Content uploaded by Marla M Holt
Author content
All content in this area was uploaded by Marla M Holt on Oct 01, 2014
Content may be subject to copyright.
NOAA Technical Memorandum NMFS-NWFSC-89
U.S. DEPARTMENT OF COMMERCE
National Oceanic and Atmospheric Administration
National Marine Fisheries Service
Sound Exposure and
Southern Resident Killer Whales
(Orcinus orca):
A Review of Current Knowledge and Data Gaps
February 2008
NOAA Technical Memorandum
NMFS Series
The Northwest Fisheries Science Center of the National
Marine Fisheries Service, NOAA, uses the NOAA Techni-
cal Memorandum NMFS series to issue informal scientific
and technical publications when complete formal review
and editorial processing are not appropriate or feasible due
to time constraints. Documents published in this series may
be referenced in the scientific and technical literature.
The NMFS-NWFSC Technical Memorandum series of the
Northwest Fisheries Science Center continues the NMFS-
F/NWC series established in 1970 by the Northwest &
Alaska Fisheries Science Center, which has since been
split into the Northwest Fisheries Science Center and the
Alaska Fisheries Science Center. The NMFS-AFSC Techni-
cal Memorandum series is now being used by the Alaska
Fisheries Science Center.
Reference throughout this document to trade names does
not imply endorsement by the National Marine Fisheries
Service, NOAA.
This document should be cited as follows:
Holt, M.M. 2008. Sound exposure and Southern Resident
killer whales (Orcinus orca): A review of current knowl-
edge and data gaps. U.S. Dept. Commer., NOAA Tech.
Memo. NMFS-NWFSC-89, 59 p.
NOAA Technical Memorandum NMFS-NWFSC-89
U.S. DEPARTMENT OF COMMERCE
National Oceanic and Atmospheric Administration
National Marine Fisheries Service
Sound Exposure and
Southern Resident Killer Whales
(Orcinus orca):
A Review of Current Knowledge and Data Gaps
Marla M. Holt
Northwest Fisheries Science Center
Conservation Biology Division
2725 Montlake Boulevard East
Seattle, Washington 98112
February 2008
ii
Most NOAA Technical Memorandums
NMFS-NWFSC are available online at the
Northwest Fisheries Science Center
web site (http://www.nwfsc.noaa.gov)
Copies are also available from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
phone orders (1-800-553-6847)
e-mail orders (orders@ntis.fedworld.gov)
Table of Contents
List of Figures............................................................................................................................................... v
List of Tables ..............................................................................................................................................vii
Executive Summary ..................................................................................................................................... ix
Acknowledgments........................................................................................................................................xi
Abbreviations and Acronyms ....................................................................................................................xiii
Introduction................................................................................................................................................... 1
Basic Concepts .........................................................................................................................................1
Frequency Bandwidth............................................................................................................................... 2
Southern Resident Killer Whales..................................................................................................................4
Killer Whale Sound Production and Function .............................................................................................. 6
Characterizing Sound and Propagation......................................................................................................... 8
Sound Propagation Variables ...................................................................................................................8
Ambient Noise..........................................................................................................................................8
Whale Watching Sound Propagation...................................................................................................... 12
Other Anthropogenic Sound Sources .....................................................................................................14
Modeling Sound Propagation ................................................................................................................. 15
Auditory Capabilities and Auditory Effects of Sound Exposure ................................................................ 18
Audiograms and Basic Auditory Function ............................................................................................. 18
Hearing Sensitivity ................................................................................................................................. 19
Other Capabilities ................................................................................................................................... 21
Auditory Masking, Critical Ratios, and Critical Bandwidths.................................................................22
Spatial Overlap ....................................................................................................................................... 23
Hearing Loss Due to Sound Exposure.................................................................................................... 25
Behavioral Changes in the Presence of Unwanted Sound .......................................................................... 29
Behavioral Patterns................................................................................................................................. 29
Active Sonar and other Anthropogenic Sounds...................................................................................... 30
Vocal Response to Background Noise ...................................................................................................31
Strandings and other Nonauditory Effects of Sound Exposure ..................................................................32
Zones of Influence ......................................................................................................................................34
iii
Zone of Audibility .................................................................................................................................. 34
Zone of Responsiveness .........................................................................................................................36
Zone of Masking and Effects on the Active Space of Sound Emissions................................................ 37
Zone of Hearing Loss or Injury .............................................................................................................. 42
Likelihood of Acoustic Impacts on the SRKW Population ........................................................................ 45
Risk Assessment ..................................................................................................................................... 45
Extent of Masking Effects ...................................................................................................................... 46
Interaction with Nonacoustic Variables.................................................................................................. 47
Recommendations for Future Work............................................................................................................ 49
Masking Effect Assessment.................................................................................................................... 49
Behavioral Response .............................................................................................................................. 50
Likelihood of Auditory Injury ................................................................................................................ 50
Conclusions................................................................................................................................................. 52
References...................................................................................................................................................53
iv
List of Figures
Figure 1. Critical habitat designation for Southern Resident killer whales .................................................5
Figure 2. Cumulative distributions of 2-second average SPLs.................................................................. 10
Figure 3. Spectrum levels in Haro Strait in July 2005 and November 2005..............................................11
Figure 4. Received spectral levels of a container ship traveling at 21 knots as it passed 442 m from
the recording equipment ............................................................................................................................. 14
Figure 5. Spectrum source levels of all vessels at cruise speeds recorded on the broadband
hydrophone system .....................................................................................................................................15
Figure 6. Behavioral and physiological audiograms based on averaged thresholds for two female
killer whales ................................................................................................................................................ 20
Figure 7. Equal loudness contours of human subjects depicting the level of a comparison tone
required to match the perceived loudness of a 1,000 Hz tone presented at different levels .......................22
Figure 8. Behavioral audiograms plus electrophysiological thresholds below 2 kHz for the killer
whale, bottlenose dolphin, and beluga whale .............................................................................................28
Figure 9. Killer whale behavioral audiogram and one-third octave ambient levels in a sea state of 0...... 35
Figure 10. Predicted maximum horizontal detection ranges at 50 kHz for a killer whale at the surface
echolocating on a Chinook salmon at 65 m depth for various noise conditions.........................................42
v
vi
List of Tables
Table 1. Source levels for whale watching vessels at different speeds, calculated received levels,
and measured received level of a container ship......................................................................................... 39
Table 2. Predicted maximum horizontal detection ranges at 50 kHz for a killer whale at the surface
echolocating on a Chinook salmon at 65 m depth, reduction in range relative to ambient Haro Strait
measurement, and percent of reduction in range ........................................................................................41
vii
viii
Executive Summary
This document reviews what is currently known about potential acoustic impacts on
endangered Southern Resident killer whales (SRKWs). Killer whales (Orcinus orca) use sound
for echolocation, social communication, and passive listening. Ambient noise, including that
from natural and anthropogenic sources, has the potential to interfere with the reception and use
of these important biological sounds. Significant sources of anthropogenic sounds that
contribute to ambient background noise in critical habitats of SRKWs include sonar, acoustic
harassment devices, vessel traffic, and construction noise.
Most measurements of ambient sounds made in SRKW habitat are greatly influenced by
vessel traffic that, at close ranges, raises noise levels significantly above ambient levels. In order
to address potential acoustic impacts, particularly from anthropogenic sources, this document
reviews parameters of sound that are pertinent to the auditory capabilities of killer whales and
various studies on noise effects in killer whales and other dolphins. The latter includes auditory
ramifications such as auditory masking or hearing loss and behavioral effects such as disruption
of foraging events or avoidance of an area. With this information, the document then
incorporates information on the soundscape of SRKW habitat and defines zones of audibility,
responsiveness, masking, and hearing loss and addresses the likelihood of acoustic impacts on
the SRKW population.
Lastly, recommendations are made for future work in order to address gaps in
information that, if available, would increase confidence in predicting the likelihood of acoustic
impacts on SRKWs.
ix
x
Acknowledgments
Sue Moore, Brad Hanson, Dawn Noren, Candice Emmons, and Mike Ford provided
valuable feedback on writing this report. This work was funded by the National Research
Council Postdoctoral Research Associateships Program and the Northwest Fisheries Science
Center.
xi
xii
Abbreviations and Acronyms
ABR auditory brainstem response
AHD acoustic harassment device
CR critical ratio
dB decibel
DL difference limen
DI directivity index
EL echo level
Hz hertz
kHz kilohertz
µPa micropascal
ms milliseconds
MTTS masked temporary threshold shift
NL noise level
Pa pascal
PAL passive acoustic listener
PCB polychlorinated biphenyl
PTS permanent threshold shift
RMS root-mean-square
RL received level
SL source level
SEL sound exposure level
SPL sound pressure level
SRKW Southern Resident killer whale
TL transmission loss
TS target strength
TTS temporary threshold shift
VTOSS Vessel Traffic Operations Support System
xiii
xiv
Introduction
Southern Resident killer whales (SRKWs) are an endangered population of
approximately 85 individuals that spend the summer in inland waters surrounding British
Columbia and Washington State. These killer whales (Orcinus orca) are fish eaters that
typically feed on Pacific salmon (Oncorhynchus spp.). They actively use sound for echolocation
and vocal communication. Additionally, they can glean information about the environment such
as the presence of prey from passive listening. Both natural and anthropogenic sounds have the
potential to impact the use of biologically important acoustic signals by SRKWs. Concern about
anthropogenic sound exposure, such as those produced by vessels or military sonar, has provided
the impetus to study and describe the acoustic environment that Southern Residents inhabit and
the effects of sound exposure on their auditory system, behavior, and physiology.
The purpose of this paper is to review what is currently known about killer whale
auditory capabilities and the use of sound by killer whales, the characteristics of sound in their
environment, and effects of sound exposure in killer whales and other dolphins in order to
address potential acoustic impacts on the SRKW population. As is the case for all marine
mammal groups, it is extremely difficult to address acoustic effects that might have indirect or
small but consistent consequences at the population level, as opposed to those with immediate
and sometimes extreme outcomes. The life history patterns and habitats of marine mammals in
general make this assessment extremely challenging (NRC 2005). In most cases, there is
insufficient empirical data on which to draw in order to address acoustic impacts at the
population level. A review of the current data and their limitations allows the opportunity to
assess data gaps, which are summarized at the end of this document.
Basic Concepts
Given the various ways that sound energy is quantified, it is necessary to review some
basic acoustical concepts. Sound is essentially generated when a vibrating object sets molecules
in a medium adjacent to that object into motion. Sound amplitude or what is perceived as
loudness is directly related to the amount of pressure generated by the vibrating object. In a
compressible medium, the motion of molecules produces positive pressure where there is
condensation and negative pressure where there is rarefaction of molecules. The intervals of
condensation and rarefaction typically occur in a cyclical fashion. In a plane progressive wave
of sound (when the acoustic pressure is the same in all planes perpendicular to the direction of
propagation), the instantaneous pressure, p, generated in a compressible fluid can be described
by
p = ρcu (1)
where ρ equals the fluid density, c equals the speed of sound, and u equals the particle velocity.
Acoustic pressure is typically measured as the root-mean-square (RMS) pressure average over
the duration of the sound. For impulsive sound such as pile driving strikes or biosonar clicks,
peak sound pressure (the range from zero to the greatest pressure of the signal) or peak-to-peak
sound pressure (the range of the most positive to the most negative pressure of the signal) are
often reported instead, since it is difficult to define an appropriate duration over which to average
the signal’s pressure (Madsen 2005). Pressure is typically reported in units of pascals (Pa) or
micropascals (µPa). In a plane progressive wave, sound intensity is described by the sound
power per unit area and is a product of the sound pressure and particle velocity by
I = pu (2)
and substituting u from first equation, intensity of the sound, I, is related to p by
I = p(p/ρc) = p2/ρc (3)
where p is the RMS pressure average over the duration of the sound. Intensity is typically
reported in units of watts per square meter. Sound levels are most often described in units of
decibel (dB), which is traditionally defined as a power or intensity ratio. Sound intensity level in
decibels is as follows:
dB = 10 log10 (I1/I2) (4)
where I1 is the intensity of the sound of interest and I2 is a reference intensity. In the case of a
plane wave, sound pressure which is typically what is measured by a microphone or hydrophone
may also be used to measure the sound’s magnitude in dB. Because sound intensity is
proportional to pressure squared, sound pressure level (SPL) in dB is given by
dB = 10 log10 (p12)/( p22) = 20 log10 (p1/ p2) (5)
where p1 is the pressure of the sound of interest and p2 is typically the standard reference pressure
for a given medium. In water the reference is usually 1 µPa. SPLs in this document are
referenced to the underwater convention (re 1 µPa) based on RMS measurements unless
otherwise noted. This reference pressure is different from the standard used to measure sound
pressure levels in air. Thus a dB (re 1 µPa) underwater is not equivalent to a dB (re 20 µPa)
measured in air. Pulsed sounds such as explosions, seismic air gun pulses, or pile driving
impacts are often measured in terms of their energy and not just pressure or intensity. Energy
measures include time as a dimension and are also used to quantify sound exposure when both
amplitude and duration of exposure is important. Energy is proportional to the time integral of
the pressure squared and in dB sound exposure levels (SELs) has the units of dB re 1 µPa2s.
Frequency Bandwidth
Amplitude, intensity, or energy measurements in dB are always dependent on the
measurements integrated across a frequency bandwidth. Broadband SPL measurements (overall
SPL) will be different from those based on one octave, one-third octave, narrower band, and
spectral density level measurements. For example, sound pressure spectral density gives the
mean squared pressure of a sound within a given frequency bandwidth divided by the
measurement bandwidth and the units are pressure square per hertz. The decibel unit based on
spectral density level is dB re 1 µPa2 /Hz (or re 1 µPa /Hz1/2). On the other hand, one-third
octave levels in dB are based on the mean square pressure level for each one-third octave band.
2
An octave is a factor of two in frequency and sound levels are often reported in one-third
octave bands because the effective filter bandwidth of the auditory system in humans and some
animals is approximated by one-third octave (Richardson et al. 1995, but see the Auditory
Capabilities section and Auditory Masking subsection below). The term noise in this document
usually refers to the general definition of unwanted sound. For a further review of these
concepts, see Urick (1983), Richardson et al. (1995), and the Web site http://www.dosits.org/
science/intro.htm, hosted by the University of Rhode Island.
3
Southern Resident Killer Whales
Killer whales are the largest cetacean in the dolphin family, Delphinidae. The three
identified ecotypes of killer whales in the northeastern Pacific Ocean are residents, transients,
and offshores. While there is considerable overlap in their geographic range, these ecotypes are
genetically distinct and do not appear to interbreed. The differences between ecotypes also
extend to their morphology, foraging ecology, behavior, and acoustic repertoire. For example,
residents are generally fish eaters while transients are generally mammal eaters (Ford et al. 2000)
and less is known about the diet of offshores. Residents tend to live in larger, more stable groups
consisting of multigenerational, matrilineal-related kin, while transients live in smaller, less
stable groups usually consisting of females and a few offspring (Ford et al. 2000). Residents
tend to be more vocal, particularly when foraging and socializing, while transients are
acoustically cryptic presumably because their prey can hear within the frequency range of their
sound emissions (Barrett-Lennard et al. 1996, Deecke et al. 2002, Deecke et al. 2004).
Along the U.S. and Canadian west coast, there are currently four communities of resident
killer whales that have been identified: Northern, Southern, Southern Alaska, and Western
Alaska Residents (Krahn et al. 2004). SRKWs consist of three pods—J, K, and L pod—found
during the late spring to early autumn in the inland waters of Washington State and British
Columbia. Members are individually identified based on photo identification records of natural
markings. Like all marine mammals, they are long-lived and slow to mature.
Both male and female resident killer whales of the area do not become sexually mature
until an average age of 15 years and females produce an average of 5.5 surviving offspring
(Olesiuk et al. 1990). Annual population censuses indicate that SRKW numbers experienced a
population decline in the mid to late 1990s. This distinct population segment of killer whales has
been listed under the Endangered Species Act of 1973 (ESA) as endangered (NMFS 2005b) with
three factors believed to be related to their decline: food availability, contaminant loads, and
vessel and noise interactions (Krahn et al. 2004).
As part of the ESA listing, NOAA’s National Marine Fisheries Service (NMFS) is
responsible for designating critical habitat for SRKWs. As shown in Figure 1, this includes the
following areas but excludes spaces around U.S. military sites: 1) core summer area of U.S.
waters up to the border of Canada and surrounding the San Juan Islands, 2) Puget Sound
excluding Hood Canal, and 3) U.S. waters within the Strait of Juan de Fuca (NMFS 2006). The
following two sections will review information on the use of sound by killer whales in general
and the acoustic scene of the areas defined as critical habitat for SRKWs in particular.
4
Figure 1. Critical habitat designation for Southern Resident killer whales (NMFS 2006).
5
Killer Whale Sound Production and Function
Killer whales produce a wide variety of clicks, whistles, and pulsed calls (Schevill and
Watkins 1966, Ford 1989, Thomsen et al. 2001). Clicks are echolocation signals that are
produced individually or in click trains. Individual clicks produced by Northern Resident killer
whales are relatively broadband, short (0.1–25 milliseconds [ms]), and range in frequency from 8
to 80 kHz with an average center frequency of 50 kHz and an average bandwidth of 40 kHz (Au
et al. 2004). These broadband biosonar signals are predominantly used for sensing objects such
as prey in the environment and are produced by whales foraging on salmon at peak-to-peak
source levels ranging from 195 to 225 dB re 1 µPa at 1 m (Au et al. 2004). There may be a
considerable amount of variation in source level and frequency content of killer whale biosonar
emissions depending on prey type. For example, Norwegian killer whales feeding on herring
produced echolocation clicks that were lower in mean center frequency and source level,
although there was some overlap in the range of these biosonar parameters (Simon et al. 2007).
SRKWs also feed on salmon and it is likely that their biosonar emissions are similar to those of
Northern Residents.
Whistles are tonal, nonpulsed signals that are relatively longer in duration (0.06–18
seconds [s]) and lower in frequency (0.5–10.2 kHz, Thomsen et al. 2001). Whistles are most
often heard during close-range social activities but not as often during foraging and traveling
(Thomsen et al. 2002). Whistles produced by Northern Residents typically have source levels
ranging from 133 to 147 dB re 1 µPa at 1 m (Miller 2006).
Pulsed calls are the most commonly observed type of sounds emitted by killer whales and
are categorized as discrete (stereotyped), variable, or aberrant (Ford 1989). These calls produced
by both Northern and Southern residents are relatively long (600–2,000 ms), appear
harmonically rich, and range in frequency between 1 and 10 kHz; but, those with high frequency
components may contain harmonics up to 30 kHz (Ford 1989, Miller 2002). Variable calls are
produced at source levels ranging from 133 to 165 dB while stereotyped calls are produced at
source levels ranging from 135 to 168 dB re 1 µPa at 1 m (Miller 2006). These calls are most
often used when killer whales are foraging and traveling and likely function to maintain social
cohesion among pod members in the absence of other sensory information (Ford 1989, Miller
2002).
Killer whale communities have different discrete call repertoires, both among and
between ecotypes (Ford 1991, Ford and Ellis 1999). Discrete calls are highly stereotyped and
repetitive, and have pod-specific qualities that are stable over time (Ford 1989, Foote 2005). For
example, each of the three pods among the Southern Resident population uses one or two
discrete call types more than 50% of the time (Ford 1989, Foote 2005).
Given the biological significance of sound for biosonar and social functions in killer
whales, it is important to address how ambient noise, from a variety of sources acting alone or in
6
combination with other threats, might impact the population structure of SRKWs. In order to
assess noise impacts on this population, it is necessary to consider several parameters of sound
that are pertinent to the auditory capabilities of these whales. Such factors include amplitude,
duration, and spectral characteristics of the source, as well as how it propagates in the
underwater environment.
7
Characterizing Sound and Propagation
Ambient noise is essentially the background din. Several sources of sound contribute to
ambient noise levels in the ocean and coastal marine environments (for review, see Richardson et
al. 1995). Natural sounds include those produced from abiotic and biotic sources. Abiotic
sounds arise from activities related to weather (i.e., wind, waves, and rain), seismic activity,
underwater slides, and currents. Some of these sources can substantially increase ambient noise
levels. For example, heavy precipitation can raise levels from a few to 20 dB re 1µPa2/Hz or
more between 1 and 20 kHz (Wenz 1962, Nystuen et al. 1993). Biotic sources of noise include
those produced by marine mammals, fish, and snapping shrimp (Alpheus and Synalpheus spp.).
For example, ambient noise in some areas is dominated by snapping shrimp in frequencies that
overlap with those of echolocating dolphins (Au et al. 1985).
Anthropogenic sounds that contribute to overall ambient noise include active sources
(e.g., air guns, sonars) and those that are by-products of various human activities in the ocean.
The latter include oil drilling, construction, and vessel traffic. Active sources include those used
for military tactics, seismic surveys, fisheries, and oceanographic research. In the areas proposed
as SRKW critical habitat, almost all of these sources of underwater sound are present.
Sound Propagation Variables
Sound propagated in sea water is subject to a number of variables that degrade the
signal’s amplitude over distance and thus affect potential received levels. The transmission loss
of sound occurs due to two primary factors: spreading and attenuation. If a source is allowed to
spread in all directions equally, then a spherical spreading law may be used to describe
transmission loss. This might be an appropriate assumption if a source emanates in deep water
in the open ocean. However, even oceanographic features of the water column can create less
than spherical spreading conditions. This is because the speed of sound varies based on
temperature, salinity, and pressure (depth).
Sound will also be attenuated as it travels over a distance due to absorption of sound by
sea water and scattering due to particles. Sound is also reflected at the sea surface and from the
bottom, creating instances of constructive or destructive interference. The details of spreading
and attenuation of underwater sound are reviewed in Urick (1983). In shallow water
environments such as those inhabited by SRKWs, the parameters that affect sound propagation
can vary considerably in space and time. The acoustic environment in shallow water, therefore,
is often described as complex.
Ambient Noise
Several investigations have reported ambient noise measurements in the areas designated
as SRKW critical habitat and adjacent Canadian waters known to be important SRKW habitat.
8
Many of these studies have focused on describing sound levels generated by vessel propulsion.
Noise from vessel propulsion is usually generated by propeller cavitation (producing broadband
noise), propeller “singing” (producing tonal sounds and harmonics related to the propeller blade
rate), and auxiliary machinery (such as pumps and rotating shafts) with the amount of noise
related to vessel size, speed, and mode of operation (Richardson et al. 1995). Data from the
following peer-reviewed articles and government reports* are examined in this section: Veirs and
Veirs (2005), Nystuen (2006), Erbe (2002), Hildebrand et al. (2006), Laughlin (2005), and Jones
and Wolfson (2006).
Veirs and Veirs (2005) report average sound pressure levels (SPL dB re 1µPa measured
from 0.1 to 15 kHz) and power spectra (converted to noise spectral density levels in dB re
1µPa2/Hz) of ambient sound recorded on the west side of San Juan Island in Haro Strait (lat
48°33′25″ N, long 123°10′23″ W). All measurements were reported as received levels between
0.1 and 15 kHz, thus propagation loss was not determined from source to receiver. The
frequency range of these measurements is pertinent for the lower frequency range of killer whale
hearing relevant to the reception of communication calls (see the Auditory Capabilities section
below). Measurements were made on hydrophones that were cross-calibrated with a calibrated
hydrophone and projector rented from the Naval Undersea Warfare Center in Newport, Rhode
Island.
Veirs and Veirs (2005) reported average SPLs over hourly, daily, and monthly time
periods from April 2004 to November 2005. Half-hourly averaged SPLs ranged from 95 to 130
dB with an overall half-hour SPL average of 115 dB. Broadband ambient sound levels were
highly influenced by large vessels, such as commercial ships, that increased SPLs between 20
and 25 dB over a 10–30 minute period and to a lesser extent by smaller vessels, such as motor
boats, that increased SPLs by 15–20 dB (Veirs and Veirs 2005). During the summer (July and
August), these smaller vessels contributed more to the overall ambient levels during the day,
raising hourly SPL averages by 2–4 dB compared to nonsummer daytime hours. As a
consequence, there was a more pronounced diurnal pattern in the summer but it only amounted
to about a 2 dB difference in 12-hour averaged SPLs.
Nighttime SPLs did not exhibit much seasonal variation, since larger ships operate during
all months of the year. Monthly averaged SPLs ranged from 114.5 to 117.5 dB and were
generally lowest from November through April and highest from June through August.
Cumulative distributions of SPLs were also provided from archived measurements taken as 2-
second averages (Figure 2). These distributions illustrate that 2-second SPL averages are greater
than or equal to 120 dB 50% of the time during summer days, 30% of the time during summer
nights, and 20% of the time during the winter (Veirs and Veirs 2005).
* Government research reports are based on work contracted through the Northwest Fisheries Science Center Marine
Mammal Program and go through a limited peer review process. These reports are online at http://www.nwfsc.noaa
.gov/research/divisions/cbd/marine_mammal/research.cfm or can be requested by contacting the report author(s).
9
0
10
20
30
40
50
60
70
80
90
100
80 90 100 110 120 130 140
Cumulat i ve
%
Cumulative %
Summer Day
SPL (dB re 1μPa)
Summer Night
Winter Averageage
Summer day
Summer night
Winter average
SPL (dB re 1µPa)
Figure 2. Cumulative distributions of 2-second average SPLs (0.1–15 kHz). Summer day includes SPL
measurements during the hours of 0800–2000 and summer night includes SPL measurements
from the other 12 of the 24 hour cycle (based on Figure 6 of Veirs and Veirs 2005).
Figure 2. Cumulative distributions of 2-second average SPLs (0.1–15 kHz). Summer day includes SPL
measurements during the hours of 0800–2000 and summer night includes SPL measurements
from the other 12 of the 24 hour cycle (based on Figure 6 of Veirs and Veirs 2005).
Ambient noise levels were further described by frequency and showed similar trends for
ambient levels recorded underwater elsewhere (Wenz 1962). For both July and November 2005,
the highest spectrum levels occurred for the lowest frequencies, with peaks of about 82 dB re
1µPa2/Hz occurring between 400 and 500 Hz, then generally decreased at a rate of 5 dB for each
doubling of frequency (i.e., per octave) from 500 Hz to 5 kHz and at a rate of 6 dB for each
doubling of frequency from 5 to 20 kHz (Figure 3). Between 1 and 20 kHz, there was a more
pronounced difference in spectral levels (up to 5 dB re 1µPa2/Hz) between daytime and
nighttime hours in July compared to November. Below 1 kHz, very little seasonal variation in
spectral levels was observed (Veirs and Veirs 2005).
Ambient noise levels were further described by frequency and showed similar trends for
ambient levels recorded underwater elsewhere (Wenz 1962). For both July and November 2005,
the highest spectrum levels occurred for the lowest frequencies, with peaks of about 82 dB re
1µPa2/Hz occurring between 400 and 500 Hz, then generally decreased at a rate of 5 dB for each
doubling of frequency (i.e., per octave) from 500 Hz to 5 kHz and at a rate of 6 dB for each
doubling of frequency from 5 to 20 kHz (Figure 3). Between 1 and 20 kHz, there was a more
pronounced difference in spectral levels (up to 5 dB re 1µPa2/Hz) between daytime and
nighttime hours in July compared to November. Below 1 kHz, very little seasonal variation in
spectral levels was observed (Veirs and Veirs 2005).
Ambient sounds were also recorded from acoustic moorings called passive aquatic
listeners (PALs) during the late spring and summer of 2005 from two areas off Cape Flattery,
Washington (Nystuen 2006). Mean sound spectra from these moorings showed that close ships
dominated the sound field below 10 kHz while rain and drizzle were the dominant sound sources
above 20 kHz during a week in April 2005. As the dominant sound source, relative spectral
density levels of close ships decreased with increasing frequency between 1 and 10 kHz while
sound from rain had a slighter decrease with increasing frequency for spectral density levels
between 20 and 50 kHz (Nystuen 2006). Furthermore at this location, shipping noise dominated
the sound field approximately 10% to 30% of the time but was dependent on weather; that is,
when the weather was poor, less shipping noise was present. From April to July the shipping
activity increased while rain activity, which dominated the sound field from 1 to 10% of the
time, decreased (Nystuen 2006).
Ambient sounds were also recorded from acoustic moorings called passive aquatic
listeners (PALs) during the late spring and summer of 2005 from two areas off Cape Flattery,
Washington (Nystuen 2006). Mean sound spectra from these moorings showed that close ships
dominated the sound field below 10 kHz while rain and drizzle were the dominant sound sources
above 20 kHz during a week in April 2005. As the dominant sound source, relative spectral
density levels of close ships decreased with increasing frequency between 1 and 10 kHz while
sound from rain had a slighter decrease with increasing frequency for spectral density levels
between 20 and 50 kHz (Nystuen 2006). Furthermore at this location, shipping noise dominated
the sound field approximately 10% to 30% of the time but was dependent on weather; that is,
when the weather was poor, less shipping noise was present. From April to July the shipping
activity increased while rain activity, which dominated the sound field from 1 to 10% of the
time, decreased (Nystuen 2006).
10
a)
40
45
50
55
60
65
70
75
80
85
90
100 1000 10 10000000 0
Frequency (Hz)
2
dB re 1μPa/Hz
1/
dB re 1µPa/Hz
Noon
Daily
Monthly
Nightly
Midnight
100 1,000 10,000 100,000
Frequency (Hz)
b)
40
45
50
55
60
65
70
75
80
85
90
100 1000 10000 100000
Frequency (Hz)
dB re 1μPa/Hz
1/2
Noon
Daily
Monthly
Nightly
Midnight
dB re 1µPa/Hz
100 1,000 10,000 100,000
Frequency (Hz)
Figure 3. Spectrum levels in Haro Strait in a) July 2005 and b) November 2005 (based on Figure 7 and
Figure 8 in Veirs and Veirs 2005).
11
Whale Watching Sound Propagation
Erbe (2002) reported sound level measurements in the presence and absence of whale
watching vessels in Brotchi Ledge outside of Victoria Harbor (Canada) and in Haro Strait on the
west side of San Juan Island (United States). Sound levels of whale watching inflatable and
noninflatable motorboats were based on recordings made between 0.1 and 20 kHz and lasting
10–15 seconds. Vessel source levels were estimated using propagation models based on ray
theory which included variables based on sound speed profiles, absorption loss by the sediment,
and frequency-dependent absorption by ocean water (Erbe 2002).
In general, inflatables with larger engines had higher source levels, ones with stern drives
were considerably quieter than those with outboards, and inflatables were slightly louder than
motorboats with inboards or stern drives. The relationship between source level and vessel
speed was nonmonotonic and attributed to vessel speed measurement error of the radar gun (Erbe
2002). At speeds of approximately 50 km/hour, average broadband source levels measured
between 0.1 and 20 kHz were 162 and 159 dB re 1 µPa at 1 m for inflatables and motorboats,
respectively (Erbe 2002). Noise levels were higher by about 15–27 dB re 1 µPa (based on one-
twelfth octave band level analysis) in the presence of about five vessels (mix of inflatables and
motorboats operating within 400 m of the hydrophone) compared to those measured in their
absence (at a sea state of one-half, Erbe 2002).
Hildebrand et al. (2006) also reported source level measurements (as spectra in dB re 1
µPa2/Hz at 1 m) for a variety of vessels of different sizes, propulsion systems, and operational
speeds in Haro Strait. Measurements include those made from a calibrated broadband recording
system (1–75 kHz). The extended frequency range of the vessel noise is pertinent for the higher
frequency range of killer whale hearing relevant to the reception of echolocation signals (see the
Auditory Capabilities section below). Received levels were measured at ranges that varied
between 125 and 442 m for idle, normal cruise speed (17–31 knots), and power acceleration to
full speed. Source level spectra for 10-second averages were calculated at 1 m assuming
spherical spreading loss and using absorption coefficients from the equation of Ainslie and
McColm (1998).
For power up conditions, the 10-second interval was chosen to include the highest sound
pressure level. Noise spectra were not sufficiently above background levels to make accurate
measurements of vessel noise levels under controlled conditions while vessels were idle despite
efforts to collect data under conditions where no other vessels in the local vicinity were present.
Thus data collected under idle conditions were not considered to be accurate measurements of
real vessel noise (Hildebrand et al. 2006).
Noise source spectra from four whale watching vessels reported by Hildebrand et al.
(2006) are summarized here. Boat A is a 28-foot fiberglass monohull with twin 90-horsepower,
four-stroke outboard motors. Only the power up measurements produced spectral levels that
were sufficiently above ambient noise levels. Under these traveling conditions, source spectral
levels were approximately 118 dB re 1 µPa2/Hz at 1 m between 4 and 15 kHz, then generally
decreased at a rate of approximately 5 dB per octave up to 50 kHz. Another small whale
watching vessel, Boat B with a 29-foot aluminum monohull and twin 225-horsepower outboard
motors, produced source levels at a cruise speed of 24 knots that were significantly above
12
ambient conditions between 100 and 300 Hz and between 6 and 75 kHz. Another medium-sized
whale watching vessel, Boat C, is a 38-foot aluminum catamaran with jet drives. While cruising
at a speed of 31 knots, Boat C produced source spectral levels that were generally lower than the
smaller vessel with no measurable rotating equipment noise at the lower frequencies. The
difference between idle and power acceleration noise was less than 10 dB.
Source levels for a larger whale watching vessel, Boat D, a 50-foot monohull vessel with
three inboard/outboard (stern) drives, produced the highest source spectral levels that generally
ranged between 110 and 145 dB re 1 µPa2/Hz at 1 m between 60 and 1,000 Hz, with higher
peaks at some frequencies in this range while idling. At the same speed, this vessel produced
source levels that ranged between 115 and 110 dB re 1 µPa2/Hz at 1 m between 1 and 5 kHz,
then generally decreased at a rate of 5 dB/octave between 5 and 75 kHz. When Boat D was
operating at a cruise speed of 23 knots, source spectral levels generally ranged between 125 and
145 dB re 1 µPa2/Hz at 1 m between 60 and 1,000 Hz, with higher peaks at some frequencies in
this range as well. At the same speed, this vessel produced source spectral levels of about 125
dB re 1 µPa2/Hz at 1 m between 1 and 20 kHz, then generally decreased at a rate of
approximately 10 dB/octave up to 75 kHz.
When Boat D was powering to cruise, source spectral levels generally ranged between
122 and 145 dB re 1 µPa2/Hz at 1 m between 60 and 1,000 Hz, with the highest peak of 148 dB
re 1 µPa2/Hz at 1 m between 100 and 200 Hz. At the same speed, Boat D produced source
spectral levels that ranged between 120 and 128 dB re 1 µPa2/Hz at 1 m between 1 and 20 kHz,
then generally decreased at a rate of approximately 10 dB/octave up to 60 kHz (Hildebrand et al.
2006).
Hildebrand et al. (2006) also opportunistically measured the source spectral levels of a
290 m long Korean container ship, the MV Hanjin Marseilles. Source spectral levels ranged
from 135 to 165 dB re 1 µPa2/Hz at 1 m between 60 and 1,000 Hz when measured from the low
frequency hydrophone and showed a more or less linear decrease in level of approximately 10
dB/octave across this frequency range. At 442 m from the recording equipment, the received
spectrum levels of the container ship ranged between 42 and 110 dB re 1 µPa2/Hz between 0.06
and 75 kHz. Even at the highest frequencies, received levels were approximately 20 dB above
ambient levels (that were measured on a different day) as shown in Figure 4 (Hildebrand et al.
2006).
A comparison of source spectral levels for all vessels operating at cruise speed is shown
in Figure 5 (Hildebrand et al. 2006). Above 2 kHz, the largest whale watching vessel, Boat D,
produced the highest source spectral levels, while Boat C produced the lowest source spectral
levels. Boat C presumably produced the lowest noise levels at higher frequencies because of its
jet drive system. The MV Hanjin Marseilles produced surprisingly significant levels of noise
above 2 kHz that, with the exception of Boat D, were higher in level compared to other whale
watching vessels.
From the results of Hildebrand et al. (2006) and Erbe (2002), noise generated by whale
watching vessels as well as other vessel types is dependent on a combination of size, engine type,
and operating speed.
13
Pressure spectrum level dB re 1uPa
2
/Hz
Ambient 5/31 at 12:30
MV Hanjin Marseilles at 442 m
Ambient 5/31 at 1:00
Frequency (Hz)
Figure 4. Received spectral levels of a container ship (MV Hanjin Marseilles) traveling at 21 knots as it
passed 442 m from the recording equipment. Ambient conditions the day after the Hanjin passed
in the Haro Strait as well as average sea states of 1 and 3 in deep water reported by Urick (1983)
are also shown for comparison (adapted from Hildebrand et al. 2006).
Other Anthropogenic Sound Sources
While most studies thus far reviewed have focused on vessel sounds, other sources of
anthropogenic sounds in SRKW habitat include those associated with marine construction such
as pile driving and dredging. Sound exposure from pile driving has received particular attention
given the loud impulsive (transient with rapid rise time) nature of such construction activity. As
with other types of sounds, exposure from pile driving will depend on a number of factors
including the size and type of pile driving equipment, the intensity of the driving activity, its duty
cycle, and the local environment. Sound level measurements were made of pile driving activity
associated with the restoration of the Friday Harbor ferry terminal. Average peak levels ranged
between 180 and 215 dB re 1 µPa between 0 and 10 kHz (average RMS levels: 166–196 dB re 1
µPa, average SELs: 171–187 dB re 1 µPa2s) that varied depending on the condition (Lauglin
2005).
In some cases, mitigation measures are taken to reduce the overall level of construction
noise introduced into the environment. For example, air bubble curtains, in which a perforated
air hose is placed outside the pile, generates bubbles as air is passed through the hose. Bubble
curtain designs include single, dual, multiple ring, and tree curtains. In ideal cases, the bubble
curtain effectively absorbs and reduces the magnitude of the pressure waves generated from
14
Boat A
Boat B
Boat C
Boat D
Ship
Boat D
Ship
Boat B
Boat A
Boat C
Figure 5. Spectrum source levels of all vessels at cruise speeds recorded on the broadband hydrophone
system (adapted from Hildebrand et al. 2006).
driving the pile into the ground (Wursig et al. 2000). However, many factors can influence the
effectiveness of the bubble curtain such as water currents, bathymetry, and tide levels. The
average reduction in sound levels from multiple ring bubble curtain deployment during pile
driving at the Friday Harbor ferry terminal ranged between 1 and 3 dB and the maximum
reduction was 16 dB (Laughlin 2005). The results depended on the individual piles driven, the
extent of the bubble curtain, and external factors such as ambient noise, given that these tests
were not conducted in controlled conditions (Laughlin 2005).
Modeling Sound Propagation
Because many areas frequented by SRKWs are complex shallow water acoustic
environments, it is necessary to consider how transmission loss from source to receiver will
affect the sound levels potentially reaching killer whales in the area. Jones and Wolfson (2005)
modeled propagation of sound emanating from large vessels in the main shipping lanes in the
Haro Strait, focusing on numerical estimates in the open channel. Predicted values of received
sound levels were compared to field measurements made by PALs collected in spring and
summer 2004 (Nystuen 2006). Information on the position, course, and speed of large vessels
(≥ 20 m in length) was obtained from the Vessel Traffic Operations Support System (VTOSS) of
the Canadian Coast Guard Services. Information for 27 May–30 June 2004 from VTOSS
illustrated that 24% of the vessels larger than 20 m were bulk carriers, 15% were tugs, 13% were
15
container ships, 10% were ferries, 9% were fishing vessels, and a smaller percentage were a
variety of other vessels types.
Correlation of specific acoustic recordings with data from the VTOSS database provided
information about the sound propagation of ships based on type, location, speed, and orientation.
For example, when a large cargo ship passed to the side of the PAL mooring at a range of 1 km,
received levels were approximately 85 and 75 dB (re 1µPa) at 3.6 and 10.4 kHz, respectively. In
contrast, when the same ship passed to the side of the PAL mooring at a range of 4–5 km, the
spectral levels of the same frequencies were about 20 dB lower.
In Jones and Wolfson (2006), modeling and analysis were focused at the frequency of 3.6
kHz to represent the frequency ranges between 1 and 10 kHz. Acoustic propagation modeling
was described using two-dimensional parabolic equation numerical methods in which only
reverberation in the forward direction was considered while backscattering was neglected (Jones
and Wolfson 2006). This model is especially vulnerable to uncertainties in bathymetry, sea
surface roughness, bottom substrate, and variables affecting sound speed profiles.
Characteristics of these parameters along with appropriate assumptions were incorporated
in the model for analysis. Sound speed profiles obtained from conductivity, temperature, and
depth measurements collected in May and June in 1990 to 2002 illustrated relatively little
variation in sound speed with depth and locations in the Haro Strait (Jones and Wolfson 2006).
Model simulations were performed for a single ship at 10 ship positions, each position separated
by 1 km as it passed by the stationary mooring. The model was run with three different
categories of sediment type (smoother sand/mud sediment, rough sand/mud sediment, rough
rock/sand sediment) and three levels of wind speed/sea surface roughness (0, 5, and 10 m/s).
In all cases, the model results were consistent with the measured results when an
estimated ship source level of 175 dB at 3.6 kHz was used. However, the results for the different
cases of sediment and sea surface conditions varied within 3 dB of each other, implying these
parameters do not strongly influence model results (Jones and Wolfson 2005). Furthermore,
variability of the model results did not increase with increased sea surface roughness. The
investigators hypothesize that the short distance between ship and receiver results in bathymetric
effects that dominate the results while bottom and surface conditions contribute little to
propagation effects.
Model results were also obtained for the average positions of ships traveling in the
northern shipping lanes of the Haro Strait using VTOSS database information. The closest
distance to the mooring in this situation was approximately 2 km while the farthest was
approximately 7 km. Predicted received levels at the farthest point were lower by about 15 dB at
3.6 kHz compared to the closest point based on this model. Given the complexity of the acoustic
environment of the Haro Strait, the report recommends that acoustic modeling is best used as a
complement to field measurements (Jones and Wolfson 2006). Other recommendations include
correlating seasonal and regional measurements of background sound levels with records of
killer whale locations and direct measurements of broadband source levels of large ships in the
main shipping lanes.
16
To summarize, most measurements of ambient sounds levels in SRKW habitat are
strongly influenced by vessel traffic that can reach broadband received levels of up to 130 dB
and source levels that average approximately 160 dB (Erbe 2002, Veirs and Veirs 2005). In
addition various vessels, including those used for whale watching activities as well as larger
ships, can raise noise levels significantly above ambient levels even at frequencies up to 75 kHz
(Hildebrand et al. 2006). Furthermore, data model comparisons illustrate that propagation
models may accurately predict levels of vessel noise potentially received by killer whales at a
particular location. However, given the complex shallow water environment, models should be
used only to complement and cannot substitute for field measurements.
17
Auditory Capabilities and Auditory Effects
of Sound Exposure
The reports summarized in the previous section focus on describing sounds in SRKW
core summer habitats. In order to assess noise impacts, it is necessary also to consider several
parameters of sound that are pertinent to the auditory capabilities of the whales. Such factors
include hearing sensitivity of sound based on amplitude, duration and temporal factors,
frequency, and how exposure of sound affects baseline sensitivity and other auditory functions.
This section reviews what is known about the hearing capabilities and auditory effects of sound
exposure in killer whales. Because of the limitations associated with time and cost of conducting
laboratory research on marine mammals, there is a dearth of information about some aspects of
auditory capabilities in killer whales, particularly in terms of auditory effects of noise. A number
of studies have investigated such effects on smaller delphinids and appropriate information will
also be reviewed to supplement current knowledge in this area.
Audiograms and Basic Auditory Function
There are several experimental approaches that have been used to determine the auditory
capabilities of marine mammals including killer whales. A basic assessment of hearing is to
determine sensitivity as a function of frequency in order to plot an audiogram. One of the most
direct approaches to measure this involves behaviorally training subjects to respond in a
particular way when they hear a particular sound and plot performance as a function of the
sound’s amplitude (behavioral psychophysics). Another approach involves electrophysiological
measurements of hearing sensitivity, which requires measuring auditory brainstem responses
(ABRs) as a function of the sound’s amplitude. One advantage of the latter approach is that it
typically takes less time to measure a threshold, which can be desirable in some instances, for
example, when hearing recovery functions are measured after noise exposure.
Both of these approaches have experimental limitations that should be noted. For
example, not all listening environments are the same and ambient noise levels of the testing area
should be measured and noted to avoid “noise-limited” measurements of hearing sensitivity.
Additionally, the temporal summation (or temporal integration) properties of the auditory system
influence sensitivity to a sound. Sounds shorter than some critical value or time constant are
generally less detectable than longer signals; that is, sensitivity typically decreases nonlinearly
for signals shorter than the time constant. Time constants for tonal stimuli in mammals typically
vary between 30 and 800 ms and appear to be relatively consistent between several taxa (Fay
1988).
Although time constants have not been determined for killer whales, Johnson (1968)
measured the dependence of thresholds on signal duration in a bottlenose dolphin (Tursiops
truncatus) and reported time constants that ranged from 220 ms at 4 kHz to 30 ms at 45 kHz.
18
Johnson (1991) also measured time constants in a beluga whale (Delphinapterus leucas) for a 60
Hz single tone pulse and reported a time constant of 20 ms. The lower frequency results are
consistent with those measured in humans (Plomp and Bouman 1959). Assuming that the
temporal summation properties of killer whales follows the typical delphinid pattern, thresholds
measured with signals shorter than these time constants will not be directly comparable to those
measured with longer signals. For example, thresholds measured by electrophysiological
methods are orders of magnitude shorter (and often below these time constants) than those
measured by behavioral methods.
Hearing Sensitivity
In a behavioral psychophysical setting, hearing sensitivity is usually defined as a
threshold. The concept of a threshold is a statistical probability of an animal detecting the sound
at a particular level some percentage of the time (often 50 or 75% correct depending on the
procedure). The physical parameters of the stimulus are not the only variables that affect
threshold measurements of a subject. Nonsensory cues such as reinforcement contingencies and
expectancies can significantly affect the measured threshold. For example, in most sound
detection tasks, the signal level is reduced beyond some detection point and signal-present and
signal-absent trials are both often used to keep the animal “honest” in reporting the presence of a
signal.
If an animal experiences punishment for responding to a signal when no signal is present,
then the animal might become more conservative in reporting that it heard the signal. That is,
the individual might adopt a tendency to report “no” when uncertain. This would lead to a
higher measured hearing threshold or perhaps an underestimate of hearing sensitivity. For these
reasons, it is necessary to account for animal response biases during sensory tasks (for further
discussion see Schusterman 1974, Renouf 1991). This type of response behavior is not only
important to consider in terms of how animals perform in laboratory settings but also how
responses of free-ranging animals might be affected by different motivational states. For
example, if there is disturbance in a unique feeding habitat where food is seasonally limited,
animals may tolerate exposure because they are motivated to feed and may not have suitable
alternatives.
Another consideration of lab-based results is that very few individuals are typically
tested. As with any population, auditory capabilities are expected to show individual variation
based on factors including genetics, age, sex, and exposure to pathogens, ototoxic drugs, and
noise. For example, in humans and presumably other mammals, high-frequency hearing loss or
presbycusis is common with age (Yost 2000). Thus it is expected that a larger number of older
animals would be less sensitive to higher frequencies than younger animals.
Two laboratory studies on the hearing sensitivity of captive killer whales have been
conducted (Hall and Johnson 1972, Szymanski et al. 1999). One subadult male was tested in the
earlier study and an audiogram was obtained for 8-second pure tones ranging in frequency
between 500 Hz and 31 kHz. Greatest sensitivity or a lowest threshold of 30 dB re 1 µPa for this
individual occurred at 15 kHz, the observed upper limit of hearing sensitivity was 32 kHz,
thresholds below 10 kHz were probably noise limited, and this individual was suspected to have
high-frequency hearing loss (Hall and Johnson 1972). Both behavioral and electrophysiological
19
thresholds were obtained in two adult females in a later study (Szymanski et al. 1999). These
individuals had much higher and more plausible upper frequency hearing limits of about 120
kHz compared to other delphinids than the individual in the Hall and Johnson (1972) study.
Both behavioral and electrophysiological audiograms appeared similar but behavioral
thresholds were typical lower (Figure 6). However, electrophysiological thresholds were
measured using much shorter signals (1 or 0.5 ms durations) than those for behavioral thresholds
(2 s). Furthermore, the investigators reported electrophysiological thresholds based on dB peak-
to-peak measurements and behavioral thresholds based on dB RMS measurements. If
comparable dB measurements were computed, then the average differences of thresholds
between the two methods would be reduced. The frequency of best sensitivity averaged for both
subjects was 20 kHz and range of best sensitivity (±10 dB from lowest threshold) was 18–42
kHz (Szymanski et al. 1999).
One caveat of this study was that behavioral thresholds might be underestimates of
hearing sensitivity because the subjects were reinforced after missing a signal if they waited for a
recall tone and false alarm rates were reportedly so low that it is possible that the animals were
trained to conservatively report the signal. Despite these experimental caveats, the audiogram
based on Szymanski et al. (1999) is the best available data to determine hearing sensitivity of
killer whales in quiet conditions.
20
30
40
50
60
70
80
90
100
110
120
Behavioral - dB RMS
Physiological - dB peak-to-peak
Threshold (dB re: 1 µPa)
Threshold (dB re 1 µPa)
100 1000 1 1 10000 00000 000000
Frequency
100 1,000 1 1 1,0,000 00,000 000,000
Frequency (Hz)
Figure 6. Behavioral and physiological (ABR) audiograms based on averaged thresholds for two female
killer whales as reported by Szymanski et al. 1999. (Reprinted from Figure 5c in Szymanski et al.,
copyright 1999, with permission from Acoustical Society of America.)
20
Other Capabilities
Other auditory capabilities that are pertinent to sound exposure impacts include
directional hearing and sound localization, frequency and intensity discrimination, and loudness
perception. For example, animals must be able to localize the source in order to avoid a noxious
sound. Unfortunately, there are no studies that have measured these auditory capabilities in
killer whales, but results from studies on other delphinids are available to supplement current
knowledge. Bottlenose dolphins have very good localization abilities; they can resolve angular
separation of clicks within a fraction of a degree and pure tones between 6 and 100 kHz with a
few degrees (≤ 5º) in both the horizontal and vertical planes (Renaud and Popper 1975). This
ability is most likely related to use of echolocation for sensing objects in the underwater
environment.
Frequency discrimination involves the ability to perceive two tones as being separate in
frequency or a pure tone being of constant frequency versus frequency modulated. This is
usually measured via the difference limen (DL), which is typically defined as the difference
between a reference sound frequency and that of the sound frequency that is just perceived as
different. Because this discrimination ability is dependent on the reference value, discrimination
is proportional to the standard and defined as the just noticeable difference or relative DL as a
percentage of the standard (Richardson et al. 1995, Yost 2000). Bottlenose dolphins seem to
outperform most mammals at frequency discrimination. Behavioral tests indicate that relative
DLs are 0.21–0.81% from 2 to 130 kHz (Thompson and Herman 1975). Intensity discrimination
is also very good in bottlenose dolphins, who can detect level differences of 0.35 to 2 dB for pure
tones and clicks (reviewed in Richardson et al. 1995).
For the most part, audiograms provide information about hearing sensitivity in quiet
conditions; however, loudness perception is another important consideration of hearing and
acoustic exposure impacts, particularly for predicting behavioral reactions to sounds and onset of
auditory injury in mammals. The perception of loudness in humans is usually conducted through
loudness-matching experiments. In these tasks, subjects are asked to match the loudness of a
tone at one frequency to that of another frequency, then the results are plotted as level in dB of
the comparison tone in equal loudness level contours. In people, equal loudness contours
parallel the audiogram for SPLs but flatten out as SPL increases (Figure 7). Thus loudness
perception can be approximated but not fully predicted by an audiogram.
In determining noise exposure for people, idealized equal loudness curves have been used
to calculate weighting functions, which emphasize some frequencies (at best sensitivity) and
deemphasize other frequencies. Essentially, the weighting functions are used to filter sound for
calculating exposure risk because weighting functions improve dose-response functions in
humans (e.g., correlation between noise exposure and annoyance response, Leatherwood et al.
2002). No equal loudness curves are currently available for delphinids or other marine
mammals, so audiograms have been used to develop interim weighting function despite the fact
that the effectiveness of this approach has not been tested (Miller et al. 2005, Southall et al.
2007).
21
Comparison level
(
dB re 20
μ
Pa
)
Original standard
Comparison frequency (Hz)
Figure 7. Equal loudness contours of human subjects depicting the level of a comparison tone required to
match the perceived loudness of a 1,000 Hz tone presented at different levels (threshold, 20, 40,
60, 80, and 100 phons or dB SPL). Each of the above curves represents an equal loudness
contour based on ISO Standard 226: 2003 revision. The original 40 phon standard is labeled
(source: http://en.wikipedia.org/wiki/Equal-loudness_contour).
Auditory Masking, Critical Ratios, and Critical Bandwidths
While audiograms provide information about hearing sensitivity in quiet conditions, free-
ranging animals usually need to detect biologically important sounds in noisy environments.
There are two types of auditory effects of noise exposure that will be reviewed here. One effect
is a simultaneous or masking effect while the other effect is a residual or hearing loss effect that
can be temporary or permanent (Yost 2000). In most cases, background noise must sufficiently
overlap in frequency, level, duration, and direction of the target signal for these auditory effects
to be fully realized.
Any sound that reduces the audibility of another sound of interest has the potential to
mask that sound. In practice, a target sound needs to be a certain level above the ambient noise
in order for the target to be detected. This concept has been defined as the critical ratio (CR), or
the difference in dB of the background ambient (masking) noise spectrum level and the amount
by which a signal must exceed the background level in order to be audible. For example, a CR
of 15 dB means that a signal must be 15 dB above the spectrum level of the noise in order for it
to be detected. CRs across a number of mammal groups tend to be independent of the masker
spectrum level and tend to increase with increasing frequency (Fay 1988). A few studies have
measured CRs in the bottlenose dolphin, false killer whale (Pseudorca crassidens), and beluga
22
whale (reviewed in Richardson et al. 1995). The results from these studies indicated that
delphinid subjects have CRs ranging from 17 to 20 dB below 1 kHz to about 40 dB at
approximately 100 kHz. Killer whales have CRs that ranged from about 20 dB at 10 kHz to 40
dB at 80 kHz (Bain et al. unpublished data, cited in Bain and Dahlheim 1994).
The amount of frequency overlap between a target signal and masker is probably one of
the more important considerations of auditory masking in noisy environments. For example,
pure tones are masked by sounds at frequencies near the tone frequency. Sounds at frequencies
outside the bandwidth of the auditory filter typically do not affect audibility unless the noise
level is very high. The limit of the frequency spread of the noise in its ability to mask a signal at
a particular frequency is called the critical bandwidth. There are a number of different ways that
critical bandwidths have been estimated. An indirect way, called the equal-power method,
assumes that the signal power must be equal or exceed the total noise power in the masking band
in order for it to be audible (Fletcher 1940). According to this assumption, the masking
bandwidth (in Hz) is estimated as 10(CR/10).
More direct ways to estimate critical bandwidths include measuring masked thresholds as
a function of the bandwidth of the noise masker (band-narrowing technique) or using notched
(band-reject) noise maskers and varying the bandwidth of the notch. These direct methods
typically result in critical bandwidths that are usually but not necessarily wider than those
estimated using the equal-power model for many mammals (Sharf 1970, Au and Moore, 1990,
Southall et al. 2003a). It is now generally accepted that the indirect method using the equal-
power model has limited accuracy in estimating critical bandwidths within a given species (see
Southall et al. 2003b, Yost and Shofner 2005).
While there are no direct critical bandwidth measurements available for killer whales, a
few studies have measured critical bandwidths in other delphinids using both the band-narrowing
and notched-noise techniques. Using the band-narrowing technique, Au and Moore (1990)
reported critical bandwidths of 17, 25, and 45 kHz for pure-tone frequencies of 30, 60, and 120
kHz, respectively, in a bottlenose dolphin. Using the notched noise technique, both Lemonds et
al. (2000) and Finneran et al. (2002a) estimated much narrower critical bandwidths (reported as
equivalent rectangular bandwidths). In the former study, critical bandwidths were 16% at 40
kHz and about 11% between 60 and 100 kHz in a bottlenose dolphin (Lemonds et al. 2000). In
the latter study, critical bandwidths were 12 and 17% of the center frequency at 20 and 30 kHz,
respectively, in two bottlenose dolphins and 9 and 15% at 20 and 30 kHz, respectively, in a
beluga whale (Finneran et al. 2002a). These values roughly correspond to one-sixth of an octave
wide and, like CRs, are frequency dependent with critical bandwidths tending to increase with
increasing frequency (Yost 2000).
Spatial Overlap
Spatial overlap between a target sound and masker is another important consideration of
auditory masking. The directional hearing system helps ameliorate the masking effect of noise,
provided that the masker itself has a direction. That is, when a signal and masker are spatially
separate, masked thresholds are lower or sensitivity is better compared to when signals and
maskers are collocated. The auditory phenomenon, termed “spatial release from masking” or
“spatial unmasking,” can significantly improve signal detection of both simple and complex
23
sounds. For example, Bain and Dahlheim (1994) measured auditory thresholds for tones at 4, 8,
and 20 kHz masked by low frequency noise with energy primarily between 500 Hz and 5 kHz in
two captive killer whales. The signal always occurred directly in front of the subject (0, 0°)
while the masker was positioned at 0 (collocated), 90, or 150° in the horizontal plane or 30°
above or below the subject. Although it is unclear whether any sound energy from the noise
masked the 8 and 20 kHz signal, masked thresholds were generally lower (ranging between 0
and 40 dB) when the signal and masker were spatially separate compared to when they were
collocated. Masked thresholds were lowest (on average by −7 to −24 dB) when the signal and
masker were separated by 90 and 150° relative to when the signal and masker were collocated.
These effects were largest for the higher frequencies, but in most cases there were large
interindividual differences (Bain and Dahlheim 1994).
In another investigation using a bottlenose dolphin as a subject, masked thresholds were
measured with spatial separation of signal (at 30, 60, and 120 kHz) and masker (band pass of 30
and 150 kHz) in both the horizontal and vertical planes (Au and Moore 1984). As in the killer
whale investigation, masked threshold differences of the dolphin were generally largest
(thresholds were lowest) for the highest frequencies at both azimuth and elevation (Au and
Moore 1984). CRs are generally measured with the signal and masker emanating from the same
point in space. Thus masking effects in more real world situations would not be as substantial
when signals and maskers have different spatial configurations relative to each other as data from
critical ratio investigations suggest.
Temporal factors can also cause release from masking in some cases, particularly when
two separate sources of noise (maskers) across different frequencies are amplitude modulated in
a coherent way. When the signal overlaps in frequency with one of the two maskers, detection
of the signal improves when the two maskers are amplitude modulated coherently compared to
when they are modulated incoherently. Such an auditory phenomenon is called comodulation
masking release. This effect may be related to the degree of amplitude modulation of the two
maskers. For example, when only one masker is present which overlaps in frequency with the
target signals, detection of the signal is easier for a masker that has a larger degree of amplitude
modulation (and periods of lower energy), compared to when a masker has less amplitude
modulation even when the overall sound energy level of the maskers are equivalent (Brumm and
Slabberkoorn 2005). No marine mammal studies on comodulation masking release have been
conducted, but the effect seems to occur across a wide range of taxa including humans and birds
and is likely to occur in marine mammals as well.
While most investigations of auditory masking have measured thresholds of pure tones
masked by white noise, very few studies have determined the effects of noise on the reception of
natural sounds. Bain and Dahlheim (1994) reported thresholds of pure tones, killer whale clicks,
and a call masked by recorded vessel noise from oil-spill clean up operations in Alaska’s Prince
William Sound after the 1989 Exxon Valdez oil spill. While masking of pure tones occurred for
all frequencies below 20 kHz, very little masking of the call and clicks was observed; however, a
27 kHz pure tone was added to the call to facilitate performance transfer and it is unclear
whether or not this improved detection (Bain and Dahlheim 1994). Given that the click and call
plus high frequency tone were likely outside the bandwidth of the vessel noise, it is not
surprising that these results were observed.
24
Hearing Loss Due to Sound Exposure
Another auditory effect of sound exposure is hearing loss. There are temporary and
permanent forms of hearing loss. Temporary hearing loss or temporary threshold shift (TTS)
involves recovery of baseline hearing over a period of time and occurs from physiological
fatigue of the hair cells of the inner ear. The magnitude of the threshold shift depends on the
amplitude, duration, temporal pattern, frequency, and energy content of fatiguing sound. For
example, exposure to a more intense but shorter sound may result in similar TTS compared to
that measured after exposure to a less intense but longer sound. With some types of sound
exposure, subjects may experience large but fully recoverable shifts in threshold (≈40 dB) that
become asymptotic with longer exposure durations (asymptotic TTS). The largest threshold
shifts often occur at frequencies one-half to one octave above the frequency of exposure (Yost
2000).
Permanent hearing loss or permanent threshold shift (PTS) does not show recovery over
time and is the manifestation of auditory injury. PTS results from damage or death of the hair
cells in the inner ear. PTS can occur from repeated exposures that induce TTS or from a single
intense exposure. The relationship between TTS and PTS is often quite complex despite intense
research in humans and terrestrial models. For example, even very large threshold shifts (up to
40 dB) resulting from sound exposure can be fully recovered. For ethical reasons, PTS is not
studied in marine mammals. TTS has been investigated to describe how temporary hearing loss
is dependent on the type, frequency, amplitude, and duration of the exposure, and to model PTS
from TTS data despite the complicated relationship between the two. This work is typically
done in a laboratory setting in which trained animals are tested for baseline hearing and exposed
to noise, then hearing is tested again to measure a threshold shift.
Both behavioral and electrophysiological methods have been used to measure TTS.
While there are no TTS studies on killer whales, there are a handful of TTS investigations on
bottlenose dolphins and beluga whales. These include measuring TTS after exposure to
broadband noise (Nachtigall et al. 2003, 2004), tones (Schlundt et al. 2000, 2006, Finneran et al.
2005, 2007), and impulsive sounds (Finneran et al. 2000, 2002b). Since TTS is dependent on
both amplitude and duration of exposure, sound exposure level (SEL in dB re 1 µPa2s) is a
common way to quantify exposure, because it describes both duration and amplitude in one
metric. However, SEL inherently assumes that different combinations of amplitude and duration
that yield equivalent SELs (i.e., equal energy) result in the same amount of TTS. Some data
measured in delphinids suggest that this may be an appropriate assumption under some
conditions (Finneran et al. 2005).
Using a behavioral procedure to measure threshold shifts of a 7.5 kHz tone in a
bottlenose dolphin, an average TTS of 11 dB resulted after exposure to broadband noise having a
flat spectrum between 4 and 11 kHz at 179 dB re 1 µPa for 41 to 54 minutes (Nachtigall et al.
2003). This corresponded to an SEL of 213.8 dB re 1 µPa2s when the average exposure duration
of 50 minutes was used. Using an electrophysiological approach to measure threshold shifts at 8,
11.2, 16, 22.5, and 32 kHz in the same subject with the same fatiguing stimulus at 160 dB re 1
µPa for 30 minutes (SEL of 192.6 dB re 1 µPa2s), the largest threshold shifts occurred after 5
minutes of exposure (TTS5) at frequencies between 8 and 16 kHz. The maximum threshold shift
of 8 dB occurred at 16 kHz with shifts equal to approximately 5 and 6 dB at 8 and 11.2 kHz,
25
respectively. Recovery to baseline hearing levels was approximately 1.5 dB per doubling of time
(Nachtigall et al. 2004).
TTS was also measured in bottlenose dolphins and belugas whales exposed to 1-second
tones ranging between 3 and 75 kHz and at levels between 182 and 201 dB re 1 µPa. Masked
temporary threshold shifts (MTTS) at frequencies equal to or higher than (by one-half or one
octave) the exposure frequency ranged between 6 and 17 dB (Schlundt et al. 2000). Only
threshold shifts greater than or equal to 6 dB were reported as noise induced TTS because this
value seemed sufficiently above the 3–4 dB variability of baseline hearing thresholds. The
beluga whale exhibited equivalent threshold shifts (between 6 and 8 dB) at higher exposure
levels to that of the dolphin when the same exposure and hearing test frequencies was used
between the subjects. Only masked thresholds could be measured since behavioral testing was
conducted in San Diego Bay, where ambient noise levels were high and variable, so masking
noise (between 3 and 100 kHz and varied in spectral levels between 63 and 95 dB re 1µPa2/Hz)
was used to create a consistent noise floor during preexposure, exposure, and postexposure
conditions (Schlundt et al. 2000).
Because the presence of masking noise often results in smaller threshold shifts in
terrestrial mammals, these results reported by Schlundt et al. (2000) were questioned and testing
was replicated in a quiet pool again using a behavioral response procedure. In this study, two
bottlenose dolphins were exposed to a 3 kHz tone at levels ranging between 100 and 200 dB re 1
µPa for 1–8 seconds, corresponding to SELs of 100 to 203 dB re 1 µPa2s. In most cases,
significant threshold shifts (different from control exposures) measured 4 minutes after exposure
(TTS4) were observed at 4.5 kHz for SELs above 190 dB re 1 µPa2s. No significant TTS was
observed at the same frequency as the exposure up to levels of 200 dB re 1 µPa2s (Finneran et al.
2005). For SELs above 200 dB re 1 µPa2s, recovery was not complete by 10 minutes, but
hearing sensitivity typically returned to baseline values within a day. All data summarized for
exposures to continuous noise or tones (nonimpulsive sounds) in bottlenose dolphins and beluga
whales showed that, in most cases, significant threshold shifts occurred at SELs greater than or
equal to 195 dB re 1 µPa2s despite differences in exposure duration, sound pressure level,
experimental approaches, and subjects (Finneran et al. 2005).
More recent work has extended exposure durations at the same frequency up to 128
seconds corresponding to SELs of up to 217 dB re 1 µPa2s, and preliminary results indicated that
TTS4 of up to 23 dB occurred with full recovery within 30 minutes (Schlundt et al. 2006).
Hearing sensitivity was measured both behaviorally and electrophysiologically in a bottlenose
dolphin after exposure to a 20 kHz tone that ranged in SELs of 203–206 dB re 1 µPa2s in a more
recent study by Finneran et al. (2007). Threshold shifts up to 40–45 dB using the physiological
approach were observed with frequency-dependent results. The largest shifts were observed at
30 kHz, followed by 40 and 20 kHz with no measurable TTS occurring at 10, 50, 60, and 70
kHz. Threshold shifts were smaller (19–33 dB) and recovery time was faster when the
behavioral method was used even when the exposure and the time taken to measure hearing
sensitivity after exposure was comparable between the two approaches (Finneran et al. 2007).
Such differences are likely attributable to inherent differences between the two approaches.
Further investigations are required to work out these differences and to interpret TTS results
using evoked potential audiometry (Finneran et al. 2007).
26
A few studies have also investigated the effects of impulsive sounds (transient sounds
having rapid rise times and high peak levels) on MTTS in bottlenose dolphins and beluga
whales. In these studies, behavioral methods were used to measure masked hearing thresholds
before and after exposure in San Diego Bay. No MTTS (defined as ≥ 6 dB) was observed in
either species after exposure to sounds simulating underwater explosions with SELs up to 179
dB re 1 µPa2s, but behavioral responses above 156 dB re 1 µPa2s were sometimes noted
(Finneran et al. 2000). In another study, MTTS was measured after exposure to single
underwater impulses from a seismic water gun. MTTSs of 7 and 6 dB were measured at 0.4 and
30 kHz, respectively, in a beluga whale after exposures of 186 dB re 1 µPa2s. Recovery
occurred within 4 minutes after exposure. No MTTS greater than or equal to 6 dB was observed
in the bottlenose dolphin up to SELs of 188 dB re 1 µPa2s (Finneran et al. 2002b).
It is unclear what effect the masking noise had on the threshold shifts from impulsive
sound exposure because these experiments were not repeated in quieter conditions. In some
cases, physiological responses to sound exposure were also investigated (Romano et al. 2004)
and these are reviewed in the Nonauditory Effects section below.
The TTS investigations on delphinids reviewed here involved sound exposures with
constant exposure levels. However, free-ranging animals are often exposed to sounds that are
intermittent in time and space. While very little work has been done on the issue of sound
exposure intermittence on TTS in marine mammals, such exposures in people yield somewhat
complicated results, partly because of partial recovery during the quieter periods. For very short
interruptions or small fluctuations, TTS is proportional to the average exposure level. With more
variability in sound exposure levels, TTS is not a consistent function of the total energy of the
exposure (Ward 1997).
Another consideration is that even between similarly related species, TTS at a given SEL
can vary dramatically. This might be due to differential sensitivity at the frequencies of the
sound exposure. Bottlenose dolphins and beluga whales have similar audiograms (Schlundt et
al. 2000) which might explain why similar sound exposure levels produced comparable amounts
of TTS. Hearing sensitivity at both 4 and 20 kHz is roughly 10–20 dB better in the killer whale
than in the bottlenose dolphin, as shown in Figure 8. However, these results are based on the
performance of only a few individuals (n = 1 or 2). Some investigators argue that it might be
more appropriate to normalize SEL relative to baseline hearing sensitivity to account for these
differences (Kastak et al. 2005). Thus while TTS results of the bottlenose dolphin and beluga
whale provide the best available data to assess potential hearing loss due to sound exposure in
killer whales, differences in hearing sensitivity at a given frequency might produce more or less
TTS as a result of sound exposure at that frequency.
27
20
30
40
50
60
70
80
90
100
110
120
130
140
150
Killer whale
Bottlenose dolphin
Beluga whale
1 1 1 1 100 10 00 000 0000 000 000000
equency
Threshold (dB re: 1 µPa)
Fr
Threshold (dB re 1µPa)
10 100 1,000 1 1 1 0,000 00,000 ,000,000
ency Frequ (Hz)
Figure 8. Behavioral audiograms plus electrophysiological thresholds below 2 kHz for the killer whale
(based on Szymanski et al. 1999), bottlenose dolphin (based on Johnson 1967), and beluga whale
(based on White et al. 1977, Awbry et al. 1988, Johnson et al. 1989).
28
Behavioral Changes in the Presence of
Unwanted Sound
Other effects of sound exposure include behavioral changes that might have
repercussions on life functions of affected animals. Behavioral responses to sound exposure can
be highly variable and depend not only on factors related to the characteristics of the sound
source, its transmission, and background ambient noise levels, but also on individual, group,
population, and species-level differences arising from factors such as age, sex, dependent
offspring presence, hearing sensitivity/loudness perception, activity patterns, motivational states,
previous experience that may have resulted in habituation or sensitization, and noise tolerance
(NRC 2003, 2005).
There have been many studies investigating the effects of sound exposure on the behavior
of marine mammals, including those on killer whales. These responses may be short-term
changes such as changing swimming direction, dive duration, or vocal behavior, or long-term
changes such as avoiding a once popular area for foraging, breeding, or socializing (Morton and
Symonds 2002, Foote et al. 2004, Bain et al. 2006, Williams and Ashe 2006, Williams et al.
2006). Most of these studies involve investigating behavioral changes in the presence of
anthropogenic sources relative to some baseline behavioral measurement. However, in some
cases, it is unclear whether the physical presence of the sound-producing device (such as a
moving vessel) or the sound itself caused behavioral changes.
Behavioral Patterns
Behavioral activities of SRKWs were observed in the presence and absence of vessels
within 1,000 m over the summers of 2003–2005 (Bain et al. 2006). Activity budgets (foraging,
rest, traveling, and socializing) were reported at two sites along the west side of San Juan Island,
a north site (lat 48°30.561′N, long 123°8.494′W) and a south site (lat 48°27.421′N, long
122°59.401′W). Whales spent more time traveling and less time foraging in the presence of
boats within 100 and 400 m of the focal whale group than in their absence. No difference
between foraging, resting, and traveling between study sites was found, but whales spent more
time socializing in the north than in the south site. They were more likely to continue foraging
when boats were absent compared to when they were present within 100 and 400 m, but it was
possible the responses at 400 m were a result of the effect at 100 m. Whales also traveled in less
direct paths and had longer average durations between breaths when vessels were present
compared to when they were absent within 1,000 m. Surface active behavior was not notably
more frequent in the presence of vessels than in their absence within 1,000 m, and there was no
significant difference in swim course or speed due to vessel traffic (Bain et al. 2006).
Williams et al. (2006) also found that vessel presence changed the behavioral patterns of
Northern Resident killer whales in Johnstone Strait, British Columbia. Whales spent
significantly less time feeding and more time traveling, socializing, and resting in the presence of
29
vessels than in their absence within eight zones, four within a reserve where no boats were
allowed and four adjacent zones. While the overall energetic demand was estimated to be 3%
higher in the presence of boats, the lost opportunity to feed had a larger energetic effect
estimated to be about a 28% decrease in energetic gain (Williams et al. 2006).
The significance of these results as well as those of Bain et al. (2006) is contingent on the
ability of researchers to accurately discriminate different behavioral states of free-ranging killer
whales. Williams and Ashe (2006) also conducted controlled experiments in which adult focal
male Northern Residents in Johnstone Strait were tracked when no boats were within 1,000 m,
then in the presence of a few (1–3) or many (>3) boats within 1,000 m. Males swam in less
direct paths in the presence of a few boats within 1,000 m, but swam in more direct paths when
there were many boats present within 1,000 m. If swimming directedness had been analyzed
simply in the presence or absence of boats, then this difference in behavior would not have been
apparent. An investigation of Southern Resident behavior with varying number of boats would
also be useful to determine boat number effects. However, unlike in the Northern Residents,
there are fewer opportunities to observe Southern Resident behavior when no boats or a few
boats are near them.
Active Sonar and other Anthropogenic Sounds
There was also a documented case of exposure by J pod to military active sonar by the
USS Shoup on 5 May 2003 in the Haro Strait (U.S. Navy 2004). The USS Shoup, a guided
missile destroyer, was equipped with a AN/SQS-53C(V)4 hull-mounted sonar with kingfisher
mine avoidance system that produces frequency-modulated signals between 2.6 and 3.3 kHz at a
nominal source level of 235 dB re 1 µPa at 1 m. This system’s midfrequency tactical sonar is
designed to detect, localize, and avoid mines and other objects. The goal of the Navy on that day
was to train ship personnel to use the sonar system in a “sweep channel” exercise in which 1–2
second signals were emitted once every 28 seconds from 1123 to 1438 hours (U.S. Navy 2004).
During part of the exercise, J pod was in the area and observed by several researchers familiar
with SRKW behavior who described the whales as exhibiting “abnormal behavior” including
bunching as a group close to shore (NMFS 2005a).
The sonar signals were recorded by Val Veirs on bottom mounted hydrophones and
recordings were provided to NMFS, but the dynamic range of the system prevented accurate
assessments of the highest received levels. Thus the Navy Research Laboratory estimated
received levels relative to the whales. Mean received levels relative to J pod were estimated to
range between 121 and 175 dB re 1 µPa and likely received SELs were 169.1–187.4 dB re 1
µPa2s with a worse case range estimated to be 177.7–195.8 dB re 1 µPa2s. NMFS (2005a)
reported that it was likely that these received levels were audible to the whales and that
reverberation from the complex underwater environment may have reduced the ability of the
whales to localize the sound source.
TTS onset was estimated based on TTS studies conducted on dolphins (reviewed in the
Hearing Loss Due to Sound Exposure subsection above) and except in the worse case scenario,
NMFS (2005a) stated that it was not likely that the whales experienced temporary or permanent
hearing loss (based on a TTS onset level of 195 dB re 1 µPa2s), but that the received levels were
high enough to induce a behavioral reaction that was consistent with observer reports.
30
31
Longer-term effects of sound exposure were observed in both resident and transient killer
whales in the Broughton Archipelago, British Columbia, when acoustic harassment devices
(AHDs) were in operation. The AHD signals were designed to repel harbor seals from salmon
farms by producing a 10 kHz signal at a source level of 194 dB (re 1 µPa at 1m). It was
estimated that the received levels of the AHD signals would reach ambient noise levels at about
50 km from the source. The number of days killer whales were sighted in the area was
significantly reduced after the AHD broadcasts, compared to before (Morton and Symonds
2002).
Other anthropogenic sounds described in SRKW critical habitat are those associated with
construction activity, particularly pile driving given its relatively loud and impulsive acoustic
signature at close ranges. Although behavioral effects of pile driving or other marine
construction noise on killer whales has not been reported, Wursig et al. (2000) found that Indo-
Pacific humpback dolphins (Sousa chinensis) significantly increased their group swim speed
during pile driving even with the deployment of a bubble curtain, but such human activity did
not have a long-term overall effect on dolphin abundance in the area.
Vocal Response to Background Noise
Another response to sound exposure involves changes in vocal behavior. In humans, an
unconscious response to speak louder with higher background noise levels is called the Lombard
effect/response (Lombard 1911) and is often observed when people wear headphones and talk
while listening to loud music. Many animals also call louder, longer, more often, or at different
frequencies in the presence of masking noise (Brumm and Slabbekoorn 2005). For example,
beluga whales in the St. Lawrence River responded to increases in ambient noise from boat
traffic by increasing call amplitude (Schiefele et al. 2005), call repetition rate, and frequency of
the call (Lesage et al. 1999). SRKWs significantly increased the duration of their primary
stereotyped call in the presence of boats compared to in their absence in the more recent time
periods, which correlated with a dramatic increase in the number of whale watching boats in
their habitat (Foote et al. 2004). In laboratory settings, bottlenose dolphins increased the average
number of clicks emitted per trial as white noise levels were increased during an echolocation
target detection task (Au et al. 1982).
Biogenic noise from other animals also results in vocal compensation effects. For
example, Antarctic killer whales shifted the frequency modulated points of their calls below or
above the frequency range of predominant leopard seal (Hydruga leptonyx) calls during the
season when leopard seals were calling (Mossbridge and Thomas 1999). A captive beluga whale
produced louder and higher frequency echolocation clicks when it was moved from San Diego
Bay, California, to Kaneohe Bay, Hawaii, which had ambient noise levels that were typically 12–
17 dB higher than in San Diego due to snapping shrimp presence (Au et al. 1985). Thus when it
comes to sound production, there are strategies that animals use to attempt to compensate for
increased ambient sound levels. When animals exhibit such vocal compensation behaviors, it
can be interpreted that auditory masking is a potential challenge, particularly when ambient
sound levels reach a point where animals can no longer compensate by increasing their vocal
output or other sound emission parameter.
Strandings and other Nonauditory Effects
of Sound Exposure
An extreme behavioral outcome to sound exposure is stranding. In some cases, cetacean
strandings have coincided in space and time with military sonar exercises. There are three well-
documented cases, one in Greece in 1996, one in the Bahamas in 2000, and one in the Canary
Islands in 2002, and all involved the mass stranding of 11–14 individual beaked whales (and
sometimes other cetacean species). While there is a general consensus among the scientific
community that military sonar was a likely causal link to the strandings, the mechanisms which
caused the observed pathologies are still a mystery (NRC 2005). For example, results from
necropsies performed on 10 of the 14 individuals involved in the Canary Island stranding event
revealed extensive bubble formation and tissue damage in various organs. A number of
hypotheses including decompression sickness and acoustically mediated bubble growth due to
tissue supersaturation of nitrogen (via rectified diffusion) were proposed, but are highly debated
explanations (Crum and Mao 1996, Fernandez et al. 2004, 2005, Jepson et al. 2003, Piantodosi
and Thalmann 2004).
There are no documented cases of killer whale strandings that coincide with military
sonar exercises; however, several harbor porpoises (Phocoena phocoena) stranded before, on,
and after the USS Shoup exercise on 5 May 2003. Forensic examination revealed no evidence of
acoustic trauma but inadequate preservation of the samples produced equivocal results (NMFS
2005a).
Other less extreme, but potentially serious, nonauditory effects of noise include
physiological effects related to stress. Most studies investigating these effects have been
conducted on terrestrial mammals and include changes in heart, respiration, and metabolic rate,
and immune and reproductive function.
A few studies have investigated the production of stress hormones including
catecholamines (such as norepinephrine, epinephrine, and dopamine) and aldosterone
(implicated in a longer-term stress response in marine mammals) as well as decreased immune
function resulting from controlled sound exposures in delphinid subjects. For instance, Thomas
et al. (1990) found no difference in the levels of catecholamines in beluga whales after exposure
to playbacks from an oil drilling platform. Romano et al. (2004) also measured the effects of
sound exposure on nervous system and immune function in laboratory subjects of ongoing TTS
studies (Finneran et al. 2002b).
Norepinephrine, epinephrine, and dopamine levels were significantly higher in the beluga
whale after high-level (< 183 dB re 1 µPa2s) compared to low-level (>183 dB re 1 µPa2s) or
control exposures of impulses from a seismic water gun. The bottlenose dolphin did not show
differences in catecholamine levels but did have higher aldosterone levels and lower absolute
monocyte (cells of the immune response) counts after experimental exposures compared to those
32
measured after control exposures (Romano et al. 2004). It appears that sound exposure in some
instances results in measurable differences in the production of stress hormones and cells related
to immune function that could have undesirable physiological effects in free-ranging delphinids.
33
Zones of Influence
The previous sections have reviewed studies on the effects of sound exposure in killer
whales and other odontocetes. These studies have shown that 1) marine mammals, including
killer whales and closely related species, are vulnerable to auditory effects of noise exposure
including masking and hearing loss effects; 2) individuals and groups of killer whales, including
Southern Residents, exhibit behavioral changes that are consistent with avoidance responses in
the presence of vessels and AHD signals; 3) physiological changes with undesirable effects on
endocrine and immune function in smaller dolphins have resulted from sound exposure; and 4) in
extreme cases, beaked whale strandings have been linked spatiotemporally with midfrequency
military sonar exercises.
One objective of this review is to use results from previous investigations to assess
potential impacts of sound exposure from various sources on SRKWs. An approach to this
assessment is to estimate the zone where an acoustic effect is expected based on available data
(Richardson et al. 1995). Four zones are defined based on different response effects. In general,
these are listed in order of smaller radii of influence relative to the acoustic source as follows: 1)
zone of audibility is the area where the received levels of sound exposure are high enough that a
killer whale can hear it; 2) zone of responsiveness is the area where the received level is such
that the whale responds behaviorally or physiologically; 3) zone of masking is the area where
received levels can mask the detection of sound emissions for biosonar and vocal
communication, prey sounds, or other biologically important sounds; and 4) zone of hearing loss
or injury is the area where received levels (or sometimes more appropriately SELs) are high
enough to cause temporary or permanent hearing loss.
Richardson et al. (1995) took the approach of describing these zones by distance, but
accurate assessments of range require adequate knowledge of sound propagation characteristics.
Here, zones of influence are by and large described by received levels given the complex shallow
water environments inhabited by Southern Residents and the uncertainty involved in predicting
propagation loss in space and time in these areas. This focus on received levels ignores the
influence of temporal factors such as duration, pulse repetition rate, or duty cycle of the sound
exposure that also likely contribute to potential response effects. Where information is available
and appropriate, such as in the discussion of zones of hearing loss, SELs that take the duration of
exposure into account will also be discussed.
Zone of Audibility
Whether or not a sound produced by human activity will be audible to a Southern
Resident killer whale depends on the interplay between source, path, and receiver variables (for
review, see Richardson et al. 1995). Source variables include source level, duration, and
frequency of the sound; path variables include properties of transmission loss that affect received
levels; and receiver variables include hearing sensitivity and responsiveness of the whales with
respect to the received sound. If the received sound level falls below the detection threshold of
34
the animal, then the signal is said to be “threshold limited,” while if it is above the detection
threshold but below the ambient noise level at all frequencies that killer whales are sensitive to, it
is said to be “noise limited” (Richardson et al. 1995). The absolute hearing sensitivity of killer
whales is generally below the typical one-third octave ambient noise level for frequencies less
than 3 kHz and greater than 70 kHz (arrows in Figure 9) in low ambient noise situations such as
in calm conditions (e.g., a sea state of 0) in about 200 m of water. Thus killer whale hearing will
be threshold limited in these conditions.
Whales will likely be threshold rather than noise limited at lower frequencies. This
prediction is based on the assumption that the critical bandwidths of the killer whale are
approximated by one-third of an octave, which is a typical assumption with lack of direct
measurements of critical bandwidths. More direct predictions can be made using the critical
ratios of 20 dB re 1µPa at 10 kHz and 40 dB re 1µPa at 80 kHz (Bain et al. unpublished data,
cited in Bain and Dahlheim 1994) and spectrum levels measured in the areas that SRKWs
frequent such as those reported by Veirs and Veirs (2005). For example, the received level of a
10 kHz tonal signal would have to be approximately 78 dB in order to be detected at night in
Haro Strait because ambient spectrum levels at this frequency are 58 dB re 1µPa2/Hz and the
critical ratio is 20 dB.
20
30
40
50
60
70
80
90
100
110
120 Killer whale thresholds
SS 0 1/3 Octave levels
100 1000 10 1 100000 00000 0000
Frequency
Threshold (dB re: 1 µPa)
Threshold (dB re 1µPa)
100 1,000 1 1 1,0,000 00,000 000,000
Frequency (Hz)
Figure 9. Killer whale behavioral audiogram based on Szymanski et al. 1999 (reprinted from Figure 5c in
Szymanski et al., copyright 1999, with permission from Acoustic Society of America), and one-
third octave ambient levels in a sea state of 0 from Richardson et al. 1995 (reprinted from Figure
5.1b in Richardson et al., copyright 1995, with permission from Academic Press). Dashed lines
are extrapolated values and the two × marks indicate physiological thresholds to supplement
information on sensitivity at the lower frequencies. Arrows indicate the frequency boundaries at
which hearing would be noise versus threshold limited.
35
The degree to which hearing will be noise limited will depend on the ambient conditions
which are quite variable. Based on the Szymanski et al. (1999) audiograms, Erbe (2002)
estimated that vessel noise levels based on one-twelfth octave band analysis was sufficiently
loud enough to exceed the hearing thresholds of killer whales at frequencies above 2 kHz. It was
estimated that a fast inflatable operating at speed of 51 km/hour and a source level of 162 dB re
1µPa would be audible to killer whales over a range of about 16 km in a sea state of one half. A
slow inflatable operating at a speed of 10 km/hour would be audible over a shorter range of 1 km
in the same sea state (Erbe 2002). The model used assumed that the limit of audibility occurs
when the critical band level becomes equal to or less than the ambient noise. However, one-
twelfth octave critical bandwidths were assumed across all frequencies based on Fletcher’s
(1940) equal power assumption. The zone of audibility for this boat noise might be different if
direct measurements of killer whale critical bandwidths were available.
Zone of Responsiveness
In this section, the zone of responsiveness is defined as the area where the received level
is such that killer whales would respond behaviorally. Most studies that illustrate statistically
significant effects on the behavior of killer whales in the presence of anthropogenic sound
provide insufficient information related to the source or received sound levels, frequencies, and
duration of exposure, making it extremely challenging to predict the zone of responsiveness.
The level of ambient noise, which can fluctuate in space and time and the resulting effective
signal-to-noise ratio, also adds complexity to this issue. Furthermore, reactions based on other
sensory information such as visual cues cannot be ruled out in the absence of controlled studies
(such as those involving playbacks).
Based on the little information available, it appears that killer whales will avoid an area
within about 4 km of an AHD producing a 10 kHz signal at a source level of 194 dB re 1 µPa at
1 m (Morton and Symonds 2002, NMFS 2005a). The transmission loss characteristics of the
waters surrounding the AHD broadcasts are unknown; but, if we assume that at the most it would
take the form of spherical spreading loss and at the least cylindrical spreading loss, the likely
range for received levels at 4 km can be determined. The validity of these spreading loss
assumptions is tentative. Using the following equation to calculate received levels,
RL = SL – TL (6)
where RL is the received level, SL is the source level of the signal, and TL is the transmission
loss, where TL = 20 log R + αR for spherical spreading or TL = 10 log R + αR for cylindrical
spreading, and where α is the absorption loss at 10 kHz and is approximately 0.00118 dB/m
(Urick 1983), this would correspond to a AHD received level ranging from 117 to 153 dB re 1
µPa.
Based on the response of J pod to the sonar exercise by USS Shoup in May 2003,
Southern Residents are predicted to behaviorally respond to frequency modulated signals
between 2.6 and 3.3 kHz at received levels estimated between 121 and 175 dB re 1 µPa and
received SELs of 169.1–187.4 dB re 1 µPa2s.
36
In the absence of data on the effects of vessel noise levels on killer whale behavior, Erbe
(2002) assumed that a received broadband sound pressure level of 120 dB re 1µPa based on
Richardson et al. (1995) would produce a behavioral reaction by killer whales. A behavioral
response was predicted to occur when the vessel was within a 200 m and a 50 m range for an
inflatable operating at 51 km/hour and 10 km/hour, respectively. Effects on the activity budgets
of Southern Residents in the presence of vessels within 400 m as reported by Bain et al. (2006)
agree with these predictions by Erbe (2002). In addition, the lower limits of the estimated
received levels of both AHDs and midfrequency sonar signals that produced behavioral
responses in Northern and Southern Resident killer whales fall within this range.
Furthermore, if it is assumed that received levels above 120 dB would produce a
significant behavioral response, then this would occur 50% of the time during summer days at
the position of the recording hydrophones based on 2-second sound pressure level averages
reported in the Veirs and Veirs (2005) study. This type of criterion based on one broadband
sound pressure level measurement does not take into account frequencies of best hearing
sensitivity of killer whales. For example, most energy of the received levels measured by Veirs
and Veirs (2005) falls below 1 kHz, which is outside the range of best hearing sensitivity of
killer whales (Szymanski et al. 1999). Weighting functions that emphasize frequencies of best
hearing sensitivity while deemphasizing those of least hearing sensitivity based on the killer
whale audiogram (or equal loudness level contours if available) would be useful to measure
exposure levels in order to address these frequency effects.
Zone of Masking and Effects on the Active Space
of Sound Emissions
The active space of a signal is the range over which it can be detected. Any sound that
masks the reception of a communicative signal, such as conspecific call or echolocation click,
will reduce the active space of that signal. Detection of natural signals a few dB below ambient
noise levels has been demonstrated in humans and some marine mammals (Richardson et al.
1995). Captive killer whales have demonstrated that detection thresholds of clicks and calls are
not impeded by boat noise, although in some cases, frequency overlap was not a likely
contributing factor (Bain and Dahlheim 1994). The sonar equation provides a useful way to
describe the active space of an acoustic signal. To determine the range at which a sound might
be heard by a listener, the sonar equation may be defined as follows:
RL = SL – TL – NL + DI (7)
where NL is the noise level and DI is the directivity index of the auditory system (Urick 1983).
In most cases where background noise exists, the received level would have to exceed the
CR in order for detection to take place. In the case of auditory masking, NL increases, reducing
the received level of the signal. If the source level of a signal is not below the hearing threshold,
the excess level (in dB) can be used to calculate the distance over which it might be heard after
transmission loss has been taken into account. Assumptions about transmission loss vary
considerably as well as the bandwidth appropriate for analyzing background noise. Based on a
one-twelfth octave bandwidth analysis, Erbe (2002) estimated an inflatable operating at a speed
of 51 km/hour would mask the reception of a killer whale call to the point of being
37
unrecognizable (when spectral components at 4.7 and 5.8 kHz were inaudible) within a range of
14 km. An inflatable operating at a speed of 10 km/hour was predicted to mask killer whale calls
within a 1 km range. However, the killer whale call level incorporated into the model was based
on received levels (up to 124 dB re 1µPa) and not source levels which are known to be
considerable higher (135–168 dB re 1 µPa at 1 m, Miller 2006). Thus the range of masking
effects is likely overestimated in this instance.
Miller (2006) reported that the mean active space of stereotyped calls with high-
frequency components was 10–16 km in sea state 0 while the mean active space of stereotyped
calls without high-frequency components, whistles, and variable calls was 5–9 km. In these
cases, a killer whale call was assumed to be detectable by another whale when the received level
in at least 1 one-third octave band exceeded the hearing threshold or was 6 dB below the
background noise level, whichever was greater. Relative to the sea state of 0 in which noise
spectral levels are about 26 dB re 1µPa2/Hz lower, estimated active space of calls in a sea state of
6 were 75%, 83%, and 91% lower for stereotyped calls, variable calls, and whistles, respectively,
using Fletcher’s (1940) equal power assumption (Miller 2006). The frequency and time structure
of signals relative to that of the noise as well as the directional hearing system might result in
increased sensitivity to these signals, but these considerations were not incorporated in the
models for simplicity (Erbe and Farmer 2000, Erbe, 2002, Miller 2006).
For echolocation signals, the sonar equation is modified to include two-way transmission
loss as well as the target strength (TS) of the ensonified object as follows:
RL = SL – 2TL + TS – NL +DI (8)
Au et al. (2004) used these equations to estimate the horizontal range at which a foraging
killer whale would be able to perceive biosonar echoes off prey in quiet conditions. This was
based on a model of Chinook salmon (Oncorhynchus tshawytscha) target strength and an
estimated echo level (EL) received by a whale using the following equation:
EL = SL – 2TL + TS (9)
Based on these assumptions and using dB peak-to-peak levels, it was estimated that an
echolocating killer whale would receive echoes of a Chinook salmon at 65 m depth between 29
and 33 dB above threshold at a horizontal range of 100 m. In noisier conditions, the noise level
along with the DI must also be considered. The DI of a killer whale was estimated to be 21 dB at
50 kHz based on a scaling factor to convert the DI of a bottlenose dolphin, since no direct
measurements of DI in killer whales were available (Au et al. 2004). For moderately heavy rain,
the ambient noise spectral density level was estimated to be 52 dB re 1µPa2/Hz and it was
predicted that the echo level would not extend above the noise level until the whale was within a
horizontal range of 40 m from the prey item (Au et al. 2004). Thus the active space of the
echolocation signal was reduced considerably from quiet to noisier conditions created by heavy
rain. It is important to note that the quiet condition assumed a complete lack of ambient noise,
which is not realistic. Free ranging animals will never be in an environment that is completely
lacking of ambient noise.
38
The analysis of active space for echolocation signals can be extended for conditions when
vessels are operating in the vicinity of killer whales. Based on data from Hildebrand et al.
(2006), spectral levels of vessel and Haro Strait ambient noise were used to estimate the
horizontal detection range of killer whales echolocating on Chinook salmon. For this analysis,
source spectral density levels at 50 kHz for cruise and power up operating speeds were available
to predict received levels of a whale at 100, 200, and 400 m from two whale watching vessels.
Received levels for these conditions are based on the following equation:
RL = SL − TL = SL − 20 log R + αR (10)
where SL is the source level at 50 kHz for a given vessel at a given speed, R is the range in
meters, and α is the absorption loss at 50 kHz and is approximately 0.016 dB/m (Urick 1983) as
shown in Table 1.
Table 1. Source levels for whale watching vessels at different speeds, calculated received levels
assuming spherical spreading loss for three distances, and measured received level of a container
ship as reported by Hildebrand et al. (2006).
Vessel Speed
SL at 50
kHz dB
re 1µPa^2/Hz Range (m)
Calculated RL
at 50 kHz
in dB
re 1µPa^2/Hz
Measured RL
at 50 kHz
in dB
re 1µPa^2/Hz
Boat B Cruise 101 100 59.4
101 200 51.8
101 400 42.6
Power 92 100 50.4
92 200 42.8
92 400 33.6
Boat C Cruise 93 100 51.4
93 200 43.8
93 400 34.6
Power 95 100 53.4
95 200 45.8
95 400 36.6
Boat D Cruise 111 100 69.4
111 200 61.8
111 400 52.6
Power 104 100 62.4
104 200 54.8
104 400 45.6
Ship (Hanjin
Marseilles)
21
knots
107 100 65.4
107 200 57.8
442 48
Ambient (measured on day after MV Hanjin Marseilles pass) 26
39
Transmission loss was assumed to be spherical which is likely only accurate in deep
water. The range of distances between whales and vessel noise is pertinent based on the
voluntary “Be Whale Wise” guidelines, which included maintaining cruising speeds of less than
7 knots (13 km/hour) within 400 m of the whales and a “no go” zone within 100 m of the whales
(http://www.nwr.noaa.gov/Marine-Mammals/upload/BeWhaleWise.pdf). Additionally, the
received spectral density level at 50 kHz of a container ship at 442 m was used. The NL that a
killer whale auditory system would receive was calculated as
NL = N0 + BW – DI (11)
where N0 is the noise spectral density of the noise, BW is the received bandwidth, and DI is the
directivity index used in the previous investigation (Au et al. 2004).
This NL was based on RMS measurements, so 9 dB were added to convert to peak-to-
peak levels as in the previous analysis by Au et al. (2004). The maximum horizontal range (m)
was taken as the horizontal distance between a whale at 1 m and a salmon at 65 m of depth in
which the echo levels off the fish would extend into the noise floor. Echo levels were
determined by the same methods as in Au et al. (2004). Ambient levels in the Haro Strait were
those reported by Hildebrand et al. (2006) and are comparable to ambient levels between a sea
state of 1 in deep water (Urick 1983). It was assumed that only the noise from one vessel was
present and other sources, including natural ones, were not included in the calculation of NL.
The results of this analysis are shown in Table 2 and Figure 10.
Compared to ambient levels in the Haro Strait in which the maximum horizontal
detection range was predicted to be 400 m, boat noise generated by cruise and power up speeds
up to 400 m relative to the whales was predicted to significantly reduce the active space of an
echolocation click at 50 kHz. At a distance of 400 m between vessel and whale, Boat B at power
up speed and the Boat C at cruise speed were predicted to have the least impact (maximum
horizontal detection range was 250 m, resulting in a 150 m reduction in active space compared to
Haro Strait ambient levels), while Boat D at cruise was predicted to have the greatest impact that
was similar to the masking effects of heavy rain (maximum detection range was 40 m, thus the
active space was effectively reduced by 360 m).
The predicted distances at which noise produced by cruising (> 20 knots) whale watching
vessels would approach ambient levels in Haro Strait are 650, 950, and 1,400 m for Boat C, Boat
B, and Boat D, respectively. The zones of masking for these vessel types are defined within
these ranges. Even the large container ship (MV Hanjin Marseilles) passing at 442 m relative to
a whale was predicted to reduce the active space of the biosonar signal by 340 m relative to Haro
Strait ambient conditions.
The above analysis is oversimplified because it only considers one frequency and
echolocation clicks are broadband. Whales most likely use acoustic information across a range
of frequencies to detect prey through biosonar. Additionally, as stated by Au et al. (2004), the
DI assumes that the noise is isotropic (independent of direction) and vessel noise at this range is
likely anisotropic. For example, the whales might be able to swim to some depth to avoid vessel
noise at the surface, but such behavior might also have energetic costs. The target strength of a
40
Table 2. Predicted maximum horizontal detection ranges at 50 kHz for a killer whale at the surface
echolocating on a Chinook salmon at 65 m depth, reduction in range relative to ambient Haro
Strait measurement, and percent of reduction in range relative to ambient Haro Strait
measurement.
Noise condition
Horizontal echo
detection range
(m)
Reduction in
range re ambient
Haro Strait (m)
% reduction in
range re ambient
Haro Strait
No noise 650 – –
Sea state of 1 450 – –
Ambient Haro Strait 400 – –
Heavy rain 40 – –
Boat B cruise at 100 m 20 380 95
Boat B cruise at 200 m 40 360 90
Boat B cruise at 400 m 100 300 75
Boat B power at 100 m 50 350 88
Boat B power at 200 m 100 300 75
Boat B power at 400 m 250 150 38
Boat C cruise at 100 m 40 360 90
Boat C cruise at 200 m 100 300 75
Boat C cruise at 400 m 250 150 38
Boat C power at 100 m 40 360 90
Boat C power at 200 m 80 320 80
Boat C power at 400 m 200 200 50
Boat D cruise at 100 m 0 400 100
Boat D cruise at 200 m 20 380 95
Boat D cruise at 400 m 40 360 90
Boat D power at 100 m 10 390 98
Boat D power at 200 m 30 370 93
Boat D power at 400 m 80 320 80
Hanjin (ship) at 100 m 0 400 100
Hanjin (ship) at 200 m 20 380 95
Hanjin (ship) at 442 m 60 340 85
Chinook salmon was modeled using tonal sonar signals and a model of a fish (Au et al. 2004)
and not with killer whale-like echolocation signals and real fish to determine target strength
values (see Au et al. 2007). Furthermore, information on target detection thresholds in killer
whales in quiet conditions is lacking, so interpreting appropriate detection ranges predicted from
this analysis is unknown.
The range at which killer whales would detect, discriminate, and consider pursuing prey
is debatable as discussed in Au et al. (2004). For example, even if the detection of prey is
predicted at 400 m in quiet ambient conditions (Figure 10), it does not necessarily mean that
whales will choose to pursue the prey. For these reasons, the data from this analysis should be
interpreted with caution, given the lack of information about predator-prey interactions in free-
ranging killer whales.
41
0
50
100
150
200
250
300
350
400
450
500
Sea State of 1
Ambient Haro Strait
Heavy Rain
Boat B cruise at 100m
Boat B cruise at 200m
Boat B cruise at 400m
Boat B power at 100m
Boat B power at 200m
Boat B power at 400m
Boat C cruise at 100m
Boat C cruise at 200m
Boat C cruise at 400m
Boat C power at 100m
Boat C power at 200m
Boat C power at 400m
Boat D cruise at 100m
Boat D cruise at 200m
Boat D cruise at 400m
Boat D power at 100m
Boat D power at 200m
Boat D power at 400m
Hanjin (ship) at 100m
Hanjin (ship) at 200m
Hanjin (ship) at 442m
Max. Horizontal Detection Range (m)ion Range (m)
) Max. horizontal detection range (m
Noise Condition
Figure 10. Predicted maximum horizontal detection ranges at 50 kHz for a killer whale at the surface
echolocating on a Chinook salmon at 65 m depth for various noise conditions.
Figure 10. Predicted maximum horizontal detection ranges at 50 kHz for a killer whale at the surface
echolocating on a Chinook salmon at 65 m depth for various noise conditions.
Noise condition
Zone of Hearing Loss or Injury Zone of Hearing Loss or Injury
The zone of hearing loss or injury is the area where received levels or sound exposure
levels are high enough to cause temporary or permanent hearing loss. Studies on TTS in various
marine mammal groups are needed (NRC 2003) because of their utility to predict auditory injury
or PTS. Hearing loss effectively reduces the active space of important signals such as
conspecific calls or biosonar signals. That is, whales that experience temporary or permanently
elevated hearing thresholds must be closer to a source to detect it. TTS data from studies
conducted on bottlenose dolphins and beluga whales have been used to predict the zones of
hearing loss (TTS) and injury (PTS) in killer whales. In particular, the TTS studies that have
used white noise as fatiguing stimuli (Nachtigall et al. 2003, 2004) are most appropriate for
extrapolating potential hearing loss in killer whales due to vessel noise exposure, those using
pure tone fatiguing stimuli (Finneran et al. 2005, 2007) are most appropriate for sonar exposure,
and those using impulsive sounds (Finneran et al. 2000, 2002b) are most appropriate for pile
driving activity.
The zone of hearing loss or injury is the area where received levels or sound exposure
levels are high enough to cause temporary or permanent hearing loss. Studies on TTS in various
marine mammal groups are needed (NRC 2003) because of their utility to predict auditory injury
or PTS. Hearing loss effectively reduces the active space of important signals such as
conspecific calls or biosonar signals. That is, whales that experience temporary or permanently
elevated hearing thresholds must be closer to a source to detect it. TTS data from studies
conducted on bottlenose dolphins and beluga whales have been used to predict the zones of
hearing loss (TTS) and injury (PTS) in killer whales. In particular, the TTS studies that have
used white noise as fatiguing stimuli (Nachtigall et al. 2003, 2004) are most appropriate for
extrapolating potential hearing loss in killer whales due to vessel noise exposure, those using
pure tone fatiguing stimuli (Finneran et al. 2005, 2007) are most appropriate for sonar exposure,
and those using impulsive sounds (Finneran et al. 2000, 2002b) are most appropriate for pile
driving activity.
Erbe (2002) used preliminary data from the Nachtigall et al. (2003) study as cited in Au
et al. (1999) in which a TTS of 12–18 dB at 7.5 kHz occurred in a bottlenose dolphin after
exposure to an octave band of noise at 179 dB re 1µPa for approximately 50 minutes (resulting
in an SEL of 213 dB re 1µPa2s). This corresponded to a sensation level of 96 dB (amount above
the pure tone threshold of the subject at 7.5 kHz). Within a few meters, octave band noise levels
of whale watching boats were usually less than 96 dB above the threshold of killer whales in this
bandwidth. Therefore, Erbe (2002) used data from humans and other terrestrial animals to scale
down the amount of TTS and assumed that exposure levels of 68–74 dB above threshold for the
Erbe (2002) used preliminary data from the Nachtigall et al. (2003) study as cited in Au
et al. (1999) in which a TTS of 12–18 dB at 7.5 kHz occurred in a bottlenose dolphin after
exposure to an octave band of noise at 179 dB re 1µPa for approximately 50 minutes (resulting
in an SEL of 213 dB re 1µPa2s). This corresponded to a sensation level of 96 dB (amount above
the pure tone threshold of the subject at 7.5 kHz). Within a few meters, octave band noise levels
of whale watching boats were usually less than 96 dB above the threshold of killer whales in this
bandwidth. Therefore, Erbe (2002) used data from humans and other terrestrial animals to scale
down the amount of TTS and assumed that exposure levels of 68–74 dB above threshold for the
42
same exposure period would result in a 5 dB threshold shift. At that time, the accuracy of this
assumption was questionable because no TTS was observed in the bottlenose dolphin for
exposure levels of 87 dB above threshold (Au et al. 1999). It was predicted that a TTS of 12–18
dB would result from exposure of a fast inflatable (traveling at 51 km/hour) within a 10 m range
after 30–50 minutes. A TTS of 5 dB was predicted from exposure of the same inflatable over the
same duration within 450 m range. Exposure to a slower moving inflatable (10 km/hour) would
result in TTS of 5 dB after 30–50 minutes within a 20 m range (Erbe 2002).
Based on more current work by Nachtigall et al. (2003), a TTS of 11 dB at 7.5 kHz
would result from exposure to a fast inflatable within a 10 m range after 50 minutes of
continuous exposure. In another study using the same octave band of noise for exposure, an
electrophysiological approach was used to measure postexposure hearing thresholds more
rapidly and over a wider range of frequency. Lower amplitude and shorter duration exposures
produced measurable TTS (Nachtigall et al. 2004) compared to a lack of TTS measured through
behavioral techniques (Nachtigall et al. 2003), but differences in results from these two
approaches might be related to methodological issues that are still being resolved (Finneran et al.
2007). Based on the current data, a killer whale would experience a TTS of 5 dB at
approximately the same frequency (8 kHz) and a TTS of 8 dB at 16 kHz after 30 minutes of
continuous exposure within a range from the vessel corresponding to 70 dB above hearing
sensitivity.
Based on TTS results in which a 3 kHz pure tone was used for exposure, onset TTS one-
half octave higher than the frequency of exposure (4.5 kHz) was predicted at sound exposure
levels of 195 dB re 1µPa2s (Finneran et al. 2005). Killer whales, however, are about 10–15 dB
more sensitive than bottlenose dolphins within this frequency range, but these behavioral results
are only based on a one or two individuals (Johnson 1967, Szymanski et al. 1999). When
differences in hearing sensitivity are taken into account, sound exposure levels that are predicted
to result in killer whale TTS onset should be considered to be lower by approximately 10–15 dB
within this frequency range. This is a conservative estimate given the lack of direct
measurements of TTS in killer whales.
Pile driving activity in Friday Harbor had an average SEL that ranged between 171 and
187 dB re 1 µPa2s (Lauglin 2005). Exposure to a seismic water gun resulted in masked TTS of 7
and 6 dB at 0.4 and 30 kHz, respectively but no TTS greater than or equal to 6 dB at 4 kHz
occurred in a beluga whale after a comparable SEL of 186 dB re 1 µPa2s (peak-to-peak pressures
of 226 dB re 1 µPa). No MTTS greater than or equal to 6 dB was observed in the bottlenose
dolphin at 0.4, 4, or 30 kHz up to SELs of 188 dB re 1 µPa2s (peak-to-peak pressures of 228 dB
re 1 µPa, Finneran et al. 2002b). While the absolute thresholds of the bottlenose dolphin and
beluga whale are very similar at 0.4 and 4 kHz, the beluga whales hearing is better by about 10
dB at 30 kHz (Johnson 1967, White et al. 1977, Awbry et al. 1988, Johnson et al. 1989). The
killer whale’s threshold at 4 kHz is about 18 dB better while at 30 kHz is 5 dB worse than the
beluga whale’s thresholds (Syzmanski et al. 1999). No absolute threshold measurements at 0.4
kHz are available in the killer whale. With these sensation level differences in mind, killer
whales exposed to the same stimulus might not exhibit a threshold shift greater than or equal to
6 dB at 30 kHz but might experience TTS at 4 kHz. The magnitude of TTS at this frequency
cannot be predicted with accuracy without further data.
43
The range of auditory injury or PTS in killer whales exposed to boat noise has been
predicted by Erbe (2002). An approach based on those used for humans was taken in which a
weighting function used to predict PTS from broadband exposures was modeled. The killer
whale weighting function was taken as the killer whale audiogram subtracted from the critical
band levels of noise and the energy was integrated across all frequencies. A PTS of 2–5 dB was
predicted for whales exposed within 1 km of a fast inflatable or within 50 m of a slow inflatable,
continuously for an 8 hour day, 5 days a week, over 50 years (Erbe 2002). Another approach to
predict PTS onset from TTS data is to extrapolate curves of TTS as a function of weighted and
unweighted levels (depending on the exposure type) up to a value that would likely be associated
with PTS in terrestrial mammals (i.e., TTS ≥ 40 dB). This approach is being considered for a
number of groups based on hearing function in order to define noise exposure criteria for free-
ranging marine mammals (Southall et al. 2007).
44
Likelihood of Acoustic Impacts on the
SRKW Population
Determining whether sound exposure results in measurable differences in hearing or
behavior at the individual or group level has been explored in a few species (Richardson et al.
1995, Nowacek et al. 2007, Finneran et al. 2007). Determining whether the effect is biologically
significant, that is, having overall deleterious effects at the level of the population, is difficult
(NRC 2005). Currently available information is sparse enough to preclude an accurate
assessment of the deleterious effects of sound exposure and related cumulative effects at the
individual as well as the population level for Southern Resident killer whales. However, the
following discussion, based on the framework provided in NRC (2005), may serve as an outline
for recommendations for future work (discussed in the next section) while suggesting some
provisional approaches for acoustic impact assessment in the interim period.
Risk Assessment
One approach is that based on risk assessment. Risk assessment requires that the
problems are clearly identified and the probability of exposure and the types of biological effects
are defined, leading to an estimate of risk. There are many types of potential biological effects,
but the focus should be on those that have repercussions on life functions that ultimately affect
population status as outlined in the Population Consequences of Acoustic Disturbance model
(NRC 2005).
Both the risk of behavioral disturbance and auditory injury at the population level require
specific information regarding the type, amplitude, and duration of the exposure; the range over
which the exposure might affect free-ranging animals including sound propagation properties;
the fraction of the population affected (e.g., through dose-response curves); the intensity of the
reaction; and the ease with which animals of the population might recovery from the exposure
(such as finding alternative habitats or recovering hearing function).
In many cases, it is very difficult to interpret the biological significance of behavioral
disturbances. For example, does the disturbance result from the acoustic signal resembling the
sounds of a predator, prey, or conspecific? Or is the response just indicative of annoyance that
scales to exposure level? It is often impossible to determine differences in the reasons why
animals respond in a particular way because of a lack of power to determine causal relationships
between exposure and behavioral changes (NRC 2005). For healthy populations, recovery from
sound exposure disturbance might be relatively easy, having no long-term effects on that
population. However, for small populations that are constrained by other factors that affect
survival, such as food availability, recovery from sound exposure disturbance that affects
foraging success, for example, might not be as easy.
45
Related to these considerations is the area affected versus the habitat available
(Richardson et al. 1995). The core habitat of Southern Residents is particularly concentrated on
the west side of San Juan Island during the summer (Hauser 2006), and it is considered a central
area for feeding (Baird and Hanson 2004, Hanson et al. in prep.). This is also an area where few
or no parts of the range are absent of human-produced noise. Broadband received levels (0.1–15
kHz) were more than 120 dB re 1 µPa approximately 50% of the time during summer days in
2005 in this area due to vessel traffic (Veirs and Veirs 2005). This broadband level has been
used to estimate zones of behavioral responses in resident killer whales without direct empirical
evidence (Erbe 2002).
Assuming these measurements are representative of summer exposure levels on a regular
basis and that Southern Residents would show a behavioral response to these exposure levels,
then it is likely that behavioral effects of noise exposure would be manifested on a regular basis.
However, these broadband SPLs reported by Veirs and Veirs (2005) do not include the
frequency range of best hearing sensitivity of killer whales (18–42 kHz, Syzmanski et al. 1999).
Extent of Masking Effects
The masking effects of sound exposure on communication signals (pulsed calls) are
likely to extend up to several kilometers away in some conditions. While whale watching
vessels may shut down when within a few hundred meters of the whales, they commonly motor
around the area adjacent to the whales and within the zone of masking when arriving,
repositioning, or exiting the area. Other vessels, such as larger ships and fishing vessels, as well
as sound from natural sources contribute to particularly high levels of ambient noise in SRKW
summer habitat. Thus masking of communication signals is probably a consistent challenge for
Southern Residents. Pulsed calls are most often heard while killer whales are foraging and
traveling (Ford 1989, Miller 2002). Some individuals, particularly females, share prey items
(Ford and Ellis 2006) and calls used for communication during cooperative foraging could be
masked by a variety of sound sources.
The levels of many anthropogenic sounds, such as vessel traffic, are strongest at
frequencies below 10 kHz (Figure 5). A common assumption is that masking of echolocation
signals is not much of a concern compared to communication signals because echolocation
signals are strongest above 20 kHz (Au et al. 2004). However, killer whale hearing is tuned to
higher frequencies and whales probably need to hear faint echoes from ensonified objects. Thus
any sounds that increase background levels at higher frequencies above ambient levels have the
potential to decrease the range at which echolocation activities are effective, including those
associated with foraging.
The analyses of active space of echolocation signals in the presence of both natural and
anthropogenic sources clearly illustrate potential impacts in terms of the reductions in the range
at which killer whales can detect prey items. The fact that whale watching vessels idling at 200
m produced source spectral levels that were comparable to ambient levels (Hildebrand et al.
2006) suggests that an idling vessel at this distance alone would not reduce the active space of
echolocation signals. However, this assumption needs to be investigated further. According to
the “Be Whale Wise” guidelines, 400 m is the shortest distance where vessels can power up to
normal cruising speeds relative to whales. At this range, vessel noise reduced the horizontal
46
detection range of a 50 kHz echolocation signal by as much as 360 m (from 400 to 40 m) relative
to ambient conditions. The distance at which noise from vessel operations would be reduced to
ambient is considerably larger than 400 m for the vessel measured by Hildebrand et al. (2006).
Additional noise level measurements of vessels of different propulsion types operating at slower
speeds (less than 7 knots) in SRKW habitat are needed to evaluate potential masking effects
further, particularly those that include high frequency measurements.
Most studies that aim to identify biological effects of noise exposure do not consider
cumulative acoustic effects. Even if only a very small number of individuals are exposed to
sounds that have the potential for auditory injury, other individuals in the area might still be
affected by masking or behavioral disturbance, given the larger range at which a sound source
would have an impact. Thus the assessment of sound exposure impacts should not ignore less
severe events. For example, if a severity score is calculated based on models used for
determining the Potential for Biological Removal, the scores must reflect the cumulative effects
to determine total number of individuals affects and the effects on the population should be
determined from this total number (NRC 2005).
Furthermore, there are other auditory processes besides detection that are affected by
noise exposure such as sound localization, discrimination, and recognition. For example, if a
killer whale detects the call of a conspecific but cannot localize or recognize the call, then the
function of the call for social cohesion may be impaired. There are very few studies on the
effects of noise on these types of auditory processes in marine mammals. Nonetheless, these
types of impacts should also be considered.
Interaction with Nonacoustic Variables
Sound exposure effects might also interact with nonacoustic variables that have been
identified as possible factors related to the population decline of SRKWs. For example, when
they are within 100–400 m of vessels, there is evidence that they switch from foraging behaviors
to other behaviors such as traveling (Bain et al. 2006). It is unknown whether this behavioral
change is in response to the presence of the vessels, to the noise produced by vessels, to other
variables, or some combination of these factors. In addition, most vessels, such as those used for
whale watching, operate during the day and there is evidence that foraging behavior is also
diurnally dependent (Baird et al. 2005). Thus Southern Residents might be temporally and
spatially restricted for activities related to foraging. There are cases in which individuals
including those with dependent offspring have physical signs of poor body condition before they
disappear and are assumed to have died. These observations indicate that starvation is a threat.
Many marine mammals go through periods when food is less abundant. Entering such
periods with insufficient body reserves can have significant biological effects at the level of the
individual as well as for an endangered population. Acoustic variables that affect caloric intake
in SRKWs would affect population recovery. Other anthropogenic effects might also interact
with the metabolic effects of periods of low caloric intake such as the mobilization of lipophilic
contaminants. For example, fasting northern elephant seals (Mirounga angustirostris) showed
higher levels of mobilized polychlorinated biphenyls (PCBs) into the bloodstream during the end
of their postweaning fast compared to before the fast. Additionally, PCB concentrations in the
blood were higher in leaner animals, suggesting that they might be more at risk to potential toxic
47
effects (Debier et al. 2006). The interaction of acoustic effects with other threats on the SRKW
population is of significant concern. However, it is extremely difficult to link potentially small
but consistent changes in response to noise exposure to population level effects.
48
Recommendations for Future Work
Future studies that would increase confidence in predicting potential acoustic impacts on
SRKWs include those that accurately measure source characteristics and sound propagation
effects, particularly of anthropogenic sounds in core habitat. For example, knowing the source
level, frequency content, and typical duration of exposure can enable better predictions of what
the received levels might be at the whales’ location. Because the propagation of vessel sounds is
not equal in all directions, a study that measured propagation from the bow, sides, and stern of
the vessel would be able to determine if received levels are higher for a particular orientation and
how this might interact with directional hearing and the ability to detect signals in noise. Such
information could be used to mitigate killer whale and whale watching vessel interactions,
particularly with respect to maneuvering around the whales.
Killer whale hearing sensitivity has been described for frequencies between 1 and 100
kHz. The whales will likely hear a sound if received levels are above the hearing threshold of
the animal and if background levels are sufficiently low enough to allow it. The assessment of
the zones of audibility and masking are contingent on knowledge of the auditory bandwidths of
killer whales. As a first approach, one-third octave levels could be used but direct measurements
of critical bandwidths across a wide range of frequencies within the hearing range is necessary to
assess the validity of using one-third octave band analysis to describe sound sources and assess
their impacts. In the interim, it may be useful to bracket the potential effects by using an upper
and lower range of probable critical bandwidths of killer whales, such as one-third and one-
twelfth octave analyses (e.g., see Southall 2003b). For these reasons, studies that include noise
level measurements in SRKW habitat should report sound pressure levels in broadband, one-
third octave band, one-twelfth octave band, and spectral density levels whenever possible.
Masking Effect Assessment
Further assessment of masking effects on echolocation signals would be useful,
particularly for more types of whale watching vessels as well as other vessels in core habitat.
Because the energy of most noise produced by vessels is concentrated at the lower frequencies, it
is tempting to assume that the amount of noise generated at the higher frequencies will not affect
killer whales. This assumption might be based in part on an anthropomorphic bias, since humans
do not hear functionally above 20 kHz yet laboratory evidence shows that killer whales are most
sensitive at and above this frequency. It is necessary to consider the whales’ subjective
perceptual world (the “Umwelt”) that they experience.
Currently, there are very few available measurements of ambient noise levels above 20
kHz (see Hildebrand et al. 2006). The existing data suggests that maintaining slow cruise speeds
within 400 m of the whales and shutting vessel engines down completely when within 100 m of
a whale is sufficient to reduce the effects of masking of echolocation signals used for foraging.
Predicting zones of masking will depend on the vessel type, operating speed, and propulsion type
as well as the number of vessels actively motoring. Whale watching vessels operating at normal
49
cruise speeds even beyond 400 m are predicted to reduce the active space of the 50 kHz
echolocation signal, but there is difficulty in interpreting appropriate detection ranges. Further
information about noise levels above 20 kHz for vessels of different sizes, propulsion types, and
operating speeds, particularly for speeds less than 7 knots, would be useful to further assess the
impacts of anthropogenic sounds on killer whales and how regulatory guidelines such as those
based on “Be Whale Wise” may or may not be effective in reducing such potential impacts on
animals using higher frequency biosonar for prey detection.
Information about the target strength of key prey items such as various Pacific salmon
species using more realistic signals as well as investigations on predator-prey interactions in
free-ranging killer whales are needed to predict the potential masking impacts of anthropogenic
sources on echolocation and foraging behavior. Studies that focus on foraging efficiency and
sound production associated with foraging in the presence and absence of anthropogenic noise
would provide additional information that could be used for this assessment.
Behavioral Response
Predicting the zone of responsiveness is dependent on knowing the quantitative aspects of
the exposure relative to the whales. As is the case for most cetacean species (see Nowacek et al.
2007), there is either a lack of information altogether or too broad a range of received levels that
are potentially correlated with behavioral responses to provide confidence regarding what future
exposure might result in behavioral avoidance or displacement in SRKWs. A broadband
received level of 120 dB re 1 µPa without respect to frequency content and the sensitivity of the
animals has sometimes been used as a generalized rule of thumb to define the onset of a
behavioral response to noise (e.g., Erbe 2002). Two-second SPL averages were greater than or
equal to 120 dB (re 1 µPa measured from 0.1 to 15 kHz) 50% of the time during summer days in
Haro Strait (Veirs and Veirs 2005).
One interpretation is that killer whales would potentially behaviorally or physiologically
respond to sounds in the areas associated with summer foraging at least 50% of the time. This
assumption requires further empirical evaluation, particularly since killer whales are most
sensitive at frequencies between 18 and 42 kHz. Studies that investigate behavioral responses to
vessel traffic should at the least report received sound levels, if not source levels, and preferably
sound exposure levels to address the temporal aspects of the exposure. Moreover, it is unknown
whether behavioral responses to vessel effects result from the presence of the vessel, the sounds
it produces, or some combination of these factors. Future studies that tease apart these effects,
such as controlled exposure experiments in which recordings of the vessel sounds are played
back in the absence of the vessel to determine behavioral changes, are also needed to make more
accurate predictions of the zone of responsiveness.
Likelihood of Auditory Injury
Determining the likelihood of auditory injury is probably one of the more imperative
assessments of sound exposure in SRKWs. Even if the potential for such exposure is low
relative to other possible acoustic risk factors, the uncertainty in predicting this type of auditory
effect in an ESA-listed species warrants further research efforts. Auditory injury is usually
estimated from studies on TTS because inflicting permanent hearing loss in marine mammals is a
50
significant violation of the Marine Mammal Protection Act of 1972. No studies have measured
TTS in killer whales but a few well documented studies have measured TTS in bottlenose
dolphins and beluga whales. From these studies, TTS onset has been estimated at a sound
exposure level of 195 dB re 1 µPa2s for midrange exposure and test frequencies. Given the
sensation level differences between killer whales, bottlenose dolphins, and beluga whales at
these frequencies, this onset SEL criterion might not be conservative enough to prevent auditory
injury in killer whales. TTS studies in killer whales would provide direct evidence to support or
refute these assumptions.
Research focused on assessing potential population effects of sound exposure is probably
the most important (albeit most challenging) research priority. Dose-response curves, in which
the percentage of individuals that show responses to sound exposure such as behavioral
disturbance or onset-TTS as a function of sound exposure level, are needed to make accurate
assessments at the level of the population (NRC 2005). Additionally, studies that measure equal
loudness contours in killer whales would be extremely useful to determine weighting functions
that could be used when measuring ambient noise in order to better predict behavioral and
auditory disturbance. An approximation to this would be to use a weighting function based on
the killer whale audiogram.
Finally, there is a need for more information in published, peer-reviewed formats. For
example, a working group is in the process of publishing noise exposure criteria for marine
mammals that includes the development of interim weighting functions for different groups.
These groups are categorized based on hearing abilities given the variety of auditory sensitivities
at different frequencies across the different marine mammal taxa (Southall et al. 2007). Such
documents would provide tools for policy makers and natural resource managers to make more
confident decisions about potential acoustic impacts from human activity on SRKWs and other
marine mammal populations.
51
Conclusions
This report offers the following conclusions on sound exposure and SRKWs:
• Given the chronic nature of vessel traffic in areas designated as critical habitat of SRKWs
and the potential for vessel noise to mask communication and echolocation signals of
SRKWs, more measurements of these sounds within the relevant frequency range of
killer whale hearing (i.e., 1–100 kHz) are needed. Such studies should further investigate
the effects of vessel size, propulsion type, operating speed, and vessel orientation on the
sound levels emitted.
• Some vessel noise, depending on the vessel size, propulsion type, and operating speed,
was predicted to significantly reduce the range at which echolocating killer whales could
detect salmon in the water column. Interpretations of these effects are limited by a dearth
of information on predator-prey interactions, such as the typical ranges at which killer
whales detect and pursue salmon in quiet conditions. Studies that elucidate this
information would enable a more accurate assessment of the impacts of vessel noise as
well as other anthropogenic sources on the foraging success of echolocating killer
whales.
• The ability to assess zones of audibility and masking are dependent on knowing the
appropriate bandwidth to quantify anthropogenic sounds. The effective filter bandwidth
of the killer whale auditory system is unknown and has been estimated through indirect
methods. However, these methods are often not accurate compared to direct
measurements. Direct measurements of killer whale auditory filter bandwidths are
needed to provide more confidence in assessing impacts of anthropogenic sounds.
• Assessment of auditory injury in killer whales is estimated through surrogate species such
as the bottlenose dolphin. However, differences in hearing sensitivity between the two
species may result in inaccurate assessments of TTS in killer whales. Investigations that
measure TTS directly in killer whales would provide very valuable data to assess the
potential of auditory injury from sound exposure.
• Predicting the zone of responsiveness is dependent on knowing the quantitative aspects of
the exposure relative to the whales. There is either a lack of or too broad a range of
received levels that are potentially correlated with behavioral responses to provide
confidence regarding what future exposure might result in behavioral effects in SRKWs.
Investigations that quantify both the received levels of the sound exposure (using
appropriate metrics and within the appropriate frequency range) and the behavioral
response of free-ranging killer whales are needed to assess zones of responsiveness.
These studies should ideally measure the proportion of animals affected from controlled
exposures.
52
References
Ainslie, M. A., and J. G. McColm. 1998. A simplified formula for viscous and chemical absorption in
sea water. J. Acoust. Soc. Am. 103:1671–1672.
Au, W. W. L., and P. W. B. Moore. 1984. Receiving beam patterns and directivity indices of the Atlantic
bottlenose dolphin, Tursiops truncatus. J. Acoust. Soc. Am. 75:255–258.
Au, W. W. L., and P. W. B. Moore. 1990. Critical ratio and critical bandwidth for the Atlantic bottlenose
dolphin. J. Acoust. Soc. Am. 88:1635–1638.
Au, W. W. L., D. A. Carder, R. H. Penner, and B. L. Scronce. 1985. Demonstration of adaptation in
beluga whales echolocation signals. J. Acoust. Soc. Am. 77:726–730.
Au, W. W. L., R. H. Penner, and J. Kadane. 1982. Acoustic behavior of echolocating Atlantic bottlenose
dolphins. J. Acoust. Soc. Am. 71:1269–1275.
Au, W. W. L., P. E. Nachtigall, and J. L. Pawlowski. 1999. Temporary threshold shift in hearing induced
by an octave band of continuous noise in the bottlenose dolphin. J. Acoust. Soc. Am.
106:2251(A).
Au, W. W. L., J. K. B. Ford, J. K. Horne, and K. A. Newman Allman. 2004. Echolocation signals of
free-ranging killer whales (Orcinus orca) and modeling of foraging for Chinook salmon
(Oncorhynchus tshawytscha). J. Acoust. Soc. Am. 115:901–909.
Au, W. W. L., K. J. Benoit-Bird, and R. A. Kastelein. 2007. Modeling the detection range of fish by
echolocating bottlenose dolphins and harbor porpoises. J. Acoust. Soc. Am. 121:3954–3962.
Awbry, F. T., J. A. Thomas, and R. A. Kastelein. 1988. Low frequency underwater hearing sensitivity in
belugas (Delphinapterus leucas). J. Acoust. Soc. Am. 84:2273–2275.
Bain, D. E., and M. E. Dahlheim. 1994. Effects of masking noise on detection thresholds of killer
whales. In T. R. Loughlin (ed.), Marine mammals and the Exxon Valdez, p. 243–257. Academic
Press, San Diego, CA.
Bain, D. E., R. Williams, J. C. Smith, and D. Lusseau. 2006. Effects of vessels on behavior of Southern
Resident killer whales (Orcinus spp.) 2003–2005. NMFS Contract AB133F05SE3965.
(Available from D. E. Bain, Friday Harbor Laboratories, Univ. Washington, 620 University
Road, Friday Harbor, WA 98250.)
Baird, R. W., M. B. Hanson, and L. M. Dill. 2005. Factors influencing the diving behaviour of fish-
eating killer whales: Sex differences and diel and interannual variation in diving rates. Can. J.
Zool. 83:257–267.
Baird R. W., and M. B. Hanson. 2004. Diet studies of Southern Resident killer whales: Prey sampling
and behavioral cues of predation. NMFS Contract AB133F-03-SE-1070. (Available from
Cascadia Research Collective, 218½ W. Fourth Ave., Olympia, WA 98501.)
53
Barrett-Lennard, L. G., J. K. B. Ford, and K. A. Heise. 1996. The mixed blessings of echolocation:
Differences in sonar use by fish-eating and mammal-eating killer whales. Anim. Behav. 51:553–
565.
Brumm, H., and H. Slabbekoorn. 2005. Acoustic communication in noise. In P. J. B. Slater, C. T.
Snowden, H. J. Borckmann, T. J. Roper, and M. Naguib (eds.), Advances in the study of
behavior, Vol. 35, p. 151–207. Academic Press, San Diego, CA.
Crum, L. A., and Y. Mao. 1996. Acoustically enhanced bubble growth at low frequencies and its
implications for human diver and marine mammal safety. J. Acoust. Soc. Am. 99:2898–2907.
Debier, C., C. Chalon, B. J. Le Boeuf, T. deTillesse, Y. Larondelle, and J. Thome. 2006. Mobilization of
PCBs from blubber to blood in northern elephant seals (Mirounga angustirostris) during the post-
weaning fast. Aquat. Toxicol. 80:149–157.
Deecke, V. B., J. K. B. Ford, and P. J. B. Slater. 2004. The vocal behaviour of mammal-eating killer
whales: Communicating with costly calls. Anim. Behav. 69:395–405.
Deecke, V. B., P. J. B. Slater, and J. K. B. Ford. 2002. Selective habituation shapes acoustic predator
recognition in harbour seals. Nature 420:171–173.
Erbe, C. 2002. Underwater noise of whale-watching boats and potential effects on killer whales (Orcinus
orca), based on an acoustic impact model. Mar. Mamm. Sci. 18:394–418.
Erbe, C., and D. M. Farmer. 2000. A software model to estimate zones of impact on marine mammals
around anthropogenic noise. J. Acoust. Soc. Am. 108:1327–1331.
Fay, R. R. 1988. Hearing in vertebrates: A psychophysical databook, Hill-Fay Associates, Winnetka, IL.
Fernández, A., M. Arbelo, R. Deaville, I. A. P. Patterson, P. Castro, J. R. Baker, E. Degollada, H. M.
Ross, P. Herráez, A. M. Pocknell, E. Rodríguez, F. E. Howie, A. Espinosa, R. J. Reid, J. R. Jaber,
V. Martin, A. A. Cunningham, and P. D. Jepson. 2004. Pathology: Whales, sonar and
decompression sickness (reply). Nature 428 (15 April 2004).
Fernandez, A., J. F. Edwards, F. Rodriguez, A. Espinosa de los Monteros, P. Herraez, P. Castro, J. R.
Jaber, V. Martin, and M. Arbelo. 2005. “Gas and fat embolic syndrome” involving a mass
stranding of beaked whales (family Ziphiidae) exposed to anthropogenic sonar signals. Vet.
Pathol. 42:446–457.
Finneran, J. J., C. E. Schlundt, D. A. Carder, J. A. Clark, J.A. Young, J.B. Gaspin, and S. H. Ridgway.
2000. Auditory and behavioral responses of bottlenose dolphins (Tursiops truncatus) and a
beluga (Delphinapterus leucas) to impulsive sounds resembling distant signatures of underwater
explosions. J. Acoust. Soc. Am. 108:417–431.
Finneran, J. J., C. E. Schlundt, D. A. Carder, and S. H. Ridgway. 2002a. Auditory filter shapes for the
bottlenose dolphin (Tursiops truncatus) and the white whale (Delphinapterus leucas) derived
with notched noise. J. Acoust. Soc. Am. 112:322–328.
Finneran, J. J., C. E. Schlundt, C.E. Dear, D. A. Carder, and S. H. Ridgway. 2002b. Temporary shift in
masked hearing thresholds in odontocetes after exposure to single underwater impulses from a
seismic water gun. J. Acoust. Soc. Am. 111:2929–2940.
Finneran, J. J., D. A. Carder, C. E. Schlundt, and S. H. Ridgway. 2005. Temporary threshold shift in
bottlenose dolphins (Tursiops truncatus) exposed to mid-frequency tones. J. Acoust. Soc. Am.
118:2696–2705.
54
Finneran, J. J., C. E. Schlundt, B. Branstetter, and R. L. Dear. 2007. Assessing temporary threshold shift
in a bottlenose dolphin (Tursiops truncates) using multiple simultaneous auditory evoked
potentials. J. Acoust. Soc. Am. 122:1249–1264.
Fletcher, H. 1940. Auditory patterns. Rev. Mod. Phys. 12:47–65.
Foote, A. D. 2005. Correlates of variability in killer whale stereotyped call repertoires. Master’s thesis.
University of Durham, Durham, UK.
Foote, A. D., R. W. Osborne, and A. R. Hoezel. 2004. Whale-call response to masking boat noise.
Nature 428:910.
Ford, J. K. B. 1989. Acoustic behavior of resident killer whales (Orcinus orca) off Vancouver Island,
British Columbia. Can. J. Zool. 67:727–745.
Ford, J. K. B. 1991. Vocal traditions among resident killer whales (Orcinus orca) in coastal waters of
British Columbia. Can. J. Zool. 69:1454–1483.
Ford, J. K. B., and G. M. Ellis. 1999. Transients: Mammal-hunting killer whales of British Columbia,
Washington, and southeastern Alaska. University of British Columbia Press, Vancouver.
Ford, J. K. B., and G. M. Ellis. 2006. Selective foraging by fishing-eating killer whales (Orcinus orca)
in British Columbia. Mar. Ecol. Prog. Ser. 316:185–199.
Ford, J. K. B., G. M. Ellis, and K. C. Balcomb. 2000. Killer Whales, 2nd edition. University of British
Columbia Press, Vancouver.
Hall, J. D., and C. S. Johnson. 1972. Auditory thresholds of a killer whale Orcinus orca Linnaeus. J.
Acoust. Soc. Am. 51:515–517.
Hanson, M. B., R. W. Baird, C. Emmons, J. Hempelmann, G. S. Schorr, J. Sneva, D. Van Doornik, K.
Ayres, S. K. Wasser, K. C. Balcomb, K. Balcomb-Baratok, and M. J. Ford. In prep. Species and
stock identification of prey selected by endangered “Southern Resident” killer whales in their
summer range. (Available from M. B. Hanson, Northwest Fisheries Science Center, 2725
Montlake Blvd. E., Seattle, WA 98112.)
Hauser, D. D. W. 2006. Summer space use of Southern Resident killer whales (Orcinus orca) within
Washington and British Columbia inshore waters. Master’s thesis. Univ. Washington, Seattle.
Hildebrand, J., M. McDonald, J. Calambokidis, and K. Balcomb. 2006. Whale watch vessel ambient
noise in the Haro Strait. NMFS Contract NA17RJ1231. (Available from Univ. California San
Diego, Scripps Institution of Oceanography, Joint Institute for Marine Observations, 9500 Gilman
Dr., La Jolla, CA 92093.)
Jepson, P. D., M. Arbelo, R. Deaville, I. A. P. Patterson, P. Castro, J. R. Baker, E. Degollada, H. M. Ross,
P. Herraez, A. M. Pocknell, F. Rodriguez, F. E. Howie, A. Espinosa, R. J. Reid, J. R. Jaber, V.
Martin, A. A. Cunningham, and A. Fernandez. 2003. Gas-bubble lesions in stranded cetaceans.
Nature 425:575–576.
Johnson, C. S. 1967. Sound detection thresholds in marine mammals. In W. N. Tavolga (ed.), Marine
bioacoustics, Vol. 2, p. 247–260. Pergamon, Oxford.
Johnson, C. S. 1968. Relation between absolute threshold and duration-of-tone pulses in the bottlenose
porpoise. J. Acoust. Soc. Am. 43:757–763.
55
Johnson, C. S. 1991. Hearing thresholds for a periodic 60-kHz tone pulses in the beluga whale. J.
Acoust. Soc. Am. 89:2996–3001.
Johnson, C. S., M. W. McManus, and D. Skaar. 1989. Masked tonal hearing thresholds in the beluga
whale. J. Acoust. Soc. Am. 85:2651–2654.
Jones, C. D., and M. A. Wolfson. 2006. Acoustic environment of the Haro Strait: Preliminary
propagation modeling and data analysis. APL-UW Tech. Memo. 03-06. Univ. Washington,
Applied Physics Laboratory, Seattle.
Kastak, D., B. L. Southall, R. J. Schusterman, and C. Reichmuth Kastak. 2005. Underwater temporary
threshold shift in pinnipeds: Effects of noise level and duration. J. Acoust. Soc. Am. 118:3154–
3163.
Krahn, M. M., M. J. Ford, W. F. Perrin, P. R. Wade, R. P. Angliss, M. B. Hanson, B. L. Taylor, G. M.
Ylitalo, M. E. Dahlheim, J. E. Stein, and R. S. Waples. 2004. Status review of Southern
Resident killer whales (Orcinus orca) under the Endangered Species Act. U.S. Dept. Commer.,
NOAA Tech. Memo. NMFS-NWFSC-62.
Lauglin, J. 2005. Underwater sound levels associated with restoration of the Friday Harbor ferry
terminal. Washington Dept. Transportation, Seattle.
Leatherwood, J. D., B. M. Sullivan, K. P. Shepherd, D. A. McCurdy, and S. A. Brown. 2002. Summary
of recent NASA studies of human response to sonic boom. J. Acoust. Soc. Am. 111:586–598.
Lemonds, D. W., W. W. L. Au, P. E. Nachtigall, and H. L. Roitblat. 2000. High-frequency auditory filter
shapes in an Atlantic bottlenose dolphin. J. Acoust. Soc. Am. 108:2614(A).
Lesage, V., C. Barrette, M. C. Kingsley, and B. Sjare. 1999. The effect of vessel noise on the vocal
behavior of belugas in the St. Lawrence River estuary, Canada. Mar. Mamm. Sci. 15:65–84.
Lombard, E. 1911. Le signe de l’elevation de la voix. Ann. Mal. Oreille Larynx 37:101–119.
Madsen, P. T. 2005. Marine mammals and noise: Problems with root mean square sound pressure levels
for transients. J. Acoust. Soc. Am. 117:3952–3957.
Miller, J. H., A. E. Bowles, B. L. Southall, R. L. Gentry, W. T. Ellison, J. J. Finneran, C. R. Greene Jr., D.
Kastak, D. R. Ketten, P. L. Tyack, P. E. Nachtigall, W. J. Richardson, and J. A. Thomas. 2005.
Strategies for weighting exposure in the development of acoustic criteria for marine mammals. J.
Acoust. Soc. Am. 118:2019(A).
Miller, P. J. O. 2002. Mixed-directionality of killer whale stereotyped calls: A direction of movement
cue? Behav. Ecol. Sociobiol. 52:262–270.
Miller, P. J. O. 2006. Diversity in sound pressure levels and estimated active space of resident killer
whale vocalizations. J. Comp. Physiol. A. 192:449–459.
Morton, A. B., and H. K. Symonds. 2002. Displacement of Orcinus orca (L.) by high amplitude sound
in British Columbia, Canada. ICES J. Mar. Sci. 59:71–80.
Mossbridge, J. A., and J. A. Thomas. 1999. An “acoustic niche” for Antarctic killer whale and leopard
seal sounds. Mar. Mamm. Sci. 15:1351–1357.
56
Nachtigall, P. E., J. L. Pawloski, and W. W. L. Au. 2003. Temporary threshold shifts and recovery
following noise exposure in the Atlantic bottlenosed dolphin (Tursiops truncatus). J. Acoust.
Soc. Am. 113:3425–3429.
Nachtigall, P.E., A. Y. Supin, J. L. Pawloski, and W. W. L. Au. 2004. Temporary threshold shifts after
noise exposure in the bottlenose dolphin (Tursiops truncatus) measured using evoked auditory
potentials. Mar. Mamm. Sci. 20:673–687.
NMFS (National Marine Fisheries Services). 2005a. Assessment of acoustic exposures on marine
mammals in conjunction with USS Shoup active sonar transmissions in the eastern Strait of Juan
de Fuca and Haro Strait, Washington, 5 May 2003. NMFS Office of Protected Resources, Silver
Spring, MD.
NMFS (National Marine Fisheries Services). 2005b. Endangered and threatened wildlife and plants:
Endangered status for Southern Resident killer whales. Federal Register [Docket No.
041213348-5285-02, 18 November 2005] 70(217):69903–69912.
NMFS (National Marine Fisheries Services). 2006. Endangered and threatened species; designation of
critical habitat for Southern Resident killer whales. Federal Register [Docket No. 060228057-
6283-02, 29 November 2006] 71(229):69054–69070.
NRC (National Research Council). 2003. Ocean noise and marine mammals. The National Academies
Press, Washington, DC.
NRC (National Research Council). 2005. Marine mammals populations and ocean noise. The National
Academies Press, Washington, DC.
Nowacek, D. P., L. H. Thorbe, D. W. Johnstone, and P. L. Tyack. 2007. Response of cetaceans to
anthropogenic noise. Mamm. Rev. 37:81–115.
Nystuen, J. A. 2006. Marine mammals monitoring for NW Fisheries. NMFS Contract N00024-02-D-
6602. Univ. Washington, Applied Physics Laboratory, Seattle.
Nystuen, J. A., C. C. McGlothin, and M. S. Cook. 1993. The underwater sound generated by heavy
rainfall. J. Acoust. Soc. Am. 93:3169–3177.
Olesiuk, P. F., M. A. Bigg, and G. M. Ellis. 1990. Life history and population dynamics of resident killer
whales (Orcinus orca) in the coastal waters of British Columbia and Washington State. Rep. Int.
Whaling Comm., Spec. Issue 12:209–243.
Piantadosi, C. A., and E. D. Thalmann. 2004. Pathology: Whales, sonar, and decompression sickness.
Nature 428 (15 Apr 2004).
Plomp, R., and M. A. Bouman. 1959. Relation between hearing threshold and duration of tone pulses. J.
Acoust. Soc. Am. 31:749–758.
Renaud, D. L., and A. Popper. 1975. Sound localization by the bottlenose porpoise (Tursiops truncatus).
J. Exp. Biol. 63:569–585.
Renouf, D. 1991. Sensory reception and processing in Phocidae and Otariidae. In D. Renouf (ed.), The
behaviour of pinnipeds, p. 345–394. Chapman and Hall, London.
Richardson W. J., C. R. Greene, C. I. Malme, and D. H. Thomson. 1995. Marine mammals and noise.
Academic Press, San Diego, CA.
57
Romano, T. A., M. J. Keogh, C. Kelly, P. Feng, L. Berk, C. E. Schlundt, D. A. Carder, and J. J. Finneran.
2004. Anthropogenic sound and marine mammal health: Measures of the nervous and immune
systems before and after intense sound exposure. Can. J. Fish. Aquat. Sci. 61:1124–1134.
Scheifele, P. M., S. Andrew, R. A. Cooper, M. Darre, F. E. Musiek, and L. Max. 2005. Indication of a
Lombard vocal response in the St. Lawrence River beluga. J. Acoust. Soc. Am. 117:1486–1246.
Schevill, W. E., and W. A. Watkins. 1966. Sound structure and directionality in Orcinus (killer whale).
Zoologica 51:70–76.
Schlundt, C. E., R. L. Dear, D. A. Carder, and J. J. Finneran. 2006. Growth and recovery of temporary
threshold shifts in a dolphin exposed to mid-frequency tones with durations up to 128 s. J.
Acoust. Soc. Am. 120:3227(A).
Schlundt, C. E., J. J. Finneran, D. A. Carder, and S. H. Ridgway. 2000. Temporary shift in masked
hearing thresholds of bottlenose dolphins, Tursiops truncatus, and white whales, Delphinapterus
leucas, after exposure to intense tones. J. Acoust. Soc. Am. 107:3496–3508.
Schusterman, R. J. 1974. Low false-alarm rates in signal detection by marine mammals. J. Acoust. Soc.
Am. 55:845–848.
Sharf, B. 1970. Critical bands, In J. V. Tobias (ed.), Foundations of modern auditory theory, Vol. I,
p. 159–202. Academic Press, New York.
Simon, M., M. Wahlberg, and L. A. Miller. 2007. Echolocation clicks from killer whales (Orcinus orca)
feeding on herring (Clupea harengus). J. Acoust. Soc. Am. 121:749–752.
Southall, B. L., R. J. Schusterman, and D. Kastak. 2003a. Auditory masking in three pinnipeds: Aerial
critical ratios and direct critical bandwidth measurements. J. Acoust. Soc. Am. 114:1660–1666.
Southall, B. L., R. J. Schusterman, and D. Kastak. 2003b. Acoustic communication ranges for northern
elephant seals (Mirounga angustirostris). Aquat. Mamm. 29:202–213.
Southall, B. L., A. E. Bowles, W. T. Ellison, J. J. Finneran, R. L. Gentry, C. R. Green Jr., D. Kastak, D.
R. Ketten, J. H. Miller, P. E. Nachtigall, W. J. Richardson, J. A. Thomas, and P. L. Tyack. 2007.
Marine mammal noise exposure criteria: Initial scientific recommendations. Aquat. Mamm.
33:414-521.
Szymanski, M. D., D. E. Bain, K. Kiehl, S. Pennington, S. Wong, and K. R. Henry. 1999. Killer whale
(Orcinus orca) hearing: Auditory brainstem response and behavioral audiograms. J. Acoust. Soc.
Am. 106:1134–1141.
Thomas, J. A., R. A. Kastelein, and F. T. Awbrey. 1990. Behavior and blood catecholamines of captive
belugas during playbacks of noise from an oil drilling platform. Zoo Biol. 9:393–402.
Thomsen, F., D. Franck, and J. K. B. Ford. 2001. Characteristics of whistles from the acoustic repertoire
of resident killer whales (Orcinus orca) off Vancouver Island, British Columbia. J. Acoust. Soc.
Am. 109:1240–1246.
Thomsen, F., D. Franck, and J. K. B. Ford. 2002. On the communicative significance of whistles in wild
killer whales (Orcinus orca). Naturwissenschaften 89:404–407.
Thompson, R. K. R., and L. M. Herman. 1975. Underwater frequency discrimination in the bottlenosed
dolphin (1–140 kHz) and the human (1–8 kHz). J. Acoust. Soc. Am. 57:943–948.
58
59
Urick, R. J. 1983. Principles of underwater sound, 3rd edition. McGraw-Hill, New York.
U.S. Navy, Pacific Fleet. 2004. Report on the results of the inquiry into allegations of marine mammal
impacts surrounding the use of active sonar by USS Shoup (DDG 86) in the Haro Strait on or
about 5 May 2003. Online at: http://www.acousticecology.org/docs/SHOUPNavyReport0204.pdf
[accessed 24 January 2008].
Veirs, V., and S. Veirs. 2005. Average levels and power spectra of ambient sound in the habitat of
Southern Resident orcas. Report to NOAA/NMFS/NWFSC. Online at http://www
.coloradocollege.edu/dept/ev/Research/Faculty/OVALItems/pdf_Papers/051204noaa
-haro_noise_final.pdf [accessed 23 January 2008].
Ward, W. D. 1997. Effects of high-intensity sound. In M. J. Crocker (ed.), Encyclopedia of acoustics, p.
1497–1507. John Wiley and Sons, New York.
Wenz, G. M. 1962. Acoustic ambient noise in the ocean: Spectra and sources. J. Acoust. Soc. Am.
34:1936–1956.
White Jr., M. J., J. Norris, D. Ljungblad, K. Baron, and G. di Sciara. 1977. Auditory thresholds of two
beluga whales (Delphinapterus leucas). Tech. Rep. 78-108. Hubbs Sea World Research
Institute, San Diego, CA.
Williams, R., and E. Ashe. 2006. Northern Resident killer whale response to vessels varied with number
of boats. NMFS Contract AB133F04SE0736. (Available from R. Williams, Pearse Island, Box
193, Alert Bay, BC V0N1A0, or E. Ashe, 2103 N. 54th St., Seattle, WA 98103.)
Williams, R., D. Lusseau, and P. S. Hammond. 2006. Estimating relative energetic costs of humans
disturbance to killer whales (Orcinus orca). Biol. Conserv. 133:301–311.
Wursig, B., C. R. Greene Jr., and T. A. Jefferson. 2000. Development of an air bubble curtain to reduce
underwater noise of percussive piling. Mar. Environ. Res. 49:79–93.
Yost, W. A. 2000. Fundamentals of hearing, 4th edition. Academic Press, San Diego, CA.
Yost, W. A., and W. Shofner. 2005. The critical band and the critical ratio. J. Acoust. Soc. Am.
117:2582(A).
Most NOAA Technical Memorandums NMFS-NWFSC are available online at the
Northwest Fisheries Science Center web site (http://www.nwfsc.noaa.gov).
NOAA Technical Memorandum NMFS-NWFSC-
88 Olson, O.P., L. Johnson, G. Ylitalo, C. Rice, J. Cordell, T. Collier, and J. Steger. 2008. Fish habitat use
and chemical contaminant exposure at restoration sites in Commencement Bay, Washington. U.S. Dept.
Commer., NOAA Tech. Memo. NMFS-NWFSC-88, 117 p. NTIS number pending.
87 Keller, A.A., B.H. Horness, V.H. Simon, V.J. Tuttle, J.R. Wallace, E.L. Fruh, K.L. Bosley,
D.J. Kamikawa, and J.C. Buchanan. 2007. The 2004 U.S. West Coast bottom trawl survey of groundfish
resources off Washington, Oregon, and California: Estimates of distribution, abundance, and length composi-
tion. U.S. Dept. Commer., NOAA Tech. Memo. NMFS-NWFSC-87, 134 p. NTIS number pending.
86 Keller, A.A., V.H. Simon, B.H. Horness, J.R. Wallace, V.J. Tuttle, E.L. Fruh, K.L. Bosley,
D.J. Kamikawa, and J.C. Buchanan. 2007. The 2003 U.S. West Coast bottom trawl survey of groundfish
resources off Washington, Oregon, and California: Estimates of distribution, abundance, and length composi-
tion. U.S. Dept. Commer., NOAA Tech. Memo. NMFS-NWFSC-86, 130 p. NTIS number pending.
85 Norman, K., J. Sepez, H. Lazrus, N. Milne, C. Package, S. Russell, K. Grant, R.P. Lewis, J. Primo,
E. Springer, M. Styles, B. Tilt, and I. Vaccaro. 2007. Community profiles for West Coast and North
Pacific fisheries–Washington, Oregon, California, and other U.S. states. U.S. Dept. Commer., NOAA Tech.
Memo. NMFS-NWFSC-85, 602 p. NTIS number pending.
84 Brand, E.J., I.C. Kaplan. C.J. Harvey, P.S. Levin, E.A. Fulton, A.J. Hermann, and J.C. Field. 2007.
A spatially explicit ecosystem model of the California Current’s food web and oceanography. U.S. Dept.
Commer., NOAA Tech. Memo. NMFS-NWFSC-84, 145 p. NTIS number PB2008-102578.
83 Hecht, S.A., D.H. Baldwin, C.A. Mebane, T. Hawkes, S.J. Gross, and N.L. Scholz. 2007. An overview
of sensory effects on juvenile salmonids exposed to dissolved copper: Applying a benchmark concentration
approach to evaluate sublethal neurobehavioral toxicity. U.S. Dept. Commer., NOAA Tech. Memo. NMFS-
NWFSC-83, 39 p. NTIS number PB2008-102577.
82 Helser, T.E., I.J. Stewart, C.E. Whitmire, and B.H. Horness. 2007. Model-based estimates of abundance
for 11 species from the NMFS slope surveys. U.S. Dept. Commer., NOAA Tech. Memo. NMFS-NWFSC-
82, 145 p. NTIS number PB2008-102576.
Recent NOAA Technical Memorandums
published by the
Northwest Fisheries Science Center
