The ‘peck/no-peck decision-maker’ in the black-headed gull chick

Abstract

1. 1. In companion papers a model called the Choice Threshold Model was described, and its applicability to percentage preference data tested. The present paper extends the treatment to absolute numbers of responses in a choice situation. 2. 2. It is postulated that while the Choice Threshold Model functions as a ‘directing’ decision-maker, determining how choices are to be directed, the determination of when and how often such choices shall occur may be the function of a separate ‘go/no-go decision-making mechanism’, the rate of whose activity is represented by the symbol K. 3. 3. Data on rates of pecking at coloured stimuli by black-headed gull chicks Larus ridibundus, are used to determine the relationship between K and the stimulus situation. 4. 4. It is found that K is determined only by the most preferred stimulus available. 5. 5. Thus a stimulus affects the animal in two different ways. By influencing the ‘directing decision-maker’ (the Choice Threshold Model), it affects the relative number of times it is chosen compared with other simultaneously available stimuli. By influencing the ‘go/no-go decision-maker’, which it only does when it is the most preferred of the available stimuli, it affects the absolute rate of choosing regardless of which stimulus is chosen. This dichotomy is reminiscent of that of classical ethology between ‘directing’ and ‘releasing’ roles of stimuli. 6. 6. Paradoxically it is found that stimuli which are preferred with respect to their ‘directing’ function tend to have low ‘releasing’ value and vice versa. 7. 7. A quantitatively precise prediction concerning the relative merits of two methods of studying selective responsiveness, simultaneous or successive presentation of stimuli, can be derived from the model if it is assumed that K is constant. In the light of the findings about the dependence of K on the stimulus situation, this prediction has to be modified. The present data are used in an indirect way to verify the modified prediction. 8. 8. The actual values of K (i.e. ‘releasing’ value) for each colour are estimated from the data, and are plotted graphically against the corresponding threshold values (i.e. ‘directing’ value). The inverse relationship appears to be a regular one, for the points all fall on a smooth curve. 9. 9. These findings can all be explained by a Lorenzian hydraulic type of model. Various versions of the model have been simulated, and the parameters of one version adjusted to achieve a precise fit to the gull data.