Tadpole antipredator responses change over time: what is the role of learning and generalization?

摘要

Many prey animals quickly learn the identity and risk level associated with unknown predators and this provides them with an immediate survival advantage. However, as time passes the information they learned about the risk level associated with the predator decreases in value. As the certainty of the information decreases, there should be a point in time when the prey should stop responding to the information. Here, we conditioned tadpoles to recognize a tiger salamander, fire belly newt, or goldfish and found that they responded to each of the 3 predators with an equal response both 1 day and 8 days postconditioning. Subsequently, we reconditioned each of the groups to recognize tiger salamanders alone and found that the duration of time for which the tadpoles responded to the tiger salamander cue was influenced by what the tadpoles learned in the past. Tadpoles conditioned twice to the tiger salamander retained their response the longest, whereas tadpoles taught goldfish and then tiger salamander responded to the tiger salamander for the least time. Those learning the newt and then the tiger salamander retained their response to the tiger salamander for an intermediate amount of time, indicating that information gained through predator generalization influences the retention of responses to predators. Our results highlight the complex algorithm used by animals to acquire, encode, and use information from their environment.