Complementarity of passive samplers and biomonitoring tools to evaluate surface water contamination by pesticides in agricultural catchments

Grab sampling may be scarcely adequate to evaluate the impact on surface waters of pesticide used in agriculture because of its poor spatiotemporal representativeness compared to the high variability of the contamination patterns. Indeed, it is still not easy to link the effect of agricultural practice changes to water quality at the catchment scale. Furthermore, biological indicators may also be useful in addition to pesticide concentration determination to evaluate their impact on ecosystems. In this way, the contribution of new integrative tools has been evaluated. Passive samplers and biological indicators have been applied together at the catchment scale so as to identify their coherence and complementarity. The tested chemical tools are the passive-SBSE (Stir Bar Sorptive Extraction) dedicated to hydrophobic to moderately hydrophilic compounds and the POCIS (Polar Organic Chemical Integrative Sampler) dedicated to hydrophilic compounds. As concerns biological approach, it consisted in the in situ deployment of i) caging methodology with a highly ecological relevant amphipod species, Gammarus fossarum, coupled to the measurement of toxicity markers (acetylcholine esterase inhibition, feeding rate) and ii) the study of microbial and macro-invertebrate communities (measurement of leaf decomposition rates) as well as phototrophic periphytic communities (measurement of Pollution Induced Community Tolerance). These tools have been applied during june 2014 (corresponding to vineyard and corn period of application) in two contrasted agricultural catchments. The Ardières basin situated in the Beaujolais region (France) is submitted to viticulture pressure and a marked pesticide contamination has already been highlighted. Three measurement stations have been chosen so as to search for a contamination gradient from upstream to downstream. The Coise basin is located in the Monts du Lyonnais (France) in a zone of mixed farming breeding with a predominance of meadows, corn and cereals. Three measurement stations have been selected on three sub-catchments corresponding to differentiated pesticide use. In total, 49 pesticides (28 herbicides, 9 fungicides and 12 insecticides) have been analyzed as well as potential confusing parameters (metal(oïd)s, nutrients, temperature, luminosity, water flow). Results show the high coherence of chemical and biological tools responses. All of them were able to reflect the spatial contamination gradients as well as the specific contamination and impact associated to each agro-pedo-climatical context. A weekly exposure also permitted to take into account the temporal variability of the contamination according to flood occurrence. Finally, their sensitivity, reactivity and integration capacity make these tools relevant and promising to assess agricultural pesticide impact on surface water in addition to grab sampling.