Abstract
Metals mining conducted during ages has left an environmental legacy of mining waste containing high levels of arsenic (As). The weathering of those waste generates AMD, which promotes the spread of As in aquatic ecosystems, resulting in negative impacts on human and environmental health. Despite that some methods exist to treat AMD, there is a need for more sustainable and cost-effective strategies adapted to historic mines and As-rich AMD, all of which being the benefits of bioremediation. This thesis is focused on the evaluation of two bioremediation approaches to treat As-rich AMD : 1) bio-oxidation of iron (Fe) and arsenic (As), which leads to their co-precipitation and 2) sulphate-reduction, which leads to the precipitation of As, Zn and Fe in the form of sulfides. The performances and bacterial community dynamics of field-scale bioreactors were monitored during one year, individually and coupled, for the treatment of the As-rich AMD from the Carnoulès mine. Water and precipitates were characterized using geochemical tools (elemental geochemistry and speciation) and environmental genomics (total and active bacterial communities analyzed by 16S rRNA metabarcoding, arsenite-oxidizing bacteria quantified by qPCR targeting aioA gene).The bioxidation bioreactors removed 43 ± 11 % of Fe and 67 ± 10 % of As from an effluent containing up to 111 mg/L As and 1067 mg/L Fe, using a residence time of 9 h. The sulfate-reduction bioreactor showed efficient glycerol to H2S conversion, with nearly 100 % As and Zn removal and less than 20 mg/L of organic carbon release, while decreasing residence time from 29 to 4 days. For both approaches, the residence time, the pH, the temperature, the physico-chemistry of the AMD and the biomass carrier (only for the biooxidation) are the main factors which influence the structure of the bacterial community in the bioreactors. The results suggest that the resilience and functional redundancy of the bacterial communities conferred robustness and stability to the treatment systems. Finally, the present thesis provides fundamental data and knowledge to progress towards a sizing of AMD bioremediation facilities based on knowledge-based practice rather than empirical practice.