Is the Pathogenic Ergot Fungus a Conditional Defensive Mutualist for Its Host Grass?

Is he Pa hogenic E go Fungus a Condi ional De ensi e Mu ualis o I s Hos G ass? Pauliina P. Wa ¨li 1 *, Piippa R. Wa ¨li 1,2 , Ka i Saikkonen 3 , Juha Tuomi 1 1Depa men o Biology, Uni e si y o Oulu, Oulu, Finland, 2Kola i Uni , Finnish Fo es Resea ch Ins i u e, Kola i, Finland, 3Plan P oduc ion Resea ch, MTT Ag i ood Resea ch Finland, Jokioinen, Finland Abs ac I is well ecognized, ha ou comes o mu ualis ic plan -mic oo ganism in e ac ions a e o en con ex dependen and can ange om mu ualis ic o an agonis ic depending on condi ions. Ins ead, seemingly pa hogenic associa ions a e gene ally conside ed only ha m ul o plan s. The e go ungus (Cla iceps pu pu ea) is a common seed pa hogen o g asses and ce eals. E go scle o ia con ain alkaloids which can cause se e e oxici y in mammals when inges ed, and hus he ungal in ec ion migh p o ide p o ec ion o he hos plan agains mammalian he bi o es. Theo e ically, he ne e ec o e go in ec ion would posi i ely a ec hos seed se i he cos is no oo high and he de ensi e e ec is s ong enough. Acco ding o ou empi ical da a, his si ua ion is plausible. Fi s , we ound no s a is ically signi ican seed loss in wild ed escue (Fes uca ub a) in lo escences due o e go in ec ion, bu he seed succession dec eased along inc easing numbe o scle o ia. Second, in a ood choice expe imen , sheep showed a oidance agains o age con aining e go . Thi d, he equency o e go -in ec ed in lo escences was highe in sheep pas u es han su ounding ung azed a eas, indica ing a p o ec i e e ec agains mammalian g azing. We conclude ha , al hough e go can p ima ily be ca ego ized as a plan pa hogen, e go in ec ion may some imes ep esen indi ec bene icial e ec s o he hos plan . E go may hus se e as a condi ional de ensi e mu ualis o i s hos g ass, and he pa hogenic in e ac ion may ange om an agonis ic o mu ualis ic depending on he si ua ion. Ci a ion: Wa ¨li PP, Wa ¨li PR, Saikkonen K, Tuomi J (2013) Is he Pa hogenic E go Fungus a Condi ional De ensi e Mu ualis o I s Hos G ass? PLoS ONE 8(7): e69249. doi:10.1371/jou nal.pone.0069249 Edi o : Dipshikha Chak a o y, Indian Ins i u e o Science, India Recei ed Ap il 12, 2013; Accep ed June 11, 2013; Published July 10, 2013 Copy igh : ß2013 Wa ¨li 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. Funding: This s udy was suppo ed by g an s om he Ma ja a and Eino Kolli Founda ion, Ol i Founda ion, Oska O ¨ lund’s Founda ion and Thule Ins i u e (PPW), and he Academy o Finland (p ojec n o 127140) and Ella and Geo g Eh n oo h Founda ion (PRW). The unde s had no ole in s udy design, da a collec ion and analysis, decision o publish, o p epa a ion o he manusc ip . Compe ing In e es s: The au ho s ha e decla ed ha no compe ing in e es s exis . * E-mail: [email protected] In oduc ion In e ac ions be ween plan s and mic oo ganisms ha e adi- ionally been cha ac e ized based on he o en isible p ima y e ec o he mic obe on plan i ness. Pa hogens clea ly deple e hos esou ces and ac as ha m ul an agonis s o he plan . Mu ualis s, such as ce ain endophy es and myco hizas, o e he hos some bene icial se ice ou weighing he consump ion o he hos esou ces. These ela ionships a e no s aigh o wa d in na u e, and he condi ional cha ac e is ics o symbio ic (in he sense o li ing oge he ) species associa ions a e well ecognized oday. The p esumed mu ualis s may no always be ad an ageous o hos i ness, and he o al i ness e ec s may a y om bene icial o an agonis ic depending on he condi ions [1–5]. By con as , he possible condi ional aspec o seemingly pa hogenic in e ac ions is much less discussed. Mu ualis ic e ec s o mainly an agonis ic associa ions ha e p e iously been sugges ed in ela ion o plan adap a ions o he bi o y, o g azing may imp o e plan i ness h ough o e compensa ion in some si ua ions ( o discussion and e e enc- es, see [6–8]). Some animal-pa asi e in e ac ions ha e also been shown o u n bene icial o he hos in ce ain special condi ions ( e . in [9]). Simila si ua ions a e likely in plan -pa hogen in e ac ions as well, and se e al possible ac o s could con ibu e o such an al e na i e ou come. Fi s ly, a pa hogen may ha e addi ional and o en di e se subsidia y e ec s on i s hos (di ec o indi ec , ac ing on hos physiology o ecology [10]) ha can be di icul o unco e [4]. On he o he hand, he mani es a ion and impac o he e ec s on hos i ness may a y depending on condi ions (i.e. in e ac ing geno ypes, o he in e ac ing species and he g ow h condi ions o he sys em) [1], [3], [5], [11], [12]. Mo eo e , a hos ’s compensa ion and e en o e compensa ion o p ima y cos may al e he ini ial si ua ion [13]. The e go ungus, Cla iceps pu pu ea (F .) Tul., is a common seed pa hogen o empe a e g asses and ce eals. E go in ec s single g ass lo e s and de elops a ungal issue called a scle o ium ins ead o a g ass seed. E go is de ined as a plan pa hogen because i deple es he hos ’s esou ces and causes di ec seed loss o he hos plan (e.g., [14], [15]). Howe e , e go scle o ia con ain a a ie y o alkaloids, many o which a e oxic o mammals. Inges ion o e go wi hin g ass o g ain p oduc s can cause se e e and e en ually le hal in oxica ion in bo h ca le and humans. Acco dingly, e go has been esponsible o se ious poisoning epidemics in human his o y and is s ill a cause o economic losses in g ain p oduc ion, especially because he scle o ia ha e o be emo ed om he in ec ed seed se s ( e . in [16–18]). Se e al plan inhabi ing ungi, like g ass endophy es, ha e been sugges ed o play a pa in he he bi o e de ence o hei hos plan s [19], [20]. In such a case, he dec ease in he pala abili y o a plan o he bi o es is o en due o PLOS ONE | www.plosone.o g 1 July 2013 | Volume 8 | Issue 7 | e69249 oxic me aboli es p oduced by he ungus, ep esen ing a o m o ’acqui ed chemical de ence’ [21]. G ass g azing mammals ha e p e iously been shown o a oid ea ing g asses in ec ed wi h endophy ic ungi ha p oduce e go alkaloids (e.g., [22]). Al hough e go is usually conside ed me ely as a pa hogen, such a p o ec i e e ec has been specula ed o occu also wi h e go [17], [19], [23– 25]. In his s udy, we p esen (1) a simple model showing how he cos s o e go in ec ion, in e ms o los iable seeds, may be balanced by p o ec i e e ec s agains g azing. Second (2), we p esen empi ical da a o he suppo o he model. The cos s o e go in ec ion a e es ima ed wi h seed da a on wild ed escue (Fes uca ub a L. sl.). The demons a ion o he ac ual p o ec i e e ec o e go in ec ion is done ia sheep g azing: We compa ed he abundance o e go on ed escue inside and ou side sheep pas u es. I he in ec ion has p o ec i e e ec s on he hos plan , e go in ec ed in lo escences should be less equen ly ea en and he e o e ela i ely mo e abundan inside he pas u es. Fu he , we p esen expe imen al esul s on domes ic sheep ood choice in a si ua ion whe e animals we e p o ided e go - ee and, al e na- i ely, e go -con aining o age. P edic ions: how does e go in luence plan i ness? We conside a condi ion whe e e go in ec ion would be a ou able o i s hos by compa ing he a e age i nesses o e go - in ec ed and non-in ec ed plan s. Because e go is oxic o mammalian g aze s, i is highly likely ha e go in luences he isk o g azing on he hos plan . I is well-known ha he p esence o a oxic plan can dec ease g azing isk o i s neighbou s [26–28]. Acco dingly, we can p opose a hypo hesis ha a seed in ec ed by e go may p o ide associa ional de ence o o he seeds in he e go -in ec ed in lo escence. I so, hen he ne e ec s o e go on hos i ness, ela i e o a non-in ec ed plan , depends on how many seeds i will lose due o he in ec ion (cos o in ec ion) and how many i will sa e due o educed seed p eda ion. Assume ha B is he a e age numbe o seeds pe plan in he absence o g azing and ungal in ec ion. I h is he isk o g azing (0#h#1) and ais he ela i e dec ease in seed numbe pe plan when being g azed (0#a#1), he expec ed i nesses o non- in ec ed and in ec ed plan s will be B(1-ha) and B(1- b)[1-ha(1-d)], espec i ely, whe e bdeno es he ela i e decline in iable seeds pe plan i in ec ed by e go (0#b#1), and d is he ela i e decline in he isk o g azing (h) i he plan has been in ec ed by e go (0#d#1). The in ec ion will be bene icial o he hos i in ec ed plan s ha e on a e age highe i ness han non-in ec ed, o o he wise exp essed as, b had 1{ha1{dðÞ Thus, e go can bene i he plan i he cos o in ec ion is low enough and e go in ec ion has su icien ly high de ensi e e ec agains seed p eda ion (Figu e 1). This heo e ical a gumen implies wo main ques ions conce n- ing he i ness consequences o e go on he hos plan . Fi s , is low cos o in ec ion a easible assump ion in he wild? Secondly, is he e any e idence ha e go could ha e a de ence e ec in a ou o he hos plan ? I he answe s a e posi i e, i is possible ha he e go -plan ela ionship could ep esen a case o ‘‘de ensi e mu ualism’’, gene ally de ined as a mu ually bene icial ela ionship be ween wo species in which one p o ec s he o he om an enemy ha causes a i ness loss o he la e (e.g., [19]). In de ensi e mu ualism, he species ha p o ides p o ec ion may o may no i sel incu a ne i ness cos in esponse o he o he (Figu e 2A and 2B, acco dingly). In he case o g ass- ungi in e ac ions, he i s al e na i e (Figu e 2A) would co espond o a cos ly pa hogen and he o he one (Figu e 2B), o example, o an asexual Epichloe¨ endophy e ha causes no isible o appa en cos s o i s hos . In he i s scena io he symbion i sel educes hos i ness in he absence o he enemy (Figu e 2A, h = 0), and hence does no sa is y he gene al de ini ion o mu ualism in such condi ions. Howe e , he si ua ion changes in he p esence o enemies because now also a cos ly symbion may imp o e hos i ness in ela ion o symbion - ee hos s (Figu e 2A, h = 1). Acco dingly, Fellous & Sal audon [9] ha e amed he concep o ‘‘condi ionally help ul pa asi es’’, e e ing o pa asi es p o iding bene icial i ness e ec s o a hos in some special condi ions and being ha m ul in o he s. These cases a e cong uen wi h he concep s o symbio ic ela i ism [20] and mu ualism-pa asi ism con inuum [29], whe e he ecological ou come o symbiosis is seen o a y om pa asi ic o mu ualis ic among di e en en i on- men s. Ma e ials and Me hods S udy sys em The ungus Cla iceps pu pu ea has o e 400 hos species wi hin he amily Poaceae, including ye and o he economically impo an ce eals and o age g asses. E go is common in empe a e clima es and i occu s also in sub opical and a c ic egions [30], [31]. E go in ec ion is es ic ed o a single o a y/seed, bu one o se e al o he seeds in he in lo escence may ca y scle o ia due o sepa a e p ima y in ec ions o seconda y in ec ions by conidia p esen in ‘‘honeydew’’ p oduced by he ungus in he ea ly s age o in ec ion [14]. We used ed escue, Fes uca ub a, as he hos plan and domes ic sheep as he he bi o e in o de o oughly es ima e he cos s and possible de ensi e e ec s o e go in ec ion in na u e. F. ub a is a common g ass species in he wild as well as in ag icul u al habi a s in no he n Scandina ia, and is equen ly in ec ed wi h e go (Wa¨li e al., unpublished). The sheep pas u es s udied a e semi- Figu e 1. E ec o e go on hos i ness. Pa ame e space whe e e go in ec ion has a posi i e (below he lines) o nega i e (abo e he lines) ne e ec on hos i ness. b= ela i e loss o seeds due o e go in ec ion, d = p o ec i e e ec o e go in ec ion, ha= cos o he bi o y. doi:10.1371/jou nal.pone.0069249.g001 E go Pa hogen as De ensi e Mu ualis PLOS ONE | www.plosone.o g 2 July 2013 | Volume 8 | Issue 7 | e69249 na u al meadows domina ed by F. ub a and u ilized o o a ional g azing. In o a ional g azing animals a e mo ed om pas u e o pas u e du ing he g azing pe iod. Thus, he pas u e ege a ion is no ully consumed and g asses may lowe and p oduce seed du ing and ou side g azing pe iods. E ec o e go on ed escue seed numbe F. ub a indi iduals a e mainly pe ennial and long-li ed clonal colonies o med o se e al ame s lowe ing in di e en yea s. The e o e, we used he quan i y o seeds p oduced pe ame (in lo escence), a i ness co ela ing measu e, o es ima e he cos o e go in ec ion. We collec ed ma u e in lo escences o F. ub a om se en semi-na u al meadows and one i e side popula ion in no he n Finland in Sep embe 2008 and 2009 (Appendix S1). Depending on he size and in ec ion in ensi y o he popula ion, ou o 20 e go -in ec ed and unin ec ed in lo escences we e collec ed andomly om he popula ion. The numbe o seeds, e go s and emp y lo e s we e eco ded om each in lo escence, cons i u ing he o al amoun o lo e s pe in lo escence. The ull lo e s we e dissec ed o e eal possible small-sized scle o ia. E ec o g azing on e go in ec ion equencies The e ec o g azing on e go in ec ion equencies was es ima ed by sampling F. ub a in lo escences om sheep pas u es in no he n Scandina ia (Appendix S2). We collec ed 24–61 samples om each o 6 sepa a e loca ions in Sep embe 2006 and 2008. In lo escences we e collec ed a andom om mode a ely g azed sheep pas u es and om co esponding ung azed a eas ou side he pas u e ence. Inside he pas u es he o e all amoun o F. ub a in lo escences was diminished due o g azing and hus he densi y o F. ub a in lo escences di e ed be ween he g azed and ung azed si es. Sheep eac ion o e go con aining eed To es whe he la ge he bi o es a oid, o in he sho e m lea n o a oid ea ing e go , we conduc ed ood choice expe imen s wi h domes ic sheep a MTT Ag i ood Resea ch Finland, Animal P oduc ion Resea ch, Jokioinen. The single-day expe imen was ca ied ou indoo s in a sheep shed in Decembe 2009. The p e e ence o e go -con aining and e go - ee o age was es ima - ed in a pai wise es . E go scle o ia o ye (Secale ce ale) we e mixed wi h o age pelle s in a 1:4 olume a io. Hal li es o e go - con aining and e go - ee con ol pelle s we e o e ed on simila plas ic ays placed side by side on he loo . A single sheep (male, n = 6) was allowed o app oach he ays a a ime, and was allowed o choose be ween he ays one o ou imes. Sheep we e allowed o isually examine, smell and ouch he eeds, bu a me al ence placed abo e he pelle s p e en ed ac ual ea ing. In he es , he ac ions o posi i e choice (sheep ying o ea o age) o ejec ing he eed (a e isual o o he cue) we e eco ded. Ac i i y was eco ded as a posi i e choice (ea ing decision) when sheep kep hei head in he ay, ac i ely ying o ea o o e 3 seconds, and he es was con inued each ime un il he sheep selec ed one o he ays. Six ounds o he es we e conduc ed, and he a angemen o ays was changed andomly be ween animals. E hical s a emen The sheep used in his esea ch we e expe imen al animals o Animal P oduc ion Resea ch o MTT Ag i ood Resea ch Finland, Jokioinen, kep wi h ins i u ional pe mi s in acco dance wi h The Finnish Ac on Animal Expe imen a ion. The na ional Animal Expe imen Boa d o Finland was consul ed, and no speci ic pe mi s we e equi ed o he ood choice es desc ibed in his s udy, as he sheep we e no le o inges e go , and hus he es did no mee he c i e ia o an animal expe imen desc ibed in The Finnish Ac on Animal Expe imen a ion (62/2006) as "ca ying ou such expe imen s, es s, esea ch o in es iga ions on animals. which may cause pain, su e ing, dis ess o las ing ha m compa able a leas o he pain caused by he in oduc ion o a needle". No p o ec ed o endange ed species we e used in he ield s udies, and p i a e owned land was accessed wi h landowne s’ pe mission. S a is ical analyses The logis ic eg ession (binomial dis ibu ion and logi link unc ion) o e en s/ ials da a was employed o h ee sepa a e es s o es ima e 1) whe he he p opo ional e go scle o ia amoun in in lo escences di e ed among eigh wild g ass popula ions (popula ion as ixed ac o ) and 2) whe he success ul seeds om all lo e s di e ed among hese g ass popula ions and among g ass in lo escences wi h and wi hou e go in ec ion (popula ion and e go in ec ion as ixed ac o s). We es ed u he 3) whe he he numbe o e go scle o ia as con inuous a iable a ec ed he p opo ion o success ul seeds om all lo e s in in lo escences. In Figu e 2. Hos i ness and con ex dependency in de ensi e symbiosis. A e age hos i ness wi h o wi hou a p o ec i e symbion in ela ion o isk o a ack. De ensi e mu ualism assumes ha he symbion imp o es hos i ness in he p esence o he enemy (h = 1), howe e he symbion i sel may (A) o may no (B) incu a i ness cos o he hos in he absence o he enemy (h = 0). doi:10.1371/jou nal.pone.0069249.g002 E go Pa hogen as De ensi e Mu ualis PLOS ONE | www.plosone.o g 3 July 2013 | Volume 8 | Issue 7 | e69249 his case, he scale pa ame e was es ima ed by he squa e oo o de iance/do . To compa e e go in ec ion equency o F. ub a in lo escences in g azed and ung azed a eas and among si es, he da a o e go incidence in g ass in lo escences was analysed as e en / ial da a wi h logis ic eg ession (binomial dis ibu ion and logi link unc ion), he e en being he p esence o one o mo e e go scle o ia in a F. ub a in lo escence examined and he ial being a F. ub a in lo escence collec ed and examined. G azing ( wo le els) and collec ion si e (six le els) we e used as ixed ac o s in he model. The ood choice da a was analysed as e en / ial da a wi h logis ic eg ession (binomial dis ibu ion and logi link unc ion), he e en being he posi i e ood choice and he ial being he app oach o he o age ay. We es ed he e ec s o each h ee ac o s (e go , sheep indi idual and es ound) in he ood ay choice o sheep in sepa a e analyses due o he low numbe o eplica es. Because posi i e ood choice was eco ded e e y ime when he con ol ay was app oached, he da a o app oaching he e go con aining ay was used o es he di e ences among sheep indi iduals. In case o es ounds, he scale pa ame e was es ima ed by he squa e oo o de iance/do . The analyses we e pe o med using SAS 9.3, wi h he GENMOD p ocedu e. Resul s E ec o e go on ed escue seed p oduc ion In wild g ass popula ions he p opo ional e go scle o ia con en in in lo escences did no di e among eigh wild g ass popula ions in Finnish Lapland (x 2 = 6.44, P,0.17). The o e all numbe o lowe s and he seed p oduc ion (numbe o seed) a ied g ea ly: in e go in ec ed in lo escences he seed numbe a ied be ween 0 and 69, and in unin ec ed be ween 0 and 73, and he o al amoun o lowe s pe in lo escence was be ween 17 and 152 in in ec ed and be ween 22 and 139 in unin ec ed in lo escences. The p opo ional seed p oduc ion a ied signi i- can ly be ween popula ions (x 2 = 41.61, P,0.0001) (see Appendix S1 o means and con idence in e als in seed p oduc ion and e go con en in in lo escences in each popula ion examined). P esence o e go in ec ion had no signi ican e ec on success ul seed p oduc ion (x 2 = 1.20, P= 0.27), bu he inc ease in p opo - ion o e go ic lo e s (as a con inuous a iable) dec eased he p opo ion o success ul seed (x 2 = 4.45, P= 0.035). E ec o g azing on e go in ec ion equencies E go in ec ion incidence in ed escue in lo escences di e ed among g azed and ung azed a eas (x 2 1 = 26.18, P,0.0001) and among collec ion si es (x 2 5 = 51.24, P,0.0001). F equency o e go in ec ed in lo escences was highe in pas u es (43 %) han su ounding non-g azed a eas (16 %). E go in ec ion equencies o g azed and ung azed a eas in di e en collec ion si es a e p esen ed in Appendix S2. Sheep eac ion o e go con aining eed Sheep ac ions a he e go -con aining ay di e ed om ac ions a he con ol ay (x 2 = 47.53, P,0.0001). All he sheep made posi i e choice (ea ing decision) e e y ime hey app oached he con ol ay, bu some a oidance was de ec ed wi h he e go - con aining ay. Ac ions did no di e signi ican ly be ween ounds (x 2 = 3.32, P=0.65), bu sheep indi iduals di e ed in hei eac ions (x 2 = 48.64, P=0.0001). Fou o he six ams clea ly selec ed agains e go -con aining eed, when wo nea ly always made he posi i e choice om he ay hey i s app oached (Figu e 3). When aking a e ages o e ounds o indi idual sheep and o e six indi iduals, he animals had an almos equal p obabili y o app oaching con ol o e go y ays (Figu e 4). The mos p onounced di e ence was ound conce ning how he ams esponded o ood quali y. When app oaching he con ol ay, he sheep always made a posi i e choice and ne e shi ed away o he e go y ay. In con as , when hey i s app oached he e go y ay, hey chose ha ay only in 65 % o cases, and shi ed o he con ol ay in 35 % o cases. Because o he shi om he e go y o he con ol ay, con ol ood was selec ed mo e o en (69 %) as compa ed o 31 % o he e go y ood. Conce ning he a iabili y in he a e age esponses be ween indi idual sheep, indi idual A sligh ly mo e equen ly (62 % cases) app oached he con ol ay and always ejec ed he e go y ood (Figu e 3). On he o he hand, indi iduals E and F app oached mo e equen ly (78 % and 65 %, espec i ely) he e go y ay and nea ly ne e changed o ano he ay (Figu e 3). Discussion Ou model p edic s ha , in spi e o possibly losing some seeds due o in ec ion, a hos plan may bene i om he e go -p o ided p o ec ion agains g aze s, and by hese means sa e mo e seeds han i loses in he p esence o g aze s. Ou empi ical obse a ions con i m ha his hypo hesis is plausible. Fi s , he seed loss o a ame caused by he in ec ion is a he small o e en non-exis en i only a ew scle o ia pe in lo escence a e p oduced. Secondly, he e go equencies we e highe in sheep pas u es han su ounding ung azed a eas. This di e ence may well be he consequence o selec i e o aging, whe e sheep ha e a ou ed unin ec ed lowe heads and a oided in ec ed ones. This is in acco dance wi h he esul s o he ood choice expe imen , whe e sheep o en changed om a scle o ia-con aining eed o a scle o ia- ee al e na i e and ne e ice e sa. Di ec cos s and hos con ol o e in ec ion Acco ding o ou ield da a, e go may no sel -e iden ly cause ma ked di ec cos s o i s hos . By con as , clea ly nega i e e ec s o e go on g ain and seed p oduc ion (numbe and weigh o g ain/seed) ha e been epo ed wi h ag icul u al c ops and Ken ucky blueg ass (Poa p a ensis) g own o seed [32], [33]. Howe e , he esul s om cul i a ed g asses and ce eals may no co espond o he si ua ion in na u e, due o he selec i e b eeding, geno ypic uni o mi y o hos popula ions and a i icial, a he cons an g ow h condi ions, like high densi y and high nu ien concen a ions [2], [3], [34]. Wi h wild sal ma sh Spa ina species, nega i e, bu simila ly o ou esul s, also neu al and e en posi i e e ec s o e go in ec ion on seed se ha e been epo ed, depending on he in ec ion le el, hos species and/o di e en habi a s [35], [36]. In addi ion o seed numbe , he o al cos s o e go in ec ion may include addi ional e ec s, like educed weigh , quali y and iabili y o unin ec ed seeds in he in lo escence [32], [37]. We did no es o he e ec s on iabili y o seeds o seedling es ablishmen in his esea ch, bu in s udies wi h wild Spa ina oliosa and cul i a ed Ken ucky blueg ass e go in ec ion caused no e ec on seed ge mina ion, al hough Ken ucky blueg ass seed s o abili y was somewha impai ed [32], [36]. The seemingly mino cos s o e go in ec ion may be due o se e al ac o s: e.g. hos esis ance and ole ance o e go , o e ing he hos ways o con ol and compensa e o he in ec ion. Resis ance can be seen as ac o s limi ing he chance and ex en o E go Pa hogen as De ensi e Mu ualis PLOS ONE | www.plosone.o g 4 July 2013 | Volume 8 | Issue 7 | e69249 in ec ion (e.g., [38], [39]). C. pu pu ea has an excep ionally wide ange o hos s including he en i e sub amily Pooideae [30]. E en hough being o g ea in e es , inding comple ely esis an cul i a s o a ie ies o g ass and ce eal e go hos species has no been e y success ul. No speci ic esis ance genes ha e been de ec ed in ce eal c ops [40], bu gene ic a iabili y in e go esis ance has been ound e.g. in Ken ucky blueg ass and ye [37], [41]. This indica es ha he e has no been s ong selec ion o esis ance o e go in he pooid amily, u he implying ha he ne cos s o e go in ec ion o he hos a e no high in gene al. Ins ead, hos g asses may ha e adap ed he abili y o es ic e go in ec ion o a ole able le el. This is in acco dance wi h ou esul s, as he in ec ed in lo escences mos commonly bo e only one o wo scle o ia. One mechanism o es ic ion could be he escape o in ec ion by asynch onous lowe de elopmen wi hin g ass in lo escences, as i is known o a ec , o example, seed p eda ion [42]. The di e ences o e go esis ance in ce eals indica e he po en ial o limi he size and he esou ces consumed by an e go scle o ium. The de ec ed low le el o e go in ec ion wi hin g ass in lo escences may also al e na i ely o pa ly esul om a ious en i onmen al ac o s such as wea he and a ailable sou ces o in ec ion. Fu he , g asses seem o ha e a high ole ance o e go , o example, he abili y o mi iga e o o se he ad e se i ness e ec s o e go in ec ion (e.g., [38], [39]). Acco ding o Ja oz & Da elos [43], a lo al in ec ion may no di ec ly educe plan ep oduc i e e o because plan s can o en p oduce mo e lowe s han ma u e seeds. Such was he case wi h ou da a, as he seed p oduc ion o Figu e 3.Indi idual sheep eac ions o e go y eed. Single ams’ (A–F) ac ions ela ing o e go -con aining eed on di e en ounds o a ood choice es . The column colou s indica e he p opo ion o ac ions wi hin ials: whi e, e go y eed no app oached; g ey, e go y eed ejec ed; black, e go y eed chosen. doi:10.1371/jou nal.pone.0069249.g003 E go Pa hogen as De ensi e Mu ualis PLOS ONE | www.plosone.o g 5 July 2013 | Volume 8 | Issue 7 | e69249 wild ed escue (when compa ed o he lowe amoun ) was o e all e y low and no all lo e s p oduced seed. Thus, e go may no ac ually always educe seed numbe by aking a place o a seed, and he hos may ha e he po en ial o alloca e esou ces o unin ec ed lowe s ins ead o in ec ed ones. Sal audon e al. [13] ha e specula ed abou ep oduc i e compensa ion, and e en o e compensa ion, o a plan o a p edic able a ack by pa asi es, simila ly o he compensa ion o he bi o y. Wi h e go his could be ia delaying he de elopmen o lo e s and/o using esou ces limi edly in popula ions wi h high p essu e o pa hogen a ack. ‘‘In case’’ o e go in ec ion g asses migh hen be able o alloca e spa ed esou ces o ep oduc ion wi hin, o also be ween in lo escences, since many pe ennial g ass species p oduce se e al lowe ing shoo s a di e en imes du ing a g owing season. Acco dingly, Raybould e al. [35] ha e p oposed ha e go could encou age seed se by causing changes in he hos ’s esou ce alloca ion by inc easing he in es men o lowe heads and seed p oduc ion a he expense o o he plan unc ions. A simila mechanism is sugges ed o cause some posi i e e ec s o closely e go - ela ed Epichloe¨ endophy es on seed p oduc ion du ing some li e s ages o i s hos g asses [44], [45]. The physiological mechanism o his change in esou ce alloca ion may be due o ungal compounds ac ing like plan ho mones, as some ungal endophy es may p oduce auxin-like plan -g ow h egula o s [46]. He bi o y and indi ec bene i s o in ec ion We ound clea di e ences in o e all e go in ec ion equencies be ween mode a ely g azed pas u es and su ounding ung azed a eas. The highe in ec ion incidence in pas u es indica es ha g azing somehow a ec s he p opo ion o e go -in ec ed in lo es- cences. One p obable explana ion could be selec i e g azing p e e ing he unin ec ed g ass in lo escences wi h e go -in ec ed ones le unea en. Documen ing ac ual plan selec ion in he ield by g azing animals is needed o a mo e accu a e con i ma ion o he exac le el o hos ille and gene p o ec ion by e go . Ou ood choice expe imen suppo ed he hypo hesis o selec i e g azing, as he o e all choice agains e go -con aining o age was signi ican . The sheep indi iduals howe e di e ed in hei ac ions, wi h ou ou o six ams clea ly a oiding he e go - con aining o age, and wo always ying o ea om he con aine hey i s app oached. The a iance in beha iou among indi iduals, which is common in choice expe imen s on lea ning beha iou (e.g., [47]), may be due o he indi idual cha ac e s and abili ies, ei he in adjus ing o sho - e m expe imen al se ings o in he abili y o de ec and/o a oid e go . Wi h he la e , he e may also ha e been di e ences in e go - ela ed his o y be ween he ams, e.g. p e iously encoun e ing e go in he pas u e. The animals we e also no allowed o ac ually ea he o ages in his es , so possible a e sion lea ning mus ha e aken place ea lie . Mammalian he bi o es a e mos likely able o dis inguish he p esence o e go . Mammals can de ec and a oid endophy e- in ec ed plan s, e en by hei alkaloid p o ile [48]. This is possibly due o he bi e as e o alkaloids [19], and e go scle o ia can con ain high doses o simila alkaloids [16], [19], [49]. E go is also claimed o ha e a dis inc i e smell de ec able e en by man ( e . in [23]). Mo eo e , he ully de eloped da k-colou ed scle o ia o C. pu pu ea a e usually la ge han seeds and o en cu ed, which makes hem clea ly isible when p o uding om lo e s. Le - Yadun and Halpe n [25] sugges ed ha he colou o e go could e en p o ide a isual aposema ic signal o mammalian he bi- o es. The au ho s also specula ed abou he possibili y o de elopmen o ood a e sion owa ds e go -in ec ed g asses. Inges ion o e go scle o ia in low le els is no le hal, bu can cause symp oms which make de elopmen o ood a e sion possible [25]. Bakau & B yden [50] ha e de ec ed a ian disc imina ion o e go y ood in a longe - e m ood choice s udy in which bi ds had he possibili y o also inges e go . Wi h ou da a, i is impossible o p opose he ac ual mechanism o e go a oidance. Howe e , om a plan ’s pe spec i e, he eason and mechanism o e go -induced a oidance is i ele an . I he p esence o e go nega i ely a ec s he seed p eda ion p obabili y o an in lo escence, he plan gains a p o ec i e e ec . Acco dingly, e go may well p o ide de ence o o he ille s o he hos plan , and e en o neighbou ing plan s. In ou model, we ha e assumed ha e go in ec ion causes di ec loss o seeds, and hence be a (i.e. he ela i e decline in iable seeds pe plan i in ec ed by e go ) can achie e only posi i e alues. The a gumen can be expanded so ha be a may also esul in nega i e alues indica ing ha e go -in ec ed plan s p oduce mo e seeds han non-in ec ed plan s. In such a case e go would be bene icial o he plan wi hou any p o ec i e e ec s. In ac , he bi o es could heo e ically e en be sligh ly a ac ed by e go in ec ed plan s and s ill e go would be bene icial o he hos as a as he loss o seeds due o he g ea e isk o g azing is smalle han he o e all imp o ed seed p oduc ion o e go in ec ed plan s. In his s udy we ound only a e y low cos associa ed wi h e go in ec ion in e ms o seed numbe pe in lo escence, so his scena io migh be possible in some e go - g ass species combina ions. E go – a condi ionally help ul pa hogen In his s udy, we only e e o i ness e ec s on he hos on he ecological ime scale. The alkaloids p oduced by e go , which a e he cause o he hos -acqui ed chemical de ence, ha e p obably e ol ed o p o ec he ungus i sel , and ha e a p o ec i e e ec on he hos as a by-p oduc . Such a case can be de ined as ‘‘by- p oduc mu ualism’’ [7]. Rega ding he e olu iona y his o y o he ela ionship, ou s udy is no su icien o claim ha he in e ac ing species ha e e ol ed special adap a ions o ecei e ‘‘mu ualis ic bene i s’’ om each o he (see [6], [8]). Such a coe ol ed mu ualis ic symbiosis is mos easily de eloped when ansmission o he symbion is e ical (e.g. Epichloe¨ g ass endophy es), whe eas acco ding o some species in e ac ion models in ho izon ally ansmi ed symbion s, like e go , highe i ulence would e ol e mo e easily (e.g., [51]). On he o he hand, mu ualism is commonly de ec ed among ee-li ing species, e.g. among ec omyco hizal ungi and plan s as well as pollina ing insec s and plan s, and hus, e iden ly, e ical ansmission is no a necessa y condi ion o mu ualism o e ol e [52]. Ac ually each Figu e 4. O e all eac ion o sheep o e go y eed. A decision ee ep esen ing he a e age p obabili ies o ams’ (n = 6) app oaches and decisions be ween e go y and con ol o age in a ood choice es . The inal choice signi ies he al e na i e e en ually chosen o ea ing. doi:10.1371/jou nal.pone.0069249.g004 E go Pa hogen as De ensi e Mu ualis PLOS ONE | www.plosone.o g 6 July 2013 | Volume 8 | Issue 7 | e69249 plan is a pa o a complex web o in e ac ing mu ualis s, bo h e ically ansmi ed symbion s and ee-li ing o ganisms, co- e ol ing oge he [53]. Fellous & Sal audon [9] ha e discussed he e olu ion o pa asi ic in e ac ion in o a mu ualis ic ela ionship in ela ion o condi ionally help ul pa asi es. They sugges , ha i he bene icial i ness e ec is s ong, e en he a ely occu ing mu ualis ic si ua ion may selec agains esis ance o he pa asi e in ec ion. Acco ding o Fellous & Sal audon [9], he in e ac ion may e ol e owa ds mu ualis ic symbiosis especially i he pa asi e p o ides a comple ely new ai o unc ion o he hos . E go -g ass in e ac ion demons a es his scena io, since g asses usually do no ha e chemical de ence mechanism agains he bi o es, bu cope wi h he bi o es by ole a ing g azing [19] and wi h weak phy oli h (silica bodies) de ence [54]. Toxic e go p o ides he hos wi h a new ai , he acqui ed chemical de ence, compa ed o unin ec ed g asses. G asses seem no ha e e ol ed s ong esis ance agains e go , which indica es he lack o s aigh o wa d selec ion agains e go in ec ion. In many g assland ecosys ems g asses expe ience ecu en g azing, which is he si ua ion whe e g asses would bene i om any addi ional p o ec ion agains he bi o y. We conclude ha e go in ec ion in ol es di ec cos s o he hos g ass, bu he cos s may no necessa ily be no able in wild g ass popula ions. E go may p o ide he hos wi h indi ec , ecological bene i s ia acqui ed chemical de ence in condi ions we e mammalian he bi o es a e p esen . Thus, we sugges ha he e go -g ass sys em is an example o condi ional pa hogen-hos in e ac ion in which he ou come may all on di e en pa s o a con inuum om pa asi ism o mu ualism depending on he speci ic si ua ion. The in e ac ion may be de ined as condi ional de ensi e mu ualism. The p o ec i e e ec p o ided by e go is concen a ed on ep oduc ion and ep oduc i e issue, in lo escence and he seed se , whe e he i ness e ec s a e mos p onounced. This is he case especially in condi ions whe e la ge he bi o es consume mos o he in lo escences. In ac , he de ensi e e ec o e go has ac ed agains human exploi a ion, as hea ily a ec ed seed and g ain se s ha e been, and s ill a e, disposed o . Ou a gumen could pa ly explain why g asses a e commonly suscep ible and a he ole an o e go and no e ec i e esis ance mechanisms ha e e ol ed o e ime. This migh p o ide new insigh in o plan b eeding p og ams o pooid ce eals, whe e e go esis ance is one o he goals. Suppo ing In o ma ion Appendix S1 Seed p oduc ion o e go in ec ed and unin ec ed ed escue ame s. Loca ion, coo dina es, habi a , collec ion da es and mean p opo ion o seeds in e go -in ec ed and e go - ee in lo escences in each g ass popula ion. (DOCX) Appendix S2 E go incidence in ed escue in lo escences in sheep pas u es and co esponding ung azed a eas. Loca ion, coo dina es, habi a , collec ion yea , e go in ec ion equency and sample size. (DOCX) Acknowledgmen s We would like o hank he s a a he Animal P oduc ion Resea ch o MTT Ag i ood Resea ch Finland o assis ance, Aa on Be gdahl o e ising he language and Simcha Le -Yadun and Simon Fellous o hei aluable commen s on he manusc ip . Au ho Con ibu ions Concei ed and designed he expe imen s: PRW JT PPW KS. Pe o med he expe imen s: PPW PRW. Analyzed he da a: PRW. W o e he pape : PPW PRW JT KS. Re e ences 1. Johnsson NC, G aham JH, Smi h FA (1997) Func ioning o myco hizal associa ions along he mu ualism-pa asi ism con inuum. New Phy ol 135: 575– 585. 2. Saikkonen K, Fae h SH, Helande M, Sulli an T (1998) Fungal endophy es: A con inuum o in e ac ions wi h hos plan s. Annu Re Ecol Sys 29: 319–343. 3. Saikkonen K, Wa¨li PR, Helande M, Fae h SH (2004) E olu ion o endophy e- plan symbioses. T ends Plan Sci 9: 275–280. 4. Leugn TLF, Poulin R (2008) Pa asi ism, commensalism, and mu ualism: explo ing he many shades o symbiosis. Vie Milieu 58: 107–115. 5. Rod iguez Es ada QE, Jonke s W, Kis le HC, May G (2012) In e ac ion be ween Fusa ium e icillioides,Us ilago maydis,andZea mays: An endophy e, a pa hogen, and hei sha ed plan hos . Fung Gene Biol 49: 578–587. 6. Ja¨ emo J, Tuomi J, Nilsson P, Lenna sson T (1999) Plan Adap a ions o he bi o y: mu ualis ic e sus an agonis ic coe olu ion. Oikos 84: 313–320. 7. Ag awal AA (2000) O e compensa ion o plan s in esponse o he bi o y and he by-p oduc bene i s o mu ualism. T ends Plan Sci 5: 309–313. 8. de Mazancou C, Lo eau M, Dieckmann U (2005) Unde s anding mu ualism when he e is adap a ion o he pa ne . J Ecol 93: 305–314. 9. Fellous S, Sal audon L (2008) How can you pa asi es become you allies? T ends Pa asi ol 25: 62–66. 10. Pa ida-Ma ı ´nez LP, Heil M (2011) The mic obe- ee plan : ac o a i ac ? F on Plan Sci 2: 100. 11. Sapp J (2004) The dynamics o symbiosis: an his o ical o e iew. Can J Bo 82: 1046–1056. 12. Cheplick GP, Fae h SH (2009) The ecology and e olu ion o he g ass- endophy e symbiosis. USA: Ox o d Uni e si y P ess. 256 p. 13. Sal audon L, He´ aude V, Shyko JA (2008) A abidopsis haliana and he Robin Hood pa asi e: a chi al ous oomyce e ha s eals i ness om ecund hos s and bene i s he poo es one. Biol Le 4: 526–529. 14. Alexopoulos CJ, Mims CW, Blackwell M (1996) In oduc o y mycology. USA: John Wiley & Sons, Inc. 868 p. 15. Ag ios GN (2005) Plan pa hology. 5 h edi ion. USA: Else ie Academic P ess. 952 p. 16. K ska R, C ews C (2008) Signi icance, chemis y and de e mina ion o e go alkaloids: A e iew. Food Addi Con am A 25: 722–731. 17. Haa mann T, Rolke Y, Giesbe S, Tudzynski P (2009) E go : om wi chc a o bio echnology. Mol Plan Pa hol 10: 563–577. 18. Sco PM (2009) E go alkaloids: ex en o human and animal exposu e. Wo ld Myco oxin J 2:141–149. 19. Clay K (1988) Fungal endophy es o g asses: a de ensi e mu ualism be ween plan s and ungi. Ecology 69: 10–16. 20. Whi e JF, To es MS (2009) In oduc ion: symbiosis, de ensi e mu ualism, and a ia ions on he heme. In: Whi e JF, To es MS, edi o s. De ensi e mu ualism in mic obial symbiosis. USA: CRC P ess. pp. 3–8. 21. Cheplick GP, Clay K (1988) Acqui ed chemical de ences in g asses: he ole o ungal endophy es. Oikos 52: 309–318. 22. an San en E (1995) Animal p e e ence o all escue du ing ep oduc i e g ow h in he sp ing. Ag on J 84: 979–982. 23. Pa be y DG (1996) T ophism and he ecology o ungi associa ed wi h plan s. Biol Re 71: 473–527. 24. Tenbe ge KB (1999) Biology and li e s a egy o he e go ungi. In: K ˇen V, C ak L, edi o s. E go , he genus Cla iceps. Ams e dam: Ha wood Academic Publishe s. pp. 25–56. 25. Le -Yadun S, Halpe n M (2007) E go (Cla iceps pu pu ea) – An aposema ic ungus. Symbiosis 43: 105–108. 26. Tuomi J, Augne M (1993) Syne gis ic selec ion o unpala abili y in plan s. E olu ion 47: 668–672. 27. Ka ban R (1997) Neighbou hood a ec s a plan ’s isk o he bi o y and subsequen success. Ecol En omol 22: 433–439. 28. Rau io P, Be g all UA, Kes i K, Tuomi J, Leima O (2012) Food selec ion by he bi o es and neighbou hood e ec s in he e olu ion o plan de ences. Ann Zool Fenn 49: 45–57. 29. B ons ein JL (1994) Condi ional ou comes in mu ualis ic in e ac ions. T ends Ecol E ol 9: 214–217. 30. Pazˇou o a´S,Olsˇo ska´ J, Linka M, Kolı ´nska´ R, Fliege M. (2000) Chemo aces and habi a specializa ion o Cla iceps pu pu ea popula ions. Appl En i on Mic obiol 66: 5419–5425. 31. G anmo A (1993) The pe ec s a e o Cla iceps pu pu ea in No he n No way. Pola lokken 17: 439–446. 32. Chas ain TG (1992) Rela ionship o e go o Ken ucky blueg ass seed p oduc ion and quali y. J Appl Seed P od 10: 7–10. E go Pa hogen as De ensi e Mu ualis PLOS ONE | www.plosone.o g 7 July 2013 | Volume 8 | Issue 7 | e69249 33. Mikaliu ¯nai e˙ R, Dabke ic ˇius Z (2009) The sp ead o e go (Cla iceps pu pu ea)on Poaceae plan s and incidence on ce eals in Li huania. Zemdi bys e 96: 246–259. 34. Saikkonen K, Saa i S, Helande M (2010) De ensi e mu ualism be ween plan s and endophy ic ungi. Fungal Di e s 41: 101–113. 35. Raybould AF, G ay AJ, Cla ke RT (1998) The long- e m epidemic o Cla iceps pu pu ea on Spa ina anglica in Poole Ha bou : pa e n o in ec ion, e ec s on seed p oduc ion and he ole o Fusa ium he e ospo um. New Phy ol 138: 497–505. 36. Fishe AJ, DiTomaso JM, Go don TR, Aege e BJ, Ay es DR (2007) Sal ma sh Cla iceps pu pu ea in na i e and in aded Spa ina ma shes in No he n Cali o nia. Plan Dis 91: 380–386. 37. Alde man SC, Coa s DD, C owe FJ (1996) Impac o e go on Ken ucky blueg ass g own o seed in No heas e n O egon. Plan Dis 80: 853–855. 38. Roy BA, Ki chne JW (2000) E olu iona y dynamics o pa hogen esis ance and ole ance. E olu ion 54: 51–63. 39. Oli e TH, Lea he SR, Cook JM (2009) Tole ance ai s and he s abili y o mu ualism. Oikos 118: 346–352. 40. Tudzynski P, Sche e J (2004) Cla iceps pu pu ae: molecula aspec s o a unique pa hogenic li es yle. Mol Plan Pa hol 5: 377–388. 41. Mi di a V, Miedane T (2009) Resis ance o e go in sel -incompa ible ge mplasm esou ces o Win e Rye. J Phy opa hol 157: 350–355. 42. E iksson O (1995) Asynch onous lowe ing educes seed p eda ion in he pe ennial o es he b Ac aea spica a. Ac a Oecol 16: 195–203. 43. Ja osz AM, Da elos AL (1995) E ec s o disease in wild plan popula ions and he e olu ion o pa hogen agg essi eness. New Phy ol 129: 371–387. 44. Olejniczak P, Lembicz M (2007) Age-speci ic esponse o he g ass Puccinellia dis ans o he p esence o a ungal endophy e. Oecologia 152: 485–494. 45. Wa¨li PR, Helande M, Nissinen O, Leh onen P, Saikkonen K (2008) Endophy e in ec ion, nu ien s a us o he soil and du a ion o snow co e in luence he pe o mance o meadow escue in sub-a c ic condi ions. G ass Fo age Sci 63: 324–330. 46. Siegel MR, La ch GCM, Johnson MC (1987) Fungal endophy es o g asses. Annu Re Phy opa hol 25: 293–315. 47. De enne A (2010) Shi s in pos disc imina ion g adien s wi hin a s imulus dimension based on bila e al acial symme y. J Exp Anal Beha 93: 485–494. 48. Panaccione DG, Cipole i JR, Sedlock AB, Blemings KP, Scha dl CL, e al. (2006). E ec s o e go alkaloids on ood p e e ence and sa ie y in abbi s, as assessed wi h gene-knockou endophy es in pe ennial Ryeg ass (Lolium pe enne). J Ag ic Food Chem 54: 4582–4587. 49. Uhlig S, Vikø en T, I ano a L, Handeland K (2007) E go alkaloids in No wegian wild g asses: a mass spec ome ic app oach. Rapid Commun Mass Spec om 21: 1651–1660. 50. Bakau WJK, B yden WL (1998) E go ism and eed a e sion in poul y. In: Ga land T, Ba AC, edi o s. Toxic Plan s and O he Na u al Toxican s. Walling o d: CAB In e na ional. pp. 493–496. 51. Law R, Lewis DH (1983) Bio ic en i onmen s and he main enance o sex – some e idence om mu ualis ic symbioses. Biol J Linn Soc 20: 249–276. 52. Wilkinson DM & She a TN (2001) Ho izon ally acqui ed mu ualisms, and unsol ed p oblem in ecology? Oikos 92: 377–384. 53. Thompson JN (2006) Mu ualis ic webs o species. Science 312: 372–373. 54. Massey FB & Ha ley SE (2006) Expe imen al demons a ion o he an ihe bi o e e ec s o silica in g asses: impac s on oliage diges ibili y and ole g ow h a es. P oc R Soc B 273: 2299–2304. E go Pa hogen as De ensi e Mu ualis PLOS ONE | www.plosone.o g 8 July 2013 | Volume 8 | Issue 7 | e69249