In freshwater systems, the abundance and diversity of long-distance migratory fish are limited by the maintenance of longitudinal connectivity and natural flow regimes of rivers. Using a graph-based view of each riverscape, we analyzed the river fragmentation process and overlapped with the probable number of migratory species in each remaining stretch. Applying this methodology in basins, we assess historical and potential scenarios quantifying segment extensions free from dams determining the fish richness based on the available habitat for each species. The highest number of migratory species was observed in fragments longer than 100 km. In the future scenario, there was an increase in the number of fragments, with an increase in the number of stretches shorter than 50 km, inadequate to maintain most of the migratory species. Segments of the highest classification order and located in the longest lotic fragments were considered the most important for the species habitat conservation. Dam construction in these segments could seriously affect the ecological processes at a regional level. The proposed analyses enable to approach basins with high diversity of species and nonsalmonid species, supporting the lack of base data concerning those areas, and determine priorities for studies and conservation.
Keywords: Barriers; Connectivity; Dams; Ecological network; Longitudinal connectivity; Stream network.