Introduction

The application of isotope-based methods has become well established for water-resource assessment, development and management in the hydrological sciences, and is now an integral part of many water quality and environmental studies (Clark and Fritz 1997; Cook and Herczeg 2000). To meet the drinking-water needs for future generations, sustainable watershed management and more detailed knowledge about recharge processes are essential. Environmental isotopes and chemical tracers are valuable tools for investigating recharge processes and groundwater-flow pathways in hydrogeological systems. Comparison of the stable isotopic compositions of precipitation and groundwater provides an excellent tool for evaluating recharge mechanisms (Clark and Fritz 1997; Jones and Banner 2003). Numerous studies using hydrochemistry and stable isotopes of water have been used to characterize recharge processes in different hydrogeological environments. For example, Barbieri et al. (2005) characterized groundwater flow paths and recharge elevations in a karst aquifer of central Italy by a combined study of hydrochemistry and stable isotopes (2H, 18O and 87Sr/86Sr). Marfia et al. (2003) combined the analysis of major and minor ions, δ13C and stable isotopes to investigate the origin and the hydrogeochemical evolution of surface and groundwater in a karst-dominated geologic setting in Central America.

Several investigations on the isotopic composition of rain have been carried out in volcanic areas such as East Maui, Hawaii, Cheju Island, Korea, Tahiti-Nui, French Polynesia, Mount Etna, Italy (Lee et al. 1999; Scholl et al. 2002; D’Alessandro et al. 2004) and Stromboli, Italy (Liotta et al. 2006).

The data collected previously by Paternoster et al. (2008), identify, in general terms, the groundwater isotopic characteristics within the Vulture basin (Fig. 1) in Italy, and were used in the present study. From that work, the local meteoric water line (LMWL; δD‰ = 6.56 δ18O + 4.12) was obtained, along with the weighted local meteoric water line (WLMWL; δD‰ = 4.92 δ18O –9.70), computed using the mean values weighted by the rainfall amount, identifying the monthly isotopic composition of precipitation over a two-year period at rain gauge stations.

Fig. 1
figure 1

a Geological map of central-southern Italy (from Bonardi et al. 2009, modified). b Geological setting of Mt. Vulture area (base geological map by Giannandrea et al. 2004, modified). The geological map is provided in Italian Gauss-Boaga, Zone Est coordinates, using the Roma Datum of 1940

Full size image

The objective of this study is to demonstrate how natural tracers, combined with hydrogeological and hydrochemical data, can be used to get a refined understanding of recharge mechanisms and recharge elevations of basin aquifers and to better understand the hydrogeochemical evolution of groundwater. The present methodology, as demonstrated by the example of the Mt. Vulture basin, constitutes a powerful tool to better define a conceptual hydrogeological model for complex hydrogeological systems. The improved understanding of groundwater flow patterns, especially within such complex volcanic and/or sedimentary environments, is fundamental for the preservation and sustainable management of water resources.

Study area

Geology

Mt. Vulture is an isolated cone (1,320 m above sea level (a.s.l.), stratovolcano-shaped, of Quaternary age, situated along the external edge of the Apennine Chain, close to the western portion of the Bradanic foredeep on the northeastern sector of the Basilicata region (Italy) (Fig. 1). Volcanic activity took place from the middle Pleistocene to upper Pleistocene, starting about 0.73 million years (Ma) ago and ending about 0.13 Ma ago (Brocchini et al. 1994; Buettner et al. 2006). The volcanic products consist of 700-m thickness of dominantly undersaturated silica pyroclastic deposits and subordinate lava flows arranged in radial banks with respect to the summit of Mt. Vulture (Brocchini et al. 1994; Serri et al. 2001; Giannandrea et al. 2004). The peripheral sectors of the volcanic structure are characterized by the presence of the fluvio-lacustrine deposits from Pliocene to lower Pleistocene age, with intercalations of pyroclastic layers (Fiumara di Atella Super-synthems, Giannandrea et al. 2006; Fig. 1). The oldest pre-Miocene bedrock units consist principally of deep-sea sediments belonging to units ranging from early Triassic to lower-middle Miocene (Boenzi et al. 1987; Principe and Giannandrea 2002). Beneath the Meso-Cenozoic substratum units, radiolarians and limestones of the Apulian platform are found to a depth of about 5 km (La Volpe et al. 1984). The fracture-fault systems characterized by NW–SE, NNW–SSE and E–W normal faults (Beneduce and Giano 1996; Schiattarella et al. 2005) involve the Mt. Vulture volcanic products and the bedrock units as well, highlighting a concentric-radial pattern jointed to the volcanic events, which controlled the area drainage system evolution, principally oriented in NW–SE and E–W directions (Ciccacci et al. 1999).

Climate

According to UNESCO/FAO (1963), the study area has a temperate Mediterranean climate, with moderately hot summers and cold winters. It shows a mean annual rainfall of about 750 mm year−1 (data based on observations from 1964 to 2006, Hydrographic Service of Civil Engineers of Puglia Region) with a maximum amount of rainfall from November to January (Fig. 2a). The maximum rainfall amounts are associated with the highest elevations of the study area (Fig. 2a). Spilotro et al. (2000) estimated an average annual precipitation of about 850–650 mm year−1 and a potential evapotranspiration of about 580 mm year−1. The annual average temperature for the Vulture area is about 13°C (data from 1964 to 2006), with a maximum from June to August (22°C) and a minimum between December and February (~5°C) (Fig. 2b).

Fig. 2
figure 2

a Mean monthly precipitation based on 40 years of observation (data from Hydrographic Service of Civil Engineers of Puglia Region, measured at Melfi gauge station, elevation 580 m a.s.l.). The upward arrows indicate the months characterized by maximum precipitation amount; while the downward arrows indicate the months characterized by minimum precipitation amount. The map shows the yearly rainfall distribution (contours) over the study area (data from Hydrographic Service of Civil Engineers of Puglia Region) in relation to topographic elevation. b Mean monthly temperature for 40 years of observation (data from Hydrographic Service of Civil Engineers of Puglia Region, measured at Melfi gauge station, elevation 580 m a.s.l.). The upward arrows indicate the months characterized by highest temperatures, while the downward arrows indicate the months characterized by lowest temperature

Full size image

Hydrogeology

The Mt. Vulture basin represents one of the most important aquifer systems of southern Italy. The aquifer core is mainly constituted of pyroclastic and subordinate lava flow layers, with different permeabilities which locally give rise to distinct aquifer layers. The principal hydrogeological complexes are volcanic products, with high-medium permeability values composed mainly of pyroclastic deposits and subordinate lava flows, which are the principal host aquifer rocks, of the Mt. Vulture basin (Fig. 3). The porous and fractured lava flows show a hydraulic conductivity (K) of about 10−1 cm s−1, while in the tuff and the incoherent pyroclastic deposits, with intergranular porosity, K is less (≈ 10−3 cm s−1). In the fluvio-lacustrine gravel deposits of the Fiumara di Atella Super-synthems, K is ≈ 10−2 cm s−1 (Spilotro et al. 2005). The structural hydraulic parameters and anisotropy features of the aquifer are the major factors that control groundwater flow pathways. Toward the S–SE, the volcanic aquifer thickness decreases considerably, leading to groundwater discharge where Fiumara di Atella Super-synthems deposits constitute a local extension of the volcanic aquifer with a relatively low groundwater circulation. Flow direction and rates are controlled primarily by the properties of the rock matrix and secondarily by the existing fracture network. The bedrock units are the marly-clayey complex, the calcareous-marly complex and arenaceous-conglomeratic-clayey complex showing less permeability. The different permeabilities of the volcanic products, fluvio-lacustrine deposits and the sedimentary bedrock allowed the construction of a detailed hydrogeological map (Fig. 3). The permeability of the aquifer host rocks varies from the highest to the lowest elevations of the Mt. Vulture basin. The juxtaposition of these hydrogeological features leads to different groundwater flowpaths and helps to better identify the main processes affecting the geochemistry of the groundwaters.

Fig. 3
figure 3

Hydrogeological map of the Mt. Vulture basin showing the main hydrogeological complexes (by Spilotro et al. 2006; modified) and the sampling points used for this study. The contour interval for the piezometric surface is 30 m. The map and the location data are provided in UTM Zone 33 coordinates, using the European Datum of 1950. Geology base map by Giannandrea et al. (2004)

Full size image

Two conceptual hydrogeological models have been proposed for the Mt. Vulture aquifer. Celico and Summa (2004) hypothesize two independent hydrogeological basins—one in the S–SE Monticchio-Atella sector, and one in the northeast Melfi-Barile area. The proposed southeastern groundwater basin is located between the most important faults of the volcano: the Grigi Valley-Fosso del Corbo fault to the north, and an unnamed fault in the south (Schiattarella et al. 2005). The proposed northeastern basin is the Melfi-Barile, characterized by radial drainage. The presence or absence of permeable rocks locally determines the existence of more inter-communicating basal layers. Recently Spilotro et al. (2005, 2006) suggested a new conceptual hydrogeological model in which the Mt. Vulture volcanic edifice is a huge aquifer where the spring regime shows a slightly seasonal variation of flow. As suggested by Spilotro et al. (2006), the surface drainage network constitutes the natural limits of the water catchments, where the pseudo-tectonic structure of Grigi Valley–Fosso del Corbo Fault is the only drainage axis widely affecting preferential groundwater flow and no evidence for the presence of the southern fault as a subordinate drainage axis was observed. The permeability and anisotropy features of the volcanic aquifer are the result of the magmatic and pyroclastic sequences and their successive arrangement and tectonic-volcanogenic deformation, creating locally shallower groundwater flows (Spilotro et al. 2006). As identified by Spilotro et al. (2006), the spring regime shows a slightly seasonal variation of flow, with higher values in spring and lower values in autumn.

Although different conceptual models of the Mt Vulture basin have been developed, they are mostly based on hydrochemical, hydraulic and geological observations. This study demonstrates how far the use of environmental tracers and their application to determine recharge sources and elevations of groundwater provide an efficient tool to improve the current understanding of hydrogeological systems.

Methodology, sampling and analytical procedures

This study reports analytical data for 48 groundwater sources (Fig. 3), located at elevations ranging from 352 to 1,000 m a.s.l., and including drilled wells used for irrigation and drinking-water supply and some springs. Sampling sites are located mainly in the W–NW, S–SE and N–NE sectors of the study area, principally within the volcanic units. Some sampling points of the S–SE and N–NE sectors are located within the fluvio-lacustrine sediments from Pliocene to lower Pleistocene age. Details of the sampling points are given in Table 1 (all location data are provided in UTM Zone 33 coordinates, using the European Datum of 1950). Seventeen sampled springs are characterized by elevation ranging from 352 to 960 m a.s.l. Thirty-one groundwaters samples were taken from the operating vertical and horizontal wells of several companies and private owners. The well depth and/or length is between 22 and 220 m and many of them are constructed inside the volcanic units (pyroclastic and effusive rocks).

Table 1 Description of sample collection sites. All location data are provided in UTM Zone 33 coordinates, using the European Datum of 1950. Elevation in m a.s.l. (meters above sea level); NA not available; well depth depth (meters) for vertical wells; well length length (meters) for horizontal wells
Full size table

A complete hydraulic dataset of the inventory performed by Regione Basilicata (1987, 1989; elevation of water levels in public/private wells and springs) has been integrated with measurements carried out in the present study using a global positioning system instrument (5-m precision) and piezometric level sensor (LP10) with high accuracy.

The water samples were collected during three sampling campaigns, from June 2007 to June 2008, for stable isotopes (18O, D) and major constituents (Ca2+, Mg2+, Na+, K+, Cl−, SO 2−4 , NO −3 ).

Measurements for water pH, temperature, electrical conductivity (EC), Eh and discharge were determined with high-resolution multiparametric probes (Idronaut, Ocean Seven 305; WTW–Tetracon 325). All water samples were stored in high-density polyethylene (HDPE; 50 ml) bottles with watertight caps, after filtration through a 0.45 μm Millipore filter for cation and anion analysis. Samples for cation analysis were preserved by acidifying to pH ~2 with concentrated HNO3. Alkalinity was determined in the field by titration with HCl (0.1 M). After sampling, all samples were stored at 4°C. Major ion determinations were carried out in the Gaudianello Spa laboratory on un-acidified (F−, Cl−, NO −3 and SO 2−4 ) and acidified (Na+, K+, Ca2+, Mg2+) water samples with separate aliquots by ion chromatography (Dionex CX-100). Ionic balance was computed for each sample taking into account major species. All samples exhibited imbalances lower than 5%.

The isotope investigations (δD and δ18O) were carried out for 48 groundwater locations, at which samples were collected one to three times during the present survey. Water samples were also analyzed at the Alfred Wegener Institute in Potsdam using a common equilibration technique with a Finnigan MAT Delta-S mass spectrometer equipped with two equilibration units for the online determination of hydrogen and oxygen isotopic composition (Meyer et al. 2000). The isotopic composition is based on international reference materials, V-SMOW (Vienna Standard Mean Ocean Water), as standard. The external errors of long-term standard measurements for hydrogen and oxygen are better than 0.8 and 0.1‰, respectively. Stable isotope compositions are reported, hereafter, as δ-values in parts per thousand (‰), calculated with respect to the V-SMOW international standard. The δ values are given by:

$$ \delta \left( {{\raise0.5ex\hbox{$\scriptstyle 0$}\kern-0.1em/\kern-0.15em\lower0.25ex\hbox{$\scriptstyle {00}$}}} \right) = 1000{\raise0.5ex\hbox{$\scriptstyle 0$}\kern-0.1em/\kern-0.15em\lower0.25ex\hbox{$\scriptstyle {00}$}} \cdot \left( {\frac{{{{\text{R}}_{\rm{sample}}} - {{\text{R}}_{{\rm{V}} - {\rm{SMOW}}}}}}{{{{\hbox{R}}_{{\rm{V - SMOW}}}}}}} \right) $$

where R sample and R V-SMOW represent the ratios of heavier to lighter isotopes (2H/1H or 18O/16O). R sample and R V-SMOW are the isotope ratios in the sample and the standard respectively. The sample is described as depleted in the heavier isotopes if the δ values are lower (more negative), and in contrast it is enriched in the heavier isotopes if the δ values are higher (more positive).

Results

Groundwater flowpaths

The existing groundwater flow information comes from water discharge points, including 106 springs and 119 shallow and deep wells (data from census of Regione Basilicata 1987, 1989,) consisting of unpublished static water level data from private wells (Gaudianello SpA; Traficante Srl, Itala Srl, Cutolo Srl) and public wells (Consortium of Melfi, National Irrigation Corporation) and the distribution of springs and their elevations. No significant groundwater level changes were observed when comparing the three samplings, from June 2007 to June 2008, confirming the suggestion of Spilotro et al. (2005) that the levels are fairly consistent.

A piezometric-surface map has been drawn using the geostatistical method of ‘ordinary kriging’ using SURFER software (Golden Software). As shown in Figs. 4 and 5, groundwater flowing within the volcanic aquifer moves from the highest to the lowest elevations along radial streamlines, illustrated by the flow arrows that are perpendicular to the lines of constant head. However, in some limited areas the different slope of the flow arrows is due to the automatic interpolation of the ordinary kriging geostatistical methods which do not show the real situation.

Fig. 4
figure 4

Water-table map, obtained using ordinary kriging (SURFER software, Golden Software), shows the irregularities of the piezometric surface. The circles drawn with black dotted lines encompass the irregularities of the constant head contour lines. The Grigi Valley–Fosso del Corbo Fault is shown as a red line. The map and the location data are provided in UTM Zone 33 coordinates, using the European Datum of 1950. The contour interval used for the piezometric surface is 40 m. The arrows show the groundwater flowpaths

Full size image
Fig. 5
figure 5

Piezometric elevation map showing the piezometric surface gradients (using SURFER). All the location data are provided in UTM Zone 33 coordinates, using the European Datum of 1950

Full size image

The radial deposition of the volcanic products favors groundwater flow from the highest elevations towards the periphery areas, as proposed by Spilotro et al. (2006). The irregularities in the radial symmetry of the aquifer are indicated by the equipotential lines. In the NW and SE sectors, these differences are probably due to the presence of the most important pseudo-tectonic structure (Grigi Valley–Fosso del Corbo Fault) with a NE–SE trend; while in the orthogonal directions the presence of lower permeability hydrogeological complexes lead to minor changes in the piezometric surface (Fig. 4).

The map/graph of the piezometric profile line of the volcanic aquifer shows that there are quite steep vertical groundwater gradients (Fig. 5). The present hydrogeological and hydraulic studies give emphasis to the conceptual hydrogeological model proposed by Spilotro et al. (2006). The isotopic investigation discussed in the following contributes additional knowledge of the Mt. Vulture volcanic aquifer.

Groundwater–chemical composition

The ionic compositions of the analyzed water samples are reported in Table 2. Observed groundwater temperatures range from 10°C (cold water) to 19.8°C (slightly thermal waters). ‘Thermal water’ is a term that was introduced by Nathenson et al. (2003) for springs which do not meet the numerical criterion of Reed (1983) but have temperature higher than non-thermal springs and usually have also dissolved constituents normally found in thermal waters. At the mean and low elevations, the temperatures of the groundwater (mean values 16.3°C) following the shallow flowpaths, were slightly above the mean annual air temperature (13.5°C).

Table 2 Dissolved contents of groundwater. Temperature, EC, pH, and Eh were measured in situ with a multiparametric probe that provided values assuming a water temperature of 20°C. The values reported are the arithmetic means of three samplings; BDL below detection limit; NM not measured
Full size table

The analyzed shallow and deep groundwater samples have variable pH values and redox state (Eh) which range from slightly acidic to neutral (pH 5.4–7.5) and from slightly reduced to oxidized (–36 mV< Eh< +185 mV), respectively; probably reflecting differences in circulation paths, water discharge and residence times within the aquifer. The increase in water acidity is principally due to the dissolution of CO2, which is the principal gas of magmatic origin. The high-dissolved CO2 contents in the groundwater of the volcanic aquifer are probably due to the ongoing active magmatic-mantle outgassing (Caracausi et al. 2009; Paternoster 2005). The mean EC values vary between 0.19and 17.97 ms cm−1. From the 1980s until 2006, the value of the EC increased more than 15% (data from census of Regione Basilicata 1987, 1989), probably due to decreases in total annual precipitation. The mean TDS values of the individual sites range from 208 to 18,000 mg L−1.

As shown in the Stiff and Piper diagrams (Fig. 6), the Vulture groundwaters generally display a chemical composition from bicarbonate alkaline-earth (group 1) to sulphate-bicarbonate alkaline (group 2). The principal dissolved anion bicarbonate is resulting from the reaction of dissolved CO2 to form HCO3 (Stumm and Morgan 1996). In contrast to soil CO2 in the unsaturated zone, the deep CO2 is directly injected in the saturated zone, with a high initial CO2 content, and therefore a low initial pH value. The low-pH water is very aggressive towards the host volcanic rocks, leaching their more soluble components. As a consequence, major ion constituents are progressively brought into solution, pH increases to the typical values measured in Vulture groundwaters, and CO2 is partially converted to bicarbonate. The dissolution of the host rocks depends on contact time between the rocks and water. Deep CO2 produces a fast and extensive enlargement of the fracture systems in the host rocks principally in the saturated zone and it may create a specific organization of flow patterns in the saturated zone (Annunziatellis et al. 2008 and references therein).

Fig. 6
figure 6

Relative amounts of major ions analyzed in springs and wells, plotted with a Stiff diagrams on the map, and b a Piper plot. All the location data are provided in UTM Zone 33 coordinates, using the European Datum of 1950

Full size image

Most of the water samples show high concentrations of Na+, K+, and Ca2+. The Na-excess found in a few analyzed water samples may be due to the hydrolysis of Na-silicates and also due to the exchange of Ca2+ for Na+, on the surfaces of clay-minerals (Na-smectite). This indicates an intensive water–rock interaction process, in agreement with results of Paternoster et al. (2009).

A few springs with the highest TDS values and a sulphate-bicarbonate alkaline composition showed higher mineralization, with elevated contents of SO 2−4 , Na+, Cl−, and dissolved CO2. As reported by Paternoster et al. (2009), the δ34S (SO 2−4 ) isotopic compositions of groundwaters with the highest concentration of SO 2−4 , displays sulphur isotopic values similar to those measured by Marini et al. (1994) in the magmas, supporting a main origin from the leaching of mineral weathering products such as feldspathoids, belonging to the sodalite group found in the volcanic host rocks. This similarity of sulphur isotopic values is due to the longer residence times of groundwater within the host rocks, representing the local extension of the volcanic aquifer where fluvio-lacustrine sediments with intercalated pyroclastic layers could contain entrapped brackish groundwaters. Along the flow path, from the mountainous area to the lower elevations close to the Atella River (Fig. 1), which drains the south part of the Vulture area, the groundwaters have the highest content of dissolved ions (Fig. 6a). In the W–NW sector, the groundwaters are characterized by lower mineralization values showing a bicarbonate alkaline-earth composition as a consequence of the interaction between volcanic rocks and groundwaters of meteoric origin, flowing at high-mid elevations, characterized by flowpaths with shorter residence times. At the lower elevations in the S–SE sectors, the springs near the base of the aquifer and occasionally in contact with the fluvio-lacustrine deposits, with intercalations of pyroclastic layers, are characterized as sulphate-bicarbonate alkaline composition with high salinity. This hydrogeochemistry variation is due to the deeper and lengthy flow pathways with longer groundwater residence time.

The hydrogeochemical differences between waters at the highest elevations, which probably are the main recharge areas, and the samples taken at the lowest elevations are consistent with the conceptual hydrogeological model previously described, from the core to the boundaries of the aquifer, indicating an important role played by aquifer radial symmetry.

Stable isotopic analysis (δ18O and δD)

Comparison of the stable isotopic compositions of precipitation and groundwater provides an excellent tool for evaluating recharge areas (Clark and Fritz 1997; Jones and Banner 2003). The principal groundwater recharge areas of the Mt. Vulture basin are identified using the yearly mean isotopic composition of water from precipitation samples collected monthly over a 2-year period by Paternoster et al. (2008). Using five rain gauges, elevations ranging from 320 to 1285 m a.s.l., the local meteoric water line (LMWL) and the weighted local meteoric water line (WLMWL) were defined. The rainwater isotopic composition showed a wide range of variation from –12.2 to –2.9‰ for δ18O and from –79 to –19‰ for δD, due to seasonal and elevation effects, but sometimes one effect prevails over another.

Stable isotopic composition of precipitation: seasonal and elevation effects

Dansgaard (1964) recognized the inverse correlation between the δ18O and δD of precipitation with temperature. With respect to the seasonal effect, Fig. 7 shows that the precipitation in summer months, when the air temperature values are higher, is characterized by heavier isotopic composition. On the other hand, winter precipitation has lighter isotopic composition.

Fig. 7
figure 7

Mean air temperature and δ18O. The monthly air temperatures are referred to 2002 (from Paternoster et al. 2008). The grey bars represent the mean air temperature and the black and blue line/circles represent the isotopic composition. The arrows indicate the months characterized by lowest and highest temperatures correlated to the lighter and heavier isotopes composition, respectively

Full size image

The elevation effect is also clearly observed. The δ18O and δD values become lighter with increasing elevation, whereas lower level sites such as rain gauge A (320 m a.s.l.), located at the bottom of the volcanic aquifer (with a mean yearly δ18O value of –6.9‰ and a mean yearly δD values of –42.1‰) and rain gauge B (520 m a.s.l.; δ18O = –7.5‰, δD = –46.4‰) showed more positive values (Fig. 8). Paternoster et al. (2008) observed a good correlation between the isotopic composition of rainfall and the sample’s elevation, defining a gradient of 0.17‰ δ18O/100 m, and 0.84‰ δD/100.

Fig. 8
figure 8

a Plot of δ18O in rain vs. elevation (m a.s.l.) of the five rain gauge stations (monthly rain samples collected during 2002 by Paternoster et al. 2008); b plot of δD in rain vs. elevation (m a.s.l.) of the five rain gauge stations (monthly rain samples collected during 2002 by Paternoster et al. 2008); c map showing the location of the rain gauge stations

Full size image

Groundwater isotopic composition: spatial distribution

According to previous studies in many temperate regions, as well as in this study, recharge rates appear to be the highest during winter and early spring when the soils are saturated, and vegetation is dormant (Wenner et al. 1991; Clark and Fritz 1997; Dennis et al. 1997; Winograd et al. 1998; Paternoster et al. 2008). In the Vulture area, recharge is minimal during summer when most of the short-duration and high-intensity rainfall is lost through direct surface runoff and/or is returned to the atmosphere due to the medium-high temperatures and high evapotranspiration rates (Spilotro et al. 2005). This confirms that the main recharge occurs during the winter and spring months.

The average values of the isotopic content in groundwater during the present study ranged from –8 to –10.2‰ for δ18O and from –53 to –65‰ for δD. The large numbers of stable isotope analyses performed for this study have been reported in Table 3. The observed variations in the measured isotopic values are often within the analytical error ranges (δ18O and δD are better than 0.8 and 0.10‰, respectively). These isotopic compositions are similar to mean isotopic values detected by Paternoster et al. (2008).

Table 3 Isotopic composition of each groundwater samplings. The minimum, maximum, and mean values of the three samplings are reported. NA not available, wells out of order. The external errors of long-term standard measurements for hydrogen and oxygen are better than 0.8 and 0.10‰, respectively
Full size table

No significant seasonal variation in the isotopic ratios is observed. However, values for a few water points (8, 13, 47 and 48) located in the W–NW sector fall slightly outside of the acceptable analytical error (Fig. 9). These springs show a high degassing rate (bubbling gases by Paternoster 2005) which could results in a shift of the δ18O isotopic composition towards negative values in the liquid phase. These slightly seasonal fluctuations in the δ18O and δD values are probably due also to the shorter groundwater flow pathway of the W–NW sector confirming the hydrogeological setting of the considered areas. Under these conditions, it is possible to assume that the hydrological features of the aquifer make the water bodies relatively homogeneous, which, as a consequence, are not influenced by the seasonal variation of the meteoric recharge.

Fig. 9
figure 9

Seasonal fluctuation of the groundwater isotopic compositions listed in Table 3. Error bars for each sample are ±0.1‰ for δ18O and ±0.8‰ for δD. Sample numbers (No. Sample) are as given in Table 3 but arranged by sector. The circled values falling slightly outside of the acceptable analytical error are relative to the springs which show a high degassing rate resulting in a shift of the δ18O isotopic composition towards negative values in the liquid phase

Full size image

By plotting the mean isotopic composition on a δD vs. δ18O scattergram, it can be seen (in Fig. 10) that all the samples fall between the Eastern Mediterranean meteoric water line (EMMWL) and the global meteoric water line (GMWL). It is observed that most of the groundwater samples follow the LMWL, in particular fitting the WLMWL (Paternoster et al. 2008) which defines more precisely the meteoric end-member in the local hydrological cycle, providing that the investigated groundwaters are meteoric in origin. However, one sample (sample No. 1, Table 3) is located to the left of the EMMWL, clearly suggesting a depletion in 18O and D (mean values of –10.4 and –59‰, respectively) with respect to the other groundwater samples of the area, and also falling outside of the range defined by the LMWL (Paternoster et al. 2008). The values obtained for this sample would indicate a higher recharge elevation compared with the mean recharge elevations of the other groundwater samples. This difference in δ18O and δD composition highlights the fact that the hydrological and structural setting probably involves the coexistence of different groundwater flowpaths influencing isotopic groundwater features. The water circulation is probably conditioned by a tectonic structure responsible for the ascent of deeper fluids. As suggested by Paternoster et al. (2008), it is likely that groundwater associated with sample No.1 (see Fig. 10) was recharged under different (e.g. colder) climatic conditions with respect to the present day, justifying the different isotopic values. Secondary processes that occur after the falling of rain such as evaporation and evapotranspiration, do not appear to change the isotope values of groundwater significantly, so that isotope ratios can be used as tracers.

Fig. 10
figure 10

Isotopic groundwater values of Mt. Vulture area. Global meteoric water line (GMWL) from Craig (1961); Eastern Mediterranean meteoric water line (EMMWL) from Gat and Carmi (1971); Local meteoric water line (LMWL) from Paternoster et al. (2008); Weighted local meteoric water line (WLMWL) from Paternoster et al. (2008)

Full size image

Figure 11 compares the piezometric map (shown previously in Fig. 4) with the spatial distribution of δ18O and δD in groundwater, countered using SURFER. As it is possible to note, in some sectors (such as the W–NW and central areas) the shape of the piezometric contours with constant head (Fig. 11a) fits the isotopic contours. This probably indicates that the groundwater flowing in the west-central sector is characterized by shorter flowpaths directly recharged from the rainfall. In the N–NE and S–SE sectors, the shape of the piezometric contours does not show a good fit with the isotopic contours, suggesting a more complex hydrogeological context where the flowpaths are probably longer and deeper than in the W–NW and central areas. The initial isotopic features are slightly modified, probably due to the highest groundwater residence times within the aquifer host rocks.

Fig. 11
figure 11figure 11

Isotopic contour maps drawn (using SURFER). a shows the piezometric surface contours and the area delineated in (b) and (c). Contour maps for b δ18O and c δD in groundwater samples; blue circles indicate the sampling points; blue lines indicate the volcanic aquifer limit. The map and the location data are provided in UTM Zone 33 coordinates, using the European Datum of 1950

Full size image

Discussion: recharge and discharge patterns of the Mt. Vulture volcanic aquifer system

The recharge elevation was calculated using the equation relating elevation and isotope ratios of precipitation with isotopic gradients of the Mt. Vulture area (–0.17‰ for δ18O/100 m and –0.84‰ for δD/100 m) reported by Paternoster et al. (2008). Recharge elevations for groundwater at the sampling points were calculated by resolving the Paternoster equation (δ18O‰ = –0.0017 H –7.28), with elevation (H) defining the corresponding δ18O values of the sampling points. The mean inferred recharge elevations are compiled (Table 4) and illustrated in Fig. 12. The measured elevations represent the spring elevations and the water-table elevations ( m a.s.l.) of the shallow and deep wells.

Fig. 12
figure 12

Inferred recharge elevations. δ18O values of springs and groundwater vs. elevation. The dotted line and the equation represent the regression line of δx, where x = O measurements in precipitation at different elevations from Paternoster et al. (2008)

Full size image
Table 4 Inferred recharge elevation computed using the equation relating elevations and isotope ratios of precipitation with isotopic gradients of Mt. Vulture area (–0.17 and –0.84‰ for δ18O/100 and δD/100 m, respectively) reported by Paternoster et al. (2008). For explanation see reported text. The table contains measured spring and well water-level elevations
Full size table

From the computed inferred recharge elevations (Fig. 12), three different sectors were defined to improve the conceptual hydrogeological model of the studied area. The sampling points of the S–SE sector, located at the bottom of the volcanic aquifer at low elevations (between 394 and 450 m a.s.l.), show an inferred mean recharge elevation ranging from 555 to 1,300 m a.s.l., which occur in the high and medium elevations of the area close to Mt. San Michele and Mt. Vulture (Fig. 13). The S–SE sector has an inferred mean recharge elevation that suggests a remote recharge area with relatively long residence times. The inferred recharge elevations of some sampling points (1, 2, 44 and 46; see Fig. 13) do not seem to be compatible with the local radial flow system shown by the water-table map of Fig. 4. In fact, there is a large difference between the water-table elevation and the estimated mean elevation of the recharge area. The water isotopic composition could result from a mixing of shallow and deep waters, probably influenced by tectonic structures, controlling groundwater flowpaths in the discharge area belonging to a deeper aquifer that was recharged under different (e.g. colder) climatic conditions with respect to the present day (Paternoster et al. 2008).

Fig. 13
figure 13

Recharge and discharge patterns of the Mt. Vulture volcanic aquifer system. The map and the location data are provided in UTM Zone 33 coordinates, using the European Datum of 1950. Rectangles with dashed black line show the three main sectors within the Mt. Vulture hydrogeological system. The S–SE and N–NE sectors are discharge areas; the W–NW sector is a recharge area. Question mark on the dashed flow lines shows that recharge elevations are uncertain. Mt. Vulture and Mt. San Michele are the main places of the investigated area

Full size image

The N–NE sector represents a minor discharge sector where the recharge area of the outflow points show an inferred elevation ranging from 900 to 1,300 m a.s.l., where spring and well water-table elevations range between 300 and 530 m a.s.l. Therefore, the N–NE sector is characterized by shorter groundwater flowpaths than the S–SE sector. In contrast, the W–NW sector of the Vulture aquifer shows mean inferred recharge elevations, ranging from 700 to 1,300 m a.s.l., compatible with the measured elevations of the springs and well water levels.

Based on this study, the S–SE and N–NE sectors represent the principal out-flowing areas where the deeper groundwater discharge is, respectively, about 7 and 3 km away (Fig. 13). In the W–NW sector, which represents the main recharge area, the groundwater flows are characterized by shorter flowpaths. Some springs discharge locally derived groundwater that circulates at shallow depths in the subsurface.

By using the mean inferred-recharge elevation obtained from the integrated isotopic analysis, it is possible to show graphically the recharge and discharge patterns to give a simplification of the conceptual hydrogeological model of the groundwater flowpaths in the investigated area (Fig. 13). The integration of the isotopic and hydraulic data with hydrogeochemical analysis, permits the hydrogeochemical tracing of the groundwater flows in the Mt. Vulture volcanic aquifer.

Concluding remarks

The integrated study of the rainwater isotopic composition and groundwater hydrogeochemical and isotopic characteristics of the Mt. Vulture area have highlighted the following points:

  1. 1.

    The geochemical characteristics of groundwater found in the investigated aquifer show heterogeneity among different sectors. The two identified hydrogeochemical water types reflect the water–interaction processes taking place within the host aquifer rocks.

  2. 2.

    The original composition of the recharge rainwater is modified by low–temperature leaching of the host volcanic rocks. The chemical results show the occurrence of two distinct groundwater types. The first water type displays a bicarbonate alkaline–earth composition. The second water type, flowing in the N–NE and principally in the S–SE sectors, displays a sulphate-bicarbonate composition with the highest Na and SO4 contents, possibly related to their circulation in fluvio-lacustrine deposits, with intercalations of feldspathoid-rich pyroclastic layers. This different composition is due to the complex geological and hydrogeological contexts of the aquifer.

  3. 3.

    Using the isotopic gradient (0.17‰ for δ18O/100 m), three main sectors within the Mt. Vulture hydrogeological system can be distinguished:

    • The W–NW sector, which constitutes the main recharge area of the studied aquifer, showing an isotopic signature similar to rainwater. The slightly seasonal fluctuations of a few water points, characterized by the occurrence of a high degassing rate (bubbling gases), could results in a shift of the δ18O isotopic composition towards negative values in the liquid phase. These slightly seasonal fluctuations in the δ18O values are probably also due to the shorter groundwater flowpath of the W–NW sector, confirming the hydrogeological setting of the considered areas. The recharge contributing to groundwater appears to be derived from a mean elevation ranging from 700 to 1,100 m a.s.l., close to the measured elevation of the springs and well water levels.

    • The second and third sectors represent the main discharge areas toward the S–SE and N–NE areas respectively. No evidence for seasonal variation of the groundwater isotopic composition was observed. The mean elevation of the recharge area of the deeper groundwater is about 1,100 m a.s.l. which is a hydrologically reasonable value for the Mt. Vulture area. This implies that the deeper waters that are recharged at higher elevations have longer flowpaths. This confirms that most flow in the Mt. Vulture region conforms to the radial pattern, but from the highest to the lowest elevations there are some irregularities in the flow. Based on this study, the S–SE and S–SE sectors represent the principal out-flowing areas where the deeper groundwater discharge is derived from about 5 km away.

  4. 4.

    This investigation improved the previous conceptual hydrogeological model proposed by Spilotro et al. (2005, 2006), which, at large scale, defined the radial streamlines of the groundwater flowpath. The present hydrogeochemical, hydraulic and integrated isotopic study allowed better definition of the recharge and discharge patterns of the Mt. Vulture volcanic aquifer system. By means of this study, the W–NW area, at the highest elevations, was found to be the main recharge area close to the drainage axis (Grigi Valley-Fosso del Corbo fault), widely affecting the preferential groundwater flow, where several operating wells of companies and private owners are located. The flowing groundwater moves along radial flowpaths toward the lowest elevations in the S–SE and S–SE sectors but with differences in length and depth. This fact is supported also by the groundwater hydrogeochemical differences from the highest to the lowest elevations of the volcanic aquifer.

As mentioned previously, several hydrogeochemical and stable isotope studies have been performed to characterize recharge and discharge processes in quite simple hydrogeological environments. The studies in volcanic aquifers were found to be more complex. The complex hydrogeochemical characteristics of groundwater at a volcanic site have resulted in the development of stable-isotope and chemical-tracer tools for investigating recharge processes and groundwater flow pathways of these hydrogeological systems. Given the importance of the studied area, the present investigation can be useful to the studies of volcanic aquifers comparable to the Mt. Vulture system.

The presented work can contribute to policies of conservation and management, also with time-based criteria, in sensitive recharge areas which should be protected for drinking-water supply and other uses. Therefore, this study may be used as an aid in regional planning to establish groundwater management rules.