Karen Warkentin knew she'd have some interesting questions to answer when she saw a hungry snake attacking a clutch of near-term red-eyed treefrog eggs in the lab. As the predator started tucking in to the eggs, tiny frog embryos began tumbling from the clutch, even though they should have waited another 2 days before hatching. Warkentin eventually discovered that the tadpoles were making a tough decision: to escape the snake by fleeing to the water, even though they are much more vulnerable to aquatic predators at such an early age. Intrigued by the youngster's decision, Warkentin was curious to find out which cues had triggered their evacuation. Warkentin began to suspect that vibrations, generated by the snake's assault, prompted the treefrog's bid for freedom, but why didn't other less sinister vibrations send the youngsters tumbling free too? Curious to know how the embryos distinguished a life-threatening attack from vibrations caused by rain or rustling leaves,Warkentin and her student Michael Caldwell decided to see what makes vibration sequences scary for red-eyed treefrog embryos(p. 1376).

Travelling to the Smithsonian Tropical Research Institute at Gamboa,Panama, Warkentin and Caldwell collected frogspawn from trees growing over a local pond. Back in the lab, the team waited until the eggs were 5 days old before attaching a vibrating probe to the clutch to shake the embryos up. Teaming up with Gregory McDaniel, a vibrations engineer, Warkentin designed 32 white noise vibration patterns, with bursts of vibration ranging from 0.5–20 s interspersed with gaps ranging from 0.5–100 s. Exposing egg clutches to the vibrations, the team recorded how many embryos were scared enough to hatch during the following 10 minutes.

Analysing the embryo's escapology, Warkentin realised that the frogs weren't responding to the percentage of time filled with vibration or the length of the time cycle that the pattern repeated over. However, the vibration duration and gap between vibration bursts had a profound effect on the embryo's desire to hatch; 0.5 s bursts combined with 1.5-2.5 s gaps were very scary, with three quarters of the embryos deciding to take their chances in the water, but combining a scary 0.5 s burst with a lengthy gap wasn't at all scary. `Vibration duration and interval appear to function as two necessary elements of a composite cue' says Warkentin. The team also realised that the embryos sometimes waited for up to a minute after the vibrations started before beginning to hatch. Warkentin explains that the treefrog eggs are secured with jelly and so are quite tough for the snake to tear loose,giving the embryos enough time to sample several vibration cycles before making their life or death decision.

So how do the embryos sense these seismic events? Warkentin isn't sure. She explains that it is possible that the embryo's lateral line neuromasts pick up the vibrations, or that the embryos simply sense the signal by sloshing around within their capsules. But that is one of the unresolved questions that will keep her returning to Panama for years to come.

Warkentin, K. M., Caldwell, M. S. and McDaniel, J. G.(
). Temporal pattern cues in vibrational risk assessment by embryos of the red-eyed treefrog, Agalychnis callidryas.
J. Exp. Biol.