One-Trial Versus Multi-Trial Learning for a Predator Encountering a Model- Mimic System

Abstract

A stochastic model of the strategy of a predator faced with a model-mimic prey system is developed and solved. This model is closely related to one developed by Estabrook and Jespersen (1974) except that it embodies a multi-trial learning process rather than a single-trial learning assumption. Comparison of the 2 models reveals that a single-trial strategy is always superior (for the predator) to a multi-trial strategy, thus providing theoretical justification for the use of the single-trial learning assumption. In addition, it is shown that for either model the prey can, by adjusting their spatial distributions but not their absolute numbers, force a well-adapted predator to ignore them, provided the model is sufficiently unpalatable.