Causes and consequences of variation in dispersal in wood frogs

Background/Question/Methods

            Dispersal is a fundamentally important trait that influences ecological processes on scales from the individual to the community.  Many classical ecological and evolutionary models assume constant rates of dispersal, but these predictions can change in dramatic ways when variation in dispersal is included. Further, the evolution of dispersal is shaped by the costs of dispersal, yet tests for costs of dispersal remain uncommon. I tested for the causes and consequences of variation in dispersal using a long-term data set on wood frogs (Rana sylvatica). Metamorphosing wood frogs were marked at six ponds over five years, and then recaptured as adults that had either returned to their natal pond or dispersed to a new pond for breeding.  I investigated three factors hypothesized to affect dispersal in wood frogs:  conspecific density, summer rainfall, and mass at metamorphosis.  Increased conspecific density may increase rates of dispersal as an adaptive response to minimize competition.  Summer rainfall may reduce the costs of dispersal and increase dispersal rates.  Mass at metamorphosis is associated with increased locomotor performance and larger frogs may be better able to dispersal.  Additionally, I tested whether dispersing individuals incurred costs by comparing survival and adult mass of dispersing and non-dispersing individuals.

Results/Conclusions

            Dispersal probability increased with both larger mass at metamorphosis and greater summer precipitation.  This is consistent with larger animals having greater dispersal capacity, as well as greater precipitation reducing the costs of dispersal.  However, there was no evidence for density-dependent dispersal.  This was particularly surprising because density-dependence strongly affected larval and postmetamorphic survival, and recent theoretical work suggests density-dependent dispersal is likely to evolve under a wide range of conditions.  There was no evidence for dispersal incurring a growth cost to dispersing individuals, as there was no difference in the average size of dispersing and non-dispersing individuals.  However, dispersing individuals were less likely to survive to subsequent breeding seasons relative to non-dispersing individuals.  This is consistent with dispersal incurring a survival cost. The conditional dispersal observed here indicates that ecological and evolutionary models based on constant dispersal rates may be inappropriate for predicting the dynamics of more complex ecological systems.  Further, the costs of dispersal likely shape the evolution of conditional dispersal.