Content uploaded by Bahadır Akman
Author content
All content in this area was uploaded by Bahadır Akman on Feb 12, 2015
Content may be subject to copyright.
125
A new subspecies of Lyciasalamandra antalyana from Turkey
All articles available online at http://www.salamandra-journal.com
© 2014 Deutsche Gesellscha für Herpetologie und Terrarienkunde e.V. (DGHT), Mannheim, Germany
SALAMANDRA 50(3) 125–132 30 October 2014 ISSN 0036–3375
A new subspecies of Lyciasalamandra antalyana
(Amphibia: Salamandridae) from the Lycian Coast, Turkey
B A O G
1) Zoology Section, Department of Biology, Faculty of Science, Ege University, 35100 Bornova, Izmir, Turkey
2) Hauptstr. 33, 65527 Niedernhausen, Germany
Corresponding author: B A, email: bahadirakm@hotmail.com
Manuscript received: 20 February 2014
Accepted 23 May 2014 by S L
Abstract. A new subspecies of the Lycian salamander Lyciasalamandra antalyana is described from Yağcavillage (Antalya
province) and Burdur province on the Lycian Coast, Turkey. It is distinguished from the nominotypical form by its dorsal
colouration, multivariate morphometrics, and mitochondrial molecular markers.
Key words. Urodela, Lyciasalamandra antalyana gocmeni ssp. n., SrDNA gene, Turkey.
Introduction
Ten species and three subspecies have so far been de-
scribed (G et al. ) in the genus Lyciasalamandra
(V S ). eir occurrence is limited to
a narrow area along the Lycian Coast in southern Turkey
and some Aegean islands. All species and subspecies have
relatively small and mostly allopatric distributions. In spite
of their geographic proximity, species and subspecies dier
signicantly in colouration, and populations may exhibit
substantial dierences colour and pattern even within taxa
(V S ).
During the last two decades, several new species of
Lycia salamandra and a large number of new populations
have been recorded (e.g., V et al. , G et
al. , G A ). Especially in Lyciasala-
mandra antalyana (B B, ), the known
distribution was substantially extended towards the north.
Just a small distance to the north from the type locality at
Hurma village, recorded specimens are characterized by
pronounced dierences in terms of colouration (V et
al. ). Specimens from these newly discovered popu-
lations (e.g., Yağcavillage and ermessos) exhibit a ful-
ly striped yellow pattern, while topotypical animals from
Hurma are mainly yellow in the region of the eyes, nostrils
and parotoids (Fig. ).
A et al. () published a compilation of the
known L. antalyana populations and discussed if an as-
signment to subspecic status of the newly discovered yel-
low-coloured populations might be warranted. Based on
molecular data, multivariate analyses of morphometrics,
and the striking colouration, we here suggest the popula-
tions from Yağca village and farther north in the Burdur
province as a new subspecies.
Material and methods
Field trips took place between February of and No-
vember of . Geographical coordinates for sample loca-
tions as given in Table were computed with a Magellan XL
GPS. Digital colour photos were taken of all live specimens.
Specimens were euthanised by a ethanol injection into
their body cavity and subsequently xed in ethanol.
Additional specimens were collected from the type locality
of the species (Hurma/Antalya), as well as from Gökdere
and Hacısekililer for comparison with the new populations
from Yağca (Antalya) and Burdur. Specimens are deposited
at e Zoology Museum of Harran University, Şanlıurfa,
Turkey (ZMHRU; Tab. ). All currently known localities of
L. antalyana are illustrated in Figure .
We used the morphometric data published by A et
al. () to perform multiple discriminate analyses (DC)
with PAST version . (H et al. ). We prefer
the DC to a principle component analysis (PCA) since DC
is specically designed for testing the power of characters
to distinguish between pre-dened groups (here: poten-
tial subspecies; S R ). We studied males
and females separately to make provision for potential
dierences among sexes as was previously described for
other Lyciasalamandra species (e.g., G et al. ,
G A ). Obvious juveniles could eas-
ily be distinguished from adults by their smaller size and
the lack of the hedonic gland; they were excluded from
126
B A O G
analyses. We rst used the raw measurements of A
et al. (): rostrum–anus length (RA), tail length (TL),
nostril–eye distance (NED), distance between nostrils
(DBN), eye diameter (ED), head length (HL), head width
(HW), parotoid length (PL), parotoid width (PW), fore
limb length (FLL), hind limb length (HLL), and distance
between fore- and hindlimbs (DFHL). In a second analy-
sis, we employed ratios of these measurements calculated
against RA (PERCRA values) to correct for the overall size
of the specimens. We analysed a total of and males of
L. a. antalyana and L. a. gocmeni ssp. n., respectively, and
females of each subspecies.
A segment of base pairs (bp) in length of the Sr-
RNA gene was sequenced (for DNA extraction, primers,
PCR conditions and sequencing details, see V et al.
) for specimens collected at new localities of the
new subspecies and provided by A et al. ():
Kavacık ( specimen), Kızılseki (), Kocaaliler (), Ko-
caaliler () and Kırkgözhan, Yağca () (GenBank acces-
sion number of the single new haplotype: KJ; www.
ncbi.nlm.nih.gov). We aligned them with homologous
sequences from V et al. () of additional Lycia-
salamandra populations as well as Salamandra infraim-
maculata, Mertensiella caucasica, Neurergus crocatus and
Pleuro deles poireti as outgroup taxa (GenBank accession
numbers: EU, EU, EU, EU,
EU, EU, EU, EU, EU,
EU, EU, EU, EU, EU,
EU, and EU) using the clustal W option in
Mega . (T et al. ). We used jModeltest, ver-
sion .. (D et al. ), to select the best t out of
nested substitution models. e HKY+G model (gam-
ma shape parameter α = .) was selected based on the
Bayesian Information Criterion (BIC). However, HKY is
Figure 1. Map showing localities of Lyciasalamandra antalyana antalyana (solid circles) and L. a. gocmeni ssp. n. (asterisks); numbers
are according to Table 1.
127
A new subspecies of Lyciasalamandra antalyana from Turkey
not supported by Mega . and we therefore used the Ta-
mura-Nei substitution model (T N ), since
the HKY model is a special feature of the Tamura-Nei
model that does not distinguish between substitution rates
for the two types of transformations.
According to V et al. (), a neighbour-joining
(NJ) tree of Lyciasalamandra mitochondrial genes pro-
duced equivalent results of maximum likelihood, maximum
parsimony, and Bayesian inference trees. We therefore only
calculated an NJ tree from , bootstrap replicates using
Mega . and applying the selected substitution model.
Results
In terms of colour and pattern, specimens from Yağca
and Burdur populations dier from Hurma populations
(Figs, ). Yağca and Burdur populations have a more dis-
continuous pattern than Hurma populations. Additionally,
comparing the Yağca and Burdur populations, the discon-
tinuous dorsal pattern of Burdur specimens is dominated
by the dark background, while yellow is dominant in the
Yağca population. e dorsal colouration patterns of either
dier clearly from the nominotypical form. Yağca and Bur-
dur specimens have yellow dorsal markings that also occur
on the upper side of the tail. In contrast, specimens from
Hurma are yellow only in the region of the eyes, nostrils
and parotoids (Figs , ). All other taxa of Lyciasalamandra
do not exhibit any such yellow coloration.
Based on our raw data, the multiple discriminate func-
tion allowed to correctly classify both males and females
from Yağca and Burdur versus other populations only at
and , respectively (Fig. ). e probability that males
and females of both taxa belong to the same group was
. and ., respectively. However, when using ra-
tios (Fig. ), the percentages of correctly classied speci-
mens of both sexes were in males and in females
(psame group = . + ., respectively).
Table 1. Geographic and some climatic information on the localities of Lyciasalamandra antalyana antalyana and L. a. gocmeni ssp. n.,
as well as museum numbers of the specimens used. e numbers in brackets correspond to the localities shown in Figure 1. Asterisk:
at the time of collecting.
Museum numbers
(ZMHRU) Locality Altitude
(m a.s.l.)
Latitude
(DMS)
Longitude
(DMS)
Collection date
number of specimens
2012/6 North of Kocaaliler [1] 750 37°20’N 30°42’E 29.03.2012
1 juv.
2012/5 Hamartaşı/Kocaaliler [2] 720 37°20’N 30°42’E 01.04.2012
5 (2♂♂, 3♀♀)
2012/42 Hamartaşı/Kocaaliler [2] 720 37°20’N 30°42’E 10.04.2012
2 (1♂, 1 juv.)
2012/9 Kavacık [3] 653 37°17’N 30°44’E 01.04.2012
2 (1♂, 1♀)
2012/44 Kavacık [3] 653 37°17’N 30°44’E 10.04.2012
1 juv.
2012/8 North of Kızılseki [4] 400 37°16’N 30°45’E 01.04.2012
4 (2♂♂, 1♀, 1 juv.)
2012/7 West of Kızılseki [5] 438 37°15’N 30°44’E 01.04.2012
1♀
2012/43 Ortacamevki/Kızılseki [6] 541 37°15’N 30°44’E 10.04.2012
1 juv.
2012/2 Kırkgözhan, Yağca [7] 348 37°06’N 30°34’E 07.03.2012
10 (2♂♂, 6♀♀, 2 juv.)
2012/3 Kırkgözhan, Yağca [8] 350 37°06’N 30°34’E 12.03.2012
2 (1♂, 1 juv.)
2012/4 Çığlık [9] 313 37°03’N 30°33’E 12.03.2012
3 (1♂, 1♀, 1 juv.)
2011/82 Hurma 99 36°51’N 30°35’E 25.02.2011
12 (1♂, 3♀♀, 8 juv.)
2012/1 Hurma 99 36°51’N 30°35’E 06.03.2012
17 (8♂♂, 8♀♀, 1 juv.)
2013/173 Gökdere 50 36.82925°N 30.55599°E 24.11.2013
1 (1 juv.)
2013/175 Hacısekililer 581 36.80321°N 30.48953°E 26.11.2013
3 (1♂, 1♀, 1 juv.)
128
B A O G
Figure 2. (a) Male of Lyciasalamandra antalyana gocmeni ssp. n. from the type locality, Kırkgözhan, Yağca; (b) male, (c) female, and
(d)juvenile of L. a. gocmeni ssp. n. from Kızılseki.
Figure 3. Male (a), female (b) and juvenile (c) of Lyciasalamandra antalyana antalyana from the type locality, Hurma village; some
variation in colour pattern is shown in (d).
129
A new subspecies of Lyciasalamandra antalyana from Turkey
Of the base pair positions, ve were diagnostic be-
tween the two L .antalyana lineages (= sequence di-
vergence in uncorrected p-distances). Most of the new-
ly sequenced specimens shared the haplotype found by
V et al. () in specimens from Yağca village. Only
one additional haplotype with one mutation was found at
Kırkgözhan, Yağca village. In the NJ tree, the haplotypes
of the new subspecies form a clade of their own with
bootstrap support (Fig. ).
Five diagnostic base positions within the studied -bp
fragment of the SrRNA mitochondrial gene, dierenc-
es surfacing from a multivariate analysis of morphomet-
ric ratios, and a distinctive colouration clearly separate the
suggested new subspecies from the nominotypical form.
Based on these results, we describe the Yağca and Burdur
populations as:
Lyciasalamandra antalyana gocmeni ssp. n.
Holotype and type locality (Figs a, ): Adult male, ZMHRU
/- from Kırkgözhan/Yağca, Antalya Province, Tur-
key, m above sea level (°’.”N, °’.”E).
Leg. March by B. G, B. A, N. İ,
O.G and M. V.
Paratypes: specimens collected from Antalya (Yağca and
Çığlık populations) and Burdur (Kavacık, Kızılseki, and
Kocaaliler populations) were deposited as paratypes. For
locality details and collection numbers see Table .
Diagnosis: Lyciasalamandra antalyana gocmeni ssp. n. dif-
fers from all other Lyciasalamandra species and subspecies
by having yellow dorsal and supracaudal markings in life.
Figure 4. Multiple discriminate analysis of males and females of L. a. antalyana (black) and L. a. gocmeni ssp. n. (grey) based on raw
data; males: Hotelling’s t² = 54.834, F = 1.924; females: Hotelling’s t² = 10.81, F = 0.4504.
Figure 5. Multiple discriminate analysis of males and females of L. a. antalyana (black) and L. a. gocmeni ssp. n. (grey) based on ratios;
males: Hotelling’s t² =58.741, F = 3.0916; females: Hotelling’s t² = 43.601, F = 2.5764.
130
B A O G
In contrast, specimens of L . a. antalyana are yellow only
in the region of the eyes, nostrils and parotoids. No other
taxon of Lyciasalamandra shows any such yellow coloura-
tion.
Description of the holotype: e body shape is equivalent
to other species of Lyciasalamandra. Head at, longer than
broad (HW/HL .). Snout rounded. Parotoids long and
narrow (PW/PL .), the posterior part broader than the
anterior part. Gular fold distinct. e cloacal region shows
a very slight swelling, and the pads on the upper arm are
quite developed. In prole, the tiny thorns on the dorsal
side are visible. e nger-like projection above the base
of the tail measures about . mm tall; it is pointed and
curved forward at its free end.
In life, the colouration of the dorsum including the head
and upper jaws was brownish-red, especially on the paro-
toids that were dotted with black dermal pores on each
side. On each upper eyelid, there was a thin black cross bar.
e interparotoid, interorbital and internasal spaces were
yellow with brown ecks (Fig. a).
e ground colour of the dorsal side of the trunk had
an interrupted light yellow pattern. is pattern extended
backwards into the median region of the trunk. Two broad
brown stripes that were intersected by transverse brown
bars extended along the dorsolateral sides of the trunk.
Yellow was dominant on the dorsum. e legs and the tail
were esh-coloured. e proximal half of each leg had a
Figure 6. NJ tree of the Tamura-Nei distances of Lyciasalamandra antalyana, showing the monophyletic position of L. a. gocmeni;
numbers at branches indicate bootstrap support ≥ 70% for 2,000 replicates.
Figure 7. Type locality of Lyciasalamandra antalyana gocmeni
ssp.n. at Kırkgözhan, Yağca.
131
A new subspecies of Lyciasalamandra antalyana from Turkey
light yellow maculation, and a light brown reticulation was
found on elbows and knees. Tiny ecks were present on
the dorsal side of some nger joints. On the dorsal side of
the tail, which had rows of black dermal pores, there were a
light brown and yellow maculation. e lower parts of the
trunk, tail and legs were esh-coloured. e throat region
had a yellow tinge.
Measurements of the holotype (in mm): Total length
(ToL) .; RA and length of trunk (LT) . and .,
respectively; TL .; NED .; DBN .; ED .; HL
.; HW .; PL .; PW .; FLL .; HLL .;
DFHL .
Paratype variation: Variation observed in some morpho-
metric characters and ratios vis-à-vis the holotype are
summarized separately for adults and juveniles in Table .
Sexual dimorphism was observed within the population
(p ≤ .) regarding HL, PL, PW and FLL in raw data or
PERCRA values. In addition, the projection at the base of
the tail in nine male specimens ranged between . and
. mm with an average of . mm. As far as the colour
pattern is concerned, the description given for the holo-
type largely applies to the other males and females as well
(Fig.). e discontinuous light yellow parts in the pattern
of adult females are broader and more pronounced than
that of adult males. Juveniles sport a darker ground colour
with a more intensely yellow and brown pattern on their
dorsum. From this series and the specimens observed in
the eld, it would appear that the pattern as well as the dor-
sal background colour change with age and depending on
sex. Both females and juveniles lack any protuberance at
their tail bases and have smooth or less swollen cloacae.
Habitat, geographic distribution, and ecology: We found
Lyciasalamandra antalyana gocmeni ssp. n. under rocks in
a karst-dominated landscape. In the dry months of sum-
mer, the animals use hollow spaces in the karst or heaps of
rock for sheltering. e hillside is usually vegetated with
plane (Platanus orientalis) and pine trees (Pinus brutia).
Activity outside of the shelters is highly dependent on tem-
peratures and humidity. All localities recorded lie with-
in the potential distribution predicted by R et al.
(). e habitat of the new subspecies extends into the
Taurus mountain range at the western border of the An-
talya lowlands. In the east, the area is bordered by the Aksu
valley. In the west and north, the forest-free tablelands of
the Taurus do not allow further expansion.
Derivatio nominis: e name of the new subspecies has
been chosen in honour of the Turkish herpetologist, Pro-
fessor Dr. B G. Besides his contributions to
herpetology, he is a role model and a good friend.
Acknowledgements
anks to M V for providing unpublished DNA data.
References
A, B., B. G, N. İ, D. Y (): Range
extension of Lyciasalamandra antalyana. –Biharean Biologist,
: –.
B, M. I. B (): e subspecic status of the
population of Mertensiella luschani (Steindachner) in the An-
talya region of Southwestern Anatolia. – Ege Üniversitesi Fen
Fakültesi İlmi Raporlar Serisi, : –.
D, D., G. L. T, R. D D. P ():
jModel Test . More models, new heuristic and parallel com-
puting. – Nature Methods, : .
Table 2. Some morphometric characters (in mm) and ratios of
Lyciasalamandra antalyana gocmeni ssp. n. 1: Values in raw data;
2: Values in PERCRA; N: number of specimens; SD: Standard
deviation; the other abbreviations of characters are given in Ma-
terial and methods.
Juveniles (N = 10)
Mean±SD (Range)
Adults (N = 22)
Mean±SD (Range)
ToL 1 86.50±9.76 (77.00–108.00) 130.68±15.65 (101.00–153.00)
2 1.79±0.08 (1.72–1.93) 1.79±0.06 (1.66–1.89)
RA 1 48.40±5.52 (42.00–60.00) 72.95±7.67 (60.00–86.00)
LT 1 33.07±3.36 (29.13–40.12) 48.25±5.39 (39.50–58.17)
2 0.69±0.05 (0.61–0.76) 0.66±0.02 (0.61–0.69)
TL 1 38.10±5.02 (33.00–48.00) 57.73±8.49 (40.00–68.00)
2 0.79±0.08 (0.72–0.93) 0.79±0.06 (0.66–0.89)
NED 1 2.27±0.32 (1.67–2.92) 2.92±0.41 (2.22–3.68)
2 0.05±0.01 (0.04–0.06) 0.04±0.00 (0.03–0.05)
DBN 1 3.72±0.32 (3.24–4.26) 5.09±0.54 (4.13–5.92)
2 0.08±0.00 (0.07–0.08) 0.07±0.01 (0.06–0.08)
ED 1 3.28±0.42 (2.84–4.14) 4.27±1.27 (3.26–9.78)
2 0.07±0.01 (0.05–0.09) 0.06±0.02 (0.05–0.16)
HL 1 13.58±0.87 (12.41–15.37) 17.61±1.49 (15.12–20.94)
2 0.28±0.02 (0.26–0.32) 0.24±0.01 (0.22–0.26)
HW 1 8.75±0.74 (7.61–9.89) 11.42±0.88 (9.88–12.84)
2 0.18±0.02 (0.16–0.21) 0.16±0.01 (0.14–0.18)
PL 1 6.37±0.77 (5.28–7.25) 8.59±0.78 (7.31–10.29)
2 0.13±0.02 (0.11–0.16) 0.12±0.01 (0.10–0.14)
PW 1 2.24±0.39 (1.68–2.77) 3.22±0.46 (2.61–4.11)
2 0.05±0.01 (0.04–0.06) 0.04±0.00 (0.04–0.05)
FLL 1 15.94±1.87 (13.82–19.49) 21.72±1.65 (18.20–24.75)
2 0.33±0.02 (0.31–0.36) 0.30±0.02 (0.26–0.34)
HLL 1 18.10±2.75 (15.00–23.47) 26.19±2.16 (21.79–28.92)
2 0.37±0.03 (0.34–0.41) 0.36±0.02 (0.34–0.40)
DFHL 1 26.83±2.86 (22.23–31.52) 39.91±4.84 (31.54–47.10)
2 0.56±0.05 (0.50–0.65) 0.55±0.03 (0.49–0.58)
HW/HL 1 0.64±0.04 (0.56–0.71) 0.65±0.04 (0.59–0.74)
TL/TBL 1 0.44±0.02 (0.42–0.48) 0.44±0.02 (0.40–0.47)
PW/PL 1 0.35±0.03 (0.28–0.38) 0.37±0.04 (0.30–0.44)
NED/HL 1 0.17±0.02 (0.13–0.19) 0.17±0.02 (0.13–0.20)
132
B A O G
G B., H. A D. Y (): A new Lycian
salamander, threatened with extinction, from the Göynük
Canyon (Antalya, Anatolia), Lyciasalamandra irfani n. sp.
(Urodela: Salamandridae). – North-Western Journal of Zoo-
lo gy, : –.
G, B. B. A (): Lyciasalamandra arikani n. sp.
& L. yehudahi n. sp. (Amphibia: Salamandridae), two new Ly-
cian salamanders from Southwestern Anatolia. – North-West-
ern Journal of Zoology, : –.
G, B., M. V, B. A, O. G, N. İ, M.
A. O (): New records of the Turkish Lycian salaman-
ders (Lyciasalamandra, Salamandridae). – North-Western
Journal of Zoology, : –.
H, Ø., D. A. T. H, P. D. R (): PAST: Pale-
ontological statistics soware package for education and data
analysis. – Palaeontologia Electronica, : –.
R, D., S. L, M. Ö, S. B, K. E
M. V (): A novel method to calculate climatic niche
similarity among species with restricted ranges – e case of
the terrestrial Lycian salamanders. – Organisms, Diversity and
Evolution, : –.
S, R. R. F. J. R (): Biometry: the principles and
practice of statistics in biological research. ird edition. –
Freeman, New York.
T, K. M. N (): Estimation of the number of nu-
cleotide substitutions in the control region of mitochondrial-
DNA in humans and chimpanzees. – Molecular Biology and
Evolution, : –.
T, K., G. S, D. P, A. F S. K
(): MEGA: Molecular evolutionary genetics analysis ver-
sion .. – Molecular Biology and Evolution, : –.
V M., S. S, R. Z, A. S A. M
(): A molecular phylogeny of ‘true’ salamanders (family
Salamandridae) and the evolution of terrestriality of repro-
ductive modes. – Journal of Zoological Systematics and Evo-
lutionary Research, : –.
V, M., I. B, O. G, A. K, M. Ö M. R.
T (): A revision of population designation and geo-
graphic distribution of the Lycian salamander Mertensiella
luschani (Steindachner, ). – Zoology in the Middle East,
: –.
V, M. S. S (): When non-monophyly re-
sults in taxonomic consequences – the case of Mertensiella
within the Salamandridae (Amphibia: Urodela). – Salaman-
dra, : –.
V, M., E. L, M. Ö, A. K, I. B, R. M. P-
S. S (): Cracking the nut: geographi-
cal adjacency of sister taxa supports vicariance in a polytomic
sala mander clade in the absence of node support. – Molecular
Phylogenetics and Evolution, : –.
