Putting μ/g in a new light: plasticity in life history switch points reflects fine-scale adaptive responses

Life history theory predicts that organisms with complex life cycles should transition between life stages when the ratio of growth rate (g) to risk of mortality (µ) in the current stage falls below that in the subsequent stage. Empirical support for this idea has been mixed. Implicit in both theory and empirical work is that the risk of mortality in the subsequent stage is unknown. However, some embryos and larvae of both vertebrates and invertebrates assess cues of post-transition predation risk and alter the timing of hatching or metamorphosis accordingly. Furthermore, although life history switch points of prey have traditionally been treated as discrete shifts in morphology or habitat, for many organisms they are continuous transitional periods within which the timing of specific developmental and behavioral events can be plastic. We studied red-eyed treefrogs (Agalychnis callidryas), which detect predators of both larvae and metamorphs, to test if plastic changes during the process of metamorphosis could reconcile the mismatch between life history theory and empirical data and if plasticity in an earlier stage transition (hatching) would affect plasticity at a subsequent stage transition (metamorphosis). We reared tadpoles from hatching until metamorphosis in a full-factorial cross of two hatching ages (early- vs. late-hatched) and the presence or absence of free-roaming predators of larvae (giant water bugs) and metamorphs (fishing spiders). Hatching age affected the times from oviposition to tail resorption and from hatching to emergence onto land, but did not alter responses to predators or developmental stage at emergence. Tadpoles did not alter their age at emergence or tail resorption in response to larval or metamorph predators, despite the fact that predators reduced tadpole density by ~30%. However, developmental stage at emergence and time needed to complete metamorphosis in the terrestrial environment were plastic and consistent with predictions of the "minimize µ/g" framework. Our results demonstrate that likely adaptive changes in life history transitions occur at previously unappreciated timescales. Consideration of plasticity in the developmental timing of ecologically important events within metamorphosis, rather than treating it as a discrete switch point, may help to reconcile inconsistencies between empirical studies of predator effects and expectations of long-standing ecological theory.