Spatially separated populations of the same species often exhibit correlated fluctuations in abundance, a phenomenon known as spatial synchrony. Dispersal has been shown both theoretically and experimentally to increase synchrony. However, the types of dispersal patterns investigated tend to be fairly simplistic: either emigrants from a given population are equally likely to move to any other population (global dispersal), or else they only move to the nearest populations (nearest-neighbor dispersal). Natural dispersal patterns tend to be "fat-tailed": most individuals disperse short distances, but a minority disperse very long distances. What is the effect of fat-tailed dispersal on spatial synchrony? On the one hand, occasional long-distance dispersal might synchronize widely separated populations. On the other hand, individuals that disperse long distances would otherwise have dispersed short distances. Occasional long-distance dispersal therefore might reduce spatial synchrony among nearby populations without occurring at a sufficiently high rate to generate synchrony among widely separated populations.
To test these alternative hypotheses, we assembled replicate metapopulations of the protist predator Euplotes and its prey Tetrahymena in laboratory microcosms. Each metapopulation comprised 15 populations. We subjected metapopulations to one of two dispersal treatments, imposed by pipetting culture medium and the organisms in it between culture vessels. Eight of the metapopulations experienced nearest-neighbor dispersal. The other eight experienced "small world dispersal", a form of fat-tailed dispersal: dispersers had a 30% chance of dispersing to a randomly chosen population instead of to their nearest neighbor.
Results/Conclusions
Compared to nearest-neighbor dispersal, small world dispersal significantly increased spatial synchrony at long distances without decreasing it at short distances. Mean synchrony (cross-correlation) for small world dispersal was 0.81±0.02 while for nearest neighbor dispersal mean synchrony was 0.70±0.02. Distance decay of synchrony occurred only for nearest-neighbor dispersal. Using mathematical models and data from previous experiments, we show that these results are due to the nonlinear relationship between dispersal rate and synchrony in our system. Increasing long distance dispersal at the expense of nearest neighbor dispersal increases synchrony of widely separated populations while still leaving sufficient nearest-neighbor dispersers to produce maximally high synchrony between neighboring populations. However, our models show that the opposite result (occasional long-distance dispersal reduces short-distance synchrony without increasing long-distance synchrony) is also possible, depending on the shape of the dispersal rate-synchrony relationship. The shape of this relationship depends on system-specific ecological details.