Simulating dispersal of reintroduced species within heterogeneous landscapes

This paper describes the development and application of a spatially explicit, individual based model of animal dispersal (J-walk) to determine the relative effects of landscape heterogeneity, prey availability, predation risk, and the energy requirements and behavior of dispersing organisms on dispersal success. Significant unknowns exist for the simulation of complex movement behavior within heterogeneous landscapes. Therefore, initial simulations with J-walk examined the relative effect of landscape patterns and species-specific characteristics on dispersal success. Differences in landscape pattern were simulated by random generation of fractal maps with average available energy (i.e. prey) and predation risk expressed as a function of habitat type. Variation in species-specific patterns were then simulated by a series of scenarios that varied the response of dispersing individuals to habitat heterogeneity, including: habitat selection to maximize energy intake, habitat selection to minimize predation risk, or habitat selection contingent on energy reserves. Results showed that significant shifts in dispersal could be related to (1) the unique spatial arrangement of habitat within each map, (2) changes in relative prey abundance, and (3) variation in the relationship between energy availability and predation risk. Hypothetical management scenarios were used to identify critical data needed to assure the persistence of reintroduced populations of American martens (Martes americana). (C) 2003 Elsevier B.V. All rights reserved.