Little is known about natural extinction rates of freshwater fish populations in the absence of catastrophic events. Theory (Lande 1993) and empirical evidences demonstrate that extinction risk increases as population size decreases because of a higher likelihood of stochastic extinction. Population size is therefore the best predictor of extinction probability but is difficult to estimate in the field for many fish species. Fortunately freshwater fish populations frequently occur in insular habitats (e.g., lakes, catchments), and area of occupancy (e.g., lake area) could be used as a rough proxy for population size. In this WP we will therefore quantify the relationship between the area occupied by a fish population and its expected lifetime, the aim being to assess the consequences of habitat shrinking due to human driven fragmentation. Because restoring connectivity may cause upstream dispersal of both native and non-native species, we will also incorporate effects of key invasive predators and competitors in addition to area (Englund et al. 2009).

In a fragmented world the persistence of a species at a regional scale depends on the interplay between local extinction and colonization, processes dealt with by metapopulation theory. To assess the persistence of metapopulations in differently fragmented rivers-capes and under different scenario of future fragmentation, we will combine extinction and colonization models from WP1. The results of this WP will be used in Task 4.2 (WP leader Bernard Hugueny).