absence of observational learning, it is difficult to conceive how Mertesian mimicry might evolve.
Broadly, the present experiments suggest that a knowledge of stimulus effectiveness in laboratory studies of direct and vicarious avoidance learning will further understanding of aposematic signals used in model-mimic systems. Provided that laboratory experiments of visually mediated avoidance learning fit within a broader ecological context, one might predict that the stronger the avoidance response (e.g., the greater the resistance to extinction) in the laboratory, the more effective (and, perhaps, the more common) the stimulus complex in nature. This prediction is consistent with the observation that aposematic animals use both color and pattern cues to deter potential predators ( Cott, 1940). Strong avoidance acquisition after experience with a model prey (or after observation of prey sampling by a conspecific) could deter even brief sampling excursions by birds, thus decreasing the likelihood of encountering a mimic (and weakening the learned association). Strong acquisition might also permit the number of mimics to exceed the number of models without damage to the mimetic complex ( Brower, 1960).
Learned avoidance and preference by birds for some foods is mediated by visual cues. Learning can occur either through direct experience or observation. For avoidance, aposematic colors such as red are more effective conditioned stimuli than cryptic colors such as green. For preference, both red and green appear to be equally effective. Ecologically, the differential effectiveness of color is predictable, because the aposematic colors are used frequently by animals to advertise unpalatability, whereas the cryptic colors serve this function rarely. An unanswered question is the importance of pattern cues as a variable influencing direct and observationally acquired preference and avoidance. The potential significance of striping is suggested by the observation that aposematic animals often are striped as well as brightly colored. The present experiments were designed to address this issue using redwinged blackbirds as subjects. Experiments 1 and 2 assessed whether complex stimuli (pattern and color cues) elicited greater resistance to extinction than simple stimuli (pattern or color cues) in avoidance learning. Experiment 3 investigated whether complex stimuli might facilitate resistance to extinction of color preference responding. Experiments 4 and 5 assessed whether complex stimuli would enhance learned avoidance of visual stimuli associated with merely unpalatable food. Both direct and vicariously acquired avoidance were more resistant to extinction when complex stimuli were employed. Resistance was enhanced regardless of whether the unconditioned stimulus was malaise or unpalatability. Conversely, resistance to extinction of color preferences was not affected by the use of complex stimuli. These results