species on the rocks: systematics and biogeography of the...
TRANSCRIPT
Molecular Phylogenetics and Evolution 85 (2015) 208–220
Contents lists available at ScienceDirect
Molecular Phylogenetics and Evolution
journal homepage: www.elsevier .com/locate /ympev
Species on the rocks: Systematics and biogeography of the rock-dwellingPtyodactylus geckos (Squamata: Phyllodactylidae) in North Africaand Arabia
http://dx.doi.org/10.1016/j.ympev.2015.02.0101055-7903/� 2015 Elsevier Inc. All rights reserved.
⇑ Corresponding author at: Department of Biology, Villanova University, 800Lancaster Avenue, Villanova, PA 19085, USA.
E-mail address: [email protected] (M. Metallinou).
Margarita Metallinou a,⇑, Jan Cervenka b, Pierre-André Crochet c, Lukáš Kratochvíl b, Thomas Wilms d,Philippe Geniez c, Mohammed Y. Shobrak e, José C. Brito f, Salvador Carranza a
a Institute of Evolutionary Biology (CSIC – University Pompeu Fabra), Passeig Marítim de la Barceloneta 37-49, 08003 Barcelona, Spainb Faculty of Science, Charles University in Prague, Department of Ecology, Vinicná 7, CZ-122 44 Prague, Czech Republicc CEFE UMR 5175, CNRS, Université de Montpellier, Université Paul-Valéry Montpellier, EPHE, 1919 route de Mende, 34293 Montpellier cedex 5, Franced Zoologischer Garten Frankfurt, Bernhard-Grzimek-Allee 1D, 60316 Frankfurt am Main, Germanye Biology Department, Faculty of Science, Taif University, 888, Taif, Saudi Arabiaf CIBIO, Centro de Investigação em Biodiversidade e Recursos Genéticos da Universidade do Porto, Instituto de Ciências Agrárias de Vairão, R. Padre Armando Quintas,4485-661 Vairão, Portugal
a r t i c l e i n f o a b s t r a c t
Article history:Received 28 August 2014Revised 7 February 2015Accepted 16 February 2015Available online 24 February 2015
Keywords:Multilocus phylogenyReptilesArid environmentsAllopatryUndescribed diversityTaxonomy
The understanding of the diversity of species in the Palearctic and the processes that have generated it isstill weak for large parts of the arid areas of North Africa and Arabia. Reptiles are among their mostremarkable representatives, with numerous groups well adapted to the diverse environments. ThePtyodactylus geckos are a strictly rock-dwelling genus with homogeneous morphology distributed acrossmountain formations and rocky plateaus from the western African ranges in Mauritania and the Maghrebto the eastern tip of the Arabian Peninsula, with an isolated species in southern Pakistan. Here, we use abroad sampling of 378 specimens, two mitochondrial (12S and cytb) and four nuclear (c-mos, MC1R,ACM4, RAG2) markers in order to obtain the first time-calibrated molecular phylogeny of the genusand place its diversification in a temporal framework. The results reveal high levels of intraspecific vari-ability, indicative of undescribed diversity, and they do not support the monophyly of one species (P.ragazzii). Ptyodactylus species are allopatric across most of their range, which may relate to their highpreference for the same type of structural habitat. The onset of their diversification is estimated to haveoccurred in the Late Oligocene, while that of several deep clades in the phylogeny took place during theLate Miocene, a period when an increase in aridification in North Africa and Arabia initiated.
� 2015 Elsevier Inc. All rights reserved.
1. Introduction
The biogeography of North Africa and the Sahara-Sahel has beenthe subject of extensive research, often using reptiles as modelorganisms (Carranza et al., 2008; Pook et al., 2009; Metallinouet al., 2012; Rato et al., 2012; Brito et al., 2014). The neighboringArabian Peninsula has received less attention, albeit being veryinteresting from a biogeographical point of view, with diverselandscapes and numerous endemics (Arnold, 1977, 1980a andother articles in the same volumes; Edgell, 2006; Carranza andArnold, 2012; Cox et al., 2012). These studies have addressed inter-esting questions on the origin and relationships between North
African and Arabian reptile faunas, the relative times of speciationevents in Sahara and Arabia and have revealed patterns ofcolonization and diversification in these arid environments.These areas combine a variety of different habitats fromMediterranean and transitional Afro-tropical to very arid desert,all of which have been colonized by different groups of reptiles(e.g. Arnold, 1980b, 1986; Sindaco and Jeremcenko, 2008; Geniezet al., 2011; Metallinou et al., 2012; Sindaco et al., 2013; Šmídet al., 2013a). Notably, they include regions that have been predict-ed to harbor higher species richness than is currently known(Ficetola et al., 2013) and evidence of previously undescribeddiversity is continuously coming to light (e.g. Busais and Joger,2011a; Geniez et al., 2011; Carranza and Arnold, 2012; Šmídet al., 2013b; Metallinou and Carranza, 2013; Badiane et al.,2014; Vasconcelos and Carranza, 2014).
M. Metallinou et al. / Molecular Phylogenetics and Evolution 85 (2015) 208–220 209
The geckos of the genus Ptyodactylus Goldfuss (1820) extendtheir distribution across a very large area of North Africa and theMiddle East and also include a single known representative fromsouthern Pakistan (Sindaco and Jeremcenko, 2008; Fig. 1). Like othergroups of reptiles from these arid areas [e.g. Acanthodactylus,Chalcides (Sindaco and Jeremcenko, 2008)], they can occupy differ-ent climatic regions, from the mesic Mediterranean strips to ariddeserts, and they also extend to the savannah region of the Sahelsouth of the Sahara desert. Ptyodactylus geckos are easily identifiableby their toe morphology: their slender digits end in a triangular padthat bears multiple divided lamellae in a fan-like pattern with aretractile claw in the distal expansion. They are relatively large(most of the species between 75 and 100 mm of snout-vent length;Heimes, 1987; Werner and Sivan, 1993; Schleich et al., 1996; Baha ElDin, 2006; Gardner, 2013) with slender body, large head andrelatively long limbs. All Ptyodactylus species are rock-climbing spe-cialists and exploit very similar structural habitats: rocky substratessuch as cliffs, boulders and caves (Baha El Din, 2006). They lack anyfemoral or preanal pores and can communicate using sound, whichin most species consists of a series of loud multiple clicks (Werner,1965; Frankenberg, 1974; Werner and Sivan, 1994). Although theirvertical pupil denotes a mainly nocturnal activity, most of thespecies can also have diurnal activity and one species, P. puiseuxiBoutan, 1893, is mainly diurnal, with morning and afternoon activitypeaks (Werner and Sivan, 1994; Disi et al., 2001). Ptyodactylusspecies are mainly allopatric or parapatric, but in the few areaswhere two species coexist, there is a clear temporal or spatial parti-tioning of foraging activity patterns (Baha El Din, 2006; Kratochvíl &Cervenka, pers. obs.).
Taxonomic work on Ptyodactylus has been hampered by thesuperficial similarity of its species (Arnold, 1980a; Baha El Din,2006). Although many Ptyodactylus taxa were described duringthe 19th century (often as ‘‘variations’’ of P. hasselquistii), for manyyears authors followed a conservative approach, treating mostforms as subspecies of P. hasselquistii (Donndorff, 1798)(Anderson, 1898; Loveridge, 1947; Werner, 1965; Kluge, 1967).After the works of Arnold (1986), Heimes (1987) and Schleichet al. (1996), a total of six species were recognized: two fromNorth Africa (P. oudrii Lataste, 1880, P. ragazzii Anderson, 1898),three the Middle East (P. hasselquistii, P. guttatus Heyden, 1827, P.puiseuxi), and one species restricted to south Pakistan (P. homolepisBlanford, 1876). Werner and Sivan (1993, 1994) studied the mor-phology and ecology of the three species occurring within theLevant (P. guttatus, P. puiseuxi and P. hasselquistii) and delineatedwith precision their distribution and sympatric areas. More recent-ly, Baha El Din (2006) ‘‘tentatively’’ elevated P. siphonorhinaAnderson (1896) from Egypt to species rank, hence recognizingseven species in total. Nevertheless, limited access to material fromthe extensive range of the genus hindered further systematic work,paucity even more pronounced in Arabia. Perera and Harris(2010a) and Froufe et al. (2013) demonstrated that P. oudrii andP. ragazzii, respectively, include several geographically concordantgenetic lineages that could correspond to distinct species.However, given the lack of material for genetic and morphologicalanalysis from the complete species’ range, they did not proposeany formal taxonomic changes. A taxonomic study lacking contex-tualization was recently published by Nazarov et al. (2013), com-plicating the picture even more by describing two new speciesbased on very few specimens collected from single localities inJordan (P. ananjevae Nazarov, Melnikov, Melnikova, 2013) andsouth Oman (P. dhofarensis Nazarov, Melnikov, Melnikova, 2013),and another one from four geographically very close localities inthe Hajar Mountains of North Oman (P. orlovi Nazarov, Melnikov,Melnikova, 2013). The lack of information on the distribution ofthese three species (reported only from the type localities or their
vicinity) has forced Gardner (2013) in the most recent field guideto treat them as part of the Ptyodactylus hasselquistii species com-plex until their taxonomy and distribution are clarified.
In this work, we assemble for the first time a geographicallycomprehensive dataset (Fig. 1) including representatives of allbut one (P. homolepis) Ptyodactylus members, integrating a largeamount of new data with information from all previous works onthe genus in one multilocus phylogenetic study. We investigatethe relationships among the currently known forms and the diver-gent lineages reported, assessing the overall diversity of the genus.We reconstruct these relationships within a calibrated timeframe,which allows us to explore the biogeographical history of thiswidely distributed group. Finally, we provide the essential frame-work and taxonomic information for a further systematic assess-ment of the genus Ptyodactylus.
2. Materials and methods
2.1. Taxon sampling and nomenclature
A total of 378 representatives of all but one currently acceptedPtyodactylus species from 220 different localities throughout NorthAfrica and the Middle East were included in the study (Fig. 1). Datafor 26 specimens of P. oudrii (Perera and Harris, 2010a) and 22 of P.ragazzii (Froufe et al., 2013) were retrieved from GenBank (seebelow). Based on recent phylogenetic assessments and availabilityof material, a specimen of Asaccus gallagheri was used as outgroup(Gamble et al., 2008, 2011, 2012; Pyron et al., 2013). Informationon all specimens included in the analyses is provided in Table S1,Supplementary material.
We follow nomenclature from Sindaco and Jeremcenko (2008)and Baha el Din (2006). Like Gardner (2013), as a result of the lackof information on the distribution, unavailability in publicrepositories of the DNA sequences used for the descriptions, andfailure to compare the new species with the relevant material (P.hasselquistii from across its distribution range), we treat the threespecies described recently by Nazarov et al. (2013) as part of theP. hasselquistii species complex until their taxonomic status isrevised (work in progress).
2.2. DNA extraction and sequencing
Genomic DNA was extracted from ethanol-preserved tail, ton-gue or liver tissue using either the DNeasy Blood & Tissue Kit(Qiagen, Valencia, CA, USA) or the SpeedTools Tissue DNAExtraction kit (Biotools, Madrid, Spain). A fragment of the 12SrRNA (12S) mitochondrial gene was PCR-amplified and sequencedin both directions for 329 samples, while 12S sequences weredownloaded from GenBank for the remaining 48 samples. Giventhe very high level of genetic variability detected in the 12S marker(see Results) and the lack of prior systematic information regardingeither morphological or genetic features of the organisms, theGMYC method (see next section) was used to select a subset of145 samples for further gene sampling: of these, 139 samples werenewly sequenced for an additional mitochondrial gene fragment,cytochrome b (cytb), and four nuclear ones: oocyte maturation fac-tor MOS (c-mos), melanocortin 1 receptor (MC1R), acetylcholiner-gic receptor M4 (ACM4) and recombination activating protein 2(RAG2). Available sequences of c-mos and ACM4 for the remaining6 samples were downloaded from GenBank. Primers and PCR con-ditions are listed in Table S2, Supplementary material. Contigassembly and translation to aminoacids was done in Geneious5.3 (Drummond et al., 2010), and heterozygote sites were codedaccording to IUPAC ambiguity codes.
A B150
170
169
149
160
168161
162
157
159158
155154
153
167166
152151
164163
156165
194
193
192
203
214
213212
202
206
201
195
205
204
196
197
200199211
207208
198
210209
216
217
80
215
19
1
17
3 2
18
5
29
4
2827
16 26
1525
14
2412
13
119
2310
2220
821
67
172
186
174 175179180
187173
176
181
183182
188
184
177185
171
Saharan AtlasHigh Atlas
Anti-
Atlas
Tassili
Hoggar
AirMountains
AssabaMountains
WadiRumGulf of Suez
150 3000 Km250 5000 Km
178
30˚N
20˚N
10˚N
10˚W 0˚ 10˚E 20˚E 30˚E 40˚E 50˚E
Clade B
P. guttatus
P. puiseuxi
P. siphonorhinaP. ragazzii
P. togoensis*
P. oudrii
Clade A
P. hasselquistii species complex
C
190189
191
323334
35
494839
50 5152 40
4746453843
44
41 4237
36
143141
144
145 146
147
148
142
130134
133132
131
135117
120118
121
119
6072
53
137
138139
140
136
116
104
115
106
107
108
127
105 123126
125
128129 124
11461
113
30
31
5668
6755
57
6364
62
6665
111
100
101
102
109
99112
110
98
122
54
73
103
58
69
7059
71
818
59597
94
96
93
92
89
79
77
87 88
78
80
86
75
84
76
83
7485
82
90
91
ArabianSea
PersianGulf Gulf
of Oman
Red Sea
Gulfof Aden
Rub al Khali
Dhofar
Hadhramaut
Mountains
Asir
Mountains
Hejaz
Mountains
Gulf of Suez
Hajar
Mountains
250 5000 Km
P. guttatus
P. puiseuxi
P. siphonorhina
P. ragazzii
P. togoensis*
P. oudrii
P. hasselquistiispecies complex
A B
C
Fig. 1. Sampling localities of the specimens included in this study. Type localities of the different species are marked with star symbols of the same color as the species. In theupper left map, known distribution ranges of Ptyodactylus species. Colored ranges are drawn qualitatively based on Baha El Din (2006) and Sindaco and Jeremcenko (2008) andupdated with new records reported herein. Sketched area represents the distribution range of ‘‘P. ragazzii’’, which is herein shown to be polyphyletic. Type localities of species ofthe P. hasselquistii complex described in Nazarov et al. (2013), for which no ranges are known, are represented by grey star symbols. (⁄) P. togoensis, formerly a synonym of P.ragazzii, is herein assigned to the western populations of ‘‘P. ragazzii’’, which is shown to be polyphyletic (see Sections 3.2 and 4.3.1 for more details). Colors refer also to Fig. 2, andlocality numbers to Fig. 2, Supplementary Fig. S1 and Supplementary Table S1.
210 M. Metallinou et al. / Molecular Phylogenetics and Evolution 85 (2015) 208–220
M. Metallinou et al. / Molecular Phylogenetics and Evolution 85 (2015) 208–220 211
2.3. Phylogenetic analyses and topological tests
Multiple sequence alignments were performed with the onlineapplication of MAFFT v.7 (Katoh and Toh, 2008) applying defaultsettings to parameters (Auto strategy, Gap opening penalty: 1.53,Offset value: 0.0). The models of molecular evolution were selectedunder the AIC information criterion using jModelTest v2.1.3(Guindon and Gascuel, 2003; Darriba et al., 2012). The best-fit par-titioning scheme and models of molecular evolution for the datasetincluding six concatenated loci were selected with PartitionFinderv.1.1.0 (Lanfear et al., 2012) under the same criterion and with thefollowing parameters set: branchlengths: unlinked; models:mrbayes; search: greedy; for the protein-coding genes, the firstand second position were tested as one partition, and the thirdas another (see all models selected in Table 1). For the taxa accept-ed herein, uncorrected p-distances with pairwise deletion for thetwo mitochondrial and four nuclear gene fragments were estimat-ed in MEGA v.5 (Tamura et al., 2011) using the reduced dataset of145 selected samples.
Both the 12S dataset and the concatenated dataset of the select-ed specimens (see above) were analyzed with maximum likelihood(ML) and Bayesian inference (BI) methods. Models and partitionswere set as shown in Table 1. ML analyses were done withRAxML 7.4.2 (Stamatakis, 2006) as implemented in raxmlGUI(Silvestro and Michalak, 2012), performing 100 ML inferences.Node support was assessed with bootstrap analysis (Felsenstein,1985) of 1000 replicates. The software BEAST v.1.7.5 (Drummondand Rambaut, 2007) was used for BI analyses, where the meanoverall rate was fixed to 1.0. Two individual runs of 4 � 107 gen-erations were performed with a sampling frequency of 4000. Thefollowing models and prior settings were applied, otherwise theywere left by default: Constant Size coalescent tree prior; randomstarting tree; base substitution prior Uniform (0, 100); alpha priorUniform (0, 10). In the analysis of the concatenated dataset, parti-tions and clock models were unlinked and the xml file wasmanually modified to set Ambiguities = ’’true’’ for the nuclear genefragments partition in order to account for variability in theheterozygous positions, instead of treating them as missing data.
Table 1Basic information on genetic markers and datasets.
Dataset N� ind.a Gene fragment N� seq.b
12S 377 12S 377
Six-marker concatenated 145 12S 144cytb 127cmos 134mc1r 122acm4 113rag2 109
7 tips BEAST dating 7 12S 7cytb 6cmos 7mc1r 7acm4 6rag2 6
72 tips BEAST dating 72 12S 72cytb 59cmos 62mc1r 57acm4 53rag2 49
a Number of individuals.b Number of sequences.c Alignment length.d Number of variable positions.e Number of parsimony-informative positions.f Partition scheme and model selected in PartitionFinder or jModelTest (see Section 2
Posterior trace plots and effective sample sizes of the runs weremonitored in Tracer v1.5 (Rambaut and Drummond, 2007) toensure convergence. The results of the individual runs were com-bined in LogCombiner discarding 10% of the samples and the ultra-metric tree was produced with TreeAnnotator (both provided withthe BEAST package).
The general mixed Yule-coalescent model (GMYC) (Pons et al.,2006; Fujisawa and Barraclough, 2013) analysis was performedusing the SPLITS package (Ezard et al., 2009) in R (RDevelopment Core Team, 2014) in order to objectively identifythe divergent lineages in the Ptyodactylus phylogeny. We usedthe latest version of GMYC (Fujisawa and Barraclough, 2013) andtested both the single-threshold and the multiple-thresholdapproach on the 12S ultrametric tree obtained (see above). Basedon the results of this analysis, one or a few specimens of eachGMYC entity were selected for further gene sampling.
The Approximately-Unbiased (AU) (Shimodaira, 2002) and theShimodaira-Hasegawa (SH) (Shimodaira and Hasegawa, 1999)tests were employed to test the monophyly of ‘‘P. ragazzii’’. Per-sitelog likelihoods for the unconstrained (best) ML tree and a treewhere all samples identified as ‘‘P. ragazzii’’ were forced in a mono-phyletic group were estimated in RAxML as above. P-values wereestimated in CONSEL (Shimodaira and Hasegawa, 2001).
2.4. Estimation of divergence times
Absolute divergence times were estimated in a Bayesian frame-work with the application of previously calculated mean rates ofmolecular evolution for the two mitochondrial markers 12S (mean:0.00755, stdev: 0.00247, substitutions per site per million years)and cytb (mean: 0.0228, stdev: 0.00806) (Carranza and Arnold,2012). The mean rate of the mitochondrial 12S marker calculatedby Carranza and Arnold (2012) coincides with the mean rate calcu-lated for the geckos of the genus Stenodactylus based on partiallydifferent calibration points (mean: 0.00701) (Metallinou et al.,2012). We performed two sets of analyses applying only the 12Srate, and applying both 12S and cytb rates, to compare the results.As a result of the high level of hidden diversity uncovered in this
Len.c Var.d PIe Scheme and modelf
403 216 209 GTR + I + G
406 211 202 12S+(1st + 2nd)cytb: GTR + I + G395 206 198 3rd cytb: GTR + I + G415 55 46 nuclears: GTR + I + G668 94 84431 44 31410 46 35
390 131 77 GTR + G395 149 66 TrN + G415 22 7 HKY + G668 47 14 GTR + G431 16 6 HKY + G410 19 6 HKY
405 209 186 GTR + I + G394 202 189 GTR + I + G415 51 39 HKY + G668 85 70 GTR + I + G431 37 23 TrN + I410 39 23 GTR + G
).
212 M. Metallinou et al. / Molecular Phylogenetics and Evolution 85 (2015) 208–220
study (see below), which complicates an accurate knowledge of thespecies limits, we chose to conduct concatenated gene tree ana-lyses testing two different combinations of the number of termi-nals to explore if this affected the results of the calibrationanalyses. A first combination including one representative fromeach one of the seven species considered in the present study(see Fig. 1), and a second including one representative from eachone of the lineages identified by the GMYC analysis (72 specimensin total). These concatenated datasets were analyzed partitioningper gene fragment with BEAST v.1.7.5, performing three runs of10 � 107 generations with a sampling frequency of 10,000. Meanrates for the mitochondrial markers were used for calibration(Uncorrelated Lognormal clock) with a normal prior as explainedabove, models as shown in Table 1, a Yule prior was selected forthe tree and the rest of the parameter priors and the xml filemodifications were as in previous BEAST analyses. An additionalrun was performed for all the analyses sampling entirely fromthe prior in order to ensure that the data were sufficientlyinformative.
2.5. Biogeographical analyses
A Bayesian discrete phylogeographic approach was used for thereconstruction of ancestral areas in BEAST 1.7.5, using the BayesianStochastic Search Variable Selection (BSSVS) (Lemey et al., 2009).The same dataset as in the 72-clade dating analysis was analyzed(see above). Geographical traits were attributed to the specimensincluded in the dataset based on their locality information and afive-region delimitation (Fig. 3) based on the biogeographical par-titioning of Africa with multivariate methods (Linder et al., 2012)and relevant biodiversity and herpetological literature (Šmídet al., 2013a; Brito et al., 2014). An exponential prior was usedfor the discrete location state rate (locations.clock.rate) withmean = 1.0 and offset = 0, and two runs of 20 � 107 generationswith a sampling frequency of 20,000 were performed. All other set-tings were as above.
3. Results
3.1. Taxon sampling and data matrices
The 12S dataset included 377 terminals (excluding the out-group) and 403 base pairs (bp), while the six-marker concatenateddataset included 145 terminals and a total of 2725 bp. Table 1 listsinformation on all markers and datasets used in the phylogeneticand dating analyses with their corresponding numbers of variableand parsimony informative positions, as well as models of molecu-lar evolution selected for each one. The uncorrected p-distances forthe mitochondrial and nuclear gene fragments between and withinthe seven species considered in the present work are summarizedin Table S3 of the Supplementary material.
3.2. Phylogenetic analyses and topological tests
The ML and BI analyses of the 12S dataset yielded almost iden-tical results, except for few, not well-supported, nodes (Fig. S1,Supplementary material). The level of genetic variability withinPtyodactylus is very high and this is reflected in the results of theGMYC analysis that recognized 72 different lineages with thesingle threshold approach and 116 with the multiple thresholdone (see Table S1). The result of the likelihood ratio test wassignificant (p < 0.05) in both cases indicating that the null model(i.e. single population) could be rejected. Herein, we consider theresults of the former estimation (72 clades), as it is probable that
non-uniform geographic sampling in our dataset is causingoversplitting (Talavera et al., 2013).
The levels of intraspecific variability in the 12S mitochondrialgene ranged between 1.49% and 11.31%, for the eastern clade of‘‘P. ragazzii’’ and the P. hasselquistii species complex, respectively.The levels of interspecific variability for the same marker rangedbetween 9.4% for the comparison between P. siphonorhina and P.guttatus, and 20.4% for the comparison between the two clades of‘‘P. ragazzii’’ (see below) (Table S3, Supplementary material).
In order to ensure a good representation of the geneticvariability and good geographical coverage in further multilocusanalyses, a total of 145 specimens were selected based on theresults of the GMYC analysis and taking into account the geo-graphical distribution of the localities. Of the 72 GMYC entitiesrecovered, only seven (six single specimens clades and one includ-ing two specimens) were not represented in the multilocus datasetas a result of the failure to amplify the nuclear gene fragments forthese samples (see Table S1, Supplementary material). Thesummary of the results of ML and BI analyses of this concatenateddataset are presented in Fig. 2. The topologies inferred were, onceagain, almost identical, with the main difference being the groupthat branches first in each case: in the ML analysis this was theclade formed by the West African populations of ‘‘P. ragazzii’’, likein the analyses of the 12S dataset with both methodologies, but inthe BI analysis it was P. oudrii, although the clade grouping allother species was not supported (posterior probability, pp: 0.60).
The species ‘‘Ptyodactylus ragazzii’’ as traditionally recognized(i.e. including both East and West African populations, see below)was not recovered as monophyletic in any of the analyses carriedout. The samples from East Africa (Ethiopia and Djibouti), closestto the type locality in present-day Eritrea and Ethiopia (Fig. 1), andthe samples from the western part of the species’ distribution(Mauritania, Mali, Burkina Faso, Benin, Niger, Algeria) form twowell-supported clades that branch in different places of the phy-logeny, making ‘‘P. ragazzii’’ polyphyletic (Figs. 1 and 2). The resultsof the topological tests indicate that, based on our multilocus data-set, the monophyly of this species can be clearly rejected (AU:0.006, SH: 0.011). In order to resolve the polyphyly of ‘‘P. ragazzii’’,we herein separate the western populations as a distinct species,to which we tentatively apply the name P. togoensis, which wasintroduced as a ‘‘variety’’ of P. hasselquistii from Northern Togo byTornier (1901) and synonymized with P. ragazzii by Werner(1919). The type locality of this form is found approximately115 km west of our southernmost sample from this area (see Fig. 1).
The clade of P. siphonorhina and P. guttatus, and their sisterrelationship with P. puiseuxi (all together forming clade C) arewell-supported in all analyses. Lastly, all the 12S and concatenateddataset analyses recovered the P. hasselquistii species complex asmonophyletic, even though support for this relationship was lowin ML analyses. Within this highly variable group, there are 37GMYC lineages (see Table S1, Supplementary materials), whichare divided into two well-supported clades. Clade A is distributedacross most of the species’ range, including Sudan and Egypt,across the Nile Valley and the northwestern coast of the Red Sea;northern, central and the western coast of the Arabian Peninsula;some inland parts of the Asir Mountains in southern Arabia andacross the Hajar Mountain range in Oman and the U.A.E. (Fig. 1).Clade B is distributed across the Asir Mountains (from Taif inSaudi Arabia down to southwestern Yemen), the coastal parts ofsouthern Yemen and the Hadhramaut, east to Dhofar in southOman (Fig. 1).
3.3. Divergence time estimation analyses
As shown in Table 2, the differences between the estimatedages using one representative of each species considered in the
ZFMK-E38_Egypt_1
NHMC.80.3.49.1_Syria_180
BEV.T3757_Jordan_13
JEM650_Yemen_136
Poud-P2_Morocco_157
SPM002962-105_Jordan_7
ZFMK 60844_Syria_178
NHMC.80.3.47.15_Egypt_2
IBE-S10231_Saudi Arabia_58
J 06/04_Jordan_174
SUD12/2010-50_Sudan_84
Prag-3038_Mauritania_200
IBE-S10115_Saudi Arabia_124
DNA44_Ethiopia_190
Poud-P5_Morocco_161
IBE-S10126_Saudi Arabia_57
DJI-02_Djibouti_189
S10632_Algeria_192
Jor44_Jordan_8
Poud-P24_Morocco_155
TW1008_Jordan_175
JOR003_Jordan_15
IBE-S10169_Saudi Arabia_59
IBE-S10315_Saudi Arabia_123
IBE-S10099_Saudi Arabia_55
IBE-S10012_Saudi Arabia_113
TW1009_Jordan_172
UAE37_UAE_40
TW1007_Jordan_175
Poud-908_Algeria_150
Prag-3379_Mauritania_198
NHMC.80.3.48.6_Syria_183
AO138_Oman_141
IBE-S10047_Saudi Arabia_101
BEV.T519_Algeria_193
IBE-S4198_Egypt_81
JEM122_Yemen_133
BEV.10369_Egypt_79
SPM002388-87_Mali_
BEV.8972_Egypt_80
IBE-S10306_Saudi Arabia_115
990J_Syria_179
IBE-S10290_Saudi Arabia_116
PTY12_Morocco_162
991J_Syria_182
JEM508_Yemen_117
JEM466_Yemen_131
BEV.7211_Egypt_93
Poud-P22_Morocco_158
BEV.8802_Israel_4
JOR042_Jordan_8
07-12M_Morocco_154
Poud-P18_Morocco_156
ZFMK 60862_Syria_184
NHMC.80.3.47.8_Jordan_6
JOR100_Jordan_173
07-09M_Morocco_153
ZFMK 70963_Oman_38
BEV.7210_Egypt_215
IBE-S10282_Saudi Arabia_99
DJI-01_Djibouti_189
IBE-S10324_Saudi Arabia_127
DJI-03_Djibouti_189
IBE-S10276_Saudi Arabia_56
SPM002378-2_Egypt_76
TW1000_Morocco_
JOR071_Jordan_12
BEV.9002_Egypt_77
IBE-S10272_Saudi Arabia_53
SPM002970-1_Egypt_76
IBE-S10286_Saudi Arabia_31
IBE-S10384_Saudi Arabia_102
BEV.T518_Algeria_193
BEV.8105_Morocco_152
BEV.T3039_Benin_195Prag-429_Burkina Faso_196
BEV.T3958_Jordan_95
JEM132_Yemen_132
BEV.2279_Niger_203
TW1005_Jordan_11
IBE-S10305_Saudi Arabia_103
SPM001852_Egypt_
OM04/2010-54_Oman_142
BEV.8486_Israel_94
NHMC.80.3.47.9_Jordan_14
BEV.8870_Mali_197
Jor35_Jordan_5
TW1006_Jordan_9
TW1001_Morocco_
Prag-349_Niger_202
JEM424_Yemen_130IBE-S10297_Saudi Arabia_125
BEV.T394_Egypt_3
06-14b_Israel_171
IBE-S10187_Saudi Arabia_100
ZFMK 60930_Syria_181
Prag-430_Burkina Faso_196
BEV.10359_Egypt_78
BEV.10156_Algeria_194
JEM110_Yemen_140JEM115_Yemen_138
ZFMK 60845_Syria_178
IBE-S10145_Saudi Arabia_54
JOR128_Jordan_16
JEM37_Yemen_60
Jor10_Jordan_15
AO60_Oman_37
JOR040_Jordan_5
OM7_Oman_39
JEM515_Yemen_118
Prag-409_Niger_201
JOR036_Jordan_5
BEV.T702_Algeria_149
IBE-S993_Morocco_159
OM04/2010-56_Oman_142
BEV.9000_Egypt_75
Poud-P17_Morocco_151
SPM002952-48_Egypt_216
JEM639_Yemen_137
IBE-S10320_Saudi Arabia_122
Prag-3128_Mauritania_199
UAE22_Oman_32
BEV.8998_Egypt_74
BEV.12105_Lebanon_177
JEM62_Yemen_143
BEV.7244_Egypt_92
IBE-S10171_Saudi Arabia_113
IBE-S10372_Saudi Arabia_31
SUD12/2010-98_Sudan_82
JOR050_Jordan_10
TMHC179_Ethiopia_191
UAE41_Oman_36
SUD12/2010-31_Sudan_83
ZFMK87104_Saudi Arabia_30
11022012A_Morocco_160
IBE-S10109_Saudi Arabia_73
NHMC.80.3.49.15_Jordan_176
TW1011_UAE_34
IBE-S10197_Saudi Arabia_98
IBE-S10263_Saudi Arabia_115
JEM559_Yemen_119IBE-S10085_Saudi Arabia_113
IBE-S10048_Saudi Arabia_31
TW1030_UAE_33
ZFMK 87179_Saudi Arabia_61
96
94
99
97
78
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
92
92
78
74
97
81
82
93
99
99
76
96
83
82
99
73
99
95
100
87
100
92
90
73 97
99
97
95
74
83
99
9299
100
9983
Clade A
Clade B
Clade C
P. oudrii
P. hasselquistiicomplex
P. siphonorhina
P. guttatus
P. puiseuxi
P. ragazzii
0.1
P. togoensis*
100
100
100100
99
88
100
100
100
96
99
Fig. 2. ML tree of the genus Ptyodactylus inferred using the six-locus concatenated dataset. Numbers next to the node refer to bootstrap support (values above 70), blackcircles indicate posterior probability in BI analysis above 0.95 and empty circles mark nodes that are not recovered in the latter analysis. The tree was rooted using Asaccusgallagheri. (⁄) P. togoensis, formerly a synonym of P. ragazzii, is herein assigned to the western populations of ‘‘P. ragazzii’’, which is shown to be polyphyletic (see Sections 3.2and 4.3.1 for more details). Codes of the tips represent Specimen Number, followed by Country of origin, followed by locality code, as illustrated in Fig. 2. Detailed informationon all samples is listed in Supplementary Table S1. Inset picture shows a specimen of P. cf. hasselquistii from Oman.
M. Metallinou et al. / Molecular Phylogenetics and Evolution 85 (2015) 208–220 213
Table 2Comparison of divergence time estimations.
Node 72 clades 7 species
12Sonly
12S + cytba 12Sonly
12S + cytb
Root 25.66(16.10–38.03)
27.07(17.66–37.96)
24.43(12.61–39.18)
26.53(13.45–42.23)
togoensis 14.68(8.44–22.11)
15.47(9.52–22.42)
– –
oudrii 12.02(7.09–17.77)
12.69(8.06–18.10)
– –
hasselquistii + ragazzii + CladeC
18.09(10.96–26.23)
19.10(12.51–26.75)
18.93(9.70–30.20)
20.59(10.46–32.71)
ragazzii + Clade C 16.37(10.08–24.15)
17.29(11.44–24.71)
15.64(7.85–24.78)
17.01(8.91–27.37)
Clade C 9.57(5.83–14.26)
10.10(6.38–14.38)
8.68(4.45–13.76)
9.47(4.90–15.13)
siphonorhina + guttatus 6.02(3.58–9.00)
6.36(3.87–9.09)
5.35(2.64–8.60)
5.83(2.89–9.44)
puiseuxi 5.52(3.13–8.25)
5.83(3.54–8.53)
– –
siphonorhina 2.96(1.61–4.50)
3.13(1.85–4.60)
– –
guttatus 2.72(1.43–4.26)
2.88(1.62–4.39)
– –
hasselquistii complex 16.20(9.62–23.40)
17.11(11.18–24.12)
– –
Clade A 11.34(6.74–16.53)
11.99(7.66–16.83)
– –
Clade B 8.90(5.40–13.24)
9.40(5.81–13.32)
– –
a This set of dates is referred to in the text.
214 M. Metallinou et al. / Molecular Phylogenetics and Evolution 85 (2015) 208–220
present study (Fig. 2) and one representative of each of the 72GMYC lineages (Table S1, Supplementary materials) are not sub-stantial. The posterior mean rates for the mitochondrial markerswere 0.0258 for cytb and 0.0104 for 12S. Given the relatively highvalue of the standard deviation applied to the calibration rates ofboth 12S and cytb genes, the values of the 95% highest probabilitydensities (HPDs) were quite wide for all the nodes in the phylogeny(see Table 2). Also, the ages estimated applying solely the 12S rateor both 12S and cytb rates were quite similar, with the latter beingslightly older. Therefore, herein, we indicate the ages estimatedusing both rates and the 72-clades dataset, where more nodesare represented. This estimation suggests that diversification inthe genus Ptyodactylus started in the Late Oligocene (�27 Ma ago).
3.4. Ancestral area reconstruction
The results of the ancestral area reconstruction analysis usingthe BSSVS approach are presented in Fig. 3, on an annotated treewithin a temporal framework. Node ages were very similar tothe dating-only analysis of the same dataset. The area inferred tohave the highest posterior probability is attributed to each ances-tral node of the tree and in most cases posterior area probabilitywas above 0.80. The ancestral area for the P. hasselquistii radiationwas Arabia, and that of the clade grouping P. ragazzii and Clade Cwas most probably somewhere in northeastern Africa or the
Levant region. The ancestor of these two sister clades was mostprobably Arabian, but could also have been in the latter region.Finally, for the two most basal nodes in the phylogeny, westernNorth Africa sums the highest probability as the ancestral area,but inferences are far from conclusive.
4. Discussion
This study represents the first comprehensive systematic workon the genus Ptyodactylus, providing an insight on the phylogenyand systematics of this widely distributed group of North Africaand Arabia, based on multilocus molecular data. Taxonomic sam-pling has been shown to be very important in phylogenetic, speciesdelimitation, diversification and other studies (Dimitrov et al.,2012; Talavera et al., 2013; Ruane et al., 2014). Here, the samplingcovers the largest part of the known distribution range of the spe-cies (Heimes, 1987; Werner and Sivan, 1994; Sindaco andJeremcenko, 2008; see Fig. 1). For three of the members of thisgroup, P. puiseuxi, P. siphonorhina and P. ragazzii (see Results),DNA sequences are provided for the first time. Moreover, weinclude here an exhaustive taxonomic sampling of P. hasselquistii,for which DNA sequences from only three specimens were previ-ously available, and we investigate the patterns of high geneticdiversity in this and other species.
4.1. Habitat, morphology and allopatry in the genus Ptyodactylus
Despite the broad distribution of the genus Ptyodactylus (Fig. 1),all of its members present conserved morphologies, in terms of sizeand shape, and none has clearly adapted to different structuralhabitats that are very diverse across their range (Arnold, 1977,1980a; Werner and Sivan, 1994). In contrast to Ptyodactylus, sever-al other geckos from the same geographical areas, like Tarentola,Stenodactylus, Pristurus or members of the Arid clade ofHemidactylus (sensu Carranza and Arnold, 2006) comprise speciesthat are adapted to living on different substrates and make differ-ent habitat use (Carranza et al., 2000; Carranza and Arnold, 2006;Arnold, 2009; Metallinou et al., 2012). Even the highly conservedNorth African genus Tarentola presents a ground-dwelling memberwith very distinct morphology (Carranza et al., 2002), and thestrictly climbing geckos of the genus Asaccus include some small-sized members with rather distinct morphologies (Arnold andGardner, 1994; Gardner, 1994). This suggests that a strong selec-tive pressure might be acting to maintain this morphology forthe rock-climbing Ptyodactylus while, at the same time, this con-figuration of characters may hinder their versatility to exploit dif-ferent substrates. Indeed, it has been shown that habitat use iscorrelated with morphological evolution in rock-dwelling lizardsof the family Scincidae (Goodman et al., 2008) and that theextreme morphological divergence between two closely relatedspecies of the gecko genus Rhoptropus, a sand-dwelling and a rock-dwelling one, is related to locomotor performance on different sub-strates (Higham and Russell, 2010).
This morphological conservatism observed in Ptyodactylus hasprobably played a critical role in shaping the diversification anddistribution patterns of the members of the genus. All the speciesare largely allopatric or parapatric and very few areas of sympatryexist where a maximum of two species are present (Baha El Din,2006; Disi et al., 2001; Sindaco and Jeremcenko, 2008; Wernerand Sivan, 1993, 1994; see also Fig. 1). This contrasts with the alsonocturnal geckos of the genus Stenodactylus and the members ofthe Arid clade of Hemidactylus, which present very large areas ofsympatry among species, including up to three sympatric speciesand, in the case of Hemidactylus, up to four sympatric species insome areas of Socotra Island (Carranza and Arnold, 2012; Gómez-
region
Maghreb
E Sahel
Arabia
NE Africa & Levant
W Sahel
0510152025
JEM37_Yemen_60
991J_Syria_182
BEV8105_Morocco_152
Poud_P22_Morocco_158
IBE_S10109_Saudi_Arabia_73
BEV8870_Mali_197
SPM001852_Egypt_
NHMC80_3_48_6_Syria_183
IBE_S10047_Saudi_Arabia_101
IBE_S10012_Saudi_Arabia_113
ZFMK_60844_Syria_178
IBE_S10093_Saudi_Arabia_114
UAE41_Oman_36
IBE_S10272_Saudi_Arabia_53
IBE_S4198_Egypt_81
IBE_S10324_Saudi_Arabia_127
IBE_S10384_Saudi_Arabia_102
JEM62_Yemen_143
ZFMK_70963_Oman_38
IBE_S10197_Saudi_Arabia_98
SUD12/2010_98_Sudan_82
BEV7210_Egypt_215
UAE37_UAE_40
IBE_S10115_Saudi_Arabia_124
AO125_Oman_145
ZFMK_70961_Oman_47
07_09M_Morocco_153
IBE_S10297_Saudi_Arabia_125
BEV7244_Egypt_92
BEV_T519_Algeria_193
JOR128_Jordan_16
IBE_S10064_Saudi_Arabia_54
JEM466_Yemen_131
DNA44_Ethiopia_190
JEM424_Yemen_130
Poud_P18_Morocco_156
BEV_T702_Algeria_149
JEM559_Yemen_119
IBE_S10099_Saudi_Arabia_55
JEM650_Yemen_136
11022012A_Morocco_160
JEM113_Yemen_139
MHNG_2678_90_Mali_206
TW1009_Jordan_172
IBE_S993_Morocco_159
Poud_P24_Morocco_155
TMHC179_Ethiopia_191
IBE_S10305_Saudi_Arabia_103
IBE_S10290_Saudi_Arabia_116
SPM002962_105_Jordan_7TW1006_Jordan_9
JEM110_Yemen_140
DJI_02_Djibouti_189
Prag_429_Burkina_Faso_196
TW1011_UAE_34
BEV_T3039_Benin_195
BEV_T3958_Jordan_95
IBE_S10263_Saudi_Arabia_115
Prag_349_Niger_202
JOR036_Jordan_5
Prag_3128_Mauritania_199
BEV10156_Algeria_194Prag_409_Niger_201
IBE_S10145_Saudi_Arabia_54
IBE_S10096_Saudi_Arabia_102
JEM508_Yemen_117
Poud_P17_Morocco_151
ZFMK_60930_Syria_181
IBE_S10048_Saudi_Arabia_31
Poud_908_Algeria_150
06_14b_Israel_171
SPM002952_48_Egypt_216
P. togoensis*
P. oudrii
P. ragazzii
P. puiseuxi
P. siphonorhina
P. guttatus
P. h
asse
lqui
stii complex
A
B
C
D
E
Fig. 3. BEAST chronogram with ancestral area reconstruction using the BSSVS approach. Brach colors refer to areas as pictured in the lower left inset map. Branch width isproportional to the ancestral area posterior probability. Nodes that receive less than 0.80 probability are accompanied by a pie chart detailing the probability of each inferredarea (highest to lowest clockwise). Codes of the tips represent Specimen Number, followed by Country of origin, followed by locality code, as illustrated in Fig. 1. Detailedinformation on all samples is listed in Supplementary Table S1.
M. Metallinou et al. / Molecular Phylogenetics and Evolution 85 (2015) 208–220 215
Díaz et al., 2012; Metallinou et al., 2012; Šmíd et al., 2013a).Similarly, the diurnal geckos of the genus Pristurus also exhibitlarge areas with up to three sympatric species (Arnold, 2009;Sindaco and Jeremcenko, 2008).
4.2. General biogeographic patterns of Ptyodactylus
The characteristic and exclusive toe morphology of Ptyodactylusamong phyllodactylid geckos is accepted as a strong support forthe monophyly of the genus (Loveridge, 1947), which has also beenrecovered recently by molecular data, although not all the repre-sentatives of the genus were included (Gamble et al., 2008, 2011,2012; Pyron et al., 2013). Heimes (1987) in his revision ofPtyodactylus hypothesized that the Middle East acted as the centerof diversification of the genus, based on the relative species rich-ness, and from there they spread westwards to North Africa andthe Sahel and eastwards to Pakistan through the Strait ofHormuz. He did not provide, however, any insight into the sistergroup of Ptyodactylus. Recent molecular studies on the phylogeny
of Phyllodactylidae (Gamble et al., 2008, 2011, 2012; Pyron et al.,2013) have also failed to give an unequivocal answer to this ques-tion. In our phylogenetic and biogeographical reconstruction, thetwo west North African lineages are the first to branch out, withP. togoensis as a sister clade to the remaining Ptyodactylus (Figs. 2and 3). The ancestral area reconstruction favors a western NorthAfrican (either Maghreb or Sahel) origin and basal branching forPtyodactylus, but only slightly more than an eastern one, so resultswith the present data should be considered inconclusive. The pres-ence of several long branches in the chronogram (Fig. 3) may beindicative of unsampled and/or extinct clades, which would furthercomplicate ancestral inference.
The divergence time estimations (Table 2) indicate thatPtyodactylus have a rather old origin in the Late Oligocene.According to our results, Ptyodactylus started diversifyingapproximately 27 Ma ago (17.7–38, 95% HPD). This suggests thattheir similarity in terms of morphology and habitat use is not theresult of a recent origin and rapid dispersal of the genus acrossits current range but it is rather the result of a high degree of
216 M. Metallinou et al. / Molecular Phylogenetics and Evolution 85 (2015) 208–220
conservatism regarding these two aspects. The estimated time isvery similar to that inferred for the onset of diversification inStenodactylus (29.5 Ma [20.7–39.2, 95% HPD]) (Metallinou et al.,2012), and for the members of the Arid clade of Hemidactylus(29.1 Ma [19.2–40.3, 95% HPD]) (Šmíd et al., 2013a). This was atime of high geological instability in North Africa and Arabia as aresult of the onset of major seismic and volcanic events in theAfar region, affecting eastern Ethiopia, northeast Sudan and south-west Yemen (Menzies et al., 1992). These volcanic and tectonicevents triggered the formation of some of the most relevant andcomplex physiographic features in the contact zone betweenAfrica and Arabia, like the Gulf of Aden, the Red Sea and the eleva-tion of the Afro-Arabian rift-flanks to heights above 3600 m(Menzies et al., 1992; Bosworth et al., 2005; Autin et al., 2010),which had a major impact on the fauna and flora of the area (e.g.Pook et al., 2009; Thiv et al., 2010; Vargas-Ramírez et al., 2010;Portik and Papenfuss, 2012).
Diversification within the main clades of Ptyodactylus started inthe middle Miocene, and may be related to the progressive aridifi-cation of North Africa, which took place at the time (Flower andKennett, 1994). This phenomenon has been shown to have had aprofound effect on the diversification of biota (Douady et al.,2003; Guillaumet et al., 2008; Thiv et al., 2010; Gonçalves et al.,2012; Metallinou et al., 2012). In the cases of rock-dwelling ani-mals like Ptyodactylus or Agama (Gonçalves et al., 2012), the expan-sion of continuous sands probably acted as a vicariant agent,restricting their populations into areas of hard substrate (Britoet al., 2014; Fig. 1), although less in Agama, which are much moregeographically widespread (Leaché et al., 2014) than Ptyodactylus,probably due to their higher tolerance to temperature and sub-strate types. It is interesting to notice that the same expansion ofsand surfaces that promoted vicariance in the rock-dwellingPtyodactylus, facilitated dispersal in sand-adapted species, likethe geckos of the genus Stenodactylus and the scincids of the genusScincus and some sand-diving species of the genus Chalcides(Carranza et al., 2008; Metallinou et al., 2012). The situation ofPtyodactylus in Arabia is very similar to that of western NorthAfrica, with populations tightly linked to rocky environments andwith large desert areas like the Rub al Khali or flat arid plainsacting as barriers to dispersal (Fig. 1). Diversification in this areastarted approximately 12 Ma ago in clade A and 9.4 Ma ago in cladeB; dates that coincide with the allopatric speciation of the fivemorphologically similar Arabian species of the genus Stenodactylus(clade B in Metallinou et al., 2012). However, since little informa-tion is available on the geological and climatic history of theinterior of Arabia, the biogeographic patterns in both Ptyodactylusand Stenodactylus are very difficult to interpret.
4.3. Phylogeny and systematics of Ptyodactylus
The samples analyzed in this study fall within six major cladesin the Ptyodactylus phylogeny. Three of them, P. oudrii, P. togoensisand P. ragazzii, are distributed in North Africa and the Sahel; twoare found in northeastern Africa, the Levant and Arabia (clades Aand C) and one is exclusive to southern and southwestern Arabia(clade B) (Figs. 1 and 2). The monophyly of each one of these cladesreceives very high support in all the analyses, but not all relation-ships among them are equally supported (Fig. 2). The sister rela-tionship between P. ragazzii and clade C is supported in ML butnot BI analyses. Also, there is a topological difference betweenthe results of ML and BI analyses of the concatenated datasetregarding the first clade to branch in the phylogeny. In the follow-ing sections, we describe the major findings and phylogeneticstructure in each one of the aforementioned clades and explorethe inter- and intraspecific relationships of the Ptyodactylusmembers.
4.3.1. Polyphyletism of ‘‘Ptyodactylus ragazzii’’One of the most important results of our study is that the taxon
known as ‘‘P. ragazzii’’ is composed of two deep lineages thatbranch independently in the phylogeny (Figs. 1 and 2). As men-tioned above, this is confirmed by a topological test in which thealternative topology of monophyly is clearly rejected based onour dataset. The systematic issue caused by this non-monophylyis resolved herein by resurrecting the name Ptyodactylus togoensisTornier (1901) for the West African populations. Ptyodactylusragazzii was described on the basis of specimens from at least threelocalities in present-day Eritrea and Ethiopia (Anderson, 1898).Despite the distance of a few hundred kilometers between these(approximate) and our sampled localities (Fig. 1), given the extend-ed distribution of ‘‘P. ragazzii’’ across the Sahel, it can be at presentjustified to attribute P. ragazzii to our samples from Eastern Africa,pending further investigations in the area. It should be noted thatalthough the spelling for this name is ‘‘ragazzi’’ in the speciesdescription (Anderson, 1898, p. 62), a corrigendum issued simulta-neously with the original work confirms ‘‘ragazzii’’ as the correctspelling and this is accepted as valid according to theInternational Code of Zoological Nomenclature (Article 32.5).
Froufe et al. (2013) analyzed samples of P. togoensis (under thename P. ragazzii) and discussed the presence of three major lin-eages in their phylogeographic study: a western lineage from theTagant plateau and the Assaba mountains, a southern one fromthe western Sahel and an eastern one from the Aïr mountains inNiger. In our study, the second lineage is shown to extend furthernorthwest reaching the edge of the sands of the Sahara desert(localities 206 & 197 in Fig. 1). Our samples from the Hoggar andTassili mountains are assigned to the eastern lineage mirroringthe pattern uncovered in the lizard Agama tassiliensis from thesame mountain formations (Gonçalves et al., 2012). Interestingly,with the addition of new samples in this study, the geographicaldistance between the western and the southern lineages nowequals that between the southern and the eastern one.Nevertheless, the genetic distance between the last two is muchsmaller (data not shown), and in the phylogenetic tree these lin-eages present very little divergence, while the western cladeappears to be the most divergent (Fig. 2). The presence of sandydeserts immediately to the east of the Assaba Mountains couldbe responsible for this isolation, but without samples from theapparently more suitable area to the south a possible connectioncannot be discarded.
Apart from this specific area where further sampling is neededin order to confirm this pattern, the greater Sahel area from wherethere are abundant records of ‘‘P. ragazzii’’ (Sindaco andJeremcenko, 2008; Fig. 1) should be explored in order to establisha possible contact zone between P. togoensis and P. ragazzii. Giventhe presence of large unsampled areas within the distribution ofthe genus Ptyodactylus in the Sahara-Sahel, additional overlookedgenetic diversity should be expected.
4.3.2. New data on Ptyodactylus oudriiAnother Ptyodactylus species that exhibits deep genetic
intraspecific divergence is P. oudrii from the Maghreb, which ledPerera and Harris (2010a) to postulate the existence of a speciescomplex. Here, we include previous sequence data (with codes‘‘Poud’’, Fig. 2 and Fig. S1, Supplementary material) and provideexpanded taxonomic and genetic sampling that confirm the exis-tence of four major lineages with distinct geographic distributionwithin this group: an eastern lineage from the Saharan AtlasRange, two lineages from the High Atlas Range (one in the westand one in the east), and one from the Anti-Atlas. Our additionalsamples extend the range of the eastern lineage, previously knownonly from the eastern part of the Saharan Atlas Range in Algeria, toEastern Morocco. This range expansion opens the possibility that a
M. Metallinou et al. / Molecular Phylogenetics and Evolution 85 (2015) 208–220 217
contact zone exists between the eastern lineage and the lineageinhabiting the eastern parts of the High Atlas (locality 160 inFig. 1). The two other lineages inhabit the Western High Atlas –again with a possible contact zone with the previous lineage –and the Anti-Atlas.
The Saharan Atlas lineage is the most divergent one and speci-mens from the gap of 250 km that separates it from the easternHigh Atlas lineage would help circumscribe their limits.Ptyodactylus geckos are rock-dwellers, so this area with plenty ofrocky formations does not apparently present any barrier to theirdispersal. Instead, we hypothesize that reproductive rather thandispersal barriers exist between members of these two lineages.Importantly for future taxonomic considerations, the type localityof P. oudrii is found only 35 km north of our sampling locality169 (Fig. 1) included in the Saharan Atlas lineage.
The three lineages distributed across the High Atlas and theAnti-Atlas show a longitudinal geographical pattern of highdivergence (Figs. 1 and 2). Several groups of reptiles like Timontangitanus (Perera and Harris, 2010b), the high altitude lizards ofthe genus Atlantolacerta (Barata et al., 2012a), the geckos of thegenus Quedenfeldtia (Barata et al., 2012b), the agamid Agamaimpalearis (Brown et al., 2002; Gonçalves et al., 2012) and theterrapin Mauremys leprosa (Fritz et al., 2005, 2006) also presentrelatively high levels of genetic variability across the High Atlas andthe Anti-Atlas, although there is not a clear pattern across all taxa.
According to our divergence time estimations, the four deep lin-eages identified in P. oudrii originated approximately 12.7 Ma ago,a date much older than that inferred for some of the other reptilespecies mentioned above (A. impalearis, M. leprosa, T. tangitanus)and that roughly coincides with some estimations of the majoruplift of the Atlas Mountains (Gomez et al., 2000; Piqué et al.,2002). Nevertheless, this geological process was extended in timeand very complex (Frizon de Lamotte et al., 2008), and it is difficultto build specific scenarios on how the landscape of the AtlasMountains and processes during its formation have affected thediversity.
To conclude, the P. oudrii complex constitutes a very interestingmodel to explore diversification processes in the biodiversity-richAtlas Mountains, once its species limits become well defined.With the new data on the range of the lineages presented here,we pinpoint the crucial areas for completing the taxonomic sam-pling across potential contact zones, which will enable a fine-scalegenetic multilocus approach focused on the degree of reproductiveisolation between the lineages, ideally combined with a morpho-logical assessment of the different forms.
4.3.3. Ptyodactylus guttatus species group (Clade C)This clade represents a small radiation from the Levant that
originated approximately 10 Ma ago (Table 2) and resulted in threemainly allopatric or parapatric species (Figs. 1 and 2; Werner andSivan, 1994; Baha El Din, 2006). Thanks to its geographic locationat the crossroad of three continents, the Levant has long been con-sidered an area of great biogeographic interest (Werner and Sivan,1993; Šmíd et al., 2013a,b). The realization of this significancetogether with the relatively good accessibility of the area have con-tributed to a better knowledge of its faunal elements (Disi et al.,2001; Baha El Din, 2006; Sindaco and Jeremcenko, 2008), whichis also reflected on the taxonomy of the Ptyodactylus geckos.With the revisionary works of Werner and Sivan (1993, 1994)and Baha El Din (2006), the three species of this group, P. puiseuxi,P. guttatus and P. siphonorhina, became relatively well known froma morphological, ecological and physiological point of view.Genetic data on this group are presented here for the first timeand do not contradict the current taxonomical hypothesis basedon morphology and ecology.
Ptyodactylus siphonorhina was «tentatively elevated to the speci-fic level» recently (Baha El Din, 2006; p. 71) based on morpho-logical and ecological evidence that differentiated it from itssupposed sister species, P. guttatus. Although it had previouslybeen considered a nomen nudum by Werner and Sivan (1994), fol-lowing Wermuth (1965), or a synonym of P. guttatus by Heimes(1987), this latter author already observed different patterns inprotein electrophoresis analyses between P. guttatus from NorthAfrica and from Jordan and interpreted it as an old separationbetween the two populations (Heimes, 1987). Here, the geneticdistinctiveness of P. siphonorhina is confirmed based on a limited,but very well distributed across the species’ range, number of sam-ples which also includes a sample from very close to its type local-ity (Figs. 1 and 2). The two sister species are mainly distributed oneither side of the Gulf of Suez but they are sympatric in parts of thenorthern Eastern Desert (Baha El Din, 2006). Although, like allother Ptyodactylus, both species are mainly found on vertical rockysurfaces, boulders, under ledges and in caves, they have differentecological preferences that support their specific distinctiveness.While P. siphonorhina is found on rocks in highly arid open desert,P. guttatus is mainly found on hills, in areas of modest aridity (BahaEl Din, 2006). Despite its relatively large distribution range, thelevel of intraspecific genetic variability of P. siphonorhina is moder-ate (4.2% in the 12S; Table S3, Supplementary materials), whilethat of P. guttatus, which shows a more restricted distributionrange (Fig. 1), is even lower (2.4%). In this latter species, only thepopulation from Wadi Rum in Jordan seems to be genetically dis-tinct (Figs. 1 and 2).
The third taxon of this small radiation of Levantine Ptyodactylusis P. puiseuxi. Our dense sampling surrounds its type locality inpresent-day Israel (between loc. 171 and 185, Fig. 1). Distributedfurther north than P. guttatus, these two species can be distin-guished on the basis of morphology and present a mainly parapatricdistribution in Israel, with just a single site of sympatry that hasbeen considered a consequence of a recent human introductionof P. guttatus into the range of P. puiseuxi (Werner and Sivan,1993). Interestingly, in the same work, few specimens from thecontact zone presented intermediate states for some charactersand were hypothesized to be hybrids. Hybridization experimentsconfirmed that F1 hybrids could be obtained, albeit with low success,but not F2 (Werner and Sivan, 1996). Among the samples analyzedherein, there is no indication of presence of a hybrid, as bothP. guttatus and P. puiseuxi present unique alleles for all loci (datanot shown). Apart from their morphology, the two species alsodiffer in their behavior, ecology, physiology and biology, withdifferences in the activity patterns (P. puiseuxi diurnal and P. guttatusdiurnal-nocturnal) and vocalization being the most relevant ones(Werner and Sivan, 1994). Of all three species of clade C, P. puiseuxiis the one that presents the highest level of intraspecific diversitybased on our dataset (Table S3, Supplementary material) withthe samples from Syria grouping in a divergent clade (Figs. 1 and2 and S1 Supplementary material). Importantly, three names, cur-rently synonymized with P. puiseuxi, were proposed for specimenscoming from a single locality, at the ruins of Palmyra in Syria,which coincides with locality 182 of our sampling (Figs. 1 and 2):Ptyodactylus bischoffsheimi Boutan, 1893, P. montmahoui Boutan,1893 and P. barroisi Boutan, 1893. Bibliographic research revealsthat two more names have been placed under the synonymy ofP. puiseuxi: P. lobatus sancti-montis Barbour, 1914 (type locality8 km south of locality 177; Figs. 1 and 2) and P. lobatus syriacusPeracca, 1894 (part) (Werner and Sivan, 1994) (type localitybetween localities 27 (P. guttatus) and 187 (P. puiseuxi) of oursampling (Figs. 1 and S1 Supplementary material). Since P. lobatussyriacus had also been placed under the synonym of P. guttatus, theallocation of this name in the current systematics of Ptyodactylusis unclear and remains to be established. To conclude, further
218 M. Metallinou et al. / Molecular Phylogenetics and Evolution 85 (2015) 208–220
sampling mainly in the area between the two main genetic andgeographical clusters (in northern Lebanon and eastern Syria)would be needed to clarify if the divergent lineage from Syriamay represent a different form.
4.3.4. Ptyodactylus hasselquistii species complexPreliminary information on the morphological variability of P.
hasselquistii in Arabia became available with the work of Arnold(1980a, 1986) on the understudied herpetofauna of thePeninsula. He perceived it as intraspecific geographical variationand outlined a draft pattern of differentiation between animalsfrom northwest Arabia along the Hejaz mountains, from the Asirmountains, from Dhofar and the Hajar Mountains and from centralnorthern Arabia (Arnold, 1986). It therefore became evident that asystematic revision with comprehensive material of the speciesrange was warranted for a proper delimitation of the differentforms of this complex.
In this study, taxonomical sampling in the P. hasselquistii com-plex is one of the main deliverables in order to sufficiently docu-ment the genetic variability and provide the necessary materialfor a study using species delimitation methods (Metallinou et al.,unpubl. results). Our data uncovers for the first time the existenceof very high levels of genetic divergence within the P. hasselquistiispecies complex (11.3% in 12S and 18.2 in cytb; Table S3,Supplementary material), which is also supported by the identifi-cation of 37 GMYC lineages, two more than among all the remain-ing species of the genus included in our study. We consistentlyrecover two deep clades both in the mitochondrial and concatenat-ed phylogenies (Figs. 2 and S1, Supplementary material) that wename clades A and B. The genetic distance between these twoclades (14.1% in the 12S) is much higher than other interspecificdistances in the genus (Table S3, Supplementary material). Theonset of the diversification in the P. hasselquistii species complexdates back to approximately 17 Ma ago, while in clades A and B,12 Ma and 9.4 Ma ago, respectively (Table 2). These last ages arecomparable with the age of clade C (10.1 Ma) that comprises theradiation of the three species from the Levant (see above).
Among our sampling, the type locality of P. hasselquistii (Cairo,Egypt) is represented in our analyses by locality 81 belonging toclade A (Figs. 1 and 2). Two of the taxa described recently byNazarov et al. (2013) can also be tentatively allocated to this cladethanks to the proximity with some of our specimens. The typelocality of P. orlovi is found 28 km southwest of locality 38 and ofP. ananjevae, 52 km east of locality 5 (Figs. 1 and 2). Regarding P.dhofarensis, its type locality (Wadi Ayun) coincides with locality145 (Figs. 1 and S1, Supplementary material) of a sample com-prised in clade B. Nevertheless, as a result of the high level ofgenetic diversity in the P. hasselquistii species complex and theunavailability in public repositories of the cytochrome oxidase Isequences of the type specimens by Nazarov et al. (2013), no con-clusions regarding the limits of these taxa can be drawn at thispoint but will be the subject of a separate in-depth study.
The mountains of the Arabian Peninsula harbor a great diversityof flora and fauna (Arnold, 1972, 1977, 1980a and other articles inthe same volumes) and current systematic studies in reptiles areuncovering an important overlooked diversity in this area (e.g.Arnold and Gardner, 1994; Gardner, 1994; Busais and Joger,2011a; Carranza and Arnold, 2012; Metallinou and Carranza,2013; Šmíd et al., 2013a, b; among others). However, given the dif-ficult accessibility and political instability in the region, there arestill several parts that remain largely unexplored. It is thereforeexpected that unrecognized diversity is comprised within thePtyodactylus species complexes, the delimitation of which willrequire comprehensive analyses of genetic and morphological data(Metallinou et al., unpubl. results) similar to the work carried outon the geckos of the genus Hemidactylus of the Arabian Peninsula
(Busais and Joger, 2011a,b; Carranza and Arnold, 2012; Šmídet al., 2013a,b, 2015).
Acknowledgments
The authors would like to thank the people who donatedsamples or helped in the field: F. Amat, M. Aoutchiki, E.N. Arnold, O.Chaline, M. Cheylan, L. Chirio, A. Cluchier, D. Donaire, D. Escoriza,A. Foucart, M. Geniez, E. Gómez-Díaz, F. Hulbert, Y. Mansier, D.Modry, O. Peyre, J. Renoult, V. and X. Rufray, S. Scholz, M. Siol, F.Speyser, J. Viglione, and especially S. Baha El Din for theP. hasselquistii sample from the type locality, J. Šmíd for P. hasselquistiisamples from Sudan, P. Lymberakis for several P. puiseuxi samples,A. Nistri and T. Mazuch for P. ragazzii samples from Djibouti andEthiopia, respectively. We would like to thank E. Planas for invalu-able advice and help with the illustrations. Special thanks are dueto the members of the Nature Conservation Department of theMinistry of Environment and Climate, Sultanate of Oman fortheir help and support and for issuing collecting permits (Refs.:08/2005; 16/2008; 38/2010; 12/2011). We thank A.K. Nasher forsupport and Environment Protection Agency, Sana’a, Republic ofYemen for permits (Ref. 10/2007). We are thankful to theDeanship of academic research at Taif University for funding thesample collection in Saudi Arabia (Grant no. 1-433-2108). Thiswork was supported by Grant CGL2012-36970 from theMinisterio de Economía y Competitividad, Spain (co-funded byFEDER). M.M. was supported by a FPU predoctoral grant from theMinisterio de Educación, Cultura y Deporte, Spain (AP2008-01844). JCB is supported by Fundação para a Ciência e Tecnologia(IF/00459/2013). We thank two anonymous reviewers for helpfulcomments on an earlier version of the manuscript.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.ympev.2015.02.010.
References
Anderson, J., 1896. A Contribution to The Herpetology of Arabia: With a PreliminaryList of The Reptiles and Batrachians of Egypt. RH Porter, London, 122 pp.
Anderson, J., 1898. Zoology of Egypt. Reptilia and Batrachia. Bernard Quaritch,London, lxv+371 pp.
Arnold, E.N., 1972. Lizards with northern affinities from the mountains of Oman.Zool. Meded. 47, 111–128.
Arnold, E.N., 1977. Little-known geckoes (Reptilia: Gekkonidae) from Arabia withdescriptions of two new species from the Sultanate of Oman. Sci. Results OmanFlora Fauna Survey 1975, 81–110.
Arnold, E.N., 1980a. The reptiles and amphibians of Dhofar, southern Arabia. J.Oman Stud., Spec. Rep. 2, 273–332.
Arnold, E.N., 1980b. Reptiles of Saudi Arabia: a review of the Lizard GenusStenodactylus (Reptilia: Gekkonidae). Fauna of Saudi Arabia 2, 368–404.
Arnold, E.N., 1986. A key and annotated check list to the lizards andamphisbaenians of Arabia. Fauna of Saudi Arabia 8, 385–435.
Arnold, E.N., 2009. Relationships, evolution and biogeography of Semaphore geckos,Pristurus (Squamata, Sphaerodactylidae) based on morphology. Zootaxa 2060,1–21.
Arnold, E.N., Gardner, A.S., 1994. A review of the Middle Eastern leaf-toed geckoes(Gekkonidae: Asaccus) with descriptions of two new species from Oman. Faunaof Saudi Arabia 14, 424–441.
Autin, J., Leroy, S., Beslier, M.O., D’Acremont, E., Razin, P., Ribodetti, A., Bellahsen, N.,Robin, C., Al Toubi, K., 2010. Continental break-up history of a deep magma-poor margin based on seismic reflection data (northeastern Gulf of Adenmargin, offshore Oman). Geophys. J. Int. 180, 501–519.
Badiane, A., Garcia-Porta, J., Cervenka, J., Kratochvíl, L., Sindaco, R., Robinson, M.D.,Morales, H., Mazuch, T., Price, T., Amat, F., Shobrak, M.Y., Wilms, T., Simó-Riudalbas, M., Ahmadzadeh, F., Papenfuss, T., Cluchier, A., Viglione, J., Carranza,S., 2014. Phylogenetic relationships of Semaphore geckos (Squamata:Sphaerodactylidae: Pristurus) with an assessment of the taxonomy of Pristurusrupestris. Zootaxa 3835, 33–58.
Baha El Din, S., 2006. A Guide to the Reptiles and Amphibians of Egypt. Cairo andNew York, The American University in Cairo Press, xvi+359 pp.
M. Metallinou et al. / Molecular Phylogenetics and Evolution 85 (2015) 208–220 219
Barata, M., Carranza, S., Harris, D.J., 2012a. Extreme genetic diversity in the lizardAtlantolacerta andreanskyi (Werner, 1929): a montane cryptic species complex.BMC Evol. Biol. 12, 167.
Barata, M., Perera, A., Martínez-Freiría, F., Harris, D.J., 2012b. Cryptic diversitywithin the Moroccan endemic day geckos Quedenfeldtia (Squamata:Gekkonidae): a multidisciplinary approach using genetic, morphological andecological data. Biol. J. Linn. Soc. 106, 828–850.
Blanford, W., 1876. On some lizards from Sind, with descriptions of new species ofPtyodactylus, Stenodactylus and Trapelus. J. Asiatic Soc. Bengal 45, 18–26.
Bosworth, W., Huchon, P., McClay, K., 2005. The Red Sea and Gulf of Aden Basins. J.Afr. Earth Sci. 43, 334–378.
Boutan, L., 1893. Mémoire sur les reptiles rapportés de Syrie par le Docteur Théod.Barrois. 1re Partie. Genre Ptyodactyle. Revue Biologique du Nord de la France 5,1–36.
Brito, J.C., Godinho, R., Martínez-Freiría, F., Pleguezuelos, J.M., Rebelo, H., Santos, X.,Vale, C.G., Velo-Antón, G., Boratynski, Z., Carvalho, S.B., Ferreira, S., Gonçalves,D.V., Silva, T.L., Tarroso, P., Campos, J.C., Leite, J.V., Nogueira, J., Álvarez, F.,Sillero, N., Sow, A.S., Fahd, S., Crochet, P.-A., Carranza, S., 2014. Unravellingbiodiversity, evolution and threats to conservation in the Sahara-Sahel. Biol.Rev. 89, 215–231.
Brown, R., Suárez, N., Pestano, J., 2002. The Atlas mountains as a biogeographicaldivide in North-West Africa: evidence from mtDNA evolution in the Agamidlizard Agama impalearis. Mol. Phylogenet. Evol. 24, 324–332.
Busais, S., Joger, U., 2011a. Three new species and one new subspecies ofHemidactylus Oken, 1817 from Yemen (Squamata, Gekkonidae). VertebrateZool. 61, 267–280.
Busais, S., Joger, U., 2011b. Molecular phylogeny of the gecko genus HemidactylusOken, 1817 on the mainland of Yemen: (Reptilia: Gekkonidae). Zool. MiddleEast 53, 25–34.
Carranza, S., Arnold, E.N., 2006. Systematics, biogeography, and evolution ofHemidactylus geckos (Reptilia: Gekkonidae) elucidated using mitochondrialDNA sequences. Mol. Phylogenet. Evol. 38, 531–545.
Carranza, S., Arnold, E.N., 2012. A review of the geckos of the genus Hemidactylus(Squamata: Gekkonidae) from Oman based on morphology, mitochondrial andnuclear data, with descriptions of eight new species. Zootaxa 3378, 1–95.
Carranza, S., Arnold, E.N., Mateo, J., López-Jurado, L., 2000. Long-distancecolonization and radiation in gekkonid lizards, Tarentola (Reptilia:Gekkonidae), revealed by mitochondrial DNA sequences. Proc. Roy. Soc. Lond.B Biol. Sci. 267, 637.
Carranza, S., Arnold, E.N., Mateo, J., Geniez, P., 2002. Relationships and evolution ofthe North African geckos, Geckonia and Tarentola (Reptilia: Gekkonidae), basedon mitochondrial and nuclear DNA sequences. Mol. Phylogenet. Evol. 23, 244–256.
Carranza, S., Arnold, E.N., Geniez, P., Roca, J., Mateo, J., 2008. Radiation, multipledispersal and parallelism in the skinks, Chalcides and Sphenops (Squamata:Scincidae), with comments on Scincus and Scincopus and the age of the SaharaDesert. Mol. Phylogenet. Evol. 46, 1071–1094.
Cox, N., Mallon, D., Bowles, P., Els, J., Tognelli, M., 2012. The Conservation Status andDistribution of Reptiles of the Arabian Peninsula. By: IUCN, Gland, Switzerlandand Cambridge, UK and the Environment and Protected Areas Authority.
Darriba, D., Taboada, G.L., Doallo, R., Posada, D., 2012. JModelTest 2: more models,new heuristics and parallel computing. Nat. Methods 9, 772–772.
Dimitrov, D., Lopardo, L., Giribet, G., Arnedo, M.A., Alvarez-Padilla, F., Hormiga, G.,2012. Tangled in a sparse spider web: single origin of orb weavers and theirspinning work unravelled by denser taxonomic sampling. Proc. Roy. Soc. B: Biol.Sci. 279, 1341–1350.
Disi, A., Necas, P., Modry, D., Rifai, L., 2001. Amphibians and reptiles of theHashemite Kingdom of Jordan: An atlas and field guide. Ed. Chimaira, Frankfurt,408 pp.
Donndorff, J.A., 1798. Zoologische Beytraege zur xiii. Ausgabe des LinneischenNatursystems. Weidmannschen Buchhandlung, Leipzig, vi+980 pp.
Douady, C.J., Catzeflis, F., Raman, J., Springer, M.S., Stanhope, M.J., 2003. The Saharaas a vicariant agent, and the role of Miocene climatic events, in thediversification of the mammalian order Macroscelidea (elephant shrews).Proc. Natl. Acad. Sci. 100, 8325–8330.
Drummond, A.J., Rambaut, A., 2007. BEAST: Bayesian evolutionary analysis bysampling trees. BMC Evol. Biol. 7, 214.
Drummond, A.J., Ashton, B., Buxton, S., Cheung, M., Cooper, A., Heled, J., Kearse, M.,Moir, R., Stones-Havas, S., Sturrock, S., 2010. Geneious v5.1. Available from<http://www.geneious.com>.
Edgell, H.S., 2006. Arabian Deserts: Nature, Origin and Evolution. Springer,Dordrecht, lxvii+592 pp.
Ezard, T., Fujisawa, T., Barraclough, T., 2009. Splits: Species’ Limits by ThresholdStatistics. R Package Version, 1.
Felsenstein, J., 1985. Confidence limits on phylogenies: an approach using thebootstrap. Evolution 39, 783–791.
Ficetola, G.F., Bonardi, A., Sindaco, R., Padoa-Schioppa, E., 2013. Estimating patternsof reptile biodiversity in remote regions. J. Biogeogr. 40, 1202–1211.
Flower, B.P., Kennett, J.P., 1994. The middle Miocene climatic transition: eastAntarctic ice sheet development, deep ocean circulation and global carboncycling. Palaeogeogr. Palaeoclimatol. Palaeoecol. 108, 537–555.
Frankenberg, E., 1974. Vocalization of males of three geographical forms ofPtyodactylus from Israel (Reptilia: Sauria: Gekkoninae). J. Herpetol., 59–70
Fritz, U., Fritzsch, G., Lehr, E., Ducotterd, J.-M., Mueller, A., 2005. The AtlasMountains, not the Strait of Gibraltar, as a biogeographic barrier forMauremys leprosa (Reptilia: Testudines). Salamandra 41, 97–106.
Fritz, U., Barata, M., Busack, S.D., Fritzsch, G., Castilho, R., 2006. Impact of mountainchains, sea straits and peripheral populations on genetic and taxonomicstructure of a freshwater turtle, Mauremys leprosa (Reptilia, Testudines,Geoemydidae). Zoolog. Scr. 35, 97–108.
Frizon de Lamotte, D., Zizi, M., Missenard, Y., Hafid, M., El Azzouzi, M., Maury, R.C.,Charrière, A., Taki, Z., Benammi, M., Michard, A., 2008. The Atlas System. In:Michard, A., Saddiqi, O., Chalouan, A., Frizon de Lamotte, S. (Eds.), ContinentalEvolution: The Geology of Morocco Structure, Stratigraphy, and Tectonics of theAfrica-Atlantic-Mediterranean Triple Junction. Springer-Verlag, Berlin,Heidelberg, xviii+424 pp.
Froufe, E., Gonçalves, D.V., Brito, J.C., Harris, D.J., 2013. Nuclear and mitochondrialmarkers reveal the existence of several geographically concordant lineageswithin a Sahelian gecko species, Ptyodactylus ragazzii. Amphibia-Reptilia 34,85–93.
Fujisawa, T., Barraclough, T.G., 2013. Delimiting species using Single-Locus Data andthe Generalized Mixed Yule Coalescent Approach: a revised method andevaluation on simulated data sets. Syst. Biol. 62, 707–724.
Gamble, T., Bauer, A.M., Greenbaum, E., Jackman, T.R., 2008. Out of the blue: a novel,trans-Atlantic clade of geckos (Gekkota, Squamata). Zoolog. Scr. 37,355–366.
Gamble, T., Bauer, A.M., Colli, G.R., Greenbaum, E., Jackman, T.R., Vitt, L.J., Simons,A.M., 2011. Coming to America: multiple origins of New World geckos. J. Evol.Biol. 24, 231–244.
Gamble, T., Greenbaum, E., Jackman, T.R., Russell, A.P., Bauer, A.M., 2012. Repeatedorigin and loss of adhesive toepads in geckos. PLoS ONE 7, e39429.
Gardner, A.S., 1994. A new species of Asaccus (Gekkonidae) from the mountains ofnorthern Oman. J. Herpetol., 141–145
Gardner, A.S., 2013. The Amphibians and Reptiles of Oman and the U.A.E. Chimaira,Frankfurt am Main, 480 pp.
Geniez, P., Padial, J.M., Crochet, P.-A., 2011. Systematics of north African Agama(Reptilia: Agamidae): a new species from the central Saharan mountains.Zootaxa 3098, 26–46.
Goldfuss, G.A., 1820. Reptilia. In: Schubert, G.H. (Ed.), Handbuch derNaturgeschichte zum Gebrauch bei Vorlesungen, Handbuch der Zoologie, vol.3. J. L. Schrag, Nuernberg, pp. 121–181.
Gomez, F., Beauchamp, W., Barazangi, M., 2000. Role of the Atlas Mountains(northwest Africa) within the African-Eurasian plate-boundary zone. Geology28, 775–778.
Gómez-Díaz, E., Sindaco, R., Pupin, F., Fasola, M., Carranza, S., 2012. Origin andin situ diversification in Hemidactylus geckos of the Socotra Archipelago. Mol.Ecol. 21, 4074–4092.
Gonçalves, D.V., Brito, J.C., Crochet, P.-A., Geniez, P., Padial, J.M., Harris, D.J., 2012.Phylogeny of North African Agama lizards (Reptilia: Agamidae) and the role ofthe Sahara desert in vertebrate speciation. Mol. Phylogenet. Evol.64, 582–591.
Goodman, B.A., Miles, D.B., Schwarzkopf, L., 2008. Life on the rocks: habitat usedrives morphological and performance evolution in lizards. Ecology 89, 3462–3471.
Guillaumet, A., Crochet, P.-A., Pons, J.M., 2008. Climate-driven diversification in twowidespread Galerida larks. BMC Evol. Biol. 8, 32.
Guindon, S., Gascuel, O., 2003. A simple, fast, and accurate algorithm to estimatelarge phylogenies by maximum likelihood. Syst. Biol. 52, 696–704.
Heimes, P., 1987. Beitrag zur Systematik der Faecherfinger (Sauria: Gekhonidae:Ptyodactylus). Salamandra 23, 212–235.
Heyden, C.H.G., 1827. Reptilien. In: Rüppell, E. (Ed.) Atlas zu Reise im noerdlichenAfrika. H. L. Brönner, Frankfurt a.M., pp. 1–24.
Higham, T.E., Russell, A.P., 2010. Divergence in locomotor performance, ecology, andmorphology between two sympatric sister species of desert-dwelling gecko.Biol. J. Linn. Soc. 101, 860–869.
Katoh, K., Toh, H., 2008. Recent developments in the MAFFT multiple sequencealignment program. Brief. Bioinform. 9, 286–298.
Kluge, A.G., 1967. Higher taxonomic categories of gekkonid lizards and theirevolution. Bull. Am. Museum Natur. History 135, 1–60.
Lanfear, R., Calcott, B., Ho, S.Y., Guindon, S., 2012. PartitionFinder: combinedselection of partitioning schemes and substitution models for phylogeneticanalyses. Mol. Biol. Evol. 29, 1695–1701.
Lataste, F., 1880. Diagnoses de reptiles nouveaux d’Algerie. Le Naturaliste 1, 306–307.
Leaché, A.D., Wagner, P., Linkem, C.W., Böhme, W., Papenfuss, T.J., Chong, R.A., Lavin,B.R., Bauer, A.M., Nielsen, S.V., Greenbaum, E., Rödel, M.-O., Schmitz, A.,LeBreton, M., Ineich, I., Chirio, L., Ofori-Boateng, C., Eniang, E.A., Baha El Din,S., Lemmon, A.R., Burbrink, F.T., 2014. A hybrid phylogenetic–phylogenomicapproach for species tree estimation in African Agama lizards with applicationsto biogeography, character evolution, and diversification. Mol. Phylogenet. Evol.79, 215–230.
Lemey, P., Rambaut, A., Drummond, A., Suchard, M., 2009. Bayesian phylogeographyfinds its roots. PLoS Comput. Biol. 5, e1000520.
Linder, H.P., de Klerk, H.M., Born, J., Burgess, N.D., Fjeldså, J., Rahbek, C., 2012. Thepartitioning of Africa: statistically defined biogeographical regions in sub-Saharan Africa. J. Biogeogr. 39, 1189–1205.
Loveridge, A., 1947. Revision of the African lizards of the family Gekkonidae. Bull.Museum Compar. Zool. Harvard 98, 1–469.
Menzies, M.A., Baker, J., Bosence, D., Dart, C., Davison, I., Hurford, A., Al’Kadasi, M.,McClay, K., Nichols, G., Al’Subbary, A., Yelland, A., 1992. The timing ofmagmatism, uplift and crustal extension: preliminary observations fromYemen. Geol. Soc., Lond., Spec. Publ. 68, 293–304.
220 M. Metallinou et al. / Molecular Phylogenetics and Evolution 85 (2015) 208–220
Metallinou, M., Carranza, S., 2013. New species of Stenodactylus (Squamata:Gekkonidae) from the Sharqiyah Sands in northeastern Oman. Zootaxa 3745,449–468.
Metallinou, M., Arnold, E.N., Crochet, P.-A., Geniez, P., Brito, J.C., Lymberakis, P., BahaEl Din, S., Sindaco, R., Robinson, M., Carranza, S., 2012. Conquering the Saharaand Arabian deserts: Systematics and biogeography of Stenodactylus geckos(Reptilia: Gekkonidae). BMC Evol. Biol. 12, 258.
Nazarov, R., Melnikov, D., Melnikova, E., 2013. Three New Species of Ptyodactylus(Reptilia; Squamata; Phyllodactylidae) from the Middle East. Russian J.Herpetol. 20, 147–162.
Perera, A., Harris, D.J., 2010a. Genetic variability within the Oudri’s fan-footed geckoPtyodactylus oudrii in North Africa assessed using mitochondrial and nuclearDNA sequences. Mol. Phylogenet. Evol. 54, 634–639.
Perera, A., Harris, D.J., 2010b. Genetic variability in the ocellated lizard Timontangitanus in Morocco. Afr. Zool. 45, 321–329.
Piqué, A., Tricart, P., Guiraud, R., Laville, E., Bouaziz, S., Amrhar, M., Ouali, R.A., 2002.The Mesozoic-Cenozoic Atlas belt (North Africa): an overview. Geodin. Acta 15,185–208.
Pons, J., Barraclough, T.G., Gómez-Zurita, J., Cardoso, A., Duran, D.P., Hazell, S.,Kamoun, S., Sumlin, W.D., Vogler, A.P., 2006. Sequence-based speciesdelimitation for the DNA taxonomy of undescribed insects. Syst. Biol. 55,595–609.
Pook, C.E., Joger, U., Stümpel, N., Wüster, W., 2009. When continents collide:Phylogeny, historical biogeography and systematics of the medically importantviper genus Echis (Squamata: Serpentes: Viperidae). Mol. Phylogenet. Evol. 53,792–807.
Portik, D.M., Papenfuss, T.J., 2012. Monitors cross the Red Sea: the biogeographichistory of Varanus yemenensis. Mol. Phylogenet. Evol. 62, 561–565.
Pyron, R.A., Burbrink, F.T., Wiens, J.J., 2013. A phylogeny and revised classification ofSquamata, including 4161 species of lizards and snakes. BMC Evol. Biol. 13, 93.
R Core Team, 2014. R: A language and environment for statistical computing. RFoundation for Statistical Computing, Vienna, Austria.
Rambaut, A., Drummond, A.J., 2007. Tracer v1. 4.Rato, C., Carranza, S., Harris, D.J., 2012. Evolutionary history of the genus Tarentola
(Gekkota: Phyllodactylidae) from the Mediterranean Basin, estimated usingmultilocus sequence data. BMC Evol. Biol. 12, 14.
Ruane, S., Bryson, R.W., Pyron, R.A., Burbrink, F.T., 2014. Coalescent speciesdelimitation in milksnakes (genus Lampropeltis) and impacts on phylogeneticcomparative analyses. Syst. Biol. 63, 231–250.
Schleich, H.H., Kästle, W., Kabisch, K., 1996. Amphibians and Reptiles of NorthAfrica: Biology, Systematics, Field Guide. Koeltz Scientific Books, Königstein(Germany), 630 pp.
Shimodaira, H., 2002. An approximately unbiased test of phylogenetic treeselection. Syst. Biol. 51, 492.
Shimodaira, H., Hasegawa, M., 1999. Multiple comparisons of log-likelihoods withapplications to phylogenetic inference. Mol. Biol. Evol. 16, 1114–1116.
Shimodaira, H., Hasegawa, M., 2001. CONSEL: for assessing the confidence ofphylogenetic tree selection. Bioinformatics 17, 1246.
Silvestro, D., Michalak, I., 2012. RaxmlGUI: a graphical front-end for RAxML. Org.Diver. Evolut. 12, 335–337.
Sindaco, R., Jeremcenko, V.K., 2008. The Reptiles of the Western Palearctic. EdizioniBelvedere, Latina (Italy), 579 pp.
Sindaco, R., Venchi, A., Grieco, C., 2013. The Reptiles of the Western Palearctic. 2.Annotated Checklist and Distributional Atlas of the Snakes of Europe, NorthAfrica, Middle East and Central Asia. Edizioni Belvedere, Latina (Italy), 543 pp.
Šmíd, J., Carranza, S., Kratochvíl, L., Gvozdík, V., Nasher, A.K., Moravec, J., 2013a. Outof Arabia: a complex biogeographic history of multiple vicariance and dispersalevents in the gecko genus Hemidactylus (Reptilia: Gekkonidae). PLoS ONE 8,e64018.
Šmíd, J., Moravec, J., Kratochvíl, L., Gvozdík, V., Nasher, A.K., Busais, S.M., Wilms, T.,Shobrak, M.Y., Carranza, S., 2013b. Two newly recognized species ofHemidactylus (Squamata, Gekkonidae) from the Arabian Peninsula and Sinai,Egypt. ZooKeys 79.
Šmíd, J., Moravec, J., Kratochvíl, L., Nasher, A.K., Mazuch, T., Gvozdík, V., Carranza, S.,2015. Multilocus phylogeny and taxonomic revision of the Hemidactylusrobustus species group (Reptilia, Gekkonidae) with descriptions of three newspecies from Yemen and Ethiopia. Syst. Biodivers. 1, 1–23.
Stamatakis, A., 2006. RAxML-VI-HPC: maximum likelihood-based phylogeneticanalyses with thousands of taxa and mixed models. Bioinformatics 22, 2688.
Talavera, G., Dinca, V., Vila, R., 2013. Factors affecting species delimitations with theGMYC model: insights from a butterfly survey. Methods Ecol. Evol. 4, 1101–1110.
Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Kumar, S., 2011. MEGA5:molecular evolutionary genetics analysis using maximum likelihood,evolutionary distance, and maximum parsimony methods. Mol. Biol. Evol. 28,2731–2739.
Thiv, M., Thulin, M., Hjertson, M., Kropf, M., Linder, H.P., 2010. Evidence for avicariant origin of Macaronesian-Eritreo/Arabian disjunctions in CampylanthusRoth (Plantaginaceae). Mol. Phylogenet. Evol. 54, 607–616.
Tornier, G., 1901. Die crocodile, schildkroten und eidechsen in Togo. Archiv furNaturgeschichte 1901, 65–88.
Vargas-Ramírez, M., Vences, M., Branch, W.R., Daniels, S.R., Glaw, F., Hofmeyr, M.D.,Kuchling, G., Maran, J., Papenfuss, T.J., Siroky, P., Vieites, D.R., Fritz, U., 2010.Deep genealogical lineages in the widely distributed African helmeted terrapin:evidence from mitochondrial and nuclear DNA (Testudines: Pelomedusidae:Pelomedusa subrufa). Mol. Phylogenet. Evol. 56, 428–440.
Vasconcelos, R., Carranza, S., 2014. Systematics and biogeography of Hemidactylushomoeolepis Blanford, 1881 (Squamata: Gekkonidae), with the description of anew species from Arabia. Zootaxa 3835, 501–527.
Wermuth, H., 1965. Liste der rezenten Amphibien und Reptilien: Gekkonidae,Pygopodidae, Xanthusiidae, Das Tierreich 80; Berlin, xxii+246 pp.
Werner, F., 1919. Wissenschaftliche Ergebnisse der mit Unterstützung derKaiserlichen Akademie der Wissenschaften in Wien aus der Erbschaft Treitlvon F. Werner unternommenen zoologischen Expedition nach dem Anglo-Aegyptischen Sudan (Kordofan) 1914. Bearbeitung de Denkschriften derKaiserlichen Akademie der Wissenschaften/Mathematisch-Naturwissenschaftliche Classe, vol. 96, pp. 437–509.
Werner, Y., 1965. Ueber die israelischen Geckos der Gattung Ptyodactylus und ihreBiologie. Salamandra 1, 15–25.
Werner, Y.L., Sivan, N., 1993. Systematics and zoogeography of Ptyodactylus(Reptilia: Sauria: Gekkonidae) in the Levant: 1. Biometry of three species inIsrael. Revista Española de Herpetología 7, 47–64.
Werner, Y.L., Sivan, N., 1994. Systematics and zoogeography of Ptyodactylus(Reptilia: Sauria: Gekkonidae) in the Levant: 2, Taxonomy, with a review ofecology and zoogeography. Revista Española de Herpetología 8, 105–122.
Werner, Y.L., Sivan, N., 1996. Systematics and zoogeography of Ptyodactylus(Reptilia: Sauria: Gekkonidae) in the Levant: 3, Experimental and naturalhybrids of P. guttatus and P. puiseuxi. Israel J. Herpetol. 42, 185–202.