Groundwater contamination by pesticides and their metabolites: intrinsic properties and hydrodynamic constraints
Résumé
Enhanced groundwater monitoring has highlighted the contamination by pesticides and their metabolites which can be more problematic than parent compounds. The fate of pesticides has widely been studied in soil, where major processes such as sorption and degradation lead to reduction of the flux likely to leach into the groundwater. Thanks to laboratory experiments, identifying the key parameters controlling these processes (role of organic carbon, pH influence on ionic pesticide, formation of bound residues,...) is now possible. As these processes are not instantaneous, it appears necessary to evaluate the impact of hydrodynamic by studying, for example, the leaching of these molecules into soil columns. However, soil must not be the only compartment investigated. Indeed when dealing with groundwater, it is essential to also consider both the unsaturated and saturated zones. Studying processes in the unsaturated zone is a challenge, especially when its thickness exceeds several meters. However drilling of non-destructive cores (without fluid) allows the collection of geological materials that can be considered for laboratory experiments (batches, respirometers, percolation through undisturbed columns,...). While a long-term and regular monitoring is essential to describe the groundwater contamination, it is also crucial to implement multidisciplinary approaches to understand the transfer of these substances into aquifers. Assessment of groundwater quality by statistical approaches such as the Mann-Kendall test allows the description of trends, but cannot predict the impact of remediation measures applied to reduce the contamination. Taking into account hydrodynamics by using the age of the water (calculated by tritium, CFCs and SF6 techniques) may be useful to understand spatial and temporal trends of pesticide concentrations in groundwater at both aquifer and regional scales. Geochemical approaches are also required to understand how water and solutes migrate across the aquifer. More recently, the characterization of the stable isotope signatures (e.g. δ13C, δD) of pesticides showed promise to discriminate sources of contamination and processes taking place in the aquifers (degradation vs. dilution, biodegradation vs. chemical degradation). Lastly, modelling appears a way to evaluate the impact of the superimposed processes occurring in the soil-unsaturated-saturated zones continuum, provided that each compartment is considered separately. Examples resulting from different lab and field experiments will be presented here in order to demonstrate how the implementation of multidisciplinary approaches leads to a better description and understanding of the pesticide fate and their impact on groundwater quality.