Beyond productivity- Effects of extreme weather events on ecosystem processes and biotic interactions

Beyond p oduc i i y: E ec s o ex eme wea he e en s on ecosys em p ocesses and bio ic in e ac ions Disse a ion zu E langung des akademischen G ades D . e . na . o geleg de Fakul ä ü Biologie, Chemie und Geowissenscha en de Uni e si ä Bay eu h on F au Julia Wal e (M.A.) gebo en am 01.09.1981 in Memmingen Die o liegende A bei wu de un e de Be euung on P o . Anke Jen sch in de Zei on Mai 2008 bis Feb ua 2011 am Helmhol z Zen um ü Umwel o schung-UFZ in Leipzig, und on Janua 2012 bis Ap il 2012 am Leh s uhl ü S ö ungsökologie an de Uni e si ä Bay eu h ange e ig . Volls ändige Abd uck de on de Fakul ä ü Biologie, Chemie und Geowissenscha en de Uni e si ä Bay eu h genehmig en Disse a ion zu E langung des akademischen G ades eines Dok o s de Na u wissenscha en (D . e . na .). Disse a ion einge eich am: 11.04. 2012 Zulassung du ch die P ü ungskommission: 11.09.2012 Wissenscha liches Kolloquium: 25.10.2012 Am ie ende Dekanin: P o . D . Bea e Lohne P ü ungsausschuss: P o . Anke Jen sch (E s gu ach e in) P o . Ch is iane We ne Pin o (Zwei gu ach e in) P o . Michael Hauhs (Vo si zende ) P o . Thomas Foken P o . John Tenhunen Table o Con en s 1. Sho summa y o he hesis/ Ku ze Zusammen assung de Dok o a bei .......................................... 1 2. Backg ound o he hesis..................................................................................................................... 6 2.1. Clima e change and ex eme wea he e en s ............................................................................... 6 2.1.1. Tempe a u e ex emes .......................................................................................................... 8 2.1.2. P ecipi a ion ex emes .......................................................................................................... 8 2.2. Plan and ecosys em esponse owa ds ex eme wea he e en s................................................ 10 2.2.1. Mo phological and physiological esponse o single plan s o a ious clima ic s ess ypes . 11 Plan esponse o hea ............................................................................................................... 11 Plan esponse o os .............................................................................................................. 11 Plan esponse o d ough ......................................................................................................... 12 Plan esponse o hea y ain all................................................................................................ 12 2.2.2. Impac o ex eme wea he e en s on plan communi ies and ecosys ems......................... 13 Obse a ional s udies ............................................................................................................... 13 Expe imen al e idence on ex eme wea he e en s and plan communi ies ............................ 14 3. On his hesis..................................................................................................................................... 18 3.1. Objec i es o his hesis ............................................................................................................. 18 3.2. Ou line o manusc ip s............................................................................................................... 19 3.3. Eme ging esea ch ques ions ..................................................................................................... 21 3.3.1. Resilience and s ess memo y............................................................................................. 21 3.3.2. Ex eme wea he e en s and ecosys em p ocesses a mul iple le els................................. 22 3.3.3. Clima e change and land use .............................................................................................. 23 Lis o manusc ip s and decla a ion o own con ibu ion...................................................................... 24 P esen a ions o my wo k a con e ences.............................................................................................. 28 Cu iculum o he pos g adua e school HIGRADE ............................................................................. 29 Acknowledgemen s............................................................................................................................... 30 Re e ences o he In oduc ion .............................................................................................................. 31 Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining .................. p oduc i i y.................................................................................................................... 37 Manusc ip 2: How do ex eme d ough and plan communi y composi ion a ec hos plan me aboli es and he bi o e pe o mance?....................................................................... 71 Manusc ip 3: Ecological s ess memo y and c oss s ess ole ance in plan s in he ace o clima e ex emes......................................................................................................................... 90 Manusc ip 4: Do plan s emembe d ough ? Hin s owa ds a d ough -memo y in g asses............... 105 Manusc ip 5: Cold ha diness o Pinus nig a A nold as in luenced by geog aphic o igin, wa ming, and ex eme summe d ough ...................................................................................... 121 Manusc ip 6: Inc eased ain all a iabili y educes biomass and o age quali y o empe a e g assland la gely independen o mowing equency .................................................................. 145 Manusc ip 7: Combined e ec s o mul i ac o clima e change and land-use on decomposi ion in empe a e g assland ..................................................................................................... 167 Synopsis .............................................................................................................................................. 191 Sho summa y o he hesis 1 1. Sho summa y o he hesis/ Ku ze Zusammen assung de Dok o a bei Unde global clima e change, ex eme wea he e en s, such as hea wa es, d ough o hea y ain spells, a e p ojec ed o inc ease in magni ude and equency. As hese may a ec ege a ion and ecosys ems mo e han g adual shi s in mean clima ic pa ame e s, in es iga ing he consequences o ex eme wea he e en s ecen ly became an impo an issue in clima e change esea ch. The main ocus o mos expe imen s in es iga ing e ec s o ex eme wea he e en s on ege a ion is on p ima y p oduc i i y. In ou expe imen in a i icially plan ed communi ies, e en an ex eme d ough o 1000-yea ecu ence did no ha e e ec s on abo e- o below-g ound biomass p oduc ion om 2005-2010 (manusc ip 1). Thus, he main objec i es o his hesis we e (1) o in es iga e i ex eme wea he e en s ha e an e ec on ecosys em unc ions1 beyond p oduc i i y, (2) o es i such a high esis ance o esilience2 in esponse o d ough ega ding p oduc i i y also exis s in mo e na u ally g own plan communi ies and (3) o u he elucida e possible mechanisms o he su p isingly la ge s abili y o he plan communi ies. To in es iga e hese objec i es, se e al expe imen al s udies we e conduc ed in a i icially plan ed, as well as in na u ally g own g assland communi ies and consequences o ex eme wea he e en s o ecosys em p ocesses, such as decomposi ion and he bi o y we e in es iga ed. In a po expe imen , i was s udied, i g ass plan s eac imp o ed owa ds epea ed d ough when compa ed o a i s d ough and hus e eal a kind o d ough memo y. Such a memo y migh be one possible, bu up un il now widely neglec ed mechanism o esilience. E en hough biomass p oduc ion emained s able in ou expe imen in a i icially plan ed communi ies (manusc ip 1), biomass quali y was se e ely a ec ed by ex eme d ough , he eby s ongly a ec ing he de elopmen o a he bi o e ca e pilla eeding on d ough -exposed lea es (manusc ip 2). Fu he , plan compounds o he hos plan depended on he composi ion o he plan communi y i was g own in. This in u n esul ed in s ong e ec s on he la al mo ali y o he bi o es eeding on such plan s. In con as o he s udy in a i icially plan ed communi ies (manusc ip 1), abo eg ound ne p ima y p oduc i i y (ANPP) was educed in na u ally composed g assland in esponse o ex eme ain all a iabili y, including an ex eme d ough ollowed by hea y 1 Ecosys em unc ions: P ocesses ha in ol e mo e han one ecosys em o ophic le el and a e impo an o he main enance o he whole ecosys em (e.g. decomposi ion, which is impo an o nu ien u no e , o p o iding ood o good quali y o sus ain ood webs) 2 Resilience is unde s ood he e as he ime equi ed o e u n o a s eady-s a e ollowing dis u bance (Holling (1973); Gunde son (2000)) Sho summa y o he hesis 2 ain all (manusc ip 6). Fo age quali y was al e ed by d ough . Fu he mo e, mowing equency s ongly al e ed o age quali y and biomass p oduc ion, bu did no in e ac wi h ain all a iabili y and hus did nei he bu e , no ampli y e ec s o ex eme ain all a iabili y. Despi e e ec s o ain all a iabili y on ANPP, g assland showed high esilience a e d ough ollowed by hea y ain, as e ec s we e la ge sho ly a e he ex eme e en , bu did no pe sis un il a second ha es la e in he yea . In na u al g assland, ain all a iabili y and d ough also a ec ed ecosys em p ocesses, he e li e decomposi ion, beyond p oduc i i y (manusc ip 7). D ough ollowed by hea y ain pulses dec eased decomposi ion a es. Decomposi ion in mo e equen ly mown meadows was mo e ulne able owa ds d ough exposu e. Win e wa ming and addi ional win e ain had no long- e m e ec on decomposi ion. To conclude, p ojec ed inc eases in d ough equency unde clima e change may inhibi decomposi ion and al e nu ien and ca bon cycling along wi h soil quali y in empe a e g assland, whe eas a educ ion o snow co e leading o mo e a iable soil su ace empe a u es may coun e ac inc eased decomposi ion unde win e wa ming. In his hesis, an ecological s ess memo y as one possible mechanism o esilience is de ined as any esponse o a single plan a e a s ess expe ience ha imp o es he eac ion o he plan owa ds u u e s ess expe ience and which is assessed on a whole plan le el (manusc ip 3). This hesis u he p o ides e idence o a d ough memo y in g ass plan s (manusc ip 4): Plan s epea edly subjec ed o d ough showed imp o ed pho o-p o ec ion and a highe a e o li ing biomass when compa ed o plan s aced wi h hei i s d ough . Simila ly, ee seedlings exposed o d ough in summe e ealed highe os esis ance du ing win e , p o iding e idence o a long-las ing “c oss-s ess-memo y” (manusc ip 5). To sum up, he hesis shows ha ex eme wea he e en s, e en hough nei he se e ely a ec ing biomass p oduc ion in a i icially composed, no in na u ally g owing communi ies in he long- e m, exe s ong in luence on physiological o biogeochemical pa ame e s, such as plan compounds o soil bio ic ac i i y. These changes in u n modi y ecosys em unc ions beyond p oduc i i y, o example he bi o y o decomposi ion, possibly al e ing bio ic in e a ions and nu ien cycling. Fu he mo e, he indings imply ha plan s exhibi a s ess memo y a e s ess exposu e, which may be one mechanisms leading o a high s abili y and esilience upon equen s ess. Ku ze Zusammen assung de Dok o a bei 3 Ku ze Zusammen assung de Dok o a bei Im Zuge des globalen Klimawandels we den ex eme We e e eignisse, wie Hi zewellen, Dü en ode S a k egene eignisse seh wah scheinlich häu ige und auch in ensi e we den. Da diese Vege a ion und Ökosys eme s ä ke beein lussen können als g aduelle Ände ungen in klima ischen Du chschni spa ame e n, is die Un e suchung de Konsequenzen ex eme We e e eignisse in le z e Zei e s ä k in den Fokus de Klimawandel o schung ge e en. Das Haup augenme k de meis en Expe imen e, die Folgen ex eme We e e eignisse ü die Vege a ion un e suchen, lieg au de P imä p oduk i i ä . Inne halb unse es Expe imen s in küns lich zusammengese z en Gemeinscha en wu de die obe - und un e i dische Biomassep oduk ion du ch eine ex eme Dü e nich beein luss (Manusk ip 1). Dahe sind die Ziele diese A bei , zu un e suchen, (1) ob ex eme We e e eignisse einen E ek au Ökosys em unk ionen1, auße de einen quan i a i en P oduk ion on Biomasse haben, (2) ob die E gebnisse de hohen S abili ä in den küns lich zusammengese z en A engemeinscha en auch ü die na ü lich gewachsenen G ündlandbes änden gel en und (3) mögliche Mechanismen de e s aunlichen S abili ä de P lanzengemeinscha en nähe zu beleuch en. Da ü wu den meh e e Expe imen e in küns lichen und na ü lichen P lanzengemeinscha en du chge üh , in denen Folgen ex eme We e e eignisse ü Ökosys emp ozesse, wie z. B. S euabbau ode He bi o ie, un e such wu den. In einem Top expe imen wu de auße dem un e such , ob G asp lanzen besse mi eine wiede hol en Dü e im Ve gleich zu eine e s en Dü e umgehen können, und dami eine A Dü egedäch nis au weisen. Ein solches Gedäch nis könn e ein mögliche , abe bishe wenig e o sch e Mechanismus on Resilienz2 sein. Obwohl die Biomassep oduk ion küns lich zusammengese z e Gemeinscha en s abil blieb (Manusk ip 1), ände e sich die Biomassequali ä s a k du ch ex eme Dü e. Dadu ch e ände e sich die En wicklung eine phy ophagen Raupe, wenn sie Blä e aß, die eine Dü e ausgese z wo den wa en (Manusk ip 2). Auße dem beein luss e die A enzusammense zung de Gemeinscha , in de die Fu e p lanze wuchs, die P lanzeninhal ss o e, was die S e be a e de La en e ände e. 1 P ozesse, die meh als eine Ökosys emebene be e en und die ü die Au ech e hal ung des gesam en Sys ems no wendig sind 2 Benö ig e Zei , um nach eine S ö ung wiede einen s abilen Zus and zu e eichen (Holling (1973); Gunde son (2000)) Ku ze Zusammen assung de Dok o a bei 4 Im Gegensa z zu S udie in küns lich zusammengese z en Gemeinscha en (Manusk ip 1) wu de die obe i dische Ne op imä p oduk ion (NPP) du ch den Ein luss on ex eme Niede schlags a iabili ä , also ex eme Dü e ge olg on s a kem Regen, eduzie (Manusk ip 6). Auch die Fu e quali ä wu de du ch die Dü e modi izie . Des Wei e en beein luss e die Mahd equenz Fu e quali ä und Biomassep oduk ion. Alle dings konn e die Mahd equenz die E ek e de ex emen Niede schlags a iabili ä wede abpu e n, noch e s ä ken; es gab keine In e ak ion zwischen den beiden Fak o en. T o z de E ek e de Niede schlags a iabili ä au die NPP zeig e sich wiede eine hohe Resilienz on G ünland nach Dü e und S a k egen, da die nega i en E ek e di ek nach dem ex emen We e e eignis seh s a k wa en, abe nich bis zu zwei en E n e Ende des Jah es anhiel en. Auch im na ü lichen G ünland wu den Ökosys emp ozesse, hie S euabbau, neben de P oduk i i ä beein luss : Dü e, ge olg on S a k egen, e inge e S euabbau a en. S euabbau in ö e gemäh en Wiesen wu de du ch die Dü e s ä ke beein äch ig . Win e e wä mung und zusä zlich applizie e Win e niede schlag ha en keine lang is igen E ek e au den Abbau. Zusammen assend läss sich sagen, dass die o he gesag e Zunahme on Dü en den S euabbau behinde n und dadu ch in Näh s o - und Kohlens o k eislau eing ei en könn e. Das Tauen de Schneedecke bei Win e e wä mung üh e zu eine e höh en Va iabili ä de Bodenobe lächen empe a u und könn e dami e höh en Abbau a en du ch Win e e wä mung en gegen wi ken. Die o liegende A bei de inie ökologisches S essgedäch nis als die S essan wo eine Einzelp lanze, die die Reak ion diese gegenübe wiede hol em S ess e besse . Ein solches S essgedäch nis könn e ein mögliche Mechanismus on Resilienz sein (Manusk ip 3). Die A bei zeig e s e Hinweise au ein Dü egedäch nis bei G assp lanzen. P lanzen, die wiede hol e Dü e ausgese z wa en wiesen einen besse en Schu z o oxida i em S ess und dadu ch meh lebende Biomasse au als P lanzen, die das e s e Mal eine Dü e ausgese z wu den. Auch wa en Baumkeimlinge, die im Somme eine Dü e e uh en, im Win e os esis en e , was au ein „C oss-S essgedäch nis“ hinweis (Manusk ip 5). So zeig diese A bei , dass ex eme We e e eignisse, selbs wenn sie die Biomassep oduk ion nich s a k ode lang is ig beein lussen, physiologische ode biogeochemische Pa ame e , wie z. B. P lanzeninhal ss o e ode die Ak i i ä de Boden auna, e ände n. Diese Ände ungen modi izie en wiede um Ökosys em unk ionen, wie He bi o ie ode S euabbau, wodu ch mögliche weise lang is ig in bio ische In e ak ionen ode S o k eisläu e eingeg i en wi d. Wei e hin leg diese A bei nahe, dass Ku ze Zusammen assung de Dok o a bei 5 P lanzen, nachdem sie S ess ausgese z wa en, ein S essgedäch nis en wickeln können, das zu e höh e S abili ä und Resilienz un e häu igen S esse eignissen üh . Backg ound o he hesis 6 2. Backg ound o he hesis 2.1. Clima e change and ex eme wea he e en s Ins umen al empe a u e eco ds show ha a wa ming o he clima e sys em o e he las cen u y is unequi ocal (Hulme, 2005; Blenkinsop and Fowle , 2007; T enbe h e al., 2007). The global mean su ace empe a u es ha e isen by 0.74 ° C (±0.18 ° C) on a e age om 1905-2006 (Hulme, 2005; Blenkinsop and Fowle , 2007; T enbe h e al., 2007). Wa ming was mos p onounced o e land egions, especially o e he no he n hemisphe e du ing win e and sp ing (T enbe h e al., 2007). In Ge many empe a u es ha e isen by 1° C om 1901-2000, wi h a mo e p onounced wa ming du ing win e (Schönwiese e al., 2005; Zebisch e al., 2012). O he empe a u e indices, such as he global sea le el ise o a ound 17 cm in he las cen u y, he educ ion o snow co e in he no he n hemisphe e o he widesp ead glacie e ea a e consis en wi h he eco d showing ising empe a u es (T enbe h e al., 2007). Fu he mo e, global wa ming is accele a ing quickly: he wa ming a e o 0.13° C pe decade om 1955-2005 is almos double he wa ming a e o 0.07 ° C pe decade o 1906-2005 (Beie kuhnlein and Foken, 2008; T enbe h e al., 2007). I is now widely acknowledged, ha an h opogenic g een house gas emissions accoun o he la ges pa o obse ed wa ming since p eindus ial imes and ha he obse ed wa ming can no be explained by in e nal o cing o na u al ex e nal adia i e o cing only (Hege l e al., 2007; T enbe h e al., 2007). Fu u e p ojec ions indica e a u he wa ming o be ween 1.1° C and 6.4 ° C un il 2100, depending on he emission scena io used in he model. E en i CO2 emissions we e held cons an on he le el o he yea 2000 (which is al eady no ul illed), empe a u es con inued o ise o a leas he i s hi d o he 21s cen u y (Meehl e al., 2007). Along wi h ising empe a u es, o he componen s o he clima e sys em, o example p ecipi a ion, a e obse ed and p ojec ed o change. Modi ica ions in he magni ude, as well as in he equency and du a ion o ex eme wea he e en s a e o inc easing conce n: Such changes may occu bo h h ough changes in he mean o in he a iabili y o he dis ibu ion o a clima e a iable, causing disp opo ionally la ge changes in he equency o in ensi y o wea he ex emes, compa ed wi h he changes in he mean (Meehl e al., 2000b; Nicholls and Alexande , 2007) (Fig. 1). Ex eme wea he e en s a e mo e and mo e esponsible o a la ge pa o clima e ela ed damage o socie y and ecosys ems (Field e al., 2012). Backg ound o he hesis 13 oxygen, o example by de eloping ai oo s. Plan s no adap ed o looding can some imes acclima e by his ological modi ica ions, like ae enchymes. O he wise, hey expe ience hypoxia o e en anoxia. This causes e men a ion ins ead o espi a ion in he oo cells, which es ic s g ow h by a as deple ion o s o ed ca bohyd a es. Lac a e and e hanol accumula e and migh cause cell damage by inc easing acidi y. A e e-ae a ion plan s migh su e oxida i e damage by o ma ion o ROS (Schulze e al., 2005; Lambe s e al., 2008). O en myko hiza a e damaged in hypoxic soils, which impai s he plan s nu ien supply. 2.2.2. Impac o ex eme wea he e en s on plan communi ies and ecosys ems Obse a ional s udies Besides physiological and mo phological al e a ions in single plan s, clima ic a iables a ec species dis ibu ion and anges, phenological li e cycle e en s, communi y composi ion and species in e ac ions (Hughes, 2000; Visse and Holleman, 2001; Wal he e al., 2002). Many obse a ional s udies documen he e ec o he g adual wa ming on ege a ion: Polewa ds o upwa ds ange shi s in esponse o wa ming ha e been obse ed o a ious species, e.g. an upwa d shi o he eeline and o alpine plan s in Eu ope in he las decades (Hughes, 2000; Wal he e al., 2002; Pa mesan and Yohe, 2003; Thuille , 2007). The ising empe a u es also led o phenological shi s in many plan species, o ins ance o an ea lie onse o bud bu s o lowe ing (Wal he e al., 2002). Wa me condi ions o en ma ch he needs o in asi e plan s, ha can possibly es ablish mo e apidly and mo e widesp ead unde new condi ions. An inc ease o he mophilic in asi e species has been documen ed in se e al ecosys ems (Wal he e al., 2002). Clima e change may also lead o species ex inc ions, wi h species in moun ain habi a s o he Medi e anean especially endange ed (McCa y, 2001; Thomas e al., 2004; Thuille e al., 2005; Sch ö e e al., 2005). Compa ed o obse a ions o he e ec s o g adual wa ming o plan communi ies, popula ions and species dis ibu ion, obse a ional s udies in es iga ing he consequences o ex eme wea he e en s a e a e, as he occu ence o ex eme clima ic e en s is also a e (Meehl e al., 2000a; Gu schick and Bassi iRad, 2003; Jen sch e al., 2007). Rapid ca as ophic shi s in communi y composi ion o en ollow dis u bances caused by ex eme clima ic e en s (e.g. s o ms)(Sche e e al., 2001). E en less d ama ic e en s may cause changes in species compe i i e and acili a i e in e ac ions (Be ness and Callaway, 1994; Jen sch e al., 2007). Fo ins ance, compe i ion in ensi ies in plan -plan in e ac ions unde ex eme d ough (Tielbo ge and Kadmon, 2000; Ludwig e al., 2004; Maes e and Co ina, 2004). Thus, na u ally occu ing d ough s cause long-las ing shi s in plan communi y Backg ound o he hesis 14 composi ion (Allen and B eshea s, 1998; B eshea s e al., 2005; Muelle e al., 2005). D ough u he educes o es esilience and p oduc i i y and is p ojec ed o inc ease ee mo ali y (Thompson e al., 2009; Llo e e al., 2004; Noo me s e al., 2008; Allen e al., 2010). Many ee-species in he Medi e anean a e p ojec ed o dec ease hei dis ibu ion due o mo e se e e d ough s (Sch ö e e al., 2005). In mesic g assland, howe e , inc eased p ecipi a ion a iabili y, leading o longe d y pe iods ollowed by mo e ex eme ain all e en s, p omo ed plan coexis ence and hus s abilized di e si y (Adle e al., 2006). The ex eme summe hea wa es in Cen al- and Wes e n Eu ope in 2003 and in Eas e n Eu ope in 2010, accompanied by se e e d ough , caused c op ailu e and Eu ope- wide educ ions in p ima y p oduc i i y (Ciais e al., 2005; Ba ioped o e al., 2011). Wa m spells du ing win e ha e also been obse ed o cause damage, as hey may lead o a loss o os acclima ion and hus inc eased damage upon ecu ing os . A win e “hea wa e” in 2007 in no he n Scandina ia, accompanied by hawing, led o ex ensi e damage o he dominan dwa -sh ubs (Bokho s e al., 2009). S imbeck e al. (1995) ound ha a na u al haw du ing midwin e caused deha dening o mon ane ed sp uce. As global wa ming ad ances he beginning o he g owing season, inc easing damage caused by la e os e en s has been obse ed (Gu e al., 2008). Expe imen al e idence on ex eme wea he e en s and plan communi ies As obse a ional e idence on he impac s o ex eme wea he e en s is limi ed, se e al con olled ield-expe imen s assessed e ec s o ex eme clima ic e en s on na u al o a i icially composed ege a ion. The ad an ages o well-conduc ed expe imen s5 a e he possibili y o inco po a e con ol ea men s and o minimize he in luence o con ounding ac o s. Howe e , as such a educ ionis app oach implies a he a i icial condi ions a ely ound in eali y, he ans e o expe imen al e idence on complex, na u al sys ems migh be limi ed. In he beginning o expe imen al clima e change esea ch (1990s), s udies es ing e ec s o ex eme wea he e en s on plan communi ies we e sca ce (Jen sch e al., 2007; manusc ip 1) and he majo i y o he expe imen s implemen ed changes in wea he ends, such as wa ming o inc eased CO2. Un il 2006, esea ch in es iga ing he e ec s o ex eme e en s accoun ed o only one i h o he expe imen al clima e change s udies published 5 Well-conduc ed expe imen s should include p ope con ol ea men s a ying only he ac o s udied, should wo k wi h enough eplica es o ensu e s a is ical powe and should andomly assign ea men s and eplica es. Fu he , ea men a i ac s and biases caused by he expe imen conduc o s ha e o be a oided (Hu lbe , 1984) Backg ound o he hesis 15 (Jen sch e al., 2007). Mos expe imen s assessed abo eg ound p oduc i i y as main esponse pa ame e and in es iga ed e ec s o d ough (manusc ip 1). P ecipi a ion manipula ions: Expe imen ally applied d ough dec eased g assland p oduc i i y in some s udies (Mo ec o e al., 2004; an Ruij en and Be endse, 2010; de Boeck e al., 2011). Howe e , p oduc i i y was o en only a ec ed in esponse o d ough in a id habi a s (Gilgen and Buchmann, 2009; Mi anda e al., 2009) o in gene ally d y yea s (Bloo e al., 2010). The VULCAN expe imen s assessing da a a sh ubland si es ac oss Eu ope, also ound a end o educed biomass p oduc ion a e d ough only a he d ie si es (Penuelas e al., 2004; Penuelas e al., 2007). In mesic g assland, d ough o en had no long- e m e ec s on p oduc i i y (Naud sa e al., 2011), which was also ound o he EVENT I expe imen (manusc ip 1). Despi e o en no ha ing la ge e ec s on p oduc i i y, d ough al e s belowg ound p ocesses, e.g. by educing soil espi a ion (EVENT I and CLIMOOR expe imen : Emme e al., 2004; K eyling e al., 2008; Sowe by e al., 2008; Tobe man e al., 2008; manusc ip 1). Se e al s udies did no es he di ec e ec s o d ough , bu he e ec s o inc eased ain all a iabili y ( ewe , bu la ge e en s, including long d y in e als and hea y ain spells) on g assland pa ame e s. Some s udies showed a la ge e ec o mean annual p ecipi a ion on p oduc i i y (Ba e e al., 2002; Chou e al., 2008), while o he s ound ain all a iabili y o be a mo e impo an d i e o ANPP (Knapp e al., 2002; Fay e al., 2003). In he Rain Manipula ion Plo s (RaMPs) expe imen a Konza P ai ie Biological S a ion in Kansas, USA, a educ ion in soil espi a ion, plan CO2 up ake (Ha pe e al., 2005) soil wa e con en (Fay e al., 2003) and p oduc i i y (Fay e al., 2003; Knapp e al., 2002) and an inc ease in soil ni ogen a ailabili y and in plan di e si y (Knapp e al., 2002) was ound in empe a e con inen al g assland unde inc eased ain all a iabili y (la ge bu ewe ain all e en s wi h a cons an o e all ain all amoun (Heisle and Wel zin, 2006). Heisle - Whi e e al. (2008, 2009) ound a dec ease in p oduc i i y a he empe a e pa o a ansec and an inc ease in he semi-a id end unde ewe , bu la ge ain all e en s. In a Cali o nian g assland, changes in p ecipi a ion pa e ns caused changes in ophic in e ac ions, e.g. a educ ion in consume abundance on a longe ime scale ha o e ode di ec , au ecological sho - e m e ec s (Su le e al., 2007). The d ough s udies no applying compensa ing ain pulses show ha a id sys ems o mesic sys ems in d y yea s a e mo e ulne able owa ds d ough . Thus, a su icien o e all ain all amoun seems o be impo an o g assland eco e y, which was also ound in ou Backg ound o he hesis 16 s udy wi hin he EVENT II expe imen (manusc ip 6). The impac o d ough in a id ecosys ems seems o depend la gely on o e all ain all amoun o he occu ence o se e al la ge ain pulses. Expe imen s es ing e ec s o hea y ain all e en s on ege a ion a e a e. In he EVENT I expe imen hea y ain all e en s had only mino e ec s on p oduc i i y (K eyling e al., 2008). Tempe a u e manipula ions: Expe imen s applying no only g adual wa ming, bu ex eme hea pulses, a e sca ce. A none e al. (2011) ound only sho - e med e ec s o an expe imen al hea wa e on he p oduc i i y o he dominan g ass species in allg ass-p ai ie o Oklahoma, bu no changes in mos o he s udied species. In cold biomes plan s pe o med be e du ing a wa ming pulse, bu wo se a e wa ds, possibly due o a loss o cold esis ance and subsequen highe s ess le els unde he ecu ing cold (Ma chand e al., 2005; Ma chand e al., 2006; Bokho s e al., 2009), whe eas esh li e decomposi ion was una ec ed by wa ming pulses (Bokho s e al., 2010). In he EVENT I expe imen , epea ed soil eeze- haw cycles caused an inc ease in p oduc i i y o empe a e g assland (K eyling e al., 2010). Howe e , lagged s ess e ec s in hea h communi ies diminished biomass wo ege a ion pe iods a e applying wa ming pulses (K eyling e al., 2010). Combined manipula ions o mul iple clima ic a iables: Few expe imen s apply mul iple, combined clima ic s esso s on ege a ion: The CLIMAITE p ojec (Mikkelsen e al., 2008) applying ele a ed CO2, d ough and wa ming as single ac o s and in combina ion on sh ubland sys ems in Denma k ound mos ly smalle esponses o nu ien cycling o he combined ea men s han o he single ea men s. Ne e heless, he u u e clima e scena io combining all ac o s led o educed N u no e (La sen e al., 2011). G ime e al. (2008) ound a la ge long- e m esis ance o in e ile, es ablished g assland in esponse o wa ming, d ough s and wa e addi ions o e 13 yea s. A mesocosm expe imen including he baceous species in Belgium also applied hea wa es and d ough as single and combined ac o s (de Boeck e al., 2011; an Pee e al., 2004). They ound ha nega i e e ec s o d ough on CO2 exchange, g ow h, su i al and biomass p oduc ion we e exace ba ed by hea wa es, whe eas hea wa es alone had no e ec , due o anspi a i e cooling. The summa ized esul s demons a e ha in ensi ying d ough s migh educe p oduc i i y and also ag icul u al yield, especially unde al eady d y condi ions, wi h smalle Backg ound o he hesis 17 o no e ec s in mesic g assland. Fu he mo e, some s udies show ha ex eme e en s al e ne ca bon balance and soil p ocesses, he eby al e ing nu i ional pa hways and soil quali y. The e is an u gen need o u he combine mul iple clima ic s esso s, as e ec s o such mul i ac o expe imen s migh poin in o ally di e en di ec ions as expec ed ou o he esponse owa ds single ac o s (Mikkelsen e al., 2008). S udies in es iga ing pa ame e s o he han p oduc i i y and soil espi a ion a e needed o elucida e e ec s on bio ic in e ac ions and ecosys em p ocesses on mul iple le els. On his hesis 18 3. On his hesis 3.1. Objec i es o his hesis The p e ailing esponse pa ame e o mos expe imen s in es iga ing e ec s o ex eme wea he e en s on ege a ion is p ima y p oduc i i y. In he EVENT I expe imen , in which s a is ically ex eme wea he e en s we e applied on a i icially plan ed communi ies o a ying species- and unc ional di e si y, he ex eme wea he e en s did su p isingly no cause la ge and de imen al changes in g assland p oduc i i y (K eyling e al., 2010). The applied ea men s could consequen ly no be called “ex eme clima ic e en s” sensu Smi h (Smi h, 2011a), as, al hough being ex eme in hei magni ude and leng h ela i e o he e e ence pe iod, hey did no cause an ex eme esponse o plan communi ies, such as widesp ead species mo ali y o communi y b eakdown. Howe e , al hough no se e ely a ec ing p oduc i i y, he wea he ea men s caused mo e sub le changes on a physiological and biogeochemical le el ha a e summa ized in manusc ip 1. Sligh changes, o example in plan me abolic compounds can a ec mul iple ecosys em p ocesses and le els, o ins ance by dec easing pala abili y o he bi o es o by changing decomposi ion a es, which in u n al e s ophic in e ac ions and nu ien cycling. Thus, one objec i e o his hesis was o elucida e how ex eme wea he e en s a ec ecosys em unc ions beyond p oduc i i y, such as plan -he bi o e in e ac ions o decomposi ion. Especially mesic g assland communi ies a e o en e y s able when aced wi h ex eme d ough (see sec ion 2.2.2.), which was also shown in he EVENT I expe imen . Ye , he unde lying mechanisms o such a high s abili y a e no well unde s ood. Ano he objec i e o his hesis is o u he elucida e possible mechanisms o he su p isingly la ge esis ance o esilience o plan s and plan communi ies when aced wi h ex eme wea he e en s. He e, he ocus is on a possible s ess memo y, as up un il now i is unclea , how plan s and plan communi ies eac when s ess is applied epea edly o e a ela i ely sho ime span. On he one hand, his migh lead o a s ep-wise educ ion in he abili y o eco e , un il a o al b eakdown o he sys em (Sche e e al., 2001). On he o he hand, s ess acclima ion may lead o a pe sis ing inc ease in s ess esis ance, a mechanisms ha could be ega ded as kind o s ess memo y. The conside a ion o no only an inc eased e en magni ude, bu also o an inc eased equency o e en s is u gen ly needed in s udying clima ic ex emes (Smi h, 2011b). On his hesis 19 The EVENT I expe imen is highly con olled in e ms o species composi ion, as he plan ed communi y composi ions we e kep cons an o e he yea s by pe iodically weeding. To in es iga e i he indings o high s abili y in he a i icially composed plan communi ies can be con e ed o mo e na u al sys ems, he EVENT II expe imen was es ablished on a semi-na u al meadow in 2008. He e, no only ain all was manipula ed, bu also di e en land-use scena ios we e implemen ed. This expe imen was also designed o answe he ques ion whe he he e ec s o d ough o hea y ain a e caused by an o e all al e a ion in mean annual ain all amoun , o by inc eased ain all a iabili y (la ge , bu ewe ain all e en s) unde cons an annual ain all amoun . Fo his eason, in EVENT II ain all amoun was kep cons an om 2009 onwa ds and only he size o and he in e als be ween he ain all e en s we e a ied. To sum up, he main objec i es o his hesis we e (1) o in es iga e i ex eme wea he e en s ha e an e ec on ecosys em unc ions beyond p oduc i i y, (2) o es i he high s abili y o esilience in esponse o d ough ega ding p oduc i i y also exis s in mo e na u ally g own plan communi ies and (3) o u he elucida e possible mechanisms o he su p isingly la ge esis ance o esilience o he plan communi ies. 3.2. Ou line o manusc ip s The i s manusc ip summa izes 5 yea s o d ough esea ch in he a i icially plan ed g assland communi ies o EVENT I. Ex eme d ough had no e ec on abo eg ound- o belowg ound p oduc i i y. Ne e heless, se e al o he physiological and biogeochemical pa ame e s we e a ec ed. I physiological changes on a lea le el in luence o he ecosys em le els and p ocesses in he long- e m had hus o be in es iga ed. The second manusc ip he e o e deals wi h changes in lea compounds caused by ex eme d ough and esul ing e ec s on he bi o es eeding on such lea es. A second ocus o his s udy was o elucida e e ec s o plan communi y composi ion on lea compounds and, as a consequence, he bi o e de elopmen . The s udy showed ha changes in g ass compounds caused by se e e d ough a ec ed he bi o es eeding on such g ass: Ca e pilla s ed wi h d ough -subjec ed lea es showed signi ican ly highe su i al, a longe du a ion o la al de elopmen and highe pupal weigh . Fu he , plan compounds o ou a ge g ass depended on he composi ion o he plan communi y i was g own in, which in u n a ec ed he bi o e de elopmen : La ae eeding on species- iches communi ies wi hou legumes showed he highes mo ali y, which was closely linked o low p o ein con en in hese lea es. This s udy p o ides e idence ha e en qui e sub le changes in plan s caused by d ough o communi y composi ion a e able o in luence bio ic in e ac ions and may e en lead o desynch onisa ion On his hesis 20 o ophic and phenological adjus men s unde clima e change. Fu he mo e, as clima e change is likely o a ec plan communi y composi ion, his will u he a ec lea quali y and hus plan -he bi o e in e ac ions. The second objec i e o his hesis was o elucida e possible mechanisms o he high s abili y o g assland p oduc i i y unde clima ic ex emes. In he i s h ee yea s o he EVENT I expe imen , a d ough o 100-yea ecu ence was applied (leading o 32 days o consecu i e d ough ), and in he nex yea s, a d ough o 1000 yea ecu ence was applied. In e e y yea , he same plo s we e subjec ed o d ough . One possible mechanisms o esilience migh be ha he communi ies buil up an ecological memo y ha helped hem o cope wi h d ough in he ollowing yea s. As ecological memo y on a communi y le el is di icul o assess, we ocused on an ecological s ess memo y on a single plan le el. Su p isingly ew s udies in es iga ed i whole plan s a e able o emembe s ess and o eac imp o ed owa ds a ecu en s ess e en . This issue is especially impo an as equency o ex eme wea he e en s is p ojec ed o inc ease unde clima e change (Smi h, 2011b). Fu he , a common de ini ion o s ess memo y o ecologis s is missing. The hi d manusc ip hus i s de ines he concep o an ecological s ess memo y on a whole plan le el, e iews he ew exis ing s udies indica ing s ess memo y a e clima ic s ess (d ough , os , hea ) and discusses possible mechanisms o an ecological s ess memo y, including epigene ic ones. A d ough memo y in g ass plan s was in es iga ed wi hin a po -expe imen in which one g oup o plan s was subjec ed o a single d ough and he o he o ecu en d ough (manusc ip 4). This s udy p o ided e idence ha g ass plan s a e able o emembe d ough e en a e a ha es and esp ou ing and o show a highe pe cen age o li ing biomass, due o imp o ed pho op o ec ion, when compa ed o plan s subjec ed o hei i s d ough . Simila ly, he expe imen pe aining o manusc ip 5 es ed os ha diness o Pinus nig a ju eniles and showed ha plan s exposed o d ough du ing summe e ealed highe os ha diness in win e (manusc ip 5). As bo h, os and d ough s ess, in ol e dehyd a ion s ess, i migh well be ha an ecological c oss-s ess memo y was in ol ed he e. Plan os ha diness in his s udy was ela ed o a highe concen a ion o ca bohyd a es. Con en o ca bohyd a es is also o en inc eased unde d ough (e.g. manusc ip 2). Thus, he c oss-s ess memo y indica ed in manusc ip 5 migh be ela ed o he as e syn hesis o soluble ca bohyd a es. To es i he indings o he a i icially composed plan communi ies also hold unde mo e ealis ic condi ions, an ex eme d ough was also applied on na u ally g own g assland communi ies in he EVENT II expe imen . He e, e ec s o inc eased ain all a iabili y (changes in iming and dis ibu ion o ain all, bu no in o e all ain all sum) on he On his hesis 21 p oduc i i y and some aspec s o o age quali y o es ablished g assland we e in es iga ed. In con as o he indings in a i icially plan ed communi ies (manusc ip 1), ANPP and o age quali y we e educed in na u ally composed g assland in esponse o ex eme d ough ollowed by hea y ain all e en s (manusc ip 6). Mowing equency s ongly al e ed o age quali y and biomass p oduc ion, bu did did nei he bu e , no ampli y e ec s o ex eme ain all a iabili y on p oduc i iy, as i did no in e ac wi h ain all a iabili y manipula ions. Despi e e ec s o ain all a iabili y on ANPP, g assland showed high esilience a e ex eme sp ing d ough ollowed by hea y i iga ion, as e ec s we e la ge sho ly a e he ex eme e en , bu did no pe sis un il a second ha es la e in he yea , when no di e ences be ween he ain all a iabili y manipula ions appea ed. In he p eceding yea , when he ex eme sp ing d ough was no ollowed by i iga ions and hus also ecei ed he smalles o e all amoun o wa e , nega i e e ec s on p oduc i i y we e la ge and emained un il he second ha es in la e summe . Then, o me ly d ough exposed communi ies s ill showed educed biomass p oduc ion. This highligh s he impo an ole o a su icien o e all amoun o ain all o eco e y p ocesses in empe a e g assland and is in acco dance wi h he d ough s udies men ioned in sec ion 2.2.2., showing se e ely ad e se e ec s o d ough p ima ily in d y yea s o in a id biomes. As his hesis in es iga es e ec s o ex eme wea he e en s on ecosys ems beyond p oduc i i y, manusc ip 7 epo s indings o a long- e m decomposi ion expe imen conduc ed wi hin EVENT II. Ex eme d ough educed li e decomposi ion when li e bags we e exposed o d ough o six weeks wi hin an 11 mon h pe iod. Su p isingly, low ain all a iabili y wi h egula i iga ion dec eased decomposi ion. Addi ional win e ain accele a ed decomposi ion, whe eas win e wa ming had no e ec on decomposi ion, bu educed snow co e and inc eased a iabili y o su ace empe a u es. Mo e equen mowing s ongly s imula ed decomposi ion, which could be a ibu ed o changes in li e quali y. Howe e , he s imula ing e ec o equen mowing was absen unde ex eme ain all a iabili y including d ough . P ojec ed inc eases in d ough equency unde clima e change may inhibi decomposi ion and al e nu ien and ca bon cycling along wi h soil quali y. Especially decomposi ion in in ensi ely managed g assland appea s ulne able owa ds d ough . 3.3. Eme ging esea ch ques ions 3.3.1. Resilience and s ess memo y O en, and also in ou s udy (manusc ip 1) g assland shows a su p isingly la ge esis ance o esilience owa ds d ough . Mechanisms o esilience emain o be elucida ed. On his hesis 22 One likely mechanism is a s ess memo y o plan s ha ende s hem less ulne able o epea ed s ess e en s (manusc ip s 3, 4, 5). Howe e , i such a mechanisms exis s also unde na u al condi ions and also on la ge scales, e.g. on a communi y le el, is ye o be in es iga ed, especially as indings o manusc ip 6 imply ha g assland esilience unde mo e na u al condi ions migh be diminished unde gene ally d y condi ions. Possible mechanisms o a s ess memo y a e la gely unknown. Join esea ch o ecologis s and molecula biologis is needed o elucida e possible epigene ic mechanisms. Fi s s udies al eady showed he he i abili y o acqui ed s ess ole ance (see manusc ip 3). Besides ecological s ess memo y, o he unde lying physiological and biogeochemical p ocesses ha se e o main ain p oduc i i y and migh hus be mechanisms o communi y s abili y and eco e y ha e o be iden i ied and add essed in u u e esea ch. Main aining ecosys em esilience is o majo impo ance o mi iga e and p e en ca as ophic consequences o global clima e change. 3.3.2. Ex eme wea he e en s and ecosys em p ocesses a mul iple le els Up un il now, he main esponse pa ame e s udied in esea ch on ex eme clima ic e en s is p ima y p oduc ion (manusc ip 1). Howe e , e en i p ima y p oduc ion emains s able, o he physiological and biogeochemical pa ame e s a e changed unde ex eme wea he e en s (manusc ip s 1, 2, 7). Such changes, e.g. ood plan quali y migh se iously in e e e in ecosys em synch onisa ion and ecosys em unc ioning. Fu he wo k o s udy long- e m e ec s o ex eme wea he e en s on, e.g. bio ic in e ac ions o biodi e si y is needed o es ima e consequences o wea he ex emes and o enable policy make s o p e en des abiliza ion o es ablished ood-webs and o seize measu es o adap a ion. How he bi o es migh eac o changes in hei hos plan in mo e na u al condi ions han he ones desc ibed in manusc ip 2 and whe he specialis s migh eac di e en ly compa ed o gene alis he bi o es also needs u he esea ch. We showed ha win e wa ming did no inc ease decomposi ion, due o loss o snow insula ion and inc eased su ace empe a u e a iabili y. How decomposi ion migh be a ec ed by summe wa ming, also in combina ion wi h d ough condi ions, needs u he s udy. Long- e m changes in soil bio ic ac i i y unde mo e equen mowing needs o be add essed, o ind explana ions o he highe ulne abili y o decomposi ion owa ds d ough in mo e equen ly mown communi ies. P elimina y esul s o he EVENT expe imen s also indica e s ong e ec s o hea y ain all on bio ic in e ac ions, such as myco hiza o decompose auna. As hea y ain e en s a e expec ed o inc ease in he u u e, bu a e a ely s udies ye , mo e in es iga ions a e needed o look a e ec s o hea y ain on ecosys em unc ions. Cu iculum o pos g adua e school 29 Cu iculum and c edi poin s o he pos g adua e school HIGRADE and awa d cou se name du a ion own con ibu ion c edi poin s In oduc ion o wa e esou ces and aqua ic ecosys em managemen 3 days ac i e pa icipa ion and homewo k 1 In oduc ion in o biodi e si y sciences 2 days ac i e pa icipa ion 1 Ad anced cou se e es ial ecosys em unc ions and biodi e si y 3 days ac i e pa icipa ion 2 Ad anced cou se p o eomics 5 days lab wo k and analysis 2 Semina on land-use con lic s and conse a ion o na u al esou ces 1 day p esen a ion and ac i e pa icipa ion 1 Applica ion cou se “Land-use con lic s and conse a ion o na u al esou ces in he Banaue egion o No he n-Luzon/ Philippines 12 days conduc ion o ield expe imen and w i ing o inal epo 3 So Skills: P esen a ions in Englisch 2 days ac i e pa icipa ion including sho p esen a ions 1 So Skills: Scien i ic W i ing 3 days ac i e pa icipia ion including w i ing o sho sec ions 1 So Skills: G an Aquisi ion 1 day pa icipa ion 0.25 Fou alks a in e na ional con e ences p epa a ion and p esen a ion o alks 2 Fou p esen a ions in he UFZ  epa a io semina and one pos e p esen a ion a he UFZ Topic I con e ence p epa a ion and p esen a ion o alks 1.5 O ganisa ion o UFZ doc days 2009 se e al days planning o loca ion, ac i i ies, alks, schedule 0.75 S a is ics: Da a Analysis and Modelling using R 6 days ac i e pa icipa ion 1.25 Publica ion o a icles in ISI-lis ed jou nals p epa a ion o manusc ip s, i s and co esponding au ho 2 Pa icipa ion a he compe i ion “Wissenscha Ve s ehen” and AWARD o he 3 d place Finals we e held one day  epa a ion o a icle and 15 minu e p esen a ion 1 20.75 Acknowledgemen s 30 Acknowledgemen s: I hank - P o . Anke Jen sch o gi ing me he oppo uni y o do my phD wi hin he g oup, o gi ing me space o de elop own in e es s and ideas, o p o iding a wo king a mosphe e ha makes ha d wo k easie , o gi ing me he chance o p esen my wo k a con e ences and o being he amily- iendlies supe iso - my second supe iso , D . Ha ald Auge, o many ui ul discussions and ideas o my esea ch - he depa men o Conse a ion Biology a he Helmhol z-Cen e o En i onmen al Resea ch o my nice “win e -home”, D . Klaus Henle o gi ing me he oppo uni y o wo k in Bay eu h du ing summe and he g adua e school HIGRADE o a lo o in e es ing cou ses and ainings and o unding some esea ch s ays - P o . Ca l Beie kuhnlein o gi ing me he possibili y o use (wo)man-powe and echnical esou ces o his g oup and o as and help ul edi ing o my manusc ip s - D . G ego Aas and he s a o he Ecological Bo anical Ga den o he Uni e si y o Bay eu h o hei suppo o he whole expe imen - P o . Wol am Beyschlag (Uni e si y o Biele eld), Jun.P o . Ch is iane We ne -Pin o (Uni e si y o Biele eld), P o . John Tenhunen (Uni e si y o Bay eu h), D . Uwe Rasche (FZ Jülich) and Ha ald Auge (UFZ Halle) o p o iding us hei echnical equipmen and D . Sonja Lö le (LFE Ebe swalde) o he help and suppo wi h he lea chemical analysis - nume ous s uden wo ke s and in e ns wi hin he EVENT- expe imen s o hei help in main enance, ea men execu ion and o doing measu emen s wi h me, e en “p e-dawn”, especially o Roman Hein, Lau a Nagy, Inés Pas o , Julia Gommola, Jan Taucha, Julia Smi h, Da id Eichenbe g and Clesio Gomes da Sil a - all he echnicians o hei gene al help in he expe imen , and especially Ch is ian Schemm and Ch is ine Pilsl o hei help in he lab and Reinhold S ahlmann o his help wi h “compu e -s u ” - all my co-phDs (Ke s in, Lau a, Da id, Roman, Daniel, Jan) and ou pos -doc, Jü gen K eyling, o making ield wo k un, o physical and psychological suppo and o many use ul discussions - my pa en s, sis e s iends and Roman and Smilla (I hope hey all know wha o ) Re e ences 31 Re e ences o he In oduc ion: Adle , P.B., HilleRisLambe s, J., Ky iakidis, P.C., Guan, Q., Le ine, J.M., 2006. Clima e a iabili y has a s abilizing e ec on he coexis ence o p ai ie g asses. P oceedings o he Na ional Academy o Sciences o he Uni ed S a es o Ame ica 103, 12793–12798. Alexande , L., Zhang, X., Pe e son, T., Caesa , J., Gleason, B., Klein Tank A.M.G., Haylock, M., Collins, D., T ewin, B., Rahimzadeh, F., Tagipou , A., Rupa Kuma , K., Re adeka , J., G i i hs, G., Vincen , L., S ephenson, D.B., Bu n, J., Aguila , E., B une , M., Taylo , M., New, M., Zhai, P., Rus icucci, M., Vazquez- Agui e, J., 2006. Global obse ed changes in daily clima e ex emes o empe a u e and p ecipi a ion. Jou nal o Geophysical Resea ch- A mosphe es 111, 1–22. Allan, R., Soden, B., 2008. A mosphe ic wa ming and he ampli ica ion o p ecipi a ion ex emes. Science 321, 1481–1484. Allen, C.D., B eshea s, D.D., 1998. D ough -induced shi o a o es -woodland eco one: Rapid landscape esponse o clima e a ia ion. P oceedings o he Na ional Academy o Sciences o he Uni ed S a es o Ame ica 95, 14839–14842. Allen, C.D., Macalady, A.K., Chenchouni, H., Bachele , D., McDowell, N., Venne ie , M., Ki zbe ge , T., Rigling, A., B eshea s, D.D., Hogg, E.H., Gonzalez, P., Fensham, R., Zhang, Z., Cas o, J., Demido a, N., Lim, J.-H., Alla d, G., Running, S.W., Seme ci, A., Cobb, N., 2010. A global o e iew o d ough and hea - induced ee mo ali y e eals eme ging clima e change isks o o es s. Fo es Ecology and Managemen 259, 660–684. Allen, M., Ing am, W., 2002. Cons ain s on u u e changes in clima e and he hyd ological cycle. Na u e 419, 224–232. A none, J.A., Jasoni, R.L., Lucchesi, A.J., La sen, J.D., Lege , E.A., She y, R.A., Luo, Y., Schimel, D.S., Ve bu g, P.S., 2011. A clima ically ex eme yea has la ge impac s on C4 species in allg ass p ai ie ecosys ems bu only mino e ec s on species ichness and o he plan unc ional g oups. Jou nal o Ecology 99, 678–688. Ba e , J.E., McCulley, R.L., Lane, Bu ke, I.C., Lauen o h, W.K., 2002. In luence o clima e a iabili y on plan p oduc ion and N-mine aliza ion in Cen al US g asslands. Jou nal o Vege a ion Science 13, 383–394. Ba ioped o, D., Fische , E.M., Lu e bache , J., T igo, R., Ga cia-He e a, R., 2011. The Ho Summe o 2010: Red awing he Tempe a u e Reco d Map o Eu ope. Science 332, 220–224. Ba holy, J., Pong acz, R., Gelybo, G., Szabo, P., 2008. Analysis o expec ed clima e change in he Ca pa hian Basin using he PRUDENCE esul s. Idoja as 112, 249–264. Beie , C., Emme , B., Gunde sen, P., Tie ema, A., Penuelas, J., Es ia e, M., Go don, C., Go issen, A., Llo ens, L., Roda, F., Williams, D., 2004. No el app oaches o s udy clima e change e ec s on e es ial ecosys ems in he ield: d ough and passi e nigh ime wa ming. Ecosys ems 7, 583–597. Beie kuhnlein, C., Foken, T., 2008. Klimawandel in Baye n. Auswi kungen und Anpassungsmöglichkei en. Bay eu he Fo um Ökologie 113, Bay eu h. Benis on, M., S ephenson, D.B., Ch is ensen, O.B., Fe o, C.A.T., F ei, C., Goye e, S., Halsnaes, K., Hol , T., Jylha, K., Ko i, B., Palu iko , J., Schoell, R., Semmle , T., Wo h, K., 2007. Fu u e ex eme e en s in Eu opean clima e: an explo a ion o egional clima e model p ojec ions. Clima ic Change 81, 71–95. Be ness, M.D., Callaway, R., 1994. Posi i e In e ac ions in Communi ies. T ends in Ecology and E olu ion 9, 191–193. Blenkinsop, S., Fowle , H.J., 2007. Changes in Eu opean d ough cha ac e is ics p ojec ed by he PRUDENCE egional clima e models. In e na ional Jou nal o Clima ology 27, 1595–1610. Bloo , J., Pichon, P., Falcimagne, R., Leadley, P., Soussana, J.F., 2010. E ec s o wa ming, summe d ough , and CO2 en ichmen on abo eg ound biomass p oduc ion, lowe ing phenology, and communi y s uc u e in an upland g assland ecosys em. Ecosys ems 13, 888–900. Bobe g, F., Be g, P., Thejll, P., Gu owski, W.J., Ch is ensen, J.H., 2010. Imp o ed con idence in clima e change p ojec ions o p ecipi a ion u he e alua ed using daily s a is ics om ENSEMBLES models. Clima e Dynamics 35, 1509–1520. Boeck, H.J. de, D eesen, F.E., Janssens, I.A., Nijs, I., 2011. Whole-sys em esponses o expe imen al plan communi ies o clima e ex emes imposed in di e en seasons. New Phy ologis 189, 806-817. Bohne , H.J., 2000. Wha makes desicca ion ole able? Genome biology 1, 1–4. Bokho s , S., Bje ke, J., Melillo, J., Callaghan, T., Phoenix, G., 2010. Impac s o ex eme win e wa ming e en s on li e decomposi ion in a sub-A c ic hea hland. Soil Biology and Biochemis y 42, 611–617. Bokho s , S.F., Bje ke, J.W., Tomme ik, H., Callaghan, T.V., Phoenix, G.K., 2009. Win e wa ming e en s damage sub-A c ic ege a ion: consis en e idence om an expe imen al manipula ion and a na u al e en . Jou nal o Ecology 97, 1408–1415. B eshea s, D., Cobb, N., Rich, P., P ice, K., 2005. Regional ege a ion die-o in esponse o global-change- ype d ough . P oceedings o he Na ional Academy o Sciences o he Uni ed S a es o Ame ica 102, 15144– 15148. Re e ences 32 B i a, K.R., an de Sch ie , G., Jones, P.D., 2009. We and d y summe s in Eu ope since 1750: e idence o inc easing d ough . In e na ional Jou nal o Clima ology 29, 1894–1905. Cha es, M.M., Pe ei a, J.S., Ma oco, J., Rod igues, M.L., Rica do, C.P., Oso io, M.L., Ca alho, I., Fa ia, T., Pinhei o, C., 2002. How plan s cope wi h wa e s ess in he ield. Pho osyn hesis and g ow h. Annals o Bo any 89, 907–916. Chou, W.W., Sil e , W.L., Jackson, R.D., Thompson, A.W., Allen-Diaz, B., 2008. The sensi i i y o annual g assland ca bon cycling o he quan i y and iming o ain all. Global Change Biology 14, 1382–1394. Ch is ensen, O., Ch is ensen, J., 2004. In ensi ica ion o ex eme Eu opean summe p ecipi a ion in a wa me clima e. Global and Plane a y Change 44, 107–117. Ciais, P., Reichs ein, M., Vio y, N., G anie , A., Ogée, J., Alla d, V., Aubine , M., Buchmann, N., Be nho e , C., Ca a a, A., Che allie , F., Noble , N. de, F iend, A.D., F iedlings ein, P., G ünwald, T., Heinesch, B., Ke onen, P., Knohl, A., K inne , G., Lous au, D., Manca, G., Ma eucci, G., Miglie a, F., Ou ci al, J.M., Papale, D., Pilegaa d, K., Rambal, S., Seu e , G., Soussana, J.F., Sanz, M.J., Schulze, E.D., Vesala, T., Valen ini, R., 2005. Eu ope-wide educ ion in p ima y p oduc i i y caused by he hea and d ough in 2003. Na u e 437, 529–533. Coppola, E., Gio gi, F., 2010. An assessmen o empe a u e and p ecipi a ion change p ojec ions o e I aly om ecen global and egional clima e model simula ions. In e na ional Jou nal o Clima ology 30, 11–32. Dai, A., T enbe h, K.E., Qian, T., 2004. A global da ase o Palme D ough Se e i y Index o 1870-2002: Rela ionship wi h soil mois u e and e ec s o su ace wa ming. Jou nal o Hyd ome eo ology 5, 1117–1130. Eas e ling, D.R., E ans, J., G oisman, P., Ka l, T.R., Kunkel, K., Ambenje, P., 2000. Obse ed Va iabili y and T ends in Ex eme Clima e E en s: A B ie Re iew. Bulle in o he Ame ican Me eo ological Socie y 81, 417–425. Emme , B.A., Beie , C., Es ia e, M., Tie ema, A., K is ensen, H.L., Williams, D., Penuelas, J., Schmid , I., Sowe by, A., 2004. The esponse o soil p ocesses o clima e change: Resul s om manipula ion s udies o sh ublands ac oss an en i onmen al g adien . Ecosys ems 7, 625–637. Fay, P.A., Ca lisle, J.D., Knapp, A.K., Blai , J.M., Collins, S.L., 2003. P oduc i i y esponses o al e ed ain all pa e ns in a C-4-domina ed g assland. Oecologia 137, 245–251. Field, C., Ba os, V., S ocke , T., Qin, D., Dokken, D., Ebi, K., Mas and ea, M., Mach, K., Pla ne , G.-K., Allen, S., Tigno , M., Midgley, P. (Eds.), 2012. Managing he Risks o Ex eme E en s and disas e s o Ad ance Clima e Change Adap a ion. A Special Repo o Wo king G oups I and II o he In e go e nmen al Panel on Clima e Change. Camb idge Uni e si y P ess, Camb idge, UK and New Yo k, USA. Fowle , H.J., Eks oem, M., 2009. Mul i-model ensemble es ima es o clima e change impac s on UK seasonal p ecipi a ion ex emes. In e na ional Jou nal o Clima ology 29, 385–416. Gilgen, A.K., Buchmann, N., 2009. Response o empe a e g asslands a di e en al i udes o simula ed summe d ough di e ed bu scaled wi h annual p ecipi a ion. Biogeosciences 6, 2525–2539. G ime, J.P., F idley, J.D., Askew, A.P., Thompson, K., Hodgson, J.G., Benne , C.R., 2008. Long- e m esis ance o simula ed clima e change in an in e ile g assland. P oceedings o he Na ional Academy o Sciences o he Uni ed S a es o Ame ica 105, 10028–10032. G o man, P., D iscoll, C., Fahey, T., Ha dy, J., Fi zhug, R., Tie ney, G., 2001. Colde soils in a wa me wo ld: a snow manipula ion s udy in a no he n ha dwood o es ecosys em. Biogeochemis y 56, 135–150. G oisman, P., Knigh , R., Eas e ling, D.R., Ka l, T.R., Hege l, G., Razu ae , V., 2005. T ends in in ense p ecipi a ion in he clima e eco d. Jou nal o Clima e 18, 1326–1350. Gu, L.P.J., Hanson, W.M., Pos , D.P., Kaise , B., Yang, R., 2008. The 2007 eas e n US sp ing eeze: inc eased cold damage in a wa ming wo ld? BioScience 58, 253–262. Gunde son, L.H., 2000. Ecological esilience - in heo y and applica ion. Annual Re iew o Ecology and Sys ema ics 31, 425–439. Gu schick, V.P., Bassi iRad, H., 2003. Ex eme e en s as shaping physiology, ecology, and e olu ion o plan s: owa d a uni ied de ini ion and e alua ion o hei consequences. New Phy ologis 160, 21–42. Ha pe , C.W., Blai , J.M., Fay, P.A., Knapp, A.K., Ca lisle, J.D., 2005. Inc eased ain all a iabili y and educed ain all amoun dec eases soil CO2 lux in a g assland ecosys em. Global Change Biology 11, 322–334. Haylock, M., Goodess, C., 2004. In e annual a iabili y o Eu opean ex eme win e ain all and links wi h mean la ge-scale ci cula ion. In e na ional Jou nal o Clima ology 24, 759–776. Hege l, G., Zwie s, F., B aconno , P., Gille , N., Luo, Y., Ma engo, J., O sini, A., Nicholls, N., Penne , J., S o , P., 2007. Unde s anding and A ibu ing Clima e Change. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Ma quis, M., A e y , K., Tigno , M., Mille , H. (Eds.), Clima e change 2007: he physical science basis. Con ibu ion o wo king g oup I o he ou h assessmen epo o he in e go e nmen al panel on clima e change. Camb idge Uni e si y P ess, Camb idge, UK and New Yo k, USA, pp. 663–745. Heisle , J., Wel zin, J., 2006. Va iabili y ma e s: owa ds a pe spec i e on he in luence o p ecipi a ion on e es ial ecosys ems. New Phy ologis 172, 189–192. Re e ences 33 Heisle -Whi e, J.L., Blai , J.M., Kelly, E.F., Ha money, K., Knapp, A.K., 2009. Con ingen p oduc i i y esponses o mo e ex eme ain all egimes ac oss a g assland biome. Global Change Biology 15, 2894– 2904. Heisle -Whi e, J.L., Knapp, A.K., Kelly, E.F., 2008. Inc easing p ecipi a ion e en size inc eases abo eg ound ne p ima y p oduc i i y in a semi-a id g assland. Oecologia 158, 129–140. Holling, C., 1973. Resilience and s abili y o ecological sys ems. Annual Re iew o Ecological Sys ems 4, 1–23. Hough on, J., Ding, Y., G iggs, D., Nogue , M., an de Linden, P., Dai, X., Maskell, K., Johnson, C., 2001. Clima e Change 2001: The Scien i ic Basis. Con ibu ion o Wo king G oup I o he Thi d Assessmen Repo o he In e go e nmen al Panel on Clima e Change. Camb idge Uni e si y P ess, Camb idge, UK and New Yo k, USA. Hough on, J., Mei a Filho, L., Callande , B., Ha is, N., Ka enbe g, A., Maskell, K., 1995. Clima e Change 1995- The science o clima e change.Con ibu ion o WGI o he Second Assessmen Repo o he In e go e nmen al Panel on Clima e Change. Camb idge Uni e si y P ess, Camb idge, UK and New Yo k, USA. Hough on, J., Callande , B., Va ney, S., 1992. Clima e Change 1992: The Supplemen a y Repo o he IPCC Scien i ic Assessmen . Camb idge Uni e si y P ess, Camb idge, UK, New Yo k, USA and Vic o ia, Aus alia. Hughes, L., 2000. Biological consequences o global wa ming: is he signal al eady appa en ? T ends in Ecology and E olu ion 15, 56–61. Hulme, M., 2005. Recen clima e ends. In: Lo ejoy, T., Hannah, L. (Eds.), Clima e change and biodi e si y. Yale Uni e si y P ess, Yale, pp. 31–40. Hun ing on, T., 2006. E idence o in ensi ica ion o he global wa e cycle: Re iew and syn hesis. Jou nal o Hyd ology 319, 83–95. Hu lbe , S., 1984. Pseudo eplica ion and he design o ecological expe imen s. Ecological monog aphs 54, 187– 211. Iglesias, A., Qui oga, S., Schlicken iede , J., 2010. Clima e change and ag icul u al adap a ion: assessing managemen unce ain y o ou c op ypes in Spain. Clima e Resea ch 44, 83–94. Jacob, D., 2009. Regionalisie e Szena ien des Klimawandels-Regional clima e change scena ios. Raum o schung und Raumo dnung 69, 89–96. Janska, A., Ma sik, P., Zelenko a, S., O esna, J., 2010. Cold s ess and acclima ion - wha is impo an o me abolic adjus men ? Plan Biology 12, 395–405. Jen sch, A., K eyling, J., Beie kuhnlein, C., 2007. A new gene a ion o clima e-change expe imen s: e en s, no ends. F on ie s in Ecology and he En i onmen 5, 365–374. Jylhä, K., F onzek, S., Tuomen i a, H., Ca e , T.R., Ruos eenoja, K., 2008. Changes in os , snow and Bal ic sea ice by he end o he wen y- i s cen u y based on clima e model p ojec ions o Eu ope. Clima ic Change 86, 441–462. Ka l, T.R., Knigh , R.W., 1998. Secula ends o p ecipi a ion amoun , equency, and in ensi y in he Uni ed S a es. Bulle in o he Ame ican Me eo ological Socie y 79, 231–241. Keyan ash, J., D acup, J., 2002. The quan i ica ion o d ough : An e alua ion o d ough indices. Bulle in o he Ame ican Me eo ological Socie y 83, 1167–1180. Kiou sioukis, I., Melas, D., Ze e os, C., 2010. S a is ical assessmen o changes in clima e ex emes o e G eece (1955-2002). In e na ional Jou nal o Clima ology 30, 1723–1737. Klein Tank, A., Können, G., 2003. T ends in indices o daily empe a u e and p ecipi a ion ex emes in Eu ope, 1946-99. Jou nal o Clima e, 3665–3680. Knapp, A.K., Fay, P.A., Blai , J.M., Collins, S.L., Smi h, M.D., Ca lisle, J.D., Ha pe , C.W., Danne , B.T., Le , M.S., McCa on, J.K., 2002. Rain all a iabili y, ca bon cycling, and plan species di e si y in a mesic g assland. Science 298, 2202–2205. Knapp, A.K., Beie , C., B iske, D.D., Classen, A.T., Luo, Y., Reichs ein, M., Smi h, M.D., Smi h, S.D., Bell, J.E., Fay, P.A., Heisle , J.L., Lea i , S.W., She y, R., Smi h, B., Weng, E., 2008. Consequences o mo e ex eme p ecipi a ion egimes o e es ial ecosys ems. BioScience 58, 811–821. Kod a, E., S einhaeuse , K., Ganguly, A., 2011. Pe sis ing cold ex emes unde 21s -cen u y wa ming scena ios. Geophysical Resea ch Le e s 38, doi:10.1029/2011GL047103. K eyling, J., 2010. Win e clima e change: a c i ical ac o o empe a e ege a ion pe o mance. Ecology 91, 1939–1948. K eyling, J., Beie kuhnlein, C., Jen sch, A., 2010. E ec s o soil eeze- haw cycles di e be ween expe imen al plan communi ies. Basic and applied Ecology 11, 65–75. K eyling, J., Wenigmann, M., Beie kuhnlein, C., Jen sch, A., 2008. E ec s o ex eme wea he e en s on plan p oduc i i y and issue die-back a e modi ied by communi y composi ion. Ecosys ems 11, 752–763. Lambe s, H., Chapin III, F., Pons, T., 2008. Plan Physiological Ecology. Sp inge , New Yo k. La sen, K.S., And esen, L.C., Beie , C., Jonasson, S., Albe , K.R., Ambus, P.E., A ndal, M.F., Ca e , M.S., Ch is ensen, S., Holms up, M., Ib om, A., Kongs ad, J., an de Linden, L., Ma aldo, K., Michelsen, A., Mikkelsen, T.N., Pilegaa d, K.I., P iemé, A., Ro-Poulsen, H., Schmid , I.K., Sels ed , M.B., S e nbak, K., Re e ences 34 2011. Reduced N cycling in esponse o ele a ed CO2, wa ming, and d ough in a Danish hea hland: Syn hesizing esul s o he CLIMAITE p ojec a e wo yea s o ea men s. Global Change Biology 17, 1884–1899. Li, Y., Ye, W., Wang, M., Yan, X., 2009. Clima e change and d ough : a isk assessmen o c op-yield impac s. Clima e Resea ch 39, 31–46. Llo e , F., Sisca , D., Dalmases, C., 2004. Canopy eco e y a e d ough dieback in holm-oak Medi e anean o es s o Ca alonia (NE Spain). Global Change Biology 10, 2092-2099. Lopez-Mo eno, J.I., Vicen e-Se ano, S.M., Angulo-Ma inez, M., Begue ia, S., Kenawy, A., 2010. T ends in daily p ecipi a ion on he no heas e n Ibe ian Peninsula, 1955-2006. In e na ional Jou nal o Clima ology 30, 1026–1041. Ludwig, F., Dawson, T.E., P ins, H.H., Be endse, F., K oon, H. de, 2004. Below-g ound compe i ion be ween ees and g asses may o e whelm he acili a i e e ec s o hyd aulic li . Ecology Le e s 7, 623–631. Maes e, F.T., Co ina, J., 2004. Do posi i e in e ac ions inc ease wi h abio ic s ess? - A es om a semi-a id s eppe. P oceedings o he Royal Socie y o Se ies B- Biological Sciences 271, S331-S333. Ma chand, F.L., Kockelbe gh, F., an de Vij e , B., Beyens, L., Nijs, I., 2006. A e hea and cold esis ance o a c ic species a ec ed by successi e ex eme empe a u e e en s? New Phy ologis 170, 291–300. Ma chand, F.L., Me ens, S., Kockelbe gh, F., Beyens, L., Nijs, I., 2005. Pe o mance o High A c ic und a plan s imp o ed du ing bu de e io a ed a e exposu e o a simula ed ex eme empe a u e e en .Global Change Biology 11, 2078-2089. Ma chand, P., 1996. Li e in he cold. An in oduc ion o win e ecology. Uni e si y P ess o New England, Hano e , New Hampshi e, USA. Ma engo, J.A., Rus icucci, M., Penalba, O., Renom, M., 2010. An in e compa ison o obse ed and simula ed ex eme ain all and empe a u e e en s du ing he las hal o he wen ie h cen u y: pa 2: his o ical ends. Clima ic Change 98, 509–529. McCa y, J., 2001. Ecological consequences o ecen clima e change. Conse a ion Biology 15, 320–331. Meehl, C., Zwie s, F., E ans, J., Knu son, T., Mea ns, L., Whe on, P., 2000a. T ends in ex eme wea he and clima e e en s: issues ela ed o modeling ex emes in p ojec ions o u u e clima e change. Bulle in o he Ame ican Me eo ological Socie y 81, 427–436. Meehl, C., Ka l, T., Eas e ling, D.R., Changnon, S., Pielke J ., R., Changnon, D., E ans, J., G oisman, P., Knu son, T., Kunkel, K., Mea ns, L., Pa mesan, C., Pulwa y, R., Roo , T., Syl es, R., Whe on, P., Zwie s, F., 2000b. An in oduc ion o ends in ex eme wea he and clima e e en s: obse a ions, socioeconomic impac s, e es ial ecological impac s, and model P ojec ions. Bulle in o he Ame ican Me eo ological Socie y 81, 413–416. Meehl, G.A., S ocke ,T.F., Collins, W.D., F iedlings ein, P., Gaye, A.T., G ego y, J.M., Ki oh, A., Knu i, R., Mu phy, J.M., Noda, A., Rape , S.C.B., Wa e son, I.G., Wea e , A.J., Zhao, Z.-C, 2007. Global clima e p ojec ions. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Ma quis, M., A e y , K., Tigno , M., Mille , H. (Eds.), Clima e change 2007: he physical science basis. Con ibu ion o wo king g oup I o he ou h assessmen epo o he in e go e nmen al panel on clima e change. Camb idge Uni e si y P ess, Camb idge, UK and New Yo k, USA, pp. 748–845. Mikkelsen, T.N., Beie , C., Jonasson, S., Holms up, M., Schmid , I.K., Ambus, P., Pilegaa d, K., Michelsen, A., Albe , K., And esen, L.C., A ndal, M.F., B uun, N., Ch is ensen, S., Danbaek, S., Gunde sen, P., Jo gensen, P., Linden, L.G., Kongs ad, J., Ma aldo, K., P ieme, A., Riis-Nielsen, T., Ro-Poulsen, H., S e nbak, K., Sels ed, M.B., So ensen, P., La sen, K.S., Ca e , M.S., Ib om, A., Ma inussen, T., Miglie a, F., S e d up, H., 2008. Expe imen al design o mul i ac o clima e change expe imen s wi h ele a ed CO2, wa ming and d ough : he CLIMAITE p ojec . Func ional Ecology 22, 185–195. Min, S.-K., Zhang, X., Zwie s, F.W., Hege l, G.C., 2011. Human con ibu ion o mo e-in ense p ecipi a ion ex emes. Na u e 470, 378–381. Mi anda, J.D. de, Padilla, F.M., Laza o, R., Pugnai e, I.F., 2009. Do changes in ain all pa e ns a ec semia id annual plan communi ies? Jou nal o Vege a ion Science 20, 269–276. Mo ec o , M., Mas e s, G., B own, V., Cla ke, I., Taylo , M., Whi ehouse, A., 2004. Changing p ecipi a ion pa e ns al e plan communi y dynamics and succession in an ex-a able g assland. Func ional Ecology 18. 648-655. Mo iondo, M., Bindi, M., Kundzewicz, Z.W., Szwed, M., Cho ynski, A., Ma czak, P., Radziejewski, M., McE oy, D., W e o d, A., 2010. Impac and adap a ion oppo uni ies o Eu opean ag icul u e in esponse o clima ic change and a iabili y. Mi iga ion and Adap a ion S a egies o Global Change 15, 657-679. Muelle , R.C., Scudde , C.M., Po e , M.E., Talbo T o e , R., Geh ing, C.A., Whi ham, T.G., 2005. Di e en ial ee mo ali y in esponse o se e e d ough : e idence o long- e m ege a ion shi s. Jou nal o Ecology 93, 1085–1093. Munne-Bosch, S., Aleg e, L., 2000. Changes in ca o enoids, ocophe ols and di e penes du ing d ough and eco e y, and he biological signi icance o chlo ophyll loss in Rosma inus o icinalis plan s. Plan a 210, 925–931. Re e ences 35 Naud sa, K., an den Be gen, j., Janssens, I., Nijs, I., Ceulemans, A., 2011. Does an ex eme d ough e en al e he esponse o g assland communi ies o a changing clima e? En i onmen al and Expe imen al Bo any 70, 151–157. Newman, G.S., A hu , M.A., Mulle , R.N., 2006. Abo e- and belowg ound ne p ima y p oduc ion in a empe a e mixed deciduous o es . Ecosys ems 9, 317–329. Nicholls, N., Alexande , L., 2007. Has he clima e become mo e a iable o ex eme? P og ess 1992-2006. P og ess in Physical Geog aphy 31, 77-87. Noo me s, A., McNul y, S.G., DeFo es , J.L., Sun, G., Li, Q., Chen, J., 2008. D ough du ing canopy de elopmen has las ing e ec on annual ca bon balance in a deciduous empe a e o es . New Phy ologis 179, 818–828. O‘Go man, P.A., Schneide , T., 2009. The physical basis o inc eases in p ecipi a ion ex emes in simula ions o 21s -cen u y clima e change. P oceedings o he Na ional Academy o Sciences o he Uni ed S a es o Ame ica 106, 14773–14777. Pa cellie , A., Gu buxani, S., Schmi , E., Sola y, E., Ga ido, C., 2003. Hea shock p o eins, cellula chape ones ha modula e mi ochond ial cell dea h pa hways. Biochemical and Biophysical Resea ch Communica ions 304, 505–512. Pa mesan, C., Yohe, G., 2003. A globally cohe en inge p in o clima e change impac s ac oss na u al sys ems. Na u e 421, 37–42. Penuelas, J., Go don, C., Llo ens, L., Nielsen, T., Tie ema, A., Beie , C., B una, P., Emme , B., Es ia e, M., Go issen, A., 2004. Nonin usi e ield expe imen s show di e en plan esponses o wa ming and d ough among si es, seasons, and species in a no h-sou h Eu opean g adien . Ecosys ems 7, 598–612. Penuelas, J., P ie o, P., Beie , C., Cesa accio, C., Angelis, P. de, Da o, G. de, Emme , B.A., Es ia e, M., Ga adnai, J., Go issen, A., Lang, E.K., K oeel-Dulay, G., Llo ens, L., Pellizza o, G., Riis-Nielsen, T., Schmid , I.K., Si ca, C., Sowe by, A., Spano, D., Tie ema, A., 2007. Response o plan species ichness and p ima y p oduc i i y in sh ublands along a no h-sou h g adien in Eu ope o se en yea s o expe imen al wa ming and d ough : educ ions in p ima y p oduc i i y in he hea and d ough yea o 2003. Global Change Biology 13, 2563–2581. Raisanen, J., Joelsson, R., 2001. Changes in a e age and ex eme p ecipi a ion in wo egional clima e model expe imen s. Tellus Se ies A-Dynamic Me eo ology and Oceanog aphy 53, 547–566. Rapacz, M., Plazek, A., Niemczyk, E., 2000. F os de-acclima ion o ba ley (Ho deum ulga e L.) and meadow escue (Fes uca p a ensis Huds.). Rela ionship be ween soluble ca bohyd a e con en and esis ance o os and he ungal pa hogen Bipola is so okiniana (Sacc.) Shoem. Annals o Bo any 86, 539–545. Rigby, J.R., Po po a o, A., 2008. Sp ing os isk in a changing clima e. Geophysical Resea ch Le e s 35. Robock, A., Vinnikow, K., S ini asan, G., En in, J., Hollinge , S., Sp e anskaya, N., Liu, S., Namkhai, A., 2000. The global soil mois u e da a bank. Bulle in o he Ame ican Me eo ological Socie y 81, 1281–1299. Sakai, A., La che , W., 1987. F os su i al o plan s. Sp inge , Be lin, London. Salinge , M., 2005. T ends in ex eme wea he and clima e e en s: issues ela ed o modeling ex emes in p ojec ions o u u e clima e change. Clima ic Change 70, 9–29. Schae , C., Jend i zky, G., 2004. Clima e change: Ho news om summe 2003. Na u e 432, 559–560. Sche e , M., Ca pen e , S., Foley, J.A., Folke, C., Walke , B., 2001. Ca as ophic shi s in ecosys ems. Na u e 413, 591–596. Schönwiese, C.-D., S aege , T., S. T ömel, S., 2005. Klimawandel und Ex eme eignisse in Deu schland. Klimas a usbe ich 2005, Deu sche We e diens . h p://www.geo.uni- ank u .de/iau/klima/PDF_Da eien/Sw_e _al-Klimawandel_Ex em-KSB2005.pd Sch ö e , D., C ame , W., Leemans, R., P en ice, I., A aujo, M., A nell, N., Bondeau, A., Bugmann, H., Ca e , T., G acia, C.A., La Vega-Leine , A. de, E ha d, M., Ewe , F., Glendining, M., House, J., Kankaanpää, S., Klein, R., La o el, S., Lindne , M., Me zge , M., Meye , J., Mi chell, T., Regins e , I., Rounse ell, M., Saba é, S., Si ch, S., Smi h, B., Smi h, J., Smi h, P., Sykes, M., Thonicke, K., Thuille , W., Tuck, G., Zaehle, S., Zie l, B., 2005. Ecosys em se ice supply and ulne abili y o global change in Eu ope. Science 310, 1333–1337. Schulze, E., Beck, E., Mülle -Hohens ein, K., 2005. Plan Ecology. Sp inge , Be lin, London. Smi h, M.D., 2011a. An ecological pe spec i e on ex eme clima ic e en s: a syn he ic de ini ion and amewo k o guide u u e esea ch. Jou nal o Ecology 99, 656–663. Smi h, M.D., 2011b. The ecological ole o clima e ex emes: cu en unde s anding and u u e p ospec s. Jou nal o Ecology 99, 651–655. Sowe by, A., Emme , B., Tie ema, A., Beie , C., 2008. Con as ing e ec s o epea ed summe d ough on soil ca bon e lux in hyd ic and mesic hea hland soils. Global Change Biology 14, 2388–2404. S imbeck, G., Schabe g, P., Hayes, D., Shane, J., Hawley, G., 1995. Midwin e deha dening o mon ane ed sp uce du ing a na u al haw. Canadian Jou nal o Fo es Resea ch 25, 2040–2044. Su le, K.B., Thomsen, M.A., Powe , M.E., 2007. Species in e ac ions e e se g assland esponses o changing clima e. Science 315, 640–642. Re e ences 36 Tebaldi, C., Hayhoe, K., A blas e , J.M., Meehl, G.A., 2006. Going o he ex emes. Clima ic Change 79, 185– 211. Thomas, C., Came on, A., G een, E., Bakkenes, M., Beaumon , L., Collingham, Y., E asmus, B., Fe ei a Siquie e, M. de, G ainge , A., Hannah, L., Hughes, L., Hun ley, B., an Jaa s eld, A., Midgley, G., Miles, L., O ega-Hue a, M., Pe e son, T., Philipps, O., Williams, S., 2004. Ex inc ion isk om clima e change. Na u e 427, 145–148. Thomashow, M., 1999. Plan cold acclima ion: F eezing Tole ance Genes and Regula o y Mechanisms. Annual Re iews in Plan Physiology and Plan Molecula Biolology 50, 571–599. Thompson, I., Mackey, B., McNul y, S., Mossele , A., 2009. Fo es esilience, biodi e si y, and clima e change. A syn hesis o he biodi e si y/ esilience/s abili y ela ionsship in o es ecosys ems. Sec e a ia o he con en ion on biodi e si y. CBD Technical Se iex no. 43, Mon eal. Thuille , W., 2007. Clima e change and he ecologis . Na u e 448. 550-552 Thuille , W., La o el, S., A aujo, M., Sykes, M., P en ice, I., 2005. Clima e change h ea s o plan di e si y in Eu ope. P oceedings o he Na ional Academy o Sciences o he Uni ed S a es o Ame ica 102, 8245–8250. Tielbo ge , K., Kadmon, R., 2000. Tempo al en i onmen al a ia ion ips he balance be ween acili a ion and in e e ence in dese plan s. Ecology 81, 1544–1553. Tobe man, H., E ans, C.D., F eeman, C., Fenne , N., Whi e, M., Emme , B.A., A z, R.R.E., 2008. Summe d ough e ec s upon soil and li e ex acellula phenol oxidase ac i i y and soluble ca bon elease in an upland Calluna hea hland. Soil Biology and Biochemis y 40, 1519–1532. T enbe h, K., Jones P.D, Ambenje P., Boja iu, R., Eas e ling, D., Klein Tank, A., Pa ke , D., Rahimzadeh, F., Renwick, J., Rus icucci, M., Soden, B., Zhai, P., 2007. Obse a ions: Su ace and A mosphe ic Clima e Change. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Ma quis, M., A e y , K., Tigno , M., Mille , H. (Eds.), Clima e change 2007: he physical science basis. Con ibu ion o wo king g oup I o he ou h assessmen epo o he in e go e nmen al panel on clima e change. Camb idge Uni e si y P ess, Camb idge, UK and New Yo k, USA, pp. 235–336. T enbe h, K.E., Dai, A., Rasmussen, R.M., Pa sons, D.B., 2003. The changing cha ac e o p ecipi a ion. Bulle in o he Ame ican Me eo ological Socie y 84, 1205–1217. an Pee , L., Nijs, I., Reheul, D., Cauwe , B. de, 2004. Species ichness and suscep ibili y o hea and d ough ex emes in syn hesized g assland ecosys ems: composi ional s physiological e ec s. Func ional Ecology 18, 769–778. an Ruij en, J., Be endse, F., 2010. Di e si y enhances communi y eco e y, bu no esis ance, a e d ough . Jou nal o Ecology 98, 81–86. Visse , M., Holleman, L., 2001. Wa me sp ings dis up he synch ony o oak and win e mo h phenology. P oceedings o he Royal Socie y o Se ies B- Biological Sciences 268, 289–294. Wal he , G., Pos , E., Con ey, P., Menzel, A., Pa mesan, C., Beebee, T., F omen in, J.-C., Hoegh-Guldbe g, O., Bai lein, F., 2002. Ecological esponse o ecen clima e change. Na u e 416, 389–395. Wang, H., Fu, R., Kuma , A., Li, W., 2010. In ensi ica ion o summe ain all a iabili y in he Sou heas e n Uni ed S a es du ing ecen decades. Jou nal o Hyd ome eo ology 11, 1007–1018. Wasilewska, A., Vlad, F., Si ichand a, C., Redko, Y., Jammes, F., Valon, C., F ey, N.F.d., Leung, J., 2008. An upda e on abscisic acid signaling in plan s and mo e … Molecula Plan 1, 198–217. Woldendo p, G., Hill, M., Do an, R., Ball, M., 2008. F os in a u u e clima e: modelling in e ac i e e ec s o wa me empe a u es and ising a mosphe ic CO2 on he incidence and se e i y o os damage in a empe a e e e g een (Eucalyp us pauci lo a). Global Change Biology 14, 294–308. Zebisch, M., G o hmann, T., Sch ö e , D., Hasse, C., F i sch, U., C ame , W., 2012. Clima e Change- Klimawandel in Deu schland. Vulne abili ä und Anpassungss a egienklimasensi i e Sys eme-UBA Fo schungsbe ich , www.umwel da en.de/publika ionen/ pd -l/2947.pd . Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 37 Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y Jou nal o Ecology, 2011, 99: 689–702. Anke Jen sch1*, Jue gen K eyling2, Michael Elme 3, Ellen Gellesch2, B uno Glase 4, Ke s in G an 2, Roman Hein2, Ma co La a4, Heyda Mi zae6, S e anie E. Nadle 2, Lau a Nagy1, Denis O ieno6, Ka in P i sch7, Uwe Rasche 8, Ma in Schädle 9, Michael Schlo e 7, B ajesh K. Singh10, Ju a S adle 9, Julia Wal e 11, Camilla Wells ein2, Jens Wöllecke3 and Ca l Beie kuhnlein2 1 Dis u bance Ecology, Uni e si y o Bay eu h, D-95440 Bay eu h, Ge many. E-mail: anke.jen sch@uni- bay eu h.de 2 Biogeog aphy, Uni e si y o Bay eu h, 95440 Bay eu h, Ge many. 3 Soil P o ec ion and Recul i a ion, BTU Co bus, Kon ad-Wachsmann-Allee 6, 03046 Co bus, Ge many. 4 Soilphysics, Uni e si y o Bay eu h, 95449 Bay eu h, Ge many 5 Soil Biogeochemis y, Ma in-Lu he Uni e si y o Halle-Wi enbe g, on Seckendo pla z 3, 06120 Halle, Ge many. 6 Plan Ecology, Uni e si y o Bay eu h, 95440 Bay eu h, Ge many. 7 Te es ial Ecogene ics, Helmhol z Zen um Muenchen, Ge man Resea ch Cen e o En i onmen al Heal h, Ingols aed e Lands aβe 1, 85764 Neuhe be g, Ge many. 8 Ins i u e o Chemis y and Dynamics o he Geosphe e, ICG-3, Phy osphe e, Resea ch Cen e Jülich, Leo- B and -S ., D-52425 Jülich, Ge many. 9 Communi y Ecology, Helmhol z-Cen e o En i onmen al Resea ch – UFZ, Theodo -Liese -S . 4, 06110 Halle, Ge many. 10 Cen e o Plan s and En i onmen , Uni e si y o Wes e n Sydney, Pen i h Sou h DC, NSW, Aus alia. 11 Conse a ion Biology, UFZ-Helmhol z, Cen e ü En i onmen al Resea ch, Pe mose s . 15, 04318 Leipzig, Ge many. Co esponding au ho : P o . D . Anke Jen sch, E-mail: [email p o ec ed]e Running i le: D ough e ec on mul iple ecosys em se ices Summa y 1. S udying he e ec s o ex eme clima ic o wea he e en s such as d ough and hea wa es on biodi e si y and ecosys em unc ions is one o he mos impo an ace s o clima e change esea ch. In pa icula , p ima y p oduc ion is amoun ing o he common cu ency in ield expe imen s wo ldwide. Ra ely, howe e , a e mul iple ecosys em unc ions measu ed in a single s udy in o de o add ess gene al pa e ns ac oss di e en ca ego ies o esponses and o analyse e ec s o clima e ex emes on a ious ecosys em unc ions. 2. We se up a long- e m ield expe imen , whe e we applied ecu en se e e d ough e en s annually o i e consecu i e yea s o cons uc ed g assland communi ies in cen al Eu ope. Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 38 The 32 esponse pa ame e s s udied we e closely ela ed o ecosys em unc ions such as p ima y p oduc ion, nu ien cycling, ca bon ixa ion, wa e egula ion and communi y s abili y. 3. Su p isingly, in he ace o se e e d ough , abo e- and below-g ound p ima y p oduc ion o plan s emained s able ac oss all yea s o he d ough manipula ion. 4. Ye , se e e d ough signi ican ly educed below-g ound pe o mance o mic obes in soil indica ed by educed soil espi a ion, mic obial biomass and cellulose decomposi ion a es as well as myco hiza ion a es. Fu he mo e, d ough educed lea wa e po en ial, lea gas exchange and lea p o ein con en , while inc easing maximum up ake capaci y, lea ca bon iso ope signa u e and lea ca bohyd a e con en . Wi h ega d o communi y s abili y, d ough induced complemen a y plan –plan in e ac ions and shi s in lowe phenology, and dec eased in asibili y o plan communi ies and p ima y consume abundance. 5. Syn hesis. Ou esul s p o ide he i s ield-based expe imen al e idence ha clima e ex emes ini ia e plan physiological p ocesses, which may se e o egula e ecosys em p oduc i i y. A po en ial eason o di e en dynamics in a ious ecosys em se ices acing ex eme clima ic e en s may lie in he empo al hie a chy o pa e ns o as e sus slow esponse Such da a on mul iple esponse pa ame e s wi hin clima e change expe imen s os e he unde s anding o mechanisms o esilience, o syne gisms o decoupling o biogeochemical p ocesses, and o undamen al esponse dynamics o d ough a he ecosys em le el including po en ial ipping poin s and h esholds o egime shi . Fu u e wo k is needed o elucida e he ole o biodi e si y and o bio ic in e ac ions in modula ing ecosys em esponse o ex eme clima ic e en s. Keywo ds: below-g ound, compe i ion, decomposi ion, in asion, lea chemis y, mic obial, phenology, plan –clima e in e ac ions, p ecipi a ion change, p oduc i i y In oduc ion Cu en ly, knowledge abou ecological esponses o clima e change is based la gely on e ec s o clima ic ends such as g adual wa ming, p ecipi a ion change and CO2 en ichmen . Howe e , he magni ude and equency o ex eme clima ic o wea he e en s such as se e e d ough , hea wa es, hea y ain and la e os e en s a e expec ed o inc ease in he nea u u e (IPCC 2007; O’Go man & Schneide 2009). Thus, p edic ions o e ec s o clima e Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 45 d ough . The maximum quan um e iciency o pho osys em II was calcula ed as F /Fm. Va iable luo escence (F ) and maximum luo escence (Fm) we e measu ed be o e dawn. Va iable luo escence was calcula ed as Fm-F0, Fm being he maximum luo escence o he da k-adap ed lea a e applying a sa u a ing ligh pulse and F0 being he s eady-s a e luo escence yield o he da k-adap ed lea (Maxwell & Johnson 2000). To enable a compa ison be ween absolu e luo escence alues, a luo escence s anda d ma e ial was measu ed be o e dawn and calcula ed as F /Fm (F = Fm – F0) (Maxwell & Johnson 2000). Absolu e F0 and Fm alues we e aken o sepa a e he e ec s o pho odamage, becoming appa en wi h an inc ease o F0, om he e ec s o pho op o ec ion ela ed o enhanced non- pho ochemical quenching, becoming appa en wi h a dec ease in Fm (Wal e e al., 2011). Lea gas exchange Ca bon dioxide assimila ion (A) a he lea was moni o ed in A. ela ius in all he g assland communi ies. (No da a could be ob ained om H. lana us in he pa icula yea o da a mining due o i s lea e s a us.) A se ies o weekly measu emen s we e ca ied ou using a po able gas exchange sys em (LI-6400, LI-Co , Lincoln, NE, USA). A se o 3 g ass u s on each plo we e iden i ied and ma ked o measu emen s. On any measu emen day, 2-3 sui able lea blades selec ed om each o he u s pe plo we e se pa allel in he cu e e, wi h hei uppe su aces well exposed so ha hey we e ully illumina ed du ing measu emen s. E e y u n o measu emen s las ed one o wo minu es, when a s eady s a e was a ained and a se o 10 eadings pe measu emen logged a 10-s in e als. The selec ed lea es we e ma ked and simila lea es we e moni o ed ei he du ing midday (12:00 o 14:00 h) o h oughou he day ( om sun ise o sunse ), when diu nal cou se measu emen s we e conduc ed. The measu ed lea es we e hen excised a he end o he measu emen pe iod and he lea a ea (LA) o he sec ion o lea enclosed in he cu e e de e mined using lea a ea me e CI-202 CID, Camas, WA, USA. Lea a ea in o ma ion was hen used o s anda dize he lea gas exchange da a. Soil espi a ion In si u a es o soil espi a ion we e measu ed using a po able CO2 in a ed gas analyse (EGM-4, PP Sys ems, Amesbu y, USA) linked o a soil espi a ion chambe (SRC-1, PP Sys em, Amesbu y, USA). A he beginning o he ege a ion pe iod, pe manen PVC colla s (10 cm diame e , 5 cm heigh , ligh g ey colou ) we e ins alled in e e y plo wi h a 1- cm edge abo e soil su ace o ealize a closed sys em when he soil espi a ion chambe was placed on he colla du ing measu emen . The day be o e each measu emen , all abo e- g ound ege a ion was emo ed om he colla using scisso s. Du ing he ime ame o 8:00 Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 46 o 12:00, he soil espi a ion chambe was placed o 240 seconds on he colla o e e y plo . An in e nal an ealized he e en dis ibu ion o ai and he in a ed gas analyse moni o ed he build-up o CO2 wi hin he sys em. The a es o soil espi a ion we e de e mined om his by i ing a quad a ic equa ion o he change in CO2 concen a ion wi h ime. Fo his s udy, we analysed he soil espi a ion a es a second 240 o each high-di e si y g assland plo including A. ela ius, H. lana us, P. lanceola a and G. p a ense on he las day o d ough manipula ion. Maximum lea and canopy up ake a es Ne ecosys em CO2 exchange was measu ed wi h chambe s on 40 × 40 cm ames es ablished on each o he ea men plo s. Daily cou se o ne ecosys em CO2 exchange (NEE) was measu ed using manually ope a ed, closed gas exchange canopy chambe s. Ligh - esponse cu es depic ing he ne pho osyn he ic CO2 up ake a e (A) o plan s a any measu ing ime we e ob ained om lea -le el gas-exchange measu emen s by i ing an empi ical ec angula hype bola model (Gilmano e al. 2005): NEE = (α+Q / αQ-β) - γ, whe e α is he ini ial slope o he ligh - esponse cu e and an app oxima ion o he canopy ligh u iliza ion e iciency (mol CO2/ mol PAR), β is he maximum CO2 up ake capaci y (µmol m–2 s–1), Q is he pho osyn he ically ac i e adia ion (PAR, in µmol m–2 s–1), and γ is an app oxima ion o he a e age day ime ecosys em espi a ion (lmol m–2 s –1). An app oxima ion o maximum canopy up ake capaci y was ex apola ed om lea -le el measu emen s. Canopy ne ecosys em exchange a e (NEE) was es ima ed om lea pho osyn he ic a e a sa u a ing ligh in ensi ies (i was shown ha A a PAR = 2000 µmol m– 2 s–1 co ela es well wi h canopy NEE). Maximum g oss p ima y p oduc i i y (GPPmax) was calcula ed as: GPPmax = NEE2000 – Reco, whe e A2000 is he maximum lea pho osyn he ic a e a a sa u a ing le el o ligh in ensi y andReco is he co ec ed espi a ion e m (γ) ob ained om he model. Nu ien cycling In si u decomposi ion a e o cellulose Biological ac i i y o soil auna and mic oo ganisms was de e mined indi ec ly om he decay o cellulose using mini-con aine ubes (K eyling e al. 2008a). In o al, 864 mini- con aine s we e illed wi h 0.2 g o cellulose (poo in phospho us, Schleiche & Schüll, Dassel, Ge many) each, closed wi h a 2-mm mesh, and pu in o con aine ubes, consis ing o 12 mini-con aine s each. Two ubes we e bu ied ho izon ally 1 cm below soil su ace in each g assland plo . A e 94 days, one ube pe plo was ha es ed, whe eas he o he s we e Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 47 ha es ed a e 186 days. A e ca e ul cleaning and d ying, he decay o cellulose was de e mined by sub ac ing inal ashes- ee d y mass om ini ial d y mass (105 °C). Myco hizal coloniza ion One comple e plan indi idual o P.lanceola a was aken om each plo on he las day o d ough using a soil co e sample wi h 5 cm diame e (Eijkelkamp; Ne he lands). This pa icula species was chosen, because p e-analysis e ealed highe e ec s o d ough on myco hizaal coloniza ion o P. lanceola a han on ha o o he species es ed. Roo s we e cu o and ixed in o malin-alcoholic-acid (50 % E hanol, 40 % H2O, 7.5 % o malin, 2.5 % acidic acid), and s ained wi h 5 % blue ink inega solu ion a e boiling in 10 % KOH. A e wa ds, myco hiza ion a ios we e de e mined by scanning 15 cm ine oo s o each sample o a buscules and esicules unde a mic oscope (400×) using he “magni ied in e sec ion me hod” (McGonigle e al. 1990). Soil mic obial ni ogen pool Soil mic obial ni ogen was ex ac ed om esh soil acco ding o a modi ied chlo o o m umiga ion–ex ac ion me hod (B ookes e al. 1985). A e chlo o o m umiga ion (24 h a oom empe a u e), dissol ed o ganic and mic obial N was ex ac ed wi h 50 mL 0.5 M K2SO4 and quan i ied (DIMA TOC-100, Dima ec, Essen, Ge many). Mic obial biomass and ela i e abundance o mic obial g oups we e measu ed using phospholipid a y acid (PLFA) analysis as desc ibed (Singh e al. 2006). Po en ial soil enzyme ac i i ies Fo soil enzyme ac i i y measu emen s, enzymes in ol ed in ca bon, ni ogen and phospho us cycling we e selec ed, hus add essing impo an mic obial soil unc ions (Wald op & Fi es one 2006). The enzyme ac i i ies es ed we e acid phospha ase clea ing o ganically bound phospha e, cellobiohyd olase, β-xylosidase and β-glucosidase ela ed o he deg ada ion o plan cell wall componen s and N-ace ylglucosaminidase ep esen ing chi inases ha deg ade chi in om ungal o a h opod o igin. Soil samples o de e mining soil enzyme ac i i ies we e collec ed immedia ely a e inishing he d ough manipula ions (K eyling e al. 2008a). Fou samples pe plo (dep h 0 – 5 cm) we e combined, mixed and kep a 4 °C un il u he p ocessing wi hin 4 weeks a e sampling. Soil suspensions (0.4 g esh soil in 40 mL H2O) we e p epa ed om each sample. The assay is based on he enzyma ic clea age o he below-de ailed me hylumbelli e one (MU) coupled subs a es and he subsequen de ec ion o MU eleased du ing incuba ion. In b ie , 50 μL pe well o soil suspensions ( h ee eplica es each sample) we e dispe sed in mic opla es and 100 μL o subs a e solu ions we e added o s a he eac ions. A e s opping he eac ion wi h 100 μL Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 48 o 2.5 M T is bu e and cen i uga ion, MU concen a ions we e de e mined on a luo escence spec ome e a exci a ion/emission wa eleng hs o 365/450 nm, espec i ely. The ollowing enzyme subs a es we e used wi h he incuba ion imes gi en: MUF-phospha e, 20 min; MUF-xyloside, 1 h; MUF-cellobiohyd o u ane, 1 h; MUF-N-ace yl-β-glucosaminide, 40 min; MUF-β-glucoside, 1 h. Subs a e concen a ions in he incuba ion mix we e 500 μM excep o MUF-cellobiohyd o u ane wi h 400 μM. To accoun o quenching and o calcula e he amoun o MUF eleased, calib a ion cu es we e included wi h 50 μL o soil samples as in he incuba ion wells and MUF-solu ions o gi e a inal amoun o 0 - 500 pmol pe well. Nega i e con ols o au o luo escence o subs a es we e also included. Enzyme ac i i ies a e exp essed as MUF- elease pe g am soil d y weigh pe hou . Plan -a ailable soil ni a e and ammonium Plan -a ailable ni ogen was ex ac ed om ou homogenized, sie ed (< 2 mm), mixed samples o he uppe soil laye (0-10 cm) o each plo sampled in July using a 1 M KCl solu ion a e il a ion (Ro h, Ka ls uhe Ge many, Typ 15 A Blauband) (K eyling e al. 2010). Ni a e and ammonium we e quan i ied using low injec ion analysis (FIA, MLE D esden FIA-LAB). Lea ca bon o ni ogen a io Lea ca bon (C), lea ni ogen (N) and C:N a ios we e measu ed om mixed samples o wo sun-exposed lea es o i e indi idual plan s pe species and plo , sampled in July (K eyling, Beie kuhnlein & Jen sch 2010). The samples we e o en-d ied o 48 h a 75 °C. The d y lea es we e ball-milled and subsamples o 1 mg analysed wi h an elemen al analyse in a mass spec ome e using ConFlo III in e ace. Plan -a ailable ni ogen was ex ac ed om ou homogenized, sie ed (2 mm) and il e ed (Ro h, Ge many, Typ 15A Blauband) mixed samples o he uppe soil laye (0–10 cm) o each plo using a 1 M KCl solu ion. Lea p o ein con en To al p o ein con en in µg pe mg esh weigh was de e mined as a p oxy o nu i i e alue o he legume key species H. lana us, which was g owing in all plo s. One lea sample pe plo was aken on he las day o d ough ea men , ozen in liquid ni ogen and eeze-d ied o de e mine p o ein-bound amino acids. Amino acids o he p o ein ac ion we e ex ac ed. Amino acid concen a ions we e measu ed wi h an ion exchange ch oma og aph (Bio onik, amino acid analyse LC 3000) and p o ein con en was calcula ed by pooling he con en o each amino acid in he p o ein ac ion. Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 49 Lea ni ogen iso ope signal Equally aged, sou h- acing lea es o A. ela ius we e collec ed and o en-d ied a 60 °C o 48 h, and hen ine-milled. Na u al abundance o δ15N and o al ni ogen concen a ion we e analysed using an elemen al analyse (EA 3000, Eu oVec o , I aly) coupled online o a ConFlo III in e ace (The mo Elec on, B emen, Ge many) connec ed o an iso ope- a io mass spec ome e (MAT 253, The mo Elec on, B emen, Ge many) . The δ15N alues we e calcula ed as: δ15N [‰] = ( Rsample/Rs anda d ) -1)*1000, whe e R ep esen s he a io o 15N:14N iso opes. As s anda d, (ni ogen in) ai was used. Communi y esponses In asibili y In asibili y o he expe imen al communi ies was eco ded h ee imes pe yea : be o e and a e he d ough manipula ions in ea ly summe , and in all (K eyling e al. 2008c). In ading plan indi iduals we e collec ed om he inne squa e me e o each plo and subsequen ly sepa a ed by species. Remo al ook place only a e he i s ue lea es (a e he co yledons) eme ged, bu mos specimens we e conside ably olde han his and clea ly es ablished in he s and. A his poin in de elopmen , we expec ed ha numbe o indi iduals gi e a measu e o es ablished in ade s a he han chance ge mina ions. Fo each plo , he numbe o indi iduals was de e mined. The plan ed a ge species o he expe imen we e emo ed om he subsequen analysis. Tes s con i med ha ge mina ion om he soil seed bank was negligible a e one yea . Thus, in asibili y was only based on species in ading om he ma ix ege a ion. Plan composi ional change The measu emen s o abo e-g ound species-speci ic co e (s. abo e) we e used o e alua e shi s in he species abundance dis ibu ions o he a i icial plan assemblages. Composi ional change o each indi idual plo was e alua ed by compa ing he species abundance dis ibu ion a each ime s ep o he ini ial species abundance dis ibu ion ( i e weeks a e plan ing) by he B ay–Cu is index. Compe i i e e ec / acili a i e e ec The Rela i e Neighbou E ec calcula es he e ec o neighbou s ela i e o he plan wi h he g ea es pe o mance: RNE = Pcon -Pmix/x wi h x = Pcon i Pcon > Pmix and x = Pmix i Pmix> Pcon , whe e RNE = Rela i e neighbou e ec (-1 ≤ RNE ≤ +1), Pcon = pe o mance Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 50 pe plan o a plan g owing alone ,Pmix = pe o mance pe plan o a plan g owing in mix u e. Nega i e alues indica e acili a ion, and posi i e alues indica e compe i ion (Ma kham & Chanway 1996). Senescence Tissue die-back was quan i ied by co e measu emen s o s anding-dead plan o gans (K eyling e al. 2008d). A pin-poin me hod was applied, eco ding he p esence o plan o gans in gene al and he p esence o each species sepa a ely a 100 e ically inse ed s eel needles. These alues we e ea ed as pe cen age co e . The measu emen was epea ed ou imes o e he cou se o he ege a ion pe iod. Va iabili y in leng h o lowe ing Fo each species, weekly obse a ions o he lowe ing s a us o ou indi iduals pe plo and species we e ca ied ou (Jen sch e al. 2009). Indi iduals we e coun ed as ‘ lowe ing’ when he an he s we e isible in a leas one lowe . Flowe ing leng h was calcula ed as he di e ence be ween he da es o he 25 and 75 pe cen ile o he lowe ing cu e o e ime. Va iabili y in leng h o lowe ing was ob ained as he s anda d de ia ion be ween all species o each ea men (d ough and con ol) sepa a ely. S a is ical signi icance o di e ence in a iabili y was e alua ed by he Le ene es . Va iabili y in lowe phenology Flowe phenology was ob ained om he same da a as leng h o lowe ing (see abo e). As a su oga e, he mid- lowe ing da e was calcula ed o each species and plo , i.e. he da e o he 50 pe cen ile o he lowe ing cu e o e ime. Va iabili y in lowe phenology was exp essed as he s anda d de ia ion be ween all species o each ea men (d ough and con ol) sepa a ely. S a is ical signi icance o di e ence in a iabili y was e alua ed by he Le ene es . Resis ance o he bi o y (phenol con en ) Fo analysis o o al soluble ca bohyd a es and o al phenolics, h ee mixed samples o a leas wo plan s pe plo we e aken a he end o he d ough pe iod, immedia ely ozen in liquid ni ogen and lyophilized (n=15). Thi y milig ams we e ex ac ed in 50 % me hanol. To al soluble ca bohyd a es we e analysed using he an h one me hod wi h glucose as a s anda d. Ex inc ion was measu ed a 620 nm. To al phenols we e analysed using FolinCiocal eu’s eagen and ca echin as a s anda d and measu ing ex inc ion a 750 nm. Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 51 P ima y consume abundance Richness was sampled in June in one ci cula a ea (40 cm diame e ) in each g assland plo using a D-Vac suc ion sample (eco ech GmbH, Bonn, Ge many). Fo each plo , he sampling bag was emo ed and all sampled ma e ial was s o ed in e hanol. A h opod samples we e quan i ied as he o al numbe o indi iduals and iden i ied a leas o o de le el. Howe e , some axa we e iden i ied o he amily le el ( amilies wi hin he Coleop e a, Hemip e a, mos Hymenop e a) and in one case o genus le el (Psylliodes [Ch ysomelidae]. The use o highe axonomic le els has been shown o p oduce a good app oxima ion o o al species ichness (Biaggini e al. 2007). S a is ical Analyses Linea Models combined wi h analysis o a iance (ANOVA) we e applied o es o signi ican di e ences be ween g oups a single poin s o ime, while aking he spli -plo design in o accoun . Homogeneous g oups o ac o combina ions (d ough manipula ion, ege a ion ype, di e si y le el) we e iden i ied by Tukey HSD pos hoc compa isons. Le el o signi icance was se o p<0.05. S a is ical signi icance o di e ence in a iabili y o leng h in lowe ing was e alua ed by he Le ene es . Fo ime se ies, Linea Mixed-E ec s Models we e employed o es o e ec s o d ough manipula ion and di e si y and hei espec i e in e ac ions while aking he spli -plo design and he epea ed measu es in o accoun ( ime used as andom ac o ). When no signi ican in e ac ion was ound, he model was simpli ied by using only he d ough manipula ions as ixed e ec s and ime as andom e ec . Signi icance o di e ences (p < 0.05) was e alua ed by Ma ko Chain Mon e Ca lo sampling o 1000 pe mu a ions. Linea Mixed-E ec s Models we e conduc ed wi h he unc ion ‘lme ’ (Ba es & Sa ka 2007). P io o s a is ical analysis, da a was log- o squa e- oo - ans o med, i condi ions o no mali y we e no me , o o imp o e homogenei y o a iances. Bo h cha ac e is ics we e es ed by examining he esiduals e sus i ed plo s and he no mal qq-plo s o he linea models. All s a is ical analyses we e pe o med using R. Resul s The e ec s o d ough on all measu ed ecosys em p ope ies a e summa ized in Fig. 1 using esponse a ios o s anda dize he e ec size o he se e e d ough ea men . Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 52 Figu e 1: E ec s o ecu en se e e d ough e en s on 32 esponse pa ame e s o ganized in o ecosys em unc ions. All da a we e collec ed a he EVENT I expe imen al si e (Jen sch, K eyling & Beie kuhnlein 2007) in Cen al Eu ope du ing he yea s 2005 - 2009. A pa ame e is ma ked as signi ican ( illed black ba ), i da a o a leas one yea showed signi ican di e ences be ween d ough and ambien condi ions (ANOVA). Da a shown ep esen maximum e ec s om yea s wi h highes d ough e ec s, a e aged o e all h ee expe imen al g assland communi ies. Fo e e ences o published de ails please e e o Ma e ials and me hods sec ion Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 53 Wa e egula ion Se e e d ough signi ican ly educed soil mois u e du ing he manipula ion pe iods in all yea s (Figs 1, 2). Figu e 2 Soil mois u e in he EVENT expe imen a -2 o -7 cm du ing manipula ion (ligh g ey boxes) and eco e y a e ex eme d ough o con ol (black line) and d ough (g ey line). MJJA = May, June, July, Augus . Plan a ailable wa e is shown be ween he dashed lines: pe manen wil ing poin (pF = 4.2) and ield capaci y (pF = 1.8). See Ma e ials and me hods o echnical de ails. A high a iabili y bo h wi hin yea s and be ween yea s is e iden due o in e -annual a iabili y o p ecipi a ion (Table 2). E en hough absolu e minima in soil mois u e we e simila o d ough and con ol in mos yea s, soil mois u e o he d ough plo s emained conside ably longe below he app oxima e pe manen wil ing poin (pF = 4.2) o he soil subs a e. The manipula ion e ec anished wi hin days o all yea s excep 2009, whe e a lag phase o abou wo mon hs un il Augus occu ed. Table 2 Tempe a u e and p ecipi a ion sums (added daily amoun ) o each yea un il he s a o he d ough manipula ion and he espec i e al e a ion om he long- e m mean (1971-2000, da a: Ge man Wea he Se ice s a ion Bay eu h) yea empe a u e sum (1 Janua y o s a o manipula ion) ela i e change o empe a u e sum compa ed o long- e m mean (%) p ecipi a ion sum (1 Janua y o s a o manipula ion) ela i e change o p ecipi a ion sum compa ed o long- e m mean (%) 2005 824.7 -3 259.7 -9 2006 394.7 -38 208.3 +10 2007 978.7 +77 258.6 +9 2008 757.6 +40 282.2 +19 2009 574.9 +4 246.4 +4 Fu he , d ough dec eased lea wa e po en ial, while inc easing lea ca bon iso ope signal in some species (Figu e 1). P ima y p oduc ion A he le el o he g assland communi y o ecosys em, espec i ely, local, annually ecu en 100-yea and 1000-yea ex eme d ough e en s had no signi ican e ec on a ious Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 54 p ocesses ha con ibu e o p ima y p oduc ion in any o he i e yea s om 2005 o 2009 (Figs 1, 3). Su p isingly, nei he abo e-g ound p ima y p oduc ion (ANPP), no g een co e o ege a ion o below-g ound p oduc ion eco ded as oo leng h in he main oo ing ho izon we e a ec ed by d ough (Figs 1, 3). Figu e 3: (a) Abo e-g ound Ne P ima y P oduc ion (ANPP) , (b) co e o g een biomass, and (c) oo leng h o e i e g owing seasons (mean ± SE o e all species composi ions in g assland, n = 15 pe da a poin ). An as e isk ma ks signi ican ea men e ec s (ANOVA, Tukey HSD pos hoc compa ison: p < 0.05) Fu he , he e was no signi ican d ough e ec on biomass p oduc ion o he ni ogen- ixing plan L. co nicula us (Figu e 1). Ca bon ixa ion D ough inc eased he maximum up ake capaci y (GPPmax) in g assland by 36 % (Figu e 1). The soil espi a ion a e (Reco calcula ed by he model was lowe unde d ough han unde ambien condi ions. Soil espi a ion was sligh ly bu no signi ican ly dec eased a he end o he d ough . 2005 2006 2007 2008 2009 0 100 200 300 400 Yea ANPP [g m-² a-1] 0 20 40 60 80 100 G een co e [%] Ambien D ough Roo leng h [cm 4 cm-2] n.s. n.s. (a) (b) 0 1 2 3 4 5 6(c) n.s. Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 61 Conclusion Ou expe imen al da a demons a e ha wea he ex emes ini ia e ecosys em- egula ing unc ions such as wa e and nu ien cycling, gas exchange and composi ional dynamics while main aining p ima y p oduc ion. They indica e an impo an con ibu ion o ecological complexi y o he main enance o p oduc i i y in he ace o inc eased empo al clima e a iabili y and ex ao dina y wea he e en s. Howe e , single species eac ions can no be ansla ed di ec ly o he communi y and ecosys em le el. A po en ial eason o di e en d ough impac s on a ious ecosys em p ope ies may lie in he empo al hie a chy o as e sus slow esponse pa e ns. In ou empe a e g assland, we obse ed he ollowing esponse dynamics wi hin hal a decade o ecu en d ough e en s: e y as al e a ion o soil mois u e s a us, subsequen as change in nu ien cycling and gas exchange, slow species-speci ic esponse in p ima y p oduc ion, ine ia in communi y p oduc i i y. Such da a on mul iple esponse pa ame e s wi hin clima e change expe imen s os e he unde s anding o mechanisms o esilience, o syne gisms o decoupling in biogeochemical p ocesses, and o undamen al esponse dynamics o d ough a he ecosys em le el. As i was he case wi h he open ques ions on he consequences o he c isis o biodi e si y, we see his complexi y in s udying impac s o clima e ex emes as a new chance o a boos in ecological heo y. Addi ionally, comp ehensi e s udies on he complex esponses will help de eloping coping s a egies o he adap ed managemen o hese ecosys ems. Fu u e challenges consis o analysing esponses o mul iple ecosys em unc ions and a mul iple le els o o ganiza ion wi h he goal o assessing how hey in e ac o in luence eme gen ecosys em p ope ies, such as ecosys em unc ion and s abili y. The obse ed s abili y in p ima y p oduc ion in he ace o ecu en se e e d ough does no mean ha he esponses a he ecosys em le el a e null. On he con a y, he obse ed changes in ecosys em egula ing unc ions in e ms o gas exchange, nu ien cycling, wa e egula ion and communi y s abili y sugges a p ominen ole o ex eme wea he e en s in ecosys em esponse o clima e change. Howe e , modelling he beha iou o ecosys ems du ing and a e ex eme clima ic e en s a la ge spa ial scales and o e longe pe iods o ime equi es mo e in-dep h knowledge on possible esponse mechanisms a he le el o plan communi ies. Po en ial epigene ic, physiological o ophic esponses need o be igo ously u he explo ed expe imen ally. Labo a o y s udies on molecula mechanisms ha e o be ela ed o s udies Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 62 wi h he same species in he ield. Field s udies mus in eg a e a ious le els o unc ional di e si y (Beie kuhnlein e al. in p ess). Pheno ypical di e si y o popula ions has o be conside ed. Li e cycles o plan species and coho s can be o c ucial impo ance. G adien s in soil ypes ha e o be in eg a ed. Then, we can each a be e unde s anding o he mechanisms ha a e ini ia ed in plan communi ies by ex eme e en s. Fu u e wo k is needed o elucida e he ole o biodi e si y and o bio ic in e ac ions in modula ing ecosys em esponse o ex eme wea he e en s. Fu he , we need mo e da a on impac s o clima e ex emes on mul iple ecosys em p ope ies om a ious ecosys ems and biomes, in o de o os e he sea ch o gene ali y ac oss di e en ca ego ies o esponse. He e, a majo challenge is o assess he speed o esponse ac oss a ious pa ame e s, including long- e m eedbacks, i.e. caused by a ni ogen-dependen eedback on p oduc i i y (Haddad e al. 2002). Gene ally, scien is s a e challenged by ela ing he ecosys em p ope ies measu ed (he e: ne ecosys em exchange, biomass abo e and below g ound, ca bon ixa ion by pho osyn hesis, nu ien a ios) o ecosys em unc ions and se ices, such as p oduc i i y, ca bon ixa ion, nu ien cycling, decomposi ion and wa e egula ion. Measu ing ecosys em se ices is a as -de eloping esea ch a ea wi h many deba es on how o assess he se ices adequa ely. Acknowledgemen s: The con ibu ion o a ious wo king g oups o he measu emen s in he EVENT expe imen gi es us a unique oppo uni y o b ing bi s and pieces oge he . We hank J. Bö che -T eschkow, M. Ewald, N. He old, Z. Hussein Y. Li, M. Mede e , C. Mülle , L. Muelle , S. Neugebaue , D. P ab, K. Simmnache , H. Skiba, S. Wal he , M. Wenigmann, D. Wul and many s uden helpe s o assis ance wi h da a mining in he ield and ui ul discussions. Re e ences: Allan, R.P. & Soden, B.J. (2008) A mosphe ic wa ming and he ampli ica ion o p ecipi a ion ex emes. Science, 321, 1481–1484. And esen, L.C., Michelsen, A., Jonasson, S., Schmid , I.K., Mikkelsen, T.N., Ambus, P. & Beie , C. (2010) Plan nu ien mobiliza ion in empe a e hea hland esponds o ele a ed CO2, empe a u e and d ough . Plan Soil, 328, 381–396. Bal ane a, P., P is e e , A.B., Buchmann, N., He, J.-S., Nakashizuka, T., Ra aelli, D. & Schmid, B. (2006) Quan i ying he e idence o biodi e si y e ec s on, ecosys em unc ioning and se ices. Ecology Le e s, 9, 1146-1156. Ba es, D.M., Sa ka , D. (2007) lme4: Linea mixed-e ec s models, R package e sion 0.9975-13, URL h p://www.R-p ojec .o g. Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 63 Beie kuhnlein, C., Jen sch, A., Thiel, D., Willne , E. & K eyling, J. (in p ess) P o enances o Eu opean g ass species espond speci ically o wa ming and ex eme d ough . Jou nal o Ecology. Special Issue Ex eme E en s. Bell, J.E., She y, R. & Luo, Y. (2010) Changes in soil wa e dynamics due o a ia ion in p ecipi a ion and empe a u e: An ecohyd ological analysis in a allg ass p ai ie. Wa e Resou ces Resea ch, 46. Biaggini, M., Conso i, R., Dappo o, L., Dellacasa, M., Pagge i, E. & Co i, C. (2007) The axonomic le el o de as a possible ool o apid assessmen o a h opod di e si y in ag icul u al landscapes. Ag ic. Ecosys . En i on., 122, 183–191. B ookes, P.C., Landman, A., P uden, G. & Jenkinson, D.S. (1985) Chlo o o m umiga ion and he elease o soil ni ogen: A apid di ec ex ac ion me hod o measu e mic obial biomass ni ogen in soil. Soil Biology and Biochemis y, 17, 837-842. Buckland, S.M., Thompson, K., Hodgson, J.G. & G ime, J.P. (2001) G assland in asions: e ec s o manipula ions o clima e and managemen . J Appl Ecol, 38, 301–309. Ciais, P., Reichs ein, M., Vio y, N., G anie , A., Ogee, J., Alla d, V., Aubine , M., Buchmann, N., Be nho e , C., Ca a a, A., Che allie , F., De Noble , N., F iend, A., F iedlings ein, P., G unwald, T., Heinesch, B., Ke onen, P. Knohl, A., K inne , G., Lous au, D., Manca, G., Ma eucci, G., Miglie a, F., Ou ci al, J., Papale, D., Pilegaa d, K., Rambal, S., Seu e , G., Soussana, J., Sanz, M., Schulze, E.-D., Vesala, T. & Valen ini, R. (2005) Eu ope-wide educ ion in p ima y p oduc i i y caused by he hea and d ough in 2003. Na u e, 437, 529-533. Cos anza, R. R. d'A ge, R. de G oo , S. Fa be , M. G asso, B. Hannon, S. Naeem, K. Limbu g, J. Pa uelo, R.V. O'Neill,R. Raskin, P. Su on, and M. an den Bel . 1997. The alue o he wo ld's ecosys em se ices and na u al capi al. Na u e, 387, 253-260. De Boeck, H., Lemmens, C.M.H.M., Za alloni, C., Gielen, B., Malchai , S., Ca nol, M., Me ckx, R., an den Be ge, J., Ceulemans, R. & Nijs, I. (2008) Biomass p oduc ion in expe imen al g asslands o di e en species ichness du ing h ee yea s o clima e wa ming. Biogeosciences, 5, 585-594. De mody, O., Wel zin, J.F., Engel, E.C., Allen, P. & No by, R.J. (2007) How do ele a ed CO2, wa ming, and educed p ecipi a ion in e ac o a ec soil mois u e and LAI in an old ield ecosys em? Plan Soil, 301, 255–266. Emme , B.A., Beie , C., Es ia e, M., Tie ema, A., K is ensen, H.L., Williams, D., Penuelas, J., Schmid , I. & Sowe by, A. (2004) The esponse o soil p ocesses o clima e change: Resul s om manipula ion s udies o sh ublands ac oss an en i onmen al g adien . Ecosys ems, 7, 625–637. Fay, P.A., Ca lisle, J.D., Knapp, A.K., Blai , J.M.& Collins S.L. (2000) Al e ing ain all iming and quan i y in a mesic g assland ecosys em: Design and pe o mance o ain all manipula ion shel e s. Ecosys ems, 3, 308- 319. Fay, P.A., Ca lisle, J.D., Knapp, A.K., Blai , J.M. & Collins, S.L. (2003) P oduc i i y esponses o al e ed ain all pa e ns in a C-4-domina ed g assland. Oecologia, 137, 245–251. Fishe , B., R.K. Tu ne , P.Mo ling (2009) De ining and classi ying ecosys em se ices o decision making. Ecological Economics, 68, 643-653. Gilmano , T.G., Tieszen, L.L., Wylie, B.K., Flanagan, L.B., F ank, A.B., Ha e kamp, M.R., Meye s, T.P. & Mo gan, J.A. (2005) In eg a ion o CO2 lux and emo ely-sensed da a o p ima y p oduc ion and ecosys em espi a ion analyses in he No he n G ea Plains: po en ial o quan i a i e spa ial ex apola ion. Global Ecol. Biogeog ., 14, 271–292. Gleeson, D.B., Mulle , C., Ma, W., Banja ee, S., Sicilliano, S. & D.V. Mu phy (2010) In luence o wa e on a chaeal and bac e ial ammonia oxidise communi y dynamics and ni ogen cycle p ocesses. Soil Biology & Biochemis y, doi:10.1016/j.soilbio.2010.06.020) G ime, J.P., B own, V.K., Thompson, K., Mas e s, G.J., Hillie , S.H., Cla ke, I.P., Askew, A.P., Co ke , D. & Kiel y, J.P. (2000) The esponse o wo con as ing limes one g asslands o simula ed clima e change. Science, 289, 762–765. Gu schick, V.P. & Bassi iRad, H. (2003) Ex eme e en s as shaping physiology, ecology, and e olu ion o plan s: owa d a uni ied de ini ion and e alua ion o hei consequences. New Phy ologis , 160, 21-42. Haddad, N.M., Tilman, D. & J.M.H. Knops (2002) Long- e m oscilla ions in g assland p oduc i i y induced by d ough . Ecology Le e s, 5, 110-120. Hec o , A. & Bagchi, R. (2007) Biodi e si y and ecosys em mul i- unc ionali y. Na u e, 448, 188–190 IPCC (2007) Clima e Change 2007: The Physical Science Basis. Con ibu ion o Wo king G oup I o he Fou h Assessmen Repo o he In e go e nmen al Panel on Clima e Change. Camb idge Uni e si y P ess, Camb idge. Jen sch, A. (2006) Ex eme clima ic e en s in ecological esea ch. F on ie s in Ecology and he En i onmen , 4, 235-236. Jen sch, A. & Beie kuhnlein C. (2008) Resea ch on ie s in clima e change: E ec s o ex eme me eo ological e en s on ecosys ems. Comp es Rendus Geoscience, 340, 621-628. Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 64 Jen sch, A. & Beie kuhnlein, C. (2010) Simula ing he u u e esponses o ecosys ems, key species and Eu opean p o enances o expec ed clima ic ends and e en s. No a Ac a Leopoldina NF, 112 (384), 89-98. Jen sch, A., K eyling, J. & Beie kuhnlein, C. (2007) A new gene a ion o clima e change expe imen s: e en s, no ends. F on ie s in Ecology and he En i onmen , 5, 365-374. Jen sch, A., K eyling, J., Boe che -T eschkow, J. & Beie kuhnlein, C. (2009) Beyond g adual wa ming: ex eme wea he e en s al e lowe phenology o Eu opean g assland and hea h species. Global Change Biology, 15, 837-849. Kahmen, A., Pe ne , J. & Buchmann, N. (2005) Di e si y dependen p oduc i i y in semi-na u al g asslands ollowing clima e pe u ba ions. Func ional Ecology, 19, 594-601. Knapp, A.K., Beie , C., B iske, D.D., Classen, A.T., Luo, Y., Reichs ein, M., Smi h, M.D., Smi h, S.D., Bell, J.E., Fay, P.A., Heisle , J.L., Lea i , S.W., She y, R.A., Smi h, B. & Weng, E. (2008) Consequences o mo e ex eme p ecipi a ion egimes o e es ial ecosys ems. Bioscience, 58, 811-821. K eyling, J., Ju asinski, G., G an , K., Re ze , V., Jen sch, A. & Beie kuhnlein, C. (in p ess) Win e wa ming pulses s ongly a ec he de elopmen o plan assemblages in empe a e g assland and hea h. Global Change Biology. K eyling, J., Beie kuhnlein, C. & Jen sch, A. (2010) E ec s o soil eeze- haw cycles di e s ongly be ween a i icial ege a ion ypes. Basic and Applied Ecology, 11, 65-75. K eyling, J., Beie kuhnlein, C. P i sch, K., Rado ski, M., Schlo e , M., Wöllecke J. & Jen sch, A. (2008a) Soil bio ic p ocesses emain su p isingly s able in ace o 100-yea ex eme wea he e en s in expe imen al g assland and hea h. Plan & Soil, 308, 175-188. K eyling J, Beie kuhnlein C, P i sch K, Schlo e M, Jen sch A (2008b): Recu en soil eeze- haw cycles enhance plan p oduc i i y. New Phy ologis , 177, 938-945. K eyling, J., Ellis, L., Beie kuhnlein, C. & Jen sch, A. (2008c) Bio ic esis ance and luc ua ing esou ces a e addi i e in de e mining in asibili y o g assland and hea h communi ies exposed o ex eme wea he e en s. Oikos, 117: 1524-1554. K eyling, J., Wenigmann, M., Beie kuhnlein, C. & Jen sch, A. (2008d) E ec s o ex eme wea he e en s on plan p oduc i i y and issue die-back a e modi ied by communi y composi ion. Ecosys ems, 11, 752-763. Loik, M.E., B eshea s, D.D., Lauen o h, W.K. & Belnap, J. (2004) A mul i-scale pe spec i e o wa e pulses in d yland ecosys ems: clima ology and ecohyd ology o he wes e n USA. Oecologia, 141, 269–281. Ma kham, J.H. & Chanway, C.P. (1996) Measu ing plan neighbou e ec s. Func ional Ecology, 10, 548-549. Maxwell, K. & Johnson, G.N. (2000) Chlo ophyll luo escence - a p ac ical guide. Jou nal o Expe imen al Bo any, 51, 659–668. McGonigle, T.P., Mille , M.H., E ans, D.G., Fai child, G.L. & Swan, J.A. (1990) A new me hod which gi es an objec i e-measu e o coloniza ion o oo s by esicula a buscula myco hizal ungi. New Phy ologis , 115, 495-501. Mi zaei, H., K eyling, J., Hussain, Z., Li, Y., Tenhunen, J., Beie kuhnlein, C. & Jen sch, A. (2008) One ex eme d ough e en enhances subsequen ca bon up ake in expe imen al g assland communi ies. Jou nal o Plan Nu i ion and Soil Science, 171, 681-689. Moise, E.R.D. & Hen y, H.A.L. (2010) Like mo hs o a s ee lamp: exagge a ed animal densi ies in plo -le el global change ield expe imen s. Oikos, 119, 791–795. Mo ec o , M.D., Mas e s, G.J., B own, V.K., Cla ke, I. P., Taylo , M.E. & Whi ehouse, A.T. (2004) Changing p ecipi a ion pa e ns al e plan communi y dynamics and succession in an ex-a able g assland. Func ional Ecology, 18, 648-655. Newman ,G.S., A hu , M.A., Mulle , R.N. (2006) Abo e- and belowg ound ne p ima y p oduc ion in a empe a e mixed deciduous o es . Ecosys ems, 9, 317–329. Niklaus, P.A., S ocke , R., Kö ne , C. & Leadley, P.W. (2000) CO2 lux es ima es end o o e es ima e ecosys em C seques a ion a ele a ed CO2. Func ional Ecology, 14, 546–559. O’Go man, P.A. & Schneide , T. (2009) The physical basis o inc eases in p ecipi a ion ex emes in simula ions o he 21s -cen u y clima e change. PNAS, 106, 14773-14777. Olesen, J.E. & Bindi, M. (2002) Consequences o clima e change o Eu opean ag icul u al p oduc i i y, land use and policy. Eu opean Jou nal o Ag onomy, 16, 239-262. Penuelas, J., Go don, C., Llo ens, L., Nielsen, T., Tie ema, A., Beie , C., B una, P., Emme , B., Es ia e, M. & Go issen, A. (2004) Nonin usi e ield expe imen s show di e en plan esponses o wa ming and d ough among si es, seasons, and species in a no h-sou h Eu opean g adien . Ecosys ems, 7, 598-612. P is e e , A.B. & Schmid, B. (2002) Di e s iy-dependen p oduc ion can dec ease he s abili y o ecosys em unc ioning. Na u e, 416, 84-86. Po s, D.L., Huxman, T.E., Cable, J.M., English, N.B., Ignace, D.D., Eil s, J.A., Mason, M.J., Wel zin, J.F. & Williams, D.G. (2006) An eceden mois u e and seasonal p ecipi a ion in luence he esponse o canopy- scale ca bon and wa e exchange o ain all pulses in a semi-a id g assland. New Phy ol, 170, 849–860. Sch ö e , D., C ame , W., Leemans, R., P en ice, I.C., A aújo, M.B., A nell, N.W., Bondeau, A., Bugmann, H., Ca e , T.R., G acia, C.A., Vega-Leine , A.C.D.L., E ha d, M., Ewe , F., Glendining, M., House, J.I., Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 65 Kankaanpää, S., Klein, R.J.T., La o el, S., Lindne , M., Me zge , M.J., Meye , J., Mi chell, T.D., Regins e , I., Rounse ell, M., Saba é, S., Si ch, S., Smi h, B., Smi h, J., Smi h, P., Sykes, M.T., Thonicke, K., Thuille , W., Tuck, G., Zaehle, S. & Zie l, B. (2005): Ecosys em se ice supply and ulne abili y o global change in Eu ope. Science, 310, 1333-1337. Shaw, M.R., Za ale a, E.S., Chia iello, N.R., Cleland, E.E., Mooney, H.A. & Field, C.B. (2002) G assland esponses o global en i onmen al changes supp essed by ele a ed CO. Science, 298, 1987–1990. She y, R.A., Zhou, X.H., Gu, S.L., A none, J.A., Schimel, D.S., Ve bu g, P.S., Wallace, L.L. & Luo, Y.Q. (2007) Di e gence o ep oduc i e phenology unde clima e wa ming. P Na l Acad Sci USA, 104, 198–202. Shi, L., Gu enbe ge , M., Ko ke, I. & R.Hampp (2002) The e ec o d ough on myco hizas o beech (Fagus syl a ica L.): changes in communi y s uc u e, and he con en o ca bohyd a es and ni ogen s o age bodies o he ungi. Myco hiza,12(6),303-11. Singh, B.K., Reid, E., O d, B., Po s, J., & Mila d, P. (2006) In es iga ing mic obial communi y s uc u e in soils by physiological, biochemical and molecula inge p in ing me hods. Eu opean Jou nal o Soil Sciences. 57, 72-82 Singh, B.K., Ba dge , R.D., Smi h, P & Reay, D.(2010) Mic oo ganisms and clima e change: eedbacks and mi iga ion op ions. Na u e Re iews Mic obiology, 8, 779-790 S Clai , S.B., Sudde h, E.A., Fische , M.L., To n, M.S., S ua , S.A., Sal e, R., Egge , D.L. & Acke ly, D.D. (2009) Soil d ying and ni ogen a ailabili y modula e ca bon and wa e exchange o e a ange o annual p ecipi a ion o als and g assland ege a ion ypes. Glob Change Biol, 15: 3018–3030. S e nbe g, M., B own, V.K., Mas e s, G.J. & Cla ke, I.P. (1999) Plan communi y dynamics in a calca eous g assland unde clima e change manipula ions. Plan Ecol, 143, 29–37. S i , M. & Schulze, E.D. (1994) Does Rubisco con ol he a e o pho osyn hesis and plan g ow h? An exe cise in molecula ecophysiology. Plan , Cell and En i onmen , 17, 465–487. Su le, K.B., Thomsen, M.A. & Powe , M.E. (2007) Species In e ac ions Re e se G assland Responses o Changing Clima e. Science, 315, 640 – 642. Swemme , A.M., Knapp, A.K. & Snyman, H.A. (2007) In a-seasonal p ecipi a ion pa e ns and abo e-g ound p oduc i i y in h ee pe ennial g asslands. Jou nal o Ecology, 95, 780-788. Tennan , D. (1975) A es o a modi ied line in e sec me hod o es ima ing oo leng h. Jou nal o Ecology, 63, 995-1001. Tilman, D. (1996) Biodi e si y: popula ion e sus ecosys em s abili y. Ecology, 77(2), 350-363. T illo, N., Fe nandez, R.J. (2005) Whea plan hyd aulic p ope ies unde p olonged expe imen al d ough : s onge decline in oo -sys em conduc ance han in lea a ea. Plan Soil, 277, 277–284. an Ruij en, J. & Be endse, F. (2010) Di e si y enhances communi y eco e y, bu no esis ance, a e d ough . Jou nal o Ecology, 98, 81-86. Wald op, M. & Fi es one, M. (2006) Seasonal dynamics o mic obial communi y composi ion and unc ion in oak canopy and open g assland soils. Mic ob. Ecol., 52, 470–479. Wal e , J., Beie kuhnlein, C., Hein, R., Nagy, J., Rasche , U., Willne , E. & Jen sch, A. (2011): Do plan s emembe d ough ? Hin s owa ds a d ough memo y in g asses. En i onmen al and Expe imen al Bo any 71,34-40. Wang, Y., Yu, S. & Wang J. (2007) Biomass-dependen suscep ibili y o d ough in expe imen al g assland communi ies. Ecology Le e s, 10, 401–410. Wel zin, J.F., B idgham, S.D., Pas o , J., Chen, J.Q. & Ha h, C. (2003) Po en ial e ec s o wa ming and d ying on pea land plan communi y composi ion. Glob Change Biol, 9, 141–151. Wiiliams, M.A. (2007) Response o mic obial communi ies o wa e s ess in i iga ed and d ough -p one allg ass p ai ie soils. Soil Biology and Biochemis y, 39, 2750-2757 Williams, M.A. & Xia, X. (2009) Cha ac e iza ion o he wa e soluble soil o ganic pool ollowing he ewe ing o d y soil in a d ough -p one allg ass p ai ie. Soil Biology and Biochemis y, 41, 21-28 Wohl ah , G., An ang, C., Bahn, M., Haslwan e , A., Newesely, C., Schmi , M., D osle , M., P adenhaue , J. & Ce nusca, A. (2005) Quan i ying nigh ime ecosys em espi a ion o a meadow using eddy co a iance, chambe s and modelling. Ag . Fo es Me eo ol., 128, 141–162. Wo m, B., Ba bie , E.B., Beaumon , N., Du y, J.E., Folke, C., Halpe n, B.S., X Jackson, J.B.C. , Lo ze, H.K., Micheli, F., Palumbi, S.R., Sala, E., Selkoe, K.A., S achowicz, J.J. & Wa son, R. (2006) Impac s o biodi e si y loss on ocean ecosys em se ices. Science, 314, 787-790. Xiang, S.-R., Doyle, A., Holden, P.A. & Schimel, J.P. (2008) D ying and ewe ing e ec s on C and N mine aliza ion and mic obial ac i i y in su ace and subsu ace Cali o nia g assland soils. Soil Biol Biochem, 40, 2281–2289. Za alloni, C., Gielen, B., De Boek, H.J., Lemens, M.H.M.C., Ceulemans, R. & Nijs, I. (2009) G ea e impac o ex eme d ough on pho osyn hesis o g asslands exposed o wa me clima e in spi e o acclima ion. Physiologia Plan a um, 136, 57-72. Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 66 Online Suppo ing In o ma ion Table S1: Sea ch i ems o sea ching he ISI Web o Science® Da abase o publica ions on wea he e en s and clima e ex emes. As e isks a e place holde s wi hin he sea ch s ing Main ca ego y Sea ch i ems in ISI Web o Knowledge F os * os e en * OR se e e nigh os * OR *sp ing os * OR eeze- haw* OR haw- eeze* OR la e os * OR se e e os * OR *g ound os * OR ex eme os * OR ex eme cold* Hea & d ough hea wa e* OR hea wa e* OR se e e hea * OR * empe a u e e en * OR d y spell* OR ex eme hea * OR *win e wa ming* OR wa m* win e OR summe d ough * OR sp ing d ough * OR au umn d ough * OR se e e d ough * S o m ex eme s o m* OR *win e s o m* OR hu icane* OR yphoon* OR cyclone* OR o nado* OR s o m su ge* OR *winds o m* OR *wind s o m* OR * opical s o m* OR ice s o m* Hea y ain ex eme lood* OR summe lood* OR ex eme ain* OR o en ial ain* OR ex eme p ecipi a ion OR * ain all e en * OR hea y ain* OR hail* OR we spell* Ex eme ex eme e en * OR ex eme wea he e en * OR clima * ex eme* OR ex eme me eo ological e en * OR ex eme wea he * OR ex eme clima * e en * Table S2: Links o sea ching he ISI Web o Science® Da abase o publica ions on wea he e en s and clima e ex emes Links Sea ch i ems in ISI Web o Knowledge Main i ems in “Topic” F os , Hea & d ough , S o m, Hea y ain, Ex eme AND in “Topic” we land* OR loodplain* OR pea * OR bog* OR en* OR swamp* OR mi e OR g assland* OR meadow* OR pas u e* OR hea h* OR sh ubland* OR o es * OR woodland* OR und a OR aiga OR sa anna* OR ma sh* OR s eppe OR dese * OR aqua ic* OR limn* *bi d* OR a ian OR insec * OR bu e ly* OR bee le* OR a h opod* OR mo h* OR amphibian* OR ep ile* OR mollusc* OR mollusk* OR e eb a e* OR *in e eb a e* OR mammal* AND in “Topic” clima * change* OR global change* OR clima * wa ming NOT in “Topic” palaeo* OR paleo* OR pleis ocene OR holocene Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 67 Table S3: Sampling yea s o all esponse pa ame e s p esen ed in Figu e 1. Gi en a e da a om yea s wi h maximum d ough e ec Ecosys em p ope y Yea o sampling wi h maximum d ough e ec Ecosys em se ice Abo e-g ound p oduc ion (ANPP) 2005 Ni ogen ixing plan s 2009 Plan co e 2009 P ima y p oduc ion Below-g ound biomass 2007 shoo / oo - a io 2006 Gas exchange Maximum ca bon up ake capaci y 2005 Pho osyn he ic pe o mance 2008 Lea gas exchange 2007 Soil espi a ion 2010 Nu ien cycling Decomposi ion a e 2007 Myco hiza ion a e 2008 Soil mic obial biomass 2008 Soil enzyme ac i i y 2006 Plan a ailable NO3- 2008 Plan a ailable soil NH4 2008 soil mic obial N 2009 Lea C/N- a io 2009 Lea p o ein con en 2009 Lea ca bohyd a e con en 2009 Lea ni ogen iso ope signal 2007 1° consume abundance 2008 Soil mois u e 2009 Wa e egula ion Lea wa e po en ial 2008 Lea ca bon iso ope signal 2007 Communi y esponses In asibili y 2006 Plan composi ional change 2005 Senescence 2006 Va iabili y in leng h o lowe ing 2006 Va iabili y in lowe phenology 2008 Resis ance o he bi o y 2009 Compe i i e e ec 2007 Facili a i e e ec 2007 Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 68 (a) Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 69 (b) (c) Manusc ip 1: Clima e ex emes ini ia e ecosys em egula ing unc ions while main aining p oduc i i y 70 (d) Figu e S1: Resea ch on ecological e ec s o clima e ex emes and wea he e en s based on publica ions ound in he ISI Web o Science ( o sea ch de ails see Table 2) (a) Tempo al de elopmen o he numbe o publica ions on clima e ex emes (n=380) in he las wo decade (shown is only he las decade); o al yield 1134 pee - e iewed pape s (b) S udied ex eme wea he e en s (n=464 inlcuding double o iple assignmen s) o he ele an pee - e iewed pape s (n=380) yielded by he li e a u e s udy. 24 publica ions did no speci y he e en . (c) Resea ch ac i i y in he h ee main biomes by p opo ion o publica ions based on 380 pee - e iewed pape s pa icula ly s udying e ec s o clima e ex emes on ecosys em unc ions. G assland includes dese s, pea and we lands. Sh ubland includes und a. Any one pape may ha e been assigned o mul iple subjec a eas. (d) S udied e ec s o ex eme wea he e en s on ecosys em p ope ies a anged by ecosys em se ices and unc ions based on 380 pee - e iewed pape s pa icula ly s udying e ec s o clima e ex emes on ecosys em unc ions Manusc ip 2: How do ex eme d ough and plan communi y composi ion a ec hos plan me aboli es and he bi o e pe o mance 77 and he le els we e shi ed e e y second day, pu ing he lowes le el o he highes place and all o he le els one le el lowe . The plan lea es om one plo we e ed o he la ae in ou pe i dishes (8 la ae in o al). Ca e pilla s we e able o ea ad libi um, as special ca e was aken ha he g ass inside one pe i dish was ne e o ally consumed. Lea es we e eplaced a leas e e y second day. We eco ded he mo ali y o he la ae, he de elopmen ime un il pupa ion, he weigh o he pupae one day a e pupa ion and pupal mo ali y. Th ough he isola ion in clima e chambe s we we e able o a ibu e he esponses in he bi o e pe o mance o di e ences in plan compounds, as opposed o when eeding is conduc ed unde ield condi ions and i is impossible o disen angle he e ec s o plan nu ien s, compensa o y eeding and di ec wea he e ec s on he bi o es (Go e de e al. 2002). S a is ical analysis Chemical lea ai s we e analyzed using a wo-way ANOVA wi h wea he ea men and communi y composi ion as ixed ac o s. Addi ionally, we included he numbe o columns and ows as andom ac o s. This au oma ically implemen s he nes ing o composi ion wi hin ea men -blocks in he mixed e ec model (Fa away 2006; Do mann and Kühn 2008). I se e al samples pe plo we e aken, as was he case wi h phenolics, ca bohyd a es and condensed annins, hen he plo numbe was addi ionally included as a andom ac o in he mixed model, o a oid any pseudo- eplica ion. De elopmen al ime and pupal weigh we e analyzed using linea mixed e ec models wi h he pe i dish nes ed wi hin he plo nes ed wi hin he ea men block as a andom ac o , in addi ion o he ow and he column o he ea men blocks as andom ac o s. La al and pupal mo ali y we e analyzed using gene alized mixed e ec models wi h binomial dis ibu ion and o he wise he same model o mula as o o he de elopmen al ai s. The signi icance le els in he mixed e ec models we e e alua ed by Ma ko Chain Mon e Ca lo sampling o 1000 pe mu a ions (Baayen 2009; package language R).The signi icance o he ixed ac o s o he gene alized mixed e ec models was de e mined by compa ing he null model, wi hou any ac o s, o he simples ac o ial model, in which non- signi ican e ms had been emo ed by backwa ds s epwise selec ion. P io o all analyses, da a we e ans o med acco dingly, i he assump ions o ANOVA, homogenei y o a iances and no mali y, we e no me (C/N a io and ni ogen con en : log- ans o med; RWC: a csin- squa e oo - ans o med). To de e mine he ela ionship be ween chemical lea ai s and de elopmen ai s, we applied hie a chical pa i ioning, as lea chemical ai s a e o en collinea (Schädle e al. Manusc ip 2: How do ex eme d ough and plan communi y composi ion a ec hos plan me aboli es and he bi o e pe o mance 78 2003). In hie a chical pa i ioning, he independen in luence and he join in luence ( he in luence om being co ela ed o ano he explana o y a iable) o explana o y a iables is calcula ed by compa ing he model- i s o models wi h and wi hou he pa icula a iable (Mac Nally 2002; Do mann and Kühn 2008). To de e mine co ela ions be ween he di e en de elopmen ai s, a co ela ion analysis was used o co ela e pupal weigh wi h de elopmen al ime, and logis ic eg ession was used o in es iga e he ela ionship be ween pupal mo ali y and pupal weigh . All s a is ical analyses we e pe o med using R 2.11.0 (R De elopmen Co e Team 2010). Fo mixed e ec models we used he so wa e package lme4 (Ba es & Mechle 2010), and o mul iple pos -hoc compa isons he package mul comp was used (Ho ho n e al. 2008). Resul s Soil mois u e The ege a ion pe iod o he yea 2009 (Ap il 1s – Oc obe 31s ) wi h a o al sum o 459 mm o p ecipi a ion was sligh ly we e han he long- e m a e age p ecipi a ion sum o 437 mm o he ime pe iod 1971-2000 (Da a: Ge man Wea he Se ice). Soil mois u e ell mo e quickly du ing he i s hal o he d ough pe iod compa ed o he second hal , bu ose quickly again a e he d ough pe iod was o e (Fig.1). Day o he yea 140 150 160 170 180 0,00 0,05 0,10 0,15 0,20 0,25 0,30 D ough Con ol Soil mois u e [ ol % ] 0 5 10 15 20 25 A e age Tempe a u e [°C ] Day o he yea 140 150 160 170 180 0,00 0,05 0,10 0,15 0,20 0,25 0,30 D ough Con ol D ough Con ol Soil mois u e [ ol % ] 0 5 10 15 20 25 A e age Tempe a u e [°C ] Fig. 1 Cou se o soil mois u e in d ough -exposed plo s (black ci cles) and con ol plo s (da k-g ey squa es), and a e age daily empe a u es, assessed a a heigh o 1, 20 m (ligh g ey ba s). Da a a e shown om he i s day o he d ough manipula ion (day o he yea 140=May 20 h, 2009) un il wo days a e he ex eme d ough ended, indica ed by he black e ical line (day o he yea 182=July 1s , 2009). The a e age o hou ly eadings om i e senso s pe ea men we e aken he e (n=5) Manusc ip 2: How do ex eme d ough and plan communi y composi ion a ec hos plan me aboli es and he bi o e pe o mance 79 The e ec o d ough and communi y composi ion on chemical lea ai s D ough signi ican ly dec eased RWC by 8 % (Fig. 2a), p o ein con en by 23 % (Fig. 2b), ni ogen concen a ions by 26 % (Table 3) and phenols by 7 %, when compa ed o he con ol ea men (Fig. 2c) (see Table 2 o s a is ical de ails). Fu he mo e, d ough signi ican ly inc eased he C/N a io by 24 % and he soluble ca bohyd a es by 32 % (Fig. 2d, e). Condensed annins howe e we e no al e ed by d ough manipula ion (Table 2). CD RWC [%] 020 40 60 80 100 (a) CD P o ein con en [µg/mg FW] 012345 (b) CD Phenols [nmol/mg] 0 50 100 150 200 (c) CD C/N a io [%] 0 10203040 (d) CD Ca bohyd a es [nmol/mg] 0 100 200 300 (e) Wea he ea men CD RWC [%] 020 40 60 80 100 (a) CD P o ein con en [µg/mg FW] 012345 (b) CD Phenols [nmol/mg] 0 50 100 150 200 (c) CD C/N a io [%] 0 10203040 (d) CD Ca bohyd a es [nmol/mg] 0 100 200 300 (e) Wea he ea men CD RWC [%] 020 40 60 80 100 (a) CD P o ein con en [µg/mg FW] 012345 (b) CD Phenols [nmol/mg] 0 50 100 150 200 (c) CD C/N a io [%] 0 10203040 (d) CD Ca bohyd a es [nmol/mg] 0 100 200 300 (e) CD RWC [%] 020 40 60 80 100 CD RWC [%] 020 40 60 80 100 (a) CD P o ein con en [µg/mg FW] 012345 CD P o ein con en [µg/mg FW] 012345 (b) CD Phenols [nmol/mg] 0 50 100 150 200 (c) CD C/N a io [%] 0 10203040 CD C/N a io [%] 0 10203040 (d) CD Ca bohyd a es [nmol/mg] 0 100 200 300 CD Ca bohyd a es [nmol/mg] 0 100 200 300 (e) Wea he ea men Fig. 2 Di e ences in a) RWC (n=5), b) he p o ein con en (n=5), c) he o al soluble phenol con en (n=15), d) he C/N a io (n=5) and e) he o al soluble ca bohyd a es (n=15) in lea es o Holcus lana us unde d ough (ligh g ey ba s, D) compa ed o con ol (whi e ba s, C). As e isks indica e he le el o signi icance: * p<0.05; **p<0.01; ***p<0.001. Means ± 1 SE a e shown Communi y composi ion signi ican ly a ec ed he con en o condensed annins and phenols (Table 2) and ma ginally signi ican ly a ec ed he con en o soluble ca bohyd a es (p=0.07) and p o eins (p=0.01). Table 2 ANOVA esul s o he e ec s o d ough ea men , communi y composi ion and, i signi ican , hei in e ac ion on RWC, o al soluble p o ein con en , ni ogen concen a ions, C/N a io (n=5, espec i ely), o al soluble ca bohyd a es, o al soluble phenols and condensed annins (n=15, espec i ely). Values o in e ac ion a e only shown, when signi ican , as o he wise, he in e ac ion e m was excluded om he mixed model (d =deg ees o eedom, MS=mean sum o squa es). As e isks indica e le el o signi icance: `p<0.01; * p<0.05; **p<0.01; ***p<0.001 RWC p o ein ni ogen C/N ca bohyd a es phenols cond. annins d MS F d MS F d MS F d MS F d MS F d MS F d MS F d ough 1 0.18 28.4*** 1 7.51 7.4* 1 1.02 65.1*** 1 1.13 76.6*** 1 106533 68.3*** 1 5871 11.3** 1 76.4 0.6 composi ion 3 0.01 1.4 3 2.29 2.26` 3 0.01 0.7 3 0.01 0.8 3 3755 2.4` 3 2547 4.9** 3 516.6 4.2** in e ac ion - - - - - - 3 0.06 4.1* 3 0.04 2.6` - - - - - - - - - esiduals 30 0.01 32 1.01 32 0.02 32 0.01 97 1558 97 518 97 124.5 Manusc ip 2: How do ex eme d ough and plan communi y composi ion a ec hos plan me aboli es and he bi o e pe o mance 80 The condensed annins in H. lana us om 4+ communi ies we e signi ican ly lowe han he condensed annins in lea es om 2- and 4- communi ies (Fig. 3a). Phenols we e educed in H. lana us g owing in legume communi ies (4+), when compa ed o H. lana us om monocul u es (1-) (Fig. 3b). Ca bohyd a es we e ma ginally signi ican ly inc eased in legume communi ies (4+) when compa ed o wo-species communi ies wi hou legume (2-) (Fig. 3c), while p o ein con en was ma ginally signi ican ly lowe in ou -species communi ies wi hou legume (4-) when compa ed o legume communi ies (4+)(Fig. 3d). Fig. 3 E ec s o plan communi y composi ion on a) he con en o condensed annins (n=15), b) he o al soluble phenol con en (n=15), c) he o al soluble ca bohyd a es (n=15) and d) he o al soluble p o eins (n=5) in lea es o Holcus lana us (1-: monocul u e, 2-: wo g asses, 4-: 2 g asses, 2 he bs; 4+: wo g asses, 1 he b, 1 legume). Di e en le e s indica e signi ican di e ences be ween he communi ies (p<0.05). E ec s o ca bohyd a e and p o ein con en we e only ma ginally signi ican (p<0.1) As a as lea ni ogen concen a ion was conce ned, communi y composi ion signi ican ly in e ac ed wi h he d ough ea men , as ni ogen concen a ion was educed in all communi ies excep o he 4+ communi ies in esponse o d ough (Table 3). Table 3 In e ac i e e ec s o communi y composi ion and he d ough ea men on he ni ogen concen a ion in lea es o H. lana us (n=5). Means ± 1 s anda d de ia ion a e gi en. As e isks behind he communi y label indica e a signi ican d ough e ec in he espec i e communi y (p<0.05) composi ion 1- * 2- * 4- * 4+ con ol 1.99 ± 0.31 1.90 ± 0.31 1.88 ± 0.31 1.69 ± 0.30 d ough 1.26 ± 0.32 1.45 ± 0.32 1.26 ± 0.32 1.51 ± 0.31 Cond. Tannins [nmol/mg] 010 20 30 40 1- 4- 2- 4+ Phenols [nmol/mg] 0 50 100 150 200 1- 4- 2- 4+ Ca bohyd a es [nmol/mg] 0100 200 300 P o ein con en [µg/mg FW] 012345 (a) (b) (d)(c) Communi y composi ion ab aa b aab b ab Cond. Tannins [nmol/mg] 010 20 30 40 1- 4- 2- 4+ Phenols [nmol/mg] 0 50 100 150 200 Phenols [nmol/mg] 0 50 100 150 200 1- 4- 2- 4+ Ca bohyd a es [nmol/mg] 0100 200 300 1- 4- 2- 4+ 1- 4- 2- 4+ Ca bohyd a es [nmol/mg] 0100 200 300 P o ein con en [µg/mg FW] 012345 (a) (b) (d)(c) Communi y composi ion ab aa b aab b ab Manusc ip 2: How do ex eme d ough and plan communi y composi ion a ec hos plan me aboli es and he bi o e pe o mance 81 De elopmen ai s o S .li o alis and hei ela ion o lea chemical ai s The communi y composi ion o he a ge g ass was ound o ha e a highly signi ican e ec on la al mo ali y, wi h la ae eeding om 4- plo s showing a signi ican ly highe mo ali y han in all o he communi ies (Fig. 4). (P=0.007; Chisq= 12.2). The d ough ea men did no a ec la al mo ali y. Fig. 4 E ec o he plan communi y composi ion, in which H. lana us g ows, on he mo ali y o he 320 S. li o alis la ae. Di e en le e s indica e signi ican di e ences be ween he communi ies (p<0.05) (n=20/ ea men combina ion) The de elopmen ime un il pupa ion was highly signi ican ly inc eased in hose la ae ha we e ea ed on lea es subjec ed o d ough (p=0.0003; F= 14.7; d =1). Fu he mo e, he d ough ea men was ound o signi ican ly in e ac wi h communi y composi ion ega ding de elopmen ime un il pupa ion (p=0.015; F= 3.8; d =3), as he la ae eeding on H. lana us om he 2- communi ies eac ed wi h he g ea es inc ease du ing he la al s age unde d ough (Table 4). Table 4 Di e ences in he du a ion o la al s age o 320 S. li o alis la ae eeding on H. lana us lea es unde he d ough (ligh g ey ba s, D) and he con ol (whi e ba s, C) ea men s om ou di e en communi y composi ions. D ough signi ican ly inc eased de elopmen al ime wi h s onges e ec s in 2- communi ies (n=20/ ea men combina ion). Means ± 1 s anda d de ia ion a e gi en composi ion 1- 2- 4- 4+ con ol 25 ± 2.5 22 ± 2.4 25 ± 2.3 24 ± 2.3 d ough 25 ± 2.3 27 ± 2.4 27 ± 2.5 26 ± 2.4 The la ae eeding on d ough plan s we e also ound o ha e signi ican ly hea ie pupae (p=0.033; F= 4.7; d =1) (Fig. 5a) and e ealed a signi ican ly lowe pupal mo ali y (p=0.007; Chisq= 7.2) (Fig 5b), i espec i e o communi y composi ion. Mo ali y [%] 02040 60 80 100 Communi y composi ion 1- 4- 2- 4+ aa b a Mo ali y [%] 02040 60 80 100 Mo ali y [%] 02040 60 80 100 Communi y composi ion 1- 4- 2- 4+ 1- 4- 2- 4+ aa b a Manusc ip 2: How do ex eme d ough and plan communi y composi ion a ec hos plan me aboli es and he bi o e pe o mance 82 Fig. 5 Di e ences in (a) he pupal weigh and (b) he pupal mo ali y o S. li o alis la ae ed wi h H. lana us lea es ou o he d ough (ligh g ey ba s, D) and he con ol (whi e ba s, C) ea men s. As e isks indica e he le el o signi icance: *P 0.05; **P 0.01 (n = 20/ ea men combina ion) Hie a chical pa i ioning showed ha p o ein con en had he g ea es posi i e in luence on he su i al o la ae (Table 5). Table 5 Hie a chical pa i ions o he e ec s o lea chemical ai s on de elopmen du a ion, pupal weigh , la al mo ali y and pupal mo ali y o 320 la ae. The o al explained a iance (R2), he indi idual e ec on he explained a iance, and he join e ec on explained a iance a e gi en. The la e quan i ies he e ec ha can be explained by he co ela ion o a speci ic independen a iable wi h o he independen a iables. + o – behind he mos impo an pa i ions o one pa ame e indica e whe he he pa ame e s we e nega i ely o posi i ely co ela ed o he lea chemical. RWC=Rela i e wa e con en ; C/N=C/N a io; ni ogen=ni ogen concen a ion; ca bon=ca bon concen a ion; p o ein= o al soluble p o eins ; phenols= o al soluble phenols ; annins= condensed annins ; ca bos= o al soluble ca bohyd a es RWC C/N ni ogen ca bon p o ein phenols annins ca bos de elopmen o al 0.143 0.259+ 0.210- 0.014 0.001 0.036 0.029 0.078 ime independen 0.113 0.187 0.118 0.019 0.085 0.013 0.014 0.03 join 0.03 0.071 0.091 -0.005 -0.084 0.024 0.015 0.048 pupal weigh o al 0.229- 0.024 0.04 0.031 0.079 0.094 0.077 0.04 independen 0.178 0.025 0.025 0.035 0.068 0.047 0.048 0.02 join 0.051 -0.001 0.015 -0.003 0.012 0.047 0.029 0.021 la al mo ali y o al 0.01 0.003 0.001 0.001 0.047- 0.006 0.005 0.009 independen 0.009 0.009 0.005 0.004 0.065 0.007 0.01 0.007 join 0.001 -0.006 -0.004 -0.002 -0.018 -0.001 -0.005 0.002 pupal mo ali y o al 0.031 0.097- 0.086 0.017 0 0.049 0.096- 0.042 independen 0.014 0.034 0.028 0.013 0.01 0.023 0.072 0.015 join 0.017 0.063 0.058 0.004 -0.009 0.026 0.023 0.027 The de elopmen ime was mainly posi i ely in luenced by he C/N- a io, and hus nega i ely by he ni ogen con en , indica ing ha he de elopmen ook longe , he less ni ogen was in he lea es (Table 5). Pupal weigh was ound o be nega i ely a ec ed by RWC. Pupal su i al, and hus adul eclosion was posi i ely a ec ed by con en o condensed annins and by C/N a io. Pupal weigh and de elopmen ime we e no co ela ed (Pea sons Pupal weigh [g] 00.02 0.06 0.10 0.14 CD Wea he ea men Pupal mo ali y [%} 020 40 60 80 CD (a) (b) Pupal weigh [g] 00.02 0.06 0.10 0.14 CDCD Wea he ea men Pupal mo ali y [%} 020 40 60 80 CDCD (a) (b) Manusc ip 2: How do ex eme d ough and plan communi y composi ion a ec hos plan me aboli es and he bi o e pe o mance 83 co ela ion coe icien : -0.019; p=0.88). Pupal su i al was posi i ely ela ed wi h pupal weigh (p=0.023; logis ic eg ession). Discussion The ex eme 42-day d ough did no only a ec he wa e con en o he a ge g ass species, bu also esul ed in changes o almos all o he lea chemical ai s ha we assessed. These changes o he lea chemicals also clea ly a ec ed he de elopmen ai s o he he bi o e ca e pilla . Howe e , in con as o ou expec a ions, d ough did no inc ease de ensi e compounds in he g ass. Fu he mo e, he d ough did no esul in a wo se o e all he bi o e pe o mance, as adul eclosion and pupal weigh we e e en highe o la ae ed om d ough ea ed plan s. The communi y composi ion o he a ge g ass also a ec ed some aspec s o he lea chemical composi ion, bu changes canno clea ly be linked o inc eased compe i ion o highe s abili y in mo e di e se communi ies. Di e ences mainly occu ed in legume communi ies: H. lana us g owing in legume communi ies showed no e ec s o d ough on he lea ni ogen concen a ion and had he lowes con en o condensed annins and phenols. The communi y composi ion o he a ge g ass a ec ed he mo ali y o he he bi o e, wi h highes mo ali y in la ae eeding on he g ass g owing in ou -species communi ies wi hou legume, in which also a end owa ds lowe p o ein con en was appa en (4-). Soil mois u e and plan s ess The ex eme d ough condi ions we e accompanied by a ma ked dec ease in soil mois u e o e he d ough pe iod, and his in u n clea ly caused plan s ess. In he yea 2009 we did no di ec ly quan i y plan s ess le els, e.g. by de e mining chlo ophyll con en o maximum quan um yield. Howe e , a educ ion in lea ela i e wa e con en in ou a ge g ass in all communi ies, along wi h ma ked changes in lea chemicals, indica e acclima ion p ocesses and s ess eac ions, showing ha he plan s expe ienced s ess (Sinclai and Ludlow 1985; Cha es e al. 2002). E ec o d ough and communi y composi ion on chemical lea ai s A dec ease in p o eins and ni ogen, along wi h an inc ease in he C/N a io unde se e e d ough condi ions has also been obse ed in o he s udies in es iga ing d ough e ec s on lea chemicals (Shu e e al. 1998; Liu e al. 2008). Howe e , his is no in Manusc ip 2: How do ex eme d ough and plan communi y composi ion a ec hos plan me aboli es and he bi o e pe o mance 84 acco dance wi h he so-called “plan s ess hypo hesis”, which assumes a ailable ni ogen o inc ease unde plan s ess (Whi e 1984). Ne e heless, as ni ogen up ake is linked o wa e up ake, a dec ease in ni ogen up ake, and he e o e also p o ein con en unde d ough , is no su p ising. An inc ease in ca bohyd a es unde d ough could be a ibu ed o osmo ic adjus men in he cou se o d ough acclima ion. Soluble ca bohyd a es om s a ch deg ada ion ac as compa ible solu es o p e en u go loss in plan cells (Cha es e al. 2002; Regie e al. 2009). The dec ease in o al phenolics unde d ough does no suppo he idea ha plan s unde s ess use he su plus om ca bohyd a es (due o es ic ed g ow h while pho osyn hesis is s ill assimila ing ca bon) o accumula e mo e C-based de ence compounds (He ms and Ma son 1992). Howe e , a educ ion o phenols in esponse o d ough has also been epo ed by Shu e e al. (1998). In ou s udy, he need o osmo ic adjus men unde ex eme d ough migh ha e been a eason no o accumula e phenols, bu a he in es he su plus o ca bon-based compounds in soluble ca bohyd a es. As communi y composi ion can al e esou ce pa i ioning be ween plan s, i migh also change he o age quali y o lea issue. Bo h phenols and condensed annins we e ound o be lowes in he communi ies ha included one legume species. This kind o educ ion in de ence compounds can be explained by he po en ially highe ni ogen a ailabili y in legume communi ies caused by N2- ixing in oo -nodules, enabling highe g ow h a es a he expense o lowe de ence by C-based compounds (He ms and Ma son 1992). The end owa ds highe p o ein con en in lea es om legume communi ies compa ed o ou -species communi ies wi hou legumes suppo he idea o an inc ease in ni ogen a ailabili y in legume communi ies and inc eased compe i ion o ni ogen in communi ies con aining ou species, bu no legume. Mo eo e , labelling s udies indica e a di ec up ake o legume-de i ed ni ogen by g asses (Gubsch e al 2011). O e all, he ni ogen concen a ion o g asses g owing in legume communi ies (4+) was no a ec ed by d ough , bu showed changes in esponse o d ough in he o he communi ies. The highe s abili y in his communi ies and he possible e iliza ion e ec o he legume suppo o he s udies which ha e shown an enhancemen o he ni ogen a ailabili y o plan s g owing in legume communi ies (Spehn e al. 2002; Tempe on e al. 2007; Dybzinski e al. 2008). De elopmen al ai s o S. li o alis and hei ela ion o lea chemical ai s The mo ali y o la ae was high, as he expe imen was s a ed immedia ely a e ha ching o he la ae, when hey a e qui e ulne able. Fu he mo e, H. lana us seemed o be a sub-op imal ood sou ce o S. li o alis, as la ae he same age om he same egg s ain Manusc ip 2: How do ex eme d ough and plan communi y composi ion a ec hos plan me aboli es and he bi o e pe o mance 85 g ew be e when eeding on Plan ago lanzeola a and T i olium p a ense, unde o he wise simila condi ions. Ne e heless, as we expec ed mo ali y o be high, due o he esul s om a p io ial expe imen , we included enough eplica es in o de o ob ain a subs an ial da a se o s a is ical analysis. The epo ed changes in lea me aboli es, ela ed o he d ough ea men and di e ences in he communi y composi ions, clea ly had an e ec on he bi o e pe o mance. Mo ali y o la ae was highe in ou -species communi ies wi hou a legume compa ed o he o he communi ies. Hie a chical pa i ioning showed ha he mo ali y a e depended mos ly on he p o ein con en . These da a hin owa ds a cen al ole o p o eins o he su i al o he ea ly ins a s. O he s udies, oo, showed poo la al su i al unde low ni ogen concen a ions (Mye s and Pos , 1981; Ca es 1987; de B uyn e al. 2002). La al de elopmen up o pupa ion was signi ican ly longe o la ae ed on d ough - s essed plan s, which migh inc ease p eda ion isk and hus mo ali y unde na u al condi ions (Ben ey and Denno 1997). In acco dance wi h Fische and Fiedle (2000) and wi h Mo ehouse and Ru owski (2010), he p olonged de elopmen al ime o he la ae was linked o a educed N-a ailabili y in d ough -s essed lea es. I migh be ha la ae ed longe on N- limi ed g ass o each a ce ain “g ow h a ge ” (Raubenheime and Simpson 1997). Such compensa o y eeding on low quali y issue may be a common phenomenon (Schädle e al. 2007b). D ough also had posi i e e ec s on he bi o e pe o mance, i espec i e o communi y composi ion: The inc eased pupal weigh migh be explained by a highe up ake o ene gy, as ca bohyd a e con en s in lea es inc eased unde d ough and as la ae ed longe on d ough plan s, possibly caused by lowe ed ni ogen con en s. Thus, he “imbalanced die ” when eeding on d ough s essed plan s caused i s ly compensa o y longe eeding, o each a ce ain ni ogen le el necessa y o de elopmen , and secondly led o a highe ene gy up ake and highe pupal weigh s (Raubenheime and Simpson 1997). RWC was nega i ely ela ed o pupal weigh , as p esumably wa e dilu es nu ien s o ca bohyd a es in well- wa e ed plan s. The inc eased pupal weigh was co ela ed wi h a lowe pupal mo ali y o he la ae ha had been ed om d ough -s essed plan s, which is in acco dance o o he s udies (Fische and Fiedle 2000). I should be ei e a ed ha he samples o he lea chemical analysis we e aken on he las wo days o he d ough ea men , whe eas he la al de elopmen las ed om app oxima ely en days be o e o en days a e he d ough . Thus, la ae ed on lea es ha expe ienced milde s ess le els in hei la e s ages, which migh ha e alle ia ed he e ec s o Manusc ip 2: How do ex eme d ough and plan communi y composi ion a ec hos plan me aboli es and he bi o e pe o mance 86 ex eme d ough on lea chemical ai s (Hube y and Denno 2004). Fu he s udies wi h a highe empo al esolu ion would make i possible o in es iga e he cou se o changes in lea ai s o e he whole ange o plan s ess esponses. Ne e heless, he p esen s udy e eals ha di e en de elopmen pa ame e s can be di e en ly in luenced by ex eme d ough , e en hough hese only las ed o a ound hal o he du a ion o hei la al de elopmen . Fu he mo e we show ha plan communi y composi ion (and he e o e also compe i ion and he p esence o speci ic plan unc ional ai s) al e s lea me abolism and he eby a ec s he bi o es. Conce ning he deba e on he hypo heses on plan -s ess (Whi e1984) e sus plan - igou (P ice 1991) we con i m o he s udies which show he dependence o he bi o e pe o mance on plan -s ess le el (Schei s and de B uyn 2005), in es iga ed ai s (Co nelissen e al. 2008) and insec - eeding guild (La sson 1989; Ko iche a e al. 1998; Hube y and Denno 2004), as ou esul s di e om o he d ough expe imen s using di e en eeding guilds (EnglishLoeb e al. 1997; de B uyn e al. 2002). Di e ences o o he s udies may also be a ibu ed o ou speci ic expe imen al condi ions: Plan s we e g own in expe imen al communi ies in he ield, no in isola ed po s in he g eenhouse. This ensu es mo e ealis ic plan g ow h condi ions. In addi ion, he he bi o es in ou s udy ha e no been g own on he plan s, bu we e ed in clima e chambe s, o mo e clea ly ela e he ob ained esul s o changes in plan me aboli es. Ano he eason o di e ences o o he s udies and p ominen hypo heses migh be ha he plan s we e se e ly s essed, bu also eleased om s ess and eco e ed du ing he eeding expe imen . Acco ding o Hube y and Denno (2004) his migh ha e caused di e ences in he bi o e pe o mance when compa ed o eeding expe imen s unde cons an s ess. Ou s udy p o ides addi ional e idence ha he ni ogen limi a ion hypo hesis, s a ing ha high ni ogen con en s a e bene icial o he bi o e pe o mance (Whi e1984) does no seem o be widely applicable o e all de elopmen ai s and s ages (Fische and Fiedle 2000). Simila o Fische & Fiedle (2000) ou da a p o ide hin s ha highe p o ein o ni ogen con en s a e bene icial o a as e a e o de elopmen and highe la al su i al, al hough hey migh educe adul eclosion. Conclusion Ou indings sugges ha ex eme d ough s, which a e p ojec ed o inc ease in equency wi h clima e change, can also a ec he de elopmen o he bi o es. This is p ima ily caused by a Manusc ip 3: Ecological s ess memo y and c oss s ess ole ance in plan s in he ace o clima e ex emes 93 Impo an aspec s a e no only he pe sis ence o species and subs a e, bu also he possibili ies o species o ecolonize (Nys om and Folke, 2001). An o e iew on he a ious unde s andings o ecological memo y is p o ided by Golinski e al. (2008). The complexi y o he p ocesses in ol ed in he b oad de ini ion o ecological memo y including empo al pa e ns up o e olu iona y scales and spa ial pa e ns up o landscape le els (Thompson e al., 2001), may limi i s applicabili y in ecology as well as he possibili y o assess and measu e such a memo y. I hus appea s necessa y o apply a educ ionis amewo k o os e ou unde s anding o he impo ance o ecological memo y in imes o global change (see Table 1 o examples and le els o ecological memo y a e dis u bances). In he ollowing, we ocus on ecological s ess memo y o single indi iduals as a s a ing poin . Ecological s ess memo y is de ined he e as any esponse o a single plan a e a s ess expe ience ha imp o es he esponse o he plan owa ds u u e s ess expe ience and which is assessed on a whole plan le el (Fig. 1). nega i e esponse neu al esponse imp o ed esponse no s ess no s ess s ess s ess acclima ion eco e y s ess memo y eco e y acclima ion s ess memo y Æimp o ed pe o mance damage lagged s ess e ec s collapse esponse o s ess ime nega i e esponse neu al esponse imp o ed esponse no s ess no s ess s ess s ess acclima ion eco e y s ess memo y eco e y acclima ion s ess memo y Æimp o ed pe o mance damage lagged s ess e ec s collapse esponse o s ess ime Fig.1 Plan s ess esponse unde single and epea ed s ess wi hou acclima ion leads o s ess damage (nega i e esponse), exhibi ing acclima ion (neu al esponse) and exhibi ing an addi ional ecological s ess memo y (imp o ed esponse). Acclima ion helps o p e en s ess damage and o p omo e eco e y, despi e o en leading o educed g ow h du ing s ess. An ecological s ess memo y exis s, when he plan keeps a so o “s ess imp in ” a e s ess exposu e ha imp o es plan esponse o ecu en s ess compa ed o plan s wi hou s ess memo y. Lagged s ess e ec s a e de imen al e ec s ha occu some ime a e he s ess occu ed. S ess damage may lead o e en g ea e damage o comple e collapse when ecu en s ess is applied Manusc ip 3: Ecological s ess memo y and c oss s ess ole ance in plan s in he ace o clima e ex emes 94 An ecological s ess memo y migh in ol e he pe sis ence o acclima ion mechanisms and p o ec i e subs ances. Howe e , i is no acclima ion pe se, as such a memo y emains ac i e long a e he s ess has been applied, and enables he plan o espond quicke and mo e adequa e o a ecu en s ess e en . I hus equi es a pe sis ing “imp in ” modi ying u u e s ess esponse. Fo acclima ion, he plan does no need o expe ience eal s ess, as, e.g. o os acclima ion, ce ain en i onmen al cues a e su icien o igge acclima ion wi hin he seasonal li e cycle pe o mance. Ecological s ess memo y has a empo al dimension and, in his sense can only be s udied a e he s ess s opped and he plan eco e ed, e.g. ook up i s p e-s ess me abolism again and epai ed o compensa ed damage. A e applying ecu en s ess, ecological s ess memo y should lead o an imp o ed pe o mance when compa ed o plan s wi hou a pe sis ing s ess memo y. Con as ingly, lagged o delayed s ess e ec s a e de imen al e ec s o single s ess e en s ha become clea ly appa en only a e some ime, e.g. when he plan dies o a communi y collapses (Fig. 1). The esponse o ees o d ough , o ins ance, is o en exp essed in inc eased mo ali y bu his may happen e en se e al yea s a e he d ough e en (Bigle e al., 2007). Likewise, al e a ions in soil os e en s can lead o inc eased mo ali y o dwa -sh ubs a e mo e han one yea , bu wi h no appa en e ec s in he i s yea a e he s ess e en (K eyling e al., 2010). Such lagged esponses clea ly indica e ca y-o e e ec s in i ness which a e no easily de ec able di ec ly a e he e en and which may explain he indings o educed esilience upon epea ed s ess e en s. Fo example, Llo e e al. (2004) in es iga ed he impac o ecu en d ough (1985 and 1995) on esp ou ing and die-back in Que cus ilex. They ound a p og essi e loss o indi idual esilience upon ecu en d ough , as he abili y o su i e and esp ou was educed compa ed o he i s d ough . Muelle e al. (2005) examined Pinus edulis and Junipe us monospe ma mo ali y a e ex eme d ough episodes and also ound a educ ion o esilience and a highe mo ali y a e o a ecu en d ough in 2002 compa ed o he p e ious d ough in 1996. I may ha e been ha de imen al e ec s ha e pe sis ed e en a e many yea s and plan s may no ha e been eco e ed be o e he ollowing s ess exposu e: S a ch s ocks in ligno ube s o Q. ilex we e ound no o be es o ed o hei p e-s ess alues e en 10 yea s a e an ex eme d ough (Lopez e al., 2009). Thus, o clea ly dis inguish such lagged s ess e ec s om an ecological s ess memo y i may be necessa y o expe imen s o apply a ecu en s ess e en and o compa e he answe o ecu en ly s essed plan s o single s essed plan s. Only when he pe o mance o ecu en s ess is imp o ed when compa ed o single s essed plan s, he de ini ion allows calling his an ecological s ess memo y. Manusc ip 3: Ecological s ess memo y and c oss s ess ole ance in plan s in he ace o clima e ex emes 95 In he ollowing, we e iew mechanisms o d ough ole ance and e idences o d ough memo y, o os ole ance and os memo y as well as o hea s ess ole ance and hea s ess memo y. 3. D ough ole ance and d ough memo y Plan s a e able o acclima e o d ough s ess, he eby inc easing hei d ough ole ance. Mechanisms o acclima ion include he accumula ion o osmop o ec i e p o eins, like dehyd ins (Bohne , 2000; Lambe s e al., 2008), he accumula ion o soluble suga s (Lambe s e al., 2008; Wal e e al., 2012), a educ ion o he pho osyn he ic appa a us along wi h addi ional mechanisms o p e en damage by eac i e oxygen species (Munne-Bosch and Aleg e, 2000) and he accumula ion o compa ible solu es (p oline, be aine) (Bohne , 2000). Changes in gene exp ession ha accompany d ough acclima ion a e o en ABA- media ed, and up egula ed genes include genes o he LEA amily (la e emb yogenesis abundan ) (Bohne , 2000; Lambe s e al., 2008). In addi ion o physiological changes, pheno ypic and mo phological esponses can be ini ia ed du ing d ough s ess, such as an inc eased oo o shoo a io o he de elopmen o oo s in deepe soil laye s (Newman e al., 2006). Some ecen indings indica e he exis ence o an ecological d ough memo y: Wal e e al. (2011) ound an inc ease in pho op o ec ion in single g ass plan s unde epea ed d ough when compa ed o plan s ha we e no subjec ed o d ough p e iously, e en se e al weeks a e he i s d ough was applied and a e he plan s we e comple ely cu and eg own (Fig. 2A). Simila ly, Ona e e al. (2011) showed ha U ica dioica subjec ed o combined d ough and nu ien de iciency in hei ju enile phase e ealed imp o ed d ough s ess ole ance in ma u e lea es, especially in ep oduc i e shoo s. Unde labo a o y condi ions, Goh e al. (2003) ound ha A abidopsis haliana epea edly subjec ed o high le els o abscisic acid (ABA), also in ol ed in d ough s ess signaling and esponse, led o a o ma ion o ecological s ess memo y, as gene exp ession was changed in esponse o ollowing s ess e en s compa ed o non- ea ed plan s. Knigh e al. (1998) obse ed changes in d ough s ess-induced calcium-signaling a e plan s had encoun e ed ei he osmo ic o oxida i e s ess p e iously. Manusc ip 3: Ecological s ess memo y and c oss s ess ole ance in plan s in he ace o clima e ex emes 96 Fig. 2 Examples o (A) ecological d ough memo y (adap ed om Wal e e al. 2011), (B) ecological hea memo y (adap ed om Whi le e al. 2009) and (C) o an ecological c oss-s ess memo y (adap ed om K eyling e al. 2012b). Plan s we e ei he uns essed (ligh g ay ba s) o subjec ed o d ough s ess (A, C) o hea s ess (B) (da k g ay ba s; p e ea men indica ed on he y-axis) and pe o mance was measu ed unde subsequen d ough (A), hea (B) o os (C) s ess (as indica ed by inse s) in he same plan s (A, C) o in he F 3 gene a ion (B). Di e en le e s indica e signi ican di e ences be ween plan s wi h and wi hou p e-s ess acco ding o he sou ces Cuk e al. (2010) showed ha modi ied ac i i y o an ioxida i e enzymes (ca alase, asco ba e pe oxidase), which a e o en also up egula ed unde d ough , is inhe i ed o he nex gene a ion o A. haliana. I seems likely ha ma e nal plan s inhe i ing s ess ole ance would e eal an ecological s ess memo y hemsel es. To sum up, he e is e idence ha ce ain physiological p ocesses in plan s a e modi ied by o me s ess e en s. These modi ica ions can be decisi e in ace o epea ed e en s and may s imula e a as e s a o p o ec i e mechanisms and inc eased s ess ole ance and compensa ion. Howe e , s udies in es iga ing his opic a e a e and mo e s udies in es iga ing d ough memo y in di e en species and also unde mo e na u al condi ions and s udies compa ing he esponse o ecu en d ough manipula ions o he esponse o single d ough manipula ions a e needed in he u u e. 50 Seed numbe pe plan P and F1 23°C P and F1 30°C F3: hea 500 1500 a b (B) Li ing biomass [%] 0 10 20 30 40 50 60 70 a b d ough No p e iousd ough P e iousd ough (A) a b No p e iousd ough P e iousd ough os Cold ha diness [LT50 °C] -40 0 -10 -20 -30 (C) 50 Seed numbe pe plan P and F1 23°C P and F1 30°C F3: hea 500 1500 a b (B) 50 Seed numbe pe plan P and F1 23°C P and F1 30°C F3: hea 500 1500 a b (B) Li ing biomass [%] 0 10 20 30 40 50 60 70 a b d ough No p e iousd ough P e iousd ough (A) Li ing biomass [%] 0 10 20 30 40 50 60 70 a b d ough No p e iousd ough P e iousd ough (A) a b No p e iousd ough P e iousd ough os Cold ha diness [LT50 °C] -40 0 -10 -20 -30 (C) a b No p e iousd ough P e iousd ough os Cold ha diness [LT50 °C] -40 0 -10 -20 -30 a b No p e iousd ough P e iousd ough os Cold ha diness [LT50 °C] -40 0 -10 -20 -30 (C) Manusc ip 3: Ecological s ess memo y and c oss s ess ole ance in plan s in he ace o clima e ex emes 97 4. F os ole ance and os memo y In egions whe e subze o empe a u es a e eached, pe ennial plan s show he po en ial o acclima e o os o educe os damage, caused by in acellula ice c ys als and dehyd a ion. As apoplas ic ice o ma ion leads o cell dehyd a ion, d ough acclima ion and os acclima ion o en in ol e he same mechanisms, like accumula ion o soluble suga s o ansc ip ion o dehyd ins and LEA-genes (Lambe s e al., 2008; Janska e al., 2010). F os acclima ion is igge ed by low empe a u e and by changes in he pho ope iod (Thomashow, 1999; Janska e al., 2010). Ha dening usually akes se e al weeks, while deha dening, i.e. he loss o os ha diness, can occu wi hin hou s a e empe a u e inc ease (Rapacz e al., 2000), lea ing he plan s ulne able o sho - e m la e os e en s du ing he g owing season o a e win e wa ming e en s. The equency o os days and nigh s is expec ed o dec ease in a ious biomes unde global clima e wa ming (IPCC 2007), ye , an inc ease in minimum empe a u e o e win e is unlikely (Kod a e al., 2011). Obse ed and p ojec ed educ ion in snow co e , which ac s as insula ion o many plan s, in combina ion wi h mo e a iable ai empe a u es may u he exace ba e he equency o os s ess in many no he n egions (K eyling, 2010). As global wa ming may egionally lead o an ea lie deha dening and onse o he g owing season, he isk o la e os damage is likely o inc ease, when he iming o la e os e en s is no changing (Rigby and Po po a o, 2008;Woldendo p e al., 2008). Fu he mo e, global wa ming migh lead o mo e equen eeze- haw cycles du ing win e , possibly associa ed wi h (pa ial) deha dening a e especially wa m win e days, leading o os damage a u he sub-ze o empe a u es (Bokho s e al., 2009). I is well es ablished ha plan s a e able o “ emembe “ low empe a u es o e a ce ain ime span, as e naliza ion, he p omo ion o sp ing lowe ing a o ed by low empe a u es equi es some so o win e -memo y (Sung and Amasino, 2005). He e, we a gue ha cold ole ance acclima ion is no ecological s ess memo y i sel , as he plan s ha den unde low empe a u es bu wi hou expe iencing os s ess. Cold acclima ion a he is an e olu iona y esponse o a oid os s ess. Expe ience o os s ess, o example du ing eeze- haw cycles o unde la e os e en s, could enable he plan o eac di e en ly o he nex os spell, e en wi hou p io acclima ion. Tahkoko pi e al. (2007) ound inc eased an hocyanin le els in sp ing in newly g own s ems o Vaccinium my illus a e plan s had been subjec ed o os s ess in win e , Manusc ip 3: Ecological s ess memo y and c oss s ess ole ance in plan s in he ace o clima e ex emes 98 compa ed o plan s no expe iencing os s ess be o e. This s ongly indica es an ecological os s ess memo y on a whole plan le el. Unde labo a o y condi ions, Knigh e al. (1996) ound a modi ied calcium signa u e in A. haliana a e plan s had expe ienced a cold shock be o e. Calcium ac s as a second messenge in low empe a u e signaling and may he e o e igge al e ed esponse o epea ed os . Howe e , Polle e al. (1996) showed ha sp uce needles su i ing a sp ing os e en e ealed less an ioxida i e enzymes and pigmen s in he ollowing all, hus p obably showing educed os p o ec ion despi e a ansien inc ease in an ioxida i e enzymes a e sp ing os was applied. This shows ha os s ess may lead o a loss o esilience a he han o he o ma ion o a “posi i e” ecological os memo y. To sum up, ew s udies hin owa ds he exis ence o an ecological os s ess memo y (Tahkoko pi e al. 2007; Knigh e al. 1996), al hough he esponse o epea ed os s ess has no ye been in es iga ed. I emains o be elucida ed i os s ess expe ience helps o su i e subsequen os s ess, as one s udy also indica es a dec ease in os esis ance a e os s ess (Polle e al. 1996). 5. Hea s ess ole ance and hea s ess memo y On a cell-le el, hea s ess acclima ion is a he well unde s ood: Upon exposu e o ex emely high empe a u es, exp ession o no mal housekeeping genes is s opped and hea shock p o eins (HSP), which ac o p e en p o ein damage o pho o-oxida ion and which epai al eady dena u a ed p o eins (chape ones) a e inc easingly syn hesized (Pa cellie e al., 2003; Baniwal e al., 2004; Ko ak e al., 2007). Fu he mo e, compa ible solu es like p olin o be aine ac o s abilize p o eins (Schulze e al., 2005). Unde global wa ming i is e y likely ha hea wa es will inc ease in equency and magni ude o e mos land a eas, as indica ed, e.g., by he Eu opean mega-hea wa es in 2003 and 2010 (Schä e al., 2004; Ba ioped o e al., 2011; IPCC 2011). Ye , s udies in es iga ing an ecological hea s ess memo y a e widely lacking in plan ecology. In e es ingly, a ansgene a ional ecological s ess memo y was ound o mild hea s ess: Plan s om he F3 gene a ion showed a hea -speci ic i ness imp o emen when he pa en al and F1 gene a ion had been ea ed wi h mild hea (30°C), e en when he F2 gene a ion was g own unde no mal condi ions (Whi le e al., 2009) (Fig. 2B). As o all Manusc ip 3: Ecological s ess memo y and c oss s ess ole ance in plan s in he ace o clima e ex emes 99 ansgene a ional s udies, he exis ence o ecological s ess memo y wi hin one plan gene a ion was no es ed, bu is implied by he inhe i ance o s ess ole ance. To conclude, we ound no s udy in es iga ing an ecological hea s ess memo y in single plan s. Ne e heless, he esul s by Whi le e al. (2009) indica e he exis ence o a hea s ess memo y imp o ing hea ole ance a e hea s ess was al eady expe ienced in he pas . 6. C oss-s ess memo y As os , hea and d ough s ess all in ol e cell dehyd a ion, acclima ion mechanisms a e pa ly he same (Beck e al., 2007). I is hus possible ha acclima ion and o ma ion o a s ess memo y o one kind o s ess also p e en s damage by o he s esso s, p o iding c oss- s ess memo y and ole ance. Fo ins ance, os ole ance o local popula ions o eco ypes, espec i ely, is ela ed o d ough ole ance (Blodne e al., 2005). Mo e speci ically, exposu e o an ex eme d ough e en in he p eceding yea was ound o suppo la e os ole ance in g ass species (K eyling e al., 2012a) and maximum os ha diness in ju enile Pinus nig a (K eyling e al., 2012b) (Fig. 2C). Ano he o m o c oss-s ess ole ance is inc eased he bi o e esis ance a e he expe ience o abio ic s ess like d ough , caused by an inc ease in C-based seconda y me aboli es upon g ow h es ic ion due o abio ic s ess (He ms and Ma son, 1992). I is unclea ye , how long such modi ica ions o seconda y compounds a e main ained. Howe e , Ag awal (2002) showed ha he p ogeny o ma e nal plan s a acked by he bi o es also e ealed highe induced esis ance owa ds he bi o es. This implies he i abili y and hus a o m o s ess memo y o induced de ense, al hough i is unclea i his could also ac o p e en d ough , hea o os damage. 7. Possible mechanisms behind an ecological s ess memo y As possible mechanisms o an ecological s ess memo y, B uce e al. (2007) sugges he accumula ion o ansc ip ion ac o s o p o eins o acili a e a as esponse upon epea ed s ess exposu e as well as epigene ic mechanisms, such as his one modi ica ions o chemical changes a he DNA (me hyla ion, ace yla ion) ha a e inhe i ed h ough mi o ic o e en meio ic cell di isions (Bossdo e al., 2008; Chinnusamy e al., 2008; Boyko and Ko alchuk, 2011). Ano he possibili y is he accumula ion o p o ec i e subs ances. Howe e , his is no likely o be e y impo an , as syn hesis o p o ec i e subs ances is cos ly and o en p e en s Manusc ip 3: Ecological s ess memo y and c oss s ess ole ance in plan s in he ace o clima e ex emes 100 no mal g ow h (He ms and Ma son, 1992). Today, i is well es ablished ha plan s ess induces epigene ic changes (Goh e al., 2003; Chinnusamy e al., 2008). I was shown ha epigene ic changes upon UV-C and lagellin exposu e (Molinie e al., 2006), upon TMV ( obacco-mosaic- i us)-exposu e (Boyko e al., 2007), upon pa hogen o he bi o e a ack o low nu ien s (Ve hoe en e al., 2010) and upon ni ogen de iciency (Kou e al., 2011) a e inhe i ed. Fu he mo e, inhe i ed DNA-hypome hyla ion in ice seedlings inc eased pa hogen esis ance (Akimo o e al., 2007) and inhe i ed epigene ic changes upon TMV exposu e inc eased pa hogen esis ance in p ogenies (Ka hi ia e al., 2010) Howe e , no gene al he i abili y o epigene ic changes upon s ess exposu e was ound in A. haliana, implying ha ansgene a ional epigene ic memo y seems o be es ic ed o special condi ions (Pecinka e al., 2009). The esul s o Tahkoko pi e al. (2007) hin owa ds epigene ic changes as unde lying mechanisms o an ecological os memo y, as he new s ems g owing in sp ing ne e expe ienced os s ess hemsel es bu e ealed modi ica ions (see Sec ion 4). Thus, in o ma ion had o be con eyed h ough mi o ic di isions. A u he possibili y o e ain an ecological s ess memo y a e changes in phenology o mo phology o he plan ha emain s able o e longe ime scales han me e changes in he accumula ion o p o ec i e subs ances. Shi s in oo o shoo a io in esponse o d ough (Kalapos e al., 1996; Kahmen e al., 2005) o win e wa ming pulses (K eyling e al., 2008) a e one ob ious mo phological esponse wi h implica ions o u u e d ough ole ance. Fu he mo e, speci ic lea a ea can be adap ed o d ough condi ions (Kalapos e al., 1996), he eby educing wa e loss o his issue also o u u e d ough e en s (see Table 1 o examples o mechanisms o an ecological s ess memo y). 8. Resea ch challenges S udies in es iga ing ecological s ess memo y a e a e. Mos s udies on he du a ion and he i abili y o plan s ess a e conduc ed on a cellula le el and ocus on gene ic o epigene ic aspec s. In such s udies, ime spans be ween he ini ial and he epea ed s ess is usually es ic ed o only se e al hou s o days (B uce e al., 2007). Mo e ecologically ele an esea ch and assessmen s o s ess ole ance and ecological s ess memo y a e needed. Fu he mo e, mul igene a ional epigene ic s udies should conside consequences o ex eme wea he e en s mo e p ominen ly. In o de o e alua e he ecological ele ance o ecological s ess memo y, assessing ecologically meaning ul pa ame e s a he plan le el in con olled lab expe imen s needs o Manusc ip 3: Ecological s ess memo y and c oss s ess ole ance in plan s in he ace o clima e ex emes 101 be accompanied by ield expe imen s and obse a ions a e na u ally occu ing ex emes. He e, long- e m ield- and moni o ing s udies in es iga ing he esponse o na u ally occu ing ex emes in na u al plan communi ies migh be e y aluable. To simul aneously elucida e unde lying p ocesses and mechanisms, coope a ion be ween ecologis s and molecula biologis s a e u gen ly needed. Fo ins ance, i is s ill unclea whe he unseasonable os damage in sp ing in luences acclima ion in he ollowing all and i his migh be de imen al o bene icial. In pa icula he absolu e deg ee and he empo al s abili y o ecological s ess memo y equi es a en ion because ecological s ess memo y could play an impo an ole in s abilizing ecological communi ies in he ace o mo e equen ex eme clima ic e en s. He e, we sugges ha ecological s ess memo y can, a leas pa ly, explain he su p isingly weak e ec s o epea ed ex eme d ough e en s on he p oduc i i y o g assland communi ies (Jen sch e al. 2011). He i abili y o bene icial s ess memo y o ollowing gene a ions, po en ially by epigene ic p ocesses, is ano he highly ele an aspec o ou unde s anding o ecological esponse o mo e equen ex eme clima ic e en s. To conclude, he e a e e idences o he exis ence o an ecological s ess memo y. Howe e , mechanisms and consequences a e no ye well in es iga ed. A s ess memo y o single plan s migh ac o s abilize plan communi ies unde equen clima ic ex emes and migh inc ease esilience. I migh be e en possible o mi iga e de imen al e ec s o ex eme e en s by a i icially applying milde s ess on a small scale, e.g. o ag icul u e. Howe e , i is no ye clea i , on a ield and landscape scale, lagged de imen al e ec s migh lead o a educ ion o esilience unde epea ed ex eme wea he e en s ha ou weigh possible posi i e e ec s o an ecological s ess memo y. Fu he mo e, inc eased acclima ion owa ds ecu ing s esso s migh educe mo ali y o plan s, bu migh as well educe ag icul u al yield i plan s educe hei pho osyn he ic ac i i y o p e en damage. Hope ully, u u e esea ch will con ibu e o elucida e mechanisms and consequences o an ecological s ess memo y. Acknowledgemen s: We hank wo anonymous e iewe s o aluable hin s and commen s o cla i y ou concep s and imp o e ou manusc ip . Re e ences: Ag awal, A.A., 2002. He bi o y and ma e nal e ec s: mechanisms and consequences o ansgene a ional induced plan esis ance. Ecology 83, 3408–3415. Akimo o, K., Ka akami, H., Kim, H.-J., Ogawa, E., Sano, C.M., Wada, Y., Sano, H., 2007. Epigene ic inhe i ance in ice plan s. Annals o Bo any 100, 205–217. Manusc ip 3: Ecological s ess memo y and c oss s ess ole ance in plan s in he ace o clima e ex emes 102 Allen, C.D., B eshea s, D.D., 1998. D ough -induced shi o a o es -woodland eco one: Rapid landscape esponse o clima e a ia ion. P oceedings o he Na ional Academy o Sciences o he Uni ed S a es o Ame ica 95, 14839–14842. Baniwal, S.K., Bha i, K., Chan, K.Y., Fau h, M., Ganguli, A., Ko ak, S., Mish a, S.K., No e , L., Po , M., Scha , K.D., T ipp, J., Webe , C., Zielinski, D., Koskull-Do ing, P. on, 2004. Hea s ess esponse in plan s: a complex game wi h chape ones and mo e han wen y hea s ess ansc ip ion ac o s. Jou nal o Biosciences 29, 471–487. Ba ioped o, D., Fische , E.M., Lu e bache , J., T igo, R., Ga cia-He e a, R., 2011. The ho summe o 2010: ed awing he empe a u e eco d map o Eu ope. Science 332, 220–224. Beck, E.H., Fe ig, S., Knake, C., Ha ig, K., Bha a ai, T., 2007. Speci ic and unspeci ic esponses o plan s o cold and d ough s ess. Jou nal o Biosciences 32, 501–510. Beng sson, J., Angels am, P., Elmq is , T., Emanuelsson, U., Folke, C., Ihse, M., Mobe g, F., Nys om, M., 2003. Rese es, esilience and dynamic landscapes. Ambio 32, 389–396. Bigle , C., Ga in, D.G., Gunning, C., Veblen, T.T., 2007. D ough induces lagged ee mo ali y in a subalpine o es in he Rocky Moun ains. Oikos 116, 1983–1994. Blodne , C., Sk oppa, T., Johnsen, O., Polle, A., 2005. F eezing ole ance in wo No way sp uce (Picea abies [L.] Ka s .) p ogenies is physiologically co ela ed wi h d ough ole ance. Jou nal o Plan Physiology 162, 549–558. Bohne , H.J., 2000. Wha makes desicca ion ole able? Genome Biology 1, 1–4. Bokho s , S.F., Bje ke, J.W., Tomme ik, H., Callaghan, T.V., Phoenix, G.K., 2009. Win e wa ming e en s damage sub-A c ic ege a ion: consis en e idence om an expe imen al manipula ion and a na u al e en . Jou nal o Ecology 97, 1408–1415. Bossdo , O., Richa ds, C.L., Pigliucci, M., 2008. Epigene ics o ecologis s. Ecology Le e s 11, 106–115. Boyko, A., Ko alchuk, I., 2011. Genome ins abili y and epigene ic modi ica ion-he i able esponses o en i onmen al s ess? Cu en Opinion in Plan Biology 14, 260–266. Boyko, A., Ka hi ia, P., Zemp, F.J., Yao, Y., Pog ibny, I., Ko alchuk, I., 2007. T ansgene a ional changes in he genome s abili y and me hyla ion in pa hogen-in ec ed plan s ( i us-induced plan genome ins abili y). Nucleic Acids Resea ch 35, 1714–1725. B uce, T.J.A., Ma hes, M.C., Napie , J.A., Picke , J.A., 2007. S ess ul "memo ies" o plan s: e idence and possible mechanisms. Plan Science 173, 603–608. Chinnusamy, V., Gong, Z., Zhu, J.-K., 2008. Abscisic acid-media ed epigene ic p ocesses in plan de elopmen and s ess esponses. Jou nal o In eg a i e Plan Biology 50, 1187–1195. Cuk, K., Gogala, M., Tkalec, M., Vidako ic-Ci ek, Z., 2010. T ansgene a ional s ess memo y in A abidopsis haliana(L.)Heynh.: an ioxida i e enzymes and HSP70. Ac a Bo anica C oa ica 69, 183–197. Eas e ling. D.R., Meehl, G.A., Pa mesan, C., Changnon, S.A., Ka l, T.R., Mea ns, L.O., 2000. Clima e ex emes: obse a ions, modeling, and impac s. Science 289, 2068–2074. Fay, P.A., Ca lisle, J.D., Knapp, A.K., Blai , J.M., Collins, S.L., 2000. Al e ing ain all iming and quan i y in a mesic g assland ecosys em: Design and pe o mance o ain all manipula ion shel e s. Ecosys ems 3, 308– 319. Goh, C.H., Nam, H.G., Pa k, Y.S., 2003. S ess memo y in plan s: a nega i e egula ion o s oma al esponse and ansien induc ion o d22 gene o ligh in abscisic acid-en ained A abidopsis plan s. Plan Jou nal 36, 240– 255. Golinski, M., Bauch, C., A land, M., 2008. The e ec s o endogenous ecological memo y on popula ion s abili y and esilience in a a iable en i onmen . Ecological Modelling 212, 334–341. Hege l, G.C., Hanlon, H., Beie kuhnlein, C., 2011. Elusi e ex emes. Na u e Geoscience 4, 142-143. He ms, D.A., Ma son W. J., 1992. The dilemma o plan s - o g ow o o de end. Qua e ly Re iew o Biology 67, 283–335. IPCC, 2007. Clima e change 2007: The Physical Science Basis. Con ibu ion o Wo king G oup I o he ou h assessmen epo o he in e go e nmen al panel on clima e change. Camb idge Uni e si y P ess, Camb idge & New Yo k. IPCC, 2011. IPCC Special Repo on Managing he Risks o Ex eme E en s and Disas e s o Ad ance Clima e Change Adap a ion. Fac s Shee . h p://www.ipcc.ch/news_and_e en s/docs/s ex/SREX_ ac _shee .pd (las accessed No embe 28 h, 2011). Janska, A., Ma sik, P., Zelenko a, S., O esna, J., 2010. Cold s ess and acclima ion-wha is impo an o me abolic adjus men ? Plan Biology 12, 395–405. Jen sch, A., K eyling, J., Beie kuhnlein, C., 2007. A new gene a ion o clima e-change expe imen s: e en s, no ends. F on ie s in Ecology and he En i onmen 5, 365–374. Jen sch, A., K eyling, J., Elme , M., Gellesch, E., Glase , B., G an , K., Hein, R., La a, M., Mi zae, H., Nadle , S., Nagy, L., O ieno, D., P i sch, K., Rasche , U., Schädle , M., Schlo e , M., Singh, B., S adle , J., Wal e , J., Wells ein, C., Wöllecke, J., Beie kuhnlein, C., 2011. Clima e ex emes ini ia e ecosys em- egula ing unc ions while main aining p oduc i i y. Jou nal o Ecology 99, 689–702. Manusc ip 4: Do plan s emembe d ough ? Hin s owa ds a d ough memo y in g asses 109 Ge many (49°55´19´´N,11°3455``E, 365 m asl). Tubes we e embedded in o he homogenized soil (loamy sand consis ing o 82 % sand, 13 % sil , 5 % clay o a dep h o 80cm). Plan s we e kep unde na u al ambien condi ions o wo yea s and ha es ed wice pe yea . Fu he mo e, ubes we e pe iodically weeded. In Ap il 2009, ubes, including he soil, we e a anged unde a ain-ou -shel e o he expe imen . To a oid he la e al low o wa e in o he po s, hey we e placed on pla es on a plas ic shee . The anspa en ain-ou shel e was le open a he side up o 80 cm, allowing ai exchange nea o he su ace and hus a oiding g eenhouse-e ec s. Plan s we e subjec ed o wo di e en ea men s and a anged in a comple ely andomized design: 28 plan s in he ecu en d ough ea men ( wo plan s died in he wo yea s be o e he expe imen s a ed) we e subjec ed o an ea ly d ough in June 2009 (D1), whe eby wa e was comple ely wi hheld o 16 days om June 3 d un il June 18 h. The same plan s we e subjec ed o a la e d ough (D2), whe eby wa e was wi hheld o 16 days om Sep embe 4 h o Sep embe 19 h. The ecu en d ough ea men was compa ed o a single d ough ea men : 27 eplica es ( h ee plan s died p io o he expe imen ) we e wa e ed egula ly e e y hi d day wi h 300ml ain wa e (C1) while he i s d ough pe iod was applied o double-s essed plan s. C1 plan s we e exposed o hei i s d ough in Sep embe (C2), concomi an o he second d ough o he ecu en d ough ea men (See Fig. 1 o an o e iew). Fig. 1 O e iew on he expe imen al ime cou se and on applied ea men s. In June, d ough was applied o 16 days o D1 plan s, while C1 plan s we e wa e ed egula ly. All plan s we e wa e ed a e he 16 h day un il he onse o he la e d ough in Sep embe . In Sep embe all plan s (D2 and C2) we e subjec ed o he d ough o 16 days and we e wa e ed a e ha . Only in compa ing plan s subjec ed o a second, la e d ough (D2) o plan s expe iencing hei i s d ough (C2) a he same poin in ime, we could p e en con ounding o po en ial d ough memo y e ec s wi h seasonali y o iming e ec s. A compa ison be ween he esponse o he i s d ough in June and he esponse o he second d ough in Sep embe is hus no alid o in es iga e po en ial d ough memo y e ec s. We also did no ha e a well- wa e ed con ol in Sep embe , as we we e in e es ed in a po en ial “d ough -memo y”, which can only be in es iga ed by compa ing single-s essed wi h double-s essed plan s. To quan i y e ec s o a single, ea ly d ough (D1), d ough plan s we e compa ed o well- wa e ed plan s du ing he i s d ough pe iod (C1). All plan s we e wa e ed wi h he same June 3 d-18 h Sep . 4 h-19 h D1 C1 D2 C2 d ough ed wa e ed June 3 d-18 h Sep . 4 h-19 h D1 C1 D2 C2 d ough ed wa e ed Manusc ip 4: Do plan s emembe d ough ? Hin s owa ds a d ough memo y in g asses 110 amoun o wa e (300 ml, e e y hi d day) un il he onse o he expe imen and in be ween he wo d ough ea men s. 2.2. Abo eg ound biomass Abo eg ound biomass was ha es ed on July 5 h, 17 days a e he i s d ough (D1 and C1) ended, and on Oc obe 9 h, 21 days a e he second d ough ended (D2 and C2). A e his ime, e e sible d ough damages should ha e been eco e ed. As we only compa e D2 wi h C2 plan s and D1 wi h C1 plan s, he di e ence o ou days in eco e y ime a e d ough does no ha e an e ec on he esul s. Plan s we e cu 4 cm abo e he g ound in o de o simula e common managemen echniques in meadows, so ed in o li ing (g een) and dead biomass. Dead biomass was de ined as wil ed, b own plan pa s ha los chlo ophyll. Biomass was d ied a 70° C o 72 hou s and weighed. Pe cen age o dead biomass was calcula ed as pe cen age o o en-d ied, dead biomass in ela ion o o e all o en-d ied biomass o indi idual plan s. 2.3. Rela i e lea wa e con en (RWC) Rela i e lea wa e con en was de e mined in he a e noon o he 13 h day o he i s and second d ough ea men (June 15 h and Sep embe 16 h), acco ding o Ba s and Wea he ley (1962).The second lowes lea o each plan was cu , s o ed in a mois ened plas ic bag o anspo , and immedia ely weighed o de e mine esh weigh (FW). Lea es we e s o ed in dis illed wa e a 4° C o e nigh and weighed he nex mo ning o de e mine u gid weigh (TW). A e wa ds lea es we e d ied a 70° C and he d y weigh (DW) was de e mined. RWC was calcula ed as: 100* )( )( (%) DWTW DWFW RWC − − = 2.4. Chlo ophyll a luo escence Chlo ophyll a luo escence was eco ded using a pulse-ampli ude-modula ed pho osyn hesis yield analyze (PAM 2000 and Mini-PAM) (Wal z, E el ich, Ge many) wi h a lea clip holde as desc ibed by Bilge e al. (1995). The second o hi d ully-expanded lea es we e measu ed on ou di e en blades o one indi idual. Fou measu emen s pe plan we e a e aged o u he analysis. We ob ained p edawn luo escence alues (be ween 2:00 and 4:00) a he end o he i s d ough ea men , h oughou he second d ough pe iod and Manusc ip 4: Do plan s emembe d ough ? Hin s owa ds a d ough memo y in g asses 111 h oughou he ea ly eco e y phase a e he second d ough . The maximum quan um e iciency o pho osys em II was calcula ed as F /Fm. Va iable luo escence (F ) and maximum luo escence (Fm) we e measu ed be o e dawn. F was calcula ed as Fm-F0, Fm being he maximum luo escence o he da k adap ed lea a e applying a sa u a ing ligh pulse and F0 being he s eady s a e luo escence yield o he da k adap ed lea (Maxwell and Johnson, 2000). To enable a compa ison be ween absolu e luo escence alues, a luo escence s anda d ma e ial was measu ed be o e and a e each measu ing cycle. S anda d measu emen s we e used o no malize he luo escence alues ob ained and o calib a e he wo di e en PAMs in use. P edawn measu emen s o luo escence a he da k adap ed lea allow d awing conclusions abou unde lying p ocesses which al e plan pho osyn he ic pe o mance and abou pho oinhibi o y damage and non-pho ochemical quenching (Maxwell and Johnson, 2000). Absolu e F0 and Fm alues we e aken o sepa a e he e ec s o pho odamage, becoming appa en wi h an inc ease o F0, om he e ec s o pho op o ec ion ela ed o enhanced non-pho ochemical quenching, becoming appa en wi h a dec ease in Fm (Osmond e al., 1993; A aus e al., 1998; Maxwell and Johnson, 2000). 2.5. Lea gas exchange The ne CO2 assimila ion a e (Pn) and anspi a ion we e measu ed a midday (be ween 11:30 and 13:30), when d ough s ess should be a i s maximum, due o high empe a u e and i adiance. I was measu ed on he second, ully de eloped lea o each plan using a gas-exchange sys em (Li-6400, Li-Co , Lincoln, NE, USA) equipped wi h a CO2 ca idge o adjus and main ain cons an CO2 o 400 μmol mol–1 ai wi hin he lea cu e e. Gas exchange measu emen s we e conduc ed on clea days wi hou clouds o main ain cons an PPFD. A e eaching s eady-s a e pho osyn hesis, da a we e logged. The lea a ea was es ima ed simul aneously by measu ing he lea wid h and la e on used o co ec alues o ne pho osyn hesis, as lea blades did no ill he whole lea cu e e. Gas exchange was measu ed unde ambien ligh condi ions a he end o he i s d ough pe iod, in he ea ly and la e d ough pe iod and in he ea ly eco e y phase a e he d ough . 2.6. S a is ical analysis To de e mine signi ican di e ences be ween single and ecu en d ough ea men s, analyses o a iance we e pe o med o all a iables o each sampling da e. We de ined “ ea men ” as a ixed ac o . “P o enance” was a andom ac o in his expe imen , as Manusc ip 4: Do plan s emembe d ough ? Hin s owa ds a d ough memo y in g asses 112 p o enances we e chosen andomly ou o a la ge popula ion o p o enances, and as we we e no in e es ed in he speci ic p o enances, bu in he whole popula ion o ou plan s (Do mann and Kühn, 2009). We examined he esiduals agains i ed plo s and no mal qq-plo s p io o each analysis o es whe he he assump ions o ANOVA, homogenei y o a iances and no mali y, could be me (Fa away, 2006). I his was no he case, da a we e powe ans o med ( luo escence da a), log- ans o med (absolu e biomass da a) o a csin- ans o med ( ela i e wa e con en ) acco dingly. All s a is ical analyses we e pe o med using R 2.11.0 (R De elopmen Co e Team 2010). Fo mixed e ec models we used he so wa e package nlme (Pinhei o e al., 2008). 3. Resul s 3.1. E ec s o he i s d ough (D1) Tempe a u e [°C] 10 12 14 16 18 20 22 24 26 28 0 2 4 6 8 10 12 14 P ecipi a ion [mm] 0 5 10 15 20 8 10 12 14 16 18 20 22 24 26 A e age Tempe a u e Maximum Tempe a u e Day o expe imen 0 5 10 15 P ecipi a ion (a) (b) Tempe a u e [°C] 10 12 14 16 18 20 22 24 26 28 0 2 4 6 8 10 12 14 P ecipi a ion [mm] 0 5 10 15 20 8 10 12 14 16 18 20 22 24 26 A e age Tempe a u e Maximum Tempe a u e Day o expe imen 0 5 10 15 P ecipi a ion (a) (b) Tempe a u e [°C] 10 12 14 16 18 20 22 24 26 28 0 2 4 6 8 10 12 14 P ecipi a ion [mm] 0 5 10 15 20 8 10 12 14 16 18 20 22 24 26 A e age Tempe a u e Maximum Tempe a u e A e age Tempe a u e Maximum Tempe a u e Day o expe imen 0 5 10 15 P ecipi a ion (a) (b) Fig. 2 Cou se o daily maximum empe a u e (g ey squa es), daily a e age empe a u e (black ci cles) and p ecipi a ion (da k g ey ba s) a he s udy si e du ing he expe imen al pe iods in June (a) and Sep embe (b). Ve ical black line indica es he s a o he d ough ea men , g ey dashed line indica es when ewa e ing s a ed. Manusc ip 4: Do plan s emembe d ough ? Hin s owa ds a d ough memo y in g asses 113 Tempe a u e and p ecipi a ion da a du ing he expe imen al pe iod in June a e shown in Figu e 2. Ea ly d ough ea men (D1) in June signi ican ly educed he ela i e lea wa e con en measu ed a he end o he d ough ea men compa ed o he well-wa e ed con ol (C1) by a ound 22% (P<0.001; Fig. 3a). C1 D1 Rela i e lea wa e con en [%] 020 40 60 80 100 (a) Pn [µmol CO 2 m -2 s -1 ] 02 4 6 8 10 12 (d) C1 D1 F /Fm 0.0 0.2 0.4 0.6 0.8 1.0 (c) C1 D1 C1 D1 Li ing biomass [%] 0204060 80 (b) C1 D1 Rela i e lea wa e con en [%] 020 40 60 80 100 (a) Pn [µmol CO 2 m -2 s -1 ] 02 4 6 8 10 12 (d) C1 D1 F /Fm 0.0 0.2 0.4 0.6 0.8 1.0 (c) C1 D1 C1 D1 Li ing biomass [%] 0204060 80 (b) Fig. 3 E ec s o he i s d ough (D1) in June (da k g ey) compa ed o well-wa e ed con ol plan s (C1) (ligh g ey) on ela i e wa e con en o lea es (a), pe cen age o li ing biomass (b), maximum quan um e iciency F /Fm, measu ed p edawn (c), and ne pho osyn hesis Pn, measu ed du ing midday (d). Rela i e lea wa e con en (a) was measu ed h ee days be o e he end o he d ough ea men , maximum quan um e iciency F /Fm (c) and ne pho osyn hesis Pn (d) we e eco ded on he las day o he d ough ea men . To al abo eg ound ha es was conduc ed 17 days a e he end o he i s d ough ea men and he pe cen age o dead biomass (d y weigh ) (b) was calcula ed. Means and SE a e shown, as e isks indica e signi icance (*, P≤ 0.05; **, P≤ 0.01; ***, P≤ 0.001). Ea ly d ough ea men (D1) had no e ec on he o al abo eg ound biomass when compa ed wi h he well-wa e ed con ol ea men (C1) (da a no shown). Howe e , d ough Manusc ip 4: Do plan s emembe d ough ? Hin s owa ds a d ough memo y in g asses 114 educed he pe cen age o li ing biomass signi ican ly by a ound 15% (P=0.03; Fig. 3b). On he las day o ea ly d ough ea men (D1) he pho ochemical e iciency (F /Fm) o plan s unde d ough was signi ican ly educed compa ed o he well-wa e ed con ol (P=0.004; Fig. 3c). Plan s unde ea ly d ough ea men (D1) exhibi ed educed ne pho osyn hesis by 58% on he las day o he ea ly d ough ea men (P<0.001; Fig. 3d). 3.2. E ec s o ecu en d ough e en s (D2) compa ed o a single d ough e en (C2) in Sep embe 3.2.1. Tempe a u e, Rela i e lea wa e con en and p oduc i i y in he second d ough pe iod A e age daily empe a u es du ing he expe imen al pe iod in Sep embe anged be ween 10°C and 20°C (Fig. 2b). The maximum empe a u e was excep ionally high wi h 25.5° C on he 6 h day o he expe imen in Sep embe and lowes on he 11 h day o he expe imen wi h 12.3° C. The wi hholding o wa e was e lec ed in he ela i e wa e con en o he lea es, which was educed o a ound 65% bu did no e eal any di e ences be ween single and ecu en ly d ied plan s (D2 s. C2: P=0.38, Fig. 4a). C2 Rela i e lea wa e con en [%] 02040 60 80 D2 (a) Li ing biomass [%] 0 10203040506070 C2 D2 (b) C2 Rela i e lea wa e con en [%] 02040 60 80 D2 (a) Li ing biomass [%] 0 10203040506070 C2 D2 (b) Fig. 4 Lea wa e s a us a he end o he second d ough pe iod 13 days a e wa e had been wi hheld (a), and pe cen age o li ing biomass o he second ha es , h ee weeks a e he end o he d ough ea men (b). Means +/- 1 SE a e shown, as e isk indica es signi icance o di e ence (*, P≤ 0.05). Plan s in single and ecu en d ough ea men s p oduced an almos equal amoun o o e all (li ing and dead) abo eg ound biomass (1.47g in ecu en d ough ea men (D2) and 1.49g in single d ough ea men (C2)). Howe e , he pe cen age o li ing biomass, was inc eased signi ican ly by 7% in plan s subjec ed o ecu en d ough (D2) compa ed o plan s expe iencing hei i s d ough (C2) (P=0.048) (Fig. 4b). To al li ing biomass was inc eased Manusc ip 4: Do plan s emembe d ough ? Hin s owa ds a d ough memo y in g asses 115 by 10% in ecu en d ough ea men , al hough his e ec did no p o e o be signi ican (0.91g in ecu en d ough ea men and 0.82g in single d ough ea men , P=0.18). 3.2.2. Pho osyn he ic pa ame e s Wi h p og essi e d ough s ess, F /Fm in he single and ecu en d ough ea men declined, eaching a minimum o he double-s essed plan s (D2) on he 14 h day o he expe imen , wo days be o e ewa e ing (Fig. 5a). Single-s essed plan s (C2) al eady eached minimal quan um e iciency on he 11 h day o he d ough and F /Fm alues ose again a e ha . The loss o lea wa e and pho ochemical e iciency unde ex eme d ough was e lec ed in a decline o ne pho osyn hesis by mo e han 60% compa ed o ne pho osyn hesis be o e he d ough ea men s a ed (D2 and C2) (Fig. 5b). Fig. 5 Cou se o maximum quan um e iciency F /Fm, measu ed p edawn (a) and ne pho osyn hesis, measu ed du ing midday (b) in A. ela ius subjec ed o ecu en o single d ough be o e and du ing he d ough phase and ewa e ing in Sep embe . Da k g ey dashed line indica es he s a o he d ough , black dashed line indica es he end o he d ough and he s a o ewa e ing in Sep embe . Means +/- 1 SE a e shown, as e isk indica es signi icance o di e ence be ween single and ecu en d ough ea men s on single days (*, P≤ 0.05). G asses unde ecu en d ough (D2) showed lowe maximum quan um e iciency compa ed o plan s exposed o a single d ough (C2), om he ele en h day o he expe imen un il he end o measu emen s, 10 days a e he onse o ewa e ing (Fig. 5a). This educ ion was signi ican on he las day o measu emen s unde he d ough (14 h day o he Rewa e ing 345678 Pn [µmolCO 2 m -2 s -1 ] Day o expe imen 1 4 7 10 13 16 19 22 25 0.4 0.5 0.6 0.7 0.8 F /Fm C2- single D2- ecu en D ough (a) (b) Rewa e ing 345678 Pn [µmolCO 2 m -2 s -1 ] Day o expe imen 1 4 7 10 13 16 19 22 25 0.4 0.5 0.6 0.7 0.8 F /Fm C2- single D2- ecu en D ough (a) (b) Manusc ip 4: Do plan s emembe d ough ? Hin s owa ds a d ough memo y in g asses 116 expe imen , P=0.05) and on he i s day a e ewe ing (17 h day o he expe imen , P=0.02) (Fig. 5a). Du ing d ough and in he pos -d ough eco e y phase, plan s subjec ed o single (C2) and ecu en d ough (D2) did nei he di e signi ican ly ega ding ne pho osyn hesis, no anspi a ion ( anspi a ion da a no shown). Howe e , on he 14 h day o he expe imen , ne pho osyn hesis o g asses subjec ed o ecu en d ough (D2) was 25% lowe compa ed o g asses being subjec ed o hei i s d ough (C2), bu his e ec was no signi ican (P=0.11). A close look a luo escence pa ame e s on bo h o hese days wi h signi ican educ ions in F /Fm (14 h and 17 h day o expe imen ) e ealed ha he dec ease o F /Fm in plan s subjec ed o ecu en d ough can be explained by a dec ease in Fm a he han by an inc ease in F0 (Table 2). Two days be o e ewa e ing, Fm was educed by a ound 20% (14 h day o he expe imen , P=0.07). F0 was non-signi ican ly educed by a ound 6% (P=0.57). Fm was educed by a ound 10% on he i s day o measu emen s a e ewe ing (17 h day o he expe imen , P=0.3) as opposed o F0, which was educed by only 0.5% in plan s subjec ed o ecu en d ough (P=0.94) (Table 2). Table 2 Maximum luo escence (Fm) and s eady s a e luo escence (F0) wo days be o e ewa e ing (14) and one day a e ewa e ing (17) in he ecu en and single d ough ea men du ing he expe imen al pe iod in Sep embe . Means +/- 1 SE a e shown (n=5). day o expe imen 14 17 Fm single 0.696 ± 0.0031 0.793 ± 0.0026 ecu en 0.553 ± 0.0030 0.718 ± 0.0025 F0 single 0.157 ± 0.0001 0.170 ± 0.0001 ecu en 0.148 ± 0.0001 0.169 ± 9.9e-5 F /Fm o bo h ea men s eco e ed g adually a e he d ough ea men ended, eaching p e- d ough alues en days a e ewa e ing had s a ed, on he 26 h day o he expe imen (Fig. 5a). One week a e ewe ing, ne pho osyn hesis had been almos comple ely es o ed, showing educ ions o only 9% compa ed o p e-d ough alues. 4. Discussion This s udy in es iga ed, whe he A. ela ius plan s o six mid- and eas e n Eu opean p o enances can emembe d ough s ess o e an en i e ege a ion pe iod e en a e a ha es . We hypo hesized ha plan s would no show di e en pe o mance unde ecu en d ough . This hypo hesis was no con i med, as g asses esponded consis en ly di e en in ecu en d ough as compa ed o a single d ough , indica ing enhanced pho op o ec ion. Manusc ip 4: Do plan s emembe d ough ? Hin s owa ds a d ough memo y in g asses 117 Su p isingly, his e ec pe sis ed a e o al abo eg ound biomass ha es and eg ow h. The obse ed changes in eac ion o ecu en d ough a e no in acco dance wi h indings indica ing educed esis ance o esilience a e ha ing al eady been exposed o d ough s ess be o e (Llo e e al., 2004; Za alloni e al., 2008 ). Howe e , hese we e conduc ed unde ield condi ions and Za alloni e al. (2008) in es iga ed communi y esponses. Thus, di e en ou come o he expe imen s a e no su p ising. The indings a e in acco dance wi h epo ed s ess imp in e ec s o s ess memo y (B uce e al., 2007). Howe e , o ou knowledge, no s udy has al eady p o ided e idence ha g asses do emembe d ough s ess e en a e a ha es and can exhibi imp o ed pe o mance in he ace o epea ed abio ic s ess o e such a long du a ion. S ess imp in and acclima ion we e p e iously mos ly epo ed o las o “se e al days” (B uce e al., 2007). Unde se e e d ough , g asses expe iencing ecu en d ough showed educed maximum quan um e iciency F /Fm (Fig. 5a). This was mainly ela ed o educ ions in maximum luo escence, indica ing enhanced dissipa ion o ligh ene gy o p e en pho odamage (Maxwell and Johnson, 2000). Osmond e al. (Osmond e al., 1993) and A aus e al. (A aus e al., 1998) sugges ha a co ela ion o educed F /Fm wi h an inc ease o F0 can be in e p e ed as ch onic pho oinhibi o y damage due o he deg ada ion o he D1 p o ein in he eac ion cen e s. By con as , a cons an F0 and dec easing Fm alues, as in ou s udy, poin owa ds pho oinhibi ion ela ed o enhanced non-pho ochemical quenching ia he Xan hophyll cycle and hus indica e pho op o ec ion (A aus e al., 1998). Ano he possible explana ion o dec eased F /Fm caused by educed Fm alues migh be a educ ion o chlo ophyll. Un o una ely, we did no measu e chlo ophyll con en in ou s udy. Howe e , a educ ion o chlo ophyll can be conside ed as a ea u e o acclima ion, as i educes he possibili y o pho odamage because o an excess o ene gy (Munne-Bosch and Aleg e, 2000). Ne pho osyn hesis did no e eal any di e ences be ween g asses unde ecu en (D2) and single d ough (C2), bu showed a end owa ds lowe pho osyn hesis in plan s ecei ing ecu en s ess unde ex eme d ough (Fig. 5b). This is in acco dance wi h he educed pho ochemical e iciency in plan s ecu en ly expe iencing d ough (D2). The esul s o he abo eg ound biomass suppo he hypo hesis o enhanced pho op o ec ion o double-s essed plan s, as he pe cen age o li ing biomass was signi ican ly inc eased in plan s expe iencing hei second d ough , al hough o al abo eg ound biomass o o al li ing biomass we e no signi ican ly al e ed. Plan s can adap o d ough by enhancing oo g ow h. Howe e , ela i e wa e con en o he lea es was no signi ican ly enhanced in plan s expe iencing ecu en d ough , Manusc ip 4: Do plan s emembe d ough ? Hin s owa ds a d ough memo y in g asses 118 indica ing ha he obse ed esul s can no be explained by changes in oo biomass o imp o ed wa e up ake mechanisms. Ecophysiological measu emen s did no e eal any consis en di e ences be ween ecu ing (D2) and single d ough (C2) in he pos d ough eco e y phase. Maximum quan um e iciency wo days a e ewe ing in g asses subjec ed o ecu ing d ough was signi ican ly lowe compa ed o g asses subjec ed o hei i s d ough . This was mo e likely ela ed o inc eased s ess le els unde d ough a he han a lowe eco e y a e. Some s udies indica ed ha eco e y a e depends on expe ienced s ess le el (Miyashi a e al., 2005; Resco e al., 2009). The signi ican ly lowe pe cen age o dead biomass in plan s subjec ed o ecu ing d ough can be a sign o quicke eco e y, bu is mo e likely o be a sign o imp o ed pho op o ec ion, as discussed abo e. In acco dance o o he s udies, he eco e y o ecophysiological pa ame e s was qui e as , almos eaching p e-s ess le els a e abou one week (Galle e al., 2007; Galmes e al., 2007). We did no elucida e unde lying molecula o biochemical mechanisms o acclima ion in his s udy, as we we e in e es ed in he e ec s o ecu en d ough on ag icul u al ele an pe o mance pa ame e s. Thus, we can only hypo hesize abou po en ial long-las ing acclima ion p ocesses in he g asses. The obse ed pheno ypic plas ici y could be ei he explained by belowg ound dynamics o by long-las ing changes in gene exp ession, ende ing he plan s mo e pe missi e o eac quickly o ecu en s ess, e.g. epigene ic p ocesses (Aubin-Ho h and Renn, 2009;Molinie e al., 2006; Bi d, 2007; Bossdo e al., 2008). Ve hoe en e al. (Ve hoe en e al., 2010) ecen ly showed ha s ess induces changes in me hyla ion pa e ns and ha hese pa e ns a e he i able. An in es iga ion o he changes in me hyla ion pa e ns as a esponse o d ough and a link o obse ed me hyla ion pa e ns o s ess esponse a e e y p omising. Fu he mo e, we could only in es iga e six p o enances o A. ela ius plan s, which o igina ed mainly om a eas in Eu ope wi h qui e simila clima ic condi ions. O he , mo e di e en p o enances we e no su i ing in su icien eplica es o ou s udy. Ne e heless, an ex ension o ou expe imen o o he p o enances and plan g oups seems p omising, as hey may e eal di e en acclima ion pa e ns and he e o e also di e en esponses o ecu en d ough . 5. Conclusion To conclude, ou s udy indica es ha g asses unde d ough e ain a long-las ing s ess imp in ha acili a es a as e and mo e p o ec i e esponse owa ds a ecu en d ough . Manusc ip 5: Cold ha diness o Pinus nig a A nold as in luenced by geog aphic o igin, wa ming, and ex eme summe d ough 125 He e, eigh p o enances o P. nig a om au och honous o igins and om sou he n Ge many we e es ed o hei cold ha diness in a common ga den expe imen in sou he n Ge many. We hypo hesized ha (1) cold ha diness di e s be ween p o enances, wi h p o enances om colde o igins displaying supe io cold ha diness, and ha (2) cold ha diness is a ec ed by clima ic expe iences o he indi iduals wi h d ough inc easing cold ha diness and wa ming dec easing cold ha diness. We u he expec ed ha (3) di e ences in cold ha diness be ween p o enances a e physiologically- ela ed o he con en o soluble ca bohyd a es and lipid composi ion o he needles, and ha (4) he (sub-) medi e anean species P. nig a is less os - ole an han ee species na i e o Cen al Eu ope, while (5) cold ha diness o he ine oo s o P. nig a is high compa ed o cold ha diness o i s oliage as i na u ally occu s in egions wi hou con inuous snow co e . 2. Ma e ial and me hods Ju eniles o P. nig a om eigh p o enances h oughou Eu ope we e exposed o di e en clima e change scena ios (wa ming and ex eme d ough ) in a common ga den expe imen . Cold ha diness was de e mined by he Rela i e Elec oly e Leakage me hod (REL) in wo consecu i e win e s. The expe imen was es ablished in Bay eu h, Ge many (49°55’19” N, 11°34’55” E) in Ma ch 2009. The long- e m mean annual empe a u e o he si e is 8.2°C, whe eas long- e m mean annual p ecipi a ion is 724 mm. 2.1. Expe imen al design Eigh p o enances o P. nig a (Figu e 1; Table 1) we e ob ained as seeds and cul i a ed a he Ba a ian Ins i u e o Fo es Seeding and Plan ing (ASP) in Teisendo , Ge many om Ap il 2008 o Ap il 2009. Manusc ip 5: Cold ha diness o Pinus nig a A nold as in luenced by geog aphic o igin, wa ming, and ex eme summe d ough 126 Figu e 1: O igins o he a ge p o enances (open ci cles) wi hin he dis ibu ion o P. nig a (black lines and do s o agmen ed popula ions Isaje e al., 2004). X indica es he expe imen al si e. Abb e ia ions o p o enances a e speci ied in Table 1. G ey scales display he mean minimum empe a u e o he pe iod 1950 o 2000 wi h a 5’ spa ial esolu ion (Hijmans e al., 2005). These p o enances a e pa o an in e na ional long- e m p o enance ial which s a ed in 2009 (Hube , 2011). The p o enances s em om au och honous popula ions o P. nig a excep o he p o enance om Zellingen, Ge many, which was in oduced om Aus ian sou ces in 1909. Subspecies iden i ies o he p o enances a e assigned geog aphically and mo phologically (Table 1), as gene ic analyses a e no ye a ailable (Hube , 2011). Table 1 O igins o a ge p o enances used in he expe imen wi h co esponding clima ic in o ma ion. Skie: Iden i ica ion numbe in an in e na ional p o enance ial (Hube , 2011). MAT: Mean Annual Tempe a u e; MinT: Mean Minimum Tempe a u e; MAP: Mean Annual P ecipi a ion; P ecip. Seasonali y: Coe icien o a ia ion in mean mon hly p ecipi a ion sum. All clima e da a o he pe iod 1950 o 2000 om wo ldclim (Hijmans e al., 2005). P o enance Coun y Subspecies Skie No h Eas Al i ude (m) MAT (°C) MinT (°C) MAP (mm) P ecip. Seasonali y DE Ge many nig a 01 49°53'17" 09°43'16" 290 9.2 -3.1 587 18 AU Aus ia nig a 07 47°46'00" 16°11'00" 369 8.4 -4.9 712 33 YU Se bia nig a 12 43°49'39" 19°35'22" 866 8.7 -5.6 964 17 HR C oa ia nig a/ dalma ica 14 43°26'00" 17°13'00" 256 13.2 1.2 1108 33 IT.N I aly nig a 17 45°42'00" 13°49'00" 372 11.4 -1.2 1212 17 IT.S I aly la icio 19 39°18'08" 16°20'22" 1500 9.0 2.2 1300 48 FR1 F ance nig a 23 44°09'10" 05°52'30" 549 10.7 -2.9 789 16 FR2 F ance la icio 24 44°24'18" 03°58'39" 581 10.8 -0.9 745 19 Manusc ip 5: Cold ha diness o Pinus nig a A nold as in luenced by geog aphic o igin, wa ming, and ex eme summe d ough 127 The C oa ian p o enance s ems om a loca ion e y close o one o he ew au och honous s ands o P. nig a subspecies dalma ica and i s assignmen o he subspecies nig a is somewha ques ionable. The seedlings we e anspo ed o Bay eu h and indi idually plan ed in o 4-li e plas ic po s illed wi h sandy sil (pH 7.3, o al C 1.9%, o al N 0.15%, plan a ailable NO3--N 22.5 mg l-1; plan a ailable NH4+-N 1.8 mg l-1). Selec ion o he plan s occu ed andomly o each p o enance om all hose plan s ali e a he plan ing da e. The mean plan size a he s a o he expe imen was 12.2 cm ± 2.5 cm SD. The po ed indi iduals we e exposed o he ully c ossed h ee old ac o ial combina ion o (1) a d ough manipula ion (d ough and con ol) and (2) a con inuous wa ming manipula ion (wa ming and e e ence) and (3) he p o enance ea men (eigh p o enances). The wo clima e ea men s we e c ossed esul ing in ou clima e manipula ions (con ol, d ough , wa ming, wa ming & d ough ), ha we e eplica ed h ee imes, esul ing in 12 expe imen al uni s in o al. The p o enance ea men was nes ed wi hin each expe imen al uni . Each p o enance was u he eplica ed wi h se en plan s pe expe imen al uni (nes ed eplica es), esul ing in 21 plan s pe ac o ial combina ion o he h ee- ac o ial design and 672 plan s o e all. Each expe imen al uni was co e ed by a single ain-ou shel e (11 m by 7 m, 3.8 m high) cons uc ed o a s eel ame (GlasMe all Rieme GmbH) and co e ed wi h a anspa en polye hylene shee (0.2 mm, SPR5, He mann Meye GmbH) enabling an almos 90% pene a ion o pho osyn he ically-ac i e adia ion. The edge o he ain-ou shel e s was a a heigh o 80 cm. The con ol i iga ion egime simula ed he local daily 30-yea a e age p ecipi a ion. The applica ion was ca ied ou wice a week wi h collec ed ain wa e . The d ough ea men consis ed o 42 days wi hou p ecipi a ion, which ep esen s he local s a is ical 1000-yea ecu ence d ough e en . D ough du a ion was no a p io i se be o e he manipula ions. We moni o ed plan pe o mance du ing he ea men and would ha e s opped he ea men when ei he 66% o he plan s showed wa e s ess symp oms (discolo a ion o oliage) o when 33% o he plan s exhibi ed le hal s ess, o when he local 1000-yea ex eme would be eached. The la e condi ion was se because we assume ha e en s wi h mo e han 1000 yea ecu ence ime a e no oo ealis ic e en when changing equencies o ex emes due o clima e change a e acknowledged (Schä e al., 2004). The same p o ocol was applied in a pa allel expe imen wi h ou g ass species (Beie kuhnlein e al., 2011), which all showed se e e d ough symp oms a e abou 20 days, emphasizing he high d ough ole ance o Pinus nig a. The d ough ea men s a ed on May 27 h 2009 and esul ed in he soil mois u e Manusc ip 5: Cold ha diness o Pinus nig a A nold as in luenced by geog aphic o igin, wa ming, and ex eme summe d ough 128 alling below he pe manen wil ing poin (pF = 4.2) o he soil app oxima ely h ee weeks a e he s a o he ea men (Figu e 2). In he e-we ing phase each indi idual in he d ough ea men s ecei ed 240, 280 and 300 ml on h ee days wi hin one week (in o al 820 ml o 36 mm). Following ha , he po s we e i iga ed acco ding o he con ol p ecipi a ion ea men . To al amoun o p ecipi a ion in he d ough ea men was 13% lowe han in he con ol o e he yea . The d ough was simula ed in he i s yea o he expe imen only. Th oughou he second yea , all plan s ecei ed con ol i iga ion. Figu e 2: Tempe a u e a mean plan heigh , snow co e and soil mois u e (-2.5 o -7.5 cm) o e he cou se o he expe imen . Sampling da es a e indica ed by a ows. The wa ming ea men was pe o med con inuously un il Oc obe in he i s yea o he expe imen and om Ap il o he end o he expe imen in Janua y o he second yea . The wa ming manipula ion ook place bo h passi ely (wind-shel e s which educed he wind speed by 70 % and black loo -co e s e sus whi e loo co e s) and ac i ely (IR- adia ion wi h app oxima ely 30 W pe m²), which inc eased he ai empe a u e a plan heigh by 1.6°C on a e age when he wa ming ea men was a ec ed (Figu e 2). Maximum di e ences we e 5.2°C (single measu emen s) o 3.6°C o daily mean empe a u e. The ou h ea men was a combina ion o d ough and wa ming. The wa ming inc eased he d ough e ec , educing he soil mois u e by ano he 1.5% on a e age (Figu e 2). Manusc ip 5: Cold ha diness o Pinus nig a A nold as in luenced by geog aphic o igin, wa ming, and ex eme summe d ough 129 Du ing he i s win e o he expe imen , plan s we e kep ou side he shel e s in a sand bed om Oc obe o Ap il. Figu e 2 illus a es ha he plan s we e co e ed by snow du ing he coldes pa s o he i s win e . Fo he second win e , plan s we e kep inside he shel e s wi h he wa ming ea men ongoing. 2.2. Response pa ame e s Cold ha diness was quan i ied by a sligh ly modi ied e sion o he ela i e elec oly e leakage me hod (REL) o ex-si u samples acco ding o S imbeck e al. (2007): P e- es s e ealed no di e ences in absolu e alues when he samples we e ozen wi h o wi hou 1ml solu ion con aining an ice nuclea o , p esumably because he su ace o he samples was we and oze a a ound 0°C anyhow. Fu he mo e, highe eezing a es we e applied. A a a e o 0.6°C/h (S imbeck e al. 2007) i would ha e aken 3.5 days o each ou minimum empe a u e, whe eas commonly he a e o 6°C/h is applied (e.g. Su inen 1992, Schabe g 2008). Two needles om he cu en yea we e sampled pe indi idual in mid-win e o bo h yea s (Janua y 20 h in 2010 and Janua y 31s in 2011), insed wi h de-ionized wa e , and cu o 0.5 cm. Samples om he se en nes ed eplica es pe p o enance and expe imen al uni we e combined o o m one mixed sample, homogenized and subsequen ly di ided in o se en subsamples subjec ed o di e en empe a u e le els o one hou (+4.5°C, -7.5°C, -14.5°C, - 23°C, -33°C; -40°C, -196°C (liquid N)) using a con olled en i onmen chambe (Lich - The mos a e Typ 1301, RUMED) and a manually con olled chain o eeze s sequen ially a he lowes empe a u es. Ini ial elec oly e leakage was de e mined in 16 ml 0.1% / T i on X-100_Bides a e 24 h and he inal elec oly e leakage was de e mined a e au ocla a ion o he samples. Elec oly e leakage was quan i ied by he conduc i i y o he solu ion a 25°C measu ed wi h a WTW inolab pH/Cond 720. Cold ha diness is exp essed as he LT50 o each mixed sample, es ima ed by non-linea eg ession o he REL e sus he empe a u e le els using he o mula by Ande son e al. (1988): (1) YT is he REL a empe a u e T, Ymin is he asymp o ic alue o he esponse a iable in uninju ed issue, Ymax is he asymp o ic alue a maximum low- empe a u e s ess, k ep esen s he s eepness o he esponse cu e, and Tm is he midpoin o he symme ical cu e (an es ima e o LT50). Cu e i ing was ca ied ou using a quan ile eg ession and he unc ion “nl q()” om he so wa e package “quan eg” (Koenke , 2006). Manusc ip 5: Cold ha diness o Pinus nig a A nold as in luenced by geog aphic o igin, wa ming, and ex eme summe d ough 130 The mul i ude o di e en echnical p o ocols o REL used in he li e a u e ( eezing wi h o wi hou addi ional solu ion, a ious eezing a es and du a ions, e c.) limi s he compa abili y be ween s udies s ongly. Howe e , he ela i e di e ences wi hin a p o ocol should be obus and mo e o less independen o e.g. eezing a es (Su inen e al 1992). The e o e, we s ick o he in e p e a ion o ela i e di e ences wi hin ou s udy and minimize he discussion o absolu e alues. Cold ha diness o needles om he cu en yea o e minal buds was addi ionally de e mined o ju enile and adul indi iduals o he mos impo an local ee species (P. syl es is, Picea abies, Fagus syl a ica and Que cus pe aea). Samples we e aken on 26 h o Janua y in he i s win e . Th ee mixed samples o se en indi iduals each we e ob ained om a nea by o es (lowland si e: 350 m asl) and, o he ju enile s age o he wo coni e s, om a highland si e (Walds ein, Fich elgebi ge, 760 m asl) abou 50 km no hwes o he expe imen al si e. In he second win e , ca e ully exca a ed ine oo s o wo p o enances (FR1 and IT.S) exhibi ing low and high cold ha diness in hei oliage in he i s yea we e analyzed o hei cold ha diness by applying he same p o ocol as o he needles. Mean annual minimum empe a u es o he pe iod 1950 o 2000 (mean empe a u e o he coldes day o he yea s 1950-2000) o each geog aphic o igin o he p o enances we e e ie ed om wo ldclim (Hijmans e al., 2005) and used as indica o s o minimum empe a u es (Table 1). We assume ha he ela i e di e ences be ween geog aphic o igins ha e been sui ably e lec ed, al hough hese alues exceed he absolu e minimum empe a u es due o daily a e aging ( o ou expe imen al si e he minimum empe a u e based on wo ldclim is -3.5°C while he absolu e annual minimum empe a u es be ween 1998 and 2011 a an hou ly esolu ion anged be ween -10.8 and -25.5°C). An o dina y leas squa es eg ession be ween his indica o and cold ha diness was applied o he con ol ea men in o de o de ec local adap a ion o la e os e en s. Soluble ca bohyd a es we e quan i ied in he i s win e o wo p o enances exhibi ing low and high cold ha diness, espec i ely (FR1 and IT.S). Mixed samples o one needle om he se en plan s pe expe imen al uni we e aken, immedia ely ozen in liquid ni ogen and s o ed a -30°C. F ozen ma e ial was g ound in a ball mill; soluble ca bohyd a es o 20 mg o plan ma e ial we e ex ac ed in 50 % me hanol and analyzed using he an h one me hod (Klebe e al., 1997). Ex inc ion was measu ed a 620 nm. We used known concen a ions o Glucose as a s anda d. Manusc ip 5: Cold ha diness o Pinus nig a A nold as in luenced by geog aphic o igin, wa ming, and ex eme summe d ough 131 Lipid composi ion was ob ained o he same wo p o enances (FR1 and IT.S). Epicu icula wax lipids including alkanes as he mos abundan wax componen we e eco e ed by insing needles o 60 seconds in dichlo ome hane (DCM), which esembles s anda d echniques using chlo o o m (Radle and Ho n, 1965). A e emo al o he epicu icula wax lipids, needles we e g ound o a ine powde using a ball mill (Re sch M200). In e nal waxes we e eco e ed by s anda d Soxhle ex ac ion using a mix u e o DCM/Me hanol (93:7) (Wiesenbe g e al., 2010). Ex ac s o in e nal and epicu icula waxes we e d ied and sequen ially sepa a ed using solid-phase ex ac ion in o lipid ac ions including a y acid and alkane ac ions (Wiesenbe g e al., 2010). The ollowing sec ion only discusses he esul s o he alkane ac ion as a ep esen a i e componen o epicu icula wax lipids and he a y acids o in e nal lipids as he dominan compound class o cell memb anes. O he ac ions we e also analyzed, bu no signi ican di e ences in hei dis ibu ion pa e ns we e ob ained. Aliquo s o deu e ia ed s anda ds (D39C20 acids and D50C24 alkane, espec i ely) we e added o he lipid ac ions o compound iden i ica ion and quan i ica ion. Fa y acid ac ions we e de i a ized using BSTFA (N,O- Bis ( ime hylsilyl) i luo oace amide) o 1h a 80°C, whe eas alkanes did no equi e any u he p epa a ion. All ac ions we e measu ed using gas ch oma og aphy coupled wi h lame ioniza ion de ec ion (Agilen 7890). In addi ion o lipid dis ibu ion pa e ns, molecula p oxies we e also de e mined o e alua e he di e ences be ween p o enances and clima e manipula ions. The a e age chain leng h (ACL) o lipids is in luenced by lipid biosyn hesis and egula es he wa e epellency o hyd ophobic hyd oca bon chains o a y acids in cell memb anes as well as a y acids and alkanes in epicu icula waxes (Kola ukudy e al., 1976). Ini ially, he ACL was used o di e en ia e plan and mic obial sou ces o o ganic ma e in e es ial sedimen s (B ay and E ans, 1961): ACL = Σ (zn * n) / Σ (zn) (2) whe e n is he numbe o ca bons and zn he amoun o a y acids o alkanes wi h n ca bons. Ano he pa ame e o ob ain he di e ences in he lipid biosyn hesis as a ec ed by en i onmen al s ess is he p edominance o odd e sus e en alkanes, he so-called ca bon p e e ence index (CPI: Kola ukudy e al., 1976): CPI = [(Σ n-C25-33 odd / Σ n-C24-32 e en) + (Σ n-C25-33 odd / Σ n-C26-34 e en)]/2 (3) The deg ada ion o alkanes and a less e ec i e syn hesis o he p edominan odd alkanes in waxes lead o a educ ion o he CPI unde en i onmen al s ess (Wiesenbe g e al., 2008). Manusc ip 5: Cold ha diness o Pinus nig a A nold as in luenced by geog aphic o igin, wa ming, and ex eme summe d ough 132 An analysis o a iance (ANOVA) combined wi h linea mixed e ec models we e applied o es o he main and in e ac i e e ec s o he h ee ac o s: p o enance, d ough , and wa ming on cold ha diness (LT50), soluble ca bohyd a e con en and lipid composi ion. Including he expe imen al uni as a andom ac o accoun ed o he spli -plo design (Pinhei o and Ba es, 2004). Da a we e log ans o med o imp o e he homogenei y o a iances and he no mali y o esiduals i necessa y. All s a is ical analyses we e conduc ed wi h he so wa e R 2.11.1 (R De elopmen Co e Team, 2010) and he addi ional packages “nlme” and “quan eg”. 3. Resul s 3.1. Local adap a ion in cold ha diness Mean cold ha diness di e ed be ween he p o enances by abou 10°C in bo h win e s (Figu e 3). Figu e 3: Cold ha diness (LT50) as a ec ed by geog aphic o igin (p o enances; le ) and p eceding clima ic condi ions (d ough and wa ming in in e ac ion; igh ) in he i s (uppe panel) and second (lowe panel) yea o he expe imen . ANOVA- esul s a e p o ided pe yea wi h signi ican e ec s in bold. Mean alues and s anda d e o s a e shown o 84 indi iduals pe ba o he p o enances and 168 indi iduals pe ba o he clima e ea men s. No e ha he d ough manipula ion only ook place in he i s summe o he expe imen . The wa ming ea men s opped h ee mon hs be o e sampling in he i s yea while unning h oughou sampling in he second yea . Manusc ip 5: Cold ha diness o Pinus nig a A nold as in luenced by geog aphic o igin, wa ming, and ex eme summe d ough 133 The LT50 alues anged be ween a minimum o -21.2°C o p o enance FR2 and a maximum o -32.1°C o p o enance FR1 in he i s win e ; and be ween a minimum o -23.2°C o p o enance FR2 and a maximum o -33.1°C o he p o enance om Se bia (YU) in he second win e . P o enances om colde o igins gene ally displayed supe io cold ha diness (Figu e 4). Signi ican co ela ions ( ² = 0.77 in he i s win e and ² = 0.80 in he second win e ) be ween cold ha diness and mean minimum win e empe a u e a he o igins we e ound o he au och honous p o enances in bo h yea s i he p o enance om C oa ia was excluded om he analyses. When included, no signi ican co ela ion was ound in he i s win e , while he co ela ion was weake ( ² = 0.59) while s ill emaining signi ican in he second win e . Figu e 4: Local adap a ion in cold ha diness depending on he mean minimum empe a u e a he o igin. Linea eg essions a e shown o all au och honous p o enances (DE, open iangle, no included) excluding he p o enance om C oa ia (HR, open ci cle), as i s au och honous s a us is ques ionable. Resul s o he eg ession including he p o enance om C oa ia a e gi en in pa en heses. Cold ha diness (LT50) displays he mean o he con ol ea men pe p o enance (n = 21). 3.2. Clima ic expe iences al e cold ha diness Cold ha diness was a ec ed by he clima ic expe iences o he indi iduals. The ex eme summe d ough inc eased cold ha diness by 3.9°C on a e age in he i s win e and he e was a non-signi ican end in he same di ec ion in he second win e a e he d ough manipula ion (Figu e 3). Unexpec edly, he summe wa ming om he i s yea esul ed in inc eased cold ha diness o 3.4°C on a e age while he yea - ound wa ming o he second yea esul ed in no signi ican e ec , al hough he end ollowed he same di ec ion as in he i s yea . In e es ingly, he d ough and he wa ming e ec in he i s yea we e no addi i e (ANOVA, in e ac ion be ween d ough and wa ming: F = 16.0; p = 0.004), esul ing in lowe Manusc ip 5: Cold ha diness o Pinus nig a A nold as in luenced by geog aphic o igin, wa ming, and ex eme summe d ough 134 cold ha diness in he un ea ed a ian and compa able, high cold ha diness in he o he h ee clima e manipula ions (Figu e 3). 3.3. Physiological easons o a ying cold ha diness The amoun o soluble ca bohyd a es in he needles inc eased by 25.9% in a p o enance exhibi ing high cold ha diness compa ed o a p o enance showing low cold ha diness (Table 2; ANOVA: F = 15.3; p = 0.004). The d ough manipula ion had no signi ican e ec on he ca bohyd a e concen a ion (F = 0.0; p = 0.889). Table 2: Compa ison o ca bohyd a e con en and a e age chain leng h (ACL) o a y acids o cu en yea needles be ween wo p o enances exhibi ing low (IT.S) and high (FR1) cold ha diness. Samples aken in he i s win e o he expe imen , mean ± s anda d de ia ion p o ided, n = 3. Cold ha diness o sou ce Con ol D ough Cold ha diness (LT50 in °C) high -25.8 ± 2.3 -36.7 ± 1.0 low -17.5 ± 1.6 -26.8 ± 4.5 Soluble ca bohyd a es (‰TM) high 118.4 ± 9.4 111.9 ± 9.5 low 89.0 ± 15.9 93.7 ± 3.0 ACL o epicu icula wax a y acids high 18.8 ± 0.2 17.7 ± 0.2 low 18.1 ± 0.0 17.8 ± 0.1 ACL o in e nal a y acids high 17.8 ± 0.5 17.6 ± 0.1 low 17.5 ± 0.1 17.1 ± 0.3 ACL o epicu icula wax alkanes high 27.5 ± 0.1 27.6 ± 0.0 low 27.5 ± 0.1 27.5 ± 0.1 CPI o epicu icula wax alkanes high 9.3 ± 0.2 10.4 ± 0.2 low 12.0 ± 0.6 11.9 ± 0.4 The composi ion o in e nal a y acids (ACL) as main componen s o cell memb anes did no di e signi ican ly be ween a p o enance exhibi ing high cold ha diness and a p o enance showing low cold ha diness (F = 3.7; p = 0.092). Likewise, no e ec o he d ough manipula ion was ound (F = 1.7; p = 0.222). Fo he epicu icula wax lipids, he p o enance wi h he high cold ha diness was cha ac e ized by a sligh ly highe ACL (3.9 %; F = 5.2; p = 0.051) han ha wi h he low cold ha diness. The d ough ea men led o a gene al dec ease in ACL (F = 22.0; p = 0.002), which was s onge o he plan s wi h a high (5.5 %) a he han a low cold ha diness (1.5 %; in e ac ion be ween p o enance and d ough manipula ion: F = 7.7; p = 0.024). Manusc ip 5: Cold ha diness o Pinus nig a A nold as in luenced by geog aphic o igin, wa ming, and ex eme summe d ough 141 selec ion o os - ole an p o enances could be based on he cu en clima ic condi ions wi hin he species anges. Howe e , ou inding ha clima ic expe iences wi hin he li e o single plan s al e cold ha diness indica es ha p o enance ials unde con ol condi ions may be misleading unde changing clima ic mean and ex eme condi ions. The mul i ude o possible clima ic a iables o be selec ed o and unce ain ies conce ning u u e clima es imply ha he sea ch o bes -adap ed p o enances should no be he only s a egy. In addi ion, managemen ac ions which p omo e gene ic di e si y (e.g. suppo ing na u al egene a ion and addi ion o gene ically di e se ma e ial) a e c ucial as gene ic di e si y enables o ganisms o con inue adap ing and e ol ing o new condi ions wi hin one o se e al gene a ion cycles (Hosius e al., 2006; Schabe g e al., 2008b). Fu he mo e, he ole o he bi o es and diseases unde changing clima e equi es de ailed in es iga ions. Fo ins ance, a needle bligh known as he “ ed band disease” (Do his oma sep ospo a) is epo ed o inc ease in impo ance o e ecen yea s in P. nig a (Isaje e al., 2004), a de elopmen ha may be ela ed o clima e change (Wa e al., 2011). Ul ima ely, ee species esponses should be ega ded in he con ex o popula ions unde compe i i e p essu e. The ad an age o common ga den expe imen s is ha hey can de ec he spec um o possible species-speci ic esponses. Ne e heless, he e is a need o es he ob ained esul s in communi ies whe e he compe i i e balance migh ampli y o bu e esponses. 5. Conclusions Cold ha diness o Pinus nig a oliage is highly a iable be ween p o enances and shows signs o local adap a ion o p e ailing minimum empe a u es a he o igin. Bo h se e e d ough e en s and summe wa ming can inc ease cold ha diness, indica ing ha he in e ac ion o di e en clima e pa ame e s leads o unexpec ed esul s and ha win e su i al can be al e ed by clima ic e en s du ing he g owing season. Physiologically, cold ha diness is ela ed o soluble ca bohyd a e con en and lipid composi ion. In e es ingly, a ia ion o cold ha diness o he needles wi hin he (sub-) medi e anean species P. nig a was highe han be ween his species and o he species common o he empe a e zone o Cen al Eu ope. Taken oge he , ou esul s imply ha he cold ha diness o he oliage o P. nig a is adap i e o long- e m g owing condi ions a he o igin (gene ic he i age) and o sho - e m al e a ions o hese condi ions (indi idual plas ici y), while i s hin s sugges ha cold ha diness o he Manusc ip 5: Cold ha diness o Pinus nig a A nold as in luenced by geog aphic o igin, wa ming, and ex eme summe d ough 142 oo s is high and p obably no unde selec i e p essu e cu en ly. Ou da a om mid-win e sugges s ha below- and abo e-g ound cold ha diness o selec ed p o enances appea o be well adap ed o cul i a ion in empe a e egions as an adap a ion s a egy agains he ad e se e ec s o clima e change in d y habi a s. Howe e , wi h espec o la e sp ing and ea ly au umn os e en s, he empo al pa e n o os ha diness wi h po en ial in a-speci ic di e ences should be in es iga ed in mo e de ail. Be o e ansloca ions a e ecommended, u he in es iga ions a e equi ed, e.g. explo ing he ole o bio ic in e ac ions unde changing clima ic condi ions. Gene ally, wi hin-species di e si y should be conse ed a he species le el and imp o ed in an h opogenically ounded s ands in o de o allow o adap ion o clima e change. 6. Acknowledgemen s: This s udy was unded by he Obe ankens i ung (OF a_02631) in coope a ion wi h he "Ba a ian Clima e P og amme 2020" in he join esea ch cen e “FORKAST” and he Ba a ian S a e Minis y o he En i onmen and Public Heal h (ZKL01Ab 7_18456). We hank Ch is ian Schemm, Elke König, S e an König, Ch is ine Pilsl and nume ous s uden wo ke s and in e ns o hei ou s anding help du ing he ield wo k. Re e ences: Ande son, J.A., Kenna, M.P., Talia e o, C.M., 1988. Cold ha diness o Midi on and Ti g een be mudag ass. Ho science 23, 748–750. Bakh , J., Bano, A., Dominy, P., 2006. The ole o abscisic acid and low empe a u e in chickpea (Cice a ie inum) cold ole ance. II. E ec s on plasma memb ane s uc u e and unc ion. J Exp Bo 57, 3707–3715. Big as, F.J., Ryyppo, A., Linds om, A., Sa in, E., 2001. Cold acclima ion and deacclima ion o shoo s and oo s o coni e seedlings. In: Big as, F.J., Colombo, S.J. (Eds.), Coni e cold ha diness. Kluwe , Do d ech , pp. 57–88. Blodne , C., Sk oppa, T., Johnsen, O., Polle, A., 2005. F eezing ole ance in wo No way sp uce (Picea abies [L.] Ka s .) p ogenies is physiologically co ela ed wi h d ough ole ance. J Plan Physiol 162, 549–558. Bol e, A., Czajkowski, T., Kompa, T., 2007. The no h-eas e n dis ibu ion ange o Eu opean beech - a e iew. Fo es y 80, 413–429. B ay, E., E ans, E., 1961. Dis ibu ion o n-pa a ins as a clue o he ecogni ion o sou ce beds. Geochim Cosmochim Ac a 22, 2–15. Cape, J.N., Pe cy, K.E., 1993. En i onmen al-in luences on he de elopmen o sp uce needle cu icles. New Phy ol 125, 787–799. Cla k, J.S., 2010. Indi iduals and he a ia ion needed o high species di e si y in o es ees. Science 327, 1129–1132. Eccel, E., Rea, R., Ca a a, A., C isci, A., 2009. Risk o sp ing os o apple p oduc ion unde u u e clima e scena ios: he ole o phenological acclima ion. In J Biome eo ol 53, 273–286. Ganse , D., Backes, K., Kakuba i, Y., 1999. Al i udinal and seasonal a ia ion o os esis ance o Fagus c ena a and Be ula e manii along he Paci ic slope o M . Fuji, Japan. J Ecol 87, 382–390. G o man, P.M., D iscoll, C.T., Fahey, T.J., Ha dy, J.P., Fi zhugh, R.D., Tie ney, G.L., 2001. Colde soils in a wa me wo ld: A snow manipula ion s udy in a no he n ha dwood o es ecosys em. Biogeochemis y 56, 135–150. Hijmans, R.J., Came on, S.E., Pa a, J.L., Jones, P.G., Ja is, A., 2005. Ve y high esolu ion in e pola ed clima e su aces o global land a eas. In J Clima ol 25, 1965–1978. Manusc ip 5: Cold ha diness o Pinus nig a A nold as in luenced by geog aphic o igin, wa ming, and ex eme summe d ough 143 Holliday, J.A., Ralph, S.G., Whi e, R., Bohlmann, J., Ai ken, S.N., 2008. Global moni o ing o au umn gene exp ession wi hin and among pheno ypically di e gen popula ions o Si ka sp uce (Picea si chensis). New Phy ol 178, 103–122. Hosius, B., Leinemann, L., Konne , M., Be gmann, F., 2006. Gene ic aspec s o o es y in he cen al Eu ope. Eu op J Fo es Res 125, 407–417. Hube , G., 2011. In e na ionale Baye ische Schwa zkie e n-He kun s e such 2009/2010 - Neue Tes s ü Schwa zkie e n-He kün e in Baye n im Hinblick au den Klimawandel. Fo s a chi , in p ess. Isaje , V., Fady, B., Seme ci, H., Andono ski, V., 2004. EUFORGEN Technical Guidelines o gene ic conse a ion and use o Eu opean black pine (Pinus nig a). In e na ional Plan Gene ic Resou ces Ins i u e, Rome, I aly. Jagielska, A., Cwalina, M., P us-Glowacki, W., 2007. Gene ic di e si y o he black pine‘s s ands (Pinus nig a) A n. Sylwan 151, 23–31. Jalili, A., Jamzad, Z., Thompson, K., A aghi, M.K., Ash a i, S., Hasaninejad, M., Panahi, P., Hooshang, N., Azadi, R., Ta akol, M.S., Palizda , M., Rahmanpou , A., Fa ghadan, F., Mi hossaini, S.G., Pa aneh, K., 2010. Clima e change, unp edic able cold wa es and possible b akes on plan mig a ion. Global Ecol Biogeog 19, 642–648. Kim, K.S., Pa k, S.H., Jenks, M.A., 2007. Changes in lea cu icula waxes o sesame (Sesamum indicum L.) plan s exposed o wa e de ici . J Plan Pysiol 164, 1134–1143. Klebe , H.-P., Schlee, D., Schöpp, W., 1997. Biochemisches P ak ikum. Me hoden ü S udium, P axis, Fo schung. G. Fische , Jena. Kod a, E., S einhaeuse , K., Ganguly, A., 2011. (2011), Pe sis ing cold ex emes unde 21s -cen u y wa ming scena ios. Geophys Res Le , doi:10.1029/2011GL047103. Koenke , R., 2006. Quan ile Reg ession. R package e sion 4.02. h p://www. -p ojec .o g. Kola ukudy, P., C o eau, R., Buckne , J., 1976. Biochemis y o plan waxes. In: Kola ukudy, P. (Ed.), Chemis y and Biochemis y o Na u al Waxes. Else ie , Ams e dam, pp. 290–347. Kölling, C., 2007. Klimahüllen ü 27 Waldbäume. AFZ - De Wald 23, 1242–1245. K eyling, J., 2010. Win e clima e change: a c i ical ac o o empe a e ege a ion pe o mance. Ecology 91, 1939–1948. K eyling, J., Bi ne , T., Jaeschke, A., Jen sch, A., S einbaue , M.J., Thiel, D., Beie kuhnlein, C., 2011a. Assis ed coloniza ion - a ques ion o ocal uni s and ecipien locali ies. Res o Ecol 19, 433-440. K eyling, J., Hen y, H.A.L., 2011. Vanishing win e s in Ge many: soil os dynamics and snow co e ends, and ecological implica ions. Clim Res 46, 269–276. K eyling, J., Thiel, D., Nagy, L., Jen sch, A., Hube , G., Konne , M., Beie kuhnlein, C., 2011b. La e os sensi i i y o ju enile Fagus syl a ica L. di e s be ween sou he n Ge many and Bulga ia and depends on p eceding ai empe a u e. Eu op J Fo es Res, DOI: 10.1007/s10342-011-0544-y. Ko ne , C, Paulsen, J. 2004. A wo ld-wide s udy o high al i ude eeline empe a u es. J Biogeog 31: 713–732. Kozlowski, T.T., Palla dy, S.G., 2002. Acclima ion and adap i e esponses o woody plan s o en i onmen al s esses. Bo Re 68, 270–334. Kuse , J.E., Ching, K.K., 1980. P o enance a ia ion in phenology and cold ha diness o wes e n hemlock seedlings. Fo es Sci 26, 463–470. Ma ei, M., Badino, S., Bossi, S., 2004. Chemo axonomic signi icance o lea wax n-alkanes in he Pinales (Coni e ales). Jou nal o Biological Resea ch 1, 3–19. Mancuso, S., 2000. Elec ical esis ance changes du ing exposu e o low empe a u e measu e chilling and eezing ole ance in oli e ee (Olea eu opaea L.) plan s. Plan Cell En i on 23, 291–299. McKenney, D., Pedla , J., O‘Neill, G., 2009. Clima e change and o es seed zones: Pas ends, u u e p ospec s and challenges o ponde . Fo es Ch on 85, 258–266. Moelle ing, E.R., Mu han, B., Benning, C., 2010. F eezing ole ance in plan s equi es lipid emodeling a he ou e chlo oplas memb ane. Science 330, 226–228. Mo in, X., Ameglio, T., Ahas, R., Ku z-Besson, C., Lan a, V., Lebou geois, F., Miglie a, F., Chuine, I., 2007. Va ia ion in cold ha diness and ca bohyd a e concen a ion om do mancy induc ion o bud bu s among p o enances o h ee Eu opean oak species. T ee Physiol 27, 817–825. Nkongolo, K.K., Michael, P., G a on, W.W., 2002. Iden i ica ion and cha ac e iza ion o RAPD ma ke s in e ing gene ic ela ionships among Pine species. Genome 45, 51–58. Pa mesan, C., Yohe, G., 2003. A globally cohe en inge p in o clima e change impac s ac oss na u al sys ems. Na u e 421, 37–42. Pinhei o, J.C., Ba es, D.M., 2004. Mixed-e ec s models in S and S-PLUS. Sp inge , New Yo k. P io , S.A., P i cha d, S.G., Runion, G.B., Roge s, H.H., Mi chell, R.J., 1997. In luence o a mosphe ic CO2 en ichmen , soil N, and wa e s ess on needle su ace wax o ma ion in Pinus palus is (Pinaceae). Am J Bo 84, 1070–1077. Manusc ip 5: Cold ha diness o Pinus nig a A nold as in luenced by geog aphic o igin, wa ming, and ex eme summe d ough 144 R De elopmen Co e Team, 2010. [R: A Language and En i onmen o S a is ical Compu ing. R e sion 2.11.1. R Founda ion o S a is ical Compu ing. ISBN 3-900051-07-0, URL h p://www.R-p ojec .o g, Vienna, Aus ia. Radle , F., Ho n, D.H., 1965. Composi ion o g ape cu icle wax. Aus J Chem 18, 1059-&. Repo, T., Mononen, K., Al ila, L., Pakkanen, T.T., Hanninen, H., 2008. Cold acclima ion o peduncula e oak (Que cus obu L.) a i s no he nmos dis ibu ion ange. En i on Exp Bo 63, 59–70. Saenz-Rome o, C., Tapia-Oli a es, B.L., 2008. Gene ic a ia ion in os damage and seed zone delinea ion wi hin an al i udinal ansec o Pinus de oniana (P. michoacana) in Mexico. Sil a Gene 57, 165–170. Sakai, A., Weise , C.J., 1973. F eezing esis ance o ees in No h Ame ica wi h e e ence o ee egions. Ecology 54, 118–126. Schabe g, P.G., Hennon, P.E., D'Amo e, D.V., Hawley, G.J., 2008a. In luence o simula ed snow co e on he cold ole ance and eezing inju y o yellow-ceda seedlings. Glob Change Biol 14, 1282–1293. Schabe g, P.G., DeHayes, D.H., Hawley, G.J., Nijensohn, S.E., 2008b. An h opogenic al e a ions o gene ic di e si y wi hin ee popula ions: Implica ions o o es ecosys em esilience. Fo es Ecol Manag 256, 855– 862. Schä , C, Vidale, PL, Lu hi, D, F ei, C, Habe li, C, Linige , MA, Appenzelle , C. 2004. The ole o inc easing empe a u e a iabili y in Eu opean summe hea wa es. Na u e 427: 332–336. Shaw, M.R., Za ale a, E.S., Chia iello, N.R., Cleland, E.E., Mooney, H.A., Field, C.B., 2002. G assland esponses o global en i onmen al changes supp essed by ele a ed CO2. Science 298, 1987–1990. Shephe d, T., G i i hs, D.W., 2006. The e ec s o s ess on plan cu icula waxes. New Phy ol 171, 469–499. S imbeck, G.R., Kjellsen, T.D., Schabe g, P.G., Mu akami, P.F., 2007. Cold in he common ga den: compa a i e low- empe a u e ole ance o bo eal and empe a e coni e oliage. T ees-S uc Func 21, 557– 567. Su inen, M.L., Pal a, J.P., Reich, P.B., 1992. Seasonal di e ences in eezing s ess esis ance o needles o Pinus nig a and Pinus esinosa - e alua ion o he elec oly e leakage me hod. T ee Physiol 11, 241–254. Thomas, C.D., Came on, A., G een, R.E., Bakkenes, M., Beaumon , L.J., Collingham, Y.C., E asmus, B. F. N., de Siquei a, M. F., G ainge , A., Hannah, L., Hughes, L., Hun ley, B., an Jaa s eld, A. S., Midgley, G.F., Miles, L., O ega-Hue a, M.A., Pe e son, A.T., Phillips, O.L., Williams, S.E., 2004. Ex inc ion isk om clima e change. Na u e 427, 145–148. Thomashow, M.F., 1999. Plan cold acclima ion: F eezing ole ance genes and egula o y mechanisms. Ann Re Plan Physiol Plan Molec Biol 50, 571–599. Va elides, C., B o as, G., Va elides, Y., 2001. P o enance a ia ion in Pinus nig a a h ee si es in No he n G eece. Ann Fo es Sci 58, 893–900. Va us, S., Walsh, J.E., Chapman, W.L., Po is, D., 2006. The beha io o ex eme cold ai ou b eaks unde g eenhouse wa ming. In J Clima ol 26, 1133–1147. Visnjic, V.C., Doh enbusch, A., 2004. F os esis ance and phenology o Eu opean beech p o enances (Fagus syl a ica L. ). Allg Fo s Jgdz g 175, 101–108. Wa , M.S., Ganley, R.J., K i icos, D.J., Manning, L.K., 2011. Do his oma needle bligh and pi ch canke : he cu en and u u e po en ial dis ibu ion o wo impo an diseases o Pinus species. Can J Fo es Res 41, 412–424. Wiesenbe g, G.L.B., Schmid , M.W.I., Schwa k, L., 2008. Plan and soil lipid modi ica ions unde ele a ed a mosphe ic CO2 condi ions: I. Lipid dis ibu ion pa e ns. O g Geochem 39, 91–102. Wiesenbe g, G.L.B., Gocke, M., Kuzyako , Y., 2010. Fas inco po a ion o oo -de i ed lipids and a y acids in o soil - E idence om a sho e m mul iple (CO2)-C-14 pulse labelling expe imen . O g Geochem 41, 1049–1055. Zobel, B.J., Wyk, G. ., S ahl, P., 1987. G owing Exo ic Fo es s. Wiley, New Yo k, N.Y. Manusc ip 6: Inc eased ain all a iabili y educes biomass and o age quali y o empe a e g assland la gely independen o mowing equency 145 Manusc ip 6: Inc eased ain all a iabili y educes biomass and o age quali y o empe a e g assland la gely independen o mowing equency Ag icul u e, Ecosys ems and En i onmen , 2012, 148: 1-10. Julia Wal e a*, Ke s in G an b∗, Ca l Beie kuhnleinc, Jü gen K eylingc, Michael Webe d, Anke Jen schb aConse a ion Biology, Helmhol z Cen e o En i onmen al Resea ch- UFZ, Pe mose s aße 15, 04318 Leipzig, Ge many, Julia.wal e @u z.de, phone:+49-341-2351654, ax: :+49-341-2351470, co esponding au ho bDis u bance Ecology, Bay eu h Uni e si y, 95440 Bay eu h, Ge many cDepa men o Biogeog aphy, Bay eu h Uni e si y, 95440 Bay eu h, Ge many dDepa men o Plan Physiology, Bay eu h Uni e si y, 95440 Bay eu h, Ge many Highligh s 1. G assland was subjec ed o inc eased ain all a iabili y and mowing equency. 2. Inc eased ain all a iabili y educes g assland p oduc i i y and o age quali y. 3. Mo e equen mowing ini ially inc eases and la e on dec eases p oduc i i y. 4. Mowing egime does mos ly no in e ac wi h ain all a iabili y manipula ions. 5. Su icien o e all ain all amoun is impo an o g assland esilience. Abs ac Clima e models indica e ha global wa ming will s imula e a mosphe ic exchange p ocesses and inc ease ain all a iabili y, leading o longe d y pe iods and mo e in ense ain all e en s. Recen s udies sugges ha bo h he magni ude o he ain all e en s and hei equency may be as impo an o empe a e g assland p oduc i i y as he annual sum. Howe e , un il now in e ac i e e ec s be ween land managemen p ac ice, such as mowing equency, and ain all a iabili y on p oduc i i y and o age quali y ha e no been s udied in de ail. He e, we p esen he da a om a ield expe imen (EVENT II) in which a Cen al- Eu opean g assland was subjec ed o inc eased sp ing ain all a iabili y (low, in e media e and ex eme ain all a iabili y wi hou any change o he ain all amoun ) and inc eased mowing equency ( ou imes compa ed o wice a yea ). We assessed biomass p oduc ion, ∗ au ho s con ibu ed equally o he publica ion Manusc ip 6: Inc eased ain all a iabili y educes biomass and o age quali y o empe a e g assland la gely independen o mowing equency 146 o age quali y pa ame e s, oo -leng h and shoo - oo a io. Enhanced sp ing ain all a iabili y educed midsumme p oduc i i y and he lea N and p o ein concen a ions o a a ge species, bu did no exe any long- e m e ec s on biomass p oduc ion and o age quali y in la e summe . Howe e , he inc eased sp ing ain all a iabili y educed abo eg ound ne p ima y p oduc i i y by 15 %. Mo e equen mowing inc eased p oduc i i y in he i s yea o he s udy, bu dec eased p oduc i i y a he end o he second yea , showing a decline in he po en ial o o e compensa ion a e a his o y o mo e in ense mowing. Gene ally, mo e equen mowing dec eased he shoo - oo a io and inc eased he concen a ion o lea N. Inc eased mowing equency nei he bu e ed, no ampli ied he ad e se e ec s o ain all a iabili y on p oduc i i y, bu made lea N concen a ions in ea ly summe mo e esponsi e o al e ed ain all pa e ns. These esul s highligh he ac ha e en ela i ely small and sho - e m al e a ions o ain all dis ibu ion can educe p oduc ion and o age quali y, wi h li le bu e ing capaci y o al e ed mowing equency. Compa isons wi h p oduc i i y da a om he i s yea o he s udy, in which bo h, ain all dis ibu ion and ain all amoun we e modi ied, demons a e he c ucial ole o su icien mois u e (annual ain all amoun ) o g assland esilience: In his i s yea , nega i e e ec s o ex eme ain all a iabili y las ed un il he end o he yea . To conclude, inc eased ain all a iabili y unde clima e change will likely a ec ag icul u al yield in empe a e meadows. Managemen s a egies o bu e hese e ec s ha e ye o be de eloped. Keywo ds: EVENT II expe imen , ex eme wea he e en , ainou -shel e , o age quali y, Alopecu us p a ensis, T i olium p a ense 1. In oduc ion Clima e change is p ojec ed o modi y no only annual p ecipi a ion sum, bu also o esul in mo e ex eme ain all egimes in many pa s o he wo ld (IPCC 2007; Jen sch and Beie kuhnlein, 2008). This will cause mo e se e e d ough pe iods as well as an inc ease in he equency and magni ude o ex eme p ecipi a ion e en s (T enbe h e al., 2003, Min e al., 2011). E idence is moun ing ha he equency and se e i y o d ough s and ex eme p ecipi a ion e en s has al eady inc eased o e ecen decades in many egions (Blenkinsop and Fowle , 2007; Haylock and Goodess, 2004; IPCC 2007). P ima y p oduc i i y and ecosys em unc ioning in e es ial ecosys ems a e s ongly in luenced by he annual amoun o p ecipi a ion (Sala e al., 1988). Howe e , ecen esea ch sugges s ha ain all a iabili y may exe an e en s onge in luence on ecosys em Manusc ip 6: Inc eased ain all a iabili y educes biomass and o age quali y o empe a e g assland la gely independen o mowing equency 147 unc ioning, whe e especially empe a e g assland sys ems seem o be esponsi e o changes in ain all a iabili y. In g assland, mo e ex eme ain all egimes (less, bu mo e in ense ain all e en s) a ec ANPP (abo eg ound ne p ima y p oduc i i y) (Ba e e al., 2002; Fay, 2009; Heisle -Whi e e al., 2009; Knapp e al., 2002), ca bon cycling (Chou e al., 2008; Fay, 2009; Ha pe e al., 2005) and N mine aliza ion (Ba e e al., 2002, Heisle -Whi e e al., 2009). The la e may in u n a ec lea quali y in e ms o N o p o ein con en . La ge educ ions in ANPP ha e been shown in mesic g assland in esponse o mo e ex eme ain all pa e ns (Fay e al., 2003; Heisle -Whi e e al., 2009; Knapp e al., 2008). In addi ion o he ain all amoun and a iabili y, land managemen s a egies, such as mowing equency, can a ec p oduc i i y and lea li e quali y in managed g assland. Mo e equen cu ing is known o inc ease lea N con en . Howe e , whe he o no mowing inc eases o dec eases he p oduc i i y o g assland depends on he mowing in ensi y, e.g. mowing his o y, mowing equency and cu ing heigh (G een and De ling, 2000; McNaugh on, 1979; Tu ne e al., 1993; Weigel e al., 2009). Mowing o de olia ion is likely o al e he esponse o ain all a iabili y by al e ing plan communi y composi ion (Swemme and Knapp, 2008). Fu he mo e, a educ ion o anspi a i e issue al e s wa e up ake and consump ion and he e o e eac ion owa ds ain all (Hei schmid e al., 1999; McNaugh on, 1979; Yang and Midmo e, 2004). Cu en ly, a knowledge gap exis s on how land managemen p ac ices, such as mowing equency, a e in e ac ing wi h mo e ex eme ain all egimes: Inc eased mowing equency migh bu e he e ec s o ain all a iabili y on g assland, diminishing he ampli ude o he esponse owa ds ain all ex emes (Swemme and Knapp, 2008). A s udy by Be nha d -Röme mann e al. (2011) indica es ha clima e pa ame e s ge less impo an o biomass p oduc ion unde in e media e mowing equencies. Howe e , land managemen s a egies migh also ampli y he e ec s o ain all a iabili y. To ou knowledge, his is he i s s udy o expe imen ally manipula e mowing and ain all pa e ns in Eu opean managed g assland (meadows) in o de o iden i y any po en ial in e ac ions be ween ain all a iabili y and mowing equency. The p ima y objec i es o ou s udy we e (1) o in es iga e he ac o ially-combined e ec s o inc eased sp ing ain all a iabili y and inc eased mowing equency on he p oduc i i y and he o age quali y o semi-na u al, Cen al-Eu opean empe a e g assland and (2) o de e mine, whe he mowing equency ampli ies o bu e s he e ec s o ain all a iabili y on biomass p oduc ion and lea quali y o a a ge species. We conduc ed a ield expe imen in which we al e ed he empo al dis ibu ion and he magni ude o he ain all e en s, bu no he o e all ain all sum. To assess po en ial in e ac ions be ween ain all Manusc ip 6: Inc eased ain all a iabili y educes biomass and o age quali y o empe a e g assland la gely independen o mowing equency 148 a iabili y and mowing equency, we c ossed he ac o ain all a iabili y wi h he ac o mowing equency ( wo o ou imes pe yea ). In he p e ious yea , we al e ed he o al ain all amoun s along wi h he al e a ions in ain all a iabili y. This enables a compa ison be ween he e ec s o he al e ed o al ain all amoun s and dis ibu ion and he e ec s o al e ed ain all a iabili y unde cons an o al ain all amoun s. We hypo hesized ha (i) inc eased ain all a iabili y nega i ely a ec s p oduc i i y and lea quali y, as has been shown o o he mesic g asslands, (ii) inc eased ain all a iabili y alone can cause changes in p oduc i i y ha a e compa able o changes caused by al e a ions in bo h, a iabili y and he annual sum o ain all oge he , (iii) mo e equen mowing inc eases p oduc i i y and o age quali y, as has been shown o mo e equen , bu s ill mode a e mowing equencies, (i ) mo e equen mowing bu e s ad e se e ec s o inc eased ain all a iabili y on p oduc i i y and lea quali y, as g ow h esponses migh be synch onized and less esponsi e o ain all changes a e mowing. 2. Ma e ial & Me hods 2.1 S udy si e The s udy was conduc ed wi hin he EVENT II expe imen in a semi-na u al g assland in he Ecological Bo anical Ga den o he Uni e si y o Bay eu h, Ge many, Cen al Eu ope (49°55´19´´N, 11°3455``E, 365 m asl) (Jen sch & Beie kuhnlein, 2010). Communi ies a e domina ed by all g asses, especially Alopecu us p a ensis L. (meadow ox ail). The egional clima e is empe a e and mode a ely con inen al, wi h a mean annual empe a u e o 8.2 °C (1971–2000), and daily means anging be ween -19.6 and 27.6. The mean annual p ecipi a ion o 724 mm (1971–2000) has a bimodal dis ibu ion wi h a majo peak in June/July and a second peak in Decembe /Janua y (da a: Ge manWea he Se ice). The expe imen was ins alled on a semi-na u al, es ablished meadow. Fo mo e han 20 yea s p io o he expe imen , he meadow was mown wice pe yea and no e ilized. The ec angula ly shaped expe imen al a ea has a o al heigh di e ence o 95 cm wi hin he diagonal om sou hwes o no h eas , and abou 7 cm om sou heas o no h wes . The soil o he expe imen is classi ied as S agnosol wi h a sandy-loamy Ap-ho izon o abou 30 cm dep h, a s ongly loamy Sw-ho izon (20 cm) and a sandy-clayey Sd-ho izon (>40 cm). Plan oo s mainly occu in he uppe 15 cm, wi h almos no oo s pene a ing below he A-ho izon, mean pH- alue is 5.9. Manusc ip 6: Inc eased ain all a iabili y educes biomass and o age quali y o empe a e g assland la gely independen o mowing equency 149 2.2 Expe imen al Design The EVENT II expe imen was es ablished in 2008. The expe imen al design consis s o wo ac o ially-c ossed ac o s: (1) manipula ion o he empo al dis ibu ion and magni ude o ain all e en s in he g owing season and (2) manipula ion o mowing equency. We implemen ed h ee scena ios o ain all a iabili y ea men s in 2008 and 2009, assigned o he same plo s: (1) low ain all a iabili y wi h weekly i iga ion, ensu ing a con inuous wa e supply, (2) in e media e ain all a iabili y, wi h na u al ambien ain all a iabili y and (3) ex eme ain all a iabili y, including an ex eme sp ing d ough . Table 1 A e age soil mois u e [ ol %], a ia ion coe icien (CV) o soil mois u e [%], numbe o ain all e en s exceeding 1 mm, he sum o he ain all amoun [mm] and he a ia ion coe icien (CV) o daily ain all amoun [%] in 2008 and 2009. pa ame e yea ime span low mid ex eme na u al a e age soil mois u e 2008 26/05-30/10 2008 (158 days) 29 21 19 2009 01/04-31/10 2009 (214 days) 30 29 25 2009 01/04-17/05 2009 (47 days) 42 40 36 2009 -29/06 2009 (43 days) 29 29 21 2009 -09/08 2009 (43 days) 32 31 27 2009 -28/10 2009 (80 days) 23 24 22 CV soil mois u e 2008 26/05-30/10 2008 (158 days) 20 31 38 2009 01/04-31/10 2009 (214 days) 35 35 37 2009 01/04-17/05 2009 (47 days) 6 8 13 2009 -29/06 2009 (43 days) 21 24 30 2009 -09/08 2009 (43 days) 25 29 26 2009 -28/10 2009 (80 days) 35 36 35 no. o e en s 2008 26/05-30/10 2008 (158 days) 61 53 45 2009 01/04-31/10 2009 (214 days) 80 71 60 2009 01/04-17/05 2009 (47 days) 19 16 16 2009 -29/06 2009 (43 days) 17 13 1 2009 -09/08 2009 (43 days) 17 16 17 2009 -28/10 2009 (80 days) 27 25 25 p ecipi a ion sum 2008 26/05-30/10 2008 (158 days) 445.2 334.8 296.1 334.8 2009 01/04-31/10 2009 (214 days) 596.8 596.8 596.8 458.5 2009 01/04-17/05 2009 (47 days) 130.2 99.5 99.5 99.5 2009 -29/06 2009 (43 days) 102.5 108.1 36.6 77.4 2009 -09/08 2009 (43 days) 164.2 152.2 223.7 127.1 2009 -28/10 2009 (80 days) 199.9 205 205 154.5 CV p ecipi a ion 2008 26/05-30/10 2008 (158 days) 164 183 204 183 2009 01/04-31/10 2009 (214 days) 204 256 297 227 2009 01/04-17/05 2009 (47 days) 227 280 280 280 2009 -29/06 2009 (43 days) 156 248 656 183 2009 -09/08 2009 (43 days) 192 205 223 220 -28/10 2009 (80 days) 203 275 275 192 1 Highes alues in each ca ego y a e in bold. 2 Values o he ege a ion pe iod 2008 and 2009 and o he ime spans be ween he compensa ion i iga ion ea men s in 2009 o he di e ing ain all a iabili y ea men s a e gi en. Values o 2009 shown o e one ime span begin wi h a compensa ion i iga ion and exclude he ollowing compensa ion i iga ion, as he la e is only e ec i e o soil mois u e and biomass o he ollowing pe iod. Manusc ip 6: Inc eased ain all a iabili y educes biomass and o age quali y o empe a e g assland la gely independen o mowing equency 150 In 2008, he i s yea o he s udy, o al g owing season amoun o ain all and a iabili y o ain all we e al e ed. This made i possible o assess di ec d ough e ec s, as he ex eme ain all a iabili y ea men also ecei ed leas o al ain all (see Table 1 o an o e iew o e soil mois u e and ain all pa ame e s in bo h yea s). In 2009, he main yea o he s udy, we con olled he amoun o ain all o e he g owing season o all ea men s and manipula ed only he dis ibu ion o ain all, in o de o isola e he e ec o ain all a iabili y. All ain all a iabili y ea men s we e adjus ed o he o al 597 mm o ain all o he low a iabili y ea men in ou compensa ion i iga ions (Table 2). Thus, no only he leng h o he d y in e als, bu also he magni ude o ain all pe e en was changed. The low ain all a iabili y ea men ecei ed a leas he 30-yea weekly a e age ain all each week. The ege a ion pe iods om 1971 o 2000 se ed as a e e ence (da a: Ge man Wea he Se ice). Missing amoun s on na u al ain all we e added i he weekly ain all was less han he long- e m a e age o he same week. This ea men ensu ed con inuous wa e a ailabili y. I weekly ain all exceeded he long- e m sum, i was no sub ac ed o he nex i iga ion. Fo 2008, he o e all ain all amoun o 553 mm on he low ain all a iabili y ea men (na u al plus i iga ed ain all) wi hin he ege a ion pe iod (Ap il 1s -Oc obe 30 h) exceeded he 30-yea -a e age by 94 mm. In 2009, he o al amoun o 597 mm, i iga ed on all ea men s by applying compensa ion i iga ions (see below), exceeded he 30-yea -a e age by 138 mm. Bo h yea s consequen ly esemble a he we yea s. The in e media e ain all a iabili y ea men emained unde ambien condi ions wi hou any ea men , excep o he compensa ion i iga ions applied in 2009 which adjus ed ain all sum o he low ain all a iabili y ea men a ou poin s o ime (Table 2). Thus, in 2009 he in e media e ain all a iabili y ea men ecei ed he ambien ain all plus he compensa ion i iga ions. Table 2 Amoun and iming o compensa ion i iga ion [mm] in 2009 on he ex eme and in e media e ain all a iabili y ea men s gi en o apply he same o e all ain all amoun on all ea men s o e he ege a ion pe iod. T ea men da e ex eme mid May 18 h 36.6 36.6 June 30 h 96.5* 25.1 Augus 10 h 60.3 60.3 Oc obe 28 h 32 32 *applied on wo consecu i e days