Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
1 Introduction Sound is an important tool used by ocean scientists to study the topography of the seafloor and its substructure; the direction and speed of ocean currents; and the size, shape, and number of organisms in the ocean. Four fundamental proper- ties of sound transmission are important to understand as background for this report: 1. The transmission distance of sound in seawater is determined by a combi- nation of geometric spreading loss and an absorptive loss proportional to the sound frequency (see Box 1.1~. Thus, attenuation of sound increases as its ~ . frequency increases. 2. The speed of sound is proportional to the temperature of the seawater through which it passes. 3. The sound intensity decreases with distance from the sound source. Gen- erally, the decrease in sound intensity ranges between 1/r (r = distance from the source) and 1/r2 (spherical spreading), depending on characteristics of the sound source location and transmission paths, although sound intensity can decrease even more under certain conditions. Thus, a sound level may be as much as 60 dB lower than that of the source level at 1 km from the source (see Figure 1.1~. Because of the wave properties of sound and propagation conditions, waves from different sources or refracted and reflected waves from a single source can con- verge and either add to or cancel each other, so that simple geometric models of spreading do not always predict actual sound fields in the ocean. This is espe- cially true in shallow water. 9
10 MARINE MAMMALS AND LOW-FREQUENCY SOUND 4. The strength of sound is measured on a logarithmic scale, 10 logic I/Iref (I = intensity] ); therefore, 180 dB is 10 times less intense than 190 dB, and 170 dB is 100 times less intense than 190 dB. Because of property 1, only low-frequency sounds are useful for studying large-scale processes over long distances, such as the structure of the ocean over scales of hundreds to thousands of kilometers. For example, sound has been used to study circulation patterns in the ocean using tomographic techniques analogous to the CAT [computerized axial tomography] scan technology used in medicine (Munk and Wunsch, 1979; Munk et al., 1995~. Likewise, sound is used in geophysical studies to characterize the subsurface structure of the seafloor. The decrease in sound transmission distance with increasing frequency also has implications for marine mammal communication because only low-frequency vocalizations can travel long distances. Because of property 2, sound speed can be used to infer the average temperature of the water volume through which the sound waves have passed. Scientists are using the relationship of the speed of sound and water temperature to infer whether global warming is occurring. The Acoustic Thermometry of Ocean Climate (ATOC) experiment is monitoring the 1Intensity is considered the fundamental quantity of sound, but it is seldom measured. Instead, pressure is normally measured. The two are related by I = p2/pocO, where p is the time-averaged pressure, pO is the density of the medium, and cO is the sound speed in the medium. An acoustic wave whose pressure is 1 pPa has an intensity of 0.64 * 10-22 watts/cm2. For transient signals, it is more meaningful to refer to the energy flux density (E) of the acoustic wave. The energy flux density is the time integral of the instantaneous intensity.
INTRODUCTION Sca Surface / / 1,000 m \~/ 135 did 2~)~}(} m 1 2~) `1D I (31) m 1 55 d] Source Level ~ 95 dB \ \ \ - - / FIGURE 1.1 Calculated received levels at various distances from the ATOC source, based on spherical spreading, assuming no reflections from the sea surface or bottom. Received levels are affected by conditions in the ocean, where the source is deployed (on the bottom or in the water column) and thus reflection from the sea surface and the seafloor, and directionality of the source. Spherical spreading is a proper assumption at these distances until sound waves reach a boundary (see Urick, 1983~. travel time of sound between sources off the coasts of Hawaii and California to receivers around the Pacific Ocean (see Figure 1.2) for a variety of purposes (see section below on "The ATOC Concepts. Ambient noise levels vary both from place to place in the ocean and over time at each location. The relative frequency bands also vary, due in part to the nonrandom distribution of vocal animals and human-generated noise. Measure- ments and predictions of ambient noise in the ocean were made by Knudsen et al. (1948) and Ross (1976~. Natural ambient noise levels increase as frequency decreases and are related to the sea state. Ross (1976) reported that the ambient noise in areas of heavy shipping activity could range between 85-95 dB (1 Hz bandwidth), peaking at a frequency of about 100 Hz. High sea states can produce similar levels of ambient noise. Some whales, seals, and fish use low-frequency sound to communicate and to sense their environments (Tyack, 1998~. For example, baleen whales and
2 Dual vl~ 1 j MARINE MAMMALS AND LOW-FREQUENCY SOUND it: - \ 30° Gus\\'-.\ a ° ~60° FIGURE 1.2 ATOC sources at Pioneer Seamount and Kauai, showing transmission paths to receivers in different parts of the Pacific Ocean. some toothed whales are known to use and respond to low-frequency sound emitted by other individuals of their species (McDonald et al.,1995; Edds-Walton, 1997; Ljungblad et al., 1997; Stafford et al., 1998~. Sharks and some other fish species are able to sense and react to low-frequency sound (Myrberg et al., 1976; Myrberg, 1990~. Therefore, it is possible that human-generated low-frequency sound can interfere with the normal behavior of some marine animals and there is some evidence that this occurs (Myrberg, 1978, 1980, 1990; Richardson et al., 1995~. Serious misunderstandings of the potential effects of sound of various intensities on marine mammals have occurred because the levels of sound inten- sity in water and in air have not been consistently (or in some cases, correctly)
INTRODUCTION 13 referenced to the International System of Units (SI) standards2 that have been established and in-water sound levels have been misunderstood to be comparable directly to in-air levels, with which most people are more familiar. Air-water comparisons are inherently misleading. THE ATOC CONCEPT Individuals studying the ocean or using it for scientific, commercial, or military purposes use underwater sound as a major tool to monitor and explore the ocean's contents and boundaries. Sound also enters the ocean as a result of natural environmental processes, biological activity, and human activities unre- lated to study of the ocean, such as the propulsion noise of ships (Table 1.1~. Although all kinds of sounds are used, many applications have used sound in the 1- to 100-Hz frequency range because absorption of these sound frequencies by seawater is minimal, variability caused by the environment is somewhat reduced, and long-range propagation is possible, making underwater or subbottom remote sensing feasible. An example of the scientific use of low-frequency sound in the ocean was the Heard Island Feasibility Test (HIFT), in which sound was transmitted from one array of sources with a sound level of 221 dB (rms)3 and a frequency of 57 Hz (rms) through the ocean to a number of receivers over distances of up to 16,000 km (Baggeroer and Munk, 1992~. A major goal of HIFT was to serve as a prototype for regular observations of the speed of sound in the ocean as a direct means of measuring the rate of ocean warming due to global climate change. The regular observations were proposed as the long-term ATOC experiment. The issue of global warming is of major significance to scientists, policymakers, and citizens worldwide, yet it has been difficult to determine the extent of atmospheric and oceanic warming based on observations of global air and sea surface temperatures.4 The advantages of long-distance sound transmis- sions in the ocean are that (1) low-frequency sound waves pass through and thus sample a wide range of ocean depths between the source and the receivers, (2) the summed effects of random variability along the transmission path due to eddies and variations in ocean currents are minimized, and (3) longer-distance 2ANSI S1.8-8-1989 (ASA 84-1989), Revision of S1.8-1969 (R 1974), Reaffirmed by ANSI on July 29, 1997. In the International system of units (SI), acoustic pressure is expressed in watts per square meter, but the do notation is used more commonly at the present time. 3The amplitude of pulsed sounds is typically expressed as ``peak-to-peak,, (one cycle of the sine wave) or ``zero-to-peak,, (one-half cycle). continuous sounds may be expressed as ``root-mean- square (rms), which is the square root of the time average of the square of a quantity; for a periodic quantity the average is taken over one complete cycle (Lapedes, 1974). 4The NRC recently issued a report on the measurement of atmospheric global warming (NRC, 2000).
4 MARINE MAMMALS AND LOW-FREQUENCY SOUND TABLE 1.1 List of Some Anthropogenic Sounds, Including Sources, Frequencies, and Levels Source Frequency at Highest Level Source Level at Highest Level 1/3-Octave Band (Hz) 1/3-Octave Band (dB re 1 pPa @ 1 m) 5-m Zodiac inflatable boata 6,300 Bell 212 helicopterb 16 152 159 Large tanker 100 + 125 177 Icebreaker 100 183 ATOC 75 195c Air gun array (32 guns) 50 2lod HIFT 50 + 63 221e Military search sonar 2,000-5,000 230+ SOURCE: Richardson et al. (1995, Table 6.9). aSpeed and horsepower of engines were not given in Richardson et al. (1995). bAircraft flyover source levels were computed by Malme et al. (1989) for a standard altitude of 1,000 ft (305 m). For consistency with other sound sources, these values were changed to a reference range of 1 m by adding 50 dB. CNumbers provided by ATOC investigators from actual transmissions, rather than from Richardson et al. (1995). dAnecdotal evidence suggests that airgun arrays may reach source levels of 240 dB. eNumbers provided by Heard Island Feasibility Test (HIFT) investigators. transmissions may allow the detection of smaller changes in temperature. The designers of ATOC hope to conduct ATOC transmissions from Kauai for at least 5 additional years to make a quantitative assessment of the role that acoustic thermometry can play in an integrated ocean-observing system for ocean weather and climate in the North Pacific Ocean (P. Worcester, Scripps Institution of Oceanography, personal communication, 1999~. Results from the initial ATOC transmissions indicate that this technique shows promise for at least one of its planned applications, namely, to use ocean temperature measurements from ATOC to constrain climate models (ATOC Consortium, 1998~. ATOC Consor- tium investigators compared sea-level estimates derived from historic averages, ATOC-based tomography, recent direct measurements, results from a general circulation model (GCM), and data from a satellite-based altimeter. Combina- tions of the GCM, altimeter, and ATOC data show that the GCM alone underes- timates the magnitude of the seasonal sea surface heat flux cycle. However, despite the usefulness of acoustic tomography programs like ATOC and other research uses of low-frequency sound in the ocean, concern exists that adding more sound to the ocean could harm marine mammals, sea turtles, and other organisms, as the following section will describe. It is necessary to sample the ocean's temperature frequently enough to be able to distinguish any trend of temperature increase amidst the "noise" of random
INTRODUCTION 15 variations created by temperature, tides, and mesoscale ocean structure.5 Infor- mation gained from ATOC-like techniques cannot be replaced by measurements from satellites because satellites can only sense features of the surface layer of the ocean. The temperatures from the full depth of the ocean can be measured in situ, but such measurements are limited in number and frequency because of the cost and limited number of oceanographic ships, moorings, and drifters available, compared to the great volume of the ocean. The planned Array for Real-time Geostrophic Oceanography (ARGO) drifter program will provide a new and innovative means of measuring ocean interior temperatures over large scales (albeit not in an integrated, synoptic manner), and may provide a complement to, and possibly a replacement for, ATOC-type acoustic measurements of water temperature. LOW-FREQUENCY SOUND AND MARINE VERTEBRATES It is well known that some marine organisms produce low-frequency sounds and/or can hear such sounds. For example, there is evidence that baleen whales (such as finback [Balaenoptera physalus], blue [Balaenoptera musculus], and humpback whales [Megaptera novaeangliae]) communicate using low-frequency sound (e.g., reviewed in Edds-Walton, 1997~. Table 1.2 shows the frequency range and dominant frequencies of the vocalizations of a sample of baleen whales, toothed whales, and seals. The geographic extent of the use of low-frequency sounds by baleen whales is being monitored on an experimental basis in the Atlantic and Pacific oceans using a novel source of data the Integrated Under- sea Surveillance System (IUSS) formerly known as the SOund SUrveillance System (SOSUS) which was originally designed for tracking submarines. The IUSS has allowed the tracking of individual whales in at least a few cases by triangulating the positions of vocalizations over time (Clark, 1995; Stafford et al., 1998; Watkins et al., 2000~. Such data are important in determining the migra- tion behavior of individual whales and in assessing whether human influences change these pelagic migrations. Richardson et al. (1995) present a comparison of the audiograms of some species of marine mammals (see Figure 1.3~. Addi- 5As Peter Worcester, ATOC principal investigator, explained to the Committee in writing in 1999 "The required duty cycle is actually set by the need to avoid aliasing of rapidly changing oceano- graphic phenomena. If high-frequency phenomena are sampled at too low a rate, they will errone- ously appear in subsequent analyses as low-frequency variability. In general one needs to sample at more than twice the highest frequency containing significant energy to avoid aliasing. In the case of the ocean, mesoscale variability has timescales from a week to a few months, and so needs to be sampled every few days. The tides are of course even higher frequency, but because their frequen- cies are well known, they can be sampled adequately using a frequency of approximately 1 day out of every few days. This combination of ocean phenomena led to a 2 percent duty cycle being used, consisting of 1 day with six 20-minute transmissions at 4-hour intervals to adequately sample tidal variability, occurring every fourth day to adequately sample ocean mesoscale variability."
16 MARINE MAMMALS AND LOW-FREQUENCY SOUND TABLE 1.2 Frequencies Used in Communication and Echolocation by Selected Manne Mammals Species Frequency Range (Hz)a Dominant Frequencies (Hz)b Selected Baleen Whales Gray Whale adults 20-2,000 20- 1,200 calf clicks 100-20,000 3,400-4,000 Humpback Whale 30-8,000 120-4,000 Finback Whale 14-750 20-40C Minke Whale 40-2,000 60-140d Southern Right Whalee 30-2,200 50-500 Bowhead Whale 20-3,500 100-400 Blue Whale Atlanticf 1 0-20h Pacificg 10-390 16-24 Selected Toothed Whales Sperm Whale (clicks) 100-30,000 2,000-16,000 White Whalei whistles 260-20,000 2,000-5,900 clicks 40,000-120,000 Killer Whale whistles 1,500-18,000 6,000-12,000 clicks 1,200-25,000 Bottlenose Dolphin whistle 800-24,000 3,500- 14,500 clicks) 1,000- 150,000 30,000- 130,000 Selected Seals California Sea Lion (in air) <1,000-<8,000 500-4,000 Harbor Seal (in air) <100-150,000+ <100-40,000 Gray Seal 100-40,000 100- 10,000 SOURCE: Modified from Richardson et al. (1995). aThe frequency range listed is the lowest to highest frequencies listed by Richardson et al. (1995) and more recent authors. Gaps in the ranges are not shown. bDominant frequencies are essentially the bandwidth of sound that has the greatest energy. They do not include all the frequencies produced, since there may be many weak harmonics. CEdds (1988). dEdds-Walton (2000). eAlthough few recordings exist, the northern right whale repertoire is likely to be similar. fPublished data are too limited to give the frequency range for this population. "Stafford and Fox (1996). hEdds (1982). iW. Perrin and D.W. Rice, both NMFS experts in taxonomy, verified that individuals of the species Delphinapterus leucas can be called white whales, belugas, or belukhas. White whale is used throughout this report. iRidgway and Au (1999).
INTRODUCTION 17 tional information about the effects of low-frequency sound on marine mammals is contained in Chapters 2 and 3. Low-frequency sounds are used by other marine vertebrates, including sharks and bony fish (Myrberg, 1972, 1978, 1980,1990~. Sharks are attracted to sources emitting such sounds as possible food indicators (e.g., Myrberg, 1978), and many species of fish use low-frequency sounds for communication (e.g., Demski et al., 1973~. ORIGIN OF STUDY As a result of issues raised by HIFT, the Office of Naval Research (ONR) requested in 1992 that the National Research Council examine the state of knowl- edge of the effects of low-frequency sounds on marine mammals and assess the trade-offs between the benefits of underwater sound as a research tool and the possible harmful effects on marine mammal populations of introducing addi- tional low-frequency sound into the ocean. In 1994 the NRC issued a report, Low Frequency Sound and Marine Mammals: Current Knowledge and Research Needs, which concluded that (1) very little is known about the effects of low- frequency sound on marine mammals and (2) it is difficult to establish regulatory policy in the absence of data regarding such effects (see Appendix B for the executive summary from that report). The report included a series of recommen- dations about the kinds of research needed to fill the gaps in our knowledge. Subsequent to HIFT, the ATOC program was proposed with a mission to make regular measurements of the travel times of low-frequency sound through- out the Pacific Ocean (Figure 1.2~. As a result of concerns about the effects of low-frequency sound added to the ocean by ATOC, the ATOC program con- ducted the first several years of transmissions under a permit to test the effects of ATOC sound sources on marine mammals through a Marine Mammal Research Program (MMRP). The Defense Advanced Research Projects Agency requested that the NRC update the information contained in its 1994 report based on the MMRP and other results.6 In addition, the NRC was asked to ascertain how data acquired since 1994 fulfill the research needs described in the 1994 report. An interim NRC report published in 1996 provided guidance to the MMRP in the midst of its observational studies. The director of the ONR program (Robert Gisiner) and the principal investigators of the MMRP (Christopher Clark and Daniel Costa) briefed the NRC's Committee to Review Results of ATOC's Ma- rine Mammal Research Program in 1996 and 1999 and participated in subsequent . . Open c 1scusslons. The Committee summarizes and comments on the results of the MMRP in Chapter 2. Chapter 3 is devoted to updating the research priorities first identified 6The MMRP formed its own advisory board to provide independent advice to MMRP investiga- tors regarding MMRP needs, plans, schedules, and research results.
18 Q 160- ~ 120- m - o a:, 80- t60 _` ~ 120 m - o Hi, 80 MARINE MAMMALS AND LOW-FREQUENCY SOUND Underwater Audiograms of Hair Seals T ;~>,:. . . ,. . ~ · ~ _ ~ I ..~.. HarborJM .: ~ . - HarborJT tic .. ,. i-] ~i. ~ . Harbor/K tG - . + /~ Ringed .' / 1: - At Harp ...~... . Monk . 1 1 ~ I 1 ,000,000 4 JO Coon - ''io,ooo ~oo,ooo Frequency (Hz) Underwater Audiograms of Eared Seals C Sea LionlS C Sea LioniK N Fur Seal/hA N Fur Seal/B 1 1 T I ~ I T '-i I ~ r- I t~ I - T I I I I T 1 1 )O 1,000 1O,OOO 10O,000 1,00O,OOO Frequency (Hz) FIGURE 1.3 Audiograms of representative seal and toothed whale species. Source: Richardson et al. (1995~; used with permission from Academic Press. References for these data are given in the source document. In most cases, the data represent measure- ments on one or two individuals of the species. NOTE: Complete audiograms should be U-shaped. If not, hearing should be tested at higher or lower frequencies, as necessary. For example, the audiograms of the true seals (shown as "hair seals" in the figure) appear to be truncated at lower frequencies.
INTRODUCTION 140- 100- a' m at} 0, 60- 20- 140- it ~ 100- a) m _ o u, Sol 19 Underwater Audiograms of Odontocetes w. ,( Sax ''' 1 .. _ , ,,. ~1 I -I I I I 111' -- -' - l~~-l~T r-rll' - I I I ~ 11~1. 10 100 1,000 1 O'OOO Frequency (Hz) ., Am. ~- .' Y . Am, 14 ->c- - Beluge Killer Wh Herb Porp _ Beiji . . . . 100,000 1,000,000 Underwater Audiograms of Odontocetes x~ No .~ i, NC_ or At"= \ '. " ~ - - 1- ~ 1 1 1 ~ I 1 ,000 1 0,000 Frequency (Hz) A., Bat [:)ol/J ...~... Bot Dol/L False Kil Risso's D Bo~u I - 1--1-- 1 T 1 T 1 TTT| 1 00,000 1 ,000,000 20- . . . 10 100 _ 1 1 111111-
20 MARINE MAMMALS AND LOW-FREQUENCY SOUND in the 1994 NRC report, based on data obtained from research conducted by the MMRP, as well as the results of other relevant research such as that sponsored by the ONR program on marine mammals and ocean acoustics. Based on this comparison of recent research achievements and research needs listed in the 1994 report, the present report identifies areas in which gaps in our knowledge con- tinue to exist. Chapter 4 discusses regulatory issues, such as how permits for acoustic and marine mammal research are issued. Chapter 5 draws together the Committee's findings and provides recommendations based on these findings.
