Phylogenetic analysis of cryptic speciation in the polychaete Pygospio elegans

This is an elec onic ep in o he o iginal a icle. This ep in may di e om he o iginal in pagina ion and ypog aphic de ail. Au ho (s): Ti le: Yea : Ve sion: Please ci e he o iginal e sion: All ma e ial supplied ia JYX is p o ec ed by copy igh and o he in ellec ual p ope y igh s, and duplica ion o sale o all o pa o any o he eposi o y collec ions is no pe mi ed, excep ha ma e ial may be duplica ed by you o you esea ch use o educa ional pu poses in elec onic o p in o m. You mus ob ain pe mission o any o he use. Elec onic o p in copies may no be o e ed, whe he o sale o o he wise o anyone who is no an au ho ised use . Phylogene ic analysis o c yp ic specia ion in he polychae e Pygospio elegans Kesäniemi, Jenni; Rawson, Paul D.; Lindsay, Sa a M.; Kno , Emily Kesäniemi, J., Rawson, P., Lindsay, S., & Kno , E. (2012). Phylogene ic analysis o c yp ic specia ion in he polychae e Pygospio elegans. Ecology and E olu ion, 2 (5), 994-1007. doi:10.1002/ece3.226 Re ie ed om h p://onlinelib a y.wiley.com/doi/10.1002/ece3.226/ ull 2012 Phylogene ic analysis o c yp ic specia ion in he polychae e Pygospio elegans J. E. Kes¨ aniemi1,P.D.Rawson 2,S.M.Lindsay 2& K. E. Kno 1 1Depa men o Biological and En i onmen al Science, Uni e si y o Jy ¨ askyl¨ a, P.O. Box 35, FI-40014, Finland 2School o Ma ine Sciences, Uni e si y o Maine, O ono, Maine 04469-5751 Keywo ds COI, de elopmen al mode, la ae, popula ion s uc u e. Co espondence Jenni E. Kes¨ aniemi, Depa men o Biological and En i onmen al Science, Uni e si y o Jy ¨ askyl¨ a, P.O. Box 35, FI-40014, Finland. Tel: +358 40 805 3872; E-mail: [email p o ec ed] Funded by a g an om he Jenny and An i Wihu i Founda ion and he Cen e o Excellence in E olu iona y Resea ch (Uni e si y o Jy ¨ askyl¨ a). K. E. K. acknowledges ui ul discussions wi h membe s o he Ma ie Cu ie Ini ial T aining Ne wo k Specia ion unded by he Eu opean Union. Recei ed: 26 Oc obe 2011; Re ised: 16 Janua y 2012; Accep ed: 19 Janua y 2012 Ecology and E olu ion 2012; 2(5): 994–1007 doi: 10.1002/ece3.226 Abs ac De elopmen in ma ine in e eb a e species can ake place h ough a a ie y o modes and la al o ms, bu wi hin a species, de elopmen al mode is ypically uni- o m.Poecilogony e e s o hep esenceo mo e hanonemodeo de elopmen wi hin a single species. T ue poecilogony is a e, howe e , and in some cases, ap- pa en poecilogony is ac ually he esul o a ia ion in de elopmen mode among ecen ly di e ged c yp ic species. We used a phylogene ic app oach o examine whe he poecilogony in he ma ine polychae e wo m, Pygospio elegans,is he e- sul o c yp ic specia ion. Popula ions o wo ms iden i ied as P. elegans exp ess a a ie y o de elopmen al modes including plank onic, b ooded, and in e media e la ae; hese modes a e ound bo h wi hin and among popula ions. We examined sequence a ia ion among pa ial mi ochond ial cy och ome c oxidase subuni I sequences ob ained o 279 indi idual wo ms sampled ac oss b oad geog aphic and en i onmen al scales. Despi e a la ge numbe o unique haplo ypes (121 haplo- ypes om 279 indi iduals), sequence di e gence among Eu opean samples was low (1.7%) wi h mos o he sequence a ia ion obse ed wi hin popula ions, el- a i e o he a ia ion among egions. Mo e impo an ly, we obse ed common haplo ypes ha we e widesp ead among he popula ions we sampled, and he wo mos common haplo ypes we e sha ed be ween popula ions di e ing in de elop- men al mode. Thus, ou esul s suppo an ea lie conclusion o poecilogony in Pelegans. In addi ion, p edominan ly plank onic popula ions had a la ge numbe o popula ion-speci ic low- equency haplo ypes. This inding is la gely consis en wi h in e species compa isons showing high di e si y o species wi h plank onic de elopmen al modes in con as o low di e si y in species wi h b ooded de elop- men al modes. In oduc ion Mos ma ine in e eb a es ha e complex li e cycles and show a di e se ange o la al de elopmen al modes. De elopmen- al mode is o en de ined as disc e e ca ego ies desc ibing cha ac e is ics o la ae, o la al ypes (Le in and B idges 1995). Fo example, la ae can be plank onic (pelagic) o ben hic, eeding o non eeding, b ooded o ee-li ing, and a combina ion o mul iple desc ip o s is o en necessa y o a comple e de ini ion o de elopmen al mode (e.g., McEdwa d and Janies 1993; Collin 2003; Ra and By ne 2006). De el- opmen al mode is an impo an aspec o in e eb a e li e his o ies, wi h wide- anging consequences a ec ing, o ex- ample, de elopmen ime, mo ali y, and dispe sal po en ial (Le in and B idges 1995). Unde s anding he consequences and e olu ion o di e en de elopmen al modes is, on one hand, aided by ou endency o ca ego ize i as disc e e ypes. On he o he hand, such de ini ions may also lead us o o e - look in e media e o acul a i ely a ying o ms ha do no i de ini ions o disc e e de elopmen al modes (Allen and Pe ne 2007). Many di e en de elopmen al modes may be obse ed wi hin gene a o la ge axonomic g oups, bu ypically only one de elopmen al mode exis s wi hin a single species. In a e cases, species may exp ess wo o mo e de elopmen modes. The e m poecilogony (Gia d 1905 ci ed in K ug 994 c 2012 The Au ho s. Published by Blackwell Publishing L d. This is an open access a icle unde he e ms o he C ea i e Commons A ibu ion Non Comme cial License, which pe mi s use, dis ibu ion and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed and is no used o comme cial pu poses. J. E. Kes¨ aniemi e al. No C yp ic Specia ion in Pygospio elegans 2009) has been used o desc ibe such de elopmen al mode polymo phism. In poecilogonous species, mul iple de elop- men al modes a e obse ed, ei he wi hin o among di e en popula ions o a single species. T ue poecilogony has been documen ed wi hin spionid wo ms (e.g., S eblospio bene- dic i, Le in 1984, and Bocca dia p oboscidia, Gibson 1997; Oya zun e al. 2011) and in sacoglossan sea slugs ( e iewed in K ug 2007, 2009). Howe e , in a numbe o cases, wha we e o iginally desc ibed as poecilogonous species ha e u ned ou o be mo phologically c yp ic species wi h species-speci ic de elopmen al modes (see Hoagland and Robe son 1988). The a i y o ue poecilogony has led some au ho s o sug- ges ha he e a e cos s associa ed wi h polymo phic de el- opmen and ha poecilogony is a ansien s age o specia ion co-occu ing wi h de elopmen al mode ansi ions (Gibson and Gibson 2004; Ellingson and K ug 2006). Al e na i ely, poecilogony migh be an ad an ageous plas ic esponse, and a po en ial be -hedging s a egy, o enhance o sp ing suc- cess in he ace o changing en i onmen al condi ions (K ug 2007). One possible poecilogonous species is Pygospio elegans Clapa ` ede, a small, seden a y, ube-building spionid poly- chae e wo m, widely dis ibu ed in he no he n hemi- sphe e (Muus 1967; Ange 1984). A e in e nal e iliza ion (Hanne z 1956), emales deposi emb yos and yolky nu se eggs in capsules inside he ma e nal ube. Di e en la ae eme ge om he capsules depending on he ela i e numbe o emb yos and nu se eggs laid by he mo he ; he e a e no ini ial di e ences in emb yo size (S¨ ode s ¨ om 1920; Hanne z 1956; Rasmussen 1973; Ange e al. 1986; Blake and A no sky 1999, pe s. obs.). He e, we de ine plank onic la ae as hose ha eme ge when hey a e 3-se ige s long ( ypically >20 em- b yos laid pe capsule wi h ew o no nu se eggs). The la ae de elop long swimming se ae and ac i ely swim and eed in he wa e column (Hanne z 1956). B ooded la ae, on he o he hand, do no ha e swimming se ae and emain inside he capsules o a longe pe iod subsis ing only on nu se eggs ( ypically one o wo emb yos laid pe capsule, Fig. 1). These la ae lack a pelagic phase du ing de elopmen and me a- mo phose in o ju eniles soon a e hei eme gence om he capsules a 14–20 se ige s. An in e media e ype o la a also occu s (4–10 emb yos laid pe capsule; Hanne z 1956, pe s. obs.). A e eme gence a app oxima ely 10 se ige s, hese la - ae ha e a sho pelagic phase. Despi e hei di e ences, all la al ypes me amo phose in o mo phologically and eco- logically iden ical adul s. Moni o ing ep oduc ion in P. elegans is labo ious and has been done exhaus i ely in only a ew popula ions. The e ha e been some obse a ions o di e en la al o ms si- mul aneously wi hin a single popula ion (Rasmussen 1973; Gudmundsson 1985, pe s. obs.), p o iding some e idence ha P. elegans is a ue poecilogonous species. Howe e , whe he o no a single indi idual can p oduce mul iple la - Figu e 1. B ooded Pygospio elegans la ae in capsules ( om ¨ Angs ¨ o, Finland). The capsules (app ox. 0.5 mm long, each con aining one o wo la ae) a e isible a e b eaking down he sand ube. Pho o c edi : Jenni Kes¨ aniemi. al ypes is no clea (bu , see Fig. 30 in Rasmussen 1973). Hanne z (1956) and Rasmussen (1973) hypo hesized ha de- elopmen al mode polymo phism in P. elegans is in ac a i- a ion wi hin a single de elopmen al mode, e lec ing plas ic esponses o en i onmen al a ia ion. This hypo hesis was based on obse a ions ha in some popula ions di e en la ae a e p oduced seasonally. Howe e , nei he simul ane- ous no seasonal p oduc ion o di e en la ae in a single popula ion is uni e sal. Mo e commonly, among popula- ion di e ences in de elopmen al mode a e no ed, and some popula ionsha ee enbeenconside ed“ ixed” o apa ic- ula de elopmen al mode since no o he modes ha e been obse ed du ing epea ed sampling om hese popula ions (Ange 1984; Mo gan e al. 1999; Bolam 2004, pe s. obs.). The p esence o “ ixed” popula ions di e ing in de elopmen al mode aises suspicion ha c yp ic species may be p esen . This suspicion was s eng hened when Ange (1984) ound ha expe imen al exposu e o wo ms om se e al “ ixed” popula ions o di e en salini ies and empe a u es did no induce a change in de elopmen al mode. No co ela ions be ween o he en i onmen al a iables and de elopmen al mode ha e been no ed in he li e a u e, bu ew expe imen- al es s ha e been pe o med. Changes in densi y and ood supply did no induce changes in de elopmen al mode in P. elegans collec ed om Somme Bay, F ance (Mo gan 1997), bu in No h Ame ica, low densi y has appa en ly inc eased he equency o asexual ep oduc ion in P. elegans (Wilson 1983). To cla i y he species s a us o P. elegans popula- ions, Mo gan and colleagues (1999) examined popula ion c 2012 The Au ho s. Published by Blackwell Publishing L d. 995 No C yp ic Specia ion in Pygospio elegans J. E. Kes¨ aniemi e al. s uc u e among ou po en ially “ ixed” popula ions in he English Channel di e ing in de elopmen al mode. They ound high gene ic simila i y and po en ially high gene low among he P. elegans popula ions, and concluded ha he species is poecilogonous. Ne e heless, due o he limi ed scope o hei s udy and he a i y o poecilogony, he ques- ion o poecilogony e sus c yp ic specia ion s ill emains. We add essed his ques ion by su eying a ia ion in a po - ion o he mi ochond ial gene cy och ome c oxidase subuni I using haplo ype ne wo k and phylogene ic me hods, and using a DNA sequence-based c i e ion ad oca ed in DNA ba coding s udies o assess he p esence o c yp ic species. Ou samples co e ed bo h a b oad geog aphical a ea and a ange o en i onmen al condi ions. Fo some popula ions, he e we e also da a a ailable ega ding he p edominan de- elopmen al mode among indi iduals. The la ge da ase also allowed us o in es iga e wi hin-popula ion di e si y in ou s udy popula ions. We hypo hesized ha P. elegans is indeed a poecilogonous species, despi e appa en di e gence o pop- ula ions in de elopmen al mode. Ma e ials and Me hods Sample collec ion and molecula me hods Adul P. elegans we e collec ed be ween 2007 and 2010 om 14 loca ions in Eu ope (Fig. 2) and h ee loca ions in he Uni ed S a es (eas coas : Maine and wes coas : Washing on). In Eu ope, popula ions om he Bal ic Sea (Finland, Ge many, Denma k, Sweden), Wadden Sea ( he Ne he lands, Schie monnikoog Island), No h Sea (Edinbu gh, UK), he English Channel (Plymou h, UK, and Somme Bay, F ance), Whi e Sea (Russia), and he No h A lan ic Ocean (Iceland) we e sampled (Fig. 2, Table 1). Se e al colleagues enabled he collec ing e o (see Acknowledgemen s). A mos lo- ca ions, he samples we e collec ed om he shallow in e - idal zone (0.1–1 m). The wo samples om he Finnish a chipelago (¨ Angs¨ oandF ˚ a ¨ o) we e collec ed by scuba om 2–5 m deep wa e . Samples om Ge many we e collec ed om 18-m dep h. A he ime o collec ing, he adul wo ms and sand ubes we e examined o signs o la ae o egg capsules and hen p ese ed in e hanol (94–99%). Using hese obse a ions, and in o ma ion om p e ious s udies o P. elegans’ ep o- duc ion and de elopmen (i.e., Rasmussen 1973; Mo gan e al. 1999; Bolam 2004), we cha ac e ized he sampling loca- ions by he di e en la al de elopmen al modes obse ed (Table 1). This cha ac e iza ion is en a i e, since we we e unable o su ey all popula ions exhaus i ely, bu ep esen s ou bes knowledge o he p edominan de elopmen al mode in he popula ions. Addi ional sampling a he same si es has con i med ou cha ac e iza ion o de elopmen al mode (pe s. obs.) bu a some si es we ha e no obse ed any signs o sexual ep oduc ion and so a p edominan de elopmen al mode is no known. F om he Eu opean samples, genomic DNA was ex ac ed using he DNeasy Blood and Tissue ex ac ion ki (Qiagen, Ge many) and a KingFishe magne ic p ocesso (The mo- Scien i ic, MA, USA). A 600-bp agmen o he cy och ome c oxidase subuni I (COI) gene was ampli ied using species- speci ic p ime s (PeCox1 F 5– TAT AGG CCT TTG ATC AGG AAC – 3,PeCox1R5 – AGG GTC TCC GCC TCC TGT – 3). Polyme ase chain eac ions (PCRs) we e pe - o med in 20 μL eac ions con aining 1 μLo heDNAex- ac , 3 mM MgCl2(Bio ools, Spain), 200 μM o each dNTP (Fe men as, Ge many), 0.5 μMo eachp ime (TAGCopen- hagen, Denma k), 0.1 U o Taq polyme ase, and 1 X o PCR Bu e (Bio ools). Reac ion condi ions included an ini ial de- na u a ion s ep a 94◦C o 2min, hen35cycleso dena - u a ion a 94◦C o 15 s, annealing a 55◦C o 15s,and ex ension a 72◦C o 45 s, ollowed by a inal ex ension a 72◦C o 2 min. Fo sequencing, he PCR p oduc s we e ea ed wi h Exonuclease I and Sh imp alkaline phospha ase (Fe men as), cycle sequenced in bo h di ec ions using he BigDye .3.1 ki , and isualized wi h an ABI 3130xl Gene ic Analyze and Sequencing Analysis .5.2. so wa e (all Applied Biosys ems, CA, USA). DNA ex ac ion, ampli ica ion, and sequencing o he No h Ame ican samples ollowed simila p o ocols, bu se- quencing a i ac s a he 5endo he esul ingsequences educed he leng h o high-quali y sequence eads o hese samples. To be conse a i e, we analyzed a sho e agmen o he COI gene (567 bp) when No h Ame ican samples we e included. In analyses in ol ing only he Eu opean samples, he 600-bp agmen was used. Haplo ype ne wo k and phylogene ic analyses Sequences we e aligned using he Clus alW op ion o MEGA 4 (Tamu a e al. 2007). Fo hese analyses, he 567-bp agmen o he COI gene was used and all indi iduals we e included. To examine he ela ionship be ween he haplo- ypes, a minimum spanning ne wo k was cons uc ed wi h A lequin .3.5.1.2. (Exco ie and Lische 2010) and isual- ized wi h HapS a (Teache and G i i hs 2010). Fo phylogene ic analyses, a single ep esen a i e o each haplo ype was used. JModel es (Posada 2008) was used o ind he op imal model o sequence e olu ion o he COI da a (selec ed using he Akaike in o ma ion c i e ion, AIC). The gene al ime e e sible model wi h in a ian posi ions and gamma-dis ibu ed a es (GTR + I + G) was selec ed and used in ee econs uc ion. Sequence di e gence was es ima ed wi h MEGA 4 using a gamma shape pa ame- e o 0.637 (acco ding o JModel es ) and he Tamu a Nei 996 c 2012 The Au ho s. Published by Blackwell Publishing L d. J. E. Kes¨ aniemi e al. No C yp ic Specia ion in Pygospio elegans Figu e 2. Eu opean sampling si es labeled acco ding o hei abb e ia ions in Table 1. Si es FIA ( ¨ Angs ¨ o) and FIF (F˚ a ¨ o) a e loca ed in he Finnish a chipelago, app oxima ely 20 km apa . Regional g ouping o popula ions o he hie a chical AMOVA analysis a e indica ed wi h numbe ed supe sc ip s: 1. No he n Bal ic Sea: Finland, 2. Sou he n Bal ic Sea: Denma k, Ge many, Sweden, 3. No h Sea + Wadden Sea + English Channel: UK, F ance, he Ne he lands, and 4. No h A lan ic Ocean: Iceland. subs i u ion model since he GTR model is no a ailable in MEGA 4. Fo ee econs uc ion, we explo ed bo h maximum like- lihood and Bayesian analyses. Bayesian analysis was con- duc ed wi h M Bayes .3.1.2. (Ronquis and Huelsenbeck 2003). MCMC (Ma ko Chain Mon e Ca lo) chains (one cold and h ee hea ed chains) we e un o 4 million gene a- ions, ees we e sampled e e y 100 gene a ions, and 25% o he ees we e disca ded in he bu nin. All pa ame e s we e es ima ed in he analysis. Pos e io p obabili ies we e used o assess clade suppo , wi h 80% used as he minimum cu o . Maximum likelihood analysis was conduc ed wi h PhyML 3.0. (Guindon and Gascuel 2003). All pa ame e s we e es- ima ed in he analysis excep he gamma shape pa ame e , which was se o 0.637 acco ding o he esul s om JMod- el es . Boo s ap analysis wi h 1000 eplica es p o ided an es ima e o clade suppo , wi h 70% used as he minimum cu o . A e analysis, ees we e oo ed along he lineage leading o mos o he No h Ame ican haplo ypes (also he longes b anch). T ees we e isualized using FigT ee .1.2.2. (h p:// ee.bio.ed.ac.uk/so wa e/ ig ee/). Analysis o gene ic di e si y Ou gene ic di e si y analyses ocused on popula ions wi h su icien sample sizes o making obus es ima es, so he c 2012 The Au ho s. Published by Blackwell Publishing L d. 997 No C yp ic Specia ion in Pygospio elegans J. E. Kes¨ aniemi e al. Table 1. Sampling loca ion in o ma ion, popula ion codes, di e si y measu emen s, and obse ed la al modes o he popula ions. N=numbe o indi iduals in he gene ic analysis, H=haplo ype di e si y, and π=nucleo ide di e si y. Box indica es he Eu opean popula ions wi h su icien sample size used in di e si y and demog aphic analyses as well as hie a chical analyses o popula ion s uc u e (AMOVA). G oups ( egions) de ined o AMOVA analysis a e shaded. Region Loca ion Code N*No. o haplo ypes HπObse ed la al mode Eu ope No he n Bal ic Sea ¨ Angs ¨ o, Finland FIA 22 4 0.260 0.0017 B ooded F˚ a ¨ o, Finland FIF 21 10 0.890 0.0129 No known Hanko, Finland FIH 19 6 0.778 0.0128 No known Sou he n Bal ic Sea Ge many GER 22 16 0.909 0.0118 No known Velle up, Denma k DKV 20 12 0.916 0.0141 B ooded, in e media e He sle , Denma k DKH 20 11 0.916 0.0154 B ooded, in e media e Ro ig, Denma k DKR 21 6 0.710 0.0116 In e media e, plank onic Gullma jo d, Sweden SWE 21 5 0.633 0.0080 No known Wadden sea he Ne he lands NET 23 14 0.822 0.0107 In e media e, plank onic No h sea D um sands, UK UKD 20 19 0.995 0.0130 Plank onic English Channel Somme Bay, F ance FRA 23 22 0.996 0.0141 Plank onic Plym Bay, UK UKP 24 20 0.975 0.0153 Plank onic A lan ic Ocean Iceland ICE 20 3 0.511 0.0009 B ooded, in e media e Whi e Sea Russia RUS 3 1 0.000 0.0000 No known No h Ame ica A lan ic (eas ) Lubec, ME NAE1 7 3 0.733 0.0360 No known Lowe’s co e, ME NAE2 6 3 0.810 0.0210 No known Paci ic (wes ) False Bay, WA NAW 7 1 0.000 0.0000 No known *Based on 600 bp COI agmen in Eu opean popula ions, bu based on 567-bp agmen in No h Ame ican popula ions. Russian sample (n=3) and he No h Ame ican samples (n=6–7) we e excluded. In hese analyses, he 600-bp ag- men o he COI gene was used. Haplo ype di e si y and nucleo ide di e si y o each popula ion we e calcula ed wi h A lequin .3.5.1.2. (Exco ie and Lische 2010), which was also used o es ima e popula ion s uc u e (ST) iaahi- e a chical analysis o molecula a iance (AMOVA). In he AMOVA analysis, sequences we e g ouped acco ding o geo- g aphical egions ( ou g oups: No he n Bal ic Sea; Sou h- e n Bal ic Sea; No h Sea + Wadden Sea + English Channel; and No h A lan ic Ocean; 10,000 pe mu a ions). Popula ion s uc u e was also in es iga edusing BAPS 5.3 (Co ande and Tang 2007), a Bayesian model-based clus e ing me hod ha can use sequence da a. In hese analyses, he maximum num- be o clus e s (K) was se om wo o 13, and o each he analysis was un 10 imes. In he end, he Kwi h he highes likelihood was chosen o desc ibe he samples. Explo a o y analyses es ed whe he di e ences in haplo- ype and nucleo ide di e si y measu es we e e iden among he Eu opean popula ions wi h di e en de elopmen al mode. He e, plank onic popula ions (UKP, UKD, FRA, see Table1)we ecompa ed opopula ions ha p oduceb ooded o in e media e ype la ae (FIA, DKV, DKH, DKR, NET, ICE). This compa ison is con ingen on ou de ini ion o p edominan de elopmen al mode (see Table 1), so popu- la ions whe e de elopmen al mode is no known (FIF, FIH, GER, SWE, and RUS) we e excluded. Fo hese compa isons, Mann–Whi ney U es s we e pe o med using PASW S a is- ics 18 (SPSS, Inc., 2009, Chicago, IL, www.spss.com). To assess i Eu opean popula ions (excluding RUS) ha e gone h ough a ecen popula ion expansion, Fu’s Fsneu al- i y es was calcula ed. Fu’s es (which is based on he haplo- ype dis ibu ion; Fu 1997) was used because i is hough o be be e a e ealing signs o popula ion expansion han Tajima’s D es (Fu 1997; Schneide and Exco ie 1999). Tajima’s D(Tajima 1989) and Fu and Li’s F(Fu and Li 1993) we e also calcula ed o es o neu ali y o he sequences. Misma ch dis ibu ions, he equencies o obse ed pai - wise di e ences be ween haplo ypes, we e calcula ed o each Eu opean popula ion. Also R2(Ramos-Onsins and Rozas 2002) and aggedness s a is ics ( g, Ha pending 1994) wi h con idence in e als based on coalescen simula ions we e calcula ed o de ec expansion (10,000 pe mu a ions and he a es ima ed om he da a we e used in he coalescen simula ions). Lowe R2and g alues a e expec ed o a pop- ula ion g ow h scena io (Ha pending 1994; Ramos-Onsins and Rozas 2002). These analyses we e pe o med in DnaSP 4.0 (Rozas e al. 2003). Resul s Polymo phism and haplo ype di e si y A o al o 279 P. elegans indi iduals om 14 Eu opean loca ions we e sequenced. F om his sample, 121 unique 998 c 2012 The Au ho s. Published by Blackwell Publishing L d. J. E. Kes¨ aniemi e al. No C yp ic Specia ion in Pygospio elegans haplo ypes o he COI gene agmen (600 bp) we e iden- i ied. An expanded da ase included 20 addi ional indi id- uals om h ee No h Ame ican loca ions o a o al o 299 sequences, 567 bp in leng h, wi h 123 unique haplo- ypes (GenBank accession numbe s JN033571–JN033693). The mos common haplo ype, EUNA10, was sha ed by 36 Eu opean indi iduals and was ound in all h ee Danish popula ions, Iceland, Sweden, and Plym Bay in he UK (English Channel). This haplo ype was also obse ed in wo ms sampled om bo h popula ions on he Eas Coas o he Uni ed S a es (NAE). The second mos common hap- lo ype, EU11, was ound in Denma k, Finland, F ance, he Ne he lands, and he Whi e Sea, Russia (35 indi iduals). No e ha bo h EUNA10 and EU11 we e ound in popula ions di - e ing in de elopmen al mode (see Table 1). These wo mos common haplo ypes also we e ound wi hin he whole sam- ple ange in Eu ope and comp ise 25% o all indi iduals sequenced. We obse ed a la ge numbe o low equency haplo ypes wi hin loca ions in Eu ope. Ou o 121 haplo ypes, 98 we e de ec ed only once in he Eu opean da ase ( om only one indi idual o he 279 sequenced). Nine y pe cen o he hap- lo ypes (109 ou o 121) we e ound in only one popula ion (11o hesewe e ound ommo e hanoneindi idual).Pop- ula ions om he No h Sea, English Channel, and Wadden Sea had he highes pe cen age o popula ion-speci ic low- equency haplo ypes. The Bal ic Sea popula ions (Finland, Denma k, Ge many, Sweden)sha edmanyhaplo ypes(se en ou o 12 sha ed haplo ypes a e ound only in he Bal ic Sea), and only one haplo ype (EU8) was sha ed exclusi ely among he h ee popula ions in he UK and F ance. In mos pop- ula ions, haplo ype di e si y was high (Table 1). Howe e , wo Eu opean popula ions had low di e si y wi h mos in- di iduals sha ing he same haplo ype. In ¨ Angs¨ o, Finland, 19 o 22 indi iduals sampled (86%) sha ed an iden ical haplo- ype (EU6) and in he sample o 20 indi iduals om Iceland, 13 sha ed haplo ype EU1 and six sha ed haplo ype EUNA10. O e all, popula ions wi h p edominan ly plank onic la ae had highe haplo ype di e si y han popula ions ha also p oduced o he la al ypes (N=9, U=0.000, z=–2.334, P=0.020). Howe e , nucleo ide di e si y was no signi i- can ly di e en (N=9, U=4.5, z=–1.167, P=0.243). Hap- lo ype di e si y in he No h Ame ican samples was some- wha lowe han in mos o he Eu opean samples (Table 1), bu No h Ame ican sample sizes we e also ela i ely small and so es ima es o di e si y om hese popula ions may no be eliable. Mean sequence di e gence (Tamu a Nei model) wi hin he o al Eu opean da ase was 1.7%. Di e gence be ween he Eu opean and No h Ame ican haplo ypes was no ice- ably highe : 5.3% (o 6.1% when excluding EUNA10, he haplo ype ha is sha ed wi h he Eu opean samples). Mean sequence di e gence wi hin he o al No h Ame ican da ase was 3.1%, highe han wha we obse ed om he Eu opean sample. Haplo ype ne wo k and phylogene ic analyses Figu e 3 shows he minimum spanning haplo ype ne wo k as calcula ed in A lequin. The low sequence di e gence among haplo ypes is e lec ed in he ne wo k and haplo ypes om di e en popula ions a e in e mingled. A lequin de ec ed many al e na i e connec ions among he Eu opean haplo- ypes due o he low le el o di e gence be ween hem, bu g aphing all possible al e na i e connec ions would ha e made he ne wo k un eadable. The mos common hap- lo ype, EUNA10, was ound om almos all popula ions and o he linked haplo ypes came om Iceland, he No h Sea, he English Channel, and he Sou he n Bal ic Sea, bu no om Finland. The o he common haplo ype, EU11, is mul iple mu a ional s eps away om EUNA10. Mo eo e , he o he No h Ame ican haplo ypes we e no connec ed closely o EUNA10 and we e clea ly di e en om he Eu o- pean sequences. The Eu opean haplo ype closes o he clus- e o No h Ame ican haplo ypes is om he Ne he lands (Fig. 3), and al e na i e connec ions (also 26 mu a ional s eps o NAW NAE) a e om he UK ( wo haplo ypes om UKP, one om UKD; no shown). O e all, he minimum span- ning ne wo k included a la ge numbe o small nodes de- pic ing he high equency o single on haplo ypes no ed ea - lie , and some medium equency haplo ypes (obse ed in wo o eigh indi iduals) ha we e de ec ed in one popula- ion only. These single on and low- equency haplo ypes a e widesp ead h oughou he ne wo k. Phylogene ic analyses esul ed in simila ee opologies ega dless o which ee econs uc ion me hod was used, he e o e, only he esul s om he maximum likelihood analysis a e discussed and shown (Fig. 4). As in he hap- lo ype ne wo k, he e was a clea sepa a ion be ween he Eu opean haplo ypes and mos No h Ame ican haplo ypes (o he han EUNA10) and he Eu opean clade was well sup- po ed by boo s ap analysis (Fig. 4). Wi hin he Eu opean clade, he e was e y li le di e gence and only a ew g oups we e clea ly esol ed wi h high boo s ap suppo (Fig. 4, do s a suppo ed nodes). The lack o boo s ap suppo a mos nodes indica es limi a ions o he da a o esol ing ela ion- ships o he P. elegans haplo ypes. Howe e , his analysis also e eals clus e s de ec ed in he haplo ype ne wo k. Fo exam- ple, one well-suppo ed g oup con ains almos all he Finnish ¨ Angs¨ o haplo ypes (3 ou o 4; EU6, FIA43, FIA44). Ano he well-suppo ed g oup con ains indi iduals om Ge many and all o he h ee Danish popula ions, e en hough o he haplo ypes om hese popula ions we e also dis ibu ed else- whe e in he phylogene ic ee. In addi ion, mos Swedish (excep one) and all Icelandic haplo ypes we e included in a c 2012 The Au ho s. Published by Blackwell Publishing L d. 999 No C yp ic Specia ion in Pygospio elegans J. E. Kes¨ aniemi e al. Figu e 3. Haplo ype ne wo k o he 123 COI haplo ypes de ec ed in P. elegans. Ci cle size is p opo ional o haplo ype equency and haplo ypes wi h mo e han one indi idual a e also named. Small black ci cles ep esen unde ec ed in e media e haplo ypes and lines connec ing ci cles ep esen one mu a ional s ep unless o he wise speci ied. Ci cles a e colo ed o ep esen sampling si es. Haplo ypes ound om mo e han one loca ion a e colo ed as pie cha s wi h p opo ionally sized wedges ep esen ing he haplo ype equency in each popula ion. O als wi h dashed ou lines enci cle clus e s in he haplo ype ne wo k which we e also de ec ed in phylogene ic analyses wi h s ong (70% o g ea e ) boo s ap suppo . well-suppo ed g oup, which also included EUNA10, one o he mos common haplo ypes also sampled om No h Ame ica. The wo mos commonly encoun e ed haplo- ypes (EUNA10 and EU11) did no g oup oge he (Fig. 4, as e isks). Popula ion s uc u e and demog aphic analyses Fo he es o egional subdi ision o sequence di e si y in Eu ope (AMOVA), he popula ions we e a anged in o ou g oups acco ding o geog aphical egion (1. No he n Bal ic Sea: Finland, 2. Sou he n Bal ic Sea: Denma k, Ge many, Sweden, 3. No h Sea +Wadden Sea +English Channel: UK, F ance, he Ne he lands, and 4. No h A lan ic Ocean: Iceland, see Fig. 2). These esul s (Table 2) showed ha mos o he a ia ion was ound wi hin popula ions (69.8%, P> 0.001) and ha di e en ia ion among he egions was sig- ni ican al hough small (accoun ing o 8% o he molecula a iance, P=0.001). Addi ional signi ican a ia ion among popula ions wi hin each egion (22.2%, P>0.001) indica ed ha s uc u e may also be p esen on smalle spa ial scales. Analysis using he p og am BAPS de ec ed se en gene ic clus e s in ou da a (wi h he p obabili y o 0.99). Clus e s we e no based on sampling loca ion, each clus e con aining indi iduals om ou o en sampling loca ions. Two clus e s we e s ic ly Bal ic, one con aining mos o he indi iduals sampled om he Finnish ¨ Angs¨ o si e and he o he con aining mos o he Ge man samples, al hough Ge man indi iduals we e also placed in o ou o he clus e s (Table 3). Demog aphic analyses sugges ed he e has been ecen popula ion expansion in he popula ions om he UK and F ance. In hese popula ions, Fu’s Fs alues we e nega i e and signi ican . Unimodal misma ch dis ibu ion cu es o hese popula ions also indica e ha a ecen popula ion expansion 1000 c 2012 The Au ho s. Published by Blackwell Publishing L d. J. E. Kes¨ aniemi e al. No C yp ic Specia ion in Pygospio elegans Figu e 4. Maximum likelihood ee o COI haplo ypes de ec ed in P. elegans. Nodes ma ked wi h black do s indica e clades esol ed wi h boo s ap alues o 70% o highe . Dashed lines connec haplo ype names o he b anches and should no be in e p e ed as b anch leng h. Numbe s in he b acke s ollowing he haplo ype name indica e he numbe o indi iduals obse ed wi h ha haplo ype. When no numbe is indica ed, he haplo ype was sampled only once. As e isks indica e he wo mos common haplo ypes: EUNA10 and EU11. Cen al colo wheel indica es egion o sampling: ligh g ay =Bal ic Sea (No he n + Sou he n); da k g ay =Iceland; black =No h Sea, English Channel and Wadden Sea; and whi e =sha ed Eu opean haplo ypes. No e, he No h Ame ican haplo ypes in he ci cula phylogeny a e in he cen e o he colo wheel. c 2012 The Au ho s. Published by Blackwell Publishing L d. 1001