We present research investigating how, with in depth knowledge of the community, microbial communities may be harnessed for bioremediation of hazardous water contaminants. We focused on the bacterial reduction of perchlorate, a common water contaminant. For this we studied the structure and capabilities of a novel sulfur-utilizing, perchlorate-reducing bacterial (SUPeRB) consortium. Initially, we characterized the minimal consortium that retained functional capabilities, using 16S rRNA and functional gene analysis. A diverse functional consortium dominated by Beta-Proteobacteria of the family Rhodocyclaceae and sulfur-oxidizing Epsilon-Proteobacteria was found. We also examined the optimal growth conditions under which perchlorate degradation occurred and uncovered the upper limits of this function. Bacterial isolates were screened for function and the presence of functional genes. We expanded to bioreactor studies at bench- and pilot-scale, and first used a perchlorate-reducing, bench-scale bioreactor to probe the stability of the microbial ecosystem. During stable reactor function, a core consortium of Beta- and Epsilon-Proteobacteria reduced perchlorate and the co-contaminant nitrate. A disturbance of the vi consortium led to a failure in function and to higher system diversity. This suggests that the SUPeRB consortium was not metabolically flexible and high population diversity was necessary for a return to stable function. In a pilot-scale bioreactor we determined that the SUPeRB consortium could stably degrade low levels of perchlorate to below the EPA maximum recommended limit. Field conditions, such as temperature extremes and intermittent perchlorate feed, did not negatively impact overall function. When all reactor consortia were compared we observed that the volume of the reactor and the initial inoculum were not as important to stable reactor function as the acclimatization of the consortium to the system and maintenance of favorable conditions within the reactor. In summary we found that the SUPeRB consortium successfully degraded perchlorate in multiple systems. The study of this novel consortium expands our knowledge of the metabolic capabilities of perchlorate-reducing bacteria and suggests potential evolutionary pathways for perchlorate-reduction by microorganisms. The SUPeRB consortium may be used to establish bioremediation systems for perchlorate and other environmental contaminants.

Dans les écosystèmes forestiers pauvres en nutriments, les minéraux du sol constituent la principale source de nutriments inorganiques nécessaires à leur bon fonctionnement. Néanmoins ces nutriments ne sont pas directement accessibles aux racines des arbres. C'est l'action conjointe de facteurs abiotiques, comme le pH ou la circulation de l'eau, et biotiques comme les racines ou les microorganismes du sol dont les bactéries, qui vont conduire à l'altération de ces minéraux. A ce jour, nos connaissances sur les communautés bactériennes impliquées dans le processus d'altération et leur distribution dans des sols forestiers restent limitées, notamment à des habitats tels que la rhizosphère et la mycorhizosphère. Les objectifs de cette thèse étaient de caractériser les communautés bactériennes colonisant les minéraux du sol et leur aptitude à altérer les minéraux et enfin d'identifier les gènes bactériens impliqués. La combinaison d'approches cultivable, non cultivable et de biogéochimie sur des minéraux enterrés pendant 4 ans dans un sol forestier, a démontré que leur surface était colonisée par des communautés bactériennes spécifiques, capables d'altérer les minéraux et présentant des capacités métaboliques limitées, suggérant que ce support pourrait être considéré comme un habitat : la minéralosphère. La relation minéral/bactéries a été testée in situ via un amendement minéral sur une plantation et a mis en évidence l'impact de la disponibilité en nutriments sur la structuration des communautés bactériennes capables d'altérer les minéraux. L'étude génétique réalisée sur la souche modèle PML1(12) a révélé l'implication de plusieurs mécanismes dans la fonction altération / In nutrient-poor forest ecosystems, minerals are the main source of inorganic nutrients for the long lasting functioning of the forests. However, these nutrients are not directly accessible to the tree roots. It is the joined action of abiotic factors, such as pH and water circulation, and biotic factors such as tree roots and soil microorganisms, and notably bacteria, which leads to the solubilisation of these minerals. To date, our knowledge of the bacterial communities involved in the mineral weathering process and their distribution in forest soils is very limited and remains restricted to habitats such as the rhizosphere and mycorrhizosphere. The goals of this PhD thesis were to characterise the mineral associated bacterial communities, their ability to weather minerals and finally to identify the bacterial genes involved in the mineral weathering process. The combination of geochemical, cultivation-dependent and -independent approaches applied on minerals grounded in a forest soil during 4 years, revealed that the mineral associated bacterial communities were specific, able to weather minerals and had restricted metabolic abilities. These results suggest that minerals could be considered as a true ecological habitat: the mineralosphere. The mineral/bacteria relationship was tested in situ through a mineral amendment applied on a small-scale plantation, which has highlighted that the nutrient availability impacted the functional structure of the mineral weathering bacterial communities. At least, random mutagenesis applied on a model mineral weathering bacterial strain revealed that its mineral weathering ability resulted from several molecular mechanisms