In the European Union, nitrate vulnerable zone (NVZ) should be designed for the mitigation of nitrate (NO3-) contamination caused by agricultural practices. Before establishing new NVZ, the sources of NO3-; must be recognized. A geochemical and multiple stable isotopes approach (hydrogen, oxygen, nitrogen, sulfur and boron) and statistical tools were applied to define the geochemical characteristics of groundwater (60 samples), calculate the local NO3-; threshold and assess potential sources of NO3-; contamination in two study areas (hereafter Northern and Southern), located in a Mediterranean environment (Sardinia, Italy). Results of the integrated approach applied to two case study, permits to highlight the strengths of integrating geochemical and statistical methods to provide nitrate source identification as a reference by decision makers to remediate and mitigate nitrate contamination in groundwater.Hydrogeochemical features in the two study areas were similar: near neutral to slightly alkaline pH, electrical conductivity in the range of 0.3 to 3.9 mS/cm, and chemical composition ranging from Ca-HCO3-; at low salinity to Na-Cl-; at high salinity. Concentrations of NO3-; in groundwater were in the range of 1 to 165 mg/L, whereas the nitrogen reduced species were negligible, except few samples having NH4+ up to 2 mg/L. Threshold values in the studied groundwater samples were between 4.3 and 6.6 mg/L NO3-;, which was in agreement with previous estimates in Sardinian groundwater.Values of delta S-34 and delta O-18(SO4) of SO42-; in groundwater samples indicated different sources of SO42-; . Sulfur isotopic features attributed to marine SO42-; were consistent with groundwater circulation in marine-derived sediments. Other source of SO42-; were recognize due to the oxidation of sulfide minerals, to fertilizers, manure, sewage fields, and SO42-; derived from a mix of different sources.Values of delta N-15 and delta(ONO3)-O-18 of NO3-; in groundwater samples indicated different biogeochemical processes and NO3-; sources. Nitrification and volatilization processes might have occurred at very few sites, and denitrification was likely to occur at specific sites. Mixing among various NO3-; sources in different proportions might account for the observed NO3-; concentrations and the nitrogen isotopic compositions. The SIAR modeling results showed a prevalent NO3-; source from sewage/manure. The delta 11B signatures in groundwater indicated the manure to be the predominant NO3-; source, whereas NO3-; from sewage was recognized at few sites. Geographic areas showing either a predominant process or a defined NO3-; source where not recognize in the studied groundwater. Results indicate widespread contamination of NO3-; in the cultivated plain of both areas. Point sources of contamination, due to agricultural practices and/or inadequate management of livestock and urban wastes, were likely to occur at specific sites.
