Evaluation of soil water losses under irrigation saving techniques in a semi-arid region in Tunisia Open Access

In the Mediterranean regions, especially in Tunisia, water is the main factor determining crop production. Continuous changes in climate conditions are characterized mainly by a decrease in water availability and an increase in temperature (IPCC 2021). By 2050, the amounts of precipitation in the countries of North Africa would be reduced by 20–50% compared to the current average values (Sarr 2012). In fact, Tunisia, with an average rainfall of less than 300 mm/year and an available water volume of less than 1,000 m³/inhabitant/year, is considered a country with limited water resources (Meddi & Eslamian 2021). To face this challenge, irrigation is not the only valid option, it is imperative to adopt appropriate agronomic management strategies that focus on improving the efficiency of agricultural water use (Sampathkumar et al. 2021).

Conventional irrigation, based on the maximum evapotranspiration requirements of crops (ETM), is used by farmers under conditions of limited or unrestricted water availability. This method of irrigation is now classified as a luxury water use that can be reduced with little or no drop in yield (Wanniarachchi & Sarukkalige 2022). Several irrigation water saving strategies have been used in recent years to improve water productivity such as supplemental irrigation (Satognon et al. 2021) and deficit irrigation (Elhani et al. 2019). Other scientific alternatives are also possible: the integration of antiperspirant products in the soil that can reduce the consumption of irrigation water, the limitation of the phenomena of percolations and water runoff by incorporating materials into the soil with high water retention ‘polymers’, and the use of drought-tolerant varieties (Wilson et al. 2009). With regard to this multitude of the aforementioned strategies, some researchers have oriented themselves through the knowledge base on the physiological and morphological mechanisms of the plant toward the development of the new technique ‘partial root zone drying’ (PRD) (Xie et al. 2012; Qin et al. 2018; Martínez-Romero et al. 2019).

In Tunisia, potato cultivation occupies an area of 25.6 thousand hectares, or 15% of the areas reserved for market gardening. It occupies the second place after tomato cultivation. The production of potatoes as a strategic crop in Tunisia requires more and more frequent use of water and therefore requires more efficient management of irrigation. PRD irrigation, which stands out for its potential to save water, its optimization is of major interest, especially in conditions of limited water availability.

Numerical modeling of the water storage in soils is a key factor in optimizing irrigation management (Kanzari et al. 2020). Among the modeling tools, the Hydrus-1D model (Šimůek et al. 2016) constitutes a promising software for managing the irrigation under different techniques such as drip irrigation (Yang et al. 2019) and deficit irrigation (González et al. 2015). However, a lack of research studies is noticed according to the use of the Hydrus-1D model in irrigation with the PRD technique (Nouna et al. 2016).

The objective of this study is to evaluate (i) the performance of the Hydrus-1D model for simulating water flow under irrigation saving techniques in the conditions of semi-arid Tunisia; (ii) the soil water storage and deep percolation and (iii) the effect of the increase of temperature on the soil water dynamics.

The choice of plant material was made on a seasonal potato variety ‘Spunta’. The planting of the potato is carried out on 20 February 2021, in a plot of 1,393.2 m². The plot is divided into three blocks, each block has 10 experimental units, and each unit has an area of 46.44 m². Inside the unit, we have six lines the spacing between them is 0.83 m, the length of a line is 9 m and the distance between the plants on the same line is 0.3 m so the density planting is 4.01 plants/m². The used experimental design is the Complete Random Block with one main factor which is water regime with four levels of irrigation: FI (Full Irrigation 100%): irrigation with 100% of crop water requirements. DI (Irrigation 75% ETM): Regulated deficit irrigation with the application of 75% of water crop requirements from the tuberization stage. PRD_Right (Irrigation 50% of water crop requirements on the right side): Irrigation by partial drying of the roots. PRD_Left (Irrigation 50% of water crop requirements on the left side): Irrigation by partial drying of the roots. Drip irrigation was used for all treatments to deliver the required amounts of irrigation water.

The gravimetric method allows the monitoring of the water stock in the soil. It consists of taking soil samples at different periods of the potato cycle and at different depths using an auger. This method was used to monitor the soil water dynamic by measuring the soil water profile in the corresponding dates for the calibration and the validation of the model in the next section.

The approach of parameter estimation involves the minimization of an objective function that considers all deviations between the measured and simulated data, with the simulated results controlled by the adjustable parameters to be optimized. The parameter vector b includes the parameters α, n, θ_r, θ_s and K_s. Each inverse problem was restarted several times with different initial estimates of optimized parameters and the run with the lowest value of the objective function was assumed to represent the global minimum (Kanzari et al. 2021).

The simulation period is 98 days, from 20 February 2021 until 28 May 2021. The dates of output are the 11th, 24th, 57th, 71st, 88th and 98th day. The profile is simulated up to a depth of 60 cm. Initial conditions can be expressed in terms of water content or pressure load and in terms of initial concentration. In our study, the initial state was expressed in terms of soil water content. Water content measurements were taken every 20 cm. The depth of 60 cm was simulated with a uniform intermodal space of 0.2 cm.

Climatic data (reference evapotranspiration and rainfall) are obtained from the climatic station located at the Potato and Artichoke Technical Center (CTPTA) in ‘Saïda’. The evapotranspiration values were estimated from the Mabia. ET₀ software using the following parameters: temperatures (maximum and minimum in °C), relative humidity (maximum and minimum in %) and the wind speed (m/s).

The study of the variation in soil water content showed:

• – An increase in soil moisture in the soil layer up to 30 cm in depth. The texture of the soil is more clayey and thus increases the retention of water.

• – A common profile, that of 18 May 2021 and whose water content values are quite close to those of the average field capacity (37%). The measurement of this profile was carried out only 2 days after the irrigations. The soil remained quite moist except for the case of the PRD_Right where the soil is in a drier state. The roots under water stress caused by the PRD irrigation strategy quickly absorbed the supplied amounts of water.

• – That in the case of the FI, where 100% of the crop water requirements were applied for irrigation, the dynamics of soil water content reflected the cycles of saturation/drying out under the effect the irrigation doses, the evaporation and root water uptake without any stress. The soil dries out quickly after surface irrigation, while in the deepest layers, the soil humidity values showed less variation and revealed deep water percolation (Michot et al. 2003; Er-Raki et al. 2021).

• – In the case of DI, where 75% of the crop water requirements were provided during irrigation, the water profiles showed a lower drying rate than that in the FI case and the soil retained more moisture, especially in the deepest layers. The reduction water quantities had the effect of reducing the infiltration beyond the root zone.

• – In the case of the PRD irrigation strategies, the soil water profiles indicated a more significant state of drying due to the reduction in the doses of irrigation. In the case of PRD_Left, the value of the water content of the soil at the level up to 30 cm in depth did not vary. The root system of crops can be more developed on the left side and therefore water consumption is more optimal (Wu et al. 2022).

The calibration of Hydrus-1D was carried out over a total duration of 98 days with a daily time step (from 20 February 2021 to 28 May 2021). The times of the output files are the 11th, 24th, 57th, 71st, 88th and 98th day and correspond to the water profiles produced in the case of the FI irrigation treatment. The geometric domain is 60 cm in depth.

The hydraulic properties (α, n, θ_r, θ_s and K_s) were estimated by inverse modeling. During the optimization process, the values of these coefficients were adjusted so that the measured soil water content values were as close as possible to the simulated values. The calibrated soil hydraulic parameters are shown in Table 1.

As a direct result of the successful calibration and validation, the Hydrus-1D model is able to provide soil water storage variations in the studied unsaturated zone of the soil (Figure 5).

After tuberization (approximately 62 days), the variation in SWS for the PRD_Right showed a shift in the level of the irrigation peaks compared to the other treatments. Root development on this side may be the cause.

This effect is also studied according to the variation of the bottom cumulative fluxes in Figure 8.

To deal with the problems of water scarcity in arid and semi-arid regions, the use of water saving techniques at the plot scale is an obligation in order to optimize crop yields by minimizing the irrigation water consumption and maximization of production. The Hydrus-1D model was able to simulate water dynamics, as indicated by the low calculated NRMSE values (Karlsson et al. 2015). The good correlation between the measured and simulated soil water content values shows that the inverse modeling succeeded in optimizing the hydrodynamic parameters. As a direct result of the successful calibration and validation of the Hydrus-1D model, variations in soil water storage in the unsaturated zone were investigated. The obtained showed results that after tuberization (approximately 62 days), the variation in the water stock for PRD_Right showed a lag in the irrigation peaks compared to the other treatments. Root development on this side may be the cause. In addition, the study of the measured water profiles revealed the same result. However, to better understand the evolution of soil water under PRD irrigation technique, it is necessary to conduct in-depth study with a soil water movement modeling tool such as that of Hydrus-2D in order to understand the evolution of soil water dynamics under PRD irrigation technique.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.