port and ocean engineering under arctic conditions - Poac.com

PORT AND OCEAN
ENGINEERING
UNDER ARCTIC CONDITIONS
VOLUME 11
SYMPOSIUM ON NOISE AND MARINE MAMMALS
Edited by
Symposium Organizers
and Editors
W.M. SACKINGER. Ph. D.. P.E. J.L. IMM
M.O. JEFFRIES. Ph. D. S.D. TREACY
The Geophysical Institute
Minerals Management Service
University of Alaska Fairbanks
U.S. Department of the Interior
Anchorage, Alaska
The Geophysical Institute
University of Alaska Fairbanks
Fairbanks, Alaska

Beluga whale (Delphinapterus leucas) in new ice.
Photo Credit: Naval Ocean Systems Center

Copyright @ 1988 by the Geophysical Institute, University of Alaska Fairbanks. All rights
reserved. No part of this publication may be reproduced, stored in a retrieval system, or
transmitted in any form or by any means, electronic, mechanical, photocopying, recording,
or otherwise, without the prior written permission of the publisher, the Geophysical
Institute, University of Alaska Fairbanks, Fairbanks, Alaska 99775-0800, U.S.A.
ISBN 0 - 915360 - 06 - 3

PREFACE
The series of conferences on Port and Ocean Engineering under Arctic
Conditions (POAC) is organized biennially by national POAC committees under the
long-term policy direction of the POAC International Committee. Previous POAC
conferences have been held in Norway (21, Canada (21, Iceland, Finland, Greenland
and Alaska. The Ninth Conference (POAC-87) in the POAC series was held at the
University of Alaska Fairbanks, Alaska, USA from August 17-2 1, 1987. This multi-
volume book, entitled "Port and Ocean Engineering Under Arctic Conditions", is a
compilation of the papers written for and presented at POAC-87.
A total of 224 people registered for POAC-87 and 122 papers were presented
during 14 sessions. The sessions were: Arctic Database; Ice Properties;
Icebreaking Vessels; Ice Modelling; Arctic Port Design; Geotechnical; Ice-structure
Interaction; Ice Morphology; Ice Dynamics; Ice, Climate and Forecasting; Spray Ice;
Remote Sensing; and two special symposia on Noise and Marine Mammals, and
SteelIConcrete Composite Structural Systems.
Papers submitted to POAC-87 were reviewed and edited prior to publication. All
the papers in this book have been refereed by two, three, or more reviewers, and then
edited, to try to ensure a consistent and high standard for technical content, style
and format for publication. Once accepted for publication, authors submitted a
camera-ready copy of their papers. The majority of papers in this book were
verbally presented at POAC-87; a few authors were unable to attend the conference,
but their papers have been published since they met the necessary review and editorial
standards.

ACKNOWLEDGEMENTS
Many individuals and organizations contributed to the success of POAC-87 and to
the publication of this book.
The conference SPONSORS were:
University of Alaska Fairbanks
Geophysical Institute, University of Alaska Fairbanks
Minerals Management Service, Technology Assessment and Research Program
National Science Foundation
Minerals Management Service, Environmental Studies Branch, Alaska OCS Region
Alaska Oil and Gas Association, Lease Planning and Research
Committee, Member Companies:
Amoco Production Company
ARC0 Alaska, Inc.
BP Alaska Exploration Inc.
Chevron USA, Inc.
Conoco, Inc.
Elf Aquitaine Petroleum
Exxon Company, USA
Marathon Oil Company
Mobil Oil Corporation
Shell Western E & P, Inc.
Standard Alaska Petroleum Company
Unocal Corporations
and the CO-SPONSORS were:
American Society of Civil Engineers
Alaska Academy of Engineering and Science
Centre for Frontier Engineering Research (C-FER)
Le Comite Arctique International
The long-term policy of the conferences on Port and Ocean Engineering under
Arctic Conditions is directed by the POAC INTERNATIONAL COMMITTEE (1987):
Prof. Per Tryde
Technical University of Denmark (President)

Mr. Alf Engelbrektson
VBB-SWECO Engineers, Stockholm, Sweden (Vice President)
Prof. Per Bruun
The Norwegian Institute of Technology. Trondheim, Norway
(Secretary General)
Prof. William M. Sackinger
University of Alaska Fairbanks, Fairbanks, Alaska, USA
(Past President)
Dr. Pauli Jumppanen
Oy Wartsila Ab. Helsinki, Finland (Past President)
Prof. Bernard Michel
Lava1 University, Quebec, Canada (Past President)
Mr. K.R. Croasdale
Esso Resources Canada, Calgary, Alberta, Canada
Prof. G.R. Peters
Memorial University of Newfoundland, St. John's,
Newfoundland, Canada
Dr. K. Takekuma
Nagasaki Technical Institute/Mitsubishi Heavy
Industries, Nagasaki, Japan
Dr. -1ng. Joachim Schwarz
Hamburgische Schiffbau-Versuchsanstalt, Hamburg, Germany
Mr. G. Viggoson
Vita og Hafnamala Stjarinn, Reykjavik, Iceland
Dr. E. Enkvist
Wartsila Arctic Research Centre, Helsinki, Finland
Dr. T. Carstens
Norwegian Hydrodynamics Labs, Trondheim, Norway
Prof. Xu Ji-zu
Tianjin University, Tianjin, China
Dr. W.F. Weeks
University of Alaska Fairbanks, Fairbanks, Alaska, USA

POAC-87 was organized by the U.S. NATIONAL ORGANIZING COMMITTEE:
Prof. W.M. Sackinger, Chairman; University of Alaska Fairbanks, Fairbanks,
Alaska
Mr. Muhammed A. Ali
Chevron Corporation, San Francisco, California
Prof. F. Lawrence Bennett
University of Alaska Fairbanks, Fairbanks, Alaska
Mr. Chris Birch
State of Alaska Department of Transportation, Fairbanks, Alaska
Mr. Irving Boaz
Shell Oil Company, Houston, Texas
Comdr. Lawson W. Brigharn
U.S. Coast Guard, Boston, Massachusetts
Mr. David Chiang
Science Applications International Corp., McLean, Virginia
Prof. Jin S. Chung
Colorado School of Mines, Golden, Colorado
Mr. Roger Colony
University of Washington, Seattle, Washington
Dr. M.J. Feifarek
Marathon Oil Company, Houston, Texas
Mr. Joseph Galate
Enertech Engineering & Research Company, Houston, Texas
Prof. Ben. C. Gerwick, Jr.
University of California-Berkeley, Berkeley, California
Mr. H. Glenzer, Jr.
State of Alaska Department of Transportation, Fairbanks, Alaska
Mr. Roger Herrera
Standard Alaska Production Company, Anchorage, Alaska
Mr. Malcolm W. Howard
BP Petroleum Development Ltd., London, United Kingdom
Mr. Jerry Imm
Minerals Management Service, Anchorage, Alaska
vii

Dr. Martin 0. Jeffries
University of Alaska Fairbanks, Fairbanks, Alaska
Dr. Jerome B. Johnson
USA CRREL, Ft. Wainwright, Alaska
Mr. Austin Kovacs
USA CRREL, Hanover, New Hampshire
Dr. Thomas Kozo
US Naval Academy, Annapolis, Maryland
Prof. Charles Ladd
Massachusetts Institute of Technology, Cambridge, Massachusetts
Dr. Malcolm Mellor
USA CRREL, Hanover, New Hampshire
Dr. Thomas Osterkamp
University of Alaska Fairbanks, Fairbanks, Alaska
Mr. Dennis Padron
Han-Padron Associates, New York. New York
Dr. Robert S. Pritchard
Ice Casting, Inc., Seattle, Washington
Prof. Louis Rey
Le Comite Arctique International, Monte Carlo, Monaco
Ms. Patricia Sackinger
Fairbanks, Alaska
Mr. Terry Setchfield
Exxon Production Research Company, Houston, Texas
Prof. Lewis Shapiro
University of Alaska Fairbanks, Fairbanks, Alaska
Dr. Harold Shoemaker
US Department of Energy, Morgantown, West Virginia
Dr. Charles E. Smith
Minerals Management Service, Reston, Virginia
Mr. Rodney Smith
Minerals Management Service, Anchorage, Alaska
Dr. Walter Spring
Mobil Research and Development Corporation, Dallas, Texas

Prof. William Stringer
University of Alaska Fairbanks, Fairbanks, Alaska
Mr. Larry Sweet
University of Alaska Fairbanks, Fairbanks, Alaska
Prof. Shyam Sunder
Massachusetts Institute of Technology, Cambridge,
Massachusetts
Mr. Stephen D. Treacy
Minerals Management Service, Anchorage, Alaska
Mr. Michael Utt
Unocal Corporation, Brea, California
Dr. Ken Vaudrey
Vaudrey & Associates, San Luis Obispo, California
Mr. Robert Visser
Belrnar Engineering and Management Service Co., Redondo Beach, California
Dr. Vitoon Vivatrat
Engineering Science Inc., Houston, Texas
Dr. W.F. Weeks
University of Alaska Fairbanks, Fairbanks, Alaska
Prof. Gunter Weller
University of Alaska Fairbanks, Fairbanks, Alaska
Dr. J. Patrick Welsh
Naval Ocean Research and Development Activity, Hanover, New Hampshire
Mr. Jonathan Widdis
State of Alaska Department of Transportation, Fairbanks, Alaska
Dr. Jay Wiedler
Brown and Root USA, Houston, Texas
An important and vital task in the organization of POAC-87 and preparation of
papers for publication was the review and evaluation of abstracts and papers. In
addition to all members of the International Committee and the U.S. National Organiz-
ing Committee, the reviewers included:
Dr. H. Burcharth, University of Aalborg, Denmark
Dr. A. Chen, Exxon Production Research Company, Houston, Texas

Mr. Li Fu-cheng , University of Alaska Fairbanks, Fairbanks, Alaska
Dr. James U. Kordenbrock. David Taylor Research Center, U.S. Navy
Mr. Donald Kover, David Taylor Research Center, U.S. Navy
Dr. C.-H. Luk, Exxon Production Research Company, Houston, Texas
Dr. Lasse Makkonen, Technical Research Centre of Finland
Dr. A. L. Mindich, Mirza Engineering Inc., Chicago, Illinois
Mr. J. Poplin, Exxon Production Research Company, Houston, Texas
Dr. T. D. Ralston, Exxon Production Research Company, Houston, Texas
Dr. Philip A. Sackinger, Massachusetts Institute of Technology, Cambridge,
Massachusetts
Dr. A. Wang. Exxon Production Research Comp,any, Houston, Texas
Also helping with conference organization and the preparation of this book were
Kathryn Coffer, Nancy Smoyer, Jan Dalrymple and Kim Morris. Day-to-day conference
administration and co-ordination was by the Conferences and Institutes Office,
University of Alaska-Fairbanks (Nancy Bachner and staff). Special thanks are due
to Dr. S.-I. Akasofu, Director, Geophysical Institute, and to Dr. P.J. O'Rourke,
Chancellor, University of Alaska-Fairbanks, for their encouragement and financial
support; the encouragement of Dr. Harold D. Shoemaker of the U.S. Department of
Energy was also appreciated.
To our sponsors and co-sponsors, the International and National Organizing
Committees, and all those individuals who helped make POAC-87 and the
publication of this book possible, our grateful thanks.
William M. Sackinger
Martin 0. Jeffries
Fairbanks
January 1988

FOREWORD
Many marine mammal species found in arctic waters have important
relationships with ice. Many are pagophilic, using ice as a platform to
haul out (ringed, spotted, bearded, ribbon, and harp seals; and walrus)
or to hunt and scavenge (polar bear and arctic fox). Some marine
mammals relate to ice as a floating barrier around, through, and under
which their seasonal migrations proceed (bowhead whales, beluga, and
narwhal) or as an encroaching northern border that may ultimately prompt
an annual migration to warmer waters (gray whale).
Marine mammals in ice-covered waters are subject to continual
auditory input from a highly active acoustic environment. Some input
is airborne, such as that which occurs at or above the water's surface (e.g.,
on land or ice as experienced by pinnipeds and polar bears). However,
marine mammals mostly hear sounds that are generated, transmitted,
and/or received underwater. Acoustic input may derive from natural
sources including wave active, seismic activity, ice movement and
breakage, as well as sounds produced by the above species and other
biota. In addition, noise produced by various human activities
contYibutes to overall loading of the acoustic environment. These
anthropogenic noises are essentially a by-product of shipping, oil and gas
exploration and development, fishing vessels, various activities of coastal
communities, and activities of other marine industries.
Ice movement and breakage produce pervasive,and at times
explosive,noise in arctic waters. The presence of sea ice also partially
controls the underwater acoustic environment by providing a rough
reverberative ceiling for local sound waves. Even when ice recedes, it
continues to affect the underwater acoustic environment as surface layers
warm up differentially in the water column, causing sound waves to refract
downward. This refraction results in higher propagation loss in shallow
water through increased contact of sound waves with bottom topography.
Depending on the pattern, sound frequency, and intensity of the sound
source in combination with ambient oceanographic features (e.g., water
depth, salinity, presence of ice, underwater substrate), various sounds may
be discernible to these species above ambient-noise conditions, thereby
potentially influencing individual animals to some degree. Such

influences on marine mammals may include changes in behavior such
as curious attraction or avoidance reactions, changes in physiological rates,
and interference with communicative or echolocative efforts.
These topics were the focus of the "Symposium on Noise and Marine
Mammals in Ice-Covered Waters." The symposium, held August 18,
1987 at the University of Alaska Fairbanks, was coordinated by the
Minerals Management Service and was part of the 9th International
Conference on Port and Ocean Engineering Under Arctic Conditions. The
concept of this Symposium was originally suggested by Dr. Louis Rey,
and the Symposium was co-sponsored by Le Comite Arctique International.
The titles and authors of scientific papers presented at the symposium
are as listed in the Table of Contents which follows, with the exception
of the paper "Possible effects of ambient noise on the ability of the Bowhead
Whale, Balaena mysticetus, to discern under-ice reverberations from their
calls" (William T. Ellison and Christopher Clark). This paper was
presented by Mr. Charles I. Malme at the request of the authors, but is not
included in this volume.
A panel discussion on the subject of potential directions for research
was held immediately following presentation of the submitted papers. Two addi-
tional papers, entitled "Evidence of Glacial Seismic
Events in the Acoustic Environment of Humpback Whales" (Paul R. Miles and Charles
I. Malme) and "Review of Studies on the Effects of
Man-Induced Noise on Marine Mammals of the Bering, Chukchi, and
Beaufort Seas and How the Results Have Been Applied to Federal Oil and
Gas Management Decisions" (Cleveland J. Cowles and Jerry L. Imm),
although not presented at the symposium, are included in this volume
as useful background information.
The authors of these papers are to be commended for their contributions to the
Symposium and to this volume.
Jerry Imm
Stephen D. Treacy
Minerals and Management Service
Anchorage
March 1988
xu

TABLE OF CONTENTS
VOLUME I1
..............................................................
Preface iii
Acknowledgements ........................................ .............. v
..............................................................
Foreword xi
REVIEW OF STUDIES ON THE EFFECTS OF MAN-INDUCED NOISE ON MARINE MAMMALS OF
THE BERING, CHUKCHI, AND BEAUFORT SEAS AND HOW THE RESULTS HAVE BEEN APPLIED
TO FEDERAL OFFSHORE OIL AND GAS MANAGEMENT DECISIONS
Cleveland J. Cowles and Jerry L. Imm. ........................................ 1
UNRESOLVED ASPECTS CONCERNING THE INFLUENCE OF NOISE ON MARINE MAMMALS
J.M.Terhune ............................................................ 9
EFFECTS OF INDUSTRIAL ACTIVITIES ON RINGED SEALS IN ALASKA. AS INDICATED BY
AERIAL SURVEYS
Kathryn J. Frost and Lloyd E. Lowry. ..................................... 15
RESPONSES OF RINGED SEALS (Phoca hispida) TO NOISE DISTURBANCE
Brendan P. Kelley, John J. Burns and Lori T. Quakenbush ................ . . 27
RESPONSES OF MIGRATING NARWHAL AND BELUGA TO ICEBREAKER TRAFFIC AT THE
ADMIRALTY INLET ICE-EDGE, N.W.T. IN 1986
Susan E. Cosens and Larry P. Dueck ........................................ 39
OBSERVATIONS OF FEEDING GRAY WHALE RESPONSES TO CONTROLLED INDUSTRIAL
NOISE EXPOSURE
Charles I. Malme, Bernd Wursig, James E. Bird and Peter Tyack ........................ 55
INDUSTRY OBSERVATIONS OF BOWHEAD WHALES IN THE CANADIAN BEAUFORT SEA,
1976-1985
JohnG.WardandE.Pessah ........................................... . 75
MASKEDDETECTIONTHRESHOLDSFORTHEBELUGAANDBOTTLENOSEDOLPHIN
Charles W. Turl, Ralph H. Penner and W.W.L. Au ............ . . . . . . . . 89
EVIDENCE OF GLACIAL SEISMIC EVENTS IN THE ACOUSTIC ENVIRONMENT OF HUMPBACK
WHALES
Paul R. Miles and Charles I. Malme .......................................... 95
PANEL DISCUSSION
Stephen D. Treacy .
Author List ..................................... . . . . . . . . . . . . . . . Ill

REVIEW OF STUDIES ON THE EFFECTS OF MAN-INDUCED NOISE ON MARINE
MAMMALS OF THE BERING, CHUKCHI, AND BEAUFORT SEAS AND HOW THE
RESULTS HAVE BEEN APPLIED TO FEDERAL OFFSHORE OIL AND GAS
MANAGEMENT DECISIONS
Abstract
Since 1980, the U. S. Minerals
Management Service has managed and
funded a variety of studies of the
potential effects of man-induced noise
on marine mammals of the Bering,
Chukchi, and Beaufort Seas. The
purpose of such studies is to provide
information needed for informed
decisionmaking pertaining to environ-
mentally sound leasing and management
of offshore oil and gas development on
the Alaska Outer Continental Shelf
(OCS). Many of the noise/ marine
mammal-interaction studies have been
used in establishing lease specifica-
tions or regulations for offshore
operations in Federal lease areas.
Results also have been important in the
resolution of litigation pertaining to
OCS oil and gas leasing and explora-
tion. Specific examples of how results
have been applied are presented and
future Alaska information needs in this
discipline are discussed.
Introduction
Since 1980 the Minerals Management
This is a reviewed and edited version of apaper submit-
ted to the Ninth International Conference on Port and
Ocean Engineering Under Arctic Conditions, Fairbanks,
Alaska, USA. August 17-22, 1987.
Cleveland J. Cowles
Jerry L. Imm
Minerals Management Service, Anchorage, Alaska, USA
Service (MMS) has funded a variety of
studies of man-induced noise effects on
marine mammals of the Alaska Outer
Continental Shelf (OCS). Within the
context of pertinent legislation, MMS
identified the various sources of noise
associated with offshore oil and gas
activities and designed studies to
examine potential effects. One under-
lying basis for selection of research
topics to date has been application of
a key- species approach (described in
Cowles and Im, 1980). In many cases,
the results of these studies have been
used in management decisions such as
formation of regulations or mitigating
measures. This has affected the way
industry has been permitted to operate
in the Bering, Chukchi, and Beaufort
Seas. We briefly review study efforts
and link significant findings with
management decisions.
General Review of Past MMS Noise-
Effects Studies
Of the several noise sources from
offshore oil and gas activities, some
of the most common ones are ships
(icebreakers, drillships, geophysical
ships, conventional and icebreaking
ships), work boats, aircraft (fixed
wing and rotary wing), dredges, drill-
ing rigs, production rigs, and
pipelines--used either directly in
drilling and production activities or

indirectly in support of such opera-
tions. For clarity of discussion, the
studies and respective marine mammal
species are discussed below by noise-
source categories.
Aircraft: Starting in 1980, bowhead
whales (Balaena mysticetus) were
studied in the presence and absence of
industrial stimuli; as a part of this
research, bowhead whale behavior in
response to fixed-wing and helicopter
noise was studied under different
operating conditions. Overt reactions
to the fixed- wing observation aircraft
were sometimes conspicuous when the
aircraft was below 457 m a.s.1..
uncommon at 457 m, and generally
undetectable at 610 m (Richardson,
1983, 1985).
With helicopters, no overt
responses at approximately 153 m a.s.1.
were noted during two field experiments
and three opportunistic observations
from 1981 through 1984. No significant
changes in blow intervals were found;
thus, there is no conclusive evidence
that single helicopter passes (at
greater than 153 m a.s.1.) disturb
bowhead whales that are below the
surf ace.
Malme et al. (1983, 1984) per-
formed systematic studies of whale
responses to aircraft and reported that
gray whales (Eschrichtius robustus)
tended to avoid a location where
recorded helicopter noise was played
back into the water. However, the
playback rate of one simulated pass
every 10 s to 2 min greatly exceeded
typical helicopter-traffic rates along
routes to offshore industrial sites
(Richardson, 1985).
Boats: Vessel traffic is a major
source of potential disturbance to
bowhead whales near areas being
explored or developed by the petroleum
industry. In general, of all the
stimuli presented to bowheads in our
studies, "small"-boat traffic elicited
the greatest variety of responses
(Richardson, 1985). The studies showed
that bowheads demonstrated a strong or
frequent reaction to boats at a dis-
tance of 1 to 2 km. Other baleen
whales have shown considerable toler-
ance of boats but often have avoided
rapidly or erratically moving vessels.
Baker et al. (1982) found changes in
the respiration rates and diving
behavior of humpback whales (Megaptera
novaeangliae) when boats were within
about 900 m; vessels that approached
closely and moved erratically had the
greatest effects. Sorenson et al.
(1984) found evidence that "squid-
eating" toothed and beaked whales were
less common near boats than elsewhere;
no such effect was found for "fish-
eating" cetaceans, including some
baleen whales (Richardson, 1985).
The long-term effects of boat
disturbance on whales are especially
difficult to assess. The MMS hopes to
develop some behavioral information
regarding bowhead whalelicebreaker
interaction as part of future
research that is mentioned below.
Dredging: Pertinent research has
included both actual noise from dredge
operations and playback of dredging
noise. For actual dredging, the noise
was detectable for at least several
kilometers; and bowheads seemed to
behave normally within the ensonified
zone. For playbacks, bowheads
responded to strong dredge noise even
when the noise level was increased
gradually over time.
Drilling: Bowheads have been studied
near operating drillships, well within
the zone where drillship noise is
clearly detectable. General activities
of these animals seemed normal; and
there was no conclusive evidence that
the noise affected surfacing, respira-
tion, or dive cycles.
The sightings near drillships
showed tolerance of drilling but did
not prove that bowheads are unaffected
by drillships. Playback experiments
showed that some bowheads reacted,
although not strongly, to drillship
noise at intensities equivalent to
several kilometers from a real
drillship (Richardson, 1985). The
results for summering bowheads were
generally consistent with reactions of
migrating gray whales to the same
drillship noise (Malme et al., 1983,
1984). Migratory gray whales
approaching the sound source tended to
change speed and course only slightly.

Migratory gray whales were exposed to
drillship-noise levels in the 50- to
315-Hz band at levels of 110, 117 and
122 dB re 1 uPa, respectively. Play-
back exposure of feeding gray whales
did not produce clear evidence of
disturbance or avoidance behavior at
received levels below 110 dB. Thus, at
a conservative 120-dB threshold, a
typical drillship would affect feeding
gray whales at a range of about 300 m.
Other observations indicate that
reaction thresholds of bowhead and gray
whales to playbacks of drilling noise
are similar (Richardson, 1985).
Observations of beluga whale
(Delphinapterus leucas) responses to
playbacks of oil-drilling sounds
indicate that the direction of whale
movement and general activity (feeding,
traveling) are not greatly affected by
these sounds, especially if the sound
source is constant. Whales continued
to move in the direction they were
travelling before playbacks began. On
several occasions, beluga whales within
2 km of the sound source appeared to
feed during playback experiments.
Whales also approached and quickly
passed closely by the underwater
speaker while sounds were being pro-
jected (Stewart, Aubrey, and Evans,
1983).
Now that data are available, the
MMS' approach has evolved from the
field-experiment phase to the site-
specific, predictive modeling of whale
response to drilling operations. These
studies (discussed later) for post-sale
analyses have addressed a diversity of
technology actually used in drilling
operations in arctic waters.
Production and Pipelines: The MMS has
not directly studied these two noise
producers but hopes to in the near
future by using controlled-playback
experiments in the spring lead system
near Point Barrow. Research work would
be conducted "downstream" of the whale
hunting and censusing locations to
avoid any disturbance to the whales and
subsistence whaling activities. The
MMS anticipates that coordination with
local organizations will be as impor-
tant a component of this study as is
the technical work.
Seismic Boats: Studies of endangered
whale and seismic boat interactions
have been a particularly crucial aspect
of the MMS studies program. Actual and
potential litigation, seismic-
exploration prohibition (e.g., eastern
Alaskan Beaufort Sea in 1981), and
indirect human effects (e.g., conflicts
with aboriginal whaling) have led to
complex studies spanning three species
(bowhead, gray, and humpback whales)
and as many oceanic provinces (Pacific,
Bering, and Arctic).
In general, considering the
source-level intensity (245-250 dB re 1
uPa) of most airgun impulses, the
avoidance responses of the latter
species are relatively near-field
phenomena. As a result of cooperative
studies by MMS-sponsored researchers
and geophysical operators, we have
found that bowheads in Alaskan waters
will not avoid operating seismic boats
at ranges farther than 3.5 to 5.0 tan
and at received levels less than 160 to
170 dB re 1 uPa (Ljungblad et al.,
1985; Richardson, 1985). Migratory
gray whale avoidance of operating
seismic boats will occur at a received
peak pressure level of 164 dB re 1 uPa
and at a range of about 2.5 km (Malme
et al., 1984). Because this latter
gray whale study may have affected sea
otters in the study area (USDOI, FWS,
1982), sea otters also were studied.
It was found that sea otters displayed
little, if any, reaction to a wide
array of sound sources, including
seismic boats at ranges of 900 m to
1.6 tan. Other studies with smaller
airgun experiments showed no clear
evidence of avoidance by feeding
humpbacks exposed at effective pulse-
pressure levels of 172 dB re 1 'uPa
(Malme et al., 1985). Feeding gray
whales near St. Lawrence Island,
Alaska, avoided a single-airgun
exposure when average pulse levels
reached 173 dB re 1 uPa. In specific
''typical" locations that were analyzed,
gray whale avoidance would occur at a
range of about 3 km (Malme et al.,
1986).
The work summarized above repre-
sents the primary information base of
past and future decision processes
pertaining to offshore seismic

exploration/endangered whale issues.
On-Ice Seismic Exploration: In the
late 1970's and early 1980's. analyses
of biological data suggested that
spring distribution and reproduction of
ringed seals may have been affected by
on-ice seismic exploration that was
conducted primarily by "Vibroseis"
methodology. Beginning in 1981, MMS
sponsored a series of field experiments
and other studies that examined change
in ringed seal behavior, ice-lair use,
distribution, and the affected acoustic
environment. The overall results of
these efforts (see Burns and Kelly,
1983; Frost, Burns, and Lowry, 1985)
showed that abandonment or altered use
of seal lairs occurred mainly within
150 m of seismic lines. Comparison of
seal densities, based on aerial or
ground surveys, in "seismic" and
"control" areas produced mixed results.
Detailed radio-telemetry studies of
seal behavior (Kelly, Quackenbush, and
Rose, 1986) in Kotzebue Sound, Alaska,
and further acoustical analyses there
(Cummings, Holiday, and Lee, 1984)
enhanced our understanding of potential
effects on ringed seals. Monitoring
studies that addressed regional abun-
dance and distribution offshore Alaska
were ongoing through spring 1987.
Pre-Lease Decisions Utilizing Noise-
Effects Studies
Endangered Species Act Section 7
Consultation: One of the most impor-
tant applications of the noise-ef f ects
studies has been their relevance to
Endangered Species Act (ESA) section 7
compliance, particularly to the evolu-
tion of Biological Opinions on proposed
lease sales provided to MMS by the
National Marine Fisheries Service
(NMFS). Early NMFS opinions on pro-
posed lease sales in the Beaufort Sea
concluded that, "There is too little
information to determine whether the
lease sale and all resulting activities
are likely to jeopardize the continued
existence of the bowhead" (USDOC, NOAA,
NMFS, 1980). Thus, NMFS asserted that
in the face of insufficient information
on oil-spill and noise effects, MMS
could not clearly avoid jeopardy of the
bowhead population. Recently, however,
a Biological Opinion on activities in
the same region concluded, "The leasing
and exploration phases of Lease Sale 97
are not likely to jeopardize the
continued existence of any endangered
or threatened marine species" (USDOC,
NOAA, NMFS, 1987). In supporting
comments pertaining to their assessment
of noise effects in the Sale 97 area,
NMFS stated, "This opinion is based on
the best available information includ-
ing noise effects studies on bowhead
whales summering in the Canadian
Beaufort Sea" (USDOC, NOAA, NMFS,
1987).
In addition to enhancing these
generalized analyses, other ESA-related
operational consultations have bene-
fited. Study recommendations, conser-
vation measures, and reasonable and
prudent alternatives expressed in
Biological Opinions also must take new
information into account.
Seismic-Vessel Exploration-Permit
Requirements: As mentioned previously,
after concern for seismic-vessel
effects on the fall migration of
bowhead whales reached unprecedented
levels in the early 19801s, entire
offshore areas of the Beaufort Sea were
closed to seismic exploration if whales
were known to be present. This
approach to seismic-vessel management
lowered profit expectations consider-
ably, and vessel owners sought relief.
Essentially they questioned closures on
the basis of presence when
effects on whales had not been demon-
strated. Subsequently, and with
improved information in hand on bowhead
responses to seismic boats--especially
data determined by acoustically
oriented field experiments (Ljungblad
et al., 1985; Richardson, 1985)~
permits for Beaufort Sea seismic
operations were revised under terms
that allow exploration beyond known
whale-response distances. For example,
a seismic permit for Beaufort Sea
exploration now typically requires
special precautions of the vessel
operator during bowhead migrations:
"After the beginning of the whale
migration, seismic vessels can operate
their high energy sources only when
visibility exceeds 3 miles. During
periods of fog, darkness, or weather
conditions which limit visibility to

less than 3 miles, the seismic sound
sources must be shut down. Operations
cannot be initiated or resumed until an
area with a radius of 5 miles from the
vessel is clear of whales. This may
require the use of aircraft."
The specific distances referred to
are derived primarily from the result
of MMS studies that established thres-
hold distances related to bowhead whale
disturbance.
Probably one of the most publi-
cized applications of noise-effects
studies in resolving seismic vessel1
endangered whale conflicts was their
use in litigation decisions on proposed
St. George Basin Lease Sale 70. This
area, just north of Unimak Pass in the
southern Bering Sea, is suspected
habitat of right whales (Balaena
glacialis) and is adjacent to the
spring and fall primary migration route
of gray whales. Following a suit filed
to block the sale (Village of False
Pass v. Watt, D. Alaska, Cir. No.
A83-176) the U. S. District Court,
Alaska, ruled (aspects of this ruling
were later overturned) that the Secre-
tary of the Interior could not execute
leases until:
(1) A "Worst-case" analysis of
seismic effects on gray and right
whales or a supplementary Environmental
Impact Statement (EIS) evaluating
effects of preliminary seismic
exploration was prepared.
(2) The Final Notice of Sale or
other order must include restrictions
implementing reasonable and prudent
alternatives contained in the relevant
Biological Opinion or justification
that such restrictions are not
necessary.
Subsequently, MMS issued a supple-
mental EIS that focused heavily on
recent studies results and revised
restrictions on seismic operations. In
obtaining NOAA concurrence on the
adequacy of a draft "Notice to Lessees"
(NTL) for protecting whales from
potential seismic effects, Good (1983)
wrote :
"The Minerals Management Service
believes these restrictions are more
than adequate to protect gray and right
whales. Deep seismic surveys have been
conducted in the Bering Sea since the
early 1970's. The National Marine
Fisheries Service routinely allows deep
seismic surveys to come within 20 miles
of Unimak Pass during the gray whale
-
d a deep-seismic system, far more
powerful than a high resolution system.
The whales came as close as three miles
to the airguns before 'some possible
changes in the swimming patterns of
cow-calf pairs were observed. . .
' (Emphasis added .)
As is evident above, important
information obtained from a noise-
effects study was assimilated directly
into the OCS management-decision
process. The preliminary results
referred to were later reported in
Malme et al. (1984). Following the
issuance of the NTL and the supple-
mental EIS, the case was closed.
On-Ice Seismic Exploration:
Similarly, studies results have
affected the decision process regarding
permits and regulations of on-ice
seismic explorations. At first, in the
face of substantial uncertainty about
the extent of potential effects on
ringed seals, conservative management
approaches (such as seasonal termina-
tion of all seismic activities) were
implemented by regulatory agencies.
Industry groups were, of course,
concerned and sought changes through
legislative processes. In 1982,
regulations governing the small take of
marine mammals incidental to specified
activities under section 101(a)(5) of
the Marine Mammals Protection Act were
proposed to deal specifically with the
ringed seal issue. Among information
considered (Federal Register, 1982a),
NMFS cited the findings of "Burns et
al., (1981)" and other results of
MMS-sponsored, June 1981 studies.
These studies were instrumental in
showing that although on-ice seismic
activities may affect a small number
(less than 1,000) of seals in the area
covered by seismic activities, the
"taking" would have a negligible impact
on the 2.5 million-animal population.

Subsequently, a final rule on the
matter (Federal Register, 1982b) was
issued to allow "small takes" of ringed
seals. This action provided a new
framework by which MMS could structure
permits for on-ice seismic exploration,
and which ultimately enhanced seismic-
exploration opportunities while
providing appropriate protection for
this valued species. We believe that
environmental-studies results pertain-
ing to the acoustic environment and
this issue were particularly instru-
mental in resolving this issue.
Mitigating Measures: Much of the
various types of information obtained
from noise-effects studies has an
important but difficult-to-quantify
influence on environmental assessment
and, ultimately, the lease conditions
related to offshore oil and gas
development. Many noise-effects issues
are now better understood; therefore,
environmental analyses (in ESA section
7 consultation, EIS's, and exploration
plan reviews) are better supported with
scientific results. Almost all Alaska
lease sales are now accompanied by
NTL's formulated on the basis of
noise-effects-studies results.
Post-Lease Decisions Utilizing
Noise-Effects Studies
ESA Section 7 Consultation:
Biological Opinions on pre- and post-
lease operations--issued by NMFS after
consultation with MMS--are phased and,
as such, have been recognized by the
Federal courts. Prior to any approval
of post-lease development plans,
consultation between MMS and NMFS will
have occurred and the resulting Biolog-
ical Opinion, which takes into account
noise studies available up to that
time, will have been prepared.
Monitoring: For several years
industry has been required to have in
place a whale-observation program while
conducting exploration-drilling activi-
ties during the bowhead migration.
Seasonal drilling restrictions have
been waived if appropriate studies are
ongoing and if industry has met other
requirements. The determination of
Zones of Influence (21) around opera-
tions within which bowhead whales are
considered likely to react to acoustic
stimuli has provided a frame of refer-
ence for these observation programs.
These are challenging studies to design
and carry out, since there are many
uncontrolled variables, i.e., the
changing acoustic environment, the
changing chemical and physical environ-
ments, and the changes in whale
behavior due to factors that cannot be
differentiated from man-induced
changes. McLaren et al. (1986) is an
example of an industry-sponsored
monitoring study; other studies are in
progress. The MMS continues to provide
regionwide aerial monitoring that
shares data with site-specific studies.
Mitigating Measures: There have
been several mitigating measures
utilized on the Alaska OCS that are
directly related to the potential for
acoustic disturbance of whales, other
marine mammals, and birds. These
studies usually involve reconmending
horizontal and vertical separation of
operations from individual animals or
aggregations, as well as cessation of
activities until the animals have
departed a 21.
Post-Sale Environmental Assess-
- ment: One study designed for post-sale
application is "Prediction of Drilling
Site-Specific Interaction of ~ndustriai
Acoustic Stimuli and Endangered Whales
in the Alaskan Beaufort Sea." In order
to enhance environmental assessment
predictive techniques and accuracy,
this study measured sound characteristics
at drilling sites and used
models derived from previous studies to
predict response zones. Results will
be useful in future exploration-plan
reviews and other post-sale
applications.
The MMS recently contracted to
study the Davis Strait bowhead stock,
which is relatively pristine and free
from man-induced noise disturbance,
with the goal to compare the "normal"
behavior of those whales to the
"normal" behavior of the Western Arctic
stock. The latter stock has been
exposed to human activities in both the
Canadian and U.S. Beaufort Seas for
over 10 years. We may be able to
determine if cumulative human and
disturbance effects are evident in the

Western Arctic bowhead stock.
Future Needs and Applications
The MMS is presently preparing to
perform a study on the "Effects of
Production Activities on Bowhead
Whales" in the Chukchi and Beaufort
Seas to determine if bowhead whales
will be affected by noise in their
spring migration along the Chukchi
coast (near Point Barrow) and into the
Beaufort Sea. This study will help us
to find another answer to the long list
of acoustic-stimuli questions associ-
ated with the spectrum of oil and gas
operations. These studies usually
require more than 1 year to arrive at
satisfactory findings.
Even with the completion of this
study, there are other noise-effects
studies that may be required in the
future. One of the compelling reasons
for this lack of specificity is that
many results are not derived from
easily controlled experiments. The MMS
generally is working on wild, protected
species in a harsh environment--not
capturing and lab testing large marine
mammals or other protected species.
Instead, reliance on field observations
by trained experts who also can measure
acoustic parameters is required. Team
approaches will continue to be used to
establish the degree of relationship
between offshore operations and
wildlife .
Conclusion
Over the past several years, MMS
has commissioned an array of stuuies
dealing with noise disturbance.
Results from many of these studies have
influenced important decision processes
to protect marine mammal and bird
populations while simultaneously
fostering orderly oil and gas resource
development.
Literature Cited
Baker, C.S., L.M. Herman, B.G. Bays,
and W.F. Stifel. 1982. The Impact of
Vessel Traffic on the Behavior of
Humpback Whales in Southeast Alaska.
Unpublished report prepared by Kewalo
Basin Marine Mammals Lab. , Honolulu,
HI, for USDOC, NOAA, National Marine
Mammals Laboratory, Seattle, WA. 39
PP .
Burns, J.J. and B.P. Kelly. 1983.
Studies of Ringed Seals in the Alaskan
Beaufort Sea During Winter: Impacts of
Seismic Exploration. Annual Report,
Outer Contimental Shelf Environmental
Assessment Region (OCSEAP) Research
Unit (RU) 232.
Cowles, C. J. and J.L. Imm. 1980.
Endangered Species Research: A
Rationale for the Selection of a
Research Strategy. In: Proceedings of
the Interagency ~eetin~ to Review,
Coordinate, and Plan Bowhead Whale
Research, other Cetacean Research, and
Related Research Bearing Upon the Conservation
and Protection of Endangered
Marine Species in Alaska and Elsewhere.
USDOI, Bureau of Land Management. pp.
69-79.
Cummings, W.C., D.V. Holliday, and B.J.
Lee. 1984. Potential Impacts of
Man-Made Noise on Ringed Seals:
Vocalizations and Reactions. Environ-
mental Assessment of the Alaskan
Continental Shelf. Final Report of
Principal Investigators, RU 636.
Tracer Doc. No. T-84-06-008-U.
124 pp.
Federal Register. 1982a. Regulations
Governing Small Takes of Marine
Mammals Incidental to Specified Activi-
ties, 47(42):9027-9030 (March 3, 1982).
Federal Register. 1982b. Regulations
Governing Small Takes of Marine
Mammals Incidental to Specified Activi-
ties, 47(96):21248-21259 (May 18,
1982).
Frost, K.J., J.J. Burns, and L.F.
Lowry. 1985. Distribution, Relative
Abundance, and Potential Displacement
of Ringed Seals in Alaska. Abstract
from the Proceedings of the Sixth
Biennial Conference on the Biology of
Marine Mammals.
Good, A.H. 1983. Letter from A.H.
Good, Associate Solicitor, Energy and
Resources, USDOI, to Dr. Anthony J.
Calio, Deputy Administrator, USDOC,
NOAA; dated May 13, 1983.

Kelly, B.P., L.T. Quakenbush, and J.R.
Rose. 1986. Ringed Seal Winter
Ecology and Effects of Noise Disturb-
ance. Environmental Assessment of the
Alaskan Continental Shelf. Final
Report of Principal Investigators, Part
2, RU 232. 83 pp.
Ljungblad, D.K., B. Wursig, S.L.
Swartz, and J.M. Keene. 1985. Obser-
vations of the Behavior of Bowhead
Whales (Balaena mysticetus) in the
Presence of Operating Seismic Explora-
tion Vessels in the Alaskan Beaufort
Sea. OCS Study MMS 85-0076. Report
prepared by SEACO, Inc. Anchorage, AK:
USDOI, MMS, Alaska OCS Region. 51 pp.
Malrne, C.I., P.R. Miles, C.W. Clark, P.
Tyack, and J.E. Bird. 1983. Investi-
gations of the Potential Effects of
Underwater Noise from Petroleum Indus-
try Activities on Migrating Gray Whale
Behavior. Report No. 5366 prepared by
Bolt, Beranek, and Newman, Inc.,
Cambridge, MA, for USDOI, MMS, Alaska
OCS Region, Anchorage, AK. 134 pp.
Malme, C.I., P.R. Miles, C.W. Clark, P.
Tyack, and J.E. Bird. 1984. Investi-
gation of the Potential Effects of
Underwater Noise from Petroleum Indus-
try Activities on Migrating Gray Whale
Behavior, Phase 11. January 1984
Migration. Report No. 5586 prepared by
Bolt, Beranek, and Newman, Inc., for
USDOI, MMS, Alaska OCS Region,
Anchorage, AK. 185 pp.
Malme, C.I., P.R. Miles, P. Tyack, C.W.
Clark, and J.E. Bird. 1985. Investi-
gation of the Potential Effects of
Underwater Noise from Petroleum Indus-
try Activities on Feeding Humpback
Whale Behavior. Report No. 5851
prepared by Bolt, Beranek, and Newman,
Inc., for USDOI, MMS, Alaska OCS
Region, Anchorage, AK. 100 pp.
Malme, C.I., B. Wursig, J.E. Bird, and
P. Tyack. 1986. Behavioral Responses
of Gray Whales to Industrial Noise:
Feeding Observations and Predictive
Modeling. Environmental Assessment of
the Alaskan Continental Shelf. Final
Report of Principal Investigators, RU
675. BBN Report No. 6265. 164 pp.
McLaren, P.L., C.R. Greene, W.J.
Richardson, and R.A. Davis. 1986.
Bowhead Whales and Under-Water Noise
Near a Drillship Operation in the
Alaskan Beaufort Sea, 1985. Report
prepared by LGL Limited and Greenridge
Sciences for UNOCAL Corp. 137 pp.
Richardson, W.J., ed. 1983. Behavior,
Disturbance Responses, and Distribution
of Bowhead Whales, Balaena mysticetus,
in the Eastern Beaufort Sea. 1982.
Unpublished report prepared by LGL
Ecological Research Associates, Inc.,
Bryan, TX, for USDOI, MMS, Reston, VA.
357 pp.
Richardson, W.J., ed. 1985. Behavior,
Disturbance Responses, and Distribution
of Bowhead Whales, Balaena mysticetus,
in the Eastern Beaufort Sea. 1980-84.
OCS Study MMS 85-0034. Prepared by LGL
Ecological Research Associates, Inc.,
for USDOI, MMS. 306 pp.
Sorensen, P.W., R.J. Medved, M.A.M.
Hyman, and H.E. Winn. 1984. Distribu-
tion and Abundance of Cetaceans in the
Vicinity of Human Activities Along the
Continental Shelf of the Northwestern
Atlantic. Marine Environmental
Research 12:69-81.
Stewart, B.S., F.T. Awbrey, and W.E.
Evans. 1983. Beluga Whale,
Delphinapterus leucas, Responses to
Industrial Noise in Nushagak
Bay, Alaska: 1983. Environmental
Assessment of the Alaskan Continental
Shelf. Final Report of Principal
Investigators, RU 629. Hubbs-Sea World
Research Institute, Technical Report
No. 83-161. 9 pp.
USDOI, FWS, 1982. Section 7 Consulta-
tion (1-1-83-F-1). Studies on Southern
Sea Otter Response to Acoustic Stimuli.
December 14, 1982. 6 pp.
USDOC , NOAA, NMFS. 1980. Endangered
Species Act Section 7 Consultation -
Biological Opinion for Proposed Outer
Continental Shelf Oil and Gas Lease
Sale in the Nearshore Beaufort Sea and
All Resulting Activities (Sale BF).
June 24. 1980. 19 pp.
USDOC, NOAA, NMFS. 1987. Endangered
Species Act Section 7 Consultation and
Biological Opinion for Oil and Gas
Leasing and Exploration - Beaufort Sea
Sale 97. May 20, 1987. 22 pp.

Abstract
UNRESOLVED ASPECTS CONCERNING THE INFLUENCE
OF NOISE ON MARINE MAMMALS
J. M. Terhune
University of New Brunswick, Saint John, New Brunswick, CANADA
The importance and utilization of
underwater sounds in the lives of marine
manuals varies greatly. Some species
are virtually silent and inhabit areas
where sound transmission is possible
over only short distances. Other
species vocalize year round and
inhabit areas where sounds can be
detected at great distances. The
detection of purposefully produced
vocalizations and natural "noises" can,
in theory, be measured or predicted
using current knowledge and technology.
The influence of man-made sounds on the
detection of "natural" sounds can be
determined. While noise exposure models
can be constructed, these models will be
restricted to the detection of sounds.
The myriad of possible consequences of
industrial noises masking natural sounds
anchor introducing frightening acoustic
stimuli is largely unknown. Until the
role of sound in a natural setting is
known for a species, it will be
difficult to predict (or measure) the
in luence of a perturbed situation.
Efforts must be made to link short and
This is a reviewed and edited version of apaperpresented
at the Ninth International Conference on Port and Ocean
Engineering Under Arctic Conditions, Fairbanks, Alaska,
USA, August 17-22, 1987. Q The Geophysical Institute,
University of Alaska, 1987.
long- term life history factors with
noise exposure models. This will
require the development of new
approaches and technologies as well as
utilization of available techniques.
Introduction
There is great variability in the
prevalence and nature of underwater
vocalizations of marine mammals. The
very long range calls of some cetaceans
(Payne and Webb, 1971) contrast markedly
with the echo-location pulses of others.
Presumed communicative vocalizations of
seals exhibit much variability. Weddell
seals (Leptonychotes weddelli) call
throughout the year, over long distances
and have a wide variety of call types
(Thomas and Kuechle, 1982) . The harbor
seal (Phoca vitulina), however, is
almost silent. Fig. 1 illustrates the
variable nature of the vocal behavior of
a few phocids. Using the nature of the
vocal activities as an index, it would
follow that the importance of underwater
comnunication to the species also varies
greatly. To seme extent, this diversity
may reflect the acoustical properties of
the areas which the various species
inhabit. For example, harbor seals
frequent coastal areas which are
characterized by very shallow water,
irregular bottom features, islands,
turbidity, upright vegetation in the

water etc. These features will severely
limit the transmission distance of
underwater sound. During underwater
recording sessions in an estuary, I
often heard the airborne noises of small
fishing boats well before I could detect
them with a hydrophone. Polar seals in
ice-covered waters are not so limited
and long range inter-animal
communication is possible. Weddell
seals have been heard (through
hydrophones) at distances of almost 30
km (Thomas and Kuechle, 1982). Ambient
noises generated by ice, wind and rain
will interfere with the cmunication
channels. In addition, species with
high population concentrations or very
distant calling ranges, may well mask
each other (Terhune and Ronald, 1986).
The evolutionary pressures of these
noises may well have shaped the vocal
communication channels utilized by the
various marine mammals.
Ñ
Weddell
Short bharbor w
Seasonal
CALL TYTnP "ria
Few Many
Fig. 1. Variability of seal vocalization
patterns.
Detection of Sounds
Cmunication is limited by the
sensitivity and decoding abilities of
the receiver. To date, only a few
aspects of marine mama1 hearing have
been examined. These include the
sensitivity to pure tones, critical
ratios (e.g. Moore and Schusterman,
1987), directional hearing (e.g.
Terhune, 1974) and upper frequency
limits (e-g. Terhune and Ronald, 1976).
These, and other studies, have examined
only a few individuals of a few species
and the results must be interpreted in
the "broad brush" sense only. A recent
study (Terhune, unpublished results)
suggests that seals and bottlenosed
dolphins (Tursiops truncatus) process
short duration sounds somewhat
differently (porpoise data from Johnson,
1968). A directional hearing study
suggests that lew frequencies and pure
tones cannot be located accurately
(Terhune, 1974). This implies that many
sounds could not be cross correlated
(cocktail party effect). This will have
direct bearing on the masking influence
of noises. To date, masking studies
have had the test signal and masking
sound originate from the same sound
source. This effectively reduces the
possibility of separating the signal
from the noise by cross-correlation
techniques. Thus, the sets of values
reported in Moore and Schustennan (1987)
may be higher than would be the case if
a high frequency and somewhat irregular
signal (i.e. not a continuous pure tone)
from one source were masked by a noise
from another direction.
Because some vocal marine mama1
species have essentially evolved under
various types of noisy conditions, some
vocalizations may be somewhat pre-
disposed to overcoming noise. The very
long (45-60 sec), frequency modulated
call of the bearded seal (Ray et al.,
1969) and the repetitive, harp seal
calls which increase in loudness and(or)
frequency toward their finish (Watkins
and Schevill, 1979) will be more
detectable under noisy conditions than
short duration calls. The evolution of
very different types of calls also
suggests that acoustical communication
will be utilized for different purposes
in various species. While harp seals
may depend upon their myriad of calls
for locating the reproductive herd
and/or courtship (Terhune and Ronald,

1986), ringed seals may utilize sounds
in association with territorial defense
(Stirling, 1973).
There are many unanswered questions
concerning the capabilities of marine
manmals to detect and recognize various
sounds. Until further information can
be obtained, we can only utilize
extrapolations from other species or
assume that the various systems are
noise limited and thus may be inferred
from an examination of ambient noises.
Sound detection thresholds of baleen
whales clearly must be dealt with in
this latter manner. Many technically
difficult problems could be investigated
and reasonable approximations on the
limits of detectability or recognition
generated. For example, equal loudness
curves would permit the establishment of
"weighting curves" similar to the "A, B,
and C" curves established for humans.
This would be inportant in assessing
perceived levels of sounds. Because
high frequencies are absorbed to a
greater extent than low frequencies, the
spectrum of vocalizations will change
with distance. Thus, Weddell seal
vocalizations of 30 km distance will
sound appreciably different than the
same calls at close range. This is
analogous to the human situation of
hearing someone speaking on the
telephone. In the Weddell seal case,
the researchers could identify the calls
of seals 30 krn away. We do not know at
what distance the sounds would be
detectable, and still recognizable, to
the seals however. The question of
which statistical level to use when
considering the influence of ambient
noises must also be addressed. If the
sound levels can be accurately
described, it should be possible to
consider the various proportions of
noise and quiet periods. In some
instances, the term "ambient signal" may
better describe the ambient "noise" as
the marine mammal may be obtaining some
information relating to navigation,
location of open water etc. by listening
to specific sounds. Although many
experiments examining these and similar
questions would be costly and time
consuming to perform, they are
technically possible and will likely be
completed in due time.
Situation of the Listener
The problem of assessing possible
consequences of industrial noises is
made particularly difficult because it
not only requires information on the
noise but also on the situation of the
listener. In the human case, public
protests concerning airport noises are
ccinmon and have resulted in the
initiation of a number of mitigative
measures specifically aimed at reducing
the absolute noise levels and timing of
the noises. Many humans, however, are
kncwn to actively seek out or
voluntarily tolerate dangerously loud
sounds. Persons attending "rock
concerts" or patronizing a bar that has
entertainment, tolerate noise levels
that are known to cause permanent
hearing loss. The disturbing noises
generated by a late night party may
depend upon whether or not you were
invited!
High frequency acoustic scaring
devices have been employed to protect
fish or crops from mammalian and avian
pests. These devices have generally
been found to be ineffectual against
rodents and rabbits (Wilson and
McKillop, 1986) and harbor seals
(Geiger, 1985) . Although these devices
shew some initial success, the mammals
in question may habituate to the sound,
have different initial tolerance levels
and/or the intensity involved does not
cause unconditioned aural pain (Geiger ,
1985). In the case of using a sound
deterrent device to protect salmon
caught in gill nets, the sound may
frighten the seals and sea lions away
initially but, without further negative
reinforcement, may come to serve as a
dinner bell. Harbor seals that haul out
onto offshore rocks in the Bay of Fundy,
Canada, have habituated to traffic
noises from a road a half kilometer
away. The seals remained on the
haul-out site when large transport
trucks went by but raced to the water
when a researcher inadvertently slmd
a car door. These seals have a regular

haul-out cycle and will tend to return
to the rocks if a disturbance occurs
early in the cycle, but will remain in
the water if they are disturbed late in
the cycle. Thus sounds which will
frighten marine mamnals in the short
term are variable and the context in
which the sound is detected is
important.
Frightening Sounds
Some sounds (and
can generate extreme
Very loud noises may
related activities)
fear in a manmal.
terrorize an animal
and result in long-lasting behavioral
changes. For example, many dogs and
other domestic animals are known to be
frightened of thunderstorms or to become
"gun-shy". I have observed instances in
which harbor seals have been shot at and
subsequently vacated a haul-out ledge
for a month or so. Overall, however,
seals in the Bay of F'undy have cone to
tolerate much human-related disturbance.
During daylight, they are constantly
vigilant when hauled-out , while
(protected) harbor seals in California
are thought to sleep (Terhune, 1985).
Harbor seals return to the Bay of F'undy
each spring following an annual
migration to the United States (where
they are legally protected). In spite
of being purposefully harassed, these
seals remain in the area and don't swim
back to the U.S.A. (which they could do
in a day or two). Over time, through
genetic selection, a marine mammal
population may become more tolerant of
disturbance than their ancestors.
Through more active selection, a similar
situation has occurred with domestic
animals.
Life History Linkage with Noise Exposure
The role of acoustical communication
(both active and passive) in the lives
of marine manmals must be known to
properly utilize a noise exposure model.
The variability of marine mama1
vocalizations suggest that it would be
necessary to construct species specific
noise exposure models. Once the normal
situation is properly determined and the
noise model constructed, the final step
(and probably most difficult) will be to
assess the situation with regard to
various life history parameters. That
is, the receiver end of the noise
exposure model must be calibrated.
Various studies have noted short-term
behavioral responses. Will the noise
exposure be reflected in long term
recruitment, mortality, longevity and/or
distribution of the species? Life
history studies will, more likely,
reflect the influence of a number of
perturbatinq factors simultaneously. It
may not be possible to detect an
influence resulting frcm a single mode
of disturbance. Another unknown is the
cumulative effect of stresses (Geraci
and St. Aubin, 1980 ) . A multi-faceted
increase in human activity in an area
could possibly result in a greater
overall impact than would be predicted
by the sum of all the impacts of
individual factors. The myriad of
caplex interactions may mask all but
the most extreme problems.
Life history studies of all marine
manual species are far from complete.
There is still controversy over the
population estimates of harp seals
(Holt, 1987) even though this is
probably one of the best known of all
marine mama1 populations. Technical
problems of locating the animals and the
wide range of variation associated with
food, ice conditions, weather, age
structures, etc. result in population
assessment with large standard errors.
Marine manmal studies are occasionally
confounded by the secrecy associated
with naval operations and studies. The
routes and noises produced by many
vessels (e.g. nuclear submarines) are
known but not available to the public.
Assessment of a noise disturbance model
would begin with an unknown amount of
disturbance already occurring.
In spite of all of the problems
noted above, it should be possible to
relate sane life history factors to a
noise exposure model. This would
require a long term set of species by
species studies conducted by a stable,
experienced workforce. Until such
studies link biological consequences to

noise exposures, the value of noise
exposure models will be marginal at
best.
References
Geiger, A.C. 1985. Evaluation of seal
harassment devices to protect
salmon in gillnet fisheries.
Abstract. - "Sixth Biennial
Conference on the Biology of
Marine ~ammals,"Vancouver, B.C.
Geraci, J.R., and D.J. St. Aubin. 1980.
Offshore petroleum resource
development and marine mammals: a
review and research recommendations.
Mar. Fish. Rev. 42: 1-12.
Holt, S.J. 1987. Letter to the editor.
Bulletin, Can. Soc. Zool. 18: 9.
Johnson, C.S. 1968. Relation between
absolute threshold and duration-of
tone pulses in the bottlenosed
porpoise. J. Acoust. Soc. Amer.
43: 757-763.
Sterlinq, I. 1973. Vocalization in the
ringed seal (Phcca hispida). J.
Fish. Res. Bd. Can. 30: 1592-1594.
Ray, C., W.A. Watkins and J. J. Burns.
1969. The underwater sonq of
Erignathus (bearded seal ).
Zoologia 54: 79-83.
Terhune, J.M. 1974. Directional hearinq
of a harbor seal in air and water.
J. Acoust. Scc. Amer. 56:
1862-1865.
Terhune, J.M. 1985. Scanning behavior
of harbor seals on haul-out sites.
J. Mammal. 66: 392-395.
Terhune, J.M. and K. Ronald. 1976. The
upper frequency limit of ringed
seal hearing. can. J. Zool. 54:
1226-1229.
Terhune, J.M. and K. Ronald. 1976.
Distant and near range functions
of harp seal underwater calls.
Can. J. Zool. 64: 1065-1070.
Thomas, J.A. and V.B. Kuechle. 1982.
Guantitative analvsis of Weddell
seal (Leptonych;tes weddelli)
underwater vocalizations at
McMurdo Sound, Antarctica. J.
Acoust. Soc. Amer. 72: 1730-1738.
Watkins, W.A. and W.E. Schevill. 1979.
Distinctive characteristics of
underwater calls of the harp seal,
Phoca groenlandica, during the
breeding season. J. Acoust. Soc.
Amer. 66: 983-988.
Wilson, C.J. and I.G. McKillop. 1986.
An acoustinq scaring device tested
against European rabbits. Wildl.
Soc. Bull. 14: 409-411.
Discussion
B. MORRIS: Would you speculate on the
purpose of bearded seal vocalizations?
You stated that bearded seals have only
one type of call that they use year-
round.
J. TERHUNE: Watkins and Burns (Zoologica
54:79-83, 1969) believe that the males
call curing the breeding season to
proclaim territory and (or) breeding
condition. At other times of the year
the seals are apparently solitary. Calls
may then act as a means by which the
seals can space themselves and not
inadvertently cluster in an area. The
calls might also permit the animals to
stay in contact (an acoustic herd) even
though dispersed.
S. TREACY: Do harbor seals echo-locate?
J. TERHUNE: I don't think so. The seals
do not emit trains of clicks which would
enhance the amount of information they
could gather. I don't believe that
emitting a single click would work,
especially on a moving target.

Abstract
EFFECTS OF INDUSTRIAL ACTIVITIES ON RINGED SEALS
IN ALASKA, AS INDICATED BY AERIAL SURVEYS
Kathryn J. Frost
Lloyd F. Lowry
Alaska Department of Fish and Game, Fairbanks, Alaska, USA
The preferred pupping habitat of
ringed seals (Phoca hispida) is the
stable shorefast ice which also provides
a convenient platform for some types of
industrial activity. Concern for the
possible effects of on-ice industrial
activities on ringed seals has resulted
in restrictions on industry and in
research to evaluate the problem.
Aerial surveys were conducted in 1970,
1975-77, and 1981-82, and the data were
used to compare ringed seal abundance in
"industrial" and "control" areas.
Results were equivocal and sometimes
contradictory.
In 1985-87 a major program of
aerial surveys was conducted to monitor
the ringed seal population off Alaska
and to continue investigating possible
effects of industrial activities.
Studies around artificial islands in the
central Beaufort Sea suggest some
displacement of seals within 2 nm
(nautical miles) of the islands.
Comparisons of industrial and control
blocks indicated that seals were more
This is a reviewed and edited version of apaper presen ted
at the Ninth International Conference on Port and Ocean
Engineering Under Arctic Conditions, Fairbanks, Alaska,
USA, August 17-22, 1987.
abundant in the industrial block whether
or not industrial activity had occurred.
Historical data indicate that seal
density in the Beaufort Sea was high in
1975, decreased greatly by 1977, and has
subsequently increased. Although these
trends correlate with levels of
industrial activity (high in the late
1970's and early 1980ts, then decreasing
greatly in 1985-87), the changes
occurred in areas both with and without
activity and are, therefore, probably
due to some other cause. Other studies
in addition to aerial surveys are needed
in order to understand ringed seals and
how they may be affected by human
activities.
Introduction and Background
Ringed seals (Phoca hispida) are
the most abundant marine mammals found
in seasonally ice-covered waters of
northern Alaska. These seals are an
important subsistence species for
coastal residents of northern Alaska and
are a major ecological component of the
arctic and subarctic marine fauna. They
prey on small fishes and crustaceans
(Lowry et al. 1980) and are the major
prey of polar bears (Ursus maritimus)
(Smith 1980). Ringed seals compete for
food with other pinnipeds, as well as
seabirds, arctic cod (Boreogadus saida),

and bowhead whales (Balaena mysticetus)
(Lowry et al. 1978-t and Lowry
1984).
Ringed seals normally spend winter
and spring on and under extensive
unbroken shorefast ice. They maintain
breathing holes through the shorefast
ice, and in spring bear their young in
subnivean lairs on top of the ice (Smith
and Stirling 1975). The shorefast ice
also provides a convenient platform on
which various aspects of petroleum
development can be conducted, including
construction and maintenance of winter
ice roads and airstrips, and seismic
exploration. Areas most suitable for
industrial activity may also support
relatively high densities of ringed
seals.
In June 1970, Burns and Harbo
(1972) conducted the first extensive
aerial surveys of ringed seals in
shorefast ice areas of the Chukchi and
Beaufort seas. The principal objectives
of the research were to develop survey
techniques and to gather baseline
information on ringed seal distribution
and density. However, since seismic
exploratory activities were ongoing in
the study area, an attempt was made to
determine whether the surveys could
detect any effect of seismic activities
on seal distribution. (Profiling was
conducted using dynamite charges -
maximum charge of 50 pounds - buried a
minimum of 1 foot per pound, with a
minimum burial of 20 feet to generate
seismic waves which were used to
determine subsurface geological
profiles. In waters deeper than 3
fathoms, activities were terminated on
15 March.) Locations of seismic lines
were plotted and "undisturbed" and
"disturbed" areas in the central
Beaufort Sea were chosen for analysis.
Comparisons based on surveys flown on
9 June showed a slightly higher density
in the disturbed area while the reverse
was true on 13 June. Based on their
data, the authors concluded that seismic
operations such as were being conducted
under state regulations had not
appreciably displaced ringed seals.
From 1971 to 1974, ringed seal
surveys were not conducted. During that
time development of alternative seismic
energy sources other than explosives was
encouraged. Surveys conducted using
air-guns and vibroseis equipment were
allowed to operate in water deeper than
3 fathoms after the 15 March cutoff
date.
Extensive aerial surveys were again
conducted in June of 1975, 1976, and
1977, principally to investigate the
possible magnitude of annual
fluctuations in ringed seal abundance
along the Beaufort Sea coast (Burns and
Eley 1978). Specific tests of the
effects of on-ice human activities were
not included in the survey design since
the objective was an extensive,
broad-scale assessment of abundance.
However, a substantial increase in
on-ice seismic activity was evident both
to permitting agencies and to ringed
seal survey personnel, and a request was
therefore made to use these data to
compare seal densities in areas with and
without extensive seismic survey
activity. The comparisons consistently
showed a lower density of seals in
"seismic areas" than in adjacent
"controls." The magnitude of this
difference ranged from 227 to 88% with
an average difference of 51% for the 3
years (Burns and Kelly 1982).
Therefore, the best available data
indicated that displacement was
occurring, and beginning in 1979 a
cutoff date of 20 March was imposed on
operations in water deeper than 3
fathoms. The cutoff date was intended
to avoid disturbance of ringed seals
during the primary pupping period.
However, it had a severe impact on
industry by restricting the potential
duration of their operations and
eliminating the optimum working period
in terms of daylight, weather, and ice
conditions. Therefore, in 1981 a
program began as part of the Outer
Continental Shelf Environmental
Assessment Program to clarify and
quantify the possible impacts of on-ice
seismic exploration on ringed seals.
Intensive aerial surveys were one
component of that program.
From 2 to 9 June 1981, 12 aerial
survey flights were conducted in the
Beaufort Sea with emphasis on areas of
intense seismic activity (Burns et al.
1981). On 3 days, surveys were flown

directly along seismic shot lines and on
"control" lines which were parallel to
and midway between the seismic lines.
On 2 days, seal density was higher on
the control lines, while on the third,
density was higher on the seismic lines.
Comparisons similar to those previously
done for 1975-77 data were also made
between 2 sets of seismic and adjacent
control blocks. In both cases, seal
densities were virtually identical in
the seismic and control blocks.
From 26 May to 4 June 1982, aerial
surveys were again flown along seismic
lines and on control lines midway
between the seismic lines (Burns and
Kelly 1982). Statistical comparisons
were made of the density of seals on
seismic and control lines for 8 flights
made on 7 days. A significant
difference was found in only 1
comparison, and in that case seal
density was higher on the seismic
transects than on the controls. There
were no significant differences when
data for all flights were combined.
While the results were sometimes
equivocal or even contradictory, these
studies, in aggregate, indicated that
on-ice seismic activity of the type and
intensity conducted at that time did not
result in large-scale displacement of
ringed seals in the central Beaufort
Sea. However, the fact remained that
ringed seals are abundant and
ecologically important along the Chukchi
and Beaufort sea coasts, and that their
preferred pupping habitat, the shorefast
ice, also provides a convenient platform
for industrial activities. There was a
clear need to develop accurate and
repeatable techniques for assessing
ringed seal abundance, as well as to
determine what factors influence ringed
seal distribution. Therefore, beginning
in 1985, the Minerals Management Service
through the National Oceanic and
Atmospheric Administration Outer
Continental Shelf Environmental
Assessment Program funded a 3-year study
to monitor the ringed seal population
off Alaska and to continue investigating
the possible effects of industrial
activities on ringed seals. The results
of that study are presented in this
paper.
Methods
Aerial survey design
In order to gather the type of
baseline data needed for a monitoring
program, we chose to conduct an
extensive survey covering all of the
shorefast ice between Kotzebue Sound and
Barter Island. The study area was
divided into 11 sample units that
corresponded to sectors used in previous
surveys (Burns and Harbo 1972; ADF&G
unpubl.). Sector boundaries were marked
by identifiable landmarks such as capes,
points, villages, or radar installations
(Figure 1). Surveys were conducted over
the shorefast ice between 20 May and
16 June, beginning in Kotzebue Sound and
proceeding north and east to Barter
Island. The Chukchi Sea was surveyed
from 20 to 31 May, and the Beaufort Sea
from 27 May to 16 June, to coincide with
ice conditions which were optimal for
sighting seals and with the peak period
of seal haul-out.
The surveys were flown between 1000
and 1600 hours true local time, the time
of day when maximum numbers of seals are
known to haul out (Burns and Harbo 1972;
Smith and Hammill 1981). The obser-
vation platform was a Twin Otter
aircraft equipped with oversize bubble
windows, radar altimeter, and Omega-GNS
500 Global Navigation System, flown at a
true ground speed of 110-130 knots.
Survey altitude was usually 500 ft in
the Chukchi Sea and 300 ft in the
Beaufort Sea.
Strip width varied according to
altitude and was determined by
pre-calculated inclinometer angles which
were marked on the windows. At 500 ft,
the transects began 0.125 nm out from
the centerline and extended out to
0.5 nm for a net width of 0.375 nm
(2,250 ft). At 300 ft, the inclinometer
angles remained the same and the
effective strip width was reduced to
0.225 nm (1,350 ft). Two observers sat
on either side of the aircraft just
forward of the wings. A third person
recorded beginning and ending points of
transects, ice conditions, and weather.
Each observer counted the seals in the
strip on his or her side of the aircraft
and also made note of industrial

CHUKCHI SEA
ALASKA
BEAUFORT SEA t
Kolfbuf Sound FAST ICE / PACK ICE
EDGE
activity such as artificial islands, ice
roads, seismic lines, and airstrips.
All data were recorded by 1-minute
intervals.
The survey was flown according to a
stratified random strip transact design,
with transects spaced 2 nm between
centerlines. Transects were flown along
lines of latitude in the Chukchi Sea and
lines of longitude in the Beaufort Sea,
from shore to the approximate seaward
edge of the shorefast ice. Within each
sector, 60% of the possible transects
were randomly selected and flown; in
some sectors all or some of the selected
lines were replicated for comparative
purposes.
Data analysis
Counts of seals at cracks and at
holes were added separately for each
1-minute interval. The lengths of
transects were calculated from beginning
and ending GNS positions and divided by
total elapsed time to obtain ground
speed. The area surveyed per minute
interval was calculated by multiplying
speed by strip width. Each minute
interval, therefore, was assigned data
on latitude and longitude (of the
beginning point), area (nm 2 ), local
Figure 1. Map of northern
Alaska showing sectors
used for design and
analysis of ringed seal
aerial surveys, and an
example of lines flown
during a survey.
time, counts of seals at holes and at
cracks, and ice and weather conditions.
Densities were calculated by
dividing the number of seals counted by
the area surveyed (Cochran 1977).
Variance of the density was calculated
by using the model unbiased o-stimator
(Cochran 1977, formula 6.27), modified
to account for total sampling area
(Estes and Gilbert 1978).
The possible effects of industrial
activity were examined by comparing
densities of seals in areas with and
without activity such as ice roads,
seismic trails, or artificial islands.
The shortest straight-line distances
from artificial islands to each minute
sighting block were determined by
comparing positions for each interval to
position for the islands. Densities
were then calculated for 2-nm concentric
circles centered at the artificial
islands, out to a distance of 10 nm.
Since the islands were less than 10 nm
apart and interactive effects were
possible, a density in relation to all
islands was also calculated using the
minimum distance from any of the 3
islands for each 1-minute sighting
block.

Densities were calculated for an Table 1. The density of ringed seals at
industrial and 2 adjacent control holes in relation to distance from 3
blocks. All sightings within 10 nm of artificial islands in the Beaufort Sea,
land were used in this comparison. June 1985-1987.
"Industrial" blocks were areas which
included artificial islands, ice roads,
and seismic trails. "Control" blocks of
1985
similar size were delineated to the east Distance (am)
and west of the industrial area. Island Survey 0-2 2-4 4-6 6-8 8-10
Results and Discussion
Data were obtained for 3 artificial
islands : Seal, Northstar, and
Sandpiper, for all 3 years of the survey
(Table 1). In 1985, all 3 of the
islands were active: Seal was engaged
in drilling operations and Northstar and
Sandpiper were under construction.
During the 1985 surveys, the central
Beaufort Sea adjacent to the artificial
islands was surveyed twice, on 7 and
11 June. Analysis of density with
distance from the islands indicated that
for all comparisons the density of seals
at holes was 20%-80% lower within 2 nm
of the islands than it was 2-4 nm away.
During the 1986 surveys Seal Island
was inactive and had been so all winter:
Northstar was inactive at the time of
survey but had been in operation through
late spring; and Sandpiper was currently
active. The area was surveyed before
break-up on 6 June, and after break-up
had commenced on 13-16 June. Unlike
1985, there was no clear trend in
density with distance from the islands
for either survey; results for
individual islands were contradictory.
Near Northstar (active until April) the
density for both surveys was slightly
lower (3%-15%) within 2 nm of the island
than between 2-4 nm. Near Sandpiper the
density was higher within 2 nm of the
island on one survey, and lower on the
other.
During winter and spring of
1986-87, all 3 artificial islands were
inactive. As in previous years, the
islands were surveyed twice in 1987, on
6 and 11 June. There was no consistent
trend in seal density with distance from
the 3 non-operational islands. Seals
were more numerous near Seal Island,
less numerous near Northstar, and
differed between the 2 surveys at
Sandpiper.
Seal 85-1 0.7 1.2 1.1 1.7 1.3
85-2 - 1.9 1.0 3.3 2.2
Northstar 85-1 0.8 1.6 2.2 1.4 0.9
85-2 0.8 1.0 5.8 1.5 1.5
Sandpiper 85-1 0.6 3.1 1.0 1.0 1.1
85-2 2.6 4.4 1.8 1.9 1.6
1986
Distance (nm)
Island Survey 0-2 2-4 4-6 6-8 8-10
Seal 86-1 6.1 5.8 4.6 2.3 5.1
86-2 - 4.6 6.5 5.0 5.6
Northstar 86-1 5.0 5.2 6.8 4.2 2.1
86-2 5.0 5.9 5.7 8.8 5.3
Sandpiper 86-1 8.3 3.3 6.5 3.2 3.6
86-2 5.2 6.2 6.8 9.1 9.1
1987
Distance (nm)
Island Survey 0-2 2-4 4-6 6-8 8-10
Seal 87-1 - 1.1 2.9 2.7 5.5
87-2 14.4 9.5 10.4 5.9 4.8
Northstar 87-1 1.1 3.3 5.6 4.1 5.2
87-2 3.8 8.4 14.2 6.3 6.1
Sandpiper 87-1 7.1 7.6 2.2 4.2 3.9
87-2 6.8 5.5 6.6 5.2 11.9
Interpretation of the data
regarding trends in density around
individual islands was complicated and
the utility of such data limited by
several factors: sample sizes were
small (17-80 nm 2 total per survey),
particularly within 2 nm of the islands
where the sample for a survey usually
consisted of 1-3 minutes (1-6 nm 2 ) of
data; the islands were close enough
together (particularly Seal and

Northstar islands which were only 4 nm
apart) for interactive effects to occur;
and not all islands were in similar
operational status either within or
between years. Consequently, the data
set shown in Table 1 could not be
treated as 18 replicate tests of the
effect of an artificial island on seal
density.
To address the first two of these
problems we determined the minimum
distance from any island in the data set
from each survey (Table 2). In 5 of the
6 comparisons, the density of seals at
holes was 12%-72% lower within 2 nm of
any island than it was 2-4 nm away.
Inspection of the raw data indicated
that for the single exception (survey
86-1) the higher density at 0-2 nm was
probably an artifact of the way position
was assigned to the minute survey
interval. Although the density of seals
was lower near the islands in both 1985
when all islands were active and 1987
when none were active, the magnitude of
the difference was much greater during
activity (50%-70%) than in its absence
(12%-30%).
Table 2. The density of ringed seals at
holes in relation to distance from any
of 3 artificial islands in the Beaufort
Sea, June 1985-1987.
Distance from any island (nm)
Survey nm 2 0-2 2-4 4-6 6-8 8-10
A block comparison of industrial
and adjacent control areas was also done
for all 3 years. In 1985, industrial
activity, including seismic lines, ice
roads, and islands, was widespread,
resulting in an industrial block
approximately 60 nm across. In 1986,
the only obvious activities were the
artificial islands and associated ice
roads, resulting in an industrial block
which was only 16 nm across (Figure 2).
During 1987 surveys there was no
offshore industrial activity; however,
data were analyzed according to the 1986
industrial and control blocks for
comparative purposes.
70 45 "
70 15 .,
150 148
WEST CONTROL
r----- -- .--.
BEAUFORT SEA
EAST CONTROL
Figure 2. Map showing locations of
artificial islands in sector B3 of the
Beaufort Sea and the 1986 industrial and
control blocks.
In both 1985 and 1986 the density
of total seals was significantly higher
in the industrial block than in the
control blocks (Figure 3). In 1987, in
the absence of any offshore industrial
activity, density in the "industrial"
block was also higher than either
control, suggesting that some
characteristics other than the presence
or absence of activity were responsible
for the difference.
Aerial surveys of ringed seals in
1985-1987 were the most extensive and
systematic ever conducted in Alaska, and
the first for which between-year
statistical comparisons were possible.
Data from those years demonstrated
substantial year-to-year variability in
ringed seal densities (Table 3).

0 WEST CONTROL
0 EASTCONTROL
Figure 3. Density of ringed seals
(total seals/nm 2 ) in industrial and
control blocks in the central
Beaufort Sea, June 1985-1987.
Table 3. Comparison of ringed seal
densities (total seals/nm 2 ) on the
shorefast ice of the Chukchi and
Beaufort seas based on surveys conducted
in 1985-1987. Data from the Chukchi Sea
in 1987 are not yet analyzed.
Density
Sector 1985 1986 1987*
* Preliminary data.
Between 1985 and 1986, observed density
of total seals hauled out on the Chukchi
Sea shorefast ice increased 607 from 2.9
to 4.7 seals/nm 2 . Increases in
individual sectors ranged from 307-907..
In the Beaufort Sea, the overall
increase was 12%, from 3.0 to 3.3
seals/nm 2 , with the westernmost sector
near Barrow decreasing 77 and the
central sectors increasing 20%-30%. The
causes for such inter-annual variation
are unknown. While relationships
between seal abundance and physical
parameters such as ice deformation and
extent of shorefast ice do exist and may
explain small-scale differences in the
distribution and abundance of seals
(Frost et al. 1985, 1987), they cannot
account for the large observed inter-
annual differences . We have no measure
of biological parameters such as prey
availability, which may be a major
factor in determining overall ringed
seal distribution and abundance in a
given year.
Historical data also indicate
substantial year-to-year variability in
the occupancy of nearshore areas by
ringed seals. Data are available for
the Alaskan Beaufort Sea since 1970
(Burns and Harbo 1972; Burns and Eley
1978; Burns et al. 1981; Burns and Kelly
1982; Frost et al. 1985). During that
period, the density of ringed seals on
the shorefast ice of the Beaufort Sea,
as a whole, dropped from a high of 3.3
seals/nm 2 in 1975, to a low of 1.1
seals/nm 2 in 1977, and subsequently
increased steadily to 3.3 seals/nm2 by
1986. The density in any particular
year ranged from 507 below to 40% above
the mean density for 8 years of surveys.
In the Canadian Beaufort Sea near
Tuktoyaktuk, ringed seal densities have
fluctuated from 557 above to 70% below
the long-term mean in a far less regular
manner than the Alaskan Beauf ort Sea
(Stirling et al. 1981; Kingsley 1986).
Such annual and long-term
variability demonstrate the need for
regular and relatively extensive
coverage of areas in which smaller-scale
comparisons are being made. For
example, the density of ringed seals in
the central Beaufort Sea (sectors B2 and
B3) decreased in the mid- to late 1970's
and subsequently increased in the
mid-1980's (Figure 4). This could be
attributed to changes in industrial
activity, which intensified in the late
1970's and early 19801s, then gradually
decreased. However, the western
Beaufort Sea (sector Bl), which
experienced little or no seismic or
other industry activity, showed the same
fluctuations in density during this time
period. Furthermore, the major decline

in density which occurred in the study
area between 1975 and 1977 also occurred
in the Canadian Beaufort Sea (Stirling
et al. 1981).
00 J
70 72 74 76 78 80 82 34 86
YEAR
Figure 4. Density of ringed seals in 3
sectors of the Beaufort Sea based on
aerial surveys conducted in 1970-1986.
(Data from Burns and Harbo 1972; Burns
and Eley 1978; Burns et al. 1981; Burns
and Kelly 1982; Frost et al. 1986.)
While aerial surveys are useful in
monitoring long-term trends in abundance
over large areas, they are not well-
suited to detecting small-scale
differences in geographically restricted
areas. In this study, aerial survey
data indicated a possible local effect
of artificial islands on the density of
ringed seals. However, interpretation
was complicated by the fact that the
minimum sighting unit was 1 minute or
2 nm; land and the edge of shorefast
ice, which may both affect seal
densities, were variable distances from
the 3 islands; and the precision of
navigational equipment sometimes varied
by  1 nm. In analyses of industrial
and control blocks, the greatest
difficulties were in obtaining an
accurate measure of industrial activity
and in designating comparable control
blocks. There is considerable east-west
variability in the Beaufort Sea in ice
topography, extent of shorefast ice, and
bathymetry. Control and industrial
blocks were not necessarily comparable
simply because they were adjacent, as is
indicated by higher densities in the
"industrial" blocks with or without
industrial activity.
In aggregate, analyses of histori-
cal and recent aerial survey data
emphasize the importance of matching
research technique to the question at
hand. Our data indicate that in 1985-86
there were no apparent broad-scale
effects of industrial activity on the
density of ringed seals as measured by
aerial surveys. The data do not
discount local effects which would be
more appropriately detected by other
techniques, nor do they discount the
possibility that regional effects could
occur at different levels of industrial
activity. Most aerial surveys conducted
during peak years of industrial activity
in the central Beaufort Sea did not have
sampling effort or design suitable for
statistical analyses of differences
between relatively small areas. By
conducting on-ice studies, Burns and
Kelly (1982) found that although aerial
surveys showed no significant difference
in densities along seismic and control
lines, the rate of alteration or
refreezing of lairs and breathing holes
within 150 m of seismic lines was
approximately double the rate at
distances greater than 150 m. Kelly et
al. (1986) also reported results of
on-ice studies which indicated that
ringed seals do respond to disturbance.
Conclusions
1. Based on aerial surveys conducted
in 1985-87, the density of ringed seals
was lower within 2 nm of artificial
islands in the central Beaufort Sea than
it was 2-4 nm away. Although this was
true in years with and without activity
on the islands, the difference was
greatest when they were all active.
Interpretation of data was complicated
by possible interactive effects of
distance from the islands, from land,
and from the shorefast ice edge; the
accuracy of navigational equipment; and
other biological or physical parameters
that could affect densities but which
were unknown to us.
2. Comparisons of industrial and
adjacent control blocks in 1985 and 1986
indicated that the density of ringed
seals was significantly higher in the
industrial blocks in both years. A
similar comparison of the same blocks in
1987 when there was no industrial

activity also showed that seals were
more numerous in the industrial block,
which indicates that factors other than
the presence or absence of industrial.
activity caused the difference.
3. There was a steady increase in the
density of ringed seals in the central
Beaufort Sea from 1985-87 which occurred
concurrently with a decrease in
industrial activity. In our opinion,
the two are probably not related, since
a similar increase in density occurred
in the western Beaufort Sea where no
industrial activity took place.
4. While our studies do not show any
broad-scale effect of industrial
activities on ringed seal abundance and
distribution in the Beaufort Sea, that
does not imply that such effects cannot
occur. Our data relate only to the
types and levels of activities that
occurred during the study period.
Continued monitoring of ringed seal
populations is warranted. Studies of
ringed seal ecology are needed in order
to explain the causes of the "natural"
fluctuations in density that have been
documented.
5. Aerial surveys are a useful means
of detecting long-term trends in ringed
seal abundance and comparing regional
trends in abundance over large areas.
However, aerial surveys alone are not
adequate for determining the effects of
industrial activity. Such studies
should combine aerial surveys with
on-ice studies that monitor the use and
fate of breathing holes and lairs on a
finer scale.
Acknowledgements
This study was funded by the
Minerals Management Service, Department
of the Interior, through an Interagency
Agreement with the National Oceanic and
Atmospheric Administration, Department
of Commerce, as part of the Alaska Outer
Continental Shelf Environmental
Assessment Program, contract numbers
02-5-022-53, 03-5-022-69 (RU #232),
NA-81-RAC-00045, and 84-ABC-00210.
The authors especially want to
acknowledge the logistic support
provided by the NOAA Office of Aircraft
Operations in the form of the NOAA Twin
Otter and its excellent flight crews,
particularly Commander Dan Eilers who
served not only as Chief Pilot all 3
years, but as a valuable member of the
scientific team. Thanks go to Jim
Gilbert, Howard Golden, Sue Hills, Dawn
Hughes, and Val Uhler for their many
hours as observers, recorders, and idea
generators; to George Lapiene for taking
care of sometimes complicated logistics
planning; and to Jesse Venable for
helping to translate hours of survey
data into analyzed results. Special
thanks to John Burns who, for better or
worse, got us started surveying ringed
seals and laid the ground work for this
project over 15 years ago.
References
Burns, J. J. and T. J. Eley. 1978. The
natural history and ecology of the
bearded seal (Erignathus barbatus)
and the ringed seal (Phoca hispida).
Paces 99-162 in Environmental Assessment
ofthe ~laskan~ontinental Shelf, Annual
Reports, Vol. 1, Outer Continental
Shelf Environmental Assessment Program,
Boulder, CO.
Burns, J. J. and S. J. Harbo, Jr. 1972.
An aerial census of ringed seals,
northern coast of Alaska. Arctic
25:279-290.
Burns, J. J. and B. P. Kelly. 1982.
Studies of ringed seals in the Alaskan
Beaufort Sea during winter: impacts of
seismic exploration. Annual report
RU #232 to Outer Continental Shelf
Environmental Assessment Program,
Juneau, AK. 57 pp.
Burns, J. J., L. F. Lowry, and K. J.
Frost. 1981. Trophic relationships,
habitat use, and winter ecology of
ice-inhabiting phocid seals and
functionally related marine mammals
in the Arctic. Annual report RU #232 to
Outer Continental Shelf Environmental
Assessment Program, Juneau, AK. 81 pp.
Cochran, W. G. 1977. Sampling
techniques, John Wiley & Sons, Inc.,
NY. 428pp.
Estes, J. A. and J. R. Gilbert. 1978.
Evaluation of an aerial survey of

Pacific walruses (Odobenus rosmarus
divergens). J. Fish. Res. Bd. Can.
35:1130-1140.
Frost, K. J. and L. F. Lowry. 1984.
Trophic relationships of vertebrate
consumers in the Alaskan Beaufort
Sea. Pages 381-401 & P. W. Barnes,
D. M. Schell, and E. Reimnitz, eds.
The Alaskan Beaufort Sea - Ecosystems
and Environments. Academic Press,
NY.
Frost, K. J., L. F. Lowry, and J. J.
Burns. 1985. Ringed seal monitoring:
relationships o f distribution,
abundance, and reproductive success to
habitat attributes and industrial
activities. Interim report 1985 -
RU #667 to Outer Continental Shelf
Environmental Assessment Program,
Juneau, AK. 85 pp.
Frost, K. J., L. F. Lowry, and J. R.
Gilbert. 1987. Ringed seal monitoring:
relationships o f distribution,
abundance, and reproductive success to
habitat attributes and industrial
activities. Interim report 1986 -
RU #667 to Outer Continental Shelf
Environmental Assessment Program,
Juneau, AK. 53 pp.
Kelly, B. P., L. T. Quakenbush, and J.
R. Rose. 1986. Ringed seal winter
ecology and effects of noise
disturbance. Final report (Part 2)
RU #232 to Outer Continental Shelf
Environmental Assessment Program,
Juneau, AK. 83 pp.
Kingsley, M. C. S. 1986. Distribution
and abundance of seals in the Beaufort
Sea, Amundsen Gulf, and Prince Albert
Sound, 1984. Environmental Studies
Revolving Funds Report No. 025. 16 pp.
Lowry, L. F., K. J. Frost, and J. J.
Burns. 1978. Food of ringed seals and
bowhead whales near Point Barrow,
Alaska. Can. Field-Nat. 92:67-70.
Lowry, L. F., K. J. Frost, and J. J.
Burns. 1980. Variability in the diet
of ringed seals, Phoca hispida, in
Alaska. Can. J. Fish. Aquat. Sci.
37:2254-2261.
Smith, T. G. 1980. Polar bear
predation of ringed and bearded seals in
the land-fast sea ice habitat. Can. J.
2001. 58:2201-2209.
Smith T. G. and M. 0. Hammill. 1981.
Ecology of the ringed seal, Phoca
hispida, in its fast ice breeding
habitat. Can. J. 2001. 59:966-981.
Smith, T. G. and I. Stirling. 1975.
The breeding habitat of the ringed seal
(Phoca hispida). The birth lair and
associated structures. Can. J. Zool.
53:1297-1305.
Stirling, I., M. C. S. Kingsley, and W.
Calvert. 1981. Seals in the Beaufort
Sea 1974-1979. Report prepared for Dome
Petroleum Limited, Esso Resources Canada
Limited, and the Department of Indian
and Northern Affairs. Can. Wildl.
Serv., Edmonton, Alberta. 70 pp.
Discussion
B. KELLEY: Given the great annual
variability in ice and snow conditions
that appear to affect the proportion
of basking seals, how useful do you
feel aerial survey data are for
examining year-to-year changes in seal
density?
K. FROST: Although annual variability
in ice and snow conditions can be
considerable, that variability is
generally apparent to survey
observers. As long as surveys are
conducted not only during the same
general time period, but also under
similar ice conditions (e.g., before
extensive snow melt and cracking of
the fast ice occur), year-to-year
comparisons should be useful.
However, relatively large geographic
areas should be included in
comparisons in order to avoid apparent
differences caused by small-scale
local variability. Data should be
screened to ensure that surveys have
not been conducted under post-breakup
conditions, which result in extremely
high densities. The best indicators
of these conditions are a large
percentage (greater than 30%) of seals
at cracks and a high incidence of
groups of more than 2 seals at holes.
Within a year, surveys of some part of

the study area should be replicated
several times at 3 to 5 day intervals
to ensure that surveys are conducted
under suitable conditions and before
haul-out patterns have begun to change
with the onset of breakup.
T. NEWBURY: You described data on the
number of seals around drill sites,
and you implied that there might be a
direct effect on these seals due to
activity. Won't there also be an
indirect effect due to the influence
of activity on seal predators,
primarily polar bears?
K. FROST: Indirect effects on predators
are possible, but it is difficult to
predict what they might be. Any effects
would depend on the characteristics of
the activities and the specifics of
predator and prey populations in the area
(including density, age, and sex
composition, or other factors). In 300
hours of surveys over the fast and pack
ice of northern Alaska, no polar bear
kill sites were seen in the very
nearshore region where drill sites were
located.
V. R. NERALLA: Did you conduct any
studies on the appearance of ringed seals
and environmental conditions (e.g., air
temperature, wind speed and wind
direction, etc.)?
K. FROST: Most investigators who have
surveyed ringed seals have considered the
effects of environmental conditions. The
survey protocol which specified that
surveys be conducted between 10 a.m. and
4 p.m. (sun time) and when wind speed was
less than 20 knots incorporated what this
and other studies have found about ringed
seal behavior relative to weather. In
order for a study to specifically test
the effects of weather, surveys would
have to be conducted under poor as well
as good weather conditions.
Data obtained in our study are presented
in Frost et al. 1985 and 1987. It should
be noted that it is difficult to make
specific correlations between
environmental conditions and seal haul-
out patterns for two major reasons: ( 1)
environmental variables are interactive
and do not always produce the same
results (for example, strong winds on
bright, sunny days may not inhibit haul-
out, whereas more moderate winds on cold
days may greatly reduce the number of
seals hauling out); and, (2) it is
extremely difficult to obtain on-ice
measurements of temperature and wind
speed or direction during surveys.
Conditions at shore-based weather
stations or those at survey altitude may
be considerably different than conditions
over the ice at ground level.

RESPONSES OF RINGED SEALS (Phoca hispida) TO NOISE DISTURBANCE
Abstract
Brendan P. Kelly
Institute of Marine Science, University of Alaska Fairbanks, Alaska, USA
John J. Burns
Living Resources Inc., Fairbanks, Alaska, USA
Lori T. Quakenbush
Institute of Marine Science, University of Alaska Fairbanks, Alaska, USA
The effects of on-ice industrial noises
on ringed seals (Phoca hispida) were
investigated to determine the extent to
which such disturbance increases the rates
at which seals abandon breathing holes and
lairs. In the spring of 1982, breathing holes
and lairs were abandoned three times as
often within 150 m of recent seismic survey
lines as were structures at greater distances
from the same lines. Subnivean structures
were abandoned at equal rates within and
beyond 150 m of control lines. Aerial
surveys of ringed seals conducted in the
Beaufort Sea in 1981 and 1982, however,
showed no consistent differences in the
density of basking seals in transects
centered over seismic survey lines and in
intervening transects.
The rate of abandonment of subnivean
seal structures was compared over six years.
In undisturbed areas, the abandonment rate
was 4.0% in shore-fast ice and 12.9% in
drifting ice. Among seal structures
subjected to industrial noise in the shore-fast
ice, the rate was 13.5%, and with the
addition of repeated examinations of
structures by investigators the rate was
32.5%.
This is a reviewed and edited version of a paper presented
at the Ninth International Conference on Port and Ocean
Engineering Under Arctic Conditions, Fairbanks, Alaska,
USA, August 17-22, 1987. @ The Geophysical Institute,
University of Alaska, 1987.
Radio-tagged seals departed their lairs
in response to snow machines within 2.8 km,
human footfalls as far away as 600 m, a skier
as far away as 400 m, and in response to a
helicopter flying 5 km from the lair at an
altitude of 152 m, and during helicopter
landings or takeoffs as far away as 3 km.
Ringed seals abandon breathing holes
and lairs in response to naturally occurring
conditions such as minimal snow cover,
shifting ice, and the activities of predators.
They abandon those sites at higher rates in
response to anthropogenic noises. Seals
would be most adversely affected by noise
disturbance in late March through June
when the amount of time they spend out of
the water is increasing and movements,
especially of females and their dependent
young, are limited to small areas.
Introduction
Potential effects on marine mammals
of anthropogenic noises include physical
harm from extremely loud noises,
interference with vocal communication,
increased levels of stress, and displacement
from local areas (Rausch 1973; Geraci and
St. Aubin 1980; Schusterman and Moore
1980; Norris 1981; Stewart 1981; Ronald and
Dougan 1982; Mansfield 1983; Kelly et al.
1986). Displacement has the most potential
for widespread and long-term effects and has
been a focus of our investigations. Ringed

seals (Phoca hispida) are the most adapted of
northern pinnipeds to inhabiting thick,
relatively stable sea ice, and their ability to
maintain holes through the ice permits them
to occupy areas of complete ice cover year-
round. That adaptation allows ringed seals
to exploit resources from which other
pinnipeds are largely excluded during
winter, but it also makes them more
vulnerable to predation by polar bears
(Ursus maritimus) and arctic foxes (Alopex
lagopus). Mortality of ringed seal pups can
be substantial due to that predation (Smith
1976; Smith 1987; Kelly et al. 1987).
Occupation of areas of extensive ice cover,
especially shore-fast ice, also makes ringed
seals more vulnerable to human activities;
for thousands of years the species was a
major resource for coastal Eskimos and it
remains important in modern Eskimo
culture and economy (Hall 1866; Boas 1888;
Stefansson 1913; Manning 1944; McLaren
1958a; Cox and Spiess 1980; Wenzel 1984;
Smith 1987). In recent times, petroleum
exploration and development activities have
taken place in ringed seal habitat. Gravel
island construction and exploration for oil
using seismic profiling have overlapped
spatially and temporally with ringed seal
whelping and breeding areas. Both gravel
island construction and seismic profiling
involve operating heavy trucks and
bulldozers on the shore-fast ice. Seismic
profiling further entails imparting
substantial amounts of low frequency sound
energy into the oceanic crust (and
incidentally the water column and overlying
ice) and recording the reflected signals.
Concern about the effects of
disturbance in Alaskan waters was first
expressed by subsistence hunters on the
Seward Peninsula. They reported decreased
harvests of ringed seals in an area subjected
to offshore gold exploration in the 1960's
(Burns and Kelly 1982). More recently,
seismic profiling during oil exploration has
presented a greater potential for disturbance
since it involves considerable noise energy
and affects extensive areas. Explosives were
the main signal sources until their offshore
use was banned in 1977 b an
Administrative Order of the A aska
Department of Natural Resources.
Subsequently, "air guns," "water guns," and
Vibroseis machines have been used to
generate source signals.
When we began this study in 1981, the
only data available with which to examine
the hunters' suggestion that ringed seals
r
were displaced by noise disturbances were
the results of aerial surveys conducted
between 1970 and 1977 (Burns and Harbo
1972; Burns and Eley 1978). In those
surveys, lower densities of seals were
observed in the vicinity of coastal
settlements than in adjacent near-shore
areas. They believed that those differences
were greater than could be accounted for by
removal of seals, since hunting of ringed
seals was greatly reduced compared to
earlier times. They speculated that human
activity, especially snow machine travel,
was displacing seals from those areas.
Examination of the aerial data from the
Beaufort Sea for indications of differences in
seal densities inside and outside of areas
affected by seismic exploration has yielded
conflicting results. Burns and Harbo (1972)
reported similar densities of seals in both
areas during the 1970 survey, but Burns et
al. (1981) re-examined the data from 1975-
1977 and concluded that densities of seals in
"seismic" areas were consistently lower than
in undisturbed areas. None of those surveys
were designed to test for indications of
displacement, however, and the
retrospective partitioning of the data into
disturbed and undisturbed areas was
unsatisfactory.
For the present study, we combined onice
surveys of subnivean seal structures
(breathing holes and lairs) using trained
Labrador retrievers, aerial surveys of
basking seals, and radio telemetr to
quantify the reactions of ringed sea s to
seismic profiling that employed the
Vibroseis method and to other
anthropogenic noises. Our objectives were to
(1) determine the effect of seismic profiling
activities on ringed seal distribution, (2)
determine the behavioral responses of
ringed seals occupying lairs to
anthropogenic noise, (3) compare the rates of
abandonment of subnivean structures in
disturbed and undisturbed areas, and (4)
assess the significance of abandonment of
subnivean structures in terms of the
numbers and distribution of alternative
structures available to individual seals. Our
primary measures of disturbance were the
relative densities of basking seals along and
immediately adjacent to seismic "shot lines,"
the rate of abandonment of subnivean
structures as a function of distance from
seismic lines, and changes in haulout
frequency and duration in areas subjected to
seismic profiling. Rates of short-term and
long-term displacement resulting from noise
disturbance were assessed relative to
f

natural rates established using surveys and
telemetric studies conducted between 1981
and 1987.
Methods
Aerial surveys
Aerial surveys were conducted along
the Beaufort Sea coast of Alaska (Figure 1)
between 2 and 9 June 1981 and between 25
May and 4 June 1982. The 1981 surveys
were conducted from a twin engine fixed-
wing aircraft (Grumman Goose) equipped
with a Global Navigation System (GNS). A
Bell 204 helicopter, also with GNS, was used
for the 1982 surveys. Two observers counted
all seals visible within 0.5 nm of each side of
the aircraft while flying at an altitude of
500 ft.
Figure 1. Map of Alaska showing locations mentioned in text.
2 9
Between 2 and 9 June 1981, we
surveyed 2,880 nm of transect lines divided
into three groups; those parallel to the coast
between Smith Bay (70°55'N 154O20W) and
Barter Island (70°08'N 143'40'W), those
centered over seismic lines that had been
surveyed during the previous few months,
and control lines centered between those
seismic line transects. In 1982, we
conducted aerial surveys between 25 May
and 4 June. The total length of those
transects was 1,083 nm, again divided into
those along seismic trails, those along
control transects, and a series duplicating
some of the transects surveyed in June 1981.
Subnivean seal structure surveys
Trained Labrador retrievers were used
to find seal-made structures (subnivean lairs
and breathing holes) in 10 surveys between
so"
50Â

1982 and 1987. The method was similar to
the way Eskimo hunters used sled dogs to
locate seal holes (Hall 1866). Canadian
workers adapted the method for biological
sampling (Smith and Stirling 1975) and one
of us (BPK) learned the method from them
in 1981. The dogs ran in front of a snow
machine at the direction of the dog handler.
When they detected seal odor, they followed
the scent to its source and indicated the
location of the seal structure by digging in
the snow above it. Generally, the dogs were
directed to run perpendicular to the wind
direction to maximize the area of detection.
The search pattern varied de ending
on the objectives of work during the different
field efforts. The dogs generally searched
either along lines established by heavy
equipment and snow machines or at random
within pre-selected areas, usually near field
camps or stations established to monitor
radio-tagged seals. An exception was in a
drift ice survey in 1987 when the dogs
searched along tracks of polar bears.
Seal structure surveys in 1982 were
limited to the shore-fast ice of the Beaufort
Sea, primarily in the vicinity of Reindeer
Island and Seal Island, a man-made gravel
island (Figure 2). In that effort, searches
were primarily along seismic lines or control
lines delineated by snow machine tracks.
Surveys in 1983 were conducted again in the
Reindeer Island area and at numerous
shore-fast ice locations from Norton Sound
in the Bering Sea north to Point Barrow in
the Chukchi Sea (Figure 1). In 1984, shore-
fast ice was surveyed in Kotzebue Sound and
drifting ice was surveyed elsewhere in the
Chukchi Sea. Shore-fast ice was surveyed in
the vicinity of Point Barrow in 1985, 1986,
and 1987. The 1987 surveys also included
efforts in the Beaufort Sea east of Point
Barrow; on shore-fast ice between the man-
made gravel island, Tern Island (70°17'N
147O28'W) and Narwhal Island (70°24'N
147'30'W) and at several locations in
drifting ice.
At a minimum, each structure was
probed with an aluminum rod, 1 cm in
diameter. Most structures were partially
uncovered to permit examination and
measurements after which they were
carefully re-covered. When examined,
structures were classified as breathing
holes, simple resting lairs, multi-chambered
lairs, or birth lairs, and notations were made
of the dimensions, physical setting, and
indications of predator activity. The status
of each structure was recorded as open, if
the hole through the ice was maintained by
the seal to its maximal diameter; frozen, if
the entire hole was refrozen; or, in the case of
lairs as altered, if access to the lair was
obstructed by partial freezing of the access
hole or by a collapsing ceiling. Structures
that were classified as frozen were
considered to have been abandoned by the
seals.
In most instances, structures were
examined when first located. In 1982,
however, many structures were probed when
first located, but they were not examined
further until a subsequent revisit. In that
year, approximately 72% of the structures
were visited two or more times.
The examination of a structure was
considered to constitute a disturbance, thus
all examinations subsequent to the initial
one were of structures previously exposed to
anthropogenic disturbance and were
categorized as such.
Monitoring of haulout activity
Fourteen ringed seals were live-
captured at breathing holes in the shore-fast
ice; three in the Beaufort Sea in 1982, six in
the Beaufort Sea in 1983, and five in
Kotzebue Sound in 1984 (Kelly et al. 1986).
The weight, sex, and minimal age, as
determined by claw annuli (McLaren
1958b), of each seal was recorded. VHF
radio transmitters were glued to the pelage
of the dorsum, midway between the base of
the tail and the region of maximal girth,
before each seal was released at its capture
site.
The unique frequency of each deployed
transmitter was monitored from a nearby
camp every half-hour in 1982 and every hour
in 1983 and 1984 for up to 2.5 months
between March and early June. Signals
could be received only when the transmitters
were above the ice surface, thus indicating
that the seals were out of the water.
Haulout bouts of 13 of the radio-tagged seals
were monitored after their release; no
signals were received from one of the seals
tagged in 1983. Whenever feasible, the
exact location of the signal sources,
indicating the location of lairs or other
haulout sites, was determined. Those
determinations were accomplished by skiing
or walking around the signal source while
monitoring the signal with a hand-held,
directional antenna.

SEISMIC SURVEY LINES, /
0 1000 2000 3000 4000 METERS - Seismic survey lines
p Subnivean seal structures
Figure 2. Subnivean structures used by radio-tagged ringed seals and seismic survey lines in
1982 (north of Reindeer Island) and 1983 (south of Reindeer Island). (After Kelly et al. 1986.)
When radio-tagged seals were in their
lairs and subjected to anthropogenic sounds,
notations were made of their behavioral
responses (departed or remained in lair). In
April 1983, a simulated seismic survey was
conducted near Reindeer Island in an area
occupied by radio-tagged seals. The survey
consisted of four seismic lines; A, B, C, and D
depicted in Figure 2. Four machines
travelled over each seismic line. A drill
truck used a power auger to bore holes
through the ice, generally every 67 m along
the survey lines. A bulldozer (D6
Caterpillar) then leveled the ice along the
survey lines. Every 67 m, the ice surface
was vibrated ten times in 16 second sweeps
from 10 to 70 Hz by a Vibroseis machine. A
fuel truck followed at the end of the convoy.
Lines A and B were vibrated on 21 April,
lines C and D were vibrated on 22 April, and
line A was vibrated a second time on 27
April. Additionally, the behavioral
responses of seals in lairs to the sounds of
I
helicopters, snow machines and other
equipment operating on the ice, and people
walking or skiing on the ice were
documented opportunistically.
Results
Aerial surveys
^
Results of our aerial surve s were
conflicting with regard to the e fects of
seismic survey activities on seal
distribution. In 1981, we observed an
average of 1.3-1.4 ringed seals per nm2 on
the shore-fast ice between Point Barrow
(71°23.2'N 156'27.2W) and Oliktok Point
(70°30.0'N 14g052.6W) and an average of
1.1 ringed seals per nm2 between Oliktok
Point and Barter Island (70°08.1'N
142O24.7'W), similar to the densities
observed in four surveys conducted between
1970 and 1978 (Burns et al. 1981).

The observed densities of basking seals
along seismic lines and intermediate control
lines on three days are shown in Table 1.
Densities along the two sets of lines differed
significantly only on 3 June when densities
along the seismic lines were 58% of those
along control lines.
Concentrations of seals basking along
newly opened cracks (as opposed to at
breathing holes) appeared unexpectedly
early in June 1981 and increased from 13.9%
of all seals sighted on 4 June to 16.5% on
7 June and 22.8% on 8 June. The indication
was that seals were leaving breathing holes
and lairs maintained through the winter in
favor of haulout sites along newly opened
cracks. Therefore, we suspected that the 5
and 9 June surveys were less representative
of the early spring distribution of seals than
was the 3 June survey. Thus, in 1982, we
scheduled aerial surveys to begin in late
May in the hope of obtaining relative
densities that were more representative of
early spring distribution.
A replicated survey track from Cape
Halkett (70°48'N 152Oll'W) to a point
offshore of Prudhoe Bay yielded 1.28 seals
per nm2 in 1981 and 1.84 seals per nm2 in
1982, not a significant difference (t = 1.03,
df = 32, p > 0.10). The 1982 effort also
included eight flights in which a series of
seismic and control lines were surveyed
(Table 2). Observed densities along seismic
and control lines did not differ si
except on 26 May when more sea lifi~tly, s (1 00 per
nm2) were observed along seismic lines than
along control lines (0.48 seals per nm2) (t =
2.24, df = 13, p < 0.05).
Responses of seals to anthropogenic noises
Haulout bouts of the radio-tagged
ringed seals were monitored for 3 to 10
weeks (Table 31, and we documented the
behavioral responses of seals that were
hauled out in lairs when exposed to a variety
of anthropogenic noises. Single observations
were obtained during the approach of a
seismic convoy, a hovercraft, and a dog. A
seal (GI83) departed his lair when a seismic
convoy was 0.64 km away. On another
occasion, the same seal departed his lair
when a dog approached within 5 m of the
lair. Another seal (SA82) remained in her
lair when a hovercraft passed at a distance of
2.5 km.
Responses of seals to helicopter noise
was variable. Responses to helicopters
landing and taking off (i.e., when the
helicopters were applying maximal power
and lift close to the surface) were noted six
times. On two occasions, at distances of 1.0
and 3.0 km, the seal departed. On four
occasions, all at distances greater than
2.5 km, the seals remained in their lairs.
Seals departed lairs in five of 14 cases in
response to airborne helicopters. In one
case, the helicopter was directly over the lair
at an altitude of 152 m, and in another case
it was 5 km away at that same altitude when
the seal departed. The closest approaches of
airborne helicopters that were tolerated by
seals in lairs were 0.6 km at an altitude of
122 m and directly overhead at an altitude of
762 m.
Nine observations of the seals'
responses to operating snow machines were
obtained. One seal remained in its lair on
two occasions when snow machines were
operatin 0.5 km distant. Three other seals
departe

Table 2. Ringed seal densities observed along adjacent seismic and control transects during
aerial surveys in 1982.
Date
26 May
29 May
30 May
31 May
31 May
1 June
3 June
4 June
Seismic Transects Control Transects
Density Length Density Length
(sealslnm2) (nm) (sealslnm2) (nm) T Test
lairs were noted. In nine cases, the seals
remained in the lairs, including four cases in
which the people approached to within
0.2 km. In 12 cases where the seals
responded by departing, people were
walking 0.1 to 0.6 km from the lairs. In each
of four cases in which a person walked
within 0.1 km, the seal departed from its
lair.
People on skis approached lairs 26
times. Seals remained in lairs during five of
six approaches by skiers to within 0.2 km.
Four departures were observed, one at
0.2 km, two at 0.3 km, and one at 0.4 km.
In all instances in which seals departed
lairs in response to noise disturbance, they
subsequently reoccupied the lair. The
breathing holes and lairs known to have
been used by the three female seals radio-
tagged in 1982 (SA82, BA82, BE821 were
within an extensive grid of seismic lines that
had been vibrated a few weeks before the
Table 3. Ringed seals radio-tagged in the Beaufort Sea and Kotzebue Sound . (After Kelly et al.
1986.)
- -
Age (yrs) First Last Known
Seal indicated Weight Date signal signal minimal
no. Sex by claws (kg) tagged received received no. of lairs

seals were tagged (Figure 2). Each seal
maintained at least one lair and one
breathing hole within the grid of seismic
lines. The breathing holes and lair access
holes passed through 2 m of ice and
presumably had been maintained since
freeze up, or shortly thereafter. The
breathing holes ranged from 19 to 129 m
from the nearest seismic line; the lairs
ranged from 250 to 700 m from the nearest
seismic line. All of those structures
remained in active use until at least early
June, approximately two months after the
seismic convoy had left the area.
Of the seals radio-tagged in 1983, three
(two males and one female) occupied lairs
within the grid of simulated seismic survey
lines (Figure 2). One of the males (TI83)
tended to haul out late at night in April and
was never in his lair during the daytime
hours that the seismic convoy operated
(Kelly et al. 1986). The frequency and
duration of his haulout bouts and their
locations showed no significant changes in
relation to the seismic operation.
The second male (GI83) was in his lair
when the seismic equipment approached on
21 April and his departure, when the convoy
was 0.64 km distant, was mentioned earlier.
The next signal from him was heard on
23 April when he briefly hauled out at a
different site than the one he departed two
days earlier. Thereafter, five additional
haulout bouts by that seal were recorded, at
least two of them from the lair he departed
in response to the convoy. No signals were
received from his transmitter after 26 April
when he hauled out briefly (less than one
hour). On 17 May, examination of the lair
he had occupied during the seismic survey
indicated its continuing use as a haulout
site, but we were unable to ascertain
whether he or some other seal was using the
lair at that time.
The female (LR83) using the 1983
seismic area was ta ged after the survey
was completed. Her F our haulout sites were
within the seismic line grid, and her birth
lair probably was in use prior to the seismic
survey (Figure 2). She continued to use that
lair as late as 4 June, more than one month
after the survey.
Fate of seal structures in areas of industrial
activity
While lair use by the radio-tagged
seals appeared to be interrupted only briefly
by anthropogenic disturbance, we did
observe cases of abandonment in our
examination of other structures. We found
evidence that the examination by
investigators and the activities associated
with seismic profiling and gravel island
construction increased the rates of
abandonment. Data on these points were
obtained in 1982 on the shore-fast ice of the
Beaufort Sea.
Of 37 structures that were opened and
examined when first found, 46% were frozen
or altered when revisited. Another 59
structures were only probed when first
found, and 22% of those structures were
frozen or altered when revisited. The
difference in the proportion of structures
frozen or altered was significant (G = 6.35,
df = 2, p < 0.05).
The fate of 110 structures was
investigated as a function of their distance
from seismic lines and a gravel island under
construction. Within 150 m of the seismic
lines, 14/48 (29.2%) structures were
abandoned, compared to 4/37 (10.0%) of the
structures beyond 150 m of the same lines.
The difference was statistically significant
(G = 5.53, df = 1,O.Ol < p < 0.025).
Within 8 km of the Seal Island
construction site, the incidence of
abandonment was 8/25 (32.0%), similar to
the rates close to seismic lines. Near the
island construction site, no differences were
detected in abandonment rates within and
beyond 150 m of the search lines.
We observed varying rates of
abandonment in over 700 seal structures
examined between 1982 and 1987. Our
samples were grouped according to ice type,
the amount of anthropogenic disturbance,
and the number of examinations by the
investigators. That breakdown resulted in
the four samples shown in Table 4. They
are; (1) 93 structures from the drifting ice,
not subjected to unnatural noise
disturbance; (2) 471 structures from shore-
fast ice and not subject to human
disturbance; (3) 148 structures from shore-
fast ice and subjected to "on-ice" industrial
activity; and (4) 107 of the above 148
structures after being subjected to two or
more investigator examinations as well as
industrial activities.
On the shore-fast ice with no
significant anthropogenic disturbances
(sample 2), only 4.0% of the structures were

Table 4. Rates of abandonment (freezing of breathing or access holes) of ringed seal structures
in four samples collected between 1982 and 1987.
Anthropogenic Number Percent
Sample disturbance N frozen frozen
1. Drifting ice
Chukchi &
Beaufort seas
1984 & 1987
2. Shore-fast ice
Bering, Chukchi,
& Beaufort seas
1983-1987
3. Shore-fast ice
Beaufort Sea
1982
4. Shore-fast ice
Beaufort Sea
1982
None 93 12 12.9%
None
frozen. This represents what we consider to
be the natural rate of abandonment on
shore-fast ice during our study.
Of the structures in the shore-fast ice
that were subjected to industrial noise
(sample 31, 13.5% were abandoned when
first examined. The difference in
abandonment rates between the industrially
disturbed sample (3) and the undisturbed
sample (2) was highly significant (X2 =
17.14, df = 1, p < 0.001). In sample 4,
which includes 107 of the structures from
sample 3 that were subjected to multiple
examinations as well as to industrial
activities, the abandonment rate was 32.7%,
indicating a significant increase due to the
investigator's activities (X2 = 12.42, df = 1,
p < 0.001).
Sample 1 includes 93 structures from
the drifting ice and is included for
comparison with the shore-fast ice samples.
Compared to the latter habitat, the drifting
ice is less stable. Furthermore, a high
pro ortion of ringed seal structures in the
?
dri ting ice are opened by polar bears, a
source of natural disturbance similar to that
of our opening a lair. In the 1987 driftiice
sample, 22/39 (56.4%) of the structures had
been visited by bears, but that proportion
was biased since the sample was collected
Seismic surveys, 148 20 13.5%
island building
Seismic surveys, 107 35 32.7%
island building,
& investigator
examinations
while following bear tracks (Kelly et al.
1987). The 1984 drifting ice sample was
random, however, and of 54 structures in
that sample, nine (16.6%) were visited by
bears. The rate of abandonment of
structures in the drift ice (sample 1) was
12.9%, similar to that found on the shore-
fast ice in 1982 (sample 3). Only one other
sample not subjected to industrial
disturbance showed rates of abandonment
similar to those in the drifting ice. One of
seven undisturbed, shore-fast ice samples
exceeded 5% abandonment. That sample,
collected near Barrow in 1986, had a 12.8%
abandonment rate and was associated with a
very high incidence of arctic fox activity.
Discussion
The responses of ringed seals to noise
disturbance were quite variable as indicated
by the behavior of radio-tagged seals and by
the rates of abandonment of seal structures
near and at various distances from human
activities. Some structures remained in
active use despite close proximity to seismic
survey lines, snow machine trails, gravel
island construction, and helicopter flight
paths. Other structures were abandoned
quickly when exposed to noises at greater
distances. That variation probably is due in

part to differences in the noise environment
that are difficult to measure. For example,
helicopter noise is muffled on warm, cloudy,
snowy, or windy days and is loudest in clear,
calm, cold conditions. A snow machine, or
person on foot or skis, produces different
kinds and levels of noise when the snow is
very cold and hard or windblown compared
to newly fallen, relatively warm or soft
snow. Snow machines travelling over
smooth ice sound different than those over
rough ice. Also, the seals' sensitivity to
anthropogenic noise may lessen when
background noise, such as from wind-driven
snow or ice movement is high.
In spite of an array of variables not
accounted for, it is apparent that ringed
seals in lairs are aware of sound intrusions,
and they generally react to mechanical
conveyances at greater distances than they
do to eople on foot or on skis. The
indivi I ual variation in their reactions,
however, makes it difficult to define
"critical" distances for noise disturbances.
Although we found fewer active seal
structures within 150 m of seismic lines
than beyond that distance, we cannot say
how the rate of abandonment changed
within that range, which was chosen on the
basis of sample size, rather than distance
se. -
Gravel island construction appeared to
result in displacement of ringed seals at
rates similar to those observed close to
seismic survey lines. Our data suggested
that the radius of disturbance was greater
around Seal Island when it was under
construction than was the radius around
seismic exploration, but the data are
insufficient for determining the distance
from the island at which the incidence of
abandonment began to decrease.
The displacement of some seals within
two hundred meters of seismic lines
probably results in little, if any, increased
mortality since, as we reported elsewhere
(Kelly 1985; Kelly et al. 1986), individual
seals use more than a single lair and as
many as four or five lairs each. Our
telemetric studies indicated that the
distances between lairs used by individual
seals averaged 572 m for females and 2,018
m for males (Kelly et al. 1986). We do not
know if mortality would be likely to occur if
individual seals were displaced completely
from the areas containing all of their lairs.
At the very least, such an event would be
likely to increase intra-specific strife by
forcing displaced seals to use structures
maintained by other seals.
That ringed seals respond to noise
disturbances by fleeing into the water
probably is the result of their subjection to
predation by polar bears and arctic foxes.
Weddell seals (Levtonychotes weddelli),
which breed on the shore-fast ice of
Antarctica, have evolved in the absence of
surface predators and are much less readily
disturbed (Stirling 1977; Kooyman 1981).
The rates of ringed seal structure
abandonment that we observed in areas of
noise disturbance were more than three
times greater than the overall rates in
undisturbed areas but similar to the rates in
areas of frequent predator activity.
Increasing the frequency with which
ringed seals flee lairs may increase stress
levels and energy demands at times when
rest is important to their well-being. Lair
occupation becomes increasingly frequent
and longer in duration throu hout the
spring months (Kelly et a . 1986),
apparently due to the seals' need to maintain
high epidermal temperatures while
replacing their pelage (Feltz and Fay 1966).
Of potentially greater importance are
the effects of disturbances that cause
structures to be completely abandoned. That
occurrence would be deleterious especially
for nursing pups. Furthermore, females
with nursing young are more susceptible to
disturbance in lairs by virtue of their more
frequent and extended haulout bouts (Kelly
et al. 1986). Short of abandoning a pup,
female seals can take them through the
water to alternate lairs (Smith and Stirling
1975; Taugbd 1982). If a newborn pup is
forced into the water, however, it may not
survive the resultant heat loss. At birth,
ringed seal pups do not have the insulating
blubber layer that protects older seals from
excessive heat loss when submerged. Pups
that do survive swimming through the water
to an alternate lair would have to expend
significant amounts of their energy reserves
in order to maintain core temperature while
drying (Taugbd 1982). Those pups would be
easier prey for polar bears and arctic foxes
and would be less able to withstand other
stresses.
Our investigation focused on the effects
of noise disturbance on the seals' use of lairs
and breathing holes. From our telemetric
studies, we know that seals spend the
majority of the time in the water under the
f?

ice (Kelly et al. 1986). Little is known about
their activities under the ice, although much
of it must involve feeding and, perhaps,
territorial defense. Sound is readily
conducted throu h the ice into the water,
f
and the effects o noise disturbance on seals
under the ice remains unknown. Recent
experiments with captive ringed seals
suggest that ambient noise provides a
critical navigational cue to seals swimming
under ice in total darkness (Wartzok et al.
1987).
Acknowledgements
Much of this work was funded by the
Minerals Management Service, Department
of the Interior through an Interagency
Agreement with the National Oceanic and
Atmospheric Administration, Department
of Commerce, as art of the Alaskan Outer
Continental Shelf Environmental
Assessment Program. Additional support
was provided by the Department of Wildlife
Management, North Slope Borough,
Barrow, Alaska, and the U. S. Fish and
Wildlife Service. We are grateful to those
organizations and to a great many people
who assisted in the field work and provided
logistic support. We learned a great deal
about ringed seals and the use of trained
dogs for locating their subnivean structures
from T. Smith and M. Hammill of the Arctic
Biological Station, Ste. Anne de Bellevue,
Quebec and J. Memorana, Holman, N. W. T.
Personnel of Shell Western E & P, Inc.,
Geophysical Services, Inc., Western
Geophysical Co., Sefel Geophysical Co.,
NANA Regional Corporation, U. S. Coast
Guard - Port Clarence LORAN Station,
711th Aircraft and Warning Squadron Radar
Site - Cape Lisburne, Telonics, Inc., and the
NOAA Helicopter Corps were extremely
helpful to our efforts.
References
Boas, F. 1888. The central Eskimo. Bur.
Amer. Ethnology, 399-699.
Burns, J. J. and Eley, T. J. 1978. The
natural history and ecolow of the bearded
seal (~rignathus barbatus) and the ringed
seal (Phoca hispidm. Dep. Commer.,
NOMXSEAP Environ. Assess. Alaskan
Continental Shelf, Ann. Rep. Year Ending
March 1978,l: 99-162.
Burns, J. J. and Harbo, S. J. 1972. An aerial
census of ringed seals, northern coast of
Alaska. Arctic 25: 279-290.
Burns, J. J. and Kelly, B. P. 1982. Studies of
ringed seals in the Alaskan Beaufort Sea
during winter: impacts of seismic
exploration. Unpubl. ann. rep., OCSEAP
Res. Unit 232, to U.S. Dep. Commer.,
NOAA, Off. Mar. Pollut. Assess., Juneau,
Alaska, 57 pp.
Burns, J. J., Kelly, B. P. and Frost, K. J.
1981. Studies of ringed seals in the Beaufort
Sea during winter. Executive Summay, 31
October 1981. OCSEAP Res. Unit 232, to
U.S. Dep. Commer., NOAA, Off. Mar. Pollut.
Assess., Juneau, Alaska. 81 pp.
Cox, S. L. and Spiess, A. 1980. Dorset
settlement and subsistence in northern
Labrador. Arctic 33: 659-669.
Feltz, E. T. and Fay, F. H. 1966. Thermal
requirements in of epidermal cells from
seals. CryobioT3: 261-264.
Geraci, J. R. and St. Aubin, D. J. 1980.
Offshore petroleum resource development
and marine mammals: a review and
research recommendations. Mar. Fish. Rev.
42: 1-12.
Hall, C. F. 1866. Arctic researches and life
among the Esquimaux: being the narrative
of an expedition in search of Sir John
Franklin, in the years 1860,1861, and 1862.
Harper and Bros. Publ., N. Y., 595 pp.
Kelly, B. P. 1985. Dependence by ringed
seals on subnivean lairs. Sixth Biennial
Conference on the Biology of Marine
Mammals. (Abstract).
Kelly, B. P., Quakenbush, L. and. Rose, J. R.
1986. Ringed seal winter ecology and effects
of noise disturbance. Final rep., OCSEAP
Res. Unit 232, Part 2, to U.S. Dep. Commer.,
NOAA, Natl. Ocean Serv., Ocean Assess.
Div., Alaska Office, Anchorage, Alaska,
91 PP.
Kelly, B. P., Amstrup, S. C., Gardner, C. and
Quakenbush, L. T. 1987. Predation on
ringed seals in the western Beaufort Sea.
Seventh Biennial Conference on the Biology
of Marine Mammals, 5-9 December 1987,
Miami, Florida. (Abstract).
Kooyman, G. L. 1981. Weddell seal
(Leptonychotes weddelli Lesson, 1826). 12

"Handbook of marine mammals" (S. H.
Ridgway and R. J. Harrison F.R.S., eds.).
Vol. 2: Seals. Academic Press, London, 275-
296.
Manning, T. H. 1944. Hunting implements
and methods of the present-day Eskimos of
north-west Hudson Bay, Melville Peninsula,
and south-west Baffin Island. Geom. J. 103:
10-152.
Mansfield, A. W. 1983. The effects of vessel
traffic in the arctic on marine mammals and
recommendations for future research. Can.
Tech. Rep. Fish. and Aquatic Sci. No. 1186.
Ste-Anne-de-Bellevue, Quebec, 97 pp.
McLaren, I. A. 1958a. The economics of
seals in the eastern Canadian arctic. Fish.
Res. Board Can. Circ. 1,93 pp.
McLaren, I. A. 1958b. The biology of the
ringed seal (Phoca hispida Schreber) in the
eastern Canadian arctic. Fish. Res. Board
Can. Bull. 118: 97 pp.
Norris, K. S. 1981. Marine mammals of the
arctic, their sounds and their relation to
alterations in the acoustic environment by
man-made noise. In "The question of sound
from icebreaker operations: the proceedings
of a workshop February 23-24, 1981",
Toronto, Ontario (N. M. Peterson, ed.).
Arctic Pilot Proj., Petro-Canada, Calgary,
Alberta, 304-309.
Rausch, R. L. 1973. Post mortem findings
in some marine mammals following the
Cannikin test on Amchitka Island. Ms.
Atomic Energy Comm., Las Vegas, Nev.
86 PP.
Ronald, K. and Dougan, J. L. 1982. The ice
lover: biology of the harp seal (Phoca
groenlandica). Science 215: 928-933.
Schusterman, R. J. and Moore, P. W. B.
1980. Noise disturbance and audibility in
innipeds. J. Acoust. Soc. Am. 70, Suppl. 1,
583.
Smith, T. G. 1976. Predation of ringed seal
pups (Phoca his ida) by the arctic fox
(Alopex lagopu*n. J. Zool. 54: 1610-
1616.
Smith,
+
T. G. 1987. The ringed seal, Phoca
his ida, of the Canadian western Arctic.
an. Bull. Fish. and Aquat. Sci. 216.
Ottawa, 81 pp.
Smith, T. G. and Stirling, I. 1975. The
breeding habitat of the ringed seal (Phoca
hispida). The birth lair and associated
structures. Can. J. Zool. 53: 1297-1305.
Stefansson, V. 1913. My life with the
Eskimos. MacMillan Co., New York, 538 pp.
Stewart, B. S. 1981. Behavioral response of
northern elephant seals and California sea
lions on San Nicolas Island. California. to
loud impulse noise. J. ~cbust. Soc. Am.
Suppl. 1, Vol. 70, S84.
Stirling, I. 1977. Adaptations of Weddell
and ringed seals to exploit the polar fast ice
habitat in the absence or presence of surface
predators. In "Adaptations within antarctic
ecos stems" (G. A. Llano, ed.). Proceedings
of the Third Symposium on Antarctic
Biology. Gulf Publishing Co., Houston,
Texas, 741-748. 1252 pp.
Taugb01, G. 1982. Ringed seal
thermoregulation, energy balance and
development in early life, a study on P s
hispida in Kongsfj., Svalbard. Unpubl.
thesis, 102 pp. (Transl. from Norwegian by
Can. Fish. Aquat. Sci., No. 5090, 1984,
109 pp.)
Wartzok, D., Elsner, R., Kelly, B. P.,
Mimken, G. and Quakenbush, L. T.. 1987.
Visual, acoustic, and vibrissal contributions
to under-ice navigation by ringed seals,
Phoca his ida. Seventh Biennial Conference
of Marine Mammals, 5-9
December 1987, Miami, Florida. (Abstract).
Wenzel, G. 1984. Archaeological evidence
for prehistoric Inuit use of the sea-ice
environment. In "Sikumiut: the people who
use sea ice" (A. Cooke and E. V. Alstine,
eds.). Canadian Arctic Resources
Committee, Ottawa, 41-59. 153 pp.

Abstract
RESPONSES OF MIGRATING NARWHAL AND BELUGA
TO ICEBREAKER TRAFFIC AT THE ADMIRALTY INLET ICE-EDGE,
N.W.T. IN 1986
A mobile ice-based camp was
maintained at the mouth of Admiralty
Inlet from 22 May to 10 July 1986.
Aerial surveys of migrating whales
were flown, using a Bell 206 Jet
Ranger helicopter, both when ships
were absent from the ice-edge and
whales were presumed to be undisturbed
and when ships were in the vicinity of
the ice-edge. Numbers, distribution
and behaviour of both narwhal and
beluga were recorded. Data indicated
that day-to-day variation in numbers
and distribution of both species was
substantial, making assessment of the
effects of ship traffic and assocl-.zed
noise on numbers difficult. Observa-
tion of undisturbed whales indicated
that migratory patterns, group size
and general activity of narwhal dif-
fered significantly from those of bel-
uga. In the presence of ship traffic
beluga showed an increase in non-
directed movement, a decline in
directed movement and a decline in
inactivity. Narwhal showed no change
in non-directed movement, an increase
in slow directed movement and a slight
This is a reviewed and edited version of a paper
presented at the Ninth International Conference on Port
and Ocean Engineering Under Arctic Conditions, Fair-
banks, Alaska, USA, August 17-22, 1987.
Susan E. Cosens
Larry P. Dueck
Fisheries and Oceans, Winnipeg, Manitoba, CANADA
decline in inactivity. Changes in
group structure and orientation in re-
sponse to ship traffic also differed
between the two species. Overall, re-
actions of both species appeared to be
less pronounced than was described in
a previous study. This difference may
be attributable to differences in ice
conditions.
Introduction
As a result of recent interest
in developing resources in the Arctic,
there is concern that disturbance by
shipping traffic and other industrial
activity may have detrimental effects
on marine mammal populations (see
Mansfield 1983). Both industry and
government have directed effort to ad-
dress these concerns. Most of the re-
search has taken place in the Beaufort
Sea area where exploration for hydro-
carbons, including seismic and drill-
ing operations, has been most inten-
sive. Research has focussed on large
cetaceans, such as the bowhead whale
(Balaena mysticetus) (eg. Ljungblad et
al. 1985; Richardson 1985; Richardson
et al. 1985). The beluga whale
(Delphinapterus leucas) has also been
the subject of similar studies in the
western Arctic (see Ford 1977; Fraker
1983).

Exploration for hydrocarbons in
the eastern Arctic has been limited
but other industrial activities, in-
cluding mining, have required the use
of icebreakers and tankers in the
Northwest Passage. The Arvik and Nan-
asivik mines, for example, send ship-
ments of ore to Europe each summer.
Government and industry have cooperat-
ed in efforts to extend the shipping
season in the eastern Arctic by bring-
ing the icebreaking tanker MV Arctic
to Nanasivik at the north end of
Baffin Island, Northwest Territories,
during the fast ice season. More
effort has been directed toward bring-
ing the ship in before break-up rather
than during the winter freeze-up per-
iod.
Early shipping raises potential
problems of interference with cetacean
migrations through Lancaster Sound.
In 1982, a study of responses of nar-
whal (Monodon monoceros) and beluga to
icebreaker traffic was begun by LGL at
the Admiralty Inlet ice edge. At that
time, the MV- Arctic, accoipanied by a
Canadian Coastguard icebreaker, was
making one round trip through the fast
ice of Admiralty Inlet at the end of
June, about the time that the odonto-
cete migration begins to peak. Re-
sults from this study indicated that
(1) odontocetes moved away from loca-
tions where icebreakers were generat-
ing noise, (2) initial avoidance re-
sponses occurred when ships were over
40 km from the ice edge and (3) beluga
and narwhal were behaviourally differ-
ent in their responses to icebreaker
disturbance.
Intending to expand the work
that LGL conducted, we began our study
in 1986. We intended to monitor odon-
tocete migration patterns over a long-
er period of time than had been done
previously, and to compare non-
disturbance with disturbance pat-
terns. We wanted to quantify be-
haviour patterns of both beluga and
narwhal such that statistical compari-
sons of behaviour between species and
between disturbance conditions could
be made. We conducted 27 helicopter
surveys between 22 May and 6 July and
collected data on the behaviour of
beluga and narwhal during two passages
of a single ship and one passage of
two ships through the fast ice. This
paper summarizes the data on numbers,
distribution, activity and orientation
collected during these surveys.
Materials and Methods
We established a mobile ice-
based camp at the mouth of Admiralty
Inlet (73O45'N 84'05'W) on 22 May and
conducted aerial surveys from 24 May
to 6 July. Our surveys included three
observers in a Bell 206 Jet Ranger
helicopter that flew at an altitude of
about 230 m and a speed of about 160
km/hour. We planned to proceed east
from our camp along the floe edge to
Navy Board Inlet, north until we were
several kilometres from the floe edge,
west to Cape Crauford, then return to
camp. This flight path would have
covered the mouth of Admiralty Inlet
both close to and several kilometers
from the shear line. Actual flight
paths were variable (Fig. 1). Pack
ice was heavy, and our survey route
usually followed open leads which were
irregular in distribution. We fre-
quently made return flights halfway or
more across Lancaster Sound, following
the edge of the pack ice. Occasional-
ly, Lancaster Sound was completely
covered in pack ice so large areas of
open water were infrequently encount-
ered. There was no land-fast ice edge
across Lancaster Sound in 1986, thus
our westerly flights were limited only
by fuel requirements.
Whales observed within a 250 m
strip on either side of the helicopter
were identified to species, counted
and, when possible, classed to sex and
gross age (adult, sub-adult, new-
born). Their ongoing behaviour,
orientation and general movement speed
was noted. We also recorded ice con-
ditions by estimating % ice cover
within the 500 m survey strip. Chang-
es in ice conditions, helicopter alti-
tude, flight speed and bearing were
recorded on cassette recorders.
Ongoing behaviour was classified
into four categories: (1) directed
movement - whales maintained constant

Lancaster
Sound
Figure 1. a) A survey made on 10 June
that followed our preferred flight
path. b) A survey made on 25 May that
deviated from our preferred flight
path. Hatched areas represent land
fast ice.
direction during swimming movements,
(2) circling - whales changed the di-
rection in which they were swimming,
(3) deep dive - whales disappeared
from view after a dive, 4) resting -
whales remained stationary showing no
locomotory behaviour. The latter cat-
egory was further classified by noting
location of whales relative to the
water surface: (1) back exposed -
backs breaking the water surface, (2)
hanging - body totally submersed below
the water surface.
We estimated the bearing of
sighted whales relative to true
north. Orientation relative to ships
present in the survey area during dis-
turbance surveys were calculated later
in the laboratory. Movement speed of
whales was classified either as slow
or moderate to fast. Fast-moving
whales exhibited more vigorous fluke
strokes and raised more of their backs
above the water than did slow-moving
whales. Fast movement was always di-
rected and often synchronous within
the group.
We conducted surveys during non-
disturbance periods at various times
of day to determine whether there was
any evidence of diurnal variation in
behaviour and to control for differen-
ces in time of day between disturbance
and non-disturbance surveys (Table
1). We could not predict when ships
would arrive at the ice edge, but were
able to increase the chances that con-
trol and disturbance surveys would be
done at comparable times of day.
We used the Canadian Tide and
Current Tables, with Dundas Harbour on
Devon Island (Fig. 1) as our point of
Table 1. Distribution of control and
disturbance surveys relative to time
of day.
Time Number of Surveys
Control Disturbance
(no ship) (ship)

eference, to determine the tide state
for each survey (Table 2).
Weather conditions were recorded
at least twice daily. We used a hand
held anemometer to record wind speed
and a sling psychrometer to record dry
bulb temperature. During surveys we
noted obstructions to visibility of
whales including fog, light snow and
sea state conditions.
We used a rather broad defini-
tion of disturbance by including ships
within 130 km of the floe edge. Given
that whales were expected to occur
throughout Lancaster Sound during a
ship passage, we did not want to ex-
clude the possibility that disturbed
whales might move ahead of the ship as
it travelled eastward toward Admiralty
Inlet. We conducted 16 control sur-
veys where ships were not active in
the vicinity of the Admiralty Inlet
ice edge and 11 disturbance surveys
Table 2. Number of surveys conducted
during different types of tidal activ-
ity. Note that more control than dis-
turbance surveys were conducted during
ebb tides.
Rising High Ebb Low
Nondist. 2 3 7 4
Dist. 3 4 2 2
when ships were either moving through
Lancaster Sound or the fast ice of
Admiralty Inlet (Table 3). We main-
tained radio contact with both the =
-
des Grosseliers and the MV Arctic and
were able to obtain updates of ship
positions at the time of our surveys.
Table 3. Approximate ship positions during disturbance surveys conducted at Admiralty
Inlet during 1986.
Survey Date Identification Ship Location Activity
--
11
12
3 June
4 June
MV Arctic
--
CCG des Grosseliers
MV Arctic
moving west
moving
icebreaking through
floe edge
13
14
17
18*
2 3
6 June
8 June
11 June
12 June
25 June
--
MV Arctic
--
CCG des Grosseliers
--
MV Arctic
CCG -- des Grosseliers
--
CCG des Grosseliers
icebreaking
moving south
stationary
returning - moving north
along ship track
moving west
icebreaking through
floe edge
24
Lady Franklin
27 June --
CCG des Grosseliers
Lady Franklin
73'19'N 85'24'W
as above
following
icebreaking moving south
following
2 5
26
27
2 July
3 July
6 July
--
CCG des Grosseliers 74O26'N 82O38'W moving east
--
MV Arctic
73'42'N 84'43'W icebreaking through
floe edge
-- MV Arctic
returning - moving north
through floe edge
- - - - - --
*MV Arctic at mouth of Lancaster Sound.

Results
Aerial survey conditions
Weather conditions were similar
for both control and disturbance sur-
veys. No significant differences
occurred in dry bulb air temperature,
wind speed, wind direction or minimum
visibility (Table 4). Sea state,
estimated on the Beaufort Scale,
ranged from5 to 4 during control sur-
veys and 0-3 during disturbance sur-
veys. One possible source of differ-
Table 4. Summary of weather conditions
during 16 control and 11 nondisturbance
surveys. -
Survey Type
(0
*
i2 I,
Control Disturbance
Temperature
( OC) -1.6 3.8 +0.8 2.1
Wind dir O
220 89.4 243 124.0
Wind speed
(kn) 7.5 5.5 4.1 3.3
Minimum
visibility
(km) 20.7 8.2 16.2 5.8
ence between control and disturbance
surveys was the relatively larger num-
ber of non-disturbance surveys (Table
2) done during ebb tides. Control and
disturbance surveys also differed sig-
nif icantly (x =904.6, df=4, p

veys compared to 1182 during disturb-
ance surveys. This was equivalent to
averages of 32 (SD=32.1) and 107
(SD=211.1) narwhal/survey seen during
control and disturbance periods re-
spectively.
May June July
Survey Date
Figure 3. Abundance of migrating nar-
whal seen during control surveys (c)
and disturbance surveys (d).
Beluga were seen less regularly
and occurred in fewer numbers than did
narwhal (Fig. 4). The largest number
of beluga found during any given sur-
vey was 109, seen on 9 June, when the
peak in numbers was recorded. We saw
259 beluga in total during non-
disturbance and 198 during disturbance
surveys, or averages of 16 (SD=27.5)
and 18 (SD=30.2) beluga/survey respec-
tively.
Narwhal were absent from only
two surveys, neither of which were
disturbance surveys. Beluga were absent
during 5 of 16 non-disturbance
surveys and during 7 of 11 disturbance
surveys. T is difference was not sig-
9
nificant (x = 2.74, df=l, p>.05). On
five of the seven disturbance surveys
when beluga were absent, the ship was

Narwhal were seen in more open water
during disturbance than during non-
disturbance surveys.
Beluga distribu on also dif-
^-
fered significantly (x = 46.4, de4,
p

Table 6. Percentages of migrating narwhal and beluga involved in different activities
during control and disturbance surveys.
Activity*
% Narwhal % Beluga
Control Disturbance Control Disturbance
Directed movement 34 30 3 6 2 1
Slow directed movement 16 2 6 38 39
Back exposed 2 4 25 2

fairly evenly among other directions.
Orientation differed significantly
from random (x = 278.6, df=6,
p

of ships. We would predict signifi-
cant correlations between independent
variables and survey date, if this al-
ternate hypothesis is a likely expla-
nation. Related to survey date for
narwhal is numbers of whales observed,
where the last three disturbance sur-
veys produced the most whales. The
last survey found 714 narwhal, an
order of magnitude more than any num-
ber seen in a non-disturbance survey.
It is possible that some of the obser-
ved behavioural changes could be att-
ributable to the sheer number of nar-
whal in the area as a consequence of
interactions among groups or individ-
uals. We would predict correlations
between behaviour and total numbers
seen, if this factor is important.
A third potentially confounding
variable wa.s ice cover. By the time
the last disturbance surveys were
done, the pack ice was much reduced
and open water predominated. Non-dis-
turbance surveys were generally con-
ducted wher. ice cover was heavy. Sig-
nificant associations between depend-
ent variables and ice cover would sug-
gest that variation in ice cover could
explain observed results.
The fourth alternate hypothesis
is that differences in time of day be-
tween non-disturbance and disturbance
surveys can explain observed results.
This possibility results from the un-
even distribution of surveys across
time of day. The problem arises from
the time period 10:OO-14:OO where 6
control surveys but only 1 disturbance
survey were conducted. Other time
periods were more evenly sampled. To
test for the effect of sampling bias,
correlational analyses were again
used.
We did not consider tides to be
a confounding variable for narwhal.
Although control surveys sampled more
ebb tides than did disturbance sur-
veys, two of these control surveys did
not locate any narwhal. Surveys from
which we had behavioural data were re-
latively evenly distributed across
tides. Sampling for beluga was less
easy to control. Non-disturbance sur-
veys where beluga were actually seen
were mainly done during ebb and low
tides. This was not the case during
disturbance surveys where only one
survey was done during an ebb tide.
The lower number of disturbance sur-
veys and the absence of beluga from
many of these surveys resulted in sam-
pling bias. Behavioural differences
in beluga across treatments might be
attributable to tides. A difference
in means across tidal categories in
non-disturbance surveys would suggest
that this variable cannot be ruled out
as a factor contributing to differenc-
es between control and disturbance
surveys. Our samples, however, were
too few to do such a comparison in a
meaningful way.
Group size of narwhal
The correlation between mean
group size and survey date was not
significant during control surveys
(r,, = .09, N = 14, p>.05), but was
significant and positive during dis-
turbance surveys (r = .69, N = 11,
p

off at 4 whales (Fig. 5). We do not
know whether this trend also occurs
during non-disturbance conditions.
Total numbers cannot, therefore, be
ruled out as a factor contributing to
differences in mean group size between
control and disturbance surveys.
Total Narwhal
Figure 5. Relationship between mean
group size and total numbers of nar-
whal during control (c) and distur-
bance (d) surveys.
Mean group size did not vary
with time of day, thus time of day ef-
fects did not contribute to observed
differences in mean group size between
treatment groups. We also compared
group size in different categories of
ice cover and found no significant as-
sociation between group size and % ice
cover. Variation in ice cover was
therefore not a contributing factor to
variation in mean group size across
treatments.
Distribution relative to ice cover
Significant differences in dis-
tribution of both narwhal and beluga
relative to ice cover were found be-
tween control and disturbance sur-
veys. These changes in distribution
resulted in narwhal being found in
more open water and beluga in less
open water during disturbance than
during control surveys. Seasonal
changes in ice alone do not explain
observed variation in whale distribu-
tion because the distribution of
whales was not random relative to ice
cover. Whales may change their habi-
tat preferences during the course of
the migratory period, but this hypo-
thesis cannot be tested with the pre-
sent data.
General activity
Both narwhal and beluga showed
behavioural differences between con-
trol and disturbance surveys. Narwhal
activity did not vary significantly
with survey date nor time of day in
either treatment category. There is
no indication that behaviour varied
with ice type during non-disturbance
surveys, but during disturbance sur-
veys, the proportion of narwhal engag-
ed in directed movement is somewhat
lower than predicted and back exposed
is somewhat higher than predicted in
40 to 80% ice than in other ice types
(Table 8). The percentage of narwhal
for which behaviour was unknown was
also higher for these categories.
Data are really too limited to indi-
cate anything stronger than a trend at
this time.
Table 8. Percent of narwhal showing
directed movement (all DM) and back
exposed (BE) behaviour relative to ice
type.
Ice Class Control Disturbance
All DM BE Unkn. All DM BE Unkn.
* During disturbance surveys, no nar-
whal were seen in 20-40% ice cover.

Behavioural activity of narwhal
was compared with total numbers of in-
dividuals seen. For slow directed
movement, there was a significant pos-
itive correlation (r=.65, N=14, p

Although our data were not suf-
ficient to be conclusive, they do sug-
gest that beluga abundance on the
study area depended on ship traffic.
Their absence during most of the dis-
turbance surveys would indicate the
beluga are more likely than are nar-
ance. Finley and Davis (1984) state
that beluga moved further away and re-
mained away from the disturbance area
longer than did narwhal. Our results
are consistent with these observa-
tions.
Beluga showed no significant
variation in mean group size between
control and disturbance surveys. Of
course, the low numbers seen during
disturbance surveys makes comparison
difficult but we cannot, at the pre-
sent time, suggest that group size was
affected by disturbance from vessel
activity. Barber and Hochheim (1986)
conducted a photographic survey from a
Twin Otter aircraft along the
Admiralty Inlet floe edge during a
disturbance period. They located 5
groups of beluga containing 3-6 ani-
mals. Only 8% of beluga were seen as
solitary individuals. Although this
percentage was much smaller than ours,
the upper limit of 38 beluga in one
group was similar to our value of 32.
They also found, consistent with our
observations, that narwhal gr ups ty-
pically did not contain largelnumbers
of individuals. They recorde a maxi-
m of 7 narwhal in one group?
From photographs, ~arber and
Hochheim (1986) were able to measure
group densities (number of whaleslm )
and found that narwhal tend d to form
more compact (.09 whaleslm , SD=.06,
N=6 groups) than did beluga (-05
whaleslm , SD=.09, N=9 groups). More
data are needed to verify that differ-
ences between species are statistical-
ly significant and to determine whe-
ther group cohesiveness varies with
disturbance level.
Changes in mean group size of
narwhal were clearly correlated with
changes in total numbers of narwhal
seen. Although we do not claim to
have seen every narwhal in the study
area, we expect that our data are in-
dicative of relative numbers. The
?
correlation between group size and
numbers suggests that narwhal may be
aggregating when they can locate other
narwhal. It would be valuable to com-
plete the data set on undisturbed nar-
whal to see if the curve reaches the
same plateau at 4 individuals. If
shipping noise reduces the ability of
narwhal to locate one another, then we
would expect the control group to pro-
duce a larger mean group size when
large numbers of narwhal are present,
than does the treatment group. In e-
ffect, we would expect the maximum
mean group size to be greater than
four. Alternatively, in the face of a
reduced signallnoise ratio, narwhal
might tend to clump more closely, to
counteract the effect of a reduced
signal range. We cannot presently re-
ject the null hypothesis, that ship-
ping traffic has no effect on mean
group size.
Our data on activity patterns
during non-disturbance periods gener-
ally agree with observations made by
Finley et al. (1984). They found that
90% of beluga in pre-disturbance sur-
veys were engaged in directed move-
ment, and that narwhal tended to be
more often engaged (19%) in other
types of activities. Our proportions
were not as high, but we did find that
more beluga (76%) were involved in
directed movement than were narwhal
(50%).
Because of poor weather, Finley
et al. (1984) were unable to collect
equivalent aerial survey data on be-
haviour during the approach of the
ship. They did do observations from
the ice but did not quantify behaviour
during disturbance periods. Their ob-
servations suggest that beluga re-
sponded to ship activity by rapid
directed movement whereas narwhal
tended to sink beneath the water sur-
face, lie quietly near the floe edge,
moving out of the area only slowly.
Our results from disturbance periods
do not agree with those of Finley et
al. However, the comparison is limit-
ed by the different methods of data
collection and our small sample of
disturbed beluga. Our data suggest
that directed movement by beluga de-
clined, that circling increased; and

that slow directed movement by narwhal the habituation hypothesis. However,
increased. the critical test, as described ear-
Her, has yet to be made.
Our results do, however, confirm
that beluga and narwhal behave differ-
ently under normal conditions and re-
spond differently to disturbance by
ship traffic. Our observations sug-
gest that responses were less intense
in 1986 than they were in 1982 to
1984. Again, there are several possi-
ble explanations: 1) both narwhal and
beluga have been exposed to shipping
in Lancaster Sound, and are beginning
to habituate, 2) the presence of heavy
pack ice provided cover and the lack
of land fast ice across Lancaster
Sound enabled easy escape by disturbed
whales, thus responses to ship
approach were less intense than when
whales were trapped against ice in
relatively open water. Ice conditions
might also explain apparent differenc-
es in non-disturbance behaviour pat-
terns between the two studies. These
hypotheses could easily be tested by
monitoring responses to ship passage
under ice conditions more typical of
1982-84 than 1986.
The differences in distribution
by ice type between control and dis-
turbance surveys may or may not be re-
lated to disturbance by ship traffic.
It is possible that habitat preferenc-
es change over the migratory period.
Further observations of undisturbed
whales are necessary to clarify this
question.
Our data on orientation are con-
sistent with those of Finley et al.
(1983, 1984) and Miller and Davis
(1984) in that most whales avoided the
ship. Our results also support their
observation that responses to ships
begin to occur when vessels are 45 to
60 km away. Narwhal and possibly bel-
uga showed large changes in orienta-
tion relative to the ship when it was
nearer than 50 km. This suggests that
habituation may not be occurring. One
would expect whales to react at a
closer proximity if they were getting
used to vessel activity. This does
not appear to be the case, although
reactions seem to be less intense than
have been previously observed. The
orientation data favour rejection of
Barber and Hochheim (1986) re-
corded orientations of photographed
whales relative to a ship and found
that beluga were less variable in
their orientation, and appeared to be
more directed in their movements away
from the ship than were narwhal.
These results are again consistent
with Finley and Davis (1984). Thus,
it appears that beluga are more in-
clined than narwhal to vacate an area
of disturbance. This does not sug-
gest, however, that narwhal are less
disturbed by vessel activity. The
two species are behaviourally dif-
ferent, even during the spring migra-
tory period, and interspecific differ-
ences include reactions to disturbance
by vessel traffic.
Acknowledgements
Research funds were provided
through the Northern Oil and Gas
Action Program (NOGAP). Additional,
much appreciated field support was
provided by the Polar Continental
Shelf Program in Resolute Bay, N.W.T.
We also thank Glenn Williams, the
G.N.W.T. Natural Resources Officer in
Arctic Bay, N.W.T. ,for his advice and
generous donation of storage space
during the field portion of the stu-
d ~ . Ipilee Koonoo and Timothy
Sangoyak were our guides and field as-
sistants during our stay on the ice.
Val Churney and Dave Yablecki also as-
sisted with data collection and camp
maintenance. We thank the field crew
for maintaining cheerful dispositions
in spite of the often long working
hours and frequent camp moves and we
thank our helicopter pilots for their
capable flying skills. We also appre-
ciated the warm welcome we received in
Arctic Bay and the support provided by
the community. Finally, we thank the
Canadian Coastguard, in particular
Captain Dave Johns, for their coopera-
tion and support during the study, and
Allan Sneyd of Canarctic Shipping for
his assistance and information provid-
ed during the planning phase of the
project.

References
Barber. D. and Hochheim. K. 1986.
Results of aerial photographic surveys
for disturbance reactions of ceta-
ceans: Admiralty Inlet, N.W.T. Re-
port prepared by E.M.S.I. for Canada
Department of Fisheries and Oceans,
Central and Arctic Region, 30 p.
Finlev. -. K.J. and Davis. R.A. 1984.
Reactions of beluga whales and narwhals
to shin traffic and ice-breaking
alone ice-edges in the eastern Cana-
-
ed, King City, Ontario, for Cana-
da Department of Indian Affairs and
Northern Development, 42 p.
Finlev. K.J.. Greene. C.R. and Davis.
-.
R.A. 1983. A study of ambient noise,
ship noise, and the reactions of nar-
whals and belugas to the MV Arctic
breaking ice in Admiralty Inlet,
N.W.T. - 1982. Report by L.G.L. Lim-
i ted, Toronto, for Canada Department
of Indian Affairs and Northern Devel-
opment, 108 p.
Finley, K.J.. Miller. G.W.. Davis.
-.
R.A. and Greene, C.R. 1984. Respon-
ses of narwhals (Monodon monoceros)
and belugas (Delphinapterus leucas) to
ice-breaking ships in Lancaster Sound
- 1983. Report by L.G.L. Limited,
King City, Ontario, for Canada Depart-
ment of Indian ~ffairs and ~orthern
Development, 117 p.
Ford, J. 1977. White whale offshore
exploration acoustic study. Report by
F.F. Slaney & Co. for Imperial Oil
Ltd.
Fraker, P.M. 1983. The white whale
monitoring program, Mackenzie Estu-
ary. Part I. Migration, distribution
and abundance of whales and effects of
industry activities on whales. Report
by L.G.L., Sidney, B.C. for Esso
Resources Canada Limited, Dome Petro-
leum and Gulf Canada Resource, Inc.
.,
Ljungblad, D.K., Wursig, B., Swartz,
S.L. and Keene, J.M. 1985. Observations
on the behaviour of bowhead
whales (Balaena mysticetus) in the
presence of operating seismic exploration
vessels in the Alaskan Beaufort
Sea. - Report by SEACO, Inc. for United
States Minerals Management Service, 53
Mansfield, A.W. 1983. The effects of
vessel traffic in the Arctic on marine
mammals and recommendations for future
research. Can. Tech. Rept. of Fish
Aq. Sci. No. 1186, 97 p.
Miller, G.W. and Davis, R.A. 1984.
Distribution and movements of narwhals
and beluga whales in response to ship
traffic at the Lancaster Sound ice
edge - 1984. Report by L.G.L. Limit-
ed,King City, Ontario, for Canada De-
partment of Indian Affairs andNorthern
Development, 34 p.
Richardson, W.J. 1985. Behaviour,
disturbance responses and distribution
of bowhead whales Balaena mysticetus
in the Eastern Beaufort Sea, 1980-84.
Report by L.G.L. Ecological Research
Associates, Inc. for U.S. Minerals
Management Service, 306 p.
Richardson, W.J., Fraker, M.A.,
Wursig, B. and Wells, R.S. 1985. Be-
haviour of bowhead whales Balaena mys-
ticetus summering in the Beaufort Sea:
Reactions to industrial Activities.
Biological Conservation, 32: 195-230.
Discussion
T. ALBERT: How do you define "critical
distance?" What distance does it occur
in beluga and narwhal and under what
conditions? For example, do beluga
respond at say 30 km from an icebreaker?
S. COSENS: Finley et al. (1984) indi-
cated that narwhal and beluga in Lan-
caster Sound began showing avoidance
reactions to ships that were about 50 km

away from them. To assess whether there extreme responses in 1986 may be related
was any evidence for habituation we to both the absence of an ice edge
compared the orientation of narwhal and blocking westward movement, and the
beluga to ships greater than 50 km away presence of pack ice providing cover.
to those less than 50 km away. Alterna-
tive methods, such as correlation
analysis of the relationship between
orientation and ship distance, could also
be used.
To address your example, beluga do
show avoidance responses at 30 km from an
icebreaker. Ironically, one of our
problems in assessing beluga behaviour
was the general absence of beluga in our
surveys when ships were active at the ice
edge.
K. FROST: Did you note any obvious
response by belugas or narwhals to your
survey aircraft? In Alaska, in 1987
surveys, we noticed some apparent behav-
ioural changes when our aircraft circled
at 1000 ft.
S. COSENS: We flew our surveys at 700
ft, an altitude that Larry Dueck found,
from previous work, to be non-disruptive.
We did initially attempt to circle groups
of whales to obtain more detailed behav-
ioural observations than were possible
with straight-line passes. We found,
however, that whales reacted to the
circling helicopter, so we discontinued
our attempts to collect data in this way.
C. MALME: Did you determine the noise
levels at which whales began to show
avoidance behaviour in response to ships?
S. COSENS: We did record both underwater
ambient and ship noise. When we complete
analysis of these recordings we should be
able to estimate noise levels to which
narwhal and beluga react.
J. WARD: In 1982 when Finley started his
work and first saw the extreme escape
responses, the ice edge was just to the
west of Admiralty Inlet. Where was the
ice edge during your study?
S. COSENS: There was no landfast ice
across Lancaster Sound in our vicinity in
1986. Whales were free to continue
migrations westward at all times during
our study. There was, however, extensive
pack ice present during late May and
early June. I think that the absence of

Abstract
OBSERVATIONS OF FEEDING GRAY WHALE
RESPONSES TO CONTROLLED INDUSTRIAL NOISE EXPOSURE
Charles I. Malme
BEN Laboratories Inc., Cambridge, Massachusetts, USA
Bernd Wursig
Moss Landing Marine Laboratories, Moss Landing, California, USA
James E. Bird
University of Maryland, College Park, Maryland, USA
Peter Tyack
Woods Hole Oceanographic Institution, Woods Hole, Massachusetts, USA
A field study was conducted on the
potential effects of underwater noise
from petroleum industry activities on
feeding and summering gray whales. This
study was performed near Southeast Cape,
St. Lawrence Island, Alaska, in August
1985, using a 100 cu. in. air gun source
and an underwater projector for playback
of drillship noise. Sound source levels
and acoustic propagation losses were
measured to permit estimation of the
sound exposure levels at whale sighting
positions. The surface-dive patterns
and blow rates of whales were determined
by observation of focal groups. A
computer-aided analysis of whale
sighting data was performed to determine
patterns under pre-exposure, exposure,
and post-exposure conditions. For the
air gun source there was an 0.5
probability that the whales would stop
feeding and move away from the area when
the average pulse levels reached 173 dtf
(re 1 microPasca1). The 0.1 probability
of feeding interruption was estimated to
This is a reviewed and edited version of a paper presented
at the Ninth International Conference on Port and Ocean
Engineering Under Arctic Conditions, Fairbanks, Alaska,
USA, August 17-22, 1987. @ The Geophysical Institute,
University of Alaska, 1987.
occur at 163 dB, but whale responses
were highly variable. Most whales
returned and resumed feeding after the
air gun vessel had moved on. Playback
of drillship noise did not produce clear
evidence of disturbance or avoidance
behavior for levels below 110 dB. Pos-
sible avoidance occurred for exposure
levels approaching 119 dB.
Background
The exploration and development of
energy resources in the outer contin-
ental shelf (OCS) regions of the con-
tinental U.S. and Alaska is introducing
the factor of man-made noise into this
environment. Since gray whales
(Eschrichtius robustus) migrate and feed
throughout the western OCS regions of
the U.S. and Alaska, their activities
may potentially be affected by increased
levels of industrial noise. The objec-
tive of this study was to determine the
character and degree of response of
feeding gray whales to playback of
industrial noise (drillship sounds) and
to sound from an air gun seismic
exploration source. A complete descrip-
tion of this study can be found in Malrne
et al. (1986).

Previous studies
There have been very few controlled
experiments involving drillship play-
backs to non-migrating baleen whales.
Richardson et al. (1985a) and Richardson
et al. (1985b) found some evidence for
bowhead whale (Balaena mysticetus)
avoidance at distances of 4 to 5 ton from
the playback vessel, with the received
sound level (RSL) at the closest whales
ranging from approximately 100 dB to 113
dB. They note, however, that because of
the limited number and short duration of
the playbacks, more experiments are
needed and that their results "...must
be considered preliminary." (Richardson
et al. 1985a, p. 222.) Malme et al.
(1985) conducted two drillship playback
experiments on feeding humpback whales
(Megaptera novaeangliae) in Frederick
Sound, Alaska. There were no consistent
responses of whales at ranges to the
sound source of >0.5 km with RSL >116
dB.
Richardson et al. (1986) conducted
air gun experiments on non-migrating
bowhead whales using a single 0.66-1
Bolt air gun. During three experiments
in 1981 and 1983, involving a moving
source, they found no evidence of
avoidance at distances from 3 to 5 km
with RSL near the whales > 118 to 133
dB. In 1984, two experiments were
conducted using a stationary source.
Results showed that at 0.2 to 1.2 km and
2 to 4.5 km with RSL described as
"intense" (not measured because of
sonobuoy overload) and 124 to 131 dB,
respectively, whales moved away from the
source vessel. Malme et al. (1985)
conducted single air gun (100 cu. in.)
experiments on feeding humpback whales
in Frederick Sound, Alaska. They found
no overall pattern of avoidance, with
RSL up to 172 dB. However, observers
did note startle responses by whales at
air gun onset on three occasions with
RSL at 150 dB to 169 dB at ranges up to
3.2 km.
Migrating whale studies
Malme et al. (1983) found that dur-
ing playbacks of a variety of industrial
noise stimuli to southbound migrants,
each sound stimulus caused a statistic-
ally significant response and that each
of these responses was different when
compared to control conditions.
Patterns of response appeared to vary
predictably as a function of received
sound level. Responses generally
involved avoidance of the sound source,
based on track deflection scores for
whales exposed to playbacks of drilling
platform, helicopter and production
platform sounds and a drop in speed for
whales exposed to drilling platform,
drillship, semisubmersible and heli-
copter sounds. During drilling platform
and helicopter sound playbacks, apparent
avoidance of the source area out to
about 250 m was noted with sound levels
at this range approximately 111 to
118 dB.
During January 1984, similar
industrial noise playbacks were con-
ducted on southbound migrating gray
whales off the coast of California
(Malme et al. 1984). An analysis pro-
cedure was developed which permitted
determination of the probability of
avoidance of the region near the play-
back source. This measure showed that
avoidance behavior began at sound
exposure levels of around 110 dB for the
overall signal and was greater than 80%
for regions with signal levels higher
than 130 dB. Some variation among the
various playback stimuli was observed,
with the drillship producing the great-
est avoidance and the production plat-
form the lowest, for levels between 110
and 125 dB. However, for levels between
125 and 130 dB, the reactions to all
playback stimuli were comparable.
Malme et al. (1983) conducted
experiments with seismic exploration
sources on northward migrating mother/
calf pairs during April and May 1983
using a stationary and towed single air
gun and a 40 gun towed array. Overall,
results showed that the most predictable
responses of the whales to air gun
activity occurred at received levels
>160 dB re 1 pPa when the air gun source
was within 2 km of the animals.
Small sample sizes prevented
definite quantification of response for
average pulse pressure levels between
140 and 160 dB, but analysis showed that
some behavioral changes did occur at
these levels. In general, whales would
slow down and turn away from the source.
In several cases, groups were seen swim-

ming into the surf zone and also posi-
tioning themselves in the sound shadow
of a rock, island, or outcropping.
There were significant differences,
independent of range or level of
exposure, in milling indices ,* speed
indices for groups prior to exposure and
those same groups during exposure to the
air gun noise. There were also signifi-
cant differences in milling indices and
speed indices for groups during exposure
and after exposure to air gun noise.
During the southbound January 1984
migration, seismic experiments were
conducted using both a stationary single
air gun and a towed single air gun.
During stationary air gun experiments,
whales avoided the sound source area by
moving further offshore or inshore of
the air gun vessel. This avoidance
response was first detected 2 km north
of the vessel and persisted until the
whales were at least 2 km south of the
vessel. No identifiable avoidance
response was observed during moving air
gun experiments. However, these
experiments were of short duration and
sample sizes were low.
The probability of avoidance
analysis for the stationary air gun
source showed that the threshold of
avoidance behavior occurred for average
pulse pressure levels of approximately
164 dB. This was somewhat higher than
the level of 160 dB which was observed
to produce changes in the migration
behavior of mother/calf pairs during the
April and May 1983 field experiments.
Introduction
Based on a review of recent litera-
ture and discussions with researchers
working on feeding gray whales in the
northern Bering Sea (Wursig, et al.
1983, 1986; Thomson 1984), we decided to
conduct our studies in the nearshore
waters off Southeast Cape, St. Lawrence
Island, Alaska. The project was con-
*Milling index is a measure of the
direct- ness or linearity of the route
taken by the whale from point (xi, yl)
to (xn, yn) and is calculated by
dividing net speed by cumulative speed.
ducted in the latter half of August,
1985.
The study concentrated on the area
around St. Lawrence Island, especially
near Southeast Cape (Fig. 1). Gray
whales, apparently feeding, as evidenced
by mud plumes, were located in the area
of Kialegak Point, Southeas

generally below or near the bottom of
the surface layer of warmer, less saline
water.
Playback procedure
The acoustic playback system was
designed to provide sound levels and
frequency response capable of realistic-
ally simulating a broad range of
petroleum industry activities. In order
to boost the low frequency response of
the projector system, two USN/USRD Type
J-13 projectors were used to provide
response down to 32 Hz. In addition to
the two low frequency projectors, a
USN/USRD Type F-40 projector was used to
provide high frequency sound above 2
kHz. Electrical equalization and cross-
over networks were used to enable all of
the projectors to be driven from a 300-
watt power amplifier. As a result of
the use of two low frequency projectors
and the electronic equalization network,
the useful response of the system
extended from 32 Hz to 20 kHz. The
playback system and its response curve
are shown in Fig. 2.
160 ,
During a playback sequence, a pre-
recorded, 15-min. duration, industrial
noise stimulus on a cassette tape was
used to generate a test signal. Two
cassette recorders coupled to a fader
control permitted uninterrupted
continuous sound for as long as
desired. Playback periods of 30 min to
1 hr were generally used.
Air gun source characteristics
E COMBINED PROJECTOR RESPONSE WITH EQUALIZATION
(CONSTANT RMS 113 OCTAVE BAND LEVEL INTO EQUALIZER.
el
REF. DRIVE LEVEL. 1A RMS IN 200 Hz 113 OCTAVE BAND)
Deuda Power
C~wtto Lmd Amplifier
llydropho~
The air gun used for the seismic
signal tests was a 100 cu. in. Western
Geophysical gun operated at 4500 psi.
The peak source level was 226 dB re luPa
at 1 m (125 Hz bandwidth). A typical
pressure signature and spectrum are
shown in Fig. 3. The firing rate used
was 6 pulses per min. This gun was
operated from the NANCY-H, an 80-ft (24
m) oil-industry cargo/supply vessel.
Since the peak pressure of the
signal from the air gun is influenced
strongly by multipath propagation
conditions, we have found the average
.... --9---
FREQUENCY (Hz)
Clou-Over
Circuit
PROJECTOR Hvdrophoni
INSTRUMENTATION
Figure 2. Playback instrumentation.

.
ft.
*
a
a
: O
ft.
- 1.1
9 320
Time. m f c.
Figure 3. Air gun signature and
spectrum, 100 cu. in., 4000 psi, range
200 m, depth 10 m.
pulse pressure level to be a useful
measure of the received level of the
transient signals from an air
gun.This quantity is a measure of the
effective energy of a noise pulse in
terms of an average pressure level
defined as (Urick 1983, Sec. 4.4)
where
pc = the specific acoustic
impedance of water
p(t) = the original pulse pressure
waveform
-
p = the average pulse pressure
T = the average pulse duration
(the time required for
p 2 (t) to decay to less than
13.5% of the initial value).
Generally, it is more convenient to
express acoustic pressure in logarithmic
terms. Consequently, the average pulse
pressure level is defined as
where
Transmission loss analysis
The transmission loss data obtained
using the air gun were analyzed using a
computer-implemented least-squares
technique which determines the best-fit
values for two parameters in the
received level model (Eq. 3, Table 1).
The values of Ls' and Ar are determined
by this technique using measured data.
When the source level is calibrated, the
effect of the local bottom and surface
conditions on sound propagation can be
determined as a local "anomaly" where:
LSt = Ls + \l (dB) (4)
Here, Ls is the pressure level measured
at 1 m from the source and A is the
local anomaly resulting from bottom and
surface reflection effects.
Whale behavior observation
Whale behavior data were obtained
by close observation of focal whale
groups, recording surf acing, dive and
blow information. In addition, tracking
of the focal groups was performed using
a two-vessel triangulation procedure or
a land-based theodolite when weather
permitted. The experimental procedure
involved location of feeding whales,
observation of behavior during a control
period with the support vessels present,
observation of behavior during an
experiment period with the sound
stimulus on, and observation of behavior
during a post-experiment control
period. Generally, several of these
sequences were performed each day.
Whales were considered to be un-
disturbed during non-experimental days

Table 1.
for St.
Sound transmission parameters
Lawrence Island air gun
1986) under control
conditions: 1) Blow
and experimental
Interval - time
experiments. between respirations while the whale is
at the surface; 2) Length of Surfacing -
*B
*re **a v
time that the whale is at the surface
Dtte/Time I~B) M-.A. I ~ I discounting shallow submergences between
respirations; 3) Length of Dive - time
8/22/1443-1600 -4 17 20 that the whale is below the surface
between surfacings; 4) Number of Blows
a/22/1731-1745 2 144 20 per Surfacing; and 5) Blow Rate - the
8/24/1722-1754 -3 20 10
number of blows per minute calculated
from length of surfacing, length of dive
8/24/2015-2024 o 30 12 and number of blows per surfacing.
8/25/1221-1254. 7 54 14 We also noted if whales were
engaged in the following activities: 1)
*Determined from data using the method of least-sauares. Feeding - the presence of mud, birds
and/or regular surfacing and diving in
SOUND TRANSHISSION EQUATION the same location; 2) Travelling -
RSL - L^+A^-s log E^-1s log R-A~R-A~R/B^-~~ ( d ~
re lu?a)
concerted movement in a particular
direction; 3) Milling - movement at or
(31 near the surface accompanied by many
direction changes; 4) Socializin - two
where -~ ~
or more whales w i t h i n d l e n g t h
RSL - (7-8 m) of each other and interacting;
Received sound level at range R (dB re luPa)
and, 5) Surface Active Behavior
Ls - -
Source level [dB re 1 "Pa at 1 m)
breaching, pectoral slapping, etc.
R - Range in km
Because of small sample sizes, we were
~ y . mlecular (volumetric) absorption ( d ~
Ar - Reflection loss at surface and bottom (dB - meters per
km)
A, - Change in effective source level due to proximity of
surface and/or bottom (dB) (local anomaly).
-41 = Conversion constant (5 log 23-15 log m/km)
E - ( H + H:)/2 where Hs - fieptf. at source (n) and H:
depth at receive: la).
-
when large boats were not moving in the
study area and during the first pre-
disturbance control periods of each
experimental day. We did not consider
subsequent control periods of experi-
mental days as undisturbed for the
purposes of surfacing-dive behavior
analysis, since the data indicate that
such subsequent control periods may not
have represented a true undisturbed
situation, but instead whales were
potentially affected by the previous
experiment of that day.
To assess the possible effects of
air gun and drillship operations on the
behavior of gray whales on the feeding
grounds, we measured the following
surfacing, respiration and dive cycle
variables (after Wursig et al. 1984,
per km) unable to compare statistically the
frequency of these behaviors during
control and experimental conditions.
Measurement of whale positions and whale
movement patterns
Limited visibility conditions for
most of the field period did not permit
land-based observations. As a result,
most whale positions were ascertained by
triangulating with a shipboard theodo-
lite and binocular compasses, a tech-
nique developed by Malme et al. (1985)
to study feeding humpback whales in
Frederick Sound, Alaska. This procedure
was developed from land-based theodolite
tracking procedure (see Wursig 1978 and
Tyack 1981). The ship-based technique
requires obtaining two concurrent bear-
ings to a whale using a theodolite on a
primary observation vessel and a binoc-
ular compass on a secondary observation
vessel (a Zodiac in this case). The
range between the two observation
vessels is obtained using radar. A
Loran C system on the primary vessel
provides a geographic position refer-
ence. The primary observation vessel
for the study was the BIG VALLEY, a 90-
ft (27 m) fishing/ utility vessel. This
procedure is shown in Fig. 4.

CHALE
9
OR VESSEL
PRIMARY OBSERVATION VESSEL SECONDARY OBSERVATION VESSEL
LORANC ff.1 BINOCULAR-COMPASS I#,.#, 1
RADAR fro) RADIO
THEODOLITE W, I
RADIO
CALCULATE rw È
Figure 4. Whale tracking using observa-
tions from two vessels.
Results
Acoustic data
Ambient noise in the test area was
generally low and controlled by wind-
generated sea noise. Sound transmission
was found to be more efficient than is
usual for shallow water areas with a
sandlsilt bottom because of the probable
presence of a sub-bottom rock layer.
Measurements of received level at
several depths and ranges did not show
the depth dependence expected to be
produced by the observed strong downward
refracting gradients. This was probably
a result of the shallow water which
ranged from 15 to 25 m in depth.
Reflections and general scattering from
the bottom and probable sub-bottom
layers produced generally reverberant
received signals. While no specific
sub-bottom information has been obtained
for the St. Lawrence test area,
MacKensie (1973) reported underlying
layers of granitic and basaltic rock at
depths of 3 to 10 m for an area lying to
the east of the island.
Acoustic exposure estimation
Since some variation in sound
transmission was observed for the
several test areas used, specific data
from each test area were used in
prediction of the sound exposure levels
for whale sightings.
The results of analysis of the
transmission loss measurements are
summarized in Table 1. The values of
and Ar shown in the table were used
together with Eq. (3) to estimate the
exposure levels at the whale sighting
positions for the air gun experiments.
A comparison of measured data with the
received average pulse pressure level
versus range characteristic predicted by
Eq. (3) is shown in Fig. 5A.
Figure 5. Comparison of average pulse
pressure data with predictions of
empirical propagation propagation
model. (Source - 100 cu. in. air gun at
4500 psi).

Surfacing-dive behavior, observation
results
The frequency distributions of the
five surfacing, dive and respiration
characteristics used in the analysis are
shown in Fig. 6. Blow interval and blow
rate approximate a normal distribution,
while the distributions of the other
three characteristics are highly skewed.
Consequently, blow interval and blow
rate were analyzed with parametric
testing procedures (by analysis of
variance and Student-Newman-Keuls
multiple comparisons tests), while
length of surfacing, length of dive, and
number of blows per surfacing were
analyzed with non-parametric methods (by
Kruskal-Wallis , Mann-Whitney-U and non-
parametric multiple comparisons; Zar
1974, Sokal and Rohlf 1969).
There were significant differences
in surfacing-dive characteristics
between the condition of no known
disturbance and the potential disturb-
ances of drillship playbacks and air gun
experiments (Table 2 and Fig. 6).
During drillship playbacks, blow
interval decreased and length of surfac-
ing, length of dive and number of blows
per surfacing all increased. For air
gun sounds, the response was opposite to
that of drillship, with blow interval
increasing and the other three primary
characteristics decreasing.
Interestingly, blow rate did not change
from the undisturbed condition, because
increases or decreases in blow interval
time made up for shifts in lengths of
surfacings and dives.
Figures 7 and 8 show these summary
data in more detail. For drillship
playback experiments, the surfacing-dive
characteristics stay at a "disturbed"
level within a one-half hour period
after exposure of whales to drillship
sounds. Whales shift their surfacing-
dive characteristics close to the pre-
disturbance level in the 30 to 60 minute
period after exposure. They even appear
to overshoot the presumed undisturbed
level, with blow interval higher and the
other three primary characteristics
lower, than during the presumed un-
disturbed situation (Fig. 7). Responses
of whales to air gun do not tend to go
back to the presumably undisturbed
condition within one hour of air gun
sounds, especially for blow intervals
and length of dives. These data
indicate that air gun sounds have a
longer-term effect on the behavior of
primarily feeding gray whales than do
drillship sounds (Fig. 8). A caution is
necessary, however: drillship sounds
were made by playbacks which may have
some differences in sound character-
istics from real drillships and air gun
sounds were supplied by only one air gun
instead of the many often used during
seismic mapping activities.
Although relatively few surfacing-
dive data were collected during the
short field season, some interesting
trends have emerged. In general, blow
intervals decreased during drillship
sounds and length of surfacing, length
of dive and number of blows per surfac-
ing increased. This trend indicates
that whales are cycling through their
basic surfacing-dive patterns more
slowly while subjected to drillship
sounds. They returned to a pre-
disturbance level relatively quickly,
usually after about one-half hour post
disturbance. Blow rate altered
little. Kesponses to the air gun were
different. Whales increased blow
intervals and tended to decrease length
of surfacing, length of dive and number
of blows per surfacing. They were more
likely to alternate feeding with travel,
or travel away from the sound source.
This trend was especially strong on
several occasions when we noticed a
definite cessation of feeding and
movement away from the sound source.
Recovery to "normal" levels was less
rapid than for drillship sounds, but
tended to occur about one hour after
disturbance.
Summary of movement patterns
Drillship playback
Two playback tests for which whale
movement data are available suggest that
the whales did not alter their movement
patterns with RSL at 103 to 110 dB and
the BIG VALLEY as close as 1.1 km. In
one case, a whale continued to feed in
the same general area during both con-
trol and experimental periods. However,
during one pre-control period, whales
appeared to respond to the presence of
the BIG VALLEY, thus complicating

,.
240
192
- IB
-
0 144 -
0
-1 tT -
L -
Â¥^ 96
-
48 -
-
15 30 45
Blow Interval (a)
1
0
0 1 2 3 4
Length of Dive (min)
l l l l l l ~ ~
Length of Surfacing (mm) 72
-
-
-
Blow Rote (number/min)
T-24
se - ? 6 8
N - 489
I I 1 , l ~
x - 0..
'd - o w 2
N - 486
l I l
7
Ll ~ ~ ~ ~ ~
L
46 0 0
2 4 6 8 10
Number of Blows per Surfacing
r
12
Figure 6. Frequency distribution of surfacing-dive data on undisturbed whales. See
text for definition of undisturbed.

Table 2. Summary statistics for undisturbed whales and whales during drillship
playbacks and air gun experiments.
Experironfcal No. of Blows/ length of Surfacing length of Dive Blew Rate
Situation Blew Interval (8) Surfacing (rain) (mini (No./Min. 1
Undisturbed 14.2 6.44 811 2.4 1.68 409 0.44 0.442 406 1.80 1.158 494 1.17 0.530 480
Airgun 16.5 6.01 147 2.0 1.40 135 0.38 0.430 135 1.54 1.081 134 1.20 0.570 131
interpretation of results. During two
other playback tests, whales in the
vicinity of the BIG VALLEY did move out
of the general area, but we were unable
to obtain track data on individual
whales and, therefore, KSL for specific
focal animals are not available.
However, for one of these latter two
experiments, KSL at the whales moving
out of the area was estimated at 108 to
119 dfl at distances of approximately 1
km to 0.3 km, respectively. Results of
drillship playbacks during the present
study appear consistent with our earlier
findings .
Air gun
Alterations in whale movement pat-
terns and/or feeding behavior were noted
during each of the six air gun experi-
ments. Table 3 summarizes the behavior
of eight of the nine focal whales under
observation during the experiments.
Responses were noted at RSL ranging from
149 dB to 176 dB at distances up to
approximately 4 km. However, in one
case, RSL reached a peak of 165 db with
the NANCY H 0.7 km distant with very
little, if any, response observed. We
did observe the cessation of feeding
with apparent movement away from the
experimental vessel during air gun sound
exposure on five occasions. However, in
three of these cases, the whales resumed
feeding either during the experiment
(one case) or during the post-control
period (two cases). In the remaining
two cases, one whale stopped feeding
with apparent movement away from the
experimental vessel (Whale A, AG 3) and
continued to move out of the area during
the post-control period; the other whale
(Whale L) stopped feeding during AG 5,
but we do not have information on its
pre-control movement pattern.
Most of the responses involved
either an abrupt change in direction or
an increase in speed with apparent
movement away from the experimental
vessel. On one occasion a whale
spyhopped* several times in apparent
response to increasing RSL. We did note
that in three and possibly four cases
(marked with an asterisk in Table 3)
whales showed a response to the
operating air gun at a time coinciding
with the NANCY H moving past the whale's
position, at which point the whales were
experiencing peak RSL.
In order to derive a general guide-
line for estimating the probable
behavioral response of summering and
feeding gray whales to air gun noise, it
is necessary to examine the summary of
individual whale responses presented
previously in Table 3. On the basis of
the information presented in this table,
the summary cumulative distribution
function shown in Fig. 9 was developed.
It includes only those whales for which
a definite interruption of feeding
activity was observed. If a whale
*Raising the anterior portion of the
body so that the eyes are above the
water.

** n s ***
8 * , , I ,
a,- w-
UMXST ALL OS POST OS
UNOIST ALL OS POST OS
ns , ns , ns
a,- w-
ALL 0s POST 0s
Figure 7. Summary statistics for undisturbed whales, and whales during and after
drillship playbacks. Center bars denote means, boxes denote 95% confidence intervals,
bars denote 1 standard deviation above and below the mean, and numbers denote sample
size. Asterisks show significance levels of probability: * = 0.05, ** = 0.01, *** =
0.001, ns = not significant.

0
I *** ns
I ns
-30 nun -0 fnin
UNDIST ALL AG POST AG
UNOIST ALL AQ POST AG
6 I I I I I
UNOIST ALL AG POST AG
UNDIST ALL AG POST AG
Iff
T
1 ns , ns
-30 -on
n s
I
10.0 "."
UNDIST ALL AG POST AG
Figure 8. Summary statistics for undisturbed whales, and whales during and after air
gun experiments.
66

Table 3.
1985.
Summary of focal whale response to air gun experiments, 22, 24, 25 August
Focal Pre-Control
Mute Activity
*Itesponse when N.H. broadside to whale.
1H/V Nancy I.
21^/V ~ig Valley
Feeding Movement away, steeped novament
as AS acproadied, m feeding
Feeding Direction change, nave towards
and then away E m N.H.
feeding before and after move
Feeding, Group split at KG onset, A move
joined Mule E mrth 6 east offshore, feeding
to 1731
Feeding Feeding with inshore movement
as N.H. noved past (possibly
feeding related)
Feeding Feeding, turn, speed increase,
dive, fluke out, reeume feeding
Feeding, m Joined by Mule N, BOIB feeding,
track plot royhcpe, wtheast movement
Unknown Joined Mula L, feeding
feeding Feeding to 1605, novamerot
parallel to N.H. then offshore
resumed feeding after the air gun vessel
had moved away or stopped firing, the
corresponding original response exposure
level is marked "F".
The resulting cumulative distribution
can be seen to be somewhat skewed, having
z
u
I-
3 " -
.a -
a: .,
an interpolated median value of 173 dB and
a calculated mean value of 169.6 dB. If
the data values shown are considered to be
representative samples of the true
acoustic response statistics which might
"1
3
u
2
1-1
-
-6
"
.4 -
.3 -
be obtained with more extensive testing, .2-
Return to Pre-
Control area,
feeding
Feeding, Movement
possibly affected
by 6.V.
Continued offshore
m n t
feeding sane general
area
Feeding u n general ~
area
Group aplit, L
increase speed ming
southeast out of area,
no feeding
Feeding, group split
Offshore to 1715, then
back to original loca-
tion and feeding
o . r s . , . v - - u * q , .
it is useful to calculate the confidence
limits of the acoustic response measures
determined by the present data. We need
to estimate how well the data represent
. - 1 -
3
p
170 160 IS0
the range of expected feeding gray whale
responses to air gun noise disturbance.
RVERRGE PULSE LEVEL, dB re 1 uPa
A distribution-free confidence
interval test for the median was developed
by Thompson (19361. This test provides a Figure 9. Cumulative distribution for
means of calculating the confidence level
of a median estimate based on a number of
observed feeding disturbance. (Data from
Table 3. F - whale returned and resumed
samples from a parent population having an feeding .)

unknown distribution form. The results of
applying this test to the data shown in
Fig. 9 give a confidence estimate of 68%
that the true median (0.5) response level
lies between 170 and 175 dB and a 94%
confidence estimate that it lies within
the interval of 163 dB to 177 dB.
Discussion
Acoustic data
The transmission loss measured in the
St. Lawrence Island area was lower than
that measured off the California coast
during a previous study of migrating gray
whales (Malme et al. 1983, 1984). A
comparison of the characteristics of the
two areas for average pulse pressure
propagation can be made by examining Fig.
5A and Fig. 5B. A shallow sub-bottom
layer of rock probably causes the con-
siderably better sound propagation
conditions observed off St. Lawrence
Island since the bottom composition
according to chart information is
sand/silt for both areas.
Behavioral data
For both types of experimental
stimuli, subsequent experiments of a day
appeared to be affected by the earlier
experiments. This took both the form of
surfacing-dive data not always going back
to a pre-disturbance level after the first
experiment of the day, and whales at times
reacting less strongly to a subsequent
experiment. This is not a firm con-
clusion, however, because many other
factors such as time of day, presence of
one or two boats in the area, and general
behavior of the whales may have served as
confounding factors. Interestingly,
number of blows per surfacing, length of
surfacings, and length of dives were all
lower during the present study than for
presumed undisturbed gray whales studied
in July and September 1982 in the same
area (Wursig et al. 1986). We wonder
whether our present results may have been
affected by the presence of at least one
large vessel near the whales at almost all
times, unlike the situation in 1982, when
observations were generally made from a
small skiff > 1 km distant from the mother
ship. This possibility of a level of
disturbance even during presumed
"undisturbed" situations does not negate
our results, however, since industrial
disturbance is likely to be accompanied by
the presence of larger vessels in real
situations.
Disturbance reactions during air gun
playbacks were very similar to the reac-
tions found for surfacing-dive character-
istics of bowhead whales when subjected to
air gun sounds (Richardson et al. 1985a,
Ljungblad et al. 1985, Richardson et al.
198b). In bowheads, blow intervals in-
creased and length of surfacing, length of
dive and number of blows per surfacing all
decreased during air gun firing. The same
basic behavioral shift from feeding or
milling prior to air gun sounds to travel-
ing away from the sound source was noted
for bowheads during several experiments
with full-scale seismic vessels (Ljungblad
et al. 1985).
A possible bias exists in our data.
We used number of surfacing-dive variables
encountered to calculate degrees of free-
dom for statistical analyses, without
regard for possible dependence of data in
surfacing-dive sequences of individual
animals. If such dependence is strong,
our degrees of freedom used to calculate
statistics are overestimates of actual
allowable degrees of freedom. We have not
been able to determine amount of depend-
ence within parts of our serial data,
however, and we therefore present degrees
of freedom as directly related to sample
sizes of surfacing and dive variables.
Future work may help clarify potential
dependence of our time-series data.
Movement observations
The number of wnales observed under
experimental conditions was low throughout
the field season. This was due mainly to
the late starting date of the project,
which resulted in a low number of whales
present in the study area coupled with
adverse viewing conditions. Because many
whale groups were far offshore during much
of the field period, it was generally not
possible to use land-based theodolite
tracking of individual whales in combina-
tion with small boat observations as was
accomplished by Wursig et al. (1983,
1986). The use of this method would have
increased the number of whale groups
tracked, since land-based observers could
have concentrated on 3 to 4 groups
simultaneously, whereas the two-boat
method most often employed required BIG

VALLEY observers to focus only on the one
to two groups under observation by Zodiac
personnel in order to obtain whale move-
ment data. Movement data were obtained
during two drillship playback experiments
but the whale numbers were low. During
several air gun experiments, we have
extended detailed observations, including
both surfacing/respiration data and track
plots. As examples we describe two air
gun experiments (AG1 and AG4) for which
overall behavioral patterns are fairly
complete in Appendix A.
Conclusions
It is difficult to compare experi-
mental results concerning migrating gray
whales with those of feeding gray whales.
Different behavioral responses were
measured in feeding and migrating gray
whales. The pattern of gray whale respon-
ses may scale not only with KSL, but also
rate of change of RSL or movement of the
sound source. Both of these parameters
varied with moving vs. stationary air gun
sources. A priori one may expect the
response of gray whales to noise stimuli
to be a function of behavioral state as
has been pointed out by Brodie (1981) and
Richardson et al. (1985). However, the
results of our studies on the behavioral
responses of migrating and feeding gray
whales to drillship sound playback and air
gun operations indicate measurable
responses at similar exposure levels.
Drillship playback
Analysis of the sighting data for the
combined drillship playback experiments
showed that a number of whales were
exposed to levels that produced avuiaance
behavior for migrating gray whales (110 to
120 dB). No definite pattern of avoidance
of the source area was observed. However,
until more testing is performed at higher
exposure levels, we believe that the
application of the probability of avoid-
ance results for migration activity would
provide a conservative response estimate
for feeding activity. For the purpose of
estimating zones of influence, we will
consider that exposure of feeding gray
whales to noise levels of 110 dB or more
(from a continuous stationary source, such
as from a drillship) would result in pos-
sible avoidance of the region near the
source and exposure to levels of 120 dB or
more would probably cause avoidance of the
area by more than one-half of the gray
whales.
Air gun noise
More data on focal whales under
control and experimental conditions are
needed before firm conclusions regarding
the effects of air gun operations on
feeding gray whales can be made. The
present data set shows that feeding gray
whales can respond in a variety of ways to
a moving, single air gun and that these
responses can occur at RSL ranging from
149 dB to 176 dB, with whale distance up
to 4 km from the source.
The cumulative distribution of whales
observed to interrupt feeding activity as
a function of average pulse pressure level
shown previously in Fig. 9 was somewhat
skewed. For skewed distributions, the
median is a better estimator for the
expected value than is the mean (Zar 1974,
p. 24). Thus, an average peak pressure
level of 173 dB will be considered as the
level of air gun noise at which 50% of
feeding gray whales will probably inter-
rupt feeding activity. Based on the data
shown in Fig. 9 and on the confidence
limit calculation, 163 dA will be con-
sidered as the air gun noise level which
will probably cause 10% of feeding gray
whales to interrupt feeding activity.
Comparing these values with the prob-
ability of avoidance values obtained for
migrating gray whales, we find that a 0.1
probability of avoidance occurred for an
air gun noise level of 164 dB and a 0.5
probability of avoidance occurred for a
level of 170 dB. The acoustic sensitivity
of gray whales to air gun noise when
feeding is thus apparently not greatly
different from their sensitivity while
migrating.
Acknowledgement
This study was funded by the Minerals
Management Service through an interagency
agreement with the National Oceanic and
Atmospheric Administration, as part of the
Outer Continental Shelf finvironmental
Assessment Program.

References
Baker, C.S., L.M. Herman, B.G. Bays, and
G.B. Bauer 1983. The impact of vessel
traffic on the behavior of humpback whales
in Southeast Alaska: 1982 season. Report
from the Kewalo Basin Marine Mammal
Laboratory, University of Hawaii, Honolulu
for U.S. National Marine Fisheries
Service, Seattle, WA. 30 p. and tables
and figures.
Brodie, P.F. 1981. Energetic and
behavioural considerations with respect to
marine mammals and disturbances from
underwater noise. p. 287-290 In: N.M.
Peterson (ed.), The question of sound from
icebreaker operations: Proceedings of a
workshop. February 22-23, 1981, Toronto.
Arctic Pilot Project, Petro-Canada,
Calgary, Alberta. 350 p.
Dean, F.C., C.M. Jurasz, V.P. Palmer, C.H.
Curby, and D.L. Thomas 1985. Analysis of
humpback whale (Megaptera novaeangliae)
blow interval data Glacier Bay, Alaska,
1976-1979. Report from the University of
Alaska, Fairbanks, for U.S. National Park
Service, Anchorage, AK. 224 p. and second
volume of figures.
Harvey, J.T. and B.R. Mate 1984. Dive
characteristics and movements of radio-
tagged gray whales in San Ignacio Lagoon,
Baja California Sur, Mexico. p. 561-575
In: M.L. Jones, S.L. Swartz, and S.
Leatherwood (eds .) , The gray whale
Eschrichtius robustus. Academic Press,
Orlando, FL. 600 p.
Ljungblad, D.K. tf. wUrsig, S.L. Swartz,
and J.M. Keene 1985. Observations on the
behavior of bowhead whales (Balaena
mysticetus) in the presence of operating
seismic exploration vessels in the Alaskan
Beaufort Sea. Report from SEACO, Inc.,
San Diego for the Minerals Management
Service, Anchorage, AK. 53 p. and
appendices.
Mackensie, K.V., 1973. Long-range
propagation in the shallow Bering Sea, J.
Acoust. Soc. Am. 54(4):1066-1080.
Malme, C.I., P.R. Miles, C.W. Clark, P.
Tyack, and J.E. Bird 1983. Investigations
of the potential effects of underwater
noise from petroleum industry activities
on migrating gray whale behavior. Report
No. 5366, report prepared by tfolt Beranek
and Newman Inc., Cambridge, MA, for the
Minerals Management Service, Anchorage,
AK. Various paging.
Malme, C.I., P.R. Miles, C.W. Clark, P.
Tyack, and J.E. Bird 1984. Investigations
of the potential effects of underwater
noise from petroleum industry activities
on migrating gray whale behavior - Phase
11: January 1984 migration. Report No.
5586, report prepared by Bolt Beranehc and
Newman Inc., Cambridge, MA, for tne
Minerals Management Service, Anchorage,
AK. Various paging.
Malme, C.I., P.R. Miles, P. Tyack, C.W.
Clark, and J.E. Bird 1985. Investigation
of the potential effects of underwater
noise from petroleum industry activities
on feeding humpback whale behavior.
Report No. 5851, report prepared by BBN
Laboratories Inc., Cambridge, MA, for the
Minerals Management Service, Anchorage,
AK. Various paging.
Malme, C.I., B. w;rsig, J.E. Bird, and P.
Tyack 1986. Behavioral responses of gray
whales to industrial noise: feeding
observations and predictive modeling.
Report No. 6265, report prepared by BBN
Laboratories Inc., Cambridge, MA, for
NUAA, Anchorage AK. Various paging.
Richardson, W.J., B. w;rsig, and C.K.
Greene, Jr. 1986. Reactions of bowhead
whales, Balaena mysticetus, to seismic
exploration in the Canadian Beaufort Sea.
J. Acoust. Soc. Am. 79(4):1117-1128.
Kichardson, W.J., M.A. Fraker, B. w;rsig,
and R.S. Wells 1985. Behaviour of bowhead
whales Balaena mysticetus summering in the
Beaufort Sea: Reactions to industrial
activities. Biol. Conserv. 32(3):195-230.
Richardson, W.J., R.S. Wells, and
B. Wursig 1985. Disturbance responses of
bowheads, 1980-84. p. 89-196 In: W.J.
Richardson (ed.), Behavior, disturbance
responses and distribution of bowhead
whales Balaena mysticetus in the eastern
Beaufort Sea, 1980-84. Report from LGL
Ecol. Res. Assoc., Inc., Bryan, TX, for
the Minerals Management Service, Reston,
VA. 306 p.
Sokal, R.R. and F.J. Rohlf 1969. Biometry
- The principles and practice of
statistics in biological research. W.H.
Freeman & Co., San Francisco, CA. 776 p.

Thompson, W.R. 1936. On confidence ranges
for the median and other expectation
distributions for populations of unknown
distribution form. Ann. Math. Statist.
7: 122.
Thomson, D.H. (ed.) 1984. Feeding ecology
of gray whales (Eschrichtius robustus) in
the Chirikof Basin, summer 1982. Report
from LGL Alaska Research Associates, Inc.,
Anchorage, AK for NOAA, Juneau, AK. 222 p.
Tyack, P. 1981. Interactions between
singing humpback whales and conspecifics
nearby.
116.
~~~~~
Behav. Ecol. Sociobiol. 8(1):105-
Urick, K.J. 1983. Principles of under-
water sound for engineers, 3rd edition,
McGraw-Hill, New York. 423 p.
w;rsig, B. 1978. On the behavior and
ecology of bottlenose and dusky dolphins.
Ph.0. Dissertation, State University of
New York at Stony firook. 326 p.
wursig, B., E.M. Dorsey, M.A. Fraker, K.S.
Payne, W.J. Kichardson, and K.S. Wells
1984. Behavior of bowhead whales, Balaena
mysticetus, summering in the Beaufort Sea:
surfacing, respiration, and dive char-
acteristics. Can. J. Zool. 62(10):1910-
1921.
w

START TIME: 73000
STOP TIME : U4000 LEGEND
b =big
=
22 Aug 1985 Pre-AG 1 Control
I 3.0 4 0 5.0 6 0
Kilometers Eost
Figure 10. Track plot of whale E during
pre-AGI control on 22 August.
incides with the first appearance of the
sun all day, thereby making mud at the
water's surface more easily visible to
observers). At this time, the whale was
approximately 0.5 km northeast of its pre-
AG 1 control position. The whale stayed
in this same general area, feeding,
through the post-AG 1 control period.
Zodiac personnel noted that after the end
of AG 1, other whales were also moving
inshore to the general area of Whale E.
AG 4, 24 August, 1929-2026
Whale B was observed to be feeding
during the control period between AG 3 and
AG 4. Figure 13 shows the movement pat-
tern of this whale relative to the NANCY H
during AG 4. At the onset of AG 4, RSL at
Whale B was 159 dB with the NANCY H 1.75
km to the north. Whale B continued to
feed and between 1942-1954 was moving
slowly to the north, toward the NANCY H,
which was motoring southward. During this
period, RSL was increasing and at 1957 it
-I. 6 (
START TIME: U4000
STOP TIME : 160000
22 Aug 1985 AG 1
2.0 3.0 4 0 5 0
Kilometer; Eost
LEGEND
A =big
= non
n= e
Figure 11. Track plot of whale E during
AGl on 22 August.
had reached 176 dM with the NANCY H
0.18 Km directly offshore of Whale B. At
this point, observers noted that the whale
had turned and was moving rapidly to the
south, diving with flukes out. This was
the first time during the entire period of
observation that Whale B displayed a full
fluke out upon diving, and this action was
unusual since the whale was in shallow
water (depth < 9 m) in which fluke outs do
not normally occur. The whale continued
to move south, and at 2002 another full
fluke out was noted. At this point, RSL
had reached a peak of 177 dB with the
NANCY H 0.17 km distant. Mud was observed
with this dive, and Whale B was presumed
to be feeding. The whale continued moving
slowly to the south until approximately
2011, at which time it began to mill. By
2015, mud was again associated with Whale
B, and the whale continued to feed
throughout the remainder of AG 4, staying
in the same general location. RSL at
Whale B was decreasing during this period
and by the end of the experiment was 159
dB, with the NANCY H 1.80 tan to the south-
east of the whale's location. Whale B
continued to feed during the post AG 4
control and was last observed at 2042.

START TIME: 160000
LEGEND
STOP TIME : 171000
= non
x=
22 Aug 1985 Post-AG 1 Control
3.0 4.0 & 0 I
I 1.0 4.0 5 0
Kilometers East-
Figure 12. Track plot of whale E during
post-AGI on 22 August.
START TIME: 192900
STOP TIME : 202600
24 Aug 1985 AG 4
LEGEND
A = non
*= b
Figure 13. Track plot of whale B during
AG4 on 24 August.

Abstract
INDUSTRY OBSERVATIONS OF BOWHEAD WHALES
IN THE CANADIAN BEAUFORT SEA, 1976-1985
In 1976 Canadian Marine Drilling
Ltd. (Canmar) began exploratory
drilling in the deeper offshore
waters of the Canadian Beaufort Sea
using drillships supported by
icebreaking supply boats and
icebreakers. As part of the terms
and conditions of the approval for
this activity, Canmar personnel were
requested to record incidental
sightings of wildlife in general and
of marine mammals in particular
during the course of their offshore
exploration activities. This paper
summarizes all bowhead whale
sightings recorded by Canmar
personnel since 1976 and discusses
these sightings with respect to the
bowhead whale issue in the Canadian
Beaufort Sea. Namely, it has been
suggested that exploration activities
in the Canadian Beaufort have caused
bowheads to decrease their use of the
area where the exploration activity
has occurred. An alternative
suggestion is that the distribution
of bowheads in the Canadian Beaufort
This is a reviewed and edited version of a paper
presented at the Ninth International Conference on
Port and Ocean Engineering Under Arctic Conditions,
Fairbanks, Alaska, USA, August 17-22, 1987. @ The
Geophysical Institute, University of Alaska, 1987.
John G. Ward
E. Pessah
Dome Petroleum Limited, Calgary, Alberta, CANADA
Sea has varied from year to year
because of a varying distribution of
the zooplankton on which bowheads
feed. The Canmar incidental
sightings are examined in relation to
the first of these hypotheses and to
the various other studies that have
been carried out on bowheads in the
Canadian Beaufort Sea. It is
concluded that the incidental
sightings, together with the results
of other studies, do not support the
suggestion of a trend to decreasing
use of the exploration area by
bowheads.
Introduction
In 1976, Canadian Marine
Drilling Ltd. (Canmar) began
exploratory drilling in the deeper
(20-65 m) offshore waters of the
Canadian Beaufort Sea region (Figure
1) using drillships supported by
icebreaking supply boats and
icebreakers. Prior to 1976, all
exploratory drilling activity in the
region had been located onshore or in
nearshore waters adjacent to the
Mackenzie Delta and Tuktoyaktuk
Peninsula. The nearshore drilling
activity had been carried out from
artificial islands constructed in
waters less than 15 m deep.

Figure 1. The Canadian Beaufort Sea.
As part of the terms and conditions
of the government approval of
offshore drilling with drillships,
Canmar was requested to have their
personnel record incidental sightings
of wildlife during the course of the
exploration activities in the
Beaufort Sea. This activity was
initiated in 1976 and has been
continued each year since that time.
One species of wildlife that
occurs in the eastern Canadian
Beaufort Sea and that is of
articular interest is the bowhead
whale (Balaena mysticetus). It is
classified as an endangered species
in both Canada and the U.S. The
western Arctic population of this
species is the largest remaining
stock and is estimated to number
approximately 4,400 individuals (IWC
1986). This stock of bowheads
winters in the Bering Sea and summers
in the eastern Beaufort Sea. During
the spring and fall migration through
the Chukchi and western Beaufort seas
adjacent to Alaska, the whales are
subject to a subsistence hunt by
Alaska Eskimos.
Since 1980, systematic aerial
surveys of bowhead whale distribution
have been carried out annually in the
southeastern Beaufort Sea during the
last half of August and first half of
September (e.g. Davis et al. 1982;
Duval et al. 1986). Also, beginning
in 1980 and continuing until 1984, a
major bowhead whale behaviour study
was carried out in the southeasteri
Beaufort Sea to examine the
behavioural responses of bowheads to
oil exploration activities

(Richardson 1985). The latter study
raised a concern about exploration
activities in the Canadian Beaufort
Sea.
The issue as it presently exists
for summering bowheads in the
Canadian Beaufort Sea has been
described by Duval et al. (1986) as
follows. "Two hypotheses have been
formulated as part of the Beaufort
Environmental Monitoring Project
(BEMP) to explain the annual
variability in the distribution of
bowhead whales observed in the
southeastern Beaufort Sea since
systematic aerial surveys were
initiated in 1980 (INAC and
Environment Canada 1984, 1985). One
hypothesis suggests that activities
of the oil and gas industry have
caused, or contributed to, the
exclusion of bowheads from the
industrial zone (the lexclusion
hypothesis1). The other hypothesis
suggests that the distribution of
bowheads is determined by physical
and biological oceanographic factors,
particularly those influencing the
distribution and abundance of
zooplankton (the 'food hypothesis1).
During the most recent [I9861 BEMP
workshop addressing the bowhead whale
(INAC and Environment Canada 1987),
it was concluded that testing of
these hypotheses is unlikely to be
possible within a sound statistical
framework and, therefore, must rely
on the lweight-of-evidencel from past
and future research and monitoring
efforts directed at this species.''
The purpose of this paper is to
examine the Canmar incidental
sightings of bowhead whales in
relation to the exclusion hypothesis
and to the various studies that have
been carried out on bowheads in the
eastern Beaufort Sea.
Methods
The recording of wildlife
sightings by Canmar has varied to
some extent between years and types
of vessels. Prior to 1980, only
drillship personnel were involved in
the recording of sightings of
wildlife, whereas in 1980 and
subsequent years, personnel on both
drillships and support vessels
recorded sightings of wildlife.
Most observation activity by
personnel both on the drillships and
on the support vessels was
opportunistic in nature. However, on
the drillships, limited systematic
observations were also carried out.
Ice observers on the drillships
carried out a 10 minute wildlife
watch every four hours during a 12
hour shift, if work, weather and
light conditions permitted. When two
ice observers were on board each
drillship to provide 24 hour ice
reporting coverage, up to six watches
per day could be conducted.
Generally, only three or four such
watches were undertaken each day.
Since 1981, ice observers have
carried out either 10 or 15 minute
long watches, with the length of
watch used on each drillship being
the same throughout the drilling
season each year.
As summarized in the next
section, most whale sightings
resulted from opportunistic
observations. Because the actual
level of observational effort
expended by personnel on various
vessels in making these sightings is
unknown, there is no good comparative
measure of overall effort on a
year-to-year basis. However, it is
suggested that the number of Canmar
vessels operating in the offshore
waters each year can serve as an
approximate indicator of
observational effort made each year.
The number of vessels are shown in
Table 1.
Results
The Canmar sightings of bowhead
and unidentified whales since 1976
are summarized in Table 2 according
to whether they were reported by
personnel on the drillships or on the
various support vessels. The
sightings of unidentified whales have
been included because it is felt the
majority of these will be bowhead
rather than beluga whales. The
distinctive white or cream colour of

TABLE 1
Number of Canmar Vessels in Offshore Waters of the Eastern Beaufort Sea
VESSEL TYPE 1976 ----------
1977 1978 1979 1980 1981 1982 1983 1984 1985
Drillships 1* 3** 3 4 4 4 4 4 4 4
SupplylStandby*** - - - - 17 22 18 17 17 18
DredgesIBarge Tugs 0 0 0 0 0 4 3 5 1 0
* Three drillships were present but only one was involved in recording wildlife
observations.
** Three drillships recorded sightingsy but records were available only for one ship.
*** Personnel on these vessels did not record sightings prior to 1980.
TABLE 2
Summary of Number of Sightings (and Total Numbers seen1) of Bowhead and Unidentified
Whales Reported by Canmar Personnel during the Period August 1 - September 10 for the
Years 1976 - 1985.
Drillships Support vessels 2
Bowhead Unidentified
~otal~
Supply
Drillships Vessels
1 Numbers in parentheses are total number of whales seen for all sightings.
Data from ~ome/Canmar (1978Â 1979, l98Oy 1982Â 1984ay 1984b, 1985) and Marex
(1977).
2 Includes supply/standby vessels and dredgeslbarge tugs shown in Table 1.
3 Bowhead and unidentified whales.
4 Four sightings of bowheads in 1980 and one sighting in 1981 were made during
the systematic watches. All other sightings of both bowhead and unidentified
whales resulted from opportunistic observations.

adult beluga whales would make them
relatively easy to identify at close
range* but their comparatively small
size would make them difficult to
detect at greater distances (Norton
and Harwood 1986; Duval et al.
1986). Hence it is probable that
belugas* if seen* would generally be
identified to species. The much
larger bowheads would be easier to
detect at greater distancesy but
their black colouration would
probably create doubt as to species
for some observers if the whale was a
long distance away. Hence bowheads
are more likely to be seen at greater
distances from the vessel than beluga
whales* but such whales may be
classed often as unidentified whales.
Only sightings made during the
period August 1 to September 10 have
been included in Table 2. This time
frame was selected because it
corresponds to the one selected by
Richardson et al. (1985) to show the
zones of industrial activity in the
southeastern Beaufort Sea and the
qualitative depictions of the
abundance and distribution of
bowheads in this region for the years
1980 - 1984.
With the exception of 19773 1980
and l98ly drillship personnel
reported very few sightings of whales
each year (Table 2). On the other
hand* personnel on support vessels
reported the majority of the whale
sightings in all years since 1980
when they began reporting sightings.
However* when the differences in
numbers of drillships versus numbers
of support vessels is considered*
then the numbers of sightings are
more similar for the two types of
vessels. Like the drill ship^^
support vessels reported largest
numbers in 1980 and l98ly and lower
numbers in 1982 - 1985. However*
unlike the drill~hips~ the numbers of
sightings recorded by support vessels
in 1982-85 as compared to 1980-81 did
not show as large a decrease as that
which occurred for drillships.
Discussion
The Canmar incidental sightings
data provide a reasonable indication
of the relative abundance of bowhead
whales within the overall industrial
zone and contribute to the
"weight-of-evidence'' with respect to
the validity of the 'exclusion
hypothesis'. Before considering
these two aspects y we review the
results of studies conducted during
the two bowhead research initiatives
started in 1980 and previously
mentioned in the Introduction.
Bowhead Research Studies
The two bowhead research
programs that were started in the
eastern Beaufort Sea in 1980 have
provided a substantive systematic
data base on the relative abundance
and distribution of bowheads in the
Canadian Beaufort Sea in August and
early September y from 1980 to the
present. Richardson et al. (1985)
summarized the data from these
research studies in the form of maps
that showed all sightings of bowheads
in the southeastern Beaufort Sea
during 10-day periods from August 1
to September loy for the years 1980 -
1984. They also summarized the zones
of industrial activity on these maps
for the same periods.
Figure 2 contains composite maps
adapted from Richardson et al. (1985)
that show the zones of industrial
activity and the relative abundances
and distributions of bowheads seen
during the entire period August 1 -
September 10 for the years 1980 -
1985. They also include the
locations of drillship activity. The
map that is included for 1985 is
based on the results presented by
hval et al. (1986). The maps are
based on the technique used by
Richardson et al. (1985) to summarize
the relative abundance and
distribution of bowheads during
individual 10-day periods for the
years 1980-84. Areas of abundance
that are designated on these maps as
few* moderate and many are those
with3 respectively3 widely separated
sightings of 1-3 whales* many
sightings of 1-3 whales* and large
groups of whales (Richardson et al.
1985). The technique is highly
qualitative and is presented here

Bowhead Abundance
Many Moderate Few None
--- Industrial Zone Boundary A Canrnar Drillship Location
August 1 - September 10, 1981- -"'''
August 1 - September 10, 1982
,,w ,W' we' ,>.
Beeufo,!
-.
'=' ,,? >w
Beaufort
sea 8
August 1 - September 10, 1983
'"" UP' >w 82.' ,
8ea"rorf
sea 8
Figure 2. Relative Abundances and Distributions of Bowheads in the Eastern Beaufort
Sea, 1980 - 1985.

primarily to provide the reader with
an overall impression of where
bowheads were seen in general in the
years 1980-85 and where
concentrations (I. e. "many "1 occurred
specifically, relative to the
location of industry activity.
Richardson et al. (1985)
described the relative abundances and
distributions indicated by the survey
data from 1980 - 1984 as follows.
"Over the 1980 - 1982 period, bowhead
distribution overlapped progressively
less with the area of offshore
dredging, construction and drilling.
Bowheads were abundant within the
main industrial area in 1980, much
less abundant there in 1981, and
virtually absent in 1982. Maximum
numbers in the main industrial area
in 1983 were slightly greater than in
1982, and there was some further
increase in 1984". Richardson et al.
(1987) concluded that "utilization of
the main industrial area decreased
markedly from 1980 to 1982 and then
increased slightly from 1982 to
1983-84." On the basis of an
apparent overall downward trend from
1980 to 1984, Richardson et al.
(1985, 1987) suggested that industry
activity may be affecting utilization
of the industrial zone by bowheads.
A significant assumption of this
apparent trend is that 1980 and 1981
are representative of the historical
"normal" bowhead distribution in the
Canadian Beaufort.
Although Figure 2 indicates that
in the years 1982 - 1985 most
bowheads occurred outside the
industrial zone, localized
concentrations did occur within the
zone in 1983 and 1984. In both
years, these localized concentrations
occurred in the offshore area of
east-central Mackenzie Bay,
approximately along the 20 m depth
contour line.
Prior to 1980, limited aerial
surveys north of Kugmallit Bay in
1978 and 1979, and incidental
sightings by biologists and industry
personnel (non-Canmar) in that area
in 1976 - 1977, provide less detailed
but useful additional data.
According to Richardson et al.
(19851, "The fragmentary data from
1976 - 1979 indicate that many
bowheads were seen in the middle of
the main industrial area [i.e., north
of Kugmallit Bay] in early August of
1976 and 1977, but not in 1978 or
1979. Bowheads apparently entered
the industrial area in early
September of 1978, but in 1979 there
were very few sightings at any
time." Richardson et al. (1985,
1987) concluded that "the presence of
many whales in 1980, after a period
of apparent scarcity in 1978 - 1979,
casts doubt on the suggestion that
there is a trend for decreasing
utilization of the main industrial
area." As discussed below, the
evidence indicates that the apparent
trend is likely an artifact of the
data initially used to suggest the
trend.
Canmar Incidental Sightings
A comparison of the incidental
sightings results (Table 2) and the
systematic aerial survey results show
that the Canmar incidental sightings
for the years 1980 - 1985 provide an
indication of the relative abundance
of bowheads in the industrial zone
that is reasonably similar to that
shown by the aerial surveys (Table
3). Namely, bowheads were abundant
in the industry zone in 1980 and
1981, and relatively scarce there in
1982 - 1985. Even if the
unidentified whales, which are
suspected to be largely bowheads,
were added to Table 3, the trend does
not change. The similarity between
these two data sources provides the
confidence to use the 1976 - 1979
Canmar incidental sightings data to
make statements about the presence of
bowheads in the industrial zone in
those years (Figure 3).
The small numbers of bowhead
sightings by Canmar personnel in 1978
and 1979 (Table 2) indicate that
bowheads were not abundant in the
industrial zone during the August 1 -
September 10 period in those two
years. Richardson et al. (1985) made
the same suggestion based on other
incidental sightings and limited
survey coverage in the area
immediately north of Kugmallit Bay in

- Year
1980
1981
1982
1983
1984
1985
TABLE 3
Comparison of the Canmar Incidental Sightings
and the Systematic Aerial Surveys of Bowhead Whales
Support August Aerial survey2 September Aerial survey2
Drillship Vessel Industry Industry
sightingsl sightingsl zone 3 Entire Area zone3 Entire Area
1 Numbers in parentheses indicate total number of bowhead whales seen for all
sightings. (Values from Table 2).
2 Numbers include both bowheads seen on the transects and those seen beyond the
transects. Data taken from Renaud and Davis (19811, Davis et al. (19821, Harwood
and Ford (1983), McLaren and Davis (19851, Harwood and Borstad (19851, and Duval et
al. (1986).
3 Industry zone for each year same as one shown on Figure 2.
4 1980 survey area mostly in the industry zone.
those two years. It is expected that
drillships would have recorded more
whale sightings during the August 1 -
September 10 period in those two
years as was experienced in 1980 and
1981 if the whales had been
relatively abundant in the industrial
zone. For example, in 1979, Explorer
IV personnel recorded five sightings
of approximately 30 bowheads in total
at the Natsek site west of Herschel
Island after September 10. In that
same year and month Ljungblad (1981),
who was carrying out bowhead surveys
for the U.S. Eernment, also
recorded significant numbers of
bowheads 56 km west of Natsek. A
drillship was present at Natsek in
1978 also, from September 3 to
October 14, but no whale sightings
were reported. In 1978, the four
sightings of bowheads and
unidentified whales by Canmar
personnel (Table 2) consisted of one
sighting from the Tarsiut site and
three sightings from the Ukalerk
site. Only beluga sightings were
recorded at the Nerlerk and Kopanoar
sites (Dome/Canmar 1978). The
sightings in the Ukalerk area were
immediately offshore of Kugmallit Bay
where Fraker (1978) also reported
bowhead sightings in early September
1978. Thus the drillship sightings
in 1978 and 1979 agree with the
results of other studies in those two
years and provide additional evidence
of the relative absence of bowheads
in the industrial zone in those two
years.
It is suggested that in 1977,
bowheads were also relatively scarce
in the overall industry zone,
although unlike 1978 and 1979,
bowheads were locally abundant in the
area north of Kugmallit Bay and Toker
Point. ~ome/Canmar (1979) reported
that most whale sightings in 1977
occurred at the Ukalerk drill site
located NNE of Kugmallit Bay, and
unpublished information indicated
that bowheads were sighted in the
Ukalerk area by Canmar personnel from
August 20 to September 20. All of
this is consistent with the findings
summarized by Richardson et al.
(1985). Drillships were also located
at the Nektoralik and Kopanoar sites
in 1977 (Figure 3).
The relatively large number of

9
Auaust 1 - September 10.1976
August 1 - September 10,1978
ACanmar Dnllsh#p Location
Beauloff
ACanmar Dnllship Location
Sea k
Yukon \ .~tl\'^L\
iugust 1 - September tu, iyrf
kcanmar Dnllship Location
Beaulort
sea 8
Figure 3. Offshore Industrial Area, 1976 - 1979.
sightirigs of bowheads and
unidentified whales recorded in 1977
differed from those in 1980 and 1981
in that the sightings apparently
occurred mostly or entirely at one
site, while in 1980 and 1981 the
sightings were distributed among
three and four drillships,
respectively (Dome/Canmar 1980,
1982). This suggests that the
distribution in 1980 and 1981 was
relatively widespread in the
industrial area, while in 1977 it was
localized in the eastern corner of
the exploration area in the vicinity
of outer Kugmallit Bay and Ukalerk.
It is possible that the concentration
of bowheads in 1977 was in fact
widespread off the Tuktoyaktuk
Peninsula and that only the western
edge of the concentration overlapped
the exploration area.
In 1976, there was only one
August 1 - September 10, 1979
A Canmar Dnllship Location
124
Beaufort
SM
8W inr' ***
'*' 1M"' IMC' lie ,
Beauloft
sea i
report of 50+ unidentified whales in
two pods seen 29 km SE of Kopanoar on
August 18 by Canmar personnel
travelling on a helicopter (Marex
1977). In view of the number of
whales seen, it is possible this
sighting was of belugas rather than
bowheads. No whales were reported
from the Tingmiark location between
July 20 and August 5, which is
located immediately north of
Kugmallit Bay where Fraker (1977)
first reported bowheads on August 3.
Fraker (1977) also noted that 1976
was the first time in five years of
beluga whale studies that bowheads
had been observed within his study
area. Although less certain, it
seems likely that within the
industrial zone, bowheads were only
locally abundant in the area north of
Kugmallit Bay and Toker Point.
The reason for the relatively

large numbers of sightings by support
vessels in 1982-85 compared to
1980-81 is unknown and cannot be
explained by differences in numbers
of such vessels (Table 1). Possibly
because the support vessels move
about within the industry zone,
personnel have a greater opportunity
to encounter either small localized
concentrations of whales or transient
whales that may be present in the
industry zone for only a very brief
time.
In summary, the Canmar
incidental sightings of bowhead
whales indicate that bowhead whales
were relatively abundant and widely
distributed in the industrial zone in
1980 and 1981, and relatively scarce
in the years 1982 - 1985 and 1977 -
1979. Localized concentrations
occurred in the industrial zone in
1983 and 1984, and probably in 1977.
The situation is less clear for 1976,
but bowheads also may have been
relatively scarce in the overall
industrial zone in that year except
for a localized concentration in the
Kugmallit Bay area. The above
interpretation differs from that of
August
Richardson et al. (1985, 1987) who
equated the bowhead sightings north
of Kugmallit in 1976 and 1977 to the
widespread abundance observed in the
industrial zone in 1980. In view of
the results of the Canmar incidental
sightings, this seems unlikely.
Historical Whaling Records
Studies of whaling records from
the late 1800's and early 1900's also
provide useful information on the
distribution of bowheads relative to
the "industry zone". Townsend (1935
-
in Dalheim et al. 1980) summarized
the locations where bowheads were
taken by Yankee whalers in the
eastern Beaufort in August and
September 1848 - 1919. That summary
shown in Figure 4 indicates that most
bowheads were taken in areas approxi-
mately west of longitude 138OOO'W and
east of 131°00'W A few whales (6%)
were taken in the intervening area
which now encompasses the primary
exploration area.
Fraker and Bockstoce (1980)
analyzed the locations of both
sightings and captures of bowheads by
September
I
'LÑ. U.S.A. 1 Canada
Figure 4. Locations where bowheads were harvested by Yankee whalers in the Eastern
(Canadian) Beaufort Sea in August and September, 1848 - 1919 (adapted from
Townsend 1935).

commercial whalers in the period 1891
- 1906. Prior to 1900, August
sightings or captures of bowheads
were mainly in areas east of
longitude 131°00' or west of
138°00' as reported by Townsend
(1935). Approximately 16% of the
sightings or captures in August
occurred in what is now the main
industrial zone. After 1900, when
the stock was presumably
substantially depleted, all of the
relatively few August sightings or
captures were in the eastern Beaufort
east of 13Z0W longitude or in
Amundsen Gulf.
Thus it is interesting that the
approximate area known as the main
Industrial zone today coincides with
an area in which few bowheads were
taken in the period 1848 - 1919. It
is generally assumed that the Yankee
whalers concentrated their hunt for
bowheads in areas where they were
most likely to encounter the whales.
This then strongly suggests that
historically bowheads were also not
present in the zone in large
numbers. In light of this
information, it seems even less
likely that the exclusion hypothesis
is a valid explanation of the bowhead
distribution observed in recent years.
Conclusions
The Canmar incidental sightings
data provide further evidence towards
determination of the relative
abundance or absence of bowheads
within the industrial zone. During
the period from 1976 to 1985 the
incidental sightings data, together
with the results of other studies,
suggest that bowheads were widely
abundant in the exploration zone only
in 1980 and 1981, that localized
concentrations occurred in the
industrial zone in 1976, 1977, 1983,
and 1984, and that bowheads were
largely absent from this zone in
1978, 1979, 1982 and 1985.
Historical whaling records indicate
that in the late 1800's and early
1900's bowheads were also not
abundant in the area now known as the
main industrial area.
It is concluded, therefore, that
the available information from 1976
to 1985 and the historical whaling
information do not support the
suggestion of a trend for decreasing
utilization of the industrial zone by
bowheads as a result of oil and gas
exploration activities. As a result,
the "weight-of-evidence" suggests
that the exclusion hypothesis is
likely invalid.
References
Dalheim, M., T. Bray and H.
Braham. 1980. Vessel survey for
bowhead whales in the Bering and
Chukchi Seas, June-July 1978. Marine
Fisheries Review 42:51-57.
Davis, R.A., W.R. Koski, W.J.
Richardson, C.R. Evans and W.G.
Alliston. 1982. Distribution,
numbers and productivity of the
western Arctic stock of bowhead
whales in the eastern Beaufort Sea
and Amundsen Gulf, summer 1981.
Prep. by LGL Limited, Toronto,
Ontario. Prep. for Dome Petroleum
Limited, Calgary, Alberta, and Sohio
Alaska Petroleum Company, Anchorage,
Alaska. 135 p.
Dome/Canmar. 1978. Canmar
wildlife observation program,
Beaufort Sea - 1978 season.
Unpublished Dome/Canmar Technical
Report, Dome Petroleum Ltd., Calgary.
Dome/Canmar. 1979. Environmental
observations in the Beaufort Sea,
1979 season. Wildlife Report.
Unpublished Dome/Canmar Technical
Report, Dome Petroleum Ltd., Calgary.
Dome/Canmar. 1980. Environmental
observations in the Beaufort Sea,
1980 season. Wildlife Report.
Unpublished Dome/Canmar Technical
Report, Dome Petroleum Ltd.,
Calgary. 71p.
Dome /Canmar. 1982. Wildlife
observations in the Beaufort Sea,
1981 season. Unpublished Dome/Canmar
Technical Report, Dome Petroleum
Ltd., Calgary. 77p.
Dome/Canmar. 1984a. Wildlife

observations in the Beaufort
Sea, 1982 season. Unpublished
~omefCanmar Technical Report, Dome
Petroleum Ltd., Calgary.
72p.~ome/Canmar. 1984b.
Wildlife observations in the
Beauf ort Sea, 1983 season.
Unpublished DomefCanmar Technical
Report, Dome Petroleum Ltd.,
Calgary. 70p.
Dome / Canmar. 1985. Wildlife
observations In the Beaufort
Sea, 1984 season. Unpublished
DomefCanmar Technical Report, Dome
Petroleum Ltd., Calgary. 70p.
Duval, W.S. (ed.), L. W. Harwood,
P. Norton, J. Cubbage, J.
Calambokidis, G.A. Borstad, J.C.
Chemiawsky and R. Kerr. 1986.
Distribution, abundance and age
segregation of bowhead whales
relative to industry activities and
oceanographic features in the
southeast Beaufort Sea, August -
September 1985. Environmental
Studies Revolving Funds Report 057,
Ottawa.
Fraker, M.A. 1977. The 1977
whale monitoring program,
Mackenzie estuary, N.W.T.
Unpublished report by F.F. Slaney and
Company Ltd., Vancouver, B.C., for
Imperial Oil Ltd., Calgary.
Fraker, M.A. 1978. The 1978 whale
monitoring program, Mackenzie
Estuary, N.W.T. Unpubl. Rep. by F.F.
Slaney & Co., Vancouver, for Esso
Resources Canada Ltd., Calgary. 28 p.
Fraker, M.A. and J.R. Bockstoce.
1980. Summer distribution of
bowhead whales in the eastern
Beaufort Sea. Marine Fisheries
Review 42:57-64.
Harwood, L. and G.A. Borstad.
1985. 1984 Beaufort bowhead whale
monitoring study. Unpublished
report by ESL Environmental Sciences
Limited and G.A. Borstad Ltd. for the
Environmental Studies Revolving Fund
and Indian and Northern Affairs
Canada. ESRF 192-14-02.
Harwood, L.A. and J.K.B. Ford.
1983. Systematic aerial surveys of
bowhead whales and other marine
mammals in the southeastern Beaufort
Sea, August-September 1982.
Unpublished report by ESL
Environmental Sciences Ltd. for Dome
Petroleum Ltd. and Gulf Canada
Resources Inc.
Indian and Northern Affairs Canada
(INAC) and Environment Canada.
1984. Beaufort Sea Environmental
Monitoring Project, 1983 - 1984.
Final Report. Prepared by LGL Ltd.,
ESL Environmental Sciences Ltd. and
ESSA Ltd. 292p.
Indian and Northern Affairs
Canada (INAC) and Environment
Canada. 1985. Beauf ort
Environmental Monitoring Project,
1984-1985. Final Report. Prepared
by ESL Environmental Sciences Ltd.,
LGL Ltd., ESSA Ltd., Arctic
Laboratories Ltd. and Arctic Sciences
Ltd. 162p.
Indian and Northern Affairs Canada
(INAC) and Environment Canada.
1987. Beauf ort Environmental
Monitoring Project, 1985 - 1986 Final
Report. Environmental Studies No.
40. Prepared by LGL Ltd., ESL
Environmental Services Ltd., ESSA
Ltd., Arctic Laboratories Ltd., and
Arctic Sciences Ltd. 199p.
IWC (International Whaling
Commission). 1986. Report of the
Sub-committee on protected species
and aboriginal subsistence whaling.
Report of the International Whaling
Commission Annex H. Vol. 36: 95-111.
Ljungblad, D.K. 1981. Aerial
surveys of endangered whales in the
Beaufort Sea, Chukchi Sea and
northern Bering Sea. NOSC Technical
Document 449, Naval Ocean Systems
Center, San Diego, California. 302p.
Marex. 1977. Environmental
Observations - Beaufort Sea 1976.
Appendix 5. Wildlife Observations.
Unpublished report prepared by Marex
for Canadian Marine Drilling Ltd.,
Calgary.
McLaren, P.L. and R.A. Davis.
1985. Distribution of bowhead
whales and other marine mammals in

the southeast Beaufort Sea,
August-September 1983. Prep. by LGL
Limited, Toronto, Ontario. Prep. for
Environmental Studies Revolving
Funds. COGLA-ESRF-001. 62 p.
Norton, P. and L.A. Harwood.
1986. Distribution, abundance and
behaviour of white whales in the
Mackenzie Estuary. Environmental
Studies Revolving Funds, Report No.
036. Ottawa. 73p.
Renaud, W.E. and R.A. Davis.
1981. Aerial surveys of bowhead
whales and other marine mammals off
the Tuktoyaktuk Peninsula, N.W.T.,
August- September 1980. Prep. by LGL
Limited, Toronto, Ontario. Prep. for
Dome Petroleum Limited, Calgary,
Alberta. 55 p.
Richardson, W.J.
Behavior, disturbance responses
and distribution of bowhead whales
Balaena mysticetus in the eastern
Beaufort Sea, 1980-84. Unpubl. Rep.
from LGL Ecol. Res. Assoc., Inc.,
Bryan, Texas for U. S. Minerals
Management Service, Reston, VA. 306p.
Richardson, W.J., R.A. Davis, C.R.
Evans and P. Norton. 1985.
Distribution . of bowheads and
industrial activity, 1980-84. p.
255-306 In: W. J. Richardson (ed. 1,
Behavior, disturbance responses and
distribution of bowhead whales
Balaena mysticetus in the eastern
Beaufort Sea, 1980-84. Unpublished
Report from LGL Ecological Research
Associates, Inc. for U.S. Minerals
Management Service, Reston, VA. 306p.
Richardson, W.J., R.A. Davis, C.R.
Evans, D.K. Ljungblad and P.
Norton. 1987. Surimer distribution
of bowhead whales, Balaena
mysticetus, relative to oil industry
activities in the Canadian Beaufort
Sea, 1980 - 84. Arctic 40(2): 93 -
104.
Townsend, C.H. 1935. The
distribution of certain whales as
shown by logbook records of American
whaleships. Zoologica (N.Y.119, 50p.
Discussion
T. ALBERT: Is it appropriate to infer
bowhead distribution from catch location
data if there is no indication of the
level of effort? For example, were the
various areas "searched" in a similar
manner? Using catch per unit of effort
(CPUE) seems OK but just using catch
location data does not seem appropriate.
J. WARD: Yes, I believe it is appro-
priate to infer bowhead distribution from
catch location data with the appropriate
qualifiers. As we indicated in the
paper, we have made the assumption that
commerical whalers would have concen-
trated their efforts in areas where they
were most likely to encounter bowheads.
During the commercial whaling period, I
would be surprised if whalers did not
look for whales in the area now known as
the industrial zone as well as the other
areas where they did take many whales.
The fact that relatively few whales were
taken from the area off the Mackenzie
Delta suggests to me that even then few
whales used the "industrial zone" area.
D. DICKENS: I am interested to know
whether you have attempted to correlate
your observations with ice severity both
in terms of the extent of open water in
the so-called "Industrial Zone" as we11
as the availability of relatively clear
passage via leads and shore openings
along the Alaskan coast.
J. WARD: The short answer is "no." Such
an analysis would not be useful because
of the limited nature of the incidental
sightings. Other studies provide some
information on the subject of ice condi-
tions and bowhead whale distribution.
These include the whale surveys carried
out for Minera1.s Management Service in
the Alaska Beaufort Sea, the whale survey
carried out in the Canadian Beaufort Sea
for government and industry, and a 1985
study of the effects of oceanographic
features on bowheads carried out for the
Canadian Environmental Studies Revolving
Fund.

T. NEWBURY: You have described inciden-
tal sightings from vessels. The data is
important because it shows that some
whales come close to vessels. However,
your extrapolation with the incidental
sightings to statements about broad-scale
bowhead distribution is not correct. The
main reason is that you have made no
comparable effort to collect incidental
sightings away from vessels. You do
compare the number of incidental sight-
ings with the number of sightings during
broad-scale aerial surveys, but the
correlation is very weak. Further, there
are no estimates of effort for the aerial
surveys, so the correlation between
incidental sightings and aerial sightings
may be simply due to changing levels of
effort. To summarize, the incidental
sightings show that whales come
close to vessels, but the incidental
sightings are not useful for analyses of
changes in broad-scale distribution.
J. WARD: I belive the incidental sight-
ings do provide useful information on
changes in distribution of bowheads
within the industrial zone, as I have
described in this paper. Your use of the
term "broad-scale bowhead distribution" I
assume means bowhead distribution within
the entire southeast Beaufort Sea. We
have not used the ship-based incidental
sightings to extrapolate beyond the
industrial zone as suggested by your
question. With respect to your comment
on level of effort, aerial survey effort
for the surveys used in the comparisons
has been relatively similar from year to
year with the exception of 1980 when the
area surveyed was quite limited. Inciden-
tal observation effort will have varied
with the level of exploration activity
and the location of wellsites. Neverthe-
less, there is an apparent correlation
between the two types of observations in
year-to-year relative numbers of bow-
heads. The results of the incidental
sightings parallel those of the aerial
surveys. Perhaps more importantly, the
incidental sightings provide additional
information on the distribution of
bowheads in the industrial zone in the
years prior to 1980 when broader scale
offshore aerial surveys were first
carried out.
In summary, I agree that the incidental
sightings show that some whales come
close to vessels, but I disagree with
your comment that they are not useful for
analyses of changes in distribution.
When combined with the results of other
studies, the incidental sightings do
assist in providing qualitative insights
into bowhead distribution.

Abstract
MASKED DETECTION THRESHOLDS FOR THE BELUGA
AND BOTTLENOSE DOLPHIN
A beluga and an Atlantic bottlenose
dolphin were trained to detect a sphere
target with varying amounts of masking
noise at three distances. Target
detection performance as a function of
masking noise level was determined at
each target range for both species. The
Echo-to-Noise ratio (Ee/No)max for the
beluga was approximately 10 dB better
than the dolphin for all ranges. The
difference in performance between the
two species may be due to critical
bandwidth, signal processing capability,
or a different echolocation strategy.
Variations in the animal detection
performance across the three ranges were
consistent with target strength and
transmission loss differences.
Introduction
There has been little effort to
compare the echolocation capabilities of
different cetacean species in the same
task (Nachtigall, 1980). Differences in
procedures and experimental design have
made cross- species comparisons
This is a reviewed and edited version of apaperpresented
at the Ninth International Conference on Port and Ocean
Engineering Under Arctic Conditions, Fairbanks, Alaska,
USA, August 17-22, 1987.
Charles W. Turl
Ralph H. Penner
W. W. L. Au
Naval Ocean System Center, Kailua, Hawaii, USA
difficult. Au et al. (1986) reported
that there are some differences in the
echolocation systems of the beluga and
Tursiops. They reported that the
transmitted beam of the beluga is
narrower than the dolphin, but that
their echolocation signals had similar
wave-shapes and frequency spectra (Au et
al., 1985). The beluga emitted higher
peak frequencies and higher intensity
signals in Kaneohe Bay than in San Diego
Bay. Au et al. (1974) suggested that
bottlenose dolphins changed their
signals in order to compensate for the
higher ambient noise in Kaneohe Bay.
Penner et al. (1986) suggested that the
beluga used a narrower beam to minimize
the effects of a noise masker by
receiving target echoes via a surface
reflected path while echolocating. The
objective of this study was to directly
compare a beluga and a bottlenose
dolphin using the same experimental
apparatus, procedure, targets and
masking noise.
Methods
The experiment was conducted in
Kaneohe Bay, Hawaii using an adult male
beluga (Dl 575) and an adult male
bottlenose dolphin (Tt 8). The experi-
mental configuration and procedure used
were similar to those of Au and Penner
(1981). The beluga, stationed in a 40 cm

Figure 1. Experimental configuration
with the beluga in the hoop station.
(The rubber mat is not shown in the
figure) .
dia hoop, is depicted in Figure 1.
Targets were thin-walled stainless
steel, water-filled spheres with
diameters of 7.62 cm (target strength
-28.3 dB) and 22.86 cm (target strength
-17.4 dB). Target ranges of 16.5 and 40
m were used with the 7.62 cm sphere and
80 m with the 22.86 cm. The center of
the hoop station and the target depth
were both at 1 m.
Masking noise was projected from an
Edo-Western 6166 spherical transducer
initially located, in line with the
target, 4 m from the hoop for the 16.5 m
target range. For the 40 m and 80 m
target ranges, the noise hydrophone was
located 5 m from the hoop. The spectrum
of the masking noise measured at the
hoop was relatively flat from 40 to 160
kHz (Figure 2). A rubber mat extended
approximately 0.3 m into the water and
attached to the end of the pen directly
between the hoop and the noise projector
to block surface-reflected paths (see
figure 3 of Penner et al., 1986).
TARGET
CLICK
NOISE
HYDROPHONE
I .-'t
Echolocation signals of both
animals were measured with and Apple I1
microprocessor system described by Au et
al. (1982). A click hydrophone (Edo-
Western 6166), located 2 m from the hoop
station, was used to detect each click.
The results of the echolocation signal
measurements obtained in this experiment
are presented in Turl et al. (1987).
A trial started when the animal
was stationed in front of the
experimenter, with the acoustic screen
in the raised position, the target out
of the water, and the masking noise
turned off. An underwater signal cued
the animal to turn and swim across the
pen and insert its head into the
stationing hoop. The target was either
gently lowered into the water or left
out, the masking noise was turned on and
the acoustic screen was lowered. The
animal could echolocate for as long as
it desired. When the animal finished
echolocating, it backed out of the hoop
and responded by striking one of two
response paddles to indicate whether or
not it detected the target.
Six ten-trial blocks with an equal
number of randomized target present and
absent trials comprised a session. Five

0 100 200
FREQUENCY (kHz)
Figure 2. Frequency spectrum of the
masking noise used in this experiment.
noise levels (No) in 3 dB increments
were used during testing. The received
masking noise levels at the hoop station
are shown in Figure 2. The first and
last 10-trial blocks were always
conducted at the minimum noise level.
The noise levels of the other blocks
were randomized for each session. Each
animal participated in one session per
day; the animal which participated in
the first session would be used in the
second session of the following day and
vice versa.
Results
The target detection performance of
each animal, as a function of the
masking noise level for each distance,
is shown in Figure 3. The beluga's
performance exceeded the bottlenose
dolphin by 8 to 13 dB a t all three test
distances. The beluga's 75% correct
response thresholds occurred at n ise
levels of 85, 72 and 63 dB re 1 Pa /Hz
for target ranges of 16.5, 40 and 80 m,
respectively. The dolphin's correspond-
ing 75% correct response threshold
occurred at 72, 59 and 55 dB, respec-
tively. For both animals, response
accuracy decreased as the noise level
increased.
Large fluctuations in the animal's
emitted source levels usually occurred
5
 BELUGA
BOTTLENOSE DOLPHIN
-
40 50 60 70
MASKING NOISE SPECTRUM LEVEL
(dB re 1 pPa2/Hz)
Figure 3. Beluga and bottlenose
dolphin's performance data as a function
of the masking noise level at 16.5, 40
and 80 meters. The target range (R) and
spher size at each distance is shown in
each figure.

BELUGA BOTTLENOSED
DOLPHIN
Figure 4. Beluga and bottlenose
dolphin's performance data at three
distances plotted as a function of the
echo signal-to-noise ratio.
in most click trains with typical
variations of 15 dB between the minimum
and maximum signal levels. Since it is
not known what signal levels the animal
used to detect targets, we used the
maximum level of each signal in the
trial in order to calculate the echo
energy flux density (Ee) returning to
the animal (e.g. Au and Penner, 1981).
Such a technique will provide the upper
limit of the E/N, or the maximum E ~ /
No available to the animal.
The performance data from Figure 3
is replotted (Figure 4) as the echo-
signal-to-noise ratio ((Ee/No)max).
Also shown are the two Tursiops
measurements of Au and Penner (1981 ).
At the 75% correct response threshold
the (Ee/NOlmax was approxi-
@ 16.5 M
40.0 M
@ 80.0 M
 AU & PENNER
mately 1.0 dB for the beluga at the
three target distances. For the
bottlenose dolphin, (Ee/No)max was
approximate 10 dB.
Discussion and conclusions
Our results clearly indicate that
the beluga's detection performance in
masking noise was superior to that of
the dolphin. Differences in the
animals' projected signal amplitudes
cannot explain the performance
differences, since the relative emission
signal amplitudes of both animals were
similar. If the beluga had a narrower
critical bandwidth at frequencies
between 100 and 120 kHz, the amount of
noise it received would be less than
that of the dolphin and this could
possibly explain the difference in
comparative detection performance.
Another possibility which might
affect the difference in detection
performance could be the echolocation

emission rate strategies used by each
animal (Turl et al., 1987). The beluga
emitted more clicks per trial than the
dolphin and emitted clicks at a higher
rate. This indicates that the number of
echoes available for processing to the
beluga exceeded that available to the
dolphin by a factor from 1.25 to 2 and
suggests that the beluga may be
processing more information per unit
time than the dolphin. Since the
animals were not constrained to
echolocate in any particular manner, one
must wonder why the dolphin did not also
increase its repetition rate and use
more clicks if that would be to its
advantage. Given differences in
habitats and prey it is likely that
different species of toothed whales have
evolved different capabilities and use
different strategies depending on
environmental differences.
In this study masked detection
thresholds of a beluga were 8 to 13 dB
better than that of a bottlenose
dolphin. Possible explanations for the
difference in the detection sensitivity
between the beluga and bottlenose
dolphin may include differences of
critical bandwidth, echolocation
emission rate strategies and signal
processing capabilities. We believe
that it is unlikely that any one of
these reasons could singly account for
the differences in performance; more
likely, a combination of factors are
involved. The beluga lives in an
arctic-subarctic environment while the
bottlenose dolphin inhabits more
temperate-tropical waters, but both
species tend to inhabit shallow coastal
water and both feed on a variety of prey
organisms. Maybe the beluga's superior
performance in the described experiment
represents an adaptation to an arctic
ice-covered environment with occasional
high noise levels. Further investi-
gations are required in order to isolate
which mechanisms are relevant to the
beluga's enhanced performance as
compared to the dolphin in a high noise
echolocation detection task.
References
Au, W.W.L., Floyd, R.W., Penner, R.H.,
and Murchison, A.E. 1974. Propagation
measurements of echolocation signals of
the Atlantic bottlenose dolphin,
Tursiops truncatus Montagu, in open
water. J. Acoust. g. &. 56:1180-
1290.
Au, W.W.L., Floyd, R.W. and Haun, J.E.
1978. Propagation of Atlantic bottle-
nose dolphin echolocation signals. J.
Acoust. SOC. Amer. 64:411-422.
---
Au, W.W.L. 1980. Echolocation signals
of the Atlantic bottlenose dolphin
(Tursiops truncatus) in open waters.
-
In: "Animal Sonar Systems" (R.G. Busnel
and J.F. Fish, eds.). Plenum Press, New
York, 251-282.
Au, W.W.L. and Penner, R.H. 1981.
Target detection in noise bv Atlantic
bottlenose dolphins. L. Acoust. g.
Amer. 70:687-693.
Au, W.W.L., Penner, R.H. and Kadane, J.
1982. Acoustic behavior of echolocating
Atlantic bottlenose dolphin. L. Acoust.
SOC. Amer. 71:1269-1275.
--
Au, W.W.L., Penner, R.H. and Turl, C.W.
1987. Propagation of beluga whale
echolocation signals. L. Acoust. x.
&. (in press).
Au, W.W.L., Carder, D.A., Penner, R.H.
and Scronce, B.L. 1985. Demonstration
of adaptation in beluga whale
echolocation signals. J. Acoust. z.
Amer. 77:726-730.
-
Nachtigall, P.E. 1980. Odontocete
echolocation performance on object size,
shape and material. In "Animal Sonar
Systems" (R.G. ~usneland J.F. Fish,
eds.). Plenum press, New York, 71-95.
Penner, R.H., Turl, C.W. and Au, W.W.L.
1985. Target detection by the beluga
using a surface-reflected path. J.
Acoust. SOC. Amer. 80(6): 1842-1843.
---
Turl, C.W., Penner, R.H. and Au, Whitlow
W. L. 1987. Comparison of target
detection capabilities of the belutra " and
bottlenose dolphin. J. Acoust. g.
(in press).
e.

Discussion
T. ALBERT: Have any other dolphins or
toothed whales been looked at which
receive and send at the same time?
W. TURL: The other dolphins which have
been tested behave similarly to the
bottlenose dolphin.

EVIDENCE OF GLACIAL SEISMIC EVENTS IN THE ACOUSTIC ENVIRONMENT OF
HUMPBACK WHALES
Abstract
High level impulses of underwater
sound occurring as frequently as 3-4
times per minute were detected in
Glacier Bay, Alaska during a measurement
project to define the underwater
acoustic environment in that region.
Analysis of these impulses demonstrates
signal-to-noise ratios as high as 40 dB,
significant broadband energy from below
20 Hz to above 2 kHz and the presence of
pure tone components. The character of
the impulses is relatively invariant
within the Bay area.
We hypothesize that these events
are associated with the many active
glaciers in the region and are the
result of stick-slip action generating
seismic energy at the ice-rock inter-
face. That energy is then propagated
through the bedrock and radiated as
sound into the water through the walls
and/or bottom of the fjords. Acoustic
instrumentation included a sonobuoy and
a hydrophone separated by about 3.2 km.
Analysis of simultaneously recorded data
provides directional information relat-
ing to the source of the events. Data
This is a reviewed and edited version of a paper submit-
ted to the Ninth International Conference on Port and
Ocean Engineering Under Arctic Conditions, Fairbanks,
Alaska, USA, August 17-22, 1987.
Paul R. Miles
Charles I. Maime
BBN Laboratories Inc., Cambridge, Massachusetts, USA
are presented for this phenomenon,
relating it to findings of others who
used geophones on and near glaciers in
other regions.
Background and Hypothesis
This paper resulted from a larger
study, administered by the Marine Mammal
Laboratory of the National Marine
Fisheries Service, to measure and define
the underwater acoustic environment of
Glacier Bay National Park in Southeast
Alaska during the summer of 1981. Major
funding for the overall study (Malme, et
al. 1982) originated with the National
Park Service which had become concerned
regarding fluctuations in the population
of feeding humpback whales in the area.
Potential detrimental influence on hump-
back whales of man-made noise in Glacier
Bay, especially noise due to increased
traffic from cruise ships, small boats
and aircraft, was one driving force in
establishment of the project.
The purposes of the acoustic study,
therefore, were to:
1) determine the natural and man-
made contributions to the ambient noise,
with emphasis on areas frequented by
humpback whales,
2) examine sound propagation and
reverberation properties of representa-

tive areas, such as Frederick Sound/
Stephens Passage for comparison to
conditions in Glacier Bay,
3) obtain radiated noise spectra
from ships, small boats, and aircraft
operating in these areas,
4) examine any unusual acoustic
differences which may exist between the
two field locations.
With respect to the latter element
of the study, two natural phenomena were
observed in Glacier Bay which were not
detected in the Frederick Sound/Stephens
Passage area:
1) frequency high-level acoustic
impulsive events were recorded which had
strong low-frequency components and
which were varying in event rate from
day-to-day, possibly on a diurnal basis,
2) very high level effervescent
sound levels were measured in the
vicinity of tidewater glaciers and
floating glacial ice.
The first phenomenon is the subject of
this paper, although a brief commentary
on the effervescence phenomenon will
also be included.
We hypothesize that the high-level
impulsive underwater sound events, which
occur as frequently as 3-4 times per
minute, are associated with the many
glaciers in the region and are the
result of stick-slip action generating
seismic energy at the icelrock
interface.
Physical Environment
The characterization of the
acoustic environment of Glacier Bay
required specific measurements to be
made throughout the region as well as
reference to historical data. Figure 1
summarizes the 38 stations which were
occupied in July and early August 1981
and provides information regarding the
Glacier Bay regional characteristics.
The acoustic data reported here were
acquired at Station 27-27A and in Queen
Inlet (Station 36).
The dynamic glaciological aspects
of Glacier Bay are readily evident when
observing the many active glaciers and
icefields of the region and the fjords
making up the various inlets. Depths in
the steep-walled fjords vary from about
50 meters to 430 meters. The entire
region of the present Park was covered
with ice as recently as 1794 when
Vancouver sailed by the mouth of the
bay. The retreat of the glaciers since
that time has been rapid with the rate
varying somewhat over the last
200 years. Retreat rates of 400 meters/
year, and greater have been recorded
(Muir Glacier has recently shown retreat
rates as high as about 2 kml~ear!).
Some advancing has also been observed as
well as some relative inactivity as a
glacier stalls on a sill. The release
of the weight of the ice from the land
mass has resulted in isostatic rebound
rates as high as about 4 cm per year in
the Bartlett Cove area and an overall
average of about 2 cm/year for the Park
area (Hale and Wright, 1979).
The regional geology of the area is
complex and has been summarized by
MacKevett et al. (1971). Bedrock
stratigraphic age varies from Early
Silurian through Late Tertiary with
granitic intrusives and metamorphic
formations. Foliated granitic bedrock
such as diorite and granodiorite and
detrital clastic rocks such as gray-
whacke and shale are major rock types
with common zones of limestone and
volcanic rocks. Faulting trends to the
northwest with the Fairweather Fault
being the major fault zone located along
the coast west of the Fairweather moun-
tains. Other fault zones are located
along the West Arm of Glacier Bay, along
the east shore of the Park and with some
cross-faulting in the Tlingit Point-
Tidal Inlet area.
The U.S. Geological Survey com-
pleted an extensive seismic survey of
the bottom and subbottom topography of
Glacier Bay (B. Molnia, personal com-
munication). The USGS records reveal
the presence of a series of sills and
basins throughout the arms and inlets.
The terminal moraine of the original
glacier at the entrance to the Park
provides a hard rocky botto'm that is
continually swept clean of silt by tidal

Figure
currents (as high as 7 knots in the
narrows). The bottom there and over
other sills is acoustically reflective.
The basins, on the other hand, have been
filled with fine silty deposits varying
in thickness from 50 to 150 m. Figure 2
provides sketches of the sediment layers
at several positions within the bay. In
these areas, the bottom is acoustically
absorptive. The sides of the arms and
inlets are steep and rocky with no
acoustically important sediments.
Water temperature and salinity as a
function of depth and area within the
bay area were measured primarily because
these data are required in the deriva-
tion of sound speed profiles to provide
insight into sound propagation char-
acteristics. Increased surface tempera-
tures due to high insolation rates and
reduced salinity due to melting glacial
ice were observed. Temperature and
salinity profiles up-bay stabilize at
depths greater than about 30 meters. At
the southern area of the bay where con-
siderable mixing occurs, the profiles

ICr0.s - Chtnnell
UPPER WEST ARM NEAR
QUEEN AND RENDU INLETS
i
ISWI INEI
(WEST1 (EAST!
0 1 2 3nmt
LOWER MUIR INLETQADAMS INLET
APPROX 3 MILESNORTH OF
WILLOUGHBV ISLAND
Ref: B. Molnia. USGS. Unoublished
Note: There is an approximate 30: 1 EAST OF WILLOUGHBV ISLAND
wrtical/horizontal display distortion.
Reflector 1
Reflector 2
ISWI INEI
Figure 2. Rough sketches of sediment thicknesses in Glacier Bay.
are stable to the surface. Sound speed
profiles resulting from Wilson's
Equation,
(where T is water temperature (OC) and S
is salinity in parts per thousand) are
shown in Figure 3, demonstrating the
stability of the physical characteris-
tics of the water column below 30
meters, in particular. The measurement
system used on this project did not
permit measurements deeper than 50
meters although historical data demon-
strate a repeatable profile with a
slightly negative gradient of 0.06 m/sec
per meter of depth down to 200 m depth.
Sound generated from a source located at
a depth greater than 30 meters will be
propagated with a slightly downward
refracting ray path angle.
directional general purpose hydrophone
system located at the research vessel
and an SSQ57A sonobuoy system with the
radio receiver located on the research
vessel.
The general purpose system was bat-
tery operated permitting complete shut-
down of the research vessel propul-
sion system and accessories to minimize
the possibility of self-contamination of
acoustic data. The H56 hydrophone,
owned and calibrated by the U.S. Navy,
has a frequency range of from 10 Hz to
65 kHz and is flat from 10 Hz to 20 kHz.
The data recorder used was a two channel
NAGRA IV SJ stereo recorder with an
additional channel for voice annotation.
The frequency response was flat from
0.02 to 20 kHz. The system was used for
acquisition of vessel radiated noise
signatures and for obtaining short term
background noise measurements.
Instrumentation The synoptic ambient noise measure-
ment system used an SSQ57A sonobuoy.
Two acoustic measurement systems This permitted the acquisition of 8
were used as shown in Figure 4; an omni- hours of continuous ambient noise data

-(STA N0.461811
?30 - -(SA No.28) 7/26
I
fc
w
'40-
50 -
60 I
JOHNS HOPKINS INLET-
SOUND SPEED PROFILES IN GLACIER M Y
Figure 3. Sound speed profiles in Glacier Bay.
u
GENERAL PURPOSE ACOUSTIC MEASUREMENT SVSTEM
-v RÇouw
SYMOTTIC AMBIENT NOISE MEASUREMENT SVSTEM
DATA KDUCTKMI BYBTEM
Figure 4. Acoustic measurement and data
reduction systems.
I I 1 I / .
using an r-f link from a remotely
emplaced buoy with the data transmitted
to the research vessel, located up to
four miles away. Data were acquired
with this system in the 0.01 to 20 kHz
band. Synoptic data from the sonobuoy
were tape recorded for five minutes
every half-hour for the duration of the
8 hour period. Continuous data were
recorded on a strip chart recorder for
the full 8-hour period.
Figure 4 also outlines the data
analysis system consisting of a Nicolet
446A narrowband spectrum analyzer pro-
viding data in the form of underwater
sound pressure level as a function of
frequency and a graphic level recorder
which provides a continuous plot of log
amplitude of the acoustic signal as a
function of time.
Acoustic Impulse Events
The acoustic impulse events were
heard and recorded throughout Glacier
Bay as long as the ambient noise levels
permitted detection. The signals were
heard in near-glacier regions, such as
Queen Inlet and Geikie Inlet as well as
in areas 'which -were relatively remote
from glaciers such as the Marble Islands
and Bartlett Cove. The Bartlett Cove
events usually could be heard only at

2 TYPICAL SEISMIC EVENT SIGNATURES NEAR N. MARBLE ISLAND
(log AMPLITUDE) (712111 150)
.
~
- . - ~
-. - . - 4 .. ~ - - . ~.
.
~-
I
-. _1 ... - -~
-- .
EXAMPLE OF SYNOPTIC AMBIENT log AMPLITUDE VS TIME RECORD, STATION 27,7124
Figure 5. Typical seismic event signatures and an example of synoptic ambient vs
time record, Station 27, 7/24/81.
night when ambient levels were very
low. The events with their typical
booming impulsive sound with a tail-off
in sound level were also heard in 1982
at Point Adolphus, across Icy Strait
from the Glacier Bay entrance and about
24 n.m. or 44 km to Brady Glacier (the
nearest). Pt. Adolphus is an Important
feeding area for humpback whales.
Figure 5 provides two strip chart
recordings of such events. The lower
record is a 401ninute continuous record
of the events as recorded by the sono-
buoy system on 24 July 1981 in the early
afternoon. Event rates as high as 4-5
transients per minute are evident. At
other times and days the event rate
could be as low as one event every 2 or
3 minutes. The upper figure, obtained
on a high speed and high pen response
recorder demonstrates the most common
characteristics of these events. That
is, signal-to-noise ratios as high as
40 dB at the onset of the fast rise-
time events were frequently observed
with a tail-off of sound level lasting
as long as 7 seconds. A pure tone 63 Hz
component, usually occurring about 1.4
seconds after first arrival, was detect-
ed frequently. Occasional emergent
impulses were also seen, peaking in
level at about 3 seconds from first
arrival and diminishing in level into
the background after about another
3 seconds. Figure 6 shows similar data
obtained in Queen Inlet on two other
days, which is further up-bay than
Station 27 in the Marble Islands. The
nearest tidewater glacier to Station 27
is about 45 km away. In Queen Inlet,
the nearest tidewater glacier is 16 km

QUEEN INLET, 7/16/1935, FREQUENT HIGH LEVEL EVENTS
QUEEN INLET, 7/29/1130, SPORADIC EVENTS WITH TONALS
Figure 6. Sporadic impulsive events with tonals recorded in Queen Inlet, 7/29/81 at
1130.
away and the closest valley glacier is
9 km away. Notice in particular that
the peak levels of the impulses received
at the two sites are very similar and in
the order of 115 dB//uPa and that the
general character of the signatures is
the same. Also note the pure tone com-
ponent in the Queen Inlet data and the
similarity of the overall signature with
that obtained at Station 27, 18 miles or
33 km down-bay. One difference between
the signatures from the two sites is the
duration of the signatures. The more
distant site (Figure 5) showed 6-7 sec.
event duration and the Queen Inlet
events in Figure 6 are 4-4.5 sec. long.
These differences possibly may be due to
reverberation but also probably relate
to differences in travel time of differ-
ent seismic wave types and dispersion
effects. One common qualitative element
in all of these events is that immedi-
ately following reception of peak
sound pressure level, there is a "roll-
ing thunder" type sound which varies in
intensity but reduces in level until it
is lost in the ambient background.
Figure 7 presents three ambient
noise spectra measured in different
areas of the bay. One signature of an
impulsive event is shown (Queen Inlet).
The upper curve (STA 24) demonstrates
the loudness of the effervescent noise

@-",e~
1
STA. 24 MUIR GLACIER
--
STA %QUEEN INLET
1 0 0 2 0 0 3 0 0 4 < 1 0 5 0 0 à ‡ O O T O O a h à ˆ 0 0
FREQUENCY IHzl
Figure 7. Upper Glacier Bay ambient noise (low frequency).
referred to earlier. Levels are as much
as 10 dB noisier than the broadband
sound typically experienced in the open
ocean for a sea state 6 condition. SS6
corresponds to sustained 28-33 knot
winds and 17 to 26 ft wave heights.
Clearly, the broadband ice sound has the
capability of masking significant other
sources of acoustic energy, except at
frequencies below about 150 Hz. We
established through observation and
others have established, such as Dewart
(1968), that the sound is due to multi-
tudes of small bubbles frozen into the
glacial ice that have been compressed
over time and then released with a
popping or frying sound as the water
melts away the surface of the ice. The
lowest curve in the figure shows a
typical spectrum under the quietest
condition obtained.
Figures 8 and 9 are narrowband
spectra of two impulsive events obtained
near North Marble Island (near Sta 27)
and in Queen Inlet, respectively.
Notice in particular the low frequency
tonal components at about 63, 126, and
189 Hz and good signal-tonoise ratio
beyond 1 kHz (Fig. 8) and out to 2 kHz
in Fig. 10. The cause of the pure tone
components in the transients is unknown
although, we feel it may be a resonance
effect associated with the generation of
shear energy at the ice/ rock interface.
One brief experiment was performed
at Station 27 with the hope of providing
some insight regarding the source of the
impulsive events. The H56 hydrophone
was deployed to a 25 meter depth at the
research vessel, located approximately
3.2 km away from the sonobuoy and the
impulsive events recorded simultaneously
with the sonobuoy data on a 2-channel
tape recorder. Range to the sonobuoy
was determined acoustically by dropping
a weighted light bulb over the side of
the research vessel where it sank until
hydrostatic pressure caused the bulb to
implode. The resulting sharp sound
transient was recorded from both the H56
hydrophone and the sonobuoy. Measuring
the time difference and knowing the
speed of sound permits determination of
range between the two sensors. Figure
10 sketches the arrangement showing that
a time difference of 2.2 seconds was
realized, corresponding to a sensor
spacing of 3234 meters (using a sound
speed of 1470 m/sec). Similarly, single
impulsive events are recorded

-
100
. 2
"isa
3
Ã
80
w
?
E 70
h
'g m
a
E
g so
a BACKGROUND
0 0.2 0.4 0.8 0.8 1.0 1.2 1.4 1.6 1.8 2.0
FREQUENCY (kHz1
Figure 8. Narrowband spectrum analysis of an underwater impulsive event at
N. Marble Island, Glacier Bay.
-
s
3
2
? -
loo
-i m
>
w ¥
8"
h
w
^ so
40 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.8 1.8 2.0
FREQUENCY IkHzl
Figure 9. Narrowband spectrum analysis of an underwater impulsive event in Queen
Inlet, Glacier Bay.
103

an extra ambiguity for broadside
arrivals.
la1 HYDROPHONE SEPARATION CALIBRATION
1 (BULB SOURCE)
Based on the arrival directions
shown in Fig. 11, we feel that the most
likely source of these events is
associated with the glaciers in the
Brady Ice Field complex to the west
of the measurement site. However, we
cannot totally rule out White Glacier
and Casement Glacier to the northeast
because of the small sample of data.
I (bl IMPULSIVE EVENT FIRST ARRIVALS
I
SONOBUOY
Figure 10. Determination of direction
of arrival for impulsive events.
simultaneously. Also sketched in Figure
10 is the first arrival of such an
event, demonstrating that the sonobuoy
registered the event first. The cosine
of the wavefront arrival angle (with an
ambiguity) is the ratio of the arrival
times . Figure 11 illustrates that
ambiguity and includes with it a poten-
tial angular error of about 5' due to
imprecise information on the azimuthal
orientation of the research vessel/
sonobuoy line of sight, and to some
fluctuation in the time differences in
the first arrivals of the impulsive
events at the two hydrophones. Of the
events measured in this way, all showed
arrival time differences to be within
1.3 to 1.4 seconds. If seismic impulses
were radiating from the bottom of the
fjord, arrivals at the two hydrophones
would be nearly simultaneous, providing
Discussion
-At-
L
0
1.4m-
~ ~ ~ ~ ~
05 1.0 1.5 2 0 2.5
ARRIVAL TIME DIFFERENCE (SECONDS)
~
H 56
~ I
3 0
~
We have stated that we feel that
these events are related to seismic
energy being generated at the icelrock
interface ~ g beneath ~ ~ active a glaciers ~ l of ~ the ~
region. We do not have information
regarding the various degrees of
~
activity of the many glaciers, only that
1cl GEOMETRY
all of the tidewater glaciers were
demonstrating activity. In the West Arm
LINEARIZED IMPULSE
area, Reid, Lamplugh, and Johns Hopkins
glaciers are all fed by the Brady Ice
Field. Charpentier and Geike, which are
now valley glaciers, are located close
to the fjords. In the vicinity of
Station 27, all the nearby potential
seismic sources are valley glaciers;
White, Adams and Morse, and Casement to
the north and northeast.
Dewart (1968) has measured
impulsive events using seismometers on
and near the ice of Kaskawulsh Glacier
in the Yukon. Impulse rates varying
from 1 to 62 per hour were measured and
were quite short in duration (0.5 to
2 seconds). Further, he noted a diurnal
variation in event rate with changes in
numbers of events per hour similar to
what we observed. While he did not
mention observing pure tones in associa-
tion with these events, he does refer to
work by Oelsner on the Lovenbreen
Glacier in West Spitzbergen where he
observed impulses (which he calls
-eigenimpulses") having a dominant fre-
quency around 60 Hz. He also relates
the event rate to temperature fluctua-
tions, which correlate with run off
fluctuations (lubrication).
Weaver and Malone (1979) discuss
extensive seismometer measurements made
in the immediate vicinity of ,glaciers in
the North Cascade mountains of

 SHORT AMBIENT. SVP,
Figure 11. Approximate angle of arrival of impulsive events at Station 27.
Washington. They measured many impulsive
events having a 1-2 Hz dominant fre-
quency and event rates which varied from
thee per day to three per minute depend-
ing on the particular glacier being
monitored. At one station (Longmire) on
Mt. Rainier, they reported that through
a 10-year period, consistent correlation
existed between event rate and time of
year; low rates in winter and high rates
In the summer when more run-off occurs,
permitting more glacial motion through
lubrication. Magnitudes of the events
based on the earthquake energy relation-
ship developed by Gutenberg and Richter
in Richter (1958), varied from 1.2 to
2.4 representing seismic energies at th
source of these events of 1 0 to 10 5
Joules. They report further that only
the very largest surface avalanches
produced a relatively small response at
stations a few hundred meters away.
Crevassing impulses, as referenced by
Weaver and Malone (1979) and reported by
Neave and Savage (1970) generate in the
order of 1 Joule of seismic energy.

Based on their extensive study, Weaver
and Malone conclude that the impulses
which they monitored were due to stick-
slip action through a shear sliding
mechanism at the ice/rock interface.
Van Wormer and Berg (1973) report
that weak P phases and well developed
monochromatic non-dispersive shear-wave
trains are often recorded at University
of Alaska's seismological station SCM in
southern Alaska. They conclude that
these events probably originate at or
near tidewater glaciers, in Prince
William Sound. Frequencies of the shear
wave trains in this case were in the
order of 1 to 2 Hz. Earthquake magni-
tudes of 1.6 to 3.0 were recorded and
event rates at the remote measurement
facility were in the 2.5 to 3 per hour
range. They hypothesize that these
events are associated with release of
seismic energy at the ice/rock interface
in the order of 107.5 Joules. The
authors compute that a stick-slip motion
of about 3 mm of a glacial section hav-
ing an area of 1 km 2 is sufficient to
generate the earthquake magnitudes
observed.
If we assume that the pure tone
elements of the impulsive events
measured in Glacier Bay are due to
shear-wave energy generated at the
ice/rock interface we have a means for
estimating the focal distance to the
events. Shear waves do not propagate in
water, but they can be converted to
compressional waves at major changes in
impedance or at major discontinuities.
Measuring the time difference between
the first arrival and the pure tone
component of the signatures in Figure 5
provides a P-S arrival difference of
approximately 1.3 seconds. Jeffreys and
Bullen (1958) provide a table of seismic
wave arrival times for surf ace focus,
near earthquake events. Their data for
compressional and shear wave arrivals in
granite bedrock, demonstrate that the
distance to a seismic event, through the
rock medium, for a time difference of
1.3 seconds is approximately 11.1 km or
6.0 n.m. Referring back to Figure 11
and assuming that the radiating
rock/water interface of a fjord wall is
where the shear-to-compressional energy
conversion takes place and that the
radiating face is approximately at the
bar mark on the 50Â angle of arrival
line, a 6 n.m. distance to the originat-
ing seismic event is at the end of the
arrow shown. In this estimate, there-
fore, it appears that the source of the
impulsive events measured at Station 27
is associated with the Reid Glacier,
near Its eastern edge. Similarly, for
the ambiguous arrival angle to the
northeast, the range vector indicates
that any source in that direction could
not be associated with a glacier. While
we do not have directivity information
for the Queen Inlet events, we do have a
P-S At estimate from Figure 6 (at the
bottom) of approximately 1.1 seconds
which translates, according to Jeffreys
and Bullen, to a distance of 9.4 km or
about 5 n.m. A probable source of the
Queen Inlet events, theref ore, is
Carroll Glacier, immediately to the
north of the measurement site.
With the knowledge of mean sound
pressure levels of these events as
received at Station 27, it is possible
to develop a rough estimate of the
seismic energy released at the source.
Our sound transmission loss measurements
demonstrated a 15 log (range) dependency
for spreading loss in water and Ewing,
Jardetsky and Press (1957) indicate a
similar dependency for body waves (shear
and compressional) propagating over
short distances for surface focus
events.
Sound pressure is related to
radiated acoustic power (W) by the
relationship,
where r is the distance from the source
in meters, p is the r.m.s. acoustic
sound pressure in pascals, and p and c
are the density and sound speed of
water. Radiated acoustic power at the
water/rock interface for a sound
pressure at the research vessel of 100
dB//uPa 18 0.17 watts based on computing
the sound pressure at the radiating rock
face 13 n.m. away and using the above
relationship. Sound pressure at the
water/rock interface is estimated from

where po
research
pressure
is the sound pressure at the
vessel and p is the predicted
at the rock face r meters away.
For 110 dB//pPa, the acoustic power at
the rock face is 1.7 watts. Predicted
energy at the source of the seismic
event is:
w
Eb = 10 log(;) + 15 log R,
(dB//l watt), (for body waves)
where 11 is the rock-to-water energy con-
version efficiency and R is the distance
in meters through the rock material to
the source. Converting these dB values
to power and multiplying by the seismic
impulse duration in seconds provides the
watt-sec impulse power at the source.
Energy in the more conventional seismic
units (joules or ergs) is obtained by
application of standard unit conversion
factors. We have assumed that solid
friction losses in rock for this case
are small compared to spreading losses.
Further, if we assume a 5% seismic
energy-to-underwater sound energy con-
version efficiency and that the 6 n.m.
rock transmission path length is reason-
able, we compute the event energies and
earthquake magnitudes shown in the
following table. The earthquake energy/
magnitude relationship used for shallow
focus local events and body waves is
that provided by Richter (19581,
log E = 9.9 + 1.92 M -0.024M
L L
where E is energy in ergs and ML is
local earthquake magnitude.
Source Energy Estimates*
Mean SPL at
Hydrophone
(Initial
Impulse for
1 second
duration) 100 dB//pPa 110 dB//pPa
Body Waves:
Energy 4 x 10 Joules 4 x 1 0 Joules
Earthquake
Magnitude M = 1.9 M = 2.5
*See text for glacially related seismic
event magnitude estimates by others.
Conclusions
Seismic events (primarily body
waves) due to stick-slip action at the
ice/rock interface in glaciers in
Glacier Bay probably are the cause of
underwater acoustic impulsive events
which often occur as frequently as 3 to
4 times per minute and may have a
diurnal and seasonal variation in event
rate.
The question regarding the possi-
bility of a microseismic cause of these
impulsive events due to isostatic
rebound and resulting fracturing and
slippage in existing fault zones is not
resolved. A wintertime measurement
period may help to answer this question
since it is expected that stick-slip
action of glaciers would tend to be
suppressed in the winter and rebound
rate would remain relatively unchanged.
More synoptic measurements should
be performed using both a hydrophone
array and a seismometer array with
simultaneous recording to help remove
assumptions and ambiguities which were
used in this analysis.
References
Dewart, G. 1968, "Seismic Investigation
of Ice Properties and Bedrock Topography
at the Confluence of Two Glaciers,
Kaskawulsh Glacier, Yukon Territory,
Canada," Institute of Polar Studies,
Report No. 27.
Ewing, W.M., W.S. Jardetsky, and F.
Press. 1957, Elastic Waves in Layered
Media, McGraw Hill.
Hale, L.Z. and R.G. Wright. 1979, "The
Glacier Bay Marine Ecosystem; A
Conceptual ecological Model," National
Park Service.
Jeffreys, H. and K.E. Bullen. 1958,
Seismological Tables, British
Association for Advancement of Science.
MacKevett, E.M., Jr., D.A. Brew, C.C.
Hawley, L.C. Huff, J.G. Smith. 1971,
"Mineral Resources Glacier Bay National
Monument, Alaska," Geological Survey

Professional Paper #632, U.S. Geological
Survey.
Malme, C.I., P.R. Miles, P.T. McElroy.
1982, "The Acoustic Environment of
Humpback Whales in Glacier Bay and
Frederick SoundIStephens Passage,
Alaska," BBN Report No. 4848.
Neave, K.G. and J.C. Savage. 1970,
"Icequakes on the Athabasca Glacier, "
Jour. Geoph. Res. (75), No. 8, pp. 1351-
1362.
Richter, C.F. 1958, Elementary
Seismology, W.H. Freeman Co.
Vanwormer, D. and E. Berg. 1973,
"Seismic Evidence for Glacier Motion,"
Journal of Glaciology (12), No. 65, pp.
259-265.
Weaver, C.S. and S.D. Malone. 1979,
"Seismic evidence for Discrete Glacier
Motion at the Rock-Ice Interface,"
Journal of Glaciology (23), No. 89, pp.
171-184.

Following presentation of the
technical papers, symposium speakers
participated in a panel discussion on
potential directions for research rela-
tive to noise and marine mammals.
Suggested topics for additional
research (and the name of the panelist
or audience member initiating each
topic) are summarized as follows:
1. Investigation of the hearing and
vocalization of seals and small ceta-
ceans relative to frequency thresholds
and band widths (J. Terhune) ;
2. Selection of a particular species
(e.g., ringed seals) from which data on
ecological effects may be extrapolated
to other species (B. Kelly);
3. Tracking of a large whale species
(e.g., humpback whale) to test acoustic
responses in a manner similar to work
performed by Sam Ridgeway (W. Turl);
4. Further analysis of bowhead whale
distribution data to test the theory
that whales may be excluded from
particular industrial zones (Tom
Newbury, Minerals Management Service
(MMS) ) ;
5. Use of synoptic monitoring systems
(without observers) to collect data on
the physical acoustics of particular
areas (C. Malme);
6. Consideration of the basic behav-
ioral responses of marine mammals to
long-term noise (S. Cosens); and
7. Radio-tagging large cetaceans to
the extent that this is feasible (Tom
Albert, North Slope Borough (NSB)).
The merits of selecting a particular
key species for extrapolation of
certain results to other species
(suggestion 2 above) were discussed
further. Major points made during this
discussion were the following:
PANEL DISCUSSION
Stephen D. Treacy
Minerals Management Services, Anchorage, Alaska, USA
a. A scientific benefit derives
from long-term research continuity with
a single species (K. Frost).
b. Key species should be select-
ed based to some extent on the ease by
which information can be collected and
on their wide geographic distribution
(S. Cosens). Key species should be
selected based, in part, on the extent
of existing good data (Tom Newbury,
MMS). Existing data, however, may be
poor and may require a great amount of
initial attention (B. Kelly).
c. An overall problem with the
key species approach is that, at least
in the case of Tursiops sp. where
long-term data exist over 20 years, the
data would not extrapolate to other
species (W. Turl).
d. The key species approach may
reflect economic reality . In marine
mammal science (B. Kelly).
e. Any key species had better be
highly similar for data extrapolation
to apply (W. Turl). Extrapolation is
difficult even from area to area for
the same species CJ. Ward).
One qualifying statement on the topic
of bowhead whale exclusionary zones
(suggestion 4 above) was that analyses
of bowhead distribution data showing
areas where impact (or lack of impact)
occurred should indicate the degree of
variability. The accuracy of statis-
tical tests used will determine the use
of such results (Tom Albert, NSB).
Major points made about radio-tagging
of marine mammals (suggestion 7 above)
were:
a. Radio-tagging is an excellent
approach: it provides reams of data
quickly on difficult management and
scientific issues (J. Terhune).

. Radio-tagging is expensive
and provides a small sample of individ-
ual animals tagged (B. Kelly).
c. Radio-tagging of polar bears
has helped greatly in determining off-
shore denning of polar bears (J. Ward).
This is because the polar bear is one
species that does have a large percent-
age of tagged individuals (B. Kelly).

AUTHOR LIST
Au.W.W.L ........................
Bird . J . E . ........................
Burns. J . J . .......................
89
55
27
Cosens. S . ........................ 39
Cowles. C . J . ...................... 1
Dueck. L . P ........................ 39
Frost. K . J . ........................ 15
Imm. J . L . ........................ 1
Kelley. B . P . ....................... 27
Lowry. L . F . ....................... 15
Malme. C . I . ....................... 55:95
Miles. P . R . ....................... 95
Penner.R.H . ...................... 89
Pessah . E ......................... 75
Quakenbush. L . T . ................ 27
Terhune. J . M ...................... 9
Treacy. S . D . ................... 109
Turl. C . W . ....................... 89
Tyack. P . ......................... 55
Ward. J . G . ....................... 75
Wursig. B . ..................... 55
Ill