Love in the litter: The effect of leaf litter species on amphibian oviposition site selection

Background/Question/Methods   Oviposition site selection strongly influences the fitness of offspring and the composition of communities. In forests, tree leaf litter serves as the dominant source of nutrients and energy input to aquatic ecosystems. Furthermore, mesocosm experiments strongly suggest that inputs of different leaf litter species highly impact the survival and fitness of larval amphibians in these habitats owing to differences in litter chemistry. We asked whether the presence of leaf litter from three native and dominant tree species (maple, Acer spp.; oak, Quercus spp.; or eastern hemlock, Tsuga canandensis) could affect amphibian oviposition site selection. These litter species vary greatly in chemistry and previous work has demonstrated many strong differences in abiotic conditions induced by each litter species. Prior work with gray tree frog (H. versicolor) tadpoles, which oviposit in ephemeral habitats immediately adjacent to forests, suggests that growth rate and size at metamorphosis are greater in ponds containing maple and hemlock, but lower in ponds containing oak. Thus, we predicted that H. versicolor would oviposit more eggs in ponds containing maple or hemlock litter than in ponds containing oak litter. To test this, we distributed aquatic mesocosms containing one of the three litter species along the edge of a forest. The mesocosms were left uncovered throughout the breeding period, checked three times each week for eggs, and removed all oviposited eggs.

Results/Conclusions   Oviposition events occurred multiple times during the breeding period and there was a significant effect of litter on number of eggs oviposited in each mesocosm. H. versicolor oviposited the most eggs in maple replicates, which received an average of 150% more eggs than oak replicates and an average of 330% more eggs than hemlock replicates. This partially agreed with our predictions, but the greater oviposition rate in oak litter treatments relative to hemlock treatments was not expected. This may be explained by lack of hemlock trees near the experimental site and consequent lack of familiarity with this litter type, although the role of familiarity in oviposition site selection has yet to be investigated. Overall, these results point to a potential mechanism driving species distribution within and across landscapes.