Thermal Imaging at Plant Level to Assess the Crop-water Status in Almond Trees (cv. Guara) Under Deficit Irrigation Strategies

1 The mal imaging a plan le el o assess he c op-wa e s a us in almond ees (c . 1 Gua a) unde de ici i iga ion s a egies 2 3 Ga cía-Teje o I.F.1*, Rubio, A.E2., Viñuela, I1., He nández, A1., Gu ié ez-Go dillo, S1., Rod íguez-Pleguezuelo, C.R.3, 4 Du án-Zuazo V.H.3 5 1 Ins i u o Andaluz de In es igación y Fo mación Ag a ia, Pesque a y de la P oducción Ecológica (IFAPA). Cen o “Las 6 To es – Tomejil”. C a. Se illa-Cazalla Km. 12,2. 41.200. Alcalá del Río, Se illa, Spain. 7 2 Facul ad de Biología. Depa amen o de Biología Vege al y Ecología. Uni e sidad de Se illa. A enida de Reina 8 Me cedes s/n. 41012. Se illa, Spain. 9 3 Ins i u o Andaluz de In es igación y Fo mación Ag a ia, Pesque a y de la P oducción Ecológica (IFAPA). Cen o 10 “Camino de Pu chil”. Apdo. 2027, 18080, G anada, Spain 11 12 *e-mail: i an .ga cia@jun adeandalucia.es 13 Abs ac 14 Almond (P unnus dulcis Mill.) has been adi ionally associa ed o ma ginal land cul i a ion and ain- ed 15 ag icul u e in Sou h Spain. Howe e , in he las yea s, his c op is being p og essi ely in oduced in mo e 16 p oduc i e ag icul u al a eas wi hin he Guadalqui i i e basin, whe e he a ailable wa e esou ces a e 17 no enough o sa is y he adequa e c op-wa e equi emen s. Conside ing his limi a ion, a mo e p ecise 18 i iga ion scheduling o maximize he yield is equi ed. In a ed he mal imaging eme ges as al e na i e o 19 o he adi ional me hodologies o assess he c op-wa e s a us, especially when de ici i iga ion (DI) 20 s a egies a e being applied. The aim o his s udy was o de ine he me hodology o assess he almond 21 wa e s a us by means o he mal in o ma ion. The ial was conduc ed du ing 2014, du ing he ke nel- illing 22 pe iod, in an almond expe imen al o cha d (SW Spain), wi h 5-yea -old ees, subjec ed o h ee i iga ion 23 egimes: i) a ull-i iga ion ea men (C-100), which ecei ed 100% o ETC; ii) a egula ed de ici i iga ion 24 (RDI-50), which ecei ed 100% o ETC excep du ing he ke nel illing pe iod, when his ea men was 25 i iga ed wi h 50% o ETC; iii) and a low- equency de ici i iga ion ea men (LFDI), which ecei ed 100% 26 o ETC excep du ing he ke nel illing pe iod, when i was subjec ed o con inuous pe iods o i iga ion- 27 es ic ion, de ined in e ms o he h eshold alues o shaded lea wa e po en ial (Ψlea ). Th ee daily cu es 28 o canopy empe a u e (TC), s oma al conduc ance o wa e apou (gs) and Ψlea wi h measu emen s a 8:00, 29 11:00, 14:00, 17:00 and 20:00 we e de eloped. Addi ionally, C op Wa e S ess Index (CWSI), empe a u e 30 di e ence be ween canopy and he su ounding ai (ΔTcanopy-ai ), and he ela i e index o s oma al 31 conduc ance (IG) ob ained a di e en scales (canopy and ow) we e es ima ed. Signi ican co ela ions o 32 in a ed he mal in o ma ion s. Ψlea and gs we e ob ained (p ≤ 0.05 and p ≤ 0.01), in pa icula , by using 33 he he mal eadings aken a 11:30, 14:30 a 17:30 h, especially obus we e he ela ionships ob ained 34 be ween TC and CWSI wi h Ψlea a 11:30 h; and be ween TC and CWSI wi h gs, and Ψlea a 14:30 h. Finally, 35 2 conside ing he in a ed he mal moni o ing p ocedu e ( eadings a ee and ow le el), simila alues o TC 36 we e ob ained, and he e o e, he images aken a ow le el o e ed a be e in o ma ion wi h a highe 37 easibili y in e ms o image p ocessing. 38 Keywo ds: The mog aphy, he mal indexes, wa e s ess, lea gas exchange and lea wa e po en ial. 39 1.- In oduc ion 40 I iga ed ag icul u e in he Sou h o Eu ope, and mo e conc e ely in semi-a id a eas such as Andalusia (S 41 Spain), is c ucial o hei de elopmen , especially in hose u al egions wi h a lowe economic po en ial. In 42 his line, o he case o Andalusia, i iga ed ag icul u e gene a es mo e han 60% o u al employmen s, and 43 ep esen s 64% o ag icul u al p oduc ion. Cu en ly, 1,176,000 ha a e de o ed o i iga ed ag icul u e, 44 co esponding o 24% o o al Andalusian ag icul u al su ace, and his being 33% o he i iga ed ag icul u e 45 in Spain (ARA, 2011). 46 Clima ic condi ions in his a ea a e cha ac e ized by he sca ci y and i egula i y o ain all, coinciding he 47 d y pe iod wi h he season o highes e apo anspi a ion. Mo eo e , he las o ecas p edic ions a gue 48 signi ican wa e esou ces deple ions; wi h an impo an declining in he soil wa e ese es, mo e accused 49 pe iods o ain all es ic ions and inc easing in he a e age empe a u es (IPCC, 2014). In his ag eemen , 50 i is expec ed ha his si ua ion p omo es an imbalance be ween he i iga ion demand and he a ailable 51 wa e esou ces in he Medi e anean ag icul u e (Daccache e al. 2012, Olesen e al. 2011). This ac will 52 suppose an impo an cons ain o he compe i i eness be ween ag icul u e and o he mo e p oduc i e 53 sec o s such as he indus y o ou ism. In addi ion, he in oduc ion o al e na i e c ops in o de o maximize 54 he p o i abili y o ag oecosys ems will be equi ed, oge he wi h di e en s a egies o imp o e he 55 ag icul u al wa e managemen (Ga cía-Teje o e al. 2014a). 56 In his con ex , almond (P unus dulcis Mill.) is he hi d c op in e ms o su ace in Spain, ep esen ing globally 57 almos 40%, and 84% wi hin he EU. Howe e , only 5% o he global p oduc ion is de eloped in Spain 58 (FAOSTAT, 2016). Conc e ely, he su ace o almond in Andalusia is abou 152,000 ha, and wi hin hem, 59 95% a e associa ed o ma ginal and ain- ed ag icul u e because o he clima e limi a ions, whe e annual 60 ain alls does no exceed o 300 mm wi h low nu yields (CAPDR, 2016). Howe e , in he las ew yea s, he 61 ag icul u al su ace de o ed o almond c op has signi ican inc eased, specially, in a eas whe e his c op 62 was no adi ionally cul i a ed, hese new o cha ds being cul i a ed unde in ensi e and i iga ion p ac ices. 63 Thus, almond can be ound unde e y di e en ag icul u al sys ems om he mos ma ginal si ua ions o 64 he mos in ensi e o cha ds, which p omo es a wide ange o yields ( om 150 o 2,600 kg ha-1) (CAPDR, 65 2016). 66 Acco ding o Goldhame and Fe e es (2016), i iga ion is he mos limi ing ac o o his c op, wi h c op 67 wa e - equi emen s oscilla ing be ween 900 and 1,350 mm (Goldhame and Gi ona, 2012). In his 68 ag eemen , Goldhame and Fe e es (2016) epo ed alues close o 4,000 kg ha-1 (depending on he 69 cul i a ) o i iga ion doses a ound 1,250 mm, wi h yield educ ions close o 14% when he i iga ion doses 70 3 we e close o 1,000 mm. Mo e ecen ly, López-López e al. (2018) in a long- e m expe ience de eloped in 71 he p o ince o Có doba (Andalusia, Sou h Spain), epo ed maximum yield alues ( 2,500 kg ha-1) in 72 ma u e almond ees (c . Gua a), when hese ees we e i iga ed ecei ing he maximum c op wa e 73 equi emen s (close o 10,000 m3 ha-1). 74 In spi e o his, almond is conside ed a d ough - esis an c op because o i s xe omo phic p ope ies 75 (To ecillas e al. 1996), and many au ho s ha e epo ed di e en esul s ela ed o he e ec s o de ici 76 i iga ion (DI) s a egies (Pue o e al., 2013; Phoga e al., 2013; 2018; Spinelli e al., 2016; among o he s). 77 Mo e ecen ly, López-López e al. (2018) discussed he e ec s o wa e de ici s in almond ees in e ms o 78 wa e use, e alua ing di e en de ici i iga ion (DI) s a egies du ing h ee consecu i e yea s. These au ho s 79 ound ha almond ees unde di e en mode a e DI s a egies we e able o keeping canopy olumes simila 80 o hose ees ha we e ully i iga ed, hese being di ec ly ela ed wi h he almond capabili y o ob ain yield 81 alues unde mode a e de ici i iga ion simila o hose epo ed by ully i iga ed ees; his ac being 82 accompanied wi h simila soil wa e deple ions and anspi a ion le el. 83 Taking in o accoun he maximum c op-wa e demand, he wa e sca ci y in semi-a id a eas, and he p ope 84 esponse o his c op o mode a e wa e s ess, DI would be a sui able al e na i e o each equilib ium 85 be ween he a ailable wa e esou ces and a p ope c op de elopmen wi h inal yields able o ensu e he 86 compe i i eness and easibili y o his c op (Ga cía-Teje o e al., 2016a). Howe e , he applica ion o DI 87 s a egies equi es a p ope knowledge abou he c op physiological s a us, wi h he aim o ensu ing he 88 co ec c op de elopmen wi hou signi ican comp omising he yield and ui -quali y, especially when wa e - 89 s ess is applied in di e en c op s ages (Spinelli e al., 2016). In his sense, acco ding o Pue o e al. (2013), 90 when a DI s a egy is applied in ui ees, his is mainly de eloped supplying a speci ic wa e wi hholding, 91 aken as e e ence he c op wa e equi emen s by means o he c op e apo anspi a ion (ETC), wi hou 92 aking in o accoun he e ec s o canopy a chi ec u e, he deg ee o canopy co e o he soil managemen 93 (among o he s); o wi hou conside ing he c op physiological s a us when his wa e s ess is applied. In 94 his ega d, he mos p ope i iga ion scheduling should conside he whole o soil-plan -a mosphe e sys em; 95 al hough in e ms o ep esen a i eness, he li e componen (plan ) would be o e ing he mos aluable 96 in o ma ion, inasmuch as his e lec s he mos in eg a i e in o ma ion, mainly in e ms o inal yield. 97 T adi ionally, c op wa e moni o ing has been de eloped by using punc ual measu emen s o s em (Ψs em) 98 o lea (Ψlea ) wa e po en ial a midday o p e-dawn (Ψpd) (Shackel, 2011; No es e al., 2005) o moni o ing 99 he gas-exchange pa ame e s such as anspi a ion (E), s oma al conduc ance (gs) o ne pho osyn he ic 100 a e (A) (Gomes-La anjo e al., 2006). 101 Acco ding o Remo ini and Massai (2003), S em is no only a p ope indica o o plan -wa e s a us as well 102 as he c op p oduc i i y. In he same ein, Mi ás-A alos e al. (2016) epo ed ha wa e po en ial is a 103 sui able indica o o almond wa e s a us, al hough i s use ulness is educed, because o a minimum numbe 104 o eplica ions a e equi ed, and he ep esen a i eness in he whole plan is educed. 105 In he las yea s, he use o emo e sensing in ag icul u e, and mo e conc e ely, in a ed he mal imaging o 106 moni o he c op wa e s a us has been p og essi ely in oduced (Cos a e al., 2013). This echnique has 107 4 been p ope ly desc ibed as a good me hodology o c op-wa e moni o ing in di e en woody c ops such as 108 ci us (Ga cía-Teje o e al., 2011; González-Dugo e al., 2014); young almonds (Ga cía-Teje o e al., 2012), 109 ines (Ga cía-Teje o e al., 2016b) o oli es (Egea e al., 2017). This echnique is based on he lea ene gy 110 balance. When a wa e s ess si ua ion is applied, plan s esponds wi h a pa ial s oma al closu e, educing 111 he s oma al conduc ance, limi ing he lea anspi a ion and p omo ing an a enua ion o he e apo a i e 112 cooling p ocess, esul ing in highe lea / canopy empe a u e alues (Jones, 1999; 2004). 113 This echnique can be applied a di e en moni o ing scales, om “lea o canopy” o “o cha d o basin” le el 114 (Poble e-Eche e ía e al., 2014; 2016). The selec ion o he mos p ope me hodology will be ela ed wi h 115 he desi ed goal and he economic a ailabili y (Cos a e al., 2013). In his sense, he use o he mog aphy 116 a o cha d scale by using sa elli es images, allows o ake decisions ela ed o c op a iabili y o i iga ion 117 scheduling, bu some cons ain s mus be aken in o accoun . On one hand, he mal images aking by 118 sa elli es ha e he di icul y o depending o he momen in which he sa elli e passes abo e he o cha d; 119 and on he o he hand, he spa ial and spec al esolu ion is no p ope . These cons ain s could be sol ed 120 by using o unmanned ae ial ehicles (UAVs), despi e i s economically es ic ions. In his sense, he use o 121 he mal images a o cha d scale, aken by means o UAVs, equi es ha ing he p ope echnology; and his 122 ac can inc ease he cos o his ool, becoming less accessible he use o his echnology. By he con as , 123 hese senso s can be used a plan le el, wi h he mal came as much mo e p o i able, easing he 124 accessibili y o his echnique by he i iga ion communi ies o echnicians. 125 Likewise, he main cons ain s o his echnique a e ocused in he image p ocessing (many imes equi ing 126 high ime consuming), and he co ec in e p e a ion o he in a ed he mal in o ma ion (Ga cía-Teje o e al., 127 2015a). Because o his, many imes di e en ela ionships be ween in a ed he mal in o ma ion and o he 128 physiological pa ame e s such as gs, A, E, o Ψs em a e equi ed (Jones 2004; Jones e al., 2009), al hough 129 hese ela ionships a e no always enough obus because o he high dependence o he me eo ological 130 condi ions (Jones, 1999; 2004), he moni o ing p oceedings (Cos a el al., 2013), he cul i a (Cos a e al., 131 2012; Ga cía-Teje o e al., 2016b) o e en, he c op phenological s age (Cohen e al., 2015). 132 Up o day, se e al au ho s ha e de eloped s a egies o op imize his echnique, de eloping di e en 133 p o ocols and s a egies o ake he mal eadings unde ield condi ions (Jones e al., 2009; Pou e al., 2014; 134 Poble e-Eche e ía e al., 2014, 2016, Ga cía-Teje o e al., 2012, 2016b) and desc ibing di e en 135 ela ionships be ween in a ed he mal in o ma ion and physiological pa ame e s. 136 We hypo hesize ha he mog aphy could be a sui able echnique o moni o almond wa e s a us, especially 137 when his is subjec ed o DI p og ams. The aim o he p esen wo k was o e alua e he pe o mance o 138 he mog aphy unde ield condi ions a wo moni o ing le els (plan and ow) o assess he c op wa e s a us 139 in almond ees (c . Gua a), de e mining he bes momen o he day o ob ain he he mal in o ma ion and 140 he mos obus he mal index o in e p e p ope ly he c op-wa e s a us. 141 142 143 144 5 2. Ma e ial and me hods 145 2.1. Expe imen al si e 146 The ial was conduc ed du ing 2014 in an expe imen al o cha d o almonds (P unus dulcis Mill. D.A. Webb 147 c . Gua a, g a ed on o GF677), loca ed in he Guadalqui i i e basin (37º 30’ 47’’ N; 5º 58’ 2’’ O) (Se ille, 148 SW Spain). Plan ed in 2009, he ees we e spaced 6 x 7 m, and d ip i iga ed using wo pipe lines wi h 149 emi e s o 2.3 L h-1, and 14 emi e s pe ee. The soil is sil y loam, ypical Flu isol (USDA, 2010), 2.5 m 150 deep, e ile, and low ino ganic ma e con en (< 15.0 g kg-1). The oo s a e loca ed p edomina ely in he 151 i s 50 cm o soil, co esponding o he in ended we ing dep h, al hough hese exceed mo e han one me e 152 in dep h. Soil-wa e con en alues a ield capaci y (–0.033 MPa) and wil ing poin (–1.5 MPa) we e 0.35 153 and 0.12 m3 m–3 espec i ely, wi h an allowable soil-wa e deple ion le el o 0.27 m3 m–3. 154 The clima ology in he s udy a ea is a enua ed meso-Medi e anean, wi h an annual ET0 a e o 1,400 mm 155 and accumula ed ain all o 540 mm, mainly dis ibu ed om Oc obe o Ap il. 156 157 2.2. I iga ion ea men s 158 Th ee i iga ion ea men s we e applied: i) a ull i iga ed ea men (C-100), which ecei ed 100% o he 159 c op e apo anspi a ion (ETc) du ing he i iga ion pe iod (60 – 304 day o he yea , DOY), ii) a egula ed 160 de ici i iga ion (RDI-50), which ecei ed 100% o ETC excep du ing he ke nel illing pe iod and p e-ha es ; 161 when his ea men was i iga ed a 50% o ETC. Acco ding o his, he ke nel- illing pe iod ook place om 162 171 o 227 DOY and p e-ha es om 228 o 243 DOY; his pe iod coinciding wi h he ime in which he 163 ke nel has inished i s g ow h and he nu spli pe iod begins, jus be o e he i iga ion wi hholding (250 DOY) 164 se en days be o e he ha es ing (257 DOY). iii) and a low- equency de ici i iga ion (LFDI) which ecei ed 165 he 100% ETc du ing he i iga ion pe iod, excep du ing he ke nel- illing s age and p e-ha es ; when his 166 ea men was i iga ed acco ding he egis e ed alues o Ψlea measu ed in shaded lea es. In his sense, 167 du ing he ke nel- illing pe iod ( om 171 o 227 DOY) his ea men was subjec ed o i iga ion- es ic ion 168 cycles wi h he ollowing i iga ion dynamic: Once s a ed he ke nel- illing pe iod, i iga ion was sup essed, 169 ill eaching alues o Ψlea close o -2.0 MPa. Then, ees we e e-wa e ed wi h he same pe iodici y and 170 amoun o wa e as C-100 (app oxima ely du ing 5 - 7 days) ill eaching simila alues o Ψlea o hose 171 egis e ed in C-100. Once his h eshold alue was eached, his ea men was subjec ed o a new es ic ion 172 pe iod un il he h eshold o Ψlea (~ -2.0 MPa) was again su passed. This dynamic o i iga ion- es ic ion 173 cycles was main ained du ing whole s age o ke nel illing pe iod un il ha es ing. 174 I iga ion doses we e calcula ed acco ding o he me hodology p oposed by Allen e al. (1998), ob aining he 175 alues o e e ence e apo anspi a ion acco ding o he Penman-Mon ei h equa ion; by using a wea he 176 s a ion ins alled in he same expe imen al o cha d; and using he c op coe icien s ob ained by Ga cía-Teje o 177 e al. (2015b), which anged be ween 0.6 and 1.2. Acco ding o his, i iga ion doses applied o C-100, RDI- 178 50 and LFDI we e 6,850, 4,400 and 4,180 m3 ha-1, espec i ely (Table 1). 179 180 181 6 2.3. Plan measu emen s 182 Du ing he expe imen al pe iod, h ee daily cu es o canopy empe a u e (TC), s oma al conduc ance o 183 wa e apou (gs) and lea wa e po en ial (Ψlea ) we e ob ained du ing he ke nel illing and p e-ha es 184 pe iod. These eadings we e aken a 08:30, 11:30, 14:30, 17:30 and 20:00 h local ime, du ing he days 185 29 h July (Cu e 1) (210 DOY); 5 h Augus (Cu e 2) (217 DOY) and 27 h Augus (Cu e 3) (239 DOY). These 186 days coincided wi h he i iga ion es ic ion pe iods o LFDI, wi h he aim o egis e ing he c op physiological 187 s a us du ing pe iods o maximum wa e s ess in his ea men . In his sense, Cu e 2 was de eloped a 188 week a e Cu e 1. The eason was ha , when Cu e 1 was de eloped, LFDI has been subjec ed o se en 189 days o i iga ion es ic ion. Taking in o accoun he ob ained esul s du ing his cu e, i was decided o 190 ex end his pe iod once mo e week, in o de o egis e he c op physiological esponse unde a si ua ion o 191 maximum s ess. Finally, be ween Cu e 2 and 3, he e was a eco e y pe iod ( om 218 o 225 DOY), being 192 he Cu e 3 de eloped a e 14 days wi hou i iga ion (in simila condi ions a Cu e 2). 193 Table 2 shows he alues o ai empe a u e (Tai ), ela i e humidi y (RH), and apou p essu e de ici (VPD) 194 egis e ed du ing he sampling days and o each moni o ing hou . 195 Measu emen s o Ψlea we e conduc ed by using a p essu e chambe (Soil Mois u e Equipmen Co p., S a. 196 Ba ba a, CA, USA), moni o ing 12 ees pe i iga ion ea men (one lea pe ee), loca ed in he no h side 197 o he ee and being o ally ma u e, esh and shaded, a 1.5 m o heigh , app oxima ely. Addi ionally, he 198 s oma al conduc ance o wa e apo (gs), was measu ed in hese same ees, by using a po ome e SC-1 199 (Decagon De ices, INC, WA, USA), on one lea comple ely exposed o he sun pe moni o ed ee, and a 200 1.5 m o heigh . 201 TC was measu ed by using a The maCam (Fli SC660, Fli Sys ems, USA, 7-13 μm, 640x480 pixels) 202 h oughou he day (8:30h, 11:30h, 14:30h, 17:30h, and 20:00h local ime), wi h emissi i y (ε) se a 0.96. 203 Each pixel co esponds o an e ec i e empe a u e eading (Jones, 2004). Two me hodologies we e es ed 204 o moni o he canopy empe a u e: i) 12 images we e aken a ee le el (one image pe ee assessed, 205 hee being he same ees in which he measu emen s o Ψlea and gs we e de eloped), o each daily cu e, 206 ea men and momen o he day), and ii) du ing Cu es 1 and 2, he mal images we e aken a ow le el, 207 so ha , he ees moni o ed in he same image we e subjec ed o he same i iga ion ea men (Fig. 1). 208 These images a ee le el we e aken in he sunli side o he ees, wi h he image placed a 2 m o he 209 canopy (Fig. 1). Backg ound empe a u e was de e mined by measu ing he empe a u e o a c umpled 210 shee o aluminium oil placed close o he lea es o in e es using ε= 1 (Jones e al. 2002). To acili a e he 211 u he analysis o hese images, a cooled whi e sc een was used as backg ound, his being placed behind 212 o each moni o ed ee o simpli y he isola ion o he canopy su ace h ough image p ocessing. 213 The mal images a ee le el we e analysed wi h he so wa e de eloped by Ga cía-Teje o e al. (2012). This 214 so wa e allows o emo e hose a eas o pixels conside ed s em and he backg ound (Fig. 2). 215 Fo he case o he images aken a ow le el, hese we e analysed using he so wa e The maCam 216 Resea ch P o (Fli Sys ems, USA), selec ing a speci ic a ea on he le and on he igh and ob aining he 217 a e age alue o TC o each a ea (Fig. 3). This me hodology is much as e han he p e ious desc ibed by 218 7 Ga cía-Teje o e al. (2012), al hough i does no disc imina es he ep esen a i e a eas wi h he same 219 easibili y, and he a eas selec ion is done acco ding o he isual pe spec i e o he ope a o . 220 Conside ing he TC alues ob ained a ee le el, h ee di e en he mal indica o s we e calcula ed: he 221 di e ence be ween canopy and he su ounding ai (ΔTcanopy-ai ), he c op wa e s ess index (CWSI), and he 222 index o he ela i e s oma al conduc ance hese being calcula ed as ollows (Cos a e al., 2013): 223 224 ΔTcanopy-ai = TC - Tai (1) 225 226 𝐶𝑊𝑆𝐼 = ∆𝑇𝑐𝑎𝑛𝑜𝑝𝑦−𝑎𝑖𝑟− ∆𝑇𝑤𝑒𝑡 ∆𝑇𝑑𝑟𝑦− ∆𝑇𝑤𝑒𝑡 (2) 227 228 𝐼𝐺= ∆𝑇𝑑𝑟𝑦− ∆𝑇𝑐𝑎𝑛𝑜𝑝𝑦−𝑎𝑖𝑟 ∆𝑇𝑐𝑎𝑛𝑜𝑝𝑦−𝑎𝑖𝑟− ∆𝑇𝑤𝑒𝑡 (3) 229 whe e ΔTcanopy-ai , ∆Td y and ∆Twe a e he di e ences be ween canopy and ai empe a u e o he c op in 230 he momen o he measu emen , when he c op has he s oma a ully closed and when i is ully anspi ing, 231 espec i ely. TC is he canopy empe a u e and Tai he empe a u e o he su ounding ai . 232 To ob ain he e e ence alues o ∆Twe , he e was es ima ed he non-wa e s ess baseline (ΔTcanopy-ai = a 233 + b*VPD) acco ding o Idso e al. (1981), using a ∆Td y alue equal o 5 ºC, as i was p oposed by Jackson 234 e al. (1981). Non-wa e s ess baseline was es ima ed using he canopy empe a u e eadings ob ained 235 om ull i iga ed ees (C-100). 236 237 2.4. Expe imen al design and s a is ical analysis 238 The expe imen al design was o andomized blocks, wi h ou eplica ions pe i iga ion ea men . Each 239 eplica ion had 15 ees (3 ows and 5 ees pe ow), being moni o ed he h ee cen al ows o each 240 eplica ion (n=12). 241 Fo each measu emen day, an explo a o y desc ip i e analysis o da a (Ψlea , gs and TC) was conduc ed by 242 applying a Le ene’s es o check he a iance homogenei y o he s udied a iables. Signi ican di e ences 243 be ween i iga ion ea men s (p ≤ 0.05) in he s udied a iables we e iden i ied by applying a one-way 244 ANOVA and a Tukey’s es o ea men sepa a ion, wi h he SPSS s a is ical so wa e (SPSS Inc., 15.0 245 S a is ical package; Chicago, IL, USA). 246 To e alua e he non-wa e s ess baselines, a linea co ela ion analysis was made (n = 15). To e alua e 247 he ela ionships be ween a iables, a linea co ela ion analysis be ween he alues o he mal indica o s 248 (TC, ΔTcanopy-ai , CWSI and IG) and he c op physiological a iables (Ψlea and gs) was made, by using he 249 a e age alues o each ea men and sampling ime (n = 9). The ob ained co ela ion coe icien s we e 250 8 used o iden i y which would be he bes ime o ca y ou TC eadings and he mos ep esen a i e he mal 251 index as a p oxy o c op physiology ai s. 252 Finally, compa a i e s udy be ween he TC eadings aken a ee and ow le el was conduc ed by means o 253 a linea co ela ion analysis be ween hese alues, using he a e age alues o each ea men and he 254 whole da a ob ained du ing he wo i s daily cu es (n=30). 255 256 3. Resul s and discussion 257 3.1. Daily e olu ion o c op physiological s a us 258 Figu e 4 shows he e olu ion o Ψlea , gs, and TC measu ed a ee le el du ing he h ee daily cu es 259 de eloped du ing he i iga ion pe iod in which he wa e s ess egimes we e imposed. On o e all, as he 260 clima ic condi ions along he day became mo e ad e se, Ψlea eached mo e nega i e alues, wi h a inal 261 eco e y a he end o he day. By con as , gs inc eased du ing he i s eadings un il eaching a maximum 262 poin in which a signi ican dec ease was obse ed, his coinciding wi h he momen o he day in which he 263 clima ic condi ions we e mo e ex eme. A e his poin a sligh eco e y o gs was ound wi h he las 264 measu emen s o he day. In ela ion o TC his a iable showed a mo e dependen end on he clima ic 265 condi ions along he day, eaching he maximum alues in hose momen s in which he Tai alues we e he 266 highes . Du ing cu es 1 and 2, he lowes alues o Ψlea we e eached a 17:30 h, coinciding wi h he highes 267 VPD alues egis e ed du ing hese days; and wi h he momen s in which he TC alues we e maximum. 268 Conside ing he ob ained alues o each ea men , no di e ences we e obse ed a 8:30 h, bu hese we e 269 appea ing along he day wi hou obse ing a o al eco e y be ween he DI ea men s and C-100 a 20:00 270 h. I is ema kable ha he obse ed di e ences in e ms o Ψlea we e highe du ing he Cu e 2, his being 271 associa ed wi h he mo e se e e clima ic condi ions de ec ed and he imposed wa e es ic ion pe iod o 272 LFDI in his cu e, which had been p olonged o a u he se en days, in compa ison o Cu e 1. 273 Rega ding o gs, du ing Cu e 1, all he ea men s showed a g owing endency, eaching he maximum 274 alues a 14:30 h (VPD = 2.61 kPa). Howe e , du ing he Cu e 2, he maximum alues we e obse ed a 275 11:30 h (VPD = 1.82), om which gs dec eased, showing a pa ial eco e ing in C-100 a he end o he day. 276 This di e ence obse ed o he case o gs could be associa ed wi h he mo e se e e clima ic condi ions 277 egis e ed du ing he Cu e 2, in compa ison o he p e ious one. Finally, i is no iceable ha he deple ion 278 in TC was accompanied wi h a sligh eco e y o gs and he sligh eco e y o Ψlea and gs du ing he eadings 279 a 20:00 h. 280 Rega ding o he alues ob ained du ing he Cu e 3, i was ob ained h ee weeks a e Cu e 2, when 281 clima e condi ions we e simila o hose obse ed in he p e ious one, and LFDI was subjec ed o 15 days 282 o i iga ion es ic ion. In his sense, i was obse ed a simila end o ha de ec ed in Cu e 2, wi h he 283 highes alues o gs obse ed a 11:30 (VPD = 1.31 kPa), wi h a signi ican educ ion in all he ea men s a 284 14:30 h, ollowed by a pa ial eco e y a 17:30 h, and a new descend a he end o he day. This dec easing 285 end occu ed a 14:30 h, being his esponse associa ed wi h a pa ial s oma al closu e, when clima ic 286 condi ions, specially he VPD alues a e s ongly ele a ed. E en mo e, his descend in he alues o gs 287 9 p omo ed ha , he eadings o Ψlea be ween 11:30 and 14:30 we e simila , and he pa ial eco e y o gs a 288 17:30 was accompanied wi h a signi ican lessen alues o Ψlea . 289 Rela ing o he TC eadings, hese we e highly de e mined by he clima ic condi ions. On o e all, TC eadings 290 in he h ee s udied ea men s we e below o ai empe a u e (Tai ), excep he eadings aken a 08:30 and 291 11:30 h o he Cu es 2 and 3. The highes di e ences in TC be ween ea men s we e de ec ed specially 292 in he eadings aken a 11:30, 14:30 and 17:00, al hough hese we e no as pa en as o he case o Ψlea . 293 On o e all, and aking in o accoun he moni o ed physiological a iables, i can be assumed ha Ψlea was 294 he pa ame e ha e lec ed he highes di e ences be ween ea men s. In his sense, du ing he Cu e 1, 295 signi ican di e ences we e obse ed be ween C-100 and he emaining ea men s a 11:30 and 14:30 h, 296 wi h an ab up descend in he eadings conduc ed a 17:30 (<-2.0 MPa), wi hou di e ences be ween he 297 h ee i iga ion ea men s. Du ing he Cu e 2, he Ψlea alues egis e ed in C-100 we e signi ican di e en 298 han hose egis e ed in he emaining ea men s du ing all day (excep a 8:30 h), no being eached he 299 h eshold alue o -1.5 MPa in C-100. Finally, i also d aws a en ion ha , du ing Cu e 3, C-100 eached 300 again Ψlea alues close o -2.0 MPa, as i was i ed o he Cu e 1. 301 I is ema kable ha , whe eas Ψlea was able o show signi ican di e ences be ween ea men s, his ac 302 was no as pa en in e ms o gs, because o he low capaci y o almond o egula e he s oma al closu e 303 unde d ough condi ions. In his ega d, almond ees p esen a as eco e y o wa e po en ial, bu a delay 304 in he alues o gs as i has been s a ed by au ho s such as To ecillas e al. (1996) o Rome o e al. (2004). 305 In his line, in physiological e ms, when almond is subjec ed o a mild- o-mode a e wa e s ess si ua ion a 306 s oma al conduc ance educ ion is no as pa en as he e ec s in e ms o wa e po en ial because o i s low 307 capabili y o egula ing he s oma a when a wa e s ess si ua ion is applied, as i has been discussed by 308 some au ho s such as Wa ingue e al. (1990), Egea e al. (2011) o Eichi (2013). In his ag eemen , 309 p e iously o obse e a signi ican educ ion in gs, almond esponds wi h signi ican descends in e ms o 310 lea o s em wa e po en ial, (Ga cía-Teje o e al, 2012, 2015b). Consequen ly, almond would be able o 311 main ain accep able le els o gs (p omo ing signi ican descends in he c op-wa e po en ial) bu , keeping 312 op imum alues o ca bon assimila ion, pho osyn he ic a e, and hence inc easing he in insic wa e -use 313 e iciency (McCu chan and Shackel, 1992; Rouhi e al., 2007). 314 Gomes-La anjo e al. (2006) epo ed alues o Ψlea o di e en cul i a s, which anged be ween -1.72 and 315 -2.0 MPa in Glo ie a; -1.71 and -2.40 MPa in Fe agnes; -1.91 and -2.34 MPa in F ancoli; -1.97 and -2.26 316 MPa in Lau anne, and -1.88 and -1.92 MPa in Masbo e a. In his line, hese alues co espond o 317 measu emen s done a midday in well-wa e ed ees, which a e in line wi h he h eshold ange be ween - 318 1.5 and -2.0 MPa conside ed and ob ained in his wo k o C-100. 319 Ob iously, his wa e po en ial deple ion a ec s o lea gas exchange. In his sense, o ull i iga ed 320 condi ions, daily cycle o gas exchange is almos cons an when no adia ion limi a ion occu (To ecillas e 321 al., 1988; Klein e al., 2001; Rome o e al., 2006) and apou p essu e de ici (VPD) is no highe han 2 KPa 322 (Rome o e al., 2006). Howe e , in ou case, he alues o VPD we e highe han his alue du ing he h ee 323 cu es when he eadings we e aken be ween 11:30 and 20:00 h, which would explain he daily a ia ion 324 16 conse ación del agua en ambien es semiá idos. Tocina Es udios Locales, Re is a de In es igación 535 Local, 5, 175-190. 536 Ga cía-Teje o, I.F., Cos a, J.M., Egip o, R., Lima, R.S.N., Du án, V.H., Lópes, C., Cha es, M.M. 2016b. 537 The mal da a o moni o c op-wa e s a us in i iga ed Medi e anean i icul u e. Ag ic. Wa e 538 Manage. 176, 80-90. 539 Ga cía-Teje o, I., He nández, A., Padilla-Díaz, C.M., Diaz-Espejo, A., Fe nández, J.E. 2017. Assessing plan 540 wa e s a us in a hedge ow oli e o cha d om he mog aphy a plan le el. Ag ic. Wa e Manage, 541 188, 50-60. 542 Goldhame , D.A., Fe e es, E. 2016. Es ablishing an almond wa e p oduc ion unc ion o Cali o nia using 543 long- e m yield esponse o a iable i iga ion. I ig. Sci. pp. 1–11. doi:10.1007/s00271-016-0528-2 544 (on line i s ). 545 Goldhame , D.A., Gi ona, J. 2012. C op yield esponse o wa e : almond’ In: S edu o P, Hsiao TC, Fe e es 546 E, Raes D (eds) FAO i iga ion and d ainage Pape No. 66. Food and Ag icul u e O ganiza ion o he 547 Uni ed Na ions, Rome, pp 358–373 548 Gomes-La anjo, J., Cou inho, J.P., Galhano, V., Co dei o, V. 2006. Responses o i e almond cul i a s o 549 i iga ion: Pho osyn hesis and lea wa e po en ial. Ag ic Wa e Manage. 83, 261-265. 550 Gonzalez-Dugo, V., Za co-Tejada, P., Fe e es, E. 2014. Applicabili y and limi a ions o using he c op wa e 551 s ess index as an indica o o wa e defici s in ci us o cha ds. Ag ic. Fo . Me eo ol. 198-199, 94- 552 104. 553 Idso, S.B., Jackson, R.D., Pin e , P.J.J., Regina o, R.J., Ha field, J.L. 1981. No malizing he s ess deg ee- 554 day pa ame e o en i onmen al a iabili y. Ag ic. Me eo ol. 24, 45–55. 555 IPCC. 2014. Clima e change 2014. Impac s, adap a ions and ulne abili ies. Pa B. Regional aspec s. 556 Ba os, V & Field CB (eds). Camb idge Uni e si y P ess, 688 pp. 557 Jackson, R.D., Idso, S.B., Regina o, R.J., Pin e , P.J. 1981. Canopy empe a u e as a c op wa e -s ess 558 indica o . Wa e Resou ces Res. 17, 1133-1138. 559 Jiménez-Bello, M.A., Balles e , C., Cas el, J.R., In igliolo, D.S. 2011. De elopmen and alida ion o an 560 au oma ic e mal imaging p ocess o assessing plan wa e s a us. Ag ic. Wa e Manage. 98, 1497- 561 1504. 562 Jones, H.G. 1999. Use o in a ed he mome y o es ima ion o s oma al conduc ance as a possible aid o 563 i iga ion scheduling. Ag ic. Fo es Me eo ol. 95, 139-149. 564 Jones, H.G. 2004. I iga ion scheduling: ad an ages and pi alls o plan -based me hods. J. Exp. Bo . 55, 565 2427-2436. 566 Jones, H.G., S oll, M., San os, T., de Sousa, C., Cha es, M.M., G an , O.M. 2002. Use o in a- ed 567 he mog aphy o moni o ing s oma al closu e in he ield: applica ion o he g ape ine. J. Exp. Bo . 568 53, 2249-2260. 569 17 Jones H.G., Se aj, R., Lo eys, B.R., Xiong, L., Whea on, A., P ice, A.H. 2009. The mal in a ed imaging o 570 c op canopies o he emo e diagnosis and quan i ica ion o plan esponses o wa e s ess in he 571 ield. Func ional Plan Biology 36, 978–979. 572 Klein, I, Espa za, G., Weinbaum, S.A., DeJong, T.M. 2001. E ec s o i iga ion dep i a ion du ing he ha es 573 pe iod on lea pe sis ence and unc ion in ma u e almond ees. T ee Physiol. 21, 1063-1072. 574 López-López M., Espada o , M., Tes i, L., Lo i e, I.J., O gaz, F., Fe e es, E. 2018. Wa e use o i iga ed 575 almond ees when subjec ed o wa e de ici s. Ag ic. Wa e Manage. 195, 84-93. 576 Mi ás-A alos, J.M.; Pé ez-Sa mien o, F.; Alcobendas, R.; Ala cón, J.J.; Mounze , O.; Nicolás, E. Using 577 midday s em wa e po en ial o scheduling de ici i iga ion in mid–la e ma u ing peach ees unde 578 Medi e anean condi ions. I ig. Sci. 2016a, 34, 161–173. 579 McCu chan, H., Shackel, K.A. 1992. S em-wa e Po en ial as a Sensi i e Indica o o Wa e S ess in P une 580 T ees (P unus domes ica L. c . F ench). J. Am. Soc. Ho . Sci. 117, 607–611. 581 No es, P.A., Pé ez-Pas o , A., Egea, G., Coneje o, W., Domingo, R. 2005. Compa ison o changes in s em 582 diame e and wa e po en ial alues o de ec ing wa e s ess in young almond ees. Ag ic. Wa e 583 Manage. 77, 296–307. 584 Mölle , M., Alchana is, V., Cohen, Y., Me on, M., Tsip is, J., Nao , A., Os o sky, V., ,Sp in sin, M., Cohen, 585 S. 2007. Use o he mal and isible image y o es ima ing c op wa e s a us o i iga ed g ape ine. 586 J. Exp. Bo . 58, 827–838. 587 Olesen, J.E., T nka, M., Ke sebaum, K.C., Skjel åg, A.O., Seguin, B., Pel onen-Sainio, P., Rossi, F., Kozy a, 588 J., Micale, F. 2011, Impac s and adap a ion o Eu opean c op p oduc ion sys ems o clima e change. 589 Eu op. J. Ag on. 34, 96-112. 590 Phoga , V., Skewes, M.A., McCa hy, M.G., Cox, J.W., Simunek, J., Pe ie, P.R. 2017. E alua ion o c op 591 coe icien s, wá e po oduc i i y and wá e balance componen s o wine g apes ii iga ed a di e en 592 de i i le els by a sub-su ace d ip. Ag ic. Wa e Manage. 180, 22-34. 593 Poble e-Eche e ía, C., O ega-Fa ías, S., Zúñiga, M. 2014. Use o in a ed he mog aphy on canopies as 594 indica o o wa e s ess in A bequina oli e o cha ds. Ac a Ho , 1057, 399-404. 595 Poble e-Eche e ía, C., Sepúl eda-Reyes, D., O ega-Fa ias, S., Zúñiga, M., Fuen es, S. 2016. PLan wá e 596 s ess de ec ion based on ae ial and e es ial in e ed he mog aphy: a s udy case om ineya d 597 and oli e o cha d. Ac a Ho 1112, 141-146 598 Pou, A., Diago, M.P., Med ano, H., Baluja, J., Ta daguila, J. 2014. Valida ion o he mal indices o wa e 599 s ess s a us iden i ica ion in g ape ine. Ag ic. Wa e Manage. 134, 60-72. 600 Pue o, P., Domingo, R., To es, R., Pé ez-Pas o , A., Ga cía-Riquelme, M. 2013. Remo e managemen o 601 de ici i iga ion in almond ees base on maximum daily unk sh inkage. Wa e ela ions and yield. 602 Ag ic. Wa e Manage. 126, 33-45. 603 Remo ini, D., Massai, R. 2003. Compa ison o wa e s a us indica o s o young peach ees. I ig. Sci. 22, 604 39-46. 605 18 Rouhi, V., Samson, R., Lemeu , R., Van Damme, P. 2007. Pho osyn he ic gas exchange cha ac e is ics in 606 h ee di e en almond species du ing d ough s ess and subsequen eco e y. En i on. Exp. Bo . 607 59, 117–129. 608 Rome o, P., Bo ía, P., Ga cía, F. 2004. E ec s o egula ed de ici i iga ion unde subsu ace d ip i iga ion 609 condi ions on ege a i e de elopmen and yield o ma u e almond ees. Plan and Soil 260, 169– 610 181. 611 Rome o, P., Ga cía, J., Bo ía, P. 2006. ‘Cos –bene i analysis o a egula ed de ici -i iga ed almond o cha d 612 unde subsu ace d ip i iga ion condi ions in Sou h-eas e n Spain’, I ig. Sci. 24, 175–184. 613 Shackel, K. 2011. A plan -based app oach o de ici i iga ion in ees and ines. Ho Sci. 46, 173–177. 614 Spinelli, G.M., Snyde , R.L., Sanden, B.L., Shackel, K.A. 2016. Wa e s ess causes s oma al closu e bu 615 does no educe canopy e apo anspi a ion in almond. Ag ic. Wa e Manage. 168, 11-22. 616 Tes i, L., Goldhame , D.A., Inies a, F., Salinas, M. 2008. C op wa e s ess index is a sensi i e wa e s ess 617 indica o in pis achio ees. I ig. Sci. 26, 395–405. 618 To ecillas, A., Ruiz-Sanchez, M.C., del Amo , F., León, A. 1988. Seasonal a ia ions on wa e ela ions o 619 Amygdalus communis L. unde d ip i iga ed and non i iga ed condi ions. Plan and Soil. 106, 215- 620 220. 621 To ecillas, A., Ala cón, J.J., Domingo, R., Planes, J., Sánchez-Blanco, M.J. 1996. S a egies o d ough 622 esis ance in lea es o wo almond cul i a s. Plan Sci. 118, 135–143. 623 USDA. 2010. Keys o soil axonomy (11 h Edi ion). Uni ed S a es Depa men o Ag icul u e, Na u al 624 Resou ce Conse a ion Se ice, 334 pp. 625 Wa inge , A., Heilmeie , H., Ha ung, W., Schulze, E.D. 1990. Daily and seasonal cou ses o lea 626 conduc ance and abscisic acid in he xylem sap o almond ees (P unus dulcis (Mille ) D.A.Webb) 627 unde dese condi ions. New Phy ol. 116, 581–587. 628 Za co-Tejada, P.J., Be ni, J.A.J., Suá ez, L., sepulc é-Can ó, G., Mo ales, F., Mille , J.R. 2009. Imaging 629 chlo ophyll luo escence wi h an ai bone na ow-band mul ispec al came a o ege a ion s ess 630 de ec ion. Remo e Sens. En i on. 113, 1262-1275. 631 632 633 634 635 636 637 638 19 639 640 FIGURES 641 642 643 644 Figu e 1. Example o he mal images a plan (le ) and ow ( igh ) le el 645 646 647 648 Figu e 2. Example o image p ocessing using he so wa e de eloped by Ga cía-Teje o e al (2012). 649 On he igh , he ini ial he mal imaging; on he le , a bi map image, in which he black a ea 650 ep esen s he pixels o he he mal image conside ed o calcula e he canopy empe a u e. 651 652 653 654 Figu e 3. Example o image p ocessing a ow le el using he The maCam Resea ch P o (Fli 655 Sys ems, USA). 656 657 20 658 Figu e 4. Daily cu es o lea wa e po en ial (Ψlea ), s oma al conduc ance (gs) canopy empe a u e 659 (Tc) and ai empe a u e (Tai ) in almond ees subjec ed o di e en i iga ion doses: C-100, ull 660 i iga ed ea men ; RDI-50, egula ed de ici i iga ion; LFDI, low- equency de ici i iga ion. Le e s 661 a, b, and c show signi ican di e ences be ween C-100, RDI-50 and LFDI ea men s, espec i ely 662 (p<0.05). 663 664 665 666 21 Figu e 5. Non-wa e s ess baseline (∆Tcanopy-ai ) = a*VPD + b. Da a ob ained o he DOYs 210, 217 667 and 239 and using he eadings aken a 8:30, 11:30, 14:30, 15:30 and 20:00. 668 669 670 671 Figu e 6. Rela ionships be ween canopy empe a u e eadings (TC), he di e ence be ween 672 canopy and ai empe a u e (ΔTcanopy-ai ) and c op wa e s ess index (CWSI) wi h lea -wa e 673 po en ial (Ψlea ). 674 675 676 Figu e 7. Rela ionships be ween canopy empe a u e eadings (TC) a ee and ow le el. 677 678 679 680 681 682 22 TABLES 683 684 Table 1. Clima ic condi ions, wa e equi emen s and i iga ion doses applied du ing he season 685 Pe iod (DOY) Tai (ºC) RH (%) Rain all (mm) ET0 (mm) KC ETC (mm) C-100 (mm) RDI-50 (mm) LFDI (mm) 60 o 90 13.53 70.18 55.6 92.11 0.3 27.63 0 0 0 91 o 120 17.95 72.96 35.6 114.84 0.55 34.74 8.03 8.03 8.031 121 o 151 21.42 54.06 12.6 175.79 0.9 118.66 109.21 109.21 109.21 152 o 181 23.40 58.88 7.4 176.94 1.05 167.21 161.66 108.32 107.88 182 o 212 25.24 58.81 0.2 184.24 1.15 190.69 190.54 91.46 77.85 213 o 243 26.05 54.06 0 173.82 1.15 179.90 179.90 88.15 79.56 244 o 273 22.66 78.06 175.8 104.99 0.8 75.59 23.02 23.02 23.02 274 o 304 19.93 77.67 73.2 79.86 0.7 50.31 12.5 12.5 12.5 DOY. day o he yea ; Tai . a e age ai empe a u e; RH. a e age ela i e humidi y. ET0. e e ence 686 e apo anspi a ion; KC. c op coe icien ; ETC. c op e apo anspi a ion; C-100. con ol ea men ; SDI-50. 687 egula ed de ici i iga ion a 50% o ETC du ing he ke nel illing pe iod; LFDI. low- equency de ici i iga ion 688 du ing he ke nel illing pe iod. 689 690 691 Table 2. A e age alues o ai empe a u e (Tai ). ela i e humidi y (RH) and apou p essu e de ici 692 (VPD) egis e ed du ing he daily cu es 693 Cu e 1 (210 DOY) Cu e 2 (217 DOY) Cu e 3 (239 DOY) Hou Tai (ºC) RH (%) VPD (kPa) Tai (ºC) RH (%) VPD (kPa) Tai (ºC) RH (%) VPD (kPa) 08:30 24.4 63 0.94 16.6 63 0.70 21.3 60 1.01 11:30 25.3 69 0.99 27.4 50 1.82 27.4 64 1.31 14:30 34.1 51 2.61 33.4 45 2.82 37.5 31 4.43 17:30 34.5 42 2.29 42.1 21 6.49 41.0 25 5.81 20:00 31 41 1.84 34.1 27 3.90 38.8 26 5.10 Tai . a e age ai empe a u e; RH. a e age ela i e humidi y; VPD. apou p essu e de ici 694 695 696 697 698 699 700 701 702 703 23 Table 3. Pea son’s co ela ion coe icien s be ween he mal in o ma ion and he s udied 704 physiological a iables 705 Hou TC ∆Tcanopy-ai CWSI IG 8:30 gs -0.32* ns -0.40* ns Ψlea ns ns ns ns 11:30 gs ns ns ns ns Ψlea -0.85** -0.69* -0.85** ns 14:30 gs -0.70* ns -0.82** ns Ψlea -0.39* ns -0.69* ns 17:30 gs ns -0.70* -0.62* ns Ψlea -0.39* ns -0.34* 0.74** 20:00 gs -0.75* ns ns ns Ψlea ns ns ns ns TC. canopy empe a u e; ∆Tcanopy-ai . di e ence be ween canopy and ai empe a u e; CWSI. c op- 706 wa e s ess index; IG. ela i e index o s oma al conduc ance; gs. s oma al conduc ance o wa e 707 apou ; Ψlea . lea -wa e po en ial in shaded lea es. * and ** show signi ican ela ionships a 708 con idence le el o 95 and 99%. espec i ely. 709 710 711 712 713 714 715 716 717 718 719 720 721 9 Table 4. Example o alse-colou ed images aken a ee and ow le el du ing he Cu e 1 in he di e en i iga ion ea men s and momen o he day. The alues o canopy empe a u e (TC) co espond o he a e age o i e measu emen s aken o each ea men and momen o he day. T ee le el Row le el C-100 RDI-50 LFDI C-100 RDI-50 LFDI 08:30 TC (ºC) 21.7 21.6 21.4 21.3 21.3 21.4 11:30 TC (ºC) 23.3 23.6 23.4 23.5 23.6 23.4 14:30 TC (ºC) 26.9 27.0 29.9 27.4 28.6 29.1 17:30 TC (ºC) 32.5 30.9 30.0 32.3 30.7 31.0 20:00 TC (ºC) 29.2 28.6 28.9 29.1 28.6 29.0 10 Table 5. Example o alse-colou ed images aken a ee and ow le el du ing he Cu e 2 in he di e en i iga ion ea men s and momen o he day. The alues o canopy empe a u e (TC) co espond o he a e age o i e measu emen s aken o each ea men and momen o he day. T ee le el Row le el C-100 RDI-50 LFDI C-100 RDI-50 LFDI 08:30 TC (ºC) 20.9 20.5 22.2 21.2 21.2 21.4 11:30 TC (ºC) 27.4 28.6 29.1 27.6 28.7 29.0 14:30 TC (ºC) 30.2 32.5 32.9 30.7 32.8 33.1 17:30 TC (ºC) 31.8 32.2 31.1 30.5 30.5 31.2 20:00 TC (ºC) 31.1 31.5 31.9 31.3 31.7 32.2