Dispersal and Chaos in Spatially Structured Models: An Individual-Level Approach

An individual-level stochastic model of a single-species metapopulation model is presented. When density dependence is low, the local population dynamics are stable or cyclic. Increased coupling between local populations causes an increase in variability of local population values by amplifying demographic stochasticity. However, at an ensemble level, increased coupling induces greater spatial correlation and hence a reduction in population variability. When density dependence is high, the local dynamics are chaotic and extinction is common in unlinked populations. Increasing coupling does not lead to spatial correlation, and so ensemble persistence is enhanced, since local population extinctions can be revived by migration from neighbouring populations. The model predicts that increasing the coupling between neighbouring populations has no effect on the likelihood of observing chaos, both at a local and an ensemble level.