Fungal-mediated multitrophic interactions - do grass endophytes in diet protect voles from predators?

Fungal-Media ed Mul i ophic In e ac ions - Do G ass Endophy es in Die P o ec Voles om P eda o s? Susanna Saa i 1,2 *, Janne Sundell 3,6 , O so Hui u 4 , Ma jo Helande 2 , Elise Ke oja 5 , Hannu Ylo ¨nen 3 , Ka i Saikkonen 1 1Plan P oduc ion Resea ch, MTT Ag i ood Resea ch Finland, Jokioinen, Finland, 2Depa men o Biology, Uni e si y o Tu ku, Tu ku, Finland, 3Depa men o Biological and En i onmen al Science, Konne esi Resea ch S a ion, Uni e si y o Jy a ¨skyla ¨, Konne esi, Finland, 4Suonenjoki Resea ch Uni , Finnish Fo es Resea ch Ins i u e, Suonenjoki, Finland, 5Se ices Uni , MTT Ag i ood Resea ch Finland, Jokioinen, Finland, 6Lammi Biological S a ion, Uni e si y o Helsinki, Lammi, Finland Abs ac Plan -associa ed mic o-o ganisms such as myco oxin-p oducing endophy es commonly ha e di ec nega i e e ec s on he bi o es. These e ec s may be ca ied o e o na u al enemies o he he bi o es, bu his has been a ely explo ed. We examined how eeding on Neo yphodium endophy e in ec ed (E+) and endophy e ee (E2) meadow yeg ass (Sche odonus p a ensis) a ec s body mass, popula ion size and mobili y o sibling oles (Mic o us le is), and whe he he die media es he ulne abili y o oles o leas weasel (Mus ela ni alis ni alis) p eda ion. Because leas weasels a e known o be ol ac o y hun e s, we also examined whe he hey a e able o dis inguish ol ac o y cues o oles ed on E+and E2die s. Nei he body mass o oles no popula ion size di e ed be ween die s. Howe e , con a y o ou p edic ion, leas weasels p eyed mo e o en on oles ed wi h E2g ass han on oles ed wi h E+g ass. The mobili y o oles ed on E+g ass was educed compa ed o oles ed on E2g ass, bu his e ec was un ela ed o isk o p eda ion. Leas weasels appea ed unable o dis inguish be ween exc emen odou s o oles be ween he wo ea men s. Ou esul s sugges ha consump ion o endophy ic g ass is no di ec ly dele e ious o sibling oles. Wha ’s mo e, consump ion o endophy es appea s o be ad an ageous o oles by educing isk o mammalian p eda ion. Ou s udy is hus he i s o demons a e an e ec o plan -associa ed mic obial symbion s on he bi o e-p eda o in e ac ions in e eb a e communi ies. Ci a ion: Saa i S, Sundell J, Hui u O, Helande M, Ke oja E, e al. (2010) Fungal-Media ed Mul i ophic In e ac ions - Do G ass Endophy es in Die P o ec Voles om P eda o s? PLoS ONE 5(3): e9845. doi:10.1371/jou nal.pone.0009845 Edi o : Jon Moen, Umea Uni e si y, Sweden Recei ed No embe 9, 2009; Accep ed Ma ch 1, 2010; Published Ma ch 24, 2010 Copy igh : ß2010 Saa i 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 is pa o Endophy ic Fungi in Ecological Plan P oduc ion P og amme no. 110658, unded by he Academy o Finland, Niemi Founda ion, Ma ja a and Eino Kolli Founda ion and Ol i Founda ion. 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 . h p://www.aka. i/ i/A/, h p://www.ol isaa io. i/web/ol isaa io/, h ps://www.kollinsaa io.com/. The e is no URL o Niemi Founda ion. 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 p o ec ed] In oduc ion Al hough mic obial in e ac ions wi hin and be ween ophic le els a e ubiqui ous, hei oles ha e emained la gely igno ed in communi y scale s udies [1]. Mic obial symbion s ha e p o ound pheno ypic e ec s on hei hos s which may cascade upwa d h ough ood webs [2]. Fo ins ance, plan s hos mul iple symbio ic mic obes, including myco hizal ungi, endophy ic ungi and bac e ia ha may a ec he pe o mance o he bi o es h ough symbion -p oduced oxins which accumula e in he bi o e issue and he eby di ec ly ha m p eda o s ollowing inges ion o he p ey. In addi ion, e ec s on highe ophic le els may mani es , e.g., ia changes in he bi o e densi ies, popula ion dynamics, body size o beha iou o he bi o es [3], [4]. One g oup o mic obial symbion s ha a e known o a ec mul i ophic in e ac ions a e ungal g ass-endophy es in he genus Neo yphodium. They a e known o in ec 20–30% o all g ass species [5], o ming sys emic and asymp oma ic in ec ions h oughou he ae ial pa s o he hos plan , including he seeds, and he eby allowing e ical dispe sal o he endophy e om one plan gene a ion o ano he [6]. G ass-endophy es may ha e mul i a ious e ec s on he bi o e communi ies [7]. Consump ion o Neo yphodium endophy e o igin myco oxins has long been known o cause se e e li es ock diso de s, including symp oms anging om embling o s agge ing and se e e muscle spasms ha cause animals o collapse [8], [9]. Also smalle e eb a e he bi o es, such as oden s, a e commonly nega i ely a ec ed by endophy e inges ion. These e ec s include, e.g., dec ease in popula ion densi y [10], lowe ed body mass [11], inc eased oxici y-induced mo ali y [12] and supp ession o ep oduc ion and g ow h [13]. The alkaloids p oduced by endophy es may also ha e nega i e e ec s on he na u al enemies o in e eb a e he bi o es [14], [15]. Howe e , expe imen al s udies on endophy es and hei e ec s on highe ophic le els a e s ill sca ce [2], [16], and no s udy has examined he cascading e ec s o ungal symbion s o g asses on e eb a e ood chains. He e, we examined how eeding on endophy e (Neo yphodium uncina um (Gains, Pe ini & Schmid ) Glenn, Bacon, P ice & Hanlin) in ec ed (E+) o endophy e ee (E2) meadow yeg ass (Sche odonus p a ensis (Huds.) P. Beau ex. Lolium p a ense) a ec s body mass and popula ion size o sibling oles (Mic o us le is Mille ex M. ossiaeme idionalis Ogne ) and whe he he E+die in luences he ulne abili y o oles o p eda ion by hei mos impo an na u al enemy, he leas weasel (Mus ela ni alis ni alis L.). Because leas weasels a e known o be ol ac o y hun e s [17], we also examined whe he leas weasels a e able o dis inguish ol ac o y PLoS ONE | www.plosone.o g 1 Ma ch 2010 | Volume 5 | Issue 3 | e9845 cues o oles ed on E+and E2die s. Based on p e ious s udies we p edic ed ha i) consump ion o myco oxic E+g ass has di ec nega i e e ec s on oles and ha ii) hese e ec s in luence he ulne abili y o oles o hei mos impo an na u al enemy, he leas weasel (Mus ela ni alis ni alis L.). Fu he mo e, i endophy e- con aining die a ec s he chemical composi ion o ole u ine, as demons a ed by Hui u e al. (2008), we p edic ha iii) ol ac o y hun ing leas weasels migh be able o disc imina e and p e e he scen o weake ole p ey ed on E+g ass om hose ed on E2 g ass. Resul s Consump ion o E+g ass did no ha e nega i e e ec s on ole popula ion size o body mass. In he ole popula ion expe imen he es ima ed median di e ence in he minimum numbe o oles ali e was only 0.5 indi iduals (95% CI o di e ence = 26–+7; p = 1.0), in a ou o he E2g ass popula ions (means and s anda d de ia ions, oles ali e: E+13.269.07, E21365.15) In he ole body biomass expe imen , emales ed on E+g ass had on a e age 0.2 g (95% CI o di e ence = 21.0 – +0.7; p = 0.66) lowe and males 0.3 g (95% CI = 20.4 – +1.1; p = 0.35) highe body mass compa ed o oles ed on E2g ass (means and s anda d de ia ions, ole body mass: E+ emale 21.3263.10, E2 emale 21.4962.75, E+male 25.6663.64 and E2male 25.3163.66). Voles main ained on E+g ass exhibi ed lowe mobili y in he p eda ion expe imen han E2 oles (es ima ed median di e ence in he ac i i y o oles = 13.3 bel c ossings/ hou in a ou o E2 oles, 95% CI = 2.7–35.4; Fig. 1). Leas weasels we e mo e likely o cap u e oles ed on E2g ass han oles ed on E+g ass (14 oles ed on E2cap u ed e sus h ee oles ed on E+cap u ed in he 17 success ul ials; Fig. 2). Howe e , ac i i y o oles did no explain he suscep ibili y o oles o leas weasel p eda ion, no did ole sex, leng h o he eeding pe iod o body mass di e ence be ween he oles ed on E+and E2die s in he beginning o he expe imen (Table 1). Leas weasels appea ed unable o dis inguish be ween he ol ac o y cues o oles ed on he wo g ass ypes; 10 leas weasels chose he bedding o a ole ed on E+g ass and 11 chose he bedding o a ole ed on E2g ass (Fig. 2). Discussion Al hough E+g asses a e commonly hough o be chemically p o ec ed agains he bi o es [7], [16], ou esul s wi h sibling oles and meadow escue do no suppo his no ion unanimously. Con a y o ou p edic ions, we did no ind nega i e e ec s o an E+die ei he on popula ion size o body mass o sibling oles. I appea s ob ious ha Neo yphodium-in ec ed meadow yeg ass is no pa icula ly oxic o sibling oles, a leas wi hin a ime ame o a ew mon hs. Endophy e-in ec ed meadow yeg ass has p e iously been shown o dec ease body mass in a closely ela ed ole species, he ield ole (Mic o us ag es is L.) in labo a o y condi ions [11]. I is he e o e plausible ha ole ance o loline myco oxins a ies among di e en e eb a e he bi o e species. The disc epancy be ween his and ea lie s udies may also s em om a ia ion in myco oxin p oduc ion, which is known o be dependen on en i onmen al condi ions [7], [16]. Con a y o ou p edic ions, leas weasels p eyed mo e o en on oles ed wi h E2g ass han on oles ed wi h E+g ass. Voles ha had consumed E2g ass we e also mo e mobile han oles ha had consumed E+g ass. Al hough high mobili y is o en associa ed wi h inc eased p eda ion isk [18], he deg ee o ole mobili y was un ela ed o he p ey selec ion beha iou o leas weasels in ou expe imen . Reduced mobili y migh be expec ed i myco oxins had educed he physiological well-being o oles o a poin o apa hy. Howe e , his is no plausible in he ligh o ou expe imen s, as oles did no lose body mass o show educed popula ion g ow h. Voles exhibi an a ay o beha iou s in hei a oidance o p eda o s. Many o hese a e ela ed o mobili y, o example leeing and eezing [19]. The la e beha iou was equen ly Figu e 1. Mobili y o oles. Di e ences in mobili y o sibling oles as de ined by numbe s o bel c ossings pe hou o oles ed on endophy e in ec ed (E+) o endophy e ee (E2) g asses. The uppe and lowe bounda ies o he box indica e he uppe and lowe qua iles, espec i ely. The ho izon al line deno es he median. Ve ical ba s ep esen he ails o he dis ibu ion. Medians o he g oups a e connec ed wi h do ed line. Filled ci cles ep esen mean alues. Mild ou lie s a e ma ked wi h open ci cles. The numbe o eplica es is 24. doi:10.1371/jou nal.pone.0009845.g001 G ass Endophy es in Food Chain PLoS ONE | www.plosone.o g 2 Ma ch 2010 | Volume 5 | Issue 3 | e9845 obse ed in encoun e s be ween E+ oles and weasels, so i is plausible ha educed mobili y was ela ed o eezing unde p eda ion isk. Fu he mo e, as weasels did no di e en ia e be ween odo s o oles main ained on he di e en die s, we ega d di e ences in he oles’ an ip eda o y beha iou he mos pa simonious explana ion o he obse ed pa e ns in weasel p ey selec ion. Howe e , his easoning is indica i e a bes , as un o una ely speci ics o ole a oidance beha iou we e no eco ded. We also canno conclude how ole mobili y o e all, ega dless o ea men , was a ec ed by he p esence o p eda o s, since mobili y was no measu ed in he absence o p eda o s in he sys em. Fu he mo e, eezing may ha e a ec odou compounds in bedding and we a e indeed unawa e o whe he he oxic compounds we e ansmi ed o he u ine. Conclusions In ou s udy we we e able o demons a e indi ec posi i e e ec s o mic obial plan symbion s on a e eb a e he bi o e ia educed p eda ion. Simila e ec s ha e been p e iously demon- s a ed wi h in e eb a e p eda o s and pa asi es as na u al enemies in ood webs whe e he bi o es eed on endophy e in ec ed plan s [14], [15], [20]. In cases whe e species can ole a e myco oxins p oduced by he endophy e and a e less a isk o p eda ion due o endophy e consump ion, he ne e ec o he endophy e on he hos g ass will be nega i e. The e o e he adi ional iew o endophy es as de ensi e plan mu ualis s may no hold i a hi d ophic le el is included. Thus, ou esul s p o ide e idence ha he na u e o he ela ionship be ween g ass endophy es and hei hos s may ange om mu ualism o pa asi ism depending on he complexi y o he ood web. Ma e ials and Me hods E hics s a emen All p ocedu es in ol ing oles we e ca ied ou in acco dance wi h he Ac on he Use o Animals o Expe imen al Pu poses Figu e 2. P eda ion and ole odou p e e ence o weasel. The e ec s o endophy e in ec ed (E+) and endophy e ee (E2) g ass die s on p eda ion and ole odou p e e ence o leas weasel. Es ima ed pe cen ages o cap u ed oles (n = 17) and odou p e e ence (n = 21). doi:10.1371/jou nal.pone.0009845.g002 Table 1. Resul s o i ing logis ic eg ession models o he da a o leas weasel p eda ion. Explana o y a iable Coe icien ^ bb P- alue o H 0 :b~~0Odds Ra io ORðÞ* ) 95% CI o OR{ ) Di e ence be ween E- and E+ oles in he numbe o bel c oss- ings pe hou ( = ac i i y E2 2ac i i y E+ ) 0.01 0.25 1.01 0.99–1.04 Sex o ole 20.60 0.65 0.55 0.00–6.04 Leng h o he eeding pe iod (days) 0.14 0.19 1.15 0.95–1.40 Di e ence be ween E2and E+ oles in weigh ( = weigh E2 2weigh E+ ) 0.12 0.81 1.13 0.42–3.22 *)The es ima ed odds a io OR~exp ^ bb  :Fo quan i a i e explana o y a iables 100 OR{1ðÞindica es he pe cen change in he odds o E+ ole being cap u ed o each 1-uni inc ease in he explana o y a iable. Fo example, o e e y one day inc ease in he leng h o he eeding pe iod, he odds o E+ ole being cap u ed inc eases by 15%. Fo ca ego ical a iable sex he OR o 0.55 implies ha o emales he odds o E+ ole being cap u ed is 0.55 imes he odds o males, i.e. 45% lowe . {)The 95% con idence in e al o he OR indica es he ange o alues wi hin which he odds a io om 95 o 100 simila s udies would be expec ed o all. The 95% CI also indica es he p ecision o he es ima ed OR. doi:10.1371/jou nal.pone.0009845. 001 G ass Endophy es in Food Chain PLoS ONE | www.plosone.o g 3 Ma ch 2010 | Volume 5 | Issue 3 | e9845 es ablished by he Minis y o Ag icul u e and Fo es y, Finland. The s udy was app o ed and supe ised by he Animal Expe imen Commi ee o Finland (License numbe : STH393A). Species Meadow yeg ass, he expe imen al plan species, is one o he mos impo an o age g asses in Finland. I is a na i e g ass species in Eu ope which occu s commonly ou side o ag onomic use in meadows, oadsides and was elands in Finland [21]. Se e al widely used meadow yeg ass cul i a s in Finland a e commonly in ec ed Neo yphodium uncina um endophy e [22], which g ows sys emically in all pa s o he hos plan . N. uncina um p oduces lolines which may cause a iable esponses in in e eb a es and small e eb a es [7], [11] bu he loline appea s o be non- oxic o la ge mammal he bi o es [23]. The sibling ole is a common and widely dis ibu ed species in sou he n and wes e n Finland. The indi iduals used in he expe imen we e labo a o y-bo n indi iduals, whose pa en s we e apped om na u al popula ions in nea by ields o MTT Ag i ood Resea ch Finland, Jokioinen (60u489150N, 23u299 100E), in au umn 2006. P io o he expe imen , oles we e housed in ca. 60640640 cm 3 cages (3–5 same sex indi iduals in a cage) and p o ided wi h ad libi um po a oes, wa e and wice a week wi h endophy e ee esh g ass cu om he wild. Bedding was p o ided in he o m o wood sha ings and hay. Tempe a u e in he labo a o y was ca. 20uC and pho ope iod 16 h ligh : 8 h da k. The leas weasel is a common specialis p eda o o oles and hei single mos impo an sou ce o mo ali y in na u al popula ions [24], [25]. The leas weasels used in he expe imen we e ei he i s gene a ion labo a o y-bo n indi iduals om he Konne esi Resea ch S a ion o he Uni e si y o Jy a¨skyla¨, Finland, o apped om he wild bu kep unde simila condi ions like lab-bo n ones o weeks be o e he expe imen . P io o he p eda ion and ol ac o y expe imen s, he leas weasels we e housed indi idually in 60680660 cm 3 cages in an ou doo shel e and p o ided wi h a oos e chick pe day wi h one as ing day a week, and occasionally oles o he gene a Mic o us and Myodes. Bedding was p o ided in he o m o wood sha ings and hay. Vole popula ion size The expe imen al ield was es ablished o s udy he impo ance o endophy es on he popula ion de elopmen o oles. Seeds (cul i a ‘Kaspe ’) we e ob ained in a ious seed lo s om seed p oduc ion a ms ia he Plan P oduc ion Inspec ion Cen e, Loimaa, Finland. As samples, we s ained 50 seeds pe seed lo and examined hem mic oscopically o endophy e s a us [26]. We chose wo seed lo s o he expe imen : one unin ec ed (E2,0% endophy e equency) and he o he in ec ed (E+, 79% endophy e equency). We sowed E+and E2seeds in a ield in i e plo pai s (each plo 39625 m 2 ) so ha we andomized E+and E2 ea men s sepa a ely wi hin each pai . The ield was es ablished in Jokioinen in May 2006. Each plo was su ounded wi h a shee me al ence in o de o keep he expe imen al oles inside and oles o na u al popula ions and small mammal p eda o s ou o he expe imen al a eas. The shee me al was embedded 60 cm below g ound while 60 cm emained abo e g ound. Be o e sowing, he ield was e ilized wi h cow manu e (30 000 kg/ha) and again in June 2007 wi h a comme cial e ilize [16:9:22 (N : P : K) wi h mic onu ien s, Kemi a, p oduc numbe : 0647334]. All o he plan s excep o meadow yeg ass we e egula ly oo ed up om he ield du ing he expe imen . Endophy e in ec ion s a us o nine plan indi iduals pe ield plo was e i ied be o e he expe imen using immunoplo assay o de ec monoclonal an ibodies speci ic o Neo yphodium (Phy osc een Immunoplo Ki #ENDO7973, Ag inos ics, Wa kins ille, Geo - gia, USA). Alkaloid ex ac s o he plan s we e analysed [27], [28] and E+plan s we e de ec ed o ha bo ac i e endophy e in ec ions p oducing loline alkaloids. Vole indi iduals (50 males and 50 emales) we e andomly selec ed om a la ge pool o males and emales, all o which we e sexually ma u e and had a body mass o .20 g. Indi iduals we e andomly assigned ei he o an E+o an E2 ea men . The ea e , in Augus , i e male and i e emale oles we e eleased in o each o he en enclosu es. Fou and a hal mon hs la e , which app oxima es he annual leng h o he ep oduc i e pe iod o oles in Finland, ole popula ion sizes we e es ima ed using Ugglan mul iple li e cap u e aps (G ahnab, Sweden). Fi e aps we e placed in each enclosu e unde plywood shel e boxes o educe exposu e o he elemen s. The aps we e bai ed wi h ca o s and checked wice a day o h ee days. Popula ion size was es ima ed o each enclosu e wi h he minimum numbe ali e –me hod [29]. Vole body mass Voles (72 indi iduals) we e selec ed om a la ge pool o males and emales, all o which had a body mass o .16 g, and housed singly in cages. The selec ed oles we e assigned in o 36 pai s (15 emale and 21 male pai s) based on simila i y in body mass. A his s age, oles we e p o ided ad libi um po a o and wa e . On he ollowing day, we assigned he ole indi iduals andomly ei he o an E+o an E2die ea men wi hin each pai and eco ded hei body mass o he nea es 0.1 g wi h elec onic scales. The mean di e ences in he body mass (body mass E2 2body mass E+ ) o emales and males (and s anda d de ia ions o he di e ences) we e 0.01 g (0.73 g) and 0.03 g (0.44 g), espec i ely. A e his, we emo ed po a oes om he cages and p o ided he co e- sponding expe imen al die s, namely ad libi um esh meadow yeg ass h ee imes a day cu om espec i e E+o E2plo s in he ield (see ‘‘Vole popula ion size’’). The body mass o he oles was again eco ded when he oles had been main ained on he expe imen al die o se en days. P ey p e e ence o weasel P io o he p eda ion expe imen six emale and 18 male ole pai s (one ed on E+g ass and one ed on E2g ass wi hin each pai ) we e ed wi h he expe imen al die s as in he body mass bioassay o 7–30 days. The leng h o he eeding pe iod a ied be ween ole pai s due o logis ical easons bu was eco ded o use a co a ia e in he analyses. The expe imen was conduc ed a he Konne esi Resea ch S a ion o he Uni e si y o Jy a¨skyla¨ (62u379400N, 26u179150 E), Finland, in an expe imen al enclosu e (10610 m 2 ) on a ield na u ally ege a ed by meadow plan species ( he ield did no include meadow yeg ass). The enclosu e was di ided in o six sec o s sepa a ed by ca. 30 cm wide sho -g ass bel s (cu 2 cm abo e g ound su ace) o enable moni o ing o ole mobili y. One expe imen al ial consis ed o exposing bo h a ole ed on E+and a ole ed on E2g ass om he same eeding pai o leas weasel p eda ion in he enclosu e. Thus, a eplica e consis ed o a pai o oles ha we e bo h he same sex, simila in weigh in he beginning o he expe imen and had been on he expe imen al die o he same pe iod o ime. The oles we e ma ked wi h ibe s ips o di e en colo s (15 cm long, 2 cm wide) bound a ound he pel is, whe e i does no hinde mo emen s, and eleased in o he cu middle bel be ween he sec o s o he enclosu e. We andomized he s ip colo be ween oles ed on E+and E2g ass in e e y eplica e. Simul aneously wi h eleasing oles, we placed a G ass Endophy es in Food Chain PLoS ONE | www.plosone.o g 4 Ma ch 2010 | Volume 5 | Issue 3 | e9845 leas weasel in an Ugglan li e ap in he same middle bel whe e he oles we e eleased. The leas weasel was eleased om he ap i e minu es a e eleasing he oles. We obse ed and eco ded he numbe o bel c ossings o he oles in he enclosu e un il one o he oles was cap u ed by a leas weasel. A single ial las ed om 15 min o 9 hou s. Se en ou o he 24 ials had o be e mina ed o ex e nal easons (da kness a nigh , hea y ain e c.) be o e he weasel had cap u ed a ole. Di e en oles and leas weasels we e used in e e y eplica e. A e each ial we apped he su i ing ole (o bo h oles i a eplica e had o be e mina ed) and emo ed hem om he enclosu e back o he labo a o y. Odou p e e ence o weasel We s udied whe he leas weasels, ha a e known o be ol ac o y hun e s [19], dis inguish and p e e ol ac o y cues o oles ed wi h E+g ass om oles ed wi h E2g ass. As a sou ce o odou we used u ine- and aeces-soaked ole bedding ma e ial om cages o oles ha had been eeding on ei he E2o E+g ass o he weasel p eda ion expe imen s (see abo e). The collec ed bedding ma e ial was s o ed in ai - igh plas ic bags a 222uC o ca. h ee mon hs. The bedding was hawed in he bags a oom empe a u e be o e he expe imen . The ole odou expe imen was ca ied ou in a Y-maze a ena [19], which consis ed o h ee anspa en Pe spex (Pe spex, Ro e dam, The Nea he lands) plas ic ubes o 80 mm inside diame e and 80, 60 and 60 cm leng h, o ming a Y. The weasel en e ed he 80 cm long ube and came o a bi u ca ion o ,60u. The e i had o choose o con inue in o ei he o he 60 cm ubes un il i eached a a ge ‘‘nes box’’ con aining ole odou a he end o he ube. The nes boxes we e small labo a o y cages (25610610 cm 3 ) co e ed wi h a Pe spex oo . The ends o he Pe spex ubes we e sepa a ed om he nes box by a Pe spex doo wi h 12 holes o 4-mm diame e o allow ai low. In each nes box we placed a 2610610 cm 3 wi e mesh baske illed wi h ole bedding om ei he he E+o he E2 ea men . We andomized E+and E2bedding om he same ole eeding pai o ei he end o he Y-maze p io each ial. Thus, we collec ed he E+and E2bedding in each eplica e om oles ha had been on he expe imen al die o he same amoun o ime. The en ance o he ube was a wooden box, in which we acclima ized he weasel o i e minu es p io o he ial. The box was sepa a ed om he ube by a Pe spex doo ha could be opened om ou side he expe imen a ion oom wi h a mono il- amen line. The doo had holes o allow ai low om he a ena o he weasel box du ing he acclima iza ion pe iod. Abo e he a ena we moun ed in a ed ligh sou ces and a came a ha was connec ed o a moni o in an adjacen oom whe e he beha io o he weasel was moni o ed on sc een. Al oge he , we es ed nine emale and 12 male weasels. All weasels slowly app oached he ube and selec ed one o he ubes a he bi u ca ion. The es ended when he weasel eached he end o ei he b anch o he Y- ube and sni ed he holes a he doo sepa a ing he ube om he nes box. A e each ial we cleaned all pa s o he a ena wi h wa e and ca. 50% e hanol. S a is ical me hods The esponse a iables used in he s a is ical analyses and o he de ails o he expe imen s a e summa ized in Table 2. We analyzed he body mass o oles sepa a ely o emales and males because a ia ion in male body mass was highe han in emales. The E+and E2die ea men s we e compa ed using a pai ed - es and 95% con idence in e al (CI) o he mean di e ence [30]. The analyses we e pe o med by he MIXED p ocedu e in e sion 9.1.3 o he SAS/STAT so wa e. The es o he da a we e analyzed h ough exac s a is ical me hods because he da a we e small and non-no mally dis ibu ed. We based he s a is ical analyses o leas weasel p eda ion on logis ic eg ession models o bina y da a. Le Yi~1i E+ ole was cap u ed and Yi~0i E2 ole was cap u ed in pai i. Each bina y ou come a iable Yiis assumed o ha e a Be noulli dis ibu ion wi h pa ame e pi, whe e piis he p obabili y o E+ ole being cap u ed in a pai i. The alue pi~0:5indica es an equal p obabili y o cap u e o E+and E2 oles. When examining whe he leas weasels p e e E+ oles o e E2 oles as p ey, he logis ic eg ession model had he ollowing o m: log pi=1{pi ðÞ½~að1Þ whe e ais a cons an and he a io pi=1{pi ðÞis he odds o E+ Table 2. Summa y o he de ails o he expe imen s. Expe imen Numbe o pai s Numbe o oles Males Females C i e ia o pai ing Response a iable in he s a is ical analysis Vole popula ion size 5 25 25 Same sex, app oxima ely he same ini ial body mass. Di e ence in he numbe o oles be ween E+and E2in a enclosu e pai a e ou and a hal mon hs. Vole body mass 36 42 30 Same sex, app oxima ely he same ini ial body mass. Di e ence in body mass (g) be ween E+and E2in a ole pai a e eeding he oles se en days. P ey p e e ence o weasel 24 36 12 Same sex and leng h o he eeding pe iod, app oxima ely he same ini ial body mass. Di e ence in he numbe o bel c ossings pe hou be ween E+ and E2in a ole pai P ey p e e ence o weasel 17 *) 24 10 Same sex and leng h o he eeding pe iod, app oxima ely he same ini ial body mass. E+ ole cap u ed in a pai o oles ed on E+and E2(yes, no) Odou p e e ence o weasel 21 *) 42 0 Same sex and leng h o he eeding pe iod, app oxima ely he same ini ial body mass. Bedding o E+ ole chosen in a E+and E2bedding pai (yes, no) *)Also he numbe o leas weasels. doi:10.1371/jou nal.pone.0009845. 002 G ass Endophy es in Food Chain PLoS ONE | www.plosone.o g 5 Ma ch 2010 | Volume 5 | Issue 3 | e9845 ole being cap u ed in a pai i. On i ing he model o he da a, an es ima e o a,^ aa, is ob ained, and he es ima ed cap u e p obabili y (p opo ion) ^ pp€ ii ~exp ^ aaðÞ=1zexp ^ aaðÞ g . The 95% con idence limi s o piwe e compu ed simila ly om he es ima ed limi s o a. When examining he dependence o pion he alues xio each po en ial explana o y a iable (one a a ime), he model was o he ollowing o m: log pi=1{pi ðÞ½~azbxið2Þ whe e aand bwe e he unknown pa ame e s es ima ed by he da a. We i ed he models by using he app oach o condi ional exac in e ence [31], and pe o med he analyses wi h e sion 8 o he LogXac so wa e. The analysis o he bina y choice da a o ole odou was based on he co esponding model han he model (1) abo e. We es ed he di e ence in he amoun o cap u ed oles pe enclosu e be ween E+and E2g asses and also he di e ence in he bel c ossings o oles/hou by using he exac Wilcoxon signed anks es , and es ima ed he median di e ence and i s 95% con idence in e al (CI) by he Hodges-Lehmann p ocedu e [32]. We pe o med he analysis wi h e sion 8 o he S a Xac so wa e. Acknowledgmen s We hank Pa¨i i Leh onen, Maija Sinisalo, Tanja Jyla¨nki, Lenka T eba icka and Ma ko Haapakoski o hei ield assis ance and Ag i ood Resea ch Finland Jokioinen and Konne esi Resea ch S a ion o use o labo a o y and ou doo enclosu e acili ies o he expe imen s. Au ho Con ibu ions Concei ed and designed he expe imen s: SS JS OH MH HY KS. Pe o med he expe imen s: SS JS. Analyzed he da a: EK. Con ibu ed eagen s/ma e ials/analysis ools: SS JS HY. W o e he pape : SS JS OH MH HY KS. Re e ences 1. Ag awal AA, Acke ly DD, Adle F, A nold AE, Cace es C, e al. (2007) Filling key gaps in popula ion and communi y ecology. F on Ecol En i on 5: 145–152. 2. Cheplick GP, Fae h S (2009) Ecology and e olu ion o he g ass-endophy e symbiosis. USA: Ox o d Uni e si y P ess. 256 p. 3. Woo on JT (1993) Indi ec e ec s and habi a use in an in e idal communi y: in e ac ion chains and in e ac ion modi ica ions. Am Na 141: 71–89. 4. an Veen FJF, Mo is RJ, God ay HCJ (2006) Appa en compe i ion, quan i a i e ood webs, and he s uc u e o phy ophagous insec communi ies. Annu Re En omol 51: 187–208. 5. Leuch mann A (1992) Sys ema ics, dis ibu ion, and hos speci ici y o g ass endophy es. Na Tox 1: 150–162. 6. Clay K (1990) Fungal endophy es o g asses. Annu Re Ecol Sys 21: 275–297. 7. Saikkonen K, Leh onen P, Helande M, Ko iche a J, Fae h SH (2006) Model sys ems in ecology: dissec ing he endophy e-g ass li e a u e. T ends Plan Sci 11: 428–433. 8. Bacon CW, Po e JK, Robbins JD, Lu ell ES (1977) Epichloe¨ yphina om oxic all escue g asses. Appl and En i on Mic obiol 34: 576–581. 9. Ho eland CS (1993) Impo ance and economic signi icance o he Ac emonium endophy es o pe o mance o animals and g ass plan s. Ag ic Ecosys En i on 44: 3–12. 10. Coley AB, F ibou gh HA, Pel on MR, Gwinn KD (1995) E ec s o all escue endophy e in es a ion on ela i e abundance o small mammals. J En i on Qual 24: 1044–1044. 11. Hui u O, Helande M, Leh onen P, Saikkonen K (2008) Consump ion o g ass endophy es al e s he ul a iole spec um o ole u ine. Oecologia 156: 333–340. 12. Cono e MR (1998) Impac o consuming all escue lea es wi h he endophy ic ungus, Ac emonium coenophialum, on meadow oles. J Mammal 79: 457–463. 13. Du ham WF, Tannenbaum MG (1998) E ec s o endophy e consump ion on ood in ake, g ow h, and ep oduc ion in p ai ie oles. Can J Zool 76: 960–969. 14. Ba ke GM, Addison PJ (1996) In luence o cla icipi aceous endophy e in ec ion in yeg ass on de elopmen o he pa asi oid Mic oc onus hype oae Loan (Hymenoph e a: B aconidae) in Lis onou us bona iensis (Kusche ) (Coleop e a: Cu culionidae). Biol Con ol 7: 281–287. 15. Ha¨ i SA, K auss J, Mu¨lle CB (2008) T ophic cascades ini ia ed by ungal plan endosymbion s impai ep oduc i e pe o mance o p asi oids in he second gene a ion. Oecologia 157: 399–407. 16. Saikkonen K, Saa i S, Helande M (2010) De ensi e mu ualism be ween plan s and endophy ic ungi? Fungal Di e si y (Doi: 10.1007/s13225-010-0023-7). 17. Ylo¨nen H, Sundell J, Tiilikainen R, Ecca d JA, Ho ne T (2003) Weasels’ (Mus ela ni alis ni alis) p e e ence o ol ac o y cues o he ole (Cle h ionomys gla eolus). Ecology 84: 1447–1452. 18. Banks PB, No dahl K, Ko pimaki E (2000) Nonlinea i y in he p eda ion isk o p ey mobili y. P oc R Soc Lond B 267: 1621–1625. 19. Sundell J, Hy o¨nen H (2004) Beha iou and choise o e uge by oles unde p eda ion isk. Beha Ecol Sociobiol 56: 263–269. 20. Bul man TL, McNeill MR, Goldson SL (2003) Isola e-dependen impac s o ungal endophy es in a mul i ophic in e ac ion. Oikos 102: 491–496. 21. Ha¨me -Ah i L, Suominen J, Ul inen T, Uo ila P (1988) Re keilykas io (Field lo a in Finland), 4 h edi ion. Helsinki: Finnish Museum o Na u al His o y, Bo anical Museum. 530 p. 22. Saikkonen K, Ahlholm J, Helande M, Leh ima¨ki S, Niemela¨inen O (2000) Endophy ic ungi in wild and cul i a ed g asses in Finland. Ecog aphy 23: 360–366. 23. Clay K, Scha dl C (2002) E olu iona y o igins and ecological consequences o endophy e symbiosis wi h g asses. Am Na 160: 99–127. 24. Ko pima¨ki E, No dahl K, Rin a-Jaska i T (1991) Responses o s oa s and leas weasels o luc ua ing ood abundance: is he low phase o he ole cycle due o mus elid p eda ion? Oecologia 88: 552–561. 25. No dahl K, Ko pimaki E (1995) Mo ali y ac o s in a cyclic ole popula ion. P oc R Soc Lond B 261: 49–53. 26. Saha DC, Jackson MA, Johnson-Cicalese JM (1988) A apid s aining me hod o de ec ion o endophy ic ungi in u and o age g asses. Phy opa hology 78: 237–239. 27. Jus us M, Wi e L, Ha mann T (1997) Le els and issue dis ibu ion o loline alkaloids in endophy e-in ec ed Fes uca p a ensis. Phy ochemis y 44: 51–57. 28. Woldemichael GM, Wink M (2002) Concomi an occu ence o py olizidine and quinolizidine alkaloids in he hemipa asi e Osy is alba L. (San alaceae). Biochem Sys Ecol 30: 264–276. 29. K ebs CJ (1999) Ecological me hodology, 2nd edi ion. San F ancisco: Benjamin Cummings. 94–95. 30. Li le TM, Hills FJ (1978) Ag icul u al expe imen a ion: design and analysis. New Yo k: John Wiley & Sons. 350 p. 31. Hi ji KF, Meh a CR, Pa el NR (1987) Compu ing dis ibu ions o exac logis ic eg ession. Jasa 82: 1110–1117. 32. Cono e WJ (1999) P ac ical nonpa ame ic s a is ics, hi d edi ion. New Yo k: John Wiley & Sons. 584 p. G ass Endophy es in Food Chain PLoS ONE | www.plosone.o g 6 Ma ch 2010 | Volume 5 | Issue 3 | e9845