Extinction thresholds for species in fractal landscapes

Predicting species' responses to habitat loss and fragmentation is one of the greatest challenges facing conservation biologists, particularly if extinction is a threshold phenomenon. Extinction thresholds are abrupt declines in the patch occupancy of a metapopulation across a narrow range of habitat loss. Metapopulation-type models have been used to predict extinction thresholds for endangered populations. These models often make simplifying assumptions about the distribution of habitat (random) and the search for suitable habitat sites (random dispersal). We relaxed these two assumptions in a modeling approach that combines a metapopulation model with neutral landscape models of fractal habitat distributions. Dispersal success for suitable, unoccupied sites was higher on fractal landscapes for nearest-neighbor dispersers (moving through adjacent cells of the landscape) than for dispersers searching at random (random distance and direction between steps) on random landscapes. Consequently, species either did not suffer extinction thresholds or extinction thresholds occurred later, at lower levels of habitat abundance, than predicted previously. The exception is for species with limited demographic potential, owing to low reproductive output (R-o' = 1.01), in which extinction thresholds occurred sooner than on random landscapes in all but the most clumped fractal landscapes (H = 1.0). Furthermore, the threshold was more precipitous for these species. Many species of conservation concern have limited demographic potential, and these species may be at greater risk from habitat loss and fragmentation than previously suspected.