Response of freshwater periphyton to pesticides, their mixture and transformation products: New insight from untargeted metametabolomics

International audience

English. Pesticides are widely used in agriculture and ends up into aquatic environment where they may affect exposed organisms. All along this transfer process from field to river, pesticides can be degraded by abiotic processes and generate transformation products (TPs) for which the identity and the ecotoxicity is mostly unknown. Among these organisms, aquatic periphytons are complex assemblages of microorganisms with a short life cycle (e.g. microalgae, bacteria, etc.) playing a key role in aquatic ecosystems (e.g. primary production). Thus, they are relevant to investigate the impact of chemical contamination at the community level. Despite increasing knowledge on the impact of chemical stress, there is a still of paucity information about the effect of mixture of pesticides and their TPs on these communities. To such an end, metabolomics appears as choice method by prodiving a comprehensive picture of the molecular phenotype. In this context, our study aims to evaluate the toxicity of the herbicide terbuthylazine (TBA) and its transformation products on periphyton through the implementation of a multi-descriptors approach. To do so, periphyton was exposed during 4 weeks to TBA, to one of its commercially available degradation product (TBA-desethyl), both at 5 and 150 µg/L and to weathered TBA (wTBA, homemade by photodegradation from TBA at 5 and 150 µg/L). To enhance the environmental realism, periphyton was exposed to TBA-desethyl in mixture with 2 ubiquitous contaminants, glyphosate and AMPA (GA) at environmental concentrations (0.1 and 0.3 µg/L, respectively). A control (without any contaminants) and GA conditions were also set-up to allow comparison between conditions. First, most descriptors (photosynthetic yield, biomass, heterotrophic enzymatic activities) were strongly affected by TBA and not or slightly by the transformation products, being the TBA-desethyl or the wTBA or the mixture with glyphosate. Controversy, multivariate analyses (PCA and HCA) showed discrepancies in the metabolomics profile between each conditions, TBA 150µg/L being the most discriminated while both 5 and 150 µg/L of TBA-desethyl and wTBA were also separated from the controls. Nevertheless, all conditions with GA had a similar metametabolome profile suggesting no mixture effect. Overall, this study confirms the higher sensitivity of the metametabolome to micropollutants and so its relevance to highlight potential impact of TPs and mixtures on aquatic microbial communities.