Human Vocal Commands Verify Audio Discrimination Ability in the Steller Sea Lion Eumetopias jubatus

Abstract

We report results of experiments using human vocal commands to investigate how well a Steller sea lion (Eumetopias jubatus) can discriminate different sounds. The participant, Hama, a 12-year-old female at Kinosaki Marine World, Japan, was raised by four trainers. We investigate Hama’s ability to discriminate 10 vocal commands (e.g. Iya-iya, Okay etc.) used in daily performances in two experiments: 1, commands given directly to the sea lion by trainers, but out of her sight; and 2, vocal commands given to the sea lion by trainers, but using recorded commands over a loudspeaker. In the experiment 1, Hama was able to discriminate all 10 commands, but her accuracy to discriminate one command was significantly lower than others (p < .05). In the experiment 2, Hama was able to discriminate all but one command; accuracy rates among commands were variable, but the accuracy rate for one was significantly lower than the others, and those for three commands were significantly higher than others(p < .05). These experiments demonstrate that Hama can discriminate most commands based on their acoustic properties, regardless of which of four trainers spoke them out of her sight or if they were relayed from a recording. This ability suggest that the Steller sea lion E. jubatus might use sound for their own complex communication.