In ﬂ uence of the soil water retention curve type and magnetic water treatment on lettuce irrigation management responses

This study aimed to evaluate the in ﬂ uence of the magnetized water use on the lettuce irrigation management responses, and based on generated data, to evaluate by simulation the in ﬂ uence of soil water retention curve type on the lettuce irrigation management responses. This work was divided into three stages: 1 – determination of ﬁ eld and laboratory soil water retention curves; 2 – lettuce crop irrigation management experiment using soil water retention curve with ﬁ eld data, evaluating different soil water tensions to start irrigation and different water types (magnetically treated water, and non-treated water); and 3 – estimate of the irrigation management responses (simulation) using the soil water retention curve performed in the laboratory (using non-treated water), compared with the experimental results (stage 2). The use of magnetically treated water determined the soil moisture maintenance for a longer time and less irrigation events, leading to less water being applied and electricity consumption. The use of soil water retention curve derived from the ﬁ eld data determined less water and electric energy consumption in the Lucy Brown lettuce irrigation, in comparison to the simulated use of the soil water retention curve from laboratory data.


INTRODUCTION
Over the years, water availability has been a concern of many economic sectors. In irrigated agriculture, researchers have been currently concerned with the development of techniques and equipment for water use optimization.
Irrigation with magnetically treated water has been scientifically evaluated over the years, and has brought positive improvements on water use reduction, qualitative and productive benefits for some crops production (Maheshwari & Grewal ; Surendran et al. ; Yusuf & Ogunlela ).
According to some authors, the magnetic treatment modifies the physical structure of water molecules, and their chemical composition, altering the hydrogen bonds, van der Waals forces, and sizes of the water clusters, decreasing the surface tension and increasing the viscosity Based on the exposed information, it is expected that the use in irrigation of water subjected to magnetic treatment, associated with the use of the soil water retention curve with data from field testing for irrigation management, will determine reduction in water amount and in electricity consumption. In this sense, this study aimed to evaluate the influence of the magnetized water use on the lettuce irrigation management responses, and based on generated data, to evaluate by simulation the influence of soil water retention curve type on the lettuce irrigation management responses. Initially, two forms of soil water retention curve construction were performed (in field, and in laboratory) -Stage 1. The objective was to evaluate the influence of the soil water retention curve determination method on the lettuce crop irrigation management responses. Subsequently, a lettuce crop irrigation management experiment was performed using the soil water retention curve made by field data -Stage 2. It was evaluated different combinations of soil water tensions to start irrigation and types of water (magnetically treated water, and non-treated water). Finally, using the data generated in the field experiment, an estimate (simulation) of the irrigation management responses was performed, using the soil water retention curve performed in the laboratory -Stage 3. These results were compared with the Stage 2 results.

Location of the experiment and water magnetic treatment
The field essays were conducted in a greenhouse (  The measured air temperatures inside the greenhouse during the experiment were: maximum of 31.0 ± 2.4 C, minimum of 19.0 ± 1.8 C, and average of 25.0 ± 1.5 C.
A Sylocimol Residencial magnetizer from the Timol Indústria e Comércio de Produtos Magnéticos (Uberlândia -MG, Brazil) was used to magnetize the water. This equipment is composed of alternating magnets and covered by a stainless-steel protection, which submits the water to a magnetic field of 3,860 Gauss, with the capacity to magnetize 1,000 L in one hour of exposure (TIMOL ). The equipment was allocated and kept inside a 500 L water tank throughout the experiment.

Determination of the soil water retention curve
The soil of the experimental area was classified as Typic Hapludox (Soil Survey Staff ). Table 1 shows the results of the chemical and physical soil analysis, referring to the depths of 0-20 cm and 20-40 cm.
To determine the soil water retention curve in the field, two beds in the area (1.2 × 2.4 m) were selected: one for the magnetically treated water evaluation (MW), and one for non-treaded water evaluation (OW). The beds were delimited by a PVC sheet 20 cm high above the ground surface to avoid water runoff, and 50 cm below the ground surface to prevent lateral water flow. The objective was to promote only drainage and evaporation of water during the test ( Figure 2).
To evaluate the soil water retention characteristics that plants were submitted, the soil water retention curves were made after incorporating the planting fertilization. In order to promote repetitive assessment, in each bed six tensiometers were installed at 12.5 cm (depth defined to indicate irrigation), six at 25 cm (to check the amount of water in the effective depth of the lettuce roots), and three at 40 cm (evaluation depth to ensure saturation of the soil profile).
The soil was subjected to saturation (zero reading on the tensiometer at all depths) using 30 drip emitters with flow rate of 4 L h À1 . Once saturation condition was reached, water application was interrupted, and the beds were covered with impermeable material to prevent water evaporation and allow only drainage. The beds were kept covered until field capacity condition occurred in the soil. After reaching equilibrium (field capacity condition), the impermeable material was removed in order for the soil to dry until the recommended reading limit of the tensiometer, i.e. À80 kPa. The maintenance of the water column inside the tensiometers was performed routinely.
Tensiometers readings were performed concomitantly with the soil moisture measurement in laboratory by the standard method (gravimetric method). The soil water retention curve was determined to a depth of 12.5 cm. Soil volumetric water content was obtained from gravimetric water content multiplying it by the soil bulk density after  bed making (1.14 ± 0.05 g cm À3 ). Sampling was more frequent on the first day (0, 3, 6, 12 and 24 hours after saturation), and later evaluations were daily.
A digital tensimeter (model TENSIMETER, from the company Hidrodinâmica Irrigation. URL: http://tensiometro. com.br/) was used to read the tensiometers. Equation (1) was used to determine the soil water tension or the matric potential from the reading of the tensiometer.
where Ψ m is the soil water tension or matric potential (kPa), L is the tensiometer reading (kPa), and h is the water column height in the tensiometers (cm).
The F Test (p 0.05) was performed to evaluate the water type (MW and OW) influence on the behavior of the field soil water retention curves.
A soil water retention curve was also determined in laboratory, using soil cores made from deformed soil samples, from saturated condition to a tension of À98 kPa, being À0.98, À1.96, À3.92, À5.88, À7.84 and À9.8 kPa in hanging-water funnel, and À98 kPa in pressure plate apparatus. To this, we used six soil deformed samples submitted to a 2 mm sieve, from 0 to 0.20 m deep to prepare the soil cores with bulk density of 1.14 g cm À3 , the same founded in field conditions. In order to compare its hypothetical irrigation management responses with the results from the actual use of the soil water retention curve performed in field (comparison only with non-treated water).
For the field (magnetically treated water -MW, and non-treated water -OW) and laboratory data (non-treated water -OW), regression analysis (p 0.05) were performed to define the most appropriate mathematical model of the relationship between volumetric soil moisture and soil water tension. Linear models, second degree polynomial, power and logarithmic models were evaluated.

Lucy Brown lettuce irrigation management
The irrigation management evaluation was carried out in being the irrigation process performed to raise the soil moisture to field capacity condition. As a limit of À80 kPa is indicated for proper functioning of the tensiometers (Azevedo and Silva 1999), À70 kPa was defined as the treatment with the greatest distance from the field capacity tension. The soil water retention curve used for the crop management was the one derived from the field with non-treated water. In one repetition of each experimental combination (type of water and soil water tension), three tensiometers at 12.5 cm was used to manage the irrigation. The readings of the tensiometers were performed daily at 9 am and 3 pm.
A drip irrigation system (distribution uniformity -DU equal to 95.4%) was used for irrigation, with self-regulating dripper ClickTif NaanDanJain (average flow of 2.14 ± 0.08 L h À1 ), with a spacing of 30 cm (16 emitters per bed).
Additionally, two irrigation lines were used per bed, spaced by 60 cm. Equation (2) was used to estimate the irrigation water depth. The irrigation time was calculated by Equation (3). It was assumed 95% application efficiency, and according to Yuri et al. () an effective depth of the root system of 25 cm (250 mm).
where LI is the irrigation depth (mm), θ cc is the volumetric soil moisture at field capacity (cm 3 cm À3 ), θ l is the volumetric soil moisture at the time of reading (cm 3 cm À3 ), z is the effective depth of the root system (mm), E a is the application efficiency (decimal), and DU is the uniformity of distribution of the irrigation system (decimal).
where T is the irrigation time (min), A is the bed area (m 2 ); q a is the average flow of the emitters (L h À1 ), and n e is the number of emitters per bed.
Electric energy consumption was estimated for each treatment by Equation (4)

Irrigation management responses
Evaluation of field data The field data were analyzed to evaluate the influence of the irrigation water magnetic treatment on the lettuce irrigation management responses. For the irrigation management was used the soil water retention curve constructed in field. We evaluated the following irrigation management responses: soil water tension data during the experiment for each treatment; number of irrigation events (NI); average interval between irrigations (Ii); total applied irrigation depth during the crop cycle (ΣLI); and electric energy consumption estimate (EC).
For the soil water tension data comparison between each treatment, an Analysis of Variance on Ranks was used, by the Kruskal-Wallis Test with 5% of probability, followed by the Tukey's Test, also at 5% of significance for pairwise multiple comparison.

Simulation of irrigation management responses
Using the soil water tension data measured during the experiment, simulation of irrigation management responses was estimated using the soil water retention curve made in laboratory. The objective was to evaluate the influence of the soil water retention curve data source on the lettuce irrigation management responses. Total applied irrigation depth during the crop cycle (ΣLI) and electric energy consumption estimate (EC) was estimated.

Soil water retention curve
Soil water retention curve performed in field Table 2 presents the summary of the regression analysis (p 0.05) of the soil moisture data (θ l ) as a function of the time (t) for the field test.
All mathematical models presented significant fitting to the data for both types of water (except the linear model for magnetized water use). The third-degree polynomial model was selected for both types of water due to the higher value of the determination coefficient (greater than 85%). Figure 3 shows the behavior of the data fit and the respective model for each treatment.
Comparing the fitted mathematical models using the F Test (p 0.05) it was found significant differences between data generated by the models, demonstrating that the type of water significantly influenced the behavior of soil Almost all mathematical models represented significantly (p 0.05) the soil moisture data as a function of the soil water tension for both types of water (except the linear model for magnetized water use). The logarithmic model was selected to represent the soil water retention curve for both types of water. Figure 4 shows the behavior of the data fitting for each treatment, as well as the fitted mathematical models.
Comparing the fitted mathematical models by the F Test (p 0.05) it was not found any significant differences between data generated by those, demonstrating that the type of water did not significantly influence the soil water retention behavior.
Considering the estimated values of soil moisture at field capacity (0.356 cm 3 cm À3 for magnetized water, and 0.340 cm 3 cm À3 for non-treated water) for the fitted mathematical models presented in Figure 3, tensions equivalent to 7.22 kPa for magnetized water and 10.54 kPa for nontreated water were observed. Although there are no significant differences between models, it was interesting *significant at 5% probability, nsnot significant at 5% probability. to note that the soil moisture in field capacity using water subjected to magnetic treatment was obtained with less retention force (31.5% less). According to some authors, the cause of this effect may be associated with the surface tension decrease and water viscosity increase (Toledo As there was no statistical difference between the curves changing the type of water, the soil water retention curve performed in field conditions for non-treated water was selected for the irrigation management due its higher value of determination coefficient for the curve fitting (R 2 ¼ 94.3%). Considering the soil water tension equal to 10 kPa in module to represent the field capacity in the irrigation management, this resulted in a soil moisture of 0.342 cm 3 cm À3 . Table 4 shows the summary of the regression analysis (p 0.05) of the fitting of the volumetric moisture data as a function of the soil water tension, referring to the laboratory analysis.

Soil water retention curve performed in the laboratory
Almost all mathematical models represented the data significantly (p 0.05), being the exception the linear model. It was selected the power model ( Figure 5) due its highest value of determination coefficient (R 2 ¼ 94.1%).
Applying the F Test (p 0.05) between the fitted models of soil water retention curve performed in the field and in the laboratory, both with non-treated water, it was observed that the obtained models were statistically different.
Assuming the same soil water tension value regarding the field capacity found in the field test (10 kPa in module), this would result in a volumetric soil moisture *significant at 5% probability, nsnot significant at 5% probability.

Irrigation management responses
Soil water tension data during the experiment, number of irrigations (NI) and average interval between irrigations (Ii) Figure 6 shows the estimated soil water tensions (Ψ m ) during the experiment for the different experimental combinations.
It was possible to observe less data variability using magnetized water in almost all treatments.   Kareem & Adeniran () also observed a reduction in  Ψ ccsoil water tension referring at field capacity (10 kPa). Total applied irrigation depth during the crop cycle (∑LI) Table 7 shows the Tukey Test result (p 0.05) to evaluate the total applied irrigation depth for different combinations of soil water tention to start irrigation and types of water, considering the soil water retention curve built in the field (actual applied values).
It was observed a significant reduction in the total applied irrigation depth using magnetized water for irrigation (with T2 treatment as exception). The reduction was 34.62, 18.62 and 26.34% for T1, T3 and T4, respectively. Selim et al. () also observed reduction in the amount of water applied (25% reduction) for wheat crops irrigated with magnetically treated water compared to non-treated water.
Regarding the influence of soil water tension to start irrigation on total irrigation water depth, it was observed that T3 > T1 > T2 > T4 for non-treated water, and T2 > T3 > T1 > T4 for magnetized water. In both types of water, T4 treatment obtained the lowest irrigation depths. Koetz et al.
() points out that generally, longer intervals between irrigations determine lower total applied irrigation depth.
Electric energy consumption (EC) kWh m À2 cycle À1 ), due to the lower total applied irrigation depth and longer interval between irrigations. The highest consumption was reached using non-treated water with 40 kPa for soil water tension to start irrigation (0.0177 kWh m À2 cycle À1 ), being the treatment that demanded a larger total applied irrigation depth.

Simulation of irrigation management responses
Comparison between applied and simulated irrigation depth during the crop cycle Table 9 shows the Tukey Test result (p 0.05) to evaluate the soil water retention curve data source (fieldapplied  Electric energy consumption (EC)

Practical observations
The result of this research presents, to the scientific and growers Community, a discussion that should be the focus of today's research on irrigated agriculture, which is the rational and sustainable use of water. In terms of water sub- The use of soil water retention curve derived from the field data determined less water and electric energy consumption in the Lucy Brown lettuce irrigation, in comparison to the simulated use of the soil water retention curve derived from laboratory data. This finding presents itself as an opportunity for the growers themselves to make their retention curves in the irrigation area, with the potential to achieve high irrigation efficiencies, providing savings in water and electricity.