scieee Open visual document viewer

Interactive effects of past and present environments on overwintering success : a reciprocal transplant experiment

Oksanen, Tuula A.,Koivula, Minna,Koskela, Esa,Mappes, Tapio,Soulsbury, Carl D.

Full text

In e ac i e e ec s o pas and p esen en i onmen s on o e win e ing success—a ecip ocal ansplan expe imen Tuula A. Oksanen1, Minna Koi ula2, Esa Koskela3, Tapio Mappes1& Ca l D. Soulsbu y3 1Cen e o Excellence in E olu iona y Resea ch, Depa men o Biological and En i onmen al Science, Uni e si y o Jy ¨ askyl¨ a, Finland 2MTT, Bio echnology and Food Resea ch, Biome ical Gene ics, Jokioinen, Finland 3Depa men o Biological and En i onmen al Science, Uni e si y o Jy ¨ askyl¨ a, Finland Keywo ds Delayed densi y dependence, li e his o y, Myodes gla eolus, popula ion dynamics, ecip ocal ansplan expe imen Co espondence Tuula A. Oksanen, Cen e o Excellence in E olu iona y Resea ch, Depa men o Biological and En i onmen al Science, P.O. Box 35, FI-40014 Uni e si y o Jy ¨ askyl¨ a, Finland; E-mail: [email p o ec ed] Funded by he Academy o Finland g an s o T.A.O., M.K., E.K. and T.M. and he Cen e o Excellence in E olu iona y Resea ch in he Uni e si y o Jy ¨ askyl¨ a. Recei ed: 11 May 2011; Re ised: 28 Oc obe 2011; Accep ed: 31 Oc obe 2011 Ecology and E olu ion 2012; 2(5): 899–907 doi: 10.1002/ece3.82 Abs ac Li e-his o y ai s a e in luenced by en i onmen al ac o s h oughou he li espan o an indi idual. The ela i e impo ance o pas e sus p esen en i onmen on indi idual i ness, he e o e, is a ele an ques ion in popula ions ha ace he challenge o empo ally a ying en i onmen . We s udied he in e ac ing e ec s o pas and p esen densi y on body mass, condi ion, and su i al in enclosu e popula ions o he bank ole (Myodes gla eolus) using a ecip ocal ansplan design. In connec ion wi h he cyclic dynamics o na u al ole popula ions, ou hypo hesis was ha indi iduals bo n in low-densi y enclosu es would do be e o e win e ing in low-densi y enclosu es han in high-densi y enclosu es and ice e sa. Ou esul s show ha he e ec o summe (pas ) densi y was s ong especially on su i al and body mass. The esponse o body mass o summe densi y was nega i e in bo h win e (p esen ) densi y g oups, whe eas he esponse o su i al p obabili y was nonlinea and di e ed be ween he win e densi y g oups. In pa icula , ou da a show a end o highe o e win e ing success o indi iduals o igina ing om he lowes summe densi ies in low win e densi y and ice e sa. We he e o e conclude ha he capaci y o indi iduals o espond o a change in densi y was cons ained by he delayed densi y-dependen e ec s o en i onmen expe ienced in he pas . These e ec s ha e he po en ial o con ibu e o ole popula ion dynamics. Possible mechanisms media ing he e ec s o pas en i onmen in o p esen pe o mance include bo h in insic and en i onmen al ac o s. In oduc ion Condi ions du ing ea ly de elopmen , such as en i onmen- al o pa en al e ec s, can ha e long- e m consequences on indi idual li e his o ies (Becke man e al. 2002). Mo eo e , many species a e capable o de eloping apid esponses o changes in hei immedia e en i onmen such as bu s s o compensa o y g ow h (Me cal e and Monaghan 2001). Va i- a ion in indi idual li e his o ies may he e o e be conside ed an ou come o he in e ac ion be ween genes and en i on- men al ac o s bu also an in e ac ion be ween en i onmen al ac o s ope a ing du ing he di e en phases o he li e cycle (Monaghan 2008). This idea was i s adop ed by esea che s ocusing on compensa o y g ow h, ha is accele a ed g ow h a e a pe iod o es ic ed de elopmen ha enables indi- iduals o ca ch up wi h/ o hei non es ic ed conspeci ics (Ho nick e al. 2000; Me cal e and Monaghan 2001, 2003). The impo ance o compensa o y g ow h on indi idual li e his o ies has been widely examined, o example, in ela ion o he leng h o li espan (Ozanne and Hales 2004; Inness and Me cal e 2008), dominance s a us (Royle e al. 2005), and me abolism (C iscuolo e al. 2008). In a mo e gene al ecological con ex , he li e-his o y e ec s o changes in ood quali y ha e been demons a ed by se e al ood manipula- ion s udies (e.g., Plais ow e al. 2006; Tabo sky 2006; Ba e e al. 2009; Helle e al. in p ess). Food esou ces and die quali y a e en i onmen al ac o s ha ypically a y ac oss he li espan o an indi idual, and a e he e o e biologically meaning ul a ge s o he esea ch on delayed li e-his o y e ec s. Addi ionally, ano he po en- ially impo an a iable in his con ex is popula ion densi y (Becke man e al. 2002). Fo many species, including small c 2012 The Au ho s. Ecology and E olu ion published by Blackwell Publishing L d. This is an open access a icle unde he e ms o he C ea i e Commons A ibu ion-NonComme cial License, which pe mi s use, dis ibu ion and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed and is no used o comme cial pu poses. 899 Su i al E ec s o Pas and P esen Densi ies T. A. Oksanen e al. mammals, he li e-his o y e ec s o popula ion densi y a e well documen ed and conside ed o be e olu iona ily sig- ni ican . Fo example, high popula ion densi y is known o supp ess sexual ma u a ion (P e o –Jullia d e al. 1999; E - gon e al. 2001a), dec ease ep oduc i e success (Koskela e al. 1999), and educe he g ow h a e o ju eniles (Os eld and Canham 1995). In cyclical popula ions, he e is commonly obse ed pa e n known as he Chi y e ec (Chi y 1967; Boons a and K ebs 1979), in which indi iduals a e la ge du ing he inc ease and peak phase o luc ua ions han in he declining and low phase. In he con ex o p edic able densi y luc ua ions, i has e en been sugges ed ha emales may modi y he pheno ype o hei o sp ing acco ding o he densi y o he popula ion o imp o e hei i ness (Lacey 1998). This could be achie ed h ough adap i e ma e nal e ec s, ha is, an adap i e ela ionship be ween o sp ing pheno ype and he en i onmen expe ienced by he mo he (Rossi e 1996; Mousseau and Fox 1998; Ma shall and Ulle 2007). A he indi idual le el, densi y-induced changes in he li e-his o y ai s o small mammals a e mos likely based on pheno ypic plas ici y ha enables quick esponses o changes in he en i onmen (Ag ell e al. 1995; No dahl and Ko - pim¨ aki 2002). Fo example, E gon e al. (2001b) conduc ed an ex ensi e ield ansplan expe imen by mo ing ield oles (Mic o us ag es is L.) be ween si es ha di e ed in a e age o e win e ing body mass. Thei esul s showed ha ans- plan ed oles did no e ain he cha ac e is ics o hei sou ce popula ion and ha he immedia e en i onmen he e o e had an o e iding ole in shaping he body size. Mo eo e , as ep oduc ion s a ed ea lie in si es wi h highe body mass, adjus ing body size o he immedia e en i onmen ca ied a clea bene i by enabling ep oduc ion concu en ly wi h he es o he popula ion and con ibu ing o he popula ion g ow h a e. Howe e , acco ding o ou knowledge, he e ha e been no p e ious expe imen s on small mammals using a ecip ocal ansplan design ha has manipula ed densi y. We aimed o s udy he e ec s o pas and p esen densi ies on he body mass, condi ion, and su i al o young bank oles (Myodes gla eolus Sch ebe ) (Fig. 1), by designing an expe imen whe e indi iduals bo n in an enclosu e popula- ion o 8, 12, 16, 20, o 24 adul indi iduals we e ansplan ed in ei he a low (9–10 adul indi iduals) o a high (18 adul indi iduals) densi y enclosu e o o e win e . We hypo he- sized ha a change in popula ion densi y would lead o a misma ch be ween an immedia e adap i e esponse and u- u e en i onmen , and he eby o lowe ed indi idual i ness (Ba eson e al. 2004). This idea pa allels he concep o p e- dic i e adap i e esponses (PAR) o human e olu ion, which s a es ha o ganisms p ese hei physiology acco ding o he cues o hei p ena al en i onmen in expec a ion ha pa ic- ula physiology will ma ch hei u u e en i onmen (Gluck- man e al. 2005). In ou s udy, howe e , we canno quan- Figu e 1. S udy species, he bank ole Myodes gla eolus is a small oden species common in no he n Eu ope. The main habi a s a e o es s and ields, and he die ypically consis s o o bs, shoo s, seed, be ies, and ungi. Pho o c edi : Ma i Laine. i y how he popula ion densi y expe ienced by he mo he s ansla ed in o he in au e ine en i onmen expe ienced by he s udy animals. In ole popula ions, densi y and densi y- ela ed ac o s (e.g., a ailabili y o ood esou ces and ee b eeding e i- o ies) ha e been ound o a ec se e al li e-his o y ai s such as ma u a ion, ep oduc i e success, and suscep ibil- i y o he cos s o ep oduc ion (Bujalska 1985; Koskela e al. 1999; P e o –Jullia d e al. 1999; Oksanen e al. 2007; Mappes e al. 2008). Popula ion densi y is he e o e likely o ha e a subs an ial e ec on he de elopmen o young indi iduals. I was p edic ed ha he summe densi ies expe ienced by indi- iduals du ing ea ly de elopmen would in luence hei li e- his o y ajec o ies, and ha he ecip ocal ansplan design would challenge he indi iduals o espond o he changed condi ions wi hin he limi s o hei capaci y o u he phe- no ypic plas ici y. Mo e speci ically, we es ed i indi iduals bo n in low summe densi ies would do be e o e win e ing in low win e densi y han in high win e densi y and ice e sa. This is meaning ul in he con ex o ole popula ion cycles because i indi iduals ha a e bo n du ing he peak phase o luc ua ions ha e ela i ely low i ness du ing he declining phase, i could speed up he decline in popula ion densi y and so ampli y he luc ua ions. Ma e ials and Me hods S udy si e Expe imen al animals we e young bank oles aised in en- closu e colonies a Konne esi Resea ch S a ion. Colonies 900 c 2012 The Au ho s. Ecology and E olu ion published by Blackwell Publishing L d. T. A. Oksanen e al. Su i al E ec s o Pas and P esen Densi ies o igina ed om labo a o y-bo n descendan s o wild indi- iduals o iginally cap u ed a ou s udy si e (Konne esi, cen- al Finland 62◦37N, 26◦20E). The e was densi y a ia ion in he enclosu es (8, 12, 16, 20, o 24 adul indi iduals pe 0.2-ha enclosu e) ha had been es ablished o an ea lie en- closu e expe imen (see Oksanen e al. 2007). These densi ies we e all ela i ely high and co espond o he peak phase o he mul iannual popula ion cycles o na u al bank ole pop- ula ions (Yoccoz e al. 2001). Howe e , due o una oidable na u al mo ali y he densi y o eigh indi iduals was consid- e ed a sui able s a ing poin o he low-densi y ea men . Densi y o 24 indi iduals (120 indi iduals/ha) co esponds o a e y high bu no unusual densi y in na u al popula- ions. These densi ies, ha is he densi ies p e ailing when he expe imen al animals we e bo n a e e e ed o as sum- me densi ies. Sex a io o adul indi iduals in he enclosu es was 1:1. The 11 0.2-ha s udy enclosu es we e si ua ed in an old ield. To moni o he animals, 20 mul iple-cap u e li e aps we e dis ibu ed in each enclosu e in a 5 ×4 g id wi h a dis ance o 10 m be ween aps. Each ap was co e ed wi h a gal anized shee me al chimney ha educed exposu e o p ecipi a ion and empe a u e ex emes.Enclosu e ences we e cons uc ed o 1.25-m high gal anized shee me al ha was embedded 0.5 m in o he g ound. The ences we e high enough o en- close he s udy popula ions, bu did no p e en p eda ion by mammalian and a ian p eda o s. The oles we e dependen on na u ally occu ing ood esou ces excep du ing apping pe iods, when he aps we e bai ed wi h oa s and sun lowe seeds. Recip ocal ansplan expe imen Following he end o he b eeding season in Oc obe , all in- di iduals (adul indi iduals and he o sp ing bo n du ing he b eeding season) we e apped om he enclosu es and ans e ed o he labo a o y. O sp ing om he i s li e o each adul emale we e measu ed wi h an elec onic scale o hei body mass ( o nea es 0.01 g) and wi h a digi al calipe o hei head wid h ( o nea es 0.1 mm) and eleased back in o six enclosu es in wo di e en densi ies o o e win e . In o al, his ecip ocal ansplan expe imen included 73 indi iduals eleased a ca. 91 days o age (42 males and 31 emales). Fou o he win e enclosu es had low densi y (nine o 10 indi iduals) and wo win e enclosu es had high densi y (18 indi iduals). These densi ies a e e e ed o as he win e densi ies. Ra ios o males and emales in each enclosu e we e adjus ed as close o 1:1 as possible. App oxima ely one-hal o he indi iduals o igina ing om each summe densi y class (8, 12, 16, 20, o 24) we e assigned o low win e densi y enclosu es and he o he hal was assigned o high win e densi y enclosu es. Indi iduals o igina ing om he same li - e (24 li e s in o al) we e di ided in o di e en ea men s Table 1. Numbe o indi iduals ecip ocally ansplan ed om summe densi ies o win e densi ies. Summe densi ies a e 8, 12, 16, 20, and 24 indi iduals pe enclosu e in he pa en al gene a ion and win e densi ies a e ei he 9–10 (low) o 18 (high) expe imen al indi iduals pe enclosu e. Summe densi y Win e densi y Enclosu e Densi y Low High 1823 7822 11 8 8 8 21221 91266 31622 10 16 3 3 42000 52033 62443 82455 To al 168 37 36 and eplica e enclosu es o andomize he e ec s o common o igin and o a oid inb eeding. The bi h o hese indi iduals had been moni o ed in labo a o y du ing an ea lie expe i- men by Oksanen e al. (2007), and he bi h da e, mo he , and numbe o siblings o hese indi iduals we e he e o e known. The emales and hei li e s we e eleased back in o he enclosu e, he day ollowing he bi h o he pups and he o al ime he emale spen in he labo a o y was on a e age ou o i e days. S udy design is desc ibed in Table 1. Indi iduals assigned o he high and low win e densi y g oups did no di e in hei age, body mass a au umn, o he size o he li e in which hey had been ea ed (mean age [days] ±SE, high: 91.59 ± 0.84,low: 91.39±1.27,Independen samples - es : =0.136, n=73, P>0.8; mean body mass [g] ±SE, low: 16.8 ±0.2, high: 16.9 ±0.2, Independen samples - es : =–0.219, n= 73, P>0.8; mean li e size ±SE, low: 5.4 ±0.2, high: 5.3 ± 0.2, Independen samples - es : =0.200, n=73, P>0.8). Rep oduc i e his o y o hese indi iduals was no eco ded du ing he b eedingseason; howe e ,basedon he bi hda es, each indi idual had had a possibili y o ep oduce a leas 1–2 imes be o e he b eeding season was o e . The animals we e apped again in sp ing (Ma ch) be o e he nex b eeding season s a ed and ans e ed back o he labo a o y whe e hei bodymassandheadwid hwasmeasu ed.Condi ionwas es ima ed as a s anda dized esidual om a linea eg ession o body mass on head wid h (Schul e–Hos edde e al. 2005). Indi iduals ha we e no caugh a e an in ensi e apping e o we e eco ded dead. The p o ocol and he p ocedu es employed we e e iewed and app o ed by he expe imen al animal commi ee o he Uni e si y o Jy ¨ askyl¨ a. c 2012 The Au ho s. Ecology and E olu ion published by Blackwell Publishing L d. 901 Su i al E ec s o Pas and P esen Densi ies T. A. Oksanen e al. Da a analysis Da a we e analyzed using linea (LMM) o gene alized (GLMM) linea -mixed e ec s models (Bolke e al. 2009), using he lme4 (Ba es and Maechle 2009) package in R (R De elopmen Co e Team 2009). Sex ( emale/male) and win- e densi y (low/high) we e used as ixed ac o s and summe densi y and li e size as con inuous co a ia es. The p obabil- i y o su i ing o he end o he expe imen (sp ing) was an- alyzed wi h GLMM wi h binomial dis ibu ion and logi link unc ion. Su i al did no i a linea esponse wi h summe densi y (see Fig. 2A), so we used a quad a ic e m (summe densi y2) o e alua e i su i al a ied nonlinea ly wi h sum- me densi y. Models including he quad a ic e m we e signi - ican ly be e han wi hou i (LR es : χ2=9.67, P=0.008). We s a ed om a model ha included all ele an ac o s as main e ec s and hei wo-way in e ac ions. The model was hen hie a chically simpli ied by emo ing he in e ac ions and main e ec s wi h nonsigni ican P- alues. Howe e , in he analyses o log ans o med body mass and condi ion, we decided o include he e ec s o sex and sex-win e den- si y in e ac ion, because hey a e biologically meaning ul due Figu e 2. E ec s o summe and win e densi y on su i al. (A) The p opo ion o indi iduals ha su i ed a low (open ba s) and high ( illed ba s) win e densi y and (B) he p edic ed p obabili y o su - i al o e he win e . Open ci cles and dashed line, low win e densi y; illed ci cles and solid line, high win e densi y. Equa ion o he cu e o he low-densi y g oup is y =–0.0099x2+ 0.2989x – 1.4345 and equa- ion o he cu e o he high densi y g oup is y =–0.0017x2+ 0.0853x – 0.2883. o he sexual size dimo phism obse ed in he s udy species (Koskela e al. 2009). Models ou pu s we e examined o homogenei y and no - mali y o esiduals. P- alues o Gaussian e o dis ibu ions we e es ima ed by compa ison o a p obabili y dis ibu ion ob ained by 10,000 Ma ko chain Mon e Ca lo simula ions (Baayen 2007). Fo all a iables, we epo e ec sizes and 95% con idence in e als (CI) (Nakagawa and Cu hill 2007). In he analysis o da a collec ed in Oc obe , he uni o sam- plesizeisali e (n=24 o ansplan ed li e s o 57 o all li e s bo n du ing he summe ) and in he analysis o da a collec ed in Ma ch, he uni o sample size is an indi idual (n=35 [body mass and condi ion] o 73 [su i al]). Resul s Body mass o he ansplan ed indi iduals be o e he ecip- ocal ansplan p ocedu e had a s a is ically signi ican neg- a i e ela ionship wi h summe densi y and li e size (li e mean o body mass [g] ±SE: 16.8 ±0.2, indi idual ange: 13.7–20.5; LMM 10 enclosu es, 24 li e s: summe densi y: =–2.31, P=0.038, =–0.52, 95% CI =–0.75/–0.06; li e size: =–3.67, P=0.002, =–0.70, 95% CI =–0.84/–0.35; andom e ec o s udy enclosu e: <0.001 ±<0.001. The s ong in luence o densi y on li e size was demons a ed by a s a is ically signi ican nega i e co ela ion be ween sum- me densi y and li e size o e all 57 li e s whose bi h had been obse ed in he labo a o y (Spea man’s ho =–0.468, n=57, P≤0.001). Condi ion, howe e , was clea ly no e- la ed o ei he o hese co a ia es ( ange: –1.9 o 2.88; LMM: 10 enclosu es, 24 indi iduals, summe densi y: =–0.704, P=0.480, =–0.18, 95% CI =–0.57/0.32; li e size: = –1.272, P=0.232, =–0.32, 95% CI =–0.65/0.19; andom e ec o s udy enclosu e: 0.248 ±0.498). Fo y-eigh pe cen o he ansplan ed indi iduals su - i ed o e he win e ( om Oc obe o Ma ch). Su i al was signi ican ly a ec ed by summe densi y (Table 2). The e- sponse o su i al o summe densi y a ied nonlinea ly in bo h win e densi y g oups (Table 2, Figs. 2A, B). Though he in e ac ion e m be ween summe and win e densi ies was only close o signi ican (Table 2), i s ongly sugges ed ha he dis ibu ion o su i al p obabili ies o indi iduals bo n in high densi ies was highe a high win e densi ies, while he con e se was ue o low summe densi ies (Figs. 2A, B). The esponses o body mass and condi ion o summe den- si y we e linea and nega i e in bo h win e densi y g oups (Figs. 3A, B). The esponse o body mass was s a is ically sig- ni ican , whe eas he esponse o condi ion was weake and only ma ginally signi ican (Table 3). In e ac ion e ms be- ween summe and win e densi ies in hese models we e no signi ican and only sugges ed a weak end o an in e ac i e e ec in body mass (Table 3). 902 c 2012 The Au ho s. Ecology and E olu ion published by Blackwell Publishing L d. T. A. Oksanen e al. Su i al E ec s o Pas and P esen Densi ies Table 2. P obabili y o su i al. Binomial GLMM wi h summe densi y (SD) and win e densi y (WD) as independen a iables (n=73). The in e cep co esponds o an indi idual in low win e densi y. Random e ec o s udy enclosu e is included in o he model (es ima ed pa am- e e o a iance componen ±SD: 0.297 ±0.545). E ec size ( )and noncen al 95% con idence in e als (CI) a e shown o each a iable. Con en ions o e ec sizes: small e ec , =0.10, medium e ec , = 0.30, la ge e ec , =0.50 (Cohen 1988). Es ima e SE zP CI In e cep −8.543 3.305 – – WD 4.988 4.512 1.11 0.269 0.11 −0.09/0.30 SD 1.343 0.490 2.74 0.006 0.27 0.07/0.44 SD (quad a ic) −0.045 0.016 −2.84 0.004 −0.28 −0.09/–0.45 WD ×SD −0.964 0.658 −1.46 0.143 0.15 −0.05/0.34 WD ×SD −0.038 0.021 1.81 0.070 0.19 −0.02/0.37 (quad a ic) Discussion We conduc ed a simple enclosu e expe imen whe e bank oles aised in a ying popula ion densi ies we e ansplan ed ei he in low o high densi y o o e win e . Ou aim was o s udy he in e ac ion be ween summe and win e densi y on pheno ypic a ia ion in body mass, condi ion, and su i al. The ela i e impo ance o pas and p esen en i onmen on indi idual i ness has p e iously been s udied mainly by ma- nipula ing ood esou ces while, acco ding o ou knowledge, his was he i s expe imen u ilizing a densi y-manipula ion app oach. Bank ole popula ionsinFennoscandiashowbo h seasonal and mul iannual cyclic luc ua ions in popula ion densi y (Kallio e al. 2009), and indi idual oles, he e o e, equen ly ace densi y- ela ed changes in hei en i onmen (Hansson and Hen onen 1985; Ko pim¨ aki e al. 2005). The ac ual densi y luc ua ions, howe e , canno be easily in eg a ed in o expe imen al designs, and p e ious s udies on densi y- dependen e ec s on indi idual cha ac e is ics ha e mos ly elied on co ela i e da a (Tkadlec and Zejda 1998; No dahl and Ko pim¨ aki 2002) and compa isons be ween popula ions wi h di e en densi ies ei he in na u al (P e o –Jullia d e al. 1999; E gon e al. 2001a) o in semina u al en i onmen s (Os eld e al. 1993; Os eld and Canham 1995; Koskela e al. 1999). Al hough ou design ei he does no include densi y luc ua ions compa able o na u al ole cycles, i indica es he po en ial o indi iduals o espond o changes in popula ion densi y, which is impo an in he con ex o delayed li e- his o y e ec s (Becke man e al. 2002). Delayed li e-his o y e ec s a e po en ially a key mechanism in linking en i on- men al condi ions o popula ion le el esponses such as ma- e nally media ed densi y e ec s o he ole cycles (Inchaus i e al. 1998). We he e o e o mula ed ou hypo hesis in he con ex o ole popula ion cycles, and p edic ed ha indi id- uals bo n in low densi ies would do be e o e win e ing in low-densi y han in high-densi y en i onmen and ice e sa. This could con ibu e o he cycles, o example, by speeding up he decline phase o he cycle. Acco ding o ou esul s, he e ec o summe densi y was s ong on mos o he measu ed a iables. Summe densi y had a nega i e e ec on he size o he li e s om which he expe imen al indi iduals o igina ed and on he body mass o hese indi iduals be o e he ecip ocal ansplan . Mo e- o e , summe densi y induced a nonlinea esponse in he p obabili y o su i al and had a nega i e e ec on he body mass a e he o e win e ing pe iod (Tables 2, 3 and Figs. 2, 3A). The aw da a sugges ha he o e win e ing success o indi iduals o igina ing om he lowes summe densi ies was be e in low win e densi y han in high win e den- si y and ice e sa (Fig. 2A). The in e p e a ion based on Table 3. Indi idual body mass and condi ion in sp ing. LMM ou pu s o he e ec s o sex, summe densi y (SD), and win e densi y (WD) (n= 35). In e cep co esponds o a emale in low win e densi y. Random e ec o s udy enclosu e is included in o he models (es ima ed pa ame e o a iance componen ±SD: 0.004 ±0.061, 0.088 ±0.297, espec i ely). E ec size ( ) and noncen al 95% con idence in e als (CI) a e shown o each a iable. Con en ions o e ec sizes: small e ec , =0.10, medium e ec , =0.30, la ge e ec , =0.50 (Cohen 1988). Fixed e ec s Es ima e SE P CI Body mass In e cep 3.021 0.145 – – Sex 0.048 0.044 0.63 0.669 0.10 −0.21/0.39 WD −0.322 0.145 −1.39 0.113 −0.22 −0.48/0.10 SD −0.0113 0.007 1.85 0.047 0.29 −0.02/0.53 Sex ×WD 0.082 0.094 0.87 0.315 0.14 −0.18/0.42 WD ×SD 0.010 0.009 1.53 0.094 0.24 −0.07/0.50 Condi ion In e cep 0.628 1.348 – – Sex 0.523 0.518 1.01 0.331 0.17 −0.17/0.46 WD −2.472 1.543 −1.60 0.120 −0.27 −0.53/0.07 SD −0.094 0.050 −1.88 0.065 0.31 −0.02/0.56 Sex ×WD 0.546 0.642 0.85 0.403 0.15 −0.19/0.44 WD ×SD −0.085 0.058 1.46 0.152 0.25 −0.09/0.51 c 2012 The Au ho s. Ecology and E olu ion published by Blackwell Publishing L d. 903 Su i al E ec s o Pas and P esen Densi ies T. A. Oksanen e al. Figu e 3. E ec o summe densi y on body mass and condi ion. A) Body mass in sp ing and (B) body condi ion in sp ing. Figu es a e p oduced om he aw da a and he cu es i ed a e linea . ou bes i ing model (Table 1, Fig. 2B), howe e , is no as s aigh o wa d. The in e ac i e e ec s o summe and win e densi ies on he p obabili y o o e win e ing su i al sugges ha he esponse o he su i al p obabili y o he summe densi y was di e en be ween he low and high win e den- si y g oups. The esponse in he low win e densi y g oup was clea ly quad a ic wi h a peak su i al p obabili y a a summe densi y o app oxima ely 16 (Fig. 2B). In e p e ing he esul s om he poin o iew o he high win e den- si y g oup is mo e complica ed as appea s ha ou summe densi y se ies was no long enough o ca ch he ull shape o he cu e o he peak alue o he p edic ed p obabili y o su i al (Fig. 2B). Tendency o ela i ely low su i al in low densi y may be explained by he possible bene i s o inc eased social ole ance du ing he non ep oduc i e season ha has been obse ed in se e al ole species. Fo example, oles a e less e i o ial, less ac i e, and sha e nes s and hoa ds o ood (e.g., Webs e and B ooks 1981; Wol and Lidicke J 1981; Ecca d e al. 2011). In high-densi y condi ions, hese beha - io s may ansla e in o mo e e ec i e he mo egula ion and co-ope a i e de ense agains o he species. The o e all win- e su i al a e (48%) was simila o a su i al a e om an ea lie s udy (51%) conduc ed in he same s udy enclosu es (Oksanen e al. 2001), sugges ing ha no hing excep ional happened, o example, in he wea he condi ions du ing he win e . The nega i e e ec o li e size on body mass be o e o e - win e ing (i.e., age o ca. 91 days) was expec ed as a ade-o be ween bank ole li e size and o sp ing g ow h and body size has been con i med in nume ous ea lie s udies (Mappes e al. 1995; Koskela 1998; Oksanen e al. 2001, 2002; Mappes and Koskela 2004). Howe e , as li e size was nega i ely e- la ed o popula ion densi y as well, in heo y, he e ec o popula ion densi y on body mass could ha e been posi i e. The nega i e e ec o summe densi y on body mass he e- o e sugges ed ha densi y e ec s we e no media ed by li e size exclusi ely, bu ha densi y had mo e di ec e ec s on body mass as well. Compa ed o p e ious s udies wi h simila designs, ou e- sul s only weakly suppo he indings o he ield ole ans- plan expe imen (E gon e al. 2001b), which sugges ed an o e iding ole o he immedia e en i onmen in shaping li e- his o y ai s and he eby enabled disca ding in insic mech- anisms as an explana ion o a ia ion in hem. Recen ly, an o e iding e ec o an immedia e en i onmen has been e- p o e d in ano he ole ansplan expe imen as well. Helle e al. (in p ess) s udied he long- e m e ec s o ju enile and adul en i onmen s manipula ed by ood suplemen a ion in enclosu e popula ions o he bank ole. Thei s udy showed ha he ep oduc i e success o emales was de e mined by he quali y o he adul en i onme whe eas he su i al o males and he cha ac e is ics o he li e s we e de e mined by he ju enile en i onmen . In ou expe imen , he im- media e en i onmen seemed o challenge he e ec o pas en i onmen only in he e ms o o e win e ing su i al as demons a ed by he shi in he dis ibu ion o he p edic ed su i alp obabili ies(Fig. 2B).Innon-mammalians udysys- ems, Tabo sky (2006) and Ba e e al. (2009) s udied he in e ac ing e ec s o die a y condi ions du ing di e en de- elopmen al pe iods on he li e-his o y ai s in cichlids and cock oaches, espec i ely. Bo h o hese s udies conclude ha he li e-his o y ai s measu ed we e in luenced by ju enile g ow h condi ions a he han by esou ce a ailabili y la e in li e. The esul s o he cu en s udy suppo hei conclusions wi h he excep ion o he weak in e ac ion be ween he pas and he p esen densi ies in body mass. These a he con a- dic o y esul s sugges ha he ela i e signi icance o pas e sus p esen condi ions may a y be ween s udy species as well as ai s, mechanisms, and condi ions unde exami- na ion. Building up a gene al unde s anding abou delayed li e-his o y e ec s, he e o e, will no be an easily a ainable goal. The s onges suppo o ou esul s comes om an in e - eb a e sys em (Plais ow e al. 2006). A ood manipula ion expe imen in soil mi es showed ha he in e gene a ional e ec s o pa en al nu i ional condi ions can be con ex de- penden and ha e complex e ec s on popula ion dynamics (Plais ow e al. 2006). Fu he mo e, he pe sis ence and sig- ni icance o he e ec s o pa en al ood en i onmen (low, 904 c 2012 The Au ho s. Ecology and E olu ion published by Blackwell Publishing L d. T. A. Oksanen e al. Su i al E ec s o Pas and P esen Densi ies medium, o high) a ied be ween he ood en i onmen s o he descendan s, and ha he e ec s o pa en al en i onmen s we e mos p onounced when he descendan en i onmen was no es ic ing, ha is, in high ood condi ions (Plais ow e al. 2006). This esul is compa able o ou esul s on su - i al p obabili y, which sugges ed ha he e ec o summe densi y on su i al p obabili y di e ed be ween he win e densi y g oups and ha he esponse was mo e dis inc when he e was less compe i ion o e esou ces, ha is when he win e densi y was low. Ou da a does ha e some sho comings; in pa icula , due o he limi ed numbe o indi iduals a ailable o he ecip- ocal ansplan expe imen , i was no possible o include he co a iance among indi iduals bo n in o he same li - e o he co a iance among indi iduals o igina ing om he same enclosu e in he s a is ical models. Mo eo e , i is una oidable ha indi iduals a ailable o he ansplan ep- esen a non- andom subse o he o sp ing bo n du ing he summe as iabili y selec ion occu s o e he b eeding season. Howe e , ou esul s on body mass and he p obabil- i y o su i al gene ally seem o suppo he hypo hesis ha he li e-his o y ajec o ies o bank oles a e in luenced by popula ion densi y expe ienced du ing ea ly de elopmen . Fu he mo e, ou esul s on su i al p obabili y sugges ha a la ge-scale change in popula ion densi y has a po en ial o in luence indi idual i ness. The esul s he e o e sugges ha delayed densi y-dependen e ec s may indeed play a ole in he cyclic dynamics o ole popula ions. In enclosu e en i onmen , possible mechanisms media ing he e ec s o ma e nal en i onmen in o o sp ing pe o mance include ma e nal e ec s such as li e size- ela ed ac o s, ma e - nallyde i ed immuni y,andea lyp og ammingo indi idual me abolism. Howe e , pu ely en i onmen al e ec s such as p eda ion o en i onmen al pa hogens canno be uled ou ei he . Mo eo e , he esul s sugges ha in p edic able en i on- men s, i would be possible o indi iduals o imp o e hei i ness by adjus ing hei ep oduc i e e o (i.e., li e size) o he o hcoming densi y condi ions (Gluckman e al. 2005). Howe e , in unp edic able en i onmen s, he capaci y o in- di iduals o espond o changes in popula ion densi y may be cons ained by he delayed densi y-dependen e ec s o he pas en i onmen . In bank oles, his is likely o inc ease he complexi y o easons leading o cyclic popula ion dy- namics and he di icul ies aced in explaining hem. The e- sul s, howe e , a e no only in e es ing in he con ex o ole popula ion dynamics and he possible in insic and ex in- sic causes con ibu ing o hem, bu also in conside ing o he species including humans, which ace he challenge o empo- ally a ying en i onmen , and he eby a possible misma ch be ween immedia e adap i e esponses and u u e en i on- men (Ba eson e al. 2004; Gluckman e al. 2005; Ricka d and Lummaa 2007; on Bonsdo e al. 2011). Acknowledgmen s We hank Konne esi Resea ch S a ion and he Expe imen- al Animal Uni o he Uni e si y o Jy ¨ askyl¨ a o p o iding he acili ies, and Nigel G. Yoccoz as well as he anonymous e iewe s o use ul commen s on he p e ious e sions o he pape . The s udy was inancially suppo ed by he Academy o Finland (g an numbe s: 104568 and 108955 o T.A.O.; 71425, 72896 o M.K.; 115961 and 119200 o E.K.; and 118603, 109165, and 204284 o T.M.) and he Cen e o Excellence in E olu iona y Resea ch in he Uni e si y o Jy ¨ askyl¨ a. Re e ences Ag ell, J., S. E linge, J. Nelson, C. Nilsson, and I. Pe sson. 1995. Delayed densi y-dependence in a small- oden popula ion. P oc. R. Soc. Lond. B. 262:65–70. Baayen, R. H. 2007. Analyzing linguis ic da a: a p ac ical in oduc ion o s a is ics using R. Camb idge Uni . P ess, Camb idge, U.K. Ba e , E. L. B., J. Hun , A. J. Moo e, and P. J. Moo e. 2009. Sepa a e and combined e ec s o nu i ion du ing ju enile and sexual de elopmen on emale li e-his o y ajec o ies: he h i y pheno ype in a cock oach. P oc. R. Soc. B 276:3257–3264. Ba es, D., and M. Maechle . 2009. lme4: linea mixed-e ec s models using S4 classes, R package e sion 0.999375-31. A ailable a h p://lme4. - o ge. -p ojec .o g/. Ba eson, P., D. Ba ke , T. Clu on-B ock, D. Deb, B. D’Udine, R. A.Foley,P.Gluckman,K.God ey,T.Ki kwood,M.M.Lah , e al. 2004. De elopmen al plas ici y and human heal h. Na u e 430:419–421. Becke man, A., T. G. Ben on, V. Kai ala, and P. Lundbe g. 2002. Popula ion dynamic consequences o delayed li e-his o y e ec s. T ends Ecol. E ol. 17:263–269. Bolke ,B.M.,M.E.B ooks,C.J.Cla k,S.W.Geange,J.R. Poulsen, M. H. H. S e ens, and J. -S. S. Whi e. 2009. Gene alized linea mixed models: a p ac ical guide o ecology and e olu ion. T ends Ecol. E ol. 24:127–135. Boons a, R., and C. J. K ebs. 1979. Viabili y o la ge- and small-sized adul s in luc ua ing ole popula ions. Ecology 60:567–573. Bujalska, G. 1985. Regula ions o emale ma u a ion in Cle h ionomys species, wi h special e e ence o an island popula ion o C. gla eolus. Ann. Zool. Fennici 22:331–342. Chi y, D. 1967. The na u al selec ion o sel - egula o y beha iou in animal popula ions. P oc. Ecol. Soc. Aus . 2:51–78. Cohen, J. 1988. S a is ical powe analysis o he beha io al sciences. 2nd ed. L E lbaum Associa es, Hillsdale, NJ. C iscuolo, F., P. Monaghan, L. Nasi , and N. B. Me cal e. 2008. Ea lynu i ionandpheno ypicde elopmen :‘ca ch-up’ g ow h leads o ele a ed me abolic a e in adul hood. P oc. R. Soc. B 275:1565–1570. c 2012 The Au ho s. Ecology and E olu ion published by Blackwell Publishing L d. 905 Su i al E ec s o Pas and P esen Densi ies T. A. Oksanen e al. Ecca dJ.A.,K.Fey,B.A.Cape s,andH.Yl ¨ onen. 2011. B eeding s a e and season a ec in e speci ic in e ac ion ypes: indi ec esou ce compe i ion and di ec in e e ence. Oecologia167:623–633. E gon, T., J. L. MacKinnon, N. C. S ense h, R. Boons a, and X. Lambin. 2001a. Mechanisms o delayed densi y-dependen ep oduc i e ai s in ield oles, Mic o us ag es is: he impo ance o inhe i ed en i onmen al e ec s. Oikos 95:185–197. E gon, T., X. Lambin, and N. C. S ense h. 2001b. Li e-his o y ai s o oles in a luc ua ing popula ion espond o he immedia e en i onmen . Na u e 411:1043–1045. Gluckman, P. D., M. A. Hanson, and H. G. Spence . 2005. P edic i e adap i e esponses and human e olu ion. T ends Ecol. E ol. 20:527–533. Hansson, L., and H. Hen onen. 1985. G adien s in densi y a ia ions o small oden s: he impo ance o la i ude and snow co e . Oecologia 67:394–402. Helle, H., E. Koskela, and T. Mappes. Li e in a ying en i onmen s: expe imen al e idence o delayed e ec s o ju enile en i onmen on adul li e his o y. J. Anim. Ecol., in p ess. Ho nick, J. L., C. V. Eanaeme, O. G´ e a d, I. Du asne, and L. Is asse. 2000. Mechanisms o educed and compensa o y g ow h. Domes . Anim. Endoc inol. 19:121–132. Inchaus i, P., and L. R. Ginzbu g. 1998. Small mammals cycles in no he n Eu ope: pa e n and e idence o a ma e nal e ec hypo hesis. J. Anim. Ecol. 67:180–194. Inness, C. L. W., and N. B. Me cal e. 2008. The impac o die a y es ic ion, in e mi en eeding and compensa o y g ow h on ep oduc i e in es men and li espan in a sho -li ed ish. P oc. R. Soc. B 275:1703–1708. Kallio, E. R., M. Begon, H. Hen onen, E. Koskela, T. Mappes, A. Vahe i, and O. Vapalah i. 2009. Cyclic han a i us epidemics in humans – P edic ed by oden hos dynamics. Epidemics 1:101–107. Ko pim¨ aki, E., K. No dahl, O. Hui u, and T. Klemola. 2005. P eda o -induced synch ony in popula ion oscilla ions o coexis ing small mammal species. P oc. R. Soc. B 272:193–202. Koskela, E., T. Mappes, and H. Ylonen. 1999. Expe imen al manipula ion o b eeding densi y and li e size: e ec s on ep oduc i e success in he bank ole. J. Anim. Ecol. 68:513–521. Koskela, E., T. Mappes, T. Niskanen, and J. Ru kowska. 2009. Ma e nal in es men in ela ion o sex a io and o sp ing numbe in a small mammal – a case o T i e s and Willa d heo y? J. Anim. Ecol. 78: 1007–1014. Koskela, E. 1998. O sp ing g ow h, su i al and ep oduc i e success in he bank ole: a li e size manipula ion expe imen . Oecologia 115:379–384. Lacey, E. P. 1998. Wha is an adap i e en i onmen ally induced pa en al e ec ? Pp. 54–66 in Mousseau, T. A. and C. W. Fox, eds. Ma e nal e ec s as adap a ions. Ox o d Uni . P ess, New Yo k. Mappes, T., and E. Koskela. 2004. Gene ic basis o he ade-o be ween o sp ing numbe and quali y in he bank ole. E olu ion 58:645–650. Mappes,T.,E.Koskela,andH.Yl ¨ onen. 1995. Rep oduc i e cos s and li e size in he bank ole. P oc. R. Soc. Lond. B. 261:19–24. Mappes,T.,M.Koi ula,E.Koskela,T.A.Oksanen,T.Sa olainen, and B. Sine o. 2008. F equency and densi y-dependen selec ion on li e-his o y s a egies – a ield expe imen . PLoS One 3:e1687. Ma shall, D. J., and T. Ulle . 2007. When is a ma e nal e ec adap i e? Oikos 116:1957–1963. Me cal e, N. B., and P. Monaghan. 2001. Compensa ion o a bad s a : g ow now, pay la e ? T ends Ecol. E ol. 16:254–260. Me cal e, N. B., and P. Monaghan. 2003. G ow h e sus li espan: pe spec i es om e olu iona y ecology. Exp. Ge on ol. 38:935–940. Monaghan, P. 2008. Ea ly g ow h condi ions, pheno ypic de elopmen and en i onmen al change. Philos. T ans. R. Soc. B. 363:1635–1645. Mousseau, T. A., and C. W. Fox. 1998. The adap i e signi icance o ma e nal e ec s. T ends Ecol. E ol. 13:403–407. Nakagawa, S., and I. C. Cu hill. 2007. E ec size, con idence in e al and s a is ical signi icance: a p ac ical guide o biologis s. Biol. Re . 82:591–605. No dahl, K., and E. Ko pim¨ aki. 2002. Changes in popula ion s uc u e and ep oduc ion du ing a 3-y popula ion cycle o oles. Oikos 96:331–345. Oksanen, T. A., E. Koskela, and T. Mappes. 2002. Ho monal manipula ion o o sp ing numbe : ma e nal e o and ep oduc i e cos s. E olu ion 56:1530–1537. Oksanen, T. A., M. Koi ula, E. Koskela, and T. Mappes. 2007. The cos o ep oduc ion induced by body size a bi h and b eeding densi y. E olu ion 61:2822–2831. Oksanen, T. A., P. Jonsson, E. Koskela, and T. Mappes. 2001. Op imal alloca ion o ep oduc i e e o : manipula ion o o sp ing numbe and size in he bank ole. P oc. R. Soc. Lond. B. 268:661–666. Os eld, R. S., and C. D. Canham. 1995. Densi y-dependen p ocesses in meadow oles: an expe imen al app oach. Ecology 76:521–532. Os eld, R. S., C. D. Canham, and S. R. Pugh. 1993. In insic densi y-dependen egula ion o ole popula ions. Na u e 366:259–261. Ozanne, S. E., and C. N. Hales. 2004. Li espan: ca ch-up g ow h and obesi y in male mice. Na u e 427:411–412. Plais ow, S., C. Lapsley, and T. Ben on. 2006. Con ex -dependen in e gene a ional e ec s: he in e ac ion be ween pas and p esen en i onmen s and i s e ec on popula ion dynamics. Am. Na . 167:206–215. P e o -Jullia d,A.,H.Hen onen,N.G.Yoccoz,andN.C. S ense h. 1999. Delayed ma u a ion in emale bank oles: op imal decision o social cons ain ? J. Anim. Ecol. 68:684–697. 906 c 2012 The Au ho s. Ecology and E olu ion published by Blackwell Publishing L d. T. A. Oksanen e al. Su i al E ec s o Pas and P esen Densi ies R De elopmen Co e Team. 2009. 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, Aus ia. A ailable om h p://www. -p ojec .o g/. Ricka d, I. J., and V. Lummaa. 2007. The p edic i e adap i e esponse and me abolic synd ome: challenges o he hypo hesis. T ends Endoc inol. Me ab. 18:94–99. Rossi e , M. 1996. Incidence and consequences o inhe i ed en i onmen al e ec s. Ann. Re . Ecol. Sys . 27:451–476. Royle,N.J.,J.Linds ¨ om, and N. B. Me cal e. 2005. A poo s a in li e nega i ely a ec s dominance s a us in adul hood independen o body size in g een swo d ails Xiphopho us helle i. P oc. R. Soc. B 272:1917–1922. Schul e-Hos edde, A. I., B. Zinne , J. S. Milla , and G. J. Hickling. 2005. Res i u ion o mass-size esiduals: alida ing body condi ion indices. Ecology 86:155–163. Tabo sky, B. 2006. Mo he s de e mine o sp ing size in esponse o own ju enile g ow h condi ions. Biol. Le . 2:225–228. Tkadlec, E., and J. Zejda. 1998. Densi y-dependen li e his o ies in emale bank oles om luc ua ing popula ions. J. Anim. Ecol. 67:863–873. on Bonsdo , M. B., T. Ran anen, S. Sipil¨ a, M. K. Salonen, E. Kajan ie, C. Osmond, D. J. P. Ba ke , and J. E iksson. Bi h size and childhood g ow h as de e minan s o physical unc ioning in olde age. Ame ican Jou nal o Epidemiology Ad ance Access published No embe 9, 2011, doi: 10.1093/aje/kw 270. Webs e , A. B., and R. J. B ooks. 1981. Social beha io o Mic o us pennsyl anicus in ela ion o seasonal changes in demog aphy. J. Mammal. 62:738–751. Wol , J. O., and Z. Lidicke J . 1981. Communal win e nes ing and ood sha ing in aiga oles. Beha . Ecol. Sociobiol. 9:237–240. Yoccoz,N.G.,N.C.S ense h,H.Hen onen,andA.-C. P ´ e o -Jullia d. 2001. E ec s o ood addi ion on he seasonal densi y dependen s uc u e o bank ole Cle h ionomys gla eolus popula ions. J. Anim. Ecol. 70:713–720. c 2012 The Au ho s. Ecology and E olu ion published by Blackwell Publishing L d. 907