Visual search in barn owls: from feature to conjunction search

Visual search is the process of searching for something interesting in a cluttered environment. It is well studied in humans, but not in non-primate species. This study provides a comprehensive overview of visual search in barn owls using a novel methodology: the OwlCam. The OwlCam is a tiny, lightweight camera that can be mounted onto the head of barn owls to record first person videos from these birds. Due to the very limited eye movements of barn owls this provides an easy method of gaze tracking with only one camera in these birds which is not possible in most other species. Six experiments to investigate overt visual search processes in barn owls are described. Experiment 1 investigates basic overt attentional processes in barn owls. The question was, whether barn owls are attracted to the same conspicuous stimuli as humans? This was shown by confronting barn owls with arrays of differently oriented stimuli containing one odd target among similar distracters. Barn owls were attracted to the odd stimulus, as humans are. Experiments 2 and 3 prove that barn owls have a pop-out effect. Pop-out is the ability of visual systems to detect and localize certain visual targets equally efficiently regardless of the complexity of visual distraction. This was tested by training the owls to search for a unique target in feature-search arrays. These feature searches contained several quantities of uniform distracters and one odd target, different from the distracters in either orientation (experiment 2) or intensity (experiment 3). The search time and number of saccades until the target was found did not change at different set sizes, showing that barn owls have a pop-out effect. Experiment 4 was similar to experiment 3 but with inverted target and distracters to test whether barn owls have search asymmetries. That may be the case, if a search task’s difficulty changes when the target and distracters are inverted. The data showed that this was the case. Finally, experiment 5 and 6 were conjunction-search tasks. These are difficult search tasks, in which the target is a unique combination of two features, orientation and intensity, and the distracters are the other feature combinations. The owls could learn to do these tasks, and indeed it took them much longer to find the targets than in the simples feature searches, and search time depended on array size. Thus, this study shows that barn owls match human performance in several visual searches. Therefore, a pop-out effect or complex visual searches do not require the elaborate cortical structure found in mammalian brains. Furthermore, this study shows that barn owls wearing OwlCams could be a model system to study visual search and overt attention in non-primate species