Hydrogeochemistry and Groundwater Quality Assessment in the High Agri Valley (Southern Italy)

The High Agri Valley (southern Italy) is one of the largest intermontane basin of the southern Apennines a ﬀ ected by intensive agricultural and industrial activities. The study of groundwater chemical features provides much important information useful in water resource management. In this study, hydrogeochemical investigations coupled with multivariate statistics, saturation indices, and stable isotope composition ( δ D and δ 18 O) were conducted in the High Agri Valley to determine the chemical composition of groundwater and to de ﬁ ne the geogenic and anthropogenic in ﬂ uences on groundwater quality. Twenty-four sampling point ( including well and spring waters) have been examined. The isotopic data revealed that groundwater has a meteoric origin. Well waters, located on recent alluvial-lacustrine deposits in shallow porous aquifers at the valley ﬂ oor, are in ﬂ uenced by seasonal rainfall events and show shallow circuits; conversely, spring waters from ﬁ ssured and/or karsti ﬁ ed aquifers are probably associated to deeper and longer hydrogeological circuits. The R -mode factor analysis shows that three factors explain 94% of the total variance, and F1 represents the combined e ﬀ ect of dolomite and silicate dissolution to explain most water chemistry. In addition, very low contents of trace elements were detected, and their distribution was principally related to natural input. Only two well waters, used for irrigation use, show critical issue for NO 3- concentrations, whose values are linked to agricultural activities. Groundwater quality strongly a ﬀ ects the management of water resources, as well as their suitability for domestic, agricultural, and industrial uses. Overall, our results were considered ful ﬁ lling the requirements for the inorganic component of the Water Framework Directive and Italian legislation for drinking purposes. The water quality for irrigation is from “ good to permissible ” to “ excellent to good ” although salinity and relatively high content of Mg 2+ can occasionally be critical.


Introduction
Groundwater resource is one of the most challenging current and future issues of worldwide concern. The ever-increasing rate of population growth and the inherent water supply demand have led to intensive water exploitation. Groundwater is of great importance for domestic, drinking, irrigation, and industry purposes especially where the water resource is availability scarce. Groundwater quality depends on natural processes such as rock-water interaction, climatic conditions, geological context, and anthropogenic activities. Natural pro-cesses including the mineral precipitation or dissolution, ionexchange, redox condition, residence time, and mixing between different water type may have a great impact on groundwater quality [1]. Anthropogenic activities such as rapid urbanization, industrialization, and intensive agricultural activities have caused a deterioration in water quality worldwide [2][3][4][5]. Groundwater contamination can be persisted a long time due to the low flow rate of groundwater in an aquifer and may involve major ions and trace elements [6]. High levels of contaminants, exceeding guideline values, can turn water to be unsuitable for drinking, irrigation, fishing, and recreation [7], causing serious adverse effects on human and biota health [8]. The release of elements from rocks depends upon physical (temperature, residence time, flow rate) and chemical factors (namely, weathering of rockforming minerals, precipitation of sparingly soluble secondary minerals, and presence of reactive gas species such as CO 2 ), i.e., [9,10]. The redox process can have an important role in determining the partitioning between aqueous solution and solid phases, taking into account that some elements are present in nature in different oxidation states. In southern Apennines (southern Italy), the High Agri Valley (hereafter HAV) is characterized by a relevant presence of intensive agricultural activities in predominantly hilly environment due to the occurrence of groundwater and fertile soils and a unique forest heritage with a rich biodiversity that have stimulated tourism growth. In recent decades, together with an ancient rural economy based on agriculture, wood, and dairy production, new industrial activities have been undertaken. Hydrocarbon extraction, related to the largest onshore oil field in Western Europe, is the most important industrial activity and has actually an oil production capacity of about 80.000 barrels/day [11]. In the HAV, the groundwater represents an important freshwater resource, used as drinking water supply for agricultural and industrial purposes. Usually, the anthropogenic activities influencing groundwater systems contribute to a decrease in water availability, both directly through variations in aquifer recharge and indirectly through changes in groundwater quality and use. The evaluation and management of groundwater resources require an understanding of hydrogeochemical features of the aquifers.
The aim of this research, representing the first step of a comprehensive hydrogeochemical characterization of the area, is to understand the geochemical evolution of groundwater in the HAV in order to guarantee reliable supply for all purposes and define a sustainable groundwater management strategy, also considering the growing industrialization that could cause pollution phenomena. The main objectives are to (1) determine the geochemical processes controlling the chemical composition, (2) define the geogenic and anthropogenic influences on groundwater quality, and (3) compare the concentrations of some inorganic elements to values established by the World Health Organization and the Italian legislation for drinking and irrigation purposes.

Climate, Geology, and Hydrogeology
HAV area (Basilicata region, southern Italy) extends for about 140 km 2 and is 30 km long and 12 km wide. According to the Köppen classification, the climate of the area is of warm-summer Mediterranean type, characterized by coldhumid winter and hot-dry summer. In the study area, the mean annual precipitation value (calculated in 2005-2015 range) was of about 1000 mm/y, while the mean annual temperature was of 12.5°C. The coldest month was January with average temperature ranging between 3°C and 4°C. Hot and arid conditions during summer produce periods of drought during July and August. The climate of the area, characterized by high seasonal precipitation rates in autumn and spring and snowy winter, allows the greatest amount of the aquifer recharge. From a geological point of view, HAV is a NW-SE trending Quaternary basin located in the axial zone of the southern Apennines, an east-verging fold-and-thrust belt developed as an accretionary wedge due to the eastward migration of the compressional tectonics in the Apennine Arc (Late Oligocene-Early Pleistocene) [12,13]. The HAV evolution, since the Middle Pleistocene, is controlled by a still active NE-SW extensional tectonic regime [14], as witnessed by NW striking high-angle normal and oblique faults bordering the basin, which represent the main seismogenic structures in the area [15,16]. The HAV is characterized by Quaternary continental deposits overlying a Pre-Quaternary substratum (Figure 1(a)). In the south-west side of the basin, the substratum consists of Mesozoic to Cenozoic shallow-water and slope carbonates of the Campania-Lucania Platform over thrusted on coeval pelagic successions (Lagonegro Units), while to south-east the substratum is characterized by Tertiary siliciclastic sediments of the Gorgoglione Flysch and Albidona Formation [13,17]. A Quaternary synorogenic succession, consisting of Lower Pleistocene-Holocene continental clastic sediments, mainly coarse-grained, fills the HAV basin [18] and references therein. Finally, the most recent Pleistocene and Holocene deposits are represented by terraced alluvium, alluvial fans, and recent to present-day alluvial sediments (Figure 1(a)).
The HAV hydrogeology is characterized by a high groundwater content favoured by climate conditions. Two different kinds of aquifers are recognized: fissured and/or karstified aquifers in the Apennine units of the substratum and porous multilayer aquifers developed in the Quaternary succession [19]. Both types host high volumes of groundwater storage, with the richest amount located in Pre-Quaternary substratum aquifers usually subdivided in different hydrogeological complexes according to the structural, lithological, and permeability features [20]. These aquifers mainly occur in highly fractured limestones (Carbonate Platform) and in the underlying "Calcari con Selce" Fm (Lagonegro Units), both characterized by high permeability (Figure 1(b)). Porous aquifers occur in gravelly-sandy permeable deposits of the Quaternary succession [21], where the larger groundwater body is recognized in a multilayered and semiconfined aquifer. Carbonate karst and fractured aquifers play a leading role in the overall hydrogeological system, since their groundwater resources supply the detrital-alluvial ones occurring at the valley bottom. Finally, most springs from HAV, generally showing an average flow rate greater than 5 l/s, are at the contact between permeable limestone and dolostone (Lagonegro Units and Carbonate Platform) and impermeable Plio-Pleistocene siliciclastic and recent alluvial-lacustrine deposits [22].

Materials and Methods
3.1. Sampling and Analyses. Twenty-four water samples were collected during a single survey carried out between March and April 2016 in the HAV. Nine of them were taken from springs of the fissured and/or karstified aquifers located on Pre-Quaternary carbonate rocks and sediments belonging to the Lagonegro Units and fifteen from wells on recent alluvial-lacustrine deposits in shallow porous aquifers at the valley floor with a phreatic level ranging from 1 to 6 meters 2 Geofluids from the surface ( Figure 1). The investigated springs are located on the right bank of the Agri River, except for S3 site, in the left river side. Most of springs, managed by the Acquedotto Lucano public agency, are used for drinking purposes, whereas well waters, mostly belonging to private individuals, are used for irrigation purposes. Temperature, pH, electric conductivity (EC, measured at 25°C), and redox potential (Eh) were measured in situ using a high-resolution multiparametric probes (Hanna Instruments HI-9828), after calibration in the laboratory by means PTB/NIST traceable standard solutions. Total alkalinity was determined in situ by titrating unfiltered samples with 0.01 N HCl, and methyl-orange was indicator. All water samples were filtered in situ at 0.45 μm membrane (MF-Millipore) and then stored in low-density   [15]) where localization of the sampling sites (springs and wells) is displayed. Schematic cross section (b) of the northern part of study area is drawn (modified after Carbone et al. [17]).    EC: electrical conductivity measured at 25°C; T: temperature; Eh: redox potential; TDS: total dissolved solid; d.l.: detection limit; b.d.l.: below detection limit; IB: ionic balance; SAR: sodium adsorption ratio by Richards [26]; MAR: magnesium adsorption ratio by Raghunath [27]; %Na: sodium percentage by Todd and Mays [28]; RSBC: residual sodium bicarbonate by Gupta and Gupta [29]; PI: permeability index by Ragunath [27]. ) and cations (Na + , K + , Mg 2+ , and Ca 2+ ) were determined by ionic chromatography. Minor and trace elements (Li, B, Rb, Sr, Ba, V, and Cu) were determined by inductively coupled plasma-optical emission spectroscopy (ICP-OES). All trace element determinations were performed with the external standard calibration method, using NIST and SLRS standard reference materials for calibration. The precision of the analytical results was estimated by running triplicate analyses every ten samples. Uncertainty of measurements was ≤5-7% for all trace elements. Dissolved SiO 2 was measured by VIS spectrophotometry upon reaction with ammonium molybdate in acid media (and treatment with oxalic acid) to form a yellow silicomolybdate complex, whose absorbance was read at 410 nm [23]. Major element analyses were performed at the Analytical Chemical laboratory of the University of Basilicata; trace elements were determined using a Thermo Scientific™ iCAP™ 7200 at Gaudianello Company laboratory. The oxygen and hydrogen isotopic compositions were analysed at National Institute of Geophysical and Volcanology (INGV, Palermo, Italy) on unfiltered water samples using Analytical Precision AP 2003 and Finnigan MAT Delta Plus IRMS devices, respectively. The isotopic ratios are expressed as the deviation per mil (δ‰) from the reference V-SMOW. The uncertainties 1σ were ±0.1% for δ 18 O and ±1% for δD.

Geochemical
Modeling. The Geochemist's Workbench software (GWB 8.0, [24]), implemented with the Thermoddem database [25], was used to calculate saturation indices (SI) for the main mineral phases presents in the aquifer and the construction of activity diagrams. Saturation indices (SI) are defined as SI = log ðIAP/KtÞ, where IAP is the ion activity product of the mineral-water reaction and Kt is the thermodynamic equilibrium constant at the measured temperature. Thus, SI = 0 indicates a thermodynamic equilibrium state, and values > 0 denote oversaturation and <0 undersaturation.

Irrigation Quality Parameters.
In order to determine the suitability of the investigated groundwater for irrigation purposes, the following parameters were evaluated: (i) The sodium adsorption ratio (SAR), by Richards [26]: (ii) The magnesium adsorption ratio (MAR), by Raghunath [27]: (iii) The sodium percentage (%Na), by Todd and Mays [28]: (iv) The residual sodium bicarbonate (RSBC), by Gupta and Gupta [29]: (v) The permeability index (PI), by Raghunath [27]: where concentrations are expressed in meq/l.

Results
Temperature, pH, EC, Eh, and the total dissolved solids (hereafter TDS) and major elements are listed in Table 1 together with the geographical coordinates of the sampling points. The EC, Eh, and TDS box-and-whisker plots are shown in Figure

Geofluids
Box-and-whisker plots of the major constituents (cations and anions) are given in Figure 3.
V and Cu are characterized by very low contents and in many cases are below the detection limit (b.d.l.: 0.2 and 0.4 μg/l, respectively). Consequently, these elements will not be further included and treated in the discussion. Li and Rb are detected in most samples with values between b.d.l. and Based on the Piper diagram, the investigated waters show a homogenous distribution with bicarbonate alkaline-earth composition ( Figure 5).
Another useful index for water classification is Ionic Salinity or Total Ionic Salinity (TIS) that shows the sum of anion and cation total, expressed in meq/l [30]. Iso-TIS lines are reported in Figure 6 (SO 4 2vs. HCO 3 -+ Cl -) where HAV waters falling in the 8 to 20 meq/l range because of their similar features.
The investigated waters showed δ 18 O values between -7.2 and -8.9 ‰ and those of δD from -44 to -53 ‰. The waters from the karst and fissured aquifers generally showed more negative isotopic values (down to −8.9 and−53‰) than waters hosted by shallow porous aquifers, the latter displaying a wide range of variation, from −8.5 to −7.2‰ and from −44 to−52‰ for δ 18 O and δ 2 Η, respectively. 6 Geofluids selected trace elements (Sr, Ba, B). Factors were extracted after varimax rotation using the STATGRAPHICS 18 package. This operation was performed using a standardized correlation matrix, classical type of factoring, thereby equally weighting all the variables during factor calculations. The communalities provide an index of the efficiency of the proposed set of factors [31], and the magnitude of the communalities calculated in this study suggests that most of the original variance is still accounted for by the present set of factors. Three factors explain 94% of the total variance in the selected database (Table 3).

Discussion
Usually, in terrains where the carbonate lithological facies is dominated by the occurrence of dolomite, the bicarbonate alkaline-earth hydrofacies, with higher Mg 2+ values than Ca 2+ one, can be ascribed to dolomite dissolution. It is well known that the dissolution of dolomite can be expressed as follows: and the water-dolomite interaction should release equally charged amounts of Ca 2+ + Mg 2+ and HCO 3 − , with a ðCa 2+ + Mg 2+ /HCO 3 − Þ ratio close to 1. Accordingly, in our case, the correlation between (Ca 2+ +-Mg 2+ and HCO 3 − ) is highly significant (r = 0:86, p value < 0.001, Figure 7). Only P7 sample falls away from general trend, maybe because of its anthropogenic component (see discussion below). However, both the (Ca 2+ +Mg 2+ /HCO 3 − ) and Mg 2+ /Ca 2+ ratios, as meq/l, are above 1, claiming for additional source(s) of Mg 2+ in groundwater, as also suggested by the finding that several samples are oversaturated with respect to dolomite or close to the SI = 0 value (Table 2, Figure 8).

Geofluids
The correlation between dissolved silica and Mg 2+ is significant (r = 0:65, p < 0:01), clearly indicating silicate dissolution as additional source of Mg 2+ in solution. The incongruent dissolution of silicates reacting with CO 2 derived from microorganism respiration and/or mineralization of soil organic matter can promote a Mg 2+ surplus into the groundwater [32] and references therein. F1 thus represents the combined effect of dolomite and silicate dissolution on the release of Mg 2+ in solution.
The second factor (F2; var:% = 15:2) includes significant and positive weightings for the largely soluble ions Na + , Cl -, and SO 4 2-. Overall, the investigated waters have Na/Cl meq/ l ratio > 1, suggesting that Na + in groundwater may derive from weathering of mineral phases different from the common NaCl salt. The positive correlation between dissolved silica and Na + (r = 0:66, p < 0:01) associated with negative values of the feldspars SI (albite and anorthite, Table 2) suggests that the silicate incongruent dissolution is ongoing process allowing the release of Na + in solution.
Further, the log a (H 4 SiO 4 ) aq vs. log a (Na + /H + ) activity diagram (Figure 9) shows that the studied waters are in equilibrium with kaolinite, a T-O clayey mineral with a minor cation exchange capacity with respect to other clays, that does not include in its lattice interlayered cations such as Na + . This, in turn, means that kaolinite cannot contribute to narrow down the Na + content in HAV groundwater. As  for Cl -, its geogenic abundance in groundwater increases with groundwater age [33], and excluding salt dissolution and marine spray supply at inland [34] and references therein, Clmay mostly derive from anions exchange at the clay mineral edges. In fact, depending on the pH of solution and due to protonation reactions at the broken bonds of both octahedral and tetrahedral layers, the clay mineral edges may be positively charged, promoting anions adsorption, e.g., [35]. The likely lack of sulphate minerals (anhydrite and gypsum) in HAV lithologies excludes a SO 4 2origin through salt solubilization. Although oxidative reaction affecting suphides (e.g., pyrite FeS 2 ) dispersed in rocks, even in small amounts, may represent a source for SO 4 2in groundwater, the use of N-rich fertilizers may also originate most of the SO 4 2in solution from the recycling of the groundwater used in irrigation. As shown in Figure 10  , two different trends are clearly recognized. In the first one, involving exclusively well waters (except for S25 sample), the NO 3 -/SO 4 2ratio is close to 1, and NO 3 contents are generally >4 mg/l. P5 and P7 samples have the highest NO 3 and SO 4 2-values. Although there are no baseline data on the NO 3 geogenic level in the area, it is known that nitrate concentrations above 4 mg/l can be referred to anthropogenic contamination, e.g., [36,37]. In our case, this supports the idea of an anthropogenic origin for SO 4 2-, mostly due to agriculture fertilizers. The second trend is outlined by a few wells showing larger SO 4 2contents and low NO 3 contents and suggests a possible geogenic origin for SO 4 2due to oxidative reaction affecting suphides. In the third factor (F3; var:% = 11:2), only boron has a significant weight. Boron contents in groundwater may be affected by anthropogenic activities [38] and interaction with evaporite levels, e.g., [39] and references therein. In the aqueous environment, B is a very mobile element occurring as boric acid (B(OH) 3 ) in dilute aqueous solution at pH < 7 and as prevailing metaborate anion (B(OH) 4 − )at pH > 10, e.g., [40] and references therein. Boron may have several possible natural and anthropogenic sources in inland aquifers including, for instance, leaching of geologic materials as well as domestic wastewater, e.g., [41] and references therein. The low B concentration observed in the HAV groundwater (well below 100 μg/l) appears to exclude any significant anthropogenic source. F3 thus likely accounts for the solubilization of B by leaching of sedimentary deposits of marine and nonmarine origin, e.g., [42] and references therein.

Isotopic Constraints.
Water isotopes are used to define water origin, recharge areas, circulation paths, and mixing or exchange processes [43][44][45]. The δD and δ 18 O values are displayed in Figure 11, together with the Northern Calabria    [48]). Spring waters are well-correlated (r = 0:95, p value < 0.0001, Figure 11(a)) and fall between SIMWL and NCMWL, while well waters show a wider data distribution (r = 0:66, p value = 0.013, Figure 11(b)) falling between GMWL and NCMWL. Both water types are fed by rainwater and have a meteoric origin.
The isotope data variation, measured in well waters, is probably linked to seasonal rainfall effect since the shallow porous aquifer has a relatively short and surface hydrogeological circuits through the gravelly-sandy permeable deposits of the quaternary succession.
Spring waters from karst and fissured aquifers showing a narrow range of O and 2 H isotopic values can be used to deduce the mean elevation of the recharge area through the equation proposed for southern Italy by Vespasiano et al. [46]: and a range of 1030 to 1540 meter a.s.l mean elevation of the recharge areas has been estimated. The inferred range is consistent with the elevation of the springs located in the highly fissured limestones (Carbonate Platform) and in the underlying "Calcari con Selce" Fm (Lagonegro Units). In addition, S11, S12, and S13 springs with more negative isotope data have longer and deeper hydrogeological circuit likely due to the local geological complexity affecting the recharge area to the west and north-west [17].

Groundwater for Irrigation and Drinking Uses.
The groundwater quality is of fundamental relevance for irrigation and drinking purposes [3], also in the northern bank of the Mediterranean, affected by climate change evolving toward semiarid to arid conditions, e.g., [49] and references therein. The amount of dissolved ions affects the agricultural productivity influencing both the growth of plants and soil structure [50]. Some important irrigation quality parameters (such as SAR, MAR, %Na, RSBC, and PI) are largely used for determining the suitability of groundwater for a proper agricultural uses [8,51,52] and references therein. Irrigation water quality parameters of the analysed groundwater are shown in Table 1. The %Na vs. EC plot [53] provides a mostly adopted method for rating irrigation water. In Figure 12, the HAV groundwater falls in the "excellent to good" and to a lesser extent in the "good to permissible" fields. Similarly, the PI indicates that the HAV groundwater is suitable for irrigation, being ranked as class I (PI: 30-50%; total concentration: 13-39 meq/l). However, in the SAR vs. EC diagram [26], the HAV groundwater is ranked as C2S1 and C3S1 (medium to high salinity hazard and low sodium hazard, Figure 12).
Excess salinity reduces the osmotic activity of plants limiting the absorption of water and nutrients from the soil of exchange able sodium, e.g., [54]. This suggests that efforts, including leaching and proper drainage, are needed in order to control the salinity hazard, especially for those waters, representing a significant part of HAV groundwater dataset, having EC higher than 750 μs/cm. Accordingly, some concerns also arise from the RSBC values, mostly related to the relatively high HCO 3 content, that in most samples is >5 meq/l, implying that HAV  groundwater cannot be classified as fully satisfactory for irrigation purposes according to Gupta and Gupta [29]. RSBC exceeding 5 meq/l may cause an alkalizing effect, claiming for a bicarbonate neutralization when long-term irrigation purposes are required [55]. An additional alkalizing effect is also due to high Mg 2+ contents in groundwater. Usually, Ca 2+ and Mg 2+ are in equilibrium in most waters, and Raghu-nath [27] and Gupta and Gupta [29] suggested that MAR values exceeding 50% indicate magnesium hazard as soils become more alkaline [56], favoring the decrease of phosphorous availability [57]. In the HAV groundwater, MAR is always >70% indicating magnesium hazard and suggesting long-term magnesium monitoring in the area in order to plan alkalinity mitigation policies.   Figure 12: Plots of calculated values of %Na and SAR vs. EC of groundwater samples (after Wilcox [53] and [26], respectively). The symbols are as Figure 5. Geofluids In Italy, groundwater quality for drinking use is regulated by D. Lgs 31/2001 [58] (acceptance of the 98/83/EU directive [59]) which provides for the measurement of indicator parameters (such as odor, color, taste, pH, and hardness) and chemical and microbiological parameters. In this study, a preliminary assessment of groundwater quality for drinking use was proposed considering only some indicator and chemical parameters analysed. The drinking water quality was compared with the conditions set by the World Health Organization 2004 (hereafter WHO 2004 [60]) and the Italian legislation (D. Lgs. 31/2001). The relative maximum admissible concentrations (hereafter MAC) imposed for protecting of groundwater resources by the Italian legislation [58], and the desirable limit (DL) and maximum permissible limit (MPL) prescribed by WHO for drinking purposes are shown in Table 4.
Most water samples have TDS below DL of 500 mg/l as defined by the WHO, and overall HAV groundwater has TDS below MPL of 1500 mg/l as defined by the WHO specification for drinking water. In most samples, HCO 3 exceeds the MPL defined by the WHO (240 mg/l). Na + , Cl -, and SO 4 2-are below the respective MPL (Na + = 200 mg/l; Cl − = 600 mg/l; SO 4 2− = 400 mg/l). NO 3 contents are always below the MAC of 50 mg/l except for two water samples (P5 and P7) having very high NO 3 values (58 and 203 mg/l, respectively). Some environmentally relevant trace elements, including B, V, and Cu, are characterized by very low contents, below the respective MAC and MPL. Ba and B are below the guideline values pro-vided by the WHO for drinking water whereas Sr 2+ exceeds the MPL (0.5 mg/l) in two water samples (P5 and P7). Our data allows to state that P5 and P7 water samples, used only for irrigation use, show critical issue for Sr 2+ and NO 3 concentrations. As previously said, high NO 3 values are wellrelated to SO 4 2contents and are linked to human activities, including mainly agricultural activities.
As to strontium values, it is well known that Sr-bearing minerals (mainly carbonate and sulphate) are distributed in a number of rock types, and Sr 2+ release from minerals and rocks to water by weathering processes is very common [61]. Strontium, in fact, readily reacts with water and oxygen to form insoluble mineral phases or create complexes with carbonate and silicate depending on the water mineralization. Therefore, high Sr 2+ concentrations may indicate an anthropogenic input by agricultural activities such as the use of fertilizers, carbonate additives, manure (i.e., cattle and poultry) and dumping, and industrial wastes [62]. Excluding P5 and P7 samples, a good and positive correlation between Sr 2+ and HCO 3 is evident for the analysed dataset (r = 0:75, p < 0:01 ), suggesting that Sr 2+ contents are linked to carbonate phases. Conversely, as P5 and P7 waters show high NO 3 and SO 4 2-contents, in addition to the Sr 2+ ones, a contribution from agricultural practices cannot be excluded for these waters. Although a microbiological characterization is required for defining the overall groundwater quality, based on the analysed inorganic component, all HAV groundwaters are suitable for drinking use, except for two well waters.

Conclusions
Groundwater is an essential water resource for drinking and irrigation uses in the HAV. In this study, hydrogeochemistry coupled with multivariate statistics, saturation indices, and stable isotope composition was used to assess the geochemical processes controlling the groundwater chemistry. Consequently, groundwater quality and its suitability for drinking and agricultural use were evaluated. All the examined groundwater has a meteoric origin although some springs show long and deep flow than the other ones. The main geochemical process affecting water chemistry is the dolomite and silicate dissolution that is also controlled by the concentration and distribution of trace elements. The SO 4 2-/NO 3 ratios suggest that an anthropogenic contamination, mostly associated with the use of agriculture fertilizers, cannot be excluded for some water samples (P5 and P7 samples). The suitability of water for drinking purposes was evaluated by comparing different chemical parameters with those reported by the World Health Organization [29] and the Italian legislation guidelines. Our results demonstrate that most of HAV groundwater is chemically suitable for drinking use with respect to the analysed inorganic chemical elements. As to agricultural use, the % Na and PI indicate that groundwater is generally useful for irrigation, although SAR, MAR, and RSBC highlight a medium to high salinity hazard. In light of the recent increase in human activities (i.e., industrialization and intensive agricultural practices), this study represents a warning for the local authorities providing significant insights to delineate a successful policy for management of groundwater resources.  12 Geofluids

Data Availability
The data used to support the findings of this study are included within the article.

Disclosure
Part of the paper is related to Maria Assunta Musto Master degree thesis (University of Basilicata, Potenza-Italy).

Conflicts of Interest
There are no conflicts of interest to declare.