Water and Nutrient Balance in an Ornamental Cascade Cropping System

: Seedlings of Chrysalidocarpus lutescens , Dracaena deremensis and Dracaena marginata were grown in plastic containers ﬁlled with sphagnum peat-moss to assess the effects of three different water systems on plant growth, water saving and nutrient removal during the experimental period. The experiment lasted for 8 weeks and consisted of three water systems. These consisted of an open draining system fertigated with a standard nutrient solution (system T 0 ) and two closed systems: sequential reuse of the leachate (system T 1 ) and sequential reuse of the leachate with the addition of H 2 O 2 (system T 2 ). Over the course of the experiment, samples of water and supplies generated in each water treatment were collected weekly, and from these data water volume and nutrient loads were calculated. The addition of H 2 O 2 to the leachate resulted in an enhancement in plant dry weight in Dracaena deremensis and Dracaena marginata. Regarding anion loads (Cl − , NO 3 − , H 2 PO 4 − , SO 42 − ) in these water systems, there was a removal rate of 42%, 28%, 27% and 28%, respectively, in the closed systems compared to the open system. For the cation loads (Na + , K + , Ca 2+ and Mg 2+ ) in these water systems, there was a removal rate from 28% to 29%, respectively, in the closed systems compared to the open system. O 2 to the leachate enhanced the growth of these species, resulting in a higher dry weight. The comparison between open and closed systems revealed a greater water saving along with greater nutrient removal in the closed systems compared to the open draining system. These results suggest that future experiments on cascade cropping systems should include different volumes of H 2 O 2 to check if the ameliorative effect is even better.


Introduction
The increasing surface of greenhouses in Southern Europe and the Mediterranean region has led to several side effects such as environmental pollution, mainly associated with the release of water and nutrients as wastewater into the environment [1,2]. This has become even more aggravated, since the majority of soilless cropping systems in the greenhouses in this area are open or free draining-draining the leachate directly into the soil [3,4]. The leakage of nutrients, such as nitrate and phosphorous, from fertigation in greenhouses into the environment is usually greater than the thresholds established by environmental guidelines, causing the contamination of water resources [5,6].
High-tech, sophisticated greenhouse systems can obtain a return in the use of inputs such as water and nutrients, through, for instance, the collection of the drainage generated by a crop for further irrigation in another, or the same crop [7,8]. Nevertheless, the implementation of this type of methodology has several disadvantages, such as the high investment cost and the continuous replenishment of recirculating nutrient solutions due to the increasing electrical conductivity of the groundwater used for irrigation [9,10]. Moreover, reduced yields associated with the increase in salinity in the root zone, and the risk of pathogen propagation throughout the fertigation system, entail more drawbacks for its implementation [11,12]. Another management strategy for reusing the drainage and reducing the pollution generated is the sequential reuse of the leachate to irrigate increasingly salt-tolerant crops [13]. This fertigation strategy is known as a cascade cropping system, and it is based on the collection of the drainage from beneath one crop to fertigate the salt-tolerant crop that is above it in the series, aiming to almost entirely reduce the volume of leachate of the last crop [14]. It is worth mentioning that in this fertigation system, each subsequent crop must be able to tolerate the accumulated salts from the previous crop in order to avoid plant damage and death [15].
Although the implementation of this type of growing system (cascade cropping) is still in its early stages, based on recent literature, there are some models with different combinations of open and closed systems. For instance, Muñoz et al. [16] conducted an experiment with horticultural crops in which beef tomato was the donor crop, or the first stage of the cascade cropping system, and lettuce, tomato and endive were then the receiving crops, or second stage of the cascade cropping system. Similarly, García-Caparrós et al. [17] conducted a cascade cropping system that involved simultaneously growing one horticultural and one ornamental species, with Cucumis melo being the donor crop and Rosmarinus officinalis the receiving crop. Analogously, Plaza et al. [18,19] carried out an experiment with Citrullus lanatus being the donor crop and Cordyline fruticose the receiving crop. In this latter experiment with horticultural and ornamental species, the differences between an open and closed systems under the implementation of the cascade cropping system were also assessed.
Besides the disinfectant power of hydrogen peroxide, its employment in irrigation water or fertigation to improve oxygenation in the root zone is increasing among growers, since it improves both the yield and the crop quality [20][21][22]. Nevertheless, there is little information about the supplies of H 2 O 2 in an ornamental cascade cropping system and its effects on crops. Therefore, in this study, a pot experiment with Chrysalidocarpus lutescens, Dracaena deremensis and Dracaena marginata was established in order to discern the effects of different water treatments on plant growth, water saving and nutrient removal under cascade cropping systems, including the application of H 2 O 2 in fertigation treatment.

Plant Material and Growing Conditions
Previous experiments with similar conditions were conducted with the same species in the facilities of the University of Almeria (36 • 49 N, 2 • 24 W). Seedlings of Chrysalidocarpus lutescens, Dracaena deremensis and Dracaena marginata were purchased from a local nursery and then transplanted into 1.5 L plastic containers filled with sphagnum peat-moss. The climatic conditions during the experimental period were recorded with HOBO SHUTTLE sensors (model H 08-004-02, Onset Computer Crop., Bourne, MA, USA). The recordings reported an average temperature of 16.5 ± 1.5 • C, a relative humidity (RH) of 55.6 ± 2.9%, and photosynthetically active radiation (PAR) of 55.4 ± 4.4 µmol m −2 s −1 . All the experiments lasted for 8 weeks following the advice given by local nursery growers to produce saleable plants of all three species.

Experimental Water Systems
The experiment consisted of three water systems. These included an open draining system fertigated with a standard nutrient solution (system T 0 ), and two closed systems: sequential reuse of the leachate (system T 1 ) and sequential reuse of the leachate with the addition of H 2 O 2 (system T 2 ). Each water system consisted of four replicates with four plants (one plant per pot) per species, and stage with a planting density of 6 plants per m 2 . In the T 0 water system, seedlings of each species were fertigated with a standard nutrient solution for ornamental plants (fertigation treatment 1) recommended by Jimenez and Caballero [23]. In the two closed systems, C. lutescens plants (first stage) were fertigated with the standard nutrient solution (fertigation number 1), but in each closed system the leachate generated by this species was used in a different way. In water system (T 1 ), the sequential reuse treatment, the drainage of C. lutescens plants (fertigation number 2) was used directly to fertigate D. deremensis, from which the drainage generated (fertigation number 3) was reused to fertigate D. marginata. In the case of T 2 , fertigation numbers 2 and 3 were supplied with H 2 O 2 (1.2 M) at 1% (v/v) to result in fertigation treatments numbers 4 and 5, respectively ( Figure 1). and Caballero [23]. In the two closed systems, C. lutescens plants (first stage) were fertigated with the standard nutrient solution (fertigation number 1), but in each closed system the leachate generated by this species was used in a different way. In water system (T1), the sequential reuse treatment, the drainage of C. lutescens plants (fertigation number 2) was used directly to fertigate D. deremensis, from which the drainage generated (fertigation number 3) was reused to fertigate D. marginata. In the case of T2, fertigation numbers 2 and 3 were supplied with H2O2 (1.2 M) at 1% (v/v) to result in fertigation treatments numbers 4 and 5, respectively ( Figure 1).

Chemical Composition of Fertigation Treatments
Collection of the drainage discharged by each species in the different fertigation treatments was performed weekly. The nutrient solutions and drainage samples consisted of 5 mL aliquots, filtered using membrane filters (0.45 µm), and frozen for further analysis. To determine the pH and EC values of the samples obtained, both a pH meter and a conductivity meter were used (models Milwaukee pH52 and C66) (Milwaukee Instruments, USA), respectively. The determination of the chemical composition of the nutrient solution or drainage (anions and cations concentration) was performed by high-performance liquid chromatography (HPLC) (model 883 Basic IC Plus, anions ion exchange column model Metrosep A SUPP 4, cations ion exchange column model Metrosep C4 100, IC conductivity detector range (0-15,000 mS cm −1 ); Metrohm, Herisau, Switzerland)) as described by Csaky and Martinez-Grau [24].
After recording these determinations, and together with the values of volume noted for the nutrient solution and drainage generated (expressed in L m −2 ), the amounts of nutrient supplied and discharged from each crop (expressed in grams per m −2 ) were calculated by multiplying them. The comparison between the water systems was conducted taking into account the distribution ratio of each water system and the planting density.

Model Development
The comparison between the water treatments required the determination of the number of plants and the water supplies needed in each water system, which were calculated following the equations given by Garcia-Caparros et al. [17,25]. Data used for the

Chemical Composition of Fertigation Treatments
Collection of the drainage discharged by each species in the different fertigation treatments was performed weekly. The nutrient solutions and drainage samples consisted of 5 mL aliquots, filtered using membrane filters (0.45 µm), and frozen for further analysis. To determine the pH and EC values of the samples obtained, both a pH meter and a conductivity meter were used (models Milwaukee pH52 and C66) (Milwaukee Instruments, USA), respectively. The determination of the chemical composition of the nutrient solution or drainage (anions and cations concentration) was performed by high-performance liquid chromatography (HPLC) (model 883 Basic IC Plus, anions ion exchange column model Metrosep A SUPP 4, cations ion exchange column model Metrosep C4 100, IC conductivity detector range (0-15,000 mS cm −1 ); Metrohm, Herisau, Switzerland)) as described by Csaky and Martinez-Grau [24].
After recording these determinations, and together with the values of volume noted for the nutrient solution and drainage generated (expressed in L m −2 ), the amounts of nutrient supplied and discharged from each crop (expressed in grams per m −2 ) were calculated by multiplying them. The comparison between the water systems was conducted taking into account the distribution ratio of each water system and the planting density.

Model Development
The comparison between the water treatments required the determination of the number of plants and the water supplies needed in each water system, which were calculated following the equations given by Garcia-Caparros et al. [17,25]. Data used for the model and its calibration were collected from previous experiments conducted under similar conditions. From the results obtained in these experiments, the following equations were defined in order to determine the number of plants needed in each stage of the different water Treatments (1), and the nutrient solution supplies for the species in each stage of the different drainage water Treatments (2): The use of subscript ( i−1 ) refers to the previous stage in the different drainage water systems.
The main inputs of this model are the following: water uptake of the species in each stage of the different water systems (Wupt i ), and the percentage of leachate in the previous stage of the different drainage water systems (Xlea i ). The main outputs of the model are the number of plants needed in each stage of the different drainage water systems (P i ), and water supply for the species in each stage of the different drainage water systems (Wsup i ).
Finally, with the values obtained using Equations (1) and (2), the distribution ratio (DR) was determined. In our experiment, the results obtained showed that the distribution ratio (DR) of each species (expressed as number of plants per m 2 ) that can be simultaneously grown in each water system was 1/1.34/0.42 for both system T 1 and T 2 , which was also applied to T 0 to compare the data recorded between the different water systems.

Statistical Analysis
The experiment was analysed as a completely randomized block design, where the values obtained were considered as independent replicates. Unifactorial variance (ANOVA) was performed, and Fisher's Least significant difference (LSD) tests (p < 0.05) were used to assess the differences between the water treatments. All the statistical analyses were performed using Statgraphics Centurion XVI.II (Statpoint Technologies, Inc., Warrenton, VA, USA).

Chemical Composition of Fertigation Treatments
The chemical composition of the fertigation treatments showed a significant increase in the pH and EC of fertigation treatments Nos. 2, 3, 4 and 5 compared to fertigation treatment No. 1, mainly associated with the increase in concentrations of Cl − , NO 3 − , Na + , K + , Ca 2+ and Mg 2+ (Table 1).

Plant Biomass
Fertigation of D. deremensis and D. marginata with T 1 resulted in a decline of 13% and 20% in DW, respectively, compared to the control treatment (T 0 ), whereas the fertigation of these species with T 2 resulted in a significant increase of 16% and 18% in DW, respectively, compared to T 0 ( Table 2).

Application of Model Development to Water and Nutrient Management
Considering the water volume and nutrient loads of each system, systems T 1 and T 2 both consumed 45.2 L m −2 , resulting in a water saving of 28% compared to T 0 (62.9 L m −2 ). The T 0 system generated a leachate volume of 17.7 L m −2 , whereas neither T 1 nor T 2 generated any leachate.
The anion loads in the drainage were also calculated; comparing T 1 and T 2 with T 0 , there was a removal of 42% of Cl − (9.7 g m −2 in T 0 and 5.6 g m −2 in T 1 and T 2 ), 28% of NO 3 − (23.4 g m −2 in T 0 and 16.8 g m −2 in T 1 and T 2 ), 27% of H 2 PO 4 − (4.2 g m −2 in T 0 and 3.1 g m −2 in T 1 and T 2 ), and 28% of SO 4 2− (12.1 g m −2 in T 0 and 8.6 g m −2 in T 1 and T 2 ), respectively. There was no generation of pollution in T 0 , whereas in T 1 and T 2 there was a discharge to the environment of 12.5 g m −2 of Cl − , 11.6 g m −2 NO 3 − , 1.3 g m −2 of H 2 PO 4 − and 2.7 g m −2 SO 4 2− . As for the cation loads, the comparison of T 1 and T 2 against T 0 reported that there was a removal of 28% Na + (3.8 g m −2 in T 0 and 2.7 g m −2 in T 1 and T 2 ), 28% K + (7.4 g m −2 in T 0 and 5.3 g m −2 in T 1 and T 2 ), 28% Ca 2+ (5.0 g m −2 in T 0 and 3.6 g m −2 in T 1 and T 2 ), and 29% Mg 2+ (2.1 g m −2 in T 0 and 1.5 g m −2 in T 1 and T 2 ), respectively. There was no generation of pollution in T 0 , whereas in T 1 and T 2 there was a discharge to the environment of 3.8 g m −2 of Na + , 4.3 g m −2 of K + , 4.6 g m −2 of Ca 2+ and 1.8 g m −2 of Mg 2+ (Figure 2).  is an open draining system fertigated with standard nutrient solution and two closed systems (sequential reuse of the leachate (system T1) and sequential reuse of the leachate with the addition of H2O2 (system T2)). Blue bar is the water supply, grey bar is the drainage water and white bar is the reusable drainage water for the irrigation of the next crop in the system.

Discussion
The chemical analysis of the different water systems revealed a significant increase in EC and pH values as well as in Cl − , NO3 − , Na + , K + , Ca 2+ and Mg 2+ concentrations in the fertigation treatments compared to the standard nutrient solution. The increasing value of these parameters can be associated with the reuse of the drainage and the accumulation of these nutrients due to differences in water uptake between the treatments, as reported by Carmassi et al. [26] and Grewal et al. [27]. The lowest values of H2PO4 − and SO4 2− reported in fertigation treatments 2 and 4 can be associated with a preferential uptake of these nutrients for Dracaena deremensis, as reported Zulfiqar et al. [28].
Fertigation with the sequential reuse of the leachate resulted in a clear reduction in dry weight in D. deremensis and D. marginata plants. Nevertheless, the additional supply of H2O2 to the leachate notably increased the dry weight in both species. Similar results concerning the ameliorative effect of the addition of H2O2 have been reported in different crops such as wheat [29], zucchini, soybean, and cotton [30], mainly due to the increase in oxygenation conditions.
It is necessary to point out that the addition of H2O2 did not result in significant differences in water saving and nutrient removal between the closed systems assessed (T1 and T2). The comparison between the closed systems (systems T1 and T2) and the open system (T0) showed a clear reduction in water volume as well as nutrient loads. The reduction in water volume in our experiment (28%) was greater compared to the results is an open draining system fertigated with standard nutrient solution and two closed systems (sequential reuse of the leachate (system T 1 ) and sequential reuse of the leachate with the addition of H 2 O 2 (system T 2 )). Blue bar is the water supply, grey bar is the drainage water and white bar is the reusable drainage water for the irrigation of the next crop in the system.

Discussion
The chemical analysis of the different water systems revealed a significant increase in EC and pH values as well as in Cl − , NO 3 − , Na + , K + , Ca 2+ and Mg 2+ concentrations in the fertigation treatments compared to the standard nutrient solution. The increasing value of these parameters can be associated with the reuse of the drainage and the accumulation of these nutrients due to differences in water uptake between the treatments, as reported by Carmassi et al. [26] and Grewal et al. [27]. The lowest values of H 2 PO 4 − and SO 4 2− reported in fertigation treatments 2 and 4 can be associated with a preferential uptake of these nutrients for Dracaena deremensis, as reported Zulfiqar et al. [28]. Fertigation with the sequential reuse of the leachate resulted in a clear reduction in dry weight in D. deremensis and D. marginata plants. Nevertheless, the additional supply of H 2 O 2 to the leachate notably increased the dry weight in both species. Similar results concerning the ameliorative effect of the addition of H 2 O 2 have been reported in different crops such as wheat [29], zucchini, soybean, and cotton [30], mainly due to the increase in oxygenation conditions.
It is necessary to point out that the addition of H 2 O 2 did not result in significant differences in water saving and nutrient removal between the closed systems assessed (T 1 and T 2 ). The comparison between the closed systems (systems T 1 and T 2 ) and the open system (T 0 ) showed a clear reduction in water volume as well as nutrient loads. The reduction in water volume in our experiment (28%) was greater compared to the results obtained by other researchers in cascade cropping systems (around 18%) [17,25]. An increase in the reduction in the water volume for irrigation is highly advantageous in areas of water scarcity, such as the Mediterranean area [31]. Considering high pollutants, such as nitrate and phosphates, the comparison between the water systems revealed a clear removal of these nutrients with values of around 28%, which are slightly lower in comparison to the data reported in other cascade cropping systems (around 30%) [17,25]. These high rates of nitrate and phosphate removal discharged into the environment are beneficial for the adequate maintenance of water resources, since both nutrients are the main cause of water eutrophication processes [32]. In our experiment, in the case of sulfates (28%), the rate of removal is similar to the range proposed by other researchers in a cascade cropping system with Cucumis melo and Rosmarinus officinalis [32] in the different water systems assessed (from 27 to 39%). Regarding chloride, the removal rate in our experiment (42%) was much greater than in previous experiences [17,25], with values of around 12%. This higher rate of chloride removal can be associated with the uptake of these salts for an enhancement of crop growth, as reported by Garcia Caparros et al. [33], since the choice of the species for a cascade cropping system is mainly based on the species' increasing salt tolerance capacity.
As for the cation loads, it is necessary to point out that in our experiment the rate of removal between the different cations assessed was similar, with values of around 28%. These removal values were within the ranges noted by other researchers in similar growing systems, such as those reported by Garcia-Caparros et al. [17,25]. The removal of cations, especially sodium, is of crucial importance in order to avoid the salinization of the soil and the aquifers [34]. Moreover, the higher rate of removal of these cations can be ascribed to the nutrient requirements of the species selected, which are involved in crucial physiological processes in plant growth and development [35][36][37].

Conclusions
The results obtained show that fertigation with the sequential reuse treatment resulted