Molecular phylogeny of the Western Palaearctic Cordulegaster taxa...

Molecular phylogeny of the Western PalaearcticCordulegaster taxa (Odonata: Anisoptera:Cordulegastridae)

ELSA FROUFE1*, SÓNIA FERREIRA2,3, JEAN-PIERRE BOUDOT4, PAULO C. ALVES2,3,5

and DAVID JAMES HARRIS2,3

1CIIMAR, Centro Interdisciplinar de Investigação Marinha e Ambiental, Rua dos Bragas, 289,4050-123 Porto, Portugal2CIBIO/InBIO, Centro de Investigação em Biodiversidade e Recursos Genéticos da Universidade doPorto, R. Padre Armando Quintas, 4485-661 Vairão, Portugal3Departamento de Biologia, Faculdade de Ciências da Universidade do Porto, 4099-002 Porto,Portugal4Immeuble Orphée, Apt 703, Cidex 62, 78 rue de la Justice, F-54710 Ludres, France5Wildlife Biology Program, University of Montana, Missoula, MT 59812, USA

Received 9 August 2013; revised 29 August 2013; accepted for publication 29 August 2013

Although Odonata are a key component of many freshwater ecosystems, their taxonomy and evolutionary historyis still far from being well resolved. In the present study, we report the first molecular phylogeny for the WesternPalaearctic Cordulegaster genus (Odonata: Anisoptera: Cordulegastridae). We sequenced fragments of bothmitochondrial and nuclear genes [cytochrome c oxidase I (COI) and Internal Transcribed Spacer-1 (ITS-1)] fromeight species and 13 subspecies, from western, southern and central Europe, Turkey, and Morocco. Our datasupport the existence of two major groups corresponding to the traditional boltonii- and bidentata-groups. Bothgroups are monophyletic based on COI sequences and the distinctiveness of Cordulegaster princeps, Cordulegastertrinacriae, Cordulegaster picta and Cordulegaster heros relative to Cordulegaster boltonii, and Cordulegasterhelladica and Cordulegaster insignis relative to Cordulegaster bidentata, is confirmed. All species are alsomonophyletic for ITS-1, with the exception of Cordulegaster helladica buchholzi, which shares the haplotype withC. insignis. Although moderate levels of genetic diversity were found within C. boltonii, there was no clearseparation among the four subspecies, with the exception of the populations of Cordulegaster boltonii algirica fromNorth Africa. Similarly, no genetic differentiation was found between the two subspecies of C. bidentata,Cordulegaster bidentata bidentata and Cordulegaster bidenta sicilica. © 2013 The Linnean Society of London,Biological Journal of the Linnean Society, 2014, 111, 49–57.

ADDITIONAL KEYWORDS: Cordulegaster bidentata – Cordulegaster boltonii – Europe – mitochondrial DNA– Morocco – nuclear DNA – Turkey.

INTRODUCTION

Odonata are a key component of many freshwaterecosystems. Their dependence on freshwater forreproduction, their role as predators, and their sen-

sitivity to changes in the environment make themvaluable indicators of the overall wetland habitatsquality. Studies on Odonata have increased signifi-cantly in the last decades and they are currentlyregarded as model organisms for ecological and evo-lutionary research (Córdoba-Aguilar, 2008). Never-theless, the Odonata taxonomy and evolutionaryhistory is still far from being well resolved, even in*Corresponding author. E-mail: [email protected]

bs_bs_banner

Biological Journal of the Linnean Society, 2014, 111, 49–57. With 3 figures

© 2013 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 111, 49–57 49

mailto:[email protected]

Europe where the number of species is small andwhere a large number of studies have been conducted(Dijkstra & Kalkman, 2012).

The genus Cordulegaster (Odonata: Anisoptera:Cordulegastridae) is a good example of a phylo-genetically poorly understood group. With a Holarcticdistribution, this genus comprises approximately 30species. Regardless of the controversial placementand/or validity of several species, to date, only thevalidity of Cordulegaster bilineata has been testedwith molecular markers (Pilgrim, Roush & Krane,2002). Cordulegaster has eight species currently rec-ognized in the Western Palaearctic, which are tradi-tionally divided into two groups: the boltonii- and thebidentata-groups (Boudot, 2001). These groups differin morphology and ecological traits. Morphologicaldistinctness is found in the superior abdominalappendages of males and in the ovipositor of females.Ecologically, boltonii-group habitats vary from hilland mountain springs and from brooks to rivers,whereas the species of the bidentata-group are con-fined to upper courses of brooks as well as to seepageand spring habitats. Despite their widespread distri-bution and complex subspecific division, the validityof some species and some subspecies is still ques-tioned and genetic diversity within WesternPalaearctic species of Cordulegaster remainsunknown (Boudot et al., 2009). This obscures conser-vation priorities and hampers the development ofadequate conservation measures because, of the eightspecies in the Western Palaearctic, three are classi-fied with threatened categories and four are NearThreatened (Kalkman et al., 2010; Samraoui et al.,2010). In the present study, we sequenced the inter-nal transcribed spacer 1 (ITS-1) and part of themitochondrial (mt)DNA cytochrome c oxidase I (COI)gene in order to determine: (1) the phylogenetic rela-tionships between all the Cordulegaster species of theWestern Palaearctic region and (2) the genetic dis-tinctiveness of most subspecies.

MATERIAL AND METHODSSPECIMENS EXAMINED

A total of 112 samples of all Western PalaearcticCordulegaster species covering the majority of specieswith ranges from Morocco and southern, western andcentral Europe and south-western Anatolia wereanalyzed (Fig. 1, Table 1, see also Supporting infor-mation, Table S1). Representatives of 13 subspecies,from 59 different locations, were assigned morphologi-cally to subspecific taxa sensu Boudot (2001) andBoudot & Jacquemin (1995), whenever possible(Table 1 and Fig. 1). As outgroups, representatives ofAnotogaster sieboldi Selys, 1854, Chlorogomphus

okinawensis Ishida, 1964, Chlorogomphus brunneusOguma, 1926, Chlorogomphus iriomotensis Ishida,1972, Onychogomphus forcipatus (Linnaeus, 1758)and Anax imperator Leach, 1815 were selected(Table 1).

DNA EXTRACTION, AMPLIFICATION AND SEQUENCING

Total genomic DNA was isolated from a leg of eachindividual in accordance with a high-salt protocol(Sambrook, Fritsch & Maniatis, 1989). Two partialDNA fragments, mtDNA COI and ITS-1 were ampli-fied using the polymerase chain reaction (PCR). TheCOI was amplified using the LCO1490 and HC02198primers (Folmer et al., 1994) producing approxi-mately 700-bp fragments and the ITS-1 with 18S-forward and 5.8S-reverse primers (Pilgrim et al.,2002), producing approximately 900-bp fragments.The PCR conditions (25-μL reactions) were: 19 μL ofH2O, 2.5 μL of 10× Promega Buffer B, 0.5 μL of10 mM of each primer, 1.5 μL of 25 mM MgCl2, 0.5 μLof 10 mM dNTPs, 0.1 μL of Promega Taq DNA poly-merase, and 0.5 μL of 100 ng per μL DNA template.The cycle parameters were: initial denaturation at94 °C for 3 min, denaturation at 94 °C (30 s), anneal-ing at 56 °C (45 s) for both COI and ITS-1, andextension at 72 °C (45 s) repeated for 35 cycles, witha final extension of 5 min at 72 °C. Amplified DNAfragments were purified and sequenced (most withboth primers) using the same primers as in the PCR,following ABI PRISM BigDye Terminator protocols.Sequences were visualized on a 310 AppliedBiosystem DNA Sequencing Apparatus. All sequenceswere submitted to GenBank (Table 1).

PHYLOGENETIC ANALYSIS

COI alignment was performed manually usingBIOEDIT, version 5.0.9. (Hall, 1999). Only the differ-ent COI haplotypes were included in the phylogeneticanalyses. The alignment of the Cordulegasterhaplotypes was analyzed using maximum likelihood(ML) and Bayesian inference (BI) methods, includingthe outgroups. The best-fit model of nucleotide sub-stitution evolution under corrected Akaike informa-tion criterion was estimated using JMODELTEST,version 0.1.1 (Posada, 2008). Model GTR + G waschosen and used in the ML phylogenetic analyses. MLtrees were built in PHYML (Guindon et al., 2010)with 1000 bootstrap replicates and searching forthe best-scoring ML tree. Phylogenetic BI was per-formed using MrBayes, version 3.1.2 (Ronquist &Huelsenbeck, 2003). Two independent runs of 107

generations were sampled at intervals of 100 genera-tions, producing a total of 100 000 trees. Burn-in wasdetermined upon convergence of log likelihood and

50 E. FROUFE ET AL.

© 2013 The Linnean Society of London, Biological Journal of the Linnean Society, 2014, 111, 49–57

parameter estimation values using TRACER, version1.5 (Rambaut & Drummond, 2003).

Estimates of average evolutionary divergence oversequence pairs within and between groups, usinguncorrected p-distances (p-dist), were calculatedbased on the number of base differences per site fromaveraging over all sequence pairs within each groupand estimation of net average between groups ofsequences, respectively in MEGA, version 4.0.2(Tamura et al., 2007).

Sequences from ITS-1 were aligned in MAFFT(Katoh & Miyata, 2005) and the alignment wasanalyzed using ML and BI methods, in accordancewith the same methodology for the COI fragment.Because these sequences were found to be generallyconserved (preliminary results; data not shown),unrooted networks were constructed using statistical

parsimony (Templeton, Crandall & Sing, 1992) toevaluate the relationships among these closely-related haplotypes. This analysis was implemented inTCS, version 1.21 (Clement, Posada & Crandall,2000) with a connection limit of 95% and indelstreated as a fifth state. The same methodology wasused with COI in order to better evaluate variationwithin species.

RESULTS

The COI alignment consisted of 112 DNA sequencesfrom Cordulegaster specimens and six outgroup speci-mens (Table 1). Aligned sequences had a total length of570 bp, with 163 polymorphic and 139 parsimonyinformative sites and high levels of nucleotide variabil-ity (HD = 0.965, π = 0.06349, within Cordulegaster). No

Figure 1. Map showing the location of sequenced samples of Cordulegaster species in the Western Palaearctic. Popu-lation codes match those used in Table 1. Yellow, Cordulegaster bidentata; red, Cordulegaster boltonii; green,Cordulegaster princeps; violet, Cordulegaster trinacriae; orange, Cordulegaster heros; brown, Cordulegaster helladica;light blue, Cordulegaster insignis; dark blue, Cordulegaster picta.

WESTERN PALAEARCTIC CORDULEGASTER TAXA 51


Tab

le1.

Lis

tof

sam

ples

:spe

cies

,su

bspe

cies

,nu

mbe

rof

spec

imen

s(N

),po

pula

tion

code

,cou

ntr

y,IU

CN

regi

onal

Red

Lis

tth

reat

ened

cate

gori

es(E

uro

pean

/Nor

thA

fric

an),

and

cyto

chro

me

cox

idas

eI

(CO

I)an

dIn

tern

alT

ran

scri

bed

Spa

cer-

1(I

TS

-1)

hap

loty

pes

Spe

cies

Su

bspe

cies

NP

opu

lati

onC

oun

try

Hap

loty

pe

IUC

NC

OI

ITS

Cor

du

lega

ster

bid

enta

tabi

den

tata

51,

4F

ran

ce,

Ital

yN

Th

1,h

2,h

3,h

4Q

Cor

du

lega

ster

bid

enta

tasi

cili

ca4

5–7

Ital

yS

icil

yN

Th

5,h

6,h

7O

Cor

du

lega

ster

bid

enta

ta13

2–3,

8–14

Ger

man

y,M

onte

neg

ro,

Alb

ania

,G

reec

eN

Th

2,h

8–h

15O

,P

Cor

du

lega

ster

bolt

onii

iber

ica

1018

–23

Spa

inL

Ch

19–h

24E

,F

Cor

du

lega

ster

bolt

onii

imm

acu

lifr

ons

525

Fra

nce

,P

ortu

gal

LC

h16

,h

21,

h25

A,

BC

ord

ule

gast

erbo

lton

iibo

lton

ii4

16F

ran

ce,

Por

tuga

lL

Ch

17,

h21

,h

26A

,B

,E

Cor

du

lega

ster

bolt

onii

1124

,28

–32

Au

stri

a,G

erm

any,

Ital

y,S

pain

LC

h21

,h

27–h

28C

,D

Cor

du

lega

ster

bolt

onii

algi

rica

1217

,33

–36

Mor

occo

,S

pain

LC

h18

,h

29–h

31E

,G

Cor

du

lega

ster

hel

lad

ica

hel

lad

ica

337

,38

Gre

ece

EN

h32

–h33

S,

TC

ord

ule

gast

erh

ella

dic

abu

chh

olzi

639

Gre

ece

EN

h34

RC

ord

ule

gast

erh

ella

dic

a3

40G

reec

eE

Nh

35L

,S

Cor

du

lega

ster

her

ospe

lion

ensi

s3

41–4

2G

reec

eN

Th

37L

Cor

du

lega

ster

insi

gnis

643

–45

Turk

eyE

Nh

38–h

40R

Cor

du

lega

ster

pict

a2

46G

reec

eV

Uh

41M

,N

Cor

du

lega

ster

prin

ceps

1747

–53

Mor

occo

NT

h42

–h49

H,

I,J

Cor

du

lega

ster

trin

acri

ae8

54–5

9It

aly

NT

h50

–h51

KA

not

ogas

ter

sieb

old

i1

Japa

nC

hlo

rogo

mph

us

bru

nn

eus

1Ja

pan

Ch

loro

gom

phu

sir

iom

oten

sis

1Ja

pan

Ch

loro

gom

phu

sok

inaw

ensi

s1

Japa

nA

nax

impe

rato

r1

Por

tuga

lO

nyc

hog

omph

us

forc

ipat

us

1M

oroc

co

52 E. FROUFE ET AL.


indels and no stop codons were observed, after trans-lating all sequences to amino acids.

The final COI alignment used in the phylogenyreconstruction consisted of 57 distinct haplotypes,including the outgroups. The tree topologies resultingfrom ML and BI approaches were congruent andreceived high support values (> 0.9) for most relation-ships between species (Fig. 2). The Cordulegastergenus is clearly separated into two distinct clades,corresponding to the boltonii- and bidentata-groupsand monophyly is evident with strong nodal supportfor all the species, with the relationships betweenspecies in the boltonii-group being better resolved(Fig. 2). The boltonii-group includes five Cordulegasterspecies: Cordulegaster boltonii, Cordulegaster trinac-riae, Cordulegaster princeps, Cordulegaster heros andCordulegaster picta, whereas Cordulegaster bidentata,Cordulegaster insignis and Cordulegaster helladicagrouped together to form the bidentata-group. Rela-tionships within the boltonii-group are better resolvedin the BI analysis than in the ML; furthermore,C. boltonii and C. trinacriae are sister taxa (89 Bayes-ian posterior probability but no bootstrap support),with the Moroccan endemic C. princeps being relatedto this pair (again with higher Bayesian posteriorprobability than bootstrap support). Cordulegasterheros and C. picta, from the Balkans and Turkey aresister taxa, basally related to the three previousmembers of the group. Within the bidentata-group,relationships between the three species were not wellsupported in both BI and ML analysis. The maximumpairwise sequence divergence (uncorrected distances)was retrieved between C. princeps and C. bidentata(10%) and the minimum value (5.1%) was observedbetween C. helladica and C. insignis (Table 2).

Although the number of haplotypes found withinthe more widespread species C. boltonii andC. bidentata was quite high (16 and 15, respectively),the haplotype networks showed little correspondencewith either geography or currently defined subspe-cies. Within C. boltonii, a notable exception are thespecimens of Cordulegaster boltonii algirica fromNorth Africa, which form a distinct group separatedfrom all other haplotypes (h29, h30 and h31) (Fig. 3).Specimens from Cordulegaster boltonii iberica andCordulegaster boltonii immaculifrons did not formdistinct units and shared the commonest haplo-types. The network has the classical star topologywith a central haplotype (h21) shared by the threeother subspecies: Cordulegaster boltonii boltonii,C. b. iberica and C. b. immaculifrons (Fig. 3). Moreo-ver, h21 is present in individuals from several loca-tions in Iberia, France, Germany and Austriacovering most of the range of the species. The fourspecimens from central Italy (Tuscany) are repre-sented by two exclusive haplotypes (h27 and h28) T

able

2.E

stim

ates

ofev

olu

tion

ary

dive

rgen

ceov

erse

quen

cepa

irs

betw

een

grou

ps(A

)(n

um

ber

ofba

sedi

ffer

ence

spe

rsi

tefr

omav

erag

ing

over

alls

equ

ence

pair

sbe

twee

ngr

oups

(SE

esti

mat

esar

esh

own

abov

eth

edi

agon

al)

and

esti

mat

esof

aver

age

evol

uti

onar

ydi

verg

ence

over

sequ

ence

pair

sw

ith

ingr

oups

(B)

(nu

mbe

rof

base

diff

eren

ces

per

site

from

aver

agin

gov

eral

lse

quen

cepa

irs

wit

hin

each

grou

p)

AC

ord

ule

gast

erbo

lton

iiC

ord

ule

gast

ertr

inac

riae

Cor

du

lega

ster

prin

ceps

Cor

du

lega

ster

her

osC

ord

ule

gast

erpi

cta

Cor

du

lega

ster

bid

enta

taC

ord

ule

gast

erin

sign

isC

ord

ule

gast

erh

ella

dic

aB

Cor

du

lega

ster

bolt

onii

–0.

009

0.00

90.

011

0.01

0.01

10.

011

0.01

10.

021

Cor

du

lega

ster

trin

acri

ae0.

053

–0.

009

0.01

10.

010.

011

0.01

10.

011

0.00

2

Cor

du

lega

ster

prin

ceps

0.05

50.

057

–0.

011

0.01

0.01

10.

011

0.01

0.00

9C

ord

ule

gast

erh

eros

0.07

90.

082

0.08

–0.

009

0.01

10.

010.

010

Cor

du

lega

ster

pict

a0.

073

0.07

10.

080.

054

–0.

011

0.01

0.01

0C

ord

ule

gast

erbi

den

tata

0.09

10.

081

0.1

0.09

0.09

2–

0.00

90.

009

0.01

9C

ord

ule

gast

erin

sign

is0.

092

0.07

50.

093

0.08

10.

081

0.06

2–

0.00

80.

011

Cor

du

lega

ster

hel

lad

ica

0.08

70.

081

0.08

90.

081

0.07

90.

062

0.05

1–

0.04



54 E. FROUFE ET AL.


exhibiting some distinctiveness (three steps). TheC. bidentata haplotype network does not corroboratethe distinctiveness of Cordulegaster bidenta sicilica(h5, h6 and h7) from the nominotypical subspeciesand indicates the specimens from South Italy (h3 andh4) as the most genetically distinct ones (five muta-tions) (Fig. 3). The haplotypes of Cordulegasterhelladica helladica and Cordulegaster helladicabuchholzi do not connect and are not shared withC. insignis.

The ITS-1 alignment consisted of 79 DNAsequences from Cordulegaster specimens (Table 1), asa result of the non-amplification of some individuals.Aligned sequences had a total length of 490 bp withno heterozygote individuals found; regarding indels,two 4–8-bp indels were found within the boltonii-group and five 2–8-bp indels within the bidentata-group. The diversity observed in the ITS-1 sequencesin general supports the species distinctions estimatedwith the COI sequence data, with one noteworthyexception: all the C. h. buchholzi analyzed share thehaplotype with C. insignis (haplotype R) (Fig. 2). Fur-thermore, only two pairs of species are connectedunder the 95% parsimony criterion used, by aminimum of eight steps: C. boltonii and C. princeps,and C. insignis and C. helladica (Fig. 2).

DISCUSSION

The European Odonata fauna is particularly wellknown, yet taxonomic uncertainties remain in severalgenera (Dijkstra & Kalkman, 2012), includingCordulegaster. We examined for the first time bothmitochondrial and nuclear genes sequences for allWestern Palaearctic species of this genus. Our datasupport the existence of two major groups correspond-ing to the boltonii- and bidentata-groups in agree-ment with morphological and ecological traits.Furthermore, all species are monophyletic and thediversity observed in the ITS-1 sequences supportsthe species distinctions estimated with the COIsequence data, with one notable exception: all the

C. h. buchholzi analyzed share the haplotype withC. insignis from Turkey. Cordulegaster helladica andC. insignis are morphologically distinct species, beingidentified by the shape of the posterior face of theoccipital triangle (Lohmann, 1993). They occur inrelatively close geographical areas (populations 39,43–45; Fig. 1) and therefore this shared haplotypemight be the result of hybridization, or the mainte-nance of an ancestral polymorphism. However, thelow number of the specimens analyzed does not allowa more comprehensive assessment of the taxonomicimplications of this shared haplotype.

Regarding subspecies validity and within the wide-spread C. boltonii, only the North African popula-tions of C. b. algirica are genetically distinct, withthe European specimen not grouping together,despite its morphological similarity. Specimens fromC. b. iberica and C. b. immaculifrons did not formdistinct units and shared the commonest haplotypes.The value of maintaining these taxa at the subspe-cies level is doubtful. Although various studiesaddress the usefulness of subspecies (Braby,Eastwood & Murray, 2012), in the case of these formsthat show genetic overlap and only colour patternvariations, there is little obvious taxonomic or con-servation value. The validity of C. bidenta sicilica isalso questionable because our data do not corroborateits distinctiveness from C. bidentata bidentata. Bycontrast, there are two haplogroups separated by4.0% divergence (COI) in C. helladica, namely C. h.helladica from the Peloponnese and C. h. buchholzifrom the Cyclades islands. This divergence is thehighest observed within a species in the presentstudy, and at a level similar to that between someaccepted species. This result, together with the IUCNRed List of threatened categories of C. helladica andits subspecies status (Boudot, 2010; Kalkman et al.,2010), reinforces the need of more sampling and theuse of additional markers for a detailed assessmentof the Greek and Turkish Cordulegaster taxa. Thiswould allow clarifying conservation priorities andassisting the development of adequate conservationmeasures.

It appears likely that North Africa was colonizedtwice from Iberia, with one older colonization by acommon ancestor of C. princeps and C. boltonii andone more recent by C. boltonii. Other taxa, suchas wall lizards (Podarcis spp.) show a similar pattern(Kaliontzopoulou et al., 2011), although the widevariety of colonization patterns across the Straits fordifferent organisms (e.g. Habel, Dieker & Schmitt,2009; Habel et al., 2010) have led to the hypothesisthat evolutionary histories unique to each specieshave produced this array of diverse scenarios(Jaramillo-Correa et al., 2010). As a result of the widedistribution of the commonest haplotypes, especially

Figure 2. Left: phylogenetic tree obtained by Bayesianinference (BI) analysis of 112 cytochrome c oxidase Isequences (570 bp) of Cordulegaster specimens and severaloutgroups. Support values (%) are given as bootstrapsupport/Bayesian posterior probability. The tree topologiesresulting from Maximum Likelihood and BI approacheswere congruent. Right: haplotype (TCS) networks showingthe relationships of Cordulegaster specimens, inferredfrom 79 individuals sequenced for 490-bp ITS nuclearsequences. Circle size is proportional to the observedhaplotype frequencies and black points represent unob-served haplotypes and potential intermediates.◀



for C. boltonii, determination of refugium locations isdifficult without further sampling and the use offaster evolving markers.

To conclude, our results generally support thecurrent taxonomy of Cordulegaster in the WesternPalaearctic, although particularly in Greece adetailed study is needed to determine the taxonomicstatus of the endangered subspecies, which showa high level of differentiation in COI but also anincongruent pattern with ITS-1. The two ecologicalgroups, the boltonii- and the bidentata-groups,are well-supported genetic entities, with relation-ships between species in the boltonii-group beingbetter resolved. Moderate genetic diversity withinC. bidentata and C. boltonii species suggests thatlarge populations survived in diverse refugia duringthe last ice ages, although little geographicaldifferentiation/substructuration is presently observed.No evidence was found supporting the validity ofC. b. iberica, C. b. immaculifrons and C. bidentasicilica, thus, conservation efforts can be appropri-ately focused at the species level in Europe for bothC. boltonii and C. bidentata.

ACKNOWLEDGEMENTS

The authors wish to thank the four reviewers forhelpful remarks and suggestions that improved thequality of the manuscript. We are grateful to our

colleagues who sent samples for the genetic analysis:Andrea Corso, Dávid Murányi, Elisa Riservato, GeertDe Knijf, Julia Wrede, Klaus Guido Leipelt, MilošJovic, Sonke Hardersen, Takuya Kiyoshi, VincentKalkman and Wolfgang Schweighofer. This researchwas partially supported by Pest-C/MAR/LA0015/2011and Pest-e/MAR/2A0015/2011. EF has a contract(Programme Ciência 2008) and SF has a researchfellowship (SFRH/BD/65038/2009), both from FCT.DJH is supported by FEDER through the competeprogramme, the project ‘Genomics and EvolutionaryBiology’ cofinanced by North Portugal RegionalOperational Program (ON.2) under NSRF throughthe European Regional Development Fund.

REFERENCES

Boudot J-P. 2001. Les Cordulegaster du Paléarctique occi-dental: identification et répartition (Odonata, Anisoptera,Cordulegastridae). Martinia 17: 1–34.

Boudot J-P. 2010. Cordulegaster helladica. In: IUCN 2012.IUCN Red List of threatened species, Version 2012.1. Avail-able at: http://www.iucnredlist.org

Boudot J-P, Jacquemin G. 1995. Revision of Cordulegasterboltonii (Donovan, 1807) in southwestern Europe and north-ern Africa, with description of C. b. iberica ssp. from Spain(Anisoptera: Cordulegastridae). Odonatologica 24: 149–173.

Boudot J-P, Kalkman VJ, Azpilicueta M, Amorín T,Bogdanovic A, Cordero Rivera A, Degabriele G,

Figure 3. Haplotype (TCS) networks showing the relationships of Cordulegaster specimens, inferred from 82 individualssequenced for 570-bp cytochrome c oxidase I (COI) sequences, and the subspecies distinctiveness. Circle size isproportional to the observed haplotype frequencies and black points represent unobserved haplotypes and potentialintermediates. Red, parsimony network of the 16 haplotypes observed in Cordulegaster boltonii: plain, Cordulegasterboltonii boltonii; dashed line, Cordulegaster boltonii algirica; cross hatched, Cordulegaster immaculifrons; squares,Cordulegaster boltonii iberica. Yellow: parsimony network of the 15 haplotypes observed in Cordulegaster bidentata: plain,Cordulegaster bidentata bidentata; dashed, Cordulegaster bidentata sicilica. Light blue: parsimony network of the threehaplotypes observed in Cordulegaster insignis. Brown: parsimony networks of COI 5 haplotypes observed in Cordulegasterhelladica: plain, Cordulegaster helladica helladica; dashed, Cordulegaster helladica buchholzi. For individual codes, seeTable 1.

56 E. FROUFE ET AL.


http://www.iucnredlist.org

Dommanget J-L, Ferreira S, Garrigós B, Jovic M,Kotarac M, Lopau W, Marinov M, Mihokovic N,Riservato E, Samraoui B, Schneider W. 2009. Atlas ofthe Odonata of the Mediterranean and North Africa.Libellula Supplement 9: 1–256.

Braby ML, Eastwood R, Murray N. 2012. The subspeciesconcept in butterflies: has its application in taxonomy andconservation biology outlived its usefulness? BiologicalJournal of the Linnean Society 106: 699–716.

Clement M, Posada D, Crandall KA. 2000. TCS: a com-puter program to estimate gene genealogies. MolecularEcology 9: 1657–1660.

Córdoba-Aguilar A, ed. 2008. Dragonflies and damselflies:model organisms for ecological and evolutionary research.Oxford: Oxford University Press.

Dijkstra K-DB, Kalkman VJ. 2012. Phylogeny, classificationand taxonomy of European dragonflies and damselflies(Odonata): a review. Organisms Diversity & Evolution 12:209–227.

Folmer O, Black M, Hoeh W, Lutz R, Vrijenhoek R. 1994.DNA primers for amplification of mitochondrial cytochromec oxidase subunit I from diverse metazoan invertebrates.Molecular Marine Biology and Biotechnology 3: 294–299.

Guindon S, Dufayard JF, Lefort V, Anisimova M,Hordijk W, Gascuel O. 2010. New algorithms andmethods to estimate maximum-likelihood phylogenies:assessing the performance of PhyML 3.0. Systematic Biology59: 307–321.

Habel JC, Dieker P, Schmitt T. 2009. Biogeographicalconnections between the Maghreb and the Mediterraneanpeninsulas of southern Europe. Biological Journal of theLinnean Society 98: 693–703.

Habel JC, Rödder D, Scalercio S, Meyer M, Schmitt T.2010. Strong genetic cohesiveness between Italy and theMaghreb in four butterfly species. Biological Journal of theLinnean Society 99: 818–830.

Hall TA. 1999. Bioedit: a user friendly biological sequencealignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symposium Series 41: 95–98.

Jaramillo-Correa JP, Grivet D, Terrab A, Kurt Y,De-Lucas AI, Wahid N, Vendramin GG,González-Martínez SC. 2010. The Strait of Gibraltar as amajor biogeographic barrier in Mediterranean conifers: acomparative phylogeographic survey. Molecular Ecology 19:5452–5468.

Kaliontzopoulou A, Pinho C, Harris DJ, Carretero MA.2011. When cryptic diversity blurs the picture: a cautionarytale from Iberian and North African Podarcis wall lizards.Biological Journal of the Linnean Society 103: 779–800.

Kalkman VJ, Boudot J-P, Bernard R, Conze K-J,De Knijf G, Dyatlova E, Ferreira S, Jovic M, Ott J,Riservato E, Sahlén G. 2010. European Red List of drag-onflies. Luxembourg: Publications Office of the EuropeanUnion.

Katoh K, Miyata T. 2005. MAFFT version 5: improvement inaccuracy of multiple sequence alignment. Nucleic AcidsResearch 33: 511–518.

Lohmann H. 1993. Revision der Cordulegastridae. 2.Beschreibung neuer Arten in der Gattungen Cordulegaster,Anotogaster, Neallogaster und Sonjagaster (Anisoptera).Odonatologica 22: 273–294.

Pilgrim EM, Roush SA, Krane DE. 2002. Combining DNAsequences and morphology in systematics: testing the valid-ity of the dragonfly species Cordulegaster bilineata. Heredity89: 184–190.

Posada D. 2008. jModelTest: phylogenetic model averaging.Molecular Biology and Evolution 25: 1253–1256.

Rambaut A, Drummond A. 2003. Tracer: MCMC traceanalysis tool. Oxford: University of Oxford.

Ronquist F, Huelsenbeck JP. 2003. MrBayes version 3.0:Bayesian phylogenetic inference under mixed models.Bioinformatics 19: 1572–1574.

Sambrook J, Fritsch EF, Maniatis T. 1989. Molecularcloning: a laboratory manual. New York, NY: Cold SpringHarbor Press.

Samraoui B, Boudot J-P, Ferreira S, Riservato E, JovicM, Kalkman VJ, Schneider W. 2010. The Odonata ofNorth Africa: Red List status and distribution. In: García N,Cuttelod A, Abdul Malak D, eds. The status and distributionof freshwater biodiversity in Northern Africa. Gland: IUCN,51–70.

Tamura K, Dudley J, Nei M, Kumar S. 2007. MEGA4:molecular Evolutionary Genetics Analysis (MEGA) softwareversion 4.0. Molecular Biology and Evolution 24: 1596–1599.

Templeton AR, Crandall KA, Sing CF. 1992. A Cladisticanalysis of Phenotypic Associations with haplotypesinferred from restriction endonuclease mapping and DNAsequence data. 3. cladogram estimation. Genetics 132: 619–633.

SUPPORTING INFORMATION

Additional Supporting Information may be found in the online version of this article at the publisher’s web-site:

Table S1. List of samples: species, sample collection code, collection sites (population, latitude, longitude andcountry), and GenBank accession codes.



Molecular phylogeny of the Western Palaearctic Cordulegaster taxa...

Documents

Transcript of Molecular phylogeny of the Western Palaearctic Cordulegaster taxa...