Taxonomical use of floral scent data in apomictic taxa of Hieracium and Sorbus derived from hybridization

Feulner, Martin

Taxonomical use of floral scent data in apomictic taxa of Hieracium and Sorbus derived from hybridization

Feulner, Martin

Published: 2013-01-01
Repository: epub.uni-bayreuth.de
Language: de

Full text

Taxonomical use o lo al scen da a in apomic ic axa o Hie acium and So bus de i ed om hyb idiza ion Disse a ion zu E langung des Dok o g ades D . e . na . an de Fakul ä Biologie/Chemie/Geowissenscha en de Uni e si ä Bay eu h o geleg on Ma in Feulne , aus Bay eu h Bay eu h, 2013 Die o liegende A bei wu de on Mai 2006 bis Janua 2013 am Leh s uhl P lanzensys ema ik de Uni e si ä Bay eu h un e Be euung on F au P o . D . Sig id Liede-Schumann und He n P o . D . S e an Dö e l ange e ig . Volls ändige Abd uck de on de Fakul ä ü Biologie, Chemie und Geowissenscha en de Uni e si ä Bay eu h genehmig en Disse a ion zu E langung des akademischen G ades eines Dok o s de Na u wissenscha en (D . e . na .). Disse a ion einge eich am: 08.02.2013 Zulassung du ch die P ü ungskommission: 13.02.2013 Wissenscha liches Kolloquium: 04.07.2013 Am ie ende Dekan: P o . D . Bea e Lohne P ü ungsausschuss: P o . D . Sig id Liede-Schumann (E s gu ach e ) P o . D . Be ina Engelb ech (Zwei gu ach e ) P o . D . S e an Dö e l P o . D . Heike Feldhaa PD D . G ego Aas This disse a ion is submi ed as a ‘Cumula i e Thesis’ ha includes ou publica ions: wo published a icles, one submi ed a icle, and one a icle in p epa a ion o submission. Lis o Publica ions 1) Feulne M., Schuhwe k F., Dö e l S. 2009: Flo al scen analysis in Hie acium subgenus Pilosella and i s axonomical implica ions. Flo a 204: 495–505. 2) Feulne M., Schuhwe k F., Dö e l S. 2011: Taxonomical alue o in lo escence scen in Hie acium s. s . Biochemical Sys ema ics and E olu ion 39: 732–743. Submi ed ( e ision unde e iew): 3) Feulne M., Liede-Schumann, S., Me e, U., Weig A., Aas, G.: Gene ic s uc u e o So bus la i olia (Lam.) Pe s. axa endemic o No he n Ba a ia. Submi ed o Plan sys ema ics and e olu ion, PLSY-D-12-00168. In p epa a ion o submission o O ganisms Di e si y & E olu ion: 4) Feulne , M., Poin ne , S., Heuss, L., Aas, G., Dö e l, S.: Flo al scen and i s co ela ion wi h gene ic da a in So bus axa. Publica ions no included in o his hesis: 5) Feulne , M., Mösele , B.M., Nezadal, W. 2001: In og ession und mo phologische Va iabili ä bei de Blauen Himmelslei e , Polemonium cae uleum L. in No dbaye n, Deu schland, Feddes Repe o ium, 112: 231–246. 6) Feulne , M., Konne , M. 2007: Au och hone Weiß annen o kommen in den Schluch en F änkische Keupe gebie e. Diskussionsbei ag zu de en gene ische S uk u , A enauss a ung, waldbauliche Behandlung und Ka ie ung in Na u a 2000, Waldoekologie online 4: 91–110. Decla a ion o con ibu ion o publica ions The hesis con ains ou esea ch a icles. Mos o he esea ch wo k p esen ed in his hesis was ca ied ou by mysel a he Uni e si y o Bay eu h including all sample collec ions, mos analy ic wo k and he mos s a is ics unde p o ound suppo o P o . D . S e an Dö e l, PD D . Ul ich Me e, D . Al ons Weig, PD D . G ego Aas and P o . D . Sig id Liede-Schumann. I p epa ed he manusc ip s unde conside a ion o he commen s o all coau ho s. 1s publica ion Feulne M., Schuhwe k F., Dö e l S. (2009): Flo al scen analysis in Hie acium subgenus Pilosella and i s axonomical implica ions. Flo a 204: 495–505. The ield wo k was done by mysel , da a analysis was done by mysel unde he p o ound suppo o PD D . S e an Dö e l. No be Meye and D . F anz Schuhwe k con ibu ed o species selec ion, ga e p o ound ad ice abou locali ies o endemic Hie acium axa and helped wi h species iden i ica ion in he ield and o he ba ium specimens. I p epa ed he manusc ip by ecognizing he p o ound commen s o my co-au ho s. 2nd publica ion Feulne M., Schuhwe k F., Dö e l S. (2011): Taxonomical alue o in lo escence scen in Hie acium s. s . Biochemical Sys ema ics and Ecology 39: 732–743. The ield wo k was conduc ed by mysel , da a analysis was done by mysel unde he p o ound suppo o PD D . S e an Dö e l. D . Jochen Mülle , No be Meye and D . F anz Schuhwe k ga e p o ound ad ice abou locali ies o endemic Hie acium axa and helped wi h species iden i ica ion in he ield and o he ba ium specimens. I p epa ed he manusc ip by ecognizing he commen s o my co-au ho s. 3 d publica ion Feulne , M., Liede-Schumann, S., Me e, U., Weig A., Aas G.: O igin and gene ic s uc u e o h ee So bus la i olia (Lam.) Pe s. axa endemic o No he n Ba a ia. Submi ed o Plan sys ema ics and e olu ion, PLSY-D-12-00168-1. The plan ma e ial collec ion was conduc ed by mysel unde he suppo o D . G ego Aas (EBG Bay eu h) who had he idea o his esea ch. AFLP Labo a o y wo k was conduc ed by Michaela Hochholze and D . Al ons Weig (bo h DANECO Bay eu h). Ch omosome coun s we e conduc ed unde supe ision o PD D . Ul ich Me e. Da a analysis was done by mysel unde he suppo o D . Al ons Weig and P o . D . Sig id Liede-Schumann. I p epa ed he manusc ip by ecognizing he commen s o my co- au ho s. 4 h publica ion Feulne , M. Poin ne S., Heuss L., Aas G., Dö e l S.: Co ela ion be ween axonomic g oupings o So bus mic ospecies based on lo al scen and gene ic da a. In p epa a ion o submission o O ganisms Di e si y & E olu ion. The scen collec ion was done by he bachelo candida es Lisa Heuss, S e an Poin ne and mysel . Da a analysis was done by mysel unde he p o ound suppo o PD D . S e an Dö e l. I p epa ed he manusc ip by ecognizing he commen s o my co-au ho s. Con en Summa y 1 Zusammen assung 3 1. Gene al In oduc ion 5 2. Aims o esea ch 8 3. Synopsis 9 3.1 Ma e ial and Me hods 9 3.2 Resul s and Discussion 17 3.3 Conclusion and pe spec i es 23 3.4 Re e ences 25 Publica ions 31 1. Flo al scen analysis in Hie acium subgenus Pilosella and i s axonomical implica ions. 31 2. Taxonomical alue o in lo escence scen in Hie acium s. s . 43 3. Gene ic s uc u e o h ee So bus la i olia (Lam.) Pe s. axa endemic o No he n Ba a ia. 56 4. Flo al scen and i s co ela ion wi h gene ic da a in So bus axa. 78 Danksagung 95 Eidess a liche E klä ung 96 Summa y Summa y Scen in es iga ions a e mainly ca ied ou in he con ex o he ecological unc ion o scen componen s o pollina o a ac ion, o s udy hei geog aphical a iabili y o hei e olu ion. In con as , p o iles o in lo escence scen compounds we e a ely used o phylogene ic analyses and axonomy. So a no in es iga ions a e a ailable ha ocus on scen o apomic ic plan g oups and i s alue o he axonomy o hese g oups. Apomic s p oduce e ile seeds wi hou pollina ion, ei he om soma ic cells o he nucellus o un educed emb yo sac cells. O e all, apomixis occu s sca e ed o e he whole Angiospe m ee in ea ly as well as la e b anching amilies. Rosaceae and As e aceae a e pa icula ly ich in apomic ic axa. In hese amilies gene a such as Hie acium and So bus con ibu e a lo o local species biodi e si y in Cen al Eu ope because o hei high numbe o apomic ic species and axa, which a e o en endemic. Mos membe s o hese gene a buil up polyploid mic ospecies swa ms ini ia ed by hyb idiza ion e en s. The e icula e s uc u e o hese axon complexes leads o axonomic di icul ies ha can no be sol ed by mo phology alone. The e o e, molecula o chemical ma ke s a e needed o in es iga e he pa en age o such axa and hei collec i e species membe ship. An impo an s a ing poin o his esea ch was he inding ha a i icial hyb ids o Ci us p oduce a combina ion o he lea and peel ola iles o hei pa en s. This led o he hypo hesis ha na u al hyb ids migh likewise p oduce scen pa e ns combined o he ola iles o hei pa en s. In lo escence scen s we e in es iga ed om 64 mainly Ba a ian axa o he genus Hie acium (Subgenus Pilosella and Subgenus Hie acium). In So bus (Rosaceae) we ocussed on h ee axa o he S. la i olia agg., endemic o No he n Ba a ia, S. adeana, S. co digas ensis and S. anconica and hei pa en al species om he S. a ia agg ega e (agg.), as well as S. o minalis. Samples we e collec ed wi h dynamic headspace me hod. Subs ances we e iden i ied by gas ch oma og aphy coupled o mass spec ome y (GC-MS). Scen da a we e e alua ed by using a ious clus e me hods. In Hie acium, addi ional e icula ion analyses we e applied ha ace con lic ing signals in a phylog am which can be in e p e ed as hyb idiza ions be ween he axa in ol ed. In So bus addi ional AFLP (Ampli ied F agmen Leng h Polymo phism) analyses we e ca ied ou om he popula ions and indi iduals ha we e s udied o scen and gene ic and scen da a we e co ela ed. In Hie acium, he in lo escense scen consis ed mainly o sesqui e penes, mono e penes, alipha ics and a oma ics. In he lowe scen o So bus besides hese subs ances also ni ogene-con aining subs ances we e ound. 1 Summa y In Hie acium (bo h subgene a) as well as in So bus, axa o hyb id o igin showed a mixed scen pa e n compa ed wi h he pa en al axa. In many cases, he pa en al axa ha had been sugges ed by mo phological in es iga ions o e ealed by gene ic in es iga ions could be con i med by scen . In Hie acium subgenus Pilosella howe e , based on scen da a, some c i ical subspecies could be shown o belong o ano he collec i e species (e.g. H. bauhini ssp. hispidissimum) han he one ha had been p oposed o mo phological easons. In gene al, scen pa e ns co ela e well wi h mo phological o gene ical g oupings, bo h in Hie acium and in So bus. Re icula e scen analyses in Hie acium showed ha some axa a e p obably de i ed om mo e han wo ances o s. Scen da a e lec ed e en axonomic pa e ns on a highe le el, i.e. he sec ional le el. In Hie acium s.s . wo main g oups we e ound, he high g owing and la e lowe ing axa such as H. umbella um and H. lae iga um on he one hand, and he low g owing ea lie lowe ing axa such as H. mu o um and H. bi idum on he o he hand. The AFLP s udy e ealed ha he selec ed membe s o he So bus la i olia agg. a e gene ically clea ly di e en ia ed and mos ly o clonal s uc u e. They a e mo e closely ela ed o S. a ia han o S. o minalis. The S. a ia agg. has a complex s uc u e. Besides S. a ia s.s . and S. pannonica he e a e also in e media e plan s wi h a ini y (a .) o ei he he one o he o he o hese axa. S. co digas ensis was de i ed om in e media es a . S. a ia s.s ., and bo h S. adeana and S. anconica de i ed om in e media es a . S. pannonica o S. pannonica i sel . Flo al scen o he same So bus axa was in es iga ed and compa ed wi h AFLP da a on indi idual as well as on popula ion le el. Co ela ion analysis e ealed a e y high co ela ion be ween scen and AFLP da a on indi idual and popula ion le el. O e all, his wo k shows ha in wo un ela ed plan complexes, Hie acium and So bus, which bo h ha bou a high numbe o apomic ic species de i ed by hyb idiza ion, scen is o high axonomical alue. The main eason o his co ela ion may be ha mos axa in he in es iga ed g oups possess mixed scen pa e ns om hei pa en s o p ogeni o s because hey a e o hyb id o igin. In addi ion, in aspeci ic a iabili y o scen pa e ns is low wi hin apomic ic axa due o hei clonali y, simpli ying he axonomic use o scen da a. Fu he mo e, he ole o pollina o media ed selec ion o scen is educed because o apomixis. 2 Zusammen assung Zusammen assung Schwe punk de Un e suchungen on Blü en- und In lo eszenzdü en is ih e ökologische Funk ion bei de Anlockung on Bes äube n, ih e geog a ische Va iabili ä , ode es we den e olu i e Aspek e des Du es un e such . Nu wenige S udien beschä igen sich dagegen mi Dü en und ih em Po en ial zu Au klä ung axonomische F ages ellungen. Insbesonde e bei apomik ischen P lanzeng uppen gib es auße den hie o geleg en Un e suchungen bei Hie acium und So bus bishe keine wei e en Un e suchungen zu diesem Thema. Bei Apomik en e olg die Emb yobildung ohne Be uch ung aus eine soma ischen Nucelluszelle ode eine unbe uch e en, un eduzie en Emb yosackzelle. Apomixis inde man in Eu opa gehäu bei As e aceen und Rosaceen und hie insbesonde e bei den Ga ungen Hie acium (As e aceae) und So bus (Rosaceae). Viele apomik ische Taxa diese Ga ungen gehö en zu polyploiden Kleina enschwä men, die du ch Hyb idisie ung en s anden sind. Au g und e ikula e Me kmals e eilung sind bei solchen P lanzena en die Ausgangssippen sowie die Zugehö igkei on Kleina en zu g öße en Einhei en o aus de Mo phologie allein nich zu e lässig abzulei en. An küns lich e zeug en Ci us-Hyb iden konn e nachgewiesen we den, dass sie neben wenigen neuen Dü en eine Mischung aus den jeweiligen el e lichen Dü en besi zen. Dahe e schien es in e essan , bei P lanzenkomplexen, die zu g oßen Teilen au na ü liche Hyb idisie ung zu ückgehen, die Zusammense zung de Dü e zu un e suchen und de en Nu zen ü sys ema ische F ages ellungen zu e o schen. In de o liegenden A bei wu den Blü en- und In lo eszenzdü e on 64 iel ach endemischen Taxa de Ga ung Hie acium (Subgenus Pilosella und Subgenus Hie acium) un e such . Bei So bus wu den d ei in No dbaye n endemische Ve e e de S. la i olia-G uppe (So bus adeana, S. co digas ensis, S. anconica) un e such , die du ch Hyb idisie ung zwischen S. a ia und S. o minalis en s anden sind. Du wu de im Gelände mi els de „Dynamic Headspace“ Me hode abgesaug und mi Gasch oma og aphie gekoppel mi Massenspek ome ie (GC- MS) analysie . Die Dü e wu den au g und on Ähnlichkei en in Beziehung geb ach , dabei kamen bei Hie acium auch Re ikula ionsanalysen, welche Hyb idisie ungs- E eignisse au decken können, zum Einsa z. Bei So bus wu den bishe ehlende gene ische Un e suchungen mi els AFLP (ampli ied agmen leng h polymo phism) Analysen du chge üh . Dies e möglich e eine di ek e Ko ela ion on Du - und gene ischen Da en au Popula ions- und eilweise auch Indi iduenebene. Bei Hie acium wu de de In lo eszenzdu o allem on Sesqui e penen, Mono e penen, Fe säu ede i a en, und einigen A oma en bes imm . Im Du on So bus 3 Synopsis – Ma e ial and Me hods Hie acium ha zianum ssp. pseudo anconicum Ha z e Zahn, endemic o No he n F anconia on limes one ocks, Walbe la. Hie acium caesium F ., a a e elic species o he Sou he n F anconian Ju a on limes one ocks, Essing. Hie acium schneidii x pilosella, a a e spon aneous hyb id be ween H. schneidii Schack e Zahn and H. pilosella L. H. schneidii Schack e Zahn, endemic o he No he n F ankonian Alb. Pla e 1: Examples o in es iga ed Hie acium axa. 10 Synopsis – Ma e ial and Me hods Fu he mo e, subspecies a ibu ion o he one o he o he collec i e species some imes seems o be subjec i e (i.e. H. bauhini ssp. hispidissimum). Such uni s a e doub ul because i is no clea ye whe he hey cons i u e na u al uni s o a e o poly opic o igin (c . Schuhwe k 2002). The e o e, i is impo an o in es iga e o which collec i e species some doub ul subspecies belong. The hola c ic genus So bus (Pla e 2) comp ises a la ge a ie y o in e media e species e ol ed om hyb idiza ion be ween common and widely dis ibu ed axa such as So bus o minalis, So bus aucupa ia and So bus a ia. In some a eas o Eu ope endemic hyb ids ha e been de eloped ha a e s abilized by apomixis (Ká pá i 1960, Düll 1961, Meye e al. 2005, Lepší e al. 2009, Rich e al. 2010, Robe son e al. 2010). These hyb id species a e dis ibu ed mos ly in calca eous a eas o Eu ope. In So bus, besides diplospo y and apospo y (Jankun and Ko anda 1987), ano he apomixis ype, pseudogamy occu s, in which pollina ion is necessa y o induce ui se (Jankun and Ko anda 1987, Campbell and Dickinson 1990). Acco ding o Meye e al. (2005) sel ing is su icien o induce ui se in pseudogamous So bus. He e, So bus adeana, S. co digas ensis and S. anconica we e in es iga ed which a e endemic in No he n Ba a ia (Meye e al. 2005, Aas and Kohles 2011). They ha e a small dis ibu ion a ea and ypically occu along o es ma gins o in e y open o es s ages. Thei pa en al lineages a e S. o minalis and S. a ia agg. The la e has a complica ed phylogene ic s uc u e. I comp ises besides obliga o y sexual axa such as S. a ia s.s ., also se e al acul a i e o obliga e apomic ic lineages such as S. pannonica and S. g aeca. So bus pannonica is a xe omo phic membe o S. a ia agg. and mo e widesp ead han S. a ia s.s . in he no he n F anconian Alb (Ku zelnigg 1995, Meye e al. 2005). I is a non-lec o ypi ied axon which comp ises p esumably apomic ic mo pho ypes illing he mo phological gap be ween S. a ia s.s . and S. g aeca (Spach) Loddiges ex Schaue (Ká pá i 1960, Ku zelnigg 2005, Meye e al. 2005). So bus g aeca is ano he xe omo phic membe o he S. a ia agg., mainly dis ibu ed in he Medi e anean lo al egion. I ep oduces sexually o is a acul a i e apomic (Ku zelnigg 1995). I is unce ain whe he S. g aeca occu s in he s udy a ea a all (Düll 1961, Ku zelnigg 1995, 2005), bu indi iduals ha can be a ibu ed mo phologically o S. g aeca we e ound in he No he n F anconian Alb (own obs.), ye , i is di icul o delimi a e his elemen agains S. pannonica. 11 Synopsis – Ma e ial and Me hods So bus pannonica Ká pá i, on he edge o pine o es s in he No he n F anconian Alb, Obe ails eld. So bus anconica Bo nm. ex Düll, on he edge o pine o es s, endemic o No he n F anconia, Obe ails eld. F ui s o So bus pannonica Ká pá i, in he Sou he n F anconian Ju a, Deining. So bus adeana N. Mey., endemic o he No he n F anconian Alb, Modschiedel. Pla e 2: Examples o in es iga ed So bus axa. 12 Synopsis – Ma e ial and Me hods In he p esen hesis h ee main me hods we e applied: Scen collec ion and analysis using dynamic headspace and gas ch oma og aphy coupled o mass spec ome y (GC-MS), ampli ied agmen leng h polymo phism (AFLP) analyses and mo phological analyses. Vola ile collec ion and chemical analyzes o lo al/in lo escence scen (publica ion 1, 2 and 4) Flo al scen was collec ed in he ield using he dynamic headspace me hod desc ibed by Dö e l and Jü gens (2005), and Dö e l e al. (2005a,b). Capi ula (Hie acium) o in lo escences (So bus) we e enclosed wi hin a polyes e o en bag (Toppi s) and he emi ed ola iles we e apped in an adso ben ube h ough he use o a memb ane pump (ASF Thomas, Inc.). As abso ben ube, we used Ch oma oP obe qua z mic o ials o Va ian Inc. (leng h: 15 mm; inne diame e : 2 mm), cu he closed end, illed hem wi h a mix u e (1:1) o 3 mg Tenax–TA (mesh 60–80) and Ca bo ap (mesh 20–40), and ixed he adso ben mix u e in he ial wi h glass wool. Simul aneous collec ions o bo h he lowe scen and su ounding ai a e used o dis inguish be ween lo al compounds and ambien con aminan s. In So bus we used g een lea samples as blank, so we could e eal he lo al scen by subs ac ing hese subs ances, whe eas in Hie acium we collec ed he in lo escence scen because we used he su ounding ai as blank. Fo each axon, one sample o wo o six indi iduals was collec ed. In Hie acium, sampling was ca ied ou on esh and newly opened in lo escences ( he capi ula o Hie acium a e composed o many lo e s wi h mos (90%) o hem open a ime), be ween 11 a.m. and 3 p.m, he pe iod wi h he mos in ensi e scen emission (as de e mined by he human nose; Feulne , unpublished da a). In So bus, he 100–150 lowe s o he pseudo- umbels bloom in pa allel. He e, headspaces co e ed single in lo escences, each. In bo h s udy plan s, scen was collec ed o 3 o 5 minu es a e a ime o 3 o 10 minu es, whe e he scen accumula ed in he closed bag. Headspace samples we e analyzed on a Va ian Sa u n 2000 mass spec ome e coupled o a Va ian 3800 gas ch oma og aph equipped wi h a 1079 injec o (GC-MS) as desc ibed ea lie (Dö e l and Jü gens 2005, Dö e l e al. 2005). The GC-MS da a we e p ocessed using he Sa u n So wa e package 5.2.1. Componen iden i ica ion was ca ied ou using he NIST 02 mass spec al da abase, o MassFinde 2.3, and con i med by he compa ison o e en ion imes wi h published da a (Adams 1995). Iden i ica ion o indi idual componen s could be con i med by he compa ison o bo h mass spec um and GC e en ion da a wi h hose o au hen ic s anda ds. 13 Synopsis – Ma e ial and Me hods Scen da a analysis (Publica ion 1,2 and 4) Fo bo h Hie acium and So bus da a se s, pai wise quali a i e simila i ies we e calcula ed using he Jacca d simila i y index. The signi icance o di e ences in scen p o iles among axa was assessed by ANOSIM wi h 10,000 andom pe mu a ions based on hese simila i y ma ices using P ime Ve sion 5 and 6 (Cla ke and Go ley 2001, 2006). In Hie acium, a e icula ion ne wo k analysis was conduc ed wi h he p og am - ex, e sion 4.0a1 (Maka enko 2001) o analyse he ela ionships among he axa. This me hod allows o isualize ela ionships o species in e connec ed wi h mo e han one ances o (Legend e and Maka enko 2002), which is impo an o analysing g oups, such as Hie acium, wi h many axa o hyb id o igin (see Feulne e al. 2009). In his app oach, a neighbou joining ee was cons uc ed using a dissimila i y ma ix (1-Jacca d), and homoplasies we e made isible by so-called e icula ion lines. Those homoplasies poin owa ds hyb idiza ion o in og ession (Legend e and Maka enko 2002). In Hie acium, in addi ion o he p esence and absence o compounds, we also calcula ed he a e age ela i e (pe cen age o o al) amoun o scen compounds o he single axa. In So bus, he Jacca d ma ix was used o clus e he scen da a wi h UPGMA using P ime Ve sion 5 and 6 (Cla ke and Go ley 2006). Addi ionally in aspeci ic a iabili y o scen da a was compa ed among species using PERMDISP in PRIMER Ve sion 6 (Cla ke and Go ley 2006). Molecula me hods, DNA ma ke Sample collec ion and DNA ex ac ion (Publica ion 3) Lea samples o So bus we e aken in May and June 2010. Immedia ely a e ha es ing hey we e placed in plas ic bags and pu in a box wi h ice o anspo a ion. A he same day, lea es we e washed wi h e hanol in he labo a o y and ozen in an ex ac ion ube a –80 ° C un il ex ac ion. F ozen lea samples (40 – 70 mg, 1 – 2 cm2) we e ex ac ed using widesp ead ex ac ion sys ems and plan ki s (NucleoMag 96 Plan ki ; Mache y-Nagel, Dü en, Ge many, Fas P ep®-24 Tissue Homogenize (MP Biomedicals Eu ope, Illki ch, F ance). The pu i ied genomic DNA was dilu ed en old and used o all subsequen PCR eac ions. 14 Synopsis – Ma e ial and Me hods AFLP (Publica ion 3) AFLP is a inge p in ing me hod ha allows disc imina ion be ween indi iduals (Vos 1995). I is help ul o closely ela ed plan g oups and can de ec clonal s uc u es de i ed i.e. by apomic ic ep oduc ion (Vos 1995). Fo AFLP is o undamen al impo ance o ind app op ia e speci ic p ime s o he second speci ic PCR s ep. Fo he p elimina y p ime sea ch 24 p ime combina ions we e es ed, and he ollowing six combina ions we e hen selec ed o his s udy because hey yielded he bes esul s in species di e en ia ion: MCAA/E-ACG, M-CAC/E-ACG, M- CAC/E-ACA, M-CAT/E-ACG, M-CTC/E-ACG, MCTT/E-ACG. AFLP Da a analysis (Publica ion 3) The eac ions we e sepa a ed on a e ical elec opho esis sys em (4200 Sequence Analysis Sys em, Li-Co Biosciences, Bad Hombu g) oge he wi h DNA size ma ke s (50–700 bp Sizing S anda d, Li-Co Biosciences, bad Hombu g). AFLP banding pa e ns we e e alua ed using GeneMa ke 1-95 so wa e (So Gene ics). Band classes we e calcula ed wi h a ole ance ac o o 0.1 %. A neighbou joining (NJ) analysis o he p esence and absence ma ix was conduc ed (Nei-Li dis ance), ollowed by a boo s ap (BS) analysis a e in e node oo ing wi h 1000 eplica es using he p og am TREECON (Van de Pee and De Wach e 1994). Fo da a o S. a ia agg., we addi ionally applied model-based clus e ing (P i cha d e al. 2000) using he p og am STRUCTURE (h p://p i ch.bsd.uchicago.edu/s uc u e.h ml) in o de o e ie e he mos likely numbe o g oups wi hin he S. a ia agg. In o de o in es iga e gene ic a iabili y, he numbe o polymo phic loci and Nei´s gene di e si y “NGD” (Nei 1972) we e calcula ed wi h Popgene (Yeh and Yang 1999). As a measu e o he gene ic dis ance be ween axa we calcula ed Nei´s s anda d gene ic dis ance (Ds) using he p og am POPGENE (Yeh and Yang 1999). Ch omosome coun s (Publica ion 3) In So bus, ch omosome numbe s we e coun ed om oo ip me is ems o cul i a ed p ogeny o S. co digas ensis, S. adeana, S. anconica, and S. pannonica (one seedling pe axon), g own in he Ecological Bo anical Ga den and ha es ed in May 2010. Some o he seedlings we e also included in he AFLP analysis. The esh oo ips we e p e ea ed in 0.002 hyd oxychinoline (4h s), ixed in CARNOY´s solu ion and s ained in ca mine 15 Synopsis – Ma e ial and Me hods ollowing Snow (1963). F om he s ained oo ips we p epa ed squash p epa a ions in 45% ace ic acid, and obse ed soma ic me aphase in he mic oscope. Co ela ion analyses be ween scen and AFLP da a (Publica ion 4) Co ela ion analyses be ween in scen and AFLP da a we e made on indi idual le el and popula ion le el. Fo co ela ion analyses on popula ion le el p esence-absence da a occu ing a leas in one indi idual om bo h da a se s - AFLP (see Feulne e al. 2013, submi ed) and scen - we e used. Simila i y ma ices (Jacca d) we e calcula ed and hese we e he inpu o he RELATE co ela ion analysis (Spea man Rank co ela ion, 10,000 pe mu a ions) in PRIMER Ve sion 6 (Cla ke and Go ley 2006). 16 Synopsis – Resul s and Discussion 3.2 Resul s and Discussion Flo al scen analysis in Hie acium subgenus Pilosella and i s axonomical implica ions (Publica ion 1). In Hie acium subgen. Pilosella lo al scen s o 27 p edominan ly Ba a ian in e media e species, mos ly o he collec i e species Hie acium calodon, H. zizianum and H. densi lo um we e in es iga ed wi h dynamic headspace me hod. Re icula e analyses we e applied o depic hyb id specia ion by isualizing ela ions be ween samples placed a om each o he in a ee (Maka enko 2001). Al oge he , 56 lo al scen compounds we e iden i ied, mainly a oma ics, a y acid de i a i es and mono-, homo- and sesqui e penes. The chemical pa e ns we e ound o be axon-speci ic and a e hus o axonomical alue. The esul ha he basic (non-hyb id) species o Hie acium subgen. Pilosella such as H. piloselloides, H. echiodes o H. cymosum we e well sepa a ed by scen unde lined he u ili y o scen o axonomical in es iga ions in Hie acium. Fo many species o p esumed hyb id o igin, e icula ion analyses o scen da a allowed insigh s in o hei pa en age. One example is H. allax ssp. du ise um. This axon is an in e media e be ween H. cymosum and H. echioides. In he scen ee i clus e s close o H. cymosum, bu he e icula ion analysis connec ed i wi h H. echioides. The eason is ha i has scen compounds o bo h axa, on his basis i is possible o de ec pa en al axa (see abo e). In he scen ee, di e en subspecies o one axon (e.g. H. zizianum) o en clus e ed oge he . Excep ions ha e o be e alua ed and can ha e axonomical implica ions. Hie acium bauhini ssp. hispidissimum did no clus e wi h o he membe s o H. bauhini bu wi h H. densi lo um. In consequence i should be assigned o H. densi lo um and no o H. bauhini. The e a e mo phological cha ac e s ha suppo his placemen , sugges ing ha he e alua ion o some mo phological cha ac e s needs o be econside ed in he classi ica ion o Hie acium (in his case he cymose in lo escence s uc u e). Ano he in e es ing esul is ha in he clus e analysis some subspecies o H. densi lo um do no g oup oge he wi h he o he membe s o his g oup, bu a he wi h he H. echioides de i a es, such as H. calodon and H. allax. They also show some mo phological a ini ies o sec ion Echioides such as dense, hick and cu ed b is le-like hai s on he s em. The subspecies H. densi lo um ssp. cymosi o me and H. densi lo um ssp. psammo ophicum may ha e been de i ed om axa o sec ion Echioides, as could been de ia ed om hei s ong hick and cu ed b is le-like hai s, oo. Hie acium subgen. Pilosella axa s ongly de ia ed in hei numbe s o scen compounds (12-30 compounds). In hyb ids his may lead o an o e es ima ion o he 17 Synopsis – Resul s and Discussion in luence o one o he pa en s. Ne e heless, his can make in luences o species de ec able which a e no appa en in mo phology. An example o his aspec is he p ima y hyb id H. schneidii x pilosella. This species is ha d o di e en ia e agains H. piloselli lo um by mo phology alone. This axon was ound and iden i ied he e by lo al scen and unpublished RAPD-ma ke s (Gebaue and Feulne 2008, unpub.) o he i s ime. I g ows as a spon aneous hyb id among i s pa en s. O e all, scen pa e ns implica e ha only a low numbe o axa may be he ances o s o mos o he hyb id axa. In e es ingly, his has been sugges ed al eady by Zahn (1921-1923, 1930-35), and he p esumably pa en al species in ol ed ha e been named as basic species by him (c . Zahn 1921–1923, 1930–35). The placemen o H. caespi osum in he scen ee close o H. zizianum (an in e media e species) speaks agains i s s a us as “basic species”. I sha es also mo phological ai s wi h H. zizianum such as he s aigh hai s on he s em. Also Tichomi o (2000) conside ed H. caespi osum as hyb id be ween H. onegense (syn: H. caespi osum ssp. b e ipilum, an eas e n dis ibu ed species) and H. lac ucella. I would be in e es ing in u he s udies o in es iga e he scen o H. onegense o con i m his hypo hesis. Taxonomical alue o in lo escence scen in Hie acium s. s . (Publica ion 2) Publica ion 2 deals wi h he axonomical alue o in lo escence scen in Hie acium s. s . in Cen al Eu ope. Hie acium s. s . comp ises a as numbe o mos ly apomic ic axa p esumably o igina ed om hyb idiza ions in he pas . In lo escence scen s o 37 axa om se en sec ions o Hie acium subgen. Hie acium we e in es iga ed by headspace analyses. O e all, 58 di e en scen compounds belonging o a oma ics, sesqui e penes, homo e penes, mono e penes and a y acid de i a i es we e ound. As in H. subgen. Pilosella (publica ion 1) in lo escence scen was ound o be highly axon-speci ic in Hie acium subgen. Hie acium. Taxonomy sugges ed by scen pa e ns was compa ed wi h esul s om gene ic s udies (Feh e e al. 2009) ha include many axa in es iga ed he e by scen . Feh e e al. (2009) iden i ied wo main g oups, e med “wes e n clade” and “eas e n clade” in Hie acium s.s . using sequence da a o chlo oplas and mi ochond ial ma ke s (Feh e e al. 2009). This di e en ia ion was explained by di e en glacial e ugia (Feh e e al. 2009). The scen s udy iden i ied he same main g oups as in Feh e e al. (2009). Howe e , we ound ha hese g oups a e iden ical wi h wo dis inc mo pho- and lowe ypes, he high-g owing and la e- lowe ing-one, such as H. umbella um and H. lae iga um and he low-g owing and ea lie lowe ing mo pho ype, such as H. mu o um 18 Synopsis – Resul s and Discussion and H. bi idum. Some subs ances such as linalool and linalool oxide we e ound dominan ly in species wi h high g ow h, whe eas mono e penes such as e pinolene we e a he ypical o he low-g owing mo pho ypes. The low-g owing scen g oup comp ises sec ions such as O eadea, Hie acium and Bi ida, he high-g owing g oup comp ises he sec ions D epanoidea, T iden a a and Hie acioides. Membe s o he low-g owing species g oups such as H. wiesbau ianum o H. glaucinum may be o poly opic o igin since hei axa o en clus e ed in e mingled in he scen clus e . In con as , in he scen ee mos subspecies o H. mu o um a e placed nex o each o he and he e o e a e mos likely monophyle ic. Also H. bi idum is mainly placed in a g oup o i s own despi e some mo phologically de ia ing membe s (i.e. Hie acium bi idum ssp. s enolepis a . alde loccosum). In e es ingly, he in es iga ed H. bi idum axa o H. bi idum g ex bi idum and g ex subcaesium a e nes ed be ween H. mu o um (low g owing) and H. glaucum (high g owing) in he scen ee. This in e media e posi ion o a whole species g oup was p oposed al eady by Koch (1838) and Zahn (1906) due o i s mo phological in e mediacy and is con i med he e by scen da a. This inding shows ha lo al scen can con i m he axonomical posi ion o species e en a highe ank (i.e. sec ional le el). In H. anconicum (in e media e be ween H. mu o um and H. bupleu oides) in e popula ion di e ences could be ound. Popula ions om Baden-Wü embe g we e close o H. bupleu oides whe eas popula ions om F anconia we e close o H. mu o um. The same esul s we e e ealed by AFLP s udies (Feulne , unpublished da a). This esul was unexpec ed and shows ha hyb id axa in Hie acium s.s . could be in luenced by in og ession. The scen s udy e eals an in e media e posi ion o some axa hi he o conside ed as basic species such as H. lachenalii and H. lae iga um, suppo ing he esul s o Feh e e al. (2009) based on molecula da a. The scen s udy shows ha he in e media e axa such as H. lachenalii, H. saxi agum, H. caesium we e de i ed by mul iple hyb idiza ion e en s be ween a e y es ic ed numbe o membe s o he wo mo phological g oups (high- g owing and low-g owing). This imi a es a clinal a ia ion de i ed by s epwise e olu ion as was desc ibed by so-called educ ion lines (i.e. Hie acium umbella um - H. lae iga um - H. lachenalii - H. mu o um, comp. Zahn 1921–1923, 1930–1935). Al hough he species o Hie acium subgen. Hie acium a e olde hyb ids, hei phylogeny is no concealed by po en ial mu a ions. Thus, lo al scen composi ion is a highly conse ed ai in Hie acium s.s . 19 Synopsis – Re e ences Dö e l, S., Jü gens, A. (2005): Spa ial ag ance pa e ns in lowe s o Silene la i olia: Lilac compounds as ol ac o y nec a guides? Plan Sys ema ics and E olu ion 255: 99–109. Dö e l, S., Füssel, U., Jü gens, A., Aas, G. (2005a): 1,4-Dime hoxybenzene, a lo al scen compound in willows ha a ac s an oligolec ic bee. Jou nal o Chemical Ecology 31: 2993–2998. Dö e l, S., Wol e, L., W., Jü gens, A. (2005b): Quali a i e and quan i a i e analyses o lowe scen in Silene la i olia, Phy ochemis y 66: 203–213. Düll, R. (1961): Die So bus-A en und ih e Bas a de in Baye n und Thü ingen. Be ich e de Baye ischen Bo anischen Gesellscha 34: 11–65. Faeg i, K., an de Pijl, L. (1979): The p inciples o pollina ion ecology, Ox o d, New Yo k, To on o, Sydney, Pa is, F ank u , Pe gamon P ess L d. Feh e , J., Gemeinholze , B., Ch ek, J. J ., B äu igam, S. (2007): Incong uen plas id and nuclea DNA phylogenies e eal ancien in e gene ic hyb idiza ion in Pilosella hawkweeds (Hie acium, Cicho ieae, As e aceae). Molecula Phylogene ics and E olu ion 42: 347–361. Feh e , J., K ak, K., Ch ek, J. J . (2009): In a-indi idual polymo phism in diploid and apomic ic polyploid hawkweeds (Hie acium, Lac uceae, As e aceae): disen angling phylogene ic signal, e icula ion, and noise. BMC E olu iona y Biology 9: 239. Fens e , C.B., A mb us e , W.S., Wilson, P., Dudash, M.R., Thomson, J.D. (2004): Pollina ion synd omes and lo al specializa ion. Annual Re iew o Ecology, E olu ion and Sys ema ics 35: 375–403. Feulne , M., Schuhwe k, F., Dö e l, S. (2009): Flo al scen analysis in Hie acium subgenus Pilosella and i s axonomical implica ions. Flo a 204: 495–505. Feulne , M., Schuhwe k, F., Dö e l, S. (2011): Taxonomical alue o in lo escence scen in Hie acium s. s . Biochemical Sys ema ics and Ecology 39: 732–743. Füssel, U., Dö e l, S., Jü gens, A., Aas, G. (2007): In e - and in aspeci ic a ia ion in lo al scen in he genus Salix and i s implica ion o pollina ion. Jou nal o Chemical Ecology 33: 749–765. Gadella, T.W.J. (1984): Cy ology and he mode o ep oduc ion o some axa o Hie acium subgenus Pilosella. P oceedings o he Koninklijke Nede landse Akademie an We enschappen 87: 387–399. 26 Synopsis – Re e ences Gancel, A.L., Ollé, D., Olli aul , P., Lu o, F., B illoue , J.M. (2002): Lea and peel ola ile compounds o an in e speci ic ci us soma ic hyb id [Ci us au an i olia (Ch is m.) Swing. + Ci us pa adisi Mac ayden]. Fla ou and F ag ance Jou nal 17: 416–424. Gögle , J., S ökl, J., S amko a, A., Twele, R., F ancke, W., Cozzolino, S., Co is P., Sc ugli, A., Ayasse, M. (2009): Ménage à ois - wo endemic species o decep i e o chids and one pollina o species. E olu ion 63: 2222–2234. Heuss, Lisa (2011): Bes äube und Blü endü e e schiedene So bus Taxa. Bachelo hesis am LS P lanzensys ema ik Uni e si ä Bay eu h, unpublished. Hö andl, E., Hojsgaa d, D. (2012): The e olu ion o apomixis in angiospe ms: A eapp aisal. Plan Biosys ems 146: 681–693. Jankun, A., Ko anda, M. (1987): Emb yological s udies o So bus 2. apomixis and o igin o So bus bohemica. P eslia 59: 97–116. Junke , R.R., Blü hgen, N. (2008): Flo al scen s epel po en ially nec a - hie ing an s. E olu iona y Ecology Resea ch 10: 295–308. Junke , R.R., Loewel, C. G oss, R., Dö e l, S., Kelle A., Blü hgen, N. (2011): Composi ion o epiphy ic bac e ial communi ies di e s on pe als and lea es. Plan Biology 13: 918–924. Ká pá i, Z. (1960): Die So bus-A en Unga ns und de ang enzenden Gebie e. Feddes Repe o ium 62: 71–334. Kessle , D.I., Baldwin, T. (2007): Making sense o nec a scen s: he e ec s o nec a seconda y me aboli es on lo al isi o s o Nico iana a enua e. The Plan Jou nal 49: 840–854. Knudsen, J.T., Tolls en L. (1993): T ends in lo al scen chemis y in pollina ion synd omes: lo al scen composi ion in mo h-pollina ed axa. Bo anical Jou nal o he Linnean Socie y 113: 263–284. Knudsen, J.T., Tolls en, L. (1995): Flo al scen in ba -pollina ed plan s: a case o con e gen e olu ion. Bo anical Jou nal o he Linnean Socie y 119: 45–57. Koch, D.W.D.J. (1838): Synopsis de deu schen und Schweize Flo a., Wilmans, F ank u a. M. 840 p. Kol unow, A., G ossniklaus, U. (2003): Apomixis: A de elopmen al pe spec i e. Annual Re iew o Plan Biology 54: 547–574. Kol unow, A., Johnson, S.D., Okada, T. (2011): Apomixis in hawkweed: Mendel’s expe imen al nemesis. Jou nal o Expe imen al Bo any: 1–9. 27 Synopsis – Re e ences K ahulco á, A., K ahulec, F., (2000): O sp ing di e si y in Hie acium subgen. Pilosella (As e aceae): new cy o ypes om hyb idisa ion expe imen s and om open pollina ion. F agmen a Flo is ica e Geobo anica Polonica 45: 239–255. K ahulco á, A., K ahulec, F., Chapman, H.M. (2000): Va ia ion in Hie acium subgen. Pilosella: wha do we know abou i s sou ces? Folia Geobo anica 35: 319–338. Ku zelnigg, H. (1995): So bus. In: Cone H.J. e al. (eds): Gus a Hegi, Illus ie e Flo a on Mi eleu opa 4 (2B): Spe ma ophy a: Angiospe mae: Dico yledones 2 (3). Rosaceae 2. Blackwell, Be lin: 328–385. Ku zelnigg, H. (2005): So bus L. em. C an z – Mehlbee e, Ebe esche, Elsbee e, Speie ling. In: Jäge , E.J., Ro hmale , K.W. (eds): Exku sions lo a on Deu schland 4, Ge äßp lanzen, K i ische Band, 9 h edn. Spek um, Heidelbe g: 403–406. Legend e, P., Maka enko , V. (2002): Recons uc ion o biogeog aphic and e olu iona y ne wo ks using e iculog ams. Sys ema ic Biology 51: 199–216. Le in, R.A., McDade, L.A., Raguso, R.A. (2003): The sys ema ic u ili y o lo al and ege a i e ag ance in wo gene a o Nyc aginaceae. Sys ema ic Biology 52: 334– 351. Lepší, M., Ví , P., Lepší, P., Boublík, K., Kolář, F. (2009): So bus po ae-bohemicae and So bus albensis, wo new endemic apomic ic species ecognized based on a e ision o So bus bohemica. P eslia 8: 63–89. Maka enko , V. (2001): T-REX: Recons uc ing and isualizing phylogene ic ees and e icula ion ne wo ks. Bioin o ma ics 17: 664–668. Me xmülle , H. (1982): Hie acium schneidii – Ein unbekann e Baye ische Endemi . Be ich e de Baye ischen Bo anischen Gesellscha 53: 91–95. Meye , N., Meie o , L., Schuwe k, H., Ange e , O. (2005): Bei äge zu Ga ung So bus in Baye n. Be ich e de Baye ischen Bo anischen Gesellscha Sonde band: 5–216. M áz, P., Ch ek, J.J., Feh e , J., Plačko á, I. (2005): Ra e ecen na u al hyb idiza ion in Hie acium s.s . - e idence om mo phology, allozymes and chlo oplas DNA. Plan Sys ema ics and E olu ion 255: 177–192. Nägeli, C.V., Pe e , A., (1885): Die Hie acien Mi el-Eu opas. Monog aphische Bea bei ung de Piloselloiden mi besonde e Be ücksich igung de mi eleu opäischen Sippen. München. Nei, M. (1972): Gene ic dis ance be ween popula ions. Ame ican Na u alis 106: 283–291. Nogle , G.A. (1984): Game ophy ic apomixis. In: Joh i, B M. (ed.): Emb yology o angiospe ms. Sp inge -Ve lag, Be lin: 475–518. 28 Synopsis – Re e ences Plepys, D., Iba a, F., Lö s ed , C. (2002): Odou -media ed nec a o aging in he sil e Y mo h, Au og apha gamma (Lepidop e a: Noc uidae): beha iou al and elec ophysiological esponses o lo al ola iles. Oikos 99: 75–82. P i cha d, J.K., S ephens, M., Donnelly, P. (2000): In e ence o popula ion s uc u e using Mul ilocus Geno ype Da a. Gene ics 155: 945–959. Raguso, R.A. (2001): Flo al scen , ol ac ion, and scen d i en o aging beha iou . Cogni i e Ecology o Pollina ion; Animal Beha iou and Flo al E olu ion (eds L. Chi ka & J.D. Thomson). Camb idge Uni e si y P ess, Camb idge: 83–105. Raguso, R.A., Schlumpbe ge , B.O., Kaczo owski, R.L., Hol s o d, T.P. (2006): Phylogene ic ag ance pa e ns in Nico iana sec ions Ala ae and Sua eolen es. Phy ochemis y 67: 1931–1942. Rich, T., Hous on, L., Robe son, A. (2010): Whi ebeams, Rowans and Se ice T ees o B i ain and I eland A Monog aph o B i ish and I ish So bus L. BSBI Handbook 14, London. Robe son, A., Rich, T.C.G., Allen, A.M., Hous on, L, Robe s, C., B idle, J.R., Ha is, S.A., Hiscock, S.J. (2010): Hyb idiza ion and polyploidy as d i e s o con inuing e olu ion and specia ion in So bus. Molecula Ecology 19: 1675–1690. Rosenbaumo á, R., K ahulco á, A., K ahulec, F. (2012): The in iguing complexi y o pa henogenesis inhe i ance in Pilosella ub a (As e aceae, Lac uceae). Sexual Plan Rep oduc ion 25: 185–196. Schä le , I., Balao, F., Dö e l, S. (2012): Flo al and ege a i e cues in oil-sec e ing and non-oil-sec e ing Lysimachia species. Annals o Bo any 110: 125–138. Schies l, F.P., Dö e l, S. (2012): The e olu ion o lo al scen and ol ac o y p e e ences in pollina o s: coe olu ion o p e-exis ing bias? E olu ion 66: 2042–2055. Schuhwe k, F., (2002): Some hough s on he axonomy o Hie acium. Be ich e de Baye ischen Bo anischen Gesellscha 72: 193–198. Schuhwe k, F. (2008): Zu Gliede ung de Hie acium bi idum-G uppe in Baye n und benachba en Gebie en. – In: Dengle , J., Dolnik, C. & T epel, M. (eds.): Flo a, Vege a ion und Na u schu z zwischen Schleswig-Hols ein und Südame ika – Fes sch i ü Klaus Die ßen zum 60. Gebu s ag. Mi eilungen de A bei sgemeinscha Geobo anik ü Schleswig-Hols ein und Hambu g 65: 375– 392. Snow, R. (1963): Alcoholic hyd ochlo ic acid-ca mine as s ain o ch omosomes in squash p epa a ions. S ain Technology 38: 9–13. 29 Synopsis – Re e ences S eine , K.E., Kaise , R., Dö e l, S. (2011): S ong phylogene ic e ec s on lo al scen a ia ion o oil-sec e ing o chids in Sou h A ica. Ame ican Jou nal o Bo any 98: 1663–1679. Talen , N. (2009): E olu ion o game ophy ic apomixis in lowe ing plan s: an al e na i e model om Maloid Rosaceae. Theo y in Biosciences 128: 121–138. Tichomi o , V.A. (2000): The genus Pilosella (As e aceae) in he lo a o bela us. Bo anicheskii Zhu nal 5: 104–114. Van de Pee , Y., De Wach e , Y. (1994): TREECON o Windows: a so wa e package o he cons uc ion and d awing o e olu iona y ees o he Mic oso Windows en i onmen . Compu e Applica ions in he Biosciences 10: 569–70. Ve eecken, N.J., Cozzolino, S., Schies l, F.P. (2010). Hyb id lo al scen no el y d i es pollina o shi in sexually decep i e o chids. BMC E olu iona y Biology 10:103. Vos, P., Hoge s, R., Bleeke , M. (1995): AFLP: a new echnique o DNA inge p in ing. Nucleic Acids Resea ch 23: 4407–4414. Whi ehead, M.R., Peakall, R. (2009): Flo al scen in a whole plan con ex . In eg a ing lo al scen , pollina ion ecology and popula ion gene ics. Func ional Ecology 23: 863–874. Yeh, F., Yang, R. (1999): POPGENE. Mic oso Window-based F eewa e o Popula ion Gene ic Analysis. Ve sion 1.31, Uni e si y o Albe a: Edmon on, Albe a, Canada. Zahn, K. H. (1906): Die Hie acien de Schweiz. Neue Denksch i en de Allgemeinen Schweize ischen Gesellscha ü Gesamm en Na u wissenscha en 40: 165–728, Zü ich. Zahn, K. H. 1921–1923. Composi ae – Hie acium. In: Engle , A. (ed.), Das P lanzen eich. Regni ege abilis conspec us. Ve lag on Wilhelm Engelmann, pp. 1–1146. Zahn, K. H. 1930–1935. Hie acium 2. In: Asche son, P. e al. (eds), Synopsis de mi eleu opäischen Flo a, zwöl e Band, zwei e Ab eilung. Ve lag on Geb üde Bo n aege , pp. 1–790. 30 Publica ions Publica ion 1 1. Flo al scen analysis in Hie acium subgenus Pilosella and i s axonomical implica ions. Published in Flo a 2009 31 Au ho 's pe sonal copy Flo a 204 (2009) 495–505 Flo al scen analysis in Hie acium subgenus Pilosella and i s axonomical implica ions Ma in Feulne a, , F anz Schuhwe k b , S e an Do¨ e l a a Depa men o Plan Sys ema ics, Uni e si y o Bay eu h, Uni e si a ¨ ss 30, 95440 Bay eu h, Ge many b Bo anische S aa ssammlung Mu ¨nchen, Menzinge S aße 67, D-80638 Mu ¨nchen Recei ed 5 Ap il 2008; accep ed 30 June 2008 Abs ac Species- ich Hie acium subgen. Pilosella is well-known o a high deg ee o endemism and in a-speciﬁc di e en ia ion including many subspecies (‘‘mic ospecies’’) o e y es ic ed dis ibu ion. In Hie acium subgen. Pilosella ﬂo al scen s o 27 p edominan ly Ba a ian species, mos ly o Hie acium calodon, H. zizianum and H. densi lo um, a e in es iga ed he e. Flo al scen composi ions we e s udied by GC-MS analysis o dynamic headspace samples. Al oge he , 56 ﬂo al scen compounds we e iden iﬁed, mainly benzenoids, a y acid de i a i es, mono e penes, homo e penes and sesqui e penes. The chemical pa e ns we e ound o be axon-speciﬁc and a e hus o axonomical alue. The da a suppo some ea angemen s a subspeciﬁc le el, such as he inclusion o H. bauhini subsp. hispidissimum in H. densi lo um. These ea angemen s a e suppo ed by mo phological da a. The adi ional species concep s, howe e , a e mos ly co obo a ed by ou scen da a. 2008 Else ie GmbH. All igh s ese ed. Keywo ds: Plan e penes; Chemo axonomy; Mic ospecies; Re icula ion; Apomic ic plan s In oduc ion The main ques ion o ﬂo al scen in es iga ion ega ds in e ac ions o ﬂowe ing plan s and hei pollina o s. Flo al scen componen s a e well-known o playing an impo an ole as a ac an s o pollina o s (e.g., Dobson e al., 2005;Do¨ e l e al., 2006;Plepys e al., 2002). Flo al scen compounds can also unc ion as epellen o he bi o es o pa hogens. Howe e , Le in e al. (2003) poin ou ha hese compounds may no be unc ional in e e y case, because hey a e o en by-p oduc s o he me abolism. Flo al scen s can be species-speciﬁc, helping o main ain species in eg i y ia hei e ec s o pollina o p eselec ion. Di e ences among species in ﬂo al scen composi ion we e de ec ed in a la ge ange o plan amilies (Knudsen e al., 2006). Howe e , p oﬁles o ﬂo al scen compounds we e a ely used o phylogene ic analyses and axonomy. An objec ion o he use o ﬂo al scen pa e ns in sys ema ics is ha con e gence o ﬂo al scen s may play an impo an ole in plan s in gene al, caused by simila pollina o p essu es ac ing on plan s o independen o igin (Dobson e al., 2005). Ne e - heless, compa a i e s udies o ﬂo al scen and DNA da a in o chids (Ba kman, 2001;Williams and Whi en, 1998) as well as in Nyc aginaceae (Le in e al., 2003) e ealed ha he ee opologies gene a ed by ARTICLE IN PRESS www.else ie .de/ﬂo a 0367-2530/$ - see on ma e 2008 Else ie GmbH. All igh s ese ed. doi:10.1016/j.ﬂo a.2008.06.003  Co esponding au ho . E-mail add esses: [email p o ec ed] (M. Feulne ), [email p o ec ed] (F. Schuhwe k), s e an.doe e l@uni-bay eu h.de (S. Do¨ e l). 32 Au ho 's pe sonal copy ARTICLE IN PRESS Table 1. Vouche and locali y in o ma ion o plan ma e ial used in his s udy. Taxon Vouche Numbe o samples pe popula ion H. lac ucella Wall . Neus a¨d lein, 491580N, 111250; 15.7.2006, Feulne 1 (UBT) 2 H. pilosella L. Neus a¨d lein, 491580N, 111250; 30.6.2004, Feulne 2 (UBT) 1 Obe ails eld, 491490N, 111210O; 21.6.2005, Feulne 3 (UBT) 2 O ¨kol. Bo . Ga en Bay eu h, 491560N, 111350O; 25.6.2004, Feulne 4 (UBT) 3 H. bauhini Schul . subsp. bauhini Bindlach, 501000N, 111370O; 12.6.2006, B ei eld and Feulne 5 (UBT) 2 H. bauhini Schul . subsp. hispidissimum (Rehm.) Zahn Go¨ßweins ein, 491460N, 111190O; 10.6.2005, Feulne and Bolze 6 (He ba Bolze 2) 3 H. piloselloides Zahn subsp. p aeal um (Vill. ex Gochna ) Po ens ein, 491460N, 111250O; 25.6.2005, Feulne 7 (UBT) 2 Velbu g, 491130N, 111400O; 23.6.2005, Feulne 8 (UBT) 1 H. cymosum L. subsp. cymosum E e zhausen, 491020N, 111590O; 1.7.2006, Feulne 91 O ¨kol. Bo . Ga en Bay eu h, 491560N, 111350O; 10.6.2005, Feulne 10 (UBT) 2 H. caespi osum Dumo . subsp. caespi osum O ¨kol. Bo . Ga en Bay eu h, 491560N, 111350O; 15.6.2006, Feulne 11 (UBT) 2 Weiden, 491400N, 121100O; 1.7.2006, Feulne 12 (UBT) 1 Zeil, 501010N, 101350O; 25.6.2006, Feulne 13 (UBT) 1 H. echioides Lumn. subsp. echioides B andenbu g, B odowin, 521550N, 131560O; 15.7.2006, Feulne 15 (UBT) 3 H. densi lo um Tausch subsp. psammo ophicum (Schack and Zahn) Al do , 491330N, 111350O; 10.6.2006, Feulne 16 (UBT) 2 H. densi lo um Tausch subsp. bauhini olium (NP.) Velbu g, 491130N, 111400O; 11.6.2005, Feulne and Meye (He ba Meye 1) 2 Hie acium densi lo um Tausch subsp. umbelli e um (Na¨geli & Pe e ) Go schl. Po ens ein, 491460N, 111250O; 1.7.2005, Bolze (He ba Bolze 5) 3 H. densi lo um Tausch subsp. och ocephaloides (Ha z and Zahn) Lochau, 491580N, 111230O; 15.7.2006, Feulne 18 (UTB) 2 H. densi lo um Tausch subsp. cymosi o me (NP.) Velbu g, 491130N, 111400O; 7.2005, Meye 99-13.3a (M) 2 H. calodon phyllopho um NP. Go¨ßweins ein, 491460N, 111190O; 1.7.2005, Feulne and Bolze (He ba Bolze 1) 3 H. calodon Tausch ex Pe e subsp. pseudo allax Tou on Neudo , 501040N, 111160O; 15.7.2006, Feulne 19 (UTB) 3 H. calodon ‘‘Ra ensbu g’’ Thu¨nge sheim, 491520N, 91500O; Schuhwe k 95/27 & Meie o (M) 3 H. schneidii Schack and Zahn Obe ails eld, 491490N, 111210O; 8.6.2005, Feulne 21 (UTB) 3 Waischen eld, 491500N, 111210O; 8.6.2005, Bolze (He ba Bolze 4) 2 Zauppenbe g, 491490N, 111230O; 9.6.2006, Bolze (He ba Bolze 5) 1 H. allax F oel. subsp. du ise um NP. Regensbu g, 491010N, 121070O; 2005, Schuhwe k s. n . (Me xmu ¨lle 33371) (M) 2 H. zizianum Tausch subsp. pachyphyes Zahn E e zhausen, 491020N, 111590O; 1.7.2006, Feulne 24 (UTB) 2 H. zizianum Tausch subsp. zizianum S eiﬂing, 491490N, 111240O; 29.5.2005, Feulne 25 (UTB) 2 Pegni z-Bahnho , 491450N, 111320O; 17.7.2006, Feulne 26 (UTB) 1 H. zizianum Tausch subsp. adenocymige um Ge s l. and Zahn Ki chaho n, 491500N, 111230O; 15.7.2007, Feulne 28 (UTB) 3 H. spu ium Chaix subsp. ubula um (Vollm.) Zahn Wel enbu g, 481530N, 111490O; 3.6.2006, Schuhwe k 87/36 and Lippe (M) 3 H. allacinum F. W. Schul z subsp. allacinum U ¨ ingen, 491470N, 91430O; 20.5.2005, Feulne and Meye 29 (UTB) 4 H. allacinum F. W. Schul z S eiﬂing, 491490N, 111240O; 15.6.2005, Bolze and Feulne (He ba Bolze 3) 3 H. schneidii x H. pilosella Obe ails eld, 491490N, 111210O; 7.2005, Feulne 31 (UTB) 2 Po ens ein, 491460N, 111250O; 7.2005, Feulne 32 (UTB) 1 H. glome a um F oel. subsp. glome a um Ho ,501180N, 111540O; 2006, Feulne 33 (UTB) 1 Cobu g, 501150N, 101570O; 2006, Feulne 34 (UTB) 1 H. au an iacum L. subsp. au an iacum Obe s do , 471240N, 101160O; 2006, Feulne 35, (UTB) 2 O ¨kol. Bo . Ga en Bay eu h, 491560N, 111350O; 2005, Feulne 36 (UTB) 3 M. Feulne e al. / Flo a 204 (2009) 495–505496 33 Au ho 's pe sonal copy non-coding DNA ma ke s a e in some cases cong uen wi h hose based on ﬂo al scen da a. The me hods o analyse scen pa e ns o axonomi- cal ques ions a e mani old (Ba kman, 2001;Le in e al., 2003). The simples me hod is he use o p esence/ absence da a o di e en ﬂowe ola iles. Fo axono- mical ques ions, di e en ia ion be ween compounds o ﬂo al and g een pa s o he plan s seems o be negligible (c . Le in e al., 2003). Conside a ions on he axonomic alue o ﬂo al ag ance analysis concen a ed so a on ou b eeding g oups, while inb eeding g oups and g oups wi h e icula e ela ionships ha e no been conside ed. Re icula ion in Scandina ian species o Hie acium subgen. Pilosella was p o en by Tyle ’s (2005) in es iga- ions using isoenzyme ma ke s. He ound no speciﬁc pa e ns o isoenzymes use ul o species disc imina ion, and concluded ha gene ﬂow mus be common among he species. Re icula ion in Hie acium subgen. Pilosella comp ises a high deg ee o pu a i e hyb id species wi h in e media e cha ac e s o en gene ically isola ed by apomixis. Fu he , di e en ep oduc i e modes, such as allogamy, au ogamy o apomixis can occu oge he in he same capi ula (K ahulco a´ e al., 2000). Apomic- ic elemen s e en can in og ess as K ahulco a´ and K ahulec (2000) ha e ound in a iﬁcial c ossing expe imen s in ol ing pen aploid apomic ic Hie acium species. In his case, he apomic ic pen aploids jus se e as pollen dono s, whe eas he e a- and diploids ac as ecipien s. A la ge numbe o species in Hie acium subgen. Pilosella a e poo ly cha ac e ized mo phologically. Di ﬁcul ies o iden i y he elemen s o he complex ha e s imula ed Hie acium axonomis s o desc ibe ‘‘collec- i e species’’–binding oge he o mal subspecies o mic ospecies–al hough many o hese elemen s beha e in na u e as ﬁxed apomic ic species (c . Schuhwe k, 2002). Na¨geli and Pe e (1885) p oposed he ﬁ s axonomic concep o Hie acium which is s ill in use in Cen al Eu ope, whe eas in o he pa s o Eu ope (e.g., Scandina ia and Russia) al e na i e concep s a e applied (c . Schuhwe k, 2002). The concep o Na¨geli and Pe e (1885) is based on he idea ha he e a e some well dis inguishable species, he so called ‘‘basic species’’, wi h unique mo phological cha ac e s. How- e e , mos o he elemen s in Hie acium subgen. Pilosella show in e media e cha ac e s, he e o e, hey a e ea ed as ‘‘in e media e species’’ by Na¨geli and Pe e (1885). The au ho s conside ed some o hem o be o hyb id o igin. In his s udy, we in es iga e 27 axa mos ly o Ba a ian o igin (Table 1), ep esen ing abou one hi d o he Ba a ian species o subgen. Pilosella. In he axa s udied, bo h spon aneous hyb ids g owing oge he wi h a leas one pa en al species only, and pu a i e hyb ids showing independen ai s in mo phology, ecology o dis ibu ion a e included. He e, we ocus especially on H. densi lo um, H. zizianum, H. allax and H. calodon,all in e media es be ween he main species H. cymosum, H. echioides, H. piloselloides and H. bauhini. Some axa, e.g., H. densi lo um subsp. cymosi o me, H. densi lo um subsp. bauhini olium o H. bauhini subsp. hispidissimum display mo phological cha ac e s ques ioning hei membe ship in he desc ibed collec i e species. Fu he mo e hese axa a e e y a e o e en endemic o Ba a ia. Membe s o he ‘‘Echinina’’ species g oup (H. allax, H. calodon and H. schneidii) a e well-known glacial elic s (Me xmu¨ lle , 1982) and, he e o e, o special in e es . In Hie acium subgen. Pilosella many species a e expec ed o be o hyb id o igin. Agains his back- g ound, i will be es ed whe he ﬂo al scen subs ances a e use ul ma ke s o de ec ion o hyb ids and hei o igin. Ma e ial and me hods S udy plan s This pape deals wi h axa o Hie acium subgen. Pilosella in Ge many wi h special a en ion o species- ich Ba a ia. All accessions, au ho s o axa, ouche specimens and locali ies a e gi en in Table 1. Taxa o a subspecies le el we e gi en when hei s a e was conﬁ med by p e ious s udies, i.e. Schuhwe k (2002); Schuhwe k and Lippe (1997, 2002) and Go schlich (1996). Vola ile collec ion Flo al scen was collec ed in he ﬁeld using he dynamic headspace me hod desc ibed by Do¨ e l and Ju¨ gens (2005), and Do¨ e l e al. (2005). Capi ula we e enclosed wi hin a polyes e o en bag (Toppi s s ) and he emi ed ola iles we e apped in an adso ben ube h ough he use o a memb ane pump (ASF Thomas, Inc.) o eigh minu es. As abso ben ube, we ook Ch oma oP obe qua z mic o ials o Va ian Inc. (leng h: 15 mm; inne diame e : 2 mm), cu he closed end, ﬁlled hem wi h a mix u e (1:1) o 3 mg Tenax-TA (mesh 60–80) and Ca bo ap (mesh 20–40), and ﬁxed he adso ben mix u e in he ial wi h glass wool. Simul a- neous collec ions o bo h he ﬂowe scen and su ound- ing ai we e used o dis inguish be ween ﬂo al compounds and ambien con aminan s. Sampling was ca ied ou on esh and newly opened capi ula, be ween 11 a.m. and 3 p.m., he pe iod wi h he mos in ensi e scen emission (Feulne , unpublished da a). In mul i-headed synﬂo escences mo e han one open capi ulum was sampled. ARTICLE IN PRESS M. Feulne e al. / Flo a 204 (2009) 495–505 497 34 Au ho 's pe sonal copy Chemical analysis The samples we e analysed using a Va ian Sa u n 2000 mass spec ome e , and a Va ian 3800 gas ch oma og aph wi h a 1079 injec o , ha had been ﬁ ed wi h he Ch oma oP obe ki . This ki allows he he mal deso p ion o small amoun s o solids o liquids con ained in qua z mic o ials (Mic o-SPE; c . Ami a and Dagan, 1997). The injec o spli en is opened (1/20) o ﬂush all ai om he sys em and closed a e 2 min; simul aneously, he injec o is hea ed om 40 1C ( empe a u e du ing he ﬁ s 2 min) wi h a a e o 200 1C/min o 200 1C; his empe a u e is held o 4.2 min, a e which he spli en opens (1/10) and he injec o cools down. A ZB-5 column (5% phenyl polysiloxane) was used o he analyses (60 m long, inne diame e 0.25 mm, ﬁlm hickness 0.25 mm, Phenomenex). Elec onic ﬂow con- ol was used o main ain a cons an helium ca ie gas ﬂow o 1.8 ml min 1 . The GC o en empe a u e was held o 7 min a 40 1C, hen inc eased by 6 1C/min o 250 1C and held o 1 min. The MS in e ace was 260 1C and he ion ap wo ked a 175 1C. The mass spec a a e aken a 70 eV (in EI mode), wi h a scanning speed o 1 scan s 1 om m/z 30 o 350. Da a analysis The GC-MS da a we e p ocessed using he Sa u n So wa e package 5.2.1. Componen iden iﬁca ion was ca ied ou using he NIST 02 mass spec al da abase, o MassFinde 2.3, and conﬁ med by he compa ison o e en ion imes wi h published da a (Adams, 1995; Da ies, 1990). Iden iﬁca ion o indi idual componen s could be conﬁ med by he compa ison o bo h mass spec um and GC e en ion da a wi h hose o au hen ic s anda ds. S a is ical analysis Fo s a is ical analyses, a simila i y ma ix (So¨ ensen simila i ies) was cons uc ed using a p esence/absence ma ix o scen da a. The signiﬁcance o di e ences in scen p oﬁles among axa was assessed by ANOSIM (Cla ke and Go ley, 2001), wi h 10,000 andom pe mu a ions. Fo u he analyses, he axa samples (mos ly 2–5 indi iduals pe axon) we e me ged by coun ing a single subs ance, i i was ep esen ed in a leas hal o he samples. To analyse he ela ionships among he axa, we conduc ed a e icula ion ne wo k analysis wi h he p og am - ex Ve sion 4.0a1 (Maka enko , 2001). Fo analysis o g oups wi h many species o hyb id o igin, no mal ee models a e no sui able, because hey canno depic ela ionships o species in e connec ed wi h mo e han one ances o (Legend e and Maka enko , 2002). The e o e, hese au ho s de eloped he so wa e - ex Ve sion 4.0a1 (Maka enko , 2001;Legend e and Maka enko , 2002), which is applied o his s udy. The la e p og am is calcula ing a dis ance educ ion be ween he p elimina y joined a angemen s by a special algo i hm made isible by adding e icula ion lines. The dashed lines a e symbols o homoplasy in he da a se (Maka enko , 2001;Legend e and Maka en- ko , 2002). T- ex canno dis inguish be ween he easons o homoplasy, be i gene ﬂow ia hyb idisa ion e en s o con e gence. Resul s A o al o 56 di e en ﬂo al scen compounds we e ound in he 27 in es iga ed Hie acium axa (Table 2). The e a e ema kable di e ences among he axa conce ning he numbe o scen componen s anging om 19 componen s in H. pilosella o 31 in H. bauhini subsp. bauhini. The iden iﬁed compounds belong o benzenoids, sesqui e penes, mono e penes and a y acid de i a i es. The mos commonly occu ing compounds we e (Z)-3-Hexen-1-ol ( ound in all in es iga ed species), D-Limonene, (E)-b-Ocimene, Ylangene (all ound in 27 o 28 axa), (Z)-Ocimene (26 axa), (E)-4,8-Dime hyl- 1,3,7-nona iene and Me hylsalicyla e (in 25 axa). Mos scen samples we e domina ed by (E)-b-Ocimene (a e - age amoun 20%), (Z)-3-Hexen-1-ol (15%) and (E)-4,8 Dime hyl-1,3,7-nona iene (11%). The ela i e amoun o (Z)-3-Hexen-1-ol was highes in H. zizianum subsp. pachyphyes wi h 67%; (E)-4,8-Dime hyl-1,3,7-nona- iene eached he highes ela i e amoun in H. calodon x H. allacinum (36%), and (E)-b-Ocimene in H. densi lo um subsp. psammo ophicum (49%). Membe s o H. densi lo um emi ed a b oade ange o mono e penes, and -Phelland ene, l-Fenchone and Linalool eached high ela i e amoun s in hese axa. Fo H. cymosum subsp. cymosum o H. zizianum subsp. zizianum high alues o sesqui e penes as Copaene and Ylangene we e ound. In H. zizianum subsp. zizianum, H. piloselloides subsp. p aeal um and H. allacinum subsp. allacinum highe amoun s o acids (Hexanoic acid and Oc anoic acid) we e iden iﬁed. (Table 2) The p esen -absen da a o scen a e highly speciﬁc in he in es iga ed Hie acium axa (ANOSIM R- alue ¼0.769, po0.01), allowing he iden iﬁca ion o axa by ﬂo al scen da a. The e o e, he samples o one axon we e be me ged o he e icula ion analysis. The esul o he e icula ion analysis is shown in Fig. 2. The ee consis s o h ee main g oups, in which he axa a e a anged in mos cases as expec ed by mo phology. Howe e , some di e ences o he concep ARTICLE IN PRESS M. Feulne e al. / Flo a 204 (2009) 495–505498 35 Au ho 's pe sonal copy Swing.+Ci us pa adisi Mac ayden]. Fla ou F ag ance J. 17, 416–424. Go schlich, G., 1996. Hie acium. In: Sebald, O., Seybold, S., Philippi, G., Wo¨ z, A. (Eds.), Die Fa n- und Blu¨ enpﬂanzen Baden-Wu¨ embe gs, ol. 6. Ulme , S u ga , pp. 393–535. Ju¨ gens, A., Do¨ e l, S., Me e, U., 2006. The chemical na u e o e id ﬂo al odou s in s apeliads (Apocynaceae–Asclepia- doideae–Ce opegieae). New Phy ol. 172, 452–468. Knudsen, J.T., E iksson, R., Ge shenzon, J., S ( ahl, B., 2006. Di e si y and dis ibu ion o ﬂo al scen . Bo . Re . 72, 1–120. K ahulco a´, A., K ahulec, F., 2000. O sp ing di e si y in Hie acium subgen. Pilosella (As e aceae): new cy o ypes om hyb idisa ion expe imen s and om open pollina ion. F agm. Flo . Geobo . 45, 239–255. K ahulco a´, A., K ahulec, F., Chapman, H.M., 2000. Va ia- ion in Hie acium subgen. Pilosella (As e aceae): wha do we know abou i s sou ces? Folia Geobo . 35, 319–338. Legend e, P., Maka enko , V., 2002. Recons uc ion o biogeog aphic and e olu iona y ne wo ks using e iculo- g ams. Sys . Biol. 5, 199–216. Le in, R.A., McDade, L.A., Raguso, R.A., 2003. The sys ema ic u ili y o ﬂo al and ege a i e ag ance in wo gene a o Nyc aginaceae. Sys . Biol. 52, 334–351. Maka enko , V., 2001. T-REX: econs uc ing and isualizing phylogene ic ees and e icula ion ne wo ks. Bioin o - ma ics 17 (7), 664–668. Me xmu¨lle , H., 1982. Hie acium schneidii—Ein unbekann e baye ische Endemi . Be . Baye . Bo . Ges. 53, 91–95. Na¨geli, C.V., Pe e , A., 1885. Die Hie acien Mi el-Eu opas. Monog aphische Bea bei ung de Piloselloiden mi beson- de e Be u¨cksich igung de mi eleu opa¨ischen Sippen, Mu¨nchen. Plepys, D., Iba a, F., Lo¨ s ed , C., 2002. Vola iles om ﬂowe s o Pla an he a bi olia (O chidaceae) a ac i e o he sil e Y mo h, Au og apha gamma (Lepidop e a: Noc ui- dae). Oikos 99, 69–74. Schuhwe k, F., 2002. Some hough s on he axonomy o Hie acium. Be . Baye . Bo . Ges. 72, 193–198. Schuhwe k, F., Lippe , W., 1997. Ch omosomenzahlen on Hie acium (Composi ae, Lac uceae) Teil 1. Send ne a 4, 181–206. Schuhwe k, F., Lippe , W., 2002. Ch omosomenzahlen on Hie acium (Composi ae, Lac uceae) Teil 4. Send ne a 8, 167–194. Tyle , T., 2005. Pa e ns o allozyme a ia ion in No dic H pilosella.. Plan Sys . E ol. 250, 133–145. Wes ich, P., 1989. Die Wildbienen Baden-Wu¨ embe g. Ulme , S u ga , pp. 373–374. Williams, N.H., Whi en, W.M., 1998. Molecula phylogeny and ﬂo al ag ances o male euglossine bee-pollina ed o chids: a s udy o S anhopea (O chidaceae). Plan . Spec. Biol. 14, 129–136. Zahn, K.H., 1921–1923. Composi ae-Hie acium. Das Pﬂan- zen eich 4 (280). W. Engelmann, Leipzig. Zahn, K.H., 1930–1935. Hie acium. Synopsis de mi eleu - opa¨ischen Flo a. Geb . Bo n a¨ge , Leipzig, 12 (1). ARTICLE IN PRESS M. Feulne e al. / Flo a 204 (2009) 495–505 505 42 Publica ion 2 2. Taxonomical alue o in lo escence scen in Hie acium s. s . Published in Biochemical Sys ema ics and E olu ion 2011 43 Au ho 's pe sonal copy Taxonomical alue o inﬂo escence scen in Hie acium s. s . Ma in Feulne a , * , F anz Schuhwe k b , S e an Dö e l a a Depa men o Plan Sys ema ics, Uni e si y o Bay eu h, Uni e si ä ss . 30, 95440 Bay eu h, Ge many b Bo anische S aa ssammlung München, Menzinge S aße 67, D-80638 München, Ge many a icle in o A icle his o y: Recei ed 7 Janua y 2011 Accep ed 25 June 2011 A ailable online 5 Augus 2011 Keywo ds: Inﬂo escence ola iles Re icula ion analysis As e aceae Apomic ic species complex Ancien hyb idiza ion abs ac In Cen al Eu ope Hie acium s. s . comp ises a as numbe o mos ly apomic ic axa p esumably o igina ed om hyb idiza ion in he pas . Inﬂo escence scen s o 37 axa om 7 sec ions o Hie acium subgenus Hie acium we e in es iga ed by headspace analysis. O e all, 58 di e en scen compounds belonging o benzenoids, sesqui e penes, homo- e penes, mono e penes and a y acid de i a i es we e ound. The scen pa e ns we e used o pe o m a neighbou joining and e icula ion analysis and he esul s a e discussed agains he backg ound o cu en axonomy. The scen clus e ing e ealed a clea seg e- ga ion be ween sec ions D epanoidea, T iden a a and Hie acioides agains membe s o he sec ions Hie acium, O eadea and Biﬁda. The scen ee eﬂec ed dis inc mo phological and phenological g oups in Hie acium and we e cong uen wi h ac ual gene ic g oupings. Ac ual sec ion ci cumsc ip ions we e suppo ed wi h some excep ions conce ning he sec ions O eadea and Vulga a. Re icula ion analyses o scen da a eﬂec ed he hyb id s a us o in e media es, such as Hie acium anconicum,Hie acium caesium, as well as axa o Hie acium wiesbau ianum and Hie acium glaucinum. The da a also poin ed owa ds a hyb id o igin o membe s o some pu a i e non-hyb id axa such as Hie acium lachenalii, which is in acco dance wi h ecen molecula s udies. The axonomical use ulness o scen da a in dominan ly apomic ic axa is discussed. Ó2011 Else ie L d. All igh s ese ed. 1. In oduc ion In es iga ions o ﬂo al scen deal mainly wi h ques ions conce ning he in e ac ion be ween ﬂowe s and pollina o s, and many s udies we e conduc ed in o de o iden i y subs ances a ac i e o pollina o s (e.g. Plepys e al., 2002; Dö e l e al., 2006). In con as o his ﬁeld o esea ch, he e a e only ew s udies dealing wi h he use o scen da a o axonomy (Williams and Whi en, 1998; Ba kman, 2001; Le in e al., 2003). Feulne e al. (2009) demons a ed o he ﬁ s ime ha such analyses a e use ul o in es iga ing e icula e complexes as hose o Hie acium subgen. Pilosella. In his species g oup, whe e ecen hyb idiza ion occu s e y o en, inﬂo escence scen pa e ns we e ound o be axon-speciﬁc and use ul o axonomic conside a ions (Feulne e al., 2009). Fu he mo e, pa en al axa o hyb ids could be iden iﬁed on he basis o scen compounds (Feulne e al., 2009). Subgenus Hie acium is dis ibu ed in empe a e a eas o No h Ame ica, Asia and Eu ope (B äu igam, 1992), howe e , i is mos di e se in Eu ope (Zahn, 1922–1938). I is a e icula e complex, bu in con as o Hie acium subgen. Pilosella, ecen hyb idiza ion is a e and was only epo ed om Sou he n and Eas e n Eu ope so a (M áz e al., 2005; M áz and Paule, 2006; Feh e e al., 2007). In Cen al *Co esponding au ho . Tel.: þ49 921 55 2459; ax: þ49 921 55 2786. E-mail add esses: [email p o ec ed] (M. Feulne ), [email p o ec ed]wn.de (F. Schuhwe k), S e an.doe e l@uni-bay eu h.de (S. Dö e l). Con en s lis s a ailable a ScienceDi ec Biochemical Sys ema ics and Ecology jou nal homepage: www.else ie .com/loca e/biochemsyseco 0305-1978/$ –see on ma e Ó2011 Else ie L d. All igh s ese ed. doi:10.1016/j.bse.2011.06.012 Biochemical Sys ema ics and Ecology 39 (2011) 732–743 44 Au ho 's pe sonal copy Eu ope, mos axa a e iploid apomic s and seem o ha e a isen om ancien hyb idiza ion (Schuhwe k, 2002; Feh e e al., 2009). He e, we es ed whe he scen pa e ns a e use ul o axonomic conside a ions in a p esumable ancien hyb id complex using Hie acium s. s . as a model. In Middle Eu ope nea ly all Hie acium axa a e polyploid apomic s ( i- and e aploids), which p oduce seeds wi hou pollina ion (B äu igam and G eu e , 2007). In sexual plan s pollina o medi- a ed selec ion o ﬂo al scen appea s o be impo an (Salzmann e al., 2007), such p ocesses should become edundan in apomic s, whe e ﬂo al scen s ha e no unc ion in pollina o a ac ion and he e o e may be unde elaxed selec ion. The e o e i is likely ha mu a ions could lead o a loss o gain o scen compounds, excep hose scen subs ances which ha e unc ions o he s han pollina o a ac ion (e.g. epellen s agains ﬂo i o es, an i-pa hogenes). O he ﬂo al ea u es may also be unde elaxed selec ion in Hie acium, and indeed, many apomic s a e known o be male s e ile and don’ p oduce pollen a all ( S o cho á e al., 2002, own in es iga ions). Such changes may especially be e iden in axa ha a e ancien hyb ids and apomic s o many gene a ions. The e o e, i may be mo e di ﬁcul o iden i y pa en al axa in Hie acium s. s . by scen da a han in g oups such as Hie acium subgen. Pilosella whe e ecen hyb idiza ion s ill occu s (comp. Feulne e al., 2009). Taxonomy in Hie acium s. s . is complica ed. Many au ho s sepa a e he wo subgene a o Hie acium in o wo gene a (B äu igam and G eu e , 2007). The axonomic concep o “basic species”und “in e media e species”is s ill used o Hie a- cium, which acco ding o Zahn (1922–1938), helps in in e p e ing he la ge numbe o axa wi h cha ac e s in e media e among wo o mo e species. I u he assigns he high numbe o subspecies o mic ospecies o so-called “collec i e species”. Howe e , collec i e species a e only a heo e ical cons uc ion, because he mic ospecies o subspecies a e he eal axa (Schuhwe k, 2002). Taxa in es iga ed he e belong o “basic”species o Hie acium such as H. schmid ii,H. bupleu oides,H. mu o um and Hie acium biﬁdum and “in e media e” axa such as H. anconicum,Hie acium glaucinum and H. wiesbau ianum. We in es i- ga ed whe he scen pa e ns eﬂec he cu en axonomic concep and he in e p e a ion o axa as in e media e, and can he composi ion o ola iles con ibu e o unde s anding a pu a i e hyb id o igin. Fu he mo e, we ask whe he scen simila i ies a e in acco dance wi h ac ual gene ic in es iga ions (Feh e e al., 2009) and cu en sec ion delimi a ion. 2. Me hod 2.1. S udy plan s All accessions, au ho s o axa, ouche specimens and locali ies o he in es iga ed Hie acium s. s . axa a e gi en in Table 1. Taxon iden iﬁca ion ollows Zahn (1922–1938). As addi ional in o ma ion g ow h o m and hai ypes o he in oluc um (b is led, s ella e and glandula ) a e gi en in Table 1.H. wiesbau ianum axa o Thu ingia we e de e mined by Jochen Mülle , Jena. The ecen nomencla u e changes in his subgenus (G eu e , 2007) we e no applied, since no all axa used in his s udy we e published ollowing he new nomencla u e. Sec ion ci cumsc ip ion (Table 2) ollows Go schlich (2009), o axa no included in he la e s udy, we used he concep o Sell and Wes (1976). 2.2. Vola ile collec ion Inﬂo escence scen was collec ed in he ﬁeld using a s anda d dynamic headspace me hod as desc ibed in Feulne e al. (2009). Fo each axon wo o six indi iduals we e collec ed. Sampling was ca ied ou on esh and newly opened capi ula (one capi ulum pe plan and sample), be ween 11 a.m. and 3 p.m, he pe iod wi h he mos in ensi e scen emission (Feulne , unpublished da a). 2.3. Chemical analysis The samples we e analysed on a Va ian Sa u n 2000 mass spec ome e , and a Va ian 3800 gas ch oma og aph wi h a 1079 injec o , ha had been ﬁ ed wi h he Ch oma oP obe ki . This ki allows he he mal deso p ion o small amoun s o solids o liquids con ained in qua z mic o ials (Mic o-SPE; c . Ami a and Dagan, 1997). The injec o spli en was opened (1/20) o ﬂush any ai om he sys em and closed a e 2 min; he injec o was hea ed wi h 40  C o 2 min, and empe a u e was hen inc eased wi h a a e o 200  C/min o 200  C; his empe a u e was held o 4.2 min, a e which he spli en opened (1/10) and he injec o cooled down. A ZB-5 column (5% phenyl polysiloxane) was used o he analyses (60 m long, inne diame e 0.25 mm, ﬁlm hickness 0.25 m m, Phenomenex). Elec onic ﬂow con ol was used o main ain a cons an helium ca ie gas ﬂow o 1.8 ml min 1 . The GC o en empe a u e was held o 7 min a 40  C, hen inc eased by 6  C pe min o 250  C and held o 1 min. The MS in e ace was 260  C and he ion ap wo ked a 175  C. The mass spec a a e aken a 70 eV (in EI mode) wi h a scanning speed o 1 scan s 1 om m/z30 o 350. M. Feulne e al. / Biochemical Sys ema ics and Ecology 39 (2011) 732–743 733 45 Au ho 's pe sonal copy Table 1 Vouche and locali y in o ma ion o plan ma e ial used in his s udy. All accessions o igina e om Ge many. Addi ional in o ma ion abou g ow h o m and hai co e o he in oluc um is gi en. Taxon Species code Vouche Numbe o scen samples pe popula ion G ow h o m and ﬂowe ing ime Hai ypes o in oluc um H. bupleu oides C.C. Gmel. ssp. bupleu oides bb Ba a ia, Gößweins ein, 49  46 0 N, 11  20 0 E, 30.7.2008, Feulne 95 (UBT) 2 all-g owing and la e ﬂowe ing s ella e, sligh ly b is led S ei be g, 49  48 0 N, 11  11 0 E, 30.7.2008, Feulne 40 (UBT) 2 H. glaucum ssp. isa icum (Nägeli ex J. Ho m.) Nägeli e Pe e gli Ba a ia, Ruhpolding, Fischbach al, 47  42 0 N, 12  39 0 E, 10.8.2007, Feulne 41 (UBT) 3 all-g owing and la e ﬂowe ing s ella e H. anconicum (G iseb.) Zahn w Ba a ia, Fo chheim, Walbe la, 49  42 0 N, 11  09 0 E, 30.7.2008, Feulne 101 (UBT) 3 all-g owing and la e ﬂowe ing s ella e, b is led and glandula h Egglo s ein, Ha d , 49  42 0 N, 11  15 0 E, 30.7.2008, Feulne 42 (UBT), 2 b Beu en, 48  32 0 N, 9  15 0 E, 20.7.2008, Feulne 43 (UBT) 3 H. ha zianum Zahn h Ba a ia, Fo chheim, Walbe la, 49  42 0 N, 11  09 0 E, 30.7.2008, Feulne 43 (UBT) 3 all-g owing and la e ﬂowe ing s ella e, b is led and glandula H. schmid ii ssp. kalmu inum (Zahn) Go schlich sk Ba a ia, Ka ls ad , Kalbens ein, 49  58 0 N, 9  46 0 E, 29.5.2008, Schuhwe k 86/181 (M). 3 low-g owing and ea ly ﬂowe ing s ella e, b is led and spa sely glandula H. schmid ii ssp. coma ulum (Jo d. ex Bo eau) Go schlich sc Ba a ia, Naila, Höllen al, 50  20 0 N, 11  41 0 E, 21.6.2007, Feulne 45 (UBT) 4 low-g owing and ea ly ﬂowe ing sligh ly s ella e, b is led and glandula H. glaucinum ssp. medium (Jo d.) O. Bolòs & Vigo gm Ba a ia, Ka ls ad , Kalbens ein, 49  58 0 N, 9  46 0 E, 29.5.2008, Feulne 46 (UBT) 2 low-g owing and ea ly ﬂowe ing glandula H. glaucinum ssp. simila um (Jo d. ex Bo eau) Zahn gs Ba a ia, Kne zgau, 49  59 0 N, 10  37 0 E, 29.5.2008, Feulne 100 (UBT) 1 low-g owing and ea ly ﬂowe ing b is led and glandula Ü ingen, S einb uch, 49  46 0 N, 9  43 0 E, 29.5.2008, Feulne 47 (UBT) 1 H. glaucinum ssp. oegocladum (Jo d. ex Bo eau) Soó go Ba a ia, Schmidmühlen, Aichae Be g 49  16 0 N, 11  56 0 E, 5. 6. 2009, Feulne 72 2 low-g owing and ea ly ﬂowe ing glandula H. glaucinum Jo d. gpo Ba a ia, Po ens ein, Felsenbad 49  46 0 N, 11  24 0 E, Feulne 48 (UBT) 2 low-g owing and ea ly ﬂowe ing b is led and glandula H. glaucinum g ex cine ascens (Jo d.) Zahn gci dol Ba a ia, Döhlau, 49  57 0 N, 11  39 0 E, 20.5.2008, Feulne 49 (UBT) 2 low-g owing and ea ly ﬂowe ing glandula H. glaucinum g ex cine ascens (Jo d.) Zahn gci mun Ba a ia, Münne s ad , 50  15 0 N, 10  11 0 E, 15.5.2008, Feulne 50 (UBT) 2 low-g owing and ea ly ﬂowe ing glandula Hie acium glaucinum ssp. p asiophaeum (A .-Tou . e Gau ie ) G eu e gp Ba a ia, Fo chheim, Eh enbü g, 49  42 0 N, 11  09 0 E, Feulne 73 (UBT) 3 low-g owing and ea ly ﬂowe ing spa sely b is led and glandula Weismain, Neudo , 50  34 0 N, 11  15 0 E, Feulne 77 (UBT) 3 H. onosmoides F . o Ba a ia, Ka ls ad , Kalbens ein, 49  58 0 N, 9  46 0 E, 15.5.2008, Feulne 51 (UBT) 2 all-g owing and la e ﬂowe ing b is led, spa sely glandula and s ella e H. wiesbau ianum ssp. semicine ascens Bo nm. e Zahn ws Ba a ia, Münne s ad , 50  15 0 N, 10  11 0 E, 29.5.2008, Feulne 52 (UBT) 2 low-g owing and ea ly ﬂowe ing s ella e, b is led and glandula Hie acium [wiesbau ianum] pa imacula um Jochen Müll. wp Thu ingia, Jena, Haselbe g, 50  55 0 N, 11  32 0 E, 5.6.2008, Feulne 53 (UBT) 2 low-g owing and ea ly ﬂowe ing s ella e, b is led and glandula H. [wiesbau ianum ssp. jenzigense a .] euwiesbau iani o me (Schack e Zahn) Jochen Müll. we Thu ingia, Jena, Haselbe g, 50  55 0 N, 11  32 0 E, 5.6.2008, Feulne 54 (UBT) 2 low-g owing and ea ly ﬂowe ing s ella e, b is led and glandula M. Feulne e al. / Biochemical Sys ema ics and Ecology 39 (2011) 732–743734 46 Au ho 's pe sonal copy Table 1 (con inued ) Taxon Species code Vouche Numbe o scen samples pe popula ion G ow h o m and ﬂowe ing ime Hai ypes o in oluc um H. wiesbau ianum ssp. jenzigense Bo nm. e Zahn wj Thu ingia, Jena, Haselbe g, 50  55 0 N, 11  32 0 E, 5.6.2008, Feulne 55 (UBT) 2 low-g owing and ea ly ﬂowe ing s ella e, b is led and glandula H. wiesbau ianum ssp. niphan hodes Bo nm. e Zahn wn Thu ingia, Schi newi z, 50  51 0 N, 11  32 0 E, 5.6.2008, Feulne 56 (UBT) 2 low-g owing and ea ly ﬂowe ing s ella e, b is led and glandula H. [somme el ii] c inicaesium (Schack e Zahn) Jochen Müll. wc Thu ingia, Schi newi z, 50  51 0 N, 11  32 0 E, 5.6.2008, Feulne 57 (UBT) 2 low-g owing and ea ly ﬂowe ing s ella e, b is led and glandula H. wiesbau ianum ssp. ape o um Bo nm. e Schack ex Zahn wap Thu ingia, Leu a al, 50  52 0 N, 11  34 0 E, 5.6.2008, Mülle 9794 2 low-g owing and ea ly ﬂowe ing s ella e, b is led and glandula H. wiesbau ianum ssp. a noldianum Zahn wa Ba a ia, Eichs ä , S einb uch, 48  54 0 N, 11  10 0 E, 19.7.2008, in cul . Meye , 20.7.2008 1 low-g owing and ea ly ﬂowe ing s ella e, b is led and glandula Al do Ti ing, 48  59 0 N, 11  17 0 E, in cul . Meye , 20.7.2008 1 Mühlheim Gailach al, 48  51 0 N, 10  59 0 E, in cul . Meye , 20.7.2008 1 H. wiesbau ianum s l. Uech . ex Baeni z wsch Ba a ia, A nsbe g, Schambach al, 48  54 0 N, 11  22 0 E, 5.6.2009, in cul . Meye 2 low-g owing and ea ly ﬂowe ing s ella e, b is led and glandula Hie acium biﬁdum ssp.s enolepis a . aldeﬂoccosum (Vollm.) Zahn bs Ba a ia, Hei zenho en, Öde G ainbe g, 49  08 0 N, 11  56 0 E, 5.6.2009, Feulne 74 (UBT) 1 low-g owing and ea ly ﬂowe ing s ella e, b is led and spo adic glandula Deue ling, 49  02 0 ,11  54 0 , 5.6.09, Schuhwe k 09/100 2 H. biﬁdum Ki . ex Ho nem.g ex biﬁdum bb Ba a ia, Ruhpolding, Fischbach al, 47  41 0 N, 12  39 0 E, 10.8.2007, Feulne 61 (UBT) 2 low-g owing and ea ly ﬂowe ing s ella e, b is led H. biﬁdum Ki . ex Ho nem.g ex biﬁdum bbk Ba a ia, K ögels ein, 49  58 0 N, 11  16 0 E, 6.6.2007, Feulne 62 (UBT) 2 low-g owing and ea ly ﬂowe ing s ella e, b is led H. biﬁdum ssp. basicunea um Zahn bsc Ba a ia, Fo chheim, Eh enbü g, 49  42 0 N, 11  09 0 E, Feulne 63 (UBT) 3 low-g owing and ea ly ﬂowe ing s ella e, b is led and glandula H. mu o um ssp. sil ula um (Jo d.) Zahn msi Ba a ia, Be zens ein, Bu g uine 49  41 0 N, 11  25 0 E, 20.6.2008, Feulne 64 (UBT) 2 low-g owing and ea ly ﬂowe ing glandula H. mu o um L. mk Ba a ia, Kup e be g, 50  08 0 N, 11  35 0 E, 15.6.2007, Feulne 65 (UBT) 2 low-g owing and ea ly ﬂowe ing glandula H. mu o um L. mw Ba a ia, Fo chheim, Eh enbü g, 49  42 0 N, 11  09 0 E, Feulne 80 (UBT) 2 low-g owing and ea ly ﬂowe ing glandula H. mu o um L. mne Ba a ia, Neus äd lein, Ho lache, 49  58 0 N, 11  25 0 E, 10.6.2008, Feulne 66 (UBT) 2 low-g owing and ea ly ﬂowe ing glandula H. mu o um L. mb Ba a ia, Neudo , Bä en al, 50  34 0 N, 11  15 0 E, 20.6.2006, Feulne 90 (UBT) 1 low-g owing and ea ly ﬂowe ing glandula H. saxi agum F . ssp. du ii Zahn sd Ba a ia, S einb uch Gu enbe g, 50  09 0 N, 11  34 0 E, 15.6.2007, Feulne 67 (UBT) 2 all-g owing and la e ﬂowe ing sligh ly b is led and glandula Hie acium caesium F . ca Ba a ia, G ä enbe g, Almos 49  40 0 N, 11  21 0 E, 20.6.2008, 15. 6. 1996, Wagenknech s. n.(M) 2 all-g owing and la e ﬂowe ing s ella e, b is led, sligh ly glandula Hie acium caesium F . ssp. caesium cc Ba a ia, Essing, 48  56 0 N, 11  47 0 E, 20.8.2008, Feulne 68 (UBT) 2 all-g owing and la e ﬂowe ing s ella e, b is led (con inued on nex page) M. Feulne e al. / Biochemical Sys ema ics and Ecology 39 (2011) 732–743 735 47 Au ho 's pe sonal copy 2.4. Da a analysis The GC–MS da a we e p ocessed using he Sa u n So wa e package 5.2.1. Componen iden iﬁca ion was ca ied ou using he NIST 08 mass spec al da a base, o MassFinde 3, and conﬁ med by compa ison o e en ion imes wi h published da a (Adams, 2007). Iden iﬁca ion o indi idual componen s was conﬁ med by compa ison o bo h mass spec um and GC e en ion da a wi h hose o au hen ic s anda ds. 2.5. S a is ical analysis A simila i y ma ix (Sö ensen simila i ies) was cons uc ed based on he p esence/absence o compounds. The signiﬁcance o di e ences in scen p oﬁles among axa was assessed by ANOSIM wi h 10,000 andom pe mu a ions using P ime (Cla ke and Go ley, 2001). Fo u he analyses, he samples o a speciﬁcspecies(2–6 indi iduals pe axon) we e me ged. A single subs ance was ea ed as p esen in a axon i i occu ed in a leas hal o he indi idual samples. To analyse he ela ionships among he axa we conduc ed a e icula ion ne wo k analysis wi h he p og am - ex, e sion 4.0a1 (Maka enko , 2001). This me hod allows o isualize ela ionships o species in e connec ed wi h mo e han one ances o (Legend e and Maka enko , 2002), which is impo an o analysing g oups, such as Hie acium, wi h many axa o hyb id o igin (see Feulne e al., 2009). In his app oach, a neighbou joining ee was cons uc ed using a dissimila i y ma ix (1-Jacca d), and homoplasies a e made isible by so-called e icula ion lines. Those homoplasies poin owa ds hyb idiza ion o in og ession (Legend e and Maka enko , 2002). In addi ion o he p esence and absence o compounds, we also calcula ed he a e age ela i e (pe cen age o o al) amoun o scen compounds o he single axa (see Table A.1). Table 2 Sec ions o Hie acium s. s ., acco ding o Go schlich (2009) and Sell and Wes (1976) o he in es iga ed species. Basic species p in ed in bold. Fo in e media e species he o mula o he axa o which hey a e mo phologically in e media e (comp. Zahn,1922–1938) a e gi en in b acke s. Sec ion axa Hie acium sensu Go schlich (2009) H. mu o um O eadea (F .) A .-Tou . sensu Go schlich (2009) H. schmid ii H. glaucinum (schmid ii-mu o um) O eadea (F .) A .-Tou . sensu Sell and Wes (1976) H. saxi agum (lachenalii >schmid ii) H. onosmoides (lachenalii <schmid ii) Biﬁda (A .-Tou .) Clapham sensu Go schlich (2009) H. biﬁdum H. wiesbau ianum (biﬁdum-schmid ii/glaucinum) Vulga a (G iseb.) Willk. & Lange sensu Sell and Wes (1976) H. lachenalii H. caesium (lachenalii-biﬁdum) D epanoidea Monnie sensu Sell and Wes (1976) H. glaucum H. bupleu oides H. anconicum (bupleu oides-mu o um) H. ha zianum (lae iga um- anconicum) Hie acioides Dumo . sensu Sell and Wes (1976) H. umbella um T iden a a (F .) A .-Tou . sensu Sell and Wes (1976) H. lae iga um Table 1 (con inued ) Taxon Species code Vouche Numbe o scen samples pe popula ion G ow h o m and ﬂowe ing ime Hai ypes o in oluc um (con inued on nex page) Hie acium lachenalii C. C. Gmel. l Ba a ia, Neus äd lein, Ho lache 49  58 0 N, 11  25 0 E, 20.6.2007, Feulne 69 (UBT) 2 all-g owing and la e ﬂowe ing glandula Mis elbach, Buchs ein, 49  55 0 N, 11  32 0 E, 12.6.08, Feulne 70 (UBT) 2 Naila Höllen al 50  20 0 N, 11  41 0 E, 9.6.08, Feulne 71 (UBT) 2 Hie acium lae iga um Willd. lae Ba a ia, Neus äd lein, Ho lache, 49  58 0 N, 11  25 0 E, 14.7.2008, Feulne 72 (UBT) 2 all-g owing and la e ﬂowe ing sligh ly sho glandula Rehaue Fo s , 50  16 0 N, 12  40 0 E, Feulne 73 (UBT) 1 Hie acium umbella um L. u Ba a ia, Neus äd lein, Ho lache 49  58 0 N, 11  25 0 E, 20.8.2007, Feulne 78 (UBT) 2 all-g owing and la e ﬂowe ing glab ous M. Feulne e al. / Biochemical Sys ema ics and Ecology 39 (2011) 732–743736 48 Au ho 's pe sonal copy 3. Resul s 3.1. Scen compounds and pa e ns A o al o 58 di e en inﬂo escence scen compounds we e p esen in he Hie acium axa (Table A.1). The compounds belong o benzenoids, sesqui e penes, homo e penes, mono e penes and a y acid de i a i es. Mos nume ous subs ances we e sesqui e penes wi h nea ly hal o he numbe o iden iﬁed subs ances. The e we e ema kable di e ences among he axa conce ning he numbe o scen componen s, anging om 22 componen s in H. glaucinum ssp. p asiophaeum o 40 in Hie acium umbella um. The scen pa e ns (p esence/absence o compounds) we e highly axa speciﬁc (ANOSIM: R¼0.86; p¼0.01), allowing he iden iﬁca ion o axa by scen da a. Low-g owing axa wi h ze o o one s em lea (excep H. lachenalii, H. onosmoides and H. caesium) and an ea lie lowe ing ime Tall-g owing axa wi h wo o many s em lea es and a la e lowe ing ime schmid ii kalmu inum glaucinum cine ascens gci mun wiesbau ianum semicine ascens wiesbau ianum ape o um glaucinum simila um wiesbau ianum jenzigense c inicaesium wiesb. s. l. glaucinum cine ascens gci dol glaucinum oegocladum onosmoides glaucinum medium wiesbau ianum niphan hodes wiesbau ianum a noldianum bi idum s enolepis alde loccosum glaucinum gpo wiesbau ianum wsch mu o um sil ula um mu o um mk mu o um mb lachenalii glaucinum p asiophaeum mu o um mw mu o um mne pa imacula um wiesb. s. l. euwiesbau iani o me wiesb. s. l. schmid ii coma ulum bi idum bi idum bb bi idum bi idum bbk caesium ca bi idum basicunea um caesium caesium saxi agum du ii umbella um ha zianum lae iga um anconicum w anconicum h anconicum b glaucum isa icum bupleu oides Fig. 1. Neighbou joining ee (1-Jacca d) wi h e icula ions added o he basic addi i e ee limi ed o 17, numbe o objec s n¼40 (by - ex V. 4.1, Maka enko , 2001). Basic species a e p in ed in bold. M. Feulne e al. / Biochemical Sys ema ics and Ecology 39 (2011) 732–743 737 49 Au ho 's pe sonal copy The mos abundan and commonly occu ing compounds we e (Z)-3-hexen-1-yl ace a e (a e age ela i e amoun 27%), (Z)-3-hexen-1-ol (19%), me hyl salicyla e (10%), (E)-4,8-dime hyl-1,3,7-nona iene (6%), (E)- b -ocimene (4.6%), a -copaene (4.4%), 2-phenyle hyl alcohol (3.4%), linalool, limonene, and phenylace aldehyde (Table A.1). 3.2. G ouping o axa in he neighbou joining ee on he basis o scen simila i y In he neighbou joining ee (Fig. 1) he in es iga ed 37 axa clus e ed in ou main- and se e al subg oups, based on he p esence and absence o compounds. O e all, scen da a we e –despi e some excep ions (see below) –cong uen wi h he g ow h o m and ﬂowe ing ime ( all g ow h, many s em lea es and la e ﬂowe ing ime e sus low g ow h, ze o o one s em lea and ea lie ﬂowe ing ime, see Table 1) o he axa s udied. We did no ﬁnd exclusi e compounds o he all-g owing o o he low-g owing axa g oup as a whole. Howe e , (Z)-Linalooloxid u anoid occu ed in mos o he all-g owing axa and only in a ew o he o he mo pho ypes. p-Me hylanisole occu ed only in he all-g owing axa o sec ion D epanoidea (H. anconicum, H. bupleu oides and H. glaucum ssp. isa icum; comp. Table A.1). In he majo i y o he low-g owing axa e pi- nolene was p esen , bu missing in mos o he all-g owing mo pho ypes. The all-g owing g oup comp ised membe s o ou sec ions, among hem D epanoidea (e.g., H. bupleu oides,H. anco- nicum), T iden a a and Hie acioides (see Fig. 1,Table 2). A well seg ega ed second g oup was buil by low-g owing axa (e.g., H. biﬁdum,H. schmid ii, H. mu o um,H. glaucinum and H. wiesbau ianum), inwhich membe s o H. biﬁdum and Hie acium caesium we e sis e o he emaining axa. Howe e , his second g oup also con ains a ew high-g owing mo pho ypes, i.e., Hieacium lachenalii, H. caesium,andH. onosmoides (see Fig. 1). H. lachenalii clus e ed wi h H. mu o um, whe eas he pu a i e de i a i es o H. lachenalii (i.e., H. onosmoides,H. caesium) clus e ed wi h H. glaucinum and H. biﬁdum, espec i ely. 3.3. Iden iﬁca ion o scen homoplasies and e icula e ela ionships o axa Re icula ions linked in mos cases mo phologically in e media e hyb id axa wi h hei p obable pa en s e en i hey did no buil a g oup in he neighbou joining ee. As an example, H. lachenalii g oups oge he wi h H. mu o um, bu is connec ed by a e icula ion line o Hie acium lae iga um. Indeed, H. lachenalii sha es i s in oluc um-indumen (dense glandula hai s) wi h H. mu o um while i s alle g ow h o m esembles H. lae iga um (see Table 1). H. caesium ca (see Table 1) clus e ed wi h H. biﬁdum bu a e icula ion line linked i o H. lachenalii, indica ing scen homoplasies in acco dance wi h i s mo phological in e media y be ween H. biﬁdum and H. lachenalii. I possesses s ella e hai s on he in oluc um like H. biﬁdum, bu wo s em lea es, as i is o en ound also in H. lachenalii (Table 1). The pu a i e o igin o he endemic species H. anconicum by hyb idiza ion be ween H. bupleu oides and H. mu o um (see Table 2) was pa ly conﬁ med by ou analysis, because a e ic- ula ion line linked he g oup wi h one axon o H. mu o um bu addi ionally also wi h some membe s o he H. biﬁdum/H. caesium g oup. Taxa o a speciﬁc sec ion (e.g. Biﬁda, O eadea o Hie acia) did no clus e oge he in all cases, ins ead, sec ional subg oups occu ed in di e en places o he ee (Fig. 1). Howe e , i e icula ions be ween axa (e.g. be ween H.wiesbau ianum and H. biﬁdum o H. glaucinum and H. mu o um) a e aken in o conside a ion, scen esul s ﬁ be e wi h he sec ion concep acco ding o Sell and Wes (1976) and Go schlich (2009). 4. Discussion 4.1. Subgenus speciﬁc componen s In Hie acium subgen. Hie acium many componen s we e iden iﬁed which we e al eady ound in ou p e ious s udy o subgenus Pilosella (Feulne e al., 2009). Subs ances such as (Z)-3-hexen-1-yl ace a e o (Z)-3-hexen-1-ol domina e he scen in bo h subgene a (con . Feulne e al., 2009), mono e penes, such as enchone could be ound only in subgenus Pilosella (see Feulne e al., 2009). Subs ances ha could be ound so a only in subgenus Hie acium a e among o he s p-me hylanisole, ca one, py oids and u anoids o linalool oxid and g - e pinene. Some subs ances a e much mo e widesp ead in Hie acium s. s . han in H. subgen. Pilosella (e.g. benzeneace aldehyde, linalool). 4.2. Con o mi y be ween scen g ouping and cu en axonomy Membe s o he sec ion D epanoidea clus e closely oge he (Fig. 1) which is in s ong acco dance wi h cu en sec ional classiﬁca ion (c . Zahn, 1922–1938, S ace, 1998; Go schlich, 2009) as well as molecula in es iga ions (Feh e e al., 2009). Only he in e media e H. ha zianum clus e ed close o i s pu a i e second pa en al axon H. lae iga um. The close g ouping o he low-g owing axa, including pu a i e de i a i es o H. lachenalii, is in acco dance wi h high mo phological and gene ic simila i y among hese ypes (Feh e e al., 2009). Fu he mo e, he weak co ela ion o scen g oups and sec ional g oups wi hin he low-g owing axa poin s owa ds a complex and p esumably polyphyle ic e olu ion. Ne e heless, wi hin his g oup, scen da a e ealed a clea seg ega ion o H. mu o um agains membe s o H. biﬁdum. The e o e, ou da a conﬁ m he ac ual axonomical ea men o Go schlich (2009), assigning H. biﬁdum and H. mu o um o di e en sec ions, and no combining i in one, as p oposed by Sell and Wes (1976). Scen da a conﬁ m he esul o Feh e M. Feulne e al. / Biochemical Sys ema ics and Ecology 39 (2011) 732–743738 50 Au ho 's pe sonal copy e al. (2009), sugges ing ha H. lachenalii is no a ue “basic”species bu a hyb id be ween a membe o he “Wes e n clade” (H. mu o um,H. biﬁdum, e c.) and he H. umbella um g oup (H. umbella um,H. lae iga um and o he s). Simila ly, ou da a poin owa ds a hyb id o igin o H. lachenalii wi h pa icipa ion o H. mu o um and –ﬁ ing wi h he genomic da a o Feh e e al. (2009) - he H. umbella um/H. lae iga um g oup (Fig.1). F om scen da a, H. lachenalii could no e ie ed as pa en o axa like H. onosmoides, H. saxi agum ssp. du ii o H. caesium ssp. caesium, which had p e iously been conside ed as in e media es o H. lachenalii.H. saxi agum ssp. du ii acco ding o scen da a has a close ela ionship wi h H. umbella um/H. lae iga um, which is suppo ed by i s ela i ely poo -hai y in oluc um co e , co esponding mo e o H. lae iga um han o H. lachenalii. O he examples o he con o mi y be ween scen -g ouping and molecula sequence da a a e he icini y o H. bupleu - oides/H. glaucum and H. umbella um/H. lae iga um, all belonging o he “Eas e n clade”sensu Feh e e al. (2009) as well as he close g ouping o H. schmid ii, H. mu o um and H. biﬁdum, which all a e membe s o he “Wes e n clade”(Feh e e al., 2009). O e all, he con o mi ies o ou esul s wi h esul s based on gene ic analyses gi e s ong suppo o he axonomic eliabili y o scen da a in Hie acium. The dis an placemen o H. schmid ii ssp. coma ulum and ssp. kalmu inum indica es di e gen e olu ion o bo h subspecies (see Fig. 1), in acco dance wi h hei conside ably de ia ing mo phology. H. schmid ii ssp. kalmu inum, which clus e s oge he wi h H. wiesbau ianum and H. glaucinum axa, has indeed mo e common mo phological ea u es wi h H. wiesbau ianum axa (e.g. p esence o b is led hai s on he uppe lea side and densely s ella e hai s on he in oluc um) han wi h H. schmid ii ssp. coma ulum (e.g. glab ous uppe lea side and sligh ly s ella e hai s on he in oluc um), making his clus e ing eliable. Fu he mo e, he wo H. schmid ii subspecies also beha e s ongly di e en acco ding hei ecology: H. schmid ii ssp. kalmu inum g ows on calca eous and H. schmid ii ssp. coma ulum only on acidic soil (e. g. Schuhwe k, 1990). As a consequence H. schmid ii ssp. kalmu inum will be keyed ou as H. wiesbau ianum by B äu igam and Schuhwe k (in p ess). All in es iga ed H. glaucinum axa which a e mo phologically ei he close o H. mu o um (e.g. ssp. oegocladum, ssp. medium)o oH. schmid ii (ssp. p asiophaeum, ssp. simila um) g oup nex o H. mu o um and H. schmid ii ssp. coma ulum. The e o e, scen clus e ing does no a ou any sec ional delimi a ion, nei he ha o Sell and Wes (1976), uni ing H. mu o um and H. glaucinum in sec ion Hie acium, no ha o Go schlich (2009),a ﬁlia ing i o sec . O eadea (Table 2). H. glaucinum and mos H. wiesbau ianum axa a e close neighbou s o clus e pa ly in e mingled. This migh be due o hei common pa en , H. schmid ii (Table 2). Some au ho s (e.g. Zahn,1922–1938), howe e , discuss H. glaucinum ins ead o H. schmid ii as pa en al axon o H. wiesbau ianum. Bo h ideas a e suppo ed by ou da a, since some o he H. wiesbau ianum axa (e.g. H. pa imacula um) neighbou H. schmid ii ssp. coma ulum, o he s (e.g., H. wiesbau ianum ssp. semicine ascens, ssp. niphan hodes)H. glaucinum. One ob ious example o he la e scena io is H. wiesbau ianum ssp. semicine ascens which neighbou s H. glaucinum ssp. cine ascens mun (Fig. 1). Bo h axa sha e unique mo phological ea u es, such as he mul i- headed inﬂo escences and he s em lea mo phology. The o he pu a i e pa en o H. wiesbau ianum is H. biﬁdum (Table 2). Wi h excep ion o H. biﬁdum ssp. s enolepis a . aldeﬂoccosum (bs ), bo h axa g oups a e well sepa a ed by scen : H. wiesbau ianum axa and bs emi (E)- b -ocimene, which is absen om he o he in es iga ed H. biﬁdum axa, whe eas benzeneace aldehyde is only emi ed by H. biﬁdum ssp.biﬁdum and ssp. basicuneaum (Table A.1). In e es ingly, also in AFLP analyses (Reisch/Meye , ined.) H. biﬁdum ssp. s enolepis a . aldeﬂoccosum g oups close o axa o H. wiesbau ianum han o axa o H. biﬁdum. In conclusion, ou da a conﬁ m he hyb id o igin o mo phologically in e media e axa such as H. glaucinum, H. wiesbau ianum, H. anconicum and also sugges a hyb id o igin o H. lachenalii so a conside ed as ”basic” axon. One common assump ion is ha many Hie acium s. s . axa o igina ed du ing and sho ly a e he las ice age, a ime when hyb idiza ion be ween sexual “basic”species appea ed o be equen (Zahn, 1922–1938, Feh e e al., 2009). Mos o he species s udied a e apomic s in which inﬂo escence/ﬂowe scen does no longe ha e a unc ion as pollina o a ac an . The e o e, scen p oﬁles may be aced wi h mu a ional chances ha do no unde lay conse ing pollina o media ed selec ion (see chap e 1). Howe e , scen mu a ions did no seem o occu in such a es ha hey conceal he acks o phylogeny. O he wise i would no ha e been possible o iden i y in e media es by scen o o ﬁnd a co ela ion wi h axonomy. The e o e, scen p oﬁles as well as o he ﬂo al ea u es (e.g. inﬂo escence mo phology) a e a he conse ed in apomic ic Hie acium. This may eﬂec he ela i ely young age o many apomic ic axa. Also Feh e e al. (2009) concluded ha , because o an ex emely low le el o ITS a ia ion, mos species e ol ed in he Qua e na y, a ime o apid specia ion in Hie acium s. s . Ano he eason o scen conse a ion may be ha a la ge ange o subs ances may ha e unc ions o he s han pollina o a ac ion (e.g. epellen s agains ﬂo i o es, an i-pa hogenes; comp. Piche sky and Ge shenzon, 2002). Ne e heless, he e a e some a e epo s abou ecen gene ﬂow in pu a i e apomic ic Hie acium (see Chapman e al., 2004; Tyle and Jönsson, 2009), which can only be media ed by insec pollina o s. In such cases, pollina o media ed selec ion o inﬂo escence scen may s ill be o some impo ance. Acknowledgemen We hank P o . Sig id Liede-Schumann o suppo ing his s udy. We hank D . Jochen Mülle o de e mining he H. wiesbau ianum axa o Thu ingia. No be Meye and D . Al ed Bolze helped iden i ying axa in he ﬁeld. D . Ul ich Me e ga e aluable commen s on he manusc ip . M. Feulne e al. / Biochemical Sys ema ics and Ecology 39 (2011) 732–743 739 51 da a con i med he hyb id o igin o he S. la i olia axa, hey we e 1.3 o 1.5 imes mo e closely ela ed o S. a ia agg. han o S. o minalis. The S. a ia agg. showed a complica e gene ic s uc u e and ell in o ou main g oups, wo in e media e g oups besides S. pannonica and S. a ia s.s . Some p ogeny o S. pannonica was mo e a iable han expec ed and clus e ed pa ly wi h o he g oups indica ing gene low wi hin S. a ia agg. Di e en subg oups o he S. a ia agg ega e may be pa en al o he S. la i olia axa, con ibu ing o he ema kable gene ic dis ances be ween hem. Key wo ds So bus, No he n Ba a ia, gene ic s uc u e, pa en al species iden i ica ion, AFLP In oduc ion In Cen al Eu ope, he genus So bus L. (Rosaceae) includes se e al widesp ead, diploid (2n = 34; Lilje o s 1955, Düll 1959) and sexually ep oducing species, namely S. a ia (L.) C an z, S. o minalis (L.) C an z, and S. aucupa ia L. Spon aneously sexually ep oducing and ou -c ossing hyb ids can be ound be ween S. a ia and S. o minalis and be ween S. a ia and S. aucupa ia; howe e , hese hyb ids a e a e (comp. Aas e al. 1994, Meye e al. 2005). Besides such uns able hyb ids (i.e. S. x omen ella Gand., S. x pinna i ida (Sm.) Düll), an imp essi e numbe o s able hyb ids ha e been desc ibed as endemic apomic ic mic ospecies om many a eas o Eu ope. A high di e si y o endemic So bus mic ospecies has been epo ed especially om B i ain (Rich e al. 2010, Robe son e al. 2010), he Czech Republic (Ka pa i 1960, Lepší e al. 2009), and om pa s o sou he n Ge many, in pa icula Thu ingia and No he n Ba a ia (Düll 1961, Meye e al. 2005). Mos o hese mic ospecies show a limi ed dis ibu ion, bu con ibu e conside ably o he local species di e si y and he e o e a ac inc easing no ice o species p o ec ion e o s (Meye e al. 2005). The S. la i olia (Lam.) Pe s. agg ega e comp ises mic ospecies de i ed by hyb idiza ion be ween S. a ia agg. and S. o minalis (Düll 1961, Challice and Ko anda 1978, Aas e al. 1994). Membe s o his agg ega e p esumably o igina ed poly opically in he pos glacial pe iod (Düll 1961). The o igin o he No he n Ba a ian S. la i olia axa was no ye in es iga ed in de ail (see Meye e al. 2005). Pa en al species iden i ica ion is complica ed by he ac ha in he s udy a ea he S. a ia agg ega e (agg.) consis s o a wide ange o axonomically no su icien ly in es iga ed o ms (see below). In Ba a ia, so a 17 mic ospecies o S. la i olia agg. ha e been ecognized (c . Meye e al. 2005). He e, we ocus on h ee mic ospecies endemic o he no he n F anconian Ju a (Ge many, 58 Ba a ia), namely So bus anconica Bo nm. ex Düll, S. co digas ensis N. Mey. and So bus adeana N. Mey. These axa a e e y simila , hey di e o some ex en in lea and ui mo phology, mainly in he size and shape o he lea es, he numbe o la e al eins and he colo and shape o he ui ( o de ails see Meye e al. 2005). Mo phologically, hey esemble So bus a ia agg. mo e han So bus o minalis (Meye e al. 2005). All h ee in es iga ed S. la i olia- axa a e dis ibu ed pa apa ically in he s udy a ea (see Meye e al. 2005). The dis ibu ion a eas o S. adeana and S. co digas ensis a e si ua ed close o each o he (dis ance abou 10 km) and a e es ic ed o a ew squa e kilome e s only (Meye e al. 2005, Aas and Kohles 2011), whe eas S. anconica has a much wide ange in he F anconian Alb mo e sou hwa d and dis an om he wo o he axa. The axa egula ly g ow sympa ically wi h species om he So bus hyb ida g oup (hyb id axa be ween S. a ia and S. aucupa ia), S. o minalis o wi h membe s o S. a ia agg., such as S. pannonica Ká pá i. So bus pannonica is a xe omo phic membe o S. a ia agg., and is mo e widesp ead in he no he n F anconian Alb han S. a ia s.s . (Ku zelnigg 1995, Meye e al. 2005). I is a non- ypi ied axon, which comp ises p esumably apomic ic mo pho ypes illing he mo phological gap be ween S. a ia s.s . and S. g aeca (Spach) Loddiges ex Schaue (Ká pá i 1960, Ku zelnigg 2005, Meye e al. 2005). So bus g aeca, ano he xe omo phic membe o S. a ia agg., is mainly dis ibu ed in sou he n, sou heas e n and eas e n Eu ope. I can ep oduce sexually o acul a i ely apomic ic (Ku zelnigg 1995). I is unce ain whe he S. g aeca occu s in he s udy a ea (Düll 1961, Ku zelnigg 1995, 2005), bu indi iduals ha a e e y simila o S. g aeca ha e been ound in he no he n F anconian Alb (own obs.); ye , i is di icul o delimi a e S. g aeca agains S. pannonica. In his pape he AFLP echnique is used o in es iga e he gene ic s uc u e o S. la i olia axa and i s p obable pa en s in no he n F anconia. Ques ions add essed include (1) how wide is he gene ic dis ance be ween he S. la i olia axa; (2) do S. la i olia axa ha e a clonal s uc u e and do hey ep oduce as apomic s; (3) which membe o S. a ia agg. - besides S. o minalis - is mos likely pa en al o hese S. la i olia axa? Addi ional ch omosome coun s gi e insigh s in he cy ology o he axa. 59 Ma e ial and Me hods Plan ma e ial Fo AFLP-analyses we collec ed lea ma e ial om So bus co digas ensis (one o one locali y) and S. adeana (a one o he h ee known locali ies, comp. Meye e al. (2005)) and om wo popula ions o S. anconica. Sampled indi iduals we e chosen andomly. Lea ma e ial was also collec ed om ou o se en popula ions o pa en al species co- occu ing o coming close o he mic ospecies dis ibu ion a ea. Se en plan s o S. o minalis om di e en locali ies we e ha es ed. F om he S. a ia agg. we collec ed samples a wo popula ions o S. pannonica. Plan s om one si e (Neudo ) a e simila o S. g aeca ( oundish lea es as b oad as long, se a ion as long as b oad, comp. Düll 1961, Ku zelnigg 1995) and plan s om he locali y "Ko digas " a e a ilia ed o he ypical o m o S. pannonica e med “ ennis acke ” by local bo anis s because i has obo a e oblong lea es. Fu he mo e, we collec ed plan s om ou popula ions in he con ac a ea o S. a ia and S. pannonica in he no h-wes e n pa o he F anconian Alb. The e, popula ions included plan s ha could be clea ly a ilia ed o S. a ia s.s ., bu also plan s wi h mo phological simila i y o S. pannonica. Such in e media es we e also ound wi hin he ange o S. co digas ensis. These in e media es we e mo phologically de ia ing om he hin-lea ed S. a ia s.s . by ough lea es and a iable lea shapes anging om o a e o o bicula . F om S. pannonica hey di e ed in usually ha ing mo e eins on hei lea es. Addi ionally, ma e ial om seedlings o he in es iga ed axa we e included in o his s udy (Table 1). To his pu pose, seeds o h ee mo he ees o S. adeana, S. co digas ensis, S. anconica, S. pannonica and S. a ia-S. pannonica in e media es we e ha es ed in au umn 2009 om he same popula ions chosen o he in es iga ion o adul s. Seeds we e ge mina ed and plan s we e g own in he Ecological-Bo anical Ga dens (Uni e si y o Bay eu h, EBG). Fou seedlings om each mo he ee we e analyzed. Table 1 Taxa, si e locali y and ouche in o ma ion o he indi iduals analysed. Taxon Locali y // Gauss K üge coo dina es Indi iduals Taxon code, Vouche numbe Au obahn Rossdo // 4437013/ 5539920 5 a ia70-74 Wa endo // 4434614/ 5543670 5 a ia75-79 G a enhäusling // 4437746/ 5542273 5 a ia80-84 So bus a ia (L.) C an z Rossdo // 4437884/ 5540559 5 a ia85-89 60 Ko digas // 4443782/ 5551720 5 panK8-panK10, panK13, panK34 S. pannonica Ká pá i Neudo // 4447194/5546549 4 panK22, panK24, panK28, panK29 O sp ing S. pannonica Ko digas // 4443782/ 5551720 12 (MT 7) panO 154-157 (MT 8) panO 150-153 (MT 9) panO 146-149 S. adeana N. Mey. Neudo // 4447194/5546549 7 ade004, ade005, ade006, ade023, ade025, ade027, ade030 O sp ing S. adeana Neudo // 4447194/5546549 11 (MT 3) adeO 162- adeO 165 (MT 4) adeO 166, adeO 167, adeO 169 (MT 6) adeO 090- adeO 093 S. anconica Bo nm. ex Düll B ünnbe g // 4457310/ 5520910 2 anc067, anc069 Muggendo // 4447515/ 5518390 5 anc044, anc047, anc 051, anc053, anc064 O sp ing S. anconica Muggendo // 4447515/ 5518390 12 (MT 2) ancO 102- ancO 105 (MT 3) ancO 122- anc O 125 (MT 4) ancO 126- ancO 129 S. co digas ensis N. Mey. Ko digas // 4443782/ 5551720 8 co d003, co d011, co d014, co d017, co d018, co d020, co d031, co d039 O sp ing S. co digas ensis Ko digas // 4443782/ 5551720 12 (MT 1) co dO 110- co dO 113 (MT 2) co dO 106- co dO 109 (MT 7) co dO 130- co dO 133 Go elho // 4451879/5530456 1 o 001 Neudo // 4447194/5546549 1 o 021 Hainbach // 4451892/ 5530457 1 o 002 Muggendo // 4447515/ 5518390 2 o 042, o 046 S. o minalis (L.) C an z Ko digas // 4443782/ 5551720 2 o 012, o 033, o 043 61 Molecula me hods DNA ex ac ion Lea samples we e aken in May and June 2010. Immedia ely a e ha es ing hey we e placed in plas ic bags and pu in a box wi h ice o anspo a ion. A he same day, lea es we e washed wi h e hanol in he labo a o y and ozen in an ex ac ion ube a –80 ° C un il ex ac ion. F ozen lea samples (40 – 70 mg, 1 – 2 cm2) we e blended in 200 µl ex ac ion bu e (NucleoMag 96 Plan ki ; Mache y-Nagel, Dü en, Ge many, con aining 5 µl RNAse A) wi h a Fas P ep®-24 Tissue Homogenize (MP Biomedicals Eu ope, Illki ch, F ance) o 40s a a speed o 6m/s. Insolubles we e pelle ed a 15,000 x g o 5 min a oom empe a u e. Genomic DNA was p epa ed om he supe na an using he NucleoMag 96 Plan ki adap ed o he KingFishe au oma ed pu i ica ion sys em (The mo Scien i ic, Langenselbold, Ge many). De ails o he nucleic acid pu i ica ion p ocedu e a e p esen ed in Table A1 o he suppo ing in o ma ion. The pu i ied genomic DNA was dilu ed en old and used o all subsequen PCR eac ions. AFLP analysis AFLP analysis was conduc ed ollowing he me hod o Vos (1995) using he IRDye Fluo escen AFLP Ki o La ge Plan Genome Analysis (Li-Co -Biosciences, Bad Hombu g, Ge many). All eac ions we e conduc ed as desc ibed in he Li-Co applica ion manual, bu adap ed in he ollowing manne : a) 200 ng genomic DNA we e used o he EcoRI/MseI es ic ion diges ion, which was ex ended o 16 hou s; b) a 1:4 (ins ead a 1:10) dilu ion o he adap o -DNA liga ion mix u e was used o p eampli ica ion eac ions; c) a 1:40 dilu ion o he p eampli ica ion eac ion was used o selec i e ampli ica ion. Fo a p elimina y p ime sea ch 24 p ime combina ions we e es ed, and he ollowing six combina ions we e hen selec ed o his s udy because hey yielded he bes esul s in species di e en ia ion: M-CAA/E-ACG, M-CAC/E-ACG, M-CAC/E-ACA, M-CAT/E- ACG, M-CTC/E-ACG, M-CTT/E-ACG. Image collec ion and analysis The eac ions we e sepa a ed on a e ical elec opho esis sys em (4200 Sequence Analysis Sys em, Li-Co Biosciences, Bad Hombu g) oge he wi h DNA size ma ke s (50–700 bp Sizing S anda d, Li-Co Biosciences, bad Hombu g). AFLP banding pa e ns we e e alua ed using GeneMa ke 1-95 so wa e (So Gene ics) and a p esence-absence ma ix 62 was cons uc ed. Also, unique bands we e sco ed and we e no le ou om analysis. Band classes we e calcula ed wi h a ole ance ac o o 0.1 %. Ch omosome coun s Ch omosome numbe s we e coun ed om oo ip me is ems o one seedling, espec i ely om h ee o ou mo he ees (MT) o S. co digas ensis (MT 2,5,6,7), S. adeana (Neudo MT 3,4,6), S. anconica (MT 2,3,4, s.n.) and S. pannonica (locali y Neudo , MT 2,5, locali y Ko digas MT 8,9), g own in he EBG and ha es ed in May 2010. Some o he seedlings we e also included in he AFLP analysis (Table 1). The esh oo ips we e p e ea ed in 0.002 hyd oxychinoline (4h s), ixed in CARNOY´s solu ion and s ained in ca mine a e Snow (1963). F om he s ained oo ips we p epa ed squash p epa a ions in 45% ace ic acid, and obse ed soma ic me aphase pla es in he mic oscope. S a is ical analyses A neighbou joining (NJ) analysis o he p esence - absence ma ix was conduc ed (Nei Li dis ance), ollowed by boo s ap (BS) analysis a e in e node oo ing wi h 1000 eplica es using he p og am TREECON (Van de Pee and De Wach e 1994). The ee was oo ed wi h a S. o minalis indi idual. The axon clades o subg oups (in case o S. a ia agg.) e ealed we e used o all u he calcula ions (Fig. 1). Fo da a o S. a ia agg., we addi ionally applied model-based clus e ing (P i cha d e al. 2000) using he p og am STRUCTURE (h p://p i ch.bsd.uchicago.edu/s uc u e.h ml) in o de o e ie e he mos likely numbe o g oups wi hin he S. a ia agg ega e. Fo AFLP da a he ecessi e allele c i e ion was used and se o 1. Da a we e analysed as diploid because we did no know he exac ploidy le els o he plan s, which a ies be ween di- and e aploid. A o al o 10 independen uns wi h K se o 2-10 using he admix u e model op ion wi h co ela ed equencies (p io mean FST ¼ 0.1 equal o all popula ions) we e pe o med. The mos likely numbe o g oups is cha ac e ized by a maximum pos e io p obabili y Ln P(D) and he highes s abili y o esul s e ealed om each o en uns (comp. P i cha d e al. 2000; Falush e al. 2003, Guge li e al. 2008). A bu n-in o 50,000 s eps ollowed by 50,000 i e a ions ga e s able esul s a e es ing di e en bu n-in pe iods and i e a ions. STRUCTURE calcula es also he p opo ion o an indi idual geno ype o igina ing om each o he K g oups (= q). The indi iduals we e assigned o each o he K g oups using a h eshold o q o 0.3-0.8 o highe . 63 A P incipal Coo dina e analysis (PCo) o he da a was conduc ed wi h PRIMER (Jacca d Index) (Cla ke and Go ley 2001). In o de o in es iga e gene ic a iabili y, he numbe o polymo phic loci and Nei´s gene di e si y “NGD” (Nei 1972) we e calcula ed wi h POPGENE (Yeh and Yang 1999) se ing he p og am ou ines o a diploid, dominan ma ke da a se . I was assumed ha he NGD o p obable apomic ic axa should be clea ly lowe han he one o sexual axa (comp. Nybom and Ba ish 2000), and he NGD o seedlings o apomic s should no exceed he NGD o he adul s. As a measu e o he gene ic dis ance be ween axa we calcula ed Nei´s s anda d gene ic dis ance (Ds) using he p og am POPGENE (Yeh and Yang 1999). Fo he S. la i olia axa he p opo ions be ween he gene ic dis ances o S. a ia agg. and o S. o minalis we e calcula ed o ind ou o which pa en hey a e gene ically mo e closely ela ed. Fo he S. a ia agg. we calcula ed he dis ances o he ou subg oups e ealed in he NJ ee (see Fig. 1). We also es ed he calcula ion using mo e subg oups as indica ed by Bayesian clus e ing, howe e , he esul s did no de ia e and he e o e we no show hem he e. Vouche s udy Vouche s (la e al shoo s), ha we e simul aneously collec ed wi h he ma e ial o AFLP analyses om indi iduals o each S. a ia agg. subg oup we e deposi ed in he he ba ium UBT, and mo phologically analysed ega ding shape o b oades lea es, numbe o eins and se a ion (Fig. 3). Resul s AFLP analyses 578 ma ke s we e iden i ied om six p ime combina ions. The pe cen age o polymo phic loci o he S. la i olia axa a ied be ween 6.75 % and 22.32 % (Table 3). Wi h 52.42 % he in e media e S. a ia - pannonica g oup a . S. a ia s.s . eached he highes pe cen age o polymo phic loci. S. a ia s.s . eached 43.25 %, S. o minalis 26.99 % (Table 2). Neighbou joining- ee and Bayesian clus e ing o S. a ia agg. In he neighbou joining ee (Fig. 1) he S. la i olia axa g ouped be ween S. a ia agg. and S. o minalis, bu close o S. a ia agg. The indi iduals o S. a ia agg. and S. o minalis we e much mo e s ongly dissimila han he indi iduals o he pu a i e mic ospecies S. adeana, S. co digas ensis and S. anconica (Fig. 1). All S. la i olia accessions cons i u ed 64 highly suppo ed g oups o hei own (BS 100 o S. adeana, 84 o S. co digas ensis, and 97 o S. anconica), and we e clea ly sepa a e om each o he . The o sp ing o he S. la i olia axa clus e ed wi h he adul s and he p ogeny o a single mo he ee clus e ed oge he . The So bus a ia agg. clade was s ongly suppo ed (BS 99) and could be di ided in o ou g oups, he S. pannonica g oup, he S. a ia s.s . g oup and wo g oups o in e media es (Fig. 1). One g oup o in e media es (a ia76, a ia77, a ia79, panO 150, panO 151, panO 153, panO 156) was e ie ed mo e closely o he S. pannonica g oup and is he e o e called “in e media e g oup a ine (a .) pannonica” (Fig. 1). Ano he g oup was e ie ed mo e closely o S. a ia s.s ., i is he e o e e e ed o as “in e media e g oup a . S. a ia”. This g oup consis s o wo subclades ( he one wi h a ia074, a ia075, a ia082, a ia086 and he o he one wi h a ia070, a ia073, a ia078) (Fig. 1). The o sp ing o wo mo he ees o S. pannonica is pa ly g ouped wi hin S. pannonica and pa ly wi hin he in e media e g oup a . pannonica (Fig. 1). In he Bayesian clus e ing o he S. a ia agg. he Ln P(D) was highes and s anda d de ia ion o pos e io p obabili ies was lowes o se en g oups, (Ln P(D) o K = 7 be ween -7915 and -8085; 10 uns). The assignmen o indi iduals o se en g oups was e y cons an and he g oups we e in gene al in good acco dance wi h he g oups and subg oups e ealed by he NJ ee. Bayesian g oups we e iden ic o he NJ ee subg oups o a ia074, a ia075, a ia082 (see abo e), a ia087, a ia088, a ia089 o a ia076, a ia077, a ia079, panno 150, pannO 151, pannO 153. Indi idual a ia086 o med a g oup o i s own. 65 Fig. 1 Neighbou joining ee wi h Nei Li dis ance and boo s ap alues wi h 1000 eplica es using TREECON. O sp ing indi iduals a e shadowed da k-g ay. 66 Vouche s udy Vouche s udies o S. a ia agg. e ealed ha he in e media es a . S. a ia s.s . de ia ed om S. a ia s.s . (Fig. 3a) by a hicke lea ex u e. Thei lea shape was a iable, o a e o obo a e (Fig. 3c). The in e media es a . S. pannonica (Fig. 3d) had mo e lea eins (abou 10–11) han S. pannonica (8 eins, Fig. 3b) and he se a ion o he lea ma gins ex ended o he lea base; u he mo e, he lea shape was oundish ins ead o obo a e and he lea base was ounded and no cunea e as in S. pannonica. Nei´s gene di e si y Nei´s gene di e si y (NGD) o he axa and subg oups o S. a ia agg. is shown in Table 2. Gene di e si y alues o he S. la i olia g oup (adul s and seedlings) we e a he simila . They a ied o S. adeana be ween 0.029 and 0.083, o S. co digas ensis be ween 0.051 and 0.088, and o S. anconica be ween 0.041 and 0.088 (see Table 2). The gene ic di e si y in he S. a ia agg ega e di e s ema kably. I was high o he plan s iden i ied as S. a ia s.s . (0.129) and o plan s o he in e media e g oup wi h a ini y o S. a ia s.s . (0.177). I was lowe o he in e media e S. a ia-S. pannonica g oup wi h a ini y o S. pannonica (0.098) and e y low o S. pannonica (be ween 0.027 and 0.070). Gene ic di e si y was also ela i ely low in he second pu a i e pa en al axon S. o minalis (NGD 0.089). None o he p ogeny was ound o be iden ical – possibly a consequence o band ep oducibili y, because also unique bands we e sco ed (c . Ma e ial and Me hods). Howe e , he gene ic di e si y o he p ogeny o S. anconica and S. adeana did no exceed he alues o he adul s. NGD exceeded somewha he alues o he adul s in case o S. pannonica p ogeny o mo he ee 8, (NGD 0.070) and S. co digas ensis p ogeny o mo he ee 2 (NGD 0.088) (Table 2). 67 a) S. a ia s.s . (a ia087*) 12 eins, lea ex u e hin, lea ma gin se a ed om he base, lea base ounded. b) S. pannonica (pannN029*) 8 eins, lea ex u e hick, b oade han long se a ed, lea base cunea e, no se a ed. c) In e media e a . S. a ia (a ia075*) 11 eins, lea ex u e mode a ely hick, lea ma gin se a ed om he base, lea base cunea e. d) In e media e a . S. pannonica (a ia079*) 11 eins, lea ex u e hick, lea ma gin se a ed om he base, lea base ounded. Fig. 3 Lea shape o he b oades lea o la e al shoo s om selec ed ouche s o S. a ia agg. in es iga ed in his s udy incl. addi ional in o ma ion abou lea cha ac e s and g oup a ilia ion (* o abb e ia ions see Table 1). Acknowledgmen s The inancial suppo o his s udy by he Obe ankens i ung Bay eu h (OFS P ojec -No. 03086) is g a e ully acknowledged. We hank Michaela Hochholze (DNA Analy ics, Uni e si y o Bay eu h) o conduc ing AFLP labo a o y wo k, and F ancesco Balao o suppo ing s a is ical analysis. Ma in Lepší and No be Meye dese e hanks o de e mina ion o he ba ium ma e ial. 74 Li e a u e Aas G, Kohles M (2011) Ve b ei ung, Häu igkei und Ve jüngung on So bus co digas ensis (Ko digas -Mehlbee e) in de nö dlichen F ankenalb. Tuexenia 31:59 - 71 Aas G, Maie J, Bal isbe ge M, Me zge S (1994) Mo phology, isozyme a ia ion, cy ology, and ep oduc ion o hyb ids be ween So bus a ia (L.) C an z and S. o minalis (L.) C an z. Bo Hel 104:195–214 Challice J, Ko anda M (1978) Fla onoids as ma ke s o axonomic ela ionships in he genus So bus in Eu ope. P eslia 50:305-320 Cla ke KR, Go ley RN (2001) P ime 5: Use Manual/Tu o ial. P ime -E L d. Düll R (1959) Unse e Ebe eschen und ih e Bas a de. Wi enbe g. Ziemsen. Düll R (1961) Die So bus-A en und ih e Bas a de in Baye n und Thü ingen. Be Baye Bo Ges 34:11-65 Guge li F, B odbeck S, Holde egge R (2008) U ili y o mul ilocus geno ypes o axon assignmen in s ands o closely ela ed Eu opean whi e Oaks om Swi ze land. Ann o Bo -London 102:855-863 Ká pá i Z (1960) Die So bus-A en Unga ns und de ang enzenden Gebie e. Feddes Repe . 62:71-334 Ku zelnigg H (1995) So bus. In: Cone HJ e al (eds): Gus a Hegi, Illus ie e Flo a on Mi eleu opa 4 (2B): Spe ma ophy a: Angiospe mae: Dico yledones 2 (3). Rosaceae 2. Blackwell, Be lin, pp 328-385 Ku zelnigg H (2005) So bus L em C an z – Mehlbee e, Ebe esche, Elsbee e, Speie ling. In: Jäge E J, Ro hmale KW (eds) Exku sions lo a on Deu schland 4, Ge äßp lanzen, K i ische Band 9 h edn. Spek um, Heidelbe g, pp 403-406 Lepší M, Ví P, Lepší P, Boublík K, Kolář F (2009) So bus po ae-bohemicae and So bus albensis, wo new endemic apomic ic species ecognized based on a e ision o So bus bohemica. P eslia 8:63–89 Lilje o s A (1955) Cy ological s udies in So bus. Ac a Ho i Be giani 17:47-113 Meye N, Meie o L, Schuwe k H, Ange e O (2005) Bei äge zu Ga ung So bus in Baye n. Be Bay Bo G Sonde band:5-216 Nei M (1972) Gene ic dis ance be ween popula ions. Ame ican Na u alis 106:283–291 P i cha d JK, S ephens M, Donnelly P (2000) In e ence o popula ion s uc u e using Mul ilocus Geno ype Da a. Gene ics 155:945-959 75 Rich T, Hous on L, Robe son A (2010) Whi ebeams, Rowans and Se ice T ees o B i ain and I eland A Monog aph o B i ish and I ish So bus L. BSBI Handbook 14, London. Robe son A, Rich TCG, Allen AM, Hous on L, Robe s C, B idle JR, Ha is S A, Hiscock JS (2010) Hyb idiza ion and polyploidy as d i e s o con inuing e olu ion and specia ion in So bus. Mol Ecol 19:1675–1690 Snow R (1963) Alcoholic hyd ochlo ic acid-ca mine as s ain o ch omosomes in squash p epa a ions. S ain Technol 38:9-13 Talen N (2009) E olu ion o game ophy ic apomixis in lowe ing plan s: an al e na i e model om Maloid Rosaceae. Theo y Biosci 128:121-38 Van de Pee Y, De Wach e Y (1994) TREECON o Windows: a so wa e package o he cons uc ion and d awing o e olu iona y ees o he Mic oso Windows en i onmen . Compu Applic Biosci 10:569-70 Vos P, Hoge s R, Bleeke M (1995) AFLP: a new echnique o DNA inge p in ing. NAR 23: 4407–4414 Yeh F, Yang R (1999) POPGENE. Mic oso Window-based F eewa e o Popula ion Gene ic Analysis. Ve sion 1.31, Uni e si y o Albe a: Edmon on, Albe a, Canada 76 Suppo ing in o ma ion Table S1: Pla e layou and ins umen se ings o DNA pu i ica ion ia BindIT 3.1 KingFishe so wa e. A e s ep 7, elu ed DNA was ans e ed o plas ic cups and kep a -20°C. Se ings S ep 1 Bind MC2 S ep 2 Wash MC3 S ep 3 Wash MC4 S ep 4 Wash E hanol S ep 5 Wash MC5 S ep 6 Elu ion MC6 S ep 7 Disposal o magne ic beads Pla e A B C D E F G Layou 92 µl supe na an 8 µl magne ic beads 100 µl bu e 200 µl bu e 200 µl bu e 200 µl 80% E hanol 200 µl bu e 20 µl bu e P ecollec no no no no no no Release ime [mm:ss] no elease 00:30 00:30 00:30 no elease 00:15 00:20 Beginning Release speed - as as as - as as Pause o manual handling no no no no no no Mixing ime [mm:ss] 05:00 01:00 01:00 01:00 01:00 10:00 Mixing/ pause Mixing speed medium as as as medium medium Pos mix no no no no no no Collec coun 3 3 3 3 3 6 End Collec ime [s] 1.5 1.5 1.5 1.5 1.5 1.5 77 Publica ion 4 4. Flo al scen and i s co ela ion wi h gene ic da a in So bus axa. In p epa a ion o submission o O ganisms Di e si y & E olu ion 78 Flo al scen and i s co ela ion wi h gene ic da a in So bus axa Ma in Feulne a,*, S e an Poin ne a, Lisa Heussa, G ego Aasb, S e an Dö e l a,c aDepa men o Plan Sys ema ics, Uni e si y o Bay eu h Uni e si ä ss . 30, 95440 Bay eu h, Ge many * Au ho o co espondence: [email p o ec ed] b Ecological-Bo anical ga dens, Uni e si y o Bay eu h Uni e si ä ss aße 30, 95447 Bay eu h, Ge many email: g ego[email p o ec ed] c P esen add ess: Uni e si ä Salzbu g, FB O ganismische Biologie Hellb unne s asse 34, 5020 Salzbu g email: [email p o ec ed] Abs ac S a is ical analyses be ween lo al scen and gene ic ma ke da a o es ing hei axonomical co ela ion a e a e. We in es iga ed in lo escence scen pa e ns o apomic ic So bus la i olia mic ospecies, So bus anconica, S. adeana and S. co digas ensis endemic o no he n Ba a ia, o igina ed by hyb idiza ion and hei pa en al axa wi h dynamic headspace me hod. The scen da a (p esence-absence o compounds) we e used o cons uc an UPGMA ee, and o calcula e a simila i y ma ix o co ela e hem, bo h on indi idual as well as popula ion le el, wi h AFLP da a published in an ea lie s udy. Scen analyses showed a o al o 68 chemical subs ances, among hem a oma ic compounds, mono- and sesqui e penes, alipha ics, and ni ogen con aining compounds. Scen pa e ns we e axon speci ic, and he numbe o scen componen s di e ed among axa. Co ela ions wi h AFLP da a on popula ion and indi idual le el a e highly signi ican , indica ing ha scen and gene ic da a a e highly cong uen in he plan s s udied. Scen he e o e is a use ul ma ke o axonomical ques ions in So bus. Keywo ds Flo al scen , apomixis, AFLP, So bus, axonomy, co ela i e analysis; 79 In oduc ion S udies o sexual (i.e. Le in e al. 2003; Raguso e al. 2006), bu also o apomic ic species complexes (Feulne e al. 2009; 2011) e ealed ha axonomic conclusions based on scen can be highly cong uen wi h hose based on gene ic ma ke s. The main unc ion o lo al scen is he a ac ion o pollina o s (Dö e l e al. 2006; Plepys e al. 2002). The e o e, scen da a may be in luenced by pollina o media ed selec ion (Knudsen and Tolls en 1993; Plepys e al. 2002; Dö e l e al. 2005). One consequence o his is he e olu ion o pollina ion synd omes which means ha plan species pollina ed by he same guild o animals ha e simila pheno ypes o hei lo al cha ac e is ics including scen (Faeg i and an de Pijl 1979; Fens e e al. 2004; Dobson e al. 2005). Al hough scen is in luenced by pollina o -media ed selec ion and coe olu ion, in mos s udies i was shown ha only a limi ed numbe o subs ances ha e key unc ions in a ac ing pollina o s (Dö e l e al. 2006, S ensson e al. 2010, Bu ge e al. 2012), whe eas o he subs ances may be mo e de e mined by phylogeny han by pollina o -media ed selec ion (S eine e al. 2011; Schä le e al. 2012). The e a e some examples in which scen da a suppo ed axonomy and e ealed a good con o mi y wi h DNA da a (i.e. Le in e al. 2003; Raguso e al. 2006). Howe e , o ou bes knowledge, a s a is ical app oach o he axonomical alue o scen by de ailed co ela ions be ween scen da a and da a om gene ic ma ke s was so a only p esen ed once in a s udy dealing wi h Oph ys (O chidaceae, S ökl e al. 2008). He e, howe e , signi ican co ela ion be ween bo h da a se s was no ound (S ökl e al. 2008). Scen da a in apomic s may beha e di e en ly compa ed o sexual species, because apomic ic plan s p oduce seeds wi hou e ilisa ion and do no ely on pollina ion (Nogle 1984; Jankun and Ko anda 1987; Talen 2009). The e o e, pollina o -media ed selec ion in luencing scen pa e ns is o mino impo ance in apomic s (comp. Feulne e al. 2009; 2011). Fu he mo e, he in a-indi idual gene ic a iabili y o apomic s is ex emely low, he e o e, scen pa e ns also may be s ongly iden ical be ween indi iduals and popula ions o he same apomic ic axon. Apomixis is o en coupled wi h hyb id specia ion (Talen 2009) as i is he case in he So bus la i olia g oup (Rosaceae). So bus la i olia axa o igina ed om hyb idiza ion be ween S. a ia agg. and S. o minalis (Düll 1961; Rich e al. 2010, Feulne e al. 2013; submi ed). Among membe s o he So bus la i olia agg. a e many axa endemic o es ic ed egions in i.e. G ea B i ain, Czech Republic, and Ge many (Düll 1961; Lepší e al. 2009, Meye e al. 2005; Robe son e al. 2010; Rich e al. 2010). 80 In e es ingly, i has been shown in o he hyb id complexes such as Ci us, Oph ys and Hie acium, ha he scen consis s mainly o a mix u e o scen componen s o he pa en al species (Gancel e al. 2002; Ve eecken e al. 2010; Feulne e al. 2009; 2011), and only a low numbe o new compounds. The e o e, scen analyses may be a aluable ool o pa en al species iden i ica ion, and indeed, he axonomic eliabili y o scen pa e ns was shown o be high in g oups o igina ed by hyb idiza ion (Feulne e al. 2009; 2011). He e, we in es iga e he scen o he apomic ic mic ospecies S. adeana, S. co digas ensis, and S. anconica belonging o he S. la i olia agg ega e endemic o No he n Ba a ia and occupying e y small pa apa ic dis ibu ion a eas. In a o me s udy he hyb id s a e, he in aspeci ic a iabili y and he gene ic s uc u e o pa en al axa was in es iga ed wi h AFLP analyses (Feulne e al. 2013; submi ed). In he p esen s udy, based on he same ma e ial in es iga ed in he AFLP s udy (Feulne e al. 2013; submi ed), we in es iga ed lo al scen composi ion and co ela ed scen clus e ing wi h he AFLP ee, o es ima e he co ela ion be ween bo h da a se s. Ma e ial and me hod S udy plan s We collec ed scen om S. adeana (one o one known popula ion, comp. Meye e al. 2005), S. co digas ensis (one o app oxima ely h ee known popula ions), S. anconica ( wo o > 50 known popula ions) as well om pa en al axa such as S. a ia s.s ., S. pannonica and S. o minalis (comp. Feulne e al. 2013; submi ed). F om he S. a ia agg. also in e media es be ween S. a ia s.s . and S. pannonica wi h a ini y o S. a ia s.s . (comp. Feulne e al. 2013; submi ed) we e included in o he s udy. These plan s a e appela ed as a . a ia. Fo all axa, AFLP da a we e a ailable om he same popula ions (Feulne e al. 2013, submi ed) and in 11 cases AFLP da a and scen da a we e collec ed om he same indi iduals (Table 1, Feulne e al. 2013, submi ed). Fo u he in o ma ion abou axonomy, popula ion s uc u e and ecology o he axa in es iga ed, see Feulne e al. (2013, submi ed). 81 Table 1: Taxa, locali y and ouche in o ma ion o he indi iduals analysed. (* compa e Feulne e al. 2013, submi ed). Taxon Locali y / Gauss K uege coo dina es / Vouche numbe o indi iduals o scen sampling numbe o AFLP samples om he same popula ion/indi idual as used o scen sampling* So bus a ia (L.) C an z G a enhäusling 4437746/ 5542273, Feulne 200–207 (UBT) 2 5/0 In e media es a . a ia s.s . Au obahn Rossdo 4437013/ 5539920, Feulne 208–217 (UBT) 4 5/0 Ko digas 4443782/ 5551720, Feulne 218–227 (UBT) 4 5/1 Neudo 4447194/5546549, Feulne 228–234 (UBT) 1 4/0 S. pannonica Ká pá i B ünnbe g 4457310/ 5520910, Feulne 235 (UBT) 1 – S. adeana N. Mey. Neudo 4447194/5546549 Feulne 236 (UBT) 4 7/0 S. anconica Bo nm. ex Düll B ünnbe g 4457310/ 5520910 Feulne 237–239(UBT) 1 2/0 Mugggendo 4447515/ 5518390 Feulne 240–245 (UBT) 4 5/3 S. co digas ensis N. Mey. Ko digas 4443782/ 5551720, Feulne 246–253 (UBT) 8 7/8 Neudo Bä en al 4447194/5546549, Feulne 254 (UBT) 1 1/0 Hainbach // 4451892/ 5530457, Feulne 255 (UBT) 1 0/1 S. o minalis (L.) C an z Ko digas 4443782/ 5551720, Feulne 256–259 (UBT) 2 3/0 Vola ile collec ion In lo escence scen was collec ed in he ield using a s anda d dynamic head-space me hod as desc ibed in Feulne e al. (2009). Fo each axon, wo o six indi iduals we e sampled. 82 Sampling was ca ied ou on esh and newly opened in lo escences (one in lo escence pe plan and sample), be ween 11 a.m. and 3 p.m, he pe iod wi h he mos in ensi e scen emission (as de e mined by he human nose; Feulne , unpublished da a). Scen samples o lea es and su ounding ai we e collec ed as con ol o each locali y and popula ion in es iga ed. Chemical analysis The samples we e analysed on a Va ian Sa u n 2000 mass spec ome e , and a Va ian 3800 gas ch oma og aph wi h a 1079 injec o , ha had been i ed wi h he Ch oma oP obe ki . This ki allows he he mal deso p ion o small amoun s o solids o liquids con ained in qua z mic o ials (Mic o-SPE; c . Ami a and Dagan, 1997; Dö e l e al. 2005). The injec o spli en was opened (1/20) o lush any ai om he sys em and closed a e 2 minu es; he injec o was hea ed wi h 40 °C o 2 min, and he empe a u e was hen inc eased wi h a a e o 200 °C/min o 200 °C; his end empe a u e was held o 4.2 min, a e which he spli en opened (1/10) and he injec o cooled down. A ZB-5 column (5% phenyl polysiloxane) was used o he analyses (60 m long, inne diame e 0.25 mm, ilm hickness 0.25 µm, Phenomenex). Elec onic low con ol was used o main ain a cons an helium ca ie gas low o 1.8 ml min-1. The GC o en empe a u e was held o 7 min a 40 °C, hen inc eased by 6 °C pe min o 250 °C and held o 1 min. The MS in e ace was 260 °C and he ion ap wo ked a 175 °C. The mass spec a a e aken a 70 eV (in EI mode) wi h a scanning speed o 1 scan s-1 om m/z 30 o 350. Da a analysis The GC-MS da a we e p ocessed using he Sa u n So wa e package 5.2.1. Componen iden i ica ion was ca ied ou using he NIST 08 mass spec al da a base o MassFinde 3, and con i med by compa ison o e en ion imes wi h published da a (Adams 2007). Iden i ica ion o indi idual componen s was con i med by compa ison o bo h mass spec um and GC e en ion da a wi h hose o au hen ic s anda ds. S a is ical analysis A clus e analysis (UPGMA) based on Jacca d simila i y index (calcula ed using he p esence-absence o compounds) was cons uc ed in PRIMER Ve s. 6 (Cla ke and Go ley 2006). The axon speci icy o scen was es ed wi h ANOSIM (10.000 pemu a ions) in 83 educed in luence o pollina o selec ion his ac is leading o a s ong co ela ion be ween scen da a and gene ic da a, as has been shown abo e. Mos s udies compa ing scen and molecula da a a e based on O chidaceae. Howe e , his is a s ongly con as ing sys em including many decep i e plan s whe e pollina o selec ion plays a key ole (Salzmann e al 2007; Ve eecken e al. 2010). Decep i e plan s usually ha e an inc eased a iabili y o scen (Salzman e al. 2007), likely o a oid ha pollina o s can easily lea n o disc imina e be ween he ewa d and he mimic (Acke man e al. 2011). The e o e i is no as onishing ha in Oph ys no co ela ion o scen wi h gene ic da a was ound (S ökl e al. 2008). Table 3: P esence/absence o lo al scen ola iles, occu ing in n o all indi iduals in es iga ed (n/n) o 7 So bus axa. a ia a . a ia a ia s.s . pann- onica o mi- nalis ade- ana co digas -ensis anco- nica a oma ics benzaldehyde 4/4 2/2 6/6 4/4 4/4 8/8 5/5 benzeneace aldehyde 3/4 1/2 3/6 3/3 4/4 4/8 - me hyl benzoa e 4/4 2/2 6/6 - 4/4 8/8 5/5 2-phenyle hyl alcohol 4/4 2/2 6/6 4/4 4/4 8/8 5/5 me hyl phenylace a e 4/4 2/2 6/6 - 4/4 8/8 3/5 me hyl salicyla e 4/4 2/2 6/6 4/4 4/4 8/8 5/5 anisaldehyde 2/4 1/2 2/6 1/4 - - - alipha ics me hyl iso ale a e 4/4 2/2 5/6 - 4/4 - 2/5 2,3-bu andiol 4/4 2/2 6/6 - - - - (Z)-3-hexen-1-ol 4/4 2/2 6/6 3/4 4/4 8/8 5/5 me hyl hexanoa e - - 4/6 - 4/4 8/8 3/5 me hyl (3Z)-hex-3-enoa e - - 2/6 - 4/4 8/8 2/5 me hyl 2-hyd oxy-3-me hylpen anoa e 4/4 2/2 6/6 - 4/4 - - (Z)-3-hexen-1-ol ace a e 4/4 2/2 6/6 4/4 4/4 8/8 5/5 ace ic acid hexyl es e - - 4/6 3/4 1/4 8/8 - (E)-2-hexen-1-ol ace a e - - 4/6 - - - 5/5 oc anal 4/4 2/2 6/6 4/4 4/4 - 5/5 Homo e penes (E)-4,8-Dime hyl-1,3,7-nona iene 4/4 2/2 6/6 4/4 4/4 8/8 5/5 I egula mono e penes 4-oxoisopho one epoxide 4/4 2/2 6/6 4/4 4/4 8/8 4/5 4-oxoisopho one 4/4 2/2 6/6 4/4 4/4 8/8 5/5 dihyd ooxoisopho one 4/4 2/2 6/6 4/4 4/4 8/8 5/5 mono e penes α-pinene 4/4 2/2 6/6 4/4 4/4 8/8 5/5 camphene 3/4 2/2 6/6 2/4 4/4 8/8 - β-pinene 4/4 2/2 6/6 3/4 4/4 8/8 5/5 (Z)-ocimene 4/4 2/2 6/6 4/4 4/4 8/8 5/5 limonene 4/4 2/2 6/6 4/4 4/4 8/8 5/5 90 eucalyp ol 4/4 2/2 6/6 4/4 4/4 8/8 5/5 dihyd o-5-me hyl-5- inyl-2(3H)- u anone 4/4 2/2 5/6 1/4 4/4 7/8 - (E)-ß-ocimene 1/4 1/2 5/6 3/4 - 5/8 5/5 (Z)-a busculone 4/4 2/2 6/6 1/4 1/4 - - (E)-a busculone 4/4 2/2 6/6 1/4 - - - (Z)-linalol-oxid u anoid 3/4 1/2 4/6 3/4 3/4 - 5/5 (E)-linalol-oxid u anoid 4/4 2/2 6/6 3/4 4/4 - 1/5 lilac aldehyde A 4/4 2/2 5/6 - - - - Lilak aldehyd B+C 4/4 2/2 6/6 - 3/4 8/8 - lilac aldehyde D 4/4 2/2 6/6 - 3/4 3/8 - unkn MT1402 3/4 - 6/6 - - 1/8 4/5 lilac alcohol A 4/4 2/2 2/6 - 4/4 2/8 - lilac alcohol BC 4/4 2/2 6/6 - 4/4 4/8 - lilac alcohol D 4/4 2/2 4/6 - 4/4 2/8 2/5 isomen hone 4/4 1/2 4/6 4/4 3/4 4/8 1/5 lilac de i a i e 4/4 2/2 6/6 - 4/4 - - linalool 4/4 2/2 6/6 4/4 4/4 - 3/5 N-con aining subs ances 3-py idineca boxaldehyde 4/4 2/2 5/6 4/4 4/4 8/8 - amyl/isoamyl-py ole - - - 3/4 4/4 2/8 - phenylace oni ile 4/4 2/2 6/6 4/4 4/4 8/8 5/5 me hyl nico ina e 2/4 - - 2/4 4/4 8/8 - unk-N1364 m/z 125, 81, 39 4/4 2/2 6/6 3/4 4/4 8/8 1/5 unk-N1377 m/z 151, 94 - - - 2/4 2/4 - - Unk-N1498 m/z 117,91,65,50,39 4/4 2/2 6/6 - 3/4 8/8 2/5 unk-N1530 m/z 117,91,59,50 4/4 2/2 6/6 - 4/4 7/8 2/5 Indole 4/4 2/2 6/6 - 3/4 - - 1-ni o-2-phenyle hane 4/4 2/2 6/6 - 4/4 8/8 3/5 Sesqui e penes c . α-longipinene 2/4 - 6/6 - - - - unk-ST1684 m/z 204,161, 91, 69, 55 1/4 - 2/6 - - - - unk-ST1690 m/z 161,119, 85, 73, 58 - - 6/6 1/4 1/4 4/8 5/5 α-Copaene 3/4 - 4/6 - - 3/8 - β-bou bonene 3/4 2/2 6/6 1/4 4/4 5/8 5/5 unk-ST1711 m/z 161,123, 81, 67, 55 2/4 - 6/6 - - 4/8 - unk-ST1732 m/z 161,139, 93, 79 2/4 - - - - - - Longi olene 4/4 2/2 6/6 4/4 3/4 8/8 1/5 Isocomene 2/4 2/2 6/6 - 1/4 8/8 - (E)-β-ca yophyllene 4/4 2/2 6/6 4/4 3/4 6/8 5/5 α-gu junene 4/4 2/2 6/6 - 3/4 5/8 3/5 unk-ST1808 m/z 204,161,143,133,105 - - 2/6 1/4 1/4 4/8 1/5 unk-ST1831 m/z 204,189,161,133,119, 105 c . Ge mac ene D 3/4 1/2 6/6 1/4 2/4 6/8 4/5 unk-ST1838 m/z 161,93,41 2/4 1/2 6/6 - 1/4 5/8 5/5 uniden i ied unk m/z 112,140,181 4/4 2/2 5/6 - 4/4 - - 91 Re e ences Aas, G. & Kohles, M. (2011). Ve b ei ung, Häu igkei und Ve jüngung on So bus co digas ensis (Ko digas -Mehlbee e) in de nö dlichen F ankenalb. Tuexenia 31, 59–71. Acke man, J. D., Cue as, A. A. & Ho , D. (2011). A e decep ion-pollina ed species mo e a iable han hose o e ing a ewa d? Plan Sys ema ics and E olu ion 293: 91–99. Adams, R. P. (2007). Iden i ica ion o Essen ial Oil Componen s by Gas Ch oma og aphy/Mass Spec ome y, ou h ed. Allu ed Publishing Co po a ion, Ca ol S eam, IL. Ami a , A. & Dagan, S. (1997). A di ec sample in oduc ion de ice o mass spec ome y s udies and gas ch oma og aphy mass spec ome y analyses. Eu opean Mass Spec ome y, 105–111. Campbell, C. S. & Dickinson T. A. (1990). Apomixis, Pa e ns o Mo phological Va ia ion, and Species Concep s in sub am. Maloideae (Rosaceae). Sys ema ic Bo any 15, 124–135. Cla ke, K. R. & Go ley, R. N. (2006). P ime 6: Use Manual/Tu o ial. P ime -E L d. Dö e l, S. & Jü gens, A. (2005). Spa ial ag ance pa e ns in lowe s o Silene la i olia: Lilac compounds as ol ac o y nec a guides? Plan Sys ema ics and E olu ion 255, 99–109. Dö e l, S., Wol e, L., W. & Jü gens, A. (2005). Quali a i e and quan i a i e analyses o lowe scen in Silene la i olia. Phy ochemis y 66, 203–213. Dö e l, S., Jü gens, A., Sei e , K., Laube, T., Weißbecke , B. & Schü z, S. (2006). Nu se y pollina ion by a mo h in Silene la i olia: he ole o odou s in elici ing an ennal and beha iou al esponses. New Phy ologis 169, 707–718. Düll, R. (1959). Unse e Ebe eschen und ih e Bas a de. Wi enbe g. Ziemsen. Düll, R. (1961). Die So bus-A en und ih e Bas a de in Baye n und Thü ingen. Be ich e Baye ische Bo anische Gesellscha 34, 11–65. Feulne , M., Liede-Schumann, S., Me e, U., Weig A. & Aas, G. (2013). Gene ic s uc u e o So bus la i olia (Lam.) Pe s. axa endemic o No he n Ba a ia – submi ed o Plan sys ema ics and e olu ion (PLSY-D-12-00168). Feulne , M., Schuhwe k, F. & Dö e l, S. (2009). Flo al scen analysis in Hie acium subgenus Pilosella and i s axonomical implica ions. Flo a 204, 495–505. Feulne , M., Schuhwe k F. & Dö e l S. (2011). Taxonomical alue o in lo escence scen in Hie acium s. s . Biochemical Sys eme ic and E olu ion 39, 732–743. 92 Gancel, A. L., Ollé, D., Olli aul , P., Lu o, F. & B illoue , J. M. (2002). Lea and peel ola ile compounds o an in e speci ic ci us soma ic hyb id [Ci us au an i olia (Ch is m.) Swing.+Ci us pa adisi Mac ayden]. Fla ou and F ag ance Jou nal 17, 416–424. Jankun, A. & Ko anda M. (1987). Emb yological s udies o So bus 2. apomixis and o igin o So bus bohemica. P eslia 59, 97–116. Ká pá i, Z. (1960). Die So bus-A en Unga ns und de ang enzenden Gebie e. Feddes Repe o ium 62, 71–334. Knudsen, J. T. & Tolls en, L. (1993). T ends in lo al scen chemis y in pollina ion synd omes: lo al scen composi ion in mo h-pollina ed axa. Bo anical Jou nal o he Linnean Socie y 113, 263–284. Ku zelnigg, H. (1995). So bus. In Cone , H. J. e al. (Eds.) Gus a Hegi, Illus ie e Flo a on Mi eleu opa 4 (2B): Spe ma ophy a: Angiospe mae: Dico yledones 2 (3). Rosaceae 2. (pp. 328–385). Blackwell, Be lin. Ku zelnigg, H. (2005). So bus L em C an z – Mehlbee e, Ebe esche, Elsbee e, Speie ling. In Jäge , E. J & Ro hmale , K. W. (Eds.) Exku sions lo a on Deu schland 4, Ge äßp lanzen, K i ische Band 9 d edn. (pp 403–406). Spek um, Heidelbe g. Lepší, M., Ví , P., Lepší, P., Boublík, K. & Kolář, F. (2009). So bus po ae-bohemicae and So bus albensis, wo new endemic apomic ic species ecognized based on a e ision o So bus bohemica. P eslia 8, 63–89. Le in, R. A., McDade, L. A. & Raguso, R.A. (2003). The sys ema ic u ili y o lo al and ege a i e ag ance in wo gene a o Nyc aginaceae. Sys ema ic Biology 52, 334– 351. Meye , N., Meie o , L., Schuwe k, H. & Ange e , O. (2005). Bei äge zu Ga ung So bus in Baye n. Be ich e Baye ische Bo anische Gesellscha Sonde band, 5–216. Nogle , G. A. (1984). Game ophy ic apomixis. In Joh i, B M. (Ed.) Emb yology o angiospe ms. (pp. 475–518). Sp inge -Ve lag, Be lin. Plepys, D., Iba a, F. & Lö s ed , C. (2002). Odou -media ed nec a o aging in he sil e Y mo h, Au og apha gamma (Lepidop e a: Noc uidae): beha iou al and elec ophysiological esponses o lo al ola iles. Oikos 99, 75–82. Raguso, R. A., Le in, R. A., Foose, S. E., Holmbe g, M. W. & McDade, L. A. (2003). F ag ance chemis y, noc u nal hy hms and pollina ion‘‘synd omes’’ in Nico iana. Phy ochemis y 63, 265–284. 93 Rich, T., Hous on, L. & Robe son, A. (2010). Whi ebeams, Rowans and Se ice T ees o B i ain and I eland A Monog aph o B i ish and I ish So bus L. BSBI Handbook 14, London. Robe son, A., Rich, T. C. G., Allen, A. M., Hous on, L., Robe s, C., B idle, J. R., Ha is, S. A. & Hiscock, J. S. (2010). Hyb idiza ion and polyploidy as d i e s o con inuing e olu ion and specia ion in So bus. Molecula Ecology 19, 1675–1690. Salzman, C. C., Na della, A. M., Cozzolino, S. & Schies l, F.P. (2007). Va iabili y in lo al scen in ewa ding and decep i e o chids: The Signa u e o pollina o -imposed selec ion? Annals o Bo any 100, 757–765. S ökl, J., Schlue e , P. M., S uessy, T. F., Paulus, H. F., Assum, G., & Ayasse, M. (2008). Scen a ia ion and hyb idiza ion cause he displacemen o a sexually decep i e o chid species. Ame ican Jou nal o Bo any 95, 472–481. S ensson, G.P., Okamo o, T., Kawaki a, A., Go o, R. and Ka o, M. (2010). Chemical ecology o an obliga e pollina ion mu ualism: es ing he ‘p i a e channel’ hypo hesis in he B eynia-Epicephala associa ion. New Phy ologis 186: 995–1004. Ve eecken N. J., Cozzolino S. & Schies l F.P. (2010). Hyb id lo al scen no el y d i es pollina o shi in sexually decep i e o chids. BMC E olu iona y Biology 10, 103. 94 Danksagung Danksagung Ohne die Hil e zahl eiche Pe sonen wä e diese A bei nich möglich gewesen. Den nächs genann en Kollegen bin ich ü ih e Un e s ü zung bei diese A bei besonde s zu Dank e p lich e : D . And eas Jü gens, P o . D . S e an Dö e l, PD D . Ul ich Me e, P o . D . Sig id Liede Schumann, PD D . G ego Aas. D . And eas Jü gens üh e mich in die Me hode de Blü endu un e suchungen ein. E bo mi zusammen mi Ul ich Me e, Sig id Liede-Schumann und S e an Dö e l am Leh s uhl P lanzensys ema ik die Möglichkei , meine Fo schungen übe Blü endü e zu beginnen. G oße Dank da ü . Ebenso g oß is de Dank an S e an Dö e l, de mich die P omo ionszei hindu ch imme mi Ra und Ta anlei e e in allen F agen und um das, was man un e Headspace n „D eck“ nenn , dessen Anwesenhei abe manchmal auch du ch meh maliges Kop waschen nich gleich „blank“ gepu z we den konn e. Da übe hinaus un e ande em ü den sogenann en s a is ical suppo , ohne den es nich s gewo den wä e. G oße Dank an Ul ich Me e, de du ch s e ige Un e s ü zung en scheidend zu Ve besse ung de axonomischen Quali ä de A bei en bei ug und mi Ra und Ta zu Sei e s and. P o . D . Sig id Liede-Schumann Dank da ü , dass sie mi meine A bei am Leh s uhl P lanzensys ema ik du chzu üh en e möglich e, mich imme un e s ü z e und die End edigie ung meine Publika ionen übe nahm. PD D . G ego Aas danke ich ü die Zusammena bei in Sachen Mehlbee e und so manches T öp chen Elsbee enlikö , manchmal auch Ka ee. Dank an D . F anz Schuhwe k, No be Meye , D . Gün he Go schlich und D . Jochen Mülle bei de Un e s ü zung zu den Hie acien und iele in e essan e Exku sionen. Dank an Michaela Hochholze und D . Al ons Weig ü die Du ch üh ung de AFLP Analysen und das s e e o ene Oh . Dank an D . I mga d Schä le , meine Zimme kollegin ü die iele p ak ische Hil e und Tipps wäh end meine A bei . Ih , Anne Heiduk und Sab ina Sa le bin ich auch ü das Ko ek u enlesen dankba . Dank an D . Hannah Bu ge ü die Hil e bei den schwäbischen Hie acien. Dank an die Regie ung on Obe anken (He Neumann), die meine A bei zu sel enen endemischen Hie acium und So bus-A en ö de en und un e s ü z en. Dank an die Obe ankens i ung ü die Finanzie ung de AFLP-Analysen. Ganz he zlich möch e ich mich bei dem Leh s uhl eam ü die kollegiale Be ei scha bedanken. Danke an Angelika Täube , Ba ba a Me h, Pe a K aus, Ma gi Gebaue und Tina Leis ne . Die Gä ne des Ve auens wa en Reinha d K ug, Heike Deinlein, Sabine Ba hel, und Geo g Seidle . Besonde s möch e ich mich auch bei meine Mu e ü die Un e s ü zung bedanken. 95 Eidess a liche E klä ung Eidess a liche E klä ung Hie mi e klä e ich, dass ich die A bei selbs ändig e ass und keine ande en als die on mi angegebenen Quellen und Hil smi el benu z habe. Fe ne e klä e ich, dass ich ande wei ig mi ode ohne E olg nich e such habe, diese Disse a ion einzu eichen. Ich habe keine gleicha ige Dok o p ü ung an eine ande en Hochschule endgül ig nich bes anden. Bay eu h, den 96