Water Use and Leaf Nutrient Status for Terraced Cherimoya Trees in a Subtropical Mediterranean Environment

Water scarcity in many semi-arid agricultural areas, in particular for the Mediterranean basin, is promoting changes in irrigated agriculture, with alternative strategies being introduced for water-use optimization. The coast of Granada and Malaga (Southeast Spain) is an economically important area for subtropical fruit cultivation. This intensively irrigated agriculture is characterized by requiring extra amounts of water and the adoption of sustainable practices to improve agricultural water management. A two-season experiment was conducted to assess (1) the water use in terraced cherimoya (Annona cherimola Mill. cv. Fino de Jete) orchards under conventional and organic production systems with drainage lysimeters, and (2) the impact on fruit yield and nutritional effects between the two considered production systems. Crop coefficient (Kc) values for cherimoya were 0.60–0.66, 0.64–0.71, and 0.48–0.62 at flowering, fruit set, and fruit growth, respectively. Fruit yield was similar in both systems, ranging from 47.1 for conventional to 44.1 kg tree−1 for organic farming, averaging 13.2 and 12.3 t·ha−1, respectively. No differences between these systems were observed in terms of leaf nutrient status, with variations in the N, P, and K contents during the different phenological stages. The N, P, and K lessen during flowering and fruit growth; the highest levels of these nutrients were fixed at harvest. These patterns were the opposite in Ca and Mg, ascribable to the antagonism between K and both Ca and Mg. Thus, these findings highlight the need to establish the optimal use of irrigation water with respect to crop requirements, thereby encouraging sustainable subtropical farming in terraces.


Introduction
The reduction of available water resources has become a global problem, aggravated by climate change [1]. The problems arising under this scenario will be accentuated in the countries of the Mediterranean basin, where the increase in temperature will be even more pronounced, and rainfall will be reduced, both characterized by greater temporal and spatial variability [2]. These changes in climate, together with an increase in the atmospheric concentration of CO 2 , will have direct adverse effects on ecosystems and agricultural production [3], especially regarding the availability and distribution of irrigation water [4,5]. In addition to temperature and rainfall, relative humidity is also expected to have a great influence on the water-use efficiency [6].
(K C ), fruit yield, and mineral nutrition throughout the production cycle under conventional and organic production systems in a subtropical Mediterranean environment.

Experimental Area and Drainage Lysimeters
The experiment was conducted on orchard terraces of cherimoya located in the Mediterranean coast near Almuñécar (Granada, Southeast Spain) (36 • 48 00" N, 3 • 38 0" W) at an elevation of 150 m above sea level (a.s.l.). The study terrace was a reverse-sloped bench-terrace type with a toe drain measuring 150-180 m long, with a platform of 2-3 m wide, and the talus 3-5 m high ( Figure 1A).

Experimental Area and Drainage Lysimeters
The experiment was conducted on orchard terraces of cherimoya located in the Mediterranean coast near Almuñécar (Granada, Southeast Spain) (36°48′00″ N, 3°38′0″ W) at an elevation of 150 m above sea level (a.s.l.). The study terrace was a reverse-sloped bench-terrace type with a toe drain measuring 150-180 m long, with a platform of 2-3 m wide, and the talus 3-5 m high ( Figure 1A). The soil was characterized by 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 total N, 14.6 mg·kg −1 total P, and 178.7 mg·kg −1 available K [16]. For the soil profile from 0.10 to 0.90 m, the soil water content at field capacity θF (0.33 atm) and at permanent wilting point θW (15 atm) had average values of 0.23 and 0.11 cm 3 ·cm −3 , respectively.
Cherimoya (A. cherimola cv. Fino de Jete) trees, twenty years old, were planted in single rows spaced 7 m × 5 m apart (approx. 280 trees·ha −1 ; Figure 1A,B). In all studied trees, about 100-120 flowers were hand pollinated with pollen from the same cultivar in both conventional and organic plantations. Cherimoya flowers were pollinated in the morning after storing pollen overnight to pollinate newly opened flowers.
Cherimoya (A. cherimola cv. Fino de Jete) trees, twenty years old, were planted in single rows spaced 7 m × 5 m apart (approx. 280 trees·ha −1 ; Figure 1A,B). In all studied trees, about 100-120 flowers were hand pollinated with pollen from the same cultivar in both conventional and organic plantations. Cherimoya flowers were pollinated in the morning after storing pollen overnight to pollinate newly opened flowers.
The whole experimental orchard, as well as the trees studied with the lysimeters, was managed according to commercial practices in the area, using conventional and organic fertilization and routine cultivation techniques. The mineral fertilizer application rate (N:P 2 O 5 :K 2 O 1:0.40:0.70) was 680, (190 kg N·ha −1 ), 272 (174 kg P 2 O 5 ·ha −1 ), and 476 (160 kg K 2 O·ha −1 ) g tree −1 , respectively, to simulate the contributions made by farmers in the area, who usually used KNO 3  The lysimeters were located on the terraces as a part of the orchard with mature cherimoya trees at full production ( Figure 1C). The drainage lysimeters were replicated twice for each system production, containing one tree and 7.5 m 2 in area (3.0 m × 2.5 m), 1.0 m deep, bounded on the sides by nylon-reinforced polyethylene, and 35 m apart. Irrigation 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.
Reference evapotranspiration (ET 0 ) is the amount of water lost by evapotranspiration from a hypothetical reference crop with an assumed crop height, fixed surface resistance, and an albedo that is maintained under optimal water and nutrient conditions. ET 0 was estimated by the Penman-Monteith equation [17], and climatic data were obtained from a meteorological station 80 m from the drainage lysimeters. The ET C was estimated weekly with the soil-water balance (SWB) [18], which is as follows: where P EF is the effective precipitation (mm) [19,20], I is the irrigation amount (mm), U is the upward capillary flow into the root zone (mm), R is the runoff (mm), D W is the downward drainage out of the root zone (mm), and ∆S the volumetric change in soil water stored in the 0-90 cm soil layer (mm). The downward flow (D W ) was determined using the drainage lysimeter. Soil water content (∆S) was measured twice weekly using the Frequency Domain Reflectometry (FDR) system, at 10, 20, 30, 50, 70, and 90 cm soil depth in the lysimeters. The FDR used was a commercial device with a hand-held capacitance probe (Diviner-Sentek Pty Ltd., Sentek Sensor Tecchnologies, Stepney, South Australia, Australia). This instrument comprises a data display connected by cable to a portable probe rod with one sensor attached. The upward movement of water (U) was evaluated in agreement with Darcy's law [21,22], which could be considered negligible in the water balance equation. The surface runoff (R) was also negligible because the lysimeters were placed within the platform of terraces with 0% slope.
Finally, the crop coefficient (K C ) was weekly calculated with the following equation: The water-use efficiency (WUE) was adjusted by the equation where Y is the fruit yield (kg·ha −1 ) and ET C is the total actual evapotranspiration over the growing season (mm). Finally, throughout the study period in both conventional and organic plantations, foliar samples were collected for chemical analyses [16], and at the end of each survey, the yield of seven trees was determined, and weight and size of 15 fruits per tree were measured using calipers. Figure 2A displays the reference crop evapotranspiration (ET 0 ) throughout the study period for both production systems, showing a typical Mediterranean pattern. The ETc increased in agreement with the crop water demand, especially during development of fruits in the tree ( Figure 2B). Overall, ETc was higher during the summer months, especially in July, with the maximum monthly average ETc for cherimoya being 4.6 mm day −1 or 164.7 L·tree −1 day −1 during the fruit-set phenological stage.

Crop Coefficients (Kc)
Horticulturae 2019, 5, x FOR PEER REVIEW 5 of 11 ETc was higher during the summer months, especially in July, with the maximum monthly average ETc for cherimoya being 4.6 mm day −1 or 164.7 L·tree −1 day −1 during the fruit-set phenological stage. Figure 2C shows the changes in the average Kc for cherimoya over the two monitoring seasons estimated by the water balance from experimental lysimeters for both productions systems. The Kc at three main growing stages (flowering, fruit set, and fruit growth) was fitted by a polynomial function (Kc vs. Julian days). During the phenological stages of flowering, fruit set, and fruit growth, average Kc values were 0.63, 0.68, and 0.55, respectively. After fruit harvest, the Kc for cherimoya trees decreased quickly to 0.23. The Kc is closely related to crop type and management practice, which may influence the plant development rate and ground coverage throughout vegetative growth [17,23]. The average annual Kc value for the cherimoya trees during the irrigation period (May-October) was 0.55. The pooled Kc values for cherimoya trees for both systems was the highest in summer 0.57, and intermediate in spring and autumn at 0.48 and 0.39, respectively. These values provide a useful base for designing the irrigation timetable in drip-irrigation systems for terraced cherimoya plantations.
Until now, Kc values for cherimoya trees have not been defined, especially for orchard terraces. According to the findings, the amounts of water required for the fruit production period increased from the initial stage to the mid-season and decreased at the end of the late season. The most water was required at the flowering and fruit setting stages and comparatively less was required in the initial and  Figure 2C shows the changes in the average Kc for cherimoya over the two monitoring seasons estimated by the water balance from experimental lysimeters for both productions systems. The Kc at three main growing stages (flowering, fruit set, and fruit growth) was fitted by a polynomial function (Kc vs. Julian days). During the phenological stages of flowering, fruit set, and fruit growth, average Kc values were 0.63, 0.68, and 0.55, respectively.
After fruit harvest, the Kc for cherimoya trees decreased quickly to 0.23. The Kc is closely related to crop type and management practice, which may influence the plant development rate and ground coverage throughout vegetative growth [17,23]. The average annual Kc value for the cherimoya trees during the irrigation period (May-October) was 0.55. The pooled Kc values for cherimoya trees for both systems was the highest in summer 0.57, and intermediate in spring and autumn at 0.48 and 0.39, respectively. These values provide a useful base for designing the irrigation timetable in drip-irrigation systems for terraced cherimoya plantations.
Until now, Kc values for cherimoya trees have not been defined, especially for orchard terraces. According to the findings, the amounts of water required for the fruit production period increased from the initial stage to the mid-season and decreased at the end of the late season. The most water was required at the flowering and fruit setting stages and comparatively less was required in the initial and maturity (harvest) stages, with evapotranspiration being the dominant factor governing crop water requirements. The Kc is a widely used parameter in the estimation of water consumption by plants. Therefore, it is essential to determine this value for the local conditions in which it will be used since the water flow dynamics are a joint function of local climatic factors [17,24]. In this context, the phenological stage, cultivar, local climate, and soil conditions resulted in different Kc values that maintained the amount of water within the capacity level for absorption and utilization by the plant root system. Table 2 shows the impact of irrigation on cherimoya fruit yield for both production systems during the monitoring period. The average irrigation water applied in the orchard terraces for conventional and organic treatments during the first and second seasons were 3970 and 4124 m 3 ·ha −1 , respectively. Although the average fruit yield under the conventional (47.1 kg·tree −1 ) system was higher than the organic (44.1 kg·tree −1 ) system, these differences were not significantly different during the two monitoring seasons (p > 0.01). The yield in the organic plots during the first season could be attributed to the residual effect of the conventional fertilizer applied in previous years, as the plant nutrients from organic manures need time to be available for uptake by the trees. Therefore, the average fruit yield for conventional and organic farming in terraced chirimoya plantations (~280 trees·ha −1 ) were 14,308 and 13,468 kg·ha −1 , respectively. The water-use efficiency for conventional and organic production systems was 3.6 and 3.3 kg·m −3 , respectively. Table 2. Impact of irrigation (100% ET C ) on fruit yield and water-use efficiency for the two production systems. In relation to fruit size, we found no significant differences, although the fruits from conventional systems seemed to be larger.

Irrigation, Fruit Yield, and Water-Use Efficiency
The irrigation system is based on economic and technical factors, the most critical of which are the cost of the facilities and availability of the required quantity and quality of water. In this line, the economic cost of installation and technical maintenance of a drip irrigation system is high, especially in orchard terraces. The cost is usually the first key aspect that influences the decision of farmers who normally use the cheapest system. However, the cost is not necessarily related to the effectiveness of the system. In this sense, irrigation by gravity through furrows or flood irrigation are the cheapest systems. However, the water distribution is rarely uniform, which means they are inefficient. On the Granada coast, flood irrigation is widespread in most cherimoya plantations in the fertile flat areas. In the Almuñécar area, the water price per m 3 in plantations located on terraces ranges from 0.25 to 0.30 €. Drip irrigation systems have several advantages, such as the reduction of water loss through evaporation and reduction of weed development. Economic investment for the maintenance of the system (filters and various irrigation accessories) is mandatory for proper operation, and in the same way high quality water is required to avoid any possibility of suspended materials clogging the system.
Due to the importance of subtropical crops in Andalusia and its geographical advantage in the European market, it is important to study water relationships in this particular subtropical environment, since water is scarce, and optimization of water use is essential. In this context, studies are needed to understand the impacts of the water reduction supply on yields and fruit quality.
Spain is the only cherimoya producer that can provide the European markets with high quality fruit instead of traditional temperate fruit. In this sense, fruit conservation is difficult due to the highly perishable nature of fruit and the predominance of a single cultivar, Fino de Jete, that represents almost 95% of the cultivated area. One of the most important challenges for this crop is the introduction of new cultivars that expands the varietal range. The almost absolute dependence on cv. Fino de Jete, which concentrates production during October and November, creates challenges for commercialization. The technique called "superpoda" (super pruning) has been implemented, which allows production of cherimoya in winter and spring. This involves the early defoliation of the entire tree before the natural fall of the leaves; after the old period has passed, tree flowering and harvest are slightly advanced. However, its manual execution makes cultivation more expensive and profitability is not completely ensured. An increase in pollen conservation would help produce spring fruits, which would help improve shortage of flowers.
On the other hand, organic farming perceptions are divergent, but a strong consensus exists on its eco-friendly nature and inherent ability to protect the environment in contrast to the conventional system [14]. In general, organic food production costs are higher as organic farming is labor intensive and labor is costly. However, the quality of products is also higher. Efforts have been made to encourage organic farming overall and to market organic food, including cherimoya fruits. The increasing demand for organic food products, as well as the policies adopted by many governments to encourage the export of organic agri-products, are the factors driving the increase in organic food farming, which has the potential to strengthen the local economy as well as consumer health.

Leaf Mineral Content of Cherimoya Trees
The content of nutrients in relation to phenological stages throughout the production cycle for conventional and organic cherimoya trees were almost similar. In this sense, Figure 3 depicts the overall pattern of pooled leaf mineral contents for both systems, revealing similar trends with respect to N, P, and K use, with post-harvest accumulation and low levels during flowering and fruit growth, especially for N and K. These trends were the opposite for Ca and Mg. Fe, Zn, Mn, Cu, and B concentrations rose during harvest and fell during flowering. The lower energy demand of cherimoya trees during harvest presumably encouraged the accumulation of reserves. The greater foliar K (1.98%) content during this stage could provoke Mg translocation to other organs, as pointed out by Adiscott [25]. The flowering process reduces the N concentration, probably due to cell division and elongation of new spring shoots [26]. Also, the lower foliar P during flowering may be ascribed to the production of nucleic acids and coenzymes, which are fundamental for photosynthesis. K also fell during flowering, possibly because of the translocation toward new cherimoya tissues, where K is crucial to activate enzymes and regulate osmotic pressure.
The lower macro-and micro-nutrient content concentrations in some stages would also be affected by the dilution factor between foliar mass and growing fruits. The harvesting of cherimoya fruits during the harvest period activates a recovery process in foliar mineral content. In short, two clearly differentiated stages during the cycle were observed: A first stage of reduction of the foliar mineral levels during the most critical stages, followed by another of recovery once the harvesting of the fruits had begun. For some plant nutrients (P, K, and Ca), the seasonal factor was significant, with greater contents found at the harvest stage.
foliar K (1.98%) content during this stage could provoke Mg translocation to other organs, as pointed out by Adiscott [25]. The flowering process reduces the N concentration, probably due to cell division and elongation of new spring shoots [26]. Also, the lower foliar P during flowering may be ascribed to the production of nucleic acids and coenzymes, which are fundamental for photosynthesis. K also fell during flowering, possibly because of the translocation toward new cherimoya tissues, where K is crucial to activate enzymes and regulate osmotic pressure.   Table 3 shows the results for the analysis of the variance (ANOVA) concerning the effect of the production system and season on the average macro-and micro-nutrient leaf content; differences between individual means were tested using the LSD test at p < 0.01. In this sense, regarding the effects of the production systems (conventional vs. organic), no significant differences were recorded between the systems, except for Zn and Cu. However, there was a clear tendency for greater levels of NPK under conventional farming. The traditional fertilization practices in orchard terraces usually lead to excessive fertilizer applications [27]. These are unavoidable for maintaining high productivity and growth, although promoting environmental degradation and having negative impacts on crop development [28]. Under the rational use of mineral fertilizers, plant nutrients can be incorporated only when needed to endorse normal growth and productivity and an economic response to its application is generated. Although leaf nutrient analysis is the methodology used for diagnosing nutritional status in crops and the need for further mineral or organic amendments, this information is insufficient for determining the amount of nutrients that should be incorporated in the field. In this context, the reestablishment of plant nutrients removed in harvested fruits and pruning residues could be enough for the tree to maintain growth and productivity, or at least an approximation of the real requirements. However, it is crucial to consider the plant nutrients already present in the soil, and if they are in available form for uptake by root system of trees. Tree reserves in storage organs, mineralization of the soil organic matter, reuse of nutrients, nutrient supplied by the irrigation water, and other factors have to be considered.
Consequently, plant nutrient removal from orchards could be essential for estimating tree consumption and to obtain some information about to the amount of plant nutrients to be incorporated if the leaf nutrient status reveals the need for amendments.
In general, the leaf mineral content in cherimoya trees was within the ranges considered as normal by many authors [29][30][31]. Therefore, none of the nutrients could have acted as limiting factors.
On the other hand, recycled wastewater can provide both water and nutrients, and therefore could be an important tool to mitigate the adverse impact of climate change as was reported by Trimmer and Guest [32]. Also, slow and intermittent application of nutrients with recycled wastewater can increase nutrient-use efficiency [33]. However, there should be special attention to the water and nutrient balance according to Elliott and Jaiswal [34]. In addition, as stated by Long et al. [35], an elevated CO 2 concentration would result in higher yield but a lower nutritional value of food [36].

Conclusions
Severe climatic events, combining water scarcity and droughts, are predicted to increase in intensity, frequency, and geographic extent as a result of climate change. To successfully grow crops given this adverse scenario, farmers will need to adjust to less available water. The application of efficient water management strategies is key for increasing agricultural water productivity in areas with scarce water. This is the case for subtropical Mediterranean farming, particularly terraced cherimoya orchards in Southeast Spain. In this study, conventionally applied irrigation (100% ET C ) could be reduced and a water-saving program could be promoted. In this sense, the Kc values for cherimoya offer useful information for optimizing irrigation management and adjusting irrigation volume and frequency to the crop water demand. Our findings highlight the importance of rational water use in orchard terraces to promote sustainable agricultural development for subtropical fruit trees.
Regarding the crop response to the production systems, we found no significant differences between conventional and organic farming, either in fruit yield or leaf nutrient status, at least for the studied period. Thus, subtropical fruit production is feasible under precision plant nutrition and irrigation management, and continuous efforts are needed to advise farmers about rational use of water and plant nutrients according to crop requirements.