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.
~ Findings and Recommendations Or The Committee discussed a wide range of topics related to its charge. This chapter presents the results of the Committee's review of the Marine Mammal Research Program (MMRP), identification of important future research and obser- vations, specification of desirable regulatory reforms, and ideas for comprehen- sive monitoring and regulation of sound in the ocean. RESULTS OF THE MMRP Findings: Tests of the Acoustic Thermometry of Ocean Climate (ATOC) source were authorized under permits for the MMRP experiment. Although the MMRP was allowed some control over the specific days that the California transmissions took place, MMRP was retrofitted to an operational use of the ATOC source and was not designed to investigate the effects of the source most effectively. This situation illustrates problems that can be encountered in studies designed to monitor animals in the area where a noise source has been placed and is operated for other reasons, rather than optimizing transmissions to accomplish a specific research objective. As a consequence, the results of the MMRP do not conclu- sively demonstrate that the ATOC signal either has an effect or has no effect on marine mammals in the short- or long-term. In view of the lack of data for marine mammals exposed to the ATOC signal at received levels above 137 dB, and the incomplete analyses of much of the data collected off the Kauai source, the Committee could reach no conclusions about the effects of the ATOC source at the level of 195 dB. Data from tests of oil industry noises (Malme et al., 1983) and the low-frequency active (LFA) sonar tests (Tyack and Clark, 1998) indicate that these kinds of signals can alter the normal behavior of migrating gray whales, 73
74 MARINE MAMMALS AND LOW-FREQUENCY SOUND and data from MMRP are suggestive of a behavioral change in humpback whales and sperm whales exposed to 130 dB or greater from the ATOC signal. Thus, there is a potential for altering the normal behavior of marine mammals capable of hearing low-frequency sounds, such as those produced by the ATOC source, LEA sonar, and commercial shipping. The biological significance of short- and long-term exposure cannot be extrapolated from the limited data acquired during the short-term MMRP studies. Redistribution of marine mammals from tradi- tional feeding or breeding areas was not observed, but cannot be ruled out. Recommendations: If ATOC does continue, a marine mammal monitoring and research component should be required, but the marine mammal research program should have the flexibility to design and perform playback experiments opti- mized to yield information about biologically significant effects of the source on marine mammals. In general, any long-term experiment involving high source level ensonification of large areas of the ocean should take precautions to mini- mize exposure of marine mammals to dangerous levels of sound. Such precau- tions could include one or more of the following: · Design the physical structure of the source to minimize the potential exposure of marine organisms, if this is technically feasible. For example, a physical barrier could be erected around the source, like a radome on a radar facility. · Install sensors to shut down the source automatically when marine mam- mals are too close. · Make the source level and duty cycle as low as possible, given the objec- tives of the research. · Install the source away from large concentrations of marine mammals. · Design the ramp-up period to make it possible for marine mammals to avoid received levels that would cause temporary threshold shift (TTS). If the Kauai ATOC transmissions are continued, it will be important to continue ship- and air-based measurements of marine mammals within the 130- dB zone around the source. The Committee was told that ATOC investigators plan to continue aerial observations. Observations of marine mammals near the Kauai source should include humpback whales as well as other species. Vessel- based observations and aerial surveys are complementary and both methods should be used. The inability of shore-based observers on Kauai to observe humpback whales near enough to the ATOC source to be exposed to sound levels that would be likely to produce biologically significant behavioral responses indicates that shore-based observations should not be continued for studying the effects of the Kauai source. Shore-based methods are useful, however, when animals exposed to the sound levels of interest can be viewed easily from shore (e.g., Malme et al., 1983, 1984; Frankel and Clark, 1998a).
FINDINGS AND RECOMMENDATIONS 75 Long-term observations should be conducted near the Kauai source. Priority should be given to specific studies targeted to resolve areas of critical uncer- tainty. Since the source may be deployed for at least 5 more years to meet the ATOC objectives, long-term studies of animals near the source are a priority. If the source is to be operated for decades as part of an operational ocean-observing system, it is important to determine whether there are some animals that are resident in the area, because in such a case small effects might accumulate over time to have a larger impact. A study should be designed specifically to deter- mine whether there are resident animals in the source area and to study whether the ATOC source changes the pattern of residency for identified individual animals on day-to-day and interannual timescales. Such observations would be more sensitive indicators of habitat usage for resident populations than the more general comparisons of the distributions of sightings of each species conducted by the MMRP. A vessel-based study to photo-identify marine mammals sighted near the source could be used to test whether there are any resident individuals or popula- tions. If carried out over the years ATOC hopes to operate, such an observation program could provide more information on the status of the population (as did similar data for baleen whales off California; Calambokidis, 1996, 1999~. The vessel conducting this photo-identification work could also combine visual and acoustic monitoring of marine mammals and monitor for any unexpected effects of ATOC transmissions with a statistically designed sampling program. In addi- tion, ATOC has a responsibility to design a well-controlled study on the effects of ATOC transmissions on humpback whales within the 130-dB zone around the source. This would probably need to be vessel-based in order to follow animals near the source. It could use the same vessel involved with the photo-identifica- tion work. 2. Conduct aerial surveys around the ATOC source. If aerial surveys are conducted early enough in the season to document the migration of humpback whales into the area, there will be some evidence as to whether individual ani- mals enter the area around the receiver and later are repelled by the source. Individual animals are impossible to track for long from an airplane, but a broader coverage can usually be accomplished using aircraft rather than shore- or ship- based methods. Aerial surveys also allow documentation of the distributions of sperm whales and other species that are more difficult to observe from vessels to determine whether there is cause for concern about long-term exposure of these species. 3. Use radio-tagging and tracking, and recoverable data recorders. The use of such tags to study the effect of the California source on elephant seals provided a wealth of data on diving behavior and received levels of sound. Although tags and data recorders are harder to place on whales than on elephant seals and harder to retrieve later, they could be useful in providing the same kind of information
76 MARINE MAMMALS AND LOW-FREQUENCY SOUND on received level and diving behavior, with a potential for recording whale vocalizations. 4. Conduct studies on the effects of ramped-up signals. Ramped-up signals are used worldwide in high-energy seismic surveys as a common-sense measure, but their effectiveness has not been tested scientifically. Funding should be designated from the ATOC program or other sources for competitive grants to study the responses of seals and whales to a ramp-up of a noxious sound. Since baleen whales and sperm whales cannot be tested in controlled studies in captivity with the ATOC signal, and since most other species probably cannot hear it, a higher-frequency ATOC-like sound could be used. For example, responses of captive dolphins or white whales could be used to document the kinds of responses or lack thereof to a ramped-up signal. Do they ignore it or do they approach to investigate the new sound? How many exposures does it take (if ever) for them to leave the area as the ramp-up begins? Do they habituate to the ramped-up signal? At what level should the ramp-up start, in order to avoid startling the animals but still be heard above background noise (e.g., should the ramp-up phase start with a lower source level than 165 dB)? This would not be a perfect experiment because it could be argued that baleen whales might not respond the same way as small toothed whales, nor wild animals like captives, but such an experiment should provide insight into the potential behavior of other marine mammals. The Minerals Management Service (MMS) is funding a study in the Gulf of Mexico region of signal ramp-up that is designed to repel small toothed whales and seals before seismic surveys using airguns are started. Data from this study should be reviewed by ATOC investigators. FUTURE RESEARCH AND OBSERVATIONS The Committee reiterates the research recommendations made in NRC (1994, 1996~. Although progress has been made on many of these topics (e.g., TTS), all remain valid and were discussed in detail in Chapter 3. In addition to these research topics, research priorities are identified below and suggestions made for the appropriate institutional structure for selecting, funding, and overseeing re- search. The federal agencies and Congress should determine whether these ac- tivities are of high enough priority to merit reprogramming of existing funds or appropriating new funds. The speed with which these research and observation activities are implemented will depend on the urgency felt by the public, Con- gress, and federal agencies for gaining the knowledge necessary to address the goals of both protecting marine mammals and carrying out desirable human activities that add sound to the ocean.
FINDINGS AND RECOMMENDATIONS 77 Priority Studies Findings: The typical pattern for funding research on the effects of noise on marine mammals is for a group responsible for producing noise to be required to provide data on the effects of its operations. This leads the group to fund projects closely tied to the specific signals, areas, times of planned operations, and species of special concern to that project. For example, the primary approach of the MMRP has been to study one anticipated sound stimulus at three anticipated source sites. A problem with this approach is that results cannot be extrapolated readily to other stimuli or sometimes to other sites. There are hundreds of sources of potential concern, and it may be more efficient to develop a more comprehensive research program using a set of representative stimuli to more easily allow any users to model the predicted response to their own stimulus. Sound is multidimensional and cannot be characterized fully by a single measure, for example, peak intensity or frequency. Thus, in considering (1) the auditory capability of a species, (2) its response to a particular sound, and (3) regu- latory guidelines for exposure of animals to sound, the full dimensionality of sound should be taken into consideration. In this regard, factors that must be considered include the temporal and spectral characteristics of the sound, includ- ing the intensity, duration, duty cycle, frequency, bandwidth, rise time, temporal structure and similarity of any of these dimensions to biologically relevant sounds, as well as sensitivity of the relevant species to sounds with those characteristics. Decisions based on such parameters should be made more objective by combining parameter values in mathematical decision models. Such risk assess- ment models have yet to be developed and tested for marine mammals and sound because of a lack of basic information about how sound characteristics are related to species-specific hearing sensitivity. An additional complication of modeling the effects of sound on marine mammals arises in predicting the received levels at the animals, particularly in shallow water, because of reflection off the sea sur- face and seafloor and unexpected areas of anomalous temperatures, salinities, and densities. Such factors can cause differences between predicted and actual received levels as great as 30 dB (J. Lynch, Woods Hole Oceanographic Institution, personal communication, 1999~. This potential problem with acoustic propaga- tion models creates a responsibility for major noise generators to calibrate their model-predicted sound levels against measured levels. Recommendations: With regard to future research to study the effects of human- generated sound on marine mammals, the Committee supports the recommenda- tion of NRC (1994) that there is a need for planned experiments designed to relate the behavior of specific animals to the received level of sound to which they are being exposed. Very few studies have succeeded in this aim. Because studies of ocean acoustics and marine mammal behavior are very challenging, successful experiments will require a closer collaboration between biologists and acousti-
78 MARINE MAMMALS AND LOW-FREQUENCY SOUND clans than has been the case in the past for many field studies. Success will also require continued refinement of techniques for making acoustic and visual obser- vations, such as methods for locating vocalizing marine mammals and develop- ment of tags that can monitor received levels at the tagged animal. To move beyond requiring extensive study of each sound source and each area in which it may be operated, a coordinated plan should be developed to explore how sound characteristics affect the responses of a representative set of marine mammal species in several biological contexts (e.g., feeding, migrating, and breeding). Research should be focused on studies of representative species using standard signal types, measuring a standard set of biological parameters, based on hearing type (Ketten, 1994), taxonomic group, and behavioral ecology (at least one species per group; Box 5.1~. This could allow the development of mathematical models that predict the levels and types of noise that pose a risk of injury to marine mammals. Such models could be used to predict in multi- dimensional space where TTS is likely (a "TTS potential region") as a threshold of potential risk and to determine measures of behavioral disruption for different species groups. Observations should include both trained and wild animals. The results of such research could provide the necessary background for future envi- ronmental impact statements, regulations, and permitting processes. The uncertainty in predictions of received sound levels hinders the applica- tion of models of marine mammal responses to sound and will require three complementary approaches: (1) development of better acoustic propagation models; (2) development of better observing systems to gather the data needed in models; and when the first two are not feasible, (3) development of better systems to observe ambient sound and transient noise pollution events in the ocean. Any research that includes relatively loud sound sources should monitor sound levels around the source site to gather data to calibrate its acoustic propagation models. The locations of ATOC sites were dictated by requirements for the ATOC sound to reach many preplaced receivers and thus were not ideal for observing marine mammals. Pioneer Seamount is 88 km offshore, and observations on the north shore of Kauai were complicated by frequent high winds, waves, and bad weather. In general, studies designed to observe the effects of sound on marine mammals should be conducted in areas of high animal density for easy and less costly observation. Potential investigators should not transmit the noise until preexposure control data on individual subjects are obtained, and great care should be taken to ensure that ensonified marine mammals are not significantly damaged physiologically. Acoustic studies focused on topics other than marine mammals should try to keep sound sources away from marine mammal "hotshots," even if this compli- cates logistics, increases costs, and/or decreases the efficiency of the experi- ments. In the case of the MMRP, acoustic and marine mammal studies with different goals were linked, leading to the proposal to place the sources in areas with high concentrations of marine mammals. Continuation of the oceanographic
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: ~~d ~~S4~d~ !~.!~...!~.!~! ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ art- - - - -a ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ -a ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ amp- ~ ~ ~ ~ ~ ~ -a ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ a. ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ -a ~ ~ ~ ~ ~ ~ ~ amp- ~ ~ -amp- ~ ~ ~ ~ ~~Resea~h~and~o~bse~at}~o~n~s~shou~ld~be~=nd~u~d~on~least~onO~s~pecl~es~ln~=h~ -:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:---:-:-:-:-:-:-:---------:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:---:-:-:-:-:---:-:-:-:-:-:-:-:-:-:-:---:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:---:-:-:-:-:-:-:---:-:-:-:-:-:-:-:-:-:-:- ~4~.~Pyg~my~a~n~d~dwa~s~pe~rm~whal~es~and~po~rpo~lses~(~h~i~gh~-fmq~uency~[g~re~e~r~th~an~ ........................................... ......................................... .............................................. ~$1in~d4~1IId~lnil~lI~s~lg~nils~s~h~I~u~I~d~b4~d~IvI~lopId~I~tI~l~n~g~wllh~l~n~d~lyl~d~u~l4~s~o1~ . .. ~th4~ab~ov4~s~4n~sp~cl~s~g~u~ps~.~Th~4se~Si~g~n~al~s~s~h~ou~ld~4mu~lat4~the~Si~g~n~al~s~u~s~4d~ :::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: :::::::::::::: :::::::: ::::::::::::::::::::::::::::: "::::: .::::::::::~:::::: : .:::::::::::::::::::::::::::::::~::::::: a::::::: ~~ or human :IVI :l~es~l~n~ n~e~ocean~l~n~c Us ~l~ng~l~mp~u~ sent Con l~n~uous~sou~r~s~ ~~ !:~:~:~:~:~:~:~:~:~:~:~:~:~:~:::::~:~:~:~:~:~:~:::~:~:~:~:~:~:~:~:~:~:~:~:~:~:::~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:::~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:::~:~:~:~:~:~:~:::~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~:~: 33333333333333433333333~ `~3~13~^ ^;3~33^~^v~ ~I^~;^^~^ ~^3~ ~^^~3~3 ^~;^~^~ ~ . ~ ~ I::I ~ =~' ~ ~ ~ ~ ~ 0 - an ~ I-O:IV' I~ - 0 ~= ~ ~ ~ -- owl I=' ~ ~ ~ ~ "~ -- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ~2~.~Co~ntl~n~Uo~u~req~UenCv-~m~od~u~lated~(~LF^~and~o1~h~e~r~s~0~n~a~rs~)~am~p~1~1t~u~dd-~ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ~modI~II1Id~(drl~Il~ln9~hOs~In~l@~I~l~III~nds~s~hl~p~eng~lnIs),~b~sdbI~nd~lshl~p~nolse~ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -- - -- - - - - - - - ~Wh~on~tosti~ng~r0~p~s~on~tiv0~s~peci~s~s~ral~d~l~ e~re~n~t~b~iol~og~i~ca~l~pa~me~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ !... ~ ~ ~ ~ ~ ~ !... ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ !... .................. - - - - - - - -. - ~~shou~ld~ba~meas~u~md~4s~b~s~lS~r~futu~ra~gu~latlons~4n~d~l~nd~lvl~d~u~l~parm~litl~ng~ :::::::::::::::::::::::: : :::::::::::::: ::::::::::: ::: ::::::::::::::::::::::::::::::::::::::: :::::::: ::::::: . . e+; A ,^^~.~,^^~,~^~^* I, !,!.^^^;^~! ... . . ~ a. 0' , ,~, , ma, ,`Y a, ,~ a, .~. .. =~ ha, ,, . , ~~Ilol~Ig~=ll~y~s~lg~I~lflII~nt~bI~h-~lo~wl1~h~th~I~lItI~ntl~I~l~lI~ch~I~ngI~dIm~Ig~ph~lI~
80 MARINE MAMMALS AND LOW-FREQUENCY SOUND component of ATOC would involve less risk to marine mammals if the source were moved, but this could be prohibitively expensive and would alter the data series. Even if scientists use sound responsibly, however, this does not guarantee the protection of essential marine mammal habitat, because commercial users are not subject to the same permitting requirements. Studies of wild marine mammals should include careful determination of their locations, coupled with improved sampling and modeling of acoustic propa- gation to estimate received sound levels accurately. Alternatively, acoustic data loggers could be mounted on individual animals to record (1) the sounds (and their levels) to which the animals are exposed; (2) their vertical and horizontal movements; and (3) the sounds produced by the animals, including physiological sounds such as breathing and heartbeats. Preliminary analysis of MMRP data from tagged elephant seals demonstrated the usefulness of this approach in assess- ing behavioral response to ATOC sounds at Pioneer Seamount (Costa et al., 1999) and showed potential for use in other comparable studies. Acoustic data loggers will be particularly important for research on deep-diving mammals, whose behavior and exposure cannot be monitored systematically at depth in any other way. A central theme of this report is that the task of developing predictive models of acoustic conditions that would harm marine mammals could be simplified by partitioning research among a small number of species that are representative in their hearing capabilities and sensitivities of larger groups of marine mammals. Box 5.1 describes the priority species groups, signal characteristics, and biological response parameters that should be investigated. Richardson et al. (1995) summarized studies of marine mammal responses to human-generated sounds, particularly those associated with oil exploration and shipping. Some of these studies reported a significant difference between levels of pulsed versus more continuous sounds required to evoke a response in whales. To evoke the same level of response in migrating gray whales, a pulsed air gun sound required levels 50 dB higher than a diverse array of low-frequency con- tinuous sources. This result is unexpected based on human hearing capabilities. How do marine mammals respond to signals with durations between the pulsed air gun noise (pulses separated by 7 to 15 seconds) and more continuous sounds? Another important question is: How do marine mammals respond when the received level is the same from two sources at different distances? This would help to discriminate whether marine mammals generally respond to received level (as was the case in the Phase II LEA study), estimated range to source, the gradient of acoustic energy over distance, and/or other sound characteristics. Response to Stranded Marine Mammals Findings: Observed behavioral responses of baleen whales to human-generated sound have all been reactions to sounds that are near or above the noise back-
FINDINGS AND RECOMMENDATIONS 81 ground (Richardson et al., 1995~. Behavioral response thresholds in the range of 100 to 110 dB (Dahlheim and Ljungblad, 1990; Frankel et al., 1995) are consid- erably greater than human underwater threshold hearing levels. As has been suggested by Frankel et al. (1995) and Richardson et al. (1995), field observations of acoustic response thresholds probably have been limited by background noise rather than being indicative of true hearing thresholds. The NRC (1994, p. 57) recommended the organization of a Stranded Whale Auditory Test (SWAT) team to obtain audiometric data from stranded or ensnared whales using recently developed electrophysiological techniques and instrumen- tation (e.g., Dolphin, 1997) for evoked potential studies (Hall, 1992~. Some thresholds for killer whales and dolphins have been obtained using evoked poten- tial methods (Popov and Supin, 1998; Popov and Klishin, 1998; Popov et al., 1998; Szymanski et al., 1995, 1998), but further research on methods of evoked potential audiometry are required for the application of the SWAT team approach to large whales. Evoked potential audiograms, even on a few animals (e.g., using the SWAT team approach), would resolve the issue of whether baleen whale thresholds are below the observed behavioral response. Levels producing TTS often are stated as decibels above absolute sensitivity, so knowing the audiogram would be important for regulatory decisions and policy making if noise levels are being controlled based on TTS. Determining the hearing capabilities of the marine mammals in the categories listed in Box 5.1 is an exceptionally high priority for future research, and plans for such studies should be developed and implemented as soon as possible. The ultimate goal of such studies should be to provide information on hearing sensitivity, TTS, nonauditory effects, and other biological parameters listed in Box 5.1 to help in the determination of sound levels that might affect marine mammal hearing or significantly alter their behavior. Physiological methods such as the auditory brainstem response (ABR) pro- vide baseline data for use in evaluating promising behavioral techniques such as playbacks like those employed by Dahlheim and Ljungblad (1990) and others and thus are an important complement to behavioral techniques. So far these playback methods have produced thresholds that are on the order of 50 dB less sensitive than thresholds at the most sensitive frequencies obtained in the labora- tory setting using ABR techniques with smaller species. Recommendations: The concept of SWAT teams recommended in NRC (1994) and NRDC (1999) should be implemented by funding trained scientists and asso- ciating them with stranding networks. The Office of Naval Research (ONR) partially funded a small effort to support the activities of a SWAT team, but the hardware and field methods are not yet adequate for wide testing. The ONR program manager (R. Gisiner) estimates that a considerable, but not unreason- able, amount of hardware and software design and testing will be needed (about 1 to 2 years of effort) before a system capable of regular operation under the
82 MARINE MAMMALS AND LOW-FREQUENCY SOUND SWAT team approach is feasible. However, this activity should be expanded to at least two teams, one on the east coast and one on the west coast of the United States. The teams should be responsible for (1) necropsy of suspected/possible marine mammal victims of sound injury (to be able to show whether sound caused the injuries or deaths) and (2) testing of hearing on stranded or entangled live animals. There is a need to expand the pool of individuals capable of doing this kind of work and capable of relating ear anatomy to function. An immediate need is for funding a specialist in evoked potential audiometry to develop improved methods applicable to large whales. A postdoctoral fellowship might be the most economical way to achieve this goal. NMFS and/or ONR should include funding for such work in the next budget cycle. Alternative possibilities for studying hearing in animals that are not kept in captivity also should be explored, such as placing a tag with electrodes on the head of a free-swimming whale and playing sound to the animal in a quiet environment. Multiagency Research Support Findings: Most marine mammal studies are funded from mission-oriented sources. At this time the greatest source of funding for marine mammal research is ONR. However, by its nature, ONR-funded research tends to be focused on questions of practical importance to the Navy and is not necessarily responsive to the broad interests of scientists seeking to learn more about the basic biology of marine mammals. Scientist-driven fundamental research could significantly improve our understanding of hearing and the effects of low-frequency sound on marine mammals, as well as our overall understanding of the acoustic behavior of these animals. Recommendations: If government funding shortages and priorities continue to constrain budgets for marine mammal research in the foreseeable future, manage- ment of sound in the ocean should remain conservative (and should incorporate management of all sources of human-generated noise in the sea, including indus- trial sources), in the absence of required knowledge. If government regulators need better information on which to base decisions, they should take such steps as necessary to provide increased funding for marine mammal research and to improve the ways that needed research is identified, funded, and conducted. Acquiring better information is often complicated because the regulatory parts of agencies like NMFS and FWS are separated from research, and funded research may not necessarily match research needed by regulators. It is imperative that the research and regulatory arms of NMFS and FWS maintain good linkages within these two agencies, and that priority is given to research needed by regulators in each agency. Government agencies with basic science missions (e.g., National Science Foundation [NSF] and National Institutes of Health [NIH]) should fund marine mammal research at the levels needed to answer fundamental questions
FINDINGS AND RECOMMENDATIONS 83 about hearing anatomy and physiology. Mission agencies with responsibilities related to marine mammals (e.g., ONR, National Oceanic and Atmospheric Administration [NOAA], MMS, U.S. Geological Survey [USGS]) should also fund basic research (notwithstanding ONR's limitations under the Mansfield Amendment), in the spirit of the recommendation of NRC (1992) that "federal agencies with marine-related missions find mechanisms to guarantee the continu- ing vitality of the underlying basic science on which they depend" (p. 28~. Such research should receive the same level of peer review as other basic research and be competitive with such programs for funding. Because marine mammal research is quite expensive, multiagency funding may be necessary to spread the costs. Alternatively, multiple parts of the same agency may need to cooperate in order to provide sufficient funds. Multidisciplinary Research Teams and Peer Review Findings: The MMRP suffered from the lack of an interdisciplinary group for planning its research. Further, the MMRP was added to an existing research program, rather than being designed to fulfill its own objectives. The agenda for the acoustic oceanographic component of ATOC required different transmission schedules from those that would have been optimal for marine mammal research. For example, the geographic location, depth, and duty cycle (in part) of the source were determined by the needs of ATOC, not the needs of the MMRP. As a consequence, the biological data that resulted were not optimal for answering the fundamental biological questions raised by the ATOC transmissions. Recommendation: Consideration should be given to establishing a multi- investigator program to study the effects of sound on marine mammals, funded by consortia of government agencies, non-governmental organizations, shipping, and hydrocarbon exploration and production industries. These consortia should include individuals, organizations, and companies in nations that share marine mammal stocks and sound-producing activities with the United States (e.g., Canada, Mexico, nations of the North Atlantic Treaty Organization). Such consortia could be initiated through a workshop to bring together the interested communities. The design and implementation of auditory research on marine mammals ideally should be an interdisciplinary enterprise. Valuable contribu- tions can be made by physical acousticians on the choice of sound stimuli to be used, by electronics experts on the choice and calibration of transducers for presenting the stimuli, by marine biologists on the choice of species and/or the best season and location for testing, by psychoacousticians on the testing proce- dures, and by statisticians on initial design and eventual data analysis and presen- tation. Without collaboration among specialists within these various disciplines, there is a greater probability that expensive and time-consuming projects will contain errors that preclude an unambiguous interpretation of the results. These
84 MARINE MAMMALS AND LOW-FREQUENCY SOUND projects are sufficiently complex that one or two individuals cannot reasonably be expected to have the full range of knowledge necessary for success. The logistical difficulties, permitting issues, and expense of such research demand advanced planning in all these areas. If such a research program is established, it should use a public Request for Proposal (RFP) process that results in proposals from more than one research team and is modeled after the peer-review processes used by NSF and NIH. Conversely, some research should continue to be funded through the less conser- vative ONR model, which provides program managers with greater latitude to fund more innovative science. A spectrum of funding styles is useful. The REP should be well advertised to encourage ideas and proposals from a wide range of researchers and institutions (including foreign participants), rather than relying on a set of traditional investigators. The goal of the process should be to optimize the selection of hypotheses, methods, and design and to identify the best performers) (e.g., best track record in previous work) for the proposed work. It is to the advantage of the sponsors to implement programs of broad-based peer review for such proposals. Future research on marine mammals unquestionably would profit from a broad-based review of the plans developed by multi- disciplinary teams and evaluated by a peer-review process that is objective and independent. Such a review should determine whether the proposing investiga- tive teams did the following adequately: · identified basic problems); · established specific hypotheses to be tested, with appropriate methods for data reduction, data presentation, and statistical analysis; · identified optimal experimental methods and test conditions (including geographic location of study); and · evaluated the power of the proposed experimental design. Because long research projects often need to adjust to experience gained in field programs and learning about what kinds of observations are practical and achievable, it is important to provide advice from an outside review team later in the life of a project. Sponsors of research need to be aware that studies funded and led by one special interest are vulnerable to concerns about conflict of interest. For example, research on the effects of smoking funded by NIH is likely to be perceived to be more objective than research conducted by the tobacco industry. Concern for peer review, efficiency, and independence argues for having an agency such as NSF take the lead in managing an interagency research program on the effects of . . noise on manne mamma s. Agencies that fund such applied research should ensure that adequate fund- ing for analysis and plans for peer review are in place before a research award is made. Analysis might be speeded by employing a larger team for analysis and
FINDINGS AND RECOMMENDATIONS 85 involving this team in planning the observations to make them as easy as possible to analyze later. Although publication in peer-reviewed journals is the standard by which most research is judged, applied research output from projects like the MMRP is not necessarily suitable for publication in available academic journals and the results may need to be used for regulatory decisions within a shorter amount of time than the normal journal paper cycle. Timely peer review of such studies might be better accomplished by conducting a mail and/or panel review of results by an independent group established specifically for this purpose. Population-Level Audiograms Findings: Ridgway and Carder (1997) published the first evidence that auditory capabilities in bottlenose dolphins may vary with sex and may change with age, similar to observations in humans (Ward, 1997~. These data reinforce the recom- mendation of NRC (1994) that audiograms should be obtained for many indi- viduals in a population to determine the normal range of hearing capability and the effects of aging. Because of the difficulty and expense of training and main- taining large numbers of animals, most studies collect data from one or two individuals of a particular species. Although individual differences have been noted (e.g., Terhune and Turnbull, 1995; Schlundt et al., 2000), no single study has used the same methods for multiple individuals of both sexes and varied ages. In addition, Ridgway and Carder (1997) reported that a young dolphin apparently had been deafened due to disease and had survived in the wild although deaf and mute. Clearly, there is a range of normal hearing among individuals, and even deficits may not prove fatal for individuals of social species. The major barrier to large-scale testing of the hearing of many individuals of the same species has been the need to train each individual to respond to sounds in measurable ways. The further development of audiometric procedures based on auditory evoked potentials (Hall, 1992; Szymanski et al., 1999) would eliminate that problem. Recommendations: Federal agencies should sponsor studies on the hearing abilities of both free-swimming and stranded animals. Population-level audio- grams of many individuals (such as are performed for humans; see Yost and Killion, 1997) are necessary for establishing the baseline of hearing capabilities and normal hearing loss in marine mammals, as also recommended in NRC (1994~. Stranded animals should be assessed to determine if their hearing is "normal." Data are needed to provide comparisons that would allow an evalua- tion of how common hearing deficits may be among stranded animals. The development of population-level audiograms will require the perfection and wide use of auditory evoked potential techniques, to eliminate the need to train all tested animals. However, if the cost and techniques limit widespread auditory evoked potential measurements of captive animals, a good sample of multiple animals (different ages and both genders) of the same species should be tested.
86 MARINE MAMMALS AND LOW-FREQUENCY SOUND National Captive Marine Mammal Research Facility Findings: There are few sources of trained marine mammals and few facilities available to academic (or even government) scientists for closely controlled research on the hearing capabilities of captive marine mammals and on sick or injured marine mammals being rehabilitated for release back into the wild. The costs of capturing, training, and maintaining marine mammals are great, meaning that anyone working with a marine mammal must make a long-term commitment to its care and well-being. Unlike work with small lab mammals or farm animals, marine mammal research requires decades of obligation to the animals, including experienced trainers and veterinarians and long-term care of trained and retired animals. Some rare marine mammals will only become available for study at such facilities that have received them after rescues. Lack of specialized research facilities hinders the priority studies described earlier. For example, a rare oppor- tunity to obtain evoked potential audiometric data from a stranded baleen whale (a juvenile gray whale) was lost when competing demands for access made it impossible to get sufficient time with the animal to test its hearing. This experi- ence emphasizes the need for not only better methods of obtaining auditory information, but also for gaining access to species for which few or no auditory data exist. Currently, there is only one site in the United States (and perhaps the world) that has the facilities and animals that could be used in such studies. This site is operated by the U.S. Navy in San Diego, California.1 Even this site has its limitations, however, in that investigators must be U.S. citizens to work with the trained animals. Although some research on the hearing abilities of marine mammals could possibly be conducted at public aquariums, research commit- ments at aquariums usually are secondary to public display requirements. In addition, they are not able to do research with animals in the open sea, as is possible with the highly trained animals maintained by the Navy. Such training takes years and is beyond the capabilities or interests of public aquariums. The ocean science community has a variety of different centers and shared-used facilities, for example, the Ocean Drilling Program, that could provide a model for a national captive marine mammal research facility. Recommendations: If the studies described in Chapter 3 and Box 5.1 are of sufficient priority to reduce uncertainties in the regulation of human-generated sound in the ocean, federal agencies should establish a national facility for the study of marine mammal hearing and behavior. The Committee believes that such a facility might be established at relatively little incremental cost by enhance- ment of the existing Navy facility. 1The one committee member associated with this facility did not take a position on whether the facility could or should be expanded and made available beyond Navy scientists.
FINDINGS AND RECOMMENDATIONS 87 The facility for captive marine mammal research would have animals for "hire" by investigators funded for peer-reviewed research. Offset funds would come from individual grants and researchers, but the funding base for such a facility should not be provided solely by such offsets. Allocation of space, animals, and facility resources should be determined by a broad-based review board on the basis of the quality and significance of the proposed research. An additional virtue of establishing a national captive marine mammal research facility is that the total number of marine mammals removed from the wild would be minimized. Investigators could apply for support for short- or long-term study of the animals at this facility, from the range of agencies funding marine mammal research, at costs that would not have to include long-term maintenance of the animals. Such a facility should include the capability to work with trained ani- mals in the open ocean. The Navy's Marine Mammal Program facility in San Diego keeps marine mammals and already has trained animals and expertise in maintaining them. Its role potentially could be expanded to provide a more widely accessible national facility, including unclassified research. If such a facility is operated by the Navy, it will be necessary to ensure that research data are not restricted from publication. Establishment of a facility to promote field studies could also enable research recommended in this report, but such a facility would be more expensive and a lower priority than a national facility for research on trained, captive animals. REGULATORY REFORM Findings: The existing permit system for acoustic research is unnecessarily restrictive in some aspects and not comprehensive enough in other regards. It is not scientifically defensible to apply general source standards for permit require- ments (120 dB for continuous sound, 160 dB for intermittent sound, and 180 dB for sounds of all frequencies and durations) for all species and all sound charac- teristics under all possible conditions. Until NMFS publishes new acoustic guide- lines, current NMFS policy recommends applying for a small-take authorization for sound-producing research activities that have the potential to harass, injure, or kill a marine mammal (K. Hollingshead, NMFS, personal communication, 1999~. Different species have different sensitivities and susceptibilities, and sound effects may accumulate as new sources are added. In the absence of information, managers rightly have chosen to be cautious, at least in regard to permitting ocean science research. However, as noted in NRDC (1999), there is virtually no regulation of sound produced by large commercial ocean-going vessels: "The worst polluter, shipping, is also the least regulated, while a comparative light- weight [in terms of the amount of sound put into the ocean], scientific research, is far more strictly scrutinized" (p. 13~. This combination of cautious regulation of minor activities, coupled with lack of regulation of major noise sources, will not,
88 MARINE MAMMALS AND LOW-FREQUENCY SOUND in the long run, adequately protect marine mammals from potentially deleterious effects of noise and could unnecessarily impede important acoustic research. Recommendations: Congress should change the Marine Mammal Protection Act (MMPA) and/or NOAA should change the implementing legislation of the MMPA to allow incidental take authorization based solely on negligible impact on the population. Research should be undertaken to allow the definition of Level A harassment to be related to the TTS produced in a species, when known. Level B harassment should be limited to meaningful disruption of biologically significant activities that could affect demographically important variables such as reproduction and longevity. COMPREHENSIVE MONITORING AND REGULATION OF SOUND IN THE OCEAN Findings: Protecting marine mammals from significant adverse impacts clearly requires a broader application of regulations. There is a global increase of sound levels in the sea resulting from shipping, recreation, aircraft, and naval operations as well as research (Urick, 1986~. It is important to characterize the existing ambient sound field in terms of the levels, frequencies, and locations of sources, especially in terms of areas particularly important for marine mammals (i.e., the "hotshots" of NRDC, 1999~. Such a characterization of the ambient noise field will provide a context for determining when, where, and with what characteristics new sources could be added. Recommendations: Noise monitoring is important and acoustic hotspots should be identified. Fortunately, ambient noise data exist for a variety of locations, which could provide time series and baselines for additional monitoring. Exist- ing data should be identified and made accessible through a single easy-to-access source. Like marine mammal research programs, funding for noise monitoring should be awarded based on responses to a request for proposals and careful evaluation of the costs and benefits of the proposed systems. The opening of the existing IUSS for whale research was important for demonstrating the power of bottom-mounted hydrophore arrays, but the IUSS may or may not provide the best system for the acoustic monitoring tasks envisioned here, given that it was designed for an entirely different purpose. The first step in comprehensive monitoring and regulation of sound in the ocean should be to attempt to characterize the existing ambient sound field in the ocean and to characterize the sources that contribute to it. Monitoring of baseline sound levels should be carried out, particularly in critical habitats of acoustically sensitive or vulnerable species or in habitats critical to specific life stages, such as breeding and calving areas. Protection of marine mammals from subtle or long- term effects of harassment cannot be achieved through regulation of individual
FINDINGS AND RECOMMENDATIONS 89 "takes." An alternative habitat-oriented approach is required to protect marine mammals from the cumulative impacts of noise pollution, chemical pollution, physical habitat loss, and fishing. Such an approach requires monitoring of the status of marine mammal populations along with the quality of critical habitats, including the acoustic quality. Account should be taken of the populations in- volved; it is sensible to protect more rigorously species that are more endangered (e.g., northern right whales) than those that are less at risk. Basic research regarding what is significant about critical habitats and what factors have population-level effects for example, food supply, water quality, and noise levels and characteristics will prove much more effective for protecting marine mammals than merely attempting to regulate individual human activities that may potentially cause changes in the behavior of an individual marine mammal. NMFS regulations should encompass the entirety of noise pollution and other threats to marine mammals.
