Page 42 - Summer2017

Sound Production in Aquatic Mammals
Odontocetes (Toothed Whales, Including Dolphins and Porpoises)
Odontocetes generate high frequencies (ultrasound) that they use for navigation and finding prey, as well as other sounds used for communication (listen to audio recordings at http:// acousticstoday.org/twhales). Controlled experiments with hydrophones and blindfolded dolphins proved they emit clicks while navigating and finding targets (Kellogg, 1958; Norris et al., 1961). The mechanism of sound production, however, was controversial. At first, it seemed logical to assume that the larynx was the source (Purves, 1966; Purves and Pilleri, 1983) because this was the origin of sound in other ultrasonic vocalizers such as bats, rodents, and shrews. However, the complexity of the nasal region begged an explanation, and many researchers proposed that it functions in generating sound (reviewed in Cranford, 2000). Controlled experiments on live dolphins and detailed dissections (reviewed in Cranford and Amundin, 2003) proved that odontocetes use the nasal region to generate these clicks (see below).
Odontocetes also have a larynx (Figures 3B and 5), but it is not clear what sounds, if any, it can make. Although it is not the generator of echolocation ultrasounds, it too is a com- plex structure with an undetermined function. This com- plexity suggests that it may have a role in producing some sounds, most likely for communication (Reidenberg and Laitman, 1988).
The odontocete larynx can remain open while still protect- ing the respiratory tract. Its unique shape is reminiscent of a snorkel, allowing air to flow from the lungs to the na- sal region while blocking off incursions of water from the mouth (Figure 3B). The epiglottic and corniculate cartilages are elongated into a tubular shape and sealed by the tissues of the soft palate and posterior pharyngeal wall (that are formed into an encircling sphincter; Reidenberg and Lait- man, 1987). This interlock fully protects the larynx by sepa- rating the respiratory passageway (inside the lumen) from the digestive tract (that loops around it), thereby allowing both systems to function simultaneously. This higher level of protection indicates that the larynx must have an essential, and nearly constant, role for airflow (Reidenberg and Lait- man, 1994). Otherwise, simply closing the vocal folds would be sufficient to protect the larynx underwater, as likely occurs during feeding in other aquatic mammals. This level of pro- tection is even better than that seen in mysticetes (Figure 3A).
Why would odontocetes need to use the respiratory tract while underwater? Clearly, they are not breathing while submerged, so the logical answer is that airflow is necessary for sound production. There are three possible scenarios for such acoustic respiratory function. One is that air flows through the larynx and causes vibrations of laryngeal tis- sue to generate sound. Another is that air is simply chan- neled uninterrupted to the nasal region so it can be used to generate sounds there. Finally, elevating the larynx may increase pressure in the nasal cavity, driving air upward through sound-generating tissues located there. This piston- like movement was observed through video endoscopy in a live dolphin while it produced click sounds (Cranford et al., 2011). In all three scenarios, water can surround the larynx without causing drowning. Even swimming upside-down, open-mouthed, while chasing a fish would not cause drown- ing or compromise the flow of air needed to generate echo- location clicks.
The odontocete larynx contains vocal folds, but they are usually fused into a single midline fold (Reidenberg and Laitman, 1988). The vocal fold can be tensed, relaxed, and erected into the airstream where it may vibrate passively as air rushes along either side (Figure 5A), similar to how the reed of a woodwind instrument works. The vocal fold attachment points have migrated during the reshaping of the larynx. The ventral (inferior) attachment has moved ros- trally (forward) from the thyroid cartilage to the epiglottic cartilage, and the dorsal (superior) attachment has moved caudally (rearward) and ventrally with elongation of the ary- tenoid cartilage. The result is a fold rotated approximately 90° (clockwise if viewed from the left side of the animal; Fig- ure 3B) compared with that of a typical terrestrial mammal (Figure 1). This places the fold parallel to the trachea and therefore parallel to the airflow, rather than perpendicular to it as in other mammals. Curiously, vocal folds are rotated in the opposite direction in mysticetes (Figure 3A). Although mysticetes and odontocetes both have vocal folds lying par- allel to the trachea, they are oriented nearly 180° apart! This indicates an evolutionary divergence between odontocetes and mysticetes from the basic mammalian pattern.
The pathway for sound transmission from the larynx is not known. Vibrations may travel up through the skull to the forehead or rostrum, or ventrally through the throat tissues in a pattern similar to the sound conduction path proposed for mysticetes, hippos, pinnipeds, and sirenians. Interesting
40 | Acoustics Today | Summer 2017