Does presentation format influence visual size discrimination in tufted capuchin monkeys (Sapajus spp.)?

Most experimental paradigms to study visual cognition in humans and non-human species are based on discrimination tasks involving the choice between two or more visual stimuli. To this end, different types of stimuli and procedures for stimuli presentation are used, thus stressing the necessity to compare data obtained with different methods. The present study assessed if, and to what extent, capuchin monkeys' ability to solve a size discrimination problem is influenced by the type of procedure used to present the problem. Capuchins' ability to generalise knowledge across different tasks was also evaluated. We trained eight adult tufted capuchin monkeys to select the larger of two stimuli of the same shape and different size by using pairs of food items (Experiment 1), computer images (Experiment 1) and objects (Experiment 2). Our results indicated that monkeys reached the learning criterion faster with food stimuli compared to both images and objects. They also employed consistently less trials with objects than with images. Moreover, female capuchins had higher levels of accuracy to acquisition with food stimuli than with images. Finally, capuchins did not immediately transfer the solution of the problem acquired in one task condition to the other conditions. Overall, these findings suggest that - even in relatively simple visual discrimination problems where a single perceptual dimension (i.e., size) has to be judged - learning speed strongly depend on the mode of presentation.