Acoustic Properties of Humpback Whale Songs Sound production and detection are critical to the lives of marine organisms. Organisms including mysticetes or baleen whales use sound as their primary means of learning about their environment and also as means of communicating, navigating and foraging for food (NRC, 2003). Consequently, there is a growing concern that sounds produced by humans (anthropogenic sounds) may interfere with these essential activities by polluting the waters with sounds that could potentially injure the hearing organs of these mammals. In a report issued by The National Research Council in 2003, the council identified a number of critical objectives for future research on ocean noise and marine mammals. One of these recommendations urged research teams to "describe the distribution and characteristics of sounds generated by marine mammals & #8230;within behavioral contexts" (NRC, 2003, 128). Au et al. (2006) sought to achieve this objective in a study that examined the acoustic properties of humpback whales in order to determine potential sound exposure of nearby whales. By studying the physics and physiology of sound production in this species of baleen whales, the authors developed insight into the amount of anthropogenic sound humpback whales typically tolerate.

The structural components of whale songs have been studied extensively. A whale song has been defined as a cyclical pattern of themes made of a distinct repetition of phrases (Payne, Tyack & Payne, 1983). These phrases consist of short song units that can be classified into two groups, those of high-frequency harmonics and some tonal quality and those that are broadband with grunts or gurgle-like qualities (Au et al., 2006). The authors of the study were particularly interested in the different source levels of these isolated units to determine the intensity limit of sound production. The authors measured a maximum source level of 173 dB. This measurement was based on sounds received by a vertical panel of 5 hydrophones emerged at seven meter intervals. Using an eight-channel Teac TASCAM (Model DA-78HR), the song was recorded and analyzed by a root mean square equation for source level. The distance between the whale and the hydrophone panel was determined by taking time of arrival differences of the signals between the center hydrophone and the two neighboring hydrophones (Au et al., 2006).

This conclusion has important implications because the authors could use it to estimate the level of sounds (sound intensity) that other whales are exposed to in the presence of a singer. Previous studies have cited a number of interactive behaviors between singers and other whales. Darling and Berube (2001) report singers stopping their song to join another group, singers with adult companions and with a cow and calf. It, therefore, appears reasonable to assume that singers consistently sing in the presence of other whales, as close as two whale lengths away (Darling & Berube, 2001). To estimate the source level that whales are exposed to, the authors assume a whale length of twelve to fourteen meters. Taking into account the distance between whales and sound conduction in water, the authors conclude that a whale may be exposed to sounds having a root mean square source level of 147 dB. This conclusion presumes that whales do not reduce the sound intensity as they approach another whale.

The authors further conclude that these higher source level units are less frequent in the whale song themes. The most intense units were longer broadband units that were described as grunts and gurgles (Au et al., 2006). Unit A, for instance, had the highest average source level but was recorded much less frequently in the themes than the short, low-frequency downsweep unit F. More generally, only four out of the twenty-seven recorded units were above 170 dB. While companion whales may be exposed to high source level notes, it remains unclear what phrases are used for what occasion and whether a singer will use phrases and themes with less high source level notes in the presence of another whale. The authors make their conclusion about the sound exposure of companion whales under the assumption that singer whales do not alter their singing behavior in the presence of another whale.

The author further examined the hearing ability of humpback whales. By the same logic applied to their first research objective, the authors reasoned that a determination of the high-frequency harmonics in a signal might help them predict the upper frequency of hearing in humpback whales. Based on the harmonic structure of the spectrogram, and the general assumption that animals hear the totality...

The authors provide ample evidence for their frequency calculations and cite that the use of new portable digital recorders generally have longer bandwidth than older cassette tape recorders. While old recorders were unable to capture such high frequency units, the new technology used by the authors allow them to capture these new units and significantly add to the understanding of the acoustic properties of humpback whale songs. The three units that show frequencies exceeding 24 kHz do not feature frequently in the common song themes.
The authors also make novel conclusions about the physiology of sound production in humpback whales. Based on the direction of the higher frequency harmonics, the beam of the sound appears to be directed slightly above the horizon. Given that whales often suspend themselves with the longitudinal axis of their bodies between 0 and 75 degrees to the vertical, this evidence implies that the sound generator is producing sounds at a steep angle above the whale's head (Au et al., 2006). The evidence presented in the study is tentative and inconclusive, however. According to the authors, the exact position of the whales with respect to the vertical and therefore the angle above the whale's head was not measured. The authors found that sound fields began to show directivity in the high-frequency sound units. The location of the sound generator may have significant implications for sound propagation through water because the orientation of the whale affects how sound waves propagate in the horizontal direction.

This finding bears significantly on the larger topic of whale sound production and behavior of whale singers towards other whales. Based on the directivity of sound propagation, it appears that the orientation of the whale towards its companions may influence the frequency of the harmonic and the average sound level that whales are exposed to.

This study makes significant contributions to the knowledge of the acoustic properties of humpback whales. The highest source level measured in the study was slightly lower than the 175 dB and 190 dB measured by Thompson, Cummings and Ha (1986), for a moan and grunt respectively. The study, therefore tentatively corroborates the earlier findings. However, Au and colleagues conduct a more comprehensive analysis of the source levels of individual song units. Each unit in a song theme is associated with a range of source levels. Unit A, for instance has a large range of source levels, peaking in whale E4 at 173 dB and reaching an intensity trough of 164 dB. While it does not feature often in the song themes, it is a versatile sound unit used frequently at a high intensity source level. While little is know about the meaning of the sounds, this information may serve as a precursor to more detailed covariance analysis with whale orientation and physiological behavior.

According to a study conducted by Richardson and colleagues, the source levels of anthropogenic sounds commonly exceed the limit of any sound produced by humpback whales (Richardson, Greene, Malme & Thomson, 1995). Supply ships generate source levels of 181 dB, larger tankers of approximately 186 dB, and icebreakers of up to 193 dB. While these sound intensities were all measured at a one-meter distance and sound intensity decreases with distance from the source, the average exposure would still be extreme. Au et al., calculated that sound intensities decrease by approximately 28 dB over 12 to 14 meters. Humpback whales were commonly found between 15 and 25 meters beneath the surface of the ocean and would be within range of the anthropogenic sound. A humpback whale would thus be exposed to an approximately 145 dB source level. While this is not as intense as the 173 dB maximum source level recorded in this study, it helps to put this finding into an anthropomorphic sense. Even if a person is shouting only close to maximum strength, a listener is likely to suffer hearing damage over an extended period of exposure.

Further, frequencies of sounds generated by ships and industrial activity are within the range of sound frequencies generated by humpback whales. Based on the authors' assumption that animals can hear the same frequency sounds that they produce, the anthropogenic sounds most certainly interfere with the sounds produced by humpback whales. The findings of this study confirm the concerns that scientists presented in the National Research Council on the effect of anthropogenic sounds and could be used to argue for stricter guidelines for tanker travel.…