Non-random breeding site selection is consistent with tadpole performance in neotropical pastures and forests

Background/Question/Methods

Modified and intact habitats often differ in amphibian species richness, abundance, and composition, but the mechanisms underlying these patterns are poorly understood. Quantifying intraspecific and interspecific variation in vital rates across habitat types can improve our understanding of species occupancy within modified landscapes. I address whether adult frogs select breeding pools based on distance to the forest-pasture edge, and whether patterns of pool selection are consistent with tadpole performance for two model species, Engystomops pustulosus and Dendrobates auratus. Both experiments were conducted using artificial pools in pasture, edge, and forest habitats in the Osa Peninsula, Costa Rica.  I identified the species of eggs and tadpoles and quantified the number of deposition events for each species. In addition, I quantified the survival, growth, and development of E. pustulosus and D. auratus tadpoles.   

Results/Conclusions

Seven species deposited eggs or tadpoles in pools across the pasture-forest edge. Engystomops pustulosus selected pasture and edge pools, whereas members of the Dendrobatidae and Aromobatidae selected forest pools. The breeding sites that these species selected were consistent with E. pustulosus and D. auratus tadpole performance in pasture, edge, and forest. Survival of D. auratus tadpoles was low in pastures (< 5%) and higher in edge and forest (> 18%); in contrast, E. pustulosus survival was equally high in all habitats (> 80%). Dendrobates auratus metamorphs were largest in edges but larval period did not differ among habitats. Metamorphs of E. pustulosus were largest and larval period was shortest in pastures. My results suggest that adult frogs select breeding sites where tadpole performance is highest. Therefore, both adult behavior and tadpole vital rates contribute to patterns of frog occupancy in a mosaic of forest and pasture.