Variability in the water footprint of arable crop production across European regions

wa e A icle Va iabili y in he Wa e Foo p in o A able C op P oduc ion ac oss Eu opean Regions Anne Gobin 1,*, Ku Ch is ian Ke sebaum 2, Jose Ei zinge 3, Mi osla T nka 4,5, Pe Hla inka 4,5, Joze Takáˇc 6, Joop K oes 7, Domenico Ven ella 8, Anna Dalla Ma a 9, Johannes Deels a 10, B anisla a Lali´c 11, Pa ol Nejedlik 12, Simone O landini 9, Pi jo Pel onen-Sainio 13, A i Rajala 13, T iin Saue 14, Le en ¸Saylan 15, Ruzica S iˇce ic 16, Višnja Vuˇce i´c 17 and Ch is os Zoumides 18 1Flemish Ins i u e o Technological Resea ch (VITO), 2400 Mol, Belgium 2Leibniz Cen e o Ag icul u al Landscape Resea ch (ZALF), 15374 Münchebe g, Ge many; [email p o ec ed] 3Ins i u e o Me eo ology, Uni e si y o Na u al Resou ces and Li e Sciences, 1180 Vienna, Aus ia; [email p o ec ed] 4Global Change Resea ch Ins i u e, The Czech Academy o Sciences, 61300 B no, Czech Republic; [email p o ec ed] (M.T.); [email p o ec ed] (P.H.) 5Depa men o Ag osys ems and Bioclima ology, Mendel Uni e si y in B no, 61300 B no, Czech Republic 6Na ional Ag icul u al and Food Cen e, Soil Science and Conse a ion Resea ch Ins i u e, B a isla a 82713, Slo akia; [email p o ec ed] 7 Wageningen En i onmen al Resea ch (Al e a), 6700 AA Wageningen, The Ne he lands; joop.k oes@wu .nl 8Consiglio pe la Rice ca in Ag icol u a e L’analisi Dell’economia Ag a ia, Uni à di Rice ca pe i Sis emi Col u ali degli Ambien i Caldo-A idi, 70125 Ba i, I aly; domenico. en ella@c ea.go .i 9Depa men o Ag i ood P oduc ion and En i onmen al Sciences (DISPAA), Uni e si y o Flo ence, 50155 Flo ence, I aly; [email p o ec ed] (A.D.M.); [email p o ec ed] (S.O.) 10 No wegian Ins i u e o Bioeconomy Resea ch (NIBIO), 1431 Ås, No way; [email p o ec ed] 11 Facul y o Ag icul u e, Uni e si y o No i Sad, No i Sad 21000, Se bia; [email p o ec ed] 12 Ea h Science Ins i u e o Slo ak Academy o Science, B a isla a 84005, Slo akia; [email p o ec ed] 13 Na u al Resou ces Ins i u e Finland (Luke), 31600 Jokioinen, Finland; [email p o ec ed] (P.P.-S.); [email p o ec ed] (A.R.) 14 Es onian C op Resea ch Ins i u e, Tallinn Technical Uni e si y, Jõge a 48309, Es onia; [email p o ec ed] 15 Depa men o Me eo ology, Facul y o Ae onau ics and As onau ics, Is anbul Technical Uni e si y, 34469 Is anbul, Tu key; [email p o ec ed] 16 Facul y o Ag icul u e, Uni e si y o Belg ade, Zemun-Belg ade 11080, Se bia; s [email p o ec ed] 17 Me eo ological and Hyd ological Se ice, Zag eb 10000, C oa ia; [email p o ec ed] 18 Ene gy, En i onmen & Wa e Resea ch Cen e , The Cyp us Ins i u e, Nicosia 2121, Cyp us; [email p o ec ed] *Co espondence: [email p o ec ed]; Tel.: +32-14-336775 Academic Edi o : Ashok K. Chapagain Recei ed: 24 Oc obe 2016; Accep ed: 31 Janua y 2017; Published: 8 Feb ua y 2017 Abs ac : C op g ow h and yield a e a ec ed by wa e use du ing he season: he g een wa e oo p in (WF) accoun s o ain wa e , he blue WF o i iga ion and he g ey WF o dilu ing ag i-chemicals. We calib a ed c op yield o FAO’s wa e balance model “Aquac op” a ield le el. We collec ed wea he , soil and c op inpu s o 45 loca ions o he pe iod 1992–2012. Calib a ed model uns we e conduc ed o whea , ba ley, g ain maize, oilseed ape, po a o and suga bee . The WF o ce eals could be up o 20 imes la ge han he WF o ube and oo c ops; he la ges sha e was a ibu ed o he g een WF. The g een and blue WF compa ed a ou ably wi h global benchma k alues (R 2 = 0.64–0.80; d = 0.91–0.95). The a iabili y in he WF o a able c ops ac oss di e en egions in Eu ope is mainly due o a iabili y in c op yield ( c = 45%) and o a lesse ex en o a iabili y in c op wa e use ( c = 21%). The WF a iabili y be ween coun ies ( c = 14%) is lowe han he a iabili y be ween seasons ( c = 22%) and be ween c ops ( c = 46%). Though modelled yields Wa e 2017,9, 93; doi:10.3390/w9020093 www.mdpi.com/jou nal/wa e Wa e 2017,9, 93 2 o 22 inc eased up o 50% unde sp inkle i iga ion, he wa e oo p in s ill inc eased be ween 1% and 25%. Con on ed wi h d ainage and uno , he g ey WF ended o o e es ima e he con ibu ion o ni ogen o he su ace and g oundwa e . The esul s showed ha he wa e oo p in p o ides a measu able indica o ha may suppo Eu opean wa e go e nance. Keywo ds: wa e oo p in ; a able c ops; ce eals; Eu ope; c op wa e use; yield 1. In oduc ion The wa e oo p in (WF) concep has c ea ed awa eness o sus ainable wa e use ollowing a global assessmen o na ional p oduc ion, consump ion and in e na ional ade [ 1 ]. T adi ional wa e consump ion s a is ics ha e been gi en o di e en sec o s, such as domes ic, ag icul u al and indus ial wa e use, bu hese show li le abou how much wa e is ac ually used. The wa e oo p in p o ides a way o compa e wa e use o egions, sec o s, commodi ies and na ions. Leading wo k in unde s anding wa e a ailabili y and isk has come om he ood indus ies h ough he analysis o wa e quan i ies ha companies use h oughou hei supply chain. Wi h wa e being inhe en ly local, he wa e oo p in calcula ions highligh he isks o local exploi a ions ha could po en ially dis up bo h business ope a ions and he su ounding communi y. Wa e is a p ecious commodi y, ce ainly in d ough -p one egions and a imes o d ough in any pa o he wo ld. The economic cos o d ough has been eno mous. In 2003, combined d ough and hea wa es led o 30% educ ion in p ima y p oduc i i y [ 2 ], and an es ima ed 13 billion € loss in Eu opean ag icul u al p oduc ion [ 3 ]. Wi h wa e sho ages al eady h ea ening g ow h, he u u e o Eu ope’s ag icul u e will be ied closely o wa e a ailabili y. In addi ion clima e models p ojec ha sou he n Eu ope will ace inc eased d ough and cen al Eu ope p olonged d y spells [ 4 , 5 ] equen ly combined wi h hea wa es [ 6 ]. The ising popula ion, coupled wi h inc easing demands by he ag icul u e and ene gy indus ies p esen s an in e dependen ela ionship o en e e ed o as he wa e – ood–ene gy nexus; he demand o wa e will likely ou weigh supply by 2050 unless changes in ood and ene gy p e e ences a e implemen ed [ 7 ]. While access o wa e has been ecognized as a basic human igh , he inc easingly high demand o wa e esou ces should be alued acco ding o i s supply. The WF is closely linked o he concep o i ual wa e , which is he olume needed o p oduce a commodi y o se ice. Impo ing i ual wa e can be pe cei ed as a pa ial solu ion o p oblems o wa e sca ci y, pa icula ly in d y egions [ 8 ]. Na ional, egional and global wa e and ood secu i y can be imp o ed when wa e -in ensi e commodi ies a e aded om places whe e hey a e economically iable o places hey a e no . Food impo o e s an al e na i e o educe p essu e on domes ic wa e esou ces and enables mo e p oduc i e wa e use as exp essed by he WF o ood [ 9 ]. O he esea ch has aken a li e cycle assessmen (LCA) app oach o e alua e he wa e oo p in o p oduc s, p ocesses and o ganisa ions as ini ia ed by [ 10 ]. Subsequen ly, an ISO 14046 s anda d was se o speci y he p inciples, equi emen s and guidelines [ 11 ]. The ISO s anda d may in oduce complexi y by c ea ing wa e oo p in s o each en i onmen al impac , e.g., o wa e a ailabili y, sca ci y, eu ophica ion and eco- oxici y, ac oss he li e cycle o a p oduc which is beyond he c op wa e oo p in ha his esea ch ocuses on. The WF o c ops o ms he basis o WF es ima ions o c op p oduc s and de i ed commodi ies [ 12 ]. In e ms o wa e olumes used, he c op WF es ima ions conside h ee majo sou ces o wa e , i.e., wa e om ain (g een WF), i iga ion (blue WF) and wa e o dilu ing chemicals (g ey WF) [ 13 ]. In a compa ison o di e en i iga ion and wa e conse a ion me hods o ou loca ions [14], i was concluded ha a combina ion o d ip i iga ion and syn he ic mulching allowed o he la ges educ ion in he WF o maize, po a o and oma o. The in e -annual a iabili y o he c op WF highligh ed in e alia he impo ance o inc eased yields o 22 c ops o he pe iod 1978–2008 in Wa e 2017,9, 93 3 o 22 China [ 15 ]. Unde s anding he a iabili y is a p e equisi e o making p ojec ions o good wa e go e nance unde di e en scena ios o global change. Ou s udy con ibu es o unde s anding he a iabili y o he WF ac oss egions, soils and annual wea he condi ions in Eu ope. We hypo hesize ha he a iabili y in he wa e oo p in o a able c ops ac oss di e en egions in Eu ope is mainly due o a iabili y in c op yield and o a lesse ex en o a iabili y in c op wa e use. The e o e, he objec i es o his s udy we e o quan i y he a iabili y in wa e used o g ow a able c ops ac oss di e en egions in Eu ope; o es ima e hei yield a iabili y; o es ablish he a iabili y in he WF o hese di e en c ops; and, o compa e he esul s wi h benchma k alues om global model es ima es as in [ 16 ]. Unde s anding he sou ces o a iabili y in he WF is impo an o elucida e wa e consump ion pa e ns in ela ion o c op p oduc ion, which in u n enables mo e e icien wa e managemen and ag icul u al wa e go e nance wi hin he amewo k o a wa e – ood–ene gy nexus. 2. Ma e ials and Me hods 2.1. Da a We collec ed empe a u e, ain all, wind speed, sola adia ion and ela i e humidi y da a om 41 me eo ological s a ions ac oss di e en egions in Eu ope o he pe iod 1992–2012 (Table 1; Figu e 1). Re e ence e apo anspi a ion was calcula ed using he modi ied Penman-Mon ei h app oach [ 17 ]. The clima ological diag ams o empe a u e, p ecipi a ion and e apo anspi a ion o hese loca ions demons a e a wide a ia ion in wea he condi ions (Figu e 2) and soils (Appendix A). The dominan soil ype(s) o 45 loca ions we e desc ibed in e ms o ex u e; chemical composi ion; olume ic wa e con en a sa u a ion, ield capaci y and wil ing poin o di e en soil ho izons up o 1.5 m o o an impe ious laye . Wi h he excep ion o polde egions, g oundwa e was absen and wa e leaching om he oo zone was discha ged as d ainage. In each loca ion majo a able c ops we e selec ed o calcula ing he wa e oo p in (Table 1). Table 1. Me eo ological s a ions and c ops pe egion (Loca ion see Figu e 1). Coun y Region Me eo S a ions 1Majo C ops 2 AT Ma ch eld G oss Enze sdo ,Fuchsenbigl WHB, BAR, MAZ, SBT BE Flande s Koksijde,Gen , Ukkel, Pee WHB, BAR, MAZ, SBT, POT CY Coun y Nicossia, Pa os, La naca WHD, POT, BAR, MAZ CZ Eas e n Czech Domaninek,Lednice,Ve o any WHB, BAR, MAZ, RAP DE-1 Mä k. Ode land Munchebe g, Manschnow WHB, BAR, SBT, RAP, POT, MAZ DE-2 No h-Eas Lowe Saxony B aunschweig WHB, BAR, SBT EE Coun y Kuusiku, Ta u, Tallinn, Võ u, Pä nu, Väike-Maa ja, Ku essaa e WHB, BAR, POT, RAP FI-1 Häme Jokioinen BAR, WHB, BAR, POT, RAP FI-2 Sou h Finland Mikkeli,Ylis a o,Laukaa,Piikio BAR, WHB, BAR, POT, RAP HR Kop i nica-K iže ci K iže ci MAZ IT-1 Foggia Foggia WHD, SBT IT-2 Val d’O cia Radico ani WHB, WHD, BAR NO Sou h Eas e n No way Sø åsjo de BAR NL Fle oland Lelys ad WHB, POT, SBT, MAZ PL Mazo ia D ˛ab owice WHB, BAR, POT, SBT, RAP SK Danube Lowland B a isla a-le isko, Hu bano o, Ni a, Jaslo ske Bohunice WHB, BAR, MAZ SR Voj odina Rimski Sance i WHD, MAZ, SBT, POT TR Th ace Edi ne, Kı kla eli, Teki da˘g WHD, WHB, BAR, MAZ No es: 1 In bold a e me eo ological s a ions loca ed in he icini y o expe imen al ields; 2 BAR is ba ley (Ho deum ulga e L.); MAZ is maize (Zea mays L.); POT is po a o (Solanum ube osum L.); SBT is suga bee (Be a ulga is L.); RAP is oilseed ape (B assica napus L.); WHB is common whea (T i icum aes i um L.); and, WHD is du um whea (T i icum u gidum L.). Wa e 2017,9, 93 4 o 22 Wa e 2017,9,934o 22   Figu e1.Loca iono di e en me eo ologicals a ionsac ossEu ope. 2.2.C opWa e Use FAO’s“Aquac op”model e sion5.0[18]wasused ocalcula e hec opwa e  oo p in .The g ow hmoduleise apo anspi a iond i en,whe ec op anspi a ion(T)iscon e ed obiomass h oughawa e p oduc i i ypa ame e [19,20].Thee apo a i epowe o  hea mosphe e(ET0)is con e ed oac uale apo anspi a ion(ET)andsepa a edin onon‐p oduc i ewa e  luxes,i.e.,soil e apo a ion(E),andp oduc i ewa e  luxes,i.e.,c op anspi a ion(T).Soilmois u econdi ions de e mineE om hesoilsu aceno co e edbycanopy[19,20].C opcanopyexpands om seedling oma u i yasde e minedbyaccumula edg owingdeg eedays. C opcalenda andg ow hcha ac e is icswe ecollec ed o  hemajo a ablec opsineach loca ion(Table1).Thec opg ow hpa ame e swe ese usingexpe imen al ieldda acollec ed o  each egion(AppendixA,[21]).Fo  egionswi hou expe imen al ieldda aa ailable,c opg ow h pa ame e swe ede i ed om a me s’ ields’da a. Allwea he ,soilandc opinpu da a(Figu e2;AppendixA)we einse edin o hemodel.The model’sphenologicalmodulewas uning owingdeg eedays ocap u ec opg ow hdynamics du ing heg owingseason.Rain edmodel uns o  hedi e en loca ionswe e ollowedby sp inkle i iga ion uns,a 80% ieldcapaci y,andacco ding olocal a mp ac ices.The e o e, egionswhe enoi iga ionwas epo edwe eexcluded om hei iga ionmodel uns.  Figu e 1. Loca ion o di e en me eo ological s a ions ac oss Eu ope. 2.2. C op Wa e Use FAO’s “Aquac op” model e sion 5.0 [ 18 ] was used o calcula e he c op wa e oo p in . The g ow h module is e apo anspi a ion d i en, whe e c op anspi a ion (T) is con e ed o biomass h ough a wa e p oduc i i y pa ame e [ 19 , 20 ]. The e apo a i e powe o he a mosphe e (ET0) is con e ed o ac ual e apo anspi a ion (ET) and sepa a ed in o non-p oduc i e wa e luxes, i.e., soil e apo a ion (E), and p oduc i e wa e luxes, i.e., c op anspi a ion (T). Soil mois u e condi ions de e mine E om he soil su ace no co e ed by canopy [ 19 , 20 ]. C op canopy expands om seedling o ma u i y as de e mined by accumula ed g owing deg ee days. C op calenda and g ow h cha ac e is ics we e collec ed o he majo a able c ops in each loca ion (Table 1). The c op g ow h pa ame e s we e se using expe imen al ield da a collec ed o each egion (Appendix A, [ 21 ]). Fo egions wi hou expe imen al ield da a a ailable, c op g ow h pa ame e s we e de i ed om a me s’ ields’ da a. All wea he , soil and c op inpu da a (Figu e 2; Appendix A) we e inse ed in o he model. The model’s phenological module was un in g owing deg ee days o cap u e c op g ow h dynamics du ing he g owing season. Rain ed model uns o he di e en loca ions we e ollowed by sp inkle i iga ion uns, a 80% ield capaci y, and acco ding o local a m p ac ices. The e o e, egions whe e no i iga ion was epo ed we e excluded om he i iga ion model uns. Wa e 2017,9, 93 5 o 22 Wa e 2017,9,935o 22  No he nEu ope(EE,FI,NO)  Sø åsjo de ,No way Jokioinen,Finland Wes e nEu ope(BE,DE,NL)  Lelys ad,TheNe he lands(NL) Gen ,Belgium(BE) Figu e 2. Con . Wa e 2017,9, 93 6 o 22 Wa e 2017,9,936o 22  Cen alEu ope(AT,CZ,DE,SK)  G ossEnze sdo ,Aus ia(AT) Hubano o,Slo akia(SK) Sou hEu ope(CY,HR,IT,SR,TR)  Foggia,I aly(IT) Teki dağ,Tu key(TR) Figu e2.Clima ologicaldiag ams o di e en me eo ologicals a ionsalongab oad ansec inEu ope o  hepe iod1992–2012.Pisp ecipi a ion(mm);ET0is e e ence e apo anspi a ion(mm);Tmeanisa e age empe a u e(°C).A wole e code e e s o hecoun ies.  Figu e 2. Clima ological diag ams o di e en me eo ological s a ions along a b oad ansec in Eu ope o he pe iod 1992–2012. Pis p ecipi a ion (mm); ET0 is e e ence e apo anspi a ion (mm); Tmean is a e age empe a u e (◦C). A wo le e code e e s o he coun ies. Wa e 2017,9, 93 7 o 22 2.3. Wa e Foo p in Calcula ions I iga ed ag icul u e ecei es wa e om i iga ion (blue wa e ) and om p ecipi a ion (g een wa e ), while ain ed ag icul u e only ecei es g een wa e . G een wa e is o igina ed by p ecipi a ion and is he soil wa e held in he unsa u a ed zone a ailable o plan s, while blue wa e e e s o he manageable wa e in i e s, lakes, we lands and aqui e s [ 22 ]. The g een WF and blue WF e lec he ain ed and i iga ed c op wa e use pe ha es ed c op wi h calcula ion me hods es ablished by [ 13 ]. The g ey WF accoun s o wa e used o dilu e nu ien pollu ion o mee ambien wa e quali y s anda ds; o easons o compa ison we ocused on ni ogen pollu ion [16]. WFg een =10·∑lgp d=1ETd,g een Y(1) WFblue =10·∑lgp d=1ETd,blue Y(2) WFg ey =[[∝·AR]/[cmax −cna ]] Y(3) whe e ETd is he daily e apo anspi a ion in mm · day −1 , accumula ed o e he leng h o he g owing pe iod (lgp, in days), unde ain ed (g een) and i iga ed (blue) condi ions. The ac o 10 con e s wa e dep hs om millime es in o wa e olumes pe land su ace (m 3· ha −1 ). The nomina o e lec s c op wa e use in m3·ha−1, whe eas he denomina o (Y) is c op yield in Mg·ha−1. The g een wa e e apo anspi a ion unde i iga ed condi ions was es ima ed as he o al e apo anspi a ion simula ed in a scena io wi hou i iga ion. The blue wa e e apo anspi a ion equalled he o al e apo anspi a ion simula ed in he scena io wi h i iga ion minus he simula ed g een wa e e apo anspi a ion. Fo he g ey WF, we assumed ha he ni ogen ac ion ( α ) ha eached ee lowing wa e bodies h ough leaching o uno equalled 10% o he applica ion a e (AR in kg · ha −1· yea −1 ). Fe ilize applica ion a es we e educed signi ican ly in he Eu opean Membe S a es ollowing he in oduc ion o he Ni a es Di ec i e in 1991 and he Wa e F amewo k Di ec i e in 2000. Repo ing mechanisms a e in place so ha ni ogen applica ion a es and de i ed g oss ni ogen balances a e a ailable om Eu os a o he pe iod 1992–2012 [ 23 ]. Fe ilize consump ion a es a e a ailable pe hec a e o a able land in he Wo ld Bank da abase [ 24 ]. We assumed d inking wa e s anda ds o wa e quali y wi h a di e ence be ween maximum accep able and na u al backg ound concen a ion (cmax −cna ) o 10 mg·L−1[16]. 2.4. Yield S a is ics Yield is an impo an componen o he WF. Yields, a ea and p oduc ion o whea , ba ley, g ain maize, po a o, suga bee and oilseed ape di e ed dis inc ly ac oss he di e en egions in Eu ope, as shown o 2012 egional s a is ical yields (Figu e 3). The ha es ed p oduc ion o ce eals in 2012–2015 in he EU-28 was es ima ed a one nin h o global ce eals p oduc ion; whea (44%–47%), maize (21%–22%) and ba ley (19%–20%) accoun o a high sha e [ 25 ]. Despi e a Eu opean-wide sys em o p oduc ion quo a, suga bee emains he mos impo an oo c op o no h-wes e n Eu ope. Po a o p oduc ion is mo e widely sp ead ac oss he di e en Eu opean Membe S a es, as e lec ed by he p esence o yield da a in di e en egions (Figu e 3). Oilseed ape, he main oilseed c op ac oss Eu ope, showed an upwa d end in p oduc ion du ing he las decade due o i s use o bioene gy pu poses [ 25 ]. Regional s a is ical yields we e compa ed wi h modelled yields assuming a humidi y o 14% o ce eals, 80% o oo c ops and 9% o oilseed ape [25]. Wa e 2017,9, 93 8 o 22 Wa e 2017,9,938o 22    Figu e3.Yields(Mg∙ha −1 ) o majo a ablec opsac oss heEu opean egions o  heyea 2012basedon egionals a is ics.A wole e code e e s o hecoun y ha  he egionbelongs o.  Figu e 3. Yields (Mg · ha −1 ) o majo a able c ops ac oss he Eu opean egions o he yea 2012 based on egional s a is ics. A wo le e code e e s o he coun y ha he egion belongs o. Wa e 2017,9, 93 9 o 22 2.5. S a is ical Analysis The s a is ical analysis was done in R using he co e unc ionali ies [ 26 ] and he hyd oGOF package [ 27 ]. Common s a is ical measu es we e used o desc ibe he da ase s. The coe icien o a ia ion (c ), i.e., he a io o he s anda d de ia ion o he mean exp essed in %, was used o compa e he sp ead o a iables. The Pea son co ela ion coe icien ( 2 ) was used as a measu e o s eng h o an associa ion be ween wo a iables. S a is ical me ics o desc ibe he ag eemen be ween modelled and s a is ical yields and be ween ou and benchma k WFs we e he mean a e age e o (MAE), he oo mean squa e e o (RMSE) and he index o ag eemen (d) [ 27 ]. The eg ession lines on he g aphs and he associa ed coe icien o de e mina ion (R 2 ) we e p o ided as a measu e o how well he s a is ical yields o he benchma k WFs we e app oxima ed by ou modelled esul s. 3. Resul s The wa e oo p in (WF) o a able c ops ac oss di e en egions in Eu ope showed a la ge a iabili y. We p esen ed his la ge a iabili y in ela ion o he di e en componen s ha comp ised he wa e oo p in : e apo a ion and anspi a ion; biomass and yield; and, he g een, blue and g ey WF. Since hese componen s we e in insically linked o he wa e balance, a gene al compa ison was made o he majo wa e balance inpu and ou pu . 3.1. Wa e Balance The wa e balance was d i en by e e ence e apo anspi a ion, calcula ed om sola adia ion, wind speed, empe a u e and ela i e humidi y using he modi ied Penman–Mon ei h equa ion [ 17 ]. In all s udied egions (Table 1), he e e ence e apo anspi a ion was highe han he p ecipi a ion accumula ed o e he g owing season o sp ing sown c ops (Figu e 4). Fo au umn sown c ops his di e ence was less p onounced. In no he n and wes e n Eu opean egions cumula i e p ecipi a ion was highe han cumula i e e apo anspi a ion du ing he g owing season o he pe iod 1992–2012. Simula ed sub-su ace d ainage was in all cases highe han simula ed su ace uno , bu his di e ence was no always signi ican (Figu e 5). A su plus on he wa e balance led o highe uno and d ainage du ing he g owing season, and ice e sa o a de ici . Due o highe p ecipi a ion du ing win e a su plus occu ed du ing he g owing season o au umn sown c ops (Figu e 5). Wa e 2017,9,939o 22  2.5.S a is icalAnalysis Thes a is icalanalysiswasdoneinRusing heco e unc ionali ies[26]and hehyd oGOF package[27].Commons a is icalmeasu eswe eused odesc ibe heda ase s.Thecoe icien o  a ia ion(c ),i.e., he a ioo  hes anda dde ia ion o hemeanexp essedin%,wasused ocompa e hesp eado  a iables.ThePea sonco ela ioncoe icien ( 2)wasusedasameasu eo s eng ho an associa ionbe ween wo a iables.S a is icalme ics odesc ibe heag eemen be weenmodelledand s a is icalyieldsandbe weenou andbenchma kWFswe e hemeana e agee o (MAE), he oo  meansqua ee o (RMSE)and heindexo ag eemen (d)[27].The eg essionlineson heg aphs and heassocia edcoe icien o de e mina ion(R2)we ep o idedasameasu eo howwell he s a is icalyieldso  hebenchma kWFswe eapp oxima edbyou modelled esul s. 3.Resul s Thewa e  oo p in (WF)o a ablec opsac ossdi e en  egionsinEu opeshowedala ge a iabili y.Wep esen ed hisla ge a iabili yin ela ion o hedi e en componen s ha comp ised hewa e  oo p in :e apo a ionand anspi a ion;biomassandyield;and, heg een,blueandg ey WF.Since hesecomponen swe ein insicallylinked o hewa e balance,agene alcompa ison wasmadeo  hemajo wa e balanceinpu andou pu . 3.1.Wa e Balance Thewa e balancewasd i enby e e encee apo anspi a ion,calcula ed omsola  adia ion, windspeed, empe a u eand ela i ehumidi yusing hemodi iedPenman–Mon ei hequa ion[17]. Inalls udied egions(Table1), he e e encee apo anspi a ionwashighe  han hep ecipi a ion accumula edo e  heg owingseasono sp ingsownc ops(Figu e4).Fo au umnsownc ops his di e encewaslessp onounced.Inno he nandwes e nEu opean egionscumula i ep ecipi a ion washighe  hancumula i ee apo anspi a iondu ing heg owingseason o  hepe iod1992–2012. Simula edsub‐su aced ainagewasinallcaseshighe  hansimula edsu ace uno ,bu  his di e encewasno alwayssigni ican (Figu e5).Asu pluson hewa e balanceled ohighe  uno  andd ainagedu ing heg owingseason,and ice e sa o ade ici .Due ohighe p ecipi a ion du ingwin e asu plusoccu eddu ing heg owingseasono au umnsownc ops(Figu e5).  Figu e4.P ecipi a ion(Pinmm)and e e encee apo anspi a ion(ET0inmm)du ing heg owing seasono au umnandsp ingsownc opsac oss heEu opean egions o  hepe iod1992–2012.A wole e code e e s o hecoun y ha  he egionbelongs o. Figu e 4. P ecipi a ion (Pin mm) and e e ence e apo anspi a ion (ET0 in mm) du ing he g owing season o au umn and sp ing sown c ops ac oss he Eu opean egions o he pe iod 1992–2012. A wo le e code e e s o he coun y ha he egion belongs o. Wa e 2017,9, 93 16 o 22 Wa e 2017,9,9316o 22   Figu e13.G eywa e  oo p in (inm3∙Mg−1) o modelledands a is icala ableyields o  hepe iod 1992–2012and o  ou di e en ni ogenapplica ion a es.No e hedi e encesinscalebe ween c ops.A wole e code e e s o hecoun y ha  he egionbelongs o. 4.Discussion Weused he“Aquac op”model oes ima ec opg ow hande apo anspi a ionunde bo h ain edandi iga edcondi ions.Thismodelhasbeende eloped osimula eyield esponse owa e  unde wa e ‐limi edcondi ions[18–20].Regions,c opsandsoils ha a esensi i e od yspellsand d ough p o ide o awa e ‐limi eden i onmen .Re iewso modelbeha iou showmixed esul s wi h espec  owa e usee icienciesandyields[28].Anin e compa isono eigh modelsshowed be ween13%and19%unce ain yin hees ima iono e apo anspi a ion(“Aquac op”:c =15%), andbe ween13%and34% o  anspi a ion(“Aquac op”:c =24%) o whea [21]. Theg eenWFshowed hela ges  a iabili y o ce eals(c =51%–52%),closely ollowedby po a o(c =48%); helowes  a iabili ywas o oilseed ape(c =36%)andsuga bee (c =28%) (Table3).Fo alla ablec ops, heyieldismo e a iable(   =45%;c inTable2) han hec op e apo anspi a ion(   =21%;c inTable3).Thisclea lydemons a es heimpo anceo yieldsand hei  a iabili y o  hewa e  oo p in .Simila  o he indingso [29], oo and ube c opsha e muchlowe WFsascompa ed oce ealsandoilseedc ops(Table3),owing oacombinede ec o  highe yieldsandhighe mois u econ en sa ha es .Ce ealsandoilseedc opsha eamuch smalle ha es able ac iono  he o albiomassp oducedpe su acea ea,and he e o eha e la ge wa e  oo p in s. Be ween hedi e en  egionsinEu ope,highyieldingwes e nEu opean egionsha eWFs ha canbeup osix imeslowe  han heWFso  egionsinno he no sou he nEu ope (Figu es9,11and13).A h ee oldinc easecanoccu be weenseasons,ce ainlyin egionswi h a iableyields.Ananalysiso  a iance(ANOVA)demons a edclea e ec so c ops,coun ies, seasonsand hei  ac o ialin e ac ionson heg eenwa e  oo p in (p<0.001).Thela ges  a iabili y be weenseasonsis o  hesou he n(CY,TR)andno he ncoun ies(EE,FI,NO)and o WHD, BARandRAP.The a iabili ybe weencoun ies(   =14%; ange:7%–26%)islowe  han he a iabili ybe weenseasons(   =22%; ange:10%–50%)and he a iabili ybe weenc ops (   =46%; ange:29%–52%). Table3.G eenwa e  oo p in (WF)(WFginm3∙Mg−1)ande apo anspi a ion(ETinmm) o  he majo a ablec opsinEu ope o  hepe iod1992–2012.Fo c opabb e ia ionsseeFigu e3. C opWFg∙mWFg∙sWFg∙c ET∙mET∙sET∙c  RAP18576613640510325 WHB1108580524598719 WHD1414720513756217 BAR901458513378224 MAZ590304523737320 POT15775483326419 SBT6719293518625 No es:Whe emdeno esmean,ss anda dde ia ionandc coe icien o  a ia ion(%). Figu e 13. G ey wa e oo p in (in m 3· Mg −1 ) o modelled and s a is ical a able yields o he pe iod 1992–2012 and o ou di e en ni ogen applica ion a es. No e he di e ences in scale be ween c ops. A wo le e code e e s o he coun y ha he egion belongs o. 4. Discussion We used he “Aquac op” model o es ima e c op g ow h and e apo anspi a ion unde bo h ain ed and i iga ed condi ions. This model has been de eloped o simula e yield esponse o wa e unde wa e -limi ed condi ions [ 18 – 20 ]. Regions, c ops and soils ha a e sensi i e o d y spells and d ough p o ide o a wa e -limi ed en i onmen . Re iews o model beha iou show mixed esul s wi h espec o wa e use e iciencies and yields [ 28 ]. An in e compa ison o eigh models showed be ween 13% and 19% unce ain y in he es ima ion o e apo anspi a ion (“Aquac op”: c = 15%), and be ween 13% and 34% o anspi a ion (“Aquac op”: c = 24%) o whea [21]. The g een WF showed he la ges a iabili y o ce eals (c = 51%–52%), closely ollowed by po a o (c = 48%); he lowes a iabili y was o oilseed ape (c = 36%) and suga bee (c = 28%) (Table 3). Fo all a able c ops, he yield is mo e a iable ( c = 45%; c in Table 2) han he c op e apo anspi a ion ( c = 21%; c in Table 3). This clea ly demons a es he impo ance o yields and hei a iabili y o he wa e oo p in . Simila o he indings o [ 29 ], oo and ube c ops ha e much lowe WFs as compa ed o ce eals and oilseed c ops (Table 3), owing o a combined e ec o highe yields and highe mois u e con en s a ha es . Ce eals and oilseed c ops ha e a much smalle ha es able ac ion o he o al biomass p oduced pe su ace a ea, and he e o e ha e la ge wa e oo p in s. Table 3. G een wa e oo p in (WF) (WFg in m 3· Mg −1 ) and e apo anspi a ion (ET in mm) o he majo a able c ops in Eu ope o he pe iod 1992–2012. Fo c op abb e ia ions see Figu e 3. C op WFg·m WFg·s WFg·c ET·m ET·s ET·c RAP 1857 661 36 405 103 25 WHB 1108 580 52 459 87 19 WHD 1414 720 51 375 62 17 BAR 901 458 51 337 82 24 MAZ 590 304 52 373 73 20 POT 157 75 48 332 64 19 SBT 67 19 29 351 86 25 No es: Whe e m deno es mean, s s anda d de ia ion and c coe icien o a ia ion (%). Be ween he di e en egions in Eu ope, high yielding wes e n Eu opean egions ha e WFs ha can be up o six imes lowe han he WFs o egions in no he n o sou he n Eu ope (Figu es 9,11 and 13) . A h ee old inc ease can occu be ween seasons, ce ainly in egions wi h a iable yields. An analysis o a iance (ANOVA) demons a ed clea e ec s o c ops, coun ies, seasons and hei ac o ial in e ac ions on he g een wa e oo p in (p< 0.001). The la ges a iabili y be ween seasons is o he sou he n (CY, TR) and no he n coun ies (EE, FI, NO) and o WHD, BAR and RAP. The a iabili y be ween coun ies ( c = 14%; ange: 7%–26%) is lowe han he a iabili y be ween seasons ( c = 22%; ange: 10%–50%) and he a iabili y be ween c ops ( c = 46%; ange: 29%–52%). Wa e 2017,9, 93 17 o 22 The b eakdown o he c op wa e oo p in s in di e en componen s enabled a be e unde s anding o he di e en con ibu ing ac o s in ol ed. A compa ison be ween ou calcula ions o he Eu opean egions and wa e oo p in benchma ks o c op p oduc ion p o ided by [ 16 ], e ealed a good ag eemen o he g een WF (R 2 = 0.80; d = 0.95), a easonably good ag eemen o he blue WF (R 2 = 0.64; d = 0.91) and a lowe ag eemen o he g ey WF (R 2 = 0.25; d = 0.73), whe e R 2 is he coe icien o de e mina ion and d is he index o ag eemen [ 27 ]. O e all he bes i was ob ained o he g een WF (Figu e 14). All WF we e highly in luenced by yield so ha only well calib a ed models able o model yield can be success ully deployed o es ima e he WF. Yield a iabili y de e mined he WF a iabili y. Wa e 2017,9,9317o 22  Theb eakdowno  hec opwa e  oo p in sindi e en componen senabledabe e  unde s andingo  hedi e en con ibu ing ac o sin ol ed.Acompa isonbe weenou  calcula ions o  heEu opean egionsandwa e  oo p in benchma ks o c opp oduc ionp o ided by[16], e ealedagoodag eemen  o  heg eenWF(R2=0.80;d=0.95),a easonablygood ag eemen  o  heblueWF(R2=0.64;d=0.91)andalowe ag eemen  o  heg ey WF(R2=0.25;d=0.73),whe eR2is hecoe icien o de e mina ionanddis heindexo ag eemen  [27].O e all hebes  i wasob ained o  heg eenWF(Figu e14).AllWFwe ehighlyin luenced byyieldso ha onlywellcalib a edmodelsable omodelyieldcanbesuccess ullydeployed o es ima e heWF.Yield a iabili yde e mined heWF a iabili y.  Figu e14.Compa isono g een,blueandg eywa e  oo p in s(WF)o  hiss udywi hbenchma k WF[16]inm3∙Mg−1 o a ablec ops.Theblueline ep esen s heiden i yline. C opg ow handp oduc iona emos lya ec edby hedis ibu iono g eenwa e du ing he g owingseason.Bluewa e di ec lyin luences heyieldp o idedwa e isa ailable o i iga ion: wemodelledyieldinc easeso up o50%whichhighligh  hebene i so i iga ion.Wa e sca ci y, exace ba ed u he byclima echange,isanissueo majo conce nina idandsemi‐a id egions wi hse iousimpac son oodsecu i y,sus ainabili yandeconomy.The ac  ha  hemajo i yo  a ailablewa e isconsumedbyag icul u alac i i ies,pa icula lyina idandsemi‐a idcoun ies, unde pins heneed o moni o ingand educingwa e consump ionpa e nsinag icul u ala eas. Ou modelledyieldinc easesdidno  esul inlowe c opWFsunde sp inkle i iga ion;d ip i iga ionmay esul inlowe WFsascalcula edby[14].TheWFo ag icul u alc opsallows o  decisionmakingandbe e managemen o  hewa e po en ial. Applied oag icul u alp oduc ion heg eyWFis heamoun o  eshwa e  equi ed o  he assimila iono anypollu an ,incasuni ogen uno due oag icul u alc opp oduc ion.Ni ogen applica ion a esdi e conside ablybe ween egions,and egula ionsa einplace olimi  heinpu  o exampleinni ogen ulne ablezoneso na u econse a iona eas[30]. O he impo an sou ceso  a ia iona ec op ype, a mingsys em,soil ype,and he ain all‐ uno  egime.When uno andd ainagewe e akenin oaccoun (Figu e5), heg eyWF showedalo mo e a iabili ybe ween heyea sandcouldbeaslowasze odu ingsomeyea s o  summe c ops.Inaddi ion, hepollu ionandhenceg eywa e isa ibu ed oasinglec op he eby neglec ing he oleo ac opin he o a ion.Fo exampleoilseed apehadahighg eyWF,despi e hec op’scapaci y odeple eni ogen om hep e iousc opbe o ewin e and he e o e educeN leaching.Acon a yexampleis hehighleaching isko ba esoildu ingwin e p io  osuga bee . Thesee ec swe eno inco po a edin heapplica ionso  heg eywa e  oo p in o c ops[13,16]. Recen applica ionsconcen a edonani ogenbalance obudge up akeandlosses,anda i eda  highe es ima eso ni ogen‐ ela edwa e pollu ionin i e basinsowing odi e encesin compu a ionalme hods[31].The e o e heg eyWFshouldbecompa edwi hcau ionbe ween s udiesandag icul u alsys ems. Figu e 14. Compa ison o g een, blue and g ey wa e oo p in s (WF) o his s udy wi h benchma k WF [16] in m3·Mg−1 o a able c ops. The blue line ep esen s he iden i y line. C op g ow h and p oduc ion a e mos ly a ec ed by he dis ibu ion o g een wa e du ing he g owing season. Blue wa e di ec ly in luences he yield p o ided wa e is a ailable o i iga ion: we modelled yield inc eases o up o 50% which highligh he bene i s o i iga ion. Wa e sca ci y, exace ba ed u he by clima e change, is an issue o majo conce n in a id and semi-a id egions wi h se ious impac s on ood secu i y, sus ainabili y and economy. The ac ha he majo i y o a ailable wa e is consumed by ag icul u al ac i i ies, pa icula ly in a id and semi-a id coun ies, unde pins he need o moni o ing and educing wa e consump ion pa e ns in ag icul u al a eas. Ou modelled yield inc eases did no esul in lowe c op WFs unde sp inkle i iga ion; d ip i iga ion may esul in lowe WFs as calcula ed by [ 14 ]. The WF o ag icul u al c ops allows o decision making and be e managemen o he wa e po en ial. Applied o ag icul u al p oduc ion he g ey WF is he amoun o eshwa e equi ed o he assimila ion o any pollu an , in casu ni ogen uno due o ag icul u al c op p oduc ion. Ni ogen applica ion a es di e conside ably be ween egions, and egula ions a e in place o limi he inpu o example in ni ogen ulne able zones o na u e conse a ion a eas [30]. O he impo an sou ces o a ia ion a e c op ype, a ming sys em, soil ype, and he ain all- uno egime. When uno and d ainage we e aken in o accoun (Figu e 5), he g ey WF showed a lo mo e a iabili y be ween he yea s and could be as low as ze o du ing some yea s o summe c ops. In addi ion, he pollu ion and hence g ey wa e is a ibu ed o a single c op he eby neglec ing he ole o a c op in he o a ion. Fo example oilseed ape had a high g ey WF, despi e he c op’s capaci y o deple e ni ogen om he p e ious c op be o e win e and he e o e educe N leaching. A con a y example is he high leaching isk o ba e soil du ing win e p io o suga bee . These e ec s we e no inco po a ed in he applica ions o he g ey wa e oo p in o c ops [ 13 , 16 ]. Recen applica ions concen a ed on a ni ogen balance o budge up ake and losses, and a i ed a highe es ima es o ni ogen- ela ed wa e pollu ion in i e basins owing o di e ences Wa e 2017,9, 93 18 o 22 in compu a ional me hods [ 31 ]. The e o e he g ey WF should be compa ed wi h cau ion be ween s udies and ag icul u al sys ems. P ac ices o educe he WF o c op p oduc ion s a wi h awa eness o he WF o di e en c op managemen sys ems. A ansi ion o less wa e -demanding c ops wi h highe wa e p oduc i i ies o highe wa e use e iciencies o e s oppo uni ies o op imize plan wa e use. Soil and wa e conse a ion echniques and wa e sa ing i iga ion me hods, e.g., d ip i iga ion and de ici i iga ion, could u he educe wa e demands [ 14 ]. Ad anced echniques lowe he c op wa e demand, bu canno ma kedly dec ease he WF; achie ing mo e s able and highe yields, howe e , can. The high dependency on yield wa an s s a egies o inc ease ag icul u al p oduc i i y which is accomplished h ough b eeding p og ams and/o h ough op imizing esou ces use du ing he c op g ow h season. The g ey WF is pa ly egula ized h ough he Wa e F amewo k and Ni a es Di ec i es wi h designa ed ni a e ulne able zones and limi a ions on ni ogen and phospho us applica ions [ 30 ]. Ensu ing ha wa e quali y is minimally a ec ed o e s good pe spec i es o nu ien sma p ecision a ming. O e all he wa e oo p in and i s assessmen p ocess helps es ablish a g ea e awa eness o wa e consump ion pa e ns among di e en s akeholde s in ol ed. 5. Conclusions We calcula ed he g een and blue wa e oo p in wi h FAO’s “Aquac op” model, and he g ey wa e oo p in on he basis o ni ogen applica ion a es o six majo a able c ops in 45 loca ions ac oss Eu ope o he pe iod 1992–2012. The WF o ce eals is la ge han he WF o ube and oo c ops owing mainly o he di e ence in yield and mois u e con en a ha es be ween hese c op ypes. Since yield has a la ge a iabili y han c op wa e use, yield es ima es a e o pa amoun impo ance o he c op WF. The WF o whea , o example, can be up o i e o six imes la ge in no he n and sou he n Eu ope as compa ed o high yielding wes e n Eu opean egions. The WF a iabili y be ween c ops was la ge han he a iabili y be ween seasons and in u n la ge han he a iabili y be ween coun ies. Yield inc eases unde sp inkle i iga ion we e no high enough o educe he wa e oo p in . Wa e sa ing i iga ion and soil conse a ion echniques, howe e , may esul in WF educ ions. The g een and blue WF, bu no he g ey WF, compa ed a ou ably wi h in e na ionally a ailable benchma k alues. Con on ed wi h d ainage and uno , he g ey WF ended o o e es ima e he con ibu ion o ni ogen o he su ace and g oundwa e . O he ag o-hyd ological me hods o calcula e he g ey WF esul ed in e en la ge alues which poin s o cau ion when compa ing di e en s udies. The la ge a iabili y be ween c ops, egions and seasons; and be ween yields and wa e use as majo componen s o he WF highligh s he impo ance o c op yield a iabili y. The wa e oo p in is a measu able indica o ha may suppo Eu opean wa e go e nance. Acknowledgmen s: The au ho s acknowledge unding om COST ES1106, Belgian Science Policy con ac numbe SD/RI/03A, JPI FACCE MACSUR (D.M. 24064/7303/15), 2812ERA 147 ( om BLE Ge many), he Czech p ojec LD13030, he Minis y o Educa ion, You h and Spo s o he Czech Republic wi hin he Na ional Sus ainabili y P og am I (NPU I), g an numbe LO1415, and con ac n III43007 o he Minis y o Educa ion and Science o he Republic o Se bia. The au ho s hank hei na ional me eo ological se ices and ins i u es o p o iding wea he da a. Suppo o da a collec ion is acknowledged om Sabina Thale o Aus ian da a, E a Pohanko á o Czech da a and Top ak Aslan o Tu kish da a. Au ho Con ibu ions: Anne Gobin, Ku Ch is ian Ke sebaum, Jose Ei zinge and Mi ek T nka concei ed and designed he expe imen s o calcula ing he wa e oo p in . “Aquac op” model uns we e pe o med by Anne Gobin. All au ho s con ibu ed o da a collec ion and in e p e a ion o he esul s. Con lic s o In e es : The au ho s decla e no con lic o in e es . Appendix A The dominan soil ype(s) and me eo ological s a ions o each egion a e p o ided in Table A1, oge he wi h e e ences o ele an da ase s. C op cha ac e is ics used o calib a ion a e p o ided in Table A2. Wa e 2017,9, 93 19 o 22 Table A1. Soil hyd ological p ope ies o he opsoil o di e en loca ions ac oss Eu ope. CTRY Loca ion Soil Type Tex u e 1FC (%) WP (%) Po e Space Re e ences AT G oss Enze sdo Che nozem Sil Loam 35 21 43 [32] AT G oss Enze sdo Pa ache nozem Sandy Loam 28 8 39 AT G oss Enze sdo Flu isol Clay Loam 35 22 42 AT Fuchsenbigl Calca ic Che nozem Sil Loam 38 23 53 BE Koksijde Calca ic & Gleyic Flu isol Ma ine Clay 39 23 50 [33–35] BE Gen Albelu isol Sandy Loam 22 10 47 BE Pee Podzol Loamy Sand 16 8 46 BE Ukkel Lu isol Sil y Loam 34 12 49 CY La naca Ch omic Ve isol Clay 42 28 50 [36,37] CY Nicosia Ve ic-Ch omic Lu isol Clay Loam 38 24 46 CY Pa os Eu ic Flu isol Loam 32 18 46 CZ Domaninek Dys ic Cambisol Loam 30 15 47 [38] CZ Lednice Che nozem Sil Loam 35 16 49 CZ Ve o any Che nozem Sil Loam 33 14 47 DE Manschnow Flu ic Gleysol Clay loam 39 15 46 [39] DE Manschnow Cambisol Sandy Loam 31 9 40 [39] DE Manschnow Podzol Sandy Loam 14 5 42 [39] DE Münchebe g Eu ic Cambisol Loamy sand 26 11 36 [40,41] DE B aunschweig Lu isol Sandy Loam 24 6 46 [41] EE Kuusiku Calcic Lu isol Sil Loam 28 7 40 [42] EE Väike-Maa ja Calca ic Cambisol Sandy Loam 28 8 45 EE Ta u Mollic Cambisol Loam 30 9 48 EE Võ u S agnic Lu isol Loamy Sand 20 6 42 EE Tallinn Haplic Albelu isol Sand 16 3 44 EE Ku essaa e Gleysol Clay 35 22 50 EE Pä nu Gleysol Clay loam 32 20 48 FI Jokioinen Haplic Umb isol Sil loam 35 21 45 [43] FI Mikkeli Mollic Cambisol Sandy Loam 28 7 42 FI Ylis a o Ve i-Gleyic Cambisol Sil Loam 35 15 48 FI Laukaa Eu ic Regosol Sil y Clay 46 25 55 FI Piikiö Ve ic Cambisol Clay Loam 36 22 48 HR K iže ci Gleyic Lu isol Sil loam 36 12 41 [44] IT Foggia Allu ial e isol Clay Loam 42 24 55 [45] IT Radico ani Ve ic Cambisol Sil y Clay 42 27 51 [46,47] NL Lelys ad Gleyic Flu isol Ma ine Clay 36 16 45 [48] NO Sø åsjo de Gleyic Podzolu isol Sil Loam 37 20 50 [49] PL D ˛ab owice Podzol Loamy Sand 23 17 40 SK Jasl.Bohunice Che nozem Sil y Loam 34 14 44 [50] SK Ni a Lu isol Clay Loam 36 17 44 SK B a isla a Flu isol Sandy Loam 32 12 44 SK Hu bano o Phaeozem Clay Loam 35 18 44 SR Rimski Sance i Che nozem Loam 34 17 51 [51] TR Ki kla eli Cambisol Sandy Clay Loam 35 17 42 [52] TR Teki da˘g Flu ic Cambisol Sandy Clay Loam 39 28 46 [53] TR Edi ne Cambisol Clay Loam 37 23 41 [53] No es: 1Soil ex u e is classi ied acco ding o he USDA nomencla u e. Table A2. Plan ing (P) and ha es ing (H) da es o a able c ops in he di e en egions. Region * WHB/WHD BAR RAP MAZ POT SBT P; H P; H P; H P; H P; H P; H AT 12/10; 30/7 25/3; 30/6 7/5; 26/9 16/4; 5/9 12/4; 18/8 BE 15/10; !/8 15/10; 15/7 15/9; 15/7 1/5; 30/9 10/4; 30/9 10/4; 15/10 CY 15/11; 30/5 15/11; 4/5 15/1; 24/5 CZ 3/10; 30/7 30/3; 25/7 28/8; 20/7 30/4; 15/9 DE-1 2/10; 30/7 20/9; 15/7 28/8; 24/7 16/5; 15/9 DE-2 25/10; 30/7 25/9; 25/6 15/4; 30/9 EE 30/8; 10/8 25/4; 3/8 1/5; 15/9 5/5; 10/9 FI-1 30/8; 20/8 15/5; 20/8 15/5; 10/9 15/5; 10/9 FI-2 10/9; 15/8 10/5; 15/8 10/5; 1/9 15/5; 5/9 HR 29/4; 3/10 IT-1 15/11; 20/6 22/3; 18/8 IT-2 15/11; 15/7 15/11; 10/7 Wa e 2017,9, 93 20 o 22 Table A2. Con . Region * WHB/WHD BAR RAP MAZ POT SBT P; H P; H P; H P; H P; H P; H NL 20/10; 30/7 5/9; 17/7 30/4; 15/10 25/4; 20/9 10/4; 15/10 NO 25/4; 15/8 PL 20/9; 14/7 24/4; 16/7 28/8; 17/8 15/4; 30/9 15/4; 20/9 SK 7/10; 20/7 24/3; 10/7 20/4; 3/9 15/4; 15/9 SR 15/11; 10/7 20/4; 30/9 30/3; 10/7 30/3; 15/10 TR 15/11; 30/6 30/9; 15/6 9/4; 20/8 15/3; 20/8 No es: * Fo he name o he egion see Table A1. Re e ences 1. Hoeks a, A.Y.; Mekonnen, M.M. The wa e oo p in o humani y. P oc. Na l. Acad. Sci. USA 2012 ,109, 3232–3237. [C ossRe ] [PubMed] 2. 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.; e al. 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 2005,437, 529–533. [C ossRe ] [PubMed] 3. COPA-COGECA. COPA-COGECA. Assessmen o he impac o he hea wa e and d ough o he summe o 2003 on ag icul u e and o es y. In Fac Shee s o he Commi ee o Ag icul u al O ganisa ions in he Eu opean Union and he Gene al Commi ee o Ag icul u al Coope a ion in he Eu opean Union; COPA-COGECA: B ussels, Belgium, 2003. 4. Dai, A. Inc easing d ough unde global wa ming in obse a ions and models. Na . Clim. Chang. 2013 ,3, 52–58. [C ossRe ] 5. T nka, M.; Olesen, J.E.; Ke sebaum, K.C.; Skjel ag, A.O.; Ei zinge , J.; Seguin, B.; Pel onen Sainio, P.; Ro e , R.; Iglesias, A.; O landini, S.; e al. Ag oclima ic condi ions in Eu ope unde clima e change. Glob. Chang. Biol. 2011,17, 2298–2318. [C ossRe ] 6. T nka, M.; Hla inka, P.; Semeno , M.A. Adap a ion op ions o whea in Eu ope will be limi ed by inc eased ad e se wea he e en s unde clima e change. J. R. Soc. In e ace 2015,12, 20150721. [C ossRe ] [PubMed] 7. Dame au, K.; Pa , A.G.; an Vlie , O.P. Wa e sa ing po en ials and possible ade-o s o u u e ood and ene gy supply. Glob. En i on. Chang. 2016,39, 15–25. [C ossRe ] 8. Allan, J.A. Vi ual wa e : A s a egic esou ce, global solu ions o egional de ici s. G ound Wa e 1998 ,36, 545–546. [C ossRe ] 9. Zoumides, C.; B uggeman, A.; Hadjikakou, M.; Zacha iadis, T. Policy- ele an indica o s o semi-a id na ions: The wa e oo p in o c op p oduc ion and supply u iliza ion o Cyp us. Ecol. Indic. 2014 ,43, 205–214. [C ossRe ] 10. Ridou , B.G.; P is e , S. A e ised app oach o wa e oo p in ing o make anspa en he impac s o consump ion and p oduc ion on global eshwa e sca ci y. Glob. En i on. Chang. 2010 ,20, 113–120. [C ossRe ] 11. In e na ional O ganiza ion o S anda diza ion. ISO 14046: 2014: En i onmen al Managemen : Wa e Foo p in —P inciples, Requi emen s and Guidelines; In e na ional O ganiza ion o S anda diza ion: Gene a, Swi ze land, 2014. 12. Chenowe h, J.; Hadjikakou, M.; Zoumides, C. Quan i ying he human impac on wa e esou ces: A c i ical e iew o he wa e oo p in concep . Hyd ol. Ea h Sys . Sci. 2014,18, 2325–2342. [C ossRe ] 13. Mekonnen, M.M.; Hoeks a, A.Y. The g een, blue and g ey wa e oo p in o c ops and de i ed c op p oduc s. Hyd ol. Ea h Sys . Sci. 2011,15, 1577–1600. [C ossRe ] 14. Chukalla, A.D.; K ol, M.S.; Hoeks a, A.Y. G een and blue wa e oo p in educ ion in i iga ed ag icul u e: E ec o i iga ion echniques, i iga ion s a egies and mulching. Hyd ol. Ea h Sys . Sci. 2015 ,19, 4877–4891. [C ossRe ] 15. Zhuo, L.; Mekonnen, M.M.; Hoeks a, A.Y. The e ec o in e -annual a iabili y o consump ion, p oduc ion, ade and clima e on c op- ela ed g een and blue wa e oo p in s and in e - egional i ual wa e ade: A s udy o China (1978–2008). Wa e Res. 2016,94, 73–85. [C ossRe ] [PubMed] Wa e 2017,9, 93 21 o 22 16. Mekonnen, M.M.; Hoeks a, A.Y. Wa e oo p in benchma ks o c op p oduc ion: A i s global assessmen . Ecol. Indic. 2014,46, 214–223. [C ossRe ] 17. Allen, R.G.; Pe ei a, L.S.; Raes, D.; Smi h, M. C op E apo anspi a ion—Guidelines o Compu ing C op Wa e Requi emen s; FAO I iga ion and D ainage Pape ; Food and Ag icul u e O ganiza ion: Rome, I aly, 1998; p. D05109. 18. FAO C op-Model o Simula e Yield Response o Wa e . A ailable online: h p://www. ao.o g/n /wa e / aquac op.h ml (accessed on 23 Augus 2016). 19. S edu o, P.; Hsiao, T.C.; Raes, D.; Fe e es, E. AquaC op—The FAO c op model o simula e yield esponse o wa e : I. Concep s and unde lying p inciples. Ag on. J. 2009,101, 426–437. [C ossRe ] 20. Raes, D.; S edu o, P.; Hsiao, T.C.; Fe e es, E. Aquac op he FAO c op model o simula e yield esponse o wa e : II. Main algo i hms and so wa e desc ip ion. Ag on. J. 2009,101, 438–447. [C ossRe ] 21. Ke sebaum, K.C.; K oes, J.; Gobin, A.; Takác, J.; Hla inka, P.; T nka, M.; Ven ella, D.; Giglio, L.; Fe ise, R.; Mo iondo, M.; e al. Assessing he unce ain y o model based wa e oo p in es ima ion using an ensemble o c op g ow h models on win e whea . Wa e 2016,8, 571. [C ossRe ] 22. Rocks öm, J.; Falkenma k, M.; Ka lbe g, L.; Ho , H.; Ros , S.; Ge en, D. Fu u e wa e a ailabili y o global ood p oduc ion: The po en ial o g een wa e o inc easing esilience o global change. Wa e Res. 2009 ,45, W00A12. [C ossRe ] 23. Eu os a . G oss Nu ien Balance. A ailable online: h p://ec.eu opa.eu/eu os a /cache/me ada a/EN/ aei_p _gnb_esms.h m (accessed on 23 Augus 2016). 24. Wo ld Bank, Fe ilize Consump ion Ra es pe Hec a e o A able Land. A ailable online: h p:// da abank.wo ldbank.o g/da a/ epo s.aspx?sou ce=2&se ies=AG.CON.FERT.ZS&coun y (accessed on 23 Augus 2016). 25. Eu os a . Handbook o Annual C op S a is ics. Eu opean Commission-Eu os a , Di ec o a e E Sec o al and Regional S a is ics, Uni E1 Ag icul u e and Fishe ies. A ailable online: h p://ec.eu opa.eu/eu os a / cache/me ada a/Annexes/ap o_acs_esms_an1.pd (accessed on 23 Augus 2016). 26. R Co e G oup. The R P ojec o S a is ical Compu ing. A ailable online: h ps://www. -p ojec .o g/ (accessed on 20 No embe 2016). 27. Zamb ano-Bigia ini, M. hyd oGOF: Goodness-o -Fi Func ions o Compa ison o Simula ed and Obse ed Hyd ological Time Se ies. 2011. R Package Ve sion 0.3-2. A ailable online: h p://CRAN.R-p ojec .o g/ package=hyd oGOF (accessed on 23 Augus 2016). 28. E e , S.R.; Tolk, J.A. In oduc ion: Can Wa e Use E iciency Be Modeled Well Enough o Impac C op Managemen ? Ag on. J. 2009,101, 423–425. [C ossRe ] 29. Mekonnen, M.M.; Hoeks a, A.Y. The G een, Blue and G ey Wa e Foo p in o C ops and De i ed C op P oduc s. Value o Wa e Resea ch Repo Se ies No. 47; UNESCO-IHE: Del , The Ne he lands, 2010. A ailable online: h p://www.wa e oo p in .o g/Repo s/Repo 47-Wa e Foo p in C ops-Vol1.pd (accessed on 23 Augus 2016). 30. Vel ho , G.L.; Lesschen, J.P.; Webb, J.; Pie zak, S.; Mia kowski, Z.; Pin o, M.; K os, J.; Oenema, O. The impac o he Ni a es Di ec i e on ni ogen emissions om ag icul u e in he EU-27 du ing 2000–2008. Sci. To al En i on. 2014,468, 1225–1233. [C ossRe ] [PubMed] 31. Mekonnen, M.M.; Hoeks a, A.Y. Global g ay wa e oo p in and wa e pollu ion le els ela ed o an h opogenic ni ogen loads o esh wa e . En i on. Sci. Technol. 2015 ,49, 12860–12868. [C ossRe ] [PubMed] 32. Ei zinge , J.; T nka, M.; Seme ádo á, D.; Thale , S.; S obodo á, E.; Hla inka, P.; Siska, B.; Takáˇc, J.; Mala inská, L.; No áko á, M.; e al. Regional clima e change impac s on ag icul u al c op p oduc ion in Cen al and Eas e n Eu ope–ho spo s, egional di e ences and common ends. J. Ag ic. Sci. 2013 ,151, 787–812. [C ossRe ] 33. Gobin, A. Impac o hea and d ough s ess on a able c op p oduc ion in Belgium. Na l. Haza ds Ea h Sys . Sci. 2012,12, 1911–1922. [C ossRe ] 34. Gobin, A. Modelling clima e impac s on a able yields in Belgium. Clim. Res. 2010,44, 55–68. [C ossRe ] 35. Gobin, A. The wa e oo p in o Belgian a able c ops. I al. J. Ag ic. Me eo ol. 2015,4, 91–97. 36. Zoumides, C.; B uggeman, A.; Zacha iadis, T.; Pashia dis, S. Quan i ying he poo ly known ole o g oundwa e in ag icul u e: The case o Cyp us. Wa e Resou . Manag. 2013,27, 2501–2514. [C ossRe ] Wa e 2017,9, 93 22 o 22 37. Came a, C.; Zomeni, Z.; Nolle , J.; Zissimos, A.; Ch is o o ou, I.; B uggeman, A. A high esolu ion map o soil ypes and physical p ope ies o Cyp us: A digi al soil mapping op imiza ion. Geode ma 2017 ,285, 35–49. [C ossRe ] 38. Hla inka, P.; T nka, M.; Ke sebaum, K.; Ce mak, P.; Pohanko a, E.; O s_ag, M.; Poko ný, E.; Fische , M.; B nický, M.; Zalud, Z. Modelling o yields and soil ni ogen dynamics o c op o a ions by HERMES unde di e en clima e and soil condi ions in he Czech Republic. J. Ag ic. Sci. 2014,152, 188–204. [C ossRe ] 39. Ke sebaum, K.C.; Nendel, C. Si e-speci ic impac s o clima e change on whea p oduc ion ac oss egions o Ge many using di e en CO2 esponse unc ions. Eu . J. Ag on. 2014,52, 22–32. [C ossRe ] 40. Mi schel, W.; Ba kusky, D.; Hu nagel, J.; Ke sebaum, K.C.; Nendel, C.; Laacke, L.; Luzi, K.; Rosne , G. Cohe en mul i- a iable ield da a se o an in ensi e c opping sys em o ag o-ecosys em modelling om Münchebe g, Ge many. Open Da a J. Ag ic. Res. 2016,2, 1–10. [C ossRe ] 41. Kollas, C.; Ke sebaum, K.C.; Nendel, C.; Mane ski, K.; Mülle , C.; Palosuo, T.; A mas-He e a, C.M.; Beaudoin, N.; Bindi, M.; Cha eddine, M.; e al. C op o a ion modelling–a Eu opean model in e compa ison. Eu . J. Ag on. 2015,70, 98–111. [C ossRe ] 42. Kadaja, J.; Saue, T. E ec s o di e en i iga ion and d ainage egimes on yield and wa e p oduc i i y o wo po a o a ie ies unde Es onian empe a e clima e. Ag ic. Wa e Manag. 2016,165, 61–71. [C ossRe ] 43. Pel onen-Sainio, P.; Jauhiainen, L.; Hakala, K. C op esponses o empe a u e and p ecipi a ion acco ding o long- e m mul i-loca ion ials a high-la i ude condi ions. J. Ag ic. Sci. 2011,149, 49–62. [C ossRe ] 44. Vuˇce i´c, V. Modelling o maize p oduc ion in C oa ia: P esen and u u e clima e. J. Ag ic. Sci. 2011 ,149, 145–157. [C ossRe ] [PubMed] 45. Ven ella, D.; S ellacci, A.M.; Cas ignanò, A.; Cha eddine, M.; Cas ellini, M. E ec s o c op esidue managemen on win e du um whea p oduc i i y in a long e m expe imen in Sou he n I aly. Eu . J. Ag on. 2016,77, 188–198. [C ossRe ] 46. Dalla Ma a, A.; O lando, F.; Mancinii, M.; Guasconi, F.; Mo ha, R.; Qu, J.; O landini, S. A simpli ied index o an ea ly es ima ion o du um whea yield in Tuscany (Cen al I aly). Field C ops Res. 2015 ,170, 1–6. [C ossRe ] 47. Guasconi, F.; Dalla Ma a, A.; G i ono, D.; Mancini, M.; O lando, F.; O landini, S. In luence o clima e on du um whea p oduc ion and use o emo e sensing and wea he da a o p edic quali y and quan i y o ha es s. I al. J. Ag ome eo ol. 2011,3, 21–28. 48. Van Bakel, P.J.T.; Massop, H.T.L.; K oes, J.G.; Hoogewoud, J.; Pas oo s, R.; K oon, T. Ac ualisa ie Hyd ologie oo STONE 2.3. In Aanpassing Rand oo waa den en Pa ame e s, Koppeling Tussen NAGROM en SWAP, en Plausibili ei s oe s; WO -Rappo 57; We elijke Onde zoeks aken Na uu & Milieu (MNP), Al e a: Wageningen, The Ne he lands, 2008. 49. Deels a, J.; K æ nø, S.H.; Skje dal, R.; Vandsemb, S.; Egges ad, H.O.; Lud igsen, G.H. A Gene al Desc ip ion o Sku e ud Ca chmen ; Jo d o sk (Now NIBIO) Repo No. 61/05; Bio o sk: Ås, No way, 2005. 50. Takáˇc, J.; Skalský, R.; Mo á ek, A.; Klikušo ská, Z.; Bezák, P.; Bá dyo á, M. Spa ial Pa e ns o Ag icul u al D ough E en s in Danube Lowland in he 1961–2013 Pe iod. In P oceedings o he In e na ional Scien i ic Con e ence owa ds Clima ic Se ices, Ni a, Slo akia, 15–18 Sep embe 2015. 51. S iˇce i´c, R.; Cosi´c, M.; Dju o i´c, N.; Peji´c, B.; Maksimo i´c, L. Assessmen o he FAO “Aquac op” model in he simula ion o ain ed and supplemen ally-i iga ed maize, suga bee and sun lowe . Ag ic. Wa e Manag. 2011,98, 1615–1621. 52. ¸Saylan, L.; Çalda˘g, B.; Bakano˘gulla ı, F. In es iga ion o Po en ial E ec s o Clima e Change on C op G ow h by C op G ow h Simula ion Models; TUBITAK P ojec No.: 1080567; TUBITAK: Is anbul, Tu key, 2012; p. 258. (In Tu kish) 53. Çaki , R. Is anca (Yıldız) Da˘gı Güneyinde Ye Alan Ve isol O dosu Top akla ının Top ak Taksonomisine Gö e Beli lenmesi, Top ak e Su Mühendisli˘gi Yönünden ˙ I delenmesi. Ph.D. Thesis, T akya Üni e si esi, Edi ne, Tu key, 1997. © 2017 by he au ho s; licensee MDPI, Basel, Swi ze land. This a icle is an open access a icle dis ibu ed unde he e ms and condi ions o he C ea i e Commons A ibu ion (CC BY) license (h p://c ea i ecommons.o g/licenses/by/4.0/).