Table Of ContentRESEARCHARTICLE
Towards a DNA Barcode Reference Database
for Spiders and Harvestmen of Germany
JonasJ.Astrin1*,HubertHöfer2*,JörgSpelda3*,JoachimHolstein4*,SteffenBayer2,
LarsHendrich3,BernhardA.Huber1,Karl-HinrichKielhorn5,Hans-JoachimKrammer1,
MartinLemke6,JuanCarlosMonje4,JérômeMorinière3,BjörnRulik1,MaltePetersen1,
HannahJanssen1,ChristophMuster7
1 ZFMK:ZoologischesForschungsmuseumAlexanderKoenig,Bonn,Germany,2 SMNK:Staatliches
MuseumfürNaturkundeKarlsruhe,Karlsruhe,Germany,3 ZSM:ZoologischeStaatssammlungMünchen,
München,Germany,4 SMNS:StaatlichesMuseumfürNaturkundeStuttgart,Stuttgart,Germany,5 Karl-
a11111 HinrichKielhorn,Berlin,Germany,6 MartinLemke,Lübeck,Germany,7 ZoologischesInstitutundMuseum,
UniversitätGreifswald,Greifswald,Germany
*[email protected](JJA);[email protected](HH);[email protected](JS);joachim.
[email protected](JH)
Abstract
OPENACCESS
Citation:AstrinJJ,HöferH,SpeldaJ,HolsteinJ, AspartoftheGermanBarcodeofLifecampaign,over3500arachnidspecimenshavebeen
BayerS,HendrichL,etal.(2016)TowardsaDNA
collectedandanalyzed:ca.3300Araneaeand200Opiliones,belongingtoalmost600spe-
BarcodeReferenceDatabaseforSpidersand
cies(median:4individuals/species).Thiscoversabout60%ofthespiderfaunaandmore
HarvestmenofGermany.PLoSONE11(9):
e0162624.doi:10.1371/journal.pone.0162624 than70%oftheharvestmenfaunarecordedforGermany.Theoverwhelmingmajorityof
speciescouldbereadilyidentifiedthroughDNAbarcoding:mediandistancesbetweenclos-
Editor:MatjaKuntner,ScientificResearchCentreof
theSlovenianAcademyofSciencesandArt, estspecieslayaround9%inspidersand13%inharvestmen,whilein95%ofthecases,
SLOVENIA intraspecificdistanceswerebelow2.5%and8%respectively,withintraspecificmediansat
Received:June3,2016 0.3%and0.2%.However, almost20spiderspecies,mostnotablyinthefamilyLycosidae,
couldnotbeseparatedthroughDNAbarcoding(althoughmanyofthempresentdiscrete
Accepted:August25,2016
morphologicaldifferences).Conspicuouslyhighinterspecificdistanceswerefoundineven
Published:September28,2016
morecases,hintingatcrypticspeciesinsomeinstances.Anewprogramispresented:DiS-
Copyright:©2016Astrinetal.Thisisanopen
tatscalculatesthestatisticsneededtomeetDNAbarcodereleasecriteria.Furthermore,
accessarticledistributedunderthetermsofthe
newgenericCOIprimersusefulforawiderangeoftaxa(alsootherthanarachnids)are
CreativeCommonsAttributionLicense,whichpermits
unrestricteduse,distribution,andreproductioninany introduced.
medium,providedtheoriginalauthorandsourceare
credited.
DataAvailabilityStatement:Allrelevantdataare
withinthepaperanditsSupportingInformationfiles.
Sequencesareappendedasalignment,theirtrace
Introduction
filesareavailablefromBOLD(http://www.
boldsystems.org/).TheDiStatsscriptisavailable
Long-termmonitoringofbiodiversityisoneofthemostimportantchallengesinconservation
throughGitHub(https://github.com/mptrsen/distats)
biology.To evaluatetheconservationstatusandanthropogenicimpactofhabitats,sufficient
andthroughtheZFMKhomepage(www.zfmk.de/en/
research/research-centres-and-groups/distats). knowledgeonspeciescompositionofnaturalenvironmentsisneededonaregionallevel.For
manyifnotmostinvertebratetaxa,wearestillfarfromachievingthisgoal.Onepromising
Funding:TheGBOLprojectisfinancedbythe
approachtomeetthischallengeisDNAbarcoding[1],atechniquethatusestheeasytohomol-
GermanFederalMinistryofEducationandResearch
(BMBF#01LI1101).Partofthesequencescontributed ogize,well-quantifiable,discretetaxonomiccharacterscontainedinDNAsequencedatafor
byZSMresultedfromfundingbytheBavarianState standardized,rapid,andrelativelycheapspeciesidentification.DNAbarcodingdependson
PLOSONE|DOI:10.1371/journal.pone.0162624 September28,2016 1/24
DNABarcodingofSpidersandHarvestmeninGermany
MinistryofEducationandCulture,Scienceandthe lowlevelsofintraspecificvariationcoupledwithmarkedgeneticdifferentiationbetweenspe-
Arts(BarcodingFaunaBavarica,BFB).The cies(the'barcodinggap',investigatedinspidersin[2–4]).
sequencingworkofZSMwassupported,inpart,by
Withmorethan45,800describedspecies[5],spidersareamongthemostdiverseanimal
fundingfromtheGovernmentofCanadatoGenome
orders[6].Theyareabundantinallterrestrialhabitats.Asubiquitouspredators,theyoccupya
CanadathroughtheOntarioGenomicsInstitute,while
theOntarioMinistryofResearchandInnovationand keypositioninfoodwebs.Manyspeciesshowpreferencesforspecifichabitatstructuresor
NSERCsupporteddevelopmentoftheBOLD environmentalfactors,e.g.temperature,humidity,shading[7],whichturnsthemintopotential
informaticsplatform.Thefundershadnoroleinstudy indicators[8].Easytoobserveanddocument,spidersareseenasamodelgroupforecological
design,datacollectionandanalysis,decisionto
studies[9,10].
publish,orpreparationofthemanuscript.
ThespiderfaunaofGermany,comprisingapproximately1000species[11],iswellknown,
CompetingInterests:Theauthorshavedeclared andchecklistsandredlistsofendangeredspecieshavebeenpublishedforGermanyandmost
thatnocompetinginterestsexist.
ofitsfederalstates(see[11]andreferencestherein).The'ArachnologischeGesellschafte.V.'
(www.arages.de)offersregularlyupdatedoccurrencemaps,basedonasteadilygrowingdata-
base.Therefore,spidersareregularlyusedinhabitatassessments,biodiversityinventories,and
ecologicalstudies(e.g.[12–18]).Spidersareparticularlypromisingasindicatorsofsustainable
forestmanagement[19],habitatstructure[20],successionalstages[21,22],orconservation
value[23,24].Therehavebeenseveralattemptstoclassifyspidersaccordingtotheirhabitator
nichepreferencesinGermanyorCentralEurope[7,25–30]andtousethesedatatoclassify
habitatsorassesshabitatqualitybyidentifyingtheproportionofrare,endangered,stenotopic,
orcharacterspecies(e.g.[18,31,32]).IdentificationofGermanspidersisfacilitatedbythe
onlinekeysforspidersofEuropeatwww.araneae.unibe.ch[33].However,morphologicaliden-
tificationtospecieslevelrequiresadultspecimensinmostinstances.About80–200spiderspe-
ciescanoccurinanear-naturalhabitatinGermany,ofwhichonlyasmallfractioncanbe
directlyrecordedandidentifiedinthefield(pers.obs.,H.Höfer,C.Muster).Foranecolog-
icallymeaningfulassessmentoraclosetocompleteinventory,muchmoretimeneedstobe
investedtocapture,process(oftenmeaningdissectionofsexualorgans)andidentifythe
(adult)spiders,requiringconsiderableexpertise.Regularly,severalspecimensremainthathave
tobecheckedbythefewavailabletaxonomicspecialistswithsufficientknowledgeonmorpho-
logicalvariabilityintherespectivespeciesandwithaccesstoreferencecollections.
Withsome6500speciesworldwide,harvestmen(Opiliones)constitutethethird-largest
orderofarachnids[34].Currently,52specieshavebeenrecordedfromGermany[35].The
omnivorousharvestmenconstitutearegularcomponentofterrestrialfaunas,withhighestden-
sitiesindampandshadedhabitats[36].Theiruseinappliedandecologicalstudiesisexplained
bytheexistenceofbothstenotopicspecieswithstrictmicrohabitatrequirements(andoften
limitedgeographicranges)andinvasivespeciesthatexhibitimmensecolonizationpotential
[37,38].DeterminationofmostGermantaxaisreliablyachievableusingtheworkof[39].
However,recentstudieshaverevealedhighlevelsofcrypticdiversityinCentralEurope[40–
42],suggestingapromisingperspectiveforDNAbarcodinginthistaxon.
TheuseofmitochondrialCOIbarcodes[43]fromanextensivereferencedatabaseofspider
andharvestmenspecieswillaidnon-specialistsinthedeterminationofthesegroups.Species
thathavehithertobeenproblematicorevenimpossibletoidentifymorphologically–eitherin
generalorforaparticularsex–maybereliablydiscriminated.EventhoughnotfrequentinGer-
many, therearestillmanyspiderspeciesinwhichoneofthe(dimorphic)sexesisstill
unknown,andbarcodingcanprovidethelinkbetweensexes(demonstratede.g.in[44]).More-
over,disputedinstancesofsynonymymayberesolved[45].Notleast,aconsiderableadvantage
ofbarcodingisthepossibilitytoidentifyjuvenilespecimens[3,46–49].Thiswillnotonlymake
inventoriesmorecomplete,butwillalsoallowspecies-levelinclusionofjuvenilesintoecologi-
calanalyses.Tapping intothisrichmaterialresourcewillallowstudyingmoreecologicalques-
tionswithoutthenecessityforexhaustiveandexpensivesampling.Afuturebroadapplication
ofroutineDNAbarcodinginspidersisfacilitatedthroughmass-trapping,sincesomeofthe
PLOSONE|DOI:10.1371/journal.pone.0162624 September28,2016 2/24
DNABarcodingofSpidersandHarvestmeninGermany
samplingsolutionsemployedintrapspreserveDNAwellenoughforbarcoding[50,51].The
methodfurtherholdsthepotentialtorevealcrypticspeciesortoidentifycaseswheremorpho-
logicalplasticitymayhavebeenover-interpreted.Barcodingmaythusactasacatalystfor
alphataxonomy[52].Whileintrogressionevents[53],retentionofancestralpolymorphisms
[49],nuclearmitochondrialpseudo-genes[54]orendosymbionts(Wolbachia bacteriaetc.)
[55,56]allposepotentialproblemstoDNAbarcodingapproaches,agrowingnumberofstud-
iesshowthegeneralfeasibilityofDNAbarcodingforarachnids[2–4,46,47,57–66].
TheGermanBarcodeofLife(GBOL)campaignisimplementedbyanationalnetworkofca.
20biodiversityresearchinstitutionsandmorethan200taxonspecialists[67].Itpursuesthe
goaltoestablishaDNAbarcodelibraryofasmanyanimal,fungalandplantspeciesaspossible
thatoccurinGermany.Theprojectaimsatcollecting,ifpossible,tenspecimensperspecies,
fromlocationsasdistinctaspossiblethroughoutthecountryinordertocapturegeneticvari-
ability.Somespecieswithwiderrangesmayalsoincludespecimenscollectedinneighboring
countries.
NaturalhistorycollectionsconstitutethecoreinfrastructureofGBOL,takingintoaccount
thatbarcodingprojectsproduceavaluablelegacyofvouchers(morphologicalspecimensand
molecularsamplesalike)whichbecomerelevantinsubsequentstudiesduetothehighquality
oftheunderlyingtaxonomicassignmentsandgranularityofthemetadata.Thesevouchers
formthephysicalfoundationthatfuturemonitoringprojectswillbebasedon,warrantingcon-
tinuoustestability,validation,andcoherentexpansionofthebarcodingreferencedatabase–
ideallyforcenturiestocome.
WithintheGBOLconsortium,arachnidshavereceivedwideattention,asnolessthanfour
GBOLinstitutesandtheirrespectiveexternalarachnologistpartnerscollaborateintensivelyon
compilinganationalmolecularinventoryofspidersandharvestmen:StaatlichesMuseumfür
NaturkundeKarlsruhe(SMNK),StaatlichesMuseumfürNaturkundeStuttgart(SMNS),Zool-
ogischesForschungsmuseumAlexanderKoenig(ZFMK),andZoologischeStaatssammlung
München(ZSM).AcariarebeinginvestigatedinanotherGBOLsubproject(bySenckenberg
MuseumfürNaturkundeGörlitz,SMNG).
Forspiders,country-focused,taxonomicallybroadbarcodingdatasetshavesofarbeenpub-
lishedforCanada[61](1018speciescovered),forSloveniaandforSwitzerland[66](together
298species)and,inapilotproject,fortheNetherlands[68](31species)(foralistofongoing
Europeanprojects,seehttp://www.araneae.unibe.ch/barcoding/content/15/Barcoding-of-
European-spiders).Thepresentstudycontributesthefirstdatasetofaspiderbarcodingcam-
paignforGermanyandthefirstdatasetworldwideofthiskindforharvestmen.
MaterialsandMethods
Sampling
Forthisstudy,3537arachnidspecimens,3339Araneaeand198Opiliones,weresampledfrom
Germany(91%ofthematerial)andneighboringcountries.WithinGermany,24%ofthespeci-
menswerecollectedinBaden-Württembergand13%inSchleswig-Holstein.MostotherGer-
manstateswererepresentedby6–10%oftheGermanspecimenseach.Thuringia(0.1%),
Hesse(1%)andRhineland-Palatinate(2%)werelesswellrepresented,aswerethecity-states
andthecomparativelysmallSaarland(1%).Fig1illustratesthesamplingpattern.
SMNKandexternalpartnersareresponsibleforandcontributed14%ofthespecimens,
SMNSandpartners10%,ZFMKandpartners56%,ZSMandpartners20%.
To date598morphologicalarachnidspecies(561spp.inspidersvs.37inharvestmen)in
269genera(246vs.23)and50families(44vs.6)couldbeintegrated.Settingthisintorelation
PLOSONE|DOI:10.1371/journal.pone.0162624 September28,2016 3/24
DNABarcodingofSpidersandHarvestmeninGermany
Fig1.Geographicsamplingofarachnidspecimensunderlyingthepresentstudy.ImageproducedusingGPSVisualizer
(www.gpsvisualizer.com).
doi:10.1371/journal.pone.0162624.g001
PLOSONE|DOI:10.1371/journal.pone.0162624 September28,2016 4/24
DNABarcodingofSpidersandHarvestmeninGermany
Fig2.Numberofspeciessampledperfamily(excludingfamiliesrepresentedonlyby1or2speciesinthisstudy).Numbersabovebarsare
percentagesshowingspeciescoverageforGermany,asderivedfromthechecklistsmentionedinthetext.Familynamesprefixedwith"O."belongto
Opiliones,allothersarespiderfamilies.
doi:10.1371/journal.pone.0162624.g002
withtheGermanchecklists[11,35],speciescoverageis57%forspidersand71%forharvest-
men.SpeciesnumbersareplottedforthemorefrequentfamiliesinFig2.
Thespeciesinthedatasetwererepresentedby6individualsonaverage(median:4).Almost
19%,i.e.112species,were'singletons'.With48specimens,Pardosalugubriswasthespecies
withmostindividuals;allotherspecieswererepresentedby30orfewerindividuals(seeFig3).
Mostindividuals(98%)werecollectedspecificallyforGBOLbetweentheyears2011and
2015.Theoldestspecimenprocessedinthisstudywascollectedin2003.
Collectingwasmostlydonebyhand,andmostspecimenswerekilledandpreserveddirectly
in96%or100%ethanol.7%ofthespecimenswereinitiallypreservedin70%water-dilutedeth-
anoland8%werecollectedinpropyleneglycol.Thelatterwasusedascapturefluidinpitfall
traps;soonafteridentification,tissueforDNAextractionwastransferredtoabsoluteethanol.
Allmaterialusedinthisstudyispropertyofthefederalstatesoftheinvolvedinstitutions.
Materialacquiredbytheseinstitutionsisonlyacceptedafteracheckthatitwascollectedin
compliancewithnationalandinternationallaws,regulationsandconventionsandthatthe
materialisfreefromthirdpartyrights.Furthermore,inordertobecomecertifiedasaGBOL
collector,itisrequiredtoaccepttheproject'sgeneraltermsandconditions,whichdemand
abidingbytheregulationsoftheConventiononBiologicalDiversityandnationallegislations.
Fieldworkpermitswereissuedbythefollowingauthorities:BayerischesStaatsministeriumfür
PLOSONE|DOI:10.1371/journal.pone.0162624 September28,2016 5/24
DNABarcodingofSpidersandHarvestmeninGermany
Fig3.Numberofspecimenssampledperspecies.19%ofthespecieswere'singletons',whilethemedianrepresentationlayat4individualsper
species.
doi:10.1371/journal.pone.0162624.g003
UmweltundGesundheit,München;RegierungspräsidiumStuttgart;Regierungspräsidium
Karlsruhe;Struktur-undGenehmigungsdirektionKoblenz;KreisverwaltungRhein-Sieg-Kreis,
AmtfürNatur-undLandschaftsschutz;AmtfürUmwelt,Verbraucherschutz undLokale
Agenda,UntereLandschaftsbehörde,Bonn;NationalparkamtMüritz,Hohenzieritz;Biosphär-
enreservatsverwaltungNiedersächsischeElbtalaue,Hitzacker;NationalparkforstamtEifel,
Schleiden-Gemünd;AmtfürdasBiosphärenreservatSüdost-Rügen,Putbus;Landesamtfür
Umwelt,NaturschutzundGeologieMecklenburg-Vorpommern,Güstrow;Landesamtfür
Landwirtschaft,UmweltundländlicheRäumeSchleswig-Holstein,Flintbek;LandratKreis
HerzogtumLauenburg;LandratKreisRendsburg-Eckernförde,FachdiensteuntereNat-
urschutzbehörde.Thepermitscoverstateforests,publiclandandprotectedareasaswellasthe
fivespeciesofArachnidaprotectedinGermany:Arctosacinerea,Dolomedesfimbriatus,Dolo-
medesplantarius,EresuscinnabarinusandPhilaeuschrysops.
Altogether,over100collectorscontributedmaterial.Fielddataforallanalyzedspecimens
canbeaccessedinS1Table.Juvenilespecimensanalyzedbelongtotaxathatareeasilyidentifi-
ablealsoinjuvenilestage(e.g.basedoncoloration)orforwhichproblematic('look-alike')con-
genersdonotoccurinthestudyarea.Juvenilesthatclusteredconspicuouslyinthetreewere
removedfromthedataset.
Allmorphologicalspecimenvouchersandalsomolecularvouchers(DNAandoftentissue)
aredepositedatandareavailablefromthefollowingfourGermanpubliccollections(perma-
nentrepositories):StaatlichesMuseumfürNaturkundeKarlsruhe(SMNK),Staatliches
MuseumfürNaturkundeStuttgart(SMNS),ZoologischesForschungsmuseumAlexanderKoe-
nig(ZFMK),Bonn,ZoologischeStaatssammlungMünchen(ZSM).Allvouchernumbersare
giveninS1Table.ThevoucherIDsinS1Table aswellasthenamesinthetreesincludethe
institutionalcode,sothattheassociationofagivensampletooneoftheGBOLpartnerinsti-
tutescanbeeasilyestablished.
SequencedataareavailableonBOLD[69]viaDOIdx.doi.org/10.5883/DS-GBOLARA,and
onGenBank.Specimendatawillalsobeaccessible,alongsidespecimenimages,throughthe
GBOLportal(www.bolgermany.de).
PLOSONE|DOI:10.1371/journal.pone.0162624 September28,2016 6/24
DNABarcodingofSpidersandHarvestmeninGermany
Molecularmethods
Analyseswereperformedmostlyinthreeseparatelaboratories:ZFMK,SMNS,andCanadian
CentreforDNABarcoding(CCDB)inGuelph.SMNSandZFMKusedtheirownfacilities(up
tothepointofsequencing),SMNKsampleswereprocessedatbothSMNSandZFMK,ZSM
samplesatCCDB.
Total genomicDNAwasusuallyisolatedfromlegs.Inverysmallspecimens(especially
manyLinyphiidae)atZFMKandSMNS,DNAwasextractednon-destructivelyfromwhole
specimenswhichwererecoveredafterlysis(cf.[70]).
AtZSM,singlelegswereremovedfromeachspecimenandsentin96welllysisplatesto
CCDBforstandardizedDNAextraction,PCRamplificationandbidirectionalSangersequenc-
ing.CCDBlabprotocolsareavailableunderwww.ccdb.ca/resources.php.
AtZFMKandSMNS,silica-basedmethodswereemployedtoextractDNA.SMNSfollowed
theprotocolby[71]withPallAcroPrepTM96filterplates(PallCorporation,PortWashington,
NY, USA),whileaQiagen(Hilden,Germany)BioSprint96magneticbeadextractorandcorre-
spondingkitswereusedatZFMK.
Polymerasechainreactionforthe5'partofthemitochondrialcytochromecoxidasesubunit
1(COI)genewascarriedout,atZFMK,intotalreactionmixesof20μl,including2μlofundi-
lutedDNAtemplate,0.8μlofeachprimer(10pmol/μl),andstandardamountsofthereagents
providedwiththe'MultiplexPCR'kitfromQiagen(Hilden,Germany).AtSMNS,PCRreac-
tionsof25μlvolumecontained4μlDNA,5unitsofTaq KAPA extrapolymerase(KAPA-
BIOYSYTEMS,Boston,USA),10μlof5xKAPA Taq extrabuffer,3μlofMgCl (25mM;both
2
solutionsprovidedbythemanufacturer),1μlofeachprimer(10pmol/μl),and1μlof10mM
dNTPmix(Bioline,Luckenwalde,Germany).
ThePCRprimersused(alsoforsequencing)aregiveninTable 1.
Thermalcyclingwasperformed,atZFMK,onAppliedBiosystems2720ThermalCyclers
(LifeTechnologies,Carlsbad,CA,USA),usingaPCRprogramwithtwocyclesets,asa
Table1. Listofprimersusedforamplificationandsequencingofthe5'partofthemitochondrialCOIgene.
Primername Sequence Publication Usedat
LCO1490 5'-GGTCAACAAATCATAAAGATATTGG Folmeretal.1994 SMNS,CCDBforZSM,ZFMK
HCO2198 5'-TAAACTTCAGGGTGACCAAAAAATCA Folmeretal.1994 SMNS,CCDBforZSM,ZFMK
LepF1 5'-ATTCAACCAATCATAAAGATATTGG Hebertetal.2004 CCDBforZSM
LepR1 5'-TAAACTTCTGGATGTCCAAAAAATCA Hebertetal.2004 CCDBforZSM
C_LepFolF cocktailofLepF1andLCO1490 www.boldystem.org/index.php/Public_Primer_PrimerSearch CCDBforZSM
C_LepFolR cocktailofLepR1andHCO2198 www.boldystem.org/index.php/Public_Primer_PrimerSearch CCDBforZSM
LCO1490-JJ 5'-CHACWAAYCATAAAGATATYGG Astrin&Stüben2008 ZFMK
HCO2198-JJ 5'-AWACTTCVGGRTGVCCAAARAATCA Astrin&Stüben2008 ZFMK
LCO1490-JJ2 5'-CHACWAAYCAYAARGAYATYGG new ZFMK
HCO2198-JJ2 5'-ANACTTCNGGRTGNCCAAARAATCA new ZFMK
LCO1490-JJ4a 5'-CNACNAAYCAYARRGAYATYGG new ZFMK
HCO2198-JJ4a 5'-AIACYTCNGGRTGICCAAARAATC new ZFMK
LCO1490-JJ4 5'-CIACIAAYCAYAARGAYATYGG new ZFMK
HCO2198-JJ4 5'-ANACTTCNGGRTGNCCAAARAATC new ZFMK
Species—alsomanyothersthanarachnids—withstronglymodified bindingsitescouldusuallybesuccessfullyamplifie datZFMKwithasetofnewly
(manually)designed,highlydegenerateprimers(mostoftenusingcombinationLCO1490-JJ2&HCO2198-JJ2).ThecombinationLCO1490-JJand
HCO2198-JJconstitutesthestandardsetofprimersusedatZFMK.ThestandardsetofprimersusedatCCDBforZSMwasthecombinationC_LepFolF
andC_LepFolR,acocktailconsistingoftheprimerslistedabove.
doi:10.1371/journal.pone.0162624.t001
PLOSONE|DOI:10.1371/journal.pone.0162624 September28,2016 7/24
DNABarcodingofSpidersandHarvestmeninGermany
combinationofa'touchdown'anda'step-up'routine:firstcycleset(15repeats):35sdenatur-
ationat94°C,90sannealingat55°C(−1°Cpercycle)and90sextensionat72°C.Secondcycle
set(25repeats):35sdenaturationat94°C,90sannealingat45°C,and90sextensionat72°C.
AtSMNS,PCRamplificationwascarriedoutinaLabcyclerbySensoQuest(Göttingen,Ger-
many).PCRconditionswere:35cyclesof60sdenaturationat93°C,90sannealingat50°Cand
60sextensionat72°C.
PCRproductsweresubsequentlysentforbidirectionalSangersequencingtovariouscom-
panies:ZFMKtoBGI(HongKong,China)andMacrogen(Amsterdam,Netherlands),SMNS
toLGCGenomics(Berlin,Germany)in2013andfrom2014ontoGATCBiotech(Konstanz,
Germany).
DNAsequencealignmentwasperformedusingparallelizedMAFFTver.7.123[72].PAUP(cid:3)
ver.4.0b10[73]wasusedforp-distancetransformationsandforevaluatingbasecomposition
andinformationcontent.Statisticalparsimonynetworks[74]werecalculatedwiththeTCS
algorithm[75]inPopART(http://popart.otago.ac.nz).Statisticalevaluationofthedatawas
performedusingSPSS,R(boxplots),SpeciesIdentifierver.1.7.7–3[76](extractionof'splits'
and'lumps'),andthePerlscriptDiStats(intraspecificdistances,individualizeddataonclosest
speciespairsandonmostdistantcongeners).DiStatshasbeendevelopedforthisstudyandis
available,includingdocumentation,underGitHub(https://github.com/mptrsen/distats)and
throughtheZFMKhomepage(www.zfmk.de/en/research/research-centres-and-groups/
distats).ThescriptusesFASTA asinputformat,calculatesp-distancesorK2Pdistancesand
canbeparallelizedinordertoprocesslargedatasets.Itcanproducetwooutputfiles:atable
withstatisticsforeachspeciesandoptionallyalsothematrixofallpairwisedistancesinthe
dataset.Foranalignmentcontaining1000COIbarcodesequences,theanalysiswilltake
around6minuteswhenusingasinglethread(ona3.4GHzprocessor).DiStatshasanalgorith-
miccomplexity(O)ofapproximatelyO(n2),whichmeansthatruntimeincreasesexponen-
tiallywiththenumberofinputsequences(n).UsingmultipleCPUthreadsreducestherun
timebyafactorof1/c,wherecisthenumberofthreads.
PAUP wasalsousedforreconstructionofapheneticneighbor-joining(NJ;[77])treeasa
quickmolecularidentificationcheck.PhylogeneticreconstructionsusingMaximumLikelihood
(ML;[78])wereperformedwithRAxMLver.7.3.0[79].EvolutionarymodelselectionfortheML
analysiswasimplemented,usinghierarchicallikelihoodratiotesting,inModelGeneratorver.
0.85[80]andindicatedGTR+I+Γasthebest-fittingmodel[81].TheCOIdatasetwasparti-
tionedtotreat3rdcodonpositionsseparatelyfrom1stand2ndpositions.Theanalysiswasrun
for1milliongenerationsandincluded1000bootstrapreplicates.FortreerootingpurposesinNJ
andMLanalyses,wechoseamitesequencefromBOLDasoutgroup(seeS1Alignment).
Results
AverageCOIsequencelengthforthe3537sequenceswas650bp.To accommodatemany
slightlyshortersequences,whileavoidinggeneticdistanceartifacts,alignmentlengthwassetto
653bp.Sequencesshorterthan500bpwereexcludedfromtheanalysis.Theshortestincluded
sequencewascomposedof509residues.
Thedatasetcomprised2099distincthaplotypes,meaningthat1438sequenceswerenon-
unique.
Amongnucleotides,therewasacompositionalbiastowardsAT: 67.5%,whichiscloseto
levelspreviouslyreportedforspiders(e.g.[4,82,83]).Indetail,overallbasecompositionwas:A
25.3,C13.3,G19.2,T42.2%.
Altogether,5,572,791pairwisedistanceswerecomputedforspiders;ofthese,17,867were
intraspecificdistances.ForOpiliones,therewere19,503pairwisedistancesofwhich896were
PLOSONE|DOI:10.1371/journal.pone.0162624 September28,2016 8/24
DNABarcodingofSpidersandHarvestmeninGermany
Table2. Estimatorsusedtocharacterizegeneticdistancestructureinthedataset.
[%] intraspecific interspecific
median mean range 95thperc. median mean range 5thperc.
Araneae 0.3 0.7 0.0–10.1 2.5 17.5 17.4 0*-28.2 13.6
Opiliones 0.2 1.3 0.0–8.9 8.1 19.3 19.4 7.0–30.1 13.6
Aran.K2P 0.3 0.7 0.0–11 2.6 20.0 20.0 0.0–35.6 15.0
Opil.K2P 0.2 1.3 0.0–9.5 8.7 22.6 22.7 7.5–38.6 15.1
Theuppertworowsindicateuncorrecteddistancesforspidersandharvestmen,respectively,whilethethirdandfourthrowsgiveK2Pdistances(asrequired
forabarcodedatarelease).Medianandmeandistancesaregivenforbothintraspecific andinterspecific distances,alongwiththerangebetweenthe
smallestandlargestobservationintherespectivedatacategory.*:Therewerecasesofsharedhaplotypesamongspecies,seetext.
doi:10.1371/journal.pone.0162624.t002
distancesbetweenconspecificspecimens.Table2givesanoverviewofintra-andinterspecificdis-
tances,separatedbyorder.Table3summarizesdistancesforallclosestspeciespairsandforthe
mostdistantcongenericpairs;S2andS3Tablesillustratetheseinmoredetail,givingindividual
statisticsbytaxon.S2Table(spiders)andS3Table(harvestmen)furthermoreindicateintraspecific
distancerangesandcentraltendenciesforallanalyzedspecies.S4Tableindividuallyliststhehigh-
estintraspecificdistancesinthedataset,whileS5Tablegivesthelowestinterspecificdistances.
Therangecoveredbyintraspecific(p-)distanceswassimilarforspiders(0–10%)andhar-
vestmen(0–9%),withanarithmeticmeanof0.7%inspidersand1.3%inharvestmen.The
influenceofhighoutlierswasstrongerinthecomparativelysmallharvestmandataset(median
at0.2%vs.meanat1.3%),causedmostlybythedeepsplitswithinMitopusmorio(Fabricius,
1779)andPhalangiumopilioLinnaeus,1758(bothdiscussedbelow),butalsoinNemastoma
lugubre(Müller,1776)(seeS4Table).
Theinterspecificdistancerangevariedforthetwoarachnidorders:0.0–28%inspiders,
7–30%inharvestmen.Infourcases,haplotypesweresharedamongnominalspiderspecies
(seeS5Table,alldiscussedbelow):EnoplognathalatimanaHippa&Oksala,1982/E.ovata
(Clerck,1757);Pardosalugubris(Walckenaer, 1802)/P.saltansTöpfer-Hofmann, 2000;Tibel-
lusmaritimus(Menge,1875)/T.oblongus(Walckenaer, 1802);Xysticusaudax(Schrank,
1803)/X.cristatus(Clerck,1757).Interspecificarithmeticmeanswere17%forspiders,19%
forharvestmen.
Table3. Statisticsforclosestspeciespairsandmostdistantcongenericpairs.
[%] closestspeciespairs mostdistantcongenerpairs
range median(ofalldist.forspeciespairs) range median(oflargestdistances)
Araneae 0.0*-20.1(22.0) 9.2 1.8–20.2 11.8
Opiliones 7.0–21.8(22.8) 13.8 12.6–18.8 14.9
AraneaeK2P 0.0–21.0(28.6) 9.9 1.9–23.8 12.8
OpilionesK2P 7.5–24.0(29.0) 15.5 13.7–22.0 16.5
Theuppertworowsindicateuncorrecteddistancesforspidersandharvestmen,respectively,whilethethirdandfourthrowsgiveK2Pdistances(asrequired
forabarcodedatarelease).Therangefortheclosestspeciespairsindicatestheminimumandmaximumamongallsmallestpairwisedistancesbetween
closestspeciespairs.Ifaclosestspeciespairisrepresentedbyseveralindividuals,theremaybelargerdistancesaswell:themaximalclosestspecies
distanceintherespectivedatasetisindicatedinparentheses.Whilefortheclosestspeciespairsnoclassific atorial(genus)backgroundinformationwas
used,thelasttwocolumnsinthistableorientthemselvesatdistancesfromrepresentativeswithinthesamegenus(butdifferentspecies).Therangeforthe
mostdistantcongenerpairsextractstheextremesamongthesemaximalpairwisedistancesbetweenfarthestcongenericspeciespairs.Informationonthe
individualclosestspeciespairsandontherespectivemostdistantcongenericspeciespairsisgiveninS2TableforspidersandS3Tableforharvestmen.
*Therewerecasesofsharedhaplotypesamongspecies,seetext.
doi:10.1371/journal.pone.0162624.t003
PLOSONE|DOI:10.1371/journal.pone.0162624 September28,2016 9/24
DNABarcodingofSpidersandHarvestmeninGermany
Fig4.Boxplotofp-distancesfortheorderAraneae.Sortedbydistancecategory:betweenspecimensofdifferentgenera(allogeneric),between
specimensbelongingtodifferentspecies,buttothesamegenus(congeneric),andbetweenspecimensthatbelongtothesamespecies(intraspecific).
Boxesindicateinterquartilerange(IQR:betweenupper[Q3]andlower[Q1]quartile).Blackbarsdesignatemedians,whiskersindicatevalueswithin
1.5×IQRbeneathQ1or1.5×aboveQ3.Circlesdepictoutliers(aboveorbelow1.5×IQR).
doi:10.1371/journal.pone.0162624.g004
Meanintraspecificdistancesvariedconsiderablyamongfamilies.Forthebest-represented
spiderfamilies,theselaybetween0.4%(Agelenidae,Lycosidae,Philodromidae)and1.0%(Tet-
ragnathidae).Clubionidae,LinyphiidaeandTheridiidaehadmeanintraspecificdistancesof
0.6%,Gnaphosidae0.7%,Araneidae0.8%,SalticidaeandThomisidae0.9%.
Table 2andtheboxplots(Figs4and5)indicatethatauniversalbarcodinggapisabsent
fromthedataset.However,mostofthespeciesseparatewell;whenignoringthe5%most
extremeoutliers(ahypotheticalscenarionotsurpassingtheusualsignificancethreshold),the
barcodinggapforharvestmenwouldspan5.5%andforspiderseven11%(seeTable 2).The
mediandistanceforclosestspeciespairswas9%inspidersand13%inharvestmen.
ApheneticreconstructionusingNeighborJoining(NJ;S1Fig)andaphylogeneticrecon-
structionusingMaximumLikelihood(ML;S2Fig)deliveredtreesinwhichthespeciesover-
whelminglyformedmonophyleticclusters(butseediscussion–severalofthecaseswith
conspicuousdistanceswerealsorecoveredasparaphyleticandpolyphyletic).MLbootstrap
analysispredominantlyindicatedveryhighsupportforspecies-levelnodes,butdidusuallynot
allowmuchinsightintodeepertreetopology.
Discussion
ManypreviousstudiesusingCOIhaveshownthatspeciesdifferentiationviaDNAbarcoding
isgenerallyfeasibleandpromisinginarachnids(mostlyspiders:e.g.[2–4,46,47,57–66]–butsee
PLOSONE|DOI:10.1371/journal.pone.0162624 September28,2016 10/24
Description:As part of the German Barcode of Life campaign, over 3500 arachnid .. and the Perl script DiStats (intraspecific distances, individualized data on closest .. See S1 Table for more details on individual specimens in the tree. (PDF).