Experiments On Mimicry: I. the Response of Wild Birds To Artificial Prey

Artificial prey, consisting of a piece of pastry placed on a coloured card, were placed randomly on an area of suburban lawn. Wild birds, mostly Starlings, were allowed to eat them. The birds were first conditioned to accept an edible green control and were then presented with equal numbers of controls and a red model, which was made distasteful by soaking the pastry in quinine hydrochloride. The birds rapidly became conditioned to eating the controls and leaving the models; this fact, and their observed reactions, show that they found the models unpalatable. The birds were then presented with controls, models and edible mimics in the ratio I : 2 : I. There were four kinds of mimic, one perfectly resembling the model, the others differing in a change of colour (yellow), or the addition of a black bar, or both. One type of mimic was offered in each trial, and the type changed according to a randomised block design. Perfect mimics were almost completely protected from predation. An analysis of the variance in the predation of the four types of mimic show that colour is of great importance in determining the rate of predation (yellow mimics being eaten more than red), that the black bar is of lesser importance, and that the birds' reaction to both these changed with time. The analysis is thought to indicate, but not prove, that the birds 'generalise' (desist somewhat from eating even the poor mimics). The red mimic with the black bar was eaten significantly more than the perfect red mimic, showing that the birds eat a slightly imperfect mimic more often than a perfect one. Wild birds thus behave in a way which could maintain Batesian mimicry and promote its gradual evolution in their natural prey.