Aerial Visual Shape Discrimination and Matching-To-Sample Problem Solving Ability of an Atlantic Bottlenose Dolphin

摘要

Using a simultaneous matching-to-sample paradigm, two-dimensional geometric shapes of various configurations were presented to a bottlenose dolphin in an aerial visual discrimination task. Large differences in perimeter lengths between any two types of shapes presented together generally resulted in better discrimination performance. Alternatively, small differences in perimeter lengths of any two shape types presented together generally resulted in poor discrimination performances. The absence of a monotonic function between perimeter differences and performance accuracy indicated that other form parameters may be contributing factors in the discrimination process. Data were collected to determine if position biases influenced response behavior. The results were negative. Stimulus preference was also shown to have little effect on choice-making by the dolphin. Complete changes of the discriminanda were made to assess conceptual learning by the dolphin, and the results indicated that transfer did not occur between sets of stimuli when subsequent sets of geometric shapes differed greatly from the previously learned set. Comparison of baseline data with a reacquisition curve of the same problem presented 6 months later clearly indicated a function typical of problem-specific rather than conceptual learning.