hungarian team visited the island in November
1999 and April 2001, to carry out herpetofaunal assessments related to all eight species.
Whereas no specimen of the Burrowing Boa was encountered, good populations of
Telfair's and Bojer's Skinks, as well as of the Ornate Day Gecko, could be estimated.
Several specimens of DurrelPs Night Gecko, the Keel-scaled Tree Boa, and 33 specimens
of Gunther's Gecko were observed.
Key words: Round Island, Mauritius, herpetofauna, endemic reptiles, conservation
Received 26 February 2002; accepted 27 July 2002

78 Biota 3/i-a, 2002 KORSOS & TROCSANYI
INTRODUCTION
The tiny, volcanic Round Island with its
151 hectares lies 20 km to the north of
Mauritius in the Indian Ocean. It is
inhabited by eight species of reptiles (no
amphibians), five of which are endemic,
occurring nowhere else in the world.
The original vegetation, hardwood
ebony and teak forests, was clear-cut in
the 18th century. Introduced goats and
rabbits were eliminated relatively recently
(in 1979 and 1986, respectively) to
protect the remnants of the palm savannah,
then the whole island was designated
as a nature reserve with extremely
limited access for scientific research only.
In 1989, a conservation management
plan was set up for the island to restore
its vegetation, to control the invasion of
alien species, and to monitor the population
changes of its native and endemic
flora and fauna.
The aim of this paper is to summarize in
brief the information presently available
on the remarkable herpetofauna of
Round Island, to give an illustrative list of
the eight known reptile species, and to
provide a comprehensive literature
(without references) to initiate further
herpetological studies. The scientific
results of our expeditions will be published
elsewhere.
MATERIAL AND METHODS
With the help of the local authorities, the
National Parks and Conservation Service
(NPCS) of the Ministry of Agriculture,
Food, Technology & Natural Resources,
and the Mauritian Wildlife Fund (MWF),
a Hungarian team visited the island in
November 1999, and later, with the support
of Fauna and Flora International, in
April 2001. Five and ten days were spent
on Round Island, respectively, the necessary
material being carried over by governmental
helicopters, following the
strict expedition protocol set up previously
by NPCS and MWF. There is no
permanent research station on the island
at present, so expedition members have
to bring everything themselves (incl.
food, fresh water, tents, sleeping bags,
etc.), and, similarly, have to remove all
waste material when leaving the island.
Usually there are four management
expeditions organized by NPCS every
year, with the main purpose being to
monitor plant reintroduction and vegetation
recovery.
Surveys and censuses of herpetofauna
were made every full day during the
early morning, late afternoon and
evening hours; midday was left out
because of the extremely hot period
when all reptiles appeared to rest.
Gunther's geckos were caught and
measured on all occasions if possible; the
same applied for Casarea dussumieri,
whereas other geckos and skinks
observed were only noted during the
census. In the second visit to the island,
special methods were used to rediscover
Bolyeria multocarinata; the results will be
discussed in another paper.
RESULTS AND DISCUSSION
Round Island reptiles include three geckos
(Gekkonidae), three skinks
(Scincidae), and two snake species (the
only known members) of the family
Bolyeriidae, according to the following
list (endemic forms are marked by an
asterisk):
Gekkonidae:
Gunther's Gecko Phelsuma guentheri
Boulenger, 1885*
Ornate Day Gecko Phelsuma ornata
Gray, 1825
Durrell's Night Gecko Nactus serpensinsula
durrelli Arnold & Jones, 1994*
Scincidae:
Telfair's Skink Leiolopisma telfairii
(Desjardins, 1831)*
Bojer's Skink Congylomorphus bojerii
(Desjardins, 1831)
Bouton's Skink Cryptoblepharus boutonii

KORSOS & TR6CSANYI BJOta Ti-a. 2002 79
(Desjardins, 1831)
Bolyeriidae:
Keel-scaled Tree Boa Casarea dussumieri
(Schlegel, 1837)*
Burrowing Boa Bolyeria multocarinata
(Boie, 1827)*
Whereas no specimen of the Burrowing
Boa was encountered (the last observation
of this species dates back to 1974)
during our two expeditions, relatively
good populations of Telfair's and Bojer's
Skinks, as well as of the Ornate Day
Gecko could be estimated. Several specimens
of Durrell's Night Gecko, the Keelscaled
Tree Boa, and 33 specimens of
Gunther's Gecko - exceeding the results
of all previous expeditions - were
observed and measured accordingly.
Bouton's Skinks, in accordance with the
former observations, seemed to be confined
to the rocky shoreline surfaces. The
Mauritian nature conservation authorities
put enormous efforts into the
restoration of the native vegetation of
Round Island, which could help to stabilize
the different reptile populations.
During the second expedition (April
2001) our aim was to search for evidence
of the presence or extinction of
the Burrowing Boa (Bolyeria multocarinata
(Boie, 1827)). This animal has been
seen only four times in the 20th century,
and since its last observation in 1974 no
signs of its survival have been discovered.
Because our searches, too, were
unsuccessful in this respect, we consider
this snake species to be extinct.
In our detailed reports (available through
the National Parks and Conservation
Service, Reduit, Mauritius) conclusions
were drawn about the status of and the
possible threats to the unique herpetofaunal
assemblage of Round Island. We
also proposed to the Mauritian nature
conservation authorities to organize
more regular and standardized surveys
to follow the changes in the different
reptile populations.
Acknowledgements
We would like to extend our sincere thanks to the following persons for their assistance
during our visit to the island: Youssef Mungroo, Vishnu Bachraz, Krishna Puttoo,
Shyamduth Ramrekha, Vishal Nundlaul (all from the Ministry of Agriculture, Food,
Technology & Natural Resources, National Parks and Conservation Service, Reduit,
Mauritius), Carl Jones and Ashok Khadun (Mauritian Wildlife Appeal Fund, Port Louis,
Mauritius). Our travel could not have been accomplished without the financial help of
Comp Travel Ltd. (Budapest, Hungary), the Fauna & Flora International (Great Britain),
and the Mauritius Travel & Tourist Bureau Ltd. (Floreal, Mauritius). Special thanks are
due to our wives, Zita Zachar & Agnes Varga, for helping and encouraging us in the field
(as well as before and after the expedition).
REFERENCES
ARNOLD, E. N. 1980: Recently extinct reptile populations from Mauritius and Reunion,
Indian Ocean. Journal of Zoology, London 191: 33-47.
ARNOLD, E. N. & JONES, C. G. 1994: The night geckos of the genus Nactus in the
Mascarene islands with a description of the distinctive population on Round
Island. Dodo, Jersey Wildlife Preservation Trust 30: 119-131.
BLOXAM, Q. & VOKINS, M. 1978: Breeding and maintenance of Phelsuma guentheri
(Boulenger, 1985) at the Jersey Zoological Park. Dodo, Jersey Wildlife
Preservation Trusts 15: 82-91.
BULLOCK, D. J. 1977: Round Island - A tale of destruction. Oryx 14: 51-58.
BULLOCK, D. J. 1986: The ecology and conservation of reptiles on Round Island and

80 Biota 3/1-2,2002 KORSOS & TROCSANYI
Gunner's Quoin, Mauritius. Biological Conservation 37: 135-156.
BULLOCK, D. & NORTH, S. 1975: Report of the Edinburgh University Expedition to Round
Island, Mauritius, July and August 1975. Unpublished report, University of
Edinburgh.
BULLOCK, D. & NORTH, S. 1984: Round Island in 1982. Oryx 18: 36-41.
DASZAK, P. 1994: The 1993 Raleigh International Round Island Expedition, including a survey
of intestinal parasites collected from animals of Round Island, Mauritius and
Rodrigues and observations on Round Island reptiles. Unpublished report,
Kingston University.
DULLOO, M. E., BULLOCK, D. J. & NORTH, S. 1996: Report of the expedition to Round
Island and Gunner's Quoin, Mauritius, July/August 1996. Unpublished report,
Mauritius Wildlife Foundation.
GARBUTT, N. 1992: The reptiles of Round Island, Mauritius. Herptile 17: 157-170.
GARBUTT, N. 1992: The reptiles of Round Island, Mauritius. Vivarium 4: 14-18, 32, 33.
GARBUTT, N. 1992: The Round Island gecko: A most unusual Phelsuma. Dactylus 1: 17-21.
JONES, C. G. 1988: Round Island boa eats Serpent Island gecko. Oryx 22: 180.
JONES, C. G. 1993: The ecology and conservation of Mauritian skinks. Royal Society of Arts
and Sciences, Mauritius 5: 71-95.
JONES, C. G. & HARTLEY, J. 1995: A conservation project on Mauritius and Rodrigues: An
overview and bibliography. Dodo, Jersey Wildlife Preservation Trust 31: 40-65.
KORS6S, Z. & TR6CSANYI, B. 2000: Report on the Hungarian Herpetological Expedition to
Round Island, Mauritius, November 1999. Unpublished report, Hungarian
Natural History Museum & University of Pecs.
KORSOS, Z. & TROCSANYI, B. 2001: Population assessment of Gunther's Gecko in its natural
habitat, Round Island, Mauritius. 15th Annual Meeting of the Society of
Conservation Biology, University of Hawaii, Hilo, USA, July 29-August 1, 2001
MERTON, D. V., ATKINSON, I. A. E., STRAHM, W, JONES, C., EMPSON, R. A.,
MUNGROO, Y, DULLOO, E. & LEWIS, R. 1989: A management plan for the
restoration of Round Island, Mauritius. Jersey Wildlife Preservation Trust.
NORTH, S. G. & BULLOCK, D. J. 1986: Changes in vegetation and populations of introduced
mammals of Round Island and Gunner's Quoin, Mauritius. Biological
Conservation 37: 99-117.
NORTH, S. G., BULLOCK, D. J. & DULLOO, M. E. 1994: Changes in the vegetation and reptile
populations on Round Island, Mauritius, following eradication of rabbits.
Biological Conservation 67: 21-28
OWADALLY, A. W. & LAMBERT, M. 1988: Herpetology in Mauritius. A history of extinction,
future hope for conservation. Bulletin of the British Herpetological Society 23:11-
20.
RAMREKHA, S. 1999: Expedition on Round Island, 17th-25th August 1999. Unpublished
report, National Park and Conservation Service, Mauritius.
TONGE, S. 1990: The past, present and future of the herpetofauna of Mauritius. Bulletin of
the Chicago Herpetological Society 25: 220-226.
TR6CSANYI, B. 1997: The artificial environment and activity: Some aspects of the captive
maintenance of the Round Island Gecko (Phelsuma guentheri). Thesis, University
of Kent & Jersey Wildlife Preservation Trust.
VINSON, J. 1953: Some recent data on the fauna of Round and Serpent Islands. Proceedings
of the Royal Society of Arts and Sciences, Mauritius 1: 253-257.
VINSON, J. & VINSON, J.-M. 1969: The saurian fauna of the Mascarene islands. Bulletin of
the Mauritius Institute 6: 203-320.
VINSON, J.-M. 1975: Notes on the reptiles of Round Island. Bulletin of the Mauritius
Institute 8: 49-67.

KORSOS
& TROCSANYI Bio la 31-3, 2003 81
2. Palm savannah: typical
vegetation of the
western part of Round
Island
1. Round Island from the air
3. Gunther's Gecko
Pheisuma guentheri
Boulenger, 1885

82 Biota 3/1-2, 2002 KORSOS & TROCSANYI
5. Ornate Day Gecko
Phelsuma ornata Gray,
1825
- .
4. Gunther's Gecko
Phelsuma guentheri
Boulenger, 1885
6. Durrell's Night
Gecko Nactus serpensinsula
durrelli
Arnold & Jones, 1994

KORSOS & TROCSANYI BJOla 3/1-2, 2002 83
8. Bojer's Skink
Congylomorphus
bojerii (Desjardins,
1831)
7. Telfair's Skink
Leioiopisma telfairii
(Desjardins, 1831)
9. Bouton's Skink
Cryptoblepharus boutonii
(Desjardins,
1831)

84 Biota 3/1-2, 2002 KORSOS & TROCSANYI
11. Keel-scaled Tree
Boa Casarea dussumieri
(Schle gel, 1837).
Adult's head
10. Burrowing Round Island Boa Bolyeria multocarinata (Bole,
1827)
The only museum specimen in Mauritius
12. Keel-scaled Tree
Boa Casarea dussumieri
(Schlegel, 1837).
Juvenile colouration

KUTRUP & YILMAZ Biota 3/i-a, 2002 85
Preliminary data on some new
specimens of Vipera barani
collected from Trabzon
(Northeastern Turkey)
Bilal KUTRUP1 & Nurh ay at YILMAZ2
1Karadeniz Technical University, Faculty of Arts and Sciences, Dept. of Biology, 61080,
Trabzon, Turkey
E-mail: kutrup@ktu.edu.tr
2Karadeniz Technical University, Faculty of Rize, Arts&Sciences, Dept. of Biology, 53100,
Rize, Turkey
E-mail: nurhayat@ktu.edu.tr
Abstract
This study was carried out to find viper specimens in Trabzon, situated in northeastern
Turkey. During the study period, a total of 9 viper specimens were collected from new
locations in Arpaozu (Caykara), Balhca, Sugeldi (Of), and Camlik (Vakfikebir). For morphometric
studies, these specimens were examined, and for all of these specimens, 16
different items of data were collected. In addition, collected data were compared with
data from other vipers (Vipera barani and Vipera pontica) which are described from
Rize, Artvin and Adapazan. Trabzon viper specimens are characterised by partial fragmentation
of frontal and parietals, high ventrals and loreals, fewer subcaudals and yellow
green tail tips than in Vipera barani and Vipera pontica. When comparing the subalpine
population (Arpaozu) with other lowland ones (Of, Yomra and Vakfikebir), differences
in apical and dorsal patterns could be detected. The Arpaozu population had
one apical, which is normally found only in Vipera ursiniiand occasionally in the Vipera
kaznakovi complex (V. kaznakovi, V. dinniki and V. darevski). On the other hand, the
number of apicals was 2 or 3 in Vipera barani and V. pontica. Also, the number of zigzag
bands was low in the Arpaozu specimens (48-49 instead of 62-67 in the lowland
specimens). As a result, it is clearly indicated that the new viper specimens from Trabzon
show some similarities in colour patterns as well as scalation characters with V. barani
and V. pontica. However, the number of circumoculars, loreals and crown scales is lower
in our specimens than in both species. Taking into consideration the differences and
similarities, the Trabzon specimens are most similar to V. barani
Key words: Reptilia, Viperidae, Vipera barani, Trabzon
Received 30 August; accepted 20 December 2001

86 3/1-2, 2OO2 KUTRUP & YILMAZ
INTRODUCTION
Recent studies of the systematics of
vipers in the Caucasus have shown that
the taxonomy of this group is rather complex.
It has been reported that there are
two viper groups in this region, Vipera
kaznakovi and Vipera ursinii (Vedmederja
et al. 1986, Nilson et al. 1995). Vipera
barani has been evaluated to be an
important species in the Vipera kaznakovi
group as Vipera pontica. Only Vipera
kaznakovi is well defined and restricted in
distribution and geographically separated
from the other species in this region.
Among these species, Vipera pontica
shows clear similarities with Vipera barani
in having high fragmentation of the
frontal and parietals, in the number of
ventrals as well as by their yellow green
tail (Billing et al. 1990).
Initially, Vipera barani Bohme and Joger,
1983, was originally described from a single
specimen - a melanistic female from
Sapanca (Adapazan) which is situated on
the southwest coast of the Black Sea.
Then, Frazen and Heckes identified and
described Vipera barani from Rize
(Nilson, pers. comm.). Recently Baran et
al. (1997) published information about
three viper specimens that they considered
as being V. pontica from
Camlyhemsin (Rize). Nevertheless, some
German herpetologists (Frazen and
Heckes) believe that the Camlyhemsin
population should be evaluated as V.
barani
The taxonomy of the Vipera barani and
Vipera pontica has been confusing and
contradictory (Hoggren et al. 1993,
Nilson et al. 1994, 1995, Joger et al.
1997). The original question of whether
the specimens belonging to the Eastern
Black Sea Region (Rize and Trabzon) and
those of Adapazan are identical must be
answered. For this purpose, DNA
sequences of the vipers caught from
Adapazan and Rize have been studied by
Frazen and Heckes. There is another
question; could it be that what we have is
a series of populations within the subgenus
pelias along the Anatolian Black
Sea coast? The new specimens caught
from Trabzon could fit into such a series.
On the other hand, Camlyhemsin and
Adapazan specimens were evaluated as
two different groups (V. barani in berus
group and V. pontica in aspisfgroup) by
Joger et al. (1997). They also stated that
the melanistic specimens belonging to V.
barani show similarities with the aspis
group. Recently, there has been a trend
to evaluate both species as the same
taxon.
Despite intensive searching by several
herpetologists, no additional Vipera
barani specimen was captured.
Examining more specimens belonging to
Viper from new localities in Trabzon is
aimed at gaining a broad perspective
including colour pattern and pholidosis
characteristics.
MATERIAL AND METHODS
The study was conducted from April to
September of 1999 and 2000, and joint
field trips were made in different parts of
Trabzon to capture the samples of viper.
A total of nine specimens (5 females and
4 males) were captured from four different
locations. Two adult females were
caught from the locality of Arpaozu, situated
approximately 20 km south of
Uzungol (Trabzon). This habitat was separated
from the others by having steep
wooded mountain slopes with many big,
rocky outcrops at high altitudes (more
than 2000m).
The other specimens were caught in the
lowland Black Sea coast locations of
Ballica, Sulgeldi (Of), Cinarh (Yomra) and
Camhk (Vakfikebir). One adult male was
captured from Ballica, situated approximately
63 km east of Trabzon and another
adult male was encountered in Sugeldi,
situated 5 km to the eastern part of
Ballica. One adult female was caught in

KUTRUP & YILMAZ 3/1-2, 2O02 87
Cmarli, which is nearby in the east of
Trabzon. Two adult males, one adult and
young females were retrieved from
Camlik which is 68 km west of Trabzon.
The locality of Cumhuriyet is approximately
one km from the sea with an altitude
of 80 m., while the Ballica and
Sugeldi localities are two or three km
from the sea, with an altitude of 250-330
m. Although Ballica and Sugeldi specimens
were captured near the tea plant
Camellia sirennis, the Cmarli and
Cumhuriyet specimens were captured on
short grassland populated by Oak
Quercussp., Alder Alnus sp. and hazelnut
Corylus sp. trees.
Pattern and coloration were examined
and noted for five adult and three immature
specimens. In addition, black and
white slides were taken. Then specimens
were fixed with 70% ethanol. For morphometric
studies, 9 vipers were examined,
and for all of these specimens 16
different items of data were obtained.
Total and tail length were measured. In
addition, the number of ventrals, subcaudals,
anterior, mid-body and posterior
dorsal scale rows, apical plates, supralabials,
sublabials, circumocular scales, loreals,
chanthals, crown scales and zig-zag
spirals in the dorsal band were counted. A
further rostral index (height/breadth) was
calculated. The division of parietals and
frontals was noted. This information was
used in morphological description, taxonomical
analyses and comparison with
the other Vipera barani and Vipera pontica
captured from different locations.
RESULTS AND DISCUSSION
The new specimens from lowland populations
of Yomra, Of and Vakfikebir show
some differences in terms of scalation and
colour pattern in comparison to the subalpine
population (Arpaozu). Also, our
specimens were compared to Vipera
barani and Vipera pontica, which were
described from Adapazan and Rize, in
order to clarify these morphological similarities
and differences within both
species. In addition, they were compared
to the Camlyhemsin specimens, which
had not yet received a clear taxonomic
position (Table 1).
Table 1. External morphology of the Trabzon viper specimens collected from the
Arpaozu, Of, Yomra and Vakfikebir and related taxa (V. barani and V. pontica)
Characters
Tail length (mm)
Total length (mm)
Ventral
Subcaudal
Dorsal scales
Neck
Mid-body
Pasterior
Apicals
Chantals
Rostral index
Loreals
Supralabials
Sublabiais
Circumoculars
Crown scales
Zig-zag band
Arpaozu
2(2)
52-73
444-590
142
28-29
25-26
21
17
1
2-2
1.12-1.24
4-5
8-9
9-11
8-9
18-24
48(49)
Of
2(6")
82-83
518-543
140-141
33-37
24-25
21
17-18
2
2-2
1.05-1.18
5-5
9-9
11-11
10-13
22-25
67-*
Yomra
1(2)
78
634
148
29
26
21
17
2
2-2
1.10
4-5
9-8
10-11
11
26
56
2(2)
22-69
176-627
149
30-31
23-25
21
17
2
2-2
1.05-1.07
5-4
8-9
10-11
9-10
22-25
62-65
= Character was not visible because of damaged specimen
Vakfikebir
2(

88 Biota 3 -2, 2OO2 KUTRUP & YILMAZ
As can be seen in Table 1, the number of
ventrals in the specimens caught from
Trabzon falls between 141 and 145,
except for two specimens collected from
Yomra and Vakfikebir. These specimens
have a high number of ventrals (148-
149), a feature which is not counted in V.
barani and V. pontica. The number of
subcaudals in males is higher than in
females (33-38 instead of 28-31 in
females). Also, it is obvious that males
have longer tails than females.
Consequently, the rate of tail length to
total length, which is 0.11-0.13 in
females and 0.15-0.17 in males, confirms
this result. The Arpaozu population (two
females) shows remarkable differences
from the other lowland populations in
having one apical (Figure 1), a feature
which is normally found only in Vipera
ursinii, and occasionally in the V. kaznakovicomplex
(V. kaznakovi, V. dinniki
and V. darevski). In contrast, one apical
was not reported either in V. barani or in
Figure 1. Dorsal view of the head of the
adult female from Arpaozu
patterns of the Arpaozu specimens are
very different from the members of this
group (V. kaznakovi complex).
The number of crown scales (interchantals
+intersupraoculars ) is much lower in
Trabzon specimens (18-26) than in the V.
pontica (34), V. barani (25-40) and
Camlyhemsin specimens (25-33). But
only a young male from Vakfikebir differs
from the others in having higher crown
scales (33) than in the V. barani and V.
pontica specimens.
Additionally, the Trabzon specimens have
lower loreal counts (4-5, 5-4 and 5-5)
(Figure 2). Four of the specimens examined
have 5-5 loreals like the
Camlyhemsin ones (5-5), while the others
show similarities with V. barani (3-6).
Furthermore, our specimens differ from
Figure 2. Lateral view of the head of the
adult male from Balhca
the V. pontica, V. barani and
Camlyhemsin specimens in having fewer
circumoculars (9-11 instead of 10-14 in
\. pontica V. (Table barani, 1). 11-12 However, in V. pontica colour and 10-12
in the Camlyhemsin specimens) On the
other hand, the Balhca specimens show
clear similarities in circumocular counts
(12-13) with V. barani (figure 2).
Head scales between the rostral and posterior
end of the parietal area, as well as
temporal, are not keeled as in V. barani
and V. pontica. The head morphology of
the Trabzon specimens differs from that
of V. pontica in having much more blunt
and expanded snouts than any member

KUTRUP & YILMAZ Biola '31-2,2002 89
Figure 3. Dorsal view of the adult Ballica
male
of the berus complex (Figure 1).
According to Billing et al. (1990), V. pontica
has a pronounced snout Also our
specimens do not have the markedly
raised snouts which are normally found in
several Caucasian viper populations such
as V. pontica (Nilson et al. 1995).
A great number of different colour
morphs was expressed in the Caucasian
vipers (Billing et al. 1990, Hoggren et al.
1993, Nilson et al. 1995, Kutrup 1999).
Although the Arpaozu locality, which was
situated in high subalpine mountain belts
(2000m) such as the Vipera dinniki locality
(Nilson et al. 1994, 1995, Hoggren et
al. 1993), has very different climatic factors
and plant species, specimens belong-
ing to Arpaozu did not show any significant
differences in terms of colour and
pattern from the lowland specimens.
Only the number of spirals in the dorsal
zig-zag band was lower in the Arpaozu
specimens (48-49 instead of 62-67 in the
Of, Yomra, and Vakfikebir specimens,
respectively) (Figure 3, Table 1). This high
number of zig-zag bands (67) was normally
found in the high populations of V.
dinniki (Nilson et al. 1995, Joger et al.
1997). On the other hand, the number of
zig-zag bands which was observed in
adult samples was lower than in young
ones. This could be given an ontogenetic
explanation, as pattern often fades with
age (Nilson et al. 1995).
Only two melanistic specimens (one
female from Arpaozu and one male from
Sugeldi) were observed during the study
period (Figure 1). In these samples, a
black coloration with high melanin production
covered all other colour patterns.
In contrast, dorsal patterns belonging to
unmelanistic specimens consisted of light
brown in adults and dark brown in the
young, and broad, black-bordered zigzag
spirals with a total of 48-67 in both
adults and young. Also, many dark
blotches, which were vertically placed
and not connected to the zig-zag band,
were observed along the body sides.
Head patterns of two black bands
extended from the posterior end of the
parietal area to the sides of the neck and
they had wide, black-bordered bands,
which ran from the eyes to the neck
along the upper sides of the supralabials
(Figure 2). All unmelanistic specimens
had a blotch which was approximately
elliptical on the neck, just as in V. barani
from Adapazan (Joger et al. 1997) . The
edges of the supralabials, rostral , chanthals
and supraoculars were white. All of
the samples had black ventrals. Although
bigger white spots were only seen on the
ventral of the head in the melanistic samples,
many large and small white spots

90 Biota 3/1-2,2002 KUTRUP & YILMAZ
extended from mentale to mid body area
in the unmelanistic samples. These white
spots also ran along the sides of the ventrals
to the anale in young samples. The
yellow coloration, which was seen at the
posterior area of the subcaudals, was fairly
distinct in all specimens.
An examination of the morphology of the
different specimens clearly indicate that
Trabzon vipers show some differences in
terms of colour patterns as well as scalation
characteristics from V. barani and V.
pontica. On the other hand, these vipers
show similarities with V. barani and V.
pontica in scalation as well as colour patterns.
According to the scalation, the
specimens of Trabzon vipers may be
belong to either V. barani or V. pontica.
But the colour patterns of our melanistic
specimens are more similar to those in V.
barani than to those in V. pontica.
However, the number of crown scales
(interchanthals + intersupraoculars), loreals
and circumoculars is lower in our specimens
than in both other species.
Taking into account the differences and
similarities as indicated above, Trabzon
samples have more similarities with V.
barani than with V. pontica. Ultimately,
we evaluate our vipers as Vipera barani.
Furthermore, our viper samples can be
evaluated as belonging to a subspecies of
V. barani (Joger, pers. comm.).
It must be borne in mind that we would
need a larger series of molecular data to
find out more.
REFERENCES
BARAN, I., TOSUNOGLU, M., KAYA, U. & KUMLUTAS, Y. 1997: Camlyhemsin (Rize
Civarmm Herpeto- faunasi Hakkinda. Tr. j. of Zoology 21: 409-416.
BILLING, H., NILSON, G. & SATTER, U. 1990: Vipera pontica sp. n., a new species in the
kaznakovi group (Reptia-Viperidae) from north-eastern Turkey and adjacent
Transcaucasiana. Zool. Scripta 19: 231-237.
B6HME, W. & JOGER, U. 1983: Eine neue Arts des Vipera berus -Komlexes aus der Turkei.
Amphibia-Reptilia 4: 265-271.
HOGGREN, M., NILSON, G., ANDREN, C, ORLOV, N. L & TUNIEV, B. S. 1993: Vipers of
the Caucasus. Natural History and Systematic Review. Herpetological Natural
History 1: 11-19.
JOGER, U., LENK, P., BARAN, I., BOHME, W., ZEIGLER, T, HEIDRICH, P. & WINK, M. 1997:
The phylogenetic position of Vipera barani and of V. niloskii within the Vipera
berus complex. Herpetologica Bonnensis: 185-194.
KUTRUP, B. 1999: The Morphology of Vipera ammodytes transcaucasiana (Reptilia,
Viperidae) Specimens Collected from Murgul (Artvin. Turkey), Tr. J. of Zoology
23: 433-438.
NILSON, G., ANDREN, C. & SZYDLAR, Z. 1994: The systematic position of the Common
Adder, Vipera berus (L.) (Reptilia, Viperidae), in north Korea and adjacent regions.
Bonn. Zool. Beitr. 45: 49-56.
NILSON, G., TUNIEV, B., ORLOV, N. , HOGREN, M. & ANDREN, C. 1995. Systematics of
the Vipers of the Caucasus: Polymorphism or Sibling Species. Asiatic
Herpetological Research 6: 1 -26.
VEDMEDERJA, V. L., ORLOV, N. L. & TUNIEV, B. S. 1986: On the taxonomy of three viper
species of the Vipera kaznakovi complex. In: Ananjeva, N. & Borkin, L. (eds).
Systematics and Ecology of Amphibians and Reptiles. Proceedings of the
Zoological Institute, Leningrad: 55-65. In Russian.

MEESKE, SCHNEEWEISS & RYBCZYNSKI Biota 3/i-a, 2002 91
Reproduction of the European
Pond Turtle Emys orbicularis in
the northern limit
of the species range
A.CM. MEESKE1, N. SCHNEEWEISS2 & KJ.
RYBCZYNSKi2
1Zentrum fuer Naturschutz, Von-Siebold-Str.2r D-37075 Goettingen
E-mail: mmeeske@gwdg.de
2Naturschutzstation Rhinluch, Nauener Str. 68, D-16833 Linum
E-mail: agena@t-online.de
Abstract
Incubation conditions and ecological requirements of nesting areas of the European
Pond Turtle Emys orbicularis were investigated in the summers of 1997-2001 in one
local population in southwest Lithuania. Females were trapped and each individual was
weighed, measured and colourmarked. Additionally, some females were tagged with
radio transmitters to find out nest sites. Nesting started between the middle of May and
the first week of June. The nesting period extended over 14-20 days. Open areas on
sandy dry grassland were used as nesting areas. Nest sites had expositions of 80-270°
and were found on flat, slight or strongly inclined ground. The average vegetation cover
amounted to 50%. 77% of females which were observed in at least three years nested
in the same nesting area and 23% of females in different nesting areas. The mean nest
size averaged 12,5 eggs. Females with larger body sizes frequently deposited a larger
number of eggs. From captures of juveniles, successful reproduction for several years
could be proved in the 90s. Suitable temperatures in summer and the following winter
allow successful incubation.
Keywords: Testudines, Emydidae, Emys orbicularis, reproduction, Lithuania
Received 8 February2002; accepted 28 July 2002

92 Bio la 3 i-a, 2002 MEESKE, SCHNEEWEISS & RYBCZYNSKI
INTRODUCTION
The European Pond Turtle Emys orbicularis
is a threatened species, especially in
the northern range of its distribution
(Juszczyk 1987, Zemanek 1988, 1991,
Schiemenz & Gunther 1994, Fritz 1996,
Fritz & Gunther 1996, Fritz 2000,
Schneeweiss & Fritz 2000). In particular,
during the last two centuries a heavy
decline of European Pond Turtle was
noticed. Lithuania represents the northern
border of the species range (Fritz
1996), where the subspecies orbicularis
lives (Fritz 1992, 1998). Along this northern
border the climate, including short
summers and cold winters, is very stressful
to reproduction (Schneeweiss &
Jablonski 2000).
Up to the present, different sites with
occurrences of turtles were found mainly
in the southern parts of the country
(Balciauskas et al. 1997), but there is a
Figure 1. Geographic position of the study area
60° N
BAlClC SGA
lack of information about reproduction.
The actual distribution and sizes of the
local populations is still unknown.
Nevertheless, young turtles were noticed
in different locations in Lazdijai District in
southwestern Lithuania, which is a proof
of reproduction success in these areas.
The present study examines incubation
conditions and ecological requirements of
nesting areas in the view of incubation
success under natural conditions.
STUDY AREA
The study area is located in southwestern
Lithuania near the northeastern border
with Poland (40 km SSW Alytus, Lazdijai
District; 23°90'E, 54°40'N) (see Figure 1).
This area experiences a cold continental
climate, including warm summers and
cold winters (Pearce & Smith 1993). The
area contains different kinds of water
bodies, partly seasonally flooded wet-

MEESKE,
SCHNEEWEISS & RYBCZYNSKI Biota 3/i-a, 2002 93
Figure 2. Map of the study area
lands and sandy, dry areas, deciduous
forests, pine forests and extensively used
agricultured land. The local population
inhabits two main ponds (A, G) and also
uses several smaller and mainly temporary
water bodies (B, C, D, E, F, H, I, K,
M). Four nesting areas (I-IV) are situated
in front of pine forests (see fig. 2).
MATERIAL AND METHODS
Before the start of the nesting period,
females were captured in aquatic baited
traps (method by Servan 1986) and in
land traps and were weighed, measured
(method by Fritz 1989, 1992, 1995) and
colourmarked for identification. Some
females were tagged with radio transmitters
glued on the carapace. Localizations
of turtles using a receiver (Stabo XR 100)
connected to a hand operated unidirec-
nesting area
field
way
border of reserve
tional antenna were recorded to find out
nesting sites. Females were located by
walking in the direction indicated by the
antenna (Homing-in-on-the-Animal)
(White & Garrott 1990). During the nesting
period the nesting areas were controlled
every afternoon and evening to
notice the number of females, their nesting
behaviour and the nest locations.
The degree of vegetation cover was estimated
(1m2 around the nest). Expositions
were noticed with a compass and inclinations
with a protractor.
RESULTS
97% of the females (30 of 31) were
mature and 94% of the females (29 of 31)
were observed nesting at least once.
The body sizes and body masses of mature
females turtles are shown in Table 1.

94 3/1-2, 2OO2 MEESKE, SCHNEEWEISS & RYBCZYNSKI
Figure 3. During nesting period 2000 daily observations of females and their nests in
the investigation area in comparison with the daily maximum air temperature
Date
Beginning, duration and course of oviposition
period
In four years (1997r 1999-2001) the
nesting season started between the middle
of May and the first week of June and
continued 14-20 days.
On various days during the nesting period
the number of females wandering in
the investigation area was higher than
the number of observed ovipositions (see
Figure 3). Some females could not be
observed nesting in the known nesting
areas. Other females did not nest the first
day of emergence from the water body.
Table 1. Body sizes and body masses of mature females
23
26
24
22
20
18 ?
• 16 £
14 g
12 |
10 £
S
6
4
2
131 observed Females
E23 observed Nests
Maximum Air Temperature °c
Nesting Site Fidelity
Animals inhabiting home pond A mainly
used nesting areas II and III, whereas turtles
inhabiting home pond G preferred
nesting areas I and IV and unknown
areas. Female turtles showed nesting site
fidelity. 77% of females observed over
three years (10 of 13) nested in the same
nesting area (nesting site fidelity within
300 m). 31% of the females (4 of 13)
showed fidelity within 20 m. 23% of the
females (3 of 13) changed their nesting
area.
During five observation years, female
Mean Range SD n
Carapace length 170mm 146-189 mm 10.19 30
Carapace width 136mm 120-152 mm 8.064 30
Carapace height 72mm 54-86 nun 5.69 30
Body mass 916 g 630-1252 g 365.1 25

MEESKE, SCHNEEWEISS & RYBCZYNSKI Biota 3/1-2,2002 95
Table 2. Variation of nest sizes of female turtles with at least two nests in different years.
Female
KUC-22
KUC-5
KUC-12
KUC-3
KUC-6
KUC-9
KUC-16
KUC-34
KUC-33
KUC-10
KUC-17
KUC-4
KUC-21
KUC-23
Length of
carapace (mm)
156
160
163
169
170
173
176
176
177
181
181
182
187
188
1997
8
13
12
11
-
12
?
-
-
13
12
9
20
12
Number o
1998
-
?
10
13
9
11
-
9
12
15
11
14
16
-
"eggs laid
1999
-
-
9 9
Figure 4. Comparison of females' carapace sizes with their clutch sizes in
four observation years
155 160 165 170 175 180 185 , 190 r 195 „
Carapaxlength (mm)
KUC-17 nested in three different nesting
areas (nesting area I: 1997/98; nesting
area II: 1999/2000; nesting area III:
2001). Nesting area II included most of all
nests. Before and after the first change of
the nesting area, KUC-17 showed nesting
site fidelity and laid eggs within 10 m in
two subsequent years. Female KUC-12
searched every year for the same nesting
area (III), which is divided into two suitable
main parts. KUC-12 regularly
9
9
14
16
10
-
12
14
17
?
2000
10
11
12
13
11
13
15
17
-
-
11
14
-
15
• Number of laid Eggs 1997
- Number of laid Eggs 1998
" Number of laid Eggs 1999
Number of laid Eaas 2000
changed these locations and consequently
used part I in 1997, 1999 and 2001
and part II in 1998 and 2000.
Clutch size
The females deposited an average of 12,5
eggs (range = 7-20, SE = 2,59, n = 46).
Larger females frequently laid a larger
number of eggs (see figure 4). Variation
of nest sizes within individuals in different
years is shown in Table 2.

96 Biota 3/1-2,2002 MEESKE, SCHNEEWEISS & RYBCZYNSKI
Nesting females were older than 25
years, up to very old females for which
age was not determinable. The smallest
females had the smallest clutch sizes
(female KUC-22: 8 eggs in 1997; female
KUC-19: 7 eggs in 2000), while one of
the largest and oldest females, KUC-21,
laid the maximum number of 20 eggs in
1997 (see Table 2).
Nest site locations
The females nested in open areas on
spongy sandy ground on sandy dry grassland
(n = 57) and in five cases on loamy
ground. Around the nests, vegetation
was found which showed sunny and dry
conditions or sandy ground for this area.
Index species are Hieradum pilosella,
Artemisia scoparia and Sedum acre. The
degree of vegetation cover ranged
between 30-90% (mean = 55%, n = 16).
The females chose places with inclinations
between 0-20°. 45% of the nests
(14 of 31) were on flat ground (0-5°),
19% (6 of 31) on slight inclined (5-10°),
and 36% (11 of 31) on strongly inclined
ground (10-20°).
The expositions of 50 nest sites amounted
to 80-270°.
Reproduction success
Young turtles from different years in the
90s were recorded in the study area. They
hatched before 1997 or after the summer
of 1999. These results in successful incubation
were noticed only for the nests
Table 3. Results of incubation success for nests without predation from August. All nests
were protected against predators, controls of nests in May 2000 (*nest controlling in
October 1999).
Female
KUC-6*
KUC-12
KUC-14
KUC-16
KUC-20
KUC-23
KUC-24
Clutch size
9
12
>7
14
12
?
?
Unfertilized
eggs%
44
0
0
0
0
0
1 egg
Dead er
before
hatching
0
0
0
0
0
0
1 embryo
from 1999 during 1997-2000.
After the warm summer in 1999, some
nests showed signs of hatching in
September and at the beginning of
October, but most of the juveniles left
the nests during the following spring of
2000. 7 nests were checked after hatching
and showed a high hatching rate up
to 100% (see Table 3).
8 nests from 1997 were dug out in the
middle of September during possible
hatching time. Only 1 juvenile of a controlled
79 eggs successfully left the egg
shell before digging. Nests from 1998
were not checked. Generally, no hatch-
ibryo%
after hatching
11-22
0
0
0
8
0
0
Remarks
33% hatched successfully
100% hatched successfully
100% hatched successfully
100% hatched successfully
92% hatched successfully
100% hatched successfully
> 70% hatched successfully
ing was observed. The next spring after
the snow melted, 60% of the nests were
destroyed by predators. 7 nests from the
year 2000 (controlled in 2001) contained
78% fertilized eggs (range: 30-100%),
but no hatchling left the egg shell or the
breeding chamber respectively. The
young turtles died before hatching in
summer and autumn or during winter.
DISCUSSION
Beginning and duration of oviposition
period
Studies in central Poland showed that
weather conditions influence the onset

MEESKE, SCHNEEWEISS & RYBCZYNSKI Biota 3/1-2,2002 97
of the nesting period (Zemanek & Mitrus
1997) and the duration of the nesting
period (Zemanek 1988)r which explains
the differences in beginning and duration
of nesting periods in Lithuania.
Generally, females of the European Pond
Turtle deposited eggs in East Germany
(Breitenstein 1973, Andreas & Paul 1998,
Schneeweiss et al. 1998) and in Poland
(Jablonski 1992, Jablonski & Jablonska
1998, Zemanek & Mitrus 1997, Mitrus &
Zemanek 1998, 2000) at the end of May
and in June. The results of this study support
these findings. The short summer in
the northern ranges does not allow later
nesting (e.g. in July), because the time for
incubation becomes too short. In southern
regions, nesting in July is not unusual
(Fritz & Gunther 1996).
Nesting Site Fidelity
77% of the females had nests within <
300 m distance from the previous nest.
31 % of the females nested within 20 m.
23% of the females searched for different
nesting areas, but presumably, more
specimens shifted nesting areas. In two
subsequent years, female KUC-10
deposited eggs in the same nesting area.
In the third year KUC-10 showed searching
behaviour in the same area but was
not observed nesting there.
Several Lithuanian females showed nesting
site fidelity. Nesting site fidelity for
European Pond Turtle is mentioned by
Wermuth (1952) and Schneeweiss &
Steinhauer (1998) for Germany, by
Jablonski & Jablonska (1998) and Mitrus
& Zemanek (2000) for Poland, and by
Rovero & Chelazzi (1996) for Italy,
although there are no standardized definitions
and no figures about nesting site
fidelity. In Austria, European Pond Turtle
females laid eggs within 14.5 m in 2 years
(Roessler 2000). Various females of
Chrysemys picta marginata in Canada
laid eggs within 10 m of their previous
nest (Christens & Bider 1987). Congdon
et al. (1983) investigated an inter-annual
fidelity of 73% of Emydoidea blandingi
females to the nest area in Michigan and
Loncke & Obhard (1977) of 92% of
Chelydra serpentina females in Canada.
The Lithuanian results are approximately
coincident with the observations of other
species.
During five years of observation, female
KUC-17 nested in three different nesting
areas. The substrate in the first nesting
area (I) seemed to be very unsuitable
because of the very hard loamy ground.
The reason for her next change from a
suitable nesting area (II) to another nesting
area (III) is unknown. Regular shifting
of nesting areas could be a kind of reproduction
strategy in comparison with nesting
site fidelity. In various nesting areas
different predation pressure on nests or
different incubation conditions could
occur. The shifting strategy also helps in
searching for new nesting areas if old
places were disturbed by overgrowth of
bushes and trees or were destroyed by
man. The known nesting areas of the 80s
in the same study area are partly unsuitable
because of overgrowth today
(Gruodis, verbal information, 2001), so
the turtles were forced to search for other
places. Schneeweiss & Steinhauer (1998)
and Mitrus & Zemanek (2000) also
assume that females change nesting
areas when old places are destroyed. But
Schneeweiss & Steinhauer (1998) also
observed that if suitable places were not
on females' migration routes, the females
used unsuitable places.
Clutch size
In Lithuania the mean clutch size averaged
12.5 eggs (range: 7-20 eggs) over 4
years and was independent of the beginning
or the duration of nesting periods. In
East Germany Etnys orbicularis females
deposited an average number of 12.7
eggs (Schneeweiss et al. 1998). This
result agrees with the Lithuanian data.

98 Biota 3/1-2,2002 MEESKE, SCHNEEWEISS & RYBCZYNSKI
Roessler (2000) found average clutch
sizes of 12.4 eggs in Austria, but there
some females nested twice a year. The
result of the Lithuanian range of clutch
sizes is similar to observations of clutch
sizes of 10-18 eggs in East Germany
(Andreas & Paul 1998) and in Poland
(Jablonski 1992, Zemanek & Mitrus
1997, Mitrus & Zemanek 1998).
Nest site locations
In Lithuania similar results for nest site
locations were found in comparison with
other regions. The Lithuanian females
mainly used open areas with sandy dry
grassland for nesting (57 of 62 nests). In
East Germany 18 of 32 nests were on
xerothermic sandy grassland and all nests
were on sandy or sandy-loamy areas
(Schneeweiss et al. 1998). In the study
area the vegetation around the nests
describes typical characteristics for nesting
areas of European Pond Turtle
(Meeske 1997). The degree of vegetation
cover ranged between 30-90% (mean =
55%, n = 16). A vegetation cover of nest
sites of between 5-80% was observed in
East Germany (Schneeweiss et al. 1998)
and in Austria between 80-90% (Roessler
2000). In Lithuania, in cases of high vegetation
cover the plants were mainly
species with short growth which do not
impair the sun's radiation on the nest.
The ecological requirements of nesting
areas do not clearly differ in the northern
species range.
Isberg (1929) mentions that the substrate
type is an important incubation factor. If
there is a spongy substrate as in
Lithuania, the heat of sunshine can better
permeate through upper strata and
humidity can ooze away or evaporate
easily.
The factors of solar exposure and inclination
have a strong impact on the vegetation
and consequently on the ecoclimate
(Schaefer 1992). The nests were in places
with flat, slight or strongly inclined
ground. The degree of slope has an effect
on the intensity of radiation. Additionally,
sloping areas have the advantage that
longer rain periods will do little harm to
the clutches.
Reproduction success
In spring 2000 most of the young turtles
left the nests from 1999. In Lithuania, the
main hatching time is in spring, as in
other northern regions where European
Pond Turtle lives (East Germany: Andreas
& Paul 1998; Poland: Zemanek 1992,
Zemanek & Mitrus 1997, Mitrus &
Zemanek 1998, 2000). After melting of
the snow in spring, the situation of water
bodies in size and number is clearly more
favourable than in autumn. The hatchlings
have shorter ways to the next water
body and can choose between various
kinds of water bodies with diverse depth,
temperature, hiding sites and food selection.
Fritz & Giinther (1996) describe
"spring hatching" in comparison with
"autumn hatching" as a second kind of
reproduction strategy, because after leaving
the nests, juveniles have a long period
of growth before hibernating.
In different years in the 90s juveniles
hatched in Lithuania, but during the
study only the warm summer 1999 and
the gentle winter 1999/2000 guaranteed
successful reproduction. Consequently,
reproduction success strongly depends on
weather conditions. The results of this
study support the findings from northeast
Germany and eastern Poland that in the
northern species range, favourable climatic
conditions in the summer and the
following winter are necessary for reproduction
(Schneeweiss & Jablonski 2000).
In the northern range successful reproduction
does not happen every year.
That the actual reproduction success is
sufficient for a survival of these populations
in Lithuania may become clear in
another study.

MEESKE, SCHNEEWEISS & RYBCZYNSKI Biota 3/1-2,2002 99
Acknowledgements
Thanks to all people, institutes, and organisations who helped to render these investigations
possible, especially Dr. Linas Balciauskas (Vilnius), Dr. Uwe Fritz (Dresden), Dr.
Pranas Mierauskas (Vilnius), Prof. Dr. Miihlenberg (Goettingen), Richard Podloucky
(Hildesheim), Dr. Arunas Pranaitis (Meteliai), and to DAAD (German Academic
Exchange Service), DGHT (German Society for Herpetology), Institute of Ecology
(University Vilnius), Meteliai Regional Pare Service, Ministry of Environment (Vilnius),
Universitaetsbund Goettingen, Center for Nature Conservation (University Goettingen).
REFERENCES
ANDREAS, B. & PAUL, R. 1998: Clutch size and structure of breeding chambers of Emys o.
orbicularis in Brandenburg. In: Fritz, U., Joger, U., Podloucky, R. & Servan, J.
(eds.). Proceedings of the EMYS Symposium Dresden 96. Mertensiella,
Rheinbach 10: 29-32.
BALCIAUSKAS, L., TRAKIMAS, G., JUSKAITIS, R., ULEVICIUS, A. & BALCIAUSKIENIE, L.
1997: Lietuvos zinduoliu, varliagyviu ir ropliu atlasas. Projekto Lietuvos ekologine
ivairovex grupe Ekologijos institutas Vilnius: 112s.
BREITENSTEIN, W. 1973: Eiablage der Europaischen Sumpfschildkrote. Aquarien Terrarien
20: 130S.
CHRISTENS, E. & BIDER, J.R. 1987: Nesting Activity and Hatching Success of the Painted
Turtle (Chrysemys picta marginata) in Southwestern Quebec. Herpetologica 43:
55-65.
CONGDON, J.D., TINKLE, D.W., BREITENBACH, G.L. & VAN LOBEN SELS, R.C. 1983:
Nesting Ecology and Hatching Success in the Turtle Emydoidea blandingi.
Herpetologica 39: 417-429.
FRITZ, U. 1992: Zur innerartlichen Variabilitat von Emys orbicularis (Linnaeus, 1758), 2.
Variabilitat in Osteuropa und Rededefinition von Emys orbicularis (Linnaeus,
1758) und E. o. hellenica (Valenciennes, 1832) (Reptilia: Testudines: Emidae).
Zoologische Abhandlungen des Staatlichen Museums fur Tierkunde Dresden 47:
37-77.
FRITZ, U. 1995: Zur innerartlichen Variabilitat von Emys orbicularis (Linnaeus, 1758), 5a.
Taxonomie in Mittel-Westeuropa, auf Korsika, Sardinien, der Apeninnen-
Halbinsel und Sizilien und Unterartengruppen von E. orbicularis (Reptilia:
Testudines: Emydidae). Zoologische Abhandlungen des Staatlichen Museums fur
Tierkunde Dresden 48: 185-242.
FRITZ, U. 1996: Zur innerartlichen Variabilitat von Emys orbicularis (Linnaeus, 1758), 5b.
Intraspezifische Hierarchic und Zoogeographie (Reptilia: Testudines: Emydidae).
Abhandlungen des Staatlichen Museums fur Tierkunde Dresden 49: 31-71.
FRITZ, U. 1998: Introduction to zoogeography and subspecific differentiation in Emys orbicularis
(Linnaeus, 1785). In: Fritz, U., Joger, U.p Podloucky, R. & Servan, J. (eds.)
Proceedings of the EMYS Symposium Dresden 96. Mertensiella, Rheinbach 10:1-
27.
FRITZ, U. 2000: Verbreitung, Formenvielfalt und Schutz der Europaischen Sumpfschildkrote
Emys orbicularis (L.). In: Hodl, W. & Roessler, M. (eds.). Die Europaische
Sumpfschildkrote. Biologiezentrum d. OO. Landesmuseums, Stapfia69: 13-20.
FRITZ, U. & GUNTHER, R. 1996: Europaische Sumpfschildkrote - Emys orbicularis
(Linnaeus, 1758). In: Gunther, R. (ed.). Die Amphibien und Reptilien
Deutschlands. Gustav Fischer Verlag, Jena: 825 s.
ISBERG, O 1929: Das ehemalige Vorkommen der Sumpfschildkrote (Emys orbicularis) in

100 Biota 3/i-a, 2002 MEESKE, SCHNEEWEISS & RYBCZYNSKI
Schweden und damit zusammenhangende klimatische Erscheinungen. Arkiv for
zoologi 21 A: 1-54.
JABLONSKI, A. 1992: Zolw Blotny. in (1992): Polska czerwona ksiega zwierzat. Warszawa:
231-323.
JABLONSKI, A. & JABLONSKA, S. 1998: Egg-laying in Emys orbicularis (L.), in Leczynsko -
Wlodawskie Lake District (Eastern Poland). In: Fritz, U., Joger, U., Podloucky, R.
& Servan, J. (eds.). Proceedings of the EMYS Symposium Dresden 96.
Mertensiella, Rheinbach 10: 141-146.
JUSZCZYK, W. 1987: Plazy i gady krajowe. Czesc 3. Gady Reptilia. PWN (Panstwowe
Wydawnictwo Naukowe), Warszawa: 214 s.
LONCKE, D.J. & OBHARD, M.E. 1977: Tag Success, Dimensions, Clutch Size and Nesting
Site Fidelity for the Snapping Turtle, Chelydra serpentina, (Reptilia, Testudines,
Chelydridae) in Algonquin Park, Ontario, Canada. Journal of Herpetology 11:
243-244.
MEESKE, M. 1997: Nesting ecology of Emys orbicularis in South Lithuania. Acta Zoologica
Lituanica, Institute of Ecology, Vilnius, Vol. 7 Biodiversity: 138-142.
MITRUS, S. & ZEMANEK, M. 1998: Reproduction of Emys orbicularis (L.) in Central Poland.
In: Fritz, U., Joger, U., Podloucky, R. & Servan, J. (eds.). Proceedings of the EMYS
Symposium Dresden 96. Mertensiella, Rheinbach 10: 187-191.
MITRUS, S. & ZEMANEK, M. 2000: Distribution and biology of Emys orbicularis (L.) in
Poland. In: Hodl, W. & Roessler, M. (eds.). Die Europaische Sumpfschildkrote.
Biologiezentrum d. OO. Landesmuseums, Stapfia69: 107-118.
PEARCE, E. A. & SMITH, C.G. 1993: The world weather guide. 3. Auflg.. Helicon Publishing
Ltd, Oxford
ROESSLER, M. 2000: Die Fortpflanzung der Europaischen Sumpfschildkroten Emys orbicularis
(L.) im Nationalpark Donau-Auen (Niederosterreich). In Hodl, W. & Roessler,
M. (eds.). Die Europaische Sumpfschildkrote. Biologiezentrum d. OO.
Landesmuseums, Stapfia69: 145-156.
ROVERO, F. & CHELAZZI, G. 1996: Nesting migrations in a population of the European
pond turtle Emys orbicularis (L.) (Chelonia Emydidae) from central Italy.
Ethology, Ecology & Evolution 8: 297-304.
SCHAEFER, M. 1992: Worterbucher der Biologie: Okologie. 3., uberarb. u. erw. Auflg.,
Gustav Fischer Verlag, Jena: 433 S.
SCHIEMENZ, H. & GUNTHER, R. 1994: Verbreitungsatlas der Amphibien und Reptilien
Ostdeutschlands (Gebiet der ehemaligen DDR). Rangsdorf: 143 S.
SCHNEEWEISS, N. & STEINHAUER, C. 1998: Habitat use and migrations of a remnant population
of the European pond turtle, Emys o. orbicularis (Linnaeus, 1785),
depending on landscape structures in Brandenburg, Germany. In: Fritz, U., Joger,
U., Podloucky, R. & Servan, J. (eds.). Proceedings of the EMYS Symposium
Dresden 96. Mertensiella, Rheinbach 10: 235-243.
SCHNEEWEISS, N., ANDREAS, B. & JENDRETZKE, N. 1998: Reproductive ecology data of
the European pond turtle (Emys o. orbicularis) in Brandenburg, Northeast
Germany. In: Fritz, U., Joger, U., Podloucky, R. & Servan, J. (eds.). Proceedings of
the EMYS Symposium Dresden 96. Mertensiella, Rheinbach 10: 227-234.
SCHNEEWEISS, N. & FRITZ, U. 2000: Situation, Gefahrdung und Schutz von Emys orbicularis
(L.) in Deutschland. In Hodl, W. & Roessler, M. (eds.). Die Europaische
Sumpfschildkrote. Biologiezentrum d. OO. Landesmuseums, Stapfia 69: 133-
144.
SCHNEEWEISS, N. & JABLONSKI, A. 2000: The reproduction of Emys orbicularis in relation
to climatic factors in Northeast Germany and Eastern Poland. In: Buskirk, J.,

MEESKE, SCHNEEWEISS & RYBCZYNSKI 3/1-2,2002 101
Cheylan, A/I., Duguy, R., Fritz, U.r Jablonski, A., Keller, C, Pieau, C., Podloucky,
R., Servan J. & Taskavak, E. (eds.). Proceedings of the 2. International Symposium
on Emys orbicularis. Edition Soptom, Chelonii 2: 83-87.
SERVAN, J. 1986: Utilisation d'un nouveau piege pour I'etude des populations de Cistude
d'Europe Emys orbicularis (Reptilia: Testudines). Rev. Ecol. (Terre Vie) 41: 111-
117.
WERMUTH, H. 1952: Die Europaische Sumpfschildkrote. Neue Brehm-Bucherei 81,
Wittenberg-Lutherstadt: 40 S.
WHITE, G.C. & GARROTT, R.A. 1990: Home range estimation. In: White, G.C. & Garrott,
R.A. (eds.). Analysis of Wildlife Radio-Tracking Data. Academic Press, INC., San
Diego, U.S.A.: 145-182.
ZEMANEK, M. 1988: Skladanie jaj przez zolwie blotne, Emys orbicularis (L), w Polsce
Srodkowej w warunkach naturalnych. Przeglad Zoologiczny XXXII: 405-417.
ZEMANEK, A/I. 1991: Wystepowanie zolwia blotnego, Emys orbicularis (L.) w Polsce i
zagadnienia jego ochrony. Przyglad Zoologiczny. XXXV: 337-347.
ZEMANEK, M. & MITRUS, S. 1997: Biologia i ochrona Zollwia blotnego Emys orbicularis w
Wojewodztwie radomskim. Chronmy Przyrode Ojczysta 53/1: 67-83.

AAENEGON & SALVIDIO Biota 3/1-2,2002 103
Notes on habitat, egg-laying and
first record of the advertisement
call of Hyperolius kihangensis
Michele MENEGON1 & Sebastiano
SALVIDIO2
'Museo Tridentino di Storia Naturale, Via Calepina 14, C. P. 393, 1-38100 Trento,
Italia,
2Dipartimento per lo Studio del Territorio e delle sue Risorse (DIP.TE.RIS), Corso
Europa26, Genova, 1-16132 Italia.
Abstract
Hyperolius kihangensis is a small treefrog endemic to the Udzungwa Scarp Forest of SE
Tanzania. It breeds only in a restricted area characterised by flooded swamps with shallow
water within dense highland forests. During a herpetological survey conducted in
1998-1999, some ecological data were obtained and the male advertisement call was
recorded at the type locality. The call of H. kihangensis is composed of a series of four
discrete squeaks of low amplitude, with the fundamental frequency of the note at about
2,72 kHz, and with a dominant frequency of about 3.4 kHz. The advertisement call of
H. kihangensis is unusual within the genus because of its peculiar tonal quality. A few
Hyperolius species show similar features in their breeding call, for example, H. kachalo-
/aefrom Zambia and H. viridigulosus fram West Africa.
Key words: advertisement call, Eastern Arc forests, Hyperolius kihangensis, Tanzania
Received 3 September 2001; accepted 1 April 2002

104 Biota 3/i-a, 2002 MENEGON & SALVIDIO
INTRODUCTION
Hyperolius kihangensis has recently been
described in a restricted area in
Udzungwa Scarp Forest Reserve, Eastern
Arc Mountains, Tanzania (preliminarily
described by Schi0tz & Westergaard in
Schi0tz 1999, then formally by Schi0tz &
Westergaard 2000). This endemic
treefrog is small, the male reaching 19
mm, and the female 26 mm in total
length. A brown hourglass (i.e. eightshaped)
dorsal spot characterizes its dorsal
colouration in both sexes (Schi0tz
1999, Schi0tz & Westergaard 2000).
Males have granular dorsal skin and a
gular flap. Swollen white spots, consisting
mainly of iridophores, are also present in
a consistent proportion (i.e. about 25%)
of the population, both in males and
females (Westergaard et al. 2000).
Little is known about the biology of this
species (Schi0tz & Westergaard in Schi0tz
1999) and the breeding call of the male
has not been recorded or described.
During two field expeditions in the
Udzungwa Scarp Forest Reserve, conducted
between December 1998 and
January 1999, we had the opportunity to
collect more specimens of this treefrog, to
observe egg clutches, and to record its
call directly in the field. The aim of this
paper is to describe some aspects of the
reproductive behaviour and the advertisement
call of this little known
Tanzanian treefrog.
MATERIAL AND METHODS
Thirty days of field search were conducted
in Udzungwa Scarp Forest Reserve,
Iringa Region, between December 1998
and January 1999. The site's geographical
coordinates were obtained with a
Magellan 3000 global position system
(GPS) navigator.
The advertisement call of Hyperolius
kihangensis was recorded with a Sony
"TCM 1000a" tape recorder and a Shure
prologue directional microphone. Sound
analysis was carried out on an Apple
Macintosh G4 personal computer. The
signal was sampled at 44.100 Hertz and
16 bit resolution. Canary 1.2.4 software
(Charif et al. 1993) was used for the
measurement of temporal and spectral
parameters of the call, and to generate
audiospectrograms. Frequency information
was obtained through Fast Fourier
Transform (FFT size: 4096 pt.). The call
variables analysed were note duration,
internote duration, fundamental frequency,
dominant frequency, and changes in
dominant frequency. For each parameter
mean, standard deviation and range are
given.
RESULTS
Habitat
The study site is located in the Udzungwa
Scarp at about 1780 m a.s.l (08°22'24"S,
35°58'34"E) and is characterised by a
typical highland vegetation (Moreau
1935). As reported by Schi0tz (1999) and
Schi0tz & Westergaard (2000),
Hyperolius kihangensis was found only in
a small swampy area inside the forest in
proximity to the Kihanga stream. During
our field research about 30 specimens of
both sexes were observed, and 9 (7 males
and 2 females) were collected. Treefrogs
were found on leaves between 30 and
about 200 cm above the ground. The
small swamp where the frogs were found
covered a surface of about 1500 square
metres. This area is more open than the
surrounding canopy forest, and is completely
covered by grasses and shrubs.
During the day the swamp received a
greater quantity of sunlight than the
neighbouring shaded areas, and some
random air temperature measurements
showed relatively higher mean temperatures
in the swamp (27.1 °C) in comparison
with the surrounding forest (25°C).
Several specimens, mating pairs and egg
clutches, of which two contained tadpoles,
were observed. Two egg clutches

MENEGON & SALVIDIO Biota 3/i-a, 2002 105
were adherent to leaves and grasses
above or not far from shallow, stagnant
water and contained 47 to 76 eggs,
respectively. The eggs had a black and a
white pole and the slime was clear.
During the day, Hyperolius kihangensis
remained on branches or leaves in shade
with the limbs close to the body; fingers
and toes were kept between the body
and the substrate. This diurnal sleeping
posture is functional in reducing the surface
area exposed to the air and thereby
the evaporative water loss (Duellman &
Trueb 1986).
Six treefrog species have been listed by
Schi0tz & Westergaard (2000) for this
forest. These are: Hyperolius kihangensis,
Hyperolius puncticulatus, Hyperolius
spinigularis, Afrixalus uluguruensis,
Leptopelis parkeri and Leptopelis barbouri.
During our field researches we
observed all of these species, and in addition
two specimens of Leptopelis uluguruensis
were collected, thus confirming
the previous doubtful record for the
Udzungwa Mountains (Schi0tz 1999).
Description of the advertisement
call
During the month in the field, only once
did we hear the breeding call of H.
kihangensis, in spite of several hours of
night listening. The call of a Hyperolius
kihangensis male (SVL 18.2 mm) was
recorded after almost an hour of attempts
during the night (January 2nd 1999, 9.45
p.m., air temperature = 19,7°C, relative
air humidity 97%). Just two sequences of
four notes were recorded from this
voucher specimen that is now conserved
in the herpetological collection of the
Museo Tridentino di Scienze Natural! of
Trento (collection number MTSN 086TA),
as well as all original pictures regarding
the species.
The advertisement call of the male was a
low pitched voice, composed of a
Figure 1. Audiospectrogram and waveform of a complete sequence of
four notes of the breeding call of Hyperolius kihangensis.
0.2 0.4 0.6 0.9 1£ IjB
S 0-0 02 0.4 Q& J.4 1j6

106 Biota 3/1-2,2002 AAENEGON & SALVIDIO
Figure 2. Frequency spectrum and audiospectrogram of a single
note of the breeding call of Hyperolius kihangensis.
Table 1. Summary of numerical parameters of Hyperolius
kihangensis advertisement call recorded at 19.7°C. kHz =
kiloHertz, ms = milliseconds, S.D. = standard deviation.
Call feature
Fundamental frequency (kHz)
Dominant frequency (kHz)
Changes in dominant frequency
(kHz)
Note duration (ms)
Internote duration (ms)
sequence of four discrete notes. Every
note had a slight variation in frequency
(i.e., frequency modulation) and a peculiar
tonal quality. The audiospectrogram
depicts these very characteristic features
of the note (Figure 1). Two consecutive
calls were recorded and analyzed; each is
a sequence of four discrete notes, similar
Mean ± S.D.
2.076 ±0.1 96
3.368 ±0.175
782.5 ±113.7
57 ± 6.23
35 1.97 ±24.54
Range
2.225 - 3.208
3.048 - 3.578
666 - 956
48.9 - 65.2
321.9-381
in fundamental and dominant frequency
as well as note duration and frequency
modulation. The intercall duration was
about 20 seconds. As depicted in Figure
2, the fundamental frequency of the
note was about 2.72 kHz and the dominant
frequency 3.38 kHz. Each note had
three to four peaks of intensity. The

MENEGON & SALVIDIO Biota 3/1-2,2002 107
power spectrum and the audiospectrogram
of a single note of the call were
analyzed and are shown in Figure 2. The
fundamental and dominant frequencies
were highlighted and it is possible to
appreciate that the note duration was
about 56.5 ms. A summary of several call
parameters is given in Table 1.
DISCUSSION
During our herpetological research in the
Udzungwa Scarp Forest Reserve, several
mating pairs and egg clutches of H.
kihangensis were observed. However,
only one single male was heard calling
and recorded. The very low frequency of
calling, combined with the brevity and
quietness of the advertisement call of this
species, could be reasons for the difficulty
of recording vocalisations of H.
kihangensis males. Thus, there is a possibility
that other Hyperolius species (H.
tanneri and H. spinigularis) with welldeveloped
gular sacs but with unknown
voices (Schi0tz 1999, Schi0tz &
Westergaard 2000) could emit short and
very low-pitched calls, not easily recognizable.
With respect to its breeding habitat and
reproductive behaviour, H. kihangensis is
quite different from all other Hyperolius
species known for the Eastern Arc forests
of Africa. Instead, it is comparable to
some small, West African forest treefrogs
such as H. zonatus and the larger H.
bobirensis (Schi0tz 1999).
The advertisement call of Hyperolius
kihangensis is unusual within the genus,
because of its peculiar tonal quality. A
few species, such as H. kachalolae from
Zambia and the West African H. viridigulosus,
show similar features in their
breeding calls. The calls of these species
are, however, clearly formed of a series of
harmonics. In the former, the fundamental
and the dominant frequency are the
same, while in the latter the harmonics
have a complex pattern of intensity maxima
(Schi0tz .1999).
Acknowledgements
These data are part of a herpetological research project authorized by COSTECH
(research permit # 98-028-CC-98-13). Special thanks are due to David Mover (Iringa)
and Prof. Kim Howell (University of Dar es Salaam) for helping us to obtain research
permits and for many useful suggestions. We wish to thank Paolo Pedrini, Ivan
Farronato and Martin Pickersgill for their comments on the preliminary version of the
manuscript
REFERENCES
CHARIF, R. A., MITCHELL, S. & CLARK, C. W. 1996: Canary 2.0 Users' manual. Cornell
Laboratory of Ornithology, Ithaca, New York, USA.
DUELLMAN, W.E. & TRUEB, L. 1986: Biology of Amphibians. Johns Hopkins University
Press, Baltimore & London.
MOREAU, R. E. 1935: A synecological study of Usambara, Tanganyka Territory, with particular
reference to birds. Journal of Ecology 23: 1-43.
SCHI0TZ, A. 1975: The Treefrogs of Eastern Africa. Steenstrupia, Copenhagen, Denmark.
232 pp.
SCHI0TZ, A. 1999: Treefrogs of Africa. Edition Chimaira, Frankfurt am Mein, Germany. 350
PP-
SCHI0TZ, A. & Westergaard, M. M. 2000: Notes on some Hyperolius (Anura: Hyperoliidae)

108 BJOla 3 1-2, 2002 MENEGON & SALVIDIO
from Tanzania, with supplementary information on two recently described
species. Steenstrupia 25: 1-9
WESTEGAARD, M. M., BRESCIANI, J. & BUDTZ, P. E. 2000: Structural aspects of white spots
on dorsal skin of Hyperolius kihangensis (Amphibia: Hyperoliidae). African
Journal of Herpetology 4: 73-77.

NAGY JOGER, GUICKING & WINK 3/1-2, 2OO2 109
Phylogeography of the European
Whip Snake Coluber (Hierophis)
viridiflavus as inferred from
nucleotide sequences of the
mitochondrial cytochrome b gene
and ISSR genomic fingerprinting
Zoltan Tamas NAGY1, Ulrich JOGER2, Daniela
GUICKING1 & Michael WINK1
1lnstitut fur Pharmazie und Molekulare Biotehnolgie, Universitat Heidelberg, Im
Neuenheimer Feld 364, D-69120 Heidelberg, Germany
2Hessisches Landesmuseum Darmstadt, Zoologische Abteilung, Friedensplatz 1, D-
64283 Darmstadt, Germany
Abstract
The intraspecific phylogeography of European Whip Snake Coluber (Hierophis) viridiflavus
was reconstructed using complete sequences (1117 bp) of the protein-coding
mitochondrial cytochrome b gene. C. (H.) gemonensis (Laurenti) and C. (H.) caspius
Gmelin were used as outgroups. Additionally, a microsatellite-based genomic fingerprint
method, ISSR, was employed to check for gene flow between populations. Two clearly
different genetic clades could be identified within European Whip Snake, a western one
occurring in France, Switzerland and Italy west of the Apennines, and an eastern one
found in Croatia, eastern and southern Italy. The latter one could be further subdivided
into three subgroups, two of which occur in southernmost Italy only (southern Calabria
and Sicily). These two distinct entities were probably formed during a long continuous
settlement of the European Whip Snake in these climatically favourable areas, whereas
the more northern populations experienced enormous shrinking of their ranges during
cold periods in the Pleistocene. Potential glacial refuge areas are discussed.
Key words: European Whip Snake, Coluber (Hierophis) viridiflavus, phylogeography
Received 16 November 2001; accepted 5 February 2002

110 >ta 3/1-2, 2002 NAGY ,JOGER, GUICKING & WINK
INTRODUCTION
The taxonomy and systematics of the
European Whip Snake, Coluber
(Hierophis) viridiflavus Lacepede, 1789,
and of the other species of the genus
Coluber (sensu lato), have been intensively
debated in the last centuries.
Although the European whip snake exists
exclusively in Europe (Heimes 1993,
Naulleau 1997), it shows remarkable
morphological variation. Many of these
characteristics can change from population
to population (Schatti & Vanni
1986), and therefore a multitude of morphological
forms (insular and colour
forms first of all) at species and subspecies
levels have been reported in the
literature (Suckow 1798, Bonaparte
1833, De Betta 1874, Boulenger 1893,
Mertens & Muller 1928, Mertens &
Wermuth 1960, Kramer 1971,
Capolongo 1984, etc.). While Bruno
(1975 & 1980) considers that two separate
taxa ("viridiflavus" and "carbonarius"
- on the basis of melanism) at species
level might be possible, Schatti & Vanni
(1986) among others question the existence
of subspecies - due to lack of evidence
in pholidotic and hemipenis-morphological
data. The occurence of the
Table 1. Tissue samples included in the present study.
* - no proper data available
Sample Species Country of origin
P7 (1) C. (Hjviridifloms Croatia
J19(5) " France
J36(3) " Italy
J39(5) " France
MO (5)
DOS (7)
DG9(7)
DG16(9)
DG17

NAGY JOGER, GUICKING & WINK Biota 3/i-a, 2002 111
Figure 1. Map with localities of specimens used for the analysis.
Switzerland, see Table 1 and Figure 1 for
details). In most cases a blood sample
was taken from the tail of living specimens
at the capture site (Joger & Lenk
1997); in addition, different tissue samples
were taken (a small piece of liver,
heart, or tail) from recently killed animals
(snakes that were run over by cars). The
distribution area of the European whip
snake (Heimes 1993, Naulleau 1997) was
more or less covered by sampling - with
the exception of the Western islands of
the Mediterranean Sea, such as Corsica,
Sardinia etc. Samples from the two closely
related species (Schatti 1988): Coluber
(Hierophis) gemonensis (Laurenti, 1768)
and Coluber (Hierophis) caspius (Gmelin,
1789), were analyzed as outgroups.
2. Isolation of DMA
Tissue samples were stored until processing
either in 80% ethanol or in EDTA
buffer (Arctander 1988). The isolation of
total genomic DMA was carried out with
proteinase K digestion, followed by
cleaning steps with phenol-chloroform
and guanidine-thiocyanate (Sambrook et
al. 1989). The extracted DNA was dissolved
and stored in Tris-EDTA (10 mM
Tris and 1mM EDTA).
3. PCR and nucleotide sequencing
The whole cytb gene was amplified with
polymerase chain reaction in an end volume
of 50 ul. PCR primers are listed in
Table 2. PCR was carried out under standard
conditions: initial denaturating (4
min at 94°C), 31-33 cycles with denaturating
(45 s, 94°C), elongation (2 min,
72°C) and annealing step (50 s, 41-
45°C), followed by a final elongation step
for 10 min at 72°C. PCR products were
stored at 4°C, and were directly used as
templates in a cycle sequencing reaction
(Lenk et al. 1999). Typical conditions
were: 3 min at 94°C, 26 cycles with
denaturating (30 s, 94°C) and annealing
and elongation step (45 s, 60°C). The
sequencing products were analyzed in an
automated sequencer ALFExpress II
(Amersham Pharmacia Biotech).
Sequences were checked for errors and

112 oiola 3/1-2,2002 NAGY rJOGERr GUICKING & WINK
Table 2. Primers used for PCR amplification and sequencing (cy: primer used in cycle
sequencing only).
Primer
L14724
smiA (L-14S46)
mtE (H-15556)
smtF (H-16060)
L-15570cy
H-15305cy
Source
Meyer eial. (1990)
after Kocher et al. ( 1989),
Leak &Wink (1997)
Lenketal. (2001)
modified after Wink (1995)
This study
This study
aligned manually.
The statistical analyses of the nucleotide
sequences were carried out with the
PAUP* 4.0b8 (Swofford 2001) programme
package using maximum parsimony
method with heuristic search and
bootstrap analysis.
4. ISSR-PCR fingerprinting
ISSR (Inter simple sequence repeat)
genomic fingerprinting was performed to
verify results obtained from mitochondrial
cytochrome b sequences and to evaluate
potential hybridisation. ISSR-PCR
uses single primers designed from short
tandem repeats to amplify stretches of
DNA between adjacent microsatellites.
Given that microsatellites are scattered
throughout the genome (Tautz & Renz
1984) and in such density that adjacent
microsatellites lie within the limits of Taq
polymerase processing, during ISSR-PCR
a large number of fragments is generated.
Once separated on polyacrylamide
gel, ISSR-PCR products appear as polymorphic
multiband fingerprints.
Five ISSR-primers that are known to generate
polymorphic fingerprints in several
vertebrate taxa were tested under varying
PCR conditions on Coluber
(Hierophis) samples. (CA)io- and
(GACA)4-primers gave most useful
results. The reactions were carried out
according to the protocol of Wink et al.
Sequence (S'-3')
CGA AGC TTG ATA TGA AAA ACC ATC
CAA CAT CTC AGC ATG ATG AAA CTT CG
AAT AGG AAG TAT CAT TCT GGT TTG AT
TCA GTT TTTGGT TTA CAA GAC CAA TG
GAY AAA ATC CCA TTY CAC CC
AAT GAT ATT TGT CCT CAT GG
(2000). In a PCR volume of 25 ul, 50-100
ng of total DNA were used as template,
plus 6 pmol (GACA)4 or 20 pmol (CA)io
primer, 1,5 mM MgCb, 0,1 mM of dCTP,
dGTP, and dTTP, 0,045 mM dATP, 1 uCi
33P-alpha-dATP, 2,5 ul 10x amplification
buffer and 1 unit Taq polymerase
(Amersham Pharmacia Biotech). PCR
programmes were set for 5 min at 94°C,
followed by 26 cycles of 30 s at 94°C, 20
s at annealing-temperature, and 50 s at
72°C. After completion, the temperature
was set to 72°C for 5 min and then lowered
to 4°C for further storage. Optimal
annealing-temperatures, as found out by
a 2°C temperature gradient PCR (40°C-
60°C), were 40°C for (CA)io-primer and
55°C for (GACA)4-primer. PCR products
were separated on a denaturing Sequagel
matrix at 65 W for 4 h and visualised by
autoradiography. Fingerprint patterns
were evaluated visually.
RESULTS
1. Analysis of cytochrome b sequences
The complete mitochondrial gene cytb
(1117 base pairs) was amplified from 24
samples, including the outgroups.
The phylogeographic analysis of Coluber
(Hierophis) viridiflavus revealed four haplotype
groups in two main clades, which
are clearly separated genetically and by
geographic areas (Figure 2). These clades
are supported with significant bootstrap

NAGY ,JOGERr GUICKING & WINK 3/1-2, 2002 113
values (96-100%).
Among our sequences 57 variable
nucleotides (5.10% of the cytb gene,
excluding outgroups) were found (Figure
3); the two main groups are separated by
- at least - 2.95% (33 different
nucleotides). They represent a Western
and an Eastern clade. The Western group
(clade W in Figure 2) includes animals
from France, Switzerland and Italy west
of the Apennines (i.e. Tuscany province
and Riviera coast). This group possesses a
quite uniform haplotype - only single
nucleotide mutation sites were detected -
which appears to be monomorphic without
further subdivisions.
The Eastern group is divided into further
subgroups: 1. Italy, east of the Apennines
and the island Krk (clade E on Figure 2),
2. southern Calabria, and 3. Sicily (both
of them represented in clade S). At the
Eastern border of the distribution area live
- sympatrically with C. (H.) gemonensis -
the Slovenian and Croatian populations
of C. (H.) viridiflavus. Our sample from
the island Krk shows the closest relationship
with the eastern Italian samples. In
contrast, the Southern (South-Eastern)
populations belong to a clearly separated
group: the South Calabrian and Sicilian
populations differ not only from the
northeastern ones, but also from each
other.
2. ISSR-PCR patterns
ISSR-primers (GACA)4 and (CA)io produced
informative fingerprints with several
polymorphic bands. ISSR profiles
were more diverse with the (GACA)4primer
(Figure 4) than with the (CA)ioprimer
(not shown). Both primers generate
a series of species-specific bands as
Figure 2. Phylogeny of the cytochrome b gene of Coluber (Hierophis) viridiflavus. Strict
consensus tree resulted from maximum parsimony analysis, tree length=251. Bootstrap
proportions >50 (100 replications) are presented.
100
cyt6-MP
100
W
99
Krk, CRO (P7)
Canossa, I (155}
F. Staffora, I (DG77)
T. Platano, I (DG66)
T. Platano, I (DG65)
N-Calabria, I (DG5S}
Torre S. Genaro, 1 (DG22)
Mte St. Angelo, I (DG21)
Lago di Lesina, I (DG17}
Lago di Lesina, 1 (DG16)
Grado, \(t>G9)
Grade, I (DCS;
9gj- Sicily, I (DG38)
63 PL Sicily. I (DG45)
I S-Calabria, S-Cala I (DG53)
Quincy, F (J19)
Canguers, F (Al)
R. L'AIlondon, CH (DG86)
Rapallo, I (DG7I)
Quincy, F (J40)
Quincy, F (J39)
Tuscany, I (J36)
C. (H.) gemonensis, CRO
C. (H.) cospius, OR

114 Biota 3 1-2. 3002 NAGY ,JOCER, GUICKING & WINK
Figure 3. Variable nucleotides of C. (H.) viridiflavus samples (cytochrome b gene with
position numbering). E: eastern group; W: western group; Sic: Sicily; Cal: South-
Calabria; var: variable sequences with single nucleotide mutations.
*c
*c
#C
*c_
SC
K
vi
vi
vi
vi
vi
vi
id
id
id
id
id
id
1111
111 1111222233 3344444445 5555666666 7778888999 9990000
2244458122 4569056811 3834566671 3457146889 0261447011 5790467
4527814709 7658487828 0127302577 1929829341 8934366658 5492721
E
E var
E~Sic
E Cal
W
W
var
GCACACGCGA TTTGGCTCGG
. . .G.TRT AT.T . .AC
. . .G. . .TA- .C...A.C. . .AC
ATGGG.A. .G
ATGGG.A.
TGTTATAGTC
R
GGACCAATTC
Y W
T.G.
G.C
ACGAGGCAAT ATGAGTC
D
Y
. . . GA. . . . . AG . . .
. G.A .C.
. .CAA.C.AA CA.CGA.AC. AA.T CT GT . .AA.G.C ... .A.T
G Y. CAA.C.AA CA.CGA.AC . AA.T. . CT . . GT. .AAYGMC . . . .A.T
... . irrn n/-n ,. . , ...
Figure 4. ISSR PCR fingerprint made with
, ° ,-.,„ . .. , , , , . .,
(GACA)4 primer. See text for details.
'
well as several group-specific bands
.... . ,. ° £, K , . ,
(Figure 4. indicated with numbers) that
s . ,' ,.„
support the differentiation into a western
(Figure 4, bands No. 1, 2, 3, 4) and an
eastern clade (bands No. 5: several bands
in a specific size range, to a certain extent
also No. 6, 7). Western and eastern
clades, as suggested by ISSR-analysis,
conform to those suggested by
cytochrome b sequences.
Furthermore, ISSR patterns suggest a
\ f--•- * hybrid status for some samples. This is
evident in samples J55 and DG77, both
from Northwestern Italy. According to
i cytochrome b sequences, both samples
7 belong to the eastern clade, whereas
_ _. ,. _^__ „ ISSR-profiles show both eastern clade
"* specific bands (Figure 4, no. 5) as well as
• "" rzim;^_ 6 western clade specific bands (in sample
•^^V^"-—-^35— 4 DG77 No 2 and 3^ in samp|e J55 No 1;
^"^•"•-^--"""^ '-. ' 2, and 3). These results suggest recent
gene flow between the eastern and western
clade in Northwestern Italy.
The separation of the Sicilian and southern
Caiabrian samples (DG38, DG45 and
DG53, respectively) from the rest of the
eastern clade is less evident according to
*•==»* ~ ISSR profiles than it is according to
cytochrome b sequences. The band pattern
of the Sicilian sample DG45 may
show genetic introgression from the
west, too (Figure 4, bands No. 2, 3).
DISCUSSION
The sequence analysis (confirmed by

NAGY ,JOGER, GUICKING & WINK Biota 3/i-a, 2002 115
ISSR-PCR fingerprinting) unambiguously
shows that there are discrete phylogeographic
units within the distribution range
of the European Whip Snake. Spatial and
historic factors might be responsible for
their isolation.
Two bigger groups are apparently separated
on the genetic level: The Western
and Eastern populations are divided by
the mountain ranges of the Alps and the
Apennines acting as strict geographic
barriers (Figure 1). Apparently this species
does not cross mountain ranges easily
(though it has been reported from elevations
above 1,800 m [Heimes 1993]).
Results based on cytb sequences support
the conclusion that the Western group
might be monotypic; however, no samples
from the Western Mediterranean
islands (see above), or from the Northern
border of the distribution zone (e.g.
Luxembourg) were available for our
research. An important aspect is that the
distribution of this colubrid species is continuous
between France and the Western
foot of the Apennines; it includes a recent
connection along the narrow coastal strip
of Cote d'Azur and Riviera di Ponente.
The mitochondria! sequences confirm a
close relationship between Coluber
(Hierophis) gemonensis and Coluber
(Hierophis) viridiflavus. This result suggests
also a Western-Eastern divergence:
Coluber (Hierophis) viridiflavus is representative
of the central Mediterranean,
while Coluber (Hierophis) gemonensis of
the eastern Mediterranean, respectively.
In addition, it is also possible that both of
them share a common ancestor with C.
(H.) caspius and C. (H.) jugularis, as
Schatti (1988) believes.
As the sister species are distributed in the
Balkans, it is justified to assume the geographic
origin of C. (H.) viridiflavus at the
eastern edge of its distribution range.
Many eastern European species that
reach Italy do not cross the Apennines; a
vertebrate example is the Hooded Crow
Corvus corone comix. In the turtle Emys
orbicularis, genetically different subspecies
inhabit eastern and western Italy
(Lenketal. 1999).
The population living on Krk island - at
the eastern border of the distribution
zone - shows a petty divergence compared
to Italian populations which belong
to the eastern group. A continuous presence
of the species during the Pleistocene
is, however, unlikely in this northern part
of its range. Therefore it is probably
derived from a southern Italian stock (The
Balkans' own glacial refugia were probably
occupied by C. (H.) gemonensis).
North-South movements of such reptiles
along the coasts of Italy are unhindered.
The existence of South(east)ern subgroup^)
is, however, again paralleled by
Emys orbicularis (Lenk et al. 1999) as well
as by the snakes of the Elaphe longissima
complex (Lenk & Wuster 1999). Such
common zoogeographical patterns can
be traced back to the glacial zoogeographic
history in all likelihood (Lenk et
al. 1999, Schatti 1988).
The southern Italian and Sicilian regions
could have functioned as refugia during
cold periods of the Pleistocene. A refuge
for the eastern group could have been at
the Gulf of Taranto and in Calabria. Our
two Calabrian samples belong to slightly
different haplotypes: DG53 ("South"),
and DG55 ("North"), respectively. Both
are different from the Sicilian sample. This
variation could be explained by a longer
continuous evolutionary history in these
possible glacial refuges. A possible refuge
for the western group could have been in
the region of Naples-Salerno. However,
we could not check samples from that
area.
The evolution of the southern populations
could also continue unhindered during
glacial periods. This may have resulted
in higher genetic differences, in other
words, in a mosaic of genetically different
forms. A few of these forms (in fact only

116 Biota 3/i-a, 2002 NACY ,JOGER, GUICKING & WINK
one on each side of the Apennines)
would have used the climatically favorable
interglacials to spread into more
northern areas. Pleistocene fossils attributed
to C. (H.) viridiflavus have been
found in southern Germany and Austria;
in the Pliocene, this species reached the
territories of today's Czech Republic and
Poland (Ivanov 1997, Schatti 1988).
Climatic conditions during the glacial
periods forced it to retreat into the above
named refuges in southern Italy. After the
last glacial period, re-immigration to the
northern Adriatic region (eastern group)
as well as to France and Belgium (western
group) would have happened in a comparatively
short time, which may not
have allowed a significant genetic differentiation
of northern populations to
occur.
ISSR-PCR has useful applications for
examining group specificity and
hybridization events. Because of its
detection of nuclear genetic rearrangements,
it provides a valuable complement
to data obtained from the maternally
inherited mitochondrial genome. It is
especially powerful in detecting
hybridization events (Wink et al. 2000).
In our study, it allowed us to re-examine
the separation of Coluber (Hierophis)
viridiflavus samples into a western and an
eastern clade, which is strongly supported
by ISSR-analysis, as well as suggesting a
gene flow between both groups in
Northwestern Italy. However, ISSR-fingerprints
provided less evidence for the
separation of Sicilian and South Calabrian
samples from other eastern clade samples
than is suggested by cytochrome b
sequence data.
Because of the evident, but not complete,
genetic isolation, we recommend once
more the recognition of two subspecies
of the European whip snake sensu
Mertens & Wermuth (1960). If further
studies in the contact zones of both subspecies
prove that an intrinsic genetic
barrier exists, the two main clades could
even be raised to species level (see Joger
et al. 1998). The applicable name for the
western populations is Coluber
(Hierophis) viridiflavus viridiflavus
Lacepede, 1789 (terra typica: southern
France). The Eastern stocks were traditionally
assigned to the subspecies
Coluber (Hierophis) viridiflavus carbonarius
Bonaparte 1833 (nomen conservandum,
terra typica restricta Mertens &
Muller 1928: Monti Euganei, Padua,
Italy); however, the subspecific name has
no more (or has far less) ethymological
importance (an allusion to the melanism).
The border between both subspecies is
the Apennines. The South Italian populations
are less differentiated within the
Eastern group (1.16% [13 different
nucleotides] and 1.34% [15] by
Calabrian and Sicilian samples, respectively,
at cytb level), so in our view taxonomical
consequences are not justified.
This, however, would mean that older
names apply to the eastern subspecies:
Coluber (Hierophis) profulax Costa, 1828
(terra typica Aspromonte, southern
Calabria) or C. (H.) xanthurus
Rafinesque-Schmaltz, 1810 (terra typica
Sicily). An even older name is C. (H.) sardus
Suckow 1798 (Terra typica Sardinia).
As we have not studied Sardinian material,
we cannot decide to which form it
applies. In all cases, the morphology of
the European whip snake has to be studied
in more detail in order to find reliable
characters to define subspecies.
Acknowledgements
We would like to thank Toni Amann, Anna Hundsdorfer, Luc Legal, Peter Lenk and
Werner Mayer for providing tissue samples, as well as Edoardo Razzetti for helpful comments.
This study was financially funded by DAAD (research scholarship for the first
author) and DFG (Jo-134-7 and Wi-719/18).

NAGY JOGER, GUICKING & WINK Biota 3/1-2, 2002 117
REFERENCES
ARCTANDER, P. 1988: Comparative studies of avian DNA by restriction fragment polymorphism
analysis. J. Ornithol. 129: 205-216.
BONAPARTE, C.L. 1833: Iconografia della fauna Italica (1832-1841). ll(5), Roma. Tipogr.
Salviucci. iv+264 pp., 54 pis.
BOULENGER, G.A. 1893: Catalogue of the snakes in the British Museum, I., London. Brit.
Mus. (Nat. Hist), 448 pp., 28 pis.
BRUNO, S. 1975: Note riassuntive sull'erpetofauna dell'isola di Montecristo (Archipelago
Toscano). Studi sulla fauna erpetologica italiana, XX. Lav. Soc. ital. Biogeogr. (n.s.)
V: 98 pp., 45 figs.
BRUNO, S. 1980:1 serpenti del Veneto. I. A/lorfologia, tassonomia, geonemia. Lav. Soc. ven.
Sci. nat., 5 (suppl.): 70 pp.
CAPOLONGO, D. 1984: Note sull'erpetofauna pugliese. Atti Soc. ital. Sci. nat. Milano civ.
Stor. nat. Milano 125: 189-200.
DE BETTA, E. 1874: Fauna d'ltalia. In: Parte V, Rettili ed Anfibi. Vallardi, R, Ed, Milano.
HEIMES, P. 1993: Coluber viridiflavus Lacepede, 1789 - Gelbgriine Zornnatter. In: Bohme,
W. (ed.): Handbuch der Reptilien und Amphibien Europas, Vol. 3/I. Wiesbaden:
177-198.
IVANOV, M. 1997: Vlvoj kenozoicke hadi fauny Evropy. In: Hladilove, S. (ed.): Dynamika
vztahu marinnfho a kontinentalm'ho prostredi. Brno: 59-91.
JOGER, U. & LENK, P. 1997: Entnahme und Behandlung von Blutproben fur molekulargenetische
Untersuchungen in der Feldherpetologie. In: Henle, K. & Veith, M.
(eds): Naturschutzrelevante Methoden der Feldherpetologie. Mertensiella,
Rheinbach 7: 329-340.
KOCHER, T.D., THOMAS, W.K., MEYER, A., EDWARDS, S.W., PAABO, S., VILLABLANCA,
F.X. & WILSON, A.C. 1989: Dynamics of mitochondrial DNA evolution in animals:
Amplification and sequencing with conserved primers. PNAS 86: 6196-
6200.
KRAMER, E. 1971: Revalidierte und neue Rassen der europaischen Schlangenfauna. Lav.
Soc. ital. Biogeogr. (n.s.) I (1970): 667-675.
LENK, P., FRITZ, U., JOGER, U. & WINK, M. 1999: Mitochondrial phylogeography of the
European pond turtle, Emys orbicularis (Linnaeus 1758). Mol. Ecol. 8:1911-1922.
LENK, P., KALYABINA, S., WINK, M. & JOGER, U. 2001: Evolutionary relationships among
the true vipers (Reptilia: Viperidae) inferred from mitochondrial DNA sequences.
Mol. Phyl. Evol. 19: 94-104.
LENK, P. & WINK, M. 1997: A RNA/RNA heteroduplex cleavage analysis to detect rare
mutations in populations. Mol. Ecol. 6: 687-690.
LENK, P. & WUSTER, W. 1999: A multivariate approach to the systematics of Italian rat
snakes of the Elaphe longissima complex (Reptilia: Colubridae): Revalidation of
Camerano's Callopeltis longissimus var. lineata. Herpetol. J. 9: 153-162.
MERTENS, R. & MULLER, L. 1928: Liste der Amphibien und Reptilien Europas. Abh. senckb.
naturf. Ges. 41: 1-62.
MERTENS, R. & WERMUTH, H. 1960: Die Amphibien und Reptilien Europas (Dritte Liste).
Frankfurt am Main, 264 pp.
MEYER, A., KOCHER, T.D., BASASIBWAKI, P. & WILSON, A.C. 1990: Monophyletic origin
of Lake Victoria cichlid fishes suggested by mitochondrial DNA sequences. Nature
347: 550-553.
NAGY, Z.T. & KORS6S, Z. 2001: Morphological studies on grass snakes, Natrix natrix (L.) in
a Hungarian population. In: Lymberakis, P., Valakos, E., Pafilis, P. & Mylonas, M.
(eds): Herpetologia Candiana, S.E.H., Irakleio, 2001: 107-114
NAULLEAU, G. 1997: Coluber viridiflavus Lacepede, 1789. In: Case, J.-P. (ed.): Atlas of

118 Biota 3/1-2.2002 NAGY ,JOGER, GUICKING & WINK
Amphibians and Reptiles in Europe. Paris: 342-343.
SAMBROOK, J., FRITSCH, E.F. & MANIATIS, T. 1989: Molecular Cloning: a Laboratory
Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York.
SCHATTI, B. 1988: Systematik und Evolution der Schlangengattung Hierophis Fitzinger,
1843 (Reptilia, Serpentes). Dissertation, Zurich.
SCHATTI, B. & VANNI, S. 1986: Intraspecific variation in Coluber viridiflavus Lacepede,
1789, and validity of its subspecies (Reptilia, Serpentes, Colubridae). Revue
suisseZool. 93:219-232.
SUCKOW, G.A. 1798: Anfangsgrunde der theoretischen und angewandten
Naturgeschichte der Thiere. Leipzig, Weidmann'sche Buchhandlung.
SWOFFORD, D.L. 2001: PAUP*. Phylogenetic Analysis Using Parsimony (*and Other
Methods), Version 4.0b8. Sinauer Associates, Sunderland.
TAUTZ, D. & RENZ, M. 1984: Simple sequences are ubiquitous repetitive components of
eukaryotic genomes. Nucleic Acids Res. 25: 4127-4138.
WINK, M. 1995: Phylogeny of Old and New World vultures (Aves: Accipitridae and
Cathartidae). Inferred from nucleotide sequences of the mitochondrial
cytochrome b gene. Z. Naturforsch. Journal of Bioscience 50c: 868-882.
WINK, M., GUICKING, D. & FRITZ, U. 2000: Molecular evidence for hybrid origin of
Mauremys iversoni Pritchard et McCord, 1991, and Mauremys pritchardi
McCord, 1997 (Reptilia: Testudines: Bataguridae). Zool. Abh. Mus. Tierkd.
Dresden 51: 41-49.

1
PASTOREILI, LAGHI & SCARAVELLI Biota 3/1-2. 2002 119
Seasonal activity and spatial
distribution of a Speleomantes
italicus population in a natural
cave
Christian PASTORELLI1, Paolo LAGHI2 & Dino
SCARAVELLI3
Y Cerchia di S. Egidio 2205, 47023 Cesena (FC), Italy
E-mail: pastorellic@libero.it
2v. Bruno C. Garibaldi 22, 47100 Forli, Italy
E-mail: spelerpes@libero.it
3Riserva Naturale Orientata e Museo di Onferno, p. Roma 1, 47855 Gemmano (RN),
Italy
E-mail: rnoonf@tin.it
Abstract
The biology of a Speleomantes italicus population inhabiting the natural cave "Grotta
del Tritone" has been studied since 1998. The cave opens at 810 m a.s.l. and it is located
in the northern Apennine slope (Forli-Cesena province, Emilia-Romagna region,
Italy). Data on environmental parameters (inside and outside temperature, relative
humidity and light intensity) and activity of Speleomantes italicus (number, size and
location of the individuals) have been collected from February 1999 to January 2001 by
monthly surveys. The hypogean observable activity of 5. italicus in the study site
showed a strong seasonal pattern, with the highest number of specimens in late spring
(May) and early fall (September). The authors also analysed seasonal variation of spatial
distribution as a function of temperature, humidity and light intensity. Salamanders
were observed only in the first part of the cave (up to about 25 meters from the
entrance); the animals also seemed to prefer the lighted parts of the cave, while juveniles
were found at a shorter distance from the entrance than adults. During summer,
salamanders were active deeper in the cavity than in fall or spring, and in winter they
were found near the entrance and sometimes outside. Superficial activity was scarce
and close to cave entrance.
Key words: population ecology, seasonal activity, spatial distribution, Plethodontidae,
Speleomantes italicus
Received 3 August 2002; accepted 26 September 2002

120 Biota 3/i-a, 2002 PASTORELLI, LAGHI & SCARAVELLI
INTRODUCTION
European Plethodontid salamanders of
the genus Speleomantes are fully terrestrial
and live in moist and cool environments
such as caves and soil interstices.
They are found from south-eastern
France to central Italy and on the island
of Sardinia (Lanza et al. 1995). To date,
seven species (three continental and four
insular) and one insular subspecies of
Speleomantes have been recognised on
the basis of morphological and genetic
data (Lanza et al. 1986, 1995, 2001,
Nascetti et al. 1996). Notwithstanding
the great interest that this genus holds,
studies on eco-ethology, population
dynamics and life history of
Speleomantes started only recently and
regarded few species (Cimmaruta et al.
1999, Mutz 1998, Pastorelli et al. 2001,
Salvidio 1990, 1991, 1992, 1993a,
1993b, 1996, 1998; Salvidio et al. 1994,
Voesenek et al. 1987).
Speleomantes italicus is a species endemic
to the northern and central Apennines
(Lanza et al. 1995) and is found in natural
and artificial caves and in interstitial
habitats.
The biology of a Speleomantes italicus
population inhabiting a natural cave in
the northern Apennines has been studied
since 1998 (Pastorelli et al. 2001). This
paper reports data about seasonal activity
and spatial distribution of the salamanders
as a function of environmental
parameters studied from February 1999
to January 2001.
MATERIALS AND METHODS
The study site is a natural cave called
"Grotta del Tritone" [43°53'52" N -
0°29'02"W (Rome)] located on the
northern Apennine slope (Forli-Cesena
province, Emilia-Romagna Region, Italy).
The cave opens north-eastward at 810 m.
a.s.l., in marl and sandstones. It reaches a
total extension of about 50 meters, on a
sub-horizontal horizon. (Figure 1). The
dominant vegetation around the study
site is oak-tree woodland, mainly composed
of coppices of Quercus cerris and
Ostrya carpinifolia with a poor leaf soil.
Since September 1998 the study site was
sampled monthly (except June 1999) and
records included individual data as well as
environmental data.
Salamanders were searched for by two
persons outside the cave near its
Figure 1. The cave "Grotta del Tritone". On the left the black circle indicates the location
of the study site in Italy and in grey is pointed out Speleomantes italicus range.
GROTTA DHL TRiTONE

PASTORELLI, LAGHI & SCARAVELLI Biota 3 1-2, 2OO2 121
entrance, and inside on the walls and
floor and, where possible, on the cave
vault for about one hour per session. For
each specimen captured the following
data were recorded: date and hour of
observation, sex, distance from cave
entrance and from cave floor (in meters).
All specimens with a body length wider
than the smallest male (41 mm) with a
mental body gland (cf. Lanza 1959) were
considered to be adult salamanders.
Environmental data include weather conditions,
temperatures and relative humidity.
The latter two were measured at the
entrance (Te1, URe1) and at the following
distances from it: 5 meters outside the
cave (Te2, URe2) and 7 (Ti1, URi1), 15
012, URi2) and 21 metres (Ti3, URi3)
inside. Two minimum-maximum mercury
thermometers were placed to record temperatures
between consecutive samples,
respectively at Te2 and Ti1 sampling locations.
Finally, in July 1999, March, June
and December 2000 maximum light
intensity was recorded for each linear
meter inside the cave to the point at
which light reached the 0.0 lux instrumental
value.
RESULTS
Environmental parameters
During the study period, the mean minimum
temperature outside the cave varied
from -11 °C to 15°C, while the mean minimum
internal temperature varied from
3°C to 10°C. During the same period
maximum temperatures varied from 10°C
to 37°C outside and from 8°C to 11°C
inside. The highest variation was
observed for Te 2 (mean = 13.129; SD =
7.467) followed by Te 1 (mean = 12.000;
SD = 3.949), Ti 3 (mean = 9.574; SD =
1.878), Ti 1 (mean = 8.973; SD = 1.489),
and Ti 2 (mean = 8.914; SD = 1.269).
Inside the cave the relative humidity varied
from 80% to 100%, while outside it
Figure 2. Light intensity variations recorded inside the cave (deep zone)
VD r-- oo
« •— — — — cs cs
Distance of the sampling point from the entrance (m)
• 12-Mar-OO: overcast sky -02-Jun-OO: clear sky

122 Biota 3/1-2,2002 PASTORELLI, LAGHI & SCARAVELLI
fluctuated between 30 % and 100%. The
highest variation was recorded for URe 2
(mean = 68.714; SD = 21.504) followed
by URe 1 (mean = 78.579; SD = 15.222),
URi 3 (mean = 94.389; SD = 4.680), URi
2 (mean = 94.762; SD = 4.230), and URi
1 (mean = 95.143; SD = 3.525).
Light intensity decreased to about 10 lux
at 7 metres inside the cave, then it gradually
reached 0.0 lux at 21 m from the
entrance (Figure 2).
Seasonal activity
The total number of salamanders captured
was 204 in the year 1999 and 181
in the year 2000.
The highest number of animals was captured
in May (74 and 45 specimens in
1999 and 2000 respectively) and
September (45 and 36 specimens in 1999
and 2000 respectively). The lowest num-
ber of active salamanders was recorded in
December 2000 and January 2001 (only
3 active specimens in both cases) and no
salamanders were found in January 2000.
This activity pattern did not differ significantly
between the two years (Mann-
Whitney test T: 135.5 and P: 0.854).
Activity of females, males, and both sexes
pooled did not vary meaningfully
between the two years (Mann-Whitney
test respectively with P>0.5 in all cases).
Salamanders activity was correlated with
environmental parameters recorded during
each sampling session. The number of
active juveniles, females, males, and
adults was not significantly correlated to
the relative humidity recorded inside and
outside the cave (Spearman's correlation
coefficient with P > 0.05 in all cases). The
abundance of active salamanders was
significantly correlated to internal tern-
Figure 3. Regression between number of specimens captured and temperature measured
7 meters far from the entrance.
8 _ 9 . 1

PASTORELLI, LAGHI & SCARAVELLI Biota 3/1-2,2002 123
peratures (Figure 3), except for males, for
which abundance was significantly correlated
with Ti 2 and Ti 3 only (P < 0.05 in
all cases). A meaningful correlation was
also found between the number of active
salamanders and external temperatures
(P< 0.05) except for males (P > 0.05);
juvenile abundance was also not significantly
correlated to Te2 (P > 0.05).
Spatial distribution
Taking into account the whole body of
captures, the animals were found at a
mean distance of 6.5 m from the
entrance, and 51.3% of captures were
made within the first 7 m of cave extension.
Only 5 captures (1.4%) were made
1 m outside the cave and only one
(0.3%) more than 25 metres inside. This
distribution pattern did not vary qualitatively
between the two years.
Juveniles were found at a mean distance
of 3.9 m from the entrance, and 88.5%
of observations were made within the
first 7 m. Only 2.3% of observations (4
captures) were recorded 1 m outside the
cave.
The mean distance from the entrance of
adults was 8.6 m; 66.7% of observations
were recorded within 4 to 12 m from the
entrance and only 8.5% were found
within the first 4 m from the entrance.
Only one adult salamander (0.6%) was
found more than 25 m from the entrance
and only one male specimen (0.6%) 1 m
outside the cave.
The mean distance from the entrance
recorded for females was 8.6 meters,
while it was 8.8 meters for males.
Significant variations were recorded
between female and male distribution in
the year 1999 (P < 0.05), but not in 2000
(P > 0.05) or when years were pooled (P
> 0.05). Superficial activity outside the
cave was observed only in December
1999 and November 2000, close to the
entrance, when temperatures inside and
outside the cave were very similar. There
Figure 4. Seasonal variations of spatial distribution. Standard errors are reported.
13,0
12,0
11,0
10,0
9,0
8,0
7,0
6,0
5,0
4,0
3,0
2,0
',0
0,0
Spring Summer FaH Winter
D Juveniles d Females D Males

124 Biota 3/i-a, 2002 PASTORELLI, LAGHI & SCARAVELLI
was no significant correlation between
the spatial distribution of females and
males and light intensity. However, the
spatial distribution of all specimens captured
inside the cave was meaningfully
correlated to light intensity (Spearman's
correlation coefficient, P < 0.01 in all
cases except for females: P < 0.05).
The highest mean distance from the
entrance was recorded in summer (7.6
m), followed by fall (6.6 m), spring (6.4
m), and winter (3.9 m) respectively. This
pattern did not show variations between
females and juveniles, while males were
found deeper in the cave in spring (Figure
4).
DISCUSSION
Environmental parameters
Inside the study cave, environmental variations
were lower than those recorded
outside; thermal and hygrometric differences
between the various sampling
points inside the cave were on average
low and generally due to different microhabitat
conditions. During the study period,
relative humidity inside the cave
remained high throughout the year.
Because of the direction of the cave's
exposure, light penetrated only the first
part of the cavity; then its intensity gradually
decreased, becoming 0.0 at 21
meters inside.
Seasonal activity
Speleomantes italicus hypogeal observable
activity showed great variations during
the year, with the highest number of
salamanders captured in May and
September and the lowest in January and
December. These data are in contrast
with those recorded for Speleomantes
strinatii in an artificial tunnel. In fact, this
latter population was more active in July
and August (Salvidio et al. 1994).
Salamander activity pattern was similar
between years, but in 2000 a smaller
number of salamanders was captured.
This could be due to the variation in environmental
factors between years, or perhaps
to human disturbance; further data
are needed to clarify this observation.
Our data show a direct influence of inside
temperatures on salamander activity;
again, these results appear different from
those obtained for S. strinatii, in which
the influence of internal temperatures
seemed to be indirect, as the highest correlation
coefficients were obtained for
temperatures recorded outside the tunnel
or near its entrance (Salvidio et al. 1994).
Spatial distribution
Plethodontid salamanders are usually
found in the twilight tract of a cave near
its entrance (Cimmaruta et al. 1999,
Lanza 1946, 1999, Pastorelli et al. 2001,
Salvidio et al. 1994). As already observed
for S. strinatii (Salvidio et al. 1994), in S.
italicus juveniles concentrate near the
entrance of the cave and adult specimens
in deeper zones.
The hypothesis of a spatial segregation
between juveniles and adults due to
behavioural interference is persuasive,
and it has already been demonstrated in
the American plethodontid genus
Desmognathus (Colley et al. 1989) and
verified by Salvidio & Pastorino (2002)
even for Speleomantes strinatii.
The overall spatial distribution of salamanders
was positively correlated with
light intensity. According to Roth (1976),
Speleomantes are able to feed in complete
darkness using only chemical cues.
Notwithstanding this statement, is not
unlikely that, when possible, cave salamanders
prefer to forage in the twilight
by means of visually guided prey catching
behaviour.
Similar to S. strinatii (Salvidio et al. 1994),
even in a S. italicus population external
activity was scarce and in proximity to the
cave entrance; the poor leaf soil produced
by the coppiced wood near the cave
probably cannot provide enough mois-

PASTORELLI, LAGHI & SCARAVELLI Biota 3/i-a, 2002 125
ture to allow salamanders to move outside,
except during very humid periods.
Salamander distribution inside the cave
varies seasonally, as already observed by
Lanza (1946). In winter, when external
conditions are similar to internal ones,
salamanders are active at a short distance
from the entrance, while in fall and spring
the animals are found at a greater distance.
In summer the animals are active
deeper inside the cave, where environmental
variations are lower.
Acknowledgements
The authors thank Benedetto Lanza and Sebastiano Salvidio for useful suggestions.
REFERENCES
CIMMARUTA, R., FORTI, G., NASCETTI, G. & BULLINI, L. 1999: Spatial distribution and
competition in two parapatric sibling species of European plethodontid salamanders.
Ethology Ecology & Evolution, 11: 383-398.
COLLEY, S. A., KEEN, W. H. & REED R. W. 1989: Effects of adult presence on behav- iour
and microhabitat use of juveniles of a Desmognathine salamander. Copeia, 1989
(1):1-7.
LANZA, B. 1946: L'Hydromantes Gistel in Toscana e notizie sui suoi costumi (Amphibia;
Caudata; Pletodontidae). Archo zoologico italiano, 31: 219-237.
LANZA, B. 1959: II corpo ghiandolare mentomiero dei "Plethodontidae" ("Amphibia,
Caudata"). Monitore zool. Ital. 67: 15-53.
LANZA, B., NASCETTI, G. & BULLINI, L. 1986: A new species of Hydromantes from Eastern
Sardinia and its genetic relationships with the other Sardinian Plethodontids
(Amphibia, Urodela). Bollettino Museo Regionale di Scienze Natural!, Torino 4:
261-289.
LANZA, B., CAPUTO, V., NASCETTI, G. & BULLINI, L. 1995: Morphologic and genetic studies
of the European plethodontid salamanders: taxonomic inferences (genus
Hydromantes). Bollettino Museo Regionale di Scienze Naturali, Monografie XVI,
Torino.
LANZA, B., LEO, P., FORTI, G., CIMMARUTA, R, CAPUTO, V. & NASCETTI, G. 2001:
Descrizione preliminare dello Speleomantes imperialis sarrabusensis subsp. n.
(Amphibia: Caudata: Plethodontidae). Atti del III Congresso Nazionale S.H.I.,
Pianura 13: 83-84.
MUTZ, T. 1998: Haltung und Zuchtdes Sardischen Hohlensalamanders Hydromantes imperialis
(Stefani, 1969) und einige Beobachtungen zur Okologie der Europaischen
Hohlensalamander. Salamandra 34: 167-180.
NASCETTI, G., CIMMARUTA, R., LANZA, B. & BULLINI L. 1996: Molecular Taxonomy of
European Plethodontid Salamanders (Genus Hydromantes). Journal of
Herpetology 30: 161-183.
PASTORELLI, C, LAGHI, P. & SCARAVELLI, D. 2001: Studi preliminari sull'ecologia di
Speleomantes italicus (Dunn, 1923) nell'Appennino Tosco-Romagnolo. Atti del
III Congresso Nazionale SHI (Pavia, 2000), Pianura 13: 347-351.
ROTH G., 1976: Experimental analysis of the prey catching behaviour of Hydromantes italicus
Dunn (Amphibia, Plethodontidae). J. Comp. Physiol. 109: 47-58
SALVIDIO, S. 1990: Regime alimentaire d'une population epigee de Speleomantes ambrosii
(Caudata, Plethodontidae) de la Ligurie Centrale (Italie septentrionale). Bulletin
de la Societe Herpetologique de France 54: 69-72.
SALVIDIO, S. 1991: Habitat ed attivita stagionale delle popolazioni interstiziali di
Speleomantes ambrosii nell'alta Val Bisagno (Liguria Centrale). Rivista

126 Biota 3/1-2,2002 PASTORELLI, LAGHI & SCARAVELLI
Piemontese di Storia Naturale 12: 69-74.
SALVIDIO, S. 1992: Diet and food utilization in a rock-face population of Speleomantes
ambrosii (Amphibia, Caudata, Plethodontidae). Vie Milieu 44: 35-39.
SALVIDIO, S. 1993a: Life history of the European Plethodontid Salamander Speleomantes
ambrosii (Amphibia, Caudata). Herpetological Journal 3: 55-59.
SALVIDIO, S. 1993b: Struttura di popolazione del geotritone Speleomantes ambrosii. Suppl
Ric. Biol. Selvaggina XXI: 517-520.
SALVIDIO, S.r LATTES, A., TAVANO, M., MELODIA, F. & PASTORINO, M. V. 1994: Ecology
of a Speleomantes ambrosii population inhabiting an artificial tunnel. Amphibia-
Reptilia, 15:35-45.
SALVIDIO, S. 1996: L'ecologia del Pletodontidi europei: stato delle ricerche sul geotritone
Speleomantes ambrosii. Atti del 1 ° Convegno italiano di Erpetologia montana.
Studi trentini di Scienze Naturali, Acta Biologica 71: 133-136.
SALVIDIO, S. 1998: Estimating abundance and biomass of a Speleomantes strinatii
(Caudata, Plethodontidae) population by temporary removal sampling.
Amphibia-Reptilia 19: 113-124.
SALVIDIO, S. & PASTORINO, M.V. 2002: Spatial segregation in the European plethodontid
Speleomantes strinatii in relation to age and sex. Amphibia-Reptilia 23: 505-
510.
THORN, R. 1969: Les salamandres d'Europe d'Asie et d'Afrique du Nord. Paul Lechevalier,
Paris.
VOESENEK, L. A. C. J., ROOY (VAN), P. T. J. C. & STRIJBOSCH, H. 1987: Some autoecological
data on the Urodeles of Sardinia. Amphibia-Reptilia 8: 307-314.

PASTORELLI, LAGHI & SCARAVELLI Biota 3/1-2.2002 127
Speleomantes antipredator
strategies: a review and new
observations
Christian PASTORELLI1, Paolo LAGHI2 & Dino
SCARAVELLI3
Y Cerchia di S. Egidio 2205, 47023 Cesena (FC), Italy
E-mail: pastorellic@libero.it
2v. Bruno C. Garibaldi 22, 47100 Forli, Italy
E-mail: spelerpes@libero.it
3Riserva Naturale Orientata e Museo di Onferno, p. Roma 1, 47855 Gemmano (RN),
Italy
E-mail: rnoonf@tin.it
Abstract
In the genus Speleomantes some antipredator adaptations are known, such as aposematic
coloration, noxious skin secretions, body elevation and tail undulation, and immobility.
During a long-term ecological study of Italian Cave Salamander in the cave
"Grotta del Tritone" in the northern Apennine mountains (Forll-Cesena province,
Emilia-Romagna region, Italy), the authors observed some adult Speleomantes italicus
displaying biting behaviour when taken by hand or forceps at the tail tip. Biting, as a
defensive behaviour, is known for a number of plethodontid salamanders such as
Aneides, Desmognathus, Gyrinophilus and Plethodon, but it has never before been
observed and described in European plethodontid salamanders. Moreover, we observed
the first case of damage to human skin by tail base gland secretions of Speleomantes
supramontis, supporting the hypothesis of a defensive role for the tail base glands in the
genus Speleomantes.
Key words: Antipredator strategies, defensive biting behaviour, noxious skin secretions,
Speleomantes, Plethodontidae.
Received 3 August 2002; accepted 26 September 2002

128 Biota 3/1-a, -M PASTORELLI, LAGHI & SCARAVELLI
INTRODUCTION
Prey may respond evolutionary to predator
pressure either by removing themselves
from the foraging microhabitat of
the predators (predator avoidance mechanisms)
or by reducing the probability of
successful predation when they are within
the perceptual field of the predators
(antipredator mechanisms) (Brodie et al.
1991). Antipredator mechanisms are
widely found in plethodontid salamanders.
Aposematic coloration, noxious skin
secretions, body elevation, tail undulation,
immobility, biting, rapid escape
movements, tail autotomy, and other
mechanisms have been extensively studied
among the American members of this
family (Brodie 1977, 1983, Brodie et al.
1989, Garcia-Paris & Deban 1995,
Labanick 1984). However, antipredator
strategies of European plethodontid
Speleomantes are still poorly known.
MATERIALS AND METHODS
A long-term ecological study of Italian
Cave Salamander has been carried out
since 1998 in the cave "Grotta del
Tritone" [43°53'52" N - 0°29'02"W
(Rome)], in the northern Apennine
mountains (Forli-Cesena province, Emilia-
Romagna region, Italy). The aim of this
research is to investigate severa aspects
of the life history of Speleomantes italicus,
such as seasonal activity, spatial distribution,
individual displacements, population
structure and size, diet, growth
rates and reproductive behaviour
(Pastorelli et al. 2001). During field work
the authors observed some adult
Speleomantes italicus displaying biting
behaviour when taken by hand or forceps
at the tail tip. This defensive behaviour
Figure 1. The observed sequence of the defensive biting behaviour displayed by
Speleomantes italicus (explanation in the text).

PASTORELLI, LAGHI & SCARAVELLI Biota 3/i-a, 2002 129
was filmed with a digital video camera.
During another sampling session conducted
in Sardinia in 1999, the authors
observed the first case of damage to
human skin by tail base gland secretions
of Speleomantes supramontis.
RESULTS AND DISCUSSION
The biting behaviour was observed in
four Speleomantes italicus adult females.
(A) The animal, caught at the tail tip, was
hanging from forceps or hands and rested
immobile for a few seconds.
(B) Then it tried to free itself with rapid
coiling-uncoiling movements, sometimes
while urinating or producing noxious skin
secretions mainly from its tail base.
(C: 1, 2, 3) The salamander often opened
its mouth, and sometimes bit its own tail
or the forceps.
(D) After some seconds the stressed animal
returned to its initial position and
remained stationary or, more rarely,
repeated the biting sequence after a
while.
Biting as defensive behaviour is known
for a number of plethodontid salamanders,
such as Aneides, Desmognathus,
Gyrinophilus and Plethodon (Brodie et al.
1989), but it has never before been
described for European plethodontid
salamanders. Desmognathus quadramaculatus
use biting to repulse attacks of
snakes, such as Thamnophis sirtalis
(Brodie etal. 1989). Speleomantes italicus
may take some advantage by biting
behaviour against Anguis fragilis and
Natrix snakes (Lanza, 1999a). Also, the
adhesive nature of skin secretions of Cave
Salamanders can be used against snakes,
as quoted for Plethodon and Ensatina by
Arnold (1982).
Only adult females displayed the biting
behaviour. Although these are preliminary
data, the hypothesis of a correlation
between biting and nest defence in
Speleomantes should be taken into
account and verified, as it has been
experimentally demonstrated in some
American plethodontids (Bachmann
1984, Horn etal. 1990).
Body coiling has been also observed in
several plethodontids, such as Aneides,
Batrachoseps, Desmognathus, Ensatina,
Gyrinophilus, Hydromantes, Plethodon,
Pseudotriton, and Bolitoglossa (Garcia-
Paris & Deban 1995). In Hydromantes
platycephalus, which is closely related to
Speleomantes, this generalized escape
strategy, associated with tucking limbs
close to the body, results in a peculiar
rolling antipredator escape behaviour,
similar to that displayed by the anuran
genus Oreophrynella; this may represent
a convergence between two distantly
related, taxa that both occur on rocky
slopes (Garcia-Paris & Deban 1995). A
similar antipredator behaviour was also
observed in Speleomantes italicus.
Indeed, this species displayed limb tucking
when coiling as a consequence of disturbance,
and sometimes dropped down
from the cave walls to the floor; thus, the
occurrence of rolling escape or falling
escape strategies in this genus should be
verified. In S. italicus we also observed
coiling, abundant skin secretions in the
tail base region, immobility (Lanza
1999a), and escape by rapid snake-like or
coiling-uncoiling movements.
During sampling session, conducted by
the authors in Sardinia in spring 1999,
several specimens of all four Sardinian
Speleomantes species were observed. In
Speleomantes imperialis the authors
observed abundant skin secretions at the
tail base region, body coiling, body elevation,
and tail undulation, similar to the
antipredator strategies reported in Lanza
(1999a).
During the handling of a Speleomantes
supramontis adult specimen that was
taken by forceps at the trunk region, its
tail base secretions mixed with urine were
accidentally sprayed on the eyelid of one
of the authors, causing irritation and

130 Biota PASTORELLI, LAGHI & SCARAVELLI
scald-like dermatological symptoms
(Figure 2). Human skin responses to salamander
toxins began only few seconds
after the contact and lasted about ten
days. It is worth noting that absorption of
salamander venom was reduced by
promptly washing the eyelid with fresh
water.
The skin secretions of plethodontid salamanders
are known to be very irritating
to human mucous membranes. Hansen
(1990) reported a severe reaction, includ-
ing temporary blindness, suffered by a
human after handling an adult
Hydromantes platycephalus. Toxicology
of these secretions has been studied only
for S. italicus (Benedicenti & Polledro
1899) and S. strinatii (Ph\sa\\x 1918). This
is the first observed case of damage to
human skin by tail base gland secretions
(although mixed with urine) of S. supramontis,
supporting the hypothesis of its
effective defensive function (Brizzi et al.
1991).
Figure 2. Human skin's reaction to Speleomantes supramontis tail base secretion include
scald-like symptoms (in the rectangle).
Acknowledgements
We thank Prof. Benedetto Lanza for useful suggestions and literature, Franca Monti and
Luciano Cicognani who realized the digital film, and ST.E.R.N.A. company for technical
support.
REFERENCES
ARNOLD, S. J. 1982: A quantitative approach to antipredator performance: Salamander
defence against snake attack. Copeia 1982: 247-253.

PASTORELLI, LAGHI & SCARAVELLI Biota 3/1-2.2002 131
BACHMANN, M. D. 1984: Defensive behaviour of brooding female red-backed salamanders
(Plethodon cinereus). Herpetologica 40: 436-443.
BENEDICENTI, A. & POLLEDRO ,O. 1899: Sulla natura e sulla azione fisiologica del veleno
dello Spelerpes fuscus. Atti della regia Accademia dei Lincei, 296 Rendiconti 8
(1st semester): 413-418.
BRIZZI, R.r CALLONI, C. & DELFINO, G. 1991: Tail base glands in European plethodontid
salamanders, with some comments on their biological and phylogenetic significance.
Amphibia-Reptilia 12: 357-372.
BRODIE, E. D. JR. 1977: Salamander antipredator postures. Copeia 1977: 523-535.
BRODIE, E. D. JR 1983: Antipredator adaptations of salamanders: evolution and convergence
among terrestrial species. In: Margaris N. S., Arianoutsou-Faraggitaki M. &
Reiter R. J. (eds). Plant, animal, and microbial adaptations to terrestrial environment.
Plenum Publishing Corporation; New York.
BRODIE, E. D. JR., DOWDEY, T. G. & ANTHONY, C. 1989: Salamander antipredator strategies
against snake attack: biting by Desmognathus. Herpetologica 45:167-171.
BRODIE, E. D. JR., FORMANOWICZ, D. R. JR. & BRODIE III, E: D. 1991: Predator avoidance
and antipredator mechanisms: distinct pathways to survival. Ethology
Ecology & Evolution 3: 73-77.
GARCIA-PARIS, M. & DEBAN, S. M. 1995: A novel antipredator mechanism in salamanders:
rolling escape in Hydromantes platycephalus. Journal of Herpetology 29: 149-
151.
HANSEN, R. W. 1990: Hydromantes platycephalus (Mount Lyell Salamander). Toxicity.
Herpetological Review 21: 91.
HOM, C. L, WILLITS, N. H. & CLARK, C. W. 1990: Fitness consequences of nest defence in
plethodontid salamanders: predictions of a dynamic optimisation model.
Herpetologica 46: 304-319.
LABANICK, G. M. 1984: Anti-predator effectiveness of autotomized tails of the salamander
Desmognathus ochrophaeus. Herpetologica 40: 110-118.
LANZA, B. 1999a: Speleomantes ambrosii (Lanza, 1955) - Ambrosis Holensalamander. In:
Grossenbacher K. & Thiesmeier B. (eds). Handbuck der Reptilien und Amphibien
Europas, Band 4/1, Schwanzlurche (Urodela) I (Hynobiidae, Proteidae,
Plethodontidae, Salamandridae I: Pleurodeles, Salamandrina, Euproctus,
Chioglossa, Mertensiella). AULA Verlag, Wiesbaden: 91-135.
LANZA, B. 1999b: Speleomantes imperialis (Stefani, 1969) - Duftender Holensalamander.
In: Grossenbacher K. & Thiesmeier B. (eds). Handbuck der Reptilien und
Amphibien Europas, Band 4/1, Schwanzlurche (Urodela) I (Hynobiidae,
Proteidae, Plethodontidae, Salamandridae I: Pleurodeles, Salamandrina,
Euproctus, Chioglossa, Mertensiella). AULA Verlag, Wiesbaden: 155-163.
PASTORELLI, C., LAGHI, P. & SCARAVELLI, D. 2001: Studi preliminari sull'ecologia di
Speleomantes italicus (Dunn, 1923) nell'Appennino Tosco-Romagnolo. Atti del
III Congresso Nazionale SHI (Pavia, 2000). Pianura 13: 347-351.
PHISALIX, M. 1918: Les venins cutanes du Spelerpes fuscus Gray. Bulletin du Museum
national d'Histoire naturelle 24: 92-96.
STEFANI, R. 1969: La distribuzione geografica e I'evoluzione del geotritone sardo
(Hydromantes genei Schleg.) e del geotritone continentale europeo
(Hydromantes italicus Dunn). Archo zoologico italiano 53: 207-243.

POLYNOVA & POLYNOVA Biota 3/1-2.2002 133
Tail autotomy as an index of
human influence on the
AlsophySax pipiens population in
the Bogdino-Baskunchak state
reserve
Galina V. POLYNOVA1 & Olga E. POLYNOVA2
'Russian Peoples' Friendship University, Department of Ecology, Pavlova 8/5, Moscow,
113093, Russia
2Moscow State University, Department of Geography, Vorobyovy Gory, MOSCOW,
119899, Russia
E-mail: olgapol@aport.ru
Abstracts
We started our investigations in the Alsophylax pipiens Pall, population in August 1998
and we concluded it in August 2000. The first part of our research dealt with the undisturbed
part of the population, and the second in a place where human influence is very
high because of frequent excursions around the reserve territory. We collected the main
population data in both territories and our comparison gave interesting results. We
compared several sex-age groups in both parts of the population and counted the percentage
of tail autotomy in each group. In almost all groups in the disturbed part of the
population this percentage is high. Therefore, tail autotomy can be an index of human
pressure on ecosystems.
Keywords: population, sex-age groups, tail autotomy
Received 15 March 2002; accepted 3 August 2002

134 Biota 3/1-2,2002 POLYNOVA & POLYNOVA
INTRODUCTION
Alsophylax pipiens Pall, is the only species
of the Gekkonidae family inhabiting
Russia. The whole of the species' area in
Russia is represented by the territory of
the isolated population inhabiting the
Bolshoe Bogdo mountain in the northeastern
part of the Astrakhan region,
which is situated beside a beach with the
largest salt deposits in the world, on the
shore of Lake Baskunchak. This is where
Alsophylax pipiens was described for the
first time in 1813 by the great Russian scientist
and traveller, Peter Simeon Pallas.
In 1997 the environs of Lake Buskunchak
and the mountain Bolshoe Bogdo
acquired the status of the Bogdino-
Baskunchak state reserve.
Our investigations of the Alsophylax pipiens
population started in 1995 as the
part of the Astrakhan Herpetological
expedition, financed by the MacArthur
and ISAR Foundations (Polynova &
Bozshansky 1995, 1998a, 1998b,
Bozshansky & Polynova 1997,
Bozshansky & Polynova 1998, Polynova
& Polynova 2000). We continued our
work in August 1998 and in August 2000
with the Student Ecological Expedition of
Figure 1. Size by groups (1-5) of males and subadults (1998).
«
6-,
5 -
I ^
B 3
d o
Z z "
1 -
Q
ihi
-
•
.
• - -
' • :
5
*,
i
2
11
j
f
':
'
3
the Russian Peoples' Friendship
University, Ecological Department. The
main purpose of the expedition was to
investigate human influence on the
reserve's ecosystems. The mountain
Bolshoe Bogdo is the centre of the recreational
burden of the reserve's territory.
This article deals with the data gathered
during the 1998 and 2000 field seasons
and presents some interesting results.
MATERIAL AND METHODS
We used the following methods: catching,
measuring, sex determining, marking
with colour and finger cutting off; mapping
of lizards' meetings and movements;
biotopes description and daily activity
registration. The last two data will form
the basis of further publications.
The young lizards that we couldn't distinguish
by sex rate were put into the
subadult group (Figure 1, group no. 1).
All of them spent one winter. We based
our division of adult males and females
into size-age groups on their body length.
It is well known that adult lizards grow
very slowly. That is why the variation line
of their body length is continuous.
Nevertheless, the male data from 1998
?
\
s
>
J
"
4
,
"
-••H
|" ' - -
-
. .
t
i K
t °
.< ^^
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
Body size (mm)

POLYNOVA & POLYNOVA Biota 3/1-2,2002 135
Figure 2. Size groups of females (1998)
34 -^' 35 -" 36 -: !-37, •*•*' 38 39
*«,, ";v Body size (mm)
show markedly distinguished groups
(Figure 1, no. 2-5). The division of female
groups is based on their variation line
(Figure 2) and by analogy with the distinguished
size groups of males. This
method is imperfect to some extent, but
Table 1. Body length of sex and age groups (public data)
Sex and age groups
Juveniles
Juveniles
Juveniles
Subadults
Subadults
Just adults
Just adults
Just adults
Adult males
Adult males
Adult females
Adult females
Body length (mm)
16-18
15-17
18-20
20-26.6 (22.41±0.60)
19-24 (21. 5±0.2)
25-28
31-32
34
33-36 (35.0±0.3)
22.5-33-0(31.310.5)
31-33 (33.110.5)
30-40(36.110.5)
as the species is so small and rare, we
considered it to be impossible to investigate
lizards' age by the classic methods of
comparing annual bone rings. Table 1
shows public data on this species' body
length. The differences between subadult
Authors
Scherbak & Golubev, 1986
Smirnovetal, 1985
Chernov 1947
Brushko 1995
Shammakov 1981
Andrushkol9555 Scherbak
& Golubev 1986
Brushko 1995
Smirnovetal. 1985
Kubykin 1975
Brushko 1995
Brushko 1995
Kubykin 1975

136 Biota 3/i-a, 2002 POLYNOVA & POLYNOVA
Table 2. Sex-age groups and tail autotomy percentage (August 1998).
No. of
group
1
2
3
4
5
6
7
8
Sex and age
Subadults
S3
33
•/ *//v5

POLYNOVA & POLYNOVA Biota 3/1-2,2002 137
Figure 4. Size by groups (4-7) of females (2000).
females (Figure 1,2). The average sizes of
groups are presented in Table 2. Other
authors (Scherbak & Golubev 1986) also
show the presence of several sex-age
groups in this species.
The average male/female sex-ratio is
1.4:1. The same-size male/female sexratio
(groups no. 4 and no. 6) is 2:1. The
older sex groups do not coincide, but on
the whole we see the prevalence of
females over males. The prevalence of
males in the whole population is due to
young adult males (groups no. 2 and no.
3). These males are probably from two
clutches of the same year. We base this
probability on the information about two
clutches known in this species (Scherbak
& Golubev 1986).
We see that in the youngest part of the
population there is a prevalence of males,
then the sex-ratio changes to 1:1. As time
goes by, the number of the oldest females
prevails over the number of the oldest
males. Our data are quite representative,
because we took them from the wide territory
of the population.
The next body data - the length of the tail
- needs special attention. It includes little
information about the lizard's age
because of the high tail autotomy percent,
but it has another value. Tail autotomy
percentage increases from subadult
animals (14%) to mature females (100%)
(Table 2). This increase is quite natural:
those lizards who enjoy a longer life have
more accidents. There are some public
data about tail autotomy levels in the
Alsophylax pipiens populations (16.7% -
Shammakov 1981, 30% - Anan'eva &
Muhtabayar 1997).
The next stage of our investigations in
August 2000 was devoted to the most
disturbed part of the species' population.
It lies in the northeastern territory, at the
top of the mountain. There we caught,
measured, and marked 70 lizards (Table
3, Figure 3, 4).
As we can see, the northeastern part of
the population also has 7 sex-age groups:
1 subadult group, 2 groups of adult males
and 4 groups of adult females. Subadults
were longer than in 1998, and adult
males formed only two distinguished
groups: the younger male group No. 2
unites two groups from 1998 (no. 2, no.
3, Table 2). The next size-age male group
No. 3 was the same as in 1998 (No.4,
Table 2). Finally, we couldn't find the oldest
male group in 2000. At the same
time, females also formed a continuous
variation line, which we divided approximately
into 4 groups (by analogy with
previous data and due to several maximums).
It was more than in 1998. We

138 Biota 3/i-a, 2002 POLYNOVA & POLYNOVA
found younger females No.4, and other
female groups were approximately the
same in both seasons.
It is obvious that there were some
changes in the male/female sex-ratio.
The average sex -ratio was 1.6:1 (males :
females), the same-age youngest
male/female sex - ratio was 4.3:1 (males
no. 2 and females no. 4, Table 3), then it
changed to 1:1 (males no. 3 and females
no. 5 and no. 6) and, finally, there were
no oldest males corresponding to the oldest
females.
We can now see that the data of 1998
and 2000 demonstrate the prevalence of
males in the whole population. This is
because of the prevalence of males in the
youngest group. Then we see the increasing
proportion of females in older groups
and the prevalence of females over males
in the oldest part. We couldn't find such
detailed information in public data. Some
authors (Andrushko 1955, Scherbak &
Golubev 1986) point out 1:1 sex -ratio in
this species.
Sex-age groups and sex-ratio data from
the 1998 and 2000 field seasons have
much in common, but tail autotomy
materials differ. As we have shown, the
tail autotomy percentage of the undisturbed
southeastern population group
naturally increases from the youngest to
the oldest animals (Table 2). At the same
time, the tail autotomy percentage of the
northwestern part is very high in all sexage
rate groups, with exception of the
youngest females: no. 4 (Table 3). It is
well known that tail autotomy in natural
lizard populations can indicate natural
predation pressure (Pianka 1970, Medel
et. al., 1988 and others), lizard density,
aggressive interactions of competitive
individuals, and other factors.
The comparison of these two population
groups shows that: 1) the density of
lizards in the northwestern disturbed area
is lower than in the southeastern undisturbed
area, so aggressive interactions
between competitive individuals must be
lower too; 2) natural predation (our
observations) is the same; 3) the main
difference between these areas that can
influence tail autotomy is the recreational
burden. The high recreational burden of
this region consists of dozens of tourists
and their cars. On weekends the number
of visitors increases by three or four
times. Lizards practically live under
tourists' feet and our experience shows
the ease of lizards' tail loss.
The man-made pressure level can be
undoubtedly measured for almost every
animal and plant species that suffers it.
But we know some indicator species that
are most suitable for this aim. To our
minds, the tail autotomy level in lizard
populations may be one of them. This
index can be of value in factoring the disturbance
caused by the recreational burden.
We think that such an index is more
important, especially in reserves, than
investigations in already disturbed
ecosystems, but it needs further research.

POLYNOVA & POLYNOVA Biota 3/1-2,2002 139
REFERENCES
ANAN'EVA, N.B. & MUNHBAYAR, K. 1997: Pisklivyy gekkonchik. In: Zemnovodnye i presmykauschiesya
Mongolii. Presmykauschiesya. Kollektivnaya monografiya.
Sovmestnaya rossiysko-mongorskaya ekspediciya. RAN, AN Mongolii. KMK Ltd.
M.: 14-27.
ANDRUSHKO, A.M. 1955: Presmykauschiesya Kazahskogo nagor'ya i ih hozyaystvennoe
znachenie. Uchen. zap. Leningr. un-ta. Ser. biol. nauk 181: 19-43.
BOZHANSKIY, AT. & POLYNOVA, G.V. 1998: Proekt regional'nogo spiska reptiliy Krasnoy
knigi Astrahanskoy oblasti. In: Problemy sohraneniya bioraznoobraziya aridnyh
regionov Rossii. Materialy mezhdunarodnoy nauchno-prakticheskoy konferencii.
Volgograd, Rossiya, 11-17 sentyabrya 1998: 57-59.
BRUSHKO, Z.K. 1995: Yaschericy pustyn' Kazahstana. Izd-vo Konzhyk, Almaty: 231 p.
KUBYKIN, R.A. 1975: Ekologo-faunisticheskiy obzor reptiliy ostrovov oz. Alakol' (Vost.
Kazahstan). Izv. AN KazSSR. Ser. biol. 3: 10-16.
POLYNOVA, G.V. & POLYNOVA, O.E. 2000: Problemy sohraneniya gerpetofauny
Astrahanskoy oblasti. In: Aktual'nye problemy ekologii i prirodopoPzovaniya. Izdvo
RUDN, M.: 65-70.
SMIRNOV, S.I., SHKUNOV, V.F., KUDAKINA, E.I. 1985: Gekkony Severnogo Prikaspiya. In:
Voprosy gerpetologii. Nauka, L.: 195-196.
CHERNOV, S.A. 1947: Materialy k gerpetofaune Kazahskogo nagor'ya, severnogo
poberezh'ya Balhasha i gor. Kin-Tau. Izv. AN KazSSR. Ser. zool., vyp. 6.: 120-124.
SHAMMAKOV, S. 1981: Presmykauschiesya ravninnogo Turkmenistana. Ylym, Ashhabad:
311 p.

RACCA Biota 3/1-2.2002 141
The conservation of the Agile Frog
Rana dalmatina in Jersey
(Channel Islands)
Laura RACCA
The Durrell Institute of Conservation and Ecology, University of Kent, Canterbury, Kent,
CT2 7NS, UK
E-mail: Iauracca1@yahoo.it
Abstract
The Jersey population of the Agile Frog has been declining in both range and number
since the early 1900s. At the present time, there is only one site on the island that supports
a natural agile frog population. Since the 1980s, a captive breeding, reintroduction
and habitat management programme involving various organisations has been trying
to arrest this dramatic decline in agile frog numbers in the wild. All parties involved
agreed that without a clearer understanding of the ecology and habitat requirements of
the agile frog in Jersey, it was difficult to develop management techniques for the protection
and improvement of key habitat areas. The objectives of the study are to (1)
estimate the number of adult frogs through a mark - recapture programme; (2) monitor
hatching success and recruitment; (3) study the interactions with other species present
in the pond (toads, newts, invertebrates and grass snakes) and in the surrounding
area (grass snakes, small mammals and birds); (4) determine the frog's overwintering
preferences; (5) monitor the captive breeding programme and carry out an investigation
of potential reintroduction sites; (6) develop optimal conditions for captive rearing
of tadpoles and consider a health monitoring protocol prior to release of captive raised
frogs; (7) compare the habitat requirements of the Agile Frog in Jersey and in northern
France. Preliminary results of the study obtained from the observations collected during
the first fieldwork season are presented here.
Key words: conservation, decline, ecology, Jersey, Rana dalmatina, Species Action Plan.
Received 6 September 2001; accepted 24 October 2001

142 Biota 3 i-->, RACCA
INTRODUCTION
The island of Jersey is situated 21 kilometres
off the north-west coast of France
and 128 kilometres from the English
coast. Jersey is the largest of the Channel
Islands with a surface area of 117 square
kilometres and an official population of
approximately 85,000. The island has its
own government, called the States of
Jersey, which comprises 53 elected members.
Jersey also has its own system of
local administration, fiscal and legal systems,
and courts of law. Jersey is neither
part of the United Kingdom nor a colony.
It is not represented in the United
Kingdom Parliament, whose Acts extend
to Jersey only if the island expressly
agrees that they should do so. The island
has allegiance to the British Crown. Jersey
is not part of the European Union (States
of Jersey 2001).
The Environmental Services Unit (ESU) is
the department within the States of
Jersey Planning and Environment
Committee involved in almost every
aspect of environmental management in
Jersey. On the island, certain ecosystems
are protected by a designation under the
planning law. They are called Sites of
Special Interest (SSI) and are monitored
by the ESU to ensure that the conditions
that make them »special« are not lost
(States of Jersey 2001).
The following amphibians and reptiles are
found in Jersey (Tonge 1986): Agile Frog
Rana dalmatina Common Toad Bufo
bufo, Palmate Newt Triturus helveticus,
Grass Snake Matrix natrix, Slow Worm
Anguis fragilis, Green Lizard Lacerta bilineata
and Wall Lizard Podarcis muralis.
They are all protected under schedule 1
of the Conservation of Wildlife (Jersey)
Law 2000 (Jersey Legal Information
Board 2001).
Figure 1. North Slack, Ouaisne' Common. The only site in the island that supports a
wild population of Rana dalmatina. The terrestrial habitat around the slack is a developing
heathland.

RACCA Biota 3/i-a, 2002 143
DISTRIBUTION AND DECLINE OF THE
AGILE FROG
Apart from Jersey, the agile frog is not
found anywhere else in the British Isles
(Grossenbacher 1997). The Jersey population
of the agile frog has been declining
in both range and numbers since the
early 1900s (Gibson & Freeman 1997). In
the 1970s, the frog could be found in
only seven localities, and by the 1980s
this had dropped to only two sites (Tonge
1986). In 1987, a spill of aldicarb (an
agricultural pesticide) in one of the sites
caused the loss of one of the two remaining
populations of frogs (Gibson &
Freeman 1997). Therefore, at the present,
there is only one site in the island, a
coastal heathland, which supports a population
of Rana dalmatina in the wild
(Figure 1).
Possible causes of agile frog decline in
Jersey are the following (Agile Frog
Group 2001):
1) Habitat loss / fragmentation. Jersey is
a densely populated island and habitat
fragmentation continues through development
for housing and the associated
services required. In addition, small-scale
turnover of semi-natural habitats, including
conversion of heathland and marginal
farmland to agricultural land, mean
that important habitat areas are continually
being lost.
2) Water quality. Jersey, not being part of
the European Union, does not have to
follow its guidelines for the maintenance
of water quality. In Jersey, pollution of
groundwater is caused by two main
sources, agriculture and domestic wastes.
Agricultural nutrients and sewage from
defective or poorly located soakaways are
implicated in the pollution of the groundwater
by the abnormally high nitrate concentrations
observed. In addition, pesticide
residue has often been detected in
run-off from arable farmland.
3) Pollution events. As previously mentioned,
a spillage of aldicarb in Noirmont
pond in 1987 had a lethal effect on many
frog and toad eggs and adults, and basically
halved the agile frog's range
overnight. Another pollution event was a
recent contamination of an area close to
the airport by run-off from airport firefighting
practice. In addition, there are
countless small scale pollution events
occurring on a frequent basis throughout
Jersey which threaten both aquatic and
terrestrial habitats.
4) Water shortage. Agriculture is the principal
land use in Jersey and covers 58%
of the total land area of the island. Land
is intensively farmed due to the excellent
soil quality, and the scarcity of land frequently
leads to double cropping. The
industry exacts a high demand from the
island's water supply and often improves
land drainage to ensure rapid run-off.
Additionally, many houses are still not
connected to a main water supply, and
are instead reliant on the ground water.
These agricultural and domestic factors
may have caused the lowering of the
water table that is now being experienced
in Jersey, which has resulted in the loss of
many ephemeral ponds.
5) Predation pressures. Apart from their
natural predators (palmate newts, some
aquatic macro-invertebrates, small mammals,
grass snakes, etc), frogs and toads
are now also being predated by pets,
which are increasing in numbers with the
growth in human population. Domestic
cats and polecat-ferrets are both plentiful
as feral pests. More recently, large numbers
of feral ducks have invaded many
available water bodies, including the agile
frog's only natural breeding site.
6) Small population effects. The Jersey
population of the agile frog has, and still
is, going through a severe Cbottleneckx
effect. At the present time, the factors
listed above are likely to be the most
influential in determining the immediate
survival of the agile frog. In the longer
term, however, any surviving population

144 Biota 3/1-2,2002 RACCA
could become exposed to genetic risk
factors, such as inbreeding depression
(Caughley 1994).
CAPTIVE BREEDING PROGRAMME OF
THE AGILE FROG IN JERSEY
A collaborative programme incorporating
captive-breeding, reintroduction and
habitat management started in the late
1980s in order to try to arrest the potentially
terminal decline in agile frog numbers
in Jersey. The organisations involved
are the Environmental Services Unit
(ESU), the Herpetology Department of
the Durrell Wildlife Conservation Trust,
and the Zoology Section of the Societe
Jersiaise. The main objectives of the programme
are (Agile Frog Group 2001) (1)
to maintain self-sustaining captive safety
net populations in the event of the extirpation
of the species from the remaining
natural breeding site and (2) to generate
surplus tadpoles and froglets for reintroduction
into existing and potential new
sites.
Figure 2 shows the total number of egg
clumps found in the wild and number of
clumps taken for the captive breeding
programme from 1990 to 2001. It is not
clear if the lack of clumps in the wild dur-
ing 1994, 1995, and 1996 was real or if
spawn were missed during the surveys
because of the unavailability of surveyors.
REINTRODUCTION AND INTRODUC-
TION SITES
Grosnez pond
This site (Figure 3) has never supported
populations of agile frogs. It was chosen
as an introduction site because it had
good quality water and was a breeding
site for toads and newts (Agile Frog
Group 2001). The introduction started in
1994 and the first clump of frog eggs was
found two years later. Breeding has reoccurred
in 1998, 1999, and 2000.
Tadpoles were released in the pond for
the last time in 2000 (Racca 2000).
Noirmont pond
This site (Figure 3) was an historic breeding
pond for the agile frog until 1987,
when a spill of an agricultural pesticide
polluted its water. It was chosen as a reintroduction
site after the results of water
analyses indicated that the pollution
problem had been solved and after
observing that introduced toad tadpoles
developed successfully to metamorpho-
Figure 2. Total number of egg clumps found in the wild and numer of clumps taken for
the captive rearing programme, from 1990 to 2001.
a, 12-]
1 10-
1 8
? 4-
-i 2 -
i 0
0 No of chimps found at Ouaisne'
D No of clumps taken for captive rearing programme

RACCA Biota 3/1-2, 2002 145
Figure 3. Grosnez pond (A) and enclosure in Noirmont pond (B).
sis. Frog tadpoles, taken from captive
breeding sites, were introduced last year
for the first time (Racca 2000) and then
again this year.
Population status and reproductive success
of the Agile Frog in Jersey
In order to produce an effective conservation
programme for the agile frog in
Jersey, it is necessary to investigate the
threats to this species and its ecology and
behaviour. For this reason, on February
2001, I started a three year study on the
ecology and conservation of Rana dalmatina
in Jersey. The objectives of the
study are (1) to estimate the number of
adult frogs through a mark - recapture
programme; (2) monitor hatching success
and recruitment; (3) study interactions
with other species present in the pond
(toads, newts, invertebrates, and grass
snakes) and in the surrounding area
(grass snakes, small mammals, and birds);
(4) determine the frog's overwintering
preferences; (5) monitor the captive
breeding programme and carry out an
investigation of potential reintroduction
sites; (6) develop optimal conditions for
captive rearing of tadpoles and consider a
health monitoring protocol prior to
release of captive raised frogs; (7) compare
the habitat requirements of the agile
frog in Jersey and in northern France.
These first, preliminary results were
obtained from the data collected during
this year's fieldwork season, from
February to July 2001. Surveys were carried
out every night between 22.00-
24.00hrs during the breeding season (mid
January to mid March). In total, 10 male
frogs were caught and PIT-tagged
(Camper & Dixon 1988). No females
were caught. However, from the number
of egg clumps found in the breeding
Figure 4. Estimated mortality of eggs during embryonic phase, in the wild (left) and in
captivity (right). North Slack, South Slack and Sump are the water bodies where egg
clumps were found at Ouaisne' (left) . Site A and site B are captive breeding sites; the
first is a private garden pond and the latter is an enclosure at the Jersey Zoo (right).
1 111 ULLi
North Slack South Slack
• % hatched eggs D% d*ad oggs at hatching

146 Biota 3/1-2,2002 RACCA
Table 1. Summary of development of wild spawn and tadpole.
Number of clumps
Estimated number of eggs
First clump laid
Minimum embryonic developmental time (days)
First transformation
Minimum tadpole developmental time (days)
Total number of froglets caught
ponds (n = 7), it can be assumed that at
least this number of breeding females this
year visited the site. This year, only two
clumps were laid in captivity. The highest
egg mortality during the embryonic
phase was 40% in the wild and 50% in
captivity (Figure 4). The main predators
of the tadpole phase in the wild were
aquatic macro-invertebrates such as
Water Boatmen Notonecta glauca, Water
Spiders Agryroneta aquatica, Great
Diving Beetles Dytiscus marginalis, Silver
Water Beetles Hydriphilus piceus, and
dragonfly nymphs Odonata. These data
were collected by doing monthly standardised
netting sessions in the water
bodies where the tadpoles were found.
Palmate newts were also caught in the
ponds, but in very small numbers. Only
three sightings of grass snakes were
noted during the whole period of fieldwork.
Drift fences and pitfall traps were
used in order to catch newly metamorphosed
frogs. Once caught, they were
marked with waterproof ink scratched
with a needle on the skin of one of their
back legs. Sixty-eight froglets were
caught and most of them were recaptured
at least once.
Table 1 summarises the data collected
about wild spawn clump development
during this year's fieldwork.
7
2950
23/2/01
27
15/6/01
66
68
Conclusions
A Species Action Plan for the agile frog
has just been published in Jersey. It was
written and published by the Jersey Agile
Frog Group and its main objectives can be
summarised as follows:
1) To ensure that all existing natural,
introduction and reintroduction sites are
actively protected.
2) To increase the number of sites
through introductions / reintroductions.
3) To maintain a self-sustaining captive
population of frogs at a minimum of
three locations.
4) To further investigate the threats to
and ecology and behaviour of the agile
frog in Jersey.
5) To increase the conservation profile
and level of awareness of the agile frog's
plight in Jersey.
The situation of the agile frog in Jersey is
certainly worrying. On the other hand, an
effort to try and change this situation is
being made, with the publication of the
Species Action Plan and the involvement
of the ESU, the Durrell Institute of
Conservation and Ecology (DICE), the
Durrell Wildlife Conservation Trust, the
Zoology Section of the Societe Jersiaise
and, in the near future, with financial help
from local businesses.

RACCA Biota 3/i-a, 2002 147
Acknowledgements
I would like to thank my supervisor Dr. Richard Griffiths for his constant help. I am also
grateful to all the members of the Agile Frog Group for their enthusiastic involvement
in this project. A grant from the Jersey Ecology Trust Fund and contributions from the
firm Mourant de Feu & Jeune and the Environmental Services Unit (States of Jersey
Planning & Environment Committee) funded this year's PhD study.
REFERENCES
AGILE FROG GROUP 2000: Species Action Plan: the agile frog (Rana dalmatina) in Jersey.
Environmental Services Unit, Jersey, 26 pp.
CAMPER, J.D & DIXON, J.R 1988: Evaluation of a microchip marking system for amphibians
and reptiles. Texas Parks and Wildlife Department, Research Publications 7100-
159. Austin, Texas.
CAUGHLEY, G. 1994: Directions in conservation biology. Journal of Animal Ecology, 63 (2):
215-244.
GROSSENBACHER, K. Rana dalmatina. In: Case, J.R, Cabela, A., Crnobrnja-lsailovic, J.,
Dolmen, D., Grossenbacher, K., Haffner, P., Lescure, J., Martens, H., Martinez
Rica, J.P., Maurin, H., Oliveira, M.E., Sofianidou, T.S., Veith, M. & Zuiderwijk, A.
(eds.). 1997: Atlas of Amphibians and Reptiles in Europe. SHE & SPN-MNHN,
Paris.
GIBSON, R. & FREEMAN, M. 1997: Conservation at home; recovery programme for the
agile frog Rana dalmatina in Jersey. Dodo Journal of Wildlife Preservation Trust
33:91-104.
Jersey Legal Information Board 2001: http://www.jerseylegalinfo.je, accessed July 2001.
RACCA, L. 2000: Agile frog in Jersey: a field study (March-June 2000). Environmental
Services Unit, States of Jersey Planning and Environment Committee (unpublished).
States of Jersey 2001: http://www.gov.je, accessed July 2001.
TONGE, S. 1986: The herpetofauna in Jersey. British Herpetological Society Bulletin 17: 18-
19.

SALVIDIO, ALARIO, PASTORINO & FERRETTI Biota 3/1-2, 2002 149
Seasonal activity and abundance
of Speleomantes ambrosii in cave
habitats
Sebastiano SALVIDIO1f Gabriele ALARIO1,
Mauro Valeric PASTORINO2 &
Mirko FERRETTI3
1DIP.TE.RIS-Universita di Geneva, Corso Europa 26, 1-16132 Geneva, Italia
E-mail: salvidio@dipteris.unige.it
2Gruppo Speleologico Ligure "A. Issel" - Busalla (GE), Italia
3Gruppo Speleologico Lunense, Sarzana (SP), Italia
Abstract
In this study, seasonal activity, population structure and abundance of three S. ambrosii
populations were studied in three natural caves in the Province of La Spezia, NW Italy.
The annual pattern of activity showed two peaks, one in spring and the other in
autumn. The demographic structure was analysed on the basis of body-size frequency
distributions, and was composed of three well separated body-size groups. The smallest
component corresponded to immatures in their first year of life, the intermediate
one to immatures aged two, and the largest comprised subadults and adults aged three
years or more. Temporary removal experiments were used to estimate population abundance.
Results showed that S. ambrosii populations had relatively high probability of
capture, ranging from 0.51 to 0.88. These results are in good accordance with previous
data obtained on S. strinatii living inside an artificial cavity. Thus, cave salamanders that
concentrate inside underground habitats are highly exposed to disturbance and collection.
Key words: Cave habitat, population structure, Speleomantes ambrosii, removal
method.
Received 3 September 2001; accepted 20 February 2002

150 Biota 3/1-2,2002 SALVIDIO, ALARIO, PASTORINO & FERRETTI
INTRODUCTION
The cave salamander Speleomantes
ambrosii (Lanza 1955) is endemic to a
small karstic area in Eastern Liguria and
North-western Tuscany. It displays a wide
altitudinal distribution, from 30 m a.s.l. to
about 1700 m in the Apuan Alps, and it is
found in extremely diverse habitats from
coniferous forests and maquis to rocky
areas, and it usually occurs in the leaf litter,
under stones along streams, and in
underground habitats (Lanza et al. 1995,
Lanza, 1999). The species is listed in
Annex II of the "Habitat Directive"
92/43/ECC and is also protected by
national (DPR 357/97) and regional legislation
(L.R. Liguria 4/92 and L.R.
Toscana 56/00). However, little is known
of the biology and ecology of this
endemic species. Forti et al. (1997) and
Cimmaruta et al. (1999) studied the distribution
of S. ambrosiiand S. strinatii'm a
parapatric area where these species live in
close contact. In this area, S. ambrosii
seems confined to the more arid zones, in
contrast to the more competitive S. strinatii
(Cimmaruta et al. 1999).
The aim of this study was to collect basic
data on the seasonal activity, abundance,
and population structure of S. ambrosii
living in underground habitats.
MATERIAL AND METHODS
Activity and abundance of S. ambrosii
were studied in three natural karstic caves
in the Province of La Spezia, Eastern
Liguria: Grotta del Papero (near Ricco del
Golfo) at 205 m a.sl, Grotta Lunga di S.
Antonio (near Pignone) at 230 m a.s.l.,
and Grotta di Cassana (near Cassana) at
190 m a.s.l. Salamanders were surveyed
monthly, from September 2000 to August
2001 in Grotta del Papero and Grotta
Lunga di S. Antonio, while in Grotta di
Cassana surveys were conducted from
April to August 2001.
Animals active on the cave walls were
spotted with the aid of a headlamp and
counted without further disturbance.
Population abundance was estimated by
temporary removal methods (Bruce
1995, Salvidio 1998, 2001). In temporary
removal experiments, all animals captured
are physically removed from the
population, kept in a holding area, and
relocated when the study is completed.
When basic assumptions are met, (e.g.,
population closure, equal sampling effort,
equal catchability, and effective reduction
of the population after each search), the
statistics calculated from removal data are
reliable. Salamanders were measured to
the nearest millimetre, from the tip of the
snout to the posterior part of the cloaca
(SVL). These SVL measurements generated
polymodal frequency distribution histograms
that were analysed with the
FAO-ICLARM Stock Assessment Tools
(FiSAT) computer programme (Gayanilo
et al. 1996) which enables the decomposition
of mixed length-frequency distributions
into their Gaussian components by
means of Bhattacharya's (1967) log-differences
method. Abundance was estimated
by means of the programme CAP-
TURE routine Mbh (White et al. 1982).
This software gives the population estimate
Ni, its standard error (S.E.), and, if at
least three removal samplings are used, it
allows a test for homogeneity of capture
probabilities with a x2 goodness of fit test.
During this study no direct mortality of
the removed salamanders occurred, and
at the end of the third removal sampling
all salamanders were released at their
capture sites.
RESULTS
Seasonal activity
The activity patterns in the three caves
were similar (Figure 1), and annual trends
were highly correlated in Grotta del
Papero and Grotta di S. Antonio
(Spearman rank correlation coefficient: rs
= 0.90, n =12; P < 0.01), suggesting that
similar extrinsic (e.g., climate, food abun-

SALVIDIO, ALARIO, PASTORINO & FERRETTI Biota 3/1-2,2002 151
Figure 1. Seasonal pattern of underground activity in three cave populations of
Speleomantes ambrosii.
Number of individuals
Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug
dance) and/or intrinsic (e.g., physiological
cycle) factors were acting in all sites
simultaneously.
Abundance of cave populations
The statistics resulting from the temporal
removal experiments are given in Table 1.
Capture probabilities were always higher
than 0.20, a value considered adequate in
estimating population abundance in
removal experiments (White et al. 1982).
These results indicate that the assumption
of closure were met. However, the numbers
of removed salamanders from both
Grotta del Papero and Grotta Lunga di S.
Antonio were too low to represent well
structured populations. On the contrary,
70 specimens were captured inside
Papero
S. Antonio
Cassana
Grotta di Cassana, suggesting that this
latter cave was a more favourable habitat.
Population structure
All the salamanders captured in July 2001
removal experiments (n = 106) were
pooled to obtain an adequate sample to
be analysed. Body-size measurements
generated a polymodal frequency distribution
histogram that was successfully
decomposed into three separated gaussian
components (Figure 2). The two
smallest body-size groups (mean SVL =
25.99, S.D. 2.67, and mean SVL = 41.11,
S.D. 4.13) were composed of immature
salamanders, while the largest one (mean
SVL = 60.28, S.D. 5.44) was a mixture of
Table 1. Speleomantes ambrosii removal data and population size estimates. Ni =
Number estimated; S.E. = standard error; (Chi-square test for homogeneity of capture
probabilities (P).
Year 2000
Grotta di S. Antonio
Grotta del Papero
Year 2001
Grotta di S. Antonio
Grotta del Papero
Grotta di Cassana
1st capture
June 25
19
24
1st capture
July?
10
13
42
2M capture
June 27
3
3
2 capture
July 9
2
8
15
3"" capture
June 29
0 1
31 capture
July 11
2
1
13
Ni
22
28
14
22
79
S.E.
0.21
0.34
0.82
1.25
6.27
X2 P Capture
probability
0.52 >0.05 0.88
0.58 >0.05 0.85
2.10
2.77
2.35
>0.05
>0.05
>0.05
0.70
0.67
0.51

152 3/1-2, 2OO2 SALVIDIO, ALARIO, PASTORINO & FERRETTI
Figure 2. Speleomantes ambrosii population structure. The curves separating each
Gaussian component were calculated by the FiSAT software (see text).
20 24 28 32 36 40
subadults (i.e.r large immatures) and
adults (i.e., gravid females and males with
mental glands). If females reach sexual
maturity at about 58 mm in SVL, as in S.
strinatii (Salvidio 1993), then the sex ratio
of the adult population was not different
from 1 (17 males/9 females; Chi-square =
2.46, P > 0.05).
DISCUSSION
In the study caves, seasonal activity was
clearly bimodal, peaking in summer and
autumn, and during winter few active
salamanders were observed on the cave
walls. Overall, this pattern of activity was
similar to that of a S. strinatii population
inhabiting a rock-face habitat (Salvidio
1993). This activity pattern was in agreement
between sites, suggesting that the
same extrinsic or intrinsic factors were
acting in all caves, which were in fact at
similar altitudes and therefore exposed to
similar climatic conditions.
The demographic structure was composed
of three Gaussian components,
N-106
48 52 53 60 64 68
and two cohorts of immatures in their
first and second year of life were
observed. The growth increment
between these two components was
about 15 mm, a higher value in relation
to that reported for a rock-face population
of S. strinatii, in which juvenile
growth was about 10-13 mm in the first
year (Salvidio 1998). Within the largest
body size group, it was not possible to
separate other age classes, as growth rate
decreases in larger (and older) individuals
having reached sexual maturity.
In the caves Grotta del Papero and Grotta
Lunga di S. Antonio, few salamanders
were removed from the cave walls, suggesting
that these sites hosted only a limited
proportion of salamanders living in
the surrounding karst system. On the
contrary, in the cave Grotta di Cassana,
the abundance of salamanders was adequate
to obtain effective estimates of
population abundance. Overall, 70 salamanders
were removed and capture
probabilities were relatively high (i.e.,

SALVIDIO, ALARIO, PASTORINO & FERRETTI BJOta 3/1-2, 2002 153
0.51) suggesting that about half of the
population was exposed to capture during
each capture session. This value was
similar to that obtained for S. strinatii living
inside an artificial tunnel (mean value
0.639; Salvidio 2001). Thus, Grotta di
Cassana was a favourable habitat for
salamanders that were highly exposed to
disturbance and collection, at least during
their underground peaks of activity.
These results have implications for conservation,
as those caves hosting large
numbers of salamanders should be closed
or monitored to avoid disturbance and
overcollection of the endemic S. ambrosii.
Acknowledgements
Capture permits were obtained from the Italian A/linistero deH'Ambiente - Servizio
Conservazione della Natura (prat. n. SCN/2D/98/14345 and SCN/2D/2001/378). We
thank Edoardo Razzetti for his field help during the July 2000 sampling.
REFERENCES
BRUCE, R. C. 1995: The use of temporary removal sampling in a study of population dynamics
of the salamander Desmognathus monticola. Australian Journal of Ecology
20: 403-412.
CIMMARUTA, R., FORTI, G., NASCETTI, G. & BULLINI, L. 1999: Spatial distribution and
competition in two parapatric sibling species of European plethodontid salamanders.
Ethology Ecology and Evolution 11: 383-398.
FORTI, G., CIMMARUTA, R., NASCETT, G. & BULLINI, L. 1997: Parapatria e competizione
in pletodontidi del genere Hydromantes. S.ltE Atti 18: 121-124.
GAYANILO, Jr., F.C., SPARRE, P. & PAULY, D. 1996: The FAO-ICLARM Stock Assessment
Tools (FiSAT) User's Guide. FAO Computerized Information Series (Fisheries), 8.
Food and Agriculture Organisation of the United Nations. Rome.
LANZA, B. 1999: Plethodontidae - Lungenlose Salamander. In: Grossenbacher, K. &
Thiesmeier, B. (eds.). Handbuch der Reptilien und Amphibien Europas. Aula
Verlag, Wiesbaden: 77-204.
LANZA, B., CAPUTO, V., NASCETTI, G. & BULLINI, L. 1995: Morphological and genetic
studies of the European plethodontid salamanders: taxonomic inferences (genus
Hydromantes). Torino, Mus. reg. Sci. nai., Monografie 16, Torino.
SALVIDIO, S. 1993: Life history of the European plethodontid salamander Speleomantes
ambrosii. Herpetological Journal 3: 55-59.
SALVIDIO, S. 1998: Estimating abundance and biomass of a Speleomantes strinatii
(Caudata: Plethodontid) population by temporary removal sampling. Amphibia-
Reptilia19: 113-124.
SALVIDIO, S. 2001: Estimating terrestrial salamander abundance in different habitats: efficiency
of temporary removal methods. Herpetological Review 32: 21-24.
SALVIDIO, S. LATTES, A., TAAVANO, M., MELODIA, F. & PASTORINO, M.V. 1994: Ecology
of a Speleomantes ambrosii population inhabiting an artificial tunnel. Amphibia-
Reptilia 15: 35-45.
WHITE, G.C., ANDERSON, D.R., BURNHAM, K.P. & OTIS, D.L. 1982: Capture-recapture
and Removal Methods for Sampling Closed Populations. Los Alamos National
Laboratory 8787 NERP, Los Alamos, New Mexico. 235 pp.

SCALI & GENTILLI Biota 3/1-2.2002 155
A comparison of main heathlands
in northern Italy and their
importance for amphibian
populations
Stefano SCALP & Augusto GENTILLI2
'Museo Civico di Storia Naturale di Milano, C.so Venezia 55 - 1-20121 Milano
E-mail: scaliste@iol.it
2Dipartimento di Biologia Animale, Universita di Pavia, P.zza Botta 9-10, 1-27100 Pavia
E-mail: augusto.gentilli@unimib.it
Abstract
The importance of heathlands as habitat for reptiles has been stressed by many authors.
Few studies about their importance for amphibians, on the other hand, are available
and the significance of these habitats is probably underestimated. Heathlands are considered
priority natural habitats by the European Union in the Habitats Directive and so
their ecology is particularly important to understand.
The authors conducted some herpetological research in the main heathlands in the Po
Plain (Northern Italy) and combined new data with bibliographic information. Six
heathland areas were investigated in Piedmont and Lombardy, and three to seven
amphibian species were found; the complete list of taxa recorded in all the heathlands
11 is the following: Triturus carnifex, T. vulgaris meridionalis, Pelobates fuscus insubricus,
* f Bufo bufo, B. viridis, Hyla intermedia, Rana dalmatina, R. latastei, R. temporaria, R. synklepton
esculenta.
Some information about water presence and permanence, vegetation structure and
coverage, soil type and distance from the Alps is also presented. Data were analysed
using multivariate techniques. Two areas were very different from the others: the first
one is a deteriorating heathland, devoid of woods and permanent water, with a low
amphibian species diversity; the second one is the southernmost area and it is characterised
by the presence of many scattered trees. The other four zones were quite similar,
even if some differences could be observed in amphibian species composition and
habitat structure.
Key words: Amphibians, conservation, heathland, Northern Italy
Received 2 October; accepted 8 February 2002

156 Biota 3 1-5 SCALI & GENTILLI
INTRODUCTION
Heathlands are considered priority natural
habitats by the European Union in the
Habitat Directive 92/43/CEE, Enclosure
1, CORINE habitat code 31.2 because
they have strongly declined throughout
Europe (Beebee 1996). Therefore, knowledge
of their fauna is particularly important
for conservation projects.
The importance of heathlands as habitat
for reptiles has been stressed by many
authors (Spellerberg 1989, Stumpel
1992, Scali 1995, Zuffi 1987a, b). Few
studies about their importance for
amphibians, on the other hand, are available
(Scali 1993, Beebee 1996, Denton &
Beebee 1996) and the significance of
these habitats is probably underestimated.
Many amphibian species that live in
heathlands are considered very important
taxa in the Habitats Directive: Triturus
earn if ex, Pelo bates fuscus insubricus,
Bufo calamita, Bufo viridis, Hyla arborea,
Hyla intermedia, Rana arvalis, Rana dalmatina,
Rana latastei and Rana lessonae.
MATERIALS AND METHODS
The authors conducted some herpetological
research in the main heathlands of
the western Po Plain (Northern Italy) during
the last ten years, and they combined
new data with bibliographic information
(Pozzi 1980). Six heathland areas were
investigated in Piedmont and Lombardy:
1. Piano Rosa, 2. Rovasenda, 3. Lonate
Pozzolo, 4. Mombello, 5. Ca' del Re, and
6. Salvadorino (Figure 1).
We prepared a matrix using the following
variables: area, presence of the different
taxa, total number of taxa, altitude, distance
from the Alps, presence and number
of permanent and temporary ponds,
presence of some plant species (Pinus
sylvestris, Betu/a pendu/a, Populus sp.,
Quercus sp., Pteridium aquilinum), presence
of woods in or around the area, and
type of soil (clay or gravel). A summary of
all the variables and their abbreviations
(in parentheses) is as follows: i) heathland
ID (Area), ii) distance from the Alps
(Alps), iii) number of permanent ponds
Figure 1. Location of the study areas in Northern Italy. Heathlands are assigned by same
numbers as in tables 1 and 2.
__ Chattllon
VALLFD.MlSTft
TraverseHi. "*ji
'Atogna Campo
y
- Breia \ Angara •.
'• '"' Gozzano'1
D Vv V^Sesto Sesto Catencte
A8argcise-;ia , . '-*-.--
.c\*ne
v-.^
DCantV
n SonmaLambaVdO: ' Tr8dale Mar^nDC«ne
"Cogsfale . 1 V: ...•., ° >JGallarete
; _0urttengo
" i -;/ * ~ *i' Pralungo.
Peninefiflo
Gattinaran
•;
Ghemme
n .
I1 '-ffusK Arsizio, .'-
N :;,, i '
°0lerao .nhonado'Lognalio .
^_
3_
09ena90
^Bislla _ --UJsona* -suano caston!)Primb ' Nerviam,
:ambgijano •* ... :•. . '1-1^1 \n Jnveruno ° °
% "a'vrea
J^eroJo Canevese
Socca Cana»t5« ; ^CaluSD
.
FelBto"
Jofrazjo Vterron* ^h4maleiata " .Undwna
.CV.L.biano
" ° a
^S^t^N ^ r , r-Cassolnovo
TrOnzano VerceSese Bbrgo.Vercelli '•; °
D - - ..c ..-'•-".«!. . "7 vigevano

SCALI & GENTILLI Biota 3/i-a, 2002 157
Table 1. Habitat characteristics of the six analysed heathlands. Presence of plant species
is coded as follows: 0=absent, 1=present; soil is coded as 1=clay, 2=gravel.
Area
1
2
3
4
5
6
Alps
(Km)
5
15
30
25
25
65
Permsites
(N)
1
1
0
2
1
1
Tempsites
(N)
2
1
6
4
3
0
(Permsites), iv) Number of temporary
ponds (Tempsites), v) presence of permanent
water (Permwater), vi) presence of
temporary water (Tempwater), vii) presence
of Pinus sylvestris (Pinus), viii) presence
of Betula pendula (Betula), ix) presence
of Populus sp. (Populus), x) presence
of Quercus sp. (Quercus), xi) presence
of Pteridium aquilinum (Pteridium),
xii) percentage of woods in the area
(Intwoods), xiii) percentage of woods
around the area (Extwoods), xiv) type of
soil (Soil). The six areas were coded as
wet or dry heathlands: the former ones
have clayey and waterproof soils, with
hygrophilous vegetation, while the latter
ones have gravelly and permeable soils
and plants typical of dry areas. Two types
of degradation can be found in these
heathlands: A) a lack of management
that leads to a succession through to
woodland mainly composed of exotic
species, such as Robinia pseudoacada
and Prunus serotina, B) a heavy human
disturbance deriving from many kinds of
activities (i.e. road traffic, pollution, water
drainage, buildings, etc.).
The six studied areas (Figure 1) are
described briefly below and in Table 1:
1. Piano Rosa (300 m a.s.l.): it is a medium-sized,
well preserved, wet heathland
near the Alps, with sparse woods and
large open areas. It is surrounded by
woods and agricultural lands. A small
number of permanent and temporary
ponds are present.
2. Rovasenda (220 m a.s.l.): it is a large,
well preserved, wet heathland. Human
Permwater Tempwater Pinus
(N) (N)
1
1
0
1
0
0
0
1
1 ( ) 0
Betula
1
1
0
1
1
1
Populus
1
0
0
1
1
1
Quercus
disturbance is low and many different
microhabitats are present. A stream and
some permanent and temporary ponds
are present.
3. Lonate Pozzolo (220 m a.s.l.): it is a
large, deteriorating, wet area with small,
well preserved portions. Human disturbance
is quite heavy, due to military
activities and road construction; a progressive
growth of wood is also present.
A thin woodland belt is present around
the area. Only a small number of temporary
ponds are present.
4. Mombello (210 m a.s.l.): it is a medium-sized,
deteriorating, wet heathland
once used as a clay quarry. Human disturbance
is quite low and caused by quarry
activities on the borders and wood
growth. The area is characterised by the
presence of many little rises covered by
shrubs and woods. Two large, permanent
and many temporary ponds are present.
5. Ca' del Re (210 m a.s.l.): it is a medium-sized,
well preserved, wet heathland
located near the previous area. Human
disturbance is low. Some sparse woods
are present in the area and a deteriorating
woodland belt surrounds the heathland.
A large permanent pond and some temporary
ponds are present.
6. Salvadorino (70 m a.s.l.): it is a very
small, well preserved, dry area with some
isolated trees and shrubs. It is surrounded
by typical plain woods and channels.
Only one small permanent pond is present.
Data were analysed using two multivariate
methods (cluster analysis with the
0
1
0
0
0
I

158 Biota 3/1-2,2002 SCALI & GENTILLI
Table 2. Presence of the ten Amphibian species in the six areas.
Area
1
2
3
4
5
6
Tricar
1
1
0 1
1
0
Trivul
1
1
0
0
1
1
Pelftis
0 1
0
0
0
0
Bufbuf
J
0
0
0
0 1
Bufvir Hylint
0
0 1
1
0
0
UPGMA method and correspondence
analysis) with the MVSP 3.1 for Windows
package. These two techniques were
chosen to obtain some groups of related
data; cluster analysis was used to group
the heathlands in relation to habitat characteristics
and presence of different
species. Correspondence analysis allowed
us to group different variables and
species to understand the relationships
between them.
Figure 2. Dendrogram of the grouped
heathlands in relation to habitat characteristics
and species composition
(UPGMA method).
0,52 0,6 0,68 0,76 0,84 0,92 1
-2
-5
-4
-1
RESULTS
Three to seven amphibian species were
found in the six heathlands (Table 2); the
complete list of taxa recorded in all the
heathlands is as follows: Triturus carnifex,
T. vulgaris meridionalis, Pelobates fuscus
insubricus, Bufo bufo, B. viridis, Hyla
intermedia, Rana dalmatina, R. latastei, R.
Randal
1
0
0 1
1
1
Ranlat
0
0
0
0
1
1
Ranesc
1
1
1
1
1
1
Rantem
J
0
0
0
0
0
No. of species
7
5
3
5
6
6
temporaria, and R. synklepton esculenta.
A cluster analysis was conducted on the
basis of the data reported in Tables 1 and
2 and the results are showed in Figure 2.
Two areas (3 and 6) are well separated by
a group of similar areas (1,2,4 and 5).
The presence of all species was related to
environmental variables using the correspondence
analysis to highlight habitat
preferences of amphibians in these
heathlands. The results of this analysis are
reported in Figure 3. Three main clusters
were highlighted: i) Bufo viridis was
grouped with the presence and the number
of temporary sites, ii) Rana latastei
and Bufo bufo were grouped with natural
woodlands internal to studied areas, iii)
Triturus vulgaris, Hyla intermedia and
Rana synklepton esculenta were associated
with permanent sites and surrounding
woodlands. The other species appear to
be less associated with the considered
variables.
DISCUSSION
The six heathland areas of the western Po
Plain comprise good habitat for amphibians
and the presence of ten species was
recorded altogether. The minimum
recorded number of species was three
and the maximum was seven. It is particularly
interesting that some endangered
taxa - such as Pelobates fuscus insubricus
and Rana latastei ~ are present. These
species usually live in wood land areas,
but can also colonize some well-preserved
heathlands.
Our analyses showed a certain similarity
in environmental characteristics and

SCALI & GENTILLI )ta 3/1-2, 2OO2 159
amphibian species composition between
the six considered areas.
Only two areas were very different from
the others: the first one, (3), is a deteriorating
heathland, devoid of woodland
and permanent water, with a low
amphibian species diversity. The scarcity
of amphibians in this zone confirms their
sensitivity to habitat degradation. The
second area, (6), is the southernmost area
and it is characterised by the presence of
many scattered trees; the different habitat
structure probably creates some
microclimatic conditions that increase
amphibian diversity. The other four zones
are quite similar, even if some differences
could be observed in amphibian species
composition and habitat structure.
In particular, the presence of the Green
Toad is correlated to temporary ponds, in
accordance with the information available
for this species (Nollert & Nollert
1992, Scali 1995). The Common Toad
and the Italian Agile Frog are present only
in heathlands with internal woodlands,
characterized by typical lowland trees (i.e.
Quercus robur) (Borkin & Veith 1997,
Grossenbacher 1997). The Smooth Newt,
the Italian Treefrog and the Edible Frog
need external woodlands, with large and
sunny permanent ponds (Lanza 1983).
Rana temporaria is only influenced by the
proximity to the Alps, because it usually
lives in mountainous areas in Italy (Lanza
1983). The Italian Spadefoot Toad was
found only once by Pozzi (1980) in
Rovasenda heathland; its presence should
be considered occasional, because this
species usually prefers areas with sandy
soils (Lanza 1983). It is quite difficult to
explain the habitat preferences of the
Italian Crested Newt; in particular, those
preferences which do not appear clearly
related to the variables used in the present
study.
Our results confirm that heathlands could
be important for the preservation of some
amphibian populations and for the maintenance
of high amphibian diversity.
Figure 3. Association between the variables calculated with correspondence analysis
(Axis 1: Eigenvalue = 0.160, Percentage of variance = 74.7; Axis 2: Eigenvalue = 0.029,
Percentage of variance = 13.5).
CA variable scores

160 Biota 3/1-2,2002 SCALI & GENTILLI
REFERENCES
BEEBEE, TJ.C. 1996: Ecology and Conservation of Amphibians. Chapman & Hall, London:
214 pp.
BORKIN, L.J. & VEITH, M. 1997: Bufo bufo (Linnaeus, 1758). In: Case, J.P., Cabela, A.,
Crnobrnja-lsailovic, J., Haffner, P., Lescure, J., Martens, H., Martinez Rica, J.P.,
Maurin, H., Oliveira, T., Sofianidou, T.S., Veith, M. & Zuiderwijk, A. (eds.), Atlas
of Amphibians and Reptiles in Europe. Societas Europea Herpetologica &
Museum National d'Histoire Naturelle (IEGP/SPN), Paris: 118-119.
DENTON, J.S. & BEEBEE, TJ.C. 1996: Habitat occupancy by juvenile natterjack toads (Bufo
calamita) on grazed and ungrazed heathland. Herpetological Journal 6: 49-52.
GROSSENBACHER, K. 1997: Rana latastei Boulenger, 1879. In: Case, J.P., Cabela, A.,
Crnobrnja-lsailovic, J., Haffner, P., Lescure, J., Martens, H., Martinez Rica, J.P.,
Maurin, H., Oliveira, T., Sofianidou, T.S., Veith, M. & Zuiderwijk, A. (eds.), Atlas
of Amphibians and Reptiles in Europe. Societas Europea Herpetologica &
Museum National d'Histoire Naturelle (IEGP/SPN), Paris: 146-147.
LANZA, B. 1983: Anfibi e Rettili. Guide per il riconoscimento delle specie animali delle acque
interne italiane. 27. Anfibi, Rettili (Amphibia, Reptilia) [Collana del progetto finalizzato
"Promozione della qualita deH'ambiente". AQ/1/205]. Roma; Consiglio
Nazionale delle Ricerche.
NOLLERT, A. & NOLLERT, C. 1992: Die Amphibien Europas: Bestimmung - Gefahrdung -
Schutz. Franckh-Kosmos, Stuttgart.
POZZI, A. 1980: Anfibi e Rettili della brughiera di Rovasenda (Piemonte). AQ/1756-67.
Quaderni sulla "Struttura delle Zoocenosi terrestri". C.N.R., Roma: 467-472.
SCALI, S. 1993: Osservazioni su Rana latastei e Triturus vulgaris meridionals nel Parco
delle Groane (Lombardia, Italia). In: Ferri, V. (ed.). Atti del Primo Convegno
Italiano sulla Salvaguardia degli Anfibi, Quaderni della Civica Stazione
Idrobiologica Milano 20: 109-116.
SCALI, S. 1995: Amphibians and reptiles of Groane Regional Park (Lombardy, NW Italy).
First census and ecological notes. In: Llorente, G.A., Montori, A., Santos, X. &
Carretero, M.A. (Eds.). Scientia Herpetologica. Asociacion Herpetologica
Espanola, Barcelona: 307-311.
SPELLERBERG, I.F. 1989: An assessment of the importance of heathlands as habitats for reptiles.
Botanical Journal of,the Linnean Society 101: 313-318.
STUMPEL, A.H.P. 1992: Reptile management problems in Netherlands heathlands. In:
Korsos, Z. & Kiss, I. (eds.), Proc. Sixth Ord. Gen. Meet. S.E.H, Budapest, 1991:
421-424.
ZUFFI, M. 1987a: Anfibi e Rettili del Parco lombardo della Valle del Ticino: risultati preliminari
e proposte gestionali. Quaderni Civica Stazione Idrobiologica, Milano 14: 7-
65.
ZUFFI, M. 1987b: Su alcune stazioni di Podarcis sicula campestris (De Betta, 1857) della
Lombardia occidentale (Reptilia Lacertidae). Atti della Societa italiana di Scienze
natural! e del Museo civico di Storia naturale, Milano 128: 169-172.

SCALI, CORTI, GENTILLI, LUISELLI, RAZZETTI & ZUFFI Biota 3/i-a, 2002 161
Continental versus Mediterranean
European Whip Snake Hierophis
viridiflavus: a morphometric
approach
Stefano SCALI1, Claudia CORTPf Augusto
GENTILLI3, Luca LUISELLI4, Edoardo
RAZZETTI5 & Marco A.L. ZUFFI5*
'Museo Civico di Storia Naturale, c.so Venezia 55, 20121 Milano-ltaly
2Dip. Biologia Animale e Genetica, University of Florence, via Romana 17, 50125
Florence-Italy
3Dip. Biologia Animale, University of Pavia, p.za Botta 9, 27100 Pavia, Italy
"Istituto di Studi Ambientali "Demetra" (E.N.I. S.p.A.), via dei Cochi 48/B, 1-00133
Rome, Italy; and F.I.Z.V., via Cleonia 30, 1-00152 Rome, Italy; and Department of
Biological Sciences, Rivers State University of Science and Technology, P.M.B. 5080, Port
Harcourt (Rivers State)-Nigeria
^Corresponding author: Museo di Storia Naturale e del Territorio, University of Pisa, via
Roma 79, 1-56011 Calci (Pisa)-ltaly
E-mail: marcoz@museo.unipi.it.
Abstract
Hierophis viridiflavus, a wide distributed European colubrid, has been subjected to taxonomic
studies, with particular reference to morphological aspects of subspecific variation.
Other aspects of biology, ecology, and ethology are still anecdotal. Because of its
wide distribution and extreme abundance around the Tyrrhenian and Ligurian Seas, it
could be of great interest to test if any eventual geographic variation of morphological
features could be related to different ecological adaptations (i.e.: dietary habits, reproductive
strategies). We focused our attention on the occurence of: a) larger body sized
Hierophis viridiflavus in peninsular and continental (i.e. the Po valley area, pre-Alpine
areas) Italy, b) smaller body sized Hierophis viridiflavus on small Mediterranean islands
if compared to those with larger body size on large Mediterranean islands and mainland
areas. The main set of results clearly indicate a strong differentiation among the considered
populations: Hierophis viridiflavus of continental and north peninsular Italy do
not differ from the populations of the Ligurian and Tyrrhenian islands. On islands, the
Whip snakes of Corsica and of small islands in the Tuscan Archipelago display a significantly
high number of ventrals while in the Sardinian and Sicilian populations a lower
number of ventrals has been found; the lowest number of ventrals is displayed by the
Sicilian population. These morphological differences suggest the occurrence of separate

162 BJOta 3/i-a, 2002 SCALI, CORTI, GENTILLI, LUISELLI, RAZZETTI & ZUFFI
evolutionary patterns. Therefore, as a consequence we propose to assign the Sardinian
populations to Hierophis viridiflavus sardus (Suckow, 1798) and the Sicilian populations
to Hierophis viridiflavus xanthurus (Rafinesque Schmaltz, 1810).
Key words: morphometrics, snakes, Hierophis viridiflavus, Mediterranean distribution
Received 19 October 2001; accepted 6 April 2002
INTRODUCTION
The European Whip snake, Hierophis
viridiflavus, has a southwestern European
distribution, inhabiting a great variety of
habitats from sea level up to 2000 m a.s.l.
(Heimes 1993, Naulleau 1997). Formerly
divided into three different subspecies,
Hierophis viridiflavus viridiflavus, H.v.carbonarius
and H.v. kratzeri, it has been
quite recently subjected to a taxonomic
revision by Schatti & Vanni (1986): these
authors suggested that most of the variability
observed is due to very high phenotypic
variation, and therefore this
species must be considered monotypic
(Schatti & Vanni 1986, Heimes, 1993).
On the other hand, very recent (in the
present issue) genetic studies carried out
by Joger and co-workers show that two
clearly different genetic groups could be
identified, a western one occurring in
France, Switzerland and Italy west of the
Apennines, and an eastern one found in
Croatia, Slovenia and eastern Italy (Nagy
et al. 2001). Also, an increasing interest
regarding ecology and biology of the
European whip snake is also now evident
(Ciofi & Chelazzi 1991, Capizzi et al.
1995, Capula et al. 1997, Springolo &
Scali 1998, Scali & Montonati 2000, Zuffi
2001 a), but other aspects of life history
traits are still anecdotal (Zuffi 2001 b,
Zuffi et al. 2000). Because of the wide
distribution of this species and its abundance
around the Tyrrhenian and
Ligurian Sea coasts, it could be of great
interest to test whether any geographic
variation of morphological features could
be related to different ecological adaptations
(i.e.: dietary habits, reproductive
strategies; Zuffi et al. 2000), as in the
case of mainland versus large or small
island populations. Recently, new interest
has been addressed to morphological
variation and ecological plasticity of some
Mediterranean reptiles (Delaugerre &
Cheylan 1992, Corti et al 2001, Luiselli &
Zuffi 2002). In particular we focused our
attention on: a) larger body sized H. viridiflavus
of continental and peninsular Italy,
b) smaller body sized H. viridiflavus of
small Mediterranean islands if compared
to those larger body sized specimens of
large islands and continental areas, in
order to understand which biological
structures could be related to life history
traits, such as reproductive patterns or
dietary habits (Delaugerre & Cheylan
1992, Zuffi 2001 a).
MATERIALS AND METHODS
Whip snakes were examined i) from the
mainland (Lombardy and Tuscany, continental
and peninsular Italy respectively);
ii) from large islands (Corsica, Sardinia,
Sicily); iii) from small islands (Elba,
Montecristo, Pianosa, Capraia-Tuscan
Archipelago; Tavolara, Caprera, Mai di
Ventre-Sardinian satellite islands) (Figure
1). Standard measurements on 326 specimens
of H. viridiflavus from herpetological
collections of the Universities of
Florence, Milan and Pisa and from private
collections were considered (± 1 mm):
snout-vent length (SVL), tail length (tL),
head length (HL), head width (HW),
interorbital length (INT-ORB), number of
ventrals (VS), and number of subcaudals
(SCS). Data are presented as average ± 1
SD. The measurements were natural log-

SCALI, CORTI, GENTILLI, LUISELLI, RAZZETTI & ZUFFI Biota 3/1-2, 2002 163
Figure 1. Study area and localities. 1 =
Tuscany; 1bis= Lombardy; 2=Elba;
3=Pianosa; 4=Montecristo; 5=Corsica;
6=Sardinia; 7=Sicily.
transformed to reach linearity and tested
for normality. We used the residuals of
head size parameters against SVL to
avoid any potential effect caused by
snake size; Student t tests were used for
the means between sexes. MANOVA was
used for (i) island vs. continent, and (ii)
between all studied localities (only those
with sample size > 4). ANOVA (with
post-hoc multirange tests, LSD test,
Lesser Significant Differences), to highlight
the significant group of variation
(only those with sample size > 4). Tests
were two-tailed, and set at ( = 0.05, with
SPSS 8.0.
RESULTS
Intersexual comparison
Whip snakes showed strong sexual size
dimorphism in all characters (all with P <
0.01), but HL (P > 0.05), with males evidently
larger and longer (SVL: 819.4 ±
147.5 mm, n = 179; HL: 25 ± 6 mm, n =
116; HW: 16.6 ± 3.6 mm, n = 105; INT-
ORB: 10 + 1.5 mm, n = 91) than females
(SVL: 728.5 ± 147.9 mm, n = 77; HL:
24.4 ± 6.1 mm, n = 54; HW: 13.5 ± 3.5
mm, n = 52; INT-ORB: 8.5 + 1.4 mm, n =
34), and with a lower number of VS
(203.8 ± 21.3 vs. 210.5 ± 14.6). TL and
SCS were excluded because of the high
frequency of damaged or injured tails.
Mainland-island comparison
Whip snake females were too scarce
(sample size < 4) from many localities for
any valid statistical approach to be carried
out Whip snake males appeared significantly
different in VS (F = 5.278, P <
0.01, df = 4, R2 = 0.315, Radj = 0.255)
and in one covariate (HL, F = 10.581, P <
0.01, df = 1) between mainland and
island (large plus small islands) categories,
but not because of geographic area (F =
3.834, P > 0.05).
All localities comparison (with N > 4)
Whip snake males showed clearcut differences
in VS (F = 4.931, P < 0.001, df = 9;
R2 = 0.52, Radj = 0.414), with no variation
of covariates, but a significant influence
of considered areas (F = 3.733, P <
0.01). Among these, snakes from
Montecristo, Pianosa, Elba and Corsica
have a significantly higher number of
ventral scales if compared to those from
Sardinia and Sicily (Figure 2).
Furthermore, the latter show the lowest
number of ventral scales of the Whip
snake populations considered.
DISCUSSION
Sexual dimorphism is well documented in
Hierophis viridiflavus, with larger and
longer males than females (Springolo &
Scali 1998, Scali & Montonati 2000). It
has been long questioned whether there
are possible advantages in having sexes
differ in head or body size. Shine (1994)
argued that the larger male is a common
pattern in species with evident male-male
combats; Bonnet et al. (1999) found that
Whip snake males are favoured in movement
and long displacements by having a
higher muscle mass if compared to

Biota 3/1-2, 2002 SCALI, CORTI, GENTILLI, LUISELLI, RAZZETTI & ZUFFI
Figure 2. Geographical variation of ventral scales (average ± 1 SD) of Hierophis viridiflavus
males.
o
M
220
210
200
190
N- 12 5 25 13 16 5 S 16
Tuscany Montecristo Sardinia Pianosa
Elba Corsica Sicily Lombardy
females. The occurrence of geographical
variation in Whip snake males in the
Mediterranean islands considered shows
that j) the high number of VS found in
Montecristo, Pianosa and Corsica and the
low number of VS found in Sardinia and
Sicily populations respectively are not
correlated to the area size, ii) the
observed variability of Hierophis viridiflavus
follows a north-south oriented pattern.
This variability in Hierophis viridiflavus
shares the same zoogeographical
history of many Mediterranean animal
populations (Corti et al. 1991, Masseti
1993). Island snakes could have colonised
the whole area during the Messinian
salinity crisis (6 MY). They may have differentiated
during several climatic
changes according to independent
trends. Sicilian whip snakes resemble the
snakes of southern Italy, such as those in
Calabria and Apulia (unpubl. data), with
constant melanism, reduced body size
and significant low number of ventral and
caudal scales. The pattern of scale variation
underlines an actual status of differ-
ent body size structure (Saint Girons
1978, Padoa 1981), a genetically determined
character (Shine 2000). This significant
difference throughout the species
distributive area, particularly evident in
the island populations considered, in
accordance with preliminary data sets
and suggestions presented by Nagy et al.
(2001), leads us to reconsider that some
of the Mediterranean populations of
Hierophis viridiflavus may be separated
taxonomic units. We therefore suggest
that the Sardinian populations could be
tentatively assigned to Hierophis viridiflavus
sardus (Suckow, 1798) and the
Sicilian populations to Hierophis viridiflavus
xanthurus (Rafinesque Schmaltz,
1810) (see Table 1 for morphometric
data), even if a further interdisciplinary
approach of both morphometric and
molecular techniques (see Nagy et al.
2001) should be a necessary step of
investigation.

SCALI, CORTI, GENTILLI, LUISELLI, RAZZETTI & ZUFFI 3/1-2, 2OO2 165
Table 1. Morphometric data of Sardinian and Sicilian proposed subspecies.
Hierophis viridiflavus sardus
Hierophis viridiflavus xanthurus
sex
male
female
male
SVL
n=16
749.1±96.4
558-860
n=3
656.7±55.9
620-721
n=5
545±215.3
372-860
tL
n=ll
278.2±51.3
192-334
n=3
217.7±27.6
189-244
n=5
176.4±56.6
121-271
VS
n=16
200.2±4.2
193-207
n=3
215.3±4.5
211-220
N=5
195.4±2.3
193-199
Acknowledgements
We thank the Italian Ministry of Agriculture and the National Park "Arcipelago
Toscano" for permission to collect data on the Natural Reserve "Isola di A/lontecristo",
the EU INTERREGG II project Corse-Toscana (Province of Livorno and University of Pisa)
for granting most of the research; the Parco Naturale di Migliarino, S.Rossore
Massaciuccoli (Pisa) for permission to enter most protected areas; the Parco Naturale
del Ticino and the Parco Naturale di Groane in Lombardy for permission to collect data;
and the Herpetological Department of the California Academy of Sciences (CAS) of San
Francisco (USA), the Museums of Natural History of Milan, Morbegno (Sondrio),
Genua, Florence and Pisa. We also wish to thank all colleagues, friends, co-workers and
students who provided useful information on preserved snakes in private collections.
Finally, we wish to thank the Museum of Natural History and Territory (University of
Pisa) for grants to MALZ.
REFERENCES
BONNET, X., NAULLEAU, G. & SHINE, R. 1999: The dangers of leaving home: dispersal and
mortality in snakes. Biological Conservation 89: 39-50.
CAPIZZI, D., LUISELLI, L. CAPULA, M. & RUGIERO, L. 1995: Feeding habits of a
Mediterranean community of snakes in relation to prey availability. Rev. Ecol.
(Terre Vie) 50: 353-363.
CAPULA, M., FILIPPI, E., LUISELLI, L. & TRUJILLO-JESUS, V. 1997: The ecology of the
Western Whip snake, Coluber viridiflavus (Lacepede, 1789), in Mediterranean
Central Italy (Squamata: Serpentes: Colubridae). Herpetozoa 10: 65-79.
CIOFI, C. & CHELAZZI, G. 1991: Radiotracking of Coluber viridiflavus using external transmitters.
J. Herpetol. 25: 37-40.
CORTI, C., LUISELLI, L. & ZUFFI M.A.L. 2001: Observations on the natural history and morphometrics
of the Montpellier snake, Malpolon mospessulanus, on Lampedusa
Island (Mediterranean Sea). Herpetological Journal 11: 79-82.
CORTI, C., NISTRI, A., POGGESI, M. & VANNI, S. 1991: Biogeographical analysis of the
Tuscan herpetofauna (Central Italy). Revista Espanola de Herpetologia 5: 51-75.
DELAUGERRE, M. & CHEYLAN, M. 1992: Atlas de repartition des batracien et reptiles de
Corse. Ecole Pratique des Haute Etudes. Oikema: Pamplona.
HEIMES, P. 1993: Coluber viridiflavus Lacepede, 1789 - Gelbgrune Zornatterln. In: Bohme,
W. (ed.), Handbuch der Reptilien und Amphibien Europas, Band 3/I. Schlangen
(Serpentes) I. Aula-Verlag, Wiesbaden: 177-198.
LUISELLI, L. & ZUFFI, M.A.L. 2002: Female life-history traits of the aspic viper Vipera aspis

BiQta 3/1-2, 2002 SCALI, CORTI, GENTILLI, LUISELLI, RAZZETTI & ZUFFI
and sand viper (V. ammodytes) from the Mediterranean region. In: Schuett,
G., & Hoggren, M. Beds.). Biology of Vipers. CPG/Biological Science Press: 279-
284.
MASSETI, M. 1993: Post-Pleistocene variations of the non-flying terrestrial mammals on
some Italian islands. Suppl. Ric. Biol. Selvaggina, 21: 209-217.
NAGY Z.T., JOGER U., GUICKING D. AND WINK, M. 2001: Phylogeography of the
European Whip snake Coluber (Hierophis) viridiflavus, inferred from nucleotide
sequences of the mitochondrial cytochrome B gene and ISSR genomic fingerprintings.
11th OGM SEH, Zalec, Slovenia, Abstracts. Biota 2, Supplement: 38.
NAULLEAU, G. 1997: Coluber viridiflavus (Lacepede, 1789). In: Case, J.P., Cabela, A.,
Crnobrnja-lsailovic, J., Dolmen, D., Grossenbacher, K., Haffner, P., Lescure, J.,
Martens, H., Martinez Rica, J.P., Maurin, H., Oliveira, M.E., Sofianidou, T.S.,
Veith, M., and Zuiderwijk, A.(eds.). Atlas of Amphibians and Reptiles in Europe.
Societas Europaea Herpetologica & Museum National d'Histoire Naturelle
(IEGP/SPN), Paris: 342-343.
PADOA, E. 1981: Manuale di Anatomia Comparata dei vertebrati. Feltrinelli, Milano.
RAFINESQUE SCHMALTZ, C.S. 1810: Caratteri di alcuni nuovi generi e nuove specie di animali
e piante della Sicilia, con varie osservazioni sopra i medesimi. Sanfilippo,
Palermo.
SAINT GIRONS, H. 1978: Morphologie externe comparee et systematique des Viperes d'
Europe (Reptilia, Viperidae). Rev. suisse Zool. 85: 565-595.
SCALI S. & MONTONATI S. 2000: Analisi multivariata del dimorfismo sessuale in due specie
di colubrini italiani (Coluber viridiflavus ed Elaphe longissima: Reptilia,
Serpentes, Colubridae) su basi biometriche. Atti I Congresso Nazionale della
Societas Herpetologica Italica, Torino 1996. Mus. reg. Sci. nat. Torino: 429-434.
SCHATTI, B., & VANNI, S. 1986: Intraspecific variation in Coluber viridiflavus Lacepede,
1789, and validity of its subspecies (Reptilia, Serpentes, Colubridae). Rev. suisse
Zool., 93: 219-232.
SHINE, R, 1994: Sexual size dimorphisem in snake revisited, Copeia 1994: 326-346.
SHINE, R. 2000: Vertebral numbers in male and female snakes: the roles of natural, sexual
and fecundity selection. J. Evol. Biol. 13: 455-465.
SPRINGOLO, M. AND SCALI, S. 1998: Sexual dimorphism and ontogenetic changes in
Coluber viridiflavus: a head morphometric analysis. In: Miaud, C. & Guyetant, R.
Current studies in herpetology. Proc. 10th OGM Societas Europaea
Herpetologica. Le Bourget du Lac-France: 413-417.
SUCKOW, G.A. 1798: Anfansgsrunde der theoretischen und angewandten Naturgeschichte
der Thiere. Weidmannischen Buchandlung, Leipzig
ZUFFI, M.A.L. 2001 a: Diet and morphometrics of Coluber (=Hierophis) viridiflavus on the
island of Montecristo (Tyrrhenian Sea, Italy). Herpetological Journal 11:123-125.
ZUFFI, M.A.L. 2001 b: Morphological and functional variability of the whip snake (Hierophis
viridiflavus) from Corsica: a comparative approach to other Mediterranean
islands and mainland. Interregg II - Programme de cooperation scientifique -
Corte, Pise, Sassari, Calgari. Bilan et perspectives. Congres Environment et identite
en Mediterranee, Actes., Corte 13-16 Giugno 2000: 109-112.
ZUFFI, M.A.L., CORTI, C. & LUISELLI, L.M. 2000: Significato adattativo della variabilita morfologica
nel biacco (Hierophis viridiflavus) del Mediterraneo. 61° Congresso
Unione Zoologica Italiana, S. Benedetto del Tronto (AP), 24-28 settembre 2000,
Abstracts, 1 pag.

SUROVA Biota 3/i-a, 2002 167
The role of frog egg aggregations
as a control of abiotic factors
Galina. S. SUROVA
Moscow State University, Biological Faculty, Department Theory of Evolution
E-mail: avs@avs.bio.msu.su
Abstract
Differences in oxygen concentration, temperature and pH between clutches of frog egg
aggregations in Rana temporaria and ambient water were estimated. It was suggested
that these parameters could have an effect on egg mortality. We found that oxygen
concentration, temperature and pH inside the clutch aggregations are different from
those in the ambient water. Usually the temperature is higher within clutches so the
accumulation of heat by eggs may have adaptive importance. Oxygen concentration is
lower and pH becomes a little higher at the end of embryonic life; this last phenomenon
should be the result of natural physiological processes. Nevertheless, investigative
parameters reflect the singularity of a particular locality in the pond.
Key words: Rana temporaria, clutches; aggregations; temperature; concentration of
oxygen
Received: 28 February 2002; accepted 17 May 2002

168 Biota 3/i-a, 2002 SUROVA
INTRODUCTION
In the middle latitudes brown frogs lay
their eggs in large aggregations. It is a
current opinion that such aggregations
are condensers of heat (Seale 1982) and
protect embryos and tadpoles from
morning frost. The aggregation diminishes
the effect of daily temperature fluctuations
and prevents the overcooling of
larvae (Waldman 1982). The thick layer
of slime protects embryos and still inactive
tadpoles from invertebrate predators
(Rough 1976) and the pernicious influence
of acid marsh waters (Freda 1986)
Thus, aggregations are a perfect protection
against external factors of the environment.
The larger the aggregation, the
more strongly expressed its protective
effect is (Waldman 1982).
On the other hand, the negative effect of
aggregation is also well known: individuals
from aggregated clutches die more
often than those from solitary ones,
because of high density (Surova &
Severtzov 1985). The factors affecting
mortality in tadpole aggregations when
the larvae become free-swimming are
now well known as an "effect of group",
i.e. they have an exclusively biogene origin
(Schwartz et al. 1976). Other factors
of tadpole and egg mortality during early
ontogeny have been very poorly investigated.
There is only supposition that
death can be caused by hypoxia and possible
poisoning by the eggs' own metabolism
products in the limited space of
aggregation (Surova & Severtzov 1985),
as updating of the environment occurring
here only by diffusion (Burggren 1985)
can be insufficient. There are no concrete
data in this area.
We undertook an attempt to characterize
temperature, oxygen and acid modes in
large aggregations of clutches of the
Common Frog Rana temporaria in two
constant ponds at the Biological station of
Moscow State University near
Zvenigorod, Moscow region, Russia.
MATERIAL AND METHODS
Characteristics of habitat
Pond 1 is situated in a water-meadow
near the Moscow river and is sunny practically
all the time. The shallow places
where the clutch aggregations were
located were usually very warm. The
study was carried out on two aggregations
(N1 and N2), located near each
other. Pond 2 is smaller than 1. It is a part
of a wood ravine which retains water and
is very shady. The aggregations of clutches
here (N3) were only occasionally
warmed by the sun during the daytime.
Measurements of parameters of
environment
Measurements began at the end of the
spawning period, when most frogs have
already laid their eggs (stage of development
up to the middle of the gastrula).
The data were taken over a period of two
weeks. Measurements in both ponds
were stopped when the tadpoles reached
the stage of free-swimming larva (stages
of development according to Dabagjan &
Sleptsova (1975)).
The measuring instruments were positioned
at different points of an aggregation:
at the centre and the edges, on the
surface and at various depths (from 3 up
to,. 9 points .for one measurement).
Simultaneous parameters were measured
in ambient water at a distance 15-20 cm
away from an aggregation on the surface
and at the same depth (also 3-9 points).
All measurements were taken at approximately
the same points every time. The
final results are given on the diagrams as
average values on all points for one
measurement. Differences were tested
with a t-test. The acidity was measured
by a portable field pH-meter, temperature
and content of oxygen (in mg/l) - by
a portable oxymeter.
RESULTS
In pond 1 spawning began on April 10.

SUROVA Biota 3/1-2,2002 169
Table 1. Characteristics of development and pH in the aggregations.
Stages of development according to Dabagjan & Sleptsova (1975):
0 - beginning of cleavage; 7 - middle of cleavage; 20 - neurula; 26 - 28 - tail bud; 29
- 30 - hatching; 33 - 37 - external gills; 38 - 39 - beginning of development buds of the
legs; 39 - free swimming larvae
Nof
Aggregation
1
2
3
Measurement
date
IV
3V
5V
10V
14V
IV
3V
5V
10V
14V
IV
3V
5V
10 V
14V
Age (days)
17
19
22
28
32
17
19
22
28
32
5
7
9
14
18
By the 17th of April several aggregations
of clutches had formed. Two of them, situated
in the warmest part of the pond,
were taken for study. Aggregation 1 had
a total area of 1.25 m2, number of clutches
- 484, density - 387.2 clutches/m2.
Aggregation 2 differed little in size (1.05
m2), but it had a smaller number of
clutches (174), as a near-bottom layer
was absent, and consequently their density
was much smaller: (165.7 cl./m2). In
the colder pond #2 spawning began on
April 15, and a single, but very large
aggregation was formed only by May 1,
with an area of about 6 m2. (1.5m _ 4m),
number of clutches - 611, and density -
101.8 cl./m2.
The water in pond 1 was neutral all the
time; in pond 2 it was alkaline (Table 1).
The pH value in aggregations follows the
same pattern, i.e. the acid mode in aggregations
of clutches corresponds to the
PH
in the water
6.9
7.0
7.0
7.0
-
-
7.1
7.0
6.9
-
-
7.8
7.8
7.7
-
pH in the
aggregation
8.2
6.9
6.8
-
-
-
7.7
6.9
6.7
-
-
7.9
7.7
7.6
-
Stages of
development
27-28
29-30
33-37
38-39
39
27
29-30
33-37
38-39
39
0-13
7-20
26-28
29-30
33-37
acid mode of the pond. However, after
hatching, the medium in aggregations
became a little bit more acid. This is obviously
connected with the beginning of
active feeding and excretion of products
of metabolism. pH values in the pond and
in aggregations are favourable and are far
from the threshold value of acidity conducive
to deviations of development and
embryo mortality (pH of 5.0 and lower;
Freda 1986). The temperature and oxygen
mode changes are given on the diagrams
(Figure 1 - 3). In both ponds, temperature
in aggregations was 1-3
degrees higher than in ambient water.
This is especially appreciable in aggregation
2, where the difference is significant
through the whole ontogeny period
(Figure 2). In aggregation 1 the significant
differences were observed only at
the stage of hatching and at the stage of
free-swimming larva in a very dense,

170 'Biota 3/1-2,2002 SUROVA
Figure 1. Temperature and oxygen content
in ambient water and in the aggregation
N 1.
Legend: temperature; 1- in aggregation,
2 - in the water;
Oxygen; 3 - in aggregation, 4 - in the
water, * - p < 0.05; ** - p < 0.01.
May. I May. 3 May. 5 May. 7 May.9 May II May. 13
Figure 2. Temperature and oxygen content
in ambient water and in the aggregation
N 2.
For legend see figure 1.
May, I May. 3 May, 5 May. 7 May. 9 May 11 May. 13
Figure 3. Temperature and oxygen content
in ambient water and in the aggregation
N 3.
For legend see figure 1.
May. 1 May. 3 May, 5 May, 7 May, 9 May II May. 13
obviously thermal aggregation of tadpoles
(Figure 1). In the cooler pond 2, the
temperature in the aggregation was significantly
higher than in ambient water
(Figure 3). This relation was reversed on
May 10. It happened because the temperature
fell dramatically in a short time.
The temperature in the ponds became
lower, but in pond 2 it cooled down much
more strongly (up to 8.7°C, but in pond 1
upto10.5°C).
The change of oxygen mode in aggregations
also has a general tendency: the
level of oxygen in aggregations of clutches
is always lower than in ambient water.
In aggregation 2 these differences are
significant for all measurements (Figure
2). In aggregation 1, at the egg stage and
at hatching the level of oxygen in clutches
and in the surrounding water has no
significant differences, although the
means are very different (see Figure 1).
After hatching the deviation becomes
less, and the difference in oxygen content
in aggregations and ambient water
becomes significant. In aggregation 3 the
oxygen level was different at all times
except at the stage of tail budding (May
5), when the level of oxygen in the
aggregation and in the water was practically
equal (Figure 3). On the same diagram
it is clearly visible that at a practically
constant level of oxygen saturation
in the water from May 5 up to May 14,
the oxygen level in an aggregation of
clutches, and then tadpoles, is steadily
reduced, reflecting increasing consumption
of oxygen during the process of
development.
DISCUSSION
The consumption of oxygen by embryo
and larvae considerably reduces its level
in the aggregation. On one hand, our
study has obviously shown that the living
conditions of aggregation are closely connected
with ambient values. On the other
hand, the aggregations of clutches form-

SUROVA 3/1-2, 2OO2 171
ing thick layers of mucilage, and after
hatching - aggregations of larvae with
high density, are an independent system
with temperature higher than ambient, a
little more acidic medium and a lower
content of oxygen. The accumulation of
heat by eggs and dark coloured tadpoles
has obvious adaptive importance, as was
already mentioned above.
Such a strong downturn of temperature
as occurred in pond 2 on May 10 could
probably not be compensated for by any
accumulation of heat in an aggregation.
This indicates that the heat-sink ability of
an egg mass is not absolute; it has some
limits which depend on ambient temperature.
In other words, the higher the temperature,
the better expressed is the ability
of an aggregation to accumulate and
to keep heat.
The other parameters we tested probably
have no direct adaptive importance but
most likely reflect natural physiological
processes in embryos and larvae. Such a
small increase of acidity as we observed
can not render oppressive action upon
growth and development. If this does
occur, it is more likely because of excretion
specific metabolites under a high
density of tadpoles (Schwartz et al.
1976), but the pH level is not affected by
this factor.
The essential decrease of ambient oxygen
level - especially just after hatching - can
result in hypoxia and delay of ontogeny.
However, it is necessary to note that
hypoxia, naturally arising in aggregations,
serves as a physiological trigger for hatching
(Petranka et al. 1982). Certainly, oxygen
level and pH which conduce to
oppression can only be shown in further
laboratory and field experiments. But the
data given here show that the fluctuations
of temperature, oxygen and pH in
aggregations do not go beyond the limits
capable of causing the death of larvae.
Nevertheless, our earlier study showed
that the presence of aggregations could
result in 70% larvae mortality (Surova &
Severtzov 1985). Thus, the role of aggregation
in the life of embryos and larvae is
ambiguous and encourages us to continue
to study the balance between the
adaptive protective effect of aggregations
and the high price for selection of resistance
to high density.
Conclusions
1. In aggregations the temperature is
always higher than in ambient water.
However, during a long and strong
enough decrease of temperature, the
aggregation can not effectively retain
heat.
2. The level of oxygen in an aggregation
is always lower than in ambient water.
The content of oxygen in an aggregation
is constantly reduced with the passage of
time. This is not always connected to parallel
changes in environment, but more
likely by its consumption by larvae.
3. The acidity in an aggregation does not
differ greatly from the acidity in water,
only increasing slightly during late stages
of development.
4. The changes of investigated parameters
in an aggregation are closely connected
to their changes in ambient water.
They have their own character for the
particular pond and place of aggregation
locality.
5. Nevertheless, an aggregation is a more
or less independent system. It has the
original profile of parameters of internal
abiotic factors that, in a number of cases,
can have adaptive importance.

172 Biota 3/i-2, SUROVA
REFERENCES
DABAGYAN, N.B. & SLEPTZOVA, L.A. 1975: Common frog Rana temporaria. Obyekti
biologii razvitiya. Moskva. Nauka: 442-462. In Russian.
SCHWARTZ, S.S., PYASTOLOVA, O.A., DOBRINSKAYA, A.A. & RUNKOVA, G.G. 1976:
Effect of group in the populations of water animals and chemical ecology.
Moskva. Nauka. 152 p. In Russian.
SUROVA, G.S. & SEVERTZOV, A.S. 1985: The mortality of the common frog (Rana temporaria)
in the early ontogenesis and its factors. Zool. Journ. 64: 61-71.
BURGGREN, W. 1985: Gas exchange, metabolism, and "ventilation" in gelatinous frog egg
masses. Physiol. Zool. 58: 503-514.
FREDA, J. 1986: The influence of acidic pond water on amphibians: a review. Water Air and
Soil Pollut. 30: 439-450.
PETRANKA, J.W., INST, J.I. & CRAWFORD, E.G. 1982: Hatching of amphibian embryos: the
physiological trigger. Science 217: 257-259.
POUCH, F.H. 1976: Acid precipitation and embryonic mortality of spotted salamanders,
Ambystoma maculatum. Science 192: 68-70.
SEALE, D.B. 1982: Physical factors influencing oviposition by the woodfrog, Rana sylvatica,
in Pennsylvania. Copeia 3: 627-635.
WALDMAN, B. 1982: Adaptive significance of communal oviposition in woodfrogs (Rana
sylvatica). Behav. Ecol. and Sociobiol. 10: 169-174.

YOROS, KORSOS & SZALAY _ BlOta 3/1-2, 2003 _ 173
A comparative morphological
study of the two Hungarian
discoglossid toad species
Bombina spp
Judit VOROS1, Zoltan KORSOS1 &
Ferenc SZALAY2
'Hungarian Natural History Museum, Baross u. 13, H-1088 Budapest, Hungary
E-mail: jvoros@zoo.zoo.nhmus.hu, korsos@zoo.zoo.nhmus.hu
2Szent Istvan University, Faculty of Veterinary Science, Department of Anatomy and
Histology, Istvan u. 2, H-1077 Budapest, Hungary
E-mail: fszalay@univet.hu
Abstract
In order to study the hybridization of the two Hungarian discoglossid toad species (Firebellied
Toad, Bombina bombina and Yellow-bellied Toad, Bombina variegata), external
morphological features (body length, head length, head width, forelimb length, femur
length, tibia length, and foot length) of 271 specimens from 18 different localities of
Hungary were examined. In the hybrid zone (located in the northern region of
Hungary) between the characteristic ranges of the two species, intermediate individuals
can be found showing intermediate morphological characters. A new videomorphological
method was developed to quantify the external morphological characters, mainly
the colour pattern on the belly (circularity, mean area, and mean circumference of
patches, patch density, area ratio, ratio of mean area and mean circumference, colour
density), in order to facilitate reliable identification of pure and hybrid specimens. With
the examination of specimens collected from different regions of the country, a hierarchical
phenetic classification was performed to demonstrate the situation of the
Bombina hybrid zones in Hungary.
Key words: Bombina bombina, Bombina variegata, hybridization, morphological characters,
videoanalysis, colour pattern, phenetic classification
Received 30 November 2001'; accepted 12 February 2002

174 3/1-2, 2OO2 VOROS, KORSOS & SZALAY
INTRODUCTION
Natural hybridization between the two
European toad species, Fire-bellied Toad
Bombina bombina and Yellow-bellied
Toad Bombina variegata, was first reported
in Hungary by Mehely (1892). The
ranges of Fire-bellied Toad and Yellowbellied
Toad meet in a wide front extending
from Austria along the southern edge
of the Danube Valley to the Black Sea and
surrounding the Carpathian Mountains
along their foothills (Szymura 1993).
Hungary is an interesting part of the
Bombina hybrid zones. In the Great Plain,
Fire-bellied Toad is very common, occuring
in most the ponds. However, only isolated
Yellow-bellied Toad populations can
be found in the hilly regions of the country
(above 300 m a.s.l.). In the regions
where the ranges of the two species
meet, they interbreed and form hybrid
populations. Bombina hybridization in
Hungary has been investigated in detail in
the Matra Mountains and the Aggtelek
Karst by Gollmann (1987 and 1986,
respectively). It has been reported from
the Mecsek Mountains (Mehely 1904),
the Biikk Mountains (Dely 1996), and
observed in the Zemplen Mountains as
well (Voros 2000).
In our study we investigated individual
species identification of Bombina populations
throughout Hungary, using external
morphological characters. We compared
samples from populations of both the
typical habitats and from those with
intermediate characters.
We examined whether hybrid populations
can be found in the presumed
hybrid zones and what percentage of the
measured specimens can be considered
as hybrid. We also determined the most
reliable character among 12 for identification
of the two Bombina species.
MATERIAL AND METHODS
Our study area contained 18 different
study localities in Hungary (Figure 1).
These were both typical Fire-bellied Toad
and Yellow-bellied Toad habitats.
Localities where the ranges of the two
species meet, and hence interbreed, were
preferred when selecting sampling sites.
Altogether 271 Bombina specimens were
studied, using two morphological methods.
Seven morphometric characters
were measured in the field, with an accuracy
of 0.1 millimetre: body length, head
length, head width, forelimb length,
femur length, tibia length, and foot
length.
A short video recording was shot in the
field of the ventral surface of every specimen,
which was later analysed by a
newly developed computer program to
describe details of the colour pattern of
the belly. This method was chosen
because, superficially, the two species can
be distinguished most easily by observations
of the colour and pattern of the
ventral surface. Video tapes were played
and paused at each individual. A rectangle
on the belly of the specimens
(between the four limbs) was selected by
the program, and this part of the ventral
surface was analysed. The exact characters
calculated by the computer program
were: circularity, mean area, and mean
circumference of patches, patch density,
area ratio (ratio of the colour
patches/background), ratio of mean area
and mean circumference, and colour density.
The computer program will be
described in detail and published separately,
and will be available in the future.
Multivariate discriminant analyses were
carried out for morphometric and videomorphological
characters, first separated,
then combined together. Before the
analysis we grouped all specimens into
Fire-bellied Toad or Yellow-bellied Toad
species according to the classical morphological
methods (Mehely 1891). There
were geographical regions of Hungary
where we could find clearly identifiable
individuals, such as the Hortobagy,

VORC3S, KORSOS & SZALAY Biota 3/i-a, 2002 175
Tihany, and Dinnyes for Fire-bellied Toad,
and the Mecsek Mts. for Yellow-bellied
Toad, and we could compare the populations
from these regions to the others. A
phenetic classification (unweighted pairgroup
average method with Euclidean
distances) of the 18 sampling sites was
applied to define the hierarchical relation
of the populations. All analyses were carried
out using Statistica for Windows
(Version 5).
RESULTS
Discriminant analysis
A discriminant analysis was applied for
the two feature group (morphometric
and videomorphological), first separated
and then together. The first part of the
results shows the significance of the characters
in question affecting the segregation
of the groups.
ed in the mean circumference of the
patches, area ratio, colour density, forelimb
length, and tibia length (Table 1).
The other result of the discriminant analysis
was the misclassification probability, in
case of identification based on classical
morphological features, into two species.
The classification results show that preidentification
with the classical morphological
method corresponds by 92.46%
to identification with the computer program
based on the analysis of the colour
patterns of the belly (Figure 2), but by
only 87.44% with morphometric measurements
(Figure 3.). If we take into consideration
all 12 features, the correspondence
is 95.48% (Figure 4).
At the same time, the misclassification
probabilities show the overlap between
the two pre-identified groups (i.e. the
two species). These percentages (7.54%,
Table 1. The significance and the Wilks' Lambda values of the 12 characters.
Characters
Mean area
Mean circumference
Patch density
Area ratio
Ratio of the area and circ.
Colour density
Circularity
Head length/ head width
Forelimb length
Femur length
Tibia length
Foot length
According to the significance and the
Wilks' Lambda values (Wilks' Lambda is a
statistic for the differentiation of two statistical
populations; its value is 1 if the centroids
of the two groups are indistinguishable,
and 0 at maximal divergence (Podani
1997), significant differences were detect-
p-values
0.326
0.0001
0.136
0.0003
0.224
0.0003
0.143
0.259
0.001
0.027
0.01
0.117
Wilks' lambda values
0.246
0.269
0.248
0.262
0.247
0.262
0.248
0.247
0.259
0.252
0.254
0.248
12.56%, 4.52%) indicate the misclassification
of the videomorphological method
(Figure 2-4).
Phenetic classification
To understand the relationships between
the populations, we made a phenetic

176 Biota 3/1-2,2002 VOROS, KORS6S & SZALAY
Figure 1. Sampling localities: 1. Aggtelek,
2. Zemplen, 3. Beregi Plains, 4.
Hortobagy, 5. Matra, 6. Kiskunsag, 7.
Pilis, 8. Vertes, 9. Lake Velence, 10. River
Drava, 11. Mecsek, 12. Lake Kis-Balaton,
13. Balatonfenyves, 14. Tihanyi
Peninsula, 15. Bakony, 16. Orseg, 17.
Hansag, 18. Szigetkoz.
classification of the samples according to
locality (samples from 38 populations).
The phenogram (Figure 5) gave a hierarchical
connection clearly defining the two
Bombina species. There are specimens
from closely situated study plots which
belong to different species.
DISCUSSION
Among the videomorphological characters,
the mean circumference of patches,
the area ratio, and colour density
(p=0.0001, 0.0003, 0.0003), and among
the morphometric characters, the forelimb
length and tibia length (p= 0.0027,
0.01) were found to be most reliable for
the differentiation of the two species.
Regarding a single individual, consideration
of all 12 characters proved to be the
most reliable method for species identification.
According to our study of the Aggtelek
region (north-eastern Hungary) there are
hybrid populations. The samples of
Aggtelek were taken from closely situated
sites, and because the specimens
belong to both groups (species), they
have intermediate morphological characters.
The ranges of the two species meet
here and individuals interbreed in these
habitats. The Pilis Hill (close to Budapest)
also appeared to be an interesting locality,
since specimens from this site had an
uncertain position in the phenogram.
To be able to define the hybrid zones of
Bombina species, we would like to study
the genetic background of the morphological
characters in the future.
Figure 2. The overlap in case of the analysis of colour patches of the belly. The striped
columns show Fire-bellied Toad Bombina bombina, the blank columns show Bombina
variegata. Overlap (misclassification probability) 7.54%.
30
o 15
N N
-4,5 -4,0 -3,5 -3,0 -2,5 -2,0 -1,5 -1,0 -0,5 0,0 0,5 1,0 1,5 2,0 2,5 3.0 3.5 4,0

VOROS, KORSOS & SZALAY Biota 3/i-a, 2002 177
Figure 3. The overlap in case of the morphometric measurements. The striped columns
show Fire-bellied Toad Bombina bombina, the blank columns show Bombina variegata.
Overlap (misclassification probability) 12.56%.
-5,0 -5,0 -4,0 -3,0 -2,0 -1,0 0,0 1,0 2,0 3,0 4,0
-5,5 -4,5 -3,5 -2,5 -1,5 -0,5 0,5 1,5 2,5 3,5 4,5
Figure 4. The overlap in case of analysis by all 12 features. The blank columns show
Fire-bellied Toad Bombina bombina, the striped columns show Bombina variegata.
Overlap (misclassification probability) 4.55%.
-5.5 -4.5 -3,5 -2,5 -1,5 -0.5 0,5 1,5 2,5 3,5
-5,0 -4,0 -3,0 -2,0 -1,0 0,0 1,0 2,0 3,0
Expected
Normal
Expected
Normal

178 Biota 3/1-2,2002 VOROS, KORSOS & SZALAY
Figure 5. Phenogram (UPGMA, Euclidean distances) of the samples according to localities.
The upper group corresponds to Fire-bellied Toad Bombina bombina, the lower
group to Bombina variegata.
Uertes
Plaiul
Hortobag
T ihany
Dinnyes
ftggtell
T iszahat
Szerenle
Hortob2
Baja
Kisbal
Mariaf
Hortobl
Ki skunsa
Taljand
Sz iget 1
Bakony
Hansag
Sz iget 2
Pilis
Ronania
Matra
Ujhutal
UjhutaS
Ujhuta4
Regec
flggtelS
Ohuta
Ujhuta2
Gone 2
Goncl
Zenplenl-
Mecsek i
Mecsek 2
Ujhuta3
UjhutaG
Acknowledgements
We would like to thank the Nature Conservation Directorates of the Balaton Upland
and the Duna-lpoly National Parks for granting study permits in their respective territories.
We are grateful to numerous persons not listed here who have made possible the
collection of field data, and, furthermore, especially to Giinter Gollmann (University of
Vienna) for sharing his literature, and to Gabor Herczeg and Mihaly Foldvari (Budapest)
for technical support.
REFERENCES
DELY, O. GY. 1996: Amphibians and reptiles of the Biikk Mountains. In: Mahunka, S. (ed):
The fauna of the Bukk National Park. Vol. 2. Hungarian Natural History Museum,
Budapest: 535-572.
GOLLMANN, G. 1987: Bombina bombina and Bombina variegata in the Matra mountains:
New data on distribution and hybridization. Amphibia-Reptilia 8: 213-224.
f GOLLMANN, G., ROTH, P. & HODL, W. 1986: Hybridization between the Fire-bellied toads
Bombina bombina and Bombina variegata in the karst regions of Slovakia and
Hungary: morphological and allozyme evidence. Journal of Evolutionary Biology
1:3-14.
MEHELY, L. 1891: A magyar fauna Bombinatorjai es egy uj Triton (Molge) faj hazankbol.
MTA Mathematikai es Termeszettudomanyi Kozlemenyek 24: 553-574. (in

VOROS, KORSOS & SZALAY BJOta 3/1-2,2002 179
Hungarian)
MERELY, L. 1892: Beitrage zur Kenntnis der Bombinator-Arten, sowie deren Standorte und
Verbreitung in Ungarn. Mathematische und Naturwissenschaftliche Berichte aus
Ungarn 10: 55-79.
MF.HELY, L. 1904: A Mecsekhegyseg es a Kapela herpetologiai viszonyai. Allattarvi
Kozlemengek 3: 241-289. In Hungarian.
PODANI, J. 1997: Bevezetes a tobbvaltozos biologiai adatfeltaras rejtelmeibe. -Budapest,
412 pp. In Hungarian
STATISTICA FOR WINDOWS, Version 5., StatSoft, Inc.
SZYMURA, J. M. 1993: Analysis of hybrid zones with Bombina. In: Harrison, R. G. (ed.).
Hybrid zones and the evolutionary process. Oxford University Press, New York,
Oxford, pp. 261-289.
VOROS, J. 2000: A comparative morphological study on Hungarian Fire-bellied and Yellowbellied
toad (Bombina spp.) populations. M. Sc. Thesis, Szent Istvan University,
Faculty of Veterinary Science, Institut for Zoology, Budapest, 38 pp. In Hungarian.

ZAVADIL & SIZLING Biota 3/1-2. 2002 181
Morphological variability in the
newts of the Cristatus group
Vit ZAVADIL1 & Arnost L SIZLING1,2
Agency for Nature Conservation and Landscape Protection of the Czech Republic,
Kalicnicka 4-6, Praha 3, 130 00, Czech Republic
E-mail: zavadil@nature.cz,
2Faculty of Science at Charles' University, Department of Philosophy and History of
Science, Vinicna 7, Praha 2, 128 44, Czech Republic
E-mail: arnost.l.sizling@seznam.cz
Abstract
Between 1993 and 2000, a sample of 646 individuals belonging to the T. cristatus
group were collected and measured at sites in the Czech and Slovak Republics.
Additionally, 26 individuals of T. carnifex from Italy and Slovenia, 8 of T. dobrogicus
from Hungary and Rumania, and 6 of T. karelinifrom Turkey were collected for the purposes
of comparison with the newts determined as T. carnifex and T. dobrogicus in the
Czech and Slovak Republics. All of these individuals were measured with manual callipers
under anaesthesia by the same person to the nearest 0.05 mm in the standardised
manner.
The quantities measured were W (weight), L (snout-vent-length), Led (length of tail),
L.c. (length of head), Pa (average length of forelimbs), Pp (average length of
hindlimbs), LiE (interlimb distance), and Ltc (width of head). The colours of the body
and the existence of the crest were also recorded. The hypothesis that the Wolterstorff
Index (Wl) - which is counted as 100*Pa/LiE - separates the crested newt group into
single species has been tested. As expected, T. dobrogicus is the only species that Wl is
able to discriminate. However, we received the most significant results using modified
Wl calculated as 100*Pa/(LiE+6) for males and 100*Pa/(LiE+22) for females.
Key words: news, Cristatus group, Wolterstorff Index
Received 28 February 2002; accepted 10 May 2002

182 Biota "5/i-a, aooa ZAVADIL & SIZLING
INTRODUCTION
Until recently, Triturus cristatus newts
were supposed to be the only representatives
of the T. cristatus group in the
Czech Republic (Rocek 1992). In 1993,
Triturus dobrogicus was morphologically
determined in south Moravia at the confluence
of the Morava and Dyje rivers;
proof in accordance with ELFO followed
almost immediately (Zavadil et al. 1994).
In 1997, newts showing features typical
Table 1. Measured values.
of T. carnifex were found near the town
of Znojmo, at a site close to the border
with Austria (Zavadil & Reiter unpubl.).
Three years later, morphological analysis
conducted by Pialek et al. (2000) proved
that the observed population belongs to
T. carnifex. ELFO analysis by Horak &
Pialek (1999) determined the subpopulations
as hybrids T. carnifex x T. cristatus
and T. cristatus x T. dobrogicus.
This evidence of morphological variability
Short Name Fool Name Unit
W Weight grams
L snout-vent-length millimetres
Led length of tail millimetres
L.c.l length of head millimetres
L.c. 2 length of head millimetres
Pa average length of forelimbs millimetres
Pp average length of hindlimbs millimetres
LiE interlimb distance millimetres
Ltc width of head millimetres
Sp. oc. inter eyes distance millimetres
Cr. max. cd. maximally width of tail millimetres
Lc2
Lc 1
Figure 1. Measured values. See also Table 1.
Led

ZAVADIL & SIZLING BJOLa 3/1-2, aooa 183
motivated us to start research at multiple
sites in the Czech and Slovak Republics.
In this paper we try to answer the following
questions, which can help resolve
issues concerning the status of T. cristatus
behind the Alps and T. dobrogicus at the
edge of its range:
Are the Czech populations of T. carnifex
and T. dobrogicus morphologically different
from the Czech population of T.
cristatus?
Are the Czech populations of T. carnifex
and T. dobrogicus (which are considered
as marginal or isolated) morphologically
different from the populations of T.
carnifex and T. dobrogicus populations in
core areas of their distributions?
Is there any geographic variability in body
shape within the T. cristatus group?
MATERIAL AND METHODS
Between 1993 and 2000, 646 individuals
belonging to the T. cristatus group were
measured at sites in the Czech and Slovak
Republics. For the purposes of comparison
with the newts determined there as
T. carnifex and T. dobrogicus, an additional
26 individuals of T. carnifex from
Slovenia and Italy, 8 of T. dobrogicus
from Hungary and Rumania, and 6 of T.
karelinii from Turkey were examined.
Since inaccuracy may be caused when
measurements are taken by different persons
or on animals conserved in alcohol
(Arntzen & Wallis 1994), all the individuals
were measured with manual callipers
by the same person to the nearest 0.05
mm in the standardised manner and
under anaesthesia. The anaesthetic used
was 0.8% water solution of phenoxyethanol
(0.8 millilitre of phenoxyethanol
in 1 litre of water).
In accordance with a suggestion of
Klepsch (1994), measurements were
taken of weight (W), snout-vent-length
(L), length of tail (Led), length of head
Figure 2. The localities; large rings show sites on which the large samples were collected.

184 Biota 3/1-2,2002 ZAVADIL & SIZLING
(I.e.), average length of forelimbs (Pa),
average length of hindlimbs (Pp), interlimb
distance (LiE), width of head (Ltc),
distance between eyelids (Sp. oc) and
maximum width of tail (cr. max. cd.) (see
Table 1, Figure 1). Furthermore, the
colouring of the body and the existence
of the crest were recorded.
During the first three years, the studied
sites in the Czech Republic were chosen
randomly. Later on, as we focused on differences
between the T. cristatus and T.
dobrogicus species, we concentrated on
localities of T. cristatus remote from areas
inhabited by T. dobrogicus and on localities
situated at higher altitudes.
Since 1997, we have focused on sites
around the town of Znojmo, where the T.
carnifex had been found; since 1998 we
have examined sites in Slovakia situated
on the border between the ranges of T.
cristatus and T. dobrogicus (see Figure 2).
We used the Analysis of Principal
Component (PCA) for comparison of
samples from locations within the former
Yugoslavia (Kalezic 1990) and the
Wolterstorff Index (Wl) for comparison
with the results presented by Wolterstorff
(1924), Arntzen & Wallis (1994) and
Pialek et al. (2000).
Since neither of the methods could satisfactorily
show the existence of a morphometric
difference between T. cristatus
and T. carnifex, we decided to use the
Discriminant Analysis (DA) and the Cube
Analysis (dividing the W-L-Lcd-L.c.-Pa-
Pp-LiE-Ltc-Sp.oc.-cr.max.cd. space into
three coherent areas consisting of bond
cubes; see Appendix).
However, the best results were reached
by a modification of Wl, which we
named Intensified Wolterstorff Index
(i-WI; see below).
RESULTS AND DISCUSSION
Multivariate analyses
Three accomplished multivariate analyses
(PCA, Discriminant Analysis and Cube
Analysis - see Appendix) were used to
measure morphological and geographical
differences between studied species. The
results of PCA are concordant with the
works of Kalezic et al. (1990) and Pialek
& Zavadil (1997); however, the clusters
shown in our study were larger and consequently
overlapped. To obtain the best
result using Discriminant Analysis we
reduced measured variables so that they
were linearly independent (we used all
variables except W and L) and used a
quadratic model because of inequality of
covariance matrices and transformed
axes to normalize its distribution and
eliminate allometry. However, even when
all assumptions of DA were met as accurately
as possible, we did not obtain significantly
better results (measured in percentage
of successful determination) than
we obtained using only Pa and LiE variables.
This result suggests that these two
variables are adequate, but in the end we
constructed and carried out the Cube
Analysis to guarantee this result.
Therefore, we concentrate on these two
variables in the following text.
Wl analysis
Wolterstorff Index (Wl) analysis is a classical
method used to identify species
belonging to the T. cristatus group. It is
defined as 100*Pa/LiE. Though the Wl is
very helpful for the determination of T.
dobrogicus, it is not satisfactory for distinguishing
between T. cristatus and T.
carnifex (Kalezic et al. 1990, Arntzen &
Wallis 1994, 1999). The problem is that
the Wl decreases with growing body
length (see Figure 3 where the Wl corresponds
to the slope of the dotted lines).
When studying the Pa-LiE space (see
Figure 3), we realised that the decrease in
observed Wl was related to ineffective
distinction between the two species.
There is linearity between Pa and LiE, but
the borderlines/regression lines do not
attain zero. This could be caused by a dif-

ZAVADIL & SIZLING 3/1-2, 2OO2 185
Figure 3. Males-top, Females-bottom. There is a difference between Wl dotted lines
and i-WI full lines in Figs, a, b: rings correspond to T. carnifex, filled circles to T. cristatus,
triangles to T. dobrogicus and sharps to T. karelinii; sharps inside the circles mark
external individuals of T. ca. (I, SLO) and diamonds external individuals of T. do. (H,
RO).
[interlimb Distance (LIE) in mms
full lines (i-WI) does not attain zero
dotted lines (Wl) attain zero
ferent rate of growth in the larvae stage.
It is shown on Figure 3 where the dotted
lines correspond to Wl and the coloured
lines correspond to the intensified Wl,
which is calculated as 100*Pa/(LiE+22)
for females and 100*Pa/(LiE+6) for males
(see Appendix). The classical way to eliminate
allometry helping the logarithm
Interlimb Distance (LiE) in mms
function appears to be less powerful.
Box and Whisker plot of the i-WI is
shown in Figure 4 and the effectiveness
of discrimination in Table 2.
If the i-WI should turn out to increase
with growing length measures, it is probably
caused by the simplification of its
original form, which is ATN(Pa/(LiE+22))

186 Biota 3/i-a, 2002 ZAVADIL & SIZLING
Figure 4. Males-top, Females-bottom. Box and whisker plot of i-WI for different localities.
The size of sample "n" is for males 12 for site 1041, 21 for1161, 13for1171,31
for 1181, 15 for 1211, 12 for 1291, 6 for 1351, 8 for 1401, 22 for 1411, 4 for 1421, 6
for 1451, 15 for 1491, 6 for 2021, 7 for 2051, 7 for 2061, 11 for 2071, 1 for 3011, 1
for 4011, 11 for 5011, 2 for 6021 and 4 for 7011. For females the size of sample "n"
is 4 for 1041, 17 for 1171, 22 for 1181, 14 for 1211, 10 for 1291, 9 for 1351, 8 for
1391, 4 for 1401, 30 for 1411, 9 for 1451, 9 for 1481, 5 for 1491, 9 for 2021, 9 for
2061, 6 for 2071, 3 for 3011, 3 for 4011, 10 for 5011, 2 for 6021 and 2 for 7011.
cu
1041 1171 1211 1351 1411 1451 2021 2061 3011 5011 7011
1161 1181 1291 1401 1421 1491 2051 2071 4011 6021
1041 1181 1291 1391 1411 1481 2021 2071 4011 6021
1171 1211 1351 1401 1451 1491 2061 3011 5011 7011

ZAVADIL & SIZLING Biota 3 i-2,2002 187
Table 2. Effectiveness of i-WI Analysis.
MALES; i-WI = 100*Pa/(LiE+6) % of Successful
Determination
Species T. ca. T. cr. T.do.
T. ca. 45 9 0
83
T. cr. 7 135 0
95
T.do. 0 4 68
94
FEMALES; i-WI = 100*Pa/(LiE+22)
T. ca. T. cr. T.do.
T. ca. 35 7 0
83
T. cr. 1 154 8
94
T.do. 0 0 67
100
Table 3. Threshold values for Wl and i-WI according to different authors; many of these
were taken from graphs
WI - Threshold value
Males
Females
T. karelinii
T. carnifex
T. cristatus
T. dobrogicus
T. karelinii
T. carnifex
T. cristatus
T. dobrogicus
Our Study i-WI
from to
55.5
45.5
32
35.3
28.7
22
63
55.5
45.5
40
35.3
28.7
Our Study Wl
from to
70 75
60 75
50 69
38 55
60 65
50 65
32 61
30 ' 45
for females and ATN(Pa/(LiE+6)) for
males, where ATM is the mathematical
function arc tangent. However, there is
good statistical evidence that the i-WI
increases with growing width sizes (Pa,
Pp, Ltc, P < 0.05). All results remained
unchanged when the data for subadult
individuals had been added.
The thresholds of Wl and i-WI can be
found in Table 3. The evidence that the
Wl is higher for males than for females is
much stronger for samples collected
within single sites than for samples collected
on more sites (see Figure 4).
Geographical variability
We have no evidence for morphological
difference between the Czech population
of T. carnifex and populations of T.
Wolterstorff 1924
From to
69 82
63 73
59.8 65
45 52
67.5 72
52 64
49 54
34 45
Kalezicetal.1990 Arntzen & Walis 1994
from to from to
69 81 57 67
63 73 49 64
55 65 44 62
45 51 35 46
61 72 57 67
54 65 49 64
45 54 44 62
34 45 35 46
carnifex south of the Alps and T. dobrogicus
in the Czech and Slovak Republics
and the same species abroad. This is
because we have no evidence that for the
same species, the difference between the
Czech population and populations
abroad is bigger than the difference
among Czech populations (see Figure 4).
Although there are sites where the same
species of the T. carnifex group show a
statistically different body shape (see
Figure 4), it would not be correct to speak
about geographical variability. The reason
is that no correspondence was found
between the shape (i-WI and the results
of the Discriminant Analyses) and any
geographical attribute (altitude, the distance
between sites or any direction, as,
for example, East, West, SE, etc.).

188 Biota 3/i-a, 2002 ZAVADIL & SIZLING
A sound hypothesis seems to be that the
morphological differences between particular
sites can be caused by different
densities of larvae predators (Schmidt &
Van Buskirk2001).
Appendix
The Analysis of Principal Components
(PCA) transforms the axes of the examined
orthogonal space in order to show
clusters of the studied samples. If it is not
possible to see any clusters, there is no
evidence for their absence. The PCA was
used for data analyses in the works of
Kalezic et al. (1990) and Pialek & Zavadil
(1997).
The Discriminant Analysis calculates a
number D for a case (a single newt), and
for each group of a studied sample (TV
cristatus, T. carnifex, T. dobrogicus). The
group with the lowest D is assumed to be
the group into which the case (newt)
belongs.
The purpose of the Cube Analysis is to
divide morphometric space into two continuous
parts, which contain either T.
cristatus or T. carnifex individuals. In the
case of ten-dimensional space it would be
very difficult, so in the beginning we
reduced the data with the help of PCA
(the axes with the lowest variability are
eliminated). After data reduction, we
divided the reduced space into optimal
size cubes, in the same way that biologists
divide maps into quadrants for
quadrant mapping. The optimal size of
the cubes was taken empirically by testing
various sizes. The cubes create three
continuous areas: the first one contains
only T. carnifex, the second one only T.
cristatus, and the third one consists of
cubes in which both species could be
found. This algorithm was applied subsequently
on the third area.
Wl and i-WI analyses. In Figure 1 in
Arntzen & Wallis (1994) the T. cristatus
and T. marmoratus are plotted in Pa-LiE
state space. Making the best line for distinguishing
between the two species
(error of determination in percent is the
lowest) and extrapolating it, we realized
that the line does not attain zero and so it
can not represent a threshold for the Wl.
This is because the formula Wl =
1QO*Pa/LiE is the other form for a line
which attains zero. The form of this line
can be written as Pa = (WI/100)*LiE.
This phenomenon is also clearly apparent
in our data (see Figure 3). It means that
threshold lines in the Pa-LiE state space
do not attain zero, and the common form
of these lines must be Pa =
(i-WI/100)*LiE+b, where b represents
the point of intersection with the Pa axis.
This equation can be written as i-WI =
100*Pa/(LiE+c) where c is computed for
males and females separately (b =
c*[i-WI/100]). The number i-WI we call
intensified Wolterstorff index.
Acknowledgements
Special thanks to J. Pialek, D. Weber-Obdrzalkova, Jan Sula and Lukas Kratochvfl for
their assistance and to P. Belansky, D. Cogalniceanu, P. Dolezal, A. Horak, M. Kaftan, J.
Kautman, S. Koukal, A/I. Macholan, K. Poboljsaj, Puky M., A. Reitr, K. Rozfnek, R.
Rozinek, A. Ruxova, R. Zajfcek and J. Zima for their help during the collection of the
data sample.
REFERENCES
ARNTZEN, J., W. & WALLIS, G., P. 1994: The "Wolterstorff index' and its value to the taxonomy
of the Crested Newt superspecies. In: Bischoff, W., Bohme, W. &
Bottcher, I. (eds.): Okologie und Stammesgeschichte der Schwanzlurche. Abh.

ZAVADIL & SIZLING Biota 3/1-2,2002 189
Ber. Naturk., Magdeburg 17: 57-66
ARNTZEN,)., W. & WALLIS, G.f P. 1999: Geographic variation and taxonomy of Crested
Newts (Triturus cristatus superspecies): morphological and mitochondrial DNA
data. Contrib. Zool. 68: 181-203.
HORAK, A.r PlALEK, J. 1999: Genetic structure of crested newts (Triturus cristatus superspecies)
in the Czech Republic. Abstr. in 10th OGM SEN, Irakleio, 6.-10.
September 1999: 70-71.
KALEZlC M.r L.r DZUKlC G., STAMENKOVlC S., CRNOBRNJA, J. 1990: Morphometrics of
the crested newt (Triturus cristatus complex) from Yugoslavia: relevance for taxonomy.
Arch. biol. nauka, Beograd 42: 17-37.
KLEPSCH, L. 1994: Zur Artdifferenzierung der Kammolche (Tr/furuscr/steft/s-Artenkreis) im
Waldviertel: Morphometrische und molekulargenetische Untersuchungen.
Diplomarbeit, Universitat Wien.
PlALEK , J., ZAVADIL, V. 1997: Morphological differentiation between members of the
Triturus cristatus superspecies in the Czech Republic. In: Rocek, Z. & Hart, S.
(eds.): Herpetology '97: 162.
PlALEK, J., ZAVADIL, V., VALICKOVA, R. 2000: Morphological evidence for the presence of
Triturus carnifex in the Czech Republic. Folia Zoologica 49: 33-40.
ROCEK, Z. 1992: Triturus cristatus (Laurenti, 1768) - colek velky. In: Barus, V. & Oliva, O.
(eds.): Obojzivelnici, Amphibia. Fauna CSFR, Vol. 25. Academia, Praha: 115-122.
SCHMIDT, B. R. & VAN BUSKIRK, J. 2001: Verhalten, Wachstum und Morphologie von
Kammolch-Larven in der An- und Abwesenheit von Libellenlarven. In: Krone, A.
(ed.): Der Kammolch (Triturus cristatus) Okologie und Bestandssituation. Rana,
Sonderheft 4, Rangsdorf.
WOLTERSTORFF, W. 1924: Uberfcicht der Unterarten und Formen des Triton cristatus Laur.
Bl. Aquar.Terrarkde 33: 120-126.
ZAVADIL, V., PlALEK, J. & KLEPSCH, L. 1994: Extension of known range of Triturus
dobrogicus: electrophoretic and morphological evidence for presence in the
Czech Republic. Amphibia-Reptilia 15: 329-335.

ZUFFI, CENTILUI, RAZZETTI & SCALI Biota 3/1-2. 2002 191
Transition-hybridization areas in
parapatric species of Vipera aspis
group from northern Italy
Marco A.L ZUFFI1*P Augusto GENTILLI2,
Edoardo RAZZETTI2 & Stefano SCALP
'"Corresponding author: Museo di Storia Naturale e del Territorio, University of Pisa, via
Roma 79, 1-56011 Calci (Pisa)-Italy
E-mail: marcoz@museo.unipi.it
2Dip. Biologia Animale, University of Pavia, p.za Botta 9, 27100 Pavia, Italy
3Museo Civico di Storia Naturale, c.so Venezia 55, 20121 Milano-ltaly
Abstract
There is evidence of some contact between the parapatric species of Vipera aspis, Italian
V. aspis and Vipera atra Meisner, 1820 along a North-South oriented sector, from about
8°45' to 9° 20' Longitude East. V. atra from north-western Italy is characterized by a
high number of ventral scales and dorsal bars, and by a distinct morphological pattern
of the hemipenes that has elongated lobes, mostly covered by short and small spines,
while V. aspis has a relatively lower number of ventral scales and dorsal bars, and
hemipenes that show large lobes with the upper part completely covered by a calyculate
area. The intermediate specimens appear to be more similar to V.aspis than to V.
atra.
Key words: Vipera aspis (species group), northern Italy, contact zones.
Received: 3 October 2001; accepted 12 April 2002

192 Biota 3/1-2,2002 ZUFFI, GENTILLI, RAZZETTI & SCALI
INTRODUCTION
Paleontological and molecular biology
data indicate that the small vipers of the
western Palaearctic region, including the
Sand viper, Vipera ammodytes, and the
Asp viper, Vipera aspis, represent a complex
of phylogenetically closely related
taxa (Herrmann & Joger 1997, Lenk et al.
2001).
The Asp viper, with regards to colour
morphs and dorsal markings, is one of the
most variable snakes of the Palaearctic
region (Physalix 1968, Brodmann 1987).
Among the several subspecies which
have been described in the past (cf. Saint
Girons 1997), only atra, frandsciredi,
hugyi, zinnikeri and the nominal aspis
have been accepted (Zuffi & Bonnet
1999) (Figure 1). More recently, some of
these taxa have been recognised as good
species (Zuffi 2002) and they form part of
interest of this contribution. Inter and
intra-population analysis of life history
traits represents one of the most fascinating
approaches towards the description
and the comprehension of phenotypic
plasticity (Seigel & Fitch 1985, Bonnet &
Naulleau 1996, Shine 2000, Luiselli &
Zuffi 2002). The high morphological variability
of the Asp viper, and the wide
range of habitats in which this species
lives (Saint Girons 1997), offer an opportunity
for multidisciplinary studies in evolutionary
biology.
The analysis of the number of ventral and
caudal scales of northern Italian Vipera
aspis shows, within a North-South oriented
sector (from about 8°45 to 9°20'
Longitude East), that V. aspis specimens
display intermediate characters between
those of V. atra and of Italian V. aspis.
This geographic sector extends from the
western and eastern sides of the Ticino
river, from the Maggiore lake to the confluence
with the Po river; from the Ticino-
Po confluence southwards to the begin-
Figure 1. Present distribution of European Vipera aspis subspecies.
40° lat N
frandsciredi
500 Km

ZUFFI, GENTILLI, RAZZETTI & SCALI Biota 3/1-2,2002 193
Figure 2. Contact zone between Vipera atra and Italian V. aspis. Circles = atra; Boxes =
Italian aspis; Crosses = intermediate individuals
ning of the Apennines; and along a relatively
large area that connects, along a
virtual line, the towns of Voghera,
Tortona, Novi Ligure, and Genoa.
This contribution presents the first available
morphological data on V. aspis populations
of two distinct species, i.e. Vipera
atra and Italian Vipera aspis from northwestern
Italy, along a western-eastern
transect, passing through the border
between the species (Figure 2). We present
data documenting the parapatric
occurrence of two species of V. aspis
along with a limited number of intermediates.
MATERIALS AND METHODS
173 vipers (55 V. atra. 104 V. aspis, 14
intermediate specimens), (Zuffi & Bonnet
1999, Zuffi 2002) were studied for the
following characteristics: snout-vent
length (SVL), tail length (tL), total length
(TL), ventral scales (VS), subcaudal scales
(CS), total scales (TOTS), bars or blotches
or spots on the dorsal left side (BARS).
Because of strong sexual dimorphism,
analyses were considered for each sex
separately. 22 vipers were analysed for
hemipenial morphology as described in
Zuffi (2002). Moreover, 35 individuals
from the Enrica Calabresi herpetological
collection (University of Florence) were
examined for differences in skull morphology.
Calabresi (1924) pointed out
that the shape of parietal and basisphenoid
bones are quite characteristic in
Italian V. aspis subspecies. We used differences
in the parietal and basisphenoid
morphology to evaluate the skulls.
Parietal bone shows a different pattern
for each of the two species: i) in V. atra
the parietal does not show any bony
crest; ii) in V. aspis of Italy lateral bony
crests are typically present, and they are
convergent posteriorly contacting each
other in "V" shaped crests. Basisphenoid
bones show different patterns in each
species: i) in V. atra the proximal portion
of the bone, at the level of the postorbitals,
is a romboid shaped crest that con-

194 Biota 3/1-2,2002 ZUFFI, GENTILLI, RAZZETTI & SCALI
tinues posteriorly as a longitudinal narrow
crest, forming a single process, ending at
the basioccipital suture; in ii) V. aspis
such a ventral pattern is generally absent;
the proximal part of the basisphenoid in
V. aspis possesses a narrow longitudinal
crest that continues, in the distal part of
the basisphenoid bone, as a small romboid
shaped crest. Body size parameters
were analysed to evaluate the variation
along the transect, without regard to taxonomic
status of each specimen. After
this we used additional characteristics to
assign taxonomic status as V. atra or V.
aspis or as intermediate. The original data
set was log-transformed to avoid the
effect of allometry; data were then tested
for normality. We used SPSS 8.0 PC to
analyse our data. All tests were two tailed
and we chose P = 0.05 as our level of significance.
RESULTS
Body size features
There was no significant difference in SVL
or in TL between V. atra and V. aspis
males and females. Vipera atra males differed
in CS and TOTS when compared to
females (males CS: 43.7 ± 3.1, N=23,
males TOTS: 198 ± 5, N = 23; vs. females
CS: 37.6 ± 3.4, N = 28, females TOTS:
192.3 ± 5, N = 27; Student t test = 6.569,
49 df, P < 0.001, and t = 4.001, 48 df, P
< 0.0001). Vipera aspis males differed
significantly from females in VS (146.6 ±
3.4, N = 35, vs. 150.4+ 3.7, N = 66,
Student t test = -4.813, 103 df, P =
0.0001), CS (42 ± 2.5, N = 32 vs. 33.9 ±
2.3, N = 65, Student t test = 15.894, 99
df, P = 0.0001, p), TOTS (188.4 ± 3.9, N
= 31, vs. 184.3 ± 4.1, N = 63, Student t
test = 4.726, 96 df, P = 0.0001), and
BARS (40.7 ± 2.1, N = 24, vs. 43.7 ± 3.8,
N = 36, Student t test = -2.435, 62 df, P
= 0.018) respectively. Vipera aspis males
of the intermediate form differed from
females in CS (44 ± 2.5, N = 5, vs. 36.1 ±
2.1, N = 8, Student t test = 3.8, 15 df, P
= 0.002).
Vipera atra males possessed a significantly
higher number of ventrals, subcaudals,
totals and dorsal bars either of Italian V.
aspis specimens (VS, TOTS, BARS, P =
0.0001; CS, P = 0.031; One-Way
ANOVA, LSD multirange test), or those
of intermediate specimens (VS, TOTS,
BARS, P = 0.0001). On the other hand,
V. aspis of Italy and the individuals of the
intermediate form were close to each
other for VS (P = 0.095); CS (P = 0.786);
TOTS (P = 0.259) and BARS (P = 0.22).
Vipera atra females had a significantly
higher number of ventrals, subcaudals,
totals and dorsal bars than V. aspis specimens
(VS, CS, TOTS.BARS, P = 0.0001),
as well as the number of ventrals, totals
and dorsal bars of the intermediate specimens
(VS, P = 0.004; TOTS, P = 0.002;
BARS, P = 0.016).
Skull morphology
We did not observe any appreciable sexual
differences in the skull morphology of
either the parietal or basisphenoid bones.
In the study area we found that the V. a.
atra pattern of the parietal bone occurs in
87.5% of the aspis classified individuals
(N = 8); the V. aspis pattern occurs in
76% of V. aspis classified individuals (N =
21). The pattern of the intermediate
specimens (N = 6) was similar to V. atra
in 33.3% of the cases, and to V. aspis in
66.6% of the cases.
Hemipenial morphology
Hemipenes of French V. aspis show well
separated lobes, relatively longer than the
basal segment; the basal segment possesses
large basal spines reaching the
bifurcation of the sperm groove and
medium to small sized spines up to last
third to fourth of the lobe with a relatively
large calyculate area (Domergue 1962:
98, Fig. 10, Case 1968: 98, Fig. 3, Joger
et al., 1997, Zuffi 2002). In comparison,
V. atra hemipenes have well separated

ZUFFI, GENTILLI, RAZZETTI & SCALI 3/1-2 195
lobes that are relatively longer than the
basis, with large basal spines and many
small spines covering both asulcated and
sulcated sides up to the apex, with a very
reduced calyculate area. Italiann Vipera
aspis shows strong similarities to French
V. aspis: it has similar basal segment and
spines, medium to small spines and a relatively
well defined calyculate area.
However, its hemipenes are, on average,
larger than those of the nominal and the
other subspecies. The only one intermediate
specimen examined for this characteristic
shows an intermediate hemipenial
morphology, sharing the features of both
V. atra and V. aspis of Italy.
DISCUSSION
Vipera aspis Is highly polymorphic regarding
dorsal patterns of colouration and
markings (Brodmann 1987, Zuffi &
Bonnet 1999). Our investigation indicates
that V. atra and Italian V. aspis are distin-
guishable by the analysis of ventrals, dorsal
bars, morphology of skull bones, and
hemipenial morphology. It is only in the
contact zone that some of them are individuals
showing an intermediate pattern,
more like V. aspis of Italy.
Based on the fact that few individuals are
intermediates in morphology, our results
suggest some restriction in gene flow
between V. atra and Italian V. aspis.
Previous works seem to indicate a strong
genetic separation between these two
forms (Pozio 1980) as well as at a morphological
level (Zuffi 2002) as suggested
by the marked difference of hemipene
morphology between these taxa. Genetic
and inter-fertility studies have not been
done yet but will provide additional evidence
to help determine the systematic
status of these two taxa.
REFERENCES
BONNET, X. & NAULLEAU, G. 1996: Catchability in snakes: consequences for estimates of
breeding frequency. Can. J. Zool. 74: 233-239.
BRODMANN, P. 1987: Die Giftschlangen Europas und die Gattung Vipera in Afrika und
Asien. Kummerly-Frey.
CALABRESI, E. 1924: Ricerche sulle variazioni della Vipera aspis Auct. in Italia. Boll. 1st. Zool.
R. Univ. Roma, 2: 78-127.
DOMERGUE, Ch. A. 1962: Observations sur le penis des Ophidiens. Bull. Soc. Sci. nat phys.
Maroc42:87-105.
GASC, J.-P. 1968: Morphologic des hemipenis chez Vipera ursinii ursinii (Bonaparte) et discussion
biogeograpique sur la repartition des especes du genre Vipera en Europe
occidentale. Bull. Mus. Nat. Hist. Nat. 40: 95-101.
HERRMANN, H.-W. & JOGER, U., 1997: Evolution of viperine snakes. In: Thorpe, R.S.,
Wuster, W. & Malhotra, A. Eds. Venomous Snakes: Ecology, Evolution and Snake
Bite. Symp. zool. Soc. London 70: 43-61.
JOGER, U., LENK, P., BARAN, I., BOHME, W., ZIEGLER, T, HEIDRICH, P. & WINK, M. 1997:
The phylogenetic position of Vipera barani and V. nikolskii within the Vipera
berus complex. In: Bohme, W., Bischoff, W. & Ziegler, T. (eds). Herpetologia
Bonnensis. Bonn (SEH), Germany: 185-194.
LENK, P., KALYABINA, S., WINK, M. & JOGER, U. 2001: Evolutionary relationships among
the true vipers (Reptilia: Viperidae) inferred from Mitochondria! DNA sequences.
Molecular Phylogenetics Evolution 19: 94-104.

196 Biota 3/i-a, ZUFFI, GENTILLI, RAZZETTI & SCALI
LUISELLI, L. & ZUFFI, M.A.L. 2002: Female life-history traits of the aspic viper (Vipera aspis)
and sand viper (V. ammodytes) from the Mediterranean region. In: Schuett
G.W., Hoggren M., Douglas M.E. & Greene H.W. (eds.). Biology of Vipers.
CPG/Biological Sciences Press, Carmel, Indiana: 279-284.
PHISALIX, M. 1968: La livree des viperes de France. Bull. Mus. Nat. Hist. Nat., Paris 4: 661-
676.
POZIO, E. 1980: Contributo alia sistematica di Vipera aspis (L.) mediante analisi elettroforetica
delle proteine contenute nel veleno. Natura. Soc. ital. Sci. nat., Museo civ.
Stor. nat. e Acquario civ., Milano, 71: 28-34.
SAINT GIRONS, H., 1997: Vipera aspis (Linnaeus, 1758). In: Case, J.P., Cabela, A.,
Crnobrnja-lsailovic, J., Dolmen, D., Grossenbacher, K.r Haffner, P., Lescure, J.,
Martens, H., Martinez Rica, J.P., Maurin, H., Oliveira, M.E., Sofianidou, T.S.,
Veith, M. & Zuiderwijk, A. (eds). Atlas of Amphibians and Reptiles in Europe.
Societas Europaea Herpetologica and Museum National d'Histoire Naturelle
(IEGP/SPN), Paris: 386-387.
SHINE, R. 2000: Vertebral number in male and female snakes: the roles of natural, sexual
and fecundity selection. J. Evol. Biol. 13: 49-86.
SEIGEL, R.A., & FITCH, S. 1985: Annual variation in reproduction in snakes in a fluctuating
environment. Journal of Animal Ecology 54: 497-505.
ZUFFI, M.A.L. 2002: A critique of the systematic position of the asp viper subspecies [Vipera
aspis aspis (Linne, 1758), Vipera aspis atra Meisner, 1820, Vipera aspis francisciredi
Laurenti, 1768, Vipera aspis hugyi Schinz, 1833, Vipera aspis zinnikeri
Kramer, 1958]. Amphibia-Reptilia 23: 191-213.
ZUFFI, M.A.L. & BONNET, X. 1999: Italian subspecies of the asp viper, Vipera aspis: patterns
of variability and distribution. Ital. Journal of Zoology 66: 87-95.

NOVE KNJIGE/BOOK REVIEWS
Cabela, A., Grillitsch, H. & Tiedemann
(eds.)2001:
Atlas zur Verbreitung und Okologie der
Amphibien und Reptilien in Osterreich.
880 pages.
Hardback, ISBN 3-85457-586-6
Copies can be ordered by e-mail:
groeger@ubavie.gv.at
Price: 66.9 EURO (plus postage)
This massive book, measuring 21 x 27 cm
and weighing much more than 2 kg,
reflects the efforts of a team of three
main researchers at the Natural History
Museum of Vienna who directed the
fieldwork of more than 400 field workers,
edited the copy of the 12 authors who
wrote particular chapters (18), and synthesized
all the cartographic information.
The most substantial part of the atlas,
pages 164-610, includes accounts of 37
species (21 amphibian and 16 reptile
species), from Salamandra atra (p. 164-
174) to Vipera ursinii (605-610). On
average, about 12 pages is devoted to
each species.
The result of these intensive labours is an
extremely handsome book, illustrated
with beautiful colour photographs (157!)
of each species and their habitats, numerous
graphs and tables, and, of course,
innumerable maps that include not only
details of species distribution in Austria,
but also topography. For each species
extensive ecological and phenological
data is shown in various graphs.
Moreover, vertical distribution and sympatric
species are also given.
An unusual, but very useful, feature is
that this Atlas contains some special
chapters, such as an identification key for
groups of reptiles and amphibians, even
for eggs and larvae of amphibians with,
again, excellent black and white drawings
of amphibian larvae. At the end a special
Biota '31-2, 197
chapter is devoted to strategies for protection
of both groups. There are also
some very interesting chapters about fossils
and archaeological finds, and probably
the most complete list of references
ever seen in 39 pages!
The Atlas is a valuable source of information
for all those who are interested in
amphibians and reptiles. It is not an ordinary
atlas; it is much more! With much
ecological data presented in various
ways, it is also extremely useful as an
ecological handbook. All European herpetologists
should read this book, since it
includes so much information and so
many ideas.
I must point out that the Atlas is a real
craftsman's tool. I regret, however, that
at least a few brief English summaries
were not included.
Milan Vogrin

198 Biota 3/i-a,
EDITOR ACKNOWLEDGEMENTS
I would like to make full acknowledgement of the generosity to the members of
Editorial Board and of the referees in both their time and expertise. I should like to thank
all those who have sent in colour slides for possible reproduction on the covers.
List of referee (alphabetical order) for volume 1, 2 and 3
1. Maurizio Biondi (Coppito, Italy)
2. Luigi Boitani (Rome, Italy)
3. Massimo Capula (Rome, Italy)
4. Dan Cogalniceanu (Bucharest, Romania)
5. Jelka Crnobrnja-lsailovic (Belgrade, FR Yugoslavia)
6. Helmut Faber (Graz, Austria)
7. Simone Fattorini (Rome, Italy)
8. Paolo Galeotti (Pavia, Italy)
9. Justin Gerlach (Cambridge, United Kingdom)
10. Giinter Gollmann (Wien, Austria)
11. Richard Griffiths (Kent, United Kingdom)
12. Kurt Grossenbacher (Bern, Switzerland)
13. Ulrich Joger (Darmstadt, Germany)
14. Janusz Kloskowski (Lublin, Poland)
15. Zoltan Korsos (Budapest, Hungary)
16. Luca Luiselli (Rome, Italy)
17. Jean Christophe de Massary (Paris, France)
18. Valentin Perez- Mellado (Salamanca, Spain)
19. Juha Merila (Helsinki, Finland)
20. Claude Miaud (Le Bourget du Lac, France)
21. David Mifsud (Basel, Switzerland)
22. Deanna H. Olson (Corvallis, Oregon, USA)
23. Zbynek Rocek (Prague, Czech Republic)
24. Luca Salvati (Rome, Italy)
25. Sabastian Salvido (Genova, Italy)
26. Josef Schmidler (Miinchen, Germany)
27. Andrej Sorgo (Race, Slovenia)
28. Piotr Tryjanowski (Poznan, Poland)
29. Reuven Yosef (Eilat, Israel)
30. Marco Zuffi (Pisa, Italy)
Milan Vogrin

INSTRUCTIONS
TO CONTRIBUTORS
Biota publishes papers from all fields of
biology and ecology in their widest
sense.
It is open for authors from all countries.
The language of papers is English or
Slovene (with English summary).
Types of papers
The journal publishes original scientific
papers, short communications, review
articles, book reviews, special issues containing
selected and edited papers dealing
with a specific theme or based on a
conference or workshop.
The submission of a paper obliges the
author(s) not to submit the same paper
elsewhere. Manuscripts are submitted to
reviewers for evaluation of their significance
and soundness. Authors will generally
be notified of acceptance, rejection,
or need for revision within three months.
The referees remain anonymous.
Decisions of the editor are final.
The form of the manuscript
The paper should consist of title, author's
name and address, abstract (not exceed
250 words), key words (no more than six)
main text (i.e. introduction, material and
methods, results, discussion) followed by
acknowledgements, references, tables
and figures. Authors of scientific taxa
should be omitted. Manuscripts should
be double-spaced on one side of the
paper, with wide margins all round. The
text of a manuscript should be without
special style settings and footnotes.
Headings should be on separate line, and
their hierarchy should not exceed three.
Authors are urged to have the manuscript
revised by "native speaker", if necessary.
Tables and figures should be numbered
consecutively according to the text. Each
should be on a separate sheet of paper.
Biota 3/1-2,2002 199
Each caption to a figure or table should
start with the number of the figure/table,
e.g. Figure 1., Table 2. Tables and figures
should be numbered in order of appearance
in the text.
Illustrations must be submitted cameraready
for copying, usually as laser printouts.
All internal structures, letters,
graphic symbols must be well readable (at
least 2 mm high) after size reduction.
References. In the list of references, the
following usage should be followed:
Journal: SURNAME, A.B. Year: Title. Full
Journal name, Volume: pagination.
Book: SURNAME, A.B. & SURNAME, C.
Year: Title. Publisher, Place.
Chapter: SURNAME, A., SURNAME & B.
SURNAME, C. Year: Chapter title. In:
Editors. Book or Proceeding title.
Publisher, Place: pagination.
In the text, citations should give the
author's name and the year of publication,
e.g. Surname (1998) or (Surname
1998) or Surname & Surname (1998).
Where there are three or more authors,
the first authors name plus "et al" must
be given, e.g. Surname et al. (1998,
1999). Where two or more papers abbreviate
to the same citation (i.e. two or
more papers produced by the same
authors in the same year), use "a", "b",
"c", etc. in the order of their first appearance,
e.g. (Surname 1998a, b).
Reprints and proof
Proof will be send to the main author
only once and it should be returned to
the Editor without delay. Corrections
should be limited to typographical errors.
The editors reserve the right to correct
the proofs themselves, using the accepted
version of the typescript, if the
author's corrections are overdue and the
journal would otherwise be delayed.
Proofs should be checked very carefully. It
is the correspondence author's responsi-

NAVODILO AVTORJEM
Biota objavlja prispevke s podrocja
biologije in ekologije v najsirsem pomenu
besede. Prispevki so lahko v angleskem
ali v slovenskem jeziku z daljsim
angleskim povzetkom (enako naslovi
tabel in grafov), ki ga pripravi avtor sam.
Vrste prispevkov
V Bioti je mogoce objavljati izvirne
znanstvene clanke, kratke notice, pregledne
clanke in predstavitve novih knjig.
Biota sprejema tudi prispevke z razlicnih
konferenc in posvetov. Z objavo se avtorji
obvezejo, da ne bodo nikjer objavili
enakega prispevka. Vsi prispevki bodo
predlozeni v recenzijo dvema recenzentoma.
Recenzenti ostanejo anonimni.
Avtorji bodo obvesceni o sprejemu, zavrnitvi
ali reviziji prispevka predvidoma v
treh mesecih. Odlocitev urednika je
dokoncna.
Oblika prispevka
Prispevek mora vsebovati naslov, imena
avtorjev in njihove naslove, izvlecek (do
250 besed), kljucne besede (do sest) ter
glavni tekst (uvod, material in metode,
rezultati, diskusija) ki mu sledi zahvala,
literatura, tabele in grafi. Prispevek predlozite
v dveh izvodih, z dvojnim medvrstnim
razmakom in s sirokimi robovi.
Odstavki naj bodo med seboj loceni s
prazno vrstico. V kolikor je prispevek
pisan v angleskem jeziku, avtorju priporocamo,
da ga pregleda "native
speaker".
Tabele in slike (grafi, fotografije, risbe)
naj bodo ostevilcene po zaporedju, kot se
pojavljajo v besedilu. Vsaka tabela in graf
morata biti na svojem listu. Tabela ali graf
se morata priceti z zaporedno stevilko,
npr. Slika 1, Tabela 2.
llustracije naj bodo tiskane na laserski
tiskalnik. Risbe so lahko narisane tudi s
Biota 3/1-2,2002 201
crnim tusem na paus papirju. Crke,
stevilke in simboli morajo biti velike vsaj 2
mm.
Literatura
Vsi uporabljeni viri morajo biti citirani
med tekstom. Literaturo uredite po
abecednem redu prvega avtorja in glede
na letnico izdaje:
Revija: PRIIMEK, A.B. Leto: Naslov. Polno
ime revije, letnik: strani.
Knjiga: PRIIMEK, A.B. & PRIIMEK, C.
Leto: Naslov. Izdajatelj, kraj.
Poglavje: PRIIMEK, A., PRIIMEK, B. &
PRIIMEK, C. Leto: Naslov poglavja. V:
Urednik(i). Naslov knjige ali zbornika.
Izdajatelj, kraj: strani.
V tekstu citiramo na naslednji nacin:
Priimek (1998) ali (Priimek 1998) ali
Priimek & Priimek (1998). Ce so vec kot
trije avtorji pa: Priimek et al. (1998,
1999). V primeru, ce citiramo vec del
istega avtorja, objavljenih v enem letu,
posamezno delo oznacimo s crkami a, b,
c, itd., npr. (Priimek 1998a, b).
Korektura in separati
Prvi odtis prispevka urednik poslje
glavnemu avtorju v korekturo. Avtor je
dolzan vrniti popravljeno besedilo v
najkrajsem moznem casu. Sirjenje obsega
besedila ob korekturah ni dovoljeno.
Glavni avtor prejme 30 separatov in izvod
revije, kjer je bil objavljen prispevek, brezplacno.
Original in dve kopiji prispevka vkljucno s
tabelami, slikami in grafi posljite na
naslov (v kolikor prispevek posiljate po
elektronski posti, datoteko shranite kot
"obogateno besedilo" - Rich Text Format
(rtf.):
BIOTA
Milan Vogrin, Zg. Hajdina 83c,
SI-2288 Hajdina, Slovenia
Fax: 02 788 30 51
E-mail: milan.vogrin@guest.ames.si

Klub svetnic in svetnikov LDS Zalec
POKRAJINSKI PODJETNISKI FORUM
Ul. Ivanke Uranjek 1, Zalec

Drustvo za proucevanje ptic
in varstvo narave
Society of bird research
and nature protection
Ptujska c. 91
SI - 2327 Race
Slovenia

V P
Drustvo varuhov okoija Radoziv
Environmental society Radoziv
Ul. Ivanjke Uranjek 1
SI-3310 Za/ec
Slovenia

forum
Danube Forum
Hanulova 5/D
844 40 Bratislava
www.de-forum.org

TO ADVERTISE IN THIS JOURNAL
SEND AN E-MAIL TO
milan.vogrin@guest.arnes.si