Testing assumptions about amphibian movement behavior with field experiments

Background/Question/Methods

Stewardship of species in fragmented landscapes often seeks to mitigate the effects of isolation on populations. This is often done by increasing landscape connectivity, which is intended to facilitate the dispersal of individuals among different populations. Landscape connectivity could be assessed with dispersal data, but dispersal is often difficult to directly observe, so data on movement behavior are frequently used to assess landscape connectivity. Studies of landscape connectivity frequently make 3 assumptions regarding movement behavior: individuals within a population behave in the same way, individuals respond to forest canopy structure rather than groundcover, and related species can be viewed as surrogates for one another. We test these assumptions with a series of three field experiments on amphibians.  First, to assess the role of individual variation on movement behavior, we examined the effect of larval density on movement rates of ornate chorus frogs. Second, to assess the response of amphibians to canopy structure and groundcover, we examined the response of ornate chorus frogs and eastern newts to the removal of groundcover. Finally, to test the role of surrogate species, we compared the movement rates and habitat preferences of two related species, the southern leopard frog and the gopher frog. 

Results/Conclusions

Results from these experiments indicate that assumptions regarding movement behavior in studies of landscape connectivity may not be met. The first experiment showed that ornate chorus movement behavior was affected by larval density. Chorus frogs raised at high densities moved faster in high quality habitat than in low quality habitat, but chorus frogs raised at low densities moved faster in low quality habitat than in high quality habitat. Ornate chorus frogs and eastern newts also responded to the manipulation of groundcover in the second experiment by moving faster in areas where groundcover was removed. Finally, gopher frogs moved faster than southern leopard frogs, and gopher frogs preferred to move into open habitats while leopard frogs did not show a significant preference for forest or open habitat.  We show that amphibian movements can respond strongly to environmental factors that are implicitly assumed to have no effect on movement behavior in studies of landscape connectivity. These results highlight the need to consider potential sources of variation in movement behavior when assessing landscape connectivity.