Dispersal decisions underlie the spatial dynamics of metacommunities. Prey individuals may disperse to reduce the risk of either predation or starvation, and both of these risks may depend on conspecific density. Surprisingly, there is little theory examining how dispersal rates should change in response to the combined effects of predation and changes in conspecific density. We develop such a model and show that, under certain conditions, predators may induce dispersal at low prey densities but not high prey densities. We then experimentally manipulate the density of the ciliate Paramecium aurelia and the perceived presence of its predator, the flatworm Stenostomum virginiamum, in a two-patch metacommunity to parameterise the model. Paramecium dispersed in response to Stenostomum at low densities, but they reduced their dispersal in response to predation risk at high predator densities. By applying our model to the empirical data, we show that this switch in dispersal strategy, linked to increases in prey density, occurred because predators increased the difficulty or risk of dispersal. Together, the model and experiment reveal that the effects of predators on dispersal are contingent on prey density. Previous studies have sometimes reported an increase in dispersal rate when predation risk is elevated, and other times a decrease in dispersal rate. Our demonstration of a switch point, with predation risk increasing dispersal at low prey densities but reducing dispersal above a threshold of prey density, may reconcile the diversity of prey dispersal behaviours reported in these previous investigations and observed in nature.
- Non-consumptive effects
- Trophic interactions