Submitted 30 August 2017Accepted 2 February 2018Published 20 February 2018

Corresponding authorAndreacutes S Quinterossebasquintgmailcom

Academic editorJane Hughes

Additional Information andDeclarations can be found onpage 23

DOI 107717peerj4404

Copyright2018 Portelli and Quinteros

Distributed underCreative Commons CC-BY 40

OPEN ACCESS

Phylogeny time divergence andhistorical biogeography of the SouthAmerican Liolaemus alticolor-bibroniigroup (Iguania Liolaemidae)Sabrina N Portelli and Andreacutes S Quinteros

UNSa-CONICET Instituto de Bio y Geociencias del NOA Rosario de Lerma Salta ArgentinaThese authors contributed equally to this work

ABSTRACTThe genus Liolaemus comprises more than 260 species and can be divided in twosubgeneraEulaemus andLiolaemus sensu stricto In this paperwepresent aphylogeneticanalysis divergence times and ancestral distribution ranges of the Liolaemus alticolor-bibronii group (Liolaemus sensu stricto subgenus) We inferred a total evidence phy-logeny combining molecular (Cytb and 12S genes) and morphological characters usingMaximum Parsimony and Bayesian Inference Divergence times were calculated usingBayesian MCMC with an uncorrelated lognormal distributed relaxed clock calibratedwith a fossil record Ancestral ranges were estimated using the Dispersal-Extinction-Cladogenesis (DEC-Lagrange) Effects of some a priori parameters of DEC were alsotested Distribution ranged from central Peruacute to southern Argentina including areasat sea level up to the high Andes The L alticolor-bibronii group was recovered asmonophyletic formed by two clades L walkeri and L gracilis the latter can be splitin two groups Additionally many species candidates were recognizedWe estimate thatthe L alticolor-bibronii group diversified 145Myr ago during theMiddleMioceneOurresults suggest that the ancestor of the Liolaemus alticolor-bibronii groupwas distributedin a wide area including Patagonia and Puna highlands The speciation pattern followsthe South-North Diversification Hypothesis following the Andean uplift

Subjects Biogeography Taxonomy ZoologyKeywords Andean uplift Ancestral range Lizards Liolaemus Total evidence

INTRODUCTIONThe genus Liolaemus WIEGMAN 1834 currently includes 262 species (updated from(Abdala Quinteros amp Semham 2015) distributed between Tierra del Fuego in southernArgentina and northern central Peru The taxonomic composition and phylogeneticrelationships of this genus have varied over the years Laurent (1983) proposed dividingit into two main groups (subgenera) Liolaemus sensu stricto (or Chileno group) andEulaemus (or Argentino group) Furthermore taxonomic studies on these two groupshave led to the recognition of numerous subgroups Recently Troncoso-Palacios et al(2015) proposed the division of the Liolaemus sensu stricto subgenus into two mainsections the L chiliensis section and the L nigromaculatus section The L alticolor-bibroniigroup was attributed to the Liolaemus sensu stricto subgenus (Espinoza Wiens amp Tracy

How to cite this article Portelli and Quinteros (2018) Phylogeny time divergence and historical biogeography of the South AmericanLiolaemus alticolor-bibronii group (Iguania Liolaemidae) PeerJ 6e4404 DOI 107717peerj4404

2004 Lobo 2005 Quinteros 2012 Quinteros 2013) and the L chiliensis section (Troncoso-Palacios et al 2015 Panzera et al 2017) The group was first defined byOrtiz (1981) as theL alticolor-walkeri group containing three species (L alticolor BARBOUR 1909 L walkeriSHREVE 1938 and L tacnae SHREVE 1941) whereas the L bibronii group was definedby Cei (1986) consisting of L bibronii BELL 1843 L exploratorum CEI amp WILLIAMS1984 and L sanjuanensis CEI 1982 The taxonomic composition of these groups increasedand varied in last years Depending on the criteria methods and dataset used specieswere assigned to either more specific groups (L alticolor or L bibronii groupmdash (Lobo ampEspinoza 1999 and Martinez Oliver amp Lobo 2002) or more inclusive groups such as theL alticolor-bibronii group (Espinoza Wiens amp Tracy 2004 Lobo 2005 Quinteros 2012Aguilar et al 2013) In a contribution made by Quinteros (2012) L alticolor group wasre-described including two new species resulting in a total number of 27 species Morerecently the number of species was corrected to 30 (Martiacutenez et al 2011 Aguilar et al2013 Quinteros et al 2014 Abdala Quinteros amp Semham 2015) Morando et al (2007)performed a phylogeographic study on several populations of L bibronii and L gracilisBELL 1843 concluding that L bibronii is in fact a species complex with many candidatespecies (some of them described by Martiacutenez et al (2011) Quinteros (2012) Quinteroset al (2014) and Abdala Quinteros amp Semham (2015)) Using morphological charactersQuinteros (2013) recovered the L alticolor-bibronii group as monophyletic and foundsister group relationships to L gravenhorsti Furthermore Aguilar et al (2013) describedthree new species from Peru and performed a phylogenetic analysis of local species ofthe L alticolor-bibronii group In summary the L alticolor-bibronii group is composedof 30 species widely distributed in an outstanding altitudinal gradient from sea level tomore than 5000 masl (Abdala amp Quinteros 2014) between Santa Cruz Province (southernArgentina) Chile and Ancash (central Peru) (Quinteros 2012 Quinteros 2013 Aguilar etal 2013) Due to these traits the L alticolor-bibronii group is an excellent model to inferthe historical events which possibly molded the current distribution of its members

Until now there have not been any proposals about the origin or diversification of theL alticolor-bibronii group based on phylogenies or quantitative frameworks studying itsbiogeography There are related studies such as Cei (1979) who characterized Patagoniaas an active center of origin and dispersion including Liolaemus as an example of arecent adaptive radiation Other authors applied phylogenies but failed to use explicitbiogeographical methodology Lobo (2001) showed a cladogram of the chiliensis group(Liolaemus sensus stricto subgenus) including distribution areas of the species These areasare similar to those outlined by Roig Juntildeent (1994) adding some but not including otherssuch as the Puna Young-Downey (1998) performed a Brooks Parsimony Analysis (BPABrooks 1990) over a Liolaemus phylogeny while Schulte et al (2000) predicted speciesdistribution based on molecular phylogeny Even though no explicit methodology wasincluded the latter study can be considered an ancestral areas analysis More details onthe biogeography of the of the Liolaemus sensu stricto subgenus are given by Diacuteaz Goacutemez ampLobo (2006) who applied three methods (Fitch optimization DIVAmdashRonquist 1997 andWeighted Ancestral Areas AnalysismdashHausdorf 1998) to reconstruct the ancestral area ofthe subgenus

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 230

In this study we perform phylogenetic analyses in search of patterns which couldhave molded the current distribution of the species We use quantitative and explicitmethodology and estimate time divergence to investigate the historical biogeography ofthe Liolaemus alticolor-bibronii group

MATERIALS AND METHODSTaxon sampling and characters usedWe included 30 species of the Liolaemus alticolor-bibronii group taken primarily fromQuinteros (2013) with recent species additions by Martiacutenez et al (2011) Aguilar et al(2013) Quinteros et al (2014) and Abdala Quinteros amp Semham (2015) 15 populations ofuncertain taxonomic status and eight outgroup species (increasing in nine the number ofterminal taxa included by Quinteros 2013) See File S1 for details

The phylogenetic analyses included typical morphological characters (Lobo 2001Lobo 2005 Quinteros 2012 Quinteros 2013) and sequences of Cytb and 12S genes takenfrom the literature (Espinoza Wiens amp Tracy 2004 Morando et al 2007 Victoriano et al2008 Aguilar et al 2013 Troncoso-Palacios et al 2016 Olave et al 2011 see File S1 forspecimensrsquo GenBank accession numbers) Sequences were aligned and edited with MEGAv7026 (Kumar Stecher amp Tamura 2016) Morphological characters were coded intodiscrete and continuous based on their variation Discrete characters were coded as BinaryPolymorphic Binary Multistate and Polymorphic Multistate while continuous characterswere coded lsquolsquoas suchrsquorsquo following Goloboff Mattoni amp Quinteros (2006) methodology

We included 960 bp from 12S 809 bp from Cytb and 167 morphological charactersupdating the list by Quinteros (2013) see File S2 for details

Phylogenetic analysesWe conducted Maximum Parsimony (MP) and Bayesian Inference (BI) analyses usingmatrices including morphology + Cytb + 12S characters Alternative analyses excludingmorphology andor continuous characters were also performed

Maximum Parsimony analyses were implemented in TNT v11 (Goloboff Farris ampNixon 2003 Goloboff Farris amp Nixon 2008 Goloboff amp Catalano 2016) under equal andimplied weights (Concavity value= 3 4 5 and 6) Continuous characters were included inMP analyses only since TNT is the only software supporting them Runs were performedusing traditional search Tree Bisection Reconnection (TBR) with 500 replicates and saving20 trees each Additionally we performed an analysis using the New Technology Searchtool implemented in TNT (Sectorial Search Tree Fusing Tree Drifting and Ratchet)with 50 replicates hitting the best score at least 20 times Support was measured underSymmetric Resampling with 500 replicates and a 33 deletion

For BI we selected the best-fitting model using jModel Test 304 (Posada 2008) Thebest-fitting model for CytB and 12S individually was GTR+G whereas GTR (GTR+0+I)fitted best for the concatenate genes Bayesian Inference analyses were carried out in MrBayes v31 (Ronquist amp Huelsenbeck 2003) We run PartitionFinder v111 (Lanfear et al2012) to detect the best partition scheme including both genes (without partition matrix)Calculations were run twice for 10 million generations each resulting in a final average

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 330

standard deviation of Split frequencies below 005 sampling trees every 1000 generationsand using four simultaneous chains (one cold and three hot) in each run Convergencesof the chain to stationary distribution were confirmed using Tracer v16 (Rambaut etal 2014) We discarded as burn-in the first one thousand sampled trees that were notwithin the stationary distribution of log likelihoods Trees and posterior probabilities weresummarized using the lsquolsquo50 majority rulersquorsquo consensus method

Time divergence estimatesAge of nodes and substitution rates were simultaneously estimated (for both topologies BIand MP) using Bayesian MCMC (Marcov Chain Monte Carlo) approach as implementedin BEAST v240 (Drummond amp Rambaut 2007) We used a fossil assigned to Eulaemus(Albino 2008) representing the earliest record of this subgenus to place a mean prior of20 Myr on the tree height Since the fossil was assigned to the Eulaemus subgenus withoutspecific status in our study we must place it as an outgroup taxon sister of the speciesmembers of the Liolaemus sensus stricto subgenus which are the focal species Divergencetimes in BEAST were estimated according to Fontanella et al (2012) (1) a lognormal priorwas employed for fossil calibrations (Hedges amp Kumar 2004) (2) a Yule speciation processwith a random starting tree was used for the tree prior (3) an uncorrelated lognormaldistributed relaxed clock (UCLD) model was employed allowing evolutionary ratesto vary along branches within lognormal distributions (Drummond et al 2006) Threeindependent runs of 50000000 generations each were performed with sampling every5000 generations The three separate runs were then combined (following removal of 10burn-in) using Log Combiner v20 (Drummond amp Rambaut 2007) Adequate samplingand convergence of the chain to stationary distribution were confirmed by inspection ofMCMC samples using Tracer v16 (Rambaut et al 2014) The effective sample size (ESS)values of all parameters were greater than 200 which was considered a sufficient levelof sampling The sampled posterior trees were summarized using Tree Annotator v20(Drummond amp Rambaut 2007) to generate a maximum clade credibility tree (maximumposterior probabilities) and to calculate the mean ages 95 highest posterior density(HPD) intervals posterior probabilities and substitution rates for each node The BEASTtopology was visualized with Fig Tree v12 (Rambaut 2008)

Biogeographical analysesIn order to assign the ancestral geographic areas of distribution the regionalization ofSouth America according to Morrone (2001) was used Briefly Morrone (2001) detailedbiogeographic regions sub regions and provinces of Latin America and the Caribbeanpointing out characteric taxa and their predominant vegetation Even though those areasare defined by contemporary species they are a useful methodological tool to avoid the useof randomly defined areas We chose the following provinces based on the georeferenceddistribution of the species included (Fig 1) A Desierto Peruano Costero B Puna CYungas D Atacama E Coquimbo F Prepuna G Monte H Chaco I Pampa J SantiagoK Maule and L Patagonia Central To infer the processes which modeled the currentdistribution of the species of the L alticolor-bibronii group we used a parametric method

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 430

Figure 1 Map of South America showing the biogeographic regions used Biogeographic regions ofMorrone (2001) employed in DEC analyses (A) Desierto Peruano Costero (B) Puna (C) Yungas (D)Atacama (E) Coquimbo (F) Prepuna (G) Monte (H) Chaco (I) Pampa (J) Santiago (K) Maule (L)Patagonia Central and (M) Patagonia Subandina

Full-size DOI 107717peerj4404fig-1

implemented in RASP (Nylander et al 2008 Yu Harris amp He 2010) This method is basedon Dispersal-Extinction-Cladogenesis (DEC) models which require information about asingle ultrametric-dated phylogeny and distributional information of extant species

For DEC we used the two calibrated trees recovered in BEAST (based on BI and MPtopologies) and applied two different adjacency matrices (1) unconstrained and (2)

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 530

Table 1 Probabilities of dispersal cost Probabilities between the areas (AndashL) employed in the DEC anal-ysis under Matrix II See lsquoMaterials and Methodsrsquo for area names

A B C D E F G H I J K L

A 1 08 06 08 06 04 06 06 04 04 02 04B 08 1 08 08 08 08 08 08 06 06 04 06C 06 08 1 06 06 06 06 08 06 04 02 04D 08 08 06 1 08 06 06 06 04 06 04 04E 06 08 06 08 1 08 06 04 04 06 06 06F 04 08 06 06 08 1 08 06 06 08 06 08G 04 08 06 06 06 08 1 08 08 06 06 08H 02 08 08 06 04 06 08 1 08 04 04 06I 02 06 06 04 04 06 08 08 1 04 06 08J 04 06 04 06 06 08 06 04 04 1 08 08K 02 04 02 04 06 06 06 04 06 08 1 08L 04 06 04 04 06 08 08 06 08 08 08 1

Table 2 Probabilities of dispersal cost Probabilities between the areas (AndashL) employed in the DEC anal-ysis under Matrix III See lsquoMaterials and Methodsrsquo for area names

A B C D E F G H I J K L

A 1 0001 0001 08 06 0001 0001 0001 0001 04 02 0001B 0001 1 08 0001 0001 08 08 08 06 0001 0001 06C 0001 08 1 0001 0001 06 06 08 06 0001 0001 04D 08 0001 0001 1 08 0001 0001 0001 0001 06 04 0001E 06 0001 0001 08 1 0001 0001 0001 0001 08 06 0001F 0001 08 06 0001 0001 1 08 06 06 0001 0001 08G 0001 08 06 0001 0001 08 1 08 08 0001 0001 08H 0001 08 08 0001 0001 06 08 1 08 0001 0001 06I 0001 06 06 0001 0001 06 08 008 1 0001 0001 08J 04 0001 0001 06 08 0001 0001 0001 0001 1 08 0001K 02 0001 0001 04 06 0001 0001 0001 0001 08 1 0001L 0001 06 06 0001 0001 08 08 06 08 0001 0001 1

constrained (See lsquoDiscussionrsquo for details) We also included three different lsquolsquoarea-dispersalrsquorsquomatrices a first analysis (Matrix I) was set to allow a dispersal event between everygeographic area without any cost a second analysis (Matrix II) was set to apply dispersalcost where we randomly assigned probability values of dispersion between areas takinginto account the closeness between them The values were 08 for adjacent areas 06 forareas separated by one intermediate area 04 for areas separated by two intermediate areas02 for areas separated by three intermediate areas and 008 for areas separated by fourintermediate areas (Table 1) A third analysis (Matrix III) was set to apply dispersal costplus geographic barriers cost considering probability values from the second analysis plusvalues assigned to geographic barriers (in this case the only clear geographical barrier wasthe Andes mountain range) When a geographic barrier was present a probability value ofclose to 0 was assigned (0001mdashTable 2)

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 630

Figure 2 Congruence between theMaximum Parsimony (MP) and Bayesian Inference (BI) analyses Some members of the L lemniscatus clade is recovered as a sister group of the L walkeri clade under MPNumbers above nodes correspond to Posterior Probabilities (Only Probabilities over 095 are shown)Numbers under nodes correspond to Symmetric Resampling values

Full-size DOI 107717peerj4404fig-2

RESULTSPhylogeny of the Liolaemus alticolor-bibronii groupThe analyses showed two topologies one recovered under MP and the other under BIThe topologies found are congruent in the main groups (SPR distance 06277mdashFig 2)however we also detected incongruence within them

The Liolaemus alticolor-bibronii group was recovered in both MP and BI analyses withdifferent composition and with high support (Figs 3 and 4) In the BI tree a clade formedby the L gravenhorsti group and the L lemniscatus group was recovered as sister of theL alticolor-bibronii group On the other hand the MP analysis recovered the L lemniscatusgroup as a member of the L alticolor-group but some terminal taxa were not recovered asmembers despite their previously assignment to the L alticolor-bibronii group (L paulinaeDONOSO BARROS 1961 L sp4 L sp5 and L sp14)

Within the Liolaemus alticolor-bibronii group two main groups were recovered theL gracilis and L walkeri group (Figs 2ndash4) The L walkeri group formed by terminal taxaof northern distribution ranges (North Argentina Bolivia and Peru) was recovered withhigh support in BI but moderate MP support The L gracilis group on the other hand wasrecovered with high support in MP but low support in BI The latter group contains twoclades (1) a clade formed by species distributed in central-southern Argentina (referred toas L bibronii sensu stricto clade) and (2) the L robertmertensi group

The taxonomic composition of each group recovered under MP and BI is listed inTable 3

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 730

Figure 3 Topology recovered with Bayesian InferenceNumbers above nodes correspond to PosteriorProbability values (only values over 095 are shown)

Full-size DOI 107717peerj4404fig-3

As we mentioned above Liolaemus paulinae previously considered member of theL alticolor-bibronii group was not recovered as a member of the group The same topologywas recovered for L sp14 (sister taxon of L paulinae) Liolaemus araucaniensis MUumlLLERamp HELLMICH 1932 did not show relations to any group suggesting this species could berelated to the L belli group

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 830

Figure 4 Topology recovered withMaximum ParsimonyNumbers above nodes correspond to Sym-metric Resampling values (only values over 60 are shown) Note that the L lemniscatus is paraphyletic andsome species of this group are nested inside the L walkeri clade

Full-size DOI 107717peerj4404fig-4

Divergence time estimatesDivergence time estimates of the Liolaemus alticolor-bibronii group were obtained using atime-calibrated tree with BEAST including MP and BI topologies (Figs 5 and 6)

DatingThe two topologies resulted in two time calibrated trees with similar times of divergenceOur results suggest that according to the BI topology the Liolaemus gravenhorsti group

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 930

Table 3 Taxonomic composition of the main groups Species and populations members of the main groups recovered inside the Liolaemusalticolor-bibronii group under Maximum Parsimony and Bayesian Inference

Group Maximum Parsimony Bayesian Inference

L gravenhorsti L chiliensis L cyanogaster L gravenhorsti L schroederi L chiliensis L cyanogaster L exploratorum L gravenhorstiL paulinae L schroederi

L lemniscatus L fuscus L lemniscatus L curicencis L fuscus L lemniscatus L nitidusL pseudolemniscatus

L alticolor-bibronii L gracilis group L walkeri group L gracilis group L walkeri groupL walkeri L alticolor L chavin L incaicus L lemniscatus group

L pachacutec L sp3 L tacnae L walkeri L wariL alticolor L chaltin L chavin L incaicus L pachacutecL puna L pyriphlogos

L gracilis L bibronii sensu stricto group L robertmertensi group L bibronii sensu stricto group L robertmertensi group L sp2L sp4 L sp5

L bibronii sensustricto

L abdalai L aparicioi L bibronii L cyaneinotatusL exploratorum L gracilis L gracilis3 L gracilis4L pyriphlogos L ramirezae 1 L ramirezae2L sanjuanensis L saxatilis L sp6 L sp7 L sp8 L sp9L sp10 L sp15 L tandiliensis L variegatus L yalguaraz

L abdalai L bibronii L chungara L cyaneinotatus L sp6L sp7 L sp8 L sp9 L sp10 L sp13 L yalguaraz

L robertmertensi L bitaeniatus L chaltin L chungara L pagaburoi L punaL ramirezae 3 L robertmertensi L sp1 L sp2 L sp11L sp12 L sp13 L walkeri 5 L yanalcu

L aparicioi L bitaeniatus L gracilis L pagaburoiL ramirezae L robertmertensi L sanjuanensis L saxatilisL sp11 L sp12 L sp15 L tandiliensis L variegatusL yanalcu

(sister group of the L alticolor-bibronii group) diverged in the Early Miocene 7 Myr (95highest posterior density interval (HPD) 932ndash489) ago whereas in the MP topologythis clade diverged 325 Myr (95 HPD 482ndash171) ago This clade is formed by speciesdistributed in Argentina and Chile

We estimated the divergence time of the Liolaemus alticolor-bibronii group to MiddleMiocene around 1284 Myr ago for BI (95 HPD 2001ndash564) and 1405 Myr ago for MP(95 HPD 2356ndash454)

Considering the clades of the Liolaemus alticolor-bibronii group we estimated that theL gracilis clade diverged 1049 Myr (95 HPD 1706ndash385) ago for BI being 1239 Myr(95 HPD 1742ndash725) for MP This clade splits into two clades the L bibronii sensustricto clade and the L robertmertensi group The L bibronii sensu stricto clade initiated itsdivergence 468 Myr (95 HPD 725ndash211) ago in BI whereas for MP this clade diverged982 Myr (95 HPD 1185ndash578) ago The L robertmertensi group does so 937 Myr (95HPD 1568ndash301) ago in BI and 42 Myr (95HPD 684ndash167) ago for MP topology Thelatter originated two clades one distributed in central-southern Argentina which diverged675 Myr (95HPD 1029ndash323) ago and the other distributed in Northwestern Argentinaand Bolivia diverging 562 Myr (95HPD 745ndash381) ago these clades are recovered onlyunder BI

The Liolaemus walkeri clade dates back to the Late Miocene around 1154 Myr (95HPD 2026ndash28) ago in our BI topology 1278 Myr (95 HPD 1965ndash587) ago inMP Liolaemus walkeri group includes a clade distributed exclusively in Peru (with anorigin estimated at 1038 Myrmdash95 HPD 1815ndash255 in BI and 1072 Myrmdash95 HPD1932ndash212 forMP) as well as a clade distributed mainly in Bolivia (origin estimated around245 Myrmdash95 HPD 426ndash062mdashrecovered only with BI) While the topology recovered

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 1030

Figure 5 Times of divergences estimates for the L alticolor-bibronii group under BI topology Ultrametric tree scaled in Myr Numbers and hor-izontal bars on nodes represent posterior probabilities values and 95 credibility intervals (A) L bibronii sensu stricto group (B) L robertmertensigroup (C) L gracilis group (D) L walkeri clade C+ D L alticolor-bibronii group (E) L gravenhorsti group (F) L lemniscatus group

Full-size DOI 107717peerj4404fig-5

under MP L lemniscatus diverged 315 Myr (95 HPD 505ndash123) ago from the L walkericlade the BI topology shows the L lemniscatus group as a member of the L gravenhorstigroup (outside of the L alticolor-bibronii group) diverging 35 Myr (95 HPD 585ndash122)ago

Reconstruction of ancestral distributionWe compared six different scenarios in DEC for each topology The best Likelihood scoreswere obtained with the unconstrained adjacency matrix (Table 4) Based on this we showthe results recovered with the unconstrained adjacency matrix for both topologies (BI and

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 1130

Figure 6 Times of divergences estimates for the L alticolor-bibronii group under MP topology Ultrametric tree scaled in Myr Numbersand horizontal bars on nodes represent posterior probabilities values and 95 credibility intervals (A) L bibronii sensu stricto group (B)L robertmertensi group (C) L gracilis group (D) L walkeri clade C+ D L alticolor-bibronii group (E) L gravenhorsti group

Full-size DOI 107717peerj4404fig-6

MP) We obtained the same biogeographic scenarios for all three dispersion cost matricesused (Figs 7 and 8) Alternative scenarios obtained with the constrained adjacency matrixcan be found in File S3

Divergence times ancestral range and major probabilities of the main groups recoveredare shown in Table 5

The ancestral area of the Liolaemus alticolor-bibronii and the L gravenhorsti groupscorrespond to the range BJ (Puna and Santiago p (1) and FJ (Prepuna and Santiago p065) for the BI and MP topology respectively Due to dispersion andor vicariance eventsthe L alticolor-bibronii group occupies areas in Argentina whereas the L gravenhorstiremains confined to Chile

Our results suggest that the ancestral area of the Liolaemus alticolor-bibronii group isformed by the areas BL (Puna and Patagonia Central p 067ndashBI) and F (Prepuna p041ndashMP) Subsequently a vicariance event for BI and a dispersal event (MP) originatedthe L gracilis group and the L walkeri clade With lower probabilities the analyses showsthe following ancestral areas for the L alticolor-bibronii group B (Puna p 016) for BIand BF (Puna and Prepuna p 026) for MP the range BH (Puna and Chaco p 017) forBI and FJ (Prepuna and Santiago p 023) or FL (Prepuna and Patagonia Central p 009)for MP

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 1230

Table 4 Comparison between the results obtained in DEC applying two different adjacency matricesNumber of alternative ancestral areas obtained from the main groups applying constrained and uncon-strained adjacency matrix MI MII and MIII correspond to dispersion coast and geographic barriers ma-trix (see Materials and Methods for details) Global maximum Likelihood scores (minuslnL) and dispersal andextinction rates estimated

Global dispersal Global vicariance Global extinction

Bayesian InferenceMatrix I 33 10 0Matrix II 39 23 8

Constrainedadjancencymatrix Matrix III 46 22 12

Matrix I 31 16 0Matrix II 36 16 2

Unconstrainedadjancencymatrix Matrix III 31 16 0Parsimony

Matrix I 37 18 1Matrix II 35 21 6

Constrainedadjancencymatrix Matrix III 37 22 6

Matrix I 37 16 0Matrix II 36 21 6

Unconstrainedadjancencymatrix Matrix III 37 20 6

minuslnL Dispersal rates Extinction rates

Bayesian InferenceMatrix I 15641 0024 0013Matrix II 15422 0015 0018

Constrainedadjancencymatrix Matrix III 16132 0022 0020

Matrix I 14791 0012 0002Matrix II 14747 0009 0008

Unconstrainedadjancencymatrix Matrix III 14554 0015 0021Parsimony

Matrix I 19625 0011 0013Matrix II 16796 0021 0026

Constrainedadjancencymatrix Matrix III 17670 0035 0025

Matrix I 16575 0004 0012Matrix II 16339 0008 0014

Unconstrainedadjancencymatrix Matrix III 15728 0002 0021

The Liolaemus gracilis group has its origins in L (Patagonia Central p 05) for bothtopologies A dispersion event inside the L gracilis group originated the L robertmertensiand the L bibronii groups The first one shows theHL range (Chaco and Patagonia Centralp (1) as its ancestral area for BI and area B (Puna p 036) for MP whereas the L bibroniisensu stricto remains in L Patagonia Central (p= 1 BI p= 057 MP)

The Liolaemus walkeri group has B (Puna p 1) as its ancestral area for BI and BF (Punaand Prepuna p 087) for MP After a dispersion event (BI) and a vicariance event (MP)the L walkeri group split into two clades One occupied the range BC (Puna and Yungasp 084) in BI and B (Puna p 1) in MP and the other remained in B (Puna p 068) for

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 1330

Figure 7 Ancestral area of distribution (unconstrained adjacency matrix-BI topology) The analysis was run by randomly dividing the entiretime span of evolution of the Liolaemus alticolor- bibronii group in three periods (18ndash55 Myr 55ndash125 Myr 125 Myr-present) Pie charts of eachnode depict the relative probabilities of ancestral arearanges (showed in Legend) Letters above nodes represent ancestral ranges Circles around piecharts represent events blue circle dispersal event green circle vicariance See material and methods for area names shown in South America mapTime axis (in Myr) is annotated with major geological events

Full-size DOI 107717peerj4404fig-7

both topologies In the topology recovered under MP the L walkeri clade includes theL lemniscatus group whose ancestral area corresponds to the EF range (Coquimbo andPrepuna p 1) but after an extinction event is currently restricted to E (Coquimbo)

DISCUSSIONPhylogeny of the Liolaemus alticolor-bibronii groupWe recovered the Liolaemus gravenhorsti group as sister taxon of the Liolaemus alticolor-bibronii group These results are in accordance with previous morphologicalmdash (Lobo 2005Diacuteaz Goacutemez amp Lobo 2006 Quinteros 2013) and molecularmdash(Schulte et al 2000 Schulte2013 Pyron Burbrink amp Wiens 2013 Zheng amp Wiens 2016) based phylogenies

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 1430

Figure 8 Ancestral area of distribution (unconstrained adjacency matrix-MP topology) The analysis was run by randomly dividing the entiretime span of evolution of the Liolaemus alticolor- bibronii group in three periods (18ndash55 Myr 55ndash125 Myr 125 Myr-present) Pie charts of eachnode depict the relative probabilities of ancestral arearanges (showed in Legend) Letters above nodes represent ancestral ranges Circles around piecharts represent events blue circle dispersal event green circle vicariance See material and methods for area names shown in South America mapTime axis (in Myr) is annotated with major geological events

Full-size DOI 107717peerj4404fig-8

Relationships within the Liolaemus alticolor-bibronii group were previously studiedby Quinteros (2013) and Aguilar et al (2013) In his research Quinteros (2013) includes33 terminal taxa which belong to the L alticolor-bibronii group performing a moreinclusive study On the other hand the molecular-based phylogeny by Aguilar et al (2013)was focused on Peruvian terminal taxa (including eight additional taxa members of theL alticolor-bibronii group) Aguilar et al (2013) described the L alticolor-bibronii group asparaphyletic (see below) Morando et al (2004) and Martiacutenez et al (2011) performedphylogeographic analyses focusing on populations of L bibronii and L gracilis andL bibronii respectively The relationships found in the present study are similar to thosedetected byMorando et al (2004) andMartiacutenez et al (2011) despite the higher number ofterminal taxa included

The topology recovered by Quinteros (2013) showed low support and low resolutionin the clades inside the Liolaemus alticolor-bibronii group Nevertheless we agree with

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 1530

Table 5 Mean node ages (Ma) and 95 highest posterior density interval (HPD) obtained for the Lalticolor-bibronii group and the main clades inside The ancestral areas inferred with a probability ge 05with the unconstrained adjacency matrix are also shown See Fig 5 for location of groups

Bayesian Inference Maximum Parsiomny

Group Age (95HPD)

AncestralArearange

Age (95HPD)

AncestralArearange

Ancestor of the L gravenhorstiand alticolor-bibronii groups

1402 Myr(1820ndash105)

BJ 1 1525 (2008ndash950)

FJ 065

L alticolor-bibronii 1284 Myr(2001ndash564)

BL 067 1405 Myr(2356ndash454)

F 041

L walkeri 1154 Myr(2026ndash28)

B 1 1278 Myr(1965ndash587)

BF 087

L gracilis 1049 Myr(1706ndash385)

L 05 1239 Myr(1742ndash725)

L 05

L robertmertensi 937 Myr(1568ndash301)

HL 1 42 Myr(684ndash167)

B 036

L bibronii sensu stricto 468 Myr(725ndash211)

L 1 982 Myr(1185ndash578)

L 057

(Quinteros 2013) who detected the L lemniscatus and the L robertmertensi groups nestedinside the L alticolor-bibronii group (Parsimony tree Fig 3) Previous studies found bothgroups outside of the L alticolor-bibronii group (Cei 1986 Cei 1993 Lobo 2001 Lobo2005 Diacuteaz Goacutemez amp Lobo 2006) Schulte et al (2000) Schulte (2013) Pyron Burbrinkamp Wiens (2013) while Zheng amp Wiens (2016) found L robertmertensi settled inside theL alticolor-bibronii group but recovered the L lemniscatus group outside Our BI treeis congruent with the latter topologies The L alticolor-bibronii group contains in twomain clades (the L gracilis and L walkeri groups) which was not recovered previouslyReviewing the literature the most similar topology is described by Aguilar et al (2013)which recovered a Peruvian clade (similar to our L walkeri clade) a clade with speciesmembers of the L alticolor-bibronii group and of the L monticola and L pictus groups

The Liolaemus walkeri group recovered here shows the same composition as the Peruvianclade ofAguilar et al (2013) as well as species distributed in Bolivia and northern Argentina(Figs 2 and 3) As we brieflymentioned above themain differences between our parsimonyand BI trees is the location of the L lemniscatus group Using parsimony this group isrecovered inside the L alticolor-bibronii group and nested within the L walkeri groupwhereas in the BI tree it is located outside the L alticolor-bibronii group as a sister ofthe L gravenhorsti group The topology recovered under BI is more congruent regardingspecies distribution since the species of the L lemniscatus group are usually found incentral Argentina and Chile (as are the species members of the L gravenhorsti group)The species of the L walkeri group on the other hand are distributed in northwesternArgentina Bolivia and Peru

The Liolaemus gracilis group has not been previously proposed as such Again the BItree shows more biogeographical congruence than the parsimony tree Nevertheless thetwo main clades (L bibronii sensu stricto and L robertmertensi groups) which form theL gracilis group are recovered in both analyses In addition the BI tree recovers L gracilis

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 1630

L robertmertensi L sanjuanensis L saxatilis and L tandiliensis as closely related (as inMorando et al 2004 Martiacutenez et al 2011 Aguilar et al 2013 Pyron Burbrink amp Wiens2013 Zheng amp Wiens 2016)

The greater congruence between the BI and previous studies compared to the parsimonytree can be explained by similarities of the dataset given that they share the same genesequences (Morando et al 2004 Martiacutenez et al 2011 Aguilar et al 2013)

The Liolaemus lemniscatus group was recovered inside the Liolaemus alticolor-bibroniigroup (nested in the L walkeri group) under MP but outside (sister of the L gravenhorstigroup) under BI In previous molecular-based phylogenies members of the L lemniscatusgroup were foundmore related to speciesmembers of the L nigroviridis andor L monticolagroup (Panzera et al 2017 Pyron Burbrink amp Wiens 2013 Schulte et al 2000 Schulte2013) Moreover Pyron Burbrink amp Wiens (2013) Schulte et al (2000) and Schulte (2013)recover the L alticolor-bibronii group as sister of the L gravenhorsti group (L chiliensisgroup of Schulte 2013) with both more closely related to the L elongatus-kriegi complexesthan to the L lemniscatus and L nigroviridis groups In our analyses under BI (excludingmorphology data) the L lemniscatus group is paraphyletic and is recovered outside theL alticolor-bibronii group Unfortunately we cannot include species members of theL nigroviridis andor the L monticola groups in the present study This will be are searchavenue to further investigate in future studies

Taxonomic implicationsWe included 15 populations of uncertain taxonomic status These populations are currentlyassigned to known species members of the L alticolor-bibronii group but differ in somemorphological character states According to our study (BI and Parsimony trees) most ofthem can be considered candidate species Liolaemus sp4 and L sp5 are morphologicallyclose to L araucaniensis In both trees BI and Parsimony L sp4 and L sp5 are not closelyrelated to L araucaniensis (which appears basal) This result in addition to the varyingcharacter states converts these terminal taxa into candidate species Liolaemus bibroniiwas characterized as a species complex byMorando et al (2007)Martiacutenez et al (2011) andQuinteros (2013) Over the last few years three species previously confused with L bibroniihave been described L cyaneinotatus (Martiacutenez et al 2011) L abdalai (Quinteros 2012)and L yalguaraz (Abdala Quinteros amp Semham 2015) In this study we include L sp6L sp7 L sp8 L sp9 L sp10 and L sp15 all of which are assigned to L bibronii Some ofthese terminal taxa were included in the phylogeographic study of Morando et al (2007)Our results are consistent with those of Morando et al (2007) Martiacutenez et al (2011)and Quinteros (2013) concluding that despite the newly described species L bibronii stillcorresponds to a complex of species with several candidate species Liolaemus robertmertensiwas described by Hellmich (1964) but since its description no further taxonomic studiesrelated to it In this study we included three populations attributed to L robertmertensi Lsp11 L sp12 and L sp13 Liolaemus sp12 has previously been included in molecular-basedphylogenies (Schulte et al 2000 Espinoza Wiens amp Tracy 2004) whereas L sp11 and L s13have only been included in a morphology-based phylogeny as L robertmertensi (Quinteros2013) In both analyses (Bayesian and Parsimony) performed here these terminal taxa are

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 1730

not closely related to L robertmertensi while still being members of the L robertmertensigroup Liolaemus sp14 corresponds to a population distributed near L tandiliensis Thisterminal taxon was discovered only recently and is currently under description

Inclusion of morphological and continuous charactersin the analysesWe performed MP and BI analyses excluding morphological data (Figs S4 and S5)Nevertheless similar main groups are recovered UnderMP the Liolaemus alticolor-bibroniigroup is paraphyletic and includes members of the L gravenhorsti group BI topologies aresimilar to those detected under MP The species members of the L gravenhorsti group arenested inside the L alticolor-bibronii group Also main clades of the L alticolor-bibroniigroup are recovered yet the relationships inside the L gracilis group are not resolvedThe L lemniscatus group is recovered paraphyletic Many authors mention homoplasticmorphology (Alvarez et al 1999 Escobar Garciacutea et al 2009 Mott amp Vieites 2009 Muelleret al 2004 among others) but there is also evidence of highly homoplasious mtDNA data(Engstrom Shaffer amp McCord 2004) Jarvis et al (2014) performed a phylogenetic analysisin a very broad scale of major orders of birds 418 million base pairs (14536 loci) finding atopology with high values of bootstrap support Gene trees (of every individual locus) in thesame paper found that none of them was congruent to the species tree suggested by Hahnamp Nakhleh (2015) Here the inclusion of morphological data increases clade resolutionand support the monophyly of some groups

The study of continuous characters is limited toMP analysis as it can only be performedby TNT software Consequently we tested which role continuous characters play inthe recovered topology by excluding them Without continuous characters the Liolaemusalticolor-bibronii group appears polyphiletyc The topology recovers the main groups underMP and BI but their composition include species members of the L gravenhorsti group (seeFig S6) The use of continuous characters analyzed as such was employed in many previousstudies (Abdala 2007 Aacutelvarez Moyers Areacutevalo amp Verzi 2017 Barrionuevo 2017 BardinRouget amp Cecca 2017 Naacutejera-Cortazar Aacutelvarez Castantildeeda amp De Luna 2015 Quinteros2013 among many others) In the present study continuous characters employed inmorphological analyses assisted to recover the monophyly of the L alticolor-bibronii groupand supported the monophyly of the L gravenhorsti group When comparing the resultsobtained under BI and MP (without continuous characters) we find that continuouscharacters as well as the optimality criterion influence the different topologies

The use of a priori component from DECImplementation of DEC models in Lagrange allows biogeographic analyses to inferbiogeographic events under different probabilities taking into account the distancesamong areas and the presence of geographic barriers In this study we applied twodifferent adjacency matrices and three area-dispersal matrices recovering different generalpatterns of diversification and species distribution in the Liolaemus alticolor-bibronii groupwhich allow inference of its evolutionary history

Chacoacuten amp Renner (2014) explored the influence of the user-defined componentsapplicable in DEC-Lagrange (adjacency matrix area-dispersal matrix) They concluded

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 1830

that results were most influenced by adjacency matrices while the areandashdispersal matrixand dispersal probabilities had negligible effects In this study we applied three differentmatrices with dispersal probabilities Using the unconstrained adjacency matrix theancestral reconstructions were very congruent among each other we found only onedifference between the three matrices applied In their study Chacoacuten amp Renner (2014)found that an unconstrained adjacency matrix returned the best results (higher Likelihoodscore) Contrastingly results by Ree amp Smith (2008) and Clark et al (2008) show a betterdata fit of the constrained matrix In this study we tested both an unconstrained aswell as a constrained adjacency matrix The number of assigned ambiguous ancestralareas possible in each node was higher in the constrained than in unconstrained analyses(Table 4) Furthermore the results obtained using the unconstrained matrix display higherLikelihood scores than the constrained matrix (Table 4) Based on these findings we agreewith Chacoacuten amp Renner (2014) concluding that the adjacency matrix influences the resultsin a DEC-Lagrange study but at the same time when applying a constrained matrix thedispersal probabilities matrix had an effect on our results

Patterns of diversification and biogeographic implicationsThe use of two different phylogenetic topologies to estimate the time of divergences andreconstruction of ancestral distribution returned similar results (see Table 5) Despite thedifferences recovered in the composition of main groups their divergence times werecongruent The exceptions were the Liolaemus robertmertensi and L bibronii sensu strictogroups even though an overlap exists The ancestral area reconstruction also shows somedifferences between the topologies but overall most results display great congruence Themain difference is in the ancestral area of the L alticolor-bibronii group The BI topologyrecovers the range BL (Puna and Patagonia Central) whereas the MP topology recoversthe area F (Prepuna) The first one is congruent with previous studies in Liolaemus (Cei1979 Diacuteaz Goacutemez 2011 Schulte et al 2000)

The fossil-calibrated dating analysis indicates that the initial divergence within Eulaemusoccurred approximately 1808 Myr ago during the Early Miocene which roughlycorresponds to previous studies (Fontanella et al 2012) The origin and diversification ofthe Liolaemus alticolor-bibronii group could have started around the Early-Middle Miocene(13ndash14 Myr ago) which is consistent with results by (Schulte 2013 Medina et al 2014Zheng amp Wiens 2016) During that period many dramatic changes occurred eventuallyleading to the Andes uplift The Central Andes lifted quickly geologically speaking from thelateMiocene to the Early Pliocene (Whitmore amp Prance 1987Gregory-Wodzicki 2000)Theuplift separated the East from the West of the continent and affected the dispersion ofmany taxa and potentially induced the differentiation between populations (Elias et al2009) Our results support previous studies suggesting an influence of the Andean uplifton the diversification of South American taxa (Schulte et al 2000 Hoorn et al 2010Pincheira-Donoso et al 2013)

Regarding its evolution and distribution it can be assumed that the Liolaemus alticolor-bibronii group colonized areas along the lower Andes Our results indicate that the ancestralarea of the group extended from the Patagonian Andes to the northern Andes (Argentina

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 1930

Bolivia and Peruacute) This is in accordance with paleontological evidence the oldest knownfossil of Liolaemus corresponds to Patagonia in the Miocene era at the Gaiman Formationin Chubut Argentina (Albino 2008)

Our findings are also consistent with those of Diacuteaz Goacutemez (2011) whose DIVA(Dispersion-Vicariance-Ronquist 1997) analysis located the Liolaemidae commonancestorrsquos distribution from Peru to Patagonia along the Andes and arid regions ofSouth America Cei (1979) concluded that Patagonia was the center of origin for at leastfour Liolaemus groups describing two main faunal regions of Patagonia (1) septentrionalregion (older Patagonia) and (2) meridional region (Santa Cruz Province) Our resultsagree with Cei (1979)rsquos hypothesis as the first region described in his paper comprisesareas of the Andes and Patagonia included in the distribution areas calculated in thepresent study On the other hand our results disagree with those published by Schulte etal (2000) who found the areas of Andes occidental lower lands and eastern lower landsto be ancestral area of the Liolaemus sensu stricto subgenus which correspond to SierrasPampeanas Maulina and Chile Central in our study

According to Bremer (1992) the ancestral area of a taxon is not necessarily to a singleplace but may be equal or larger than the area presently inhabited by the taxon Resultsof an ancestral area analysis will display a taxonrsquos ancestral distribution range which isequal or smaller than the sum of distribution ranges of its descendants Including a fossilgroup neither distributed in ancestral nor current areas of the studied taxa may result inan ancestral area larger than the sum of the descendantsrsquo distributions In our study weused a fossil record attributed to Liolaemus (Albino 2008) in an area already present in theanalysis

The common ancestor of the clade formed by the Liolaemus alticolor-bibronii groupplus (L gravenhorsti + L lemniscatus groups) had its divergence around 14ndash15 Myr agoThis calibration coincides with the beginning of the Andes uplift (Donato et al 2003)and our DEC analyses (Figs 7ndash8) which suggest that vicariance event could have ledto the current clade distribution in Chile (L gravenhorsti and L lemniscatus groups)Argentina Bolivia and Peru (L alticolor-bibronii group) Because of their geographyorogeny and vast biodiversity the Andes Mountains were a subject of interest of manydistribution and evolution studies (Castroviejo-Fisher et al 2014 Chaves amp Smith 2011Goicoechea et al 2012 Torres-Carvajal et al 2016) Based on a phylogenetic analysis ofProctoporusRiama lizardsDoan amp Schargel (2003) proposed the south-to-north speciationhypothesis (SNSH) which predicts a pattern of cladogenesis of Andean species followingthe rise of the Andes with basal lineages occurring in southern areas and derived ones innorthern areas Our results gained with BI topology agree with the hypothesis of Doan ampSchargel (2003) since the ancestral area recovered for the L alticolor-bibronii group includesPatagonia and the derived groups show a northern distribution On the other hand ourresults obtained with MP topology reject the SNSH since the ancestral range of theL alticolor-bibronii group corresponds to Prepuna and the derived groups are distributedacross both north and south of Prepuna The SNSHwas also rejected for Proctoporus lizardsbased on a phylogenetic analysis with increased character and taxon sampling (Goicoechea

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 2030

et al 2012) as well as for other organisms such as glass frogs (Castroviejo-Fisher et al2014)

The Liolaemus gravenhorsti group has its origins 772 Myr (95 HPD 932ndash621) agomatching the continued Andes uplift At that time (around 62 Myr ago) lakes and fluvialdeposits were affected by a deformation event possibly the final uplift stage reaching thepresent average altitude of 4400 masl (Garzione et al 2008) which may have triggered adispersal or a vicariance event of L chiliensis and L cyanogaster towards Argentina

The divergence time of the Liolaemus alticolor-bibronii group was estimated around12ndash14 Myr ago which corresponds to the formation of the Atacama Desert approximately147 Myr ago based on the lack of accumulation of cupric deposits in northern Chile(Alpers amp Brimhall 1988) Our results of BI topology suggest that following the desertrsquosformation a vicariance event could have split the L alticolor-bibronii group into two mainclades distributed North (L walkeri clade) and South (L gracilis group) of the AtacamaDesert A similar scenario was recovered with MP topology but instead of a vicarianceevent dispersal events led to the L walkeri group to the north and the L gracilis group tothe south

The diversification of the Liolaemus gracilis group approximately 10ndash12 Myr agocoincides with the uplift the Austral Andes (altitude gt4000 masl) which reached theirmaximum elevation around 8ndash12 Myr ago (Miocene-Pliocene) (Hartley 2003) Thisgeological event may have dispersed the species of the L gracilis group giving origin totwo clades one distributed in central-northern Argentina (L robertmertensi group) and aclade distributed in central-southern Argentina (L bibronii sensu stricto clade) SimilarlyCosacov et al (2010) report a phylogeographic break for Calceolaria poliriza in southernMendoza possibly caused by a landscape discontinuity (Ramos amp Kay 2006) during theuplift of the Andes in the Late Miocene (11Myr) Inside the L gracilis group the L bibroniisensu stricto clade originates 5 Myr ago while the grouprsquos more terminal clades divergein the Pleistocene (25 Myr) and are distributed in Patagonia Seacutersic et al (2011) studiedphylogeographical patterns of plants and vertebrates (including some Liolaemus species)from Patagonia Specifically Seacutersic et al (2011) argued that the phylogeographic break ofL bibronii could be concordant with the Rio Limay basin which drains (or that in the pastdrained) the East Andean watershed crossing the Patagonian steppe to the Atlantic OceanOther studies (Morando Aacutevila amp Sites Jr 2003 Morando et al 2007 Aacutevila Morando ampSites Jr 2006) claimed that three species of Liolaemus (among them L bibronii) shared avicariant pattern produced by the coastline shifts along the Atlantic at the northern andsouthern rim of the Somuncuraacute Plateau

The Liolaemus robertmertensi group (recovered under BI) is distributed in SierrasPampeanas (Central and northwestern Argentina) and Sierras Subandinas (northwesternArgentina) A possible explanation for the present distribution of the grouprsquos Astereceaespecies is proposed by Crisci et al (2001) they found a historical pattern that relatedTandilia to Ventania Mahuidas Sierras Pampeanas and Sierras Subandinas to the Westand Uruguay and southern Brazil to the East Crisci et al (2001) hypothesize that theendemism of these mountainous chains is a result of generally arid conditions during theTertiary andor Quaternary geologic periods in southern South America which eventually

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 2130

led to an isolation and differentiation of these populations in the more elevated areasThe ancestral area recovered under MP for the L robertmertensi group is Puna and itsdiversification time is approximately 42 Myr This species group currently inhabits theSierras Subandinas These mountains showed a second deformation cycle at 45 Myr withan out of sequence growth event (Hernaacutendez amp Echavarria 2009) This can explain thevicariance and dispersal events that the species of the group experienced

The Liolaemus walkeri clade started its diversification 11ndash12 Myr ago in an area whichtoday corresponds to the Andean Plateau This plateau rose from 2500 masl to 4000 maslaround 10 Myr ago during the Miocene-Pliocene (Gubbels Isacks amp Farrar 1993 Gregory-Wodzicki 2000 Hartley 2003) under an arid to semiarid climate conditions (Hartley2003) The L walkeri clade is formed by two clades distributed in central and southernPeru Bolivia and northern Argentina This distribution could be due to a dispersal eventfrom the Puna to Yungas (the northern distributed clade) while the other clade remainedin the Puna The clade split follows a similar pattern as the plant genus Distichia whichseparated due to the plateaursquos climate conditions into clades in northern Argentina andChile (towards the South) and Ecuador and Colombia (towards the North) (RamirezHuaroto 2012)

The desertification process from the Miocene to Pliocene may have caused the rangeexpansion of the ancestors of the species members of the Liolaemus alticolor-bibroni groupSubsequently the aridhumid cycles which followed the glacial and interglacial period ofthe Pliocene and Pleistocene produced expansion and retraction of arid and damp habitatsacting as geographic barriers and causing fragmentation and speciation of the extant taxa

In conclusion we describe the possible ancestral area of the Liolaemus alticolor-bibroniigroup as a large area which includes Patagonia and the Puna highlands during the LowerMiocene and the Pleistocene The inclusion of species members of the L monticola and Lnigroviridis groups will probably modify the ancestral area-range of the L alticolor-bibroniigroup

ACKNOWLEDGEMENTSWe are grateful to F Lobo JM Diacuteaz Goacutemez C Abdala M Quipildor S Ruiz T Hibbardand F Arias for their assistance in the field lab andor discussion of ideas related to thisstudy We thank T Hibbard and M Schulze for improving the English style We would liketo thank S Kretzschmar G Scrocchi and E Lavilla (FML) J Williams (MLP) J Faivovichand S Nenda (MACN) JC Acosta A Laspiur and E Sanabria (UNSJ) L Vega (UNdMP) RAguayo and F Valdivia (CBGR) J Aparicio A Kirigin and M Ocampo-Vallivian (CBF) HVoris and A Resetar (FMNH) L Ford and D Frost (AMNH) J Cadle and J Rosado (MCZ)J Wiens and E Censky (CMNH) T Reeder and R Etheridge (SDSU) K De Queiroz and RHeyer (USNM) for allowing access to their collections We thank Willi Heniing Societyfor making TNT freely available We thank two anonymous reviewers whose commentsgreatly improves the MS

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 2230

ADDITIONAL INFORMATION AND DECLARATIONS

FundingThis studywas supported by grants from theAgenciaNacional de InvestigacionesCientificas(PICT-2015-1398) Consejo de Investigaciones de la Universidad Nacional de Salta(CIUNSa 2013) and by a Graduate Fellowship from Consejo Nacional de InvestigacionesCientiacuteficas y Teacutecnicas (S Portelli) The funders had no role in study design data collectionand analysis decision to publish or preparation of the manuscript

Grant DisclosuresThe following grant information was disclosed by the authorsAgencia Nacional de Investigaciones Cientificas PICT-2015-1398Consejo de Investigaciones de la Universidad Nacional de Salta (CIUNSa 2013)Consejo Nacional de Investigaciones Cientiacuteficas y Teacutecnicas

Competing InterestsSebastiaacuten Quinteros is a researcher of IBIGEO-CONICET Associate Editor of Cuadernosof Herpetologiacutea and Secretary of the Asociacioacuten Herpetoloacutegica Argentina Sabrina Portelliis a Doctoral scholar of CONICET

Author Contributionsbull Sabrina N Portelli and Andreacutes S Quinteros conceived and designed the experimentsperformed the experiments analyzed the data contributed reagentsmaterialsanalysistools prepared figures andor tables authored or reviewed drafts of the paper approvedthe final draft

Data AvailabilityThe following information was supplied regarding data availability

The raw data is included in the Supplemental Files

Supplemental InformationSupplemental information for this article can be found online at httpdxdoiorg107717peerj4404supplemental-information

REFERENCESAbdala CS 2007 Phylogeny of the boulengeri group (Iguania Liolaemidae Liolaemus)

based on morphological and molecular characters Zootaxa 15381ndash84Abdala CS Quinteros AS 2014 Los uacuteltimos 30 antildeos de estudios de la familia de

lagartijas maacutes diversa de Argentina Actualizacioacuten taxonoacutemica y sistemaacutetica deLiolaemidae Cuadernos de Herpetologiacutea 28(2)55ndash82

Abdala CS Quinteros AS Semham RV 2015 A new species of Liolaemus of the Liolae-mus alticolor-bibronii Group (Iguania Liolaemidae) from Mendoza Argentina SouthAmerican Journal of Herpetology 10(2)104ndash115 DOI 102994SAJH-D-14-000331

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 2330

Aguilar CWood Jr PL Cusi JC Guzmaacuten A Huari F LundbergM Sites Jr JW 2013Integrative taxonomy and preliminary assessment of species limits in the Liolaemuswalkeri complex (Squamata Liolaemidae) with descriptions of three new speciesfrom Peruacute ZooKeys 36447ndash91 DOI 103897zookeys3646109

Albino AM 2008 Lagartos iguanios del Colhuehuapense (Mioceno Temprano) deGaiman (provincia del Chubut Argentina) Ameghiniana 45775ndash782

Alpers C Brimhall G 1988Middle Miocene climatic change in the Atacama Desertnorthern Chile evidence from supergene mineralization at La Escondida GeologicalSociety of America Bulletin 1001640ndash1656DOI 1011300016-7606(1988)100lt1640MMCCITgt23CO2

Aacutelvarez A Moyers Areacutevalo RL Verzi DH 2017 Diversification patterns and size evolu-tion in caviomorph rodents Biological Journal of the Linnean Society 121(4)907ndash922DOI 101093biolinneanblx026

Alvarez Y Juste J Tabares E Garrido-Pertierra A Ibantildeez C Bautista JM 1999Molecular phylogeny and morphological homoplasy in fruitbatsMolecular Biologyand Evolution 16(8)1061ndash1067 DOI 101093oxfordjournalsmolbeva026195

Aacutevila LJ MorandoM Sites Jr JW 2006 Congeneric phylogeography hypothesizingspecies limits and evolutionary processes in Patagonian lizards of the Liolaemusboulengeri group (Squamata Liolaemini) Biological Journal of the Linnean Society89241ndash275 DOI 101111j1095-8312200600666x

Bardin J Rouget I Cecca F 2017 The phylogeny of Hildoceratidae (CephalopodaAmmonitida) resolved by an integrated coding scheme of the conch Cladistics3321ndash40 DOI 101111cla12151

Barrionuevo JS 2017 Frogs at the summits phylogeny of the Andean frogs of the genusTelmatobius (Anura Telmatobiidae) based on phenotypic characters Cladistics3341ndash68 DOI 101111cla12158

Bremer K 1992 Ancestral areas a cladistic reinterpretation of the center of originconcept Systematic Biology 41436ndash445 DOI 101093sysbio414436

Brooks DR 1990 Parsimony analysis in historical biogeography andcoevolution ethod-ological and theoretical update Systematic Zoology 3914ndash30 DOI 1023072992205

Castroviejo-Fisher S Guayasamin JM Gonzalez-Voyer A Vilaacute A 2014 Neotrop-ical diversification seen through glassfrogs Journal of Biogeography 466ndash80DOI 101111jbi12208

Cei JM 1979 Remarks on the South American lizard Liolaemus anomalus Koslowskyand the synonymy of Phrynosaura werneriMuumlller (Reptilia Lacertilia Iguanidae)Journal of Herpetology 13183ndash186 DOI 1023071563926

Cei JM 1986 Reptiles del centro centro-oeste y sur de la ArgentinaMuseo Regionale diScienze Naturali Torino Monografie 41ndash527

Cei JM 1993 Reptiles del noroeste nordeste y este de la ArgentinaMuseo Regionale DiScienze Naturale Torino Monografie 141ndash949

Chacoacuten J Renner SS 2014 Assessing model sensitivity in ancestral area reconstructionusing LAGRANGE a case study using the Colchicaceae family Journal of Biogeogra-phy 411414ndash1427 DOI 101111jbi12301

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 2430

Chaves JA Smith TB 2011 Evolutionary patterns of diversification in the Andeanhummingbird genus AdelomyiaMolecular Phylogenetics and Evolution 60207ndash218DOI 101016jympev201104007

Clark JR Ree RH AlfaroME KingMGWagnerWL Roalson EH 2008 A comparativestudy in ancestral range reconstruction methods retracing the uncertain histories ofinsular lineages Systematic Biology 57693ndash707 DOI 10108010635150802426473

Cosacov A Seacutersic AN Sosa V Johnson LA Cocucci AA 2010Multiple periglacialrefugia in the Patagonian steppe and post-glacial colonization of the Andes thephylogeography of Calceolaria polyrhiza Journal of Biogeography 371463ndash1477DOI 101111j1365-2699201002307x

Crisci JV Freire-E S Sancho G Katinas L 2001Historical biogeography of theAsteraceae from Tandilia and Ventania mountain ranges (Buenos Aires Argentina)Caldasia 2321ndash41

Diacuteaz Goacutemez JM 2011 Estimating ancestral ranges testing methods with a clade ofNeotropical lizards (Iguania Liolaemidae) PLOS ONE 6e26412DOI 101371journalpone0026412

Diacuteaz Goacutemez JM Lobo F 2006Historical biogeography of a clade of Liolaemus (IguaniaLiolaemidae) based on ancestral areas and dispersal-vicariance analysis (DIVA)Papeacuteis Avulsos de Zoologia 46261ndash274 DOI 101590S0031-10492006002400001

Doan TM Schargel WE 2003 Bridging the gap in Proctoporus distribution a newspecies (Squamata Gym-nophthalmidae) from the Andes of Venezuela Herpeto-logica 5968ndash75 DOI 1016550018-0831(2003)059[0068BTGIPD]20CO2

DonatoM Posadas P Miranda-Esquivel DR Ortiz Jaureguizar E Cladera G2003Historical biogeography of the Andean region evidence from Listroderina(Coleoptera Curculionidae Rhytirrhinini) in the context of the South Amer-ican geobiotic scenario Biological Journal of the Linnean Society 80339ndash352DOI 101046j1095-8312200300243x

Drummond AJ Ho SYW Phillips MJ Rambaut A 2006 Relaxed phylogenetics anddating with confidence PLOS Biology 4e88 DOI 101371journalpbio0040088

Drummond AJ Rambaut A 2007 BEAST Bayesian evolutionary analysis by samplingtrees BMC Evolutionary Biology 7214 DOI 1011861471-2148-7-214

Elias M JoronMWillmott K Silva-Brandatildeo KL Kaiser V Arias CF Gomez PintildeerezLM Uribe S Brower AVZ Freitas AVL Jiggins CD 2009 Out of the Andespatterns of diversification in clearwing butterfliesMolecular Ecology 181716ndash1729DOI 101111j1365-294X200904149x

Engstrom TN Shaffer HB McCordWP 2004Multiple data sets high homoplasy andthe phylogeny of softshell turtles (Testudines Trionychidae) Systematic Biology53(5)693ndash710 DOI 10108010635150490503053

Escobar Garciacutea P Schoumlnswetter P Fuertes Aguilar J Nieto Feliner G SchneeweissGM 2009 Five molecular markers reveal extensive morphological homoplasy andreticulate evolution in the Malva alliance (Malvaceae)Molecular Phylogenetics andEvolution 50226ndash239 DOI 101016jympev200810015

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 2530

Espinoza REWiens JJ Tracy CR 2004 Recurrent evolution of herbivory in small cold-climate lizards breaking the ecophysiological rules of reptilian herbivory Proceedingsof the National Academy of Sciences of the United States of America 10116819ndash16824DOI 101073pnas0401226101

Fontanella FM Olave M Avila LJ Sites Jr JW MorandoM 2012Molecular dating anddiversification of the South American lizard genus Liolaemus (subgenus Eulaemus)based on nuclear and mitochondrial DNA sequences Zoological Journal of theLinnean Society 164825ndash835 DOI 101111j1096-3642201100786x

Garzione CN Hoke GD Libarkin JCWithers S Mac Fadden B Eiler J Ghosh P MulchA 2008 Rise of the Andes Science 3201304ndash1307 DOI 101126science1148615

Goicoechea N Padial JM Chaparro JC Castroviejo-Fisher S Dela Riva I 2012Molecular phylogenetics species diversity and biogeography of the Andean lizardsof the genus Proctoporus (Squamata Gymnophthalmidae)Molecular Phylogeneticsand Evolution 65953ndash964 DOI 101016jympev201208017

Goloboff P Catalano S 2016 TNT version 15 with a full implementation of phyloge-netic morphometrics Cladistics 32221ndash238 DOI 101111cla12160

Goloboff P Farris J Nixon K 2003 TNT tree analysis using new technology Programand documentation Available at httpwwwlilloorgarphylogeny tnt

Goloboff P Farris J Nixon K 2008 TNT a free programa for phylogentic analysisCladistics 24774ndash786 DOI 101111j1096-0031200800217x

Goloboff PA Mattoni CI Quinteros AS 2006 Continuous characters analyzed as suchCladistics 22589ndash601 DOI 101111j1096-0031200600122x

Gregory-Wodzicki KM 2000 Uplift history of the central and northern andes a reviewGeological Society of America Bulletin 1121091ndash1105DOI 1011300016-7606(2000)112lt1091UHOTCAgt20CO2

Gubbels RW Isacks BL Farrar E 1993High-level surfaces plateau uplift and forelanddevelopment Bolivian central Andes Geology 21695ndash669DOI 1011300091-7613(1993)021lt0695HLSPUAgt23CO2

HahnMW Nakhleh L 2015 Irrational exuberance for resolved species trees Evolution707ndash17 DOI 101111evo12832

Hartley A 2003 Andean uplift and climate change Journal of the Geological Society1607ndash10 DOI 1011440016-764902-083

Hausdorf B 1998Weighted ancestral area analysis and a solution of theredundant dis-tribution problem Systematic Biology 47445ndash456 DOI 101080106351598260815

Hedges SB Kumar S 2004 Precision of molecular time estimates Trends in Genetetic20242ndash247 DOI 101016jtig200403004

HellmichW 1964 Uumlber eine neue Liolaemus-Art aus den Bergen von CatamarcaArgentinien Senckenb Biologica 45505ndash507

Hernaacutendez R Echavarria L 2009 Faja plegada y corrida subandina del noroesteargentino estratigrafiacutea geometriacutea y cronologiacutea de la deformacioacuten Revista de laAsociacioacuten Geoloacutegica Argentina 6568ndash80

Hoorn CWesselinghter Steege FP Bermudez H MoraMA Sevink A Sanmartiacuten JSanchez-Meseguer I Anderson A Figueiredo CL Jaramillo JP Riff C Negri D

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 2630

Hooghiemstra FR Lundberg H Stadler J Sarkinen T Antonelli A 2010 Amazoniathrough time Andean uplift climate change landscape evolution and biodiversityScience 330927ndash931 DOI 101126science1194585

Jarvis ED Mirarab S Aberer AJ Li B Houde P Li C Ho SY Faircloth BC Nabholz BHoward JT 2014Whole-genome analyses resolve early branches in the tree of life ofmodern birds Science 3461320ndash1331 DOI 101126science1253451

Kumar S Stecher G Tamura K 2016MEGA 7 molecular evolutionary genetics analysisversion 70 for bigger datasetsMolecular Biology and Evolution 331870ndash1874DOI 101093molbevmsw054

Lanfear R Calcott B Ho SYW Guindon S 2012 PartitionFinder combined selection ofpartitioning schemes and substitution models for phylogenetic analysesMolecularBiology and Evolution 29(6)1695ndash1701

Laurent RF 1983 Contribucioacuten al conocimiento de la estructura taxonoacutemica del geacuteneroLiolaemusWiegmann (Iguanidae) Boletin de la Asociacion Herpetologica Argentina115ndash18

Lobo F 2001 A phylogenetic analysis of lizards of the Liolaemus chiliensis group(Iguania Tropiduridae) Journal of Herpetologic 11137ndash150

Lobo F 2005 Las relaciones filogeneacuteticas en el grupo chiliensis de Liolaemus (IguaniaLiolaemidae) Sumando nuevos caracteres y taxa Acta Zoologica Lilloana 4967ndash89

Lobo F Espinoza RE 1999 Two new cryptic species of Liolaemus (Iguania Tropiduri-dae) from northwestern Argentina resolution of the purported reproductivebimodality of Liolaemus alticolor Copeia 1999122ndash140 DOI 1023071447393

Martiacutenez LE Aacutevila LJ Peacuterez CHF Peacuterez DR Sites Jr JW MorandoM 2011 A newspecies of Liolaemus (Squamata Iguania Liolaemini) endemic to the Auca Mahuidavolcano Northwestern Patagonia Argentina Zootaxa 301031ndash46

Martinez Oliver I Lobo F 2002 Una nueva especie de Liolaemus del grupo alticolor(Iguania Liolaemidae) de la puna saltentildea Argentina Cuadernos de Herpetologiacutea16(1)47ndash64

Medina CD Avila LJ Sites Jr JW MorandoM 2014Multilocus phylogeography of thepatagonian lizard complex Liolaemus kriegi (Iguania Liolaemini) Biological Journalof the Linnean Society 113256ndash269 DOI 101111bij12285

MorandoM Aacutevila LJ Baker J Sites Jr JW 2004 Phylogeny and phylogeography of theLiolaemus darwinii complex (Squamata Liolaemidae) evidence for introgression onincomplete lineage sorting Evolution 58842ndash861DOI 101111j0014-38202004tb00416x

MorandoM Aacutevila LJ Sites Jr JW 2003 Sampling strategies for delimiting species genesindviduals and populations in the Liolaemus elongatus-kriegi complex (SquamataLiolaemidae) in Andean-Patagonian South America Systematic Biology 52159ndash185DOI 10108010635150390192717

MorandoM Aacutevila LJ Turner CR Sites Jr JW 2007Molecular evidence for a speciescomplex in the Patagonian lizard Liolaemus bibronii and phylogeography of theclosely related Liolaemus gracilis (Squamata Liolaemini)Molecular Biology andEvolution 43952ndash973 DOI 101016jympev200609012

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 2730

Morrone JJ 2001 Biogeografiacutea de Ameacuterica Latina y el Caribe Vol 3 Zaragoza MampTndashManuales amp Tesis SEA 148 pp

Mott T Vieites DR 2009Molecular phylogenetics reveals extreme morphologicalhomoplasy in Brazilian worm lizards challenging current taxonomyMolecularPhylogenetics and Evolution 51(2)190ndash200 DOI 101016jympev200901014

Mueller RL Macey JR Jaekel MWake DB Boore JL 2004Morphological ho-moplasy life history evolution and historical biogeography of plethodontidsalamanders inferred from complete mitochondrial genomes Proceedings of theNational Academy of Sciences of the United States of America 101(38)13820ndash13825DOI 101073pnas0405785101

Naacutejera-Cortazar LA Aacutelvarez Castantildeeda ST De Luna E 2015 An analysis ofMyotispeninsularis (Vespertilionidae) blending morphometric and genetic datasets ActaChiropterologica 17(1)37ndash47 DOI 10316115081109ACC2015171003

Nylander JAA Olsson U Alstroumlm P Sanmartiacuten I 2008 Accounting for phylogeneticuncertainty in biogeography a Bayesian approach to dispersalndashvicariance analysis ofthe thrushes (Aves Turdus) Systematic Biology 57257ndash268DOI 10108010635150802044003

OlaveM Avila LJ Sites Jr JW MorandoM 2011 Evidence of hybridization in theArgentinean lizard Liolaemus gracilis and Liolaemus bibronii (IguaniLiolaemini) anintegrative approach based on genes and morphologyMolecular Phylogenetics andEvolution 61381ndash391 DOI 101016jympev201107006

Ortiz JC 1981 Estudio multivariado de las especies de Liolaemus del grupo nigromac-ulatus (Squamata Iguanidae) Anales del Museo de Historia Natural de Valparaiacuteso14247ndash265

Panzera A Leacheacute AD DrsquoElia G Victoriano PF 2017 Phylogenomic analysis of theChilean clade of Liolaemus lizards (Squamata Liolaemidae) based on sequencecapture data PeerJ 5e3941 DOI 107717peerj3941

Pincheira-Donoso D Tregenza T Matthew J Hodgson DJ 2013 The evolution ofviviparity opens opportunities for lizard radiation but drives it into a climatic cul-de-sac Global Ecology and Biogeography 22857ndash867 DOI 101111geb12052

Posada D 2008 jModelTest phylogenetic model averagingMolecular Biology andEvolution 251253ndash1256 DOI 101093molbevmsn083

Pyron RA Burbrink FTWiens JJ 2013 A phylogeny and revised classification ofSquamata including 4161 species of lizards and snakes Evolutionary Biology 1393DOI 1011861471-2148-13-93

Quinteros AS 2012 Taxonomy of the Liolaemus alticolor-bibronii group (IguaniaLiolaemidae) with descriptions of two new species Herpetologica 68100ndash120DOI 101655HERPETOLOGICA-D-10-000651

Quinteros AS 2013 A morphology-based phylogeny of the Liolaemus alticolorndashbibroniigroup (Iguania Liolaemidae) Zootaxa 36701ndash32

Quinteros AS Valladares P Semhan R Acosta JL Barrionuevo S Abdala CS2014 A new species of Liolaemus (Iguania Liolaemidae) of the alticolor-bibronii

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 2830

group from northern Chile South American Journal of Herpetology 920ndash29DOI 102994SAJH-D-13-000401

Rambaut A 2008 FigTree tree figure drawing tool version 122 Available at http treebioedacuk software figtree (accessed on 18 December 2018)

Rambaut A SuchardMA Xie D Drummond AJ 2014 Tracer v16 Available at http beastcommunity tracer

Ramirez Huaroto DW 2012 Filogenia y Biogeografiacutea del geacutenero Distichia Nees ampMeyen (juncaceae) en los Andes Tropicales MSc thesis Universidad InternacionalMenendez Pelayo

Ramos VA Kay SM 2006 Overview of the tectonic evolution of the Southern CentralAndes of Mendoza and Neuqueacuten (35ndash39S Latitude) Geological Society of AmericaSpecial Paper 4071ndash18

Ree RH Smith SA 2008Maximum likelihood inference of geographic range evolutionby dispersal local extinction and cladogenesis Systematic Biology 571 4ndash14DOI 10108010635150701883881

Roig Juntildeent S 1994Historia biogeograacutefica de Ameacuterica del Sur australMultequina3167ndash203

Ronquist F 1997 Dispersal-vicariance analysis a new approach to thequantification ofhistorical biogeography Systematic Biology 46195ndash203 DOI 101093sysbio461195

Ronquist F Huelsenbeck JP 2003Mr Bayes 3 Bayesian phylogenetic inference undermixed models Bioinformatic 191572ndash1574 DOI 101093bioinformaticsbtg180

Schulte JA 2013 Undersampling taxa will underestimate molecular divergence dates anexample from the South American Lizard Clade Liolaemini International Journal ofEvolutionary Biology 20131ndash12 DOI 1011552013628467

Schulte JA Macey JR Espinoza RE Larson A 2000 Phylogenetic relationships in theiguanid lizard genus Liolaemus multiple origins of viviparous reproduction andevidence for recurring Andean vicariance and dispersal Biological Journal of theLinnean Society 6975ndash102 DOI 101111j1095-83122000tb01670x

Seacutersic AN Cosacov A Cocucci AA Johnson IA Pozner R Avila LJ Sites Jr JWMorandoM 2011 Emerging phylogeographical patterns of plants and terrestrialvertebrates from Patagonia Biological Journal of the Linnean Society 103475ndash494DOI 101111j1095-8312201101656x

Torres-Carvajal O Lobos SE Venegas PJ Chaacutevez G Aguirre-Pentildeafiel V ZuritaD Echevarriacutea LY 2016 Phylogeny and biogeography of the most diverseclade of South American gymnophthalmid lizards (Squamata Gymnoph-thalmidae Cercosaurinae)Molecular Phylogenetics and Evolution 9963ndash75DOI 101016jympev201603006

Troncoso-Palacios J Elorza AA Puas GI Alfaro-Pardo E 2016 A new species ofLiolaemus related to L nigroviridis from the Andean highlands of Central Chile(Iguania Liolaemidae) ZooKeys 55591ndash114 DOI 103897zookeys5556011

Troncoso-Palacios J Schulte JA Marambio-Alfaro Y Hiriart D 2015 Phenotypicvariation phylogenetic position and new distributional records for the poorly known

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 2930

Liolaemus silvai Ortiz 1989 (Iguania Iguanidae Liolaemini) South AmericanJournal of Herpetology 1071ndash81

Victoriano PF Ortiz JC Benavides E Adams BJ Sites Jr JW 2008 Comparativephylogeography of codistributed species of Chilean Liolaemus (Squamata Tropiduri-dae) from the central-southern Andean rangeMolecular Ecology 172397ndash2416DOI 101111j1365-294X200803741x

Whitmore TC Prance GT 1987 Biogeography and quaternary history in TropicalAmerica Oxford Monographs on Biogeography 1st edition New York OxfordUniversity Press 224 pp

Young-Downey AR 1998 Phylogenetic studies on Liolaemus (Sauria Tropiduridae) aninterpretation based on molecular data and a biochemical test of a biogeographichypothesis PhD thesis University of Miami Coral Gables Florida 85 pp

Yu Y Harris AJ He XJ 2010 S-DIVA (statistical dispersalndashvicariance analysis) a toolfor inferring biogeographic historiesMolecular Phylogenetic Evolution 56848ndash850DOI 101016jympev201004011

Zheng YWiens JJ 2016 Combining phylogenomic and supermatrix approachesand a time-calibrated phylogeny for squamate reptiles (lizards and snakes) basedon 52 genes and 4162 speciesMolecular Phylogenetics and Evolution 94537ndash547DOI 101016jympev201510009

Portelli and Quinteros (2018) PeerJ DOI 107717peerj4404 3030