Seasonal changes of maximum daily shrinkage reference equations for irrigation scheduling in olive trees: influence of fruit load

Agricultural Water Management Manuscript Draft Manuscript Number: AGWAT3330 Title: Seasonal changes of maximum daily shrinkage reference equations for irrigation scheduling in olive trees: influence of fruit load Article Type: Research Paper Keywords: LVDT; RDI; trunk diameter fluctuations. Abstract: Maximum daily shrinkage (MDS) is the parameter of daily cycle of trunk diameter most widely suggest in irrigation scheduling of several fruit trees. However, as in other plant-measured approach, the irrigation decision may be difficult due to the influence of the environment in the values obtained. Reference equations of MDS have been established in order to avoid the effects of environmental conditions. Such equations are usually related with simple meteorological data, in order to estimate easily MDS values in full-irrigated conditions. This work studies the influence of the fruit load and the inter-annual variations in the reference equation of MDS in olive trees. These reference equations were calculated during 4 seasons in a full-irrigated orchard and the equations were validated with the data of a different season. The values of MDS were related with vapour pressure deficit (VPD) and temperature obtained near the experimental orchard. In addition, meteorological data were considered as mean daily or as midday values. The validation of the equations were made using the fits with all the meteorological data considered (midday and mean daily of VPD and temperature). In each meteorological data, in addition, two different fit, one according fruit load and other with the complete pool data were used. The equations fit were significantly different each season in all the meteorological data considered. Although, seasons with similar fruit load were more similar. In both meteorological data considered (VPD and temperature) the midday values improve the fit respect to mean daily values. The reference equations in which temperature was used obtained best fit that the ones calculated with VPD. No significant differences were found in the validation when equations according with fruit load or using the complete pool data were compared. The limitations and usefulness of these reference equations is also discussed. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 1 Seasonal changes of maximum daily shrinkage reference equations for irrigation 1 scheduling in olive trees: influence of fruit load 2 3 A. Morianaa, F. Morenob,, I.F. Girónb, W. Conejeroc, M.F. Ortuñoc, M. Corella, A. 4 Torrecillasc,d 5 6 aEscuela Universitaria de Ingeniería Técnica Agrícola. University of Seville, Sevilla, 7 Spain 8 bInstituto de Recursos Naturales y Agrobiología (CSIC), P.O. Box 1052, E-41080 9 Sevilla, Spain 10 cDpto. Riego. Centro de Edafología y Biología Aplicada del Segura (CSIC). P.O. Box 11 164, E-30100 Espinardo (Murcia), Spain 12 dUnidad Asociada al CSIC de Horticultura Sostenible en Zonas Áridas (UPCT- 13 CEBAS), Paseo Alfonso XIII s/n. E-30203 Cartagena (Murcia), Spain 14 * Corresponding author: [email protected] Phone: (+34) 954624711; Fax: (+34) 15 954624002 16 17 18 19 20 21 22 23 24 *Manuscript Click here to download Manuscript: Moriana et al.doc Click here to view linked References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 2 Abstract 25 Maximum daily shrinkage (MDS) is the parameter of daily cycle of trunk diameter most 26 widely suggest in irrigation scheduling of several fruit trees. However, as in other plant- 27 measured approach, the irrigation decision may be difficult due to the influence of the 28 environment in the values obtained. Reference equations of MDS have been established 29 in order to avoid the effects of environmental conditions. Such equations are usually 30 related with simple meteorological data, in order to estimate easily MDS values in full- 31 irrigated conditions. This work studies the influence of the fruit load and the inter- 32 annual variations in the reference equation of MDS in olive trees. These reference 33 equations were calculated during 4 seasons in a full-irrigated orchard and the equations 34 were validated with the data of a different season. The values of MDS were related with 35 vapour pressure deficit (VPD) and temperature obtained near the experimental orchard. 36 In addition, meteorological data were considered as mean daily or as midday values. 37 The validation of the equations were made using the fits with all the meteorological data 38 considered (midday and mean daily of VPD and temperature). In each meteorological 39 data, in addition, two different fit, one according fruit load and other with the complete 40 pool data were used. The equations fit were significantly different each season in all the 41 meteorological data considered. Although, seasons with similar fruit load were more 42 similar. In both meteorological data considered (VPD and temperature) the midday 43 values improve the fit respect to mean daily values. The reference equations in which 44 temperature was used obtained best fit that the ones calculated with VPD. No significant 45 differences were found in the validation when equations according with fruit load or 46 using the complete pool data were compared. The limitations and usefulness of these 47 reference equations is also discussed. 48 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 3 Keyword: LVDT, RDI, trunk diameter fluctuations. 49 50 1. Introduction 51 Irrigated agriculture is actually the largest fresh water consumer in the world. In the last 52 decades, olive production in the Mediterranean region has intensified, and the 53 traditional rainfed crop is now frequently irrigated (Eris and Barut, 1995). The scarcity 54 of water supplies and the increasing demand of other water-user sectors impose to the 55 Mediterranean agriculture an increasing pressure to limit its water consumption, and so 56 there is a constant need to improve the water use by the crops using better irrigation 57 management (Fereres and Evans, 2006). Among the tools that olive growers can use to 58 achieve this goal are more precise irrigation scheduling methods which involve the 59 determination of water requirements by crop and/or the application of regulated deficit 60 irrigation. 61 Measurement of the plant water condition may be useful for irrigation 62 scheduling because of its dynamic nature, which is directly related with climatic and 63 soil conditions, as well as crop productivity (Fereres and Goldhamer, 2003; Goldhamer 64 et al., 2003). 65 The trunk or stem of all plants presents daily cycles of swelling and shrinking 66 that is known as trunk diameter variations (Kozlowski, 1967). Continuous records of 67 stem diameter have been proposed as a management tool for irrigation scheduling 68 (Huguet et al., 1992; Cabibel and Isberie, 1997; Cohen et al., 2001; Goldhamer and 69 Fereres, 2001). In a recent paper Ortuño et al. (2010) have reviewed the state of the art 70 regarding the use of trunk diameter variations derived parameters for irrigation 71 scheduling in woody crops. As so far as we know, Goldhamer and Fereres (2004) were 72 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 4 the first to demonstrate that is possible to develop a deficit irrigation schedule based 73 only on maximum daily trunk shrinkage (MDS) in almond trees. García Orellana et al. 74 (2007), Velez et al. (2007) and Ortuño et al. (2009c) confirmed that in citrus MDS is a 75 good indicator for scheduling deficit irrigation. Other useful parameter derived from the 76 trunk daily cycles of swelling and shrinking is the trunk growth rate (TGR) as defined 77 by Goldhamer and Fereres (2001) that can be used for irrigation scheduling of fruit 78 trees. 79 The use of the absolute values of the plant-based water status indicators could be 80 meaningless and thus we need to obtain reference values for these indicators. Reference 81 values can be obtained by maintaining trees under conditions of non-limiting soil water 82 supply. At the same time is necessary to develop reference equations to help us to 83 interpret the values of a plant-based water status indicator. These reference equations 84 can be obtained by relating their values in trees under non-limiting soil water conditions 85 with evaporative demand of the atmosphere (Moreno et al., 2006; Conejero et al., 86 2007b; Ortuño et al., 2009b and 2010). 87 MDS values can be affected by several factors, such as tree age (Moriana and 88 Fereres, 2004), phenological period (Marsal et al., 2002; Intrigliolo and Castel, 2004; 89 Moriana and Fereres, 2004; Conejero et al., 2007b) and fruit load (Conejero et al., 2010; 90 Marsal et al., 2002; Intrigliolo and Castel, 2006). In olive trees the alternate bearing can 91 be a factor that can affect MDS values. In a recent paper by Moriana et al. (2010) have 92 shown that MDS is no the best indicator for optimal irrigation scheduling in olive trees 93 but can be a good tool to be used in deficit irrigation scheduling. In this case, the stress 94 level will be indicated by MDS values lower than the one obtained in the base lines or 95 reference equations. 96 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 5 The objectives of this paper were: (1) to obtain reference equations of MDS for 97 olive trees based on its relation with the evaporative demand of the atmosphere; (2) to 98 study the interannual variation of the reference equations, and (3) to evaluate the 99 influence of fruit load on the MDS vs evaporative demand parameters relationships. 100 101 2. Material and Methods 102 2.1. Description and design of the experiment 103 Experiments were conducted at La Hampa, the experimental farm of the Instituto de 104 Recursos Naturales y Agrobiología (CSIC), which is located at Coria del Río near 105 Seville (Spain) (37º17’’N, 6º3’W, 30m altitude) during 5 consecutive seasons (from 106 2005 to 2009). The sandy loam soil (about 2 m deep) of the experimental site was 107 characterized by a volumetric water content of 0.33 m3 m-3 at saturation, 0.21 m3m-3 at 108 field capacity and 0.1 m3m-3 at permanent wilting point, and 1.30 (0-10cm) and 1.50 109 (10-120 cm) g cm-3 bulk density. 110 The experiment was performed on 37-year-old olive trees (Olea europaea L cv 111 Manzanillo). Tree spacing followed a 7m x 5m square pattern. Pest control and 112 fertilization practices were those commonly used by the growers and no weeds were 113 allowed to develop within the orchard. 114 Irrigation was carried out during the night by drip using one lateral pipe per tree 115 row and five emitters per plant, delivering 3 L h-1 each. Plants irrigation requirements 116 were determined according to daily reference evaotranspiration (ETo) and a crop factor 117 based on the time of the year and the percent of ground area shaded by the tree canopy 118 (Fernández et al., 1998). During the experimental period (from end of april until 119 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 6 beginning of October), total crop evapotranspiration (ETc) was 430 mm (2005), 413 120 mm (2006), 414 mm (2007), 430 mm (2008), 392 mm (2009). 121 During the experimental period, olive trees were irrigated daily above their 122 water requirements in order to obtain non-limiting soil water conditions. A total amount 123 of water (rainfall not included) of 476 mm (2005), 442 mm (2006), 410 mm (2007), 124 644mm (2008), 605mm (2009), measured with in-line water meters, was applied during 125 the experiment. 126 The design of the experiment was completely randomized with four replications, 127 each replication consisting of the three adjacent rows of five trees. Measurements were 128 made in the inner tree of the central row of each replicate, the other trees served as 129 borders. 130 2.2 Measurements 131 Micrometeorological 30 min data, namely air temperature, solar radiation, air relative 132 humidity and wind speed at 2 m above the soil surface were collected by an automatic 133 weather station located some 40 m from the experimental site. Daily reference 134 evapotranspiration (ETo) was calculated using the Penman-Monteith equation (Allen et 135 al., 1998). Mean daily vapour pressure deficit (VPDm) was calculated from the maen 136 daily vapour pressure and relative humidity (Goldhamer and Fereres, 2001). 137 Trunk diameter fluctuations were measured throughout the experimental periods 138 in four trees, using a set of linear variable displacement transducer (LVDT) (model 139 DF±2.5 mm, accuracy ±10 m, Solartron Metrology, Bognor Reis, UK) attached to the 140 trunk, with a special bracket made of Invar, an alloy of Ni and Fe with a termal 141 expansion coefficient close to zero (Katerji et al., 1994). Measurements were taken 142 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 7 every 10 s and the datalogger (model CR10X with AM 416 multiplexer, Campbell Sci. 143 Ltd., Logan, USA) was programmed to report 30 min means. Maximum 144 The data obtained during the five seasons were analyzed taking into account the 145 years with low fruit load (2005, 2007 and 2009) and years with full fruit load (2006 and 146 2008). Data from 2009 were used to validate the relationships obtained in previous 147 years. Linear regression analysis was carried out to explore relationships between 148 variables (MDS and climatic variables). Differences between regression lines were 149 determined with a T-test of the slope and y-intercept. 150 151 3. Results 152 The MDS vs mean daily temperature relationship during the four years of the 153 experiment showed the best fit in a lineal form (Table 1 and Fig. 1a). The increase in 154 temperature produces an increase in the MDS in a rate around 0.04 mm ºC-1. The range 155 of variations in mean daily temperature was wide enough for the Seville conditions of 156 olive growth and varied from around 10 to 30ºC. The equations of each year for mean 157 daily temperature are shown in Table 1. All the equations were significantly different in 158 the slope and the intercept. The coefficient of determinations was significant in all the 159 years but low, except in the 2008 season when it was clearly higher (r2=0.82). When the 160 data were grouped in full fruit load (FFL) and low fruit load (LFL) years there were no 161 significant differences in the slope but it was in the intercept. The LFL equations tended 162 to lower values of MDS than the FFL equations when the same mean daily temperature 163 is considered. 164 When the temperature considered is the ones that occurred at midday the scatter 165 of the points is reduced (Table 1 and Fig. 1b) in comparison with that of mean daily 166 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 8 temperature (Fig. 1a). The range of variations in temperature (Fig. 1b) is similar to that 167 in Fig. 1a, and changes from around 20 to 40ºC. The equations in each year were 168 significantly different between them, as in the case of mean daily temperature. The 169 coefficients of determination were slightly higher than the ones of mean daily 170 temperature (Table 1). The equations of LFL and FFL years were significantly different 171 for the intercept but not for the slope. As in the data of Fig. 1a, the values during LFL 172 year tended to be lower than the ones of the FFL year when the same range of 173 temperature is considered. 174 The relationship between MDS and VPD was also lineal. The increase in VPD 175 produces an increase in the MDS in full irrigated conditions (Fig. 2). When the mean 176 daily VPD is considered the range of data were from near 0 to 4 KPa (Fig. 2a). The 177 equations of each year were significantly different in the intercept and the scatter was 178 slightly higher than in the midday temperature relationship (Table 1 and Figs. 1b and 179 2a). There were also significant differences between the equations when they were 180 grouped in FFL and LFL. The MDS in FFL year tended to higher values than in LFL 181 year when the same mean daily VPD is considered. 182 The scatter in the MDS vs VPD relationship is slightly reduced when the values 183 at midday (Figure 2b) are considered instead of the daily average (Figure 2a). Although 184 the coefficient of determination was slightly higher than the mean daily VPD, they were 185 lower than the ones obtained in midday temperature relationships (Table 1). The range 186 of variations of midday VPD was also higher than mean daily VPD and it extended until 187 6 KPa (Fig. 2). There were significantly differences in the intercept but not in the slope 188 of the equations between years. There were also significant differences between the 189 equations when they were grouped in full fruit load and low fruit load years (Table 1). 190 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 15 Genard, M., Fishman, S., Vercambre, G., Huguet, J.G., Bussi, C., Besset, J., Habib, R., 331 2001. A biophysical analysis of stem and root diameter variations in woody 332 plants. Plant Physiol. 126, 188-202. 333 Goldhamer, D.A., Fereres, E. 2001. Irrigation scheduling protocols using continuously 334 recorded trunk diameter measurements. Irrig. Sci. 20, 115-125. 335 Goldhamer, D.A., Fereres, E. 2004. Irrigation scheduling of almond trees with trunk 336 diameter sensors. Irrig. Sci. 23, 11-19. 337 Goldhamer, D.A., Fereres, E., Salinas, M. 2003. Can almond trees directly dictate their 338 irrigation needs? Calif. Agric. 57, 138-144. 339 Herzog, K.M., Häsler, R., Thum, R., 1995. Diurnal changes in the radius of a subalpine 340 Norway spruce stem: their relation to the sap flow and their use to estimate 341 transpiration. Trees 10, 94-101. 342 Huguet, J.G., Li, Sh., Lorendeau, J. Y., Pelloux, G., 1992. Specific micromorphometric 343 reactions of fruit trees to water stress and irrigation scheduling automation. J. 344 Hort. Sci. Bio. 67, 631-640 345 Intrigliolo, D.S. and Castel, J.R., 2004. Continuous measurement of plant and soil water 346 status for irrigation scheduling in plum. Irrig. Sci. 23, 93–102. 347 Intrigliolo, D.S. and Castel, J.R., 2006. Usefulness of diurnal trunk shrinkage as a water 348 stress indicator in plum trees. Tree Physiol. 26, 303–311. 349 Intrigliolo, D.S. and Castel, J.R., 2007. Crop load affects maximum daily trunk 350 shrinkage of plum trees. Tree Physiol. 27, 89–96. 351 Katerji, N., Tardieu, F., Bethenod, O., Quetin, P., 1994. Behavior of Maize stem 352 diameter during drying cycles: comparison of two methods for detecting water 353 stress. Crop Sci. 34, 165-169 354 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 16 Kozlowski, T.T. 1967. Diurnal variations in stem diameters of small trees. Bot. Gaz. 355 123, 60-68. 356 Marsal, J., Gelly, M., Mata, M., Arbone´ s, J., Rufat, J., Girona, J., 2002. Phenology and 357 drought affects the relationship between daily trunk shrinkage and midday stem 358 water potential of peach trees. J. Hort. Sci. Biotechnol. 77, 411–417. 359 Moreno, F., Conejero, W., Martín-Palomo, M.J., Girón, I.F., Torrecillas, A., 2006. 360 Maximum daily trunk shrinkage reference values for irrigation scheduling in 361 olive trees. Agric. Water Manage. 84, 290–294. 362 Moriana, A. and Fereres, E., 2002. Plant indicators for scheduling irrigation for young 363 olive trees. Irrig. Sci. 21:83-90. 364 Moriana, A. and Fereres, E., 2004. Establishing references values of trunk diameter 365 fluctuations and stem water potential for irrigation scheduling of olive trees. 366 Acta Hort. 664:407-412. 367 Moriana, A., Orgaz, F., Fereres, E., Pastor, M., 2003. Yield responses of a mature olive 368 orchard to water deficits. J. Amer. Soc. Hort. Sci. 128:425-431 369 Moriana, A., Girón, I.J., Martín-Palomo, M.J., Conejero, W., Ortuño, M.F., Torrecilla, 370 A., Moreno, F. 2010. New approach for olive trees irrigation schedulings using 371 trunk diameter sensors. Agric. Water Manage. 97, 1822-1828. 372 Ortuño, M.F., Brito, J.J., Conejero, W., García-Orellana, Y, Torrecillas, A. 2009a. 373 Using continuously recorded trunk diameter fluctuations for estimating water 374 requirements of lemon trees. Irrig. Sci. 27, 271-276. 375 Ortuño, M.F., Brito, J.J., Conejero, W., García-Orellana, Y, Torrecillas, A. 2009b. 376 Maximum daily trunk shrinkage and stem water potencial baselines for irrigation 377 scheduling of lemon trees. Irrig. Sci. 27, 121-127. 378 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 17 Ortuño, M.F., García-Orellana, Y., Conejero, W., Pérez-Sarmiento, F., Torrecillas, A. 379 2009c. Assessment of maximum daily trunk shrinkage signal intensity threshold 380 values for deficit irrigation in lemon trees. Agric. Water Manage. 96, 80-86. 381 Ortuño, M.F., Conejero, W., Moreno, F., Moriana, A., Intrigliolo, D.S., Biel, C., 382 Mellisho, C.D., Pérez-Pastor, A., Domingo, R., Ruíz-Sánchez, M.C., Casadesus, 383 J., Bonany, J., Torrecillas, A. 2010. Could trunk diameter sensors be used in 384 woody crops for irrigation scheduling? A review of current knowledge and 385 future perspective. Agric. Water Manage. 97, 1-11. 386 Velez, J.E., Intrigliolo, D.S., Castel, J.R. 2007. Scheduling deficit irrigation of citrus 387 trees with maximum daily trunk shrinkage. Agric. Water Manage. 90, 197-204. 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 18 Figure Captions 403 Figure 1. Relationship of MDS with the mean daily temperature (a) and the midday 404 temperature (b) during four consecutive seasons (2005 to 2008). ■ 2005; ● 2006; □ 405 2007; ○ 2008. The regression equations obtained with each season, the “FFL (full fruit 406 load)” and “LFL (low fruit load)” season and the total pool of data is presented in Table 407 1. 408 409 Figure 2. Relationship of MDS with the mean daily VPD (a) and the midday VPD (b) 410 during four consecutive seasons (■ 2005; ● 2006; □ 2007; ○ 2008). Line represent the 411 fit of all the data. The regression equations obtained with each season, the “FFL (Full 412 fruit load)” and “LFL (Low fruit load)” season and the total pool of data is presented in 413 Table 1. 414 415 Figure 3. Validation of the reference equations with the measured data of MDS in full 416 irrigated trees during 2009 season. The equations used are the ones obtained with the 417 mean daily temperature (a) and the midday temperature (b). White circle are the 418 equation obtained with the LFL (low fruit load) years and black circle are the equation 419 using the pool data, total equation (see Table 1 for equations). In all the cases the 420 relationship between MDS measured and estimated are significantly different from the 421 line 1:1. 422 423 Figure 4. Validation of the references equations with the measured data of MDS in full 424 irrigated trees during 2009 season. The equations used are the ones obtained with the 425 mean daily VPD (a) and the midday VPD (b). White circle are the equations obtained 426 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 19 with the LFL (Low fruit load) years and black circle are the equations with the pool 427 data, “total equation (Table 1). In all the cases the relationship between MDS measured 428 and estimated are significantly different from the line 1:1. 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 20 Table 1. Equations, coefficient of determinations (r2) obtained in the relationships of 451 Figs. 1 and 2. Each season is presented and, in addition, the results when they are 452 grouped according to the crop load (LFL, low fruit load; FFL, full fruit load). “Total” is 453 the equation considering all data (four seasons). RMSE: residual mean squared error. N: 454 number of data. Statistic Dif: statistical differences between equations 455 456 Season Equations r2 RMSE N Statistic Dif. MDS vs Mean Temperature 2005 -0.79+0.053X 0.68*** 0.09 105 All of them 2006 -0.26+0.035X 0.48*** 0.12 111 Statistical 2007 -0.52+0.038X 0.65*** 0.09 160 Different 2008 -0.59+0.046X 0.82*** 0.08 175 LFL (05&07) -0.67+0.046X 0.68** 0.10 265 Intercept different FFL (06&08) -0.55+0.045X 0.73*** 0.10 286 Total -0.59+0.045X 0.67*** 0.11 551 MDS vs Midday Temperature 2005 -0.78+0.042X 0.79*** 0.07 105 All of them 2006 -0.37+0.033X 0.67*** 0.09 111 Statistical 2007 -0.57+0.034X 0.73*** 0.08 160 Different 2008 -0.61+0.038X 0.85*** 0.07 175 LFL (05&07) -0.65+0.037X 0.80*** 0.08 265 Intercept different FFL (06&08) -0.58+0.038X 0.78*** 0.09 286 Total -0.58+0.037X 0.73*** 0.10 551 MDS vs Mean VPD 2005 0.03+0.16X 0.69*** 0.08 105 2006 0.34+0.16X 0.57*** 0.10 111 Intercept 2007 0.09+0.20X 0.63*** 0.10 163 different 2008 0.08+0.24X 0.78*** 0.09 167 LFL (05&07) 0.17+0.13X 0.67*** 0.10 268 All of them FFL (06&08) 0.16+0.22X 0.64*** 0.12 278 Different Total 0.23+0.14X 0.48*** 0.14 546 MDS vs Midday VPD 2005 -0.07+0.13X 0.82*** 0.06 105 Intercept 2006 0.30+0.12X 0.68*** 0.09 111 different 2007 0.06+0.13X 0.67*** 0.09 160 2008 0.11+0.14X 0.51*** 0.13 167 LFL (05&07) 0.13+0.10X 0.66*** 0.10 265 All of them FFL (06&08) 0.16+0.14X 0.53*** 0.13 278 different Total 0.20+0.10X 0.45*** 0.14 543 457 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 21 458 459 Table 2. Best fits of the relationship between MDS observed and estimated using 460 different meteorological variables. The adjusted validated were obtained from 2005 to 461 2008 (Table 1), while the data used to compared such validations were measured during 462 2009 season (n=148). LFL, low fruit load equation. “Total” is the equation considering 463 all data (four seasons). RMSE: residual mean squared error 464 465 466 Eq. validated Equations RMSE r2 LFL Mean Temperature Y= 0.11+0.74X 0.10 0.66*** “Total” Mean Temperature Y=0.17+0.73X 0.09 0.66*** LFL Midday Temperature Y=0.09+0.81X 0.06 0.85*** “Total” Midday Temperature Y=0.15+0.80X 0.06 0.85*** LFL Mean VPD Y=0.20+0.44X 0.05 0.75*** “Total” Mean VPD Y=0.26+0.48X 0.05 0.75*** LFL Midday VPD Y=0.17+0.54X 0.05 0.81*** “Total” Midday VPD Y=0.24+0.54X 0.05 0.81*** 467 468 469 470 471 472 473 474 475 476 477 Mean daily temperature (ºC) 0 10 20 30 40 50 MDS (mm) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Midday temperature (ºC) 0 10 20 30 40 50 MDS (mm) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 a b Figure 1 Mean daily VPD (kPa) 01234567 MDS (mm) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 a Midday VPD (kPa) 01234567 MDS (mm) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 b Figure 2 MDS measured (mm) 0.0 0.2 0.4 0.6 0.8 1.0 MDS estimated (mm) 0.0 0.2 0.4 0.6 0.8 1.0 a MDS measured (mm) 0.0 0.2 0.4 0.6 0.8 1.0 MDS observed (mm) 0.0 0.2 0.4 0.6 0.8 1.0 b Figure 3