Diadromous species exhibit a life-cycle implying migrations between freshwater and ocean. Diadromy is observed in many fish lineages suggesting that adaptive traits are associated with this strategy such as better dispersal ability and more efficient tracking of the spatio-temporal variability of the environment. One of the hypotheses that have been put forward to explain the evolution and persistence of diadromy states that the purpose of migrations is to select environmental conditions that will optimize pre-reproductive growth. Growing at sea and reproducing in river (anadromy) is supposed to be favored when freshwater primary productivity is lower in freshwater than in the nearby sea. The reverse (growing in freshwater, reproducing in ocean; catadromy) is expected when productivity is higher in sea than in freshwater. According to our study, conducted at a global scale, this hypothesis holds true. In addition it is shown that diadromous species are overrepresented in species poor rivers and/or in those that have been open to colonization after the retreat of glaciers after the last glacial maximum. Because of their dispersal ability, diadromous species are also over-represented on oceanic islands. Biodiversity on islands is classical topic of biogeography but yet few studies have dealt with freshwater fishes. In our study of fish communities from the Polynesian islands, we demonstrate the generality of a model initially built to explain terrestrial biodiversity on oceanic islands. As predicted by this model, species richness peaks at peaks at intermediate island age. Also revealed by our results are a positive relationship between island elevation and species richness and a decrease in species richness as the distance from the nearest biodiversity hotspot increases. Diadromy has been a successful strategy over geologic times but in face to human activities it seems to contribute to extinction proneness, as exemplified by the numerous diadromous species red listed by IUCN. Paradoxically, our study about the biological and ecological traits shared by the freshwater fishes that have been recently globally extinct does not suggest that diadromy has been a major factor. This study points to endemism and restricted geographic range size as the major determinants of extinction. It seems that because of their dispersal abilities, diadromous species tend to have widespread historical distributions and, even if many species are declining, this prevented most of them from being entirely extirpated. This is well exemplified by the European sturgeon (Asipenser sturio), formerly widely distributed over Europe and now restricted to one reproducing population in the Garonne River, because of dramatic population collapses after 1850. To have a better idea of the fate of this species before 1850, we analyzed archaeozoological records. According to our analyses, sturgeons started to decline a long time ago, about 2500 years ago. Using reconstructed past temperatures, our analyses show that a high temperature increases the probability of finding this species in archeological remains but cannot explain entirely the observed decline. This suggests that human activities impacted sturgeon populations well before the industrial revolution. This thesis by synthesizing species occurrence data over different spatial and temporal scales contributed to a better knowledge about the diversity and vulnerability of diadromous fishes.
keywords: Macroecology, Diadromy, Extinction, Oceanic island, Archaeozoology