Most species have complex life cycles in which they are vulnerable to different suites of predators as they shift habitats through ontogeny. While these predators, separated in time and space, do not interact directly, they may still interact indirectly via their effects on the density and traits of their shared prey. Further, whether the effects of early predators alter subsequent predator-prey interactions may depend upon predator identity and environmental context. Here we examine sequential predator effects on the leaf-breeding treefrog, Agalychnis callidryas. Arboreal egg predators reduce initial tadpole density and also induce embryos to hatch earlier at smaller sizes. We explore how these egg predator effects play out given functionally distinct aquatic predator species and across environments that differ in resource availability. First, we predict the effects of sequential predators on larval mortality and growth using simulations, parameterized from short term experiments, which quantify tadpole size- and density-specific growth at different resource levels and mortality in the presence of two different size-limited predators, dragonfly nymphs and giant water bugs. We then compared these predicted results with larval mortality and growth observed in mesocosm experiments where we manipulated hatchling size and density, and aquatic predators and resources.

Results/Conclusions

In the first mesocosm experiment we manipulated hatching size and initial density and the presence of giant water bugs and dragonfly nymphs and measured survival and growth over 28 days. Predators reduced survival and increased growth in a manner consistent with model predictions. Hatchling size and initial density had no effect on survival, although these were predicted. Further, in the experiment early hatched tadpoles grew faster in the presence of predators, a result that was not predicted by the simulation. In the second experiment we manipulated hatching size, resource availability and the presence of dragonfly nymphs and measured survival and growth over 30 days. Dragonflies reduced survival and dragonflies and high resources increased growth in a manner consistent with predictions. Hatching early reduced survival, irrespective of predator environment, a result not predicted by the simulation. Our results highlight that by developing a functional understanding of the relationships between prey size and density and their effects on mortality and growth we can develop a predictive framework for understanding sequential predator effects. Further, such a framework can reveal when additional processes, beyond size- and density-specific mortality and growth, are needed to understand predator-prey interactions.