Identification of the water stress level in olive trees during pit hardening using the trunk growth rate indicator.

1 Identification of the water stress level in olive trees during pit hardening using the trunk growth rate indicator. Corell, M1, Martín-Palomo, MJ1, Girón, I2, Andreu, L.1, Trigo, E1, López-Moreno, YE1, Torrecillas, A3, Centeno, A4, Pérez-López, D4, Moriana, A1 1Universidad de Sevilla, Departamento de Ciencias Agroforestales, 2Consejo Superior de Investigaciones Científicas, IRNAS, 3Consejo Superior de Investigaciones Científicas, CEBAS. 4Universidad Politécnica de Madrid, Departamento de Producción Vegetal:Fitotecnia ABSTRACT Water scarcity is generating an increasing interest in deficit irrigation scheduling. The 1 trunk diameter fluctuations are daily cycles that have been suggested as tools for irrigation 2 scheduling. The trunk growth rate (TGR) was suggested as the best indicator for olive 3 trees during pit hardening. The aim of this work is to clarify how the TGR could be used 4 to identify water stress levels. The experiment was performed during the 2017 season, in 5 a commercial, super-high-density orchard in Carmona (Seville, Spain). Four different 6 irrigation treatments were performed according to midday stem water potential values 7 and TGR. The data obtained were very variable and both indicators presented a wide 8 range of water status throughout the season. The maximum trunk diameter data clearly 9 showed the pattern of the trees water status but the comparison between treatments and 10 the identification of the water stress level was not possible. The average TGR was linked 11 to the midday stem water potential, but with a minimum amount of data. Irrigation 12 scheduling based on the average TGR was difficult because of the great increases in some 13 2 daily TGR values. For clarity, the pool of data was grouped by midday stem water 14 potential. These water stress levels were characterized using the weekly frequency of 15 TGR values. The increase of water stress reduced the frequency of values between -0.1 16 and 0.3mm day-1 from 60% to less than 25%. Moderate water stress levels increased the 17 percentage of values lower than -0.3mm day-1 from 7% to 37%. The most severe water 18 stress conditions increased the TGR values between -0.3 and -0.1mm day-1 from 16% up 19 to 22%. 20 Keywords: Regulated deficit irrigation, trunk diameter fluctuation, water relations. 21 3 INTRODUCTION 22 Deficit irrigation has been traditionally scheduled using a percentage of the crop 23 evapotranspiration (Behobudian and Mills, 1997). For most fruit trees, this approach 24 allowed knowing the yield effect of water stress based on the moment when this 25 restriction is applied (Behobudian and Mills, 1997). However, the trees response to 26 irrigation restrictions depends on the phenological stage (when), water stress level (how), 27 and duration (how long) of the water stress (Hsiao, 1990). When these three factors are 28 considered, regulated deficit irrigation (RDI) is actually performed and discrepancies 29 between works can be explained (i.e. peaches, Girona, 2002). In the last years, several 30 irrigation works have focused on the different water stress indicators reported and even 31 suggested specific irrigation scheduling in fruit trees considering these tools (Steduto et 32 al 2012, García-Tejero and Durán-Zuazo, 2018). But the environmental effect on most of 33 these indicators limits their suitability for application in commercial orchards. 34 Irrigation restrictions during pit hardening in olive trees have no yield effect under 35 moderate water stress conditions (Goldhamer, 1999). The fruit development is affected 36 by a midday stem water potential (SWP) lower than -2MPa during this period, although 37 the fruit size is recovered after adequate rehydration (Girón et al 2015). Severe water 38 stress levels, with midday water potential values lower than -4MPa, had a limiting yield 39 effect (Moriana et al 2003; Iniesta et al 2009; Fernández et al 2013, among others). 40 Therefore, there are SWP data suggesting suitable water stress levels for this fruit tree. 41 However, the SWP is not an automatic continuous tool and its usefulness is limited in the 42 smart agriculture. Trunk diameter fluctuations are daily cycles of swelling and shrinkage 43 that provide several indicators, reported as early signals of water stress in olive trees 44 (Moriana and Fereres, 2002). The trunk growth rate (TGR), the difference between two 45 consecutive daily peaks (Goldhamer et al, 1999), has been reported as the only indicator 46 4 really useful in olive irrigation scheduling (Moriana and Fereres, 2002, Moriana et al 47 2010, Fernández et al 2011). The typical seasonal pattern of maximum diameter in fully- 48 irrigated mature olive trees is an increase before pit hardening and an almost constant 49 value thereafter (Moriana et al 2003). This is reflected in positive TGR values before pit 50 hardening and almost null thereafter. Daily TGR values, however, even in fully-irrigated 51 conditions, are extremely changeable and difficult to understand. Several works reported 52 the average TGR during pit hardening as an indicator of water stress (Moriana et al 2013; 53 Girón et al 2015). But the average TGR does not benefit from the main advantage of this 54 measure, the continuous monitoring. Girón et al (2016) suggested that part of this great 55 variation in daily TGR is related to large variations of the daily vapour pressure deficit 56 (VPD), and this relationship could be inversely proportional. A large reduction in VPD 57 between days increases the TGR value. Recently, Corell et al (2017) reported that only 58 daily TGR data lower than -0.1mm day-1 were related to water stress conditions. In that 59 work, authors suggested that the maximum diameter picture indicates the water stress 60 pattern but the average TGR for a period and the daily TGR would be the indicators to 61 consider for irrigation scheduling (Corell et al 2017). Using these results, the aim of this 62 work was to schedule regulated deficit irrigation based on average and daily TGR values. 63 The daily TGR would show the moment when the water stress starts, while the average 64 TGR would be the target value to establish different levels of water stress. 65 MATERIAL AND METHODS 66 Site description and experimental design 67 The experiment was conducted during 2017 at the commercial farm "Morillo" located in 68 Carmona (37.49ºN, -5.67ºW, Seville, Spain). The olive (Olea europaea L cv Arbequina) 69 orchard is a super-high density (4*1.5m) plantation, with 11 years of age at the beginning 70 of the experiment. Trees were irrigated each other day with one line of drips (3.4l h-1) 71 5 separated 0.4m. The effective depth of the soil is very variable between plots and went 72 from 0.4 to more than 1m. The soil texture is sand-loam, with a high percentage of 73 carbonate (higher than 25%) and pH around 8.4. The organic matter at 0-40cm is around 74 1.8%, with an adequate level of P2O5 and K2O. 75 The statistical design included randomized complete blocks with 4 repetitions and 76 4 irrigation treatments. The experimental plot included 3 rows of around 20 trees each, 77 and the measurements were obtained from the central row. The amount of water applied 78 was measured with a water meter in each plot. The irrigation treatments were defined 79 based on the pit hardening period. The beginning of the pit hardening period was 80 estimated according to Rapoport et al (2013) around day of the year (DOY) 160. To 81 summarise, this date was defined as the moment when a change in the slope of the 82 longitudinal fruit growth was measured. The end of pit hardening period was on DOY 83 245. This date was selected considering the possible harvest day (early November). The 84 irrigation treatments were: 85  Control. Trees were irrigated to obtain an optimum water status 86 throughout the season with around 100% ETc (crop evapotranspiration). 87 Irrigation problems reduced the amount of water applied in some plots 88 from day of the year (DOY) 199 to 208, with a reduction in midday stem 89 water potential. 90  Sustained deficit irrigation (SDI). Irrigation was scheduled to distribute 91 150mm throughout the season. During pit hardening the amount of water 92 supplied was almost constant and, after this period, it was reduced slightly. 93  Regulated deficit irrigation 1 (RDI-1). Irrigation was scheduled using 94 the midday stem water potential (SWP) and the trunk growth rate (TGR). 95 Before pit hardening, water was applied only when SWP values lower than 96 6 -1.2MPa were measured. During pit hardening, this threshold of SWP was 97 lowered to -2MPa. In this period, the TGR values were considered 98 following the Corell et al (2017) and Girón et al (2015) recommendations. 99 Thus, two indicators were considered: daily TGR and average TGR. The 100 average TGR was calculated using all the data available for each tree 101 during pit hardening. Water was applied when the daily TGR was lower 102 than -0.1mm day-1 (according to Corell et al 2017) and the average TGR 103 was lower than -0.031mm day-1 (double the value suggested by Girón et 104 al 2015). From pit hardening, trees were irrigated when the TGR average 105 was lower than -0.016mm day-1 (according to Girón et al 2015). The TGR 106 average was calculated with all the data from pit hardening. 107  Regulated deficit irrigation 2 (RDI-2). Irrigation scheduling as RDI-1 108 but with a higher level of water stress during pit hardening and limited 109 seasonal amount of water (150 mm). During pit hardening, the SWP 110 threshold was -3MPa and the average TGR was -0.062mm day-1. 111 The amount of water applied was estimated taking into account the difference between 112 the threshold considered and the value measured. The trees were not irrigated if the 113 difference was lower than 10%. The maximum irrigation rate was 4mm day-1, which was 114 the maximum crop evapotranspiration estimated for the average season. This value was 115 applied when differences were higher than 30%. When they were between 20-30%, the 116 irrigation was 2mm day-1 and just 1mm day-1 for the interval 10-20%. 117 Meteorological conditions throughout the experiment 118 Weather data during the season were obtained from the "Villanueva de Rio y Minas" 119 station in the Andalusian Weather Stations Network (Fig. 1). This station is located 120 approximately 25km away from the experimental orchard. Data during 2017 were typical 121 7 of Mediterranean zones, with null rainfall during the summer period and warm winters. 122 Maximum potential evapotranspiration values (ETo), higher than 6mm day-1, were 123 measured from the end of Spring until mid-August. The average ETo during the pit 124 hardening period (DOY 160-245) was 6.18mm day-1 with null rainfall. During phase I, 125 extending from shoot sprouting until pit hardening (DOY 46-160), the average ETo was 126 4mm day-1 and the total rainfall was 136.5mm. However, during the recovery period, 127 from pit hardening until the 1st of November, rainfall was very scarce, 25.6mm, while 128 the ETo was still high with an average of 3.7mm day-1. The total rainfall this year was 129 very low, 277.9mm, considering the seasonal average (539mm, AEMET, 2018). 130 Measurements 131 The water relations of the trees were studied in combination with the soil moisture, leaf 132 gas exchange and midday stem water potential measurements. The soil moisture was 133 measured with FDR sensors (Echo20 HS10, Decagon Device, USA). Measurements were 134 made in four plots per treatment. The FDR probes were placed in the irrigation line, about 135 30cm from an emitter (Fernández et al., 1991). Data were obtained at 0.2m and 0.4m 136 depth. The leaf gas exchange was measured with the midday leaf net photosynthesis using 137 an infrared gas analyser (CI-340, CID BioScience, USA) in one fully expanded sunny 138 leaf per tree. The water potential was measured at midday in one leaf per tree, using the 139 pressure chamber technique (Scholander et al., 1965). The leaves near the main trunk 140 were covered with aluminium bags at least one hour before measurements were taken and 141 a pressure pump was used (PMS model 1000). 142 Trunk diameter fluctuations are a daily cycle of shrinkage and swelling in which 143 different indicators can be estimated. The most common ones are the maximum daily 144 shrinkage (MDS) and the trunk growth rate (TGR) (Ortuño et al., 2010). The MDS values 145 are not an early indicator of water stress (Moriana and Fereres, 2002) and only the TGR 146 8 values were considered. The TGR is the difference between two consecutive daily 147 maximums (Goldhamer et al. 1999), the TGR on day “n” is the difference between the 148 maximum daily diameter for day “n+1” and for day “n”. 149 The trunk diameter fluctuations were measured in one tree per repetition using a 150 band dendrometer (5m accuracy, D6, UMS, Germany) attached to the main trunk. The 151 band dendrometer works like a beam when bending. The trunks were measured using the 152 nodes of a wireless sensor with a network topology for easy installation and maintenance. 153 The band rested on a part of the trunk surface. The ends of the band were joined with 154 Invar steel, an alloy of Ni and Fe with a thermal expansion coefficient close to zero 155 (Katerji et al., 1994), the band circled the trunk. A Teflon net below the steel prevented 156 friction with the bark surface. Each band dendrometer was plugged into a node (Widhoc 157 smart solution SL, Spain) near the sensor. These nodes were integrated by two different 158 parts. One being the measurement interface, and the other the processing, recording and 159 communication system. The nodes generated a stabilized power supply of 10Vdc to the 160 band dendrometer. The data from each sensor node were sent wirelessly to cloud. Ten 161 measurements of each band dendrometer were taken every hour. 162 Data analyses were performed with ANOVA and the mean separation was made 163 using a Tukey’s test with the Statistix (SX) program (8.0). Significant differences were 164 considered when p-level<0.05 in both tests. Calculations of the p-level were performed 165 considering the F-test of variance equality. When conditions of variance equality could 166 not be obtained, a decrease in the degree of freedom and, therefore, a more restrictive p- 167 value was calculated. The number of samples measured is specified in the text and figures. 168 169 RESULTS AND DISCUSSION 170 9 The pattern of water applied and the crop evapotranspiration (ETc) are shown in Fig. 2. 171 The amount of water applied in the Control treatment was clearly different from the rest 172 of treatments in the experiment. In Control trees, slightly lower values than the calculated 173 ETc were applied from DOY 160 to 220. However, such reductions were lower than 20% 174 when maximum differences occurred (between DOY 199 to 208). From DOY 220, the 175 Control irrigation was higher than 100% ETc and reached a maximum value of 176 approximately 500mm. The water applied in deficit treatments was very similar and the 177 main differences occurred from the mid-pit hardening period. From DOY 200, until the 178 end of this period, the water applied to RDI 2 was slightly lower than RDI 1 and SDI. 179 After pit hardening, both RDI 1 and 2 received a higher amount of water, at similar rates; 180 while SDI received a clearly lower amount in comparison. Overall, the water applied in 181 RDI 2 and SDI was almost the same, around 150mm, while RDI 1 received a slightly 182 higher amount with 185mm. 183 The soil moisture at 0.2 (Fig. 3a) and 0.4m (Fig. 3b) was very changeable 184 throughout the experiment and it presented very high variability within treatments. No 185 significant differences were found throughout the experiment at any of the depths. The 186 trends of the treatments were very erratic and there were no clear patterns in any of the 187 depths considered. The data varied from 0.25 to 0.4m3 m-3. The Control data were the 188 most constant with typical maximum values from DOY 230. In this treatment, there was 189 a clear reduction of soil humidity in the period DOY 200-230. For the rest of treatments, 190 only RDI 1 at 0.4m depth presented data generally lower than Control until the end of the 191 experiment from DOY 160. 192 The pattern of midday stem water potential (SWP) is showed in Figure 4. Before 193 pit hardening (DOY 160), the SWP values were very similar and higher than -1.5MPa. In 194 the period of pit hardening (from DOY 160 until 243), the maximum level of water stress 195 16 demand (Girón et al 2016). The olive tree physiology is closely related to the evaporative 345 demand. The daily evolution of leaf conductance is linked to the daily pattern of VPD 346 (Angelopoulos et al., 1996). These variations on the trunk diameter could be related to 347 general modifications of the tree water. Díaz-Espejo et al (2018) suggest that under no 348 water stress conditions, olive trees regulate the water status in order to minimize the leaf 349 dehydration using an isohydric response. In this way, the water in the trunk reservoir 350 could be involved in part of this regulation. The trunk diameter variations are produced 351 mainly by changes in the water content and growth of the bark (Brough et al., 1986) and 352 they are considered a water reservoir in the tree (Simonneau et al., 1993). On the other 353 hand, water stress conditions could be identified with negative values below -0.1mm day- 354 1 of daily TGR as Corell et al (2017) reported. However, the greatest negative daily TGR 355 were measured under moderate water stress level (level 2 and 3) instead of severe 356 conditions. These high rates of dehydration in the tree are in agreement with the high 357 capacity of dehydration reported in the literature for this species (among others, 358 Angelopoulos et al., 1993, Moriana et al. 2003, Iniesta et al., 2009, Díaz-Espejo et al 359 2018). According to Díaz-Espejo et al (2018), at these moderate water stress level (from 360 -1.4 to -4MPa) the hydraulic conductivity of the tree is almost unaffected, and the 361 stomatal closure and osmotic adjustment are the more intensive physiological responses. 362 This partial embolism resistance of olive trees has been partially related to extraxylematic 363 components (Díaz-Espejo et al 2018) which could be associated to the TGR variation. 364 CONCLUSIONS 365 The trunk growth rate (TGR) provided information about the moment when the water 366 stress occurs and the level of the water stress. The different parameters used did not 367 always present clear or complete information. The maximum trunk diameter presented 368 clear information about the pattern of the irrigation treatments and it help identify 369 17 conditions of water stress, but not the level of water stress. The comparison between 370 treatments was not possible only with the maximum diameter data. The average TGR was 371 related to the midday stem water potential but this indicator alone is not a useful irrigation 372 scheduling tool because a minimum number of data is needed (i.e. weekly in the present 373 work). The average TGR evaluates a period of irrigation but it is difficult to use for daily 374 irrigation scheduling as the water stress target level. Daily TGR values were very 375 changeable and the same midday water potential measurement was associated to very 376 different daily TGRs. However, all these changes can help explain the water status of the 377 trees. Frequency and absolute values of daily TGR are tools for continuous irrigation 378 scheduling. 379 Acknowledgement 380 This research was supported by the Agencia Española de Investigación (AEI) and the 381 Fondo Europeo de Desarrollo (FEDER) project AGL2016-75794-C4-4-R. Authors thank 382 to Finca La Florida for their helpful assistant during the experiment, specially to Mr Javier 383 León. 384 385 References 386 AEMET (Meteorological Spanish Agency). 2018. 387 (http://www.aemet.es/es/serviciosclimaticos/datosclimatologicos/valoresclimatol 388 ogicos). 389 Archer, P, Cohen, M, Ameglio, T, Valancogne, C, Ant¢n, A 1997. Trunk diameter 390 variations in relation to walnut water potential. III International Symposium on 391 sensors in horticulture. Tiberiades. Israel. pp17-22. 392 18 Behboudian, MH and Mills, TM 1997. Deficit irrigation in deciduous orchards. 393 Horticultural Reviews 21, 105-131 394 Brough, DW, Jones, HG, Grace, J. 1986. Diurnal changes in water content of the stems 395 of apple trees, as influenced by irrigation. Plant, Cell Environ. 9, 1-7. 396 Corell, M, Martín-Palomo, MJ, Pérez-López, D, Centeno, A, Girón, I, Moreno, F, 397 Torrecillas, A, Moriana, A. 2017. Approach for using trunk growth rate (TGR) in 398 the irrigation scheduling of table olive orchards. Agric Water Manage 192, 12-20 399 Fernández, J.E., Moreno, F., Cabrera, F., Arrue, J.L., Martín-Aranda, J. 1991. Drip 400 irrigation, soil characteristics and the root distribution and root activity of olive 401 trees. Plant Soil. 133, 239-251. 402 Fernández, E., Moreno, F., Martín-Palomo, M.J., Cuevas, M.V., Torres-Ruiz, J.M., 403 Moriana, A. 2011. Combining sap flow sensor and trunk diameter measurements 404 to assess water needs in mature olive orchards. Environ Exp Bot 72,330-338 405 Fernández, JE, Pérez-Martín, A, Torres-Ruiz, JM, Cuevas, MV, Rodriguez-Dominguez, 406 CM, Elsayed-Farg, S, Morales-Sillero, A, Garcia, JM, Hernandez-Santana, V, 407 Diaz-Espejo, A. 2013. A regulated deficit irrigation strategy for hedgerow olive 408 orchards with high plant density. Plant Soil 372,279-295. 409 Garcia-Tejero, I and Duran-Zuazo, VH. 2018. Water scarcity and sustainable agriculture 410 in semiarid environment. Tools, strategies and challenges for woody crops. 411 Academic Press, Elsevier, London. 412 Girón, I.F., Corell, M., Martín-Palomo, M.J., Galindo, A., Torrecillas, A., Moreno, F., 413 Moriana, A. 2015. Feasibility of trunk diameter fluctuations in the scheduling of 414 regulated deficit irrigation for table olive trees without references trees. Agric. 415 Water Manage. 161: 114-126. 416 19 Girón, IF; Corell, M; Martín-Palomo, MJ; Galindo, A.; Torrecillas, A.; Moreno, F; 417 Moriana, A. 2016. Limitations and usefulness of maximum daily shrinkage 418 (MDS) and trunk growth rate (TGR) indicators in the irrigation scheduling of table 419 olive trees. Agric. Water Manage 164,38-45 420 Girona, J. 2002. Regulated deficit irrigation in Peach. A global analysis. Acta Hort. 592, 421 335-342. 422 Goldhamer, D.A., 1999. Regulated deficit irrigation for California canning olives. Acta 423 Hortic. 474, 369-372. 424 Goldhamer, D.A., Fereres, E., Mata, M., Girona, J., Cohen, M., 1999. Sensitivity of 425 continuous and discrete plant and soil water status monitoring in peach tress 426 subjected to deficit irrigation. J. Amer. Soc. Hort. Sci. 124, 437-444. 427 Hsiao, TC. 1990. Measurements of plant water status. Chap. 9. In: Irrigation of 428 agricultural crops. (Eds: Steward,BA; Nielsen,DR) (Agronomy Monograph, 30.) 429 American Society of Agronomy, Madison, pp 243-279. 430 Iniesta, F, Testi, L, Orgaz, F, Villalobos, F. 2009. THe effects of regulated and continuous 431 deficit irrigation on the water suse, growth and yield of olive trees. Europ. J. 432 Agronomy 30, 258-265 433 Katerji, N., Tardieu, F., Bethenod, O., Quetin, P., 1994. Behavior of Maize stem diameter 434 during drying cycles: comparison of two methods for detecting water stress. Crop 435 Sci. 34, 165-169. 436 Moriana, A. and Fereres, E. 2002. Plant indicators for scheduling irrigation for young 437 olive trees. Irrig. Sci. 21, 83-90 438 Moriana, A., Orgaz, F., Fereres, E., Pastor, M., 2003. Yield responses of a mature olive 439 orchard to water deficits. J. Amer. Soc. Hort. Sci. 128: 425-431. 440 20 Moriana, A; Girón, I; Martín-Palomo, MJ; Conejero, W; Ortuño; MF; Torrecillas, A; 441 Moreno, F. 2010. New approach for olive trees irrigation schedulings using trunk 442 diameter sensors. Agric. Water Manage 97:1822-1828. 443 Moriana, A., Pérez-López, D., Prieto, M.H., Ramírez-Santa-Pau, M., Pérez-Rodríguez, 444 J.M., 2012. Midday stem water potential as a useful tool for estimating irrigation 445 requirements in olive trees. Agric. Water Manage. 112: 43-54. 446 Moriana, A.; Corell, M; Girón, IF; Conejero, W.; Morales, D.; Torrecillas, A.; Moreno, 447 F. 2013. Regulated deficit irrigation based on threshold values of trunk diameter 448 fluctuation indicators in table olive trees. Sci Hort 164,102-111 449 Ortuño, M.F., Conejero, W., Moreno, F., Moriana, A., Intrigliolo, D.S., Biel, C., 450 Mellisho, C.D., Pérez-Pastor, A., Domingo, R., Ruiz-Sánchez, M.C., Casadesus, 451 J., Bonany, J., Torrecillas, A., 2010. Could trunk diameter sensors be used in 452 woody crops for irrigation scheduling?. A review of current knowledge and future 453 perspectives. Agric. Wat. Management 97,1-11. 454 Rapoport, H.F., Pérez-López, D., Hammami, S.B.M., Aguera, J., Moriana, A., 2013. Fruit 455 pit hardening: physical measurements during olive growth. Ann. Appl. Biol. 163, 456 200-208. 457 Scholander, P.F., Hammel, H.T., Bradstreest, E.A., Hemmingsen, E.A., 1965. Sap 458 pressure in vascular plant. Science 148, 339-346. 459 Simonneau,T; Habib,R; Goutouly,JP; HUguet,JG (1993): Diurnal changes in stem 460 diameter depend upon variations in water content: direct evidence in peach trees. 461 J. Exp. Bot. 44, 615-621. 462 Steduto, P, Hsiao, TC, Fereres, E, Raes, D. 2012. Crop yield response to water. FAO 463 irrigation and drainage paper nº 66. FAO, Rome. 464 465 21 Figure captions 466 Fig. 1. Seasonal pattern of potential evapotranspiration (ETo) and rainfall during 2017 467 season. Vertical lines limit the period of pit hardening. Source: Villanueva Rio y Minas 468 station. Andalusian Climatic Network. 469 Fig. 2. Seasonal pattern of rain, estimated crop evapotranspiration and water applied in 470 each treatment. Each point is the average of 4 values. Vertical bars represent the standard 471 error. Vertical lines limit the pit hardening period. 472 Fig.3. Seasonal pattern of soil moisture at 20 and 40cm depth. Each point is the average 473 of 4 values. Vertical bars represent the standard error. Vertical lines limit the pit hardening 474 period. 475 Fig. 4. Seasonal pattern of midday stem water potential. Each point is the average of 4 476 values. Vertical bars represent the standard error. Vertical lines limit the pit hardening 477 period. Asterisks indicate significant differences (p<0.05, Tukey Test). 478 Fig. 5. Seasonal pattern of midday net photosynthesis rate. Each point is the average of 479 4 values. Vertical bars represent the standard error. Vertical lines limit the pit hardening 480 period. Asterisks indicate significant differences (p<0.05, Tukey Test). 481 Fig. 6. Seasonal pattern of maximum diameter. Each point is the average of 4 values. 482 Vertical lines indicate the end of the pit hardening period. Asterisks significant 483 differences in trunk growth rate (p<0.05, Tukey Test). 484 Fig. 7. Relationship between (a) average trunk growth rate in two different periods vs 485 midday stem water potential at the end of each period (Y=0.060+0.027X; R2=0.61***; 486 Error=0.024; n=29; circled data not included); (b) seasonal average trunk growth rate vs 487 seasonal average midday stem water potential (Y=0.070+0.033X; R"=0.64***; 488 Error=0.016; n=14; circled data not included) (c) weekly average trunk growth rate vs 489 midday stem water potential at the end of that week. Each point is an individual plot 490 value. Circle points indicate plots where more than 50% of data were lost. 491 Fig. 8. Pattern of daily trunk growth rate (TGR) and midday stem water potential (SWP) 492 during two weeks in three individual plots. 493 Fig. 9. Relationship between midday stem water potential (SWP) before irrigation and 494 weekly average of trunk growth rate (TGR) after irrigation. Solid triangles are data for 495 22 individual plots with SWP lower than -3MPa. Empty triangles are data for individual 496 plots greater than -2MPa. The empty and solid circle is the average of all the values in 497 each group. Horizontal and vertical bars are the standard error of TGR and SWP in each 498 data group. 499 Fig. 10. Weekly frequency of daily TGR values at various levels of water stress. Water 500 stress conditions have been described with midday stem water potential values ()greater 501 than -1.4MPa, between -1.4 and -2.5MPa, between -2.5 and -4MPa and lower than - 502 4MPa. Daily TGR were grouped in higher than 0.3mm day-1, between -0.1 and 0.3mm 503 day-1, between -0.1 and -0.2mm day-1 and lower than -0.3mm day-1. All data for individual 504 plots were considered and grouped according to  values. Different letters in each range 505 of daily TGR indicate significant differences between the water stress level for the same 506 range of daily TGR values (p<0.05, Tukey Test). 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 23 Table 1. Average trunk growth rate (TGR, mm day-1) and standard error during the pit 523 hardening period, oil accumulation and whole season. Each point is an average of 4 524 values. There were no significant differences between treatments. (Tukey Test p<0.05) 525 526 527 528 529 530 531 532 533 534 Pit Hardening Oil accumulation Whole season Control 0.024±0.011 0.033±0.010 0.027±0.011 RDI 1 -0.006±0.014 -0.003±0.031 -0.006±0.015 RDI 2 -0.032±0.011 0.009±0.011 -0.015±0.008 SDI -0.007±0.054 0.051±0.055 0.002±0.017