Consumptive effects are stronger but more variable than non-consumptive effects in predator–prey systems

Andrew Kramer and Marc J. Lajeunesse, Department of Integrative Biology, University of South Florida, Tampa, FL, Travis Flock, Integrative Biology, University of South Florida, Tampa, FL

Background/Question/Methods
Predators reduce prey fitness through consumption and by inducing trait changes. Studies have identified a variety of non-consumptive effects of fear and found morphological and behavioral responses to predators can have substantial effects on fitness. In fact, previous meta-analyses of these studies found that prey fitness was as strongly affected by non-consumptive effects as consumptive effects. This surprising synthesis led to increased focus on the role that fear plays in predator–prey dynamics. Given the subsequent proliferation of research, we revisited the relative fitness cost of traits involved in non-consumptive effects and the cost of consumption, and asked whether the magnitude of effects caused by fear of predators differs among systems. We surveyed the literature and used a trivariate meta-analysis to jointly synthesize and model the multivariate effects of consumptive and non-consumptive effects on prey. We distinguished between studies that measured “true” consumptive effects, in which prey could not adjust behavior or morphology, and the more common case where the predator effect necessarily integrated consumptive and non-consumptive effects. Each study was classified with the type of non-consumptive effect, the measure of fitness used, and the class and habitat of the prey species.
Results/Conclusions
We identified extracted 1126 effects sizes from 182 studies with appropriate experimental design and data availability. We found that consumptive effects have twice the negative effects on prey and that consumptive effects were more variable. Non-consumptive effects were strongest in aquatic systems and were not necessarily stronger in pairings with strong consumptive effects. Further, we found that total predator effects were non-additive and, in fact, were less than consumptive effects—which is biological intuitive, otherwise trait modifications due to fear would not improve fitness. This suggests estimating the total predator effect requires measuring the combination of effects and increasing focus on how much non-consumptive effects reduce consumption by predators.