Spatial synchrony and persistence in fragmented landscapes: Models and experiments

Background/Question/Methods

Most species in nature comprise populations and subpopulations that are spatially subdivided to at least some degree. The overall effect of such subdivision on global extinction risk is challenging to predict, because it reflects the net outcome of numerous countervailing factors. Smaller populations are at greater extinction risk due to demographic stochasticity, but the more independent populations there are, the lower the global extinction risk. Dispersal can lead to re-establishment of previously-extinct populations and rescue populations that would otherwise go extinct. But dispersal also is thought to synchronize the fluctuations of different populations, thereby reducing the effective number of independent populations and increasing global extinction risk. While numerous experimental studies manipulate factors such as population size, number of populations, and dispersal rate and examine effects on local and global extinction risk, few attempt direct quantification and testing of the mechanisms  linking these factors to extinction. We have been using stochastic predator-prey models and experiments in a tractable model system (laboratory microcosms of protists) to quantify the mechanisms driving spatial synchrony and extinction risk. I will review our previous work, show some new results, and place our work in the context of other studies in both artificial and natural systems.

Results/Conclusions

Results of our experimental work support several theoretical predictions. Dispersal only synchronizes cycling populations. It does so by forcing the populations to cycle in phase, a phenomenon known as phase locking. Dispersal cannot synchronize non-cyclic population fluctuations arising from demographic and environmental stochasticity. Since most natural populations do not cycle, the implication is that dispersal-induced synchrony, and the associated increase in global extinction risk, may not be a major concern in most natural systems. However, local and global extinction risk in our system appear to be largely independent of dispersal rate or spatial synchrony. Similarly, in nature spatially-synchronized cycles often persist for extremely long periods, suggesting that synchrony-induced increases in global extinction risk are of minor importance even in cycling systems.  A better understanding of the links between synchrony, local extinction risk, and global extinction risk requires quantification of the rate at which dispersal re-establishes synchrony in the face of desynchronizing demographic stochasticity and other spatially-localized perturbations. In our experiments, higher rates of dispersal lead to more rapid establishment of phase locking (within as little as 1.5 cycle periods from an initially anti-synchronous state), but the relationship between dispersal rate and realized levels of synchrony is nonlinear.