Content uploaded by Spahr Webb
Author content
All content in this area was uploaded by Spahr Webb on Apr 13, 2015
Content may be subject to copyright.
Content uploaded by Spahr Webb
Author content
All content in this area was uploaded by Spahr Webb on May 27, 2014
Content may be subject to copyright.
Blue and fin whales observed on a seafloor array
in the Northeast Pacific
Mark A. McDonald, John A. Hildebrand, and Spahr C. Webb
Scripps Institution of Oceanography, University of California at San Diego, La Jolla, California
92093-0205
(Received 1 April 1994; accepted for publication 21 March 1995)
Calling blue and fin whales have been tracked using relative travel times and amplitudes from both
direct and multipath arrivals to a seafloor array of seismometers. Calls of three fin whales swimming
in the same general direction, but several kilometers apart, are believed to represent communication
between the whales because of signature differences in call character, an alternating call pattern, and
coordination of call and respiration times. Whale call tracks, call patterns, call character, and
swimming speeds were examined during periods with and without the presence of noise. Noise
sources included airguns, when the whales were subject to sound levels of up to 143 dB P-P
(peak-to-peak) re: 1 pPa over the 10 to 60-Hz band, and transits of merchant ships, when the whales
received continuous levels up to 106 dB rms re: I /•Pa over the 10 to 60-Hz band (115 dB P-P).
Whale responses associated with these noises remain arguable. ¸ 1995 Acoustical Society of
America.
PACS numbers: 43.80.Jz, 43.80.Ka, 43.80.Nd
INTRODUCTION
Many baleen whales produce loud low-frequency under-
water sounds a significant percentage of the time, providing
a practical tool with which to study whale distribution and
movements (Watkins and Wartzok, 1985; Nishimura and
Conlon, 1994; Clark, 1994). Studies of whales have used
arrays of acoustic sensors to determine swimming speed and
direction as well as to monitor call interaction between
whales (Patterson and Hamilton, 1964; Cummings etal.,
1968; Watkins and Schevill, 1977; Cummings and Holliday,
1985). Temporal patterns in acoustic call sequences provide a
measure of respiration times (Cummings et al., 1986) while
call characteristics may separate stocks or groups within the
species (Winn et al., 1981; Ford and Fisher, 1983). Further
study of whale calls may allow them to be used to monitor
behavioral changes associated with man-made noise sources.
The data used for this study were recorded with a seaf-
loor seismometer array having an aperture of about 10 km,
allowing measurement of directionality, apparent acoustic
velocity, relative amplitude and absolute amplitude of sig-
nals. Whale calls were detected during approximately 10%
of the eleven day recording period (a seismology experiment
provided these whale recordings incidentally). Whale calls
were most easily identified when observed in repetitive se-
quences, typically lasting for hours. With the 128-Hz sam-
pling rate used in our seafloor recording system, only low-
frequency sounds, such as those produced by blue and fin
whales, were recorded. Whale calls were detectable at ranges
up to 30 km during this study, although only calls within
about 15 km of the array were analyzed because of the
higher signal-to-noise ratio.
I. METHODS
A. Recording instruments
The study site is about 500 km offshore from Astoria,
Oregon (Fig. 1), in 2400 m of water on the southern Juan de
Fuca Ridge, about 60 km north of the Blanco Fault Zone.
During August of 1990, eight seafloor seismometer recording
packages were deployed with 4 to 6 km between adjacent
instruments. Data were recorded internally on optical disks
and examined after instrument retrieval. There were two in-
strument deployments during the study, each for about 5.5
days. The location of the instruments on the seafloor was
known to within a few tens of meters and instrument clock
drifts were known to within about 10 ms, as discussed in
McDonald et al. (1994). Imprecision in the instrument tim-
ing and navigation is negligible relative to the errors in call
locations resulting from picking errors in the arrival times of
the whale calls. The call arrival time picking errors limit call
position accuracy to several hundred meters theoretically and
to some lesser accuracy, on the order of 1 km, as practiced in
these analyses. The number of instruments used in plotting
the call tracks varied, but was never less than four.
The primary sensor used for recording the signals under
discussion was the vertical component of a 3-axis seismom-
eter (Mark Products L-4, 1-Hz natural period). The response
for this sensor is nearly flat to particle velocity from 1 to
over 100 Hz. The electronics system response was low pass
filtered at 60 Hz because of the 128-Hz sampling rate used.
Background ambient ocean noise levels are higher at low
frequency, so the system gain was 20 dB lower at 1 Hz
relative to 10 Hz. The spectrograms and amplitude plots
shown in this paper are not corrected for system roll-off be-
low 10 Hz and above 60 Hz. Signal amplitudes were con-
verted to dB re: 1 /xPa using correlations between the seis-
mic sensor package and calibrated hydrophones (Benthos
AQ-1), as performed during previous seafloor experiments in
similar water depths where the hydrophones were attached to
the seafloor recording package. No hydrophones were de-
ployed during this study. Experiments using both vertical
seismometers and hydrophones have shown a higher signal-
to-noise ratio for whale calls on seismometers than on hy-
drophones. We suggest this is partially due to the direction-
712 J. Acoust. Soc. Am. 98 (2), Pt. 1, August 1995 0001-4966/95/98(2)/712/10/$6.00 ¸ 1995 Acoustical Society of America 712
130øW 125øW 48 ø
N
46 ø
N
44 ø
N
42 ø
FIG. 1. The observation site was located on the Juan de Fuca Ridge about
500 km offshore Oregon in 2400 m of water.
ality of the vertical component seismometer and the high
incidence angles associated with the whale signallg.
B. Signal localization methods
Different methods are used to locate short transient
sound sources (fin whale calls), long transient somces (blue
whale calls) and earthquakes. The short (one second) tran-
sient sounds were most easily located, because multipath ar-
rivals are separated in time from the direct arriva;[ and from
each other. Figure 2(a) shows an example of the received
signal from a one second transient sound (fin whale call) on
18
16
'" 14
z
<•
rr
12
10
6 8 10 12 14 16
TIME (seconds)
SOURCE
RE3EIVER
FIG. 2. (a) The amplitude display of a 1-s duration fin whale pulse received
on each of four seafloor recorders at different ranges. The ear:test arrivals
are the direct water path and the later arrivals are bounce paths between the
seafloor and sea surface. (b) A cross-section diagram illustrating the direct
and bounce paths seen in (a).
150 [•. • .... 3rd bounce
r•,•,,.•{•.• . . _. A 2nd b .....
145 • • + lstbounce
115
110
105 ambient with moderate %o'•
shipping
10 20 30 40
RANGE (km) 50
FIG. 3. The amplitude of an airgun source versus range for the direct water
path and bounce paths as received on a seafloor recorder during this study.
four array elements. Figure 2(b) shows the corresponding
raypaths for each arrival. Since most whale calls occurred
outside of the array, the relative travel time information al-
lows computation of the azimuth to the source. Relative am-
plitudes of the several multipath arrivals allow range estima-
tion from each instrument. The amplitude of the direct arrival
in Fig. 2(a) decreases steadily with range, while the first
bounce path amplitude reaches a maximum at about 12 km
and the second bounce path provides a relatively higher am-
plitude signal at longer ranges.
Direct and multipath amplitude ratios are plotted versus
range from airgun pulses. during this study in Fig. 3. The
airgun pulses are sufficiently similar to the whale calls under
discussion to expect similar reflection coefficients and simi-
lar multipath amplitude 'variations as a function of range.
Most of the energy in these airgun pulses is in a frequency
band between 10 and 35 Hz and has a duration of several
hundred milliseconds. The observed amplitude variation as a
function of range can be explained in terms of the downward
refraction of the direct path due to the water sound-speed
profile and by the change in seafloor reflection coefficient as
a function of incidence angle. We compared these airgun data
to whale data to estimate the range of the whale from each
instrument. Range information has been combined with rela-
tive arrival time information to determine best estimates for
call locations. Each location is overdetermined by using
range data from several instruments, resulting in better loca-
tion estimates. Because the reflection coefficients are site de-
pendent and the sound-speed profile is both site dependent
and time variant it was important that the airgun amplitude
data were gathered at the same site and at nearly the same
time as the fin whale calls.
Transient signals having durations greater than several
seconds (blue whale calls} have overlapping multipath arriv-
als making it more difficult to use multipath relative ampli-
tudes for range estimation. When the source is within 10 km
of the array or inside the array, relative arrival times locate
the source adequately. At ranges beyond 10 km from the
array, the bearing detemfination is quite accurate but the
range determination, using only arrival time data, becomes
poor. By calculating the average source level of well-located
calls of a given type, the amplitude versus range relationship
713 J. Acoust. Soc. Am., Vol. 98, No. 2, Pt. 1, August 1995 McDonald et el.: Whales observed on a seafloor array 713
derived from the airgun data was used to estimate range.
Earthquake ranges are readily determined using relative ar-
rival times of the compressional and shear wave energy trav-
eling in the rock (e.g., Mailick and Fraser, 1990) while their
bearing is determined by relative arrival times on multiple
instruments.
II. RESULTS
A. Recorded sounds
The observed signals can be described in three catego-
ries; seismic, biologic, and man-made. The most common
signals are those in the seismic category, including those
from seafloor hydrothermal flow (Sohn et al., 1992), earth-
quake T-phases (Walker and Bernard, 1993) and earthquake
body waves. In the biologic category are whale calls and the
so called "fishbumps" (Buskirk etal., 1981), which can
sometimes be identified as biologic because they occur more
fi'equently at certain times of day and certain water depths.
Man-made signals observed were primarily those associated
with ships. Less common signals in this type of data include
tidal current flow noise (Duennebier etal., 1981; Ambos
et al., 1985) and volcanic tremor (Talandier and Okal, 1984;
Talandier and Okal, 1987).
We analyzed two sequences of repetitive, two part, ap-
proximately 20 Hz, whale calls from the eleven days of re-
cordings. These two part calls consist of: A 19-s signal (type
A) followed by a 24-s interval, a different 19-s signal (type
B), followed by a 60-s interval before the first signal (type A)
of the next pair occurs. These signals lasted for 10.5 h (Au-
gust 18, 0030 to 1100 Local) and 5.2 h (August 19, 0640 to
1150 Local), consisting of about 375 and 180 calls, respec-
tively. The 10.5-h sequence was in progress when data re-
cording began and continued until the whale was beyond the
detection range of the array. There were also other faint 19-s
signals which were not analyzed because of low signal-to-
noise ratio, but which have the same characteristics.
There were two sequences of 1-s duration whale calls
near 20 Hz, repeating at an average interval of 19 s, except
for pauses which typically last 150 s. These sequences lasted
2.2 h (August 21,0410 to 0620 Local) and 8.0 h (August 28,
2010 Local to August 29, 0410 Local), consisting of about
1000 and 4000 calls, respectively. Only the 8.0-h sequence
was analyzed because of the poor signal4o-noise ratio on the
2.2-h sequence. We also recorded signals from 65 regional
and local earthquakes, several of which occurred during a
whale call sequence.
III. CALLS ASSOCIATED WITH BLUE WHALES (B.
musculus)
A. Spectra
The 19-s duration signals are recognized as blue whales
because of a similarity to signals recorded in the presence of
blue whales (Cummings and Thompson, 1971; Edds, 1982;
Thompson et al., 1987; Ailing et al., 1991; Alling and Payne,
unpublished manuscript), and because of considerable un-
published navy research on these signals (Cummings and
Thompson, 1994). Similar recordings are reported that lack
60 t
)_ 40•
•- •o 1
0 10 20 30 40 50 80 70
TIME (seconds)
FIG. 4. The spectrogram and corresponding lime series record of a typical
blue whale call pair. The .speclrogram was made using a filler bandwidlh of
1.55 Hz and a Iime window of four seconds. The average duration and
slandard deviation for each portiou of the call pair sequence is shown from
one sequence of 132 call pairs. The call is divided inlo lwo portions, paris A
and B.
visual corroboration (Weston and Black, 1965; Kibblewhite
et al., 1967; Northrup et al., 1968; Northrup et al., 1971;
Thompson, 1965; Thompson etal., 19791 Thompson and
Friedl, 1982) and we understand that other recordings with
visual corroboration remain unpublished as this correlation is
well accepted in the whale acoustics field.
A typical two part blue whale call waveform and corre-
sponding spectrogram tire shown in Fig. 4. The first 19-s call
segment (part A) shows a series of six spectral lines about
1.5 Hz apart with the lowest at 17.5 Hz. Spectral lines like
this can be generated by a pulsed tone and have been referred
to as pulsive (Watkins, 1964). A pulselike amplitude modu-
lation is evident in the waveform display. Spectrograms com-
puted for more than 100 call pairs l'mm the August 18 se-
quence reveal remarkably little variability in character.
The second segment (part B) in the blue whale call pair
is probably not amplitude modulated. The amplitude vari-
ability seen in the records appears to be the result of con-
structive and destructive interference of sevmal reflected
paths and the direct travel path as evidenced by differences
in the tilne-versus-amplitude character of the same signal
observed at different ranges. The fundamental tone in this
call begins at 19 Hz and sweeps down in frequency to 18 Hz
in the first 3 to 4 s (Fig. 4). The 18-Hz tone is then carried
until the last 5 s where the dominant tone sweeps down to 17
Hz. Since the sampling rate was 128 Hz there may be seg-
ments or components of the call sequence above 60 Hz
which were not recorded. Higher frequency segments and
components have been reported by Cummings and Thomp-
son (1971), Thompson et al. (1979), and Ailing et al. (1991),
although the whales recorded by Ailing et al., may be pygmy
blue whales (B. musculus bt, vicauda).
B. Temporal pattern in calls
Only the August 18 blue whale call series was suitable
for analysis in terms of the gaps between calls, since airgun
shot noise made it difficult to pick the beginning and end of
714 J. Acoust. Soc. Am., VoL 98, No. 2, Pt. 1, August 1995 McDonald et aL: Whales observed on a seafloor array 714
24O
o
• •80
.J
_J
o
LU 120
J
< 6o
z
I I I I
60 120 180
TIME (minutes)
FIG. 5. The time interval between successive blue whale call pairs is typi-
cally 60 s but after about five calls there is a longer 060 s) interval, iuter-
preted to be a breathing pause.
each call from the August 19 series. Spectrograms were com-
puted and call start and stop times were picked for 134 con-
secutive pairs having a high signal-to-noise ratio. All but one
of the 134 call pairs followed a stereotype pattern consisting
of a 19-s pulsive signal followed by a 24.5-s gap and a 19-s
monotonic signal. One unusual call was missing part B and
the next call pair began 45 s later.
A plot of interval time between call pairs (Fig. 5) shows
a longer time after every five or six calls, which may repre-
sent a respiration pattern. This would correspond to a typical
dive time of 660 s and a breathing time of 160 s. Similar
patterns in call intervals were reported by Thompson et aL
(1987) where call pairs occurred in series of two to six with
a median repetition interval (breathing time) of 131 s. Edds
(1982), however, reports exceptions to the pattern of calling
during dives and not calling during breathing and reports
surface slicks during the call sequences demonstrating the
whale to be at shallow depth during some calls.
C. Call tracking
Call locations were calculated [Fig. 6(a)] at regular time
intervals for a portion of the August 18 series by fitting
smoothed curves through the amplitude and relative arrival
time data from four seafloor instruments. Changes in both
bearing and received call amplitude [Fig. 6(b)] occur
smo6thly, suggesting relatively continuous movement of the
whale during the call series. Each track could have been
generated by either a single whale or a group of whales trav-
eling together if separations were less than one kin. This
series contains no overlapping calls which would be indica-
tive of a second whale at the same location. Other recordings
(Thompson, 1965, in Urick, 1983, p. 219) have shown over-
lapping calls.
The speed of the whale(s) was 6 km/h averaged over the
first track and. 10 km/h averaged over the second, while the
715 J. Acoust. Soc. Am., Vol. 98, No. 2, Pt. 1, August 1995
a
• 6o
o
z 55
•' 50
i-
ra 45
o
ß Position 0200 0030
0600
ob Call Amplitude at Element 6
120 ., •6• ["
e u
110 • P I t
0 2 4 6
tocol time (hours)
FIG. 6. (a) The track of a blue whale with local time shown adjacent to each
of the points for which position was calculated. (b) The blue whale call
amplitude changes with time. Each point plotled represents one blue whale
call pair, so the gaps iudicate breaks in the call sequence.
•-h interval ranged from 4.5 to 13 km/h.
average speed per •
All speeds were computed between the points marked with
times on the track plot of Fig. 6(a). The whale(s) may be
swimming an irregular course between the computed points
implying a higher speed. Reported observed speeds for blue
whales are 2 to 6.5 km/h while feeding, 5 to 33 km/h while
cruising or migrating and a maximum speed of 20 to 48 km/h
when being chased or harassed (Yochem and Leatherwood,
1985).
IV. CALLS ASSOCIATED WITH FIN WHALES (B.
physalus)
A. Correlation of 1-s signal with fin whales
The l-s, frequency-downswept, pulses recorded during
this study (Fig. 7) are known to be typical fin whale calls
181
164 FIRST
REFLI !CTION
DIRECT
PATH
SECOND
REFLECTION THIRD
'1 • REFLECTION
TIME (seconds)
FIG. 7. Spectrogram and time series for a typical fin whale call, the first
sound at 1.3 s is the direct water arrival and subsequent sounds are the
multipath arrivals of the same signal. The speclmgram was made using a
filter bandwidth of 6.19 H/and a time window of 2 s.
McDonald et al.: Whales observed on a seafloor array 715
(Schevill et al., 1964; Thompson et al., 1979; Watkins et al.,
1987; Richardson et al., 1991; Thompson et al., 1992). Iden-
tification of acoustic signals from the fin whale is primarily
from combined visual observations and recordings off the
east coast of the U.S. where there is a seasonal pattern of
signals common off Bermuda only in winter and common in
the St. Lawrence estuary and off the New England coast only
in summer (Edds, 1988; Watkins, 1981).
B. Temporal pattern in calls
We describe call series durations and rest intervals, as-
sociated with the respiration cycle of fin whales (Cummings
et al., 1986), using the terms: Pulse interval, rest, and gap, as
defined by Watkins et al. (1987). The average pulse interval
of the fin whales, excluding rests, is 19 s averaged over 467
calls from one continuous series, having no gaps (exception-
ally long rests), during the 28-29 August series. This 19-s
pulse interval is similar to prior observations (Watkins et al.,
1987). The rest durations, corresponding to breathing times
and the pulse series durations corresponding to dive times,
suggest typical dive times of 600 s and breathing times of
150 s, from this study. Pulse interval times of 900 and 720 s
with corresponding breathing times of 150 and 120 s have
been reported respectively by Patterson and Hamilton (1964)
and by Watkins et al. (1987). These acoustically interpreted
durations are longer than the 201 and 90 s average dive times
and the 90 and 55 s average breathing times from visual
observations, reported by Stone et al. (1992) and by Edds
and MacFarlane (1987), although, other factors such as the
activity of the whale and the water depth, make these com-
parisons difficult to interpret.
C. Call tracks and whale interaction
The 28-29 August series of fin whale calls indicates
repeat interactions among three whales located several kilo-
meters apart. Figure 8 shows the time series and correspond-
ing spectrogram from several minutes of a typical fin whale
call series with mt/itiple whales. At this compressed time
scale it is not possible to see the downsweep associated with
each pulse, but the frequencies and bandwidth show the dis-
tinctire signature a•socmted with each of three whales.
Whale "a" in Fig. 8 has a signature consisting of a down-
sweep from 18 to 14 Hz, while whale "b" has a 25- to 16-Hz
downsweep signature, and whale "c" a 37- to 22-Hz down-
sweep signatur e . These signatures were very consistent
throughout th'e sequence of 467 calls which were examined
in detail. That the whales are interacting, rather than just
independently calling, is suggested by the consistent alter-
nate spacing between calls with the calls never overlapping,
the distinctive call signature of each whale and the apparent
synchronization of respiration. The time series records from
the four seafloor instruments demonstrate the separation be-
tween these whales, with different instruments showing the
highest amplitude for the whale nearest that instrument.
For the sequence of 467 calls analyzed in detail, the
whale labeled "a" in Fig. 8 produced 46% of the total calls,
"b" produced 41% and "c" produced 13%. Within these 467
calls there are only six occasions when a whale called twice
40-
3a--
a
60 120 180
TIME (seconds)
FIG. 8. Time series for elements S1, S4, S5, and S7, and spectrogram for
element S1 from several minutes of a typical fin whale call series with
multiple whales. The frequency and bandwidth differences make up the
distinctive signatures associated with three whales labeled "a," "b," and
"c." That the whales are interacting rather than just independently calling is
suggested by the consistent alternate spacing between calls with nonover-
lapping calls and a distinctive call signature for each whale. The time series
records from the four seafloor instruments demonstrate the separation be-
tween the three whales with different instruments showin• the highest am-
plitude for the nearest whale.
without an intervening countercall, including the rest gaps in
this sequence. Of these occasions, whale "a" called consecu-
tively only once, that after a particularly long rest gap, such
that the two calls were some 5 min apart. Whale "b" twice
called consecutively and whale "c" on three occasions. On
the occasions when whale "c" repeated its call without a
response, it called two, three, and five times consecutively.
Cursory examination of other similar data sets from the same
area suggest this to be a common patten], where calls of the
type produced by whale "c" are fewer in total number but
are most often repeated consecutively.
Figure 9 shows the tracks of the three whales over the
1.5 h while they were nearest the array. The whales traveled
side-by-side rather than single-file and no whale appeared to
be leading. The whales tracked in Fig. 9 were swimming at
speeds of 5 to 14 km/h as averaged over the «-h intervals
shown. Swimming speeds for fin whales from visual data are
reported as 10 to 16 km/h for long periods when in transit
and 20 km/h or more for shorter periods (Watkins, 1981).
The lowest frequency whale, "a," swam most quickly, called
most often and never called twice in a row; while the highest
frequency whale, "c," swam slowest, called least often and
called consecutively most often.
716 J. Acoust. Soc. Am., Vol. 98, No. 2, Pt. 1, August 1995 McDonald et al.: Whales observed on a seafloor array 716
40
z
•- 3o
•5
20 30
o 4
35 40 45 50
KILOMETERS EAST
• Arroy Element s• Whole 'b" Pos;tion
FIG. 9. Tracks of the three fin whales ("a," "b." and "c") from Fig. 8 travel
together, showing their positions at equivalent times connected by dashed
lines. The whales appear to travel side by side rather than in a single file.
Whale "a" with the lowest frequency signature. travels the fas, est (circles),
whereas whale "c" with the highest frequency signature, travel4 the slowest
(triangles).
V. WHALE RESPONSE TO MAN-MADE AND SEISMIC
NOISE IN THE OCEAN
A. Significance
Concern for the welfare of marine mammals has focused
attention on the effects man-made noise may have on the
behavior, communication, or general welfare of wh;des. Un-
der the Marine Mammal Protection Act, increasingly strict
guidelines for the use of man-made noise in the oceans are
being applied as a precaution against disruption of whale
behavior (Holman, 1989; Green et al., 1994). Natural tran-
sient noises in the ocean are primarily associated with earth-
quakes and volcanic eruptions while man-made transients
include shipping, geophysical surveying with airguns, under-
water explosions and hydraulic sound sources such as are
used with acoustical oceanography experiments.
B. Airgun noise
We conducted a seismic refraction survey w th a four
airgun array having a total capacity of 1600 cubic in., fired at
1800 psi. These airguns are individually larger than those
typically used in oil exploration, resulting in lower maximum
sound levels at lower frequency, near 15 Hz. The airgun
array produced about 215 dB P-P re: I /.tPa at I m over a 10-
to 60-Hz band (a suboptimal sound-pressure level I:ecause of
depth and towing speed constraints) as estimated from seaf-
loor measurements. The array was nearly symmeU'iz in azi-
muth. The directionality of the airgun away was not mea-
sured, but this source level measurement is appropriate for
estimating received level at the whales because the seafloor
reflected paths will be the loudest received at the whales.
Figure 10 shows a blue whale track during airgun operations.
The dashed lines connect the airguns location with the whale
location for the matching time. This whale, in Fig. 10, was
z 50
• 40 -o ?:.._.•
-• 30 --- ;:.o
40 60 80
K LOMEIEES
b
• 124[Coll Amplitude ol Element 8
116•
• 6 8 10 12
Locol time •hours)
FIG. 10. (a) A blue wh•e tr•k during airgun operations. The dashed lines
connect the location of the airgan ship with the whale location at the core-
spuncling t, me. The whale sta•ed its call sequence well within the trachng
range of the •ay when the •rgun ship was 15 km distant (07•) h local
time). The wh•e lollowed a putreit track until it stopped calling at a r•ge
of 10 •. After a gap in the call •equence, the whale track moved diagongly
away from the ship. (b) The r•eived whale call amplitude changed with
time. Each point plotted •prescnts one blue whale call pair. so the gaps
in,cate pauses in the call sequcace.
moving slightly faster than the whale tracked in Fig. 6(a),
with an average speed of 10 km/h, varying from 7 to 13
km/h. The whale started its call sequence well within the
tracking range of the array when the airgun ship was 15 km
distant (0700 h local time). The whale closed on the ship
following a pursuit track until it stopped calling at a range of
10 km. At this point, the ship was moving about 10 km/h and
was beginning to increase its distance from the whale; the
sound level of the airguns then was 143 dB P-P re: 1 /xPa
over a 10- to 60-Hz band at the whale. After a gap in the call
sequence, a new call series, presumably by the same whale,
was again located 10 km from the ship, suggesting it had
taken a track generally paralleling the ship. The series of
positions after 0930 shows the whale moving diagonally
away from the ship. Comparing tt•is track with that ol' Fig.
6(a), it appears the whale may have been approaching the
ship intentionally, or perhaps was unaffected by the airgun
ship. More data of this type will be needed to draw conclu-
sions about the affect of such ndise on blue whale behavior.
Studies of bowhead and gray whale behavior in the presence
of airgun noise indicate avoidsthee at broadband levels of
about 160 to 170 dB 0-P re: I /.•Pa (Malme etal., 1984:
Richardson etaL, 1986; Richardson etal.. 1991; Tyack,
1993).
C. Ship-generated noise
Noise created by the research vessel did not significantly
increase the background noise level between 10 and 60 Hz at
the study site (except during airgun operations), but regular
passages of larger merchant ships were observed to increase
10- to 60-Hz noise levels by as much as a factor of 20 in
amplitude (26 dB) relative to times when no merchant ships
717 J. Acoust. Soc. Am., Vol. 98, No. 2, Pt. 1, August 1 [,95 McDonald et aL: Whales observed on a seafloor array 717
ß Max•. Amp. 64 Events
UI T-Phase 54 Events
105 115 125 135 145 155 165
P-P Pressure (dB re 1 gtPa)
FIG. 11. The peak amplitudes of body-phase or t-phase pressure signals,
whichever is greater, and the t-phase pressure signals from the 64 earth-
quakes observed in this study. At low amplitudes the smaller number of
detected events is attributed to detection limitations related to the variability
of background noise from local shipping.
were nearby. The blue whale track ill•ustrated in Fig. 6 could
not be located after 0800 h because an approaching merchant
vessel increased the noise level so that it was difficult to pick
the arrival times and amplitudes of the blue:wh•le calls. This
merchant vessel came from the north-northwest, passing
about 5 km west of the nearest array element •t 0940 h when
the whale was approximately 5 km west of the vessel. Sound
levels at the whale were about 106 dB rms re: 1/•Pa over the
10- to 60-Hz band (115 dB P-P), yet the blue whale contin-
ued to call as before. In contrast, avoidance behaviors have
been observed in beluga whales at ship noise levels of only
94 to 105 dB rms re: 1,/•Pa in the 20- to 1000-Hz band from
ships 35 to 50 km distant (Finley et el., 1990). Bowhead
whale avoidance behaviors have been observed in half the
animals when exposed to 115 dB rms re: 1 /•Pa broadband
drillship noises (Richardson eta!., 1990), but behavioral re-
actions are considered to vary depending on the characteris-
tics of the noise, whale activity and the physical situation
(Richardson and Greene, 1993).
D. Earthquake-generated noise
Earthquakes are believed to have generated ocean noise
at similar frequencies and magnitudes throug,hout the evolu-
tion of whales, and thus are a background noise to which the
whales are presumably adapted. The frequency of occurrence
of earthquakes producing transient sounds of given ampli-
tude in the ocean can be predicted using observations such as
were collected in this study together with the empirical rela-
tionship by which the frequency of earthquake occurrence
rises by a factor of l0 (a close approximation) for each de-
crease in eaxthquake magnitude (Frohlich and Davis, 1993).
The histogram shown in Fig. I 1 provides a reference level
upon which the empirical relationship is applied for this re-
glen of the north Pacific. Similar north Pacific data has been
reported by Johnson and Jones (1978), Jones and Johnson
(1978), Hyndman and Rogers (1981), and Fox et al. (1994)
and there are at least 20 similar published data sets for other
areas of the Pacific.
The data of Fig. 11 have been divided into two types,
tertiary-phase (T-phase) energy which is transmitted in the
water and b9dy-phase energy transmitted in the rock, enter-
ing the wate• near the receiver. The earthquakes that gener-
ated the signals recorded during this stud• occurred at ranges
as far as 170 km and as near as 1.1 km, with most at ranges
near 65 kin, the range from the study sii• to the Blanco
Transform Fault Zone. Body-phase energy provides the high-
est amplitude sound in the water at near ranges, while at
longer ranges the T-phases dominate.showing energy above
background levels from 3 to 35 Hz (Walker et al., 1992)
regardless of the earthquake range. Different regions in the
ocean will be dominated by one or the other, depending on
local seismicity levels.
Using this information,. it is possible to estimate the
earthquake sound level a whale will hear with a given fre-
quency of occurrence. If we assume whale hearing has
evolved to withstand the maximum level of earthquake gen-
erated sound, occurring once per lifetime, this level suggests
how loud a man-made sound might be w.i. thout causing per-
manent damage. For purposes of the calculation, we assume
our study site receives earthquake noise typical of that re-
ceived by the hypothetical whal$. We believe the earthquake
noise at our study site to be above average for the worlds
oceans, but not extremely so. We also assume for the calcu-
lation that the pressure amplitude produced in the ocean is
directly related, one to one on a 1ogarithn•ic scale, with
earthquake magnitude increase. Extrapolating from our
sound level of 140 dB P-P re: I /zPa in the 10- to 60-Hz
band, from earthquakes once per day (Fig. 11), we calculate
a sound level exposure of 204 dB P-P re: •/•Pa in the 10- to
60-Hz band for earthquakes occurring once per lifetime (50
years). In the less likely occurrence where the whale hap-
pened to be directly over the earthquake, the sound exposure
level would be 231 dB P-P re: 1 /xPa. We anticipate that
SOSUS data (Gagnon et ai., 1993; Nishimura and ConIon,
1994; Fox vt al., 1994) will soon provide direct answers to
the question of maximum earthquake sound pressure levels
detected over a period of years, avoiding the assumptions
implicit iri our calculation. Very distant earthquakes, such as
the very large and deep Bolivian event in 1994, have no
bearing on this calculation, as that event produced no signifi-
cant energy above 2 Hz in the region of this study. Possible
effects of an intermediate range earthquake on gray whales is
discussed by Maline et ai., 1989.
Figure 12 shows earthquake sound energies, both
T-phase and body-phase during a fin whale call series. There
is no disruption of the call interaction in this case when the
sound level from the earthquake is 121 dB P-P re: 1 /•Pa
over the 10- to 60-Hz band at the whale. When extrapolated
back to the Blanco Transform Fault Zone, sound levels from
this earthquake would be 27 dB higher (148 dB P-P re: 1
/xPa) within a few kilometers of the epicenter. The range
between the earthquake and the whales is practically the
same as the ear. t. hquake to recording instrument range at the
time scale of this plot.
718 J. Acoust. Sec. Am., VoL 98, No. 2, Pt. 1, August 1995 McDonalc• et al.: Whales observed on a seafloor array 718
o 60 TIME (seconds)
FIG. 12. Earthquake noise, including both t-phase and body phases during a
fin whale call sequence. There appears to be no disruption of the call inter-
action in this case when the noise level from the earthquake is 121 dB
re: I /sPa over the 10- to 60-Hz band, at the whale.
Vl. DISCUSSION
A. Noise pollution of the deep ocean
Noise pollution, herein defined as anthropogenic noise,
in the deep sound channel is basically of two types: Transient
noises and background noise levels. Natural transients are
primarily associated with earthquakes and volcanic e•up-
tions, while manmade transients include geophysical survey-
ing with airguns, underwater explosions and hydraulic
sources such as those used for acoustical oceanography ex-
periments. First we discuss the possible anthropogenic ef-
fects on background noise levels in the deep sound channel.
It is impmtant to keep in mind during this discussion that the
sound channel has a profound effect: 20-Hz ambient noise
levels in the sound channel are about 30 times louder in
amplitude (30 dB) than noise levels at typical seafloor
depths, such that while local wind noise (associated with
breaking waves) may set the background noise level at the
seafloor (McCreery et aL, 1993; Shooter et al., 1990), the
background level in the deep sound channel is associated
with distant sources, including wind noise, shipping, seis-
micity and distant whale calls.
The analysis of Payne and Webb (1971) proposes that
propeller driven shipping has changed the background noise
levels of the entire worlds' oceans, negatively affecting the
ability of the 20-Hz whale calls to be heard at long range.
More recent work challenges that hypothesis, suggesting that
sources other than shipping may cause the observed back-
ground levels at 20 Hz. Ambient noise levels in the deep
ocean sound channel near the 20-Hz frequency range are
bounded on the low side (below 5 Hz) by nonlinear wave-
wave interaction, a phenomena which is relatively well mea-
sured and understood (Kibblewhite, 1985; Webb, 1992).
Noise levels between 5 and 50 Hz have traditionally been
attributed to shipping noise (Wenz, 1962), but may be alter-
natively and/or regionally sourced from high latitude winds
(Bannister, 1986). Some authors have suggested other noise
sources in this frequency range including atmospheric turbu-
lence (Wilson, 1979; Copeland, 1993), lightning (Du-
brovskiy and Kosterin, 1993), and glacial ice flow (Cope-
land, 1993), but we consider these sources improbable or
insignificant.
Many observations of ambient noise levels have a peak
at 20 Hz which is usually attributed to whale calls and as-
sumed to be caused by whales relatively local to the obser-
vation site even though the whale calls may not be distinc-
tive in the time series data (Kibblewhite etal., 1976;
Copeland, 1993). The important question is whether the
source of background noise near this 20-Hz peak is shipping,
a source the whales did not evolve with and may not be
readily adapting to. Our calculations of earthquake T-phase
noise levels in the sound channel, when extrapolated back to
a frequency of occurrence of nearly continuous noise (the
distinction between background and transient noise becomes
blurred), suggest seismicity may be the dominant noise
source at 20 Hz in some otherwise quiet regions of the ocean.
This is significant because earthquakes would have been
present throughout the period of whale evolution. A discus-
sion of the assumptions used in these calculations would be
too lengthy for this paper and recent SOSUS data may also
provide better data upon which to base this type of calcula-
tion.
A good example of transient noise pollution is that from
the seismic airgun array. Transient noises associated with
geophysical surveying off the California coast have been
readily recorded on land seismometer arrays 6100 km distant
in Polynesia after traveling via the deep sound channel with
little dispersion (Okal and Talandier, 1986). These noises,
however, were probably not even heard by a whale near the
surface in the mid-Pacific, because of the trapping of the
sound in the deep sound channel. This type of noise will be
heard only by the whales which dive below several hundred
meters depth and to those in the polar regions where the
sound channel shallows, assuming some land mass has not
blocked the path. Prediction of airgun noise propagation is
also complicated by the environment near the airguns, which
in this instance may have been well suited for introduction of
the airgun noise into the deep sound channel by downslope
conversion where the sound is reflected off the seafloor, pro-
viding an efficient means of entering the deep sound channel
(Jensen et aL, 1994).
B. Why blue and fin whales call
Speculations on why blue and fin whales call at 20 Hz
have focused primarily on the question of communication
with, or at least the broadcasting of relative location infor-
mation to other whales of like species (Payne and Webb,
1971). The use of 20-Hz signals for depth finding sonar is
somewhat discounted by the observation of calls when the
whales are moving slowly and calling frequently, making the
information content redundant. The whales would be ex-
pected to perceive the depth from the 20-Hz echos, but the
signal is louder than would be necessary and a higher fre-
quency signal would seem a more logical choice for this
purpose. Relative location infmmation is undoubtedly impor-
tant to survival of the species if breeding pairs are to meet.
Recent developments in long distance call tracking of blue
whales using navy hydrophone arrays (Gagnon et al., 1993)
has raised speculation that the whales are horizontally echo-
sounding the island of Bermuda from ranges up to 1000
miles (Clark, 1993).
719 d. Acoust. Soc. Am., Vol. 98, No. 2, Pt. 1, August 1995 McDonald et aL: Whales observed on a seafloor array 719
The information contained in the multipath amplitude
ratios, as used to track fin whales in this study, suggests that
fin whales could measure oceanic sound speed profiles by
countercalling among the pod. The direct path amplitude
relative to the bounce path amplitudes would provide a mea-
sure of the magnitude of downward refraction associated
with the shallow (upper 500 m) sound-speed profile. Suffi-
cient information could be obtained by the whale receiving
the signal if the source were at least several km distant from
the calling whale. Countercalls could provide measurements
of variability in the sound-speed profile which would not be
possible simply by listening to the echo of their own call,
because a horizontal travel path amplitude would be added
for comparison with reflected path amplitudes. Extraction of
water sound-speed and corresponding temperature profiles
may be possible using the principles of matched field pro-
cessing (Jensen etal., 1994). Changes in the depth and
sharpness of the thermocline may be estimated from these
amplitude ratios and may help the whale locate food.
VII. CONCLUSIONS
The use of passive seafloor arrays to track and monitor
the calls of passing whales has advantages over other meth-
ods, such as radio tracking or visual observations, not only in
cost but also because seafloor recording arrays are unobtru-
sive, unlike methods, using ships or aircraft as observing
platforms, which may interfere with whale behavior. The re-
cordings from this study further show the use of signature
calls among fin whales, as suggested by Watkins (1981).
These recordings provide a minimum measure of how many
whales were present in an area and provide information on
the character and pattern of their calls, which may eventually
lead to association with specific behaviors and/or the separa-
tion of population groups by characteristic calls. Tracking of
the calls provides direction and speed of travel information
which may prove complementary to future efforts towards an
acoustic census of pelagic baleen whale populations.
As demonstrated during the eleven days of this study,
acoustic recordings provide a measure of the low-frequency
noise levels that whales are exposed to and observations of
any response to such noise. The observed noise levels at the
whale during this study were 143 dB P-P re: I /xPa over the
10- to 60-Hz band for airgun noise, 106 dB rms re: 1 ttPa
over the 10- to 60-Hz band for ship noise and 121 dB P-P re:
1 /zPa over the 10- to 60-Hz band for earthquake noise.
These observations may help set the minimum level at which
a response might be expected.
ACKNOWLEDGMENTS
This research was supported by the NOAA Vents Pro-
gram. We thank LeRoy Dorman and Chris Fox for helpful
suggestions; Vince Pavlicek, Tom Deaton, and Jacques
Lemire for their assistance building the instruments, Paul
Wade for his encouragement in teaming about whales, John
Richardson, Sue Moore, Cyndy Tynan, and three anonymous
reviewers for their comments on the manuscript, and the of-
ricers and crew of the NOAA ship DISCOVERER.
Ailing, A., Dorsey, E. M.. and Gordon, J. C. D. (1991). "Blue Whales
(Balaenoptera musculus) off the Northeast Coast of Sri Lanka: Distribu-
tion, Feeding and Individual Identification," in United Nations Mar. Mam-
mal. Tech. Rep. 3, Cetaceans and Cetacean Research in the Indian Ocean
Sanctuary, edited by S. Leatherwood and G. P. Donoran (Nairobi, Kenya),
pp. 248-258.
Ambos, E. L., Hussong, D. M., and Holman, C. E. (1985). "An Ocean
Bottom Seismometer Study of Shallow Seismicity Near the Mid-American
Trench Offshore Guatemala," J. Geophys. Res. 90(B13), 11397-11412.
Bannister, R_ W. (1986). "Deep sound channel noise from high-latitude
winds," J. Acoust. Soc. Am. 79, 41-48.
Buskirk, R. E., Frohlich, C., Latham, G. V., Chen, A. T, a•d Lawton, J.
(1981). "Evidence that biological activity affecls ocean bottom seismo-
graph recordings," Mar. Geophys. Res. 5, 189-205.
Clark, C. W. (1993). "Bioacoustics of baleen whales: From infrasonics to
complex songs," J. Acoust. Soc. Am. 94, 1830, Abstract.
Clark, C. W. (1994). "Blue deep voices: Insights from Ihe Navy Whales '93
program," Whalewatcher, J. Am. Cetacean Soc., 6- I I.
Copeland, G. J. M. (1993). "Low frequency ambient noise--generalized
spectra," in Natural Physical Sources of Underw, ater Sound (Kluwer Aca-
demic, Dordrecht), pp. 17-30.
Cummings, W. C., Thompson, P. O., and Cook, R. (1968). "Underwater
sounds of migrating gray whales, Eschrichtius glaucus (Cope)," J. Acoast.
Soc. Am. 44, 1278-1281 (1968).
Cummings, W. C., and Thompson, P.O. (1971). "Underwater sounds from
the blue whale, Balaenoptera musculus," J. Acoust. Soc. Am. 50, ! 193-
1198.
Cummings, W. C.. and Holliday, D. V. (1985). "Passive acoustic location of
bowhead whales in a population census Poim Barrow, Alaska," I. Acoust.
Soc. Am. 78. 1163-1169.
Cummings, W. C., Thompson. P. O., and Ha, S. I. (1986). "Sounds from
bryde, Balaenoptera edeni, and finback, B. physalus, whales in the Gulf of
Califomia," Fish. Bull. 84(2), 359-370.
Cummings, W. C., and Thompson, P.O. (1994}. "Characteristics and sea-
sons of blue and finback whale sounds along the U.S. west coast as re-
corded at SOSUS stations," J. Acoust. Soc. Am. 95, 2853 (A).
Dubrovskiy, N. A., and Kosterin, S. V. {1993). "Noise in the Ocean Caused
by Lightning Strokes," in Natural Physical Sources of Underwater Sourut
(Klmver Academic, Dordrecht), pp. 697-709.
Duennebier, F. K., Blackinton. G, and Sutton, G. H. (1981). "Current-
generated noise recorded on ocean bottom seismometers," Mar. Geophys.
Res. 5, 109-115.
Edds, P. L. (1982). "Vocalizations of the blue whale, Balaenoptera muscu-
lus, in the St. Lawrence River," J. Mammal. 63(2). 345-347.
Edds, P. L., and MacFarlane, J. A. F. (1987). "Occurrence and general
behavior of balanopterid cetaceans summering in the St. Lawrence Estu-
ary, Canada." Can. J. Zool. 65, 1363-1376 (1987).
Edds, P. L. (1988). "Characteristics of finback Balaenoptera physalus, vo-
calizations in the St. Lawrence River," Bioacoustics 1, 131-149.
Finley, K. 1., Miller, G. W., Davis, R. A., and Green, C. R. (1990}. "Reac-
tions of Belugas, Delphinapterus leucas and narwhals, Monodon monoc-
eras, to ice-breaking ships in the Canadian High Arctic," Can. Bull. Fish.
Aqua. Sci. 224, 97-117.
Ford, J. K. B., and Fisher, H. D. (1983). "Group specific dialects of killer
whales. (Orcinus orca) in British Columbia." in Communication and Be-
havior of Whales, edited by R. Payne (Westview, Boulder), pp. 333-358.
Fox. C. G., Dziak, R. P., Matsumoto, H., and Schreiner, A. E. (1994). "Po-
tential for monitoring low-level seismicity on the Juan de Fuca Ridge
using military hydrophone arrays," Mar. Teehnol. Soc. J. 27(4), 22-30.
Frohlich, C., and Davis, S. D. (1993). "Teleseismic b Values; Or, Much Ado
About r,0," J. Geophya. Res. 98(B1), 631-644.
Gagnon G., Nishimura, C., Fleming, H., Clark, C., ConIon. D., Liechty, J.,
and Olson, L. (1993). "Results of the Whales '93 Program Utilizing SO-
SUS Array Data for Monitoring Cetaceans and T-Phases," EOS 74, 342,
Abstract.
Green, D. M., DeFerrari, H. A., McFadden, D., Pearse, J. S., Popper, A. N.,
Richardson, W. J., Ridgeway, S. H., and Tyack, P. L. (1994). "Low-
frequency Sound and Marine Mammals: Current Knowledge and Research
Needs," National Research Council Report, 96 pp., National Academy
Press, Washington, DC.
Holman, R. J. (1989). "The Marine Mammal Protection Act, A First of Its
Kind Anywhere," Oceanus 32(1), 21-28.
Hyndman, R. D., and Rogers, G. C., (1981). "Seismicity Surveys With
720 J. Acoust. Soc. Am., Vol. 98, No. 2, Pt. 1, August 1995 McDonald et aL: Whales observed on a seafloor array 720
Ocean Bottom Seismographs Off Western Canada," J. Geophys. Res.
86(B5), 3867-3880.
Jensen, E B., Kuperman, W. A., Porter, M. B., and Schmidt. H. (1994).
Computational Ocean Acoustics (AIP, Woodbury, NY).
Iohnson, S. H., and Jones, P. R. (1978). "Microearthquakes La,:ated on the
Blanco Fracture Zone With Sohobuoy Arrays," J. Geophys. Res. 83(BI),
255-261.
Jones, P. R., and Johnson, S. H. 0978). "Sonobuoy Array Measurements of
Active Faulting on the Gorda Ridge," J. Geophys. Res. 83(B7), 3435-
3440.
Kibblewhite, A. C., Denham, R. N., and Barnes, D. J. (19671. "Unusual
Low-Frequency Signals Observed in New Zealand Waters,' J. Acoust.
Soc. Am. 41, 644-655.
Kibblewhite, A. C., Shooter, J. A., and Watkins, S. L. {1976). "I;xamination
of attenuation at very low frequencies using the deep-water ambient noise
field," J. Acoust. Soc. Am. 60, 1040-1047.
Kibblewhite, A. C., and Ewans, K. C. (1985). "Wave-wave interactions,
microseisms, and infrasonic ambient noise in the ocean," J. Acoust. Soc.
Am. 78, 981-994.
Mallick, S., and Frazer, L. N. (1990). "Po/So synthetics for ,'• variety of
oceanic models and their implications for the structure ot :he oceanic
lithosphere," Geophys. J. lat. 100, 235-253.
Malme, C. L, Miles, P. R., Clark, C. W., Tyack, P., and Bird, J E. (1984).
Investigations of the potential effects of underwater noise front petroleum
industry activities on migrating gray whale behavior, Phase 11: January
1984 migration, prepared by Bolt Beranek and Newman Inc, vat. pag.,
OCS study MMS 84-025, 1J.S. Department of the Interior, Minerals Man-
agement Service, Alaska OCS Office, Anchorage, AK, N'tlS PB 86-
218377.
Malme, C. I., Miles, P R., Miller, G. W., Richardson, W. J., Roseneau, D.
G., Thomson, D. H., and Green, C. R. (1989). Analysis and ranking of the
acoustic disturbance potential of petroleum industry activitie.; and other
sources of noise in the environment of marine mammals in P, laska, pre-
pared by BBN Systems and Technologies Corporation, var. pug., OCS
study MMS 89-0006, U.S. Department of the Interior, Minerals Manage-
ment Service, Alaska OCS Office, Anchorage, AK, NTIS PB ')0-188673.
McCreery, C. S., Duennebier, F. K. and Sutton, G. H. (1993). "Correlation
of deep ocean noise (0.4-30 Hz) with wind, and the Hola Spectmm--A
worldwide constant," J. Acoust. Soc. Am. 93, 2639-2648.
McDonald, M. A., Webb, S.C., Hildebrand, J. A., Cornuelle, B. D., and Fox.
C. G. (1994). "Seismic Structure and Anisotropy of the Jmm de Fuca
Ridge at 45 ø N," J. Geophys. Res. 99(B3), 4857-4875.
Nishimura, C. E., and ConIon, D. M. (1994). "fUSS Dual Use: Vlonitodng
Whales and Earthquakes Using SOSUS," Mar. Technol. S,x'. J. 27(4),
13-21.
Nonhrnp, J. W., Cummings, W. C., and Thompson, P.O. (19rig). "20-Hz
Signals Observed in the Central Pacific," J. Acoust. Soc. Am. 43, 383-
384.
Northrop, J. W., Cummings, W. C., and Morrison, M. F. 0971) 'Underwa-
ter 20-Hz signals recorded near Midway Island," I. Acoust. Scc. Am. 49.
1909-1910.
Okal, E. A., and Talandier, J. (1986). "T-wave duration, magnitudes and
seismic moment of an earthquake-application to tsunami warning," J.
Phys. Earth 34, 19-42.
Patterson, B., and Hamilton, G. R. (1964). "Repetitive 20 cycle per second
biological hydroacoustic signals at Bermuda," in Marine Bio-Acoustics,
edited by W. N. Tavolga {Pergamon, New York), pp. 125-145
Payne, R., and Webb, D. (1971). "Orientation by means of long range
acoustic signaling in baleen whales," Ann. N.Y. Acad. Sci. 188. 110-141.
Richardson, W. J., W/irsig, B., and Green, Jr., C. R. (1986). "Reactions of
bowhead whales, Balaena mysticetus, to seismic exploration m the Cana-
dian Beaufort Sea," J. Acoust. Soc. Am. 79{4), I 117-1128.
Richardson, W. J., Wiirsig, B. and Greene, Jr., C. R. 0990). "Reactions of
bowhead whales, Balaena mysticetus, to drilling and dredging noise in the
Canadian Beaufort Sea," Mar. Environmental Res. 29, 135-160.
Richardson, W. J., Greene, Jr., Malme, C. 1., and Thomson, D. H. (1991).
Effects of noise on marine mammals, prepared by LGL EcoDgical Re-
search Associates Inc., pp. 462, OCS study MMS 90-0093, U.S. Depart-
ment of the Interior. Minerals Management Service, Atlantic OCS Region,
Hemdon, VA, 1991, NTIS PB 91-168914.
Richardson, W. 1., and Greene, Jr., C. R. (1993). "Variability in behavioral
reaction thresholds of bowhead whales to man-made underwater sounds,"
J. Acoust. Soc. Am. 94, 1848, Abstract.
Schevill, W. E., Watkins, W. A, and Backus, R. H. (1964). "The 20-cycle
signals and Balaenoptera (fin whales)," in: Marine Bio-Acou•tics, edited
by W. N. Tavolga {Pergamon, Oxford), pp. 2, 147-152.
Shooter, J. A., DeMary, T. E., and Wittenbom, A. F. (1990). "Depth depen-
dence of noise resulting from ship trat'tic and wind," [EEE I. Ocean Eng.
15(4), 292-298.
Sohn R. A., Hildebrand, J. A., Webb, S.C., McDonald, M. A., and Fox, C.
G. (1992). "Microearthquake •tudies at the Megaplume Site on the South-
ern Juan de Fuca Ridge," EOS 73, 530, Abstract.
Stone, G. S., Katona, S. K., Maiawaring, A., Allen, J. M., and Corbett, H. D.
(1992). "Respiration and surfacing rates of fin whales (Balaenoptera
physahu) observed from a lighthouse tower," Rep. lnt. What. Commun.
42, 739-745.
Talandier, J.. and Okal, E. A. (1984). "The Volcanoseismic Swarms of
1981-1983 in the Tahiti-Mehetia Area, French Polynesia," J. Geophys.
Res. 89, 11216-11234.
Talandier, J.. and Okal, E. A. (1987). Seismic Detection of Underwater Vol-
canism: The Example of French Polynesia (Birkhauser Verlag, Basel), pp.
919-950.
Thompson, E O. (1967). Marine Biological Sounds West of San Clemeate
Island, U.S. Navy Electron. Lab. Rep. 1290.
Thompson, E O., and Friedl, W. A. (1982). "A long term study of low
frequency sounds from several species of whales off Oahu, Hawaii," Ce-
Ology 45. 1-19.
Thompson, P. O., Findlay, L. T., and Vidal, O., Doublet Stereotyped and
other blue whale phonations recorded in the Gulf of California, Mexico.
Abstr. 7th Bleu. Confi Biol. Mar. Mamm., Dec. 1987, Miami, FL, p. 70.
Thompson, P. O., Findlay, L. T., and Vidal, O. (1992). "20-Hz pulses and
other vocalizations of fin whales, balaenoptera physalus, in the Gulf of
California, Mexico," I. Acoust. Sec. Am. 92. 3051-3057.
Thompson, T 1., Winn, H. E., and Perkins, E J. (1979). "Mysticere
Sounds," in Behavior of Maline Animals. Current Perspectives #• Re-
search, edited by H. E. Winn and B. L. Olla {Plenum, New 5•rk), Vol 3:
Cetaceans, pp. 403-43 I.
Tyack, P. L. (1993). "Reactions of bottlenose dolphins, Tursiops truncatus,
and migrating gray whales, Eschrichtius robustus, to experimental play-
back of low frequency man-made noise," J. Acoust. Soc. Am. 94, 1830,
Abstract.
Urick, R. J. (1983). Principles of Underrearer Sound (McGraw-Hill, New
York), pp. 377-405.
Walker. D. A., McCreery, C. S., and Yoshihisa, Y. (1992). "T-phase spectra,
seismic •no•nents, and tsunamigenesis," Bull Seismol. Soc. Am. 82(3),
1275- i 305.
Walker, D. A., and Bernard, E. N. (1993). '•Comparison of T-phase Spectra
and Tsunami Amplitudes for Fsunamigenic and Other Earthquakes," J.
Geophys. Res. 98(C7), 12557--12565.
Watkins, W. A. (1964). "Harmonic interval: Fact or artifact in spectral
analysis of pulse trains," in Marine Bio-Acoustics, edited by W. N.
Tavolga (Pergamon, New York), pp. 15-42.
Watkins, W. A. (1981). "Activities and underwater sounds of fin whales,"
Sci. Rep. Whales Res. Inst. 33, 83-117.
Watkins, W. A., and Schevill, V•: E. (1977). "Spatial Distribution of Phy-
seter catadon (sperm whales) underwater," Deep Sea Res. 24, 693-699.
Watkins, W. A., and Wartzok, D. (1985). "Sensory biophysics of marine
mammals," Mar. Mature. Sci. 1(3), 219-260.
WaLkins, W. A., Tyack, P., Moore, K. E., and Bird, J. E. (1987). "The 20-Hz
signals of finback whales (Bah•enoptera physalus)," •. Acoust. SOC. Am.
82, 1901-1912.
Webb, S.C. (1965). "The equilibrium oceanic microseism spectrum," J.
Acoust. Soc. Am. 92, 2141-2158.
Weston, D. E., and Black, R. L. "Some unusual low-frequency biological
noises underwater," Deep Sea Res. 12, 295-298.
Wenz, G. M. (1962}. "Acoustic Ambient Noise in the Ocean: Spectra and
Sources," J. Acoust. Soc. Am. 34•.12), 1936-1956.
Wilson, J. H. (1979). "Very Log, Frequency {VLF) Wind-Generated Noise
Produced by Turbulent Pressure Fluctuations in the Atmosphere near the
Ocean Surface," J. Acoust. Soc. Am. 66, 1499-1507.
Winn, H. E., Thompson, T 1., Cummings, W. C., Hain, J., Hurlnail. 1., Hays,
H., and Steiner, W. W. (1981). "Song of the Humpback Whale-Population
Comparisons," Behar. Ecol. Sc½iobiol. 8. 41-46.
Yochem, P. K., and Leatherwood, S. (1985). "Blue Whale: Bah•enoptera
musculus (Linnaeus, 1758)," in Handbook of Marine Mammals, Vol 3:
The Sireninns and Baleen Whales (Academic, London), p. 223.
721 J. Acoust. Soc. Am., Vol. 98, No. 2, Pt. 1, August 1•95 McDonald eta/.: Whales observed on a seafloor arroy 721
