The significance of examining groundwater quality is highlighted by the fact that 75% of the European Union population relies on groundwater for their water supply. In the Republic of Serbia, over 50% of the groundwater used for public water supply comes from alluvial aquifers, with 80–90% of this water originating from river water infiltration. This study investigates the origin of increased ammonium concentrations (up to 4.7 mgN/l) in the Danube alluvial aquifer near Kovin-Dubovac area, Serbia, a potential future regional water supply source surrounded by intensive agricultural production and an open coal mine. Comprehensive research involved statistical processing of physicochemical data (13 parameters from 33 sampling sites), isotopic analyses (δ15N–NH4 – 12 samples; δ34S–SO4, δ18O–SO4 in 5 samples), and microbiological tests (denitrifying, sulphate-reducing, iron-related, and heterotrophic aerobic bacteria in 15 samples). Factor analysis revealed significant positive loadings (>0.5) among indicators of autochthonous organic matter origin (Fe2+, Fetot, and TOC), geological matrix components (Na, H2S, Cl), and groundwater state parameters (pH, Eh, and NH4+). This multifaceted approach and the spatial concentration gradients of parameters associated within extracted principal components revealed the predominance of two NH4+ sources. The riparian zone is characterized by sediment organic matter mineralization, increased iron content, and natural ammonium origin (δ15N–NH4+ from +4.82‰ to +6.93‰) accompanied by conditions suitable for DNRA and sulphate reduction. Approaching to the hinterland lower iron and total organic matter content, accompanied by increased redox values revealed the signature of fertilizers application (δ15N–NH4+ −0.84‰ and −0.33‰), associated with denitrification influence. This multifaceted approach reduces ambiguity, providing a clearer interpretation of results when discerning the origin of nitrogen and aquifer potential for nitrogen loss or conservation in a reducing groundwater environment.
Read full abstract