Systematics and phylogeography of Acanthodactylusschreiberi and its relationships with Acanthodactylusboskianus (Reptilia Squamata Lacertidae)

KARIN TAMAR1 SALVADOR CARRANZA2 ROBERTO SINDACO3 JIRIacute MORAVEC4

and SHAI MEIRI1

1Department of Zoology Tel-Aviv University Tel-Aviv 6997801 Israel2Institute of Evolutionary Biology (CSIC-UPF) Passeig Mariacutetim de la Barceloneta 37-49 E-08003Barcelona Spain3Museo Civico de Storia Naturale via San Francesco di Sales 188 Carmagnola I-10022 Italy4Department of Zoology National Museum Cirkusovaacute 1740 19300 Prague 9 Czech Republic

Received 23 February 2014 revised 24 April 2014 accepted for publication 2 May 2014

Acanthodactylus is a widespread lacertid genus occurring from the Iberian Peninsula and western North Africato western India including the Middle East Cyprus and the Arabian Peninsula The genus is in dire need of ataxonomic revision and the phylogenetic relationships amongst and within its species remain unclear In particu-lar the taxonomy and relationship of the allopatric narrow-ranged Acanthodactylus schreiberi and its close rela-tive the widespread Acanthodactylus boskianus asper are poorly understood We estimated the phylogenetic andphylogeographical structure of A schreiberi across its distribution range and evaluated its relationships to A b asperusing mitochondrial and nuclear data The phylogenetic results indicate that both species are paraphyleticwith A schreiberi nested within A b asper and the subspecies A schreiberi syriacus nested within a distinct lineageof A b asper We suggest that the group is in need of a taxonomic revision because the identified lineages andgenetic diversity are incongruent with the currently recognized taxonomy We tentatively conclude that A schreiberiis restricted to Cyprus and Turkey reduced to a single form and that the populations in Lebanon and Israel belongto A b asper

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014doi 101111zoj12170

ADDITIONAL KEYWORDS convergence ndash east Mediterranean ndash ecotype ndash haplotype network ndash molecularclock ndash mtDNA + nDNA lineages ndash taxonomy

INTRODUCTION

The genus Acanthodactylus Fitzinger 1834 is com-monly known as the fringe-fingered lizards and is thelargest genus in the family Lacertidae with over 40described species (Uetz 2013) Members of this genusare small- to medium-sized diurnal terrestrial andoviparous species that inhabit semi-arid to desertecosystems from the Iberian Peninsula throughNorth Africa to the Middle East and west India in-cluding Cyprus and the Arabian Peninsula (Salvador

1982 Sindaco amp Jeremcenko 2008) Four fundamen-tal studies constructed the systematic knowledge ofAcanthodactylus mainly based on external morphol-ogy osteological characters and the morphology of thehemipenes Boulenger (1918) Salvador (1982) Arnold(1983) and Harris amp Arnold (2000) The latter threestudies divided the genus into species groups a divi-sion that is commonly used today although the as-signment of some species to groups is debated (egAcanthodactylus blanfordii Boulenger 1918 andAcanthodactylus masirae Arnold 1980 Harris amp Arnold2000) The systematics of some species groups is unclearand unstable because of high intraspecific variabilityof some species and morphological convergence of similarCorresponding author E-mail karintmrgmailcom

bs_bs_banner

Zoological Journal of the Linnean Society 2014 With 3 figures

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014 1

species (eg the description of Acanthodactylusmechriguensis Nouira amp Blanc 1999 Fonseca et al2008) Even though it is fairly easy to assign speciesto species groups the boundaries between species andrelationships within species groups are often unclearand unresolved (Salvador 1982 Arnold 1983 Harrisamp Arnold 2000 Crochet Geniez amp Ineich 2003 HarrisBatista amp Carretero 2004 Fonseca et al 2008 2009)Thus the most problematic and interesting issues inAcanthodactylus systematics are the relations amongstand within species groups the taxonomy of the genusand its biogeography

The Acanthodactylus boskianus species group is astriking case of taxonomic uncertainty Although it isa small group of only three species its geographicalrange is the largest in the genus (Salvador1982 Sindaco amp Jeremcenko 2008) It includesAcanthodactylus boskianus (Daudin 1802)Acanthodactylus schreiberi Boulenger 1878 (Salvador1982 Arnold 1983) and Acanthodactylus nilsoniRastegar-Pouyani 1998 Acanthodactylus nilsoni isknown only from western Iran (Anderson 1999)Acanthodactylus boskianus is the most widespreadspecies of its genus (sim8 000 000 km2 S Meiri unpubldata) ranging through North Africa and the Sahel thewhole Arabian Peninsula eastwards to Iran and north-wards to Turkey (Salvador 1982 Schleich Kaumlstle ampKabisch 1996 Rastegar-Pouyani 1999 Sindaco et al2000 Sindaco amp Jeremcenko 2008) Acanthodactylusboskianus has been divided into five subspeciesA boskianus boskianus (Daudin 1802) from the Niledelta and parts of Sinai A boskianus asper (Audouin1827) from much of the distribution range of the speciesA boskianus euphraticus Boulenger 1919 from IraqA boskianus khattensis Trape amp Trape 2012 from Mau-ritania and A boskianus nigeriensis Trape Chirio ampGeniez 2012 from Niger

Acanthodactylus schreiberi was described from Cypruswhere it is the only representative of Acanthodactylusand it also inhabits south-western Asia This specieshas been divided into three allopatric subspecies Thenominate subspecies A schreiberi schreiberi Boulenger1878 is endemic to Cyprus Acanthodactylus schreiberisyriacus Boumlttger 1879 inhabits isolated patches of theMediterranean coastal areas of Israel and southernLebanon (although its terra typical is givenas lsquoSyriarsquo it does not occur in modern Syria Inthe late 19th century lsquoSyriarsquo included modern-daySyria Lebanon and parts of modern-day Israel)Acanthodactylus schreiberi ataturi Yalccedilinkaya amp Goumlccedilmen2012 is known from a single coastal locality in south-ern Turkey This population was originally referred toA s schreiberi by Franzen (1998) because of the mor-phological similarity to the Cypriot form and it waslater described as a new subspecies by Yalccedilinkaya ampGoumlccedilmen (2012)

The huge geographical range of A boskianus in-cludes areas with very different climates (from sub-Mediterranean climate on the sea coasts of North Africato the hyperarid climate of Central Sahara) This widerange leads to adaptations to different environmentswith great geographical variation (Boulenger 1921Salvador 1982 Arnold 1983 Pincheira-Donoso amp Meiri2013) and consequent taxonomic confusion This problemis well known (Salvador 1982 Arnold 1983 Baha ElDin 2006) and has great effect when examining closelyrelated species in an attempt to assess their system-atic status Arnold (1983) suggested that A boskianusand A schreiberi might be sister species as they sharea relatively high number of primitive features He alsosuggested that A schreiberi may have originated asan isolate of A boskianus Previous morphological studieson the A boskianus species group indicated that therelationship between A boskianus and its sister taxonA schreiberi is far from resolved (Salvador 1982 Arnold1983) The most obvious morphological differencesbetween the Cypriot A schreiberi schreiberi and thecontinental A schreiberi syriacus are the size and degreeof keeling of the dorsal and temporal scales (Boulenger1918 1921 Salvador 1982 Arnold 1983 Franzen1998) Boulenger (1921) decided to unite A schreiberiand A syriacus until then considered different speciesas this difference is not greater than those found invariants of other species By contrast Franzen (1998)implied that those intraspecific differences indicate spe-cific distinctiveness In addition the great intraspecificmorphological variation of A boskianus means that thesecharacters fail to firmly distinguish it fromA sc syriacus Salvador (1982) presented the geo-graphical variation of A boskianus admitting that thedifferences between it and A schreiberi are unre-solved and unsatisfactory

The systematics of many lacertid lizards have re-cently been re-evaluated using molecular data (eg ArnoldArribas amp Carranza 2007 Kapli et al 2008 Greenbaumet al 2011 Ahmadzadeh et al 2012 2013) The onlymolecular phylogenetic study on the entireAcanthodactylus genus however was published by Harrisamp Arnold (2000) who suggested that the genus origi-nated in south-west Asia and later dispersed west-wards into Africa This study also indicates thatA boskianus may be paraphyletic as samples fromArabiaand Morocco formed successive basal branches (Harrisamp Arnold 2000) Four additional molecular studies onAcanthodactylus were conducted focusing onAcanthodactylus erythrurus and Acanthodactylus pardalisspecies groups in an attempt to understand the within-group systematics and relationships (Harris et al 2004Fonseca et al 2008 2009 Carretero et al 2011) Todate the only molecular study with samples of theA boskianus species group was conducted by Poulakakiset al (2013) They concluded that A s schreiberi is a

2 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

relatively recent colonist in Cyprus arriving from themainland through transmarine dispersal around085 Mya In that study based solely on 16S rRNA dataand including a single sample of A s syriacus theyfound that the examined individual branched withinthe specimens of A boskianus asper In another studyby Trape Trape amp Chirio (2012) also based solely on16S rRNA data one sample of A schreiberi formed apolytomy with the A boskianus samples These mo-lecular results present an additional dimension to thealready enigmatic taxonomic relationships between thepopulations of A schreiberi and A boskianus

The present taxonomic status of A schreiberi is there-fore unresolved as the differentiation amongst its sub-species is debated (Boulenger 1921 Franzen 1998)and the relationship with its closest relativeA boskianus should be revised

In order to clarify the systematics and to reveal thephylogenetic relationships between A schreiberi andA boskianus in the eastern Mediterranean and to de-termine the role of geological barriers in the evolu-tionary history of these two species fragments of twomitochondrial genes [12S rRNA (12S) cytochrome b(Cytb)] and three nuclear genes [melano-cortin 1 re-ceptor (MC1R) acetylcholinergic receptor Muscarinic4 (ACM4) oocyte maturation factor MOS (c-mos)] weresequenced and analysed for genetic variation We aimedto examine the genetic relationships betweenA schreiberi and the geographically close taxon thewidespread A b asper with emphasis on the rela-tions amongst the A schreiberi subspecies

MATERIAL AND METHODSDNA EXTRACTION AMPLIFICATION AND

SEQUENCE ANALYSIS

Samples of the three known subspecies of A schreiberifrom Cyprus Turkey Lebanon and Israel and samplesof A b asper from North Africa the Middle East andArabia were included in this study (Fig 1) The local-ities specimen codes and GenBank accession numbersare listed in Table 1 The genus Acanthodactylus isdivided into three clades (Harris amp Arnold 2000 PyronBurbrink amp Wiens 2013 K Tamar S Carranza RSindaco J Moravec JF Trape amp S Meriri unpubldata) hence representatives of five species from thesame clade as the A boskianus species group were usedas the closest outgroups (ie Acanthodactylus blanfordiiAcanthodactylus cantoris Acanthodactylus felicisAcanthodactylus masirae and Acanthodactylusopheodurus) In addition we used samples ofAcanthodactylus scutellatus from another clade as thedistant outgroup and used it to root the tree

Genomic DNA was isolated from ethanol-preservedtissue samples using the DNeasy Blood amp Tissue Kit

(Qiagen Valencia CA USA) All individuals were se-quenced for two mitochondrial gene fragments 12S andCytb and three nuclear gene fragments MC1R ACM4and c-mos Gene fragments were amplified and se-quenced for both strands using published primers Theprimers references and PCR conditions are listed inTable S1

Chromatographs were checked manually assem-bled and edited using GENEIOUS 536 (Biomatter Ltd)For the nuclear genes MC1R ACM4 and c-mosheterozygous individuals were identified and coded ac-cording to the International Union of Pure and AppliedChemistry (IUPAC) ambiguity codes Coding gene frag-ments (Cytb c-mos ACM4 and MC1R) were trans-lated into amino acids No stop codons were observedsuggesting that the sequences were all functional DNAsequences were aligned for each gene independentlyusing the online version of MAFFT v 6 (Katoh amp Toh2008) with default parameters In order to removeregions without specific conservation and poorly alignedpositions of the 12S rRNA we used G-blocks (Castresana2000) with low stringency options (Talavera ampCastresana 2007) Inter- and intraspecific uncor-rected p-distances and the number of variable and par-simony informative sites were calculated in MEGA v5 (Tamura et al 2011)

PHYLOGENETIC ANALYSES AND HYPOTHESIS TESTING

Phylogenetic analyses were performed for the com-plete data set simultaneously both with partitions basedon genes and partitions specified using PartitionFinderv 110 (Lanfear et al 2012) PartitionFinder was per-formed with the following parameters linked branchlength all models Bayesian information criterion (BIC)model selection all schemes search data blocks of thecomplete 12S and by codons for the other protein-coding genes (Cytb MC1R ACM4 c-mos) JModelTestv 011 (Posada 2008) was used to select the most ap-propriate model of sequence evolution under the Akaikeinformation criterion (Akaike 1973) for each parti-tion A summary of DNA partitions and relevant modelsis listed in Table 2

Phylogenetic analyses were performed using maximumlikelihood (ML) and Bayesian inference (BI) methodsML analyses were performed with RAxML v 742(Stamatakis 2006) using RAxMLGUI v 13 (Silvestroamp Michalak 2012) with a general time-reversible + Gamma distribution (GTR + G) model ofevolution parameters estimated independently for eachpartition and 100 addition replicates Reliability of theML tree was assessed by bootstrap analysis (Felsenstein1985) including 1000 replications Bayesian analyseswere performed with MrBayes v 312 (Huelsenbeck ampRonquist 2001 Ronquist amp Huelsenbeck 2003) withthe best-fitting models applied to each partitionand all

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 3

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Figure 1 Sampling localities of the Acanthodactylus schreiberi and Acanthodactylus boskianus specimens used in thisstudy with the global distribution range of the species (data modified from Sindaco amp Jeremcenko 2008 IUCN httpwwwiucnredlistorg) Locality codes and colours correlate to specimens in Table 1 and in Figures 2 and 3 (Colour versionof figure available online)

4 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

relatively recent colonist in Cyprus arriving from themainland through transmarine dispersal around085 Mya In that study based solely on 16S rRNA dataand including a single sample of A s syriacus theyfound that the examined individual branched withinthe specimens of A boskianus asper In another studyby Trape Trape amp Chirio (2012) also based solely on16S rRNA data one sample of A schreiberi formed apolytomy with the A boskianus samples These mo-lecular results present an additional dimension to thealready enigmatic taxonomic relationships between thepopulations of A schreiberi and A boskianus

The present taxonomic status of A schreiberi is there-fore unresolved as the differentiation amongst its sub-species is debated (Boulenger 1921 Franzen 1998)and the relationship with its closest relativeA boskianus should be revised

In order to clarify the systematics and to reveal thephylogenetic relationships between A schreiberi andA boskianus in the eastern Mediterranean and to de-termine the role of geological barriers in the evolu-tionary history of these two species fragments of twomitochondrial genes [12S rRNA (12S) cytochrome b(Cytb)] and three nuclear genes [melano-cortin 1 re-ceptor (MC1R) acetylcholinergic receptor Muscarinic4 (ACM4) oocyte maturation factor MOS (c-mos)] weresequenced and analysed for genetic variation We aimedto examine the genetic relationships betweenA schreiberi and the geographically close taxon thewidespread A b asper with emphasis on the rela-tions amongst the A schreiberi subspecies

MATERIAL AND METHODSDNA EXTRACTION AMPLIFICATION AND

SEQUENCE ANALYSIS

Samples of the three known subspecies of A schreiberifrom Cyprus Turkey Lebanon and Israel and samplesof A b asper from North Africa the Middle East andArabia were included in this study (Fig 1) The local-ities specimen codes and GenBank accession numbersare listed in Table 1 The genus Acanthodactylus isdivided into three clades (Harris amp Arnold 2000 PyronBurbrink amp Wiens 2013 K Tamar S Carranza RSindaco J Moravec JF Trape amp S Meriri unpubldata) hence representatives of five species from thesame clade as the A boskianus species group were usedas the closest outgroups (ie Acanthodactylus blanfordiiAcanthodactylus cantoris Acanthodactylus felicisAcanthodactylus masirae and Acanthodactylusopheodurus) In addition we used samples ofAcanthodactylus scutellatus from another clade as thedistant outgroup and used it to root the tree

Genomic DNA was isolated from ethanol-preservedtissue samples using the DNeasy Blood amp Tissue Kit

(Qiagen Valencia CA USA) All individuals were se-quenced for two mitochondrial gene fragments 12S andCytb and three nuclear gene fragments MC1R ACM4and c-mos Gene fragments were amplified and se-quenced for both strands using published primers Theprimers references and PCR conditions are listed inTable S1

Chromatographs were checked manually assem-bled and edited using GENEIOUS 536 (Biomatter Ltd)For the nuclear genes MC1R ACM4 and c-mosheterozygous individuals were identified and coded ac-cording to the International Union of Pure and AppliedChemistry (IUPAC) ambiguity codes Coding gene frag-ments (Cytb c-mos ACM4 and MC1R) were trans-lated into amino acids No stop codons were observedsuggesting that the sequences were all functional DNAsequences were aligned for each gene independentlyusing the online version of MAFFT v 6 (Katoh amp Toh2008) with default parameters In order to removeregions without specific conservation and poorly alignedpositions of the 12S rRNA we used G-blocks (Castresana2000) with low stringency options (Talavera ampCastresana 2007) Inter- and intraspecific uncor-rected p-distances and the number of variable and par-simony informative sites were calculated in MEGA v5 (Tamura et al 2011)

PHYLOGENETIC ANALYSES AND HYPOTHESIS TESTING

Phylogenetic analyses were performed for the com-plete data set simultaneously both with partitions basedon genes and partitions specified using PartitionFinderv 110 (Lanfear et al 2012) PartitionFinder was per-formed with the following parameters linked branchlength all models Bayesian information criterion (BIC)model selection all schemes search data blocks of thecomplete 12S and by codons for the other protein-coding genes (Cytb MC1R ACM4 c-mos) JModelTestv 011 (Posada 2008) was used to select the most ap-propriate model of sequence evolution under the Akaikeinformation criterion (Akaike 1973) for each parti-tion A summary of DNA partitions and relevant modelsis listed in Table 2

Phylogenetic analyses were performed using maximumlikelihood (ML) and Bayesian inference (BI) methodsML analyses were performed with RAxML v 742(Stamatakis 2006) using RAxMLGUI v 13 (Silvestroamp Michalak 2012) with a general time-reversible + Gamma distribution (GTR + G) model ofevolution parameters estimated independently for eachpartition and 100 addition replicates Reliability of theML tree was assessed by bootstrap analysis (Felsenstein1985) including 1000 replications Bayesian analyseswere performed with MrBayes v 312 (Huelsenbeck ampRonquist 2001 Ronquist amp Huelsenbeck 2003) withthe best-fitting models applied to each partitionand all

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 3

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Figure 1 Sampling localities of the Acanthodactylus schreiberi and Acanthodactylus boskianus specimens used in thisstudy with the global distribution range of the species (data modified from Sindaco amp Jeremcenko 2008 IUCN httpwwwiucnredlistorg) Locality codes and colours correlate to specimens in Table 1 and in Figures 2 and 3 (Colour versionof figure available online)

4 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Figure 1 Sampling localities of the Acanthodactylus schreiberi and Acanthodactylus boskianus specimens used in thisstudy with the global distribution range of the species (data modified from Sindaco amp Jeremcenko 2008 IUCN httpwwwiucnredlistorg) Locality codes and colours correlate to specimens in Table 1 and in Figures 2 and 3 (Colour versionof figure available online)

4 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Tab

le1

Info

rmat

ion

onth

esp

ecim

ens

use

dan

dre

late

dG

enB

ank

acce

ssio

nn

um

bers

C

odes

corr

espo

nd

tolo

cali

ties

pres

ente

din

Fig

ure

1

Cod

eS

peci

esV

ouch

erC

oun

try

Loc

alit

y12

SC

ytb

MC

1RA

CM

4c-

mos

Ab1

09dagger

Aca

nth

odac

tylu

sbo

skia

nu

sM

CC

I-R

471

Alg

eria

Tass

ili

lsquonrsquoA

jjer

KJ5

6769

4K

J567

812

KJ5

4803

7K

J547

885

KJ5

4798

7A

b171

Aca

nth

odac

tylu

sbo

skia

nu

sE

gypt

ElA

rish

S

inai

KJ5

6767

6K

J567

776

KJ5

4804

4K

J547

854

KJ5

4800

3A

b167

daggerA

can

thod

acty

lus

bosk

ian

us

Egy

ptB

alu

za

Sin

aiK

J567

727

KJ5

6779

0K

J548

063

KJ5

4785

2K

J548

014

Ab1

05

Aca

nth

odac

tylu

sbo

skia

nu

sM

CC

I-R

1566

Egy

ptB

etw

een

Ser

abit

elK

had

iman

dG

ebel

Raq

aba

Sin

aiK

J567

672

KJ5

6776

8K

J548

035

KJ5

4784

9K

J547

966

Ab1

90

Aca

nth

odac

tylu

sbo

skia

nu

sE

gypt

14km

SW

ofN

uw

eiba

aS

inai

KJ5

6769

3K

J567

773

KJ5

4805

7K

J547

856

KJ5

4795

1A

b112

Aca

nth

odac

tylu

sbo

skia

nu

sM

CC

I-R

1568

Egy

ptG

ebel

Gu

nn

aS

inai

KJ5

6769

0K

J567

771

KJ5

4803

8K

J547

851

KJ5

4795

0A

b170

A

can

thod

acty

lus

bosk

ian

us

Egy

ptS

tC

ath

erin

eS

inai

KJ5

6775

1K

J567

772

KJ5

4804

3K

J547

853

KJ5

4796

4A

b111

A

can

thod

acty

lus

bosk

ian

us

MC

CI-

R15

67E

gypt

Cro

ssro

adS

tC

ath

erin

eto

Fox

cam

pS

inai

KJ5

6768

9K

J567

770

KJ5

4805

6K

J547

886

KJ5

4798

2

Ab1

77

Aca

nth

odac

tylu

sbo

skia

nu

sE

gypt

Sh

arm

elS

hei

kh

Sin

aindash

KJ5

6777

4K

J548

047

KJ5

4785

5K

J547

965

Ab1

68dagger

Aca

nth

odac

tylu

sbo

skia

nu

sE

gypt

Mat

ruh

KJ5

6772

8K

J567

824

KJ5

4804

2K

J547

910

ndashA

b169

daggerA

can

thod

acty

lus

bosk

ian

us

Egy

ptS

idi

Bra

ni

KJ5

6772

9K

J567

813

KJ5

4806

8K

J547

872

KJ5

4801

5A

b172

daggerA

can

thod

acty

lus

bosk

ian

us

Egy

ptW

adi

El

Nat

run

KJ5

6769

9K

J567

822

KJ5

4807

9K

J547

887

KJ5

4798

6A

b120

A

can

thod

acty

lus

bosk

ian

us

Egy

pt60

kmE

ofId

fuK

J567

698

ndashK

J548

039

KJ5

4787

0K

J547

969

Ab2

79A

can

thod

acty

lus

bosk

ian

us

TA

U-R

160

58Is

rael

Wad

iR

eviv

imK

J567

673

KJ5

6776

7K

J548

051

KJ5

4791

1K

J547

952

Ab6

6A

can

thod

acty

lus

bosk

ian

us

TA

U-R

161

60Is

rael

Sh

ivta

jun

ctio

nK

J567

671

KJ5

6776

5K

J548

033

KJ5

4787

9K

J547

949

Ab2

82A

can

thod

acty

lus

bosk

ian

us

TA

U-R

162

95Is

rael

Km

ehin

KJ5

6768

2K

J567

780

KJ5

4805

3K

J547

932

KJ5

4795

4A

b64

daggerA

can

thod

acty

lus

bosk

ian

us

Isra

elR

otem

plai

nK

J567

670

KJ5

6776

4K

J548

078

KJ5

4787

5K

J547

945

Ab2

03

Aca

nth

odac

tylu

sbo

skia

nu

sH

UJ-

R-2

4055

Isra

elS

ofW

adi

Zafi

tK

J567

691

KJ5

6776

6ndash

ndashndash

Ab7

6A

can

thod

acty

lus

bosk

ian

us

TA

U-R

162

74Is

rael

Mt

Tzi

nK

J567

674

KJ5

6777

5K

J548

034

KJ5

4788

0K

J547

946

Ab7

3A

can

thod

acty

lus

bosk

ian

us

TA

U-R

160

13Is

rael

Mit

zpe

Ram

onK

J567

686

KJ5

6778

3K

J548

058

KJ5

4786

4K

J547

962

Ab7

8A

can

thod

acty

lus

bosk

ian

us

TA

U-R

160

01Is

rael

Mit

zpe

Ram

onK

J567

692

KJ5

6778

4K

J548

061

KJ5

4787

4K

J547

963

Ab8

0A

can

thod

acty

lus

bosk

ian

us

TA

U-R

160

02Is

rael

Mit

zpe

Ram

onK

J567

687

KJ5

6778

5K

J548

060

KJ5

4786

5K

J547

957

Ab2

81A

can

thod

acty

lus

bosk

ian

us

TA

U-R

162

72Is

rael

Wad

iN

ekar

otK

J567

681

KJ5

6777

9K

J548

052

KJ5

4789

2K

J547

953

Ab2

06

Aca

nth

odac

tylu

sbo

skia

nu

sH

UJ-

R-1

9646

Isra

elP

aran

KJ5

6767

7K

J567

787

ndashndash

ndashA

b12

Aca

nth

odac

tylu

sbo

skia

nu

sIs

rael

Wad

iP

aran

KJ5

6768

3K

J567

781

KJ5

4805

9K

J547

861

KJ5

4795

5A

b18

Aca

nth

odac

tylu

sbo

skia

nu

sIs

rael

Wad

iP

aran

KJ5

6768

4K

J567

782

KJ5

4806

4K

J547

884

KJ5

4795

6A

b28

Aca

nth

odac

tylu

sbo

skia

nu

sIs

rael

Wad

iP

aran

KJ5

6768

5K

J567

789

KJ5

4802

8K

J547

862

KJ5

4796

0A

b191

daggerA

can

thod

acty

lus

bosk

ian

us

Jord

anTe

llal

Heb

erK

J567

733

KJ5

6783

0K

J548

067

KJ5

4788

3K

J547

974

Ab2

33A

can

thod

acty

lus

bosk

ian

us

Jord

anP

etra

KJ5

6767

9K

J567

777

KJ5

4804

8K

J547

858

KJ5

4795

8A

b237

Aca

nth

odac

tylu

sbo

skia

nu

sN

MP

6V70

481-

2Jo

rdan

Pet

raK

J567

680

KJ5

6777

8ndash

KJ5

4788

1K

J547

959

Ab2

38

Aca

nth

odac

tylu

sbo

skia

nu

sN

MP

6V70

481-

3Jo

rdan

Pet

raK

J567

688

KJ5

6778

8ndash

KJ5

4790

8K

J547

961

Ab1

13

Aca

nth

odac

tylu

sbo

skia

nu

sM

CC

I-R

618

Jord

anP

etra

KJ5

6767

5K

J567

786

ndashndash

ndashA

b114

daggerA

can

thod

acty

lus

bosk

ian

us

MC

CI-

R62

1Jo

rdan

Wad

iR

amm

KJ5

6773

0K

J567

826

ndashK

J547

876

KJ5

4798

8

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 5

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Tab

le1

Con

tin

ued

Cod

eS

peci

esV

ouch

erC

oun

try

Loc

alit

y12

SC

ytb

MC

1RA

CM

4c-

mos

Ab1

08dagger

Aca

nth

odac

tylu

sbo

skia

nu

sM

CC

I-R

1452

(1)

Lib

yaW

adi

Mat

hke

ndu

shK

J567

736

KJ5

6780

5K

J548

036

KJ5

4785

0K

J547

967

Ab1

73

Aca

nth

odac

tylu

sbo

skia

nu

sM

auri

tan

iaB

etw

een

Zou

erat

and

Bir

Mog

hre

inK

J567

710

KJ5

6780

7K

J548

045

KJ5

4789

4K

J547

983

Ab1

74

Aca

nth

odac

tylu

sbo

skia

nu

sM

auri

tan

iaB

etw

een

Zou

erat

and

Bir

Mog

hre

inK

J567

714

KJ5

6780

8K

J548

030

KJ5

4789

5K

J547

973

Ab1

58dagger

Aca

nth

odac

tylu

sbo

skia

nu

sM

CC

I-R

1088

(4)

Mor

occo

Bet

wee

nS

aidi

aan

dM

oulo

uya

KJ5

6773

5K

J567

825

KJ5

4804

1K

J547

878

KJ5

4798

4A

b160

A

can

thod

acty

lus

bosk

ian

us

NM

P6V

7448

2M

oroc

coB

etw

een

Ait

-Kh

oujm

anan

dK

erra

ndo

uK

J567

716

KJ5

6781

9K

J548

062

ndashK

J548

001

Ab1

61dagger

Aca

nth

odac

tylu

sbo

skia

nu

sN

MP

6V74

483-

1M

oroc

coR

issa

ni

KJ5

6771

7K

J567

818

KJ5

4805

4K

J547

882

KJ5

4798

5A

b147

Aca

nth

odac

tylu

sbo

skia

nu

sN

MP

6V74

483-

2M

oroc

coR

issa

ni

KJ5

6771

5K

J567

817

ndashndash

ndashA

b175

A

can

thod

acty

lus

bosk

ian

us

Mor

occo

Ou

arza

zate

KJ5

6771

8K

J567

820

KJ5

4804

6K

J547

897

KJ5

4800

2A

b234

A

can

thod

acty

lus

bosk

ian

us

Mor

occo

65

kmE

ofO

um

El-

Ale

kK

J567

711

KJ5

6781

0K

J548

049

KJ5

4789

8K

J547

976

Ab2

35dagger

Aca

nth

odac

tylu

sbo

skia

nu

sM

oroc

coA

kka

KJ5

6771

3K

J567

811

KJ5

4806

9K

J547

899

KJ5

4797

7A

b285

daggerA

can

thod

acty

lus

bosk

ian

us

MV

ZH

erp-

2389

25N

iger

Tafo

kin

13

kmN

NE

ofA

gade

zK

J567

701

KJ5

6782

3K

J548

086

KJ5

4786

0K

J548

000

Ab1

15dagger

Aca

nth

odac

tylu

sbo

skia

nu

sO

man

2km

Sof

Liz

qK

J567

731

KJ5

6782

7K

J548

087

KJ5

4791

2K

J547

968

Ab2

31dagger

Aca

nth

odac

tylu

sbo

skia

nu

sO

man

Niz

wa

KJ5

6767

8K

J567

829

KJ5

4808

9K

J547

914

KJ5

4797

5A

b149

A

can

thod

acty

lus

bosk

ian

us

Om

an10

kmS

Eof

Ku

bara

hK

J567

732

KJ5

6782

8K

J548

088

KJ5

4791

3K

J547

971

Ab1

17A

can

thod

acty

lus

bosk

ian

us

Om

an16

kmS

ofD

uqm

KJ5

6770

7K

J567

802

KJ5

4809

1K

J547

905

KJ5

4799

0A

b116

daggerA

can

thod

acty

lus

bosk

ian

us

MC

CI-

R17

73(1

)O

man

Wad

iS

alit

KJ5

6770

6K

J567

801

KJ5

4809

0K

J547

903

KJ5

4798

9A

b159

Aca

nth

odac

tylu

sbo

skia

nu

sM

CC

I-R

1773

(2)

Om

anW

adi

Sal

itK

J567

708

KJ5

6780

3K

J548

092

KJ5

4790

6K

J547

991

Ab2

32

Aca

nth

odac

tylu

sbo

skia

nu

sO

man

4km

Nof

Raw

iyya

hK

J567

709

KJ5

6780

4K

J548

093

KJ5

4790

4K

J547

992

Ab1

48

Aca

nth

odac

tylu

sbo

skia

nu

sS

uda

nN

ofE

l-K

oin

KJ5

6770

0K

J567

821

KJ5

4804

0K

J547

877

KJ5

4797

0A

b256

A

can

thod

acty

lus

bosk

ian

us

NM

P6V

7045

0-2

Syr

iaA

rR

aqqa

hK

J567

747

KJ5

6784

2K

J548

032

KJ5

4791

5K

J548

012

Ab2

57A

can

thod

acty

lus

bosk

ian

us

NM

P6V

7047

0-1

Syr

iaA

rR

aqqa

hK

J567

748

KJ5

6784

1K

J548

080

KJ5

4788

9K

J548

013

Ab2

39dagger

Aca

nth

odac

tylu

sbo

skia

nu

sN

MP

6V72

502-

1S

yria

Qas

ral

Hay

ral

Gh

arbi

KJ5

6774

4K

J567

838

KJ5

4803

1K

J547

890

KJ5

4800

9A

b240

Aca

nth

odac

tylu

sbo

skia

nu

sN

MP

6V72

502-

2S

yria

Qas

ral

Hay

ral

Gh

arbi

KJ5

6774

5K

J567

839

KJ5

4805

5K

J547

888

KJ5

4801

0A

b255

Aca

nth

odac

tylu

sbo

skia

nu

sN

MP

6V70

443

Syr

iaS

adad

KJ5

6774

6K

J567

840

ndashK

J547

891

KJ5

4801

1A

b110

A

can

thod

acty

lus

bosk

ian

us

MC

CI-

R13

26(1

)Tu

nis

iaN

Esl

opes

ofJe

bel

Sem

mam

aK

J567

695

KJ5

6781

4K

J548

085

KJ5

4789

3K

J548

004

Ab1

57A

can

thod

acty

lus

bosk

ian

us

MC

CI-

R13

26(2

)Tu

nis

iaN

Esl

opes

ofJe

bel

Sem

mam

aK

J567

696

KJ5

6781

5ndash

ndashndash

Ab1

76dagger

Aca

nth

odac

tylu

sbo

skia

nu

sTu

nis

iaH

amm

atal

-Jar

idK

J567

697

KJ5

6781

6K

J548

070

KJ5

4790

9K

J548

005

Ab1

54dagger

Aca

nth

odac

tylu

sbo

skia

nu

sM

CC

I-R

1346

(2)

Tun

isia

33km

Sof

Tata

ouin

eK

J567

702

KJ5

6780

6K

J548

081

KJ5

4790

7K

J547

972

Ab2

36dagger

Aca

nth

odac

tylu

sbo

skia

nu

sW

este

rnS

ahar

aL

aayo

un

eK

J567

712

KJ5

6780

9K

J548

050

KJ5

4789

6K

J547

978

Ab1

56

Aca

nth

odac

tylu

sbo

skia

nu

sM

CC

I-R

823(

3)Ye

men

Sarsquo

yun

oasi

sK

J567

705

KJ5

6780

0K

J548

066

KJ5

4790

2K

J547

981

Ab1

06dagger

Aca

nth

odac

tylu

sbo

skia

nu

sM

CC

I-R

823(

4)Ye

men

Sarsquo

yun

oasi

sK

J567

703

KJ5

6779

9K

J548

065

KJ5

4790

0K

J547

980

Ab1

07dagger

Aca

nth

odac

tylu

sbo

skia

nu

sM

CC

I-R

824

Yem

enD

un

esW

ofS

hib

amK

J567

704

KJ5

6776

9ndash

KJ5

4790

1K

J547

979

As1

98

Aca

nth

odac

tylu

ssc

hre

iber

iat

atu

riM

CC

I-R

1693

(1)

Turk

eyB

otas

KJ5

6774

0K

J567

835

KJ5

4807

5K

J547

919

KJ5

4799

6

6 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

As1

99A

can

thod

acty

lus

sch

reib

eri

atat

uri

MC

CI-

R16

93(1

)Tu

rkey

Bot

asK

J567

741

KJ5

6783

6K

J548

076

KJ5

4791

8K

J547

997

As7

0A

can

thod

acty

lus

sch

reib

eri

sch

reib

eri

TA

U-R

16

151

Cyp

rus

Lar

aba

yA

kam

aspe

nin

sula

KJ5

6773

7K

J567

831

KJ5

4807

1K

J547

916

KJ5

4799

3A

s289

Aca

nth

odac

tylu

ssc

hre

iber

isc

hre

iber

iT

AU

-R

1615

0C

ypru

sL

ara

bay

Aka

mas

pen

insu

laK

J567

738

KJ5

6783

2K

J548

074

KJ5

4792

4K

J547

994

As1

18

Aca

nth

odac

tylu

ssc

hre

iber

isc

hre

iber

iN

MP

6V74

532

Cyp

rus

3km

Sof

Mer

sin

lik

Fam

agu

sta

KJ5

6773

9K

J567

834

KJ5

4807

2K

J547

917

KJ5

4799

5A

s229

daggerA

can

thod

acty

lus

sch

reib

eri

sch

reib

eri

Cyp

rus

Epi

skop

iK

J567

743

KJ5

6783

3K

J548

073

KJ5

4792

1K

J547

999

As2

30

Aca

nth

odac

tylu

ssc

hre

iber

isc

hre

iber

iC

ypru

sV

ryso

ull

esK

J567

742

KJ5

6783

7K

J548

077

KJ5

4792

0K

J547

998

As1

5A

can

thod

acty

lus

sch

reib

eri

syri

acu

sH

UJ-

R-2

3653

Isra

elC

aesa

rea

san

dsK

J567

719

KJ5

6779

1K

J548

025

KJ5

4786

6K

J547

937

As2

74A

can

thod

acty

lus

sch

reib

eri

syri

acu

sH

UJ-

R-2

3410

Isra

elH

ader

ato

Bin

yam

ina

KJ5

6772

2K

J567

792

ndashK

J547

859

KJ5

4794

7A

s121

A

can

thod

acty

lus

sch

reib

eri

syri

acu

sT

AU

-R1

6262

Isra

elR

ish

onL

e-Z

ion

san

dsK

J567

723

KJ5

6779

5K

J548

026

KJ5

4786

7K

J547

938

As2

1dagger

Aca

nth

odac

tylu

ssc

hre

iber

isy

riac

us

TA

U-R

163

98Is

rael

Hol

onsa

nds

KJ5

6772

4K

J567

798

KJ5

4802

7K

J547

873

KJ5

4794

2A

s57

Aca

nth

odac

tylu

ssc

hre

iber

isy

riac

us

TA

U-R

164

07Is

rael

Hol

onsa

nds

KJ5

6772

6K

J567

796

KJ5

4808

4K

J547

863

KJ5

4794

3A

s62

Aca

nth

odac

tylu

ssc

hre

iber

isy

riac

us

TA

U-R

164

12Is

rael

Hol

onsa

nds

KJ5

6772

5K

J567

797

KJ5

4802

9K

J547

871

KJ5

4794

4A

s4dagger

Aca

nth

odac

tylu

ssc

hre

iber

isy

riac

us

HU

J-R

-239

86Is

rael

Niz

zan

imre

serv

eK

J567

666

KJ5

6776

0K

J548

024

KJ5

4784

7K

J547

935

As5

A

can

thod

acty

lus

sch

reib

eri

syri

acu

sH

UJ-

R-2

3987

Isra

elN

izza

nim

rese

rve

KJ5

6766

7K

J567

761

KJ5

4808

2K

J547

848

KJ5

4793

6A

s204

Aca

nth

odac

tylu

ssc

hre

iber

isy

riac

us

HU

J-R

-233

21Is

rael

Ash

qelo

nK

J567

668

KJ5

6776

2ndash

KJ5

4785

7K

J547

940

As2

05A

can

thod

acty

lus

sch

reib

eri

syri

acu

sH

UJ-

R-2

3331

Isra

elA

shqe

lon

KJ5

6766

9K

J567

763

KJ5

4808

3ndash

KJ5

4794

1A

s152

A

can

thod

acty

lus

sch

reib

eri

syri

acu

sM

CC

I-R

925(

1)L

eban

onT

yre

KJ5

6772

0K

J567

793

ndashK

J547

868

KJ5

4794

8A

s153

Aca

nth

odac

tylu

ssc

hre

iber

isy

riac

us

MC

CI-

R92

5(2)

Leb

anon

Tyr

eK

J567

721

KJ5

6779

4ndash

KJ5

4786

9K

J547

939

Ab2

07A

can

thod

acty

lus

blan

ford

iiM

VZ

Her

p-23

4464

Iran

6km

NW

ofB

ampu

rS

ista

nva

KJ5

6775

4K

J567

846

KJ5

4810

1K

J547

929

KJ5

4802

0A

b208

Aca

nth

odac

tylu

sbl

anfo

rdii

MV

ZH

erp-

2460

09Ir

anS

and

dun

es7

kmN

ofB

ampu

rK

J567

755

KJ5

6784

7K

J548

102

KJ5

4793

0K

J548

021

Ac2

86A

can

thod

acty

lus

can

tori

sM

VZ

Her

p-24

8443

Pak

ista

n10

kmS

ofU

thal

B

alu

chis

tan

Pro

vin

ceK

J567

756

KJ5

6785

0K

J548

096

KJ5

4792

2K

J548

018

Ac2

87A

can

thod

acty

lus

can

tori

sM

VZ

Her

p-24

8447

Pak

ista

n45

kmN

Wof

Nag

arP

arka

rK

J567

757

KJ5

6785

1K

J548

097

KJ5

4792

3K

J548

019

Af1

97A

can

thod

acty

lus

feli

cis

CA

S22

7596

Om

an23

kmW

ofA

jdar

awt

KJ5

6773

4K

J567

845

KJ5

4810

0K

J547

928

KJ5

4800

8A

m63

Aca

nth

odac

tylu

sm

asir

aeIB

ES

7643

Om

an20

kmE

ofR

asM

adra

kah

KJ5

6775

3K

J567

849

KJ5

4809

5K

J547

925

KJ5

4801

7A

m50

Aca

nth

odac

tylu

sm

asir

aeO

man

Mas

irae

isla

nd

KJ5

6775

2K

J567

848

KJ5

4809

4K

J547

931

KJ5

4801

6A

o25

Aca

nth

odac

tylu

sop

heo

du

rus

HU

J-R

-191

89Is

rael

Tim

na

vall

eyK

J567

750

KJ5

6784

4K

J548

099

KJ5

4792

7K

J548

007

Ao7

9A

can

thod

acty

lus

oph

eod

uru

sM

CC

I-R

627

Jord

anD

iseh

KJ5

6774

9K

J567

843

KJ5

4809

8K

J547

926

KJ5

4800

6A

s11

Aca

nth

odac

tylu

ssc

ute

llat

us

TA

U-R

163

89Is

rael

Bir

Mas

has

hsa

nds

KJ5

6775

8K

J567

852

KJ5

4810

3K

J547

933

KJ5

4802

2A

s44

Aca

nth

odac

tylu

ssc

ute

llat

us

TA

U-R

164

02Is

rael

Hol

onsa

nds

KJ5

6775

9K

J567

853

KJ5

4810

4K

J547

934

KJ5

4802

3

Hap

loty

pes

(N=

59)

daggerRep

rese

nta

tive

su

sed

for

the

dive

rgen

ceti

me

anal

ysis

(N=

25)

Gen

eab

brev

iati

ons

12S

12

SrR

NA

A

CM

4ac

etyl

chol

iner

gic

rece

ptor

Mu

scar

inic

4c-

mos

oo

cyte

mat

ura

tion

fact

orM

OS

C

ytb

cyto

chro

me

bM

C1R

m

elan

o-co

rtin

1re

cept

or

Inst

itu

tion

alab

brev

iati

ons

CA

S

Cal

ifor

nia

Aca

dem

yof

Sci

ence

sU

SA

H

UJ-

R

Zoo

logi

cal

Mu

seu

m

Heb

rew

Un

iver

sity

ofJe

rusa

lem

Is

rael

IB

ES

In

stit

ute

ofE

volu

tion

ary

Bio

logy

B

arce

lon

aS

pain

M

CC

I-R

M

use

oC

ivic

odi

Sto

ria

Nat

ura

le

Car

mag

nol

a(T

orin

o)

Ital

yM

VZ

Her

pM

use

um

ofV

erte

brat

eZ

oolo

gy(U

niv

ersi

tyof

Cal

ifor

nia

B

erke

ley)

U

SA

N

MP

6V

Nat

ion

alM

use

um

(Nat

ura

lH

isto

ry)

Pra

gue

Cze

chR

epu

blic

T

AU

-R

Zoo

logi

cal

mu

seu

m

Tel

Avi

vU

niv

ersi

ty

Isra

el

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 7

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

parameters unlinked across partitions (Table 2) Twoindependent runs of 2 times 107 generations were carriedout with a sampling frequency of every 1000 genera-tions After examining the standard deviation of thesplit frequencies between the two runs and the po-tential scale reduction factor diagnostic burn-in wasperformed discarding the first 25 trees of each runand the remaining trees were combined in a majorityconsensus tree In both ML and BI alignment gaps weretreated as missing data and the nuclear gene se-quences were not phased Nodes were considered strong-ly supported if they received ML bootstrap values ge 70and posterior probability (pp) support values ge 095 (Wilcoxet al 2002 Huelsenbeck amp Rannala 2004)

A total of 59 haplotypes was identified amongst theA boskianus species group using 792 bp of the con-catenated 12S and Cytb data set (see Table 1) Haplotypenetworks were constructed for the three nuclear genesMC1R ACM4 and c-mos (only full-length sequences)SEQPHASE (Flot 2010) was used to convert the inputfiles and the software PHASE v 211 to resolve phasedhaplotypes (Stephens Smith amp Donnelly 2001 Stephensamp Scheet 2005) Default settings of PHASE were usedexcept for phase probabilities which were set as ge 07

All polymorphic sites with a probability of lt 07 werecoded in both alleles with the appropriate IUPAC am-biguity code The phased nuclear sequences were usedto generate median-joining networks using NET-WORKS v 4611 (Bandelt Forster amp Roumlhl 1999)

In order to assess alternative topologies betweenA schreiberi and A b asper topological constraints thatcould be statistically rejected were constructed We en-forced alternative topologies by hand and compared withthe unconstrained tree (best ML tree) using the ap-proximately unbiased (AU Shimodaira 2002) andShimodairaminusHasegawa (SH Shimodaira amp Hasegawa1999) tests Per-site log likelihoods were estimated inusing RAxMLGUI v 13 (Silvestro amp Michalak 2012)and P-values were calculated using CONSEL(Shimodaira amp Hasegawa 2001)

SPECIES DELIMITATION

In order to reveal the main lineages with the concat-enated analysis and as a prior for species groupingsa mitochondrial phylogeny of 59 haplotypes was per-formed with BEAST v 162 (Drummond amp Rambaut2007) without the outgroups Three individual runs were

Table 2 Information on the partitions used in the phylogenetic analyses with the different partition approaches (ie bygene and by PartitionFinder C codon) including the length model of sequence evolution selected by JModelTest andPartitionFinder and the results of the test of rate homogeneity (LRT) run in MEGA (see Material and methods)

Partition approach Partition Length (bp) Model LRT

By gene 12S sim387 GTR + I + G Not rejected (P lt 07396)Cytb 405 TrN + I + G Rejected (P lt 21819E-7)MC1R 663 GTR + I Not rejected (P lt 1)ACM4 429 HKY + I Not rejected (P lt 1)c-mos 522 TPM1uf + G Not rejected (P lt 1)

PartitionFinder ndashConcatenated

12S Cytb (C1) 2406 GTR + I + Gc-mos(C1) Cytb (C2) TrNef + I + GCytb (C3) TrN + I + GACM4 (C12) MC1R (C1) TrNMC1R (C2) F81MC1R (C3) HKY + GACM4 (C3) c-mos(C2 3) TrNef + I + G

PartitionFinder ndashmtDNA

12S Cytb (C1) 792 SYM + I + GCytb (C2) TrN + I + GCytb (C3) TrN + I + G

PartitionFinder ndashnuclear DNA

ACM4 (C12) c-mos (C12)MC1R (C1)

1614 HKY + I

MC1R (C2) F81MC1R (C3) HKY + GACM4 (C3) c-mos (C3) K80 + I

Gene abbreviations 12S 12S rRNA ACM4 acetylcholinergic receptor Muscarinic 4 c-mos oocyte maturation factor MOSCytb cytochrome b MC1R melano-cortin 1 receptorModel abbreviations F81 Felsenstein 1981 GTR general time-reversible HKY Hasegawa Kishino-Yano K80 Kimura1980 SYM symmetrical model TPM1uf Kimura three-parameter model TrN Tamura-Nei Any of these models caninclude invariable sites (+I) gamma distribution (+G) or both (+I+G)

8 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

performed for 5 times 107 generations with a sampling fre-quency of 10 000 The results were combined to inferthe ultrametric tree after discarding 10 of the samplesfrom each run Models and prior specifications appliedwere as follows (otherwise by default) for partitionsby genes and by PartitionFinder For gene partitionsGTR + I + G strict clock (12S) Hasegawa-Kishino-Yano + Invariable sites + Gamma distribution(HKY + I + G) strict clock molecular clock model (es-timate 0ndash1) (Cytb) coalescence constant size processof speciation random starting tree alpha Uniform (010) GTR Uniform For partitions by PartitionFinderGTR + I + G strict clock (partition 1 = 12S + Cytb codon1 and 2) Tamura-Nei + Gamma distribution (TrN + G)strict clock (partition 2 = Cytb codon 3) coalescenceconstant size process of speciation random starting treealpha Uniform (0 10) Parameter values both for clockand substitution models were unlinked across parti-tions For all analyses implemented in BEAST the threeruns were analysed in TRACER v 15 (Rambaut ampDrummond 2007) confirming convergence The treeswere combined in LogCombiner and TreeAnnotator(available in BEAST package) was used for the pro-duction of the final tree

For estimating species limits directly from the Bayes-ian phylogenetic tree produced with the concat-enated mitochondrial data we used the independentgeneralized mixed Yule-coalescent (GMYC) method (Ponset al 2006) The GMYC model estimated the numberof phylogenetic clusters or lsquospeciesrsquo by identifying theshifts between intraspecific (coalescence) and interspecific(diversification) branch rates (Pons et al 2006) Weperformed the GMYC function in the R v302 lsquosplitsrsquopackage (Ezard Fujisawa amp Barraclough 2009) Alikelihood-ratio test was used to determine if the GMYCmodel with a shift in the branching processes provid-ed a better fit to the data than the null model withno shifts We used a single threshold value (Monaghanet al 2009) which has already been applied success-fully to different groups of organisms (Pons et al 2006Fontaneto et al 2007 Monaghan et al 2009)

ESTIMATION OF DIVERGENCE TIMES

The lack of internal calibration points in Acantho-dactylus (no fossils are known) prevents the directestimation of time in our phylogeny Therefore we usedthe mean substitution rates and their standard errorof the same 12S and Cytb mitochondrial regions ex-tracted from a fully calibrated phylogeny of anotherlacertid group the lizards of the genus Gallotia endemicto the Canary Islands (Cox Carranza amp Brown 2010as was implemented in Carranza amp Arnold 2012) Theinferred calibration rate was estimated using the ageof El Hierro Island (Canary Islands) estimated at112 Mya (Guillou et al 1996) They assumed coloni-

zation of the island by members of the lacertid genusGallotia (Gallotia caesaris caesaris endemic to El HierroIsland) immediately after its formation from the neigh-bouring La Gomera Island (inhabited by the endemicGallotia caesaris gomerae) These two subspecies aremonophyletic sister taxa with low intraspecific vari-ability (Maca-Meyer et al 2003 Cox et al 2010) andthus suitable for calibration

For the estimation of divergence times one repre-sentative of each independent GMYC lineage was usedfrom the ultrametric tree (for the representatives seeTable 1) We used a likelihood-ratio test implementedin MEGA 52 (Tamura et al 2011) to test if the dif-ferent partitions (by genes) included in the dating analy-sis were evolving in a clock-like fashion (Table 2) Thisinformation was used to choose between the strict clockand the relaxed uncorrelated lognormal clock priorsimplemented in BEAST (Monaghan et al 2009) Thedata set included one representative from each lineagefrom the GMYC analysis using sequences from all fivepartitions (nuclear genes unphased) Three individ-ual runs were performed for 5 times 107 generations witha sampling frequency of 10 000 and the results werecombined to infer the ultrametric tree after discard-ing 10 of the samples from each run Models and priorspecifications applied were as follows (otherwise bydefault) GTR + I + G relaxed uncorrelated lognor-mal clock molecular clock model (estimate) (12S Cytb)HKY strict clock (MC1R c-mos) and TrN + I strictclock (ACM4) Yule process of speciation random start-ing tree yulebirthRate (0 1000) alpha Uniform (010) ucldmean of 12S Normal (initial value 000553mean 000553 SD 000128) ucldmean of Cytb Normal(initial value 00164 mean 00164 SD 000317) Pa-rameter values both for clock and substitution modelswere unlinked across partitions

RESULTS

The data set of this study is comprised of 19 samplesof A schreiberi 65 samples of A b asper and 11outgroup samples (Table 1 Fig 1) The data set in-cluded mitochondrial DNA (mtDNA) gene fragmentsof 12S (sim387 bp) and Cytb (405 bp) and nuclear DNA(nDNA) gene fragments of MC1R (663 bp) ACM4(429 bp) and c-mos (522 bp) totalling to sim2406 bp Thenumber of variable (V) and parsimony-informative(Pi) sites for the ingroup are listed in Table S1 Thetwo partition approaches (ie by gene and byPartitionFinder) gave similar results for both the MLand BI analyses The results of the phylogenetic analy-ses of the complete concatenated data set using MLand BI methods produced very similar topologies butdiffered to some extent at the less supported nodesat the intraspecific level (Fig 2) Separated analysesof the nuclear data sets are presented in Figure S1

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 9

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

10 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Together A b asper and A schreiberi form amonophyletic group within Acanthodactylus (Fig 2)Within the group however both taxa are paraphyleticwith A schreiberi as a whole nested within A b asperOur analyses distinguish three major clades (1) cladeA formed by A b asper from Syria (2) clade B in-cludes the two subspecies A sc ataturi from Turkeytogether with A sc schreiberi from Cyprus (3) cladeC which includes specimens of A b asper from the re-maining localities in its distribution range together withA sc syriacus from Israel and Lebanon Clade A is verywell supported and includes specimens of A b asperfrom central and northern Syria (Fig 1) splitting fromother specimens at the basal node of the group is es-timated to have occurred c 654 Mya [95 highest pos-terior density (HPD) 392ndash952 Mya] The level of geneticdifferentiation(p-distance) between these specimens and the remain-ing A b asper and all A schreiberi specimens is 37ndash46 for 12S and 107ndash119 for Cytb Clade B is alsovery well supported and includes two of the threenominal subspecies of A schreiberi A s schreiberi thenominotypical subspecies endemic to Cyprus and A sataturi from Turkey The Turkish subspecies is nestedwithin the Cypriot specimens and the two forms havelow genetic distances from each other (12S 016 Cytb123) This clade is nested between the two A b asperclades (clades A and C) in both the concatenatedand the nuclear tree although the nodes are not wellsupported Clade C is not very well supported It in-cludes a cluster of A b asper and A s syriacus Thisclade includes two inner clades that split around558 Mya (95 HPD 356ndash8 Mya) and divided into threepoorly supported geographical inner groups (Fig 2)northern Jordan and northern Oman (group C1) NorthAfrica (group C2) and samples from the Middle East(Egypt south Israel and south Jordan) with samplesfrom Yemen and southern Oman (group C3) ndash the latterincluding all specimens of the subspecies A s syriacusThe diversification within the North African group isestimated to have started around 456 Mya (95 HPD282ndash647 Mya) The IsraelminusLebanon endemic subspe-cies A s syriacus is genetically highly distinct from

A s schreiberi and A s ataturi making A schreiberiparaphyletic (p-distance 12S 431 416 Cytb 1181202 respectively)

The networks constructed for the phased haplotypesof the full length nuclear markers (MC1R ACM4 andc-mos) are presented in Figure 3 The nuclear networkanalyses show similar results for each of the three genesand closely agree with the phylogenetic tree The CypriotA s schreiberi and Turkish A s ataturi subspecies sharealleles for all three genes and both are distinct fromthe third subspecies A s syriacus Acanthodactylusschreiberi syriacus shares no alleles with the other sub-species of A schreiberi but does share alleles withA b asper for each of the genes Acanthodactylusschreiberi syriacus shares MC1R alleles with A b asperspecimens from Tunisia Syria and Israel ACM4 alleleswith Egyptian Israeli Jordanian and North Africanspecimens and c-mos alleles only with Israeli A b asperspecimens Syrian A b asper samples share one allelewith A s syriacus and two with other A b asper speci-mens from Egypt Israel and North Africa in the MC1Rone allele with an Egyptian A b asper in the ACM4and none in the c-mos gene

In order to better understand the relationships betweenA schreiberi and A boskianus we performed three to-pology tests in which we forced monophyletic group-ings (1) monophyly of A schreiberi (all three subspeciestogether) (2) monophyly of A b asper (3) monophylyof A b asper and of A schreiberi The results of thetopological tests indicate that our data set cannot rejectthe alternative hypotheses of monophyly of A schreiberi(AU P = 0091 SH P = 0062) and that of A b asper(AU P = 011 SH P = 0072) if we allow A schreiberito nest within A b asper or a monophyletic A b aspernesting within A schreiberi When forcing monophylyof both A schreiberi and of A b asper together in thesame tree the results are inconclusive (AU P = 0046SH P = 0051)

The single-threshold model in GMYC yield a topol-ogy that is clearly different from the known taxono-my The GMYC results present a total of 25 and 24ML independent lineages from the Bayesian haplotypemitochondrial phylogeny of the two species for the two

Figure 2 Maximum likelihood (ML) tree of the Acanthodactylus boskianus and Acanthodactylus schreiberi specimensinferred using 12S rRNA cytochrome b mtDNA and melano-cortin 1 receptor acetylcholinergic receptor M4 and oocytematuration factor MOS nuclear gene fragments Posterior probability in the Bayesian analysis is indicated by black dotson the nodes [values ge 095 shown for both gene partitions and partitions by PartitionFinder (PF)] and ML bootstrapsupport values are indicated in parentheses (values ge 70 shown ML ML-PF) Age estimates with BEAST are indicat-ed near the relevant nodes and include the mean and in brackets the HPD 95 confidence interval Sample codes relateto specimens in Table 1 and in Figures 1 and 3 Colours blue Acanthodactylus boskianus asper yellow Acanthodactylusschreiberi ataturi red Acanthodactylus schreiberi syriacus green Acanthodactylus schreiberi schreiberi (Colour versionof figure available online)

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 11

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

12 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

partition approaches (ie by gene and by PartitionFinderFigs S2 S3 respectively) The two partition ap-proaches gave similar clusters but at the less sup-ported nodes they differed at the positions of severallineages The single threshold GMYC result is indi-cated for a single line at 00037 Mya for the gene par-titions and at 002 Mya for PartitionFinder (verticallines in Figs S2 S3) The topology and clusters re-vealed in this analysis correspond to the lineages fromthe phylogeny of the ML and BI methods both for theparaphyly of the two species and the geographical group-ings within A b asper The GMYC results mainly differfrom the ML and BI methods in the position ofA schreiberi from Cyprus and Turkey as a sister cladeto the Syrian A b asper

DISCUSSION

We have provided a comprehensive and thorough as-sessment of the intraspecific phylogenetic relation-ships within A schreiberi and its closest relativeA b asper Our results based on mitochondrial andnuclear DNA data from 84 specimens across the entiredistribution range of A schreiberi and most of the dis-tribution range of A b asper reveal that A schreiberiis paraphyletic and nested entirely within theA boskianus subspecies

HISTORICAL BIOGEOGRAPHY

Acanthodactylus schreiberi is thought to comprisethree subspecies corresponding to three allopatricpopulations in Cyprus Turkey and IsraelminusLebanonThe Cypriot endemic nominotypical subspeciesA sc schreiberi and the Turkish subspeciesA sc ataturi cluster together (to form clade B Fig 2)nesting between A b asper clades This lineage is sisterto a clade of A b asper including A sc syriacus (cladeC Fig 2) We estimate the divergence time of theCypriotminusTurkish lineage of A schreiberi to have beenduring the late Miocene around 6 Mya although thereis no support for this split in the tree In other analy-ses using the whole genus this split is well support-ed in Bayesian analyses (K Tamar S Carranza RSindaco J Moravec JF Trape amp S Meriri unpubldata) Based on mitochondrial data Poulakakis et al(2013) found that both the Cypriot and Turkish sub-species are monophyletic and diverged from each other085 Mya (038ndash156 Mya) According to our results thisdate corresponds to an inner divergence of the

A sc schreiberi lineage rather than to the date at whichA sc schreiberi colonized Cyprus

The discrepancy in the phylogenetic relationship ofA sc schreiberi raises questions regarding the arrivalon Cyprus Cyprus originated with the raising of theTroodos Massif during the upper Cretaceous c 91 to88 Mya (Clube amp Robertson 1986 Mukasa amp Ludden1987) During the middle to late Miocene only a smallproportion of Cyprus was exposed above the Mediter-ranean (McCallum amp Robertson 1990 Robertson 1990)Towards the end of the Miocene sim596 Mya with theclosing of the passage between the Atlantic Ocean andthe Mediterranean basin the Messinian salinity crisisbegan (Krijgsman et al 1999) This resulted in thedrying up of much of the Mediterranean Sea and highsea-mounts emerged to form land bridges with thesurrounding land (Hsuuml et al 1977) By the end of theMiocene and early Pliocene sim533 Mya the passagewith the Atlantic Ocean reopened and the Mediterra-nean basin was refilled (Krijgsman et al 1999) Re-sulting from compressions raising and uplifting of thesurrounding areas towards and during the Pleisto-cene Cyprus was a complete emergent island (McCallumamp Robertson 1990) The possible connection of Cyprusto the mainland (ie to TurkeySyria) during theMessinian is debated as are suggestions of a land con-nection at later periods (Steininger amp Roumlgl 1984 Jolivetet al 2006 Bache et al 2012) Such a connection ifit existed could have provided access for terrestrialorganisms with poor overseas dispersal ability suchas lizards to colonize the island Several studies arguethat post-Messinian sea level changes are unlikely tohave formed connections between Cyprus and the main-land (Steininger amp Roumlgl 1984 Jolivet et al 2006) Thusour dating of the split between the Cypriot A schreiberiand A b asper at c 6 Mya leads us to suggest that theancestor of A s schreiberi colonized Cyprus from themainland through a land bridge connection at the be-ginning of the Messinian crisis rather than by a muchlatermore recent transmarine dispersal as suggestedby Poulakakis et al (2013) Owing to its close rela-tions with A b asper the ancestor of A schreiberi waspresumably mainland A boskianus and the cladogenesisleading to A schreiberi thus rendered A b asperparaphyletic

The Turkish subspecies A schreiberi ataturi was rec-orded for the first time by Franzen (1998) at a veryrestricted area of around 15 km of coastal strip (betweenBotas and Yukarı Burnaz Hatay Province) Owing tothe remarkable morphological similarity between

Figure 3 Haplotype networks of the nuclear gene fragments melano-cortin 1 receptor (MC1R) acetylcholinergic recep-tor M4 (ACM4) and oocyte maturation factor MOS (c-mos) with colours corresponding to Figures 1 and 2 Codes corre-late to the two alleles (ie a and b) of specimens in Table 1 Circle sizes are proportional to the number of alleles (Colourversion of figure available online)

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 13

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

A sc ataturi and the Cypriot population the speci-mens were initially identified as A sc schreiberi(Franzen 1998) Yalccedilinkaya amp Goumlccedilmen (2012) howeverdescribed this population as a new distinct subspe-cies A s ataturi presenting several differences betweenthe two in both morphology and blood-serum pro-teins The origin of A s ataturi remains uncertain asit is debated whether the newly discovered Turkishpopulation is a relict or an introduction from CyprusFranzen (1998) described this population as a pos-sible introduction from Cyprus through the harbourof Botas but Sindaco et al (2000) suggested thatit might be a relict of a previously larger popula-tion because its present distribution is similar tothat of some insects and lizards [Archaeolacerta(Phoenicolacerta) laevis and Ablepharus budaki]Yalccedilinkaya amp Goumlccedilmen (2012) proposed that A sc ataturiarrived in Turkey from the nominate population inCyprus during the Messinian crisis The phylogeneticresults haplotype networks and low levels of geneticdivergence we found suggest that the two subspeciesfrom Cyprus and Turkey have not been geneticallyisolated for a long period of time (ie they share allelesin all three nuclear genes and A s ataturi is nestedwithin A s schreiberi in the phylogeny Figs 2 3) Ourresults therefore contrast with the two latter sce-narios of a relict population or a Messinian disper-sal Both divergence time and the genetic similarityof the two subspecies agree with the original sugges-tion of Franzen (1998) that these animals were intro-duced into Turkey from Cyprus Further support forthis hypothesis is that A s ataturi is restricted to thevicinity of the Botas-Adana harbour and is absent inother suitable habitats (coastal sand dunes) wide-spread in south-eastern Turkey Its close morphologi-cal features to A s schreiberi (Franzen 1998) likewisesupport an introduction scenario

The third subspecies A s syriacus is nested withinA b asper in the concatenated mtDNA and nDNA treesand is clearly genetically distinct from the Cyprus andTurkey A schreiberi lineage The close relations ofA s syriacus with A boskianus may shed light on theorigin of the former Acanthodactylus schreiberi syriacusis distributed on stable sands of the coastal plain ofthe eastern Mediterranean in Israel and southernLebanon (Salvador 1982 Hraoui-Bloquet et al 2002Bar amp Haimovitch 2011) habitats resembling those ofA s schreiberi from Cyprus (Baier Sparrow amp Wiedl2009) The oldest divergence of the A b asper cladethat includes A s syriacus is estimated to have oc-curred during the late Miocene around 558 Mya butno further dates are available for the grouping ofA s syriacus as a result of low support values Thecoastal plain of the eastern Mediterranean was sub-merged during the late Miocene and re-emerged onlytoward the Pliocene (Nir 1970 Horowitz 1979) The

sands of the coastal plain where A s syriacus occurs(Salvador 1982 Arnold 1983 K Tamar amp S Meiripers observ) were repeatedly submerged and re-emerged during the Pleistocene sea-level changes (duringinterglacial and glacial periods respectively) A pos-sible scenario for A sc syriacusrsquos origin includes severalwaves of dispersal of Middle Eastern A boskianus whichoccurs on coarse substrates (Amitai amp Bouskila 2001Disi et al 2001 Baha El Din 2006 pers observ) towardthe Mediterranean shore Acanthodactylus boskianusasper is absent from Mediterranean climate habitatsin Lebanon and Israel It occupies only xeric zonessuggesting an invasion to the coastal plain when sandyhabitats allowed desert flora and fauna to migrate north-wards (Yom-Tov 1988) These populations adapted tosandy soils and evolved morphological features thatdistinguish them from the desert hard substrate formsof A b asper We view this as the most likely scenariogiven the biogeography the phylogenetic results andthe habitat preferences and adaptations of these lizardsAn alternative scenario according to which the an-cestor of A schreiberi originated in Cyprus and dis-persed to the shores of Israel and Lebanon (or originatedin the coastal plain of the Eastern Mediterranean anddispersed to Cyprus) we regard as far less likely Sucha scenario requires much closer genetic relationshipsbetween these two forms and is further weakened bythe close relationship between A s syriacus and thegeographically adjacent A b asper populations

Acanthodactylus boskianus asper is highly variableboth morphologically (Salvador 1982 Arnold 1983) andgenetically (this study) The subspecies is paraphyleticas A schreiberi is nested within it The topology of theA b asper tree shows four different geographical group-ings Syria (clade A) north Jordan plus north OmanNorth Africa and Middle East plus south Arabia (groupsC1 C2 C3 respectively) The different groups in thissubspecies are estimated to have first diverged duringthe late Miocene approximately 65 Mya with the splitof the Syrian population The Syrian lineage is ge-netically distant from A schreiberi and the otherA b asper specimens The nuclear networks indicatethat this group is closer to the other A b asper samplesrather than to A schreiberi The geographical splitsin the rest of the A b asper range (clade C) are esti-mated to have started around 558 Mya These groupsare supported as a distinct clade but are closely relatedto each other in both the concatenated and nuclear trees(Figs 2 S1 respectively) The diversification within thisclade is estimated to have occurred during the lateMiocene to early Pliocene when A b asper dispersedwidely west to North Africa and in Arabia The di-vergence within the North African group (group C2)is estimated to have occurred during the Pliocene ap-proximately 456 Mya with the Egyptian Nigerian andSudanese populations later dispersing west and north

14 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

in Africa This diversification correlates to the arid climatestarting in southern Sahara during the early-mid Plio-cene and later in northern Africa between the Plio-cene and the Pleistocene (Le Houeacuterou 1997) as hasbeen suggested for the dispersal of Mesalina guttulatain Africa (Kapli et al 2008) Other evidence relatesdry climate in North Africa to an earlier period around7 Mya (Schuster et al 2006) as has been suggestedfor the genus Chalcides and other reptiles (Carranzaet al 2008 Metallinou et al 2012 and reference therein)The aridification of North Africa has most likely con-tributed for the successful dispersal of A b asper westfrom south-west Asia into Africa Morphological studiesof A boskianus show relatively uniform populations inNorth Africa suggesting recent migration (Salvador1982 Arnold 1983) The other two geographical group-ings of A b asper from the Middle East and Arabia(groups C1 and C3) are located in two distinct innerclades but their location within each inner clade ispoorly supported The topology of the concatenated tree(Fig 2) shows that the group from northern Jordanand northern Oman (group C1) is closer to the NorthAfrican one than to the geographically close Middle-Eastern and south Arabian group (group C3) The taxo-nomic separation between north and south Oman hasbeen recognized in other species of reptiles and sup-ported by the topography of Oman (eg Echis coloratusand Echis omanensis Arnold Robinson amp Carranza2009) In the nuclear tree (Fig S1) these two groupsare closer to one another and with the North Africangroup form clade C Therefore the low support valuesamongst these groupings prevent an appropriate andthorough analysis of this subspecies The close rela-tionship amongst the geographical groups may reflectclose phylogenetic relationships amongst these popu-lations suggesting recent migration divergence andongoing gene flow

SYSTEMATICS AND TAXONOMIC IMPLICATIONS

The relationships within the A boskianus species groupconflict with the current known taxonomy of A schreiberiand A b asper (samples of the other subspecies ofA boskianus and of A nilsoni were unavailable for thisstudy) Both species have been found to be closelyrelated and paraphyletic The constrained topology testsexemplify the close entangled relationship between thetwo species as the separate monophyly of the two specieswas not rejected and the enforced monophyly of themboth together was inconclusive

Several causes can be responsible for paraphyly inspecies such in the A boskianus species group (Funkamp Omland 2003 and references therein) (1) inad-equate phylogenetic information (2) imperfect taxono-my (incorrectinaccurate species limits) derived frommisidentifying intraspecific variation (3) interspecific

gene flow ndash hybridization through interspecific matingand the subsequent backcrossing of hybrids into theparental populations (4) incomplete lineage sortingbecause of recent speciation events (5) unrecognizedparalogy We suggest that the relationships betweenA schreiberi and A b asper based on mitochondrialand nuclear data are most likely explained by incor-rect taxonomy probably because of the great variabil-ity of the latter species and to convergence As wasthe case in the molecular studies of the A pardalisand A erythrurus species groups (Harris et al 2004Fonseca et al 2008 2009 Carretero et al 2011 andreference therein) there are many problems with thecurrent taxonomic status of several species groups withinAcanthodactylus

Taking the molecular results of our study into accountthere are several systematic approaches to classify-ing the A schreiberiminusA b asper clade The Cypriot andTurkish populations of A schreiberi are very closelyrelated with the latter nested in the former and thetwo subspecies share nuclear alleles (Fig 3) Further-more the low uncorrected p-distance is positively cor-related with subspecies-level distances within otherlacertid species (ie 16 of Cytb in Lacerta bilineatachloronota Godinho et al 2005) We therefore con-clude that Cypriot animals were recently introducedto Turkey and that the Turkish population does notmerit a subspecific rank We suggest that A s ataturiYalccedilinkaya amp Goumlccedilmen 2012 is a junior synonym ofA s schreiberi Boulenger 1878

Regarding the relationships between A schreiberi andA b asper a few scenarios are possible One is to sinkA schreiberi within A boskianus to create one species(A boskianus) with high genetic and morphological vari-ability ranging over a broad distribution Another isfor the two taxa be regarded as a species complex (theA boskianus-schreiberi complex) until further inves-tigation on the subject However although A schreiberiis nested within A b asper the populations from Cyprusand Turkey represent a distinct evolutionary lineagewith distinct genetic and morphological features andthus it is logical to retain the specific status Two othersolutions are possible The first is to re-evaluate theSyrian populations and to consider elevating them aswell as the more divergent lineages (and subspecies)of A boskianus to specific status This would neces-sitate an examination of the phylogeny and morphol-ogy of the other four subspecies of A boskianus(A b boskianus A b euphraticus A b khattensis andA b nigeriensis) and the identification of distinctivephenotypic features in the Syrian lizards Another so-lution is to recognize the maintenance of gene flowamongst mainland populations of A b asper after thedivergence of the insular endemic A schreiberi andthus the evolutionary cohesion of the paraphyleticA b asper Arnold (1983) noted that A schreiberi may

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 15

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

have originated as an isolate from A boskianus becauseof their shared morphology and hemipenis featuresOur results support this scenario which includes thedispersal of A schreiberi to Cyprus from a mainlandpopulation that was most probably A boskianus It maybe assumed that the ancestor of the Cypriot A schreiberiafter arriving on Cyprus remained isolated for a longperiod of time and thus evolved to the modern formof A schreiberi Meanwhile the same ancestral con-tinental populations not isolated from each other con-tinued to exchange genes to varying degrees remainingA boskianus

The IsraeliminusLebanese subspecies A sc syriacus is onlydistantly related to the nominate form A sc schreiberiThis subspecies is highly phylogenetically divergent fromthe Cypriot and Turkish populations having higherp-distances (12S 4 Cytb 11ndash12) than those foundbetween other lacertid species (eg 74ndash82 of Cytbamongst Iberolacerta aranica Iberolacerta aurelioi andIberolacerta bonnali and 41ndash58 of Cytb betweenLacerta bilineata and Lacerta viridis Crochet et al2004 Godinho et al 2005 respectively) The nuclearhaplotype networks further show that Lebanese andIsraeli populations share alleles only with A b asperbut not with the nominotypical Cypriot form Arnold(1983) suggested that the geographical variation ofA boskianus reflects niche differences with animalsfrom xeric areas with dense rigid and spiny vegeta-tion having larger dorsal scales than animals from moremesic areas As was assumed for A schreiberi wesuggest that other mainland populations of A b asperwere the ancestors of the LebaneseminusIsraeli Coastal plainforms We suggest that A s syriacus is an ecomorphof A b asper that dispersed from the usual xerichabitats of the species and adapted to the new moremesic environment of the stable sands of the coastalplains of the eastern Mediterranean As a conse-quence this ecomorph converged on the morphologyof A s schreiberi which inhabits the coastal sands ofCyprus (Baier et al 2009) but still maintains differ-entiating features by having coarser dorsal scales andsharp keels (Salvador 1982 Arnold 1983 Franzen1998) This convergence led to the description ofA s syriacus as a member of A schreiberi The mor-phological assessment and the close morphological simi-larities between A b asper and A s syriacus mayexplain the wrong classification A similar erro-neous reasoning led Reed amp Marx (1959) to identifyspecimens with fine scales from Iraq as A schreiberiSalvador (1982) re-examined these specimens and as-signed them to A boskianus The morphological dif-ferences between the two forms are less prominentespecially where the two forms occur in close geo-graphical proximity in the southern coastal plainand north-western Negev Desert of Israel (Bar ampHaimovitch 2011) According to our results A s syriacus

actually belongs to A b asper being a coastal-duneecomorph convergent with but evolutionarily dis-tinct from A schreiberi Thus our preferred scenariois to treat the name Acanthodactylus schreiberi syriacusBoumlttger 1879 (which was originally described asA boskianus var syriacus by Boumlttger 1879) as a juniorsynonym of the name Acanthodactylus boskianus asper(Audouin 1827)

Recognizing A s syriacus as a junior synonym ofA b asper may have important implications for the con-servation of this coastal sand dune form which is clas-sified as critically endangered in Israel (Dolev ampPervolutzki 2004) However as the Israeli and Leba-nese coastal dune ecosystem has probably developedonly very recently during the Quaternary (Nir 1970Horowitz 1979) this form represents a remarkable caseof rapid evolutionary change It is also a remarkablecase of convergent evolution (with the CypriotA sc schreiberi) Thus we feel that these popula-tions are unique evolutionary entities that merit specialconservation efforts

The use of nuclear genes is a valuable method forestimating species divergence and lineage sorting andhelps evaluate isolated lineages and evolutionary historyThe incorporation of mitochondrial and nuclear dataprovides thorough topologies informative networks anddivergence times that reveal useful information for aproblematic taxonomy such as that of the A boskianusspecies group We have shown that phylogenetic ap-proaches to the confusing taxonomy of two closely relatedand morphologically similar species can shed light ontheir unclear relationships resolve between homoplasyand shared ancestry and identify patterns of speciesevolutionary history and biogeography

ACKNOWLEDGEMENTS

We are grateful to the following people for providingsamples for this study D Donaire B Shacham J ŠmiacutedS M Baha El Din and J Padial We thank MMetallinou and J Šmiacuted for help with the lab work andthe analyses M Novosolov for help with the figuresY L Werner B Shacham and A Levy for fruitful dis-cussions and two anonymous referees for helpful com-ments on an earlier version of this manuscript Wethank the Israel nature and park authority for issuingcollecting permits (nos 38074 38451 38489 38540)The work of J M was financially supported by the Min-istry of Culture of the Czech Republic (DKRVO 201315 National Museum 00023272) S C is funded bygrant CGL2012-36970 from the Ministerio de Economiacuteay Competitividad Spain (cofunded by FEDER) Thisstudy was funded by a Israel Taxonomic Initiative grantto S M K T is supported by an Israel Taxonomic Ini-tiative scholarship

16 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

REFERENCES

Ahmadzadeh F Carretero MA Harris DJ Perera ABohme W 2012 A molecular phylogeny of the eastern groupof ocellated lizard genus Timon (Sauria Lacertidae) basedon mitochondrial and nuclear DNA sequences Amphibia-Reptilia 33 1ndash10

Ahmadzadeh F Flecks M Roumldder D Boumlhme W Ilgaz CcedilHarris DJ Engler JO Uumlzuumlm N Carretero MA 2013Multiple dispersal out of Anatolia biogeography and evolu-tion of oriental green lizards Biological Journal of the LinneanSociety 110 398ndash408

Akaike H 1973 Information theory and an extension of themaximum likelihood principle In Petrov BN Csaki F edsSecond International Symposium on Information Theory Bu-dapest Akademiai Kiado 267ndash281

Amitai P Bouskila A 2001 Handbook of amphibians ampreptiles of Israel Israel Keter Publishing House Ltd

Anderson SC 1999 The lizards of Iran Ithaca NY Societyfor the Study of Amphibians and Reptiles

Arnold EN 1980 The scientific results of the Oman flora andfauna survey 1977 (Dhofar) The reptiles and amphibians ofDhofar southern Arabia Journal of Oman Studies SpecialReport 2 273ndash332

Arnold EN 1983 Osteology genitalia and the relationshipsof Acanthodactylus (Reptilia Lacertidae) Bulletin of the BritishMuseum (Natural History) Zoology 44 291ndash339

Arnold EN Arribas Oacute Carranza S 2007 Systematics ofthe Palaearctic and Oriental lizard tribe Lacertini (SquamataLacertidae Lacertinae) with descriptions of eight new generaZootaxa 1430 1ndash86

Arnold EN Robinson MD Carranza S 2009 A prelimi-nary analysis of phylogenetic relationships and biogeogra-phy of the dangerously venomous Carpet Vipers Echis(Squamata Serpentes Viperidae) based on mitochondrial DNAsequences Amphibia-Reptilia 30 273ndash282

Audouin JV 1827 Explication sommaire des planches dereptiles (suppleacutement) offrant un exposeacute des characteacuteresdes espegraveces In Savigny MJCL ed Description de lrsquoEacutegypteVol 1 Histoire Naturelle Paris Impeacuteriale 161ndash184

Bache F Popescu SM Rabineau M Gorini C Suc JPClauzon G Olivet JL Rubino JL Melinte-DobrinescuMC Estrada F 2012 A two-step process for the refloodingof the Mediterranean after the Messinian Salinity Crisis BasinResearch 24 125ndash153

Baha El Din SM 2006 A guide to the reptiles and amphib-ians of Egypt CairondashNew York American University in CairoPress

Baier FS Sparrow DJ Wiedl H-J 2009 The amphibiansand reptiles of Cyprus Frankfurt am Main Edition Chimaira

Bandelt H-J Forster P Roumlhl A 1999 Median-joining net-works for inferring intraspecific phylogenies Molecular Biologyand Evolution 16 37ndash48

Bar A Haimovitch G 2011 A field guide to reptiles and am-phibians of Israel Herzliya Pazbar Limited

Boumlttger O 1879 Reptilien und Amphibien aus Syrien Berichtuumlber die Senckenbergische Naturforschende Gesellschaft inFrankfurt am Main 1879 57ndash84

Boulenger GA 1919 On a new variety of Acanthodactylusboskianus Daud from the Euphrates The Annals and Maga-zine of Natural History 3 549ndash550

Boulenger GA 1878 Sur les espegraveces drsquoAcanthodactylus desbords de la Mediterraneacutee Bulletin de la Socieacuteteacute zoologiquede France 3 179ndash197

Boulenger GA 1918 Sur les leacutezards du genre AcanthodactylusWieg Bulletin de la Socieacuteteacute zoologique de France 43 143ndash155

Boulenger GA 1921 Monograph of the Lacertidœ Vol II Trus-tees of the British Museum of Natural History London

Carranza S Arnold EN 2012 A review of the geckos of thegenus Hemidactylus (Squamata Gekkonidae) fromOman based on morphology mitochondrial and nuclear datawith descriptions of eight new species Zootaxa 3378 1ndash95

Carranza S Arnold EN Geniez P Roca J Mateo J 2008Radiation multiple dispersal and parallelism in the skinksChalcides and Sphenops (Squamata Scincidae) with com-ments on Scincus and Scincopus and the age of the SaharaDesert Molecular Phylogenetics and Evolution 46 1071ndash1094

Carretero MA Fonseca MM Garcia-Munoz E Brito JCHarris DJ 2011 Adding Acanthodactylus beershebensis tothe mtDNA phylogeny of the Acanthodactylus pardalis groupNorth-Western Journal of Zoology 7 138ndash142

Castresana J 2000 Selection of conserved blocks from multi-ple alignments for their use in phylogenetic analysis Mo-lecular Biology and Evolution 17 540ndash552

Clube TMM Robertson A 1986 The palaeorotation of theTroodos microplate Cyprus in the Late Mesozoic-Early Ce-nozoic plate tectonic framework of the Eastern Mediterra-nean Surveys in Geophysics 8 375ndash437

Cox SC Carranza S Brown RP 2010 Divergence timesand colonization of the Canary Islands by Gallotia lizardsMolecular Phylogenetics and Evolution 56 747ndash757

Crochet PA Geniez P Ineich I 2003 A multivariate analy-sis of the fringe-toed lizards of the Acanthodactylus scutellatusgroup (Squamata Lacertidae) systematic and biogeographi-cal implications Zoological Journal of the Linnean Society137 117ndash155

Crochet P-A Chaline O Surget-Groba Y Debain CCheylan M 2004 Speciation in mountains phylogeographyand phylogeny of the rock lizards genus Iberolacerta (ReptiliaLacertidae) Molecular Phylogenetics and Evolution 30 860ndash866

Daudin FM 1802 Histoire naturelle geacuteneacuterale et particuliegraveredes reptiles ouvrage faisant suite agrave lrsquoHistoire naturelle geacuteneacuteraleet particuliegravere F Dufart

Disi A Modry D Necas P Rifai L 2001 Amphibians andreptiles of the Hashemite Kingdom of Jordan an atlas andfield guide Frankfurt am Main Chimaira

Dolev A Pervolutzki A 2004 Endangered species in IsraelRed list of threatened animals Vertebrates JerusalemThe Nature and Parks Authority and the Society for thePreservation of Nature

Drummond AJ Rambaut A 2007 BEAST Bayesian evo-lutionary analysis by sampling trees BMC EvolutionaryBiology 7 214

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 17

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Ezard T Fujisawa T Barraclough T 2009 Splits speciesrsquolimits by threshold statistics R package version 10-19r48Available at httpR-ForgeR-projectorgprojectssplits

Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783ndash791

Fitzinger LJFJ 1834 Acanthodactylus In Wiegman AFAed Herpetologia mexicana seu Descriptio amphibiorum NovaeHispaniae quae itineribus comitis De Sack Ferdinandi Deppeet Chr Guil Schiede in Museum zoologicum Berolinensepervenerunt Pars prima saurorum species amplectens adjectosystematis saurorum prodromo additisque multis in huncamphibiorum ordinem observationibus Berlin C G Luderitz10

Flot JF 2010 SeqPHASE a web tool for interconvertingPHASE inputoutput files and FASTA sequence align-ments Molecular Ecology Resources 10 162ndash166

Fonseca MM Brito JC Paulo OS Carretero MA HarrisDJ 2009 Systematic and phylogeographical assessment ofthe Acanthodactylus erythrurus group (Reptilia Lacertidae)based on phylogenetic analyses of mitochondrial and nuclearDNA Molecular Phylogenetics and Evolution 51 131ndash142

Fonseca MM Brito JC Rebelo H Kalboussi M LarbesS Carretero MA Harris DJ 2008 Genetic variation amongspiny-footed lizards in the Acanthodactylus pardalis groupfrom North Africa African Zoology 43 8ndash15

Fontaneto D Herniou EA Boschetti C Caprioli M MeloneG Ricci C Barraclough TG 2007 Independently evolv-ing species in asexual bdelloid rotifers PLoS Biology 5 e87

Franzen M 1998 Erstnachweis von Acanthodactylus schreiberischreiberi Boulenger 1879 fuumlr die Tuumlrkei (Squamata SauriaLacertidae) Herpetozoa 11 27ndash36

Funk DJ Omland KE 2003 Species-level paraphyly andpolyphyly frequency causes and consequences with in-sights from animal mitochondrial DNA Annual Review ofEcology Evolution and Systematics 34 397ndash423

Godinho R Crespo EG Ferrand N Harris DJ 2005 Phy-logeny and evolution of the green lizards Lacerta spp(Squamata Lacertidae) based on mitochondrial and nuclearDNA sequences Amphibia-Reptilia 26 271ndash285

Greenbaum E Villanueva CO Kusamba C Aristote MMBranch WR 2011 A molecular phylogeny of EquatorialAfrican Lacertidae with the description of a new genus andspecies from eastern Democratic Republic of the Congo Zoo-logical Journal of the Linnean Society 163 913ndash942

Guillou H Carracedo JC Torrado FP Badiola ER 1996K-Ar ages and magnetic stratigraphy of a hotspot-inducedfast grown oceanic island El Hierro Canary Islands Journalof Volcanology and Geothermal Research 73 141ndash155

Harris DJ Arnold EN 2000 Elucidation of the relation-ships of spiny-footed lizards Acanthodactylus spp (ReptiliaLacertidae) using mitochondrial DNA sequence with com-ments on their biogeography and evolution Journal of Zoology252 351ndash362

Harris DJ Batista V Carretero M 2004 Assessment ofgenetic diversity within Acanthodactylus erythrurus (ReptiliaLacertidae) in Morocco and the Iberian Peninsula usingmitochondrial DNA sequence data Amphibia-Reptilia 25227

Horowitz A 1979 The Quaternary of Israel New York Aca-demic Press

Hraoui-Bloquet S Sadek RA Sindaco R Venchi A 2002The herpetofauna of Lebanon new data on distributionZoology in the Middle East 27 35ndash46

Hsuuml KJ Montadert L Bernoulli D Cita MB EricksonA Garrison RE Kidd RB Megraveliereacutes F Muumlller C WrightR 1977 History of the Mediterranean salinity crisis Nature267 399ndash403

Huelsenbeck JP Rannala B 2004 Frequentist propertiesof Bayesian posterior probabilities of phylogenetic trees undersimple and complex substitution models Systematic Biology53 904ndash913

Huelsenbeck JP Ronquist F 2001 MRBAYES Bayesianinference of phylogenetic trees Bioinformatics 17 754ndash755

Jolivet L Augier R Robin C Suc J-P Rouchy JM 2006Lithospheric-scale geodynamic context of the Messinian sa-linity crisis Sedimentary Geology 188 9ndash33

Kapli P Lymberakis P Poulakakis N Mantziou GParmakelis A Mylonas M 2008 Molecular phylogeny ofthree Mesalina (Reptilia Lacertidae) species (M guttulataM brevirostris and M bahaeldini) from North Africa and theMiddle East another case of paraphyly MolecularPhylogenetics and Evolution 49 102ndash110

Katoh K Toh H 2008 Recent developments in the MAFFTmultiple sequence alignment program Briefings inBioinformatics 9 286ndash298

Krijgsman W Hilgen F Raffi I Sierro F Wilson D 1999Chronology causes and progression of the Messinian salin-ity crisis Nature 400 652ndash655

Lanfear R Calcott B Ho SY Guindon S 2012PartitionFinder combined selection of partitioning schemesand substitution models for phylogenetic analyses Molecu-lar Biology and Evolution 29 1695ndash1701

Le Houeacuterou HN 1997 Climate flora and fauna changes inthe Sahara over the past 500 million years Journal of AridEnvironments 37 619ndash647

Maca-Meyer N Carranza S Rando J Arnold E CabreraV 2003 Status and relationships of the extinct giant CanaryIsland lizard Gallotia goliath (Reptilia Lacertidae)assessed using ancient mtDNA from its mummifiedremains Biological Journal of the Linnean Society 80 659ndash670

McCallum J Robertson A 1990 Pulsed uplift of the TroodosMassif ndash evidence from the Plio-Pleistocene Mesaoria basinIn J Malpas EMM Panayiotou A Xenophontos C edsOphiolites oceanic crustal analogues Proceedings of theSymposium lsquoTroodos 1987rsquo Nicosia (The Geological SurveyDepartment Ministry of Agriculture and NaturalResources) 217ndash229

Metallinou M Arnold EN Crochet P-A Geniez P BritoJC Lymberakis P El Din SB Sindaco R Robinson MCarranza S 2012 Conquering the Sahara and Arabiandeserts systematics and biogeography of Stenodactylus geckos(Reptilia Gekkonidae) BMC Evolutionary Biology 12258

Monaghan MT Wild R Elliot M Fujisawa T Balke MInward DJ Lees DC Ranaivosolo R Eggleton P

18 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Tab

le1

Con

tin

ued

Cod

eS

peci

esV

ouch

erC

oun

try

Loc

alit

y12

SC

ytb

MC

1RA

CM

4c-

mos

Ab1

08dagger

Aca

nth

odac

tylu

sbo

skia

nu

sM

CC

I-R

1452

(1)

Lib

yaW

adi

Mat

hke

ndu

shK

J567

736

KJ5

6780

5K

J548

036

KJ5

4785

0K

J547

967

Ab1

73

Aca

nth

odac

tylu

sbo

skia

nu

sM

auri

tan

iaB

etw

een

Zou

erat

and

Bir

Mog

hre

inK

J567

710

KJ5

6780

7K

J548

045

KJ5

4789

4K

J547

983

Ab1

74

Aca

nth

odac

tylu

sbo

skia

nu

sM

auri

tan

iaB

etw

een

Zou

erat

and

Bir

Mog

hre

inK

J567

714

KJ5

6780

8K

J548

030

KJ5

4789

5K

J547

973

Ab1

58dagger

Aca

nth

odac

tylu

sbo

skia

nu

sM

CC

I-R

1088

(4)

Mor

occo

Bet

wee

nS

aidi

aan

dM

oulo

uya

KJ5

6773

5K

J567

825

KJ5

4804

1K

J547

878

KJ5

4798

4A

b160

A

can

thod

acty

lus

bosk

ian

us

NM

P6V

7448

2M

oroc

coB

etw

een

Ait

-Kh

oujm

anan

dK

erra

ndo

uK

J567

716

KJ5

6781

9K

J548

062

ndashK

J548

001

Ab1

61dagger

Aca

nth

odac

tylu

sbo

skia

nu

sN

MP

6V74

483-

1M

oroc

coR

issa

ni

KJ5

6771

7K

J567

818

KJ5

4805

4K

J547

882

KJ5

4798

5A

b147

Aca

nth

odac

tylu

sbo

skia

nu

sN

MP

6V74

483-

2M

oroc

coR

issa

ni

KJ5

6771

5K

J567

817

ndashndash

ndashA

b175

A

can

thod

acty

lus

bosk

ian

us

Mor

occo

Ou

arza

zate

KJ5

6771

8K

J567

820

KJ5

4804

6K

J547

897

KJ5

4800

2A

b234

A

can

thod

acty

lus

bosk

ian

us

Mor

occo

65

kmE

ofO

um

El-

Ale

kK

J567

711

KJ5

6781

0K

J548

049

KJ5

4789

8K

J547

976

Ab2

35dagger

Aca

nth

odac

tylu

sbo

skia

nu

sM

oroc

coA

kka

KJ5

6771

3K

J567

811

KJ5

4806

9K

J547

899

KJ5

4797

7A

b285

daggerA

can

thod

acty

lus

bosk

ian

us

MV

ZH

erp-

2389

25N

iger

Tafo

kin

13

kmN

NE

ofA

gade

zK

J567

701

KJ5

6782

3K

J548

086

KJ5

4786

0K

J548

000

Ab1

15dagger

Aca

nth

odac

tylu

sbo

skia

nu

sO

man

2km

Sof

Liz

qK

J567

731

KJ5

6782

7K

J548

087

KJ5

4791

2K

J547

968

Ab2

31dagger

Aca

nth

odac

tylu

sbo

skia

nu

sO

man

Niz

wa

KJ5

6767

8K

J567

829

KJ5

4808

9K

J547

914

KJ5

4797

5A

b149

A

can

thod

acty

lus

bosk

ian

us

Om

an10

kmS

Eof

Ku

bara

hK

J567

732

KJ5

6782

8K

J548

088

KJ5

4791

3K

J547

971

Ab1

17A

can

thod

acty

lus

bosk

ian

us

Om

an16

kmS

ofD

uqm

KJ5

6770

7K

J567

802

KJ5

4809

1K

J547

905

KJ5

4799

0A

b116

daggerA

can

thod

acty

lus

bosk

ian

us

MC

CI-

R17

73(1

)O

man

Wad

iS

alit

KJ5

6770

6K

J567

801

KJ5

4809

0K

J547

903

KJ5

4798

9A

b159

Aca

nth

odac

tylu

sbo

skia

nu

sM

CC

I-R

1773

(2)

Om

anW

adi

Sal

itK

J567

708

KJ5

6780

3K

J548

092

KJ5

4790

6K

J547

991

Ab2

32

Aca

nth

odac

tylu

sbo

skia

nu

sO

man

4km

Nof

Raw

iyya

hK

J567

709

KJ5

6780

4K

J548

093

KJ5

4790

4K

J547

992

Ab1

48

Aca

nth

odac

tylu

sbo

skia

nu

sS

uda

nN

ofE

l-K

oin

KJ5

6770

0K

J567

821

KJ5

4804

0K

J547

877

KJ5

4797

0A

b256

A

can

thod

acty

lus

bosk

ian

us

NM

P6V

7045

0-2

Syr

iaA

rR

aqqa

hK

J567

747

KJ5

6784

2K

J548

032

KJ5

4791

5K

J548

012

Ab2

57A

can

thod

acty

lus

bosk

ian

us

NM

P6V

7047

0-1

Syr

iaA

rR

aqqa

hK

J567

748

KJ5

6784

1K

J548

080

KJ5

4788

9K

J548

013

Ab2

39dagger

Aca

nth

odac

tylu

sbo

skia

nu

sN

MP

6V72

502-

1S

yria

Qas

ral

Hay

ral

Gh

arbi

KJ5

6774

4K

J567

838

KJ5

4803

1K

J547

890

KJ5

4800

9A

b240

Aca

nth

odac

tylu

sbo

skia

nu

sN

MP

6V72

502-

2S

yria

Qas

ral

Hay

ral

Gh

arbi

KJ5

6774

5K

J567

839

KJ5

4805

5K

J547

888

KJ5

4801

0A

b255

Aca

nth

odac

tylu

sbo

skia

nu

sN

MP

6V70

443

Syr

iaS

adad

KJ5

6774

6K

J567

840

ndashK

J547

891

KJ5

4801

1A

b110

A

can

thod

acty

lus

bosk

ian

us

MC

CI-

R13

26(1

)Tu

nis

iaN

Esl

opes

ofJe

bel

Sem

mam

aK

J567

695

KJ5

6781

4K

J548

085

KJ5

4789

3K

J548

004

Ab1

57A

can

thod

acty

lus

bosk

ian

us

MC

CI-

R13

26(2

)Tu

nis

iaN

Esl

opes

ofJe

bel

Sem

mam

aK

J567

696

KJ5

6781

5ndash

ndashndash

Ab1

76dagger

Aca

nth

odac

tylu

sbo

skia

nu

sTu

nis

iaH

amm

atal

-Jar

idK

J567

697

KJ5

6781

6K

J548

070

KJ5

4790

9K

J548

005

Ab1

54dagger

Aca

nth

odac

tylu

sbo

skia

nu

sM

CC

I-R

1346

(2)

Tun

isia

33km

Sof

Tata

ouin

eK

J567

702

KJ5

6780

6K

J548

081

KJ5

4790

7K

J547

972

Ab2

36dagger

Aca

nth

odac

tylu

sbo

skia

nu

sW

este

rnS

ahar

aL

aayo

un

eK

J567

712

KJ5

6780

9K

J548

050

KJ5

4789

6K

J547

978

Ab1

56

Aca

nth

odac

tylu

sbo

skia

nu

sM

CC

I-R

823(

3)Ye

men

Sarsquo

yun

oasi

sK

J567

705

KJ5

6780

0K

J548

066

KJ5

4790

2K

J547

981

Ab1

06dagger

Aca

nth

odac

tylu

sbo

skia

nu

sM

CC

I-R

823(

4)Ye

men

Sarsquo

yun

oasi

sK

J567

703

KJ5

6779

9K

J548

065

KJ5

4790

0K

J547

980

Ab1

07dagger

Aca

nth

odac

tylu

sbo

skia

nu

sM

CC

I-R

824

Yem

enD

un

esW

ofS

hib

amK

J567

704

KJ5

6776

9ndash

KJ5

4790

1K

J547

979

As1

98

Aca

nth

odac

tylu

ssc

hre

iber

iat

atu

riM

CC

I-R

1693

(1)

Turk

eyB

otas

KJ5

6774

0K

J567

835

KJ5

4807

5K

J547

919

KJ5

4799

6

6 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

As1

99A

can

thod

acty

lus

sch

reib

eri

atat

uri

MC

CI-

R16

93(1

)Tu

rkey

Bot

asK

J567

741

KJ5

6783

6K

J548

076

KJ5

4791

8K

J547

997

As7

0A

can

thod

acty

lus

sch

reib

eri

sch

reib

eri

TA

U-R

16

151

Cyp

rus

Lar

aba

yA

kam

aspe

nin

sula

KJ5

6773

7K

J567

831

KJ5

4807

1K

J547

916

KJ5

4799

3A

s289

Aca

nth

odac

tylu

ssc

hre

iber

isc

hre

iber

iT

AU

-R

1615

0C

ypru

sL

ara

bay

Aka

mas

pen

insu

laK

J567

738

KJ5

6783

2K

J548

074

KJ5

4792

4K

J547

994

As1

18

Aca

nth

odac

tylu

ssc

hre

iber

isc

hre

iber

iN

MP

6V74

532

Cyp

rus

3km

Sof

Mer

sin

lik

Fam

agu

sta

KJ5

6773

9K

J567

834

KJ5

4807

2K

J547

917

KJ5

4799

5A

s229

daggerA

can

thod

acty

lus

sch

reib

eri

sch

reib

eri

Cyp

rus

Epi

skop

iK

J567

743

KJ5

6783

3K

J548

073

KJ5

4792

1K

J547

999

As2

30

Aca

nth

odac

tylu

ssc

hre

iber

isc

hre

iber

iC

ypru

sV

ryso

ull

esK

J567

742

KJ5

6783

7K

J548

077

KJ5

4792

0K

J547

998

As1

5A

can

thod

acty

lus

sch

reib

eri

syri

acu

sH

UJ-

R-2

3653

Isra

elC

aesa

rea

san

dsK

J567

719

KJ5

6779

1K

J548

025

KJ5

4786

6K

J547

937

As2

74A

can

thod

acty

lus

sch

reib

eri

syri

acu

sH

UJ-

R-2

3410

Isra

elH

ader

ato

Bin

yam

ina

KJ5

6772

2K

J567

792

ndashK

J547

859

KJ5

4794

7A

s121

A

can

thod

acty

lus

sch

reib

eri

syri

acu

sT

AU

-R1

6262

Isra

elR

ish

onL

e-Z

ion

san

dsK

J567

723

KJ5

6779

5K

J548

026

KJ5

4786

7K

J547

938

As2

1dagger

Aca

nth

odac

tylu

ssc

hre

iber

isy

riac

us

TA

U-R

163

98Is

rael

Hol

onsa

nds

KJ5

6772

4K

J567

798

KJ5

4802

7K

J547

873

KJ5

4794

2A

s57

Aca

nth

odac

tylu

ssc

hre

iber

isy

riac

us

TA

U-R

164

07Is

rael

Hol

onsa

nds

KJ5

6772

6K

J567

796

KJ5

4808

4K

J547

863

KJ5

4794

3A

s62

Aca

nth

odac

tylu

ssc

hre

iber

isy

riac

us

TA

U-R

164

12Is

rael

Hol

onsa

nds

KJ5

6772

5K

J567

797

KJ5

4802

9K

J547

871

KJ5

4794

4A

s4dagger

Aca

nth

odac

tylu

ssc

hre

iber

isy

riac

us

HU

J-R

-239

86Is

rael

Niz

zan

imre

serv

eK

J567

666

KJ5

6776

0K

J548

024

KJ5

4784

7K

J547

935

As5

A

can

thod

acty

lus

sch

reib

eri

syri

acu

sH

UJ-

R-2

3987

Isra

elN

izza

nim

rese

rve

KJ5

6766

7K

J567

761

KJ5

4808

2K

J547

848

KJ5

4793

6A

s204

Aca

nth

odac

tylu

ssc

hre

iber

isy

riac

us

HU

J-R

-233

21Is

rael

Ash

qelo

nK

J567

668

KJ5

6776

2ndash

KJ5

4785

7K

J547

940

As2

05A

can

thod

acty

lus

sch

reib

eri

syri

acu

sH

UJ-

R-2

3331

Isra

elA

shqe

lon

KJ5

6766

9K

J567

763

KJ5

4808

3ndash

KJ5

4794

1A

s152

A

can

thod

acty

lus

sch

reib

eri

syri

acu

sM

CC

I-R

925(

1)L

eban

onT

yre

KJ5

6772

0K

J567

793

ndashK

J547

868

KJ5

4794

8A

s153

Aca

nth

odac

tylu

ssc

hre

iber

isy

riac

us

MC

CI-

R92

5(2)

Leb

anon

Tyr

eK

J567

721

KJ5

6779

4ndash

KJ5

4786

9K

J547

939

Ab2

07A

can

thod

acty

lus

blan

ford

iiM

VZ

Her

p-23

4464

Iran

6km

NW

ofB

ampu

rS

ista

nva

KJ5

6775

4K

J567

846

KJ5

4810

1K

J547

929

KJ5

4802

0A

b208

Aca

nth

odac

tylu

sbl

anfo

rdii

MV

ZH

erp-

2460

09Ir

anS

and

dun

es7

kmN

ofB

ampu

rK

J567

755

KJ5

6784

7K

J548

102

KJ5

4793

0K

J548

021

Ac2

86A

can

thod

acty

lus

can

tori

sM

VZ

Her

p-24

8443

Pak

ista

n10

kmS

ofU

thal

B

alu

chis

tan

Pro

vin

ceK

J567

756

KJ5

6785

0K

J548

096

KJ5

4792

2K

J548

018

Ac2

87A

can

thod

acty

lus

can

tori

sM

VZ

Her

p-24

8447

Pak

ista

n45

kmN

Wof

Nag

arP

arka

rK

J567

757

KJ5

6785

1K

J548

097

KJ5

4792

3K

J548

019

Af1

97A

can

thod

acty

lus

feli

cis

CA

S22

7596

Om

an23

kmW

ofA

jdar

awt

KJ5

6773

4K

J567

845

KJ5

4810

0K

J547

928

KJ5

4800

8A

m63

Aca

nth

odac

tylu

sm

asir

aeIB

ES

7643

Om

an20

kmE

ofR

asM

adra

kah

KJ5

6775

3K

J567

849

KJ5

4809

5K

J547

925

KJ5

4801

7A

m50

Aca

nth

odac

tylu

sm

asir

aeO

man

Mas

irae

isla

nd

KJ5

6775

2K

J567

848

KJ5

4809

4K

J547

931

KJ5

4801

6A

o25

Aca

nth

odac

tylu

sop

heo

du

rus

HU

J-R

-191

89Is

rael

Tim

na

vall

eyK

J567

750

KJ5

6784

4K

J548

099

KJ5

4792

7K

J548

007

Ao7

9A

can

thod

acty

lus

oph

eod

uru

sM

CC

I-R

627

Jord

anD

iseh

KJ5

6774

9K

J567

843

KJ5

4809

8K

J547

926

KJ5

4800

6A

s11

Aca

nth

odac

tylu

ssc

ute

llat

us

TA

U-R

163

89Is

rael

Bir

Mas

has

hsa

nds

KJ5

6775

8K

J567

852

KJ5

4810

3K

J547

933

KJ5

4802

2A

s44

Aca

nth

odac

tylu

ssc

ute

llat

us

TA

U-R

164

02Is

rael

Hol

onsa

nds

KJ5

6775

9K

J567

853

KJ5

4810

4K

J547

934

KJ5

4802

3

Hap

loty

pes

(N=

59)

daggerRep

rese

nta

tive

su

sed

for

the

dive

rgen

ceti

me

anal

ysis

(N=

25)

Gen

eab

brev

iati

ons

12S

12

SrR

NA

A

CM

4ac

etyl

chol

iner

gic

rece

ptor

Mu

scar

inic

4c-

mos

oo

cyte

mat

ura

tion

fact

orM

OS

C

ytb

cyto

chro

me

bM

C1R

m

elan

o-co

rtin

1re

cept

or

Inst

itu

tion

alab

brev

iati

ons

CA

S

Cal

ifor

nia

Aca

dem

yof

Sci

ence

sU

SA

H

UJ-

R

Zoo

logi

cal

Mu

seu

m

Heb

rew

Un

iver

sity

ofJe

rusa

lem

Is

rael

IB

ES

In

stit

ute

ofE

volu

tion

ary

Bio

logy

B

arce

lon

aS

pain

M

CC

I-R

M

use

oC

ivic

odi

Sto

ria

Nat

ura

le

Car

mag

nol

a(T

orin

o)

Ital

yM

VZ

Her

pM

use

um

ofV

erte

brat

eZ

oolo

gy(U

niv

ersi

tyof

Cal

ifor

nia

B

erke

ley)

U

SA

N

MP

6V

Nat

ion

alM

use

um

(Nat

ura

lH

isto

ry)

Pra

gue

Cze

chR

epu

blic

T

AU

-R

Zoo

logi

cal

mu

seu

m

Tel

Avi

vU

niv

ersi

ty

Isra

el

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 7

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

parameters unlinked across partitions (Table 2) Twoindependent runs of 2 times 107 generations were carriedout with a sampling frequency of every 1000 genera-tions After examining the standard deviation of thesplit frequencies between the two runs and the po-tential scale reduction factor diagnostic burn-in wasperformed discarding the first 25 trees of each runand the remaining trees were combined in a majorityconsensus tree In both ML and BI alignment gaps weretreated as missing data and the nuclear gene se-quences were not phased Nodes were considered strong-ly supported if they received ML bootstrap values ge 70and posterior probability (pp) support values ge 095 (Wilcoxet al 2002 Huelsenbeck amp Rannala 2004)

A total of 59 haplotypes was identified amongst theA boskianus species group using 792 bp of the con-catenated 12S and Cytb data set (see Table 1) Haplotypenetworks were constructed for the three nuclear genesMC1R ACM4 and c-mos (only full-length sequences)SEQPHASE (Flot 2010) was used to convert the inputfiles and the software PHASE v 211 to resolve phasedhaplotypes (Stephens Smith amp Donnelly 2001 Stephensamp Scheet 2005) Default settings of PHASE were usedexcept for phase probabilities which were set as ge 07

All polymorphic sites with a probability of lt 07 werecoded in both alleles with the appropriate IUPAC am-biguity code The phased nuclear sequences were usedto generate median-joining networks using NET-WORKS v 4611 (Bandelt Forster amp Roumlhl 1999)

In order to assess alternative topologies betweenA schreiberi and A b asper topological constraints thatcould be statistically rejected were constructed We en-forced alternative topologies by hand and compared withthe unconstrained tree (best ML tree) using the ap-proximately unbiased (AU Shimodaira 2002) andShimodairaminusHasegawa (SH Shimodaira amp Hasegawa1999) tests Per-site log likelihoods were estimated inusing RAxMLGUI v 13 (Silvestro amp Michalak 2012)and P-values were calculated using CONSEL(Shimodaira amp Hasegawa 2001)

SPECIES DELIMITATION

In order to reveal the main lineages with the concat-enated analysis and as a prior for species groupingsa mitochondrial phylogeny of 59 haplotypes was per-formed with BEAST v 162 (Drummond amp Rambaut2007) without the outgroups Three individual runs were

Table 2 Information on the partitions used in the phylogenetic analyses with the different partition approaches (ie bygene and by PartitionFinder C codon) including the length model of sequence evolution selected by JModelTest andPartitionFinder and the results of the test of rate homogeneity (LRT) run in MEGA (see Material and methods)

Partition approach Partition Length (bp) Model LRT

By gene 12S sim387 GTR + I + G Not rejected (P lt 07396)Cytb 405 TrN + I + G Rejected (P lt 21819E-7)MC1R 663 GTR + I Not rejected (P lt 1)ACM4 429 HKY + I Not rejected (P lt 1)c-mos 522 TPM1uf + G Not rejected (P lt 1)

PartitionFinder ndashConcatenated

12S Cytb (C1) 2406 GTR + I + Gc-mos(C1) Cytb (C2) TrNef + I + GCytb (C3) TrN + I + GACM4 (C12) MC1R (C1) TrNMC1R (C2) F81MC1R (C3) HKY + GACM4 (C3) c-mos(C2 3) TrNef + I + G

PartitionFinder ndashmtDNA

12S Cytb (C1) 792 SYM + I + GCytb (C2) TrN + I + GCytb (C3) TrN + I + G

PartitionFinder ndashnuclear DNA

ACM4 (C12) c-mos (C12)MC1R (C1)

1614 HKY + I

MC1R (C2) F81MC1R (C3) HKY + GACM4 (C3) c-mos (C3) K80 + I

Gene abbreviations 12S 12S rRNA ACM4 acetylcholinergic receptor Muscarinic 4 c-mos oocyte maturation factor MOSCytb cytochrome b MC1R melano-cortin 1 receptorModel abbreviations F81 Felsenstein 1981 GTR general time-reversible HKY Hasegawa Kishino-Yano K80 Kimura1980 SYM symmetrical model TPM1uf Kimura three-parameter model TrN Tamura-Nei Any of these models caninclude invariable sites (+I) gamma distribution (+G) or both (+I+G)

8 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

performed for 5 times 107 generations with a sampling fre-quency of 10 000 The results were combined to inferthe ultrametric tree after discarding 10 of the samplesfrom each run Models and prior specifications appliedwere as follows (otherwise by default) for partitionsby genes and by PartitionFinder For gene partitionsGTR + I + G strict clock (12S) Hasegawa-Kishino-Yano + Invariable sites + Gamma distribution(HKY + I + G) strict clock molecular clock model (es-timate 0ndash1) (Cytb) coalescence constant size processof speciation random starting tree alpha Uniform (010) GTR Uniform For partitions by PartitionFinderGTR + I + G strict clock (partition 1 = 12S + Cytb codon1 and 2) Tamura-Nei + Gamma distribution (TrN + G)strict clock (partition 2 = Cytb codon 3) coalescenceconstant size process of speciation random starting treealpha Uniform (0 10) Parameter values both for clockand substitution models were unlinked across parti-tions For all analyses implemented in BEAST the threeruns were analysed in TRACER v 15 (Rambaut ampDrummond 2007) confirming convergence The treeswere combined in LogCombiner and TreeAnnotator(available in BEAST package) was used for the pro-duction of the final tree

For estimating species limits directly from the Bayes-ian phylogenetic tree produced with the concat-enated mitochondrial data we used the independentgeneralized mixed Yule-coalescent (GMYC) method (Ponset al 2006) The GMYC model estimated the numberof phylogenetic clusters or lsquospeciesrsquo by identifying theshifts between intraspecific (coalescence) and interspecific(diversification) branch rates (Pons et al 2006) Weperformed the GMYC function in the R v302 lsquosplitsrsquopackage (Ezard Fujisawa amp Barraclough 2009) Alikelihood-ratio test was used to determine if the GMYCmodel with a shift in the branching processes provid-ed a better fit to the data than the null model withno shifts We used a single threshold value (Monaghanet al 2009) which has already been applied success-fully to different groups of organisms (Pons et al 2006Fontaneto et al 2007 Monaghan et al 2009)

ESTIMATION OF DIVERGENCE TIMES

The lack of internal calibration points in Acantho-dactylus (no fossils are known) prevents the directestimation of time in our phylogeny Therefore we usedthe mean substitution rates and their standard errorof the same 12S and Cytb mitochondrial regions ex-tracted from a fully calibrated phylogeny of anotherlacertid group the lizards of the genus Gallotia endemicto the Canary Islands (Cox Carranza amp Brown 2010as was implemented in Carranza amp Arnold 2012) Theinferred calibration rate was estimated using the ageof El Hierro Island (Canary Islands) estimated at112 Mya (Guillou et al 1996) They assumed coloni-

zation of the island by members of the lacertid genusGallotia (Gallotia caesaris caesaris endemic to El HierroIsland) immediately after its formation from the neigh-bouring La Gomera Island (inhabited by the endemicGallotia caesaris gomerae) These two subspecies aremonophyletic sister taxa with low intraspecific vari-ability (Maca-Meyer et al 2003 Cox et al 2010) andthus suitable for calibration

For the estimation of divergence times one repre-sentative of each independent GMYC lineage was usedfrom the ultrametric tree (for the representatives seeTable 1) We used a likelihood-ratio test implementedin MEGA 52 (Tamura et al 2011) to test if the dif-ferent partitions (by genes) included in the dating analy-sis were evolving in a clock-like fashion (Table 2) Thisinformation was used to choose between the strict clockand the relaxed uncorrelated lognormal clock priorsimplemented in BEAST (Monaghan et al 2009) Thedata set included one representative from each lineagefrom the GMYC analysis using sequences from all fivepartitions (nuclear genes unphased) Three individ-ual runs were performed for 5 times 107 generations witha sampling frequency of 10 000 and the results werecombined to infer the ultrametric tree after discard-ing 10 of the samples from each run Models and priorspecifications applied were as follows (otherwise bydefault) GTR + I + G relaxed uncorrelated lognor-mal clock molecular clock model (estimate) (12S Cytb)HKY strict clock (MC1R c-mos) and TrN + I strictclock (ACM4) Yule process of speciation random start-ing tree yulebirthRate (0 1000) alpha Uniform (010) ucldmean of 12S Normal (initial value 000553mean 000553 SD 000128) ucldmean of Cytb Normal(initial value 00164 mean 00164 SD 000317) Pa-rameter values both for clock and substitution modelswere unlinked across partitions

RESULTS

The data set of this study is comprised of 19 samplesof A schreiberi 65 samples of A b asper and 11outgroup samples (Table 1 Fig 1) The data set in-cluded mitochondrial DNA (mtDNA) gene fragmentsof 12S (sim387 bp) and Cytb (405 bp) and nuclear DNA(nDNA) gene fragments of MC1R (663 bp) ACM4(429 bp) and c-mos (522 bp) totalling to sim2406 bp Thenumber of variable (V) and parsimony-informative(Pi) sites for the ingroup are listed in Table S1 Thetwo partition approaches (ie by gene and byPartitionFinder) gave similar results for both the MLand BI analyses The results of the phylogenetic analy-ses of the complete concatenated data set using MLand BI methods produced very similar topologies butdiffered to some extent at the less supported nodesat the intraspecific level (Fig 2) Separated analysesof the nuclear data sets are presented in Figure S1

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 9

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

10 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Together A b asper and A schreiberi form amonophyletic group within Acanthodactylus (Fig 2)Within the group however both taxa are paraphyleticwith A schreiberi as a whole nested within A b asperOur analyses distinguish three major clades (1) cladeA formed by A b asper from Syria (2) clade B in-cludes the two subspecies A sc ataturi from Turkeytogether with A sc schreiberi from Cyprus (3) cladeC which includes specimens of A b asper from the re-maining localities in its distribution range together withA sc syriacus from Israel and Lebanon Clade A is verywell supported and includes specimens of A b asperfrom central and northern Syria (Fig 1) splitting fromother specimens at the basal node of the group is es-timated to have occurred c 654 Mya [95 highest pos-terior density (HPD) 392ndash952 Mya] The level of geneticdifferentiation(p-distance) between these specimens and the remain-ing A b asper and all A schreiberi specimens is 37ndash46 for 12S and 107ndash119 for Cytb Clade B is alsovery well supported and includes two of the threenominal subspecies of A schreiberi A s schreiberi thenominotypical subspecies endemic to Cyprus and A sataturi from Turkey The Turkish subspecies is nestedwithin the Cypriot specimens and the two forms havelow genetic distances from each other (12S 016 Cytb123) This clade is nested between the two A b asperclades (clades A and C) in both the concatenatedand the nuclear tree although the nodes are not wellsupported Clade C is not very well supported It in-cludes a cluster of A b asper and A s syriacus Thisclade includes two inner clades that split around558 Mya (95 HPD 356ndash8 Mya) and divided into threepoorly supported geographical inner groups (Fig 2)northern Jordan and northern Oman (group C1) NorthAfrica (group C2) and samples from the Middle East(Egypt south Israel and south Jordan) with samplesfrom Yemen and southern Oman (group C3) ndash the latterincluding all specimens of the subspecies A s syriacusThe diversification within the North African group isestimated to have started around 456 Mya (95 HPD282ndash647 Mya) The IsraelminusLebanon endemic subspe-cies A s syriacus is genetically highly distinct from

A s schreiberi and A s ataturi making A schreiberiparaphyletic (p-distance 12S 431 416 Cytb 1181202 respectively)

The networks constructed for the phased haplotypesof the full length nuclear markers (MC1R ACM4 andc-mos) are presented in Figure 3 The nuclear networkanalyses show similar results for each of the three genesand closely agree with the phylogenetic tree The CypriotA s schreiberi and Turkish A s ataturi subspecies sharealleles for all three genes and both are distinct fromthe third subspecies A s syriacus Acanthodactylusschreiberi syriacus shares no alleles with the other sub-species of A schreiberi but does share alleles withA b asper for each of the genes Acanthodactylusschreiberi syriacus shares MC1R alleles with A b asperspecimens from Tunisia Syria and Israel ACM4 alleleswith Egyptian Israeli Jordanian and North Africanspecimens and c-mos alleles only with Israeli A b asperspecimens Syrian A b asper samples share one allelewith A s syriacus and two with other A b asper speci-mens from Egypt Israel and North Africa in the MC1Rone allele with an Egyptian A b asper in the ACM4and none in the c-mos gene

In order to better understand the relationships betweenA schreiberi and A boskianus we performed three to-pology tests in which we forced monophyletic group-ings (1) monophyly of A schreiberi (all three subspeciestogether) (2) monophyly of A b asper (3) monophylyof A b asper and of A schreiberi The results of thetopological tests indicate that our data set cannot rejectthe alternative hypotheses of monophyly of A schreiberi(AU P = 0091 SH P = 0062) and that of A b asper(AU P = 011 SH P = 0072) if we allow A schreiberito nest within A b asper or a monophyletic A b aspernesting within A schreiberi When forcing monophylyof both A schreiberi and of A b asper together in thesame tree the results are inconclusive (AU P = 0046SH P = 0051)

The single-threshold model in GMYC yield a topol-ogy that is clearly different from the known taxono-my The GMYC results present a total of 25 and 24ML independent lineages from the Bayesian haplotypemitochondrial phylogeny of the two species for the two

Figure 2 Maximum likelihood (ML) tree of the Acanthodactylus boskianus and Acanthodactylus schreiberi specimensinferred using 12S rRNA cytochrome b mtDNA and melano-cortin 1 receptor acetylcholinergic receptor M4 and oocytematuration factor MOS nuclear gene fragments Posterior probability in the Bayesian analysis is indicated by black dotson the nodes [values ge 095 shown for both gene partitions and partitions by PartitionFinder (PF)] and ML bootstrapsupport values are indicated in parentheses (values ge 70 shown ML ML-PF) Age estimates with BEAST are indicat-ed near the relevant nodes and include the mean and in brackets the HPD 95 confidence interval Sample codes relateto specimens in Table 1 and in Figures 1 and 3 Colours blue Acanthodactylus boskianus asper yellow Acanthodactylusschreiberi ataturi red Acanthodactylus schreiberi syriacus green Acanthodactylus schreiberi schreiberi (Colour versionof figure available online)

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 11

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

12 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

partition approaches (ie by gene and by PartitionFinderFigs S2 S3 respectively) The two partition ap-proaches gave similar clusters but at the less sup-ported nodes they differed at the positions of severallineages The single threshold GMYC result is indi-cated for a single line at 00037 Mya for the gene par-titions and at 002 Mya for PartitionFinder (verticallines in Figs S2 S3) The topology and clusters re-vealed in this analysis correspond to the lineages fromthe phylogeny of the ML and BI methods both for theparaphyly of the two species and the geographical group-ings within A b asper The GMYC results mainly differfrom the ML and BI methods in the position ofA schreiberi from Cyprus and Turkey as a sister cladeto the Syrian A b asper

DISCUSSION

We have provided a comprehensive and thorough as-sessment of the intraspecific phylogenetic relation-ships within A schreiberi and its closest relativeA b asper Our results based on mitochondrial andnuclear DNA data from 84 specimens across the entiredistribution range of A schreiberi and most of the dis-tribution range of A b asper reveal that A schreiberiis paraphyletic and nested entirely within theA boskianus subspecies

HISTORICAL BIOGEOGRAPHY

Acanthodactylus schreiberi is thought to comprisethree subspecies corresponding to three allopatricpopulations in Cyprus Turkey and IsraelminusLebanonThe Cypriot endemic nominotypical subspeciesA sc schreiberi and the Turkish subspeciesA sc ataturi cluster together (to form clade B Fig 2)nesting between A b asper clades This lineage is sisterto a clade of A b asper including A sc syriacus (cladeC Fig 2) We estimate the divergence time of theCypriotminusTurkish lineage of A schreiberi to have beenduring the late Miocene around 6 Mya although thereis no support for this split in the tree In other analy-ses using the whole genus this split is well support-ed in Bayesian analyses (K Tamar S Carranza RSindaco J Moravec JF Trape amp S Meriri unpubldata) Based on mitochondrial data Poulakakis et al(2013) found that both the Cypriot and Turkish sub-species are monophyletic and diverged from each other085 Mya (038ndash156 Mya) According to our results thisdate corresponds to an inner divergence of the

A sc schreiberi lineage rather than to the date at whichA sc schreiberi colonized Cyprus

The discrepancy in the phylogenetic relationship ofA sc schreiberi raises questions regarding the arrivalon Cyprus Cyprus originated with the raising of theTroodos Massif during the upper Cretaceous c 91 to88 Mya (Clube amp Robertson 1986 Mukasa amp Ludden1987) During the middle to late Miocene only a smallproportion of Cyprus was exposed above the Mediter-ranean (McCallum amp Robertson 1990 Robertson 1990)Towards the end of the Miocene sim596 Mya with theclosing of the passage between the Atlantic Ocean andthe Mediterranean basin the Messinian salinity crisisbegan (Krijgsman et al 1999) This resulted in thedrying up of much of the Mediterranean Sea and highsea-mounts emerged to form land bridges with thesurrounding land (Hsuuml et al 1977) By the end of theMiocene and early Pliocene sim533 Mya the passagewith the Atlantic Ocean reopened and the Mediterra-nean basin was refilled (Krijgsman et al 1999) Re-sulting from compressions raising and uplifting of thesurrounding areas towards and during the Pleisto-cene Cyprus was a complete emergent island (McCallumamp Robertson 1990) The possible connection of Cyprusto the mainland (ie to TurkeySyria) during theMessinian is debated as are suggestions of a land con-nection at later periods (Steininger amp Roumlgl 1984 Jolivetet al 2006 Bache et al 2012) Such a connection ifit existed could have provided access for terrestrialorganisms with poor overseas dispersal ability suchas lizards to colonize the island Several studies arguethat post-Messinian sea level changes are unlikely tohave formed connections between Cyprus and the main-land (Steininger amp Roumlgl 1984 Jolivet et al 2006) Thusour dating of the split between the Cypriot A schreiberiand A b asper at c 6 Mya leads us to suggest that theancestor of A s schreiberi colonized Cyprus from themainland through a land bridge connection at the be-ginning of the Messinian crisis rather than by a muchlatermore recent transmarine dispersal as suggestedby Poulakakis et al (2013) Owing to its close rela-tions with A b asper the ancestor of A schreiberi waspresumably mainland A boskianus and the cladogenesisleading to A schreiberi thus rendered A b asperparaphyletic

The Turkish subspecies A schreiberi ataturi was rec-orded for the first time by Franzen (1998) at a veryrestricted area of around 15 km of coastal strip (betweenBotas and Yukarı Burnaz Hatay Province) Owing tothe remarkable morphological similarity between

Figure 3 Haplotype networks of the nuclear gene fragments melano-cortin 1 receptor (MC1R) acetylcholinergic recep-tor M4 (ACM4) and oocyte maturation factor MOS (c-mos) with colours corresponding to Figures 1 and 2 Codes corre-late to the two alleles (ie a and b) of specimens in Table 1 Circle sizes are proportional to the number of alleles (Colourversion of figure available online)

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 13

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

A sc ataturi and the Cypriot population the speci-mens were initially identified as A sc schreiberi(Franzen 1998) Yalccedilinkaya amp Goumlccedilmen (2012) howeverdescribed this population as a new distinct subspe-cies A s ataturi presenting several differences betweenthe two in both morphology and blood-serum pro-teins The origin of A s ataturi remains uncertain asit is debated whether the newly discovered Turkishpopulation is a relict or an introduction from CyprusFranzen (1998) described this population as a pos-sible introduction from Cyprus through the harbourof Botas but Sindaco et al (2000) suggested thatit might be a relict of a previously larger popula-tion because its present distribution is similar tothat of some insects and lizards [Archaeolacerta(Phoenicolacerta) laevis and Ablepharus budaki]Yalccedilinkaya amp Goumlccedilmen (2012) proposed that A sc ataturiarrived in Turkey from the nominate population inCyprus during the Messinian crisis The phylogeneticresults haplotype networks and low levels of geneticdivergence we found suggest that the two subspeciesfrom Cyprus and Turkey have not been geneticallyisolated for a long period of time (ie they share allelesin all three nuclear genes and A s ataturi is nestedwithin A s schreiberi in the phylogeny Figs 2 3) Ourresults therefore contrast with the two latter sce-narios of a relict population or a Messinian disper-sal Both divergence time and the genetic similarityof the two subspecies agree with the original sugges-tion of Franzen (1998) that these animals were intro-duced into Turkey from Cyprus Further support forthis hypothesis is that A s ataturi is restricted to thevicinity of the Botas-Adana harbour and is absent inother suitable habitats (coastal sand dunes) wide-spread in south-eastern Turkey Its close morphologi-cal features to A s schreiberi (Franzen 1998) likewisesupport an introduction scenario

The third subspecies A s syriacus is nested withinA b asper in the concatenated mtDNA and nDNA treesand is clearly genetically distinct from the Cyprus andTurkey A schreiberi lineage The close relations ofA s syriacus with A boskianus may shed light on theorigin of the former Acanthodactylus schreiberi syriacusis distributed on stable sands of the coastal plain ofthe eastern Mediterranean in Israel and southernLebanon (Salvador 1982 Hraoui-Bloquet et al 2002Bar amp Haimovitch 2011) habitats resembling those ofA s schreiberi from Cyprus (Baier Sparrow amp Wiedl2009) The oldest divergence of the A b asper cladethat includes A s syriacus is estimated to have oc-curred during the late Miocene around 558 Mya butno further dates are available for the grouping ofA s syriacus as a result of low support values Thecoastal plain of the eastern Mediterranean was sub-merged during the late Miocene and re-emerged onlytoward the Pliocene (Nir 1970 Horowitz 1979) The

sands of the coastal plain where A s syriacus occurs(Salvador 1982 Arnold 1983 K Tamar amp S Meiripers observ) were repeatedly submerged and re-emerged during the Pleistocene sea-level changes (duringinterglacial and glacial periods respectively) A pos-sible scenario for A sc syriacusrsquos origin includes severalwaves of dispersal of Middle Eastern A boskianus whichoccurs on coarse substrates (Amitai amp Bouskila 2001Disi et al 2001 Baha El Din 2006 pers observ) towardthe Mediterranean shore Acanthodactylus boskianusasper is absent from Mediterranean climate habitatsin Lebanon and Israel It occupies only xeric zonessuggesting an invasion to the coastal plain when sandyhabitats allowed desert flora and fauna to migrate north-wards (Yom-Tov 1988) These populations adapted tosandy soils and evolved morphological features thatdistinguish them from the desert hard substrate formsof A b asper We view this as the most likely scenariogiven the biogeography the phylogenetic results andthe habitat preferences and adaptations of these lizardsAn alternative scenario according to which the an-cestor of A schreiberi originated in Cyprus and dis-persed to the shores of Israel and Lebanon (or originatedin the coastal plain of the Eastern Mediterranean anddispersed to Cyprus) we regard as far less likely Sucha scenario requires much closer genetic relationshipsbetween these two forms and is further weakened bythe close relationship between A s syriacus and thegeographically adjacent A b asper populations

Acanthodactylus boskianus asper is highly variableboth morphologically (Salvador 1982 Arnold 1983) andgenetically (this study) The subspecies is paraphyleticas A schreiberi is nested within it The topology of theA b asper tree shows four different geographical group-ings Syria (clade A) north Jordan plus north OmanNorth Africa and Middle East plus south Arabia (groupsC1 C2 C3 respectively) The different groups in thissubspecies are estimated to have first diverged duringthe late Miocene approximately 65 Mya with the splitof the Syrian population The Syrian lineage is ge-netically distant from A schreiberi and the otherA b asper specimens The nuclear networks indicatethat this group is closer to the other A b asper samplesrather than to A schreiberi The geographical splitsin the rest of the A b asper range (clade C) are esti-mated to have started around 558 Mya These groupsare supported as a distinct clade but are closely relatedto each other in both the concatenated and nuclear trees(Figs 2 S1 respectively) The diversification within thisclade is estimated to have occurred during the lateMiocene to early Pliocene when A b asper dispersedwidely west to North Africa and in Arabia The di-vergence within the North African group (group C2)is estimated to have occurred during the Pliocene ap-proximately 456 Mya with the Egyptian Nigerian andSudanese populations later dispersing west and north

14 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

in Africa This diversification correlates to the arid climatestarting in southern Sahara during the early-mid Plio-cene and later in northern Africa between the Plio-cene and the Pleistocene (Le Houeacuterou 1997) as hasbeen suggested for the dispersal of Mesalina guttulatain Africa (Kapli et al 2008) Other evidence relatesdry climate in North Africa to an earlier period around7 Mya (Schuster et al 2006) as has been suggestedfor the genus Chalcides and other reptiles (Carranzaet al 2008 Metallinou et al 2012 and reference therein)The aridification of North Africa has most likely con-tributed for the successful dispersal of A b asper westfrom south-west Asia into Africa Morphological studiesof A boskianus show relatively uniform populations inNorth Africa suggesting recent migration (Salvador1982 Arnold 1983) The other two geographical group-ings of A b asper from the Middle East and Arabia(groups C1 and C3) are located in two distinct innerclades but their location within each inner clade ispoorly supported The topology of the concatenated tree(Fig 2) shows that the group from northern Jordanand northern Oman (group C1) is closer to the NorthAfrican one than to the geographically close Middle-Eastern and south Arabian group (group C3) The taxo-nomic separation between north and south Oman hasbeen recognized in other species of reptiles and sup-ported by the topography of Oman (eg Echis coloratusand Echis omanensis Arnold Robinson amp Carranza2009) In the nuclear tree (Fig S1) these two groupsare closer to one another and with the North Africangroup form clade C Therefore the low support valuesamongst these groupings prevent an appropriate andthorough analysis of this subspecies The close rela-tionship amongst the geographical groups may reflectclose phylogenetic relationships amongst these popu-lations suggesting recent migration divergence andongoing gene flow

SYSTEMATICS AND TAXONOMIC IMPLICATIONS

The relationships within the A boskianus species groupconflict with the current known taxonomy of A schreiberiand A b asper (samples of the other subspecies ofA boskianus and of A nilsoni were unavailable for thisstudy) Both species have been found to be closelyrelated and paraphyletic The constrained topology testsexemplify the close entangled relationship between thetwo species as the separate monophyly of the two specieswas not rejected and the enforced monophyly of themboth together was inconclusive

Several causes can be responsible for paraphyly inspecies such in the A boskianus species group (Funkamp Omland 2003 and references therein) (1) inad-equate phylogenetic information (2) imperfect taxono-my (incorrectinaccurate species limits) derived frommisidentifying intraspecific variation (3) interspecific

gene flow ndash hybridization through interspecific matingand the subsequent backcrossing of hybrids into theparental populations (4) incomplete lineage sortingbecause of recent speciation events (5) unrecognizedparalogy We suggest that the relationships betweenA schreiberi and A b asper based on mitochondrialand nuclear data are most likely explained by incor-rect taxonomy probably because of the great variabil-ity of the latter species and to convergence As wasthe case in the molecular studies of the A pardalisand A erythrurus species groups (Harris et al 2004Fonseca et al 2008 2009 Carretero et al 2011 andreference therein) there are many problems with thecurrent taxonomic status of several species groups withinAcanthodactylus

Taking the molecular results of our study into accountthere are several systematic approaches to classify-ing the A schreiberiminusA b asper clade The Cypriot andTurkish populations of A schreiberi are very closelyrelated with the latter nested in the former and thetwo subspecies share nuclear alleles (Fig 3) Further-more the low uncorrected p-distance is positively cor-related with subspecies-level distances within otherlacertid species (ie 16 of Cytb in Lacerta bilineatachloronota Godinho et al 2005) We therefore con-clude that Cypriot animals were recently introducedto Turkey and that the Turkish population does notmerit a subspecific rank We suggest that A s ataturiYalccedilinkaya amp Goumlccedilmen 2012 is a junior synonym ofA s schreiberi Boulenger 1878

Regarding the relationships between A schreiberi andA b asper a few scenarios are possible One is to sinkA schreiberi within A boskianus to create one species(A boskianus) with high genetic and morphological vari-ability ranging over a broad distribution Another isfor the two taxa be regarded as a species complex (theA boskianus-schreiberi complex) until further inves-tigation on the subject However although A schreiberiis nested within A b asper the populations from Cyprusand Turkey represent a distinct evolutionary lineagewith distinct genetic and morphological features andthus it is logical to retain the specific status Two othersolutions are possible The first is to re-evaluate theSyrian populations and to consider elevating them aswell as the more divergent lineages (and subspecies)of A boskianus to specific status This would neces-sitate an examination of the phylogeny and morphol-ogy of the other four subspecies of A boskianus(A b boskianus A b euphraticus A b khattensis andA b nigeriensis) and the identification of distinctivephenotypic features in the Syrian lizards Another so-lution is to recognize the maintenance of gene flowamongst mainland populations of A b asper after thedivergence of the insular endemic A schreiberi andthus the evolutionary cohesion of the paraphyleticA b asper Arnold (1983) noted that A schreiberi may

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 15

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

have originated as an isolate from A boskianus becauseof their shared morphology and hemipenis featuresOur results support this scenario which includes thedispersal of A schreiberi to Cyprus from a mainlandpopulation that was most probably A boskianus It maybe assumed that the ancestor of the Cypriot A schreiberiafter arriving on Cyprus remained isolated for a longperiod of time and thus evolved to the modern formof A schreiberi Meanwhile the same ancestral con-tinental populations not isolated from each other con-tinued to exchange genes to varying degrees remainingA boskianus

The IsraeliminusLebanese subspecies A sc syriacus is onlydistantly related to the nominate form A sc schreiberiThis subspecies is highly phylogenetically divergent fromthe Cypriot and Turkish populations having higherp-distances (12S 4 Cytb 11ndash12) than those foundbetween other lacertid species (eg 74ndash82 of Cytbamongst Iberolacerta aranica Iberolacerta aurelioi andIberolacerta bonnali and 41ndash58 of Cytb betweenLacerta bilineata and Lacerta viridis Crochet et al2004 Godinho et al 2005 respectively) The nuclearhaplotype networks further show that Lebanese andIsraeli populations share alleles only with A b asperbut not with the nominotypical Cypriot form Arnold(1983) suggested that the geographical variation ofA boskianus reflects niche differences with animalsfrom xeric areas with dense rigid and spiny vegeta-tion having larger dorsal scales than animals from moremesic areas As was assumed for A schreiberi wesuggest that other mainland populations of A b asperwere the ancestors of the LebaneseminusIsraeli Coastal plainforms We suggest that A s syriacus is an ecomorphof A b asper that dispersed from the usual xerichabitats of the species and adapted to the new moremesic environment of the stable sands of the coastalplains of the eastern Mediterranean As a conse-quence this ecomorph converged on the morphologyof A s schreiberi which inhabits the coastal sands ofCyprus (Baier et al 2009) but still maintains differ-entiating features by having coarser dorsal scales andsharp keels (Salvador 1982 Arnold 1983 Franzen1998) This convergence led to the description ofA s syriacus as a member of A schreiberi The mor-phological assessment and the close morphological simi-larities between A b asper and A s syriacus mayexplain the wrong classification A similar erro-neous reasoning led Reed amp Marx (1959) to identifyspecimens with fine scales from Iraq as A schreiberiSalvador (1982) re-examined these specimens and as-signed them to A boskianus The morphological dif-ferences between the two forms are less prominentespecially where the two forms occur in close geo-graphical proximity in the southern coastal plainand north-western Negev Desert of Israel (Bar ampHaimovitch 2011) According to our results A s syriacus

actually belongs to A b asper being a coastal-duneecomorph convergent with but evolutionarily dis-tinct from A schreiberi Thus our preferred scenariois to treat the name Acanthodactylus schreiberi syriacusBoumlttger 1879 (which was originally described asA boskianus var syriacus by Boumlttger 1879) as a juniorsynonym of the name Acanthodactylus boskianus asper(Audouin 1827)

Recognizing A s syriacus as a junior synonym ofA b asper may have important implications for the con-servation of this coastal sand dune form which is clas-sified as critically endangered in Israel (Dolev ampPervolutzki 2004) However as the Israeli and Leba-nese coastal dune ecosystem has probably developedonly very recently during the Quaternary (Nir 1970Horowitz 1979) this form represents a remarkable caseof rapid evolutionary change It is also a remarkablecase of convergent evolution (with the CypriotA sc schreiberi) Thus we feel that these popula-tions are unique evolutionary entities that merit specialconservation efforts

The use of nuclear genes is a valuable method forestimating species divergence and lineage sorting andhelps evaluate isolated lineages and evolutionary historyThe incorporation of mitochondrial and nuclear dataprovides thorough topologies informative networks anddivergence times that reveal useful information for aproblematic taxonomy such as that of the A boskianusspecies group We have shown that phylogenetic ap-proaches to the confusing taxonomy of two closely relatedand morphologically similar species can shed light ontheir unclear relationships resolve between homoplasyand shared ancestry and identify patterns of speciesevolutionary history and biogeography

ACKNOWLEDGEMENTS

We are grateful to the following people for providingsamples for this study D Donaire B Shacham J ŠmiacutedS M Baha El Din and J Padial We thank MMetallinou and J Šmiacuted for help with the lab work andthe analyses M Novosolov for help with the figuresY L Werner B Shacham and A Levy for fruitful dis-cussions and two anonymous referees for helpful com-ments on an earlier version of this manuscript Wethank the Israel nature and park authority for issuingcollecting permits (nos 38074 38451 38489 38540)The work of J M was financially supported by the Min-istry of Culture of the Czech Republic (DKRVO 201315 National Museum 00023272) S C is funded bygrant CGL2012-36970 from the Ministerio de Economiacuteay Competitividad Spain (cofunded by FEDER) Thisstudy was funded by a Israel Taxonomic Initiative grantto S M K T is supported by an Israel Taxonomic Ini-tiative scholarship

16 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

REFERENCES

Ahmadzadeh F Carretero MA Harris DJ Perera ABohme W 2012 A molecular phylogeny of the eastern groupof ocellated lizard genus Timon (Sauria Lacertidae) basedon mitochondrial and nuclear DNA sequences Amphibia-Reptilia 33 1ndash10

Ahmadzadeh F Flecks M Roumldder D Boumlhme W Ilgaz CcedilHarris DJ Engler JO Uumlzuumlm N Carretero MA 2013Multiple dispersal out of Anatolia biogeography and evolu-tion of oriental green lizards Biological Journal of the LinneanSociety 110 398ndash408

Akaike H 1973 Information theory and an extension of themaximum likelihood principle In Petrov BN Csaki F edsSecond International Symposium on Information Theory Bu-dapest Akademiai Kiado 267ndash281

Amitai P Bouskila A 2001 Handbook of amphibians ampreptiles of Israel Israel Keter Publishing House Ltd

Anderson SC 1999 The lizards of Iran Ithaca NY Societyfor the Study of Amphibians and Reptiles

Arnold EN 1980 The scientific results of the Oman flora andfauna survey 1977 (Dhofar) The reptiles and amphibians ofDhofar southern Arabia Journal of Oman Studies SpecialReport 2 273ndash332

Arnold EN 1983 Osteology genitalia and the relationshipsof Acanthodactylus (Reptilia Lacertidae) Bulletin of the BritishMuseum (Natural History) Zoology 44 291ndash339

Arnold EN Arribas Oacute Carranza S 2007 Systematics ofthe Palaearctic and Oriental lizard tribe Lacertini (SquamataLacertidae Lacertinae) with descriptions of eight new generaZootaxa 1430 1ndash86

Arnold EN Robinson MD Carranza S 2009 A prelimi-nary analysis of phylogenetic relationships and biogeogra-phy of the dangerously venomous Carpet Vipers Echis(Squamata Serpentes Viperidae) based on mitochondrial DNAsequences Amphibia-Reptilia 30 273ndash282

Audouin JV 1827 Explication sommaire des planches dereptiles (suppleacutement) offrant un exposeacute des characteacuteresdes espegraveces In Savigny MJCL ed Description de lrsquoEacutegypteVol 1 Histoire Naturelle Paris Impeacuteriale 161ndash184

Bache F Popescu SM Rabineau M Gorini C Suc JPClauzon G Olivet JL Rubino JL Melinte-DobrinescuMC Estrada F 2012 A two-step process for the refloodingof the Mediterranean after the Messinian Salinity Crisis BasinResearch 24 125ndash153

Baha El Din SM 2006 A guide to the reptiles and amphib-ians of Egypt CairondashNew York American University in CairoPress

Baier FS Sparrow DJ Wiedl H-J 2009 The amphibiansand reptiles of Cyprus Frankfurt am Main Edition Chimaira

Bandelt H-J Forster P Roumlhl A 1999 Median-joining net-works for inferring intraspecific phylogenies Molecular Biologyand Evolution 16 37ndash48

Bar A Haimovitch G 2011 A field guide to reptiles and am-phibians of Israel Herzliya Pazbar Limited

Boumlttger O 1879 Reptilien und Amphibien aus Syrien Berichtuumlber die Senckenbergische Naturforschende Gesellschaft inFrankfurt am Main 1879 57ndash84

Boulenger GA 1919 On a new variety of Acanthodactylusboskianus Daud from the Euphrates The Annals and Maga-zine of Natural History 3 549ndash550

Boulenger GA 1878 Sur les espegraveces drsquoAcanthodactylus desbords de la Mediterraneacutee Bulletin de la Socieacuteteacute zoologiquede France 3 179ndash197

Boulenger GA 1918 Sur les leacutezards du genre AcanthodactylusWieg Bulletin de la Socieacuteteacute zoologique de France 43 143ndash155

Boulenger GA 1921 Monograph of the Lacertidœ Vol II Trus-tees of the British Museum of Natural History London

Carranza S Arnold EN 2012 A review of the geckos of thegenus Hemidactylus (Squamata Gekkonidae) fromOman based on morphology mitochondrial and nuclear datawith descriptions of eight new species Zootaxa 3378 1ndash95

Carranza S Arnold EN Geniez P Roca J Mateo J 2008Radiation multiple dispersal and parallelism in the skinksChalcides and Sphenops (Squamata Scincidae) with com-ments on Scincus and Scincopus and the age of the SaharaDesert Molecular Phylogenetics and Evolution 46 1071ndash1094

Carretero MA Fonseca MM Garcia-Munoz E Brito JCHarris DJ 2011 Adding Acanthodactylus beershebensis tothe mtDNA phylogeny of the Acanthodactylus pardalis groupNorth-Western Journal of Zoology 7 138ndash142

Castresana J 2000 Selection of conserved blocks from multi-ple alignments for their use in phylogenetic analysis Mo-lecular Biology and Evolution 17 540ndash552

Clube TMM Robertson A 1986 The palaeorotation of theTroodos microplate Cyprus in the Late Mesozoic-Early Ce-nozoic plate tectonic framework of the Eastern Mediterra-nean Surveys in Geophysics 8 375ndash437

Cox SC Carranza S Brown RP 2010 Divergence timesand colonization of the Canary Islands by Gallotia lizardsMolecular Phylogenetics and Evolution 56 747ndash757

Crochet PA Geniez P Ineich I 2003 A multivariate analy-sis of the fringe-toed lizards of the Acanthodactylus scutellatusgroup (Squamata Lacertidae) systematic and biogeographi-cal implications Zoological Journal of the Linnean Society137 117ndash155

Crochet P-A Chaline O Surget-Groba Y Debain CCheylan M 2004 Speciation in mountains phylogeographyand phylogeny of the rock lizards genus Iberolacerta (ReptiliaLacertidae) Molecular Phylogenetics and Evolution 30 860ndash866

Daudin FM 1802 Histoire naturelle geacuteneacuterale et particuliegraveredes reptiles ouvrage faisant suite agrave lrsquoHistoire naturelle geacuteneacuteraleet particuliegravere F Dufart

Disi A Modry D Necas P Rifai L 2001 Amphibians andreptiles of the Hashemite Kingdom of Jordan an atlas andfield guide Frankfurt am Main Chimaira

Dolev A Pervolutzki A 2004 Endangered species in IsraelRed list of threatened animals Vertebrates JerusalemThe Nature and Parks Authority and the Society for thePreservation of Nature

Drummond AJ Rambaut A 2007 BEAST Bayesian evo-lutionary analysis by sampling trees BMC EvolutionaryBiology 7 214

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 17

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Ezard T Fujisawa T Barraclough T 2009 Splits speciesrsquolimits by threshold statistics R package version 10-19r48Available at httpR-ForgeR-projectorgprojectssplits

Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783ndash791

Fitzinger LJFJ 1834 Acanthodactylus In Wiegman AFAed Herpetologia mexicana seu Descriptio amphibiorum NovaeHispaniae quae itineribus comitis De Sack Ferdinandi Deppeet Chr Guil Schiede in Museum zoologicum Berolinensepervenerunt Pars prima saurorum species amplectens adjectosystematis saurorum prodromo additisque multis in huncamphibiorum ordinem observationibus Berlin C G Luderitz10

Flot JF 2010 SeqPHASE a web tool for interconvertingPHASE inputoutput files and FASTA sequence align-ments Molecular Ecology Resources 10 162ndash166

Fonseca MM Brito JC Paulo OS Carretero MA HarrisDJ 2009 Systematic and phylogeographical assessment ofthe Acanthodactylus erythrurus group (Reptilia Lacertidae)based on phylogenetic analyses of mitochondrial and nuclearDNA Molecular Phylogenetics and Evolution 51 131ndash142

Fonseca MM Brito JC Rebelo H Kalboussi M LarbesS Carretero MA Harris DJ 2008 Genetic variation amongspiny-footed lizards in the Acanthodactylus pardalis groupfrom North Africa African Zoology 43 8ndash15

Fontaneto D Herniou EA Boschetti C Caprioli M MeloneG Ricci C Barraclough TG 2007 Independently evolv-ing species in asexual bdelloid rotifers PLoS Biology 5 e87

Franzen M 1998 Erstnachweis von Acanthodactylus schreiberischreiberi Boulenger 1879 fuumlr die Tuumlrkei (Squamata SauriaLacertidae) Herpetozoa 11 27ndash36

Funk DJ Omland KE 2003 Species-level paraphyly andpolyphyly frequency causes and consequences with in-sights from animal mitochondrial DNA Annual Review ofEcology Evolution and Systematics 34 397ndash423

Godinho R Crespo EG Ferrand N Harris DJ 2005 Phy-logeny and evolution of the green lizards Lacerta spp(Squamata Lacertidae) based on mitochondrial and nuclearDNA sequences Amphibia-Reptilia 26 271ndash285

Greenbaum E Villanueva CO Kusamba C Aristote MMBranch WR 2011 A molecular phylogeny of EquatorialAfrican Lacertidae with the description of a new genus andspecies from eastern Democratic Republic of the Congo Zoo-logical Journal of the Linnean Society 163 913ndash942

Guillou H Carracedo JC Torrado FP Badiola ER 1996K-Ar ages and magnetic stratigraphy of a hotspot-inducedfast grown oceanic island El Hierro Canary Islands Journalof Volcanology and Geothermal Research 73 141ndash155

Harris DJ Arnold EN 2000 Elucidation of the relation-ships of spiny-footed lizards Acanthodactylus spp (ReptiliaLacertidae) using mitochondrial DNA sequence with com-ments on their biogeography and evolution Journal of Zoology252 351ndash362

Harris DJ Batista V Carretero M 2004 Assessment ofgenetic diversity within Acanthodactylus erythrurus (ReptiliaLacertidae) in Morocco and the Iberian Peninsula usingmitochondrial DNA sequence data Amphibia-Reptilia 25227

Horowitz A 1979 The Quaternary of Israel New York Aca-demic Press

Hraoui-Bloquet S Sadek RA Sindaco R Venchi A 2002The herpetofauna of Lebanon new data on distributionZoology in the Middle East 27 35ndash46

Hsuuml KJ Montadert L Bernoulli D Cita MB EricksonA Garrison RE Kidd RB Megraveliereacutes F Muumlller C WrightR 1977 History of the Mediterranean salinity crisis Nature267 399ndash403

Huelsenbeck JP Rannala B 2004 Frequentist propertiesof Bayesian posterior probabilities of phylogenetic trees undersimple and complex substitution models Systematic Biology53 904ndash913

Huelsenbeck JP Ronquist F 2001 MRBAYES Bayesianinference of phylogenetic trees Bioinformatics 17 754ndash755

Jolivet L Augier R Robin C Suc J-P Rouchy JM 2006Lithospheric-scale geodynamic context of the Messinian sa-linity crisis Sedimentary Geology 188 9ndash33

Kapli P Lymberakis P Poulakakis N Mantziou GParmakelis A Mylonas M 2008 Molecular phylogeny ofthree Mesalina (Reptilia Lacertidae) species (M guttulataM brevirostris and M bahaeldini) from North Africa and theMiddle East another case of paraphyly MolecularPhylogenetics and Evolution 49 102ndash110

Katoh K Toh H 2008 Recent developments in the MAFFTmultiple sequence alignment program Briefings inBioinformatics 9 286ndash298

Krijgsman W Hilgen F Raffi I Sierro F Wilson D 1999Chronology causes and progression of the Messinian salin-ity crisis Nature 400 652ndash655

Lanfear R Calcott B Ho SY Guindon S 2012PartitionFinder combined selection of partitioning schemesand substitution models for phylogenetic analyses Molecu-lar Biology and Evolution 29 1695ndash1701

Le Houeacuterou HN 1997 Climate flora and fauna changes inthe Sahara over the past 500 million years Journal of AridEnvironments 37 619ndash647

Maca-Meyer N Carranza S Rando J Arnold E CabreraV 2003 Status and relationships of the extinct giant CanaryIsland lizard Gallotia goliath (Reptilia Lacertidae)assessed using ancient mtDNA from its mummifiedremains Biological Journal of the Linnean Society 80 659ndash670

McCallum J Robertson A 1990 Pulsed uplift of the TroodosMassif ndash evidence from the Plio-Pleistocene Mesaoria basinIn J Malpas EMM Panayiotou A Xenophontos C edsOphiolites oceanic crustal analogues Proceedings of theSymposium lsquoTroodos 1987rsquo Nicosia (The Geological SurveyDepartment Ministry of Agriculture and NaturalResources) 217ndash229

Metallinou M Arnold EN Crochet P-A Geniez P BritoJC Lymberakis P El Din SB Sindaco R Robinson MCarranza S 2012 Conquering the Sahara and Arabiandeserts systematics and biogeography of Stenodactylus geckos(Reptilia Gekkonidae) BMC Evolutionary Biology 12258

Monaghan MT Wild R Elliot M Fujisawa T Balke MInward DJ Lees DC Ranaivosolo R Eggleton P

18 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

As1

99A

can

thod

acty

lus

sch

reib

eri

atat

uri

MC

CI-

R16

93(1

)Tu

rkey

Bot

asK

J567

741

KJ5

6783

6K

J548

076

KJ5

4791

8K

J547

997

As7

0A

can

thod

acty

lus

sch

reib

eri

sch

reib

eri

TA

U-R

16

151

Cyp

rus

Lar

aba

yA

kam

aspe

nin

sula

KJ5

6773

7K

J567

831

KJ5

4807

1K

J547

916

KJ5

4799

3A

s289

Aca

nth

odac

tylu

ssc

hre

iber

isc

hre

iber

iT

AU

-R

1615

0C

ypru

sL

ara

bay

Aka

mas

pen

insu

laK

J567

738

KJ5

6783

2K

J548

074

KJ5

4792

4K

J547

994

As1

18

Aca

nth

odac

tylu

ssc

hre

iber

isc

hre

iber

iN

MP

6V74

532

Cyp

rus

3km

Sof

Mer

sin

lik

Fam

agu

sta

KJ5

6773

9K

J567

834

KJ5

4807

2K

J547

917

KJ5

4799

5A

s229

daggerA

can

thod

acty

lus

sch

reib

eri

sch

reib

eri

Cyp

rus

Epi

skop

iK

J567

743

KJ5

6783

3K

J548

073

KJ5

4792

1K

J547

999

As2

30

Aca

nth

odac

tylu

ssc

hre

iber

isc

hre

iber

iC

ypru

sV

ryso

ull

esK

J567

742

KJ5

6783

7K

J548

077

KJ5

4792

0K

J547

998

As1

5A

can

thod

acty

lus

sch

reib

eri

syri

acu

sH

UJ-

R-2

3653

Isra

elC

aesa

rea

san

dsK

J567

719

KJ5

6779

1K

J548

025

KJ5

4786

6K

J547

937

As2

74A

can

thod

acty

lus

sch

reib

eri

syri

acu

sH

UJ-

R-2

3410

Isra

elH

ader

ato

Bin

yam

ina

KJ5

6772

2K

J567

792

ndashK

J547

859

KJ5

4794

7A

s121

A

can

thod

acty

lus

sch

reib

eri

syri

acu

sT

AU

-R1

6262

Isra

elR

ish

onL

e-Z

ion

san

dsK

J567

723

KJ5

6779

5K

J548

026

KJ5

4786

7K

J547

938

As2

1dagger

Aca

nth

odac

tylu

ssc

hre

iber

isy

riac

us

TA

U-R

163

98Is

rael

Hol

onsa

nds

KJ5

6772

4K

J567

798

KJ5

4802

7K

J547

873

KJ5

4794

2A

s57

Aca

nth

odac

tylu

ssc

hre

iber

isy

riac

us

TA

U-R

164

07Is

rael

Hol

onsa

nds

KJ5

6772

6K

J567

796

KJ5

4808

4K

J547

863

KJ5

4794

3A

s62

Aca

nth

odac

tylu

ssc

hre

iber

isy

riac

us

TA

U-R

164

12Is

rael

Hol

onsa

nds

KJ5

6772

5K

J567

797

KJ5

4802

9K

J547

871

KJ5

4794

4A

s4dagger

Aca

nth

odac

tylu

ssc

hre

iber

isy

riac

us

HU

J-R

-239

86Is

rael

Niz

zan

imre

serv

eK

J567

666

KJ5

6776

0K

J548

024

KJ5

4784

7K

J547

935

As5

A

can

thod

acty

lus

sch

reib

eri

syri

acu

sH

UJ-

R-2

3987

Isra

elN

izza

nim

rese

rve

KJ5

6766

7K

J567

761

KJ5

4808

2K

J547

848

KJ5

4793

6A

s204

Aca

nth

odac

tylu

ssc

hre

iber

isy

riac

us

HU

J-R

-233

21Is

rael

Ash

qelo

nK

J567

668

KJ5

6776

2ndash

KJ5

4785

7K

J547

940

As2

05A

can

thod

acty

lus

sch

reib

eri

syri

acu

sH

UJ-

R-2

3331

Isra

elA

shqe

lon

KJ5

6766

9K

J567

763

KJ5

4808

3ndash

KJ5

4794

1A

s152

A

can

thod

acty

lus

sch

reib

eri

syri

acu

sM

CC

I-R

925(

1)L

eban

onT

yre

KJ5

6772

0K

J567

793

ndashK

J547

868

KJ5

4794

8A

s153

Aca

nth

odac

tylu

ssc

hre

iber

isy

riac

us

MC

CI-

R92

5(2)

Leb

anon

Tyr

eK

J567

721

KJ5

6779

4ndash

KJ5

4786

9K

J547

939

Ab2

07A

can

thod

acty

lus

blan

ford

iiM

VZ

Her

p-23

4464

Iran

6km

NW

ofB

ampu

rS

ista

nva

KJ5

6775

4K

J567

846

KJ5

4810

1K

J547

929

KJ5

4802

0A

b208

Aca

nth

odac

tylu

sbl

anfo

rdii

MV

ZH

erp-

2460

09Ir

anS

and

dun

es7

kmN

ofB

ampu

rK

J567

755

KJ5

6784

7K

J548

102

KJ5

4793

0K

J548

021

Ac2

86A

can

thod

acty

lus

can

tori

sM

VZ

Her

p-24

8443

Pak

ista

n10

kmS

ofU

thal

B

alu

chis

tan

Pro

vin

ceK

J567

756

KJ5

6785

0K

J548

096

KJ5

4792

2K

J548

018

Ac2

87A

can

thod

acty

lus

can

tori

sM

VZ

Her

p-24

8447

Pak

ista

n45

kmN

Wof

Nag

arP

arka

rK

J567

757

KJ5

6785

1K

J548

097

KJ5

4792

3K

J548

019

Af1

97A

can

thod

acty

lus

feli

cis

CA

S22

7596

Om

an23

kmW

ofA

jdar

awt

KJ5

6773

4K

J567

845

KJ5

4810

0K

J547

928

KJ5

4800

8A

m63

Aca

nth

odac

tylu

sm

asir

aeIB

ES

7643

Om

an20

kmE

ofR

asM

adra

kah

KJ5

6775

3K

J567

849

KJ5

4809

5K

J547

925

KJ5

4801

7A

m50

Aca

nth

odac

tylu

sm

asir

aeO

man

Mas

irae

isla

nd

KJ5

6775

2K

J567

848

KJ5

4809

4K

J547

931

KJ5

4801

6A

o25

Aca

nth

odac

tylu

sop

heo

du

rus

HU

J-R

-191

89Is

rael

Tim

na

vall

eyK

J567

750

KJ5

6784

4K

J548

099

KJ5

4792

7K

J548

007

Ao7

9A

can

thod

acty

lus

oph

eod

uru

sM

CC

I-R

627

Jord

anD

iseh

KJ5

6774

9K

J567

843

KJ5

4809

8K

J547

926

KJ5

4800

6A

s11

Aca

nth

odac

tylu

ssc

ute

llat

us

TA

U-R

163

89Is

rael

Bir

Mas

has

hsa

nds

KJ5

6775

8K

J567

852

KJ5

4810

3K

J547

933

KJ5

4802

2A

s44

Aca

nth

odac

tylu

ssc

ute

llat

us

TA

U-R

164

02Is

rael

Hol

onsa

nds

KJ5

6775

9K

J567

853

KJ5

4810

4K

J547

934

KJ5

4802

3

Hap

loty

pes

(N=

59)

daggerRep

rese

nta

tive

su

sed

for

the

dive

rgen

ceti

me

anal

ysis

(N=

25)

Gen

eab

brev

iati

ons

12S

12

SrR

NA

A

CM

4ac

etyl

chol

iner

gic

rece

ptor

Mu

scar

inic

4c-

mos

oo

cyte

mat

ura

tion

fact

orM

OS

C

ytb

cyto

chro

me

bM

C1R

m

elan

o-co

rtin

1re

cept

or

Inst

itu

tion

alab

brev

iati

ons

CA

S

Cal

ifor

nia

Aca

dem

yof

Sci

ence

sU

SA

H

UJ-

R

Zoo

logi

cal

Mu

seu

m

Heb

rew

Un

iver

sity

ofJe

rusa

lem

Is

rael

IB

ES

In

stit

ute

ofE

volu

tion

ary

Bio

logy

B

arce

lon

aS

pain

M

CC

I-R

M

use

oC

ivic

odi

Sto

ria

Nat

ura

le

Car

mag

nol

a(T

orin

o)

Ital

yM

VZ

Her

pM

use

um

ofV

erte

brat

eZ

oolo

gy(U

niv

ersi

tyof

Cal

ifor

nia

B

erke

ley)

U

SA

N

MP

6V

Nat

ion

alM

use

um

(Nat

ura

lH

isto

ry)

Pra

gue

Cze

chR

epu

blic

T

AU

-R

Zoo

logi

cal

mu

seu

m

Tel

Avi

vU

niv

ersi

ty

Isra

el

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 7

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

parameters unlinked across partitions (Table 2) Twoindependent runs of 2 times 107 generations were carriedout with a sampling frequency of every 1000 genera-tions After examining the standard deviation of thesplit frequencies between the two runs and the po-tential scale reduction factor diagnostic burn-in wasperformed discarding the first 25 trees of each runand the remaining trees were combined in a majorityconsensus tree In both ML and BI alignment gaps weretreated as missing data and the nuclear gene se-quences were not phased Nodes were considered strong-ly supported if they received ML bootstrap values ge 70and posterior probability (pp) support values ge 095 (Wilcoxet al 2002 Huelsenbeck amp Rannala 2004)

A total of 59 haplotypes was identified amongst theA boskianus species group using 792 bp of the con-catenated 12S and Cytb data set (see Table 1) Haplotypenetworks were constructed for the three nuclear genesMC1R ACM4 and c-mos (only full-length sequences)SEQPHASE (Flot 2010) was used to convert the inputfiles and the software PHASE v 211 to resolve phasedhaplotypes (Stephens Smith amp Donnelly 2001 Stephensamp Scheet 2005) Default settings of PHASE were usedexcept for phase probabilities which were set as ge 07

All polymorphic sites with a probability of lt 07 werecoded in both alleles with the appropriate IUPAC am-biguity code The phased nuclear sequences were usedto generate median-joining networks using NET-WORKS v 4611 (Bandelt Forster amp Roumlhl 1999)

In order to assess alternative topologies betweenA schreiberi and A b asper topological constraints thatcould be statistically rejected were constructed We en-forced alternative topologies by hand and compared withthe unconstrained tree (best ML tree) using the ap-proximately unbiased (AU Shimodaira 2002) andShimodairaminusHasegawa (SH Shimodaira amp Hasegawa1999) tests Per-site log likelihoods were estimated inusing RAxMLGUI v 13 (Silvestro amp Michalak 2012)and P-values were calculated using CONSEL(Shimodaira amp Hasegawa 2001)

SPECIES DELIMITATION

In order to reveal the main lineages with the concat-enated analysis and as a prior for species groupingsa mitochondrial phylogeny of 59 haplotypes was per-formed with BEAST v 162 (Drummond amp Rambaut2007) without the outgroups Three individual runs were

Table 2 Information on the partitions used in the phylogenetic analyses with the different partition approaches (ie bygene and by PartitionFinder C codon) including the length model of sequence evolution selected by JModelTest andPartitionFinder and the results of the test of rate homogeneity (LRT) run in MEGA (see Material and methods)

Partition approach Partition Length (bp) Model LRT

By gene 12S sim387 GTR + I + G Not rejected (P lt 07396)Cytb 405 TrN + I + G Rejected (P lt 21819E-7)MC1R 663 GTR + I Not rejected (P lt 1)ACM4 429 HKY + I Not rejected (P lt 1)c-mos 522 TPM1uf + G Not rejected (P lt 1)

PartitionFinder ndashConcatenated

12S Cytb (C1) 2406 GTR + I + Gc-mos(C1) Cytb (C2) TrNef + I + GCytb (C3) TrN + I + GACM4 (C12) MC1R (C1) TrNMC1R (C2) F81MC1R (C3) HKY + GACM4 (C3) c-mos(C2 3) TrNef + I + G

PartitionFinder ndashmtDNA

12S Cytb (C1) 792 SYM + I + GCytb (C2) TrN + I + GCytb (C3) TrN + I + G

PartitionFinder ndashnuclear DNA

ACM4 (C12) c-mos (C12)MC1R (C1)

1614 HKY + I

MC1R (C2) F81MC1R (C3) HKY + GACM4 (C3) c-mos (C3) K80 + I

Gene abbreviations 12S 12S rRNA ACM4 acetylcholinergic receptor Muscarinic 4 c-mos oocyte maturation factor MOSCytb cytochrome b MC1R melano-cortin 1 receptorModel abbreviations F81 Felsenstein 1981 GTR general time-reversible HKY Hasegawa Kishino-Yano K80 Kimura1980 SYM symmetrical model TPM1uf Kimura three-parameter model TrN Tamura-Nei Any of these models caninclude invariable sites (+I) gamma distribution (+G) or both (+I+G)

8 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

performed for 5 times 107 generations with a sampling fre-quency of 10 000 The results were combined to inferthe ultrametric tree after discarding 10 of the samplesfrom each run Models and prior specifications appliedwere as follows (otherwise by default) for partitionsby genes and by PartitionFinder For gene partitionsGTR + I + G strict clock (12S) Hasegawa-Kishino-Yano + Invariable sites + Gamma distribution(HKY + I + G) strict clock molecular clock model (es-timate 0ndash1) (Cytb) coalescence constant size processof speciation random starting tree alpha Uniform (010) GTR Uniform For partitions by PartitionFinderGTR + I + G strict clock (partition 1 = 12S + Cytb codon1 and 2) Tamura-Nei + Gamma distribution (TrN + G)strict clock (partition 2 = Cytb codon 3) coalescenceconstant size process of speciation random starting treealpha Uniform (0 10) Parameter values both for clockand substitution models were unlinked across parti-tions For all analyses implemented in BEAST the threeruns were analysed in TRACER v 15 (Rambaut ampDrummond 2007) confirming convergence The treeswere combined in LogCombiner and TreeAnnotator(available in BEAST package) was used for the pro-duction of the final tree

For estimating species limits directly from the Bayes-ian phylogenetic tree produced with the concat-enated mitochondrial data we used the independentgeneralized mixed Yule-coalescent (GMYC) method (Ponset al 2006) The GMYC model estimated the numberof phylogenetic clusters or lsquospeciesrsquo by identifying theshifts between intraspecific (coalescence) and interspecific(diversification) branch rates (Pons et al 2006) Weperformed the GMYC function in the R v302 lsquosplitsrsquopackage (Ezard Fujisawa amp Barraclough 2009) Alikelihood-ratio test was used to determine if the GMYCmodel with a shift in the branching processes provid-ed a better fit to the data than the null model withno shifts We used a single threshold value (Monaghanet al 2009) which has already been applied success-fully to different groups of organisms (Pons et al 2006Fontaneto et al 2007 Monaghan et al 2009)

ESTIMATION OF DIVERGENCE TIMES

The lack of internal calibration points in Acantho-dactylus (no fossils are known) prevents the directestimation of time in our phylogeny Therefore we usedthe mean substitution rates and their standard errorof the same 12S and Cytb mitochondrial regions ex-tracted from a fully calibrated phylogeny of anotherlacertid group the lizards of the genus Gallotia endemicto the Canary Islands (Cox Carranza amp Brown 2010as was implemented in Carranza amp Arnold 2012) Theinferred calibration rate was estimated using the ageof El Hierro Island (Canary Islands) estimated at112 Mya (Guillou et al 1996) They assumed coloni-

zation of the island by members of the lacertid genusGallotia (Gallotia caesaris caesaris endemic to El HierroIsland) immediately after its formation from the neigh-bouring La Gomera Island (inhabited by the endemicGallotia caesaris gomerae) These two subspecies aremonophyletic sister taxa with low intraspecific vari-ability (Maca-Meyer et al 2003 Cox et al 2010) andthus suitable for calibration

For the estimation of divergence times one repre-sentative of each independent GMYC lineage was usedfrom the ultrametric tree (for the representatives seeTable 1) We used a likelihood-ratio test implementedin MEGA 52 (Tamura et al 2011) to test if the dif-ferent partitions (by genes) included in the dating analy-sis were evolving in a clock-like fashion (Table 2) Thisinformation was used to choose between the strict clockand the relaxed uncorrelated lognormal clock priorsimplemented in BEAST (Monaghan et al 2009) Thedata set included one representative from each lineagefrom the GMYC analysis using sequences from all fivepartitions (nuclear genes unphased) Three individ-ual runs were performed for 5 times 107 generations witha sampling frequency of 10 000 and the results werecombined to infer the ultrametric tree after discard-ing 10 of the samples from each run Models and priorspecifications applied were as follows (otherwise bydefault) GTR + I + G relaxed uncorrelated lognor-mal clock molecular clock model (estimate) (12S Cytb)HKY strict clock (MC1R c-mos) and TrN + I strictclock (ACM4) Yule process of speciation random start-ing tree yulebirthRate (0 1000) alpha Uniform (010) ucldmean of 12S Normal (initial value 000553mean 000553 SD 000128) ucldmean of Cytb Normal(initial value 00164 mean 00164 SD 000317) Pa-rameter values both for clock and substitution modelswere unlinked across partitions

RESULTS

The data set of this study is comprised of 19 samplesof A schreiberi 65 samples of A b asper and 11outgroup samples (Table 1 Fig 1) The data set in-cluded mitochondrial DNA (mtDNA) gene fragmentsof 12S (sim387 bp) and Cytb (405 bp) and nuclear DNA(nDNA) gene fragments of MC1R (663 bp) ACM4(429 bp) and c-mos (522 bp) totalling to sim2406 bp Thenumber of variable (V) and parsimony-informative(Pi) sites for the ingroup are listed in Table S1 Thetwo partition approaches (ie by gene and byPartitionFinder) gave similar results for both the MLand BI analyses The results of the phylogenetic analy-ses of the complete concatenated data set using MLand BI methods produced very similar topologies butdiffered to some extent at the less supported nodesat the intraspecific level (Fig 2) Separated analysesof the nuclear data sets are presented in Figure S1

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 9

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

10 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Together A b asper and A schreiberi form amonophyletic group within Acanthodactylus (Fig 2)Within the group however both taxa are paraphyleticwith A schreiberi as a whole nested within A b asperOur analyses distinguish three major clades (1) cladeA formed by A b asper from Syria (2) clade B in-cludes the two subspecies A sc ataturi from Turkeytogether with A sc schreiberi from Cyprus (3) cladeC which includes specimens of A b asper from the re-maining localities in its distribution range together withA sc syriacus from Israel and Lebanon Clade A is verywell supported and includes specimens of A b asperfrom central and northern Syria (Fig 1) splitting fromother specimens at the basal node of the group is es-timated to have occurred c 654 Mya [95 highest pos-terior density (HPD) 392ndash952 Mya] The level of geneticdifferentiation(p-distance) between these specimens and the remain-ing A b asper and all A schreiberi specimens is 37ndash46 for 12S and 107ndash119 for Cytb Clade B is alsovery well supported and includes two of the threenominal subspecies of A schreiberi A s schreiberi thenominotypical subspecies endemic to Cyprus and A sataturi from Turkey The Turkish subspecies is nestedwithin the Cypriot specimens and the two forms havelow genetic distances from each other (12S 016 Cytb123) This clade is nested between the two A b asperclades (clades A and C) in both the concatenatedand the nuclear tree although the nodes are not wellsupported Clade C is not very well supported It in-cludes a cluster of A b asper and A s syriacus Thisclade includes two inner clades that split around558 Mya (95 HPD 356ndash8 Mya) and divided into threepoorly supported geographical inner groups (Fig 2)northern Jordan and northern Oman (group C1) NorthAfrica (group C2) and samples from the Middle East(Egypt south Israel and south Jordan) with samplesfrom Yemen and southern Oman (group C3) ndash the latterincluding all specimens of the subspecies A s syriacusThe diversification within the North African group isestimated to have started around 456 Mya (95 HPD282ndash647 Mya) The IsraelminusLebanon endemic subspe-cies A s syriacus is genetically highly distinct from

A s schreiberi and A s ataturi making A schreiberiparaphyletic (p-distance 12S 431 416 Cytb 1181202 respectively)

The networks constructed for the phased haplotypesof the full length nuclear markers (MC1R ACM4 andc-mos) are presented in Figure 3 The nuclear networkanalyses show similar results for each of the three genesand closely agree with the phylogenetic tree The CypriotA s schreiberi and Turkish A s ataturi subspecies sharealleles for all three genes and both are distinct fromthe third subspecies A s syriacus Acanthodactylusschreiberi syriacus shares no alleles with the other sub-species of A schreiberi but does share alleles withA b asper for each of the genes Acanthodactylusschreiberi syriacus shares MC1R alleles with A b asperspecimens from Tunisia Syria and Israel ACM4 alleleswith Egyptian Israeli Jordanian and North Africanspecimens and c-mos alleles only with Israeli A b asperspecimens Syrian A b asper samples share one allelewith A s syriacus and two with other A b asper speci-mens from Egypt Israel and North Africa in the MC1Rone allele with an Egyptian A b asper in the ACM4and none in the c-mos gene

In order to better understand the relationships betweenA schreiberi and A boskianus we performed three to-pology tests in which we forced monophyletic group-ings (1) monophyly of A schreiberi (all three subspeciestogether) (2) monophyly of A b asper (3) monophylyof A b asper and of A schreiberi The results of thetopological tests indicate that our data set cannot rejectthe alternative hypotheses of monophyly of A schreiberi(AU P = 0091 SH P = 0062) and that of A b asper(AU P = 011 SH P = 0072) if we allow A schreiberito nest within A b asper or a monophyletic A b aspernesting within A schreiberi When forcing monophylyof both A schreiberi and of A b asper together in thesame tree the results are inconclusive (AU P = 0046SH P = 0051)

The single-threshold model in GMYC yield a topol-ogy that is clearly different from the known taxono-my The GMYC results present a total of 25 and 24ML independent lineages from the Bayesian haplotypemitochondrial phylogeny of the two species for the two

Figure 2 Maximum likelihood (ML) tree of the Acanthodactylus boskianus and Acanthodactylus schreiberi specimensinferred using 12S rRNA cytochrome b mtDNA and melano-cortin 1 receptor acetylcholinergic receptor M4 and oocytematuration factor MOS nuclear gene fragments Posterior probability in the Bayesian analysis is indicated by black dotson the nodes [values ge 095 shown for both gene partitions and partitions by PartitionFinder (PF)] and ML bootstrapsupport values are indicated in parentheses (values ge 70 shown ML ML-PF) Age estimates with BEAST are indicat-ed near the relevant nodes and include the mean and in brackets the HPD 95 confidence interval Sample codes relateto specimens in Table 1 and in Figures 1 and 3 Colours blue Acanthodactylus boskianus asper yellow Acanthodactylusschreiberi ataturi red Acanthodactylus schreiberi syriacus green Acanthodactylus schreiberi schreiberi (Colour versionof figure available online)

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 11

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

12 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

partition approaches (ie by gene and by PartitionFinderFigs S2 S3 respectively) The two partition ap-proaches gave similar clusters but at the less sup-ported nodes they differed at the positions of severallineages The single threshold GMYC result is indi-cated for a single line at 00037 Mya for the gene par-titions and at 002 Mya for PartitionFinder (verticallines in Figs S2 S3) The topology and clusters re-vealed in this analysis correspond to the lineages fromthe phylogeny of the ML and BI methods both for theparaphyly of the two species and the geographical group-ings within A b asper The GMYC results mainly differfrom the ML and BI methods in the position ofA schreiberi from Cyprus and Turkey as a sister cladeto the Syrian A b asper

DISCUSSION

We have provided a comprehensive and thorough as-sessment of the intraspecific phylogenetic relation-ships within A schreiberi and its closest relativeA b asper Our results based on mitochondrial andnuclear DNA data from 84 specimens across the entiredistribution range of A schreiberi and most of the dis-tribution range of A b asper reveal that A schreiberiis paraphyletic and nested entirely within theA boskianus subspecies

HISTORICAL BIOGEOGRAPHY

Acanthodactylus schreiberi is thought to comprisethree subspecies corresponding to three allopatricpopulations in Cyprus Turkey and IsraelminusLebanonThe Cypriot endemic nominotypical subspeciesA sc schreiberi and the Turkish subspeciesA sc ataturi cluster together (to form clade B Fig 2)nesting between A b asper clades This lineage is sisterto a clade of A b asper including A sc syriacus (cladeC Fig 2) We estimate the divergence time of theCypriotminusTurkish lineage of A schreiberi to have beenduring the late Miocene around 6 Mya although thereis no support for this split in the tree In other analy-ses using the whole genus this split is well support-ed in Bayesian analyses (K Tamar S Carranza RSindaco J Moravec JF Trape amp S Meriri unpubldata) Based on mitochondrial data Poulakakis et al(2013) found that both the Cypriot and Turkish sub-species are monophyletic and diverged from each other085 Mya (038ndash156 Mya) According to our results thisdate corresponds to an inner divergence of the

A sc schreiberi lineage rather than to the date at whichA sc schreiberi colonized Cyprus

The discrepancy in the phylogenetic relationship ofA sc schreiberi raises questions regarding the arrivalon Cyprus Cyprus originated with the raising of theTroodos Massif during the upper Cretaceous c 91 to88 Mya (Clube amp Robertson 1986 Mukasa amp Ludden1987) During the middle to late Miocene only a smallproportion of Cyprus was exposed above the Mediter-ranean (McCallum amp Robertson 1990 Robertson 1990)Towards the end of the Miocene sim596 Mya with theclosing of the passage between the Atlantic Ocean andthe Mediterranean basin the Messinian salinity crisisbegan (Krijgsman et al 1999) This resulted in thedrying up of much of the Mediterranean Sea and highsea-mounts emerged to form land bridges with thesurrounding land (Hsuuml et al 1977) By the end of theMiocene and early Pliocene sim533 Mya the passagewith the Atlantic Ocean reopened and the Mediterra-nean basin was refilled (Krijgsman et al 1999) Re-sulting from compressions raising and uplifting of thesurrounding areas towards and during the Pleisto-cene Cyprus was a complete emergent island (McCallumamp Robertson 1990) The possible connection of Cyprusto the mainland (ie to TurkeySyria) during theMessinian is debated as are suggestions of a land con-nection at later periods (Steininger amp Roumlgl 1984 Jolivetet al 2006 Bache et al 2012) Such a connection ifit existed could have provided access for terrestrialorganisms with poor overseas dispersal ability suchas lizards to colonize the island Several studies arguethat post-Messinian sea level changes are unlikely tohave formed connections between Cyprus and the main-land (Steininger amp Roumlgl 1984 Jolivet et al 2006) Thusour dating of the split between the Cypriot A schreiberiand A b asper at c 6 Mya leads us to suggest that theancestor of A s schreiberi colonized Cyprus from themainland through a land bridge connection at the be-ginning of the Messinian crisis rather than by a muchlatermore recent transmarine dispersal as suggestedby Poulakakis et al (2013) Owing to its close rela-tions with A b asper the ancestor of A schreiberi waspresumably mainland A boskianus and the cladogenesisleading to A schreiberi thus rendered A b asperparaphyletic

The Turkish subspecies A schreiberi ataturi was rec-orded for the first time by Franzen (1998) at a veryrestricted area of around 15 km of coastal strip (betweenBotas and Yukarı Burnaz Hatay Province) Owing tothe remarkable morphological similarity between

Figure 3 Haplotype networks of the nuclear gene fragments melano-cortin 1 receptor (MC1R) acetylcholinergic recep-tor M4 (ACM4) and oocyte maturation factor MOS (c-mos) with colours corresponding to Figures 1 and 2 Codes corre-late to the two alleles (ie a and b) of specimens in Table 1 Circle sizes are proportional to the number of alleles (Colourversion of figure available online)

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 13

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

A sc ataturi and the Cypriot population the speci-mens were initially identified as A sc schreiberi(Franzen 1998) Yalccedilinkaya amp Goumlccedilmen (2012) howeverdescribed this population as a new distinct subspe-cies A s ataturi presenting several differences betweenthe two in both morphology and blood-serum pro-teins The origin of A s ataturi remains uncertain asit is debated whether the newly discovered Turkishpopulation is a relict or an introduction from CyprusFranzen (1998) described this population as a pos-sible introduction from Cyprus through the harbourof Botas but Sindaco et al (2000) suggested thatit might be a relict of a previously larger popula-tion because its present distribution is similar tothat of some insects and lizards [Archaeolacerta(Phoenicolacerta) laevis and Ablepharus budaki]Yalccedilinkaya amp Goumlccedilmen (2012) proposed that A sc ataturiarrived in Turkey from the nominate population inCyprus during the Messinian crisis The phylogeneticresults haplotype networks and low levels of geneticdivergence we found suggest that the two subspeciesfrom Cyprus and Turkey have not been geneticallyisolated for a long period of time (ie they share allelesin all three nuclear genes and A s ataturi is nestedwithin A s schreiberi in the phylogeny Figs 2 3) Ourresults therefore contrast with the two latter sce-narios of a relict population or a Messinian disper-sal Both divergence time and the genetic similarityof the two subspecies agree with the original sugges-tion of Franzen (1998) that these animals were intro-duced into Turkey from Cyprus Further support forthis hypothesis is that A s ataturi is restricted to thevicinity of the Botas-Adana harbour and is absent inother suitable habitats (coastal sand dunes) wide-spread in south-eastern Turkey Its close morphologi-cal features to A s schreiberi (Franzen 1998) likewisesupport an introduction scenario

The third subspecies A s syriacus is nested withinA b asper in the concatenated mtDNA and nDNA treesand is clearly genetically distinct from the Cyprus andTurkey A schreiberi lineage The close relations ofA s syriacus with A boskianus may shed light on theorigin of the former Acanthodactylus schreiberi syriacusis distributed on stable sands of the coastal plain ofthe eastern Mediterranean in Israel and southernLebanon (Salvador 1982 Hraoui-Bloquet et al 2002Bar amp Haimovitch 2011) habitats resembling those ofA s schreiberi from Cyprus (Baier Sparrow amp Wiedl2009) The oldest divergence of the A b asper cladethat includes A s syriacus is estimated to have oc-curred during the late Miocene around 558 Mya butno further dates are available for the grouping ofA s syriacus as a result of low support values Thecoastal plain of the eastern Mediterranean was sub-merged during the late Miocene and re-emerged onlytoward the Pliocene (Nir 1970 Horowitz 1979) The

sands of the coastal plain where A s syriacus occurs(Salvador 1982 Arnold 1983 K Tamar amp S Meiripers observ) were repeatedly submerged and re-emerged during the Pleistocene sea-level changes (duringinterglacial and glacial periods respectively) A pos-sible scenario for A sc syriacusrsquos origin includes severalwaves of dispersal of Middle Eastern A boskianus whichoccurs on coarse substrates (Amitai amp Bouskila 2001Disi et al 2001 Baha El Din 2006 pers observ) towardthe Mediterranean shore Acanthodactylus boskianusasper is absent from Mediterranean climate habitatsin Lebanon and Israel It occupies only xeric zonessuggesting an invasion to the coastal plain when sandyhabitats allowed desert flora and fauna to migrate north-wards (Yom-Tov 1988) These populations adapted tosandy soils and evolved morphological features thatdistinguish them from the desert hard substrate formsof A b asper We view this as the most likely scenariogiven the biogeography the phylogenetic results andthe habitat preferences and adaptations of these lizardsAn alternative scenario according to which the an-cestor of A schreiberi originated in Cyprus and dis-persed to the shores of Israel and Lebanon (or originatedin the coastal plain of the Eastern Mediterranean anddispersed to Cyprus) we regard as far less likely Sucha scenario requires much closer genetic relationshipsbetween these two forms and is further weakened bythe close relationship between A s syriacus and thegeographically adjacent A b asper populations

Acanthodactylus boskianus asper is highly variableboth morphologically (Salvador 1982 Arnold 1983) andgenetically (this study) The subspecies is paraphyleticas A schreiberi is nested within it The topology of theA b asper tree shows four different geographical group-ings Syria (clade A) north Jordan plus north OmanNorth Africa and Middle East plus south Arabia (groupsC1 C2 C3 respectively) The different groups in thissubspecies are estimated to have first diverged duringthe late Miocene approximately 65 Mya with the splitof the Syrian population The Syrian lineage is ge-netically distant from A schreiberi and the otherA b asper specimens The nuclear networks indicatethat this group is closer to the other A b asper samplesrather than to A schreiberi The geographical splitsin the rest of the A b asper range (clade C) are esti-mated to have started around 558 Mya These groupsare supported as a distinct clade but are closely relatedto each other in both the concatenated and nuclear trees(Figs 2 S1 respectively) The diversification within thisclade is estimated to have occurred during the lateMiocene to early Pliocene when A b asper dispersedwidely west to North Africa and in Arabia The di-vergence within the North African group (group C2)is estimated to have occurred during the Pliocene ap-proximately 456 Mya with the Egyptian Nigerian andSudanese populations later dispersing west and north

14 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

in Africa This diversification correlates to the arid climatestarting in southern Sahara during the early-mid Plio-cene and later in northern Africa between the Plio-cene and the Pleistocene (Le Houeacuterou 1997) as hasbeen suggested for the dispersal of Mesalina guttulatain Africa (Kapli et al 2008) Other evidence relatesdry climate in North Africa to an earlier period around7 Mya (Schuster et al 2006) as has been suggestedfor the genus Chalcides and other reptiles (Carranzaet al 2008 Metallinou et al 2012 and reference therein)The aridification of North Africa has most likely con-tributed for the successful dispersal of A b asper westfrom south-west Asia into Africa Morphological studiesof A boskianus show relatively uniform populations inNorth Africa suggesting recent migration (Salvador1982 Arnold 1983) The other two geographical group-ings of A b asper from the Middle East and Arabia(groups C1 and C3) are located in two distinct innerclades but their location within each inner clade ispoorly supported The topology of the concatenated tree(Fig 2) shows that the group from northern Jordanand northern Oman (group C1) is closer to the NorthAfrican one than to the geographically close Middle-Eastern and south Arabian group (group C3) The taxo-nomic separation between north and south Oman hasbeen recognized in other species of reptiles and sup-ported by the topography of Oman (eg Echis coloratusand Echis omanensis Arnold Robinson amp Carranza2009) In the nuclear tree (Fig S1) these two groupsare closer to one another and with the North Africangroup form clade C Therefore the low support valuesamongst these groupings prevent an appropriate andthorough analysis of this subspecies The close rela-tionship amongst the geographical groups may reflectclose phylogenetic relationships amongst these popu-lations suggesting recent migration divergence andongoing gene flow

SYSTEMATICS AND TAXONOMIC IMPLICATIONS

The relationships within the A boskianus species groupconflict with the current known taxonomy of A schreiberiand A b asper (samples of the other subspecies ofA boskianus and of A nilsoni were unavailable for thisstudy) Both species have been found to be closelyrelated and paraphyletic The constrained topology testsexemplify the close entangled relationship between thetwo species as the separate monophyly of the two specieswas not rejected and the enforced monophyly of themboth together was inconclusive

Several causes can be responsible for paraphyly inspecies such in the A boskianus species group (Funkamp Omland 2003 and references therein) (1) inad-equate phylogenetic information (2) imperfect taxono-my (incorrectinaccurate species limits) derived frommisidentifying intraspecific variation (3) interspecific

gene flow ndash hybridization through interspecific matingand the subsequent backcrossing of hybrids into theparental populations (4) incomplete lineage sortingbecause of recent speciation events (5) unrecognizedparalogy We suggest that the relationships betweenA schreiberi and A b asper based on mitochondrialand nuclear data are most likely explained by incor-rect taxonomy probably because of the great variabil-ity of the latter species and to convergence As wasthe case in the molecular studies of the A pardalisand A erythrurus species groups (Harris et al 2004Fonseca et al 2008 2009 Carretero et al 2011 andreference therein) there are many problems with thecurrent taxonomic status of several species groups withinAcanthodactylus

Taking the molecular results of our study into accountthere are several systematic approaches to classify-ing the A schreiberiminusA b asper clade The Cypriot andTurkish populations of A schreiberi are very closelyrelated with the latter nested in the former and thetwo subspecies share nuclear alleles (Fig 3) Further-more the low uncorrected p-distance is positively cor-related with subspecies-level distances within otherlacertid species (ie 16 of Cytb in Lacerta bilineatachloronota Godinho et al 2005) We therefore con-clude that Cypriot animals were recently introducedto Turkey and that the Turkish population does notmerit a subspecific rank We suggest that A s ataturiYalccedilinkaya amp Goumlccedilmen 2012 is a junior synonym ofA s schreiberi Boulenger 1878

Regarding the relationships between A schreiberi andA b asper a few scenarios are possible One is to sinkA schreiberi within A boskianus to create one species(A boskianus) with high genetic and morphological vari-ability ranging over a broad distribution Another isfor the two taxa be regarded as a species complex (theA boskianus-schreiberi complex) until further inves-tigation on the subject However although A schreiberiis nested within A b asper the populations from Cyprusand Turkey represent a distinct evolutionary lineagewith distinct genetic and morphological features andthus it is logical to retain the specific status Two othersolutions are possible The first is to re-evaluate theSyrian populations and to consider elevating them aswell as the more divergent lineages (and subspecies)of A boskianus to specific status This would neces-sitate an examination of the phylogeny and morphol-ogy of the other four subspecies of A boskianus(A b boskianus A b euphraticus A b khattensis andA b nigeriensis) and the identification of distinctivephenotypic features in the Syrian lizards Another so-lution is to recognize the maintenance of gene flowamongst mainland populations of A b asper after thedivergence of the insular endemic A schreiberi andthus the evolutionary cohesion of the paraphyleticA b asper Arnold (1983) noted that A schreiberi may

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 15

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

have originated as an isolate from A boskianus becauseof their shared morphology and hemipenis featuresOur results support this scenario which includes thedispersal of A schreiberi to Cyprus from a mainlandpopulation that was most probably A boskianus It maybe assumed that the ancestor of the Cypriot A schreiberiafter arriving on Cyprus remained isolated for a longperiod of time and thus evolved to the modern formof A schreiberi Meanwhile the same ancestral con-tinental populations not isolated from each other con-tinued to exchange genes to varying degrees remainingA boskianus

The IsraeliminusLebanese subspecies A sc syriacus is onlydistantly related to the nominate form A sc schreiberiThis subspecies is highly phylogenetically divergent fromthe Cypriot and Turkish populations having higherp-distances (12S 4 Cytb 11ndash12) than those foundbetween other lacertid species (eg 74ndash82 of Cytbamongst Iberolacerta aranica Iberolacerta aurelioi andIberolacerta bonnali and 41ndash58 of Cytb betweenLacerta bilineata and Lacerta viridis Crochet et al2004 Godinho et al 2005 respectively) The nuclearhaplotype networks further show that Lebanese andIsraeli populations share alleles only with A b asperbut not with the nominotypical Cypriot form Arnold(1983) suggested that the geographical variation ofA boskianus reflects niche differences with animalsfrom xeric areas with dense rigid and spiny vegeta-tion having larger dorsal scales than animals from moremesic areas As was assumed for A schreiberi wesuggest that other mainland populations of A b asperwere the ancestors of the LebaneseminusIsraeli Coastal plainforms We suggest that A s syriacus is an ecomorphof A b asper that dispersed from the usual xerichabitats of the species and adapted to the new moremesic environment of the stable sands of the coastalplains of the eastern Mediterranean As a conse-quence this ecomorph converged on the morphologyof A s schreiberi which inhabits the coastal sands ofCyprus (Baier et al 2009) but still maintains differ-entiating features by having coarser dorsal scales andsharp keels (Salvador 1982 Arnold 1983 Franzen1998) This convergence led to the description ofA s syriacus as a member of A schreiberi The mor-phological assessment and the close morphological simi-larities between A b asper and A s syriacus mayexplain the wrong classification A similar erro-neous reasoning led Reed amp Marx (1959) to identifyspecimens with fine scales from Iraq as A schreiberiSalvador (1982) re-examined these specimens and as-signed them to A boskianus The morphological dif-ferences between the two forms are less prominentespecially where the two forms occur in close geo-graphical proximity in the southern coastal plainand north-western Negev Desert of Israel (Bar ampHaimovitch 2011) According to our results A s syriacus

actually belongs to A b asper being a coastal-duneecomorph convergent with but evolutionarily dis-tinct from A schreiberi Thus our preferred scenariois to treat the name Acanthodactylus schreiberi syriacusBoumlttger 1879 (which was originally described asA boskianus var syriacus by Boumlttger 1879) as a juniorsynonym of the name Acanthodactylus boskianus asper(Audouin 1827)

Recognizing A s syriacus as a junior synonym ofA b asper may have important implications for the con-servation of this coastal sand dune form which is clas-sified as critically endangered in Israel (Dolev ampPervolutzki 2004) However as the Israeli and Leba-nese coastal dune ecosystem has probably developedonly very recently during the Quaternary (Nir 1970Horowitz 1979) this form represents a remarkable caseof rapid evolutionary change It is also a remarkablecase of convergent evolution (with the CypriotA sc schreiberi) Thus we feel that these popula-tions are unique evolutionary entities that merit specialconservation efforts

The use of nuclear genes is a valuable method forestimating species divergence and lineage sorting andhelps evaluate isolated lineages and evolutionary historyThe incorporation of mitochondrial and nuclear dataprovides thorough topologies informative networks anddivergence times that reveal useful information for aproblematic taxonomy such as that of the A boskianusspecies group We have shown that phylogenetic ap-proaches to the confusing taxonomy of two closely relatedand morphologically similar species can shed light ontheir unclear relationships resolve between homoplasyand shared ancestry and identify patterns of speciesevolutionary history and biogeography

ACKNOWLEDGEMENTS

We are grateful to the following people for providingsamples for this study D Donaire B Shacham J ŠmiacutedS M Baha El Din and J Padial We thank MMetallinou and J Šmiacuted for help with the lab work andthe analyses M Novosolov for help with the figuresY L Werner B Shacham and A Levy for fruitful dis-cussions and two anonymous referees for helpful com-ments on an earlier version of this manuscript Wethank the Israel nature and park authority for issuingcollecting permits (nos 38074 38451 38489 38540)The work of J M was financially supported by the Min-istry of Culture of the Czech Republic (DKRVO 201315 National Museum 00023272) S C is funded bygrant CGL2012-36970 from the Ministerio de Economiacuteay Competitividad Spain (cofunded by FEDER) Thisstudy was funded by a Israel Taxonomic Initiative grantto S M K T is supported by an Israel Taxonomic Ini-tiative scholarship

16 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

REFERENCES

Ahmadzadeh F Carretero MA Harris DJ Perera ABohme W 2012 A molecular phylogeny of the eastern groupof ocellated lizard genus Timon (Sauria Lacertidae) basedon mitochondrial and nuclear DNA sequences Amphibia-Reptilia 33 1ndash10

Ahmadzadeh F Flecks M Roumldder D Boumlhme W Ilgaz CcedilHarris DJ Engler JO Uumlzuumlm N Carretero MA 2013Multiple dispersal out of Anatolia biogeography and evolu-tion of oriental green lizards Biological Journal of the LinneanSociety 110 398ndash408

Akaike H 1973 Information theory and an extension of themaximum likelihood principle In Petrov BN Csaki F edsSecond International Symposium on Information Theory Bu-dapest Akademiai Kiado 267ndash281

Amitai P Bouskila A 2001 Handbook of amphibians ampreptiles of Israel Israel Keter Publishing House Ltd

Anderson SC 1999 The lizards of Iran Ithaca NY Societyfor the Study of Amphibians and Reptiles

Arnold EN 1980 The scientific results of the Oman flora andfauna survey 1977 (Dhofar) The reptiles and amphibians ofDhofar southern Arabia Journal of Oman Studies SpecialReport 2 273ndash332

Arnold EN 1983 Osteology genitalia and the relationshipsof Acanthodactylus (Reptilia Lacertidae) Bulletin of the BritishMuseum (Natural History) Zoology 44 291ndash339

Arnold EN Arribas Oacute Carranza S 2007 Systematics ofthe Palaearctic and Oriental lizard tribe Lacertini (SquamataLacertidae Lacertinae) with descriptions of eight new generaZootaxa 1430 1ndash86

Arnold EN Robinson MD Carranza S 2009 A prelimi-nary analysis of phylogenetic relationships and biogeogra-phy of the dangerously venomous Carpet Vipers Echis(Squamata Serpentes Viperidae) based on mitochondrial DNAsequences Amphibia-Reptilia 30 273ndash282

Audouin JV 1827 Explication sommaire des planches dereptiles (suppleacutement) offrant un exposeacute des characteacuteresdes espegraveces In Savigny MJCL ed Description de lrsquoEacutegypteVol 1 Histoire Naturelle Paris Impeacuteriale 161ndash184

Bache F Popescu SM Rabineau M Gorini C Suc JPClauzon G Olivet JL Rubino JL Melinte-DobrinescuMC Estrada F 2012 A two-step process for the refloodingof the Mediterranean after the Messinian Salinity Crisis BasinResearch 24 125ndash153

Baha El Din SM 2006 A guide to the reptiles and amphib-ians of Egypt CairondashNew York American University in CairoPress

Baier FS Sparrow DJ Wiedl H-J 2009 The amphibiansand reptiles of Cyprus Frankfurt am Main Edition Chimaira

Bandelt H-J Forster P Roumlhl A 1999 Median-joining net-works for inferring intraspecific phylogenies Molecular Biologyand Evolution 16 37ndash48

Bar A Haimovitch G 2011 A field guide to reptiles and am-phibians of Israel Herzliya Pazbar Limited

Boumlttger O 1879 Reptilien und Amphibien aus Syrien Berichtuumlber die Senckenbergische Naturforschende Gesellschaft inFrankfurt am Main 1879 57ndash84

Boulenger GA 1919 On a new variety of Acanthodactylusboskianus Daud from the Euphrates The Annals and Maga-zine of Natural History 3 549ndash550

Boulenger GA 1878 Sur les espegraveces drsquoAcanthodactylus desbords de la Mediterraneacutee Bulletin de la Socieacuteteacute zoologiquede France 3 179ndash197

Boulenger GA 1918 Sur les leacutezards du genre AcanthodactylusWieg Bulletin de la Socieacuteteacute zoologique de France 43 143ndash155

Boulenger GA 1921 Monograph of the Lacertidœ Vol II Trus-tees of the British Museum of Natural History London

Carranza S Arnold EN 2012 A review of the geckos of thegenus Hemidactylus (Squamata Gekkonidae) fromOman based on morphology mitochondrial and nuclear datawith descriptions of eight new species Zootaxa 3378 1ndash95

Carranza S Arnold EN Geniez P Roca J Mateo J 2008Radiation multiple dispersal and parallelism in the skinksChalcides and Sphenops (Squamata Scincidae) with com-ments on Scincus and Scincopus and the age of the SaharaDesert Molecular Phylogenetics and Evolution 46 1071ndash1094

Carretero MA Fonseca MM Garcia-Munoz E Brito JCHarris DJ 2011 Adding Acanthodactylus beershebensis tothe mtDNA phylogeny of the Acanthodactylus pardalis groupNorth-Western Journal of Zoology 7 138ndash142

Castresana J 2000 Selection of conserved blocks from multi-ple alignments for their use in phylogenetic analysis Mo-lecular Biology and Evolution 17 540ndash552

Clube TMM Robertson A 1986 The palaeorotation of theTroodos microplate Cyprus in the Late Mesozoic-Early Ce-nozoic plate tectonic framework of the Eastern Mediterra-nean Surveys in Geophysics 8 375ndash437

Cox SC Carranza S Brown RP 2010 Divergence timesand colonization of the Canary Islands by Gallotia lizardsMolecular Phylogenetics and Evolution 56 747ndash757

Crochet PA Geniez P Ineich I 2003 A multivariate analy-sis of the fringe-toed lizards of the Acanthodactylus scutellatusgroup (Squamata Lacertidae) systematic and biogeographi-cal implications Zoological Journal of the Linnean Society137 117ndash155

Crochet P-A Chaline O Surget-Groba Y Debain CCheylan M 2004 Speciation in mountains phylogeographyand phylogeny of the rock lizards genus Iberolacerta (ReptiliaLacertidae) Molecular Phylogenetics and Evolution 30 860ndash866

Daudin FM 1802 Histoire naturelle geacuteneacuterale et particuliegraveredes reptiles ouvrage faisant suite agrave lrsquoHistoire naturelle geacuteneacuteraleet particuliegravere F Dufart

Disi A Modry D Necas P Rifai L 2001 Amphibians andreptiles of the Hashemite Kingdom of Jordan an atlas andfield guide Frankfurt am Main Chimaira

Dolev A Pervolutzki A 2004 Endangered species in IsraelRed list of threatened animals Vertebrates JerusalemThe Nature and Parks Authority and the Society for thePreservation of Nature

Drummond AJ Rambaut A 2007 BEAST Bayesian evo-lutionary analysis by sampling trees BMC EvolutionaryBiology 7 214

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 17

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Ezard T Fujisawa T Barraclough T 2009 Splits speciesrsquolimits by threshold statistics R package version 10-19r48Available at httpR-ForgeR-projectorgprojectssplits

Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783ndash791

Fitzinger LJFJ 1834 Acanthodactylus In Wiegman AFAed Herpetologia mexicana seu Descriptio amphibiorum NovaeHispaniae quae itineribus comitis De Sack Ferdinandi Deppeet Chr Guil Schiede in Museum zoologicum Berolinensepervenerunt Pars prima saurorum species amplectens adjectosystematis saurorum prodromo additisque multis in huncamphibiorum ordinem observationibus Berlin C G Luderitz10

Flot JF 2010 SeqPHASE a web tool for interconvertingPHASE inputoutput files and FASTA sequence align-ments Molecular Ecology Resources 10 162ndash166

Fonseca MM Brito JC Paulo OS Carretero MA HarrisDJ 2009 Systematic and phylogeographical assessment ofthe Acanthodactylus erythrurus group (Reptilia Lacertidae)based on phylogenetic analyses of mitochondrial and nuclearDNA Molecular Phylogenetics and Evolution 51 131ndash142

Fonseca MM Brito JC Rebelo H Kalboussi M LarbesS Carretero MA Harris DJ 2008 Genetic variation amongspiny-footed lizards in the Acanthodactylus pardalis groupfrom North Africa African Zoology 43 8ndash15

Fontaneto D Herniou EA Boschetti C Caprioli M MeloneG Ricci C Barraclough TG 2007 Independently evolv-ing species in asexual bdelloid rotifers PLoS Biology 5 e87

Franzen M 1998 Erstnachweis von Acanthodactylus schreiberischreiberi Boulenger 1879 fuumlr die Tuumlrkei (Squamata SauriaLacertidae) Herpetozoa 11 27ndash36

Funk DJ Omland KE 2003 Species-level paraphyly andpolyphyly frequency causes and consequences with in-sights from animal mitochondrial DNA Annual Review ofEcology Evolution and Systematics 34 397ndash423

Godinho R Crespo EG Ferrand N Harris DJ 2005 Phy-logeny and evolution of the green lizards Lacerta spp(Squamata Lacertidae) based on mitochondrial and nuclearDNA sequences Amphibia-Reptilia 26 271ndash285

Greenbaum E Villanueva CO Kusamba C Aristote MMBranch WR 2011 A molecular phylogeny of EquatorialAfrican Lacertidae with the description of a new genus andspecies from eastern Democratic Republic of the Congo Zoo-logical Journal of the Linnean Society 163 913ndash942

Guillou H Carracedo JC Torrado FP Badiola ER 1996K-Ar ages and magnetic stratigraphy of a hotspot-inducedfast grown oceanic island El Hierro Canary Islands Journalof Volcanology and Geothermal Research 73 141ndash155

Harris DJ Arnold EN 2000 Elucidation of the relation-ships of spiny-footed lizards Acanthodactylus spp (ReptiliaLacertidae) using mitochondrial DNA sequence with com-ments on their biogeography and evolution Journal of Zoology252 351ndash362

Harris DJ Batista V Carretero M 2004 Assessment ofgenetic diversity within Acanthodactylus erythrurus (ReptiliaLacertidae) in Morocco and the Iberian Peninsula usingmitochondrial DNA sequence data Amphibia-Reptilia 25227

Horowitz A 1979 The Quaternary of Israel New York Aca-demic Press

Hraoui-Bloquet S Sadek RA Sindaco R Venchi A 2002The herpetofauna of Lebanon new data on distributionZoology in the Middle East 27 35ndash46

Hsuuml KJ Montadert L Bernoulli D Cita MB EricksonA Garrison RE Kidd RB Megraveliereacutes F Muumlller C WrightR 1977 History of the Mediterranean salinity crisis Nature267 399ndash403

Huelsenbeck JP Rannala B 2004 Frequentist propertiesof Bayesian posterior probabilities of phylogenetic trees undersimple and complex substitution models Systematic Biology53 904ndash913

Huelsenbeck JP Ronquist F 2001 MRBAYES Bayesianinference of phylogenetic trees Bioinformatics 17 754ndash755

Jolivet L Augier R Robin C Suc J-P Rouchy JM 2006Lithospheric-scale geodynamic context of the Messinian sa-linity crisis Sedimentary Geology 188 9ndash33

Kapli P Lymberakis P Poulakakis N Mantziou GParmakelis A Mylonas M 2008 Molecular phylogeny ofthree Mesalina (Reptilia Lacertidae) species (M guttulataM brevirostris and M bahaeldini) from North Africa and theMiddle East another case of paraphyly MolecularPhylogenetics and Evolution 49 102ndash110

Katoh K Toh H 2008 Recent developments in the MAFFTmultiple sequence alignment program Briefings inBioinformatics 9 286ndash298

Krijgsman W Hilgen F Raffi I Sierro F Wilson D 1999Chronology causes and progression of the Messinian salin-ity crisis Nature 400 652ndash655

Lanfear R Calcott B Ho SY Guindon S 2012PartitionFinder combined selection of partitioning schemesand substitution models for phylogenetic analyses Molecu-lar Biology and Evolution 29 1695ndash1701

Le Houeacuterou HN 1997 Climate flora and fauna changes inthe Sahara over the past 500 million years Journal of AridEnvironments 37 619ndash647

Maca-Meyer N Carranza S Rando J Arnold E CabreraV 2003 Status and relationships of the extinct giant CanaryIsland lizard Gallotia goliath (Reptilia Lacertidae)assessed using ancient mtDNA from its mummifiedremains Biological Journal of the Linnean Society 80 659ndash670

McCallum J Robertson A 1990 Pulsed uplift of the TroodosMassif ndash evidence from the Plio-Pleistocene Mesaoria basinIn J Malpas EMM Panayiotou A Xenophontos C edsOphiolites oceanic crustal analogues Proceedings of theSymposium lsquoTroodos 1987rsquo Nicosia (The Geological SurveyDepartment Ministry of Agriculture and NaturalResources) 217ndash229

Metallinou M Arnold EN Crochet P-A Geniez P BritoJC Lymberakis P El Din SB Sindaco R Robinson MCarranza S 2012 Conquering the Sahara and Arabiandeserts systematics and biogeography of Stenodactylus geckos(Reptilia Gekkonidae) BMC Evolutionary Biology 12258

Monaghan MT Wild R Elliot M Fujisawa T Balke MInward DJ Lees DC Ranaivosolo R Eggleton P

18 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

parameters unlinked across partitions (Table 2) Twoindependent runs of 2 times 107 generations were carriedout with a sampling frequency of every 1000 genera-tions After examining the standard deviation of thesplit frequencies between the two runs and the po-tential scale reduction factor diagnostic burn-in wasperformed discarding the first 25 trees of each runand the remaining trees were combined in a majorityconsensus tree In both ML and BI alignment gaps weretreated as missing data and the nuclear gene se-quences were not phased Nodes were considered strong-ly supported if they received ML bootstrap values ge 70and posterior probability (pp) support values ge 095 (Wilcoxet al 2002 Huelsenbeck amp Rannala 2004)

A total of 59 haplotypes was identified amongst theA boskianus species group using 792 bp of the con-catenated 12S and Cytb data set (see Table 1) Haplotypenetworks were constructed for the three nuclear genesMC1R ACM4 and c-mos (only full-length sequences)SEQPHASE (Flot 2010) was used to convert the inputfiles and the software PHASE v 211 to resolve phasedhaplotypes (Stephens Smith amp Donnelly 2001 Stephensamp Scheet 2005) Default settings of PHASE were usedexcept for phase probabilities which were set as ge 07

All polymorphic sites with a probability of lt 07 werecoded in both alleles with the appropriate IUPAC am-biguity code The phased nuclear sequences were usedto generate median-joining networks using NET-WORKS v 4611 (Bandelt Forster amp Roumlhl 1999)

In order to assess alternative topologies betweenA schreiberi and A b asper topological constraints thatcould be statistically rejected were constructed We en-forced alternative topologies by hand and compared withthe unconstrained tree (best ML tree) using the ap-proximately unbiased (AU Shimodaira 2002) andShimodairaminusHasegawa (SH Shimodaira amp Hasegawa1999) tests Per-site log likelihoods were estimated inusing RAxMLGUI v 13 (Silvestro amp Michalak 2012)and P-values were calculated using CONSEL(Shimodaira amp Hasegawa 2001)

SPECIES DELIMITATION

In order to reveal the main lineages with the concat-enated analysis and as a prior for species groupingsa mitochondrial phylogeny of 59 haplotypes was per-formed with BEAST v 162 (Drummond amp Rambaut2007) without the outgroups Three individual runs were

Table 2 Information on the partitions used in the phylogenetic analyses with the different partition approaches (ie bygene and by PartitionFinder C codon) including the length model of sequence evolution selected by JModelTest andPartitionFinder and the results of the test of rate homogeneity (LRT) run in MEGA (see Material and methods)

Partition approach Partition Length (bp) Model LRT

By gene 12S sim387 GTR + I + G Not rejected (P lt 07396)Cytb 405 TrN + I + G Rejected (P lt 21819E-7)MC1R 663 GTR + I Not rejected (P lt 1)ACM4 429 HKY + I Not rejected (P lt 1)c-mos 522 TPM1uf + G Not rejected (P lt 1)

PartitionFinder ndashConcatenated

12S Cytb (C1) 2406 GTR + I + Gc-mos(C1) Cytb (C2) TrNef + I + GCytb (C3) TrN + I + GACM4 (C12) MC1R (C1) TrNMC1R (C2) F81MC1R (C3) HKY + GACM4 (C3) c-mos(C2 3) TrNef + I + G

PartitionFinder ndashmtDNA

12S Cytb (C1) 792 SYM + I + GCytb (C2) TrN + I + GCytb (C3) TrN + I + G

PartitionFinder ndashnuclear DNA

ACM4 (C12) c-mos (C12)MC1R (C1)

1614 HKY + I

MC1R (C2) F81MC1R (C3) HKY + GACM4 (C3) c-mos (C3) K80 + I

Gene abbreviations 12S 12S rRNA ACM4 acetylcholinergic receptor Muscarinic 4 c-mos oocyte maturation factor MOSCytb cytochrome b MC1R melano-cortin 1 receptorModel abbreviations F81 Felsenstein 1981 GTR general time-reversible HKY Hasegawa Kishino-Yano K80 Kimura1980 SYM symmetrical model TPM1uf Kimura three-parameter model TrN Tamura-Nei Any of these models caninclude invariable sites (+I) gamma distribution (+G) or both (+I+G)

8 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

performed for 5 times 107 generations with a sampling fre-quency of 10 000 The results were combined to inferthe ultrametric tree after discarding 10 of the samplesfrom each run Models and prior specifications appliedwere as follows (otherwise by default) for partitionsby genes and by PartitionFinder For gene partitionsGTR + I + G strict clock (12S) Hasegawa-Kishino-Yano + Invariable sites + Gamma distribution(HKY + I + G) strict clock molecular clock model (es-timate 0ndash1) (Cytb) coalescence constant size processof speciation random starting tree alpha Uniform (010) GTR Uniform For partitions by PartitionFinderGTR + I + G strict clock (partition 1 = 12S + Cytb codon1 and 2) Tamura-Nei + Gamma distribution (TrN + G)strict clock (partition 2 = Cytb codon 3) coalescenceconstant size process of speciation random starting treealpha Uniform (0 10) Parameter values both for clockand substitution models were unlinked across parti-tions For all analyses implemented in BEAST the threeruns were analysed in TRACER v 15 (Rambaut ampDrummond 2007) confirming convergence The treeswere combined in LogCombiner and TreeAnnotator(available in BEAST package) was used for the pro-duction of the final tree

For estimating species limits directly from the Bayes-ian phylogenetic tree produced with the concat-enated mitochondrial data we used the independentgeneralized mixed Yule-coalescent (GMYC) method (Ponset al 2006) The GMYC model estimated the numberof phylogenetic clusters or lsquospeciesrsquo by identifying theshifts between intraspecific (coalescence) and interspecific(diversification) branch rates (Pons et al 2006) Weperformed the GMYC function in the R v302 lsquosplitsrsquopackage (Ezard Fujisawa amp Barraclough 2009) Alikelihood-ratio test was used to determine if the GMYCmodel with a shift in the branching processes provid-ed a better fit to the data than the null model withno shifts We used a single threshold value (Monaghanet al 2009) which has already been applied success-fully to different groups of organisms (Pons et al 2006Fontaneto et al 2007 Monaghan et al 2009)

ESTIMATION OF DIVERGENCE TIMES

The lack of internal calibration points in Acantho-dactylus (no fossils are known) prevents the directestimation of time in our phylogeny Therefore we usedthe mean substitution rates and their standard errorof the same 12S and Cytb mitochondrial regions ex-tracted from a fully calibrated phylogeny of anotherlacertid group the lizards of the genus Gallotia endemicto the Canary Islands (Cox Carranza amp Brown 2010as was implemented in Carranza amp Arnold 2012) Theinferred calibration rate was estimated using the ageof El Hierro Island (Canary Islands) estimated at112 Mya (Guillou et al 1996) They assumed coloni-

zation of the island by members of the lacertid genusGallotia (Gallotia caesaris caesaris endemic to El HierroIsland) immediately after its formation from the neigh-bouring La Gomera Island (inhabited by the endemicGallotia caesaris gomerae) These two subspecies aremonophyletic sister taxa with low intraspecific vari-ability (Maca-Meyer et al 2003 Cox et al 2010) andthus suitable for calibration

For the estimation of divergence times one repre-sentative of each independent GMYC lineage was usedfrom the ultrametric tree (for the representatives seeTable 1) We used a likelihood-ratio test implementedin MEGA 52 (Tamura et al 2011) to test if the dif-ferent partitions (by genes) included in the dating analy-sis were evolving in a clock-like fashion (Table 2) Thisinformation was used to choose between the strict clockand the relaxed uncorrelated lognormal clock priorsimplemented in BEAST (Monaghan et al 2009) Thedata set included one representative from each lineagefrom the GMYC analysis using sequences from all fivepartitions (nuclear genes unphased) Three individ-ual runs were performed for 5 times 107 generations witha sampling frequency of 10 000 and the results werecombined to infer the ultrametric tree after discard-ing 10 of the samples from each run Models and priorspecifications applied were as follows (otherwise bydefault) GTR + I + G relaxed uncorrelated lognor-mal clock molecular clock model (estimate) (12S Cytb)HKY strict clock (MC1R c-mos) and TrN + I strictclock (ACM4) Yule process of speciation random start-ing tree yulebirthRate (0 1000) alpha Uniform (010) ucldmean of 12S Normal (initial value 000553mean 000553 SD 000128) ucldmean of Cytb Normal(initial value 00164 mean 00164 SD 000317) Pa-rameter values both for clock and substitution modelswere unlinked across partitions

RESULTS

The data set of this study is comprised of 19 samplesof A schreiberi 65 samples of A b asper and 11outgroup samples (Table 1 Fig 1) The data set in-cluded mitochondrial DNA (mtDNA) gene fragmentsof 12S (sim387 bp) and Cytb (405 bp) and nuclear DNA(nDNA) gene fragments of MC1R (663 bp) ACM4(429 bp) and c-mos (522 bp) totalling to sim2406 bp Thenumber of variable (V) and parsimony-informative(Pi) sites for the ingroup are listed in Table S1 Thetwo partition approaches (ie by gene and byPartitionFinder) gave similar results for both the MLand BI analyses The results of the phylogenetic analy-ses of the complete concatenated data set using MLand BI methods produced very similar topologies butdiffered to some extent at the less supported nodesat the intraspecific level (Fig 2) Separated analysesof the nuclear data sets are presented in Figure S1

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 9

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

10 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Together A b asper and A schreiberi form amonophyletic group within Acanthodactylus (Fig 2)Within the group however both taxa are paraphyleticwith A schreiberi as a whole nested within A b asperOur analyses distinguish three major clades (1) cladeA formed by A b asper from Syria (2) clade B in-cludes the two subspecies A sc ataturi from Turkeytogether with A sc schreiberi from Cyprus (3) cladeC which includes specimens of A b asper from the re-maining localities in its distribution range together withA sc syriacus from Israel and Lebanon Clade A is verywell supported and includes specimens of A b asperfrom central and northern Syria (Fig 1) splitting fromother specimens at the basal node of the group is es-timated to have occurred c 654 Mya [95 highest pos-terior density (HPD) 392ndash952 Mya] The level of geneticdifferentiation(p-distance) between these specimens and the remain-ing A b asper and all A schreiberi specimens is 37ndash46 for 12S and 107ndash119 for Cytb Clade B is alsovery well supported and includes two of the threenominal subspecies of A schreiberi A s schreiberi thenominotypical subspecies endemic to Cyprus and A sataturi from Turkey The Turkish subspecies is nestedwithin the Cypriot specimens and the two forms havelow genetic distances from each other (12S 016 Cytb123) This clade is nested between the two A b asperclades (clades A and C) in both the concatenatedand the nuclear tree although the nodes are not wellsupported Clade C is not very well supported It in-cludes a cluster of A b asper and A s syriacus Thisclade includes two inner clades that split around558 Mya (95 HPD 356ndash8 Mya) and divided into threepoorly supported geographical inner groups (Fig 2)northern Jordan and northern Oman (group C1) NorthAfrica (group C2) and samples from the Middle East(Egypt south Israel and south Jordan) with samplesfrom Yemen and southern Oman (group C3) ndash the latterincluding all specimens of the subspecies A s syriacusThe diversification within the North African group isestimated to have started around 456 Mya (95 HPD282ndash647 Mya) The IsraelminusLebanon endemic subspe-cies A s syriacus is genetically highly distinct from

A s schreiberi and A s ataturi making A schreiberiparaphyletic (p-distance 12S 431 416 Cytb 1181202 respectively)

The networks constructed for the phased haplotypesof the full length nuclear markers (MC1R ACM4 andc-mos) are presented in Figure 3 The nuclear networkanalyses show similar results for each of the three genesand closely agree with the phylogenetic tree The CypriotA s schreiberi and Turkish A s ataturi subspecies sharealleles for all three genes and both are distinct fromthe third subspecies A s syriacus Acanthodactylusschreiberi syriacus shares no alleles with the other sub-species of A schreiberi but does share alleles withA b asper for each of the genes Acanthodactylusschreiberi syriacus shares MC1R alleles with A b asperspecimens from Tunisia Syria and Israel ACM4 alleleswith Egyptian Israeli Jordanian and North Africanspecimens and c-mos alleles only with Israeli A b asperspecimens Syrian A b asper samples share one allelewith A s syriacus and two with other A b asper speci-mens from Egypt Israel and North Africa in the MC1Rone allele with an Egyptian A b asper in the ACM4and none in the c-mos gene

In order to better understand the relationships betweenA schreiberi and A boskianus we performed three to-pology tests in which we forced monophyletic group-ings (1) monophyly of A schreiberi (all three subspeciestogether) (2) monophyly of A b asper (3) monophylyof A b asper and of A schreiberi The results of thetopological tests indicate that our data set cannot rejectthe alternative hypotheses of monophyly of A schreiberi(AU P = 0091 SH P = 0062) and that of A b asper(AU P = 011 SH P = 0072) if we allow A schreiberito nest within A b asper or a monophyletic A b aspernesting within A schreiberi When forcing monophylyof both A schreiberi and of A b asper together in thesame tree the results are inconclusive (AU P = 0046SH P = 0051)

The single-threshold model in GMYC yield a topol-ogy that is clearly different from the known taxono-my The GMYC results present a total of 25 and 24ML independent lineages from the Bayesian haplotypemitochondrial phylogeny of the two species for the two

Figure 2 Maximum likelihood (ML) tree of the Acanthodactylus boskianus and Acanthodactylus schreiberi specimensinferred using 12S rRNA cytochrome b mtDNA and melano-cortin 1 receptor acetylcholinergic receptor M4 and oocytematuration factor MOS nuclear gene fragments Posterior probability in the Bayesian analysis is indicated by black dotson the nodes [values ge 095 shown for both gene partitions and partitions by PartitionFinder (PF)] and ML bootstrapsupport values are indicated in parentheses (values ge 70 shown ML ML-PF) Age estimates with BEAST are indicat-ed near the relevant nodes and include the mean and in brackets the HPD 95 confidence interval Sample codes relateto specimens in Table 1 and in Figures 1 and 3 Colours blue Acanthodactylus boskianus asper yellow Acanthodactylusschreiberi ataturi red Acanthodactylus schreiberi syriacus green Acanthodactylus schreiberi schreiberi (Colour versionof figure available online)

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 11

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

12 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

partition approaches (ie by gene and by PartitionFinderFigs S2 S3 respectively) The two partition ap-proaches gave similar clusters but at the less sup-ported nodes they differed at the positions of severallineages The single threshold GMYC result is indi-cated for a single line at 00037 Mya for the gene par-titions and at 002 Mya for PartitionFinder (verticallines in Figs S2 S3) The topology and clusters re-vealed in this analysis correspond to the lineages fromthe phylogeny of the ML and BI methods both for theparaphyly of the two species and the geographical group-ings within A b asper The GMYC results mainly differfrom the ML and BI methods in the position ofA schreiberi from Cyprus and Turkey as a sister cladeto the Syrian A b asper

DISCUSSION

We have provided a comprehensive and thorough as-sessment of the intraspecific phylogenetic relation-ships within A schreiberi and its closest relativeA b asper Our results based on mitochondrial andnuclear DNA data from 84 specimens across the entiredistribution range of A schreiberi and most of the dis-tribution range of A b asper reveal that A schreiberiis paraphyletic and nested entirely within theA boskianus subspecies

HISTORICAL BIOGEOGRAPHY

Acanthodactylus schreiberi is thought to comprisethree subspecies corresponding to three allopatricpopulations in Cyprus Turkey and IsraelminusLebanonThe Cypriot endemic nominotypical subspeciesA sc schreiberi and the Turkish subspeciesA sc ataturi cluster together (to form clade B Fig 2)nesting between A b asper clades This lineage is sisterto a clade of A b asper including A sc syriacus (cladeC Fig 2) We estimate the divergence time of theCypriotminusTurkish lineage of A schreiberi to have beenduring the late Miocene around 6 Mya although thereis no support for this split in the tree In other analy-ses using the whole genus this split is well support-ed in Bayesian analyses (K Tamar S Carranza RSindaco J Moravec JF Trape amp S Meriri unpubldata) Based on mitochondrial data Poulakakis et al(2013) found that both the Cypriot and Turkish sub-species are monophyletic and diverged from each other085 Mya (038ndash156 Mya) According to our results thisdate corresponds to an inner divergence of the

A sc schreiberi lineage rather than to the date at whichA sc schreiberi colonized Cyprus

The discrepancy in the phylogenetic relationship ofA sc schreiberi raises questions regarding the arrivalon Cyprus Cyprus originated with the raising of theTroodos Massif during the upper Cretaceous c 91 to88 Mya (Clube amp Robertson 1986 Mukasa amp Ludden1987) During the middle to late Miocene only a smallproportion of Cyprus was exposed above the Mediter-ranean (McCallum amp Robertson 1990 Robertson 1990)Towards the end of the Miocene sim596 Mya with theclosing of the passage between the Atlantic Ocean andthe Mediterranean basin the Messinian salinity crisisbegan (Krijgsman et al 1999) This resulted in thedrying up of much of the Mediterranean Sea and highsea-mounts emerged to form land bridges with thesurrounding land (Hsuuml et al 1977) By the end of theMiocene and early Pliocene sim533 Mya the passagewith the Atlantic Ocean reopened and the Mediterra-nean basin was refilled (Krijgsman et al 1999) Re-sulting from compressions raising and uplifting of thesurrounding areas towards and during the Pleisto-cene Cyprus was a complete emergent island (McCallumamp Robertson 1990) The possible connection of Cyprusto the mainland (ie to TurkeySyria) during theMessinian is debated as are suggestions of a land con-nection at later periods (Steininger amp Roumlgl 1984 Jolivetet al 2006 Bache et al 2012) Such a connection ifit existed could have provided access for terrestrialorganisms with poor overseas dispersal ability suchas lizards to colonize the island Several studies arguethat post-Messinian sea level changes are unlikely tohave formed connections between Cyprus and the main-land (Steininger amp Roumlgl 1984 Jolivet et al 2006) Thusour dating of the split between the Cypriot A schreiberiand A b asper at c 6 Mya leads us to suggest that theancestor of A s schreiberi colonized Cyprus from themainland through a land bridge connection at the be-ginning of the Messinian crisis rather than by a muchlatermore recent transmarine dispersal as suggestedby Poulakakis et al (2013) Owing to its close rela-tions with A b asper the ancestor of A schreiberi waspresumably mainland A boskianus and the cladogenesisleading to A schreiberi thus rendered A b asperparaphyletic

The Turkish subspecies A schreiberi ataturi was rec-orded for the first time by Franzen (1998) at a veryrestricted area of around 15 km of coastal strip (betweenBotas and Yukarı Burnaz Hatay Province) Owing tothe remarkable morphological similarity between

Figure 3 Haplotype networks of the nuclear gene fragments melano-cortin 1 receptor (MC1R) acetylcholinergic recep-tor M4 (ACM4) and oocyte maturation factor MOS (c-mos) with colours corresponding to Figures 1 and 2 Codes corre-late to the two alleles (ie a and b) of specimens in Table 1 Circle sizes are proportional to the number of alleles (Colourversion of figure available online)

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 13

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

A sc ataturi and the Cypriot population the speci-mens were initially identified as A sc schreiberi(Franzen 1998) Yalccedilinkaya amp Goumlccedilmen (2012) howeverdescribed this population as a new distinct subspe-cies A s ataturi presenting several differences betweenthe two in both morphology and blood-serum pro-teins The origin of A s ataturi remains uncertain asit is debated whether the newly discovered Turkishpopulation is a relict or an introduction from CyprusFranzen (1998) described this population as a pos-sible introduction from Cyprus through the harbourof Botas but Sindaco et al (2000) suggested thatit might be a relict of a previously larger popula-tion because its present distribution is similar tothat of some insects and lizards [Archaeolacerta(Phoenicolacerta) laevis and Ablepharus budaki]Yalccedilinkaya amp Goumlccedilmen (2012) proposed that A sc ataturiarrived in Turkey from the nominate population inCyprus during the Messinian crisis The phylogeneticresults haplotype networks and low levels of geneticdivergence we found suggest that the two subspeciesfrom Cyprus and Turkey have not been geneticallyisolated for a long period of time (ie they share allelesin all three nuclear genes and A s ataturi is nestedwithin A s schreiberi in the phylogeny Figs 2 3) Ourresults therefore contrast with the two latter sce-narios of a relict population or a Messinian disper-sal Both divergence time and the genetic similarityof the two subspecies agree with the original sugges-tion of Franzen (1998) that these animals were intro-duced into Turkey from Cyprus Further support forthis hypothesis is that A s ataturi is restricted to thevicinity of the Botas-Adana harbour and is absent inother suitable habitats (coastal sand dunes) wide-spread in south-eastern Turkey Its close morphologi-cal features to A s schreiberi (Franzen 1998) likewisesupport an introduction scenario

The third subspecies A s syriacus is nested withinA b asper in the concatenated mtDNA and nDNA treesand is clearly genetically distinct from the Cyprus andTurkey A schreiberi lineage The close relations ofA s syriacus with A boskianus may shed light on theorigin of the former Acanthodactylus schreiberi syriacusis distributed on stable sands of the coastal plain ofthe eastern Mediterranean in Israel and southernLebanon (Salvador 1982 Hraoui-Bloquet et al 2002Bar amp Haimovitch 2011) habitats resembling those ofA s schreiberi from Cyprus (Baier Sparrow amp Wiedl2009) The oldest divergence of the A b asper cladethat includes A s syriacus is estimated to have oc-curred during the late Miocene around 558 Mya butno further dates are available for the grouping ofA s syriacus as a result of low support values Thecoastal plain of the eastern Mediterranean was sub-merged during the late Miocene and re-emerged onlytoward the Pliocene (Nir 1970 Horowitz 1979) The

sands of the coastal plain where A s syriacus occurs(Salvador 1982 Arnold 1983 K Tamar amp S Meiripers observ) were repeatedly submerged and re-emerged during the Pleistocene sea-level changes (duringinterglacial and glacial periods respectively) A pos-sible scenario for A sc syriacusrsquos origin includes severalwaves of dispersal of Middle Eastern A boskianus whichoccurs on coarse substrates (Amitai amp Bouskila 2001Disi et al 2001 Baha El Din 2006 pers observ) towardthe Mediterranean shore Acanthodactylus boskianusasper is absent from Mediterranean climate habitatsin Lebanon and Israel It occupies only xeric zonessuggesting an invasion to the coastal plain when sandyhabitats allowed desert flora and fauna to migrate north-wards (Yom-Tov 1988) These populations adapted tosandy soils and evolved morphological features thatdistinguish them from the desert hard substrate formsof A b asper We view this as the most likely scenariogiven the biogeography the phylogenetic results andthe habitat preferences and adaptations of these lizardsAn alternative scenario according to which the an-cestor of A schreiberi originated in Cyprus and dis-persed to the shores of Israel and Lebanon (or originatedin the coastal plain of the Eastern Mediterranean anddispersed to Cyprus) we regard as far less likely Sucha scenario requires much closer genetic relationshipsbetween these two forms and is further weakened bythe close relationship between A s syriacus and thegeographically adjacent A b asper populations

Acanthodactylus boskianus asper is highly variableboth morphologically (Salvador 1982 Arnold 1983) andgenetically (this study) The subspecies is paraphyleticas A schreiberi is nested within it The topology of theA b asper tree shows four different geographical group-ings Syria (clade A) north Jordan plus north OmanNorth Africa and Middle East plus south Arabia (groupsC1 C2 C3 respectively) The different groups in thissubspecies are estimated to have first diverged duringthe late Miocene approximately 65 Mya with the splitof the Syrian population The Syrian lineage is ge-netically distant from A schreiberi and the otherA b asper specimens The nuclear networks indicatethat this group is closer to the other A b asper samplesrather than to A schreiberi The geographical splitsin the rest of the A b asper range (clade C) are esti-mated to have started around 558 Mya These groupsare supported as a distinct clade but are closely relatedto each other in both the concatenated and nuclear trees(Figs 2 S1 respectively) The diversification within thisclade is estimated to have occurred during the lateMiocene to early Pliocene when A b asper dispersedwidely west to North Africa and in Arabia The di-vergence within the North African group (group C2)is estimated to have occurred during the Pliocene ap-proximately 456 Mya with the Egyptian Nigerian andSudanese populations later dispersing west and north

14 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

in Africa This diversification correlates to the arid climatestarting in southern Sahara during the early-mid Plio-cene and later in northern Africa between the Plio-cene and the Pleistocene (Le Houeacuterou 1997) as hasbeen suggested for the dispersal of Mesalina guttulatain Africa (Kapli et al 2008) Other evidence relatesdry climate in North Africa to an earlier period around7 Mya (Schuster et al 2006) as has been suggestedfor the genus Chalcides and other reptiles (Carranzaet al 2008 Metallinou et al 2012 and reference therein)The aridification of North Africa has most likely con-tributed for the successful dispersal of A b asper westfrom south-west Asia into Africa Morphological studiesof A boskianus show relatively uniform populations inNorth Africa suggesting recent migration (Salvador1982 Arnold 1983) The other two geographical group-ings of A b asper from the Middle East and Arabia(groups C1 and C3) are located in two distinct innerclades but their location within each inner clade ispoorly supported The topology of the concatenated tree(Fig 2) shows that the group from northern Jordanand northern Oman (group C1) is closer to the NorthAfrican one than to the geographically close Middle-Eastern and south Arabian group (group C3) The taxo-nomic separation between north and south Oman hasbeen recognized in other species of reptiles and sup-ported by the topography of Oman (eg Echis coloratusand Echis omanensis Arnold Robinson amp Carranza2009) In the nuclear tree (Fig S1) these two groupsare closer to one another and with the North Africangroup form clade C Therefore the low support valuesamongst these groupings prevent an appropriate andthorough analysis of this subspecies The close rela-tionship amongst the geographical groups may reflectclose phylogenetic relationships amongst these popu-lations suggesting recent migration divergence andongoing gene flow

SYSTEMATICS AND TAXONOMIC IMPLICATIONS

The relationships within the A boskianus species groupconflict with the current known taxonomy of A schreiberiand A b asper (samples of the other subspecies ofA boskianus and of A nilsoni were unavailable for thisstudy) Both species have been found to be closelyrelated and paraphyletic The constrained topology testsexemplify the close entangled relationship between thetwo species as the separate monophyly of the two specieswas not rejected and the enforced monophyly of themboth together was inconclusive

Several causes can be responsible for paraphyly inspecies such in the A boskianus species group (Funkamp Omland 2003 and references therein) (1) inad-equate phylogenetic information (2) imperfect taxono-my (incorrectinaccurate species limits) derived frommisidentifying intraspecific variation (3) interspecific

gene flow ndash hybridization through interspecific matingand the subsequent backcrossing of hybrids into theparental populations (4) incomplete lineage sortingbecause of recent speciation events (5) unrecognizedparalogy We suggest that the relationships betweenA schreiberi and A b asper based on mitochondrialand nuclear data are most likely explained by incor-rect taxonomy probably because of the great variabil-ity of the latter species and to convergence As wasthe case in the molecular studies of the A pardalisand A erythrurus species groups (Harris et al 2004Fonseca et al 2008 2009 Carretero et al 2011 andreference therein) there are many problems with thecurrent taxonomic status of several species groups withinAcanthodactylus

Taking the molecular results of our study into accountthere are several systematic approaches to classify-ing the A schreiberiminusA b asper clade The Cypriot andTurkish populations of A schreiberi are very closelyrelated with the latter nested in the former and thetwo subspecies share nuclear alleles (Fig 3) Further-more the low uncorrected p-distance is positively cor-related with subspecies-level distances within otherlacertid species (ie 16 of Cytb in Lacerta bilineatachloronota Godinho et al 2005) We therefore con-clude that Cypriot animals were recently introducedto Turkey and that the Turkish population does notmerit a subspecific rank We suggest that A s ataturiYalccedilinkaya amp Goumlccedilmen 2012 is a junior synonym ofA s schreiberi Boulenger 1878

Regarding the relationships between A schreiberi andA b asper a few scenarios are possible One is to sinkA schreiberi within A boskianus to create one species(A boskianus) with high genetic and morphological vari-ability ranging over a broad distribution Another isfor the two taxa be regarded as a species complex (theA boskianus-schreiberi complex) until further inves-tigation on the subject However although A schreiberiis nested within A b asper the populations from Cyprusand Turkey represent a distinct evolutionary lineagewith distinct genetic and morphological features andthus it is logical to retain the specific status Two othersolutions are possible The first is to re-evaluate theSyrian populations and to consider elevating them aswell as the more divergent lineages (and subspecies)of A boskianus to specific status This would neces-sitate an examination of the phylogeny and morphol-ogy of the other four subspecies of A boskianus(A b boskianus A b euphraticus A b khattensis andA b nigeriensis) and the identification of distinctivephenotypic features in the Syrian lizards Another so-lution is to recognize the maintenance of gene flowamongst mainland populations of A b asper after thedivergence of the insular endemic A schreiberi andthus the evolutionary cohesion of the paraphyleticA b asper Arnold (1983) noted that A schreiberi may

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 15

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

have originated as an isolate from A boskianus becauseof their shared morphology and hemipenis featuresOur results support this scenario which includes thedispersal of A schreiberi to Cyprus from a mainlandpopulation that was most probably A boskianus It maybe assumed that the ancestor of the Cypriot A schreiberiafter arriving on Cyprus remained isolated for a longperiod of time and thus evolved to the modern formof A schreiberi Meanwhile the same ancestral con-tinental populations not isolated from each other con-tinued to exchange genes to varying degrees remainingA boskianus

The IsraeliminusLebanese subspecies A sc syriacus is onlydistantly related to the nominate form A sc schreiberiThis subspecies is highly phylogenetically divergent fromthe Cypriot and Turkish populations having higherp-distances (12S 4 Cytb 11ndash12) than those foundbetween other lacertid species (eg 74ndash82 of Cytbamongst Iberolacerta aranica Iberolacerta aurelioi andIberolacerta bonnali and 41ndash58 of Cytb betweenLacerta bilineata and Lacerta viridis Crochet et al2004 Godinho et al 2005 respectively) The nuclearhaplotype networks further show that Lebanese andIsraeli populations share alleles only with A b asperbut not with the nominotypical Cypriot form Arnold(1983) suggested that the geographical variation ofA boskianus reflects niche differences with animalsfrom xeric areas with dense rigid and spiny vegeta-tion having larger dorsal scales than animals from moremesic areas As was assumed for A schreiberi wesuggest that other mainland populations of A b asperwere the ancestors of the LebaneseminusIsraeli Coastal plainforms We suggest that A s syriacus is an ecomorphof A b asper that dispersed from the usual xerichabitats of the species and adapted to the new moremesic environment of the stable sands of the coastalplains of the eastern Mediterranean As a conse-quence this ecomorph converged on the morphologyof A s schreiberi which inhabits the coastal sands ofCyprus (Baier et al 2009) but still maintains differ-entiating features by having coarser dorsal scales andsharp keels (Salvador 1982 Arnold 1983 Franzen1998) This convergence led to the description ofA s syriacus as a member of A schreiberi The mor-phological assessment and the close morphological simi-larities between A b asper and A s syriacus mayexplain the wrong classification A similar erro-neous reasoning led Reed amp Marx (1959) to identifyspecimens with fine scales from Iraq as A schreiberiSalvador (1982) re-examined these specimens and as-signed them to A boskianus The morphological dif-ferences between the two forms are less prominentespecially where the two forms occur in close geo-graphical proximity in the southern coastal plainand north-western Negev Desert of Israel (Bar ampHaimovitch 2011) According to our results A s syriacus

actually belongs to A b asper being a coastal-duneecomorph convergent with but evolutionarily dis-tinct from A schreiberi Thus our preferred scenariois to treat the name Acanthodactylus schreiberi syriacusBoumlttger 1879 (which was originally described asA boskianus var syriacus by Boumlttger 1879) as a juniorsynonym of the name Acanthodactylus boskianus asper(Audouin 1827)

Recognizing A s syriacus as a junior synonym ofA b asper may have important implications for the con-servation of this coastal sand dune form which is clas-sified as critically endangered in Israel (Dolev ampPervolutzki 2004) However as the Israeli and Leba-nese coastal dune ecosystem has probably developedonly very recently during the Quaternary (Nir 1970Horowitz 1979) this form represents a remarkable caseof rapid evolutionary change It is also a remarkablecase of convergent evolution (with the CypriotA sc schreiberi) Thus we feel that these popula-tions are unique evolutionary entities that merit specialconservation efforts

The use of nuclear genes is a valuable method forestimating species divergence and lineage sorting andhelps evaluate isolated lineages and evolutionary historyThe incorporation of mitochondrial and nuclear dataprovides thorough topologies informative networks anddivergence times that reveal useful information for aproblematic taxonomy such as that of the A boskianusspecies group We have shown that phylogenetic ap-proaches to the confusing taxonomy of two closely relatedand morphologically similar species can shed light ontheir unclear relationships resolve between homoplasyand shared ancestry and identify patterns of speciesevolutionary history and biogeography

ACKNOWLEDGEMENTS

We are grateful to the following people for providingsamples for this study D Donaire B Shacham J ŠmiacutedS M Baha El Din and J Padial We thank MMetallinou and J Šmiacuted for help with the lab work andthe analyses M Novosolov for help with the figuresY L Werner B Shacham and A Levy for fruitful dis-cussions and two anonymous referees for helpful com-ments on an earlier version of this manuscript Wethank the Israel nature and park authority for issuingcollecting permits (nos 38074 38451 38489 38540)The work of J M was financially supported by the Min-istry of Culture of the Czech Republic (DKRVO 201315 National Museum 00023272) S C is funded bygrant CGL2012-36970 from the Ministerio de Economiacuteay Competitividad Spain (cofunded by FEDER) Thisstudy was funded by a Israel Taxonomic Initiative grantto S M K T is supported by an Israel Taxonomic Ini-tiative scholarship

16 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

REFERENCES

Ahmadzadeh F Carretero MA Harris DJ Perera ABohme W 2012 A molecular phylogeny of the eastern groupof ocellated lizard genus Timon (Sauria Lacertidae) basedon mitochondrial and nuclear DNA sequences Amphibia-Reptilia 33 1ndash10

Ahmadzadeh F Flecks M Roumldder D Boumlhme W Ilgaz CcedilHarris DJ Engler JO Uumlzuumlm N Carretero MA 2013Multiple dispersal out of Anatolia biogeography and evolu-tion of oriental green lizards Biological Journal of the LinneanSociety 110 398ndash408

Akaike H 1973 Information theory and an extension of themaximum likelihood principle In Petrov BN Csaki F edsSecond International Symposium on Information Theory Bu-dapest Akademiai Kiado 267ndash281

Amitai P Bouskila A 2001 Handbook of amphibians ampreptiles of Israel Israel Keter Publishing House Ltd

Anderson SC 1999 The lizards of Iran Ithaca NY Societyfor the Study of Amphibians and Reptiles

Arnold EN 1980 The scientific results of the Oman flora andfauna survey 1977 (Dhofar) The reptiles and amphibians ofDhofar southern Arabia Journal of Oman Studies SpecialReport 2 273ndash332

Arnold EN 1983 Osteology genitalia and the relationshipsof Acanthodactylus (Reptilia Lacertidae) Bulletin of the BritishMuseum (Natural History) Zoology 44 291ndash339

Arnold EN Arribas Oacute Carranza S 2007 Systematics ofthe Palaearctic and Oriental lizard tribe Lacertini (SquamataLacertidae Lacertinae) with descriptions of eight new generaZootaxa 1430 1ndash86

Arnold EN Robinson MD Carranza S 2009 A prelimi-nary analysis of phylogenetic relationships and biogeogra-phy of the dangerously venomous Carpet Vipers Echis(Squamata Serpentes Viperidae) based on mitochondrial DNAsequences Amphibia-Reptilia 30 273ndash282

Audouin JV 1827 Explication sommaire des planches dereptiles (suppleacutement) offrant un exposeacute des characteacuteresdes espegraveces In Savigny MJCL ed Description de lrsquoEacutegypteVol 1 Histoire Naturelle Paris Impeacuteriale 161ndash184

Bache F Popescu SM Rabineau M Gorini C Suc JPClauzon G Olivet JL Rubino JL Melinte-DobrinescuMC Estrada F 2012 A two-step process for the refloodingof the Mediterranean after the Messinian Salinity Crisis BasinResearch 24 125ndash153

Baha El Din SM 2006 A guide to the reptiles and amphib-ians of Egypt CairondashNew York American University in CairoPress

Baier FS Sparrow DJ Wiedl H-J 2009 The amphibiansand reptiles of Cyprus Frankfurt am Main Edition Chimaira

Bandelt H-J Forster P Roumlhl A 1999 Median-joining net-works for inferring intraspecific phylogenies Molecular Biologyand Evolution 16 37ndash48

Bar A Haimovitch G 2011 A field guide to reptiles and am-phibians of Israel Herzliya Pazbar Limited

Boumlttger O 1879 Reptilien und Amphibien aus Syrien Berichtuumlber die Senckenbergische Naturforschende Gesellschaft inFrankfurt am Main 1879 57ndash84

Boulenger GA 1919 On a new variety of Acanthodactylusboskianus Daud from the Euphrates The Annals and Maga-zine of Natural History 3 549ndash550

Boulenger GA 1878 Sur les espegraveces drsquoAcanthodactylus desbords de la Mediterraneacutee Bulletin de la Socieacuteteacute zoologiquede France 3 179ndash197

Boulenger GA 1918 Sur les leacutezards du genre AcanthodactylusWieg Bulletin de la Socieacuteteacute zoologique de France 43 143ndash155

Boulenger GA 1921 Monograph of the Lacertidœ Vol II Trus-tees of the British Museum of Natural History London

Carranza S Arnold EN 2012 A review of the geckos of thegenus Hemidactylus (Squamata Gekkonidae) fromOman based on morphology mitochondrial and nuclear datawith descriptions of eight new species Zootaxa 3378 1ndash95

Carranza S Arnold EN Geniez P Roca J Mateo J 2008Radiation multiple dispersal and parallelism in the skinksChalcides and Sphenops (Squamata Scincidae) with com-ments on Scincus and Scincopus and the age of the SaharaDesert Molecular Phylogenetics and Evolution 46 1071ndash1094

Carretero MA Fonseca MM Garcia-Munoz E Brito JCHarris DJ 2011 Adding Acanthodactylus beershebensis tothe mtDNA phylogeny of the Acanthodactylus pardalis groupNorth-Western Journal of Zoology 7 138ndash142

Castresana J 2000 Selection of conserved blocks from multi-ple alignments for their use in phylogenetic analysis Mo-lecular Biology and Evolution 17 540ndash552

Clube TMM Robertson A 1986 The palaeorotation of theTroodos microplate Cyprus in the Late Mesozoic-Early Ce-nozoic plate tectonic framework of the Eastern Mediterra-nean Surveys in Geophysics 8 375ndash437

Cox SC Carranza S Brown RP 2010 Divergence timesand colonization of the Canary Islands by Gallotia lizardsMolecular Phylogenetics and Evolution 56 747ndash757

Crochet PA Geniez P Ineich I 2003 A multivariate analy-sis of the fringe-toed lizards of the Acanthodactylus scutellatusgroup (Squamata Lacertidae) systematic and biogeographi-cal implications Zoological Journal of the Linnean Society137 117ndash155

Crochet P-A Chaline O Surget-Groba Y Debain CCheylan M 2004 Speciation in mountains phylogeographyand phylogeny of the rock lizards genus Iberolacerta (ReptiliaLacertidae) Molecular Phylogenetics and Evolution 30 860ndash866

Daudin FM 1802 Histoire naturelle geacuteneacuterale et particuliegraveredes reptiles ouvrage faisant suite agrave lrsquoHistoire naturelle geacuteneacuteraleet particuliegravere F Dufart

Disi A Modry D Necas P Rifai L 2001 Amphibians andreptiles of the Hashemite Kingdom of Jordan an atlas andfield guide Frankfurt am Main Chimaira

Dolev A Pervolutzki A 2004 Endangered species in IsraelRed list of threatened animals Vertebrates JerusalemThe Nature and Parks Authority and the Society for thePreservation of Nature

Drummond AJ Rambaut A 2007 BEAST Bayesian evo-lutionary analysis by sampling trees BMC EvolutionaryBiology 7 214

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 17

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Ezard T Fujisawa T Barraclough T 2009 Splits speciesrsquolimits by threshold statistics R package version 10-19r48Available at httpR-ForgeR-projectorgprojectssplits

Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783ndash791

Fitzinger LJFJ 1834 Acanthodactylus In Wiegman AFAed Herpetologia mexicana seu Descriptio amphibiorum NovaeHispaniae quae itineribus comitis De Sack Ferdinandi Deppeet Chr Guil Schiede in Museum zoologicum Berolinensepervenerunt Pars prima saurorum species amplectens adjectosystematis saurorum prodromo additisque multis in huncamphibiorum ordinem observationibus Berlin C G Luderitz10

Flot JF 2010 SeqPHASE a web tool for interconvertingPHASE inputoutput files and FASTA sequence align-ments Molecular Ecology Resources 10 162ndash166

Fonseca MM Brito JC Paulo OS Carretero MA HarrisDJ 2009 Systematic and phylogeographical assessment ofthe Acanthodactylus erythrurus group (Reptilia Lacertidae)based on phylogenetic analyses of mitochondrial and nuclearDNA Molecular Phylogenetics and Evolution 51 131ndash142

Fonseca MM Brito JC Rebelo H Kalboussi M LarbesS Carretero MA Harris DJ 2008 Genetic variation amongspiny-footed lizards in the Acanthodactylus pardalis groupfrom North Africa African Zoology 43 8ndash15

Fontaneto D Herniou EA Boschetti C Caprioli M MeloneG Ricci C Barraclough TG 2007 Independently evolv-ing species in asexual bdelloid rotifers PLoS Biology 5 e87

Franzen M 1998 Erstnachweis von Acanthodactylus schreiberischreiberi Boulenger 1879 fuumlr die Tuumlrkei (Squamata SauriaLacertidae) Herpetozoa 11 27ndash36

Funk DJ Omland KE 2003 Species-level paraphyly andpolyphyly frequency causes and consequences with in-sights from animal mitochondrial DNA Annual Review ofEcology Evolution and Systematics 34 397ndash423

Godinho R Crespo EG Ferrand N Harris DJ 2005 Phy-logeny and evolution of the green lizards Lacerta spp(Squamata Lacertidae) based on mitochondrial and nuclearDNA sequences Amphibia-Reptilia 26 271ndash285

Greenbaum E Villanueva CO Kusamba C Aristote MMBranch WR 2011 A molecular phylogeny of EquatorialAfrican Lacertidae with the description of a new genus andspecies from eastern Democratic Republic of the Congo Zoo-logical Journal of the Linnean Society 163 913ndash942

Guillou H Carracedo JC Torrado FP Badiola ER 1996K-Ar ages and magnetic stratigraphy of a hotspot-inducedfast grown oceanic island El Hierro Canary Islands Journalof Volcanology and Geothermal Research 73 141ndash155

Harris DJ Arnold EN 2000 Elucidation of the relation-ships of spiny-footed lizards Acanthodactylus spp (ReptiliaLacertidae) using mitochondrial DNA sequence with com-ments on their biogeography and evolution Journal of Zoology252 351ndash362

Harris DJ Batista V Carretero M 2004 Assessment ofgenetic diversity within Acanthodactylus erythrurus (ReptiliaLacertidae) in Morocco and the Iberian Peninsula usingmitochondrial DNA sequence data Amphibia-Reptilia 25227

Horowitz A 1979 The Quaternary of Israel New York Aca-demic Press

Hraoui-Bloquet S Sadek RA Sindaco R Venchi A 2002The herpetofauna of Lebanon new data on distributionZoology in the Middle East 27 35ndash46

Hsuuml KJ Montadert L Bernoulli D Cita MB EricksonA Garrison RE Kidd RB Megraveliereacutes F Muumlller C WrightR 1977 History of the Mediterranean salinity crisis Nature267 399ndash403

Huelsenbeck JP Rannala B 2004 Frequentist propertiesof Bayesian posterior probabilities of phylogenetic trees undersimple and complex substitution models Systematic Biology53 904ndash913

Huelsenbeck JP Ronquist F 2001 MRBAYES Bayesianinference of phylogenetic trees Bioinformatics 17 754ndash755

Jolivet L Augier R Robin C Suc J-P Rouchy JM 2006Lithospheric-scale geodynamic context of the Messinian sa-linity crisis Sedimentary Geology 188 9ndash33

Kapli P Lymberakis P Poulakakis N Mantziou GParmakelis A Mylonas M 2008 Molecular phylogeny ofthree Mesalina (Reptilia Lacertidae) species (M guttulataM brevirostris and M bahaeldini) from North Africa and theMiddle East another case of paraphyly MolecularPhylogenetics and Evolution 49 102ndash110

Katoh K Toh H 2008 Recent developments in the MAFFTmultiple sequence alignment program Briefings inBioinformatics 9 286ndash298

Krijgsman W Hilgen F Raffi I Sierro F Wilson D 1999Chronology causes and progression of the Messinian salin-ity crisis Nature 400 652ndash655

Lanfear R Calcott B Ho SY Guindon S 2012PartitionFinder combined selection of partitioning schemesand substitution models for phylogenetic analyses Molecu-lar Biology and Evolution 29 1695ndash1701

Le Houeacuterou HN 1997 Climate flora and fauna changes inthe Sahara over the past 500 million years Journal of AridEnvironments 37 619ndash647

Maca-Meyer N Carranza S Rando J Arnold E CabreraV 2003 Status and relationships of the extinct giant CanaryIsland lizard Gallotia goliath (Reptilia Lacertidae)assessed using ancient mtDNA from its mummifiedremains Biological Journal of the Linnean Society 80 659ndash670

McCallum J Robertson A 1990 Pulsed uplift of the TroodosMassif ndash evidence from the Plio-Pleistocene Mesaoria basinIn J Malpas EMM Panayiotou A Xenophontos C edsOphiolites oceanic crustal analogues Proceedings of theSymposium lsquoTroodos 1987rsquo Nicosia (The Geological SurveyDepartment Ministry of Agriculture and NaturalResources) 217ndash229

Metallinou M Arnold EN Crochet P-A Geniez P BritoJC Lymberakis P El Din SB Sindaco R Robinson MCarranza S 2012 Conquering the Sahara and Arabiandeserts systematics and biogeography of Stenodactylus geckos(Reptilia Gekkonidae) BMC Evolutionary Biology 12258

Monaghan MT Wild R Elliot M Fujisawa T Balke MInward DJ Lees DC Ranaivosolo R Eggleton P

18 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

performed for 5 times 107 generations with a sampling fre-quency of 10 000 The results were combined to inferthe ultrametric tree after discarding 10 of the samplesfrom each run Models and prior specifications appliedwere as follows (otherwise by default) for partitionsby genes and by PartitionFinder For gene partitionsGTR + I + G strict clock (12S) Hasegawa-Kishino-Yano + Invariable sites + Gamma distribution(HKY + I + G) strict clock molecular clock model (es-timate 0ndash1) (Cytb) coalescence constant size processof speciation random starting tree alpha Uniform (010) GTR Uniform For partitions by PartitionFinderGTR + I + G strict clock (partition 1 = 12S + Cytb codon1 and 2) Tamura-Nei + Gamma distribution (TrN + G)strict clock (partition 2 = Cytb codon 3) coalescenceconstant size process of speciation random starting treealpha Uniform (0 10) Parameter values both for clockand substitution models were unlinked across parti-tions For all analyses implemented in BEAST the threeruns were analysed in TRACER v 15 (Rambaut ampDrummond 2007) confirming convergence The treeswere combined in LogCombiner and TreeAnnotator(available in BEAST package) was used for the pro-duction of the final tree

For estimating species limits directly from the Bayes-ian phylogenetic tree produced with the concat-enated mitochondrial data we used the independentgeneralized mixed Yule-coalescent (GMYC) method (Ponset al 2006) The GMYC model estimated the numberof phylogenetic clusters or lsquospeciesrsquo by identifying theshifts between intraspecific (coalescence) and interspecific(diversification) branch rates (Pons et al 2006) Weperformed the GMYC function in the R v302 lsquosplitsrsquopackage (Ezard Fujisawa amp Barraclough 2009) Alikelihood-ratio test was used to determine if the GMYCmodel with a shift in the branching processes provid-ed a better fit to the data than the null model withno shifts We used a single threshold value (Monaghanet al 2009) which has already been applied success-fully to different groups of organisms (Pons et al 2006Fontaneto et al 2007 Monaghan et al 2009)

ESTIMATION OF DIVERGENCE TIMES

The lack of internal calibration points in Acantho-dactylus (no fossils are known) prevents the directestimation of time in our phylogeny Therefore we usedthe mean substitution rates and their standard errorof the same 12S and Cytb mitochondrial regions ex-tracted from a fully calibrated phylogeny of anotherlacertid group the lizards of the genus Gallotia endemicto the Canary Islands (Cox Carranza amp Brown 2010as was implemented in Carranza amp Arnold 2012) Theinferred calibration rate was estimated using the ageof El Hierro Island (Canary Islands) estimated at112 Mya (Guillou et al 1996) They assumed coloni-

zation of the island by members of the lacertid genusGallotia (Gallotia caesaris caesaris endemic to El HierroIsland) immediately after its formation from the neigh-bouring La Gomera Island (inhabited by the endemicGallotia caesaris gomerae) These two subspecies aremonophyletic sister taxa with low intraspecific vari-ability (Maca-Meyer et al 2003 Cox et al 2010) andthus suitable for calibration

For the estimation of divergence times one repre-sentative of each independent GMYC lineage was usedfrom the ultrametric tree (for the representatives seeTable 1) We used a likelihood-ratio test implementedin MEGA 52 (Tamura et al 2011) to test if the dif-ferent partitions (by genes) included in the dating analy-sis were evolving in a clock-like fashion (Table 2) Thisinformation was used to choose between the strict clockand the relaxed uncorrelated lognormal clock priorsimplemented in BEAST (Monaghan et al 2009) Thedata set included one representative from each lineagefrom the GMYC analysis using sequences from all fivepartitions (nuclear genes unphased) Three individ-ual runs were performed for 5 times 107 generations witha sampling frequency of 10 000 and the results werecombined to infer the ultrametric tree after discard-ing 10 of the samples from each run Models and priorspecifications applied were as follows (otherwise bydefault) GTR + I + G relaxed uncorrelated lognor-mal clock molecular clock model (estimate) (12S Cytb)HKY strict clock (MC1R c-mos) and TrN + I strictclock (ACM4) Yule process of speciation random start-ing tree yulebirthRate (0 1000) alpha Uniform (010) ucldmean of 12S Normal (initial value 000553mean 000553 SD 000128) ucldmean of Cytb Normal(initial value 00164 mean 00164 SD 000317) Pa-rameter values both for clock and substitution modelswere unlinked across partitions

RESULTS

The data set of this study is comprised of 19 samplesof A schreiberi 65 samples of A b asper and 11outgroup samples (Table 1 Fig 1) The data set in-cluded mitochondrial DNA (mtDNA) gene fragmentsof 12S (sim387 bp) and Cytb (405 bp) and nuclear DNA(nDNA) gene fragments of MC1R (663 bp) ACM4(429 bp) and c-mos (522 bp) totalling to sim2406 bp Thenumber of variable (V) and parsimony-informative(Pi) sites for the ingroup are listed in Table S1 Thetwo partition approaches (ie by gene and byPartitionFinder) gave similar results for both the MLand BI analyses The results of the phylogenetic analy-ses of the complete concatenated data set using MLand BI methods produced very similar topologies butdiffered to some extent at the less supported nodesat the intraspecific level (Fig 2) Separated analysesof the nuclear data sets are presented in Figure S1

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 9

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

10 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Together A b asper and A schreiberi form amonophyletic group within Acanthodactylus (Fig 2)Within the group however both taxa are paraphyleticwith A schreiberi as a whole nested within A b asperOur analyses distinguish three major clades (1) cladeA formed by A b asper from Syria (2) clade B in-cludes the two subspecies A sc ataturi from Turkeytogether with A sc schreiberi from Cyprus (3) cladeC which includes specimens of A b asper from the re-maining localities in its distribution range together withA sc syriacus from Israel and Lebanon Clade A is verywell supported and includes specimens of A b asperfrom central and northern Syria (Fig 1) splitting fromother specimens at the basal node of the group is es-timated to have occurred c 654 Mya [95 highest pos-terior density (HPD) 392ndash952 Mya] The level of geneticdifferentiation(p-distance) between these specimens and the remain-ing A b asper and all A schreiberi specimens is 37ndash46 for 12S and 107ndash119 for Cytb Clade B is alsovery well supported and includes two of the threenominal subspecies of A schreiberi A s schreiberi thenominotypical subspecies endemic to Cyprus and A sataturi from Turkey The Turkish subspecies is nestedwithin the Cypriot specimens and the two forms havelow genetic distances from each other (12S 016 Cytb123) This clade is nested between the two A b asperclades (clades A and C) in both the concatenatedand the nuclear tree although the nodes are not wellsupported Clade C is not very well supported It in-cludes a cluster of A b asper and A s syriacus Thisclade includes two inner clades that split around558 Mya (95 HPD 356ndash8 Mya) and divided into threepoorly supported geographical inner groups (Fig 2)northern Jordan and northern Oman (group C1) NorthAfrica (group C2) and samples from the Middle East(Egypt south Israel and south Jordan) with samplesfrom Yemen and southern Oman (group C3) ndash the latterincluding all specimens of the subspecies A s syriacusThe diversification within the North African group isestimated to have started around 456 Mya (95 HPD282ndash647 Mya) The IsraelminusLebanon endemic subspe-cies A s syriacus is genetically highly distinct from

A s schreiberi and A s ataturi making A schreiberiparaphyletic (p-distance 12S 431 416 Cytb 1181202 respectively)

The networks constructed for the phased haplotypesof the full length nuclear markers (MC1R ACM4 andc-mos) are presented in Figure 3 The nuclear networkanalyses show similar results for each of the three genesand closely agree with the phylogenetic tree The CypriotA s schreiberi and Turkish A s ataturi subspecies sharealleles for all three genes and both are distinct fromthe third subspecies A s syriacus Acanthodactylusschreiberi syriacus shares no alleles with the other sub-species of A schreiberi but does share alleles withA b asper for each of the genes Acanthodactylusschreiberi syriacus shares MC1R alleles with A b asperspecimens from Tunisia Syria and Israel ACM4 alleleswith Egyptian Israeli Jordanian and North Africanspecimens and c-mos alleles only with Israeli A b asperspecimens Syrian A b asper samples share one allelewith A s syriacus and two with other A b asper speci-mens from Egypt Israel and North Africa in the MC1Rone allele with an Egyptian A b asper in the ACM4and none in the c-mos gene

In order to better understand the relationships betweenA schreiberi and A boskianus we performed three to-pology tests in which we forced monophyletic group-ings (1) monophyly of A schreiberi (all three subspeciestogether) (2) monophyly of A b asper (3) monophylyof A b asper and of A schreiberi The results of thetopological tests indicate that our data set cannot rejectthe alternative hypotheses of monophyly of A schreiberi(AU P = 0091 SH P = 0062) and that of A b asper(AU P = 011 SH P = 0072) if we allow A schreiberito nest within A b asper or a monophyletic A b aspernesting within A schreiberi When forcing monophylyof both A schreiberi and of A b asper together in thesame tree the results are inconclusive (AU P = 0046SH P = 0051)

The single-threshold model in GMYC yield a topol-ogy that is clearly different from the known taxono-my The GMYC results present a total of 25 and 24ML independent lineages from the Bayesian haplotypemitochondrial phylogeny of the two species for the two

Figure 2 Maximum likelihood (ML) tree of the Acanthodactylus boskianus and Acanthodactylus schreiberi specimensinferred using 12S rRNA cytochrome b mtDNA and melano-cortin 1 receptor acetylcholinergic receptor M4 and oocytematuration factor MOS nuclear gene fragments Posterior probability in the Bayesian analysis is indicated by black dotson the nodes [values ge 095 shown for both gene partitions and partitions by PartitionFinder (PF)] and ML bootstrapsupport values are indicated in parentheses (values ge 70 shown ML ML-PF) Age estimates with BEAST are indicat-ed near the relevant nodes and include the mean and in brackets the HPD 95 confidence interval Sample codes relateto specimens in Table 1 and in Figures 1 and 3 Colours blue Acanthodactylus boskianus asper yellow Acanthodactylusschreiberi ataturi red Acanthodactylus schreiberi syriacus green Acanthodactylus schreiberi schreiberi (Colour versionof figure available online)

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 11

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

12 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

partition approaches (ie by gene and by PartitionFinderFigs S2 S3 respectively) The two partition ap-proaches gave similar clusters but at the less sup-ported nodes they differed at the positions of severallineages The single threshold GMYC result is indi-cated for a single line at 00037 Mya for the gene par-titions and at 002 Mya for PartitionFinder (verticallines in Figs S2 S3) The topology and clusters re-vealed in this analysis correspond to the lineages fromthe phylogeny of the ML and BI methods both for theparaphyly of the two species and the geographical group-ings within A b asper The GMYC results mainly differfrom the ML and BI methods in the position ofA schreiberi from Cyprus and Turkey as a sister cladeto the Syrian A b asper

DISCUSSION

We have provided a comprehensive and thorough as-sessment of the intraspecific phylogenetic relation-ships within A schreiberi and its closest relativeA b asper Our results based on mitochondrial andnuclear DNA data from 84 specimens across the entiredistribution range of A schreiberi and most of the dis-tribution range of A b asper reveal that A schreiberiis paraphyletic and nested entirely within theA boskianus subspecies

HISTORICAL BIOGEOGRAPHY

Acanthodactylus schreiberi is thought to comprisethree subspecies corresponding to three allopatricpopulations in Cyprus Turkey and IsraelminusLebanonThe Cypriot endemic nominotypical subspeciesA sc schreiberi and the Turkish subspeciesA sc ataturi cluster together (to form clade B Fig 2)nesting between A b asper clades This lineage is sisterto a clade of A b asper including A sc syriacus (cladeC Fig 2) We estimate the divergence time of theCypriotminusTurkish lineage of A schreiberi to have beenduring the late Miocene around 6 Mya although thereis no support for this split in the tree In other analy-ses using the whole genus this split is well support-ed in Bayesian analyses (K Tamar S Carranza RSindaco J Moravec JF Trape amp S Meriri unpubldata) Based on mitochondrial data Poulakakis et al(2013) found that both the Cypriot and Turkish sub-species are monophyletic and diverged from each other085 Mya (038ndash156 Mya) According to our results thisdate corresponds to an inner divergence of the

A sc schreiberi lineage rather than to the date at whichA sc schreiberi colonized Cyprus

The discrepancy in the phylogenetic relationship ofA sc schreiberi raises questions regarding the arrivalon Cyprus Cyprus originated with the raising of theTroodos Massif during the upper Cretaceous c 91 to88 Mya (Clube amp Robertson 1986 Mukasa amp Ludden1987) During the middle to late Miocene only a smallproportion of Cyprus was exposed above the Mediter-ranean (McCallum amp Robertson 1990 Robertson 1990)Towards the end of the Miocene sim596 Mya with theclosing of the passage between the Atlantic Ocean andthe Mediterranean basin the Messinian salinity crisisbegan (Krijgsman et al 1999) This resulted in thedrying up of much of the Mediterranean Sea and highsea-mounts emerged to form land bridges with thesurrounding land (Hsuuml et al 1977) By the end of theMiocene and early Pliocene sim533 Mya the passagewith the Atlantic Ocean reopened and the Mediterra-nean basin was refilled (Krijgsman et al 1999) Re-sulting from compressions raising and uplifting of thesurrounding areas towards and during the Pleisto-cene Cyprus was a complete emergent island (McCallumamp Robertson 1990) The possible connection of Cyprusto the mainland (ie to TurkeySyria) during theMessinian is debated as are suggestions of a land con-nection at later periods (Steininger amp Roumlgl 1984 Jolivetet al 2006 Bache et al 2012) Such a connection ifit existed could have provided access for terrestrialorganisms with poor overseas dispersal ability suchas lizards to colonize the island Several studies arguethat post-Messinian sea level changes are unlikely tohave formed connections between Cyprus and the main-land (Steininger amp Roumlgl 1984 Jolivet et al 2006) Thusour dating of the split between the Cypriot A schreiberiand A b asper at c 6 Mya leads us to suggest that theancestor of A s schreiberi colonized Cyprus from themainland through a land bridge connection at the be-ginning of the Messinian crisis rather than by a muchlatermore recent transmarine dispersal as suggestedby Poulakakis et al (2013) Owing to its close rela-tions with A b asper the ancestor of A schreiberi waspresumably mainland A boskianus and the cladogenesisleading to A schreiberi thus rendered A b asperparaphyletic

The Turkish subspecies A schreiberi ataturi was rec-orded for the first time by Franzen (1998) at a veryrestricted area of around 15 km of coastal strip (betweenBotas and Yukarı Burnaz Hatay Province) Owing tothe remarkable morphological similarity between

Figure 3 Haplotype networks of the nuclear gene fragments melano-cortin 1 receptor (MC1R) acetylcholinergic recep-tor M4 (ACM4) and oocyte maturation factor MOS (c-mos) with colours corresponding to Figures 1 and 2 Codes corre-late to the two alleles (ie a and b) of specimens in Table 1 Circle sizes are proportional to the number of alleles (Colourversion of figure available online)

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 13

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

A sc ataturi and the Cypriot population the speci-mens were initially identified as A sc schreiberi(Franzen 1998) Yalccedilinkaya amp Goumlccedilmen (2012) howeverdescribed this population as a new distinct subspe-cies A s ataturi presenting several differences betweenthe two in both morphology and blood-serum pro-teins The origin of A s ataturi remains uncertain asit is debated whether the newly discovered Turkishpopulation is a relict or an introduction from CyprusFranzen (1998) described this population as a pos-sible introduction from Cyprus through the harbourof Botas but Sindaco et al (2000) suggested thatit might be a relict of a previously larger popula-tion because its present distribution is similar tothat of some insects and lizards [Archaeolacerta(Phoenicolacerta) laevis and Ablepharus budaki]Yalccedilinkaya amp Goumlccedilmen (2012) proposed that A sc ataturiarrived in Turkey from the nominate population inCyprus during the Messinian crisis The phylogeneticresults haplotype networks and low levels of geneticdivergence we found suggest that the two subspeciesfrom Cyprus and Turkey have not been geneticallyisolated for a long period of time (ie they share allelesin all three nuclear genes and A s ataturi is nestedwithin A s schreiberi in the phylogeny Figs 2 3) Ourresults therefore contrast with the two latter sce-narios of a relict population or a Messinian disper-sal Both divergence time and the genetic similarityof the two subspecies agree with the original sugges-tion of Franzen (1998) that these animals were intro-duced into Turkey from Cyprus Further support forthis hypothesis is that A s ataturi is restricted to thevicinity of the Botas-Adana harbour and is absent inother suitable habitats (coastal sand dunes) wide-spread in south-eastern Turkey Its close morphologi-cal features to A s schreiberi (Franzen 1998) likewisesupport an introduction scenario

The third subspecies A s syriacus is nested withinA b asper in the concatenated mtDNA and nDNA treesand is clearly genetically distinct from the Cyprus andTurkey A schreiberi lineage The close relations ofA s syriacus with A boskianus may shed light on theorigin of the former Acanthodactylus schreiberi syriacusis distributed on stable sands of the coastal plain ofthe eastern Mediterranean in Israel and southernLebanon (Salvador 1982 Hraoui-Bloquet et al 2002Bar amp Haimovitch 2011) habitats resembling those ofA s schreiberi from Cyprus (Baier Sparrow amp Wiedl2009) The oldest divergence of the A b asper cladethat includes A s syriacus is estimated to have oc-curred during the late Miocene around 558 Mya butno further dates are available for the grouping ofA s syriacus as a result of low support values Thecoastal plain of the eastern Mediterranean was sub-merged during the late Miocene and re-emerged onlytoward the Pliocene (Nir 1970 Horowitz 1979) The

sands of the coastal plain where A s syriacus occurs(Salvador 1982 Arnold 1983 K Tamar amp S Meiripers observ) were repeatedly submerged and re-emerged during the Pleistocene sea-level changes (duringinterglacial and glacial periods respectively) A pos-sible scenario for A sc syriacusrsquos origin includes severalwaves of dispersal of Middle Eastern A boskianus whichoccurs on coarse substrates (Amitai amp Bouskila 2001Disi et al 2001 Baha El Din 2006 pers observ) towardthe Mediterranean shore Acanthodactylus boskianusasper is absent from Mediterranean climate habitatsin Lebanon and Israel It occupies only xeric zonessuggesting an invasion to the coastal plain when sandyhabitats allowed desert flora and fauna to migrate north-wards (Yom-Tov 1988) These populations adapted tosandy soils and evolved morphological features thatdistinguish them from the desert hard substrate formsof A b asper We view this as the most likely scenariogiven the biogeography the phylogenetic results andthe habitat preferences and adaptations of these lizardsAn alternative scenario according to which the an-cestor of A schreiberi originated in Cyprus and dis-persed to the shores of Israel and Lebanon (or originatedin the coastal plain of the Eastern Mediterranean anddispersed to Cyprus) we regard as far less likely Sucha scenario requires much closer genetic relationshipsbetween these two forms and is further weakened bythe close relationship between A s syriacus and thegeographically adjacent A b asper populations

Acanthodactylus boskianus asper is highly variableboth morphologically (Salvador 1982 Arnold 1983) andgenetically (this study) The subspecies is paraphyleticas A schreiberi is nested within it The topology of theA b asper tree shows four different geographical group-ings Syria (clade A) north Jordan plus north OmanNorth Africa and Middle East plus south Arabia (groupsC1 C2 C3 respectively) The different groups in thissubspecies are estimated to have first diverged duringthe late Miocene approximately 65 Mya with the splitof the Syrian population The Syrian lineage is ge-netically distant from A schreiberi and the otherA b asper specimens The nuclear networks indicatethat this group is closer to the other A b asper samplesrather than to A schreiberi The geographical splitsin the rest of the A b asper range (clade C) are esti-mated to have started around 558 Mya These groupsare supported as a distinct clade but are closely relatedto each other in both the concatenated and nuclear trees(Figs 2 S1 respectively) The diversification within thisclade is estimated to have occurred during the lateMiocene to early Pliocene when A b asper dispersedwidely west to North Africa and in Arabia The di-vergence within the North African group (group C2)is estimated to have occurred during the Pliocene ap-proximately 456 Mya with the Egyptian Nigerian andSudanese populations later dispersing west and north

14 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

in Africa This diversification correlates to the arid climatestarting in southern Sahara during the early-mid Plio-cene and later in northern Africa between the Plio-cene and the Pleistocene (Le Houeacuterou 1997) as hasbeen suggested for the dispersal of Mesalina guttulatain Africa (Kapli et al 2008) Other evidence relatesdry climate in North Africa to an earlier period around7 Mya (Schuster et al 2006) as has been suggestedfor the genus Chalcides and other reptiles (Carranzaet al 2008 Metallinou et al 2012 and reference therein)The aridification of North Africa has most likely con-tributed for the successful dispersal of A b asper westfrom south-west Asia into Africa Morphological studiesof A boskianus show relatively uniform populations inNorth Africa suggesting recent migration (Salvador1982 Arnold 1983) The other two geographical group-ings of A b asper from the Middle East and Arabia(groups C1 and C3) are located in two distinct innerclades but their location within each inner clade ispoorly supported The topology of the concatenated tree(Fig 2) shows that the group from northern Jordanand northern Oman (group C1) is closer to the NorthAfrican one than to the geographically close Middle-Eastern and south Arabian group (group C3) The taxo-nomic separation between north and south Oman hasbeen recognized in other species of reptiles and sup-ported by the topography of Oman (eg Echis coloratusand Echis omanensis Arnold Robinson amp Carranza2009) In the nuclear tree (Fig S1) these two groupsare closer to one another and with the North Africangroup form clade C Therefore the low support valuesamongst these groupings prevent an appropriate andthorough analysis of this subspecies The close rela-tionship amongst the geographical groups may reflectclose phylogenetic relationships amongst these popu-lations suggesting recent migration divergence andongoing gene flow

SYSTEMATICS AND TAXONOMIC IMPLICATIONS

The relationships within the A boskianus species groupconflict with the current known taxonomy of A schreiberiand A b asper (samples of the other subspecies ofA boskianus and of A nilsoni were unavailable for thisstudy) Both species have been found to be closelyrelated and paraphyletic The constrained topology testsexemplify the close entangled relationship between thetwo species as the separate monophyly of the two specieswas not rejected and the enforced monophyly of themboth together was inconclusive

Several causes can be responsible for paraphyly inspecies such in the A boskianus species group (Funkamp Omland 2003 and references therein) (1) inad-equate phylogenetic information (2) imperfect taxono-my (incorrectinaccurate species limits) derived frommisidentifying intraspecific variation (3) interspecific

gene flow ndash hybridization through interspecific matingand the subsequent backcrossing of hybrids into theparental populations (4) incomplete lineage sortingbecause of recent speciation events (5) unrecognizedparalogy We suggest that the relationships betweenA schreiberi and A b asper based on mitochondrialand nuclear data are most likely explained by incor-rect taxonomy probably because of the great variabil-ity of the latter species and to convergence As wasthe case in the molecular studies of the A pardalisand A erythrurus species groups (Harris et al 2004Fonseca et al 2008 2009 Carretero et al 2011 andreference therein) there are many problems with thecurrent taxonomic status of several species groups withinAcanthodactylus

Taking the molecular results of our study into accountthere are several systematic approaches to classify-ing the A schreiberiminusA b asper clade The Cypriot andTurkish populations of A schreiberi are very closelyrelated with the latter nested in the former and thetwo subspecies share nuclear alleles (Fig 3) Further-more the low uncorrected p-distance is positively cor-related with subspecies-level distances within otherlacertid species (ie 16 of Cytb in Lacerta bilineatachloronota Godinho et al 2005) We therefore con-clude that Cypriot animals were recently introducedto Turkey and that the Turkish population does notmerit a subspecific rank We suggest that A s ataturiYalccedilinkaya amp Goumlccedilmen 2012 is a junior synonym ofA s schreiberi Boulenger 1878

Regarding the relationships between A schreiberi andA b asper a few scenarios are possible One is to sinkA schreiberi within A boskianus to create one species(A boskianus) with high genetic and morphological vari-ability ranging over a broad distribution Another isfor the two taxa be regarded as a species complex (theA boskianus-schreiberi complex) until further inves-tigation on the subject However although A schreiberiis nested within A b asper the populations from Cyprusand Turkey represent a distinct evolutionary lineagewith distinct genetic and morphological features andthus it is logical to retain the specific status Two othersolutions are possible The first is to re-evaluate theSyrian populations and to consider elevating them aswell as the more divergent lineages (and subspecies)of A boskianus to specific status This would neces-sitate an examination of the phylogeny and morphol-ogy of the other four subspecies of A boskianus(A b boskianus A b euphraticus A b khattensis andA b nigeriensis) and the identification of distinctivephenotypic features in the Syrian lizards Another so-lution is to recognize the maintenance of gene flowamongst mainland populations of A b asper after thedivergence of the insular endemic A schreiberi andthus the evolutionary cohesion of the paraphyleticA b asper Arnold (1983) noted that A schreiberi may

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 15

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

have originated as an isolate from A boskianus becauseof their shared morphology and hemipenis featuresOur results support this scenario which includes thedispersal of A schreiberi to Cyprus from a mainlandpopulation that was most probably A boskianus It maybe assumed that the ancestor of the Cypriot A schreiberiafter arriving on Cyprus remained isolated for a longperiod of time and thus evolved to the modern formof A schreiberi Meanwhile the same ancestral con-tinental populations not isolated from each other con-tinued to exchange genes to varying degrees remainingA boskianus

The IsraeliminusLebanese subspecies A sc syriacus is onlydistantly related to the nominate form A sc schreiberiThis subspecies is highly phylogenetically divergent fromthe Cypriot and Turkish populations having higherp-distances (12S 4 Cytb 11ndash12) than those foundbetween other lacertid species (eg 74ndash82 of Cytbamongst Iberolacerta aranica Iberolacerta aurelioi andIberolacerta bonnali and 41ndash58 of Cytb betweenLacerta bilineata and Lacerta viridis Crochet et al2004 Godinho et al 2005 respectively) The nuclearhaplotype networks further show that Lebanese andIsraeli populations share alleles only with A b asperbut not with the nominotypical Cypriot form Arnold(1983) suggested that the geographical variation ofA boskianus reflects niche differences with animalsfrom xeric areas with dense rigid and spiny vegeta-tion having larger dorsal scales than animals from moremesic areas As was assumed for A schreiberi wesuggest that other mainland populations of A b asperwere the ancestors of the LebaneseminusIsraeli Coastal plainforms We suggest that A s syriacus is an ecomorphof A b asper that dispersed from the usual xerichabitats of the species and adapted to the new moremesic environment of the stable sands of the coastalplains of the eastern Mediterranean As a conse-quence this ecomorph converged on the morphologyof A s schreiberi which inhabits the coastal sands ofCyprus (Baier et al 2009) but still maintains differ-entiating features by having coarser dorsal scales andsharp keels (Salvador 1982 Arnold 1983 Franzen1998) This convergence led to the description ofA s syriacus as a member of A schreiberi The mor-phological assessment and the close morphological simi-larities between A b asper and A s syriacus mayexplain the wrong classification A similar erro-neous reasoning led Reed amp Marx (1959) to identifyspecimens with fine scales from Iraq as A schreiberiSalvador (1982) re-examined these specimens and as-signed them to A boskianus The morphological dif-ferences between the two forms are less prominentespecially where the two forms occur in close geo-graphical proximity in the southern coastal plainand north-western Negev Desert of Israel (Bar ampHaimovitch 2011) According to our results A s syriacus

actually belongs to A b asper being a coastal-duneecomorph convergent with but evolutionarily dis-tinct from A schreiberi Thus our preferred scenariois to treat the name Acanthodactylus schreiberi syriacusBoumlttger 1879 (which was originally described asA boskianus var syriacus by Boumlttger 1879) as a juniorsynonym of the name Acanthodactylus boskianus asper(Audouin 1827)

Recognizing A s syriacus as a junior synonym ofA b asper may have important implications for the con-servation of this coastal sand dune form which is clas-sified as critically endangered in Israel (Dolev ampPervolutzki 2004) However as the Israeli and Leba-nese coastal dune ecosystem has probably developedonly very recently during the Quaternary (Nir 1970Horowitz 1979) this form represents a remarkable caseof rapid evolutionary change It is also a remarkablecase of convergent evolution (with the CypriotA sc schreiberi) Thus we feel that these popula-tions are unique evolutionary entities that merit specialconservation efforts

The use of nuclear genes is a valuable method forestimating species divergence and lineage sorting andhelps evaluate isolated lineages and evolutionary historyThe incorporation of mitochondrial and nuclear dataprovides thorough topologies informative networks anddivergence times that reveal useful information for aproblematic taxonomy such as that of the A boskianusspecies group We have shown that phylogenetic ap-proaches to the confusing taxonomy of two closely relatedand morphologically similar species can shed light ontheir unclear relationships resolve between homoplasyand shared ancestry and identify patterns of speciesevolutionary history and biogeography

ACKNOWLEDGEMENTS

We are grateful to the following people for providingsamples for this study D Donaire B Shacham J ŠmiacutedS M Baha El Din and J Padial We thank MMetallinou and J Šmiacuted for help with the lab work andthe analyses M Novosolov for help with the figuresY L Werner B Shacham and A Levy for fruitful dis-cussions and two anonymous referees for helpful com-ments on an earlier version of this manuscript Wethank the Israel nature and park authority for issuingcollecting permits (nos 38074 38451 38489 38540)The work of J M was financially supported by the Min-istry of Culture of the Czech Republic (DKRVO 201315 National Museum 00023272) S C is funded bygrant CGL2012-36970 from the Ministerio de Economiacuteay Competitividad Spain (cofunded by FEDER) Thisstudy was funded by a Israel Taxonomic Initiative grantto S M K T is supported by an Israel Taxonomic Ini-tiative scholarship

16 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

REFERENCES

Ahmadzadeh F Carretero MA Harris DJ Perera ABohme W 2012 A molecular phylogeny of the eastern groupof ocellated lizard genus Timon (Sauria Lacertidae) basedon mitochondrial and nuclear DNA sequences Amphibia-Reptilia 33 1ndash10

Ahmadzadeh F Flecks M Roumldder D Boumlhme W Ilgaz CcedilHarris DJ Engler JO Uumlzuumlm N Carretero MA 2013Multiple dispersal out of Anatolia biogeography and evolu-tion of oriental green lizards Biological Journal of the LinneanSociety 110 398ndash408

Akaike H 1973 Information theory and an extension of themaximum likelihood principle In Petrov BN Csaki F edsSecond International Symposium on Information Theory Bu-dapest Akademiai Kiado 267ndash281

Amitai P Bouskila A 2001 Handbook of amphibians ampreptiles of Israel Israel Keter Publishing House Ltd

Anderson SC 1999 The lizards of Iran Ithaca NY Societyfor the Study of Amphibians and Reptiles

Arnold EN 1980 The scientific results of the Oman flora andfauna survey 1977 (Dhofar) The reptiles and amphibians ofDhofar southern Arabia Journal of Oman Studies SpecialReport 2 273ndash332

Arnold EN 1983 Osteology genitalia and the relationshipsof Acanthodactylus (Reptilia Lacertidae) Bulletin of the BritishMuseum (Natural History) Zoology 44 291ndash339

Arnold EN Arribas Oacute Carranza S 2007 Systematics ofthe Palaearctic and Oriental lizard tribe Lacertini (SquamataLacertidae Lacertinae) with descriptions of eight new generaZootaxa 1430 1ndash86

Arnold EN Robinson MD Carranza S 2009 A prelimi-nary analysis of phylogenetic relationships and biogeogra-phy of the dangerously venomous Carpet Vipers Echis(Squamata Serpentes Viperidae) based on mitochondrial DNAsequences Amphibia-Reptilia 30 273ndash282

Audouin JV 1827 Explication sommaire des planches dereptiles (suppleacutement) offrant un exposeacute des characteacuteresdes espegraveces In Savigny MJCL ed Description de lrsquoEacutegypteVol 1 Histoire Naturelle Paris Impeacuteriale 161ndash184

Bache F Popescu SM Rabineau M Gorini C Suc JPClauzon G Olivet JL Rubino JL Melinte-DobrinescuMC Estrada F 2012 A two-step process for the refloodingof the Mediterranean after the Messinian Salinity Crisis BasinResearch 24 125ndash153

Baha El Din SM 2006 A guide to the reptiles and amphib-ians of Egypt CairondashNew York American University in CairoPress

Baier FS Sparrow DJ Wiedl H-J 2009 The amphibiansand reptiles of Cyprus Frankfurt am Main Edition Chimaira

Bandelt H-J Forster P Roumlhl A 1999 Median-joining net-works for inferring intraspecific phylogenies Molecular Biologyand Evolution 16 37ndash48

Bar A Haimovitch G 2011 A field guide to reptiles and am-phibians of Israel Herzliya Pazbar Limited

Boumlttger O 1879 Reptilien und Amphibien aus Syrien Berichtuumlber die Senckenbergische Naturforschende Gesellschaft inFrankfurt am Main 1879 57ndash84

Boulenger GA 1919 On a new variety of Acanthodactylusboskianus Daud from the Euphrates The Annals and Maga-zine of Natural History 3 549ndash550

Boulenger GA 1878 Sur les espegraveces drsquoAcanthodactylus desbords de la Mediterraneacutee Bulletin de la Socieacuteteacute zoologiquede France 3 179ndash197

Boulenger GA 1918 Sur les leacutezards du genre AcanthodactylusWieg Bulletin de la Socieacuteteacute zoologique de France 43 143ndash155

Boulenger GA 1921 Monograph of the Lacertidœ Vol II Trus-tees of the British Museum of Natural History London

Carranza S Arnold EN 2012 A review of the geckos of thegenus Hemidactylus (Squamata Gekkonidae) fromOman based on morphology mitochondrial and nuclear datawith descriptions of eight new species Zootaxa 3378 1ndash95

Carranza S Arnold EN Geniez P Roca J Mateo J 2008Radiation multiple dispersal and parallelism in the skinksChalcides and Sphenops (Squamata Scincidae) with com-ments on Scincus and Scincopus and the age of the SaharaDesert Molecular Phylogenetics and Evolution 46 1071ndash1094

Carretero MA Fonseca MM Garcia-Munoz E Brito JCHarris DJ 2011 Adding Acanthodactylus beershebensis tothe mtDNA phylogeny of the Acanthodactylus pardalis groupNorth-Western Journal of Zoology 7 138ndash142

Castresana J 2000 Selection of conserved blocks from multi-ple alignments for their use in phylogenetic analysis Mo-lecular Biology and Evolution 17 540ndash552

Clube TMM Robertson A 1986 The palaeorotation of theTroodos microplate Cyprus in the Late Mesozoic-Early Ce-nozoic plate tectonic framework of the Eastern Mediterra-nean Surveys in Geophysics 8 375ndash437

Cox SC Carranza S Brown RP 2010 Divergence timesand colonization of the Canary Islands by Gallotia lizardsMolecular Phylogenetics and Evolution 56 747ndash757

Crochet PA Geniez P Ineich I 2003 A multivariate analy-sis of the fringe-toed lizards of the Acanthodactylus scutellatusgroup (Squamata Lacertidae) systematic and biogeographi-cal implications Zoological Journal of the Linnean Society137 117ndash155

Crochet P-A Chaline O Surget-Groba Y Debain CCheylan M 2004 Speciation in mountains phylogeographyand phylogeny of the rock lizards genus Iberolacerta (ReptiliaLacertidae) Molecular Phylogenetics and Evolution 30 860ndash866

Daudin FM 1802 Histoire naturelle geacuteneacuterale et particuliegraveredes reptiles ouvrage faisant suite agrave lrsquoHistoire naturelle geacuteneacuteraleet particuliegravere F Dufart

Disi A Modry D Necas P Rifai L 2001 Amphibians andreptiles of the Hashemite Kingdom of Jordan an atlas andfield guide Frankfurt am Main Chimaira

Dolev A Pervolutzki A 2004 Endangered species in IsraelRed list of threatened animals Vertebrates JerusalemThe Nature and Parks Authority and the Society for thePreservation of Nature

Drummond AJ Rambaut A 2007 BEAST Bayesian evo-lutionary analysis by sampling trees BMC EvolutionaryBiology 7 214

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 17

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Ezard T Fujisawa T Barraclough T 2009 Splits speciesrsquolimits by threshold statistics R package version 10-19r48Available at httpR-ForgeR-projectorgprojectssplits

Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783ndash791

Fitzinger LJFJ 1834 Acanthodactylus In Wiegman AFAed Herpetologia mexicana seu Descriptio amphibiorum NovaeHispaniae quae itineribus comitis De Sack Ferdinandi Deppeet Chr Guil Schiede in Museum zoologicum Berolinensepervenerunt Pars prima saurorum species amplectens adjectosystematis saurorum prodromo additisque multis in huncamphibiorum ordinem observationibus Berlin C G Luderitz10

Flot JF 2010 SeqPHASE a web tool for interconvertingPHASE inputoutput files and FASTA sequence align-ments Molecular Ecology Resources 10 162ndash166

Fonseca MM Brito JC Paulo OS Carretero MA HarrisDJ 2009 Systematic and phylogeographical assessment ofthe Acanthodactylus erythrurus group (Reptilia Lacertidae)based on phylogenetic analyses of mitochondrial and nuclearDNA Molecular Phylogenetics and Evolution 51 131ndash142

Fonseca MM Brito JC Rebelo H Kalboussi M LarbesS Carretero MA Harris DJ 2008 Genetic variation amongspiny-footed lizards in the Acanthodactylus pardalis groupfrom North Africa African Zoology 43 8ndash15

Fontaneto D Herniou EA Boschetti C Caprioli M MeloneG Ricci C Barraclough TG 2007 Independently evolv-ing species in asexual bdelloid rotifers PLoS Biology 5 e87

Franzen M 1998 Erstnachweis von Acanthodactylus schreiberischreiberi Boulenger 1879 fuumlr die Tuumlrkei (Squamata SauriaLacertidae) Herpetozoa 11 27ndash36

Funk DJ Omland KE 2003 Species-level paraphyly andpolyphyly frequency causes and consequences with in-sights from animal mitochondrial DNA Annual Review ofEcology Evolution and Systematics 34 397ndash423

Godinho R Crespo EG Ferrand N Harris DJ 2005 Phy-logeny and evolution of the green lizards Lacerta spp(Squamata Lacertidae) based on mitochondrial and nuclearDNA sequences Amphibia-Reptilia 26 271ndash285

Greenbaum E Villanueva CO Kusamba C Aristote MMBranch WR 2011 A molecular phylogeny of EquatorialAfrican Lacertidae with the description of a new genus andspecies from eastern Democratic Republic of the Congo Zoo-logical Journal of the Linnean Society 163 913ndash942

Guillou H Carracedo JC Torrado FP Badiola ER 1996K-Ar ages and magnetic stratigraphy of a hotspot-inducedfast grown oceanic island El Hierro Canary Islands Journalof Volcanology and Geothermal Research 73 141ndash155

Harris DJ Arnold EN 2000 Elucidation of the relation-ships of spiny-footed lizards Acanthodactylus spp (ReptiliaLacertidae) using mitochondrial DNA sequence with com-ments on their biogeography and evolution Journal of Zoology252 351ndash362

Harris DJ Batista V Carretero M 2004 Assessment ofgenetic diversity within Acanthodactylus erythrurus (ReptiliaLacertidae) in Morocco and the Iberian Peninsula usingmitochondrial DNA sequence data Amphibia-Reptilia 25227

Horowitz A 1979 The Quaternary of Israel New York Aca-demic Press

Hraoui-Bloquet S Sadek RA Sindaco R Venchi A 2002The herpetofauna of Lebanon new data on distributionZoology in the Middle East 27 35ndash46

Hsuuml KJ Montadert L Bernoulli D Cita MB EricksonA Garrison RE Kidd RB Megraveliereacutes F Muumlller C WrightR 1977 History of the Mediterranean salinity crisis Nature267 399ndash403

Huelsenbeck JP Rannala B 2004 Frequentist propertiesof Bayesian posterior probabilities of phylogenetic trees undersimple and complex substitution models Systematic Biology53 904ndash913

Huelsenbeck JP Ronquist F 2001 MRBAYES Bayesianinference of phylogenetic trees Bioinformatics 17 754ndash755

Jolivet L Augier R Robin C Suc J-P Rouchy JM 2006Lithospheric-scale geodynamic context of the Messinian sa-linity crisis Sedimentary Geology 188 9ndash33

Kapli P Lymberakis P Poulakakis N Mantziou GParmakelis A Mylonas M 2008 Molecular phylogeny ofthree Mesalina (Reptilia Lacertidae) species (M guttulataM brevirostris and M bahaeldini) from North Africa and theMiddle East another case of paraphyly MolecularPhylogenetics and Evolution 49 102ndash110

Katoh K Toh H 2008 Recent developments in the MAFFTmultiple sequence alignment program Briefings inBioinformatics 9 286ndash298

Krijgsman W Hilgen F Raffi I Sierro F Wilson D 1999Chronology causes and progression of the Messinian salin-ity crisis Nature 400 652ndash655

Lanfear R Calcott B Ho SY Guindon S 2012PartitionFinder combined selection of partitioning schemesand substitution models for phylogenetic analyses Molecu-lar Biology and Evolution 29 1695ndash1701

Le Houeacuterou HN 1997 Climate flora and fauna changes inthe Sahara over the past 500 million years Journal of AridEnvironments 37 619ndash647

Maca-Meyer N Carranza S Rando J Arnold E CabreraV 2003 Status and relationships of the extinct giant CanaryIsland lizard Gallotia goliath (Reptilia Lacertidae)assessed using ancient mtDNA from its mummifiedremains Biological Journal of the Linnean Society 80 659ndash670

McCallum J Robertson A 1990 Pulsed uplift of the TroodosMassif ndash evidence from the Plio-Pleistocene Mesaoria basinIn J Malpas EMM Panayiotou A Xenophontos C edsOphiolites oceanic crustal analogues Proceedings of theSymposium lsquoTroodos 1987rsquo Nicosia (The Geological SurveyDepartment Ministry of Agriculture and NaturalResources) 217ndash229

Metallinou M Arnold EN Crochet P-A Geniez P BritoJC Lymberakis P El Din SB Sindaco R Robinson MCarranza S 2012 Conquering the Sahara and Arabiandeserts systematics and biogeography of Stenodactylus geckos(Reptilia Gekkonidae) BMC Evolutionary Biology 12258

Monaghan MT Wild R Elliot M Fujisawa T Balke MInward DJ Lees DC Ranaivosolo R Eggleton P

18 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

10 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Together A b asper and A schreiberi form amonophyletic group within Acanthodactylus (Fig 2)Within the group however both taxa are paraphyleticwith A schreiberi as a whole nested within A b asperOur analyses distinguish three major clades (1) cladeA formed by A b asper from Syria (2) clade B in-cludes the two subspecies A sc ataturi from Turkeytogether with A sc schreiberi from Cyprus (3) cladeC which includes specimens of A b asper from the re-maining localities in its distribution range together withA sc syriacus from Israel and Lebanon Clade A is verywell supported and includes specimens of A b asperfrom central and northern Syria (Fig 1) splitting fromother specimens at the basal node of the group is es-timated to have occurred c 654 Mya [95 highest pos-terior density (HPD) 392ndash952 Mya] The level of geneticdifferentiation(p-distance) between these specimens and the remain-ing A b asper and all A schreiberi specimens is 37ndash46 for 12S and 107ndash119 for Cytb Clade B is alsovery well supported and includes two of the threenominal subspecies of A schreiberi A s schreiberi thenominotypical subspecies endemic to Cyprus and A sataturi from Turkey The Turkish subspecies is nestedwithin the Cypriot specimens and the two forms havelow genetic distances from each other (12S 016 Cytb123) This clade is nested between the two A b asperclades (clades A and C) in both the concatenatedand the nuclear tree although the nodes are not wellsupported Clade C is not very well supported It in-cludes a cluster of A b asper and A s syriacus Thisclade includes two inner clades that split around558 Mya (95 HPD 356ndash8 Mya) and divided into threepoorly supported geographical inner groups (Fig 2)northern Jordan and northern Oman (group C1) NorthAfrica (group C2) and samples from the Middle East(Egypt south Israel and south Jordan) with samplesfrom Yemen and southern Oman (group C3) ndash the latterincluding all specimens of the subspecies A s syriacusThe diversification within the North African group isestimated to have started around 456 Mya (95 HPD282ndash647 Mya) The IsraelminusLebanon endemic subspe-cies A s syriacus is genetically highly distinct from

A s schreiberi and A s ataturi making A schreiberiparaphyletic (p-distance 12S 431 416 Cytb 1181202 respectively)

The networks constructed for the phased haplotypesof the full length nuclear markers (MC1R ACM4 andc-mos) are presented in Figure 3 The nuclear networkanalyses show similar results for each of the three genesand closely agree with the phylogenetic tree The CypriotA s schreiberi and Turkish A s ataturi subspecies sharealleles for all three genes and both are distinct fromthe third subspecies A s syriacus Acanthodactylusschreiberi syriacus shares no alleles with the other sub-species of A schreiberi but does share alleles withA b asper for each of the genes Acanthodactylusschreiberi syriacus shares MC1R alleles with A b asperspecimens from Tunisia Syria and Israel ACM4 alleleswith Egyptian Israeli Jordanian and North Africanspecimens and c-mos alleles only with Israeli A b asperspecimens Syrian A b asper samples share one allelewith A s syriacus and two with other A b asper speci-mens from Egypt Israel and North Africa in the MC1Rone allele with an Egyptian A b asper in the ACM4and none in the c-mos gene

In order to better understand the relationships betweenA schreiberi and A boskianus we performed three to-pology tests in which we forced monophyletic group-ings (1) monophyly of A schreiberi (all three subspeciestogether) (2) monophyly of A b asper (3) monophylyof A b asper and of A schreiberi The results of thetopological tests indicate that our data set cannot rejectthe alternative hypotheses of monophyly of A schreiberi(AU P = 0091 SH P = 0062) and that of A b asper(AU P = 011 SH P = 0072) if we allow A schreiberito nest within A b asper or a monophyletic A b aspernesting within A schreiberi When forcing monophylyof both A schreiberi and of A b asper together in thesame tree the results are inconclusive (AU P = 0046SH P = 0051)

The single-threshold model in GMYC yield a topol-ogy that is clearly different from the known taxono-my The GMYC results present a total of 25 and 24ML independent lineages from the Bayesian haplotypemitochondrial phylogeny of the two species for the two

Figure 2 Maximum likelihood (ML) tree of the Acanthodactylus boskianus and Acanthodactylus schreiberi specimensinferred using 12S rRNA cytochrome b mtDNA and melano-cortin 1 receptor acetylcholinergic receptor M4 and oocytematuration factor MOS nuclear gene fragments Posterior probability in the Bayesian analysis is indicated by black dotson the nodes [values ge 095 shown for both gene partitions and partitions by PartitionFinder (PF)] and ML bootstrapsupport values are indicated in parentheses (values ge 70 shown ML ML-PF) Age estimates with BEAST are indicat-ed near the relevant nodes and include the mean and in brackets the HPD 95 confidence interval Sample codes relateto specimens in Table 1 and in Figures 1 and 3 Colours blue Acanthodactylus boskianus asper yellow Acanthodactylusschreiberi ataturi red Acanthodactylus schreiberi syriacus green Acanthodactylus schreiberi schreiberi (Colour versionof figure available online)

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 11

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

12 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

partition approaches (ie by gene and by PartitionFinderFigs S2 S3 respectively) The two partition ap-proaches gave similar clusters but at the less sup-ported nodes they differed at the positions of severallineages The single threshold GMYC result is indi-cated for a single line at 00037 Mya for the gene par-titions and at 002 Mya for PartitionFinder (verticallines in Figs S2 S3) The topology and clusters re-vealed in this analysis correspond to the lineages fromthe phylogeny of the ML and BI methods both for theparaphyly of the two species and the geographical group-ings within A b asper The GMYC results mainly differfrom the ML and BI methods in the position ofA schreiberi from Cyprus and Turkey as a sister cladeto the Syrian A b asper

DISCUSSION

We have provided a comprehensive and thorough as-sessment of the intraspecific phylogenetic relation-ships within A schreiberi and its closest relativeA b asper Our results based on mitochondrial andnuclear DNA data from 84 specimens across the entiredistribution range of A schreiberi and most of the dis-tribution range of A b asper reveal that A schreiberiis paraphyletic and nested entirely within theA boskianus subspecies

HISTORICAL BIOGEOGRAPHY

Acanthodactylus schreiberi is thought to comprisethree subspecies corresponding to three allopatricpopulations in Cyprus Turkey and IsraelminusLebanonThe Cypriot endemic nominotypical subspeciesA sc schreiberi and the Turkish subspeciesA sc ataturi cluster together (to form clade B Fig 2)nesting between A b asper clades This lineage is sisterto a clade of A b asper including A sc syriacus (cladeC Fig 2) We estimate the divergence time of theCypriotminusTurkish lineage of A schreiberi to have beenduring the late Miocene around 6 Mya although thereis no support for this split in the tree In other analy-ses using the whole genus this split is well support-ed in Bayesian analyses (K Tamar S Carranza RSindaco J Moravec JF Trape amp S Meriri unpubldata) Based on mitochondrial data Poulakakis et al(2013) found that both the Cypriot and Turkish sub-species are monophyletic and diverged from each other085 Mya (038ndash156 Mya) According to our results thisdate corresponds to an inner divergence of the

A sc schreiberi lineage rather than to the date at whichA sc schreiberi colonized Cyprus

The discrepancy in the phylogenetic relationship ofA sc schreiberi raises questions regarding the arrivalon Cyprus Cyprus originated with the raising of theTroodos Massif during the upper Cretaceous c 91 to88 Mya (Clube amp Robertson 1986 Mukasa amp Ludden1987) During the middle to late Miocene only a smallproportion of Cyprus was exposed above the Mediter-ranean (McCallum amp Robertson 1990 Robertson 1990)Towards the end of the Miocene sim596 Mya with theclosing of the passage between the Atlantic Ocean andthe Mediterranean basin the Messinian salinity crisisbegan (Krijgsman et al 1999) This resulted in thedrying up of much of the Mediterranean Sea and highsea-mounts emerged to form land bridges with thesurrounding land (Hsuuml et al 1977) By the end of theMiocene and early Pliocene sim533 Mya the passagewith the Atlantic Ocean reopened and the Mediterra-nean basin was refilled (Krijgsman et al 1999) Re-sulting from compressions raising and uplifting of thesurrounding areas towards and during the Pleisto-cene Cyprus was a complete emergent island (McCallumamp Robertson 1990) The possible connection of Cyprusto the mainland (ie to TurkeySyria) during theMessinian is debated as are suggestions of a land con-nection at later periods (Steininger amp Roumlgl 1984 Jolivetet al 2006 Bache et al 2012) Such a connection ifit existed could have provided access for terrestrialorganisms with poor overseas dispersal ability suchas lizards to colonize the island Several studies arguethat post-Messinian sea level changes are unlikely tohave formed connections between Cyprus and the main-land (Steininger amp Roumlgl 1984 Jolivet et al 2006) Thusour dating of the split between the Cypriot A schreiberiand A b asper at c 6 Mya leads us to suggest that theancestor of A s schreiberi colonized Cyprus from themainland through a land bridge connection at the be-ginning of the Messinian crisis rather than by a muchlatermore recent transmarine dispersal as suggestedby Poulakakis et al (2013) Owing to its close rela-tions with A b asper the ancestor of A schreiberi waspresumably mainland A boskianus and the cladogenesisleading to A schreiberi thus rendered A b asperparaphyletic

The Turkish subspecies A schreiberi ataturi was rec-orded for the first time by Franzen (1998) at a veryrestricted area of around 15 km of coastal strip (betweenBotas and Yukarı Burnaz Hatay Province) Owing tothe remarkable morphological similarity between

Figure 3 Haplotype networks of the nuclear gene fragments melano-cortin 1 receptor (MC1R) acetylcholinergic recep-tor M4 (ACM4) and oocyte maturation factor MOS (c-mos) with colours corresponding to Figures 1 and 2 Codes corre-late to the two alleles (ie a and b) of specimens in Table 1 Circle sizes are proportional to the number of alleles (Colourversion of figure available online)

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 13

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

A sc ataturi and the Cypriot population the speci-mens were initially identified as A sc schreiberi(Franzen 1998) Yalccedilinkaya amp Goumlccedilmen (2012) howeverdescribed this population as a new distinct subspe-cies A s ataturi presenting several differences betweenthe two in both morphology and blood-serum pro-teins The origin of A s ataturi remains uncertain asit is debated whether the newly discovered Turkishpopulation is a relict or an introduction from CyprusFranzen (1998) described this population as a pos-sible introduction from Cyprus through the harbourof Botas but Sindaco et al (2000) suggested thatit might be a relict of a previously larger popula-tion because its present distribution is similar tothat of some insects and lizards [Archaeolacerta(Phoenicolacerta) laevis and Ablepharus budaki]Yalccedilinkaya amp Goumlccedilmen (2012) proposed that A sc ataturiarrived in Turkey from the nominate population inCyprus during the Messinian crisis The phylogeneticresults haplotype networks and low levels of geneticdivergence we found suggest that the two subspeciesfrom Cyprus and Turkey have not been geneticallyisolated for a long period of time (ie they share allelesin all three nuclear genes and A s ataturi is nestedwithin A s schreiberi in the phylogeny Figs 2 3) Ourresults therefore contrast with the two latter sce-narios of a relict population or a Messinian disper-sal Both divergence time and the genetic similarityof the two subspecies agree with the original sugges-tion of Franzen (1998) that these animals were intro-duced into Turkey from Cyprus Further support forthis hypothesis is that A s ataturi is restricted to thevicinity of the Botas-Adana harbour and is absent inother suitable habitats (coastal sand dunes) wide-spread in south-eastern Turkey Its close morphologi-cal features to A s schreiberi (Franzen 1998) likewisesupport an introduction scenario

The third subspecies A s syriacus is nested withinA b asper in the concatenated mtDNA and nDNA treesand is clearly genetically distinct from the Cyprus andTurkey A schreiberi lineage The close relations ofA s syriacus with A boskianus may shed light on theorigin of the former Acanthodactylus schreiberi syriacusis distributed on stable sands of the coastal plain ofthe eastern Mediterranean in Israel and southernLebanon (Salvador 1982 Hraoui-Bloquet et al 2002Bar amp Haimovitch 2011) habitats resembling those ofA s schreiberi from Cyprus (Baier Sparrow amp Wiedl2009) The oldest divergence of the A b asper cladethat includes A s syriacus is estimated to have oc-curred during the late Miocene around 558 Mya butno further dates are available for the grouping ofA s syriacus as a result of low support values Thecoastal plain of the eastern Mediterranean was sub-merged during the late Miocene and re-emerged onlytoward the Pliocene (Nir 1970 Horowitz 1979) The

sands of the coastal plain where A s syriacus occurs(Salvador 1982 Arnold 1983 K Tamar amp S Meiripers observ) were repeatedly submerged and re-emerged during the Pleistocene sea-level changes (duringinterglacial and glacial periods respectively) A pos-sible scenario for A sc syriacusrsquos origin includes severalwaves of dispersal of Middle Eastern A boskianus whichoccurs on coarse substrates (Amitai amp Bouskila 2001Disi et al 2001 Baha El Din 2006 pers observ) towardthe Mediterranean shore Acanthodactylus boskianusasper is absent from Mediterranean climate habitatsin Lebanon and Israel It occupies only xeric zonessuggesting an invasion to the coastal plain when sandyhabitats allowed desert flora and fauna to migrate north-wards (Yom-Tov 1988) These populations adapted tosandy soils and evolved morphological features thatdistinguish them from the desert hard substrate formsof A b asper We view this as the most likely scenariogiven the biogeography the phylogenetic results andthe habitat preferences and adaptations of these lizardsAn alternative scenario according to which the an-cestor of A schreiberi originated in Cyprus and dis-persed to the shores of Israel and Lebanon (or originatedin the coastal plain of the Eastern Mediterranean anddispersed to Cyprus) we regard as far less likely Sucha scenario requires much closer genetic relationshipsbetween these two forms and is further weakened bythe close relationship between A s syriacus and thegeographically adjacent A b asper populations

Acanthodactylus boskianus asper is highly variableboth morphologically (Salvador 1982 Arnold 1983) andgenetically (this study) The subspecies is paraphyleticas A schreiberi is nested within it The topology of theA b asper tree shows four different geographical group-ings Syria (clade A) north Jordan plus north OmanNorth Africa and Middle East plus south Arabia (groupsC1 C2 C3 respectively) The different groups in thissubspecies are estimated to have first diverged duringthe late Miocene approximately 65 Mya with the splitof the Syrian population The Syrian lineage is ge-netically distant from A schreiberi and the otherA b asper specimens The nuclear networks indicatethat this group is closer to the other A b asper samplesrather than to A schreiberi The geographical splitsin the rest of the A b asper range (clade C) are esti-mated to have started around 558 Mya These groupsare supported as a distinct clade but are closely relatedto each other in both the concatenated and nuclear trees(Figs 2 S1 respectively) The diversification within thisclade is estimated to have occurred during the lateMiocene to early Pliocene when A b asper dispersedwidely west to North Africa and in Arabia The di-vergence within the North African group (group C2)is estimated to have occurred during the Pliocene ap-proximately 456 Mya with the Egyptian Nigerian andSudanese populations later dispersing west and north

14 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

in Africa This diversification correlates to the arid climatestarting in southern Sahara during the early-mid Plio-cene and later in northern Africa between the Plio-cene and the Pleistocene (Le Houeacuterou 1997) as hasbeen suggested for the dispersal of Mesalina guttulatain Africa (Kapli et al 2008) Other evidence relatesdry climate in North Africa to an earlier period around7 Mya (Schuster et al 2006) as has been suggestedfor the genus Chalcides and other reptiles (Carranzaet al 2008 Metallinou et al 2012 and reference therein)The aridification of North Africa has most likely con-tributed for the successful dispersal of A b asper westfrom south-west Asia into Africa Morphological studiesof A boskianus show relatively uniform populations inNorth Africa suggesting recent migration (Salvador1982 Arnold 1983) The other two geographical group-ings of A b asper from the Middle East and Arabia(groups C1 and C3) are located in two distinct innerclades but their location within each inner clade ispoorly supported The topology of the concatenated tree(Fig 2) shows that the group from northern Jordanand northern Oman (group C1) is closer to the NorthAfrican one than to the geographically close Middle-Eastern and south Arabian group (group C3) The taxo-nomic separation between north and south Oman hasbeen recognized in other species of reptiles and sup-ported by the topography of Oman (eg Echis coloratusand Echis omanensis Arnold Robinson amp Carranza2009) In the nuclear tree (Fig S1) these two groupsare closer to one another and with the North Africangroup form clade C Therefore the low support valuesamongst these groupings prevent an appropriate andthorough analysis of this subspecies The close rela-tionship amongst the geographical groups may reflectclose phylogenetic relationships amongst these popu-lations suggesting recent migration divergence andongoing gene flow

SYSTEMATICS AND TAXONOMIC IMPLICATIONS

The relationships within the A boskianus species groupconflict with the current known taxonomy of A schreiberiand A b asper (samples of the other subspecies ofA boskianus and of A nilsoni were unavailable for thisstudy) Both species have been found to be closelyrelated and paraphyletic The constrained topology testsexemplify the close entangled relationship between thetwo species as the separate monophyly of the two specieswas not rejected and the enforced monophyly of themboth together was inconclusive

Several causes can be responsible for paraphyly inspecies such in the A boskianus species group (Funkamp Omland 2003 and references therein) (1) inad-equate phylogenetic information (2) imperfect taxono-my (incorrectinaccurate species limits) derived frommisidentifying intraspecific variation (3) interspecific

gene flow ndash hybridization through interspecific matingand the subsequent backcrossing of hybrids into theparental populations (4) incomplete lineage sortingbecause of recent speciation events (5) unrecognizedparalogy We suggest that the relationships betweenA schreiberi and A b asper based on mitochondrialand nuclear data are most likely explained by incor-rect taxonomy probably because of the great variabil-ity of the latter species and to convergence As wasthe case in the molecular studies of the A pardalisand A erythrurus species groups (Harris et al 2004Fonseca et al 2008 2009 Carretero et al 2011 andreference therein) there are many problems with thecurrent taxonomic status of several species groups withinAcanthodactylus

Taking the molecular results of our study into accountthere are several systematic approaches to classify-ing the A schreiberiminusA b asper clade The Cypriot andTurkish populations of A schreiberi are very closelyrelated with the latter nested in the former and thetwo subspecies share nuclear alleles (Fig 3) Further-more the low uncorrected p-distance is positively cor-related with subspecies-level distances within otherlacertid species (ie 16 of Cytb in Lacerta bilineatachloronota Godinho et al 2005) We therefore con-clude that Cypriot animals were recently introducedto Turkey and that the Turkish population does notmerit a subspecific rank We suggest that A s ataturiYalccedilinkaya amp Goumlccedilmen 2012 is a junior synonym ofA s schreiberi Boulenger 1878

Regarding the relationships between A schreiberi andA b asper a few scenarios are possible One is to sinkA schreiberi within A boskianus to create one species(A boskianus) with high genetic and morphological vari-ability ranging over a broad distribution Another isfor the two taxa be regarded as a species complex (theA boskianus-schreiberi complex) until further inves-tigation on the subject However although A schreiberiis nested within A b asper the populations from Cyprusand Turkey represent a distinct evolutionary lineagewith distinct genetic and morphological features andthus it is logical to retain the specific status Two othersolutions are possible The first is to re-evaluate theSyrian populations and to consider elevating them aswell as the more divergent lineages (and subspecies)of A boskianus to specific status This would neces-sitate an examination of the phylogeny and morphol-ogy of the other four subspecies of A boskianus(A b boskianus A b euphraticus A b khattensis andA b nigeriensis) and the identification of distinctivephenotypic features in the Syrian lizards Another so-lution is to recognize the maintenance of gene flowamongst mainland populations of A b asper after thedivergence of the insular endemic A schreiberi andthus the evolutionary cohesion of the paraphyleticA b asper Arnold (1983) noted that A schreiberi may

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 15

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

have originated as an isolate from A boskianus becauseof their shared morphology and hemipenis featuresOur results support this scenario which includes thedispersal of A schreiberi to Cyprus from a mainlandpopulation that was most probably A boskianus It maybe assumed that the ancestor of the Cypriot A schreiberiafter arriving on Cyprus remained isolated for a longperiod of time and thus evolved to the modern formof A schreiberi Meanwhile the same ancestral con-tinental populations not isolated from each other con-tinued to exchange genes to varying degrees remainingA boskianus

The IsraeliminusLebanese subspecies A sc syriacus is onlydistantly related to the nominate form A sc schreiberiThis subspecies is highly phylogenetically divergent fromthe Cypriot and Turkish populations having higherp-distances (12S 4 Cytb 11ndash12) than those foundbetween other lacertid species (eg 74ndash82 of Cytbamongst Iberolacerta aranica Iberolacerta aurelioi andIberolacerta bonnali and 41ndash58 of Cytb betweenLacerta bilineata and Lacerta viridis Crochet et al2004 Godinho et al 2005 respectively) The nuclearhaplotype networks further show that Lebanese andIsraeli populations share alleles only with A b asperbut not with the nominotypical Cypriot form Arnold(1983) suggested that the geographical variation ofA boskianus reflects niche differences with animalsfrom xeric areas with dense rigid and spiny vegeta-tion having larger dorsal scales than animals from moremesic areas As was assumed for A schreiberi wesuggest that other mainland populations of A b asperwere the ancestors of the LebaneseminusIsraeli Coastal plainforms We suggest that A s syriacus is an ecomorphof A b asper that dispersed from the usual xerichabitats of the species and adapted to the new moremesic environment of the stable sands of the coastalplains of the eastern Mediterranean As a conse-quence this ecomorph converged on the morphologyof A s schreiberi which inhabits the coastal sands ofCyprus (Baier et al 2009) but still maintains differ-entiating features by having coarser dorsal scales andsharp keels (Salvador 1982 Arnold 1983 Franzen1998) This convergence led to the description ofA s syriacus as a member of A schreiberi The mor-phological assessment and the close morphological simi-larities between A b asper and A s syriacus mayexplain the wrong classification A similar erro-neous reasoning led Reed amp Marx (1959) to identifyspecimens with fine scales from Iraq as A schreiberiSalvador (1982) re-examined these specimens and as-signed them to A boskianus The morphological dif-ferences between the two forms are less prominentespecially where the two forms occur in close geo-graphical proximity in the southern coastal plainand north-western Negev Desert of Israel (Bar ampHaimovitch 2011) According to our results A s syriacus

actually belongs to A b asper being a coastal-duneecomorph convergent with but evolutionarily dis-tinct from A schreiberi Thus our preferred scenariois to treat the name Acanthodactylus schreiberi syriacusBoumlttger 1879 (which was originally described asA boskianus var syriacus by Boumlttger 1879) as a juniorsynonym of the name Acanthodactylus boskianus asper(Audouin 1827)

Recognizing A s syriacus as a junior synonym ofA b asper may have important implications for the con-servation of this coastal sand dune form which is clas-sified as critically endangered in Israel (Dolev ampPervolutzki 2004) However as the Israeli and Leba-nese coastal dune ecosystem has probably developedonly very recently during the Quaternary (Nir 1970Horowitz 1979) this form represents a remarkable caseof rapid evolutionary change It is also a remarkablecase of convergent evolution (with the CypriotA sc schreiberi) Thus we feel that these popula-tions are unique evolutionary entities that merit specialconservation efforts

The use of nuclear genes is a valuable method forestimating species divergence and lineage sorting andhelps evaluate isolated lineages and evolutionary historyThe incorporation of mitochondrial and nuclear dataprovides thorough topologies informative networks anddivergence times that reveal useful information for aproblematic taxonomy such as that of the A boskianusspecies group We have shown that phylogenetic ap-proaches to the confusing taxonomy of two closely relatedand morphologically similar species can shed light ontheir unclear relationships resolve between homoplasyand shared ancestry and identify patterns of speciesevolutionary history and biogeography

ACKNOWLEDGEMENTS

We are grateful to the following people for providingsamples for this study D Donaire B Shacham J ŠmiacutedS M Baha El Din and J Padial We thank MMetallinou and J Šmiacuted for help with the lab work andthe analyses M Novosolov for help with the figuresY L Werner B Shacham and A Levy for fruitful dis-cussions and two anonymous referees for helpful com-ments on an earlier version of this manuscript Wethank the Israel nature and park authority for issuingcollecting permits (nos 38074 38451 38489 38540)The work of J M was financially supported by the Min-istry of Culture of the Czech Republic (DKRVO 201315 National Museum 00023272) S C is funded bygrant CGL2012-36970 from the Ministerio de Economiacuteay Competitividad Spain (cofunded by FEDER) Thisstudy was funded by a Israel Taxonomic Initiative grantto S M K T is supported by an Israel Taxonomic Ini-tiative scholarship

16 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

REFERENCES

Ahmadzadeh F Carretero MA Harris DJ Perera ABohme W 2012 A molecular phylogeny of the eastern groupof ocellated lizard genus Timon (Sauria Lacertidae) basedon mitochondrial and nuclear DNA sequences Amphibia-Reptilia 33 1ndash10

Ahmadzadeh F Flecks M Roumldder D Boumlhme W Ilgaz CcedilHarris DJ Engler JO Uumlzuumlm N Carretero MA 2013Multiple dispersal out of Anatolia biogeography and evolu-tion of oriental green lizards Biological Journal of the LinneanSociety 110 398ndash408

Akaike H 1973 Information theory and an extension of themaximum likelihood principle In Petrov BN Csaki F edsSecond International Symposium on Information Theory Bu-dapest Akademiai Kiado 267ndash281

Amitai P Bouskila A 2001 Handbook of amphibians ampreptiles of Israel Israel Keter Publishing House Ltd

Anderson SC 1999 The lizards of Iran Ithaca NY Societyfor the Study of Amphibians and Reptiles

Arnold EN 1980 The scientific results of the Oman flora andfauna survey 1977 (Dhofar) The reptiles and amphibians ofDhofar southern Arabia Journal of Oman Studies SpecialReport 2 273ndash332

Arnold EN 1983 Osteology genitalia and the relationshipsof Acanthodactylus (Reptilia Lacertidae) Bulletin of the BritishMuseum (Natural History) Zoology 44 291ndash339

Arnold EN Arribas Oacute Carranza S 2007 Systematics ofthe Palaearctic and Oriental lizard tribe Lacertini (SquamataLacertidae Lacertinae) with descriptions of eight new generaZootaxa 1430 1ndash86

Arnold EN Robinson MD Carranza S 2009 A prelimi-nary analysis of phylogenetic relationships and biogeogra-phy of the dangerously venomous Carpet Vipers Echis(Squamata Serpentes Viperidae) based on mitochondrial DNAsequences Amphibia-Reptilia 30 273ndash282

Audouin JV 1827 Explication sommaire des planches dereptiles (suppleacutement) offrant un exposeacute des characteacuteresdes espegraveces In Savigny MJCL ed Description de lrsquoEacutegypteVol 1 Histoire Naturelle Paris Impeacuteriale 161ndash184

Bache F Popescu SM Rabineau M Gorini C Suc JPClauzon G Olivet JL Rubino JL Melinte-DobrinescuMC Estrada F 2012 A two-step process for the refloodingof the Mediterranean after the Messinian Salinity Crisis BasinResearch 24 125ndash153

Baha El Din SM 2006 A guide to the reptiles and amphib-ians of Egypt CairondashNew York American University in CairoPress

Baier FS Sparrow DJ Wiedl H-J 2009 The amphibiansand reptiles of Cyprus Frankfurt am Main Edition Chimaira

Bandelt H-J Forster P Roumlhl A 1999 Median-joining net-works for inferring intraspecific phylogenies Molecular Biologyand Evolution 16 37ndash48

Bar A Haimovitch G 2011 A field guide to reptiles and am-phibians of Israel Herzliya Pazbar Limited

Boumlttger O 1879 Reptilien und Amphibien aus Syrien Berichtuumlber die Senckenbergische Naturforschende Gesellschaft inFrankfurt am Main 1879 57ndash84

Boulenger GA 1919 On a new variety of Acanthodactylusboskianus Daud from the Euphrates The Annals and Maga-zine of Natural History 3 549ndash550

Boulenger GA 1878 Sur les espegraveces drsquoAcanthodactylus desbords de la Mediterraneacutee Bulletin de la Socieacuteteacute zoologiquede France 3 179ndash197

Boulenger GA 1918 Sur les leacutezards du genre AcanthodactylusWieg Bulletin de la Socieacuteteacute zoologique de France 43 143ndash155

Boulenger GA 1921 Monograph of the Lacertidœ Vol II Trus-tees of the British Museum of Natural History London

Carranza S Arnold EN 2012 A review of the geckos of thegenus Hemidactylus (Squamata Gekkonidae) fromOman based on morphology mitochondrial and nuclear datawith descriptions of eight new species Zootaxa 3378 1ndash95

Carranza S Arnold EN Geniez P Roca J Mateo J 2008Radiation multiple dispersal and parallelism in the skinksChalcides and Sphenops (Squamata Scincidae) with com-ments on Scincus and Scincopus and the age of the SaharaDesert Molecular Phylogenetics and Evolution 46 1071ndash1094

Carretero MA Fonseca MM Garcia-Munoz E Brito JCHarris DJ 2011 Adding Acanthodactylus beershebensis tothe mtDNA phylogeny of the Acanthodactylus pardalis groupNorth-Western Journal of Zoology 7 138ndash142

Castresana J 2000 Selection of conserved blocks from multi-ple alignments for their use in phylogenetic analysis Mo-lecular Biology and Evolution 17 540ndash552

Clube TMM Robertson A 1986 The palaeorotation of theTroodos microplate Cyprus in the Late Mesozoic-Early Ce-nozoic plate tectonic framework of the Eastern Mediterra-nean Surveys in Geophysics 8 375ndash437

Cox SC Carranza S Brown RP 2010 Divergence timesand colonization of the Canary Islands by Gallotia lizardsMolecular Phylogenetics and Evolution 56 747ndash757

Crochet PA Geniez P Ineich I 2003 A multivariate analy-sis of the fringe-toed lizards of the Acanthodactylus scutellatusgroup (Squamata Lacertidae) systematic and biogeographi-cal implications Zoological Journal of the Linnean Society137 117ndash155

Crochet P-A Chaline O Surget-Groba Y Debain CCheylan M 2004 Speciation in mountains phylogeographyand phylogeny of the rock lizards genus Iberolacerta (ReptiliaLacertidae) Molecular Phylogenetics and Evolution 30 860ndash866

Daudin FM 1802 Histoire naturelle geacuteneacuterale et particuliegraveredes reptiles ouvrage faisant suite agrave lrsquoHistoire naturelle geacuteneacuteraleet particuliegravere F Dufart

Disi A Modry D Necas P Rifai L 2001 Amphibians andreptiles of the Hashemite Kingdom of Jordan an atlas andfield guide Frankfurt am Main Chimaira

Dolev A Pervolutzki A 2004 Endangered species in IsraelRed list of threatened animals Vertebrates JerusalemThe Nature and Parks Authority and the Society for thePreservation of Nature

Drummond AJ Rambaut A 2007 BEAST Bayesian evo-lutionary analysis by sampling trees BMC EvolutionaryBiology 7 214

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 17

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Ezard T Fujisawa T Barraclough T 2009 Splits speciesrsquolimits by threshold statistics R package version 10-19r48Available at httpR-ForgeR-projectorgprojectssplits

Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783ndash791

Fitzinger LJFJ 1834 Acanthodactylus In Wiegman AFAed Herpetologia mexicana seu Descriptio amphibiorum NovaeHispaniae quae itineribus comitis De Sack Ferdinandi Deppeet Chr Guil Schiede in Museum zoologicum Berolinensepervenerunt Pars prima saurorum species amplectens adjectosystematis saurorum prodromo additisque multis in huncamphibiorum ordinem observationibus Berlin C G Luderitz10

Flot JF 2010 SeqPHASE a web tool for interconvertingPHASE inputoutput files and FASTA sequence align-ments Molecular Ecology Resources 10 162ndash166

Fonseca MM Brito JC Paulo OS Carretero MA HarrisDJ 2009 Systematic and phylogeographical assessment ofthe Acanthodactylus erythrurus group (Reptilia Lacertidae)based on phylogenetic analyses of mitochondrial and nuclearDNA Molecular Phylogenetics and Evolution 51 131ndash142

Fonseca MM Brito JC Rebelo H Kalboussi M LarbesS Carretero MA Harris DJ 2008 Genetic variation amongspiny-footed lizards in the Acanthodactylus pardalis groupfrom North Africa African Zoology 43 8ndash15

Fontaneto D Herniou EA Boschetti C Caprioli M MeloneG Ricci C Barraclough TG 2007 Independently evolv-ing species in asexual bdelloid rotifers PLoS Biology 5 e87

Franzen M 1998 Erstnachweis von Acanthodactylus schreiberischreiberi Boulenger 1879 fuumlr die Tuumlrkei (Squamata SauriaLacertidae) Herpetozoa 11 27ndash36

Funk DJ Omland KE 2003 Species-level paraphyly andpolyphyly frequency causes and consequences with in-sights from animal mitochondrial DNA Annual Review ofEcology Evolution and Systematics 34 397ndash423

Godinho R Crespo EG Ferrand N Harris DJ 2005 Phy-logeny and evolution of the green lizards Lacerta spp(Squamata Lacertidae) based on mitochondrial and nuclearDNA sequences Amphibia-Reptilia 26 271ndash285

Greenbaum E Villanueva CO Kusamba C Aristote MMBranch WR 2011 A molecular phylogeny of EquatorialAfrican Lacertidae with the description of a new genus andspecies from eastern Democratic Republic of the Congo Zoo-logical Journal of the Linnean Society 163 913ndash942

Guillou H Carracedo JC Torrado FP Badiola ER 1996K-Ar ages and magnetic stratigraphy of a hotspot-inducedfast grown oceanic island El Hierro Canary Islands Journalof Volcanology and Geothermal Research 73 141ndash155

Harris DJ Arnold EN 2000 Elucidation of the relation-ships of spiny-footed lizards Acanthodactylus spp (ReptiliaLacertidae) using mitochondrial DNA sequence with com-ments on their biogeography and evolution Journal of Zoology252 351ndash362

Harris DJ Batista V Carretero M 2004 Assessment ofgenetic diversity within Acanthodactylus erythrurus (ReptiliaLacertidae) in Morocco and the Iberian Peninsula usingmitochondrial DNA sequence data Amphibia-Reptilia 25227

Horowitz A 1979 The Quaternary of Israel New York Aca-demic Press

Hraoui-Bloquet S Sadek RA Sindaco R Venchi A 2002The herpetofauna of Lebanon new data on distributionZoology in the Middle East 27 35ndash46

Hsuuml KJ Montadert L Bernoulli D Cita MB EricksonA Garrison RE Kidd RB Megraveliereacutes F Muumlller C WrightR 1977 History of the Mediterranean salinity crisis Nature267 399ndash403

Huelsenbeck JP Rannala B 2004 Frequentist propertiesof Bayesian posterior probabilities of phylogenetic trees undersimple and complex substitution models Systematic Biology53 904ndash913

Huelsenbeck JP Ronquist F 2001 MRBAYES Bayesianinference of phylogenetic trees Bioinformatics 17 754ndash755

Jolivet L Augier R Robin C Suc J-P Rouchy JM 2006Lithospheric-scale geodynamic context of the Messinian sa-linity crisis Sedimentary Geology 188 9ndash33

Kapli P Lymberakis P Poulakakis N Mantziou GParmakelis A Mylonas M 2008 Molecular phylogeny ofthree Mesalina (Reptilia Lacertidae) species (M guttulataM brevirostris and M bahaeldini) from North Africa and theMiddle East another case of paraphyly MolecularPhylogenetics and Evolution 49 102ndash110

Katoh K Toh H 2008 Recent developments in the MAFFTmultiple sequence alignment program Briefings inBioinformatics 9 286ndash298

Krijgsman W Hilgen F Raffi I Sierro F Wilson D 1999Chronology causes and progression of the Messinian salin-ity crisis Nature 400 652ndash655

Lanfear R Calcott B Ho SY Guindon S 2012PartitionFinder combined selection of partitioning schemesand substitution models for phylogenetic analyses Molecu-lar Biology and Evolution 29 1695ndash1701

Le Houeacuterou HN 1997 Climate flora and fauna changes inthe Sahara over the past 500 million years Journal of AridEnvironments 37 619ndash647

Maca-Meyer N Carranza S Rando J Arnold E CabreraV 2003 Status and relationships of the extinct giant CanaryIsland lizard Gallotia goliath (Reptilia Lacertidae)assessed using ancient mtDNA from its mummifiedremains Biological Journal of the Linnean Society 80 659ndash670

McCallum J Robertson A 1990 Pulsed uplift of the TroodosMassif ndash evidence from the Plio-Pleistocene Mesaoria basinIn J Malpas EMM Panayiotou A Xenophontos C edsOphiolites oceanic crustal analogues Proceedings of theSymposium lsquoTroodos 1987rsquo Nicosia (The Geological SurveyDepartment Ministry of Agriculture and NaturalResources) 217ndash229

Metallinou M Arnold EN Crochet P-A Geniez P BritoJC Lymberakis P El Din SB Sindaco R Robinson MCarranza S 2012 Conquering the Sahara and Arabiandeserts systematics and biogeography of Stenodactylus geckos(Reptilia Gekkonidae) BMC Evolutionary Biology 12258

Monaghan MT Wild R Elliot M Fujisawa T Balke MInward DJ Lees DC Ranaivosolo R Eggleton P

18 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Together A b asper and A schreiberi form amonophyletic group within Acanthodactylus (Fig 2)Within the group however both taxa are paraphyleticwith A schreiberi as a whole nested within A b asperOur analyses distinguish three major clades (1) cladeA formed by A b asper from Syria (2) clade B in-cludes the two subspecies A sc ataturi from Turkeytogether with A sc schreiberi from Cyprus (3) cladeC which includes specimens of A b asper from the re-maining localities in its distribution range together withA sc syriacus from Israel and Lebanon Clade A is verywell supported and includes specimens of A b asperfrom central and northern Syria (Fig 1) splitting fromother specimens at the basal node of the group is es-timated to have occurred c 654 Mya [95 highest pos-terior density (HPD) 392ndash952 Mya] The level of geneticdifferentiation(p-distance) between these specimens and the remain-ing A b asper and all A schreiberi specimens is 37ndash46 for 12S and 107ndash119 for Cytb Clade B is alsovery well supported and includes two of the threenominal subspecies of A schreiberi A s schreiberi thenominotypical subspecies endemic to Cyprus and A sataturi from Turkey The Turkish subspecies is nestedwithin the Cypriot specimens and the two forms havelow genetic distances from each other (12S 016 Cytb123) This clade is nested between the two A b asperclades (clades A and C) in both the concatenatedand the nuclear tree although the nodes are not wellsupported Clade C is not very well supported It in-cludes a cluster of A b asper and A s syriacus Thisclade includes two inner clades that split around558 Mya (95 HPD 356ndash8 Mya) and divided into threepoorly supported geographical inner groups (Fig 2)northern Jordan and northern Oman (group C1) NorthAfrica (group C2) and samples from the Middle East(Egypt south Israel and south Jordan) with samplesfrom Yemen and southern Oman (group C3) ndash the latterincluding all specimens of the subspecies A s syriacusThe diversification within the North African group isestimated to have started around 456 Mya (95 HPD282ndash647 Mya) The IsraelminusLebanon endemic subspe-cies A s syriacus is genetically highly distinct from

A s schreiberi and A s ataturi making A schreiberiparaphyletic (p-distance 12S 431 416 Cytb 1181202 respectively)

The networks constructed for the phased haplotypesof the full length nuclear markers (MC1R ACM4 andc-mos) are presented in Figure 3 The nuclear networkanalyses show similar results for each of the three genesand closely agree with the phylogenetic tree The CypriotA s schreiberi and Turkish A s ataturi subspecies sharealleles for all three genes and both are distinct fromthe third subspecies A s syriacus Acanthodactylusschreiberi syriacus shares no alleles with the other sub-species of A schreiberi but does share alleles withA b asper for each of the genes Acanthodactylusschreiberi syriacus shares MC1R alleles with A b asperspecimens from Tunisia Syria and Israel ACM4 alleleswith Egyptian Israeli Jordanian and North Africanspecimens and c-mos alleles only with Israeli A b asperspecimens Syrian A b asper samples share one allelewith A s syriacus and two with other A b asper speci-mens from Egypt Israel and North Africa in the MC1Rone allele with an Egyptian A b asper in the ACM4and none in the c-mos gene

In order to better understand the relationships betweenA schreiberi and A boskianus we performed three to-pology tests in which we forced monophyletic group-ings (1) monophyly of A schreiberi (all three subspeciestogether) (2) monophyly of A b asper (3) monophylyof A b asper and of A schreiberi The results of thetopological tests indicate that our data set cannot rejectthe alternative hypotheses of monophyly of A schreiberi(AU P = 0091 SH P = 0062) and that of A b asper(AU P = 011 SH P = 0072) if we allow A schreiberito nest within A b asper or a monophyletic A b aspernesting within A schreiberi When forcing monophylyof both A schreiberi and of A b asper together in thesame tree the results are inconclusive (AU P = 0046SH P = 0051)

The single-threshold model in GMYC yield a topol-ogy that is clearly different from the known taxono-my The GMYC results present a total of 25 and 24ML independent lineages from the Bayesian haplotypemitochondrial phylogeny of the two species for the two

Figure 2 Maximum likelihood (ML) tree of the Acanthodactylus boskianus and Acanthodactylus schreiberi specimensinferred using 12S rRNA cytochrome b mtDNA and melano-cortin 1 receptor acetylcholinergic receptor M4 and oocytematuration factor MOS nuclear gene fragments Posterior probability in the Bayesian analysis is indicated by black dotson the nodes [values ge 095 shown for both gene partitions and partitions by PartitionFinder (PF)] and ML bootstrapsupport values are indicated in parentheses (values ge 70 shown ML ML-PF) Age estimates with BEAST are indicat-ed near the relevant nodes and include the mean and in brackets the HPD 95 confidence interval Sample codes relateto specimens in Table 1 and in Figures 1 and 3 Colours blue Acanthodactylus boskianus asper yellow Acanthodactylusschreiberi ataturi red Acanthodactylus schreiberi syriacus green Acanthodactylus schreiberi schreiberi (Colour versionof figure available online)

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 11

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

12 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

partition approaches (ie by gene and by PartitionFinderFigs S2 S3 respectively) The two partition ap-proaches gave similar clusters but at the less sup-ported nodes they differed at the positions of severallineages The single threshold GMYC result is indi-cated for a single line at 00037 Mya for the gene par-titions and at 002 Mya for PartitionFinder (verticallines in Figs S2 S3) The topology and clusters re-vealed in this analysis correspond to the lineages fromthe phylogeny of the ML and BI methods both for theparaphyly of the two species and the geographical group-ings within A b asper The GMYC results mainly differfrom the ML and BI methods in the position ofA schreiberi from Cyprus and Turkey as a sister cladeto the Syrian A b asper

DISCUSSION

We have provided a comprehensive and thorough as-sessment of the intraspecific phylogenetic relation-ships within A schreiberi and its closest relativeA b asper Our results based on mitochondrial andnuclear DNA data from 84 specimens across the entiredistribution range of A schreiberi and most of the dis-tribution range of A b asper reveal that A schreiberiis paraphyletic and nested entirely within theA boskianus subspecies

HISTORICAL BIOGEOGRAPHY

Acanthodactylus schreiberi is thought to comprisethree subspecies corresponding to three allopatricpopulations in Cyprus Turkey and IsraelminusLebanonThe Cypriot endemic nominotypical subspeciesA sc schreiberi and the Turkish subspeciesA sc ataturi cluster together (to form clade B Fig 2)nesting between A b asper clades This lineage is sisterto a clade of A b asper including A sc syriacus (cladeC Fig 2) We estimate the divergence time of theCypriotminusTurkish lineage of A schreiberi to have beenduring the late Miocene around 6 Mya although thereis no support for this split in the tree In other analy-ses using the whole genus this split is well support-ed in Bayesian analyses (K Tamar S Carranza RSindaco J Moravec JF Trape amp S Meriri unpubldata) Based on mitochondrial data Poulakakis et al(2013) found that both the Cypriot and Turkish sub-species are monophyletic and diverged from each other085 Mya (038ndash156 Mya) According to our results thisdate corresponds to an inner divergence of the

A sc schreiberi lineage rather than to the date at whichA sc schreiberi colonized Cyprus

The discrepancy in the phylogenetic relationship ofA sc schreiberi raises questions regarding the arrivalon Cyprus Cyprus originated with the raising of theTroodos Massif during the upper Cretaceous c 91 to88 Mya (Clube amp Robertson 1986 Mukasa amp Ludden1987) During the middle to late Miocene only a smallproportion of Cyprus was exposed above the Mediter-ranean (McCallum amp Robertson 1990 Robertson 1990)Towards the end of the Miocene sim596 Mya with theclosing of the passage between the Atlantic Ocean andthe Mediterranean basin the Messinian salinity crisisbegan (Krijgsman et al 1999) This resulted in thedrying up of much of the Mediterranean Sea and highsea-mounts emerged to form land bridges with thesurrounding land (Hsuuml et al 1977) By the end of theMiocene and early Pliocene sim533 Mya the passagewith the Atlantic Ocean reopened and the Mediterra-nean basin was refilled (Krijgsman et al 1999) Re-sulting from compressions raising and uplifting of thesurrounding areas towards and during the Pleisto-cene Cyprus was a complete emergent island (McCallumamp Robertson 1990) The possible connection of Cyprusto the mainland (ie to TurkeySyria) during theMessinian is debated as are suggestions of a land con-nection at later periods (Steininger amp Roumlgl 1984 Jolivetet al 2006 Bache et al 2012) Such a connection ifit existed could have provided access for terrestrialorganisms with poor overseas dispersal ability suchas lizards to colonize the island Several studies arguethat post-Messinian sea level changes are unlikely tohave formed connections between Cyprus and the main-land (Steininger amp Roumlgl 1984 Jolivet et al 2006) Thusour dating of the split between the Cypriot A schreiberiand A b asper at c 6 Mya leads us to suggest that theancestor of A s schreiberi colonized Cyprus from themainland through a land bridge connection at the be-ginning of the Messinian crisis rather than by a muchlatermore recent transmarine dispersal as suggestedby Poulakakis et al (2013) Owing to its close rela-tions with A b asper the ancestor of A schreiberi waspresumably mainland A boskianus and the cladogenesisleading to A schreiberi thus rendered A b asperparaphyletic

The Turkish subspecies A schreiberi ataturi was rec-orded for the first time by Franzen (1998) at a veryrestricted area of around 15 km of coastal strip (betweenBotas and Yukarı Burnaz Hatay Province) Owing tothe remarkable morphological similarity between

Figure 3 Haplotype networks of the nuclear gene fragments melano-cortin 1 receptor (MC1R) acetylcholinergic recep-tor M4 (ACM4) and oocyte maturation factor MOS (c-mos) with colours corresponding to Figures 1 and 2 Codes corre-late to the two alleles (ie a and b) of specimens in Table 1 Circle sizes are proportional to the number of alleles (Colourversion of figure available online)

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 13

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

A sc ataturi and the Cypriot population the speci-mens were initially identified as A sc schreiberi(Franzen 1998) Yalccedilinkaya amp Goumlccedilmen (2012) howeverdescribed this population as a new distinct subspe-cies A s ataturi presenting several differences betweenthe two in both morphology and blood-serum pro-teins The origin of A s ataturi remains uncertain asit is debated whether the newly discovered Turkishpopulation is a relict or an introduction from CyprusFranzen (1998) described this population as a pos-sible introduction from Cyprus through the harbourof Botas but Sindaco et al (2000) suggested thatit might be a relict of a previously larger popula-tion because its present distribution is similar tothat of some insects and lizards [Archaeolacerta(Phoenicolacerta) laevis and Ablepharus budaki]Yalccedilinkaya amp Goumlccedilmen (2012) proposed that A sc ataturiarrived in Turkey from the nominate population inCyprus during the Messinian crisis The phylogeneticresults haplotype networks and low levels of geneticdivergence we found suggest that the two subspeciesfrom Cyprus and Turkey have not been geneticallyisolated for a long period of time (ie they share allelesin all three nuclear genes and A s ataturi is nestedwithin A s schreiberi in the phylogeny Figs 2 3) Ourresults therefore contrast with the two latter sce-narios of a relict population or a Messinian disper-sal Both divergence time and the genetic similarityof the two subspecies agree with the original sugges-tion of Franzen (1998) that these animals were intro-duced into Turkey from Cyprus Further support forthis hypothesis is that A s ataturi is restricted to thevicinity of the Botas-Adana harbour and is absent inother suitable habitats (coastal sand dunes) wide-spread in south-eastern Turkey Its close morphologi-cal features to A s schreiberi (Franzen 1998) likewisesupport an introduction scenario

The third subspecies A s syriacus is nested withinA b asper in the concatenated mtDNA and nDNA treesand is clearly genetically distinct from the Cyprus andTurkey A schreiberi lineage The close relations ofA s syriacus with A boskianus may shed light on theorigin of the former Acanthodactylus schreiberi syriacusis distributed on stable sands of the coastal plain ofthe eastern Mediterranean in Israel and southernLebanon (Salvador 1982 Hraoui-Bloquet et al 2002Bar amp Haimovitch 2011) habitats resembling those ofA s schreiberi from Cyprus (Baier Sparrow amp Wiedl2009) The oldest divergence of the A b asper cladethat includes A s syriacus is estimated to have oc-curred during the late Miocene around 558 Mya butno further dates are available for the grouping ofA s syriacus as a result of low support values Thecoastal plain of the eastern Mediterranean was sub-merged during the late Miocene and re-emerged onlytoward the Pliocene (Nir 1970 Horowitz 1979) The

sands of the coastal plain where A s syriacus occurs(Salvador 1982 Arnold 1983 K Tamar amp S Meiripers observ) were repeatedly submerged and re-emerged during the Pleistocene sea-level changes (duringinterglacial and glacial periods respectively) A pos-sible scenario for A sc syriacusrsquos origin includes severalwaves of dispersal of Middle Eastern A boskianus whichoccurs on coarse substrates (Amitai amp Bouskila 2001Disi et al 2001 Baha El Din 2006 pers observ) towardthe Mediterranean shore Acanthodactylus boskianusasper is absent from Mediterranean climate habitatsin Lebanon and Israel It occupies only xeric zonessuggesting an invasion to the coastal plain when sandyhabitats allowed desert flora and fauna to migrate north-wards (Yom-Tov 1988) These populations adapted tosandy soils and evolved morphological features thatdistinguish them from the desert hard substrate formsof A b asper We view this as the most likely scenariogiven the biogeography the phylogenetic results andthe habitat preferences and adaptations of these lizardsAn alternative scenario according to which the an-cestor of A schreiberi originated in Cyprus and dis-persed to the shores of Israel and Lebanon (or originatedin the coastal plain of the Eastern Mediterranean anddispersed to Cyprus) we regard as far less likely Sucha scenario requires much closer genetic relationshipsbetween these two forms and is further weakened bythe close relationship between A s syriacus and thegeographically adjacent A b asper populations

Acanthodactylus boskianus asper is highly variableboth morphologically (Salvador 1982 Arnold 1983) andgenetically (this study) The subspecies is paraphyleticas A schreiberi is nested within it The topology of theA b asper tree shows four different geographical group-ings Syria (clade A) north Jordan plus north OmanNorth Africa and Middle East plus south Arabia (groupsC1 C2 C3 respectively) The different groups in thissubspecies are estimated to have first diverged duringthe late Miocene approximately 65 Mya with the splitof the Syrian population The Syrian lineage is ge-netically distant from A schreiberi and the otherA b asper specimens The nuclear networks indicatethat this group is closer to the other A b asper samplesrather than to A schreiberi The geographical splitsin the rest of the A b asper range (clade C) are esti-mated to have started around 558 Mya These groupsare supported as a distinct clade but are closely relatedto each other in both the concatenated and nuclear trees(Figs 2 S1 respectively) The diversification within thisclade is estimated to have occurred during the lateMiocene to early Pliocene when A b asper dispersedwidely west to North Africa and in Arabia The di-vergence within the North African group (group C2)is estimated to have occurred during the Pliocene ap-proximately 456 Mya with the Egyptian Nigerian andSudanese populations later dispersing west and north

14 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

in Africa This diversification correlates to the arid climatestarting in southern Sahara during the early-mid Plio-cene and later in northern Africa between the Plio-cene and the Pleistocene (Le Houeacuterou 1997) as hasbeen suggested for the dispersal of Mesalina guttulatain Africa (Kapli et al 2008) Other evidence relatesdry climate in North Africa to an earlier period around7 Mya (Schuster et al 2006) as has been suggestedfor the genus Chalcides and other reptiles (Carranzaet al 2008 Metallinou et al 2012 and reference therein)The aridification of North Africa has most likely con-tributed for the successful dispersal of A b asper westfrom south-west Asia into Africa Morphological studiesof A boskianus show relatively uniform populations inNorth Africa suggesting recent migration (Salvador1982 Arnold 1983) The other two geographical group-ings of A b asper from the Middle East and Arabia(groups C1 and C3) are located in two distinct innerclades but their location within each inner clade ispoorly supported The topology of the concatenated tree(Fig 2) shows that the group from northern Jordanand northern Oman (group C1) is closer to the NorthAfrican one than to the geographically close Middle-Eastern and south Arabian group (group C3) The taxo-nomic separation between north and south Oman hasbeen recognized in other species of reptiles and sup-ported by the topography of Oman (eg Echis coloratusand Echis omanensis Arnold Robinson amp Carranza2009) In the nuclear tree (Fig S1) these two groupsare closer to one another and with the North Africangroup form clade C Therefore the low support valuesamongst these groupings prevent an appropriate andthorough analysis of this subspecies The close rela-tionship amongst the geographical groups may reflectclose phylogenetic relationships amongst these popu-lations suggesting recent migration divergence andongoing gene flow

SYSTEMATICS AND TAXONOMIC IMPLICATIONS

The relationships within the A boskianus species groupconflict with the current known taxonomy of A schreiberiand A b asper (samples of the other subspecies ofA boskianus and of A nilsoni were unavailable for thisstudy) Both species have been found to be closelyrelated and paraphyletic The constrained topology testsexemplify the close entangled relationship between thetwo species as the separate monophyly of the two specieswas not rejected and the enforced monophyly of themboth together was inconclusive

Several causes can be responsible for paraphyly inspecies such in the A boskianus species group (Funkamp Omland 2003 and references therein) (1) inad-equate phylogenetic information (2) imperfect taxono-my (incorrectinaccurate species limits) derived frommisidentifying intraspecific variation (3) interspecific

gene flow ndash hybridization through interspecific matingand the subsequent backcrossing of hybrids into theparental populations (4) incomplete lineage sortingbecause of recent speciation events (5) unrecognizedparalogy We suggest that the relationships betweenA schreiberi and A b asper based on mitochondrialand nuclear data are most likely explained by incor-rect taxonomy probably because of the great variabil-ity of the latter species and to convergence As wasthe case in the molecular studies of the A pardalisand A erythrurus species groups (Harris et al 2004Fonseca et al 2008 2009 Carretero et al 2011 andreference therein) there are many problems with thecurrent taxonomic status of several species groups withinAcanthodactylus

Taking the molecular results of our study into accountthere are several systematic approaches to classify-ing the A schreiberiminusA b asper clade The Cypriot andTurkish populations of A schreiberi are very closelyrelated with the latter nested in the former and thetwo subspecies share nuclear alleles (Fig 3) Further-more the low uncorrected p-distance is positively cor-related with subspecies-level distances within otherlacertid species (ie 16 of Cytb in Lacerta bilineatachloronota Godinho et al 2005) We therefore con-clude that Cypriot animals were recently introducedto Turkey and that the Turkish population does notmerit a subspecific rank We suggest that A s ataturiYalccedilinkaya amp Goumlccedilmen 2012 is a junior synonym ofA s schreiberi Boulenger 1878

Regarding the relationships between A schreiberi andA b asper a few scenarios are possible One is to sinkA schreiberi within A boskianus to create one species(A boskianus) with high genetic and morphological vari-ability ranging over a broad distribution Another isfor the two taxa be regarded as a species complex (theA boskianus-schreiberi complex) until further inves-tigation on the subject However although A schreiberiis nested within A b asper the populations from Cyprusand Turkey represent a distinct evolutionary lineagewith distinct genetic and morphological features andthus it is logical to retain the specific status Two othersolutions are possible The first is to re-evaluate theSyrian populations and to consider elevating them aswell as the more divergent lineages (and subspecies)of A boskianus to specific status This would neces-sitate an examination of the phylogeny and morphol-ogy of the other four subspecies of A boskianus(A b boskianus A b euphraticus A b khattensis andA b nigeriensis) and the identification of distinctivephenotypic features in the Syrian lizards Another so-lution is to recognize the maintenance of gene flowamongst mainland populations of A b asper after thedivergence of the insular endemic A schreiberi andthus the evolutionary cohesion of the paraphyleticA b asper Arnold (1983) noted that A schreiberi may

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 15

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

have originated as an isolate from A boskianus becauseof their shared morphology and hemipenis featuresOur results support this scenario which includes thedispersal of A schreiberi to Cyprus from a mainlandpopulation that was most probably A boskianus It maybe assumed that the ancestor of the Cypriot A schreiberiafter arriving on Cyprus remained isolated for a longperiod of time and thus evolved to the modern formof A schreiberi Meanwhile the same ancestral con-tinental populations not isolated from each other con-tinued to exchange genes to varying degrees remainingA boskianus

The IsraeliminusLebanese subspecies A sc syriacus is onlydistantly related to the nominate form A sc schreiberiThis subspecies is highly phylogenetically divergent fromthe Cypriot and Turkish populations having higherp-distances (12S 4 Cytb 11ndash12) than those foundbetween other lacertid species (eg 74ndash82 of Cytbamongst Iberolacerta aranica Iberolacerta aurelioi andIberolacerta bonnali and 41ndash58 of Cytb betweenLacerta bilineata and Lacerta viridis Crochet et al2004 Godinho et al 2005 respectively) The nuclearhaplotype networks further show that Lebanese andIsraeli populations share alleles only with A b asperbut not with the nominotypical Cypriot form Arnold(1983) suggested that the geographical variation ofA boskianus reflects niche differences with animalsfrom xeric areas with dense rigid and spiny vegeta-tion having larger dorsal scales than animals from moremesic areas As was assumed for A schreiberi wesuggest that other mainland populations of A b asperwere the ancestors of the LebaneseminusIsraeli Coastal plainforms We suggest that A s syriacus is an ecomorphof A b asper that dispersed from the usual xerichabitats of the species and adapted to the new moremesic environment of the stable sands of the coastalplains of the eastern Mediterranean As a conse-quence this ecomorph converged on the morphologyof A s schreiberi which inhabits the coastal sands ofCyprus (Baier et al 2009) but still maintains differ-entiating features by having coarser dorsal scales andsharp keels (Salvador 1982 Arnold 1983 Franzen1998) This convergence led to the description ofA s syriacus as a member of A schreiberi The mor-phological assessment and the close morphological simi-larities between A b asper and A s syriacus mayexplain the wrong classification A similar erro-neous reasoning led Reed amp Marx (1959) to identifyspecimens with fine scales from Iraq as A schreiberiSalvador (1982) re-examined these specimens and as-signed them to A boskianus The morphological dif-ferences between the two forms are less prominentespecially where the two forms occur in close geo-graphical proximity in the southern coastal plainand north-western Negev Desert of Israel (Bar ampHaimovitch 2011) According to our results A s syriacus

actually belongs to A b asper being a coastal-duneecomorph convergent with but evolutionarily dis-tinct from A schreiberi Thus our preferred scenariois to treat the name Acanthodactylus schreiberi syriacusBoumlttger 1879 (which was originally described asA boskianus var syriacus by Boumlttger 1879) as a juniorsynonym of the name Acanthodactylus boskianus asper(Audouin 1827)

Recognizing A s syriacus as a junior synonym ofA b asper may have important implications for the con-servation of this coastal sand dune form which is clas-sified as critically endangered in Israel (Dolev ampPervolutzki 2004) However as the Israeli and Leba-nese coastal dune ecosystem has probably developedonly very recently during the Quaternary (Nir 1970Horowitz 1979) this form represents a remarkable caseof rapid evolutionary change It is also a remarkablecase of convergent evolution (with the CypriotA sc schreiberi) Thus we feel that these popula-tions are unique evolutionary entities that merit specialconservation efforts

The use of nuclear genes is a valuable method forestimating species divergence and lineage sorting andhelps evaluate isolated lineages and evolutionary historyThe incorporation of mitochondrial and nuclear dataprovides thorough topologies informative networks anddivergence times that reveal useful information for aproblematic taxonomy such as that of the A boskianusspecies group We have shown that phylogenetic ap-proaches to the confusing taxonomy of two closely relatedand morphologically similar species can shed light ontheir unclear relationships resolve between homoplasyand shared ancestry and identify patterns of speciesevolutionary history and biogeography

ACKNOWLEDGEMENTS

We are grateful to the following people for providingsamples for this study D Donaire B Shacham J ŠmiacutedS M Baha El Din and J Padial We thank MMetallinou and J Šmiacuted for help with the lab work andthe analyses M Novosolov for help with the figuresY L Werner B Shacham and A Levy for fruitful dis-cussions and two anonymous referees for helpful com-ments on an earlier version of this manuscript Wethank the Israel nature and park authority for issuingcollecting permits (nos 38074 38451 38489 38540)The work of J M was financially supported by the Min-istry of Culture of the Czech Republic (DKRVO 201315 National Museum 00023272) S C is funded bygrant CGL2012-36970 from the Ministerio de Economiacuteay Competitividad Spain (cofunded by FEDER) Thisstudy was funded by a Israel Taxonomic Initiative grantto S M K T is supported by an Israel Taxonomic Ini-tiative scholarship

16 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

REFERENCES

Ahmadzadeh F Carretero MA Harris DJ Perera ABohme W 2012 A molecular phylogeny of the eastern groupof ocellated lizard genus Timon (Sauria Lacertidae) basedon mitochondrial and nuclear DNA sequences Amphibia-Reptilia 33 1ndash10

Ahmadzadeh F Flecks M Roumldder D Boumlhme W Ilgaz CcedilHarris DJ Engler JO Uumlzuumlm N Carretero MA 2013Multiple dispersal out of Anatolia biogeography and evolu-tion of oriental green lizards Biological Journal of the LinneanSociety 110 398ndash408

Akaike H 1973 Information theory and an extension of themaximum likelihood principle In Petrov BN Csaki F edsSecond International Symposium on Information Theory Bu-dapest Akademiai Kiado 267ndash281

Amitai P Bouskila A 2001 Handbook of amphibians ampreptiles of Israel Israel Keter Publishing House Ltd

Anderson SC 1999 The lizards of Iran Ithaca NY Societyfor the Study of Amphibians and Reptiles

Arnold EN 1980 The scientific results of the Oman flora andfauna survey 1977 (Dhofar) The reptiles and amphibians ofDhofar southern Arabia Journal of Oman Studies SpecialReport 2 273ndash332

Arnold EN 1983 Osteology genitalia and the relationshipsof Acanthodactylus (Reptilia Lacertidae) Bulletin of the BritishMuseum (Natural History) Zoology 44 291ndash339

Arnold EN Arribas Oacute Carranza S 2007 Systematics ofthe Palaearctic and Oriental lizard tribe Lacertini (SquamataLacertidae Lacertinae) with descriptions of eight new generaZootaxa 1430 1ndash86

Arnold EN Robinson MD Carranza S 2009 A prelimi-nary analysis of phylogenetic relationships and biogeogra-phy of the dangerously venomous Carpet Vipers Echis(Squamata Serpentes Viperidae) based on mitochondrial DNAsequences Amphibia-Reptilia 30 273ndash282

Audouin JV 1827 Explication sommaire des planches dereptiles (suppleacutement) offrant un exposeacute des characteacuteresdes espegraveces In Savigny MJCL ed Description de lrsquoEacutegypteVol 1 Histoire Naturelle Paris Impeacuteriale 161ndash184

Bache F Popescu SM Rabineau M Gorini C Suc JPClauzon G Olivet JL Rubino JL Melinte-DobrinescuMC Estrada F 2012 A two-step process for the refloodingof the Mediterranean after the Messinian Salinity Crisis BasinResearch 24 125ndash153

Baha El Din SM 2006 A guide to the reptiles and amphib-ians of Egypt CairondashNew York American University in CairoPress

Baier FS Sparrow DJ Wiedl H-J 2009 The amphibiansand reptiles of Cyprus Frankfurt am Main Edition Chimaira

Bandelt H-J Forster P Roumlhl A 1999 Median-joining net-works for inferring intraspecific phylogenies Molecular Biologyand Evolution 16 37ndash48

Bar A Haimovitch G 2011 A field guide to reptiles and am-phibians of Israel Herzliya Pazbar Limited

Boumlttger O 1879 Reptilien und Amphibien aus Syrien Berichtuumlber die Senckenbergische Naturforschende Gesellschaft inFrankfurt am Main 1879 57ndash84

Boulenger GA 1919 On a new variety of Acanthodactylusboskianus Daud from the Euphrates The Annals and Maga-zine of Natural History 3 549ndash550

Boulenger GA 1878 Sur les espegraveces drsquoAcanthodactylus desbords de la Mediterraneacutee Bulletin de la Socieacuteteacute zoologiquede France 3 179ndash197

Boulenger GA 1918 Sur les leacutezards du genre AcanthodactylusWieg Bulletin de la Socieacuteteacute zoologique de France 43 143ndash155

Boulenger GA 1921 Monograph of the Lacertidœ Vol II Trus-tees of the British Museum of Natural History London

Carranza S Arnold EN 2012 A review of the geckos of thegenus Hemidactylus (Squamata Gekkonidae) fromOman based on morphology mitochondrial and nuclear datawith descriptions of eight new species Zootaxa 3378 1ndash95

Carranza S Arnold EN Geniez P Roca J Mateo J 2008Radiation multiple dispersal and parallelism in the skinksChalcides and Sphenops (Squamata Scincidae) with com-ments on Scincus and Scincopus and the age of the SaharaDesert Molecular Phylogenetics and Evolution 46 1071ndash1094

Carretero MA Fonseca MM Garcia-Munoz E Brito JCHarris DJ 2011 Adding Acanthodactylus beershebensis tothe mtDNA phylogeny of the Acanthodactylus pardalis groupNorth-Western Journal of Zoology 7 138ndash142

Castresana J 2000 Selection of conserved blocks from multi-ple alignments for their use in phylogenetic analysis Mo-lecular Biology and Evolution 17 540ndash552

Clube TMM Robertson A 1986 The palaeorotation of theTroodos microplate Cyprus in the Late Mesozoic-Early Ce-nozoic plate tectonic framework of the Eastern Mediterra-nean Surveys in Geophysics 8 375ndash437

Cox SC Carranza S Brown RP 2010 Divergence timesand colonization of the Canary Islands by Gallotia lizardsMolecular Phylogenetics and Evolution 56 747ndash757

Crochet PA Geniez P Ineich I 2003 A multivariate analy-sis of the fringe-toed lizards of the Acanthodactylus scutellatusgroup (Squamata Lacertidae) systematic and biogeographi-cal implications Zoological Journal of the Linnean Society137 117ndash155

Crochet P-A Chaline O Surget-Groba Y Debain CCheylan M 2004 Speciation in mountains phylogeographyand phylogeny of the rock lizards genus Iberolacerta (ReptiliaLacertidae) Molecular Phylogenetics and Evolution 30 860ndash866

Daudin FM 1802 Histoire naturelle geacuteneacuterale et particuliegraveredes reptiles ouvrage faisant suite agrave lrsquoHistoire naturelle geacuteneacuteraleet particuliegravere F Dufart

Disi A Modry D Necas P Rifai L 2001 Amphibians andreptiles of the Hashemite Kingdom of Jordan an atlas andfield guide Frankfurt am Main Chimaira

Dolev A Pervolutzki A 2004 Endangered species in IsraelRed list of threatened animals Vertebrates JerusalemThe Nature and Parks Authority and the Society for thePreservation of Nature

Drummond AJ Rambaut A 2007 BEAST Bayesian evo-lutionary analysis by sampling trees BMC EvolutionaryBiology 7 214

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 17

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Ezard T Fujisawa T Barraclough T 2009 Splits speciesrsquolimits by threshold statistics R package version 10-19r48Available at httpR-ForgeR-projectorgprojectssplits

Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783ndash791

Fitzinger LJFJ 1834 Acanthodactylus In Wiegman AFAed Herpetologia mexicana seu Descriptio amphibiorum NovaeHispaniae quae itineribus comitis De Sack Ferdinandi Deppeet Chr Guil Schiede in Museum zoologicum Berolinensepervenerunt Pars prima saurorum species amplectens adjectosystematis saurorum prodromo additisque multis in huncamphibiorum ordinem observationibus Berlin C G Luderitz10

Flot JF 2010 SeqPHASE a web tool for interconvertingPHASE inputoutput files and FASTA sequence align-ments Molecular Ecology Resources 10 162ndash166

Fonseca MM Brito JC Paulo OS Carretero MA HarrisDJ 2009 Systematic and phylogeographical assessment ofthe Acanthodactylus erythrurus group (Reptilia Lacertidae)based on phylogenetic analyses of mitochondrial and nuclearDNA Molecular Phylogenetics and Evolution 51 131ndash142

Fonseca MM Brito JC Rebelo H Kalboussi M LarbesS Carretero MA Harris DJ 2008 Genetic variation amongspiny-footed lizards in the Acanthodactylus pardalis groupfrom North Africa African Zoology 43 8ndash15

Fontaneto D Herniou EA Boschetti C Caprioli M MeloneG Ricci C Barraclough TG 2007 Independently evolv-ing species in asexual bdelloid rotifers PLoS Biology 5 e87

Franzen M 1998 Erstnachweis von Acanthodactylus schreiberischreiberi Boulenger 1879 fuumlr die Tuumlrkei (Squamata SauriaLacertidae) Herpetozoa 11 27ndash36

Funk DJ Omland KE 2003 Species-level paraphyly andpolyphyly frequency causes and consequences with in-sights from animal mitochondrial DNA Annual Review ofEcology Evolution and Systematics 34 397ndash423

Godinho R Crespo EG Ferrand N Harris DJ 2005 Phy-logeny and evolution of the green lizards Lacerta spp(Squamata Lacertidae) based on mitochondrial and nuclearDNA sequences Amphibia-Reptilia 26 271ndash285

Greenbaum E Villanueva CO Kusamba C Aristote MMBranch WR 2011 A molecular phylogeny of EquatorialAfrican Lacertidae with the description of a new genus andspecies from eastern Democratic Republic of the Congo Zoo-logical Journal of the Linnean Society 163 913ndash942

Guillou H Carracedo JC Torrado FP Badiola ER 1996K-Ar ages and magnetic stratigraphy of a hotspot-inducedfast grown oceanic island El Hierro Canary Islands Journalof Volcanology and Geothermal Research 73 141ndash155

Harris DJ Arnold EN 2000 Elucidation of the relation-ships of spiny-footed lizards Acanthodactylus spp (ReptiliaLacertidae) using mitochondrial DNA sequence with com-ments on their biogeography and evolution Journal of Zoology252 351ndash362

Harris DJ Batista V Carretero M 2004 Assessment ofgenetic diversity within Acanthodactylus erythrurus (ReptiliaLacertidae) in Morocco and the Iberian Peninsula usingmitochondrial DNA sequence data Amphibia-Reptilia 25227

Horowitz A 1979 The Quaternary of Israel New York Aca-demic Press

Hraoui-Bloquet S Sadek RA Sindaco R Venchi A 2002The herpetofauna of Lebanon new data on distributionZoology in the Middle East 27 35ndash46

Hsuuml KJ Montadert L Bernoulli D Cita MB EricksonA Garrison RE Kidd RB Megraveliereacutes F Muumlller C WrightR 1977 History of the Mediterranean salinity crisis Nature267 399ndash403

Huelsenbeck JP Rannala B 2004 Frequentist propertiesof Bayesian posterior probabilities of phylogenetic trees undersimple and complex substitution models Systematic Biology53 904ndash913

Huelsenbeck JP Ronquist F 2001 MRBAYES Bayesianinference of phylogenetic trees Bioinformatics 17 754ndash755

Jolivet L Augier R Robin C Suc J-P Rouchy JM 2006Lithospheric-scale geodynamic context of the Messinian sa-linity crisis Sedimentary Geology 188 9ndash33

Kapli P Lymberakis P Poulakakis N Mantziou GParmakelis A Mylonas M 2008 Molecular phylogeny ofthree Mesalina (Reptilia Lacertidae) species (M guttulataM brevirostris and M bahaeldini) from North Africa and theMiddle East another case of paraphyly MolecularPhylogenetics and Evolution 49 102ndash110

Katoh K Toh H 2008 Recent developments in the MAFFTmultiple sequence alignment program Briefings inBioinformatics 9 286ndash298

Krijgsman W Hilgen F Raffi I Sierro F Wilson D 1999Chronology causes and progression of the Messinian salin-ity crisis Nature 400 652ndash655

Lanfear R Calcott B Ho SY Guindon S 2012PartitionFinder combined selection of partitioning schemesand substitution models for phylogenetic analyses Molecu-lar Biology and Evolution 29 1695ndash1701

Le Houeacuterou HN 1997 Climate flora and fauna changes inthe Sahara over the past 500 million years Journal of AridEnvironments 37 619ndash647

Maca-Meyer N Carranza S Rando J Arnold E CabreraV 2003 Status and relationships of the extinct giant CanaryIsland lizard Gallotia goliath (Reptilia Lacertidae)assessed using ancient mtDNA from its mummifiedremains Biological Journal of the Linnean Society 80 659ndash670

McCallum J Robertson A 1990 Pulsed uplift of the TroodosMassif ndash evidence from the Plio-Pleistocene Mesaoria basinIn J Malpas EMM Panayiotou A Xenophontos C edsOphiolites oceanic crustal analogues Proceedings of theSymposium lsquoTroodos 1987rsquo Nicosia (The Geological SurveyDepartment Ministry of Agriculture and NaturalResources) 217ndash229

Metallinou M Arnold EN Crochet P-A Geniez P BritoJC Lymberakis P El Din SB Sindaco R Robinson MCarranza S 2012 Conquering the Sahara and Arabiandeserts systematics and biogeography of Stenodactylus geckos(Reptilia Gekkonidae) BMC Evolutionary Biology 12258

Monaghan MT Wild R Elliot M Fujisawa T Balke MInward DJ Lees DC Ranaivosolo R Eggleton P

18 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

12 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

partition approaches (ie by gene and by PartitionFinderFigs S2 S3 respectively) The two partition ap-proaches gave similar clusters but at the less sup-ported nodes they differed at the positions of severallineages The single threshold GMYC result is indi-cated for a single line at 00037 Mya for the gene par-titions and at 002 Mya for PartitionFinder (verticallines in Figs S2 S3) The topology and clusters re-vealed in this analysis correspond to the lineages fromthe phylogeny of the ML and BI methods both for theparaphyly of the two species and the geographical group-ings within A b asper The GMYC results mainly differfrom the ML and BI methods in the position ofA schreiberi from Cyprus and Turkey as a sister cladeto the Syrian A b asper

DISCUSSION

We have provided a comprehensive and thorough as-sessment of the intraspecific phylogenetic relation-ships within A schreiberi and its closest relativeA b asper Our results based on mitochondrial andnuclear DNA data from 84 specimens across the entiredistribution range of A schreiberi and most of the dis-tribution range of A b asper reveal that A schreiberiis paraphyletic and nested entirely within theA boskianus subspecies

HISTORICAL BIOGEOGRAPHY

Acanthodactylus schreiberi is thought to comprisethree subspecies corresponding to three allopatricpopulations in Cyprus Turkey and IsraelminusLebanonThe Cypriot endemic nominotypical subspeciesA sc schreiberi and the Turkish subspeciesA sc ataturi cluster together (to form clade B Fig 2)nesting between A b asper clades This lineage is sisterto a clade of A b asper including A sc syriacus (cladeC Fig 2) We estimate the divergence time of theCypriotminusTurkish lineage of A schreiberi to have beenduring the late Miocene around 6 Mya although thereis no support for this split in the tree In other analy-ses using the whole genus this split is well support-ed in Bayesian analyses (K Tamar S Carranza RSindaco J Moravec JF Trape amp S Meriri unpubldata) Based on mitochondrial data Poulakakis et al(2013) found that both the Cypriot and Turkish sub-species are monophyletic and diverged from each other085 Mya (038ndash156 Mya) According to our results thisdate corresponds to an inner divergence of the

A sc schreiberi lineage rather than to the date at whichA sc schreiberi colonized Cyprus

The discrepancy in the phylogenetic relationship ofA sc schreiberi raises questions regarding the arrivalon Cyprus Cyprus originated with the raising of theTroodos Massif during the upper Cretaceous c 91 to88 Mya (Clube amp Robertson 1986 Mukasa amp Ludden1987) During the middle to late Miocene only a smallproportion of Cyprus was exposed above the Mediter-ranean (McCallum amp Robertson 1990 Robertson 1990)Towards the end of the Miocene sim596 Mya with theclosing of the passage between the Atlantic Ocean andthe Mediterranean basin the Messinian salinity crisisbegan (Krijgsman et al 1999) This resulted in thedrying up of much of the Mediterranean Sea and highsea-mounts emerged to form land bridges with thesurrounding land (Hsuuml et al 1977) By the end of theMiocene and early Pliocene sim533 Mya the passagewith the Atlantic Ocean reopened and the Mediterra-nean basin was refilled (Krijgsman et al 1999) Re-sulting from compressions raising and uplifting of thesurrounding areas towards and during the Pleisto-cene Cyprus was a complete emergent island (McCallumamp Robertson 1990) The possible connection of Cyprusto the mainland (ie to TurkeySyria) during theMessinian is debated as are suggestions of a land con-nection at later periods (Steininger amp Roumlgl 1984 Jolivetet al 2006 Bache et al 2012) Such a connection ifit existed could have provided access for terrestrialorganisms with poor overseas dispersal ability suchas lizards to colonize the island Several studies arguethat post-Messinian sea level changes are unlikely tohave formed connections between Cyprus and the main-land (Steininger amp Roumlgl 1984 Jolivet et al 2006) Thusour dating of the split between the Cypriot A schreiberiand A b asper at c 6 Mya leads us to suggest that theancestor of A s schreiberi colonized Cyprus from themainland through a land bridge connection at the be-ginning of the Messinian crisis rather than by a muchlatermore recent transmarine dispersal as suggestedby Poulakakis et al (2013) Owing to its close rela-tions with A b asper the ancestor of A schreiberi waspresumably mainland A boskianus and the cladogenesisleading to A schreiberi thus rendered A b asperparaphyletic

The Turkish subspecies A schreiberi ataturi was rec-orded for the first time by Franzen (1998) at a veryrestricted area of around 15 km of coastal strip (betweenBotas and Yukarı Burnaz Hatay Province) Owing tothe remarkable morphological similarity between

Figure 3 Haplotype networks of the nuclear gene fragments melano-cortin 1 receptor (MC1R) acetylcholinergic recep-tor M4 (ACM4) and oocyte maturation factor MOS (c-mos) with colours corresponding to Figures 1 and 2 Codes corre-late to the two alleles (ie a and b) of specimens in Table 1 Circle sizes are proportional to the number of alleles (Colourversion of figure available online)

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 13

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

A sc ataturi and the Cypriot population the speci-mens were initially identified as A sc schreiberi(Franzen 1998) Yalccedilinkaya amp Goumlccedilmen (2012) howeverdescribed this population as a new distinct subspe-cies A s ataturi presenting several differences betweenthe two in both morphology and blood-serum pro-teins The origin of A s ataturi remains uncertain asit is debated whether the newly discovered Turkishpopulation is a relict or an introduction from CyprusFranzen (1998) described this population as a pos-sible introduction from Cyprus through the harbourof Botas but Sindaco et al (2000) suggested thatit might be a relict of a previously larger popula-tion because its present distribution is similar tothat of some insects and lizards [Archaeolacerta(Phoenicolacerta) laevis and Ablepharus budaki]Yalccedilinkaya amp Goumlccedilmen (2012) proposed that A sc ataturiarrived in Turkey from the nominate population inCyprus during the Messinian crisis The phylogeneticresults haplotype networks and low levels of geneticdivergence we found suggest that the two subspeciesfrom Cyprus and Turkey have not been geneticallyisolated for a long period of time (ie they share allelesin all three nuclear genes and A s ataturi is nestedwithin A s schreiberi in the phylogeny Figs 2 3) Ourresults therefore contrast with the two latter sce-narios of a relict population or a Messinian disper-sal Both divergence time and the genetic similarityof the two subspecies agree with the original sugges-tion of Franzen (1998) that these animals were intro-duced into Turkey from Cyprus Further support forthis hypothesis is that A s ataturi is restricted to thevicinity of the Botas-Adana harbour and is absent inother suitable habitats (coastal sand dunes) wide-spread in south-eastern Turkey Its close morphologi-cal features to A s schreiberi (Franzen 1998) likewisesupport an introduction scenario

The third subspecies A s syriacus is nested withinA b asper in the concatenated mtDNA and nDNA treesand is clearly genetically distinct from the Cyprus andTurkey A schreiberi lineage The close relations ofA s syriacus with A boskianus may shed light on theorigin of the former Acanthodactylus schreiberi syriacusis distributed on stable sands of the coastal plain ofthe eastern Mediterranean in Israel and southernLebanon (Salvador 1982 Hraoui-Bloquet et al 2002Bar amp Haimovitch 2011) habitats resembling those ofA s schreiberi from Cyprus (Baier Sparrow amp Wiedl2009) The oldest divergence of the A b asper cladethat includes A s syriacus is estimated to have oc-curred during the late Miocene around 558 Mya butno further dates are available for the grouping ofA s syriacus as a result of low support values Thecoastal plain of the eastern Mediterranean was sub-merged during the late Miocene and re-emerged onlytoward the Pliocene (Nir 1970 Horowitz 1979) The

sands of the coastal plain where A s syriacus occurs(Salvador 1982 Arnold 1983 K Tamar amp S Meiripers observ) were repeatedly submerged and re-emerged during the Pleistocene sea-level changes (duringinterglacial and glacial periods respectively) A pos-sible scenario for A sc syriacusrsquos origin includes severalwaves of dispersal of Middle Eastern A boskianus whichoccurs on coarse substrates (Amitai amp Bouskila 2001Disi et al 2001 Baha El Din 2006 pers observ) towardthe Mediterranean shore Acanthodactylus boskianusasper is absent from Mediterranean climate habitatsin Lebanon and Israel It occupies only xeric zonessuggesting an invasion to the coastal plain when sandyhabitats allowed desert flora and fauna to migrate north-wards (Yom-Tov 1988) These populations adapted tosandy soils and evolved morphological features thatdistinguish them from the desert hard substrate formsof A b asper We view this as the most likely scenariogiven the biogeography the phylogenetic results andthe habitat preferences and adaptations of these lizardsAn alternative scenario according to which the an-cestor of A schreiberi originated in Cyprus and dis-persed to the shores of Israel and Lebanon (or originatedin the coastal plain of the Eastern Mediterranean anddispersed to Cyprus) we regard as far less likely Sucha scenario requires much closer genetic relationshipsbetween these two forms and is further weakened bythe close relationship between A s syriacus and thegeographically adjacent A b asper populations

Acanthodactylus boskianus asper is highly variableboth morphologically (Salvador 1982 Arnold 1983) andgenetically (this study) The subspecies is paraphyleticas A schreiberi is nested within it The topology of theA b asper tree shows four different geographical group-ings Syria (clade A) north Jordan plus north OmanNorth Africa and Middle East plus south Arabia (groupsC1 C2 C3 respectively) The different groups in thissubspecies are estimated to have first diverged duringthe late Miocene approximately 65 Mya with the splitof the Syrian population The Syrian lineage is ge-netically distant from A schreiberi and the otherA b asper specimens The nuclear networks indicatethat this group is closer to the other A b asper samplesrather than to A schreiberi The geographical splitsin the rest of the A b asper range (clade C) are esti-mated to have started around 558 Mya These groupsare supported as a distinct clade but are closely relatedto each other in both the concatenated and nuclear trees(Figs 2 S1 respectively) The diversification within thisclade is estimated to have occurred during the lateMiocene to early Pliocene when A b asper dispersedwidely west to North Africa and in Arabia The di-vergence within the North African group (group C2)is estimated to have occurred during the Pliocene ap-proximately 456 Mya with the Egyptian Nigerian andSudanese populations later dispersing west and north

14 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

in Africa This diversification correlates to the arid climatestarting in southern Sahara during the early-mid Plio-cene and later in northern Africa between the Plio-cene and the Pleistocene (Le Houeacuterou 1997) as hasbeen suggested for the dispersal of Mesalina guttulatain Africa (Kapli et al 2008) Other evidence relatesdry climate in North Africa to an earlier period around7 Mya (Schuster et al 2006) as has been suggestedfor the genus Chalcides and other reptiles (Carranzaet al 2008 Metallinou et al 2012 and reference therein)The aridification of North Africa has most likely con-tributed for the successful dispersal of A b asper westfrom south-west Asia into Africa Morphological studiesof A boskianus show relatively uniform populations inNorth Africa suggesting recent migration (Salvador1982 Arnold 1983) The other two geographical group-ings of A b asper from the Middle East and Arabia(groups C1 and C3) are located in two distinct innerclades but their location within each inner clade ispoorly supported The topology of the concatenated tree(Fig 2) shows that the group from northern Jordanand northern Oman (group C1) is closer to the NorthAfrican one than to the geographically close Middle-Eastern and south Arabian group (group C3) The taxo-nomic separation between north and south Oman hasbeen recognized in other species of reptiles and sup-ported by the topography of Oman (eg Echis coloratusand Echis omanensis Arnold Robinson amp Carranza2009) In the nuclear tree (Fig S1) these two groupsare closer to one another and with the North Africangroup form clade C Therefore the low support valuesamongst these groupings prevent an appropriate andthorough analysis of this subspecies The close rela-tionship amongst the geographical groups may reflectclose phylogenetic relationships amongst these popu-lations suggesting recent migration divergence andongoing gene flow

SYSTEMATICS AND TAXONOMIC IMPLICATIONS

The relationships within the A boskianus species groupconflict with the current known taxonomy of A schreiberiand A b asper (samples of the other subspecies ofA boskianus and of A nilsoni were unavailable for thisstudy) Both species have been found to be closelyrelated and paraphyletic The constrained topology testsexemplify the close entangled relationship between thetwo species as the separate monophyly of the two specieswas not rejected and the enforced monophyly of themboth together was inconclusive

Several causes can be responsible for paraphyly inspecies such in the A boskianus species group (Funkamp Omland 2003 and references therein) (1) inad-equate phylogenetic information (2) imperfect taxono-my (incorrectinaccurate species limits) derived frommisidentifying intraspecific variation (3) interspecific

gene flow ndash hybridization through interspecific matingand the subsequent backcrossing of hybrids into theparental populations (4) incomplete lineage sortingbecause of recent speciation events (5) unrecognizedparalogy We suggest that the relationships betweenA schreiberi and A b asper based on mitochondrialand nuclear data are most likely explained by incor-rect taxonomy probably because of the great variabil-ity of the latter species and to convergence As wasthe case in the molecular studies of the A pardalisand A erythrurus species groups (Harris et al 2004Fonseca et al 2008 2009 Carretero et al 2011 andreference therein) there are many problems with thecurrent taxonomic status of several species groups withinAcanthodactylus

Taking the molecular results of our study into accountthere are several systematic approaches to classify-ing the A schreiberiminusA b asper clade The Cypriot andTurkish populations of A schreiberi are very closelyrelated with the latter nested in the former and thetwo subspecies share nuclear alleles (Fig 3) Further-more the low uncorrected p-distance is positively cor-related with subspecies-level distances within otherlacertid species (ie 16 of Cytb in Lacerta bilineatachloronota Godinho et al 2005) We therefore con-clude that Cypriot animals were recently introducedto Turkey and that the Turkish population does notmerit a subspecific rank We suggest that A s ataturiYalccedilinkaya amp Goumlccedilmen 2012 is a junior synonym ofA s schreiberi Boulenger 1878

Regarding the relationships between A schreiberi andA b asper a few scenarios are possible One is to sinkA schreiberi within A boskianus to create one species(A boskianus) with high genetic and morphological vari-ability ranging over a broad distribution Another isfor the two taxa be regarded as a species complex (theA boskianus-schreiberi complex) until further inves-tigation on the subject However although A schreiberiis nested within A b asper the populations from Cyprusand Turkey represent a distinct evolutionary lineagewith distinct genetic and morphological features andthus it is logical to retain the specific status Two othersolutions are possible The first is to re-evaluate theSyrian populations and to consider elevating them aswell as the more divergent lineages (and subspecies)of A boskianus to specific status This would neces-sitate an examination of the phylogeny and morphol-ogy of the other four subspecies of A boskianus(A b boskianus A b euphraticus A b khattensis andA b nigeriensis) and the identification of distinctivephenotypic features in the Syrian lizards Another so-lution is to recognize the maintenance of gene flowamongst mainland populations of A b asper after thedivergence of the insular endemic A schreiberi andthus the evolutionary cohesion of the paraphyleticA b asper Arnold (1983) noted that A schreiberi may

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 15

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

have originated as an isolate from A boskianus becauseof their shared morphology and hemipenis featuresOur results support this scenario which includes thedispersal of A schreiberi to Cyprus from a mainlandpopulation that was most probably A boskianus It maybe assumed that the ancestor of the Cypriot A schreiberiafter arriving on Cyprus remained isolated for a longperiod of time and thus evolved to the modern formof A schreiberi Meanwhile the same ancestral con-tinental populations not isolated from each other con-tinued to exchange genes to varying degrees remainingA boskianus

The IsraeliminusLebanese subspecies A sc syriacus is onlydistantly related to the nominate form A sc schreiberiThis subspecies is highly phylogenetically divergent fromthe Cypriot and Turkish populations having higherp-distances (12S 4 Cytb 11ndash12) than those foundbetween other lacertid species (eg 74ndash82 of Cytbamongst Iberolacerta aranica Iberolacerta aurelioi andIberolacerta bonnali and 41ndash58 of Cytb betweenLacerta bilineata and Lacerta viridis Crochet et al2004 Godinho et al 2005 respectively) The nuclearhaplotype networks further show that Lebanese andIsraeli populations share alleles only with A b asperbut not with the nominotypical Cypriot form Arnold(1983) suggested that the geographical variation ofA boskianus reflects niche differences with animalsfrom xeric areas with dense rigid and spiny vegeta-tion having larger dorsal scales than animals from moremesic areas As was assumed for A schreiberi wesuggest that other mainland populations of A b asperwere the ancestors of the LebaneseminusIsraeli Coastal plainforms We suggest that A s syriacus is an ecomorphof A b asper that dispersed from the usual xerichabitats of the species and adapted to the new moremesic environment of the stable sands of the coastalplains of the eastern Mediterranean As a conse-quence this ecomorph converged on the morphologyof A s schreiberi which inhabits the coastal sands ofCyprus (Baier et al 2009) but still maintains differ-entiating features by having coarser dorsal scales andsharp keels (Salvador 1982 Arnold 1983 Franzen1998) This convergence led to the description ofA s syriacus as a member of A schreiberi The mor-phological assessment and the close morphological simi-larities between A b asper and A s syriacus mayexplain the wrong classification A similar erro-neous reasoning led Reed amp Marx (1959) to identifyspecimens with fine scales from Iraq as A schreiberiSalvador (1982) re-examined these specimens and as-signed them to A boskianus The morphological dif-ferences between the two forms are less prominentespecially where the two forms occur in close geo-graphical proximity in the southern coastal plainand north-western Negev Desert of Israel (Bar ampHaimovitch 2011) According to our results A s syriacus

actually belongs to A b asper being a coastal-duneecomorph convergent with but evolutionarily dis-tinct from A schreiberi Thus our preferred scenariois to treat the name Acanthodactylus schreiberi syriacusBoumlttger 1879 (which was originally described asA boskianus var syriacus by Boumlttger 1879) as a juniorsynonym of the name Acanthodactylus boskianus asper(Audouin 1827)

Recognizing A s syriacus as a junior synonym ofA b asper may have important implications for the con-servation of this coastal sand dune form which is clas-sified as critically endangered in Israel (Dolev ampPervolutzki 2004) However as the Israeli and Leba-nese coastal dune ecosystem has probably developedonly very recently during the Quaternary (Nir 1970Horowitz 1979) this form represents a remarkable caseof rapid evolutionary change It is also a remarkablecase of convergent evolution (with the CypriotA sc schreiberi) Thus we feel that these popula-tions are unique evolutionary entities that merit specialconservation efforts

The use of nuclear genes is a valuable method forestimating species divergence and lineage sorting andhelps evaluate isolated lineages and evolutionary historyThe incorporation of mitochondrial and nuclear dataprovides thorough topologies informative networks anddivergence times that reveal useful information for aproblematic taxonomy such as that of the A boskianusspecies group We have shown that phylogenetic ap-proaches to the confusing taxonomy of two closely relatedand morphologically similar species can shed light ontheir unclear relationships resolve between homoplasyand shared ancestry and identify patterns of speciesevolutionary history and biogeography

ACKNOWLEDGEMENTS

We are grateful to the following people for providingsamples for this study D Donaire B Shacham J ŠmiacutedS M Baha El Din and J Padial We thank MMetallinou and J Šmiacuted for help with the lab work andthe analyses M Novosolov for help with the figuresY L Werner B Shacham and A Levy for fruitful dis-cussions and two anonymous referees for helpful com-ments on an earlier version of this manuscript Wethank the Israel nature and park authority for issuingcollecting permits (nos 38074 38451 38489 38540)The work of J M was financially supported by the Min-istry of Culture of the Czech Republic (DKRVO 201315 National Museum 00023272) S C is funded bygrant CGL2012-36970 from the Ministerio de Economiacuteay Competitividad Spain (cofunded by FEDER) Thisstudy was funded by a Israel Taxonomic Initiative grantto S M K T is supported by an Israel Taxonomic Ini-tiative scholarship

16 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

REFERENCES

Ahmadzadeh F Carretero MA Harris DJ Perera ABohme W 2012 A molecular phylogeny of the eastern groupof ocellated lizard genus Timon (Sauria Lacertidae) basedon mitochondrial and nuclear DNA sequences Amphibia-Reptilia 33 1ndash10

Ahmadzadeh F Flecks M Roumldder D Boumlhme W Ilgaz CcedilHarris DJ Engler JO Uumlzuumlm N Carretero MA 2013Multiple dispersal out of Anatolia biogeography and evolu-tion of oriental green lizards Biological Journal of the LinneanSociety 110 398ndash408

Akaike H 1973 Information theory and an extension of themaximum likelihood principle In Petrov BN Csaki F edsSecond International Symposium on Information Theory Bu-dapest Akademiai Kiado 267ndash281

Amitai P Bouskila A 2001 Handbook of amphibians ampreptiles of Israel Israel Keter Publishing House Ltd

Anderson SC 1999 The lizards of Iran Ithaca NY Societyfor the Study of Amphibians and Reptiles

Arnold EN 1980 The scientific results of the Oman flora andfauna survey 1977 (Dhofar) The reptiles and amphibians ofDhofar southern Arabia Journal of Oman Studies SpecialReport 2 273ndash332

Arnold EN 1983 Osteology genitalia and the relationshipsof Acanthodactylus (Reptilia Lacertidae) Bulletin of the BritishMuseum (Natural History) Zoology 44 291ndash339

Arnold EN Arribas Oacute Carranza S 2007 Systematics ofthe Palaearctic and Oriental lizard tribe Lacertini (SquamataLacertidae Lacertinae) with descriptions of eight new generaZootaxa 1430 1ndash86

Arnold EN Robinson MD Carranza S 2009 A prelimi-nary analysis of phylogenetic relationships and biogeogra-phy of the dangerously venomous Carpet Vipers Echis(Squamata Serpentes Viperidae) based on mitochondrial DNAsequences Amphibia-Reptilia 30 273ndash282

Audouin JV 1827 Explication sommaire des planches dereptiles (suppleacutement) offrant un exposeacute des characteacuteresdes espegraveces In Savigny MJCL ed Description de lrsquoEacutegypteVol 1 Histoire Naturelle Paris Impeacuteriale 161ndash184

Bache F Popescu SM Rabineau M Gorini C Suc JPClauzon G Olivet JL Rubino JL Melinte-DobrinescuMC Estrada F 2012 A two-step process for the refloodingof the Mediterranean after the Messinian Salinity Crisis BasinResearch 24 125ndash153

Baha El Din SM 2006 A guide to the reptiles and amphib-ians of Egypt CairondashNew York American University in CairoPress

Baier FS Sparrow DJ Wiedl H-J 2009 The amphibiansand reptiles of Cyprus Frankfurt am Main Edition Chimaira

Bandelt H-J Forster P Roumlhl A 1999 Median-joining net-works for inferring intraspecific phylogenies Molecular Biologyand Evolution 16 37ndash48

Bar A Haimovitch G 2011 A field guide to reptiles and am-phibians of Israel Herzliya Pazbar Limited

Boumlttger O 1879 Reptilien und Amphibien aus Syrien Berichtuumlber die Senckenbergische Naturforschende Gesellschaft inFrankfurt am Main 1879 57ndash84

Boulenger GA 1919 On a new variety of Acanthodactylusboskianus Daud from the Euphrates The Annals and Maga-zine of Natural History 3 549ndash550

Boulenger GA 1878 Sur les espegraveces drsquoAcanthodactylus desbords de la Mediterraneacutee Bulletin de la Socieacuteteacute zoologiquede France 3 179ndash197

Boulenger GA 1918 Sur les leacutezards du genre AcanthodactylusWieg Bulletin de la Socieacuteteacute zoologique de France 43 143ndash155

Boulenger GA 1921 Monograph of the Lacertidœ Vol II Trus-tees of the British Museum of Natural History London

Carranza S Arnold EN 2012 A review of the geckos of thegenus Hemidactylus (Squamata Gekkonidae) fromOman based on morphology mitochondrial and nuclear datawith descriptions of eight new species Zootaxa 3378 1ndash95

Carranza S Arnold EN Geniez P Roca J Mateo J 2008Radiation multiple dispersal and parallelism in the skinksChalcides and Sphenops (Squamata Scincidae) with com-ments on Scincus and Scincopus and the age of the SaharaDesert Molecular Phylogenetics and Evolution 46 1071ndash1094

Carretero MA Fonseca MM Garcia-Munoz E Brito JCHarris DJ 2011 Adding Acanthodactylus beershebensis tothe mtDNA phylogeny of the Acanthodactylus pardalis groupNorth-Western Journal of Zoology 7 138ndash142

Castresana J 2000 Selection of conserved blocks from multi-ple alignments for their use in phylogenetic analysis Mo-lecular Biology and Evolution 17 540ndash552

Clube TMM Robertson A 1986 The palaeorotation of theTroodos microplate Cyprus in the Late Mesozoic-Early Ce-nozoic plate tectonic framework of the Eastern Mediterra-nean Surveys in Geophysics 8 375ndash437

Cox SC Carranza S Brown RP 2010 Divergence timesand colonization of the Canary Islands by Gallotia lizardsMolecular Phylogenetics and Evolution 56 747ndash757

Crochet PA Geniez P Ineich I 2003 A multivariate analy-sis of the fringe-toed lizards of the Acanthodactylus scutellatusgroup (Squamata Lacertidae) systematic and biogeographi-cal implications Zoological Journal of the Linnean Society137 117ndash155

Crochet P-A Chaline O Surget-Groba Y Debain CCheylan M 2004 Speciation in mountains phylogeographyand phylogeny of the rock lizards genus Iberolacerta (ReptiliaLacertidae) Molecular Phylogenetics and Evolution 30 860ndash866

Daudin FM 1802 Histoire naturelle geacuteneacuterale et particuliegraveredes reptiles ouvrage faisant suite agrave lrsquoHistoire naturelle geacuteneacuteraleet particuliegravere F Dufart

Disi A Modry D Necas P Rifai L 2001 Amphibians andreptiles of the Hashemite Kingdom of Jordan an atlas andfield guide Frankfurt am Main Chimaira

Dolev A Pervolutzki A 2004 Endangered species in IsraelRed list of threatened animals Vertebrates JerusalemThe Nature and Parks Authority and the Society for thePreservation of Nature

Drummond AJ Rambaut A 2007 BEAST Bayesian evo-lutionary analysis by sampling trees BMC EvolutionaryBiology 7 214

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 17

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Ezard T Fujisawa T Barraclough T 2009 Splits speciesrsquolimits by threshold statistics R package version 10-19r48Available at httpR-ForgeR-projectorgprojectssplits

Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783ndash791

Fitzinger LJFJ 1834 Acanthodactylus In Wiegman AFAed Herpetologia mexicana seu Descriptio amphibiorum NovaeHispaniae quae itineribus comitis De Sack Ferdinandi Deppeet Chr Guil Schiede in Museum zoologicum Berolinensepervenerunt Pars prima saurorum species amplectens adjectosystematis saurorum prodromo additisque multis in huncamphibiorum ordinem observationibus Berlin C G Luderitz10

Flot JF 2010 SeqPHASE a web tool for interconvertingPHASE inputoutput files and FASTA sequence align-ments Molecular Ecology Resources 10 162ndash166

Fonseca MM Brito JC Paulo OS Carretero MA HarrisDJ 2009 Systematic and phylogeographical assessment ofthe Acanthodactylus erythrurus group (Reptilia Lacertidae)based on phylogenetic analyses of mitochondrial and nuclearDNA Molecular Phylogenetics and Evolution 51 131ndash142

Fonseca MM Brito JC Rebelo H Kalboussi M LarbesS Carretero MA Harris DJ 2008 Genetic variation amongspiny-footed lizards in the Acanthodactylus pardalis groupfrom North Africa African Zoology 43 8ndash15

Fontaneto D Herniou EA Boschetti C Caprioli M MeloneG Ricci C Barraclough TG 2007 Independently evolv-ing species in asexual bdelloid rotifers PLoS Biology 5 e87

Franzen M 1998 Erstnachweis von Acanthodactylus schreiberischreiberi Boulenger 1879 fuumlr die Tuumlrkei (Squamata SauriaLacertidae) Herpetozoa 11 27ndash36

Funk DJ Omland KE 2003 Species-level paraphyly andpolyphyly frequency causes and consequences with in-sights from animal mitochondrial DNA Annual Review ofEcology Evolution and Systematics 34 397ndash423

Godinho R Crespo EG Ferrand N Harris DJ 2005 Phy-logeny and evolution of the green lizards Lacerta spp(Squamata Lacertidae) based on mitochondrial and nuclearDNA sequences Amphibia-Reptilia 26 271ndash285

Greenbaum E Villanueva CO Kusamba C Aristote MMBranch WR 2011 A molecular phylogeny of EquatorialAfrican Lacertidae with the description of a new genus andspecies from eastern Democratic Republic of the Congo Zoo-logical Journal of the Linnean Society 163 913ndash942

Guillou H Carracedo JC Torrado FP Badiola ER 1996K-Ar ages and magnetic stratigraphy of a hotspot-inducedfast grown oceanic island El Hierro Canary Islands Journalof Volcanology and Geothermal Research 73 141ndash155

Harris DJ Arnold EN 2000 Elucidation of the relation-ships of spiny-footed lizards Acanthodactylus spp (ReptiliaLacertidae) using mitochondrial DNA sequence with com-ments on their biogeography and evolution Journal of Zoology252 351ndash362

Harris DJ Batista V Carretero M 2004 Assessment ofgenetic diversity within Acanthodactylus erythrurus (ReptiliaLacertidae) in Morocco and the Iberian Peninsula usingmitochondrial DNA sequence data Amphibia-Reptilia 25227

Horowitz A 1979 The Quaternary of Israel New York Aca-demic Press

Hraoui-Bloquet S Sadek RA Sindaco R Venchi A 2002The herpetofauna of Lebanon new data on distributionZoology in the Middle East 27 35ndash46

Hsuuml KJ Montadert L Bernoulli D Cita MB EricksonA Garrison RE Kidd RB Megraveliereacutes F Muumlller C WrightR 1977 History of the Mediterranean salinity crisis Nature267 399ndash403

Huelsenbeck JP Rannala B 2004 Frequentist propertiesof Bayesian posterior probabilities of phylogenetic trees undersimple and complex substitution models Systematic Biology53 904ndash913

Huelsenbeck JP Ronquist F 2001 MRBAYES Bayesianinference of phylogenetic trees Bioinformatics 17 754ndash755

Jolivet L Augier R Robin C Suc J-P Rouchy JM 2006Lithospheric-scale geodynamic context of the Messinian sa-linity crisis Sedimentary Geology 188 9ndash33

Kapli P Lymberakis P Poulakakis N Mantziou GParmakelis A Mylonas M 2008 Molecular phylogeny ofthree Mesalina (Reptilia Lacertidae) species (M guttulataM brevirostris and M bahaeldini) from North Africa and theMiddle East another case of paraphyly MolecularPhylogenetics and Evolution 49 102ndash110

Katoh K Toh H 2008 Recent developments in the MAFFTmultiple sequence alignment program Briefings inBioinformatics 9 286ndash298

Krijgsman W Hilgen F Raffi I Sierro F Wilson D 1999Chronology causes and progression of the Messinian salin-ity crisis Nature 400 652ndash655

Lanfear R Calcott B Ho SY Guindon S 2012PartitionFinder combined selection of partitioning schemesand substitution models for phylogenetic analyses Molecu-lar Biology and Evolution 29 1695ndash1701

Le Houeacuterou HN 1997 Climate flora and fauna changes inthe Sahara over the past 500 million years Journal of AridEnvironments 37 619ndash647

Maca-Meyer N Carranza S Rando J Arnold E CabreraV 2003 Status and relationships of the extinct giant CanaryIsland lizard Gallotia goliath (Reptilia Lacertidae)assessed using ancient mtDNA from its mummifiedremains Biological Journal of the Linnean Society 80 659ndash670

McCallum J Robertson A 1990 Pulsed uplift of the TroodosMassif ndash evidence from the Plio-Pleistocene Mesaoria basinIn J Malpas EMM Panayiotou A Xenophontos C edsOphiolites oceanic crustal analogues Proceedings of theSymposium lsquoTroodos 1987rsquo Nicosia (The Geological SurveyDepartment Ministry of Agriculture and NaturalResources) 217ndash229

Metallinou M Arnold EN Crochet P-A Geniez P BritoJC Lymberakis P El Din SB Sindaco R Robinson MCarranza S 2012 Conquering the Sahara and Arabiandeserts systematics and biogeography of Stenodactylus geckos(Reptilia Gekkonidae) BMC Evolutionary Biology 12258

Monaghan MT Wild R Elliot M Fujisawa T Balke MInward DJ Lees DC Ranaivosolo R Eggleton P

18 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

partition approaches (ie by gene and by PartitionFinderFigs S2 S3 respectively) The two partition ap-proaches gave similar clusters but at the less sup-ported nodes they differed at the positions of severallineages The single threshold GMYC result is indi-cated for a single line at 00037 Mya for the gene par-titions and at 002 Mya for PartitionFinder (verticallines in Figs S2 S3) The topology and clusters re-vealed in this analysis correspond to the lineages fromthe phylogeny of the ML and BI methods both for theparaphyly of the two species and the geographical group-ings within A b asper The GMYC results mainly differfrom the ML and BI methods in the position ofA schreiberi from Cyprus and Turkey as a sister cladeto the Syrian A b asper

DISCUSSION

We have provided a comprehensive and thorough as-sessment of the intraspecific phylogenetic relation-ships within A schreiberi and its closest relativeA b asper Our results based on mitochondrial andnuclear DNA data from 84 specimens across the entiredistribution range of A schreiberi and most of the dis-tribution range of A b asper reveal that A schreiberiis paraphyletic and nested entirely within theA boskianus subspecies

HISTORICAL BIOGEOGRAPHY

Acanthodactylus schreiberi is thought to comprisethree subspecies corresponding to three allopatricpopulations in Cyprus Turkey and IsraelminusLebanonThe Cypriot endemic nominotypical subspeciesA sc schreiberi and the Turkish subspeciesA sc ataturi cluster together (to form clade B Fig 2)nesting between A b asper clades This lineage is sisterto a clade of A b asper including A sc syriacus (cladeC Fig 2) We estimate the divergence time of theCypriotminusTurkish lineage of A schreiberi to have beenduring the late Miocene around 6 Mya although thereis no support for this split in the tree In other analy-ses using the whole genus this split is well support-ed in Bayesian analyses (K Tamar S Carranza RSindaco J Moravec JF Trape amp S Meriri unpubldata) Based on mitochondrial data Poulakakis et al(2013) found that both the Cypriot and Turkish sub-species are monophyletic and diverged from each other085 Mya (038ndash156 Mya) According to our results thisdate corresponds to an inner divergence of the

A sc schreiberi lineage rather than to the date at whichA sc schreiberi colonized Cyprus

The discrepancy in the phylogenetic relationship ofA sc schreiberi raises questions regarding the arrivalon Cyprus Cyprus originated with the raising of theTroodos Massif during the upper Cretaceous c 91 to88 Mya (Clube amp Robertson 1986 Mukasa amp Ludden1987) During the middle to late Miocene only a smallproportion of Cyprus was exposed above the Mediter-ranean (McCallum amp Robertson 1990 Robertson 1990)Towards the end of the Miocene sim596 Mya with theclosing of the passage between the Atlantic Ocean andthe Mediterranean basin the Messinian salinity crisisbegan (Krijgsman et al 1999) This resulted in thedrying up of much of the Mediterranean Sea and highsea-mounts emerged to form land bridges with thesurrounding land (Hsuuml et al 1977) By the end of theMiocene and early Pliocene sim533 Mya the passagewith the Atlantic Ocean reopened and the Mediterra-nean basin was refilled (Krijgsman et al 1999) Re-sulting from compressions raising and uplifting of thesurrounding areas towards and during the Pleisto-cene Cyprus was a complete emergent island (McCallumamp Robertson 1990) The possible connection of Cyprusto the mainland (ie to TurkeySyria) during theMessinian is debated as are suggestions of a land con-nection at later periods (Steininger amp Roumlgl 1984 Jolivetet al 2006 Bache et al 2012) Such a connection ifit existed could have provided access for terrestrialorganisms with poor overseas dispersal ability suchas lizards to colonize the island Several studies arguethat post-Messinian sea level changes are unlikely tohave formed connections between Cyprus and the main-land (Steininger amp Roumlgl 1984 Jolivet et al 2006) Thusour dating of the split between the Cypriot A schreiberiand A b asper at c 6 Mya leads us to suggest that theancestor of A s schreiberi colonized Cyprus from themainland through a land bridge connection at the be-ginning of the Messinian crisis rather than by a muchlatermore recent transmarine dispersal as suggestedby Poulakakis et al (2013) Owing to its close rela-tions with A b asper the ancestor of A schreiberi waspresumably mainland A boskianus and the cladogenesisleading to A schreiberi thus rendered A b asperparaphyletic

The Turkish subspecies A schreiberi ataturi was rec-orded for the first time by Franzen (1998) at a veryrestricted area of around 15 km of coastal strip (betweenBotas and Yukarı Burnaz Hatay Province) Owing tothe remarkable morphological similarity between

Figure 3 Haplotype networks of the nuclear gene fragments melano-cortin 1 receptor (MC1R) acetylcholinergic recep-tor M4 (ACM4) and oocyte maturation factor MOS (c-mos) with colours corresponding to Figures 1 and 2 Codes corre-late to the two alleles (ie a and b) of specimens in Table 1 Circle sizes are proportional to the number of alleles (Colourversion of figure available online)

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 13

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

A sc ataturi and the Cypriot population the speci-mens were initially identified as A sc schreiberi(Franzen 1998) Yalccedilinkaya amp Goumlccedilmen (2012) howeverdescribed this population as a new distinct subspe-cies A s ataturi presenting several differences betweenthe two in both morphology and blood-serum pro-teins The origin of A s ataturi remains uncertain asit is debated whether the newly discovered Turkishpopulation is a relict or an introduction from CyprusFranzen (1998) described this population as a pos-sible introduction from Cyprus through the harbourof Botas but Sindaco et al (2000) suggested thatit might be a relict of a previously larger popula-tion because its present distribution is similar tothat of some insects and lizards [Archaeolacerta(Phoenicolacerta) laevis and Ablepharus budaki]Yalccedilinkaya amp Goumlccedilmen (2012) proposed that A sc ataturiarrived in Turkey from the nominate population inCyprus during the Messinian crisis The phylogeneticresults haplotype networks and low levels of geneticdivergence we found suggest that the two subspeciesfrom Cyprus and Turkey have not been geneticallyisolated for a long period of time (ie they share allelesin all three nuclear genes and A s ataturi is nestedwithin A s schreiberi in the phylogeny Figs 2 3) Ourresults therefore contrast with the two latter sce-narios of a relict population or a Messinian disper-sal Both divergence time and the genetic similarityof the two subspecies agree with the original sugges-tion of Franzen (1998) that these animals were intro-duced into Turkey from Cyprus Further support forthis hypothesis is that A s ataturi is restricted to thevicinity of the Botas-Adana harbour and is absent inother suitable habitats (coastal sand dunes) wide-spread in south-eastern Turkey Its close morphologi-cal features to A s schreiberi (Franzen 1998) likewisesupport an introduction scenario

The third subspecies A s syriacus is nested withinA b asper in the concatenated mtDNA and nDNA treesand is clearly genetically distinct from the Cyprus andTurkey A schreiberi lineage The close relations ofA s syriacus with A boskianus may shed light on theorigin of the former Acanthodactylus schreiberi syriacusis distributed on stable sands of the coastal plain ofthe eastern Mediterranean in Israel and southernLebanon (Salvador 1982 Hraoui-Bloquet et al 2002Bar amp Haimovitch 2011) habitats resembling those ofA s schreiberi from Cyprus (Baier Sparrow amp Wiedl2009) The oldest divergence of the A b asper cladethat includes A s syriacus is estimated to have oc-curred during the late Miocene around 558 Mya butno further dates are available for the grouping ofA s syriacus as a result of low support values Thecoastal plain of the eastern Mediterranean was sub-merged during the late Miocene and re-emerged onlytoward the Pliocene (Nir 1970 Horowitz 1979) The

sands of the coastal plain where A s syriacus occurs(Salvador 1982 Arnold 1983 K Tamar amp S Meiripers observ) were repeatedly submerged and re-emerged during the Pleistocene sea-level changes (duringinterglacial and glacial periods respectively) A pos-sible scenario for A sc syriacusrsquos origin includes severalwaves of dispersal of Middle Eastern A boskianus whichoccurs on coarse substrates (Amitai amp Bouskila 2001Disi et al 2001 Baha El Din 2006 pers observ) towardthe Mediterranean shore Acanthodactylus boskianusasper is absent from Mediterranean climate habitatsin Lebanon and Israel It occupies only xeric zonessuggesting an invasion to the coastal plain when sandyhabitats allowed desert flora and fauna to migrate north-wards (Yom-Tov 1988) These populations adapted tosandy soils and evolved morphological features thatdistinguish them from the desert hard substrate formsof A b asper We view this as the most likely scenariogiven the biogeography the phylogenetic results andthe habitat preferences and adaptations of these lizardsAn alternative scenario according to which the an-cestor of A schreiberi originated in Cyprus and dis-persed to the shores of Israel and Lebanon (or originatedin the coastal plain of the Eastern Mediterranean anddispersed to Cyprus) we regard as far less likely Sucha scenario requires much closer genetic relationshipsbetween these two forms and is further weakened bythe close relationship between A s syriacus and thegeographically adjacent A b asper populations

Acanthodactylus boskianus asper is highly variableboth morphologically (Salvador 1982 Arnold 1983) andgenetically (this study) The subspecies is paraphyleticas A schreiberi is nested within it The topology of theA b asper tree shows four different geographical group-ings Syria (clade A) north Jordan plus north OmanNorth Africa and Middle East plus south Arabia (groupsC1 C2 C3 respectively) The different groups in thissubspecies are estimated to have first diverged duringthe late Miocene approximately 65 Mya with the splitof the Syrian population The Syrian lineage is ge-netically distant from A schreiberi and the otherA b asper specimens The nuclear networks indicatethat this group is closer to the other A b asper samplesrather than to A schreiberi The geographical splitsin the rest of the A b asper range (clade C) are esti-mated to have started around 558 Mya These groupsare supported as a distinct clade but are closely relatedto each other in both the concatenated and nuclear trees(Figs 2 S1 respectively) The diversification within thisclade is estimated to have occurred during the lateMiocene to early Pliocene when A b asper dispersedwidely west to North Africa and in Arabia The di-vergence within the North African group (group C2)is estimated to have occurred during the Pliocene ap-proximately 456 Mya with the Egyptian Nigerian andSudanese populations later dispersing west and north

14 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

in Africa This diversification correlates to the arid climatestarting in southern Sahara during the early-mid Plio-cene and later in northern Africa between the Plio-cene and the Pleistocene (Le Houeacuterou 1997) as hasbeen suggested for the dispersal of Mesalina guttulatain Africa (Kapli et al 2008) Other evidence relatesdry climate in North Africa to an earlier period around7 Mya (Schuster et al 2006) as has been suggestedfor the genus Chalcides and other reptiles (Carranzaet al 2008 Metallinou et al 2012 and reference therein)The aridification of North Africa has most likely con-tributed for the successful dispersal of A b asper westfrom south-west Asia into Africa Morphological studiesof A boskianus show relatively uniform populations inNorth Africa suggesting recent migration (Salvador1982 Arnold 1983) The other two geographical group-ings of A b asper from the Middle East and Arabia(groups C1 and C3) are located in two distinct innerclades but their location within each inner clade ispoorly supported The topology of the concatenated tree(Fig 2) shows that the group from northern Jordanand northern Oman (group C1) is closer to the NorthAfrican one than to the geographically close Middle-Eastern and south Arabian group (group C3) The taxo-nomic separation between north and south Oman hasbeen recognized in other species of reptiles and sup-ported by the topography of Oman (eg Echis coloratusand Echis omanensis Arnold Robinson amp Carranza2009) In the nuclear tree (Fig S1) these two groupsare closer to one another and with the North Africangroup form clade C Therefore the low support valuesamongst these groupings prevent an appropriate andthorough analysis of this subspecies The close rela-tionship amongst the geographical groups may reflectclose phylogenetic relationships amongst these popu-lations suggesting recent migration divergence andongoing gene flow

SYSTEMATICS AND TAXONOMIC IMPLICATIONS

The relationships within the A boskianus species groupconflict with the current known taxonomy of A schreiberiand A b asper (samples of the other subspecies ofA boskianus and of A nilsoni were unavailable for thisstudy) Both species have been found to be closelyrelated and paraphyletic The constrained topology testsexemplify the close entangled relationship between thetwo species as the separate monophyly of the two specieswas not rejected and the enforced monophyly of themboth together was inconclusive

Several causes can be responsible for paraphyly inspecies such in the A boskianus species group (Funkamp Omland 2003 and references therein) (1) inad-equate phylogenetic information (2) imperfect taxono-my (incorrectinaccurate species limits) derived frommisidentifying intraspecific variation (3) interspecific

gene flow ndash hybridization through interspecific matingand the subsequent backcrossing of hybrids into theparental populations (4) incomplete lineage sortingbecause of recent speciation events (5) unrecognizedparalogy We suggest that the relationships betweenA schreiberi and A b asper based on mitochondrialand nuclear data are most likely explained by incor-rect taxonomy probably because of the great variabil-ity of the latter species and to convergence As wasthe case in the molecular studies of the A pardalisand A erythrurus species groups (Harris et al 2004Fonseca et al 2008 2009 Carretero et al 2011 andreference therein) there are many problems with thecurrent taxonomic status of several species groups withinAcanthodactylus

Taking the molecular results of our study into accountthere are several systematic approaches to classify-ing the A schreiberiminusA b asper clade The Cypriot andTurkish populations of A schreiberi are very closelyrelated with the latter nested in the former and thetwo subspecies share nuclear alleles (Fig 3) Further-more the low uncorrected p-distance is positively cor-related with subspecies-level distances within otherlacertid species (ie 16 of Cytb in Lacerta bilineatachloronota Godinho et al 2005) We therefore con-clude that Cypriot animals were recently introducedto Turkey and that the Turkish population does notmerit a subspecific rank We suggest that A s ataturiYalccedilinkaya amp Goumlccedilmen 2012 is a junior synonym ofA s schreiberi Boulenger 1878

Regarding the relationships between A schreiberi andA b asper a few scenarios are possible One is to sinkA schreiberi within A boskianus to create one species(A boskianus) with high genetic and morphological vari-ability ranging over a broad distribution Another isfor the two taxa be regarded as a species complex (theA boskianus-schreiberi complex) until further inves-tigation on the subject However although A schreiberiis nested within A b asper the populations from Cyprusand Turkey represent a distinct evolutionary lineagewith distinct genetic and morphological features andthus it is logical to retain the specific status Two othersolutions are possible The first is to re-evaluate theSyrian populations and to consider elevating them aswell as the more divergent lineages (and subspecies)of A boskianus to specific status This would neces-sitate an examination of the phylogeny and morphol-ogy of the other four subspecies of A boskianus(A b boskianus A b euphraticus A b khattensis andA b nigeriensis) and the identification of distinctivephenotypic features in the Syrian lizards Another so-lution is to recognize the maintenance of gene flowamongst mainland populations of A b asper after thedivergence of the insular endemic A schreiberi andthus the evolutionary cohesion of the paraphyleticA b asper Arnold (1983) noted that A schreiberi may

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 15

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

have originated as an isolate from A boskianus becauseof their shared morphology and hemipenis featuresOur results support this scenario which includes thedispersal of A schreiberi to Cyprus from a mainlandpopulation that was most probably A boskianus It maybe assumed that the ancestor of the Cypriot A schreiberiafter arriving on Cyprus remained isolated for a longperiod of time and thus evolved to the modern formof A schreiberi Meanwhile the same ancestral con-tinental populations not isolated from each other con-tinued to exchange genes to varying degrees remainingA boskianus

The IsraeliminusLebanese subspecies A sc syriacus is onlydistantly related to the nominate form A sc schreiberiThis subspecies is highly phylogenetically divergent fromthe Cypriot and Turkish populations having higherp-distances (12S 4 Cytb 11ndash12) than those foundbetween other lacertid species (eg 74ndash82 of Cytbamongst Iberolacerta aranica Iberolacerta aurelioi andIberolacerta bonnali and 41ndash58 of Cytb betweenLacerta bilineata and Lacerta viridis Crochet et al2004 Godinho et al 2005 respectively) The nuclearhaplotype networks further show that Lebanese andIsraeli populations share alleles only with A b asperbut not with the nominotypical Cypriot form Arnold(1983) suggested that the geographical variation ofA boskianus reflects niche differences with animalsfrom xeric areas with dense rigid and spiny vegeta-tion having larger dorsal scales than animals from moremesic areas As was assumed for A schreiberi wesuggest that other mainland populations of A b asperwere the ancestors of the LebaneseminusIsraeli Coastal plainforms We suggest that A s syriacus is an ecomorphof A b asper that dispersed from the usual xerichabitats of the species and adapted to the new moremesic environment of the stable sands of the coastalplains of the eastern Mediterranean As a conse-quence this ecomorph converged on the morphologyof A s schreiberi which inhabits the coastal sands ofCyprus (Baier et al 2009) but still maintains differ-entiating features by having coarser dorsal scales andsharp keels (Salvador 1982 Arnold 1983 Franzen1998) This convergence led to the description ofA s syriacus as a member of A schreiberi The mor-phological assessment and the close morphological simi-larities between A b asper and A s syriacus mayexplain the wrong classification A similar erro-neous reasoning led Reed amp Marx (1959) to identifyspecimens with fine scales from Iraq as A schreiberiSalvador (1982) re-examined these specimens and as-signed them to A boskianus The morphological dif-ferences between the two forms are less prominentespecially where the two forms occur in close geo-graphical proximity in the southern coastal plainand north-western Negev Desert of Israel (Bar ampHaimovitch 2011) According to our results A s syriacus

actually belongs to A b asper being a coastal-duneecomorph convergent with but evolutionarily dis-tinct from A schreiberi Thus our preferred scenariois to treat the name Acanthodactylus schreiberi syriacusBoumlttger 1879 (which was originally described asA boskianus var syriacus by Boumlttger 1879) as a juniorsynonym of the name Acanthodactylus boskianus asper(Audouin 1827)

Recognizing A s syriacus as a junior synonym ofA b asper may have important implications for the con-servation of this coastal sand dune form which is clas-sified as critically endangered in Israel (Dolev ampPervolutzki 2004) However as the Israeli and Leba-nese coastal dune ecosystem has probably developedonly very recently during the Quaternary (Nir 1970Horowitz 1979) this form represents a remarkable caseof rapid evolutionary change It is also a remarkablecase of convergent evolution (with the CypriotA sc schreiberi) Thus we feel that these popula-tions are unique evolutionary entities that merit specialconservation efforts

The use of nuclear genes is a valuable method forestimating species divergence and lineage sorting andhelps evaluate isolated lineages and evolutionary historyThe incorporation of mitochondrial and nuclear dataprovides thorough topologies informative networks anddivergence times that reveal useful information for aproblematic taxonomy such as that of the A boskianusspecies group We have shown that phylogenetic ap-proaches to the confusing taxonomy of two closely relatedand morphologically similar species can shed light ontheir unclear relationships resolve between homoplasyand shared ancestry and identify patterns of speciesevolutionary history and biogeography

ACKNOWLEDGEMENTS

We are grateful to the following people for providingsamples for this study D Donaire B Shacham J ŠmiacutedS M Baha El Din and J Padial We thank MMetallinou and J Šmiacuted for help with the lab work andthe analyses M Novosolov for help with the figuresY L Werner B Shacham and A Levy for fruitful dis-cussions and two anonymous referees for helpful com-ments on an earlier version of this manuscript Wethank the Israel nature and park authority for issuingcollecting permits (nos 38074 38451 38489 38540)The work of J M was financially supported by the Min-istry of Culture of the Czech Republic (DKRVO 201315 National Museum 00023272) S C is funded bygrant CGL2012-36970 from the Ministerio de Economiacuteay Competitividad Spain (cofunded by FEDER) Thisstudy was funded by a Israel Taxonomic Initiative grantto S M K T is supported by an Israel Taxonomic Ini-tiative scholarship

16 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

REFERENCES

Ahmadzadeh F Carretero MA Harris DJ Perera ABohme W 2012 A molecular phylogeny of the eastern groupof ocellated lizard genus Timon (Sauria Lacertidae) basedon mitochondrial and nuclear DNA sequences Amphibia-Reptilia 33 1ndash10

Ahmadzadeh F Flecks M Roumldder D Boumlhme W Ilgaz CcedilHarris DJ Engler JO Uumlzuumlm N Carretero MA 2013Multiple dispersal out of Anatolia biogeography and evolu-tion of oriental green lizards Biological Journal of the LinneanSociety 110 398ndash408

Akaike H 1973 Information theory and an extension of themaximum likelihood principle In Petrov BN Csaki F edsSecond International Symposium on Information Theory Bu-dapest Akademiai Kiado 267ndash281

Amitai P Bouskila A 2001 Handbook of amphibians ampreptiles of Israel Israel Keter Publishing House Ltd

Anderson SC 1999 The lizards of Iran Ithaca NY Societyfor the Study of Amphibians and Reptiles

Arnold EN 1980 The scientific results of the Oman flora andfauna survey 1977 (Dhofar) The reptiles and amphibians ofDhofar southern Arabia Journal of Oman Studies SpecialReport 2 273ndash332

Arnold EN 1983 Osteology genitalia and the relationshipsof Acanthodactylus (Reptilia Lacertidae) Bulletin of the BritishMuseum (Natural History) Zoology 44 291ndash339

Arnold EN Arribas Oacute Carranza S 2007 Systematics ofthe Palaearctic and Oriental lizard tribe Lacertini (SquamataLacertidae Lacertinae) with descriptions of eight new generaZootaxa 1430 1ndash86

Arnold EN Robinson MD Carranza S 2009 A prelimi-nary analysis of phylogenetic relationships and biogeogra-phy of the dangerously venomous Carpet Vipers Echis(Squamata Serpentes Viperidae) based on mitochondrial DNAsequences Amphibia-Reptilia 30 273ndash282

Audouin JV 1827 Explication sommaire des planches dereptiles (suppleacutement) offrant un exposeacute des characteacuteresdes espegraveces In Savigny MJCL ed Description de lrsquoEacutegypteVol 1 Histoire Naturelle Paris Impeacuteriale 161ndash184

Bache F Popescu SM Rabineau M Gorini C Suc JPClauzon G Olivet JL Rubino JL Melinte-DobrinescuMC Estrada F 2012 A two-step process for the refloodingof the Mediterranean after the Messinian Salinity Crisis BasinResearch 24 125ndash153

Baha El Din SM 2006 A guide to the reptiles and amphib-ians of Egypt CairondashNew York American University in CairoPress

Baier FS Sparrow DJ Wiedl H-J 2009 The amphibiansand reptiles of Cyprus Frankfurt am Main Edition Chimaira

Bandelt H-J Forster P Roumlhl A 1999 Median-joining net-works for inferring intraspecific phylogenies Molecular Biologyand Evolution 16 37ndash48

Bar A Haimovitch G 2011 A field guide to reptiles and am-phibians of Israel Herzliya Pazbar Limited

Boumlttger O 1879 Reptilien und Amphibien aus Syrien Berichtuumlber die Senckenbergische Naturforschende Gesellschaft inFrankfurt am Main 1879 57ndash84

Boulenger GA 1919 On a new variety of Acanthodactylusboskianus Daud from the Euphrates The Annals and Maga-zine of Natural History 3 549ndash550

Boulenger GA 1878 Sur les espegraveces drsquoAcanthodactylus desbords de la Mediterraneacutee Bulletin de la Socieacuteteacute zoologiquede France 3 179ndash197

Boulenger GA 1918 Sur les leacutezards du genre AcanthodactylusWieg Bulletin de la Socieacuteteacute zoologique de France 43 143ndash155

Boulenger GA 1921 Monograph of the Lacertidœ Vol II Trus-tees of the British Museum of Natural History London

Carranza S Arnold EN 2012 A review of the geckos of thegenus Hemidactylus (Squamata Gekkonidae) fromOman based on morphology mitochondrial and nuclear datawith descriptions of eight new species Zootaxa 3378 1ndash95

Carranza S Arnold EN Geniez P Roca J Mateo J 2008Radiation multiple dispersal and parallelism in the skinksChalcides and Sphenops (Squamata Scincidae) with com-ments on Scincus and Scincopus and the age of the SaharaDesert Molecular Phylogenetics and Evolution 46 1071ndash1094

Carretero MA Fonseca MM Garcia-Munoz E Brito JCHarris DJ 2011 Adding Acanthodactylus beershebensis tothe mtDNA phylogeny of the Acanthodactylus pardalis groupNorth-Western Journal of Zoology 7 138ndash142

Castresana J 2000 Selection of conserved blocks from multi-ple alignments for their use in phylogenetic analysis Mo-lecular Biology and Evolution 17 540ndash552

Clube TMM Robertson A 1986 The palaeorotation of theTroodos microplate Cyprus in the Late Mesozoic-Early Ce-nozoic plate tectonic framework of the Eastern Mediterra-nean Surveys in Geophysics 8 375ndash437

Cox SC Carranza S Brown RP 2010 Divergence timesand colonization of the Canary Islands by Gallotia lizardsMolecular Phylogenetics and Evolution 56 747ndash757

Crochet PA Geniez P Ineich I 2003 A multivariate analy-sis of the fringe-toed lizards of the Acanthodactylus scutellatusgroup (Squamata Lacertidae) systematic and biogeographi-cal implications Zoological Journal of the Linnean Society137 117ndash155

Crochet P-A Chaline O Surget-Groba Y Debain CCheylan M 2004 Speciation in mountains phylogeographyand phylogeny of the rock lizards genus Iberolacerta (ReptiliaLacertidae) Molecular Phylogenetics and Evolution 30 860ndash866

Daudin FM 1802 Histoire naturelle geacuteneacuterale et particuliegraveredes reptiles ouvrage faisant suite agrave lrsquoHistoire naturelle geacuteneacuteraleet particuliegravere F Dufart

Disi A Modry D Necas P Rifai L 2001 Amphibians andreptiles of the Hashemite Kingdom of Jordan an atlas andfield guide Frankfurt am Main Chimaira

Dolev A Pervolutzki A 2004 Endangered species in IsraelRed list of threatened animals Vertebrates JerusalemThe Nature and Parks Authority and the Society for thePreservation of Nature

Drummond AJ Rambaut A 2007 BEAST Bayesian evo-lutionary analysis by sampling trees BMC EvolutionaryBiology 7 214

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 17

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Ezard T Fujisawa T Barraclough T 2009 Splits speciesrsquolimits by threshold statistics R package version 10-19r48Available at httpR-ForgeR-projectorgprojectssplits

Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783ndash791

Fitzinger LJFJ 1834 Acanthodactylus In Wiegman AFAed Herpetologia mexicana seu Descriptio amphibiorum NovaeHispaniae quae itineribus comitis De Sack Ferdinandi Deppeet Chr Guil Schiede in Museum zoologicum Berolinensepervenerunt Pars prima saurorum species amplectens adjectosystematis saurorum prodromo additisque multis in huncamphibiorum ordinem observationibus Berlin C G Luderitz10

Flot JF 2010 SeqPHASE a web tool for interconvertingPHASE inputoutput files and FASTA sequence align-ments Molecular Ecology Resources 10 162ndash166

Fonseca MM Brito JC Paulo OS Carretero MA HarrisDJ 2009 Systematic and phylogeographical assessment ofthe Acanthodactylus erythrurus group (Reptilia Lacertidae)based on phylogenetic analyses of mitochondrial and nuclearDNA Molecular Phylogenetics and Evolution 51 131ndash142

Fonseca MM Brito JC Rebelo H Kalboussi M LarbesS Carretero MA Harris DJ 2008 Genetic variation amongspiny-footed lizards in the Acanthodactylus pardalis groupfrom North Africa African Zoology 43 8ndash15

Fontaneto D Herniou EA Boschetti C Caprioli M MeloneG Ricci C Barraclough TG 2007 Independently evolv-ing species in asexual bdelloid rotifers PLoS Biology 5 e87

Franzen M 1998 Erstnachweis von Acanthodactylus schreiberischreiberi Boulenger 1879 fuumlr die Tuumlrkei (Squamata SauriaLacertidae) Herpetozoa 11 27ndash36

Funk DJ Omland KE 2003 Species-level paraphyly andpolyphyly frequency causes and consequences with in-sights from animal mitochondrial DNA Annual Review ofEcology Evolution and Systematics 34 397ndash423

Godinho R Crespo EG Ferrand N Harris DJ 2005 Phy-logeny and evolution of the green lizards Lacerta spp(Squamata Lacertidae) based on mitochondrial and nuclearDNA sequences Amphibia-Reptilia 26 271ndash285

Greenbaum E Villanueva CO Kusamba C Aristote MMBranch WR 2011 A molecular phylogeny of EquatorialAfrican Lacertidae with the description of a new genus andspecies from eastern Democratic Republic of the Congo Zoo-logical Journal of the Linnean Society 163 913ndash942

Guillou H Carracedo JC Torrado FP Badiola ER 1996K-Ar ages and magnetic stratigraphy of a hotspot-inducedfast grown oceanic island El Hierro Canary Islands Journalof Volcanology and Geothermal Research 73 141ndash155

Harris DJ Arnold EN 2000 Elucidation of the relation-ships of spiny-footed lizards Acanthodactylus spp (ReptiliaLacertidae) using mitochondrial DNA sequence with com-ments on their biogeography and evolution Journal of Zoology252 351ndash362

Harris DJ Batista V Carretero M 2004 Assessment ofgenetic diversity within Acanthodactylus erythrurus (ReptiliaLacertidae) in Morocco and the Iberian Peninsula usingmitochondrial DNA sequence data Amphibia-Reptilia 25227

Horowitz A 1979 The Quaternary of Israel New York Aca-demic Press

Hraoui-Bloquet S Sadek RA Sindaco R Venchi A 2002The herpetofauna of Lebanon new data on distributionZoology in the Middle East 27 35ndash46

Hsuuml KJ Montadert L Bernoulli D Cita MB EricksonA Garrison RE Kidd RB Megraveliereacutes F Muumlller C WrightR 1977 History of the Mediterranean salinity crisis Nature267 399ndash403

Huelsenbeck JP Rannala B 2004 Frequentist propertiesof Bayesian posterior probabilities of phylogenetic trees undersimple and complex substitution models Systematic Biology53 904ndash913

Huelsenbeck JP Ronquist F 2001 MRBAYES Bayesianinference of phylogenetic trees Bioinformatics 17 754ndash755

Jolivet L Augier R Robin C Suc J-P Rouchy JM 2006Lithospheric-scale geodynamic context of the Messinian sa-linity crisis Sedimentary Geology 188 9ndash33

Kapli P Lymberakis P Poulakakis N Mantziou GParmakelis A Mylonas M 2008 Molecular phylogeny ofthree Mesalina (Reptilia Lacertidae) species (M guttulataM brevirostris and M bahaeldini) from North Africa and theMiddle East another case of paraphyly MolecularPhylogenetics and Evolution 49 102ndash110

Katoh K Toh H 2008 Recent developments in the MAFFTmultiple sequence alignment program Briefings inBioinformatics 9 286ndash298

Krijgsman W Hilgen F Raffi I Sierro F Wilson D 1999Chronology causes and progression of the Messinian salin-ity crisis Nature 400 652ndash655

Lanfear R Calcott B Ho SY Guindon S 2012PartitionFinder combined selection of partitioning schemesand substitution models for phylogenetic analyses Molecu-lar Biology and Evolution 29 1695ndash1701

Le Houeacuterou HN 1997 Climate flora and fauna changes inthe Sahara over the past 500 million years Journal of AridEnvironments 37 619ndash647

Maca-Meyer N Carranza S Rando J Arnold E CabreraV 2003 Status and relationships of the extinct giant CanaryIsland lizard Gallotia goliath (Reptilia Lacertidae)assessed using ancient mtDNA from its mummifiedremains Biological Journal of the Linnean Society 80 659ndash670

McCallum J Robertson A 1990 Pulsed uplift of the TroodosMassif ndash evidence from the Plio-Pleistocene Mesaoria basinIn J Malpas EMM Panayiotou A Xenophontos C edsOphiolites oceanic crustal analogues Proceedings of theSymposium lsquoTroodos 1987rsquo Nicosia (The Geological SurveyDepartment Ministry of Agriculture and NaturalResources) 217ndash229

Metallinou M Arnold EN Crochet P-A Geniez P BritoJC Lymberakis P El Din SB Sindaco R Robinson MCarranza S 2012 Conquering the Sahara and Arabiandeserts systematics and biogeography of Stenodactylus geckos(Reptilia Gekkonidae) BMC Evolutionary Biology 12258

Monaghan MT Wild R Elliot M Fujisawa T Balke MInward DJ Lees DC Ranaivosolo R Eggleton P

18 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

A sc ataturi and the Cypriot population the speci-mens were initially identified as A sc schreiberi(Franzen 1998) Yalccedilinkaya amp Goumlccedilmen (2012) howeverdescribed this population as a new distinct subspe-cies A s ataturi presenting several differences betweenthe two in both morphology and blood-serum pro-teins The origin of A s ataturi remains uncertain asit is debated whether the newly discovered Turkishpopulation is a relict or an introduction from CyprusFranzen (1998) described this population as a pos-sible introduction from Cyprus through the harbourof Botas but Sindaco et al (2000) suggested thatit might be a relict of a previously larger popula-tion because its present distribution is similar tothat of some insects and lizards [Archaeolacerta(Phoenicolacerta) laevis and Ablepharus budaki]Yalccedilinkaya amp Goumlccedilmen (2012) proposed that A sc ataturiarrived in Turkey from the nominate population inCyprus during the Messinian crisis The phylogeneticresults haplotype networks and low levels of geneticdivergence we found suggest that the two subspeciesfrom Cyprus and Turkey have not been geneticallyisolated for a long period of time (ie they share allelesin all three nuclear genes and A s ataturi is nestedwithin A s schreiberi in the phylogeny Figs 2 3) Ourresults therefore contrast with the two latter sce-narios of a relict population or a Messinian disper-sal Both divergence time and the genetic similarityof the two subspecies agree with the original sugges-tion of Franzen (1998) that these animals were intro-duced into Turkey from Cyprus Further support forthis hypothesis is that A s ataturi is restricted to thevicinity of the Botas-Adana harbour and is absent inother suitable habitats (coastal sand dunes) wide-spread in south-eastern Turkey Its close morphologi-cal features to A s schreiberi (Franzen 1998) likewisesupport an introduction scenario

The third subspecies A s syriacus is nested withinA b asper in the concatenated mtDNA and nDNA treesand is clearly genetically distinct from the Cyprus andTurkey A schreiberi lineage The close relations ofA s syriacus with A boskianus may shed light on theorigin of the former Acanthodactylus schreiberi syriacusis distributed on stable sands of the coastal plain ofthe eastern Mediterranean in Israel and southernLebanon (Salvador 1982 Hraoui-Bloquet et al 2002Bar amp Haimovitch 2011) habitats resembling those ofA s schreiberi from Cyprus (Baier Sparrow amp Wiedl2009) The oldest divergence of the A b asper cladethat includes A s syriacus is estimated to have oc-curred during the late Miocene around 558 Mya butno further dates are available for the grouping ofA s syriacus as a result of low support values Thecoastal plain of the eastern Mediterranean was sub-merged during the late Miocene and re-emerged onlytoward the Pliocene (Nir 1970 Horowitz 1979) The

sands of the coastal plain where A s syriacus occurs(Salvador 1982 Arnold 1983 K Tamar amp S Meiripers observ) were repeatedly submerged and re-emerged during the Pleistocene sea-level changes (duringinterglacial and glacial periods respectively) A pos-sible scenario for A sc syriacusrsquos origin includes severalwaves of dispersal of Middle Eastern A boskianus whichoccurs on coarse substrates (Amitai amp Bouskila 2001Disi et al 2001 Baha El Din 2006 pers observ) towardthe Mediterranean shore Acanthodactylus boskianusasper is absent from Mediterranean climate habitatsin Lebanon and Israel It occupies only xeric zonessuggesting an invasion to the coastal plain when sandyhabitats allowed desert flora and fauna to migrate north-wards (Yom-Tov 1988) These populations adapted tosandy soils and evolved morphological features thatdistinguish them from the desert hard substrate formsof A b asper We view this as the most likely scenariogiven the biogeography the phylogenetic results andthe habitat preferences and adaptations of these lizardsAn alternative scenario according to which the an-cestor of A schreiberi originated in Cyprus and dis-persed to the shores of Israel and Lebanon (or originatedin the coastal plain of the Eastern Mediterranean anddispersed to Cyprus) we regard as far less likely Sucha scenario requires much closer genetic relationshipsbetween these two forms and is further weakened bythe close relationship between A s syriacus and thegeographically adjacent A b asper populations

Acanthodactylus boskianus asper is highly variableboth morphologically (Salvador 1982 Arnold 1983) andgenetically (this study) The subspecies is paraphyleticas A schreiberi is nested within it The topology of theA b asper tree shows four different geographical group-ings Syria (clade A) north Jordan plus north OmanNorth Africa and Middle East plus south Arabia (groupsC1 C2 C3 respectively) The different groups in thissubspecies are estimated to have first diverged duringthe late Miocene approximately 65 Mya with the splitof the Syrian population The Syrian lineage is ge-netically distant from A schreiberi and the otherA b asper specimens The nuclear networks indicatethat this group is closer to the other A b asper samplesrather than to A schreiberi The geographical splitsin the rest of the A b asper range (clade C) are esti-mated to have started around 558 Mya These groupsare supported as a distinct clade but are closely relatedto each other in both the concatenated and nuclear trees(Figs 2 S1 respectively) The diversification within thisclade is estimated to have occurred during the lateMiocene to early Pliocene when A b asper dispersedwidely west to North Africa and in Arabia The di-vergence within the North African group (group C2)is estimated to have occurred during the Pliocene ap-proximately 456 Mya with the Egyptian Nigerian andSudanese populations later dispersing west and north

14 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

in Africa This diversification correlates to the arid climatestarting in southern Sahara during the early-mid Plio-cene and later in northern Africa between the Plio-cene and the Pleistocene (Le Houeacuterou 1997) as hasbeen suggested for the dispersal of Mesalina guttulatain Africa (Kapli et al 2008) Other evidence relatesdry climate in North Africa to an earlier period around7 Mya (Schuster et al 2006) as has been suggestedfor the genus Chalcides and other reptiles (Carranzaet al 2008 Metallinou et al 2012 and reference therein)The aridification of North Africa has most likely con-tributed for the successful dispersal of A b asper westfrom south-west Asia into Africa Morphological studiesof A boskianus show relatively uniform populations inNorth Africa suggesting recent migration (Salvador1982 Arnold 1983) The other two geographical group-ings of A b asper from the Middle East and Arabia(groups C1 and C3) are located in two distinct innerclades but their location within each inner clade ispoorly supported The topology of the concatenated tree(Fig 2) shows that the group from northern Jordanand northern Oman (group C1) is closer to the NorthAfrican one than to the geographically close Middle-Eastern and south Arabian group (group C3) The taxo-nomic separation between north and south Oman hasbeen recognized in other species of reptiles and sup-ported by the topography of Oman (eg Echis coloratusand Echis omanensis Arnold Robinson amp Carranza2009) In the nuclear tree (Fig S1) these two groupsare closer to one another and with the North Africangroup form clade C Therefore the low support valuesamongst these groupings prevent an appropriate andthorough analysis of this subspecies The close rela-tionship amongst the geographical groups may reflectclose phylogenetic relationships amongst these popu-lations suggesting recent migration divergence andongoing gene flow

SYSTEMATICS AND TAXONOMIC IMPLICATIONS

The relationships within the A boskianus species groupconflict with the current known taxonomy of A schreiberiand A b asper (samples of the other subspecies ofA boskianus and of A nilsoni were unavailable for thisstudy) Both species have been found to be closelyrelated and paraphyletic The constrained topology testsexemplify the close entangled relationship between thetwo species as the separate monophyly of the two specieswas not rejected and the enforced monophyly of themboth together was inconclusive

Several causes can be responsible for paraphyly inspecies such in the A boskianus species group (Funkamp Omland 2003 and references therein) (1) inad-equate phylogenetic information (2) imperfect taxono-my (incorrectinaccurate species limits) derived frommisidentifying intraspecific variation (3) interspecific

gene flow ndash hybridization through interspecific matingand the subsequent backcrossing of hybrids into theparental populations (4) incomplete lineage sortingbecause of recent speciation events (5) unrecognizedparalogy We suggest that the relationships betweenA schreiberi and A b asper based on mitochondrialand nuclear data are most likely explained by incor-rect taxonomy probably because of the great variabil-ity of the latter species and to convergence As wasthe case in the molecular studies of the A pardalisand A erythrurus species groups (Harris et al 2004Fonseca et al 2008 2009 Carretero et al 2011 andreference therein) there are many problems with thecurrent taxonomic status of several species groups withinAcanthodactylus

Taking the molecular results of our study into accountthere are several systematic approaches to classify-ing the A schreiberiminusA b asper clade The Cypriot andTurkish populations of A schreiberi are very closelyrelated with the latter nested in the former and thetwo subspecies share nuclear alleles (Fig 3) Further-more the low uncorrected p-distance is positively cor-related with subspecies-level distances within otherlacertid species (ie 16 of Cytb in Lacerta bilineatachloronota Godinho et al 2005) We therefore con-clude that Cypriot animals were recently introducedto Turkey and that the Turkish population does notmerit a subspecific rank We suggest that A s ataturiYalccedilinkaya amp Goumlccedilmen 2012 is a junior synonym ofA s schreiberi Boulenger 1878

Regarding the relationships between A schreiberi andA b asper a few scenarios are possible One is to sinkA schreiberi within A boskianus to create one species(A boskianus) with high genetic and morphological vari-ability ranging over a broad distribution Another isfor the two taxa be regarded as a species complex (theA boskianus-schreiberi complex) until further inves-tigation on the subject However although A schreiberiis nested within A b asper the populations from Cyprusand Turkey represent a distinct evolutionary lineagewith distinct genetic and morphological features andthus it is logical to retain the specific status Two othersolutions are possible The first is to re-evaluate theSyrian populations and to consider elevating them aswell as the more divergent lineages (and subspecies)of A boskianus to specific status This would neces-sitate an examination of the phylogeny and morphol-ogy of the other four subspecies of A boskianus(A b boskianus A b euphraticus A b khattensis andA b nigeriensis) and the identification of distinctivephenotypic features in the Syrian lizards Another so-lution is to recognize the maintenance of gene flowamongst mainland populations of A b asper after thedivergence of the insular endemic A schreiberi andthus the evolutionary cohesion of the paraphyleticA b asper Arnold (1983) noted that A schreiberi may

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 15

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

have originated as an isolate from A boskianus becauseof their shared morphology and hemipenis featuresOur results support this scenario which includes thedispersal of A schreiberi to Cyprus from a mainlandpopulation that was most probably A boskianus It maybe assumed that the ancestor of the Cypriot A schreiberiafter arriving on Cyprus remained isolated for a longperiod of time and thus evolved to the modern formof A schreiberi Meanwhile the same ancestral con-tinental populations not isolated from each other con-tinued to exchange genes to varying degrees remainingA boskianus

The IsraeliminusLebanese subspecies A sc syriacus is onlydistantly related to the nominate form A sc schreiberiThis subspecies is highly phylogenetically divergent fromthe Cypriot and Turkish populations having higherp-distances (12S 4 Cytb 11ndash12) than those foundbetween other lacertid species (eg 74ndash82 of Cytbamongst Iberolacerta aranica Iberolacerta aurelioi andIberolacerta bonnali and 41ndash58 of Cytb betweenLacerta bilineata and Lacerta viridis Crochet et al2004 Godinho et al 2005 respectively) The nuclearhaplotype networks further show that Lebanese andIsraeli populations share alleles only with A b asperbut not with the nominotypical Cypriot form Arnold(1983) suggested that the geographical variation ofA boskianus reflects niche differences with animalsfrom xeric areas with dense rigid and spiny vegeta-tion having larger dorsal scales than animals from moremesic areas As was assumed for A schreiberi wesuggest that other mainland populations of A b asperwere the ancestors of the LebaneseminusIsraeli Coastal plainforms We suggest that A s syriacus is an ecomorphof A b asper that dispersed from the usual xerichabitats of the species and adapted to the new moremesic environment of the stable sands of the coastalplains of the eastern Mediterranean As a conse-quence this ecomorph converged on the morphologyof A s schreiberi which inhabits the coastal sands ofCyprus (Baier et al 2009) but still maintains differ-entiating features by having coarser dorsal scales andsharp keels (Salvador 1982 Arnold 1983 Franzen1998) This convergence led to the description ofA s syriacus as a member of A schreiberi The mor-phological assessment and the close morphological simi-larities between A b asper and A s syriacus mayexplain the wrong classification A similar erro-neous reasoning led Reed amp Marx (1959) to identifyspecimens with fine scales from Iraq as A schreiberiSalvador (1982) re-examined these specimens and as-signed them to A boskianus The morphological dif-ferences between the two forms are less prominentespecially where the two forms occur in close geo-graphical proximity in the southern coastal plainand north-western Negev Desert of Israel (Bar ampHaimovitch 2011) According to our results A s syriacus

actually belongs to A b asper being a coastal-duneecomorph convergent with but evolutionarily dis-tinct from A schreiberi Thus our preferred scenariois to treat the name Acanthodactylus schreiberi syriacusBoumlttger 1879 (which was originally described asA boskianus var syriacus by Boumlttger 1879) as a juniorsynonym of the name Acanthodactylus boskianus asper(Audouin 1827)

Recognizing A s syriacus as a junior synonym ofA b asper may have important implications for the con-servation of this coastal sand dune form which is clas-sified as critically endangered in Israel (Dolev ampPervolutzki 2004) However as the Israeli and Leba-nese coastal dune ecosystem has probably developedonly very recently during the Quaternary (Nir 1970Horowitz 1979) this form represents a remarkable caseof rapid evolutionary change It is also a remarkablecase of convergent evolution (with the CypriotA sc schreiberi) Thus we feel that these popula-tions are unique evolutionary entities that merit specialconservation efforts

The use of nuclear genes is a valuable method forestimating species divergence and lineage sorting andhelps evaluate isolated lineages and evolutionary historyThe incorporation of mitochondrial and nuclear dataprovides thorough topologies informative networks anddivergence times that reveal useful information for aproblematic taxonomy such as that of the A boskianusspecies group We have shown that phylogenetic ap-proaches to the confusing taxonomy of two closely relatedand morphologically similar species can shed light ontheir unclear relationships resolve between homoplasyand shared ancestry and identify patterns of speciesevolutionary history and biogeography

ACKNOWLEDGEMENTS

We are grateful to the following people for providingsamples for this study D Donaire B Shacham J ŠmiacutedS M Baha El Din and J Padial We thank MMetallinou and J Šmiacuted for help with the lab work andthe analyses M Novosolov for help with the figuresY L Werner B Shacham and A Levy for fruitful dis-cussions and two anonymous referees for helpful com-ments on an earlier version of this manuscript Wethank the Israel nature and park authority for issuingcollecting permits (nos 38074 38451 38489 38540)The work of J M was financially supported by the Min-istry of Culture of the Czech Republic (DKRVO 201315 National Museum 00023272) S C is funded bygrant CGL2012-36970 from the Ministerio de Economiacuteay Competitividad Spain (cofunded by FEDER) Thisstudy was funded by a Israel Taxonomic Initiative grantto S M K T is supported by an Israel Taxonomic Ini-tiative scholarship

16 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

REFERENCES

Ahmadzadeh F Carretero MA Harris DJ Perera ABohme W 2012 A molecular phylogeny of the eastern groupof ocellated lizard genus Timon (Sauria Lacertidae) basedon mitochondrial and nuclear DNA sequences Amphibia-Reptilia 33 1ndash10

Ahmadzadeh F Flecks M Roumldder D Boumlhme W Ilgaz CcedilHarris DJ Engler JO Uumlzuumlm N Carretero MA 2013Multiple dispersal out of Anatolia biogeography and evolu-tion of oriental green lizards Biological Journal of the LinneanSociety 110 398ndash408

Akaike H 1973 Information theory and an extension of themaximum likelihood principle In Petrov BN Csaki F edsSecond International Symposium on Information Theory Bu-dapest Akademiai Kiado 267ndash281

Amitai P Bouskila A 2001 Handbook of amphibians ampreptiles of Israel Israel Keter Publishing House Ltd

Anderson SC 1999 The lizards of Iran Ithaca NY Societyfor the Study of Amphibians and Reptiles

Arnold EN 1980 The scientific results of the Oman flora andfauna survey 1977 (Dhofar) The reptiles and amphibians ofDhofar southern Arabia Journal of Oman Studies SpecialReport 2 273ndash332

Arnold EN 1983 Osteology genitalia and the relationshipsof Acanthodactylus (Reptilia Lacertidae) Bulletin of the BritishMuseum (Natural History) Zoology 44 291ndash339

Arnold EN Arribas Oacute Carranza S 2007 Systematics ofthe Palaearctic and Oriental lizard tribe Lacertini (SquamataLacertidae Lacertinae) with descriptions of eight new generaZootaxa 1430 1ndash86

Arnold EN Robinson MD Carranza S 2009 A prelimi-nary analysis of phylogenetic relationships and biogeogra-phy of the dangerously venomous Carpet Vipers Echis(Squamata Serpentes Viperidae) based on mitochondrial DNAsequences Amphibia-Reptilia 30 273ndash282

Audouin JV 1827 Explication sommaire des planches dereptiles (suppleacutement) offrant un exposeacute des characteacuteresdes espegraveces In Savigny MJCL ed Description de lrsquoEacutegypteVol 1 Histoire Naturelle Paris Impeacuteriale 161ndash184

Bache F Popescu SM Rabineau M Gorini C Suc JPClauzon G Olivet JL Rubino JL Melinte-DobrinescuMC Estrada F 2012 A two-step process for the refloodingof the Mediterranean after the Messinian Salinity Crisis BasinResearch 24 125ndash153

Baha El Din SM 2006 A guide to the reptiles and amphib-ians of Egypt CairondashNew York American University in CairoPress

Baier FS Sparrow DJ Wiedl H-J 2009 The amphibiansand reptiles of Cyprus Frankfurt am Main Edition Chimaira

Bandelt H-J Forster P Roumlhl A 1999 Median-joining net-works for inferring intraspecific phylogenies Molecular Biologyand Evolution 16 37ndash48

Bar A Haimovitch G 2011 A field guide to reptiles and am-phibians of Israel Herzliya Pazbar Limited

Boumlttger O 1879 Reptilien und Amphibien aus Syrien Berichtuumlber die Senckenbergische Naturforschende Gesellschaft inFrankfurt am Main 1879 57ndash84

Boulenger GA 1919 On a new variety of Acanthodactylusboskianus Daud from the Euphrates The Annals and Maga-zine of Natural History 3 549ndash550

Boulenger GA 1878 Sur les espegraveces drsquoAcanthodactylus desbords de la Mediterraneacutee Bulletin de la Socieacuteteacute zoologiquede France 3 179ndash197

Boulenger GA 1918 Sur les leacutezards du genre AcanthodactylusWieg Bulletin de la Socieacuteteacute zoologique de France 43 143ndash155

Boulenger GA 1921 Monograph of the Lacertidœ Vol II Trus-tees of the British Museum of Natural History London

Carranza S Arnold EN 2012 A review of the geckos of thegenus Hemidactylus (Squamata Gekkonidae) fromOman based on morphology mitochondrial and nuclear datawith descriptions of eight new species Zootaxa 3378 1ndash95

Carranza S Arnold EN Geniez P Roca J Mateo J 2008Radiation multiple dispersal and parallelism in the skinksChalcides and Sphenops (Squamata Scincidae) with com-ments on Scincus and Scincopus and the age of the SaharaDesert Molecular Phylogenetics and Evolution 46 1071ndash1094

Carretero MA Fonseca MM Garcia-Munoz E Brito JCHarris DJ 2011 Adding Acanthodactylus beershebensis tothe mtDNA phylogeny of the Acanthodactylus pardalis groupNorth-Western Journal of Zoology 7 138ndash142

Castresana J 2000 Selection of conserved blocks from multi-ple alignments for their use in phylogenetic analysis Mo-lecular Biology and Evolution 17 540ndash552

Clube TMM Robertson A 1986 The palaeorotation of theTroodos microplate Cyprus in the Late Mesozoic-Early Ce-nozoic plate tectonic framework of the Eastern Mediterra-nean Surveys in Geophysics 8 375ndash437

Cox SC Carranza S Brown RP 2010 Divergence timesand colonization of the Canary Islands by Gallotia lizardsMolecular Phylogenetics and Evolution 56 747ndash757

Crochet PA Geniez P Ineich I 2003 A multivariate analy-sis of the fringe-toed lizards of the Acanthodactylus scutellatusgroup (Squamata Lacertidae) systematic and biogeographi-cal implications Zoological Journal of the Linnean Society137 117ndash155

Crochet P-A Chaline O Surget-Groba Y Debain CCheylan M 2004 Speciation in mountains phylogeographyand phylogeny of the rock lizards genus Iberolacerta (ReptiliaLacertidae) Molecular Phylogenetics and Evolution 30 860ndash866

Daudin FM 1802 Histoire naturelle geacuteneacuterale et particuliegraveredes reptiles ouvrage faisant suite agrave lrsquoHistoire naturelle geacuteneacuteraleet particuliegravere F Dufart

Disi A Modry D Necas P Rifai L 2001 Amphibians andreptiles of the Hashemite Kingdom of Jordan an atlas andfield guide Frankfurt am Main Chimaira

Dolev A Pervolutzki A 2004 Endangered species in IsraelRed list of threatened animals Vertebrates JerusalemThe Nature and Parks Authority and the Society for thePreservation of Nature

Drummond AJ Rambaut A 2007 BEAST Bayesian evo-lutionary analysis by sampling trees BMC EvolutionaryBiology 7 214

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 17

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Ezard T Fujisawa T Barraclough T 2009 Splits speciesrsquolimits by threshold statistics R package version 10-19r48Available at httpR-ForgeR-projectorgprojectssplits

Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783ndash791

Fitzinger LJFJ 1834 Acanthodactylus In Wiegman AFAed Herpetologia mexicana seu Descriptio amphibiorum NovaeHispaniae quae itineribus comitis De Sack Ferdinandi Deppeet Chr Guil Schiede in Museum zoologicum Berolinensepervenerunt Pars prima saurorum species amplectens adjectosystematis saurorum prodromo additisque multis in huncamphibiorum ordinem observationibus Berlin C G Luderitz10

Flot JF 2010 SeqPHASE a web tool for interconvertingPHASE inputoutput files and FASTA sequence align-ments Molecular Ecology Resources 10 162ndash166

Fonseca MM Brito JC Paulo OS Carretero MA HarrisDJ 2009 Systematic and phylogeographical assessment ofthe Acanthodactylus erythrurus group (Reptilia Lacertidae)based on phylogenetic analyses of mitochondrial and nuclearDNA Molecular Phylogenetics and Evolution 51 131ndash142

Fonseca MM Brito JC Rebelo H Kalboussi M LarbesS Carretero MA Harris DJ 2008 Genetic variation amongspiny-footed lizards in the Acanthodactylus pardalis groupfrom North Africa African Zoology 43 8ndash15

Fontaneto D Herniou EA Boschetti C Caprioli M MeloneG Ricci C Barraclough TG 2007 Independently evolv-ing species in asexual bdelloid rotifers PLoS Biology 5 e87

Franzen M 1998 Erstnachweis von Acanthodactylus schreiberischreiberi Boulenger 1879 fuumlr die Tuumlrkei (Squamata SauriaLacertidae) Herpetozoa 11 27ndash36

Funk DJ Omland KE 2003 Species-level paraphyly andpolyphyly frequency causes and consequences with in-sights from animal mitochondrial DNA Annual Review ofEcology Evolution and Systematics 34 397ndash423

Godinho R Crespo EG Ferrand N Harris DJ 2005 Phy-logeny and evolution of the green lizards Lacerta spp(Squamata Lacertidae) based on mitochondrial and nuclearDNA sequences Amphibia-Reptilia 26 271ndash285

Greenbaum E Villanueva CO Kusamba C Aristote MMBranch WR 2011 A molecular phylogeny of EquatorialAfrican Lacertidae with the description of a new genus andspecies from eastern Democratic Republic of the Congo Zoo-logical Journal of the Linnean Society 163 913ndash942

Guillou H Carracedo JC Torrado FP Badiola ER 1996K-Ar ages and magnetic stratigraphy of a hotspot-inducedfast grown oceanic island El Hierro Canary Islands Journalof Volcanology and Geothermal Research 73 141ndash155

Harris DJ Arnold EN 2000 Elucidation of the relation-ships of spiny-footed lizards Acanthodactylus spp (ReptiliaLacertidae) using mitochondrial DNA sequence with com-ments on their biogeography and evolution Journal of Zoology252 351ndash362

Harris DJ Batista V Carretero M 2004 Assessment ofgenetic diversity within Acanthodactylus erythrurus (ReptiliaLacertidae) in Morocco and the Iberian Peninsula usingmitochondrial DNA sequence data Amphibia-Reptilia 25227

Horowitz A 1979 The Quaternary of Israel New York Aca-demic Press

Hraoui-Bloquet S Sadek RA Sindaco R Venchi A 2002The herpetofauna of Lebanon new data on distributionZoology in the Middle East 27 35ndash46

Hsuuml KJ Montadert L Bernoulli D Cita MB EricksonA Garrison RE Kidd RB Megraveliereacutes F Muumlller C WrightR 1977 History of the Mediterranean salinity crisis Nature267 399ndash403

Huelsenbeck JP Rannala B 2004 Frequentist propertiesof Bayesian posterior probabilities of phylogenetic trees undersimple and complex substitution models Systematic Biology53 904ndash913

Huelsenbeck JP Ronquist F 2001 MRBAYES Bayesianinference of phylogenetic trees Bioinformatics 17 754ndash755

Jolivet L Augier R Robin C Suc J-P Rouchy JM 2006Lithospheric-scale geodynamic context of the Messinian sa-linity crisis Sedimentary Geology 188 9ndash33

Kapli P Lymberakis P Poulakakis N Mantziou GParmakelis A Mylonas M 2008 Molecular phylogeny ofthree Mesalina (Reptilia Lacertidae) species (M guttulataM brevirostris and M bahaeldini) from North Africa and theMiddle East another case of paraphyly MolecularPhylogenetics and Evolution 49 102ndash110

Katoh K Toh H 2008 Recent developments in the MAFFTmultiple sequence alignment program Briefings inBioinformatics 9 286ndash298

Krijgsman W Hilgen F Raffi I Sierro F Wilson D 1999Chronology causes and progression of the Messinian salin-ity crisis Nature 400 652ndash655

Lanfear R Calcott B Ho SY Guindon S 2012PartitionFinder combined selection of partitioning schemesand substitution models for phylogenetic analyses Molecu-lar Biology and Evolution 29 1695ndash1701

Le Houeacuterou HN 1997 Climate flora and fauna changes inthe Sahara over the past 500 million years Journal of AridEnvironments 37 619ndash647

Maca-Meyer N Carranza S Rando J Arnold E CabreraV 2003 Status and relationships of the extinct giant CanaryIsland lizard Gallotia goliath (Reptilia Lacertidae)assessed using ancient mtDNA from its mummifiedremains Biological Journal of the Linnean Society 80 659ndash670

McCallum J Robertson A 1990 Pulsed uplift of the TroodosMassif ndash evidence from the Plio-Pleistocene Mesaoria basinIn J Malpas EMM Panayiotou A Xenophontos C edsOphiolites oceanic crustal analogues Proceedings of theSymposium lsquoTroodos 1987rsquo Nicosia (The Geological SurveyDepartment Ministry of Agriculture and NaturalResources) 217ndash229

Metallinou M Arnold EN Crochet P-A Geniez P BritoJC Lymberakis P El Din SB Sindaco R Robinson MCarranza S 2012 Conquering the Sahara and Arabiandeserts systematics and biogeography of Stenodactylus geckos(Reptilia Gekkonidae) BMC Evolutionary Biology 12258

Monaghan MT Wild R Elliot M Fujisawa T Balke MInward DJ Lees DC Ranaivosolo R Eggleton P

18 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

in Africa This diversification correlates to the arid climatestarting in southern Sahara during the early-mid Plio-cene and later in northern Africa between the Plio-cene and the Pleistocene (Le Houeacuterou 1997) as hasbeen suggested for the dispersal of Mesalina guttulatain Africa (Kapli et al 2008) Other evidence relatesdry climate in North Africa to an earlier period around7 Mya (Schuster et al 2006) as has been suggestedfor the genus Chalcides and other reptiles (Carranzaet al 2008 Metallinou et al 2012 and reference therein)The aridification of North Africa has most likely con-tributed for the successful dispersal of A b asper westfrom south-west Asia into Africa Morphological studiesof A boskianus show relatively uniform populations inNorth Africa suggesting recent migration (Salvador1982 Arnold 1983) The other two geographical group-ings of A b asper from the Middle East and Arabia(groups C1 and C3) are located in two distinct innerclades but their location within each inner clade ispoorly supported The topology of the concatenated tree(Fig 2) shows that the group from northern Jordanand northern Oman (group C1) is closer to the NorthAfrican one than to the geographically close Middle-Eastern and south Arabian group (group C3) The taxo-nomic separation between north and south Oman hasbeen recognized in other species of reptiles and sup-ported by the topography of Oman (eg Echis coloratusand Echis omanensis Arnold Robinson amp Carranza2009) In the nuclear tree (Fig S1) these two groupsare closer to one another and with the North Africangroup form clade C Therefore the low support valuesamongst these groupings prevent an appropriate andthorough analysis of this subspecies The close rela-tionship amongst the geographical groups may reflectclose phylogenetic relationships amongst these popu-lations suggesting recent migration divergence andongoing gene flow

SYSTEMATICS AND TAXONOMIC IMPLICATIONS

The relationships within the A boskianus species groupconflict with the current known taxonomy of A schreiberiand A b asper (samples of the other subspecies ofA boskianus and of A nilsoni were unavailable for thisstudy) Both species have been found to be closelyrelated and paraphyletic The constrained topology testsexemplify the close entangled relationship between thetwo species as the separate monophyly of the two specieswas not rejected and the enforced monophyly of themboth together was inconclusive

Several causes can be responsible for paraphyly inspecies such in the A boskianus species group (Funkamp Omland 2003 and references therein) (1) inad-equate phylogenetic information (2) imperfect taxono-my (incorrectinaccurate species limits) derived frommisidentifying intraspecific variation (3) interspecific

gene flow ndash hybridization through interspecific matingand the subsequent backcrossing of hybrids into theparental populations (4) incomplete lineage sortingbecause of recent speciation events (5) unrecognizedparalogy We suggest that the relationships betweenA schreiberi and A b asper based on mitochondrialand nuclear data are most likely explained by incor-rect taxonomy probably because of the great variabil-ity of the latter species and to convergence As wasthe case in the molecular studies of the A pardalisand A erythrurus species groups (Harris et al 2004Fonseca et al 2008 2009 Carretero et al 2011 andreference therein) there are many problems with thecurrent taxonomic status of several species groups withinAcanthodactylus

Taking the molecular results of our study into accountthere are several systematic approaches to classify-ing the A schreiberiminusA b asper clade The Cypriot andTurkish populations of A schreiberi are very closelyrelated with the latter nested in the former and thetwo subspecies share nuclear alleles (Fig 3) Further-more the low uncorrected p-distance is positively cor-related with subspecies-level distances within otherlacertid species (ie 16 of Cytb in Lacerta bilineatachloronota Godinho et al 2005) We therefore con-clude that Cypriot animals were recently introducedto Turkey and that the Turkish population does notmerit a subspecific rank We suggest that A s ataturiYalccedilinkaya amp Goumlccedilmen 2012 is a junior synonym ofA s schreiberi Boulenger 1878

Regarding the relationships between A schreiberi andA b asper a few scenarios are possible One is to sinkA schreiberi within A boskianus to create one species(A boskianus) with high genetic and morphological vari-ability ranging over a broad distribution Another isfor the two taxa be regarded as a species complex (theA boskianus-schreiberi complex) until further inves-tigation on the subject However although A schreiberiis nested within A b asper the populations from Cyprusand Turkey represent a distinct evolutionary lineagewith distinct genetic and morphological features andthus it is logical to retain the specific status Two othersolutions are possible The first is to re-evaluate theSyrian populations and to consider elevating them aswell as the more divergent lineages (and subspecies)of A boskianus to specific status This would neces-sitate an examination of the phylogeny and morphol-ogy of the other four subspecies of A boskianus(A b boskianus A b euphraticus A b khattensis andA b nigeriensis) and the identification of distinctivephenotypic features in the Syrian lizards Another so-lution is to recognize the maintenance of gene flowamongst mainland populations of A b asper after thedivergence of the insular endemic A schreiberi andthus the evolutionary cohesion of the paraphyleticA b asper Arnold (1983) noted that A schreiberi may

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 15

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

have originated as an isolate from A boskianus becauseof their shared morphology and hemipenis featuresOur results support this scenario which includes thedispersal of A schreiberi to Cyprus from a mainlandpopulation that was most probably A boskianus It maybe assumed that the ancestor of the Cypriot A schreiberiafter arriving on Cyprus remained isolated for a longperiod of time and thus evolved to the modern formof A schreiberi Meanwhile the same ancestral con-tinental populations not isolated from each other con-tinued to exchange genes to varying degrees remainingA boskianus

The IsraeliminusLebanese subspecies A sc syriacus is onlydistantly related to the nominate form A sc schreiberiThis subspecies is highly phylogenetically divergent fromthe Cypriot and Turkish populations having higherp-distances (12S 4 Cytb 11ndash12) than those foundbetween other lacertid species (eg 74ndash82 of Cytbamongst Iberolacerta aranica Iberolacerta aurelioi andIberolacerta bonnali and 41ndash58 of Cytb betweenLacerta bilineata and Lacerta viridis Crochet et al2004 Godinho et al 2005 respectively) The nuclearhaplotype networks further show that Lebanese andIsraeli populations share alleles only with A b asperbut not with the nominotypical Cypriot form Arnold(1983) suggested that the geographical variation ofA boskianus reflects niche differences with animalsfrom xeric areas with dense rigid and spiny vegeta-tion having larger dorsal scales than animals from moremesic areas As was assumed for A schreiberi wesuggest that other mainland populations of A b asperwere the ancestors of the LebaneseminusIsraeli Coastal plainforms We suggest that A s syriacus is an ecomorphof A b asper that dispersed from the usual xerichabitats of the species and adapted to the new moremesic environment of the stable sands of the coastalplains of the eastern Mediterranean As a conse-quence this ecomorph converged on the morphologyof A s schreiberi which inhabits the coastal sands ofCyprus (Baier et al 2009) but still maintains differ-entiating features by having coarser dorsal scales andsharp keels (Salvador 1982 Arnold 1983 Franzen1998) This convergence led to the description ofA s syriacus as a member of A schreiberi The mor-phological assessment and the close morphological simi-larities between A b asper and A s syriacus mayexplain the wrong classification A similar erro-neous reasoning led Reed amp Marx (1959) to identifyspecimens with fine scales from Iraq as A schreiberiSalvador (1982) re-examined these specimens and as-signed them to A boskianus The morphological dif-ferences between the two forms are less prominentespecially where the two forms occur in close geo-graphical proximity in the southern coastal plainand north-western Negev Desert of Israel (Bar ampHaimovitch 2011) According to our results A s syriacus

actually belongs to A b asper being a coastal-duneecomorph convergent with but evolutionarily dis-tinct from A schreiberi Thus our preferred scenariois to treat the name Acanthodactylus schreiberi syriacusBoumlttger 1879 (which was originally described asA boskianus var syriacus by Boumlttger 1879) as a juniorsynonym of the name Acanthodactylus boskianus asper(Audouin 1827)

Recognizing A s syriacus as a junior synonym ofA b asper may have important implications for the con-servation of this coastal sand dune form which is clas-sified as critically endangered in Israel (Dolev ampPervolutzki 2004) However as the Israeli and Leba-nese coastal dune ecosystem has probably developedonly very recently during the Quaternary (Nir 1970Horowitz 1979) this form represents a remarkable caseof rapid evolutionary change It is also a remarkablecase of convergent evolution (with the CypriotA sc schreiberi) Thus we feel that these popula-tions are unique evolutionary entities that merit specialconservation efforts

The use of nuclear genes is a valuable method forestimating species divergence and lineage sorting andhelps evaluate isolated lineages and evolutionary historyThe incorporation of mitochondrial and nuclear dataprovides thorough topologies informative networks anddivergence times that reveal useful information for aproblematic taxonomy such as that of the A boskianusspecies group We have shown that phylogenetic ap-proaches to the confusing taxonomy of two closely relatedand morphologically similar species can shed light ontheir unclear relationships resolve between homoplasyand shared ancestry and identify patterns of speciesevolutionary history and biogeography

ACKNOWLEDGEMENTS

We are grateful to the following people for providingsamples for this study D Donaire B Shacham J ŠmiacutedS M Baha El Din and J Padial We thank MMetallinou and J Šmiacuted for help with the lab work andthe analyses M Novosolov for help with the figuresY L Werner B Shacham and A Levy for fruitful dis-cussions and two anonymous referees for helpful com-ments on an earlier version of this manuscript Wethank the Israel nature and park authority for issuingcollecting permits (nos 38074 38451 38489 38540)The work of J M was financially supported by the Min-istry of Culture of the Czech Republic (DKRVO 201315 National Museum 00023272) S C is funded bygrant CGL2012-36970 from the Ministerio de Economiacuteay Competitividad Spain (cofunded by FEDER) Thisstudy was funded by a Israel Taxonomic Initiative grantto S M K T is supported by an Israel Taxonomic Ini-tiative scholarship

16 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

REFERENCES

Ahmadzadeh F Carretero MA Harris DJ Perera ABohme W 2012 A molecular phylogeny of the eastern groupof ocellated lizard genus Timon (Sauria Lacertidae) basedon mitochondrial and nuclear DNA sequences Amphibia-Reptilia 33 1ndash10

Ahmadzadeh F Flecks M Roumldder D Boumlhme W Ilgaz CcedilHarris DJ Engler JO Uumlzuumlm N Carretero MA 2013Multiple dispersal out of Anatolia biogeography and evolu-tion of oriental green lizards Biological Journal of the LinneanSociety 110 398ndash408

Akaike H 1973 Information theory and an extension of themaximum likelihood principle In Petrov BN Csaki F edsSecond International Symposium on Information Theory Bu-dapest Akademiai Kiado 267ndash281

Amitai P Bouskila A 2001 Handbook of amphibians ampreptiles of Israel Israel Keter Publishing House Ltd

Anderson SC 1999 The lizards of Iran Ithaca NY Societyfor the Study of Amphibians and Reptiles

Arnold EN 1980 The scientific results of the Oman flora andfauna survey 1977 (Dhofar) The reptiles and amphibians ofDhofar southern Arabia Journal of Oman Studies SpecialReport 2 273ndash332

Arnold EN 1983 Osteology genitalia and the relationshipsof Acanthodactylus (Reptilia Lacertidae) Bulletin of the BritishMuseum (Natural History) Zoology 44 291ndash339

Arnold EN Arribas Oacute Carranza S 2007 Systematics ofthe Palaearctic and Oriental lizard tribe Lacertini (SquamataLacertidae Lacertinae) with descriptions of eight new generaZootaxa 1430 1ndash86

Arnold EN Robinson MD Carranza S 2009 A prelimi-nary analysis of phylogenetic relationships and biogeogra-phy of the dangerously venomous Carpet Vipers Echis(Squamata Serpentes Viperidae) based on mitochondrial DNAsequences Amphibia-Reptilia 30 273ndash282

Audouin JV 1827 Explication sommaire des planches dereptiles (suppleacutement) offrant un exposeacute des characteacuteresdes espegraveces In Savigny MJCL ed Description de lrsquoEacutegypteVol 1 Histoire Naturelle Paris Impeacuteriale 161ndash184

Bache F Popescu SM Rabineau M Gorini C Suc JPClauzon G Olivet JL Rubino JL Melinte-DobrinescuMC Estrada F 2012 A two-step process for the refloodingof the Mediterranean after the Messinian Salinity Crisis BasinResearch 24 125ndash153

Baha El Din SM 2006 A guide to the reptiles and amphib-ians of Egypt CairondashNew York American University in CairoPress

Baier FS Sparrow DJ Wiedl H-J 2009 The amphibiansand reptiles of Cyprus Frankfurt am Main Edition Chimaira

Bandelt H-J Forster P Roumlhl A 1999 Median-joining net-works for inferring intraspecific phylogenies Molecular Biologyand Evolution 16 37ndash48

Bar A Haimovitch G 2011 A field guide to reptiles and am-phibians of Israel Herzliya Pazbar Limited

Boumlttger O 1879 Reptilien und Amphibien aus Syrien Berichtuumlber die Senckenbergische Naturforschende Gesellschaft inFrankfurt am Main 1879 57ndash84

Boulenger GA 1919 On a new variety of Acanthodactylusboskianus Daud from the Euphrates The Annals and Maga-zine of Natural History 3 549ndash550

Boulenger GA 1878 Sur les espegraveces drsquoAcanthodactylus desbords de la Mediterraneacutee Bulletin de la Socieacuteteacute zoologiquede France 3 179ndash197

Boulenger GA 1918 Sur les leacutezards du genre AcanthodactylusWieg Bulletin de la Socieacuteteacute zoologique de France 43 143ndash155

Boulenger GA 1921 Monograph of the Lacertidœ Vol II Trus-tees of the British Museum of Natural History London

Carranza S Arnold EN 2012 A review of the geckos of thegenus Hemidactylus (Squamata Gekkonidae) fromOman based on morphology mitochondrial and nuclear datawith descriptions of eight new species Zootaxa 3378 1ndash95

Carranza S Arnold EN Geniez P Roca J Mateo J 2008Radiation multiple dispersal and parallelism in the skinksChalcides and Sphenops (Squamata Scincidae) with com-ments on Scincus and Scincopus and the age of the SaharaDesert Molecular Phylogenetics and Evolution 46 1071ndash1094

Carretero MA Fonseca MM Garcia-Munoz E Brito JCHarris DJ 2011 Adding Acanthodactylus beershebensis tothe mtDNA phylogeny of the Acanthodactylus pardalis groupNorth-Western Journal of Zoology 7 138ndash142

Castresana J 2000 Selection of conserved blocks from multi-ple alignments for their use in phylogenetic analysis Mo-lecular Biology and Evolution 17 540ndash552

Clube TMM Robertson A 1986 The palaeorotation of theTroodos microplate Cyprus in the Late Mesozoic-Early Ce-nozoic plate tectonic framework of the Eastern Mediterra-nean Surveys in Geophysics 8 375ndash437

Cox SC Carranza S Brown RP 2010 Divergence timesand colonization of the Canary Islands by Gallotia lizardsMolecular Phylogenetics and Evolution 56 747ndash757

Crochet PA Geniez P Ineich I 2003 A multivariate analy-sis of the fringe-toed lizards of the Acanthodactylus scutellatusgroup (Squamata Lacertidae) systematic and biogeographi-cal implications Zoological Journal of the Linnean Society137 117ndash155

Crochet P-A Chaline O Surget-Groba Y Debain CCheylan M 2004 Speciation in mountains phylogeographyand phylogeny of the rock lizards genus Iberolacerta (ReptiliaLacertidae) Molecular Phylogenetics and Evolution 30 860ndash866

Daudin FM 1802 Histoire naturelle geacuteneacuterale et particuliegraveredes reptiles ouvrage faisant suite agrave lrsquoHistoire naturelle geacuteneacuteraleet particuliegravere F Dufart

Disi A Modry D Necas P Rifai L 2001 Amphibians andreptiles of the Hashemite Kingdom of Jordan an atlas andfield guide Frankfurt am Main Chimaira

Dolev A Pervolutzki A 2004 Endangered species in IsraelRed list of threatened animals Vertebrates JerusalemThe Nature and Parks Authority and the Society for thePreservation of Nature

Drummond AJ Rambaut A 2007 BEAST Bayesian evo-lutionary analysis by sampling trees BMC EvolutionaryBiology 7 214

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 17

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Ezard T Fujisawa T Barraclough T 2009 Splits speciesrsquolimits by threshold statistics R package version 10-19r48Available at httpR-ForgeR-projectorgprojectssplits

Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783ndash791

Fitzinger LJFJ 1834 Acanthodactylus In Wiegman AFAed Herpetologia mexicana seu Descriptio amphibiorum NovaeHispaniae quae itineribus comitis De Sack Ferdinandi Deppeet Chr Guil Schiede in Museum zoologicum Berolinensepervenerunt Pars prima saurorum species amplectens adjectosystematis saurorum prodromo additisque multis in huncamphibiorum ordinem observationibus Berlin C G Luderitz10

Flot JF 2010 SeqPHASE a web tool for interconvertingPHASE inputoutput files and FASTA sequence align-ments Molecular Ecology Resources 10 162ndash166

Fonseca MM Brito JC Paulo OS Carretero MA HarrisDJ 2009 Systematic and phylogeographical assessment ofthe Acanthodactylus erythrurus group (Reptilia Lacertidae)based on phylogenetic analyses of mitochondrial and nuclearDNA Molecular Phylogenetics and Evolution 51 131ndash142

Fonseca MM Brito JC Rebelo H Kalboussi M LarbesS Carretero MA Harris DJ 2008 Genetic variation amongspiny-footed lizards in the Acanthodactylus pardalis groupfrom North Africa African Zoology 43 8ndash15

Fontaneto D Herniou EA Boschetti C Caprioli M MeloneG Ricci C Barraclough TG 2007 Independently evolv-ing species in asexual bdelloid rotifers PLoS Biology 5 e87

Franzen M 1998 Erstnachweis von Acanthodactylus schreiberischreiberi Boulenger 1879 fuumlr die Tuumlrkei (Squamata SauriaLacertidae) Herpetozoa 11 27ndash36

Funk DJ Omland KE 2003 Species-level paraphyly andpolyphyly frequency causes and consequences with in-sights from animal mitochondrial DNA Annual Review ofEcology Evolution and Systematics 34 397ndash423

Godinho R Crespo EG Ferrand N Harris DJ 2005 Phy-logeny and evolution of the green lizards Lacerta spp(Squamata Lacertidae) based on mitochondrial and nuclearDNA sequences Amphibia-Reptilia 26 271ndash285

Greenbaum E Villanueva CO Kusamba C Aristote MMBranch WR 2011 A molecular phylogeny of EquatorialAfrican Lacertidae with the description of a new genus andspecies from eastern Democratic Republic of the Congo Zoo-logical Journal of the Linnean Society 163 913ndash942

Guillou H Carracedo JC Torrado FP Badiola ER 1996K-Ar ages and magnetic stratigraphy of a hotspot-inducedfast grown oceanic island El Hierro Canary Islands Journalof Volcanology and Geothermal Research 73 141ndash155

Harris DJ Arnold EN 2000 Elucidation of the relation-ships of spiny-footed lizards Acanthodactylus spp (ReptiliaLacertidae) using mitochondrial DNA sequence with com-ments on their biogeography and evolution Journal of Zoology252 351ndash362

Harris DJ Batista V Carretero M 2004 Assessment ofgenetic diversity within Acanthodactylus erythrurus (ReptiliaLacertidae) in Morocco and the Iberian Peninsula usingmitochondrial DNA sequence data Amphibia-Reptilia 25227

Horowitz A 1979 The Quaternary of Israel New York Aca-demic Press

Hraoui-Bloquet S Sadek RA Sindaco R Venchi A 2002The herpetofauna of Lebanon new data on distributionZoology in the Middle East 27 35ndash46

Hsuuml KJ Montadert L Bernoulli D Cita MB EricksonA Garrison RE Kidd RB Megraveliereacutes F Muumlller C WrightR 1977 History of the Mediterranean salinity crisis Nature267 399ndash403

Huelsenbeck JP Rannala B 2004 Frequentist propertiesof Bayesian posterior probabilities of phylogenetic trees undersimple and complex substitution models Systematic Biology53 904ndash913

Huelsenbeck JP Ronquist F 2001 MRBAYES Bayesianinference of phylogenetic trees Bioinformatics 17 754ndash755

Jolivet L Augier R Robin C Suc J-P Rouchy JM 2006Lithospheric-scale geodynamic context of the Messinian sa-linity crisis Sedimentary Geology 188 9ndash33

Kapli P Lymberakis P Poulakakis N Mantziou GParmakelis A Mylonas M 2008 Molecular phylogeny ofthree Mesalina (Reptilia Lacertidae) species (M guttulataM brevirostris and M bahaeldini) from North Africa and theMiddle East another case of paraphyly MolecularPhylogenetics and Evolution 49 102ndash110

Katoh K Toh H 2008 Recent developments in the MAFFTmultiple sequence alignment program Briefings inBioinformatics 9 286ndash298

Krijgsman W Hilgen F Raffi I Sierro F Wilson D 1999Chronology causes and progression of the Messinian salin-ity crisis Nature 400 652ndash655

Lanfear R Calcott B Ho SY Guindon S 2012PartitionFinder combined selection of partitioning schemesand substitution models for phylogenetic analyses Molecu-lar Biology and Evolution 29 1695ndash1701

Le Houeacuterou HN 1997 Climate flora and fauna changes inthe Sahara over the past 500 million years Journal of AridEnvironments 37 619ndash647

Maca-Meyer N Carranza S Rando J Arnold E CabreraV 2003 Status and relationships of the extinct giant CanaryIsland lizard Gallotia goliath (Reptilia Lacertidae)assessed using ancient mtDNA from its mummifiedremains Biological Journal of the Linnean Society 80 659ndash670

McCallum J Robertson A 1990 Pulsed uplift of the TroodosMassif ndash evidence from the Plio-Pleistocene Mesaoria basinIn J Malpas EMM Panayiotou A Xenophontos C edsOphiolites oceanic crustal analogues Proceedings of theSymposium lsquoTroodos 1987rsquo Nicosia (The Geological SurveyDepartment Ministry of Agriculture and NaturalResources) 217ndash229

Metallinou M Arnold EN Crochet P-A Geniez P BritoJC Lymberakis P El Din SB Sindaco R Robinson MCarranza S 2012 Conquering the Sahara and Arabiandeserts systematics and biogeography of Stenodactylus geckos(Reptilia Gekkonidae) BMC Evolutionary Biology 12258

Monaghan MT Wild R Elliot M Fujisawa T Balke MInward DJ Lees DC Ranaivosolo R Eggleton P

18 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

have originated as an isolate from A boskianus becauseof their shared morphology and hemipenis featuresOur results support this scenario which includes thedispersal of A schreiberi to Cyprus from a mainlandpopulation that was most probably A boskianus It maybe assumed that the ancestor of the Cypriot A schreiberiafter arriving on Cyprus remained isolated for a longperiod of time and thus evolved to the modern formof A schreiberi Meanwhile the same ancestral con-tinental populations not isolated from each other con-tinued to exchange genes to varying degrees remainingA boskianus

The IsraeliminusLebanese subspecies A sc syriacus is onlydistantly related to the nominate form A sc schreiberiThis subspecies is highly phylogenetically divergent fromthe Cypriot and Turkish populations having higherp-distances (12S 4 Cytb 11ndash12) than those foundbetween other lacertid species (eg 74ndash82 of Cytbamongst Iberolacerta aranica Iberolacerta aurelioi andIberolacerta bonnali and 41ndash58 of Cytb betweenLacerta bilineata and Lacerta viridis Crochet et al2004 Godinho et al 2005 respectively) The nuclearhaplotype networks further show that Lebanese andIsraeli populations share alleles only with A b asperbut not with the nominotypical Cypriot form Arnold(1983) suggested that the geographical variation ofA boskianus reflects niche differences with animalsfrom xeric areas with dense rigid and spiny vegeta-tion having larger dorsal scales than animals from moremesic areas As was assumed for A schreiberi wesuggest that other mainland populations of A b asperwere the ancestors of the LebaneseminusIsraeli Coastal plainforms We suggest that A s syriacus is an ecomorphof A b asper that dispersed from the usual xerichabitats of the species and adapted to the new moremesic environment of the stable sands of the coastalplains of the eastern Mediterranean As a conse-quence this ecomorph converged on the morphologyof A s schreiberi which inhabits the coastal sands ofCyprus (Baier et al 2009) but still maintains differ-entiating features by having coarser dorsal scales andsharp keels (Salvador 1982 Arnold 1983 Franzen1998) This convergence led to the description ofA s syriacus as a member of A schreiberi The mor-phological assessment and the close morphological simi-larities between A b asper and A s syriacus mayexplain the wrong classification A similar erro-neous reasoning led Reed amp Marx (1959) to identifyspecimens with fine scales from Iraq as A schreiberiSalvador (1982) re-examined these specimens and as-signed them to A boskianus The morphological dif-ferences between the two forms are less prominentespecially where the two forms occur in close geo-graphical proximity in the southern coastal plainand north-western Negev Desert of Israel (Bar ampHaimovitch 2011) According to our results A s syriacus

actually belongs to A b asper being a coastal-duneecomorph convergent with but evolutionarily dis-tinct from A schreiberi Thus our preferred scenariois to treat the name Acanthodactylus schreiberi syriacusBoumlttger 1879 (which was originally described asA boskianus var syriacus by Boumlttger 1879) as a juniorsynonym of the name Acanthodactylus boskianus asper(Audouin 1827)

Recognizing A s syriacus as a junior synonym ofA b asper may have important implications for the con-servation of this coastal sand dune form which is clas-sified as critically endangered in Israel (Dolev ampPervolutzki 2004) However as the Israeli and Leba-nese coastal dune ecosystem has probably developedonly very recently during the Quaternary (Nir 1970Horowitz 1979) this form represents a remarkable caseof rapid evolutionary change It is also a remarkablecase of convergent evolution (with the CypriotA sc schreiberi) Thus we feel that these popula-tions are unique evolutionary entities that merit specialconservation efforts

The use of nuclear genes is a valuable method forestimating species divergence and lineage sorting andhelps evaluate isolated lineages and evolutionary historyThe incorporation of mitochondrial and nuclear dataprovides thorough topologies informative networks anddivergence times that reveal useful information for aproblematic taxonomy such as that of the A boskianusspecies group We have shown that phylogenetic ap-proaches to the confusing taxonomy of two closely relatedand morphologically similar species can shed light ontheir unclear relationships resolve between homoplasyand shared ancestry and identify patterns of speciesevolutionary history and biogeography

ACKNOWLEDGEMENTS

We are grateful to the following people for providingsamples for this study D Donaire B Shacham J ŠmiacutedS M Baha El Din and J Padial We thank MMetallinou and J Šmiacuted for help with the lab work andthe analyses M Novosolov for help with the figuresY L Werner B Shacham and A Levy for fruitful dis-cussions and two anonymous referees for helpful com-ments on an earlier version of this manuscript Wethank the Israel nature and park authority for issuingcollecting permits (nos 38074 38451 38489 38540)The work of J M was financially supported by the Min-istry of Culture of the Czech Republic (DKRVO 201315 National Museum 00023272) S C is funded bygrant CGL2012-36970 from the Ministerio de Economiacuteay Competitividad Spain (cofunded by FEDER) Thisstudy was funded by a Israel Taxonomic Initiative grantto S M K T is supported by an Israel Taxonomic Ini-tiative scholarship

16 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

REFERENCES

Ahmadzadeh F Carretero MA Harris DJ Perera ABohme W 2012 A molecular phylogeny of the eastern groupof ocellated lizard genus Timon (Sauria Lacertidae) basedon mitochondrial and nuclear DNA sequences Amphibia-Reptilia 33 1ndash10

Ahmadzadeh F Flecks M Roumldder D Boumlhme W Ilgaz CcedilHarris DJ Engler JO Uumlzuumlm N Carretero MA 2013Multiple dispersal out of Anatolia biogeography and evolu-tion of oriental green lizards Biological Journal of the LinneanSociety 110 398ndash408

Akaike H 1973 Information theory and an extension of themaximum likelihood principle In Petrov BN Csaki F edsSecond International Symposium on Information Theory Bu-dapest Akademiai Kiado 267ndash281

Amitai P Bouskila A 2001 Handbook of amphibians ampreptiles of Israel Israel Keter Publishing House Ltd

Anderson SC 1999 The lizards of Iran Ithaca NY Societyfor the Study of Amphibians and Reptiles

Arnold EN 1980 The scientific results of the Oman flora andfauna survey 1977 (Dhofar) The reptiles and amphibians ofDhofar southern Arabia Journal of Oman Studies SpecialReport 2 273ndash332

Arnold EN 1983 Osteology genitalia and the relationshipsof Acanthodactylus (Reptilia Lacertidae) Bulletin of the BritishMuseum (Natural History) Zoology 44 291ndash339

Arnold EN Arribas Oacute Carranza S 2007 Systematics ofthe Palaearctic and Oriental lizard tribe Lacertini (SquamataLacertidae Lacertinae) with descriptions of eight new generaZootaxa 1430 1ndash86

Arnold EN Robinson MD Carranza S 2009 A prelimi-nary analysis of phylogenetic relationships and biogeogra-phy of the dangerously venomous Carpet Vipers Echis(Squamata Serpentes Viperidae) based on mitochondrial DNAsequences Amphibia-Reptilia 30 273ndash282

Audouin JV 1827 Explication sommaire des planches dereptiles (suppleacutement) offrant un exposeacute des characteacuteresdes espegraveces In Savigny MJCL ed Description de lrsquoEacutegypteVol 1 Histoire Naturelle Paris Impeacuteriale 161ndash184

Bache F Popescu SM Rabineau M Gorini C Suc JPClauzon G Olivet JL Rubino JL Melinte-DobrinescuMC Estrada F 2012 A two-step process for the refloodingof the Mediterranean after the Messinian Salinity Crisis BasinResearch 24 125ndash153

Baha El Din SM 2006 A guide to the reptiles and amphib-ians of Egypt CairondashNew York American University in CairoPress

Baier FS Sparrow DJ Wiedl H-J 2009 The amphibiansand reptiles of Cyprus Frankfurt am Main Edition Chimaira

Bandelt H-J Forster P Roumlhl A 1999 Median-joining net-works for inferring intraspecific phylogenies Molecular Biologyand Evolution 16 37ndash48

Bar A Haimovitch G 2011 A field guide to reptiles and am-phibians of Israel Herzliya Pazbar Limited

Boumlttger O 1879 Reptilien und Amphibien aus Syrien Berichtuumlber die Senckenbergische Naturforschende Gesellschaft inFrankfurt am Main 1879 57ndash84

Boulenger GA 1919 On a new variety of Acanthodactylusboskianus Daud from the Euphrates The Annals and Maga-zine of Natural History 3 549ndash550

Boulenger GA 1878 Sur les espegraveces drsquoAcanthodactylus desbords de la Mediterraneacutee Bulletin de la Socieacuteteacute zoologiquede France 3 179ndash197

Boulenger GA 1918 Sur les leacutezards du genre AcanthodactylusWieg Bulletin de la Socieacuteteacute zoologique de France 43 143ndash155

Boulenger GA 1921 Monograph of the Lacertidœ Vol II Trus-tees of the British Museum of Natural History London

Carranza S Arnold EN 2012 A review of the geckos of thegenus Hemidactylus (Squamata Gekkonidae) fromOman based on morphology mitochondrial and nuclear datawith descriptions of eight new species Zootaxa 3378 1ndash95

Carranza S Arnold EN Geniez P Roca J Mateo J 2008Radiation multiple dispersal and parallelism in the skinksChalcides and Sphenops (Squamata Scincidae) with com-ments on Scincus and Scincopus and the age of the SaharaDesert Molecular Phylogenetics and Evolution 46 1071ndash1094

Carretero MA Fonseca MM Garcia-Munoz E Brito JCHarris DJ 2011 Adding Acanthodactylus beershebensis tothe mtDNA phylogeny of the Acanthodactylus pardalis groupNorth-Western Journal of Zoology 7 138ndash142

Castresana J 2000 Selection of conserved blocks from multi-ple alignments for their use in phylogenetic analysis Mo-lecular Biology and Evolution 17 540ndash552

Clube TMM Robertson A 1986 The palaeorotation of theTroodos microplate Cyprus in the Late Mesozoic-Early Ce-nozoic plate tectonic framework of the Eastern Mediterra-nean Surveys in Geophysics 8 375ndash437

Cox SC Carranza S Brown RP 2010 Divergence timesand colonization of the Canary Islands by Gallotia lizardsMolecular Phylogenetics and Evolution 56 747ndash757

Crochet PA Geniez P Ineich I 2003 A multivariate analy-sis of the fringe-toed lizards of the Acanthodactylus scutellatusgroup (Squamata Lacertidae) systematic and biogeographi-cal implications Zoological Journal of the Linnean Society137 117ndash155

Crochet P-A Chaline O Surget-Groba Y Debain CCheylan M 2004 Speciation in mountains phylogeographyand phylogeny of the rock lizards genus Iberolacerta (ReptiliaLacertidae) Molecular Phylogenetics and Evolution 30 860ndash866

Daudin FM 1802 Histoire naturelle geacuteneacuterale et particuliegraveredes reptiles ouvrage faisant suite agrave lrsquoHistoire naturelle geacuteneacuteraleet particuliegravere F Dufart

Disi A Modry D Necas P Rifai L 2001 Amphibians andreptiles of the Hashemite Kingdom of Jordan an atlas andfield guide Frankfurt am Main Chimaira

Dolev A Pervolutzki A 2004 Endangered species in IsraelRed list of threatened animals Vertebrates JerusalemThe Nature and Parks Authority and the Society for thePreservation of Nature

Drummond AJ Rambaut A 2007 BEAST Bayesian evo-lutionary analysis by sampling trees BMC EvolutionaryBiology 7 214

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 17

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Ezard T Fujisawa T Barraclough T 2009 Splits speciesrsquolimits by threshold statistics R package version 10-19r48Available at httpR-ForgeR-projectorgprojectssplits

Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783ndash791

Fitzinger LJFJ 1834 Acanthodactylus In Wiegman AFAed Herpetologia mexicana seu Descriptio amphibiorum NovaeHispaniae quae itineribus comitis De Sack Ferdinandi Deppeet Chr Guil Schiede in Museum zoologicum Berolinensepervenerunt Pars prima saurorum species amplectens adjectosystematis saurorum prodromo additisque multis in huncamphibiorum ordinem observationibus Berlin C G Luderitz10

Flot JF 2010 SeqPHASE a web tool for interconvertingPHASE inputoutput files and FASTA sequence align-ments Molecular Ecology Resources 10 162ndash166

Fonseca MM Brito JC Paulo OS Carretero MA HarrisDJ 2009 Systematic and phylogeographical assessment ofthe Acanthodactylus erythrurus group (Reptilia Lacertidae)based on phylogenetic analyses of mitochondrial and nuclearDNA Molecular Phylogenetics and Evolution 51 131ndash142

Fonseca MM Brito JC Rebelo H Kalboussi M LarbesS Carretero MA Harris DJ 2008 Genetic variation amongspiny-footed lizards in the Acanthodactylus pardalis groupfrom North Africa African Zoology 43 8ndash15

Fontaneto D Herniou EA Boschetti C Caprioli M MeloneG Ricci C Barraclough TG 2007 Independently evolv-ing species in asexual bdelloid rotifers PLoS Biology 5 e87

Franzen M 1998 Erstnachweis von Acanthodactylus schreiberischreiberi Boulenger 1879 fuumlr die Tuumlrkei (Squamata SauriaLacertidae) Herpetozoa 11 27ndash36

Funk DJ Omland KE 2003 Species-level paraphyly andpolyphyly frequency causes and consequences with in-sights from animal mitochondrial DNA Annual Review ofEcology Evolution and Systematics 34 397ndash423

Godinho R Crespo EG Ferrand N Harris DJ 2005 Phy-logeny and evolution of the green lizards Lacerta spp(Squamata Lacertidae) based on mitochondrial and nuclearDNA sequences Amphibia-Reptilia 26 271ndash285

Greenbaum E Villanueva CO Kusamba C Aristote MMBranch WR 2011 A molecular phylogeny of EquatorialAfrican Lacertidae with the description of a new genus andspecies from eastern Democratic Republic of the Congo Zoo-logical Journal of the Linnean Society 163 913ndash942

Guillou H Carracedo JC Torrado FP Badiola ER 1996K-Ar ages and magnetic stratigraphy of a hotspot-inducedfast grown oceanic island El Hierro Canary Islands Journalof Volcanology and Geothermal Research 73 141ndash155

Harris DJ Arnold EN 2000 Elucidation of the relation-ships of spiny-footed lizards Acanthodactylus spp (ReptiliaLacertidae) using mitochondrial DNA sequence with com-ments on their biogeography and evolution Journal of Zoology252 351ndash362

Harris DJ Batista V Carretero M 2004 Assessment ofgenetic diversity within Acanthodactylus erythrurus (ReptiliaLacertidae) in Morocco and the Iberian Peninsula usingmitochondrial DNA sequence data Amphibia-Reptilia 25227

Horowitz A 1979 The Quaternary of Israel New York Aca-demic Press

Hraoui-Bloquet S Sadek RA Sindaco R Venchi A 2002The herpetofauna of Lebanon new data on distributionZoology in the Middle East 27 35ndash46

Hsuuml KJ Montadert L Bernoulli D Cita MB EricksonA Garrison RE Kidd RB Megraveliereacutes F Muumlller C WrightR 1977 History of the Mediterranean salinity crisis Nature267 399ndash403

Huelsenbeck JP Rannala B 2004 Frequentist propertiesof Bayesian posterior probabilities of phylogenetic trees undersimple and complex substitution models Systematic Biology53 904ndash913

Huelsenbeck JP Ronquist F 2001 MRBAYES Bayesianinference of phylogenetic trees Bioinformatics 17 754ndash755

Jolivet L Augier R Robin C Suc J-P Rouchy JM 2006Lithospheric-scale geodynamic context of the Messinian sa-linity crisis Sedimentary Geology 188 9ndash33

Kapli P Lymberakis P Poulakakis N Mantziou GParmakelis A Mylonas M 2008 Molecular phylogeny ofthree Mesalina (Reptilia Lacertidae) species (M guttulataM brevirostris and M bahaeldini) from North Africa and theMiddle East another case of paraphyly MolecularPhylogenetics and Evolution 49 102ndash110

Katoh K Toh H 2008 Recent developments in the MAFFTmultiple sequence alignment program Briefings inBioinformatics 9 286ndash298

Krijgsman W Hilgen F Raffi I Sierro F Wilson D 1999Chronology causes and progression of the Messinian salin-ity crisis Nature 400 652ndash655

Lanfear R Calcott B Ho SY Guindon S 2012PartitionFinder combined selection of partitioning schemesand substitution models for phylogenetic analyses Molecu-lar Biology and Evolution 29 1695ndash1701

Le Houeacuterou HN 1997 Climate flora and fauna changes inthe Sahara over the past 500 million years Journal of AridEnvironments 37 619ndash647

Maca-Meyer N Carranza S Rando J Arnold E CabreraV 2003 Status and relationships of the extinct giant CanaryIsland lizard Gallotia goliath (Reptilia Lacertidae)assessed using ancient mtDNA from its mummifiedremains Biological Journal of the Linnean Society 80 659ndash670

McCallum J Robertson A 1990 Pulsed uplift of the TroodosMassif ndash evidence from the Plio-Pleistocene Mesaoria basinIn J Malpas EMM Panayiotou A Xenophontos C edsOphiolites oceanic crustal analogues Proceedings of theSymposium lsquoTroodos 1987rsquo Nicosia (The Geological SurveyDepartment Ministry of Agriculture and NaturalResources) 217ndash229

Metallinou M Arnold EN Crochet P-A Geniez P BritoJC Lymberakis P El Din SB Sindaco R Robinson MCarranza S 2012 Conquering the Sahara and Arabiandeserts systematics and biogeography of Stenodactylus geckos(Reptilia Gekkonidae) BMC Evolutionary Biology 12258

Monaghan MT Wild R Elliot M Fujisawa T Balke MInward DJ Lees DC Ranaivosolo R Eggleton P

18 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

REFERENCES

Ahmadzadeh F Carretero MA Harris DJ Perera ABohme W 2012 A molecular phylogeny of the eastern groupof ocellated lizard genus Timon (Sauria Lacertidae) basedon mitochondrial and nuclear DNA sequences Amphibia-Reptilia 33 1ndash10

Ahmadzadeh F Flecks M Roumldder D Boumlhme W Ilgaz CcedilHarris DJ Engler JO Uumlzuumlm N Carretero MA 2013Multiple dispersal out of Anatolia biogeography and evolu-tion of oriental green lizards Biological Journal of the LinneanSociety 110 398ndash408

Akaike H 1973 Information theory and an extension of themaximum likelihood principle In Petrov BN Csaki F edsSecond International Symposium on Information Theory Bu-dapest Akademiai Kiado 267ndash281

Amitai P Bouskila A 2001 Handbook of amphibians ampreptiles of Israel Israel Keter Publishing House Ltd

Anderson SC 1999 The lizards of Iran Ithaca NY Societyfor the Study of Amphibians and Reptiles

Arnold EN 1980 The scientific results of the Oman flora andfauna survey 1977 (Dhofar) The reptiles and amphibians ofDhofar southern Arabia Journal of Oman Studies SpecialReport 2 273ndash332

Arnold EN 1983 Osteology genitalia and the relationshipsof Acanthodactylus (Reptilia Lacertidae) Bulletin of the BritishMuseum (Natural History) Zoology 44 291ndash339

Arnold EN Arribas Oacute Carranza S 2007 Systematics ofthe Palaearctic and Oriental lizard tribe Lacertini (SquamataLacertidae Lacertinae) with descriptions of eight new generaZootaxa 1430 1ndash86

Arnold EN Robinson MD Carranza S 2009 A prelimi-nary analysis of phylogenetic relationships and biogeogra-phy of the dangerously venomous Carpet Vipers Echis(Squamata Serpentes Viperidae) based on mitochondrial DNAsequences Amphibia-Reptilia 30 273ndash282

Audouin JV 1827 Explication sommaire des planches dereptiles (suppleacutement) offrant un exposeacute des characteacuteresdes espegraveces In Savigny MJCL ed Description de lrsquoEacutegypteVol 1 Histoire Naturelle Paris Impeacuteriale 161ndash184

Bache F Popescu SM Rabineau M Gorini C Suc JPClauzon G Olivet JL Rubino JL Melinte-DobrinescuMC Estrada F 2012 A two-step process for the refloodingof the Mediterranean after the Messinian Salinity Crisis BasinResearch 24 125ndash153

Baha El Din SM 2006 A guide to the reptiles and amphib-ians of Egypt CairondashNew York American University in CairoPress

Baier FS Sparrow DJ Wiedl H-J 2009 The amphibiansand reptiles of Cyprus Frankfurt am Main Edition Chimaira

Bandelt H-J Forster P Roumlhl A 1999 Median-joining net-works for inferring intraspecific phylogenies Molecular Biologyand Evolution 16 37ndash48

Bar A Haimovitch G 2011 A field guide to reptiles and am-phibians of Israel Herzliya Pazbar Limited

Boumlttger O 1879 Reptilien und Amphibien aus Syrien Berichtuumlber die Senckenbergische Naturforschende Gesellschaft inFrankfurt am Main 1879 57ndash84

Boulenger GA 1919 On a new variety of Acanthodactylusboskianus Daud from the Euphrates The Annals and Maga-zine of Natural History 3 549ndash550

Boulenger GA 1878 Sur les espegraveces drsquoAcanthodactylus desbords de la Mediterraneacutee Bulletin de la Socieacuteteacute zoologiquede France 3 179ndash197

Boulenger GA 1918 Sur les leacutezards du genre AcanthodactylusWieg Bulletin de la Socieacuteteacute zoologique de France 43 143ndash155

Boulenger GA 1921 Monograph of the Lacertidœ Vol II Trus-tees of the British Museum of Natural History London

Carranza S Arnold EN 2012 A review of the geckos of thegenus Hemidactylus (Squamata Gekkonidae) fromOman based on morphology mitochondrial and nuclear datawith descriptions of eight new species Zootaxa 3378 1ndash95

Carranza S Arnold EN Geniez P Roca J Mateo J 2008Radiation multiple dispersal and parallelism in the skinksChalcides and Sphenops (Squamata Scincidae) with com-ments on Scincus and Scincopus and the age of the SaharaDesert Molecular Phylogenetics and Evolution 46 1071ndash1094

Carretero MA Fonseca MM Garcia-Munoz E Brito JCHarris DJ 2011 Adding Acanthodactylus beershebensis tothe mtDNA phylogeny of the Acanthodactylus pardalis groupNorth-Western Journal of Zoology 7 138ndash142

Castresana J 2000 Selection of conserved blocks from multi-ple alignments for their use in phylogenetic analysis Mo-lecular Biology and Evolution 17 540ndash552

Clube TMM Robertson A 1986 The palaeorotation of theTroodos microplate Cyprus in the Late Mesozoic-Early Ce-nozoic plate tectonic framework of the Eastern Mediterra-nean Surveys in Geophysics 8 375ndash437

Cox SC Carranza S Brown RP 2010 Divergence timesand colonization of the Canary Islands by Gallotia lizardsMolecular Phylogenetics and Evolution 56 747ndash757

Crochet PA Geniez P Ineich I 2003 A multivariate analy-sis of the fringe-toed lizards of the Acanthodactylus scutellatusgroup (Squamata Lacertidae) systematic and biogeographi-cal implications Zoological Journal of the Linnean Society137 117ndash155

Crochet P-A Chaline O Surget-Groba Y Debain CCheylan M 2004 Speciation in mountains phylogeographyand phylogeny of the rock lizards genus Iberolacerta (ReptiliaLacertidae) Molecular Phylogenetics and Evolution 30 860ndash866

Daudin FM 1802 Histoire naturelle geacuteneacuterale et particuliegraveredes reptiles ouvrage faisant suite agrave lrsquoHistoire naturelle geacuteneacuteraleet particuliegravere F Dufart

Disi A Modry D Necas P Rifai L 2001 Amphibians andreptiles of the Hashemite Kingdom of Jordan an atlas andfield guide Frankfurt am Main Chimaira

Dolev A Pervolutzki A 2004 Endangered species in IsraelRed list of threatened animals Vertebrates JerusalemThe Nature and Parks Authority and the Society for thePreservation of Nature

Drummond AJ Rambaut A 2007 BEAST Bayesian evo-lutionary analysis by sampling trees BMC EvolutionaryBiology 7 214

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 17

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Ezard T Fujisawa T Barraclough T 2009 Splits speciesrsquolimits by threshold statistics R package version 10-19r48Available at httpR-ForgeR-projectorgprojectssplits

Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783ndash791

Fitzinger LJFJ 1834 Acanthodactylus In Wiegman AFAed Herpetologia mexicana seu Descriptio amphibiorum NovaeHispaniae quae itineribus comitis De Sack Ferdinandi Deppeet Chr Guil Schiede in Museum zoologicum Berolinensepervenerunt Pars prima saurorum species amplectens adjectosystematis saurorum prodromo additisque multis in huncamphibiorum ordinem observationibus Berlin C G Luderitz10

Flot JF 2010 SeqPHASE a web tool for interconvertingPHASE inputoutput files and FASTA sequence align-ments Molecular Ecology Resources 10 162ndash166

Fonseca MM Brito JC Paulo OS Carretero MA HarrisDJ 2009 Systematic and phylogeographical assessment ofthe Acanthodactylus erythrurus group (Reptilia Lacertidae)based on phylogenetic analyses of mitochondrial and nuclearDNA Molecular Phylogenetics and Evolution 51 131ndash142

Fonseca MM Brito JC Rebelo H Kalboussi M LarbesS Carretero MA Harris DJ 2008 Genetic variation amongspiny-footed lizards in the Acanthodactylus pardalis groupfrom North Africa African Zoology 43 8ndash15

Fontaneto D Herniou EA Boschetti C Caprioli M MeloneG Ricci C Barraclough TG 2007 Independently evolv-ing species in asexual bdelloid rotifers PLoS Biology 5 e87

Franzen M 1998 Erstnachweis von Acanthodactylus schreiberischreiberi Boulenger 1879 fuumlr die Tuumlrkei (Squamata SauriaLacertidae) Herpetozoa 11 27ndash36

Funk DJ Omland KE 2003 Species-level paraphyly andpolyphyly frequency causes and consequences with in-sights from animal mitochondrial DNA Annual Review ofEcology Evolution and Systematics 34 397ndash423

Godinho R Crespo EG Ferrand N Harris DJ 2005 Phy-logeny and evolution of the green lizards Lacerta spp(Squamata Lacertidae) based on mitochondrial and nuclearDNA sequences Amphibia-Reptilia 26 271ndash285

Greenbaum E Villanueva CO Kusamba C Aristote MMBranch WR 2011 A molecular phylogeny of EquatorialAfrican Lacertidae with the description of a new genus andspecies from eastern Democratic Republic of the Congo Zoo-logical Journal of the Linnean Society 163 913ndash942

Guillou H Carracedo JC Torrado FP Badiola ER 1996K-Ar ages and magnetic stratigraphy of a hotspot-inducedfast grown oceanic island El Hierro Canary Islands Journalof Volcanology and Geothermal Research 73 141ndash155

Harris DJ Arnold EN 2000 Elucidation of the relation-ships of spiny-footed lizards Acanthodactylus spp (ReptiliaLacertidae) using mitochondrial DNA sequence with com-ments on their biogeography and evolution Journal of Zoology252 351ndash362

Harris DJ Batista V Carretero M 2004 Assessment ofgenetic diversity within Acanthodactylus erythrurus (ReptiliaLacertidae) in Morocco and the Iberian Peninsula usingmitochondrial DNA sequence data Amphibia-Reptilia 25227

Horowitz A 1979 The Quaternary of Israel New York Aca-demic Press

Hraoui-Bloquet S Sadek RA Sindaco R Venchi A 2002The herpetofauna of Lebanon new data on distributionZoology in the Middle East 27 35ndash46

Hsuuml KJ Montadert L Bernoulli D Cita MB EricksonA Garrison RE Kidd RB Megraveliereacutes F Muumlller C WrightR 1977 History of the Mediterranean salinity crisis Nature267 399ndash403

Huelsenbeck JP Rannala B 2004 Frequentist propertiesof Bayesian posterior probabilities of phylogenetic trees undersimple and complex substitution models Systematic Biology53 904ndash913

Huelsenbeck JP Ronquist F 2001 MRBAYES Bayesianinference of phylogenetic trees Bioinformatics 17 754ndash755

Jolivet L Augier R Robin C Suc J-P Rouchy JM 2006Lithospheric-scale geodynamic context of the Messinian sa-linity crisis Sedimentary Geology 188 9ndash33

Kapli P Lymberakis P Poulakakis N Mantziou GParmakelis A Mylonas M 2008 Molecular phylogeny ofthree Mesalina (Reptilia Lacertidae) species (M guttulataM brevirostris and M bahaeldini) from North Africa and theMiddle East another case of paraphyly MolecularPhylogenetics and Evolution 49 102ndash110

Katoh K Toh H 2008 Recent developments in the MAFFTmultiple sequence alignment program Briefings inBioinformatics 9 286ndash298

Krijgsman W Hilgen F Raffi I Sierro F Wilson D 1999Chronology causes and progression of the Messinian salin-ity crisis Nature 400 652ndash655

Lanfear R Calcott B Ho SY Guindon S 2012PartitionFinder combined selection of partitioning schemesand substitution models for phylogenetic analyses Molecu-lar Biology and Evolution 29 1695ndash1701

Le Houeacuterou HN 1997 Climate flora and fauna changes inthe Sahara over the past 500 million years Journal of AridEnvironments 37 619ndash647

Maca-Meyer N Carranza S Rando J Arnold E CabreraV 2003 Status and relationships of the extinct giant CanaryIsland lizard Gallotia goliath (Reptilia Lacertidae)assessed using ancient mtDNA from its mummifiedremains Biological Journal of the Linnean Society 80 659ndash670

McCallum J Robertson A 1990 Pulsed uplift of the TroodosMassif ndash evidence from the Plio-Pleistocene Mesaoria basinIn J Malpas EMM Panayiotou A Xenophontos C edsOphiolites oceanic crustal analogues Proceedings of theSymposium lsquoTroodos 1987rsquo Nicosia (The Geological SurveyDepartment Ministry of Agriculture and NaturalResources) 217ndash229

Metallinou M Arnold EN Crochet P-A Geniez P BritoJC Lymberakis P El Din SB Sindaco R Robinson MCarranza S 2012 Conquering the Sahara and Arabiandeserts systematics and biogeography of Stenodactylus geckos(Reptilia Gekkonidae) BMC Evolutionary Biology 12258

Monaghan MT Wild R Elliot M Fujisawa T Balke MInward DJ Lees DC Ranaivosolo R Eggleton P

18 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Ezard T Fujisawa T Barraclough T 2009 Splits speciesrsquolimits by threshold statistics R package version 10-19r48Available at httpR-ForgeR-projectorgprojectssplits

Felsenstein J 1985 Confidence limits on phylogeniesan approach using the bootstrap Evolution 39 783ndash791

Fitzinger LJFJ 1834 Acanthodactylus In Wiegman AFAed Herpetologia mexicana seu Descriptio amphibiorum NovaeHispaniae quae itineribus comitis De Sack Ferdinandi Deppeet Chr Guil Schiede in Museum zoologicum Berolinensepervenerunt Pars prima saurorum species amplectens adjectosystematis saurorum prodromo additisque multis in huncamphibiorum ordinem observationibus Berlin C G Luderitz10

Flot JF 2010 SeqPHASE a web tool for interconvertingPHASE inputoutput files and FASTA sequence align-ments Molecular Ecology Resources 10 162ndash166

Fonseca MM Brito JC Paulo OS Carretero MA HarrisDJ 2009 Systematic and phylogeographical assessment ofthe Acanthodactylus erythrurus group (Reptilia Lacertidae)based on phylogenetic analyses of mitochondrial and nuclearDNA Molecular Phylogenetics and Evolution 51 131ndash142

Fonseca MM Brito JC Rebelo H Kalboussi M LarbesS Carretero MA Harris DJ 2008 Genetic variation amongspiny-footed lizards in the Acanthodactylus pardalis groupfrom North Africa African Zoology 43 8ndash15

Fontaneto D Herniou EA Boschetti C Caprioli M MeloneG Ricci C Barraclough TG 2007 Independently evolv-ing species in asexual bdelloid rotifers PLoS Biology 5 e87

Franzen M 1998 Erstnachweis von Acanthodactylus schreiberischreiberi Boulenger 1879 fuumlr die Tuumlrkei (Squamata SauriaLacertidae) Herpetozoa 11 27ndash36

Funk DJ Omland KE 2003 Species-level paraphyly andpolyphyly frequency causes and consequences with in-sights from animal mitochondrial DNA Annual Review ofEcology Evolution and Systematics 34 397ndash423

Godinho R Crespo EG Ferrand N Harris DJ 2005 Phy-logeny and evolution of the green lizards Lacerta spp(Squamata Lacertidae) based on mitochondrial and nuclearDNA sequences Amphibia-Reptilia 26 271ndash285

Greenbaum E Villanueva CO Kusamba C Aristote MMBranch WR 2011 A molecular phylogeny of EquatorialAfrican Lacertidae with the description of a new genus andspecies from eastern Democratic Republic of the Congo Zoo-logical Journal of the Linnean Society 163 913ndash942

Guillou H Carracedo JC Torrado FP Badiola ER 1996K-Ar ages and magnetic stratigraphy of a hotspot-inducedfast grown oceanic island El Hierro Canary Islands Journalof Volcanology and Geothermal Research 73 141ndash155

Harris DJ Arnold EN 2000 Elucidation of the relation-ships of spiny-footed lizards Acanthodactylus spp (ReptiliaLacertidae) using mitochondrial DNA sequence with com-ments on their biogeography and evolution Journal of Zoology252 351ndash362

Harris DJ Batista V Carretero M 2004 Assessment ofgenetic diversity within Acanthodactylus erythrurus (ReptiliaLacertidae) in Morocco and the Iberian Peninsula usingmitochondrial DNA sequence data Amphibia-Reptilia 25227

Horowitz A 1979 The Quaternary of Israel New York Aca-demic Press

Hraoui-Bloquet S Sadek RA Sindaco R Venchi A 2002The herpetofauna of Lebanon new data on distributionZoology in the Middle East 27 35ndash46

Hsuuml KJ Montadert L Bernoulli D Cita MB EricksonA Garrison RE Kidd RB Megraveliereacutes F Muumlller C WrightR 1977 History of the Mediterranean salinity crisis Nature267 399ndash403

Huelsenbeck JP Rannala B 2004 Frequentist propertiesof Bayesian posterior probabilities of phylogenetic trees undersimple and complex substitution models Systematic Biology53 904ndash913

Huelsenbeck JP Ronquist F 2001 MRBAYES Bayesianinference of phylogenetic trees Bioinformatics 17 754ndash755

Jolivet L Augier R Robin C Suc J-P Rouchy JM 2006Lithospheric-scale geodynamic context of the Messinian sa-linity crisis Sedimentary Geology 188 9ndash33

Kapli P Lymberakis P Poulakakis N Mantziou GParmakelis A Mylonas M 2008 Molecular phylogeny ofthree Mesalina (Reptilia Lacertidae) species (M guttulataM brevirostris and M bahaeldini) from North Africa and theMiddle East another case of paraphyly MolecularPhylogenetics and Evolution 49 102ndash110

Katoh K Toh H 2008 Recent developments in the MAFFTmultiple sequence alignment program Briefings inBioinformatics 9 286ndash298

Krijgsman W Hilgen F Raffi I Sierro F Wilson D 1999Chronology causes and progression of the Messinian salin-ity crisis Nature 400 652ndash655

Lanfear R Calcott B Ho SY Guindon S 2012PartitionFinder combined selection of partitioning schemesand substitution models for phylogenetic analyses Molecu-lar Biology and Evolution 29 1695ndash1701

Le Houeacuterou HN 1997 Climate flora and fauna changes inthe Sahara over the past 500 million years Journal of AridEnvironments 37 619ndash647

Maca-Meyer N Carranza S Rando J Arnold E CabreraV 2003 Status and relationships of the extinct giant CanaryIsland lizard Gallotia goliath (Reptilia Lacertidae)assessed using ancient mtDNA from its mummifiedremains Biological Journal of the Linnean Society 80 659ndash670

McCallum J Robertson A 1990 Pulsed uplift of the TroodosMassif ndash evidence from the Plio-Pleistocene Mesaoria basinIn J Malpas EMM Panayiotou A Xenophontos C edsOphiolites oceanic crustal analogues Proceedings of theSymposium lsquoTroodos 1987rsquo Nicosia (The Geological SurveyDepartment Ministry of Agriculture and NaturalResources) 217ndash229

Metallinou M Arnold EN Crochet P-A Geniez P BritoJC Lymberakis P El Din SB Sindaco R Robinson MCarranza S 2012 Conquering the Sahara and Arabiandeserts systematics and biogeography of Stenodactylus geckos(Reptilia Gekkonidae) BMC Evolutionary Biology 12258

Monaghan MT Wild R Elliot M Fujisawa T Balke MInward DJ Lees DC Ranaivosolo R Eggleton P

18 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

Barraclough TG 2009 Accelerated species inventory onMadagascar using coalescent-based models of species delin-eation Systematic Biology 58 298ndash311

Mukasa SB Ludden JN 1987 Uranium-lead isotopic agesof plagiogranites from the Troodos ophiolite Cyprus and theirtectonic significance Geology 15 825ndash828

Nir D 1970 Geomorphology of Israel Jerusalem AcademonNouira S Blanc C 1999 Description drsquoune nouvelle espegravece

drsquoacanthodactyle de Tunisie Acanthodactylus mechriguensisn sp (Sauria Reptilia) Atti e Memorie dellrsquoEnte FaunaSiciliana 5 101ndash108

Pincheira-Donoso D Meiri S 2013 An intercontinental analy-sis of climate-driven body size clines in reptiles no supportfor patterns no signals of processes Evolutionary Biology40 562ndash578

Pons J Barraclough TG Gomez-Zurita J Cardoso A DuranDP Hazell S Kamoun S Sumlin WD Vogler AP 2006Sequence-based species delimitation for the DNA taxonomyof undescribed insects Systematic Biology 55 595ndash609

Posada D 2008 jModelTest phylogenetic model averagingMolecular Biology and Evolution 25 1253ndash1256

Poulakakis N Kapli P Kardamaki A Skourtanioti EGoumlcmen B Ilgaz Ccedil Kumlutas Y Avci A LymberakisP 2013 Comparative phylogeography of six herpetofaunaspecies in Cyprus late Miocene to Pleistocene colonizationroutes Biological Journal of the Linnean Society 108 619ndash635

Pyron RA Burbrink FT Wiens JJ 2013 A phylogeny andrevised classification of Squamata including 4161 species oflizards and snakes BMC Evolutionary Biology 13 93

Rambaut A Drummond A 2007 Tracer version 15 Avail-able at httpbeastbioedacukTracer

Rastegar-Pouyani N 1998 A new species of Acanthodactylus(Sauria Lacertidae) from Qasr-e-Shirin Kermanshah Prov-ince western Iran Proceedings of the California Academyof Sciences 50 257ndash265

Rastegar-Pouyani N 1999 First record of the lacertidAcanthodactylus boskianus (Sauria Lacertidae) Asiatic Her-petological Research 8 85ndash89

Reed CA Marx H 1959 A herpetological collection from north-eastern Iraq Transactions of the Kansas Academy of Science62 91ndash122

Robertson A 1990 Tectonic evolution of Cyprus In J MalpasEMM Panayiotou A Xenophontos C eds Ophiolites oceanicCrustal Analogues Proceedings of the Symposium lsquoTroodos1987rsquo Nicosia (The Geological Survey Department Ministryof Agriculture and Natural Resources) 235ndash250

Ronquist F Huelsenbeck JP 2003 MrBayes 3 Bayesianphylogenetic inference under mixed models Bioinformatics19 1572ndash1574

Salvador A 1982 A revision of the lizards of the genusAcanthodactylus (Sauria Lacertidae) Bonn ZoologischesForschungsinstitut und Museum Alexander Koenig

Schleich HH Kaumlstle W Kabisch K 1996 Amphibians andreptiles of North Africa Koenigstein Koeltz Scientific Books

Schuster M Duringer P Ghienne J-F Vignaud P MackayeHT Likius A Brunet M 2006 The age of the Sahara DesertScience 311 821ndash821

Shimodaira H 2002 An approximately unbiased test ofphylogenetic tree selection Systematic Biology 51 492ndash508

Shimodaira H Hasegawa M 1999 Multiple comparisons oflog-likelihoods with applications to phylogenetic inferenceMolecular Biology and Evolution 16 1114ndash1116

Shimodaira H Hasegawa M 2001 CONSEL for assess-ing the confidence of phylogenetic tree selection Bioinformatics17 1246ndash1247

Silvestro D Michalak I 2012 raxmlGUI a graphical front-end for RAxML Organisms Diversity amp Evolution 12 335ndash337

Sindaco R Jeremcenko VK 2008 The reptiles of the WesternPalearctic Latina Edizioni Belvedere

Sindaco R Venchi A Carpaneto GM Bologna MA 2000The reptiles of Anatolia a checklist and zoogeographical analy-sis Biogeographia 21 441ndash554

Stamatakis A 2006 RAxML-VI-HPC maximum likelihood-based phylogenetic analyses with thousands of taxa and mixedmodels Bioinformatics 22 2688ndash2690

Steininger FF Roumlgl F 1984 Paleogeography and palinspasticreconstruction of the Neogene of the Mediterranean andParatethys In Dixon JE Robertson AHF eds The geologi-cal evolution of the eastern Mediterranean London Geologi-cal Society London Special Publications 17 659ndash668

Stephens M Scheet P 2005 Accounting for decay of linkagedisequilibrium in haplotype inference and missing-data im-putation The American Journal of Human Genetics 76 449ndash462

Stephens M Smith NJ Donnelly P 2001 A new statisti-cal method for haplotype reconstruction from populationdata The American Journal of Human Genetics 68 978ndash989

Talavera G Castresana J 2007 Improvement of phylogeniesafter removing divergent and ambiguously aligned blocks fromprotein sequence alignments Systematic Biology 56 564ndash577

Tamura K Peterson D Peterson N Stecher G Nei MKumar S 2011 MEGA5 molecular evolutionary geneticsanalysis using maximum likelihood evolutionary distanceand maximum parsimony methods Molecular Biology andEvolution 28 2731ndash2739

Trape J-F Trape S Chirio L 2012 Leacutezards crocodiles ettortues drsquoAfrique occidentale et du Sahara Marseille IRDOrstom

Uetz P 2013 The reptile database Available at httpwwwreptile-databaseorg

Wilcox TP Zwickl DJ Heath TA Hillis DM 2002 Phylogeneticrelationships of the dwarf boas and a comparison of Bayes-ian and bootstrap measures of phylogenetic support Mo-lecular Phylogenetics and Evolution 25 361ndash371

Yalccedilinkaya D Goumlccedilmen B 2012 A new subspecies from Ana-tolia Acanthodactylus schreiberi Boulenger 1879 ataturi nssp (Squamata Lacertidae) Biharean Biologist 6 19ndash31

Yom-Tov Y 1988 The zoogeography of the birds and mammalsof Israel In Yom-Tov Y Tchernov E eds The zoogeogra-phy of Israel the distribution and abundance at a zooge-ographical crossroad Dordrecht Kluwer Academic Publishers389ndash410

ACANTHODACTYLUS SCHREIBERI PHYLOGENY 19

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014

SUPPORTING INFORMATION

Additional Supporting Information may be found in the online version of this article at the publisherrsquos web-site

Figure S1 Bayesian inference tree of the Acanthodactylus boskianus and Acanthodactylus schreiberi speci-mens inferred using melano-cortin 1 receptor (MC1R) acetylcholinergic receptor Muscarinic 4 (ACM4) and oocytematuration factor MOS (c-mos) nuclear gene fragments Posterior probability in the Bayesian analysis is in-dicated by black dots on the nodes (values ge 095 shown) and maximum likelihood bootstrap support valuesare indicated in parentheses (values ge 70 shown) Sample codes and colours correlate to specimens in Table 1and in Figures 1ndash3Figure S2 Phylogenetic tree of the generalized mixed Yule-coalescent model based on the Bayesian mtDNAhaplotype data with a single threshold model for the partitions by genes The threshold between intra- vsinterspecific variation is indicated by a vertical red lineFigure S3 Phylogenetic tree of the generalized mixed Yule-coalescent model based on the Bayesian mtDNAhaplotype data with a single threshold model for the partitions based on PartitionFinder The threshold betweenintra- vs interspecific variation is indicated by a vertical red lineTable S1 Information on the length and primers used (orientation reference and PCR conditions) for all genesin this study and the number of variable (V) and parsimony-informative (Pi) sites in the alignment calculatedfor the ingroup only

20 K TAMAR ET AL

copy 2014 The Linnean Society of London Zoological Journal of the Linnean Society 2014