Introduction

Food production for consumption and export has tremendously increased over the last 50 years. This agricultural activity used more than 40% of irrigation in the world depending on groundwater resources (Godfray et al., 2010). The situation is more serious in arid and semiarid regions. In fact, these regions are unceasingly growing and including more of the earth’s land surface (Schlesinger et al., 1990) with an area of 30% (Dregne, 1991; Scanlon et al., 2006) and contain some of the most populated areas in the world (Brown et al., 2005). Water resources evolution scenarios predict a global water shortage in the horizon of 2050 with decreasing rainfall and limited groundwater recharge (Milly et al., 2005; Döll and Florke, 2005).

Sustainability is the strategic purpose of all water policies and management plans. As water demand is growing continuously with population increase, water availability has become a crucial concern. In addition, in vulnerable arid and semiarid areas, human populations and ecosystems are relying intensively on groundwater resources to satisfy their essential water needs. The combined effect is inevitably affecting groundwater sustainability through over-pumping of aquifers, lowering of groundwater table level and speeding up quality deterioration. These facts will in turn affect water availability and cause several environmental problems such as soil salinization, lower crop yield and ecosystem degradation (Schoups et al., 2006; Praveena et al., 2012; Vadiati et al., 2018) as well as socioeconomic difficulties (Mays, 2013).

In the last two decades, special care has been given to the growing interest in groundwater sustainability, especially under arid and semiarid climate. In fact, great effort has been devoted to various interpretations depending on whether groundwater reservoirs are irreversibly or permanently lost as a result of large pumping rates that lead to environmental and ecological damages (Custodio, 2002; Morsy et al., 2018; Ranjan, 2012). Groundwater recharge was also considered as a sustainability indicator (Healy & Cook, 2002; Scanlon et al., 2002; Gleason et al., 2012). However, several studies argued that natural recharge alone cannot orient policies toward sustainable groundwater resources management (Balleau, 2013). Instead, they considered that another component should be taken into consideration: anthropogenic aquifer recharge (AAR), which comprises incidental recharge as a result of human activities that unintentionally increases recharge (Maliva, 2020).

Many studies tried to assess the anthropogenic impacts on groundwater resources in arid and semiarid regions using diverse methods. Zume and Tarhule (2011) studied the impact of anthropogenic pumping and recharge variability on an alluvial aquifer in semiarid northwestern Oklahoma using numerical simulation. Leitão et al. (2013) used water quality data with geoelectrical methods to assess the diffuse contamination of a karstic aquifer in Algarve region, Portugal. Li et al. (2019) tried to understand the hydrochemical evolution of groundwater in a regional aquifer system in Yinchuan plain in China as a result of anthropogenic activities using stable isotopes and major chemical elements. Singha et al. (2019) identified the groundwater zones that are affected by anthropogenic contaminants in Raipur District, central India, using multi-criteria decision analysis—analytical hierarchy process (MCDA–AHP) technique combined with spatial analysis (i.e., using GIS). Nonetheless, most of these studies have focused on regional groundwater flow systems; they barely investigated the impact of anthropogenic activities on groundwater resources from a local-scale viewpoint. Moreover, there is still a lack of interest in the use of an integrated approach based on a multi-discipline intersection to clarify the AAR processes in poorly monitored groundwater flow systems.

In Tunisia, where climate is dominantly arid and semiarid, the availability of shallow groundwater resources is linked usually to recharge abundance, which is related to the infiltration of rainwater. In addition, more than 40% of groundwater resources are over-used in agriculture, drinking water supply and industry (Frija et al., 2014). Consequently, a continuous decline was observed in most aquifers of Tunisia, which is associated usually with water quality deterioration (Closas et al., 2018). The Mornag aquifer, located in the northeast of Tunisia, constitutes a typical case of a stressed aquifer; it is pumped heavily to satisfy the socioeconomic development of one of the most productive provinces of Tunis. However, during the last two decades, groundwater in this region has been under several constraints related to quantity and quality. In fact, a water table drawdown of more than 10 m was noticed and associated with groundwater salinization of many parts of the aquifer (Jarraya–Horriche and Wolfgang, 2019). The plain of the Wadi El Hma constitutes the central part of the basin where a sudden increase in groundwater salinity was registered during the last 5 years. Most of the intensive agriculture activities are almost concentrated in this region. Thus, an aquifer sustainability problem is posed and needs to be understood and solved. To the authors’ best knowledge, small-scale investigation of groundwater sustainability issues was not previously addressed in this region. AAR was not also studied, and the question of groundwater salinization is still debated by the scientific and technical community dealing with the Mornag plain. Therefore, the objectives of the present work were to (a) understand the issues related to groundwater resources in a local scale along the Wadi El Hma flood plain, (b) clarify their recharge processes, (c) explain their salinity origins and (d) assess their suitability for irrigation use as principal assets toward aquifer sustainability in the region.

Study Area

The El Hma floodplain is a part of the Mornag plain located in the junction between the governorates of Ben Arous, Nabeul and Zaghouan. With an area of ~ 250 km2, the Mornag plain is located in the northeastern part of Tunisia within latitudes 36° 34′ to 36° 46′ N and longitudes 10° 6′ to 10° 24′ E (Fig. 1). More precisely, this sector is limited to the NW by the lower plain of Mejerda, to the East by the Gulf of Tunis, to the SE by the Tunisian Dorsale which includes the mountains of Sidi Salem, Zaghouan, Ressas and Bougarnine, to the West by the hills of Rades, Megrine, Bir Kassâa and Tella mount and to the SW by the hills of Khelidia. The focus area is the El Hma floodplain, which is particularly standing on the piedmonts of Ressas and Majoura Mountains. It is characterized by a flat topography ranging between 11 and 30 m (a.m.s.l.) in its downstream part and exceeding 200 m (a.m.s.l.) in its upstream zone. From the climate viewpoint, the region belongs to the semiarid stage with a mean annual precipitation of 465 mm. The most humid month is December with a mean rainfall of 72 mm, and the least is July with mean precipitation less than 7 mm. The region is characterized by mild winter and hot and dry summer. The annual average temperature is ~ 19.1 °C, and the mean annual evaporation is 1700 mm/yr (INM, 2018).

Figure 1
figure 1

Study area location

Full size image

The plain is principally drained by the El Hma River, a non-perennial channel with irregular runoff, usually so-called Wadi in arid and desertic regions of North Africa and Middle East. The upstream part of this Wadi lodges a hill dam (El Hma dam) with a capacity of 12.1 Mm3, used principally to secure irrigation water and secondary for rare artificial recharge of the aquifer throughout water releases during the humid periods. A check dam was built on the Wadi bed to reduce its water flow and enhance its infiltration (Fig. 1). The watershed of the Wadi El Hma is one of the most fertile zones of Tunis and Ben Arous provinces. It was famous for more than a century for vine, citrus, peach and olive tree orchards. However, since 2018, a sudden increase in groundwater salinity reaching more than 6 g/l was registered especially in the wells located in the vicinity of the Wadi El Hma and which was hindering crop yield in the neighboring farms. The El Hma watershed also includes some industrial activities, mainly a cement manufactory exploiting a limestone quarry located on the piedmont of the Ressas Mountain. Field investigation revealed that the factory was discharging its wastewater with a salinity of 16.5 g/l into the Wadi El Hma through a sewer pipe (Fig. 1). Mixed with the rare natural water runoff, this effluent may considerably affect the groundwater quality of the floodplain.

Geological and Hydrogeological Settings

The geological context of the Mornag plain has been the subject of several studies (Boutib, 1998; Bujalka et al., 1972; Farhat, 2011; Jauzein, 1967; Soussi et al., 2000; Turki, 1985). The region is considered as an alluvial plain where Neogene and Quaternary sediments significantly dominate. Continental and fluvial detrital formations result from the dismantling of the surrounding reliefs. From a lithostratigraphic point of view, the outcropping formations have ages ranging from Eocene to Quaternary with notable lateral variations in facies (Fig. 2). Triassic, Jurassic and Cretaceous outcrops are located at the surrounding mountains. The El Hma plain is composed dominantly of Quaternary deposits of silts, clays and sands. Quaternary formations are distributed heterogeneously in the plain, but their maximum thickness is roughly at the central part with depth exceeding 600 m (Farhat, 2011). The Wadi beds are formed by recent alluvial deposits.

Figure 2
figure 2

Geological map of the El Hma plain (from the 1:500 000 scale geological map of Tunisia, Ben Haj Ali et al. (1985))

Full size image

In regional hydrogeological scale, the Mornag plain lodges a complex multilayered aquifer composed of Quaternary alluvial deposits and sandstone lenses of Miocene and Oligocene age (Ennabli, 1980; Farhat, 2011; Mamou, 1994; Merhebene, 1998). The shallow aquifer follows a syncline structure and lies in angular discordance on the Miocene and Oligocene Formations (Moussa et al., 2020). With estimated reserves of 18.7 Mm3/yr (Jarraya–Horriche, 2004), the unconfined unit is found usually at depths ranging between 10 and 45 m (Farhat, 2011). Its natural recharge is ensured by meteoric water and the Wadi runoffs (Ennabli, 1980; Rekaya, 1987). The deep aquifer is a more explicitly multilayered system composed of ancient Quaternary deposits located in the central part of the plain, the Miocene and Oligocene sandstone series as well as the Eocene limestone formations that outcrop along the plain boundaries. Productive series are located usually at depths reaching more than 100 m (Farhat, 2011; Hechemi, 1989). In the focus area of the El Hma watershed, shallow and deep aquifers of Mio–Plio–Quaternary units are exploited mostly by dig and tube wells at depths ranging from 15 to 120 m. The transmissivity of these heterogeneous deposits ranges between 0.1 × 10−3 and 8 × 10−3 m2/s, and their storage coefficient varies between 6 × 10−3 and 7.3 × 10−5 (Rekaya, 1987; Jarraya–Horriche, 2004). The unconfined groundwater contour map of the focus area shows a shallow flow mostly influenced by topography (Fig. 3). In fact, three main flow axes can be distinguished: (i) a SW → NE flow from the recharge zone at the Majoura mount piedmont toward the central part of the plain; (ii) a S → N groundwater flow from the hill dam on the way to the plain center describing a possible underflow beneath the Wadi; and (iii) a SE → NW groundwater flow from the recharge zone at the piedmont of Ressas Mountain to the plain center. The Wadis contribute to the aquifer recharge, especially in the immediate upstream of the check dam, which was built originally to enhance the Wadi water infiltration during humid periods or artificial hill dam water releases.

Figure 3
figure 3

Spatial distribution of water samples’ total dissolved solids superposed with groundwater contour lines

Full size image

Materials and Methods

Water Sampling and Analysis

In total, 22 water samples were collected from surface water bodies (dam and Wadi) and groundwater (shallow and deep aquifers) during a field survey in June 2019. The sampling strategy consisted of taking all the existing wells in the vicinity of the Wadi. We considered a buffer zone not exceeding 500 m from the Wadi, wherein we collected the water samples. According to Ben Farhat (2011) (2011, Charef et al. (2012) and Ben Moussa et al. (2020), there are no significant seasonal variations in groundwater quality between wet and dry seasons. Therefore, a single sampling survey was considered to furnish reliable information regarding the AAR occurrence in the El Hma plain.

The water samples were taken after the stabilization of electrical conductivity (EC), and they were collected in clean high-density polyethylene (HDPE) bottles. Temperature, pH, EC and groundwater level were measured in situ. In the laboratory, bicarbonate (HCO3−) was analyzed using the titration method with sulfuric acid. Major chemicals ions (anions Cl−, SO42−, NO3− and cations Na+, Ca2+, Mg2+, K+) were analyzed by ion liquid chromatography (IC 732) and atomic absorption spectrometry (PerkinElmer AAnalyst 200) at the Laboratory of Georesources, Water Research and Technologies Center (Borj Cedria Technopark, Tunisia). Stable isotopes (18O and 2H) were analyzed using an IsoPrime, GV Instruments, mass spectrometer at the HydroSciences LAMA laboratory (Montpellier, France) with 0.06 and 0.8‰ error measurement, respectively, for oxygen-18 and deuterium. Stable isotopes contents were quantified in the usual δ notation, thus:

$$\delta^{2} H\left( \permille \right) = \left[ {\frac{{\left( {\frac{{{}^{2}H}}{{{}^{1}H}}} \right)_{{{\text{sample}}}} }}{{\left( {\frac{{{}^{2}H}}{{{}^{1}H}}} \right)_{{{\text{standard}}}} }} - 1} \right] \times 10^{3}$$
(1)
$$\delta^{18} O\left( \permille \right) = \left[ {\frac{{\left( {\frac{{{}^{18}O}}{{{}^{16}O}}} \right)_{{{\text{sample}}}} }}{{\left( {\frac{{{}^{18}O}}{{{}^{16}O}}} \right)_{{{\text{standard}}}} }} - 1} \right] \times 10^{3}$$
(2)

As shallow groundwater is mainly used for irrigation in the Wadi El Hma plain, a set of quality assessment indices were calculated (Table 1).

Table 1 Indices of groundwater suitability for irrigation: sodium adsorption ratio (SAR), percentage of sodium (%Na), permeability index (PI) and electrical conductivity (EC)
Full size table

Electrical Resistivity Tomography (ERT) Survey

ERT is a geophysical method based on the resistivity contrast. It is used to determine the resistivity of earth materials and their distribution in the subsurface. This approach is particularly applied in the case of large resistivity contrast to delineate the interface between saline and freshwater (Bauer et al., 2006; Chabaane et al., 2017; McInnis et al., 2013).

ERT measurements were performed in the same period as the piezometric and geochemical survey (June 2019) using an ABEM SAS-4000 resistivity meter. Wenner array configuration was used to delineate accurately the interface between freshwater and brackish/saline water (Baharuddin et al., 2013; Loke, 2011). It also provides a good sensitivity for vertical heterogeneities, allowing better differentiation of plumes’ structures (Doetsch et al., 2012; Kemna et al., 2002). ERT profiles were performed using 64 stainless steel electrodes set 5 m apart. The obtained resistivity data were inverted using RES2DINV software based on the smoothness-constrained least-square method (Loke & Barker, 1996; Loke & Dahlin, 2002).

Results and Discussion

Hydrochemical Data

Geochemical tracers including major chemicals and stable isotopes are essential tools for understanding the hydrogeochemical functioning of aquifers. Hydrochemical and isotopic characteristics of the waters sampled from the Wadi El Hma plain are summarized in Table 2. Water temperature and pH varied from 17 to 27 °C and 7.5 to 9.9, respectively. The EC of the groundwater sample ranged from 2.1 to 8.2 mS/cm, and that of surface water varied between 0.6 (dam) and 27 mS/cm (Wadi water). The high EC values of Wadi water may be due to saline water discharge from a neighboring cement factory, which rejects its wastewater effluent directly into the dry Wadi channel through a sewer pipe. The spatial distribution of water salinity revealed that wells located in the proximity of the Wadi and within the groundwater flow pathways particularly show high values of total dissolved solids (TDS) exceeding 6 g/l (Fig. 3). This suggests the occurrence of an AAR process consisting of infiltration of saline Wadi water and its migration within the saturated zone. For all the water samples collected during the survey, Na+ was the dominant cation and the order of mean concentration of cations was Na+  > Ca2+  > Mg2+  > K+, while Cl− was the dominant anion and the order of mean concentration of anions was Cl− > HCO3− > SO42− > NO3−.

Table 2 Physicochemical, major elements and stable isotopes data from water samples collected from the Wadi El Hma plain
Full size table

Classification of Water Samples

Classification of water samples into groups facilitates segregation of relevant analytical data to understand the sources of dissolved constituents. Alternatively, it can serve to comprehend the mineralization processes that occur during AAR. Presenting the hydrochemical data in charts and graphs like the Piper diagram helps in classifying water samples according to their geochemical facies (Piper, 1944). It is among the simplest and widely used diagrams to assess groundwater quality in many regions in Tunisia (e.g., Chekirbane et al., 2013; Lachaal et al., 2014; Mlayah et al., 2011). The Piper diagram of the water samples from the study area shows that Wadi water and brackish groundwater samples located in the vicinity of the Wadi are characterized by Na–Cl facies, whereas the remaining dam water and fresh groundwater samples were of the Ca–SO4–Cl type (Fig. 4). Focusing on the ternary diagram of anions, a tendency toward the Cl pole was observed, suggesting the existence of a mixing phenomenon as a principal conductor of water salinization. A tendency of migration toward the Na pole is observed in the cation ternary diagram, which may indicate the occurrence of cation exchange during the salinization process. In a regional flow scale, a slightly similar behavior was reported by previous studies that focused on the whole aquifer of the Mornag plain (Farhat et al., 2010; Moussa et al., 2020).

Figure 4
figure 4

Piper diagram for water samples

Full size image

Salinization Process

The bivariate diagram of sodium versus chloride shows two main groups (Fig. 5). The first group of samples indicates a flawless correlation between Na and Cl, in which the fresh shallow water and deep groundwater samples (particularly the Ca–SO4–Cl type) fall on the 1:1 line. This suggests halite dissolution as the origin of these two elements. The second group of samples comprises the saline Wadi water and brackish groundwater samples (Na–Cl type); they deviate from the halite dissolution line. Thus, besides the geological effect, another phenomenon such as salt intrusion likely governs the salinization of water in the area. The evolution from fresh groundwater (i.e., Ca–SO4–Cl type) to a salty groundwater enriched by Na+ and Cl− is proof of mixing with a saline water body, which may be the formerly infiltrated saline Wadi water in the vicinity of the check dam.

Figure 5
figure 5

Plot of Na+ versus Cl−

Full size image

The cation exchange between sodium and calcium or magnesium ions may explain the excess of Na+ concentration (Stimson et al., 2001). In addition to the dominance of Na+ and Cl− in groundwater samples belonging to the group “Shallow well (3),” these samples show relative enrichment in bicarbonate. The excess in HCO3− may cause the release of Na+ into the solution by exchange reaction with the exchange sites (Jalali, 2007). In order to confirm the cation exchange phenomena as a conducting process for groundwater salinization in the Wadi El Hma plain, the plot of Na+  + K+ – Cl− versus Ca2+  + Mg2+ – SO42− – HCO3− was examined (Fig. 6). The difference between the sum of sodium and potassium from one side and the chloride from the other side represents the amount of Na+ and K+ gained or lost relative to that provided by the geological effect (halite dissolution). In contrast, the difference between the sum of calcium and magnesium from one side and the sum of sulfate and bicarbonate from the other side points out any gained or lost Ca2+ and Mg2+ relative to that provided by gypsum, calcite and dolomite dissolution. In the absence of any exchange with saline bodies, all data should be plotted close to the origin (McLean et al., 2000). However, in the occurrence of the cation exchange phenomenon, the relation between these two parameters should be linear and significantly negatively correlated. Among the plotted samples in Figure 6, the brackish and saline groundwater samples show negative correlation of the two parameters: The increase in Na+  + K+ may be related to a decrease in Ca2+  + Mg2+ or an increase in HCO3− + SO42−. Therefore, the main process controlling the salinization of groundwater in the Wadi El Hma plain would be cation exchange. This phenomenon suggests the presence of a saline water body that consists of the infiltrated saline Wadi water. The cation exchange responsible such kind of salt intrusion may be described as:

$$\raise.5ex\hbox{$\scriptstyle 1$}\kern-.1em/ \kern-.15em\lower.25ex\hbox{$\scriptstyle 2$} {\text{Ca}}^{{{2} + }} + {\text{ Na}} - {\text{X }} \to {\text{ Na}}^{ + } + \, \raise.5ex\hbox{$\scriptstyle 1$}\kern-.1em/ \kern-.15em\lower.25ex\hbox{$\scriptstyle 2$} {\text{Ca}} - {\text{X}}_{{2}}$$
(3)
Figure 6
figure 6

Plot of Na+  + K+ – Cl− versus Ca2+  + Mg2+ – SO42− – HCO3−

Full size image

In aquifers holding clayey materials, natural softening or ion exchange may take place (Hounslow, 2018). As shown in Eq. 3, calcium is absorbed by the exchanger X whereas sodium is released to subsequently form Na–Cl type water.

Tracing of AAR with Stable Isotopes

The stable isotopes composition of the water samples from the Wadi El Hma plain is plotted in Figure 7. Isotope contents range from − 5.7 to − 1.1‰V-SMOW and between − 32.9 and − 9.5‰V-SMOW for δ18O and δ2H, respectively. For the samples depleted with a d-excess showing no or low evaporation process (> 7.5), a distinction can be made between the deep groundwater (more depleted − 5.25‰ and − 31.2‰, respectively, for oxygen-18 and deuterium (5 points)) and the shallow aquifer (− 4.49‰ and − 27.4‰, respectively, in oxygen-18 and deuterium (4 points)). Compared with the mean isotope content of rainfall at Tunis station over the 1967–2006 period (https://www.iaea.org/resources/network/global-network-of-isotopes-in-precipitation-gnip), the little more depleted value in deep aquifer (− 4.38 ‰ and − 25.4‰) could be due to the recharge zone with altitudes higher those that constitute the hills delimiting the catchment (400–700 a.m.s.l.). This formation corresponds to the deep groundwater whereas the recharge zone of shallow aquifer has altitudes lower than 200 a.m.s.l. The water flowing in the Wadi in June represents only the sewage from the cement factory and it shows isotope content close to the isotope value of the deep groundwater. In fact, the water used in production and cleaning processes in the cement factory is pumped into the deep groundwater and then rejected into the sewer system to reach the Wadi El Hma.

Figure 7
figure 7

Delta diagram (δ18O vs δ2H) of the water samples from the Wadi El Hma plain

Full size image

Based on the delta diagram, two clusters can be distinguished. The first group, which clearly deviates from the GMWL: δ2H = 8 δ18O + 10 (Craig, 1961; Rozanski et al., 1993) and the LMWL of Tunis–Carthage GNIP: δ2H = 8 δ18O + 12.4 (Zouari et al., 1985), is composed of enriched isotopes composition of the dam water and mainly phreatic groundwater samples located in the upstream region of the plain near the El Hma hill dam. These samples are also affected by evaporation, which is marked with a low d-excess (1.7–5.2) and follows a trend line with the equation: δ2H = 4.17 δ18O – 5.22 (R2 = 0.94). This suggests a possible role of aquifer recharge from the hill dam water body in the upstream zone, which presents moderate EC (2.1–2.8 mS/cm). In contrast to the middle zone near the check dam, evaporated shallow water has higher EC (3.5–5.5 mS/cm) highlighting another complementary process, which could be infiltration from the Wadi or water irrigation. With EC = 2.6 mS/cm, deep groundwater samples located in the upstream zone far away from the Wadi (A5 and A6) are not affected by evaporation. In this zone, the deep aquifer seems totally isolated from the shallow aquifer. The two evaporated samples from a deep aquifer close to the Wadi (A4 upstream and A18 middle zone) show higher mineralization (4.8–5.3 mS/cm). In this case, hydraulic connection with shallow groundwater and the Wadi has to be considered. The second group with depleted stable isotopes content includes some points in deep groundwater isolated from the surface and with moderate EC, except one (A12) with the highest EC value (8.2 µS/cm), saline Wadi water and brackish shallow aquifer samples located near the check dam in the middle part of the plain. They almost fall on the GMWL and LMWL, confirming a rapidly infiltrated meteoric water origin. Paradoxically, these samples (except isolated points in deep aquifer) are characterized by high salt content. As previously mentioned, the Wadi water is simply discharged saline water coming from the cement factory, pumped from a deep well, used in production and cleaning processes and then rejected into the sewer system to reach Wadi El Hma. The similar stable isotopes content of this effluent compared to the wells located in the vicinity of the check dam suggests the role of this effluent in AAR and groundwater salinization in this zone.

The biplot of δ18O versus Cl- is an effective tool for differentiating salinity origins. It has been applied in several AAR cases (e.g., Chekirbane et al., 2013; Kammoun et al., 2018; Mejri et al., 2018). The water samples affected by saltwater coming from the cement factory were characterized by an increase in chloride content versus depleted stable isotope content. However, the water samples affected by evaporation had an opposite behavior marked by enrichment in stable isotopes content associated with reduced chloride concentration linked to low mineralization of surface water (Clark & Fritz, 1997). Plotted in Figure 8, the water samples from the Wadi El Hma plain are clustered in two main groups: (i) dam water and fresh groundwater located in the upstream of the plain near the hill dam, which are mainly affected by evaporation; and (ii) saline Wadi water and brackish groundwater located near the check dam, where salt intrusion as salinization process is confirmed. Some points could be the result of mixing of the two processes or may be the result of infiltration of irrigated water; in contrast, other points from deep groundwater showed no evidence of anthropogenic influence.

Figure 8
figure 8

Diagram of salinization origin (δ18O vs Cl−)

Full size image

Monitoring of Brackish/Saline Groundwater Plume Using Geoelectrical Data

Geoelectrical methods such as ERT are effective tools that provide accurate images of saline and brackish water plumes in subsurface environments (e.g., Singha & Gorelick, 2005; Chabaane et al., 2018; Doetsch et al., 2012). Two ERT profiles were carried out in the most salinized part of the aquifer, in the immediate vicinity of the Wadi El Hma and just upstream of the check dam (Fig. 9). The multi-electrode ABEM SAS-4000 Terrameter was used with a set of 32 electrodes and a regular electrode spacing of 10 m. The Wenner α configuration was used because of its excellent signal-to-noise ratio (Naudet et al., 2004). The first profile (P1) with a total length of 310 m was laid along the main groundwater flow direction from the Wadi toward the left side of the plain in order to delineate the saline/brackish water plume in the affected area. The second profile (P2) was longitudinally set to the Wadi meander close to its left bank along a distance of 315 m from the check dam for the purpose of detecting the contribution of the saline Wadi water in aquifer salinization.

Figure 9
figure 9

Location of the ERT profiles

Full size image

Profile P1 showed relatively high resistivity values (40–60 Ω.m) near the Wadi left bank, which are due to the relatively dry silty sand (unsaturated vadose zone). Low resistivity values (< 10 Ω.m) appeared at 50 m far from the Wadi within depths of more than 30 m. These represent the interface between fresh and saline groundwater (Fig. 10). The deep well A12 pumps water within this plume, explaining its high EC (8.2 mS/cm). The shape of the brackish/saline water plume is not regular due to the abundant heterogeneous Quaternary deposits. In fact, located at a distance of less than 100 m from A12, the shallow well (25 m depth) A11 shows a lower EC value (5.2 mS/cm) because the saline plume emanates from deeper horizon. The low resistivity values found at 5 m depth near A11 well represent soils irrigated with the brackish water (Fig. 10).

Figure 10
figure 10

Inverted resistivity profiles: (a) Profile P1; (b) Profile P2

Full size image

Profile P2 show low values of resistivity (< 10 Ω.m) near the check dam at depths of less than 5 m. These data represent the inlet of saline water toward the saturated zone. Almost all the subsurface environment in depths more than 5 m below the Wadi bed are favorable ways of the AAR of the aquifer.

Mapping and Assessment of Water Suitability for Irrigation

Water resources in the Wadi El Hma watershed are used mainly for irrigation. Therefore, an assessment of the suitability of these waters for irrigation is essential in order to preserve soil and increase crop yield in the region. Certain indices are used to assess water stress using physical and hydrodynamic parameters of aquifers (Nayyeri et al, 2021). However, in the present case, the suitability assessment was performed using the following indices based on hydrochemical characteristics of the available water: sodium adsorption ratio (SAR), percentage of sodium hazard (%Na) and permeability index (PI).

SAR is a ratio of sodium ions to the sum of calcium and magnesium ions present in a water sample. It is used to estimate the potential of sodium ions to accumulate in the soil primarily at the expense of the other major cations as a result of regular use of sodic water (Bouwer & Idelovitch, 1987). Water suitability for irrigation using SAR was classified by Richards (1954) (Table 1). Percentage of sodium hazard (%Na) is an index proposed by Wilcox (1948) to categorize any source of water for irrigation use. Sodium cation is known to make chemical bonding with soil to reduce unsaturated water movement (Ayers and Westcot, 1985). Crop growth and yield are particularly reduced in sodium-affected soil (Todd and Mays, 2004). The use of sodium-enriched water in irrigation leads to ions tending toward clay particles by removing calcium and magnesium cations through a base-exchange reaction that reduces water movement capacity within the irrigated soils (Rengasamy & Olsson, 1993; Rhoades et al., 1968). The PI is used to assess water movement aptitude in long-term irrigated soils based on their contents in bicarbonate, sodium, calcium and magnesium ions (Doneen, 1964). Besides, water EC should obviously be taken into consideration because it represents TDS. Salt excess can increase the osmotic pressure of soil solution especially under drought conditions leading to major difficulties for plant uptake and growth (Bauder et al., 2011; Zaman et al., 2018).

The four parameters chosen to assess suitability for irrigation do not have the same importance. Therefore, a multi-criteria decision approach, namely analytic hierarchy process (AHP) developed by Saaty (2000), was applied to determine the weight of each parameter. AHP is used widely in assessing water potential in complex aquifers (e.g., Zaree et al, 2019). In the present case, a pairwise comparison matrix was established (Table 3). A 1–9 point scale was adopted to assess the relative importance in terms of suitability for irrigation among the chosen parameters. The weight values of the parameters are given in Table 3; they reflect the relative importance of each criterion. The relative weights are calculated using the eigenvector corresponding to the largest eigenvalue (Mishra et al., 2015). In addition, the consistency of the pairwise comparison matrix should be controlled by means of a control value called the consistency ratio (CR). The CR expresses the random probability of values obtained in a pairwise comparison matrix (Malczewski, 1999). If the value of CR is less or equal to 10%, the pairwise comparison decisions are sufficiently consistent (Saaty, 2005). The CR for the present matrix is 7%; thus, the obtained weights can be considered.

Table 3 Pairwise comparison matrix and weights of parameters based on AHP
Full size table

Once the SAR, %Na and PI have been calculated and reclassified using scores in a 1–4 scale (1 = good; 2 = permissible; 3 = doubtful; 4 = unsuitable), an irrigation suitability index (ISI) was calculated as:

$${\text{ISI}} = \mathop \sum \limits_{i = 1}^{3} s_{i} w_{i}$$
(4)

where si and wi represent the score and weight, respectively, of parameter i. Then, a map of spatial distribution of ISI was created using ArcGIS software (Fig. 11). This map shows clearly that all the Wadi water samples were unsuitable for irrigation. However, the dam water, deep groundwater and shallow groundwater samples located in the upstream region of the plain were mostly good to permissible for irrigation. Samples located in the vicinity of the check dam were classified doubtful regarding their suitability for irrigation.

Figure 11
figure 11

Spatial distribution of ISI in the Wadi El Hma plain

Full size image

Integrated Conceptual Modeling and Its Implications

Following the above-discussed integrated approach involving all the analytical techniques used in the present study, a conceptual model for the understanding of the hydrogeochemical functioning of the aquifer as well as for detailing the processes of groundwater recharge was developed (Fig. 12). The El Hma plain is characterized by a shallow groundwater flow type that is controlled mainly by topography. The piedmont of the Ressas Mountain and the relatively high relief areas constitute the main natural recharge zones. Wadi El Hma, which is almost dry all the year, constitutes the downstream channel where dam water is occasionally discharged to enhance aquifer recharge using a check dam. A discharged saline water with TDS of 27 mS/cm is mixed with the Wadi water. This water mixture often constitutes the main runoff in the Wadi, and it easily reaches the constructed check dam where it infiltrates and migrates toward the shallow aquifer. The saline water plume was detected at a distance of 200 m far from the check dam and at depths around 30 m. The affected wells are pumping brackish water classified as doubtful for use in irrigation. However, it is used effectively to irrigate lands and, consequently, it can be considered as an anthropogenic recharge for the local shallow aquifer.

Figure 12
figure 12

Integrated conceptual model of groundwater flow system at the Wadi El Hma plain

Full size image

The integrated approach based on hydrodynamic, geochemical and geophysical techniques was extremely useful for understanding the aquifer functioning and the impact of anthropogenic recharge sources. In fact, we proved that, in a poorly monitored hydrological system (such as the case of the El Hma plain), the combination of multi-disciplines is essential to furnish a complete picture of the system and to make it easier for decision-makers to propose suitable countermeasures for water sustainability.

Conclusions

Flood plain aquifers are critical sources of water in arid and semiarid regions of the world. However, increased pressure on these aquifers including over-pumping and AAR threatens their sustainable use. This study implemented a multidisciplinary approach based principally on a combination of geochemical tracers and geoelectrical method, and this approach was used to elucidate the AAR processes in a reduced-scale agricultural floodplain and to assess its impacts on groundwater contamination and suitability for irrigation. The obtained piezometric and geochemical data reveal an interaction between the Wadi and the aquifer, especially in the downstream part of the plain near the check dam. At these locations, the Wadi plays the role of a losing system where saline water is infiltrated throughout its permeable bed, particularly in the immediate meander located before the check dam. This area is the most affected by salt intrusion due to the presence of the existing managed aquifer recharge infrastructure (check dam). Electric resistivity tomography profiles were extremely useful for delineating and monitoring the salt water plume. Transversely and following the main groundwater flow direction, infiltrated saltwater reached more than 50 m depth and contributed to the contamination of the neighboring wells where TDS exceeded 6 g/l. Longitudinally, low resistivity values, which reveal the infiltrated saline Wadi water throughout the left bank, exist in the immediate proximity of the check dam, which represents the inlet toward the saturated zone. The brackish/saline water plume generally starts from a depth of 5 m under the Wadi bed. This plume highlights the role of the Wadi and check dam in the AAR of the aquifer. The hydrochemical data proved that cation exchange was the main salinization process of groundwater in the affected area. Isotopic tracers confirmed the evidence of salinization in this area and gave additional information regarding the recharge process in the upstream region where the hill dam contributes to aquifer recharge through direct interaction with its surface water body. Groundwater sustainability in the plain is threatened because of the present status of AAR. In fact, most of the sampled wells were classified doubtful for irrigation and the water in these wells is a threat to soils and crops.

Overall, the obtained results of this study demonstrated the important influence that anthropogenic sources exert on groundwater sustainable use in the El Hma plain. These results should draw the attention of decision-makers and stimulate their reaction to stop any contamination effluent in the Wadi and enhance the artificial recharge by hill dam water release to tentatively flush the salt and dilute its effect on the aquifer.