Print PageThe nonlethal effects of predators on prey phenotype and performance can propagate through foodwebs to affect the outcome of predator-prey interactions (trait mediated interactions; TMIs). To produce effective defensive phenotypes requires environmental cues that indicate both presence of predators and the level of risk that they impose. In aquatic organisms, chemical cues from predation events on conspecifics are key indicators of predation risk. Studies aimed at quantifying predator induced defenses have typically controlled for variation in chemical cue concentration by feeding predators a fixed biomass of prey. However, per capita risk of predation for prey is determined by the consumption rate of the predator and the density of prey in the environment, but it is unclear if biomass of prey eaten provides information about per capita risk. The objectives of this study were to determine the relative effects of controlling for the biomass versus the number of conspecific prey consumed by Anax amazili dragonfly nymph predators on the phenotypic response of red eyed tree frog tadpoles (Agalychins callidryas). To address these objective we ran short term mesocosm experiments and compared the phenotypic responses of tadpoles to chemical cues generated from predators fed different biomasses and densities of prey.
Results/Conclusions
We show that red-eyed tree frog tadpoles express graded phenotypic responses to variation in predation threat; however, increasing the numbers of prey consumed had a stronger effect on the magnitude of phenotypic response of target tadpoles than did increasing the biomass of prey consumed. This result has important implications for how we evaluate studies of risk assessment in anuran tadpoles and for how future experiments are designed. Typically, investigators attempt to minimize variation in cue concentration during an experiment by feeding predators a constant biomass of prey throughout the experiment. However, as indicated in our experiment, if the average size of prey varies then controlling the biomass of prey eaten may not maintain a consistent signature of risk over the course of an experiment. In fact maintaining a constant biomass at the expense of generating decreasing or variable densities of prey consumed to generate chemical cues can lead to unintended systematic variation in cue concentrations that are compounded through time. Indeed, such variation could be responsible for variation among studies and inhibit attempts to synthesize or generalize patterns across studies and systems.
