Functional morphology of the humpback whale (Megaptera novaeangliae) sound production system: anatomy and modeling

Supervisors: Paul White (Southampton University) & Olivier Adam (Université Paris Sud, France) Stays: Icahn School of Medicine-Mount Sinai, NY & CETAMADA, Port Barachois, Ambodifotatra, Sainte Marie, Madagascar
Male humpback whales (Megaptera novaeangliae) are the most acoustically active individuals among mysticetes. While odoncete sound production mechanism is quite well understood, and despite the recent publications by Reidenberg and Laitman, presenting the U-folds as a homologue to terrestrial mammals vocal folds, the way mysticetes are able to sing or produce social calls with their laryngeal apparatus is still unknown. Anatomical observations and acoustic signal processing have led to different hypotheses under the framework of a production-based approach. This approach has been developed in order to understand the functional morphology for sound production and to identify the factors constraining the sound output for mysticetes. In this thesis, we broaden the anatomical data acquired by Reidenberg and Laitman on larynx apparatus for 14 specimens of diferent rorqual species: sei whale (Baleanoptera borealis), fin whale (Baleanoptera physalus), minke whale (Baleanoptera acutorostrata) and humpback whale (Megaptera novaeangliae). Based on this data and on previous literature, we propose, in this work, three different functional positions (rest, breathing and rebreathing), an unidirectional regressive sound production (air flowing between the lungs and the laryngeal sac) and a new sectioning nomenclature for the different part of the U-folds: the distal section, the midsection and the corniculate flaps. Each of these sections has specific morphological and acoustical properties leading to the concept of ‘mode variation’ in the case of rorquals. Histology cuts have been sampled for the 3 parts of the folds of a fin whale specimen and the analyse for a multi-layer structure of the folds has now to be performed. Based on these new observations, we developed: (1) a project for a silicone single layer mechanical model to test the hypothesis for phonatory positions, the stiffness variation of the soft tissues and the glottal pulse shape of the mid- and distal sections, and (2) a preliminary computational model of the resonant cavities (laryngeal sac and nasal cavities) filtering the sound source according to their morphological and dynamic characteristics. The output for the impedance spectrum in the case of laryngeal sac inflation is compared to real recordings. Through the fieldwork of the author, methods for humpback whale monitoring are presented.