Introduction and study area

The assessment of geochemical element distribution and the natural background are of paramount importance in reconciling the exploitation of surface and groundwaters with the protection of the environment, including the well being of both mankind and the local fauna and flora (Darnley et al. 1995; Edmunds et al. 2002). Following this understanding the Italian Government has recognised the need to identify geochemical baselines that clarify hydrodynamics and recharge periods in large Italians sedimentary basins. This has resulted in a re-evaluation of existing geochemical data and the acquisition of new data in order to understand the factors controlling both the hydrochemical concentration levels and the distribution of potentially toxic constituents. In order to fulfil these requirements the Friuli Venezia Giulia Environmental Protection Unit (ARPA FVG) has created a network of water chemistry monitoring sites, and since 1995 more than 1,000 groundwater and surface samples have been analysed for physical and chemical characteristics. The Friuli Venezia Giulia Plain located in northeast Italy hosts well-developed alluvial aquifers composed of Eocene to Quaternary aged thick gravel and sand beds interbedded with clay and silty layers (Stefanini and Cucchi 1976, 1977; Mosetti 1983; Cucchi et al. 1998, 1999, 2000, 2006). The goal of this research is to use major ions and stable isotope data to evaluate water quality and to statistically summarise these data in order to define hydrochemical groundwater provinces. The characteristics of these provinces can then be utilised for water management by other research organisations with similar interests. In addition, this will provide local authorities with a framework, which can be used to understand the degree of pollution in groundwater resources, the recharge areas and may provide useful information for better preservation of these important water bodies.

The Friuli Venezia Giulia Plain is located in the eastern part of the Padain Plain, the largest agricultural and industrial area of northeastern Italy (Fig. 1). The studied area is bounded by the Julian and Carnic Alps to the north, the Adriatic Sea to the south and the Classical Karst of Trieste and surrounding areas form the eastern boundary (Fig. 1). The Friuli Venezia Giulia Plain groundwater systems are located in sediments that are underlain by an Eocene to Oligocene succession deposited above a 7 km thick Mesozoic carbonate platform (Cati et al. 1987; Carulli 2006). The Friuli Venezia Giulia Plain is divided into two units, the Upper Friulian and the Lower Friulian, which are separated by the resurgence belt (Fig. 1). The Upper Friulian Plain includes calcareous and dolomitic gravels with a well-developed phreatic (P) aquifer, with an aquifer transmissivity of approximately 10−2 to 10−4 m2/s in a southerly direction (Mosetti 1983; Carniel 1999). This aquifer has an average thickness of 60–80 m, reaching depths of 200 m close to the border between the plain and the Alps and surfacing in the resurgence belt. The resurgence belt (Fig. 2) is positioned in the zone of slope change (between mean slope of 5‰ for the Upper Friulian Plain and 3‰ in the Lower Friulian Plain) that is related to the last glacial maximum (LGM) deposition of gravels from glacial rivers (Fontana, 2006). The Lower Friulian Plain is characterised by alternating gravels and sand interbedded by clay and silty layers (Fig. 2) that become thicker in a southerly direction. This area has abundance in surface waters compared to the Upper Friulian Plain (Fig. 1). In the Lower Friulian Plain the groundwater is confined to a multilayer aquifer (Fig. 2). This aquifer system is hydrogeologically divided at approximately 100–120 m depth, into shallow (SC) and deep confined (DC) groundwaters by a 10−15 m thick impermeable layer of silty material dating back to the Late Pleistocene (Fontana 2006). This low permeability layer has an approximate hydraulic conductivity of 10−7 m/s (Mosetti 1983) and is almost a continuous layer from Friuli Venezia Giulia to Lombardy (Pilla 1998; Pilli 2005). Thermal aquifers with water temperatures reaching 60° have been found at depths of 550–600 m (Fig. 2) during drilling explorations and are presently partly used by local industries (Stefanini and Cucchi 1976, 1977; Mosetti 1983; Cucchi et al. 1999). These deep geothermal aquifers have not been sampled in this study.

Fig. 1
figure 1

Location of the study area (dotted line Resurgence belt, P Pluviometer, crosses phreatic well samples, triangles shallow confined well samples and circles deep confined well samples; 1 Cellina River, 2 Livenza River, 3 Tagliamento River, 4 Torre River, 5 Natisone River and 6 Isonzo River)

Full size image
Fig. 2
figure 2

North–south and east–west simplified cross sections of the study area

Full size image

The main surface drainage is the Tagliamento River, which extends across the entire length of the Friuli Venezia Giulia Plain. The headwaters are situated in the Carnic Alps and the river flows first in an eastward direction before crossing the plain to outflow in the south (Fig. 1). Numerous artificial channels carry waters from the Tagliamento River. The average discharge of the Tagliamento River, 40 km inland from the Adriatic Sea, is 90–100 m3/s (Fontana and Bondesan 2006), with a maximum discharge of 3,000–4,500 m3/s (Foramitti 1990). Other perennial rivers in the region (Fig. 1) are the Isonzo (average discharge of 67 m3/s), the Torre/Natisone (average discharge of 17 and 7 m3/s), and Cellina/Meduna rivers (average discharge of 16 and 11 m3/s (Mosetti 1983).

Materials and methods

One hundred and forty-nine wells were selected from the physical-chemical characteristics database of ARPA FVG, all of which have chemical analyses available dating from 1995 to date. Sample location, type, electric conductivity (EC), pH, total dissolved solids (TDS), temperature (T), alkalinity (Alk), total hardness (TH), sodium adsorption ratio (SAR), major ions and stable isotope (δ18O) are presented in Table 1. The water temperature, electrical conductivity (EC) and pH of each sample were measured in situ by ARPA FVG with a conductivity meter standardised to 20°C and a pH electrode previously calibrated with standard buffers. The water samples were chemically analysed by ARPA FVG by atomic absorption spectrometry (Ca2+, Na+, K+, Mg2+, accuracy 5–10%) and ion chromatography (Cl−, SO4 2−, NO3 −, HCO3 −, accuracy 5−10%). All reported values (Table 1) have an ionic balance within 10%. TDS, Alk, TH and SAR were calculated using Aquachem version 4.0 whereas the saturation indices (Table 1) were calculated with PHREEQC (Parkhurst et al. 1980).

Table 1 Chemical and isotopic composition of groundwater samples from the Friuli Venezia Giulia Plain
Full size table

A subgroup of 128 of the 149 wells was selected in 2005–2006 for stable isotope (δ18O) analyses of water, as they cover the Friuli Venezia Giulia Plain homogeneously both in terms of aquifer extensions and depths. Isotopic analyses were carried out at the University of Trieste on a VG Optima mass spectrometer. The oxygen isotopic composition (δ18O) was measured by means of a water–CO2 equilibration technique at 25°C (Epstein and Mayeda 1953). The values are reported as per mil deviations from the VSMOW standard using the conventional δ notation (Craig 1961). The standard deviation of the measurements is about ±0.02‰ (2σ).

Results

Geochemistry

Table 2 presents a univariate overview of the hydrochemical data of the study area divided as phreatic (P), shallow confined (SC) and deep confined (DC) groundwater samples, whereas hydrochemical and isotopic analyses as a function of the depth are reported in Fig. 3a and Whisker plots of major ion concentrations are reported in Fig. 3b. The dominant ion from the studied area is HCO3 and five hydrochemical facies have been found: Facies 1: Ca–HCO3; Facies 2: Ca–Mg–HCO3; Facies 3: Ca–Mg–HCO3–SO4; Facies 4: Ca–Mg–SO4–HCO3 and Facies 5: Na–Ca–Mg–HCO3 (Table 1). Facies 1 to 4 are distributed amongst the phreatic and shallow confined groundwaters, whereas the Na–Ca–Mg–HCO3-type waters (Facies 5) have been recovered only down-gradient in the deep confined aquifers (Table 1) reflecting exchange reactions by water/clay-silt layer interactions (Hendry and Wassenaar 2000). Groundwater generally shows a decrease in conductivity and mineralisation with depth (Tables 1, 2). The mean temperatures for the studied groundwater samples are similar with a minimum value of 13.5°C for the phreatic to 14.7°C for the deep confined groundwater samples (Table 2; Fig. 3a). The EC data show a decrease in mean values from 509 μS/cm in the samples collected from the phreatic wells to 432 μS/cm and to 393 μS/cm in the shallow confined and deep confined groundwater samples. A similar trend of decreasing values is found for the TDS data in which there is a decrease from 466 to 355 mg/l from the phreatic to the deep confined groundwater samples (Table 2; Fig. 3a) The pH values range from 6.9 to 8.0 (mean of 7.5) to 7.3 to 8.1 (mean of 7.7) to 7.5 to 8.0 (mean of 7.8) in the phreatic, shallow and deep confined groundwater samples, respectively (Table 2; Fig. 3a). Mean concentrations of HCO3, Mg are generally similar with no observed differences between phreatic and confined groundwaters (Table 2; Fig. 3a). Potassium (from 0.8 to 1 mg/l) and Na (from 3.5 to 8.7 mg/l) concentrations increase with increasing depths of the confined groundwater (Table 2; Fig. 3a). Chloride concentrations range from 2.1 to 25.6 mg/l (mean of 7.7 mg/l) in the phreatic samples and from 0.9 to 13.1 mg/l (mean of 4.4 mg/l) in shallow confined groundwater and from 0.5 to 6.3 mg/l (mean of 2.2 mg/l) in the deep confined aquifers (Table 2; Fig. 3a). Nitrate concentrations range from 0.2 to 50.7 mg/l (above the maximum permissible level of the Italian law of 50 mg/l), with a mean of 23.7 mg/l, in the phreatic groundwater, from 0.2 to 30.4 mg/l (mean of 11.6 mg/l) in shallow confined groundwater and from 0.1 to 23.6 mg/l (mean of 4.1 mg/l, although only 15% of these groundwater samples have values of more then 4 mg/l, samples 1324, 1299 and 1294 of Table 1), in the deep confined aquifers (Table 2; Fig. 3a). Calcium concentrations range from 44.5 to 149.6 mg/l (mean of 79.2 mg/l) in the phreatic samples and from 41.4 to 117.2 mg/l (mean of 62.5 mg/l) in shallow confined groundwater and from 30.9 to 67.6 mg/l (mean of 49.5 mg/l) in the deep confined aquifers (Table 2; Fig. 3a). The distribution of SO4 is highly variable from 0.2 to 149 mg/l and generally increases with increasing depth (Table 2; Fig. 3a). Sulphate concentrations decrease away from the Tagliamento River, which has 130 mg/l of sulphate in its water at Pinzano, reflecting the gypsum outcrops in the Carnic Alps (Carniel 1999).

Table 2 Univariate overview of the groundwater samples
Full size table
Fig. 3 
figure 3

a Evolution of hydrochemical and stable isotope parameters with increasing depth. (crosses phreatic, squares shallow confined, filled circles deep confined groundwater samples), b Whisker plots of the major ion concentrations

Full size image

Pearson’s correlation matrices (Swan and Sandilands 1995) were used to find relationships between two or more variables. Only correlations with r ≥ ±0.4 are shown (Table 3). Samples showing r ≥ 0.7 are considered to be strongly correlated, whereas samples with an r of 0.5–0.7 show moderate correlation at a significance level (p) of 0.05. Strong correlations exist among the major elements Ca, Mg, HCO3 and EC for the phreatic groundwaters, and for Ca, Cl and EC for the shallow and deep confined groundwater samples (Table 3). These relationships clearly identify the tendency to follow a similar trend of evolution for the confined groundwaters. For the phreatic samples, a strong correlation exists between Ca and Mg, Ca and HCO3, Mg and HCO3, Mg and NO3, Na and Cl and HCO3 and NO3 (Table 3). For the samples collected from the shallow confined aquifer, a strong correlation exists between Ca and K, Cl, HCO3, NO3, between K and HCO3, Cl and HCO3, NO3 and between HCO3, NO3 (Table 3). The water samples collected from the deeper aquifers show strong correlations between Ca and Mg, Ca and Cl, Mg and Na, K (negative correlation), Na and K, Cl and NO3 and a negative correlation between HCO3 and SO4 (Table 3). The major exchangeable ions present in the studied data set Na–Ca, Na–Mg and K–Na correlate positively with the phreatic and shallow confined groundwaters, whereas they have a negative correlation with the deep confined samples. It is therefore possible to postulate that the simultaneous increase/decrease in ions is largely the result of dissolution/precipitation reactions and concentration effects. The effects of ion exchange with fine, silty layers are brought to evidence by the negative correlations with the deep confined groundwater samples (Adams et al. 2001).

Table 3 Pearson’s correlation matrices for water samples collected in the deep confined aquifer
Full size table

Environmental isotopes

The δ18O of groundwater ranges from −6.25 to −8.90‰ (mean value of −7.59‰) in the phreatic groundwater samples, from −6.83 to −9.14‰ (mean value of −8.05‰) and −7.01 to −10.35‰ (mean value of −8.74‰) in the shallow and deep confined groundwater samples, respectively (Table 2; Fig. 3a). The most depleted groundwater samples (1340: −10.32‰ and 1364: −10.35‰) are collected from the southwest of the Friuli Venezia Giulia Plain (Table 1) amongst the deep confined groundwater samples. Stable isotopes in the Friuli Venezia Giulia River waters show different values according to the mean altitude of their drainage basins. The river water samples, collected from 2005 to 2006, range from −7.63 to −8.60‰ for the Natisone River (Boschin et al. 2006), from −7.68 to −9.10‰ for the Isonzo River, from −8.23 to −9.31‰ for the Tagliamento River, from −7.09 to −9.98‰ for the Livenza River and from −7.94 to −9.78‰ for the Cellina River (CAMI 2007). Rainfall samples in Friuli Venezia Giulia monitored over a 2-year period for nine pluviometers and for 1 year for an additional four pluviometers, have δ18O weighted mean values ranging from −7.09 to −10.39‰, with an average value of −9.04‰ in the mountainous areas and from −7.45 to −7.78‰ for δ18O in the Friuli Venezia Giulia Plain with a mean of −7.62‰ (Cucchi et al. 2007). The vertical isotopic gradient for the Friuli Venezia Giulia Region is 0.29‰ per 100 m, in accordance with previous findings for Friuli Venezia Giulia (Longinelli and Selmo 2003; Longinelli et al. 2006) and for neighbouring Slovenia and Croatia (Vreča et al. 2006) as well as for Austria (Kralik et al. 2006).

Geochemical contour maps

Hydrochemical zones provide useful information on plausible recharge sources and the chemical evolution of groundwater (Mahlkenecht et al. 2006). Different hydrochemical zones were recognized and mapped in the Friuli Venezia Giulia Plain by Cucchi et al. (1999). However, the six zones previously mapped have been defined without differentiating the deep confined from the shallow confined groundwater samples and also without the statistical analysis approach used in this work. The general geographical distribution of the sub-surface groundwater samples (P and SC of Tables 1–3) is shown in Figs. 4 and 5 and is based on Kriging technique of the major ion distribution using ArcGIS by Esri. The deep confined groundwater samples have been not used to create these provinces because, as highlighted in the previous section, the geochemical characteristics indicate that they are separated from the sub-surface groundwaters and therefore have different physical and chemical signatures. The sulphate distribution map (Fig. 4a) reveals that the highest concentrations are found in the centre of the Plain, adjacent to the Tagliamento River, where maximum values of 100–125 mg/l are distributed throughout a NE elongate area—along the river course. The bicarbonate distributions (Fig. 4b) reach the maximum values of 339–460 mg/l in the northeast of the Friuli Venezia Giulia where the Isonzo River flows. The nitrate concentrations (Fig. 4c) have maximum values in two areas, one located in the northwest and the other in the central region close to the resurgence area. The maximum values in these two areas (from 34 to 43 mg/l) are close to the Italian limit of 50 mg/l for groundwater samples. Figure (Fig. 4d) shows the chloride distributions along the Friuli Venezia Giulia Plain. Chlorides in the area are higher in the east, where values between 6.4 and 11.9 mg/l are recorded. The magnesium distribution map (Fig. 5a) indicates highly variable data, with maximum values of 28–32 mg/l concentrated predominantly in an N–S elongate strip approximately 20 km wide. According to the calcium distribution map (Fig. 5b), the higher Ca values are recorded in the northeast of the Friuli Venezia Giulia Plain, where the Isonzo and Natisone rivers flow. Figure (Fig. 5c) illustrates the potassium distribution map, which highlights the maximum values located at the centre of the plain where three wells record values between 1.6 and 2.7 mg/l. The sodium distribution map is presented in Fig. 5d and indicates that higher values are located in the southernmost region of the studied area.

Fig. 4
figure 4

Contour maps of selected geochemical component distribution in the Friuli Venezia Giulia Plain. a, b, c, d refer to SO4 2−, HCO3 −, NO3 −, Cl−, respectively

Full size image
Fig. 5
figure 5

Contour maps of additional geochemical component distribution in the Friuli Venezia Giulia Plain. a, b, c, d refer to Mg2+, Ca2+, K+, Na+, respectively

Full size image

Discussion and conclusions

As a general rule, water-rock interaction processes affecting the geochemical composition of water appear to be time and exchange surface dependent (Barbecot et al. 2000). Considering the temperatures generally encountered in the catchments areas, the stable isotopes of water can generally be considered as conservative and not being affected by exchanges with either soil or rock (Barth 2000; Négrel and Petelet-Giraud 2005). In the case of complex aquifers, characterised by non-homogeneous water circulation, the combination of stable isotope data and geochemical ion distribution provides information on the down-flow groundwater evolution. The shallower groundwater samples indicate a wider range of δ18O values as well as greater standard deviation values (Fig. 6), which suggest that these waters have not mixed sufficiently to homogenise variations in isotopic composition of recharge waters or that they have fast recharge rates (Cartwright and Weaver 2005). While there is a general homogenisation of the δ18O values with depth, the deep confined groundwater shows more depleted values for confined groundwater with increasing depth (δ18O = −8.74‰) as described previously for other parts of the Padanian Plain (Zuppi and Sacchi 2004; Pilla et al. 2006) as well as for many groundwater systems internationally (Weaver and Bahr 1995; Purtschart et al. 2001; Vaikmae et al. 2001; Grasby and Chen 2005). The large variations between isotopic signature of phreatic and shallow confined waters in comparison to the deep confined groundwater (Fig. 6), as well as the marked depletion in the δ18O values of the deep confined groundwater samples in relation to the sum of the total ions (Fig. 7), points to complex groundwater circulation models that have probably substantially changed during the varying temperature regimes of the Holocene-Pleistocene (in particular during interglacial and glacial periods, where ice sheets covered the study area). Significant late Quaternary sea-level fluctuations, associated with alternating cooler and wetter periods would have changed the hydraulic gradients and partially or completely disconnected the deeper parts of the aquifer systems from the more active surface circulations as found in the neighbouring Veneto region (Zuppi et al. 2004). The deep confined waters are characterised by essentially passive hydrodynamic conditions with almost no cross-transfer with the polluted shallower groundwaters. The δ18O values indicate that there is very little continuity between the aquifers and that the deeper aquifers may be, at most, only partially recharged by the present rainfall and the infiltration of local rivers. In addition to lower δ18O values, the deep confined groundwater samples are characterised by relatively high Na–K values as highlighted by the Ludwig–Langelier diagram (Fig. 8). These deep confined waters have chemical characteristics that are distinct from water samples recovered closer to modern recharge areas, suggesting that the ion exchange between the deep confined groundwater and the silty-clay layers is therefore partially responsible for the groundwater chemistry variations with depth (Figs. 3a, 8). In addition, mineral saturation calculations indicate that the shallow and deep confined groundwaters are unsaturated with respect to gypsum, anhydrite and halite, whereas dolomite and calcite reach saturation in the great majority of the samples (Table 1). The phreatic groundwater samples have values that range from unsaturated to oversaturated for dolomite and calcite, whereas these samples are always unsaturated with respect to gypsum, anhydrite and halite (Table 1).

Fig. 6
figure 6

δ18 O range and mean values for the Friuli Venezia Giulia rainfall, rivers (from Boschin et al. 2006 and CAMI 2007) and groundwater samples (this study)

Full size image
Fig. 7
figure 7

Stable isotope δ18O ratio relative to the total anions expressed in Meq/l, (crosses phreatic well samples, triangles shallow confined well samples and circles deep confined well samples)

Full size image
Fig. 8
figure 8

Major ion water composition according to the Ludwig–Langelier diagram (crosses phreatic well samples, triangles shallow confined well samples and circles deep confined well samples)

Full size image

Four different hydrochemical zones have been recognised and mapped in the Friuli Venezia Giulia Plain, based on the chemical and isotopical information of the sub-surface (phreatic and shallow confined) groundwaters (Figs. 3a, 9).

Fig. 9
figure 9

Geochemical provinces of the Friuli Venezia Giulia Plain

Full size image

Zone 1 The Cellina-Meduna Province groundwaters are characterised by depletions in δ18O values, high bicarbonate (217–460 mg/l) and nitrate (due to intensive agricultural activity) ion (17–43 mg/l) and very low sulphates (0–50 mg/l) and potassium (0.2–1.1 mg/l) concentrations. The large difference in stable isotope data indicates that in this province the sub-surface groundwater is characterised mostly by river infiltrations (Cellina and Meduna rivers have depleted values in comparison to local rainfall, Fig. 6).

Zone 2 The Tagliamento River Province sub-surface groundwaters are strongly influenced by the Tagliamento River. This zone is characterised by very high concentrations of sulphates (50–125 mg/l) and chlorides (6.4–11.9 mg/l) and low bicarbonate (140–217 mg/l) and nitrate ion concentrations (0–17 mg/l).

Zone 3 The province between the Torre and Natisone rivers is characterised by high concentrations of nitrates (17–43 mg/l) due to intense agricultural activity and high calcium values (60–100 mg/l).

Zone 4 The Isonzo River Province is characterised by δ18O values similar to local rainfall, high bicarbonate (217–460 mg/l) and chlorides (6.4–11.9 mg/l) concentrations. The isotopic data are similar to local rainfall and imply that the sub-surface groundwater is mainly recharged by local rainfall.

The difference in water chemistry and stable isotope data between the sub-surface and the deep confined groundwater provides clear indications of the potential vulnerability of the groundwater resource to surface contamination from a variety of sources, such as agricultural practices, discharge from urban areas and small-scale industries located in the region. To date, groundwater with high nitrate and chloride contents is distributed patchily in the phreatic and shallow confined groundwaters, and there is little or no apparent evidence that fertilizers and pesticides have contaminated the deep confined groundwaters. Since the groundwater contributes to the surface water, shallow and deep aquifer vulnerability to overpumping and/or contamination may result in increased risk to water resources throughout the entire catchments’ region.