Numerically exploring habitat fragmentation effects on populations using cell-based coupled map lattices

We examine habitat size, shape, and arrangement effects on populations using a discrete reaction-diffusion model. Diffusion is modeled passively and applied to a cellular grid of territories forming a coupled map lattice. Dispersal mortality is proportional to the amount of nonhabitat and fully occupied habitat surrounding a given cell, with distance decay. After verifying that our model produces the results expected for single patches of uniform habitat, we investigate heterogeneous and fragmented model landscapes. In heterogeneous single-patch systems near critical patch size, populations approach Gaussian spatial distributions with total population constrained by the capacity of the most limiting cell. In fragmented habitat landscapes, threshold effects are more complex and parametrically sensitive. The results from our experiments suggest the following: the ability to achieve persistence in hyper-dispersed patchy habitats by adding similarly fragmented patches requires meeting threshold reproduction rates; persistent metapopulations in which no local population is individually persistent appear when dispersal distances and reproduction rates are both high, but only within narrow parameter ranges that are close to extinction thresholds; successful use of stepping-stone patches to support metapopulation systems appears unlikely for passively diffusing species; elongated patches offer early colonization advantages, but blocky patches offer greater population resilience near extinction thresholds. A common theme running through our findings is that population viability estimates may depend on our ability to determine when population and habitat systems are approaching extinction threshold conditions, (C) 1999 Academic Press.