Does the type of task affect prey discrimination learning in avian predators?

Abstract

When learning to discriminate between palatable and unpalatable prey, predators may encounter the prey either simultaneously or in a sequence. The two types of discrimination tasks, which are frequently used in experiments focused on aposematism and mimicry, may considerably differ in difficulty, as they differ in the information available to predators during individual encounters with prey. Surprisingly, there are very few studies directly comparing the performance of predators between sequential and simultaneous discrimination tasks using otherwise identical experimental design, the same predators and prey. We tested the effect of the type of discrimination task on the effectiveness of discrimination learning in adult and juvenile great tits (Parus major). Birds were trained to discriminate between palatable and unpalatable artificial prey items baited with mealworms soaked either in water or quinine. Prey shape, colours and patterns were derived from real shield bug species. We compared the performance of birds tested with two discriminative cues, colour and pattern, using three discrimination tasks: (1) sequential—alternating presentation of one palatable and one unpalatable prey item, (2) two-choice—simultaneous presentation of one palatable and one unpalatable prey item and (3) multiple-choice—simultaneous presentation of multiple palatable and unpalatable prey items. Colour was a more effective discriminative cue than pattern for both adult and juvenile birds. Adults performed better than juveniles regardless of task. The birds performed equally well in the sequential and two-choice tasks, but their performance in the multiple-choice task was worse than in the other two tasks. However, these differences were only present when the birds used the pattern as a discriminative cue. The birds tested with colour, a more salient cue, performed equally well in all three tasks. Type of discrimination task may therefore affect the learning performance of predators, but the effect also depends on the saliency of a particular discriminative cue.