Techniques for Studying Vocal Learning in Bottlenose Dolphins, Tursiops truncatus

Abstract

The objective of this thesis is to develop the methods necessary for evaluating the role of learning in dolphins' natural whistle development. Bottlenose dolphins provide a unique opportunity to study social influences on vocal learning in a highly social non-human mammal. Vocal learning is critical for human language development but plays a smaller role in most non-human mammals. The methods currently used to study dolphin behavior are insufficient to evaluate social influences on learning in whistle development. The techniques necessary for such a study were developed and tested. A strategy for sampling the dolphins' behavior was developed, and a pilot study performed on captive calves. Focal samples of four mothers and their newborn calves were taken, with simultaneous acoustic recordings. The interactions were analyzed with multivariate statistics to evaluate the dolphins' relationships. Programs were developed for the automatic, unbiased extraction of whistles from recordings. Methods for categorizing whistles were compared and one was shown to perform best with all whistles. These techniques were used to compare the early acoustic environments of the calves. The combination techniques developed in this thesis allows a study of vocal learning in dolphin whistle development to be performed in a quantitative, unbiased manner.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Feb 01, 1999
Accession Number
ADA369260

Entities

People

  • Deborah R. Fripp

Organizations

  • Massachusetts Institute of Technology

Tags

Communities of Interest

  • Biomedical

DTIC Thesaurus Topics

  • Birds
  • Cells
  • Computer Programs
  • Data Analysis
  • Data Science
  • Detection
  • Fur
  • Human Behavior
  • Information Science
  • Marine Mammals
  • Oceanography
  • Psychology
  • Social Environment
  • Statistical Analysis
  • Two Dimensional
  • Wildlife

Readers

  • Marine Mammal Biology
  • Neural Network Machine Learning.