Water use and fruit yield of mango (Mangifera indica L.) grown in a subtropical Mediterranean climate

ABSTRACT

Mango (Mangifera indica L.) is one of the most important fruit crops in tropical and subtropical regions worldwide. On the coast of Granada and Malaga (SE Spain), irrigated subtropical fruit species have been introduced and cultivated on terraces with a considerable economic importance as the only European production region. The subtropical fruit production in this zone is possible with intensive irrigation on terraces, which are economically more profitable than traditional rainfed crops (almond and olive), which have been replaced or abandoned. A 2-year monitoring period was conducted using drainage lysimeters to determine the crop coefficients (Kc) and fruit yield in mango (Mangifera indica L. cv. Osteen) orchards. Also, some quality parameters such as titratable acidity, total soluble solids, and vitamin C were evaluated under these conditions. The averaged Kc values of mango trees varied within production cycle of 0.43, 0.67, and 0.63 at flowering, fruit set, and fruit growth, respectively. In this study, the fruit yield under full water requirements (100% ET_C) averaged 24.1 kg tree⁻¹, amounting to 21.2 kg ha⁻¹ mm⁻¹ in terms of water-use efficiency. The quality parameters of the mango fruits harvested in the study area were satisfactory. Thus, this study highlights the need to optimize the irrigation-water use according to actual mango requirements, thereby achieving more sustainable Mediterranean subtropical farming in orchard terraces.

KEYWORDS:

• Crop coefficient
• Mangifera indica L.
• subtropical fruits
• drainage lysimeter

Introduction

Crop terraces are among the most characteristic agricultural features in mountainous areas in many parts of the world (Tarolli et al., 2014), at the same time as being among the most visible prints of humans in the landscape (Arnáez et al., 2015). The main objectives for constructing orchard terraces are to encourage water and soil retention (Cots-Folch et al., 2006), encourage water infiltration in areas, diminish runoff velocity and thus palliate erosion (Durán et al., 2011; Li et al., 2012), allow agricultural machinery and farmers to work, and establish irrigation systems.

In particular, many mountainous areas in Spain have been terraced during the last few decades. Specifically, along the coasts of the provinces of Granada and Malaga (Andalusia, SE Spain), considerable economic investment has been made to design wide and high terraces with a regular layout for commercial purposes (Durán et al., 2003, 2013).

In SE Spain, a profitable agricultural economy has been developed based on the cultivation of subtropical crops on recently terraced slopes, since Spain is the only producer of these fruits in Europe (Barea et al., 2016; Durán et al., 2003). This is a subtropical, Mediterranean climate characterized by dry, hot summers and wet autumns and winters. According to the surface area cultivated, as well as to the economic impact, the leading crop is avocado (Persea americana Mill.), with 14,938 ha in Spain, from which 13,393 ha are located in Andalusia (90%) (ESYRCE Statistics, 2016), followed by mango (Mangifera indica L.), with 3,982 ha in Spain and 3,726 in Andalusia (94%) (ESYRCE, 2016) (Table 1).

Table 1. Exportation volume values and total transaction for mango in Spain.

Season	2010/11	2011/12	2012/13	2013/14
Mango (Mangifera indica L.)
Volume (t)	9,366	15,422	9,282	20,579
Value (thousands of €)	12,700	20,276	17,448	33,906
Avocado (Persea americana Mill.)
Volume (t)	53,916	49,803	51,829	49,807
Value (thousands of €)	98,674	95,825	92,886	107,060

Conversely, the declining availability of fresh water is becoming a worldwide crisis, worsened by climate change, mainly in the Mediterranean basin, which is characterized by dry summers with high temperatures and evapotranspiration rates, with precipitation commonly concentrated in autumn and winter but largely unpredictable in amount. The adoption of water-saving strategies by agriculture is becoming increasingly critical, especially due to shortages (García and Durán, 2018).

Nevertheless, few studies have examined the optimal water supply for subtropical Mediterranean farming. Major studies have quantified the amount of irrigation water used by fruit tree crops, with special attention on areas where this resource is scarce worldwide. Most of the water taken from the soil matrix by trees is lost to evapotranspiration, and therefore it is crucial to appraise current crop evapotranspiration (ET_C) for the entire vegetative cycle that coincides with the actual crop water demands.

Estimating the reference crop evapotranspiration (ET₀) and then adjusting by means of specific crop coefficients (K_C) for a given plant has been widely used for irrigation scheduling and reported by many authors (Allen et al. 2005; FAO, 1998; Wright, 1982). However, the crop coefficient (K_C = ET_C/ET₀) as pointed out by Allen et al. (1998) can be affected by many other factors (crop height and architecture, canopy resistance, albedo, soil surface evaporation, etc.). Several studies using the water-balance method with drainage and weighing lysimeters have been developed to estimate the K_C for different fruit trees (Abrisqueta et al., 2001; Chalmers et al., 1992; García et al., 2015; Lorite et al., 2012). Additionally, K_C should account for particular orchard conditions such as cultivar, orchard orientation, plant spacing, training system, and soil management, among other factors.

In this line, the knowledge of evapotranspiration is essential for efficient water management, and accurate predictions are needed in order to adjust irrigation volume and frequency to crop water demand. Consequently, given the importance of mango for the economy in the study area, and the complex soil-plant-water relationships in these agroecosystems, the primary objective during the two-monitoring period was to characterize water consumption without any restrictions needed by this crop cultivated on steeply sloping terrain on the Granada coast. In the present study, drainage lysimeters on terraces were used to determine: (1) water-use performance for mango (Mangifera indica L. cv. Osteen) by calculating the crop coefficients (K_C), and (2) the fruit yield and tree growth in mango, striving to improve water-use efficiency (WUE) and thus promote water savings.

Material and methods

Experimental location and mango orchard

The study was performed on orchard terraces of mango located some 7 km north of the Mediterranean coast near Almuñécar (Granada, SE Spain) (36º48′00″ N, 3º38′0″ W) at an elevation of 180 m a.s.l. The terrace, representative of those commonly found in the study area, is a reverse-sloped bench-terrace type with a toe drain measuring 160–170 m long (Figure 1). The platform was 2–3 m wide and the talus 3–5 m high. Local temperatures are subtropical to semi-hot within the Mediterranean climatic category (Elías and Ruiz, 1977). The average annual rainfall in the study area is 449.0 mm and the average temperature is 20.8°C.

Figure 1. Orchard terraces for subtropical farming in SE Spain.

The soils of the zone are Typical Xerorthent (Soil Survey Staff, 1999), with 684 g kg⁻¹ of sand, 235 g kg⁻¹ of silt and 81 g kg⁻¹ of clay, containing 9.4 g kg⁻¹ of organic matter, and 0.7 g kg⁻¹ of N, with 14.6 mg kg⁻¹ P, and 178.7 mg kg⁻¹ available K. For the soil profile from 0.10 to 0.90 m, the soil water content at field capacity θ_F (0.33 bar) and soil water content at permanent wilting point θ_W (15 bar) had mean values of 0.23 and 0.11 cm³ cm⁻³, respectively.

Bearing mango (Mangifera indica L. cv. Osteen grafted onto ‘Gomera-1ʹ) trees 15 years old, in a single row, were spaced 3 m apart with about 600–630 trees ha⁻¹. One of the scions most commonly used in south-eastern Spain, for their commercial appeal, include the cultivar from Florida (USA) “Osteen”. The experimental orchard was managed according to conventional practices in the area, using the conventional fertilization and routine cultivation techniques for diseases and insect control.

Mango drainage lysimeters

The two drainage lysimeters used for the present experiment formed part of an experimental mango plot. Each mango lysimeter corresponded to one tree 15 years old and were 6 m² (2.0 m × 3.0 m), 1.0 m deep bounded on the sides by nylon-reinforced polyethylene, and 35 m apart. The lysimeters were located on the terraces as a part of the orchard with mature trees with full production. Irrigation for each drainage lysimeter was applied by a combination of self-regulating emitters (4 L h⁻¹) in a double-line system and controlled automatically by a head-unit programmer and electro-hydraulic valves. The amounts of water applied per lysimeter were measured with flow meters.

Two experimental plots with 13 trees per row were studied, each plot including the tree and lysimeter. The five central trees of the rows were monitored for fruit yield and tree-size measurements and the other four trees served as border trees. At harvest, the total yield per tree (10 evaluated trees) was registered and fruits were collected to measure the vertical and horizontal diameters with a Vernier calliper.

Additionally, height, canopy diameter, and trunk circumference were measured 15 cm above the bud union in grafted trees. Canopy volume was calculated using the equation for one-half of a proplate spheroid:

(1) $CV=4/3\times \pi \times r^2\times 1/2\times H,$(1)

where CV is canopy volume, r is canopy radius, and H is canopy height. Trunk circumference was converted into trunk cross-sectional area: TCSA = [C²/4π], where C is the trunk circumference (cm).

Yield efficiency was estimated by dividing fruit yield by CV and by TCSA. WUE was calculated as fresh mango yield divided by total seasonal irrigation water applied.

Water balance and crop coefficient calculation

Reference evapotranspiration (ET₀) was estimated by the Penman-Monteith equation, as recommended by Allen et al. (1998). The weather data used to calculate ET₀ were taken from a weather station situated 80 m of the drainage lysimeters. The crop coefficient (Kc) was calculated with the following equation:

(2) $Kc=ETc/E{T}_0,$(2)

where ETc is the actual evapotranspiration (mm) and ET₀ is the reference evapotranspiration (mm). Here ETc is estimated with the soil-water-balance equation of Hillel (1998):

(3) $ETc=Pef+I+U+R-Dw-\mathrm{\Delta }S,$(3)

where Pef is the effective precipitation (mm), determined by USDA soil-conservation services method (Kuo et al., 2006; SCS, 1972), I the irrigation quota (mm), U the upward capillary flow into the root zone (mm), R the runoff (mm), Dw the downward drainage out of the root zone (mm), and ΔS the volumetric change of soil water stored in soil layer of 0–90 cm (mm).

The upward movement of water (U) in the loamy soil of the experimental site was estimated using Darcy’s law (Fares and Alva, 1999; De Medeiros et al. 2001), indicating that it could be considered negligible in the water balance equation. The surface runoff (R) was also negligible during the two growing seasons because the lysimeters were located in the platform of terraces with 0% slope. The downward flow (Dw) was measured by drainage lysimeter.

Finally, soil-water content was measured using a hand-held capacitance probe (Diviner-Sentek Pty Ltd.), monitoring the water content at 10, 20, 30, 40, 50, 60, 70, 80, and 90 cm soil depth. Measurements were made before and after irrigation, and heavy-rain events.

Physical and organoleptic characterization of mango fruits cv. “osteen”

Physical and organoleptic parameters were measured to characterize mango fruit in a subtropical environment. Skin, pulp, and seed percentages were evaluated in selected fruits. Also, titrable acidity was measured after diluting the fruit juice in distilled water (1:2) by titrating against sodic hydroxide 0.1 N according to the method of Liu et al. (2010) and the results were expressed as a percentage of citric acid. In addition, the determination of Vitamin C by titration was made (Redox Titration using iodate solution). The pH was evaluated by using a pH-meter (WTW 82382 Weilheim, Germany) and the total soluble solid content by the use of a refractometer (Refractometer PAL 3, ATAGO, Japan) and was expressed as a percentage.

Results and discussion

Calculation of crop coefficients (K_C)

In this study, the calculated daily crop reference evapotranspiration (ET₀) data were plotted for the 2 years (Figure 2). The data show the typical bell shape for ET₀ for this latitude and the year-to-year consistency. From June to September, no significant amount of rainfall could have affected the variability of the ET₀ readings.

Figure 2. Reference crop evapotranspiration (ET₀) in the study area during the two-year monitoring period. DOY, day of the year.

On the other hand, Figure 3 displays the crop evapotranspiration (ET_C) data, as estimated by the water-balance method in the drainage lysimeters, and twice replicated, following a pattern similar to that shown for ET₀. The ET_C increased to reach a peak in July (DOY 162–204), and then showed a certain asymmetry by falling during the fruit growth and harvest periods. However, both, ET₀ and ET_C showed noise due to several factors such as the environmental conditions and the soil in terraces that could affect the soil-water balance. This drawback could be solved in part with replications of lysimeters inside the experimental orchard.

Figure 3. Average crop evapotranspiration (ET_C) for terraced mango trees estimated in drainage lysimeters during the two-year monitoring period. DOY, day of the year.

The K_C reached an average maximum value of 0.82 (at fruit growth period DOY 183), showing a pattern comparable to that observed for ET_C, and also its turning point, which took place directly after fruit growth and harvest. A similar trend in Kc has been described for mango and other types of fruit trees (Kisekka et al., 2016). Also, in this context, Silva et al. (2007) studied an irrigated mango plantation in Brazil (average rainfall 400 mm), reporting values of ET₀ of 5.3 ± 1.03 and 4.9 ± 1.01 mm d⁻¹, and average maximum K_C values of 0.85 and 0.88 for two different seasons, these being similar to the peaks for K_C found in the present experiment.

Figure 4 shows the changes of the average crop coefficient (K_C) for mango over two monitoring seasons estimated by the water balance from experimental drainage lysimeters. The K_C values presented at three main growing stages, i.e. flowering (DOY 58–108), fruit set (DOY 114–164), and fruit growth (DOY 170–247), were fitted by a polynomial function and proved to be highly correlated (R² = 0.79).

Figure 4. Crop coefficient as a function of day of the year (DOY) for mango trees growing on orchard terraces during a two-year monitoring growing season.

Throughout the vegetative cycle the K_C varied considerably during flowering (K_C = 0.21–0.59), fruit set (K_C = 0.57–0.80), and fruit growth (K_C = 0.31–0.82), with about 49, 50, and 77 days, averaging K_C values of 0.43, 0.67, and 0.63, respectively. From DOY 240–251 ahead the fruit growth and harvest period could overlap, being highly dependent on environmental factors during a given season. After fruit harvest, K_C for the mango trees decreased quickly to 0.26.

The K_C values for mango trees were low, especially for subtropical Mediterranean areas in orchard terraces. In this sense, higher K_C values have been registered in Egypt by Mattar (2007) for flowering, fruit set, and fruit growth, with 0.66, 0.85, and 0.88, respectively. Similarly, in Brazil, Coelho et al. (2002) pointed out that the mango Kc increased from 0.39 at flowering to 0.85 during fruit growth. By contrast, Mohammad et al. (2015) reported almost unchanged values between 0.71 and 0.77 for entire production period.

The average annual value of the K_C for mango trees during the irrigation period (March-October) was 0.58 (DOY 90–271). These values provide a useful basis for designing the irrigation timetable in drip-irrigation systems. In this context, in Egypt and Brazil, Mattar (2007) reported K_C for mango tree productive cycle of 0.74 and 0.91, respectively, with bigger trees and higher fruit-yield rates than our experimental trees. Also, Teixeira et al. (2008) reported Kc ranging from 0.65 to 1.05 when average over a 20-day period, registering high values when the soil surface was frequently wetted by rain or irrigation.

Figure 5 shows the monthly K_C estimated in our experiment in comparison with other studies for mango, displaying the variability and the greater values found than on the coast of Granada, which is due basically to the size of the tree. Also, other factors involved in the production cycle of mango fruits such as local growing environment (subtropical Mediterranean climate) and management practices (orchard terraces) are crucial and should be considered in analyzing these findings.

Figure 5. Monthly average of K_C for the present study in comparison with other studies for mango trees.

Finally, according to the findings the annual water use of terraced mango trees in the framework of present study for the first and second season was 8.9 to 6.9 m³ per tree, leading to an irrigation volume of about 5,326–4,166 m³ ha⁻¹, respectively. Thus, the total average water applied during the 2-year monitoring period in terraced mango orchard was about 4,746 m³ ha⁻¹ (Table 2). Despite the importance of this fruit crop, little published data is available on the basic physiology in respect of water relations and their interactive response with the environment.

Table 2. Effect of irrigation (100% ET_C) on fruit yield, tree growth and water-use efficiency.

	Fruit				Tree growth				Yield efficiency
Irrigation (m^3 ha^−1)	Yield (kg tree^−1)	W (g)	L (mm)	W (mm)	TCSA (cm^2)	CD (m)	H (m)	CV (m^3)	(g cm^−2)	(kg m^−3)
4,746	24.1 (± 1.7)	648.4 (± 21)	141.0 (± 12)	95.0 (± 6.1)	136.0 (± 5.6)	2.9 (± 0.2)	2.9 (± 0.6)	13.9 (± 0.8)	171.2 (± 8.5)	1.3 (± 0.1)

W, weight; L, length; W, width; H, height; TCSA, trunk cross-sectional area, CD, canopy diameter; CV, canopy volume. ± Standard deviation

Thus, deficient water management due to low irrigation efficiency or inadequate irrigation scheduling in tree crops can lead to the waste of water, resulting in higher monetary costs as well as being environmentally unacceptable. In order to avoid the under- or overestimation of crop water requirements, knowledge of the water loss through evapotranspiration is crucial in the context of sustainable subtropical farming in Mediterranean area.

Fruit yield and tree growth

Over the two-year study period, fruit yield was averaged for about 600 trees per ha, and distributed in terraces of 14.5 t ha⁻¹ for the 100% ET_C (Table 2).. The trees produced 40 fruits per monitoring period, averaging a total of 24.1 kg per tree. The fruit yield in our experiment generally proved to be much lower than those reported by Avilán (1974) in Venezuela for cvs. Kent and Smith of 378 and 868 kg tree⁻¹, respectively. In this context, Iqbal et al. (2012) reported fruit yield ranging from 85.5 to 140.0 kg tree⁻¹ in Pakistan. Similarly, Smith et al. (2008) in Australia stated 36 and 181 kg tree⁻¹ for the lowest yielding and more productive trees, respectively, in a mango orchard with about 125 trees ha⁻¹ (8 m × 10 m). The Granada coast represents the climatic limit for commercially viable mango cultivation and, therefore, tree sizes and yields are far smaller.

However, due to the different planting patterns, i.e. high-density planting, the yields reached in our experiment are in a good range with respect to those from typical mango-producing areas, as reported by Bhriguvanshi et al. (2012) with mango yield in subtropical zones in India from 9.0 to 11.9 t ha⁻¹, as well as Pavel and Villiers (2004) (5.0–9.0 t ha⁻¹) and Spreer et al. (2007, 2009) (4.0–20.0 t ha⁻¹) in northern Thailand. In Brazil under full irrigation (100% ET_C) and a zone with an average of rainfall of 680 mm, the fruit yield for mango cv. Tommy Atkins with 3,450 m³ ha⁻¹ amounted to 24.0 t ha⁻¹ (dos Santos et al., 2015). For Australian conditions, Smith et al. (2008) reported fruit yield ranging between 4.5 and 22.6 t ha⁻¹.

In tropical areas of India, in agreement with Sharma et al. (2001), the high-density planting (2.5 m × 2.5 m) is standard for mango, but profitable yield for up to 14–15 years showed to a progressive decline due to the over-crowding of canopies. In our subtropical area the tree growth is reduced due to climate, and the canopy volume does not represent a significant problem with appropriate pruning management every 2-year period (Durán et al., 2003). Therefore, the feasibility of profitable mango cultivation could last longer than in the tropics.

The pomological characteristics of mango cv. Osteen in terms of weight, length and width agreed with the values reported by Rodríguez et al. (2012).

The average WUE value under this irrigation (100% ET_C) was 3.1 kg m⁻³, and therefore, in terms of evapotranspiration (WUE = Yield/ET_C) in our subtropical area amounted to 21.2 kg ha⁻¹ mm⁻¹. In Brazil, dos Santos et al. (2015) reported a WUE of 69.5 kg ha⁻¹ mm⁻¹ with higher fruit yield than along the Granada coast. It is well known that crop-water increases do not necessarily result in higher yields or WUE. Consequently, the determination of the appropriate amount of irrigation in mango is important in order to design water-saving strategies.

Regarding tree size, the height, canopy diameter, canopy volume as well as in trunk cross-sectional area values proved much lower than trees in other traditional producing mango zones (Table 2). Mirjat et al. (2011) reported mango trees with canopy diameters from 5.6 to 6.4 m in Pakistan, contrasting with 2.9 m in our study area. According to Dayal et al. (2016), considerable differences in scion canopy volume was observed due to the influence of rootstocks in different mango cultivars from 5.7 to 21.2 m³ in subtropical climatic conditions in India.

Finally, yield efficiency in terms of tree growth was 171.2 g cm⁻², and along this line in Australia, Smith et al. (2008) reported values ranging from 174.0 to 503.0 g cm⁻².

Thus, as stated above, these characteristics regarding the size of mango trees in our subtropical climate in south-eastern Spain offer the possibility for high-density planting for similar fruit yields as in tropical and other subtropical areas.

Physico-chemical characteristics for mango fruits

For this 50 mango fruits from monitored 10 trees (5 fruits per tree) were selected to assess their physico-chemical characteristics. The average fruit weight for the two seasons was 659.16 ± 101.57 g, including the percentage of seed, skin, and pulp 8.21 ± 0.79, 8.58 ± 0.82 and 84.68 ± 1.11%, respectively. Mangos reached 140.84 ± 12.41 mm length and 95.86 ± 7.11 mm width. The cv. “Osteen” gives very good results since fruit characteristics are usually within the acceptance parameters for commercialization. In this sense, Ortega et al. (2017) characterized several cultivars of mango in Ecuador, finding a fruit length of 63.3 to 125.0 mm and for width 52.7 to 85.0 mm. These researchers also evaluated the average fruit weight, calculating for cv. Kent, 465.0 g for Edward and 500.0 g for Tommy Atkins, with a high variability.

Total soluble solid was 16.34 ± 2.34ºB, and titrable acidity 0.27 ± 0.09 g acid/mg sampled, and thus the relation sugars/acidity was 74.45 ± 21.75. In this context, Ortega et al. (2017) studying different cultivars found a coefficient of variation of 0.38%, implying that there were no significant differences among cultivars, ranging from 11.30 to 16.20ºB. Thus, cv. Osteen showed good results for quality parameters, since the total sugar concentration was high compared with other studies. Islam et al. (2013) reported titrable acidity results of between 0.32 and 0.35. Also, Kamruzzaman et al. (2014) investigated the variation in certain functional factors among five commercially important Bangladeshi mango varieties such as cvs. Fazli, Langra, Ashwina, Himsagor, and Amrupali, reporting that the amount of total soluble solid of 20.13ºB was found in cv. Amrupali in contrast to other cultivars. These authors reported that the highest titratable acidity (0.798%) was observed in cv. Ashwina and lowest acid content (0.154%) was found in cv. Fazli among all cultivars and, therefore, higher values of acidity than in our experiment.

Studies in relation to mango fruit quality in subtropical Mediterranean climate is scarce, especially those regarding to the Florida mango cultivars.

The results in relation to vitamin C concentration were 20.32 ± 6.38 mg/100 ml juice. Our results are lower than those found by Cuastumal et al. (2016) (47.32 ± 3.10 mg/100 ml juice) or by Lauricella et al. (2017) (36.4 mg/100 ml juice). Despite this relatively low concentration for vitamin C, the quality parameter results for mango cv. Osteen cultivated in subtropical conditions in south-eastern Spain are highly satisfactory.

Conclusion

Spain produces mango fruits of excellent appearance and taste, and with minimal pesticide treatments. Moreover, the country’s proximity to the European market results in lower transport costs and proves more environmentally friendly, leaving a lower carbon footprint. However, the primary problem of cultivation on terraces, as for the rest of the crops in the area, is the excessive energy costs for pumping irrigation water to high levels and the price of water (~ 0.40 € m⁻³), together with the urgency to promote water-saving strategies in subtropical farming. Therefore, the traditional amount of irrigation water applied in terraced mango orchard in the study area has to be adjusted to the real requirements.

According to the findings the K_C values estimated during the two-monitoring period in the present study offer a useful tool for improving irrigation management for mango orchards located in terraces, adjusting irrigation volume and frequency to crop water demand under subtropical Mediterranean climate. Kc, in this sense were in consonance with those found by other authors for similar climates. Moreover, in the soil and climate conditions of the south of Spain, the mango orchard water requirement is not constant throughout the productive cycle. The average WUE value under this irrigation (100% ET_C) was 3.1 kg m⁻³, and therefore, in terms of evapotranspiration (WUE = Yield/ET_C) in our subtropical area amounted to 21.2 kg ha⁻¹ mm⁻¹. The fruit yield and the size of mango tree under subtropical conditions on the coast of Granada is far smaller than other tropical and subtropical areas, but this peculiarity makes it possible to grow orchards at high-density planting with production similar than those from conventional mango-producing areas. The main physico-chemical parameters of Mediterranean subtropical mango fruit highlighted in the present study during two seasons can be considered satisfactory in relation to fruits produced in traditional tropical mango areas.

Acknowledgments

Part of this publication was sponsored by the following research project: “Impact of deficit irrigation on productivity of subtropical fruit crops: tools for sustainable water stress management” (PP.AVA.AVA201601.8), and co-financed by the European Regional Development Fund (ERDF) within the Operational Programme Andalusia 2014-2020 “Andalucía is moving with Europe”.