for .

dr-e

Wrque
"---/ cANAD1AN BULLETIN OF FISHERIES AND AQUATIC SCIENCES 224
n,n_ Library / MPO - Bibeèque

dvances
in
Résearch
ON THE
bieluga\;Vhale,
DELPHINAPTERUS LEUCAS

EDITED BY
T.G. SMITH • D.J. ST. AUBIN • J.R. GERACI

CanadU
 ,5
'7
e?
Canadian Bulletin of Fisheries and Aquatic Sciences 224 if‘rpocif.

Advances in Research
on the
Beluga whale 9 0C-oko.
Delph i•n apterus eae

Pe°
Edited by

T. G. Smith
Department of Fisheries and Oceans, Arctic Biological Station
555 Blvd. St-Pierre, Ste-Anne-de-Bellevue, Quebec, Canada H9X 3R4

D.J. St. Aubin and J. R. Geraci

Wildlife Section, Department of Pathology, Ontario Veterinary College,
University of Guelph, Guelph, Ontario, Canada NIG 2W!

Scientific Excellence
Resource Protection & Conservation
Benefits for Canadians

•DEPARTMENT OF FISHERIES AND OCEANS

Ottawa 1990
 The Canadian Bulletins of Fisheries and Aquatic Sciences are designed to interpret current
knowledge in scientific fields pertinent to Canadian fisheries and aquatic environments.
The Canadian Journal of Fisheriesand Aquatic Sciences is published in annual volumes
of monthly issues. Canadian Special Publications of Fisheries and Aquatic Sciences are
issued periodically. These series are available from authorized bookstore agents and other
bookstores, or you may send your prepaid order to the Canada Communication Group —
Publishing, Supply and Services Canada, Ottawa, Ont. K 1 A 0S9. Make cheques or money
order payable in Canadian funds to the Receiver General for Canada.
Communications Directorate
Acting Director General: John Camp
Editorial and Publishing Services: Gerald J. Neville

Think Recycling!

Penser à recycler !

Printed on
recycled paper

© Minister of Supply and Services Canada 1990

Available from authorized bookstore agents, other bookstores
or you may send your prepaid order to the
Canada Communication Group — Publishing
Supply and Services Canada, Ottawa, Ont. K IA 0S9.
Make cheques or money orders payable in Canadian funds
to the Receiver General for Canada.
A deposit copy of this publication is also available
for reference in public libraries across Canada.
Cat. No. Fs 94-224E ISBN 0-660-13817-4
DFO/4426 ISSN 0706-6503

Cover photograph courtesy of Jack Orr

This publication has been peer reviewed.

Correct citation for this publication:

SMITH T. G., D. J. ST. AUBIN, AND J. R. GERACI [ed.]. 1990. Advances in research on the
beluga whale, Delphinapterus leucas. Can. Bull. Fish. Aquat. Sci. 224: 206 p.
 Contents
Abstract/Résumé v—vi
Research on beluga whales, Delphinapterus leucas: introduction and
overview. T. G. Smith, D. J. St. Aubin, and J. R. Geraci 1-6
Evolution and systematics of the beluga whale, Delphinapterus leucas, and
other odontocetes: a molecular approach. D. W. Lint, J. W. Clayton,
W. R. Li/lie, and L. Postma 7-22
The distribution and abundance of belugas, Delphinapterus leucas, in
eastern Canadian subarctic waters: a review and update. P. R. Richard,
J. R. Orr, and D. G. Barber 23-38
Distribution, abundance, and movements of beluga whales, Delphinapterus
leucas, in coastal waters of western Alaska. K. J. Frost and
L. F. Lowry 39-57
Age-length and length-weight comparisons in the beluga, Delphinapterus
leucas. D. W. Doidge 59-68
Philopatry and site tenacity of belugas, Delphinapterus leucas, hunted
by the Inuit at the Nastapoka estuary, eastern Hudson Bay. L. M. J.
Caron and T. G. Smith 69-79
Cutaneous response to implants, tags, and marks in beluga whales,
Delphinapterus leucas, and bottlenose dolphins, Tumiops truncatus.
J. R. Geraci and G. J. D. Smith 81-95
Reactions of belugas, Delphinapterus leucas, and narwhals, Monodon
monoceros, to ice-breaking ships in the Canadian high Arctic.
K. J. Finley, G. W. Miller, R. A. Davis and C. R. Greene 97-117
Echolocation abilities of the beluga, Delphinapterus leucas: a review
and comparison with the bottlenose dolphin, Tursiops truncatus.
C. W Turl 119-128
Integumentary heat loss and blubber distribution in the beluga,
Delphinapterus leucas, with comparisons to the narwhal, Monodon
monoceros. D. W. Doidge 129-140
Dynamics of tooth growth in beluga whales, Delphinapterus leucas, and
effectiveness of tetracycline as a marker for age determination.
P. F. Brodie, J. R. Geraci, and D. J. St. Aubin 141-148
Adrenal responsiveness to stimulation by adrenocorticotropic hormone
(ACTH) in captive beluga whales, Delphinapterus leucas. D. J.
St. Aubin and J. R. Geraci 149-157

III
Assessment of immunological function in captive beluga whales,
' Delphinapterus leucas: humoral response to sheep red blood cell
antigens. D. J. St. Aubin, J. R. Geraci and P. E. Shelve,' 159-164
Organochlorine contaminants in belugas, Delphinapterus leucas, from
Canadian waters. D. C. G. Muir, C. A. Ford, R. E. A. Stewart,
T. G. Smith, R. F. Addison, M. E. Zinck and P. Béland 165-190
Heavy metals and selenium in tissues of beluga whales, Delphinapterus
leucas, from the Canadian Arctic and the St. Lawrence estuary.
R. Wagemann, R. E. A. Stewart, P. Béland and C. Desjardins 191-206

iv
 Abstract

SMITH, T. G., D. J. ST. AUBIN, AND J. R. GERACI [ed.]. 1990. Advances in research on the beluga
whale, Delphinapterus leucas. Can. Bull. Fish. Aquat. Sci. 224: 206 p.

This collection of 14 articles represents a cross-section of current research on beluga

whales, Delphinapterus leucas, in North American waters.
The evolution and taxonomic status of belugas was examined using immunological
characteristics of serum albumin and electrophoretic properties of enzymes, to establish the
beluga's relationship to other small odontocetes. The findings confirm morphological studies
linking the beluga and the narwhal as the sole members of the Monodontidae.
Aerial surveys of belugas in the eastern Canadian Arctic and Alaskan waters yielded
population estimates for stocks identified by their summer distribution in these regions.
However, stock discreteness has not been established conclusively; morphometric analysis
of animals from Hudson Bay and the St. Lawrence River showed that even these isolated
populations cannot be distinguished on the basis of length—weight and age—length relation-
ships. As an aid to identifying and tracking the movements of individual whales, a number
of marking and tagging techniques were evaluated in a study on captive animals. Observa-
tions on belugas occupying the Nastapoka River reaffirmed the strong fidelity that whales
have for specific estuaries, despite disturbance and hunting pressure.
Belugas are highly social animals, and employ an elaborate vocal repertoire for commu-
nication. Comparison of their echolocating abilities with those of the bottlenose dolphin
shows the beluga's sense to be specially adapted to deal with an ice-infested habitat. Incursion
of ice-breaking ships into that environment was shown to affect herd behavior and vocaliza-
tion.
Like other cetaceans, belugas rely on blubber for insulation, but an analysis of heat transfer
through skin and blubber did not reveal any specific adaptations associated with their
seasonal occupation of warm estuaries.
Three studies were undertaken on captive whales. Tetracycline effectively marked grow-
ing teeth, and provided some insight into the deposition of calcium in dentinal formations.
Stimulation with adrenocorticotropic hormone (ACTH) elicited the characteristically strong
aldosterone response seen in other marine mammals. A distinct humoral response to red
blood cell antigen confirmed the value of this technique as a test of immune function.
Measurements of tissue levels of heavy metals and chlorinated hydrocarbon residues in
belugas from various Arctic stocks and from the St. Lawrence River revealed regional
differences, and provide important data for monitoring the accumulation of contaminants
from the environment.
 Résumé

G., D. J. ST. AUBIN, AND J. R. GERACI [ed.]. 1990. Advances in research on the beluga
SMITH, T.
whale, Delphinapterus leucas. Can. Bull. Fish. Aquat. Sci. 224: 206 p.

Cette série de quatorze articles donne un aperçu des recherches en cours sur le béluga,
Delphinapterus leucas, des eaux de l'Amérique du Nord.
L'évolution et la taxonomie de bélugas ont été étudiées par examen des caractéristiques
immunologiques de l'albumine sérique et des propriétés électrophorétiques de certains
enzymes dans le but de déterminer leurs rapports avec d'autres petits odontocètes. Les
résultats obtenus confirment ceux d'études morphologiques selon lesquelles le béluga et le
narval sont les seuls membres des Monodontidés.
Des dénombrements aériens des bélugas dans l'est de l'Arctique canadien et les eaux de
l'Alaska ont permis d'estimer les effectifs de stocks que l'on avait délimités dans ces régions
à partir de leur répartition d'été. Le caractère discret des stocks n'a cependant pu être établi
de façon concluante et des analyses morphométriques d'animaux provenant de la baie
d'Hudson et du fleuve Saint-Laurent ont montré que même ces populations isolées ne
pouvaient être différenciées à partir des relations longueur—poids et âge—longueur. Diverses
techniques de marquage et d'étiquetage ont été évaluées dans le cadre d'une étude portant sur
des animaux en captivité dans le but de trouver des moyens d'identifier et de suivre les
déplacements de certains individus. Des observations des bélugas de la rivière Nastapoka ont
permis de reconfirmer la forte tendance des baleines à occuper certains estuaires en dépit des
perturbations et des pressions de chasse.
Les bélugas sont des animaux à comportement social important qui communiquent en
utilisant un répertoire vocal élaboré. La comparaison de la capacité d'écholocation du béluga
avec celle du dauphin à gros nez a montré que le béluga était particulièrement adapté aux
habitats engorgés de glaces. Il a été montré que le passage de brise-glaces à cet endroit avait
pour effet de modifier le comportement du groupe et la vocalisation.
Comme les autres cétacés, les bélugas possèdent une couche de graisse qui leur sert
d'isolant, mais une étude du transfert de chaleur au niveau de la peau et de la graisse n'a pas
permis de noter une adaptation particulière à la présence en estuaires à eau plus chaude.
Trois études ont été réalisées sur des baleines captives. La tétracycline a permis de marquer
la croissance dentaire de sorte que certaines connaissances ont pu être acquises sur le dépôt
de calcium dans les formations de dentine. La stimulation par l'hormone adrénocorticotrope
(ACTH) a provoqué une forte réponse d'aldostérone caractéristique chez les autres mammi-
fères marins. Une réponse humorale caractéristique à l'antigène des érythrocytes a confirmé
la valeur de cette technique en tant que test de la fonction immunitaire.
La détermination des teneurs tissulaires en métaux lourds et hydrocarbures chlorés chez
des bélugas de différents stocks de l'Arctique et du fleuve Saint-Laurent a montré l'existence
d'écarts entre les régions et constitue un important moyen de contrôler l'accumulation des
contaminants présents dans le milieu.

vi
 Research on Beluga Whales, Delphinapterus leucas:
Introduction and Overview

T. G. Smith
Department of Fisheries and Oceans, Arctic Biological Station
555 Blvd. St-Pierre, Ste-Anne-de-Bellevue, Quebec, Canada H9X 324

D. J. St. Aubin and J. R. Geraci

Wildlife Section, Department of Pathology, Ontario Veterinary College,
University of Guelph, Guelph, Ontario, Canada NIG 2W]

Introduction

Beluga whales have long been in the eye of the public and continue to be a subject of
research. They among all cetaceans have the unusual habit of congregating close to shore on
a predictable schedule, thereby making them accessible to people. Perhaps for this reason,
belugas were one of the first cetaceans kept in captivity (Defran and Pryor 1980), and today
are popularly displayed in aquaria throughout the world.
Their timed movements into estuaries also made the whales easy prey for shore-based
fisheries, principally in Russia, Svalbard (Spitsbergen) and throughout the northern territo-
ries occupied by the Inuit people. Some of the catches were substantial and made available
large numbers of specimens for studies into the biology of the species. Vladykov (1944)
assembled one of the first treatises on belugas in the Gulf of St. Lawrence, and Kleinenberg
et al. (1964) did the same for the population in Soviet waters. In the Canadian Arctic,
Sergeant's (1973) data from commercial hunts in Hudson Bay and Brodie's (1971) from Inuit
subsistence catches on Baffin Island helped to provide information needed to manage the
harvests.
While hunting is still an important issue in some parts of the Arctic, other human activities,
such as those associated with offshore oil development (Smith et al. 1979), and contaminants
(Martineau et al. 1987) are demanding the attention of scientists and resource managers. This
focus continues to gain strength.
Presently the literature on beluga comprises over 600 articles touching every aspect of
beluga biology (Smith and Hammill 1990). This monograph offers 14 more, with no less
diversity. Population assessment and management issues feature prominently, as expected
from our continuing need to monitor stocks of whales. Also represented are studies on
behavior, acoustics, molecular systematics, growth, metabolism and thermoregulation,
endocrine function, tissue contaminant burdens, and the immune system - disparate topics
that come together to shape our current understanding of the beluga whale.

Systematics and Evolution

The beluga is the only representative of the genus Delphinapterus and, along with the
narwhal, Monodon monoceros, comprises the Family Monodontidae. Its relationship to other
members of the Delphinidae, particularly the Irrawaddy dolphin, Orcaella brevirostris, has
been the subject of debate (Kasuya 1973). Traditional taxonomic approaches, such as
analysis of skull morphology, have supported the early (i.e. 6-23 Myr) divergence of belugas
from delphinid odontocetes (Barnes et al. 1985). CuiTently, molecular techniques are being
used to explore the systematics of numerous mammalian groups, including the Cetacea
(Shimura and Numachi 1987). Two approaches are the immunologic characteristics of
albumin and the electrophoretic properties of marker enzymes. In this volume, Donald Lint
and co-workers characterized samples from belugas and narwhals from various sites in the
Canadian Arctic, and compared their results to those for Irrawaddy dolphins, killer whales,
Orcinus orca, Pacific white-sided dolphins, Lagenorhynchus obliquidens, and Amazon
River dolphins, Inia geoffrensis. The data support the taxonomic positions established
through morphometric analysis, linking belugas and narwhals as the only two members of
the Monodontidae and excluding the Irrawaddy dolphin from this grouping.

Population Status

A circumpolar species, belugas are found principally in waters off Canada, Alaska, Russia,
Norway and Greenland. Within Canadian waters alone, there may be at least 7 separate
beluga populations: Beaufort Sea; high Arctic; Cumberland Sound (Eastern Baffin Island);
Ungava Bay; eastern Hudson Bay; western Hudson Bay; and St. Lawrence River. Within
these areas, there may be sub-stocks which exploit the same estuaries each summer. Recent
analysis of mitochondrial DNA has shown that the females from western Hudson Bay are
genetically distinct from those in the Nastapoka estuary of eastern Hudson Bay (Helbig et al.
1990). Such genetic subdivision of stocks is known to occur in humpback whales; in which
groups of females and calves occupy certain different summering areas and return to a
common winter area where breeding and genetic mixing occur (Baker et al. 1990). The issue
of stock discreteness is vital to those managing beluga whales as part of resource utilization.
An estimated 39 000 to 58 000 belugas live in Canadian waters; by far the largest
aggregations still occur in western Hudson Bay, where at least 23 000 belugas are estimated
to spend the summer. Pierre Richard and associates with the Canadian Department of
Fisheries and Oceans studied the whales in Hudson Bay, to update the work Sergeant (1973)
had done in the late 1960's, and examined again the important question of the unity of stocks
within subregions of the Bay.
For the Alaskan population, Kathryn Frost and Lloyd Lowry's account of the distribution,
abundance, and seasonal movements of belugas summarizes nearly 15 years of observations
by the Alaska Department of Fish and Game. They identify three major areas occupied by
belugas: Bristol Bay, the Norton Sound/Yukon River delta region, and the eastern Chukchi
Sea. Populations in each of these areas are estimated to be in the range of 1 000-3 000
animals, and are thought to have been stable over the past 20-30 years.
Beyond the need for regular population censusing is a requirement to identify sub-
groupings. In the past, we have relied on rather crude indices such as maximum body size
(Sergeant and Brodie 1969). Here, Bill Doidge reexamines this approach and uses new data
to dispute the clairn that whales from the high Arctic, eastern Hudson Bay, and the St.
Lawrence River can be distinguished on the basis of body size. New and more sensitive
techniques are required. Genetic analysis promises to resolve some basic questions on stock
identity. Telemetry has proved useful for short-term tracking of animals (Frost et al. 1985),
but is limited because the signalling devices are too soon rejected. Joseph Geraci and Gary
Smith examined some of the processes which lead to rejection of substances implanted in
odontocete skin. The work shows that we are far from having achieved a method for long-
term marking.

Behavioral Ecology

The estuarine habitat of belugas has kindled considerable speculation about the adaptive
significance of this behavior (Sergeant 1973; Finley 1983). Whatever the motivation, the
attraction is certainly strong. The work by Louise Caron and Tom Smith at the Nastapoka
estuary in eastern Hudson Bay shows that individuals, identifiable by scars, return to the

2
estuary in successive years. Repeated disturbances from motor traffic and Inuit hunts may
displace the animals for periods of hours or days, but then they return to the site. This
remarkable tenacity, undeterred by hunting pressure, makes belugas vulnerable to human-
induced changes to their habitat and to over-exploitation.
Belugas are disturbed by human activities even in the open ocean, as reported by Kerry
Finley who observed the response of belugas and narwhals to ice breakers in the high Arctic.
The whales, which presumably had not been exposed to motor traffic, reacted to the noise
from ships even at remarkably long distances. Each species showed a characteristic response.
Beluga herds dispersed and the animals scattered widely, whereas narwhals tended to huddle
together and remain motionless. These differing responses seem to mimic the predator-
avoidance strategies of each species. Yet the beluga seems capable of tempering this behavior
in areas such as the St. Lawrence, Churchill and Mackenzie rivers, where they have presum-
ably become habituated to heavy maritime traffic.
The vocal repertoire of belugas has earned them the epithet "canaries of the sea." Sjare and
Smith (1986a) described their vocalizations, but could not relate the sounds to specific
situations or activities (Sjare and Smith 1986b). Studies on captive subjects might hold the
key to further understanding. Here, C. Wayne Turl of the United States Navy analyses the
echolocation abilities of trained captive belugas compared to those of the Atlantic bottlenose
dolphin, about which infinitely more is known. The more highly developed and refined
capacities of the beluga are thought to relate to the complex ice-filled winter Arctic habitat
and possibly to their use of shallow water areas in the summer.

Physiological Ecology

Belugas, as periodic inhabitants of estuaries, are clearly more tolerant of freshwater than
are most other marine odontocetes. Little is known of the adaptations that allow for this
behavior, but several studies have highlighted some important physiological transformations
associated with incursion into warmer freshwater.
Studies on thyroid hormones in free-ranging belugas have shown a distinct annual cycle
in activity, which appears to be centered around the time of estuarine occupation (St. Aubin
and Geraci 1989a). Stimulation of hormone secretion at this time is thought to promote
somatic growth, in part through mobilizing energy stores, principally blubber. Reduction of
the blubber layer can occur in the warm water of the estuaries without risking excessive loss
of body heat. Doidge measured the insulating blubber shell and rate of heat conductance in
belugas from the Nastapoka River, and used the data to estimate thermostatic energy
requirements. He found that belugas in fact appear to be overinsulated in estuaries, and hence
could readily yield some of this reserve to satisfy the needs of growing tissue.
Estuarine occupation promotes at least one other biological process in belugas: accelerated
growth of epidermis, or molting (St. Aubin et al. 1990). Passive elevation of skin temperature
in these warmer waters appears to stimulate cellular processes, and the reduction in thermal
gradient between the whales and their environment allows for greater blood flow to the skin,
particularly if the animals are overinsulated. Increased circulation delivers growth-enhan-
cing thyroid hormones which are present at elevated concentrations during this phase of the
belugas' annual cycle. All these factors combine to stimulate epidermal proliferation,
causing molt-like shedding which so far appears to be unique to belugas.

Determining Age

The exact age of animals has been used traditionally as a key piece of information to
describe physiological growth rates and vital parameters of populations. Using tetracycline
to mark the dentinal layers, Paul Brodie and associates were able to measure growth rate in
the teeth of belugas held captive for up to 10 weeks, and thereby to confirm the view that one

3
year's growth is represented by the deposition of two growth-layer-groups within a tooth. The 1
study also prov ided some new insights on the dynamics of tooth growth in the species.

Immunology and Health Evaluation

Virtually all of the information on the health of free-ranging belugas comes from studies
on stranded carcasses along the St. Lawrence (Martineau et al. 1988). There is concern that
impaired immune function underlies some of the observed conditions. Yet we know little
about the functioning of the immune system in cetaceans. Tests developed for laboratory
species have only recently been applied, and the results, though promising, are preliminary.
David St. Aubin, Joseph Geraci and Patricia Shewen report on a pilot study testing the ability
of captive belugas to mount an antibody response to sheep erythrocytes, a widely used T-cell
antigen. The approach was successful and lays the groundwork for further studies using such
techniques.
Other constituents in blood can be used to assess the health of an animal. Normal ranges
for blood constituents have been established for captive (Cornell et al. 1988) and free-ranging
belugas (St. Aubin and Geraci 1989b). Using blood, it has also been possible to track the
course of acclimation to captivity, a process that includes elements of the stress response. The
endocrine component of this response was examined in some detail by St. Aubin and Geraci,
who used adrenocorticotropic hormone (ACTH) to stimulate the secretion of hormones from
the adrenal cortex of captive belugas. Like the bottlenose dolphin (Thomson and Geraci
1986), belugas show only a modest cortisol response, in contrast to the pronounced release
of aldosterone. The apparent advantage of this adaptation is to secure electrolyte balance
during stress by retaining sodium at a time when the whales are in a low-sodium environment.

Toxicology

Tissue burdens of contaminants such as chlorinated hydrocarbons and heavy metals signal
the extent to which by-products of human activity have entered the marine environment. The
presence of xenobiotics has been documented widely in marine mammals (Aguilar 1983;
Martineau et al. 1987; Muir et al. 1988; Addison 1989) with some attempt to link pathological
conditions to the presence of contaminants (Helle et al. 1976; Reijnders 1986; Martineau et
al. 1988). Addison (1989) lias pointed out the weakness of such associations, resulting either
from lack of proper control or baseline data, or from inconsistent methods of analysis.
Standardizing methodology will improve our ability to make comparisons between different
populations. Toxicological laboratories headed by Derek Muir and Rudolph Wagemann of
the Canadian Department of Fisheries and Oceans have done so for chlorinated hydrocarbons
and heavy metals, respectively. Clear differences are apparent between the level of hydro-
carbons in Arctic belugas and those from the St. Lawrence River, and for heavy metals
among some subpopulations in the Arctic as well. Future studies may be able to incorporate
these findings as part of epidemiological studies linking contaminant levels with functional
disturbances.

Direction of Future Research

The works of Vladykov (1944) and Kleinenberg et al. (1964) established the biological
character of the beluga whale, using as a theme anatomy, feeding habits, age, growth,
reproduction, and other elements of natural history that were relevant to the management
issues of the day. Research for management purposes is still the dominant need, but, as
ev idenced by the topics covered in this monograph, current approaches are more varied and
include new technology. Data on whale abundance and distribution, long viewed as the
cornerstone of management research, become more valuable when infused with an under-

4
standing of behavior and physiology. Factors contributing to morbidity and mortality must
be identified to enable us to recognize pathological conditions that might be linked to human
activities and contaminants. The modern-day manager needs to draw on the full spectrum of
information about what influences the life and death of whales. Future studies should probe
deeply into any discipline that promises to nourish that body of knowledge.

Acknowledgments

Gary Sleno and George Horonowitsch (Dept. of Fisheries and Oceans) helped ready the
papers for publication, verifying literature cited and scrutinzing page proofs; G. Sleno also
prepared some of the figures. Dr. Michael Hammill (DFO) was involved in the initial
planning of the monograph, and was a convenor of a workshop that led to its production. We
thank Nikki Garwood and Dr. Claude Desjardins (DFO) (for Muir et al.) for translating the
abstracts, Dale-Marie Smith and Elizabeth Crossey (Univ. of Guelph), and Dora Goddard
(DFO) for secretarial assistance and Lenore Latta (Guelph) for copy editing. The Department
of Fisheries and Oceans (Quebec Region) funded the project, as part of the Programme
d'Action du St. Laurent (PASL) of the Government of Canada.

References

ADDISON, R. F. 1989. Organochlorines and ma- FROST, K. J., L. F. LOWRY, AND R. R. NELSON.
rine mammal reproduction. Can. J. Fish. Aquat. 1985. Radiotagging studies of beluga whales
Sci. 46: 360-368. (Delphinapterus leucas) in Bristol Bay, Alas-
AGUILAR, A. 1983. Organochlorine pollution in ka. Mar. Mammal. Sci. 1: 191-202.
sperm whales (Physeter macrocephalus), from HELBIG, R., P. T. BOAG, AND B. N. WHITE. 1990.
the temperate waters of the eastern North Atlan- Stock identification of beluga whales (Delphi-
tic. Mar. Pollut. Bull. 14: 349-352. napterus leucas) using mitochondrial DNA
BAKER, C. S., S. R. PALUMBI, R. H. LAMBERTSEN, markers: preliminary results. Musk Ox 37:
J. CALAMBOKIDIS, AND S. J. O'BRIEN. 1990. 122-128.
Influence of seasonal migration on geographic HELLE, E., M. OLSSON, AND S. JENSEN. 1976. PCB
distribution of mitochondrial DNA haplotypes levels correlated with pathological changes in
in humpback whales. Nature 344: 238-240. .seal uteri. Ambio 4: 261-263.
BARNES, L. G., D. P. DOMNING, AND C. E. RA Y. KASUYA, T. 1973. Systematic consideration of
1985. Status of studies on fossil marine mam- recent toothed whales based on the morpholo-
mals. Mar. Mammal. Sci. 1: 15-53. gy of tympano-periotic bone. Sci. Rep. Whales
BRODIE, P. F. 1971. A reconsideration of aspects Res. Inst. 25: 1-103 ,
of growth, reproduction and behavior of the KLEINENBERG, S. E., A. V. YABLOKOV, R. M.
white whale (Delphinapterus leucas) with BECKOVICH, AND M. N. TARASEVICH. 1964.
special reference to the Cumberland Sound, Beluga (Delphinapterus leucas). Investigations
Baffin Island population. J. Fish. Res. Board. of the species. Israel Program for Scientific
Can. 28: 1309-1318. Translations. Jerusalem. 376 p.
CORNELL, L. H., D. S. DUFFIELD, B. E. JOSEPH, MARTINEAU, D., P. BÉLAND, C. DESJARDINS, AND
AND B. STARK. 1988. Hematology and serum A. LAGACÉ. 1987. Levels of organochlorine
chemistry values in the beluga (Delphinapte- chemicals in tissues of beluga whales (Delphi-
rus leucas). J. Wildl. Dis. 24: 220-224. napterus leucas) from the St. Lawrence estua-
DEFRAN, R. H., AND K. PRYOR. 1980. The beha- ry, Quebec, Canada. Arch. Environ. Contam.
vior and training of cetaceans in captivity, p. Toxicol. 16: 137-147.
319-362. In L. T. Herman [ed.] Cetacean MARTINEAU, D., A. LAGACÉ, P. BÉLAND, R.
behavior: mechanisms and functions. John HIGGINS, D. ARMSTRONG, AND L. R. SHUGART.
Wiley & Sons, New York, NY. 1988. Pathology of stranded beluga whales
FINLEY, K. J. 1983. The estuarine habit of the (Delphinapterus leucas) from the St. Lawrence
beluga or white whale, DelphMapterus leucas. estuary, Quebec, Canada. J. Comp. Path. 98:
Cetus 4: 4-5. 287-311.

5
MUIR, D. C. G., R. WAGEMANN, N. P. GRIFT, R. J. whales. Sci. Rep. Whales Res. Inst. 38:
NORSTROM, M. SIMON, AND J. LIEN. 1988. 141-163.
Organochlorine chemical and heavy metal con- SJARE, B. L., AND T. G. SMITH. 1986a. The vocal
taminants in white-beaked dolphins (Lageno- repertoire of white whales (Delphinapterus
rhynchus albirostris) and pilot whales (Globi- leucas) summering in Cunningham Inlet,
cephala tnelaena) from the coast of Newfoun- Northwest Territories. Can. J. Zool. 64:
dland, Canada. Arch. Environ. Toxicol. Chem. 407-415.
17: 613-629. 1986b. The relationship between behavio-
REIJNDERS, P. J. H. 1986. Reproductive failure in ral activity and underwarer vocalization of the
common seals feeding on fish from polluted white whale (Delphinapterus leucas). Can. J.
coastal waters. Nature 324: 456-457. Zool. 64: 2824-2831.
ST. AUBIN, D. J., AND J. R. GERACI. 1989a. Seaso- SMITH, T. G., M. O. HAMMILL, D. W. DOIDGE, T.
nal variation in thyroid morphology and secre- CARTIER, AND G. A. SLENO. 1979. Marine
tion in the white whale (Delphinapterus leu- mammal studies in southeastern Baffin Island.
cas). Can. J. Zool. 67: 263-267. Final report to Eastern Arctic Marine Environ-
1989b. Adaptive changes in hematologic mental Studies Project. Can. MS. Rep. Fish.
and plasma chemical constituents in captive Aquat. Sci. 1552: 80 p.
beluga whales (Delphinaterus leucas). Can. J. SMITH, T. G., AND M. O. HAMMILL. 1990. A
Fish. Aquat. Sci. 46: 796-803. bibliography of the white whale, Delphinapte-
Si. AUBIN, D. J., T. G. SMITH, AND J. R. GERACI. rus leucas. Can. MS. Rep. Fish. Aquat. Sci.
1990. Seasonal epidermal molt in beluga whales 2060: 45 p.
(Delphinapterus letter's). Can. J. Zool. 68: THOMSON, C. A., AND J. R. GERACI. 1986. Corti-
359-367. sol, aldosterone, and leucocytes in the stress
SERGEANT, D. E. 1973. Biology of the white response of bottlenose dolphins (Tursiops turn-
whales (Delphinapterus leucas) in western coats). Can. J. Fish. Aquat. Sci. 43: 1010-1016.
Hudson Bay. J. Fish. Res. Board Can. 30: VLADYKOV, V. D. 1944. Etudes sur les mammifè-
1065-1090. res aquatiques. III. Chasse, biologie et valeur
SERGEANT, D. E., AND P. F. BRODIE. 1969. Body économique du marsouin blanc ou beluga (Del-
size in white whales, Delphinapterus leucas. J. phinapterus leucas)du fleuve et du golfe Saint-
Can. Fish. Res. Board Can. 26: 2561-2580. Laurent. Québec, Department des Pêcheries,
SHIMURA, E., AND K. -I. NUMACH1. 1987. Genetic Province of Quebec. 194 p.
variability and di fferentiation in the toothed

6
 Evolution and Systematics of the
Beluga Whale, Delphinapterus leucas, and Other Odontocetes:
A Molecular Approach

D.W. Lint
Department of Zoology, University of Manitoba,
Winnipeg, Manitoba, Canada R3T 2N2

J.W. Clayton', W. R. Lillie, and L. Postma

Department of Fisheries and Oceans, Freshwater Institute,
501 University Crescent, Winnipeg, Manitoba, Canada R3T 21\16

LINT, D.W., J.W. CLAYTON, W.R. LILLIE, AND L. POSTMA. 1990. Evolution and
systematics of the beluga whale, Delphinapterus leucas, and other odontoce-
tes: a molecular approach, p. 7-22. In T.G. Smith, D.J. St. Aubin, and J.R.
Geraci [ed.] Advances in research on the beluga whale, Delphinapterus leucas.
Can. Bull. Fish. Aquat. Sci. 224.

The evolution and systematics of the beluga whale (Delphinapterus leucas) and
11 other odontocetes were investigated using immunological and enzyme electro-
phoresis procedures. Rabbit antibodies were prepared to serum albumins from
Delphinapterus leucas, Monodon monoceros, 01-CillUS orca, Lagenorhynchus
obliquidens, Orcaella brevirostris,and MM geoffrensis. Reaction of these antibo-
dies with odontocete albumin antigens, by the micro-complement fixation proce-
dure, indicated a very close relationship between Delphinapterus and Monodon
(family Monodontidae), and another close relationship among Orcinus, Lageno-
rhynchus obliquidens, and Orcaella (family Delphinidae). 'nia, although closer to
the Monodontidae than to the Delphinidae, was clearly not associated with either
group.
Electrophoretic mobilities of enzymes were compared from all the species used
for the immunological comparisons, except I nia, plus Physeter macrocephalus,
Mesoplodon bidens, Phocoena phocoena, Phocoenoides dalli, Pseudorca crassi-
dens, Globicephala melas, and Lagenorhynchus albirostris. The electrophoretic
data were strongly correlated with the immunological data (r2 = 0.83) and
confirmed the branching pattern among the species revealed by the immunologi-
cal data as well as the extremely close relationship between Delphinapterus and
Monodon. Like the immunological data, the electrophoretic data associated
Orcaella with a group that includes all the representatives of the family Delphi-
nidae that were examined, and not with Delphinapterus or Monodon. The data
were discussed in terms of a phylogeny for the odontocetes and a time scale of
odontocete evolution.

Des recherches sur l'évolution et la systématique chez le marsouin blanc

(Delphinapterus leucas) et 11 autres oclontocétés furent faites utilisant des procé-
dés immunologiques et d'électrophorèses enzymatiques. Des anticorps de lapins
ont été préparés aux albumines de sérum de Delphinapterus leucas, Monodon mo-
noceros,Orcinus orca,Lagenorhynchus obliquidens,Orcaella brevirostris et Inia
geoffrensis. La réaction de ces anticorps avec les antigènes albuminés d'odonto-
cete, par la méthode de fixation micro-complément, témoignèrent d'une parenté,
très étroite entre Delphinapterus et Monodon (de la famille Monodonticlae), et
d'une autre proche parenté entre Orcinus, Lagenorhynchus, et Orcaella (famille

'Author to whom correspondence should be addressed.

7
 Delphinidae). /nia, quoique plus rapprochée du Monodontidae que du Delphinidae
n'était clairement associée ni à un groupe ni à l'autre.
Les mobilités électrophorétiques des enzimes furent comparées chez toutes les
espèces utilisées pour les comparaisons immunologiques, excepté bila, en plus de
Physeter macrocephalus, Mesoplodon bidens, Phocoena phocoena, Phocoenoi-
des dalli,Pseudorca crassidens, Globicephala me/aset Lagenorhynchus albiros-
tris. Les données électrophorétiques furent correlées très étroitement avec les
données immunologiques (r2 = 0.83) et confirment la tendance de ramification
parmi les espèces révélées par les données immunologiques en plus de la relation
très étroite entre De/phinapterus et Monodon. Tout comme les données immuno-
logiques, les données électrophorétiques associèrent Orcaella avec un groupe qui
comprend tous les représentants de la famille Delphinidae qui ont été examinés, et
non avec Delphinapterus ou Monodon. La discussion sur les données a été faite en
rapport avec une phylogénie pour les odontocétés et un barème d'évolution
d'odontocété.

Introduction

The living members of the Order Cetacea are divided into two suborders: Mysticeti (baleen
whales) and Odontoceti (toothed whales). In the suborder Odontoceti, some six (Gaskin
1982) to nine (Rice 1984) living families are recognized. Reviews by Kleinenberg et al.
(1964) and Stewart and Stewart (1989) (Delphinapterus), and Reeves and Tracey (1980)
(Monodon) accept the classical taxonomy in which the beluga, Delphinapterus leucas, and
the narwhal, Monodon monoceros, comprise the two species of the family Monodontidae.
These reviews attribute the first description of beluga to Pallas (1776) as Delphinus leucas
from animais examined at the mouth of the River Ob on the Arctic Ocean coastline of what
is now the USSR. Apparently the nameDelphinus had already been used by Linnaeus (1758)
to describe another group of small odontocetes (Barnes 1976). It seems to be unclear whether
Pallas (1776) actually considered the beluga to have affinities to the otherDelphinus species
(Kleinenberg et al. 1964). The current name Delphinapterus leucas was evidently first put
forward by Lacépède (1804) (Kleinenberg et al. 1964; Stewart and Stewart 1989). There is
a long history of debate as to the family or subfamily status of the species, and although the
assignment of Delphinapterus and Monodon to the family Monodontidae lias had many
supporters including Simpson (1945), there have also been many other differing opinions.
A complex alternative taxonomy has been advocated by Kasuya (1973). Briefly, Kasuya
proposed, on the basis of ear bone morphology, that the Irrawaddy River dolphin, Orcaella
brevirostris, merits a position of affinity with Delphinapterus such that the two species would
constitute a grouping to the exclusion of Monodon which would form a monospecific sister
group. This taxonomy has received qualified endorsement from Barnes (1976), Gaskin
(1982), and Barnes et al. (1985) who simply expand Monodontidae to include Orcaella.
Our approach consists of the use of the quantitative immunological method, microcomple-
ment fixation (MC'F), based on serum albumin, to examine the taxonomy and phylogenetic
relationships of the beluga (Champion et al. 1974; Wilson et al. 1977; Maxson and Maxson
1986). We have used the immunological comparison of odontocete serum albumins not only
because it lias been successful in investigations of many other taxa and there is therefore a
considerable body of comparative data available (Maxson and Maxson 1986) but also
because unpublished data (Vincent Sarich, pers. comm.) indicated that this method would be
appropriate. We have supplemented the information available from serum albumin immu-
nology with a comparison of electrophoretic mobilities of a number of enzymes among
accessible odontocete species (Sarich 1977; O'Brien et al. 1985).
The primary questions addressed in this paper are whether Delphinapterus is so closely
related to Monodon that together they constitute a separate family Monodontidae within the
suborder Odontoceti, or whether there is justification to associate Orcaella with either

8
Delphinapterus in particular or more generally with Monodontidae. We will also discuss
odontocete phylogeny and evolution as revealed by the serum albumin immunological and
enzyme electrophoretic data.

Materials and Methods

Odontocete Specimens
The odontocete species included in this study are presented in Table 1.Delphinapterus and
Monodon tissue and blood samples were obtained by Canada Department of Fisheries and
Oceans staff, from animals killed by native hunters in the Canadian Arctic.

TABLE I. Odontocetes represented in this study: Famillies, species abbreviations, systematic

and vernacular names, and waters where sampled.
Physeteridae
Pm— Physeter macrocephalus (Linnaeus, 1758; Husson and Holthius, 1974*)
— Sperm whale (Northwest Atlantic)

Ziphiidae
Mb— Mesoplodon bidens (Sowerby, 1804)
— Sowerby's beaked whale (Northwest Atlantic)

Platanistidae
Inia— Inia geoffrensis (deBlainville, 1817)
— Amazon river dolphin (Amazon River)

Monodontidae
Mm — Monodon monoceros (Linnaeus, 1758)
— Narwhal (Arctic)
Dl — Delphinapterus leucas (Lacépède, 1804)
—Beluga (white) whale (Arctic)

Phocoenidae
Pp — Phocoena phocoena (Linnaeus, 1758)
— Harbour porpoise (Northwest Atlantic)
Pd — Phocoenoides dalli (True, 1885)
—Dall's porpoise (Northeast Pacific)

Delphinidae
Gm — Globicephala melas (Traill, 1809; Rice, 1990*)
— Longfinned pilot whale (Northwest Atlantic)
Oo — Orcinus orca (Linnaeus, 1758)
— Killer whale (Northeast Pacific)
Pc — Pseudorca crassidens (Owen, 1846)
— False killer whale (Northeast Pacific)
La — Lagenorhynchus albirostris (Gray, 1846)
— Whitebeaked dolphin (Northwest Atlantic)
Lo — Lagenorhynchus obliquidens (Gill, 1865)
— Pacific whitesided dolphin (Northeast Pacific)
Ob — Orcaella brevirostris (Gray, 1866)
— Irrawaddy river dolphin (Southeast Indian)
* See References.

9
 Blood samples from Orcinus (Hyack) and Lagenorhynchus obliquidens (White Wings)
were obtained from the Vancouver Public Aquarium. Zinn serum, originally collected by the
late Robin Best of Instituto Nacional de Pesquisas da Amazonia, Manaus, Brazil, was
provided by Joseph Geraci, University of Guelph. Other samples came from animals stranded
or accidently entrapped in fishing gear.

Immunological Procedures

Serum albumin was recovered from isoTris (Champion et al. 1974) rinsings of Orcaella
liver and kidney tissue. Monodon liver and kidney were also treated in this way to augment
albumin derived from whole blood. Treatment of blood-rich isoTris tissue rinsing solution
with Rivanol (2-ethoxy-6, 9-diaminoacridine lactate) precipitated albumin as a
Rivanol—albumin complex (Baker et al. 1981). Albumin was released from this complex by
dialysis against 0.5 M Tris-HC1 (Baker et al. 1981). The resulting albumin solution was
desalted and concentrated to 1 mL in an Amicon 10 mL ultrafiltration stirred-cell. For the
other species, a column of Cibacron-Blue bound to Affigel agarose beads was used to
separate albumin from other major plasma proteins and hemoglobin in plasma, serum, and
whole blood. The albumin-rich fractions from the column were desalted, and concentrated
as described above for the tissue-derived protein. Both types of preparation were then
subjected to polyacrylamide—gel electrophoresis (Maxson 1973, and Linda Maxson, pers.
comm.). Albumin-containing bands, located by staining with 8-anilino- 1-naphthalene sul-
phonate, were cut out and albumin was extracted into isoTris. All procedures were conducted
at 0-4°C and all tissues, preparations, and solutions were stored at -20°C or colder.
Antibodies were prepared in rabbits, starting with four rabbits per antigen, according to
established methods for a 13 wk immunization schedule (Maxson et al. 1979). For Delphi-
napterus only, an additional set of rabbits was immunized over a 30 wk period according to
the protocol of Prager and Wilson (1971). The sera from individual rabbits were titred by the
MC'F procedure and for each antigen species the corresponding antisera were combined in
inverse proportion to titre (Champion et al. 1974). Inter-species comparisons were made with
the pooled sera except for the anti-Monodon sera, for which the titres were very low and the
serum from only one rabbit was satisfactory. Immunological distances (ID) were calculated
from homologous and heterologous slopes derived from several microcomplement fixation
curves for each determination (Champion et al. 1974; Lint 1987).

Enzyme Electrophoresis Procedures

Isozymes were resolved by the semi-micro starch—gel electrophoresis method of Tsuyuki

et al. (1966) as described by Clayton and Tretiak (1972) and detected with enzyme-specific
staining procedures as described in Table 2. Most species were represented by few
individuals and the genetic heterozygosity was very low. In these circumstances genetic
similarities (S) were determined for each species pair by simply counting, on a locus-by-locus
basis, the number of loci with identical electrophoretic mobility and then dividing by the
number of loci resolved in common in both species (Cronin and Sarich 1976). Electropho-
retie distance (D) (Sarich 1977) was calculated as:
D = —ln S

10
TABLE 2. Enzyme electrophoresis and staining conditions.
Enzyme' abbreviation, Preferred Electro.c.d'e Stain buffer Concentrationaf
E. C. No., and Name tissueb buffer and
components d
G6PD, 1.1.1.49 1 AC, pH 6.1 Tris; 0.1 M, pH 8.0
glucose-6-phosphate with: glucose-6-phosphate
dehydrogenase NADP, 0.1 (Na2 salt) 1.6
MgC1 2. 6H20 0.4
NBT 0.4
PMS 0.09

GDH, 1.1.1.47 1 AC, pH, p1-1 6.1 Tricine; 0.15 M, pH 8.0

glucose dehydrogenase glucose 200
sodium pyruvate 8.0
NAD 0.7
NBT 0.4
PMS 0.03

GLUD, 1.4.1.3 1 TEB, pH 8.0 Tris; 0.1 M, pH 8.0

glutamate dehydrogenase with: sodium glutamate 10
sucrose, 40; sodium pyruvate 1
NAD, 0.1 NAD 1.2
NBT 0.2
PMS 0.03

GOT, 2.6.1.1 m AC, pH 9.0 Tris; 0.1 M, pH 8.0

glutamic-oxaloacetic L-aspartic acid 1.5
transaminase 2-oxoglutaric acid 0.8
(aspartate pyridoxa1-5-phosphate 0.01
aminotransferase) Fast Blue BB 0.6

GPD, 1.1.1.8 m AC, pH 8.0 Tricine; 0.15 M, pH 8.0

glycerol-3-phosphate DL-glycerol-3-phosphate
dehydrogenase (Na2 salt) 20
pyruvic acid 8
NAD 0.4
NBT 0.3
PMS 0.02

GPI, 5.3.1.9 m AC, pH 9.0 Tris; 0.1 M, pH 8.0

glucose phosphate fructose-6-phosphate
isomerase (Na2 salt) 1.4
glucose-6-phosphate
dehydrogenase 1.8
MgC1 2. 6H20 0.7
NADP 0.2
NBT 0.2
PMS 0.02

11
TABLE 2: (CO/nit/Bed.)

Enzyme" abbreviation, Preferred Electro.".d'e Stain buffer Concentratioef

E C. No., and Name tissue' bu ffer and
components d
IDH, 1.1.1.42 1 AC, pH 6.1 Bicine; 0.15 M, pH 8.5
isocitrate dehydrogenase with: DL-isocitric acid
NADP, 0.1; (Na, salt) 1.3
Nonidet MgC1 -, 6H20 0.4
P 40g, 0.1 NADP- 0.3
NBT 0.2
PMS 0.03

LDH, 1.1.1.27 ni AC, pH 6.1 diethanolamine; 0.15 M, pH 9.0

lactate dehydrogenase lactic acid (pH 7.0, NaOH) 11
NAD 0.7
NBT 0.4
PMS 0.02

MDH, 1.1.1.37 ni AC, pH 6.1 diethanolamine; 0.15 M, pH 9.0

malate dehydrogenase malic acid (pH 7.0, NaOH) 16
NAD 0.7
NBT 0.4
PMS 0.02

ME, 1.1.1.40
malic enzyme m AC, pH 6.1 Tricine; 0.15 M, pH 8.0
(malate dehydrogenase, malic acid 16
oxaloacetate- MgC1 2.6H20 0.2
decarboxylating) NADP 0.3
NBT 0.4
PMS 0.02

MPI, 5.3.1.8 ni AC, pH 8.0 Tricine; 0.15 M, pH 8.0

mannosephosphate D-mannose-6-phosphate
isomerase (Ba salt) 1
glucosephosphate
isomerase 2
glucose-6-phosphate
dehydrogenase 2
MgC1 2.6H,0 0.2
NADP 0.2
NBT 0.2
PMS 0.05

PGD, 1.1.1.43 1 AC, pH 8.0 Tris; 0.1 M, pH 8.0

phosphogluconate with; 6-phosphogluconic acid
dehydrogenase NADP, 0.1; (Na3 salt) 2
Nonidet MgC1 2 6H,0 1
P 40e, 0.1 NADP - 0.5
NBT 0.5
PMS 0.05

12
TABLE 2. (Continued.)
Enzyme abbreviation, Preferred Electro.°"° Stain buffer Concentrationgf
E. C. No., and Name tissue" buffer and
components d
PGM , 5.4.2.2 1 AC, pH 6.1 Tricine; 0.15 M, pH 8.0
phosphoglucomutase glucose-1 -phosphate
(Na2 salt) 4
glucose-6-phosphate
dehydrogenase 1.6
MgC1 2'6H20 2
NADP 0.4
NBT 0.4
PMS 0.03
g Enzyme nomenclature according to Shows et al. (1987) and McAlpine et al. (1988).

" Tissue abbreviations: m, muscle; 1, liver.

• Abbreviations and references for electrophoresis buffers:

AC, Amine-citrate buffers described by Clayton and Tretiak (1972);
for pH 9.0, 2-amino-2-methy1-1,3-propanediol was used,
for LDH the buffer concentration in the starch was doubled;
TEB, Tris-EDTA-borate described by Siciliano and Shaw (1976).

d Abbreviations for buffer and stain ingredients:

Bicine, N,N-bis-(2-hydroxyethyl)-glycine;
EDTA, ethylenediaminetetraacetic acid;
NAD, nicotinamide adenine dinucleotide;
NADP, nicotinamide adenine dinucleotide phosphate;
NBT, Nitroblue tetrazolium;
Nonidet P40, ethylphenylpolyethyleneglycol,
(nonionic detergent, Registered Trademark of Shell, also known as NP 40);
PMS, phenazine methosulphate;
Tricine, N-(tris-(hydroxymethyl)-methyll-glycine;
Tris, tris-(hydroxymethyl)aminomethane;

• All concentrations of chemicals in mg•mL- ', and concentrations of linking enzymes in units•mL- '.
All stains were incubated at 22°C except GDH which was incubated at 37°C.

I' The stain recipes were adapted from the following sources:
GPD, (Clayton et al. 1973);
GPI, (Siciliano and Shaw 1976; Nielsen et al. 1989);
LDH and MDH, (Clayton and Gee 1969; Nielsen et al. 1989);
MPI, (Allendorf et al. 1977);
PGM, (Nielsen et al. 1989).
For other enzymes the present procedures represent extensive adaptations of published
procedures to produce optimum results with cetacean enzymes. The methods of
Harris and Hopkinson (1976) were used as a general guide.

g The inclusion of an unusual ingredient, the nonionic detergent Nonidet P 40,

significantly enhanced the respective isozyme stains.

13
Distance Diagrams

Unrooted distance diagrams were constructed and the associated standard deviations
determined from the molecular data (reciprocal ID's and electrophoretic D's) by the Fitch and
Margoliash (1967) method using the Runtime program by George Wilson in the PHYLIP
version 3.2 package of computer programs (J. Felsenstein, University of Washington, 1989).
The program was run 13 times on each data set according to Felsenstein's recommenda-
tions, using a procedure to add species to the trees in different orders, in an attempt to discover
the best trees. The standard deviation for phylogenetic trees is not the same as the usual one
but instead reflects the agreement between the actual distance data, in this case ID and D
values, with the distances measured between species along the branches in the distance
diagrams. The Fitch—Margoliash programs in the PHYLIP package are incompatible with
zero distances; therefore zero values were arbitrarily changed to 0.0001.

Results

Immunology

The reciprocal ID data (Table 3) clearly associate Delphinapterus and Monodon (average
ID = 2.4) to the exclusion of any of the other species tested. These same data also clearly
associate Orcaella,Orchnts and Lagenorhynchus obliquidens, (average ID = 5.9), in another
separate group. blitz is not closely associated with either group, although it is certainly closer
to Delphinapterus—Monodon (average ID = 10.7) than to the Orcinus—Lagenorhynchus
obliquidens—Orcaella group (average ID = 21.7).
Where more than one ID determination was conducted, the entries in Table 3 are the
average of all multiple determinations with their associated standard deviations. While the
standard deviations of replicate determinations are mostly less than 10%, there is substantial
disagreement among many of the reciprocal distances (Table 3). This lack of reciprocity is
typical of these kinds of ID data sets and has been quantified as percentage deviation of the
average ID value by Champion et al. (1974), and has been discussed extensively (Maxson and
Wilson 1975; Beverley and Wilson 1982).
For the data in Table 3, the percentage deviation in reciprocal ID values ranged from 0.4
to 36% and the average (13.6%) is similar to other data sets (Beverley 1979; Maxson and
Maxson 1979). Much of this lack of reciprocity centres onlnia where every ID measured with
the anti-/nia serum was greater than the corresponding reciprocal ID (Table 3). The similarity

TABLE 3. Reciprocal immunological distances determined with antisera raised against odontocete
serum albumin. Species abbreviations are given in Table 1. Entries in parentheses represent ID's
obtained with a 30—wk anti-Delphinapterus serum. Entries with ± attached represent averages of two
or more determinations; all other entries were determined from comparisons of homologous and
heterologous slopes.
Antisera
Antigens
Dl Mm Oo Lo Ob lnia
DI 1.5±.4 9.5±1.2 11.2 13.1 12.4±.2
Mal (1) 3.2±.2 9.2±1.6 9.8 12.8 13.5±.1
00 (8) 11.4±1.4 8.8±.6 5.4 4.7 23.0
Lo (10) 11.7±1.4 11.1±.2 7.4 6.5 23.5
Ob (13) 13.0±1.7 I3.1±2.3 7.1 4.5 30.0
Min (12) 8.8±2.0 8.2±2.8 18.0 20.3 15.2

14
of results with the 30-wk and 12-wk anti-Delphinapterus sera indicates that the 12-wk sera
were, as expected, satisfactory for the present kind of investigation (Champion et al. 1974).

Isozyme Electrophoresis

Only one or a few individuals were available for analysis for all species except Delphinap-
terus and Monodon, thus the enzyme electrophoretic mobility entries (Table 4) represent the
only, or the most common, allelic form observed at each locus. The 211 Delphinapterus
tested were invariant except for 1 LDHa, 1 PGM, and 11 IDH heterozygotes. Rare alleles
aside, Monodon and Delphinapterus enzymes differed in electrophoretic mobility (Table 4)
only at the mannose phosphate isomerase locus (MPI) among the 15 loci resolvable in both
species. Perhaps less expected was the identity of Globicephala and Pseudorca at 10 loci.
Inia is not included in the electrophoretic data set (Table 4) because our only Inia material
was serum from which we were able to display only lactate dehydrogenase (LDH) activity
subsequent to electrophoresis.
In Fig. 1, average immunological distances (ID) were plotted against the corresponding
electrophoretic distances (D) among the five species, Delphinapterus, Monodon, Lageno-
rhynchus obliquidens, Orcinus, and Orcaella, for which both types of data were available.
The correlation was ia = 0.83, and the slope was 15.3.

1 51—

oo
o
00
10
seru m albu m in ID

o
o
5
o
r 2 = 0.83

1 1 1 1 1 1 1 1
0
0 0.5 1.0
enzyme electro. D
Fia. 1. Albumin immunological distance (ID) plotted against
enzyme electrophoretic distance (D) for 10 pairs of odonto-
cetes; / -2 = 0.83, slope = 15.3.

15
TABLE 4. Relative electrophoretic mobilities of odontocete enzymes, approximately in mm; nd, not determined. Species abbreviations as given in Table 1. Tissue abbreviations:
m, muscle; I, liver; h, heart .
Species Enzyme abbreviation and enzyme commission number
and
(Tissue G6PD GDH GLUD GOT GPD GPI IDH LDH MDH ME MPI PGD PGM
tested) 1.1.1.49 1.1.1.47 1.4.1.3 2.6.1.1 1.1.1.8 5.3.1.9 1.1.1.42 1.1.1.27 1.1.1.37 1.1.1.40 5.3.1.8 1.1.1.43 5.4.2.2
Pm (m) +22 nd +28 nd, nd +24 +15 -14,+10 -6, nd -20,+16 +16,+20 +38 +22 +2
Mb (1, m) +22 +15 +28 -12,+19 nd +15 -14,+14 -10, nd -20,+16 +16,+20 +38 nd +10
Mm (1, m) +22 +15 +28 -12,+19 +20 +9 nd, +10 -10,+25 -20,+16 +12,+20 +40 nd +2
DI (1, m) +22 +15 +28 -12,+19 +20 +9 -12,+10 -10,+25 -20,+16 nd, +20 +38 +20 +2
Pp (1, m) +22 +15 +28 -12,+19 +20 +9 -12,+10 -10,+25 -20,+16 +16,+28 +32 +17 -5
Pd (1, m) +22 nd nd -12,+19 nd +9 -12,+10 -10,+25 -20,+16 nd, +20 +32 +28 -2
Gm (1, m) +22 nd nd -9,+24 nd +15 nd, +10 -10,+25 -20,+16 +16,+20 +40 +20 0
Oo (1, h) +22 nd nd -9,+24 nd +15 -15,+10 -10,+25 -20,+16 nd, +20 +40 +20 0
Pc (m) nd nd nd -9, +24 nd +15 nd, nd -10, +25 -20, +16 nd, +20 +40 nd 0
La (1, m) +22 +15 +28 -9,+24 +20 +15 -14,+10 -10,+25 -20,+16 +16,+20 +32 +20 0
Lo (1, m) +22 nd nd -9,+24 nd +15 -14,+10 -10,+25 -20,+16 +16,+20 +38 +20 0
Ob (1, m) +22 nd +28 -9,+24 nd +15 -13,+16 -10,+25 -20,+16 +16,+20 +26 +15 0
Distance Diagrams

The Fitch—Margoliash distance diagrams (Fig. 2a—d) are all drawn to a scale derived from
the slope of Fig. 1 (15.3), thus the index bars in Fig. 2 show that 0.1 electrophoretic D units
is the same linear distance as 1.53 serum albumin ID units. These diagrams (Fig. 2a—d) all
clearly demonstrate the strong association of Delphinapterus and Monodon. It is also clear
from each diagram that all the representatives of the family Delphinidae, including Orcaella,
are closely associated. The standard deviations for these diagrams calculated with the
PHYLIP program are shown here:

Figure Data set Percent SD No. trees

tested
2a all ID 18.7 28
2b all ID except Inia 17.1 15
2c electro. D;
5 species as in lb 11.0 15
2d all electro D;
12 species 18.2 196

The relatively large percent standard deviations for diagrams derived from the reciprocal
ID data (Fig. 2a and 2b) evidently reflect the deviations from perfect reciprocity in these data
(Table 3), considering the number of species involved.

Discussion

All the distance diagrams (Fig. 2a—d) demonstrate a close relationship betweenDelphinap-
terus and Monodon to the exclusion of any other species, thereby supporting the classical
taxonomic assignment of Delphinapterus and Monodon to a two species family Monodon-
tidae. These diagrams show that Orcaella is definitely not a member of Monodontidae but
is clearly associated with the family Delphinidae. Neither Inia, Physeter, nor Mesoplodon
appear to represent distinct outgroups to the other odontocetes in this investigation. [A
legitimate outgroup would display an essentially identical ID or D to all members of the study
group; see for example the position ofHyaena relative to the Felis cats in Collier and O'Brien
(1985).]
The addition or subtraction of taxa in the data set used to calculate these Fitch—Margoliash
distance diagrams (Fig. 2a—d) may be expected to change the relationships among the taxa.
However, the present diagrams show essentially the same branching pattern and similar
distance relationships regardless of the number of species in the data set. Evidently this is
another indication of the reliability of the data. These Fitch—Margoliash diagrams may not
be as familiar as the U-shape phylogenetic trees derived from pair group data reduction
methods [see for example Shimura and Numachi (1987)] but they are much more represen-
tative of the actual data because the branch lengths are not adjusted to an arbitrary baseline
as in the usual pair group presentation.
A similar enzyme electrophoresis study of odontocete relationships (Shimura and Numa-
chi 1987) included Pseudorca crassidens, Lagenorhynchus obliquidens, Phocoena phocoe-
na, and Phocoenoides dalli, species also represented in the present work. Of the 19 loci
examined by Shimura and Numachi, and the 18 loci tested in the present work, 13 were the
same loci. Shimura and Numachi examined large numbers of individuals and expressed their
results as Nei genetic distances (D). A plot of these Nei D's against the present band counting

17
 (a) Lo
Mm
Inia
4 DI Ob

DI
Oo
2.0 1 1D

Lo
(b)
Mm
DI —D Ob

Oo
2.0 ID

(c) Mm
Lo

DI Ob

00
0.1 electro.D

(d)

Pd Ob

Mb
0.1 electro.D
Fia. 2. Fitch—Margoliash distance diagrams derived from serum albumin immu-
nological distances (ID) and enzyme electrophoretic distances (D); species abbre-
viations are given in Table 1. (a) all immunological distance data; (b) all immuno-
logical distance data except bila; (c) electrophoretic distance data for the same
species as in (b); (d) electrophoretic data for all species. Symbols for odontocete
families:

Physeteridae ; Ziphiidae ; Platanistidae ;

Monodontidae A ; Phocoenidae 0; Delphinidae D.

18
electrophoretic D's for the interspecies distances common to the two data sets was highly
correlated (r2 = 0.98, slope = 1.50).
The similarity of these two independent data sets may be examined further by recalculation
of the Shimura and Numachi data as a Fitch—Margoliash distance diagram. If the slope (1.50)
of the Nei D against electrophoretic D plot is used to estimate distances in a Fitch—Margoliash
diagram derived from the Shimura and Numachi data, the following comparison emerges:

Phocoenidae Ziphiidae Delphinidae

to to to
Z. node P./D. node Z. node
Present results
(Fig. 2d) 0.25 0.15 0.30
Shimura and Numachi
(1987) 0.30 0.17 0.45

We interpret this remarkable similarity as further confirmation of the strength of the data
and the correctness of our interpretation. A further indication of the robustness of this
approach may be seen from the fact that in the present work Ziphiidae is represented by
Mesoplodon bidens but by Bairdius bairdii in the Shimura and Numachi (1987) study. These
observations also reaffirm one of the main points put forward by Sarich (1977): that it is as
effective to investigate taxonomy by band counting using one individual per taxon as by ana-
lyzing large numbers of individuals to get to the Nei D kind of calculation.
Heyning (1989) has presented the results of a new cladistic, parsimony analysis of
anatomical data together with a review of odontocete taxonomy based upon the organization
of the families into a number of superfamilies. In the Heyning phylogeny, Phocoenidae and
Delphinidae were sister groups at one extremity of a midline, while Ziphiidae and a somewhat
distant Physeteridae were sister groups at the opposite extremity. The other families,
Monodontidae (nearest to Phocoenidae), Iniidae, and Platanistidae, were distributed along
the midline and connected to it by relatively short character state branches.
Heyning 's phylogeny agreed with the classical superfamily taxonomy that associates the
families Monodontidae, Phocoenidae, and Delphinidae in a superfamily Delphinoidea to the
exclusion of all other families (Heyning 1989; Rice 1984). This is quite different from the
present phylogenetic trees (Fig. 2a—d) and the results of Shimura and Numachi (1987), which
were characterized by a marked separation of Delphinidae from Phocoenidae, and, in the
present work only, also from Monodontidae. Although it is difficult to compare character
state data to distance data, it is obvious that the Heyning (1989) topology and branch lengths
are very different from the present phylogeny (Fig. 2d). The only similarity is an association
between Monodontidae and Phocoenidae. It would seem that resolution of the discrepancies
between the present phylogeny (Fig. 2a—d) and the Heyning (1989) phylogeny, together with
the classical organization of the odontocetes into superfamilies, will require more data from
additional character sets.
A time scale for odontocete evolution may be approached by consideration of the present
results and the independent data set of Shimura and Numachi (1987). The slope (1.50) of a
plot of the Shimura and Numachi Nei D against the present electrophoretic D may be used
to convert the slope of the ID against electro D plot (Fig. 1, 15.3) to a slope of 10.2 for an
equivalent plot of ID against Nei D. Comparison of this result with slopes ranging from 22
to 55 for plots of serum albumin ID against Nei D presented by Wyles and Gorman (1980)
suggest that either the pesent ID's may be low or the electrophoretic D's may be high relative
to other data sets. Shimura and Numachi have already suggested that their electrophoretic Nei
D values are not high in comparison with other data sets.
Thus it seems that the ID's may be low and may produce minimal estimates of divergence
time if we use the widely accepted calibration that 1 ID unit = 0.6 million years (Myr) of

19
evolutionary divergence time (Wilson et al. 1977; Collier and O'Brien 1985). If this
calibration is used the 2.0 ID index bars in Fig. 2a and 2b become equivalent to 1.2 Myr of
divergence time and the divergence of Monodontidae from both Delphinidae and 'Ina may
each be estimated at about 6 Myr. Similarly, the divergence of Inia and Delphinidae may be
estimated at about 12 Myr.
It may also be possible to calibrate some kind of electrophoretic clock. Sarich (1977) has
suggested that for slowly evolving loci, the type used in the present work and by Shimura and
Numachi (1987), T(Myr) = 30 Nei D. If we again use the slope of 1.5 for the plot of Nei D
against electrophoretic D, presumably T(Myr) = 45 electrophoretic D units and the 0.1 D
index bar represents 4.5 Myr of evolutionary divergence time. On this basis, divergence times
may be estimated from Fig. 2d as follows: Monodontidae—Delphinidae, 23 Myr;
Monodontidae—Phocoenidae, 10 Myr; Monodontidae—Mesoplodon, 13 Myr; and
Monodontidae—Physeter, 18 Myr.
If these estimates of divergence times are interpreted as being a minimum for the ID data
and a maximum for the electrophoretic D data, then the divergence of Monodontidae and
Delphinidae for example would date somewhere between 6 Myr and 23 Myr which is the
Miocene. These estimates seem to be in reasonable agreement with the paleontological data
which show that, although primitive forms of cetaceans (archaeocetes) are found in Eocene
deposits (Barnes 1976; Whitmore and Sanders 1976), the major radiation of odontocetes is
a Miocene event (5-25 Myr) (Barnes 1984; Barnes et al. 1985). This analysis would also
suggest that intra-family odontocete divergences are more recent Pliocene events, a result in
general agreement with the paleontological data (Barnes 1984).
This analysis of the immunological and electrophoretic results has shown that the whole
set of immunological distances may be lower than those in other comparable data sets. There
is also the possibility that one or another of the present species may have produced anomalous
results. An example of a species that yielded anomalous ID's is the New-World monkey
Auras in which an ID of only 42 was observed between human and Aotus in contrast to ID's
of 39, 35 and 36, respectively, between Old-World monkeys and human, chimpanzee, and
gorilla, and ID 's of 62-70 between human and three other New-World monkeys (Sarich and
Wilson 1966). Is there some similar anomaly in our data that could reconcile the present
results with the suggestion of Kasuya (1973) of a close relationship between Orcaella and
Delphinapterus? We think not. The average ID (2.4) between Delphinapterus and Monodon
is clearly distinct and much smaller than the ID's between these species and any other species
in the present study (approximately 7-13, Table 3). This difference in ID is not likely due to
experimental error; the standard deviations where they were measured (Table 3) ranged from
0.2 to 2.8. After the Delphinapterus—Monodon pair, the next most similar group of compa-
risons was found among the Delphinidae (including Orcaella) with an average ID of 5.9
(Table 3). We have also shown that the ID data are closely correlated to our enzyme
electrophoretic data and that the two sets of data produce very similar phylogenetic trees (Fig.
2a—d). The enzyme electrophoretic data are also closely correlated with another independent
enzyme electrophoretic data set (Shimura and Numachi 1987). All this evidence suggests that
there is nothing exceptional about the Orcaella data and we conclude that the classical
assignment of Orcaella to the family Delphinidae should be maintained. Rice (1984) has also
favored this classical taxonomy in an interpretation of the morphological data in which he
suggests that while Orcaella superficially resembles a small Delphinapterus, it is more like
Delphinids in skull morphology and air-sinus system, and that the similarity of its tympano
periotic to that of Delphinapterus may be plesiomorphic or convergent. We therefore
conclude that the present data confirm the classical assignment of Delphinapterus and
Monodon to a two species family Monodontidae and the association of Orcaella with the
family Delphinidae.

20
 Acknowledgments

We thank M. Bigg, L. Maxson, E. Mitchell, and V. Sarich for advice and J. Heyning for
a copy of a recent paper. Canada Department of Fisheries and Oceans staff, including H.
Cleator, B. Dunn, W. Doidge, L. Dueck, M. Kingsley, W. Macdonald, P. Richard, M.
Roberge, T. Smith, R. Stewart, T. Strong, R. Walker, and P. Weaver, collected and
transported Delphinapterus and Monodon tissue and blood samples from animals killed by
native hunters in the Canadian Arctic. R. Baird, J. Bannister, P. Brodie, P. -Y. Daoust, D.
Hembree, K. Langelier, J. Lien, A. Read, and P. Stacey provided samples from stranded
animals or animals taken incidentally in fishing gear. T. Carmichael, M. Daniels, and D.
Tretiak provided laboratory assistance. We thank S. Cosens, S. Innes, D. St. Aubin, and two
anonymous reviewers for comments.

References
ALLENDORF, F. W., N. MITCHELL, N. RYMAN, CLAYTON, J. W., AND J. H. GEE. 1969. Lactate
AND G. STAHL. 1977. Isozyme loci in brown dehydrogenase isozymes in longnose and
trout (Salmo trutta L.): detection and interpre- blacknose dace (Rhinichthys cataractae and
tation from population data. Hereditas 86: R. atratulus) and their hybrid. J. Fish. Res.
179-190. Board Can. 26: 3049-3053.
BAKER, R. J., R. L. HONEYCUTT, M. L. ARNOLD, CLAYTON, J. W., AND D. N. TRETIAK. 1972. Amine-
V. M. SARICH, AND H. H. GENOWAYS. 1981. citrate buffers for pH control in starch gel elec-
Electrophoretic and immunological studies on trophoresis. J. Fish. Res. Board Can. 29:
the relationship of the Brachyphyllinae and the 1169-1172.
Glossophaginae. J. Mammal. 62: 665-672. COLLIER, G. E., AND S. J. O'BRIEN. 1985. A mo-
BARNES, L. G. 1984. Fossil odontocetes (Mam- lecular phylogeny of the Felidae: Immunolo-
malia: Cetacea) from the Almejas formation, gical distance. Evolution 39: 473-487.
Isla Cedros, Mexico. PaleoBios, Museum of CRONIN, J. E., AND V. M. SARICH, 1976. Molecu-
Paleontology, University of California, Ber- lar evidence for dual origin of mangabeys
keley 42: 1-46. among Old World monkeys. Nature 260:
1976. Outline of eastern north Pacific 700-702.
fossil cetacean assemblages. Syst. Zool. FITCH, W. M., AND E. MARGOLIASII. 1967. Con-
25: 321-343. struction of phylogenetic trees. Science 155:
BARNES, L. G., D. P. DOMNING, AND C. E. RAY. 279-284,
1985. Status of studies on fossil marine mam- GASKIN, D. E. 1982. The ecology of whales and
mals. Mar. Mamm. Sci. 1: 15-53. dolphins. Heinemann Educational Books Ltd.,
BEVERLEY, S. M. 1979. Molecular evolution in London. 459 p.
Drosophila. Ph.D. thesis, University of Cali- HARRIS, H., AND D. A. HOPKINSON. 1976. Hand-
fornia, Berkeley, CA. 424 p. book of enzyme electrophoresis in Human
BEVERLEY, S. M., AND A. C. WILSON. 1982. Genetics. North-Holland Publishing Compa-
Molecular evolution in Drosophila and higher ny, Amsterdam.
diptera I. Micro-complement fixation studies HEYNING, J. E. 1989. Comparative facial anatomy
of a larval hemolymph protein. J. Mol. Evol. of beaked whales (Ziphiidae) and a systematic
18: 251-264. revision among the families of extant Odon-
CHAMPION, A. B., E. M. PRAGER, D. WACHTER, toceti. Contributions in Science. Vol. 405, Na-
AND A. C. WILSON. 1974. Microcomplement tural History Museum of Los Angeles County,
fixation, p. 397-416. In C. A. Wright [cd.] Los Angeles, CA. 64 p.
and immunological taxonomy ofBiochemal HUSSON, A. M. AND L. B. HOLTHUIS. 1974. Phy-
animals. Academic Press Inc. (London) Ltd., seter macrocephalus: The valid name for the
London. sperm whale. Zool. Meded 48 (19): 205-218.
CLAYTON, J. W., W. G. FRANZIN, AND D. N. TRE- KASUYA, T. 1973. Systematic consideration of
TIAK. 1973. Genetics of glycerol-3-phosphate recent toothed whales based on the morpho-
dehydrogenase isozymes in white muscle of logy of tympano-periotic bone. Sci. Rep. Wha-
lake whitefish (Coregonus clupeaformis). J. les Res. Inst. 25: 1-103.
Fish. Res. Board Can. 30: 187-193.

21
KLEINENBERG, S. E., A. V. YABLOKOV, B. M. PRAGER, E. M., AND A. C. WILSON. 1971. The
BECKOVICH, AND M. N. TARASEVICH. 1964 dependence of immunological cross-reactivi-
(Translation 1969). Beluga (Delphinapterus ty upon sequence resemblance among lyso-
leucas). Investigation of the species, Israel zymes. I. Micro-complement fixation studies.
Program for Scientific Translations, Jerusa- J. Biol. Chem. 246: 5978-5989.
lem. 376 p. REEVES, R. R., AND S. TRACEY. 1980. Monodon
LACÉPÈDE, B. G. E. 1804. Histoire naturelle des monoceros. Mammal. Species No. 127: 1-7.
Cétacées, Plassan, Paris. xliv + 329 p. + 16 RICE, D. W. 1984. Cetaceans, p. 447-490. In S.
plates. Anderson and J. K. Jones Jr. [cd.] Orders and
LINNAEUS, C. 1758. Systema naturae per regna families of recent mammals of the world. John
tria naturae, secundum classes, ordnines, ge- Wiley and Sons, New York, NY.
nera, species cum characteribus, diferentiis, 1990. Scientific correspondence. Mar.
synonymis, locis„ Vol. 1, Laurentii Salvii, Martini. Sci. 6(4): 359-360.
Stockholm. ii + 824 p. SARICH, V. M. 1977. Rates, sample sizes, and the
LINT, D. W. 1987. A serum albumin immunologi- neutrality hypothesis for electrophoresis in
cal approach to odontocete evolution. M.Sc. evolutionary studies. Nature 265: 24-28.
thesis, University of Manitoba, Winnipeg, Man. SARICli, V. M., AND A. C. WILSON. 1966. Quan-
116 p. titative immunochemistry and the evolution of
MAXSON, L.R. 1973. A molecular approach to the primate albumins: Micro-complement fixa-
study of hylid evolution. Unpublished Ph.D. tion. Science 154: 1563-1566.
thesis, University of California, Berkeley and SHIMURA, E., AND K.-I. NUMACHI. 1987. Genetic
San Diego, CA. variability and differentiation in the toothed
MAXSON, L. R., R. HIGFITON, AND D. B. WAKE. whales. Sci. Rep. Whales Res. Inst. 38:
1979. Albumin evolution and its phylogenetic 141-163.
implications in the plethodontid salamander SHOWS, T. B., P. J. MCALPINE, C. BOLICIleIX, F. S.
genera Plethodon and Ensatina. Copeia 1979: Collins, P. M. CONNEALLY, AND 17 OTHERS.
502-508. 1987. Guidelines for human gene nomencla-
MAXSON, L. R., AND R. D. MAXSON. 1979. Compa- ture. Cytogenet. Cell Genet. 46: 11-28.
rative albumin and biochemical evolution in SICILIANO, M., AND C. R. SHAW. 1976. Separation
plethodontid salamanders. Evolution 33: and visualization of enzymes on gels, p.
1057-1062. 185-209. In I. Smith [cd.] Chromatographic
MAXSON, L. R., AND A. C. WILSON. 1975. Albu- and electrophoretic techniques. Vol. II, William
min evolution and organismal evolution in Heinemann Medical Books Ltd., London.
tree frogs (Hylidae). Syst. Zool. 24: 1-15, SIMPSON, G. G. 1945. The principles of classifica-
MAXSON, R. D., AND L. R. MAXSON. 1986. Micro- tion and classification of mammals. Bull. Am.
complement fixation: A quantitative estimator Mus. Nat. Hist. 85: 1-350.
of protein evolution. Mol. Biol. Evol. 3: STEWART, B. E., AND R. E. A. STEWART. 1989.
375-388. Delphinapterus leucas. Mammalian species
MCALPINE, P. J., C. BOUCHEIX, A. J. PAKSTIS, No. 336: 1-8.
L. C. STRANC, T. G. BERENT, AND T. B. SHows. TSUYUKI, H., E. ROBERTS, R. H. KERR, AND A. P.
1988. The 1988 catalog of mapped genes and RONALD. 1966. Micro starch gel electrophore-
report of the nomenclature committee. Cyto- sis. J. Fish. Res. Board Can. 23: 929-933.
genet. Cell Genet. 49: 4-38. WHITMORE F. C., JR., AND A. E. SANDERS. 1976.
NIELSEN, O., R. K. KELLY, W. R. LILLIE, J. W. Review of the Oligocene Cetacea. Syst. Zool.
CLAYTON, R. S. FUJIOKA, AND B. S. YONEYA- 25: 304-320.
MA. 1989. Some properties of a finite cell line WILSON, A. C., S. S. CARLSON, AND T. J. WHITE.
from beluga whale (Delphinapterus leucas). 1977. Biochemical evolution. Ann. Rev. Bio-
Can. J. Fish. Aquat. Sci. 46: 1472-1477. chem. 46: 573-639.
O'BRIEN, S. J., W. G. NASH, D. E. WILDT, M. E. WYLES, J. S., AND G. C. GORMAN. 1980. The
BUSH, AND R. E. BENVENISTE. 1985. A mole- albumin immunological and Nei electropho-
cular solution to the riddle of the giant panda's retic distance correlation: A calibration for the
phylogeny. Nature 317: 140-144. saurian genusAnolis (Iguanidae). Copeia 1980:
PALLAS, P. S. 1776. Reise durch versheidene 66-71.
Provinzen des Russichen Reichs, Vol. 3, Pt. 2,
St. Petersbourg.

22
The Distribution and Abundance of Belugas, Delphinapterus leucas,
in Eastern Canadian Subarctic Waters: A Review and Update
P. R. Richard, J. R. Orr
Department of Fisheries and Oceans, Central andArctic Region,
Freshwater Institute, 501 University Crescent,
Winnipeg, Manitoba, Canada, R3T 2N6

and D. G. Barber
Earth Observation Laboratory, Department of Geography, ENV2,
University of Waterloo, Waterloo, Ontario, Canada N2L 3GI

RICHARD, P. R., J. R. ORR, AND D. G. BARBER. 1990. The distribution and

abundance of beluga, Delphinapterus leucas, in eastern Canadian waters:
a review and update, p. 23-38. In T. G. Smith, D. J. St. Aubin, and J. R.
Geraci [ed.] Advances in research on the beluga whale, Delphinapterus
leucas. Can. Bull, Fish. Aquat. Sci. 224.

Aerial surveys document a continuous July distribution of belugas along the

coast of Hudson Bay from Eskimo Point, N.W.T., to James Bay and the occurrence
of a small July population in northern Hudson Bay, and confirm that few belugas
occupy the southeast Baffin Island coast in August, outside of Cumberland Sound.
The western Hudson Bay population is estimated to number in excess of 23 000
individuals and more than 1 300 belugas were seen along the southern Hudson Bay
coast. Northern Hudson Bay has a population estimated at more than 700 animals,
while fewer than 500 were found in southeast Baffin Island. We review arguments
pertaining to stock discreteness of subarctic beluga populations. We conclude that
evidence for geographic isolation between the eastern and western Hudson Bay
populations is inconclusive and that the stock identity of northern Hudson Bay
belugas is unknown. Southeast Baffin belugas could be a separate genetic stock,
as evidenced by their morphometric differences from other subarctic populations,
but further evidence is needed to discriminate the stocks within these populations.

Des inventaires aériens ont permis de décrire une distribution continue de

bélugas en juillet le long des côtes de la baie d'Hudson entre Eskimo Point, T.N.O.,
et la baie James, ainsi que la présence d'une petite population dans le nord de la baie
d'Hudson. Ils ont aussi permis de confirmer qu'en dehors de la baie de Cumber-
land, en août, peu de bélugas occupent les côtes du sud-est de la terre de Baffin. On
estime que la population de l'ouest de la baie d'Hudson dépasse 23 000 individus
et plus de 1 300 bélugas ont été observés tout au long des côtes du sud de la baie
d'Hudson. Le nord de la baie d'Hudson a une population de plus de 700 individus,
tandis que moins de 500 ont été trouvés dans le sud-est de la terre de Baffin. Nous
examinons différents arguments sur la séparation des stocks de bélugas sub-
arctiques et en venons à la conclusion qu'il n'y a pas de preuves concluantes pour
croire à la séparation géographique des populations de l'est et de l'ouest de la baie
d'Hudson et que l'on ne sait pas à quel stock appartient la population du nord de
la baie d'Hudson. Les bélugas du sud-est de la terre de Baffin pourraient appartenir
à un stock génétique distinct, comme le suggére des différences morphométriques
avec les autres populations sub-arctiques, mais des preuves supplémentaires sont
nécessaires avant de conclure sur l'identité de tous ces stocks.

23
 Introduction

This paper pertains to Canadian beluga populations south of the Arctic circle and north of
the Gulf of Saint-Lawrence. Summer concentrations of these belugas are commonly called
western Hudson Bay population, eastern Hudson Bay population and Cumberland Sound or
southeast Baffin Island population. The paper also reports on belugas near the Ontario coast
of Hudson Bay and in the northern part of Hudson Bay.
Stock separation and numbers of subarctic beluga stocks are not well defined. We present
results from extensive surveys and opportunistic sightings in Hudson Bay, southern Foxe
Basin and south-east Baffin Island. We give population estimates for western Hudson Bay,
northem Hudson Bay and southeast Baffin Island and report on numbers observed in southern
Hudson Bay. Our distribution data combined with recently published results of surveys in
eastern Hudson Strait and Ungava Bay (Smith and Hammill 1986) yield a new pattern of
beluga summer distribution in Hudson Bay and open new questions about stock definition
of subarctic beluga populations.

Methods

Aerial surveys covered coastal and near-offshore waters of four areas (Fig. 1): the southeast
Baffin area which includes all coastal waters of Baffin Island between Exeter Sound, in Davis
Strait, and Markham Bay, in Hudson Strait; the northern Hudson Bay area which covers
waters between Rankin Inlet and Baffin Island up to 67°N latitude; the "southern Hudson
Bay" area which covers waters from Cape Henrietta Maria at the mouth of James Bay to the
Kaskattama River (57°N, 90°W) near the Ontario—Manitoba border; the "western Hudson
Bay" area which covers waters from the Kaskattama River, in Manitoba, to Rankin Inlet,
Northwest Territories (N.W.T.). Reconnaissance surveys were flown to locate beluga
concentration areas. Census surveys centered on these areas of concentration and were
designed to estimate population size. All census surveys were systematic strip transect
surveys (Fig. 2-4). We used a systematic sampling design because it is more operationally
convenient than random surveys and is also more efficient than random sampling as a means
of mapping distribution (Caughley 1977). It is more precise to estimate population size when
sample units are unequal in size (Cochran 1977) or when sample counts are serially correlated
(Kingsley and Smith 1981), as was often the case in our surveys.
Systematic surveys were planned by delimiting a survey area, dividing it into parallel strips
of desired width (e.g.: 1.6 km in visual surveys) and selecting the desired sampling fraction
(e.g.: 10% or one transect flown for every 10 strips). The sampling fraction varied from
survey to survey, depending on available funds, but generally we tried to obtain the highest
sampling fraction affordable (25%) in known areas of beluga concentration. The first transect
was chosen by randomly selecting a number from one to 'n (eg.: from one to ten) where 'n'
is the total number of strips divided by the sampling fraction and the following were chosen
at equal distance from the first (e.g.: every 10 strips).
All reconnaissance surveys and several census surveys (Table 1) were done visually. The
1981 reconnaissance surveys were flown in a De Havilland Beaver with one observer in the
right front seat and the other in the left rear. Navigation was done by dead-reckoning using
land forms. All other visual surveys were flown in a De Havilland Twin Otter with observers
seated directly behind wing struts, one on each side of the aircraft. Struts and windows were
marked to delimit a survey strip of 800 ni on each side of the aircraft. Flight position was
determined by pilot and navigator with Omega Navigation and were regularly checked
against landmarks. Air speed and altitude were maintained at 185 km • 11-1 and 305 ni during
visual surveys, except in southeast Baffin surveys where we kept an altitude of 457 m.
Cumulative counts were recorded every five min.

24
N67

NORTHWEST Foxe Basin 22

TERRITORIES
N65 Baffin
Island
25
Davis
Keewatin Hudson Strait
4
Strait
N60
o
HUDSON BAY Ungava
Bay

5 ____

MANITOBA 28
Eastmain
30
N55 LABRA
31 DOR
ornes QUEBEC
_pay
ONTARIO
0 100 200

km
N51

W100 W95 W90 W85 W80 W75 W70 W65 W60

FIG. 1. Study area with placenames. (Note: placenames are numbered clockwise on the map.)
Belcher Islands 29 Kaskattama River 4
Cape Henrietta Maria 1 Little Whale River 30
Churchill (River & munic. of) 6 Lyon Inlet 14
Clearwater Fiord 20 Markham Bay 26
Coral Harbour (hamlet of) 17 Nastapoka River 28
Cumberland Sound 23 Nelson River 5
Daly Bay 10 Pangnirtung (hamlet of) 21
East Bay 18 Rank in Inlet (hamlet of) 9
Eskimo Point 8 Repulse Bay (& hamlet of) 13
Exeter Sound 22 Roes Welcome Sound 11
Foxe Peninsula 19 Seal River 7
Frobisher Bay 27 Severn River 3
Frozen Strait 15 Southampton Island 16
Great Whale River 31 Wager Bay 12
Hall Peninsula 24 Winisk River 2
Igaluit (munie. of) 25

25
TABLE 1. Procedures of subarctic beluga census surveys 1982-87.
Fig. Survey Survey Method Altitude Transect Strip Transect
date(s) design of recording spacing width orientation
2 July 14-15 Systematic Photographic 914m 5.5 km 1.34 km 900-270 0
1987 225X225 mm
2 July 17-18 Systematic Visual 305 m 15 km 2 X 0.8 km 40°-220°
1987
2 July 19 Systematic Photographic 1219m 5.5 km 1.88 km 90°-270°
1987 225X225 mm

3 July 23 Systematic Visual 305m blockl: 19 km 2 X 0.8 km block 1: 90°-270°

1982 block2: 11 km block 2: 0°-180°
3 July 24- Random Photographic 914m random 0.45 km 90°-270°
25, 1982 96x122 mm
3 July 26- Systematic Photographic 914m 9.4 km 0.45 km 90°-270°
27,1983 96x122 mm
3 July 27 Systematic Photographic 914m 5.5 km 1.34 km 90°-270°
&29,1984 225x225 mm
3 March18-21 Systematic Visual 305m blockl: 7.5 km 2 x 0.8 km block 1: 45°-225°
1982 block2: 13 km block2: 0°-180°
block3: 19 km block3: 0°-180°

4 August 18- Systematic Visual 457 m 9.5 km 2 X 0.8 km 90°-270°

22, 1985
4 August 16 Systematic Visual 457 m 9.5 km 2 X 0.8 km 90°-270°
&23,1986
 Photographic surveys were used to census areas of high concentration. The northern
Hudson Bay photo-surveys are an exception in that they were designed to census narwhal
concentrations but also provided us with beluga counts. We chose to conduct photo-surveys
in concentration areas because they do not suffer from problems of reliability due to observer
fatigue, boredom or bias such as are encountered in visual surveys. Beluga detectability
changes with distance are much reduced by vertical photography. The 90° angle of the camera
allows a more uniform water penetration accross the image plane compared with the oblique
angles of view of visual observers. Glare and fog are also less of a problem at high angles.
Variation in image coverage due to airplane roll can almost be eliminated by maintaining the
camera level during operation, whereas there is no way of correcting strip width variation
during visual surveys. Observer errors during film counts do exist but they can be corrected
by recounts or multiple observers. Photo-surveys are particularly useful in estuaries and other
areas where large groups or herds are encountered and visual observers have inadequate time
for an accurate count.
Photographic census surveys (Table 1) were flown in a De Havilland Twin Otter with aerial
cameras mounted in a port at the back of the aircraft. In 1982 and 1983 we used a Linhof
Aeroteknica (format 953120 mm) with a 150 mm lens and Kodak Ektachrome transparency
film. From 1984 to 1987, we used a Wild-Leitz RC8 (format 2253225 mm) with a 153 mm
lens and Kodak aerocolor 2445 negative film. Cameras were triggered by an intervalometer
at settings calculated to obtain a stream of adjoining frames. Flight position was determined
by Inertial Navigation in 1984 and Omega Navigation in all other years. Land references were
used to verify position and update navigation systems. Air speed and altitude were maintai-
ned at 185 km • h- ' and 914m, except for the 19 July 1987 survey of western Hudson Bay
which was flown at 1219 m.
The aerial film was scanned with a Nikon SMZ-1 dissecting scope on a Richards GFL-940-
MC light table. We recorded belugas identified on each frame as well as any image which
might be a whale. Frames with unidentified images were re-examined a second time for final
classification.
Surveys were flown between July 1981 and August 1988 (Fig. 2-4). Southern and western
Hudson Bay were first surveyed visually in July 1981 (Fig. 2). A second visual reconnais-
sance survey in August 1981 covered western and northe rn Hudson Bay (Fig. 2-3). From
1982 to 1984, three systematic census surveys (visual and photographic) were flown in July
in northern Hudson Bay (Fig. 3). In the southeast Baffin area, both systematic and reconnais-
sance surveys (visual and photographic) were flown in August 1985 and 1986 (Fig. 4). In July

E) July 15,
1987
Seal R.

Churchill R.

F) July 14,
Seal R. fA) July 21-24, C) July 17-18, 1987
■! 1 26-27, 1981 1987 Nelson R.
B) August
Churchill R. 25-26,1981
I

Nelson R. D) July 21, G) July 19,

1987 1987

Nelson Fl.

FIG , 2. Transect lines flown during July and August surveys of western and southern Hudson Bay.

27
1987, census surveys covered concentration areas of western Hudson Bay and a reconnais-
sance survey was flown in southern Hudson Bay (Fig. 2). Systematic and reconnaissance
visual surveys in March 1983 and in August 1988 (Fig. 3) were also flown in northern Hudson
Bay.

July 23, 1982

jC
Z2
t •

0 25
ITT
TY1

*•4.,

Southampton I.
Southampton I.

March 18-22, 1983

— March 20-21 1983

July 26-27, 1983

0 25
1""kr
in j

Southampton I. Southampton I.

Baffin I.

-- Aug. 18-25, 1981

— Aug. 13-17, 22, 1988 \J Quebec

Fia. 3. Transect lines fl own during March, July and August surveys of northe rn Hudson Bay.

28
 Estimates of the total numbers were derived from each systematic census survey. The
number of belugas and the surveyed area were totalled for each transect. In the case of photo-
surveys, the area of each transect was the sum of frame areas. Land or fog encountered on the
strip or photos were removed from the area total. Transect counts and areas were summed and
the ratio of total count over total area gave a mean density. Estimates for each census area
were obtained by multiplying the beluga density by the total census area. In cases where the
census area was divided into strata, estimates from all strata were summed to obtain an
estimate for the whole survey area.
We did not apply correction factors to the estimates to compensate for submerged animals
missed during the flights. Water clarity varied greatly within and between surveys, from the
silted waters of estuaries to the relatively clear offshore waters. Beluga diving behavior
probably also varies from one area or time to the next but we had no way of measuring
behavior during surveys. Correction factors are not yet available to compensate for these
problems. Sergeant and Hoek (1988) used the overlap of adjacent photos to obtain a minimum
correction for diving animals but our survey data is composed of non-overlapping visual
intervals or photos. Consequently, in areas of high turbidity such as the head of the Nelson
estuary, the results would tend in the direction of underestimation.
Methods described by Kingsley et al. (1985) were used to obtain estimates of variance for
systematic surveys. Confidence limits were calculated as in Smith et al. (1985). Minimum
and maximum population estimates for each survey were calculated by multiplying the upper
and lower confidence limits by the total census area.

Cumberland Sound
area
Aug. 20-23, 1985

0 1 00 Aug. 23-24, 1986

km

Baffin Island

Hall
Peninsula

Frobisher Bay
area
Meta Inco-
Aug. 16-18, 1985
gnita Peninsula
Aug. 16-18, 1986

FLOE 4. Transect lines flown during August surveys of southeast Baffin Island.

29
 In one case, on 15 July 1987, right after the systematic photo-survey of the Churchill—Seal
River area, we obtained an additional photographic pass over a large concentration near the
Seal River. The count from these photos (3 758) was added to the count (331) from the
portions of the transects which did not overlap with that pass. That total rounded to the nearest
hundred (4 100) was used in place of the lower confidence limit (1 200).

Results

During 1981 reconnaissance surveys along the coast of western Hudson Bay (Fig. 2a), we
found belugas to be widely distributed near the coast of Manitoba as well as along the coast
of Ontario and as far east of the Severn estuary as 86°409W where surveys ended. Mean
densities at the heads of the Nelson, Seal, and Churchill estuaries were, respectively, 8, 5, and
8 belugas • km-2 during July surveys and 1, 4, and 5 belugas • km-2 during August surveys.
Only 62 belugas were seen north of the Manitoba border (60°N) during the July 1981
coastal reconnaissance surveys but, during similar surveys in August 1981, 329 belugas were
found north of the border, including a small herd (128) at the mouth of the Thlewiaza River
(60°309N). The northernmost sighting in July along the southern Keewatin coast was at
latitude 60°409N while, in August, we sighted a beluga at latitude 62°209N.
In 1987, during visual and photo surveys of western Hudson Bay (Table 1, Fig. 2b—d),
belugas were distributed throughout all survey transects from Eskimo Point to the Kaskat-
tama River (Fig. 1). During the first photo survey of the Nelson estuary (Fig. 2c), we noted
that survey lines did not cover the entire beluga concentration and we later repeated the survey
over a broader area, farther north and more offshore from the estuary (Fig. 2d).
On this second photo survey of the Nelson concentration, belugas were again found
throughout transects and as far offshore as survey lines reached, 110 km from any coast and
160 km from the mouth of the river. Beluga density was highest along the two transects
closest to the mouth of the river (average 5 belugas • km-2). Beluga density on the other five
transects was lower but still substantial (average 1 beluga • km-2).
The 14-15 July 1987 photo surveys (Fig. 2c) centered around the estuaries of the Nelson
River and of the Churchill and Seal rivers, respectively. The total mean estimate for the two
surveys combined is 12 500 (Table 2) but our coverage of the concentration in or near the
Nelson estuary was incomplete. The second survey of the Nelson concentration (Fig. 2d)
covered more offshore areas and yielded a mean estimate of 19 500 belugas (Table 2). When
summed with the 15 July 1987 Churchill—Seal mean estimate of 5 600 (Fig. 2c, Table 2), this
gives a total mean estimate of 25 100 for the western Hudson Bay population.
The fact that there were 4 d between the Churchill Seal and Nelson surveys raises the issue
of possible movement of belugas between the two areas. A case in point is a 3 ni radio-tagged
male beluga released on 12 August 1988 in the Churchill river that was relocated in the
Nelson estuary five days later (P. Weaver, Dep. Fisheries and Oceans, Winnipeg, Manitoba,
personal communication). If movement of large numbers of belugas had occurred during the
interval between the two July 1987 photo surveys, counts would be biased upwards or
downwards depending on the direction of that movement. Nevertheless, the 17-18 July 1987
visual survey of approximately the sanie area (less the farthest offshore portion) gave very
similar results, a total of 23 000 belugas (Table 2). As explained earlier, these results are
uncorrected for submerged animals and therefore give negatively-biased estimates of the
population size.
During 21 July 1987 reconnaissance surveys of the Ontario coast, which covered coastal
waters from the Kaskattama River to Cape Henrietta Maria and back (Fig. 2b), we sighted
an additional 1 299 belugas. The first transect 3 km inshore had most of the belugas, a total
of 1 269 belugas, of which 232 belugas were concentrated at the estuary of the Severn River
and 393 at the estuary of the Winisk River. Along the second transect flown parallel to the
coast about 28 km offshore, we sighted a total of 30 belugas dispersed in small groups.

30
TABLE 2. Counts and estimates from systematic surveys performed in western and northern Hudson Bay (note: estimates over 100 rounded to nearest 100).
Survey Census Coverage Total Estimate 95% confidence
area (km 2) count limits
Western Hudson Bay

Nelson (14/07/1987) 3 015 24% 1 704 6 900 6 000-7 900

Churchill-Seal (15/07/1987) 3 035 24% 1 367 5 600 4 100-26 000
Total 12 500 10 100-33 900

Churchill-Seal (17/07/1987) 9 854 10% 1 184 11 300 3 300-38 700

Nelson (18/07/1987) 6 569 10% 1 209 11 700 7 700-17 800
Total 23 000 11 000-56 500

Nelson (19/07/1987) 15 022 10% 1 912 19 500 14 200-26 800

Northern Hudson Bay

(23/07/82) Strat 1 2 769 11% 98 300 50-1 400

(23/07/82) Strat 2 5 396 7% 32 400 100-2 000
Total 700 200-3 300

(26-27/07/83) 8 103 10% 104 1 000 621-1 627

(27/07/84) 5 712 17% 109 700 400-1 100

(29/07/84) 2 069 11% 11 100 30-300
Total 800 400-1 400
 In northern Hudson Bay (Fig. 3), the March survey of Roes Welcome Sound did not result
in any sighting of belugas. Similarly, no beluga was seen during the reconnaissance flight of
Frozen Strait and the mouth of Lyon Inlet. July 1982-1983 census surveys were limited to
the area around Repulse Bay-Frozen Strait. Estimates of beluga population size from these
surveys vary between 700 and 1 000 belugas (Table 2). In addition, we photographed a herd
of 143 belugas on 27 July 1983 in the estuary of the Canyon River, Southampton Island, south
of Frozen Strait.
In August 1981 and 1988, during reconnaissance surveys, belugas were dispersed in small
groups all around Southampton Island, except for one large herd of 685 photographed at the
head of East Bay on 17 August 1988. We also found 18 belugas in Foxe Basin, north of Foxe
Peninsula.
In southeast Baffin Island, during August 1985 reconnaissance and census surveys, a total
of 407 belugas were counted. The majority (398) were photographed in Clearwater Fiord, in
or near the estuary of the Ranger River. Nine were sighted near the mouth of Frobisher Bay.
There was no sighting along Hall Peninsula or along the Hudson Strait coast of Baffin Island.
During 1986 reconnaissance and census surveys, 442 belugas were photographed from the
air near the mouth of Clearwater Fiord and, at the same time, 43 were counted by a ground
crew from a cliff overlooking the estuary of the Ranger River in Clearwater Fiord. Two more
belugas were sighted by aerial observers in Cumberland Sound, near Clearwater Fiord, and
one near Frobisher Bay, off Hall Peninsula.
Again, no correction factors for submerged animals were applied to these counts but the
442 belugas photographed in 1986 were travelling in clear water, moving fast and making
short, shallow dives and we feel certain that few if any were missed by the camera. The 43
whales counted from a cliff in Clearwater Fiord were in the silted waters of the Ranger estuary
but they were relatively easy to count because the whales were well dispersed.
The difference in herd count between 1985 and 1986 is due to the fact that the 1985 count
was obtained from photographs of the herd in the silted water of the Ranger River in
Clearwater Fiord compared to the 1986 count in the clear water at the mouth of the fiord.
Although a systematic survey design was used for these surveys, a population estimate cannot
be derived from these data because of the extreme clumping of this relatively small
population.

Discussion
In managing beluga populations, it is important to know the stock size and identity of
animals taken at different hunting locations. The term "population" is used here in the sense
of the total number of belugas occupying a defined area. The term "stock" can be defined as
a management unit, a population in particular areas where harvesting occurs, or as an isolated
population, one which increases solely by reproduction and not through emigration from
other populations. In recent beluga literature, the term "stock" has meant the latter of these
definitions. Summer concentrations in or near estuaries have been hypothesized to be discrete
stocks. The problem for managers is that most hunting of subarctic belugas does not presently
take place in estuaries and it is not clear if belugas landed by any particular community belong
to one "stock" or several "stocks."
Subarctic belugas have been hypothesized to belong to four discrete "stocks": the western
Hudson Bay "stock" (Sergeant 1973, 1981), the eastern Hudson Bay "stock" (Finley et al.
1982), the Ungava Bay "stock" (Finley et al. 1982) and the Cumberland Sound or southeast
Baffin "stock" (Brodie et al. 1981; Richard and Orr 1986). The discussion that follows will
review available information pertaining to the size and discreteness of these "stocks" and the
possibility of mixing between "stocks."
Our results show that the western Hudson Bay beluga "stock" which summers along the
Manitoba coast numbered about 23 000 animals in 1987. In addition, we observed at least
1 300 belugas along the Ontario coast. These numbers far exceed previous estimates of the

32
western Hudson Bay beluga "stock" obtained from 1965 surveys (Sergeant 1973, 1981). This
difference can be explained in two ways. One is that the population has increased substan-
tially in the 22 yr that separated the two surveys. The other is that survey coverage and
methods of estimation were so different that they produced very different results. It is also
possible that both factors have contributed to this difference.
Comparison of the 1965 (Sergeant 1973, 1981) and our 1987 surveys suggests that the early
estimates may have been low for the Nelson estuary because of extrapolations from very low
coverage of the upper Nelson estuary (2.4%) and no coverage of the offshore. In the 19 July
1987 survey of the Nelson estuary, beluga density was lower in the offshore portion than in
the portion inshore corresponding to Sergeant's coverage of the estuary (approx. 1 beluga •
km-2 offshore as compared to 5 belugas • km-2 inshore). Despite that, the offshore portion
amounted to twice as many belugas because it was 10 times the area of the inshore estuarine
portion. Sampling error is also a probable factor in the 1967 survey which was flown with no
apparent sampling design. The same comments apply to the Churchill—Seal area where our
systematic coverage was greater and farther offshore.
The hypothesis that the western Hudson Bay population has increased between 1965 and
1987 cannot be tested by comparing the two studies. Errors of estimation in the 1965 survey
due to the lack of a sampling design and to low and incomplete coverage are just as likely to
produce a large discrepancy with the 1987 results as a population increase would. A greater
understanding of beluga distribution and movements and surveys with comparable coverage
and methods would have been needed to document a change in the western Hudson Bay
population's size.
Census surveys of northern Hudson Bay gave mean estimates of 700 and 1 000 belugas in
July (Table 2). In addition, individuals or small pods have frequently been reported along the
south coast of Southampton Island (Sutton and Hamilton 1932) and schools of 100 or more
have also been observed in Wager Bay in summer (Degerbol and Freuchen 1935; Anon.
1978). Our 1988-89 reconnaissance surveys found a group of over 600 belugas in East Bay
in both July and August (Richard, unpubl. data). These surveys and opportunistic sightings
suggest that a population of 1 000 or more summers in northern Hudson Bay every year.
In eastern Hudson Bay and James Bay, belugas were recently estimated to number about
1 400 and 1 200, respectively, while the Ungava Bay population was deemed too low to
derive a population estimate (Smith and Hammill 1986).
The summer distribution of belugas in southeast Baffin Island is not limited to the
concentration in Cumberland Sound. In the past, sightings in August in Frobisher Bay have
varied from a few individuals to as many as 380 (MacLaren Marex Inc. 1978, 1979, 1980).
Catches of belugas are frequently reported in summer at Iqaluit and Lake Harbour (D. Pike,
DFO Iqaluit, personal communication). These observations motivated us to survey as large
an area of southeast Baffin Island's coastal waters as possible.
Despite the wide coverage of our surveys, we found only small numbers of belugas other
than the herd seen in both years at the head of Cumberland Sound. We therefore conclude that
there were fewer than 500 belugas in the southeast Baffin summer population in 1985-86.
Considering that the annual beluga catch in Pangnirtung and Iqaluit exceeds estimates of pop-
ulation replacement rate (Richard and Orr 1986), it is probable that the population has
declined further since these surveys. This decline is consistent with an earlier decline over
several generations described by Brodie (1971).
Four lines of reasoning have been used to suggest stock discreteness of subarctic beluga
summer concentrations: morphometric differences (Sergeant and Brodie 1969; Finley et al.
1982), geographic isolation (Sergeant and Brodie 1975; Finley et al. 1982), site specificity
in estuarine use (Finley et al. 1982; Smith and Hammill 1986) and depletion through long-
term exploitation (Mitchell and Reeves 1981).
Sergeant and Brodie (1969) argued for stock isolation when they observed that western
Hudson Bay belugas had a smaller adult size and girth at age than Cumberland Sound belugas.

33
Eastern Hudson Bay and Ungava Bay belugas, on the other hand, have the same size at age
as do western Hudson Bay belugas and those captured in Hudson Strait in the fall (Finley et
al. 1982). The Hudson Strait animals are thought to be migrants from western Hudson Bay.
These assumptions have been questioned because sample sizes used for comparisons were
small and subject to ageing errors, particularly for the Ungava Bay animals, and they were
made without detailed statistical analysis (Perrin 1980; see also Doidge 1990).
The argument of geographic isolation was also used for delimiting stocks. In 1980, the
International Whaling Commission distinguished year-round resident beluga populations
and summering beluga populations which share common wintering grounds (Perrin 1980).
The western Hudson Bay beluga population, then estimated at about 10 000 (Sergeant 1973),
was thought to winter in the leads of Roes Welcome Sound and to be a resident population,
geographically isolated year-round from other populations. The eastern Hudson Bay and
Cumberland Sound populations, estimated respectively at "a few hundred" and 500 (Perrin
1980) were thought to share the same wintering area in Hudson Strait but to be separate in
summer.
Surveys in March 1982 failed to find a large concentration in northern Hudson Bay, the
hypothesized wintering area for the western Hudson Bay population (Sergeant 1973), while
surveys of Hudson Strait resulted in a much larger estimate than the eastern Hudson Bay and
Cumberland Sound populations combined (Finley et al. 1982). It was therefore concluded
that the western Hudson Bay "stock" must also winter in Hudson Strait. It was still thought
to be geographically separate from the other populations in summer (Finley et al. 1982).
Our own March surveys found no beluga in Roes Welcome Sound or Frozen Strait but local
sources have reported to us occasional winter sightings at the floe edge of Daly Bay, in Lyon
Inlet and off Cape Bylot (65°209N, 84°109 W), northern Southampton Island. Belugas have
also been seen in James Bay during winter months (Jonkel 1969) and in Hudson Bay near the
Belcher Islands (Breton et al. 1984). These reports are relatively small in numbers and do not
suggest that a large population of belugas winters in either Hudson Bay or James Bay.
Various authors have speculated that summer concentrations constitute discrete stocks
(Sergeant 1973; Brodie et al. 1981; Finley et al. 1982). Finley et al. (1982), referring to the
western Hudson Bay, eastern Hudson Bay and Ungava Bay summer concentrations,
concluded that "there is clearly no mixing among these three groups during the summer." Our
survey results combined with those of Smith and Hammill (1986) contradict that assertion.
Belugas were found throughout James Bay and were widespread between the eastern Hudson
Bay coast of Northern Quebec and the Belcher Islands (Smith and Hammill 1986). Belugas
are apparently regular summer occupants of southern Hudson Bay. Johnston (1961) mentions
that "during summer months, the schools of beluga move randomly along the coastline, with
large concentrations centered around the bigger river mouths." He reports sightings of 100
belugas at both the Severn and Winisk estuaries in July 1961. Belugas have been reported by
local Indians before spring breakup at both the Severn and Winisk estuaries (Johnston 1961).
Sergeant (1968) also mentions beluga "herds extending at least to Winisk" in late July.
Herds of 100-150 belugas have been seen near the mouth of the Winisk River on 1 July 1979
and 25 June 1982 (K. F. Abraham, Ontario Ministry of Natural Resources, Moosonee,
Ontario, personal communication). When these results are combined with our own from the
coasts of Manitoba and Ontario, it becomes clear that belugas have an essentially continuous
distribution along the Hudson Bay and James Bay coasts of the three provinces in summer.
It is unclear what route is taken by belugas to reach the western Hudson Bay and Ontario
summering areas. Although there are reports of spring sightings of belugas near the Keewatin
coast (Sergeant 1973; Reeves and Mitchell 1989), belugas are rarely seen or hunted in spring
and early summer despite frequent seal hunting trips at the floe edge. Most catches in
Keewatin communities occur in August and few occur before the end of July (Gamble 1984,
1987a,b, 1988). If 23 000 belugas migrated in May and June near the Keewatin coast, that
would not go unnoticed by residents who are always eager to hunt belugas. On the east side

34
of Hudson Bay, reports of a spring southward migration of belugas have been well
documented (Finley et al. 1982; Breton et al. 1984).
In the fall, a return northward migration occurs along both the east and west coasts of
Hudson Bay (Sergeant 1973; Finley et al. 1982) . This northward migration is apparent from
catch records and local informants (Breton et al. 1984; Gamble 1984, 1987a,b, 1988) and was
demonstrated at least for the western Hudson Bay by recoveries at Whale Cove and Repulse
Bay of belugas tagged at the Seal River (Sergeant 1973). Our 1981 surveys show a reduction
in density on the Manitoba coast in August and an extension of the range of belugas northward
along the Keewatin coast.
The stock identity of northern Hudson Bay belugas is unclear. Their distribution appears
not to overlap with western Hudson Bay belugas until late summer. According to local
hunters, belugas are seen near Repulse Bay and Frozen Strait as early as April or May
(Repulse Bay and Coral Harbour Hunters and Trappers Association, personal communica-
tion), but none are seen along the Keewatin coast before late July or August (see above).
Repulse Bay hunters also report that more belugas are seen in the fall than in summer,
presumably as a result of the western Hudson Bay return migration through that area.
Some belugas winter along the Labrador coast (Boles 1980; McLaren and Davis 1982).
They were once found in summer along the central and northern coasts of Labrador but it is
unsure how many still remain (Boles 1980). Sightings of approximately 200 belugas were
made during 8 and 10 June 1979 surveys but there is no information on numbers found later
in the summer (Boles 1980). According to Degerbol and Freuchen (1935), some Labrador
belugas move into Ungava Bay and Hudson Strait in the spring.
The above observations show that mixing between neighboring populations is possible
during a significant portion of the year. The exchange potential is greatest in winter when
most subarctic belugas occupy a small geographical area from northern Hudson Bay to
southeast Davis Strait and the Labrador Sea. More importantly, mating is thought to occur
in April or May (Sergeant 1973; Brodie 1971) while many belugas are still in their wintering
area, which means that genetic mixing through cross-breeding could also take place.
The argument that Canadian subarctic belugas may be site-specific in summer was first
suggested by Brodie et al. (1981) although the concept of homing on estuaries was implicit
in earlier works of Tomilin (1967) and Kleinenberg et al. (1969). Caron and Smith (1990)
give evidence of philopatry (site fidelity) and site tenacity for individually recognizable
belugas using the Nastapoka estuary in eastern Hudson Bay. They present data which
suggests that there were four times more adult females than males in the Nastapoka estuary
during their study and that neonates comprised 16-19% of the total number of individuals
observed. This preponderance of calves and females indicates that other age and sex classes
were underrepresented in the estuary.
Since 68-80% of animals censused were observed more than 10 km from the eastern
Hudson Bay coast (Smith and Hammill 1986), a substantial portion of the eastern Hudson Bay
population may not, at least in a given year, be as site tenacious as the river occupants. On
the other hand, if females and calves are more numerous in estuaries and more site tenacious
than males and juveniles, female-calf concentrations then form seasonally isolated stocks of
their own. If so, the survival of such estuarine female stocks would be of specific concern
because eradication of any such stock would mean reduction in the species' genetic diversity.
Another argument that has been used to infer stock discreteness is the long term decline in
catches that some populations have experienced over years of commercial exploitation
(Brodie 1971; Mitchell and Reeves 1981). This is considered the strongest line of evidence
available for the Cumberland Sound population because it suggests that the effect of
exploitation has not been moderated by recruitment from adjacent populations (Mitchell and
Reeves 1981). It was also suggested that the historical decline of commercial catches in
eastern Hudson Bay was an indication of over-exploitation of that "stock," and of an absence
of recruitment from western Hudson Bay (Finley et al. 1982; Reeves and Mitchell 1987).

35
Unfortunately, theit is no direct information on the pre-exploitation size of those populations
to verify the validity of these conclusions.
The depletion of both the eastern Hudson Bay and Ungava Bay populations appears to have
taken place in the 1800's and 1900's prior to that of Cumberland Sound (Finley et al. 1982).
As late as the mid-1970's, there was still considerable hunting pressure on those two beluga
populations (Finley et al. 1982). Belugas are now very scarce in Ungava Bay in summer
(Smith and Hammill 1986). There is some evidence that the Labrador population also
suffered a decline in numbers in the early I 900' s (Brice-Bennett 1978; Boles 1980). If
neighboring populations such as Cumberland Sound, Ungava Bay and Labrador normally
exchanged animals, the reductions in numbers which they all suffered could have prevented
the benefits of emigration from neighboring populations. Mitchell and Reeves (1981)
suggested the possibility that the Cumberland Sound population could also be part of a much
larger stock which is itself declining.
The only population which has apparently maintained large numbers during the same
period is the western Hudson Bay population. However, information available on its historic
size is not sufficiently precise to conclude on a trend in abundance (Perrin 1980; Richard and
Orr 1986; see also above discussion on its population size).
The eastern Hudson Bay—James Bay population still numbers several thousand animals
despite the heavy exploitation to which it has been subjected. It is possible that the large size
of the western Hudson Bay population could have acted as a buffer against its own depletion
and, through emigration of its animals, against a more drastic decline in the eastern Hudson
Bay population size. Yet the disappearance of historical beluga concentrations at Great
Whale and Little Whale rivers (Finley et al. 1982; Reeves and Mitchell 1987) suggests that
immigration, if it occurred, was not sufficient to buffer those two concentrations. Their
disappearance lends support to the hypothesis of philopatric estuarine stocks.
In this review, we present two recurring ideas. The first is that belugas are widespread in
distribution throughout the year and opportunities for mixing between populations are
frequent. The second is that belugas concentrate at estuaries in summer and these estuarine
concentrations could be made largely of philopatric female-calf stocks. In other words, we
might have wide-ranging males and juveniles but sedentary females and calves and the site-
tenacity of the latter would make them more vulnerable to hunting at estuaries. If these
estuarine stocks are distinct genetic stocks then there is indeed concern for their conservation
where they are hunted directly at the estuary such as in the Nastapoka River. On the other
hand, most catches of subarctic belugas today do not take place in estuaries but at locations
along the coasts of Hudson Bay and Hudson Strait where the identity of landed belugas is
much less clear.
This review shows that new approaches are needed to clarify the stock identity of subarctic
belugas hunted by about 20 N.W.T. and Quebec communities. Future studies of these
populations should concentrate on monitoring movements of individuals, using mark—recapture
or telemetric tracking, and on genetic markers (Hellbig et al. 1989) that could elucidate the
relationships between estuarine concentrations and catches at locations throughout the range
of subarctic belugas.

Acknowledgments

We thank the many people who assisted in various aspects of the surveys. Thanks particu-
larly to A. Blouw for his help in the design and operation of the photographic systems. M.
Kingsley gave us valuable advice on population estimation methods. Thanks also to T. Smith,
R. Reeves and two anonymous reviewers for their comments. Polar Continental Shelf Project
and Bradley Air Services provided continued support during the program. Hunters and
Trappers Associations from N.W.T. settlements on sub-arctic coasts were very helpful in
sharing their knowledge about the seasonal distribution of belugas.

36
 References

ANON. 1978. Wager Bay, a natural area of Cana- FINLEY, K. J., G. W. MILLER, M. ALLARD, R. A.
dian significance. Parks Canada. Department DAVIS, AND C. R. EVANS. 1982. The belugas
of Indian and Northern Affairs, Ottawa. 7 p. (Delphinapterus leucas) of Northern Quebec:
BOLES, B. K. 1980. White whales in Labrador. distribution, abundance, stock identity, catch
Report prepared for the Offshore Labrador history and management. Can. Tech. Rep.
Biological Studies (OLABS) Program. iv + Fish. Aquat. Sci. 1123: vii + 63 p.
19 p. GAMBLE, R. L. 1984. A preliminary study of the
BRETON, M., T.G. SMITH, AND B. KEMP. 1984. native harvest of wildlife in the Keewatin
The views of scientists and Inuit on biology region, Northwest Territories, Canada. Can.
and behaviour of Arctic seals and whales: Tech. Rep. Fish. Aquat. Sci. 1282: iv + 48 p.
harvesting sea mammals; management and 1987a. Native harvest of wildlife in the
conservation for the future. Canada. Depart- Keewatin region, Northwest Territories for the
ment of Fisheries and Oceans. Quebec. 50 p. period October 1983 to September 1984. Can.
BRICE-BENNETT, C. 1978. An overview of the oc- Tech. Rep. Fish. Aquat. Sci. 1543: v + 85 p.
currence of cetaceans along the northern La- 1987b. Native harvest of wildlife in the
brador coast. Report prepared for the Offshore Keewatin region, Northwest Territories for the
Labrador Biological Studies (OLABS) Pro- period October 1984 to September 1984. Can.
gram. 48 p. Tech. Rep. Fish. Aquat. Sci. 1549: v + 82 p.
BRODIE, P. F. 1971. A reconsideration of aspects 1988. Native harvest of wildlife in the Kee-
of growth, reproduction and behavior of the watin region, Northwest Territories for the
white whale, Delphinapterus leucas, with spe- period October 1985 to March 1986 and a
cial reference to the Cumberland Sound, Baf- summary for the entire period of the harvest
fin Island population. J. Fish. Res. Board. Can. study from October 1981 to March 1986. Can.
28: 1309-1318. Data Rep. Fish. Aquat. Sci. 688: y + 85 p.
BRODIE, P. F., J. L. PARSONS, AND D.E. SERGEANT. HELBIG, R., P. T. BOAG, AND B. N. WHITE. 1989.
1981. Present status of the white whale (Del- Stock identification of beluga whales (Delphi-
phinapterus leucas) in Cumberland Sound, napterus leucas) using mitochondrial DNA
Baffin Island, Rep. Int. Whaling Comm. 31: markers: Preliminary results. Muskox 37:
579-582. 122-128.
CARON, L. M. J., AND T. G. SMITH. 1990. Philopa- JONKEL, C. J. 1969. White whales wintering in
try and site tenacity of belugas, Delphinape- James Bay. J. Fish. Res. Board Can. 26:
trus leucas, hunted by the Inuit at the Nastapo- 2205-2207.
ka estuary, eastern Hudson Bay, p. 69-79. In JOIINSTON, D. H. 1961. Marine Mammal surveys
T. G. Smith, D. J. St. Aubin, and J. R. Geraci Hudson Bay 1961. Ontario Department of
[ed.] Advances in research on the beluga whale, Lands and Forests. Unpubl. manus. 32 p.
Delphinapterus leucas. Can. Bull. Fish. Aquat. [author's collection]
Sci. 224. KINGSLEY, M. C. S., AND W. SMITH. 1981. Analy-
CA UGHLEY, G. 1977. Sampling in aerial survey. sis of data arising from systematic transect sur-
J.Wildl. Manage. 41: 605-615. veys, p. 40-48. In F.L. Miller and A. Gunn
COCHRAN, W. G. 1977. Sampling techniques, 3rd [ed.] Symposium on census and inventory
ed. Wiley, New York, NY. 393 p. methods for population and habitats. Procee-
DEGERBOL, M., AND P. FREUCHEN. 1935. Mam- dings of the Northwest Section, The Wildlife
mals. Report of the Fifth Thule Expedition, Society, Banff, Alta., April 1980.
1921-1924. Vol.2, (no 4-5): 278 p. + 1 KINGSLEY, M.C.S., I. STIRLING, AND W. CAL-
folding map. VERT. 1985. The distribution and abundance of
DOIDGE, D. W. 1990. Age-length and length- seals in the Canadian high Arctic, 1980-1982.
weight comparisons in the beluga, Delphinap- Can. J. Fish. Aquat. Sci. 42: 1189-1210.
terus leucas, p.59-68. In T. G. Smith, D. J. St. KLEINENBERG, A., V. YABLOKOV, B. M. BECK 0-
Aubin, and J. R. Geraci [cd.] Advances in re- VICH, AND M.N. TARASEVICH. 1969. Beluga
search on the beluga whale, Delphinapterus (Delphinapterus leuccts): Investigation of the
leucas. Can. Bull. Fish. Aquat. Sci. 224. species. Israel Prog. Sci. Transi. Jerusalem. vi
+ 376 p.

37
MCLAREN, P. L., AND R. A. DAVIS. 1982. Winter Department of Fisheries and Oceans. Contract
distribution of Arctic marine mammals in ice- no. Fp430-8-9047/0 -I SF. y + 86 + II appen.
covered waters of eastern North America. LGL RICHARD, P.R., AND J. R. ORR. 1986. A review of
ltd, October. Report prepared for the Offshore the status and harvest of white whales (Delphi-
Labrador Biological Studies (OLABS) Pro- napterus halals) in the Cumberland Sound
gram. xiii + 151 p. area, Baffin Island. Can. Tech. Rep. Fish.
MACLAREN MAREX INC. 1978. Report on aerial Aquat. Sci. 1447: iv + 25 p.
Surveys 77-2,77-3,77-4: studies of sea birds SERGEANT, D. E. 1968. Whales, p. 388-396. In
and marine mammals in Davis Strait, Hudson C.S. Beals and D.A. Shenstone. Science, his-
Strait and Ungava Bay for Imperial Oil Ltd, tory and Hudson Bay, vol. 1 . Canada Depart-
Aquitaine Company of Canada Ltd. and Cana- ment of Energy, Mines and Resources. 502 p.
da Cities Services Ltd. Arctic Petroleum Ope- 1973. Biology of the white whales (Delphi-
rators Association, project nos 134 and 138, napterus leucas ) in western Hudson Bay. J.
February, 1978. xvi + 70 p., figures, tables and Fish. Res. Board Can. 30: 1065-1090.
appendices. 1981. On permissible exploitation rates of
1979. Report on aerial surveys of birds and Monodontidae. Rep. Int. Whaling Comm. 31:
marine mammals in the southern Davis Strait 583-588.
between April and December 1978 for Esso SERGEANT, D. E., AND P. F. BRODIE. 1969. Body
Resources Canada Ltd. and Canada Cities size in white whales,Delphinapterusleucas. J.
Services Ltd. Arctic Petroleum Operators As- Fish. Res. Board. Can. 26: 2561-2580.
sociation, project no 146, volume III: Marine 1975. Identity, abundance, and present sta-
Mammals, Febrttary 1979. xxii + 36 p., figures tus of populations of white whales, Delphi-
and tables. napterus !camas, in North America. J. Fish.
1980. Aerial monitoring of marine birds Res. Board Can. 32: 1047-1054.
and mammals: the 1979 offshore drilling pro- SERGEANT, D. E., AND W. HOEK. 1988. An update
gramme near Southeast Baffin Island for Esso of the status of white whales, Delphinapterus
Resources Canada Ltd. and Aquitaine Compa- leucas, in the Saint-Lawrence estuary, Cana-
ny of Canada Ltd. April 1980. viii + 23 p., da. Biol. Conservation 45: 287-302.
figures and tables. SMITH, T. G., AND M.O. HAMMILL. 1986. Popula-
MITCHELL, E., AND R.R. REEVES. 1981. Catch tion estimates of white whale, Delphinapterus
history and cumulative catch estimates of ini- 'camas, in James Bay, eastern Hudson Bay,
tial population size of cetaceans in the eastern and Ungava Bay. Can. J. Fish. Aquat. Sci. 43:
Canadian Arctic. Rep. Int. Whaling Comm. 1982-1987.
31: 645-682. SMITH, T.G., M.O. HAMMILL, D.J. BURRAGE,
PERRIN, W.F. 1980. Review of definition and AND G. A. SLENO. 1985. Distribution and abun-
status of stocks of the white whale, Delphinap- dance of belugas, Delphinapterus leucas, and
terus leucas , and the narwhal, Moundon mo- narwhals, Monodon monoceros, in the Cana-
noceros. In Report of the subcommittee on dian high Arctic. Can. J. Fish. Aquat. Sci. 42:
small cetaceans. Rep. Int. Whaling Comm. 30: 676-684.
117-122. SUTTON, G. M., AND W. J. HAMILTON. 1932. The
REEVES, R. R., AND E. MITCHELL. 1987. History of mammals of Southampton Island. Part II sec-
white whale (Delphinapterus lerwas) exploi- tion I. In The exploration of Southampton
tation in eastern Hudson Bay and James Bay. Island, Hudson Bay. Mem. Carnegie Mus. 12:
Can. Spec. Publ. Fish. Aquat. Sci. 95: 45 p. 3-110.
1989. History of exploitation, distribution TOMILIN, A.G. 1967. Mammals of the U.S.S. R.
and stock identity of white whales in western and adjacent countries. Cetacea Vol.9. Israel
Hudson Bay and Foxe Basin. Unpubl. rep. to Frog. Sci. Transl. Jerusalem. xxii +717 p.

38
 Distribution, Abundance, and Movements of
Beluga Whales, Delphinapterus leucas, in
Coastal Waters of Western Alaska

K. J. Frost and L.F. Lowry

Alaska Department of Fish and Game,
1300 College Road, Fairbanks, AK 99701, USA

FROST, K. J., AND L. F. Lowit y. 1990. Distribution, abundance, and movements of

beluga whales, Delphinapterus leucas, in coastal waters of western Alaska,
p. 39-57. In T. G. Smith, D. J. St. Aubin, and J. R. Geraci [cd.] Advances in
research on the beluga whale, Delphinapterus leucas. Can. Bull Fish. Aquat.
Sci. 224.

A compilation of sightings indicates that belugas occur in coastal waters of

western Alaska during the ice-free period, which ranges from year-round in parts
of Bristol Bay to July–September in the northeastern Chukchi Sea. Their distribu-
tion and abundance appears to be influenced mostly by a combination of food
availability and the presence of estuarine conditions which may be important for
molting. Sea ice, killer whales, and human activity may also affect distribution in
some areas. Major concentration areas, in which belugas can be considered as
provisional management stocks, occur in Bristol Bay, the Norton Sound/Yukon
River delta region, and the eastern Chukchi Sea (including Kotzebue Sound and the
Kasegaluk Lagoon region). These hypothetical stocks are based on distribution
and timing of sightings, and should be confirmed by tagging and genetics studies.
Tentative estimates of stock sizes are as follows: Bristol Bay — 1 000 to 1 500;
Norton Sound/Yukon Delta — 1 000 to 2 000; and eastern Chukchi Sea — 2 500
to 3 000. Limited data suggest that numbers have generally stayed stable over the
past 20-30 yr, although substantial fluctuations in local distribution have occurred
in the Kotzebue Sound region.

Une compilation basée sur l'observation visuelle dénote que les marsouins
blancs se produisent dans les eaux côtières de l'Alaska de l'ouest durant la saison
sans glace, qui s'étend sur toute l'année dans des parties de la baie de Bristol, et de
juillet-septembre dans le nord-est de la mer de Chukchi. Leur répartition et
abondance semblent être principalement influencées par une combinaison de la
disponibilité de la nourriture et de la présence de conditions estuarines qui sont
peut-être importantes pour la mue. La glace océanique, les ores et l'activité
humaine affectent peut-être aussi la distribution dans certaines zones. Des zones
de concentration majeures dans lesquelles les marsouins blancs peuvent être con-
sidérés comme des stocks pour la gestion se produisent dans la baie de Bristol, la
région de l'anse de Norton/rivière du Yukon, et de l'est de la mer Chukchi (incluant
l'anse de Kotzebue et la région du lagun de Kasegaluk. Ces stocks hypothétiques
sont calculés sur la distribution et la répartition saisonnière d'observations visuel-
les, et devraient être confirmés par l'étiquetage et des études génétiques. Des
estimations préliminaires de grosseurs de stocks sont les suivants: baie de Bris-
tol — 1 000 à 1 500; l'anse de Norton/delta du Yukon — 1 000 à 2 000; et l'est de
la mer Chukchi — 2 500 à 3 000. Des données limitées suggèrent que les nombres
sont demeurés généralement stables sur une période s'échelonnant sur les 20-30
dernières années, quoique de réelles fluctuations dans la distribution locale ont eu
lieu dans la région du détroit de Kotzebue.

39
 Introduction

Beluga whales (Delphinapterus lencas) are a common but relatively unstudied component
of the Alaskan marine mammal fauna. The majority of beluga whales in Alaska occur in the
Bering, Chukchi and Beaufort seas. Seasonal distribution and movements of this group of
whales, which we will refer to as the Bering Sea population, are complex. Most winter
sightings have been in pack ice of the central Bering Sea (Brueggeman and Grotefendt 1988;
Ljungblad et al. 1986), which is thought to be the main wintering area for the population
(Seaman and Burns 1981). Beluga distribution and movements are affected by ice (Fay 1974;
Lowry et al. 1987a), and in some circumstances winter range may extend into the northern
Bering and southern Chukchi seas.
As sea ice cover loosens in late March and April, major seasonal movements begin. Many
whales migrate northward through leads and shear zones of the Bering, Chukchi and Beaufort
seas (Fraker 1979). The belugas often travel in association with bowhead whales (Balaenn
mysticetus), and are observed and hunted during spring near settlements along the west coast
of Alaska (Seaman and Burns 1981; Braham et al. 1984). The majority of this group probably
summers in the eastern Beaufort Sea, Amundsen Gulf and the Mackenzie Delta (Fraker et al.
1978; Davis and Evans 1982). They migrate westward in August—October through offshore
waters of the Alaskan Beaufort Sea (Burns and Seaman 1988; Ljungblad et al. 1987). Due
to ice conditions and seasonal movement patterns, this group of belugas spends little time in
coastal waters of western Alaska, though they frequent nearshore areas in the eastern
Beaufort Sea and Amundsen Gulf.
This paper deals with the beluga whales that appear along the coast of western Alaska
during the open water season. Although they probably share a common wintering area with
whales from the eastern Beaufort Sea/Mackenzie stock, and may be part of the same
population, it is likely that these groups should be considered as one or more separate stocks
for management purposes. Our objectives in this paper are to compile and present all
available data on distribution and abundance of belugas in coastal waters of western Alaska,
identify concentration areas, suggest provisional management stocks based on movements
and seasonal distribution patterns, and estimate abundance and, where possible, trends in
abundance for each provisional stock.

Methods

Information on beluga distribution and abundance has been obtained from both opportu-
nistic and systematic observations. Opportunistic sightings have come from a network of
informants, including scientists, fishermen, hunters, and others who live and work along the
coast of Alaska, We have received many unsolicited reports, and regularly query certain
individuals to obtain their observations. Sources of these personal communications are listed
in Frost et al. (1983a, 1983b) and Seaman et al. (1988).
Much information has come from aerial surveys of spawning herring (Chtpea harengus)
conducted by the Commercial Fisheries Division of the Alaska Department of Fish and Game
from 1976 through 1988 (C. Whitmore, unpublished data), and aerial surveys of endangered
whales conducted by the Naval Ocean Systems Center from 1979 through 1986 (Ljungblad
et al. 1987). All beluga sightings were recorded during those surveys.
We conducted aerial surveys specifically to enumerate beluga whales at various times
between 1978 and 1987. Surveys did not cover the entire habitat of belugas along the coast
of western Alaska but were directed to specific areas at the times of year when belugas were
expected to concentrate. Surveys were of two general types depending on the type of aircraft
available and the anticipated distribution of whales. Search surveys were usually flown
parallel to the coast at a distance of 0.5-3 km offshore and an altitude of 152-457 m. One or
two observers, seated beside or behind the pilot in a fixed-wing aircraft or helicopter, looked

40
for and recorded all whales within view. Large concentrations were frequently circled and
counted or photographed. Aerial photographs were 35 mm slides taken obliquely from the
window of the aricraft. Slides were projected and all visible animals counted. Transect
surveys covered areas where belugas were thought to occur, or where concentrations were
located during search surveys. Transect surveys were flown at an altitude of 305 m at speeds
of 180-275 km/h with 1 observer on each side of the aircraft counting all whales visible within
a strip 0.9 km wide.
Sightings, whether opportunistic or made during surveys, may not detect all the animals
present in a area. Beluga whales, especially young animals, are difficult to detect in muddy
or rough water. Also, belugas often spend little time at the surface where they can be seen.
Correction factors may be developed and applied to adjust for undercounting (Sergeant 1973;
Davis and Evans 1982; Frost et al. 1985). Unless noted otherwise we have not applied any
correction factors to the data, and have presented the numbers provided by the original
observers.
Information on movements was obtained either directly from locations of radio-tagged
animals, or inferred from the sequence of occurrences of groups of animals at adjacent
localities.

Results and Discussion

Regional Distribution and Abundance

The southern limit of the normal distribution of belugas in western Alaska occurs in Bristol
Bay at approximately 56° N latitude. We know of no sightings from the Aleutian Islands,
Unimak Pass, or the Alaska Peninsula south of Port Moller (contrary to Gurevich 1980), and
local residents say that belugas do not occur in this area.
Although few sightings have occurred in the period from November through March
(Fig. 1), belugas have been reported from coastal waters of Bristol Bay northeast of Port
Moller throughout the year. Autumn and winter sightings occur infrequently which suggests
that belugas do not regularly occur in large numbers at predictable nearshore locations during
those months. In most years pack ice extends into Bristol Bay, and belugas may spend much
of their time offshore in or near the ice. Formation of shorefast ice in bays and river mouths
excludes them from some nearshore areas of Bristol Bay during winter.
Belugas begin to appear regularly in northern Bristol Bay in early April (Fig. 2). They often
swim up rivers to the edge of the ice as it begins to break up. Whales become increasingly
common in May and June with frequent sightings through August (Fig. 2-4). Most sightings
occur in Kvichak and Nushagak bays. Belugas have seldom been sighted in outer, northwes-
tern Bristol Bay.
The first sightings of the year along the Yukon and Kuskokwim deltas and in Norton Sound
usually occur in late April and early May (Fig. 2). Sighting effort in this area is limited and
belugas may often go unnoticed, especially in the muddy waters that occur near river mouths
at breakup. Most reports come from aerial surveys for herring, or from residents of coastal
villages where belugas are occasionally hunted when they appear shortly after the shorefast
ice moves away. Whales occur in this region during the summer and autumn, with most
sightings near the mouths of the Yukon River and in Norton Sound (Fig. 2-4). According to
local residents, belugas were common in the lower portion of the Kuskokwim River earlier
in this century but stopped using the lower river in the 1940's. However, during April—August
1988, frequent sightings of a few to several hundred belugas were reported in Kuskokwim
Bay and in the Kuskokwim River as much as 40 km upstream. Overall, the lack of sightings
in northwestern Bristol Bay, Kuskokwim Bay, and along the Bering Sea coast south of the
Yukon River suggests that there is little movement of belugas between Bristol Bay and the
Norton Sound/Yukon Delta region.

41
 165 150

70 WAINWRIGHT
, 70

ICY CAPE
CHUKCHI SEA
KASEGALUK LAGOON

CAPE LISBURNE

SISUALIK
KOTZEBUE SOUND

ESCHSCHOLTZ BAY

65 - 55

ALASKA

NORTON SOUND

YUKON RIVER

•
60 KUSKOKWIM RIVER ) ,JO. + 60

NUSHAGAK BAY

BERING SEA
• KVICHAK BAY
•
•
BRISTOL BAY

165 150

Sightings of belttga whales in ice-free coastal waters of western Alaska during the months of
FIG. 1.
November through March.

42
 165 150

70 A WAINWRIGF'T 70

CV CAPE
CHUKCHI SEA
KASEGALUK LAGOON

CAPE LISBURNE

SISUALIK
KOTZEBUE SOUND

ESCHSCHOLTZ BAY

65
ALASKA

NORTON SOUND
•

YUKON RIVER

so 4 k KUSKOKWIM RIVER 60

NUSHAGAK BAY

BERING SEA
KVICHAK BAY
•

BRISTOL BAY

e
165 150

Flo. 2. Sightings of beluga whales in ice-free coastal waters of western Alaska during the months of
April and May.

43
 165 150

•
70
• WAINWRIGHT 70

CHUKCHI SEA ICY CAPE

KASEGALUK LAGOON

CAPE LISBURNE

SISUALIK
KOTZEBUE SOUND

ESCHSCHOLTZ BAY
BERING STRAIT
65 65

ALASKA

NORTON SOUND • •

YUKON RIVER

60 KUSKOKWIM RIVER
60

NUSHAGAK BAY

BERING SEA
KVICHAK BAY

BRISTOL BAY

ne , 1f

165 150

FIG. 3. Sightings of beluga whales in ice-free coastal waters of western Alaska during the months of
June and July.

44
 165 150

70 WAINWRIGHT _..e"1- 70

CHUKCHI SEA eper6Y CAPE

KASEGALUK LAGOON

CAPE LISBURNE

SISUALIK
KOTZEBUE SOUND

ESCHSCHOLTZ BAY
BERING STRAIT

65 65

ALASKA

•
NORTON SOUND

YUKON RIVER

60 KUSKOKWIM RIVER 60

NUSHAGAK BAY

BERING SEA
KVICHAK BAY

BRISTOL BAY

y-
\

'

165 150

Flo. 4. Sightings of beluga whales in ice-free coastal waters of western Alaska during the months of
August through October.

45
 There have been very few recent reports of belugas in the Bering Strait region during the
open water season. Whales begin appearing in Kotzebue Sound in early to mid-June, where
they are commonly seen until early July (Fig. 3). Large groups are sometimes reported,
especially near Sisualik (in the Inupiat language meaning the place of white whales) and
Eschscholtz Bay. In late June and July whales are seen along the Chukchi Sea coast northwest
of Kotzebue and east of Cape Lisburne. Large numbers of whales appear near the barrier
islands and passes off Kasegaluk Lagoon, both south and east of Icy Cape, in July (Fig. 3).
Belugas are occasionally reported further to the northeast, especially near Wainwright, in late
July through early September (Fig. 4).
Based on the distribution and timing of sightings, four regions can be identified along the
west coast of Alaska where substantial numbers of belugas occur each year during the open
water season: Bristol Bay, Norton Sound/Yukon River delta, Kotzebue Sound, and the
northeastern Chukchi Sea near Kasegaluk Lagoon. Specific aspects of the distribution,
abundance, and movements of belugas in each of those areas are discussed in the following
sections.

Bristol Bay

The group of belugas inhabiting Bristol Bay has received more attention by scientists than
any other in Alaska. Whales come to this area to feed on rainbow smelt (Osinerus mordax)
and salmon (Oncorhynchus spp.) that occur in the rivers and bays (Brooks 1955). The
possibility that belugas could have been impacting the world's largest run of red salmon
(Oncorhynchus nerka) led to a variety of detailed studies in Kvichak and Nushagak bays
(Brooks 1954,1955,1956; Fish and Vania 1971).
We conducted studies of beluga whales in this area in May—June 1982 and April—August
1983. Observations and surveys were conducted from small boats and aircraft (Frost et al.
1984a), and VHF radiotags were applied to two whales (Frost et al. 1985). Additional surveys
were flown in Nushagak Bay by Stewart et al. (1983). A plot of all sightings made during
1982-83 (Fig. 5) shows that some areas are more heavily used than others. In Nushagak Bay
most sightings occurred in the area between the Snake and Igushik rivers and Clark's Point,
and in the north part of the bay near the intersection of the Wood and Nushagak rivers. In
Kvichak Bay most sightings occurred in the upper bay near Nakeen, and in Halfmoon Bay.
Whales have been seen in the Kvichak, Naknek, Nushagak, Igushik, Snake and Wood rivers,
sometimes as much as 50 km upstream. The two whales radiotagged in Kvichak Bay moved
between areas within this bay, but did not move into Nushagak Bay during the 11- and 14-
d tracking periods. However, seasonal shifts in abundance (Brooks 1955) and sightings of
whales in the intermediate area (Fig. 5) suggest that movement of whales between the two
bays, which are about 80 km apart, does occur.
Kvichak and Nushagak bays have a large range (7-8 m), strong currents, very muddy water,
and extensive mud flats separated by narrow river channels. During low tides belugas are
restricted to the river channels and deeper portions of the bay, while on high tides they may
range over the mudflats as well. Whales radiotagged in Kvichak Bay in 1983 commonly made
daily movements of 20-50 km, generally up and down the bay (Frost et al. 1985). Movements
were both with and against the tide with no apparent pattern. However, visual observations
of groups of whales made during the same period indicated that whales moved with the
direction of the tide 77% of the time. All movements against the tide occurred within 2 h of
a tidal change, and were usually (82%) against an ebb tide. Such a pattern would help keep
the whales within the same general portion of the bay from day to day.
We conducted nine aerial surveys from 15 April through 14 August 1983 (Frost et al
1984b). Eight were search type surveys that covered all known concentration areas in
Kvichak and Nushagak bays and the intermediate coastline. On 29 June we flew a strip
transect survey over the concentration areas in addition to flying along the coast and up the

46
 159 158 157
4

WOOD RIVEPI
KVICHAK RIVER

59 NUSHAGAK RIVER 59

SNAKE RIVER NAKEEN

HALFMOON BAY4ler
4*
IGUSHIK RIVER
wi NAKNEK RIVER
1.4
;
••• • •
•
••• KVICHAK BAY
NUSHAGAK BAY;

58 30 55 30

CAPE CONSTANTINE
BRISTOL BAY

159 158 157

FIG. 5. Kvichak and Nushagak bays showing sightings of beluga whales made during 1982-83.

rivers. Transects were spaced so that the concentration areas were completely counted. The
water was muddy and whales were visible only when breaking the surface, so we applied
correction factors to account for animais that were diving as the plane passed over. The
probability that a whale will be at the surface is equal to the mean length of a surfacing plus
the length of time an area is within the field of view, divided by the mean length of a surfacing
plus the mean length of a dive. The correction factor by which aerial survey counts can be
multiplied to derive the actual abundance of whales is the reciprocal of this probability.
The correction factors that we used were based on surfacing and dive time. data from
radiotagged whales (Frost et al. 1985), and depended on the length of time a part of the
transect was in view (which related to ground speed). Correction factors were 2.75 for a
survey speed of 180 km/h and 3.7 for a survey speed of 275 km/h. Counts were further
adjusted upward by 8% to account for dark-colored yearlings (Brodie 1971) which could not
be seen in muddy water. The number of whales estimated to occur in the area varied from 258
on 13 May to 993 on 29 June (Table 1) which suggests that animais were moving in and out
of the area and were absent during some search surveys. Nonetheless it is evident that several
hundred belugas were in the area throughout the study period, with largest numbers in
Kvichak Bay from mid-April through mid-June, and in Nushagak Bay during late June and
July. Brooks (1955) estimated there were about 1 000 belugas in Kvichak and Nushagak bays
in 1954, which is very close to our 1983 maximum estimate of 993 belugas. We have no
reason to think that the number of belugas in Bristol Bay has changed in recent years. They
are not hunted to a significant degree, and fisheries-related mortality is quite low (Frost et al.
1984b).
Neonates are not included in the counts shown in Table 1. However, it is clear that calves
are born in this area, as we found 7 dead neonates and one abortus on the beach during our
operations in Kvichak Bay in 1983 (Frost et al. 1984b).
There have been very few sightings of belugas in the northern part of Bristol Bay to the west
of Nushagak Bay. This is an area of considerable human activity particularly during the
spring and summer fishing seasons, and it is likely that we would receive reports if substantial

47
TABLE 1. Estimated numbers of beluga whales in Nushagak and Kvichak bays, April—August 1983.
Neonates are not included.

Adjusted
Date Nushagak Bay Kvichak Bay Total total"

15 April 218 474 692 747

2/5 May 41 584 625 675
17 May 85 274 359 388
31 May 27 212 239 258
14 June 49 259 308 333
24 June 182 55 237 286
29 June 347 572 919 993
14 July 496 181 677 731
14 August' 0 309 309 334
" Total estimate increased by 8% to account for dark-colored yearlings.
b Correction factor was not applied because whales were in very shallow water and the
observer thought all whales were counted.

numbers of whales utilized the area. The absence of animals during April—June is confirmed
by surveys for herring which annually cover the nearshore area from Cape Constantine to
Kuskokwim Bay. Since 1967 only the following sightings were recorded: a single animal in
June 1975, two in May 1978, and 1-11 in April—May 1988.

Norton Sound/Yukon Delta

There have been no systematic surveys of beluga whales in the Norton Sound and Yukon
River delta areas, and available information on distribution and abundance is very limited.
During aerial surveys conducted during or shortly after breakup, belugas are regularly seen
in association with herring schools (Fig. 6). Whales occur in the Sound until freezeup in
November. They occasionally move long distances up the Yukon River. In June 1982 we
received reports of 4-5 belugas in the river above the settlement of Tanana, which is 1 200 km
from the river mouth, and of a single large adult 130 km further upriver.
There may be considerable movement of whales in the Norton Sound area. Belugas appear
to feed on herring and capelin (Ma/lotus villosus) in coastal waters of Norton Sound in
May—June, move to the Yukon River mouth to feed on salmon during June—August, then
return to coastal Norton Sound to feed on saffron cod (Eleginus gracilis) in September-
November (Seaman et al. 1982, 1988). Such a pattern is also suggested by the timing of whale
hunting at coastal villages in the Sound, which occurs mostly during June and
September—October. However, some area residents think that the fall whales may not be part
of the Norton Sound group, but are "northe rn " whales on their way south for the winter.

Kotzebue Sound

There are two important beluga whale hunting sites in Kotzebue Sound, Sisualik and
Elephant Point, and most sightings of whales occur near these areas (Fig. 7). They are also
seen occasionally in Hotham Inlet, Selawik Lake, and the Buckland River. Whales arrive as
the ice cover is breaking up, appearing first in the northern portion of the Sound, then
penetrating southeastward as ice conditions allow (Burns and Seaman 1988). Belugas arrive
as early as the first week in June, are most common from mid- to late June, and are only
occasionally seen in July and August (Table 2).

48
 165 160

165 160
Fia. 6. The Norton Sound region showing beluga whale sightings made during surveys for herring,
1976-88.

165 163 161

SISUALIK
67 + _. _. •• +67
CHUKCHI SEA

KOTZEBUE SOUND
ESCHSCHOLTZ BAY
\
S. 4e ELEPHANT POINT

eilBUCKLAND RIVER

66+ 66

165 163 161

Fia. 7. The Kotzebue Sound region showing major hunting sites and the distribution of beluga
sightings, 1970-86.

49
TABLE 2. Major sightings of beluga whales in the Kitzebue Sound region, 1960-88.
Date Location Number Comments
July 60 N of Chamisso I. 900-1200
June 73 Choris Peninsula 1000+ covered area 1x9 km
11 June 78 S of Chamisso I. 20-25 first report of the year
8-9 July 78 NW Eschscholtz Bay 900-1000 calves present, largest sighting
1-2 June 79 near Cape Blossom 30+ first report of the year
12 June 79 Eschscholtz Bay 300+ first report in Eschschlotz Bay
15-25 June 79 Sisualik area many groups of 10 to 75-100
19 June 79 NW of Chamisso I. 300-600 largest sighting
13 June 80 near Kotzebue many first report of the year
15 June 80 Eschscholtz Bay many first report in Eschscholtz Bay
23 June 80 Eschscholtz Bay 800+ largest sighting
13 June 81 S of Kotzebue 22 km 1000+ first report and largest sighting
15-16 June 81 Eschscholtz Bay present first report in Eschscholtz Bay
9-12 June 82 Sisualik area many moving SE into Kotzebue Sound
21 June 82 Eschscholtz Bay 100+ first report in Eschscholtz Bay
17 June 83 Eschscholtz Bay 50+ first hunt of year
18-19 June 83 Sisualik area 25+
late June 86 Sisualik 20-30 seen on 2 d, then no more
June 86 S of Kotzebue 200 pilot report, only large sighting
July—Aug. 86 Eschscholtz Bay a few no June hunt, not enough whales
15-24 June 87 Sisualik area present hunted
20-26 June 87 N Eschscholtz Bay 20-50+ aerial survey, ice present in bay
8 Sept. 87 near Kotzebue 10+ about 1 mile off shore
June 88 near Sisualik 1 few in ice 4-5 mi off shore
26-28 June 88 Eschscholtz Bay 75-200 aerial observation
August 88 W of Sisualik 5+ unusual, taken during salmon fishing

Some information on occurrence and abundance of belugas in Kotzebue Sound can be

inferred from the magnitude of harvests at the major hunting sites, which have varied
considerably (Table 3), in combination with observations of local hunters and a knowledge
of local hunting patterns. In northeastern Kotzebue Sound, animals are harvested by hunting
from boats, and with nets that have become more common since the early 1970's (Bob Uhl,
pers. comm.). Whales taken near Sisualik are generally moving to other parts of the Sound,
and harvests may be affected both by movement patterns and by overall numbers. In contrast,
the hunting site at Elephant Point provides access to an area where large numbers of whales
regularly congregate. Whales are taken in drive hunts that are an organized effort of the entire
village of Buckland, as well as people from other settlements. Harvests were large and
relatively consistent from 1977-83, averaging 74 whales per year. According to hunters the
harvest was low in 1979 because of disturbance by boats in the constricted entrance to
Eschscholtz Bay, and a lack of coordination among hunting groups. The number of whales
sighted in 1979 was not particularly low. Hunters got together to regulate disturbance in the
area and organize hunting activities, which resulted in greater success in 1980-83. However,
beginning in 1984 and continuing through 1988, whales have been much less common in the

50
TABLE 3. Numbers of beluga whales taken by subsistence hunters in Kotzebue Sound, 1977-88.

Year NE Kotzebue Sound—Sisualik SE Kotzebue Sound-Elephant Point

1977 3 105
1978 5 90
1979 2 5
1980 13 101
1981 4 39
1982 25 129
1983 19 48
1984 31 0
1985 13 2
1986 6 3
1987 2 7
1988 +5 17

southeastern part of Kotzebue Sound (Table 2), and harvests at Elephant Point have been
greatly reduced in spite of the usual amount of hunting effort.
In response to this apparent change in the occurrence of belugas in Kotzebue Sound, we
conducted aerial surveys of the area in 1987. In order to address the possibility that shifts in
distribution and habitat utilization might explain the lack of belugas in Eschscholtz Bay,
surveys were designed to cover all of the Kotzebue Sound region. Nine surveys covering
4 130 km were flown along the coast and on north—south running transects between 13 June
and 9 July. The only whales seen were in Eschscholtz Bay on 21, 22, and 26 June, with a
maximum count of 51 animals on 22 June. Whales were in very turbid water and could only
be seen while at the surface. If surface and dive patterns were similar to what we have
recorded for whales in Bristol Bay (Frost et al. 1985) a correction factor of 3 could be applied
to account for whales below the surface, resulting in an estimate of about 150 whales in the
area in 1987. Hunters reported an increase in the number of whales in Eschscholtz Bay in
1988.
Food, primarily saffron cod, sculpins (family Cottidae), smelt, and herring, commonly
occurred in the stomachs of whales taken at Elephant Point in 1978-80, and a beluga taken
in the Buckland River had eaten at least 11 large arctic char, Salvelinus malma (Seaman et
al. 1982). Local residents report that calving occurs throughout coastal waters of Kotzebue
Sound, and both near-term and postparturient females are taken at Elephant Point (Burns and
Seaman 1988).

Northeastern Chukchi Sea

The annual occurrence of belugas near Kasegaluk Lagoon was not well-known prior to the
mid 1970's. Conversations with residents of Point Lay in 1978 indicated that large numbers
of whales occurred in the area each year (Seaman et al. 1988), and studies of beluga whales
were conducted in 1978, 1979, 1981, and 1987.
Search surveys of the nearshore area were conducted in 1978, 1979, and 1981. Although
residents of Point Lay report that belugas sometimes occur in the deeper parts of Kasegaluk
Lagoon (which is mostly less than 2 m deep), most of our sightings have been made outside
the barrier islands, particularly near passes (Fig. 8). The first sightings of the year usually
occur between 22 June and 5 July, and whales are usually seen in the area until mid-July (Frost
et al. 1983b). In 1981 we conducted five surveys between 6 and 22 July. The number of

51
 165 160
I +71
71-I- J

WAINWRIGHT

CHUKCHI SEA ICY CAPE

70 -I- .7( +70
KASEGALUK LAGOON

4 .
POINT LAY

ALASKA
69 LEDYARD BAY 69

APE SABINE
CAPE LISBURNE

165 160

Fin. 8. The Kasegaluk Lagoon region showing sightings of beluga whales made during 1978-87.

whales seen was: 6 July — 15; 8 July — 460-670; 11 July — 40-51; 15 July — 75-100; 22
July —0. Earliest sightings tend to occur at the southern end of the lagoon, and large groups
are more commonly seen further north in July. Whales sometimes occur at several widely
separated locations at the sanie time. Sequential sightings of large groups and observations
of movements indicate that whales may move fairly rapidly while in the area (Seaman et al.
1988).
Abundance of whales was estimated from aerial photographs taken in 1979 (Seaman et al.
1988) The number of belugas counted from the slides was 1 104 on 13 July and I 601 on 15
July. Survey personnel estimated that approximately 10% of the whales were outside of the
concentrations and not included in photographs. Although whales were in relatively clear and
shallow water, not all of them were close enough to the surface to be counted. Based on
sequential photographs of the same groups of whales, it was estimated that in a single
photograph, 20% were too deep to be visible. The following correction factors were therefore
. applied to the actual photographic counts; 10% for whales outside the concentration area;
20% for whales too deep to be visible; and 8% for dark-colored yearlings. This resulted in
estimates of 1 575 on 13 July and 2 282 on 15 July. Braham (1984) reported observing 9.6%
dark-colored calves along the northwest coast of Alaska in 1981. Using this instead of
Brodie's (1971) estimate of 8.0% for dark-colored yearlings would result in a slightly higher
estimate for the Chukchi Sea. It is the opinion of the authors that there is considerable
confusion between young-of-the-year and yearling calves, and that both may be included in
estimates of calves. It is likely that most small belugas that are called calves in April–June
are in fact yearling calves, not neonates. Young-of-the-year are extremely difficult to see,
particularly from the air or at any distance.
On 8 July 1987 we flew an aerial strip transect survey over a large concentration of whales
offshore from Point Lay. The entire concentration area was covered with transects spaced 1.8
km apart, and 723 whales were counted. Double counting was not considered to be a problem
since transects were very short. After completion of the strip transect survey we conducted

52
a search survey of the remainder of the area off Kasegaluk Lagoon and saw only 1 other
beluga. The water in the concentration area was relatively clear but deep, and it was evident
that not all whales in the area were counted because some were diving out of sight. Applying
correction factors of 2 and 3 would result in estimates of approximately 1 400 and 2 100
whales, which are very similar to those obtained in July 1979.
Although there is usually little or no food in the stomachs of belugas taken by hunters at
Point Lay, local residents say that whales feed in the area. Both local residents and researchers
have observed calves being born, and neonates are frequently seen (Frost et al. 1983b).
There is some evidence to suggest that the belugas that occur near Kasegaluk Lagoon have
come from Kotzebue Sound and are part of the same stock. Although there is a certain amount
of overlap within a given year, belugas generally appear along the eastern Chukchi Sea coast
in a sequence from south to north. In most of the nine years for which we have sighting
information, whales are mostly gone from Kotzebue Sound by the time the first arrivals are
seen near Point Lay. Sightings at intermediate locations indicate that whales move along the
coast between the two areas. For example, in 1987, belugas were present in Kotzebue Sound
from 14-26 June, at Cape Sabine from 27 June-1 3 July, and near Point Lay from 3-9 July.
While we think it is likely that Kotzebue Sound and Kasegaluk Lagoon whales belong to
the same stock, it appears that they do not always travel as one large group, and that not all
whales go to both areas each year. On 3 July 1982, an estimated 2 000-2 500 whales were
seen off Cape Sabine, suggesting that the entire group was travelling together. In contrast, in
early July 1978 approximately 1 000 whales were seen in Kotzebue Sound during the same
week that 1 000+ whales were seen near Kasegaluk Lagoon. In late June 1987, a maximum
of 51 whales was counted in Kotzebue Sound, while a few days later over 700 were seen off
Kasegaluk Lagoon. We consider it unlikely that such a large group was present but not
detected in Kotzebue Sound during over 4 000 km of surveys. Observations in eastern
Hudson Bay suggest that large summering concentrations may consist of several smaller
groups that arrive at different times and come and go independently (Smith and Hammitt
1986). Observations of three belugas radiotagged in a high arctic summer concentration area
in 1988 also indicated independent movements in and out of the area (Frost, Martin and
Smith, unpubl. data). Groups of belugas moving from offshore areas to the Chukchi Sea coast
may arrive at slightly different times, and late-arriving groups may first join the concentra-
tion off Kasegaluk Lagoon resulting in larger sightings later in July in the more northern area.
Overall we think that the whales occurring along the eastern Chukchi Sea coast during June
and July are interrelated, and should for now be considered a single stock for management
purposes. However, further research is necessary to confirm this relationship.

Conclusion

Identity and Abundance of Management Stocks

The data we have presented, although not entirely systematic or comprehensive, indicate
that beluga whales occur in coastal waters of western Alaska from Bristol Bay to the northern
Chukchi Sea during the open water season. However, their distribution and abundance is far
from uniform. Whales occur very rarely in some areas, occasionally in others, and regularly
in large numbers at certain locations. A generally similar pattern probably occurs throughout
the arctic and subarctic regions inhabited by belugas, but the pattern may be obscured by the
tendency of investigators to focus their efforts on known concentration areas.
In coastal waters of western Alaska, summering belugas occur primarily in four concentra-
tion areas: Bristol Bay, Norton Sound/Yukon Delta, Kotzebue sound, and Kasegaluk Lagoon.
Based on distributional information, we think that the groups in Bristol Bay and Norton
Sound/Yukon Delta should be considered as separate management stocks, while those in
Kotzebue Sound and Kaselaguk Lagoon should be considered a single stock. Behavioral

53
observations support distributional information and suggest that the same whales return to
the same concentration area each year. Within an area, belugas use the same rivers, channels,
or passes with similar timing from one year to the next, and react to human activities in a
similar manner. For example, belugas in Kotzebue Sound are extremely shy of boats, while
those in Bristol Bay regularly co-occur with an intensive fishery. However, further research,
particularly tagging and DNA studies, is necessary to confirm hypothetical stock identity.
Seaman et al. (1988) made initial estimates of the abundance of whales in each of these
provisional management stocks, based on data collected through 1983, as follows: Bristol
Bay-1 000-1 500; Norton Sound/Yukon Delta-1 000-2 000; and eastern Chukchi Sea —
2 500--3 000. Our recent information does not suggest any changes to these estimates. There
is good reason to think that the Bristol Bay stock is stable at or near its historical size. Little
is known about the Norton Sound—Yukon Delta stock, since no comprehensive surveys for
belugas have been conducted there. Our counts in June—July 1987 and other information
indicate that the number of belugas utilizing Kotzebue Sound has decreased since 1984, but
that large numbers of whales still occur each year at Kasegaluk Lagoon.

Factors Affecting Distribution and Abundance

Along the west coast of Alaska belugas move into coastal waters as soon as ice conditions
permit. In areas south of Bering Strait, belugas retnain resident throughout the open water
season, and by late summer they are hundreds of km south of the pack ice. However, in the
Chukchi Sea, pack ice may remain close by while whales are present in Kotzebue Sound and
near Kasegaluk Lagoon. We have no direct observations of belugas in the Chukchi Sea pack
ice during June and July, but movements between the coast and the pack ice have been
inferred from fluctuations in nearshore abundance at Kasegaluk Lagoon (Seaman et al.
1988). Local residents think that belugas move back and forth between the coast and the pack
ice especially when it is near the shore. Belugaà seen along the Chukchi Sea coast in July may
move to the pack ice in August-September, and join the large group of whales that is moving
westward across the Beaufort Sea from the Mackenzie Delta area (Burns and Seaman 1988).
Therefore, it appears that while belugas are a pagophilic (ice-loving) species, their attraction
to ice is weakest during summer months and is overcome to varying degrees by other factors.
These other factors seem to operate most strongly in the Bering Sea. The reasons for
pagophily by belugas, particularly in northern regions during summer, are unknown. It is
possible that belugas feed under the ice on species such as arctic cod (Boreogadus saida)
which are also pagophilic during most of the year.
Previous investigators have suggested that the main reason that belugas inhabit coastal
areas during summer months is because they are wamier and provide a thermal advantage to
all age classes (Fraker et al. 1978) and particularly to neonates (e.g. Sergeant and Brodie
1969; Fraker et al. 1979). While Fraker et al. (1978) consider that the thermal advantage is
the most reasonable explanation, they also note that other characteristics of the Mackenzie
concentration areas include low salinity and shallow water depth. The relationship to water
temperature, if it exists, is complex and quite variable. Small calves, presumed to be neonates
and most vulnerable to cold, are seen within groups of whales in the pack ice (Braham et al.
1984). At Cunningham Inlet in the Canadian Arctic many belugas, including those with
neonate calves, concentrate at the mouth of a cold, fresh river which is fed by snow melt. In
the Bering Sea, which is strongly influenced by water of North Pacific origin, summer surface
water temperatures are comparatively warm throughout. Any thermal advantage that might
occur from calving in coastal waters would probably be slight. In contrast, as the ice moves
offshore in the Chukchi Sea, radiant warming and flow from rivers may increase nearshore
temperature substantially. In July water temperatures in plumes flowing out of Kasegaluk
Lagoon through passes were as much as 2° C warmer than in adjacent marine waters, and
belugas occurred most commonly in the plumes (Seaman et al. 1988).

54
 In at least some parts of their range, belugas congregate during during the molt in shallow
areas, where they are observed to crowd together in flowing fresh water and rub on the bottom
(T. G. Smith, Arctic Biological STATION, pers. comm.). Warmer waters may be important
for rapid cell growth during the molt (Finley 1982; St. Aubin et al. 1990). Alternatively, or
in combination, warm temperatures and the presence of fresh water may attract belugas to
areas such as Kaselaguk Lagoon for molting. Turbid conditions in most parts of Alaska
preclude observations of rubbing behavior such as have been made in other coastal
concentration areas.
It is clear that beluga whales feed along the coast of western Alaska during summer months
(Seaman et al. 1982). However, the importance of summer concentration areas for feeding
appears to differ among regions. In Bristol Bay a series of very abundant anadromous and
nearshore fishes (smelt, salmon smolt, adult salmon) provides continuous feeding opportu-
nities for belugas during the open water season. A similar situation seems to occur in the
Norton Sound/Yukon Delta region with somewhat different species of prey (herring, capelin,
adult salmon, saffron cod). Although some feeding does occur near coastal concentration
areas in the Chukchi Sea, there is no evidence that nearby food resources are unusually
abundant or suitable during the 2-3 wk that belugas are present. Most belugas leave Kotzebue
Sound before the salmon arrive in July and August.
Belugas are sensitive to disturbance in certain circumstances and airborne and waterborne
noise may influence their distribution and behavior (e.g. Fraker 1977; Stewart et al. 1983;
Finley and Davis 1984). The principal disturbances to which they are presently exposed in
coastal Alaska are associated with beluga hunting, commercial fishing, coastal traffic,
industrial development, and the proximity of settlements. The intense activity associated
with the Bristol Bay red salmon fishery has had no obvious effect on beluga distribution and
movements in that area (Frost et al. 1984b). Belugas move and feed in proximity to both fixed
nets set from shore and drift nets set from boats, and commonly approach anchored but noisy
processors and freighters. Their behavior appears little affected by transiting boats or aircraft,
except in narrow river channels when they will flee from approaching outboard motorboats
(Frost and Lowry, unpubl. observation). We have less information on how fisheries and other
activities in the Northon Sound/Yukon Delta region have affected belugas, although whales
are reported to be less easy to approach since the widespread use of outboard motors (Seaman
and Burns 1981). Residents of Norton Sound have suggested that belugas are less common
nearshore in spring since the local herring fishery intensified in the mid-1980's. A major
source of disturbance in Kotzebue Sound and Kasegaluk Lagoon is that associated with
hunting. Whales nonetheless continue to return to traditional hunting areas. Near Point Lay,
a village of about 120 people, the period of beluga hunting is brief, as an adequate number
of whales can be taken in 1 or 2 drives (in 1 or 2 d). Residents have noticed no recent changes
in distribution or abundance, although they do report that belugas do not reappear in a specific
area for several days after they are hunted. In contrast, residents of Kotzebue, a settlement
of over 3 000, think that increased traffic and noise in their area have caused fewer whales
to remain for shorter periods in northeastern Kotzebue Sound. Small boat traffic and low-
flying aircraft have been blamed by hunters for preventing whales from entering Eschsholtz
Bay (Burns and Seaman 1988). Some residents cite the introduction of underwater exhaust
on outboard engines as the reason for increased avoidance of heavily travelled areas (Pete
Shaeffer, pers. comm.).
Local residents frequently state that movements of belugas in their areas are being affected
by killer whales. Killer whales occur off western Alaska and they are known to prey on beluga
whales (Lowry et al. 1987b). Fish and Vania (1971) demonstrated the effectiveness of killer
whale calls in changing beluga whale movement patterns. Occupancy of coastal areas by
belugas may somewhat reduce such interactions, especially in relatively confined estuaries
where killer whales rarely occur.

55
 We conclude that beluga whales concentrate in coastal waters of the eastern Bering Sea
primarily because of the availability and abundance of forage fishes. In these areas belugas
are not particularly responsive to disturbance, and movements appear to relate mostly to
changes in prey distribution. In the eastern Chukchi Sea, coastal areas may be important to
belugas for other reasons, such as for molting. In these areas whales appear to be more
sensitive to disturbance and, at least in the Kotzebue Sound area, their distribution and
movements may be affected by human activities.
We think it is likely that in Bristol Bay and Norton Sound, where during May—September
there are abundant prey in shallow, low-salinity areas, belugas feed and molt in the same
locations. Along the Chukchi Sea coast, belugas may feed offshore where prey are more
abundant and move into shallow coastal waters for relatively short periods in order to
facilitate the molt.

Acknowledgments

It is impossible to list the many people who contributed the data included in this study. We
thank them all, especially John Burns, Bob Nelson, Dick Russell, Glenn Seaman, Bob Uhl,
and Craig Whitmore. Support for our field studies of beluga whales was provided by the U.S.
National Oceanic and Atmospheric Administration and the U.S. Minerals Management
Service. Preparation and printing of this manuscript were supported by the Alaska Depart-
ment of Fish and Game and the Department of Fisheries and Oceans Canada.

References
BRAHAM, H. W. 1984. Review of reproduction in and Ecology. U.S. Dep. of Commerce, NOAA,
the White Whale, Delphinapterus leucas, OCSEAP Final Rep. 56: 221-357.
Narwhal, Monodon monoceros, and Irrawad- DAVIS, R. A., AND C. R. EVANS. 1982. Offshore
dy Dolphin, Orcaella brevirostris, with com- distribution and numbers of white whales in
ments on stock assessment. Rep. Int. Whaling the eastern Beaufort Sea and Amundsen Gulf,
Commn (Spec. Issue 6): 81-96. summer 1981. Unpubl. Rep. by LGL Ltd. for
BRAHAM, H. W., B. D. KROGMAN, AND G. M. SOHIO Alaska Petroleum Co., Anchorage,
CARROLL, 1984. Bowhead and white whale AK., and Dome Petroleum Ltd., Calgary,
migration, distribution, and abundance in the Alberta. 76 p.
Bering, Chukchi and Beaufort seas, FAY, F. H. 1974. The role of ice in the ecology of
1975-1978. U.S. Dept. of Commerce, NOAA, marine mammals of the Bering Sea, p.383-399.
Tech. Rep. NMFS SSRF-hyph. 778: 39 p. In D. W. Hood and E. J. Kelly [ed] Oceano-
BRODIE, P. F. 1971. A reconsideration of aspects graphy of the Bering Sea. Occ. Publ. No. 2,
of growth, reproduction and behavior of the Inst. Mar. Sci., Univ. Alaska, Fairbanks, AK.
white whale (Delphinapterus leucas), with 623 p.
reference to Cumberland Sound, Baffin Island. FINLEY, K. J. 1982. The estuarine habit of the
J. Fish. Res, Board Can. 28: 1309-1318. beluga or white whale Delphinapterus leucas.
BROOKS, J. W. 1954. Beluga investigations. Ann. Cetus 4(2): 4-5.
Rep. Alaska Dep. Fisheries, Juneau 6: 51-57. FINLEY, K. J., AND R. A. DAVIS. 1984. Reactions
1955. Beluga investigations. Ann. Rep. of beluga whales and narwhals to ship traffic
Alaska Dep. Fisheries, Juneau 7: 98-106. and ice-breaking along ice edges in the eastern
1956. Beluga investigations. Ann. Rep. Canadian High Arctic: 1982-1984. An over-
Alaska Dep. Fisheries, Juneau 8: 52-58. view. Unpubl. Rep. by LGL Ltd. for Can. Dep.
BauEccEmAN J. J., AND R. GROTEFENDT. 1988. Indian Affairs and Northern Development,
Seal, sea lion, walrus and beluga whale sur- Ottawa.
veys of the Bering Sea, 1979 and 1982. U.S. J. F., AND J. S. VANIA. 1971. Killer whale,
Dep. of Commerce, NOAA, OCSEAP Final Orcinus orca, sounds repel white whales, Del-
Rep. 56: 1-152. phinapterus leucas. Fish Bull. 69: 531-535.
BURNS, J.J., AND G. A. SEAMAN. 1988. Investiga- FRAKER, M. A. 1977, The 1977 whale monitoring
tions of belukha whales in coastal waters of program, Mackenzie estuary, N.W.T. Unpubl.
western and northern Alaska. Part II. Biology

56
 Rep. by F. F Slaney & Co., Ltd., Vancouver, 1985: with a seven year review, 1979-85.
B. C. 53 p. Naval Ocean Systems Cent. Tech. Rep. 111:
1979. Spring migration of the bowhead 142 p.
(Balaena mysticetus) and white whales (Del- 1987. Distribution, abundance, behavior,
phinapterus leucas) in the Beaufort Sea, Can. and bioacoustics of endangered whales in the
Fish. Mar. Serv. Tech. Rep. 859: vi + 36 p. Alaskan Beaufort and eastern Chukchi seas,
FRAKER, M. A., D. E. SERGEANT, AND W. HOEK. 1979-86. Naval Ocean Systems Cent. Tech.
1978. Bowhead and white whales in the sou- Rep. 1177: 187 p.
thern Beaufort Sea. Beaufort Sea Proj. Tech LOWRY, L. F., J. J. BURNS, AND R. R. NELSON.
Rep. 4: 114 p. 1987a. Polar bear, Ursus maritimus, predation
FRAKER, M. A., C. D. GORDON, J. W. MCDONALD, on belugas, Delphinapterus leucas in the Be-
J. B. FORD, AND G. CAMBERS. 1979. White ring and Chukchi seas. Can. Field-Nat. 101:
whale (Delphinapterusleucas)distribution and 141-146.
abundance and the relationship to physical and LOWRY, L. F., R. R. NELSON, AND K. J. FROST.
chemical characteristics of the Mackenzie 1987b. Observations of killer whales, Orcinus
estuary. Can. Fish. Mar. Serv. Tech. Rep. 863: orca, in western Alaska: sightings, strandings,
59 p. and predation on other marine mammals. Can.
FROST, K. J., L. F. LOWRY, AND J. J. BURNS. Field-Nat. 101: 6-12.
1983a. Distribution of marine mammals in the SEAMAN, G. A., AND J. J. BURNS 1981. Prelimina-
coastal zone of the Bering Sea during summer ry results of recent studies of belukhas in
and autumn. U.S. Dep. of Commerce, NOAA, Alaskan waters. Rep. Int. Whaling Comm. 31:
OCSEAP Final Rep. 20: 365-562. (Available 567-574.
from NOAA-OMA-OAD, Alaska Office, 701 SEAMAN, G. A., L. F. LOWRY, AND K. J. FROST.
C Street, P.O. Box 56, Anchorage, AK 99513, 1982. Foods of belukha whales (Delphinapte-
USA.) rus leucas) in western Alaska. Cetology 44:
1983b. Distribution of marine mammals in 1-19.
the coastal zone of the eastern Chukchi Sea SEAMAN, G. A., K. J. FROST, AND L. F. LOWRY.
during summer and autumn. U.S. Dep. of 1988. Investigations of belukha whales in
Commerce, NOAA, OCSEAP Final Rep. 20: coastal waters of western and northern Alaska.
563-650. (Available from NOAA-OMA- Part I. Distribution, abundance and movements.
OAD, Alaska Office, 701 C Street, P.O. Box U.S. Dep. of Commerce, NOAA, OCSEAP
56, Anchorage, AK 99513, USA.) Final Rep. 56: 153-220. (Available from
FROST, K. J., L. F. LOWRY, AND R. R. NELSON. NOAA-OMA-OAD, Alaska Office, 701 C.
1984a. Investigations of belukha whales in Street, P.O. Box 56, Anchorage, AK 99513,
coastal waters of western and northern Alaska, USA.)
1982-1983: marking and tracking of whales in SERGEANT, D. E. 1973. Biology of white whales
Bristol Bay. U.S. Dep. of Commerce, NOAA, (Delphinapterus leucas) in western Hudson
OCSEAP Final Rep. 43: 461-585. (Available Bay. J. Fish. Res. Board Can. 30: 1065-1090 ,
from NOAA-OMA-OAD, Alaska Office, 701 SERGEANT, D. E. AND P. F. BRODIE. 1969. Body
C Street, P.O. Box 56, Anchorage, AK 99513, size in white whales, Delphinapterus leucas. J.
USA.) Fish. Res. Board Can. 26: 2561-2580.
1984b. Belukha whale studies in Bristol ST. AUBIN, D. J. , T. G. SMITH, AND J. R. GERACI.
Bay, Alaska, p. 187-200. /n Proc. of the work- 1990. Seasonal epidermal molt in beluga
shop on biological interactions among marine whales, Delphinapterus leucas. Can J. Zool.
mammals and commercial fisheries in the 68: 359-367.
southeastern Bering Sea. Alaska Sea Grant STE WART, B. S., F. T. AWBREY, AND W. E. EVANS.
Rep. 84-1.300 p. 1983. Belukha whale (Delphinapterus leucas)
1985. Radiotagging studies of belukha responses to industrial noise in Nushagak Bay,
whales (Delphinapterus leucas) in Bristol Bay, Alaska. U.S. Dep. of Commerce, NOAA,
Alaska. Mar. Mammal. Sci. 1: 191-202. OCSEAP Final Rep. 43: 587-616. (Available
GUREVICII, V. S. 1980. Worldwide distribution from NOAA-OMA-OAD, Alaska Office, 701
and migration patterns of the white whale C Street, P.O. Box 56, Anchorage, AK 99513,
(beluga), Delphinapterus leucas. Rep. Int. USA.)
Whaling Comm. 30: 465-480. SMITH, T. G., AND M.O. HAMMILL. 1986. Popula-
LJUNGB LAD, D. K., S. E. MOORE, J. T. CLARKE, tion estimates of white whales, Delphinapte-
AND J. C. BENNETT. 1986. Aerial surveys of rus leucas, in James Bay, eastern Hudson Bay
endangered whales in the northern Bering, and Ungava Bay. Can. J. Fish. Aquat. Sci. 43:
eastern Chukchi, and Alaskan Beaufort seas, 1982-1987.

57
 I

58
Age—Length and Length—Weight Comparisons in the Beluga,
Delphinapterus leucas

D. W. Doidge
Department of Renewable Resources, Macdonald College of McGill University,
21,111 Bord du Lac, Ste-Anne-de-Bellevue, Quebec, Canada H9X 1C0

DOIDGE, D. W. 1990. Age—length and length—weight comparisons in the beluga,

Delphinapterus leucas, p. 59-68. /n T. G. Smith, D. J. St. Aubin, and J. R.
Geraci [ed.] Advances in research on the beluga whale, Delphinapterus
leucas. Can. Bull. Fish. Aquat. Sci. 224.

Examination of published and new data of age-length and length—weight support

the hypothesis that adult belugas, Delphinapterus leucas, vary in body size
between regions, but not to the large extent reported previously. Age—length
relations differ between sexes and regions. Adult males are slightly longer than
females. Length—weight relations were similar among the sexes and regions.
Belugas from Hudson Bay tend to be shorter, and presumably lighter than those
from Alaska, W. Greenland, the St. Lawrence River, and the MacKenzie Delta.
These differences are generally too small to classify individuals to region on the
basis of size. In eastern Hudson Bay belugas, body weight (kg) can be predicted
from length (cm) and maximum girth (cm) using the relationship
Weight = 10-433Length2 46Girth' 36 (R2=0.96, n=-17) for males, and
Weight = 10-2 .95Lengthl."Girthl . 7 ' (R 2=0.95, n=20) for females.
Lack of data on seasonal changes in food intake and energy reserves prevents
testing of the hypothesis that the small differences observed in belugas from
different regions is due to differences in winter productivity.

Une analyse de données publiées et de nouvelles données sur le rapport âge —

longueur et longueur—poids viennent appuyer l'hypothèse que l'on retrouve une
variation régionale de taille chez les marsouins blancs adultes, Delphinapterus
leucas, mais d'une échelle moindre que rapportée antérieurement. Le rapport âge
— longueur varie entre les sexes et les régions. Les mâles adultes sont légèrement
plus longs que les femelles. Le rapport longueur-poids est similaire entre les sexes
et les régions. On note que les marsouins blancs de la baie d'Hudson ont tendance
à être plus courts, et, présumément, plus légers que ceux de l'Alaska, de l'ouest
du Groenland, du fleuve St-Laurent et du delta du Mackenzie. En prenant en
considération les données sur la taille, les écarts sont généralement trop faibles
pour classifier les individus par provenance géographique. Chez les marsouins
blancs de l'est de la baie d'Hudson, le poids (kg) peut-être calculé à partir de la
longueur (cm) et du tour de taille maximal (cm) en utilisant l'équation
Poids = 10-4.33Longueur2.46Tour de taille. 36 (R2=0.96, n=17) pour les mâles, et
Poids = 10-3.95Longueurl "Tour de taille'. 2 ' (R2=0.95, n=-20) pour les
femelles.
Un manque de données sur la variation de l'alimentation saisonnière et des réserves
énergitiques ne permet pas la vérification de l'hypothèse que les petits écarts
observés chez les marsouins blancs des diverses régions sont dûes à une différence
dans la productivité hivernale marine.

59
 Introduction

The possibility of stock recognition based on body size has important management
implications for belugas, Delphinapterus leucas, since whale numbers and harvest intensity
vary greatly between regions. In the eastern subarctic, belugas that summer in Hudson Bay,
Ungava Bay and Cumberland Sound appear to overwinter together in Hudson Strait (Finley
et al. 1982; Reeves and Mitchell 1987). Presently, the degree of exchange between these
proposed stocks is unknown. Differences in body size would imply genetic discreteness and
lack of interchange between these regions.
Based on differences in extrapolated body weight, which can vary threefold for adult
males, Sergeant and Brodie (1969) conclude that three different beluga populations inhabit
eastern North America: small animals in (western) Hudson Bay, medium sized animals in the
Canadian Arctic and the St. Lawrence River, and large animals along the West Greenland
coast. They suggest winter productivity is the most important factor in determining body size
in belugas. Although Sergeant and Brodie were working with the best data available at the
time, recent data and an error in the published regression equation for Hudson Bay belugas
(D.E. Sergeant, 555 St. Pierre Blvd., Ste-Anne de Bellevue, Qué., H9X 3R4, pers. comm.)
warrant a reexamination and interpretation of body weight differences in beluga.
Here, I compare both age—length and length—weight relations for beluga by sex and
region and develop equations for predicting body weight in the eastern Hudson Bay stock
based on length and maximum girth.

Methods

Age — length data were gathered in 1980 and between 1983-87 as part of a wider study
of beluga population ecology in N. Quebec centred at the Nastapoka estuary, eastern Hudson
Bay (56°55N, 76°33W). These data also include a small number of whales harvested on the
southern side of Hudson Strait and Ungava Bay. Teeth were removed by boiling and were
stored dry. A longitudinal mid-section 0.3 mm thick was removed from the most anterior
tooth of the lower jaw (usually the left) using a Buehler Isomet slow speed saw (Buehler Ltd.,
41 Waukegan Rd., Lake Bluff, IL., USA, 60043). Sections were stored in absolute ethanol.
Growth layer groups (GLGs, terminology of Perrin and Myrick 1980) were counted in un-
stained sections using 6-10 power of a binocular microscope. Two GLGs were considered
to represent an annual increment (Sergeant 1973; Goren et al. 1987). Teeth were regarded as
unworn if the neonatal line was present in the dentine.
Age—length data, based on unworn teeth, from N.W. Alaska (Burns and Seaman 1985,
Fig. 2), Churchill' (D.E. Sergeant, 555 St. Pierre Blvd., Ste-Anne-de-Bellevue, Qué., H9X
3R4, unpubl. data) and N. Quebec (this study) had adequate sample sizes (n> 30 of each sex)
for regression-based comparisons. Logistic, parabolic, von Bertalanffy and Gompertz
growth curves (Kaufmann 1981; Innes et al. 1981; Roff 1980) were fitted to the age — length
data from each of the above sites using the NLIN procedure of PC—SAS (Ver. 6.03, Statistical
Analysis Systems, Cary, NC, USA). The Richards model was considered to be too sensitive
to imprecision in the small datasets available (Zach 1988) so was not fitted. Whales having
unworn teeth from the St. Lawrence River (Sergeant 1986, table VIII, n = 16), Bristol Bay,
Alaska (Burns and Seaman 1985, fig. 3, n = 18), the Mackenzie Delta' (D.E. Sergeant, 555
St. Pierre Blvd., Ste-Anne-de-Bellevue, unpubl. data) and West Greenland (R. Dietz,
Greenland Environment Research Institute, Tagensvej 135, 1. sal. 2200, Copenhagen,
Denmark, unpubl. data, n = 9) were compared to the 95% confidence intervals of asymptotic
length for N.W. Alaska, Churchill and N. Quebec.

' Sergeant (1973, fig. 10) shows mean values.

= Sergeant (1973, fig. 6) contains wom and unworn teeth.

60
 Thirty-seven beluga (17 male, 20 female) shot by Inuit between late June and early
September of 1985 and 1986 in the Nastapoka estuary were weighed using a 2 500 lb. capacity
scale (Model S-250, Dillon, 14620 Keswick St. Van Nuys, Ca, USA) attached to a tripod by
a 5 t capacity chain hoist. Weights were recorded to the nearest 5 lb then converted to the
nearest kg. Sex, reproductive status (pregnant, lactating or barren), standard length (Amr.
Soc. Mammal 1961) and maximum girth (which occurs near the umbilicus) were noted.
Length — weight data from the St. Lawrence River (Vladykov 1944, n = 10), Churchill,
Manitoba (Doan and Douglas 1953, n = 16), Seal River, Manitoba (Geraci, unpubl. data, n
= 9) and Cumberland Sound, Baffin Island (Brodie 1989, n = 10) had inadequate sample sizes
for regression-based comparisons. Data from these regions were compared to the 95%
confidence intervals of the length — weight relationship found for eastern Hudson Bay (n =
36).
Statistical analysis generally followed the procedure and philosophy put forth by Peters
(1983, Chapter 2). When sample size was judged adequate, within region comparisons of the
regression parameters of males and females used standard F tests (Snedecor and Cochran
1967). Most other comparisons were based on confidence intervals because sample mean,
range, variance or size often differed between data sets. Confidence intervals show the degree
of uncertainty associated with the data and facilitate assessment of practical and statistical
significance associated with any differences (SAS Institute 1987). The level of statistical
significance for Type I error was 0.05.

Results

In the age—length analyses, Gompertz and von Bertalanffy models consistently produced
the smallest and similar residual sum of the squares. Because the von Bertalanffy curves
severely underestimated body length at young ages, I chose the Gompertz equation to
describe beluga age—length relationships (Fig. 1).
Sexual dimorphism is apparent from the non-overlapping of the 95% confidence
intervals of asymptotic length for N.W. Alaska and Churchill belugas (Table 1, Fig. 2), males
being slightly longer than females. This difference was not significant in the N. Quebec
sample where confidence intervals of asymptotic length of males and females overlap.
Regional comparisons of age—length relationships are shown graphically (Fig. 2) where the
fitted curves represent larger datasets (See Table 1 for growth coefficients) and age — length
is plotted individually for the smaller datasets. Belugas from Churchill and N. Quebec
(henceforth termed Hudson Bay) show similar changes in length with age and attain similar
adult body lengths. Beluga from N.W. Alaska are significantly longer when fully grown than
beluga from Hudson Bay, the absolute difference being greater in males. Male and female
beluga from Hudson Bay also appear slightly shorter than their counterparts from Bristol Bay
(Alaska), Greenland and the St. Lawrence, but the small sample size prevents meaningful
statistical comparison.

Length—weight relationships from the Nastapoka River showed no sexual dimorphism

(comparison of slopes F133 2.09 intercepts F i 3 4.05, P>0.05) so a single regression line was
,

fitted. Body weight (kg) of eithe' r sex of beluga from eastern Hudson Bay can be estimated
from body length (cm) using the equation

Weight = 10-384 Length 2 " (R 2 = 0.92, n = 36)

Length — weight data from W. Hudson Bay and the St. Lawrence lie within the 95% C.I.
of E. Hudson Bay (Fig. 3) demonstrating that beluga do not differ significantly in length-
weight relationships between those regions. The Churchill animals were known to have been
gutted (Doan and Douglas 1953) which explains their low values (Fig 3).

61
 a 500
500
400
400
300
300
200
200
100
100
0
0 10 20 30 40 50
BODYLENGTH(CM)

o 10 20 30 40 50

BODYLENGTH(CM)
500
500
400
400
300
300
200 oet 200

100 100
B.
0
00 10 20 30 40 50 10 20 30 40 50

500 500

400 400

300 300

200 200

100 100

00 00 10 20 30 40 50
10 20 30 40 50

AGE (GLGs/2)
AGE (GLGs/2)
FIG. 1. Gompertz growth curves fitted to age—length data of beluga whales from various regions. A.
NW Alaska, B. Churchill, Manitoba, C. N Quebec. (a) females, (b) males. Data sources given in text.
Circles represent individual data points.

For the Nastapoka River data (Table 2), least squares regression showed significant
relationships between log io standard length and log io body weight in males (R2=0.95) and
females (R2=0.80). The addition of a maximum girth term improved the fit of the female
regression model (R2=0.95), but did not noticeably change the explained variance in the male
regression model (R2=0.96). Lactation did not affect female weight (ANOVA F 1 0.05, ,,

P>0.05). Body weight (kg) can be estimated from length (cm) and girth (cm) for fei-nales by

Weight = 10-196 Length'. 71 Girth'm = 0.95, n = 20) and males by

Weight = 10-4." Length°36 Girth2.46 (le = 0.96, n = 17).

Discussion

The hypothesis that belugas differ in body size between regions is supported by re-
examination of the data (Fig. 1, 2, 3), but not to the degree reported previously (Sergeant and
Brodie 1969). The conclusion that male belugas can differ by up to threefold in body weight
likely results from extrapolations made from the St. Lawrence power curve describing body
length and mass (Sergeant and Brodie 1969), and the application of an incorrect gutting factor
to Hudson Bay animals.

62
 A 500 B 500

400 400
BO DYLEN GTH(CM)

300

200

100 100

0 Hiimului,,,,,1.11.11.111111.11111
0 10 20 30 40 50 0 10 20 30 40 50

AGE (GLGs/2)
FIG. 2. Gompertz growth curves fitted to age—length data of beluga whales from NW Alaska (solid
line), Churchill, Manitoba (dots) and N Quebec (dashes) with individual data points from Bristol Bay,
Alaska (circles), the St. Lawrence River (triangles), West Greenland (squares) and the MacKenzie
Delta (asterisks). A. females, B. males. Vertical lines represent asymptotic 95% confidence intervals.
Data sources and curve coefficients given in text.

TABLE 1. Growth parameters (X ± asymptotic se) of the Gompertz growth equation fitted to age — length
data from beluga whales from various sites where
Body length (cm) = A(exp(-b.exp(-k age))) and age is GLGs/2.
Region Sex A SE b SE k SE

NW Alaska" 359 F 4 0.30 0.04 0.68 0.07 51

427 M 8 0.14 0.02 0.74 0.06 56
Churchill' 328 F 6 0.21 0.03 0.64 0.04 77
377 M 14 0.17 0.03 0.75 0.04 65
N Quebec" 330 F 6 0.27 0.06 0.60 0.10 55
349 M fo 0.28 0.05 0.71 0.07 44
Burns and Seaman (1985).
b D.E. Sergeant, pers. comm.
This study.

63
 1500
- 95% C.I. E. HUDSON BAY
1400
o ST. LAWRENCE RIVER 0
1300 0
D SEAL RIVER
1200
A CUMBERLAND SOUND
1100
* CHURCHILL
1000
BODYWEI GHT(KG)

900

800

700

600

500

400

300

200

100

0 I , , I
100 200 300 400 500

BODY LENGTH (CM)

FIG. 3. Body length—weight relationships of beluga whales

from various regions. Data sources given in text. Confidence
intervals are asymmetric due to the log—log transformation of
the data.

Age — length comparisons of N. Quebec and St. Lawrence animals showed slight
differences between sites. Mean predicted lengths of adults from N.W. Alaska are ca. 30 —
50 cm longer than Hudson Bay animals. The limited data from W. Greenland, the St.
Lawrence, the MacKenzie Delta, and Bristol Bay show a similar trend, but individuals
generally cannot be assigned to stocks on the basis of body size because of overlapping
confidence intervals of length between regions (Fig. 2).
The long body lengths attributed to Greenlandic beluga (Degerbol and Neilsen 1930) and
used for weight extrapolations by Sergeant and Brodie (1969) are not evident in present
Greenlandic harvests (R. Dietz, Greenland Environment Research Institute, Tagensvej 135,
1. sal. 2200, Copenhagen, Denmark, unpubl. data). Vladykov (1943) remarked on the lower
fluke width to body length ratios for West Greenland (Degerbol and Neilsen 1930) compared
to St. Lawrence belugas. He suggested this difference could be real or due to differing
precision of measurement. A decrease in the abundance of older age classes or possible use
of either snout to fluke tip or curvilinear lengths by Degerbol and Neilsen (1930) could
explain this discrepancy. The similarity of recent fluke width to body length ratios from W.
Greenland (R. Dietz, Greenland Environment Research Institute, Tagensvej 135, 1. sal. 2200,
Copenhagen, Denmark, unpubl. data) to those of N. Quebec and St. Lawrence beluga (Fig. 4)
implies Degerbol and Neilsen's (1930) body lengths were longer than the present accepted

64
TABLE 2. Weight, length and maximum girth of beluga whales taken at the Nastapoka River, N. Quebec.
Sex Weight Length Girth Sex Weight Length Girth
kg cm cm kg cm cm
F 345 324 176 M 681 350 118
402 319 188 375 284 172
275 252 168 518 330 200
336 273 172 284 268 152
513 319 200 885 386 254
306 262 172 341 284 178
456 332 189 438 322 190
558 338 208 552 344 188
415 339 174 304 320 100
486 334 182 470 319 204
361 316 171 388 316 162
304 293 164 238 262 137
470 320 196 459 344 178
375 299 170 363 303 160
440 301 190 64 149 100
409 310 186 290 282 161
445 347 174 F 1000+' 394 224
179 221 140
427 323 184
600 360 210
Incomplete weight, D.M. Albright, Dept. Fisheries and Oceans, Ste-Anne-de-Bellevue, Quebec, pers.
comm., 85.08.13.

standard measurement. Any bias in length measurement is further amplified in the weight
estimate because of the power nature of the curve (Fig 3).
Sergeant and Brodie (1969) assume only abdominal viscera, representing 5% of body
weight, were lost when the Churchill beluga were gutted and thus divide the original weights
(Doan and Douglas 1953) by 0.95. Comparison of the uncorrected Churchill data with the
larger sample from the Nastapoka (Fig. 3) indicates a larger than 5% difference in body
weight. Since Doan and Douglas (1953) do not explicitly state what organs were removed by
gutting, their weight data has little use in body comparisons between localities.
Comparison of Seal River and Nastapoka length — weight relationships indicate belugas are
similar in body size between western and eastern Hudson Bay. Small differences in length
— weight relations may exist between Hudson Bay belugas and belugas from other regions,
but sample sizes are presently too limited to allow meaningful statistical testing. I believe the
use of extrapolated weights for Greenlandic animals and the more than 5% gutting factor for
Churchill belugas led Sergeant and Brodie (1969) to conclude erroneously that belugas
differed greatly between region in body length — weight relationships.
Comparisons of body size between populations or stocks must consider several factors.
Differences in average body size between populations may only reflect disparity among
samples in age or sex composition rather than a true change in body size at a given age
(Scheffer 1955). Body weight without length of age data is a poor body size criterion since
it is subject to bias related to sampling technique. Hunters tend to select large animals

65
 120
MALES o
110 °
o
o
o
100
90
FLUKEWIDTH(CM)

80
70
o o

60
50
40
30
20
10
0
20 0 I30 I 0 400 500 200 300 I 400 500
BODY LENGTH (CM)

Fin. 4. Body length versus fluke width of beluga whales from the St. Lawrence River (solid line,
Vladykov 1943), W. Greenland (dotted line, Degerbol and Neilsen 1930, open circles, R. Dietz unpubl.
data) and N. Quebec (dashed lines, this study). Vertical bars represent mean ± 2 SE.

(Caughley 1977), whereas shore-tethered nets catch greater numbers of females and calves
(Sergeant and Brodie 1969; Sergeant 1973). Time of sampling may influence weight
relationships through seasonal changes in body condition, so inclusion of body girth in
weight analyses is desirable. Length—weight relations are useful for comparisons of size and
shape, but similar growth dynamics do not dictate that whales of different stocks attain the
same maximum weight or, depending on the age structure, have the same population mean
weight. In the ecological context, age—weight or age—length comparisons may be more useful
since weight at a particular age should on average reflect food availability, but tooth wear and
small harvest sizes limit the size of age based samples of beluga whales.
The similarity of length—weight and the slight differences in age—length relations between
regions implies that body mass per unit length is fairly constant, but that the rate of body
growth (per unit time) varies between regions. Levels of thyroid hormones in the blood
indicate that body growth is accelerated in the summertime when belugas frequent warmer
estuarine waters (St. Aubin and Geraci 1988). Although girth had a significant effect on body
weight in females, seasonal changes in blubber reserves are not well enough documented to
indicate if body growth is fuelled by summer feeding or mobilization of reserves built up on
the wintering grounds. Thus the data are still not adequate for testing the hypothesis that
winter productivity is responsible for small body size differences in belugas between regions.

Acknowledgments

I thank the beluga hunters of N. Quebec for access to their catch and personnel from Dep.
Fisheries and Oceans, Makivik Corp., and the University of Guelph who helped examine it.

66
Data gathered in N. Quebec in 1980 by K. Finley and others of LGL Ltd., Toronto, was made
available for analysis. R. Dietz, J.R. Geraci and D.E. Sergeant supplied unpublished data on
body size. C. Hudon and M. Hammill offered advice on SAS programming and statistics.
G.G. Beck, R. Dietz, J.R. Geraci, M.O. Hammill, L. Lowry, D.E. Sergeant, T.G. Smith and
two anonymous reviewers suggested improvements to the manuscript.
This work was supported by the Dept. of Fisheries and Oceans Canada, World Wildlife
Fund Canada's "Whales Beneath the Ice" project, the Centre for Northern Studies, McGill
University and an NSERC postgraduate scholarship.

References

AMERICAN SOCIETY OF MAMMALOGISTS, COM- INNES, S., R. E. A. STEWART, AND D. M. LA VIGNE

MITTEE ON MARINE MAMMALS. 1961. in northwest Atlantic harp seals 198.Growth
Standardized methods for measuring and Phoca groenlandica. J. Zool. (Lond.)
recording data on the smaller cetaceans. J. 194: 11-24.
Mammal. 42: 471-476. KAUFMANN, K. W. 1981. Fitting and using growth
BRODIE, P. F. 1989. The white whale or beluga curves. Oecologia 49: 293-299.
Delphinapterus leucas, p.61-68. In S.H. Ridg- PETERS, R. H. 1983. The ecological implications
way and R.J. Harrison [ed.]. The handbook of of body size. Cambridge University Press.
marine mammals. Vol. 4. Academic Press. Cambridge, U.K. 329 p.
BURNS, J.J., AND G.A. SEAMAN 1985. Investiga- PERRIN, W.F., AND A.C. MYRICK, JR. [ED.]. 1980.
tions of Belukha whales in coastal waters of Age determination of toothed whales and S ire-
western and northern Alaska. II. Biology and nians. Rep. Int. Whaling Comm. (Spec. Issue.
ecology. Final Report submitted to NOAA, 3): 229 p.
Outer Continental shelf Environmental REEVES, R. R., AND E. D. MITCIIELL. 1987. Histo-
Assessment Program. Alaska Dep. of Fish and ry of white whale (Delphinapterus leucas) ex-
Game, 1300 College Road, Fairbanks, Alaska, ploitation in eastern Hudson Bay and James
USA 99701.129 p. Bay. Can. Spec. Publ. Fish. Aquat. Sci. 95:
CAUGIILEY, G. 1977. Analysis of vertebrate popu- 45p.
lations. John Wiley and Sons, Toronto, Ont. ROFF, D. A. 1980. A motion for the retirement of
234 p. the von Bertalanffy function. Can. J. Fish.
DEGERBOL, M., AND N. L. NEILSEN. 1930. Biolo- Aquat. Sci. 37: 127-129.
giske iagttagelser over og maalinger of hvi- SAS INSTITUTE INC. 1987. SAS/STAT' Guide
dhavlen Delphinapterus leucas Pall. og dens for Personal Computers, Version 6 Edition.
fostre. (Biological observations and measure- Cary, NC, USA. SAS Institute, 1028 p.
ments of the white whale and its fetus). Medd. ST. AUBIN, D. J. AND J. R. GERACI. 1988. Seasonal
Gronland 77: 119-144. variation in thyroid morphology and secretion
DOAN, K. H., AND C. W. DOUGLAS. 1953. Beluga in the white whale, Delphinapterus leucas.
of the Churchill region of Hudson Bay. Bull. Can. J. Zool. 67: 263-267.
Fish. Res. Board. Can. 98: 27 p. SCIIEFFER, V. B. 1955. Body size with relation to
GOREN, A. D., P. F. BRODIE, S. SPOTTE, G. C. RAY, population density in mammals. J. Mammal.
H. W. KAUFMAN, A. J. GWINNETT, J. J. SCUIB- 36: 493-515.
BA, AND J. D. BUCK. 1987. Growth layer groups SERGEANT, D.E. 1973. Biology of white whales
(GLGs) in the teeth of an adult belukha whale (Delphinapterus leucas) in western Hudson
(Delphinapterus leucas) of known age: Evi- Bay. J. Fish. Res. Board. Can. 30: 1065-1090.
dence for two annual layers. Mar. Mammal. 1986. Present status of white whales Del-
Sci. 3: 14-30. phinapterus leucas in the St. Lawrence Estuary.
FINLEY, K. J., G. W. MILLER, M. ALLARD, R. A. Naturaliste can. 113: 61-81.
DAVIS, AND C. R. EVANS. 1982. The belugas SERGEANT, DE., AND P. F. BRODIE. 1969. Body
(Delphinapterus leucas) of northern Quebec: size in white whales, Delphinapterus leucas.
distribution, abundance, stock identity, catch J. Fish. Res. Board. Can. 26: 2561- 2580.
history and management. Can. Tech. Rep. SNEDECOR, G. W., AND W. G. COCIIRA N. 1967.
Fish. Aquat. Sci. 1123: 57 p. Statistical methods. Iowa State Univ. Press.,
Ames, Iowa, USA. 593 p.

67
VLADYKOV, V. D. 1943. A modification of the que de marsouin blanc ou beluga (Delphinap-
pectoral fins in the beluga from the St. Law- terus leucas) du fleuve et du golfe St. Laurent.
rence River. Naturaliste can. 70: 23-40. Dép. Pêch. Prov. Qué. 194 p.
1944. Études sur les mammifères aquati- ZACH, R. 1988. Growth curve analysis: a critical
ques. III. Chasse, biologie et valeur économi- reevaluation. Atik 105: 208-210.

68
 Philopatry and Site Tenacity of Belugas, Delphinapterus leucas,
Hunted by the Inuit at the Nastapoka Estuary, Eastern Hudson Bay

L.M.J. Caron and T.G. Smith'

Department of Fisheries and Oceans, Arctic Biological Station,
555 Blvd. St. Pierre, Ste-Anne-de-Bellevue, Québec, Canada H9X 3R4

CARON, L. M. J., AND T. G. SMITH. 1990. Philopatry and site tenacity of belugas,
Delphinapterus leucas, hunted by the Inuit at the Nastapoka estuary,
eastern Hudson Bay, p. 69-79. In T. G. Smith, D. J. St. Aubin, and J. R.
Geraci [cd.] Advances in research on the beluga whale, Delphinapterus
leucas. Can. Bull. Fish. Aquat. Sci. 224.

Observations were made of white whales, Delphinapterus leucas, in the Nasta-

poka River estuary of eastern Hudson Bay during the summers of 1983 and 1984.
The highest herd counts in 1983 and 1984 were 245 and 260 whales. The location
of belugas was found to be strongly influenced by the tide, the total number of
whales in the estuary, the state of the water surface, and the total time spent by the
whales in the estuary. Of the five size classes that could be discerned based on
lengths of grey immatures relative to adult length, neonates 1/3 length of adult)
represented 16.3% of the estuarine herd in 1983 and 19.1% in 1984. All non-white
size classes totalled 61.5% of the 1983 herd and 60.5% of the 1984 herd. In the
1984 season, the adult male to adult female sex ratio was estimated as 1:4.3. Some
whales identified by their scars were seen both years, and throughout the same
season. In 1984, an average daily presence of 30% was calculated for 24 of the most
conspicuously marked whales. Average periods of 40 and 24h were required for
the whales to reoccupy the estuary after disturbances from hunts or motor traffic.

Les marsouins blancs, ou bélugas, Delphinapterus leucas, de l'estuaire de la

rivière Nastapoka, située sur la côte est de la baie d'Hudson, ont été observés au
cours des étés 1983 et 1984. Les plus grands troupeaux dénombrés en 1983 et 1984
comprenaient de 245 et 260 bélugas respectivement. La répartition des individus
à l'intérieur de l'estuaire était influencée par la marée, le nombre total de bélugas
dans l'estuaire, l'amplitude des vagues et le temps écoulé depuis leur arrivée dans
l'estuaire. Cinq classes de tailles relatives ont été identifiées, dont celles des
nouveau-nés mesurant 1/3 de la longueur des adultes, représentait 16.3 % des
dénombrements en 1983 et 19.1 % en 1984. Les classes bélugas gris prises
globalement totalisaient 61.5 % et 60.5 % respectivement des dénombrements de
1983 et 1984. En 1984, l'estimé du rapport mâles adultes sur femelles adultes était
de 1:4,3. Des individus identifiés à l'aide de marques naturelles et de cicatrices ont
été vus chaque année et de façon répétitive au cours d'une même saison. Un taux
de la présence journalière de 30 % en 1984 a été calculé pour 24 des individus les
plus facilement reconnaissables. Des périodes moyennes de 40 et 24 h étaient
nécessaires pour le retour des bélugas dans l'estuaire après les évacuations causées
par les chasses et le déplacement des embarcations motorisées.

' Author to whom correspondence should be addressed.

69
 Introduction

White whales or belugas are the most abundant of the odontocete species occupying
Canadian arctic waters. During the summer months, belugas congregate in certain estuaries.
The occupation of traditional sites along with indications of differences in the body size of
belugas from different areas has led to the delineation of several separate stocks in Canada
(Sergeant 1986; Smith et al. 1985; Brodie et al. 1981; Sergeant 1981; Finley et al. 1982;
Fraker 1980; Vladykov 1944).
In the 19th century, herds of belugas in eastern Hudson Bay occupied the Great Whale and
Little Whale rivers by the "thousands" during the summer months, and there were reports of
herds of similar size farther north in Richmond Gulf and in the "Nistabucky" River (probably
the Nastapoka River) (Francis 1977). Richmond Fort, on Richmond Gulf, was the first beluga
hunting site used by the Hudson Bay Company in 1750, where there was a harvest of whale
skins valued for their high quality leather. Native whale hunting continued for the next 100
yr or so along the eastern Hudson Bay coast. Then, from 1854 to 1870, newer whaling
techniques using nets decimated the Great Whale and Little Whale River herds to the point
where the remaining whales would no longer use the estuaries, and the "porpoise fishery" had
to be abandoned (Francis 1977).
The Nastapoka River estuary is currently the main major summer concentration area of
beluga whales in eastern Hudson Bay, and it is also an important site of Inuit whale hunting.
Approximately 30 whales are taken at this site every year from a population on which there
is still limited information. The earliest coastal aerial survey conducted in 1978 in the Great
Whale area (Breton-Provencher 1980) counted between 242-254 belugas; subsequent
nearshore counts in 1982 showed only 162 belugas along the whole eastern Hudson Bay
coastline (Finley et al. 1982). More recent comprehensive systematic surveys including the
whole of James Bay and areas offshore as far as the Belcher Islands have yielded estimates
of some 1124-1904 belugas for the eastern Hudson Bay stocks (Smith and Hammill 1986).
In their survey the coastal concentrations alone totalled 474, of which 139 individuals were
seen at the Nastapoka estuary.
The persistent habit of estuarine aggregation makes the beluga one of the most vulnerable
arctic marine mammal species to overharvesting and destruction of critical habitat. Here we
describe the numbers and age structure of the Nastapoka belugas and their tenacity to site in
the face of hunting and human disturbance.

Materials and Methods

The Nastapoka estuary, on the east coast of Hudson Bay (56°55'N, 76°36'W) is 1.3 km long
from the river mouth to the 30m high waterfalls and 0.2 km wide in an east—west orientation.
Observations were made from a 6 m high aluminium tower located on the north shore of the
river mouth (Fig. 1). The study was conducted from 20 July to 13 August 1983 and 21 June
to 2 September 1984. In 1983 the observations were non-systematic and selective of
individuals with identifying marks. In 1984, scan samples (Altmann 1974) were taken every
2 h from 09:00 to 21:00 daily for a total of 48 d. More detailed focal samples concentrating
on individuals or groups were also made on 82 occasions totalling 390 min of observations.
The estuary was divided into four quadrats from which we recorded the frequency of the
different size classes of whales along with time, tidal height, water temperature, wind speed
and direction, surface conditions and clarity (ordinal scale 1 to 4; choppy to calm and clear)
and light conditions (scale 1 to 4; strong glare to diffuse light).
The ages of different size classes were verified by analysis of carcasses, which had been
aged by tooth sections, in the catches from Whale Cove, western Hudson Bay from 1962 to
1964 (Sergeant 1973) and catches from Cumberland Sound, eastern Baffin Island (Brodie
1967). Data were also obtained from carcasses of belugas taken in eastern Hudson Bay

70
 NASTAPOKA RIVER, NORTHERN QUEBEC

Fig. 1. The study area at the Nastapoka River estuary showing the different quadrats observed from the
tower.

(Breton-Provencher 1980). Ages were assigned to the carcasses based on counts of two
dentinal layers per year (Goren et al. 1987). Mean lengths of di fferent age classes were
calculated as proportions of the mean lengths of adult (white) females in the catch (Table 1).
Individual belugas were identified from scars which appeared for the most part to have been
old bullet wounds. Only scars located on the left side of the individual were used to minimize
error in re-identification.
Spearman rank correlations were used to test the relationship between the distribution of
belugas in the estuary and physical factors. Chi square tests were performed on the
contingency tables of the age structures, and a deviation statistic Z (Legendre and Legendre
1984) was calculated for each cell to verify which observed frequency deviated significantly
from the expected one. Mann—Whitney U tests were used to compare the means of recovery
times of whale numbers in the estuary after hunting to the recovery times after motor traffic.

Results

Numbers, Distribution, and Population Structure

In 1983, whales were present in the estuary both before and after the period of observations.
In 1984, however, the first whales of the season were seen on 23 June, but were still present
after observations ended. Numbers of belugas fluctuated daily throughout both seasons.
Some fluctuations were caused by boat traffic and whale hunting, while others were
unexplained. The highest count for 1983 was 245 and occurred on 27 July. In 1984, the
highest count, 260 belugas, was made on 25 August. During the 1984 season, the herd size
gradually increased in July and August, then started to decrease in early September (Fig. 2.)
On 22 July 1984, low altitude aerial photos taken from helicopter showed that there were
160 whales at the surface and between 20 and 30 whales visible below the water surface.
Counts made from the observation tower two hours before and one half hour after this flight
showed 175 and 145 whales in the study area respectively, closely in agreement with the
aerial estimates.

71
TABLE 1. Measured nose to tail length (cm) of belugas taken in three Canadian arctic areas with calculated propo rt ional mean lengths of different age calves (sexes lumped) to a mean
lengths of white adult females (%) in the same catches.
Mature 2 yr old 3 yr old 4 yr old 5 yr old 6 yr old
female Neonate Yearling grey grey grey grey grey
References
n n % n % Yn % n % 5n%

Western Hudson 310.5 57 151.9 948.9 198.4 16 63.9 225 16 72.5 242 6 77.9 260.7 19 84.0 273.7 15 88.1 280.1 14 90.2 (Sergeant 1973)
Bay (Whale Cove
1962-1964;
July 31 to 5 Sept.
Eastern Hudson 327.2 75 155.6 5 47.6 228.3 6 69.8 243.3 4 74.4 260.5 6 79.6 255.5 2 78.1 281.3 7 86.0 296 9 90.5 (Breton-Provencher
Bay (24 June to 1980)
25 Nov.)
Cumberland 363.8 13 166.6 1045.7 249.0 2 68.4 - - - 290.8 5 80.0 318.8 5 87.6 331 7 91.0 352 7 96.8 (Brodie 1967)
Sound (July
and Aug.)
 280-

240 -
WEEKLY MAXIMUMHERD SIZE

200-

160-

120 -

80 -

40-1

Juno 30th July 14th July 28th August 11th August 25th

Fig. 2. Seasonal change in the weekly peak number of belugas seen in the Natapoka River estuary.

In the estuary, groups of belugas were observed to swim upstream and downstream
displaying various behaviors. Spearman rank correlation between physical factors and the
median of the distribution of whales in the estuary showed that the median position of the herd
advanced and receded with the flow and ebb of the tide (R -= 0.4, P = 0.0001). It also indicated
that large herds (R = 0.34, P = 0.0001), long periods without disturbances (R = 0.23,
P = 0.0001), high waves (R = 0.32, P= 0.0001), strong northerly winds (R = 0.15, P = 0 .005),
high river water temperature (R = 0.11, P = 0.0332), and clear water (R = 0.12, P = 0.0224)
favored the occupation of the upper reaches of the estuary.
We are easily able to recognize neonate (0+) and yearling (1+) calves by their shape and
size proportional to their accompanying white adult females. Neonates which are bom prior
to entering a high arctic estuary, Cunningham Inlet, N.W.T., probably from mid-June to the
first week of July, are initially very thin, light grey to slightly pink in coloration and measure
from > 1/3 to < 1/2 of the length of their accompanying white adults. Yearling calves are
much bulkier in appearance, are the darkest of the grey immature age classes and measure
> 1/2 to 70% of the length of the white adult females. Both neonates and yearlings remain
very close to their presumed mothers, rarely ever leaving their side. The remaining immature
grey age classes, from 2+ to 6+ yr old, range from > 2/3 to <3/4 (73 — 97%) of the length
of adult females; they are loosely associated with adults and cannot be reliably separated by
visual observation into definite age categories (Table 1).
In 1984, the size class structure was estimated using scan and focal sampling techniques.
Results did not differ significantly. Scan sampling showed that 19.20% of the herd was
composed of neonates (> 1/3 lengths), 15.39% were > 1/2 length yearlings, 11.37% were >
2/3 length greys, 12.73% were >3/4 length greys, and 41.31% were white individuals.
The size class distribution was found to vary slightly, but significantly, with the season
(Table 2). The age class contributing the most to the total variation was the >2/3 length. The
abundance of this age class was high at first, then sharply decreased after the first quarter of
the season, and increased only during the last quarter. The proportions of> 1/3 lengths and

73
TABLE 2. Contingency table of the age structure variation from scan samples of the Nastapoka herd over
the 1984 season (n = 33 027). The dates listed are mean values of dates for blocks of equal sample size.
Asterisks note the values significantly different from the average. (Chi Square = 703.7002, df = 12,
P < 0.005.)
17 July 7 Aug. 20 Aug. 29 Aug. Total
Size class Percent Percent Percent Percent Average
1/3 (neonate) * 17.4 20.4 20.0 18.9 19.2
1/2 (yearling) * 13.7 * 17.4 15.1 15.4 15.4
2/3 * 17.8 " 8.7 * 6.6 12.5 11.4
3/4 * 10.9 * 14.4 12.3 13.2 12.8
White 40.2 * 39.0 * 46.1 40.0 41.2

> 1/2 lengths were slightly below the mean value during the first quarter of the season, but
their numbers increased and stabilized later on.
Only an indirect calculation of the sex ratio could be made. Observations of > 1/3 and
> 1/2 length calves were equated with the presence of an adult female. White animals accom-
panying > 2/3 length calves were also counted as adult females. From this it was calculated
that 0.322 of the Nastapoka herd was composed of adult white female animals. By
subtraction this left 0.0742 adult males, yielding a maximum sex ratio of 1:4.33. This is
unreliable since it does not include large grey animals that are known to be mature and with
calves nor does it identify the lone females which at the Nastapoka make up 16% (5/31) of
the white adult females as indicated by whales shot by Inuit. The sex ratio from killed white
animals at the Nastapoka is indicated to be unity 32:32 (Arctic Biological Station, unpublis-
hed data). However, this is probably biased towards males since the Inuit select for large
animals unaccompanied by calves both by preference and as a conservation measure.

Philopatry and Reaction to Human Disturbance

A measure of philopatry or site fidelity was derived from individually identified belugas.
In 1983 of a total of 18 individually marked animals, 14 were resighted on 2 or more days and
9 of those were again seen the following year. In 1984 over a period of 59 d, 46 individuals,
mainly white adults or large grey individuals, were recognized individually and resighted
from early July to late August.
From the resighting of identified individuals in 1984, a mean daily presence in the
Nastapoka estuary was calculated to be about 30% of the total number of days in the season
(Table 3). This average frequency of daily occupation is considered a minimal estimate
because of the difficulties inherent in reliably resighting the relatively small number of well-
marked individuals.
The Nastapoka estuary is well known among native communities as a good whale hunting
area, and at the time of this study, it was regularly visited throughout the season by hunters
of Inukjuak and Kuujjuarapik located approximately 200 km north and south respectively
of the study site. The shoreline of the estuary was also used as camping sites for caribou
and waterfowl hunters travelling in the area.
In 1984 during a period of 59 d of observations, the estuary was vacated 7 times for a total
of 170 h because of whale hunting and 6 times for 84 h because of motor boat traffic. The
estuary was vacated 18% of the time in 1984 and to 15% of the total time observed in 1983.
In the usual pattern of hunting disturbance, Inuit motor boats would suddenly arrive at the
beginning of weekends in a direction parallel to the coast at top speed and head into the estuary
in an effort to push the whales as far as possible into the shallow waters. Each boat would
then select its targets.

74
TABLE 3. Occupation rates of the Nastapoka estuary by belugas individually identified using scars and
natural markings during 1984.
Whale ID # Total observation Number of days Percent estuary
period (days) of occupation occupation
a* I 26 10 38
*3 26 14 54
*4 25 7 28
7 25 7 28
8 25 6 24
*10 24 11 46
11 23 4 15
12 24 2 8
*15 23 11 48
*17 24 8 33
19 21 2 10
20 18 8 44
22 19 5 26
*23 19 5 26
26 18 2 11
28 12 5 42
30 19 3 16
31 13 2 15
37 13 3 23
39 11 5 45
41 13 3 23
42 10 4 40
* 44 7 5 71
*45 6 1 17
Whales also identified in 1983.
Mean minimum estuary occupation = 30.46% ± 15.89%.

The majority of whales reacted to approaching motor boats by quickly leaving the estuary.
However, some whales which were further into the estuary (quadrats 1 or 2, Fig. 1) would
continue to swim upstream and would only react overtly when the boats were as close as
500 — 1 000 m away.
When small groups of whales were pursued, some whales would leap out of the water
during their escape, whereas others would lift their heads out for a brief moment before
resuming their flight. Others tried to escape by remaining motionless close to the shores, or
by swimming far upstream almost to the Nastapoka falls.
Motor boats passing approximately 1 km or more outside the estuary only caused the
whales to recede into the downstream portion, whereas motor traffic within quadrats 1 to 4
and each hunt caused the whales to completely desert the area. On a few occasions, however,
the whales were seen to depart on their own when there was no apparent disturbance. Once,
the whales left from the upstream quadrats (1 and 2) at the precise moment when an observer
was walking close to the adjacent sandy north shore of quadrat 2. On one occasion, a bullet
hitting the water in quadrat 2 was apparently the cause of a mass whale exodus from the

75
upstream quadrats. Low altitude air traffic (less than 100 m), such as helicopters, caused
whales to rapidly swim away from the aircraft. Aircraft flying higher than approximately 300
m did not appear to cause any change in swimming direction of the whales.
The effect of disturbance was measured by noting the time required for whales to return to
the estuary. We defined a partial recovery as the time required for the first whales to appear.
Total recovery was arbitrarily defined as the time required for there to be 70% of the pre-
disturbance numbers of whales in the estuary. This was done to allow for seasonal attrition
of numbers in animals that were using the estuary and to adjust for random movements away
to another area. There did not appear to be any significant differences in the mean times it took
for whales to first reappear after hunting disturbance or motor disturbance (U= 25.5, P> 0.10)
(Tables 4 and 5). However, it did on the average take significantly longer (40.1 h) for full
recovery after hunts than after simple disturbance by motor boats (24.6) (U= 13.0, P = 0.015).
The reaction of individually recognized whales appeared to vary greatly in response to
either type of disturbance. Individuals were seen to return in as little as 33 h and stay away
for as long as 574 h. For the 18 resightings of individuals after disturbances of any kind, the
average period of absence was 185 h or 7.7 d.

TABLE 4. Recovery time of whales in the Nastapoka estuary as a result of hunting activities during the
1983 and 1984 seasons. Recovery time: Partial = return of first whale to estuary, Total = return of> 70%
of the pre-disturbance count to estuary.
Date Pie- Partial Total Recovery time of
of disturbance recovery recovery identified whales
hunt count (hours) (hours) # of hours (ID no.)
23/07/83 100 35 48
09/07/84 88 19 50 —
12/07/84 51 36 35 380(8), 117(10,11)
15/07/84 35 2 25 —
15/07/84 25 26 48 —
29/07/84 114 4 45 —
08/08/84 31 48 58 144(3), 56(39)
18/08/84 209 11 12 48(15)

TABLES. Recovery time of whales in the Nastapoka estuary as a result of motor traffic activities during
the 1983 and 1984 seasons. Recovery time: Partial = return of first whale to estuary, Total = return of
> 70% of the pre-disturbance count to estuary.
Date Pre- Partial Total Recovery time of
of disturbance recovery recovery identified whales
disturbance count (hours) (hours) # of hours (ID no.)
25/07/83 70 24 33 —
03/08/83 105 28 28 —
07/07/84 101 28 36 165(1), 372(3,4)
574(7), 106(8)
10/07/84 40 20 22 —
25/07/84 117 16 41 42(l0),236(20)
169(19), 177(26)
20/08/84 129 3 5 123(23)
22/08/84 14 2 5 —
24/08/84 207 15 27 33(1), 104(3)

76
TABLE 6. Contingency table of the age structure variation over increasing periods of time elapsed since
the return of belugas to the estuary after disturbance during the 1984 season (n = 33 027). Cell values
which are significantly different from the expected value are marked with an asterisk. (Chi Square =
161.6074, df = 12, P < 0.005.)
< 30 h 30 - 75 h 75 - 135 h > 135 h
Size class Percent Percent Percent Percent Average
1/3 19.4 19.6 18.7 19.3 19.2
1/2 15.1 14.8 15.8 16.0 15.4
2/3 * 9.8 10.6 * 13.9 11.4 11.4
3/4 * 10.9 12.9 12.6 * 14.2 12.8
White 41.3 42.0 * 39.0 * 39.1 41.2

The size class distribution of the first groups of whales to reoccupy the estuary after a
disturbance was found to be significantly different from the size composition of whales
arriving later (Table 6). Adults accompanied by neonates and yearlings were among the first
whales to reoccupy the estuary after disturbances, while >2/3 and > 3/4 length greys appeared
later. The adults with the smallest immatures were more abundant throughout recovery
periods.

Discussion

Certain estuaries appear to attract belugas each summer. Until recently the function of
these summer areas of aggregation was unknown. Recent findings indicate that belugas use
the warmer waters and abrasive substrates of the estuary during a period of epidermal cell
growth which is physiologically equivalent to a real molt (St. Aubin et al. 1990). Previous
hypotheses defining them as calving (Sergeant 1973) or feeding areas (Seaman and Burns
1981) have not proven correct for Canadian arctic summer aggregations. Females and young
calves probably use the more protected nearshore areas for suckling and nurturing activities
and possibly for shelter from the elements and aquatic predators.
At the Nastapoka our average proportion of neonates was 19.2% which is one of the highest
reported so far for belugas. In other estuaries or coastal concentrations the percentages of
neonates usually range between 8 and 14% (Sergeant 1973; Heyland 1974; Davis and Finley
1979; Breton-Provencher 1980; Finley et al. 1982), and exceptionally it was reported as
18 - 23.8% (Norton and Harwood 1985). Most of those observations came from aerial pho-
tographs in which neonates might have been underestimated because of their position under
the female during suckling. Our higher proportions made from shore observation posts could
be upwardly biased because of the more rapid breathing rhythm of neonates, and greater time
spent at the surface.
Sergeant (1973), for the western Hudson Bay beluga populations, calculated a gross annual
reproduction rate of 0.13 based on reproductive tract analyses, which represents slightly less
than one calf per female every three years. While we have no independent data base with
which to compare our eastern Hudson Bay population, it seems that the high neonate
representation in the Nastapoka is likely an indication of the estuarine habitat preference of
females with newborn rather than a real difference in the reproductive rates. The lower
representation of adult males could also point to a summer segregation of sexes which
possibly utilize different feeding strategies in the inshore and offshore areas (Smith and
Hammill 1986). However, Sergeant (1973), from a consideration of various netted and shot
beluga catches, concluded that the true sex ratio was close to unity and that mortality rates
were comparable for both sexes.

77
 The return in 1984 of 9 of 14 belugas recognized in 1983 is the first positive proof of site
fidelity in this species. All but two of those individuals were white animals and five of
those were females accompanied by neonate calves. Two of the others were lone white
animals of undetermined sex and the two others were > 3/4 length grey individuals accom-
panied by neonates.
The tenacity to site displayed by the Nastapoka herd is the most striking and important
result of this study. Belugas, including individually recognizable whales, reoccupied the
estuary after hunting. This indicates a strong physiological dependance or behavioral fixation
on the particular site which is undaunted by heavy human predation. The vulnerability of the
stock to overharvesting especially because of the strong representation of sexually mature
females is clearly underlined by this pattern of behavior.
Recent analyses using mitochondria' DNA markers of belugas (Helbig et al. 1990) and
humpback whales (Baker et al. 1990) show that there is a geographic segregation of
haplotypes. Both species migrate over long distances and share common wintering grounds
with conspecifics which have spent the summer in other regions. The segregation of
haplotypes is probably a consequence of fidelity of maternal lineages to specific summer
sites, supporting the hypothesis that stocks of belugas can be defined by their summer
estuarine aggregations.
Belugas must be managed with strong consideration given to protection of the possibly
distinct stocks occupying estuaries in the summer. Protection of the estuaries as critical
habitats follows as a lcigical priority in the future conservation of the species. This is
particularly important in relation to management of exploited stocks since, as in the case at
the Nastapoka, a large number of belugas, possibly members of the same putative stock, are
killed there each year.

Acknowledgments

T.G. Smith would like to acknowledge the efforts made by the late Louise Caron. This
would have been her first published paper from a project in which she displayed remarkable
industry and persistence. We would also like to thank Dr. P. Legendre, Université de
Montréal, for help with statistical matters. Dr. B. White, Queen's University, provided us
with information on the results of the genetic analyses of this and adjacent beluga stocks. The
World Wildlife Fund, "Whales Beneath the Ice" Project provided financial support. Mimi
Breton, Pêches et Océans Canada, critically reviewed the manuscript and helped with the
French translation.

References

ALTMANN, J. 1974. Observational study of beha- BRODIE, P. F. 1967. The biology of the beluga,
vior: sampling methods. Behaviour 49: Delphinapterus lencas Pallas of Cumberland
227-267. Sound, Baffin Island. M.Sc. thesis, McGill
S. R. PALUMBI, R. H. LAMBERTSEN,
BAKER, C. S., University, Montreal, Que. 55 p.
M. T. WEINRICH, J. CALAMBOKIDIS, AND S. J. BRODIE, P. F., J. L. PARSONS, AND D. E. SERGEANT.
O'BRIEN. 1990. Influence of seasonal migra- 1981. Present status of the white whale (Del-
tion on geographical distribution of mitocho- phinapterus tenet's) in Cumberland Sound,
drial DNA haplotypes in humpback whales. Baffin Island. Rep. Int. Whaling Comm. 31:
Nature 344: 238-240. 579-582.
BRETON-PROVENCHER, M. 1980. Survey of the DAvis, R.A., AND K.J. FINLEY. 1979. Distribu-
beluga population in the Poste-de-la-Baleine tion, migration, abundance and stock identity
region (New Quebec). Unpubl. Int. Whaling of eastern arctic white whales. Unpubl. Rep.
Comm., Cambridge, (trans. from French by Int. Whaling Comm., Cambridge (IWC Doc.
Dep. Fish. & Oceans.), 56 p. SC/31/SM10), 57 p.

78
FINLEY, K.J., G. W. MILLER, M. ALLARD, R. A. SEAMAN, G. A., AND J. J. BURNS. 1981. Prelimina-
DAVIS, AND C. R. EVANS. 1982. The belugas ry results of recent studies of belukhas in Alas-
(Delphinapterus leucas) of northern Quebec: kan waters. Rep. Int. Whaling Comm. 31:
distribution, abundance, stock identity, catch 567-574.
history and management. Can. Tech. Rep. SERGEANT, D.E. 1973. Biology of white whales
Fish. Aquat. Sci. 1123: 57 p. (Delphinapterus leucas) in western Hudson
FRAKER, M.A. 1980. Status and harvest of the Bay. J. Fish. Res. Board. Can. 30: 1065-1090.
Mackenzie stock of white whales (Delphinap- 1981. On permissible exploitation rates of
tents leucas). Rep. Int. Whaling Comm. 30: Monodontidae. Rep. Int. Whaling Comm. 31:
451-458. 583-587.
FRANCIS, D. 1977. Whaling on the Eastmain. The 1986. Present status of white whales, Del-
Beaver, Summer: 14-19. phinapterus leucas, in the St. Lawrence estua-
GOREN, A. D., P. F. BRODIE, S. SPOTTE, G. C. RAY, ry. Nat. Can. (Rev. Ecol. Syst.) 113: 61-81.
H. W. KAUFMAN, A. J. GWINNETT, J. D. SCUIB- SMITH, T.G., M. O. HAmmin, D. J. BURRAGE,
B A, A ND J. D. BUCK. 1987. Growth layer groups AND G. A. SLENO. 1985. Distribution and abun-
(GLGs) in the teeth of adult belukha whale dance of belugas, Delphinapterus leucas, and
(Delphinapterus leucas) of known age: Evi- narwhals, Monodon monoceros, in the Cana-
dence of two annual layers. Mar. Mammal. dian High Arctic. Can. J. Fish. Aquat. Sci. 42:
Sci. 3: 14-30. 676-684.
HELBIG, R., P. T. BOAC, AND B. N. WHITE. 1990. SMITH, T. G., AND M. O. HAMMILL. 1986. Popula-
Stock identification of beluga whales (Delphi- tion estimates of white whale (Delphinapterus
napterus leucas) using mitochondrial DNA leucas), in James Bay, eastern Hudson Bay,
markers: Preliminary results. Musk Ox 37: and Ungava Bay. Can. J. Fish. Aquat. Sci. 43:
122-128. 1982-1987.
HEYLA ND, J. D. 1974. Aspects of the biology of ST. AUBIN, D. J., T. G. SMITII, AND J. R. GERACI.
beluga (Delphinapterus Incas Pallas) inter- 1990. Seasonal epidermal molt in beluga
preted from vertical aerial photographs, p. whales, Delphinapterus leucas. Can. J. Zool.
373-390. In G.E. Thompson [ed.], Second 68: 359-367.
Canadian Symposium on Remote Sensing, VLADYKOV, V. D. 1944. Etudes sur les mammi-
University of Guelph, Ontario, April 29 - May fères aquatiques. III. Chasse, biologie et va-
1,1974; Proceedings, Vol. 2. leur économique du Marsouin blanc ou Bélu-
LEGENDRE, L., AND P. LEGENDRE. 1984. Écolo- ga (Delphinapterus leucas) du fleuve et du
gie Numérique, Tome 2. Masson, Paris et les golfe Saint-Laurent. Contr. Dépt. Pêch. Qué.
Presses de l'Université du Québec. 235 p. 14: 1-194.
NORTON, P., AND L. A. HARWOOD. 1985. White
whale use of the southeastern Beaufort Sea,
July - September 1984. Can. Tech. Rep. Fish.
Aquat. Sci. 1401:46 p.

79
80
 Cutaneous Response to Implants, Tags, and Marks in
Beluga Whales, Delphinapterus leucas, and
Bottlenose Dolphins, Tursiops truncatus

J. R. Geraci and G. J. D. Smith

Wildlife Section, Department of Pathology, Ontario Veterinary College,
University of Guelph, Guelph, Ontario, Canada NIG 2W]

GERACI, J. R., AND G. J. D. SMITH. 1990. Cutaneous response to implants, tags, and
marks in beluga whales, Delphinapterus leucas, and bottlenose dolphins,
Tursiops truncatus, p. 81-95. In T. G. Smith, D. J. St. Aubin, and J. R.
Geraci [cd.] Advances in research on the beluga whale, Delphinapterus
leucas. Can. Bull. Fish. Aquat. Sci. 224.

To assess the value of metallic and plastic substances for tagging cetaceans, we
examined the cellular response to small test rods implanted through the skin of
beluga whales, Delphinapterus leucas, and bottlenose dolphins, Tursiops trunca-
tus. One material, a porous high density polyethylene, was retained in the dolphins
for the 12-wk duration of the study when used alone. This, and a porous composite,
Proplast I, remained in the belugas for 8 wk. When these materials were coated on
metal rods, the implants were rejected in 3-35 d. All other materials were expelled
in 3-39 d. To gain insight into the cutaneous response to tag configurations, we
implanted various prototypes; all were expelled in 0-44 d. Our findings indicate
the need to develop a composite implant using materials that initiate only a mild
inflammatory reaction, and permit rapid formation of a viable biological seal by
fibrous tissue growth into a porous material. Such implants would increase mecha-
nical stability and provide a barrier to invasive pathogens. We also investigated the
use of dye-marking in beluga whales, and found it to be unsatisfactory.

Afin de juger de la valeur de substances métalliques et plastiques pour l'étique-

tage de cétacés, nous avons examiné les réactions cellulaires aux petites tiges de
contrôle implantées chez les marsouins blancs, Delphinapterus leucas, et chez les
dauphins, Tursiops truncatus. Un matériel, le polyéthylène poreux de haute
densité, est resté en place chez les dauphins pour les douze (12) semaines de la
durée de l'étude, si utilisé seul. Ce dernier et un composite poreux, Proplast 1, ont
été retenus chez les marsouins blancs pour huit (8) semaines. Lorsque ces
matériaux ont été appliqués sur des tiges de métal, les implantations ont été rejetées
dans 3-35 jours. Tous les autres matériaux ont été rejetés dans 3-39 jours. Pour une
meilleure compréhension de la réponse cutanée aux configurations de l'étiquetage,
nous avons implanté des prototypes de designs variés; tous ont été rejetés dans
0-44 jours. Les résultats obtenus pointent au besoin de développer une implanta-
tion composite utilisant des matériaux qui initient seulement une légère réaction
inflammatoire, et permettent une formation rapide d'un scellant biologique viable
par une croissance de tissus fibreux dans un matériel poreux, assurant ainsi une
stabilité mécanique, prévenant le repliement épidermale, et érigeant une barrière
aux pathogènes envahisseurs. Nous avons aussi étudié la réaction des tissus au
marquage par la teinture et au poinçonnage par gel chez les marsouins blancs. Les
deux ont donné des résultats peu satisfaisants, d'une part dûe à l'épaisseur de
l'épiderme et de l'autre à cause d'une cicatrisation indistincte sur le fond blanc.

81
 Introduction

A cetacean that can be individually identified becomes a resource for studies on behavior,
abundance, distribution and movements. Such identification can be achieved through the use
of natural pigment patterns and scars, artificial marks, visible tags, and radio transmitters. A
variety of tags and marking techniques has been tested, as recently reviewed by Scott et al.
(1990). Some tags are implanted in the body wall (spaghetti and arrowhead types) or dorsal
fin (disks and clips) (Evans et al. 1972; Perrin et al. 1979); others encircle the caudal peduncle
(White et al. 1981). With some exceptions (Scott et al. 1990), these devices are not
dependable enough to be used routinely. They are either too difficult to apply, are rejected,
cause inordinate tissue damage, or fail because the materials are not durable or the marks
distinctive (White et al. 1981; Hobbs and Goebel 1982; Irvine et al. 1982).
We undertook this investigation to determine the tissue reactions leading to rejection of
materials embedded through cetacean skin. We had previously characterized the inflamma-
tory response and rate of healing in the skin of belugas, Delphinapterus leucas (Geraci and
Bruce-Allen 1987) and bottlenose dolphins, Tursiops truncatus (Bruce-Allen and Geraci
1985). In this study, we examined the reaction to eight metals and nine plastic materials that
have been used successfully as implants in human surgery. These materials, either metal,
plastic, or plastic-coated metal, were inserted as rods. We also tested three types of
"spaghetti" tags, with dart points embedded in the blubber and streamers fully exposed. In
addition, we examined the cutaneous reaction to dye-marking (tattooing) which, though of
marginal value in other cetaceans (White et al. 1981), might be particularly distinctive on the
white skin of D. leucas.

Materials and Methods

Animals

Ten belugas were kept at a temporary camp on the shore of western Hudson Bay, 7 km north
of the Seal River, Manitoba, in pools 8.2 m in diameter and 1.2 m deep, containing natural
seawater at a temperature of 10-12°C and salinity of 25-30 o/oo. The pools were drained and
refilled with clean water every 1-3 d. The whales were fed thawed Pacific herring (Clupea
harengus pallasi) and fresh cisco (Coregonus artedii), providing 1.7 to 2.1 x 105 J • kg body
wt- ' •
Three female bottlenose dolphins were maintained at the New England Aquarium, Boston,
in a 6.7 x 13.7 x 1.8 m deep pool containing chlorinated (<0.05 ppm free chlorine) seawater
at a temperature of 17-18°C, salinity of 32 o/oo, and a pH of 8.0. The dolphins were fed
thawed herring (Clupea harengus) and capelin (Ma/lotus villosus) providing 1.4 to 2.4 x 105
kg body wt- ' • day- '. During the study period, the animals were in good health as judged J•
by appetite, general appearance, and results of hematologic and blood chemical analyses.
The belugas were immobilized by lowering the water level in the pool until they grounded
on the vinyl pool liner. Sufficient water remained in the pool to keep the whales moist during
the hour it took to apply the tags or marks. For the study on dolphins, each animal was
captured, restrained in a standard dolphin transport stretcher, and placed on a foam pad at
poolside, where it was kept moist by sponging during the 30 min required to complete the
procedures.

Dye-Marking

Dye-marks were applied on beluga whales, using a small hand-held applicator fitted with
six 20-mm-long scored pins spaced 5 mm apart, that were coated with either black, red, or
purple tattoo ink (Spaulding & Rodgers Mfg., Inc., Voorheesville, NY). The marks were

82
applied in one of three ways: on untreated skin; on untreated skin that was then freeze-
branded; or on a site first prepared by freeze-branding to devitalize the epidermis.
The first approach was done simply by pressing the ink-coated pins perpendicularly and
to their full length into skin that had been scrubbed with alcohol and dried. To freeze the site,
either before or after applying the dye, a bronze block with a facial surface of 0.8 x 3.5 cm
was cooled in liquid N, to -196°C, and applied to dry skin for periods of up to 240 s.

Percutaneous Rods

Plastic and metal materials (Table 1), were prepared in the form of rods either 1.5 mm or
3.0 mm in diameter for direct insertion in skin. Other metal rods (1.5 and 2.9 mm in diameter)
of copel, inconel, 316 L stainless steel, Stellite, and titanium, all resistant to corrosion by
seawater (Fig. 1), were coated with various plastics. Rods with a 1-mm-thick coating of
Proplast I were prepared by Vitek, Inc., Houston, TX. A 1-gm-thick layer of ultra low
temperature isotropic carbon was vapor-deposited on rods by CarboMedics, Austin, TX. The
remaining implant rods were prepared in our laboratory: metal rods were inserted in blocks
of HDPE-70 and HDPE-120 which were then machined on a lathe to a thickness of about 1
mm; Gore-Tex, in the form of a vascular prosthesis 3 mm in diameter, was slipped over rods
and fastened with a small amount of Tecoflex 1 MP adhesive; Dacron, in the form of double
velour cardiovascular patch material 1.5 mm thick, was glued with Tecoflex 1-MP adhesive
to rods previously prepared with Tecoflex primer; and medical grade polyurethane was
applied to rods primed with Tecoflex primer, by dip-coating with a 5.5% solution of SG-60D
Tecoflex dissolved in nm-dimethylacetamide.
The uncoated metal rods were cleaned in 100% acetone, stored in 100% ethyl alcohol, and
then rinsed in a sterile 0.9% saline solution. All other implants were placed in a steam
sterilizer for 20 min, then stored dry in sealed containers. Uncoated rods were inserted in the
dorsal fin of the bottlenose dolphins and in the dorsal ridge of the beluga whales. After
scrubbing the site with alcohol, channels spaced 2 cm apart were made I cm from the leading
edge of the dorsal fin of the dolphins and 2 cm from the apex of the dorsal ridge of the whales.
A 14-gauge needle was used to core channels for materials 1.5 mm in diameter, and a
laboratory cork borer was used for materials 3 mm in diameter. One implant was inserted to
the full depth of each channel, flush with both surfaces of the fin or ridge. Control channels,
without implants, were also made in each dorsal fin.
To implant the coated rods, sites 5 cm apart on the dorsal ridge of the whales were scrubbed
with alcohol, then injected with 2% lidocaine hydrochloride, a local anesthetic. Channels
were cut through the ridge with custom-made stainless steel coring tools 3 or 5 mm in
diameter. One rod was inserted through each channel and held in place with a nylon nut on
both ends (Fig. 1).

Tags

"Spaghetti" tags were made in eight different configurations from three basic designs (Fig.
2): flat head/tubing, flat head/wire, and cone head. A flat head/tubing tag had a head stamped
from 316 L stainless steel, 0.64 mm thick, to which was glued a 1.5 cm length of 80A Tecoflex
tubing with an outside diameter of 4.5 mm and a wall thickness of 0.4 mm. A vinyl reward
message was glued inside the tube and the end heat-sealed. Some of these tags had 15-mm-
wide collars of Proplast I, Gore-Tex, or Dacron glued around the base of the tubing where
it joined the tag head. The collars were intended to provide a substrate for tissue-bonding that
would seal the surface. The flat head/wire tag used the same flat head, to which was attached
a 1.5-cm-long strand of 316 L stainless steel wire 0.76 mm in diameter, to anchor a 9-cm-long
piece of 80A Tecoflex tubing containing a reward message. Cone head/tubing tags were
made using cone-shaped heads machined from 316 L stainless steel. Like the flat head/tubing

83
TABLE 1. Materials implanted percutaneously in three bottlenose dolphins and ten beluga whales.
Material Description Supplier
HDPE-70 porous high density Porex Technologies, Fairburn, GA
-120 polyethylene;
-250 pore size: 70, 120, 250 gm

Proplast I & II porous composite of polytetra- Vitek, Inc., Houston, TX

fluoroethylene combined with
carbon (I) or aluminum oxide
(II); pore size: 100-400 gm

Gore-Tex expanded polytetrafluoro- Wm. Blair Medical Specialties, Ltd. Toronto,

ethylene; pore size: 40 gm Ontario

Dacron woven polyethylene terephtha- USCI International, Billerica, MA

late

Isotropic Carbon turbotratic form of vapor de- CarboMedics, Inc., Austin, TX

posited ultra low temperature
carbon

Tecoflex aliphatic, polyether-based Thermedics, Inc., Woburn, MA

linear polyurethane

Urethane polyether urethane BrandoFlex, Ltd., Brantford, Ontario

Aquaplast low temperature plastic WFR/Aquaplast Corp., Ramsey, NJ

Polyform low temperature plastic Rolyan Medical Products, Menomonee

Falls, WI

Copper 99% pure Cu Lewis Tool & Die, Ltd., Guelph, Ontario

Copel copper-nickel alloy Lewis Tool & Die, Ltd.

Inconel 671 nickel-chromium alloy Lewis Tool & Die, Ltd.

Monel 400 nickel-copper alloy Lewis Tool & Die, Ltd.

316 L Stainless Steel Lewis Tool & Die, Ltd.

416 Stainless Steel Lewis Tool & Die, Ltd.

Stellite cobalt-chromium alloy Lewis Tool & Die, Ltd.

Titanium 99.7% pure Ti Johnson Matthey, Inc., Seabrook, NH

tags, these also had 80A Tecoflex tubing containing a reward message attached, and some
had 15-mm-wide collars of Proplast I or Gore-Tex glued to the base of the tubing.
Five of the tags were selected at random and tested for durability. Four were clamped to
a spring balance and subjected to tension stress up to 1 600 g; two of these were submerged
in a 3% solution of NaC1 and tested daily for 5 d. The fifth tag was clamped in a customized
tensiometer, submerged in a 3% NaCI solution, and stressed repeatedly to 300 g at 1.2
cycles • s-1 for 100 000 cycles.

84
Flu. 1. (a) Plastic-coated rods immediately after being implanted in the dorsal ridge of a beluga whale.
(b) Representative rods coated with, from the top: Tecoflex, isotropic carbon, Dacron, Gore-Tex,
HDPE-120/carbon, HDPE-120/Tecoflex, HDPE-250, HDPE-70; bolts were made of nylon. (c) This
monel rod coated with HDPE-120 was removed surgically after 26 d.

In vivo trials were conducted by inserting the tags through the skin on the flanks of the
captive whales, using custom-made tag applicators. Tags were tested in groups of four, and
some were replaced following structural failure or rejection. Each whale received 8 or 12
tags; in all, 64 tags were applied.
One flat head/tubing and one flat head/wire tag were implanted in each of 13 whales that
were captured in shallow water for that purpose, restrained long enough to apply the tag, and
released. This operation to test methods for applying tags was complemented by an in vitro
study to examine the orientation of tags in the integument. Nine tags with either flat or cone
heads, and a commercially manufactured tag (model FH-69A stainless steel dart tag, Floy
Tag and Manufacturing, Inc., Seattle, WA), were inserted through the skin of a freshly killed
beluga whale taken by Inuit from Wakeham Bay, Quebec. The skin and underlying blubber
surrounding the tags was removed and transported frozen to the laboratory for dissection.

85
Flo. 2. The three basic "spaghetti" tag designs: flat head/tubing (top), flat head/wire
(middle), and cone head/tubing (bottom).

Evaluation

The appearance of marks and implant sites was recorded daily, and a photographic record
was maintained. At prescribed intervals, tissue samples were taken for histological evalua-
tion. Each biopsy site was prepared by scrubbing with 95% ethyl alcohol, then injecting into
the dermis a surrounding pattern of 1% lidocaine hydrochloride anesthetic. From dye-marked
and freeze-branded sites on beluga whales, a scalpel was used to remove 1 x 2 cm wedge-
shaped sections, incorporating epidermis and dermis, at the transition zone between the site
and adjacent normal tissue. From implant sites, small sections of channel openings were
excised immediately after an implant was expelled or removed. A sample of tissue adherent
to each plastic or plastic-coated rod was removed with the rod, so that the point of union could
be examined. Biopsy specimens from the dolphins were placed in 5 mL of universal fixative
for 72 h, then dehydrated in a graded series of ethyl alcohol, infiltrated with methacry late for
7 d, and embedded in methacrylate in a vacuum desiccator. Sections 2-5 gm thick were
stained with hematoxylin and eosin. Sub-samples from the dolphins and samples from the
whales were placed in 10% phosphate-buffered formalin and stored for several weeks. The
fixed tissues were then processed through alcohol and xylene and embedded in paraffin
blocks. Sections 10-12 1.1.M thick were stained with hematoxylin and eosin.

Results
Animals

No difficulty was encountered during the application of any of the cryogenic or dye marks,
or implantation of the materials and tags. All the animals quickly resumed normal behaviour
patterns following the procedures.

86
Dye-Marking

Five sites on beluga whales were dye-marked, seven sites were dye-marked then frozen,
and 10 other sites were frozen, allowed to thaw for 10 min, then dye-marked. The results were
similar in all cases. All sites were mildly elevated and pale 24-48 h after the procedure. The
ink was visible in the puncture marks immediately after application, but effectively disap-
peared by 24 h.
Freeze-branding, whether applied before or after the dye, resulted in shallow depressions
with normal coloration for 3 or 4 wk. Histologically, coagulation necrosis was evident in the
cells within the entire epidermal column, from the tips of the dermal papillae to the stratum.
externum. By the second week, surface layers of epidermis were sloughing, resulting in a
shallow (2-4 mm) erosion of the site. A new s. externum was forming at the junction of this
and the healthy lower layer of s. spinosum, and the s. germinativum was restored. By the third
and fourth weeks, new epidermal tissue bridged the sites, and within 7 wk, the healing process
was nearly complete.

Percutaneous Rods

Bleeding caused by the implant procedures stopped within 30 min in the dolphins, and
24-48 h in the whales. In the dolphins, the skin surrounding each implant developed a pale
gray halo up to 12 mm in diameter, which persisted for 12 wk. In the belugas, the skin
immediately surrounding the implant sites was elevated after 24 h and remained so for 1 or
2 wk.
All uncoated metal rods became loose within the first few days after being implanted. All
were extruded within 10 d in the whales, and 21 d in the dolphins (Table 2).
The plastic rods fared better. One material, HDPE-120, remained in the dolphins and the
whales for the duration of the study (Table 2). Proplast I also remained in the whales
throughout the study, but an analogue, Proplast II, was expelled from the dolphins within 7 d.
Tissue response to HDPE-120 and Proplast I was similar for both materials. However, the
sequence of events in the whales was delayed by about 2 wk, the inflammatory response was
more intense and persistent, and bacteria were present in the implant sites. Polyform
remained in the whales for the 39-d duration of the study, but histological examination
revealed bacteria, a chronic inflammatory cell infiltrate composed mainly of macrophages,
and mature s. externum surrounding the implant. These were signs that the material would
soon be expelled.
Most of the coated rods became loose within 3 to 5 d, and were expelled from 3 to 35 d after
implantation (Table 2). One of the rods was removed surgically after 26 d (Fig.lb), just before
the whale was released.
The following is a composite description of the inflammatory response to the implants,
based on serial biopsy samples from whales and dolphins. Histologically, each metal-implant
site showed an acute inflammatory response within the first week. Openings left by expelled
implants were filled with exfoliated s. externum cells, neutrophils, bacteria, and erythrocytes
within a fibrin matrix. Where implants were still present, hemorrhage and a neutrophilic
infiltrate filled the space between the s. intermedium and the implant. Cells of the s.
germinativum within 1 mm of the implants were mitotically active. The dermis contained a
marked inflammatory cell infiltrate composed mainly of neutrophils. By the end of the third
week, the tissue response had become chronic. The s. externum in the implant sites was
thickened due to proliferation and hyperplasia of the epidermal cells. The mild inflammatory
cell infiltrate in the dermis was composed of neutrophils and a few macrophages. By the end
of the sixth week, a slightly hyperplastic s. externum and some maturing scar tissue were the
only evidence that an implant had been present.

87
TABLE 2. Summary of retention of materials implanted percutaneously.
Material Maximum retention time (days)
Dolphins Whales
Plastic rods:

HDPE-120 84" 57"

HDPE-250 — 35
Proplast I — 58"
Proplast II <7 —
Urethane <7 <20
Aquaplast — 14
Polyform — 39a

Metal rods:

Copper <21
Copel <21
Inconel 671 <7 <10
Monel 400 <21 3
416 Stainless Steel <21 6
Titanium <7 <3

Coated rods:

HDPE-70 — 20
HDPE-120 — 26"
Proplast I — 27'
Proplast I/Isotropic Carbon — 35
Proplast I/Tecoflex — 35
Dacron — 31"
Gore-Tex — 25
Isotropic Carbon — 10'
Tecoflex — 28
" Implant removed surgically.

The histological reaction to plastic and plastic-coated metal was, for the most part, less
severe. After 1 wk, the s. externum was hyperplastic and edematous and, in the whales,
contained bacteria within microabscesses. The pores of the implant material contained
bacteria, exfoliated cells, and inflammatory exudate. Neutrophils were scattered throughout
the s. spinosum. Cells of the s. germinativum were mitotically active and migrating down the
surface of the implant towards the dermis. The inflammatory response in the den-nis was
confined within a radius of 7 to 10 dermal papillae from the implant, and was characterized
by congestion and an infiltrate of neutrophils. Early fibroplasia was seen in the immediate
area of the implant.
At the end of 3 wk, the s. externum adjacent to the implant was thickened due to hypertrophy
and hyperplasia (Fig. 3). S. spinosum cells in contact with the implant had begun differen-
tiating into s. externum-like cells, and coated the surface of the implant in a layer 10 to 20 cells
thick. Mitotic figures were common within the s. germinativum. Pores of the plastic within
the epidermis remained filled with exfoliated cells and bacteria. In the dermis, the implant
was surrounded by granulation tissue and an inflammatory cell infiltrate composed of
neutrophils and a few macrophages.
By 6 wk, the cells of the s. extemum adjacent to the implant were mature, while those of
the s. spinosum remained undifferentiated. Cells of the s. germinativum continued to divide
actively. The dermis contained a bed of maturing, well-vascularized scar tissue. Exfoliated

88
epidermal cells and fibroblasts were present in the peripheral pores of the implant, while
deeper pores contained macrophages, giant cells, fibrocytes, and a few budding capillaries.
Over the next 6 wk, the reaction to the implant approached a steady state (Fig. 4). Cells of
the s. externum and upper two-thirds of the s. spinosum were mature where they contacted

FIG. 3. (a) An HDPE-120 implant in dolphin skin in situ after 3 wk illustrating the histological
appearance of the implant and the response of the surrounding tissue (H&E). (b) Epidermal growth had
advanced down to the level of the bed of granulation tissue (H&E). (c) At the level of the dermis, the
pores of this implant contained newly forming fibrous tissue with few neutrophils and macrophages
(H&E).

89
the implant; the lower third of the spinous cells were not fully differentiated. Mitotic figures
in the s. germinativum became less common, and there was more maturing fibrous tissue
within the implant.

FIG. 4. (a) A completely buried HDPE implant in dolphin skin in situ after 12 wk (H&E). (b) Epidermal
cells were bridging the opening (H&E). (c) Some pores of the implant contained an inflammatory
infiltrate (H&E) composed of (d) neutrophils, tissue debris and red blood cells (H&E).

90
Tags

In laboratory tests to examine the durability of the tags, the adhesive on three of the five
tags failed at 800 to 1 000 g of stress, allowing the tubing to separate from the head. Tecoflex
1-PP adhesive did not provide a good bond, whereas Tecoflex 1-MP did when used as
prescribed by the manufacturer. This required cleaning the tag head, applying multiple
coatings of adhesive, and final activation of the adhesive by wiping both surfaces with methyl
ethyl ketone prior to bonding. The Tecoflex tubing did not break in any of the tags tested,
although a stress of 1 600 g stretched the tubing from 138 to 265 mm, an elongation of 192%.
During the in vivo trials, we found that the flat head/tubing and flat head/wire tags inserted
easily, whereas the cone head/tubing tags required that the dermis first be punctured with the
tip of a scalpel blade. In a few cases, bleeding occurred after tag insertion and appeared to stop
within 24h, but recurred intermittently for 4-7 d with cone head/tubing or flat head/wire tags.
Tags fabricated with Tecoflex 1-PP adhesive were rejected or suffered adhesive failure 0 to
11 d (average 5 d with flat head/tubing tags) and 5 to 41 d (average 20 d) (cone head/tubing
tags) after insertion. Tags fabricated with 1-MP adhesive remained intact and were rejected
3-10 d (average 5 d with flat head/tubing tags) and 3-44 d (average 20 d with cone head/
tubing tags) after insertion. Flat head/wire tags were extruded 4-13 d after implantation.
Ten of the cone head/tubing tags were monitored for movement within the integument by
measuring the length of tubing exposed immediately after insertion, and at 2-12 d intervals
thereafter. Initially, the cone head tags without a collar of porous material moved about 5 mm
deeper into the body during the first 5-10 d, and then were extruded at a rate of up to 2 mm •
d- '. Cone head tags with a collar of Proplast I also moved in and out as much as 2.3 mm •
d- '. The most stable cone head tags were those with a collar of Gore-Tex, with net extrusion
rates of 0.17 to 0.54- mm • d - , but even one of these tags moved outward at a rate of
4- mm • (I-1 .
Field trials on live whales showed that flat head/wire tags were easy to apply and that the
applicator could be removed from the animal without dislodging the tag. Flat head/tubing
tags with Dacron collars were more difficult to apply because of problems keeping the tag
loaded on the applicator, and, once applied, in removing the applicator from the whale
without dislodging the tag. Following application, all tags were tested with a gentle (300 g)
pull; none came out (Fig. 5). However, some could be pushed deeper into the body. There was
little bleeding and no visible reaction from the whales. We were able to observe the tags in
two of the whales after release. Hydrodynamic drag folded the Tecoflex tubing of the flat
head/tubing tags back against the skin, and caused the stiffer flat head/wire tags to vibrate.
A whale carcass was used to evaluate mechanical damage to tissue associated with the
implant procedure. The anchoring flange on the flat head tags was compressed during
insertion, reducing the effectiveness of the anchors. The tops of the anchor flange of the flat
head/tubing tags and the flat head/wire tags were situated about 5 and 25 mm, respectively,
below the non-papillary dermis (16 mm from the surface) (Fig. 5). One prong of the spring
wire of the cone head tags was anchored against the base of the dermis (Fig. 5). The epidermis
was pressed against the tubing, and there was some tissue destruction at the level of the
dermis, so that very little tissue was in contact with the band of porous material. The single
Floy tag tested was anchored in the blubber about 4 cm below the non-papillary dermis. The
head was designed to flip over and lie parallel to the surface after implantation, although in
the tag examined, the head was resting at a 45° angle.

Discussion

We were unable to produce a visible dye-mark in beluga whales using procedures that can
reasonably be carried out in the field. Beluga epidermis is so thick that it hid the tattoo
pigments from view; pre-treatment by freezing did not have the desired effect of replacing

91
Fro. 5. Top: Tags 4 d after insertion through the skin of the flank of a beluga whale. The holes slowly
widened around the streamers. No tag remained beyond 44 d. Lower left: Flat head tag from in vitro
study. The anchor flange (F) was loose in the blubber and not anchored against the dermis (D): E
epidermis; B blubber. Lower right: Cone head tag from in vitro study. The spring wire barbs were
positioned at (S) the base of the derrnis and appeared to be insecure anchors.

the epidermis with a translucent scar, as it does on odontocetes with a thin epidermis (White
et al. 1981; Irvine et al. 1982). We conclude that dye-marking is not a practical technique to
identify beluga whales.
Implanted rods and tags, whatever the material or configuration, and whether on whales or
dolphins, lasted only long enough to be of value as a marking tool for short-term monitoring.
One material, porous polyethylene HDPE-120, was retained in the dolphins for the 12-wk
study. This polymer, and the porous composite Proplast I, remained in the belugas for 8 wk.

92
However, all metal rods coated with the same materials were rejected within 35 d. Metals
alone lasted for even shorter periods. The prototype tags used in the pilot study of free-ranging
whales were rejected in 0-44 d in the captive animals.
We anticipated that some of the materials would be rejected. Corroding copper, for
example, releases products that impede healing (Williams 1981a). The urethane was of
industrial grade and may have released plasticizers and other toxic substances (Peppas 1982).
All of the metallic implants were lost, less because of their physical properties than the
location in which they were placed. Titanium resists corrosion, as do Inconel, monel, and
copel (American Society for Metals 1980), and is used in surgical prostheses that are deeply
implanted with no surface exposed (Williams 1981b). In this study, all were rejected within
3 wk, partly because the skin reacts naturally to sequester and expel foreign objects (Daly and
Dasse 1983), and partly because the surrounding tissue was weakened by infection with
bacteria that were carried into the wound with the implant (Dasse 1984) or deposited there
from water bathing the opening. Infection following the 416 stainless steel implant evolved
into an abscess. Thus, the corrosion-resistant property of a metal seems to be less important
to the permanence of a percutaneous implant in cetaceans than the opening it leaves behind
as a gateway for infectious organisms from the sea.
For this reason, we tested porous plastics used in surgical implants, hoping that a coating
of such material would effectively serve as a gasket to reduce the wound opening, yet
recognizing that bacteria can gain entrance through the same materials (Hall et al. 1975). We
also thought that the pores would serve their intended role to attract fibroblasts from the
dermis and blubber, eventually bonding the tissue to the implant (von Recum 1984). The ex-
periments were marginally successful. Some of the sites became infected; in others, the
plastics such as Proplast II became compressed when introduced and lost their vital porous
characteristics. However, the results with HDPE-120 and Proplast I showed the sequence of
tissue reactions one might ideally expect when these substances are implanted in odontocete
skin.
The HDPE-120 and Proplast I remained in place for 8 wk until the whales were released,
and the HDPE-120 implants were still present in the dolphins after 12 wk. By the end of the
first week in the dolphins and third week in the whales, fibroblasts began crowding the pores
of the implants; these cells were part of a mild inflammatory response. Two weeks later,
fibrous tissue was growing into the pores of the implants, and after 5 additional weeks, the
fibrous tissue was maturing. Our study ended at this point, so we were unable to determine
whether scar tissue ultimately developed, marking the final stage of maturation and the
formation of potentially sound anchoring tissue. But the findings were encouraging.
Porous plastics lack the tensile strength required of a tag. That is why we coated them onto
metal rods for another series of experiments. The results were discouraging; infection ensued,
and the mechanical forces on the rods, great enough to cause bending in some cases, collapsed
the pores. All coated rods were rejected within 5 wk.
To determine how a small tag with an exposed streamer might endure, we tested some
prototypes in the whales. All of the tags failed for a number of reasons. The bare metal was
as poor an anchor as might have been predicted from the reactions to the same materials used
as embedded rods. Furthermore, the flange intended to secure the flat head tags bent during
insertion through the tough dermis. The cone head tags were more effective, but only
extended the retention time by about one week; the tag heads continually shifted in the soft
blubber and underlying moving muscle, thereby preventing encapsulation of the head by
fibrous tissue. Tag heads with a collar of Gore-Tex were more stable, perhaps due to friction
between the collar and surrounding tissues. Bacterial infection inevitably played a role in
rejection of these tags, as well. The wire attaching the Tecoflex polyurethane tubing
streamers (which endured well) kept the wounds open, thereby promoting infection.
Conditions in the wild will undoubtedly influence the biological processes involved with
retention of implants, but it is not likely that the tags, as constructed, would be useful for long-

93
term monitoring studies. However, for tags that need not be permanent, configurations using
HDPE-120 or Proplast I may have the edge over other materials.
From these observations, we suggest the following measures to improve the effectiveness
of percutaneously implanted tags and other devices. Minimize the diameter of the head, hence
the size of the hole for access by bacteria. Embed the head deeply enough and with enough
anchors to minimize movement within the tissue. Avoid the use of sharp edges, including
wire, that can cut through tissue. Include in the design of the head a series of circular channels
for inlaying porous plastic "0" rings (we would choose HPDE-120 or Proplast I) which, when
in place, would be flush with the surface of the metal and therefore less apt to be crushed
between the anchor and the tissue. Devise a collar that provides a tight seal at the surface of
the skin to prevent bacteria from reaching the porous component. Include a slow-release,
broad spectrum antibiotic in the implant, and take measures to prevent contamination while
embedding the device.
These suggestions are intended to minimize infection and movement in tissue, two
principal forces that reduce the life of an implant. The permanence of bolts and other
percutaneous devices that anchor electronic packages might also be improved by applicable
changes in design (Grosse-Siestrup and Affeld 1984). None of these modifications will with
certainty prevent the channel in the skin from delivering the organisms that usually
undermine the attachment. That problem should be addressed in the next generation of
studies.

Acknowledgments

The studies on captive dolphins were made possible through the cooperation of J. Prescott
(Director), G. Early, J. Pearson, P. Fiorelli, J. Boggs and the "Discovery" crew at the New
England Aquarium, Boston, MA. Funding for this portion of the study came from the U.S.
Department of the Interior, Minerals Management Service contract no. 14-12-0001-29169.
We gratefully acknowledge the generous logistic and laboratory support provided by the
Department of Fisheries and Oceans (DFO), Winnipeg, Manitoba. We thank R. Moshenko,
DFO, for facilitating the field operations, and J.Orr, DFO, for supervising the camp set-up.
D. Jacobs, Ministry of Natural Resources, provided helpful advice and assistance in
Churchill. J. Hickes and his capture crew demonstrated considerable skill and patience while
selecting whales that met our requirements. For their capable assistance in the field and
laboratory, we thank K. Baker, J. Batstone, G. Beck, K. Brown, T. Friesen, L. Joe, L. Latta,
C. Thomson, and S. Waters. Drs. B. Hicks and N. Palmer helped with interpretation of the
histological findings. We are grateful to the following companies for advice, free samples,
and goods and services supplied at cost: CarboMedics, Inc., Porex Technologies, Rolyan
Medical Products, Thermedics, Inc., USCI International, Vitek, Inc., WFR/Aquaplast Corp.,
and Wm. Blair Medical Specialties, Ltd. The study was supported in part by a "Whales
Beneath the Ice" grant from the World Wildlife Fund Canada.

References

AMERICAN SOCIETY FOR METALS. 1980. Metals DALY, B. D .T., AND K. A. DASSE. 1983. Develop-
handbook. Ninth ed., Vol. 3. Properties and se- ment and in vivo testing of a chronic percuta-
lection: stainless steels, tool materials, and neous prosthesis. Artif. Organs 7(4): 454-458.
special purpose metals. American Society for DASSE, K. A. 1984. Infection of percutaneous
Metals, Metals Park, OH. 882 p. devices: prevention, monitoring and treatment.
BRUCE-ALLEN, L. J., AND J. R. GERACI. 1985. J. Biomed. Mater. Res. 18: 403-411.
Wound healing in the bottlenose dolphin EVANS, W. E., J. D. HALL, A. B. IRVINE, AND J. S.
(Tursiops truncatus). Can. J. Fish. Aquat. Sci. LEATHERWOOD. 1972. Methods for tagging
42: 216-228. small cetaceans. Fish. Bull. (U.S.) 70(1): 61-65.

94
GERACI, J. R., AND L. J. BRUCE-ALLEN. 1987. cated by results of tagging, with summary of
Slow process of wound repair in beluga whales, tagging operations, 1969-76. U.S. Dep. of
Delphinapterus leucas. Can. J. Fish. Aquat. Commerce, NOAA Tech. Rep. NMFS SSRF-
Sci. 44: 1661-1665. 737: 14 p.
GROSSE-SIESTRUP, C., AND K. AFFELD. 1984. SCOTT, M. D., R. S. WELLS, A. B. IRVINE AND
Design criteria for percutaneous devices. J. B. R. MATE. 1990. Tagging and marking stu-
Biomed. Mater. Res. 18: 357-382. dies on small cetaceans, p.489-514. In S. Lea-
HALL, C., L. ADAMS, AND J. CumoNi. 1975. De- therwood and R. R. Reeves [ed.] The bottle-
velopment of skin interfacing canula. Tran. nose dolphin. Academic Press, Inc., San Die-
Am. Soc. Artif. Inter. Organs 21: 281-284. go., CA.
HOBBS, L. J., AND M. J. GOEBEL. 1982. Bowhead VON RECUM, A. F. 1984. Applications and failure
whale radio tagging feasibility study and re- modes of percutaneous devices: a review. J.
view of large cetacean tracking. National Ocea- Biomecl. Mater. Res. 18: 323-336.
nographic and Atmospheric Administration WHITE, M. J., J. G. JENNINGS, W. F. GANDY, AND
Technical Memorandum NMFS-F/NWC-21: L. H. CORNELL. 1981. An evaluation of tag-
iii+68 p. ging, marking, and tattooing techniques for
IRVINE, A. B., R. S. WELLS, AND M. D. SCOTT. small delphinids. NOAA-TM-NMFS-SWFC-
1982. An evaluation of techniques for tagging 16: iv+142 p.
small odontocete cetaceans. Fish. Bull. (U.S.) WILLIAMS, D.F. 1981a. Electrochemical aspects
80(1): 135-143. of corrosion in the physiological environment,
PEPPAS, N. A. 1982. Structure, testing, and appli- p. 11-42. In D.F. Williams [ed.] Fundamental
cations of biomaterials, p. 465-473. In S. L. aspects of biocompatibility, Vol. 1. CRC Se-
Cooper and N. A. Peppas, [ed.] Bioma- ries in Biocompatibility, CRC Press Inc., Boca
terials:interfacial phenomena and applications. Raton, FL.
Advances in Chemistry Series, No. 199. 1981b. Titanium and titanium alloys,
American Chemical Society, Washington, DC. p. 9-94. In D.F. Williams [ed.] Biocompatibi-
PERRIN, W. F., W. E. EVANS, AND D. B. HOLTS, lity of clinical implant materials, Vol. 1. CRC
1979. Movements of pelagic dolphins (Stenel- Series in Biocompatibility, CRC Press Inc.,
la spp.) in the eastern tropical Pacific as indi- Boca Raton, FL.

95
96
 Reactions of Belugas, Delphinapterus leucas, and
Narwhals, Monodon monoceros, to Ice-Breaking Ships
in the Canadian High Arctic

K. J. Finley', G. W. Miller, R. A. Davis

LGL Limited, P.O. Box 280, King City, Ontario, Canada LOG I KO

and C. R. Greene
Greeneridge Sciences Inc., 4512 Via Huerto, Santa Barbara, CA 93110, USA

FINLEY, K. J., G. W. MILLER, R. A. DAVIS, AND C. R. GREENE. 1990. Reactions

of bel ugas, Delphinapterus leucas, and narwhals, Monodon monoceros, to
ice-breaking ships in the Canadian High Arctic, p. 97-117. hi T. G. Smith,
D. J. St. Aubin, and J. R. Geraci [cd.] Advances in research on the beluga
whale, Delphinapterus leucas. Can. Bull. Fish. Aquat. Sci. 224.

The responses of belugas and narwhals to ice-breaking ships in the Canadian

high Arctic were studied over a 3—yr period. Belugas and narwhals exhibited very
different behavioral responses to ship approaches and ice-breaking activity.
Typically, belugas moved rapidly along ice edges away from approaching ships
whereas narwhals showed no overt panic reaction. Presumed alarm vocalizations
of belugas indicated that they were aware of an approaching ship over 80 km away
and they showed strong avoidance reactions to ships approaching at distances of
35 — 50 km when received noise levels ranged from 94 to 105 dB re 11.1Pa in the
20 — 1 000 Hz band. The "flee" response of the beluga involved large herds
undertaking long dives close to or beneath the ice edge; pod integrity broke down
and diving appeared asynchronous. Belugas were displaced along ice edges by as
much as 80 km; the degree of di sp lacemen t depended on the proximity of ice edges
to coastal areas. Narwhals showed subtle responses to approaching ships; they did
not form large herds, their movements were slow or they remained motionless near
the ice edge, and they huddled together in pods, often engaging in physical contact.
Rather than actually diving, they tended to sink during submergence and dispersed
slowly along the ice edge in front of advancing ships. Ship approaches caused
narwhals to cease vocalizing temporarily. Narwhals returned to disturbance areas
much faster than belugas and resumed normal activities when received noise levels
from ice-breaking operations were as high as 120 dB. The responses of both species
at unprecedented ranges may be explained in part by the fact that no similar field
studies have been conducted in pristine marine environments with industrially-
naive populations of marine mammals.

Les réactions des marsouins blancs et des narvals aux navires brise-glaces dans
le grand nord de l'Arctique canadien ont été contrôlées sur une période de 3 ans.
Les marsouins blancs et les narvals ont exhibé des réaction très différentes de com-
portement à l'approche des navires et des activités de brise-glaces. Typiquement,
les marsouins blancs se sont déplacés rapidement le long du bord des glaces en
s'éloignant des navires qui s'approchaient, tandis que les narvals n'ont manifes-
té aucune réaction évidente de panique. Des vocalisations d'un signal présumé

' K. J. Finley Ecological Research, 10232 Summerset Place, Sidney, B.C. , Canada V8L 4X2.

97
 d'alarme des marsouins blancs ont démontré qu'ils étaient conscients de l'appro-
che d'un navire à plus de 80 km de distance et laissèrent voir de fortes réactions
d'évasion aux navires approchant à des distances de 35-50 km quand l'enregistre-
ment des niveaux de bruit était de 94 à 105 dB rel liPa (20-1 000 Hz). La réaction
d'évasion des marsouins blancs impliqua de gros troupeaux s'engageant dans de
longues plongées près ou dessous le bord des glaces; l'intégrité du troupeau se
rompit et la plongée semblait asynchrone. Les marsouins blancs étaient dispersés
le long du bord des glaces allant jusqu'à 80 km des navires; le degré du déplacement
était rattaché à la proximité du bord des glaces aux régions côtières. Les narvals ont
exhibé des réactions subtiles au rapprochement des navires; ils ne se sont pas
regroupés en gros troupeaux; leurs mouvements étaient lents ou ils restèrent
immobilisés près du bord des glaces; ils se serraient les uns contre les autres en
bandes, résultant à des contacts physiques fréquents. Plutôt que de plonger
directement, ils avaient tendance à se laisser câler et se dispersaient tranquillement
le long du bord des glaces en avant des navires qui s'approchaient. L'approche de
navires provoquait un arrêt temporaire de vocalisation chez les narvals. Les narvals
sont retournés plus vite dans les zones de perturbation que les marsouins blancs et
ont résumé les activités normales quand I 'enregistrement des niveaux de bruits de
brise-glaces étaient aussi hauts que 120 dB. Cette réaction sans précédent des deux
espèces peut, en partie, être expliquée puisqu'aucune étude similaire en chantier
n'a été effectuée dans des environnements marins avec des populations de
mammifères marins industriellement-naïves.

Introduction

Ever since the controversial cruise of the ice-breaking tanker Manhattan through the
Northwest Passage in 1969, there has been concern regarding the effects of increased marine .
wildlife of the region. In the past two decades, there have been traficonhepld
proposals to ship hydrocarbons in large ice-breaking carriers operating year-round in the
passage. The most significant new development in the region is the expansion of the
traditional summer shipping season by using reinforced carriers to serve mine sites.
One of the major environmental concerns associated with increased vessel traffic in the
arctic is the potential for deleterious effects on marine mammal populations. The effects of
excessive underwater noise levels is of particular concern since the acoustical sense of marine
mammals is of vital importance in their navigation, communication and feeding. Lancaster
Sound, the entrance of the Northwest Passage, is a major migration corridor for several
species of marine mammals, including belugas, Delphinapterus leucas, narwhals, Manado,'
monoceros, and bowhead whales, Balaena mysticetus.
In 1982, Canarctic Shipping Limited announced plans to send the ice-breaking ore carrier
MV Arctic through Lancaster Sound to the Nanisivik lead—zinc mine situated in Admiralty
Inlet, northern Baffin Island. The purpose of this exploratory voyage was to examine the
feasibility of increasing the length of the commercial shipping season to the mine by breaking
through the fast ice of Admiralty Inlet. The voyage raised concerns among local Inuit hunters
who hunt narwhals in spring along the ice edge across the mouth of Admiralty Inlet. The
hunters were worried that ship traffic would cause displacement of the narwhals and interfere
with their hunt. In response to the concerns of the hunters and to broader concerns about arctic
shipping, the present study was initiated in June 1982 to obtain preliminary data on the
responses of narwhals and belugas to the Arctic. The study was continued in June of 1983 and
1984 to take advantage of additional probes through Lancaster Sound and Admiralty Inlet by
the Arctic and accompanying ice-breakers of the Canadian Coast Guard. This paper is a
synthesis of the main findings of the study, the results of which are presented in more detail
in a technical report by LGL (1986).

98
 Methods and Materials

The spring distribution of whales in Lancaster Sound is strongly influenced by the extent
of fast ice. In "heavy" ice years such as occurred during this study, the fast ice may cover most
of the sound with a solid ice edge extending across its entrance. Since the position of the ice
edge changed from year to year, the location of study area and procedures also varied.
Our approach was to document the undisturbed behavior and distribution of the whales
along ice edges before the arrival of ships in Lancaster Sound. Whale behavior and distribu-
tion were monitored during the ships' arrival, during periods when there was ship activity in
the area, and during a post-disturbance period. Information was collected by aerial surveys,
ice-based observations and underwater recordings.

Study Areas and Protocol

The positions of the ice edge in each year are shown in Fig. 1. Changing circumstances —
weather, ice and ship — often necessitated extemporization of the planned study design. The
setting and study protocols are reviewed briefly here.
1
86° 1 84 ° \ 1 82 0

DEVON / I ISLAND

1983

74 ; 1

CAPE
CHARLES
YORKE

ADMIRALTY
BY LOT
INLET
NAVY ISLAND
BOARD
INLET

20 0 20 40 60
RCTIC'è
BAY km
(NANISIVIK 73 ; 1

184 ° 1 82° 80°

Fia. 1. Locations of the major ice edges in Lancaster Sound during the study periods in 1982-84.
Principal ice-based study areas are indicated by arrows.

99
 1982: The ice edge across Lancaster Sound was located west of Admiralty Inlet. Fast ice
covered the inlet, creating another ice edge across the mouth of the inlet. The Arctic arrived
at the Admiralty Inlet ice edge on 28 June. The ice-breaker CCG John A. MacDonald
remained offshore in Lancaster Sound because the Arctic encountered little difficulty in
breaking through the fast ice.
The 1982 study was a prototype project organized at short notice; we had only one day to
study the pre-disturbance distribution and behavior of the whales. Continuous observations
of whale behavior and regular underwater recordings were made from camps along the ice
edge during the ship's approach and were continued until after the ship reached Nanisivik
mine on 30 June. Helicopter surveys of the ice edge were conducted 22 h before, and 28 and
52 h after, the Arctic contacted the ice edge. Survey procedures are described by LGL (1986).
For safety reasons, the helicopter could not be flown over open water, and because it was not
equipped with an accurate navigation system, it was not possible to determine the positions
of whales along the ice edge accurately.
1983: The ice edge was located near the eastern entrance of Lancaster Sound and ice
conditions were severe. The Arctic arrived at the ice edge on 24 June but had great difficulty
penetrating the fast ice. The CCG John A. MacDonald arrived 2 d later and broke ice for the
Arctic into Admiralty Inlet. The ships did not reach Nanisivik until 9 July.
Information on pie-disturbance whale distribution and behavior was obtained on 5 d prior
to the arrival of the Arctic. Because the ice was so thick, the ships made very slow progress
through Lancaster Sound. Thus, levels of underwater noise remained high at the point of entry
into the ice edge for the week following the Arctic's arrival. This, coupled with the fact that
both ships ceased operations every night, created a series of disturbance periods and quiet
intervals during which the behavior and distribution of the whales were monitored. Attempts
to monitor post-disturbance behavior and distribution were complicated by the development
of major cracks in the Lancaster Sound ice sheet and the beginning of a whale hunt by local
Inuit.
Eighteen helicopter surveys of the Lancaster Sound ice edge were flown in 1983. The
surveys covered the entire width of the sound plus a small section of the adjacent coast. Some
of these surveys were incomplete due to local weather conditions. A snow storm prevented
a survey on the day that the Arctic arrived.
1984: The Lancaster Sound ice edge was again located east of Admiralty Inlet about 60 km
west of the 1983 position. The Arctic arrived at the ice edge on 28 June accompanied by the
ice-breaker CCG Louis St. Laurent. The ice-breaker accompanied the ore carrier into the ice
but the Arctic was able to break its own track into Nanisivik mine.
The 1984 study consisted of aerial surveys by fixed-wing aircraft equipped with a GNS-
500 navigation system to provide better coverage of the distributional responses of belugas
and narwhals to ships. Seven surveys were flown: the first was on the day before the ships
arrived; four surveys were flown when levels of ship activity were high at the ice edge; and
the final two surveys were flown after the ships had moved west into the ice but were probably
still audible to whales at the ice edge. The survey procedures are described by LGL (1986).

Ice-Based Observations and Underwater Recordings

Observations of whale behavior and numbers were conducted from ice-edge camps in 1982
and 1983. We recorded species and numbers of marine mammals, their general activity,
direction of movement, distance from ice edge, and sex and relative age.
Hydrophones were deployed at the ice-edge camps to record ambient noise levels in the
water, animal calls, and ship noise. Ten-minute underwater recordings were obtained every
2 h when ships were absent and usually continuously when the ships were operating nearby.
A total of 4 h of underwater recordings was obtained in 1982 and 45 h in 1983.

100
 Hydrophones were deployed at two depths, 3 and 20 m, so that any depth-related
differences in the received sounds could be measured. We used U.S. Navy model H56
standard reference hydrophones, which are wideband (to 50 kHz), low-noise hydrophones
with built-in preamplifiers. We also used low-noise bender hydrophones sensitive at
frequencies below 2 kHz. Each hydrophone had a post-amplifier with selectable gain (in
10 dB steps) up to 60 dB. Two types of tape recorders were used. The simplest was a Sony
model TC - D5M two-channel cassette recorder with a calibrated response of 10 to
16 000 Hz. Most recordings were made with a four-channel Fostex model 250 cassette
recorder with a frequency response of 10 Hz to 32 000 Hz.

Analyses of Ambient and Ship Sounds

Analysis of ambient and ship sounds was based on the use of a Hewlett-Packard model 9816
computer with an analog-to-digital converter. Several programs were used to analyze the
data. Power spectral density estimation was based on computing discrete Fourier transforms.
For convenience and consistency, received noise levels in this paper are presented in terms
of dB re 11.1Pa in the 20— 1 000 Hz band at a depth of 20m, unless specified otherwise. Most
of the sound energy of ships occurs below 1 kHz; noise levels in higher frequency bands tend
to have a stable and predictable relationship to the level in the 20 — 1 000 Hz band. Hence,
the level in this lower frequency band is a useful index to higher frequency ship noises. Most
of the communication signals of belugas and narwhals occur between 500-5 000 Hz (Ford
and Fisher 1978; Sjare and Smith 1986a).

Analyses of Whale Vocalizations

The basic vocal repertoire of the narwhal has been described by Watkins et al. (1971) and
Ford and Fisher (1978). Narwhals emit primarily two types of sounds in the frequency range
audible to the human ear: clicks and pulsed-tone calls. Clicks are relatively narrow-band
pulses of constant frequency generated in series at regular, slow (mostly < 10 s) repetition
rates with most audible energy in the 3-7 kHz band. Pulsed-tone calls are rapidly-pulsed
emissions ranging from slower ratchet-like growls to higher-pitched creaking sounds,
generated primarily at frequencies from 0.5 - 6.5 kHz. Narwhals occasionally emit tonal
sounds (whistles) with changing pitch.
The basis vocal repertoire of the beluga in the high Arctic has been described by Sjare and
Smith (1986a). Tonal sounds are the predominant vocalizations and exhibit a complex array
of characteristics. Less frequently, belugas emit pulsed-tone calls including higher-frequen-
cy screams and buzzy sounds to lower-frequency squawks, barks and blaring sounds.
Both narwhals and belugas produce broad-band clicking sounds for echolocation purposes;
much of the energy of these signals occurred at frequencies beyond the recording sensitivity
of our equipment and were not considered in this study.
In June, the underwater environment at ice edges in Lancaster Sound is usually saturated
with the vocalizations of narwhals, belugas and bearded seals (Erignathus barbatus).
Because both whale species were often present at the same time during this study, vocaliza-
tions were often intermixed on recordings. After listening to some recordings on which only
one species was present, it was possible to "predict" from all recordings which species was
present and to roughly gauge its proximity to the recording station.
In 1983, predictions of absence or presence of narwhals and belugas were tested for
accuracy by comparison with independent observations of the ice-based observers. In 86
recording sessions obtained when synchronized ice-based observations were obtained, we
correctly "predicted" which species was present in the immediate area 96% of the time. In
five cases, belugas (2 cases) or narwhals (3), were audible on the tape but the observers did
not see any animals nearby. However, in each of these cases the species in question was seen

101
in the area during recording sessions immediately preceding or following. Incorrect predic-
tions of absence occurred on two occasions when narwhals were present but apparently silent.
Detailed analysis of whale vocalizations was conducted on the recordings obtained in 1983.
Taped 30-s segments were randomly selected for spectral analyses from 10-min recording
sessions with adequate signal: noise ratios. In some sessions where ship noise predominated,
it was necessary to select quieter moments when the ship shifted between ramming and
backing modes. The selected segments were analyzed using a Spectra Color real-time
spectrum analyzer for the frequency bandwidths of 60— 6000 Hz and 60— 12000 Hz. Vocal
activity was evaluated from 77 30-s spectrographic segments; 24 of these came from the pre-
disturbance period, 33 from disturbance periods, and 20 from intervening quiet intervals.
Attempts to quantify the level of vocal activity were difficult. For example, it was seldom
possible to distinguish individual click series from the general "clatter" when narwhals were
present, but pulsed tones tended to be distinctive, allowing enumeration. Belugas often
created a cacophony that was difficult to decipher. Overlapping signals could usually be
separated on the basis of relative intensity and close attention to the real-time spectral display
but it was often impossible to discern how many components constituted a single call. At
close quarters, belugas usually produced a bewildering array of sounds that we did not
attempt to quantify.

Results
Ship and Ambient Noise Levels

The Arctic is an ice-breaking ore carrier of about 20 000 deadweight tonnes and is 209 m
long with a draft of 11 m. The ship is powered by a single diesel engine producing 14, 770
HP on a single shaft with a variable pitch propeller. The John A. MacDonald is a Canadian
Coastguard ice-breaker of about 3 685 deadweight tonnes and is 96 m long with a draft of 8.6
m. The ship is powered by nine diesel electric engines producing 15 000 HP on three
propellers.
Figure 2 shows the received noise levels of the two ships at various ranges from the
Lancaster Sound ice edge during their approaches and subsequent ice-breaking operations in
1983. The Arcti c was first detectable by a tonal peak at 105 Hz on noise spectrographs when
the ship was at a range of about 130 km. This tone was probably the second harmonic of the
major 53 Hz tone that became evident as the ship approached closer. The fundamental tone

CL
150
ARCTIC •
MACDONALD 1:1
a>
AMBIENT
co 130

LU

110
L .1

C
90 1

-80 -60 -40 -20 0 20 40 60

DISTANCE (km) FROM SOURCE

FIG. 2. Received noise levels (20-1 000 Hz band) at various ranges during the approach of the MV
Arctic and the CCG John A. MacDonald to the Lancaster Sound ice edge and during ice-breaking
operations in 1983. Vertical lines show the maximum noise levels during backing and minimum levels
during ramming. The typical range of ambient noise at the ice edge is shown.

102
was related to the cylinder firing rate of the Arctic. The ship was distinctly, though faintly,
audible to the human ear by listening to the hydrophone signal when it was at a range of 85
km and the 20— 1 000 Hz band level was 101 dB. At a range of 47 km the noise level was
105 dB. At its closest point of approach (672 m) during ice-breaking operations, the noise
levels from the Arctic reached 136 dB. Noise levels of both ships during backing were usually
10-15 dB higher than during ramming. Ship noise during ramming often merged with back-
ground noise at ranges > 25 km but was intermittently strong during backing operations at
ranges > 45 km.
In 1982 the approach of the Arctic to the Admiralty Inlet ice edge was first indicated in the
noise spectra by a tonal peak at 205 Hz. This tone and another at 53 Hz became more
prominent as the Arctic approached closer. The ship was plainly audible (noise level 94 dB)
on the hydrophone signal as it worked its way through a small pack ice field at a range of about
46 km.
Source levels for the Arctic during ice-breaking operations were estimated from recordings
obtained in 1982 when the closest distance between the ship and the hydrophone was 463
m. Maximum received noise levels at a depth of 20m during backing attained 138 dB in the
10-1 000 Hz band. Estimated source levels, assuming spherical spreading, in the backing-
up mode (191 dB re 1 pPa at 1 m) were consistently higher than in the ramming mode (184
dB). Prominent tones occurred at 53, 106 and 205 Hz during ramming; the first two were
associated with the cylinder firing rate of the diesel engine (the fundamental and second
harmonic) and the latter, to the bubbler system that functions to reduce friction between the
ship's hull and sea ice. Cavitation noise dominated the sound spectra during backing
operations. Noise levels at a depth of 3 m averaged 12 dB lower than at 20 m due to the Lloyd
Mirror Effect.
Ambient noise levels at the Lancaster Sound ice edge in 1983 ranged from 93 to 104 dB
re 1 lea in the 20-1 000 Hz band at a depth of 20 ni. Noise levels at 3 m depth averaged 9
dB lower. Noise levels at the Lancaster Sound ice edge were higher than at the Admiralty Inlet
ice edge in 1982 (85-92 dB) due to higher sea states and more animal calls. Figure 2 shows
the typical range of ambient noise levels at the Lancaster Sound ice edge. The quietest periods
occurred after the ships had arrived at the ice edge and displaced the belugas and narwhals
from around the recording station.

Normal Distribution Patterns ofBelugas and Narwhals

The chronology of spring migration of whales in Lancaster Sound is dictated by ice
conditions, particularly the position of the fast ice edge. When the whales arrive in June and
early July, coastal areas are still bound in fast ice and, in some years, Lancaster Sound may
be covered by fast ice. The position of the ice edge tends to be stable in any one season but
it is highly variable from year to year; most often it occurs west of northern Baffin Island. The
present study was conducted during a period when the ice edge was located close to the
entrance of Lancaster Sound, unusually far east of its normal position (Fig. 1). The presence
of this ice barrier caused high concentrations of whales. Figures 3 and 5 show the typical
distribution patterns of belugas and narwhals along the Lancaster Sound ice edge.
In normal ice years, when Lancaster Sound is not covered by fast ice, the peak beluga
migration occurs in late June and early July, and the whales closely follow the north shore
of the sound toward their summering grounds in the central archipelago (LGL 1983). In ice-
barrier years, many belugas congregate along the ice edge. They distribute themselves more
or less evenly along the ice edge, although they tend to concentrate along the coast of Devon
Island (Fig. 3, 5). In offshore areas, belugas remain close to the ice edge and their movements
are usually parallel to it; they appear to be waiting for breakup and their movements tend to
be causal and exploratory. Herd movements are often coordinated in one direction over a
large area. The belugas along the ice edge in June tend to be predominantly adults, often
occurring in distinctive all-white pods, believed to consist of males,

103
 Unlike belugas, which follow coastal routes or ice edges during spring migration, narwhals
are more pelagic in habit, arriving in a broad-front migration through the pack ice of northern
Baffin Bay (Davis et al. 1978). Although some narwhals move west through Lancaster Sound
as ice conditions permit, the major summering grounds of the narwhal are in the fiord
complexes of northern Baffin Island, particularly Admiralty Inlet. Thus, the existence of an
ice edge across the sound does not constitute such a barrier to migration as it does for belugas.
In mid to late June small numbers of narwhals, primarily adult males, occur along ice edges
— particularly in the southern half of the Sound adjacent to their summering areas (Fig. 3,
5) Most of the narwhals remain widely distributed in the offshore pack ice of Baffin Bay until
the fast ice begins to disintegrate.
Thus, in most years there is relatively little intermingling of narwhals and belugas because
their migration routes are segregated — belugas keeping mainly to the north shore of the
sound and the narwhals keeping to the south side and offshore. During the three years of this
study both species intermingled along the ice edge. This provided an unusual opportunity to
compare their responses to ships in the same environmental setting.

80'

DEVON ISLAND

291 B

LANCASTER
SOUND

(1
.
1294 B

102 B

LEGEND 74 ."
* OBSERVATION AND
RECORDING STATION
94 B
PRIMARY DIRECTION OF
32 N
BELUGA MOVEMENT
PRIMARY DIRECTION OF 158 B
NARWHAL MOVEMENT
77 N

BAFFIN ISLAND BYLOT

ISLAND

1 83 032 181.
80'

FIG.3. Normal distribution pattern of belugas (B) and narwhals (N) along the
Lancaster Sound ice edge, 23 June 1983.

104
Whale Responses in 1982

A total of 42 narwhals and 18 belugas were observed during an aerial survey along the
Admiralty Inlet ice edge on 27 June prior to the arrival of the Arctic. Underwater recordings
obtained from the center of the ice edge were saturated with calls of both species although
none were seen within 3 km of the recording station. Whale calls still dominated the ambient
noise at 12:15 on 28 June but a small peak in the noise spectra at 205 Hz indicated the presence
of the Arctic at about 105 km. At 14:31 the ship was 46 km from the ice edge and the noise
level was 94 dB. At this time, tonal calls of belugas dominated the higher frequencies.
However, between 14:40 and 14:50, the type of calls changed noticeably: more distinctive
pulsive calls were emitted. These calls occurred just prior to a notable movement of belugas
at the ice edge and the calls were tentatively attributed to them. Although the ice-edge
observers had not seen any whales until this time, at 15:00 many belugas were seen moving
rapidly westward along the ice edge. Approximately 200 — 300 whales passed between 15:00
and 16:25, most of them in the first 15 min. The belugas closely followed the ice edge, often
diving under the ice. Their swimming behavior appeared abnormal; they rose unusually high
out of the water and the surface-dive cycle of individuals appeared to lack synchrony with
other whales in the same pod. After 16:25, the movements of a few belugas along the ice edge
became erratic and turned eastward, apparently "cut off" from the major westward move-
ment. At this time the ship was about 13 km to the north of the ice edge and approaching
rapidly.
Small numbers of narwhals were observed at the ice edge from 16:34 until after the ship
struck the ice at 17:14. Their behavior was notably different than that of the belugas. Their
movements were slow and they frequently rested motionless close to the ice edge. In contrast
to the rapid forward-rolling dives of the belugas, narwhals tended to sink quietly beneath the
surface with little audible respiratory sound. Pods of narwhals appeared more cohesive than
those of belugas, and individuals in pods were seen contacting each other with their tusks or
flippers. Narwhals were seen within 5 km of the ship at 17:49 when it was breaking ice.
Whales and whale sounds were absent from the Admiralty Inlet ice edge throughout 29
June as the Arctic continued to break ice 10— 26 km from the ice edge. An aerial survey of
the ice edge late in the day found no whales, although 1 072 belugas and 118 narwhals were
observed along the north coast of the Brodeur Peninsula and a short section of the Lancaster
Sound ice edge. Some belugas were moving back toward Admiralty Inlet at 23:00, and by
noon on 30 June a mixed herd of 200 belugas and narwhals was observed in a "feeding frenzy"
at the centre of the ice edge. The Arctic was about 48 km from the ice edge at this time. Ship
noise was intermittently audible to the human ear on hydrophones (102 dB). Thus within 43
h of the arrival of the Arctic, both narwhals and belugas had resumed apparently normal
activities along the ice edge although the ship was still audible.

Whale Responses in 1983

Although we were unable to conduct strictly controlled disturbance experiments in 1983,

ship activity occurred on an intermittent basis, permitting observations of the reactions of
narwhals and belugas to slip sounds during discrete intervals following quiet periods. Seven
disturbance periods were recognized, varying in intensity from ship approaches to the ice
edge to subsequent start-ups and ice-breaking operations as the ships moved over 40 km into
the fast ice away from the ice edge.

Pre-Disturbance Conditions

Aerial surveys were conducted on five days prior to the arrival of the Arctic at the Lancaster
Sound ice edge in late June. Belugas were widely distributed along the ice edge with a

105
maximum count of 939 on 23 June (Fig. 3). Narwhals were consistently found along the
southern third of the ice edge where a maximum of 115 was counted on 21 June. Typically
both species were associated strongly with the ice edge; the majority occurred within 400 m
of the edge and their movements were generally parallel with it. Over 90% of the narwhals
were adult males and 61% of the belugas were thought to be males (i.e., either in distinctive
all-white "bachelor" pods or adults not accompanied by immatures).
Ice-based observers recorded movements of belugas and narwhals throughout a45 h period
prior to the arrival of the Arctic. During this period, movements were back and forth: 108
narwhals passed north and 117 passed south; 241 belugas moved north and 299 south.
Underwater recordings during the pre-disturbance period were saturated with vocalizations
of both species along with the calls of bearded seals.

Reactions of Whales to Ship Approaches

At 02:00 on 24 June, the Arctic was in northwest Baffin Bay about 250 km from the ice-
edge recording station in Lancaster Sound. It was not a factor in the ambient noise which was
dominated by water and ice noise, and bearded seal calls. The earliest indication of the
presence of the ship was at 06:00 when the Arctic was 132 km away and a weak tone at 105
Hz was present in the noise spectra.
The first evident changes in whale behavior were noted at 08:00 when the ship was 84 km
from the ice edge. Although no whales were seen in the immediate vicinity of the recording
station, several distinctive vocalizations, identical to those heard at 14:40 on 28 June 1982,
were heard. At this time, the Arctic was faintly audible on recordings although the noise level
(101 dB) was within the range of ambient noise levels during the pre-disturbance period.
Between 09:35 and 10:22 when the Arctic was 55 — 40 km to the northeast, 44 belugas
moved rapidly north past the observation site along the ice edge, showing the same "flight"
reaction as in the previous year. During this period many harsh pulsive calls, presumed to be
beluga calls, were heard. Small groups of adult narwhals were observed during this period
and their behavior, as in the previous years, was notably different from that of the belugas:
the narwhals exhibited a "freeze" response to the oncoming ship. The observers were
unaware of the proximity of the Arctic which was not expected until later in the day. The
Arctic entered the ice edge at 12:20 about 4 km south of the observation site but visibility was
limited to <1 km in a heavy snowfall.
Because of poor weather and unstable ice, we were tillable to determine how far the whales
were displaced or when they returned to the disturbance area. The Arctic continued breaking
ice until 09:07 on 25 June without making much progress. An aerial survey, conducted 2 h
after the ship shut down less than 10 km into the ice, showed that narwhals had already
returned and, in fact, were concentrated in the bight at the ship's entry point. Belugas were
just arriving in the area from the north along the ice edge.
The Arctic remained idle in the ice until 07:56 on 26 June when it began ice-breaking. Noise
levels at the ice edge reached 124 dB. The reaction of belugas to the ship's start-up was
dramatic: over 225 belugas passed the ice-edge observer in 9 min, all showing a typical "flee"
response. A few narwhals that were present in the area exhibited a "freeze" response and
eventually disappeared. Strong ship noise masked many of the whale sounds but several
distinctive pulsed calls were heard during brief lulls between ramming and backing modes
of the ship.
An aerial survey beginning 1.5 h after the ship started up showed that belugas had been
displaced 49 km north along the ice edge and were still moving rapidly away from the ship
(Fig. 4). About 1 100 northbound belugas were counted but this figure is imprecise because
the whales were tightly clumped and many were undertaking long dives beneath the ice edge.
Narwhals had also been displaced; none remained within 18 km north of the ship track.
Narwhals and belugas that departed south of the ship track had ceased travelling when they

106
 1
83 1 82' 81' • 80

DEVON ISLAND •

1040 B

LANCASTER
SOUND

147 B
7N

24 N 74 •1
E74'

1590 B
53 N
M AC7
A RC 0
Cape
Charles
Yorke

BAFFIN ISLAND BYLOT

ISLAND
to 20 40
LECECELECH km
, 83' ,82' • ,8t• 80'

Fia. 4. Distribution and movements of belugas and narwhals following the arrival
of the CCG John A. MacDonald (MAC) on 26 June 1983. The position of the MV
Arctic (ARC) is also shown.

reached the coast of Bylot Island, 20 km away, and many were milling at the ice — land
interface. Undoubtedly the whales first responded to the start-up of the Arctic but, by the time
of the survey, they may also have been influenced by the approach of the MacDonald. The
MacDonald arrived at the ice edge at 13:30 and proceeded to assist the Arctic which had shut
down after being unable to make headway.
An aerial survey beginning 1 h after the arrival of the MacDonald showed that belugas had
continued their rapid northward movement along the ice edge and were over 80 km from the
ship track near the coast of Devon Island. The closest narwhals were 37 km north of the ship
track and it was apparent, from reduced numbers, that some had dispersed offshore. Large
numbers (1 590) of belugas remained milling at the land-ice interface on the south side of the
ship track.
After their departure, belugas were not seen or heard at the observation site as the
MacDonald continued breaking ice throughout the day. However, narwhals returned to the
ship track by 19:30 when the ship was operating at a distance of 15 km. Received noise levels

107
at 20:00 were 120 dB and, during quieter intervals, a few narwhal click series were audible.
At 21:00, the maximum received noise level was 118 dB when the ship was 19 km away;
narwhal click "clatter" was strong in the quieter intervals.

Reactions of Whales to Intermittent Ice-breaking Operations of Diminishing Intensity

From 27 June to 1 July, the MacDonald and the Arctic progressed slowly westward through
the fast ice, averaging just over 4 km per day. The Arctic operated intermittently, following
the track created by the MacDonald. Although noise levels diminished, they were not
insignificant; received levels at the ice edge reached 115 dB during backing operations on 1
July when the MacDonald was about 40 km from the edge.
Except for one instance on 29 June when MacDonald returned to within 18 km of the ice
edge to assist the Arctic, we did not discern any notable influence of ship start-ups or
operations on the behavior of narwhals. On this occasion the maximum noise level, with both
ships operating, reached 121 dB, and although the narwhals left the area near the ship track
they returned 8 h later when both ships had returned westward (noise levels 111 dB). In fact,
following the arrival of the ships at the ice edge on 25 and 26 June, narwhals appeared to be
attracted to the rubble-filled ship track, and many engaged in dives directed toward the track,
possibly exploring for a westward passage. This "exploratory" activity appeared to wane
following their initial interest; fewer narwhals appeared sporadically around the track. On 30
June, most of the narwhals had moved to a feeding area along the Navy Board Inlet ice edge.
Belugas and narwhals, along with many harp seals (Phoca groenlandica),remained in a large
feeding aggregation at this location until 2 July.
Belugas continued to react as the ships slowly progressed away from the ice edge. Although
aerial survey coverage was incomplete after 27 June, due to poor weather, belugas appeared
to avoid the southern part of the Lancaster Sound ice edge until 2 July. Belugas reacted by
emitting distinctive pulsive calls during most ship start-ups, and at other times during ice-
breaking operations their vocal activity suggested a heightened level of excitement.

Whale Responses in 1984

An aerial survey on 27 June, the day before the ships arrived, documented normal patterns
of whale distribution (Fig. 5). Belugas were distributed more or less evenly along the length
of the ice edge and narwhals were distributed along the southern half of the ice edge. Both
species were widely distributed in small numbers on the offshore transects.
Major changes in whale distribution had occurred when the next survey was conducted at
08:00 on 28 June, 2 h after the Arctic arrived at the ice edge (Fig. 6). The CCG Louis St.
Laurent was approaching the ice edge; it arrived at 09:00. Virtually all the whales had vacated
the offshore waters in a large zone around the vessels. No belugas and only four narwhals
were found along a 27 km stretch of the ice edge centered in the position of the Arctic.Belugas
and narwhals 20 — 80 km north of the ships were moving rapidly north along the ice edge.
Whales over 80 km north of the ships were milling or resting, apparently not reacting to the
ships.
By 16:00 on 28 June, 9 h after the arrival of the St. Laurent, whales had begun to return
toward the area. The ships had been drifting since their arrival with their main engines off but
some shipboard machinery was operating. Narwhals had advanced substantially farther
south than had belugas but both species were moving south toward the ships.
During the night, the two ships moved into the ice a short distance and again stopped. An
aerial survey on the morning of 29 June showed that narwhals had re-occupied the ice edge
near the ships. However, the distribution of belugas along the ice edge was still affected by
the ships; most remained > 10 km north of the ships. The offshore distribution of both species
had returned to pre-disturbance patterns.

108
 B4 . 82 .

DEVON ISLAND

LEGEND
NUMBER OF WHALES BELUGA NARWHAL

l-5
6-15 0
16-50 0 LI
51-150 0
>150
0
MOVEMENTS
LANCASTER
MAJOR MINOR MILLING
SOU/VD oa o0
-°)12
@E

--?,"

1; 1

Q`•
74`71

:ac)

QQ
Q`

Cape
Charles
Yorke

5 10 20 25
km
BAFFIN ISLAND
,83 . i 82 .

Flo. 5. Normal distribution patterns and movements of belugas and narwhals before the arrival of
ships, 27 June 1984. Symbols indicate the number of whales observed in each 1—min period. "Major"
movements were indicated when >75% of the whales observed in a time period were moving in a given
direction, "minor" movements indicated 25-75% of the whales. When > 50% of the whales were not
showing directed movement, the whales were considered to be "milling."

In the early evening of 29 June, both ships started up and moved south along the ice edge,
then turned west into the ice, beginning ice-breaking operations. We were surveying the
easternmost transect line (20 — 25 km from the ice edge) when the ships started moving. The
reactions of the whales were immediate and striking. Virtually all whales began moving away
very rapidly, exhibiting porpoising behaviour and spending little time underwater. Belugas

109
 84 82 .

DEVON ISLAND

,• NUMBER OF

1-5
WHALES
LEG END
BELUGA NARWHAL

6-15 0
16-50 0
51-150 0
>150
0
itl'C MOVEMENTS
LANCASrER MAJOR MINOR MILLING
SOUND àà 00
ô 00 @E

1/6 01

746

'9sr, 74 .

V,

__APPROXIMATE LOUIS ST LAURENT TRACK

LSL
fl‘ • •

• ARC
os

6. Distribution and movements of belugas and narwhals following the arrivai of the MV Arctic
FIG.
(ARC) and CCG Louis St. Laurent (LSL) on 28 June 1984 (08:00-10:00).

were moving faster than narwhals but even the latter species was moving t'aster than
previously observed. When we surveyed the ice edge an hour later, narwhals and belugas
were moving quickly north away from the ships at distances > 20 km, nearly all the way to
the Devon Island coast 80 km away. South of the advancing ships, narwhals were moving
rapidly toward the coast of Bylot Island.
Documentation of a return to normal conditions was hindered by fog on 30 June and was
complicated by the appearance of cracks in the fast ice on 1 July. The cracks caused many

110
whales to move into the ice away from the ice edge. Nevertheless, it was apparent that 45
h after the ships had started up, beluga distribution along the ice edge and offshore had not
returned to normal. The ships were 45 — 50 km into the ice at this time. The ship noise reaching
the ice edge was unmeasured but no doubt rather weak. Beluga distribution was close to
normal 16 h later on the morning of 2 July. The distribution of narwhals on 1 July appeared
to be fairly normal, but so few whales were present that the situation was difficult to evaluate.

Influence of Ships on Whale Vocalizations

In June, the underwater environment at ice edges in Lancaster Sound is usually saturated
with the vocalizations of narwhals, belugas and bearded seals. The presence of narwhals is
evident by their click "clatter" and pulsed calls, of belugas by their bird-like whistles and
trills, and of bearded seals by their long, descending trills.
Table 1 shows the frequency of different call types of the beluga during undisturbed
situations and during periods when ships were operating at the Lancaster Sound ice edge in
1983. This repertoire is similar to that recorded at Cunningham Inlet (Sjare and Smith 1986a)
and corresponding call types are cross-referenced where possible in the table. Tonal calls
accounted for the majority of vocalizations during undisturbed periods (96% of all calls) and
during periods when ships were operating (84%). Most notable was the increase in the
occurrence of pulsed tone calls, descending tonals and chirp trains during disturbance
periods. Pulsed tone calls and descending tonals were most obvious during initial stages of
disturbance whereas chirp trains were heard during later disturbance periods. Also notable
was the decrease in the occurrence of rising and stepped tonals during disturbance periods.
In addition to changes in call types in response to disturbance, belugas also appeared to alter
the frequency emphasis or duration of some of their signals. During disturbance periods,

TABLE 1. Frequency of occurrence of various beluga vocalizations during the presence and absence of ship ice-
breaking noise at the Lancaster Sound ice edge, 1983.

% Occurrence % Composition
in quiet periods in noisy periods in quiet periods in noisy periods
Call Type (n = 31) (n = 26) (n = 313) (n = 352)
Tonal
constant ( 1 a)" 68 62 21 12
broken ( 1 b) 48 50 12 15
stepped (3a) 52 19 11 1
falling (4a) 19 46 5 13
rising (2a) 32 23 10 3
undulating regular (6a) 26 31 6 5
undulating irregular (6b) 23 54 7 9
chirp 32 31 11 7
yelp 16 35 4 5
chirp train (7) 23 69 4 13
up trills 13 4 6 2

Pulsed tones
buzz (Grp Id) 10 31 4 6
tonal scream (Grp I a) 0 23 0 7
noisy scream (Grp 3i) 0 12 0 3
a Number in parentheses refers to designation by Sjare and Smith 1986a.

111
undulating tonals began at lower frequencies (2.5 kHz ± SD 1.0 vs 5.0 ± 2.9, n = 16, 16;
P < 0.01) and covered a wider frequency range than during quiet periods. Generally, the
durations of many tonal signals increased during ship disturbance (0.9 s± SD 0.6 vs 0.7 ± 0.5,
n = 254, 235; P < 0.01). These quantifiable changes in frequency range and duration, along
with distinctive call types, gave a distinct aural impression of alarm or excitedness among
belugas in situations of disturbance.
Except for three occasions on which narwhals temporarily ceased producing click emis-
sions and pulsed-tone calls during initial phases of disturbance (ship approach and start-ups),
there was no evidence that they changed the type or structure of their signals in response to
ship noise. Pulsed-tone calls accompanied click "clatter" in 68% of 25 spectrographic
segments during disturbance periods, 80% of 20 segments in the pre-disturbance period, and
71% of 14 segments in intervening quiet periods. The mean number of pulsed calls per 30-
s segment was similar in disturbance periods (5.2 ± SD 5.8), predisturbance periods (3.8 ± SD
3.9) and quiet intervals (4.2 ± SD 4.0). There was no evidence of any change in the median
frequency or duration of pulsed calls in response to ship noise. Similarly, there was no
evidence that ship noise influenced click emissions.

Discussion

Responses of Whales to Ships

The results from this study demonstrate that narwhals and especially belugas, in the high
Arctic, are extraordinarily sensitive to shipping activity in the spring. However, the two
species responded in surprisingly different ways, considering their supposed relatedness. We
refer to these different behaviors as the "flee" response of the beluga and the "freeze"
response of the narwhal. The following behaviors constitute the general observable or
audible responses of belugas and narwhals to vessels:

Belugas Narwhals
— rapid movement slow movement or motionless

— herd formation, loss of pod integrity no herd formation; pod cohesion, body
contact
— asynchronous, shallow dives sinking submergence

— pagophilic, littoral pagophilic, pelagic

— "alarm" calls silent

The Inuit recognize these combinations of behavioral traits as "ardlingayuk" — fear of

killer whales. The primary fear/avoidance responses of these species have probably evolved
in response to predation by killer whales (Orcinus orca). The killer whale is probably the only
significant natural predator of belugas and narwhals in the high Arctic. It may be reasonable
to view the responses of belugas and narwhals to potentially threatening situations, such as
ships, from this perspective. The narwhal's strategy seems to be to remain inconspicuous —
visibly and audibly, whereas the beluga's strategy is to flock and flee.
In the particular offshore setting of this study, belugas showed much stronger apparent
reactions to ship disturbances than did narwhals. The beluga tends to be more of a shallow
water species than is the narwhal (at least in summer). The stronger reactions of belugas to
ship traffic may be indicative of their sense of vulnerability in an offshore situation. It is
interesting that the degree of displacement of belugas by ships appeared to depend on the

112
proximity of coastal waters and ice cover; animals on the side of the ship closest to shore
stopped travelling when they reached the shore — ice edge interface (e.g., 18 km in 1983),
whereas those on the offshore side of the ship continued to travel (as much as 80 km) until
they reached the opposite shore. Although narwhals initially sought cover along the ice edge
during the approaches of ships, they later tended to disperse offshore, perhaps toward pack
ice fields in Baffin Bay, their habitat for much of the year. The widespread displacement of
narwhals observed in offshore Lancaster Sound in response to ship approaches in 1984 may
have been due to lack of adequate pack ice cover. During most of the year, narwhals and
belugas inhabit areas of dense ice cover (Finley et al. 1982; LGL 1983). Ice edges provide
only limited potential for protective cover as compared to pack ice, which killer whales tend
to avoid.
The major displacement of belugas from the Lancaster Sound ice edge may also have been
due, in part, to their lack of a vested interest in remaining in the area. The ice edge acts as a
barrier to belugas and their activities in such ice-barrier years might best be described as
"passing time"— waiting for the ice to breakup so they can proceed to traditional summering
areas in the central arctic archipelago. Their normal movements along the ice edge appear
casual and exploratory. There is no evidence that ice edges constitute significant feeding
habitat for belugas, so they may have no vested interest in remaining in an area when a ship
approaches. Later in the summer, in their traditional estuarine habitat, belugas often show
persistence in using certain estuaries in spite of vessel traffic, suggesting that they have a
vested interest in remaining in specific areas (Finley et al. 1982, 1987). Belugas may then
tolerate excessive noise because there are no alternative areas that meet their ecological
requirements.
The long-range responses shown by both species to the approach of the Arctic in 1982 and
1983 are unprecedented in the marine mammal literature. This may be largely explained by
the fact that no similar field studies have been conducted in pristine marine environments
with industrially naive populations of marine mammals. High Arctic populations of whales
may be especially naive because they are exposed to very low levels of shipping activity; their
responses may be exaggerated relative to populations that are exposed to more frequent ship
traffic. Also, it is generally agreed that marine mammals are more likely to show strong
avoidance reactions to vessels if there are sudden changes in course, speed and noise levels
(see review by Richardson et al. 1989). The relatively rapid approach of the ships to the ice
edge and the sudden start-ups and variable noise levels associated with ramming and backing
modes of the ice-breakers may have evoked stronger reactions than if the noise source had
been constant. Belugas and narwhals returned to the disturbance area when noise levels were
higher than those to which they initially reacted. This suggests that their initial responses were
startle responses, and that their sensitivity to ship noise declined subsequently. In fact, apart
from their initial reactions to ship aproaches, narwhals showed little or no reaction to ship
activities such as start-ups or ice-breaking when the ships were operating within the fast ice
away from the edge; we could not discern any influence of received noise levels as high as
120 dB on the level of narwhal vocal activity or on their surface behavior.

Vocal Responses of Whales to Ships

When this study began, we were not familiar enough with the vocal repertoire of either
whale species to be able to ascribe call types with certainty to one or the other. Nonetheless,
during the approach of the Arctic to the Admiralty Inlet ice edge, a marked change was
evident in the type and quality of whale vocalizations. A large proportion of the calls were
of a noisy pulsive type (screams) that were not heard outside the disturbance period. These
calls were tentatively attributed to belugas which at the time were moving rapidly along the
ice edge in response to the approaching ship. In 1983 these pulsive "screams" were first heard
when the Arctic was approaching the ice edge at a range of 85 km and they were heard in

113
 the initial stages of virtually all ship disturbances when belugas were present and showing
"fleeing" reactions. In most cases narwhals were also present when pulsive "screams" were
heard but on one occasion only belugas were visible when these calls were heard. This, and
the association of these calls with belugas exhibiting a fleeing response in 1982 and 1983,
strongly suggests that belugas were the source; their coincidence with the panic movement
suggests that these calls represent an alarm signal. These calls are similar to the pulsed tone
sounds of belugas described by Sjare and Smith (1986a) in Cunningham Inlet. Sjare and
Smith (1986b) found no evidence that these calls were used more often in situations of alarm;
they were also heard during social behavior and herd movements. However, pulsed tone calls
show much variation in amplitude and frequency emphasis making them difficult to catego-
rize, let alone assign a functional value. It is likely that the alarm situations observed at
Cunningham Inlet (e.g. polar bears swimming in water, aircraft overflights) are not compa-
rable to an ice-breaker. Sjare (per. comm.) suggests that there may not be a specific alarm call
as such; rather, there may be a tendency for the noise component or harshness of a certain call
to increase when an animal is under stress. This is a common characteristic of mammalian
vocalizations.
Tonal vocalizations are widely used by odontocete whales in communication and probably
find their greatest expression in belugas. Tonal signals tend to be species-specific (Steiner
1981), and in some species tonal signals are believed to function as signature calls, i.e., to
identify individuals and their location (Caldwell and Caldwell 1965). The Caldwells believed
that the emotional state of individuals was transmitted through variations of rapidity,
repetition rate, loudness and/or duration of the basic signal contours. For example, dolphins
(Tursiops truncatus) recorded in a variety of stressful situations produced only "excited"
variants of their signature whistles. The Caldwells concluded that there was no special alarm
signal. However, Dreher and Evans (1964), who studied the tonal vocalizations of four
species of dolphins, found that one general call, common to all, was the downswept tonal, a
signal that was "definitely associated with great fright and disturbance." We heard similar
tonal calls along with pulsed screams in the initial phases of ship disturbance in 1983. These
calls were heard along with "excited" variants of other tonal vocalizations. The spectrum of
tonal calls used in the context of disturbance was also distinctive by the absence of certain
calls heard during undisturbed situations (e.g. rising and stepped tonals). In general, the mea-
surable characteristics of wider frequency excursions and longer durations of certain tonal
calls, the absence of some call types, and the use of pulsed tone sounds gave a distinct aural
impression of excitement or alarm by belugas in situations of disturbance.
The vocalizations of narwhals and belugas are surprisingly different considering that they
are in the same family. Low-frequency (< 10 kHz), narrow-band click signals are the
prevalent sounds produced by narwhals in Lancaster Sound along with pulsed tone calls.
Indeed, dense click "clatter" contributes significantly to the high ambient noise levels at the
Lancaster Sound ice edge. It is uncertain whether these clicks function in echolocation or in
the maintenance of social organization. Ford and Fisher (1978) suggested that low-repetition
rate click series could function in range finding and topographic analysis and that individual
identity, emotional state and position were communicated through pulsed-tone calls. Wat-
kins et al. (1971), noting the similarity between narwhal and sperm whale (Physeter catodon)
sounds, suggested that individual click series may contain information about individual
identity and position. Three properties of low-frequency clicks — regularity of repetition
rate, apparent omnidirectionality and signal locatability — might be appropriate attributes of
a communication signal. Often at the Lancaster Sound ice edge, click signals were heard for
several minutes duration in the absence of pulsed-tone calls, suggesting that the narwhals
were maintaining contact by the one mode alone. Like the sperm whale, the narwhal is a
pelagic species and appears to be a deep-diving teuthophage (squid-eater); such behavioral
and habitat similarities might favour the development of similar acoustic behavior for main-
tenance of spatial organization.

114
 The only audible response of narwhals that we could discern during ship disturbance was
a brief cessation of sound (both clicks and pulsed tones), but this occurred only in the initial
phases of disturbance when the ship was approaching the ice edge in 1982 and 1983. A shot
fired into the water by hunters located about 2 km from our recording station in 1983 had the
same transient effect on dense click "clatter" of narwhals. In fact, following the shot, the noise
level dropped from 110 dB to 103 dB. The same effect has been heard in an underwater
recording of narwhals that were being pursued by hunters in a Baffin Island fiord in late
summer of 1984 (Finley, unpubl. data). The silent behavior of narwhals corresponds to their
"freeze" reaction in situations of distress. In our experience with narwhals in stressful
situations (e.g. Finley et al. 1980; Finley and Miller 1982), we have never observed a rapid,
mass-movement of animals like belugas that might require a herd-coordinating signal. These
observations fit the "ardlingayuk" model if the narwhal's primary survival strategy is to
remain undetected by its underwater predators.

Potential for Acoustic Interference

When considering the potential effects of industrial noise upon hearing among whales, it
is important to estimate the zone of influence around the noise source. Richardson (1990) has
suggested several different criteria for defining the zone or radius of influence; (1) the radius
of detection, within which the industrial noise is detectable; (2) the radius of masking, within
which the noise affects the ability of the animal to detect other sounds; and (3) the radius of
behavioral responsiveness, within which the behavior of animals changes overtly. Several
factors are important in determining the zone of noise influence; these include the noise
frequency and source level, rate of attenuation, ambient noise levels, the hearing sensivity of
the animal at different frequencies and "critical ratio," (i.e, the intensity by which a signal
must exceed background noise in order to be heard). Because of the many sources of
variability, the radius of influence of an industrial activity will vary widely with location and
time. Also, very little is known about the function of various communication calls of whales
and the range at which the calls are used, so it is difficult to evaluate the effects of increased
noise levels on their effective communication range.
Toothed whales can hear sounds over a very wide frequency range from about 40 Hz to 150
kHz but certain bandwidths are utilized by whales for different functions; in general, high
frequency sounds are used for echolocation, and low-frequency sounds (within the human
audible range) are used in communication. At high frequencies, the beluga's echolocation
abilities are likely to be little affected by vessel noise. It is at the lower frequencies that sound
detection may be limited by vessel noise.
The audiology of the beluga has been conducted with captive animals using tonal
frequencies between 40 Hz and 125 kHz (White et al. 1978; Awbrey et al. 1988; Johnson et
al. 1989). Beluga hearing was found to be most sensitive at frequencies between 20 and 75
kHz, with an upper limit of hearing of 125 kHz. Like other odontocete whales, their peak
sensitivity correlates well with the high frequency character of their echolocation signals. At
frequencies below 10kHz, the absolute tonal thresholds of belugas steadily decline (e.g., 81
dB at 4 kHz, 102 dB at 1 kHz, Johnson et al. 1989) parallel with an increase in natural ambient
noise.
Using the audiogram of the beluga produced by White et al. (1978) and estimated noise
attenuation rates for an ice-breaking tanker, Mansfield (1983) calculated that a beluga with
an absolute auditory threshold of 75 dB at 3.85 kHz (an estimated average frequency of its
communication sounds) would first be able to detect a tanker breaking ice at full power at a
distance of 5 km. This simple calculation did not take into account the critical ratio, calculated
by Johnson et al. (1989) to be an additional 17 dB at this frequency. By similar deduction,
using Johnson et al. 's test results, we can calculate that, with an absolute auditory threshold
of 104 dB at 1 kHz and a critical ratio of 17 dB, a beluga would be unable to detect ships ice-

115
breaking at full power at ranges > 20 km, judging from the received noise levels shown in
Fig. 2. However, the present study suggests that belugas were aware of a ship's approach at
a distance of 85 km and certainly they showed overt behavioral responses at ranges > 40 km.
There are many factors which contribute to noise attenuation but these would not account for
the disparity between our field observations and theoretical calculations. We can only
suggest that the disparity is due to the application of laboratory audiograms to natural
situations. It is possible that the estimated auditory thresholds of belugas and other species
are too high, especially at frequencies below lkHz (see review by Richardson et al. 1989).
Almost all studies of audition and auditory masking of toothed whales have employed tonal
signals against a background of white noise. Relatively few of the sounds produced or
received by whales are pure tones and presumably the cetacean brain is finely tuned to
modulated signals of broader frequencies.
In their natural environment, belugas and narwhals presumably depend on passive listening
to low-level, low-frequency ambient sounds of moving and solid ice to obtain information
that is critical to their navigation and survival in ice-covered waters. Artificially increasing
the background noise could presumably interfere with this information-gathering process.
Masking by industrial noise could conceivably result in reduced navigational and foraging
capabilities in whales, thereby leading to physiological stress and reduced fitness of
populations. Such an effect would be difficult to demonstrate.
To some degree the acoustic system of belugas and narwhals must be adapted to deal with
relatively high levels of ambient noise in a pack ice environment. Studies by Au et al. (1985)
and Turl et al. (1987) have demonstrated that belugas are adept at tuning the frequency and
duration of their echolocation signals in response to changing levels of ambient noise. Our
studies suggest that belugas may also change the frequency and duration of their social
signals during periods when ship noise predominates. We are just beginning to appreciate the
sophistication of the acoustic processing system of these remarkable animals.

Acknowledgments

Principal funding and support for this study was provided by the Northern Environmental
Protection Branch of the Department of Indian Affairs and Northern Development. Additio-
nal funding and support was provided by Canarctic Shipping Ltd., Canadian Coast Guard
(Ministry of Transport), Arctic Pilot Project, Dome Petroleum Ltd., Gulf Canada Resources
Inc., Consolidex-Magnorth Ltd. and Environment Canada. We thank D. Stone, R. Hurst and
T. Langtry of DIAND and Captain I. Man of the CCG for their major roles in planning and
coordinating the study. Mr. M. Luce of Canarctic Shipping was supportive in many aspects
of the field program.

G. Silber operated recording equipment in the field. Logistic support on the ice was
provided by D. Kalluk, O. Nakitavak, Tuktu and G. Williams of Arctic Bay. B. Hough of
Polar Continental Shelf Project coordinated aircraft support and radio communications. J.
Ford of the Vancouver Public Aquarium provided advice and assistance in the analysis of
whale vocalizations. R. Ferraro of the Institute of Applied Physiology and Medicine, Seattle,
assisted in the production of real-time spectrographs. Dr. W. J. Richardson, B. Sjare, L.
Darling and an anonymous reviewer provided constructive comments on the manuscript.

116
 References

Au, W. L., D. A. CARDER, R. H. PENNER, AND B. 1986. Reactions of belugas whales and
L. SCRONCE. 1985. Demonstration of adapta- narwhals to ship traffic and ice-breaking along
tion in beluga whale echolocation signals. J. ice edges in the eastern Canadian high Arctic.
Acoust. Soc. Am. 77: 726-730. Dep. of Indian and Northern Affairs Canada,
AWBREY, F. T., J. A. THOMAS, AND R. A. Ottawa, Environmental Studies No. 37: 301 p.
KASTELEIN. 1988. Low-frequency underwater MANSFIELD, A. W. 1983. The effects of vessel
hearing sensitivity in belugas, Delphinapterus traffic in the Arctic on marine mammals and
leucas. J. Acoust. Soc. Am. 84: 2273-2275. recommendations for future research. Can.
CALDWELL, M. C., AND D. K. CALDWELL. 1965. Tech. Rep. Fish. Aquat. Sci, 1186: 97 p.
Individualized whistle contours in bottlenosed RICHARDSON, W. J. 1990. Behavioral responses of
dolphins (Tursiops truncatus). Nature 207: bowhead whales to disturbance: interpretation
434-435. and implications. Extended abstract. Fifth
DAVIS, R. A., W. J. RICHARDSON, S. R. JOHNSON, Conference on the Biology of the Bowhead
AND W. E. RENAUD. 1978. Status of the Lan- Whale. Anchorage, Alaska,
caster Sound narwhal population in 1976. Rep. RICHARDSON, W. J., C. R. GREENE, J. P. HICKIE, R.
Int. Whaling Comm. 28: 209-215. A. DAVIS, AND D. H. THOMPSON. 1989. Effects
DREHER, J. J., AND W. E. EVANS. 1964. Cetacean of offshore petroleum operations on cold wa-
communication, p. 373-393. In W. N. Tavol- ter marine mammals: a literature review, 2nd
ga [cd.], Marine bioacoustics., Vol. 1. Perga- edition. American Petroleum Institute Report
mon Press, New York, NY. No. 4370. Washington. 385 p.
FINLEY, K. J., AND G. W. MILLER. 1982. The 1979 SJARE, B. L., AND T. G. SMITH. 1986a. The vocal
hunt for narwhals (Monodon monoceros) and repertoire of white whales, Delphinapterus
an examination of harpoon gun technology leucas, summering in Cunningham Inlet,
near Pond Inlet, northern Baffin Island. Rep. Northwest Territories. Can, J. Zool. 64:
Int. Whaling Comm. 32: 449-460. 407-415.
FINLEY, K. J., R. A. DAVIS, AND H. B. SILVERMAN. 1986b. The relationship between behavio-
1980. Aspects of the narwhal hunt in the eas- ral activity and underwater vocalization of the
tem Canadian Arctic. Rep. Int. Whaling Comm. white whale, Delphinapterus leucas. Can. J.
30: 459-464. Zool. 64: 2824-2831.
FINLEY, K. J., G. W. MILLER, M. ALLARD, R. A. STEINER, W. W. 1981. Species-specific differen-
DAVIS, AND C. R. EVANS, 1982. The belugas ces in pure tonal whistle vocalizations of five
(Delphinapterus leucas) of northern Québec: western Atlantic dolphin species. Behav. Ecol.
distribution, abundance, stock identity, catch Sociobiol, 9: 241-246.
history and management. Can. Tech. Rep. Fish. TURL, C. W., R. H. PENNER, AND W. L. Au, 1987.
Aquat. Sci. 1123: 1-57. Comparison of target detection capabilities of
FINLEY, K. J., J. HICKIE, AND R. A. DAVIS. 1987. the beluga and bottlenose dolphin. J. Acoust.
Status of the belugas, Delphinapterus leucas, Soc. Am. 82: 1487-1491 ,
in the Beaufort Sea. Can. Field-Nat. 100: WATKINS, W. A., W. E. SCHEVILL, AND C. RA Y.
174-179. 1971. Underwater sounds of Mon odon (nar-
FORD, J. K. B., AND H. D. Fts HER. 1978. Underwa- whal). J. Acoust. Soc. Am , 49: 595-599.
ter acoustic signals of the narwhal (Monodon WHITE, M. J., J. NORRIS, D. LJUNGBLAD, K. BA-
monoceros). Can. J. Zool. 56: 552-560. RON, AND G. DI SCIARA. 1978. Auditory thres-
JOHNSON, C. S., M. W. MCMANUS, AND D. SKAAR. holds of two beluga whales (Delphinapterus
1989. Masked tonal hearing thresholds in the leucas). Hubbs Sea World Res. Inst. Tech.
beluga whale. J. Acoust. Soc, Am , 85: Rep. 78-109: 35 p.
2651-2654.
LGL. 1983. Biological environment of eastern
Lancaster Sound and western Baffin Bay:
components and important processes. Dep. of
Indian and Northern Affairs, Ottawa, Environ-
mental Studies No. 30: 288 p.

117
Echolocation Abilities of the Beluga, Delphinapterus leucas:
A Review and Comparison with the Bottlenose Dolphin,
Tursiops truncatus

Charles W. Turl
Naval Ocean System Cerner, Hawaii Laboratory, Kailua, HA 996734, USA

TuaL, C. W. 1990. Echolocation abilities of the beluga, Delphinapterus leucas: a

review and comparison with the bottlenose dolphin, Tursiops truncatus,
p. 119-128.1n T. G. Smith, D. J. St. Aubin, and J. R. Geraci [ed.] Advances
in research on the beluga whale, Delphinapterus leucas. Can. Bull. Fish.
Aquat. Sci. 224.

Belugas (Delphinapterus leucas) can navigate under the ice, locate food, and
find areas of open water. The results of echolocation experiments suggest that the
echolocation capabilities of belugas may be well suited to the high noise and
reverberation levels that are found in the Arctic. The beluga has a narrow
transmitted beam pattern that would be useful in under-ice navigation. The beluga
can detect targets that are masked by noise and reverberation at low signal levels.
The ability of belugas to detect echoes from a surface-reflected path may be a
strategy they have developed in an A rctic environment.
Some experiments have directly compared the beluga and the bottlenose dolphin
(Tursiops truncatus) in the same echolocation task. The results of these experi-
ments show that the beluga and bottlenose dolphin have different capabilities, and
may have different echolocation strategies. The beluga's bioacoustic abilities are
not fully known, but the information that is available suggests that their echo-
location system is well suited to function in the A rctic environment.

Le béluga (Delphinapterus leucas) peut naviguer sous la glace, localiser de la

nourriture et trouver des étendues d'eau libre de glace. Les résultats d'expérience
d'écholocation portent à croire que le pouvoir d'écholocation du béluga est bien
adapté aux niveaux de bruit et de réverbération que l'on retrouve dans l'Arctique.
Le béluga transmet un faisceau étroit qui serait utile pour naviguer sous la glace.
Il peut détecter des cibles de faibles niveaux qui sont masquées par le bruit et
les réverbérations. Le béluga peut détecter des échos réfléchis de la surface,
et c'est peut-être là une stratégie développée dans le milieu arctique.
Certaines expériences ont comparé directement le béluga et le dauphin à gros nez
(Tursiops truncatus) dans la même activité d'écholocation. Elles démontrent que
le béluga et le dauphin à gros nez ont différentes capacités et peut-être différentes
stratégies d'écholocation. Les capacités bioacoustiques du béluga ne sont pas
parfaitement connues, mais l'information disponible porte à croire que son
système d'écholocation est bien adapté au milieu arctique.

119
 Introduction

Kleinenberg et al. (1969) suggested that belugas (Delphinapterus leucas) have a well
developed "echo-sounding" ability that helps them to orient under the ice and locate areas of
ice-free water. Several experiments have been conducted to quantify the echolocation
abilities of belugas, including underwater hearing sensitivity, critical ratio thresholds, the
transmitted beam pattern and echolocation signal characteristics. The target detection
capabilities of the beluga's sonar system have been measured in a variety of ways including
target detection in masking noise, target detection in reverberation, and maximum range
detection. Measurements of source level of echolocation signals, target strength and noise or
reverberation level were made so that the echo-to-noise ratio (E/N) or the echo-to-reverbe-
ration ratio (E/R) at threshold could be determined. The beluga's ability to discriminate
between two complex planar targets has also been reported. Some echolocation experiments
have directly compared the beluga and the bottlenose dolphin (Tursiops truncatus). The
results of these experiments show some striking differences between these two species. Here
Twill review the results of the beluga sonar experiments and compare experimental data from
Tursiops when possible.
Results

Underwater Hearing and Critical Ratio

White et al. (1978) measured the underwater hearing thresholds of a male and a female
beluga using pure tone signals between 1 and 123 kHz. They reported that the maximum
hearing sensitivity was at 30 kHz for both animals. Other frequencies with thresholds below
40 dB re 1 liPa were 42.5, 60 and 85 kHz. The authors noted that the frequencies of maximum
hearing sensitivity were similar to the peak frequencies of beluga echolocation click trains
reported by Gurevich and Evans (1976) for a shape discrimination experiment. Johnson
(1964) reported that the maximum underwater hearing sensitivity of the bottlenose dolphin
was at 50 kHz (45 dB re 1 ga). Above 50 kHz the threshold increased to about 55 dB re 1 liPa
at 100 kHz.
Johnson et al. (1989) measured the masked auditory thresholds of a beluga between 40 Hz
and 115 kHz. They report that the beluga's critical ratios were about 3 dB lower than critical
ratios reported for the bottlenose dolphin (Moore and Au 1982). A low critical ratio suggests
a good frequency discrimination ability. The term critical ratio is used to quantify the effects
of noise on detection. The critical ratio is the ratio of signal power to noise spectrum level at
masked threshold.

Transmitted Beam Pattern and Echolocation Signals

Au et al. (1987) measured the transmitted echolocation beam pattern of a beluga while the
animal performed a target detection task. The animal was trained to station on a bite plate so
that its transmission beam could be measured in the vertical and horizontal planes using a
hydrophone array. The major axis of the vertical beam was directed approximately 5° above
the plane defined by the animal's teeth. The 3-dB beamwidth, in both the vertical and
horizontal plane, was approximately 6.5°. The directional character of sound emitted by a
projector can be described by its directivity index, defined as the ratio between levels
produced by a directional and omnidirectional transducer emitting equal power. The beluga's
calculated directivity index was 32.1 dB. Au (1980) reported that the 3-dB beamwidth of a
bottlenose dolphin transmitted beam was approximately 11.7° in the vertical plane and 10.7°
in the horizontal plane. The calculated directivity index of Tursiops was 26.5 dB (Au and
Moore 1984).

120
 Au et al. (1985) suggested that Delphinapterus and Tursiops (Au 1980) may change the
peak frequency and peak amplitude of their echolocation signals depending on the ambient
noise spectrum in their environment. Au et al. (1985) measured the echolocation signals of
a beluga in San Diego Bay, California and the same beluga after it had been moved to
Kaneohe Bay, Oahu, Hawaii. In San Diego Bay, the beluga emitted echolocation signals with
peak frequencies from 40 to 60 kHz and 3-dB bandwidths between 15 and 25 kHz. In
Kaneohe Bay, the beluga shifted its signals about an octave higher in frequency with peak
frequencies between 100 and 120 kHz, and 3-dB bandwidths between 30 and 40 kHz. Signal
intensities measured were about 18 dB higher than in San Diego Bay.
Au et al. (1974) measured echolocation signals of two bottlenose dolphin in Kaneohe Bay.
They reported that the animals were using echolocation signals with peak energies at
frequencies between 120 and 130 kHz compared to 30 to 60 kHz reported for Tursiops in
pools. The average peak-to-peak source levels were 30 dB higher in Kaneohe Bay.

Target Detection in Masking Noise

Turl et al. (1987) trained a beluga and a bottlenose dolphin to detect targets in masking
noise. Targets were stainless-steel, water-filled spheres 7.62 and 22.86 cm in diameter.
Target ranges of 16.5 and 40m were used with the 7.62-cm sphere and 80m with the 22.86-
cm sphere. Masking noise with a flat spectrum from 40 to 160 kHz was projected from a
spherical transducer located in front of the animal hoop station in line with the target. Target
detection performance was determined as a function masking noise level at each target range.
Each animal 's target detection performance, as a function of the masking noise level for both
targets and three distances, is shown in Fig. 1. The beluga matched the dolphin's performance
in noise levels 8-13 dB higher.

• BELUGA WHALE
• BOTTLENOSE DOLPHIN
100

90
CORRECTRES PONSE ( %)

80

70

60

R = 80 M
50 22.86 CM
SPHERE I IR=
7.62 CM R = 16.5 M
SPHERE 7.62 CM
SPHERE
4r0 510 610 710
50 60 70 80
60 70 80 90

MASKING NOISE SPECTRUM LEVEL (dB re 1 pPa 2 /Elz)

FIG. 1. (a) Beluga and bottlenose dolphin's performance data as a function of

masking noise levels at 16.5, 40 and 80 m (after Turl et al. 1987).

121
 100

90

Lu
• 80
o
a. 0 16.5 M
w 70
cc El 40.0 M
IX 80.0 M
W 60
cc • AU & PENNER (1981)
cc
0
C-) 50
BELUGA BOTTLENOSE
DOLPHIN
1
-10 -5 0 5 10 15 20

10 LOG [Ee /No ] mAx (dB)

FIG. 2. Beluga and bottlenose dolphin's performance data at three distances

replotted as a function of the maximum echo signal-to-noise ratio (from Turl et al.
1987).

The beluga's and dolphin's performance data in Fig. 1 are replotted in Fig. 2 as a function
of echo-to-noise ratio. The average value of the maximum source energy flux density (Eu) per
trial was used in the calculation. Such a technique will provide the upper limit of the Ee/No,
that is, the maximum Ee/1■10 available to the animal. The data of Au and Penner (1981) for two
Tursiops are also included in Fig. 2. At the 75% correct response threshold, the (Ee11\10) was
approximately 1.0 dB for the beluga at the three target distances. For the bottlenose dolphin
(Turl et al. 1987), (Ee/1\1 0 ) max was approximately 10 dB, consistent with the data of Au and
Penner (1981) for Tursiops.

Target Detection in Reverberation

Turl and coworkers (unpubl. data) measured a beluga's ability to detect targets placed in
front of a clutter screen. The clutter screen consisted of 400 5.1 cm diameter cork balls spaced
15.2 cm apart in a 20 X 20 array. The clutter screen was 8m from the animal's hoop station.
Targets were hollow stainless steel cylinders (3.2-cm diameter and 0.38-cm wall thickness),
14.0, 10.0, 7.0, 5.0 and 3.0 cm long. Using cylinders with the same wall thickness and
diameter, but of different lengths, the amplitude of the target echoes could be varied without
changing the echo structure.
Backscatter measurements of the clutter screen and targets were made using the procedures
given in Au and Snyder (1980). The backscatter measurements of the clutter screen were
made with the transducer located so that its 3-dB beamwidth covered approximately the same
area of the clutter screen that the beluga's beam (Au et al. 1987) would at a distance of 8 m.
The clutter screen was used in two positions: perpendicular to the beluga (90° grazing angle)
and rotated 22° (68° grazing angle). Values of the target strength based on energy (TSe) and
on maximum peak-to-peak amplitudes (TS _ ) are shown in Table 1 for the clutter screen and
the five targets. To include the entire echoP, the TSe for the clutter screen was measured for
1 ms. At 68° grazing angle, the backscatter from the clutter screen was 2 dB lower than at 90°
grazing angle.

122
TABLE I. Results of the target strength measurements of the clutter screen and cylinder targets. The
E/R PP based on the peak-to-peak amplitudes and on the energy of the incident and echo signals are
shown for clutter screen grazing angle of 90°. E/RE values were measured over a 1 ms time interval.
Target Length TS TSE E/R E/R E
(CM) (dE ) P (dB) (d BP)P (dB)
Screen
90° -25.8±1.7 -19.9±1.0
68° -27.8 ±1.3 -20.3 ±0.6

Targets
1 13.0 -17.6±.2 -16.8±,2 +8.2 +3.1
2 10.0 -19.3 ±.2 -18.3 ±.2 +6.3 +1.5
3 7.0 -22.0±.2 -21.1 ±.2 +3.8 -1.2
4 5.0 -25.1 ±.3 -23.9 ±.5 +0.7 -3.0
5 3.0 -29.3 ±.3 -27.6 ±.9 -3.6 -7.7

The beluga's detection results when the targets were located within the plane of the clutter
screen are shown in Figure 3 as a function of E/R. The E/R based on energy represents the
lowest estimate of the E/R since the entire clutter screen echo occurred within the 1 ms
integration time used. The E/R based on the peak-to-peak amplitude represents the highest
estimate of E/R. The results indicate that at the 50% detection threshold the echo-to-
reverberation ratio was -1.0 dB, based on the peak-to-peak measurements, and -5.1 dB based
on the energy measurement.
Au and Turl (1983) trained a bottlenose dolphin to detect different length cylinders in front
of a clutter screen similar to the experiment for the beluga. The results of this experiment (Fig.
3) show that when the targets were located within the plane of the clutter screen, the
bottlenose dolphin's 50% correct detection was 2.6 dB, based on the peak-to-peak measure-
ments, and 0.25 dB based on the energy measurement. The results of these two experiments
suggest the beluga can detect targets in 3.6 to 5.4 dB more reverberation than Tursiops.
100-.
CORR ECTDETECTION(%)

80

60

40 Tursiops
E/R PP (dB)
20 Beluga
E/R e (dB)

o
-10 0 10
E/R (d B)
FIG. 3. Beluga and bottlenose dolphin (after Au and Turl 1983) performance data
as a function of the echo-to-reverberation ratio. In this case, the targets were within
the plane of the clutter screen. The squares indicate echo-to-reverberation ratios
based on peak-to-peak amplitudes and the triangles indicate echo-to-reverberation
ratios based on energy.

123
Maximum Target Detection
Turl and Penner (unpubl. data) trained a beluga to detect a 7.62 and a 15.24 cm diameter
stainless steel, water-filled sphere between 40 and 120 ni. The target strength of the 7.62 and
15.24 cm spheres was -30.8 and -21.4 dB, respectively. At the 50% correct detection
threshold, the beluga's performance for the 7.62 and 15.24 cm spheres is 116 and 117 m,
respectively. The target strength difference between the 7.62 and 15.24 cm spheres is
approximately +9 dB and yet the difference of the detection threshold is 1 m. At target
distance <115 ni, the beluga's performance was greater than 90% correct detection for both
spheres. At 120m, the beluga's performance was at chance for the 7.62 and 15.24 cm spheres.
Au and Snyder (1980) reported that a bottlenose dolphin's maximum range detection with
a 7.62 cm sphere on the same target range was 113m. The results of these two experiments
indicate that the beluga and Tursiops have about the same detection range.

Echolocation Strategies
Turl and Penner (1989) reported that a beluga emitted different patterns of echolocation
click trains than the bottlenose dolphin while performing the same echolocation task. The
beluga emitted three different patterns of echolocation clicks. A Pattern I (Fig. 4A) click train
started with low amplitude clicks, followed by packets of clicks. A packet contained several
clicks with interclick intervals less than the two-way travel time (TWT) to the target; the
interpacket intervals were greater than the TWT. A Pattern II (Fig. 4B) click train consisted
of a combination of individual clicks, with some intervals less than and some greater than the
TWT. This pattern did not contain packets. Pattern III (Fig. 4C) click trains consisted of
individual clicks with interclick intervals less than the TWT.
Pattern land II click trains always started with an initial series of clicks that were less than
the TWT. This implies the beluga transmits signals and receives echoes simultaneously
which may provide the beluga with information about target presence or absence, but
probably does not provide information about target range. In Pattern I click trains the
interpacket interval is longer than the total packet duration and greater than the TWT. The
beluga may be processing all echoes in a packet before the next echo packet returns to the
animal. Beyond 80 ni target distance, the beluga emitted Pattern I and Pattern II click trains.
Pattern III click trains had a constant repetition rate of 40 to 57 ms and were emitted by the
beluga at target distances of 40 and 60m. This suggests the beluga may be processing echoes
between the transmission of echolocation signals and that this repetition rate might optimize
the analysis of information.
The bottlenose dolphin typically emits an echolocation signal and waits until the echo
returns before emitting the next signal. The bottlenose dolphin changes the pulse repetition
rate as the target distance changes so that the signal echo is during the interclick interval
(Morozov et al. 1972; Penner and Kadane 1980).
Turl et al. (1987) reported that a beluga and a bottlenose dolphin emit different echolocation
click patterns in the same masking noise experiment. The beluga and bottlenose dolphin were
the same animals that were the subjects used in Turl and Penner (1989). The targets were at
three distances from the animal (16.5, 40 and 80 m). At 16.5 and 40m, the average interclick
intervals of beluga click trains were about the same as the TWT (21 to 52 ms), whereas at
target distance of 80 m, the beluga average interclick interval was less than the TWT
ms). The bottlenose dolphin's interclick interval was greater than the TWT at the three
distances.

Target Discrimination
Gurevich and Evans (1976) trained a beluga to discriminate between two complex planar
targets. The targets were three-step pyramids constructed of polyvinyl chloride sheets each

124
 (A)

250 - INSVAL lIfIERPACKET

SERIES INTERVAL PACKETS
INTERCLIC(INTERVA L (MSEC)

200 1 1'4- ne

150 -
TWO-WAY TRAVEL ITIAE

100
1----
50

1 ■ . 1
10 zo 30 40 ' 50 1.5 3.0
CLICK ORDER
SECONDS

(B)
250
INTERCLICKINTERVAL (MSEC)

200

150 TV/CYWAY TRAVEL TIME

100

50 1111 II 11
.5 3.0
20 30
SECONDS
CLICK ORDER

250 (C)

200

5 150 TWO-WAY TRAVEL TIME

5 1.0

50

1.5 3.0
10 20 30 40 50
SECONDS
CLICK ORDER

FIG. 4. Examples of the three patterns of beluga echolocation click trains for targets at 100 m are shown
in click order (left) and in the time domain (right). A beluga pattern I click train [see (A)] plotted in the
time domain showing the relationship of the initial series, interpacket interval, and packets (after Turl
and Penner 1989). A Pattern II click train is shown in 4B and Pattern III in 4C.

13 mm thick. For the standard target, the first step was 10 X 10 cm, the second step was 7
X 7 cm and the third step was 3 X 3 cm. A selection of the standard target was considered
a correct response. The comparison targets (incorrect response) differed from the standard
targets in the size (surface area) of the third step. In a comparison of a third step of 2.9 X
2.9 cm versus the standard of 3.0 X 3.0 cm, the beluga's average correct response was 81%.
When presented with two standard or identical targets, the beluga's response dropped to 50%.
A bottlenose dolphin tested with similar targets was at chance performance on a comparison
of 2.7 X 2.7 cm versus 3.0 X 3.0 cm.

125
Target Detection Using Surface-Reflected Path

Penner et al. (1986) reported that during the initial stages of a masking noise experiment
(Turl et al. 1987) with a beluga and bottlenose dolphin, the beluga's performance did not
decrease with increasing noise. The beluga's detection performance was 85% even when the
noise level was 15 dB greater than the noise threshold level of the dolphin. Since the
interfering noise came from a point source, Penner et al. reasoned that the beluga might
realize a gain in the signal-to-noise ratio by projecting and receiving signals off the surface
as depicted in Fig. 5. The beluga's signals were measured along the whale-target axis and
these were similar to "off-axis" signal reported for the bottlenose dolphin (Au 1980). Mea-
surements using the initial test geometry (Fig. 6) showed that maximum amplitude of the
echolocation signals was approximately 7° above the animal-hydrophone-target mid-line
axis. Two methods were used to test the surface-reflected path hypothesis. First, the noise
projector was raised from 1.0 to 0.4 ni, placing it directly in the surface-reflected path; two
sessions were conducted with this geometry. Second, a screen was used to block the surface
path. The screen was a 60- X 91-cm doormat lined on one side with 0.3-cm-thick corprene
and on the other side with a loose weave .fiber packing material (commonly called horsehair).
The screen was placed 3 m from the hoop station. The screen extended to a depth of 0.6 m
below the surface of the water. Twenty test sessions were conducted with the screen in place.

NOISE SOURCE

FIG. 5. Drawing showing how the beluga might achieve a higher signal-to-noise ratio using a surface-
reflected propagation path in the original test condition (from Penner et al. 1986).

16.5 M

4M
3M
SURFACE
-t-
1M _ _ _ ___
-1 _
CLICK NOISE
NOISE TARGET
HYDROPHONE HYDR OPHONE
ACOUSTIC
SCREEN

FIG. 6. Side view of the beluga, the noise projector, and the test geometry showing a direct and surface-
reflected path (after Penner et al. 1986).

126
 In the first session, after the noise projector was raised to a depth of 0.4 m, the beluga's
correct detection decreased as the noise level increased. However, in the second session, the
beluga's detection performance increased to the level previously measured with the noise
projector at a depth of 1 m. The signal measuring hydrophone at 1 m measured higher
amplitude signals for the second session. Penner et al. said that the beluga had abandoned the
surface-reflected path and was using the direct path to maximize performance.
The beluga's detection performance was altered when the screen was used to block the
surface-reflected propagation path and the noise projector was at 1.0 m. The beluga's
detection performance sharply decreased as the noise level increased from 79 to 91 dB.
Without the screen, the beluga's performance was relatively constant (>85% correct
response) between these two levels. Penner et al. concluded that the initial test geometry used
in this experiment enabled the beluga to echolocate via a surface-reflected propagation path.

Discussion

Comparing the results of the experiments for the beluga and the bottlenose dolphin
suggests that there are differences in the sonar characteristics and target detection sensitivity
of these two species.
Kleinenberg et al. (1969) suggested that belugas locate "areas of ice-free water, finding
directions by active echo sounding of water lapping in the polynya." The edge of an ice sheet
is a source of noise caused by wave impact upon the ice floe (Urick 1984). Belugas have good
directional underwater hearing. Since maximum range detection would probably be limited
by either noise or reverberation it would seem more likely that belugas might use passive
orientation to detect the direction to open water.
Au et al. (1987) suggested that the beluga's narrow transmitted beam should provide good
spatial localization and target resolution capabilities. They further say that "by scanning over
the extent of an elongated target with a narrow beam, the animal should be able to
discriminate various echo highlight producing reflectors. Such a capability would be useful
in under-ice navigation and to find air pockets under an ice pack." Beluga's can discriminate
between targets that differ by 1 cm in test conditions; however, the cues used by this animal
to differentiate targets is not known.
The results of experiments show that belugas differ from bottlenose dolphins in their ability
to (1) detect targets in noise, (2) detect targets in reverberation, and (3) use surface-reflected
signals. Belugas also seem to differ in hearing ability and pulse production patterns. These
differences might be the result of beluga's adaptation to the Arctic acoustic environment. The
results from acoustic tests show a beluga can detect targets in masking noise about 8-13 dB
higher than the bottlenose dolphin. A beluga can detect targets in 3.6 — 5.4 dB more
reverberation than Tursiops in test conditions in which the targets are normal to the animal's
echolocation beam (e.g., looking up into the ice). The beluga's ability to receive and use
surface-reflected echoes may be useful when passing through areas of extensive ice
coverage. The beluga's critical ratio is lower than that for bottlenose dolphin by approxima-
tely 3 dB. The beluga emits packets of clicks, bottlenose dolphins do not.
Experimental results suggest that the beluga's echolocation system is well suited to
function in the Arctic environment. The ability of the beluga to detect low amplitude echoes
in high levels of masking noise and reverberation may reflect an adaptation to the high
ambient noise and reverberation levels of acoustic interference found in the Arctic. Further
investigation, including measuring the receiving beam pattern and signal processing charac-
teristics of the beluga, is required in order to isolate the mechanisms which account for the
belugas's increased detection sensitivity. Belugas may have different signal processing
capabilities than those measured for other species of odontocetes. Additional information
concerning the temporal resolving capability of the beluga could help isolate mechanisms
used by the beluga that might be relevant in this type of click emission strategy.

127
 Acknowledgments

I would like to thank the following co-workers at the Naval Ocean System Center (Hawaii
Laboratory) for their assistance in the preparation of this paper, Dr. Jeanette Thomas for her
helpful comments and P.W.B. Moore, William F. Friedl, Dr. Whitlow Au and Dr. Paul E.
Nachtigall for their review of the manuscript. The many discussions with Ralph H. Penner
are greatly appreciated.

References

Au, W. W. L., R. W. FLOYD, R. H. PENNER, AND beluga whale. J. Acoust. Soc. Am. 85:
E. MURCHISON. 1974. Measurement of 2651-2654.
echolocation signals of the Atlantic bottlenose KLEINENBERG, S. E., A. V. YABLOKOV, B. M.
dolphin, Tursiops truncatus Montagu, in open BEL'Kovicit, AND M. N. TARASEVICH. 1964.
waters. J. Acoust. Soc. Am. 56: 1280-1290. (Translation 1969). (Delphinapterus leucas).
Au, W. W. L., AND K. J. SNYDER. 1980. Long- Investigation of the species. Israel Program for
range target detection in open waters by an Scientific Translations. Jerusalem 376 p.
echolocating Atlantic bottlenose dolphin MOORE, P. W. B., AND W. W. L. Au. 1982.
(Tursiops truncatus). J. Acoust. Soc. Am. 68: Masked pure-tone thresholds of the bottlenose
1077-1084. dolphin (Tursiops truncatus). J. Acoust. Soc.
Au, W. W. L. 1980. Echolocation signals of the Am. Suppl. 1 72, S42.
Atlantic bottlenose dolphin (Tursiops trunca- MOROZOV, V. P., A. I. AKOPIAN, V. I. BURD1N, K.
tus) in open waters, p. 251-282. In R. G. A. ZAYTSEVA, AND Yu. A. SOKOVYKH. 1972.
Busnel and J. F. Fish [ed.] Animal Sonar Repetition rate of ranging signals of dolphins
Systems. Plenum Press, New York, NY. as a function of distance to target. Biofisika 17:
Au, W. W. L, AND R. H. PENNER. 1981. Target 49-55.
detection in noise by Atlantic bottlenose dol- PENNER, R. H., AND J. KADANE. 1980. Biosonar
phins. J. Acoust. Soc. Am. 70: 687-693. interpulse interval as an indicator of attending
Au, W. W. L., AND C. W. TURL. 1983. Target distance in Tursiops truncatus. J. Acoust. Soc.
detection in reverberation by an echolocating Am., Suppl. 1 S68.
Atlantic bottlenose dolphin (Tursiops trunca - PENNER, R. H., C. W. TURL, AND W. W. L. Au.
tus). J. Acoust. Soc. Am. 73: 1676-1681. 1986. Target detection by the beluga using
Au, W. W. L., AND P. W. B. MOORE. 1984. surface reflected path. J. Acoust. Soc. Am. 80:
Receiving beam pattern and directivity indices 1842-1843.
of the Atlantic bottlenose dolphin Tursiops TURL, C. W., R. H. PENNER, AND W. W. L. Au.
truncatus. J. Acoust. Soc. Am. 75: 255-262. 1987. Comparison of target detection capabi-
Au, W. W. L., D. A. CARDER, R. H. PENNER, AND lities of the beluga and bottlenose dolphin. J.
B. L. SCRONCE. 1985. Demonstration of adap- Acoust. Soc. Amer. 82: 1487-1491.
tation in beluga whale echolocation signals. J. TuRL, C. W., AND R. H. PENNER. 1989. Differen-
Acoust. Soc. Am. 77: 726-730. ces in echolocation click emission of the belu-
Au, W. W. L., R. H. PENNER, AND C. W. TURL. ga (Delphinapterus leucas) and the bottlenose
1987. Propagation of beluga echolocation dolphin (Tursiops truncatus). J. Acoust. Soc.
signals. J. Acoust. Soc. Am. 82: 807-813. Am. 86: 497-502.
GUREVICH, V. S., AND W. E. EVANS. 1976. Echo- URICK, R. J. 1984. Ambient noise in the sea.
location discrimination of complex planar Undersea Warfare Technology Office, Wa-
targets by the beluga whale (Delphinapterus shington, DC.
leucas). J. Acoust. Soc. Am. Suppl. 1 60, S5. WHITE, M. J., J. NORRIS, D. LJUNGBLAD, K. BARON,
JOHNSON, C. S. 1964. Sound detection thresholds AND G. Di SCIARA. 1978. Auditory thresholds
in marine mammals, p. 247-255. In W. N. of two beluga whales Delphinapterus leucas.
Tavolga [cd.] Marine bioacoustics Vol. 2. HSWRI Tech. Rep. No. 78-109, Sea World
Pergamon Press, New York, NY, Research Institute, San Diego, CA.
JOHNSON, C. S., M. W. MCMAN US, AND D. SK AAR.
1989. Masked tonal hearing thresholds in the

128
 Integumentary Heat Loss and Blubber Distribution in the
Beluga, Delphinapterus leucas, with Comparisons to the Narwhal,
Monodon monoceros

D.W. Doidge
Department of Renewable Resources, Macdonald College of McGill University
21,111 Bord du Lac, Ste-Anne-de-Bellevue, Quebec, Canada H9X ICO

DOIDGE, D. W. 1990. Integumentary heat loss and blubber distribution in the

beluga, Delphinapterus leucas, with comparisons to the narwhal, Mono-
don monoceros, p. 129-140. In T. G. Smith, D. J. St. Aubin, and J. R.
Geraci [ed.] Advances in research on the beluga whale, Delphinapterus
leucas. Can. Bull, Fish. Aquat. Sci. 224.

The minimum metabolic rate necessary for thermostasis was estimated from in-
tegumentary heat loss using skin and blubber thicknesses, thermal conductivities
and body girths from beluga,Delphinapterusleucas(n=22), and narwhal, Monodon
monoceros (n=5). Blubber is not uniformly distributed over the body, but forms a
thickened saddle posterior to the dorsal ridge, and smooths the outline of tendons
in the caudal region. The thinnest blubber and the thickest skin occurred on a
neonate. Thickened skin serves as a thermal buffer at birth. Standard measures of
girth at the axilla or umbilicus are highly correlated with blubber volume (R2=0.98),
indicating girth is an excellent index of body condition.
Thermal conductivity of beluga blubber was 0.102 ± 0.011 Wom".°C -1 (n=6);
skin was 0.249 ± 0.019 Wom - I.°C-1 (n=7). For eastern Hudson Bay belugas of body
lengths 149 to 401 cm, the thermostatic energy requirement (Watts) in 15°C water
equals 0.238 body length 1 • 136. Beluga and narwhal are equally insulated, but only
beluga exhibit seasonal occupation of warmer estuarine waters.

Le taux métabolique minimum nécessaire pour maintenir la thermoneutralité a

été estimé à partir de la perte de chaleur intégumentaire en se servant des épaisseurs
de peau et de graisse, des conductibilités caloriques et des tours de taille des
marsouins blancs, Delphinapterus leucas (n=22), et narval, Monodon monoceros
(n = 5). La graisse n'est pas distribuée uniformément sur la superficie du corps,
mais forme une selle plus épaisse postérieure à l'épine dorsale, et égalise le contour
des tendons dans la région caudale. La graisse la plus mince et la peau la plus
épaisse se trouvent chez les néo-natals. Une peau épaissie sert de tampon thermi-
que à la naissance. Des mesures standards de tour de taille à la hauteur axillaire ou
ombélicale sont fortement correlées avec le volume de graisse (R 2 = 0.98)
indiquant que le tour de taille est un excellent index de condition corporelle.
La conductibilité calorique de la graisse de baleine était 0.102 ± 0.011 Wom"•
°Cl (n=6); la peau était 0.249 ± 0.019 Worn".°C -1 (n=7). Pour les marsouins blancs
de l'est de la baie d'Hudson, d'une longueur de 149 à 401 cm, la demande d'énergie
thermostatique (Watts) dans l'eau de 15°C égale 0.238 de la longueur du corps '• 136 .
Le marsouin blanc et le narval sont également isolés mais seul le marsouin blanc
démontre une occupation saisonnière d'estuaires aux eaux plus chaudes.

129
 Introduction

Estimates of metabolic rates of marine mammals, especially cetaceans, have supported the
impression that these animals have high metabolic rates relative to terrestrial animals of
comparable size (Brodie 1975). Yet these estimates remain controversial since few published
studies meet the criteria for measuring basal metabolism (Lavigne et al. 1986). These criteria,
which set standard conditions for comparisons with other mammals, require that metabolic
rates be measured on mature, resting animals in a postabsorptive state at a metabolically
indifferent temperature (Kleiber 1975). Since it is often difficult or impossible to meet these
criteria when studying cetacean energetics, an alternative (but little used) approach is to
determine the thermostatic energy requirement. Kleiber (1975) defined this as the heat
production needed to maintain a constant mammalian body temperature above that of the
environment. In marine mammals, the thermostatic energy requirement equals the heat lost
by conduction through the integument when the loss via evaporative cooling in the lungs
(ca. 5%, Kanwisher and Sundnes 1966) is ignored.
The properties and function of whale blubber were explored by Parry (1949) in a study of
the structure and heat flow through fin whale (Balaenoptera physalus) and harbour porpoise
(Phocoena phocoena) blubber. He suggests blubber functions as an energy store and as in-
sulation. Others, notably Kanwisher and Sundnes (1966) and Brodie (1975), have estimated
energy requirements of cetaceans using surface area and blubber thickness. In these studies,
the term blubber is used in the commercial whaling sense which lumps blubber and skin as
one; or considers only fat thickness and not skin in the heat loss equations. However, the
thermal properties of skin and blubber are quite different, their thermal conductivities
differing by about a factor of 2 (See Results section, p. 133.). Given the thick skin of
cetaceans, especially the polar adapted species (Geraci et al. 1986), I have considered
separately the roles of skin and blubber on integumentary heat loss.
The distribution of blubber over the body affects heat loss, streamlining and buoyancy.
When coupled with measurements of body length and girth, blubber thickness data can be
used to estimate a whale's blubber volume, and if blubber density is known, blubber mass.
The correlation between standard field measurements of blubber thickness and body girth
(Am. Soc. Mammal. 1961) and blubber volume (or mass) indicates how suitable these
measures are as indices of body condition.
I examined the topography and thermal conductivity of blubber and skin and evaluated
integumentary heat loss in beluga (Delphinapterus leucas) and narwhal (Monodon monoce-
ros), two related species with distinct ecological niches. The similarities observed in
integumentary heat loss challenge the view that niche separation depends on differences in
thermal adaptation.

Methods

Specimens

Standard measurements (Am. Soc. Mammal. 1961), and detailed skin, blubber and girth
data, were obtained from 16 beluga specimens of various sizes and reproductive status in the
summers of 1984 and 1985 on the coast of eastern Hudson Bay, primarily at the Nastapoka
River (56°55'N, 76°33W). Two neonate and two adult belugas were examined at Cunnin-
gham Inlet, N.W.T. (74°07N, 93°50W') in July, 1977; one adult at Winton Bay, N.W.T.
(64°28N, 64°42'W) in August, 1978; and one neonate at Wakeham Bay, Quebec (61°36N,
71°56'W) in November, 1985. Two near-term fetuses, a neonate, a calf and an adult narwhal
from the north Baffin Island harvest were examined in summer during the mid-1970's.

130
Blubber and Skin Thickness, Volume and Mass

Whales from eastern Hudson Bay were measured by making 7 transverse cuts equally
spaced along the total length of each animal (Fig. 1). Half girth was recorded to the nearest
centimetre at each interval. Skin and blubber thickness were measured in situ to the nearest
millimetre at 7 equally spaced points dorsal-ventrally at each interval on one side of the
animal, except in the genital slit region where the most ventral measurement was not made.
Sampling protocol for the other specimens depended on time available for blubber and skin
measurements and ranged from 5 to 13 transverse intervals. These data were linearly
interpolated to yield 7 by 7 evenly spaced data matrices as described above. Only the un-in-
terpolated data from eastern Hudson Bay beluga were used to assess blubber and skin distri-
bution. Both the un-interpolated and interpolated data of narwhal and beluga were used in the
estimation of integumentary heat loss.
Whales were considered circular in cross-section for all computations of volumes and heat
loss. The cross-sectional area of skin, blubber and core at each interval was calculated using
the interval girth and mean skin and blubber thickness. Skin, blubber and core volumes were
then estimated using Simpson 's Rule (Boyce and DiPrima 1969) to evaluate the numerical
integral of the respective cross-sections over the animal's length. Blubber and skin density
were determined by dividing the weight of a sample by the volume it displaces in water.
The General Linear Models procedure of PC-SAS (SAS Inc, Cary, NC, USA) fitted least
square regressions to the length, girth and blubber data to predict blubber, core and skin
volume. The distribution patterns of skin and blubber were established by determining areas
of similar (relative) thickness using Friedman's test (ANOVA based on ranked data, See
Conover 1981).

Heat Loss

The thermal conductivity of beluga blubber and skin was estimated using a heat flux plate
(Model S387, Geneq, Montreal) to measure the heat flow through samples of known
thickness and temperature gradient. The heat source, a 4 L beaker of water, was maintained

1 2 3 4 5 6 7

INTERVAL NO.
FIG. 1. Skin and blubber sampling points spaced equally, dorso-ventrally at each of 7 equally spaced
girth intervals. Interval width is 1/8 of standard body length.

131
at 45°C by a laboratory heater pump. Water at an ambient temperature of 5°C circulated from
a 10 L water bath through a 800 cm' metal box which acted as the heat sink. Temperature
gradient and heat flux were recorded at 10 min intervals. After allowing a 20 min period for
the temperature profile to stabilize, the thermal conductivity of each sample was calculated
using the thickness of the sample at the end of the test and the mean (n > 10) of temperature
gradient and heat flux records. No correction was made for reduction in thermal efficiency
due to blood flow in live tissue (Kanwisher and Sundnes 1966).
The equations used to determine heat loss are given in Table 1. Equation (1) states Fourier's
law of heat conduction between the surface of a body and its environment (Kleiber 1975). Due
to the law of conservation of energy, the total heat loss across the skin equals the loss across
the blubber (Eq. 2). Using the above relationships, the temperature at the junction of the skin
and blubber layers can be determined (Eq. 3). This simplifies the heat transfer problem from
that of two layers (blubber and skin) to that of one layer (blubber or skin). The heat flux at
a given interval is governed by Eq. (4). Total heat loss is calculated using Simpson's rule
(Boyce and DiPrima 1969) to evaluate the numerical integral of these fluxes over the length
of the animal (Eq. 5).

TABLE 1. Equations governing integumentary heat loss.

Fourier's law of heat conduction
(1) Q = KA -STD - '

Heat loss through the blubber equals the loss through the skin
(2)

Solving for temperature at the blubber - skin junction

(3) T = G K BT + G bb SI
G;K,B + Gh•KhS

Unit flux per cm length at intervali

(4) Fi = Kii Ghi (Te-T, 1„)
Bi

Integumentary heat loss

(5) H= j.tail F
snout
where
Rate of heat flow per unit time
Thermal conductivity of insulating layer
A Surface area
87' Temperature di fference across insulating layer
D Thickness of insulating layer
Skin thickness
Blubber thickness
G Body girth

Subscripts
blubber
inner core
skin
sb skin-blubber junction (the base of the dermis)
interval number

132
 Results

Integumentary Heat Loss

The thermal conductivity of beluga blubber (summer samples) was 0.102 ± 0.011 W.m - I•
°C-1 (n = 6) and skin was 0.249 ± 0.019 W.m - '.°C- ' (n = 7). The minimum integumentary heat
loss for thermostasis in eastern Hudson Bay belugas (body length 149-401 cm) ranged from
70 to 207 W, for an ambient water of 15°C (Table 2). In the same temperature regime, high
arctic belugas and narwhals had similar heat losses to those of eastern Hudson Bay belugas
(Fig 2).
Heat loss is directly proportional to surface area (constantlengthogirth) and inversely
proportional to the thickness of insulation (in this case blubber or skin). However, length
alone predicts heat loss well (R 2 0.95). This relationship in eastern Hudson Bay belugas is:

H -= 0.238Length 11357 (n=16, P<0.001)

where H is heat loss (W) and length is straight-line distance (cm) from the tip of the snout to
the fluke notch.

500

400
HEAT LOSS(W)

300

200

100

150 200 250 300 350 400

BODY LENGTH (CM)

FIG. 2. Comparison of theoretical basal heat production based on body mass and integumentary heat
loss in 15°C water with core temperature of 35°C. Beluga: Open triangle — eastern Hudson Bay;
Asterisk— Barrow Strait, NWT; Solid triangle—Hudson Strait, N. Quebec; Open square—E. Baffin.
Narwhal: Open circle — N. Baffin. Basal metabolic heat production based on total mass (solid line) 70
Mass.75 ; Basal metabolic heat production based on mass excluding blubber (dashed line) 70 (0.6Mass). 75
where W(kg) = 0.000182 L(cm) 2." (Doidge 1990).

133
TABLE 2. Body morphometrics, integumentary heat loss in 15°C water, body surface area and status of
Monodontids from various sites.
Standard Max. Lateral' Heat Surface Statue
length girth blubber skin loss' aread
cm cm mm mm W 111 2

Beluga
E. Hudson Bay
Male 149 100 24 13 70 1.0 N
208 156 55 11 95 2.4 —
278 172 40 13 161 3.4 —
293 182 37 12 159 3.7 —
401 210 52 14 207 5.9 —
Female 298 194 52 12 148 3 •7 P
298 184 50 13 151 3.9 L
299 176 47 12 145 3.6 R
299 164 39 14 166 3.5 L
308 196 50 13 164 4.1 L
319 200 57 12 162 4 •4 P
320 182 48 14 170 4.3 L
320 196 51 11 170 4.4 L
324 176 43 13 176 4.1 L
337 184 44 12 180 4.4 L
346 207 63 14 164 4.8 B
Cunningham Inlet
Male 179 106 24 12 97 1.3 N
180 113 26 11 94 1.4 N
399 252 58 12 224 6.6 —
Female 395 242 66 12 186 6.3 L
Wakeham Bay
Male' 145 98 13 25 92 1.1 N
Winton Bay
Male 407 300 82 10 194 8.0 —

Narwhal
N. Baffin
Male 164 110 34 5 70 1.2 F
195 138 38 11 89 1.8 N
Female 223 142 45 11 93 2.1 —
353 210 41 11 188 4.6 L
Sex ? 182 114 30 10 85 1.4 F
" Measured mid-length.
b F fetus, N neonate, P pregnant, R resting, L lactating.

When maintaining body core at 35°C in 15°C water.

d Not including flippers and flukes.
' Sampled in November; all other whales sampled between late June and late August.

134
 RELATIVE THICKNESS -Pie-
FIG.3. Blubber distribution in belugas showing areas of significantly different relative thickness basec.1
on Friedman's test (n = 16).

Blubber Thickness

The general pattern of blubber distribution (Fig. 3) of eastern Hudson Bay belugas appeared
similar regardless of sex or reproductive status. The thickest blubber occurred dorsally from
mid-length posteriorly to the caudal peduncle, broadening as a slight lateral saddle. A smaller
pad of equivalent thickness lay ventrally at the caudal peduncle. The next thickest blubber
occurred laterally just behind the flipper. Behind this "flipper pad" and ventral to the thick
blubber of the dorsal ridge lay a wide region of intermediate thickness. Blubber was thinnest
in the dorso-lateral region of the neck and the lateral area of the caudal peduncle. No
measurements were made in the region of the melon or jaw.

Skin Thickness

Figure 4 shows the variation in relative skin thickness (epidermis and dermis) over the body
surface. Skin is thinnest at the caudal peduncle and in the region of the axilla, followed by
the neck (Intervals 7, 2 & 1, respectively). The thickest pad covers the anterior end of the

RELATIVE THICKNESS -lb-

FIG.4. Skin distribution in belugas showing areas of significantly different relative thickness based on
Friedman's test (n = 16).

135
dorsal ridge extending as a slight lateral saddle. Skin on the animal's side is slightly thicker
than intermediate thickness. Skin on the dorsal ridge posterior to mid-length is intermediate
in thickness. The dorsal ridge is more a product of steeply sloping sides of the animal rather
than of thickened skin. •

Body Composition and Condition

Skin, blubber and core volume can be predicted from body girth or body length. Girths at
intervals 1 to 5 (see Fig. 1) and body length each explain at least 95% of the variance in volume
when cubed and multiplied by the appropriate coefficient (Table 3). The density of blubber
(hypodermis) was 0.92 + 0.09 g cm -3 (n = 6) and that of skin 1.07 + 0.02 g cm -3
(n = 4). Direct measurements of the density of the body core were not made.
Discussion

The aquatic existence of homeotherms is generally thought to be energetically demanding

because of the high thermal conduction of water compared to air (Scholander et al. 1950).
This led many early investigators to assign high metabolic rates to marine mammals (Brodie
1975). Since the high thermal losses caused by an aquatic existence can be reduced by a
thickened blubber layer, metabolic rate need not be increased above the expected rate of
terrestrial mammals. If the warmest waters inhabited by whales are within their thermal
neutral zone, integumentary heat loss should represent the lower boundary of metabolic rate.
Heat loss through the integument depends, in part, on the mass transfer of heat via blood
flow. The heat transfer model assumes a negligible amount of heat is lost through flippers and
flukes. However, these appendages are not at environmental temperature so some additional
heat is lost even when blood flow is restricted. Bel'kovich (1961) reported belugas had the
ability to regulate blood flow through the integument using well-developed muscles in the
arterial walls, but the effect on thermal conductivity has not been quantified in belugas. The
thermal conductivity of live blubber is 35 - 65% greater than dead tissue (Kanwisher and
Sundnes 1966; Irving and Hart 1957) so the heat losses reported here for beluga are
conservative.

TABLE 3. Coefficients for estimating blubber, skin and core volume of Nastapoka belugas (n=16) from
the cube of body girths or body length.
Volume = A.10-5 Girth3 or A.10-6.Length3
(L) (cm) (cm)

Blubber Skin Core

Girth R2 R2
Interval A SE 10-2 A SE 10-2 A SE 10-2

1 6.03 0.35 95 1.77 0.10 96 11.00 0.55 96

2 2.77 0.13 96 0.81 0.04 96 5.02 0.22 97
3 2.37 0.08 98 0.68 0.03 96 4.27 0.16 98
4 2.50 0.10 98 0.73 0.03 98 4.52 0.17 98
5 4.66 0.16 98 1.35 0.06 97 8.38 0.33 98
6 12.00 0.87 92 3.48 0.28 90 21.50 1.75 90
7 39.50 6.17 71 11.40 1.83 70 70.80 11.33 70
Length 5.00 0.20 98 1.46 0.06 97 9.06 0.31 98

136
 The thermal conductivity of blubber (hypodermis) of beluga whales lies at the lower end
of the three-fold range reported for marine mammals (Table 4). Most published values of
blubber conductivity are from samples which include dermis, epidermis, and, in the case of
seals, the pelage. These heterogeneous samples thus bias heat flow estimates based solely on
blubber (fat) thickness. Thermal conductivity may vary with body location since fatty acid
composition is stratified with blubber depth (at least in fin and sei whales B.borealis, Lockyer
et al. 1985) and thermal conductivity appears to vary with fatty acid chain length (Table 4).
In belugas, the connective tissue content of blubber (and presumably the lipid content) was
fairly homogeneous over most body regions, as judged by ease of cutting and visual
appearance. The thermal conductivity of the mid-dorsal blubber is used here in the absence
of more widespread measurements and is thought to be representative of most body regions
with the exception of the head and caudal peduncle. The extent to which fatty acid
composition and thermal conductivity may vary with body location in belugas is not known.
Brodie (1975) argued that only a whale's inner core should be considered as the metabo-
lically active mass. In belugas, up to 40% of body mass can be blubber (Sergeant and Brodie
1969, pers. obs.). For thermostasis in a given environment, heat loss must be balanced by heat
production. The integumentary heat loss of belugas in the estuarine waters of Hudson Bay
(15°C) and their theoretical basal heat production based on either total or metabolically active
body mass (Fig. 2) are approximately equal for calves whereas adults are overinsulated. Even
in high arctic waters of 0°C in which the thermostatic energy requirement is increased by
75%, Kleiber's prediction for adult basal metabolism appears sufficient to meet this demand.
Larger animals may have to dissipate heat through blood vessels in the flippers and the flukes
(Bel'kovich 1961).
Basal metabolic rates based on 02 consumption reported by Kasting et al. (1989) for three
captive belugas are 1.5 — 1.8 times those predicted by Kleiber's relation. These values,
however, should be viewed with caution. The heat loss reported for the beluga named
Churchill (estimated weight 324-340 kg) was 236 W in 12— 18°C water (Kasting et al. 1989,
p. 695), a value slightly lower than the basal heat production of 263 W predicted by Kleiber's
relation, but higher than the 161 W integumentary loss which I determined for a similar sized
TABLE 4. Thermal conductivity of lipids from various sources.
Conductivity
W•m- i.°C- ' Species Source

Blubber 0.07" Mirounga leonina Bryden (1964)

0.06°' Phocoena Yasui and Gaskin (1986)
0.18° Phoca groenlandica
Phoca vitulina
Halichoerus grypus Worthy (1985)
0.20." Phoca hispida Scholander et al. (1950)
0.21' Balaenoptera physalus Parry (1949)

Fat 0.21 human Hensel et al. (1973)

0.33 Stolwijk and Hardy (1977)

Fatty 0.16 stearic acid CRC (1967)

acids 0.17 palmitic acid CRC (1967)
0.23 oleic acid CRC (1967)
a Includes epidermal, dermal and hypodermal layers.
b Flux of 133.07 kcal m.2 h- ' through 1.7 cm thickness having 14°C gradient.

Average of 4 measured in water.

137
beluga (length 278 cm) in 15°C Hudson Bay water (Table 2). The average heat flux of
47 Wm-2 (161W/3.4 ni -2) over the skin surface of the 278-cm male beluga from Hudson Bay
is similar to the 50 W.m -2 on the body of the whale called Churchill (Kasting et al. 1989). The
0 2 metabolic estimates of Kasting et al. (1989) may be biased due to the short time interval
over which expirations were collected and to the use of an assumed RQ of 0.80 rather than
the measured values.
Stromberg (1985) judged the thermogenic effect of the oxidation of fatty acids and
lipogenesis in the skin to be a useful adaptation especially in polar species of cetaceans.
Because heat flow is directly related to temperature gradient, the maintenance of higher than
ambient skin temperatures would be metabolically expensive regardless of the process used
(mass transport of heat via the blood or thermogenesis of fat oxidation). If skin is thermally
tolerant (possibly through high glycerol levels, Stromberg 1985), then temperatures in the
skin could drop, thereby reducing heat loss. The non-saturation of fatty acids in the blubber
would allow it to cool without the fat setting (Sokolov 1962). Muscle temperatures below
30°C in a net-entrapped narwhal in Tremblay Sound, Baffin Is., NWT, and a stranded beluga
at Cunningham Inlet (Doidge, unpubl. data) indicates shell cooling similar to that found in
fin whales (Brodie and Paasche 1985). The superior ability of fat to cool without losing
physiological function may be the reason that it rather than muscle is deposited subcutaneous-
ly. Thus the fat acts as an insulative layer as a "consequence of its presence" (Pond 1978).
Its insulative value does not appear to be maximised since blubber thickness (Fig. 3) relative
to body radius (girth / 27c) is not constant (Ryg et al. 1988). Blubber distribution must
therefore be influenced by additional factors, or the temperature distribution on the surface
of the inner shell must also vary within girth interval.
Bel kovich (1961) suggested the thick horny layer (stratum externum) of perinatal belugas
was characteristic of all neonatal cetaceans. A 187 cm beluga from northern Soviet waters
had skin 22 mm thick and blubber on the mid-back of 30 mm (Kleinenberg et al. 1969). Two
beluga neonates examined in northern Quebec in 1985 had similarly thickened epidermis. A
160 cm calf had mid-lateral skin of 22 mm and blubber of 24 mm. Mid-lateral skin and
blubber thicknesses of a 145 cm neonate was 25 and 10 mm, respectively. The observed color
changes of neonates (Bel' kovich 1961; Sergeant 1973) suggests the outer layer of the stratum
externum, which accounts for this extra thickness, is shed soon after birth. The outermost cells
lack the cohesiveness to remain attached to the underlying layers (D. J. St. Aubin, OVC,
University of Guelph, Guelph, Ont., pers. comm.).
Heat loss is related to the thermal conductivity and thickness of skin and blubber, but skin
has only one half the insulative value. Epidermal thickness in neonates compensates for the
reduced blubber layer. Thickened skin thus serves as a thermal buffer at birth, which is shed
and replaced by blubber. No evidence of thickened skin was observed in the two narwhal
near-term fetuses. However, the small number of animals examined precludes speculation on
how widespread the phenomenon of perinatal pachydermy is in cetaceans.
The thermal benefit of estuaries is directly proportional to the temperature gradient across
the skin and blubber to the animal's body core. Since narwhals and belugas appear to be
equally insulated (Fig. 2) and yet only beluga have an estuarine habit, estuarine use must offer
more than simply reducing heat loss. Elevated temperatures encountered by belugas in river
mouths enhance the molting process of the skin and allow for other metabolic adjustments
(St. Aubin and Geraci 1989; St. Aubin et al. 1990). It is not known if narwhal undergo the
same . rapid skin turnover without the benefit of higher water temperatures.
Although the thermostatic approach can give only an approximation of metabolic rate
under the specific conditions mentioned earlier, this method allows comparison between
individuals and species of cetaceans. Coupled with heat flow data and ultrasound measure-
ments of blubber and skin, more detailed studies of the dynamics of cetacean energetics and
thermoregulation can be undertaken using this approach on live specimens in captivity.
' Kasting et al. (1989) . Table 2 report 28 IlinT"C - ' and —T 1.8°C which is equivalent to Win -2.

138
 Acknowledgments

D.E. Sergeant and P.F. Brodie sparked my initial interest in whale blubber and heat loss.
P. Schuepp of Macdonald College provided heat flow equipment and advice on thermal
matters and numerical methods. H. Silverman with the aid of S. Attagootak supplied the Pond
Inlet narwhal data. R. Greendale, M. Hammill, I. Inookie and T. Smith helped with data col-
lection in the high arctic. D. Albright, L. Caron, G. Horonowitsch, P. Inukpuk and D. St.
Aubin provided assistance in N. Quebec. Hunters from the high arctic and N. Quebec kindly
allowed their catch to be examined. Criticism from D. Albright, G. Beck, J. Geraci, M.
Hammill, C. Lockyer, D. Sergeant, T. Smith, D. St. Aubin and two anonymous reviewers im-
proved the manuscript.
Funding was provided by Fisheries and Oceans Canada, World Wildlife Fund Canada's
"Whales Beneath the Ice" project, the Centre for Northern Studies of McGill University, and
an NSERC postgraduate scholarship.

References
AMERICAN SOCIETY OF MAMMALOGISTS, COM- DOIDGE, D. W. 1990. Age—length and length-
MITTEE ON MARINE MAMMALS. 1961. Standar- weight comparisons in the beluga, Delphinap-
dized methods for measuring and recording terus leucas, p. 59-68. In T. G. Smith, D. J.
data on the smaller cetaceans. J. Mammal. 42: St. Aubin, and JR. [cd.] Advances in
471— 476. research on the beluga whale, Delphinapterus
BECKOVICH, V. M. 1961. Question of physical leucas. Can. Bull. Fish. Aquat. Sci. 224.
thermoregulation in the beluga (Delphinapte- GERACI, J. R., D. J. Sr. AUBIN, AND B. D. HICKS.
rus leucas). Trudy soveshchaniy ikhiologi- 1986. The epidermis of odontocetes: a view
cheskoy komissii Akademii nauk SSR (Trans- from within, p. 3-21. In M.M. Bryden and
actions of Conference of the Ichthyological R.Harrison [ed.] Research on dolphins. Cla-
Commission of the USSR Academy of Scien- rendon Press, Oxford, U.K. 478 p.
ces, No. 12, p 50-59). Translation from the HENSEL, H., K. BRUCK, AND P. RATHS. 1973. HO-
Russian, Joint Publ. Res. Sel., JPRS 44,796: meotherrnic organisms, p. 503-729. In H.
9-22. NTIS, Washington, 1968. Precht, J. Christephersen, H. Hensel, and W. L.
BOYCE, W. E., AND R. C. DIPRIMA. 1969. Elemen- Larcher [ed.] Temperature and Life. Springer-
tary differential equations and boundary value Verlag, NY. 779 p.
problems. 2nd edition. John Wiley and Sons, IRVING, L., AND J. S. HART. 1957. The metabolism
Toronto, Ont. 533 p. and insulation of seals as bare skinned mam-
BRODIE, P. F. 1975. Cetacean energetics, an over- mals in cold water. Can. J. Zool. 35: 497-511.
view of intraspecific size variation. Ecology KANWISFIER, J., AND G. SUNDNES. 1966. Thermal
56: 152-161. regulation in cetaceans, p. 397-409. In K.S.
BRODIE, P., AND A. PAASCHE. 1985. Thermoregu- Norris [ed.] Whales, Dolphins and Porpoises.
lation and energetics of fin and sei whales Univ. of California Press, Los Angeles, CA.
based on postmortem, stratified temperature '789 p.
measurements. Can. J. Zool. 63: 2267-2269. KASTING, N. W., S. A. L. ADDERLEY, T. SAFFORD,
BRYDEN, M. M. 1964. Insulating capacity of the AND K.G. HEWLETT. 1989. Thermoregulation
subcutaneous fat of the southern elephant seal. in beluga (Delphinapterus leucas) and killer
Nature 203: 1299-1300. (Orcinus orca) whales. Physiol. Zool. 62:
CRC (CHEMICAL RUBBER COMPANY). 1967. Hand- 687-701.
book of chemistry and physics. 48th Edition, KLEIBER, M. 1975. The fire of life. 2nd edition.
The Chemical Rubber Company, Cleveland, Robert E. Krieger Publishing Co., NY. 453 p.
Ohio, USA. KLEINENBERG, S.E., A. V. YABLOKOV, B.M.
CONOVER, W. J. 1981. Practical nonparametric BECKOVICH, AND M. N. TARASEVITCH. 1969.
statistics. 2nd edition. John Wiley and Sons, Beluga (Delphinapterusleucas), investigation
Toronto, Ont. 493 p.

139
 of the species. Israel Program Sci. Transi., SERGEANT, DE., AND P. F. BRODIE, 1969. Body
376 p. size in white whales,De/phinapterus !twos. J.
LAVIGNE, D. M., S. INNES, G. A. J. WORTHY, K. M. Fish. Res. Board. Can. 26: 2561-2580.
KovAcs, 0.1 SCHMITZ, AND J. P. HICKIE. 1986. SOKOLOV, W. 1962. Adaptations of the mamma-
Metabolic rates of seals and whales. Can. J. lian skin to the aquatic mode of life. Nature
Zool. 64: 279-284. 195: 464-466.
LOCKYER, C.H., L.0 MCCONNELL, AND T.D. STOLWLIK, J. A.J., AND J. D. HARDY. 1977. Con-
WATERS. 1985. Body condition in terms of trol of body temperature, p. 45-68. In D.H.
anatomical and biochemical assessment of Lee [ed.] Handbook of Physiology, Section 9,
body fat in the North Atlantic fin and sei wha- Reactions to environmental agents. Am. Phy-
les. Can. J. Zool. 63: 2328-2338. siol. Soc., Bethesda, USA.
PARRY, D.A. 1949. The structure of whale blub- STROMBERG, M. W. 1985. Fat distribution in the
ber and its thermal properties. Q. J. Microsc. skin of bottlenose dolphins (Tursiops trun-
Sci. 90: 13-25. catus and Tursiops gilli). J. Morphol. 186:
POND, C.M. 1978. Morphological aspects and 315-326.
ecological and mechanical consequences of ST. AUBIN, D. J., AND J. R. GERACI. 1989. Seaso-
fat deposition in wild vertebrates. Annu. Rev. nal variation in thyroid morphology and secre-
Ecol. Syst. 9:519-570. tion in the white whale, Delphinapterus leu-
RYG, M., T.G. SMITH, AND N. A. ORITSLAND. cas. Can. J. Zool. 67: 263-267.
1988, Thermal significance of the topogra- ST. AUBIN, D. J., T. G. SMITH, AND J. R. GERACI.
phical distribution of blubber in ringed seals 1990. Seasonal epidermal molt in beluga
(Phoca hispida). Can. J. Fish. Aquat. Sci. 45: whales, Delphinapterus !meas. Can. J. Zool.
985-992. 68: 359-367.
SCHOLANDER, P. F., V. WALTERS, R. HOCK, AND WORTHY, G. A.J. 1985. Thermoregulation in
L. IRVING. 1950. Body insulation of some arc- young phocid seals. Ph.D. thesis, University
tic and tropical mammals and birds. Biol. Bull. of Guelph, Guelph, Ont. 254 p.
99:225-236. YASUI, W.Y., AND D.E. GASKIN. 1986. Energy
SERGEANT, D.E. 1973. Biology of white whales budget of a small cetacean, the harbour
(Delphinapterus leucas) in western Hudson porpoise, Phocoenn phocoena (L.). Ophelia
Bay. J. Fish. Res. Board Can. 30: 1065-1090. 25: 183-197.

140
Dynamics of Tooth Growth in Beluga Whales, Delphinapterus leucas,
and Effectiveness of Tetracycline as a Marker for Age Determination

P.F. Brodie
Department of Fisheries and Oceans, Habitat Biology Division
Bedford Institute of Oceanography, Dartmouth, Nova Scotia, Canada B2Y 4A2

J.R. Geraci and D.J. St. Aubin

Wildhfe Section, Department of Pathology, Ontario Veterinary College,

University of Guelph, Guelph, Ontario, Canada NI G 2W]

BRODIE, P. F., J. R. GERACI, AND D. J. ST. AUBIN. 1990. Dynamics of tooth growth
in beluga whales, Delphinapterus leucas, and effectiveness of tetracycline
as a marker for age determination, p. 141-148. In T. G. Smith, D. J. St.
Aubin, and J. R. Geraci [cd.] Advances in research on the beluga whale,
Delphinapterus leucas. Can. Bull. Fish. Aquat. Sci. 224.

Seven subadult beluga whales (Delphinapterus leucas) were captured in Hudson

Bay, Canada, and held in captivity for up to 10 wk after receiving an intramuscu-
lar injection of oxytetracycli ne. Sections from teeth, extracted prior to their release,
were examined under a microscope using ultraviolet light. Fluorescent marking
was evident in the teeth of all the experimental animals, and provided a time
reference from which tooth growth could be measured. After taking into account
the age-specific changes in thickness of Growth Layer Groups (GLGs), the ratio
of pre- to post-injection tooth growth provided an estimate of age. Calculated age
was in agreement with a deposition rate of two GLGs per year. The strict
seasonality of an arctic habitat is considered as a potential factor in explaining this
phenomenon. The study also demonstrated that the teeth of beluga can be effecti-
vely marked by dosages of tetracycline lower than previously used in cetaceans.

Sept marsouins blancs subadultes (Delphinapterus leucas) ont été capturés dans
la baie d'Hudson, au Canada, et gardés en captivité jusqu'à 10 semaines, après
avoir reçu une injection intramusculaire d'oxytétracycline. Avant de les remettre
en liberté, des sections de dents extraites furent examinées sous microscope
utilisant une lumière ultraviolette. Un marquage fluorescent était évident dans les
dents de tous les animaux expérimentaux, et a fourni une référence temporelle par
laquelle la croissance dentaire pouvait être mesurée. Après avoir pris en considé-
ration les changements spécifiques à l'âge dans l'épaisseur des Groupes de
Croissance en Couches (G.C.C.), le ratio de la croissance dentaire donna un estimé
de l'âge entre la pré-injection et la post-concordance avec la cadence de déposi-
tions de deux G.C.C. par année. Le précis des saisons de l'habitat arctique est
considéré comme un facteur potentiel pour expliquer ce phénomène. Cette étude
a aussi démontré que les dents des marsouins blancs peuvent être marquées
efficacement avec des dosages de tétracycline moins élevés que ceux utilisés au
préalable chez les cétacés.

141
 Introduction

Aging belugas by counting tooth layers is controversial because of disagreement as to

whether single or multiple growth-layer-groups (GLGs) (Perrin and Myrick 1980) are
deposited annually. In belugas, as many as four GLGs are deposited before the tooth erupts
and is exposed to wear. Erosion of the record of early growth can begin after as few as 15
GLGs have been laid down (Brodie 1969). Body growth rates of known-age animals have
been difficult to obtain, and have so far been limited to measurements from long-term captive
animals (Brodie 1982; Goren et al. 1987). Ideally a reference point within the calcified tissue
is required in order to establish the rate at which the tooth grows and therefore what portion
of the year is represented by a GLG.
Tetracycline administered to an animal binds with free body calcium, and is incorporated
into hard tissue that is forming at the time the drug is introduced. This becomes detectable
as fluorescence under ultraviolet illumination. Cetacean teeth have been so marked using
tetracycline in dosages of 40-60 mg/kg of body weight (Best 1976; Gurevich et al. 1980) and
25 mg/kg (Myrick et al. 1984). The technique was first applied to beluga whales in Western
Hudson Bay in 1968 by Sergeant and Brodie (1969a) who administered approximately
20 mg/kg to 93 tagged free-ranging specimens; however, none of the animals could be
recovered for subsequent analysis because of reduced hunting from that large population. In
1984 and 1985 the experiment was repeated, but in a setting that guaranteed retrieval of teeth
after a suitable period of time following administration of the drug.

Materials and Methods

Subadult beluga whales were captured in the estuary of the Seal River, Manitoba, on the
western side of Hudson Bay, and held for up to 10 wk in 8.2 ni diameter pools. The whales
were fed thawed Pacific herring (Clupea harengus pallasi) and fresh cisco (Coregonus
artedii), providing 1.7 — 2.1 x 105 J.kg body wt- 41-1 .
Four individuals in July 1984 and three in August 1985 were given intramuscular injections
of oxytetracycline (100 mg/mL), at a dosage of 10-20 mg/kg (Table 1). To achieve the
desired dose, multiple injections of up to 10 mL each were administered using a 10 cm, 18
gauge needle directed into the epaxial musculature. After 3-7 wk, one tooth was extracted
from each of six whales; three teeth were collected post mortem from the seventh. Prior to
extracting a tooth, 3-4 mL of lidocaine was injected into the gum tissue. A dental gouge was
used to free the teeth from the soft tissue attachment. Teeth were stored in NaCl, embedded
in plastic and sectioned using a Gillings-Hamco thin sectioning machine (Hamco, Rochester,
NY) and further handground on 600 grit wetpaper. The 40-50 gm thick sections were
examined using a microscope equipped with an ultraviolet light source.

Results and Discussion

In view of the small number of animals available for study, we chose not to designate any
of the whales as sham controls for the drug injections. We interpreted any fluorescent mark
as the result of tetracyline labelling. Such are the limitations of field-based research on large
marine mammals.
All teeth showed some degree of fluorescent marking, though there was variation in the
degree of marking between individuals and in different teeth from the sanie individual .
There was less intense deposition of tetracycline in the cementum, from the base of the
tooth to within 3 —5 mm of the tip, the portion of the tooth that was protruding from the
gum and was not available to deposition. Teeth from animals with the lowest dosage of
tetracycline (10 mg/kg) were as intensely marked as those from whales receiving twice
the dosage.

142
TABLE 1. Age estimates of beluga based on tooth layer counts and the ratio of dentine deposition before and after tetracycline injection.
Age estimate (yr)
Time (d) between indicated by
Whale Length injection and GLU counts tetracycline mark/
ID# Sex (cm) tooth extraction Dentine Cement total dentine
SR0384 M 254 24 4 4 2.17

SR0284 F 267 51 5—6 4 2.10

SR2385 M 979 48 6 5 2.00

SR0184 F 269 26 8—9 8 4.27

SR1685 M 274 44 8—9 9-10 2.77

5 R2185 F 295 48 10 — 12 10 4.82

 To use post-injection tooth growth as an index of age, the cumulative, median widths of
previous GLGs were compared with the thickness of substrate deposited after the tetracycline
mark. Since the post-injection period was known precisely, the ratio provided an age
estimate. This approach will underestimate the age of younger animals because GLG
thickness in beluga increases with age to about physical maturity, then declines (Fig. 1), a
trend which is apparently not found in many of the other odontocetes (Myrick 1988). The
magnitude of the error decreases with the age of the animal. Based upon the growth trend in
Fig. 1, the correction would bring the estimates more into agreement with ages based upon
dentine and cement= laminae at a ratio of two GLGs per year. In order for single GLGs to
represent annual growth, a correction of 100% would be required.
A particularly unusual finding in three of the whales was the occurrence of distinct double
bands in the dentine and, though less well defined, in the cementum (Fig. 2). One band was
deposited within the collagen matrix containing the tubule extensions of the odontoblasts,
sometimes referred to as Tomes' fibres (Miles 1967; Brodie 1970). The other was at the base
of the tubules, adjoining the pulp cavity.
Tomes' fibres are part of the mechanism by which mineralization proceeds (Miles 1967;
Brodie1970). These tubules appear to be formed coincidentally with the extrusion of collagen

1200
MALE

800
r
..

o
400

o
li ),•

>-
< FEMALE
800

[I]
400 [0] +

/, , ,
0 . ,

o 5 10 15 30 20 10 0
YOUNG OLDER
FIG. I. Thickness measurement of GLGs of male and female belugas, measured as shown in Fig. 2B.
Arrows indicate the direction in which the GLGs were measured. Teeth of younger animals, with
prenatal teeth still discernible, provided thickness measurement for the first 15 GLGs. Older animals,
with much of the tooth worn away, provided GLG thicknesses counting from the pulp cavity, towards
the worn tip. Only one older female was included in the sample. (from Brodie 1970)

144
and are pinched off in unison from the odontoblasts (Fig. 2). The collagen substrate continues
to be produced and carries with it this sheet of tubules. In unstained, longitudinal sections
of teeth it is this sheet of tubules which appears as a dark layer in transmitted light and as white
layer in reflected light. Fine laminations within the collagen, running parallel to the edge of

PRENATAL
DENTINE

CEMENT
CEMENT
LAYERS

SITE OF GLG
THICKNESS MEASUREMENT

TOMES FIBRES

FLUORESCENCE
1st BAND
\ 2ND BAND
' /
D
a

POST INJECTION LAMINATION WITHIN

PULP COLLAGEN SUBSTRATE
TOOTH GROWTH CAVITY

FIG. 2. General structure of a beluga tooth in long section, with microphotographs of features. Eight
Growth Layer Groups are depicted here. (A) Tomes' fibres as seen with transmitted light, showing the
cessation and continued extrusion of the tubules. (B) Growth Layer Group (GLG) as seen by reflected
light, which includes one dark and one light layer within the dentine. (C) Collagen substrate within the
GLG, which appears to be extruded in pulses. Tome's fibres are seen here by transmitted light, per-
pendicular to the pulp cavity. (D) Fluorescent bands formed after injection of tetracycline. Two bands
are seen here: the first being deposited immediately after injection, while the second band is formed at
the interface with the pulp cavity.

145
 the pulp cavity (Fig. 2), suggest that this substrate may be extruded in pulses and, relative to
post-injection deposition as indicated by the tetracycline mark, each pulse may represent 1-2
wk. There is evidence that tooth mineralization was in progress among the experimental
annuals at time of capture, since the Tomes' fibres seem to have been in various stages of
extrusion when the drug was administered. Single fluorescent bands were observed at the
distal ends of the tubules in some animals and mid-way along the tubule formation in others,
indicating that there were sites of mineralization along the tubules and intertubular spaces.
A simple test was conducted to ascertain whether the band at the proximal end of the
Tomes' fibres, in contact with the pulp cavity, was an "edge effect," with the tubules acting as
optic fibres, conducting fluorescence back to the surface. Under a microscope using UV light,
small pieces of the tooth section containing the proximal band were broken off. The broken
pieces continued to fluoresce, while the newly exposed edge on the tooth section did not. This
eliminated the possibility of an edge effect and indicated that the proximal band was real.
We do not know if the inner band at the pulp cavity would have eventually fused with the
outer one, or would have formed a separate fluorescent band within the extruded substrate,
because the experimental animals were released. We also cannot exclude the possibility that
the pattern might simply have been an artifact of captivity, even though only three of the
experimental animals had evidence of double banding. It is possible that some tetracycline
may have been withheld, by bonding with calcium in the bloodstream, causing a delay in
complete deposition within bones and teeth. Such an artifact might also complicate interpre-
tation of a series of tetracycline injections administered over a long period in captivity
(Myrick et al. 1988).
The occurrence of double bands suggests an unusual pattern of dentine deposition in this
species. Their teeth appear to grow continuously via collagen extrusion, punctuated by
regular periods of mineralization (Brodie 1969). These periods may coincide with seasonal
cycles in hormone secretion, such as has been observed for thyroid activity in these animals
(St. Aubin and Geraci 1989). Normal calcium deposition may be interrupted by episodes of
hypocalcemia, perhaps associated with reproduction or stress (Myrick 1988) from physical
injury (for example, from massive bullet wounds) which result in resorption and mobilization
of minerals that have already been deposited (Brodie 1970).
We do not propose that the observed artifact of double lamination of tetracycline bands is
argument for double GLGs per annum, simply that the initial mark indicates the rate of tooth
growth and the relative age. The rate at which the tooth grows, as indicated by tetracycline
marking, adds to a body of evidence that in this species, there is the potential for multiple,
rather than single, laminations to be deposited annually. Tentative classification of calves by
year class in the field, using several parameters — body growth rate (Brodie 1971) and skin
color change of a captive animal (Brodie1982) — have provided evidence for multiple, rather
than single, GLGs as annual deposits; the reason for this continues to be elusive.
The natural lamination sequences may reflect temporary consistent activities, in which the
entire herd participates, such as an early winter movement out from the coasts to feeding areas
in ice-free waters (Kleinenberg et al. 1964) and a return in early summer, half a year later, to
coastlines and estuaries. The beluga is one of the few circumpolar odontocetes subject to the
extremes of strict seasonality imposed by such a high latitude, where the year is clearly
divided into two parts (Brodie1969). The Gulf of St. Lawrence population, the most southerly
habitat for this species, is subject to essentially the same subarctic conditions. This predic-
table seasonality in belugas raises the question as to whether other species, with cyclical
patterns, for example, its cohabitant, the narwhal (Monodon monoceros) or the highly
migratory sperm whales (Physeter catodon), are similarly affected.
The study demonstrated that the teeth of beluga whales can be effectively marked by
dosages of tetracycline lower than previously used in cetaceans. The higher proportion of
blubber (as much as 40%, Sergeant and Brodie 1969b) on this arctic-adapted species may
suggest that body tissue composition should be taken into account when estimating effective

146
dosages in marine mammals. Still, even in its most concentrated form (100 mg/mL), the
volume of tetracycline solution to mark a 500-kg whale would be in the order of 50 to 100 mL,
considerably more than the 10 mL volume regarded as the safe limit for injection into a single
site. Thus, any marking program for belugas would require that the animals be temporarily
restrained for multiple injections.

Acknowledgments

We thank T. Friesen and C. Thomson (University of Guelph) and J.Orr and S. Waters (Dept.
of Fisheries and Oceans) for field assistance and maintenance of the whales. R. Moshenko,
Section Chief (DFO), coordinated logistic support for the whale capture and field camp.
Tooth sections were prepared by D. W. Doidge. The authors wish to thank the referees for
their very constructive comments. Funding for the study was provided by the Department
of Fisheries and Oceans, The World Wildlife Fund ("Whales Beneath the Ice" Program), and
the Natural Sciences and Engineering Research Council (Grant A6130 to JRG).

References

BEST, P. B. 1976. Tetracycline marking and the termination ofDelphinus delphis. Proceedings
rate of growth layer formation in the teeth of a and Workshop Report, International Confe-
dolphin (Lagenorhynchus obscurus). S. Afr. J. rence on Age Determination of Odontocetes
Sci. 72: 216-218. and Sirenians. International Whaling Com-
BRODIE, P. F. 1969. Mandibular layering in Del- mission,
phinapterus leucas and age determination. KLEINENBERG, S.E., A. V. YABLOKOV, B.M.
Nature 221: 956-958. BECKOVICH, AND M.N. TARASEVICII. 1964.
1970. Life history of the white whale, Beluga (Delphinapterus leucas). Investigation
(Delphinapterus leucas) (Pallas) in the waters of the species. Academy of Sciences of the
of Baffin Island, Canada. Ph.D. thesis. Dal- USSR, A.N. Severtson Institute of Animal
housie University, Halifax, Nova Scotia, Morphology. Nauka, Moscow. Translated from
Canada. Russian 1969 by Israel Program for Scientific
1971. A reconsideration of aspects of growth, Translations, Jerusalem. 376 p.
reproduction and behaviour of the white whale MILES, A. E. W. [ed.]. 1967. Structural and chemi-
(Delphinapterus leucas), with reference to the cal organization of teeth. Vol. 1 and 2. Acade-
Cumberland Sound, Baffin Island, population. mic Press, San Diego, CA.
J. Fish. Res. Board, Can. 28: 1309-1318. MYRICK, A. C., JR. 1988. Is tissue resorption and
1982. The beluga (Delphinapterus leu- replacement in permanent teeth of mammals
cas); Growth at age based on a captive speci- caused by stress-induced hypocalcemia?
men and a discussion of factors affecting natu- p. 379-389. In Z. Davidovitch [ed.] The bio-
ral mortality estimates. SC/33/SM7,445-447 logical mechanisms of tooth eruption and root
Rep. Intl. Whal. Comm. 32. resorption. EBSCO Media, Birmingham, Al
1989. The white whale (Delphinapterus 35233.
leucas)(Pallas,1776), p.119-144. In S. H. Ridg- MYRICK A. C. JR., E. W. SCHALLENBERGER,
way and Sir R. Harrison [ed.] Handbook of I. KANG, AND D. B. MACKAY. 1984. Calibra-
marine mammals 4, Academic Press, New tion of dental layers in seven captive Hawaiian
York, NY. 442 p. spinner dolphins, (Stenella longirostrus), ba-
GOREN, A. D., P. F. BRODIE, S. SPOTTE, G. C. RAY, sed on tetracycline labelling. Fish. Bull.
W. H. KAUFMAN, A. J. GWINNETT, J. J. SCIUB- (NOAA) 82: 227-225.
BA, AND J. D. BucK.1987. Growth layer groups MYRICK, A. C., JR., P. K. YOCHEM, AND L.H.
(GLGs) in the teeth of an adult beluga whale CORNELL. 1988. Toward calibrating dentinal
(Delphinapterus leucas) of known age: evi- layers in captive killer whales by use of tetra-
dence of two annual layers. Mar. Mamm. Sci. cycline labels. Rit. Fiskideildar. 11: 285-296.
3:14-21. PERRIN, W.F., and A. C. MYRICK. 1980. Report
GUREVICH, V. S., B. S. STEWART, AND L. H. Co K- of the Workshop. Int. Whaling Comm. Spec.
NELL. 1980. The use of tetracycline in age de- Issue 3: 1-50.

147
ST. AUBIN, D. J., AND J. R. GERACI. 1989. Season- SERGEANT, D. E., AND P. F. BRODIE. 1969a. Tag-
al variation in thyroid morphology and secre- ging white whales in the Canadian Arctic. J.
tion in the white whale,Delphinapterus leucas. Fish. Res. Board Can. 26: 2201-2205.
Can. J. Zool. 67: 263-267. 1969b. Body size in white whales, Delphi-
napterus leucas. J. Fish. Res. Board Can. 26:
2561-2580.

148
 Adrenal Responsiveness to Stimulation
by Adrenocorticotropic Hormone (ACTH)
in Captive Beluga Whales, Delphinapterus leucas

D.J. St. Aubin and J.R. Geraci

Wildlife Section, Department of Pathology, Ontario Veterinary College,
University of Guelph, Guelph, Ontario, Canada N IG 2W]

Si. AUBIN, D. J., AND J. R. GERACI. 1990. Adrenal responsiveness to stimulation

by adrenocorticotropic hormone (ACTH) in captive beluga whales, Del-
phinapterus leucas, p. 149-157. In T. G. Smith, D. J. Aubin, and J. R.
Geraci [ed.] Advances in research on the beluga whale, Delphinapterus
leucas. Can. Bull. Fish. Aquat. Sci. 224.

The effect of 75 IU of bovine adrenocorticotropic hormone (ACTH) on circu-

lating adrenal hormones and white blood cells was assessed in two beluga whales,
Delphinapterus leucas, immediately after capture, and in five others on seven
occasions during the course of 10 wk in captivity. Five control studies using saline
were carried out on four of the whales held for at least 6 wk; a sixth control injection
was performed on one whale immediately after capture. Blood samples were
collected prior to, and 0.25, 0.5, 1, 2, 3, 4, 5, 6, 12, and 24h after, injection with
ACTH or saline solution. Stimulation by ACTH significantly elevated circulating
levels of cortisol and aldosterone, and depressed total eosinophil counts. The
increase in aldosterone was proportionately greater and more consistent than for
cortisol. Peak response for aldosterone occurred 2h after stimulation, and normal
values were restored after 6h. Eosinophil counts declined steadily during the first
12 h after ACTH stimulation, and were within the normal range after 24 h. The
pattern of hormonal and cellular changes in acclimated whales was comparable to
that observed in whales tested immediately after capture.

L'effet de 75 IU de l'hormone adrénocorticotropic bovine (ACTH) sur les

hormones circulatoires surrénales et les cellules blanches sanguines fut évalué
chez deux marsouins blancs, Delphinapterus leucas, immédiatement après les
avoir capturés, et chez cinq autres, à sept occasions, au cours des 10 semaines de
captivité. Cinq études de contrôle utilisant une saline furent effectuées chez quatre
des baleines gardées pendant au moins 6 semaines; une sixième injection de
contrôle fut administrée chez une baleine immédiatement après la capture. Des
échantillons de sang furent recueillis avant, et 0.25, 0.5, 1, 2, 3, 4, 5, 6, 12,24 h après
une injection avec l'ACTH ou une solution saline. La stimulation par l'ACTH
éleva d'une façon significative les niveaux circulatoires de cortisol et d'aldosté-
rone, et a abaissée la numération éosinophile totale. L'augmentation d'aldostérone
était proportionnellement plus forte et plus uniforme que pour le cortisol. La
réaction de pointe pour l'aldosterone se produisit 2 h après la stimulation, et les
valeurs normales furent rétablies après 6 h. Le décompte d'éosinophiles déclina
d'une façon régulière durant les 12h après la stimulation pari 'ACTH, et était dans
le champ normal après 24 h. L'échantillonage des changements hormonaux et
cellulaires chez les baleines acclimatées était comparable à celle observée chez les
baleines examinées immédiatement après la capture.

149
 Introduction

An integral part of the mammalian stress response is the release of steroid hormones from
the adrenal cortex. Secretion is controlled principally by adrenocorticotropic hormone
(ACTH) produced by the anterior pituitary, which in turn is regulated by corticotropin-
releasing hormone (CRH) from the hypothalamus. Exogenous ACTH or CRH can be used
as a controlled stimulus to determine the responsiveness of the adrenal cortex in vivo. This
protocol has broad applications as a diagnostic tool and in experimental studies on a variety
of species (Paape et al. 1977; van Mourik and Stelmasiak 1984; Willard et al. 1987), including
marine mammals (St. Aubin and Geraci 1986; Thomson and Geraci 1986).
Continuous infusion of ACTH produces a biphasic response from the adrenal. Initially,
plasma levels of cortisol and aldosterone are increased, but after 48-72 h, the aldosterone
response is attenuated, an adaptation referred to as "aldosterone escape" (Gaillard et al. 1983;
Oelkers 1985). The adjustment is associated with morphologic and enzymatic changes in the
zona glomerulosa of the adrenal, reducing its capacity to synthesize and secrete aldosterone
(Mazzocchi et al. 1986; Riondel et al. 1987). We used ACTH stimulation to indirectly assess
whether a similar transformation was associated with acclimation to captivity in beluga
whales, Delphinapterus leucas.

Materials and Methods

Eight beluga whales were captured in the Seal River estuary during July, 1985 (Table 1).
Three were studied immediately after capture, then released. The others were transported to
holding facilities, where they were maintained for 10 wk in two 8.2-m-diameter pools
containing 3.8 x 10 4 L of water renewed daily from Hudson Bay. The whales were fed Pacific
herring, Clupea harrengus pallasi, supplemented with vitamins, but were fasted for 12 Il
before, and for the first 6 h of, each study.
During the latter 5 wk of the holding period, seven ACTH stimulation tests and five sham
controls were conducted (Table 1). All of the trials were begun between 07:45 and 10:15 h.
For a test, the pool was drained to a depth of less than 30 cm, a procedure that required
approximately I h. A sample of venous blood from the caudal peduncle or the flukes of each
beluga to be studied was drawn into chemically clean sterile syringes then transferred to
vacuum tubes containing heparin for plasma chemical analysis, and EDTA for hematologi-
cal determinations. A 15-cm needle was used to deliver 75 IU (0.2-0.28 IU/kg) of Cortrosyn
(Organon Canada Ltd., Toronto, Ont.) or 3 mL of saline (control) into the dorso-lateral
musculature, in line with the middle of the dorsal ridge. Thereafter, samples were nominally
taken at 15 and 30 min, and 1, 2, 3, 4, 5, 6, 12, and 2411. Specimens were held on ice until
further processed. The holding pool was refilled following the 6 and 12 h samplings, and
drained again to collect samples at 12 and 24 h post injection.

TABLE I. Schedule of ACTH stimulation tests and saline controls, expressed as the number of weeks
after capture.
Whale Sex Length Weight ACTH Test Control
(cm) (kg)
DL7 M 246 — 0 —
DL8 M 345 0 —
DL24 M 269 — — 0
DL4 M 269 275 6
DL14 F 269 270 9 7
DL16 M 274 330 10 6,10
DL21 F 295 370 5,9 10
DL23 M 272 350 6,8 10

150
 Eosinophil determinations were made within 24h of blood collection; blood smears and
total leukocyte counts were made within 48 h. Duplicate counts of eosinophils and total
leukocytes were performed using Unopette system (Becton-Dickinson) and a Neubauer
hematocytometer. Differential white blood cell counts were performed by counting 200 cells
on a Wright's stained peripheral blood smear using the cross-sectional method. To maintain
uniformity when calculating percentages, total white blood cell (direct) counts and blood
smears were made at the same time.
In the field camp, blood for hormonal and chemical analysis was centrifuged within 1 h of
collection; the plasma was transferred to sterile culture tubes and immediately frozen.
Samples from the three whales studied at the capture site were held on ice for up to 24h before
processing. Plasma chemical constituents, including electrolytes (Na, K, Cl, Ca, P, and Fe),
metabolites (glucose, cholesterol, triglycerides, urea, uric acid, and creatinine), enzymes
(creatine kinase, alkaline phosphatase, aspartate aminotransferase, alanine aminotransfe-
rase, lactate dehydrogenase, and y- glutamyl transpeptidase), protein, and steroid hormones
were analysed as previously described (St. Aubin and Geraci 1989).
Statistical comparisons were made by univariate analysis of repeated measures, according
to the Statistical Analysis System (SAS Institute, Cary, N.C.).

Results

Cortrosyn, a synthetic peptide containing 24 of the amino acids of natural ACTH, was
effective in eliciting the release of adrenal hormones in beluga whales. No attempt was made
to use graded doses to establish whether the observed response represented maximal
stimulation. We present the findings in terms of the peak response for each animal, and, in
the interest of establishing an abbreviated sampling protocol for future studies, at intervals
selected to demonstrate the most consistent changes in each of the hormonal and cellular
constituents (Tables 2 and 3). Samples collected 2h after stimulation showed maximal or near
maximal release of aldosterone and cortisol, whereas changes in eosinophils and lymphocy-
tes were most evident in the 6 h sample.

Immediate Response

The stress of chase and capture alone elicited a characteristic adrenal response. Blood
samples obtained from whale DL24 during the first 6 h after capture showed rising
concentrations of cortisol (190% after 2 and 6 h) and aldosterone (400% after 2 h) (Fig. 1,
Table 2). During this time, circulating eosinophils declined by 55% (Fig. 1), from an initial
count of 2.88 x 10 cells/L, and lymphocytes by 81%, from 5.8 x 109 cells/L.
In the two whales injected with ACTH within 1 h after capture, cortisol showed a similar
incremental increase of 100-120 nmol/L, but because the initial values were lower than in
DL24, the proportional change (360% and 420% after 3 and 5 h, respectively) was greater
(Fig. 1, Table 2). Peak concentrations of aldosterone were noted at 2 h after capture, and
reached 570% and 1280% of initial levels in the two ACTH-stimulated whales. After 6 h,
counts of circulating eosinophils had declined by 80% from 6.25 x 109 cells/L in DL7, and
by 30% from 2.28 x 109 cells/L in DL8 (Fig. 1, Table 3). Lymphocyte counts decreased from
4.0 to 0.8 x 109 cells/L in DL7, but in DL8 showed no consistent change from a relatively low
initial count of 1.1 x 109 cells/L.

Response in Acclimated Whales

The findings in whales studied 5-10 wk after capture are shown in Fig. 2, and summarized
on Tables 2 and 3. On average, the maximum percent response was lower for cortisol and
greater for aldosterone in the four ACTH trials conducted after 8-10 wk in captivity,

151
TABLE 2. Hormonal response to saline (CONT) or ACTH injection in beluga whales. Peak concen-
trations are expressed as percent of the value at the start of each trial. Numbers in parentheses show the
time, in hours, when the highest steroid concentrations were observed.
Cortisol Aldosterone
Whale Treatment Initial 2h Max % Inc. Initial 2h Max % Inc.
no. schedule" levelb sample' (time) level' sample' (time)
Immediate Response

DL7 ACTH -0 41 116 420 (5) 100 1340 1280 (2)

DL8 ACTH -0 39 135 360 (3) 140 820 570 (2)

DL24 CONT -0 132 246 190 (2) 230 950 400 (2)

Response after Acclimation

DL4 ACTH -6 135 223 310 (6) 70 300 470 (4)

DL14 CONT - 7 94 119 210 (1) 70 70 320 (1)

ACTH - 9 36 210 590 (3) 70 820 1180 (2)

DL16 CONT - 6 113 77 90 (1) 260 70 50 (1)

CONT - 10 218 105 70 (1) 80 70 110 (1)
ACTH - 10 102 80 190 (I) 70 150 2160 (2)

DL21 CONT - 10 28 69 420 (I) 200 70 35 (1)

ACTH - 5 28 39 220 (12) 70 590 850 (2)
ACTH - 9 163 201 150 (3) 70 710 520 (1)

DL23 CONT - 10 138 28 80 (1) 100 70 70 (1)

ACTH - 6 72 221 460 (3) 70 970 1410 (3)
ACTH - 8 119 174 190 (4) 70 680 970 (2)
" Time, in weeks after capture, when the whale was tested.
b Cortisol concentration in nmol/L.
Aldosterone concentration in pmol/L.

compared with the three earlier trials. However, in whales DL21 and DL23, which were
injected with ACTH on two occasions, the maximum change for both steroids was less during
the later trials (Table 2). The timing of the sham control injections had no apparent effect on
the hormonal response in the whales.
A mild adrenocortical response was associated with the manipulation required to conduct
the tests. In whales receiving a control injection of saline, mean cortisol levels ranged
between 120 and 135 nmol/L during the first hour of sampling, and thereafter declined to
roughly half of the initial values. Concentrations greater than 100 nmol/L were again noted
in samples collected after 12 and 24 h, as a further response to draining the holding pool.
Aldosterone concentrations initially followed a similar pattern; samples collected during the
first hour had concentrations of 120-170 pmol/L, whereas subsequent specimens averaged
70 pmol/L or less. In contrast to cortisol, aldosterone levels showed no further increase in the
12 and 24h samples. Mean eosinophil counts remained within 30% of initial values (2.95 +
1.50 x 109 cells/L) throughout the sampling period. Circulating lymphocytes were affected
only in the three animals with counts greater than 2.0 x 109 cells/L in samples drawn at time
0; in these animals, a 35-60% decrease in cell counts occurred after 5-6 h. Initial counts in
these whales ranged from 2.5 to 3.4 x 109 cells/L, whereas in the other two trials the values
were 1.4 and 1.9 x 109 cells/L.

152
Table 3. Changes in circulating eosinophil and lymphocyte counts following saline (CONT) or ACTH
injection in beluga whales. Minimal counts are expressed as percent of the value at the start of each trial.
Numbers in parentheses show the time, in hours, when the lowest cell counts were observed.
Eosinophils Lymphocytes
Whale Treatment Initial 6h Max % Dec. Initial 6h Max % Dec.
no. schedule levelb sampleb (time) level' sample (time)
Immediate Response

DL7 ACTH -0 6.25 1.32 21 (6) 4.0 nd 19 (5)

DL8 ACTH -0 2.28 1.59 69 (5) 0.9 0.8 89 (4)

DL24 CONT -0 2.88 1.26 44 (6) 5.8 1.1 19 (6)

Response after Acclimation

DL4 ACTH -6 1.74 0.47 9 (12) 2.8 1.9 39 (5)

DL14 CONT -7 4.92 3.78 60 (12) 1.4 1.5 93 (3)

ACTH -9 4.80 1.74 28 (12) 1.7 1.1 24 (3)

DL16 CONT - 6 1.22 0.87 64 (4) 2.5 1.0 40 (6)

CONT - 10 1.33 1.09 75 (12) 1.9 2.2 89 (3)
ACTH - 10 1.55 0.46 30 (6) 3.1 2.0 39 (5)

DL21 CONT - 10 4.24 4.48 82 (3) 3.4 3.4 59 (5)

ACTH - 5 4.99 2.50 22 (12) 1.9 1.3 58 (3)
ACTH - 9 3.73 2.33 51 (5) 2.2 3.1 73 (12)

DL23 CONT - 10 3.05 2.35 77 (6) 3.5 2.1 66 (5)

ACTH - 6 2.74 1.17 31 (12) 2.3 2.4 74 (3)
ACTH -8 2.02 1.29 59 (12) 1.8 2.0 67 (2)
" Time, in weeks after capture, when the whale was tested.
b Cell counts in 109 cells/L

Injection of ACTH significantly increased circulating levels of aldosterone (F1 8 = 8.43, P

<0.05) (Fig. 2). On average, peak aldosterone concentrations were 10-fold greaier than in
controls at each time interval; the greatest relative increase was 2140% in DL16. The large
individual variation in cortisol response obscured any significant effect of ACTH treatment
on the group as a whole (F 1 8 = 1.67). In most of the trials, a two- to three-fold increase was
observed, although cortisol Ievels were elevated nearly six-fold in DL 14. Because of the wide
variation in initial eosinophil counts (3.08 + 1.33 x 10 9 cells/L) among the whales, statistical
comparisons using actual counts at each sampling revealed no significant effect of ACTH.
However when the data were expressed as a percentage of initial values for each whale, the
changes in relative values were readily apparent and highly significant (F 1 8 = 10.05, P <
0.01). Pre-test levels were restored in the 24-h sample. In four of seven triais, lymphocyte
counts decreased by at least 30% after 5-6 h, and in two of these, did not recover by 24 h.
There were few significant changes in any of the other plasma chemical constituents or
hematological elements analysed in samples collected during the stimulation tests. In the 6-
h sample, glucose concentrations had significantly (P < 0.05) increased in the test animals
(7.0 + 0.8 mM/L, n=7 , vs. 6.1 + 0.3 mM/L at time 0, n=7 ), and decreased in the controls (6.0
+ 0.3 mM/L, n=5, vs. 6.6 + 0.5 mM/L at time 0, n=5); comparison of test vs. control for the
6-h sample showed a similar statistical difference (P < 0.05). In the control group, iron

153
 Cortiso l ( nmo l/L)
240 -
200 -
160 -
120
80 1
40
0
1.4
%In itialEosinop hils Aldostero ne (nmo l/ L)

1.2
1
0.8
0.6
0.4
0.2
0

100

80

60

40

20

Hours after Capture

Fla , 1. Plasma cortisol and aldosterone concentrations and rela-
tive changes in circulating eosinophil counts in 3 beluga whales
during the first 6 h after capture. Two whales (DL7 — +; DL8
were given 75 IU of ACTH; DL24 – – – •) received a
control injection of saline.

concentration decreased from 47.2 + 12.1 p,M/L (n=5) at time 0 to 28.3 + 2.1 jiM/L at 12 h,
and 17.3 + 4.7 IIIVI/L at 24 h (n=3 for each), though statistical comparisons are not possible
because of the small sample size. A similar trend was apparent in ACTH stimulated whales,
but wide variation in initial levels (34.3 + 15.8 111 ■4/L) obscured the significance of the
findings.

Discussion

The adrenocortical response to a single injection of ACTH in belugas was comparable to

that in bottlenose dolphins (Thomson and Geraci 1986). The most consistent feature of the
response, which distinguishes it from that in terrestrial mammals, is the large proportional
increase in aldosterone (Johnson-Kocal 1985; St. Aubin and Geraci 1986). We have
previously suggested that this pattern, which is shared to some extent with seals, is an
adaptation to the marine environment (St. Aubin and Geraci 1986).

154
 350
300 -

Cortisol (nmol/L)
250 -
200 -
150
.....

100
f ....
f.......
501
0
1.4 -
Aldosterone (nmo l/L)

1.2 -
1H
0.8 -
0.6 -
0.4 -
0.2 - +. .4 ...... „
0
1201
%Initial Eosinop hils

100-fi

80 - ...... .......

60

40

20 H
0 à 4 à 12 e 24
Time (h)

Fin. 2. Plasma cortisol and aldosterone concentration, and rela-

tive change in circulating eosinophil counts (mean and se) in 7
tests on 5 whales given 75 IU of ACTH (D), and in 5 control tests
on 4 whales ( • ).

The influence of ACTH on aldosterone secretion, though evident in other species including
people (Nicholls et al. 1975; Chan 1979; Seely et al. 1989), is particularly well developed in
cetaceans. The mechanism underlying this adaptation is unclear. Chu and Hyatt (1986)
isolated from rat adrenal a population of zona glomerulosa cells that is highly responsive to
ACTH, and perhaps similar cells are abundant in the outer cortex of the cetacean adrenal. A
directed study is clearly needed to address this hypothesis.
By comparison, the cortisol response was variable and comparatively weak, making this
hormone an unreliable index of the degree of adrenal stimulation in these animals. We
assumed that cortisol is the predominant glucocorticoid in this species, as it was shown to be
in bottlenose dolphins (Thomson and Geraci 1986), and therefore we did not examine
changes in other cortical secretions, such as corticosterone. Perhaps larger doses of ACTH
would have produced greater elevations in cortisol, but in our experience higher levels have
only been observed under pathological conditions such as terminal shock associated with
stranding. Forewarned by the deaths of two bottlenose dolphins given slightly higher doses

155
of ACTH in a previous study (Thomson and Geraci 1986), we selected a more conservative
dose for this investigation. We felt it would be imprudent to place other animals at such risk
by conducting a dose—response trial simply to establish maximal hormone release.
The systemic effects of even modest changes in plasma cortisol are reflected in the
consistent decline in circulating eosinophils, the brief elevation in glucose, and suggestion
of reduced plasma iron. In the same whales, we also noted a concurrent decline in triiodothy-
ronine levels (St. Aubin and Geraci 1988). Perhaps the metabolic systems regulating these
constituents are particularly sensitive to cortisol. Alternatively, cetacean plasma may have
a lower capacity to bind the hormone, and consequently is less able to buffer changes in free
hormone concentration. Studies in progress in our laboratory suggest that the latter may
apply.
Lymphopenia has been associated with corticosteroid administration in bottlenose dol-
phins (Medway et al. 1970), but not in those stressed by handling or stimulated by ACTH
(Thomson and Geraci 1986). In belugas, circulating lymphocytes appear to be more sensitive
to adrenocortical activity associated with initial capture (St. Aubin and Geraci 1989) than to
injection of ACTH. One consistent pattern in both control and experimental groups was that
little change occurred when basal counts were low. These low values suggest that in some
individuals lymphocyte reserves were already depleted, perhaps as a consequence of episodic
glucocorticoid secretion as the animals acclimated to captivity. Fauci and Dale (1974)
observed that hydrocortisone selectively affects T-lymphocytes, and it is possible that these
cells were already depleted in some of the initally lymphopenic animals. We made no attempt
to distinguish between T- and B-cell populations in the belugas.
Within the latter four weeks of the holding period, the two whales tested twice with ACTH
showed some attenuation of the peak aldosterone response, but not enough to suggest an
"escape" from ACTH control. It remains to be established whether continuous stimulation
by ACTH does in fact result in loss of response from the zona glomerulosa in these animals.
In view of the peculiar demands on electrolyte balance in the marine environment, such a
mechanism would appear to be maladaptive in marine mammals.

Acknowledgments

We thank T. Friesen and C. Thomson (Univ. of Guelph), J. Orr and S. Waters (Department
of Fisheries and Oceans), and J. Batstone (Churchill, Manitoba) for invaluable assistance in
the field. C. Thomson (Univ. of Guelph) performed the laboratory analyses, and M. Cattet
(Univ. of Guelph) and D. W. Doidge (McGill University) assisted with data analysis. R.
Moshenko (Section Chief, DFO) coordinated logistic support for the whale capture and field
holding facilities. This study was funded by the Department of Fisheries and Oceans, The
World Wildlife Fund ("Whales Beneath the Ice" Program), the Natural Sciences and
Engineering Research Council (Grant A6130 to JRG), and the United States Office of Naval
Research (Grant N00014-87-G-0114).

References

CHAN, J. C. M. 1979. Control of aldosterone se- FA uci, A. S., AND D. C. DALE. 1974. The effect of
cretion. Nephron 23: 79-83. in vivo hydrocortisone on s ubpopu I a ti ons
CHU, F. W., AND P. J. HYATT. 1986. Purification of human lymphocytes. J. Clin. Invest. 53:
of dispersed rat adrenal zona glomerulosa cells 240-246.
by Percoll density gradient centrifugation and GAILLARD, R. C., A. M. RIONDEL, C. A. FAVROD-
the isolation of a population of cells highly COUNE, M. B. VALLOTOIN, AND A. F. MULLER.
responsive to adrenocorticotrophin. J. Endoci- 1983. Aldosterone escape to chronic ACTH
nol. 109: 351-358. administration in man. Acta Endocrinol. 103:
116-124.

156
JOHNSON-KOCAL, T. E. 1985. Factors affecting roidogenesis and adrenal zona glomerulosa
serum cortisol and aldosterone secretion in morphology in the rat. Acta Endocrinol. 114:
captive harp seals. M.Sc. thesis, University of 47-54.
Guelph, Guelph, Ont. ST. AUBIN, D. J. AND J.R. GERACI. 1986. Adreno-
MAZZOCCHI, G., L. K. MALENDOWICZ, P. REB UF- cortical function in pinniped hyponatremia.
FAT, C. ROBBA, G. GOTTARDO, AND G. G. Mar. Mamal. Sci. 2: 243-250.
NUSSDORFER, 1986. Short- and long-term ef- 1988. Capture and handling stress suppres-
fects of ACTH on the adrenal glomerulosa of ses circulating levels of thyroxine (T4) and
the rat. A coupled sterological and enzymolo- triiodothyronine (T3) in beluga whales, Del-
gical study. Cell Tissue Res. 243: 303-310. phinapterus leucas. Physiol. Zool. 61: 170-175.
MEDWAY, W., J. R. GERACI, AND L. V. KLEIN. 1989. Adaptive changes in hematological
1970. Hematologic response to administration and plasma chemical constituents in captive
of a corticosteroid in the bottle-nosed dolphin beluga whales, Delphinapterus leucas. Can. J.
(Tursiops truncatus). J. Am. Vet. Med. Assoc. Fish. Aquat. Sci. 46: 796-803.
157: 563-565. SEELY, E. W., P. R. CONLIN, G. A. BRENT, AND R.
NICHOLLS, M. G., E. A. ESPINER, AND R. A. Do- G. DLUHY. 1989. Adrenocorticotropin stimu-
NALD. 1975. Plasma aldosterone response to lation of aldosterone: prolonged versus pulsa-
low dose ACTH stimulation. J. Clin. Endocri- tile infusion. J. Endocrinol. Metab. 69:
nol, Metab. 41: 186-188. 1028-1032.
OELKERS, W. 1985. Prolonged ACTH infusion THOMSON, C. A., AND J. R. GERACI. 1986. Corti-
suppresses aldosterone secretion in spite of sol, aldosterone, and leucocytes in the stress
high renin activity. Acta Endocrinol. 108: response of bottlenose dolphins, Tursiops trun-
91-97. catus. Can. J. Fish. Aquat. Sci. 43: 1010-1016.
PAAPE, M. J., C. DESJARDINS, A. J. GUIDRY, R. H. VAN MOURIK, S., AND T. STELMASIAK. 1984. Adre-
MILLER, AND D. R. SMITH. 1977. Response of nal response to ACTH stimulation in Rusa
plasma corticosteroids and circulating leuko- deer (Cervus rusa timorensis). Comp. B io-
cytes in cattle following intravenous injection chem. Physiol. 79A: 581-584.
of different doses of adrenocorticotropin. Am. WILLARD, M. D., K. REFSAL, AND E. THACKER.
J. Vet. Res. 38: 1345-1348. 1987. Evaluation of plasma aldosterone
RIONDEL, A. M., P. REBUFFAT, G. MAZZOCCHI, concentrations before and after ACTH admi-
G. G. NUSSDORFER, R. C. GAILLARD, L. Boc- nistration in clinically normal dogs and in dogs
K HORN, J. NUSSBERGER, M. B. VALLOTTON, with various diseases. Am. J. Vet. Res. 48:
AND A. F. MULLER. 1987. Long-term effects of 1713-1718.
ACTH combined with angiotensin II on ste-

157
 f

158
 Assessment of Immunological Function in
Captive Beluga Whales, Delphinapterus leucas:
Humoral Response to Sheep Red Blood Cell Antigens

D.J. St. Aubin and J.R. Geraci

Wildlife Section, Department of Pathology, Ontario Veterinary College,
University of Guelph, Guelph, Ontario, Canada NI G 2W1

and P. E. Shewen
Department of Veterinary Microbiology and Immunology,
Ontario Veterinary College, University of Guelph,
Guelph, Ontario, Canada NI G 2W1

ST. AUBIN, D. J., J. R. GERACI, AND P. E. SIIEWEN. 1990. Assessment of immuno-

logical function in captive beluga whales, Delphinapterus leucas: humoral
response to sheep red blood cell antigens, p. 159-164. In T. G. Smith, D.
J. St. Aubin, and J. R. Geraci [ed.] Advances in research on the beluga
whale, Delphinapterus leucas. Can. Bull. Fish. Aquat. Sci. 224.

The humoral response to sheep red blood cells (SRBC) was examined in six
beluga whales, Delphinapterus leucas, captured and maintained for 10 wk near
Seal River, Manitoba. Three were sensitized by intraperitoneal injection of the
antigen immediately after capture, and challenged by the same route 12 d later. The
others were similarly tested after 45 and 57 d, respectively, in captivity; one of the
whales in this group was released prior to receiving a challenging dose of SRBC.
A weak antibody response (1/2 dilution) to the sensitizing dose was noted in only
one of the first whales tested. Rapidly rising titers (up to 1/256) were demonstrated
after challenge in all five whales. The findings from this limited first attempt to
probe humoral immunity in a cetacean demonstrate the suitability of the approach,
and suggest that this component of the whales' immune system was unaffected by
the stress associated with acclimation to captivity.

La réaction humorale aux cellules rouge du sang d'agneau (SRBC) fut examinée
chez six marsouins blancs, Delphinapterus leucas, capturés et retenus pendant 10
semaines près de la rivière Seal au Manitoba. Trois furent sensibilisés par injection
intrapéritonéale d'un antigène immédiatement après la capture et provoqués par la
même voie 12 jours plus tard. Un contrôle similaire a été fait chez les autres en
captivité après 45 et 57 jours, respectivement; une baleine de ce groupe fut relâchée
avant d'avoir reçu une dose provocatrice de SRBC. Une faible réaction anticorps
(dilution 1/2) à la dose de sensibilisation a été notée seulement chez une des
premières baleines provoquées. Une augmentation rapide de la titrimétrie (jusqu'à
1/256) fut notée après provocation chez les cinq baleines. Les résultats de ce
premier essai préliminaire d'exploration d'immunité humorale chez un cétacé
démontre le bien fondé de cette approche et suggère que cette composante du
système immunitaire des baleines n'a pas été affectée par le stress associé avec une
acclimatisation à la captivité.

Introduction

The immune response to foreign antigens is a complex of cellular and chemical events. The
ability of an individual to mount such a response can vary according to nutritional (Sheffy
and Williams 1982) and hormonal status (Grossman 1984), and prior exposure to a given

159
challenge. A number of indirect approaches are used to gauge immune competence (Dean et
al. 1979; Muneer et al. 1988), including analysis of lymphocyte subpopulations, lymphocyte
response to mitogenic stimulation, measurement of specific immune mediators, assessment
of antibody levels, and resistance to infectious challenge.With little of such information for
cetaceans, and none for belugas, we are at a loss to determine immunocompetence in free-
ranging individuals and populations.
As a first remedial step, we undertook this study to determine the antibody response in
captive beluga whales injected intraperitoneally with sheep red blood cells (SRBC). The
whales were challenged at different times after capture to assess their ability to mount an
immune response in the face of ongoing physiological and hormonal adjustments to captivity
(St. Aubin and Geraci 1988, 1989). The study thus provides some insight into the sensitivity
of the immune system to a foreign antigen in beluga whales during early confinement.

Materials and Methods

Six beluga whales (3M, 3F) were captured in the Seal River estuary in early July, 1985, and
transported to nearby holding facilities where they were maintained for 10 wk in two 8.2-m
diameter pools containing 3.8 x 10 4 L of seawater drawn and replaced daily from Hudson Bay.
The whales were fed Pacific herring, Clupea harrengus pallasi, supplemented with thiamine
and ot-tocopherol, to provide an average of 40 kcal.ked -1 .
Blood samples drawn from veins in the caudal peduncle or tail flukes were obtained from
four whales at the time of capture, and from all six 3-4 h later after transport to holding
facilities. Thereafter, the whales were sampled at irregular intervals, as part of a routine health
assesment. Blood for hematology was collected into tubes containing the sodium salt of ethy-
lenediaminetetraacetic acid (EDTA), for plasma chemical analysis into heparinized tubes,
and for antibody tests into untreated vials. Hematological and plasma chemical constituents
were analyzed as previously described (St. Aubin and Geraci 1989).
On arrival at the field camp, three of the whales (Group A) were injected intraperitoneally,
with a 10-cm 18-gauge needle, containing 6.3 mL of a 30% (v/v) solution of SRBC. Twelve
days later, they received a second injection of 5.0 mL of an 18% solution. The other three
whales (Group B) were given the same doses, 45 and 57 d after capture.
The antigen was prepared from whole sheep blood, diluted 1:1 with Alsever's solution and
stored at 4°C for up to 1 mo. After centrifugation for 10 min at 1 000 g, the supernatant and
buffy coat were removed. The red cells were resuspended in an equal volume of phosphate
buffered saline (PBS, pH 7.4, 0..15 M), centrifuged again, and the supernatant discarded. An
appropriate volume of PBS was added to produce the final SRBC solutions.
Antibody titers were determined by the hemagglutination technique (Herbert 1978). To
inactivate complement system proteins, the serum samples were held in a water bath at 56°C
for 1 h. Duplicate serial dilutions of test serum in PBS (total volume —49 pt) were made
in U-bottom microtiter plates, to which 25 tL of a 1% solution of SRBC was added. Serum
from a rabbit immunized against SRBC was assayed concurrently as a positive control; fetal
calf serum and PBS served as negative controls. The plates were incubated at room
temperature overnight on a shaker, then examined for evidence of hemagglutination.
Antibody titer is reported as the most dilute preparation yielding at least 50% hemagglutina-
tion.

Results

A weak antibody response (1/2 dilution) to SRBC was detected in one of the three whales
in Group A 5 and 7 d after the first injection; all three animals showed a humoral reaction after

160
a challenging injection on day 12 (Fig.1). Two of the whales achieved titers of at least 1/256,
9 after challenge, whereas the third showed a more gradual rise to a titer of 1/32.
The whales in Group B, tested after more than 6 wk in captivity, reacted similarly to those
in Group A (Fig. 1). No antibodies were detected following the initial sensitizing dose in any
of the animals. At this time, one of the whales was released due to inappettance and weight
loss. In the other two, the challenging dose of SRBC produced an antibody response compa-
rable to that in the whales tested immediately after capture.
Circulating lymphocytes, calculated from total and differential white cell counts, fluctua-
ted widely (Fig. 2 and 3), with no evidence of lymphopenia. No attempt was made to
distinguish between T- and B-cell subpopulations. Counts in Group B animals at the time of
the first SRBC injection were comparable to or exceeded those at capture.

GROUP A
1/256 - •

1/64 -

)9

1/16 -
:
:

W -2- 1/4 à &

I— 0 :

5
• H-
0E
co 2 GROUP B
i=
Z 1/256

1/64

'
1/16

1/4
: /
fi'
11---r-
1 10 20 30 45
DAYS AFTER 1sT INJECTION

FIG. I. Hemagglutinating antibody titers in serum from beluga whales injected with sheep
red blood cells 0 and 12 d (Group A) or 45 and 57 d (Group B) after capture. One whale in
Group B was released prior to the second SRBC injection.

161
C IRCULATINGLYMPHOCYTES

• I . • • • • . . •
0 4 8 12 16 20 24 28

DAYS AFTER CAPTURE

FIG. 2. Circulating lymphocyte counts in three belugas injected with sheep red blood cells
within 12 h after capture, and again 12 d later (arrows).

ab

1— 5
CIRCU LATINGLYMP H OCYTES

0 20 40 60

DAYS AFTER CAPTURE

FIG. 3. Circulating lymphocyte counts in three belugas injected with sheep red blood cells
after 45 d in captivity, and again 12 d later (arrows).

162
 Discussion

Sensitization and challenge with SRBC, a T-cell dependent antigen, was effective in
eliciting an immune response from the whales. Although six weeks in captivity did not appear
to diminish the whales' ability to mount an antibody response, it would be premature to
suggest that the cumulative stress associated with a novel environment, altered diet, and
intensive handling for concurrent studies had no effect on their immune system. The small
sample size available does not permit statistical comparison of the two groups, and the test
itself measures only one aspect of immune function.
The suppressive effect of the glucocorticoid component of the stress response on the
immune system has been documented extensively (Munck et al. 1984). Cortisol and its
various analogs depress circulating lymphocyte counts, particularly T-cells (Fauci and Dale
1974). This mechanism also functions in odontocetes, including belugas (Medway et al.
1970; St. Aubin and Geraci 1989). In blood samples obtained from four whales immediately
after capture and 3-4 h later after transport to holding facilities, we observed a marked decline
in circulating lymphocytes (St. Aubin and Geraci 1989). Apparently normal levels were
restored within a week. Throughout the 10-wk period in captivity, lymphocyte counts
fluctuated, perhaps reflecting the opposing influences of antigenic stimulation and the
glucocorticoid response to confinement and handling.
Glucocorticoids also enhance the rate of gamma globulin catabolism (Levy and Waldman
1970). In the captive whales, we noted no significant change in serum concentrations of
gamma globulins (data not shown), though routine electrophoretic analysis would not have
revealed changes in immunoglobulin fractions involved in primary and secondary responses
to specific antigenic stimulation.
This preliminary study was undertaken to address the issue of stress-related immunosup-
pression in a cetacean. Further research is clearly needed on more subjects before the question
may be answered. Nevertheless, the technique of sensitization to SRBC antigen proved to be
effective in generating a response from the immune system of belugas, and with continued
application, should produce a pattern of expected responses that will be useful in assessing
the immunocompetence of captive and free-ranging whales.

Acknowledgments

We thank T. Friesen and C. Thomson (University of Guelph) for field and laboratory
assistance. Valuable field logistic support was provided by R. Moshenko (Section Chief), J.
Orr, and S. Waters (Department of Fisheries and Oceans). Funding for the study was provided
by the Department of Fisheries and Oceans, the World Wildlife Fund ("Whales Beneath the
Ice" Program), the Natural Sciences and Engineering Research Council (Grant A6130) and
the United States Office of Naval Research (Grant N00014-87-G-0114).

References

DEAN, J. H., M. L. PARDARATSINGH, AND T. R. GROSSMAN, C. J. 1984. Regulation of the immune

JERRELLS. 1979. Application of immunocom- system by sex steroids. Endocrine Rev. 5:
petence assays for defining immunosuppres- 435-455.
sion. Ann. N.Y. Acad. Sci. 320: 579-590. HERBERT, W. J. 1978. Passive hemagglutination
FAUCI, A. S., AND D. C. DALE. 1974. The effect of with special reference to the tanned cell tech-
in vivo hydrocortisone on subpopulations of nique, Chapter 20, p. 1-20. In D.M. Weir [ed.]
human lymphocytes. J. Clin. Invest. 53: Handbook of experimental immunology. 3rd
240-246 , edition. Blackwell Scientific Publ., Oxford,
U.K.

163
LEVY, A. L., AND T. A. WALDMAN. 1970. The ef- ST. AUBIN, D. J., AND J. R. GERACI. 1988. Capture
fect of hydrocortisone on immunoglobulin me- and handling stress suppresses circulating le-
tabolism. J. Clin. Invest. 49: 1679-1684. vels of thyroxine (T4) and triiodothyronine
MEDWAY, W., J. R. GERACI, AND L. V. KLEIN. (T3) in beluga whales Delphinapterus leucas.
1970. Hematologic response to administration Physiol. Zool. 61: 170-175.
of a corticosteroid in the bottle-nosed dolphin 1989. Adaptive changes in hematologic
(Tursiops truncatus). J. Am. Vet. Med. Assoc. and plasma chemical constituents in captive
157: 563-565. beluga whales, Delphinapterus leucas. Can.
MUNCK, A., P. M. GUYRE, AND N. J. HOLBROOK. J. Fish. Aquat. Sci. 46: 796-803.
1984. Physiological functions of glucocorti- SHEFFY, B. E., AND A. J. WILLIAMS. 1982. Nutri-
coids in stress and their relation to pharmaco- tion and the immune response. J. Am. Vet.
logical actions. Endocrine Rev. 5: 25-44. Med. Assoc. 180: 1073-1076.
MUNEER, M. A., I.O. FARAH, J. A. NEWMAN, AND
S.M. GOYAL. 1988. Immuno-suppression in
animals. Br. Vet. J. 144: 288-301.

164
 Organochlorine Contaminants in Belugas, Delphinapterus leucas,
from Canadian waters

D. C. G. Muiri, C.A. Ford, R.E. A. Stewart

Department of Fisheries and Oceans, Freshwater Institute,
501 University Crescent, Winnipeg, Manitoba, Canada R3T 2N6.

T.G. Smith
Department of Fisheries and Oceans, Arctic Biological Station,
555 Blvd. St. Pierre, Ste-Anne-de-Bellevue, Québec, Canada H9X 3R4

R.F. Addison, M.E. Zinck

Department of Fisheries and Oceans, Bedford Institute of Oceanography,
Dartmouth, Nova Scotia, Canada B2Y 4A2

and P. Béland
St. Lawrence National Institute of Ecotoxicology,
Rimouski, Québec, Canada G5 L 3AI

MUIR, D. C. G., C. A. Foin), R. E. A. STEWART, T. G. SMITH, R. F. ADDISON, M. E.

ZINCK, AND P. BÉLAND. 1990. Organochlorine contaminants in belugas,
Delphinapterus leucas, from Canadian waters, p. 165-190.
In T. G. Smith, D. J. St. Aubin, and J. R. Geraci [ed.] Advances in research
on the beluga whale, Delphinapterus leucas. Can. Bull. Fish. Aquat. Sei.
224.

Concentrations of organochlorine compounds were compared in blubber sam-

ples from the six stocks of beluga whales (Delphinapterus leucas) in Canadian
waters. Highest concentrations of PCB congeners (EPCB), DDT-group (DDT),
polychlorinated camphenes (PCCs), chlordane-related compounds (ECHLOR),
mirex and dieldrin were found in animals from the St. Lawrence estuary. Mean
concentrations of hexachloro-cyclohexane isomers (EHCH), total chloroben-
zenes (ECBz), and ECHLOR in females, showed few significant differences bet-
ween any of the stocks. Mean concentrations of EPCB in arctic male belugas
ranged from 2.53 ± 0.57 [tg/g in animals from Jones Sound to 4.65 ± 0.94 [tg/g (wet
wt) at Cumberland Sound. EPCB, EDDT and mirex concentrations were 25, 32
and 100-fold higher, respectively, in males from the St. Lawrence than average
levels in arctic males. The St. Lawrence animals are clearly exposed to higher
levels of PCBs as well as other pollutants than the arctic animais. Continued study
of the St. Lawrence an imals is needed because toxicological effects of organochlo-
rines are more likely to occur in this stock before becoming apparent in other
beluga stocks.

Les concentrations de certains composés organochlorés ont été comparés dans

des échantillons de tissus adipeux des six cheptels de bélugas (Delphinapterus
leucas) des eaux canadiennes. Les concentrations les plus fortes des congenères de
BPC (EBPC), du groupe DDT (DDT), des camphènes polychlorés (CPC), des

' Author to whom correspondence should be addressed.

165
 composés apparentés au chlordane (ECHLOR) de mirex et de dieldrine furent
retrouvés dans les animaux de l'estuaire du Saint-Laurent. Les concentrations
moyennes des isomères de Phexchlorocyclohexane, des chlorobenzenes totaux
(CBz) et des ECHLOR pour les femelles présentèrent peu d'écarts significatifs
entre les divers cheptels. L'écart des concentrations moyennes de EBPC chez les
bélugas mâles de l'Arctique a varié de 2,53 ± 0,57 gg/g (poids humide) au détroit
de Jones à 4,65 ± 0,94 1.tg/g au détroit de Cumberland. Les concentrations de
EPBC, de 1,DDT et de mirex étaient 25, 32 et 100 fois, respectivement, plus
élevées chez les bélugas mâles du Saint-Laurent par rapport au niveau moyen noté
chez les mâles de l'Arctique. Il est évident que les bélugas du Saint-Laurent sont
exposés à des teneurs plus élevées de BPC et d'autres éléments polluants que les
animaux de l'Arctique. Le cheptel de bélugas du Saint-Laurent doit donc faire
l'objet d'études poursuivies étant donné qu'il est plus probable que les indicences
toxicologiques des organochlorés se manifesterant en premier chez ces animaux.

Introduction

As top predators in marine food chains, belugas (Delphinapterus leucas) are an excellent
species for studying the presence and geographic variation of lipophilic organochlorine
chemicals such as polychlorinated biphenyls (PCBs) and chlorinated pesticides. At the same
time belugas, like other predatory mammals and birds, may be vulnerable to ecotoxicologi-
cal effects of some organochlorines such as reproductive failure (Reijnders 1986; Peakall
1975) and hepatic mixed function oxidase enzyme induction (Walker 1983; Safe 1984).
Analysis of stranded belugas from the Gulf of St. Lawrence estuary has shown accumulation
of PCBs and DDT-related compounds (Sergeant 1980; Martineau et al. 1987) at levels
suggested to cause the low reproductive rate of this population (Martineau et al. 1987;
Sergeant and Hoek 1988).
There has been little study of the levels or geographic variation of organochlorine
contaminants in belugas, or other small, toothed whales, inhabiting Canadian Arctic waters.
The four Arctic beluga stocks (Hudson Bay, Cumberland Sound, High Arctic (Jones Sound)
and Beaufort Sea)(Stewart 1986) could conceivably be exposed to different levels of
organochlorine pollutants depending on the remoteness of the marine and estuarine environ-
ments that they inhabit. Norstrom et al. (1988) found that polar bears in Hudson Bay had
generally higher levels of organochlorine contaminants in fat and liver tissues than did
animals further north, which they found consistent with expected greater loadings via
continental runoff and atmospheric deposition to this region. In the only report to date on
organochlorine contaminants in Arctic belugas, Addison and Brodie (1973) found 3.9 lig/g
total DDT and low levels of PCBs (<0.5 ggig detection limit) in Beaufort Sea belugas.
This survey of organochlorines in Arctic belugas was undertaken to examine possible
geographic variations of the pollutants in the various stocks and to compare the levels of types
of contaminants in Arctic animals with those in the St. Lawrence population. An additional
objective of this study was to provide information on the dietary exposure of native people
in the Arctic who utilize whale tissues as part of their traditional diet.

Materials and Methods

Samples

Beluga tissues were obtained from animals taken by Inuit hunters at five locations in the
Canadian Arctic between 1983 and 1987 (Table 1). Samples were packaged in polyethylene
bags, frozen and held at -40°C until analysis. Blubber samples from the carcasses of stranded

166
TABLE 1. Sampling location and characteristics of beluga blubber samples."

Stock Sampling site Sex N Date Age Length % Lipidb

(yr) (cm)

E. Hudson Nastapoka R. M 8 1984,85 12.0 ± 9.5 311 ± 41 86.5 ±2.2

Bay & Belcher Is. 5.0 - 33.0 278 -401 83.9 - 89.4

F 8 1984,85 11.9 ± 6.4 314 ± 31 86.4 ± 3.0

3.0 - 20.0 252 - 346 80.7 - 89.6

E. Hudson Nastapoka R. M 6 1987 15.6 ± 3.0 347 ± 22 85.8 ± 5.9

Bay 12.0 - 20.0 324 - 387 81.7 -97.2

F 6 1987 17.0 ± 6.3 316 ± 18 88.7 ± 2.9

8.0 -24.0 307 -342 84.5 - 91.6

Cumberland Pangnirtung M 6 1983 7.3 ± 6.5 415 ± 40 83.9 ± 6.2

Sound 0- 16 335 -438 77.6 - 88.0

F 6 1983 8.1 ± 7.3 375 ± 15 87.7 ± 1.5

0 - 17 361 - 395 86.4 - 90.4

W. Hudson Eskimo Point M 4 1986 13.0 ± 4.8 351 ±33 66.1 ±3.1
Bay 8.0 - 19.0 309 - 385 62.7 - 70.2

F 4 1986 10.3 ± 4.1 312 ± 18 74.1 ± 12.7

5.0- 15.0 285 -325 56.5 - 85.4

Beaufort Mackenzie M 10' 1983,87 17 431 ±25 74.1 ± 10.0

Sea Bay 13 - 20.0 390 -472 60.4 - 88.0

F 2d 1983, 87 18 361 81.7

- 77.9 - 85.6

Jones Sound M 8 1984 4.4±2.2 331 ± 71 88.4±2.8

1.0 - 7.5 253 -432 84.3 - 92.8

F 7 1984 4.6 ± 2.9 308 ± 56 87.5 ± 1.9

1.5- 10.5 254 - 424 84.5 - 90.1

St. Lawrence M 4 1986, 87 17.5 ± 9.1 416 ± 68 86.8 ± 2.7

estuary 4.0 - 23.5 315 -453 83.2- 89.6

F 5 1986, 87 15.6 ± 10.4 352 ± 43 86.6 ± 3.9

2.5 - 29.0 277 - 382 80.4 - 91.4

Retrospective Baffin Is. F 1 1966 or 67 adult - 100

Samples M 1 1966 or 67 <1 - 100
Churchill - 1" 1966 or 67 - - 100
" Mean ± standard deviation and range.
b Hexane extractable lipids.
Ages available for 2 animals only.
d Age and length available for 1 animal only.
C
Blubber oil - not fresh tissue - from male and female whales.

167
belugas in the St. Lawrence estuary were collected and preserved as described by Martineau
et al. (1987). Prior to analysis, samples were partially thawed and subsampled taking care to
exclude the exterior portion of tissue in contact with plastic.
Analyses for organochlorines were performed on blubber samples from all locations. Liver
from Cumberland Sound animals and muscle from the eastern Hudson Bay stock where also
analysed. Ages (years or growth layer groups), sex, length and maximum girth were available
for most of the animals.
In addition to fresh tissue samples, one sample of crude beluga oil obtained from the
commercial harvest of whales at Churchill (MB) and two samples of beluga blubber from
Baffin Island collected in 1966/67 (provided by Dr. R.G. Ackman, Technical University of
Nova Scotia, Halifax, N.S.) were included to provide retrospective organochlorine analysis.
The sample from Churchill was taken from the whale oil manufacturing process and had
therefore been subjected to heat whereas the two samples from Baffin Island had been
extracted at room temperature to recover fats.

Extraction and Separation

Determinations of organochlorines in blubber tissues followed the procedures described by

Muir et al. (1988a). Briefly: samples of blubber (2 g) were mixed with anhydrous sodium
sulfate and ball-milled (30 min) with hexane. The extract was centrifuged and a portion
(1/10) removed for lipid determination. The extract was chromatographed on a gel pemlea-
tion column (GPC) to separate chlorinated hydrocarbons from co-extractive lipids, and then
fractionated on Florisil into three eluates — hexane (F1), hexane:DCM (85:15) (F2), and
hexane:DCM (1:1) (F3) (Norstrom and Won 1985). The chromatography on Florisil sepa-
rated PCBs, chlorobenzenes, 4,4'-DDE and mirex in Fi from most polychlorinated camphe-
nes (PCCs), chlordane-related compounds and 4,4'-DDT in F2. F3 contained heptachlor
epoxide and dieldrin.
Twelve blubber and muscle samples from E. Hudson Bay were analysed by the method of
Addison et al. (1986). Samples were extracted by blending with hexane and the extracts were
"cleaned up" on Florisil. PCBs were confirmed by treatment of the Florisil eluates with KOH.
Liver samples were homogenized, mixed with anhydrous sodium sulfate and extracted
with hexane. The hexane extract was evaporated to small volume and a portion removed for
lipid determination. Liver extracts were subjected to GPC and Florisil chromatography as
described above.
Blubber and liver extracts were chromatographed on a 60 m x 0.25 mm DB-5 column with
H, carrier gas using gas chromatographic (OC) conditions described in previous studies
(Muir et al. 1988a). Muscle and blubber extracts from E. Hudson Bay animals were chroma-
tographed on a 30 m x 0.25 mm DB-5 column with He carrier gas.
Organochlorines were confirmed by OC-mass spectrometry (GC-MS) on a 30m x 0.25 mm
DB-5 column using HP-5970 mass selective detector (MSD). Additional mass spectrometry
was performed on an HP 5987B, operated in electron impact mode at 15 eV, at the National
Wildlife Research Centre of the Canadian Wildlife Service. GC-MS conditions were
identical to those of Norstrom et al. (1988). The presence of chlordane-related compounds,
PCCs and mirex was also confirmed by treatment of samples with HNO3 :H2SO4 (1:1) to
destroy chlorinated aromatics (PCBs, DDT, HCB) (Musial and Uthe 1983).
PCBs and organochlorine pesticides were identified and quantified as described previously
(Muir et al. 1988a). Total PCB (EPCB) was the sum of all congeners. Total chlordane
(ECHLOR) was the sum of all chlordane-related compounds including heptachlor epoxide.
PCCs were quantified by a modification of previous procedures (Muir et al. 1988a): response
factors for individual PCC peaks were calculated from the weight percent of each peak in the
total ion chromatogram of a toxaphene standard. Total PCCs was the sum of the areas of up
to 20 peaks (8 major peaks were generally observed) (Muir et al. 1990a).

168
Data Analysis

Organochlorine concentrations in beluga blubber samples were log 10 transformed prior to

calculation of means, standard errors and correlation coefficients. The general linear models
procedure of SAS (SAS Institute 1987) was used with an analysis of covariance (ANCOVA)
model to examine the main effects of location and sex, and regressor effects of length and age
on organochlorine concentrations. Comparisons between mean concentrations of each
organochlorine, ratios of various compounds, length and age at each location were made with
Tukey's test using the mean square error from the ANCOVA. Comparisons between mean
concentrations in males and females were made with the Student's t-test.

Results

Samples

A total of 88 samples of beluga blubber from six locations were analysed for organochlo-
rine contaminants in this survey (Table 1). Ages were available for samples from Jones
Sound, Cumberland Sound, Hudson Bay and the St. Lawrence; lengths were available for
almost all animals. The St. Lawrence belugas were generally older than the Arctic animals,
with a mean age of 15.6 ± 10.4 yr for females and 17.5 ± 9.1 yr for males. However, there was
wide variation in ages within all groups, and only the animals from Jones Sound were
significantly younger (P <0.05) than those from the St. Lawrence estuary. The St. Lawrence
animals included here were similar in age to the group analysed by Martineau et al. (1987).
The Beaufort Sea belugas that were analysed were significantly larger (P <0.05) than the
Hudson Bay and Jones Sound animals. Unfortunately, ages were available for only three of
the 12 Beaufort Sea animals; they were significantly older than the Jones Sound group and
were similar to the average age of the St. Lawrence animals.

Identification of Individual Contaminants in Blubber

The major organochlorine contaminants in beluga blubber from all six locations surveyed
were 1,PCBs, PCCs, E DDT and ICHLOR (Table 2). Positive identification of all major
components in blubber extracts from the Western Hudson Bay stock and from the St.
Lawrence was made by full scan mass spectrometry and by use of multiple ion chromato-
grams (Fig. 1 and 2).
A total of 60 PCB peaks (approximately 65 congeners because some co-eluted) could be
routinely determined in beluga blubber samples from all locations (Fig. 1). To facilitate
discussion, the PCB congeners are referred to by their IUPAC number (Ballschmiter and Zell
1980). Twelve congeners accounted for >50% of EPCB in all samples: tetrachloro-52 and
-66, pentachloro-95, -101, -99, and -118, hexachloro -149,-151,-153 and -138; heptachloro-
187/182 and -180 (Fig. 3 and 4).
PCC was the major organochlorine contaminant detected in blubber of Arctic belugas
(Table 2). Mean PCC concentrations were about 2-fold higher than ECHLOR and 1.5 fold-
higher than 1,PCB and 1,DDT in the Arctic animals. Fourteen PCC peaks could generally be
distinguished in beluga blubber extracts (Fig. 2). Two peaks, T2 and T12 in Fig. 2,
predominated in all samples. T2 eluted mainly in Fi along with PCBs but is shown together
with other PCC peaks in the multiple ion chromatogram because both fractions were
combined for GC-MSD analysis. Peaks T2 and T12 were previously identified in dolphin
blubber extracts as octa- and nonachloro- camphenes, respectively (Muir et al. 1988a).
The major DDT component in beluga blubber was 4,4'-DDE which represented 66% and
57% of total DDT-group compounds (DDT) in male and female St. Lawrence belugas,
respectively, and from 47 to 57% of EDDT in Arctic belugas (Table 2). Other major DDT

169
TABLE 2. Mean concentrations (J..t.g/g wet wt ± SD) of major organochlorine pesticides in beluga blubber from Canadian waters.'
Stock Sex HCB ECBz -yHCH EHCH t-noua- ICHLOR 4,4'-DDE EDDT PCC mirex dieldrin
chlor
E. Hudson M 0.30 ± 0.19 0.32 ± 0.20 0.16 ± 0.05 0.21 ± 0.06 0.59 ± 0.11 1.86 ± 0.35 1.28 ± 0.42 2.27 ± 0.68 4.13 ± 0.82 0.02 ± 0.01 0.28 ± 0.09
Bay 1984/85 0.15 - 0.66 0.17 - 0.70 0.10- 0.23 0.14 - 0.30 0.44 -0.72 1.46 - 2.41 0.65- 1.97 1.23 - 3.27 3.24 - 5.67 0.02 - 0.04 0.18 - 0.45

F 0.14 ± 0.12 0.16 ± 0.13 0.11 ± 0.03 0.15 ± 0.04 0.28 ± 0.18 0.87 ± 0.58 0.56 ± 0.42 0.98 ± 0.73 1.99 ± 1.10 0.01 ±0.01 0.14 ± 0.10
0.05 -0.39 0.07 - 0.46 0.07 - 0.15 0.09 - 0.21 0.12 - 0.55 0.38- 1.79 0.15- 1.33 0.30 - 2.25 1.07 - 3.99 0.01 ± 0.02 0.07 - 0.34

E. Hudson M 0.29 ± 0.17 0.21 ± 0.07 3.09 ± 0.54 5.53 ± 0.67 0.26 ± 0.13
Bay 1987 0.19 -0.64 0.12 - 0.34 2.27 - 3.94 4.36 - 6.33 <0.01 - 0.33

F 0.14 ± 0.08 0.16 ± 0.05 0.88 ± 0.99 1.96 ± 2.05 0.11 ± 0.16
0.09 - 0.29 0.10 - 0.24 0.18 - 2.68 0.35 - 5.64 <0.01 - 0.40

W. Hudson M 0.61 ± 0.05 0.66 ± 0.05 0.16 ± 0.05 0.24 ± 0.07 0.90 ± 0.14 2.33 ± 0.26 1.80 ± 0.23 3.13 ± 0.20 5.10 ± 0.42 0.03 ± 0.00 0.36 ± 0.07
Bay 0.58 - 0.68 0.61 -0.72 0.12 - 0.24 0.18 - 0.34 0.79- 1.09 2.03 - 2.67 1.46- 1.98 2.96 - 3.38 4.66 - 5.63 0.02 - 0.03 0.25 - 0.40

F 0.19 ± 0.18 0.23 ± 0.20 0.11 ± 0.02 0.15 ± 0.04 0.30 ± 0.30 0.85 ± 0.80 0.49 ± 0.57 0.85 ± 0.96 1.77 ± 1.41 0.01 ± 0.00 0.14± 0.12
0.08 - 0.46 0.11 -0.53 0.09 - 0.14 0.12 - 0.21 0.13 - 0.75 0.36 - 2.05 0.11 - 1.34 0.18 - 2.25 0.69 - 3.83 0.00 - 0.01 0.07 - 0.32
Retrospec- F 0.06 0.07 0.03 0.04 0.05 0.13 0.24 0.93 0.66 <0.01 0.05
tive samples'
M 0.07 0.09 0.03 0.04 0.04 0.11 0.13 0.54 0.55 <0.01 0.05

M&F 0.30 0.35 0.06 0.09 0.16 0.44 1.81 3.77 0.66 <0.01 0.17

Cumberland M 0.96 ± 0.18 1.05 ± 0.18 0.23 ± 0.08 0.39 ±0.11 1.05 ±0.27 2.38 ± 0.40 3.38 ± 1.02 6.83 ± 1.89 5.78 ± 5.39 0.01 ± 0.01 0.91 ± 0.26
Sound 0.74- 1.26 0.83- 1.34 0.10 - 0.32 0.23 - 0.56 0.79- 1.48 1.99 - 3.05 2.22 - 4.93 4.19 - 9.73 1.61 - 12.7 0.01 -0.01 0.61 - 1.19

F 0.18 ± 0.04 0.24 ± 0.06 0.23 ± 0.05 0.24± 0.06 0.18 ± 0.15 0.62 ± 0.15 0.40 ± 0.28 0.93 ± 0.55 1.77 ± 1.76 0.01 ± 0.00 0.20 ± 0.33
0.13 -0.24 0.17 - 0.32 0.14 - 0.26 0.19 - 0.32 0.07 - 0.29 0.39 - 0.78 0.16 - 0.95 0.45- 1.96 0.49 - 4.61 0.00 - 0.01 0.15 - 0.24

Beaufort M 0.59 ±0.13 0.65 ± 0.15 0.13 ± 0.04 0.23 ± 0.06 0.53 ± 0.12 1.75 ± 0.41 1.03 ± 0.44 2.20 ± 0.83 3.83 ± 1.16 0.04 ± 0.01 0.23 ± 0.05
Sea 0.40 - 0.78 0.43 - 0.85 0.06 - 0.20 0.14 - 0.32 0.32 - 0.72 1.28 - 2.56 0.15- 1.76 1.47 - 3.73 2.55 -6.62 0.02 - 0.06 0.16 - 0.34

Fd 0.29 0.33 0.10 0.17 0.21 0.67 0.35 0.67 1.38 0.02 0.10
0.16 - 0.41 0.19 -0.47 0.09 - 0.12 0.17 - 0.17 0.11 -0.31 0.44- 0.89 0.19 -0.52 0.46 - 0.88 1.12- 1.63 0.01 -0.03 0.07 - 0.14
TABLE 2. (continued)
Stock Sex HCB ICBz -yHCH t-nona- ECHLOR 4,4'-DDE E DDT PCC mirex dieldrin
chlor
Jones Sound M 0.50 ± 0.21 0.55 ± 0.22 0.12 ± 0.07 0.19 ± 0.09 0.59 ± 0.13 1.87 ± 0.44 1.07 ± 0.21 1.96 ± 0.32 4.25 ± 1.02 0.01 ± 0.00 0.34 ± 0.11
0.27 - 0.83 0.30 - 0.87 0.06 - 0.29 0.11 -0.40 0.44 -0.76 1.34 - 2.53 0.74- 1.36 1.42 - 2.34 2.96 - 5.40 0.01 -0.02 0.21 -0.48

F 0.39 ± 0.21 0.44 ± 0.23 0.10 ± 0.05 0.16 ± 0.08 0.58 ± 0.35 1.84± 1.13 1.26± 1.08 2.19 ± 1.69 3.74 ± 2.12 0.01 ±0.01 0.33 ±0.23
0.08 -0.70 0.11 -0.79 0.06 - 0.18 0.10 - 0.28 0.29- 1.32 0.97 -4.25 0.56- 3.69 1.13 -5.95 1.77 - 8.08 0.01 -0.03 0.09 -0.81

St. Lawrence M 1.34 ± 0.44 1.34 ± 0.44 0.10 ± 0.01 0.37 ± 0.11 3.67 ± 0.30 7.43 ± 0.63 65.8 ± 19.1 101 ± 32.6 14.7 ± 2.46 1.00 ± 0.64 0.93 + 0.12
estuary 0.82- 1.90 0.82- 1.90 0.09 -0.12 0.28 -0.51 3.25 -3.91 6.54 - 8.02 38.5- 80.6 52.4- 123 11.7 - 17.6 0.19 - 1.54 0.81 - 1.06

F 0.60 ± 0.43 0.60 ± 0.43 0.11 ±0.03 0.24 ± 0.10 1.86 ± 0.86 3.55± 1.99 13.9 ± 11.3 23.0± 17.3 6.34 ± 3.51 1.11 ±0.99 0.56 ± 0.31
0.22- 1.27 0.22- 1.27 0.08 - 0.14 0.12 - 0.32 0.87 - 2.82 1.50 - 6.38 1.73 - 26.8 3.95 - 42.7 2.05- 10.3 0.38 - 2.66 0.21 -0.87
• Arithmetic mean ± standard deviation and range of concentration.
• Not determined.
• Consisting of 1 adult female, 1 newborn male and a sample of beluga oil (Table 1).
• No standard deviation, N=2.
 DDT isomers 8i metabolites
1.1E8
4,4' - COD
rn/z 165+235+246
1.2E6
4,4' - ODE
Lj
Fl +F2

1.0E6

8.0E5

G.0E5

4.0E5

0 2.0E5
0

73 2.5E5

1] PCB congeners
2.0E5 -1 m/e- 258+292+326+360
+394+430
r-
1.5E5 -

- 1.0E5

5.0E41
_ 8ed

• •
20 25 30 35

Retention time (minutes)

FIG. 1. Capillary GC-MSD multiple ion chromatograms (30 m x 0.25 mm id. DB-5 column) of a beluga
blubber extract from Western Hudson Bay showing characteristic ions of tri- to octachloro biphenyls
(lower chromatogram) and DDT-related compounds. Major PCBs are identified by their IUPAC
numbers ,

components quantified in all beluga samples were 4,4'DDT, 4,4'DDD and 2,4'-DDT (Fig. 1).
2,4'-DDT and 4,4'-DDD are not separated in the chromatogram because Fi and F2 were
combined for the GC-MSD analysis; however, the 2,4'-isomer was quantified by GC because
it eluted mainly in Fi. 2,4'-DDD was also detected (Fig. 1) but was not determined by GC.
Fourteen chlordane-related compounds were routinely determined in beluga blubber
extracts from all locations (Fig. 2 and 5). Trans-nonachlor was the major chlordane-related
compound accounting for 30 — 40% of ECHLOR in Arctic animais and for 50 — 52% of
ICHLOR in the St. Lawrence stock. Cis-nonachlor and nonachlor III (1,2,2,4,5,6,7,8,8-
nonachlor; Norstrom et al. 1988) were also major chlordane components especially in Arctic
animais (Fig. 2). Ul or photoheptachlor was also more prominent in Arctic samples (Fig. 5).
The extent of transformation of chlordane is evident by the fact that cis-and trans- chlordane
were minor peaks although they form a major proportion of technical chlordane (Fig. 5).
Nonachlor III and cis-chlordane co-elute in Fig. 2 but they were separated by chromatogra-
phy on Florisil, with nonachlor III eluting in F1, and determined separately by GC.
Metabolites oxychlordane and heptachlor epoxide also formed a significant fraction of

172
 Toxaphene-related compounds
m/e 307+343
Fl+F2

112

T3

110
TI1
T4 T5
T8 TI3
Abundance

T17
TI4T16

1.0E5 –

9.0E4:
tnonachlor Chlordane-related compounds
8.0E4
mte 337+371+387+405
7.0E4 Fl+F2
c-chlordane
8.0E4

5.0E4 nonachlarIll c-nonachlor

t-chlordone l TI2
4.0E4
oxychlordane I
3.0E4

2. 0E4
T6
1 0000: L-JI CI S'23.:\
0 — 1441,14,64 41, irilt1410%4
15 28 25 38 35

Retention time(minutes)

Fia. 2. Capillary GC-MSD multiple ion chromatograms of a beluga blubber extract from Western
Hudson Bay showing characteristic ions of chlordane-related compounds (lower chromatogram) and
polychlorinated camphenes (toxaphene). GC conditions as in Fig. 1.

ECHLOR (Fig. 3). Cl is a monohydro analog of trans-chlordane while C2 and C3 are

heptachloro- compounds with mass spectra having a base peak at m/z 339 (Norstrom et al.
1988). C4 is a 5-hydro analog of trans-nonachlor. Compound "C," a heptachloro component
of technical chlordane (Fig. 5) (Miyazaki et al. 1985), is not shown in Fig. 2, but was identified
by its correct retention time and characteristic ions.
Hexachlorocyclohexane isomers (a-, B-, -y-HCH) were detected in all beluga blubber
samples at relatively low levels compared to PCBs, PCCs, ECHLOR and y, DDT (Table 2).
Hexachlorobenzene (HCB) was the major form of total chlorobenzenes (ECB z) in all beluga
samples surveyed (Table 2). In St. Lawrence animals, tetra and pentachlorobenzenes were
undetectable (<0.01 p g/g). However the detection limits were set relatively high because of
the high levels of PCBs in the same Florisil eluate. Tetra- and pentachlorobenzene isomers
constituted <10% of ICBz in Arctic beluga blubber. Mirex and dieldrin were also detected
in all samples (Table 2).

173
 14 J
12J East Hudson Bay
10 —
8—
6
4—
2—
0 I i fil — I
Percen t of Tota lPCB

14 — MI Males
12 —
St. Lawrence estuary • Females
10 —
8—
6—
4—
2— ,
. , , .
0 .

14 — Aroclor
12 —
124212541260 (111)
10 —
8—
6—
v J 1
e2LI ii _ . .
i .1 111111 . 11d .. .111_11s.A.1.11.11—r s
0
ete.eeeeteeeeeeeeeeee*eeeeee.eeeo$e.oeoeeeeeeeeeeeeeeeeeme
PCB Congener

FIG. 3. PCB congeners as a percent of EPCB in an Aroclor standard mixture and in beluga blubber from Hudson Bay and the St. Lawrence estuary.
Comparisons Among Stocks

In all beluga stocks, except for the animals from Jones Sound, mean EPCB levels were
significantly higher (t-test at P <0.01) in males than in females (Table 3). Statistical
comparisons between the stocks for PCBs and other organochlorines were therefore done by
sex. As expected from the work of Martineau et al. (1987), male and female belugas from the
St. Lawrence had significantly higher mean EPCB and 1, DDT levels (Tukey's test, P <0.05)
than mean levels in any Arctic stocks. Mean EPCB levels were 25-fold higher, for both males
and females, in the St. Lawrence animals than in the Arctic animals while for EDDT the
differences were 32-fold and 19-fold, respectively. Although only a limited number of St.
Lawrence animals are reported here, mean PCB concentrations as Aroclor equivalents (119
± 25.5 ig/g; Table 3) were similar to those reported for a set of 26 animals by Martineau et
al. (1987), who found 185 ± 79 lig/g in blubber of males and 88 ± 165 gg/g in females. The
high standard deviation for females was due to one very high result (576 l'g/g). With this
result omitted, the mean for females reported by Martineau et al. (1987) reduces to 43 ± 31
j.tg/g which is not significantly different from the mean (54.2 ± 37.61.1g/g Aroclor equivalents)
for 5 females reported in Table 3. EDDT levels reported by Martineau et al. (1987) of 91 ± 58
1.1,g/g (15 males) and 21 ± 28 lig/g (11 females) were virtually identical to those reported here
although 2,4'-DDT was not determined in the latter study. Among Arctic stocks, ANCOVA
showed a significant effect of location, but not of length or length-location interaction for
y,PCBs and EDDT in males, while female belugas showed no effects for these factors.
Comparisons of mean EPCB and EDDT levels in males showed samples from Cumberland
Sound had significantly higher PCBs (P <0.05) than those from E. Hudson Bay or Jones
Sound. EDDT levels in males from Cumberland Sound were also significantly higher than
males from the Beaufort Sea.
The EPCB/DDE ratio of 5.1 for the Beaufort Sea males was significantly higher than that
for the St. Lawrence or Hudson Bay stocks while the ratio for the Jones Sound stock was
significantly higher (P <0.05) than for the St. Lawrence stock (Table 4). No significant dif-
ferences for the EPCB/DDE ratio were observed for females. The mean DDE/EDDT ratios
for both males and females in the St. Lawrence stock were significantly higher (P <0.05) than
means for the Arctic stocks.
The large differences in concentrations observed for EPCB and EDDT between Arctic and
St. Lawrence animals were not observed for PCC and ECHLOR. PCC levels in male and
female St. Lawrence belugas (means of 14.7 ± 2.5 and 6.3 ± 3.5 p.g/g, respectively) were
significantly higher (P <0.05) than the mean levels at all Arctic locations, but differences
were only 2 to 3 fold (Table 2). Mean levels of ECHLOR in male St. Lawrence belugas (7.4
± 0.6 i'g/g) were significantly higher (P <0.05) than concentrations in all Arctic stocks by 2
to 3 fold (Table 2). ECHLOR levels in female St. Lawrence belugas (3.5 ± 2.0 ltg/g) were
not significantly different from those in females from Jones Sound (1.8 ± 1.11.1g/g).
There were no significant differences (P <0.05) in PCC levels among locations for Arctic
belugas. There was no significant effect of location, age or length using either the entire set
of data for ECHLOR in females or on separate analysis of the Arctic animals. ECHLOR/PCB
ratios for St. Lawrence belugas were 5- to 7-fold lower than those for Arctic animals (Table
4). Mean values of ECHLOR/PCB were not significantly different among Arctic stocks.
In males, EHCH levels in St. Lawrence animals were not significantly higher than in mean
levels in Cumberland Sound, Beaufort Sea and West Hudson Bay animals. Highest levels of
a-HCH were found in Cumberland Sound males. Females from Cumberland Sound had sig-
nificantly higher levels of a-HCH than St. Lawrence and Jones Sound animals. There were
no significant differences among stocks of females for HCH.
a-HCH accounted for a much higher proportion of EHCH in Arctic stocks (63% in males;
70% in females) than in St. Lawrence animals (27% in males; 45% in females). There were
no significant differences in levels in mean a-HCH or 1,HCH concentrations between males

175
TABLE 3. Mean concentrations of total PCB congeners (IPCB) and homologs (ig/g wet weight) in blubber of beluga from Canadian waters.'

Stock 'PCB C13 C14 C15 C16 C19 Cl9 CI9 Arochlor
1254:1260
E. Hudson M 2.77 ±0.51 0.03 ±0.01 0.51 ±0.09 0.76 ± 0.15 0.92 ± 0.19 0.44 ± 0.10 0.10 ± 0.02 0.008 ± 0.003 3.45 ±0.83
Bay 1984/85 2.10 - 3.70 0.01 -0.05 0.40 - 0.68 0.57- 1.03 0.65- 1.24 0.31 -0.58 0.06 - 0.12 0.005 - 0.011 2.26 - 4.78

F 1.23 ± 0.84 0.03 ± 0.02 0.23 ± 0.19 0.33 ± 0.25 0.39 ± 0.27 0.19 ± 0.11 0.05 ± 0.02 0.007 ± 0.004 1.38 ± 1.00
0.44 - 2.43 0.02 - 0.07 0.06 - 0.52 0.10 - 0.70 0.13 - 0.82 0.08 - 0.39 0.03 - 0.08 0.003 - 0.014 0.45 - 3.04

E. Hudson M _b 0.98 ± 0.23 7.99 ± 1.99

Bay 1987 0.76 - 1.42 5.48- 11.1

0.38 ± 0.38 2.53 ± 2.95

0.08 - 1.09 <0.01 -7.58

W. Hudson M 3.12 ± 0.34 0.02 ±0.01 0.65 ±0.14 0.95 ± 0.13 0.89 ± 0.04 0.49 ± 0.02 0.11 ± 0.01 0.016 ± 0.005 3.97 ± 0.29
Bay 2.75 - 3.47 0.01 -0.02 0.53 - 0.81 0.76- 1.07 0.86 - 0.95 0.46 - 0.51 0.10 - 0.13 0.010 - 0.022 3.77 -4.21

F 0.96± 1.00 0.01 ±0.01 0.18 ± 0.21 0.32 ± 0.39 0.25 ± 0.26 0.15 ± 0.13 0.04 ± 0.02 0.007 ± 0.004 1.12± 1.19
0.31 -2.44 <0.01 -0.01 0.07 - 0.50 0.09 - 0.91 0.08 - 0.64 0.04 - 0.32 0.02 - 0.06 0.004 - 0.012 0.34 - 2.88

Retrospec- F 0.49 0.04 0.10 0.13 0.13 0.08 0.02 0.001 0.41
tive samples'
M 0.26 0.03 0.08 0.08 0.06 0.02 <0.01 <0.001 0.17

M&F 1.47 0.05 0.34 0.43 0.41 0.19 0.05 0.004 1.34

Cumberland M 4.91 ± 0.25 0.07 ± 0.03 1.38 ± 0.01 1.55 ± 0.08 1.20 ± 0.20 0.61 ± 0.01 0.05 ± 0.02 <0.001 5.26 ± 0.19
Sound 4.74 - 5.09 0.04 - 0.09 1.37- 1.39 1.50- 1.61 1.06- 1.35 0.60 - 0.62 0.04 - 0.06 5.12 - 5.39

F 1.15 ± 0.41 0.07 ± 0.07 0.29 ± 0.09 0.25 ± 0.15 0.26 ± 0.13 0.18 ± 0.05 0.04 ± 0.01 <0.001 0.95 ± 0.41
0.86- 1.44 0.02 - 0.12 0.23 -0.35 0.14 - 0.36 0.16 - 0.35 0.15 - 0.22 0.04 - 0.05 - 0.65 - 1.24

Beaufort M 3.33 ± 0.85 0.06 ± 0.03 0.69 ± 0.21 0.97 ± 0.27 1.05 ± 0.26 0.46 ± 0.11 0.08 ± 0.03 0.006 ± 0.003 3.85 ± 0.94
Sea 2.32 -4.94 0.03 - 0.11 0.46- 1.16 0.69- 1.52 0.69- 1.47 0.31 -0.62 0.05 - 0.14 0.003 - 0.011 2.72 - 5.46

F4 1.23 0.05 0.20 0.32 0.42 0.20 0.04 0.004 1.41

0.83- 1.64 0.05 - 0.05 0.13 - 0.27 0.20 - 0.45 0.30 - 0.53 0.13 - 0.27 0.02 - 0.06 0.002 - 0.006 0.97- 1.85
TABLE 3. (continued)
Stock ',PCB CI, C14 CI, CI6 07 Clg Cl, Arochlor
1254:1260
Jones Sound M 2.53 ± 0.57 0.04 ± 0.01 0.67 ± 0.15 0.79 ± 0.20 0.74 ± 0.17 0.26 ± 0.07 0.03 ± 0.01 0.001 ± 0.001 2.66 ± 0.63
1.88- 3.74 0.03 -0.06 0.54- 1.00 0.60- 1.25 0.50- 1.03 0.17 -0.35 0.02 - 0.06 0.001 - 0.003 1.69 -3.63

F 2.46 1.98 0.04 ± 0.02 0.62 ± 0.52 0.75 ± 0.61 0.73 ± 0.59 0.27 ± 0.21 0.04 ± 0.03 0.001 ±0.001 2.61 ±2.13
0.95 - 6.73 0.02 - 0.09 0.24- 1.74 0.27 - 2.07 0.25 - 2.02 0.09 -0.71 0.01 -0.11 0.001 - 0.003 0.89 - 7.19

St. Lawrence M 75.8 ± 15.3 <0.01 13.9 ± 1.00 8.41 ± 1.45 27.7 ± 6.24 20.4 ± 6.23 4.70 ± 2.40 0.666 ± 0.493 119 ± 25.5
estuary 53.9 - 89.2 12.6- 14.7 6.40 - 9.54 19.0 - 33.7 11.6 - 26.4 1.25 - 6.79 <0.001 - 1.184 83.6- 144

F 37.3 ± 22.0 <0.01 5.84 ± 4.55 5.47 ± 2.55 12.7 ± 8.20 10.3±6.29 2.71 ± 1.59 0.289 ± 0.199 54.2 ± 37.6
14.5 - 68.7 1.91 - 13.1 2.67 - 8.58 4.41 -24.9 3.57- 19.5 1.20 -5.02 <0.001 - 0.514 16.4- 112
Arithmetic mean ± standard deviation and range of concentration.
b Dash indicates not determined.
Consisting of I adult female, 1 newbom male and a sample of beluga oil (Table 1).
No standard deviation, N=2.
TABLE 4. Ratios of organochlorines in beluga tissues.

Ratio
Location/tissue Sex IPCB/DDE DDE/EDDT -y/a-HCH ECHLOR/EPCB mirex/IPCB
Blubber
St. Lawrence M 1.19 ± 0.18 0.66 ± 0.06 0.52 ± 0.15 0.10 ± 0.01 0.012 ± 0.007
F 4.07 ± 2.76 0.56 ± 0.09 0.43 ± 0.17 0.10 ± 0.02 0.027 ± 0.009

Hudson Bay M 2.15 ± 0.42 0.57 ± 0.04 0.29 ± 0.04 0.70 ± 0.11 0.009 ± 0.002
F 2.78 ± 0.56 0.57 ± 0.04 0.18 ± 0.06 0.82 ± 0.19 0.012 ± 0.005

Beaufort Sea M 5.13 ±6.89 0.47 ±0.15 0.34 ± 0.08 0.53 ±0.04 0.011 ± 0.003
F 3.78 0.50 0.30 0.54 0.017

Jones Sound M 2.41 ± 0.62 0.55 ± 0.04 0.32 ± 0.05 0.76 ± 0.22 0.002 ± 0.001
F 1.98 ± 0.47 0.56 ± 0.05 0.30 ± 0.09 0.84 ± 0.28 0.006 ± 0.002

Retrospective F 2.07 0.25 0.15 0.29 <0.01

M&F 0.81 0.48 0.19 0.59 0.0005

Liverb M 2.56 0.72 0.40 . 0.33 0.002

F 6.13 ± 1.46 0.65 ± 0.03 0.20 ± 0.05 0.33 ± 0.05 0.004 ± 0.003

Muscle' M 4.76 ± 2.73 0.63 ± 0.11 <0.06 - -

F 5.87 ± 7.60 0.48 ± 0.11 <0.08 - -
a Beluga oil samples from 1966/67. See Table 1.
b Samples from Cumberland Sound.
' Samples from Eastern Hudson Bay. PCBs are Aroclor 1254 equivalents. Mirex and chlordane-related
compounds were not determined.
and females in Arctic animals (t-test at P <0.05). However, mean levels of EHCH in St.
Lawrence males were significantly higher than in females. Ratios of the y- to a- isomers of
HCH (Table 4) were significantly higher in St. Lawrence animals than in all Arctic stocks in
the case of males and all except Beaufort Sea and Jones Sound animals for females. There
were no significant differences in -y/a-HCH ratios for either sex among Arctic stocks.
There were significant effects of location but not of age or length on ECBz levels for either
sex. Mean y,CBz concentrations in St. Lawrence males were significantly higher (P <0.05)
than means for Jones Sound, E. Hudson Bay and Beaufort Sea stocks but did not differ sig-
nificantly from Cumberland Sound or W. Hudson Bay stocks. St. Lawrence females had sig-
nificantly higher mean ECBz levels than E. Hudson Bay animals.
Mean levels of mirex in St. Lawrence belugas were 50 to 100-fold higher than in Arctic
belugas (Table 2). This was the largest concentration difference between the St. Lawrence
and Arctic stocks for any organochlorine studied. Unlike most organochlorines, but similar
to a-HCH, mirex levels in males and females were not significantly different (t-test at P
<0.05). For males, mirex/EPCB ratios in St. Lawrence animals were not significantly higher
than Hudson Bay or Beaufort sea animals but they were significantly higher (P <0.05) than
ratios for the Jones Sound animals. Mirex/EPCB ratios in the Hudson Bay and Beaufort Sea
stocks were also significantly higher (P <0.05) than the ratio for Jones Sound males.
Dieldrin was present at similar levels to HCB and mirex in both Arctic and St. Lawrence
beluga blubber (Table 2). Mean levels of dieldrin in St. Lawrence males were significantly
higher than mean levels at any Arctic locations; however, levels in Jones Sound females did
not differ significantly from the St. Lawrence animals.

Organochlorine Levels in Other Tissues

Low ng/g (wet wt) levels of PCBs and organochlorine pesticides were found in liver and
muscle samples of Arctic belugas (Table 5). In liver, PCBs and PCCs were present at approxi-
mately equal concentrations in males (120 ng/g wet wt), unlike blubber in which PCCs were
more prominent (Table 2). On a lipid weight basis (liver and muscle averaged 3.4 and 1.3 %
extractable lipid, respectively), levels of organochlorines were comparable or higher than
those found in blubber. ECHLOR/PCB and PCB/DDE ratios (females only) for liver were
significantly different (t-test at P <0.05) from those calculated for blubber from the same
animals (Table 5).

Analysis of Retrospective Samples

Three beluga oil samples collected in 1966/67 contained all of the organochlorines deter-
mined in blubber samples from the mid-1980's but concentrations were generally lower. The
concentration of IDDT in oil from Churchill was higher than the mean concentration in W.
Hudson Bay beluga blubber (Table 2). 4,4'-DDD formed a major proportion of EDDT,
indicating dehydrochlorination of DDT during the preparation of the oil or during storage.
ECHLOR and PCC levels were 3- to 6-fold lower in the Churchill beluga oil sample than
mean levels of these compounds in male belugas, and 2-fold lower than means in females.
The concentration of EPCB in the oil of Baffin Island female belugas was within the range
seen in present day animals from Hudson Bay and Cumberland Sound (Table 3). DDE/
1DDT and ECHLOR/PCB ratios in the adult female sample from 1966/67 were about 2-fold
lower than found in blubber of Arctic animals from the mid-1980's (Table 4).

179
TABLE 5. Organochlorine concentrations (mean, standard deviation and range, ng/g wet weight) in beluga liver and muscle.

Tissue Sex N Lipid (%) HCB cd-ICH 4,4'-DDE /DDT ECHLOR PCC Dieldrin PCB (Aroclor)'
Muscle' M 6 1.3 ± 0.7 3.2± 1.7 1.7 ± 0.6 18.4 ± 10.2 27.6± 12.6 <0.1 67.2±22.3 8.3 ± 2.3
0.8 - 2.7 1.8 - 6.3 1.0 - 2.5 3.8 - 33.9 8.9 - 47.0 36.9 - 97.0 4.7 - 11.2

6 1.2 ± 0.2 1.7 ± 0.8 1.4 ± 0.5 5.2 ± 5.8 9.9 ± 8.9 <0.1 26.2 ± 26.0 3.4 ± 2.9
1.0- 1.3 0.7 - 3.1 0.6 - 2.0 0.7 - 16.2 1.2 - 26.4 <0.1 -68.2 0.7 - 8.6

Liver M 2 3.1 22.0 2.4 41.6 57.5 35.0 120 20.1 124 29.2
2.8 - 3.5 17.1 -27.0 1.5 - 3.2 33.8 - 49.3 48.4 - 66.5 29.1 -40.9 86.0- 153 19.1 -21.1 89.0- 160 21.5 - 36.9

3 3.7 ± 1.3 10.9 ± 5.4 4.2 ± 0.7 9.5 ± 3.6 14.4 ± 4.8 19.2 ± 3.0 60.5 ± 5.4 8.5 ± 4.8 70.3 ± 4.2 17.2 ± 0.9
2.4 - 5.0 7.5 - 17.1 3.6 - 4.9 7.1 - 13.7 11.2- 20.0 17.3 -22.7 55.5 -66 5.5 - 14.1 66.3 -74.7 16.6- 18.3
• PCB as Aroclor 1254/1260 (1:1) equivalents.
▪ Muscle of Eastern Hudson Bay animals.
• Liver from Cumberland Sound animals.
 Correlations with Age and Length

Correlations of major organochlorine levels in beluga blubber with age and length were
performed on 3 groups (the combined East and West Hudson Bay, Jones Sound and St. Law-
rence) (Table 6). Significant negative correlations between EPCB, ECHLOR, HCH,
y DDT and ICBz concentrations and age and length were observed in females from Hudson
Bay (Table 6). For males, only I DDT was significantly, positively correlated with age and
length. No significant correlations were found for the Jones Sound group except a negative
correlation of EHCH with length in males. Although only a limited number of males from
the St. Lawrence estuary were analysed, there were significant (P <0.01) positive correlations
of DDT, EPCB and ECHLOR with age in this group but not in females from this location.
TABLE 6. Correlations' of organochlorine concentrations with age and length in beluga blubber.
Location Sex N Factor EPCB EDDT ECHLOR EHCH ECBz
St. Lawrence M 4 Age 0.983* 0.997* 0.969* 0.399 0.885
estuary length 0.972* 0.998* 0.959* 0.447 0.877

F 5 Age 0.442 0.335 0.301 0.159 0.299

length 0.022 0.198 0.241 0.446 0.256

Hudson Bayb M 12 Age 0.349 0.587* 0.178 -0.020 -0.146

length 0.250 0.612* 0.034 -0.281 -0.151

F 12 Age -0.774** -0.747** -0.860** 0.758** 0.908**

length -0.609* -0.549* -0.725* 0.850** 0.852**

Jones Sound M 8 Age -0.312 -0.424 -0.241 -0.601 -0.511

length -0.046 -0.074 -0.557 -0.773* -0.632

F 7 Age -0.191 -0.148 -0.216 0.587 -0.337

length -0.221 -0.170 -0.085 -0.481 -0.475
a Correlation coefficient significant at P <0.01 (**) or P <0.05 (*).
Combined East and West Hudson Bay data (1984/86 samples).

Discussion

Sources, Pathways, and Transformations of Organochlorine Contaminants

The pattern of PCB congeners and homologs in beluga blubber differed considerably from
a mixture of Aroclor standards (Fig. 3 and 4). Levels and relative amounts of homologs were
similar in Arctic animals with tetra, penta and hexachloro congeners predominating (Ta-
ble 3). In the St. Lawrence animals, hexa and heptachloro congeners predominated (Fig. 4)
and the homolog pattern resembled a 1:1 mixture of Aroclor 1254:1260 more closely than
did the pattern in Arctic animals. The pattern of PCB congeners differed between the Arctic
and St. Lawrence males with PCB-153, -138, -187/182 and -180 (all substituted at the 2,4-
or the 2,4,5- positions on both phenyl rings) accounting for a higher proportion of EPCB in
the St. Lawrence group (Fig. 3). This pattern could be indicative of greater metabolic activity
in the St. Lawrence animals due to mixed function oxidase enzyme induction by co-planar
PCBs such that only the most recalcitrant structures remain. Tanabe et al. (1988) have
suggested that small cetaceans lack the capacity to metabolize PCBs with adjacent non-
chlorinated meta- and para-carbons (i.e. positions 3,4- or 4,5- on the biphenyl ring). The
pattern of PCB congeners seen in both Arctic and St. Lawrence belugas is very similar to that
reported for white-beaked dolphins (Lagenorhynchus albirostris) and pilot whales (Globi-

181
 East Hudson Bay
30 _

20 -

10 -

0 I
Percen t of Tota lChlordan e

Males
50-
40:
St. Lawrence estuary
• Females

30 :
20:
10:
0 Irn_t71_ • pa. el
pa. De

30-i Tech. Chlordane

20 -

10 -

0
c, de' de c9,
1111.
o 7,1,01/4"
Chlordane Compcsnent

Fia. 4. Comparison of the proportion of PCB homologs in an Aroclor 1242:1254:1260(1:1:1) standard

and in beluga blubber from Hudson Bay and the St. Lawrence estuary.

cephala malaena) from Newfoundland waters (Muir et al. 1988a). Geraci (1989) reported a
similar predominance of PCB-153 and -138 in blubber of stranded bottlenose dolphins
(Tursiops truncatus) from the U.S. central and south Atlantic coast. Tanabe et al. (1984,1988)
and Duinker and Hillebrand (1983) have reported similar PCB congeneric patterns for other
cetaceans and Japanese waters and in the North Sea, respectively.
The differences in PCB congener and homolog profiles between the St. Lawrence and
Arctic stocks may also reflect different sources and pathways of PCB contamination. In the
St. Lawrence River, sediments and biota have been contaminated as a result of direct discharge
of PCBs to the aquatic environment (Couillard 1982). Atmospheric deposition is also an
important pathway for entry of PCBs and other organochlorines into the aquatic environment
in the St. Lawrence drainage basin (Eisenreich et al. 1981). In the Arctic marine environment,
atmospheric transport, ocean currents, and inputs from north-flowing rivers are the only
pathway of introduction of PCBs. There has been some local use of PCBs, for example in
transformers in DEW line stations however, there is no evidence of direct contamination of
polar bears (Norstrom et al. 1988) or ringed seals (Muir et al. 1990b). Tanabe et al. (1988)
noted that tetra and pentachloro-biphenyls formed a greater proportion of total PCB in Dall's
porpoise (Phocoenoides dalli) from the Bering Sea than in dolphins from temperate and
tropical waters. The similar pattern of PCB homologs in Arctic beluga blubber may reflect
greater transport and deposition of the volatile tetra- and penta- congeners to the Arctic
marine environment. A diffuse atmospheric source is also suggested by the similarity in
EPCB in all beluga stocks. A latitudinal gradient in PCB homologs might be expected, for
example, between the Hudson Bay and Jones Sound animals which are separated by 15 to 20
degrees of latitude. Norstrom et al. (1988) found higher levels of PCBs in polar bears from

182
 40
East Hudson Bay
30

20

10 Ei Males

0 • Females
Pe rcen t ofTota l PC B

40
St. Lawrence estuary
30

20

10
0

40 H
Aroclor
1242:1254:1260
30 ■■■1

(1:1:1)
20 H
10 H
0
)<'\ <?). <0- 'ç•
e o

PCB Homolog
FIG. 5. Chlordane-related compounds as a percent of ECHLOR in technical chlordane and and in beluga
blubber from Hudson Bay and the St. Lawrence estuary.

West Hudson Bay than in those from the Lancaster Sound area. The lack of a gradient may
be due to the different age structure of the Jones Sound animais that were analysed, and to
the fact that the Hudson Bay animais are not residents of the Bay for much of the year but are
believed to inhabit Hudson Strait during the winter months (Smith and Hammill 1986).
The source of PCCs for both the St. Lawrence and Hudson Bay watersheds may be
atmospheric transport and deposition from the southern U.S. where PCCs were used
extensively until 1984 (Rapaport and Eisenreich 1988). PCC was an important organochlo-
rine contaminant in sea water sampled during 1986/87 at the ice island located west of Axel
Heiberg Island (Bidleman et al. 1989) and is the major organochlorine in marine fish from
Canadian Arctic waters (Muir et al. 1987). Little toxaphene was apparently used in the Great
Lakes region although PCCs are prominent contaminants in fish in these waters (Schmitt et
al. 1983). The higher levels in St. Lawrence belugas may therefore be due to the proximity
of the St. Lawrence drainage area to areas of major PCC use, in contrast to the more remote
Arctic locations. By the same reasoning, higher levels of PCCs would be expected in Hudson

183
Bay animals inhabiting waters draining northern Ontario, but this was not observed. Similar
PCC levels were observed in all Arctic stocks. Generally higher levels of PCCs were obser-
ved in Arctic char (Salvelinus alpinus) from the eastern Arctic which was consistent with the
northeastern flow of air masses from mid-latitudes in North America to the Arctic (Muir et
al. 1990c).
The pattern of PCCs in marine mammals shows extensive transformation in comparison
with the standard, making quantitation difficult (Jansson et al. 1979; Muir et al. 1988a). This
pattern may also reflect selection of heavier, less water soluble PCCs during atmospheric
transport and deposition in sea water and during accumulation by biota (Bidleman et al.
1989). PCCs were a major contaminant in dolphins and pilot whales from Newfoundland
waters (Muir et al. 1988a) and in narwhals (Monodon monoceros) (Muir et al. 1990a).
Andersson et al. (1988) found PCCs were prominent organochlorine contaminants in salmon
from the Northwest Atlantic and Greenland, and in harp seals (Pagophilus groenlandicus)
from the Gulf of St. Lawrence. The PCC/PCB ratio of 0.42 in harp seals (Andersson et al.
1988) was higher to that in St. Lawrence beluga blubber in the present study (0.19) but was
lower than ratios in Arctic belugas (mean of 1.5).
The extremely high levels of 1, DDT in St. Lawrence belugas as well as in other cetaceans
from the Canadian east coast are indicative of past heavy use of the pesticide for agricultural
and forestry uses in Eastern Canada (Gaskin et al. 1983). Deposition of DDT and related
compounds to peat bogs in northern and central Ontario and Quebec reached a maximum in
the mid-1960's coinciding with the peak production of DDT in North America (Rapaport and
Eisenreich 1988). DDE/EDDT ratios approaching 0.6, especially in St. Lawrence animals
(Table 4), suggest "old" rather than recent DDT inputs (Aguilar 1987; Martineau et al. 1987).
DDE/EDDT ratios in St. Lawrence males (but not females) were significantly higher than
ratios in Arctic animals, which suggests that DDT inputs into the Arctic may be more recent,
although the older age of the St. Lawrence animals could also be a factor. E DDT levels in
St. Lawrence belugas were similar to those in harbour porpoises (Phocoena phocoena) from
the Bay of Fundy (Gaskin 1982) and about 2-fold higher than found in white-beaked dolphins
from the Gulf of St. Lawrence (Muir et al. 1988a). Male bottlenose dolphins from the U.S.
east coast had 1.7-fold higher levels of 4,4'-DDE in blubber than St. Lawrence belugas
(Geraci 1989) while dolphins from the California coast, collected in the mid-1970's, had
about 10-fold higher concentrations (O'Shea et al. 1980).
The prevalence of chlordane-related compounds in Arctic biota is apparent from the higher
ECHLOR/PCB ratios in Arctic belugas compared with the St. Lawrence stocks. In compa-
rison with other marine mammals in the Canadian Arctic (Muir et al. 1990a) but about 4-fold
higher than in ringed seals (Muir et al. 1990c). The St. Lawrence belugas had lower levels
of chlordane-related compounds than other east coast cetaceans. White-beaked dolphins had
about 1.5- to 2-fold higher levels of ECHLOR than did St. Lawrence belugas (Muir et al.
1988a) while bottlenose dolphins from the U.S. east coast had about 5-fold higher levels of
t-nonachlor (Geraci 1989). Kawano et al. (1988) found total chlordane concentrations
(2.8 lig/g lipid wt) in Dall's porpoise from the northwest Pacific and Bering Sea that were
similar to the levels in St. Lawrence belugas.
Like PCCs and DDT, chlordane was used extensively in agriculture until the mid-1970's,
particularly in the St. Lawrence river drainage area for insect control in corn production (Von
Rumker et al. 1974). Higher levels of ECHLOR in St. Lawrence animals may reflect this, as
well as possible direct discharge and greater atmospheric deposition compared to Hudson
Bay or the Beaufort Sea. As for PCCs, however, there was no trend of declining ECHLOR
residues with increasing north latitude. There is evidence for atmospheric transport and
deposition of chlordane-related compounds to the Canadian Arctic (Gregor 1990; Hoff and
Chan 1986) although it is not as prominent in air or seawater as HCH isomers or PCCs
(Bidleman et al. 1989).

184
 The lack of differences in a-HCH between St. Lawrence and Arctic belugas is consistent
with the ubiquity of this organochlorine. a-HCH was the major form of HCH detected in air
and sea water at the ice island (Hargrave et al. 1988). Tanabe et al. (1982) showed that EHCH
levels over the Pacific Ocean peaked between 30°N and 70°N and concluded that Asia and
Southeast Asia were the main sources of HCH isomers. The significantly higher y-/a-HCH
ratios in St. Lawrence belugas may reflect the greater input to the St. Lawrence river of the
y-HCH isomer (lindane) which is still registered for use as an insecticide in Canada and the
United States. Transformation of -y-HCH to the more stable a-isomer in the atmosphere
(Oehme and Mano 1984) may also contribute to the lower -y-/a-HCH ratios in the Arctic
stocks.
EHCH levels in St. Lawrence belugas were about 2-fold lower than in white-beaked
dolphins from Newfoundland waters (Muir et al. 1988a). Mean levels of EHCH in blubber
of Arctic ringed seals from Northern Baffin Is. and Lancaster Sound were similar to levels
in belugas from the Jones Sound and Cumberland Sound stocks (Muir et al. 1988b). Tanabe
et al. (1984) and Kawano et al. (1988) have reported much higher proportions of B-HCH in
1,HCH in small cetaceans from the Northwestern Pacific than observed in belugas in the
present study. -y/a-HCH ratios in the N.W. Pacific cetaceans were 0.25 — 0.40, which is a
similar range to the ratio in Arctic belugas but lower than in St. Lawrence animals. Greater
use of technical BHC, which contains a-, B- and y-HCH, in countries bordering the western
Pacific may explain the higher proportion of the B-HCH in cetaceans in that region.
Mirex residues in St. Lawrence beluga blubber clearly distinguish these animals from
Arctic belugas and from other Canadian east coast cetaceans. Sources and levels of mirex in
St. Lawrence belugas have been discussed by Béland and Martineau (1988). These authors
noted that eels migrating down the St. Lawrence River from Lake Ontario, where relative-
ly high levels of mirex are found in sediments and biota, are the most likely source for belugas.
Ten- to 100-fold lower levels of mirex in cetaceans not feeding in the St. Lawrence estuary
indicate that atmospheric transport and deposition of mirex is a relatively unimportant
pathway for the St. Lawrence animals. Although mirex levels were low in Arctic belugas, a
distinct gradient was apparent with higher levels in animals feeding in waters influenced by
large inputs of freshwater from large rivers (Mackenzie estuary and Western Hudson Bay)
than in animals in the High Arctic and Cumberland Sound (Table 2). Mirex was also found
in the blubber of white-beaked dolphins from Canadian east coast waters at levels about 10-
fold lower than in St. Lawrence animals (Muir et al. 1988a).
HCB and ECBz levels in St. Lawrence belugas were very similar to those reported in white-
beaked dolphins (Muir et al. 1988a) but about 5- fold higher than reported in blubber of
harbour porpoises from the Bay of Fundy. Dieldrin levels in St. Lawrence belugas were
similar to those reported in white-beaked dolphins (Muir et al. 1988a) and in pilot whales and
white-sided dolphins (Lagenorhynchus acutus) from the U.S. east coast in the early 1970's
(Taruski et al. 1975) but were about 2-fold lower than those found in harbour porpoises in the
Bay of Fundy in the late 1970's (Gaskin 1982).

Age and Sex Differences in Organochlorine Levels

A higher level of PCBs and organochlorine pesticides in male than in female pinnipeds and
cetaceans is widely observed and is attributed to loss of pollutants via lactation (Aguilar
1987). Tanabe et al. (1986,1987) found that PCB and DDE residues in female short-finned
pilot whale (Globicephala macrorhynchus) and in Antarctic minke whales (Balaenoptera
acutorostrata) were positively correlated with age of immature females and negatively cor-
related with age of reproductively active females. A similar decline with increasing age was
observed for the Hudson Bay female belugas. The average age of the Hudson Bay belugas
was 10 ± 4 yr (Table 1). High Arctic females, which averaged 4.6 yr, showed no significant
correlations between residues and age and had levels similar to those in males, indicating that

185
they were too immature to have lactated. By contrast, Martineau et al. (1987) found signifi-
cant positive correlations between both male and female St. Lawrence belugas and PCB and
y DDT residues in blubber, and a similar positive correlation (not statistically significant)
was seen with the five females from the St. Lawrence that were analysed in the present work.
Such a positive trend is unusual in female cetaceans and could be caused by several factors.
It may be a reflection of the greater contamination of their diet such that successive lactations
do not eliminate a quantity equal to what they are accumulating, whereas in the Arctic
females, lactation appears to eliminate greater amounts of PCB and 1,DDT than is contribu-
ted via the diet. The positive correlation with age in the females in the St. Lawrence could also
be due to a lower frequency of lactation due to a lower reproductive rate or to the older age
of the samples which are collected exclusively from dead stranded animals. Tanabe et al.
(1986) found evidence for increasing levels in older females which they attributed to reduced
numbers of parturitions in this group. A similar relationship, whether it might be age-related
or due to reproductive effects of the contaminants, may explain the positive correlation of
PCB and E DDT levels in female belugas observed by Martineau et al. (1987).
The 4-fold and 2-fold higher levels of E DDT and EPCB, respectively, found in male than
in female St. Lawrence animals in the present study, and by Martineau et al. (1987), are
similar to the differences found between the sexes in Hudson Bay and Cumberland Sound
animals, and suggest that loss of the most lipophilic organochlorines via lactation is similar
in these populations and that, in this respect, the St. Lawrence animals are "normal."
Differences in organochlorines between sexes could also be due to diet. However, if this the
case, greater differences would be expected for mirex and a-HCH levels. Most of the
chlorinated aliphatic compounds (mirex, HCH, cis-chlordane) showed few differences bet-
ween sexes and, unlike the chlorinated aromatics, do not appear to be preferentially excreted
via lactation.

Temporal Trends in Organochlorines

Analysis of beluga oil from 1966/67 indicated that chlordane and PCCs were more
prevalent in the 1980's than during the 1960's (Table 2). Beluga oil from the W. Hudson Bay
stock had higher levels of E DDT and lower levels of PCCs and chlordane than found in the
same stock 20 yr later. Because the sex of the animals from which the oil was prepared is
unknown, no precise estimate of the increase can be given. These results are consistent with
trends between 1969 and 1984 in organochlorines in fat from Hudson Bay and Baffin Bay
polar bears (Norstrom et al. 1988), in which ECHLOR and EPCB levels increased 4.3-fold
and 1.9-fold, respectively, and E DDT declined slightly. Lower concentrations, especially of
chlorinated aliphatic compounds, could also be due to degradation during manufacture and
storage of the oil, as was indicated by relatively high 4,4-DDD levels in the sample from
Churchill.
The findings in beluga blubber and polar bear fat are not consistent with trends observed
in ringed seals in the Canadian Arctic, especially in the case of PCBs. Addison et al. (1986)
reported declines of about 60% for PCBs and 40 — 50% for 4,4'-DDE between 1972 and 1981
in blubber of ringed seals from Holman Island. EPCB concentrations in Beaufort Sea beluga
blubber appear to be higher than observed by Addison and Brodie (1973), who did not detect
PCBs in samples collected in 1972, whereas 1 DDT concentrations in adult animals (sex was
not specified) in the earlier study were similar to levels that we found in males. However, the
absence of detectable PCB in the 1972 samples may reflect lower reliability and reduced
sensitivity of PCB analysis using "packed" column gas chromatography, a less elaborate
separation technique widely used at that time. Two harbour porpoises collected in 1972 from
West Greenland had similar PCB levels (6.7 1.tg/g) to the Cumberland Sound belugas
analysed in this study (Clausen et al. 1974). Mean PCB concentrations for East Hudson Bay
animals sampled in 1984/85 and those from 1987 (Table 2 and 3) showed no significant

186
differences (t-test at P <0.05). E DDT levels in 1987 samples were about 2-fold higher than
in the earlier samples for both males and females. The greater age of the 1987 group (Table
1) may explain these differences, especially in the case of males where positive correlations
of E DDT with age were observed.
In conclusion, there are few major differences in organochlorine concentrations among
Arctic beluga stocks. This lack of geographical variation is consistent with a diffuse atmo-
spheric source of the contaminants and with wide geographic range of some stocks such as
the Hudson Bay animals. The St. Lawrence belugas, in contrast to the Arctic animals, have
much higher levels of EPCBs, E DDT, mirex, and, to a lesser extent PCCs, dieldrin and
chlordane-related compounds. The PCB and E DDT levels found in the limited number of St.
Lawrence animals analysed in this survey were similar to those reported in a larger survey
by Martineau et al. (1987). The high levels of PCBs and other organochlorines have been
suggested to be the prime cause of the low reproductive rate of the St. Lawrence population
(Béland et al. 1988; Sergeant and Hoek 1988) although this has been questioned by Addison
(1989). The St. Lawrence population has been reported to have a reproductive rate which is
less than half that in Arctic stocks (Sergeant and Hoek 1988). While no cause and effect
relationship can be unequivocally established, the comparison with Arctic belugas supports
the hypotheses put forward by those authors. The St. Lawrence animals are clearly exposed
to much higher levels of organochlorines and other pollutants, such as lead (Wagemann et
al. 1990) and polyaromatic hydrocarbons (Martineau et al. 1988), than Arctic animals. Con-
tinued study of the St. Lawrence animals is needed because toxicological effects of organo-
chlorines are clearly more likely to be expressed in this stock before becoming apparent in
other cetacean populations in Canadian waters.

Acknowledgments

We wish to thank the following individuals for help in sample collection, subsampling and
aging of animals: R. Walker and P. Weaver, Fisheries and Oceans, Winnipeg; W. Doidge,
Fisheries and Oceans, Ste-Anne-de-Bellevue, R. Plante (Filmar Inc., Bic, P.Q.) and the
Hunters and Trappers Associations of Pangnirtung, Eskimo Point and Grise Fiord. We also
thank R.J. Norstrom and M. Simon (Canadian Wildlife Service, National Wildlife Research
Centre, Hull, P.Q.) for GC-MS confirmation of PCCs and other organochlorines. Collection
and analysis of St. Lawrence specimens was supported through a grant from the Wildlife
Toxicology Fund.

References

ADDISON, R.F. 1989. Organochlorines and ma- group residues in arctic ringed seals (Phoca
rine mammal reproduction. Can. J. Fish. Aquat. hispida) between 1972 and 1981. Environ.
Sci. 46: 360-368. Sci. Technol. 20: 253-256.
ADDISON, R. F., AND P. F. BRODIE. 1973. Occur- AGUILAR, A. 1987. Using organochlorine pollu-
rence of DDT residues in beluga whales (Del- tants to discriminate marine mammal popula-
phinapterus leucas) from the Mackenzie Del- tions: A review and critique of the methods.
ta, N. W. T. J. Fish. Res. Board Can. 30: Mar. Mammal. Sci. 3: 242-262,
1733-1736. ANDERSSON, O., C-E. LINDER, M. OLSSON, L.
ADDISON, R.F., M.E. ZINCK, AND T.G. SMITH. REUTERGARDH, U-B. UVEMO, AND U. WIDE-
1986. PCBs have declined more than DDT- QUIST. 1988. Spatial differences and temporal

187
 trends of organochlorine compounds in biota search. Ontario Veterinary College, Universi-
from the northwestern hemisphere. Arch. ty of Guelph, Ont. 63 p.
Environ. Contain. Toxicol. 17: 755-765. GREGOR, D.J. 1990. Deposition and accumula-
BALLSCHMITER, K., AND M. ZELL. 1980. Baseline tion of selected agricultural pesticides in Ca-
studies of the global pollution I. Occurrence of nadian arctic snow, p. 373-386. /n D. Kurtz
organohalogens in pristine European and An- [ed.] Long range transport of pesticides Lewis
tarctic aquatic environments. Int. J. Environ. Publ, Chelsea, MI.
Anal. Chem. 8: 15-35. HARGRAVE, B.T., W.P. VASS, P.E. ERICKSON,
BÉLAND, P., AND D. MARTINEAU. 1988. The belu- AND B. R. FOWLER. 1988. Atmospheric trans-
ga whale (Delphinapterus leucas) as integra- port of organochlorines to the Arctic Ocean.
tor of the St. Lawrence basin contamination Tellus 40B: 480-493.
history. In Proceedings of the Intern. Confe- HOFF, R. M., AND K.-W. CHAN. 1986. Atmosphe-
rence on Bio-indicators: Exposure and Ef- ric concentrations of chlordane at Mould Bay,
fects. Oak Ridge National Laboratory, Knox- N.W.T. Canada. Chemosphere 15: 449-452.
ville, TN. March 20-23,1988. JANSSON, B., R. VAZ, G. BLOMKVIST, S. JENSEN,
BÉLAND, P., A. VEZINA AND D. MARTINEAU. AND M. OLSSON. 1979. Chlorinated terpenes
1988. Potential for growth of the St. Lawrence and chlordane components found in fish, guille-
(Québec, Canada) beluga whale (Delphinap- mot and seal from Swedish waters. Chemos-
(erns leucas) population based on modelling. phere 9: 181-190.
J. Cons. Int. Explor. Mer. 45: 22-32. KAWANO, M., T. INouE, T. WA DA, H. HIDAK A,
BIDLEMAN, T.F., G.W. PATTON, M.D. WALLA, AND R. TATSUKAWA. 1988. Bioconcentration
B. T. HARGRAVE, W. P. VASS, P. ERICKSON, B. and residue patterns of chlordane compounds
FOWLER, V. SCOTT, AND D. J. GREGOR. 1989. in marine animals: Invertebrates, fish, mam-
Toxaphene and other organochlorines in arctic mals and seabirds. Environ. Sci. Technol. 22:
ocean fauna: Evidence for atmospheric delive- 792-797.
ry. Arctic 42: 307-313. MARTINEAU, D., P. BÉLAND, C. DESJARDINS, AND
CLAUSEN, J., L. BRAESTRUP, AND O. BERG. 1974. A. LAGACÉ. 1987. Levels of organochlorine
The content of polychlorinated hydrocarbons chemicals in tissues of beluga whales
in arctic mammals. Bull. Environ. Contam. (Delphinapterusleucas) from the St. Lawrence
Toxicol. 12: 529-534. estuary, Québec, Canada. Arch. Environ.
COUILLARD, D. 1982. Evaluation des teneurs en Contain. Toxicol. 16: 137-147.
composés organochlorés dans le fleuve, l'es- MARTINEAU, D., A. LAGACÉ, P. BÉLAND, R.
tuaire et le Golfe Saint-Laurent, Canada. Envi- HIGGINS, D. ARMSTRONG, AND L. R. SHUGART.
ron. Poll. B3: 239-270. 1988. Pathology of stranded beluga whales
DUINKER, J. C., AND M. T. J. HILLEBRAND. 1983. (Delphinapterus !my's) from the St. Lawrence
Composition of PCB mixtures in biotic and estuary, Québec, Canada. J. Comp. Pathol. 98:
abiotic marine compartments (Dutch Wadden 287-311.
Sea). Bull. Environ. Contain. Toxicol. 31: MIYAZAKI, T., T. YAM AGISHI, AND M. MATSUMO-
25-32. TO. 1985. Isolation and structure elucidation
EISENREICH, S. J., B. B. LOONEY, AND J. D. THORN- of some components in technical grade chlor-
TON. 1981. Airborne organic contamination of dane. Arch. Environ. Contain. Toxicol. 14:
the Great Lakes ecosystem. Environ. Sci. Tech- 475-483.
nol. 15: 30-38. MUIR, D. C. G., R. WAGEMANN, W. L. LOCKHART,
GASKIN, D.E. 1982. The ecology of whales and N. P. GRIFT, B. N. BILLECK, AND D. A. MET-
dolphins. Heinemann, London. Chapter 10. . NER. 1987. Heavy metal and organic contami-
GASK IN, D.E., R. FRANK, AND M. HOLDRINET. nants in arctic marine fishes. Environmental
1983. Polychlorinated biphenyls in Harbor Studies No. 42, Indian and Northern Affairs
porpoises (Phocoena phocoena (L.)) from the Canada, Ottawa. 64 p.
Bay of Fundy, Canada, and adjacent waters, MUIR, D. C. G., R. WAGEMANN, N. P. GRIFT, R. J.
with some information on chlordane and hexa- NORSTROM, M. SIMON, AND J. LIEN. 1988a.
chlorobenzene levels. Arch. Environ. Contam. Organochlorine chemical and heavy metal
Toxicol. 12:211-219. contaminants in white-beaked dolphins
GERACI, J. R. 1989. Clinical investigation of the (Lagenorhynchus albirostris) and pilot whales
1987-88 mass mortality of bottlenose dol- (Globicephala melaena) from the coast of
phins along the U.S. central and south Atlantic Newfoundland, Canada. Arch. Environ. Toxi-
coasts. Final report to National Marine Fishe- col. Chem. 17: 613-629.
ries Service, U.S. Navy, Office of Naval Re-

188
Muu, D. C. G., R.J. NORSTROM, AND M. SIMON. lar weight chlorinated hydrocarbons to eastern
1988b. Organochlorine contaminants in arctic North America. Environ. Sci. Technol. 22:
marine food chains: Accumulation of specific 931-941.
polychlorinated biphenyls and chlordane-re- RELINDERS, P. J. H. 1986. Reproductive failure in
lated compounds. Environ. Sci. Technol. 22: common seals feeding on fish from polluted
1071-1079. coastal waters. Nature 324: 456-457.
MUIR, D.C. G., C.A. FORD, N. P. GRIFT, AND SAS Institute 1987. SAS User's Guide: Statistics
R. E. A. STEWART. 1990a. Organochlorine con- 1987 Ed., SAS Institute, Box 8000, Cary, NC.
taminants in narwhal (Monodon monoceros) SAFE, S. 1984. Polychlorinated biphenyls (PCBs)
from the Canadian Arctic, Environ. Pollut. (In and brominated biphenyls (PBBs): biochemis-
press). try, toxicology, and mechanisms of action.
MUIR, D. C. G., C. A. FORD, R. J. NORSTROM, AND CRC Crit. Rev. Toxicol. 13: 319-393.
M. SIMON. 1990b. Geographical variation in SCFIMITT, C.J., M.A. RIBICK, J.L. LUDKE, AND
organochlorine contaminants in selected Ca- T. W. MAY. 1983. National Pesticide Monito-
nadian arctic marine biota. In Proceedings of ring Program: Organochlorine Residues in
the 7th International Conference of Comité Freshwater Fish, 1976-79. U.S. Fish and Wil-
Arctique International on Global Significance dlife Service, Resource Publ. 152, Washing-
of the Transport and Accumulation of Poly- ton, DC.
chlorinated Hydrocarbons in the Arctic. (In SERGEANT, D.E. 1980. Levels of mercury and or-
press). ganochlorine residues in tissues of sea mam-
Mu I it, D. C. G., N. P. GRIFT, C. A. FORD, A. REI- mals from the St. Lawrence estuary. Int. Counc.
GER, M. HENDZEL, AND W. L. LOCKHART. Explot. Sea, Mar. Environ. Quai. Commit.,
1990e. Evidence for long range transport of C.M. 1980/E:55,10 p.
toxaphene to remote arctic and subarctic wa- SERGEANT, D. E., AND W. HOEK. 1988. An update
ters from monitoring of fish tissues, p. 329-346. of the status of white whales Dephinapterus
In D. Kurtz led.] Long range transport of pes- leucas in the St. Lawrence estuary, Canada.
ticides and other toxics. Lewis Publ, Chelsea, Biol. Conserv. 45: 287-302.
MI. STEWART, R. E. A. 1986. Species account: Beluga
MUSIAL, C. J., AND J. F. UTHE. 1983. Widespread (Delphinapterus leucas). Internal report Dep.
occurrence of the pesticide toxaphene in Ca- of Fisheries and Oceans, Winnipeg. 48 p.
nadian east coast marine fish. Intern. J. Envi- SMITH, T. G., AND M.O. HAMMILL. 1986. Popula-
ron. Anal. Chem. 14: 117-126. tion estimates of white whale, Dephinapterus
NORSTROM, R. J., M. SIMON, D. C. G. MUIR, AND leucas, in James Bay, Eastern Hudson Bay,
R.E. SCHWEINSBURG. 1988. Organochlorine and Ungava Bay. Can. J. Fish. Aquat. Sci. 43:
contaminants in arctic marine food chains: 1982-1987.
Identification, geographical distribution and TANABE, S., R. TATSUKAWA, M. KAWA NO, AND
temporal trends in polar bears. Environ. Sci. H. HIKAKA. 1982. Global distribution and at-
Technol. 22: 1063-1071, mospheric transport of chlorinated hydrocar-
NORSTROM, R. J., AND H. T. WON. 1985. Long bons: HCH (BHC) isomers and DDT com-
term preservation of egg and tissue homogen- pounds in the western Pacific, eastern Indian
ates for determination of organochlorine com- and Antarctic Oceans. J. Oceanograph. Soc.
pounds: freezing versus freeze-drying. J. As- Japan 38: 137-148.
soc. Offic. Anal. Chem. 68: 130-135. TANABE, S., H. TANAKA, AND R. TATSUKAWA.
OEHME, M., AND S. MANO. 1984. The long-range 1984. Polychlorobiphenyls, E DDT, and hexa-
transport of organic pollutants to the arctic. chlorocyclohexane isomers in the western north
Fres. Z. Anal. Chem. 319: 141-146. Pacific ecosystem. Arch. Environ. Contam.
O'SHEA, T. J., R. L. BROWNELL JR., D. R. CLARK Toxicol. 13: 731-738.
JR., W. A. WALKER, M.L. GAY, AND T. G. TANABE, S., S. MIURA AND R. TATSUKAWA. 1986.
LAMONT. 1980. Organochlorine pollutants in Variations of organochlorine residues with
small cetaceans from the Pacific and South age and sex in antarctic minke whales. Mem.
Atlantic Oceans, November 1968-June 1976. Nat. Inst. Polar Res., Spec. Issue, 41: 174-181.
Pestic. Monit. J. 14: 35-46. TANABE, S., B.G. LOGANATIIAN, A.N. SUBRA-
PEAKALL, D. B. 1975. PCB's and their environ- MANIAN, AND R. TATSUKAWA. 1987. Organo-
mental effects. CRC Crit. Rev. Environ. Con- chlorine residues in short-finned pilot whales.
trol 5: 465-508. Possible use as tracers of biological parame-
RAPAPORT, R. A., AND S. J. EISENREICH. 1988. ters. Mar. Pollut. Bull. 18: 561-563.
Historical atmospheric inputs of high molecu-

189
TANABE, S., S. WATANABE, H. KA N, AND R. TAT WAGEMANN, R., R.E. A. STEWART, P. BÉLAND,
SUKAWA, 1988. Capacity and mode of PCB AND C. DESJARDINS. 1990. Heavy metals and
metabolism in small cetaceans. Mar. Mam- selenium in tissues of beluga whales, Delphi-
mal. Sci. 4: 103-124. napterus leueas, from the Canadian Arctic and
TARUSKI, A.G., C.E. OLNEY, AND H. E. WINN. the St. Lawrence Estuary, p. 191-206. In T. G.
1975. Chlorinated hydrocarbons in cetaceans. Smith, D. J. St. Aubin, and J. R. Geraci [ed.]
J. Fish. Res. Board Can. 32: 2205-2209. Advances in research on the beluga whale,
VON RUMKER, R.E., E. W. LAWLESS, AND A.F. Delphinapterus leucas. Can. Bull. Fish. Aquat.
MEINERS. 1974. Production, distribution, use Sci. 224.
and environmental impact potential of selec- WALKER, C. H. 1983. Pesticides and birds — me-
ted pesticides. U.S. Environ. Protection Agen- chanisms of selective toxicity. Agric. Ecosys.
cy, EPA/540/1-74-001. Environ. 9: 211-226.

190
 Heavy Metals and Selenium in Tissues of
Beluga Whales, Delphinapterus leucas,
from the Canadian Arctic and the St. Lawrence Estuary

R. Wagemann and R. E. A. Stewart

Department of Fisheries and Oceans, Central and Arctic Region,
Freshwater Institute, 501 University Crescent,
Winnipeg, Manitoba, Canada R3T 2N6

P. Béland
Institut National d' Ecotoxicologie du Saint-Laurent,
310 Ave. des Ursulines, Rimouski, Quebec, Canada G5L 3AI

and C. Desjardins
Department of Fisheries and Oceans,
789 Boul. Rolland-Therrien, Longueuil, Quebec, Canada J4H 4A6

WAGEMANN, R., R. E. A. STEWART, P. BÉLAND AND C. DESJARDINS. 1990. Heavy

metals and selenium in tissues of belugas whales, Delphinapterus leucas,
from the Canadian Arctic and the St. Lawrence Estuary, p. 191-206. In T.
G. Smith, D. J. St. Aubin, and J. R. Geraci [ed.] Advances in research on
the beluga whale, Delphinapterus leucas. Can. Bull. Fish. Aquat. Sci. 224.

Samples of liver, kidney and muscle tissue from beluga (Delphinapterus leucas)
from five locations across the Canadian Arctic and the St. Lawrence Estuary were
analyzed for copper, zinc, cadmium, mercury, lead and selenium. The St. La-
wrence Estuary beluga had significantly higher levels of lead, mercury, and
selenium and significantly less cadmium than most Arctic whales. Mercury was
positively associated with selenium and age in all animals, and cadmium in kidney
was positively associated with selenium and age. Copper concentrations declined
with age. Whales from some Arctic sites appeared to have higher levels of
cadmium than from other Arctic sites.

Des échantillonages de foie, rein et masse musculaire de marsouins blancs

(Delphinapterus leucas) provenant de cinq endroits sur l'ensemble de l'Arctique
Canadien et de l'estuaire du St-Laurent furent analysés pour le cuivre, le zinc, le
cadmium, le mercure, le plomb et le sélénium. Chez le marsouin blanc de l'estuaire
du St-Laurent les niveaux de plomb, de mercure et de sélénium étaient d'une
manière significative plus élevés et les niveaux de cadmium d'une manière
significative moins élevés que dans la majorité des baleines de l'Arctique. Le
mercure était positivement associé avec le sélénium et l'âge chez tous les animaux,
et le cadmium dans le rein avec le sélénium et l'âge. Les concentrations de cuivre
déclinèrent avec l'âge. Les baleines provenant de certains sites arctiques sem-
blaient avoir des niveaux plus élevés de cadmium que ceux provenant d'autres sites
arctiques.

191
 Introduction

Marine mammals have been exposed to naturally occurring trace metals throughout their
evolutionary history. Some metals are essential for life; others, such as mercury, lead and
cadmium, are toxic to biota even at relatively low levels. They have been present in the
environment from early times, albeit at much lower concentrations (Patterson 1965; Ericson
et al. 1979), but glacial records have shown that global atmospheric fluxes of lead and
mercury have increased as a consequence of human activities (Murozumi et al. 1969; Ng and
Patterson 1981). The historic increase of lead in the environment has also been revealed by
analyses of living and fossil bivalve shells from the eastern Canadian Arctic. Modern shells
had approximately five times as much lead as fossil shells (Bourgoin and Risk 1987).
Worldwide anthropogenic cadmium emissions have also increased nearly seven-fold since
the turn of this century (Boutron and Delmas 1980), and body burdens of cadmium in humans
have increased about five-fold during the same time (Drasch 1983).
Marine mammals near polluted coastal areas have been exposed to various contaminants
as a result of human activities (Braham 1973; Harms et al. 1978; Viale 1978; Duinker et al.
1979; Hyvârinen and Sipilâ 1984). Mercury pollution from the Saguenay River, a tributary
of the St. Lawrence Estuary, has been reported (Cossa and Desjardins 1984; Gobeil et al.
1984), and fish in the international section of the St. Lawrence River had high levels of lead,
mostly alkyl lead, in 1981 (Wong et al. 1988). The pollution of the St. Lawrence Estuary and
the declining beluga (Delphinapterus leucas) stocks there (Béland et al. 1988; Martineau et
al. 1988) remain concerns among scientists (International Forum 1988) and the public
(Boychuck 1988).
Only one beluga from the St. Lawrence was analysed previously for metal contamination
(Sergeant 1980) and its mercury level was not considered high. A review of contaminants in
marine mammals (Wagemann and Muir 1984) revealed little data on Arctic belugas (Bligh
and Armstrong 1971; Beak Consultants 1978; C. T. Hansen, Greenland Environmental
Research Institute, Copenhagen, Denmark, personal communication). High metal burdens
were reported in other Arctic marine mammals, notably mercury and lead in ringed seals
(Phoca hispida) (Smith and Armstrong 1975; Wagemann 1989) and cadmium in northern
fur seals (Callorhinus ursinus) (Goldblatt and Anthony 1983) and narwhal (Monodon mo-
noceros) (Wagemann et al. 1983).
Our objective was therefore to collect and analyse samples and test the hypothesis that there
are no statistically significant differences in tissue levels of selected metals among belugas
from the St. Lawrence Estuary and five Arctic locations.

Methods

Samples of liver, kidney, and muscle were obtained from hunted animals from five
locations in the Arctic (Fig. 1) in conjunction with native hunts, and from animals found dead
in the St. Lawrence Estuary. The sampling methodology has been described previously
(Wagemann et al. 1983). Samples were excised with a clean knife and put into precleaned po-
lyethylene bags. Samples from the Arctic were frozen in the field, shipped on ice and stored
at -30°C until analyzed.
Samples were first dried and ground, then analysed for Cu, Zn, Cd, Flg, Pb and Se in
Laboratory 1 (R. Wagemann) according to Wagemann et al. (1983). The quality of data on
heavy metals in biota from this laboratory has been verified and discussed elsewhere
(Wagemann et al. 1983; Wagemann and Armstrong 1988). The St. Lawrence Estuary
samples were analysed fresh in Laboratory 2 (C. Desjardins) which was also subject to quality
control. Precision and accuracy in both laboratories were monitored by reference to standard
samples, by participation in quality check programs (Federal Interdepartmental, Internatio-
nal Atomic Energy Commission), and by monthly checks on the recovery of metals. The

192
 Fia. 1. Capture locations ( e) of belugas.
methods used in the two laboratories were similar and were verified by analysing five tissue
samples for Hg, Cd, and Pb, in each laboratory. The values obtained agreed within 10% for
Hg, 20% for Cd, and 30% for Pb. For St. Lawrence Estuary belugas, results are reported as
obtained by Laboratory 2, but converted to a dry weight basis using the average moisture
content of the tissues. (Conversion factors were 3.748 for liver; 4.623 for kidney; 3.610 for
muscle). Otherwise, moisture was determined on individual samples.
Ages of animals were determined by counts of dentinal annuli in longitudinal thin sections
of mandibular teeth, assigning two growth layer groups to a year (Goren et al. 1987). To
minimize variations, all sections were examined by the same observer, and were read up to
five times each as blind replicates.
Each set of data from the six areas for each metal and each tissue (i.e. 103 sets; Pb-kidney
data unavailable from 5 locations) were tested for normality using the Martinez and Iglewicz
(1981) test. All groups passed the test except Cu and Pb in muscle of the Mackenzie group.
One of these (Cu) passed the normality test when a single high outlier was omitted; the other
did not and was omitted from the comparison. Values outside the 99.9 % confidence interval
of the mean were considered outliers, but if they did not negate the normality test they were
retained in the calculation of the mean. The means were compared using Scheffe's (1959)
multiple means test (a = 0.05).
Robust (Andrews' sine) multiple linear regression was performed on untransformed data
(dry wt) of metals on age and metals on metals. Essentially the same associations between
the variables were found with log-transformed as untransformed data in multiple regression
of metal concentrations in other marine mammal tissues (Wagemann et al. 1983). All
variables were designated sequentially once as the dependent variable and once as the
independent variable, and variables were retained when they were significant both ways i.e.

193
as the dependent and independent variable. A critical probability level of 0.90 was chosen for
retention of any variable in the model. The t-test (o = 0.05) was used to compare two means.
Discriminant analysis was performed on all the groups with all metal variables included,
using linear discriminant functions for classification.
Some samples differed in their age distributions. No correction was necessary for metal
concentrations that were independent of age. For metals that varied with age, age was
included as a factor when comparing areas.

Results

Beluga samples were taken from five locations across the Canadian Arctic and from the St.
Lawrence Estuary between 1981 and 1987 (Table 1). The relatively small sample size from
multiple years precluded an assessment of variation with time and for statistical purposes
samples from different years were pooled for each location. The number of animals sampled
at each location and their mean age varied among sites. Whales sampled in the St. Lawrence
were significantly older than those from Eskimo Point and Grise Fiord, and the latter group
was significantly younger than the Mackenzie Delta group.
The concentration of Pb, Cd and Hg varied considerably within each group, as indicated
by the relatively large standard deviations of the group mean for each metal (Tables 2-4).
Variation in concentration was generally larger for nonessential than essential metals (Table
5). Nevertheless, the mean lead concentration in animals from the St. Lawrence Estuary was
significantly higher (P <0.05) in both liver and muscle than in any other group except the
Nastapoka group in muscle (Table 6). There were no significant differences in the mean lead
concentration among the 'Arctic locations, although the Nastapoka group had also a relatively
high mean lead concentration in both liver and muscle (0.17 and 0.19 p.g/g dry wt,
respectively). The within-group variations were relatively large (Table 5). In the St.
Lawrence Estuary group, the level of lead in the kidney was approximately the saine as in the
liver. Lead was not measured in the kidney of other groups.

TABLE 1. Mean age, standard deviations and ranges, year of collection and number (n) of belugas
collected from different locations.

Location Lat./Long. Age Year n

Mackenzie 68° 50'N 12.3±4.4' 1981 35
Delta 136° 15'W (2-22)
Mackenzie 68° 50'N 21.I±9.2' 1984 8
Delta 136° 15'W (12.5-42)
Grise Fiord 76° 35'N 5.6±4.8 1984 17
83° 14'W (1-21)
Pangnirtung 66° 06'N 11.2±3.7 1984 11
65°57'W (6-17)
Eskimo Point 61° 06'N 11.2±6.7 1984 23
93° 59'W (0-24)
Nastapoka 56° 55'N 13.2±7.7 1984 15
(Hudson Bay) 76" 33'W (1-33)
St. Lawrence 48° 26'N 17.8±9.9' 1987 8
Estuary 68° 33'W (0-30)
St. Lawrence 48° 26'N 17.4±8.5 b 1982-86 27
Estuary 68° 33'W (2.5-29)
Mean age of both groups combined = 13.9±5.5 years.
Mean age of both groups combined = 17.5±8.8 years.

194
TABLE 2. Mean levels of metals (tig/g, dry wt) and percent moisture (%M), standard deviations (),
ranges, and [n] in liver of belugas from different Canadian locations.
Mackenzie Grise Pangnirtung Eskimo Nastapoka St.Lawrence
Delta Fiord Point
Cu 50.3 39.1 60.7 117. 150. 37.3
(49.0) (21.1) (36.0) (250.) (200) (34.5)
3.0-220 3.14-96.7 27.4-153 3.68-1241 20.5-699 0.34-162
[43] [17] [11] [23] [15] [30]

Pb 0.038 0.044 0.034 0.082 0.17 0.59

(0.038) (0.028) (0.023) (0.237) (0.16) (0.63)
0.01-0.17 0.001-0.09 0.003-0.068 <0.001-1.16 0.039-0.60 0.004-2.13
[39] [16] [11] [23] [15] [30]

Cd 8.52 12.2 23.5 25.0 18.9 0.58

(5.42) (14.1) (23.9) (22.9) (9.88) (0.41)
0.28-23.6 0.64-63.3 2.74-71.2 0.03-97 3.47-39.6 <0.005-1.5
[43] [17] [11] [23] [15] [30]

Hg 44.1 8.27 18.7 24.9 38.4 126.

(45.5) (7.71) (16.8) (25.2) (48.0) (161.)
0.46-182 1.42-29.4 3.59-58.7 0.04-93 0.60-152 1.42-756
[42] [17] [11] [23] [15] [30]

Se 23.3 9.21 10.3 15.7 16.7 79.2

(19.7) (4.74) (5.57) (8.78) (7.97) (110.)
0.92-87.4 1.97-14.6 3.94-22.7 1.40-28.4 2.88-28.5 2.72-307
[43] [17] [11] [23] [10] [7]

Zn 92.0 93.6 101. 90.4 93.2 98.4

(15.5) (18.8) (25.9) (31.4) (9.99) (41.8)
50.4-123 38.3-129 67.5-145 46.8-208 79.6-118 8.62-211
[43] [17] [11] [23] [15] [30]

%M 72.8 74.8 73.0 73.7 72.8 No Data

(2.29) (2.53) (1.86) (1.90) (1.88)
66.5-78.8 68.4-78.4 69.3-76.1 69.4-78.2 70.5-77.4
[43] [17] [11] [23] [15]

The mean cadmium concentration was significantly lower in the liver and kidney of the St.
Lawrence Estuary group than in all other groups, except the Mackenzie and Grise Fiord
groups for liver, and the Grise Fiord group for kidney (Table 6). The Pangnirtung and Eskimo
Point groups had the highest cadmium concentrations in both liver and kidney and did not
differ significantly from each other. For kidney, the two groups were significantly higher
than all other groups except the Nastapoka group, and for liver the Eskimo Point group was
also significantly higher than the Mackenzie Delta group. In muscle there were no significant
differences in cadmium among any of the groups. The concentration of kidney/liver
cadmium was in the range of 3 to 5 in belugas from different Arctic localities and 12 in belugas
from the St. Lawrence Estuary.
The St. Lawrence Estuary group had the highest mean mercury concentration in the tissues
examined. Mercury levels were 3 to 15 times and 2 to 4 times higher in the liver and muscle
respectively than in all other groups, which did not differ significantly from each other. When
the values for hepatic mercury in the St. Lawrence Estuary group were compared to the
regression of mercury in liver on age for all Arctic groups (Fig. 2), 37% exceeded the upper
95% confidence limit for a single point; less than 4% of the Arctic animals exceeded this limit.

195
TABLE 3. Mean levels of metals (.1g/g, dry wt) and percent moisture (%M), standard deviations (),
ranges, and [n] in kidney of belugas from different Canadian locations.
Mackenzie Grise Pangnirtung Eskimo Nastapoka St.Lawrence
Delta Fiord Point
Cu 10.1 13.1 12.6 11.8 11.9 19.0
(1.67) (3.28) (2.86) (1.48) (1.43) (21.9)
7.3-16.7 9.32-24.6 9.37-19.4 8.82-14.8 8.66-14.3 7.5-123
[43] [17] [11] [22] [14] [30]

Pb No Data No Data No Data No Data No Data 0.69

(0.47)
0.005-1.76
[30]

Cd 44.2 42.0 106. 96.6 69.0 7.01

(20.7) (22.3) (75.0) (64.3) (35.1) (4.11)
2.9-106 5.39-91.2 16.1-277 0.052-275 17.2-133 <0.005-17.9
[43] [17] [11] [22] [14] [30]

Hg 13.1 6.81 14.4 11.3 13.4 29.5

(7.9) (4.91) (5.98) (6.26) (9.58) (42.4)
0.90-43.6 1.54-19.0 2.44-24.5 1.01-23.9 0.80-32.1 1.48-228
[43] [17] [11] [22] [14] [30]

Se 11.3 8.95 11.7 12.7 12.6 13.4

(3.45) (2.12) (2.97) (4.08) (4.21) (6.63)
4.60-25.3 5.65-12.5 6.76-16.2 3.62-19.7 5.85-20.9 5.22-22.5
[43] [17] [11] [22] [14] [8]

Zn 110. 108. 121. 110. 115. 110.

(18.6) (16.9) (25.7) (17.5) (14.0) (43.3)
73.2-170 74.8-131 89.7-185 84.1-144 94.6-151 49.9-227
[43] [17] [11] [22] [14] [30]

%M 78.4 78.2 78.2 79.1 77.5 No Data

(2.99) (2.29) (1.11) (1.19) (1.44)
72.2-92.4 71.8-80.9 75.8-79.8 77.3-81.1 75.5-79.5
[43] [17] [11] [22] [14]

For mercury in the muscle, the St. Lawrence Estuary group mean, was significantly higher
than all other means except the Pangnirtung group mean, and for liver it was also significan-
tly higher than all other groups except the Nastapoka group. The mean mercury concentration
in the kidney of the St. Lawrence Estuary group, although also higher than the others (2 to
4 times), was significantly higher only than the Grise Fiord group mean. The other groups did
not differ significantly from each other.
The St. Lawrence Estuary group also had a significantly higher mean selenium concentra-
tion in the liver compared to all other groups, which did not differ significantly from each
other. In kidney and muscle, selenium means did not differ significantly among any of the
groups.
Zinc concentration ranged within groups by a factor of two or more and there was no
significant difference in mean concentration among the six groups for the three tissues
examined. Mean copper concentration in liver and muscle did not differ significantly among
any of the groups. Copper in the kidney was highest in the St. Lawrence Estuary group and
was lowest in the Mackenzie Delta group, and only these two groups differed significantly.

196
TABLE 4. Mean levels of metals (.tg/g, dry wt) and percent moisture (%M), standard deviations (),
ranges, and [n] in muscle of belugas from different Canadian locations.
Mackénzie Grise Pangnirtung Eskimo Nastapoka St.Lawrence
Delta Fiord Point
4.58a 3.25 3.12 3.92 3.12 5.37
(5.25) (0.44) (0.26) (2.40) (0.40) (4.62)
2.53-34.0 2.68-3.96 2.70-3.51 2.66-14.1 2.51-4.03 3.03-17.6
[34] [16] [1 I ] [23] [14] [9]

Pb 0.024" 0.023 0.028 0.028 0.187 0.37

(0.025) (0.046) (0.012) (0.016) (0.338) (0.45)
0.01-0.1 <0.001-0.19 0.011-0.050 0.008-0.085 0.021-1.23 0.01-1.37
[27] [16] [11] [23] [14] [9]

Cd 0.43 0.069 0.22 0.65 0.16 0.038

(0.96) (0.049) (0.17) (2.30) (0.18) (0.044)
0.03-5.35 0.027-0.22 0.026-0.57 0.015-11.2 0.01-0.60 <0.005-0.14
[35] [16] [11] [23] [14] [9]

Hg 3.86 2.39 3.52 3.13 3.00 8.88

(5.32) (1.01) (0.80) (2.07) (2.00) (5.28)
1.16-33.4 1.13-4.49 1.56-4.64 0.36-8.68 0.43-7.00 3.2-20
[34] [16] [11] [23] [14] [9]

Se 2.20 1.17 1.15 1.77 1.52 2.40

(5.32) (0.11) (0.15) (1.12) (0.55) (1.76)
0.93-32.8 1.01-1.38 0.83-1.36 1.15-6.44 0.54-2.97 1.25-5.76
[35] [16] [11] [23] [14] [6]

Zn 80.5 68.7 83.4 65.2 74.0 81.9

(17.9) (18.0) (15.5) (15.2) (22.5) (15.3)
39.3-116 49.3-127 63.7-105 36.4-102 43.6-122 71.5-120
[35] [16] [11] [23] [14] [9]

%M 71.9 71.9 72.0 73.1 72.7 No Data

(2.32) (1.11) (1.53) (1.87) (1.37)
64.3-75.1 69.7-74.0 70.1-75.3 67.4-75.5 70.2-74.3
[35] [16] [11] [23] [14]
a
One high value (241 lig/g) excluded from mean.
One high value (2.02 tig/g) excluded from mean.

Discriminant analysis performed on all the groups with all metal variables included did not
reveal any geographic differences. Robust multiple linear regression, performed on all
groups combined, revealed a number of associations among metals and between age and
metals in the three tissues examined (Table 8). Only lead was not associated significantly with
other metals or age in the tissues examined.
Cadmium in the kidney was associated positively with selenium, zinc, and age of animals,
and negatively with mercury, and in the liver positively with zinc. In the muscle cadmium was
only associated positively with age. Mercury was strongly, positively associated with sele-
nium in all three tissues, and in both liver and kidney was strongly, positively associated with
age. Mercury in kidney was negatively associated with zinc and cadmium. Mercury in muscle
was not associated with age, but mercury and zinc were positively associated in muscle.
Selenium, in addition to the relations already mentioned, was associated positively with zinc
and cadmium in the kidney. Copper in all three tissues was strongly negatively associated
with age of animals.

197
Table 5. Coefficient of variation (%) for mean metal concentrations in liver (L), kidney (K) and muscle
(M) of belugas from various locations.
Mackenzie Grise Pangnirtung Eskimo Nastapoka St.Lawrence
Delta Fiord Point
Cu L 97.3 54.1 59.3 75.6 133.0 92.4
K 16.6 25.1 22.8 12.6 12.0 116.0
M 115.0 13.6 8.2 61.1 12.9 86.0

Pb L 99.1 63.1 69.1 104. 0 93.9 106.0

M 103.0 199,0 41,9 56.6 152.0 121.0

Cd L 63.6 116.0 101.0 91.9 52.3 70.1

K 46.8 53.2 71.0 66.5 50.8 58.8
M 221.0 71.2 79.0 352.0 109.0 117.0

Hg L 101.0 93.3 90.2 101.0 125.0 128.0

K 60.3 72.2 41.6 55.2 71.7 144.0
M 138.0 42.1 22.8 66.3 65.4 59.4

Se L 84.5 51.5 54.2 55.9 47.6 139.0

K 30.4 23.6 25.3 32.1 33.3 49.3
M 242.0 9.0 13.1 63.4 36.5 75.6

Zn L 16.9 20.1 25.7 34.7 10.7 42.5

K 16.9 15.7 21.2 15.9 12,1 39.3
M 22.2 26.2 18.6 23.3 30.4 18.7

Discussion

We tested the hypothesis that metal concentrations in three tissues were the same in belugas
from six different putative stocks. This hypothesis was rejected. Belugas from the St.
Lawrence Estuary had significantly higher lead and mercury levels in liver and muscle,
higher selenium levels in liver, and lower cadmium levels in liver and kidney than most Arctic
groups. Belugas from Eskimo Point had significantly more cadmium in their livers than did
whales from the Mackenzie Delta. Cadmium levels in kidney of whales from Eskimo Point
and Pangnirtung were significantly higher than in whales from the Mackenzie Delta and
Grise Fiord.
The sample from the St. Lawrence Estuary may have been biased with respect to
contaminant levels. The animals were found dead and the cause of death was not established.
If the high lead and mercury concentrations contributed to the whales' deaths, then our
sample from this area was biased in favour of high metal levels. If these metals did not
contribute to the death of the whales then the reported levels represent the population at large.
There are no data to evaluate these potential biases but this does not negate indications of lead
and mercury pollution in this aquatic system.

198
TABLE 6. Significant differences (Cs =0.05, not underlined) of multiple means of a metal concentration
among different locations for three tissues (Scheffe's test).
Metal Liver Kidney Muscle

Cu 621345 145326 123456

Pb 123456 No Data 23456

Cd 612534 621543 123456

Hg 123456 241536 254136

Se 123456 123456 123456

Zn 123456 123456 123456

1= Mackenzie Delta 4 = Eskimo Point

2 = Grise Fiord 5 = Nastapoka
3 = Pangnirtung 6= St. Lawrence Estuary

750-1

650

cr) 550

450

350 Do

250
0
150 0— --Er-- -----
---
0— --•
50
-&-c8
—50
0 10 20 30 40

Age, Years
Fia. 2. Regression (Hg =0.47+1.52 Age) using all groups except the St. Lawrence Estuary group, of
Hg ( lig/g, dry wt) in liver on age (years) of animals with 95% confidence bands, and the St. Lawrence
Estuary group superimposed. Each symbol represents one animal. St. Lawrence group (D); all other
groups (0).

199
 Belugas from the St. Lawrence Estuary had, on average, about 10 times more lead in liver
and muscle than did Arctic belugas (Tables 2,4). This and the presence of lead in fish sampled
from the St. Lawrence River in 1981 (Wong et al. 1988) is suggestive of lead pollution in this
aquatic system. Regression analysis showed that lead was not correlated with age in any of
the soft tissues of the belugas examined, although lead does accumulate with age in hard
tissues (Grob ler et al. 1985). We found no previous reports on lead in belugas. Compared to
narwhal (liver: 0.109 iig/g dry wt; muscle: 0.031 i.t.g/g dry wt, Wagemann et al. 1983), the lead
levels in St. Lawrence belugas were higher. In most Arctic belugas the levels were similar
to those in narwhal.
The mean mercury concentration in the liver of the St. Lawrence group was significantly
higher than in all others except Nastapoka; in muscle it was also higher than in all others
except Pangnirtung. Mercury accumulated in liver and kidney with increasing age as it does
in other marine mammals (Perttilâ et al. 1986; Norstrom et al. 1986; Wagemann et al. 1983).
The St. Lawrence Estuary whales had a higher mean age than the Eskimo Point and Grise
Fiord whales. The higher age of of the St. Lawrence animals could not account entirely for
their high mercury levels. More than a third of the animals in this group exceeded the upper
confidence limit (95%) of the hepatic mercury concentration when mercury was regressed
on age (Fig. 2). Also, St. Lawrence belugas had a high liver/kidney mercury ratio. Although
the liver is the preferred organ for mercury accumulation in mammals generally, and the
examined belugas were no exception, mercury accumulated to a greater extent in liver rela-
tive to the kidney in St. Lawrence Estuary belugas than in Arctic belugas. In belugas from
the St. Lawrence Estuary the liver/kidney ratio of the mean mercury concentration was 4.3,
and in Arctic belugas 3.4.
There are only a few previous reports of mercury levels in belugas (Table 7). Sergeant's
(1980) determination of mercury in the liver of one beluga from the St. Lawrence (36.9 ppm
=- 138 p.g/g dry wt) is not substantially different from our mean. The levels in St. Lawrence
whales were also similar to those found in Mackenzie Delta belugas in 1978 (Table 7). Levels
in the Arctic groups were similar to those previously published for western Greenland
(Hansen, C. T., personal communication, Greenland Environmental Research Institute,
Copenhagen, Denmark) and Hudson Bay (Bligh and Armstrong 1971).
Mercury concentrations in Mackenzie Delta belugas were lowest in 1972 and highest in
1978 (Table 7) and apparently declined thereafter. Drilling for oil and gas and artificial island
construction began in the Beaufort Sea in 1973 and peaked toward the end of that decade. It
is known that the formation of freshwater impoundments increases the mercury levels in fish
in such impoundments (Bodaly et al. 1984; Boucher et al. 1985). Whether or not sediment
disturbances in a more open system such as the Beaufort Sea could have a similar effect on
upper trophic levels is unknown. Although the pie -activity data (1972) are the lowest, and
the highest concentrations coincide with peak activity in the Beaufort Sea, a coincidental
correspondence cannot be ruled out in view of the small sample sizes (n = 7 for 1972, 11 = 8
for 1978) and the lack of age estimates for these samples.
Comparison with other marine mammals is helpful in obtaining a broader ecological
perspective. Mercury levels were higher in the St. Lawrence belugas than previously reported
values for northern hemisphere cetaceans except the short- finned pilot whales (Globicepha-
la macrorhyncha) from Georgia, USA. (Wagemann and Muir 1984). They were also higher
than most published values for northern pinnipeds (Sergeant 1980; Goldblatt and Anthony
1983; Ronald et al. 1984; Wagemann and Muir 1984; Helle 1985). The exceptions were
ringed seals from the Baltic (n = 6) and Lake Siamma (n = 3) (Wagemann and Muir 1984)
and from Boothnia Bay (n = 22) (Helle 1985); grey seals (Halichoerus grypus) from the
United Kingdom (n = 98) and Sable Island (n =7); harbour seals (Phoca vitulina) from New
Brunswick (n = 1), the Netherlands (n = 5), Califo rn ia (n = 4) and England (n =- 1); hooded
seals (Cystophora cristata) from the Gulf of St. Lawrence (n = 3); and bearded seals
(Erignathus barbatus) from the Canadian Arctic (n = 6) (Wagemann and Muir 1984). The

200
Table 7. Mean heavy metal concentrations (.1g/g, dry wt)" and sample sizes (n) in liver and muscle of
belugas from the Arctic.
Cd Hg Pb Zn Location, Source
Year
Liver
No Data 23.5 No Data No Data Mackenzie Lutz and Armstronge
(7) Delta, 1972 (1972, unpublished)
11.1 115 No Data 111 Mackenzie Beak Consultants
(8) (8) Delta, 1977 (1978)

No Data 33.2 No Data No Data Hudson Bay, Bligh and Armstrong

(1) 1971 (1971)

5-15 " 3.82-61.5" No Data 106-112" Western Hansen, C. T. 1989

(43) (43) (43) Greenland, (personal comm.)
1984-86
Muscle
No Data 2.56 No Data No Data Mackenzie Lutz and Armstrong'
(7) Delta, 1972 (1972, unpublished)

0.41 7.67 No Data 92.7 Mackenzie Beak Consultants

(11) (11) (8) Delta, 1977 (1978)

No Data 3.50 No Data No Data Hudson Bay, Bligh and Armstrong

(1) 1971 (1971)

No Data 1.91 No Data No Data Hudson Bay, Bligh and Armstrong

(43) 1971 (1971)

0.1" 0.03-4.73" No Data 97.5-118" Western Hansen, C.T. 1989

(43) (43) (43) Greenland, (personal comm.)
1984-86
" Converted from a wet to a dry weight basis using factors 3.748 and 3.610 for liver
and muscle, respectively.
" Median range, converted to dry weight with above-given factors.
Freshwater Institute, Contaminants and Toxicology, Winnipeg, Canada.

small sample sizes of the published data preclude detailed analysis by age, gender and
geographic location.
The mean concentration of cadmium was significantly lower in the St. Lawrence Estuary
group than in all Arctic groups except the Mackenzie (liver) and Grise Fiord (liver, kidney)
whales. Based on the strong, positive correlation of cadmium with age found in most marine
mammals (Hamanaka et al. 1982; Honda et al. 1983; Wagemann et al. 1983; Muir et al. 1988),
higher cadmium levels were expected in the significantly older St. Lawrence group than were
expected to levels in the younger Grise Fiord whales. The cadmium concentration in the St.
Lawrence group was also lower than in most other northern hemisphere marine mammals
(Goldblatt and Anthony 1983; Wagemann et al. 1983; Ronald et al. 1984; Wagemann and
Muir 1984; Hel le 1985; Steinhagen-Schneider 1986; Tohyama et al. 1986). The relatively
high mercury levels in the St. Lawrence animals may have played a role in the low cadmium
levels by competing with cadmium in metallothionein where almost all cadmium is stored
(Wagemann et al. 1984). The negative correlation between cadmium and mercury in the
kidney, the preferred organ for cadmium accumulation, would tend to support this hypothe-
sized competition.

201
 Belugas from Pangnirtung and Eskimo Point had higher cadmium levels than those from
the Mackenzie Delta (liver and kidney) and Grise Fiord (Table 6). There were no significant
differences in mean age among the three eastern Arctic groups, although the Grise Fiord mean
was low. Levels of cadmium in the Pangnirtung and Eskimo Point groups were also higher
than in belugas from western Greenland (Table 7, C.T. Hansen, Greenland Environmental
Research Institute, Copenhagen, Denmark, personal communication) and white-beaked
dolphins (Lagenorhynchus albirostris) from the coast of Newfoundland, but were lower than
in long-finned pilot whales (Globicephala melaena) from Newfoundland (Muir et al. 1988)
and narwhals from northe rn Baffin Island (Table 7, Wagemann et al. 1983). Whales from
Pangnirtung, Eskimo Point, and Nastapoka probably spend part of the year in Hudson Strait
(Brodie et al. 1981; Richard and Orr 1986; Smith and Hammill 1986), but whether or not this
has a bearing on the higher cadmium levels in their tissues is unknown.
Selenium in liver of the St. Lawrence group was the highest among all the groups. In kidney
and muscle it did not differ significantly from any of the other groups, and was significantly,
positively correlated with cadmium in kidney in the pooled sample. While selenium may
reduce cadmium toxicity (Parfzek et al. 1971; Whanger 1979; Ridlington and Whanger,
1981), a positive correlation between selenium and cadmium indicates that it does hot do so
by lowering the tissue concentration of cadmium. In the liver, however, there was no such
positive correlation and the low cadmium levels in the St. Lawrence Estuary group were
accompanied by high selenium levels in that group. The high selenium levels in the liver of
St. Lawrence Estuary animals were a direct reflection of the high mercury concentrations in
this group. Such a correlation was reported also for other marine mammals (Koeman et al.
1973; Wagemann et al. 1983; Norstrom et al. 1986).
There were no significant differences in the mean copper concentrations in liver and muscle
among any of the groups, although four animals from the Nastapoka River had inexplicably
high concentrations in liver (200-700 j_tg/g dry wt) and one animal from Eskimo Point had
an even higher concentration (1241 p.g/g dry wt). Copper in kidney was significantly higher
in the St. Lawrence Estuary group than in the Mackenzie Delta group. The reason for this
is unknown. In the pooled sample, copper declined in all tissues with age, indicating either
an actual loss of the metal from tissues with age or a dilution of it by tissue growth.
The effects of a changing chemical environment on biota are difficult to measure and
assess, and it is particularly difficult to determine sublethal effects of heavy metals on marine
mammals in the wild. High levels of nickel were found in the hair of still-born pups of ringed

TABLE 8. Associations between the dependent and independent variables (age, metals, selenium) in
beluga whale tissues (all groups combined) at 0.90 5_ P < 0.95, at 0.95 P < 0.99 (underlined), and at
0.99 P (doubly underlined).
Liver Kidney Muscle
Dependent
Variable Independent Variables

Cu - AU -Age - Age
Pb Not Assoc. No Data Not Assoc.

Cd +Zn +Age, -Hg, +Se, +Zn +Age

Hg +Age, +Se +Age., -Cd, +Se, -Zn +Se, +Zn

Se +Hg +CA, +Hg, +ZE

Zn +Cd +Cd, -Hg, +Se Not Assoc.

202
seals from Lake Saimaa in eastern Finland (Hyvarinen and Sipila 1984). Harp seals fed
methyl mercury suffered from a variety of health disorders (Freeman et al. 1975; Ramprashad
and Ronald 1977). Common seals (Phoca vitulina) fed fish from polluted waters off the coast
of the Netherlands showed signs of reproductive failure (Reijnders 1986) and vitamin A and
thyroid deficiencies (Brouwer et al. 1989). Only organochlorines were investigated in this
study although the waters from which these fish came are polluted also with heavy metals
(Reijnders 1980). Less is known about sublethal effects in marine mammals than in terrestrial
mammals, in which even low exposures to some toxic metals can affect the nervous system
(Maehara et al. 1986; Yamamura et al. 1987), behavior and learning ability, before any overt
clinical signs appear.
Our study showed that the anthropogenic pollutants, lead and mercury, have found their
way into belugas in the St. Lawrence Estuary. There is therefore reason to speculate that these
pollutants together with the high organochlorine levels in these animals (Desjardins et al.
1983a, 1983b; Martineau et al. 1987; Muir et al. 1990) may adversely affect the health of these
belugas. Although a definite connection between pollutants of any kind in the St. Lawrence
Estuary and death of animals has not been established, Martineau et al. (1985, 1986, 1987,
1988) found the condition of dead, stranded belugas in the St. Lawrence Estuary to be
consistent with organochlorine and polycyclic hydrocarbon-induced insults. Findings inclu-
ded a high incidence of tumors, gastric erosions and ulcers, as well as multisystemic diseases
and bacterial infections in the more contaminated specimens (Martineau et al. 1988). Any
adverse effects would be the result of the combined action of all toxic elements and
compounds in the tissues, acting over a relatively long time. The depuration rate of heavy
metals in belugas is unknown, but in humans the biological half-life for lead and mercury is
about 10 years (Task Group 1973; Friberg et al. 1979) and for cadmium it may be even longer,
at 10-30 years (Friberg et al. 1974; Kjellstreim and Nordberg 1978).
The population of belugas in the St. Lawrence is thought to be declining (Reeves and
Mitchell 1984; Béland et al. 1988). In spite of the closure of a chlor-alkali plant on the
Saguenay river in 1976 and an alkyl lead production plant closure in 1985 (DuPont, Canada
Ltd. at Maitland), the state of pollution of the St. Lawrence and the declining stock of belugas
there remain a concern (Boychuk 1988; International Forum 1988).

Acknowledgments

We extend sincere thanks to the Hunters and Trappers Association and hunters of
Pangnirtung, Eskimo Point and Grise Fiord for providing most of the animals, to Dr. W.
Doidge for providing samples from Nastapoka, to R. Walker and P. Weaver for sampling,
subsampling and ageing animals, to R. Plante (Filmar Inc.) for collecting and processing
samples and tooth preparations of St. Lawrence Estuary animals, and to G. Boila and D.
Savoie in Winnipeg, and R. Maltais, G. Leblanc and P. Arpin in Longueuil for analytical
work. We also thank Dr. E. Pelletier, INRS-Océanologie, Rimouski, and three anonymous
reviewers for constructive comments on an earlier manuscript, and Dr. C.T. Hansen for
providing data on belugas from western Greenland. The enthusiastic assistance with subsam-
pling and analysis by summer students is gratefully acknowledged. Collection and analysis
of the St. Lawrence specimens was supported through a grant from the Wildlife Toxicology
Fund.

203
 References

BEAK CONSULTANTS. 1978. Heavy metals pro- 1983b. Contamination de l'angille (An-
ject: Mackenzie Delta and estuary. A report for guilla rostrata) du bassin du fleuve Saint-
Imperial Oil Ltd. Calgary, Alberta, Canada. Laurent par le mirex. Rapp. can. ind. sci. ha-
BÉLAND, P., A. VÉZINA, AND D. MARTINEAU. lieut. aquat. 141: v+52 p.
1988. Potential for growth of the St. Lawrence DRASCH, G. A. 1983. An increase of cadmium
(Québec, Canada) beluga whale (Delphinap- body burden for this century - an investiga-
terus leucas) population based on modelling. tion of human tissues. Sci. Total Environ. 26:
J. Cons. Cons. Int. Explor. Mer 45: 22-32. 111-119.
BLIGH, E.G., AND F.A.L ARMSTRONG. 1971. DUINKER, J. C., M. Tn. J. HILDEBRAND, AND R. F.
Marine mercury pollution in Canada. ICES, NOLTING. 1979. Organochlorines, and metals
C.M. 1971/E 34. Fisheries Improvement in harbour seals (Dutch Wadden Sea) , Mar.
Committee. Pollut. Bull. 10: 360-364.
BODALY, R. A., R. E. HECK Y, AND R. J. P. FUDGE. ERICSON, J. E., H. SHIRAHATA, AND C. C. PATTER-
1984. Increases in fish mercury levels in lakes SON. 1979. Skeletal concentrations of lead in
flooded by the Churchill River diversion, ancient Peruvians. N. Engl. J. Med. 300:
northern Manitoba. Can. J. Fish. Aquat. Sci. 946-951.
41: 682-691. FREEMAN, H. C, G. SANGALANG, J. F. UTHE, AND
BOUCHER, R., R. SCHETAGNE, ET E. MAGNIN. K. ROLAND. 1975. Steriogenesis in vitro in the
1985. Teneur en mercure des poissons des harp seal (Pagophilus groenlandicus) without
réservoirs La Grande 2 et Opinaca (Québec, and with methyl mercury treatment in vivo.
Canada) avant et aprés mise en eau. Rev. Fr. Environ. Physiol. Biochem. 5: 428-439 ,
Sci. Eau 4: 193-206. FRIBERG, L., G.F. NORDBERG, AND V. VOUK
BOURGOIN, B. P., AND M. J. RISK. 1987. Historical [ed.]. 1979. Handbook on the toxicology of
changes in lead in the eastern Canadian Arctic, metals. Elsevier/North-Holland Biomedical
determined from fossil and modern Mya trun- Press, Amsterdam.
cata shells. Sci. Total Environ. 67: 287-291. FRIBERG, L., M. PISCATOR, G. NORDBERG, AND T.
BOUTRON, C., AND R. DELMAS. 1980. Historical KJELLSTRÔM. 1974. Cadmium in the environ-
record of global atmospheric pollution in Polar ment. 2nd ed. CRC Press, Cleveland, OH.
ice sheets. Ambio 9: 210-215. GOBEIL, C., D. COSSA, AND J. PIUZE. 1984. Niveau
BOYCHUK, R. 1988. St. Lawrence blues. Har- des concentrations en mercure dans les eaux
rowsmith, May/June issue, No. 79,36-45. du fjord du Saguenay (Québec) en octobre
BRAHAM, H. W. 1973. Lead in the California sea 1979 et septembre 1983. Rapp. tech. can.
lion (Zalophus californianus). Environ. Pol- hydrogr. sci. océan. 36: v+14 p.
lut. 5: 253-258. GOLDBLATT, C. J., AND R. G. ANTHONY. 1983.
BRODIE, P. F., J. L. PARSONS, AND D. E. SERGEANT. Heavy metals in northern fur seal (Callorhinus
1981. Present status of the white whale (Del- ursinus) from the Pribolof Islands, Alaska. J.
phinapterus leucas) in Cumberland Sound, Environ. Qual. 12: 478-482 ,
Baffin Island. Int. Whal. Comm. Rep. Comm. GOREN, A. D., P. F. BRODIE, S. SPOTTE, G. C. RAY,
31: 579-582. H. W. KAUFMAN, A. J. GWINNETT, J. J. SCIUB-
BROUWER, A., P. J. H. REIJNDERS AND J. H. KOE- BA, AND J. D. BucK. 1987. Growth layer groups
MAN. 1989. Polychlorinated biphenyl (PCB)- (GLGs) in the teeth of an adult belukha whale
contaminated fish induces vitamin A and thy- (Delphinapterus leucas) of known age: evi-
roid hormone deficiency in the common seal dence for two annual layers. Mar. Mamm. Sci.
(Phoca vitulina). Aquat. Toxicol. 15:99-106 , 3: 14-21.
COSSA, D., ETC. DESJARDINS. 1984. Evolution de GROBLER, S. R., R. J. Rossouw, AND D. KOTZE.
la concentration en mercure dans les crevettes 1985. Lead in teeth of weanling rats received
du fjord du Saguenay (Quebec) au cours de la via the maternal drinking water. Arch. Oral
période 1970-83. Rapp. tech , can. hydrogr. Biol. 30: 509-511.
sci. océan. 32: v+8 p HAMANAK A, T., T. ITOO, AND S. MISHIMA. 1982.
DESJARDINS, C., J. D. DUTIL, ET R. GÉLINAS, Age-related changes and distribution of cad-
1983a. Contamination de l'angille (Anguilla mium and zinc concentrations in the Steller
rostrata) du bassin du fleuve Saint-Laurent Sea Lion (Eumetopias juhata) from the coast
par les biphényles polychlorés. Rapp. can. ind. of Hokkaido, Japan. Mar. Pollut. Bull. 13:
sci. halieut. aquat. 144: v+56 p. 57-61.

204
HARMS, U., H. E. DRESCFIER, AND E. HUSCHEN MUIR, D. C. G., C. A. FORD, R. E. A. STEWART, T.
BETH. 1978. Further data on heavy metals and G. SMITH, R. F. ADDISON, M. E. ZINCK, AND P.
organochlorines in marine mammals from BÉLAND. 1990. Organochlorine contaminants
German coastal waters. Meeresforschung 26: in belugas, Delphinapterus leucas, from Ca-
153-161. nadian waters, p. 165-190. In T. G. Smith, D.
HELLE, E. 1985. Occurrence and effects of envi- J. St. Aubin, and J. R. Geraci [ed.] Advances in
ronmental toxins in seals in Finland. Suomen research on the beluga whale, Delphinapterus
Riista 32: 5-22 (Engl. Summary.) leucas. Can. Bull. Fish. Aquat. Sci. 224.
HONDA, K., R. TATSUKAWA, K. ITANO, N. MIYA- MUIR, D. C. G., R. WAGEMANN, N. P. GRIFT, R. J.
ZAKI AND T. FUJIYAMA. 1983. Heavy metal NORTSROM, M. SIMON, AND J. LIEN. 1988. Or-
concentrations in muscle, liver and kidney ganochlorine chemical and heavy metal con-
tissue of striped dolphin (Stenella coeruleoal- taminants in white-beaked dolphins (Lageno-
ha) and their variations with body length, rhynchus alhirostris) and pilot whales (Globi-
weight and sex. Agric. Biol. Chem. 47: cephala malaena) from the coast of Newfoun-
1219-1228. dland, Canada. Arch, Environ. Contam. Toxi-
HY VÂRINEN, H., AND T. SIPILÀ. 1984. Heavy col. 17: 613-629.
metals and high pup mortality in the Saimaa MUROZUMI, M., T. J. CHOW, AND C.C. PATTER-
ringed seal population in eastern Finland. Mar. SON. 1969. Chemical concentrations of poilu-
Pollut. Bull. 15: 335-337. tant lead aerosols, terrestrial dusts and sea salts
INTERNATIONAL FORUM ON THE FUTURE OF THE in Greenland and Antarctic snow strata. Geo-
BELUGAS. Sept. 1988, Tadoussac, Quebec. chim. Cosmochim. Acta 33: 1247-1294.
KJELLSTRÔM, T., AND G. F. NORDBERG. 1978. A No, A., AND C. C. PATTERSON. 1981. Natural con-
kinetic model of cadmium metabolism in the centrations of lead in ancient Arctic and An-
human being. Environ. Res. 16: 248-269. tarctic ice. Geochim. Cosmochim. Acta 45:
KOEMAN, J. H., W. H. M. PEETERS, C. H. M. KOUDS- 2109-2121.
TAAL-HO!, P. S. THOE, AND J. J. M. DEGOEIJ. NORSTROM, R. J., R. E. SCHWEINSBERG, AND B. T.
1973. Mercury-selenium correlations in ma- COLLINS. 1986. Heavy metals and essential
rine mammals. Nature (Lond.) 245: 385-386. elements in livers of the polar bear (Ursus ma-
MAEHARA, N., E. UCIIINO, K. TERAYAMA, H. ritimus) in the Canadian Arctic. Sci. Total En-
OHNO, AND K. YAMAMURA, 1986. Motor nerve viron. 48: 195-212.
conduction velocity (MCV) and lead content PARIZEK, J., I. ŒTÀDALOVÀ, J. KALOUSKOVÀ, A.
in sciatic nerve of lead-exposed rats. Bull. BABICK, AND J. BENE. 1971. The detoxifying
Environ. Contam. Toxicol. 37: 47-52. effects of selenium interrelations between com-
MARTINEAU, D., A. LAGACÉ, R. MASSÉ, M. MORIN, pounds of selenium and certain metals, p.
AND P. BÉLAND. 1985. Transitional cell carci- 85-122. /n W. Mertz and W.E. Cornatzer [ed.]
noma of the urinary bladder in beluga whale Newer trace elements in nutrition. Marcel
(Delphinapterus leucas). Can. Vet. J. 26: Dekker, Inc., New York, NY.
297-302. PATTERSON, C.C. 1965. Contaminated and natu-
MARTINEAU, D., A. LAGACÉ, P. BÉLAND, AND C. ral lead environments of man. Arch. Environ.
DESJARDINS. 1986. Rupture of a dissecting Health 11: 344-360.
aneurysm of the pulmonary trunk in a beluga PERTTILÂ, M., O. STENMAN, H. PYYSALO, AND K.
whale (Delphinapterus leucas). J. Wildl. Dis. WICKSTRÔM. 1986. Heavy metals and organo-
22: 289-294. chlorine compounds in seals in the Gulf of
MARTINEAU, D., P. BÉLAND, C. DESJARDINS, AND Finland. Mar. Environ. Res. 18: 43-54.
A. LAGACÉ. 1987. Levels of organochlorine RAMPRASHAD, F., AND K. RONALD. 1977. A sur-
chemicals in tissues of belugas whales (Del- face preparation study on the effect of methyl
phinapterus leucas) from the St. Lawrence mercury on the sensory hair cell population in
Estuary, Quebec, Canada. Arch. Environ. Con- the cochlea of the harp seal (Pagophilus groen-
tam. Toxicol. 16: 137-147. landicus Erxleben, 1777). Can. J. Zool. 55:
MARTINEAU, D., A. LAGACÉ, P. BÉLAND, R. 223-230.
HIGGINS, D. ARMSTRONG, AND L. R. SHUGART. REIF NDERS, P. J. H. 1980. Organochlorine and hea-
1988. Pathology of stranded belugas whales vy metal residues in harbour seals from the
(Delphinapterusleucas) from the St. Lawrence Wadden Sea and their possible effects on re-
Estuary, Quebéc, Canada. J. Comp. Path. 98: production. Netherlands J. Sea Res. 14:30-65.
287-310. 1986. Reproductive failure in common seals
MARTINEZ, J., AND B. IGLEWICZ. 1981. A test for feeding on fish from polluted coastal waters.
departure from normality based on a biweight Nature (Lond.) 324: 456-457.
estimator of scale. Biometrika 68: 331-333.

205
REEVES, R.R., AND E. MITCHELL. 1984. Catch ship of the increased level of metallothionein
history and initial population of white whales with heavy metal levels in tissues of the harbor
(Delphinapterus leucas) in the river and gulf seal (Phoca vitulina). Ecotoxicol. Environ.
of St. Lawrence, eastern Canada. Naturaliste Safety. 12: 85-94.
can. (Rev. Écol. Syst.) 111:63-121. V1ALE, D. 1978. Evidence of metal pollution in
RICHARD, P.R., AND J.R. ORR. 1986. A review of cetacea of the western Mediterranean. Ann.
the status and harvest of white whales (Delphi- Inst. Oceanogr. 54: 5-16.
napterus !emus) in the Cumberland Sound WAGEm ANN, R. 1989. Comparison of heavy me-
area, Baffin Island. Can. Tech. Rep. Fish. tals in two groups of ringed seals (Phoca
Aquat. Sci. 1447: iv+25 p. hispida) from the Canadian Arctic. Can, J.
RIDLINGTON, J.W., AND P.D. WHANGER. 1981. Fish. Aquat. Sci. 46: 1558-1563.
Interactions of selenium and antioxidants with WAGEMANN, R., AND F. A.J. ARMSTRONG. 1988.
mercury, cadmium and silver. Fundam. Appl. Trace metal determination in animal tissues:
Toxicol. 1: 368-375. an interlaboratory comparison. Talanta 37:
RONALD, K., R.J. FRANK, J. DOUGAN, R. FRANK, 545-551.
AND H.E. BRAUN. 1984. Pollutants in harp WAGEMANN R., R. HUNT, AND J. F. KLAVER-
seals (Phoca groenlandica). II. Heavy metals KAMP. 1984. Subcellular distribution of heavy
and selenium. Sci. Total Environ. 38: 153-166. metals in liver and kidney of a narwhal whale
SCHEFFE, H. 1959. Analysis of variance. John (Monodon monoceros): an evaluation for the
Wiley and Sons, New York, NY. 477 p. presence of metallothionein. Comp. Biochem.
SERGEANT, D.E. 1980. Levels of mercury and or- Physiol. 28C: 301-307.
ganochlorine residues in tissues of sea mam- WAGEMANN, R., AND D. C. G. MUIR. 1984. Con-
mals from the St. Lawrence estuary. Int. Counc. centrations of heavy metals and organochlor-
Explor. Sea ICES C. M. 1980/E:55. ines in marine mammals of northern waters:
SMITH, T. G., AND F. A.J. ARMSTRONG. 1975. overview and evaluation. Can. Tech. Rep. Fish.
Mercury in seals, terrestrial carnivores, and Aquat. Sci. 1279: v+97 p.
principal food items of the Inuit from Holman, WAGEMANN, R., N. B. SNOW, A. LUTZ, AND D. P.
N.W.T. 1975. J. Fish. Res. Board Can. 32: SCOTT. 1983. Heavy metals in tissues and or-
795-801. gans of the narwhal (Monodon monoceros).
SMITH, T. G., AND M. O. HAMMILL. 1986. Popula- Can. J. Fish. Aquat. Sci. 40 (Suppl. 2): 206-216.
tion estimates of white whales (Delphinapte- WHANGER, P. D. 1979. Cadmium effects in rats on
rusleucas) in James Bay, eastern Hudson Bay, tissue iron, selenium, and blood pressure; blood
and Ungava Bay. Can. J. Fish. Aquat. Sci. 43: and hair cadmium in some Oregon residents.
1982-1987. Environ. Health Perspect. 28: 115-121.
STEINHAGEN-SCHNEIDER, G. 1986. Cadmium and WONG, P.T.S., Y. K. CHAU, J. YAROMICH. P.
copper levels in seals, penguins and skuas HUDSON, AND M. WHITTLE, 1988. Alkyllead
from the Weddell Sea in 1982/1983. Polar contaminations in the St. Lawrence River and
Biol. 5: 139-143. St. Clair River (1981-1987). Can. Tech. Rep.
TASK GROUP ON METAL ACCUMULATION. 1973. Fish, Aquat. Sci. 1602: x+ I 34 p.
Accumulation of toxic metals with specific re- YAMAMURA, K., N. MAEHARA, H. OHNO, N. UENO,
ference to their absorption, excretion and bio- A. KOHYAMA, T. TOSHIHIRO, T. SATOH, A.
logical half-times. Environ. Physiol. B iochem. SHIMODA, AND R. KISHI. 1987. Effects of me-
3: 65-107. thylmercuric chloride of low concentration on
TOHYAMA, C., S. HIMENTO, C. WATANABE, T. the rat nervous system. Bull. Environ. Con-
SUZUKI, AND M. MORITA. 1986. The relation- tam. Toxicol. 38: 985-993.

206
 DUE DATE

11(31

RégMI «id/.

MR. • •1M

201-6503 Printed
In USA