From a perspective of assessing the environmental risk of metals, it is essential to take into account their speciation and bioavailability at the ecosystem level as well as their paritioning at the cellular level. My research tends to better understand the behavior of metals, especially mercury and selenium at the biogeochemical and cellular level. First, I am interested in how the salinity gradient in the estuarine transition zone of the St. Lawrence River can modify the speciation and bioavailability of mercury and selenium. At the cellular level, I use the subcellular partitioning approach to evaluate the potential toxicity of mercury and the effect of its antagonist, selenium on Northern Pike (Esox lucius) in a stretch of the Saint-Maurice River modified by the construction of run of the river hydroelectric dams. In order to identify biological effects in this predatory fish, the results of the subcellular distribution are coupled to the responses of various biomarkers (oxidative stress, activity of selenoenzymes and identification of melanomacrophages). Finally, I also use the subcellular partitioning approach on Yellow Perch (Perca flavescens) from Lake Saint-Pierre, an ecosystem known for its metal inputs, in order to determine whether the subcellular distribution of several metals (Fe, Mn, Cu, Cd, As, Hg) vary according to their chemical class and their essentiality.