Search

Here I summarize the most important findings of my PhD thesis. My dissertation consists of 9 chapters: An introduction, 3 peer-reviewed papers (2 published, 1 accepted), 4 manuscripts prepared for submission, and a progress report of an ongoing study. The introduction serves as a broad background and review for the topics addressed in the subsequent chapters, with discussions of these chapters where appropriate. In this thesis, I have undertaken a series of acoustic studies on two species of toothed whales, the bottlenose dolphin and the short-finned pilot whale. The bottlenose dolphin (Tursiops sp.) is one of the best known toothed whales due to studies in captivity over the last 50 years. In contrast, the short-finned pilot whale (Globicephala macrorhynchus) is a larger, deep-diving toothed whale that has been studied rather little, in part because their deep-diving ecology regularly takes them out of sight of surface observers. These species differ in the acoustic habitats they dwell in, as well as in group structure and foraging ecology. The overall aim of this thesis has been to address, in a comparative fashion, how these two species behave acoustically in the wild, and how they have adapted their vocal behaviour and sound production to their different ecological niches and habitats.Toothed whales find and capture prey using a sophisticated biosonar system. Little is known about how toothed whales use their biosonar during a complex three-dimensional task of locating and capturing prey in the wild. To alleviate this lack of knowledge, my collaborators and I investigated the echolocation behaviour of bottlenose dolphins in West Australia using calibrated hydrophone arrays. We found that the echolocation clicks used by these wild dolphins were slightly more directional but had lower source levels than clicks from trained bottlenose dolphins doing target detection tasks in net pens, and much higher source levels than dolphins in concrete pools. This adaptive sonar behaviour of toothed whales illustrates the need to record sonar parameters from dolphins in the wild rather than extrapolating from captive studies. I also found that wild dolphins actively lower their click source levels and click intervals when they approach a target. However, click intervals were stable outside a range of some 10m, indicating that the apparent source level adjustment to target range outside this range may be an active, cognitive process rather than a biophysical consequence of faster clicking rates. I carried out similar studies on the larger short-finned pilot whales using the same array deployed at the surface. My results here appeared to reveal similar source levels than found for the bottlenose dolphins despite the open habitat and larger size of the animals. However, my later investigations of data from digital acoustic tags (DTAGs: Woods Hole Oceanographic Institution) reveal that the source levels measured at the surface are much lower than those used when searching for prey several hundred meters below the surface, underscoring the need for measuring biosonar parameters in the habitat where they are actually used for foraging. Concurrent with investigations on biosonar properties, I investigated the acoustic communication signals of both species. Toothed whales communicate using several types of acoustic signals including narrow-band frequency modulated whistles and rapid series of clicks and burst pulse calls. Using a GPS linked array of receivers, I measured source levels and energy content of bottlenose dolphin whistles in a tropical, shallow habitat with high noise levels. I find that the source levels of whistles are lower than previously measured, possibly restricted by the size of the animals there. I estimated the detection range of whistles to be 5x lower than estimated for a North-Atlantic bottlenose dolphin population in a quieter habitat. I also found that the stimated metabolic cost of producing these whistles is insignificant compared to the high metabolic rate of these marine mammals, indicating that communication for bottlenose dolphins – and likely all toothed whales – is energetically cheap in terms of direct costs. Using acoustic tags (DTAGs), I find that pilot whales, on the other hand, use signals that can be detected at longer distances, primarily due to lower background noise levels in the more open habitat but also because they seem to be able to produce calls at higher amplitudes. However, their deep-diving ecology appears to impose special constraints on the communication of these animals. Toothed whales use a closed loop system of air in the nasal passages for pneumatic sound production. By using DTAGs to quantify the vocalizations of diving animals at a known depth, I demonstrate that when pilot whales descend towards foraging depths, the hydrostatic pressure negatively affects the production of communication signals. This results in lower amplitude, shorter calls at depth despite an increased distance to their social group at the surface. I show that calling ceases during the deepest part of the foraging dives, but resumes during the ascent phase, allowing foragers to re-establish acoustic contact with their social group during the ascent. These biophysical limitations further suggest that toothed whales inhaling before a dive, in sharp contrast to all pinnipeds that exhale before diving, may do so to increase the amount of air available for sound production and extend the range of depths they can cover while producing sounds pneumatically. Surprisingly, I find that the frequency content of pilot whale calls, including the time/frequency modulation patterns that seem to convey information for some toothed whales, is unaffected by depth despite the compression of air cavities inside the head of the animal. This led me to participate in an investigation of the sound production of trained bottlenose dolphins vocalizing in a mixture of oxygen and helium that alters the resonance frequency of air cavities. Using a novel signal processing technique to isolate the energy within individual harmonics of a frequency-modulated signal, we find that the fundamental frequency of whistles is unaffected by the increased resonance frequency in internal air sacs, but that energy is moved into higher harmonics. This is similar to humans speaking in a helium mixture but drastically different from a human whistling in a helium mixture. These results show that toothed whales do not actually whistle per se, but produce tonal signals by phonic lip vibrations set in motion by a controlled air flow, and where internal air spaces affect the distribution of energy in harmonics. As a consequence, the frequency contours that convey acoustic signatures in delphinids are unaffected by changes in depth as well as changes in the size of different nasal air sacs during sound production.Finally, both bottlenose dolphins and pilot whales are the subject of heavy whale watching activities and other anthropogenic noise sources, but little effort has gone into modelling the sound exposure levels that would arise from smaller vessels. I therefore measured the noise from two small vessels and modelled the masking impacts they would have on the communication signals of the two study species. I documented significant masking levels increasing with the speed of the vessels due to greater cavitation noise, as well as very high-level transients generated from gear shifts. My results show that limiting the gear shifts and keeping speeds below 5 knots would greatly reduce the masking impact from these vessels.