Geographic variation in Festuca rubra L. ploidy levels and systemic fungal endophyte frequencies

RESEARCH ARTICLE Geog aphic Va ia ion in Fes uca ub a L. Ploidy Le els and Sys emic Fungal Endophy e F equencies Se da Di ihan 1 , Ma jo Helande 1,2 , Hen y Va ¨ e 3 , Ped o E. Gundel 4 , Lucas A. Ga ibaldi 5 , J. Gonzalo N. I isa i 4 , I ma Saloniemi 1 , Ka i Saikkonen 2 1Depa men o Biology, Uni e si y o Tu ku, Tu ku, Finland, 2Na u al Resou ces Ins i u e Finland (Luke), Tu ku, Finland, 3Bo anical Museum, Finnish Museum o Na u al His o y, Uni e si y o Helsinki, Helsinki, Finland, 4IFEVA, Facul ad de Ag onomı ´a, Uni e sidad de Buenos Ai es, CONICET, Buenos Ai es, A gen ina, 5G upo de In es igacio ´n en Ag oecologı´a (AGRECO), Sede Andina, Uni e sidad Nacional de Rı ´o Neg o (UNRN) and Consejo Nacional de In es igaciones Cien ı ´ icas y Te ´cnicas (CONICET), San Ca los de Ba iloche, Rı ´o Neg o, A gen ina *[email p o ec ed] Abs ac Polyploidy and symbio ic Epichloe ¨ ungal endophy es a e common and he i able cha ac e - is ics ha can acili a e en i onmen al ange expansion in g asses. He e we examined geo- g aphic pa e ns o polyploidy and he equency o ungal endophy e colonized plan s in 29 Fes uca ub a L. popula ions om eigh geog aphic si es ac oss la i udes om Spain o no he nmos Finland and G eenland. Ploidy seemed o be posi i ely and nega i ely co e- la ed wi h la i ude and p oduc i i y, espec i ely. Howe e , he co ela ions we e nonlinea ; 84% o he plan s we e hexaploids (2n = 6x = 42), and he posi i e co ela ion be ween ploidy le el and la i ude is he esul o only ou popula ions skewing he da a. In he sou h- e nmos end o he g adien 86% o he plan s we e e aploids (2n = 4x = 28), whe eas in he no he nmos end o he g adien one popula ion had only oc oploid plan s (2n = 8x = 56). Endophy es we e de ec ed in 22 ou o he 29 popula ions. Endophy e equencies a ied among geog aphic si es, and popula ions and habi a s wi hin geog aphic si es i espec i e o ploidy, la i ude o p oduc i i y. The highes o e all endophy e equencies we e ound in he sou he nmos end o he g adien , Spain, whe e 69% o plan s ha bo ed endophy es. In no he n Finland, endophy es we e de ec ed in 30% o g asses bu endophy e equencies a ied among popula ions om 0% o 75%, being highe in meadows compa ed o i e - banks. The endophy es we e de ec ed in 36%, 30% and 27% o he plan s in Fa oe Islands, Iceland and Swi ze land, espec i ely. P ac ically all examined plan s collec ed om sou h- e n Finland and G eenland we e endophy e- ee, whe eas in o he geog aphic si es endo- phy e equencies we e highly a iable among popula ions. Common o all popula ions wi h high endophy e equencies is hea y e eb a e g azing. We p opose ha he de ec ed endophy e equencies and ploidy le els mi o pas dis ibu ion his o y o F. ub a a e he las glacia ion pe iod, and local adap a ions o pas o p e ailing selec ion o ces such as e eb a e g azing. PLOS ONE | DOI:10.1371/jou nal.pone.0166264 No embe 15, 2016 1 / 16 a11111 23(1 $&&(66 Ci a ion: Di ihan S, Helande M, Va¨ e H, Gundel PE, Ga ibaldi LA, I isa i JGN, e al. (2016) Geog aphic Va ia ion in Fes uca ub a L. Ploidy Le els and Sys emic Fungal Endophy e F equencies. PLoS ONE 11(11): e0166264. doi:10.1371/jou nal.pone.0166264 Edi o : Tzen-Yuh Chiang, Na ional Cheng Kung Uni e si y, TAIWAN Recei ed: Augus 5, 2016 Accep ed: Oc obe 25, 2016 Published: No embe 15, 2016 Copy igh : 2016 Di ihan e al. This is an open access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal au ho and sou ce a e c edi ed. Da a A ailabili y S a emen : All ele an da a a e wi hin he pape and i s Suppo ing In o ma ion iles. Funding: This wo k was suppo ed by he Academy o Finland g an s 137909, 281354 and 292732 Fund ecei e au ho : KS, as well as INTERACT (G an Ag eemen No. 262693) unde he Eu opean Communi y’s Se en h F amewo k P og amme. Compe ing In e es s: The au ho s ha e decla ed ha no compe ing in e es s exis . In oduc ion Biogeog aphic gene aliza ions on he ac o s esponsible o pa e ns o species’ anges a e la gely based on compa isons o closely ela ed species [1]. Fo example, polyploidy appea s o be posi i ely associa ed wi h la i ude, al i ude and ecen deglacia ions [2–5], and bio ic in e - ac ions ha e been connec ed o adap i e adia ion o plan s [6–8]. Al hough di e en ploidy le els ha e commonly been documen ed wi hin species as well, ex ensi e s udies on geog aphic species-speci icploidy dis ibu ions and impo ance o bio ic in e ac ions a e mos ly lacking [9], and spo adic indings a e o en con lic ing. G asses a e a pe ec model o s udies on geog aphic ploidy dis ibu ions and impo ance o bio ic in e ac ions because hey co e highe a ea o land han any o he g oup o plan s ac oss all he con inen s excep An a c ica [10]. We selec ed ed escue (Fes uca ub a L. sensu la o) as a model species o ou s udy. Fi s , i is a wild pe ennial Eu asian g ass widely dis ib- u ed and pheno ypically a iable in he No he n hemisphe e. Plan s alling in o mo phologi- cally dis inguishable ca ego ies a e o en inconsis en ly classi ied as bo h species and subspecies showing local adap a ions [11,12]. A ailable li e a u e sugges s ex ensi e hyb idiza- ion be ween (sub)species, po en ially esul ing in obse ed na u al polyploids (2n = 14, 21, 28, 42, 49, 56, 64 and 70) [11,12]. In e e ile plan s exhibi s iking mo phological a ia ion and pa o he eco ypic di e si y is sugges ed o be ela ed o he polyploidiza ion [13]. Second, ed escue is well known o i s a iable and occasionally high equencies o sys emic ungal endo- phy es [14–17]–plan associa ed ungi ha a e sugges ed o ac as de ensi e plan mu ualis s and he eby expand dis ibu ion ange o he hos g ass [7,18–21]. Bo h polyploidy and sys emic, e ically in ge mline ansmi ed Epichloë endophy es a e common g ass cha ac e is ics ha can be adap i e o a ious en i onmen al condi ions [2,4,6, 7,22–24]. Polyploidy, he mul iplica ion o he comple e se o ch omosomes, can bes ow adap i e po en ial and e olu iona y lexibili y on plan s and he eby imp o e hei compe i i e and in asi e capaci y in o no he n la i udes [4,23,25]. Possessing mo e han wo se s o ch o- mosomes can cause he e osis, shield polyploids om dele e ious e ec s o mu a ions o exam- ple by educing he incidence o homozygous ecessi es, and bu e agains inb eeding dep ession and gene ic d i [2,5,23]. Because polyploids a e usually unable o in e b eed wi h hei diploid conspeci ics, polyploidy is ecognized as one o he majo mechanisms o sympa - ic specia ion [5] and in some species he numbe o ch omosomes appea s o be posi i ely co - ela ed wi h la i ude o al i ude [2–4,26]. Recen ly, he ole o polyploidiza ion as a modula o o adap i e symbiosis be ween plan s and mic obes has been ecognized [27]. Fo example, polyploidiza ion can a ec bio ic in e ac ions h ough changes in he chemical p o ile o he plan [8]. Howe e , he empi ical e idence is a iable and pa ly con adic o y [28–30], and he ques ion how ploidy-d i en adap a ions o en i onmen al condi ions and mic obial in e - ac ions code e mine plan dis ibu ion is unknown. Simila o polyploidy, Epichloë endophy es examined in his s udy can also d i e he geo- g aphic dis ibu ion o hos g asses [31]. G ass endophy es o he genus Epichloë a e common symbion s o cul i a ed and wild Pooideae g ass species. In he symbiosis endophy ic ungus g ows sys emically and asymp oma ically h oughou he abo eg ound plan pa s, and depending on he species i may be ansmi ed ei he e ically ia hos seeds and/o ho izon- ally by sexual spo es [6,7]. In many species, he ungus is mos ly asexual and is ansmi ed e ically om mo he plan o i s o sp ing. Ve ical ansmission is supposed o p omo e ideli y be ween pa ne s and lead o mu ualis ic symbiosis because he i ness o he he i able ungus and he hos g ass is igh ly linked [32–38]. Thus, he symbioses a e commonly hough o be mu ualis ic. Nume ous s udies ha e demons a ed ha Epichloë species can inc ease g ass esilience o d ough , looding,pa hogens and he bi o es, and hus p omo e hei Polyploidy and Endophy ic Symbiosis ac oss La i ude PLOS ONE | DOI:10.1371/jou nal.pone.0166264 No embe 15, 2016 2 / 16 compe i i e abili y in g ass communi ies [14,18,21,39–41]. The de ensi e mu ualism agains he bi o es due o myco oxins appea s o p o ide he mos p e alen g ound o mu ualis ic endophy e-g ass in e ac ions [6,7,20,21,42,43]. Howe e , an inc easing numbe o empi ical s udies has e ealed ha he symbiosis can ange om an agonis ic o mu ualis ic in e ac ions, and mu ualism is less equen in wild g asses compa ed o ag onomic g asses in nu ien - ich en i onmen s [20,21,38,44,45]. This s udy aims o e eal po en ial linkages be ween geog aphic pa e ns o polyploidy and plan - ungal endophy e symbiosis. We explo e bo h ploidy le els and endophy e equencies in ed escues o e a wide ange o en i onmen s and la i udes ac oss Eu ope. The bene i s om Epichloë species a e ound o be posi i ely co ela ed wi h high nu ien a ailabili y and p o- duc i i y [20,46], and polyploidy is belie ed o inc ease owa ds less p oduc i e highe la i- udes and al i udes [4,23]. Thus, we s udied p ima y p oduc i i y ( he no malized di e ence ege a ion index, NDVI) o ou s udy si es [47]. We also hypo hesize ha ploidy le els and he equencies o endophy e symbiosis would be co ela ed, because polyploidiza ion can modu- la e plan -mic obe in e ac ions and bo h polyploidiza ion and sys emic g ass endophy es can p omo e hos i ness. Because bene i s o endophy es a e pa icula ly ob ious in high nu ien en i onmen s, we may assume ha he posi i e co ela ion should be s onge in en i onmen s wi h he highes p ima y p oduc ion. This assump ion is suppo ed by a ecen b oad-scale s udy [46] sugges ing ha p ima y p oduc ion is posi i ely associa ed wi h he occu ence o sys emic g ass endophy e symbiosis. Ma e ials and Me hods Mo e han one housand ed escue (Fes uca ub a L.) plan s in o al we e collec ed om 29 popula ions (10–70 plan s/popula ion) om eigh geog aphic si es ac oss Eu ope (Spain, Swi - ze land, sou he n Finland, Fa oe Islands, Iceland, wo a eas in no he n Finland and G een- land; Table 1). Plan s we e no collec ed om na ional pa ks o o he p o ec ed a eas equi ing pe missions. In No dic coun ies “e e yman’s igh ” gi es e e yone igh s o access, enjoy o ec ea ion and collec unp o ec ed plan s, be ies and mush ooms in a way which does no damage he en i onmen o dis u b o he s ega dless who owns o occupies he land. Plan s om Spain and Swi ze land we e collec ed om public land. Geog aphical coo dina es, al i- ude om he sea le el and ea u es o he si e we e eco ded o each popula ion (Table 1). To ensu e he p ope species iden i ica ion o he plan s and ha collec ed plan s ep esen indi idual gene s, only lowe ing indi iduals g owing a leas 10 me e s apa om each o he we e collec ed. Plan s we e dug up wi h a soil co e and placed in o plas ic bags o anspo a- ion. All he collec ed g asses we e plan ed in 250 ml po s wi h added pea and sand mix a ound he o iginal soil co e and kep in a g eenhouse in Tu ku Uni e si y Bo anical Ga den (60°26’N, 22°10’E) in ambien dayligh and 20–24°C (summe ime) and 4–8°C (win e ime) empe a u es. The collec ion si es ep esen a b oad biogeog aphical egion a ying in e ms o la i udes, al i udes, clima ic zones (con inen al, oceanic), biological selec ion p essu es such as g azing (Table 1) and seasonal changes in abio ic en i onmen al condi ions. Fo example, si es in Spain ep esen g assland and xe ophy ic o es , bo h in Medi e anean clima e cha ac e ized by summe d ough s and ainy win e s. Si es loca ed on highe la i udes a e cha ac e ized by s onge seasonal changes in day leng h and associa ed ligh quali y limi ing p ima y p oduc- ion [48], sho g owing seasons in summe and long and cold win e s. Howe e , oceanic si es (Iceland and Fa oe Islands) s ongly a ec ed by he Gul -s eam a e cha ac e ized by high p e- cipi a ion yea - ound, cool summe s and ela i ely mild win e s compa ed o he o he si es on he same la i udes. Polyploidy and Endophy ic Symbiosis ac oss La i ude PLOS ONE | DOI:10.1371/jou nal.pone.0166264 No embe 15, 2016 3 / 16 Table 1. The collec ion si es o Fes uca ub a plan s and hei a ibu es. N = numbe o collec ed plan s. Mean NDVI = mean no malized di e ence ege a ion index es ima ed om yea 2000 o 2012. In . % = pe cen age o endophy e in ec ed plan s (endophy e equency) in popula ion. 4x %, 6x % and 8x % = pe cen age o e aploids, hexaploids and oc oploids, espec i ely, in popula ion. Popula ion code Geog aphic si e Popula ion N Geog aphic coo dina es Al i ude (m a.s.l.) Mean NDVI In . % 4x % 6x % 8x % Fea u es o he si e G azing SP1 Spain Ca ´ce es 31 N 40˚12’1’’ W 5˚45’11’’ 768 0,59 81 100 0 0 Xe ophy ic o es High (ca le) SP2 Spain Salamanca 1 27 N 40˚56’20’’ W 6˚7’6’’ 863 0,57 67 78 22 0 Meadow High (ca le) SP3 Spain Salamanca 2 37 N 40˚58’24’’ W 5˚57’33’’ 812 0,51 59 81 3 16 Meadow High (ca le) SW1 Swi ze land Ande ma 26 N 46˚32’19’’ E 8˚40’31’’ 1500 0,42 23 0 96 4 Meadow wi h na u al g assland ege a ion, slope acing eas High (ca le) SW2 Swi ze land Biez 25 N 46˚37’42’’ E 8˚35’26’’ 1600 0,41 36 0 100 0 Meadow wi h na u al g assland ege a ion, slope acing no h-wes High (ca le) SW3 Swi ze land Piasca 23 N 46˚53’56’’ E 8˚42’9’’ 1850 0,19 22 0 100 0 Meadow wi h na u al g assland ege a ion, sawed e e y second yea , slope acing sou h High (ca le) FI1 Sou he n Finland Hanko 1 42 N 59˚50’23’’ E 23˚13’40’’ 0 0,55 0 2 98 0 Meadow along he coas Low FI2 Sou he n Finland Hanko 2 44 N 59˚50’27’’ E 23˚13’15’’ 0 0,53 0 0 93 7 Meadow along he coas Low FI3 Sou he n Finland Hanko 3 40 N 59˚53’0’’ E 23˚5’52’’ 0 0,43 0 5 95 0 Meadow along he coas Low FO1 Fa oe Sandoy 39 N 61˚50’11’’ W 6˚51’21’’ 69 0,42 21 3 97 0 Meadow High (sheep) FO2 Fa oe Nolsoy 41 N 62˚1’15’’ W 6˚41’8’’ 55 0,33 5 0 98 2 Meadow High (sheep) FO3 Fa oe Mykines 37 N 62˚5’51’’ W 7˚40’56’’ 125 0,16 68 0 100 0 Meadow High (sheep) FO4 Fa oe Vaga 24 N 62˚6’59’’ W 7˚26’43’’ 246 0,30 25 0 83 17 Meadow High (sheep) FO5 Fa oe Eys u oy 39 N 62˚17’24’’ W 7˚2’10’’ 316 0,34 54 5 87 8 Meadow High (sheep) FO6 Fa oe Vidoy 32 N 62˚22’3’’ W 6˚32’32’’ 148 0,31 44 9 91 0 Meadow High (sheep) IC1 Iceland Iceland 1 44 N 64˚47’34’’ W 21˚32’0’’ 390 0,28 32 2 98 0 Meadow High (sheep) IC2 Iceland Iceland 2 34 N 64˚48’52’’ W 23˚23’14’’ 10 0,34 32 0 97 3 Meadow High (sheep) IC3 Iceland Iceland 3 42 N 66˚1’21’’ W 20˚23’39’’ 38 0,30 26 5 95 2 Meadow High (sheep) GR1 G eenland G eenland 1 70 N 69˚14’59’’ W 53˚31’15’’ 0 0,16 3 0 100 0 Meadow along he coas Low GR2 G eenland G eenland 2 10 N 69˚15’27’’ W 53˚32’40’’ 0 0,19 0 0 100 0 Meadow along he coas Low FI4 No he n Finland 1 Hal i 1 22 N 69˚15’0’’ E 21˚24’36’’ 860 0,20 0 0 100 0 Meadows along i ule s abo e ee-line wi h pa chy g ass and sedge domina ed ege a ion Mode a e ( eindee ) FI5 No he n Finland 1 Hal i 2 42 N 69˚15’0’’ E 21˚19’12’’ 900 0,11 0 0 0 100 Mode a e ( eindee ) FI6 No he n Finland 1 Hal i 3 32 N 69˚16’12’’ E 21˚19’12’’ 920 0,28 0 0 100 0 Mode a e ( eindee ) FI7 No he n Finland 2 Ke o 1 34 N 69˚38’6’’ E 27˚5’1’’ 91 0,31 56 3 97 0 Meadow High ( eindee ) (Con inued) Polyploidy and Endophy ic Symbiosis ac oss La i ude PLOS ONE | DOI:10.1371/jou nal.pone.0166264 No embe 15, 2016 4 / 16 Ploidy de e mina ion Ploidy le els o he plan s we e de e mined by low cy ome y(FCM) [49]. We used known ch omosome coun s o F. ub a plan s as e e ences o he FCM esul s [49]. Plan s om di - e en popula ions (one o h ee plan s o each popula ion) we e andomly chosen o mic o- scopically de e mine cy o ype. These e e ence plan s we e g own hyd oponically o p oduce esh oo ips. Asep ically cu oo ips we e p e ea ed wi h 1% 1-alphab omonaph alene and s ained in 2% ace ic o cein solu ion [50]. The oo ips we e hen squashed in a d op o 45% ace ic acid on he slides and analyzed unde mic oscope. P epa a ions we e moun ed wi h en halan a e he me aphases we e pho og aphed. We sampled a ca. 0.5 cm 2 lea piece om each plan . The sample was chopped in a glass Pe i dish wi h an asep ic azo blade in 1 ml ice-cold nuclei isola ion bu e (LBO1 in one-s ep p ocedu e). The suspension was mixed by sucking and discha ging wi h he pipe e se e al imes and hen il e ed in o an Eppendo ube using a 50 μm nylon mesh. DNA luo och ome s ock solu ion wi h 50 μl ml -1 p opidium iodide (PI) and 50 μl ml -1 ibonuclease (RNase) was added and incuba ed on ice one hou be o e FCM analysis [51]. The 96 well pla e-based FCM p ocedu e [49] was ca ied ou using LSR II (Bech on Dickinson San Jose, USA) low cy ome- e a he Tu ku Cen e o Bio echnology, Finland. Pisum sa i um L. ‘C i ad’ (2C DNA alue = 9.09 pg) plan s ob ained om he Ins i u e o Expe imen al Bo any (Labo a o y o Molecula Cy ogene ics and Cy ome y, Olomouc, Czech Republic), we e used as an ex e nal e e ence o de e mine DNA quan i y in pic og ams (pg). In addi ion, known e aploid and hexaploid F. ub a plan s we e used o each FCM un. The FCM channel was de e mined as G 1 peak o each sample and DNA ploidy le els we e es ima ed as ollows: Ploidy le el o sample ¼G1peak lou escence o sample ðmedianÞ  G1peak lou escence o e e ence ðmedianÞ Ploidy le el o e e ence The low cy ome ic da a we e measu ed wi h Flowing So wa e e sion 2.4.1 (Pe u Te ho, Tu ku Cen e o Bio echnology, Finland; www. lowingso wa e.com). Endophy e de ec ion Fungal endophy e s a us (endophy e in ec ed E+ / endophy e ee E-) o each o he s udy plan s was de ec ed by pla ing h ee lea shea hs ɖsu ace- s e ilized by incuba ion o 1 min in 90% e hanol, 4 min in 4% sodium hypochlo i e and 30 s in 90% e hanol ɖcu in o 5 pieces and placed on po a o dex ose aga (5% PDA). The Pe i dishes we e moni o ed up o h ee weeks Table 1. (Con inued) Popula ion code Geog aphic si e Popula ion N Geog aphic coo dina es Al i ude (m a.s.l.) Mean NDVI In . % 4x % 6x % 8x % Fea u es o he si e G azing FI8 No he n Finland 2 Ke o 2 40 N 69˚43’56’’ E 27˚12’0’’ 85 0,31 75 3 85 13 Meadow High ( eindee ) FI9 No he n Finland 2 Ke o 3 34 N 69˚45’32’’ E 26˚59’19’’ 107 0,31 50 0 94 6 Meadow High ( eindee ) FI10 No he n Finland 2 Ke o 4 42 N 69˚54’36’’ E 27˚1’48’’ 73 0,27 45 2 98 0 Ri e bank High ( eindee ) FI11 No he n Finland 2 Ke o 5 31 N 69˚56’11’’ E 26˚27’45’’ 106 0,28 23 0 94 6 Ri e bank High ( eindee ) FI12 No he n Finland 2 Ke o 6 35 N 69˚56’41’’ E 26˚43’22’’ 85 0,29 20 0 100 0 Ri e bank High ( eindee ) doi:10.1371/jou nal.pone.0166264. 001 Polyploidy and Endophy ic Symbiosis ac oss La i ude PLOS ONE | DOI:10.1371/jou nal.pone.0166264 No embe 15, 2016 5 / 16 o hei sys emic endophy e ungal g ow h. When ypical whi e Epichloë ungal endophy e mycelia g ew ou om se e al lea pieces, he plan was conside ed as endophy e in ec ed [16]. The in ec ion s a us o indi idual plan s was e i ied la e by s aining and mic oscopic exami- na ion o se e al seeds o each plan . Sys emic and e ically ia hos g ass ansmi ed endo- phy ic ungi a e hos species-speci ic.Simila ly o o he s udies, we ha e iden i ied he ungus associa ed wi h ed escue as E. es ucae by compa ing he DNA sequences wi h Blas sea ches o GeneBank in ou p e ious s udies [16]. NDVI To compa e ege a ion p oduc i i y o he s udy si es quan i a i ely, we calcula ed he no mal- ized di e ence ege a ion index (NDVI) o each popula ion sepa a ely using NASA MODIS sa elli e images (Fig 1,Table 1). The NDVI accu a ely es ima es unc ional a ibu es o he eco- sys em such as abo eg ound ne p ima y p oduc ion (ANPP), i s in e -annual a ia ion and ege a ion phenology [47,52,53]. NDVI is closely and posi i ely co ela ed wi h lea a ea and he ac ion o pho osyn he ically ac i e adia ion abso bed by g een ege a ion [54–56]. NDVI da a uniquely allows si e cha ac e iza ion because i ep esen s he speci ic consequences o en i onmen al and human e ec s on ege a ion unc ioning. We ob ained NDVI alues om he MODIS p ojec h ough he MODIS global subse s ool (h p://daac.o nl.go /cgibin/MODIS/GLBVZ1Glb/modissubse o de globalcol5.pl). We used he MOD 13 Vege a ion Indices p oduc , g idded, 16-day composi e images wi h 250-m pixel size. We ex ac ed wo essen ial a ibu es o abo eg ound p ima y p oduc ion dynamics by calcula ing he NDVI a e age annual in eg al and i s in e -annual coe icien o a ia ion om Feb ua y 2000 o Decembe 2012 (Fig 1). These ai s a e known o cap u e impo an ea u es o ecosys em unc ioning [57]. S a is ical analysis We used linea models (linea eg ession) o analyse he e ec s o mean NDVI and la i ude on endophy e in ec ion equencies and mean ploidy o popula ions. No mali y o esiduals was checked g aphically and using Kolmogo o -Smi no es esul s o he models. We used a logis ic eg ession model o es ima e ploidy le el, al i ude (Al ), la i ude (La ) and hei pai -wise in e ac ions as ixed e ec s on endophy e s a us ( wo le els: E+ and E-) using binomial e o dis ibu ion (logi link). We also included popula ion as a andom e ec ( an- dom in e cep model) o accoun o he ac ha indi idual plan s we e spa ially nes ed wi hin popula ions (29 popula ions in o al). We used AIC (Akaike In o ma ion C i e ion) o selec bes - i ing models o all combina ions o ixed-e ec a iables. AIC alues we e ob ained based on maximum-likelihoodes ima es o eg ession coe icien s,because models di e ed in hei ixed s uc u e bu sha ed he same andom s uc u e ( andom in e cep s), whe eas pa ame e es ima es o inal models p esen ed in igu es we e ob ained using he es ic ed maximum likelihood me hod [58]. Models we e es ima ed using lme unc ion o he lme4 package [59] in he R so wa e [60], and AIC and Akaike’s weigh o each model o all possible models based on di e en combina ions o he p edic o a iables we e ob ained wi h he d edge unc ion o he MuMln package [61] in he R so wa e. Resul s O e all 84%, 9% and 7% o he plan s we e hexaploids (2n = 6x = 42), e aploids (2n = 4x = 28) and oc oploids (2n = 8x = 56), espec i ely. Mo e han one ploidy le el was de ec ed in 19 ou o 29 popula ions (Table 1). Ploidy le el seems o be pa ly linked wi h p oduc i i y and la - i ude; ploidy posi i ely and nega i ely associa ed wi h la i ude (p<0.0001) and p oduc i i y Polyploidy and Endophy ic Symbiosis ac oss La i ude PLOS ONE | DOI:10.1371/jou nal.pone.0166264 No embe 15, 2016 6 / 16 (p<0.0001), espec i ely (Fig 2). I is no ewo hy, howe e , ha majo i y o he plan s we e hexaploids (2n = 6x = 42), and he posi i e co ela ion be ween ploidy le el and la i ude was no linea , and he co ela ion is due o he in luence o ou de ian popula ions. In he sou h- e nmos geog aphic si e, Spain, 86% o he plan s we e e aploids (2n = 4x = 28), and in no h- e nmos Finland all o he plan s in one popula ion a high al i ude (in Hal i, 900m) we e Fig 1. Mon hly mean no malized di e ence ege a ion index (mean NDVI) alues o he Fes uca ub a collec ion si es ( o popula ion codes see Table 1) om Feb ua y 2000 o Decembe 2012. doi:10.1371/jou nal.pone.0166264.g001 Polyploidy and Endophy ic Symbiosis ac oss La i ude PLOS ONE | DOI:10.1371/jou nal.pone.0166264 No embe 15, 2016 7 / 16 oc oploids (2n = 8x = 56) (Table 1,Fig 2). Ou side hese ex emes o he la i udinal ange, hexa- ploid plan s we e dominan . In G eenland, all he plan s we e hexaploids, and in Swi ze land 98% o he plan s we e hexaploids and only 2% oc oploids (Table 1). In he o he popula ions e aploids and oc oploids we e spo adically dis ibu ed (Table 1). Endophy e in ec ions we e de ec ed in 22 ou o 29 ed escue popula ions (Fig 3). The se en o ally endophy e- ee popula ions we e he h ee popula ions om Hanko in sou he n Finland, he h ee popula ions om Hal i in no he n Finland and one popula ion om G een- land. Fu he mo e, only one in ec ed g ass was ound in he o he popula ion in G eenland (Table 1,Fig 3). O e all in ec ion equency o he s udy a ea was 29% bu equencies a ied i espec i e o ploidy (p = 0.14) (Table 2) among geog aphic si es, popula ions wi hin geo- g aphic si es and among habi a s (Fig 3,Table 1). The highes o e all in ec ion equencies we e ound in Spain (69%) whe e occu ence o in ec ions was high in all popula ions (Fig 3). A Ke o in no he n Finland, on a e age 45% o g asses we e in ec ed bu in ec ion equencies we e much highe in meadows (60%) compa ed o nea by i e banks (29%). The occu ence o in ec ions in Fa oe Islands, Iceland and Swi ze - land we e 36%, 30% and 27%, espec i ely. In hese geog aphic si es a ia ion in endophy e occu ence among popula ions was conside able only in Fa oe Islands a ying om 5% o 68% (Table 1,Fig 3). La i ude appea s no o be linked o he de ec ed a ia ion in endophy e occu ence (p = 0.65) (Table 2,Fig 4). Ins ead, in ec ion equencies appea o be associa ed wi h al i ude (p = 0.04) bu in e ac i ely wi h la i ude (p = 0.03) (Table 2). The sampling was no , howe e , designed o es he impo ance o al i ude and hus, hese esul s emain inconclusi e. In pop- ula ions collec ed om low la i udes, in Spain and Swi ze land, he endophy e equencies we e lowes in Swi ze land whe e all collec ed popula ions we e om high al i ude (Table 1). In con- as , elsewhe e in ec ion equencies a ied i espec i e o al i ude. Fo example, popula ions collec ed om G eenland and Hanko, bo h si ua ed a sea le el, and Hal i si ua ed a 900 m abo e sea le el we e endophy e- ee (Table 1) sugges ing he al i ude canno accoun o pa - e ns o endophy e occu ence. Endophy e equencies o he g ass popula ions we e no associa ed wi h p oduc i i y (p = 0.16) (Fig 5). No malized di e ence ege a ion index (NDVI) alues a ied among geog aphic si es a he han along la i ude (Table 1,Fig 6). The highes geog aphic si e speci icmean-NDVI alues Fig 2. The e ec o la i ude (a) and mean no malized di e ence ege a ion index (mean NDVI) (b) on mean ploidy o Fes uca ub a popula ions. Fi ed eg ession line in (a) is o illus a i e pu pose since s a is ical es s a e sugges i e due o p oblems o no mali y in he da a. doi:10.1371/jou nal.pone.0166264.g002 Polyploidy and Endophy ic Symbiosis ac oss La i ude PLOS ONE | DOI:10.1371/jou nal.pone.0166264 No embe 15, 2016 8 / 16 0.51, 0.56 and 0.31 we e es ima ed o sou he n Finland, Spain and Iceland-Fa oe Islands, espec- i ely (Table 1). Co esponding endophy e in ec ion equencies we e 0% (sou he n Finland), 69% (Spain), 30% (Iceland) and 36% (Fa oe Islands) demons a ing ha o e all p oduc i i y o he geo- g aphic si e is unlikely o be linked wi h endophy e in ec ion equencies in ed escue popula ions. Mon hly mean NDVI es ima es, howe e , clea ly demons a e ha p ima y p oduc ion is season- ally limi ed in all he o he s udy si es excep in Spain (Fig 1). Discussion Ou esul s do no suppo he hypo heses ha polyploidiza ion and he occu ence o sys emic ungal endophy es in ed escue show la i udinal g adien s, o ha hey a e co ela ed wi h Fig 3. Collec ion si es and endophy e in ec ion equencies o Fes uca ub a popula ions. In ec ion equency ci cles a e shown wi h he popula ion codes (Table 1). doi:10.1371/jou nal.pone.0166264.g003 Table 2. E ec s o ploidy, la i ude, al i ude and in e ac ion be ween la i ude and al i ude (La x Al ) on endophy e s a us o Fes uca ub a plan s. Es ima e S d. E o z alue P (!|z|) In e cep -4.78 5.08 0.94 0.35 Ploidy 0.03 0.02 1.48 0.14 La i ude 0.04 0.08 0.44 0.65 Al i ude 0.01 0.00 2.03 0.04* La x Al -0.00 0.00 -2.14 0.03* *, p0.05 **, p0.01 ***, p0.001 doi:10.1371/jou nal.pone.0166264. 002 Polyploidy and Endophy ic Symbiosis ac oss La i ude PLOS ONE | DOI:10.1371/jou nal.pone.0166264 No embe 15, 2016 9 / 16 51. Dolez ˇel J, G eilhube J, Suda J. Es ima ion o nuclea DNA con en in plan s using low cy ome y. Na u e P o ocols. 2007; 2: 2233–2244. doi: 10.1038/np o .2007.310 PMID: 17853881 52. Tucke CJ, Vanp ae CV, Sha man MJ, I e sum GV. Sa elli e emo e sensing o o al he baceous bio- mass p oduc ion in he Senegalese Sahel: 1980–1984. Remo e Sensing o En i onmen . 1985; 17: 233–249. 53. P ince S. Sa elli e emo e sensing o p ima y p oduc ion: compa ison o esul s o Sahelian g asslands 1981–1988. In e na ional Jou nal o Remo e Sensing. 1991; 12: 187–216. 54. Selle s PJ, Be y JA, Colla z GJ, Field CB, Hall FG. Canopy e lec ance, pho osyn hesis, and anspi- a ion. III. A eanalysis using imp o ed lea models and a new canopy in eg a ion scheme. Remo e Sensing o En i onmen . 1992; 42: 187–216. 55. Hue e AK, Didan K, Miu a T, Rod iguez EP, Gao X, Fe ei a LG. O e iew o he adiome ic and bio- physical pe o mance o he MODIS ege a ion indices. Remo e Sensing o En i onmen . 2002; 83: 195–213. 56. Pixei o G, Oes e held M, Pa uelo J. Seasonal a ia ion in abo eg ound p oduc ion and adia ion-use e iciency o empe a e angelands es ima ed h ough emo e sensing. Ecosys ems. 2006; 9: 357– 373. 57. Nemani R, Running S. Land co e cha ac e iza ion using mul i empo al ed, nea -IR, and he mal-IR da a om NOAA/AVHRR. Ecological Applica ions. 1997; 7: 79–90. 58. Zuu A, Ieno EN, Walke N, Sa elie AA, Smi h GM. Mixed e ec s models and ex ensions in ecology wi h R. New Yo k: Sp inge Ve lag; 2009. 59. Ba es D, Maechle M, Bolke B. lme4: Linea mixed-e ec s models using S4 classes. R package e - sion 0.999375–39. 2011. A ailable: h p://www.inside- .o g/packages/lme4/ e sions/0-999375-39. 60. R De elopmen Co e Team. R: a language and en i onmen o s a is ical compu ing. R Founda ion o S a is ical Compu ing, Vienna. 2011. A ailable: h p://www.R-p ojec .o g/. 61. Ba oĔK. MuMIn: mul i-model in e ence. R package, e sion 0.12.2. 2012 A ailable: h p:// - o ge. - p ojec .o g/p ojec s/mumin/. 62. Ha bed DJ. Obse a ions on popula ion s uc u e and longe i y o Fes uca ub a L. The New Phy olo- gis . 1961; 60: 184–206. 63. Bazel DR, Vica i M, Emme ich S, Filip L, Lin D, Inman A. In e ac ions be ween he bi o es and endo- phy e-in ec ed Fes uca ub a om he Sco ish islands o S . Kilda, Benecula Rum. Jou nal o Ecology. 1997; 34: 847–860. 64. Cheplick GP, Fae h S. Ecology and e olu ion o he g ass endophy e symbiosis. New Yo k: Ox o d Uni e si y P ess; 2009. 65. Colpae A, Kumpula J, Nieminen M. Reindee pas u e biomass assessmen using sa elli e emo e sensing. A c ic. 2003; 56: 147–158. 66. Thompson JN. The geog aphic mosaic o coe olu ion. Chicago: Uni e si y o Chicago P ess; 2005. Polyploidy and Endophy ic Symbiosis ac oss La i ude PLOS ONE | DOI:10.1371/jou nal.pone.0166264 No embe 15, 2016 16 / 16