Abstract
Because of their high densities and generalist feeding behaviors, the introduced frog, Eleutherodactylus coqui, has been hypothesized to consume and potentially reduce endemic invertebrates in Hawaii. To address this hypothesis, I compared E. coqui diets to invertebrate abundances in 11 sites on the Islands of Hawaii and Maui in the summer of 2004. At each site, I collected between 22 and 119 frogs from 20 × 20-m plots, and invertebrates from light traps, beating trays, and leaf litter samples. Prey items in frog stomachs were identified to order, and invertebrates collected in environmental samples were identified to the lowest taxonomic category possible. Multivariate analyses of diet content and invertebrates collected at each site suggest that most prey was from the leaf litter. Non-native ants (Hymenoptera: Formicidae) and amphipods (Amphipoda: Talitridae) comprised 30% and 22%, respectively, of the total prey items consumed. These non-native invertebrates were more abundant in stomachs of E. coqui than in the environment indicating a preference for these species. There was little evidence that E. coqui were reducing important invertebrate pests. No mosquitoes (Diptera: Culicidae) were found in stomachs, and termites (Isoptera) comprised <1% of the total prey items. Arthropod orders containing endemic species that appear most vulnerable to E. coqui predation include Acarina (mites), Coleoptera (beetles), Collembola (springtails), and Diptera (flies), which each made up >2% of the diet of E. coqui. Dominant prey items in frog stomachs differed among study sites suggesting that frogs are opportunistic feeders and forage on abundant prey items. Eleutherodactylus coqui management should focus on areas with endemic invertebrates of concern because it is these locations where E. coqui may have the greatest impact.