Anaerobic batch and flow-through experiments were performed to assess the capacity of two organic substrates to promote denitrification of nitrate-contaminated groundwater within managed artificial recharge systems (MAR) in arid or semi-arid regions. Denitrification in MAR systems can be achieved through artificial recharge ponds coupled with a permeable reactive barrier in the form of a reactive organic layer. In arid or semi-arid regions, short-term efficient organic substrates are required due to the short recharge periods. We examined the effectiveness of two low-cost, easily available and easily handled organic substrates, commercial plant-based compost and crushed palm tree leaves, to determine the feasibility of using them in these systems. Chemical and multi-isotopic monitoring (δN, δO, δS, δO) of the laboratory experiments confirmed that both organic substrates induced denitrification. Complete nitrate removal was achieved in all the experiments with a slight transient nitrite accumulation. In the flow-through experiments, ammonium release was observed at the beginning of both experiments and lasted longer for the experiment with palm tree leaves. Isotopic characterisation of the released ammonium suggested ammonium leaching from both organic substrates at the beginning of the experiments and pointed to ammonium production by DNRA for the palm tree leaves experiment, which would only account for a maximum of 15% of the nitrate attenuation. Sulphate reduction was achieved in both column experiments. The amount of organic carbon consumed during denitrification and sulphate reduction was 0.8‰ of the total organic carbon present in commercial compost and 4.4% for the palm tree leaves. The N and O isotopic fractionation values obtained (ε and ε) were -10.4‰ and -9.0‰ for the commercial compost (combining data from both batch and column experiments), and -9.9‰ and -8.6‰ for the palm tree column, respectively. Both materials showed a satisfactory capacity for denitrification, but the palm tree leaves gave a higher denitrification rate and yield (amount of nitrate consumed per amount of available C) than commercial compost.

Managed aquifer recharge (MAR) is a well-known technique for improving water quality and increasing groundwater resources. Denitrification (i.e. removal of nitrate) can be enhanced during MAR by coupling an artificial recharge pond with a permeable reactive layer (PRL). In this study, we examined the suitability of a multi-isotope approach for assessing the longterm effectiveness of enhancing denitrification in a PRL containing vegetal compost. Batch laboratory experiments confirmed that the PRL, installed in 2011, was continuing to enhance denitrification. At the field scale, changes in redox indicators along a flow path and below the MAR-PRL system was monitored over 21 months during recharge (RP) and non-recharge (NRP) periods. Our results showed that the PRL was still releasing non-purgeable dissolved organic carbon five years after installation. Nitrate concentration and isotope data indicated that denitrification was occurring under and close to the infiltration area where recharge water and native groundwater mix. Furthermore, longer operational periods of the MAR-PRL system increased denitrification. Multi-isotope analysis might be useful in identifying and quantifying denitrification in MAR-PRL systems.

Permeable reactive barriers (PRBs) filled with zero-valent iron (ZVI) are a well-known remediation approach to treat groundwater plumes of chlorinated volatile organic compounds as well as other contaminants. In field implementations of ZVI-PRBs designed to treat these contaminants, nitrate consumption has been reported and has been attributed to direct abiotic nitrate reduction by ZVI or to denitrification by autochthonous microorganisms using the dissolved hydrogen produced from ZVI corrosion. Isotope tools have proven to be useful for monitoring the performance of nitrate remediation actions. In this study, we evaluate the use of isotope tools to assess the effect of ZVI-PRBs on the nitrate fate for the further optimization of full-scale applications. Laboratory batch experiments were performed using granular cast ZVI and synthetic nitrate solutions at pH 4-5.5 or nitrate-containing groundwater (pH=7.0) from a field site where a ZVI-PRB was installed. The experimental results revealed nitrate attenuation and ammonium production for both types of experiments. In the field site, the chemical and isotopic data demonstrated the occurrence of ZVIinduced abiotic nitrate reduction and denitrification in wells located close to the ZVI-PRB. The isotopic characterization of the laboratory experiments allowed us to monitor the efficiency of the ZVI-PRB at removing nitrate. The results show the limited effect of the barrier (nitrate reduction of less than 15-20%), probably related to its non-optimal design. Isotope tools were therefore proven to be useful tools for determining the efficacy of nitrate removal by ZVI-PRBs at the field scale.

Abstract. Artificial recharge of aquifers is a technique for improving water quality and increasing groundwater resources. Understanding the fate of a potential contaminant requires knowledge of the residence time distribution (RTD) of the recharged water in the aquifer beneath. A simple way to obtain the RTDs is to perform a tracer test. We performed a pulse injection tracer test in an artificial recharge system through an infiltration basin to obtain the breakthrough curves, which directly yield the RTDs. The RTDs turned out to be very broad and we used a numerical model to interpret them, to characterize heterogeneity, and to extend the model to other flow conditions. The model comprised nine layers at the site scaled to emulate the layering of aquifer deposits. Two types of hypotheses were considered: homogeneous (all flow and transport parameters identical for every layer) and heterogeneous (diverse parameters for each layer). The parameters were calibrated against the head and concentration data in both model types, which were validated quite satisfactorily against 1,1,2-Trichloroethane and electrical conductivity data collected over a long period of time with highly varying flow conditions. We found that the broad RTDs can be attributed to the complex flow structure generated under the basin due to three-dimensionality and time fluctuations (the homogeneous model produced broad RTDs) and the heterogeneity of the media (the heterogeneous model yielded much better fits). We conclude that heterogeneity must be acknowledged to properly assess mixing and broad RTDs, which are required to explain the water quality improvement of artificial recharge basins.

Abstract. Managed Aquifer Recharge (MAR) is a technique used worldwide to increase the availability of water resources. We study how MAR modifies microbial ecosystems and its implications for enhancing biodegradation processes to eventually improve groundwater quality. We compare soil and groundwater samples taken from a MAR facility located in NE Spain during recharge (with the facility operating continuously for several months) and after 4 months of no recharge. The study demonstrates a strong correlation between soil and water microbial prints with respect to sampling location along the mapped infiltration path. In particular, managed recharge practices disrupt groundwater ecosystems by modifying diversity indices and the composition of microbial communities, indicating that infiltration favors the growth of certain populations. Analysis of the genetic profiles showed the presence of nine different bacterial phyla in the facility, revealing high biological diversity at the highest taxonomic range. In fact, the microbial population patterns under recharge conditions agree with the intermediate disturbance hypothesis (IDH). Moreover, DNA sequence analysis of excised denaturing gradient gel electrophoresis (DGGE) band patterns revealed the existence of indicator species linked to MAR, most notably Dehalogenimonas sp., Nitrospira sp. and Vogesella sp.. Our real facility multidisciplinary study (hydrological, geochemical and microbial), involving soil and groundwater samples, indicates that MAR is a naturally based, passive and efficient technique with broad implications for the biodegradation of pollutants dissolved in water.