Our selected study area falls in Tyre district. Selected sites represent Ras El-Ain springs and wells located upstream in 9 surrounding villages of the lower sub-catchment area of Ras El-Ain springs. The study sites are Ras El-Ain springs, and artesian wells used mainly for irrigation.

The area surrounding the wells is distinguished by vast fragile land, cultivated with different cultures and orchards. The wells are used mainly for irrigation and sometimes for the surrounding homes as drinking and domestic water.

Ras El-Ain is a place abound with immense springs, reservoirs and aqueducts 6 km south of Tyre, 77 km south of Beirut, in the South Governorate of Lebanon ( Figure 1 ). The place lies in a green and fertile plain, about 1 km from the seashore. It is a popular touristic destination, owing to its artesian wells fed by underground springs and collected in stone reservoirs that have been conserved through the ages ( Robinson & Smith, 1841 ). It has been the main source of water for ancient Tyre since the Phoenicians days. One of the reservoirs fed the arched aqueducts in the Roman period, which once stretched all the way to Tyre. Remains of these aqueducts, exhibiting tight and excellent masonry, with round arches and a continuous cornice above them, can still be seen ( Robinson & Smith, 1841 ), and a short stretch of the original aqueduct is still used nowadays in Tyre waterworks.

Ras El-Ain is part of the best-preserved stretches of sandy coastlines in southern Lebanon. It has considerable spectacular and recreational value, and its artesian wells are an important heritage site that give rise to several valuable freshwater habitats. It is part of Tyre Coast Nature Reserve (TCNR), which was decreed in 1998 by the Ministry of Environment. The TCNR is within the best-preserved stretches of sandy coastlines in Lebanon. The terrestrial section is 3.5 km long and covers over 3.8 km² and is divided into three zones: Touristic (sandy beach), Conservation (core area), and Agricultural (Ras El-Ain), while the Marine zone’s area is 113 km² ( Yaacoub et al., 2011 ).

Based on a synthesis of regional hydrogeological studies ( Shaban, 2020a ; UNDP-CEDRO, 2014 ; Darwish et al., 2008b ; Dubertret, 1975 ), a clear understanding of the flow patterns and recharge mechanisms in the Ras El-Ain area of South Lebanon is critical for assessing the dynamics of groundwater contamination. This coastal region features a complex hydrogeological system dominated by karstified limestone formations, which exhibit high secondary porosity due to extensive fracturing and dissolution. Groundwater in Ras El-Ain predominantly flows along NE-SW trending strike-slip faults and inclined bedding planes that dip westward, enabling lateral subsurface movement from inland recharge zones—such as the Nabatieh Plateau—toward the coastal aquifers. The major fault systems, extending for over 10 kilometers, act as primary conduits for deep groundwater flow, transporting large volumes of water with limited natural filtration. Additionally, direct recharge occurs through fractured carbonate rocks, especially in areas where the overlying soil and vegetation are sparse, allowing precipitation to infiltrate rapidly. However, interbedded marly horizons can locally impede vertical infiltration, resulting in perched water zones. The interplay of karstification, fault-guided flow, and direct recharge increases the aquifer system’s vulnerability to contamination. Pollutants from surface sources—such as agricultural runoff, wastewater, or chemical spills—can easily infiltrate through sinkholes, fractures, or unprotected wellheads, quickly reaching the groundwater table and discharging via the Ras El-Ain springs or through submarine groundwater discharge zones along the coastline. Therefore, a comprehensive understanding of the region’s hydrogeological setting is essential for effective groundwater protection and sustainable management in this environmentally sensitive area.

The methodology is based on analyzing the impact of land use on water quality. Primary data were collected through direct sampling of water from wells after surveying the farmers. Data collected from farmers included the altitude of the land, depth of water in wells, amount of water pumping, type of cultivation and the fertilizers and pesticides used. The collected data are summarized in the following table ( Table 1 ).

Concerning the sampling of water, the location coordinates of each site have been taken with a GPS. We have 24 wells located in villages in Tyre region. Locations of groundwater sample collection points are shown in the map below ( Figure 2 ). The samples were taken from the direct source of water from 24 wells and the 2 springs of Ras El-Ain. The water samples (26 samples) were stored in cool boxes and transported to the laboratory (Doctoral School of Science and Technology at the Lebanese University) for subsequent analyses. At each sampling site, approximately 1000 ml of groundwater was sampled. The following parameters were analyzed in the laboratory: pH, EC, TDS, cations (Mg²⁺, Ca²⁺, Na⁺, K⁺), anions ( ${\text{NO}}_2^{-}$, ${\text{NO}}_3^{-}$, F⁻, Br⁻, ${\text{PO}}_4^{3-}$, ${\text{SO}}_4^{2-}$, Cl⁻) and heavy metals (Pb, Mn, Fe, Cu, Zn, Cr). Before conducting any analysis at the laboratory, all samples were filtered through a micro sterilized filter and poured into sterilized flacons.

After the analysis of water sampling, the data were re-arranged for necessary statistical analysis using R statistical software. The methods used in the proposed research are described briefly in the following section. Empirical rule was used to assess the quality of groundwater sample data. The empirical rule states that approximately 95% of the measurements are within the interval of mean ± two standard deviation. In the present study, all the groundwater data were found within this limit. A total of six standard water quality indices, namely sodium absorption ratio (SAR), soluble sodium percentage (SSP), total hardness (TH), magnesium adsorption ratio (MAR), Kelly’s ratio (KR) and total dissolved solids (TDS) were calculated from groundwater quality data. The following equations were used to calculate various groundwater quality indices:

1) SAR = [Na]/(([Ca] + [Mg])/2)^1/2

2) SSP = ([Na] + [K]) × 100)/[Ca] + [Mg] + [K]

3) TH = ([Ca] + [Mg]) × 50

4) MAR = ([Mg] × 100)/([Ca] + [Mg])

5) KR = [Na]/([Ca] + [Mg])

6) TDS = 0.64 × [EC × 10⁶ (mmohs/cm)]

The aim of these analyses was also to identify the origin of groundwater quality hazards. Furthermore, non-parametric Mann–Whitney test was conducted among the sets of groundwater quality data collected from different land use zones to decipher if there were any differences in groundwater quality due to land use.

To assess the quality of water, it is critical to study its physicochemical parameters in order to be able to compare the results with standard values. The physicochemical parameters and heavy metals were compared with the WHO and FAO guidelines for drinking and irrigation water respectively. The relationship between the physicochemical parameters and heavy metals in the water samples were established using Pearson product correlation coefficient (r) at 1% level of significance. The mean values of different parameters and heavy metals’ concentration is described below.

The EC values of selected sites are calculated from TDS values. The values were between (246 - 940) ms/cm, which are within the limit of WHO standards (1500 ms/cm) and FAO standards (3000 ms/cm). EC is a good measure of salinity hazard to crops. There is a very strong relationship between conductivity and mineralization (dissolved solids and minerals in water). The values of electrical conductivity of the tested samples are within the limits; therefore, all sites are of low-medium EC level, which is within the good salinity range.

The calcium in water samples in all sites ranges between 25.98 and 99.45 mg/L which is within the WHO and FAO limits. Hence, calcium concentration in water is within the desirable ranges in these sites.

The magnesium concentrations in the studied sites range between 6.83 and 26.19 mg/L, it does not exceed the standards. Thus, no magnesium contamination was detected on these sites.

The sodium concentrations at the sites do not exceed the limits of WHO and FAO, since their values range between 1.5 and 10.6 mg/L. No contamination by this element, and the water is suitable for irrigation and for drinking.

The measured values of potassium ranged between 0.06 and 0.43 mg/L, lower than those of the WHO (12 mg/L) and FAO (20 mg/L) drinking water and irrigation water standards. Therefore, potassium content in these sites is within its normal and acceptable limits.

The concentrations of nitrite exceeded the drinking and irrigation water limits specified by the WHO and FAO (0.05 mg/L) in the following sites: (1, 2, 3, 4, 6, 8, 10, 11, 12, 13, 14, 17, 19, 20, 21, 22, 24, and 25; Figure 3 ). Only two wells (20 and 23) had low amounts of nitrite (0.005 mg/L), while well 12 had the highest concentration (0.453 mg/L). High nitrite concentration in most of the studied sites proves the high use of nitrogen fertilizers, referring to agricultural land use.

The concentrations of nitrate in the studied wells show that only site 1 exceeds the limits of both WHO (40 mg/L) and FAO (70 mg/L) with a concentration of 75 mg/L, while sites 20 (2.33 mg/L) and 21 (1.01 mg/L) show the lowest nitrate concentrations. Site 23 (44.63 mg/L) exceeds the WHO limit for drinking water. The remaining sites comply with both guidelines, including Ras El-Ain sites ( Figure 4 ). High nitrate concentrations in the majority of the sites also confirm their location in agricultural fields, and in turn their source from the excess use of nitrogen fertilizers associated with the agricultural land use.

Concerning other anions, no contamination by orthophosphate, sulfate, chloride or fluoride were detected in these sites.

The results showed that the heavy metals studied had very low or undetectable concentrations. These results are mainly attributed to the location of the study area far from the main influencers of such pollutants in groundwater such as dumpsites, industrial effluents, sewage water, etc.

The water quality indices were calculated at all 26 sampling locations. Obtained results are summarized in Table 2 .

It was found that the average TH in the groundwater is very high. Some of other indices like TDS were also comparatively high in Ras El-Ain. The box plots of different groundwater quality indices are shown in Figure 5 .

The box plots in the figure show a number of outliers in the positive direction for indices like TDS, SAR, MAR and KR. This indicates extreme values of these indices in specific locations. As the higher value of an index represents corresponding quality hazard, extreme values of these indices mean the presence of corresponding groundwater quality hazards in some locations of Ras El-Ain.

The objective of this analysis is to make the diagnosis of the typology of the water of 26 sites through the physicochemical parameters by the principal component analysis and identify the most polluted site/s.

Principal Component Analysis (PCA) is a multivariate analysis method that aims to reduce the data from the higher dimensions to lower dimensional space without losing much information. The approach of PCA is to reduce the unnecessary features, which are presented in the data and is creating or deriving new dimensions (or also referred to as components). These components are a linear combination of the original variables. This way, PCA converts a larger number of correlated variables (i.e., breaks down the data) into a smaller set of uncorrelated variables. The principal component of a data set is the direction with the largest variance.

Based on Figure 6 & Figure 7 , Axis 1 explains 29.3% of the total variation, it shows a positive correlation with ${\text{SO}}_4^{2-}$ (r = 0.85), Na⁺ (r = 0.85), Cl⁻ (r = 0.83), K⁺ (r = 0.78), EC (r = 0.75) and Br⁻ (r = 0.67). Axis 2 explains 21.8% of the total variation, it is positively correlated with F⁻ (r = 0.86), Zn (r = 0.86), Mn (r = 0.73), and negatively correlated with ${\text{NO}}_3^{-}$ (r = −0.67). Moreover, Axis 3 explains 10.2% of the total variation, it is negatively correlated with ${\text{NO}}_2^{-}$ (r = −0.79) and positively correlated with ${\text{PO}}_4^{3-}$ (r = +0.70).

Concerning the correlation between variables, the correlation circle visualizes graphically what we have already concluded: ${\text{SO}}_4^{2-}$, Na⁺, Cl⁻, K⁺, and Br⁻ are closely correlated and represent the axis 1 of this positive side with EC and Cl⁻, which are also correlated. Component 2 is mainly represented by F⁻ and Zn on its positive side which are strongly correlated. However, it is represented with ${\text{NO}}_3^{-}$ on its negative side. The component 3 is mainly represented by ${\text{PO}}_4^{3-}$and Zn on its positive side and with ${\text{NO}}_2^{-}$ on its negative side.

Additionally, the correlation matrix revealed several moderate negative correlations that provide important insights into groundwater quality dynamics in the study area. A significant inverse relationship was observed between TDS/EC and pH (r = −0.50), as well as Ca²⁺ and pH (r = −0.53), suggesting that increased mineralization and calcium concentrations are generally associated with more acidic conditions. This may reflect the dissolution of carbonate minerals or the influence of acidic contaminants such as nitrates. Similarly, nitrate ( ${\text{NO}}_3^{-}$) showed a negative correlation with pH (r = −0.49), Mn (r = −0.40), and Zn (r = −0.45), indicating that higher nitrate concentrations could be linked to slightly lower pH and increased mobility of trace metals, possibly due to redox-sensitive processes or anthropogenic inputs like fertilizers. Additionally, Ca²⁺ and TDS/EC both showed a moderate negative correlation with well depth (r = −0.56 and −0.44, respectively), implying that shallow aquifers are more affected by surface-derived inputs and are more mineralized, possibly due to intensive agricultural activity and direct recharge in karstic zones. These findings highlight the hydrogeochemical sensitivity of the Ras El-Ain aquifer system, especially under agricultural land use pressures ( Figure 7 , Figure 8 ).

According to Figure 8 , we notice that site 1 has a very high contribution (24.40); it influences the most axis 1, besides site 22 and 23 (18.09, and 13.23 respectively) and site 2 (10.57). Sites 20 and 21 influence the most axis 2 since they have the greater contribution (39.62 and 28.21 respectively) beside site 1 which has a limited impact on this component (8.20). Moreover, the site 12 influences the most axis 3 (21.05), followed by site 26 (20.87) and site 22 (16.05).

The aim of this research was to study and analyze the impact of agricultural land use, specifically the application of fertilizers and agro-chemicals on the groundwater quality in the sub-catchment area of Ras El-Ain springs and the surrounding villages in Tyre region. Ras El-Ain springs make up part of Tyre Coast Nature Reserve (TCNR), the reason behind the choice to study a substantial site belonging to a protected area. Indeed, and based on this research, Ras El-Ain is the main water source in the area, providing the whole city of Tyre and its suburbs with domestic water. In addition, this source is to be an irrigation source for the entire Ras El-Ain agricultural lands, and the remaining part flows out into the sea, making an estuary and enriching the marine ecosystem with biodiversity. The rest of the sites are wells upstream Ras El-Ain, in the lower sub-catchment basin of the springs.

Based on the above, and concerning the wells water, wells for groundwater abstraction are tremendous in the coastal plain and in the foot slopes of the adjacent mountains. They are used mainly for agricultural purposes, and are almost shallow in depth, and this is the reason for their various distribution which is estimated at more than 1000 wells/km². The shallow wells are mainly dug in the Quaternary rock formation with less than 50 m depth from terrain surface. As for the higher altitudes in the mountainous region, groundwater is stored in the rocky lithology at exceeded depths. The current field survey has been carried out to investigate water quality in the coastal plain and the surroundings. Therefore, samples were selected from 24 wells and two from Ras El-Ain Springs ( Table 3 ).

Based on the detailed geological survey data of the wells and springs examined, the distribution of wells across different aquiferous rock formations is as follows:

The majority of surveyed wells are located within the Upper and Middle Eocene aquiferous rock formations, primarily composed of marly limestone. The depth to the water surface in the wells of the Upper Eocene Aquifer ranges between 80 meters and 200 meters, whereas it is approximately 250 meters for the Middle Eocene aquiferous rock formation. The primary sources of water feeding these wells can be categorized as follows:

The evaluation of the groundwater quality was based on the evaluation of water physio-chemical parameters (pH, EC and TDS) of the samples, their content of anions ( ${\text{NO}}_2^{-}$, ${\text{NO}}_3^{-}$, Cl⁻, ${\text{SO}}_4^{2-}$, ${\text{PO}}_4^{3-}$, Br⁻ and F⁻), cations (Na⁺, K⁺, Ca²⁺, Mg²⁺), and heavy metals (Pb, Cu, Cr, Fe, Mn, Zn).

The physical analysis of water shows that acceptable ranges of pH, electrical conductivity (EC) and (TDS), which are conform with the guidelines of WHO and FAO in all sites. Based on EC and TDS values, we concluded that water is not saline, and the water is acceptable at all sites for irrigation. pH values are also within the acceptable ranges of the WHO and FAO (slightly alkaline), except for sites (16), (18), (19) and (20) that exceeded those limits.

The chemical analysis of water shows that agro-chemicals’ contamination was detectable in 17 sites with Nitrite ( ${\text{NO}}_2^{-}$) including Ras El-Ain springs, in addition to two sites contaminated with Nitrate ( ${\text{NO}}_3^{-}$). 15 sites indicated moderate contamination with bromide, and 2 sites only showed moderate contamination with fluoride.

No contamination with any other agro-chemical anions and cations, nor with any of the heavy metals was detected at any of the analyzed sites.

The main cause of pollution is clearly attributed to the intensive use of fertilizers as an important input in agricultural practices, especially nitrogen fertilizers that show in the form of nitrite and nitrate water content exceeding the acceptable FAO &/or WHO limits.

If we coincide these results of active individuals with the results of the active variables, we observe:

Although detailed records of site-specific fertilizer and pesticide application rates were not available during this study, existing literature and national assessments provide insights into typical usage patterns under similar agro-climatic and cropping conditions in Lebanon. According to the FAO and the Ministry of Agriculture of Lebanon, nitrogen-based fertilizers (urea and ammonium nitrate) are commonly used at rates ranging from 120 - 200 kg/ha/year in intensive vegetable and cereal production systems, particularly in the southern Bekaa and South Lebanon regions ( FAO, 2016 ; MoA & FAO, 2014 ). Similarly, pesticide application—including insecticides, herbicides, and fungicides—can reach up to 20 - 30 kg of active ingredient per hectare per year, especially in open-field vegetables and citrus crops, which are prevalent in the Ras El-Ain sub-catchment.

The lack of precise regulation and monitoring of chemical inputs in many smallholder systems contributes to over-application and misuse, often exceeding recommended dosages ( Saadeh et al., 2012 ). This significantly increases the risk of nitrate leaching, heavy metal accumulation, and pesticide residues in shallow groundwater, especially in karstic and fractured aquifer systems such as those in Ras El-Ain, where percolation can be rapid and direct.

Therefore, even in the absence of measured input rates, the known intensity of agricultural practices in the region—combined with the hydrogeological vulnerability—strongly supports the observed correlation between land use and groundwater contamination.