(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "North American fauna"

V 




^ 



Return vo 



lume, when bound, to 



i 



TI 



BOSTON PUBLIC LIBRARY 
BINDING INSTBUCTIONS 



u 



Author 




M 


1 


ICA*/ 


Title 












Volume 
Number 
Part 










: / 




Year 









*~ 






Call No 


5 





SPECIAL INSTRUCTIONS! 

Volume incomplete. Bind with stub 
for 



Volume incomplete. Bind without 

r (i i ** 



RETU1 B01D VOUIftk. . 



The 

TRUMPETER SWAN 



Frontispiece: A pair of trumpeters 
in flight over their breeding grounds 
on the Red Rock Lakes Refuge, 
southwestern Montana. 

BOSTON PUBLIC LIBRARY 



The 

TRUMPETER SWAN 

ITS HISTORY, HABITS, AND 
POPULATION IN THE UNITED STATES 



By 

Winston E. Banko 
Refuge Manager, Branch of Wildlije Refuges 
Bureau of Sport Fisheries and Wildlife 




NUMBER 63 



UNITED STATES 
DEPARTMENT OF THE INTERIOR 

Fred A. Seaton, Secretary 

FISH AND WILDLIFE SERVICE 

Arnie J. Suomela, Commissioner 

BUREAU OF SPORT FISHERIES AND WILDLIFE 

Daniel H. Janzen, Director 




North American Fauna, Number 63 



Published by U.S. Fish and Wildlife Service 
April 30, I960 



United States Government Printing Office • Washington • I960 



For sale by the Superintendent of Documents, U.S. Government Printing Office 
Washington 25, D.C. - Price $1.00 



CONTENTS 



Page 

Foreword 1 

Introduction 4 

Legend and tradition 4 

Description and systematics 5 

Distribution and status 8 

Geological occurrence 8 

Primitive history 9 

Early historical notes (1632-1832) 10 

Later historical notes (1833-1925) 14 

Early migration notes 25 

Spring migration (departure) 25 

Spring migration (arrival) 26 

Fall migration (departure) 27 

Fall migration (arrival) 27 

Recent occurrence, breeding, and migration reports, to 1957.. 28 

Pacific fly way 28 

Central fly way 35 

Mississippi and Atlantic flyways 37 

Habitat 38 

Breeding habitat 38 

Life zone characteristics 38 

Physical characteristics 39 

Red Rock Lakes Refuge 40 

Yellowstone National Park 48 

Copper River Basin , Alaska 52 

Wintering habitat 54 

Island Park 56 

Red Rock Lakes Refuge 58 

Yellowstone National Park 58 

National Elk Refuge 60 

Madison River 60 

Winter counts 60 

Life cycle 62 

Description 62 

Species description 62 

External appearan ce 65 

Voice 68 

Plumages and molting 70 

Flight 74 

Behavior and related characteristics 77 

Escape-distance 78 

Interspecific tolerance 79 

M emory 82 

Sensory perception 83 

Some general behavior attitudes 84 

Display 86 

Breeding 93 

Pair formation 94 



VI CONTENTS 

Page 

Breeding age — 94 

Mating fidelity 96 

Prenesting habits and behavior 99 

Territory and territorial behavior 100 

Nesting 111 

Nest sites 111 

Nest building 111 

Egg description 113 

Clutch size 114 

Egg laying 114 

Incubation 114 

Hatching data___ 115 

Renesting 116 

Cygnet development 117 

Food 122 

Feeding habits and food of young 122 

Feeding habits and food of older cygnets, immatures, and 

adults 124 

Limiting factors 130 

Egg failure 130 

Preflight cygnet mortality 131 

Mortality of immatures and adults 133 

Longevity 143 

Population 144 

Annual swan census, 1929-57 144 

Population dynamics 145 

Red Rock Lakes Refuge populations 151 

Yellowstone National Park populations. 155 

Population outside Red Rock Lakes Refuge and Yellow- 
stone National Park 160 

Summary 161 

Conclusions 163 

Discussion _. _ 163 

Management 165 

Protective legislation 165 

Captivity record. 167 

National Park Service investigations 172 

Swan management on the Red Rock Lakes Refuge _ 176 

General practices 176 

Transplanting program 178 

Banding 182 

Management recommendations _ 183 

BlBLIOGB A PHY 189 

Appendix 1 — Excerpt from "Observations on the genera of the 

swans", by Alexander Wetmore 198 

Appendix 2 — Status and distribution of trumpeter swans in the 

United States, 1954.. 199 

Appendix 3 — Measurements of trumpeter swan eggs.. 201 

Appendix 4 — Food analysis 202 

Appendix 5 — Supplementary data, annual swan census 205 

Index 207 



ILLUSTRATIONS 

Page 
Frontispiece. Pair of trumpeters in flight over their breeding 
grounds on the Red Rock Lakes Refuge, southwest- 
ern Montana. 
Figure 

1. Locations of some Hudson's Bay Company posts which en- 
gaged in swan-skin trade, 1828-84 — 18 

2. A day's bag of waterfowl in 1895 at Red Rock Lakes included 

a trumpeter swan 22 

3. Former breeding and wintering range, trumpeter swan 26 

4. Aerial view of trumpeter breeding grounds in lower Copper 
River Basin, Alaska, at confluence of Tasnuna, Bremner, and 
Copper Rivers _ 34 

5. Presently known breeding and wintering range, trumpeter 
swan — 39 

6. Map of Red Rock Lakes Migratory Waterfowl Refuge 40 

7. The Red Rock Lakes owe their stable waters to the Centennial 
Mountains which tower above them to the south, trapoing 
abundant snows that feed the numerous creeks and springs 
entering the marsh system 41 

8. Aerial view of Red Rock Lakes. Lower Lake in foreground 
is dotted with beds of bulrush. The marsh system is at the 
upper left, and the Upper Lake is in the center background — 42 

9. Trumpeter swan nest located on an old muskrat house on a 
cattail-sedge island in Lower Red Rock Lake. In the back- 
ground the Centennial Mountains wear a snow mantle normal 

for June - 43 

10. Numerous channels, sloughs, and potholes set in a bog-mat 
environment of beaked sedge typify the Red Rock Lakes 
marsh. Darker shoreline vegetation is bulrush, cattail and 
rushes - 44 

11. A trumpeter nest located in the predominant sedge environment 
of the Red Rock Lakes marsh. Stem and leaf parts of both 
sedge and cattail form the bulk of the nest material both for 

the muskrat lodge nest foundation and the nest proper 45 

12. Swan Lake is a shallow marsh sealed off from Upper Red Rock 
Lake by a natural sedge-willow anchored dike. Islands and 
peninsulas are chiefly sedge bog-mat while many extensive 

beds of spikerush also occur in these stable shallow waters — 46 

13. A trumpeter pen on her nest in a shoreline stand of pure sedge, 
Upper Red Rock Lake. The Centennial Mountains escarp- 
ment forms a chilly backdrop _ 49 

14. A female trumpeter, on her nest after returning from a feeding 
period in the Lower Lake, shakes the water from her plumage. 
The nest is located on a muskrat house behind a protective screen 
of bulrush. Note elevation of nest in tall, dense cover, 
and discoloration of swan's head and neck from contact with 
ferrous organic matter 51 

vn 



VIII CONTENTS 

Page 

15. Geographical features of the trumpeter swan breeding and 
wintering areas in the United States 52 

16. Trumpeter swan nesting site at Grebe Lake in Yellowstone 
National Park. Note exposed situation of sedge sod nest. 
Beds of wokas appear in the background, molted feathers in 

the foreground 53 

17. Aerial view of trumpeter nesting habitat in lower Tasnuna 
River Basin, Alaska. Trumpeter nest was found in small 
restricted slough in the lower right hand corner of photo 54 

18. Aerial view of Henrys Fork (North Fork) of the Snake River 
below the Railroad Ranch, Island Park, Idaho. This stretch of 
river offers habitat to wintering trumpeters. The warm Harri- 

man Springs keep these waters open even below — 30° 57 

19. Aerial view of trumpeters wintering on Henrys Fork of the 
Snake River below the Railroad Ranch. Gray birds are cyg- 
nets-of-the-year. Trumpeters from the Grande Prairie, 
Alberta, region have wintered in this area 58 

20. Aerial view of 80 trumpeters in east Culver Spring, Red Rock 
Lakes Refuge, January 1956. Air temperature —20° F. 
Note moose tracks in willow growth 59 

21. Trachea and sternum of whistling swan 66 

22. Trachea and sternum of trumpeter swan 67 

23. A pair of trumpeters on Grebe Lake, Yellowstone National 
Park, with 3 cygnets of the normal gray color phase and 2 
cygnets of the uncommon white phase 71 

24. Six trumpeters circle the open water at Culver Pond on the 
Red Rock Lakes Refuge in southwestern Montana. Seldom 
more than 6 or 8 of these birds fly together in local flights 
unless a large flock flushes together 73 

25. A pair of trumpeters show the 2 ways of carrying "landing 
gear" during flight. The normal method is by folding them 
back under the tail, but in sub-zero weather cold feet may be 
tucked up forward and be quite invisible in the warm feathers 

and down 75 

26. The normal and "exerting" neck attitudes are shown by 2 
trumpeters. Most commonly bent in this peculiar attitude 
during take-off, the neck straightens out in full flight 75 

27. Six trumpeters landing "flaps down" on Culver Pond, Red 
Rock Lakes Refuge. The feet are thrown forward before 
the moment of impact to ski the bird to a stop. Note various 

web positions guiding birds into landing 77 

28. Trumpeters on Culver Pond display the 2 methods of plumage- 
shaking, 1 employing the wings. The elevated position neces- 
sary for either position is attained by rapidly treading the 
water 84 

29. A small flock of trumpeters feeding with goldeneyes at Culver 
Pond, Red Rock Lakes Refuge. Typical drinking attitude is 
shown by swan with outstretched neck, right of center^ 85 

30. A typical mutual display of 2 wild trumpeters on wintering 
waters, Red Rock Lakes Refuge 87 

31. Trumpeters face each other in mutual display (right fore- 
ground) and a swan indulges in a wing-flapping plumage shake 
(center background while an adult bald eagle watches from a 
background snowbank in March on Red Rock Lakes Refuge 88 



CONTENTS IX 

Page 

32. Two trumpeters landing on Culver Pond are greeted by dis- 
playing swans. Band on left leg of lower bird marks it as a 
pen (female). Both Barrow's and common goldeneye ducks 

are present in this scene 90 

33. "Solo" display, trumpeter swan on Culver Pond, Red Rock 
Lakes Refuge 92 

34. Rear view of "solo" display (left) 93 

35. While a pair of trumpeters engages in mutual display (left 
center), 4 gather in a group exhibit (right). Mallards and 
pintails in the irregular foreground 93 

36. Aggressive action of trumpeter following group display, Culver 
Pond wintering waters 95 

37. Aggressive pursuit terminating a group display of 5 trumpeters, 

Red Rock Lakes Refuge 97 

38. Observations on the territorial traits of trumpeters can be made 
by a single observer over several thousand acres, since the high 
mountains provide ideal vantage points. Here the birds on 
Lower Red Rock Lake are studied from an observation post on 

the northern flank of Centennial Mountains 102 

39. Nest locations, 1954-57, Upper Red Rock Lake and Swan Lake 
marsh. (Approximately 8,000 acres in map) 108 

40. Nest locations, 1954-57, Lower Red Rock Lake. (Approxi- 
mately 3,000 acres in map) 109 

41. Newly-hatched trumpeter cygnet, Lower Red Rock Lake 118 

42. Close-up of trumpeter cygnet showing fine, grayish-white 
down .. 119 

43. Swan family at loafing site, Grebe Lake, Yellowstone National 
Park. The special foot position of the adult is commonly 
seen with mute swans. This brood is approximately a month 

old 121 

44. Trumpeter swan productivity rates, total population, 1931-57. 149 

45. Nonbreeding trumpeter swans censused at Upper Red Rock 
Lake and Lima Reservoir, 1940-57 151 

46. Trumpeter swan census, Red Rock Lakes Refuge, 1932-57 152 

47. Trumpeter swan productivity rates, Red Rock Lakes Refuge, 
1936-57 153 

48. Trumpeter swan census, Yellowstone Park, 1931-57 156 

49. Trumpeter swan productivity rates, Yellowstone Park, 1931- 

57 157 

50. Trumpeter swan census outside Red Rock Lakes Refuge and 
Yellowstone Park, 1931-57 160 

51. Trumpeter swan cygnets captured on the Red Rock Lakes by 

the Wetmore family for the live swan trade about 1900 171 

52. The U. S. Fish and Wildlife Service "snowplane" en route to the 
Culver Pond swan wintering grounds for semi-weekly feeding 
of small grain. The Centennial Mountains in background 

form the Continental Divide along their 10,000-foot crest 179 

53. These trumpeters are only part of the flock of over 200 which 
pass the late winter months on the Refuge awaiting the Spring 
break-up. The 94 trumpeters visible in this single photograph 

are more than existed in the entire United States 25 years ago_ . 184 

54. Trumpeters feed in Culver Pond with mallards, Barrows 
goldeneyes, and common goldeneyes on grain placed out by 
Refuge personnel 187 



TABLES 

Page 

1 . — -Classification of the subfamily Cygninae 6 

2. — Trumpeter swan breeding records in the United States and 

Alaska to 1925 24 

3. — Swans censused, Alaskan waterfowl inventory, January, 1949 

to 1957 32 

4. — Winter swan counts, Greater Yellowstone region, 1950 to 

1957 61 

5. — Overlapping weights and dimensions of small trumpeter and 

large whistling swans 64 

6. — Cygnet mortality at Red Rock Lakes Refuge, 1949 133 

7. — Swan census data, 1931 to 1957 146 

8. — Nonbreeding trumpeter swan populations at Upper Red Rock 
Lake, Lima Reservoir, and other important areas, 1940 to 

1957 147 

9. — Trumpeter swan production data, 1931 to 1957 148 

10. — Trumpeter swan production data, Red Rock Lakes Refuge, 

1936 to 1957 1 52 

11. — -Variations in Refuge swan nesting density 154 

12. — Trumpeter swan production data, Yellowstone Park, 1931 

to 1957 . 158 

13. — Characteristics of some Yellowstone Park lakes and their 

record of use by swans, 1931 to 1957 159 

14. — -Trumpeter swans transferred from the Red Rock Lakes 

Refuge, 1938 to 1957 180 

15. — Trumpeter swan nesting data, National Elk Refuge, Wyo- 
ming, 1944 to 1957 180 

16.— Swans banded at the Red Rock Lakes Refuge, 1945 to 1957- 182 



ILLUSTRATION CREDITS 

Credits for illustrations are as follows : Frontispiece and figures 7-14, 
18-20, 24-37, 42, 4r>-47. Winston E. Banko ; figures 2 and 44, Cecil, Wet- 
more ; figures 4 and 17, Mel Monson ; figure 1(5, David de Lancey Con- 
don; figures 23 and 43, W. Verde Watson: figure 38, U. S. Fish and 
Wildlife Service ; drawings in figures 21 and 22, Mrs. P. W. Parmalee ; 
maps and other illustrations, Shirley A. Briggs. 




1 , lfl Jfc«ll**^ 10 






FOREWORD 

The appeal of swans to man throughout history has come down to 
us in legend, custom and in many forms of art. That the world's 
largest species of this storied bird should have become nearly extinct 
in its native North America was thus especially tragic. The con- 
tinuing recovery in numbers of this beautiful and graceful symbol 
of American wilderness is a major accomplishment in wildlife 
preservation. 

The establishment of the Red Rock Lakes Migratory Waterfowl 
Refuge was the climax of this effort in the United States, and any 



Z FOREWORD 

study of the trumpeter swan necessarily focuses on the Refuge and 
the adjoining country. In this magnificent mountainous setting a few 
of the swans had survived, and the remoteness of the country has 
made it possible to maintain the wilderness environment favored by 
the birds. 

The present United States population of trumpeters is found 
mainly in a 60-mile radius encompassing parts of southwestern Mon- 
tana where the Refuge is found, eastern Idaho, and northwestern 
Wyoming, including Yellowstone National Park. 

This report on the trumpeter evolved from studies made from 1948 
to 1957 when I served first as an assistant and later as manager, of 
the Red Rock Lakes Refuge. I have also drawn extensively on the 
records of the National Park Service, the U. S. Fish and Wildlife 
Service, and the National Museum in Washington, D. C. Other perti- 
nent information bearing on the life history of the trumpeter swan in 
the United States has been extracted from published articles, unpub- 
lished reports and records, firsthand accounts and correspondence. 
This account includes a historical record of this bird in the United 
States and Alaska, an outline of its habits and characteristics in its 
native Rocky Mountain environment, and furnishes information nec- 
essary to guide its future. 

A comparable study of the trumpeter by the Canadian Wildlife 
Service has also been underway since about 1950. These investiga- 
tions are being conducted chiefly in British Columbia and Alberta 
by Mr. Ronald H. Mackay, Wildlife Biologist, who made available 
some preliminary results of swan banding studies in Alberta, and was 
helpful in many other ways. For the most part, I have dealt with 
the trumpeter in Canada only in a general way since the Canadian 
findings will probably be published later. 

I wish to express my appreciation for the cooperation received over 
the years from the many individuals and agencies who furnished in- 
formation useful in preparing this report. Generally, acknowledg- 
ment for this has been handled directly in the text, but special thanks 
and credit are due to several individuals. 

Edmund B. Rogers, former Superintendent, and David de Lancey 
Condon, Chief Naturalist, Yellowstone National Park, generously 
opened the Park files to me and furnished every help possible. Con- 
don and Walter H. Kittams, Park Biologist, reviewed the manuscript 
in its final stages. 

James Rooney of Yakima, Washington, supplied valuable biblio- 
graphic assistance. Lowell Adams, U. S. Fish and Wildlife Service, 
also gave an early review and critique of the MS. 

The information regarding the geological background of Red Rock 
Lakes, Montana, was supplied by Dr. George Kennedy, Geophysicist, 



FOREWORD 



University of California, who has an intimate knowledge of this 
region. 

Dr. Herbert K. Friedmann, U. S. National Museum, provided the 
technical physical descriptions of both species of swans native to 
North America from an unpublished manuscript. 

Joseph Flakne, Programming Director, and Marie Tremaine, Chief 
Bibliographer, Arctic Institute of North America, provided for an 
inquiry into the possible role of the swans in the economic life of 
Alaska before acquisition by the United States, and made available 
interesting and scattered notes regarding Old World swans in arctic 
Russia. 

H. Albert Hochbaum, Director, and Dr. Frank McKinney, Assistant 
Director, Delta Waterfowl Research Station, reviewed the manuscript 
and were generous with helpful comments and suggestions. Dr. 
McKinney, and indirectly, Paul Johnsgard of Cornell University, 
supplied me with the English translation (Johnsgard) of O. Hein- 
roth's (1911) classic German work on the ethology and psychology of 
the Anatidae. 

I am grateful to several residents of the Red Rock Lakes region 
who contributed to the historical aspect of this work. Credit and 
thanks are due the following residents of Monida, Montana: Cecil 
Wetmore, James F. and Alta Hanson, and A. Blaine Fordyce. Sam 
A. Trude of Island Park, Idaho, furnished information on that area in 
the early 1900's. 

In the Branch of Wildlife Research, Earl L. Atwood, Chief Bio- 
metrician, Patuxent Research Refuge, assisted in determining the 
significance of the population statistics. 

In the Branch of Wildlife Refuges, special recognition is due Rich- 
ard E. Griffith for his particular interest in the manuscript work 
which extended over a period of many years. In a large measure his 
patient and unflagging support made this report possible. The final 
draft was typed under the direction of Miss Winifred G. Baum, who 
deserves mention for positive action when time counted most. I am 
also greatly indebted to Dr. Ray Erickson and Miss Shirley A. Briggs 
for their tireless efforts in checking and editing the final draft, work 
made especially difficult under the prevailing deadline. 

It is impossible to mention everyone who has been helpful in pre- 
paring this report. 

Winston E. Banko. 

May 1, 1958. 




INTRODUCTION 



LEGEND AND TRADITION 

Of all the earth's varied avian forms, the swans have been woven 
into the cultural expressions of previous civilizations to a greater 
extent than any other group of birds, with the exception perhaps of 
the birds of prey. With their great size, migratory habits, graceful 
manner, and distinctive voice, the majestic swans have apparently 
inspired all the peoples who knew them. 

No doubt our more primitive ancestors were most interested in swans 
because of their value for food. Although wild swans were appar- 
ently never preferred, their availability and size caused them to be 
taken when other sources of meat were short. As man gradually 
developed a regard for spiritual values, swans were employed as an 
important symbolic element in the myths and religious ceremonies 
of many of the early cultures. This was true not only among the 
early peoples of northern Slavic or Nordic origin, but also in the 
regions along the Mediterranean where the long migrations of some of 
the northern-reared swans terminated. 

Although there is frequent mention of swans in Greek mythology, 
they were apparently not included among the many varieties of birds 
and animals commonly kept by the Romans. Swans were known to 
the ancient Egyptians and to the early Christian prophets. In early 
canons, the latter listed swans among the birds and animals which 
were not to be eaten. 

The use of swans as common subjects in story, myth, and ceremony 
was most prominent in the culture of the more northern races of 
mankind, where these birds entered into the ceremonies of the shamans 
of the East, the wizard men of Lapland, and the medicine men of our 



DESCRIPTION AND SYSTEMATICS 

own native Indians (Beebe, 1906: 159). The notable interest of the 
ancients in these distinctive fowl grew with the passing of centuries. 
This later expressed itself in the widely prevalent Eurasian "swan- 
maiden" legend, and more tangibly in such items as some of the early 
coins of Germany and the badge of Henry IV of England. 

Many of the legends which slowly evolved out of early European 
swan mythology are perpetuated in the fairy tales, customs, and tradi- 
tions of certain countries, being now for the most part only relics of a 
forgotten age. "Swan upping" (the taking up of swans for the pur- 
pose of pinioning and marking) is still practiced annually upon the 
Thames Eiver. Records of the English interest in swans date back 
over 800 years to the 12th century A. D. during which the complica- 
tions of swan ownership resulted in the enactment of special laws and 
regulations. After 40 years of study, Norman F. Ticehurst (1957) 
has thoroughly documented the long and interesting history of the 
mute swans in England. 

Swans have been accorded a special place in the folklore, history, 
literature, drama, arts, and musical expressions of contemporary 
peoples as well. In fact, it is difficult to name a medium of man's 
expression that does not owe a modicum of debt to these birds. In 
this respect the swans as a group are unique in the bird world. 

It is evident therefore that much of the present interest in these 
noble birds stems in part from the traditions of many previous civili- 
zations and peoples, extending far back into history. With this 
heritage in mind, we have a responsibility for the welfare of this 
group of birds, and must perpetuate such living symbols of beauty 
and grace for the enjoyment and inspiration of generations yet to 
follow. 

DESCRIPTION AND SYSTEMATICS 

The swans are similarly specialized waterfowl of the diverse and 
prolific family of the Anatidae, and as such they have been given 
status as a subfamily, the Cygninae. This group of waterfowl is 
characterized by necks as long as, or longer than, the large heavy 
body, and short strong legs and feet equipped with large webs and 
prominent nails. This combination of characteristics adapts them 
well for a specialized shallow-water existence, in which they consume 
large quantities of leafy aquatic plants, and dig and root out suc- 
culent rootstocks and tubers. 

With the exception of the black swan of Australia (Chenopis 
atratus) and the black-necked swan of South America (Cygnus 
melancon iphus) , the adult plumage of all species of swans is entirely 
white. With the further exception of the common mute swan of 
Europe (Cygnus olor), which possesses a prominent knob at the base 



G 



INTRODUCTION 



of the upper mandible in common with the black-necked swan, the 
remaining four species are all smooth-billed (that is without a basal 
knob), white-.plumaged, distinctive-voiced swans with an exclusively 
circumpolar distribution. A general account of the characteristics, 
distribution, and habits of all species of swans, with emphasis on their 
traits and requirements in captivity is given by Jean Delacour (1954: 
57-90). In this work, the plumage and appearance of both adults 
and cygnets have been accurately portrayed in full color by Peter 
Scott. 

The division of the swans of the world into 3 genera, 2 subgenera, 
and 7 species is generally accepted as most nearly logical and correct 
(Wetmore, 1951: 338). The systematics of the swans will not be 
treated here. (See appendix 1.) 

Using Wetmore's classification of the swans, which has been 
adopted by the American Ornithologists' Union, the classification 
and current principal distribution of the swans is outlined in table 1. 
The status, distribution, migration and habits of mute, Bewick's, and 
whooping swan populations in Eurasia are treated by Alfred Hil- 
precht (1956: 8-17), Witherby et al. (1939: 168-179), and G. P. 
Dementiev and N. A. Gladkov (1952) . 

Table 1. — Classification of the subfamily Cygninae 



Scientific name 


Common name 


Principal distribution 




Breeding 


Wintering 


Olor cygnus (Linnaeus) 

Olor bewickii Yarrell__ 

Olor columhianus (Ord) 

Olor buccinator (Richardson). 


Whooper Swan 

Bewick's Swan 

Whistling Swan 

Trumpeter Swan 


Northern and middle 

Eurasia. 
Northern Eurasia 

Northern North Amer- 
ica. 

Northern and middle 
North America. 

Northern and Middle 
Eurasia. 

Southern South Amer- 
ica. 

Australia and New Zea- 
land. 


Southern and middle 

Eurasia. 
Southern and middle 

Eurasia. 
Southern and middle 

North America. 
Middle North America. 

Southern and middle 


Cvgnus melancoriphus (Mol- 
ina). 


Black-necked Swan _ . . 


Eurasia. 
Southern South America. 

Australia and New Zea- 






land. 



The significant specific differences among the swans in the genus 
Olor are primarily of an anatomical nature, principally in variations 
of the tracheal route through the furculum and along the sternum. 
The variations of the trachea are also responsible for voice differences 
which, with other external differences, are of definite value to the 
field worker in making positive identifications wherever the ranges 
of the various species overlap, as they do to a limited extent over the 
circumpolar range of this genus. 

Since wild populations of the Bewick's swan (Olor beioickii) have 
never been observed in North American waters, and the whooper 



DESCRIPTION AND SYSTEMATICS 7 

swan (Olor cygnus), though formerly breeding in Greenland, has 
only three recent Continental records (St. Paul Island: Wilke, 1944: 
655; Karl Kenyon, 1949 (letter) ; Amchitka: Kenyon, 1957 (letter)), 
the American field ornithologist need be concerned mainly with dis- 
tinguishing the trumpeter (Olor buccinator) from the whistling swan 
(Olor coluinbianufi) . Although field identification of trumpeters vs. 
whistlers may be fairly certain in those cases where the bird in ques- 
tion gives voice, or where a distinct yellow spot is visible on the lores, 
it may be difficult or impossible to determine the species with certainty 
without a postmortem examination if these characteristics are not 
evident. (The subtle superficial differences between these two species 
will be treated later under the topic Life Cycle.) 

In considering the circumpolar distribution of the genus Olor, 
another point should be mentioned here. The two largest species of 
this group, the trumpeter and the whooper, according to the author- 
ities (Delacour, 1954: 72-73, 84-85; Witherby, et al., 1939: 171, 174) 
range during their breeding season principally over the interior of the 
continents with which they are associated, while the breeding ranges 
of the two smaller species, whistling and Bewick's swans, are primarily 
along the continental fringes and the islands of the Arctic Ocean. 

The numerical status of the trumpeters in North America is only 
partially understood at this time owing to the difficulty of censusing 
completely the wintering trumpeters in Alaska and British Columbia. 
The best recent estimates of this rather obscure population place it 
between 600 and 1,000 birds (Munro, 1949 : 710), while the 1957 trum- 
peter-swan census in the United States found 488 individuals in this 
country. Thus, the total continental population of trumpeters prob- 
ably numbers 1,500 or more. 

The status of whistling swans in North America is much different. 
The midwinter inventory by the U. S. Fish and Wildlife Service in 
January 1958 found 78,425. Wintering populations fluctuate and 
are about equally divided between the Pacific and Atlantic flyways 
(Stewart and Manning, 1958 : 205-207) . 



469660 O — 60- 




DISTRIBUTION AND STATUS 



GEOLOGICAL OCCURRENCE 

At one time or another in the distant past, before man appeared 
on the North American Continent, trumpeter swans must have 
occurred commonly within nearly every region of what is now the 
United States. The advance and retreat of a succession of ice ages 
in the northern hemisphere determined the distribution and status 
of this species as it did those of the other faunal elements. The 
climatic changes may also have been responsible for the passing of 
one North American species of swan from the scene. This extinct 
species has been tentatively designated Cygnus paloregonus, the 
remains of which were discovered at Fossil Lake, Oregon. (In 
addition one Eurasian species has also become extinct at some time 
in the distant past. This is the giant swan of Malta, Palaeocycnus, 
which was larger than any of the swans in existence today.) 

Remains of the progenitors of both trumpeters and whistlers have 
been identified from widely separated geologic formations in the 
United States. Alexander Wetmore (1956: 25) lists trumpeter occur- 
rences in such deposits as follows, "Modern form reported from 
Pleistocene: Aurora, Illinois; Itchtucknee River, Florida. Late 
Pleistocene : Fossil Lake, Oregon." In Illinois the trumpeter remains 
were associated with bones of the giant beaver and mastodon 
(Wetmore, 1935: 237), while in Florida remains of the trumpeter 
were found in Pleistocene material together with the bones of such 
birds as the California condor (Gymn.ogyps calif 'ornianus) , whooping 
crane {Grus americana), and jabiru stork {Jabiru mycteria) (Wet- 
more, 1931:19), all of which are of course now unknown in that 
whole region. An unusual Pleistocenic associate of the trumpeter 

8 



PRIMITIVE HISTORY 9 

identified from Fossil Lake, Oregon, deposits was a flamingo 
(Phoenicopterus copei). Dr. Herbert Friedmann (1935 : 23) has also 
recorded the presence of trumpeter bones from Kodiak Island, Alaska. 
So through the centuries the ancestors of the trumpeters existed under 
far different circumstances and in regions which today might no longer 
be considered suitable. 

PRIMITIVE HISTORY 

Long before Caucasian man made his appearance in North America, 
swans were used in various ways by many of the indigenous Indian 
tribes. The swans were undoubtedly sought principally for food, 
and today their remains are occasionally exhumed in archeological 
excavations. The bones of the trumpeter can be specifically identified 
in many such instances. 

H. K. Coale (1915: 89), in his valuable early treatise on the status 
of this species, quotes a reliable source from Ohio as stating: 

We have in our collection a great many bones of the trumpeter swan. It 
seems that this bird, although a very rare migrant at the present time, was 
here in great numbers in pre-historic time, and we find their bones in the villages 
of the old Indians, who always used the leg bone for making implements, while 
the wing bones were seldom used. I found specimens in the Baum, Bartner, 
and Madisonville village sites. 

In bibliographical material furnished by E. S. Thomas, Curator 
of Natural History of the Ohio Historical Society, various other 
authors have also reported unearthing bones of the trumpeter swan 
among kitchen-midden material from at least four ancient Indian 
village sites in that State. The remains of the trumpeters found in 
these excavations varied greatly in age, from early historic, in the 
case of the Fairport Harbor Village site, to from 2,377 to 2,750 years 
ago for the Kettle Hill Cave and Toepfner Mound sites, as estimated 
by Thomas using carbon-14 datings. 

Trumpeter bones have also been exhumed in ancient kitchen-midden 
material in Illinois. P. TV. Parmalee, Curator of Zoology at Illinois 
State Museum, wrote (correspondence) that findings in six sites cov- 
ered a time range of at least 1,500 years. Most of the swans seem to 
have been used for food, though some bones were cut into beads. At 
the Cahokia village site, near East St. Louis, about 375 trumpeter 
bones have been identified. This village is thought to have been 
vacated just before the coming of the white man. (See also Parmalee, 
1958, in Bibliography.) 

Swans also entered the lives of the early peoples by contributing to 
their dress, ceremony, and legend. The journals of a number of early 
American explorers and travelers, gathered and edited by R. G. 



10 DISTRIBUTION AND STATUS 

Thwaites (1906), contain many firsthand references to the roles 
played by these great white birds in the lives of the Indians. Like 
the plumage of the eagle, feathers of the swans were valued for their 
decorative and symbolic value. 

Although little pertinent life-history information can be learned 
from the accounts of our native swans during pre-Caucasian times, 
we do find that the swans were present and taken frequently enough 
to have entered the lives of the natives as a recognizable part of their 
culture. 

EARLY HISTORICAL NOTES (1632-1832) 

Early accounts of our native swans, and of trumpeters in particu- 
lar, are brief and scattered in the literature over a long period of time. 
A New Englander, Thomas Morton, wrote of the native swans in 
1632 (Force's Historical Tracts, vol. 2 : 46) : 

And first of the Swanne, because she is the biggest of all the fowles of that 
Country. There are of them in Merrimack River, and other parts of the Country, 
greate Store at the seasons of the yeare. The flesh is not much desired of the 
inhabitants, but the skinnes may be accompted a commodity, fitt for divers uses, 
both for fethers, and quiles. 

Although Morton gives no clue to the species identity of the swans 
which seasonally visited that part of New England in the early days, 
later accounts by other observers indicate that both trumpeters and 
whistlers probably were represented. Jeremy Belknap (1784) listed 
a New Hampshire swan with a "sound resembling that of a trumpet'', 
C. Hart Merriam (1877) thought that both were in Connecticut in 
early times, with one trumpeter reported in his day, and J. A. Allen 
(1878) stated that the trumpeter doubtless was common in Massa- 
chusetts 200 years earlier, and "may still be looked for as a straggler." 

The next early report appears to have been in a history written 
by John Lawson, Surveyor- General of North Carolina, and first 
published in 1709. It is the first to separate the trumpeter as a bird 
distinct from the lesser species and positively record its occurrence 
on the eastern seaboard. This record is as follows (Lawson, 1714: 
86) : 

Of the Swans we have two sorts : the one we call trompeters because of a 
sort of Trompeting Noise they make. These are the largest sort we have ; which 
come in great Flocks in the Winter, and stay, commonly in the fresh Rivers, 
until February, when the Spring comes on, when they go to the Lakes to breed. 
A Cygnet, that is a last year's Swan, is accounted a delicate dish, as indeed it 
is. They are known by their Head and Feathers, which are not so white as 
Old ones. 

Lawson's account is pertinent for several other reasons. First, 
we learn that the early settlers were familiar enough with Olor hue- 






EARLY HISTORICAL NOTES (1632-1832) 11 

cinator to give it the common name of "trumpeter"; second, the 
presence of this species in "large flocks" on the "fresh Rivers" gives 
us the first and only clues to its original status and winter habitat 
along the east coast; and third, the positive remark calling attention 
to the time of departure (February) when "they go to the Lakes to 
breed" suggests the nesting of this species somewhere to the east of 
the accepted eastern limits of the breeding range of this species which 
was documented later. Lawson's specific use of the term "Lakes" is 
especially interesting, inasmuch as he does not hint of the breeding 
grounds of the whistling swan in his notes on the lesser species, and 
trumpeters are indeed wholly pond or lake breeders, never known to 
nest along the banks of rivers. The trumpeter is not mentioned again 
in United States ornithological literature for another century, and so 
the possible breeding status of this species east of the Ohio River 
before settlement by the white man remains obscure. 

More than half a century was to elapse before the next perti- 
nent record of swans was left by Samuel Hearne, an employee 
of the Hudson's Bay Company. Hearne's diary contains several brief 
remarks on the swans made on his epic journey from Hudson Bay 
to the Arctic Ocean during the period 1769-72. One of his state- 
ments (Hearne, 1795: 371) not only initially documents the entry 
of the plumage of the swans into the world of commerce but records 
that great numbers of swans were taken for food by the Hudson 
Bay Indians. He states : 

In fact, the skinning of a Bear spoils the meat thereof, as much as it would 
do to skin a young porker or roasting pig. The same may be said of swans 
(the skins of which the Company have lately made an article of trade) ; other- 
wise thousands of their skins might be brought to market annually, by the 
Indians that trade with the Hudson's Bay Company's servants at the different 
settlements about the Bay. 

In the accounts which follow, the plumage of these great birds, first 
valued by the earliest colonists, gradually became and remained an 
article of frontier commerce for over a hundred years, eventually 
reaching the London fur market by the thousands of skins. This fact, 
perhaps more than any other now apparent, caused the gradual reduc- 
tion of numbers and range of both the North American swans, and 
particularly the near extinction of the trumpeter. 

Lewis and Clark appear to be the next explorers who mention 
the swans to any extent. Observations of these birds were made 
several times during the course of their transcontinental journey dur- 
ing the period 1804-6. The following brief note is found in Elliot 
Coues' edited account of this expedition (1893: 1284). The notation 
was made in northwestern Missouri on July 4, 1804, during the begin- 
ning of their ascent of the Missouri River and while nearly opposite 



12 DISTRIBUTION AND STATUS 

the present town of Atchison, Kansas, and records, "A great number 
of young; swans and geese on a lake opposite Fourth of July Creek." 
This strengthens the brief note by Widmann (1907) and Blines 
(1888) that swans once bred in north Missouri. In the Lewis and 
Clark account (Cones, 1893: 74.3-915) the two North American swans 
are correctly separated on the basis of size and voice for the second 
time. The first of these observations was penned while Lewis and 
Clark were in winter quarters at Fort Clatsop near the mouth of the 
Columbia River. This note suggests for the first time the name "whis- 
tling swan" for the lesser swans observed. The last paragraph was 
written on March 28-29, 1806, while the expedition was ascending the 
Columbia River on their return journey home. (In quotations from 
this source, interpolations in brackets are by Dr. Coues) : 

The birds which most strike our attention are the large [Ci/gntis buccinator], 
as well as the small, or whistling swan [C. columManiis], 

******* 

The small differs only [mainly] from the large in size and note ; it is about 
one-fourth less, and its note is entirely different. It cannot be justly imitated 
by the sound of letters; it begins with a kind of whistling sound, and terminates 
in a round full note, louder at the end ; this note is [not] as loud as that of the 
large species ; whence it [this small swan] might be denominated the whistling 
swan ; its habits, color, and contour appear to be precisely those of the larger 
species. These birds were first found below the great narrows of the Columbia, 
near the Chilluckittequaw nation ; that were very abundant in this neighborhood, 
and remained with the party all winter; in number they exceeded those of the 
larger species in the proportion of five to one. 

******* 

Deer Island is surrounded by an abundant growth of cottonwood, ash, and 
willow, while the interior consists chiefly of prairies interspersed with ponds. 
These afford refuge to great numbers of geese, ducks, large swan [Cygnus 
buccinator], ... In the course of the day we saw great numbers of geese, 
ducks, and large and small swans [Cygnus buccinator and C. cohimMamis], which 
last are very abundant in the ponds where the wappatoo grows, as they feed 
much on that root. 

Hans Pilder (1914 : 170) furnished some information on the trade in 
swan skins by the Hudson's Bay Company and the Canadian Company 
during the years 1806-22. According to his data the "numbers of 
swan skins which were exported by these companies were as follows: 
1806, 396 skins (Hudson's Bay Company) : 1807, 1,192 (Hudson's 
Bay Company); 1818, 2,463 (Hudson's Bay Company) plus 600 
(Canadian Company): 1820, 800 (Canadian Company); 1822, 1,800 
(Hudson's Bay Company). Since the companies merged in 1822, 
the last figure is apparently that of the combined export. 

Apparently the first record of trumpeters breeding in the United 
States is found in Dr. T. S. Roberts' (1936: 2*05) account of a journal 
entry by Count G. E. Beltrami on July 13, 1S23. Count Beltrami 



EARLY HISTORICAL NOTES (163 2-1832) 13 

accompanied Major Stephen H. Long's expedition into the Minnesota 
and Red River Valleys, encamping at that date near what is now 
called Swan Lake (Nicollet County) where Beltrami noted, "In the 
evening we halted near a little wood which lies along the banks of the 
Lake of Swans. It was the season at which these beautiful birds 
cannot fly — the old ones, because they are changing their feathers; 
the young, because they have yet only a soft down." 

William Keating (1825, vol. 1: 446), geologist with the Long ex- 
pedition, recorded that at the Lake Traverse fur post, on the border 
of Minnesota and South Dakota, 2 packs of 60 swan skins were worth 
120 Spanish dollars. This is the first reference to swans being taken 
in the United States for commercial purposes. 

Uses made of these swan skins are not itemized by these early 
writers. Delacour (1954: 76) says they were used for the manufac- 
ture of powder puffs. E. H. Forbush ( 1929 : 306 ) says that "the trade 
in swansdown offered further incentive for the destruction of the 
species." The feathers were certainly used for adornment in many 
ways, and the quills made excellent pens. John James Audubon, 
America's noted early ornithologist and artist, preferred trumpeter 
quills for drawing fine detail, as in the feet and claws of small birds, 
saying (1838 : 538) that they were "so hard, and yet so elastic, that the 
best steel pen of the present day might have blushed, if it could, to be 
compared with them." 

During the late 1820's the traffic in swan skins apparently increased. 
C. P. Wilson, editor of the Hudson's Bay Company publication, The 
Beaver, furnished additional notes regarding that Company's trade in 
swan skins. He wrote (correspondence) : 

In regard to the old sale lists . . . 5,072 skins were sold in London on 16th 
April, 1828, and on the following 10th. December 347,298 goose, swan and eagle 
quills and wings were sold. On the 29th. October that year the Company im- 
ported 4,263 swan skins from York Factory and Mackenzie River districts ; 18 
from Moose River and East Main in the southern part of James Bay ; and 26 
from the Columbia region, but no distinction is made between Trumpeters and 
Whistlers. 

In 1828, Audubon set down a significant account of an Indian swan 
hunt. These notes record for the second time the taking of swans 
specifically for their plumage in the United States proper. 1 All other 
instances of this sort have a Canadian origin. Audubon's account 
(McDermott 1942 : 154) describes the deliberate slaughter of "at least 



1 The calendar of the American Fur Company's papers of 1834-47 (Nute. 1945) gives no 
information. The records of the North-West Company, the only other big fur company 
in North America exclusive of the Hudson's Bay Company, were either amalgamated with 
those of the Hudson's Bay Company in the merger of the two concerns in 1821 or have 
been lost. Thus, there seems little likelihood that further information on this subject 
will ever come to light. 



14 DISTRIBUTION AND STATUS 

50" swans by Indians near the confluence of the Mississippi and Ohio 
Rivers (in Kentucky), the skins of which were "all intended for the 
ladies of Europe." 

A year or so after the noted English ornithologist William Yarrell 
had demonstrated a systematically reliable difference between the 
anatomy of the whooper swan and that of its smaller relative, the Be- 
wick's swan, Sir John Richardson was successful in discovering 
similar constant differences between the two closely related North 
American species (Swainson and Richardson, 1832, vol. 32: 438, 464). 
Although these two species had previously been separated on the basis 
of both size and note, and indeed the common names of trumpeter and 
whistling swan were already in use among the ornithologists of that 
day, it remained for Richardson to describe a positive method of 
identifying these two closely related species which would invariably 
serve when more superficial characteristics were either absent or in 
doubt. The differing point of anatomy discovered by Richardson, the 
form and route of the trachea through the sternum, is the only reliable 
characteristic allowing positive speciation today. 

Richardson's notes are also helpful in outlining the former range 
and distribution of this species and in stressing its importance in the 
fur trade. Richardson contributed : 

This is the most common Swan in the interior of the fur-eounties. It breeds 
as far north as latitude 61°, but principally within the Arctic Circle. ... It 
is to the trumpeter that the bulk of the Swan-skins imported by the Hudson's 
Bay Company belong. 

Elsewhere in his treatise on the northern zoology of the Continent, 
Richardson noted that the trumpeter was established across the Con- 
tinent and north to a latitude of 68°, breeding "in the interior between 
the sixtieth and sixty-eighth parallels." 

Richardson apparently was ignorant of Count Beltrami's account of 
swans breeding in Minnesota in 1823, and he apparently discounted 
Lawson's brief note that in North Carolina "they [trumpeters] go 
to the Lakes to breed" by omitting this remark in his 1831 description 
of this species, though he does mention some of Lawson's remarks 
regarding the trumpeter. We may conclude that Richardson believed 
the breeding range of this species to be confined mainly to the interior 
of Arctic Canada. 

LATER HISTORICAL NOTES (1833-1925) 

For two decades following Richardson's published description of 
the trumpeter and its range in North America, only a rather extensive 
account of the trumpeter by Audubon and a remark by Pierre Jean 
De Smet shed further light on the status of this species in the United 



LATER HISTORICAL NOTES (1833-1925) 15 

States. By this time the trumpeters of the eastern seaboard appear 
to have been exterminated, as Audubon (1838: 536-537) relates: 

the larger Swan, the subject of this article, is rarely if ever seen to the eastward 
of the mouths of the Mississippi. 

******* 

This species is unknown to my friend, the Rev. John Bachman, who, during a 
residence of twenty years in South Carolina, never saw or heard of one there ; 
whereas in hard winters the Cygnus Americanus is not uncommon, although 
it does not often proceed further southward than that State. 

Audubon (1838 : 537-538) does outline the occurrence of the trum- 
peter in the Ohio and Mississippi River valleys rather completely, 
furnishing at the same time a note on its abundance there, stating : 

The Trumpeter Swans make their appearance on the lower portions of the 
waters of the Ohio about the end of October. They throw themselves at once 
into the larger ponds or lakes at no great distance from the river, giving a 
marked preference to those which are closely surrounded by dense and tall 
canebrakes, and there remain until the water is closed by ice, when they are 
forced to proceed southward. During mild winters I have seen Swans of 
this species in the ponds about Henderson [Kentucky] until the beginning of 
March, but only a few individuals, which may have stayed there to recover from 
their wounds. "When the cold became intense, most of those which visited the 
Ohio would remove to the Mississippi, and proceed down that stream as the 
severity of the weather increased, or return if it diminished. ... I have 
traced the winter migrations of this species as far southward as the Texas, 
where it is abundant at times, ... At New Orleans . . . the Trumpeters are 
frequently exposed for sale in the markets, being procured on the ponds of the 
interior, and on the great lakes leading to the waters of the Gulf of Mexico. . . . 
The waters of the Arkansas and its tributaries are annually supplied with 
Trumpeter Swans, and the largest individual which I have examined was shot 
on a lake near the junction of that river with the Mississippi. It measured 
nearly ten feet in alar extent, and weighed above thirty-eight pounds. 

Whilst encamped in the Tawapatee Bottom when on a fur trading voyage, 
our keel boat was hauled close under the eastern shore of the Mississippi. . . . 
The great stream was itself so firmly frozen that we were daily in the habit 
of crossing it from shore to shore. No sooner did the gloom of night become 
discernible through the gray twilight, than the loud-sounding notes of hundreds 
of Trumpeters would burst on the ear ; and as I gazed over the icebound river, 
flocks after flocks would be seen coming from afar and in various directions, 
and alighting about the middle of the stream opposite to our encampment. 

Although Audubon apparently became familiar with migrating 
or wintering trumpeters during his widespread travels, and even kept 
a male in captivity for about 2 years when living at Henderson, Ken- 
tucky (1838 : 541), he never was privileged to see a nest or young of 
this species. 

A noteworthy breeding record is contained in the writings of Pierre 
Jean De Smet, who was a noted early Jesuit missionary of the Pacific 
Northwest. On April 15, 1842, De Smet was traveling with a band 
of Flathead Indian warriors and w T as encamped near Flathead Lake 



16 DISTRIBUTION AND STATUS 

in western Montana when he wrote, according to Thwaites (1906, vol. 
27:359) : 

The warriors had gone on ahead and dispersed in every direction, some to 
hunt and others to fish. . . . The warriors returned in the evening with a bear, 
goose, and six swan's eggs. 

Since western Montana, and especially the Flathead Valley, was 
later the source of many trumpeter breeding reports, this record is 
believed valid. De Smet, in 1845, reported seeing swans during 
the summer on the marshy lakes in southeastern British Columbia 
which form the source of the Columbia River (Thwaites, 1906, vol. 
29:206). This suggests that the adjacent region of the Kootenays 
at one time may have been included in the ancestral breeding range 
of this species. 

The brief notes on the native swans made by the naturalists who 
accompanied the various expeditions sent out by the Secretary of War 
during the period of 1853-55 furnish some pertinent data on the oc- 
currence of the trumpeter, chiefly in the Far West. Dr. George 
Suckley accompanied one of these surveys as naturalist westward 
from the Mississippi River to the Pacific coast, and though he re- 
cords the trumpeter near the beginning of his journey in Minnesota 
and at the end on the Columbia River, no mention is made of this 
species between these two localities. Dr. Suckley (1859: 248-249) 
recorded : 

It [trumpeter] is, like the preceding species [whistler], more abundant on the 
Columbia river than at Puget Sound. In the winter of 1853-54 I noticed im- 
mense flocks of swans, apparently of this species, collected along the shores 
of the river mentioned, and spread out along the margin of the water for a 
distance varying from an eighth to a quarter of a mile. I obtained a fine 
trumpeter swan on Pike lake [near Fort Snelling] Minnesota, in June 1853. 
They are quite common on the lakes in that vicinity in summer, breeding and 
raising their young. 

Supplementing Dr. Suckley's notes, J. G. Cooper, another natural- 
ist on this expedition added (1869 : 249) : 

The trumpeter swan associates with the preceding species [whistler] at the 
same season and in the same places. Both arrive from the north in the begin- 
ning of December, but I have not had an opportunity of noticing their departure. 

Earlier, the naturalist of another similar survey, Dr. J. S. New- 
berry, noted that farther south the swans were not as abundant as 
they were on the Columbia River, writing (1857: 100) : 

The trumpeter swan visits California and Oregon with its congeners, the 
ducks and geese, in their annual migrations, but, compared with the myriads of 
other water birds which congregate at that season in the bays and rivers of 
the west, it is always rare. Before we left the Columbia, early in November, 
the swans had begun to arrive from the north, and frequently while at Ft. 
Vancouver their trumpeting call drew our attention to the long converging lines 



LATER HISTORICAL NOTES (183 3-1925) 17 

of these magnificent birds, so large and so snowy white, as they came from 
their northern nesting places, and screaming their delight at the appearance of 
the broad expanse of water, perhaps their winter home, descended into the 
Columbia. 

This bird [whistler], considerably smaller than the last, is perhaps more 
common at the west. In California swans are much less common than on the 
Columbia, where during the winter season at least, they are exceedingly 
abundant. 

Dr. A. L. Heerman (1859 : 68) recorded the occurrence of the trum- 
peter in the Suisun and Sacramento Valleys of California, and in 
the San Francisco market. Since he made no mention of whistling 
swans, there may be some doubt about the accuracy of his 
identification. 

George Barnston, an official of the Hudson's Bay Company, verifies 
the general impression left by these observers as he relates (1862: 
7831-7832) : 

In the winter months all the northern regions are deserted by the swans, and 
from November to April large flocks are to be seen on the expanses of the large 
rivers of the Oregon territory and California, between the Cascade Range and 
the Pacific where the climate is particularly mild, and their favourite food 
abounds in the lakes and placid waters. Collected sometimes in great num- 
bers, their silvery strings embellish the landscape, and form a part of the life 
and majesty of the scene. 

Roderick MacFarlane was the next observer who made a significant 
contribution to our knowledge of the trumpeter as it existed in the 
days of long ago. MacFarlane was an experienced northern fur- 
trader, employed by the Hudson's Bay Company as a post and dis- 
trict manager for 22 years, from 1852-74, finally being appointed a 
Chief Factor of that firm in 1875. He was apparently one of the most 
qualified naturalists of "The Honorable Company" being an enthu- 
siastic and astute wildlife observer, collaborating with such other fa- 
mous early naturalists as Spencer F. Baird, Robert Kennicot, and 
Edward Preble, who w r ere also keenly interested in the fauna of the 
Arctic. 

During the period 1862-66 MacFarlane was especially active in 
collecting mammals, birds, and eggs. He w T as then stationed at Fort 
Anderson, located at latitude 68°30' N., longitude 128° W., which 
served a portion of the lower Great Mackenzie Basin fur trade. He 
collected representative fauna in the approximate area bounded to 
the north by the Arctic Ocean, to the east by the coast of Franklin 
Bay, to the south by the 67° of latitude, and to the west by the lower 
Mackenzie River — roughly a radius of about 125 miles, all within the 
Arctic Circle (66°33' N. Lat.) . 

With MacFarlane's background in mind it is interesting to review 
his notes (1891:425) : 



18 



DISTRIBUTION AND STATUS 



^ 




1>~^y 



<J53 



XV • *Fnrt Anriprtnn \ 1 



O DISTRICT HEADQUARTERS 

• TR40/N6 POST 



Fort Yukon 




\ \ -frFort, Anderson \ 

MACKENZIE \ * 
RIVER | 
DISTRICT — 

Fort Simpson-'fte. 

» -*£~- 

ATHABASCA I. / * 
DISTRICT. — -^\>* 



<~^3^Fort ResolutiorTV* p 



Fort Dunvegan^V" 



-if 



Fort Chipewyan 

Churchill 




I / (y^fork Factory 
^'.s'.7a ia^Wg^j' A&jtf 

ENGLISH RIVER '-. -p^ I TO ' Y0RK ~V • "A ft 

DISTRICT J ^-J. Norway Housi-^fc I DISTRICT \ Moose R ^ r /%*d<> 



Figure 1. — Locations of some Hudson's Bay Company posts which engaged in 
swan-skin trade, 1828-1884. 

Several nests of this species [trumpeter] were met with in the Barren Grounds 
[east of the Fort], on islands in Franklin Bay, and one containing six eggs was 
situated near the beach on a sloping knoll. ... It usually lays from four 
to six eggs, judging from the noted contents of a received total of twenty-four 
nests. 

MacFarlane was apparently able to distinguish between the nests of 
the trumpeter and those of the w-histler, as he reported the following 
information under the heading Whistling Swan: 

The maximum number of eggs taken in the twenty nests of this swan which I 
find recorded, was five, while the nest itself was always placed on the ground, 
and several were also found on the coast and islands of Liverpool and Franklin 
Bays in the Arctic Ocean. 

Later MacFarlane added (Mair and MacFarlane, 1908: 324), "For 
some time back sw r ans seem to be annually dwindling in numbers." 

With Richardson's earlier statements in mind, the following sum- 
mary by MacFarlane is particularly pertinent. Speaking of the 
Hudson's Bay Company's London sales, MacFarlane w r rote 
(1905: 754) : 

We find 57 swan skins in the above summary [of London fur sale-offerings 
1888-1897], and they no doubt belonged to the Hudson's Bay Company. 
Although no skins of Olor columbianus or Olor buccinator appear in the fur 
catalogues for 1897, 1900, 1902, or 1903 yet for many years they never failed in 
having quite a number of swan skins for sale in London. From 1853 to 1877 
they sold a total of 17,671, or an average of nearly 707 a year. There were 
seven good years (1853 to 1856, 1861, 1862, and 1867), with sales ranging be- 



LATER HISTORICAL NOTES (183 3-1925) 19 

tween 985 and 1,312 in 1854 (maximum), and seven poor years (1870 to 1877), 
with returns varying between 338 and the minimum (122) in 1877. 

Continuing, MacFarlane throws further light on the origin of some 
of the swan skins in the fur trade : 

From 1858 to 1884, inclusive, Athabasca District turned out 2,705 swan skins, 
nearly all of them from Fort Chipewyan. Mackenzie River District, according 
to a statement in my possession, supplied 2,500 skins from 1863 to 1883. From 
1862 to 1877 Fort Resolution, Great Slave Lake, contributed 798 thereof. For 
1889 Athabasca traded but 33, as against 251 skins in 1853. In 1889 and 1890 
Isle a la Crosse, headquarters of the English River District, sent out two skins 
for each outfit [post?]. 

Unfortunately, MacFarlane makes no attempt to distinguish be- 
tween the two species of swans, so that Richardson's original range 
statement cannot be compared with this record. A statement by 
Thomas Nuttall (1834: 371) that the trumpeter furnished the bulk of 
these skins is quoted by H. K. Coale (1915 : 83), but as this statement 
is in the exact wording used by Richardson the latter was no doubt the 
original source. 

From this information it is possible to trace in a rough way the 
decline of the trade in swan skins by the Hudson's Bay Company over 
a period of nearly 100 years, from 1806 to 1903. Since this trade by 
the Hudson's Bay Company began before 1772 and (as John Rich- 
ardson said in 1832) w T as principally at the expense of the 
trumpeter, the effect of such exploitation on the far-flung breeding 
populations of this species for more than 125 years must have been 
devastating and largely responsible for its extermination over vast 
regions, particularly in the heart of its Canadian breeding range. 

The relatively high price asked for trumpeter swans' eggs during 
the last decade of the 19th century also indicates the scarcity of the 
birds during this period. "The Standard Catalogue of North Amer- 
ican Birds Eggs" (Lattin, 1892) lists the cost of a single egg at $4, 
compared with a whistling swan egg price of $2.50, heath hen — $3, 
and whooping crane — $3. 

Another brief note by George Barnston (1862 : 7831) gives a definite 
swan breeding record for Eastmain Fort on James Bay. He states 
that swans are generally scarce in the Hudson Bay region, but a con- 
siderable number hatch in this area. 

Since whistling swans have never been reported nesting as far 
south as 52° N. (the latitude of Eastmain Fort), this is presumably 
a valid trumpeter breeding record, and the easternmost to come to my 
attention. 

There were three ecologically distinct regions in the United States 
in which trumpeters could be said to have once been a more or less 



20 DISTRIBUTION AND STATUS 

common breeding species in areas of suitable habitat. These regions 
were — 

(1) the Red Rock Lakes- Yellowstone-Jackson Hole region of 
southwestern Montana, northeastern Idaho, and northwestern 
Wyoming, (2) the Flathead Valley in western Montana, and (3) 
southern Minnesota and northern Iowa. Elsewhere in the United 
States, the trumpeter was recorded as a breeding species only occa- 
sionally or from widely separated locations. The prairie pothole 
country in the provinces of southern Canada and the Great Plains 
marshes of the United States were of small importance in supporting 
the total continental breeding population. Thus, little of the original 
prime breeding range of this species extended to the United States. 

During the period 1850-1900 a great many observations of the 
trumpeter swan in the United States were recorded. It was during 
this time that the work of the scientific field naturalist began to show 
up prominently in the literature. This in turn awakened scores of 
other interested observers to the value of ornithological factfinding, 
and the number of reports increased correspondingly. Except for 
notes on the breeding range, there seems little point in listing the 
scores of single occurrence records, since the trumpeter can be so 
easily confused with the whistler, and apparently often was. State 
ornithological works cover these occurrence records adequately. 

Several records of false or questionable nature have appeared in 
the literature of the past. The note by D. E. Merrill (1932 : 460) re- 
porting a trumpeter shot near Mesilla Park, New Mexico, is appar- 
ently a case of mistaken identity (J. Stokely Ligon, correspondence), 
and no skin is now available for confirmation (W. A. Dick-Peddie, 
correspondence). According to J. Van Tyne (correspondence), B. 
H. Swales, a qualified critical judge, does not believe the J. C. "Wood 
"record"' in the Auk (1908: 326), stating, "This record is worth- 
less — based entirely on memory. Wood did not know the swans or 
appreciate the value of accurate identification." Alfred M. Bailey 
(correspondence) reports that there is only one definite trumpeter 
record for the State of Colorado, that which Burnett reported 
(1916:199) as being shot near Fort Collins on November 18, 1897, 
other reports notwithstanding. 

Enough acceptable records are available from the states of Wash- 
ington, Oregon, and California in the Pacific flyway; Montana, 
Wyoming, North Dakota, Nebraska, Kansas, and Texas in the Cen- 
tral flyway; Minnesota, Wisconsin, Iowa, Illinois, Missouri, and 
Louisiana in the Mississippi flyway; and Maryland, Virginia, and 
North Carolina in the Atlantic flyway to demonstrate that the trum- 
peter still appeared as a migrant or winter resident in those states 
during the last half of the 19th century. Forbush (1912) cites a 



LATER HISTORICAL NOTES (1833-1925) 21 

report from A. S: Eldridge of Lampasas, Texas, that flocks of 75-1000 
trumpeters were seen there in the 1890's but none had been seen since 
1909. 

The continuity of the occurrence of trumpeters in the Yellow- 
stone Park region can be traced from statements made, or specimens 
secured, by Elliot Coues (1874: 544) , Dr. C. Hart Merriam (1891: 91), 
and W. C. Knight (1902: 40), to establish this general region as 
ancestral breeding range, even though their status in the Park during 
these early days is not evident. 

After the early observations cited, M. P. Skinner's random sight 
record data from 1915 to 1921 sheds some light on the number of 
swans in the Park during this period. Skinner (1925 : 154) records: 

My records of Trumpeter Swans seen in Yellowstone Park are as follows : 



May 29, 1915 4 

May 31, 1915 4 

Aug. 16, 1917 1 

June, 1919 1 



Aug. 14, 1919 2 

Sept. 6, 1919 5 

July 4, 1920 2 

May 29, 1921 1 



Skinner found a swan's nest near Lewis Lake in 1919. This estab- 
lished that trumpeters bred within the Park — a fact previously un- 
known. During the same summer, Dr. H. M. Smith, an early fisheries 
worker in the Park, reported 6 cygnets on a lake near Delusion Lake. 

A report for March 1920 from Acting Superintendent Lindsley of 
Yellowstone Park stated, "There are 20 to 30 trumpeter swans winter- 
ing in the outlet of Yellowstone Lake where there is constantly open 
water; also two wintering on Lewis River near the bridge, and two on 
Bechler River." 

Although the original status of the early swan populations inhabit- 
ing the Red Rock Lakes area is obscure, their occurrence in these 
marshes can also be traced from early times. From the 1880's (Bent, 
1925: 298) to 1896 (Brower, 1897: 138) and 1910 (French, letter) 
the early existence of these birds in that area is outlined. Informa- 
tion from older residents of the Centennial Valley confirms and ampli- 
fies these records establishing successive seasonal residence of swans 
on the Red Rock Lakes marshes since the early 1890's. This also 
agrees with information collected by Wright and Thompson (1935: 
104) , though again the actual level of any of these early populations 
was never recorded. A. C. Bent's source did report that he saw 
"quantities'' of swans and killed "many" young birds for food in the 
Centennial Valley, so they were by no means rare there. 

A later note regarding actual swan numbers in the Red Rock Lakes 
area is that left by C. S. Sperry when he surveyed these marshes as a 
waterfowl food-habits biologist for the!". S. Biological Survey in 1922. 
Sperry noted "about 15 swans" were reported on the Lakes during 
that nesting season. He further remarks that this species was 



DISTRIBUTION AND STATUS 




Figure 2. — A day's bag of waterfowl in 1895 at Red Rock Lakes included a 

trumpeter swan. 

"frequently" encountered during his week's work on these marshes in 
September of that year. 

The published notes concerning the early history of the trumpeter 
in Alaska before 1925 are both limited and brief. This is of course 
partly due to the fact that the United States did not acquire this vast 
region from Russia until 1867, but a search of the Russian records re- 
vealed little additional information. Swans do not seem to have been 
important in early Alaskan fur trade, although the taking of swans in 
arctic Russia for economic uses has apparently been practiced for 
years. 

The initial comment on the trumpeter in Alaska seems to be that left 
by Dr. Edward Adams, a competent English ornithologist, who made 
the following observation during the period 1850-51 (1878: 430) : 

Cygnus buccinator. This was the only species of Swan I met with at Michalaski 
[St. Michael]. The first appeared on the 30th. of May; but they were at no time 
numerous, from two to eight or ten keeping together. A few of them are said 
to breed here ; but most of them go further north. 

Although this note was subsequently credited by Baird, Brewer, and 
Ridgway (1884: 433), this has not been the case since that time, other 



LATER HISTORICAL NOTES (1833-1925) 23 

writers apparently overlooking this account or believing that Dr. 
Adams confused the trumpeter with the whistler. Subsequent egg 
records of trumpeters in the Norton Sound area strengthen Dr. 
Adams' original statement. 

Later, Dall and Bannister (1869: 294) stated that eggs of the 
trumpeter were obtained by a Mr. Lockhart at Fort Yukon, thus es- 
tablishing the first definite breeding record for this species in Alaska. 
One of these eggs, on deposit in the U.S. National Museum, was 
received in April or May of 1863. 

E. W. Nelson (1887 : 93) refers to DalPs note and remarks that both 
trumpeters and whistlers are to be found on the southeastern Alaskan 
coast during the migrations, and attributes the lack of knowledge re- 
garding this species in the Territory to the unexplored interior. 

In addition to these notes, there are two egg records, one in the 
Chicago Natural History Museum and another in the U. S. National 
Museum, both credited to trumpeters in Alaska and apparently un- 
published. The R. M. Barnes collection at the Chicago Natural His- 
tory Museum contains four eggs of this species which were collected 
June 28, 1902, 38 miles northeast of Cape Nome by Walter E. Bryant. 2 
The U. S. National Museum also contains an egg (one of two) col- 
lected by J. B. Chappel in Norton Bay (near Cape Denbigh) in 1867. 

There is also a clutch of 5 swan eggs in the National Museum at- 
tributed to the whistler but whose measurements, shape, and texture 
leave little doubt that they belong to the trumpeter. This collection 
is credited to A. H. Twitchell and was made June 4, 1915, at Bethel, 
Alaska. 

Table 2 outlines the former breeding range of the trumpeter on the 
continent, particularly in the United States and Alaska. The breed- 
ing range in Canada is shown in a general sense on the map, figure 3, 
based upon the preceding testimony of Richardson, MacFarlane, and 
Barnston as well as more recent information written later by J. A. 
Munro (1949 : 49), Brooks and Swarth (1925 : 38), J. D. Soper (1949 : 
240), and others. 

These records vary a great deal in value, but the fact that most 
questionable records apply to the same general region as records of 
higher caliber, or have been accepted by earlier qualified observers, 
would appear to upgrade their value. The sparse and localized nest- 
ing population of trumpeters in the United States no doubt accounts 
for the comparative paucity of U. S. breeding records. 



■U. S. Fish and Wildlife Service files refer to a set of four eggs in the R. M. Barnes 
collection with a date of June 15, 1905, but otherwise with the same data as these. Only 
"iic such set is in the collection, so the Service files must be in error on the date. Measure- 
ments of these eggs provided by Melvin A. Traylor (letter) are average for the trumpeter, 
and there seems no reason to doubt the record. 

469660 O— 60 3 



24 



DISTRIBUTION AND STATUS 






Table 2. — Trumpeter swan breeding records in the United States am 

Alaska to 1925 



State and locality 

Washington: Cherry Lake 

(Whitman Co.). 
Montana: 

Near Flathead Lake (w. 
Montana). 

Thompson River (w. Mon- 
tana). 

Clearwater drainage (Mis- 
soula Co.). 

Lake Rodgers (Flathead Co.) 

Swan Lake (Lake Co.) 

F'athead Lake (w. Montana) 

Centennial Valley (Beaver- 
head Co.). 

Flathead Valley (w. Mon- 
tana). 

Headwaters of South Fork of 
Flathead River. 

Swan Lake (Beaverhead Co.) 

Red Rock Lakes (Beaverhead 
Co.). 

Big Lake (Stillwater Co.) .... 

Big Lake (Yellowstone Co.) 
35 mi. northwest of Billings, 
Mont. 

Red Rock Lakes (Beaverhead 
Co.). 

Highland Lakes CFergus Co.) 

Idaho: 
Henrys Lake (Fremont Co.) . 
Grays Lake (Bonneville Co.). 

Icehouse Creek Reservoir 
(Fremont Co.). 
Wyoming: 
Jackson's Hole 

Lakes near head Green River. 

Near Lewis Lake (Yellow- 
stone Park). 

South of Delusion Lake in 
Yellowstone Park. 

Yellowstone (Valley) Region 
North Dakota: 

Island Lake (Barnes Co.) 

Rock Lake (Towner Co.) 

Along Red River of The 
North. 
Nebraska: 

Watt's Lake (Cherry Co.)_.. 

Swan Lake (head of the Little 
Blue; Adams Co.). 
Minnesota: 

Swan Lake (Nicollet Co.) 

Pike Lake (near Old Fort 
Snelling). 

Heron Lake (Jackson Co.)... 

Everson Lake (Meeker Co.) . 

Along Red River of The 
North. 
Iowa: 

Near Sac City (Sac Co.) 



Oakland Valley (Pottawat- 
tamie Co.). 

Near Hdwtrs. Des Moines 
River (Emmet Co.). 

Little Twin Lakes (Hancock 
Co.). 

Near Newton (Jasper Co.) 

Spirit Lake (Dickinson Co.).. 
Missouri: 

Lowland lakes near Alex- 
andria (Clark Co.). 

Northeastern Missouri 

Opposite Atchison, Kans. 
(Buchanan Co.). 



Date 



Until 1918. 



Apr. 15, 1842. 

1871 

1881 



1881 

1881 

1881 

1883-1888. 



Until 1886... 



June 10, 1896. 
1910 



1917. 
1920. 



1922 

None 



August 1877. 
1923, 1924.... 



Early 1920's. 



None... 
None ... 
1919.... 



July 19, 1919. 
None..- 



Mid 1880's. 

1895 

None. 



None. 
None. 



July 13, 1823. 
June 1853.... 



1883 

1884"orT88B 
None 



1859 

1870 

1871 

As late as 1875. 

As late as 1883- 

None 

None 



None 



None.. 

July 4, 1804. 



Authority 



Old settler. 



Pierre Jean De Smet. 

E. S. Cameron 

E. S. Cameron 



E. S. Cameron. 
E. S. Cameron . 
E. S. Cameron- 
Ed Forbes 



E. S. Cameron 

E. S. Cameron 



Cecil French _ 



Rancher's report. 
G. B. Thomas ... 



C. C. Sperry 

P. M. Sllloway (1903: 
15). 



C. E. Bendire ... 
B. Fordyce (MS.). 



S. A. Trade (MS.). 



M. P. Skinner.. 
F. V. Hayden.. 



Newspaper report . 
Alfred Eastgate... 



G. E. Beltrami. 



Thomas Miller. 
L. O. Dart 



J. A. Spurrell. 
J. A. Spurrell- 
W. C. Rice... 
J. W. Preston. 
J. W. Preston. 



Old hunters 

Lewis and Clark. 



Reference 



C. F. Yocom (1951:17). 

R. G. Thwaites (1906). 

A. C. Bent (1951: 296). 

H. K". Coale (1915:87). 

H. K. Coale (1915:87). 
H. K. Coale (191.5: 87). 
H. K. Coale (1915: 87). 
A. C. Bent (1951: 298). 

H. K. Coale (1915: 87). 

A. C. Bent (1925: 297). 

J. V. Brower (1897: 138). 
U. S. Fish and Wildll 

Service files. 
A. A. Saunders (1921:42) 
U. S. Fish and Wildl; 

Service flies. 

U. S. Fish and Wildli 

Service files. 
A. A. Saunders (1921:41) 

C. Hart Merriam (1891:9 
U. S. Fish and Wildli 

Service files. 
U. S. Fish and Wildli 

Service files. 

W. C. Knight (1902:40). 
W. C. Knieht (1902:40). 
M. P. Skinner (1920). 

A. C. Bent (1925: 297). 

E. Coues (1874:544). 

R. Reid (correspondence 
N. A. Wood (1923:23). 
W. W. Cooke (1887). 



J. M. Bates (1900: 16). 
J. M. Bates (1900: 16). 



T. S. Roberts (1936: 204). 
G. Suckley (1859: 249). 

T. S. Roberts (1936: 205). 
T. S. Roberts (1936: 205). 
W. W. Cooke (1887). 



U. S. Fish and Wildli 

Service files. 
U. S. Fish and Wildli 

Service files. 
Baird, Brewer, Ridgwf 

(1884: 432). 
R.M.Anderson (1907: 19: 

R. M. Anderson (1907: 191 

W. W. Cooke (1887). 
A. A. Mosher (1889: 66). 

J. Blines (1888:343). 

Otto Widman (1907). 
R. G Thwaites (1906). 



EARLY MIGRATION NOTES 



25 



State and locality 



Date 



Authority 



Reference 



Wisconsin: 

[ Northwest Wisconsin . 

, Southern Wisconsin... 



None 

"Early forties". 



i Jefferson County 

ndiana: Kanakee marshes 

< (Lake Co.). 

Llaska: 

I Norton Sound- . 

j Fort Yukon... 



1842-1845. 
None 



T. H. Ball. 



1850-51 

Before 1863. 



Norton Sound, near Cape 

Denbigh. 
Norton Sound, 38 mi. NE 

of Cape Nome. 



1867 

June 28, 19021. 



Mr. Lockhart. 
J. B. Chappel. 



R. M. Barnes (egg 
collection). 



Grundtvig (1895:99). 
Kumlien and Hollister 

(1903: 31). 
Kumlien and Hollister 

(1903:31). 
A. W. Butler (1897:642). 



Dr. E.Adams (1878:430). 

U.S. Natl. Mus. (egg col- 
lection). 

U. S. Natl. Mus. (egg col- 
lection). 

Chicago Natural History 
Museum. 



1 See footnote, page 233. 

From all of the foregoing information it is evident that trumpeter 
swans were once an abundant and widespread species on the continent. 
Although both of the native swans were killed by the early colonists 
(for food and plumage, the first substantial inroads into trumpeter 
populations occurred when the fur trade exploited this species over 
wide regions for more than a hundred years. The surge of western 
Settlement during the latter part of the 19th century was also partly 
responsible. The white settlers not only killed and dispossessed these 
birds in the southern portions of their breeding range, but further 
reduced or extirpated populations breeding farther north as the in- 
creased hunting pressure to the south took its toll among migrating 
flocks. 

EARLY MIGRATION NOTES 

It is believed worthwhile to present here a number of brief state- 
ments by the early observers in order to document the migration 
habits of the trumpeter when many of the original flocks were more 
or less intact. 

SPRING MIGRATION (departure) 

In regard to departure times, John Lawson (1714) noted that the 
"trompeters" along the coast of North Carolina in 1700-1701 "stay 
. . . until February . . . when they go to the Lakes to breed." Also, 
Major Long corroborated this early movement of swans generally, 
reporting (James 1823 : 191) under date of February 22, 1820, "swans, 
geese, and ducks flying up the river" while on the Missouri River in 
the northwest corner of Iowa. Lewis and Clark (Coues, 1893: 915) 
noted great numbers of both swan species still on the lower Columbia 
River on the 29th of March, though whether these were the original 
winter residents or spring migrants enroute from points farther south 
is not mentioned. 




> hypothetical eastern limit 
wintering range ^ 



Figure 3. — Former breeding and wintering range, trumpeter swan. 



In the Mississippi River drainage, John J. Audubon (1838: 537) 
corroborated the Lawson and Long testimony implying that the nor- 
mal movement north commenced in February. Observations made at 
Red Rock Lakes confirm a statement of Audubon's that trumpeters 
move back towards their breeding grounds during the late winter upon 
the advent of moderating weather. 

A. C. Bent (1925: 301) supplies a couple of late departure dates 
as follows, "Arkansas, Helena, April 29, 1891; British Columbia, 
Osoyoos, April 25." 

SPRING MIGRATION (arrival) 

Several early observations agree that the trumpeter arrives on its 
northern nesting grounds among the earliest of any of the Arctic 
avifauna. The journal of Samuel Hearne (1795: 285, 435) under 



EARLY MIGRATION NOTES 27 

date of April 12, 1772, and latitude of about 60° when just south of 
the Great Slave Lake bears this entry : 

On the twelfth we saw several swans flying to the Northward ; they were the 
first birds of passage we had seen that Spring except a few snowbirds. 
******* 

In the interior parts of the country the larger Swan [trumpeter] precedes 
every other species of water-fowl, and in some years arrive so early as the 
month of March, long before the ice of the rivers is broken up. At those times 
they always frequent the open waters of falls and rapids, where they are fre- 
quently shot by the Indians in considerable numbers. 

Richardson (Swainson and Richardson, 1832: 464) noted the early 
arrival of the trumpeter on its far northern breeding grounds, "It 
breeds . . . principally within the Arctic Circle, and in its migra- 
tion generally precedes the Geese a few days." 

Although both Barnston (1862: 7831) and Dr. Adams (1878: 430) 
agree that the trumpeter arrives on its breeding grounds in small 
flocks, the coastal arrival of trumpeters in the Norton Sound area of 
Alaska noted by Adams was preceded by both geese and ducks. 

A. C. Bent (1925: 301) lists some average dates of arrival for this 
species as follows, "Nebraska, March 16; South Dakota, April 2; 
Minnesota, Heron Lake, April 4; Saskatchewan, April 16; British 
Columbia, April 20." In the light of previous testimony, however, 
these would appear to be later than average arrivals. 

FALL MIGRATION (departure) 

Richardson (Swainson and Richardson, 1832: 438) furnishes the 
sole remark regarding this topic, writing, "Cygnus buccinator . . . 
remains later in the season [than the Geese]." This agrees with my 
observations at Red Rock Lakes where the trumpeters are con- 
sistently the last to leave the remaining patches of open water before 
the final f reezeup. 

FALL MIGRATION (arrival) 

R. G. Thwaites (1906, vol. 21: 336) presents a note from Town- 
send's narrative written at Fort Vancouver on the lower Columbia 
River, December 1, 1835, "The duck and geese, which have swarmed 
throughout the country during the latter part of the autumn, are leav- 
ing us, and the swans are arriving in great numbers." This agrees 
well with the observation by Dr. Suckley (1859: 249) in the Puget 
Sound and/or Columbia River regions, where both swans arrived 
from the north at the beginning of December. 

Dr. J. G. Cooper (1869: 83), writing of unidentified swans about 
the same period, stated, "Swans were seen in large flocks on the Co- 
lumbia River, in the Cascade Canyon, as early as October 29th, this 



28 DISTRIBUTION AND STATUS 

year (1860), and their migration southward seemed generally early. 
I saw them, however, on lakes of the Columbia Plain about the same 
time in 1853." Dr. Newberry (1857:100) mentions an early No- 
vember arrival on the Columbia. In the Mississippi flyway, Audubon 
(1838:537) reports late October arrivals usual on the lower Ohio 
River. 

A. C. Bent (1925:301) lists some fall occurrence dates as follows, 
"Minnesota, Spicer, October 8, 1913 ; Michigan, St. Clair Flats, No- 
vember 20, 1875; Washington, Douglas County, November 9, 1912; 
Colorado, Fort Collins, November 18, 1897, and November 15, 1915." 

Dr. Grinnell found trumpeter swans in October and November of 
1887 abundant at the extreme upper end of Lower St. Mary Lake near 
Glacier National Park (Bailey, 1918). 

RECENT OCCURRENCE, BREEDING, AND 
MIGRATION REPORTS, TO 1957 

Positive identification as to species, trumpeter vs. whistler, was not 
possible in most of the following reports, but they do constitute the 
opinion of trained or reliable wildlife observers and as such probably 
represent valid trumpeter records. 

Since the Red Rock Lakes Refuge was established in 1935, the in- 
formation from Montana, Idaho, and Wyoming is believed to be es- 
pecially pertinent. As might be expected from the resultant increase 
in the swan population following establishment of the Refuge, these 
birds may be moving about in this tri-State region more than is gen- 
erally realized and in some instances appear to be nesting in areas 
previously unoccupied. It is necessary to depend upon sight and/or 
sound observations in all these cases since, although over 300 trum- 
peters have been banded on the Refuge and stations of their introduc- 
tion, no band recoveries have yet been made outside the generally rec- 
ognized area of known use. (The banding data which have been 
gathered will be treated later under Management) . The known sight- 
ings of trumpeters since 1925, as well as the last previous occurrence 
report, are presented below by flyway for each State, Province, or 
Territory considered. 

PACIFIC FLYWAY 

California. Donald D. McLean (1937: 228), California Division 
of Fish and Game, gives a convincing description of a trumpeter 
which was seen in Lassen County and identified from its call, on 
November 8, 1935. The bird was seen between Grasshopper Valley 
and Termo, by Bailey Creek. 



RECENT OCCURRENCE, BREEDING, AND MIGRATION REPORTS 29 

The next previous California report seems to be that of A. L. 
Brown, who stated that the trumpeter was formerly a regular winter 
ipisitant to Lassen County (Honey Lake Valley) up to about 1910 
■ McLean, correspondence) . 

! Nevada. Frank W. Groves, present Director of the Nevada State 
JFish and Game Department, who was familiar with the trumpeter 
{from his previous experience at the Malheur National Wildlife Refuge, 
reported by letter dated December 2, 1952, to the Fish and Wildlife 
Service that he found a swan while driving east from Carson City. 
The bird was larger than a whistling swan, and lacked the yellow 
spot at the base of the bill. Groves was convinced that its call, while 
not completely typical, was that of a trumpeter. 

This lone trumpeter, if such it was, must have been a stray from 
the groups which had been transferred to the Ruby Lake Refuge in 
Nevada or the Malheur Refuge in Oregon from the Red Rock Lakes 
Refuge. There is no earlier report of the trumpeter in Nevada. 

Oregon. The only report of trumpeters within the boundaries of 
Oregon since 1925 seems to be that of Gabrielson and Jewett (1940) 
who state, "On September 7, 1929, Oberholser, Gabrielson, and Jewett 
saw a single swan at Davis Lake that, judging from its huge size, 
might have been this species. This is the only recent record of even 
its hypothetical occurrence within the State." Earlier reports and 
records of this species in Oregon precede 1900. 

Washington. Allan Brooks (1926 : 129) states that a small number 
of trumpeters cross the International Boundary from British Colum- 
bia into Washington State each year (18 in 1924). Whether this is 
still true is not known, although J. A. Munro and Ian McT. Cowan 
(1947: 55) write, "In addition to the numerous wintering bands, that 
are widely distributed [in British Columbia], a smaller number is 
transient and apparently winters south of the Canadian-United States 
boundary. Migrating birds are met with both spring and fall in 
various localities in the extreme southern part of the Province." 

J. B. Lauckhart, Chief of the Division of Game Management, 
Washington Department of Game, advised (correspondence), "We do 
have some swan killed each year during the hunting season, but we 
have never identified one as a trumpeter." The occurrence of the 
trumpeter in a wintering status in Washington is thus obscure. 

On April 7, 1939, W. C. Ralston saw and heard 20-25 trumpeters at 
Othello, Washington. 

The next earlier reports are mentioned by Stanley Jewett et al. 
(1953 : 102) of a specimen taken at Moses Lake, Washington, in 1912, 
and the skin received by Edson in 1913 from a hunter who took the 
specimen at the mouth of Nooksack River, Bellingham Bay (Whatcom 
County). 



30 DISTRIBUTION AND STATUS 






Idaho. In addition to counts made in the winter range of the 
trumpeter in the Island Park area of Fremont County, Idaho, these 
great birds are occasionally noted farther down the Snake River 
drainage and more frequently elsewhere in eastern Idaho. David 
de Lancey Condon (MS), Chief Naturalist for Yellowstone National 
Park, noted : 

On April 17, 1937, I observed five swan on a slough on Marsh Creek which is 
a tributary to the Portneuf River, 30 miles south of Pocatello, Idaho, which, 
after careful examination with field glasses were felt to be trumpeter swan. 
On April 18, 1937, a pair of trumpeter swan were watched for some time on an 
oxbow lake in the Snake River bottoms near Roberts. 

Robert Salter (1954), Game Bird Supervisor of the Idaho State 
Fish and Game Department, also set down several occurrence records 
well outside the Island Park area in eastern Idaho : 

They normally do not go further downstream along the North Fork of the 
Snake River than St. Anthony which is approximately 30 air line miles south of 
the Island Park area. We have three records in recent years of trumpeter 
swan being found in Idaho outside Fremont county. In the fall of 1943 Mr. 
Hawley Hill, District Supervisor, positively identified by dissection a trumpeter 
swan which had been illegally killed on the Snake River near Burley in Cassia 
County. On March 30, 1951, Mr. E. L. Keppner, Conservation Officer, made a 
sight and "sound" record of five trumpeter swan on Elk Horn reservoir in 
Oneida County. On January 9, during the 1952 winter inventory, 11 swans were 
seen from the air on Spring Creek, which runs into American Falls reservoir in 
Bannock County. Mr. Winston E. Banko, Refuge Manager of Red Rock Lakes 
Refuge, was in Idaho Falls and went out the next day to observe these birds. 
We were able to approach within 75 yards of three birds which then flushed. 
Banko identified them as trumpeters [from their voice]. 

Swans, apparently trumpeters from the wintering flock at the Na- 
tional Elk Refuge in Jackson Hole, Wyoming, only rarely take a 
cruise downstream as far as Swan Valley on the South Fork of the 
Snake River in Bonneville County, Idaho. My field notes, under 
date of December 12, 1956, read : 

Arno Winterfield was contacted in Swan Valley [Bonneville County, Idaho] 
regarding the former and present status of trumpeter in that area. He ad- 
vised that he came to Swan Valley in 1915 and had never seen swans there 
until the winter of 1954 or 1955 when he saw 8 or 9 flying low downstream along- 
side the Snake River south of the community of Swan Valley. As he watched, 
the birds turned and headed back upstream. He did not recall that old-time 
residents of Swan Valley had ever mentioned wintering or breeding swans in 
that area. 

U. S. Fish and Wildlife Service files hold an earlier trumpeter 
breeding record for Idaho. R. F. James, in a game warden report 
dated November 23, 1932, reports a nest on an island in the Pend 
Oreille River just below the Idaho-Washington line. 






RECENT OCCURRENCE, BREEDING, AND MIGRATION REPORTS 31 

The next previous records for Idaho are breeding accounts given 
for Grays Lake, Idaho, in 1923 and 1924 (A. B. Fordyce, MS) as 
well as on Icehouse Creek Reservoir, Fremont. County in 1920 (S. A. 
Trude, MS) , both brief manuscripts in Service files. 

In addition to these reports, R. H. Mackay (1957: 339) documents 
both sight record and band-recovery data establishing the presence 
of Alberta trumpeters among flocks of Montana birds in 1955 and 
1956 during the wintering period in the Island Park country of Idaho. 

British Columbia. Although reports of breeding trumpeters in 
British Columbia are rare, and for the most part unsubstantiated, 
hundreds of birds winter along the coast as well as in the interior. 
J. A. Munro (1949 : 709) states : 

It is not an exaggeration to say that trumpeter swans, at one time or another 
during the winter, visit most of the many lakes on Vancouver Island and along 
the mainland coast. The number fluctuates from year to year but probably 
exceeds 600 individuals. Thus, the winter population is believed to approxi- 
mate 1,000. . . . wintering populations of trumpeter swans are distributed 
over the western half of the province of British Columbia between north 
latitudes 49° and 55°. . . . Some of these frequent lakes and rivers at points 
distant from the sea ; others inhabit the lower reaches of coast streams ; and 
still others resort to the shallows and mud flats of sheltered estuaries. There 
are also interior populations, of which some winter on rapid stretches of river 
that remain open even in the coldest weather. 

Segments of the population wintering in British Columbia have 
been counted from time to time by Munro and others, but it is dif- 
ficult to secure anything like a true census due to the widely separated 
locations of wintering flocks and the relatively poor weather for 
operating aircraft which prevails over the entire region at this season. 
Also, the possible presence of whistling swans further complicates 
the matter. R. H. Mackay (correspondence, March 19, 1958) writes: 

Whistling swans winter in British Columbia to some extent. The largest 
group is a flock varying annually from 150-225 birds that winter regularly on 
the South Thompson River near Kamloops in the interior of the Province. 
Other irregular occurrences have been noted in the Lower Mainland region of 
the Fraser, on the Queen Charlotte Islands, and at Lonesome Lake where two or 
three whistlers have been recorded with the trumpeters on occasion. 

Although the breeding grounds of the trumpeters wintering in this 
Province are in the main still undiscovered, it is seen that easily half 
of the total North American population winters here, 

Alaska. For over 50 years following E. W. Nelson's (1887) state- 
ment that both species of swans were noted to occur "in migration" 
in southeastern Alaska, little was written to clarify the status of the 
trumpeter in that Territory, though collection of several clutches of 
eggs in the Norton Sound region confirmed Dall and Bannister's 
(1869) breeding record. 



32 



DISTRIBUTION AND STATUS 



A report by E. L. Kepner found in U. S. Fish and Wildlife Serv- 
ice files, and dated March 8, 1924, states: 

Camp Kora Kora ; Lake Minchumina, Kantishna District. This low lying 
section of the interior is especially adapted to the waterfowl. . . . the great 
white Trumpeter Swan is the least plentiful of all the species, but he is also 
in evidence in goodly numbers, and I have also noticed the hunters after it. 
... I believe the Treaty regulations between the U. S. and the Dominion of 
Canada afford them all the protection required. Lake Minchumina appears 
to be a favored spot for them to stop over and feed and rest on their northern 
and southern migration. 

This report implies that a remnant of the arctic breeding popula- 
tion still remained 35 years ago. Lake Minchumina is located in cen- 
tral Alaska at about latitude 64° N, If still in existence, this popula- 
tion may form a portion of those currently wintering in British 
Columbia. 

Ira N". Gabrielson's (1946:102) note was the next pertinent oc- 
currence record. This confirmed E. W. Nelson's earlier remark that 
the trumpeter was to be found in southeastern Alaska at certain sea- 
sons including wintering populations, in common with British Co- 
lumbia. 

Subsequent swan census work in Alaska, accomplished during the 
annual January waterfowl inventory since that time, indicates that 
the number of swans which actually winter in southeastern Alaska 
varies greatly, but generally is fewer than indicated in Gabrielson's 
report, whose count of March (1945) probably included at least some 
northward-bound birds actually in transit. The swan census data 
gathered by the U. S. Fish and Wildlife Service in coastal Alaska 
during the period 1949-57 are presented in table 3. 

Table 3. — Swans censused, Alaskan waterfowl inventory, January, 1949 to 1957 

[U. S. Fish and Wildlife Service records] 



Year and location 


Number 
of swans 


Year and location 


Number 
of swans 


1949: i 
Ketchikan area 


35 
4 
2 


1954: 


2 127(+50) 
34 


Petersburg area 




Sitka area.. 




2 50 (+25) 




Total 


Total 


41 


2 217(+75) 




1955: 3 


1950: 
Ketchikan area 


124 
2 

126 


32 


Craig area ... . . 




16 






59 


Total 




2 




Total 




1951: i All areas 



37 

2 49 (+30) 
12 


109 


1952: Ketchikan area.. . 


1956: Ketchikan and Petersburg areas 

combined (reduced coverage) 

19 7: Ketchikan area... . . .. 




1953: 
Ketchikan area 


15 


Craig area... 


82 






Total 


2 61 (+30) 











1 Widespread and prolonged freezing of fresh-water lakes and protected bays. 

2 Figures in parentheses are estimated numbers in addition to swans counted. 

3 Mild winter noted, many inland lakes open, vessel survey. 






RECENT OCCURRENCE, BREEDING, AND MIGRATION REPORTS 33 

In addition to the southeastern Alaskan areas listed, swans have 
been reported from time to time wintering on the Alaska Peninsula, 
Kodiak-Afognak Islands, and Prince William Sound (W. A. Elkins, 
correspondence) . Although the species of these birds has not been 
determined, it is believed that they represent trumpeters also. 

The census figures presented in table 3 are not directly comparable 
for many reasons. The exact habitat covered may not be the same 
from year to year even within a designated district, census observers 
and techniques (aerial, vessel, etc.) vary frequently, the weather no 
doubt affects distribution greatly and in unknown ways, and swans 
may not return to the same district each year. Also, the possible oc- 
currence of wintering whistlers in southeastern Alaska cannot be 
entirely dismissed at this time. 

Since small numbers of whistling swans have been known to pass 
the winter as far north as Washington, Idaho, and British Columbia, 
the data in table 3 may include some of the lesser species also, although 
there appears to be little doubt that the trumpeters are substantially 
represented here. 

Alda Orton's (1951:10) article reporting the presence of a small 
number of breeding trumpeters on the lakes of the Naha River Valley 
north of Ketchikan appears to be the first breeding record for this 
species in that area. In this instance it is interesting to note that 
the breeding and wintering range overlap as they do in the Red 
Rock Lakes and the Yellowstone Park areas to some extent. 

The first indications of a substantial breeding flock of trumpeters 
in Alaska was brought to light as a result of field work in the lower 
Copper River Basin by Melvin A. Monson (1956: 444-445). Flying 
the area comprising the convergence of the Tasnuna and Bremner 
River Valleys with that of the Copper River on August 11, 1955, Mon- 
son censused 69 adult swans and 5 broods totaling 15 cygnets. Identi- 
fication of several of this group as trumpeters had previously been 
made on the ground from voice calls, bill characteristics, and egg speci- 
mens obtained. 

Trumpeter swan investigations are currently underway by U. S. 
Fish and Wildlife personnel at the Kenai National Moose Range on 
the Kenai Peninsula, Alaska. The following information was 
gathered in the preliminary studies by David Spencer (Refuge Super- 
visor), Jim Johnson (Refuge Manager), and Jim Branson (Game 
Management Agent), and furnished by Spencer (correspondence, 
August 30, 1957) . I have summarized it as follows : 

Swans have been known to nest on the northern part of the Kenai Peninsula 
over a long period of years. From time to time a few birds have been shot by 
waterfowl hunters. Two of these which were recovered in 1951 and 1956 were 
identified as trumpeter swans. It appeared likely these were Kenai nesting 



DISTRIBUTION AND STATUS 




Figure 4. — Aerial view of trumpeter breeding grounds in lower Copper River 
Basin, Alaska, at confluence of Tasnuna, Bremner, and Copper Rivers. 



birds. Investigations in 1957 were aimed at determining the identification of 
the Kenai nesting swans and to estimate the nesting population. 

The first birds, one pair and one juvenile, were noted on the east fork of 
Moose River on April 2. The main group of swans arrived on the nesting ground 
about the third week in April. A flock of 4S adult swans observed April 30 were 
believed to be migrating birds as there was no evidence that this flock re- 
mained in the area. 

Approximately 20 pairs of swans nested on the Kenai Moose Range this year. 
An additional 10 nonbreeding birds appeared to be in the area. Three clutches 



RECENT OCCURRENCE, BREEDING, AND MIGRATION REPORTS 35 

of eggs had measurements within the trumpeter size range. The male of a nesting 
pair which was collected was identified as a trumpeter. 

The fact that only a single juvenile bird was observed to return in the spring 
suggests considerable loss among the first-year birds. The single swan (speci- 
men) collected had been previously shot. Banding is indicated as an initial 
step in management, since it will be necessary first to determine the wintering 
area of the Kenai population before steps are taken to analyze losses. 

CENTRAL FLYWAY 

Utah. At least two sight records of swans believed to be trumpet- 
ers have been made by well-qualified wildlife observers in Utah during 
the period 1925-57. While positive identification was not made in 
either of these cases, it is not believed likely that both are in error. Dr. 
Clarence Cottam (correspondence) reported that A. V. Hull, formerly 
a Service employee at the Bear River Refuge, observed a trumpeter on 
that area on June 14, 1932 ; also that in July 1940 Dr. D. I. Rasmussen 
and Leo K. Couch reported an immature trumpeter on Strawberry 
Reservoir. 

There is apparently no record that the trumpeter has ever bred 
within the State of Utah, although several early occurrence records do 
exist before 1925 (1 in 1923, 1 in 1907 or 1908, and 6 specimens captured 
in 1901 — Cottam, correspondence) . 

Wyoming. Vernon Bailey (1930 : 188) mentions a few early random 
Park trumpeter occurrence records, among them 2 breeding pairs seen 
in 1926, 1 pair on Bridger Lake and the other near Yellowstone Lake. 
George Wright and Ben Thompson (1935: 104) add, "prior to 1929 
a pair of trumpeters had been known to make unsuccessful nesting at- 
tempts at Trumpeter Lake in Lamar Valley." Wright and Thompson 
sum up the early situation : 

Early superintendents' reports have mentioned the presence of swans in the 
Park. ... but we have not been able to ascertain whether there was a period 
of interruption when the birds did not breed in the Park at all, or whether they 
simply became so scarce as to be generally overlooked. The latter is probably 
the case. 

It is a fact that in recent years there has been an increase in the number of 
trumpeter swans breeding in the Park. To a degree this increase may be more 
apparent than real, inasmuch as more attention has been focused on the swans 
than before and nesting stations recently reported may have been previously 
overlooked. 

Other recent records are given under Annual Swan Census, 1929-57. 

Two recent Wyoming reports of the trumpeter's occurrence outside 
the Yellowstone-Jackson Hole area have been made. The first is 
especially interesting. Robert L. Patterson (correspondence) fur- 
nished this as follows : 

Of . . . interest is a report of a pair of wild swans and five cygnets seen in 
September, 1953, on a small lake in the vicinity of Pathfinder Migratory Bird 



36 DISTRIBUTION AND STATUS 

Refuge in central Wyoming. We rather assume that this observation was of 
trumpeter swans although, of course, it is not verified. The observation was 
made by George Wrakestraw, one of our wildlife biologists. 

The second record is that of 2 trumpeter cygnets, banded on Lowe 
Lake in the Grande Prairie region of northwestern Alberta in the 
summer of 1956 by Canadian Wildlife personnel. The birds were 
found dead, apparently shot, near Cody, Wyoming, on or before 
October 27, 1956 (R. H. Maekay, 1957 : 339) . 

Montana. In this state no reports of trumpeters occurring outside 
their known range during the early part of the 1925-57 period are 
known. There are several interesting reports since 1950. 

K. F. Roahen (correspondence), U. S. Game Management Agent, 
reported a trumpeter found dead in poor flesh (weight 27 pounds) 
at Freeze-out Lake, Fairfield, Montana (35 miles west of Great Falls), 
on October 10, 1950. 

Henry Lentfer, taxidermist in Livingston, Montana, furnished this 
note (correspondence, January 1,1956) : 

A Mr. Hansen who works for the Montana Power Company told me about a 
pair [of swans] that evidently were about to nest on a small lake near Mystic 
Lake on the Rosebud last spring or early summer as he saw them for quite a 
while and then one flew into the powerline and was killed . . . (weight 22^ 
pounds). 

Ralph L. Hand, retired U. S. Forest Service official, also furnished 
an interesting note of a flight of about 30 swans believed to be trum- 
peters seen over Missoula, Montana, on October 31, 1953. 

Besides the winter movements of the trumpeters in the Madison 
Valley northwest of Yellowstone Park, birds occur there as far north 
as Ennis, Montana, during the spring, summer, and fall months. A 
pair or two are usually found nesting on Ennis Lake. 

Trumpeters have been reported at various places along the Beaver- 
head Valley as far north as Twin Bridges, Montana, over 100 airline 
miles northwest of the Red Rock Lakes Refuge. A pair of breeding 
swans have occasionally been reported near Twin Bridges; however, 
this report has never been verified. A number of late fall and early 
spring occurrences of trumpeters near Dillon, Montana,, have been 
reported in recent years by Joseph H. Buck, former Red Rock Lakes 
Refuge employee. 

During the winter months the occurrence of a trumpeter or two 
along the Yellowstone River outside the Park between Gardiner and 
Livingston is also occasionally observed by Park Service personnel; 
however, such sightings indicate only sporadic use of that area. 

More recent records are given under Annual Swan Census, 1929-57. 

Nebraska. During the 1956 fall hunting season a trumpeter family 
of 5, which included 3 cygnets banded in the Grande Prairie region of 



RECENT OCCURRENCE, BREEDING, AND MIGRATION REPORTS 37 

northwestern Alberta (Lowe Lake), were all shot in western Ne- 
braska. Two of the cygnets were killed on October 27— one at 
Schoolhouse Lake and one at Shoup Lake near Valentine (Cherry 
County)— while the third was found dead November 2 on the Loup 
Eiver 12 miles west of Fullerton (Nance County), having previously 
been shot. One of the adults was shot and crippled October 30 
near Fullerton while the other adult was captured wounded near 
Shelton (Buffalo County) on November 2. The crippled adults were 
taken to Grand Island where they could be cared for. No other re- 
ports for this species in Nebraska appear to have been made since the 
November 11, 1929, Holt County record (Haecker, Moser, and 
Swenk, 1945: 5). 

North Dakota. Information received from J. F. Cassel, Chairman, 
Department of Zoology, North Dakota Agricultural College, indi- 
cates that a pair of trumpeters in company with 125 whistling swans 
were observed on Slade Lake near Dawson, North Dakota, by Lee 
Pettibone on April 25, 1928. 

Russell Reid, Superintendent of the Historical Society of North 
Dakota, wrote (correspondence), "During October 1930, I observed 
two swans flying over Lake Isabel south of Dawson, North Dakota. 
These swans appeared to be exceptionally large but . . . identifica- 
tion could not be positive." 

There do not appear to be any reports from this State in the last 
25 years. 

Alberta. The presence of a breeding group of trumpeters in north- 
western Alberta in the Grande Prairie district has been known for 
some time. J. D. Soper (1949 : 240) reported 64 adults and 14 cygnets 
in this area. These birds have been under seasonal observation by 
the Canadian Wildlife Service since that time. 

A breeding pair of trumpeter swans was reported in the Cypress 
Hills region of southeastern Alberta by Robert Lister (1951: 157). 
These were observed without young in 1948 but were accompanied 
by cygnets in 1949 and 1950. These presumably winter somewhere 
in the United States. 

MISSISSIPPI AND ATLANTIC FLYWAYS 

A field report (U. S. Fish and Wildlife Service files, Patuxent 
Refuge) by B. J. Shaver dated August 31, 1937 noted two trumpeter 
swans on a small marsh lake in Beltrami County, Minnesota. These 
birds were reported to have been there all summer, but no yoimg birds 
were seen. No other records show trumpeter occurrence east of Al- 
berta or North Dakota. 



HABITAT 




i^'iW 



BREEDING HABITAT 



LIFE ZONE CHARACTERISTICS 

The trumpeter swan originally nested over a wide latitudinal range, 
roughly 1,700 miles at the greatest distance. Although the prin- 
cipal breeding grounds were reported to be mainly within the north- 
ern portions of the continent, it is apparent that this fowl formerly 
nested from at least as far east as the eastern shore of James Bay 
west to coastal Alaska. This species thus once occupied a variety of 
different ecological environments. 

Established breeding records of the trumpeter document nesting 
in the following life zones of North America : * 

1. Arctic-Alpine Zone (Mackenzie Bay, Norton Sound). 

2. Open Boreal Forest Zone (lower Mackenzie Basin, Kenai 
Peninsula) . 

3. Closed Boreal Forest Zone (upper Mackenzie Basin, Grande 
Prairie region, Yellowstone region [including Red Rock 
Lakes] ) . 

4. Aspen Parklands Zone (Cypress Hills, Alberta). 

5. Montane Pine Zone (Yellowstone region). 

6. Pacific Rainforest Zone (Naha Valley, Alaska). 

7. Eastern Deciduous Forest Zone (northeastern and northwestern 
Missouri, south central Minnesota, southern Wisconsin). 

8. Short Grass Prairie Zone (Flathead Valley, Montana; north- 
western Nebraska). 

9. Tall Grass Prairie Zone (northern Iowa, southern Minnesota). 



1 Life zone names based on unpublished information in U. S. Fish and Wildlife Service 
flies. 



38 






BREEDING HABITAT 



39 



While the range of life zones occupied by various groups of breed- 
ing trumpeters in the past has been great, the fur-trade records would 
seem to confirm that the trumpeter has been found breeding more 
typically in the Open Boreal Forest than in any other life zone. As 
the species existed near the southern limits of its breeding range in 
the United States, it was to be found nesting chiefly in the Closed 
Boreal Forest, Montane Pine, Eastern Deciduous Forest, Short Grass 
Prairie and Tall Grass Prairie Life Zones. 

PHYSICAL CHARACTERISTICS 

Although trumpeters originally lived in many different major 
groups of environments, as embraced by the life-zone concept, there 
is much evidence that this species is far more limited in the variety 
of habitat it will accept as actual breeding grounds. Its ecological 
niche may therefore be said to be as confined as its life zone range 
was generous. This characteristic would be expected in such a spe- 
cialized waterfowl. 




-fc breeding populations 

. wintering range 
Q breeding and wintering range 



Figure 5. — Presently known breeding and wintering range, trumpeter swan. 
469660 O — 60 4 



40 



HABITAT 




J^_Z 



Figure 6. — Red Rock Lakes Migratory Waterfowl Refuge. 

The following statement outlines some of the specific physical fea- 
tures of the trumpeter's breeding-habitat requirements: 

1. Stable waters possessing a relatively static level, not exhibiting 
marked seasonal fluctuations. 

2. Quiet waters of lake, marsh, or slough, not waters subject to 
obvious current or constant wave action. 

3. Shallow waters of lake or open marsh, not so deep as to pre- 
clude considerable digging and foraging for lower aquatic 
plant parts, roots, tubers, etc. 

RED ROCK LAKES REFUGE 

Perhaps the best w T ay to outline the characteristics of the trumpeter's 
present breeding habitat in the United States is to describe the 
principal nesting grounds of this species in the Red Rock Lakes 
Migratory Waterfowl Refuge. This area was acquired by the United 
States Government in 1935 and is now administered by the Bureau of 
Sport Fisheries and Wildlife principally for the perpetuation of this 
species. This splendid mountain-marsh system is located in the 
Centennial Valley (Beaverhead County) in southwestern Montana. 
Comprising about 13,000 acres of shallow lakes, productive open 
marsh, and extensive sedge meadows, it is a biotic complex not dupli- 
cated elsewhere in this country on such a grand scale. Its elevation 
of 6,600 feet, the relatively stable water supply which flows from 
numerous springs and creeks sustained by a dependable snow run-off, 
plus the exceptionally gradual gradient of the entire drainage are 
physical features which combine to perpetuate a stable virgin marsh 
of high quality, a rare feature of such generous proportions in these 
modern times. 



BREEDING HABITAT 



41 




Figure 7.— The Red Rock Lakes owe their stable waters to the Centennial 
Mountains which tower above them to the south, trapping abundant snows 
that feed the numerous creeks and springs entering the marsh system. 

Geologically speaking, the Red Rock Lakes in the Centennial Val- 
ley of southwestern Montana lie in a broad trough bounded on both 
the north and south sides by two major faults. South of the lakes 
a continuous series of en echelon faults form the north face of the 
Centennial mountain range. The north side of the trough in which 
the lakes lie is marked by a series of faults along the front of the 
Gravelly range; thus the lakes lie in a down-dropped basin which 
comprises most of Centennial Valley. The gradient of the valley 
floor of Centennial Valley is extremely slight, and erosion is not down- 
cutting the exit of the Red Rock Lakes at any great speed. There- 
fore there seems little likelihood that within the next few hundred 
years there will be any marked change in the lakes themselves; they 
will neither be filled in nor drained by erosion, though many higher 
stands of the lake shore can be seen around the margins of the valley. 
These higher stands of the lakes probably correspond to wetter cli- 
matologic periods rather than any marked difference in drainage, and 
may perhaps be correlated with the various advances of the Wisconsin 
ice sheet and the extensions and retreats of the valley glaciers into the 
bottom of the valley floor. 

Records of much dryer periods in the recent geologic past are also 
preserved. Immediately west of the Red Rock Lakes area are many 
square miles of anchored sand dunes, barchans ; these dunes are now T 
covered by vegetation and are nearly stationary. Their form and 
outline indicates that they were active moving dunes in the recent 
geologic past, and in a sense are now fossil dunes, testifying to a 
period of much low r er rainfall than now occurs. 




Figure 8. — Aerial view of Red Rock Lakes. Lower Lake in foreground is 
dotted with beds of bulrush. The marsh system is at the upper left, and the 
Upper Lake is in the center background. 

The Red Rock Lakes and marsh were the subject of a formal flora 
survey, with transect control, in 1955 and 1956 by Biologist Watson 
E. Beed of this Bureau. His work, together with that of other in- 
dividuals, forms the chief basis of present knowledge regarding the 
identification, distribution, and quantity of both submerged and 
emergent vegetation in this marsh. The following description is 
based on Mr. Beed's report. 

All of the Red Rock Lakes bottoms are composed of a mucky 
matter, being a mixture of decaying vegetation, plankton, and mineral 
soil. Because of the relative low water temperatures, seldom exceed- 
ing 50° F., decomposition is slow and the deposit of decaying material 
relatively greater than the annual rate of deposition would indicate. 
It is this residual fertility of the bottoms, which combines with favor- 
able water temperatures and levels during the short summer months, 
that produces the tremendous abundance of aquatic plants found in 
the Red Rock Lakes marsh. 



BREEDING HABITAT 



43 



The water itself, except for a stained condition in some of the bog 
bays, is normally clear except when disturbed by wind, feeding water- 
fowl, or by water movement during the spring run-off. The abun- 
dance of aquatic plants indicates a medium hard water with perhaps 
20 to 30 ppm. of bound C0 2 . The actual pTL of four samples taken 
was 8. 

With the exception of the north shore of the Upper Lake, which is 
subjected to considerable wave action, the shorelines of all waters are 
vegetated to th& water's edge, chiefly with beaked sedge (Carex ros- 
trata) 2 , and exposed mud flats are absent. Shorelines are open rather 
than timbered except the south shore of Upper Red Rock Lake, which 



2 Botanical nomenclature, which follows Beed's report, was based on Norman C. Fassett's 
Manual of Aquatic Plants, Albert S. Hitchcock's Manual of the Grasses of the United 
States, and W. E. Booth's Flora of Montana, Part 1, as well as identifications supplied 
by Loran C. Anderson, Acting Curator of the Intermountain Herbarium, Logan, Utah. 




Figure 9. — Trumpeter swan nest located on an old muskrat house on a cattail- 
sedge island in Lower Red Rock Lake. In the background the Centennial 
Mountains wear a snow mantle normal for June. 




Figure 10. — Numerous channels, sloughs, and potholes set in a bog-mat environ- 
ment of beaked sedge typify the Red Rock Lakes marsh. Darker shoreline 
vegetation is bulrush, cattail and rushes. 



is bordered by a fairly continuous belt of aspens (Populus tremu- 
loides). Along the higher better-grained soils, Engelmann spruce 
(Picea engelmanni) in a bog environment with willows (Salix spp.) 
along the immediate lake shore exist in certain locations about Upper 
Red Rock Lake. 

Arrowhead (wapato, Sagittaria lati folia) and spikerush (Eleo- 
charis macro stachycb) are the principal emergent plants found in the 
Lakes and marsh between the normal sedge shoreline and deeper water. 
The stable low water levels explain the abundance of these plants in 
certain locations. 

As might be expected, animal life flourishes in the water and around 
the shoreline. Frogs, toads, and polliwogs are extremely abundant 
locally in season, while snails and the fry of several fish species are 
also present in numbers. Water beetles, caddis flies, and rat-tailed 
maggots are also common in their appropriate ecological strata, while 



BREEDING HABITAT 



45 



the production of small Crustacea and plankton in maximum abun- 
dance is truly astounding, the water being literally alive during the 
summer season. 

Swan Lake is a small but important shallow marsh of roughly 400 
acres lying to the north of Upper Red Rock Lake and connected with 
it only by means of its drainage. Elk Springs Creek, which emerges 
from Swan Lake near the east end only a short distance from where 
it enters the Lake, is its principal source of water. The water table 
in this lake is very stable, though having no flow. Numerous emer- 
gent islands of spikerush and beaked sedge occur in Swan Lake, and 
several rather prominent beds of cattails (Typha latifolia) are also 
present. Water depths are extremely shallow, varying from a few 
inches to only about a foot in the deepest area of any size, and averag- 
ing about 6 to 10 inches over the entire floor of the lake bed. 





Figure 11. — A trumpeter nest located in the predominant sedge environment 
of the Red Roek Lakes marsh. Stem and leaf parts of both sedge and cat- 
tail form the bulk of the nest material both for the muskrat lodge nest 
foundation and the nest proper. 




Figure 12. — Swan Lake is a shallow marsh sealed off from Upper Red Rock 
Lake by a natural sedge-willow anchored dike. Islands and peninsulas are 
chiefly sedge bog-mat while many extensive beds of spikerush also occur in 
these stable shallow waters. 

Because the shallow, static nature of the water causes higher water 
temperatures than in the other areas of the Lakes, algae are con- 
spicuous, especially during the late summer months. In spite of algal 
shading, the lake bottom is very fertile and supports a great profusion 
of aquatic plants. The following plants predominated in Swan Lake 
in 1955 and 1956. Percentages of area occurrence are : water milfoil 
{MyriopTiyllum exaZbeseens) , 35; bare, 23; sago pondweed (Pota- 
mogeton pectinatus), 13; clasping-leaf pondweed (P. richardsonii) , 
12; leafy pondweed (P. foliosus), 7; slender pondweed (P. pusillus), 
4; and miscellaneous, 6. 

A glance at figure 6 will show that Upper Red Rock Lake is the 
largest lake on the Refuge; it is about 2,880 acres in expanse. It 
contains no "islands" of emergent vegetation, and even peripheral 
plants, such as bulrush or cattail, are not so prevalent as elsewhere. 



BREEDING HABITAT 47 

Water depths in the Upper Lake vary from a few inches to over 5 
feet, but almost all of the lake is less than 4 feet deep. The main 
water supply is furnished by Red Rock Creek, though the Elk 
Springs Creek-Swan Lake drainage, Tom Creek, and numerous fresh 
springs along the south side of the lake are also important and add 
a considerable flow, especially during the runoff season. Drainage 
of this lake is westward into the main marsh system. 

The very fertile bottom of the Upper Red Rock Lake supports an 
almost unbelievably abundant and luxuriant growth of aquatic plants. 
In 1955 and 1956 the following tabulations of species percentages were 
recorded: water weed (Elodea canadensis), 41; muskgrass (Chara 
spp.), 22; bare, 12; leafy pondweed, 5; sago pondweed, 4; and mis- 
cellaneous (chiefly Potamogeton spp.), 16. 

The marsh surrounding Swan Lake and extending between Upper 
and Lower Lakes comprises the largest single habitat unit within 
the Refuge, over 8,000 acres. Of this area only about 10 percent is 
open water, most of the balance being nearly pure stands of sedge in 
a bog meadow community. Water depths vary greatly but outside 
the main channels are uniformly shallow. Current in the main stream 
is very slow and even less perceptible when divided into more than 
one channel. Many isolated potholes are found within this river 
marsh area. Most of these are small but a few are of considerable 
size. 

The best of this river-marsh habitat, from a food-producing stand- 
point, is found along the stream beds of the slow-moving river and 
extensive shallow sloughs which border these outlet channels. Aquatic 
plants in the channels and sloughs of the main stream bed were found 
to occur in the following approximate order of abundance, with a 
high percentage of the area surveyed being covered: clasping-leaf 
pondweed, water milfoil, muskgrasses, sago pondweed, and arrowhead. 

Common emergent s in this area include at least two species of bur- 
reed (water burreed, Sparganium fluctuans, and S. muJtipeduncula- 
tum), hardstem bulrush (Scirpus acutus), and cattail. Ecologically, 
this unit is the most diversified of any on the Refuge, with the great- 
est variety of plant life. 

The Lower Red Rock Lake is 1,540 acres in size and is supported 
chiefly by the dependable waters of Odell Creek, which arise in the 
mountains to the south. Several permanent sedge islands are found 
near the north side of this lake, while many prominent beds of bulrush 
are scattered over the lake at large. Water depths vary generally 
from 1 to 2 feet, placing this lake in an intermediate position between 
shallow Swan Lake and the moderately deep Upper Lake. The lake 
bed was covered by plants in the following percentages : waterweed, 



48 HABITAT 

39 ; bare, 19 ; algae, 12 ; clasping-leaf pondweed, 9 ; arrowhead, 5 ; sago 
pondweed, 2 ; and miscellaneous species, 14. 

Several species of rushes (Juncus balticus, J. parryii, J. longistylus, 
etc.) and sedges (Carex festivetta, O. kelloggii, and C. laeviculmis) 
are found most frequently on firmer ground between the marsh proper 
and upland meadows. 

The northern slopes of the Centennial Mountains, which form the 
ramparts to the south of the Lakes, are covered with timber, prin- 
cipally Douglas-fir (Pseudotsuga taxi folia) and lodgepole pine (Pinus 
contorta) with a lower slope apron of aspen and willow (Salix, spp.). 
The slopes of the Gravelly Range which stretch away to the north 
of the Lakes are characterized by a more arid climate and sandier 
soil where numerous grasses and sages, principally big sagebrush 
{Artemisia tridentata), threetip sagebrush (A. tripartita), and silver 
sagebrush (A. cana) thrive. 

By far the most common waterfowl associated with the swans in the 
Red Rock Lakes marsh are the lesser scaup (A thy a afjinis), many 
hundreds of which are produced annually. Other common marsh 
nesting birds include the long-billed marsh wren (T elmatodytes pal- 
ustris), a sandhill crane (Grus canadensis) , and coot (Fulica anieri- 
cana) . 

YELLOWSTONE NATIONAL PARK 

In contrast with the Red Rock Lakes area, where the nesting terri- 
tories are contained in a single marsh system, breeding swans in 
Yellowstone Park usually exist as isolated pairs on widely separated 
waters. Without exception, each of Yellowstone's swan lakes is oc- 
cupied by only one breeding pair of birds. 

The swan lakes of Yellowstone also differ physically in many ways 
from the rather uniform Red Rock marsh nesting habitat. In the 
Park, shorelines are often timbered, feeding areas are much more apt 
to be peripheral due to deeper water areas toward the center of the 
lakes, and lake elevations are generally greater. So the breeding 
grounds of the trumpeter in Yellowstone are a more marginal habitat 
than the vast uniform marshes of the Red Rock Lakes. This is appar- 
ent in the following description, based on a letter from Condon, of 
four Park swan waters found outside the plateau region. 

The lakes of Yellowstone occupied by swan as nesting areas are 
pronouncedly different, each from the other, in their geological 
origin as well as in the general ecology of the area surrounding them. 
Trumpeter Lake, at an elevation of about 6,050 feet and with a surface 
area of about 20 acres, owes the origin of its basin to glaciation; it is 
surrounded only by a grassland-cinquefoil-sagebrush plant associa- 
tion. The waters are shallow, and one end of the lake provides cover 




* 



<*s 



m 



<?4* 




j§SO» 



Figure 13. — A trumpeter pen on her nest in a shoreline stand of pure sedge, 
Upper Red Rock Lake. The Centennial Mountains escarpment forms a chilly 
backdrop. 



50 HABITAT 

for nesting birds in the cattails, bulrushes, and sedges. A detailed 
study of the vegetation within the lake has not been made, but its 
composition is very different from the vegetation in the lakes at high 
elevations, secluded in the evergreen forests of the plateau section of 
the park. There are no large beds of wokas, (yellow pondlilies, 
Nuphar polysepalum ) . Associated with the swans on Trumpeter Lake 
are found : ruddy ducks, coots, redwinged blackbirds, yellow-headed 
blackbirds, spotted sandpipers, mallards, soras, Canada geese, long- 
billed marsh wrens, muskrats, and a variety of smaller animals. 

The Beach Spring Lagoon, with a surface area of about 29 acres, is 
off Mary Bay on Yellowstone Lake at an elevation of about 7,740 feet. 
It owes its basin to water impoundment behind a bar formed along the 
lake shore. Like Trumpeter Lake, this body of water is surrounded 
by a grassland-sagebrush vegetative complex. It does not have so 
much cover in the form of cattails, rushes, and sedges for concealment 
as does Trumpeter Lake. There are no large beds of wokas. The 
waters are shallow, and during the summer months are commonly 
visited by California gulls and white pelicans. Canada geese, buffle- 
heads, mallards, scaup, and coots are commonly seen with young on 
its waters. Marshy areas are much more extensive around this lake 
than at Trumpeter Lake. The vegetative growth in the marsh areas is 
not sufficiently tall to provide cover for swans but does provide con- 
cealment for smaller birds and mammals. 

Swan Lake, at an elevation of about 7,250 meet, owes its basin to 
glacial action. It is surrounded by a grassland-cinquefoil-sagebrush 
and sedge-marsh vegetative complex. Canada geese, coots, mallards, 
green-winged teal, scaup, goldeneyes, and buffleheads are seen with 
young on this lake. Yellow-headed blackbirds, spotted sandpipers, 
long-billed marsh wrens, and Wilson's phalaropes nest there. Musk- 
rats are common, and otters are seen at times. The waters are rela- 
tively deep with extensive marsh areas on the northern end. Sedges 
and rushes predominate on the north end. Some small clumps of 
willow are present. Wokas beds are absent. 

Geode Lake is a small lake at an elevation of about 6,150 feet with 
open rocky shores and virtually no plant cover. The waters are rela- 
tively shallow, impounded by an old beaver dam. This has silted over 
and thus established itself as a barrier which will probably retain 
water in the basin for many years to come. Very few other birds or 
animals use this lake. The absence of cover apparently discourages 
ducks and other waterfowl. There is undoubtedly overland movement 
of swans from this lake to ponds about %-mile distance under Crescent 
Hill. 

The lakes secluded in the evergreen forests of the plateaus of the 
park are, in most instances, larger in size than the open-country glacial 



/ . ' 



Figure 14.— A female trumpeter, on her nest after returning from a feeding 
period in the Lower Lake, shakes the water from her plumage. The nest is 
located on a muskrat house behind a protective screen of bulrush. Note 
elevation of nest in tall, dense cover, and discoloration of swan's head and 
neck from contact with ferrous organic matter. 

lakes. Their exact geological origin is somewhat obscure in some 
instances, but the majority of them were caused by the damming of 
old drainage basins or courses with glacial debris. Some are residual 
lakes remaining in a depressed area of a once much larger Yellowstone 
Lake. Virtually all of these lakes have beds of wokas in their shallower 
waters. In most instances there is an absence of rushes and cattails. 
This is not true of Tern Lake, Riddle Lake, or the small lake below 
Madison Junction. In some lakes beavers are present and occasionally 
their old lodges have provided a base for a nest site. Many of these 
lakes have a fringe of meadow around them consisting of grasses and 
sedges. Canada geese, mallards, goldeneyes, buffleheads, green- winged 
teal, and scaup are seen with young on most of these lakes although 
they are not present in any appreciable number. Grebes and coots 
are also found on some lakes. The sora nests at White and Tern Lakes 



52 



HABITAT 







Figure 15. — Geographical features of the trumpeter swan breeding and winter- 
ing areas in the United States. 

and also at Riddle Lake, while the common loon has been observed 
nesting at Riddle Lake in association with the trumpeter swan. 

COPPER RIVER BASIN, ALASKA 

Notes on the Copper River area were furnished by Melvin A. Mon- 
son, U. S. Fish and Wildlife Service, who discovered the trumpeters 
there in 1951. I condensed his remarks as follows : 

The Chugaeh Mountains, through which the lower Copper River passes. 
effectively divide the climate of the Copper River Basin into two categories. 
The south side of the range bordering the Gulf of Alaska has a climate maritime 
in nature with heavy precipitation and relatively mild temperature. . . . North 
of the Chugaeh Mountains the climate is colder and considerably drier. Some- 
where in between these two general categories are the climatic conditions exist- 
ing in the Bremner and Tasnuna River valleys. 




Figuke 16. — Trumpeter swan nesting site at Grebe Lake in Yellowstone National 
Park. Note exposed situation of sedge sod nest. Beds of wokas appear in 
the background, moulted feathers in the foreground. 



Because of rugged topography of the region the deep Copper River canyon 
acts as a wind funnel so that high wind velocities are common in the lower 
Bremner and Tasnuna valleys. . . . During the summer the wind blows up- 
stream . . . these strong winds pick up sand and silt from the valley floor of 
the Copper River below the confluence of the Bremner and Tasnuna Rivers and 
deposit this material as the wind divides and fans into the Bremner and Tasnuna 
Rivers. Such deposits are so great that the shallow lakes show evidence of 
fdling up. 

The Bremner River has a drainage area of 1,000 square miles and enters the 
Copper River 40 miles upstream from its mouth. . . . Only in the lower portion, 
the last 23 miles, does the river bed broaden into a relatively wide flat valley 
floor. Here there is evidence of great deposits of gravel and silt which have 
been carried down from the upper reaches. ... In this stretch, the river does 
not possess a well-defined channel. Within this lower section small shallow 
lakes are scattered through the valley floor. It is in these small individual lakes 
that considerable nesting of swans occurs. 

The Tasnuna Basin possesses characteristics similar to those described for 
the Bremner River. It has a drainage area in excess of 300 square miles and 
enters the Copper River from the west 40 miles above the mouth of this stream. 
. . . Here, as in the Bremner River, there is evidence of considerable deposits 
of gravel and silt laid down in the valley by retreating glaciers. . . . The river 
has no well-defined channel, and during high water much of the valley floor is 
flooded. Throughout the lower 10 miles are numerous shallow lakes where 
nesting swans have been observed. 

Spruce is the only valuable timber in the two valleys. It is found up to 
elevations of about 2,.">00 feet above sea level. Above the timberline the moun- 
tain slopes are covered with a dense growth of alder. . . . 

In the flat valley floors there are abundant growths of both willow and alder, 
which appear to be the dominant species. Also scattered throughout the area are 



54 



HABITAT 




Figure 17. — Aerial view of trumpeter nesting habitat in lower Tasnuna River 
Basin, Alaska. Trumpeter nest was found in small restricted slough in the 
lower right-hand corner of photo. 

limited stands of Cottonwood and birch. Grasses are abundant throughout 
much of the area, and in shallow lakes and along the lake shores in the valley 
floors there are luxuriant growths of horsetail (Equisctum spp. ). 

WINTERING HABITAT 



Within the greater Yellowstone region of southwestern Montana, 
northwestern Wyoming, and northeastern Idaho, there are five prin- 



WINTERING HABITAT 55 

cipal swan wintering districts. These are listed below in the order 
of importance, the first two being more vital to the United States 
segment of the continental trumpeter population than all the rest 
put together: 

1. Island Park area, which includes Henrys Fork of the Snake 
River and its upper tributary waters (Idaho). 

2. Eed Rock Lakes Migratory Waterfowl Refuge (Montana). 

3. Yellowstone National Park (Wyoming). 

4. National Elk Refuge, Jackson Hole (Wyoming). 

5. Madison River, and its tributary waters above the Meadow 
Lake Dam (Montana). 

All of the areas listed contain shallow-water lake, stream, and pond 
habitat with varying amounts of aquatic vegetation. Because of 
warm springs these waters do not freeze over entirely during the long 
periods of cold winter weather which normally prevail. Forays by 
trumpeters outside the greater Yellowstone region described are occa- 
sionally observed, though probably these are chiefly exploratory 
flights. Most, if not all, of the trumpeters inhabiting the United 
States are believed to winter on suitable waters within the limits of 
the districts outlined. 

These wintering grounds may be roughly divided into two mam 
groups, being either spring- fed streams, or lakes and ponds which also 
receive warm water from some source. So far as streams are con- 
cerned, water movement alone is a factor of considerable importance 
in keeping such waters open during moderately cold weather, but some 
source of warm water is a necessity because prolonged periods of cold 
are common during the winter, with daily minimum temperatures 
well below zero. In such weather, water movement alone will not 
keep solid ice from forming completely across fairly active cold rivers. 
This role assumed by warm-water springs in providing winter swan 
habitat in the greater Yellowstone region is all- important. For in- 
stance, on the main stem of Henrys Fork of the Snake River the 
abundant warm waters of Osborne, Harriman, Elk, and Big Springs 
are essential. On the Madison River within Yellowstone Park the 
warm waters from numerous geysers and springs collect in the famous 
Firehole River to keep both it and the Madison River free of solid ice 
for many miles downstream, and numerous other examples could 
be cited in the Park. Warm spring impoundments on the Red Rock 
Lakes Refuge provide the only swan wintering habitat in that area. 
Without exception the numerous large warm springs within the 
greater Yellowstone region are primarily responsible for whatever 
winter waterfowl habitat is available ; without the effect of their com- 
bined warm flow, the trumpeter as well as thousands of lesser water- 
469660 o — «»<> 5 



56 HABITAT 






fowl would be forced to migrate elsewhere, and this whole section of 
the Rockies would be nearly barren of wintering water birds. These 
warm springs provide wintering habitat for trumpeters conveniently 
near to suitable breeding grounds, a fact principally responsible for 
saving this species from extinction in the United States, sparing 
them the long dangerous migrations down heavily gunned flyways. 

Besides open water, good swan wintering habitat contains a certain 
amount of level and open terrain allowing these large birds to loaf or 
fly without restriction of visibility or movement. On the smaller 
streams this becomes especially important since the air space over 
such water is limited, and trumpeters, perhaps more than any other 
waterfowl, require ample and unrestricted air space for take-oft'. Too, 
the presence of timber growing thickly along watercourses or around 
spring ponds provides convenient perches for avian enemies (eagles), 
cover for mammalian predators, and formidable obstacles to flight. 
Unobstructed snowfields on meadows adjacent to open streams or 
ponds are regularly used as loafing sites, especially later in the winter 
when the snow hardens with settling. During this season trumpeters 
are prone to convene in large flocks and become more active socially. 

Because of these factors, the heaviest use of even a comparatively 
large open stream occurs in those areas which are not timbered or con- 
fined to a narrow canyon, and those portions of waters so restricted, 
even though containing an abundance of food at easily available levels, 
do not support their proportionate share of use. 

ISLAND PARK 

The waters of Henrys Fork of the Snake River, together with those 
of its upper tributaries, arise in the Island Park country (Fremont 
County) of northeastern Idaho along the west boundary of Yellow- 
stone National Park. Here the terrain, physical characteristics of the 
stream, and warm waters from many contributing springs combine 
to produce hundreds of acres of shallow and productive riverbed and 
pond habitat, some of which does not freeze over even in the coldest 
weather. 

The heart of this wintering area is located on or adjacent to the 
Railroad Ranch. This ranch with its adjoining lands contains the best 
and most intensively used trumpeter wintering area for its size on the 
continent. Old timers say that originally the preferred swan winter- 
ing area in this district was located on Shotgun Creek several miles to 
the north. When the desirable features of that area were eliminated 
by the completion of the Island Park Reservoir Dam in the 1930's, 
which flooded this Creek, the swans were forced to rely more heavily 
on the waters of Henrys Fork proper. (Likewise according to old 
residents the impoundment of Jackson Lake Reservoir in Jackson 



WINTERING HABITAT 



57 







— -a** 




Ht' 



K 



*«» 



•* 



Figure 18. — Aerial view of Henrys Fork (North Fork) of the Snake River 
below the Railroad Ranch, Island Park, Idaho. This stretch of river offers 
habitat to wintering trumpeters. The warm Harriman Springs keep these 
waters open even below — 30° F. 

Hole destroyed open-water areas caused by warm springs. This for- 
merly constituted a wintering area of considerable importance to 
trumpeters.) 

In the vicinity of the Railroad Ranch, the Henrys Fork is a mod- 
erately large, clear, shallow stream of relatively stable flow which 
meanders through open meadows. Several large springs and spring- 
fed tributaries provide the necessary warm water in strategic loca- 
tions to keep at least some stretches of the stream open even in the 
severe winter weather. Beds of marestail (Hippuris vulgaris), leafy 
pondweed, and sago pondweed cover the stream bed in profusion, and 
with clasping-leaf pondweed no doubt form the bulk of the trumpeters' 
winter diet in this area. 



58 



HABITAT 




Figure 19. — Aerial view of trumpeters wintering on Henrys Fork of the Snake 
River below the Railroad Ranch. Gray birds are cygnets-of-the-year. Trum- 
peters from the Grande Prairie, Alberta, region have wintered in this area. 

RED ROCK LAKES REFUGE 

Wintering habitat on the Red Rock Lakes Refuge is confined to two 
warm spring impoundments, MacDonald Pond, where the multiple 
Elk Springs furnish a plentiful supply of warm (58° F.) water, and 
Culver (Widow's) Pond where the equal but colder (41° F.) flow of 
the dual Picnic Springs is confined. At these two areas, located about 
2 miles apart at the east end of the Refuge, a combined area of from 5 
to 10 acres is normally open during the winter. Although these waters 
never freeze over entirely, the average open-water area may be reduced 
by half or less during the prolonged occurrence of —30° F. or colder 
nightly temperatures. 



YELLOWSTONE NATIONAL PARK 

Although regular counts are not taken owing to winter isolation 
factors, five main areas in Yellowstone Park are known to be regularly 
frequented or occupied by wintering trumpeter swans. A brief de- 
scription of these winter habitats follows : 

1. The Yellowstone River from the outlet of Yellowstone Lake north 
to its junction with Alum Creek normally furnishes winter quarters 
for a number of Park trumpeters as well as groups of Canada 
geese, mallards and gokleneyes. Various areas along about 2 miles 







Figure 20. — Aerial view of 80 trumpeters in east Culver Spring, Red Rock Lakes 
Refuge, January 1950. Air temperatures —20° F. Note moose tracks in 
willow growth. 

of this stream furnish natural aquatic food at available depths as 
they remain open owing to thermal activity. 

2. The Firehole and Madison Rivers offer dependable food supplies 
in open waters for the greater distance of their existence in the Park. 
A number of swans are regularly observed on the courses of these 
streams, even in midwinter, along with a number of mergansers, 
goldeneyes, mallards, and Canada geese. 

3. Shoshone Lake geyser basin contributes enough warm water to the 
west bay of that lake to create a 10-acre expanse of open water there. 
Here a pair or so of trumpeters have been known to pass at least 
part of the winter, sometimes in company with a few Canada geese. 

-1. A 5-acre expanse of water normally remains open at the north end 
of Heart Lake near the entrance of Witch Creek. These waters 
usually support a pair of swans during most of the winter along 
with a few goldeneyes. 

5. At the south end of the Park the Snake and Lewis Rivers, plus the 
open waters of Polecat Creek together with an adjacent slough, 
provide open water and aquatic food at an available depth to the 
small flock of trumpeters normally found wintering in this area. 
Mallards, geese, and mergansers are also found here with the 
trumpeters during this season. 



60 HABITAT 

NATIONAL ELK REFUGE 

Several generous warm springs on the National Elk Refuge assist 
in keeping Flat Creek open during the winter months. This habitat 
is located for the most part within the Refuge and but a short distance 
from the town of Jackson, Wyoming. It has witnessed a rather re- 
markable increase in the numbers of trumpeters which have wintered 
here during the past decade. Together with other local warm spring- 
fed water areas in Jackson Hole, the National Elk Refuge promises 
to become a wintering area of major importance to the trumpeter in 
the future, particularly if some development of warm water areas 
could be accomplished without defeating other wildlife objectives. 

MADISON RIVER 

This drainage, below Park boundaries but above Ennis Lake Dam 
and including Cliff and Wade Lakes, normally winters several small 
flocks of trumpeters. These groups trade back and forth within the 
area or visit adjoining areas as the occasion demands. Specific areas 
containing winter habitat in this district are O'Dell Creek near Ennis, 
Montana, the upper Madison River near the head of Hebgen Reservoir, 
and spring- fed portions of Wade and Cliff Lakes located to the west 
of the Madison Valley proper. 

WINTER COUNTS 

Because of the poor winter flying conditions and the isolated char- 
acter of the country, no true aerial census over the entire greater 
Yellowstone region has been made during the midwinter months. 
Fairly complete aerial or ground counts have been made as the oppor- 
tunity arose in some of the districts listed. These observations have 
been made by rangers of the National Park Service on winter patrol, 
biologists of the Idaho Fish and Game Department, U. S. Fish and 
Wildlife Service personnel, and other local observers. 

These count data are presented in table 4 in order to provide an 
index of relative use. 

Small numbers of whistling swans have been observed (voice iden- 
tification) during the midwinter months in the Island Park area. 
I have heard whistling swans on winter patrol trips in the vicinity 
of the Railroad Ranch, and Ed Kroker, foreman of the Ranch, has 
confirmed the regular occurrence of a few wintering birds in this area. 
The maximum number of the lesser species which has been noted 
is 8. These were reported by Frank Kennedy, winter keeper of the 
Elk Springs Ranch in the Island Park area. So although whistlers 
have been observed to winter in this area regularly, they are appar- 



WINTERING HABITAT 



61 



fently never present in great numbers and the data in table 4 are there- 
fore believed to represent trumpeters almost entirely. 

Table 4.— Winter swan counts, Greater Yellowstone region, 1950 to 1957 



Time of observation 



Swans observed 



Count method 



Observer 



sland Park 

(Idaho) 
Feb. 3-7, 1950. .. 



Feb. 7, 1951 _ 
Jan. 10, 1952 _ 



Area 



Feb. 2, 1953. 



Jan. 6, 1954. 



Jan. 6, 1955. _ 
Feb. 16, 1956. 
Jan. 10, 1957- 



Red Rock Lakes Refuge 
(Montana) 2 
Dec. 1949-Mar. 1950 



Dec. 1950-Mar. 1951 

Dec. 1951-JVIar. 1952 

Dec. 1952-Mar. 1953 

Dec. 1953-Mar. 1954 3 

Dec. 1954-Mar. 1955 

Dec. 1955-Mar. 1956 

Dec. 1956-Mar. 1957 

National Elk Refuge 
(Wyoming) 4 
Season 1949-50 



Season 1950-51 

Season 1951-52 

Season 1952-53 

Season 1953-54 

Season 1954-55 

Season 1955-56 

Season 1956-57 

Madison River Drainage 
(Montana) 
Dec. 1955 



Jan. 1956- 
Jan. 1956. 
Jan. 1956. 



Yellowstone National 

Park « 

Jan. 10-13, 1950 

Jan. 8-12, 1951 

Jan. 1952 

Jan. 13-15, 1953 

Jan. 3-18, 1954 

Jan. 7-18, 1955 

Jan. 9-13, 1956 

Jan. 4-9, 1957 

(Madison River Only) 



262(208)' 



257 

330 (222)i 



356. 



419 (333) • 

271 (141)1 
318 (288)i 
323 (250)i 



105-80-200-150. 



31-47-100-120... 

18-50-50-150 

35-61-75-100-— 
12-90-146-250.... 
147-140-135-250. 
55-154-325-280-. 
?-95-209-230 



10. 



13 

13 

13 

24 

30 

33 

34 _ (56)" 4 



30-50 (Cliff Lake). 

20-30 (Cliff Lake). 
6-8 (Wade Lake).. 
11 (O'Dell Creek) . 



52 

No data. 

16 

33 

56 

18 

14 

20 



Aerial . 



.do. 
.do. 



.do. 



Aerial and ground 

Aerial 

....do 

do 



Ground . 



..do... 
..do... 
..do... 
..do... 
..do... 
..do... 
..do... 



.do. 

_do. 

_do. 
.do. 
.do. 
.do. 
.do. 
.do. 



.do. 

.do. 
.do. 
.do. 



.do. 
.do. 
.do. 
.do. 
.do. 
.do. 
.do. 
.do. 



Shaw and Monaghan, Idaho 
Fish and Game Department. 
Do. 

Salter, Idaho Fish and Game 
Department. 

Misseldine, Shaw, and Nielson, 
Idaho Fish and Game De- 
partment. 

Salter, Idaho Fish and Game 
Department. Cromwell, U. 
S. Fish and Wildlife Service. 

Bizeau and Bross, Idaho Fish 
and Game Department. 

Bizeau, Idaho Fish and Game 
Department. 

Bizeau and Kaster, Idaho 
Fish and Game Department. 



U. S. Fish and Wildlife Service 
station personnel. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 



Do. 

Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 



Philip L. Wright, Montana 

State University. 
Monte Neely, local resident. 

Do. 
H. W. Baker, U. S. Fish and 

Wildlife Service. 



Unknown. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 
Do. 



i Number in parenthesis represents numbers of swan on or adjacent to Henrys Fork through the Rail- 
road Ranch or the immediate vicinity and upstream to I. P. dam. 

2 Highest count or estimate of numbers of birds seen during December, January, February, and March, 
in that order. 

3 MacDonald Pond winter-habitat development completed fall of 1953. 

4 Figure corresponds to maximum number cf swans seen about Flat Creek feeding ground during winter. 
Average number accommodated each year is probably only slightly lower except for winter of 1956-57, 
when maximum number was present only a short time, hence is shown in parenthesis. Some of the maxi- 
mum number shown in parenthesis may have been whistlers, though none were identified as such at the 
time. Identification of all individuals in entire flock of 56 was not possible. 

5 No concerted effort has been made to gather wintering trumpeter data in the Park. These figures 
represent swans seen while conducting the annual winter waterfowl inventory. 




LIFE CYCLE 



DESCRIPTION 

The large size and genera] waterfowl conformation, white color, 
and prominent long necks of our native swans identify them whether 
seen on the water or in flight. Except for the occurrence of feral mute 
swans {Cygnus olor) along the eastern seaboard in the general vicinity 
of the lower Hudson River Valley or in Michigan (Grande Traverse 
County and vicinity), only the two native North American species are 
likely to be encountered in the wild. A detailed description of the 
external appearances of both species follows. For comparison, a 
number of weights and measurements of mute, whooping, and 
Bewick's swans can be found in Hilprecht (1956: 51-54). He also 
describes the appearance, flight, and voice of these foreign swans 
(1956: 19-31). From these we may conclude that the trumpeter is 
the largest swan in the world. 

SPECIES DESCRIPTION 

Trumpeter swan, Olor buccinator (Richardson), adult (sexes 
alike) : entire plumage white, head and neck commonly with a rusty 
stain from ferrous waters; iris brown; bill black, rarely with small 
grayish or yellowish spot immediately posterior to nostril ; bill usually 
longer and broader terminally than in Olor columbianus; front edge 
of nostril usually 50 mm. or more from tip of eulmen ; four outer 
primaries emarginated terminally; and feet usually black or gray but 
sometimes tinged with brownish, yellow, or olive. 

Juvenile (sexes alike): Gray Phase (common). Brownish-gray 
especially on the head, neck, and upper back, lighter gray ventrally; 
forehead, crown, occiput, nape, and upper cheeks light rufescent 
brown ; plumage sometimes rust-stained as in the adult ; feet yellowish 
or olive gray-black; bill becoming black but with basal portion of cul- 
menary ridge behind nostril still salmon or light pink color; tomia of 
mandible dull flesh color. White Phase (rare). White down of 
young replaced directly by white feathers, identical to adult. Color 
of feet and bill as in gray phase. 
62 



DESCRIPTION 



63 



Downy Young (sexes alike) : Gray Phase (common). Head and 
neck uniformly mouse-gray ; body mouse-gray dorsally, lighter gray 
to white ventrally; feet yellowish; bill pinkish basally, dark gray 
terminally. White Phase (rare). Entire plumage white, feet yel- 
lowish, bill flesh colored. 

Adult Male: 1 Wing 545-680 (618.6) ; tail 173-191 (182) ; culmen 
from tip of frontal feathering 104-119.5 (112.5) ; tarsus 121.5-126 
(122.9) ; middle toe without claw 135-145 ( 141.1 mm.) . 2 [Adult male 
(Kenai) : total length 59 in., wingspread 87.5 in., weight 27 lb. 6 oz.; 
(Yellowstone Park) : total length 60 in., wingspread 96.5 in., weight 
27 lb. 9 oz.] 

Adult Female: Wing 604-636 (623.3) ; tail 185-207 (196) ; culmen 
from tip of frontal feathering 101.5-112.5 (107) ; tarsus 113-128.5 
(121.7); middle toe (w/o claw) 138.5-148 (143.3 mm.). 3 [Adult 
female (Yellowstone Park) : total length 58 in., wingspread 74 in.] 

For comparative purposes the external appearance of the whistling 
swan is also presented here : 

Whistling Swan, Olor columbianus (Ord), adult (sexes alike) : 
Entire plumage white, head and neck sometimes with a rusty stain 
from ferrous waters; iris brown; bill black when not with usual yel- 
low or orange-yellow spot in front of the eye, front edge of nostril 
usually less than 50 mm. from tip of culmen; four outer primaries 
emarginated terminally ; feet black or gray. 

Juvenile (sexes alike) : Entire plumage ashy gray, usually dark- 
est on the head and palest on the ventral portions of the body, some- 
times plumbeous to sooty, brownish instead of pale gray due to stain- 
ing from ferrous waters, bill basally flesh-colored with the nail and 
gape border black ; iris hazel ; tarsi and toes flesh color, livid to dusky. 

Downy Young (sexes alike) : Plumage white, tinged with car- 
tridge buif to ivory yellow especially on the head, neck, and breast; 
bill, tarsi and toes yellowish. 

Adult Male: Wing 501-569 (538); tail 162-181 (170.8) ; culmen 
from tip of frontal feathering 97-107 (102.6); tarsus 105-117.5 
(111.9) ; middle toe without claw 120-133 ( 126.4). 4 

Adult Female: Wing 505-561 (531.6) ; tail 146-186 (165.3) ; cul- 
men from tip of frontal feathering 92.5-106 (99.9) ; tarsus 99.5-115 
(107.2) ; middle toe without claw 110-126.5 (118). 5 



1 Measurements in millimeters, giving the smallest and largest of the birds examined, 
with tlie average in parentheses (25.4 mm. = l inch). 

2 Five specimens, from Idaho, Wyoming, Wisconsin, and Michigan. 

3 Three specimens, one from Montana. 

1 Eight specimens from Alaska, Maryland, Virginia, and North Carolina. 
3 Fifteen specimens from Alaska, California, Maryland, Virginia, North Carolina, and 
captivity. 



64 



LIFE CYCLE 



Table 5. — Overlapping weights and dimensions of small trumpeter and large 

whistling swans 





Olor buccinator—Minimum weights 


Olor columbianus— Maximum 


weights 




and measurements 


and measurements 










From tip 








From tip 




Age and sex 






of bill to 


From tip 






of bill to 


From tip 




Number 


Body 


anterior 


of bill to 


Number 


Body 


anterior 


ofbillto 




of speci- 


wt. 


edge of 


axis of 


of speci- 


wt. 


edge of 


axis of 




mens 


(lbs.) 


nostril 


eyes 


mens 


(lbs.) 


nostril 


eyes 








(mm.) 


(mm.) 






(mm.) 


(mm.) 


2+ vears: 


















Male. -.- 


8 


20 


50 


140 


7 


19.5 


48 


125 


Female 


14 


l(i 


50 


133 


21 


19 


49 


138 


1+ years: 


















Male.. 


8 


18 


47 


131 


2 


17 


41 


123 


Female 


4 


15 


50 


135 


7 


17 


44 


118 



EXTERNAL APPEARANCE 

If certain external characteristics are lacking or unobservable, it is 
easy to see why it is virtually impossible to distinguish these species 
positively without a postmortem examination. Richardson pointed 
out that the trumpeter tracheal route along the sternum detoured 
dorsally into the body cavity whereas the tracheal routes of other 
closely related swans did not. This detail is also naturally expressed 
in the pertinent adjacent portions of the trumpeter's anatomy, such 
as the sternum and furculum where structural modifications are neces- 
sary to accommodate the really unique dorsal tracheal detour which 
extends so prominently into the body cavity. More lately, Condon 
(MS) has focused attention on the syrinx and has been able to show 
significant differences between the trumpeter and whistling swans in 
the specimens he studied. The empirical anatomical differences be- 
tween these two species are shown graphically in figures 21 and 22. 
Since a diagnostic autopsy may be impractical from many standpoints, 
even with the specimen in hand, the means of separating trumpeters 
from whistlers solely by external means will be explored briefly here. 

The search for a valid external characteristic other than voice has 
perplexed both the field observer and the systematic scientist for well 
over a century. 

Some of the points used in the past for species differentiation are 
taken up in some detail here to provide an understanding of the facts 
involved. On this basis a positive solution of the problem may some 
day be worked out. 

With the exception of the voice criterion, 3 methods of identifica- 
tion based solely on external characteristics have been used in the 
past. These are the tail-feather count, bill coloration, and the meas- 
urements of various bill features, the last method being dependent 
upon size. A brief discussion of these 3 methods follows. 



DESCRIPTION 



65 



Tail-feather count. Many writers picked up John Richardson's 
remark that the trumpeter possessed a tail of 24 feathers, and have 
perpetuated a method of identification of such obviously limited value 
that it should, in my opinion at least, never have received more than 
passing attention in the first place. This was probably all that Rich- 
ardson intended anyway, judging by his casual note. Leonhard Stej- 
neger's early statement on this method should have pointed out once 
and for all the negative value of tail-feather counts when positive 
species determination is desired. He stated (1882 : 216-217) : 

It has often been stated as a good criterion that buccinator has twenty-four 
tail feathers in contradistinction to columbianus, which only has twenty. Inde- 
pendent of the inconvenience of this character, when the birds moult their 
rectrices. I may confess that I only in a few cases have been able to count 
twenty-four tail feathers ; and the inconstancy of the number of these feathers 
I have found pervading the whole group, this character changing individually, 
so that it is not at all to be depended upon. 

Bill coloration. Most of the references of previous writers re- 
garding this method of speciation refer to the area of the upper man- 
dible (bill) immediately in front of the eyes, known as the "lores." 
Though the whistler is usually characterized by a yellow or orange- 
yellow spot on the lores, apparently this color is sometimes lacking and 
the entire bill is black. 

On the other hand, the bill of the trumpeter is almost invariably 
black, though again this rule is not absolute. Dr. Ray Erickson, U. S. 
Fish and Wildlife Service biologist formerly at the Malheur Refuge, 
said that trumpeters kept in captivity at that Refuge sometimes ex- 
hibited an olive-yellow spot in the loral region. Apparently the 
abrasive or scuffing action given the bill when rooting out food from 
comparatively hard pond banks caused the underlying color to appear. 

In 1957, I noted that the wild trumpeters at Red Rock Lakes some- 
times exhibit a small indistinct gray spot of irregular shape behind 
the nostril. In two cases this was tinged with yellow, and in one 
instance the yellow could be discerned with the naked eye from a dis- 
tance of over 50 feet. This specimen was collected and confirmed as 
a trumpeter by a postmortem examination of the sternum. 

In conclusion it may be stated that while a completely black bill 
can represent either species, a prominent bright yellow or orange- 
yellow spot on the lores indicates a whistler. 

F. H. Kortright (1943: 77) raises another point in bill coloration, 
writing of the trumpeter, "bill, black, with narrow salmon-red streak 
on edges of mandibles, lacking in Whistling Swan." My experience 
not. only demonstrated that the salmon-red streak on the trumpeter 
was confined almost wholly to the basal section of the lower mandible 
edge, but that there was great variation in the degree to which it is 



LIFE CYCLE 




Figure 21. — Trachea and sternum of whistling swan. 



present. This variance ranges from the quite prominent "grinning 
streak" normally seen to a few very faint specks of vestigial salmon 
coloration which could be observed only upon minute and critical ex- 
amination, the preponderant color for all practical purposes being 
completely black. 

Furthermore, I have noted that some whistling swans complete 
with prominent yellow lores also exhibit the characteristic salmon- 
red "grinning" streak along the basal segment of the lower man- 



DESCRIPTION 



67 




Figure 22. — Trachea and sternum of trumpeter swan. 



dible edge, though not enough specimens have been examined in this 
regard to furnish a general rule. So, for all practical purposes, the 
salmon color usually present on the dorsal portion of the basal edge 
of the lower mandible cannot furnish a positive indicator of either 
species, since overlapping characteristics are commonly observed. 

Bill measurement. From a comparison of considerable anatomi- 
cal-measurement data on the subject, Stejneger correctly concluded 
that size alone is not sufficient to separate the two species since enough 



68 



LIFE CYCLE 



overlap exists to confuse the issue. This is true even though the 
trumpeter, the largest species of swan in the world, is on the average 
much larger than the whistler. He does state (1882: 217) : 

The position of the nostrils, those being situated more backwards in the 
Trumpeter than in the Whistling Swan, is thus the only mark by which it is 
possible to express in a short diagnosis, and which I have found constant and 
easily perceptible. 

Measurements of the bills of a number of trumpeter specimens were 
made on the Red Rock Lakes Refuge during the summer molt period. 
With the cooperation of various Fish and Wildlife Refuge person- 
nel at the Tule Lake and Sacramento Refuges, comparable data 
on whistling swans were also gathered. This information is sum- 
marized in table 5. These statistics demonstrate that, owing to the 
possibility of overlap, bill measurements alone cannot provide an ab- 
solute method of separating the species in question, though with the 
proper qualifications in mind an excellent rule of thumb can be dem- 
onstrated. Of course, in order to show overlap to the best advantage, 
maximum extremes of the weights and measurements of whistling- 
swan features and minimum extremes of the weights and measure- 
ments of the pertinent trumpeter characteristics in the appropriate 
age and sex groups were those listed. 

When the voice and bill color characteristics of a given swan are 
missing or in doubt, any swan over 1 year of age, of either sex, which 
measures 50 mm. (2 in.) or more from the tip of the bill to the front 
edge of the nostril is probably a trumpeter. If the subject measure- 
ment is less than 50 mm., identification of the species as whistling 
swan is most likely to be correct. 

VOICE 

The voices of our two species of swans differ distinctly in the adult, 
doubtless owing to the extra loop of the trachea in the sternum of the 
trumpeter. Once heard, the trumpeter's call notes should not easily 
be confused with that of any other bird, least of all that of the whis- 
tling swan, whose voice resembles at a distance a high-pitched barking 
of "wow, wow-wow." Close at hand, as described by Lewis and 
Clark (Coues, 1893 : 885), who gave to this species the common name 
of whistling swan, the "kind of whistling sound" can be heard. This 
noise "terminates in a round full note, louder at the end," the dis- 
tantly audible wow. 

Trumpeters are an expressive fowl, and their voices are often em- 
ployed to show their feelings and attitudes. During the nesting and 
brooding seasons the mated pairs are fairly mute, though individuals 
in the nonbreeding flocks remain relatively communicative all sum- 
mer. Through the fall and winter seasons the vocal natures of all age 



DESCRIPTION 



69 



classes begin to be more fully expressed. During these months most 
of the swans are loosely bound into large informal flocks, and vocal 
expression is common, individually and in an occasional synchronized 
flock effort. Though they are perfectly capable of loud hissing, this 
has only been heard from cornered flightless trumpeters. 

During the months of March and April, just before the occupation 
of the breeding grounds by the restless mated pairs, vocal efforts 
reach a climax. At this season the swans on the Refuge are still at- 
tracted to the open-water spring-heads at feeding time, though the 
pull of the breeding grounds must be growing stronger daily. The 
approach of twilight finds the large flock usually resident about 
the open water announcing with loud trumpet calls each flight of 
swans as they return from their daily visits to the still ice-locked 
lakes and marshes which will soon become their summer home. 

It is difficult, if not impossible, to describe phonetically the notes 
of the trumpeter. E. H. Forbush (1929 : 305) credits E. S. Cameron 
with the Kootenai Indian name for a swan, Ko-hoh, which when 
pronounced with a gutteral intonation is a very good reproduction 
of the notes of a trumpeter swan. The call has a definite hornlike 
quality over a wide vocal range and may be uttered from one to a 
number of times, at widely spaced intervals or in staccato fashion. 
The trumpeter gives voice perhaps most often in flight but also 
commonly while on land or floating on the water. 

In general, the warning notes of the adult are sharp and terse, 
uttered infrequently; the decoy calls are longer and more apt to be 
repeated. Voices associated with simultaneous behavioral displays 
are much more conversational and when indulged in by more than 
two swans are apt to build up rapidly in participation and volume, 
finally reaching a crescendo and then ending in longer wailing notes. 
Since the range of an adult trumpeter's calls is well over a mile, the 
combined voices of noisy flocks can be heard at a distance of several 
miles if atmospheric conditions are favorable. 

Flocked trumpeters especially may be heard after dark, and in all 
seasons, particularly on moonlight nights. At such times and when 
softened by distance their wild trumpeting calls heard afar over lake 
and marsh furnish musical reminders of their wilderness world. 

If the call notes of the adults can be described as resembling a 
horn instrument, then the immature birds have the tone of a toy 
trumpet. Higher pitched and uncertain in overall quality, they 
bear the unmistakable characteristics of adolescence. By January, 
however, in place of the high-pitched fluting of the downy and post- 
downy periods, the birds utter a hoarse off-key imitation of the adult 
trumpeter. 



70 LIFE CYCLE 

PLUMAGES AND MOLTING 

Cygnets and immatures. Young trumpeters emerge as downy 
cygnets in either of two color phases, gray and white (not albino). 
In the case of the gray individuals, by far the most common, the 
head, neck, and back are mouse-gray with the underparts quite 
white. Much the same situation prevails in the mute swan (Delacour, 
1954 : 63) except that the white phase has only been reported in captive 
birds (Hilprecht, 1956: 107-8) apparently as a result of inbreeding. 

Although white cygnets have never been seen at Red Rock Lakes, 
they are regularly observed in Yellowstone Park, where broods with 
both colors occur. Condon (MS) reports: 

Of the cygnets recorded during the period 1937-40, the following count of 
gray and white ones was secured : 1937, 23 gray and 6 white ; 1938, 3 gray 
and 1 white : 1939, 16 gray and 1 white : 1940, 16 gray and 1 white. 

Thus, over a 4-year period in the Park, 13 percent of the cygnets 
were those of the white phase. Moreover, Condon has said that the 
white down in these birds is replaced by white feathers, instead of 
those of the usual brownish-gray, so that when they have reached 
flight age they can only be told from their parents by their slightly 
smaller size, pinkish bill, and yellowish legs, which are becoming 
darkened with gray. 

Only one record of white cygnets occurring in the wild outside of 
Yellowstone Park exists, that of three cygnets censused on the Ice- 
house Reservoir (Fremont County), Idaho, in 1956. 

F. E. Blaauw (1904: 73), a Dutch aviculturist who bred the trum- 
peter in captivity at Gooilust (Holland) for more than 25 years, 
provides a good general description of the normal gray-colored young 
from his day-to-day familiarity with these birds, stating: 

The chicks are white with a grey tinge on the back. The cere is covered with 
pure white down. The bill is flesh-coloured. The down of these chicks is very 
short and dense, quite different from the longer and more fluffy down of the 
chicks of Cygnus nigricollis and C. atratux [black-necked and black swans]. 
The result is that the chicks look much smaller in comparison. ... At the 
age of about six weeks, the first feathers appear, and the birds then begin to 
grow very quickly. The first feathers are brownish-grey, without any markings 
as a rule, but one of this year's birds is remarkable for having transverse mark- 
ings on the shoulders and greater wing coverts. After the birds are feathered the 
bills gradually acquire the black colour, the black beginning at the point and at 
the forehead, and gradually increasing. Later, the middle part, which is still 
pink, gets spotted with black, and in the course of the February following the 
first summer the whole of the bill usually becomes quite black. The legs by that 
time have also gradually darkened into dusky grey, which becomes black after 
the birds are a year old. About March white feathers begin to replace the grey 
plumage, except for some fine grey spots, which are still visible on the back of 
the neck and on the head. 



DESCRIPTION 



71 




Figure 23.— A pair of trumpeters on Grebe Lake. Yellowstone National Park, 
with 3 cygnets of the normal gray color phase and 2 cygnets of the uncommon 
white phase. 

Usually the pinkish color of the cygnet's bill has become predomi- 
nantly black by the time the individual is about a year old, though a 
close examination of the upper mandible will usually still show traces 
of the typical flesh-colored pigment at least until the individual's first 
flightless molt. By the time the bird has entered its second flightless 
molt, the bill is black. 

During 11)55 and 1956 Peter Ward, of the Delta Waterfowl Research 
Station, Delta, Manitoba, noted the general molting pattern of the 
young trumpeters furnished that station by the IT. S. Fish and Wild- 
life Service in 1955, writing (correspondence) : 

Although their exact age on arrival was not known, they were assumed to be 
four weeks old. this being based on our familiarity with growth rates of other 
waterfowl. This would have given them the first feathering of scapulars and 
flanks at six weeks. Body feathering was externally complete by the tenth 
[week]. The young birds seemed very conscious of the large wings and unable 
to hold them indefinitely at their sides. 

The first juvenal plumage was a dark grey-brown and remained thus until 
early in January when a gradual moult set in. Within the month this has 
changed the neck and breast to an off-white and is still in progress. Some evi- 
dence of the same change was visible in both scapular and wing coverts at this 
time where new feathers are appearing. 

Adults. All of the closely related circumpolar species of swans are 
strictly white-plumaged birds in their adult plumage, except when a 
reddish coloration has been imparted to the head and neck feathers by 
ferrous organic compounds in the marsh or lake bottom they habitually 

469660 O— 60 6 



72 LIFE CYCLE 

frequent. At the Red Rock Lakes Refuge, swans feeding principally 
in shallow water during the summer months possess this distinctive 
reddish coloration to a much greater degree than those which custom- 
arily seek their food in deeper water. 

During the winter, this coloration is much less evident on their 
heads and necks. Then the swans are feeding primarily in flowing 
spring and river waters where relatively clean (organically) rock, 
silt, or sandy bottoms prevail. Even during that season some rusti- 
ness is invariably observed upon a close examination of the feathers 
about the head, and I have never seen a wild trumpeter entirely 
devoid of the stain. With swans living principally in very shallow- 
water marsh areas, the whole neck and head and even some of the 
ventral plumage may be startlingly reddish. 

While the leg color of most adult trumpeters can accurately be 
described as black, an off-black to distinctly gray color is also common. 
Various degrees of olive-yellow pigmentation are occasionally noted 
on the skin of the legs and feet of the adult trumpeter, but only rarely 
to the extent that the whole extremity could be said to have a yellowish 
appearance. The incidence of yellow pigment was noted on the skin 
of the legs, feet, and webs of the 102 nonbreeding swans captured 
during the summer of 1956 for fluoroscopic examination. Thirty- 
four of this group had at least some subdued olive-yellow tones in 
various skin areas. Frequently the yellow coloration was discernible 
on the skin of all three parts— legs, feet, and webs— with the most 
noticeable cases commonly found in the yearling age-class, and to 
some extent in individual birds of the 2-years-or-older age group. 

Francis D. LaNoue (MS) implied that such coloration may be 
seasonal or local in nature in Yellowstone Park swans, as he records 
an unusual instance in this regard : 

Early last spring 11 swans were seen possessed with yellow legs and feet. In 
addition all three of the dead mature birds [referred to] show the same color. 
Later in the season this color was not observed. 

Immature and adult swans regularly undergo an annual molt dur- 
ing the summer period during which time they are unable to fly 
because of a more or less simultaneous loss of primary flight feathers. 
This molt most commonly occurs among the swans at Red Rock Lakes 
during the month of July, but may be completed during June in 
some cases or be delayed until August, September, or even October. 
This flightless molting period is important from a management 
standpoint since these birds can then be conveniently captured 
for examination and banding. 

Several phases of this annual molt are as yet poorly understood. 
While the flocked nonbreeders molt more or less simultaneously, nest- 



i T 



t 




«g&& 



&** 





Figure 24.— Six trumpeters circle the open water at Culver Pond on the Red 
Rock Lakes Refuge in southwestern Montana. Seldom more than 6 or 8 of 
these birds fly together in local flights unless a large flock flushes together. 

ing pairs do not usually follow such a regular molting pattern, on 
the Red Rock Lakes breeding grounds at least. Molting of the 
breeders appears to extend over a much longer period than that of the 
nonbreeders. Molted primary wing feathers are sometimes observed 
about the nest site in early June, and individuals of mated 
pairs have been captured for banding while in a flightless condition 
as late as October. During the summer of 1954, when it was desired 
to obtain pairs of known breeding capability intact with their young 
for transfer purposes, only the pen (female) and cob (male) of 6 
pairs were simultaneously flightless and could be captured. Approxi- 
mately 25 pairs were periodically checked for flight capabilities during 
the period of July 14-28 that year, and only 3 pairs were obtained on 
each of those dates. One member of each pair was able to fly in every 
other instance. 

In the following 2 years, 1955 and 1956, even poorer luck was 
experienced. The mated-pair swan transfer program was suspended 
during those years chiefly because no more than a single known pair 
could be captured during any given attempt. 

During the early part of the molting season, where only one of a 
pair was flightless it usually has been the pen, while most late-molting 
birds checked have been cobs. Peter Scott and James Fisher 
(1953:210) observed similar conditions among the wild whooper 
swans in Iceland. 



74 LIFE CYCLE 

Molting information on Yellowstone trumpeters which is contained 
in Park reports agrees generally with that from Red Rock Lakes, 
except for sex, which was not determined. The opportunity to check 
on the duration of the flightless molt of individual birds is somewhat 
greater in the Park, however, and Condon (MS) mentions in this 
respect, "The fact that some swan have been seen in flight as late as 
July 1 and as early as August 2 indicates that the interval of time 
when they are unable to fly is in some instances rather brief, probably 
not exceeding 30 to 40 days." 

Oberhansley and Barrows (MS) record two exceptionally long 
molting periods in 1938 for the pair of trumpeters on Swan Lake : 

The female remained flightless from May 14 to October 0. a period of 
148 days, while the male was not seen in flight from June 11 to October 9, a period 
of 120 days. . . . Moulting in the pen occurred largely on or near the nest 
during the time of sitting. Most of the flight feathers were found in close 
proximity to the nest. A very small quantity of down was found on the nest 
eight days after hatching. The majority of the primary feathers were shed by 
the cob at a later date, many of them on or near the resting sites on the west 
bank of the lake. 

FLIGHT 

It is difficult to imagine a more beautiful and stirring sight in the 
whole waterfowl kingdom than a typically small flight of trumpeter 
swans as they cleave the air against a wilderness setting of dark 
conifers and the rugged Rocky Mountains. With the regular beat 
of their powerful wings and long necks undulating slightly from the 
exertion of each thrust, sometimes calling in flight but more often 
silent, they usually pass directly to their destination over the shortest 
route. If they chance to pass close to the observer, the clatter of 
their great flight quills can be heard distinctly above the usual rush- 
ing sound of moving air. 

Local flights are usually low. This is especially true over familiar 
flat marsh or water where their wing tips sometimes appear to touch 
the sedge tops, or if they are over water, to pluck small jets of water 
into the air with each powerful lift of the wing. On longer flights 
over rugged mountain country, they are to be seen at much greater 
altitudes, gaining this height with considerable effort. Once they 
approach their destination, they may descend rapidly with a roar of 
cupped wings much in the usual manner of the more agile and air- 
worthy smaller waterfowl, but their descent is usually slower and more 
gradual. 

The trumpeters prefer to fly about their Rocky Mountain en- 
vironment in small flocks, apparently family groups for the most 
part. Even when a large flock of trumpeters is roused suddenly, take- 



Figure 25.-A pair of trumpeters show the 2 ways of carrying landing gear 
during flight. The normal method is by folding them back under the tail, but 
in sub zero weather cold feet may be tucked up forward and be quite mvisible 
in the warm feathers and down. 




Figure 2<>-The normal and -exerting" neck attitudes are shown by 2 trum- 
peters. Most commonly bent in this peculiar attitude during take-off. the 
neck straightens out in full flight. 



76 LIFE CYCLE 

offs are intermittent and by loosely formed groups. Upon becoming 
airborne they make their way in loosely knit small groups or flocks, 
though often to a common destination. 

Their flying formation varies greatly, sometimes being on the order 
of an offset line, sometimes nearly abreast, or otherwise irregularly 
formed in an informal staggered formation. Even a pair in flight may 
follow no particular flight pattern. I have never observed a true 
V-formation of trumpeters. This flight formation is probably em- 
ployed on certain occasions, especially on long flights when larger 
flocks would be the rule. A high V-flight of swans, apparently trumpe- 
ters, was seen by Ralph L. Hand, as previously noted. 

H. F. Witherby et al. (1939 : 169) note of the whooper swan, "When 
travelling far flocks frequently fly in oblique lines or V's, but in 
[Outer] Hebrides parties rarely adopt any definite formation/' This 
seems to apply to what is known of the trumpeters' flights as well. 
Audubon (1838:540) notes also, "If bound to a distant place, they 
form themselves in angular lines." 

To my knowledge, the flight speed of the trumpeter has never been 
recorded. Though the swans generally have been credited with most 
astounding speeds by some writers, it is doubtful whether the 
trumpeter can, without a tailwind, fly much faster than the speed 
recorded for the whistling swan by C. S. Weiser (1933: 92). Weiser 
estimated the full speed of the latter species at from 50 to 55 miles 
an hour, when tracked by a light airplane on several passes. With ei 
strong tailwind, the trumpeter can probably approach or exceed a 
ground speed of 80 miles an hour with little difficulty. 

Their legs and feet are normally carried in a streamlined manner 
tucked under the tail, nearly reaching its end. During the winter sea- 
son, if the weather is very cold, they have been observed to fold their 
feet and legs forward in flight, carrying them quite buried in a 
warm "muff" of thick breast and down feathers (figure 25). 

The feet with their great webbed area are usually employed both 
as an aid in gaining speed on the take-off and as an effective brake 
in alighting. After sufficient flight speed is gained following the 
take-off, the required distance for which varies greatly according to 
the wind, the feet are retracted into the customary position. In 
landing, the huge webbed feet are thrust prominently forward, act- 
ing like aircraft flaps to reduce flying speed and becoming effective 
hydrofoils as the swan contacts the water, skims briefly along the 
surface, then toboggans to a swimming position. 

The long neck is often curved somewhat with the great effort re- 
quired during the take-off and flight, apparently because of abnormal 
exertion. In normal flight position, however, it is carried fully ex- 
tended, undulating slightly with the beat of the wings. 




Figure 27.— Six trumpeters landing "flaps down" on Culver Pond, Red Rock 
Lakes Refuge. The feet are thrown forward before the moment of impact to 
ski the bird to a stop. Note various web positions guiding birds into landing. 

F. E. Blaauw (1904: 74) mentions, "Young trumpeter swans when 
fully fledged are very active birds. They fly with great ease, rising 
directly from the water into the air, without running over it first 
with flapping wings as so many of the large waterfowl do." I have 
never seen this occur among trumpeters of any age class, though 
Richard Rodgers, the assistant manager at the Red Rock Lakes Refuge 
for several years, reported seeing an adult trumpeter rise directly 
into the air 'from the firm bank of MacDonald Pond. This was no 
doubt possible because of the great springing power of the powerful 
legs of these birds, but it is difficult to see how a similar take-off 
could be performed from the water unless it were very shallow over 
a hard bottom. 

BEHAVIOR AND RELATED CHARACTERISTICS 

Studies of swan behavior, per se, have never been reported, and as 
a result only a few fragmentary and scattered notes are available. 
Far from being a subject of only academic interest, certain actions 
of behavior of any species can be valuable indicators of inherent basic 
requirements. This subject should be studied by the wildlife man- 
ager in order to learn both the normal characteristic expressions of 



78 



LIFE CYCLE 



the species and those made under duress so that their management 
may be guided accordingly. 

ESCAPE-DISTANCE 

This topic, borrowed from H. Poulsen as it applies specifically to 
the swans, has been generally commented upon by some observers 
under other subject headings. The term, however, is aptly descriptive 
and so is adopted here. Poulsen (1949: 196) sums up some escape- 
distance characteristics of two Eurasian species of swans, stating: 

The escape-distance ... of Whooper Swan and Mute Swan in winter in 
shore marshes was ahout 300 metres. In open lanes in the ice in winter in the 
harbor of Copenhagen the escape-distance is much less, that is they have then 
got used to human beings and have become tame. This does not influence the 
escape-distance in their breeding places. The Danish resident Mute Swans 
have a small escape-distance towards man and often breed close to habitations. 
But in the great Swedish lakes as for instance Takern the escape-distance of 
the Mute Swan is about 1,000 metres according to B. Berg (1926). The Whooper 
Swan is a more shy bird. According to Rosenberg (1946) the escape-distance 
in its breeding places in northern Sweden is about 1,500 metres. In enclosures 
the swans gradually become tame, the escape-distance gradually decreasing. 

The escape-distance of trumpeters also varies, depending upon cir- 
cumstances and the conditioning to human activity experienced by any 
given individual. Munro (1949: 712), speaking of wild trumpeters 
in British Columbia, relates : 

On many occasions in winter I have approached within 100 yards of a group of 
trumpeter swans standing on the snow-covered ice of a river margin. . . . When 
the trumpeter swans stand on some lake beach, it is usually possible to approach 
them within easy observation distance provided some caution is used. Thus at a 
small marshy slough near Vanderhoof a single immature bird remained standing 
on a muddy shore long after a flock of whistling swans, alarmed at my approach, 
had taken wing. Again, on a slough in the Cariboo region, two young of the 
previous year were observed at close range as they fed among a scant growth of 
dead rushes, 20 to 30 yards from shore. When two of us walked towards the 
slough the swans swam to the beach, walked up on it, and began dressing their 
plumage. At our closer approach they walked again into the water and swam 
slowly towards the centre of the slough. Many such instances could be related. 

Supporting this testimony, the experiences of Frank Groves and a 
companion in "walking up to" a single trumpeter in Nevada and that 
of Donald McLean in approaching to "within 30 feet" of a trumpeter 
in California, previously related, should be recalled. Gwen Colwell 
(1948) also approached to within less than 50 feet of 5 wild cygnets of 
flight age on the Klinaklini River in British Columbia. 

The escape-distance of trumpeters on the Red Rock Lakes, where 
these birds are captured periodically for banding and examination, is 
conspicuously great, even though they are fully protected at all times 
and fed regularly through the long winters. A long escape-distance 
is also prevalent in Idaho on the swan wintering area where some 



BEHAVIOR AND RELATED CHARACTERISTICS 79 

shooting occurs. In either locality alarm is shown at the first sight of 
jman at whatever range, and the distance of their departure, whether 
by swimming or flight, probably averages well over 300 yards. Even 
in Yellowstone Park, infrequent molestation by tourists, fishermen, 
and photographers has apparently contributed to the great escape- 
distance noted there, in some instances. 

While escape-distance has a measurable value for birds on the 
ground, it seems to lose some of its meaning when airborne individuals 
are considered. For instance, trumpeters which have flushed beyond 
300 yards may fly back over the disturbing person well within 100 
yards and even approach another person much closer if he happens to 
be in the path of flight. Other waterfowl frequently act in a similar 
manner. Unfortunately the usual low-flying habits of the trumpeter 
are not greatly affected by the fright factors which contribute to a long 
flushing range, and because of this they often provide tempting targets 
to waterfowl hunters who are strategically located, practically "flying 
down the gun barrel" if the hunter is partly hidden. This characteris- 
tic, coupled with their apparent inability to profit from learned wari- 
ness, which they practice on the ground, makes them especially 
susceptible to gunfire. 

H. Poulsen (1949: 196) notes that, "The swans do not show inten- 
tion movements when they are going to fly up such as the geese and 
ducks. But the neck is held erect and the call notes are emitted 
preparatory to taking wings." Alarmed trumpeters hold their necks 
stiffly erect, with perhaps a slow nervous pumping action if flight 
is delayed, and emit single, terse, trumpetlike warning notes hinting 
of the intended escape. These short warning calls may be repeated 
while the escaping swan is taking off and for a short while in flight. 
Although the preflight behavior of swans does not include the striking 
head movements of geese and ducks, it is still a characteristic pattern 
that tells the observer that the bird is about to fly (Lorenz, 1937). 

INTERSPECIFIC TOLERANCE 

Some phases of this subject are of course closely related to the pre- 
ceding topic, and its definition as applied to the trumpeter for the 
purpose of this discussion is simply learned escape-distance, or inher- 
ent escape behavior as it may be modified by remembered experience. 
Where interspecific behavior action is related to other birds, it will 
be discussed under the more appropriate topic of territorialism. 

Condon (MS) furnishes a good general picture of the variability 
of learned escape- distance under various conditions among the Park's 
trumpeter population, under the topic "reaction to molestation or 
interference with normal solitude." He concludes, "Unless swan are 
in constant contact with man, they show a pronounced tendency to 



80 LIFE CYCLE 



be alarmed by his presence and keep at what they deem a safe dis 
tance from him." The degree to which the normal (predator) aware- 
ness of these birds in wilderness surroundings contributes to this 
caution, or is learned by individuals from previous molestation at the 
hand of man, is of course not known. 

Condon (MS) also points out that while trumpeters within Yellow- 
stone Park are principally considered birds of wilderness character, 
this species has, in certain instances, accepted man and his activities 
to the extent necessary to incubate and raise its young successfully 
in spite of considerable human activity (photographers, fishermen, 
etc.). In some cases man's activity is not introduced into the breed- 
ing habitat to any extent until after incubation has been completed, 
and in others the adjustment to molestation apparently could not be 
made and the blame for the loss was laid on man-caused activity. 

Duane Featherstonhaugh (1948: 375,379) also notes that the trum- 
peter swan is to be found successfully living and raising its young 
on waters near active farmsteads in its northwestern Alberta environ- 
ment. He even observed one pair nesting in a slough near the Grande 
Prairie airport, a principal stop on the Alaska- Edmonton run, where, 
he records, the roar of engines went unnoticed by the birds. 

At the National Elk Refuge too, where breeding trumpeters were 
successfully established by the Fish and Wildlife Service, Aimer 
Nelson, the former Refuge Manager, states (correspondence) : 

In 1944 the pair nested in the Flat Creek Marsh about 600 feet from the 
Jaekson-Moran Highway. ... one of the cygnets ventured through the woven 
wire game fence onto the highway and was killed by an automobile. In 
1948 . . . one pair nested within 400 feet of the highway. This pair of swan 
left the nesting site two days following the hatch . . . [and] on July 1 the 
birds [parents] moved their young down Flat Creek to near the Flat Creek 
bridge at the edge of town within 50 feet of the Jackson-Moran Highway, and 
in showing off their family they caused a traffic jam on the highway which 
became crowded with tourists who stopped to see them. . . . This pair managed 
to raise three of the five young. 

In summary, human activity may be tolerated within distances 
ordinarily considered within trumpeter nesting territory, certainly 
much closer than intruding swans would be permitted. (The latter 
trait will be treated later when the spatial requirements of breeding 
trumpeters are discussed.) The critical factors involved in the 
toleration of man's activity on or near the breeding territory seem 
to be not so much the actual presence of man, or even the relative 
distance at which this is experienced, as it is the degree and regularity 
of molestation. 

At the Red Rock Lakes Refuge, these birds live in quite isolated 
wilderness surroundings at a considerable distance from humanity. 






BEHAVIOR AND RELATED CHARACTERISTICS 



81 



3n occasion they are disturbed by census activities on the marsh as 
*ell as by capturing for banding, examination, etc This apparently 
3ccurs infrequently enough so the "conditioning" does no occur. 
Hence, on the Refuge the swans are extremely wary and will avoid 
anv approach by man. 

Wild trumpeters, especially the young which have apparently never 
been molested or frightened by man, have been found to be very 
tolerant of man, sometimes almost unbelievably so. Featherston- 
haugh (1948: 376) provides an interesting statement in this regara, 
telling of two stranded cygnets rescued and raised by Joe Tomshak 
in the Grande Prairie country, as he writes, "When we saw them [the 
yearling trumpeters] they were as tame as domestic fowl. They would 
waddle up to the kitchen door to beg for bread and drink from the 

rain barrel." , n . « 

An even more remarkable example is provided by Gwen Colwell 
(1948) as she describes an unusual visit of five wild trumpeter cygnets 
one winter to the meteorological station at Kleena Kleene, British 
Columbia : 

Peter, our goose, was fed his breakfast that morning on the ice ledge along 
the river bank. In no time the swans showed interest and were soon literal y 
shovelling in mouthful* of wheat. . . . They became so tame that they would 
scramble onto the ice when they saw us coming. ... At the end of the first 
week they had found their way to the house door. . . . ^YTtlTv fined 
we had all five huge birds in the telegjaph office at once. Jhey literally filled 
the room. Indeed, one felt that if they became alarmed and decided to take to 
the air, the roof would in all likelihood be carried off on their backs. 

In captivity, trumpeters have lived for long periods and even bred 
regularly in favorable environments, so the presence of man or human 
activity, in itself, cannot be said to be inimical to their existence. 
When not molested, trumpeters can tolerate considerable human ac- 
tivity and actually thrive if other factors prove favorable, lnis 
tolerance may decrease as the birds acquire a certain amount of general 
predator wariness, which increases with age. 

Trumpeters nesting in a wilderness environment are alert to any 
change in the appearance of the landscape within their nesting terri- 
tory. Any such action as the erection of a blind is regarded warily. 
Nevertheless, Marshall Edson, photographer for the Idaho Depart- 
ment of Fish and Game, reports no difficulty in having breeding 
trumpeters accept a canvas blind located at no great distance from their 
nest, but these birds were familiar with man-made activity. My 
experience on the Red Rock Lakes Refuge wilderness breeding 
grounds is to the contrary, with one nest desertion resulting when I 
moved a blind from approximately 80 yards away, where it had been 
accepted during the incubation period, to only about 40 yards distant. 
Scott and Fisher (1953: 208) report similar extreme wariness in the 



82 



LIFE CYCLE 



whooper swan on its isolated breeding grounds in central Iceland 
stating : 

Most females left the nest as soon as the intruder came into sight. . . Two 
pairs of swans showed extreme shyness when a photographic hide was erected 
more than sixty yards from their nests. One pair deserted and in the other 
case the hide was removed only just in time. 

On one occasion a pen left a nest as a result of hearing the alarm note of a 
pinkfoot [goose], while the intruders were out of sight over a ridge. She 
apparently left in a hurry as the eggs had not been covered. 

The high degree of tolerance displayed by trumpeters toward the 
presence and activity of other birds and animals persists even in the 
case of possible predators. Referring to their studies of the trumpeter 
in Yellowstone Park during the summer of 1938, Oberhansley and 
Barrows relate (MS) : 

Little concern was shown by swans at the approach of elk, moose, beaver bear 
or coyotes in observed cases and in no instance were they seen to be molested 
or greatly alarmed by any bird or animal other than man, at whose approach 
they were always alarmed. The first sight of a fisherman approaching Riddle 
Lake, for instance, would drive the swans from the small islands at the nesting 
site to the extreme north and east shore of the lake nearly a mile distant At 
other lakes reactions were similar. Cover was resorted to if sufficiently distant 
from the intruder, while on larger lakes the swans swam rapidly away from 
the shore. 

R. W Patrick (1935 : 116), observed a pair of mute swans attacking 
a bullock twice, finally driving it away. Statements by other ob- 
servers indicate that mute swans, especially in captivity, are generally 
more pugnacious than the swans of the genus Olor, though reportedly 
this trait is much less apparent in wild mutes (Witherby et al ,1939- 
175). J 

MEMORY 

The memory characteristics of trumpeters seem to play an important 
role in such actions as escape-distance. From the standpoint of the 
wildlife manager, at least the fundamentals of the trumpeter's memory 
traits should be understood so that human activity can be planned ac- 
cordingly, as this characteristic is unusually keen in this species 

Smith and Hosking (1955: 116-117), in their experimental studies 
of the aggressive displays of some birds in the laboratory, found that : 

Birds possess to a remarkable degree, a high retention of visual images 
which become deeply "imprinted" on the bird's consciousness. ... for birds 
are above all animals "eye-minded," and dominated by visual stimuli. 

This is certainly true of the trumpeter. One characteristic which 
has impressed observers who have worked with this species has been 
its ability to remember visual experiences which have frightened it in 
the past and to govern its actions accordingly. For example, in the 
year 1950 I moved a small shed from Refuge Headquarters' to the 
swan wintering grounds and placed it at the water's edge to serve 



BEHAVIOR AND RELATED CHARACTERISTICS 83 

as a swan observation post and photographic blind. During the fol- 
lowing winter or two the swans became accustomed to this rough 
ishelter, and having no reason for mistrust, accepted it as a part of the 
landscape and fed on occasion practically within its shadow. During 
the winter of 1951-52 I departed from this blind in full view of a 
large flock of trumpeters without first arranging for another party 
to preflush the flock as was customary, thinking that no harm would 
be done since it was the last observation of the season and any harmful 
effects would soon wear off. This abrupt action of stepping out of the 
blind directly into the full view of a hundred or more nearby swans 
naturally startled the flock, which, after a moment's hesitation and 
warning calls, took flight. 

This event was apparently remembered in succeeding years, at least 
until the 1955-56 season, since they were never observed to "trust" 
the small building to the original extent again, always moving off in 
a wary attitude if they inadvertently fed or moved too close. Though 
the distance to which they would approach the blind decreased steadily 
as time went on, some mistrust was still evident during the spring of 
1956, 30 feet being about as close as they would come in to feed, even 
when attempts to bait them closer were made. Apparently, the older 
birds which remembered the event communicated their feelings to 
others unconsciously by not approaching this shed too closely, for 
there must have been a considerable turnover in population during 
this 4-year period. 

Erickson has mentioned that the trumpeters held in captivity at 
the Malheur Kefuge remembered their capture and subsequent 
blood tests for avian tuberculosis, being much warier and more difficult 
to capture and handle the following season. 

Leland Stowe (1957: 223) also noted, apparently from the testi- 
mony of the Kalph Edwards family of Lonesome Lake, British Co- 
lumbia, that though wild wintering trumpeters at "The Birches" had 
taken grain from the hand of Trudy Edwards for years, after they 
observed the capture of several of their number which had been lured 
by grain into a trap, they would not accept hand-fed grain again 
during four subsequent winters. 

Trumpeters are very alert to events, and recall detail well. John 
Holman (1950) quotes Ralph Edwards of British Columbia as writ- 
ing, "The swans are very sensitive to any changes in routine, to any 
change in feeders, or even to any change in the garments worn by the 
feeder." Similar characteristics have been recognized in Canada 
geese (Hochbaum, 1955:43). 

SENSORY PERCEPTION 

While no special work has been done on this subject, it has been my 
experience that their sense of sight and hearing are very keenly de- 



-JL.V. • --• 




Figure 28. — Trumpeters on Culver Pond display the 2 methods of plumage-shak- 
ing, 1 employing the wings. The elevated position necessary for either posi- 
tion is attained by rapidly treading the water. 

veloped. Foreign sounds, such as the click of a camera shutter, are 
heard at considerable distances and may even cause an escape move- 
ment. Sight is also extremely acute, especially at great distances. 
Its long neck enables the swan to see above low forms of marsh vege- 
tation far out over the water, and its keen eyesight can detect even 
a cautious approach by an observer. 

SOME GENERAL BEHAVIOR ATTITUDES 

Submerging and Diving. When hard pressed in close pursuit and 
unable, to fly, all age classes will dive to avoid capture. The older 
cygnets as well as the immatures and adults are frequently surpris- 
ingly adept at this, and once submerged are often capable underwater 
swimmers. 

Sometimes adults also display the curious ability to submerge almost 
their entire bodies in the water while the head and neck remain upright 
in a normal position. This is apparently accomplished by changing 
the web action, thus literally pulling themselves downward in the 
water while swimming. I have seen this phenomenon employed 
occasionally by flightless Refuge swans when they were attempting to 
avoid capture by boat in deep water. After diving a number of times, 
followed by extensive swimming under water, they become winded, 
and upon successive approaches by the boat they sometimes submerge 
their buoyant bodies gradually in this manner in order to avoid diving 
until the last possible instant. 

Oberhansley and Barrows (MS) also note this behavior by swans 
in the Park, writing, "In some attempts at concealment they appeared 
to become less buoyant and drew their heads down to the level of the 
sedges in which they were hiding . . . becoming almost invisible." 

Resting. All the various postures of relaxation appear to be 
assumed most commonly w T hile out of the water. When in the water, 
individuals usually are moving about feeding, swimming, or en- 
gaging in some other definite activity. Even during the cold winter 




Figure 29.— A small flock of trumpeters feeding with goldeneyes at Culver Pond, 
Red Rock Lakes Refuge. Typical drinking attitude is shown by swan with 
outstretched neck, right of center. 

weather, when they are not actually feeding, most of their time is 
spent loafing or sleeping on the ice and snow. 

In sleeping, whether prone or erect, the trumpeter curves its long 
neck to the rear, resting its head on the back between the wings and 
with the tip of the bill usually tucked under a wing up to the nostrils. 
I have never seen the trumpeter sleep floating in the water, although 
Condon has observed this in the Park. 

Oberhansley and Barrows (MS) record of the Park swans : 
Resting attitudes observed consisted of lying on the breast on land or in 
shallow water, floating on the water, and standing on either one or both legs 
with the neck usually recurved back across the body and the head tucked under 
the wing. Two swans at Geode Lake were each observed to stand continu- 
ously upon the left leg while alternately sleeping and preening for a period 
of 54 minutes. During most of this time the right leg of each was alternately 
trailed to the rear and downward at an angle of about 15 degrees, then drawn 
into the body or used to scratch the head and neck. In each case during sleep 
the neck was curved to the right across the body and back with the head tucked 
under the right wing. The more profound sleeper completely concealed the 
eyes, while the other one was more restless, never concealing the eye and fre- 
quently raising the head to a watchful position for a short time. It required 
much more care to approach swans when resting than when they were feeding. 

Plumage-shaking. When swimming or feeding in the water, or 
shortly after coming ashore, the trumpeter commonly shakes the wa- 
ter from its feathers by one of two methods. When this action does 
not include the wings, the statement of H. Poulsen (1949: 196) ap- 
plies, "When a swan is going to shake its plumage, the movement 
starts at the tail, which is swung rapidly side to side, and then the 
movement is spreading all over the body ending at the head." If 
this is accomplished by the trumpeter while still in the water, the 
whole body is first elevated by rapid treading. In any case the neck 
is held more or less in an outstretched position. 

This method is not seen so commonly as a similar action which 
employs a full beat or two of the outstretched wings. The primary 



86 LIFE CYCLE 

feathers are also frequently adjusted following this shaking, which 
is accomplished with the neck bent in a typical "exertion" attitude. 

Attitude of Head. Erickson has noted with the captive trum- 
peters at the Malheur Refuge that the male commonly held its head 
in a nearly level position whereas the female often carried its head 
with the bill tilted slightly downward. If this holds true in wild 
flocks, it might aid the determination of sex in the field. 

Drinking. The trumpeter drinks water in the general manner of 
other fowl, the water being first drawn into the mouth by submerg- 
ing the bill and then transferred to the lower regions by elevating the 
head and neck. The latter movement is accompanied by a fairly 
rapid movement of the mandibles, a swallow of water visibly moving 
down the neck as a mobile swelling. 

Foot Attitude. Some of the swans have the curious habit of occa- 
sionally holding one foot outstretched backwards, as if drying this 
appendage. This may be done either while on the water or when 
standing on the land, and is a much observed trait of the mute swan. 
A. C. Bent (1925: 282) records an observation of this habit in the 
whistling swan, but I have only rarely seen it practiced by the wild 
trumpeter (figure 43). 

Carrying Young. Delacour and Mayr (1945 : 9) state of the swans 
generally, "They seem to be the only Anatidae which have the habit 
of taking their downy young on the back when the young are tired 
or cold. This is the usual practice with Mute and Black-necked 
Swans. It is exceptional in the other species." To my knowledge 
this trait has never been observed in the trumpeter. 

DISPLAY 

Even in a group of birds noted for their display behavior, the 
trumpeter is an expressive species. Some of this reputation of swans 
comes from the frequent "threat" attitudes of the common mute swan 
of park and zoo. While the trumpeter does not exhibit this particu- 
lar display, both its voice and prominent physical features are often 
employed in other ways to show its various feelings about aggression, 
recognition, territory, sex, etc. These emotions are expressed prin- 
cipally by the action and attitude of the wings, head and neck, gen- 
eral body demeanor, and voice, with many variations possible. But 
the true "courtship" or "nuptial" displays so prominent in the lesser 
waterfowl are notably lacking in the swans. 

An attempt might be made to classify the fundamental behavior 
postures as expressions of "recognition," "triumph," "aggression," 
etc., but since these all have a common root in their display appear- 
ance, in which the voice is an integral part, such an appraisal should 
await that time when extensive recording's of their vocal efforts have 



:.." 



*i's$?;.. 







Figure 30.— A typical mutual display of 2 wild trumpeters on wintering waters, 
Red Rock Lakes Refuge. This basic display invariably is accompanied by 
vocal expression. It is used with variations in "recognition" and "aggression", 
and may have nuptial and other connotations. 

been made in synchronization with motion pictures of their various 
actions. When these are available to the researcher for detailed com- 
parison and analysis, we may begin to understand the behavior mech- 
anism of these unique fowl. Without this, a fundamental knowledge 
of the species is lacking. 

The most common display attitude is one in which the quivering 
wings are raised horizontally and partly extended. This posture, 
when accompanied by the pertinent vocal effort and extended posi- 
tion of the head, neck, and body, is used with some variation on 
greatly different occasions. In the water or on solid footing, the 
exact position and movement of the wings may vary considerably. 
The angle of the head at the end of a usually fully extended 
upright neck is nearly level in this basic display, contrasted 
with the definitely upward-inclined head and bill of the whooper 
swan as shown during its ''mutual greeting ceremony" by E. A. Arm- 
strong (1947: 192,142). 

469660 O— 60 7 



LIFE CYCLE 




] 




Figure 31. — Trumpeters face each other in mutual display (right foreground) 
and a swan indulges in a wing-flapping plumage shake (center background) 
while an adult bald eagle watches from a background snowbank in March 
on Red Rock Lakes Refuge. 

This basic display attitude has only been seen exhibited as a mutual 
action. It is shown on occasions when aggressiveness is involved 
among flocked birds that are socially active, or during the breeding- 
nesting-brooding season when a member of a breeding pair flies back 
to its mate after successful defensive action towards another swan 
trespasser. It may also be employed first by a single adult and result 
in mutual action when aggressive tactics are employed within the 
territory. During such times, both birds apparently recognize each 
other as foes and rush together with loud staccato trumpeting, exhibit- 
ing with outstretched quivering wings and extended head and neck 



BEHAVIOR AND RELATED CHARACTERISTICS 89 

as they plunge in to do battle, or turn to escape. O. Hilden and P. 
Linkola (1955: 524) report very similar behavior in the whooper 
swan, as do Witherby et al. (1939: 170). 

Essentially the same display behavior, but with some notable varia- 
tions, commonly occurs among the individual members of late-winter- 
ing flocks as they gather about the chosen feeding areas or on their 
extensive loafing grounds of snow, which at that season has become 
packed and hardened. This attitude appears to arise from less justi- 
fied reasons than territorial protection. Commencing with bobbing 
heads and gradually increasing wing action, it develops into positive 
aggression, apparently as the result of self-assertive tendencies which 
seek expression before territorial establishment. It also occurs to 
some extent whenever large flocks of nonbreeding trumpeters gather 
together, and so may be a result of population pressure. 

One variation of this display may involve as many as 4 or 5 birds, 
if the efforts of 1 or 2 are successful in getting action started. In this 
group action, which may occur either on land or on water, the body 
position is more erect than in the simple mutual display, and the 
wings are also usually slightly more extended and are apt to be held at 
a higher angle, thus showing off the underside of the wing primaries 
to a greater extent, Treading increases their stature in the water, 
while wing movements are much more active as each swan strives to 
maintain or improve his performance in the circle of displaying indi- 
viduals. After the preliminary head-bobbing procedure, the bird 
holds its neck quite fully extended with the bill about level, facing 
its fellow demonstrators in voluble expression. 

The trumpetings of the group exhibitors are truly remarkable, even 
from the start, but especially so when they rapidly increase in tempo 
and intensity as they are joined by other swans. When the climax 
has been reached, it usually ends in lowered-head aggressiveness as 
various individuals shoot rapidly in pursuit of others in a regular 
free-for-all amid a noteworthy outburst, of trumpeting from the 
always interested onlookers. It is impossible to say whether these 
exhibitions are made up principally of any particular sex or age group 
since those observed have been of unknown sex and in the white dress 
of the adult. 

This performance is, in general, the usual aggressive expression of 
these birds, which they may assume either in defense of their breeding 
territory or in pursuit of personal victory in a spontaneous dispute. 
In the case of breeding birds on the territory, a warning display 
and calls by one or both of the resident pair usually suffices to turn 
approaching strangers away. Occasionally an intruder alights in 
an active territory of a pair, or more rarely mates from pairs in 
adjacent territories come to blows. 



90 



LIFE CYCLE 









Figure 32. — Two trumpeters landing on Culver Pond are greeted by displaying 
swans. Band on left leg of lower bird marks it as a pen (female). Both 
Barrow's and common goldeneye ducks are present in this scene. 

Aggressive action on land or water begins when an individual 
with head lowered for action propels itself decisively by wing 
or web in the direction of the trespasser. In the air the pursuing swan 
closes the distance to the trespasser by rapid flight. The victim, 
perceiving its predicament after a losing race, usually turns steeply 
upward and is followed for a brief time by the pursuer, after which 
the action is abruptly terminated by the pursuer turning back. In 
the water, actual feather pulling and wing pummeling may occur 
before one is successful in routing its opponent. Such punishment 
seldom lasts more than a few moments, but it can nevertheless be 
formidable and effective and seems always to result in the defeat of 
the trespasser. Invariably the returning victor is met in the home 
territory by its mate which joins in a mutual display of quivering 
wing, bobbing head, and staccato trumpeting, ending with reclining 
neck and wings and wailing notes. This is comparable to the 
"triumph ceremony" described by Heinroth and Lorenz for the geese 
and shelducks. 

The aggressive attitude may also be assumed by any flightless 
trumpeter when it is confronted suddenly or is pursued by man and 



BEHAVIOR AND RELATED CHARACTERISTICS 91 

is unable to escape. It is usually displayed in such cases when an 
individual is captured in a confined location or cornered on solid 
footing in an unfamiliar environment, when the aggressive posture 
or action is accompanied by considerable hissing. Swans on water 
have never been reported to take an aggressive action toward man. 
Possibly the bird feels that it is in its native element and escape is 
possible until the moment of capture. 

From the foregoing remarks it might be concluded that the trum- 
peter is a pugnacious and quarrelsome bird. This is not generally 
true, however, at least in the wild population with which I have been 
acquainted. Minor emotions are usually expressed initially by voice, 
and many times these feelings are not further developed. Hours 
have been spent observing these birds on their breeding grounds when 
not a single aggressive action was noted. During the fall months 
when the breeding birds and their families rejoin the nonbreeding 
flocks, the pair formation which apparently occurs then among ma- 
turing individuals (Delacour and Mayr, 1945 : 8) adds to the general 
unrest resulting from the change of season and impending migration, 
and more frequent emotional displays take place. This is also true 
when early spring unlocks their traditional nesting marshes, since the 
birds are still confined to limited feeding areas about the few warm 
springs. The approach of the breeding season, with its sexual stim- 
ulation and welcome seasonal change after the winter's hardships, 
ushers the population into its greatest social season. At this time 
display attitudes are commonly observed among flocked birds. As 
soon as the marshes open up and the birds disperse, the display activity 
shifts to the nesting territory and is less frequent. 

Delacour and Mayr (1015 : 9) describe the precopulatory display of 
swans : 

Swans, geese, and whistling ducks (tree ducks) have essentially the same pre- 
copulatory display, both birds of a pair repeatedly dip the whole head and neck 
until finally the female flattens herself out on the water and sinks deeper with 
the neck half extended. 

On several occasions a similar display, but without the dipping cere- 
mony, was observed on an open-water area while the birds were still 
on their wintering grounds during March. In these cases, the postur- 
ing was not followed by copulation, and perhaps expressed some other 
mood or was not even a formal display attitude. A description of 
copulation by trumpeter swans provided by Frank McKinney (corre- 
spondence) reveals that this act does not differ from that noted with 
other large species of waterfowl. 

Intraspecific strife and display among the confined trumpeters at 
the Malheur Refuge is described by Erickson, who relates (corre- 
spondence) : 




Figure 33. — "Solo" display, trumpeter swan on Culver Pond. Red Rock Lakes 

Refuge. 



A pecking order is established with each group of swans in the pond. With 
the very young cygnets, size seems to be the main factor determining the position 
of each bird in the order, but as the smaller individuals approach the previously 
larger ones in size, vigorous tussles ensue as members of the order attempt to 
maintain or improve their positions. The scuffling usually is most noticeable 
during feeding when the swans are grouped more closely. 

In an evenly matched light, the swans will approach and attempt to intimidate 
the adversary by carrying the head and neck low. If the opponent stands his 
ground, the aggressor will drive in and each will grasp the neck or feathers of 
the head, neck, or anterior part of the body of the other swan, both birds thrash- 
ing against each other, breast to breast, and each attempting to tread the other 
one under the water. This may last as long as two minutes, though usually 
the outcome is forthcoming within a few seconds, and the weaker or less experi- 
enced member of the match turns and flees, the victor grasping the rump or 
tail feathers of the loser and being towed for some distance around the pond, 
accompanied by much splashing and flailing of wings. 

The wings are less commonly used in striking the opponent, but the loser 
always uses them in running from the fight. When the loser reaches the cover 
of the willows or races up on shore, the winner will turn and face an apparently 
admiring group of other swans who have been watching the performance, will 
draw its head far back on its back, half spread its wings on each side, and 
promenade back towards the flock in an extremely pompous and swaggering 
fashion. 

The cygnets appear more interested in these engagements than the older 
swans, and will gather along the fence in their part of the enclosure and pay 
tribute to the winner in the form of rapt attention and juvenile "trumpeting" 
as the winner slowly swims by in full display. Sooner or later after the 
match, both participants will rise on the water and flap their wings. 

If the winner is a member of a mated pair, he then will rush forward, and be 
met in like fashion by his mate, both facing each other closely with quivering 
wings partly outstretched and trumpeting their congratulations in staccato 
fashion, usually ending these expressions in a longer wailing sound. 



BREEDING 



93 




Figure 34.— Rear view of "solo" display (left) . 

Judging from my observations of wild trumpeter populations, 
Erickson's report indicates that behavior becomes much more formal 
and stylized in captive birds, where certain features may be more fully 
practiced and developed. Though strikingly similar, they do not 
appear to be truly representative of similar actions among a wild 
population. A fertile field for further study in this direction awaits 
the research worker. 

BREEDING 

Although a basic knowledge of breeding characteristics and habits 
is essential to a grasp of population dynamics, very little on this 




Figure 35.— While a pair of trumpeters engages in mutual display (left center), 
4 gather in a group exhibit (right). Mallards and pintails in the irregular 
foreground. 



94 LIFE CYCLE 






subject is now known about the trumpeter in its breeding habitat in 
the United States. Only from detailed observations by various in- 
dividuals over a period of years, coupled with notes on similar char- 
acteristics of better-known species of swans, can we throw light on 
the important aspects of reproduction. 

The breeding biology of mute, whooping, and Bewick's swans, in- 
cluding notes on pair formation, courtship, territorial traits, nesting, 
and hatching success, is treated in Hilprecht (1956: 55-89). 

PAIR FORMATION 

Literally nothing is known of the ages of wild swans in their pre- 
paring associations. 

The behavior of captive trumpeters should provide clues regarding 
the formation of pairs as it occurs in the wild, and Erickson furnishes 
an interesting observation in this regard, writing (correspondence) : 

Although the cygnets form friendships during their first year of life, they 
seem to serve no sexual function, and it is as common to see three associating 
amicably with one another as it is with pairs. The first mated pairs develop 
during the third year, when they seem to resent intrusion by other single or 
paired birds. 

Writing in a more general sense of the swans as a group, Delacour 
and Mayr (1945 : 8) state: 

Pair formation, which occurs in the fall in all temperate zone swans, takes 
place without elaborate displays. According to Heinroth (1911), birds that 
are in the process of pairing swim in close proximity, press the plumage close 
to the body, and hold the neck in a peculiar position, the head appearing 
thickened. 

BREEDING AGE 

Assuming that the ages of both birds are equal, pairs do not 
appear to become firmly mated until well along in the third 
year of life. Possibly the earliest breeding of trumpeters may be 
accomplished as the pair enters the breeding season completing their 
fourth year of life. It is conceivable that if one of the paired birds is 
older, or has had previous breeding experience, successful pairing 
and nesting may be initiated somewhat earlier than would otherwise 
be the case. (W. H. Watterson [1935: 238] reports mute swans 
nesting at 2 years of age. ) 

Referring to captive trumpeters again, Erickson furnishes another 
pertinent statement on the earliest ages at which copulation was 
observed among trumpeters in the Melheur flock, writing: 

No pairs were observed copulating until in the five-year-old class, although it 
may well have occurred unobserved. One three-year-old pair, which was sepa- 
rated from the others, made a rather listless attempt at nest construction, but 
the effort was abandoned before the •'nest" had passed the "platform" stage. 
This same pair again attempted nest construction and were seen copulating two 



BREEDING 



95 




Figure 36.— Aggressive action of trumpeter following group display, Culver 

Pond wintering waters. 

years later. However, the female developed a lameness and did not continue 
nest-building, dying about three months later. 

Whatever the average minimum breeding age of trumpeters may 
be, the initial age at which a wild pair comes into breeding mood and 
nests may be influenced somewhat by the quantity of unoccupied ter- 
ritory located in suitable nesting habitat. Delacour's statement 
(1954: 72), on the four closely- related circumpolar species of swans 
may apply to the trumpeters of the Red Rock Lakes : 

Although the adult state is assumed in the third year, it seems that most of 
these swans do not breed until the fifth or sixth year, perhaps because it is 
often difficult for young birds to appropriate a nesting territory. 

We do not know if a high breeding population of swans occupying 
a limited area has any effect in postponing the initial breeding age of 
nesters. It will be shown later that while the number of breeding 
swans at Red Rock Lakes has increased in proportion to the whole 
population, it is breeding success which has declined markedly during 
these years. 

Examples of the nesting of 3 pairs of trumpeters of known ages 
may throw some light on the subject. In the first case, a pair of 
trumpeters, forming from 3 cygnets-of-the-year, which were taken 
from Red Rock Lakes to the National Elk Refuge in 1938, nested 
successfully in 1911 when they were entering their sixth year of life. 
Though these individuals were established within easy flight range of 



96 LIFE CYCLE 

the Yellowstone swan population and had the ability to fly unre- 
stricted during this period, they were seen so consistently on the Ref- 
uge following their transplanting, where none had been observed for 
many years previously, that the identity of the pair is assured. It is 
also interesting to note here, in reference to Jean Delacour's remarks, 
that while these were the only swans in the habitat, they did not 
nest until practically 6 years old. 

Two pairs of trumpeters have since bred several times in the limited 
marsh area available on the Elk Refuge. The second pair nested first 
in 1948. Since only 1 cygnet was produced in the original nesting at- 
tempt in 1944, the second pair (if natives of the Elk Refuge) must 
have been mated either from a combination of brood members of the 
1944-45 hatches, or that of 1945 when 3 cygnets were produced. In 
either case at least one of the mated pair would have bred at its hypo- 
thetical minimum breeding age, when just completing its third or 
fourth year of life. Of course, there is no proof that either of the 
mates comprising the second nesting pair were native Elk Refuge 
birds, but in view of the general tendency of the Anatidae to return 
to their natal breeding marshes following the wintering period, at 
least one and probably both mates of the second pair were Elk 
Refuge-bred birds. 

In the third instance, a pair of trumpeters which formed from 
several cygnets-of-the-year transferred to the Ruby Lake Refuge 
from Red Rock Lakes in 1949 were seen on nearby Franklin Lake 
in their third year of life, during an October 1952 aerial water- 
fowl census. On October 7, 1953, another aerial census revealed 
a pair of swans with a lone cygnet, again on Franklin Lake. 
Because identification of this pair was not confirmed by a ground 
check, and because whistling swans have arrived from the north at 
Ruby Lake as early as mid-October and no nesting swans were noticed 
on an earlier spring flight over Franklin Lake on April 16, this record, 
like the others, can be classed only as hypothetical. If it is a valid 
trumpeter nesting record, as the refuge manager assumed, these birds 
would also have begun nesting at a minimum breeding age, just com- 
pleting their fourth year. Thus the only evidence at hand, which is 
admittedly sketchy, indicates that nesting may begin as early as 
the fourth year of life (3 years, 10 months) or as late as the sixth year 
(5 years, 10 months) . 

MATING FIDELITY 

The oft-repeated truism that "swans mate for life'' has been estab- 
lished more by a lack of evidence to the contrary in semidomesticated 
captive birds than by a thorough study of the facts with wild swan 
populations. While a general rule of life-mating appears to be valid 



BREEDING 



97 




Figure 37. — Aggressive pursuit terminating a group display of 5 trumpeters, 
Red Rock Lakes Refuge. 

for swans, it is reasonable to suppose that a wild swan which loses its 
mate early in life may remate. 

H. Poulsen (1949: 197) furnishes an interesting general statement 
on the mating fidelity of swans among those pinioned in the zoological 
gardens of Copenhagen, Denmark, writing : 

The Mute Swan is strictly monogamous. Among the 60-100 tame (hand ampu- 
tated) Mute Swans kept in the parks of Copenhagen, each pair always sticks 
together for life. In the cases in which it has been attempted to pair a swan 
with a new mate instead of the dead one, no success has been achieved in contrast 
to the statement of HEINROTH (1911) that the re-pairing is not difficult. 

There is a real attraction between particular individuals of the opposite sex. 
Unfaithfulness between the mates is rare. During the last ten years only three 
cases have occurred among the tame swans in Copenhagen. Thus a wild male 
had paired with two tame females. Just as mentioned by PORTIELJE (1936) 
the pairing with the second female occurred when the first one was sitting on 
the eggs. 

In captivity, bonds of naturally monogamous species sometimes 
break down, and the trumpeter apparently is no exception. Dr. G. C. 
Low (1935: 147) records that in England: 

Polygamy in the Mute Swan is very rare indeed, and I have only heard of one 
instance, in addition to one case of a female Swan that paired with two males, 
all three living together on harmonious terms. There was for many years at 
Woburn Abbey a breeding trio of Trumpeter Swans, but I was never able to 
discover whether the odd bird was male or female. . . . Many years ago a 
ludicrous instance occurred at Abbotsbury, where two male swans [mutes] not 



98 LIFE CYCLE 






only associated together, but even built a nest every year, upon which they took 
turns in sitting ! 

Dr. J. M. Dewar (1936 : 178) published a short but interesting paper 
in which he outlined some aspects of the mating infidelity of captive 
mute swans as it has been observed in certain rare instances over the 
years in England and Europe. In his article Dr. Dewar proposes 
the apt phrase "menage a trois" to describe "the association of three 
individuals of a bird species for the purpose of nesting." Dr. Dewar 
continues his explanation of the term and its pertinency to the mute 
swan as follows: 

As a label, menage a trois is preferred to its synonym, bigamy, because 
menage a trois implies the setting-up of an establishment, which bigamy does 
not necessarily do. . . . 

Menage a trois has to be distinguished carefully from homosexuality and 
other sexual relations. Homosexuality, which has been recorded several times 
in the Mute Swan, is an association of two individuals of the same sex, leading to 
pairing and nest-building, and in the case of females the laying of eggs which are 
necessarily infertile. 

Altogether I have been able to collect from the literature six examples of 
menage a trois in the Mute Swan and to add a seventh case coming under per- 
sonal observation. 

John Ellis (1936:232), commenting upon Dr. Dewar's paper, es- 
tablishes at least one and possibly two apparently valid cases of 
menage a trois among mute swans on an English park lake. 

A similar aberrant relation existed among three trumpeters living 
south of the Red Rock Lakes Refuge on a small reservoir in Fremont 
County, Idaho. I noticed them while flying the summer census on 
August 29, 1956, in the Island Park area. Three adult swans were 
together with three cygnets on the Icehouse Creek Reservoir. All six 
birds formed a typical family group which did not break up in spite 
of repeated low flights directly over them in the pond. Conversation 
with a rancher familiar with the swans on this Reservoir later con- 
firmed that observation, and further established the fact that these 
three adult birds spent most of the spring and summer together, 
from breeding season until fall, although one departed for about a 
month during this period only to return later and remain until autumn. 
In 1957 the three adult swans returned to the Reservoir and one female 
again nested successfully. This trio was broken up later in the summer 
by the death of one of the adults. The sex of the extra bird was never 
determined. As far as is known, only one nest was built each season 
and only one brood was ever seen. Of the hundreds of pairs of wild 
trumpeters I have observed with cygnets in their breeding territories 
during the period 1948-57, this was the only record in which more 
than two adults were seen with young. 



BREEDING 



99 



PRENESTING HABITS AND BEHAVIOR 

Long before the winter season has ended, sometimes 3 months 
before any open water appears, pairs or small groups of trumpeters 
may be found loafing about their still-frozen breeding habitat on the 
Red Rock Lakes Refuge. This bleak, wintry landscape is 6 or 8 
miles from the nearest open water at the spring-fed ponds. Water is 
not generally open in the Red Rock Lakes and marsh until after mid- 
April. But both paired birds and small flocks have been seen far out 
on the vast snowhelds which overlie the breeding habitat as early as 
February 2 (in 1949) and February 21 (in 1950). This practice be- 
comes commoner as the season advances. By the time the ice on the 
Lakes is ringed with run-off water from the surrounding meadows 
in April, it is immediately occupied by the swans, which then 
no longer need return to their winter feeding waters. 

Early occupation of the breeding grounds is apparently instinctive, 
as several Yellowstone Park observers have testified to the same be- 
havior pattern. In Yellowstone these early visits to breeding habitat 
may occur considerably later than at Red Rock Lakes, since early 
April is the earliest that swans have been seen visiting their still-frozen 
Park breeding habitat. Perhaps before April other loafing areas in 
Yellowstone are more convenient to winter feeding waters, and at 
Red Rock Lakes the breeding habitat may first be occupied coinci- 
dentally as loafing grounds during the late winter months, with no 
special territorial significance. With only one exception, the early 
literature states that the trumpeter was invariably the earliest water- 
fowl to arrive on its Arctic breeding grounds in the spring, preceding 
even the geese or whistling swans. 

Whether a given pair occupies in the late winter the area to be 
claimed as a territory later has not been determined. While pairs 
often occupy the same snowfields, more consistently as the season 
advances, their positive identity as the same pair was not established 
then, or related later to the breeding pair which nested in that area. 
It would appear logical to assume that in the last few weeks pre- 
ceding the advent of open water some territorial claims are being 
staked. Specific observations of the beginning of territorial defense 
actions are lacking. At first, just the pair's presence in an area is 
enough to show that the territory is occupied. 

During the period just before the spring breakup, especially on the 
warmer days and nights, vocal efforts and flight activity reach a 
climax. The long winter with its hardships and severe cold is draw- 
ing to a close and the excitement of the breeding season is at hand. 
Singles, pairs, and small groups move frequently about from one 



100 LIFE CYCLE 

open patch of water to another, spending long periods feeding or loaf- 
ing on the ice in the warm spring sunshine. The successively later 
twilight increases this activity, and vibrant sonorous trumpetings 
from the marsh are then audible far into the night. 

TERRITORY AND TERRITORIAL BEHAVIOR 

Territorial behavior in the trumpeter is characterized by the defense 
of the "mating, nesting, and feeding ground for young," the "type 
A" territory classification of Margaret Morse Nice (1941: 471). 
Territorial defense of areas other than that concerned essentially 
with reproduction is apparently unknown in wild swan populations ; 
hence, the discussion of territory here will be confined to this defini- 
tion of the term. 

Many observers writing from firsthand experience have agreed that 
swans as a group are among the most territorially minded of any 
birds commonly kept in captivity, setting up wherever possible 
relatively large areas of habitat where various other forms of com- 
peting bird life, especially other swans, may not be tolerated during 
the breeding season. It also appears that paired swans, wild or 
"kept," differ a great deal in degree of territorial behavior, both inter- 
specifically and intraspecifically. 

H. Poulsen (1949:195, 198), writing of swans both in the wild 
and in the Zoological Gardens of Copenhagen, says : 

In the Mute Swan the well known threat display (imposing posture, Im- 
poniergehaben, HEINROTH) serving to intimidate intruders in its territory 
was observed chiefly in the adult male, but also in the female when I ap- 
proached the nest at a time when it had left it in search of food. . . . but 
never in the Whooper Swan. . . . Outside the breeding season the swans are 
sociable birds, and outside their territories they even in the breeding season 
agree well. A strange female, which settled in a territory, was seen to be 
chased off by the owning male in threat display, whereas outside the territory 
he dropped this attitude, the feathers now being close to the body. Thus a 
strange female may have two different valences (TINBERGEN, 1942). Inside 
the territory : enemy, outside the territory : female to copulate with. 

Intolerance in the mute swan is apparently modified by captivity. 
H. F. Witherby et al. (1939 : 175) remark On this : 

Mute swan, much more than Bewick's and Whooper, is naturally predisposed to 
association with man and easily tamed. O. Heinroth found that behavior of 
wild-caught birds from the Black Sea differed in no respect from that of semi- 
domestic swans. Usually aggressive and vicious towards other birds in the 
semi-artificial conditions in which it often lives, but both Naumann and E. 
Christoleit stress that this trait is less noticeable in fullv wild birds. 



In wild state usually nests on islets in swamps or in shallow water in reed 
grown lagoons, but in semi-domesticated state almost anywhere near water, 






BREEDING 



101 



sometimes in colonies of great size. In protected areas of this kind, the nests 
may be only a few yards apart. 
* 
The swannery at Abbotsbury (Dorset) is not only the largest individual con- 
gregation of Swans in this country, but is unique. Though foot-marked and 
living under conditions to some extent artificial, the birds are all fully winged 
and virtually wild. Dependent to some extent on annual fluctuations of food- 
supply, numbers vary between under 200 and over 500 pairs. 

The interspecific relationships of mute and whooping swans are 
also noted in Hilprecht (1956: 89) to vary from "extremely tolerant" 
in captivity to "far more tolerant" in the wild. 

Erickson makes an interesting comment on the trait of interspecific 
intolerance among the confined trumpeters at the Malheur Kefuge, 
stating (correspondence) : 

The trumpeters are intolerant of intrusion by whistling swans and several 
whistlers have been seriously injured when they attempted to mingle with the 
larger birds and had to be rescued or they would have been killed. Trumpeters 
seem to disregard ducks and geese. 

E. A. Armstrong (1947 : 284) notes yet another aspect of territorial 
behavior in swans, writing, "Where a pair of territorial birds, such as 
whooper swans, are constrained to remain on a lake dominated by an- 
other pair they do not breed." 

Wild trumpeters usually show territorial aggressiveness toward 
their own kind by a combination of loud calls, tremulous movements of 
half-raised wings, and general hostile body attitude. In most cases, 
the defending resident pair is located on the water when the possible 
trespasser is spotted in flight some distance away. If the distant flyer 
approaches, one or both of the swans occupying the territory assume 
the aggressive posture with their wings while trumpeting with ex- 
tended head and neck their unmistakable staccato warning notes. 
If this display is ignored by the approaching swan and a close 
passage or landing appears likely, no time is lost by the defender 
in getting into the air in immediate pursuit. 

In flight, the pursuer always seems able to overtake the departing 
intruder, though if both territorial defenders take to air, only one, pre- 
sumably the male, approaches the trespasser closely. When with 
great effort the defending bird has succeeded in overtaking the fleeing 
invader, the latter invariably climbs sharply upward, only to be fol- 
lowed closely by its pursuer. This seems to be the signal for the 
conclusion of the chase, for the successful defender, now far from its 
territory, turns and heads homeward, announcing its return with 
triumphant trumpeting. Gliding downward, it splashes into the 
water near its waiting mate where both birds approach each other 
trumpeting their mutual staccato congratulations with outstretched 



102 



LIFE CYCLE 




Figure 38. — Observations on the territorial traits of trumpeters can be made 
by a single observer over several thousand acres, since the high mountains 
provide ideal vantage points. Here the birds on Lower Red Rock Lake are 
studied from an observation post on the northern flank of Centennial Moun- 
tains. 



tremulous wings, ending with a dipping motion of their head and 
neck and longer prolonged wailing notes. A display of the whooper 
swan on similar occasions has been reported by E. A. Armstrong 
(1947:142). 

Sometimes the defending resident bird is on solid footing when a 
flying intruder draws near. In this case the wings of the hostile de- 
fender are held in an even more outstretched drooping attitude than 
is possible when it is on the water. Here, since the extending of the 
neck appears more pronounced and the body is visibly erected, an 
even more threatening attitude results than when the displaying bird 
is on the water. In either case a direct attack by air follows shortly 
if the warning is ignored. 

Though both sexes usually partake in aggressive warning displays 
toward a trespassing swan, usually only one departs from the territory 
if pursuit or eviction is necessary, but both swans may take wing when 
prompt action is made necessary by an overt violation. 



BREEDING 



103 



Some territorial demonstrations occur from before the time that 
the nest is constructed in the spring until late in the summer when the 
cygnets are half -grown, though they are not shown uniformly during 
this period. They begin sometime after territories are taken, are most 
common during nesting and early brooding, and then taper off. 

Even during the height of the breeding-nesting season, however, 
aggression and discord are not common. When breeding populations 
have been highest on the Refuge, paired swans appear to spend the 
vast majority of their time feeding, loafing, incubating eggs, caring 
for young, etc., actually geting along quite harmoniously with 
neighboring pairs if territorial boundaries are respected, as they 
usually are. Too, the large flock of nonbreeders scatter out on 
the vast shallow expanses of Upper Red Rock Lake, feeding far 
out from shore on the extensive underwater aquatic pastures or loafing 
along the shoreline wherever breeding territories are not established, 
and seldom venture into the established breeding territories. 

At the other extreme, I observed a lone trumpeter, apparently a 
"lonesome , ' nonbreeder, warned or actually evicted from five different 
territories within the space of an hour. In some cases this individual, 
which appeared to be seeking the company of its kind, would be 
allowed to settle in a distant corner of an occupied territory for a few 
moments; however, the hostility of the residents would increase visibly 
with each passing minute until cob or pen, or both, would fly at the 
trespasser in a show of aggressive eviction. 

This is not the general rule, as mated pairs seem to learn neighbor 
relations quickly, just as the nonbreeders are apparently impressed 
with the importance of staying clear of the breeding areas. By the 
time of the flightless molt in July, few acts of aggression are observed. 
Movement is naturally more restricted at this season, but even after 
the molt, when flight activity picks up, aggression never appears to 
regain its former level, though it is displayed on occasion by breeders 
throughout the rest of the summer. 

During the incubating period, mated swans rarely leave their terri- 
tory unless one makes a short aerial inspection or defensive flight. 
Sometimes the departing bird may land some distance away in a "neu- 
tral" area where either alone, or congenially with others, it feeds or 
loafs, returning to the territory some time later. Breeding pairs 
nesting along the line of flight usually make aggressive warning dis- 
plays upon seeing the commuter in flight, and this usually causes 
the vagrant mate to move on. Except for such short absences, 
the mated pair usually remains resident within its nesting territory 
until late summer or early fall. Towards the end of summer, terri- 
torial bonds definitely become weaker ; this is especially true of brood- 
less pairs which may desert their home area altogether. 



469660 O — 60 S 



104 LIFE CYCLE 

On one occasion a mated pair, together with their brood, deserted 
their territory, travelling about 2 miles across Upper Red Rock Lake 
to settle along the opposite shore. Before this change of territory, 
the brood had been reduced from 6 to 2 from unknown causes, so it is 
possible that the parents wished to move from the cause of mortality 
and/or molestation. The area selected for the new home was the shore- 
line of a rather exposed bay about midway between two other active 
but distantly spaced territories. The season being well advanced 
and the site well separated from adjacent territories, the shift of 
residence was a successful move and permanent for the season. 

Some variation of territorial behavior is evident among individual 
pairs, depending upon their familiarity with each other. Also, the 
size of the territory defended, and the subsequent spacing of nests, 
indicates the appeal of a given habitat. Once the territory is estab- 
lished, adjacent resident pairs recognize their neighbors and modify 
their aggressiveness accordingly. The pursuit distance at which the 
defenders will take aggressive action is much greater when a stranger 
has been sighted than when a mate from an adjoining territory is 
observed. 

In one unique case of tolerance, 2 pairs nested for several successive 
years a measured 885 feet from each other on the same shoreline of 
a large widening of the Red Rock River. However, the nests were 
hidden in a dense growth of bulrush and the areas used in common 
were at a minimum, each pair keeping to its end of the opening and 
the adjacent stretch of narrower river channel. The distance be- 
tween nests is usually much greater. 

Interspecific aggressive behavior by trumpeters on their nesting 
territory appears to be directed chiefly against the larger birds. 
Usually, territorial aggressiveness is not shown towards the larger 
mammals, including man. When such an intruder approaches nest 
or young, even at a distance, trumpeter pairs discreetly desert. Trum- 
peters generally disregard the ducks. Both ducks and coots swim 
unmolested among trumpeters in close association and perfect har- 
mony. Ducks and coots feed intimately within the trumpeter "family 
circle" in the territory, where they compete with cygnets for surface 
food stirred up by the parent swans. When ducks and swans are 
present in large numbers and compete excitedly for grain in the 
restricted areas of open water of the Refuge wintering areas, a trum- 
peter may take an annoyed jab at some passing duck that gets in the 
way, but this is infrequent. 

In Denmark, K. Paludan and J. Fog studied the effects of in- 
tolerance displayed by wild mute swans against ducks. This study 
was made in response to the claim by many sportsmen that the 
swans in that country interfered with the reproductive success of 



BREEDING 



105 



wild ducks. While testimonies of swan intolerance with ducks were 
noted on small and restricted sites, the swan, mallard, and gray-lag 
goose bred close together without inimical effects in other instances. 
Paludan and Fog (1956: 44) concluded, "Under no circumstances 
can the Mute Swan be of any importance to the Danish duck popula- 
tion in general." 

The trumpeter is not so tolerant of the larger water and marsh 
birds. While Erickson (correspondence) mentions that nonnesting, 
captive swans held in an enclosure at the Malheur Refuge ignore these 
lesser companions, this may not always be the case with wild pairs on 
their breeding territory. Featherstonhaugh (1948:379) writes of 
trumpeters : 

The swans also drive off any geese (in addition to other swans) that may land 
in the vicinity of a nest, but they pay no attention to ducks or other marsh birds. 
We found nests of the mallard, the ruddy duck, the lesser scaup, and the Ameri- 
can coot within a few feet of swans' nests. 

In observations made between May 11 and August 30, 1949, 1 noted 
11 instances of territorial defense, 8 against other swans and 1 each 
against a white pelican, a great blue heron, and a Canada goose. 
Charles Hotchkiss, ranger-naturalist of the National Park Service at 
Teton Park, reported to me seeing a swan chase away a sandhill crane 
which came close to the swan nest. 

Interspecific intolerance involving muskrats has been observed in 
at least two cases. W. Verde Watson, a Park Service naturalist in 
Yellowstone Park, furnishes an account of a trumpeter swan killing 
a muskrat, supposedly in defense of its cygnet brood, writing 
(1949: 49) : 

The pair of adults were lazily swimming about off shore sounding rather 
plaintive, deep toned "words" of instruction to some of the cygnets. These 
youngsters had just given some heed to the old folks and were making way 
toward them when one of the adults, with a great flurry of feathers and beating 
of the water, half-flying and half-running upon the water, bore down upon 
the muskrat intruder which was apparently swimming past some 25 to 50 feet 
farther out in the lake. Arriving at the spot where the 'rat must have been she 
beat the water furiously with both wings, seeming literally to walk about on 
the water as she darted very quickly from side to side and round and round 
following the quarry. All the while she hissed loud and angrily and periodically 
pecked vigorously at the object of her wrath. ... A quick look through the 
telephoto finder of the Leica revealed that the animal was a muskrat, and 
observations during the succeeding couple of minutes or so indicated that it 
was probably done for. . . . Thus we were really amazed when it emerged 
again and swam almost directly at the same swan that had abused it so badly 
shortly before. . . . The old bird saw it coming and with determined bearing 
swam to meet the muskrat. This time the initial shock was delivered as a 
vigorous and meticulously aimed peck followed by the same unmerciful wing 
beats and much loud hissing. . . . This encounter probably did not last over 



106 LIFE CYCLE 






30 seconds, and then when the bird sailed away from the scene and the water 
settled down the unfortunate 'rat came up and lay motionless on the surface, . 
. . . Subsequent autopsy . . . revealed a considerable bruised and bloody 
area in the neck, and it was thus considered likely that the fatal blow may have 
been the first well aimed hammerlike peck delivered at the opening of the second l 
round. 

K. O. Hart, a District Ranger in Yellowstone Park, furnishes j 
(1952: 56) a similar eye-witness account of an encounter between 
a swan and a muskrat. This time, however, the skirmish did not ? 
end in a fatality. 

The presence of pelicans, cranes, herons, geese, or muskrats, does 
not necessarily arouse aggressiveness in swans, even mated birds within 
their breeding territory. Most of these species have often been seen 
in close proximity to swans with no territorial action or other aggres- 
sive behavior resulting. Swans with young are most apt to take 
offensive action against such species. 

Despite the variations and flexibility in the defense of territories 
such action is the main factor regulating the distribution of pairs in 
an area of breeding habitat. There is no direct evidence yet that 
territorialism limits breeding numbers in all cases, although it is 
a factor in determining nesting success. 

In the complex pattern of the Red Rock Lakes marshes, territories 
are generally established where they existed the previous year, often 
with exactly the same nesting site occupied. Various observers agree 
that breeding pairs remain attached to their territories year after 
year. Oberhansley and Barrows (MS) state of their 1938 Yellow- 
stone Park observations, "At each nesting site studied, old nests were 
used and the actual time devoted to their repair was brief." Condon 
(MS), Yellowstone Park, stated, "In nesting, pairs have in most in- 
stances returned to their previous year's nests and repaired the old 
nests for re-use." 

In the contiguous marshes of the Refuge, the desirable features of 
these territories with a long breeding history are apparently recog- 
nized by other pairs who quickly appropriate areas unclaimed by 
former owners. As Albert Hochbaum (1944:78) found in his 
studies of the canvasback on prairie marshes, "I believe that 
... a delicate arrangement of terrain, if it remains relatively 
unchanged from year to year, is recognized by any individual duck 
as desirable and thus is occupied." This is apparently true of the 
trumpeter as well, since a normal population turnover would dictate 
at least some change if the selection of territories were not governed 
by inherent instincts. "What appears as desirable nesting habitat to a 
human observer may remain barren of breeding swans year after 
year, even in the face of a rising population of breeding swans, while 
territories which appear similar may be occupied. 



BREEDING 107 

During years of increasing numbers of breeding pairs at Red Rock 
I -ikes expansion has occurred chiefly into less desirable (previously 
unoccupied) habitat, not in the compression of the additional breeders 
within previously occupied territories. This may contribute to the 
low reproductive success which characterizes the years of high breed- 
ing population, but just how it does is unknown. 

If the former nesting site is missing or unusable for any reason, the 
swans select another location within essentially the same territory. 
For several years a pair nested on a bit of shallow marshland which 
protruded from an otherwise open lake shore unoccupied by other 
swans. One season this small area was destroyed by the winter's ice 
action, and with the loss of the only suitable nesting site, the otherwise 
acceptable territory has not been occupied since. 

Whatever determines the trumpeter's selection of territory, the most 
obvious result is the distant spacing of nests. For four successive 
years, during the nesting seasons of 1954-57, swan nest locations 
were plotted on aerial maps showing about 6,000 acres of water and 
marsh, about 50 percent of the nesting habitat in the Red Rock Lakes 
marsh. The Upper Red Rock and Swan Lake areas are shown in 
figure 39, and the Lower Red Rock area in figure 40. 
"This nesting distribution data is highly accurate, since the incubat- 
ino- swans are very conspicuous when viewed with a 20 X spotting 
tefescope from lookout posts on nearby hills. Most of the nests shown 
were subsequently checked by boat, and no errors were found m the 
plotting data. Any errors would probably be those of omission. A 
total of 109 nests were thus plotted during the month of June in the 
1954-57 period considered. 

Of 74 nests observed on the Upper Reel Rock and Swan Lakes 
in the 4 seasons, 57 (77 percent) were located on or within a 
few feet of previously or subsequently used sites. Three (4 per- 
cent) were occupied during all 4 seasons. Much the same situation 
prevailed on the Lower Red Rock Lake where a total of 35 sites was 
checked, with 25 (71 percent) located on substantially the same bit 
of marsh used the previous year. Here only 1 site was used all 4 years. 
Shorelines are not selected in the same proportion for nesting ter- 
ritories as is island habitat, and relatively straight shoreline is almost 
totally ignored. Whenever territories are established on compara- 
tively open shorelines, the pursuit-distance at which one of the mated 
pair will take aggressive action against an intruder is noticeably 
greater than if a bay or island is occupied. The highest concentration 
of nests per acre is found on shallow Swan Lake where the irregular 
shoreline combines with numerous stable sedge islands to provide the 
greatest variety and interspersion of water and marsh habitat. In 
1957, for instance, approximately 500 acres of water, island, and 



108 



LIFE CYCLE 




I954 = (5 I955 = Q I956=C I957=£) 

Figure 39. — Nest locations, 11)54-57, Upper Red Rock Lake and Swan Lake 
marsh. (Approximately 8,000 acres in map.) 



peripheral habitat of Swan Lake supported 7 pairs of nesting swans, 
or about 70 acres of territory per nesting pair; on deeper Lower Lake, 
where nesting sites are most frequently located atop bulrush muskrat- 
house sites, only 10 pairs nested on the 1,500 acres, or about 150 acres 
per nesting pair. The relation of shoreline to interspersed nesting 
habitat will be shown later under Population Mechanics. 



BREEDING 



109 




I954=(5 I955 = Q 1956=© 1957=© 



Figure 40.— Nest locations, 1954-57, Lower Red Rock Lake. (Approximately 

3,000 acres in map.) 

The amount of territory claimed by a mated pair of trumpeters 
may seem to be related to the quantity of food available at a given 
distance from the nest. On an open shoreline, the territory defended 
is normally greater than that of an island-nesting pair in shallow- 
water territory where a great deal of food may be available within a 
short radius of the nest. A shoreline nesting pair will, however, fre- 
quently select a small bay for the nest location where territorial de- 
fenses are limited principally to the confines of the bay itself, and m 
spite of the comparatively smaller amount of food available, the 
pursuit-distance is found to approximate that of island-nesting swans. 
So, quantity of feeding area alone does not necessarily determine either 
the size or the appeal of any given area as nesting territory. The 
arrangement of the various terrain features within the area, as well 
as the number of potential nesting pairs, are also important factors. 

The area of marsh or lake required to support a pair of trumpeters 
and their brood adequately must be but a fraction of the size of the 
territory claimed, and depends upon the quality of food produced 
as well as the quantity and availability. For instance, trumpeters 



110 LIFE CYCLE 

have nested along very narrow but productive water channels where 
food was abundant and where travel to larger waters for food would 
not appear to be necessary. They do not prefer to nest on closed 
channels or ponds of a size equal to or even larger than the channel 
if egress by swimming to larger waters is restricted, though the food 
supply within such an area may appear adequate to support a good- 
sized family. M. M. Nice ( 1941 : 469-470) , speaking generally of song 
birds, states: "Many birds, like the Song Sparrow and House Wren 
. . . , at first claim far more land than they really need; under 
pressure of competition they decrease their holdings but not beyond 
a certain point. In other cases, territorial behavior adapts itself to 
circumstances, and the type changes, so that many more pairs are 
accommodated on a certain area than would otherwise be possible." 
This may also apparently be true of the trumpeter to a limited 
extent. 

A certain amount of water space, presumably to meet flight take-off 
requirements, appears necessary within each territory, and the large 
number of potholes over the Refuge which often produce considerable 
food are not an important segment of the breeding habitat because 
of such restrictions. The marsh next to these potholes is heavily 
vegetated; thus, landing or take-off near the nest is difficult or 
impossible. 

Condon (MS) found in Yellowstone Park that: 

Of the 29 lakes used by trumpeter swan during the past 10 years, only 4 are 
smaller in size than 16 acres, and of these 4, all are 9 or more acres. This 
tendency to choose lakes of 9 or more acres indicates the need for a reasonably 
large territory for the rearing of young. . . . none of the 29 lakes used have 
harbored more than one nesting pair of trumpeter swan. Many of the lakes 
have adequately cared for several nonnesting adults, but only one nesting pair 
has been found to utilize a lake. 

Mated pairs sometimes guard territories but have never been 
observed to nest. This situation may be noted in the same territory 
from year to year, when the same pair apparently returns. It is not 
known whether these pairs have been shunted by competition to areas 
in which breeding may be inhibited for some reason, or merely 
are incapable of breeding. There does seem to be some cor- 
relation between the areas occupied in this manner and their spe- 
cific location. Too, the percentage of nonnesting pairs appears to 
increase when the breeding population is high, but no figures are avail- 
able to support this generalization. In 1957, there were 3 pairs of non- 
nested in addition to the 10 nesting pairs on Lower Eed Rock Lake, 
while the Upper Lake contained 2 pairs of nonnesters in addition to 
the 6 pairs of nesting swans. 



NE STING HI 

NESTING 
NEST SITES 

Muskrat houses furnish the great majority of swan nesting sites at 
Red Rock Lakes. On the Lower Lake where bulrush beds are plenti- 
ful, the trumpeters prefer to nest on muskrat lodges located in this 
growth, and few nests are placed on the shoreline proper. On the 
Upper Lake where bulrush is found only occasionally, all of the nests 
are located by necessity along the shore proper. 

Even while the breeding marsh remains frozen, the tops of muskrat 
houses which protrude from the snow are visited by swans and used 
as loafing sites. Whether their attraction at this season is due to their 
potential as nesting sites or to their convenient location as observation 
posts is not known, but their prebreeding season appeal is certainly 
more than just casual. 

Nests are usually located so that all or part of the site is bounded 
by a moat of water. During a year with an abnormally high spring 
water table, one swan nest was located on a solid sedge meadow, a 
measured 298 feet from the nearest waters. The dried remains of 
algae at the base of the nest, however, showed that the nest was 
actually built in a few inches of water during the runoff. The great 
majority of nests on the Refuge are placed on muskrat houses located 
on semi-floating sedge bog-mats. 

Any suitable feature located somewhat above the general level of 
the marsh terrain may be used as a nesting site. Oberhansley and 
Barrows (MS) mention in their Yellowstone report: 

Old muskrat houses furnish excellent nesting sites, and in one instance at 
Little Robinson Creek, an old beaver house was used. At Riddle Lake, beavers 
were observed in the act of converting an old swan nest into a home for them- 
selves by piling limbs and dead tree sections over it. On the other hand, old 
beaver houses form islands which were used by swans. 

Condon (MS) adds to the subject of nest locations in Yellowstone 
Park by stating: 

In some instances, small islands removed from the mainland are used as nest- 
ing sites, and islands visited at Riddle and White Lakes afforded sites for nests 
made of reeds and grasses which were piled into small mounds hollowed to 
accommodate the clutch of eggs. . . . Many nests are built on the lake shores 
in open unconcealed areas, with very little vegetative cover for concealment and 
with the nest itself consisting of only a small mound of grasses, reeds, twigs, 
and rushes. Such nests and nest sites were used at Geode, Grebe, and Madison 
Junction Lakes. 

NEST BUILDING 

Nest-building activities are shared by both sexes, though the female 
probably contributes the bulk of the effort. Nest materials on the 
Refuge consist chiefly of sedge sod and plant parts, though any ma- 



112 LIFE CYCLE 

terial nearby may be used. These materials are not moved farther 
than a neck-length away. In one instance, a pair of swans in Upper 
Red Rock Lake was observed in what seemed to be nest-construction 
activities. Both birds remained prone on a small sedge island near 
the mouth of Shambow Creek and busily picked up vegetative debris, 
rootstoeks, and plant parts with their bills, stacking these in a mound 
to the rear. This continued for perhaps 5 minutes with vigorous 
enthusiasm. Then one of the swans departed and began feeding 
some distance away, completely ignoring the activities of the other 
bird. The mate remained and continued to tug and wrestle with de- 
sirable bits of tough sedge and fibrous matter, adding these materials 
to the small heap already accumulated to the rear. It was still at the 
task when the observer left the observation post some 15 minutes 
later. 

This particular nest foundation was begun late in the season and 
was never completed. The pair of swans' which were so intent on 
its construction were not observed engaging in this activity again, 
nor did they nest at all that season, although they remained in the 
territory and defended it against other swans. 

An observation in Yellowstone Park of a pair of trumpeter swans 
engaged in nest-building activities is given by Oberhansley and Bar- 
rows (MS) : 

In repairing the nest at Swan Lake the cob and pen cooperated. The old 
nest material was rearranged by adding material from the sides of the nest 
to the top. Some of the submerged portions of the nest were brought above the 
water and other materials (chiefly sedges) near at hand were gathered and 
added. As the swans gathered sedges they would throw them to the rear in the 
general direction of the nest to be added later. This particular nest is located 
upon an old muskrat house and has been used for several years. It has a fi- 
brous, mucky base and gradually becomes coarser toward the top where the 
fibrous material blends with sedges and a few scattered leaves of willow and 
birch in the nest proper. The nest is partially surrounded and supported by 
several clumps of mountain bog birch (Betula gland nlosa) and a few willows 
(Salix, sp.) which [formed] a rather poor screen until the surrounding sedges 
had grown tall enough to serve better. All approaches are well guarded by 
water to a depth of about four feet, gradually sloping to a width of approxi- 
mately 10 feet at the bottom of the pond. The total height of the nest was 61 
inches, with top protruding 22 inches above the water line. 

Swan nests constructed on the Red Rock Lakes Refuge average 
somewhat smaller than the nest just described but the general account 
is typical. Condon (MS) records, writing again of his observations 
in Yellowstone Park : 

The type of nest constructed by trumpeter swan in Yellowstone Park varies 
greatly and the material from which it is made is governed entirely by the type 
of cover and vegetation near the nest site. . . . where an abundance of 
grasses, sedges, reeds, and rushes is available, nests of large proportions are at 
times built. Such nests have been built at Swan and Trumpeter Lakes. Their 



NESTING 11^ 

krge size is, in all probability, clue to the necessity of keeping a dry nest and 
£ dire to have the nest removed from dry land by a strip of water. Nests 
f tluTIype have been constructed only where vegetative cover and water depths 
L ave been favored with suitable sites. 

MacFarlane (1891: 425) gives a general description of the nests ol 
Lrctic trumpeters which were observed near Fort Anderson (east ol 
he Mackenzie River delta) , stating : 

Several nests of this species were met with in the Barren Grounds on islands 
n Franklin Bay, and one containing six eggs was situated near the beach on a 
Cg knoll, "it was composed of a quantity of hay, down, and feathers 
ntennixed, and this was the general mode of structure of the nests of both swans. 

Only a trace of down and a few molted feathers have been observed 
about the nest areas of trumpeters on the Red Rock Lakes area. The 
material of which the nest bowl proper is constructed differs little 
from that of which the upper part of the nest is fashioned, the finer 
leaf and stem parts of the sedges being the usual material used there 

For a month, more or less, after the eggs are hatched, the nest 
becomes a favorite site for the brooding of the cygnets at night and 
loaiino- by day. By midsummer, as a result of these activities, the 
nest is an odoriferous mound of decaying vegetation and defecation. 
The hatched egg shells and unhatched eggs, if any, are buried to a 
depth of several inches with fecal deposits in addition to the vegetative 
material which has become mashed down from the nest bowl or 
piled on the site by muskrats. Of course the nest by this time 
is unrecognizable, with usually only a few molted white feathers 
nearby to indicate that the unattractive heap was once a tidy swan s 
nest. 

EGG DESCRIPTION 

Trumpeter eggs are somewhat granular in texture, elongated ovoid 
in shape, and off-white in color when freshly laid, becoming nest- 
stained a brownish color a short while after incubation. 

In 1955, 109 eggs representing the complete clutches of 21 nests 
were measured for length and greatest diameter (see appendix 
3). Longitudinal measurements ranged from 123 mm., which was 
the longest and which had a diameter of 74 mm., to the shortest, 
which measured 104 mm., there being 4 of the latter (2 in a clutch 
of 6, 1 in a clutch of 5, and 1 in a clutch of 4) which were 69 mm., 
71 mm., 70.5 mm., and 70.5 mm., respectively, at their greatest 
diameter. 

The egg of greatest diameter measured 77.5 mm. and was 109 mm. 

long, being one of a clutch of 6, while the egg of shortest diameter 

was 68 mm. and measured 113 mm. long, being one of a clutch of 8. 

The average length of the 109 eggs checked was 110.9 mm., while 

the average diameter was 72.4 mm. It is interesting to note that 



114 LIFE CYCLE 

the average trumpeter egg from the Red Rock flock is noticeably 
smaller, 110.9 x 72.4 vs. 121.3 x 77, than the average measurements 
of 3 eggs (from 2 nests) obtained from the Copper River, Alaska, 
flock by Melvin Monson (1956: 445). The smaller egg-size ratio of 
the Red Rock Lakes trumpeters also holds true when compared with 
that reported from the Kenai trumpeter flock (Dave Spencer, cor- 
respondence). The average size of the eggs obtained at Red Rock 
Lakes compares very closely with that reported by A. C. Bent 
( 1925 : 297) in the measurement of 25 trumpeter eggs. 

CLUTCH SIZE 

Sizes of completed clutches of eggs in various trumpeter nests on the 
Red Rock Lakes Refuge show this annual variation: 1949, 12 nests 
containing 3 to 9 eggs each held a total of 61 eggs for a mean of 5.1 
eggs to the nest; 1951, 13 nests containing from 4 to 7 eggs held 73 
eggs for a mean of 5.6; 1952, 17 nests containing from 2 to 7 eggs held 
88 eggs for a mean of 5.2 ; 1955, 32 nests containing from 4 to 8 eggs 
held 157 eggs for a mean of 4.9. The 74 completed clutches observed 
over an intermittent span of 7 years contained a total of 379 eggs, the 
mean being 5.1 eggs to the nest. 

These figures agree well with the clutch size noted by MacFarlane 
(1891 : 425) among Arctic-breeding trumpeters, as he states, "It usually 
lays from 4 to 6 eggs judging from the noted contents of a received 
total of 24 nests." 

EGG LAYING 

The rate at which eggs are laid and the length of the incubation 
period were not studied on the Refuge since gathering such data on 
a single nest would unnecessarily disturb several nesting pairs. 
Other observers have made some pertinent notes. 

H. F. Witherby et al. (1939 : 177) state of the mute swan, "Accord- 
ing to O. Heinroth, eggs laid on alternate days; confirmation de- 
sirable;" which H. Poulson (1949) obligingly furnished, writing, "The 
female Mute Swan in the Zoo laid 5 eggs in 1947 and 4 eggs in 1948 on 
alternate days." 

Witherby et al. also report that whooper swans in confinement laid 
eggs on alternate days (1939: 170). For lack of specific information 
on the subject, the production of about 1 egg every 48 hours is assumed 
to be normal for the trumpeter also. 

INCUBATION 

Although Witherby et al. (1939: 177) report that in the case of 
the mute swan the male assists the female in incubating the eggs, 
this has never been observed or reported in the trumpeter. Poulsen 



NESTING 115 

(1949) corroborates Witlierby's mute-swan statement but adds that 
in the case of the whooper swan the male did not participate in 
incubation duties. 

Several records have been compiled on the length of the incubation 
period. Dr. Ward Sharp, writing of 5 trumpeter swan eggs which 
were set under bantam hens, records in a Bureau report: 

The little hens could not keep the eggs warm, faithfully as they tried. One 
end of the egg was always cool. A few days before the [other] eggs hatched 
(about 3 or 4) two eggs became warm and hatched soon afterward. The 
incubation period was about 33 days. 

Oberhansley and Barrows (MS) in their Yellowstone Park obser- 
vations also report an incubation period of 33 days, but Aimer Nelson, 
formerly Refuge Manager at the National Elk Refuge, writes of a 
period of 36-37 days. Inasmuch as the incubating period lasts more 
than 4 weeks, the variation of a few days' time would be expected 
under the varying conditions found in the wild. A normal incubation 
period of from 33 to 37 days seems reasonable. 

Again referring to the much-studied mute swan, TVitherby et al. 
(1939: 177) state, "incubation begins before completion of clutch in 
some cases." This tendency may also be present in the trumpeter to 
some extent, as unhatched eggs containing well-developed embryos 
are sometimes found in nests from which the female has departed 
with her first-hatched cygnets. 

HATCHING DATA 

The Red Rock Lakes-Yellowstone region has short, cool summers. 
Thus the average hatching date is important, since the nesting lakes 
and marshes freeze over rather early, dooming any late cygnets which 
are still flightless. 

Of 12 nests closely watched in 1949, 9 had produced broods between 
June 15 and 21. The first brood was seen on June 15, and the last 
nest under observation apparently hatched about June 26. 

In 1952 hatching was much later. On 7 nests checked that year, 
only 1 had produced young by June 17, 3 more had hatched by June 
28, and the remainder had hatched by July 3. The average hatching 
date in 1952 was later than at any time during the period 1948-52, 
presumably because mating was delayed by abnormally high water 
levels during the initial spring breakup period. Fortunately a long, 
unusually warm autumn followed, and no cygnet mortality seemed to 
result from this late hatch. Normally, hatching on the Red Rock 
Lakes marsh occurs between June 15 and June 25. 

In 1946 Sharp reported seeing the first cygnets on June 10, and 
earlier A. V. Hull mentioned, "The first egg date that we have is April 
26. Both of these records indicate unusually early nesting. 



116 LIFE CYCLE 






Condon (MS) records a late-nesting case in Yellowstone Park in 
1940. The trumpeters were not seen on the lake until May 30, and no 
cygnets appeared until July 18, probably too late to survive the fall 
freeze. 

Judging from the egg-laying rate, average clutch size, period of 
incubation, and the known hatching dates of many nests, egg-laying 
normally commences on Refuge marshes shortly after May 1 and most 
clutches are completed by May 15. Thus, nesting is underway in many 
cases before the ice has disappeared. This is many weeks before 
the ducks commence to nest in earnest. 

David Spencer (correspondence) states that on the Kenai National 
Moose Range in Alaska in 1957 most of the hatching was completed by 
June 25. This indicates that the nesting seasons on the Kenai and in 
the Red Rock Lakes marshes are coincidental to a surprising degree, 
the great difference in latitude (16°) apparently being well compen- 
sated for by the contrast in elevation (sea level vs. 6,600 ft.) 

RENESTING 

All the white northern swans raise a single brood each year, if 
successful in nesting. No instance was found in the ornithological 
literature to the contrary, either in the wild or in captivity. In cap- 
tivity, both the black swan and the black-necked swan will produce 
2 or even 3 clutches of eggs in a single year if the eggs are removed 
successively after each clutch is completed. If a brood is raised by the 
parents from the first clutch of eggs laid, this precludes a second nest- 
ing for that year. 

An unsuccessful attempt, sponsored by outside interests but per- 
mitted by the Service, was made in 1944 to increase trumpeter swan 
production artificially at the Red Rock Lakes Refuge. In this case 
a clutch of 5 eggs was removed from a wild trumpeter nest on the 
Refuge and incubated artificially. Jean Delacour (1944: 135-136) 
states of this attempt, "We had hopes that the robbed pair of swans 
would nest again, but they refused to do so, and the only supposed 
advantage of taking eggs rather than cygnets was thus denied." 

One valid case of renesting may have been found. While on a 
routine inspection of swan nests in Refuge marshes, I encountered 
2 fresh nests within 25 feet of each other. In one nest, which was 
deserted, there were the broken pieces of 5 eggs, which had apparently 
been destroyed not too long previously, while the active nest located 
nearby held 2 eggs. The smaller clutch was being incubated, and in 
addition 3 eggs were floating near the new active nest. 

Presumably when the eggs in the first nest were destroyed, possibly 
by predation, the clutch was in the process of completion, and the 
pair promptly renested close at hand. The 3 floating eggs may have 



CYGNET DEVELOPMENT 117 

been lost in the interval between the destruction of the first clutch and 
the completion of the second nest to the point of being able to hold 
eggs. Thus, a total of 10 eggs was apparently laid by the one 
female, as an association of 3 breeding swans has never been observed 
on the Refuge, and territorialism would prevent the nesting of 2 
pairs so closely. 

In the case of the failure of a trumpeter pen to renest, as reported 
by Jean Delacour, the eggs may have been removed from the nest 
after egg laying had stopped and incubation was actually under way. 
If the eggs had been quietly removed as they were laid, leaving one 
or two in the nest to restrain the inclination to desert or nest else- 
where, a much greater production by the female may have resulted. 

CYGNET DEVELOPMENT 

In common with other birds, the young of the trumpeter are 
equipped with an egg tooth believed to help in breaking out of the 
shell. Oberhansley and Barrows (MS) noted, "This sharp point is 
a special adaptation designed only for pipping the shell and dis- 
appears sometime after hatching. . . . The shells average 1 milli- 
meter in thickness and are very strong." 

Only one record of the weights of newly hatched cygnets seems to 
have been made. Regarding this, Sharp, referring to the cygnets 
artificially incubated, states, "The two cygnets weighed 7 and 7V 2 
ozs. respectively when hatched on June 19, 1944." No other figures 
are available, but these data are believed representative. 

How much time is spent by the wild pen brooding the newly hatched 
cygnets on the nest before leading them to water is not known. The 
presence of unhatched eggs or cold inclement weather might prolong 
brooding, while a complete hatch and warm weather might shorten it. 

F. E. Blaauw (1904: 74) made a note regarding the persistence of 
brooding in his captive trumpeters, stating, "During the first days of 
the life of her chicks, the old female trumpeter often retires to her 
nest for hours together, warming them under her, and she continues to 
do this during the night for a long time." 

Much attentiveness is shown the young cygnets by the parents, with 
both adults usually present. How aggressively the parents might 
defend their young from predators is not known. The treatment of 
muskrats by aroused trumpeters suggests that very vigorous action 
is possible. But the parents' close supervision of the young gives very 
few opportunities to predators. 

When the presence of humans threatens the swan family, the adults 
will usually desert their young without protective action, with little 
if any reduction of the normal escape-distance. If time and oppor- 



118 



LIFE CYCLE 



tunity allow, the adults try to lead their young into dense growths of 
emergent aquatic plants, leaving the scene unobtrusively shortly after 
hiding their cygnets. The cygnets will then usually remain quiet and 
well hidden, though sometimes flushing from their hiding place when 
deliberate search is commenced in the immediate vicinity. Scott and 
Fisher (1953: 209) report that among wild Icelandic whooping swans: 

Older cygnets would feign death when handled, having their necks in a lifeless 
attitude. This was particularly noticed in a brood estimated to be about two 
weeks old, and again in one of about four and a half weeks. . . . 

The behavior of parent whoopers with cygnets was also variable. Some 
deserted their brood and flew away at a range of several hundred yards. Others 




Figure 41. — Newly hatched trumpeter cygnet. Lower Red Rock Lake; egg in mid- 
foreground being pipped. 




Figure 42.— Close-up of trumpeter cygnet showing fine, grayish-white down. 

remained to protect their young and were photographed at less than thirty 
yards. 

When not disturbed, parent trumpeters spend most of the time 
swimming, feeding, or loafing, with the more active cygnets busily 
moving between or immediately around them. Close family ties 
result in a tightly knit family formation which reduces vulnerability 
to predators. 

This characteristic, coupled with a pair's territorial proclivities, 
virtually eliminates the possibility of two or more broods combining, 
especially in the younger age classes when both factors are more 
strongly expressed. On one occasion, however, during the August 29, 
1955, aerial census of the Refuge, a single pair was seen accompanied 

469660 O— 60 9 



120 LIFE CYCLE 

by 10 cygnets. Specific nest checks in that vicinity earlier in the year 
to determine clutch size establishes that in this instance the abnormal 
size was due to an association of at least 3 broods. The actual circum- 
stances surrounding- this unusual case are unknown. (Earlier ob- 
servations of the 4 broods nearest this occurrence had revealed that 
hatching originally produced broods of 5-3-3-1.) Sharp, in a trum- 
peter census report in Refuge files dated June 28, 1943, also reports a 
brood of 7 cygnets which he believed was a compounding of 2 broods 
of 4 cygnets seen in the same general area 12 days earlier, so such 
cases apparently do occur. 

Although swan brood amalgamation appears to be rare, older broods 
of preflight age sometimes contain individuals of marked size differ- 
ences. These cases probably represent variations in individual growth 
rates rather than examples of brood combinations, though the latter 
may occur to a limited degree. Hochbaum (correspondence) notes 
that the two Montana cygnets raised at Delta showed marked differ- 
ences in size from the start. The larger bird proved to be male, the 
smaller, female. 

The rapid growth rate of this species is assisted by the long day- 
light hours of summer which the young spend feeding in their fertile 
marsh environment. It was previously noted that the two cygnets 
hatched in captivity on the Refuge on June 19 weighed 7 and 7y 2 
ounces. On September 2 these same birds weighed 15*4 and 15% 
pounds respectively, a thirtyfold increase in 75 days. In the wild, 
growth must be at a comparable rate, as I have weighed 19-pound 
cygnets of preflight age captured for banding in early October. 

Because of the early freeze-ups on their breeding grounds, the 
average age at which cygnets are capable of flight is important. 
Oberhansley and Barrows (MS) report that in one case a cygnet 
hatched June 23 was observed on its initial flight on October 9, a 
period of 109 days. James R, Simon (1952:462) checked the flight 
capabilities of 3 trumpeter cygnets, hatched on June 16, 1951 in 
Grand Teton National Park, from October 4 to 16. Two cygnets 
could fly by the 14th, and 2 days later, 122 days from hatching, the 
whole family departed the pond on the wing. The cygnets transferred 
from the Red Rock Lakes Refuge to the Delta station in 1955 made 
initial flights when about 13 weeks old, a period of 91 days. A de- 
velopment period of from 100 to 120 days would apply to the average 
Red Rock Lakes cygnet, as young hatched about June 20 are fre- 
quently seen in flight before October 10, though seldom before Sep- 
tember 20. There is considerable variation in the individual develop- 
ment so that some cygnets, though hatching on time, might be flightless 
and vulnerable at the final freezeup. This has occurred as early as 




Figure 43.— Swan family at loafing cite. Grebe Lake, Yellowstone National Park. 
The special foot position of the adult is commonly seen with mute swans. This 
brood is approximately a month old. 

October 28 for the Red Rock Lakes and marsh, though the 16-year 
average final freezeup date compiled during the 1938-53 period, is 
November 8. 

Though family ties remain strong, the "apron strings" of the par- 
ents tend to loosen as the young grow older. On one occasion, a brood 
of cygnets was left alone for about 30 minutes while the parents moved 
briefly to another part of the lake. This occurred on August 30, when 
the young were half grown. 

Only two notes exist regarding the duration of family ties after 
the offspring's first winter. I once noted two immature trumpeters, 
in their initial flightless molt during July, closely together when a 
large loosely scattered nonbreeding flock was approached. These 
two individuals were so attracted to each other that the long pursuit 



122 LIFE CYCLE 

by boat did not separate them, as usually occurs among unrelated 
swans. Both were subsequently captured for banding, and were 
found to have been banded the previous autumn when they were pre- 
flight cygnets of the same brood. 

This observation suggests that associations of brood mates may 
persist for some time after the offspring are left by their parents 
upon the approach of the breeding season, at least until their first 
flightless molt. This period would appear to be confirmed and even 
extended by R. H. Mackay (1957: 339), who reported that three 
brood mates banded in 1955 were shot virtually together in Nebraska 
in late October 1956. Family ties in the trumpeter are apparently 
strong. 

FOOD 

Swans spend long hours in their endless quest for food and consume 
enormous amounts of succulent green vegetation when given an op- 
portunity, all parts of the various aquatic plants being taken. With- 
erby et al. (1939) mention that pondweeds, buttercups, mannagrass, 
eelgrass, waterweed, muskgrass, and clover, besides others, constitute 
the food of the Eurasian species of swans. Hilprecht (1956: 96-100) 
also mentions their foods and feeding habits. Many of these plant 
forms are an important part of the diet of our native swans as well. 

The findings from several unpublished food studies of the trumpeter 
in its Yellowstone and Red Rock Lakes environment, plus the labora- 
tory analysis of stomach contents of trumpeter casualties and pertinent 
notes by other observers are combined here to furnish a summary of 
the food of this species. Much more remains to be learned, partic- 
ularly of their winter food habits and requirements. 

FEEDING HABITS AND FOOD OF YOUNG 

From the beginning, downy cygnets are much interested in food, 
and are very active in their quest for it. For a long period follow- 
ing hatching, the swan family feeds in a tightly knit group, the adults 
usually remaining near each other in the shallow water while the 
cygnets swim busily about between them, frequently with coots or 
ducks intermixed, seeking the morsels which are stirred up by their 
parents. 

Writing from close-hand observations of swan families feeding 
on lakes in Yellowstone Park, Oberhansley and Barrows (MS) 
describe these early feeding activities : 

The first food of the cygnets was presented [brought up] to them on the sur- 
face of the water by the adults and consisted principally of aquatic beetles, 
with some insects and crustaceans, together with small quantities of the white, 



FOOD 123 

tender basal part of sedge. The young birds were very adroit in snapping 
up morsels that floated to the surface near the parent birds as they fed from 
the bottom or scratched food loose. Droppings in each case studied averaged 
over 95% animal matter for the first three weeks. Plant life became increas- 
ingly important with age and varied from sedges to water milfoil, pond weed 
and other aquatic plants as the season progressed. Beginning in August 
snail shells (Lymnaea stagnalis jugularis, Say) began to appear in the cygnet 
droppings. These snails are abundant near the east shore of the lake. 

Condon (MS) reports: 

The cygnets that have been observed during their first two weeks after 
hatching were seen to feed in very shallow waters of six inches to one foot in 
depth, and when feeding where the water was deeper, they were gathering the 
food stuffs brought to the surface from the bottom by their parents while feed- 
ing. In shallow water, where they can reach the bottom or the vegetation 
growing up from the bottom, their feeding is carried on by the simple process 
of stretching their long necks down and securing the food by working over the 
plant and animal matter at hand and selecting those desired. 

That swan cygnets tended to cut down on the animal matter in their diet 
after about five weeks was noted by their tendency to feed in deeper waters 
and to be feeding primarily on vegetative foods. Droppings collected at Grebe, 
Riddle, Geode and Swan Lakes all gave evidence of a preponderance of vegetable 
foods and very little animal. Those animal forms eaten were undoubtedly of 
the softer forms, for evidence of them did not appear in the droppings collected 
at resting areas. 

These observations agree generally with the feeding habits of the 
cygnets at the Red Rock Lakes Refuge, though close observations 
of feeding families of swans are impractical in the vast marsh sys- 
tem. On one occasion five Refuge cygnets from a single brood and 
one cygnet from another brood were recovered soon after they 
had died, presumably of exposure, and a determination of their 
stomach contents was made by the Fish and Wildlife Research 
Laboratory in Denver, Colorado. These data, as well as information 
concerning the stomach contents of a predator-killed cygnet in Yellow- 
stone, are presented in the Appendix, part 4. In the case of the 
Refuge cygnets, practically no animal matter was present, with leaf 
and stem fragments making up from 80 to 100 percent of the stomach 
contents, and the seeds of sedge and spikerush also present. In the 
single Yellowstone cygnet of comparable age, fresh water fairy 
shrimp (Eubranchipus sp.) were represented as well as sedge frag- 
ments. 

The young of the trumpeter have been successfully raised in cap- 
tivity in at least two instances, without natural parental care, and 
under varying conditions. This indicates that their early food habits 
or requirements are not critical or fixed. The first case was demon- 
strated on the Red Rock Lakes Refuge when two were raised by 
Sharp, as mentioned previously, while a pair of month-old cygnets 



124 LIFE CYCLE 

were reared at the Delta Waterfowl Research Station after having 
been transferred from their Red Rock Lakes Refuse natal en- 
vironment. 

In the case of the Refuge cygnets artificially raised by Sharp, 
bantam hens served as both the method of incubation and source 
of parental care. This attention inspired no apparent attachment to 
the foster parent, nor to the humans helping care for the cygnets. 
Initial foods fed in this case were broiler chow, milk curd, and dande- 
lion leaves. Wheat and lawn clippings were added to the diet later. 
Development of these birds appeared comparable to those in the wild 
except for a weakness of the legs, which became apparent after each 
weighed about 10 pounds, and from which they subsequently 
recovered. 

At the Delta Station the month-old trumpeters were successfully 
started on a ration consisting of duck grower pellets, with a constantly 
available supply of lesser duckweed (Lemiia minor) serving as the 
green vegetation supplement. Later the leaves and stalks of arrow- 
head were introduced and were much relished, as much as a bucketful 
being consumed by a single cygnet in a day. Upon the exhaustion 
of the local supply of arrowhead, the leaves of cattail were substituted 
with equal success. 

The young developed normally at Delta except again for an ap- 
parent weakness of the leg bones, which supported the body with 
difficulty out of water as the birds rapidly gained weight before the 
bone structure hardened. This weakness was eliminated by per- 
mitting the cygnets access to a supply of swimming water, where 
their body weight was supported much of the time by water instead 
of the individual's legs on the hatchery's cement floor. 

FEEDING HABITS AND FOOD OF OLDER CYGNETS, 
IMMATURES, AND ADULTS 

Even before the development of contour feathers changes the ap- 
pearance of the wild downy cygnets, they seem to rely progressively 
less on food material provided by the adults and become more in- 
terested in obtaining it by their own efforts. At this stage any em- 
phasis on the high protein diet of aquatic insect and crustacean life 
shifts to vegetable foods. Feeding is still accomplished very much 
as a family unit, and after the cygnets are 2 or 3 months old their 
food-gathering actions indicate that their diet is approaching that 
of their parents. 

While the feeding of the cygnets appears to be confined solely to 
water, that of the immature and adult, though predominantly aquatic 
in nature, may include a limited amount of feeding or grazing upon 
land. Audubon (1838: 540) noted, "This swan feeds principally 



FOOD 



125 



Ly partially immersing the body and extending the neck under 
vvater. . . . Often, however, it resorts to the land, and then picks 
it the herbage, not sidewise, as Geese do, but more in the manner of 
Ducks and poultry;' Swans have rarely been observed feeding on 
land on the Refuge. 

I Trumpeters most frequently feed in shallow-water areas where 
Ley are able to use their long necks to the best advantage, or if deeper 
water is encountered they may "tip up" in the manner of puddle 
ducks. Although they are quite capable of diving and swimming 
under water, they apparently resort to diving solely to escape when 

flightless. 

Oberhansley and Barrows (MS) contribute an interesting note on 
the feeding of a pair of trumpeters within Yellowstone Park, writing, 
"A typical observation on September 18 of a pair of swans feeding on 
a small lake near Madison Junction showed 2 birds suspending 15 and 
13 times respectively in 30 minutes. . . . There was no system of 
timing, sometimes both birds submerged alternately and again simul- 
taneously; neither appeared to be on guard." This description also 
typifies the feeding of trumpeters in the shallow water areas of the 

Refuge. 

Tubers and rhizomes of the various aquatic plants are a staple food 
item, along with the stems and leaves of such plants, and the swans 
spare no effort in excavating for these starchy plant parts. Their 
powerful legs, large webs, and prominent toes are especially efficient 
in stirring up the soft mud of their shallow marsh environment. 
Great holes, sometimes over a foot deep and several feet in diameter, 
on the shallow bottoms of the Red Rock Lakes marsh bear witness to 
this method of exposing food materials. 

Erickson describes the feeding of the Malheur captive flock 
(correspondence) : 

Swans are also disposed to feeding along the shorelines and river banks, where 
they gouge out and undermine the banks in search for roots and shoots. Their 
stout, muscular necks and heavy bodies aid them in performing major ex- 
cavations when undercutting the banks. In the swan pond they have been 
successful not only in eliminating much of the hardstem bulrush growth, but 
also in the destruction of the tough rootstocks of the tule. 

Many references to the food habits of the native swans are to be 
found in the early literature. It will be recalled that Lewis and 
Clark noted that both species of swans fed much on the root of the 
"wappatoo" in the Columbia River. 

Later observers confirmed Lewis and Clark's statement, 
that the swans fed much on the root of the wapato or duck 
potato. George Barnston of the Hudson's Bay Company noted that 
(1862: 7831), "In the scarcity of their favorite food, the tubers of 



126 LIFE CYCLE 

the Sagittaria sagittifolia [probably S. latifolia] , they have recourse 
to the roots of other plants, and the tender underground runners of 
grasses in the higher latitudes." J. C. Hughes (1883:283) also 
recorded that along the Columbia River the favorite food of the 
trumpeter was the wapato, listing the species as Sagittaria variabilis 
[probably S. latifolia], stating that a Mr. Allard observed a trum- 
peter which had been strangled by a large tuber of this plant which 
had become lodged in its throat. 

A. C. Bent (1925: 286) quotes Major Bendire as stating that about 
"20 small shells, perhaps half an inch in length" were found in a 
stomach of a whistling swan, while E. S. Cameron was noted to have 
observed this species to feed, presumably in Montana, as follows: 

The swans were engaged in feeding upon the soft-shelled fresh-water snails 
which abound in this lake and explain its great attraction for them. During the 
several days that I watched the swans I never saw them eat anything else, but 
doubtless they pick up vegetation as well, being accustomed to walk about in the 
grass at the mouth of Alder Creek. Marsh Lake is so shallow (only 2 feet deep 
over most of it, and 4 feet in the deepest part) that the long-necked birds can 
generally reach the mollusca without much tilting of their bodies in characteristic 
swan fashion. 

In a recent study of the food habits of whistling swans wintering 
in the Chesapeake Bay region, Robert E. Stewart and Joseph H. 
Manning (1958: 209-210) examined the gullet and stomach contents 
of 19 birds. They report : 

In the series studied, submerged aquatic plants furnished 100 percent of the 
food in fresh estuarine waters, 60 percent in brackish waters, and 47 percent in 
estuarine marsh ponds. Mollusks [chiefly long clams and Baltic macomas], 
although comprising 31 percent of the food in brackish estuarine waters, were 
not listed for other types, while rootstalks and stems of emergent marsh plants 
were important only in the estuarine marsh ponds. 

Witherby et al. (1939: 177) mention that the mute swan has been 
noted to eat "small frogs and toads, tadpoles, worms, fresh-water 
Mollusca, occasionally small fish {Alburnus) and insects with their 
larvae" in addition to its more staple diet of plant foods. Fish were 
also recovered from the gullets of several trumpeters which died at 
the Kellogg Bird Sanctuary in Michigan (Dr. Miles Pimie, corre- 
spondence), but it is doubtful that these represent a part of their 
customary diet. 

From this, and data to be presented later, it appears that the 
trumpeter will eat a variety of vegetable foods. Mollusca and 
vertebrates do not occur as a staple fare but are apparently eaten 
when readily available. 

Several observers have contributed to the knowledge of the specific 
foods of the trumpeter in Yellowstone Park. Oberhansley and Bar- 
rows (MS) state: 



FOOD 127 

Identification of two principal plant foods (from small lake near Madison 
Junction) was positive and later determined as pond weed (Potamogeton fili- 
formis) and water milfoil. . . . The droppings of adult birds consist almost 
entirely of vegetable matter, with an occasional large quantity of grit. . . . 
Four species of sedges used as food have been identified from Swan Lake, the 
must important of which is Carcx rostrata, which serves also as an excellent 
cover. . . . Seeds of various aquatic plants, principally sedges, were found 
in the stomach of the swan that died on Trumpeter Lake. One small snail shell 
and traces of crustaceans were also present. 

Condon (MS) gives a more detailed list for the swans in Yellow- 
stone Park: 

The principal plant types on which the trumpeter swans have been observed 
to feed as identified by examination of plants taken at various places feeding 
occurred are: Pondweed (Potamogeton, sp. undet), Water Milfoil (Myriophyl- 
lum, sp. undet.), Musk grass (Chara, sp. undet.), Waterweed (Elodca cana- 
densis), Duckweed (Lemna trisulca), Tules (Scirpus, sp. undet.), Spatterdocks 
(Nymphaca polysepala [Nuphar polgsepalum]), Bur-reeds (Sparganium augusti- 
folium), Wapato (Sagittaria cuncata). 

In watching the trumpeter swan feeding it was found that they eat the foliage 
of each of the above-named plants. The opportunity to examine the stomachs of 
the adult trumpeter swan for food content has been limited and only one has 
been secured in the past 2 years. This contained all vegetative matter, parts of 
which were identified as being: Muskgrass (Chara), Waterweed (Elodca), and 
Duckweed (Lemna). This stomach contained an ample quantity of quartz and 
obsidian sand grit. 

Collections of droppings . . . from Geode, Grebe, Swan, and Madison Junc- 
tion Lakes . . . were so nearly 100 percent vegetative in character that no 
animal matter was discerned. . . . 

Another interesting item was the abundance of the seeds of the spatter- 
dock ... in the droppings collected from Grebe and Madison Junction Lakes 
where large beds of these plants abound. . . . From one dropping taken at 
Grebe Lake on September 21, 1939, a total of 272 seeds of the spatterdock were 
secured and 497 small seeds from an unidentified plant. All the droppings at 
resting areas at Grebe Lake gave evidence of heavy seed eating. The failure 
to digest many of these seeds was of interest. All seeds were highly polished. 
That the preponderant percentage of the foods eaten by swans are of plant origin 
was evidenced in droppings examined at resting areas on all lakes visited. 

Dr. O. J. Murie, former U. S. Fish and Wildlife Service biologist, 
collected 17 samples of swan droppings from the shores of Grebe Lake 
on September 2, 1943, and analyzed them. Prominent plant parts in 
these remains included filamentous green alga, sedge spikes, pond 
weeds, and wokas, with traces of animal matter (Tricoptera). 

In 1938, A. V. Hull, who was manager of the Red Rock Lakes Refuge 
at the time, found the fresh carcasses of two trumpeters on the Refuge 
in December and June and sent the stomachs to the Service's Denver 
Wildlife Research Laboratory for analysis. In both, the tubers of 
sago pondweed dominated. In the December-taken bird, with the 
fragments of rootstocks, they comprised 100 percent of the contents 



128 LIFE CYCLE 

(443 tubers) while in the case of the June-taken bird 597 tubers con- 
stituted 90 percent of the contents, with leaf and stem material of the 
same plant bringing the total up to 96 percent. (See appendix 4 for 
an analytical breakdown of data.) 

Earlier, Hull submitted 3 trumpeter swan stomachs taken from 
lead-poisoning fatalities at Culver Springs April 3-7, 1937, on 
the Red Rock Lakes Refuge. White water buttercup (Ranunculus) 
and mosses (Amblystegium and Fissidens) were the most prominent 
plants represented in the 100-percent vegetable contents. ( See appen- 
dix 4 for further details.) 

Most of the foregoing information has concerned the food habits 
of the trumpeter other than during the months when they are restricted 
to the limited areas of open water in winter, about which little is 
known. 

An aquatic plant survey along portions of Henrys Fork of the Snake 
River used by wintering trumpeters indicates that the following plants 
are available (in the probable order of their importance to the 
swans) : Sago pondweed, leafy pondweed, water milfoil, white water 
buttercup (Ranunculus aquatilus), marestail, water moss (Fonti- 
nalis), clasping-leaf pondweed, and mannagrass (Glyceria elata) . 

Munro (1949: 712) notes that in British Columbia their winter 
diet includes the following items: 

From observations of feeding trumpeter swans, it has been determined that 
the following foods are attractive to them, namely, the seeds of yellow pond 
lily, Nuphar polyscpala, and water shield, Brasenia Schreberi; the tubers of 
sago pondweed, Potamogeton pectinatus, and the stems and foliage of similar 
species. Late in the season, when the supply of more desirable foods is low, 
the birds eat water moss, Fotitinalis sp., the stems and roots of sedges, Scirpus 
sp., and whatever other aquatic vegetation can be secured. The gullet of an 
adult male that died from lead poisoning at Itatsoo Lake, Vancouver Island, in 
February 1938, was packed with stems of grasses and sedges ; the stomach 
contained 12 seeds of Ceratophyllum demersum. In an examination of the 
stomach contents of seven adults and two juveniles, killed by lead poisoning on 
Vancouver Island in January 1946, Dr. I. McT. Cowan identified the following 
items, namely, stems, leaves and roots of grasses, sedges (Carex sp.), and rushes 
(Juncus sp.), seeds of Polygonum sp., Carer sp., and wild cherry (Prunus 
emarginata) . Another from Steveston, at the mouth of the Fraser River, had 
eaten grass roots, and seeds of a spike rush (Eleocharis sp.). 

Swans kept in captivity adjust their food habits to whatever is 
available. Their bulky-green-food requirements usually cannot be 
entirely met with the limited supply of aquatics available, and some 
sort of supplementary feeding becomes necessary. At the Malheur 
Refuge, where a number of trumpeters Ave re kept for several years on 
a fairly large spring-fed pond, a daily ration of barley and wheat 
supplemented their natural green foods, which they obtained either in 
the pond or by gouging roots and rhizomes from the banks. 



FOOD 



129 



At the Delta Waterfowl Research Station, the green-food require- 
ments of the trumpeters in summer are met largely from the endless 
supply of duckweed which floats through the pens, as well as the 
occasional addition of the leaves and stalks of arrowhead. Their main 
diet requirement is provided in the form of wheat, soaked and fed in 
automatic dispensers. In the winter they are fed wheat, duck pellets, 
and lettuce trimmings. 

Bulk green foods fed in quantity may be important when reproduc- 
tion is desired in captive birds. This point was stressed by Mr. C. L. 
Cunningham, a successful breeder of swans at Woodinville, Wash- 
ington, who advised that his captive swans, including the whooper, 
mute, black-necked, and black, daily consumed enormous quantities of 
fresh lawn clippings before the egg-laying season and that this 
plentiful supply of bulk green food was all-important in bringing 
his birds to a breeding condition. In this regard, the Heinroths 
(1928: 149-150) also mention that whooper swans in zoos rarely pro- 
duce fertile eggs. They cite Blaauw's success in propagating trum- 
peters, giving them abundant greens in the form of watersoldier 
(Stratiotes) before the nesting season. Similar results were obtained 
with whoopers in- Berlin by feeding cabbage and other greens. The 
nutritional requirements of the swans are unknown. This may ex- 
plain in part the generally poor breeding success experienced more 
recently with the trumpeter in captivity. 

A number of swan studies have been conducted by various govern- 
ments over the years at the insistence of fishermen and sportsmen who 
accused these birds of interfering with livelihood or sport, In Ger- 
many, a food habits investigation of mute swans was conducted by the 
Reich Health Administration before World War II (Hilprecht, 1956 : 
96). In 1951, Danish sportsmen instigated a study of mute swans in 
their country, claiming wild duck production was being adversely 
affected by the aggressive behavior of the swans on mutual nesting 
habitats (Paludan and Fog, 1956). 

In the United States, several swan studies have been carried out 
from time to time as the result of complaints to the Federal Govern- 
ment. As early as 1919, W. F. Kubichek conducted a field survey in 
Currituck Sound, N.C., for the U. S. Biological Survey, investigating 
charges that the large number of whistling swans which winter in this 
area were consuming most of the natural aquatic foods, leaving little 
or none for the ducks. A study with a somewhat similar objective is 
currently in progress on the Bear River Migratory Bird Refuge, Utah, 
under the direction of the Cooperative Wildlife Research Unit at 
Logan. The study of the food habits and status of whistling swans 
in the Chesapeake Bay area by Stewart and Manning (1958 : 203-212) 
was undertaken in part to obtain information to answer the charges 



130 LIFE CYCLE 

by shellfishers that these birds were making inroads in the commercial 
clam beds. 

In all cases, either in this country or abroad, the swans were sub- 
stantially cleared of these charges, although in certain instances some 
inimical relationships were noted. 

LIMITING FACTORS 
EGG FAILURE 

The hatching success of trumpeter swan eggs on the Red Rock 
Lakes nesting marshes has been low compared with that of 
lesser waterfowl. In 1949 for instance, 30 of 61 swan eggs laid 
in 12 nests failed to hatch — a loss of 49 percent. In 1951, the 
loss was 34 percent in the 13 nests checked, when 25 eggs remained 
in the nest from a total of 73 incubated. Examination of 178 eggs in 
36 nests in 1955 revealed that 36 percent failed to hatch for one reason 
or another. Thus, in the limited number of cases studied, hatching 
varied from, 51 to 66 percent. In Denmark the hatching rate of the 
eggs of wild mute swans was about 60 percent as reported by Paludan 
and Fog (1956: 44). In studies of the trumpeter on the Kenai 
Peninsula, Alaska, Spencer (correspondence) reports about a 65- 
percent hatching success in a limited number of clutches. 

For various reasons, the roles played by infertility and mortality of 
the embryo in the trumpeter egg are not known. Often the embryo 
has died after it was well developed. In such cases, incubation 
may have begun before egg laying was completed, or the eggs failed 
to hatch together for other reasons. The trumpeter pen then ap- 
parently departed with the cygnets which hatched first. Other causes 
of egg failure are even more obscure. These undoubtedly include 
cases of infertility or physical handicap, fatal chilling of the embryo 
after incubation has commenced, or abandonment of the nest before 
incubation has been completed. Loss of productivity due to egg fail- 
ure is a major factor in the present low production of cygnets on 
the Refuge. 

Predation is not important to hatching success on the Refuge. In 
the data presented, for a total of 61 nests observed in 1949, 1951, and 
1955, not a single case of predator-caused egg loss was observed. Only 
4 nests containing predator-destroyed eggs have been found in the 7 
seasons during which swan nests have been examined. Furthermore, 
eggs which failed to hatch remained exposed but unmolested in the 
nest long after incubation had been terminated. 

In Yellowstone Park, however, a number of random studies and 
observations made through the years show that egg destruction by 
predators apparently occurs to a significant degree. Joseph Dixon 
(1931:454) writes: " 



LIMITING FACTORS 131 

The pair of swans which Mr. Wright and Mr. Thompson watched at Tern Lake 
on June 11, 1930, were not successful in driving off the marauding ravens, for 
when the mother left to feed, a raven appeared and was observed to fly directly 
to the nest. . . . Mr. Wright recorded in his notebook what took place, as 
follows, "At first the raven just poked about in the nest with its beak. . . . 
It stuck its head down once more and pulled from an egg in the nest a long 
pink and whitish object, apparently an embryo from one of the swan eggs, and 
started to fly away with it just as the parent swans rushed back and drove it 
away from the nest." 

Condon (MS) records an incident of nest destruction by a bear, and 
George M. Wright and Ben H. Thompson (1935: 34, 35) furnish an- 
other account : 

A black object loomed by the swan nest. With field glasses glued to our eyes, 
we saw that it was an otter stretching its full length upward to peer down into 
the nest. From one side it reached out toward the center and pushed aside the 
material covering the eggs. Then the commotion started. With rapt interest, 
the otter rooted around in the dry nest material, heaving up here and digging 
in there, until it was more haystack than nest. Then the otter started to roll, 
around and around, over and over. This went on for a number of minutes. At 
frequent intervals its long neck was craned upward, and the serpentlike head 
rotated around to discover (we supposed) if the swans were returning. At 
last the otter seemed to weary of this play. It climbed from the nest to the 
outer edge, then slid off into the water. ... It never turned back, and was 
finally lost to sight. . . . Seeing that the damage was already done, and another 
year's potential swan crop for the Mirror Plateau lost irrevocably, we saw no 
further reason for caution. So we stripped off our clothes and waded out across 
the shallows. We were amazed to find all five eggs intact. There they were, 
all togethe", rolled to one side, but perfectly whole. . . . We covered the eggs 
and hurried away in confusion as huge hailstones pelted our bodies. We hoped 
that the parents would return to protect the eggs from chill. The storm ob- 
scured the scene, obliterating the next chapter in the story. Later we learned 
from ranger reports that no cygnets were raised on Tern Lake that year. Which 
meddler should shoulder the blame, the otter or the scientist? 

Even in Yellowstone Park, the low hatch is due principally to causes 
other than predation. Observers agreed that human intrusion was 
the most significant known cause of egg failure in the Park. 

PREFLIGHT CYGNET MORTALITY 

Although the relative paucity of predation records on swans of 
flying age indicates that predation is of little consequence in determin- 
ing overall swan population levels, it may be an important cause of 
death to preflight cygnets. Swan broods suffer serious losses in both 
the Red Rock Lakes and Yellowstone Park breeding populations. 
While casualties of preflight youngsters may approach or possibly 
exceed 50 percent during some years, very little is known regarding 
causes of such mortality. 

The information in table 6 is believed representative of the overall 
cygnet mortality on the Refuge during 1019. Since some of these ob- 



132 



LIFE CYCLE 



servations were accomplished at distances over 2 miles with a 20 X 
spotting scope, counts may not be absolutely accurate for any given 
count and brood. A high degree of accuracy was achieved by frequent 
observations over a period of 3 months, and as the cygnets became 
larger and moved in a less compact brood, tabulation was easier and 
more accurately made. The habit of swan families to remain in the 
nesting territory until the approach of flight age, plus the characteris- 
tic of each family to remain apart from other similar groups, makes 
the accurate tally of cygnet mortality by individual brood possible. 

In the broods under consideration, mortality was widespread and 
significant, with losses heavier among the newly hatched cygnets than 
among those approaching flying age. The causes of these losses are 
undoubtedly varied but unknown. Perhaps because the remains of a 
small cygnet resulting from a kill are few and inconspicuous and may 
be entirely consumed by predator or scavenger, few records are avail- 
able regarding specific predation on young cygnets in Red Rock's 
marshes. 

In 1949, 1 saw a large gull kill one cygnet and wound another after 
the young were separated from the parents. The old swans were dis- 
tracted by my presence in the boat, and failed to defend their young. 
Normally, the young cygnets remain very close to their parents, and 
without my presence, the attack would not have been made, or the 
parent swans would probably have repelled it. 

In Yellowstone Park the opportunities for discovering the remains 
of dead cygnets are greater than in the large marshes of the Red Rock 
Lakes. Condon (MS) furnishes a specific account: 

Those cygnets on Swan Lake, however, are thought to have been killed by 
a large male otter, whose presence there was discovered on July 10, 1939, when 
a thorough survey of this lake area was made in an endeavor to locate the lost 
cygnets, and the cause for their disappearance. One cygnet was found dead and 
floating in the lake near the northwest bank within ten yards of the swan nest- 
ing site. This cygnet, upon examination, revealed a crushed sternum, ribs, and 
two tooth punctures. . . . The otter den was located on the east bank of the 
lake and at several sites around the lake otter droppings were collected that 
contained down, bones, leg and foot skin. These fragments were examined in 
the laboratory and compared with the down and leg skin from the cygnet found. 
The comparison revealed that these undoubtedly came from a swan cygnet, and 
it is felt that in this instance the otter was responsible for the loss of part, if not 
all of the family of five cygnets from Swan Lake. 

Although otters have only rarely been reported from the Red Rock 
Lakes Refuge, minks have at times been suspected of cygnet preda- 
tion. Sharp relates in a Refuge report : 

A family of mink worked at the Idlewild boat pier during late July. I counted 
36 coots and 31 ducks and 1 young muskrat that were dragged upon the pier 
along the edges of tall overhanging sedges. A family of three cygnets that used 
this area disappeared during this time. 



LIMITING FACTORS 



133 



Table 6.— Cygnet mortality at Red Rock Lakes Refuge, 1949 



Nest 


Eggs 
hatched 






Number of cygnets observed per 


brood 




June 


July 


August 


21 


30 


6 


14 


21 


28 


6 


15 


19 


30 


No 1 


3 
3 

5 
3 
4 
4 
4 
1 
3 
1 


3 

3 

2 
4 
1 







3 
3 

5 
2 
4 

1 
4 

2 



2 
3 

3 

4 


4 

2 



2 
3 
3 
3 
4 

4 





2 
3 
3 
3 
2 

4 

2 



2 
3 
3 
3 
1 
1 
4 

2 




3 
3 

3 

1 
3 


2 



2 
3 
3 

1 
1 
2 

3 



2 
3 
3 
3 
1 
1 
2 





2 


No. 2 


3 


No 3 


2 


No. 4 


3 


No 5 --- 


1 


No. 6 


1 


No. 7 


2 


No. 8 





No. 9 


3 


No 10 





Total 




31 




















17 


Cygnet loss since previous 


5 
3 






1 
1 


2 
1 


2 
1 


1 
1 


1 
1 


1 
1 










1 






1 









With the exception of minks and skunks, no mammalian predators 
of any consequence are to be found in the dense, sedge-covered, soft 
bog-marshes of the Red Rock Lakes. Though dead cygnets are occa- 
sionally discovered on the water areas, they are invariably intact, with 
no evidence of predation. The densely vegetated shores preclude a 
systematic search. 

Although gulls, falcons, eagles, and ravens are occasionally sighted 
over the marsh during the summer, they have never been seen molest- 
ing the swans, either young or old. Presumably, the usual close 
swimming and feeding formation of the swan family presents little 
opportunity to the predator, with the size and potential defense 
capabilities of the adults an obvious deterrent. Although great 
horned owls have been mentioned as possible predators, no record 
exists. 

MORTALITY OF IMMATURES AND ADULTS 

Except for man, trumpeters in the wild appear to have few impor- 
tant natural enemies after flying age is reached. Both golden eagles 
and coyotes may take swans of any age class under certain local con- 
ditions which present favorable opportunities. Even in such cases 
it is doubtful whether the trumpeter is a normal prey of either 
eagles or coyotes with any degree of regularity. Some of the 
captive trumpeters at Malheur and Ruby Lake Refuges have been 
lost to bobcats on several occasions. The following instances outline 
the circumstances under which losses have occurred among unconnned 
trumpeters. 

The natural enemies and causes of mortality, especially parasites 
and diseases, of other swans are discussed in detail in Hilprecht 
(1956:101-107). 



134 LIFE CYCLE 



' 



Avian predation : Probably the only winged predator capable and 
willing to tackle a full-grown swan in flight is the eagle. Apparently 
even these great raptors will attempt this only occasionally. 

On several occasions, eagles have been observed to knock trumpeters 
out of the air and kill them. Sharp (1951 : 225) witnessed the follow- 
ing cases of eagle predation on cygnets of flight age concentrated on 
their Red Rock Lakes winter feeding waters, writing : 

On one occasion in late November of 1944 the writer observed one of the eagles 
[golden] make three stoops over a flock of swans on a snow covered meadow. 
The swans stood motionless and apparently had no fear. This eagle was ap- 
parently sporting over the swans as no kills had been made up to this time. . . . 
A cygnet was killed at Culver Pond on December 26, 1944, and another was struck 
in the air and killed on January 1, 1945. ... A total of three cygnets was 
killed during this winter. 

A local rancher in that area, James F. Hanson, reports occasional 
harassment of the trumpeters by eagles, but only rarely a killing. 

Henry W. Baker, Jr., Superintendent of the Federal fish hatchery 
near Ennis, Montana, provides an eyewitness account of eagles kill- 
ing a swan in the Madison Valley (correspondence) : 

When first observed, the encounter was approximately 1,000 feet elevation. 
The two eagles worked together, first one would hit the swan and then the other 
until they brought it to the ground, at which time they both attacked it . . . 
killing it in the matter of a couple minutes. I was approximately 600 to 700 
yards distant at the time. 

Ralph Edwards (Holman, 1933: 169-211), the early settler of 
Lonesome Lake, British Columbia, relates, "I have noticed during the 
hard winters that the grey ones fall prey to the eagles and starvation 
before the white ones. This winter I have seen two cases of eagles 
killing swan. The eagles do not seem to be able to catch the swan 
when they are stronger or on a straightaway flight," 

Though eagles can and apparently do kill adult swans on occasion, 
the effect on overall trumpeter numbers is believed insignificant, To 
my knowledge, no other winged predator has ever been observed to 
prey upon adult swans. 

Mammalian predation: Swans may occasionally be molested by 
coyotes, but direct evidence that they have actually killed swans still 
is lacking. Even in Yellowstone Park, where coyote populations exist 
at natural levels, reports of actual kills are nonexistent. Condon 
(1950: 1, 2) reports some ineffective coyote molestation: 

During the fall of 1947 after Swan Lake had frozen over it was common to 
see a lone pair of trumpeter swans resting on the ice. . . . The tracks in the 
snow showed that their siesta was interrupted by a coyote which . . . appar- 
ently was either persistent, hungry, or just playing a game, for it had each time 
approached cautiously to some point where it was screened by a cinque-foil 



Limiting factors 



135 



bush or clump of grass at the ice's edge and then had made a dash toward the 
resting swan. Bach time failure was its lot. 

Dr. Aclolph Murie (1940: 135), after a thorough study of coyote 
prey relations in Yellowstone Park, wrote : 

At some of the lakes where swans have been raised, coyotes are concentrated. 
At Trumpeter Lake, where seven cygnets were raised in 1936 and again in 1937, 
coyotes and coyote signs were frequently noted at the lake. ... The only evi- 
dence of waterfowl predation consisted of some remnants of a green-winged teal 
found on the bank and in one dropping. 

Dr. Murie concluded his field study and a review of available infor- 
mation on the subject within the Park by writing, "It was rather un- 
expected to find that the coyote in Yellowstone exerts no appreciable 
pressure on the trumpeter swan population. . . . The data avail- 
able at the present time indicates that the coyote does not represent an 
important mortality factor for the trumpeter swan." 

At the Red Kock Lakes Refuge, too, no definite coyote kills of 
swans have been recorded, though Hull (1939: 381) states, "During 
the winter coyotes have been known to capture adult swans in deep 
loose snow before the birds were able to get into the air or open water." 
Sharp also reported, "Ranchers hauling hay off the refuge near 
Upper Red Rock Lake reported seeing coyotes on two occasions 
flush swans off the ice or snow. In both cases, the swans were too alert 
and agile in the take-off to allow the coyote to come close to a catch." 
From the foregoing evidence it is concluded that the coyote is not 
an important predator of trumpeters, either in Yellowstone Park or 
on the Red Rock Lakes Refuge. 

Condon (MS) also relates a case where circumstantial evidence 
pointed to a bear as the suspected killer of an adult swan in Yellow- 
stone Park, writing: 

a dead male was found floating near the northwest shore of Fern Lake. ... A 
thorough examination of this bird revealed a transverse cut straight across the 
breast region 6 inches long and 1% inches deep. . . . After inspecting the area 
around Fern Lake where this swan was found, I concluded that a bear was, m 
all probability, responsible for its death. There were abundant fresh signs of 
bear around the lake and the extensive damage done the swan seemed greater 
than that which any other animal might inflict. 

From the foregoing testimony it would seem that 4-footed pred- 
ators are even less effective than their avian counterparts in preying 
upon swans of flight age. 

Hunting. From the earliest times on this continent swans have 
been taken by the native peoples by whatever method was available, 
snare, arrow, or gun. Though the primitive peoples presumably 
killed principally for food, the white man also killed the swan in 
great numbers over wide regions of its range not only for domestic 

469660 O — 60 10 



136 LIFE CYCLE 

needs but also because of the ready commercial market established 
for its down and quills. Man is, by far, the greatest enemy with 
which they have had to contend. 

Although the enactment of the Migratory Bird Treaty Act by the 
United States Congress in 1918 made it illegal to kill either species of 
our native swans, shooting and lead-poisoning still account for more 
casualties, in the case of the trumpeter at least, than the aggregate of 
all other known causes of adult mortality. 

During the last 20 years, illegal kills of trumpeter swans have oc- 
curred chiefly during the open waterfowl season in Idaho along 
Henrys Fork of the Snake River and its tributaries, and to a lesser 
extent in Montana and Wyoming. Since trumpeters commonly fly 
along the water courses at a low level, they furnish a conspicuous tar- 
get well within range of waterfowl hunters, many of which cannot 
resist the temptation to shoot, The kill is believed to have increased 
in the past decade with the appearance of greater numbers of nimrods 
afield. 

Fifteen years ago, Condon recognized this danger to the swan pop- 
ulation in the Yellowstone region, stating (MS) : 

That there is a definite mortality among the swan due to shooting by hunters 
during the open season on waterfowl is evidenced not only by the above in- 
cident [that of finding a swan containing shot in the Park] but by the arrest 
and conviction of hunters in Idaho and Montana for killing trumpeter swan. 
Many instances of this type have come to light and undoubtedly many are shot 
that no one knows about. . . . Adults known to have lost their lives at the 
hand of man far exceed the records that we have of death due to natural causes. 

In an effort to determine the gunning pressure to which the trum- 
peters are subjected while on their comparatively unprotected winter 
ranges, the U. S. Fish and Wildlife Service arranged with the Illinois 
Natural History Survey to make a fluoroscopic examination of a num- 
ber of swans during their flightless molt on the Red Rock Lakes 
Refuge. Frank Bellrose of that survey supervised this investigation 
in the summer of 1956, using a portable fluoroscope borroA^ed from 
the Delta Waterfowl Research Station, Delta, Manitoba. 

During the period July 20-28, 100 trumpeters from the Refuge's 
nonbreeding population were examined fluoroscopically by Bellrose 
to determine the presence of lead shot. In addition to this sample, 
I subsequently examined 3 birds in August, 

Of the 10-3 trumpeters thus inspected, a total of 15 (14.6 percent) 
were found to be carrying lead shot. Numbers of shot present in 
any individual bird varied from 1 to 9 and included shot sizes ranging 
from 7y 2 's to BB's with 2's and 4's most common. Of 4 swans banded 
in 1951 or earlier, 2 were carrying lead. Owing to the large size of 
these birds and the difficulty of locating the shot on the dimly il- 



LIMITING FACTORS 



137 



laminated fluoroscopic screen, some may have been missed. Thus, 
it is probable that the actual percentage of swans carrying lead shot 
was somewhat greater. . 

The actual number of trumpeters killed each year in their United 
States range is unknown, but it probably exceeds 25. The number 
of known swan casualties was 17 in 1933 (Beard et al., 1947 : 140) and 
19 in 1939 (Condon, MS) . The average number of swans known to 
have been lost to hunting during later years continues to be serious. 
The known loss was reported to be 23 in 1951, 17 in 1952, and only 9 
in 1953. Iinl955, Eddie Linck, the Idaho State Fish and Game 
Officer covering the Island Park area, estimated a total of 25 trum- 
peters illegally killed in that area alone, with 12 casualties known 
for certain. According to Walton Hester, conservation officer for the 
Island Park district in 1956, a known loss of 15 hunter-killed swans 
occurred with at least another 5 estimated as being taken. Mr. 
Hester also expressed the thought that some of the cygnets may be 
selectively shot for food, since the adults are noted for their general 
unsavoriness. Evidence supporting this view is lacking. 

The concern pi those interested in the welfare of the trumpeter 
resulted in the*apr!ointment of a new U. S. Fish and Wildlife Service 
enforcement agent in eastern Idaho in 1956, principally to cope with 
this illegal killing of trumpeters while on their main wintering 
grounds in the Island Park area. Increased local publicity in the 
past few years via radio, TV, newspapers and posters has focused 
attention on the problem. It is hoped that the increased attention 
will serve to stop the indiscriminate killing of these birds. 

Lead poisoning. This factor is irrevocably related to hunting and 
causes mortality among swans that feed, even for short or intermit- 
tent periods, in habitat shot over in previous years. Hull recovered 
4 trumpeters which died on their Kefuge feeding grounds at Culver 
Pond during late March and early April, 1937. Material from Jhese 
birds was sent to the Denver Research Laboratory where E. R. Kalm- 
bach diagnosed the fatalities of 3 as due to lead poisoning. The 
gizzard pads of all 3 exhibited the characteristic greenish coloration 
and hardening, with the contents containing 3, 11, and 17 pellets 
respectively (0.248, 0.498, and 0.857 grams of lead). In the fourth 
case, Hull made a similar diagnosis, with 19 pellets in the gizzard of 
the dead bird he examined. No other trumpeter cases have been re- 
ported in the United States. No hunting has been permitted on the 
wintering waters of the Refuge since it was established in 1935; 
hence, the possibility of further losses is lessened. 

In British Columbia, the problem of lead poisoning among winter- 
ing flocks of trumpeters has presented a recurring threat. J. A. 
Munro (1949) documented the loss of at least 9 trumpeters of the 



138 



LIFE CYCLE 



Vaseaux Lake wintering population in 1925. These birds were forced 
out of their regular winter quarters by ice and used other open waters 
which were heavily contaminated with lead shot. The stomach con- 
tents of 1 of these victims contained 451 shot, More of this group 
were believed to be subsequent victims of ingesting lead shot on this 
occasion, as only 6 of the original group of 17 returned to Vaseaux 
Lake the following winter and the flock later disappeared completely. 
Munro also records the loss of at least 13 trumpeters from a flock 
of 15 wintering on Vancouver Island in 1946. The stomach tracts 
of these victims held from 2 to 29 pellets each. 

On Idaho's lower Coeur d'Alene Eiver, a form of metallic poison- 
ing diagnostically similar to lead poisoning has resulted in irregular 
but substantial mortality in the waterfowl over the years. Whistling 
swans, stopping in migration, have occasionally suffered heavy losses 
in this area, but as contamination is local these losses have never 
involved trumpeters. 

Weather and mortality. Unseasonally cold weather undoubtedly 
affects hatching success and survival of trumpeter cygnets. These 
effects are not completely Jmown, but some influences have been 
observed and are cited here. 

Cygnet Development vs. Weather. Uneven development rates 
of cygnets are not uncommon and, when combined with late hatching, 
result in some loss when the ice forms earlier than usual. The 
average hatching date on the Red Rock Lakes Refuge is about June 
20. Over 100 days is normally required for a cygnet's development 
to flight age, and the Refuge has become icebound as early as October 
28 (twice) with November 8 the 16-year average (1938-1953). 

Final freezeup is invariably preceded by a period when ice is 
prevalent over the marsh and lakes, so development may be retarded 
m the critical preflight stage of the growing cygnet through food 
shortage and injury to the wing from hitting the ice during 
flight attempts. Although it is possible occasionally to rescue such 
birds, as well as unharmed individuals which have simply not ma- 
tured sufficiently to fly, it can hardly become a regular practice, since 
travel conditions on the new ice over the vast river and marsh area 
are often treacherous by foot and impossible by boat. 

In 1949 at least 20 cygnets on the Refuge were still incapable of 
flight by October 15. Fortunately, the final freezeup did not occur 
until November 18 that year. This additional time may have allowed 
all of these particular cygnets to develop sufficiently to fly. 

Condon raises a point about some other hazards, found in Yellow- 
stone Park : 

One factor that may contribute to cygnet mortality after they reach flying 
age is that many of the lakes on which the swans establish themselves are small 



LIMITING FACTORS 139 

and may not offer sufficient space for them to learn to fly well, and when the 
families leave these lakes they must fly for some distance over land hefore 
reaching water upon which to alight and on such occasions cygnets may become 
fatigued, fall to earth and may not again get into the air. It is felt that 
such was the fate of one of the cygnets from Geode Lake in early October 
1939 and that such mishaps may occur more often than we think where lakes 
are small and separated by rough canyons and timbered areas such as exist 
on the north side of Yellowstone Park. 

Mortality of Grown Birds. The formation of ice over normally 
open wintering waters, as a result of very low temperatures, may pre- 
vent adult swans from feeding in these traditional locations. When 
this ice remains for a prolonged period, starvation may result rather 
than movement out of the region to where food may be available. 
Such instances seem to be rare within the United States however, 
judging from the lack of evidence or reports. Dr. Olaus J. Murie 
documents such an occurrence (correspondence) : 

A number of years ago (March 22, 1932) I found two dead swans north of 
Moran in Jackson Hole, where they had been wintering in a small piece of open 
water on a stream (Second Creek) near Jackson Lake. They had evidently 
starved to death since they had eaten all the available food in that little piece 
of water. . . . The male weighed 18 lbs. 4 oz. The female 16 lbs. 12 oz. 

As a general rule in the major Island Park-Red Rock Lakes- 
Yellowstone wintering habitat, at least some open water and food are 
available to swans wintering in these warm spring-fed water areas. 
This is true even during prolonged periods of winter weather where 
nightly minimums drop below -20°F. Though food supplies are un- 
doubtedly short until moderating weather arrives, they never fail 
entirely, and I know of no other cases of trumpeter starvation in this 
country. 

In British Columbia, starvation has caused greater losses of 
wintering trumpeters. Munro (1949:713) records: 

The number of trumpeter swans dying from starvation in winters of more 
than average severity probably is a significant factor in population reduction. 
Many swans winter in sub-marginal territory where food is not abundant at 
the best of times and may become completely inaccessible at times of sub-zero 
weather, or at periods of high floods. In such times the swans stay on, in what 
probably is their ancestral wintering ground, to starve or to become, in their 
weakened state, victims of predators, rather than seek feeding grounds else- 
where. The situation is being met to some extent by artificial feeding, as has 
been noted, but winter populations in remote and inaccessible territory cannot 
be helped in this way. 

Swan mortality resulting from the icing up of swans' plumage by 
severe winter weather has also been reported from Canada, though 
never to my knowledge from the United States. The notes of O. J. 
Murie record this instance related to him by Dave Hoy, a river 
freighter familiar with wintering trumpeters: 



140 LIFE CYCLE 

Fort St. James, B. C, June 13, 1934, Mr. Hoy says that a number of years 
ago the trumpeter swans numbered 250 or more but they winter killed. He 
says the water does not freeze over for long but the short period of freezing 
may be fatal. About three years ago a number got frozen into the ice. They 
were found in that condition on a number of occasions, eaten by coyotes. Slush 
ice would gather on their wings so they could not fly. 

The icing of swans' plumage has never been reported from the Island 
Park area where slush ice is seasonally common on the streams fre- 
quented by wintering trumpeters. The water-shedding capacity of a 
swan's plumage is definitely related to the overall health of the indi- 
vidual, so it is difficult to see how the plumage of a healthy adult swan 
could collect ice, or how such an individual would allow itself to 
become frozen into the ice no matter how severe the weather might 
become. Low food supplies may play the determining role in such 
cases. 

Disease. Trumpeter swans within the United States exhibit the 
alertness and vigor associated with most wildlife populations. Dead, 
sick, or weak birds are rarely observed without reason for their 
condition being obvious, and disease does not appear to contribute 
significantly to mortality in the wild. 

While disease has apparently never been reported in adult wild 
trumpeters, small cygnets occasionally possess deformities of their 
feet. Sharp stated in a Service report. : 

Three nests did not hatch and a fourth failed due to feet deformities of the 
cygnets. The latter nest, when checked on the Upper Lake, had three dead 
cygnets and another alive on the nest. Careful study revealed that the living 
cygnet could not stand. An examination of its feet showed that they were 
pitifully deformed. Then the dead cygnets were examined, and their feet were 
deformed in a similar manner. 

A cygnet which I rescued from a gull's attacks possessed a deformed 
foot as described by Sharp. On the other hand, the broodmate of 
this cygnet, which was killed by the gull, had normal-appearing feet. 
This abnormal condition may be a cause of some mortality in 
preflight age classes, since it has not been noted with the older 
trumpeters. 

The incidence of several diseases has been reported from trumpeters 
in captivity. Confinement often increases the opportunities for sick- 
ness and infection, and facilitates the observations of afflicted birds. 
One of the adult trumpeters at the Delta Station died during the win- 
ter of 1955-56 after displaying signs of weakness and sickness, one of 
which was trouble in keeping its plumage dry and free of ice. A 
thorough postmortem examination was not possible at that time, 
though symptoms of both aspergillosis and fowl cholera were appar- 
ently present. 

Dr. E. M. Dickinson of the Veterinary Department of Oregon State 
College found and cultured bacilli typical of avian tuberculosis from 



LIMITING FACTORS 



141 



a captive trumpeter at Malheur Refuge which exhibited the lesions 
of this disease, dying there apparently of this cause. Several other 
losses in the Malheur trumpeter flock had these symptoms, although no 
successful testing for this disease was developed. Three casualties 
listed for the Ruby Lakes trumpeter flock were attributed to avian 
tuberculosis. The greatest incidence of loss in these cases seemed to 
occur in the younger age groups, especially those less than a year old, 
with a reduced rate of loss as the birds grew older. A more recent 
trumpeter loss at Malheur in 1956 was diagnosed by Dr. Dickinson as 
having aspergillosis. 

Of the captivity record of the trumpeters which have been kept at 
the New York Zoological Park, William G. Conway (correspondence) 
advised, "Four specimens are said to have died of aspergillosis; . . . 
two, probably, from botulism." Since the natural conditions neces- 
sary for botulism contamination do not normally occur in the Red 
Rock Lakes- Yellowstone region, this sickness has never been reported 
in the native trumpeter flock, but in late October 1957 two cygnets in 
Malheur's captive flock apparently died from botulism. In one case 
the diagnosis was fairly certain, while in the other the evidence was 
only circumstantial (David Marshall, correspondence). 

Dr. T. T. Chaddock (1938: 25) examined 8 wild whistling swans 
postmortem, from April-June carcsfes recoveries in Wisconsin, report- 
ing the presence of aspergillosis (in 4), pneumonia (in most) , silicosis 
(in 2), "dropsy" (in 2), besides parasites (in 5) plus other pathological 

factors. 

Parasites. Trumpeter swans, in common with other wild species, 
are frequent hosts to various parasites. Judging from the occurrence 
of parasites in other fauna, the level of any given parasite population 
is usually determined more by the general health of the host than by 
any other single circumstance. Healthy birds normally support low 
numbers of parasites, while sick or weak birds frequently exhibit 
heavy infestations. In the latter instances, while parasites may com- 
bine with other factors to hasten death, they are usually considered 
only a contributing cause of mortality. Thus, there seems sufficient 
justification to document several cases in which parasites were either 
believed to have caused mortality or were present in such numbers 
that they probably contributed substantially to the reported sickness 

or death. 

Dr. Ian McT. Cowan (1946: 248, 249) stated that the death of a 
wintering British Columbia trumpeter found in a weakened condi- 
tion near Vanderhoof, B. C, was due to gross multiple parasitism. 
He relates, "Both the cestode Hymenolepis and the filarial nematode 
Eurycerca were present in numbers apparently sufficient to induce 
pathological changes in the host, and it is not possible to determine in 
this case which of the two was most harmful." 



142 LIFE CYCLE 

In another instance Cowan found the white swan louse Ornithobius 
cygni abundant in the plumage, and 3 specimens of Eurycerca in the 
heart muscle of a trumpeter which apparently died of lead poisoning, 
as 30 pellets of lead shot were found in the gizzard of this swan. 

Parasites were noted in 5 of the 8 whistling swan carcasses exam- 
ined in Wisconsin by Chaddock (1938 : 25-27). Five cases exhibited 
pediculosis (lice infestation) around the fluff of the vent, the pres- 
ence of gizzard worm (Spirotera hmmdosa) was demonstrated, and 
the eggs of common poultry roundworm (Ascaridia liniata) were 
found when a fecal examination of the 5 internally parasitized indi- 
viduals was made. 

Leeches are common to abundant during the summer months in 
most of the waters inhabited by the trumpeters in their United States 
range. As one might expect with waterfowl accustomed to seeking 
their food in the soft mud and vegetation of such areas, leeches fre- 
quently become attached to their bodies. While these blood suckers 
apparently possess little more than nuisance value on the larger swans, 
they may be a contributing cause of mortality with small cygnets. 
Sharp reports that on one occasion : 

Cygnets taken from the north side of the Lower Lake revealed that their under 
parts had from a dozen to fifty small leeches crawling over their wet feathers. 
Swanlets on two occasions taken from this area when placed in the boat within 
a few minutes shook their heads violently and finally threw out a leech ; the 
latter were gorged with blood. These leeches were from one-half to three- 
quarters of an inch in length, and when crushed were filled with blood, as much 
as could be held without apparently bursting. One cygnet on the Upper Lake 
also disgorged a large leech. 

One of 3 cygnets transferred from the Red Rock Lakes Refuge to 
the Delta Research Station in 1955 died after only 3 days. The post- 
mortem examination was conducted by the Ontario Veterinary College 
at Guelph, Ontario, and revealed that the cygnet had died from an 
infestation of thorny headed worms (Polymorphus baschadis). These 
worms require crustaceans as an intermediate host and are fairly 
pathogenic, causing anemia and cachenia (Peter Ward, correspond- 
ence). The cygnet probably was infested at Red Rock Lakes, and 
this may be a cause of some mortality there. 

Trematodes are occasionally observed in the cloaca of Red Rock 
Lakes trumpeters when they are being examined to determine the sex. 
Two of these parasites were identified as Echinostonia revohttum. 
Feather lice (Mallophaga) are also commonly observed. Determina- 
tions of the specific identity of these parasites was made through the 
office of Dr. William Jellison, Parasitologist, National Microbiological 
Institute, at Hamilton, Montana. 

In 1950 and 1951 blood smears from various Refuge trumpeters 
were prepared and forwarded to Dr. Carlton M. Herman, Wildlife 



LONGEVITY 143 

Pathologist, Patuxent Research Refuge, for a routine examination for 
leucocytozon, but no positive report was received. 

Hilprecht (1956: 103-104) reports that swans in Europe become 
infested with parasites, the tape worm Hymenolepis aequabilis and 
the leech Protoclepsis granata. The latter becomes especially 
abundant during summers of low water levels when it is thought to 
contribute to mortality on a large scale. 

Accidents. As one might expect from such large, specialized 
fowl, trumpeters appear to be "accident prone" under certain circum- 
stances. In confinement at the Malheur Refuge, death of several swans 
resulted from drowning when their heads and necks became caught in 
an underwater section of fence while trying to obtain grain beyond 
easy reach. In the New York Zoological Park a trumpeter was 
reported to have died of injuries sustained when it was caught in a 
gate. 

In the wild state, the known accidents seem to be confined largely to 
striking power, telephone, or fence wires in flight. At least 3 cases of 
known trumpeter fatalities have been reported for the Island Park 
area from such causes, as winter fog is common along open water- 
courses. Besides these, 2 other similar cases, 1 on the Refuge and 1 in 
southeastern Montana, have occurred. In 4 of the 5 known cases the 
accident was fatal. 

LONGEVITY 

From earliest times swans were known to be very long lived. The 
ancient naturalists of Greece bore testimony to this (Evans 1903 : 121) 
as well as early English ornithologists who presumably were writing 
of the mute swan in captivity (Swann 1913: 164). Some age data 
of other species of captive and wild swans are furnished by Hilprecht 
(1956:100-101). 

The Refuge swan banding project is comparatively recent and may 
eventually contribute valuable data on the longevity of wild trumpet- 
ers. Despite the considerable success of F. E. Blaauw and others in 
breeding the trumpeter over a long period in Europe, specific records 
of longevity appear to be lacking. The Philadelphia Zoological 
Garden had eight trumpeters on display from 1895-1939, one of which 
lived for 29 years. F. H. Kortwright (1943: 80) noted that a 
trumpeter lived for 32i/ 2 years, presumably in Canada. This seems 
to be the present record for this species in captivity. 





POPULATION 

Little information was uncovered regarding the status of the swan 
population which existed in the Rocky Mountains of Montana, Wy- 
oming, and Idaho, before the specific swan surveys were initiated by 
the National Park Service in 1929. The few breeding pairs and 
immatures which then inhabited this vast region were even more 
thinly dispersed over their widely scattered mountain lake environ- 
ment than they are today. (See "Distribution and Status" for a few 
clues to precensus status of swans in Yellowstone Park and Red Rock 
Lakes Refuge.) The important general conclusion to be drawn from 
the scattered data available seems to be that this species was not 
abundant in this region since the turn of the century, and by the 
1920's must have nearly disappeared from its last breeding grounds 
in this country. 

ANNUAL SWAN CENSUS, 1929-57 

The life history studies in the Park and Refuge have progressed 
only sporadically in the three decades since the National Park Serv- 
ice's trumpeter swan survey began in the fall of 1929. Fortunately, 
the annual census which was initiated in 1931 has been effectively 
conducted during most of this period. When properly qualified, this 
accumulation of data offers a remarkably complete 27-year record of 
population levels. The large size, prominent white color, sedentary 
habits, and use of only a comparatively few water and open-marsh 
habitats combine to make these birds an ideal species to locate and 
census accurately. Too, with few exceptions, census methods and 
coverage were generally expanded in time to keep abreast of the grow- 
ing swan population. For these reasons I believe the census data 
which follow are exceptionally reliable and representative. 
144 



POPULATION DYNAMICS 



145 



Although it is assumed that little or no interchange has occurred 
between the trumpeters living in this tristate region with other popu- 
lations in Canada and Alaska, this possibility has never been thor- 
oughly explored. But in view of the attraction the waterfowl show 
for their natal marshes generally, and the absence of banding data 
to suggest the possibility of populations mixing, it is concluded that 
the factor of interchange has been insignificant thus far. 

The results of the 27 years of swan-census effort, first solely by the 
National Park Service and then with the cooperation of the Fish and 
Wildlife Service, are presented in detail in table 7 and in appendix 5. 
The distribution of the swans during their 1954 peak year census, 
and the pertinent references, are given in appendix 2. 

Not all of the census data in table 7 will be found to agree exactly 
with figures released for news publication by the Fish and Wildlife 
Service over the years. A few discrepancies were subsequently dis- 
covered in the field and corrected but were not released as a news 
followup owing to the small difference in numbers involved. In each 
case, the data in table 7 are the most comprehensive and representative 
which could be found. 

POPULATION DYNAMICS 

The rise in trumpeter numbers in the United States which began a 
quarter century ago is closely correlated with the establishment of the 
Red Rock Lakes Refuge and the increased protection and attention 
which then became possible. The relatively few departures of the 
annual census numbers from uniform rates of change can be explained 
by incomplete censuses, departures of sizable segments of the popula- 
tion, and actual variations in productivity. 1 ' Information obtained 
from the annual trumpeter censuses is presented in tables 7 to 12 and 
shown graphically in figures 44 to 50. 

As related earlier, ground counts conducted from 1931 through 
1945 were made under great handicaps of personnel and equipment 
compared with the coordinated aerial censuses which were developed 
later. Since a complete coverage of areas occupied by trumpeters 
was not accomplished during the earliest years, 1931-35, it is likely 
that the actual populations were somewhat larger than were located, 
but the numbers recorded then are believed to be highly representative. 
Since 1936, when more complete ground coverages of the trumpeter- 
occupied areas were begun, only two complete censuses, in 1950 and 
1957, produced results which did not compare well with counts made 
during the preceding and following years. The low count of im- 



1 For purposes of this discussion, "productivity" is measured by the number of cygnets 
censused annually. 



146 



POPULATION 
Table 7.— Swan census data, 1931 to 1957 





Red Rock Lakes Refuge 


Yellowstone Park 


All other areas 


Total, all areas 


Year 




























Adults 


Cyg- 


To- 


Adults 


Cyg- 


To- 


Adults 


Cyg- 


To- 


Adults 


Cyg- 


To- 






nets 


tal 




nets 


tal 




nets 


tal 




nets 


tal 


1931 ... 


(') 


(') 


(') 


18 


12 


30 


2 


3 


5 


20 


15 


35 


1932... 


19 


7 


26 


29 


2 


31 


9 


3 


12 


57 


12 


69 


1933... 


15 


9 


24 


27 


8 


35 


7 





7 


49 


17 


66 


1934... 


16 


26 


42 


16 


17 


33 


16 


6 


22 


48 


49 


97 


1935-.- 


30 


16 


46 


16 


11 


27 


(•) 


(') 


(') 


46 


27 


73 


1936... 


31 


26 


57 


38 


13 


51 


7 


2 


9 


76 


41 


117 


1937... 


34 


51 


85 


38 


26 


64 


9 





9 


81 


77 


158 


1938- - 


28 


42 


70 


40 


4 


44 


25 


9 


34 


93 


55 


148 


1939... 


50 


59 


109 


47 


17 


64 


26 





26 


123 


76 


199 


1940... 


58 


48 


106 


39 


14 


253 


26 


6 


32 


123 


68 


191 


1941... 


52 


44 


96 


44 


15 


59 


47 


10 


57 


143 


69 


212 


1942. .. 


45 


43 


88 


(') 


(') 


(') 


53 


10 


63 


98 


53 


151 


1943... 


88 


25 


113 


(') 


(') 


(') 


49 


9 


58 


137 


34 


171 


1944... 


106 


58 


164 


41 


8 


49 


60 


6 


66 


207 


72 


279 


1945... 


113 


50 


163 


(') 


(') 


(') 


67 


5 


72 


180 


55 


235 


1946... 


124 


46 


170 


43 


8 


51 


122 


18 


140 


289 


72 


361 


1947... 


131 


49 


180 


45 


8 


53 


116 


3 


119 


292 


60 


352 


1948..- 


121 


73 


194 


49 


13 


62 


142 


20 


162 


312 


106 


418 


1949. _ , 


132 


61 


193 


54 


21 


75 


162 


21 


183 


348 


103 


451 


1950-.- 


106 


40 


146 


57 


16 


73 


140 


17 


157 


303 


73 


376 


1951-.. 


170 


76 


246 


63 


11 


74 


184 


31 


215 


417 


118 


535 


1952-.. 


184 


55 


239 


58 


10 


68 


236 


28 


264 


478 


93 


571 


1953..- 


211 


38 


249 


51 


10 


61 


216 


51 


267 


478 


99 


577 


1954... 


352 


28 


380 


64 


23 


87 


144 


31 


175 


560 


82 


642 


1955... 


242 


41 


283 


58 


10 


68 


195 


44 


239 


495 


95 


590 


1956... 


293 


39 


332 


48 


9 


57 


166 


33 


199 


507 


81 


588 


1957... 


159 


45 


204 


44 


16 


60 


196 


28 


224 


399 


89 


488 



1 No census. 

2 Incomplete census. 

matures and adults in 1950 was followed by an increase in these 
combined age-classes in 1951 which was much greater than could be 
accounted for by cygnet production. It remains to be seen whether 
the same lack of agreement will occur between the 1957 census and 
that of 1958. 2 These census disparities apparently resulted from the 
movement of a substantial segment of the population out of the region 
censused (about 120 in 1950 and 100 in 1957) . 

The annual aerial census covers all possible trumpeter habitat in 
portions of three States, but not the many remote areas outside this 
region to which they conceivably might fly. Specific knowledge of 
any such outward migration is lacking at this time. Furthermore, 
no clue to their time of departure, return, or destination is now at 
hand. Their absence most likely began as a northward movement 
with the spring migration. After summering at a new location, re- 
joining the resident flocks on the traditional wintering grounds dur- 
ing the fall migration would be expected. But this is merely 
speculation. 

Since banded trumpeters from northern flocks have never been 
found among the many handled in the United States during the 
summer molt, and United States-banded swans have not been recov- 
ered outside their expected range in this country, I assume that the 

2 A total of 703 trumpeters, 565 immatures and adults and 138 cygnets, was tallied in 
1958. 



POPULATION DYNAMICS 



147 



Table 8.— Nonbreeding trumpeter swan populations at Upper Red Rock Lake, 
Lima Reservoir, and other important areas, 1939 to 1957 

[Nonpaired (flocked) immatures and adults (cygnets excepted)] 



Year 



1939. 

1940 

1941. 

1942. 

1943. 

1944. 

1945. 

1946- 

1947. 

1948. 

1949. 

1950. 

1951. 

1952. 

1953 

1954. 

1955. 

1956. 

1957. 



Upper 
Red 
Rock 
Lake 



11 

19 
18 
12 
42 
52 
66 
62 
51 
49 
50 
63 
99 
120 
135 
248 
150 
162 
77 



Lima 
Reser- 
voir ' 





ID 

9 

10 

16 

15 

27 

26 

37 

74 

69 

83 

136 

112 

7 

11 

7 

35 



Island Park 
Reservoir, 
Sheridan 
Reservoir, 
and other 
waters 



Malheur 

and Ruby 

Lake 

Refuges 



Total 



31 
22 
49 
43 
64 
79 
107 
159 
136 
164 
177 
167 
239 
310 
266 
294 
249 
233 
181 



■ Lima Reservoir was drv during 1954-1956 inclusive. Most of the resident summer population on these 
waterfapparenUy moved to Upper Red Rock Lake during this period as the latter area shows an abnor- 
mally large population during the summers of 1954-1956. 

2 No count was made in 1939. 

missing birds moved out of the region before the annual census and 
then returned in time to be included in the following year's count. 

Having mentioned the major departures from the otherwise fairly 
uniform population curves for the period 1931-57, let us look at the 
overall trend in the numbers of this species during this period. The 
main references for this discussion are tables 7 and 9. Disregarding 
the years 1935, 1942, 1943, and 1945 when complete counts were not 
secured in the Park, the two categories for which comparable informa- 
tion is available in table 7 are (1) total swan numbers and (2) cygnets. 
The data in table 9 are graphed in figure 44 to show the rates of 
change for certain segments of the population. 

In figure 44, the total population curve shows a steady and rapid 
climb (about 10 percent annually) until 1954, when 642 swans were 
counted. This peak was followed by two slightly lower counts, in 
1955 and 1956. The sharp decline in 1957 was explained earlier. The 
numbers of immatures and adults increased at a similar rate. Judg- 
ing from these rates, mortality in the immature and adult age classes 
is low. 

Surprisingly enough, the cygnet production rate has not kept pace 
with that of the mated pairs. The rising cygnet production curve 
mirrors some, but significantly not all, of the increases in breeding 
birds censused each year through 1951, when a population of approxi- 
mately 90 mated pairs was reached. After 1951 the continued rise 
of breeders was followed by a decline in cygnet production when it 



148 POPULATION 

Table 9. — Trumpeter swan production data, 1931 to 1957 



Year 


Mated 

pairs 

censused 


Cygnets 
censused 


Cygnets 

per mated 

pair 


Immatures 
and adults 
censused 


Cygnets: 
immature- 
adult ratio 


1931 




15 
12 
17 
49 




20 
57 
49 
48 


0.75 


1932 






.21 


1933 






.35 


1934 






1.02 
















23.25 




43.50 


.58 












1935 




27 

41 

77 
55 




46 
76 
81 
93 


.59 


1936 -- 






.54 


1937 






.95 


1938 






.59 
















50.00 




74.00 


.67 












1939 


46 


76 
68 
69 
53 


1.65 


123 
123 
143 

98 


.62 


1940 


.55 


1941 


47 


1.47 


.48 


1942 


.54 




Mean __________ . 










46.50 


66.50 


1.56 


121.75 


.55 








1943 




34 

72 
55 
72 




137 

207 
180 
289 


.25 


1944 


64 


1.12 


.35 


1945 


.31 


1946 


65 


1.11 


.25 










64.50 


58.25 


1.12 


203.25 


.29 








1947 


78 
74 
85 
68 


60 
106 
103 

73 


.77 
1.43 
1.21 
1.07 


292 
312 
348 
303 


.21 


1948 


.34 


1949 


.30 


1950 


.24 










76.25 


85.50 


1.12 


313. 75 


.27 








1951 


89 
84 
106 
133 


118 
93 
99 
82 


1.33 
1.11 
.93 
.62 


417 
478 
478 
560 


.28 


1952 


.19 


1953 


.21 


1954 


.15 




Mean 






103. 00 


98.00 


1.00 


483. 25 


.21 








1955 


123 
137 

108 


95 

81 
89 


.77 
.59 
.82 


495 
507 
399 


.19 


1956 


.16 


1957 


.22 










122. 66 


88.33 


.73 


467.00 


.19 









1 Numbers of mated pairs approximate and not available for periods 1931-38, 1940, 1942-43, and 1945. 

dropped from 118 young in 1951 to 93 in 1952, following which the 
annual production varied between about 80 and 100 cygnets. 

Another characteristic shown in figure 44 is the much greater pro- 
portion of mated pairs which existed in relation to nonbreeders dur- 
ing the early years of the census for which data are available (1939- 
41). This is contrasted with the ratio which existed later during 
the general population rise 1950-54. This suggests to me that the 
greatly increased proportion of nocked nonbreeders which existed 
after 1950 apparently resulted from the incapability of mated pairs 
to reduce territorial claims within limited breeding habitats to the 
degree necessary to accommodate the rising numbers of potential 
nesters. Table 8 exhibits the growth of the two main flocks of non- 
breeders which inhabit Upper Red Rock Lake and Lima Reservoir. 



POPULATION DYNAMICS 



149 



Because production is relatively stable and the period of imma- 
turity comparatively long, it is difficult to trace with certainty the 
impact of any year's production on later populations of the immature 

Figure 44.— Trumpeter swan productivity rates, total population, 1931-57. 1 




1950 



1955 



o GROUPED AVERAGES (SEE TABLES 7 AND 9) 
DOTTED LINES = LEFT SCALE, SOLID LINES=RIGHT SCALE 



1 Complete data unavailable, 1931-38, 1940, 1942-43, and 1945; mated pair data 
1931-57 approximate only. 



1 50 POPULATION 

and adult ago classes. However, the near-peak production years of 
1948, 1949, and 1951 apparently acted, possibly in concert with other 
factors, to cause the pronounced upward trend of nonbreeders in 1951, 
1952, and 1954. This, in turn, appears to have resulted in the marked 
rise in numbers of mated pairs in 1953, 1954, and 1956. If cause and 
result relationships exist here, a 5-year breeding age is indicated, at 
least during periods of high breeding populations. 

Although the rate of increase was previously higher, production 
reached a high plateau during the 1948-52 interval when a average 
of 80 pairs produced 99 cygnets annually by census time. This con- 
trasts sharply with the 5-year period which followed, 1953-57, when 
an average of 121 pairs produced only 89 cygnets yearly. In this case 
an increase of 151 percent in breeders was followed by a cygnet pro- 
duction decrease of 10 percent. This hints that the higher popula- 
tions of breeders may have depressed productivity. 

When the related ratios are graphed, that is, cygnets to mated pairs 
and cygnets to immatures and adults, the apparent inverse relationship 
between the population level and productivity becomes clearer. The 
data for 1955, 1956, and 1957 are shown separately, to bring out the 
direct relationships suggested by the grouped averages, though, of 
course, there is much more possibility for the factor of chance to enter 
in these individual cases. The element of chance in the grouped 
average curves is, for all practical purposes, insignificant. 

In figure 44, a number of other productivity relationships are also 
apparent. The grouped averages of immatures and adults combined 
are shown increasing at a constant rate from 1934 to 1954, after which 
a leveling off or decline is evident. The number of mated pairs cen- 
sused increased at a rate comparable, but not equal to that of the im- 
matures and adults. This suggests to me that a shortage of unclaimed 
breeding habitat may have caused potential mature breeders to remain 
in a flocked nonbreeding status somewhat longer, or to a somewhat 
greater degree, than would otherwise have been the case. 

Data relating to the production dynamics of the two most apparent 
United States population segments are handled under separate cate- 
gories: Red Rock Lakes Refuge populations and Yellowstone Park 
populations. (Due to a lack of comparable census data breakdowns 
for populations outside the Federal sanctuaries, it is not possible to 
study the dynamics of these populations.) It is realized that the ac- 
tivities of one population segment may affect the others, as shown by 
the following example, but it will also be demonstrated that the pop- 
ulation mechanics of the Refuge and Park flocks manifest some char- 
acteristics separately. 

A good example of the interaction existing between population 
segments is exhibited in figure 45. This shows that nonbreedins: 



POPULATION DYNAMICS 



151 



Figure 45.— Nonbreeding trumpeter swans censused at Upper Red Rock Lake 
and Lima Reservoir, 1 1940-57. 




1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 

i Lima Reservoir practically dry 1954-56 summer season. 

trumpeters which had habitually occupied the Lima Reservoir moved 
into Upper Red Rock Lake during the low water conditions in the 
Reservoir storage pool during 1954, 1955, and 1956. With the return 
of higher water conditions in the Reservoir in 1957, the census revealed 
an increased number of swans there. 
RED ROCK LAKES REFUGE POPULATIONS 

Since the breeding swans inhabiting the Red Rock Lakes Refuge 
play a major role in determining the status of the total population in 
the whole region, the census data for this area during the period 
1932-57 were broken down insofar as possible, compiled in table 10, 
and shown graphically to arithmetic scale in figure 46. Semilog 
curves are shown in figure 47 for all cases where data breakdowns 

were possible. 

Interestingly, the consistent inverse ratios expressed by breeder 
numbers in relation to production in the total population (figure 44) 
are shown to be even more pronounced for the Refuge population 
(figure 47). 

It should be explained that while a certain proportion of paired 
swans inhabiting Refuge marshes each year apparently do not nest, 
the ratio of these pairs to known nesting birds is small. It is prob- 
ably less than 10 percent for the whole marsh most years. Since so 
few nonnesting pairs were apparent in the initial phases of the study, 
their numbers were not recorded, but later, when the numbers of 
mated pairs were much higher, it is recalled that nonnesting pairs 
were more regularly observed. I therefore suspect that the numbers 

469660 O — 60 11 



152 



POPULATION 



of nonnesting pairs may increase at a disproportionate rate during 
years of high breeding population. If true, this may significantly 
affect the rate of production which I attributed to mated pairs. How- 
ever, the great majority of pairs on the Refuge nest and lay eggs, even 
in years of high population density, so the production rate decline so 



Figure 46. — Trumpeter swan census, Red Rock Lakes Refuge, 1932-57 




1932 1934 



1954 1956 



Table 10. — Trumpeter swan production data, Red Rock Lakes Refuge, 1936 to 

19571 



Year 



1936. 
1937. 
1939. 



Mean. 



1940. 
1942. 



Mean. 



1951. 
1952. 
1953. 
1954. 



Mean. 



1955. 
1956. 
1957. 



Mean. 



Mated 

pairs 

censused 



17.50 



12.50 



47.66 



.50 



Broods 
censused 



12.00 



Cygnets 
censused 



45.33 



Broods 

per mated 

pairs 



0.64 
.86 
.79 



Cygnets 

per mated 

pair 



1.86 
3.64 
3.11 



2.87 



2.53 
2.69 



2.61 



2.45 
1.77 
1.12 
.52 



1.47 



.93 
.62 

1.25 



Cygnets 
per brood 



2.89 
4.25 
3.93 



3.69 



4.00 
3.31" 



3.66 



3.80 
2.75 
2.71 
1.87 



2.78 



3.42 
2. 44 
2.65 



1 Data grouped to correspond with comparable information in table 12; unavailable prior to 1936 and for 
1938, 1941, and 1943-50. 



POPULATION DYNAMICS 



153 



evident during the years of high populations is due principally to the 
failure of eggs to hatch and young to survive. Many possible theories 
for such failures could be advanced, but the underlying causes have 
not yet been systematically investigated. 

Figure 47. — Trumpeter swan productivity rates, Red Rock Lake Refuge, 1936-57. 1 

100 

90 
80 




1931 



1935 



1940 



1945 



1955 



O GROUPED AVERAGES (SEE TABLE 10) 
DOTTED LINES = LEFT SCALE, SOLID LINES = RIGHT SCALE 



1 Data unavailable before 1936, and except for 1938, 1941, and 1943-50. 



154 



POPULATION 
Table 11. — Variations in Refuge swan nesting density 





Paired swans observed 2 , 3 


Year' 


Upper 

Lake; 2,880 

acres 


River 

Marsh; 

8,000 acres 


Swan Lake; 
400 acres 


1937 

1939 

1940 

1941 * 


6 (9) 

8 (12) 

6 (10) 

(8) 


6 (13) 
12 (26) 
12 (18) 
10 (14) 


2 (8) 
6 (8) 
2 (0) 
2 (0) 


Mean 


6.7 (9.75) 


10 (17.75) 


3 (4) 


1954 

1955 

1956 

1957 


16 (3) 
12 (14) 
12 (17) 
12 (6) 


56 (16) 
46 (14) 
54 (9) 
48 (27) 


14 (6) 
12 (0) 
18 (3) 
18 (6) 


Mean 


13 (10) 
+94 (+2) 


51 (16.5) 
+410 (-7) 


15.5 (3.75) 
+416 (-6) 





1 Census data for 1938 not available in habitat breakdown form. 

2 Paired swans observed at census time (August) for the period 
1937-41. Swans actually observed nesting (June) during the 1954- 
57 period, except for River Marsh data which are obtained from 
(aerial) census figures, as part of the marsh cannot be seen from 
lookout posts. Comparable data for Lower Red Rock Lake are 
not available because habitat boundary lines are not clear. 

3 Cygnets censused are shown in parentheses. 

4 Breakdown for Upper Lake population not available. 

It will be noted from table 10 that the averages of pertinent cygnet 
production data between the two periods of 1936-42 and 1951-57 
differ markedly. For instance, during the period of comparatively 
low breeding population (1936-42), 74 percent of the paired swans 
censused were seen with young, while broods averaged 3.7 cygnets 
each. During the 7-year period of comparatively high breeding pop- 
ulation ( 1951-57) the percentage of paired swans censused with young 
dropped to 39 percent and the average brood declined 24 percent, from 
3.7 to 2.8 cygnets per brood. Stated another way, a 250-percent in- 
crease in the average numbers of mated pairs censused on the Refuge 
between the periods 1936-42 and 1951-57 resulted in virtually no 
change in average annual cygnet production, being 45.4 cygnets dur- I 
ing the 1936-42 period and 46.0 cygnets in 1951-57. Considering the 
percent of pairs with broods for the period 1936-42 versus 1951-57, 
there is a highly significant difference between the means, the odds 
being more than 2 million to 1 against a difference as great due to 
chance; the odds against a difference as great due to chance for the 
period 1957 versus 1951-56 are 2.6 to 1. 

Table 11 shows the numbers of breeding swans which were attracted 
to each of the major habitat units when the total Refuge-paired swan 
population increased about 300 percent, from the 28-40 level to that of 
86-104 magnitude. The numbers of mated swans are shown to have 
increased by different ratios on the two characteristic habitat units 
during the periods considered. On the shoreline of the Upper Lake, 
a large open-water body where nesting is restricted to the perimeter, 



POPULATION DYNAMICS 155 

the increase in breeding pairs averaged only 94 percent, while in the 
River Marsh and Swan Lake units, where potential nest ing sites occur 
more frequently and uniformly, the increase was 410 percent and 416 
percent, respectively. This points up the relative importance of habi- 
tat composition and arrangement to breeding swans when a high 
breeding population is present. 

In figure 47, the numbers of mated pairs have shown an increasingly 
rapid rise from 1936 to 1956. The number of broods increased ini- 
tially at a rate comparable to that of mated pairs, but this trend 
shows progressively greater signs of leveling off. The total number 
of cygnets produced annually rose only slightly from 1936 to 1954, 
but declined markedly during 1955 and 1956 when the numbers of 
mated pairs were highest. 

The related productivity ratios, broods to mated pairs, cygnets to 
mated pairs, and cygnets to broods all declined at a similar rate, 
though in the opposite direction, from the increasing rate shown for 
mated pairs. When all curves are compared for 1955, 1956, and 1957, 
a strong inverse relationship between population density and produc- 
tivity is apparent. Although the element of chance may enter to some 
degree in the case of the individual years, it is outside the realm of 
expected possibility in the cases of grouped averages. 

YELLOWSTONE NATIONAL PARK POPULATIONS 

As one might expect, a breakdown of Yellowstone Park's trumpeter 
census statistics over the 1931-57 period reveals that factors similar 
to those manifest in the Refuge also have influenced populations there. 
This is evident even though the breeding habitat in the Park is much 
more varied and discontinuous than in the Refuge. 

As figure 48 shows, the Park's swan numbers during the past 27- 
year period have been characterized by several features. One has 
been the relatively constant population of paired swans supported 
prior to about 1949, considering the inadequacy of the very early 
census coverages. Another characteristic is the fairly constant rate 
at which cygnets have been produced during this long period. For 
most years since about 1949 the greater numbers of paired swans 
censused also coincided with a somewhat depressed productivity, 
though this relationship is not as strong as for the Refuge. 

Figure 49, graphed from data compiled in table 12 in a form 
similar to the preceding Refuge population study, throws further light 
on these different expressions of population dynamics. From these 
statistics it is evident that the average number of paired swans cen- 
sused in the Park during the periods of 1931-39 and 1940-50 rose only 
slightly, from 10.25 to 11.75, whereas from 1951 to 1957 the average 
rise was much more pronounced, to 15.9. This is an increase of about 



1 56 POPULATION 

Figure 48. — Trumpeter swan census, Yellowstone Park, 1931-57. 



IMMATURES AND ADULTS 
(CHIEFLY PAIRS) 





CYGNETS 




1931 1932 1934 1936 

O INCOMPLETE CENSUS 



1940 1942 

-NO CENSUS 



35 percent from the lowest pair population in the Park to the highest, 
compared with an increase of paired swans on the Refuge of over 250 
percent from 1939 to 1957. It should be noted, however, that the 
average number of cygnets per brood censused in the Park also 
dropped, from 3.0 during the 1931-39 period to 2.5 during the 1951-57 
interval, a decline of 17 percent. This is compared with a similar 
drop of 24 percent on the Refuge, from 3.0 to 2.8 cygnets per brood. 

In figure 49, the decline in the number of cygnets per mated pair 
in the Park also closely parallels a downward trend of similar pro- 
portions which occurred substantially during the same period in the 
Refuge (figure 47). While table 12 shows the average number of 
cygnets per mated pair in the Park to have declined only slightly, 
from 1.20 to 1.10, between the periods 1931-39 and 1940-50, it dropped 
to 0.72 cygnet per brood during 1951 to 1957 when the number of 
mated pairs was highest. For the Refuge, table 10 lists 5 years' data 
collected from 1936 to 1942 showing an average of 2.77 cygnets for 
each pair, this average dropping to 1.24 for the period 1951-57. 

Although the ratio of broods to mated pairs in the Park did vary 
for the periods 1931-39 versus 1940-50, there is no significant dif- 
ference between the means. For the period 1940-50 versus 1951-56, 
however, the odds against a difference as great due to chance are 18 
to 1; while for 1957 versus 1951-56, the odds are 13 to 1. 

Unlike the situation at Red Rock Lakes, the decline in average 
number of cygnets per pair in the Park has been compensated for by 



POPULATION DYNAMICS 



157 



a gradual rise in the average number of pairs having broods. These 
factors have almost exactly balanced, with the result that the average 
annual production of cygnets in the Park has remained practically 

Figure 49.— Trumpeter swan productivity rates, Yellowstone Park, 1931-57. 1 




o GROUPED AVERAGES (SEE TABLE 12) 
DOTTED LINES = LEFT SCALE, SOLID LINES = RIGHT SCALE 



'Data unavailable for 1942-43, 1945; incomplete 1940; for mated pairs, 1931-57 ap- 
proximate only. 



158 



POPULATION 



Table 12.— Trumpeter 


swan production data, Yellowstone Park, 1931 


to 1957 « 


Year 


Mated 

pairs 

censused 2 


Broods 
censused 


Cygnets 
censused 


Broods 

per mated 

pairs 


Cygnets 

per mated 

pair 


Cygnets 
per brood 


1931 


9 
7 
9 
8 


4 
1 
2 

5 


12 
2 
8 

17 


0.44 
.14 
.22 

.63 


1.33 
.29 
.88 

2.13 


3.00 
2.00 
4.00 
3.40 


1932 


1933 


1934... 




Mean __ _ 


8.25 


3.00 


9.75 


.36 


1.16 


3.10 




1936 


10 
10 
12 

17 


5 
6 
4 
5 


13 

26 
4 
17 


.50 
.60 
.33 
.29 


1.30 

2.60 

.33 

1.00 


2.60 
4.33 
1.00 
3.40 


1937 


1938 


1939 




Mean... 


12.25 


5.00 


15.00 


.43 


1.31 


2.83 




1940 


8 
10 
12 
13 


6 
6 
4 
3 


14 

15 

8 
8 


.75 
.60 
.33 
.23 


1.75 
1.50 
.66 
.62 


2.33 

2.50 
2.00 
2.66 


1941 


1944. __ 


1946 




Mean 


10.75 


4.75 


11.25 


.48 


1.13 


2.37 




1947 


14 
10 
12 

15 


3 

4 
6 
5 


8 
13 
21 
16 


.21 

.40 
.50 
.33 


.57 
1.30 
1.75 
1.07 


2.66 
3.25 
3.50 
3.20 


1948 


1949 


1950 




Mean 


12.75 


4.50 


14.50 


.36 


1.17 


3.15 




1951 


13 

10 
18 
19 


5 
6 
4 

7 


11 

10 
10 
23 


.38 
.33 
.40 
.37 


.84 
.55 
1.00 
1.21 


2.20 
1.66 
2.50 
3.29 


1952 


1953 


1954 




Mean 


15.00 


5.50 


16.00 


.37 


.90 


2.41 


1955... 


20 (28) 
17 (21) 
14 (14) 


4 
3 

7 


10 
9 
16 


.20 
.18 
.50 


.50 
.53 
1.14 


2.50 
3.00 


1956 


1957 






Mean 


17.00 


4.66 


11.66 


.29 


.72 


2.60 



1 Data unavailable 1935, 1942-43, 1945; incomplete 1940. 

2 Approximate number 1931-57; exact data 1955-57 in parentheses. 

constant over the 27-year period ; namely 12.4 cygnets annually from 
1931 to 1939, 12.9 cygnets annually from 1940 to 1950, and 12.7 young 
annually from 1951 to 1957. 

Figures 47 and 49 allow close comparisons of Park and Refuge swan 
productivity characteristics. For instance, the numbers of mated 
pairs in the Park increased at a much lesser rate than in the Refuge 
until about 1946, after which they accelerated and then fell off to a 
degree comparable with that characterizing the Refuge population. 

The total number of cygnets produced in the Park until 1954 in- 
creased at a slightly greater rate, though more irregularly, than on 
the Refuge, after which a similar leveling off is indicated in both 
environments. The number of broods censused in the Park decreased 
at only a slight rate until 1950, after which it dropped at a rate com- 
parable to that shown for the Refuge. 

As would be expected from the slower rate of increase of mated 
pairs in the Park, the productivity ratios of broods to mated pairs, 



POPULATION DYNAMICS 



159 



Table 13.— Characteristics of some Yellowstone Park lakes and their record of 
use by swans, 1931 to 1957 * 



Lake 


Acres 


Remarks 


Years of 
record 2 


Total 
number 

years 
occupied 


Total 

number 

broods 

produced 


Maximum 
number 
adults 

occupying 
lake dur- 
ing years 
of record 




66 

9 

109 
20 

13 
90 

439 

8,475 

2,730 
29 

19 

14 

2,926 

97 

20 
232 

64 

659 

40 


Adequate cover. Good food 
production. 

Cover rather poor. Food 
abundant. 

Ideal. Good food and cover... 

Adequate cover. Abundance 
of food. 

Poor cover but ample food 

Visited by too many fisher- 
men. 

Adequate cover. Good food 
production. 

West end highly favorable for 
waterfowl. 

Often disturbed by fishermen. 

Limited cover. Near high- 
way. Good food supply. 

Frequented by too many 
people. 

Seems suitable. Water level 
low in Fall. 

Visited by too many fisher- 
men. 

Cover limited. Otherwise 
lake is fine. 

Seems suitable for swans 

Seems suitable. Cover lim- 
ited. 

Large shallow areas dry up — 

Limited feeding areas 


24 
18 

23 

24 

20 
19 

24 

24 

23 
22 

22 

17 

20 

23 

13 

18 

17 
18 
22 


24 

18 

20 
17 

16 
15 

19 

18 

16 

14 

13 

7 

3 

19 

10 
12 

11 
9 
2 


13 

10 

9 
10 

9 
9 

3 

4 

3 
5 

4 

1 

2 











4 


Madison Junction __ 


2 

4 
3 






5 
4 




Whiti 


9 




4 




4 
2 

2 


Beach Spring 




2 




2 




7 


Wolfe 


6 
4 




Lilv Pad -.- 


4 
6 




Lake of the Woods. . 


Seems suitable for swans 


2 



Data obtained from Condon's (1941) MS plus annual Park and Refuge swan census reports since 
that time. 
2 Period 1931-57 except for 1942, 1943, and 1945. 

cygnets to mated pairs, and cygnets to broods, held fairly constant 
from 1934 to 1950 with much lesser expression of definite trends than 
that exhibited by the Refuge population. After 1950, however, all of 
the Park's productivity indicators declined when mated pairs in- 
creased further, though this is shown not to have occurred as consist- 
ently or at as rapid a rate as in the Refuge. 

A comparison of the Park and Refuge productivity data suggests to 
me that environment factors in the dissimilar habitat of the Park, 
other than population density, are more variable and influential in 
regulating annual production there. 

In order to illustrate the variable capacity of the Park lake habitats 
to support swan broods, a sample, of counts on breeding habitats was 
chosen from those areas on which swans have been censused for a 
period of 10 years or over. This information is shown in table 13. 
An examination of these data points up the limited capability of the 
lake habitat in the Park to support broods regularly, the upper limit 
actually being about 55 percent of the time over a long period (Madi- 
son Junction and Tern Lakes). On some other waters, in spite of 



160 



POPULATION 



regular records of swan occupancy, not a single brood has been pro- 
duced. The brief "remarks'' column does not entirely explain why 
breeding pairs select certain lakes and leave others unoccupied, nor 
why certain nesting waters are consistently more productive than 
others. 

POPULATIONS OUTSIDE RED ROCK LAKES REFUGE 
AND YELLOWSTONE NATIONAL PARK 

The annual swan tallies made over the years covering scattered 
areas contiguous to Yellowstone and Red Rock Lakes have not been 
quite as complete or consistent as on the federally administered areas. 
As a result, these data are not so reliable, especially before the aerial 
methods were employed in 1946, although they are believed to be 
highly representative. Furthermore, since census figures prior to the 
1950's were generally "lumped," it is not possible to prepare a table or 
graph of productivity data as was done in the case of Red Rock Lakes 
and Yellowstone Park populations. 

Figure 50 has been prepared from the most comprehensive infor- 
mation available in Park and Refuge files to show the increase in these 
"outside'' populations. An especially rapid population rise is shown 
for the areas outside the Refuge and Park in 1946. This particular 
increase is believed to have been more apparent than real, reflecting 
the greater accuracy and coverage of the aerial census method which 
was employed for the first time. The apparent rise in population 
numbers was greater on the "outside" areas mainly because the large 

Figure 50.— Trumpeter swan census outside Red Rock Lakes Refuge and Yellow- 
stone Park, 1931-57. 



- 


- 


IMMATURES 


AND ADULTS / \ A / 




CYGNETS /\/\ 







1931 1932 1934 1936 

O=N0 CENSUS 



POPULATION DYNAMICS 



161 



reservoirs outside the Park or Refuge boundaries were covered effi- 
ciently and completely for the first time. 

The most significant feature of figure 50 is the tendency of areas 
outside the Park and Refuge not only to hold increasingly greater num- 
bers of swans but at the same time to produce a generally rising 
number of cygnets as well. This is in contrast with the trend in Yel- 
lowstone Park and the Refuge where production recently declined. It 
is plain that habitat outside the two Federal sanctuaries is becoming 
increasingly important as the total number of swans in protected 
areas rises. 

SUMMARY, POPULATION DYNAMICS 

Total Population: 

1. The population increased at a constant rate from about 1935 until 
1954 (about 10 percent annually). Beginning in 1955 a definite 
leveling out became evident. 

2. The rates of change for the total population reflected the survival 
rate of the immature-and-adult age class to a much greater extent 
than the progressively declining cygnet production rate. Popula- 
tion turnover (mortality) in the immature and adult age classes is 
thus shown to be low. 

3. Mated pairs increased at a rate approaching, but not equal, that 
of the immature-and-adult age class. This suggests that a shortage 
of desirable unclaimed breeding territories may prolong the non- 
breeding status of potential nesters. 

4. During periods of increasing populations at high levels, cygnets 
were produced at a rate progressively less than during periods of 
increasing populations at low levels. 

5. During low levels, mated pairs comprised a much greater propor- 
tion of the total population than at high levels. 

6. There is an indication that mated pairs initially occupied breeding 
territories as 5-year-olds, at least during periods of high popula- 
tion levels. 

7. Cygnet production reached a plateau from 1948 to 1952. 

8. An inverse relationship between population levels and productivity 
is shown. 

Yellowstone Park Population: 

1. Considering the inadequate coverage of the early years, the number 
of mated pairs remained nearly constant from 1931 to about 1952. 
A definite, increase occurred beginning in 1953. 

2. Broods were produced at a nearly constant rate from 1931 to 1957. 

3. Cygnets were produced at a nearly constant rate from 1931 to 1953, 
and in ratio to the number of mated pairs present ; after 1953 the 



162 POPULATION 

number of mated pairs increased and the production of cygnets 
declined. 

4. The ratio of broods to mated pairs increased at a constant rate 
from 1931 to 1945 and then changed, but generally decreased at an 
irregular rate from 1946 to 1954. After 1954 the rate of decline 
accelerated, coincident with the further increase in mated pairs. 

5. The rate at which cygnets were produced per brood remained 
nearly constant from 1931 to 1950; a decline occurred after 1950 
with a further increase in mated pairs and the number of cygnets 
produced per brood remained low until 1957. 

6. The rate at which cygnets per mated pair were produced was nearly 
constant from 1931 to 1950 and roughly proportional to the number 
of mated pairs present; after 1950 the cygnets per mated pair rate 
changed inversely with the change in number of mated pairs 
present. 

7. Breeding habitats are "saturated" when occupied by about 15 pairs, 
greater breeding populations appear to depress productivity. 

Red Rock Lakes Populations: 

1. The number of mated pairs increased at a substantially constant 
rate during the 1936-57 period. 

2. From 1936 to 1954, the increasing rate at which broods were being 
produced slightly outweighed the consistent declines in the ratios 
of cygnets to broods and broods to mated pairs. The result was a 
slight increase in the cygnet-production rate. 

3. In 1955 and 1956, when populations of mated pairs were highest, 
cygnet production declined. The falling ratios of broods to pairs 
and/or cygnets to broods during this period resulted in the pro- 
duction drop. In 1957, when the number of mated pairs dropped, 
the production rate went up due to an increase in these same 
productivity factors. Thus, cygnet production is shown to be in- 
versely related to the number of mated pairs when high popula- 
tions are present. 

4. Breeding habitats are "saturated'' when occupied by about 40 pairs, 
greater breeding populations appear to depress productivity. 

Population Outside Federal Sanctuaries: 

1. The census data for this segment does not lend itself to the detailed 
analysis possible with other populations. 

2. The total population, and cygnet production, show consistent 
through different growth rates. 

Red Rock Lakes versus Yellowstone Park Populations: 

1. The point of diminishing productivity was reached about the same 
time (1955) in both of these habitats. In view of the long period 



CONCLUSIONS 



163 



when relatively constant numbers of mated pairs occurred in the 
Park, during which they increased about 300 percent on the Refuge, 
some' interaction between these populations is suggested. It ap- 
pears that at "supersaturated" levels, Refuge pairs may overflow 
to Park habitat, depressing the productivity rate there. The simi- 
lar drop of mated pairs in the Refuge and Park between 1955 and 
1957 leads to the same conclusion— that interaction may occur at 
high levels but not necessarily at low. 
2. Where breeding territories are contiguous, as within the Refuge, 
increasing breeding populations at "saturation" levels or below 
exert a depressive effect on the potential production rate. This 
is apparently not true in isolated habitats, as in Yellowstone Park 
where the cygnet production rate tends to change directly with the 
number of mated pairs present, unless a "supersaturated" breeding 
population is sustained when the rate declines. 

CONCLUSIONS 

1. Trumpeter swans in the tristate region of Montana, Wyoming, and 
Idaho should no longer be considered an endangered species; this 
population has increased at a constant rate for 20 years, and it now 
exhibits signs of leveling off at the maximum level this environ- 
ment will support, 

2. The rate of population change in the total population varied in- 
versely with changes in population density. This change occurred 
in two habitats of varying capacity, Red Rock Lakes and Yellow- 
stone Park, and during periods of low and high populations, hence 
the phenomena is relative and probably occurs to some degree in 
all swan-occupied habitat, 

3. The dynamics of the change in population are tied to the rate of 
productivity of mated pairs, which varied inversely with changes 
in population density. 

4. The rate of increase approached zero about 1954 when the popula- 
tion peak was reached. After that, except for 1957 when an exodus 
apparently occurred, the population fluctuated around 590, ap- 
parently nearly the maximum population which can regularly be 
sustained. Continued population and production increases on areas 
outside Federal sanctuaries will raise this estimate if further ex- 
pansion of their range occurs or more isolated breeding habitats 
within the existing environment are successfully occupied. 

DISCUSSION 

Howard and Fiske (1911) were the first to show that natural popu- 
lations tend to vary inversely with population density, and Nicholson 



164 POPULATION 

(1933 : 132-178) and Smith (1935 : 873-898) also pointed out that the 
comparative stability of natural populations is controlled dynam- 
ically, meaning that the farther that numbers rise or fall, the stronger 
is the tendency to return to the previous level (Lack, 1954). 

Lately Christian (1957: 443-62) has shown that numerous field and 
laboratory studies involving mice support a theory that the growth 
of populations may be regulated and limited by sociopsychological 
factors, (social competition) through the production of stress in some 
proportion to population density. This possibility has not been as 
thoroughly studied in wild avian populations, presumably because of 
few hints in this direction and difficulties inherent in making accurate 
censuses of entire populations; nevertheless, there are clues to such a 
theorem. Working with great tits in Holland, Kluijver (1951) 
found that both the average clutch and the proportion of pairs raising 
second broods were lower at high than low population densities; but 
it was concluded by Lack (1954) that the difference was too small 
to have any important effect on subsequent numbers. Errington 
(1945), discovered that the summer gain in Wisconsin bobwhites 
(young and adults surviving autumn) varied inversely with the num- 
bers present in April. The foregoing study, though concerned only 
with swan populations, parallels Errington's findings. 

Since by far the great majority of paired swans on Refuge marshes 
complete nests and lay eggs, population density-dependent variations 
in the number of cygnets produced may be substantially attributed to 
the hatching and survival rate. Just how the number of cygnets 
raised may be regulated by the density of mated pairs is not under- 
stood, but it seems clear that the deterioration of habitat, usually 
associated with the population declines of some wildlife species, does 
not play a major role. 





MANAGEMENT 



PROTECTIVE LEGISLATION 

The farsighted Congressional legislation which originally provided 
for the protection of wildlife in Yellowstone National Park, the Lacey 
Act of May 7, 1894, furnished essential protection for the ancestors of 
the few pairs of trumpeter swans which were discovered breeding in 
this famous Park in the summer of 1919. The early protective wild- 
life regulations which grew out of this initial National Park legisla- 
tion were forerunners of continent wide laws which first applied 
specifically to waterfowl and later to waterfowl refuges. Each link 
of legislation which protected the trumpeter was forged as part of a 
greater plan to perpetuate portions of the representative native fauna 
in their natural environment. 

Unfortunately, the passage of the second Lacey Act in 1900, the 
Weeks-McLean Law in 1913, and the Migratory Bird Treaty Act in 
1918 arrived much too late to prevent the extirpation of the trumpeter 
over most of its United States breeding range. For two decades after 
1900, a number of prominent American scientists interested in the 
problems of species survival commented on the fate of the trumpeter 
swan. William T. Hornaday (1913: 19) reported that in 1907 these 
swans were regarded as so nearly extinct that a doubting ornithologi- 
cal club of Boston refused to believe on hearsay evidence that the 
New York Zoological Park contained a pair of the living birds, 
and a committee was appointed to investigate in person and report. 
Edward Howe Forbush (1912: 175), an eminent ornithologist, 
lamented : 

The trumpeter has succumbed to incessant persecution in all parts of its 
range, and its total extinction is now only a matter of years. . . . The large 
size of this bird and its conspicuousness have served, as in the case of the 

165 



166 MANAGEMENT 

whooping crane, to make it a shining mark, and the trumpetings that were 
once heard over the breadth of a great continent, as the long converging lines 
drove on from zone to zone, will soon be heard no more. 

Passage of the Migratory Bird Treaty Act 6 years later placed a 
closed season on both species of native swans for the first time when 
it became effective in 1918. This was the first aid to survival of the 
few trumpeters which still existed outside Yellowstone Park bound- 
aries and which were to be so important in the eventual restoration of 
the species. 

In 1929, the Migratory Bird Conservation Act authorizing the 
acquisition of land for waterfowl refuges was passed by Congress. 
When supported with funds in 1934, this basic waterfowl legislation 
was as important in providing for the future increase of the United 
States trumpeter flock as the Migratory Bird Treaty Act was in 
protecting the remnant populations. 

Under the Migratory Bird Conservation Act, the Red Rock Lakes 
Migratory Waterfowl Refuge in southwestern Montana was 
established by Executive Order in 1935. This area, containing thou- 
sands of acres of historic trumpeter swan breeding habitat, was sub- 
sequently staffed by the Biological Survey, a predecessor of the 
United States Fish and Wildlife Service. . The 22,682-acre area 
originally set aside under this Order was enlarged in September of 
the same year, when about 18,000 additional acres were included in 
the Refuge in order to complete the breeding-ground acquisition and 
to bring under management certain warm spring-water areas im- 
portant to the swans during the winter months. Although several 
management problems remained, the establishment of this Refuge 
provided the upward turning point for this species in the United 
States. While the status of the trumpeters in Yellowstone Park 
before the establishment of the Refuge was marginal, it was ap- 
parently improving slowly. But the existence of this species outside 
the Park was actually in jeopardy by the early 1930 , s. 

Later, since it was discovered that irresponsible waterfowl hunters 
were killing swans under the claimed pretext of shooting snow geese, 
the hunting seasons on snow geese were closed in those States within 
the trumpeters' winter range. Service regulations, stemming from 
authority in the Migratory Bird Treaty Act, closed the snow-goose 
season first in the State of Idaho in 194f , and were modified about a 
decade later to exclude the counties where swan shooting was not a 
problem. In Montana the snow goose closure has been in continuous 
effect in Beaverhead, Gallatin, and Madison Counties from 1942 until 
the present. Similar Federal closures were initiated as a statewide 
measure in Wyoming in 1946 because of the threat posed to the newly 



CAPTIVITY RECORD 167 

established trumpeter flock introduced to the National Elk Refuge in 
Jackson Hole; however, these regulations were modified in 1955 to 
include only the pertinent areas of Teton and Lincoln Counties. 

Unless these protective measures had been taken by the Federal 
Government, this large and conspicuous bird would surely have dis- 
appeared from its native breeding grounds in this country. 

CAPTIVITY RECORD 

The history of the trumpeter in captivity is long, varied, and 
interesting. It is valuable as avicultural history, for information 
on the traits this bird displayed in captivity, its breeding record, 
and the success which was attained by some of the various individuals 
or organizations which kept it in confinement. 

Audubon (1838: 537, 541) has left us with an early note or two 
on the subject, writing : 

I have traced the winter migrations of this species as far southward as the 
Texas . . . where I saw a pair of young ones in captivity, and quite domesti- 
cated, that had been procured in the winter of 1836. 

******* 

I kept a male [trumpeter] alive upwards of two years, while I was residing 
in Henderson in Kentucky. It had been slightly wounded in the tip of the 
wing, and was caught after a long pursuit in a pond from which it could not 
escape. . . . Although at first extremely shy, it gradually became accustomed 
to the servants, who fed it abundantly, and at length proved so gentle as to 
come to my wife's call, to receive bread from her hand. "Trumpeter" . . . 
now assumed a character which until then had been unexpected, and laying 
aside his timidity became so bold at times as to give chase to my favorite 
Wild Turkey Cock, my dogs, children and servants. Whenever the gates of 
our yard happened to be opened, he would at once make for the Ohio, and it 
was not without difficulty that he was driven home again. ... in the course 
of a dark and rainy night, one of the servants having left the gate open, 
Trumpeter made his escape, and was never again heard of. 

The next record of the trumpeter's history in captivity seems to be 
that left by Baird, Brewer, and Kidgeway (1884: 432), as they state: 

A nest of this species was found by Mr. W. C. Rice at Oakland Valley, la., 
in the spring of 1871, and the Cygnets taken from it. Three of these were 
successfully raised, and were purchased for the Mount Auburn [Iowa] Cemetery, 
where they were received in December. They bore their transportation, in a 
week of unparalleled severity for the season, without injury, and were remark- 
ably docile and tame. In the summer months when at large they would leave 
their pond and seek the companionship of their keeper, whose occupation as 
painter occasionally required his presence on the grounds near their place of 
abode. If permitted, they would spend the day in his company rather than 
remain in their pond. They were perfectly and completely domesticated and 
showed no fear of any person, feeding from the hands of any stranger. This 
swan has also been domesticated in the cemetery in Cincinnati, a pair of the 

469660 O — 60 12 



1 68 MANAGEMENT 

progeny having been sent to the London Zoological Gardens, and another to 
Mount Auburn. 

Jean Delacour (1954: 78, 79) provides a recent synopsis of both the 
foreign and domestic captivity record of this species. He begins 
with the arrival of the first specimens in Europe in 1866, which were 
apparently the progeny shipped from the Cincinnati cemetery just 
mentioned : 

Trumpeter swans do very well in captivity. They soon become tame, prove 
completely hardy, easy to feed, and they breed readily. The first specimens 
to reach Europe arrived at the London Zoo in 1866, and the earliest breeding 
success in captivity took place there in 1870 when three cygnets hatched on 
July 6 and were reared on the Three-Islands Pond. Other successes followed. 
The species reached Paris, at the Jardin des Plantes, in 1873, but not until 
1880 were eggs laid, and these proved infertile. In 1885, however, five eggs 
hatched and four cygnets were reared. 

Later on, Trumpeters bred regularly in Europe, but since the beginning of 
this century, only the Duke of Bedford, in England, and F. Blaauw, in Holland, 
seemed to have possessed and raised any. 

******* 

Blaauw reared Trumpeters at Gooilust almost every year until his death 
in 1936. There were then sixteen birds left, old and young, and they all went 
to Woburn Abbey, as they had been bequeathed to the Duke of Bedford, who 
still possessed four. Unfortunately, all these birds died through lack of food 
during the 1939—15 war. . . . 

We kept only one pair at Cleres [France], in 1920 and 1921. They were two 
years old when they arrived from Gooilust, and they were placed on the lake. 
They swam and walked all over the park, at great speed. Very soon they 
became quarrelsome and attacked large birds, particularly white Rheas, which 
they pursued relentlessly over the hills. It became impossible to keep them 
at large, and as all suitable pens were already occupied, we sent them away 
to the London Zoo. 

Trumpeters kept at liberty at Woburn did not molest other birds in the huge 
park. Each pair established a territory and nested, but the young were often 
lost. At Gooilust, each pair was kept on a comparatively small pond, and bred 
regularly. 

As a result of the breaking up of F. E. Blaauw's waterfowl collec- 
tion in Holland in 1936 after his death, the disastrous results of 
World War II upon the Woburn Abbey flock of trumpeters in 
England, and Delacour's earlier abandonment of his Cleres trumpeter 
raising efforts, it is doubtful whether any of these birds remain in 
Europe or England today, except of course for the trumpeters which 
were presented to the Queen of England by the Canadian Govern- 
ment. 

Trumpeters were also apparently kept successfully in captivity in 
or near Washington, D. C, in the early days. The Patuxent Re- 
search Refuge records note that 5 sets of trumpeter eggs were in the 
collection of J. P. Norris, Jr., tagged "D. C. (captivity)" under the 



CAPTIVITY RECORD 169 

date of May 3, 189,3. No other data are furnished except that this 
report was received from A. C. Bent. 

The National Zoological Park in Washington, D. C, exhibited 
trumpeters for over 22 years, from May 24, 1000, until June 14, 1922, 
involving- a total of 8 birds. The trumpeters at the Philadelphia 
Zoological Garden have been noted under "Longevity/' 

The New York Zoological Society kept trumpeters for over 20 
years, commencing in 1899 and lasting until 1921 (W. G. Conway, 
correspondence) . A total of 22 specimens were received as purchases, 
deposits, or exchanges during this time, having been obtained from 
such States as Idaho (3 in 1889), Utah (6 in 1901), and Maine (1 in 
1901), as well as from the collections of L. C. Sanford of New Haven, 
Connecticut (4 in 1909), and F. E. Blaauw of Holland (4 in 1921). 
Apparently none of these trumpeters bred in New York. 

H. K. Coale (1915 : 89, 90) reported that, at the time of his writing, 
live Montana-originated trumpeters were in the collections of Dr. L. 
C. Sanford of New Haven, Connecticut, and Mr. John E. Thayer of 
Lancaster, Massachusetts; also that Judge R. M. Barnes of Lacon, 
Illinois, had written that he possessed 5 of the 10 trumpeters then 
known to be in captivity, though he had been unable to breed any. The 
origin of Judge Barnes's flock is unknown. 

Before the days of the enforcement of the Migratory Bird Treaty 
Act in 1918, some traffic existed in live trumpeter swans between per- 
sons able to obtain them, and those interested in their propagation or 
exhibition in captivity. 

Dr. T. S. Roberts (1936: 206) stated that an old-time resident of 
Heron Lake, Minnesota, related that pre-flight cygnets were rounded 
up and captured there for shipment to the East, presumably sometime 
after the 1850's. 

Mr. Cecil Wetmore, presently the proprietor of the Summit Hotel 
in Monida, Montana, and the oldest living pioneer of the Red Rock 
Lakes marshes, told me that about the turn of the century, when he 
was but a boy, his family captured young trumpeter cygnets for sale 
to zoos, municipal gardens, or interested individuals. The price, at 
least at one time during this period, was $50 a pair and all birds which 
remained unsold at the seasons end were liberated in order to ensure 
ample breeding stock for future years. Mr. Wetmore recalled that 
several pairs were disposed of to the Columbia Gardens in Butte, Mon- 
tana, but other purchasers had been forgotten and the pertinent 
records lost during the intervening years. 

At least one other pioneer Valley resident, Mr. Fred Hanson, cap- 
tured Red Rock Lakes swans for the live-bird trade during this period 
(Alta Hanson, correspondence). From the diary record of Lillian 
Culver it is clear that Mr. Hanson was active in supplying birds to 



170 MANAGEMENT 

the live-swan trade during the period 1896-1914, shipping them to 
various unnamed buyers located at unspecified destinations: 

1896. August 20. Fred has 15 swans and will ship tomorrow. 

September 17. Fred and Henry [Hackett, her brother] have gone after 
swan. They made $50 apiece clear on the last shipment. 

September 21. Henry left home with 12 swan this morning, he left 2. He 
has 5 at Collins and will take them. 

1897. September 12. Fred and Henry have 35 swan now. Fred expects to go 

with them soon now. 
September 19. Henry went Friday with 2 swan for King in Ohio. 
September 24. Well, Fred and Henry started at 2 p. m. with 32 swan in 

the hay rack. I hope they will not have any trouble. 

1898. August 29. Fred has 17 swan. 
November 10. Fred left with six swan. 

1899. July 25. Fred went for swan. 

1900. July 20. Home about 4 p. m. Fred and Emma gone after swan. 
September 25. B'red shipped 17 ducks. He has only 8 swan left. 

1901. October 31. Been hunting Fred's swan all day. They got out last nite. 

One died and one flew away so he has only 6 left. 

1902. October 5. Fred is ready to take swan in the morning. 
190(1 October 15. Fred went to Monida with the swan after dinner. 

1908. October 20. Matt [Reis] tore up boxes to make the swan crate and 
packed them over. He will go with the swan Friday. 
November 26. The boys got their check for swan today, $66.00. 

1911. August 31. Henery came with 10 swan and put them in the stable. We 
had dinner and I went to feed and water the swan when Fred came 
with one more. 

1914. August 31. Fred and Chester got 3 more swan. 5 new swan now. [James 
F. Hanson, son of Fred Hanson, adds that these were the last swan 
taken. Fred could not sell them and they stayed around Culver's 
pond for years.] 

Apparently one of the early day "swan-brokers'" who obtained some 
of the Red Rock trumpeters was Dr. Cecil French of Victoria, B. C, 
who wrote to the Kellogg Bird Sanctuary in 1934 (a copy was 
furnished by Dr. Pirnie) , stating : 

I have had quite a bit to do with native wild swans in my time. Until the 
days of the War [WW I] I was located in Washington, D. C, and off and on 
for several years had been engaged in collecting various birds and animals for 
Zoological institutes throughout the world. Until about the year 1913 I received 
annually, from a party in the State of Montana, from 2 to 6 young Trumpeter 
Swans, taken from the nests. These went to a collector in Boston, Mass., to the 
Duke of Bedford in England, and to a Mr. Blaauw at Hilversum, Holland. 

Patuxent Research Refuge records hold an entry that Dr. French 
reported in a letter of October 18, 1910, that trumpeter swans nested 
at Lakeview, Montana, the former postoffice address for residents of 
the Red Rock Lakes area. So apparently at least some of the early- 
day Red Rock Lakes trumpeter population found its way via Dr. 
French to the East and thence abroad to foreign aviculturists. 



CAPTIVITY RECORD 



171 




Figure 51. — Trumpeter swan cygnets captured on the Red Rock Lakes by the 
Wetmore family for the live swan trade about 1900. 

As any information regarding the whereabouts or availability of 
these rare fowl in the days before they were protected by the Migra- 
tory Bird Treaty Act could be turned to personal profit, such informa- 
tion was rarely published. No doubt the files of Dr. French and other 
early aviculturists such as F. E. Blaauw and the Duke of Bedford 
could reveal a wealth of information along these lines. 

Dr. Miles D. Pirnie (correspondence) reported that the Kellogg 
Bird Sanctuary in Michigan, with which he v T as formerly associated, 
purchased 10 or more young trumpeters from F. E. Blaauw in Hol- 
land between 1927 and 1931 at a cost of about $500 apiece. According 
to Dr. Pirnie, copulation occurred seasonally among paired indi- 
viduals but they made no attempts at nest building. The last 
pair was broken up when the male died about 1938. The sur- 
viving pen w r as then moved to the Chester K. Brooks estate near Cleve- 
land where a cob trumpeter was available. This attempt at breeding 
trumpeters in captivity also ended in failure when the pen was killed 
by a dog. Dr. Pirnie indicated that extreme emaciation due to un- 
known causes resulted in the deaths of several trumpeters at the Kel- 
logg Sanctuary. Human interference, intraspecific strife due to 
crowding, nest desertion, and malnutrition were believed to be prin- 



172 MANAGEMENT 

cipally responsible for low breeding success among the various other 
species of swans (whooper, whistling, and mute) which did breed on 
that sanctuary. 

The following statement by Dr. J. M. Derscheid (1939: 94) has 
apparently led to some speculation in the past regarding the original 
source of the trumpeter swans in Yellowstone Park. 

Mr. Blaauw told me then [upon the occasion of Derseheid's visit to Holland in 
1924] that he was not only slowly but regularly increasing the strength of that 
[trumpeter swan] stock, but that he had had much satisfaction in sending 
some of the Trumpeter Swans bred at his place to the United States Federal 
Government, with the object of restocking some American National Parks with 
this species, formerly living there, but then practically extinct. 

In spite of inquiries at the field headquarters of Yellowstone Na- 
tional Park, the likely recipient if any trumpeters had been sent, noth- 
ing could be learned which would verify this alleged transplanting. 
Regardless of Dr. Derseheid's statement, it is highly improbable that 
a Government project possessing so much inherent public interest was 
actually consummated at that early date without the usual publicity 
and deliberate recording of the facts involved. It is more likely that 
Blaauw's remark refers to the trumpeters which he shipped to the 
Kellogg Bird Sanctuary. Trumpeter swans (4) were noted in Yellow- 
stone Park as early as 1915 (M. P. Skinner, 1925: 154), but the first 
breeding trumpeters were not reported until 1919, when 2 pairs were 
discovered. No special efforts to save them from extinction were made 
by the Park Service until a decade later. 

C. G. Sibley (1938: 329) reported the hybridization of the trum- 
peter swan while confined in captivity with several other species of 
Anatidae. Drawing upon the testimony of other breeders as well 
as his own experience, Sibley lists hybrids of the trumpeter cross- 
breeding with the mute, whooper, and whistling swan and the Canada 
goose. What is perhaps of even greater significance is the indication 
that hybrids between the rather distantly related swans are some- 
times fertile, this characteristic being noted in a whistling-mute cross 
(male). The young of a whistling- whooper hybrid were also said to 
be fertile. 

NATIONAL PARK SERVICE INVESTIGATIONS 

In 1929 the National Park Service launched a comprehensive 
series of biological studies which were needed to outline the most 
pressing wildlife problems which existed in the National Parks. 
Early attention was given to determining the status of the trumpeter 
in Yellowstone Park, with the aim of saving it from extinction. 
While M. P. Skinner (1925: 153-155) had outlined the precarious 



NATIONAL PARK SERVICE INVESTIGATIONS 173 

early status of the trumpeter in the Park during the period 1915-21, 
no special attempt had been made to assure its welfare since. 

Dr. Joseph Dixon, then economic mammalogist at the University of 
California, served as the held observer during the initial phase of the 
new Park Service program and conducted the preliminary swan sur- 
vey in Yellowstone Park during the fall of 1929. In the course of 
these investigations he learned that although a pair of trumpeters had 
nested every season since 1925 on a small lake near Junction Butte, 
known locally as Trumpeter Lake, they had never been known to 
raise their young to flight age. The need for study of egg loss and 
cygnet mortality was thus indicated and plans were made to do this the 
following season. 

In the spring of 1930 Dr. Dixon returned to Yellowstone Park as 
planned, reinforced by National Park Service naturalists George 
Wright and Ben Thompson. Two nesting pairs of swans, one pair 
located at Trumpeter Lake and the other at Tern Lake, were placed 
under surveillance in order to determine the causes of egg loss and 
juvenile mortality, while a preliminary general survey of other swans 
in the Park was carried out as time and opportunity permitted. 

At the end of the 1930 nesting season, only 3 cygnets were success- 
fully raised from the total of 4 nesting pairs of trumpeters found in 
the Park that summer. The 6 cygnets which hatched at Trumpeter 
Lake were all lost, dropping out 'from unknown causes by ones and 
twos over the whole span of the breeding season in spite of several 
periods of dawn-to-dusk surveillance. The only cause of mortality 
actually observed all summer was a raven which was seen to pilfer 
the contents of a swan egg at the Tern Lake nest. Plainly, heavy 
losses of eggs and juveniles were holding the swan population in the 
Park at precariously low levels. Further investigations, coupled with 
localized predator control, were planned again for the following year. 
About this time, the Park swan restoration program came to the 
attention of some of the local residents of the nearby Red Rock Lakes 
area, where these rare fowl were still commonly killed every year 
by hunters during the waterfowl season despite the Federal law. 
Letters from at least two ranchers in the Red Rock Lakes marshes, 
Mr. Clarence Hunt and Mr. A. Hayden, protested to Park authorities 
the unnecessary shooting and killing of swans in the vicinity of their 
marshland ranches. In one of these letters Mr. Hunt reported to 
Park Ranger Frank Anderson that 7 trumpeters were killed during 
the 1930 season alone. Mr. Hayden advised the Park that the num- 
bers of swans were so reduced by hunters on the Red Rock and 
Henrys Lakes that their numbers probably totaled less than 15. Both 
ranchers made separate recommendations that something be done for 
the protection of these rare birds in the Red Rock Lakes area. 



1 74 MANAGEMENT 

From 1931 until the Red Eock Lakes Refuge was established in 
1935 the Park Service spared no labor in promoting the welfare of 
the trumpeter. To the everlasting credit of that Service this was 
true not only within the Park but in the important Red Rock Lakes 
area as well. During this critical period George Wright was placed 
in charge of the program, and he, together with his associates, Dr. 
Dixon and Ben Thompson, worked tirelessly with Park Superin- 
tendent Roger W. Toll on management measures designed to protect 
and increase the trumpeter population in the whole Yellowstone 
region. 

When George Wright was specifically assigned to the trumpeter 
problem by the National Park Service in 1931, he accepted the chal- 
lenge with characteristic enthusiasm and dedication. That summer 
he initiated the first annual swan census. The life history studies 
were also commenced at this time under his supervision. These led 
to such important discoveries as the recognition of the Red Rock 
Lakes area as a major breeding grounds of the species and the fact 
that the trumpeter wintered in Yellowstone Park, a fact apparently 
not recognized before 1932. Mr. Wright also wrote letters to officers 
of the various duck hunting clubs located about the Red Rock Lakes, 
stressing the lamentable status of the trumpeter and otherwise publi- 
cizing the problem locally. Acting in the light of Mr. Wright's 
publicity of the trumpeter's plight, the effort of Mr. Frank Conley 
of Deer Lodge, Montana, resulted in the Montana State Fish and 
Game Department's becoming interested in the swan killing problem, 
and in 1933 the Commission offered a $50 reward for information 
leading to the conviction of anyone found guilty of shooting a 
trumpeter in Montana. 

Dr. Joseph Dixon also played a key role in the early restoration of 
the trumpeter. Besides his early survey work he later served as a 
consultant. It was Dr. Dixon who originally kindled public interest 
in this species by his now historic article which appeared in the August 
1931 issue of American Forests. This not only authoritatively out- 
lined the precarious position of the trumpeter during this period but 
did much to arouse public opinion in support of the progressive Park 
program, of which little was then known. 

Ben Thompson served in the early swan restoration program chiefly 
by assisting George Wright and Dr. Dixon in the field, and later by 
outlining the trumpeter's survival problems at the first North Ameri- 
can Wildlife Conference (1936: G39-L1). 

In Yellowstone Park, management measures were taken by 
Superintendent Roger Toll as a result of recommendations based 
on field investigations by Messrs. Wright, Dixon, and Thompson. 
Local predator control of coyotes and ravens was practiced, fishing 



NATIONAL PARK SERVICE INVESTIGATIONS 175 

waters were closed to prevent molestation of nesting swans, at least 
two small nesting islands were constructed to reduce the chances of 
predation, and the life history studies were continued. In one case 
Superintendent Toll decided to relocate a main Park road that was 
being constructed in order to furnish a space "buffer zone'' to prevent 
undue disturbance of nesting swans by curious visitors. 

The climax of Park Service efforts came when J. N. "Ding" 
Darling, then chief of the Biological Survey, impelled by the Park's 
swan restoration program and favorable reports, visited the Red Rock 
Lakes area in 1934 and subsequently recommended that these lakes and 
marshes be included within the Federal waterfowl refuge system which 
was then just getting underway. Twelve years previously, in the fall 
of 1922, Charles S. Sperry, a waterfowl food-habits biologist of the 
Biological Survey, had inspected the isolated Red Rock Lakes area 
and, though he reported these marshes a wonderful breeding ground 
for all kinds of wild fowl including the trumpeter, recommended 
against acquisition owing to opposition from hunting-club interests 
which had been entrenched in that area since the turn of the century. 
In 1934, following "Ding's" recommendations, these differences were 
resolved under the urgency of the swan restoration program, and in 
the following year Red Rock Lakes Migratory Waterfowl Refuge was 
created. 

After the Refuge was established and management was undertaken 
in earnest by the Biological Survey, the Park Service did not con- 
sider the swan restoration program a finished matter. Wright, then 
Chief of the Wildlife Division of the National Park Service, and 
Superintendent Toll were both killed in a tragic automobile accident 
in 1936, but the new Superintendent, Edmund B. Rogers, established 
a firm liaison with the succeeding Red Rock Lakes Refuge managers. 
Rogers' interest led to a number of progressive swan projects during 
the next 20 years. 

Several important studies of the trumpeter were made in Yellow- 
stone Park during Rogers' tenure as Superintendent. In 1938 
Frank Oberhansley, Assistant Park Naturalist, and Maynard 
Barrows, Assistant Chief Ranger, completed a season of important 
observations resulting in their 1939 manuscript. In 1941 David Con- 
don, now Chief Naturalist of Yellowstone National Park, finished 
his manuscript on the trumpeters in the Park. Shortly afterwards 
Condon was in a position, along with refuge manager Dr. Ward 
M. Sharp, to stand in united opposition against the establishment of 
a U. S. Army mountain training and artillery center at Henrys Lake, 
Idaho. Construction of this base had actually begun before conserva- 
tionists rallied behind recommendations made by Condon and 
Sharp in order to safeguard the breeding population of trumpeters at 



1 76 MANAGEMENT 

Red Rock Lakes, which was only a few miles to the west of Henrys 
Lake, and almost within the proposed artillery-target sector. 

The significant animal loss of swans to waterfowl hunters along 
Henrys Fork of the Snake River in Idaho was also brought to light 
for the first time as a result of Condon's investigations. To re- 
duce this unnecessary mortality a cooperative educational project was 
organized in which the Park Service furnished film and information, 
and the Emergency Conservation Committee, headed by Rosalie Edge 
of New York City, supplied funds for an extended series of lectures 
throughout eastern Idaho. These were presented by George Marler, 
a seasonal Park Service naturalist. This 2-year series of lectures 
reached a total of 24,687 people, and its beneficial effects continued 
long after the project had ended. 

Park Service cooperation continues to the present time. The annual 
swan census, now accomplished by aerial methods, is carried out with 
Park Service cooperation. Park Biologist Walter Kittams has served 
as an aerial observer on swan census flights over the Park in various 
U. S. Fish and Wildlife Service aircraft every year during the past 
decade. From a population standpoint this information has been 
most valuable in assessing current trends and as a guide to habitat 
evaluations. 

SWAN MANAGEMENT ON THE RED ROCK 
LAKES REFUGE 

GENERAL PRACTICES 

When A. V. Hull assumed the initial managership of the Red Rock 
Lakes Refuge in 1935, an effective liaison was soon established with 
Park officials. Under the resulting association, ideas were exchanged 
on swan restoration measures, the annual swan census was coordi- 
nated, and both offices kept abreast of new developments. In addition, 
a number of projects designed to aid the swan program were placed 
in effect on the Refuge by Hull (1939: 378-382). 

Most of these management measures were time-proved conventional 
wildlife restoration practices, and are continued where practical on 
the Refuge today. One of the most important was the elimination 
of human activity and trespass on or near the swan breeding waters. 
This was based on standard Service regulations, and in this case 
served especially well, since trumpeters are particularly sensitive to 
irregular human activity. 

Marsh management was also begun at an early date on the Refuge. 
Grazing was limited to the conservative carrying capacity of the 
range, in order to help restore the marsh itself to a natural wilderness 
nesting environment. In more recent years the trumpeters' winter 



SWAN MANAGEMENT ON RED ROCK LAKES REFUGE 177 

habitat has been improved through the development of two warm- 
water spring- fed ponds. 

Muskrat trapping was also brought under direct regulation at an 
early date and, in fact, halted for a time. Hull found that muskrat 
houses furnished the main source of nesting sites, and overlapping 
before establishment of the Refuge was apparently a major factor in 
reducing the population of these marsh rodents to a low level. Pres- 
ent management limits muskrat trapping to removal of only the 
animals surplus to the needs of a balanced marsh habitat, with the 
need for swan nesting sites being kept particularly in mind. 

In order to supplement the few muskrat houses which were avail- 
able as swan nesting sites immediately after the establishment of the 
Refuge, a number of floating artificial nesting platforms were con- 
structed of wood on the ice, loaded with marsh hay to simulate musk- 
rat lodges, and anchored over shallow marsh waters. The swans 
accepted these devices as nesting sites, but their construction was dis- 
continued when the muskrat population again became sufficient to 
ensure a plentiful supply of natural sites. 

Coyotes were very abundant on Refuge ranges for over a decade 
following establishment of this wildlife area. Since they were then 
believed to be one of the few natural limiting factors of the swan 
population, control by hunting, trapping, and poisoning was actively- 
employed. In the fall of 1947, coyotes were brought under control 
on ranges surrounding the Refuge by the Branch of Predator and 
Rodent Control. Since their methods have proved very effective in 
suppressing the local coyote populations, no predator control measures 
have had to be carried out within the Refuge during the past 10 years. 
Subsequently, studies and observations have cast considerable doubt 
on the importance of the coyote as a factor limiting the trumpeter 
population. 

Winter feeding of small grain to trumpeters was begun on the 
Refuge during the 1936-37 season and has been carried out every 
year since that time. This program began with a seasonal consump- 
tion of about 100 bushels by the swans and other waterfowl which 
frequented Culver's Pond, an artificial impoundment located at the 
east end of the Refuge and known colloquially as the Widow's Pool. 
Here 41° F. water gushing from the dual Picnic Springs keeps a 
few acres of shallow water open during even the coldest winter 
weather. The grain which is placed out for the swans in this shallow 
pond serves as a valuable supplement to natural foods, which are cus- 
tomarily greatly reduced by the swans and the seasonal waterfowl 
migrants even before winter sets in. 

During recent years, with the increasing swan population, this 
amount has been increased to about 750 bushels. This is fed each 



178 MANAGEMENT 

season on a semiweekly schedule, with wheat and barley proving 
equally satisfactory. While seasonal feeding of wildlife is a ques- 
tionable management practice under ordinary circumstances, it is 
felt that in this case it is necessary to carry the increasing numbers 
of the rare trumpeters safely through their most difficult season when 
natural food supplies are normally either greatly depleted or entirely 
exhausted. Otherwise, especially during severe winters, starvation 
would probably occur to some extent or the movement of swans to 
other areas outside Federal sanctuaries would result. 

Present plans call for expanding the feeding program to include 
the recently created MacDonald Pond. While less water acreage 
is impounded here, the originating Elk Spring water is warmer 
(59° F.) and creates a larger open- water area during the winter. 
With the winter feeding areas at Culver and MacDonald Ponds in 
operation it is hoped that more swans can be held on the Refuge 
during the winter months, and thus reduce to some extent their 
dependence on waters outside federally protected areas during their 
critical season. 

TRANSPLANTING PROGRAM 

In view of the progress made in the swan restoration program by 
1938 it was decided that the breeding range of the trumpeter should be 
extended. Accordingly, 4 cygnets were transferred from the Red 
Rock Lakes Refuge to the National Elk Refuge near Jackson, 
Wyoming. The following year this transplanting program was en- 
larged to include the vast marshes of the Malheur National Wildlife 
Refuge in southeastern Oregon. A decade later, in 1949, the high and 
isolated mountain valley marshes of the Ruby Lake National Wildlife 
Refuge in northeastern Nevada were included in the plan. 

Table 14 documents the transfers of trumpeters from the Red Rock 
Lakes Refuge to the various areas selected for their introduction from 
1938 through 1957. Twelve of the cygnets which were originally sent 
from the Red Rock Lakes Refuge to Malheur were retransferred to 
Ruby Lake in 1947. Thus the Ruby Lake Refuge has actually received 
a total of 12 more swans, and Malheur 12 less, than table 14 indicates. 

Of the many introductions which were thus made over a period of 
20 years, only the initial group transferred to the National Elk Ref- 
uge has bred successfully to date except for the hypothetical record 
at Ruby Lake cited previously. 1 Aimer P. Nelson, formerly Refuge 
Manager at the Elk Refuge, sums up the history of these introductions 
on that area as follows (correspondence) : 



1 In 1A58, a pair of trumpeters at Malheur and another at Ruby Lake Refuse nested 
successfully, bringing off 2 and 6 cygnets respectively. 









Figure 52. — The U. S. Fish and Wildlife Service "snowplane" en route to the 
Culver Pond swan wintering grounds for semiweekly feeding of small grain. 
The Centennial Mountains in background form the Continental Divide along 
their 10.000-foot crest, 



On October 24, 1938, four cygnets were transferred and liberated on this 
refnge. Following liberation, three of the birds were always seen together while 
the fourth did not associate with the other three, and in early December the 
lone cygnet disappeared from the area and was not seen again, while the other 
three remained. 

On October 1, 1939, three additional cygnets were transferred in. These re- 
mained until the Flat Creek and marsh lands began to freeze in late November 
or early December when they also left the area and were not seen again. 

On September 23, 1941, three more cygnets were transferred to the Elk 
Refuge. When these three birds were liberated, we clipped their wing tips 
and they were seen frequently on the area until late spring, when they too 
came up missing. 

From then until 1944, the only swans seen on the area were the three adults 
that were transferred in 1938. In the case of these three they frequently left 
the refuge during the winter months and were known to be away for as long 
as two months. During their absence, they were at various times reported 
along the Snake River within the valley [Jackson Hole]. The three birds always 
returned to the refuge when the ice in the creek channel began to clear every 
spring and remained throughout the summer season. When these birds were 
2 or 3 years old, two were frequently seen together while the third remained 
to itself. In 1944 the pair nested in the Flat Creek marsh. 



180 



MANAGEMENT 



Since 1944, nesting has occurred nearly every year on the Elk Ref- 
uge, with two nests established during the seasons of 1948, 1954, and 
1955. The history of the Elk Refuge breedings follows in table 15. 

Since clutch size, hatchability, and early cygnet mortality of the 
Elk Refuge breeders are largely unknown, an accurate comparison 
of these factors with similar data from the Red Rock Lakes area is 
not possible. What little data is available indicates that losses in both 



Table 14.— Trumpeter swans transferred from the Red Rock Lakes Refuse 

1938 to 1957 



Recipient area 


Number 




Cygnets 


Adults 


Elk Refuge: 

Oct. 24, 1938 


4 
3 
3 




Oct. 1, 1939 




Sept. 23, 1941 








Total 


10 








Malheur Refuge: 

Oct. 16, 1939 


3 
6 
20 
22 
19 
7 

17 




Sept. 19, 1941 


1 


Sept. 4, 1944 


Sept. 4, 1945. 




Sept. 12, 1948... 




Julv 14, 1954 


6 (3 pairs). 

1 (nonbreeder). 


Sept. 27, 1955 


Aug. 11, 1956 


11 (nonbreeders). 


Sept. 25, 1957. 


20 






Total 


114 


19 




Rubv Lake: 

Sept. 12, 1949 


10 

1 

16 

12 




Julv 28, 1954 


6 (3 pairs). 


Sept. 26, 1955 " """ 


Oct. 10, 1956 




Aug. 2, 1957 


19 (nonbreeders). 






Total 


39 


25 




Delta Station: 

Aug. 7, 1956 


3 


3 (nonbreeders). 




Total transferred 


166 









Table 15.— Trumpeter swan nesting data, National Elk Refuge, Wyoming, 

1944 to 1957 



Year 



1944 
1945 
1946 
1947 
1948 
1949 
1950 
1951 



Pairs 
nested 



Cygnets 
initially 
observed 



Cygnets 
raised to 
flying age 



1 
3 


1,3 

3 




Year 



1952 

1953 

1954 

1955 

1956 

1957 

Total 



Pairs 


Cygnets 


nested 


initially 




observed 


1 


3 


1 


o 


2 


2,2 


2 


3.3 


(') 





1 


1 


14 


34 



Cygnets 
raised to 
flying age 



3 

2 

2,1 

3,3 






1 No nests observed. 



SWAN MANAGEMENT ON RED ROCK LAKES REFUGE 181 

the egg and preflight stages of existence are as serious as those observed 
in the Red Rock Lakes population. 

The Malheur Refuge has received a total of 133 trumpeters 
transferred from the Red Rock Lakes Refuge during the period 
1939-57. A variety of methods have been tried at Malheur during 
this period to establish a wild breeding population of these birds, 
but this goal has not yet been achieved. The chief factors contributing 
to failure in the early transplanting efforts were the practices of 
pinioning and confining the flock to a single large pool, where intra- 
specific strife and spatial competition not only created a situation 
which was unfavorable for breeding but led to significant losses from 
accidents and disease. Initial efforts to establish a breeding swan 
population by liberating the transplanted individuals directly into 
the marsh proper failed due to the dispersion and disappearance of the 
liberated birds. 

As far as the Ruby Lake Refuge effort is concerned, with the 
exception of the hypothetical single breeding record noted near 
the Ruby Lake Refuge in 1953, the efforts to establish a wild breeding 
population of trumpeters at that location have been unsuccessful for 
many of the same reasons affecting the Malheur transplants. 

In addition to the transplants involving only birds-of-the-year, a 
single attempt to achieve the desired goal at Ruby Lake and Malheur 
by transferring known breeding pairs from the Red Rock Lakes Ref- 
uge was attempted in 1954 when 3 pairs of mated trumpeters and 
their offspring were liberated in each marsh area. The adults ap- 
parently remained with their broods at their new locations for a time 
after regaining their flight feathers, but disappeared later. Perhaps 
these adults departed with their cygnets during the fall migration 
along with the whistling swans which pass through these areas dur- 
ing the autumn in considerable numbers. At any rate their fate re- 
mains unknown. 

In 1957 a new transplanting program involving both the Ruby 
Lake and Malheur Refuges was begun. Specifically designed to avoid 
many of the shortcomings of the earlier attempts, the revised pro- 
cedure employs a relatively short-term swan decoy flock, thereby 
n inimizing the bad effects of the long periods of confinement for the 
entire group. This will be employed in connection with banding and 
dyeing techniques used on the liberated birds in order to trace dis- 
persion. 

In 1955, 6 trumpeters were transferred from the Red Rock Lakes 
Refuge to the Delta Waterfowl Research Station, Delta, Manitoba, 
Canada. These birds, 3 adults and 3 cygnets, supplemented the small 
group of Canadian trumpeters already at Delta and were transferred 



182 



MANAGEMENT 



Table 16.— Swans banded at the Red Rock Lakes Refuge, 1945 to 1957 



Year 


Number 


Age class 


Location released 


194.5 


20 

5 
10 
10 
10 

3 
21 

9 
17 

9 
17 
16 

9 
10 

5 
89 
10 

1 
12 
19 
20 

2 
45 




Red Rock Lakes Refuge. 
Red Rock Lakes Refuge. 
Red Rock Lakes Refuge. 
Ruby Lake Refuge. 
Red Rock Lakes Refuge. 
Red Rock Lakes Refuge. 
Red Rock Lakes Refuge. 
Red Rock Lakes Refuge. 
Red Rock Lakes Refuge. 
Red Rock Lakes Refuge. 
Malheur Refuge. 
Ruby Lake Refuge. 
Red Rock Lakes Refuge. 
Red Rock Lakes Refuge. 
Red Rock Lakes Refuge. 
Red Rock Lakes Refuge. 
Malheur Refuge. 


1948 




1949 






Cygnets 




Non-breeders 1 


1950 






Cygnets 


1951 




1952 




1953 




1955 






Cygnets ... 




Adults. 


1956 






Immatures 




Adults 




Adults.. .. 




Immature . 




Cygnets . 


Ruby Lake Refuge. 
Ruby Lake Refuge. 
Malheur Refuge. 
National Zoological Park. 
Red Rock Lakes Refuge. 


1957 






Cygnets 




Non-breeders • 




Non-breeders ' 






Total 


369 











1 Includes both adult and immature age-classes. 

with the hope that breeding pairs could be established there, with 
methods which could eventually be applied to pertinent United States 
refuge areas. To date the transferred adults have not bred at Delta, 
though pairing has taken place. The cygnets transferred there are, 
of course, not yet sexually mature. 

In 1957 a male and a female trumpeter from the Refuge's non- 
breeding flock were transferred to the National Zoological Park in 
Washington, D. C, for exhibition and research purposes. 

BANDING 

From 1945 to 1957, 261 swans of various age classes have been 
banded and released on the Red Rock Lakes Refuge alone. This was 
done to provide facts relating to local movement or longer migration, 
causes of mortality, longevity, etc. In addition, 107 swans banded at 
Red Rock Lakes have been released elsewhere, 48 on the Malheur and 
57 on the Ruby Lake Refuges and the 2 transferred to the National 
Zoological Park. This information is shown in table 16. 

A total of 369 swans have thus been banded and released. Only 16 
recoveries have been reported as of February 14, 1957, 13 of these 
within the general Red Rock Lakes region, 2 on the Malheur Refuge, 
and 1 in the Ruby Lake marsh. There are no recoveries to reveal 
movement or migration outside their traditional known range or, in 
the case of the transferred individuals, outside the general area in 
which they were introduced. 



MANAGEMENT RECOMMENDATIONS 183 

A breakdown of the causes of death among the 16 recoveries shows 
that shooting was responsible in 5 instances. In the remaining 11 
cases, information is not available, though it. is probable that illegal 
shooting played a significant role here also. 

A period of less than 2 years existed between the date of banding 
and that of band recovery in 10 of the 11 cases where pertinent re- 
covery information was available. In the remaining example, a dura- 
tion of slightly over 3 years occurred between banding and recovery 
dates. 

In swan-banding operations on the Refuge, 27 returns of banded 
birds have been noted. The earliest return occurred when a fall- 
banded cygnet was captured again for banding in its initial post- 
juvenile molt the following summer, a period of 11 months. The 
maximum return period thus far was that of a male banded as a 
cygnet in the late summer of 1949 and recaptured in July 1957, a 
period of 7 years 11 months. 

The standard U. S. F. W. S. size 9 aluminum bird band, the largest 
made, was first used in all Refuge trumpeter-swan banding opera- 
tions, until it was discovered that this size was somewhat too small, 
and also that a significant proportion of banded captive trumpeters 
at the Malheur Refuge lost these aluminum bracelets from year to 
year. To avoid this difficulty a locking-type stainless-steel band was 
developed and placed in use in 1955. This steel lock-type band, 
carrying hand-stamped identification data, has been used exclusively 
since that time, except for 12 cygnets transferred to Ruby Lake in 
1956 which were banded with standard aluminum bands when the 
supply of steel bands was temporarily exhausted. 

A trumpeter-swan banding project has also been carried out by 
the Canadian Wildlife Service for a number of years, using the 
standard aluminum bands as well as colored plastic bands for sight 
record purposes. Except for those specific cases previously men- 
tioned, the results of this work are not yet available. 

MANAGEMENT RECOMMENDATIONS 

As has been shown, the trumpeter-swan population in the United 
States has increased manyfold during the past 30 years and is ap- 
parently now being maintained near the highest level which its year- 
round environment will support, For all practical considerations it 
has been saved from any immediate threat of extinction in this 
country; hence the prime goal is to preserve the existing habitat 
necessary to hold the substantial population gains made in recent years 
and to increase their range by transplanting to new localities. 

469660 O— 60 13 



184 



MANAGEMENT 







Sfw 



| » m 



•"p-^^wk 



.^ ' * ¥%m ( -fc3M ^spu 




Figure 53. — These trumpeters are only part of the flock of over 200 which pass 
the late winter months on the Refuge awaiting the spring breakup. The 94 
trumpeters visible in this single photograph are more than existed in the 
entire United States 25 years ago. 

Since by far the greatest proportion of breeding 1 pairs is found 
within either the Red Rock Lakes Refuge or the National Parks of 
the area (Yellowstone and Grand Teton), the maintenance of the 
quality of the habitat should not require attention beyond that now 
being given. Little need is seen at present to develop the main swan 
breeding habitat already included within these Federal sanctuaries. 
Emphasis should rather be laid on preserving and maintaining a 
seasonally balanced habitat for all age classes. 

The Red Rock Lakes Refuge marshes exist in practically a wilder- 
ness state, and as such, may be regarded as almost ideal breeding 
habitat. While a number of artificial nesting islands might be con- 
structed on the open expanse of Upper Red Rock Lake, their occupa- 
tion and defense by territorial nesting pairs would necessarily limit 
the use of these waters by large numbers of nonbreeders. Saving a 



MANAGEMENT RECOMMENDATIONS 



185 








part of the limited federally protected habitat for the flocked, non- 
breeding segment of the population is important since the only other 
suitable waters are artificial impoundments subject to drawdown with 
the usual unfavorable results. Moreover, the mere accommodation of 
a few additional pairs of breeders on the Refuge probably would not 
increase the cygnet production if the experience to date has provided 
a true index. 

If not filled in by sediments brought down by streams as a result 
of poor watershed practices, the Red Rock Lakes marsh system 
should continue to exist in virtually its present form for many years. 
Even if the threat of siltation becomes acute, the habitat could be saved 
by constructing a series of artificial pools within the Refuge along the 
main tributary streams. These could serve as silt-traps, and thus 
prevent the excessive deposition of alluvium in the already shallow 
lake beds. 

Winter feeding of small grain on the Red Rock Lakes Refuge should 
be continued to the fullest extent practicable. In addition, the possi- 



186 MANAGEMENT 

bilities of initiating winter feeding wherever concentrations of the 
trumpeters exist during this season should be studied. Since arti- 
ficial feeding during the winter apparently meets only a portion of the 
trumpeters' dietary requirements, ways of increasing the production 
or availability of the natural aquatic plants should be explored. This 
is especially applicable to the Island Park area in Fremont County, 
Idaho. In this region a comparatively few mile& of the streambed 
along Henrys Fork of the Snake River and its tributaries form the 
most important single wintering grounds for this species on the 
continent. Winter use is especially concentrated on and in the 
vicinity of the Railroad Ranch. 

Acquisition and management of an adequate portion of the swan 
wintering habitat along Henrys Fork and its tributaries would prevent 
development and hunting pressure which might adversely affect 
the trumpeters during their season of greatest vulnerability. Whether 
or not this is accomplished, special wildlife easements or regulations 
affecting these lands would assure additional protection and food for 
this rare species. A state wildlife sanctuary already exists on a por- 
tion of the Railroad Ranch wintering grounds. Specific recom- 
mendations should result from additional study of this problem in 
the field. 

Investigations into various phases of the life history of the trum- 
peter should be continued both on the Red Rock Lakes Refuge and in 
Yellowstone Park, in fact wherever the time of qualified personnel 
and opportunity exist. Priority in these studies should be placed on 
the following subjects: banding for the purpose of revealing 
trumpeter distribution and mortality; causes of the low hatehability 
of trumpeter eggs ; factors contributing to the high mortality of cyg- 
nets before flight age is reached ; and relation between the density of 
trumpeters on the breeding grounds and production. Investigations of 
specific characteristics of trumpeters held in captivity might also be 
profitably given attention : territorialism, space requirements, breed- 
ing, and dietary considerations. 

Outside their present Rocky Mountain environment, the emphasis 
should be placed on transplanting. In addition to introductions now 
being carried out at the Ruby Lake and Malheur Refuges, further 
efforts to establish a wild population should be extended to suitable 
areas within the former known breeding range of this species. 
Habitat in the Flathead Valley of Montana, the Minnesota-Iowa 
region, the Dakotas, and northwestern Nebraska may have areas suit- 
able for future plantings. 

Owing to the vulnerability of trumpeters to shooting, possibly the 
most important single requirement in transplanting, outside of favor- 




yp 



. 



*Hr>s 



Figure 54. — Trumpeters feed in Culver Pond with mallards, Barrows goldeneyes 
and common goldeneyes on grain placed out by Refuge personnel. 

able habitat factors, is provision for sufficiently protected wintering 
waters, preferably in close proximity to managed breeding grounds. 
It is difficult to see how a wild population of these slow-maturing, 
conspicuous, low-flying birds could ever become established and thrive 
if they were compelled by the freezing of their breeding grounds to 
run the waterfowl hunter gauntlet each fall in moving to far distant 
wintering grounds. 

The problems of successfully transplanting the trumpeter do not 
appear to be beyond solution if similar experience with the mute 
swan is comparable. In addition to the success of this exotic species 
in acclimating itself in both England and Denmark as a feral resident, 
during the past half century the mute swan has become firmly estab- 
lished in the thickly populated lower Hudson River Valley and 
vicinity. The mute has become established in Michigan, with a popu- 
lation wintering in Grand Traverse Bay and reportedly breeding on 
waters near East Jordan (correspondence). The numbers of these 
birds have increased steadily, about as follows : 1948 or 1949, 2 ; 1950, 
5; 1951, 8; 1952, 11; 1953, 13; 1954, 17; 1955, 24; 1956, 41. 



188 MANAGEMENT 

While the trumpeter usually lays a smaller clutch of eggs than the 
mute swan, and may be more vulnerable to shooting as well, there is 
no obvious reason why it should not respond to a sound program of 
transplanting and management if hunting losses can be eliminated or 
reduced to a low level. 

As the trumpeter population becomes more widely distributed, the 
possibility of hybridization with the mute swan may become a threat. 
With feral mute swans spreading to new areas, and with the tendency 
for swan hybrids to be fertile, some interbreeding between the two 
species is possible. This could be serious when the small total popula- 
tion of trumpeters is considered. Such a possibility must be guarded 
against, and controls be undertaken when and where the swan breeding 
populations begin to overlap. 

Because of its rarity, beauty, and other intrinsic qualities, the 
trumpeter is greatly sought for display and breeding purposes, both 
by public and private institutions and individuals. A few are being 
kept in captivity for display and observation purposes, though at 
their present level of abundance none can be provided to aviculturists 
specifically for breeding purposes, nor would their release to breeders 
generally be in keeping with the responsibilities of the U. S. Fish and 
Wildlife Service as set forth in domestic regulations and international 
migratory bird treaty agreements. It is the objective of the Bureau 
to maintain the wild population of these rare fowl at an optimum 
level — the greatest number which can be consistently supported in their 
natural environment. Any swans which may be surplus to this re- 
sponsibility should be included in the program to establish other 
wild breeding flocks, or loaned to qualified public zoological parks for 
display and breeding purposes. With continued diligent manage- 
ment and protection, the transplanting program should assure the 
continued growth in numbers and expansion of the range of the 
trumpeter swan in this country. 




BIBLIOGRAPHY 



Adams, Edward. 

1878. Notes on the birds of Michalaski, Norton Sound (Alaska). Ibis, vol. 
2, No. 8, Fourth Series. 
Allen, J. A. 

1878. A list of the birds of Massachusetts, with annotations. Bulletin of 
the Essex Institute, vol. 10. Salem, Mass. 
American Ornithologists' Union. 

1957. Checklist of North American birds. Fifth Edition. Prepared by a 
committee of the A. O. U. 691 pp. 
Anderson, R. M. 

1907. The birds of Iowa. Proceedings of the Davenport Academy of 
Science, Davenport, Iowa. 
Armstrong, E. A. 

1947. Bird display and behavior. Oxford University Press, New York. 
431 pp. 
Audubon, John J. 

1838. Ornithological biography, vol. 4. Adam and Black, Edinburgh. 618 
pp. 
Bailey, Florence Merriam. 

1918. Wild animals of Glacier National Park ; The birds. National Park 
Service, Washington. 
Bailey, Vernon. 

1930. Animal life of Yellowstone National Park. C. C. Thomas, Spring- 
field, Illinois. 241 pp. 
Baird, Spencer F., Thomas M. Brewer, and Robert Ridgway. 

1884. The water birds of North America. Memoirs of the Museum of 
Comparative Zoology at Harvard College, vol. 12. Little, Brown and 
Co., Boston. 537 pp. 
Baker, Frank C. 

1941. A study in ethnozoology of the prehistoric Indians of Illinois. Trans- 
actions American Philosophical Society, vol. 32, part 2. 
Barnston, George. 

1862. Recollections of the swans and geese of Hudson's Bay. Zoologist, 
vol. 20. 
Bates, J. M. 

1900. Additional notes and observations on the birds of northern Nebraska. 
Proceedings of the Nebraska Ornithologists' Union, First Annual Meeting 
1899. (Published 1900.) 
Beal, S. M. 

1949. Four cygnets disappear. Yellowstone Nature Notes, vol. 23, No. 1. 
Beard, Daniel B., et al. 

1947. Fading trails, the story of endangered American wildlife. Macmillan. 
279 pp. 
Beebe, C. William. 

1906. The swans. Report of the New York Zoological Society, No. 10. 
Belknap, Jeremy. 

1784. History of New Hampshire. 3 vols. Philadelphia ; also 1791-92, 
Boston. 

189 



190 BIBLIOGRAPHY 

Bent, Arthur Cleveland. 

1925. Life histories of North American wild fowl (Order Anseres). U. S. 
National Museum Bulletin 130, vol. 2. 316 pp. Reprinted by Dover Pub- 
lications, 1951. 

Blaauw, F. E. 

1904. On the breeding of some of the waterfowl at Gooilust in the year 1903. 
Ibis, vol. 4, No. 13. Eighth Series. 
Blines, Jasper. 

1888. Letter to editor. Forest and Stream, vol. 31, No. 22. 
Brooks, Allan. 

1926. The present status of the trumpeter swan. Condor, vol. 28, No. 3. 
Brooks, Allan, and H. S. Swarth. 

1925. A distributional list of the birds of British Columbia. Contribution 
No. 423, Museum of Vertebrate Zoology, University of California. Cooper 
Ornithological Club, Pacific Coast Avifauna No. 17. 
Brower, J. V. 

1897. The Missouri River- and its utmost source. Pioneer Press, St. Paul, 
Minnesota. 
Burnett, W. L. 

1916. Two trumpeter swan records for Colorado. Auk, vol. 33, No. 2. 
Butler, A. W. 

1897. The birds of Indiana. 22d annual report, Indiana Department of 
Geology and Natural Resources. (W. S. Blatchley.) 
Chaddock, T. T. 

1938. Laboratory report on whistler swan. Wisconsin Conservation Bulle- 
tin, vol. 3, No. 9. 
Christian, J. J. 

1957. A review of the endocrine responses in rats and mice to increasing 
population size including delayed effects on offspring. Naval Medical 
Research Institute Lecture and Review Series No. 57-2. 
Coale, H. K. 

1915. The present status of the trumpeter swan (Olor buccinator). Auk, 
vol. 32, No. 1. 
Oolwell, Gwen. 

1948. Trumpeters pay royal visit. Vancouver Province (magazine section) 
July 24, 1948. Vancouver, British Columbia. 
Condon, David deLancey. 

1950. An uncanny record in the snow. Yellowstone Nature Notes, vol. 24. 
No. 1. 
Cooke, Wells W. 

1887. Bird migration in the Mississippi Valley in 1884 and 1885. Division 
of Economic Ornithology and Mammalogy, U. S. Department of Agri- 
culture, Bulletin No. 2. 
Cooper, J. G. 

1869. The fauna of Montana Territory. The American Naturalist, vol. 3. 
Coues, Elliott. 

1874. Birds of the northwest. Department of the Interior, U. S. Geological 

Survey of the Territories, Miscellaneous Publication 3. 791 pp. 
1893. The history of the expedition under the command of Lewis and 
Clark. Francis P. Harper. 4 vols. 



BIBLIOGRAPHY 191 

Cowan, Ian McTaggert. 

1946. Death of a trumpeter swan from multiple parasitism. Auk, vol. 63, 
No. 2. 
pall, W. H., and H. M. Bannister. 

1869. List of the birds of Alaska, with biographical notes. Transactions 
of the Chicago Academy of Science, art. 9. 
Day, Albert M. 

1949. North American waterfowl. Stackpole and Heck, Inc. 329 pp. 
Delacour, Jean. 

1944. Government refuges are saving the trumpeter swan. Animal King- 
dom, bulletin New York Zoological Society, vol. 47, No. 6. 

1954. The waterfowl of the world, vol. I. Country Life Limited, London, 
England. 284 pp. 
Delacour, Jean, and Ernst Mayr. 

1945. The family Anatidae. Wilson Bulletin, vol. 57, No. 1. 
Dementiev, Georges P., and X. A. Gladkov. 

1952. Die vogel der Sowjetunion, vol. 4. Moscow. 
Derscheid, J. M. 

1939. The preservation of waterfowl and aviculture. Avicultural Maga- 
zine, vol. 4, No. 3. 
Dewar, J. M. 

1936. Menage a trois in the mute swan. British Birds, vol. 30, No. 6. 
Dixon, Joseph. 

1931. Save the trumpeter swan. American Forests, vol. 37, No. 8. 
Eklund, Carl R. 

1946. Mortality notes on the trumpeter swan. Auk, vol. 63, No. 1. 
Ellis, John. 

1936. Menage a trois in the mute swan (letter to the editors). British 
Birds, vol. 30, No. 7. 
Errington, Paul L. 

1945. Some contributions of a fifteen-year local study of the northern 
bobwhite to a knowledge of population phenomena. Ecological Mono- 
graphs, vol. 15, No. 1. 

Evans, A. H. 

1903. Turner on birds. Cambridge University Press. 
Featherstonhaugh, Duane. 

1948. Return of the trumpeter. Natural History, vol. 57, No. 8. 
Fisher, James, and Peter Scott. 

1953. A thousand geese. London (Houghton Mifflin, Boston, 1954:) 
Forbush, Edward Howe. 

1912. A history of the game birds, wild fowl, and shore birds of Massa- 
chusetts and adjacent States. Massachusetts State Board of Agriculture. 
622 pp. 
1929. Birds of Massachusetts and other New England States. Part 1. 
Water birds, marsh birds and shore birds. Massachusetts Department of 
Agriculture. 3 vols. 
Force's Historical Tract, vol. 2, tract 5. 
Friedmann, Herbert. 

1935. Birds of Kodiak Island. Bulletin Chicago Academy of Science, 
vol. 5, No. 3. 
Gabrielson, Ira N. 

1946. Trumpeter swans in Alaska. Auk, vol. 63, No. 1. 



192 BIBLIOGRAPHY 

Gabrielson, Ira N., and S. G. Jewett. 

1940. Birds of Oregon. Oregon State College. 650 pp. 
Grundtvig, F. L. 

1895. On the birds of Shiocton in Bovina, Outagamie County, Wisconsin, 
1881-83. Transactions of the Wisconsin Academy of Science, Arts, and 
Letters, vol. 9. 
Haecker, F. W., R. A. Moser, and J. B. Swenk. 

1945. Check-list of the birds of Nebraska. Nebraska Ornithologists' Union, 
Nebraska Bird Review, vol. 13, No. 1. 
Hart, R. O. 

1952. Trumpeter swan versus muskrat at Oxbow Lake. Yellowstone Nature 
Notes, vol. 26, No. 5. 
Hearne, Samuel. 

1795. A journey from Prince of Wales Fort in Hudson's Bay to the North- 
ern Ocean, London. (Reprinted in The Publications of the Champlain 
Society, Toronto, Canada, 1911.) 
Heerman, A. L. 

1859. Report upon birds collected on the survey. Pacific Railroad Survey 
Report, vol. 10. U. S. House of Representatives, 33d Congress, 2d Session, 
Ex. Doc. 91. 
Heinroth, Oskar, and Magdalena Heinroth. 

1911. Beitrage zur Biologie, namenthlich Ethologie und Psychologie der 
Anatiden. Proceedings of the 5th International Ornithological Congress, 
Berlin. 
1928. Die Vogel Mitteleuropas, vol. 3. Hugo Bermuhler, Berlin- 
Lichterfelde. 
Hellmayr, Charles E., and Boardman Conover. 

1948. Catalogue of birds of the American Zoological Series, Field Museum 
of Natural History, vol. 13, part I, No. 2, Publication 615. 434 pp. 
Hilden, O., and P. Linkola. 

1955. Suuri lintukirja (The big bird book), Helsinki. 750 pp. 
Hilprecht, Alfred. 

1956. Hockerschwan, Singschwan, Zwergschwan. A. Ziemsen Verlag, 
Wittenberg Lutherstadt. 151 pp. 

Hochbaum, H. Albert. 

1944. The canvasback on a prairie marsh. American Wildlife Institute. 

201 pp. 
1955. Travels and traditions of waterfowl. University of Minnesota. 
301 pp. 
Holman, John P. 

1933. Sheep and bear trails. Frank Walters. 211 p. 

1950. Concentration of trumpeter swans, Cygnus buccinator, in British 
Columbia in winter. Auk, vol. 67, No. 3. 
Hornaday, William T. 

1913. Our vanishing wildlife, its extermination and preservation. New 
York Zoological Society. 411 pp. 
Howard, L. O., and W. F. Fiske. 

1911. The importation into the United States of the parasites of the gypsy 
moth and the brown-tail moth. U. S. Department of Agriculture En- 
tomology Bulletin, No. 91. 
Hughes, J. C. 

1883. Letter to the editor, Forest and Stream, vol. 20, May 10. 



BIBLIOGRAPHY 193 

Hull, A. V. 

1939. Trumpeter swans, their management and preservation. Transac- 
tions of the Fourth North American Wildlife Conference, American Wild- 
life Institute. 
James, Edwin. 

1823. Account of an expedition from Pittsburgh to the Rocky Mountains, 
vol. 1. Philadelphia. 
Jewett, S. G., W. P. Taylor, W. T. Shaw, and John W. Aldrich. 

1953. Birds of Washington State. University of Washington Press. 
767 pp. 

Keating. William H. 

1825. Narrative of an expedition to the sources of the St. Peter's River, 
Lake Winnepeek, Lake of the Woods, &c, performed in the year 1823, 
vol. 1. Geo. B. Whittaker, London. 
Kluijver, H. N. 

1951. The population ecology of the great tit, Parus m. major L. Ardea, 
vol. 39, Nos. 1-3. 
Knight, W. C. 

1902. The birds of Wyoming. Bulletin No. 55, Wyoming Experiment Sta- 
tion, University of Wyoming. 

Kortright, F. H. 

1943. The ducks, geese, and swans of North America. Wildlife Manage- 
ment Institute. 476 pp. 
Kumlien, Ludwig, and N. Hollister. 

1903. The birds of Wisconsin. Bulletin of the Wisconsin Natural History 
Society, Nos. 1, 2, and 3. 

Lack, David. 

1954. The natural regulation of animal numbers. Oxford University Press. 
343 pp. 

Lattin, Frank H. 

1892. The Standard Catalogue of North American Birds Eggs. Third 
edition. Albion, New York. 53 pp. 
Lawson, John. 

1714. History of North Carolina. Reprinted by Observer Printing House, 
Charlotte, N. C, 1903. 171 pp. 
Lister, R. 

1951. Trumpeter swans breeding in the Cypress Hills of Alberta. Canadian 
Field Naturalist, vol. 65, No. 4. 
Lorenz, Konrad. 

1937. The companion in the bird's world. Auk, vol. 54, No. 3. 
Low, G. C. 

1935. (Edited paper by Marquess of Tavistock) The extent to which cap- 
tivity modifies the habits of birds. Bulletin of the British Ornithologists 
Club, vol. 55, session 1934-35. 
MacFarlane, Roderick. 

1891. Notes on and list of birds and eggs collected in arctic America, 1861- 

66. Proceedings of the U. S. National Museum, vol. 14. 
1905. Notes on mammals collected and observed in the northern Mackenzie 
River district, northwest territories of Canada, with remarks on ex- 
plorers and explorations of the far north. Proceedings of the U. S. 
National Museum, vol. 28. 



194 BIBLIOGRAPHY 

Mackay, R. H. 

1954. Trapping of the Queen's trumpeter swans in British Columbia. Sixth 

Annual Report of the Wildfowl Trust, 1952-53. London. 
1957. Movements of trumpeter swans shown by band returns and observa- 
tions. Condor, vol. 59, No. 5. 
Madsen, H. 

1945. On the different position of the legs of birds during flight and in cold 
weather. Dansk Ornithologisk Forenings Tidsskrift, vol. 39. 
Mair, Charles, and Roderick MaeFarlane. 

1908. Through the Mackenzie Basin (including notes on the mammals 
and birds of northern Canada ) . William Briggs, Toronto. 494 pp. 
McDermott, John F. 

1942. Audubon's journey up the Mississippi. Journal of the Illinois State 
Historical Society, vol. 35, No. 2. 
McLean, Donald D. 

1937. Some additional records of birds for northeastern California. Condor, 
vol. 39, No. 5. 
Merriam, C. Hart. 

1877. A review of the birds of Connecticut, with remarks on their habits. 
Transactions of the Connecticut Academy of Arts and Sciences, vol. 4, 
part 1. 
1891. Birds of Idaho. North America Fauna No. 5, U. S. Biological Survey. 
Merrill, D. E. 

1932. Trumpeter swan in New Mexico. Auk, vol. 54, No. 4. 
Monson, Melvin A. 

1956. Nesting of trumpeter swan in the lower Copper River Basin, Alaska. 
Condor, vol. 58, No. 6. 
Mosher, A. D. 

1882. The fauna of Spirit Lake. Forest and Stream, vol. 18, February 23. 
Munro, J. A. 

1949. Conservation of the trumpeter swan in Canada. Proceedings of the 
7th Pacific Science Congress, vol. 4. 
Munro, J. A., and I. McT. Cowan. 

1947. A review of the bird fauna of British Columbia. British Columbia 
Provincial Museum, Department of Education. 285 pp. 
Murie, Adolph. 

1940. Ecology of the coyote in the Yellowstone. Fauna of the National 
Parks of the United States, No. 4. National Park Service, U. S. Depart- 
ment of the Interior. 

Nelson, E. W. 

1884. The arrow-head in the swan. Forest and Stream, vol. 22, May 8. 
1887. Report upon natural history collections made in Alaska between 
the years 1877 and 1881. Arctic series of publications issued in connec- 
tion with the Signal Service, U. S. Army. 
Newberry, J. S. 

1857. Report on the zoology of the route. Pacific Railroad Survey Report 
1853-56, vol. 6. IT. S. House of Representatives, 33d Cong., 2d Sess., Ex. 
Doc. 91. 
Nice, Margaret Morse. 

1941. The role of territory in bird life. American Midland Naturalist, vol. 
26, No. 3. 



BIBLIOGRAPHY 195 

Nicholson, A. J. 

1933. The balance of animal populations. Journal of Animal Ecology, 

Vol. 2, No. 1. 
Nute, Grace Lee. 

1945. Calendar of the American Fur Company's papers. Annual Report of 
the American Historical Association for the Year 1944. 3 vols. 
Nuttall, Thomas. 

1834. A manual of the ornithology of the United States and Canada. 
Boston. 683 pp. 
Orton, Alda. 

1951. Long necks. The Alaska Sportsman, vol. 17, No. 9. 
Over, William H., and Craig S. Thorns. 

1920. Birds of South Dakota. Bulletin of the University of South Da- 
kota, series 21, No. 9. 
Paludan, Knud, and J0rgen Fog. 

1956. The Danish breeding population of wild living Cygnus olor in 1954. 
Danske Vildtunders0gelser, vol. 5. 

Parmalee, Paul W. 

1958. Remains of rare and extinct birds from Illinois Indian sites. Auk, 
vol. 75, No. 2. 
Patrick, R. W. 

1935. Mute swans attacking bullock. British Birds, vol. 29, No. 4. 

Pilder, Hans. 

1914. Russisch-Amerikanische Handels-Companie bis 1825. Berlin. 174 
pp. (Arctic Bibliography No. 25091.) 
Poulsen, H. 

1949. Bidrag til Svanernes Ethologi. (Contribution to the Ethology of 
the Swans, English Summary.) Dansk Ornithologisk Forenings Tids- 
skrift, vol. 42, No. 4. 
Roberts, Thomas S. 

1936. The birds of Minnesota. University of Minnesota Press. 2 vols. 

Salter, R. J. 

1954. The trumpeter swan in Idaho. Proceedings Western Association Fish 
and Game Commissioners. 
Saunders, Aretas A. 

1921. A distributional list of the birds of Montana. Pacific Coast Avi- 
fauna No. 14, Cooper Ornithological Club. 
Scott, Peter, and James Fisher. 

1953. A thousand geese. London. (Houghton Mifflin, Boston, 1954.) 
Sharp, Ward M. 

1951. Observations on predator-prey relations between wild ducks, trum- 
peter swans, and golden eagles. Journal of Wildlife Management, vol. 
15, No. 2. 

Sibley, Charles G. 

1938. Hybrids of and with North American Anatidae. 9th International 
Ornithological Congress, Rouen. 
Silloway, P. M. 

1903. Birds of Fergus County, Montana. Bulletin No. 1, Fergus County 
Free High School, Lewistown. 77 pp. 
Simon, James R. 

1952. First flight of trumpeter swans. Auk, vol. 69, No. 4. 



196 BIBLIOGRAPHY 

Skinner, Milton P. 

1920. Trumpeter swan breeding in Yellowstone Park. Condor, vol. 22, 

No. 2. 
1925. The birds of Yellowstone National Park. Roosevelt Wildlife Bul- 
letin. New York State College of Forestry at Syracuse University. 
Vol. 3, No. 1. 
1928. Yellowstone's winter birds. Condor, vol. 30, No. 4. 
Smith, H. S. 

1935. The role of biotic factors in the determination of population densities. 
Journal of Economic Entomology, vol. 28, No. 6. 

Smith, Stuart G., and Eric Hosking. 

1955. Birds fighting. Experimental studies of the aggressive displays 
of some birds. Faber and Faber, London. 125 pp. 
Soper, J. D. 

1949. Birds observed in the Grande Prairie-Peace River Region of north- 
western Alberta, Canada. Auk, vol. 66, No. 3. 
Stejneger, Leonhard. 

1882. Outlines of a monograph of the Cygninae. Proceedings of the U. S. 
National Museum, vol. 5. 
Stewart, Robert E., and Joseph H. Manning. 

1958. Distribution and ecology of whistling swans in the Chesapeake Bay 
region. Auk, vol. 75, No. 2. 
Stowe, L. 

1957. The amazing Crusoes of Lonesome Lake. Reader's Digest, February. 
Suckley, George. 

1859. Water birds. Pacific Railroad Survey Report, 1853-55. Vol. 12. 
U. S. House of Representatives, 36th Congress, 1st Session, Ex. Doc. 56. 
Swainson, William, and John Richardson. 

1832. Fauna Boreali-Americana. Part 2, The Birds. London. 
Swann, H. Kirke. 

1913. A dictionary of English and folknames of British birds. Witherby 
and Co., London. 266 pp. 
Thompson, Ben H. 

1936. The problem of vanishing species — the trumpeter swan. Proceedings 
of the North American Wildlife Conference. Report of the Special Com- 
mittee on Conservation of Wildlife Resources, 74th Congress, 2d Session, 
G. P. O. 675 pp. 

Thwaites, R. G. 

1906. Early western travels (1748-1846). Arthur Clark Co., Cleveland. 

Vols. 21, 27, 29. (32 vols.) 
Ticehurst, N. F. 

1921. A contribution to swan history. British Birds, vol 14, No. 8. 
1924. The early history of the mute swan in England. British Birds. 

vol. 18, No. 8. 
1926. An historical review of the laws, orders and customs anciently used 

for the preservation of swans in England. British Birds, vol. 19, No. 8. 
1926. On swan marks. British Birds, vol. 19, No. 11. 
1928. The office of Master of Swans. British Birds, vol. 22, No. 4. 
1930. Early records of the mute swan in Norfolk. British Birds, vol. 23, 

No. 11. 
1934. Letter to editors. British Birds, vol. 27, No. 9. 
1957. The mute-swan in England : its history and the ancient custom of 

swan-keeping. Cleaver-Hume Press Ltd., London, 136 pp. 



BIBLIOGRAPHY 197 

Watson, W. Verde. 

1949. Trumpeter swan kills muskrat. Yellowstone Nature Notes, vol. 
23, No. 5. 
Watterson, W. H. 

1935. Notes on the nesting of captive mute swans. The Wilson Bulletin, 
vol. 47, No. 3. 
Weiser, C. S. 

1933. Flying with a flock of swans. Auk, vol. 50, No. 1. 
Wetmore, Alexander. 

1931. The avifauna of the Pleistocene in Florida. Publication 3115, 

Smithsonian Institution. 
1935. A record of the trumpeter swan from the Late Pleistocene of 

Illinois. The Wilson Bulletin, vol. 47. No. 3. 
1951. Observations on the genera of the swans. Journal of the Washington 

Academy of Sciences, vol. 41, No. 10. 
1956. A cheek-list of the fossil and prehistoric birds of North America 
and the West Indies. Publication 4228, Smithsonian Institution. 
Widmann, Otto. 

1907. A preliminary catalog of the birds of Missouri. St. Louis. 288 pp. 
Wilke, Ford. 

1944. Auk, vol. 61, No. 655. 
Witherby, H. F., F. C. R. Jourdain, Norman F. Ticehurst, and Bernard W. 
Tucker. 

1939. The handbook of British birds, vol. 3. H. F. and G. Witherby, 
London. 387 pp. 
Wood, J. C. 

1908. Bird notes from southeastern Michigan. Auk, vol. 25, No. 3. 
Wood, Norman A. 

1923. A preliminary survey of the bird life of North Dakota. University 
of Michigan Museum of Zoology. Miscellaneous Publication No. 10. 
Wright, George M., and Ben H. Thompson. 

1935. Fauna of the National Parks of the United States ; wildlife manage- 
ment in the National Parks. U. S. National Park Service, Fauna Series 
No. 2. 
Yocom, C. F. 

1951. Waterfowl and their food plants in Washington. University of 
Washington Press. 272 pp. 



Appendix 1. Excerpt from "Observations on the genera of the 

swans" 

By Alexander Wetmore, Smithsonian Institution. From Journal of 
the Washington Academy of Sciences, vol. 41, No. 10, October 15, 1951. 

Externally the species of white swans are so similar that the student of study 
skins has difficulty in separating them. The comparative anatomist, . . . work- 
ing with skeletons, has no trouble whatever in dividing them into two principal 
groups on characters so evident that they cannot be disregarded. The differences 
are most apparent in the form of the trachea, sternum, and furculum. Follow- 
ing is a summary of these anatomical characters, with indication of the alloca- 
tion of the species of the Northern Hemisphere and South America : 

a. Trachea passing directly into thorax, not entering sternum; furculum simple; 

tail cuneate genus Cygnus 

Cygnus Bechstein, Orn. Taschenb., pt. 2, 1803:404. Type, by monotypy, 

Anas olor Gmelin. 
Sthenelus Stejneger, Proc. U. S. Nat. Mus. 5:184, 185. Aug. 5, 1882. 

Type, by monotypy, Anas melancoripha Molina. (Not Sthenelus 

Marschall, 1873, emendation for Sthelenus Buquet, 1860, for a genus 

of Coleoptera.) 
Sthenelides Stejneger, Auk 1 (3):235. July 1884. Type, by monotypy, 

Anas melancorphia Molina. New name for Sthenelus Stejneger 

(preoccupied). 
Euolor Mathews and Iredale, Austr. Avian Rec. 3 (5): 117. Dec. 28, 

1917. Type, by original designation, Anas olsr Gmelin. 
Species included: 

Cygnus olor (Gmelin) (skeleton examined). 
Cygnus melancoriphus (Molina) (skeleton examined), 
aa. Trachea making a loop that enters the sternum; furculum especially modified 

at symphysis to accommodate this loop; tail rounded genus Olor 

Olor Wagler, Isis, 1832:1234. Type, by subsequent designation, Cygnus 

musicus Bechstein = Anas cygnus Linnaeus (Gray, 1840). 
Clangocycnus Oberholser, Emu 8 (pt. 1); 3. July 1, 1908. Type, by 

monotypy, Cygnus buccinator Richardson. 

b. Trachea entering anterior end of sternum smoothly, without a dorsal loop. 

subgenus Olor. 
Species included: 

Olor columbianus (Ord) (skeleton examined). 
Olor cygnus (Linnaeus) (skeleton examined). 
Olor bewickii Yarrell. 
bb. Trachea making a dorsal loop as it enters sternum, protected by a bony case 
that projects into the anterior end of the body cavity 

subgenus Clangocycnus 
Species included: 

Olor buccinator (Richardson) (skeleton examined). 
The shape of the furculum and the looping of the trachea in the sternal keel 
are developed in the growing young, the loop lengthening and expanding to the 
end of the sternum as the individual becomes fully adult. This change with 
age has led to misunderstanding of the characters by some not familiar with it. 
The arrangement of the genera above, it may be noted, is identical with that 
of Stejneger in his Outlines of a monograph of the Cygninae, published in 1882. 

198 



APPENDIX 2 



199 



Appendix 2. — Status and Distribution of Trumpeter Swans in 
the United States, 1954 

[Census, August 19-September 3J 



Location 



Montana 
Red Rock Lakes Refuge: 2 
Lower Red Rock Lake.. 



Red Rock River, marsh and potholes. 



Upper Red Rock Lake and East 
Marsh. 3 

Swan Lake and adjacent marsh pot- 
holes. 

Culver Pond 



Groups 



Total 

Centennial Valley (outside Refuge): 
Red Rock River, except Blake Slough. 

Blake Slough 

Jones Reservoirs 

Passmore potholes 

Stibal pothole 

Lima Reservoir — 



2-1, 2- 
2-0, 2- 

(2-0, 2- 
2-0, 2- 
2-0, 2- 
2-0, 2- 
2-0, 2- 
2-0, 2- 
2-0, 2 

^2-0. 
2-3, 2- 
55-0. 
2-0, 2- 
2-0, 2 
2-0. 
1-0. . 



•0, 2-0, 2-0 
•0, 2-0, 2-2. 
-3, 2-0, 2-0, 
-0, 2-1, 2-0, 
-0, 1-0, 2-0, 
-0, 2-4, 2-0, 
-3, 2-0, 2-0, 
-0, 2-0, 2-2, 
-0, 2-3, 2-0, 

-0, 2-0, 199-0, 

-3, 2-3, 2-0, 
-0, 2-0, 2-0, 



Adult- 
cygnet 
ratio 



16:3 
57:16 



260:3 
18:6 

1:0 



Total. 



Beaverhead National Forest: 
Elk Lake 

Gallatin National Forest: 

Hebgen Reservoir 

Aldrich Lake 



Total. 



All other areas: 

Ennis Lake 

Conklin Reservoir (Antelope Valley) 



1-0, 11-0. 
2-8, 2-0. . 

2-0 

6-0 

18-0 

7-0 



2-2. 



2-0, 2-1. 
1-0 



352:28 

12:0 
4:8 
2:0 
6:0 

18:0 
7:0 



Total 
swans 



49:8 

2:2 

4:1 
1:0 



Total. 



Grand total ( Montana) 

Idaho 

Targhee National Forest: 

Beaver Lake 

Pond, 1 mi. northwest of Steele Lake.. 

Steele Lake (Idaho and Wyoming) 

Pond, ){ mi. northeast of Goose Lake. . 

Goose Lake 

"The Hole" 



2-0. 
2-1. 



2-0. 
2-0. 
2-0. 
1-0. 
2-0. 
2-0. 



7:3 

2:0 

2:1 



4:1 



412:40 



2:0 
2:0 
2:0 
1:0 
2:0 
2:0 



Total 

See footnotes at end of table. 
469660 O— 60 14 



11:0 



11 



200 



APPENDIX 2 



Location 


Groups 1 


Adult- 
cygnet 
ratio 


Total 

swans 


Island Park Area: 
Henrys Lake 


2-0 2-2 


4:2 

2:0 

2:4 

19:1 


6 

2 

6 

20 


Island Park Reservoir 


2-0 


Gold Lake 


2-4 


Silver Lake _ 


2-1, 6-0, 2-0, 9-0. . . 




Total 




27:7 


34 


Grand total (Idaho) 




38:7 

2:0 
2:4 
3:2 
6:0 
2:1 
2:0 
1:0 
2:0 
2:4 
2:0 
1:0 
2:0 
2:0 
2:0 
2:0 
2:0 
2:0 
4:0 

2:3 
2:0 
2:0 
2:0 
1:0 

4:3 

2:6 
3:0 
2:0 
3:0 


45 

2 
6 
5 
6 
3 
2 
1 
2 
g 


Wyoming 
Yellowstone Park: 

Pond,, south of Bunsen Peak 


2-0 

2-4 

3-2 


Geode Lake 


Trumpeter Lake. 


Fern Lake 


6-0 


Tern Lake__ __ 


2-1 


White Lake outlet 


2-0 


Solfatara Lake 


1-0 


Grebe Lake.. _ 


2-0 


Madison Junction Lake 


2-4 


Mouth of Alum Creek 


2-0 


2 


Mouth of Pelican Creek 


1-0 

2-0 


1 
2 


Beach Springs.. 


Mouth of Trail Creek 


2-0 


2 


Upper Yellowstone River 


2-0 


2 


Mouth of Chipmunk Creek 


2-0 

2-0 


2 
2 


Yellowstone Lake near Delusion Lake 


Delusion Lake 


2-0 


2 


Ponds between Delusion Lake and 


4-0 


4 


Flat Mountain Arm. 
Riddle Lake 


2-3 


5 
2 
2 


Shoshone Lake 


2-0 

2-0 


Heart Lake _ 


Pond, 1 mi. west of Beula Lake 


2-0 


2 


Pond, 2 mi. west of Boundary Creek 


1-0 


1 


patrol cabin. 
Pond, 2 mi. north of Bechler River 


4-3 


7 


Ranger Station. 
Robinson Lake.. 


2-6 


s 


Lilypad Lake 


3-0 

2-0 


3 
2 


Phone Line Lake 


Pond, south of mouth of Mountain 


3-0 


3 


Ash Creek. 






Total 




64:23 

2:4 
2:0 
4:0 


87 


Grand Teton National Park: 

Pond, 3 mi. northwest of Moran 
Emma Matilda Lake 


2-4 

2-0 

4-0 


6 
2 


Two Ocean Lake 


4 






Total 




8:4 


12 








Grand Total (National Parks) 




72:27 
10:3 

2:0 
2:0 


99 


National Elk Refuge 
Teton National Forest: 
Bridger Lake 


2-0, 2-3, 2-0, 4-0... 

2-0 

2-0 


13 
2 


Snake River 


2 






Total 




4:0 


4 



See footnotes at end of table. 



APPENDIX 3 



201 



Location 


Groups l 


Adult- 
cygnet 
ratio 


Total 

swans 


Targhee National Forest: 

Pond, 3 mi. west of Fish Lake 


2-0 

1-0. 


2:0 
1:0 
3:0 

2:0 
2:2 
2:0 


2 
1 


Pond, 1 mi. southwest of Winegar 

Lake. 
Loon Lake - - 


3-0 

2-0 

2-2 

2-0 


3 
2 


Indian Lake _ 


4 


Reservoir 1 mi. southwest of Indian 
Lake. 


2 


Total - - 




16:2 


18 








Grand total (Wyoming) 




98:32 
7:0 
5:3 


130 


Nevada 


7-0 


7 


Oregon 
Malheur Refuge 4 


5-3 


8 










560:82 


642 









i First number in combination denotes adults, second cygnets. 

2 Does not include 21 swans (13 adults and 8 cygnets) transferred to Malheur and Ruby Lake Refuges 
earlier in the year. 

3 East marsh includes only those potholes draining into the Upper Lake. 
* Transplanted swans. 

Appendix 3. — Measurements of Trumpeter Swan Eggs 

The following measurements of the eggs of trumpeter swans were 
recorded in 1955 on Red Rock Lakes marshes from a random sample 
of 21 nests which contained normal clutches. Measurements were 
made accurately to the nearest half-millimeter at the points of 
greatest length and girth. These measurements are believed to be 
representative of this species in its United States breeding habitat. 



Length Width 



Length Width 



Mm. Mm. 

Clutch No. 1: Clutch No. 3: 

Eggl 116 72.5 Eggl 

Egg2 111.5 69.5 Egg2 

Egg3 118 75 Egg3 

Egg4 119 74 Egg 4 

Egg5..._ 117.5 74 Egg5 

Clutch No. 2: Clutch No. 4 : 

Eggl 106 70.5 Eggl 

Egg 2 105.5 71 Egg 2 

Egg3 _ 106 71 Egg3 

Egg 4 106 68 Egg 4 

Egg 5 110 71 Egg 5 

Egg6 111.5 70 

Egg 7 109.5 70 

Egg 8 108 70 



Mm 


Mm. 


119.5 


75 


118.5 


74 


118.5 


74.5 


121 


76 


123 


74 


112 


73.5 


112.5 


74 


113 


73 


112 


72.5 


114.5 


72.6 



202 



APPENDIX 4 



Length Width 



Length Width 



Mm. 

Clutch No. 5: 

Eggl 108.5 

Egg 2 HI 

Egg 3 113.5 

Egg4. 113 

Egg 5 108 

Egg 6 111 

Clutch No. 6: 

Egg 1 109 

Egg 2 109.5 

Egg 3 110.5 

Egg 4 107.5 

Egg 5 108 

Clutch No. 7: 

Egg 1 108 

Egg 2 107 

Egg 3 108 

Egg 4 109 

Egg 5 110 

Clutch No. 8: 

Egg 1 112 

Egg 2 105.5 

Egg 3 109 

Egg 4 106 

Egg 5.... 110.5 

Egg 6 106.5" 

Clutch No. 9: 

Egg 1 106 

Egg 2 113 

Egg 3 110 

Egg 4 110 

Egg 5 109 

Egg 6 .__ 109 

Clutch No. 10: 

Egg 1 118 

Egg 2 117 

Egg 3 114 

Egg 4 117 

Egg 5 117 

Clutch No. 11: 

Egg 1 104.5 

Egg 2 104 

Egg 3 106 

Egg 4 107 

Egg 5 108 

Clutch No. 12: 

Egg 1 104 

Egg 2 106 

Egg 3 105 

Egg 4 105 

Egg 5 105 

Egg 6 104 



Mm. 



71 

70.5 

69 

68 

71 

71 

71 

73 

72.5 

72 

71.5 

72 

72 

72 

71.5 

73 

72.5 
71.5 
72 
71 

72.5 
72 

75 
75 
74 
73 
75 
75 

76.5 

75 

75.5 

74.5 

77 

72 

70.5 

71 

71.5 

72.5 



72 

72 

72 

70.5 

71 



Mm. 

Clutch No. 13: 

Egg 1 120 

Egg 2 116.5 

Egg 3 114 

Egg 4 117 

Clutch No. 14: 

Egg 1 H3.5 

Egg 2 113.5 

Egg 3 Hi 

Egg 4 108 

Egg 5 109 

Egg 6 HO 

Clutch No. 15: 

Egg 1 H4.5 

Egg 2 H5.5 

Egg 3 114 

Egg 4 1H.5 

Egg 5 113.5 

Egg 6 113 

Clutch No. 16: 

Egg 1 H6 

Egg 2 121.5 

Egg 3 114 

Egg 4 117 

Clutch No. 17: 

Egg 1 109 

Egg 2 107.5 

Egg 3 105.5 

Egg 4 105.5 

Egg 5 108.5 

Egg 6 109 

Clutch No. 18: 

Egg 1 no 

Egg 2 1U.5 

Egg 3 106 

Egg 4 112 

Egg 5 108 

Clutch No. 19: 

Egg 1 107 

Egg 2 106.5 

Egg 3 104.5 

Egg 4 111.5 

Clutch No. 20: 

Egg 1 109 

Egg 2 108 

Egg 3.... 104 

Egg 4 109 

Clutch No. 21: 

Eggl 115 

Egg 2 114 

Egg 3 112 



Mm. 



74.5 

72 

73 

73 

71.5 

74 

73 
73 

74 
74 



71.5 

72 

72 

72 

73 

72 
74 
74 
73 

72 
72 
70.5 
71 



Appendix 4. — Food Analysis 

The following detailed data were obtained by stomach or scat 
analysis. 

Stomach contents of predator-killed cygnet, estimated age 4 weeks, in Yellow- 
stone Park, found and examined by Condon : 

Freshwater fairy shrimp (Eitbranchipus sp.) — three complete specimens 

plus several fragments 
Carex sp. — shoot fragments 
Chara sp. — small piece 
Quartz sand grit — considerable quantity 
Stomach contents of 6 cygnets, estimated ages 3-4 weeks, found dead, 5 on 
loafing ground and 1 at nesting site, at. Red Rock Lakes Refuge (Montana) on 
July 12, 1951, by Banko (analysis by Charles C. Sperry, Section of Food Habits, 
U.S. Fish and Wildlife Service, March 19-20, 1952) : 



FOOD ANALYSIS 203 

209941 (RR1) female. 

Percentage of animal matter— ; of vegetable — 100 ; of gravel — 40. 

Contents: Leaf and stem fragments of aquatic plants— 98 percent. 
Seeds: 34 Carcx (plus fragments of a few more)— 2 percent; 1 
Eleocharis, 1 Hippuris, and 2 Myriophyllum — trace. 

209942 (RR2) sex (?) : 

Percentage of animal matter — trace ; of vegetable — 100 ; of gravel — 25. 

Contents: Leaf and stem fragments of aquatic plants— 95 percent. 
Seeds: 126 Carex (and fragments of a few more)— 5 percent; 5 Eleo- 
charis, 1 Galium, and 2 Myriophyllum — trace. 

209943 (RR 3) female, weight 449.5 gms. 

Percentage of animal matter — 0; of vegetable — 100; of gravel — 18. 
Contents: Leaf and stem fragments of aquatic plants— 80 percent. 
Seeds: 208 Carex (plus many fragments)— 20 percent; 2 Eleocharis, 

1 Galium, 2 Hippuris, 2 Myriophyllum, 2 Potamogeton, and 5 
Scirpus — trace. 

209944 (RR4) male. 

Percentage of animal matter— ; of vegetable — 100; of gravel— 75. 
Contents : Leaf and stem fragments of aquatic plants — 100 percent. 

209945 (RR 5) male, weight 381 gms. 

Percentage of animal matter — ; of vegetable — 100 ; of gravel — 20. 

Contents: Leaf and stem fragments of aquatic plants— 90 percent. 
Seeds: 86 Carcx (and fragments of about as many more)— 10 per- 
cent; 4 Eleocharis, 1 Myriophyllum, and 11 Scirpus — trace. 

209946 (RR 6) sex ( ?) , weight 447 gms. 

Percentage of animal matter — 0; of vegetable — 100; of gravel — 25. 
Contents: Leaf and stem fragments of aquatic plants— 90 percent. 
Seeds : 51 Carex (and fragments of about as many more) — 10 percent ; 

2 Eleocharis and 1 Scirpus — trace. 

Analysis of 17 samples of droppings from Grebe Lake, Yellowstone Park. 
Collected September 2, 1943, by O. J. Murie : 

Algae — About % of the droppings are hard or very hard and these almost 
invariably consist largely of filamentous green algae intermingled with 
much fine quartz grit. 

Carex — Four of the 17 envelopes contained droppings that consisted largely 
of remains of Carex spikes (the perigynia, achenes, scales and rachis). 

Potamogeton — Seeds of a Potamogeton resembling P. pusillus were found 
in one sample. It is suspected that the unidentified herbaceous vegeta- 
tion found in 6 samples (of the 17) may be largely Potamogeton. 

Nuphar polysepala — In a sample consisting largely of Carcx but also con- 
taining considerable algae, fragments of the seed wall of Nuphar polysepala 
were present to the extent of 10-20 percent of the whole. 

Sagittaria — Shredded tissues in one sample are suggestive of Sagittaria 
but definite identification could not be made. 

State of plant material— Most, possibly all, of the plant material seems to 
have been in a living state when eaten though in some cases the color of 
the substance does not seem to support this. 

Insect — Animal material was almost negligible in the dropping samples. 
However, in the one sample containing seed fragments of Nuphar to- 
gether with Carex, there were numerous specimens of caddis fly larvae 
(Trichoptera) totaling possibly 5 percent of the whole. 



204 APPENDIX 4 

Report on the examination of 2 stomachs collected on the Red Rock Lakes 
Refuse, Montana, by A. V. Hull. Analysis by Sperry : 

200172, female, killed by flying into a fence. December (early) 1938. 

Stomach— full ; gullet — full. 

Percentage of animal matter — trace; vegetable — 100; sand and gravel, 
etc.— 10. 

Contents: Potamogeton pectinatus, 443 tubers (largest being % x % 
inch) and fragments of rootstock — 100 percent ; 1 caddis larvae case- 
trace ; fragments of 2 gastropoda — trace, and possibly taken as 
"gravel", which was 70 percent fine sand. Weight of food: gullet — 
4 oz. ; gizzard — 3V 2 oz. Weight of gravel: 2 oz. (in gizzard). 
206190, sex ( ?). June 10, 1939. 

Stomach— full ; gullet— full. 

Percentage of animal matter — trace ; vegetable — 100 ; of sand and gravel, 
etc.— 15. 

Contents : Potamogeton pectinatus, 597 tubers (largest being % x % x \< 2 
inch), rootstock fragments and parte of leafy plant — 96 percent (leafy 
part about 6) ; other leaf and stem material: Ranunculus aquatilis — 
3 percent; Ceratophylliim — trace; also algae (fine stringy type) — 1 
percent. Seeds : 29 Carex, 1 Hippuris vulgaris, 3 Scripus acutus, 84 
Potamogeton, 1 Sparganium, and 1 Zannichellia — trace; 2 caddis 
larvae cases — trace. Notes on weights: food (wet) — liy 2 oz. ; sand 
and gravel — 2 oz. ; stomach and gullet tissue (wet) — 6 oz. 

Stomach examinations of trumpeters found at Picnic Springs, Red Rock Lakes 
Refuge, Montana, by A. V. Hull. Report by E. R. Kalmbach, in charge, Denver 
Laboratory, U.S. Fish and Wildlife Service: 

212892, adult. Examined by F. M. Uhler, April 2-7, 1937. 
Percentage of animal matter — 0: vegetable — 100 percent: 3 lead shot. 

gravel, etc. — 30 percent. 
Contents: leaves and stems of white water buttercup (Ranunculus sp. — 
probably 7?. trichophyllus) — 70 percent; aquatic mosses (Ambly- 
stegium sp. — 30 percent and Fissidens sp. — trace). 

212893, juvenile. Examined by C. Cottam, April 2-7, 1937. 

Percentage of animal matter — ; vegetable — 100 percent ; 17 lead shot, 
gravel, etc. 

Contents: 1 seed of Potamogeton perfoliatus ; moss plant fiber; unde- 
termined plant fiber. (The stomach was too nearly empty to give 
percentages.) 

212894, juvenile. Examined by C. Cottam, April 2-7, 1937. 

Percentage of animal matter — 1 percent: vegetable — 99 percent; 11 
lead shot, tooth of Thomomys sp., gravel, etc. — 33 percent. 

Contents: plant fiber of white water buttercup, Ranunculus (Batrach- 
iiim), probably trichophyllus — 63 percent: moss, Fissidens, probably 
grandifrons — 30 percent; moss, Calliergon sp. — 3 percent; moss, 
Amblystegium sp. — 1 percent; buds and leaves of Potamogeton 
panormitanus — 1 percent; 3 seeds of Potamogeton perfoliatus — tr. : 
7 seeds (2 species of Carex sp. — tr. ; undetermined plant fiber — 1 per- 
cent : 3 larvae of caddis fly of 3 species, one of which appeared to be 
a Hydropsychidae — 1 percent: fragments of larvae of Dytiscidae — tr. 
mollusk shell fragment — tr. (Feather fragments — 2 percent). 

On the basis of the material available I should say that each bird had 
succumbed to lead poisoning. The stomach of the adult male contained 



APPENDIX 5 



205 



three pellets of lead, weighing 0.248 gram. The juvenile female had eleven 
pellets of lead, weighing 0.498 gram, and the juvenile male, seventeen pellets, 
weighing 0.857 gram. 

In each instance there was a pronounced greenish discoloration of the 
contents, and a hardening of the gizzard pads. 



Appendix 5.— Supplementary Data, Annual Swan Census 



Period 



Fall 1929- 



Summer 1930. 



Summer 1931. 



Summer 1932. 



Summer 1933. 



Summer 1934. 



Summer 1935. 
Summer 1936. 



Summer 1937. 



Aug. 4-7, 1938...- 

Aug. 15-16, 1939- 
Aug. 16-17, 1940. 

Aug. 15-16, 1941.. 
Aug. 20-22, 1942. . 
Aug. 26-28, 1943. 
Aug. 12-18, 1944. 



Coverage; Agency 



Yellowstone Park (NPS) - 



Yellowstone Park, Jack- 
son Lake Swamp 

(NPS). 

Expanded coverage in 
Yellowstone Park 

(NPS). 



Further expanded cover- 
age in Yellowstone 
Park, also including 
Red Rock Lakes area 

(NPS). 
Coverage similar to 1932 

(NPS). 



.do. 



.do 



Yellowstone Park plus 
adjacent areas (NPS). 

R.R.L. Refuge and ad- 
jacent areas (FWS). 

Coverage similar to 1936 
(NPS and FWS). 



.do 






Coverage similar to 1938 
(NPS and FWS). 



Yellowstone Park only 
partially covered 
(NPS). 

Refuge census expanded 
to include many con- 
tiguous areas (FWS). 

Yellowstone Park only 
covered (NPS). FWS 
coverage expanded. 

No census in Yellowstone 
Park. 

FWS coverage same as 
1941. 

No census in Yellowstone 
Park. 

FWS coverage similar to 
1942. 

Yellowstone Park and ad- 
jacent areas (NPS). 

R.R.L. Refuge and ad- 
jacent areas (FWS). 



Remarks 



Survey only, complete 
census not attempted; 1 
nesting pair located 
(Trumpeter Lake). 

Survey expanded; 1 nest- 
ing pair— Trumpeter 
Lake, Tern Lake, Jack- 
son Lake swamps. 

Initial complete census 

attempted; nesting pairs 

— Yellowstone Park (4), 

Jackson Lake swamps 

(1). 

Nesting pairs— Yellow- 
stone Park (4) Red Rock 
Lakes (5), Hebgen Res- 
ervoir (1). 

Nesting pairs— Yellow- 
stone Park (3), Red 
Rock Lakes (7) (aerial 
census recommended for 
1934). 

Nesting pairs — Yellow- 
stone Park (6), Red 
Rock Lakes (7), Henrys 
Lake (3), Rock Lake, 
Wyo. (1). 

Detailed data missing 



References 



Bridger Lake and Squirrel 

Meadows included. 
Henrys Lake, Elk Lake, 

Blake Slough included. 
Coverage probably not 

quite as complete as 

1936. 

Winegar area adjacent 
Yellowstone Park in- 
cluded; both Park and 
Refuge counts con- 
ducted Aug. 4-7. 

Coverage probably equal 
to or slightly greater 
than in 1938. 

Park areas missed in 1940 
held 10 swans in 1939. 

Malheur and Elk Refuges 
included for first time. 



Areas adjacent Yellow- 
stone Park not included. 

Ennis and Elk Lakes in- 
cluded. 

No census personnel for 
park due to World War 
II and bad fire season. 

No census personnel for 
park — World War II. 



Including Indian, Loon, 
Puddle, Chain Lakes. 

Sheridan Reservoir, 
Wade, and Conklin 
Lakes. 



Joseph Llxon, 1931: 452. 
Wright and Thompson, 
1935: 104. 

W'right and Thompson, 

1935: 104. 
E. L. Arnold, ltr., August 

1930. 
O. W. Wright, ltr., May 2, 

1934. 
Q. W. Wright, ltr., May 2, 

1934. 

Do. 



Do. 



F. W. Childs, ltr., Oct. 20, 
1934. 



Unsigned report (NPS 

files). 
E. B. Rogers, ltr., Aug. 19, 

1936. 
A.V. Hull, ltr., July 28, 

1936. 
E. B. Rogers, ltr., July 24, 

1937. 
A. V. Hull, ltr., Aug. 5, 

1937. 
E. B. Rogers, ltr., Aug. 15, 

1938. 
A. V. Hull, ltr., Sept. 8, 

1938. 

E. B. Rogers, ltr., Aug. 23, 

1939 
A. V.'Hull, ltr., Aug. 19, 

1939 
E. B. Rogers, ltr., Sept. 24, 

1940. 
A. V. Hull, ltr., Oct. 31, 

1940. 



E. B. Rogers, ltr., Aug. 19, 

1941. 
A.V. Hull, ltr., Aug. 20, 

1941. 
E. B. Rogers, ltr., Aug. 18, 

1942. 
A. V. Hull, ltr.. Sept. 11, 

1942. 
E. B. Rogers, ltr., Oct. 8, 

1943. 
W. M. Sharp, ltr., Sept. 2, 

1943. 
E. B. Rogers, ltr., Sept. 1, 

1944. 
W. M. Sharp, undated ltr., 

1944. 



206 



APPENDIX 5 



Appendix 5. — Supplementary Data, Annual Swan Census — Con. 



Period 



Coverage; Agency 



Remarks 



References 



Aug. 16-31, 1945. 



Aug. 11-17, 1946. 



Aug. 10-16, 1947. 



Aug. 16-20, 1948. 



Aug. 3-5, 1949. ... 

Aug. 1-4, 1950 

July 31-Aug. 3, 1951 

July 16-21, 1952 

Aug. 3-6, 1953 

Aug. 31-Sept. 3, 1954 
Aug. 29-31, 1955.... 
Aug. 27-31, 1956.... 
Aug. 20-23, 1957—. 



No census in Yellowstone 
Park (NPS). 

R.R.L. Refuge and adja- 
cent areas (FWS). 



Yellowstone Park, great- 
est coverage to date 
(NPS). 

R.R.L. Refuge and adja- 
cent areas (FWS). 



Yellowstone Park, cover- 
age similar to 1946 
(NPS). 

Refuge and adjacent area 
coverage similar to 1946 
(FWS). 

Yellowstone Park and 
adjacent areas (NPS). 

FWS coverage of refuge 
and adjacent areas simi- 
lar to 1947. 

Coverage similar to 1948 
(NPS and FWS). 



Coverage similar to 1949 
(NPS and FWS). 



Coverage comparable to 
1950 (NPS and FWS). 



Aerial coverage, similar to 
1951 (NPS and FWS). 



Aerial coverage, similar to 
1952 (NPS and FWS). 



Aerial coverage, similar to 
1953 (NPS and FWS). 



Aerial coverage, similar to 
1954 (NPS and FWS). 



Aerial coverage, similar to 
1955 (NPS and FWS). 



Aerial coverage, similar to 
1956 (NPS and FWS). 



No census personnel 

World War II. 
Coverage similar to 1944.. 



Ground count swans on 21 
of 68 lakes checked. 



Aerial coverage for 1st 
time. 



Ground and aerial counts 
in park. 

Aerial counts in refuge and 
vicinity. 



Park, refuge, and adjacent 
areas covered aerially; 
similar to 1947, Ruby 
Lake Refuge included. 

Aerial counts all areas; 
Railroad Ranch area 
added. 

Aerial counts, all areas 



.do. 



Bridger Lake and waters 
north of Moran, Wyo., 
added. 

Upper Jackson Hole areas 
included for first time. 



Teton National Park 
added. 



Does not include 6 swans 
transferred from Red 
Rock Lakes to Delta, 
Manitoba, before census. 

Same NPS and FWS ob- 
servers conducted census 
1950-1956. 

Pair-family-group status 
tabulated for 3d year. 



E. B. Rogers, ltr., Aug. 13, 

1945. 
Refuge Narrative Report, 

May-August 1945. 
MacDonald, ltr., Oct. 1, 

1945. 
E. B. Rogers, ltr., Aug. 27, 
1946. 

Refuge Narrative Report, 

May- August 1946. W. 

M . Sharp, undated report, 

1946 
E. B. Rogers, ltr., Sept. 10, 

1947. 

Refuge Narrative Report, 

May-August 1947. 
W. N. Anderson, ltr., Sept. 

19, 1947. 
E. B. Rogers, ltrs., Aug. 24 

and Sept. 1, 1948. 
W. N. Anderson, ltr., 

Sept. 22, 1948. 

E. B. Rogers, ltr., Aug. 9, 

1949. 
Refuge Narrative Report, 

May-August 1949. 
E. B. Rogers, ltr., Aug. 11, 

1950. 
Refuge Narrative Report, 

May-August 1950. 
E. B. Rogers, ltr., Aug. 6, 

1951. 
Refuge Narrative Report, 

May-August 1951. 
E. B. Rogers, ltrs., July 22, 

1952. 
Refuge Narrative Report, 

May-August 1952. 
NPS Circular 12, Aug. 14, 

1953. 
Refuge Narrative Report, 

May-August 1953. 
NPS Circular 7, Sept. 8, 

1954. 
Refuge Narrative Report, 

May- August 1954. 
NPS Circular 11, Sept. 8, 

1955. 
Refuge Narrative Report, 

May- August 1955. 
NPS Circular 15, Sept. 7, 

1956. 
Refuge Narrative Report, 

May-August 1956. 
NPS Circular 15, Aug. 21, 

1957. 
Refuge Narrative Report, 

May-August 1957. 



INDEX 



accidents, 143. 

Adams, Edward, 22, 23, 24, 27, 189. 

Adams, Lowell, 2. 

aggression, interspecific, 104-106. 

Alaska, 2. 

Copper River Basin (includes 
Bremner and Tasnuma Rivers), 
32, 33-36, 52-54, 114. 
fossil swans, Kodiak Island, 9. 
Kenai Peninsula, 33-35, 114, 116. 

130. 
life zones, 38. 
migration, spring, 27. 
population : 

early trumpeter records, 22, 23, 

25. 
recent records, 7, 31-35, 145. 
swan census, 1949-57, 32. 
Alberta, 2, 31, 37, 38, 58, 80. 
Allen, J. A., 10, 189. 
American Fur Company, 13. 
Anderson, R. M., 24, 189. 
Arkansas, 26. 
Arkansas River, 15. 
Armstrong, E. A., 87, 101, 102, 189. 
aspergillosis, 140, 141. 
Atwood, Earl L., 3. 

Audubon, John James, 13, 15, 26, 28, 76, 
124, 167, 189. 

banding, 28, 182-183. 

bantam hens, 115, 124. 

Bailey, Alfred, 20, 28. 

Bailey, Florence, 28. 189. 

Bailey, Vernon, 17, 22, 35, 189. 

Baird, Spencer F., 17 ; with Thomas M. 

Brewer and Robert Ridgway, 24, 167- 

168, 189. 
Barnston. George, 17, 19, 23, 27, 125- 

126, 189. 
Barrows, George, see Oberhansley. 
Bates, J. M., 24, 189. 
bear, 131, 135. 



Bear River Migratory Bird Refuge, 129. 
Beard, Dan, et al., 137, 189. 
Bedford, Duke of, 168, 170, 171. 
Beebe, AVilliam, 5, 189. 
Beed, Watson E., 42. 
behavior, 77-93, 99-100. For aspects 
not listed in Contents, see under swan 
species. 
Belknap, Jeremy, 10, 189. 
Bellrose, Frank, 136. 
Beltrami, Count G. E., 12, 14, 24. 
Bent, A. C, 21, 24, 26, 27, 28, 86, 114, 

126, 169, 190. 
Biological Survey, U.S., 166, 174. 
Blaauw, F. E., 70, 77, 117, 129, 143, 168, 

170, 171, 172, 190. 
Blines, Jasper, 12, 24, 190. 
bobcat, 133. 
botulism, 141. 
Branson, J., 33. 

breeding, 93-100: see also habitat, 
age, 94-96. 
range : 

Bewick's, whooper, whistling 

swans, 7. 
trumpeter, 7. 

Canadian, 14, 18-19. 
United States, 10-11, 12- 
13, 15-16, 20, 21, 22-26, 
31-37, 38-54. 
maps, 26, 39, 52. 
Brewer, Thomas, 167-168. 
British Columbia : 
behavior, 78, 81. 
breeding, 16, 31, 32. 
captivity, 170. 
distribution and range, 2, 7, 16, 26, 

27, 29, 31, 32, 33. 
food, 128. 

lead poisoning, 137-138. 
parasites, 141. 
predation, 134. 
starvation, 139. 
wintering, 7, 29, 31, 33. 



207 



208 



INDEX 



Brooks, Allan, 29; with H. S. Swarth, 

23, 190. 
Brower, J. V., 21, 24, 190. 
Burnett, W. L., 20, 190. 
Butler, A. W., 25, 190. 

California, 16, 17, 20, 28, 29. 
Canada. (See also Alberta, British Co- 
lumbia, Delta Waterfowl Research 
Station) : 

captivity, 168. 

distribution and range, 14, 17-19, 
23, 27, 37, 145. 

interspecific tolerance, 80. 

nesting, 113. 

longevity, 143. 

swan skin trade, 12, 13. 19. 
Canadian Company, 12, 13. 
Canadian Wildlife Service. 2, 37, 183. 
Cameron, E. S., 24, 69, 126. 
captivity of trumpeter, 123-124, 167- 

172. 188 : see Malheur, Delta. 
census, annual swan, 144-161, 205-206. 
Chaddock, T. T., 141, 142, 190. 
Chenopis atratus, 5, 6; see swan, black. 
Chesapeake Bay, 126, 129. 
Chicago Natural History Museum, 23, 

24. 
Christian. J. J., 164, 190. 
clutch size, 114. 

Coale. H. K., 9, 19. 24, 169, 190. 
Colorado, 20. 28. 
Columbia River, 12. 16-17, 25, 27, 28, 

126. 128. 

Colwell, Owen, 78. 81, 190. 
Condon, David de Lancey, 2, 30, 64. 70, 
79-80, 85, 106, 110, 111, 112, 116, 123. 

127, 131, 132, 135, 136, 137. 138-139, 
175, 190. 

Connecticut, 10, 169. 

Cooke, Wells W., 24, 190. 

Cooper. J. G., 16, 27, 190. 

coot, 104-105. 

Copper River Basin, .sec Alaska. 

Coues, Elliot, 11, 12, 21, 24, 25, ON, 190. 

Cottam, Clarence, 35. 

Cowan, Ian McTaggart, 29, 128, 141-141, 

191. 
coyotes, 133, 134-135, 174, 177. 
Culver, Lillian, 169-170. 
Currituck Sound, 129. 



cygnets, trumpeter : 

captivity, 124-124, 171. 

description. 62-63, 70-71. 

development, 117-122. 

food, 122-125. 

mortality, 132, 173. 

population dynamics, 145-164. 

supervision by parents, 117, 118. 

and weather, 138-139. 

weight, 117, 120. 
cygnets, whistling, 63. 
Cygninae, description and systematica 

2, 5, 6. 
Cygnus melaneoriyhus, 5, 6 ; see swan, 

black-necked. 
Cygnus, olor, 5, 6; see swan, mute. 
Cygnus paloregonus, 8. 

Dall, W. H., and H. M. Bannister, 23, 

31, 191. 
Darling. J. N., 175. 
death, see mortality. 
Delacour, Jean, 6, 70, 91, 95, 116-117, 

168 : and Ernst Mayr, 91, 94, 191. 
Dementiev, Georges, and N. Gladkov, 

6, 191. 
Denmark, 104-105, 129, 130, 187. 
Delta Waterfowl Research Station, 3, 

71, 120, 129, 136, 140, 142, 181. 
Derscheid, J. M., 172, 191. 
De Smet, Pierre Jean, 14, 15-16, 24. 
Dewar, J. M.. 98, 191. 
Dickinson. E. M.. 140-141. 
disease, 140-141. 
display, 86-93, 95. 
distribution, trumpeters : 

United States, 6-7, 8-37, 199-201. 

Canada. 14, 17-19, 23, 27, 37, 145. 
Dixon, Joseph, 130-131, 173. 174, 175, 

191. 205. 
ducks. 101. 104-105. 

goldeneyes, 90, 187. 

mallard. 59. 93, 105, 187. 

merganser, 59. 

pintails. 93. 

ruddy, 105. 

scaup, lesser, 105. 

eagle, 56. 

bald, 88. 

golden. 133. 134. 
Edge, Rosalie, 176. 
Edwards, Ralph, 83, 134. 



INDEX 



209 



eggs, trumpeter : 

clutch size, 114. 

description, 113-114, 201-202. 

failure, 130-131. 

incubation and hatching, 114-115, 
116. 

price, 19. 
Ellis, John, 98, 191. 
Emergency Conservation Committee, 

176. 
England, 97, 168, 187. 

Queen of, 168. 
Eriekson, Ray C. 3, 65, 91-92, 94-95, 

101. 105, 125. 
Errington, Paul, 164, 191. 
escape-distance, 78-79. 
Evans, A. H., 143. 191. 

family ties, 96-98, 117-120, 121-122. 
Featherstonhaugh, Duane, 24, 32, 61, 

80, 81, 105, 136, 137, 191. 
Fish and Wildlife Service, U.S., 2, 23, 
40, 144-145, 188. 

Denver Wildlife Research Labora- 
tory, 123, 127, 137. 
Patuxent Research Refuge, 143, 
168, 170. 
flight, 73, 74-77. 
Florida, 8. 
flyways : 

Atlantic, 7, 20, 37. 
Central, 20. 35-37. 
Mississippi, 20, 28, 37. 
Pacific, 7, 20, 28-35. 
food : 

other swans, 129. 

trumpeter, 122-130, 177-178, 185- 
186. 202-205. 
foot deformities, 140. 
Forbush, Edward Howe, 20, 69, 165- 

166. 191. 
Fordyce, A. Blaine, 3, 31. 
France, 168. 
Franklin Bay, 17, 18. 
French, Cecil, 21, 24, 170, 171. 
Friedmann, Herbert K., 3, 9, 191. 

Gabrielson, Ira X., 29, and S. G. Jew- 

ett, 32, 191. 
Germany, 129. 
Glacier National Park. 28. 
goose, Canada, 105. 172. 
snow, 166. 



Grand Teton National Park, 105, 120, 

184, 200. 
Grundtvig, F. L., 25, 192. 
Groves, Frank W., 29. 
gull, 132, 133, 140. 

habitat, breeding, 19-20, 38, 39-54. 
wintering, 55-61. 
maps, 26, 39, 52. 
Haecker, F. W., R. A. Moser, and J. B. 

Schwenk, 37, 192. 
Hanson, Fred, 169-170. 
Hansion, James F., 3. 
Hanson, Alta, 3, 169. 
Hart. R. O., 106, 192. 
hatching dates, 115-116. 
Hayden, A., 24, 173. 
Hearne, Samuel, 11, 26-27. 192. 
Heerman, A. L., 17, 192. 
Heinroth; Oskar and Magdelena, 3, 90, 

114, 129, 192. 
Henrys Fork, see Snake River. 
Herman, Carlton M., 142-143. 
Hewett, Stanley, et al., 29. 
Hilden. O., and P. Linkola, 89, 192. 
Hilprecht, Alfred, 6, 62, 70, 94, 101, 129, 

133, 143, 192. 
Hochbaum, H. Albert, 3, 83, 106, 120, 

192. 
Holman, John P., 83, 134, 192. 
Hornaday, William T., 165, 192. 
Howard, L. O., and W. F. Fiske, 163, 

192. 
Hudson Bay, 11. 19. 
Hudson's Bay Company, 11, 12, 13, 17- 

19. 
Hudson River, 62, 187. 
Hughes, J. G, 126, 192. 
Hull, A. V., 33, 115, 127, 128, 135, 137. 

176, 192, 204, 205. 
Hunt, Clarence, 173. 
hunting, 135-137, 166, 173, 174, 176, 187. 
hybridization, 172, 188. 

icing of swan plumage, 139-140. 

Idaho, 2, 24, 28, 30-31, 33, 54, 60, 70, 78, 
98, 136, 137, 138, 144, 166, 169, 175, 
176, 200 ; see Island Park, Snake Riv- 
er, Railroad Ranch. 

illegal kills, 136-137, 173, 174, 176. 

Illinois, 8, 9, 20, 169. 

Illinois Natural History Survey, 136. 

incubation, 114r-115 ; artificial, 116. 



210 



INDEX 



Indiana, 24. 

Indians, 4-5, 9, 10, 11, 13-14. 

interspecific tolerance, 80-83, 91, 104- 

106. 
Iowa, 20, 24, 25, 38, 167-168, 186. 
Island Park, 30, 55, 56-67, 61, 98, 137, 

139, 140, 143, 147, 200. 

Jackson Hole. 20, 24, 30, 55, 60, 139, 

166. 
James Bay, 19. 
James, Edwin, 25, 193. 
Jellison, William, 142. 
Jewett, S. G., et al., 29, 193. 
Johnsgard, Paul, 3. 
Johnson, Jim, 33. 

Kalmbach, E. R., 137, 204-205. 

Kansas, 20. 

Keating, William, 13, 193. 

Kellogg Bird Sanctuary, 126, 170, 171- 

172. 
Kenai Peninsula,, see Alaska. 
Kennedy, George, 2. 
Kentucky, 13-14, 15, 167. 
Kenyon, Karl, 7. 
Kepner, E. L., 32. 
Kittams, Walter H., 2, 176. 
Kluijver, H. X.. 164, 193. 
Knight, W. C., 21, 24, 193. 
Kodiak Island, see Alaska. 
Kortright, F. H., 65, 143, 193. 
Kumlien, Ludwig, and X. Hollister, 25, 

193. 
Kubichek, W. F., 129. 

Lacey Act, 165. 

Lack, David, 164, 193. 

LaNoue, Francis, 72. 

Lattin, Frank H., 19, 193. 

Lawson, John, 10, 25, 26, 193. 

lead poisoning, 136, 137-138, 142, 204- 
205. 

legends. 4-5. 

legislation, protective, 165-167. 

Lewis and Clark expedition, 11-12, 24, 
25, 68, 125. 

life cycle, 62-143. 

life history studies needed, 186. 

Lima Reservoir, 147, 151, 199. 

Lister. R., 37, 193. 

London, swan-skin market, see Hud- 
son's Bay Company. 



London Zoological Gardens, 168. 

Long, Stephen H, 13, 25, 26. 

longevity, 143. 

Lorenz, Konrad, 79, 90, 193. 

Louisiana, 15, 20. 

Low, G. C, 97-98, 193. 

MacFarlane, Roderick, 17-19, 23, 113, 

114, 193 ; sec Mair. 
Mackay, Ronald H, 2, 31, 36, 122, 194. 
Mackenzie River, 17, 38, 113. 
Madison River, 36, 55, 59, 60, 61, 134. 
Madsen, H, 194. 
Maine, 169. 

Mair, Charles, and Roderick MacFar- 
lane, 18, 194. 
Malheur National Wildlife Refuge : 

accidents, 143. 

banding, 182-183. 

behavior, 83, 101, 105. 

breeding age, 94-95. 

disease, 141. 

food, 125, 128. 

management, 186. 

mortality, 133. 

population, 147, 200. 

transplanting, 29, 178, 180, 181. 
management of trumpeters, 165-188. 

recommendation, 183-138. 
Marler, George, 176. 
Maryland, 20. 
Massachusetts, 10, 169, 170. 
mating fidelity, 96. 
McDermott, John F., 13, 194. 
McKinney, Frank, 3, 91. 
McLean. Donald, 28, 29, 78, 194. 
menage a trois, 98. 
Merriam. C. Hart, 10, 21, 24. 194. 
Merrill. D. E., 20, 194. 
Michigan, 28, 62, 187, see Kellogg Bird 

Sanctuary, 
migration, 25-37, 146-147. 
Migratory Bird Conservation Act, 166. 
Migratory Bird Treaty Act, 136, 165- 

166, 169. 171. 
mink, 132. 
Minnesota, 13, 16, 20. 24. 27, 28, 37, 38, 

169, 186. 
Mississippi River, 14. 15, 26 ; see Fly- 
ways. 
Missouri, 11, 12.20,24,38. 
Missouri River, 11, 25. 
Monson, Melvin A.. 33. 52-54, 114, 194. 



INDEX 



211 



Montana, 16, 20, 24, 28, 31, 36, 38, 54, 
55, 60, 61, 134 ,136, 144, 166, 169, 170, 
174, 186, 200; see Red Rock Lakes 
Refuge. 

Mosher, A. D., 24, 194. 

mortality, 131-142, 183. 

Morton, Thomas, 10. 

Munro, J. A., 23, 29, 31, 78, 128, 137- 
138, 139, 194. 

Murie, Adolph, 135, 194. 

Murie, Olaus, J., 127, 139-40, 203. 

muskrat, 43, 51, 105, 106, 111, 112, 177. 

National Elk Refuge : 

breeding, 80, 95, 96. 

hunting, 167. 

incubation, 115. 

population, 200. 

transplanting, 178, 180. 

wintering, 30, 55, 60, 61. 
National Museum, U.S.. 2, 3, 23, 24. 
National Park Service, 2, 60, 144-145, 
172-176; see Condon. Dixon Ober- 
hansley, Thompson, Wright. 
National Zoological Park, 169, 182. 
Nebraska, 20, 24, 27, 36, 38, 186. 
Nelson, Aimer P., 80, 115, 178-179. 
Nelson, E. W., 23, 32, 194. 
Nevada, 29, 178, 200; see Ruby Lakes 

Refuge, 
nesting : 

nest building, 45, 111-113. 

nest sites and territory, 106-11, 154, 
177. 
New England, 10. 
New Hampshire, 10. 
New Mexico, 20. 
New York Zoological Park. 141, 143, 

165. 
New York Zoological Society, 169. 
Newberry, J. S., 16-17, 28, 194. 
Nice, Margaret M., 100, 110, 194. 
Nicholson, A. J., 164-165, 195. 
non-nesters, 110. 
North American Wildlife Conference. 

174. 
North Carolina, 10-11. 20, 25. 
North Dakota, 20, 24, 37, 129, 186. 
Northwest Company, 13. 
Nute, Grace Lee, 13, 195. 
Nuttall, Thomas, 19, 195. 



Oberhansley, Frank, and Maynard Bar- 
rows, 74, 82, 84, 85, 106, 111, 112, 115, 
117, 120, 122-123, 125, 126-127, 175. 

Ohio, 9, 167, 68, 170, 171. 

Ohio River, 11, 14, 15, 28. 

Olor beivickii, 6 ; see swan, Bewick's. 

Olor buccinator, 6; see swan, trumpe- 
ter. 

Olor coin m-bianu-s, 6, 7 ; see swan, whis- 
tling. 

Olor cygnus, 6 ; see swan, whooper. 

Oregon, 8, 9, 16, 20, 29, 140; see Mal- 
heur Refuge. 

Orton, Alda, 35, 195. 

otter, 132. 

Pacific Railroad Surveys, 16-17. 

pair formation, 94. 

Paleocycnus, 8. 

Paludan, Knud, and J0rgen Fog, 104- 

105, 129, 130, 195. 
parasites, 141-143. 
Parmalee, P. W., 9, 195. 
Patrick, R. W., 82, 195. 
Philadelphia Zoological Garden, 143, 

169. 
Pilder, Hans, 12, 195. 
Pirnie, Miles D., 126, 170, 171-172. 
population, 7, 61, 144; dynamics, 145- 

164. 
Poulsen, H., 78, 79, 85, 97, 100, 114-115, 

195. 
predation, 56. 

avian, 131-134. 

mammalian, 131. 133, 134-135. 
on eggs, 130-131. 
control at Yellowstone, 174. 
prenesting behavior, 99-100. 
Puget Sound, 16, 27. 

Railroad Ranch, 56, 57, 60, 61, 66, 186. 
range of trumpeter. 6. 26, 39; see dis- 
tribution. 
Red Rock Lakes area, 21-22, 160, 162, 

169-170. 173-174. 
Red Rock Lakes Migratory Waterfowl 
Refuge : 

description of refuge, 2, 38, 40-49, 

51, 102. 
establishment 1, 28. 166, 174. 
map, 40. 



212 



INDEX 



Red Rock Lakes Migratory Waterfowl 
Refuge — Continued 
trumpeters at : 

banding, 28. 1S2-1S6. 
behavior, 78-79, 80-81, 82-83, 

92. 
breeding and nesting, 20, 21, 24, 
28, 40-49, 95, 98, 99, 103-104, 
106, 107-108, 110, 111-113, 
114, 115-117, 119-120, 154, 
174. 
cygnets, 119-121, 123-124, 131- 

133. 
description, 65, 72-73. 
food, 123-124, 127-128. 
limiting factors. 131-133, 134, 
135, 136, 137, 138, 140, 142, 
143. 
management, 176-178, 184-186. 
migration, 26, 27. 
population, 145, 146, 147, 150, 

151-155, 162, 199. 
predation, 130, 132. 
transplanting, 29, 178, 180. 
wintering, 55, 58, 59, 61, 88, 138, 
139, 187. 
renesting, 116. 
Richardson, John, 14, 19, 23, 27, 64, 65 ; 

see Swainson. 
Roberts, T. S., 12, 24, 169, 195. 
Rogers, Edmund B., 2, 175, 205-206. 
Rooney, James, 2. 

Ruby Lake National Wildlife Refuge : 
banding, 182-183. 
breeding age, 96. 
disease, 141. 
population, 147, 200. 
predation, 133. 

transplanting, 29, 178-179, 181, 186. 
Russia, 22. 

Salter, Robert, 30, 195. 

Saskatchewan, 27. 

Saunders, Aretas A., 24, 195. 

Scott, Peter, and James Fisher, 73, 81- 

82, 118-119, 195. 
Sharp. Ward M., 115, 117, 120. 132-124, 

132, 134, 140, 142, 175, 195, 205. 
Sibley, Charles L., 172, 195. 
Silloway, P. M., 24, 195. 
Simon, James R., 120. 195. 
Skinner. M. P., 21, 24, 172, 196. 
skunk, 133. 



Smith, H. M., 21. 

Smith, H. S., 164, 196. 

Smith, Stuart G., and Eric Hosking, 82, 

196. 
Smithsonian Institution, National Mu- 
seum, 2, 3, 23. 
Snake River, 30, 55, 56-57, 59, 128, 136, 

176, 186. 
Soper, J. D., 23, 37, 196. 
South Dakota, 13, 27, 186. 
Spencer, David, 33, 114, 116. 
Sperry, Charles S., 21, 24, 175, 202-203, 

204. 
starvation, from freezeup, 139. 
Stejneger, Leonard, 65, 67-68, 196, 198. 
Stewart, Robert E., and Joseph H. Man- 
ning, 7, 126, 129, 196. 
Stowe, Leland, 83, 196. 
Suckley, George, 16, 24, 27, 196. 
Swainson, William, and John Richard- 
son, 14, 27, 196. 
swan : 

description, swans of the world, 

198. 
legend and tradition, 4-5. 
Bewick's, (i, 7, 14, 62, 100. 
black, 5, 6, 70, 116, 129. 
black-necked, 5, 6, 70, 86, 116. 120. 
mute : 

behavior, 82, 94. 
hill, 5. 

Classification, 6. 
description, 5-6, 62. 
egg-laying, 114. 
escape-distance, 78-79. 
feral, 62, 187. 
food, 126, 129. 
hybridization. 172, 188. 
incubation, 114, 115. 
mating fidelity, 97-98. 
territory, 100-101, 104-105. 
transplanting, 187. 
trumpeter (topics not listed in 
Contents) : 
accidents. 143. 
hill, color, 65 : size. 67-68. 
botulism. 141. 

captivity, raised in. 123-124. 
carrying young. 86. 
comparison with whistling 

swan. 7, 10, 14. 64-69. 
cygnets : see cygnet, trumpeter. 






INDEX 



213 



swan — continued 

trumpeter (topics not listed in 
( Contents ) — continued 

disease, 140-141. 

drinking, 85, 86. 

eggs, see eggs, trumpeter. 

head attitude, 86. 

hybridization, 172, 188. 

family ties, 96-98, 117-120, 
121-122. 

foot, attitude, 86, 121 ; deform- 
ities, 140. 

hunting, 22. 135-137, 166. 

icing of plumage, 139-140. 

illegal kills, 136-137, 142, 204- 
205. 

molt, 72-74. 

plumage shaking, 84, 85-86. 

predation, 130, 132, 134-135, 
174. 

productivity, 148, 149, 152, 153, 
157, 158 : .sec population dy- 
namics. 

resting, 84-85. 

sex differences, external, 86, 
120. 

starvation, 139. 

submerging and diving, 84. 

tail feather count, 65. 

trachea, 14, 67. 

weather and mortality, 138- 
140. 
whistling : 

comparison with trumpeter, 7, 
10, 14, 64-69. 

description, 6, 7, 63-64. 

disease, 141. 

distribution, 7, 33, 60. 

eggs, 19. 

escape-distance, 78. 

hybridization, 172. 

parasites, 141-142. 

population, continental, 7. 

trachea, 14, 66. 

wintering, 60-61. 
whooper : 

behavior, 87, 89, 94. 

cygnets, 118-119. 

description, 6, 14, 62. 

distribution, 6-7. 

escape-distance, 78-82. 



swan — continued 

whooper — continued 

flight, 76. 

food, 129. 

hybridization, 172. 

incubation, 115. 

interspecific tolerance, 100. 

molt, 73. 

territory, 100, 102. 
swan skins, trade in and commercial 

use, 12, 13, 19. 
Swann, H. Kirke, 143, 196. 

territorialism, 99, 100-110. 

Texas, 20, 21. 

Thomas, E. S., 9, 24. 

Thompson, Ben, 21, 35, 131, 173, 174, 

196, 205. 
Thwaites, R. G., 9, 16, 24, 27, 196. 
Ticehurst, Norman F., 5, 196. 
Toll, Roger, 174, 175. 
Townsend, John Kirk, 27. 
transplanting, 29, 178-182 ; see Malheur, 

National Elk, Ruby Lake Refuges. 
Traylor, Melvin, 23. 
Trade, Sam A., 3, 24, 31. 
tuberculosis, avian, 140-141. 

Utah, 35, 169. 

Virginia, 20. 

Ward, Peter, 71, 142. 

Washington, 20. 24, 28, 29, 33 ; see Pu- 

get Sound. 
Washington, DC, 168-169, 170. 

National Zoological Park, 169, 182. 
Watson, W. Verde, 105-106, 197. 
Watterson, W. H., 94, 197. 
weather and mortality, 138-140. 
Weeks-McLean Law, 165. 
Weiser, C. S., 76, 197. 
Wetmore, Alexander, 6, S, 198, 197. 
Wetmore. Cecil. 3, 169. 
Widmann. Otto, 12, 24, 197. 
Wilke, Ford, 7, 197. 
Wilson, C. P., 13. 
winter — 

counts, 60-61. 

management. 177-178, 179, 185-186. 

range, 10, 12, 15, 16-17. 20, 27-28, 
29-33, 36, 54-55, 144. 

maps, 26, 39, 52. 



214 



INDEX 



Wisconsin, 20, 24, 38, 141, 142. 

Witherby, H. F., et al., 6, 76, 82, 89, 
100-101, 114, 115, 122, 126, 197. 

Woburn, Abbey, 97, 168. 

Wood, J. C, 20, 24, 197. 

Wood, Norman A., 21, 197. 

Wright, George, 21, 35, 131, 173, 174, 
175, 197, 205. 

Wyoming, 2, 20, 24, 28, 35, 54, 55, 136, 
144, 166, 200; see Jackson Hole, Na- 
tional Elk Refuge, Yellowstone Na- 
tional Park. 

Yarrell, William, 14. 
Yellowstone National Park : 

description of swan habitat, 38, 48- 
53. 

breeding, 20, 24, 35, 48-52. 
wintering, 55, 58-59, 61. 
map, 52. 



Yellowstone National Park — Continued 
trumpeters in : 

behavior, 79, 82, 99, 105, 121. 

description, 70, 71, 72. 

food, 122-123, 125, 126-127. 

hatching, 115-116. 

illegal kills, 136. 

interspecific tolerance, 106. 

management, 174-176; recom- 
mendations, 184. 

molt, 74. 

nesting, 99. 

population, 146, 147, 155-160, 
161, 162-163, 172, 173, 200. 

predation, 105, 130-131, 132, 
134-135, 174-176. 

protection, 165. 

studies, 172-176. 

weather and mortality, 138- 
139. 
Yellowstone region, 21. 
Yocom, C. F., 24, 197. 



U. S. GOVERNMENT PRINTING OFFICE : 1960 O -469660 



PELAGE AND 
SURFACE TOPOGRAPHY 

I OF THE 

NORTHERN FUR SEAL 




NUMBER 64 



UNITED STATES 

DEPARTMENT OF THE INTERIOR 

FISH AND WILDLIFE SERVICE 



PELAGE AND 
SURFACE TOPOGRAPHY 

OF THE 

NORTHERN FUR SEAL 



By 

Victor B. Scheffer 

Biologist, Branch of Marine Mammals 
BUREAU OF COMMERCIAL FISHERIES 




NUMBER 64 



UNITED STATES 
DEPARTMENT OF THE INTERIOR 

Stewart L. Udall, Secretary 

FISH AND WILDLIFE SERVICE 

Clarence F. Pautzke, Commissioner 

BUREAU OF COMMERCIAL FISHERIES 

Donald L. McKernan, Director 




Boston Public Library 
Superintendent ot Documents 

1962 



North American Fauna, Number 64 



Published by U.S. Fish and Wildlife Service 
February 1962 



United States Government Printing Office • Washington • 1961 



For sale by the Superintendent of Documents, U.S. Government Printing Office 
Washington 25, D.C. - Price $1 



CONTENTS 

Page 

Abstract v 

Introduction 1 

Previous research 1 

Methods 2 

General structure of the body covering 5 

Arrangement of the body layers 5 

The skin: epidermis, dermis, and sweat glands 5 

The pilosebaceous unit: follicle, root and shaft of the hair, 

and sebaceous glands 7 

The pelage 9 

Fetal stages (sexes lumped) 10 

Black pup, newborn (sexes lumped) 14 

Synopsis of color pattern 14 

Synopsis of pelage fibers 14 

Guard hairs: larger examples 16 

Guard hairs: smaller examples 17 

Underhairs 17 

Black pup, molting (sexes lumped) 18 

Silver pup (sexes lumped) 19 

Synopsis of color pattern 19 

Synopsis of pelage fibers 19 

Guard hairs: larger examples 20 

Guard hairs: smaller examples 21 

Underfur fibers 22 

Yearling, pelagic (sexes lumped) 22 

Yearling, autumn (sexes lumped) — 23 

Three-year-old, adolescent male (bachelor) 23 

Three-year-old, adolescent female (young cow) 23 

Adult male (bull) 24 

Adult female (old cow) 24 

Variation in length of pelage fibers with age and sex 26 

Variation with season: the annual molt 26 

First molt 26 

Second molt 27 

Third molt 28 

Fourth molt 29 

Molt in adults 30 

Comparison with molt in other f urbearers 31 

The sensory vibrissae 32 

Prenatal development of the vibrissae 33 

Postnatal development of the vibrissae 34 

Pelage anomalies 35 

Color anomalies 35 

Effect of diseases, parasites, and physiological disorders 

on pelage 36 

Effect of sex abnormalities on pelage 39 

Foreign growths 39 

The Pribilof sealskin industry 40 

History of the industry 40 

Killing, skinning, blubbering, and curing 41 

Processing and marketing 42 

in 



IV CONTENTS 

The pelage — Continued 

The Pribilof sealskin industry — Continued Page 

Dimensions and weights of sealskins 45 

Strength and durability of sealskins 50 

Other features of the surface topography 51 

Features of the head 51 

Nostrils, mouth, and lips 51 

Eyelids, eye glands, and iris 51 

Ears 52 

Features of the belly 53 

Mammary gland complex 53 

Penial opening and scrotum 56 

Female external genitalia 56 

Navel and tail 56 

Features of the limbs 57 

Flippers and claws 57 

The blubber layer 59 

Summary 61 

Literature cited 65 

TABLES 

Table 1. Length and weight of male fetal seals 71 

2. Length and weight of female fetal seals 71 

3. Mean lengths of underfur and guard-hair fibers 72 

4. Length of longest vibrissa, by age and sex 72 

5. Change in color of mystacial vibrissae, with age 73 

6. Sizes of grading boards for raw, salted skins 73 

7. Sizes of male sealskins taken in early season 73 

8. Sizes of male sealskins taken in late season 74 

9. Weights of female seals, early and late summer 74 

10. Weight of fresh, male sealskin with relation to field 

length of seal 74 

11. Trade classification of raw, salted, male sealskin with 

relation to field length of seal 75 

12. Trade classification of finished, dyed, male sealskin with 

relation to field length of seal 76 

13. Trade classification of raw, salted, male sealskin with 

relation to over-all dimensions 78 

14. Trade classification of finished, male sealskin with rela- 

tion to over-all dimensions 78 

15. Length of ear from notch, by age and sex 78 

16. Yield of oil from fur seals 79 

APPENDIXES 

Appendix A' — Color notes 81 

Silver pup, male 82 

Silver pup, female 83 

Yearling, autumn, male 84 

Yearling, autumn, female 84 

Three-year-old, adolescent male (bachelor) 85 

Three-year-old, adolescent female (young cow) 86 

Adult male (bull) 86 

Adult female (old cow) 87 

Appendix B< — Glossary 89 

PLATES 
Plates 1-112 95-206 



ABSTRACT 



The midsummer population of northern fur seals, Callorhinus ursinus, 
is estimated at 1,978,000. Of this number, 1,800,000 or 91 percent, origi- 
nate on the Pribilof Islands. The Pribilof herd is capable of yielding 
80,000 to 100,000 sealskins a year. 

The pelage of the adult seal is composed of clearly defined bundles, 
each with a coarse guard hair and 35 to 40 fine underfur hairs; there 
are more than 300,000 fibers to the square inch. Each guard hair is 
accompanied by a sweat gland and two large sebaceous glands. Area 
of the haired surface of the body of the adult male is about 2.5 times 
that of the female. 

The pelage of the pup resembles that of certain land carnivores in 
having small, scattered bundles, each containing 1 to 3 fibers, some of 
the fibers being underhairs and some overhairs (guard hairs). The 
first molt, from black birthcoat to silvery, adult-type molt, occurs about 
mid-September, the second in August of the following year (on the 
yearling), the third in September of the following year (on the 2-year- 
old), the fourth and subsequent molts in late September or October. 
The molt in the adult takes 4 or 5 months. Molting has little effect on 
the commercial value of a sealskin, provided the skin has been taken 
before September. 

Dominant color of the adult pelage is light brownish gray ; most seals 
are darker on back and chest, lighter on belly, throat, and sides. Color 
patterns of the sexes are indistinguishable up to age 2 or 3 years ; color 
patterns of seals from American and Asian waters are indistinguishable. 
Colors are brighter (less brownish) in winter when the seal is at sea 
and has completed its autumnal molt. 

In addition to mutant color phases such as albino, piebald, and choc- 
olate, one may see atrichia, pediculosis, pachyderma, and other skin 
disorders ; and foreign growths, including marine algae and barnacles, 
on the guard hair. 

The flippers are naked. The only functional claws — used exclusively 
for grooming the pelage — are on the middle three digits of each hind 
flipper. The blubber on the fur seal is thinner than on phocids or hair 
seals. From a fur seal weighing 66 pounds about 0.6 gallons of blubber 
oil can be rendered. 




Fbontispiece. — Fur seals on breeding grounds, 5 July, shortly before height of 
pupping season. Harem bull in background ; cows and newborn pups in fore- 
ground. Adults are in old pelage, about ready to molt. (2824) 



INTRODUCTION 

The northern fur seal, Callorhinus ursinus (L.), breeds on islands 
of the North Pacific Ocean and adjacent seas. The midsummer, or 
maximum seasonal, population is estimated at 1,978,000 animals; of 
this number, about 1,800,000, or 91 percent, originate on the Pribilof 
Islands of Alaska. Since 1867 the United States Government has 
acted as custodian of the Pribilof herd and has regulated the taking 
of sealskins for market. In 1958 the Pribilof herd produced 78,919 
skins. It is quite certainly capable of producing 80,000 to 100,000 
skins a year, the quantity depending partly on man's selection (by age 
and sex) of the animals to be cropped and partly on natural fluctuation 
in birth rate and mortality of seals. 

During its 90-year regime, the Government has sought increasingly 
to understand the zoology of the fur seal, the better to manage the 
seal population as a national and international resource. However, 
with regard to the fur-seal pelage — the basis of production— no sus- 
tained effort to obtain zoological information was made until recent 
years. In 1940, Government biologists concerned with management 
research on the seal herd began to collect specimens and field notes, 
looking toward a report on the growth and replacement of pelage 
fibers. The present paper describes certain gross and microscopic 
aspects of the pelage in relation to sex, age, and season of year. It 
also describes, in a cursory way, other features of the surface topog- 
raphy such as flippers, ears, tail, blubber, and mammary glands. 
These features of pinniped anatomy are seldom preserved for study, 
and when they are preserved they tend to lose their original shape 
and color. It has seemed desirable, therefore, to describe and illustrate 
certain appendages, soft parts, and subdermal layers of the body 
covering in their natural condition. The scope of the present work is 
indicated by the table of contents. 

For advice in my research, I am grateful to many persons; I would 
mention especially Ford Wilke, Chief, Marine Mammal Research, 
Bureau of Commercial Fisheries, and Dr. George F. Odland, Clinical 
Instructor, University of Washington Department of Anatomy. 

Previous Research 

World literature on the subject of mammalian hair, especially the 
hair of man, domestic animals, and wild furbearers, is voluminous. 
On the structure, growth, and replacement of fur-seal pelage fibers, 



2 INTRODUCTION 

however, no scientific literature exists, partly because fur seals do not 
inhabit North Atlantic waters and have not been available to European 
zoologists for study. There are, to be sure, a number of fleeting 
references to fur-seal pelage and many popular and scientific accounts 
of the general biology of the seal. It may be helpful at this point 
to list the more important papers that have contributed a background 
to our understanding of fur-seal pelage. These papers will be 
referred to later and individually, and will be described more fully 
under Literature Cited : 

Abegglen and others (1956-58), progress reports of investigations on the 
Pribilof Islands; Baker (1957), general account of Pribilof industry; Barthol- 
omew (1951), observations of living seals; Bartholomew and Hoel (1953), 
breeding habits; Bowker (1931), leather; Clegg (1951), blubber; Fortune 
(1930), sealskins; Fouke (1949), sealskins; Fouke Fur Company (1958), popular 
account of sealskin industry; Fur Trade Review (1916), sealskins; Jordan and 
others (1898), comprehensive account; Mathur (1927), leather; Minato (1949), 
blubber; Miyauchi and Sanford (1947), blubber; Partridge (1938), leather; 
Pearson and Enders (1951), reproduction; Rand (1956), general biology of the 
South African fur seal Arctocephalus pusillus; Scheffer (1949 and later), 
various reports touching on pelage and blubber; Scheffer and Kenyon (1952), 
general account; Scheffer and Wilke (1953), growth data; Stevenson (1904), 
sealskins; Stoves (1958), northern and southern fur seals Callorhinus and 
Arctocephalus; Taylor and others (1955), comparisons of Asian and American 
fur seals; Terao (1940), leather; Thompson (1950), general account of 
Pribilof industry; U.S. Bureau of Fisheries (1916, 1917, 1922, 1938), sealskins 
and blubber; U.S. Fish and Wildlife Service (1952-57), sealskins and blubber; 
Wilber (1952) , freak blubber. 

Methods 

Specimens were collected on the Pribilof Islands, at sea off the 
American coast between California and Alaska, and at sea off Japan. 
Land specimens were taken in summer and fall, pelagic specimens in 
spring and summer, all between 1940 and 1959. Up to about 1949, the 
age of an individual seal was estimated from body size ; thereafter from 
tooth-ridge counts (Scheffer, 1950 a). It was determined directly 
when the seal happened to be wearing a metal tag (Scheffer, 1950 b). 
Government biologists have tested various methods of marking indi- 
vidual seals for study purposes. The accepted method, now being 
applied to 50,000 seals a year, is to fasten a corrosion-proof, individ- 
ually numbered metal tag to one of the flippers of the pup. In the 
past, as many as 10,000 seals a year were marked by a hot-iron brand 
which left a rectangular patch of the skin permanently denuded. 
Quick-drying, synthetic-base "traffic" type paint in yellow, blue, or 
white, applied with a swab, has been used to mark individuals tem- 
porarily in summer. Chemical depilatories have proved to be of no 
practical value, since they penetrate with great difficulty the dense, 
2-layered pelt of the seal. 



METHODS 3 

Forty fetuses were examined, of which 25 were selected as showing 
critical features of growth. About 200 pelts, mostly of known-age 
seals, were preserved by tanning and were later studied in the National 
Museum collection. Photographs (mostly at scale y 16 ) were taken of 
149 of these pelts, and hair measurments were made of 114. In Sep- 
tember 1958, on St. Paul Island, I made a special collection of bits of 
skin, in formalin, from neck, back, and belly of 76 seals of assorted 
age and sex. 

More than a million seals have been killed commercially on the 
Pribilof Islands since 1940, and their pelts have provided clues not 
only to the procession of molt in autumn but also to the incidence of 
freaks and diseased individuals in the population as a whole. 

On several occasions, starting in 1952, fur seals were brought from 
St. Paul Island to the Seattle Zoo. Here they were held in a large, 
outdoor, fresh- water pool and were subjected to shearing experiments. 
Observations were subsequently made of the rate of regrowth of under- 
fur and guard hair. (For one reason or another, the schedule of 
observations was often interrupted.) The total number of seals 
marked by shearing was eight. 

With the exception of color notes recorded in the field, studies of 
pelage were carried out in the Seattle office of Marine Mammal Re- 
search. As will be explained, the Munsell system of color notation was 
used. Under the direction of Dr. George F. Odland, median sections 
of skin, stained with haemotoxylin and eosin, were prepared by Mr. 
James Rankin. Slides of horizontal sections were prepared by tech- 
nicians of the General Biological Supply House. At one time or an- 
other, certain devices and techniques described by the following hair 
specialists were used for the present study : 

Carter (1939), horizontal sections and follicle populations ; J. I. Hardy (1935), 
cross sections of hairs by special tool; Hardy and Plitt (1940), casts of 
cuticular scales in plastic media; Mathiak (1938), cross sections of hairs by 
razor blade; Stoves (1958), many aspects of fiber microscopy; Wildman (1954), 
many aspects of fiber microscopy. 



GENERAL STRUCTURE OF 
THE BODY COVERING 

Arrangement of the Body Layers 

The layers of the body covering of the fur seal, from the outside in, 
are (1) the hairy coat or pelage, (2) the skin proper (the leather of 
the tanned pelt), (3) the panniculus adiposus or blubber, (4) the 
panniculus carnosus or discontinuous, fleshy sheet of muscle beneath 
the blubber, and (5) the tela subcutanea or loose, thin, whitish, con- 
nective tissue which binds the skin to the muscles and bones of the 
body. The layers are shown in plates 2-8. 1 (Leather technologists 
commonly use the term "epidermal area" in a collective sense for all 
the strata down to and including the sweat glands, and the term 
"corium" for the deeper, fibrous strata. Roddy (1956), for example, 
has observed that in most commercial fur skins there is no marked 
distinction between "epidermal area" and "corium.") 

The area of the haired surface of the body (that is, the surface ex- 
clusive of flippers) has been measured on two tanned pelts, as follows : 
Adult male, length of pelt 199 cm., area 1.57 sq. m. Adult female, 
length of pelt 119 cm., area 0.62 sq. m. In this sample, the male pelt 
has an area 2.5 times that of the female. Scheffer and Wilke (1953, 
p. 145) had previously concluded that the adult male outweighs the 
female about 4.5 times. 

To the student of pelage, the two most interesting parts of the body 
covering are the skin proper (epidermis, dermis, and sweat glands) 
and the pilosebaceous unit (follicle, root and shaft of the hair, and 
sebaceous glands). I shall discuss first the skin and second its out- 
growth—the hair. (For definitions of technical terms, the reader is 
referred to the glossary in appendix B, or to medical dictionaries.) 

The Skin: Epidermis, Dermis, and Sweat Glands 

Sections of skin from the back region of two 7-year-olds have been 
studied in detail (plates 9 and 10). The epidermis is not over 60 
microns thick, much thinner than the epidermis of the harbor seal, 
reported by Montagna and Harrison (1957, p. 83) as 0.5 to 1 milli- 
meter thick over the entire body. In the fur-seal epidermis, the 

1 Plates follow page 93. 



6 GENERAL STRUCTURE OF THE BODY COVERING 

stratum corneum is about 15 microns thick, appearing in 4 to 8 layers, 
more or less shattered, on prepared slides. It is sharply distinct from 
the underlying stratum malpighii which is 18 to 35 microns thick. 
The stratum malpighii consists of a superficial layer 1 to 3 cells deep 
in which a stratum granulosum and a stratum spinosum cannot be 
distinguished, and a stratum germinativum 1 cell thick. The cells of 
the stratum germinativum are more or less columnar, deeply staining ; 
those of the superficial layer are cuboidal or flattened, faintly staining. 

In life, the dermis is 3 to 4 mm. thick. While the dermis, in 
general, is thicker in adult animals than in young, and in males than 
in females, its follicular (hair root) portion does not vary appreci- 
ably in thickness. On 12 slides selected as showing true median 
sections with minimum distortion, there is surprisingly little dif- 
ference in depth from surface of skin to base of deepest hair follicle, 
measured at right angle to surface of skin. In a group of 2-year 
males, 2-year females, over-10-year females, and one old bull, the 
range in thickness (depth) was 2.0 to 2.8 mm., with an average be- 
tween 2.3 and 2.4 Tanned and buffed as "Alaska sealskin," the leather 
of a subadult male is less than 1 mm. thick ; tanned as saddle leather, 
the skin of a bull seal is about 4 mm. thick. 

An apocrine sweat gland is associated with each guard-hair follicle 
(plate 11). The gland is sinuous and unbranched. A true median 
section may expose 50 of its loops. The secretory portion of the gland 
originates deep in the dermis, beneath, and to the rear of, the guard- 
hair root. It may start at the 3-mm. level (depth), though more 
often above 2.2 mm. The secretory portion is 1 cell thick. The gland 
is largest at about the 1.4-mm. level, where a cross section may meas- 
ure 80 by 120 microns; it begins to disappear at levels between 1.0 
and 0.8 mm. Here it is replaced by the more superficial massed bulbs 
of the underfur follicles. The duct of the sweat gland is several 
cells in thickness and represents about one-quarter of the vertical 
depth of the gland, though, much less than one-quarter of the entire 
sinuous length of the gland. The duct rises through the common 
follicular bundle at the right or left side, more or less between the 
guard hair and the underfur fibers. It empties into the pilosebaceous 
funnel at the surface of the skin, above the twin exits of the sebaceous 
glands. Near the surface, the sweat-gland duct has a lumen 10 to 
15 microns wide. Here the duct is invested in a heavy epidermal 
sheath. 

Early sweat glands, nebulous and more deeply staining than ma- 
ture ones — certainly not functional — can be seen at levels between 
1.2 mm. and 1.6 mm. below the surface of the skin of the back of 
the neck of a full-term fetus. The sweat glands of a molting black 
pup, on 1 September, are adult in character. 



the skin: epidermis, dermis, and sweat glands 7 

When a heat lamp is focused on the naked flipper of a freshly 
killed seal, the black epidermis soon begins to blister. Before it does 
so, droplets appear on the surface of the skin in a fairly regular 
pattern (plate 12). These are assumed to be secretions of the sweat 
glands. 

The survival value of sweat glands beneath the dense pelage of 
the fur seal is not clear. Aoki and Wada (1951, p. 123) confirmed 
that sweat glands are present in the dog, not only in the foot pads 
but also over the body surface covered by hair. Those authors in- 
duced sweating both by drugs and by radiant heat. They con- 
cluded "that the sweat glands in the hairy skin of the dog do not 
participate actively in the central thermoregulatory mechanism, but 
. . . subserve chiefly the protection of the skin from an excessive 
rise of temperature." 

The Pilosebaceous Unit: Follicle, Root and Shaft of the 
Hair, and Sebaceous Glands 

Details of the main body pelage and vibrissae will be given in the 
next chapter. The present account is intended to provide background 
information (plates 13-31). The pelage of the fur seal is made up 
of bundles or tufts of hairs emerging from the surface of the skin 
through a common pilosebaceous funnel and orifice. The hairs are 
flattened and are generally directed hindward and downward, thus 
contributing to the sleek, streamlined profile of the body. Each hair, 
of course, originates in its own follicle. The anteriormost (upper- 
most) hair in each follicular bundle is a coarse guard hair, deeply 
rooted. Next in rank are 35 to 40 fur fibers arranged in stairstep 
fashion, the root of the fiber at the rear of the bundle being nearest 
the surface of the skin. The fur hairs originate separately, converge 
tightly at the level of the upper dermis, and diverge outside the 
body. They rise from the skin at a slope of 40° to 50° from horizon- 
tal. The number of follicular bundles per sq. mm. on skin from the 
back of adults has been estimated at 11 (on suede leather, plate 6), 
at 17 (on another sample of suede leather), and at 15 (on a horizontal- 
section slide, plate 13). Selecting 15 as a reasonable average and 
using 38 fibers per bundle as a factor, it is calculated that there are 
about 570 fibers per sq. mm., or 370,000 fibers per sq. in. 

The hair follicle is considered by most anatomists to be an invagina- 
tion of the epidermis. One can trace the stratum corneum and stra- 
tum malpighii deep into the follicle, almost to its base. At the follicu- 
lar level, the epidermis becomes the outer root sheath. An inner root 
sheath clings for a short distance up the hair root. A connective tissue 



8 GENERAL STRUCTURE OF THE BODY COVERING 

papilla enters the bulb of the root, while surrounding the papilla are 
the matrix cells or germinative cells of the hair. 

The hair consists of a swollen basal bulb and a shaft. The bulb and 
other buried regions of the hair are termed, collectively, the root. The 
shaft is free and is largely outside the body. Listed in order from the 
central axis outward, the shaft consists of a vacuolated medulla (ab- 
sent in fine hairs) , cortex (usually pigmented) , and cuticle (made up of 
overlapping scales). I have been unable to demonstrate arrectores 
pilorum, or hair-erecting muscles, in the skin of the seal. Bergersen 
(1931, p. 170) could find none in the skin of the harp seal Pagophilus. 

The paired sebaceous glands lie along the sides of, and within, the 
common follicular bundle. They attend the guard hair, not the fur 
hairs, although their secretion is shared by all members of the bundle. 
Each gland originates at about the level of the underfur bulbs, or 1.0 
to 0.8 mm. below the surface. The deeper portion of the gland is sub- 
divided into 2 or 3 shallow, roundish, irregular lobes. The upper por- 
tion is a rather smooth dome. At the level of greatest size, 0.6 to 0.4 
mm. deep, the gland may measure 90 by 150 microns. Each gland 
pours its secretion from the top directly against the right or left 
posterior side of the guard hair, at about the 0.2-mm. level. The 
lumen of the duct is 15 to 25 microns wide. 

In a full-term fetus, the sebaceous glands are well developed, up to 
40 by 70 microns in horizontal section, and apparently are functional. 
Most of them are above the 0.5-mm. level. 



THE PELAGE 

The principal aims of this chapter are to describe the gross, as well 
as the microscopic, aspects of the pelage on representative specimens 
ranging in age from fetal to old adult. Since the fur seal exhibits 
only two distinct kinds of pelage— the black birthcoat and the silver 
adult-type coat— emphasis is placed on descriptions of the black pup 
and the silver pup. Color pattern is discussed synoptically ; addi- 
tional notes are given in appendix A. Because they represent special- 
ized, nonmolting hairs, the sensory vibrissae are discussed last. The 
naked, or nearly naked, parts of the body surface are discussed in the 
next chapter. 

To illustrate changes in the pelage during prenatal life, descriptions 
of the hair primordia and hair fibers on 25 selected fetuses are given. 
To illustrate changes in the pelage during postnatal life, descriptions 
of typical individuals in each of the following classes are given (where 
no sex distinctions can be seen, male and female are treated as one) : 

Approximate duration 
Pelage class of this pelage 

black pup, newborn (sexes lumped) 2 weeks {e.g., 15-31 July). 

black pup, molting (sexes lumped) 2 months (e.g., 1 August-30 Sep- 
tember), 
silver pup, persisting as yearling, pelagic 11 months (e.g., 1 October-31 Au- 
( sexes lumped ) . gust ) . 

yearling, autumn (sexes lumped) 13 months (e.g., 1 September-30 

September.) 
3-year-old, adolescent male (bachelor) 1 year (e.g., 1 October-30 Septem- 
ber). 
3-year-old, adolescent female (young cow) __ 1 year. 

adult male (bull) 1 year. 

adult female (old cow) 1 year. 

In tracing the development of the pelage, I have usually omitted 
reference to body size since growth tables have already been published 
by Scheffer and Wilke (1953) and Scheffer (1955). An exception is 
made in the case of fetal specimens. It has seemed useful to give 
the weight of each fetus, since clearly the prenatal age of the speci- 
men from implantation cannot be known. The mean date of im- 
plantation is quite certainly in early November, a date that I have 
chosen from study of the figures in tables 1 and 2. These tables show 
length and weight of 366 fetal seals collected at sea between mid- 

9 



10 THE PELAGE 

January and the end of June. Dr. D. G. Chapman (personal cor- 
respondence) has estimated that, as of 21 November of the preceding 
year, the average fetus would have measured : male, 5.4 cm. and 10 g. ; 
female, 5.1 cm. and 10 g. 

It has seemed useful to give also the relative weight of the fetus, 
or its weight in relation to normal size at birth. SchefFer and Wilke 
(1953, p. 133, 135) measured 39 newborn seals and reported certain 
values. (See bottom row in tables 1 and 2 2 of the present report.) 
These values for mean newborn weight are used as reference points in 
describing the stage of development of the fetus. For example, a 
male fetus of 2.7 kg. is described as "0.50 MNW", or one-half mean 
newborn weight. 

Fetal Stages (Sexes Lumped) 

Early stages, between the autumn blastocyst (a pearly sphere barely 
visible to the naked eye) and the midwinter fetus (the size of a man's 
thumb) are unknown. The height of the mating season is in July. 
Three to four months later, the fertilized egg has become the blastula, 
at which time it implants in the uterine mucosa (Pearson and Enders, 
1951). As just stated, the estimated date of implantation is early 
November. 

Fetus of 23.7 g. (0.0049 MNW), female, 14 February 

This is the smallest fetus available for study (plate 32-A). Most 
of the body is smooth. On forehead and crown there is faint but 
distinct and regular pimpling. Each pimple marks the site of a hair 
primordium beneath the skin surface. With a 5 X hand lens one can 
see, through the translucent epidermis, a dark dot in each pimple. 
This dot represents a concentration of melanocytes. A regular pat- 
tern of dark dots extends along the back, though pimpling of the 
surface has not begun here. Elsewhere than on head and back, faint 
white dots, visible through the epidermis, mark the sites of primordia 
in which the elaboration of pigment has not begun. Collectively, the 
dots on the head impart a gray cast ; the rest of the body is whitish. 
The primordia resemble those in the skin of a 5-month-old human 
embryo (compare Montagna, 1956, p. 180, fig. 5). 

Fetus of 103 g. (0.021 MNW), female, 20 January 

The skin over the entire body, except flippers and other parts 
destined to remain naked, is pimpled. No body hairs have erupted. 
Pigmentation is beginning to show on the flippers in the form of 
extremely fine, scattered, dark specks. It is heaviest at the base of 
each fore and hind claw. It is barely visible on the nostrils. 

2 Tables will be found at pages 71-79. 



FETAL STAGES (SEXES LUMPED) 11 

Fetus of 131 g. (0.024 MNW), male, 25 January 

The forehead, crown, and eyelids are distinctly washed with gray. 
The nostrils are conspicuously gray. No hairs have erupted. 

Fetus of 260 g. (0.054 MNW), female, 19 January 

Extremely fine black hairs have appeared on the face, top of snout, 
around the eyes, and under the chin (plates 34 and 35). This first 
pelage could easily be overlooked if one were not looking for it. The 
hairs on the cheek posterior to the mystacial vibrissae are the longest; 
those under the chin, the smallest. No external hairs can be seen on 
back, tail, or other parts of the body. 

The largest hairs are flattened, 1.5 mm. in length, about 12 microns 
wide along most of the shaft, up to 50 microns in diameter at the flared 
base, heavily pigmented. These are the young, distal portions of 
black-pup guard hairs. 

Fetus of 312 g. (0.058 MNW), male, 6 February 

A few tiny hairs have erupted on the cheeks and above the eyes ; the 
rest of the body is naked. 

Fetus of 372 g. (0.077 MNW), female, 16 February 

This fetus was removed from an adult taken at sea and subse- 
quently held in cold storage for 8 days (plate 36). On a photograph 
of the fetus in storage, the head and flippers appear darker than the 
rest of the body. This seems to represent the beginning of conspicu- 
ous pigmentation, although it may be, instead, dark blood beneath the 
thinner-skinned parts of the body. 

Fetus of 454 g. (0.084 MNW), male, 15 February 

A few hairs on the cheeks only. 

Fetus of 575 g. (0.11 MNW), male, 13 February 

At first glance, a naked fetus, though close inspection reveals fine 
hairs over most of head and throat. The head and flippers are defi- 
nitely darker than the rest of the body. 

A horizontal section from the back of the neck of this fetus is shown 
in plate 37. It exhibits a regular pattern of hair follicles, about 10 
per sq. mm., each with a faintly pigmented hair. Some follicles reach 
a depth of 0.5 mm. Scattered among them, and outnumbering them 
10 to 20 times (depending upon how the count is made), are small, 
dark primordia without hairs. 

The hairs in their present stage of development cannot be identified 
as the tips either of guard hairs or underhairs. Some may, in fact, 
be lanugo hairs destined to be shed before the birthcoat is complete, 
4 to 5 months hence. Quite certainly each marks the site of a perma- 

553006 0—62 2 



12 THE PELAGE 

nent follicular bundle. And certainly additional bundles will appear, 
since the bundles of the birthcoat are four times more abundant than 
are the hair follicles of the present fetus. (Compare plates 37 and 
46.) 

The small, dark primordia shown in the fetus of plate 37 are dis- 
tributed singly or in clusters of 1 to 4. When in clusters, they are 
aligned with the long axis of the body, the anterior primordium being 
the largest. I do not know what these primordia represent. They 
are most likely very early stages of the birthcoat underhairs, or se- 
baceous glands (which will appear in the birthcoat), or a combination 
of the two. 

Fetus of 580 g. (0.11 MNW), male, 31 May 

This individual may have implanted very late in spring for, though 
taken in late May, it resembles a February fetus. Vellus over entire 
body, except palmar and plantar surfaces of flippers ; well developed 
only on head. The heaviest pelage is on each cheek posterior to the 
mystacial vibrissae. Here the effect is of a smoky gray wash on the 
side of the face. The blacker, heavier hairs are young guard hairs 
up to 40 microns wide at the base, distinctly medullated. The vellus 
fibers are young underhairs. 

Fetus of 595 g. (0.12 MNW), female, 6 February 

Hairs barely visible over body ; head well haired. 
Fetus of 660 g. (0.14 MNW), female, 15 February 

Hairs over entire body (plate 38-A) . 
Fetus of 1.09 kg. (0.20 MNW), male, 25 March 

Vellus has appeared on ear tips, giving a grayish cast. 
Fetus of 1.19 kg. (0.22 MNW), male, 30 March 

The pelage of the head has extended to the extreme tip of the 
snout. On the upper surface of the fore flipper, one can clearly dis- 
tinguish the haired (proximal) and nearly naked (distal) surfaces. 
A 5X lens, however, reveals fine vestigial hairing over the distal por- 
tion, destined in the adult animal to become naked. The bases of 
the mystacial vibrissae are hidden in the dense pelage of the face. 

Fetus of 1.11 kg. (0.23 MNW), female, 23 March 

The ears are well haired and gray along their full length. The 
blades of the guard hairs, which contribute virtually all of the black- 
ness of the birthcoat, are now emerging from a background of young, 
whitish underhairs. On the palmar and plantar surfaces of the flip- 
pers, especially near their edges, one can discern a few fine hairs, 
destined to disappear at birth. 



FETAL STAGES (SEXES LUMPED) 13 

Fetus of 1.23 kg. (0.23 MNW), male, 27 March 

The ears are thinly haired, whereas on a smaller specimen the ears 
are thickly haired. 

Fetus of 1.42 kg. (0.26 MNW), male, 22 March 

The only conspicuous pelage is that of the crown and face. 

Fetus of 1.45 kg. (0.27 MNW), male, 11 August 

Prematurely born, found dead on St. Paul Island. The face, crown, 
and ears are very dark gray. The upper surface of the fore flipper is 
covered with distinct vellus, most of which will be retained up to the 
normal time of birth. 

Fetus of 1.45 kg. (0.27 MNW), male, 25 March 

Although of same weight as the preceding one, this fetus is 39 cm. 
in length as against 37 cm. (plate 39). A faint gray streak has ap- 
peared along the back and around the base of the tail, marking the first 
appearance of regularly spaced, coarse, dark guard hairs. 

Fetus of 1.70 kg. (0.35 MNW), female, 12 April 

The dark streak has spread along the back and rump. Growth has 
been backward from the head and forward from the tail, leaving 
an area on the back where the streak is less prominent. The face is 
now handsomely marked with dark and light zones (plate 41-A). 

Fetus of 1.93 kg. (0.36 MNW), male, 2 April 

Generally speaking, the fetus is gray (plate 38-B). The head and 
base of tail are dark gray. The pattern of dark guard hairs against 
lighter underhairs, giving the effect of a dark wash, has spread down- 
ward from the back to the flanks. 

Fetus of 2.21 kg. (0.41 MNW), male, 2 May 

The dark effect caused by coarse guard hairs has spread to the under 
part of the body only at chin and throat (plate 40). The pelage has 
been slowest to develop on the posterior region of the chest, between 
the fore flippers. A few white-tipped guard hairs show on the 
cheeks and sides of the head, behind the ears. Whereas many mam- 
mals are marked with a dark streak along the back, the fur seal 
exhibits a dark streak for a short time only, while the fetal guard 
hairs are erupting. 

Fetus of 2.27 kg. (0.47 MNW), female, 21 April 

The dark guard hairs are approaching the chest. 

Fetus of 2.72 kg. (0.50 MNW), male, 21 April 

A fur seal delivered on the the breeding ground at this stage would 
probably survive. (An aborted fetus of 1.59 kg, was seen alive on 16 



14 THE PELAGE 

July and was picked up dead on the following day.) At first glance, 
the fetus would be called black. The black guard hairs are now 
distributed over the entire body, though thinly on the chest, which 
remains light colored. White hairs surrounded the penial opening, 
especially its posterior margin. 

Fetus of 2.44 kg. (0.51 MNW), female, 21 April 

A transverse zone of gray persists on the posterior part of the chest, 
between the flippers (plate 42) . Coarse, black guard hairs are show- 
ing for the first time here, starting along the midventral line. 
Fetus of 3.43 kg. (0.71 MNW), female, one of twins, 9 May 

All black, well haired, with a sprinkling of white hairs on sides of 
neck and throat, on posterior part of belly, and in armpits (plate 43). 
Vellus persists on the dorsal surface of the fore flipper. 

Black Pup, Newborn (Sexes Lumped) 
SYNOPSIS OF COLOR PATTERN 

At first glance, the pup appears to be all black (plate 44). Above, 
the coat is glossy black with a few scattered white hairs on forehead 
and on neck behind ears. Corners of mouth may be stained brown- 
ish by bile. Below, black, though stained brownish soon after birth ; 
slightly paler (very dark gray) on posterior region of belly. Scat- 
tered white hairs on throat and along lower lip, white crescentic spot 
at each armpit ; white spots about 1 cm. in diameter at sites of 4 mam- 
mary teats in both sexes, at penial opening, and at ventral margin of 
anus. 

The hair slope or "set" of the hair is hindward and downward from 
the snout at all stages of development of the fur seal. In related 
species, however, the birthcoat may have an attractive moire pattern. 
I have examined tanned pelts of the newborn Steller sea lion Eumeto- 
pias jubata exhibiting a hair pattern somewhat like that of lamb, kid, 
or pony. Photographic reproductions by Samet (1950, p. 356) show 
that the pelt of the newborn South American sea lion Otaria byronia, 
the "tropical seal" of the fur trade, also has a rippled pattern. 

SYNOPSIS OF PELAGE FIBERS 

In order to make a distinction, I use "underhair" for the fine-fibered 
layer of the juvenile coat and "underfur" for the homologous layer 
of the adolescent and adult coats. 

The black coat is the first pelage. It is mature (prime) at birth 
and can be plucked rather easily with one's fingers. Traces of it per- 



BLACK PUP, NEWBORN (SEXES LUMPED) 15 

sist for 2 or 3 months after birth, that is, from mid-July to end of 
September, by which time it has been replaced by the pelage of the 
silver pup, autumn. 

The black coat is a temporary body covering, quantitatively as well 
as qualitatively different from the adult coat. While both juvenile 
and adult pelages include an overlaver of guard fibers and an under- 
layer of fine fibers, the sizes and proportions of the fibers in the coats 
of pup and adult are materially different. For example, the under- 
hairs of the black coat are so thinly distributed that the newborn 
pup may become soaked to the skin in driving rain, while the underfur 
of the adult coat is dense and water-repellent. 

The black pelage has a light brownish gray basal zone merging 
gradually with a deep brownish black terminal zone. The two zones 
are about equal in width (depth). The lighter effect is contributed 
by the almost colorless underhairs, plus the pale shafts of the guard 
hairs. The darker effect is contributed by the heavily pigmented 
blades of the guard hairs. White-tipped guard hairs are rare. A 
sample area the size of a man's hand, for example, may contain none 
at all. 

The pelage contains 75 to 80 percent underhairs (shorter, finer, 
and more wavy) and 20 to 25 percent guard hairs (longer, coarser, 
and stiffer). Measured as they lie in the pelage, the mean lengths 
of the fibers are shown in table 3. The underhairs are a mixed lot 
(plate 45). Their tips line up in a ragged rank difficult to measure. 
In length, they range from to 15 mm. The finest and most abundant 
ones are fur-like, slender, wavy, without blade, and almost without 
medulla. The largest ones are miniature guard hairs. The under- 
hairs intergrade completely with the smallest guard hairs. Between 
the largest underhairs and the smallest guard hairs, however, there 
is a fairly distinct break in size, though not in shape and structure. 
All of the hairs are attenuated at the root, showing that they have 
ceased to grow. The root tends to be roundish in cross section, while 
the older portions of the shaft are distinctly flattened. All of the 
fibers taper to sharp points — some to less than 1 micron, near the 
limit of resolution. All of the fibers contain brown pigment in vary- 
ing amounts. 

A horizontal section from the back of the neck of a full-term fetus 
is shown in plates 46 and 47 (A and B). The follicular bundles are 
arranged in a honeycomb pattern, about 40 to 45 per sq. mm. Each 
bundle is embedded in a complex web of connective tissue. Each 
bundle is seen as a circle of epidermal tissue (with deeply staining 
nuclei) surrounding 1, 2, or 3 hairs. One's are least common; two's 
are most common. The hair nearest the anterior edge of the bundle 



16 THE PELAGE 

is always the largest and is a guard hair. The hairs nearest the pos- 
terior edge of the bundle may be either underhairs or small guard 
hairs. At superficial levels, to a depth of less than 0.1 mm., the hairs 
are separated from each other within the bundle by a thin, translucent 
corneal layer. Immediately below, the nucleated root-sheaths begin 
to appear. Each hair root, with its sheath, is independent until it 
reaches the common pilosebaceous opening near the surface of the 
skin. The hair roots sink to a depth of about 1 mm. At the posterior 
side of the follicular bundle, a group of deeply staining cells may be 
seen (plate 47-A). Each is the upper part of the structure which 
will, in late summer, become the adult-type underfur bundle of the 
silver pup. At depths of 0.6 to 0.8 mm., the primordia of underfur 
follicles are active (plate 47-B). The follicles are taking shape be- 
tween, and posterior to, the mature underhairs or small guard hairs 
of the black-pup pelage. Among the ordinary pup hairs, one can 
occasionally see a giant guard hair with root sheath up to 250 microns 
in diameter. Such a hair is a "premature" adult-type guard hair. 
In many stages of the fetus, one is able to see an occasional hair of 
this kind. 

GUARD HAIRS: LARGER EXAMPLES 

Slight differences between these and the guard hairs of the adult 
seal will be pointed out on page 20. A sample fiber, length 17.5 mm., 
bends backward at a point about 8 mm. from the tip. The basal half 
is nearly straight and is light gray ; the terminal half or blade is nearly 
straight and is deep brownish black. The fiber is strongly flattened 
except at its root, which is attenuated and roundish in cross section. 

The tip is about 0.8 mm. in length, very sharp. The blade is about 
8 mm. in length and 18 by 157 microns in cross section. The cross- 
section shape is crescentic, with the concave side facing posteriorly. 
The shaft is flattened-elliptical in cross section, 25 by 82 microns. The 
basal region of the shaft is slightly wider, 89 microns. The root is 
25 by 42 microns and tapers toward the base. 

The pigment is brown, distributed in fine, barely visible granules 
in the cortex along most of the shaft, becoming much heavier in the 
blade. Here it is intense brownish black, evenly distributed through 
the cortex, both in grains and in little packets of grains aligned with 
the long axis of the fiber. The tip of the fiber is distinctly pigmented ; 
the root is clear. (The next crop of fibers, the replacement crop, will 
consist of adult-type guard hairs, most of them white-tipped.) 

A medulla is absent from the tip. About 0.8 mm. from the tip, a 
blade with a conspicuous, unbroken (medium wide) lattice t} 7 pe of 
medulla begins to appear. The medulla continues to the basal end of 
the blade, where it becomes broken for a distance of 1.5 mm., then 









BLACK PUP, NEWBORN (SEXES LUMPED) 17 

again unbroken. Where the main part of the shaft is 82 microns 
wide, the medulla is 53 microns wide. The root is without medulla. 

The cuticular-scale pattern at the tip is coronal, irregular, margins 
smooth. It resembles the pattern of the underhair tip. The pattern 
of the blade is waved, irregular, margins smooth to rippled, near. The 
main part of the shaft is diamond petal, margins smooth. 

GUARD HAIRS: SMALLER EXAMPLES 

These are small shield fibers varying in length from 10 to 15 mm. 
As compared with the larger guard hairs, they are more flexible, wavy, 
and slender. They have a shorter, less conspicuous, and narrower 
blade. They usually have a swollen base which the large guard hair 
does not possess, and the medulla tends to be broken rather than un- 
broken. The color of the blade is brown, elsewhere the hairs are pale 
gray. The tip is finer, more attenuated, than on the larger guard 
hair. 

On a typical small guard hair 13 mm. long, the blade is 4 or 5 mm. 
long, with a cross section 16 by 71 microns. The cross section is more 
elliptical, less crescentic, than in larger guard hairs. Cross sections 
of the shaft (main part), shaft (basal region), and root are, respec- 
tively : 16 by 46, 28 by 68, and 18 by 24 microns. Thus, the basal 
region is nearly as wide as the blade. 

Pigment is distributed as in the larger guard hairs, though more 
sparsely. It starts in the tip, is heaviest in the blade, is lightest in 
the shaft, and is absent from the root. 

A medulla is absent from the tip and terminal part of the blade; 
broken (interrupted) narrow in the widest part of the blade; broken 
(fragmental) in the basal part of the blade; gradually increasing 
toward the root until it becomes unbroken (medium wide) lattice, as 
in the shaft of the large guard hair. A medulla is lacking in the root. 

The scale pattern is diamond petal, margins smooth, along most 
of the shaft, as it is on the larger guard-hair shaft. On the blade, 
it is more nearly wide, irregular petal. Little distinction can be 
made between the patterns of a small and a large guard hair. 

UNDERHAIRS 

The following description applies to the abundant fine hairs and 
not to the less abundant coarser hairs which resemble guard hairs. 
The underhairs are strap-shaped, clearly spiral, making 2 or 3 com- 
plete waves (as viewed in one plane), slender, and without blade. 
They range in length from 6 to 10 mm. The terminal one-third of 
the underhair fiber, corresponding to the blade of the guard hair, 
is pale golden brown; the basal two-third nearly colorless. 



18 THE PELAGE 

The tip is long and slender; blade absent; shaft cross section 12 
to 14 by 20 to 22 microns, changing but little in size along its length. 
The basal region, however, is distinctly swollen, to 24 by 40 microns. 
The root is extremely slender, 8 by 10 microns. 

Pigment, in the form of golden -brown granules, flecks, and streaks, 
is plainly visible in the tip in the terminal one-third of the fiber. 
It gradually becomes paler and more diffuse toward the base. A few 
granules can be seen even in the root. 

A medulla is present only in the swollen basal portion of the shaft. 
Here, where the shaft is 36 microns wide, the medulla is 20 microns 
wide, unbroken (medium wide) lattice. This section of medulla is 
joined on both ends by a short section of broken medulla. 

The cuticular-scale pattern is coarse pectinate near the tip, diamond 
petal, irregular, along most of the shaft; diamond petal, regular, on 
the basal swollen part. The scale margins are smooth. 

Black Pup, Molting (Sexes Lumped) 

The coat is brownish black, beginning to pale on face, flanks, and 
belly; coarser and duller than on the newborn (plates 48 and 49). 
Dorsal aspect: generally dark grayish brown (Munsell 5 YR 2/1) ; 
paler on top of snout, upper lip, and flanks. Many hairs on crown 
and back of neck are white or white-tipped. Bases of flippers are 
assuming the deep-brown color of the adult. Ventral aspect: gen- 
erally dark grayish brown (5 YR 2/1) ; paler on lower lip, chest (but 
not throat), and belly. Armpits, around mouth, and around penial 
opening stained dark orange yellow (7.5 YR 6/6) ; posterior region 
of belly stained brown; whitish spots at the mammary teats still 
visible in some specimens. 

A brief description only will be given of the fibers of the transition 
stages between black pup (newborn) and silver pup (plate 50). On 
the back pelage of a female taken on 11 August, weight 9.3 kg. (20i/*> 
lb.) adult- type underfur fibers and guard hairs are beginning to 
appear among the pup hairs. A fur buyer would say that the new 
pelage is starting to "peep," or that the pelt is in the "peepy" stage. 
Many underfur fibers — fine, nonmedullated, and in bundles — have 
moved up to the level of the pup underhair tips. Slightly below 
them, adult guard hairs have moved out in even rank, their pigmented 
blades collectively giving the effect of a distinct dark-gray band at 
the base of the pelage. This band is 4 mm. wide. Where the bases 
of the pigmented blades are still buried in the skin or leather, they 
show as black vertical streaks. 

On the crown of a female killed 22 September, weight 12.2 kg. 
(27 lb.), the shedding black pup hairs contrast sharply with the new 
crop of white-tipped guard hairs. 



SILVER PUP (SEXES LUMPED) 19 

On the back pelage of a male taken on 29 September, weight 14.7 
kg. (32'V£ lb.), the fall molt is nearly complete. In thickness and 
resiliency, as judged by one's fingers, it ranks between the pelage 
of the black pup and the silver pup. Underfur is thin. Adult-type 
guard hairs have almost completely replaced pup overhairs. A few 
unshed pup hairs show like black tree trunks in a forest of light 
underfur. The surface of the pelage now has a pepper-and-salt effect 
imparted by the white tips of the guard hairs. The dark band seen 
in the 11 August specimen has moved out beyond the underfur. The 
leather is paler, now that the guard-hair blades have passed through 
it. While at the surface of the skin no fiber of any kind contains 
pigment, intensely black pigment groups can be seen at the papillary 
level of certain fiber roots, showing through the translucent leather. 
These groups probably represent melanocytes of the forthcoming 
crop of fibers. 

Silver Pup (Sexes Lumped) 

SYNOPSIS OF COLOR PATTERN 

Dorsal parts rather uniform dark gray; face with conspicuous 
"mask"; armpits and rump patches pale (plates 51 and 52, A and B). 
First appearance of rump patches; light-colored streaks, each about 
3 by 8 cm., extending forward from the level of the knee. On ventral 
side, pale belly and anterior region of chest contrast with dark throat 
and posterior region of chest. Underfur silvery gray. Sexes indis- 
tinguishable on basis of color pattern. (Further details are given 
in appendix A. ) 

SYNOPSIS OF PELAGE FIBERS 

From the black coat — its predecessor — the silver coat differs mainly 
in having underhairs, now called underfur fibers, which are longer, 
curlier, paler, and in definite bundles ; guard hairs somewhat shorter 
and mostly white-tipped. From the autumn yearling coat — its suc- 
cessor — the silver coat differs mainly in having underfur fibers which 
are whitish rather than cinnamon-colored, and are straighter, less 
curly. 

A strip cut from the back of the silver pup shows a whitish layer 
of underfur 11 mm. wide, topped by a black band 5 mm. wide (rep- 
resenting the guard-hair blades) and a white band 1 mm. wide (rep- 
resenting the guard-hair tips). The underfur fibers are twisted in 
5 to 7 gentle waves. The pelt has a pepper-and-salt appearance. 
On the living animal, the guard hair does not readily part itself 
to show the underfur. On the black pup, by contrast, a driving 
rain may part the thin black overcoat, revealing the light underhair 



20 THE PELAGE 

as a streak from forehead to rump. Measured as they lie in the 
pelage, the mean lengths of the fibers are shown in table 3. 

The fiber population consists of underfur fibers about 97.5 percent, 
small guard hairs 2.25 percent, and large guard hairs 0.25 percent. 
The underfur fibers are very distinct from the large guard hairs. 
The small guard hairs resemble the underfur fibers in size, the large 
guard hairs in structure. Since the small guard hairs stand alone 
rather than in bundles, are medullated, and have a blade, they are 
clearly to be regarded as guard hairs rather than underfur fibers. 
In a mid-October specimen, all fibers have attenuated roots, though 
they are still growing. 

GUARD HAIRS: LARGER EXAMPLES 

These important hairs are responsible for the over-all effect of 
color, texture, and pattern of the seal's coat. Judging from photo- 
graphic illustrations by Wildman (1954), fur-seal hairs resemble most 
closely those of the American mink (Mustela vison) and th3 Asiatic 
mink or kolinsky (M. sibirica), somewhat less closely those of the ot- 
ters (Lwtra sp.). The large guard hairs of the silver pup and adult 
fur seal closely resemble, in size and structure, the large overhairs of 
the black pup. Differences are minor. For example, the pup over- 
hair tapers more suddenly to a point and is pigmented to the tip ; the 
adult guard hair tapers more gradually and usually has a colorless tip. 
A large specimen is 22 mm. long, flattened, nearly straight in dorsal 
or ventral view, bent backward near the middle in side view. The 
bend is at a point one-third to one-half the distance along the shaft 
from the tip. The terminal one-third of the shaft is widened into a 
blade. The basal half is colorless, the terminal half (except for tip) 
is brownish black. 

The tip is 1.5 mm. in length and very sharp. The blade is 8 to 10 
mm. in length, and has a flattened elliptical (boat-shaped) cross sec- 
tion whose diameters are 35 by 158 microns. The main shaft cross 
section is 32 by 118 microns. The shaft is nearly uniform in width, 
or slightly smaller toward the base. The root is 0.8 to 1.0 mm. long 
and 32 by 55 microns in diameter. 

Pigment is in granules and longitudinal streaks, evenly distributed 
throughout the medulla. It is absent in the terminal 1.5 mm. Toward 
the middle of the blade, it becomes intense brownish black, obscuring 
most other structures. It is absent from the basal half of the hair. 

The tip is nonmedullated. In the blade, the medulla is conspicuous, 
unbroken (medium wide) lattice; 65 microns wide where the blade is 
118 microns. It becomes broken (interrupted) near the middle of the 
shaft, below the blade ; then gradually increases in importance toward 
the base. At the base, it is unbroken (wide) lattice, 85 microns wide 
where the shaft is 1M microns. The root is nonmedullated. 



SILVER PUP (SEXES LUMPED) 21 

The cuticular-scale pattern is similar to that of the small guard hair 
just described. Where the scale margins are slightly crenate to 
smooth on the smaller hair, along tip and blade, they are crenate on 
the larger hair. The transition region at the base, between petal and 
mosaic, is 0.4 to 0.5 mm. in length. 

GUARD HAIRS: SMALLER EXAMPLES 

Small guard hairs are clearly distinguished — for example, by pres- 
ence of medulla and by solitary position — from underfur fibers. (In 
the black-pup coat, it will be recalled, small guard hairs intergrade 
with underhairs.) Small guard hairs outnumber the large ones by 
about 10 to 1, but are less conspicuous because of their inferior posi- 
tion, narrower width, and paler color. A typical small guard hair 
is 14 mm. long, strap-shaped, with 4 or 5 gentle waves, a distinct 
but narrow blade along the terminal one-third of the shaft, and a 
slightly swollen base. It is mainly colorless, with brown pigment 
beginning about 11 mm. from base, becoming deep brownish black 
in blade, paling toward tip, giving the effect of a colorless tip about 
0.2 mm. 

The tip of the hair is very sharp. It increases in size for a distance 
of 0.2 to 0.3 mm. toward the blade. The blade is about 3 mm. long, 
with cross-section diameters 14 by 90 microns ; cross-section shape an 
ellipse flattened on the posterior side. The main part of the shaft 
has cross-section diameters 14 to 18 by 30 to 35 microns. The basal 
one-third of the shaft is thicker than, and nearly as wide as, the 
blade, or 25 by 70 microns. The root is 1.0 to 1.2 mm. long, with 
cross-section diameters 15 by 20 microns. 

Pigment granules and streaks, evenly distributed in the cortex, 
begin about 0.2 mm. from the colorless tip. Pigmentation becomes 
intense, though never as intense as in large guard hairs, in the middle 
of the blade, about 1 mm. down from the tip. Pigment granules 
gradually disappear below the middle of the shaft, The basal half of 
the shaft is colorless. 

The medullary pattern varies with the size of guard hair. In a 
14-mm. hair, the medulla is lacking in the tip. It is broken in the 
blade, disappears in the middle third of the shaft, and becomes con- 
spicuous, unbroken (wide) lattice toward the base. At a place where 
the shaft is 53 microns wide, the medulla is 40 microns. 

The cuticular-scale pattern at the tip is coronal to irregularly waved 
mosaic, margins smooth (compare Wildman, 1954, fig. 102d, mink). 
On the blade, it is irregularly waved mosaic, margins slightly crenate 
to smooth (compare Wildman, 1954, fig. 94b, kolinsky). Along the 
shaft, it is diamond petal, margins smooth (compare Wildman, 1954, 
fig. 99a, otter). It changes rather abruptly along the basal 0.3 to 



22 THE PELAGE 

0.4 mm. of the shaft to irregularly waved mosaic, margins smooth, 
on the root (compare Wildman, 1954, fig. 94b, kolinsky). 

UNDERFUR FIBERS 

These are strap-shaped, twisting in many planes, slightly wider 
and thicker in terminal third than in basal third, nearly colorless 
along basal two-thirds and pale brown along terminal one-third ; cross 
section more roundish near base, more elliptical in terminal one-third. 
They are in bundles of 35 to 40. They stand slightly posterior to the 
guard hairs. 

The tip is less than 1 micron wide and about 1 mm. long, with a 
gradual taper. The terminal one-third of the fiber has cross-section 
diameters 6 to 7 by 12 to 15 microns; the basal one-third, 3 to 4 by 
6 to 7 microns. In between, the shaft diameters are transitional in 
size. The attenuated portion of the root lies entirely within the skin. 
It is virtually without structure; clear, smooth, and 5 microns in 
diameter. 

Pigment is distributed thinly and evenly in the terminal one-third 
of the fiber in dark-brown granules and longitudinal streaks. Pig- 
ment is absent from the basal one-half to two-thirds of the shaft. 
When mounted in o-dichlorobenzene, the basal region becomes almost 
invisible. (This mountant has a refractive index of 1.552, as com- 
pared with 1.548 for keratin.) A medulla is lacking. 

The cuticular-scale pattern, starting from the tip, is coronal to 
waived mosaic, margins smooth. Along the shaft it is pectinate, 
margins smooth. It is similar to that on the underfur fibers of mink 
(compare Wildman, 1954, figs. 94c and 102c and d). The tooth of 
each scale curls slightly outward, giving the fiber a rather rough 
silhouette. 

The remarkable ability of the underfur to trap air bubbles and 
repel water has long been known. Conrad Limbaugh, a professional 
diver, has written (personal correspondence, 1957) of an undersea 
observation of a Philippi fur seal off Isla de Guadalupe, Mexico, in 
late November : 

During our stay, the fur seal became more and more aggressive, occasionally 
rushing us with a characteristic spin, emitting strings of bubbles from various 
points on its pelt, especially from behind the ears . . . During all activities bub- 
bles streamed from various points of the pelt so that the seal left a trail of 
bubbles. 

Yearling, Pelagic (Sexes Lumped) 

The yearling, pelagic, continues to wear the coat of the silver pup, 
though with individual fibers now grown to full length (plate 53). 
It may be seen in table 3 that, between mean dates of 25 October and 



YEARLING, AUTUMN (SEXES LUMPED) 23 

22 April, while the silver pup is changing to yearling, the underfur 
and guard-hair fibers increase in length 5 or 6 percent. 

The first crop of underfur, appearing on the silver pup, is whiter 
than all subsequent crops. It may take on a slight brownish stain 
in winter, spring, and early summer. The second crop (yearling, 
autumn) is definitely more pigmented, more pinkish brown, than the 
first. And when the third crop erupts (2-year-old, autumn), it is 
cinnamon-colored like that of the adult. 

On an emaciated female yearling, weight 5.0 kg. (11 lb.) , taken alive 
on the Washington coast about 15 January, the underfur of the back 
was pinkish white (Munsell 7.5 YR 8/2), the pelage prime. On an- 
other female yearling, weight 9.5 kg. (21 lb.) , taken at sea on 29 April, 
the underfur fibers of neck and back were also 7.5 YR 8/2, those of 
the belly pinkish gray (7.5 YR 6/2) . 

Yearling, Autumn (Sexes Lumped) 

Dorsal parts medium gray ; cheeks pale ; mask conspicuous, though 
not as distinct as on the silver pup ; anterior region of chest and flanks 
pale; chest between flippers distinctly brownish (a variable charac- 
ter) ; bases of flippers and adjacent regions of body much browner 
than on the silver pup ; belly often distinctly pinkish ; color patterns 
of male and female indistinguishable (plates 54 and 55). 

As compared with the skin of older animals, the yearling skin seems 
more elastic and more lively ; it curls more readily when cut from the 
body, and it floats higher on water. The color of the underfur is not 
yet as dark as that of the adult. On a male taken 26 September, the 
fur is light grayish brown to brownish pink (7.5 YR 6/2 to 7/2) , ex- 
cept near the bases of the flippers, where it is darker. On a female 
taken 3 October, the fur is brownish pink (7.5 YR 7/2). 

Three-year-old, Adolescent Male (Bachelor) 

Dorsal parts rather uniform brownish gray, relieved by tan cheek 
and top of snout ; ventral parts alternate dark on throat and posterior 
region of chest, pale on anterior region of chest and belly. The 
juvenile mask has almost disappeared; rump patches are beginning 
to disappear; crown or "wig" hairs are beginning to lengthen and 
stand erect. (See similar pelage in plates 56, 57, and 92.) 

Three-year-old Adolescent Female (Young Cow) 

Dorsal parts rather uniform brownish gray ; ventral parts relieved 
by band of light yellowish brown across chest, along upper lip and 
cheek, and on top of snout ; belly grayish brown. Color pattern usu- 



24 THE PELAGE 

ally indistinguishable from that of the male, though the female tends 
to have paler upper lip (mustache line) and throat and never has a 
coarse wig. Rump patches are usually present on both male and fe- 
male 3-year-olds; are never seen on males older than 4 years; are 
occasionally seen on the oldest females. 

Adult Male (Bull) 

Bulls with clean pelage, resting on clean sand in late summer, give 
the impression of gray or dark gray animals. A few individuals are 
pale warm gray or brownish ; a few are almost black. As compared 
with the female, the adult male has retained only a moderately pale 
face, and has attained a light gray mane and wig. Rump patches 
have disappeared entirely (plates 58 and 59) . 

The specimen described in appendix A was rather monotonous dark 
grayish brown; ventral parts somewhat lighter and browner than 
dorsal : face not conspicuously marked ; hairs varicolored over most 
of body, giving at close range a pepper-and-salt effect. 

The guard hairs of the mane of the adult male are the largest of 
any fur-seal hairs except the vibrissae. A specimen hair is 70 mm. 
in length with a blade 89 by 240 microns in cross section ; about four 
times as long and twice as wide and thick as a large guard hair from 
an adult female. While the mane hair is an awn type, the blade does 
not flare out as abruptly as on the ordinary guard hair, but is wide 
for one-half to two-thirds the length of the shaft. The cross section 
of the shaft is, near the tip, ovoid with a concave posterior side; in 
the blade, more flattened, though still with a faint concavity; near the 
base, cigar-shaped without concavity. 

The longest mane hairs are white, without pigment. Medulla is 
absent from the base; broken (fragmental to interrupted) along most 
of the shaft, becoming unbroken (medium wide) lattice in the terminal 
10 mm. When sunlight strikes a white mane hair, it is reflected from 
the colorless cortex and the gas-filled cells of the terminal medulla, 
giving a halo effect (plate 59) . 

Adult Female (Old Cow) 

Most adult females are light-colored across face, chest, and belly. 
Virtually all have a pale band along the upper lip, from one corner of 
the mouth to the other, extending at times to the bridge of the nose, 
backward to the cheeks, and above and behind the eyes. A pale streak 
along the lower lip is not as bright as that on the upper. Seen from 
the front, the face is often mask-like, with black, naked nose in center 
and shiny black eye on each side. While some females are monotonous 
dark brownish gray, all have a paler face, chest, and belly. Rump 
patches are rarely visible. ( Plates 60-62. ) 



PELAGE HAIRS 



25 






Figure 1. — Sketches representing (left to right) : 2 large guard hairs in front and 
side view ; 2 small guard hairs in front and side view, and 3 underhairs or fur 
hairs; about X 3. (Above) Birthcoat. (Below) First adult-type pelage, on 
the silver pup of autumn. (4186 and 4187) 



26 THE PELAGE 

Variation in Length of Pelage Fibers With Age and Sex 

The data in table 3 suggest that the prime underfur fibers of the 
first crop, on the yearling, pelagic, are about as long as those of sub- 
sequent crops. The following changes in mean length throughout 
life are noted (7-year-and-older males and 7-year-and-older females, 
compared with yearlings, pelagic, sexes lumped) : Male — neck, in- 
crease 14 percent in length ; back, increase 10 percent ; belly, decrease 9 
percent. Female — neck, decrease 5 percent ; back, decrease 4 percent ; 
belly, decrease 9 percent. 

With respect to the guard hairs also, one may conclude that the 
only real changes with advancing age are those in length of the mane 
hairs and back hairs of the male. From yearling, pelagic, to adult the 
following changes in mean length are noted : Male — neck, increase 156 
percent in length ; back, increase 30 percent ; belly, increase 7 percent. 
Female — neck, increase 1 percent; back, decrease 4 percent; belly 
decrease 11 percent. 

Sex discrepancy in length of pelage fibers first appears during ado- 
lescence, in the 3- and 4-year-olds. In 7-year-old and older animals, 
the male fibers are seen to be longer than the corresponding female 
fibers by the following percentages : Neck — underfur 23 percent, guard 
hair 153 percent. Back — underfur 15 percent, guard hair 35 percent. 
Belly — underfur percent, guard hair 20 percent. The real relation, 
of course, is perhaps between length of pelage and body size, since 
the adult male is about 4.5 times as heavy as the female of the same 
age (Scheffer and Wilke, 1953, p. 145) . 

It has been mentioned that, regardless of age after 2 years and re- 
gardless of sex, the follicular or hair-root portion of the skin does not 
vary appreciably in thickness from 2.3 to 2.4 mm. 

Variation With Season: The Annual Molt 

Certain aspects of growth and replacement of hair have been de- 
scribed. It will now be helpful to discuss the whole phenomenon 
of molt, the autumnal turnover in which the pelage passes by degrees 
through "staginess" into "primeness." 

FIRST MOLT 

In the first molt, the coat of the black pup, newborn, is replaced 
by the adult-type coat of the silver pup, autumn. The molt starts 
at birth (mid-July), though it does not show on the surface for about 
2 weeks. By mid-September about half of the pups, and by mid- 
October all of the pups, have acquired a silvery gray pelage which 
is nearly adult in character. It is a warm coat that will protect the 



VARIATION WITH SEASON: THE ANNUAL MOLT 27 

pup during its first migration through the icy waters of Bering Sea 
in October and November. An individual termed a "silver pup" 
will be designated arbitrarily on the first of January as a "yearling, 
pelagic" though it undergoes no change in pelage at this time. 

SECOND MOLT 

In the second molt, the coat of the yearling, pelagic, is replaced by 
that of the yearling, autumn. The two coats are similar. It is believed 
that less than half of the members of the yearling age-class reconvene 
on the breeding grounds in their second autumn of life. Those that 
do, return late in autumn, mostly after the first of October. Members 
of all other age-classes return in nearly full numerical strength and 
much earlier in the season, from May to August, depending on their 
age. Most of the yearlings seen on land in autumn have already 
molted into their second adult-type pelage, though growth of the 
fibers has not quite ceased. 

Shearing experiments carried out on a male and female silver pup 
in Seattle Zoo throw light on the second molt. Patches were shorn 
by electric barber -clippers on 24 November 1952, when the pelage 
was judged to be prime. Length of underfur was 12 to 13 mm. ; of 
guard hair 21 to 22 mm. One month later the patches had changed 
from whitish to pale reddish brown, and had assumed a faintly 
"stubbly" texture. Both changes were probably the result of exposure 
rather than of fiber growth. Two months later, on 30 January 1953, 
the patches were seen to be unchanged. On 5 July 1953, slight re- 
growth of underfur — but not of guard hair — could be seen from a 
distance of 6 feet. By 31 July the shorn patches were covered with 
brown underfur to a depth of 3 to 5 mm. No guard hair could be seen 
or felt. On 4 September, the fur was growing rapidly; depth now 
12 to 13 mm., brownish. Sparse new guard hair was starting to 
appear. Growth of guard hair was further advanced (20 mm.) on 
the male rump patch than on the female shoulder patch (16 mm.). 
On 17 September 1953, the male weighed 20.2 kg. and the female 
15.4 kg. ; both in good health. "Apparently the underfur is fully — 
or almost fully — developed. The guard hair is about half as dense 
on the shaved spots as on the surrounding areas" (K. W. Kenyon, 
personal correspondence.). On 26 October 1953, the female died. 
The underfur of her shoulder patch was slightly browner than adja- 
cent normal underfur and had perhaps not attained its full growth. 
On 16 April 1954 (as a 2-year-old, spring), the surviving male was 
beginning to lose weight ; rump pelage normal ; underfur color light 
reddish brown (Munsell 5 YR 6/3). On 14 May 1954, he died. 

553006 O — 62 3 



28 THE PELAGE 

In summary, the new underfur started to grow in early July and 
was nearly full grown by mid- August. The new guard hair started to 
grow about mid-August and was full grown by the end of September. 

On the tanned pelts of 17 yearlings killed 7 October to 7 November 
1941, there are no black dots, signs of unprimeness, on the leather 
side. On the earliest specimen that I saw, taken 26 September, the 
pelage appears to be prime, though horizontal-section slides show 
clearly that a few new guard hairs and a few fur fibers are still grow- 
ing — that is, in certain follicular bundles, both the cylindrical root 
of the old guard hair and the elliptical blade of the new guard hair 
may be seen (plates 63 to 64), and behind the cylindrical roots of 
certain old fur fibers, the elliptical shafts, faintly pigmented, of new 
fibers may be seen. On a yearling taken 3 October, the molt is obvi- 
ously not complete. About 5 cm. anterior to base of tail a sharp 
molt line encircles the body (plate 65). Anterior to this line the 
pelage is denser and more silvery ; posterior to the line, thinner, duller, 
and browner. Fur hairs are 11 mm. in length and brownish pink on 
the anterior region; 5 mm. in length and distinctly brown on the 
posterior region. One rarely sees such a molt line on a fur seal. 

The lengths of underfur and guard-hair fibers on 20 yearlings, 
autumn, are shown in table 3. The fibers on neck and back have in- 
creased slightly in length over those of the pelagic coat ; the fibers on 
belly have increased materially. The belly fibers are, in fact, as long 
as they are ever to be on animals of older ages. The belly coat of the 
fur seal, including both underfur and guard-hair layers, is relatively 
thin over a wide area. (Why?) 

THIRD MOLT 

The third molt is experienced by seals entering their third year of 
life, that is, by "two-year-olds" (plates 66 and 67). Two-year-old 
males and females, though preadolescent, have now been caught up 
in the annual rhythm of return to the breeding grounds in summer. 
Whereas the second molt was centered in August, the third is centered 
in September. Eight specimen skins taken 20 August to 24 September 
are stagy, with short guard hairs and with dark areas on the leather 
side of the pelt. As indicated in table 3, the fibers are shorter than on 
samples of yearlings and 3-year-olds. One may deduce that most of 
the old fur and guard-hair fibers have dropped out of the 2-year pelt by 
the end of August. 

On St. Paul Island, from 7 to 23 September 1956, a sample of 122 
skins from females of all ages was graded for staginess by an em- 



VARIATION WITH SEASON! THE ANNUAL MOLT 29 

ployee of the Fouke Fur Company. The following numbers of skins 
were rejected as being commercially unusable : 

Number Number Percent 
killed rejected rejected 

2-year-olds 8 6 75 

3-year-olds 14 7 50 

4-year-olds 21 2 9 

5-year-olds 9 3 33 

6-year-olds 12 3 25 

7-year-olds 12 1 8 

8-year-olds 2 

9-year-olds 3 2 67 

10-year-olds 3 3 100 

Over 10 years old 8 3 37 

Allages 92 30 33 

The sample is too small to be very useful. It does suggest that 
September molt is more pronounced in the 2-year-olds than in older 
females. 

It has already been noted that the true cinnamon color of the adult 
underfur is first attained at the conclusion of the third molt. 

FOURTH MOLT 

The fourth molt is experienced by the 3-year-olds and is centered 
around the end of September. The 3-year-old males contribute more 
importantly than any other age-class to the annual yield of sealskins. 
Molt during the regular killing season — up to mid- August — has never 
posed a technological problem to the fur processors. 

On a horizontal section from the back of a 3-year-old male, killed 
27 September, about 1 follicular bundle in 10 contains the wide blade 
of a newly erupting guard hair. A 3-year-old female shorn in cap- 
tivity in early September 1957 had not quite recovered by 7 February 
1958. That is, the new growth of autumn 1957 had evidently started 
in late August. On the tanned skin of a bachelor killed on 23 Septem- 
ber, there are distinct dark, unprime areas. On a 3-year-old male 
killed on 27 September 1958, shed underfur was clinging abundantly to 
the hind claws. 

Records from the period 1874^77, when Pribilof natives were per- 
mitted to make autumn "food killings", are of interest here (Jordan 
and others, 1898, p. 262, 266, 267, 269) : 

Number killed Number found stagy 

17 August 134 5 

23 August 207 7 

7 October 133 All 

19 October 176 57 

24 October 104 All 

31 October 163 All 

13 December 825 A few 



30 THE PELAGE 

The food killings were made largely from the bachelor class. These 
records show that molt was conspicuous in the month of October. 

There is evidence that growth of the new underfur may begin as 
early as July. Annually in summer, from 1904 to 1932, Government 
agents on the Priblof s used to mark a few thousand bachelor seals by 
shearing the top of the head. Whenever a shorn animal appeared 
later in the summer on the killing field it was spared as a "breeding 
reserve" (Scheffer, 1950 b, p. 7-8). One year later, however, the 
guard hairs had recovered, and at this time the killing crew in- 
advertently knocked down a few large bachelors that had been shorn 
the previous summer. A veteran Fouke Fur Company employee has 
written (personal correspondence) that "after the [marked] skins 
had gone through processing and the guard hair had been removed, 
it was discovered that the underfur had not grown out to any degree." 
On the basis of modern information, this phenomenon can be ex- 
plained. For example, young underfur fibers of an individual were 
shorn in August 1930. In July 1931 the seal was killed, wearing a 
short underfur coat of two components : the stumps of the 1930 season 
and the new, short growth of 1931. 

The fibers of the 3-year-old (except, perhaps, those of the neck) 
are no longer than those of the silver pup and yearling (table 3). 
Within the 3-year class, the fibers of the female are beginning to lag 
behind those of the male. 

MOLT IN ADULTS 

There is little evidence on molt in adults with relation to calendar 
date. Quite certainly, however, it is centered in October, somewhat 
later than molt in the 3-year-old. Employees of the Fouke Fur 
Company have long recognized that stagy skins are seen more often 
in younger males (3- and 4-year-olds) than in older ones (5- and 
6-year-olds) at any given time in late summer. 

On a 5-year-old female in Seattle Zoo, shorn in early September 
1957, the pelage had recovered by 7 February of the following year. 
In other words, all of the new fibers erupted after early September. 
(The reader will recall that on a 3-year-old companion, shorn and 
killed on the foregoing schedule, the pelage had not quite recovered by 
7 February because the younger animal had started to grow a new 
crop of fibers in August. ) 

I have examined median sections of skin from adult seals, age 7 years 
or older, taken in September 1958, including 2 males and 26 females. 
On all of the sections, melanocytes in the guard-hair follicles are still 
active. No club hairs can be seen. 

In all six hind claws of an old female lying on the beach on St. 
Paul Island, 25 September 1958, there were tufts or balls of underfur, 
2 to 5 mm. in diameter, scratched from the pelt. 



VARIATION WITH SEASON! THE ANNUAL MOLT 31 

COMPARISON WITH MOLT IN OTHER FURBEARERS 

The phenomenon of molt in furbearing animals has been studied 
by Gunn (1932) in muskrat, by Bissonnette (1935, 1942) in ferret and 
weasel, by Bissonnette and Bailey (1944) in weasel, by Bassett, Pear- 
son, and Wilke (1944) in fox, by Bassett and Llewellyn (1948, 1949) in 
fox and mink, by Hall (1951) in weasel, by Rothschild and Lane 
(1957) in ermine, and by Shanks (1948) in muskrat. Molt begins in 
one or several places on the body and proceeds outward in waves. 
Most of the fibers along the front of a wave are in the same stage of 
molt. (In man, however, molting is mosaic, with new hairs popping 
up here and there to replace old ones.) New hair or fur growth is 
signaled by a massing of pigment in the follicles. This massing gives 
a dark bluish cast to the body side of the pelt. The new hairs erupt 
as the old hairs are being shed, although the momentum of new growth 
usually carries on for a while after shedding is complete. The stimulus 
or triggering mechanism in molt is change in length of daylight, 
which stimulates in turn the optic nerve, the anterior hypophysis 
(pituitary), and the hair-growth cycle. Melanocytes have already 
ceased to function while the hair shaft is still completing its growth 
in length. Thus, in a prime pelt, pigment is absent both from the 
dermis and from the attenuated base of the shaft. In some species, 
molt may be a rather long process ; in the silver fox it begins in early 
May, when simultaneously the old hairs start, to drop out and the new 
ones to appear. (Growth and maturation of the guard hair precede 
that of the underfur.) About 4% months later, by mid- September, 
the old hair is shed. The new hairs continue to grow, and pigment 
continues to leave the skin, until about the first of December, or date 
of primeness. In general, the direction of the molt is: legs, rump 
and tail, abdomen and sides, back, mane and crown. 

Molt in the fur seal is an annual event, as in the silver fox but not 
in the mink, where it is semiannual. Molt in the fur seal is rather 
slow, requiring 4 to 5 months' time. It is gradual and unobtrusive. 
Old fibers are shed singly as the new ones erupt. Thus the seal at all 
times has a warm, water-repellent coat. As previously mentioned, a 
distinct molt line is not often seen. The molt is not accompanied by 
profound changes in the epidermis. In certain phocids (elephant 
seal, Weddell seal, and monk seal), great tatters of epidermis peel 
from the body along with the newly shed hairs. No pinniped has 
more than two fundamentally distinct pelages: juvenile and adult. 
The juvenile coat may be shed, for example, in the harbor seal, shortly 
before birth. While in some species, the harp seal, for example, the 
pelage may molt several times before it assumes its mature pattern, 
its only real break in character is with the juvenile coat. 



32 THE PELAGE 

Rand (1956) has given a good description of molt in the South 
African fur seal Arctocephalus pusillus. Molt here seems to differ 
from that in the northern fur seal only in the longer time required 
for the black birthcoat to disappear. In the pup coat of Arctoce- 
phalus, change is not visible until the animal is 2 months old or older, 
and the coat persists nearly 4 months. The corresponding periods 
for Callorhinus are 6 weeks and 2i/£ months. The second molt in 
Arctocephalus, between December and February, corresponds to that 
of Oallorhirms, between June and August. (Incidentally, commer- 
cial sealing in South Africa is largely directed against the yearling 
animal, 8 to 10 months old ; while in Alaska it is directed against the 
3- and 4-year males.) 

The Sensory Vibrissae 

All mammals, with the exception of man, are reported to have, at 
some time in life, special sensory bristles on or near the face. Even 
whales may have as many as 80 such bristles, their function sup- 
posedly to detect water currents and macroplankton food (Nakai and 
Shida, 1948). According to Pocock (1914), the sensory vibrissae 
may be classified according to location as mystacial, superciliary, 
genal, submental, and interramal. The presence of all five on one 
kind of mammal is thought to be primitive. According to Noback 
(1951, p. 481), the vibrissae may also be classified as active tactile 
hairs, under voluntary control; as passive tactile hairs, not under 
voluntary control ; as follicles with a circular sinus ; as follicles with- 
out a circular sinus. 

In pinnipeds, only the mystacials and superciliaries are retained 
(plate 34). In all otariid seals, the superciliaries are unimportant; 
in phocid seals, well developed. 

I have made only a cursory examination of the vibrissal roots and 
surrounding tissues in the fur seal (plates 68, A and B, and 69). 
Br0ndsted (1931) gave 7 diagrams of the snout musculature and 
innervation of the California sea lion Zalophus, a close relative of 
the fur seal. He described the nasal muscle as being "colossal" in 
pinnipeds. According to Miller (1952 and in personal correspond- 
ence) certain features of anatomy of the dog may (?) resemble 
those of the fur seal. In the dog, the vibrissae are moved by two 
sets of muscles: levator nasolabialis and, beneath it, maxillonaso- 
labialis. The vibrissae are supplied from at least one cranial nerve : 
the trigeminal (5th cranial), mixed motor and sensory, leading to 
its maxillary division and thence through the infraorbital foramen 
to the infraorbital nerve. Whether the facial (7th cranial) nerve, 









THE SENSORY VIBRISSAE 33 

motor only, leading to the dorsal buccal nerve, plays a part also 
is uncertain. 

PRENATAL DEVELOPMENT OF THE VIBRISSAE 

On the evidence of 14 male male fetuses and 10 female fetuses, 
the formula for mystacial vibrissae in the fur seal does not vary with 
sex or age. On each side of the snout there are 5 or 6 rows contain- 
ing (from top to bottom row) the following number of vibrissae: 
or 1 ; 3, 4, or 5 ; 5 or 6' ; 5 or 6 ; 3 = 20 to 23 ( plate 70-A ) . The observed 
mean total in males is 21.4 and in females 21.1. On all of the fetuses 
there are two superciliaries on each side of the head ; the upper one 
distinctly longer than the lower. 

On the South African fur seal, Rand (1956, p. 10), has found 
up to 33 mystacial and 2 superciliary vibrissae on each side of the 
head. A common formula for mystacials is: 3-5-6-6-6-4=30. Like 
those of Callorhinus, the vibrissae of the southern seal are black at 
birth, later turning white. 

Fetus of 23.7 g. (0.0049 MNW), female, 14 February 

On this smallest fetus in the collection, the complete vibrissal 
pattern is set, though several of the individual bristles have not yet 
erupted. There are no visible signs of additional, rudimentary (ves- 
tigial) bristles. Vibrissa formula: 1-3-5-5-6-3 = 23. The outermost 
(posterolateral) vibrissa in the row-of-six is the longest, 0.8 mm. 
The inner (anteromedian) bristles are still beneath the epidermis, 
though ready to erupt. Each erupted bristle is sheathed nearly to 
its tip in a thick, translucent, tubelike extension of the skin. Black 
pigment can be seen in the larger vibrissae. It is likely present 
also in the smaller ones though obscured by epidermis. Two super- 
ciliaries can plainly be seen as white parallel lines beneath the skin, 
each 0.8 mm. in length, about ready to erupt. 

In view of the fact that the vibrissae are already well established 
on this fetus, far in advance of the guard hairs and underhairs, and 
on a body which has attained less than 1/200 of its newborn weight, 
one may deduce that the sensory bristles are very important to the 
welfare of the seal. Danforth (1925, p. 67) has noted the remark- 
able constancy of vibrissae as compared with other body hairs of 
mammals, this "being due to their early appearance and the circum- 
stances of their embryonic origin." 

Fetus of 103 g. (0.021 MNW), female, 20 January 

Mystacial vibrissa formula: 0-3^-5-5-3 = 20 (plate 70-A). The 
outer (posterolateral) vibrissae in the row-of-four and in the row-of- 
five are the longest, all about 4.6 mm. All of the mystacial vibrissae are 
dark brown, almost black/ All but the smallest have whitish tips, 



34 THE PELAGE 

suggesting that in the early stages of development the fiber starts 
to grow in advance of the melanoblasts. The superciliaries have now 
erupted and are distinctly unequal, the upper one of the pair meudur- 
ing 2.3 mm. and the lower one 1.4 mm. in length. They also are 
whitish at the tip. 

Fetus of 6.80 kg. (1.3 MNW), male, 20 June 

This is a full-term fetus taken at sea on 20 June, actually larger 
than many newborn pups. The vibrissae are glistening black, with 
the barest suggestion of pale tips on several. The largest vibrissa 
is 64 mm. in length, flattened, tapered, and slightly curved. (Vibris- 
sae are the only fur-seal hairs that taper all the way from base to 
tip.) At the base, the cross section is elliptical and measures 0.39 
by 0.55 mm. Halfway along its length, at a point where the vi- 
brissa will first fit into the slot of a Hardy sectioning device, the cross 
section is 0.28 by 0.33 mm. (plate 69). The pigment is dark brown, 
cortical, and granular; mainly in clumps which increase in size and 
darkness toward the central axis of the fiber. Toward the periphery, 
pigmentation ceases rather abruptly, leaving a clear cortical zone 
about 5 microns wide, inside the cuticle. The surface of the vibrissa 
is smooth and without scale pattern. In cross section, the cuticle is 
seen as a hyaline layer less than 1 micron thick. 

POSTNATAL DEVELOPMENT OF THE VIBRISSAE 

Measurements of the longest vibrissa on each of 31 newborn seals 
are summarized as follows (table 4) : 17 males, mean 62.6 mm. 
(51 to 75) ; 14 females, mean 57.4 mm. (52 to 63). 

In both sexes, the vibrissae grow most rapidly in the juvenile and 
adolescent years, before the end of the third year. On a yearling 
male observed in Seattle Zoo during a 6-month period, February 
to July, the longest vibrissa increased by 36.5 mm., or at a rate of 
0.20 mm. a day. Here is evidence that growth of the vibrissa is not 
confined to autumn, as is growth of the pelage proper. On the same 
individual in its second year of life, between August and mid- April, 
the longest vibrissa increased by 32 mm., or at a rate of 0.12 mm. 
a day. The vibrissae are still growing at age 8 and, so far as the 
data permit one to estimate, will continue to grow throughout life 
( plate 70-B ) . The longest recorded vibrissa on a male is 334 mm. ( 13.1 
in.) and on a female 220 mm. (8.66 in.). The record specimens were 
old animals of unknown age. The female was extremely large — 
weight 60.3 kg. or 134 lb. Comparing the mean values for newborn 
pups with selected maxi??ium values for adults, on the assumption 
that maximum values represent unbroken and relatively unworn 
bristles, one may estimate that the male vibrissa increases to about 
5.3 times its newborn length ; that of the female, 3.9 times. 



PELAGE ANOMALIES 35 

Changes in color of the vibrissae with advancing age have been 
studied with the objective of finding a useful age index. On St. 
Paul Island in September, I examined vibrissae Of 20 freshly killed 
males, ranging in age from 4 to 6 weeks (estimated) to very old. 
On the black pup and the silver pup, all vibrissae are black (Munsell 
7.5 YR 2/0). On the yearling and 2-year-old, many vibrissae are 
faded, grayish brown at the tip. On the 3- and 4-year-olds, the 
over-all color effect is blackish, though many vibrissae are mottled, 
most have faded tips, and a few of the small ones are all white. 
Through ages 5, 6, and 7, the color is increasingly whitish, though 
a few small, all-black vibrissae may be seen. (The superciliary 
vibrissae gradually become worn down to, or broken off to, the level 
of the guard hairs, and one can seldom find them on a seal older 
than 6 years.) The single, very old bull examined had all-white 
mystacial vibrissae — actually pale yellow (2.5 Y 8/4). Paling of 
the base of the vibrissa begins when the pigment cells of the follicle 
cease to function. Paling of the tip is a result of fading (bleaching) . 

In this connection it is interesting to note a statement with regard 
to human hair which, like the vibrissa of the seal, is a nonmolting 
fiber: "The fact that in rare instances hair can quickly become gray 
must be acknowledged . . . The old idea that hair is practically a 
dead tissue, cut off from the metabolic influences in the body, must 
be forsaken" (American Medical Association Journal, 1943, p. 162). 
Graying in man may result either from loss of pigment or from 
masking of pigment by gas bubbles in the cortex. Melanin can be 
bleached in vitro in the laboratory by ultraviolet radiation. Quite 
surely it can be bleached, though more slowly, by sunshine. 

The color trend with age in the female vibrissae parallels that in 
the male (table 5). The 5-year-old (entering 6th year) exhibits the 
greatest variability in color. The vibrissae in about 67 percent of 
the 5-year female class can be called "black and white." 

In July 1947, I saw mystacial vibrissae being plucked by native 
children from 3- and 4-year male seals on the killing fields of St. Paul 
Island (plate 71). One lad reported that he gathered about 500 a 
day and sold them for a cent apiece. The vibrissae were eventually 
delivered to an Oriental broker in San Francisco to be used (it is 
said) for cleaning the stems of pipes. 

Pelage Anomalies 

COLOR ANOMALIES 

Three mutant color patterns have been observed in the fur seal: 
albino, piebald, and chocolate. 



36 THE PELAGE 

The albino pup has white underhair, guard hair, and vibrissae 
(plates 72, 73-A, 73-B, 74). The parts that are black and naked on 
the normal pup (flippers, nose, and eyelids) are pinkish gray on the 
albino. The pelage becomes stained soon after birth. On a cap- 
tive which had completed its first molt into adult-type pelage in the 
Seattle Zoo, the pelage on 2 December was reddish brown, or where 
brightest it was yellowish red (Munsell 5 YR 5/6) (plate 73-B). 
One rarely sees an albino older than a pup. Once or twice each 
summer a light-brown bachelor with pink eyes and flippers may appear 
in commercial sealing drives. A female albino lived to age 4 years 
in San Diego Zoo, while a fully adult female was shot in the wild in 
August (plate 74). With regard to incidence of albinism, Edward 
C. Johnston (in U.S. Bureau of Fisheries, 1923, p. 112) reported 6 
white individuals in approximately 177,000 pups counted. Biologists 
on the Pribilofs now see perhaps 5 to 10 albinos each summer out of 
500,000 to 600,000 pups born. Observed incidence of albinism is thus 
between 1 in 30,000 and 1 in 100,000. According to Green (1947) 6 
albinos were reported in 1,672,604 coyotes Canis latrans taken over 
a 30-year period. This corresponds to 1 in 278,767. 

Piebald (white splotched) seals are seen more often than complete 
albinos (plate 75-A). One iris, neither, or both may be pinkish. 
White areas may appear on naked parts as well as on haired. A 
nearly full-term fetus had an unusual whitish belly, though the white 
was not sharply set off as it is on a true piebald. On one peculiar 
pup about 1 month old, the eyes and flippers were pink, the underhair 
white, but the guard hair grayish brown ( plate 75-B ) . The character 
for pigment in guard hair is perhaps distinct from that for underhair; 
further study is needed. 

The "chocolate" mutant is named from the appearance of the new- 
born pup. I cannot certainly identify in later life a seal that was 
chocolate at birth. Many variations of light brown, silver, and creamy 
pelage, combined with normal dark flippers and eyes, are seen on the 
breeding grounds (plate 76, A and B) . 

EFFECT OF DISEASES, PARASITES, AND PHYSIOLOGICAL 
DISORDERS ON PELAGE 

The terms "rubbed" and "mangy" are used loosely on the Pribilofs 
to describe pelage conditions where the hair fibers are sparse or patchy. 
Rubbed areas are seen on seals of all ages and both sexes. "Wherever 
the guard hair is absent, the underfur becomes soiled and matted. 
Occasionally the underfur is also absent and the naked skin is ex- 
posed. Little study has been made of such pelage anomalies. Var- 
ious degrees of "rub" or loss of guard hair are illustrated in plates 



PELAGE ANOMALIES 37 

77 (A and B) to 81. From the 6-year-old female illustrated in plate 
77-B, samples of abnormal pelage were sent, in formalin and in dried 
state, to Dr. Robert W. Menges (U.S. Public Health Service) ; similar 
specimens were also sent from another 6-year-old female. Dr. 
Menges has kindly advised that no evidence of ringworm or mites can 
be found. The diagnosis in one case is "acquired canities" and in the 
other "atrophy of skin and alopecia" (personal correspondence, 
1959). (See also Menges and Georg, 1957.) 

On a male black pup perhaps 2 or 3 weeks old, patches on the 
back of the body and neck were hairless, wrinkled, pinkish gray, 
with marginal fragments of brown, scabby, loose epidermis (plate 82, 
A and B). Seventeen lice were picked from ears, base of tail, eyelid, 
head, and denuded patches. It is presumed that the lice were 
scavengers on the denuded areas rather than causative agents in 
loss of the hair. While the lice in this case were not identified, two 
species have been collected frequently from fur seal pups: Antarc- 
tophthirus callorhini (Osborn) 1899, and ProecMnophtMrus fluctus 
(Ferris) 1916. Further study of the environmental preferences of 
the two species is proposed. 

Murray (1958, pp. 404-405) kept lice from antarctic elephant seals 
Mirounga leonina under sea water for more than 2 weeks. 

It would appear that oxygen is obtained from sea water by diffusion through 
the cuticle . . . Unlike other species of lice, Lepidophthirus macrorhini forces 
its way into the skin and creates a burrow under the stratum corneum of 
the epidermis . . . When the elephant seal moults the stratum corneum ... is 
shed intact with hairs attached. Although some lice are lost at this time 
many remain, since only the roof of the burrow which they inhabit is 
removed. 

Pachyderma, or thick skin, was observed on a bachelor killed 3 
July (plate 83). Tough, whitish connective tissue had replaced 
about nine-tenths of the blubber layer. The workman who tried to 
remove the blubber in routine fashion was obliged to reject the skin. 
The pelage was normal, the skin about five times thicker than nor- 
mal. I am reminded of an ailment of cattle in which the hair is 
lost and the skin becomes thick, dry, leathery, and deeply creased 
(U.S. Department of Agriculture, 1954). This ailment, often fatal, 
is the result of contact with highly chlorinated napthalene in cer- 
tain petroleum lubricants. Do seals suffer damage from prolonged 
contact with floating oil wastes on coastal waters ? 

Up to 1956, the skins of old female seals had not been taken 
deliberately on the Pribilof Islands. In 1956, substantial numbers 
were taken on an experimental basis. It was soon discovered that 
skins of old females tend to be of poor fur quality. Both underfill* 
and guard-hair fibers may be uneven in length and insecurely rooted. 



38 THE PELAGE 

As a result, the pelt after unhairing may present a ragged, splotchy 
appearance (plate 84). The causes — presumably physiological — of 
poor fur quality in older animals are unknown. 

The damaging effect of high temperature on fur seats is well known 
to those who work in the sealskin industry. A seal driven too quickly 
from beach to killing field, especially on a dry, warm day (60° F.), 
may die of heat prostration. Its pelt is subsequently called a "road- 
skin." "Temperatures of roadskin seals were between 42.3° and 
43.9° C. . . . Seals with temperatures above 42° C. (107.6° F.) are 
invariably prostrated or dead" (Ford Wilke and Karl W. Kenyon, 
in personal correspondence, 1951) . Roadskin seals are flayed as soon as 
possible, because about an hour after death the fur fibers loosen in 
their follicles and can be plucked with one's fingers; the pelt is then 
worthless. Initial processing is accelerated in order to get roadskins 
into brine as quickly as possible in order to prevent loss of fur and hair. 

Partridge (1938) studied horizontal layers sliced from fresh, shorn 
Alaska sealskins. He found that seals driven a long distance from 
the beach to the killing field tend to lose lipids from the "epidermal 
layer" — actually the epidermis and upper dermis — containing the 
sebaceous glands. He recorded percent lipids (of dry weight), as 
follows : 

Seal No. 1 Seal No. 2 Seal No. 4 Seal No. 3 Seal No. 5 Seal No. 6 
Length of drive. short short short long long (seal died) 

Percent lipids.. 16.7 14.2 14.0 13.1 10.8 5.1 

A fairly common, perhaps harmless, condition in which the blubber 
is reddish orange rather than white will be mentioned on page 60. 

From my diary in 1940 is taken a note on disease of the naked 
flipper : 

During branding operations in September, we noted that several hundred 
pups had warty areas % to y 2 inch in diameter on their flippers. These were 
of the same texture as the surrounding skin and were up to % inch thick. 
Occasionally a flipper with a small, round, raw sore (originating in such a 
wart?) was seen. 

Warts or blisters are commonly seen on older animals (plate 85). 
Dr. W. J. Hadlow has written (personal correspondence, 1958), 
as follows : 

On the basis of my limited examination of several cutaneous nodules ... I 
conclude that the lesion was formed from elements of the epidermal adnexa, 
probably from modified sebaceous glandular elements. The lesion somewhat 
resembled the general histologic pattern of the circumanal gland adenoma 
of the dog. A striking feature of the seal lesion was the presence of numerous 
large, acidophilic inclusions which occupied much of the cytoplasm of some 
cells. The nodules developed in the dermis and did not appear to involve 
directly the overlying epidermis, which usually was not appreciably altered. 
In this respect the lesions are not warts in the true sense of that term. 



PELAGE ANOMALIES 39 

Abegglen and others (1956, p. 14, 20) found strange circular marks 
on the fore flippers of approximately 50 bachelor seals in the 
summer of 1956. No satisfactory theory about the origin of the 
marks has been offered, although it is generally believed that they 
are natural rather than artificial. Raw, circular areas are occasionally 
seen on the fore flippers of seals taken at sea. These resemble lam- 
prey scars, though lampreys would not be likely to attack seals. 

An abnormal bachelor was killed on St. George Island in 1949. 
The left flipper had apparently been bitten off in infancy, and the 
fur had grown over the stump so tightly that when the skin was 
blubbered no armhole was apparent on the left side. Biologists in 
1947 tagged a pup with only one fore flipper. At the site of the 
missing flipper the pelage was intact and continuous. The pup moved 
successfully on land by "rock and roll." 

Seals, particularly emaciated pups, are prone to suffer an eye in- 
fection which reveals itself as a purulent whitish or yellowish dis- 
charge. The disease has not been studied. 

EFFECT OF SEX ABNORMALITIES ON PELAGE 

Upon examination of five cryptorchids (adult males with infantile 
testes), it was concluded (Scheffer, 1951) that the pelage tends to 
be smooth in texture and uniform in color along the back, as on the 
adult female (plate 86). Mane and wig are conspicuously absent, 
as is the musky male odor of the body. One cryptorchid was dis- 
covered in a harem during breeding season. Its "unmaleness" led 
it to be treated as a female by the dominant male of the harem 
(plate 87-A). Its pelt is illustrated in plate 87-B. On another 
cryptorchid, estimated age 15 years, a normal brown layer of guard 
hair was present at the base of the flippers while underfur here was 
entirely lacking. The depth of guard hair on neck, back, and belly 
is less than the depth of corresponding guard hair on the normal adult 
male (table 3) ; the depth of underfur about the same. The length 
of mystacial vibrissae and length of ear are normal. On five cryptor- 
chids, the maximum length of vibrissa was 286 mm. (compare table 
4). The maximum length of ear was 58 mm. (compare maximum 57 
mm. recorded for a normal male). 

The finding (Scheffer, 1949) of a 4-year-old female with a promi- 
nent clitoris and small ovaries has been reported. The tanned pelt of 
this individual has no unusual features. 

Supernumerary teats will be mentioned on page 53. 

FOREIGN GROWTHS 

Brown algae, red algae, and gooseneck barnacles are occasionally 
seen on the guard hair of living fur seals. Epiphytic green algae 



40 THE PELAGE 

have been reported from phocid seals, though not yet from fur seals. 
Foreign growths are more common on seals at sea than on land. 
Why do organisms attach to certain individuals but not to others ? 

Brown algae, Phaeophyceae, Ectocarpus sp. (identified by Dr. G. 
F. Papenfuss) : On a female, 2 to 4 years old, nonpregnant, shot near 
Farallon Islands, California, 12 December 1948, the pelage appeared 
curly and brownish from an extensive growth of Ectocavpus (plate 
88). The growth covered most of the body except nose, flippers, and 
belly; the parts that are commonly rubbed during grooming activi- 
ties at sea. Fragments of the dried plant, light brown in color, 
remained on the pelt throughout the tanning process. 

Red algae, Rhodophyceae, Erythrocladia sp. ("if not identical with 
the species, is closely related to E. polystromatica Dangeard" Papen- 
fuss, personal correspondence, 1945) : This plant is of fairly common 
occurrence on seals. It gives a rusty, reddish brown tint to the guard 
hair. Under a microscope, it has an attractive purplish color and 
is seen to be made up of dense clusters or disks closely appressed 
to the hairs. The first specimen from a fur seal was collected on a 
bachelor seal on St. Paul Island on 2 August 1945. Three similar 
collections made 22 June-3 August 1946 were identified by Papen- 
fuss as Erythrocladia {polystromatica^.) . 

Barnacles, Cirripedia, Lepas sp. (identified by Dr. Dora P. Henry) : 
Gooseneck barnacles are seen on perhaps one fur seal in a hundred 
taken at sea; rarely on land. A bachelor killed on St. Paul Island, 
18 July 1945, had several hundred barnacles L. hittii Leach attached 
to the rump (plate 89) . 

An emaciated yearling female, weighing only 6.4 kg. (14 lb.) was 
found dead, in fresh condition, near Grayland, Wash., on 11 April 
1948. Several hundred barnacles were attached in clusters to the 
armpits, hind flippers, fore flippers (base only), and belly. An adult 
female, weight 26 kg. (58 lb.) was found dead in the surf near Cape 
Shoalwater, Wash., on 27 April 1948. Tufts of barnacles were at- 
tached to the armpits. Dr. Henry wrote (personal correspondence) 
u The yearling female has young Lepas and larvae; the adult has 
larvae only." The female mentioned previously as having Ectocarpus 
growth had also gooseneck barnacles scattered over the body. 

The Pribilof Sealskin Industry 

HISTORY OF THE INDUSTRY 

The Pribilof seal herd has been cropped annually, so far as anyone 
knows, since the discovery of the breeding grounds in 1786-87. 
Modern, sustained-yield management began with the Treaty of 1911, 



THE PRIBILOF SEALSKIN INDUSTRY 41 

at which time all northern fur seals were brought under international 
protection and the killing of Pribilof seals was undertaken directly 
by employees of the United States Government. The Government 
now has an interest in all stages of the industry, namely, management 
of the stock, cropping, processing, and marketing. The stages after 
killing are handled primarily by a private contractor under Govern- 
ment supervision (Baker, 1957; Fouke Fur Company, 1958; Scheffer 
and Kenyon, 1952; Thompson, 1950). 

Previous to 1855 fur-seal skins were in little demand in Europe or America. 
The fur was not fashionable . . . About 1825 the unhairing and dyeing of fur- 
seal was introduced [by Denison Williams, a capmaker of Albany, N.Y.I, and 
although the article was very poor compared with the choice product of the 
present time, it was a decided advantage over the former methods of dressing. 
Between 1855 and 1870, through experiments on the part of Messrs. Oppenheim 
& Co., and of Messrs. Martin & Teichman, in London, and of Mr. George C. 
Treadwell, in Albany, the methods of dressing and dyeing fur-seal were greatly 
improved, resulting in an exquisitely soft and downy texture and rich dark- 
brown color, which was quickly adopted by the fashionable world for cloaks, 
jackets, muffs, trimmings, etc. So popular did the fur become that the demand 
quickly ran up ... to 200,000 during the eighties at greatly increased prices 
. . . Since 1870 practically the entire world's product of fur-seal skins has been 
sold in London. [Stevenson, 1904, p. 300-301] 

No Pribilof sealskins owned by the United States Government were 
shipped to London for processing or sale after 1912. On 16 December 
1913, the Government began selling sealskins at public auction in St. 
Louis, Mo. During World War I, with encouragement from the 
Secretary of Commerce, the Gibbins and Lohn Fur Skin Dressing 
and Dyeing Company developed a factory in St. Louis for processing 
sealskins. The early success of the company was partly the result of 
skills brought directly from London by Messrs. Gibbins and Lohn. 
The first lot, including 1,900 dyed pelts, was sold on 20 September 
1916 in St. Louis by Funsten Brothers and Company. Early in the 
year 1921, the Government contract with Funsten was canceled and 
a new one entered into with the Fouke Fur Company. Down to the 
present time, this firm has been the sole contractor for handling the 
United States' share of pelts from the Pribilof Islands (Fortune, 1930 ; 
Fouke, 1949; Fur Trade Review, 1916; U.S. Bureau of Fisheries, 
1916, 1917, 1922, 1938) . 

KILLING, SKINNING, BLUBBERING, AND CURING 

The "island operations" have been described by many authors (see 
above). Since about 1940 the annual harvest has consisted almost 
entirely of bachelor seals: 60,000 to 70,000 subadult males near the 
end of their third or fourth year of life. Since 1956 the harvest has 
also included a substantial number of females, from subadult to old- 



42 THE PELAGE 

adult. Future plans call for a sustained annual harvest of both males 
and females. 

On the killing field, the seal is clubbed on the head and the pelt is 
stripped from the body, along with adhering masses of blubber and 
other subcutaneous tissues. The raw pelt is washed overnight in 
running sea water. On the following day, the blubber is removed 
forcibly with a dull blubbering knife, and the tail, ears, and loose 
tatters of skin are trimmed off (plates 90-92). The pelt is then 
washed for 12 hours or more in circulating saturated brine, wrung, 
dusted with boric acid and salt, barreled, and shipped to St. Louis. 

A kind of spoilage, known as "pink," on the leather side of cured 
sealskins was seen from time to time in the early years of the industry. 
It was produced by halophilic bacteria, certain ones of which are able 
to live in saturated brine. Spoilage is now controlled by dusting each 
skin with boric acid before it is shipped. ZoBell (1946, p. 196-197) 
has given a good review of spoilage in sealskins and other marine 
products. 

PROCESSING AND MARKETING 

A sealskin shipped from the Pribilofs may not reach the auction 
block as a finished pelt for a year or two. The raw, salted skin is 
held in cold storage in the original barrel until the factory is ready 
to work it. There are more than 125 different operations in the 
Fouke process, and it takes more than 3 months for the completion 
of the dressing and dyeing of each skin, which is in work contin- 
uously (Fouke Fur Company, 1958, p. 45-46). Processing includes 
four main steps: unhairing, leathering, dyeing, and finishing. 

The raw, salted skin is first graded for size (essentially length 
and width) on a set of five graduated wooden boards, the proto- 
types of which came from London before 1916 (fig. 2 and table 6). 
As a result, each skin can then be processed with others of its size, 
all of them receiving, for example, the same measured amount of 
detergent. The five size-grades are small medium, medium, large, 
extra large, and extra extra large. In actual practice, the raw pelt 
is first graded as small, medium, or large. The "large" lot is again 
graded into large, extra large, and extra extra large as a matter of 
record and for possible research use. The three divisions of "large" 
are then regrouped for processing. 

Also at this time a particular blemish or undesirable feature may 
be noted in the record, for example, bite, bruise, scar, or flay. The 
location of the blemish may be charted on a special card (fig. 3). 

The next important step is unhairing. The skin is quickly and 
carefully preheated in a "cockle." It is then placed over a beam, 
dusted with chalk, and the guard hairs are scraped off with a tool 



THE PRIBILOF SEALSKIN INDUSTRY 



43 




Figure 2. — Outlines of the five boards used for size-grading raw, salted sealskins ; 
X % (see table 6). 



553006 0—62- 



44 



THE PELAGE 




Figure 3. — A sealskin "map" used for reference purposes in technological re- 
search. On a finished, dyed, and trimmed sealskin, the small circles represent 
ear holes ; the large ovals, fore flipper holes. 



THE PRIBILOF SEALSKIN INDUSTRY 45 

similar to a blubbering knife. At a much later stage in processing, 
the skin is passed through a "dehairing" machine which clips off the 
shorter, residual guard hairs. Poland (1892, p. 191-192) referred 
to these as "water-hairs" and described a machine for clipping them; 
basically the same machine is in use today. 

Leathering of the sealskin is essentially an oil, or chamois, dressing. 
A chemical pret aiming is not applied. The oil is derived from seal 
blubber which has cured for months in salt. Mathur (1927) found 
that this oil contains up to 7 percent free fatty acids. He concluded 
that oil tannage is essentially a process in which, through the action 
of free fatty acids (particularly oleic series), water is removed from 
the skin but connective tissue fibers are left intact. Oil tannage 
seems to be irreversible, in distinction to tannage by chrome, vege- 
table, or alum. 

The dyeing process is, to a certain extent, a trade secret. Various 
inorganic and organic reagents, as well as complex vegetable and 
animal pigments, are used to produce black, brown, and gray tones 
in the underfur. The natural curl of the fur is removed, though a 
bend or kink remains near the base of each fiber. 

The finished pelt is graded for length on boards which are ruled 
in one dimension only. However, the grading is not altogether ob- 
jective, and a very narrow skin (for example) would be graded 
slightly shorter than its actual length. The same terms — small 
medium to extra extra large — as were applied to raw, salted skins are 
applied to the finished pelt. The finished pelt is also graded for 
quality, under a system long established, as follows : 

Regular: fine, one (I), and two (II) 

Scarred : A (same as fine and one, but scarred) , and 

B ( same as two, but scarred ) . 
Three : (damaged, off-size, or very poor quality fur, virtually unsalable) . 

The pelts are sold at twice-yearly auctions in St. Louis for the 
account of the Government. 

DIMENSIONS AND WEIGHTS OF SEALSKINS 

For their practical value, certain data have been assembled on the 
sizes of sealskins at various stages in the routine flow from killing 
field to auction block. These are presented below 

The question had been raised "For seals of any specified length, 
do those individuals with greater girth tend to arrive earlier on the 
breeding ground?'' Data provided by the Fouke Fur Company for 
the period 1938-41 are given in tables 7 and 8 and are compared 
graphically in figure 4. These data shovr that the proportion of 
larger skins taken before mid-July is approximately the same as the 
proportion taken after. (In accepting this interpretation, one must 



46 



THE PELAGE 



PERCENT 
50 i- 



40 



30 



20 



10 







SM 



M 



L 



XL XXL 



Figure 4. — Comparison of sizes of male sealskins taken in early season and late 
season. Shaded column = early season ; unshaded = late. Data from bottom 
row in tables 7 and 8 ; based on 105,249 skins taken in early season and 84,277 
taken in late season. 



assume that seals of comparable length are killed each year. This 
is a reasonable assumption, in view of the fact that only seals measur- 
ing between certain established length-limits are taken by the killing 
crew. ) In 1949, an answer to the same question was sought by other 
means. During a 41-day period, 17 June-27 July, 558 bachelor skins 
were segregated according to field length of seal. The range in 
length was 38 to 51 inches. Each skin was weighed in freshly 
blubbered and wrung condition, without mask and flippers. The 
observed weights did not change appreciably, within a length class, 
from early to late season. The mean weights for the 558 skins are 
(pounds) : first quarter, 5.4; second quarter, 5.3; third quarter, 5.5; 
fourth quarter, 5.1; mean for season, 5.3. 

In the foregoing paragraph I have shown that, for bachelors 
at least, the skin areas and skin weights of early-season individ- 
uals are no greater than those of late-season individuals. Now 
attention is called to table 9, in which the whole body weights of 



THE PRIBILOF SEALSKIN INDUSTRY 



47 



early and late arriving females of known age are given. Here the 
conclusion is inescapable that the earlier individuals, within an age 
class, are heavier. The conclusions drawn in the two paragraphs are, 
of course, not incompatible. 

Relation between field length of the bachelor seal and weight of 
the fresh pelt is shown in table 10 and figure 5. The greatest vari- 
ation in skin weight is exhibited in the 44-inch animal. Seals with 
a field length of 44 inches represent 3- and 4-year-olds in almost 
equal numbers. Greatest variation in body size is commonly exhibited 
during the years of adolescence in fur seals as well as other mammals. 
Scheffer (1950 d, p. 388) found the highest coefficient of variation in 
weight of fur seal testes in ages 3 and 4. 

Relations between field length of bachelor seal and size classification 
of skin in raw-salted and in finished condition are shown in tables 
11 and 12 and in figures 6 and 7. 

An attempt to translate a given trade classification, such as "small 
medium," into dimensions of skin by inches has been made in tables 
13 and 14. (The shape and size of the grading board has previously 



POUNDS 
10 r- 



8 







♦ fit* 



I 




VA 



J |_l I I I L 



38 40 42 44 46 48 50 
INCHES 



Figure 5. — Weight of fresh male sealskin with relation to field length of seal. 
(Data from table 10; vertical lines = range, circle = mean.) 



48 



THE PELAGE 



PERCENT 

100 i- 




SIZE: SM 



M 



XL 



XXL 



D 



Figure 6. — Trade classification of raw, salted male sealskin with relation to field 
length of seal. (The "field length" in inches, snout to tip of tail on unskinned 
animal, was obtained from 523 subadult male seals, mostly ages 3 and 4 years, 
sampled at random between 18 June and 20 July 1946, on St. Paul Island. The 
fate of the pelt from each seal was followed through processing by the Fouke 
Fur Company. About one year after the kill, each pelt was classified in raw, 
salted condition according to trade size, that is : small medium, medium, large, 
extra large, or extra extra large. In the present figure, the left-hand column 
(for example) shows that, of all 41-inch seals sampled, 57% produced a small 
medium pelt, 36% medium, 5% large, and 2% extra large. ) 



been mentioned. ) The length limits given in table 14 do not entirely 
agree with those given by Bachrach (1946, p. 520). His limits are: 
small medium, below S8y 2 inches; medium, S8y 2 to 4Qy 2 ; large, 42^ 
to 45% ; extra large, 45V& to 4tSy 2 ; extra extra large, 48 1 / £ to 55 ; and 
wigs, over 55. 

In 1956 the Government began for the first time to take female 
seals in substantial numbers as part of a long-term plan. In that 
year, there were taken 22,680 females by 15 August and 26,884 



THE PRIBILOF SEALSKIN INDUSTRY 



49 



PERCENT 
100 i- 

80 - 

60 - 

40 



20 







IN-- 41 42 43 44 45 46 47 48 49 50 51 M 



size-- sm| 



M 



L 



xlQ xxlQ 



Figure 7. — Trade classification of finished, dyed male sealskin with relation to 
field length of seal. ( For explanation see fig. 6. ) 



females by 8 September. Quantitative data on female skins (size, 
weight, relation to field length and trade classification) will not be 
available for several years. Skins taken from old females in the 
1956 and 1957 seasons were generally of poor quality, with fur sparse 
and uneven in length on back of neck and belly (plate 84). As 
compared with bachelor skins, the female skins were said to be more 
elastic. During the period 4-8 September 1956, a tally was kept of 
female pelts rejected from the kill as being commercially unattractive. 
Out of 4,807 females killed, the pelts of 603, or 12.5 percent, were 
rejected ( Abegglen and others, 1956, tables S and T) . All age-classes 
from 2 to 10-plus were included, though about half of the seals killed 
were in classes 2 to 5 years. While admitting that the skins of old 
females killed in August may be difficult to process, I am inclined to 
believe that part of the difficulty stems from the fact that routine 



50 THE PELAGE 

processing methods are being applied to a new and unusual class of 
skin. (See also discussion of molt on page 26ff.) 

The weights of tanned pelts in the National Museum collection — 
3 adult males and 4 adult females — indicate that the male pelt 
without flippers may weigh up to 10 pounds; the female pelt with- 
out flippers, up to 4 pounds. These pelts were prepared for study, 
not for commercial use. 

STRENGTH AND DURABILITY OF SEALSKINS 

Bowker (1931) measured the strength and thickness of commer- 
cial sealskins dyed black or brown. Although not so specified, the 
skins were certainly those of bachelors (3- or 4-year males), buffed 
in the routine way. Results: thickness, 0.023 to 0.026 inches; 
breaking strength of a half- inch strip, 37.5 to 46.0 pounds; tensile 
strength, 3,150 to 3,680 pounds per square inch; percent stretch at 
failure, 23.5 to 38.1 ; stitch-tear, 6.1 to 9.3 pounds. Sealskin leather, 
though slightly weaker, compares favorably with the leather of 
light calfskin and sheepskin. 

Terao (1940) examined the leather of sharks, moray, California 
sea lion, common dolphin, sperm whale, and blue whale. 

The fibrous matter of the leather was found mostly running obliquely 
longitudinally and united with transversely directed one so that it forms an 
irregularly shaped rhombic network. The general feature of the network is 
evident in tbe configuration of the upper-surface which is so characteristic to 
each species as to aid its identification . . . The findings stand in harmony 
with the fact that the leathers of aquatic animals are liable to be torn trans- 
versely when used as boots, bags, etc. 

L. A. Hausman, who studied for many years the microscopy of 
hair, published (1939, p. 503) a table showing "durability" of com- 
mercial furs. He ranked highest those species having fur hairs with 
little or no medulla, as follows: Grade 100, otter and wolverine; 
grade 90, beaver ; grade 80, fur seal ; grade 70, skunk and mink ; grade 
65, raccoon (natural) ; grade 50, raccoon (dyed) ; grade 45, muskrat; 
grade 40, fox (natural) ; grade 35, oppossum; grade 25, fox (dyed), 
ermine, nutria, and lynx; grade 15, chinchilla and goat; grade 5, 
rabbit, hare, and mole. 



OTHER FEATURES OF THE 
SURFACE TOPOGRAPHY 

Features of the Head 

NOSTRILS, MOUTH, AND LIPS 

These features have been examined on four selected fetuses, and 
on newborn and adult seals, as indicated below. ( "MNW" = mean 
newborn weight; see page 10.) 
Fetus of 23.7 g. (0.0049 MNW), female, 14 February 

The nostrils are slightly open; mouth open and tongue showing. 
Fetus of 103 g. (0.021 MNW), female, 20 January 

Nostrils and mouth are open (plate 70-A) . 
Fetus of 1.42 kg. (0.26 MNW), male, 22 March 

Between the parted lips, 20 low, softly rounded bumps along the 
gumline show where most of the teeth will appear. On each side of 
the mouth there are 4 bumps above and 6 bumps below. 

Fetus of 3.31 kg. (0.69 MNW), female, 6 July 

On each side of the mouth the tips of the following teeth have 
erupted: (above) 3 incisors, 1 canine, and 3 cheek teeth; (below) 
2 incisors, 1 canine, and 2 cheek teeth; a total of 12. The third 
lower cheek tooth, which will complete the deciduous set, has yet 
to appear. 

On newborn and older seals, the rhinarium and nostrils are dark 
gray, near black (plate 98-A). The dark color continues into the 
tunnel of the nares as far as one can see, or 1-3 cm., depending on the 
size of the seal. The upper and lower lips are medium gray. The 
mouth and tongue are flesh pink, becoming red in an overheated indi- 
vidual and fading to dirty grayish pink after death (plates 55 and 
98-B). 

The tip of the tongue has a curious notch, 4 or 5 mm. deep in 
adults — a feature of all pinnipeds except the walrus. 

EYELIDS, EYE GLANDS, AND IRIS 

The eyelids are open on a fetus of 822 g. (0.17 MNW), female, 9 
March, and on all larger fetuses (plates 34, 36, 41-A, 52-A, and 55). 

51 



52 OTHER FEATURES OF THE SURFACE TOPOGRAPHY 

The eyelids are closed on smaller specimens. On newborn to adult 
seals, the eyelids are streamlined. Each is a low-relief extension of 
the facial coat; thick, muscular, and hairy (plate 99). Wide open, it 
frames a nearly circular aperture and is rimmed (in the newborn) by 
a 2- to 4-mm. wide band of naked, black, thick, wrinkled skin. The 
diameter of the eye opening is about 11 mm. in the newborn and 20 
mm. in the adult. 

Pinnipeds have tear glands but no ducts leading to the nasal pas- 
sage. According to Rabsch (1953, p. 488) the lachrymal gland is 
small, the Harderian gland large, and the individual fornix glands 
especially well developed. Alaska fur seals "cry" freely in warm, dry 
weather ; the tears run down the cheeks (plate 100) . 

The iris of the eye is dull bluish gray when the fetal eyelids open 
for the first time. It is black or brownish black at birth, and there- 
after. The eyeballs of two 4-year-old males are 43 mm. and 46 mm. 
in diameter. 

EARS 

On a fetus as small as 23.7 g. (0.0049 MNW), female, 14 February, 
the ears are well formed and nearly adult in shape. A groove or 
flexure persists for 4 to 6 weeks around the base of each ear. In a 
fetus of 1.11 kg. (0.23 MNW), female, 23 March, the groove is begin- 
ning to disappear and the ear pinna is becoming more cylindroid and 
stiffer (compare plate 101-A; also plates 34, 36 and 41-A). Thus, 
the ear is becoming more like that of a pinniped and less like that of a 
land carnivore. 

On an adult seal, the inside of the ear pinna, which can only be seen 
when the furled edges are pried apart, is smoky grayish brown, smooth 
and glossy (plates 52-A, 55, and 101-B). Externally, from the base 
toward the tip, the fur disappears and the overhairs become shorter, 
more slender, sparser, and lighter in color. The epidermis of the 
outer surface of the ear is black. The blackness is conspicuous at the 
tip and along the distal third of the ear, where the original fine hairs 
have been lost through abrasion. The pelage beneath and just behind 
the ear is paler than the surrounding hair. (Description of an old 
female; additional notes on the pelage of the ear were given in the 
preceding chapter.) 

On the evidence of 201 measurements (table 15), the growth in 
length of the ear has virtually ceased by the eighth year of life. 
Growth during postnatal life represents an increase of about 35 per- 
cent in males and 30 percent in females. 



FEATURES OF THE BELLY 53 

Features of the Belly 
MAMMARY GLAND COMPLEX 

The teats or nipples on the skin of the female and male will be 
described first, then the mammary glands of the female. On a female 
fetus of 23.7 g. (0.0049 MNW), 14 February, the position of each of 
the four mammary teats can clearly be seen as a white dot beneath 
the surface of the skin. Each dot eventually becomes a pimple, then 
a pimple within a dimple (plate 36). Vellus of the belly grows up 
around the circular, naked, flesh pink dimple. On a living, full-term 
fetus of 3.31 kg. (0.69 MNW), female, 6 July, the anterior teats are 
70 mm. apart, on a line 58 mm. anterior to the navel ; the posterior 
teats are 43 mm. apart on a line 26 mm. posterior to the navel. Each 
teat is hidden in the pelage, though its location is marked by a light 
gray dot of hair. None of the teats on the female has, at time of 
birth, risen above the black hairy coat of the belly. The anterior and 
posterior pairs of teats develop at the same rate and seem to be equally 
important throughout life. On no fetus is there evidence of rudi- 
mentary (vestigial) teats. 

The teats become more conspicuous with advancing age. On a 
silver pup as well as on older, but not yet parous, individuals, the 
teat locations are indicated by faint brown spots. Occasionally they 
can be found only after the guard hairs have been parted with one's 
fingers. They are conspicuous, dark brown spots on the pelage of 
parous females, both in and out of the nursing season (plates 102 and 
103, A and B) . On a fully adult, lactating female, the loose, blackish, 
wrinkled, naked teats can be stretched with one's fingers to a length 
of 25 mm. On a large old female, the teats were arranged as follows : 
2 lying 20 cm. apart on a line 25 cm. anterior to the navel and 2 lying 
9 cm. apart on a line 6 cm. posterior to the navel. It has been found 
that the measured distance between levels of anterior and posterior 
teats on the tanned pelt is worthless as a criterion of age. For 
example, this distance on a silver pup, length 76 cm., is the same as on 
an 8-year-old, length 125 cm. 

Of the tanned skins of 40 females examined with special attention 
to the mammary teats, 2 have an extra (fifth) teat. On the pelt of a 
5-year-old, there are 2 posterior teats on the left side, 10 cm. apart, 
aligned with the long axis of the body. On a 10-year-old, there are 
2 posterior teats on the right side, 11 cm. apart, similarly aligned. A 
10-year-old examined in the field was noted as having "5 nipples in 
use." 

On the male fur seal, the 4 teats are always hidden by the pelage. 
They are arranged around the navel as on the female (plate 104). 
On a fresh carcass of a 6-year-old, the teats were arranged as follows : 



54 OTHER FEATURES OF THE SURFACE TOPOGRAPHY 

2 lying 19 cm. apart on a line 20 cm. anterior to the navel and 2 lying 
10 cm. apart on a line 4 cm. posterior to the navel, 12 cm. anterior 
to the penial opening. One can occasionally locate the male teats 
by finding faint brownish or blackish spots on the pelage, though 
more often one must wait until the pelt has been removed and tanned ; 
then search for scarlike marks on the leather side (plate 105). 

On plucked pelts of both sexes, the location of the mammary teat 
is marked by a slight break or depression in the fur surface. Here 
the color is uniform light brown, like the surroundings. 

With the hope of finding a relation between the underlying mam- 
mary glands and the sparse, often patchy underfur of the belly, 
in September 1958 I examined the glands on a number of fresh 
carcasses. The gland-complex proves to be an extensive apron cover- 
ing the lower thorax, abdomen, and sides of the body. The sketch 
reproduced in figure 8 is based mainly on dissection, with photo- 
graphs, of an old female, age over 10 years, not lactating, killed 
on 9 September (plate 106-A). It is also based on cursory field dis- 
section of 4 copiously lactating individuals. In the old female speci- 
men, 2 liters of embalming fluid were introduced through the 4 teats. 

Using a knife, one can easily separate the glandular tissue from 
the overlying blubber. The glandular tissue is light brown and more 
fibrous; the blubber is whitish and less fibrous (plate 106-B). The 
blubber at the level of the anterior teats is 2 cm. thick ; the glandular 
tissue is 1 to 2 cm. thick. These two layers are, at the level of the 
posterior teats, 2 cm. and 1 to 1.5 cm., respectively. On the old female, 
length 137 cm., the gland-complex is about 66 cm. long and 50 cm. 
wide, along curves of the body. Posterior to the armpits the gland 
bends rather abruptly upward along the sides; anterior to the heels 
it bends less abruptly upward. It reaches forward about 21 cm. 
beyond the level of the anterior teats or almost to the level of the 
insertion of the fore flippers. It reaches hindward about 20 cm. be- 
yond the posterior teats, or almost to the level of the heels. Its total 
area is 2,189 sq. cm. (339 sq. in.). The gland-complex appears to be 
one continuous sheet, though it probably consists of 4 anastomosing 
units, 1 unit draining into each teat. (When embalming fluid was 
injected, the region around each teat swelled independently.) The 
main duct leading to the teat is thin-walled, about 5 mm. in diameter, 
and is conspicuous for only 6 cm. of its length, the deeper portion 
being buried in glandular tissue. The shape of the milk reservoir be- 
neath each teat is suggested in plate 106-A. At its margin, the gland- 
complex becomes very thin, not over 2 or 3 mm. A shipbuilder 
would say that it is "faired" into the body. Carlisle (1954) believes 
that mammary glands may have originated from sebaceous glands, 
rather than sweat glands. 



FEATURES OF THE BELLY 



55 




Figure 8. — Diagram of the mammary gland complex ; represented on a flat sur- 
face with the side flaps lifted to the plane of the thorax and belly ; reservoir 
beneath each teat swollen with embalming fluid ; distance between anterior 
teats 24 cm. ; letters represent anterior mammary teats, navel, posterior mam- 
mary teats, and vaginal opening. (4189) 



56 OTHER FEATURES OF THE SURFACE TOPOGRAPHY 

Milk capacity of the mammary gland is perhaps 2 or 3 liters. 
Milk in the bulging stomach of a nursling has been found to vary 
in amount from 1 liter in a newborn to 2.5 liters in a pup nearly 
weaned (Scheffer, 1950 c, p. 7; and Ford Wilke, MS, 1941, based on 
specimen BDM 6). To a limited extent, milk may flow spontane- 
ously from the mother. On 27 September the writer saw a cow of 
medium size scrambling over the rocks in alarm. Five to ten drops of 
milk fell in rapid succession from her left anterior teat. Wilke col- 
lected milk as it drained from the slashed mammary gland of a female 
in estrus; he later reported (1959) that it contained 46 percent fat. 

(The milk of the gray seal Halichoerus may contain up to 53 per- 
cent fat, while the suckling elephant seal Mirounga may quadruple 
in weight in 21 days ! ) 

PENIAL OPENING AND SCROTUM 

On the smallest male fetus available for study, 131 g. (0.024 MNW) , 
25 January, the site of the future penial opening is a distinct dimple 
about 0.5 mm. in diameter. On a fetus of 2.21 kg. (0.41 MNW), 
2 May, the penial opening is a craterlike prominence surrounded by 
long hairs, with a distinct opening 1 mm. in diameter. On the new- 
born pup, the opening is usually marked by a fringe of white 
hairs. On the adult, the skin of the penial opening is distinct, 
naked, and flesh pink surrounded by black (plate 57). 

The scrotum of the adult is dark gray, wrinkled, and virtually 
hairless. The testes descend in the third or fourth year, though the 
descent is less conspicuous than in land carnivores (plates 57 and 
107-A). 

FEMALE EXTERNAL GENITALIA 

Of the female external genitalia, Bartholomew and Hoel (1953, 
p. 420) have written: 

In the interval between parturition and the end of estrus [about 1 week] 
the vulva is swollen and protruding and the vestibular mucosa and vaginal 
orifice are a brilliant pink. By the time a female returns from her first trip 
to sea, the swelling is completely gone, and neither vestibular mucosa nor 
vaginal orifice are conspicuous. The entire vulva appears dark brown or black. 

On an adult female shot 18 minutes after copulation, the pink 
rosette of the vaginal opening was clearly visible from a distance of 
100 meters. Ordinarily the vestibule appears black. (Plate 108.) 

NAVEL AND TAIL 

The navel can nearly always be seen in both sexes, regardless of 
age, as a disturbance in the lay of the hair, without distinction in 
color (plates 36, 44, and 102). 



FEATURES OF THE LIMBS 57 

On a fetus as small as 23.7 g. (0.0049 MNW), female, 14 February, 
the tail is well formed and nearly adult in shape. The tail is dif- 
ficult to measure; its free length is 15 to 20 mm. on the newborn 
pup and 30 to 50 mm. on the adult (plates 44, 58, and 107-B). On 
the field record of a 4-year-old male is noted "tail 32 mm. plus 4 mm. 
hairs, though standard length is about 62 mm. when loose skin is 
pushed toward the body." As the seal reaches old age, the tail 
pelage becomes darker — at times almost black — and the underfur 
becomes sparser. (For further description of tail pelage, see pre- 
ceding chapter.) 

Features of the Limbs 

FLIPPERS AND CLAWS 

At birth, the flippers are brownish gray (Munsell 5 YR 4/1) and 
naked, except for sparse vellus on the dorsal surface of the fore 
flipper (plates 43 and 44). They soon turn almost black (plate 109). 
When wet, they are intensely black; when dry, dark grayish brown 
(5 YR 2/1) ; when very dry, as on an old bull sleeping in the sun, 
gray (plate 110). In a group color photograph, the flippers of cer- 
tain individuals may catch and reflect the light of the sky, appearing 
white or bluish. 

Forty flippers from 10 subadult male seals were cured in salt on 
St. Paul Island, 20 June 1949. (A set of 4 flippers weighs 2.5 to 
3 lb. in fresh condition.) Half of the collection was sent to a large 
national manufacturer of glues and half to a large national producer 
of photographic gelatin. The former reported (personal correspond- 
ence) that "the resulting glue liquors were dark in color and had 
a characteristic 'fishy' odor. The glue was of low test and the per- 
centage of glue obtained was much lower than for green salted stock 
obtained from cattle or other skin trimmings." The gelatin manu- 
facturer reported "we have conducted some extensive tests with this 
material but our experiments disclose that it has no value in our 
operations." 

The National Bureau of Standards also examined a barrel of 
flippers (U.S. Bureau of Fisheries, 1920) : 

The experiments made by that bureau showed a yield of glue amounting to 
G7 percent of the weight of the salted flippers. The viscosity at 40° C. of a 
10 percent solution was . . . 1.20, a little below Peter Cooper's glue, grade 1%. 

The only functional claws of the fur seal are those on the 3 inner 
toes of each hind flipper. The seal uses its claws exclusively for 
grooming the anterior parts of the body. The pup, especially, 
scratches itself frequently and vigorously during the summer molt. 
Development of the claws on five selected fetuses is described as 
follows : 



58 OTHER FEATURES OF THE SURFACE TOPOGRAPHY 

Fetus of 23.7 g. (0.0049 MNW), female, 14 February 

The sites of all 20 claws are visible. Each of the fore-claw sites 
on digits 1 to 4 is visible as a soft, whitish disc or pad of skin with 
a minute, teat-like projection. The outer hind claws are similar. 
The 3 inner hind claws, destined to become the only functional claws, 
are represented now by the largest of all the whitish pads. The claw 
rudiment on digit 5 of the fore nipper is a roundish disc below the 
surface of the skin and here it remains; it does not erupt with ad- 
vancing age of the fetus. 

Fetus of 103 g. (0.021 MNW), female, 20 January 

On each fore nipper, the first claw rudiment is distinctly conical 
and clawlike, though soft and white, 1 mm. in length. The second 
to fourth rudiments decrease in size in this order. They are also 
clawlike. The fifth is no more than a faint pimple. The hind claws 
are well formed; the middle 3 robust and about 2 mm. in length. 
(Compare plate 36.) 
Fetus of 1.09 kg. (0.20 MNW), male, 25 March 

The 3 middle claws on each hind flipper are becoming firm and 
horny ; brown with lighter tips and under parts. 

Fetus of 3.31 kg. (0.69 MNW), female, 6 July 

A full-term fetus, delivered by caesarian section and killed 9 hours 
later. The vestigial claw sites on each fore flipper are scar-like pits 
with diameters as follows (first to fifth digits) : 1, 0.5, 0.5, 0.5, and less 
than 0.5. On each hind flipper, the claws are grayish in color, with 
distal one-third to one-half thin, flexible, and semicircular in cross 
section. They may be described as follows: (first) vestigial, length 
10 mm., basal 3 mm. stiffer, distal 7 mm. softer, paper-thin; (second) 
length 20 mm.; (third) length 19 mm.; (fourth) length 18 mm.; 
(fifth) vestigial, length 5 mm. After birth, the first and fifth claws 
quickly become worn to nubbins. 

Fetus of 5.79 kg. (1.07 MNW), male, 9 July 

The hind claws on this full-term fetus are nearly as long as the 
longest recorded for the newborn: 3, 26, 26, 25, 10 cm. The claws 
dry out soon after birth, when they become brownish black along the 
basal part and horn-brown toward the thinner tip. 

On adults, the third (middle) hind claw is usually the longest. 
On a 10-year-old male, the fore claws are rudiments hidden in skin 
pits 1 to 2 mm. in diameter. The hind claws are 9, 29, 30, 26, and 4 
mm. in length on the first to fifth digits, respectively. On a 10-year- 
old female, the fore claws are also rudiments. The hind claws are 6, 
23, 24, 19, and 7 mm. in length. I do not know whether the claws 
continue to grow in later life, but I presume that they do not. 
Their position on top of the hind flipper protects them from the 



THE BLUBBER LAYER 59 

sort of attrition to which the claws of a -cat or dog are exposed ; yet 
the claws of the fur seal never exceed 30 mm. in length. (Plates 110 
and 111, A and B.) 

The Blubber Layer 

The blubber on the belly of a newborn fur seal was 3 mm. thick; 
on the belly of an old cow, 50 mm. ; and on the belly of an old bull, 
60 mm. 

On breeding bulls, some of them known to fast for at least 64 
days in summer, the skin is loose and wrinkled by mid-August. 
Clearly there has been a loss of subcutaneous fat, though no actual 
measurements of the loss have been made. McLaren (1958, p. 63) 
found in the ringed seal Pusa that blubber made up about 45 percent 
of the body weight in winter, but only 20 percent in late June. 

A reduction plant on St. Paul Island utilizes the byproducts of 
the Pribilof sealskin industry. The skinned carcasses are reduced 
to dry meal and "carcass oil", while the blubber scraped from the 
skins is reduced to "blubber oil". The yield of both kinds of oil 
over a 10-year period is shown in table 16. A subadult male seal 
weighing about 30 kg. (66 lb.) yields roughly 0.6 gal. of blubber 
oil. Apparent variation from year to year in yield of oil per skin 
is believed to result, in a large part, from inconsistent handling of 
materials in the plant. Additional data, from records kept from 
1950 to 1952, are as follows: 1 subadult male sealskin yields about 
0.13 cu. ft. of crude blubber, and 1 cu. ft. of crude blubber yields 
4.3 gal. blubber oil. Crude blubber is the plumped-up mass of fat, 
connective tissue, and muscle scraped from sealskins that have been 
allowed to soak overnight in cold sea water. On 25 July 1947, 
I weighed the crude blubber from 100 skins of "Group III" 
seals (mostly 3-year males) and from 100 skins of "Group IV" seals 
(mostly 4-year males). The average weights per skin were 11.7 
lb. and 15.2 lb., respectively. 

Thompson (1950, p. 726) gave an analysis of oil which had been 
heat rendered on St. Paul Island in 1949 from blubber scraped from 
sealskins : 

Free fatty acids (as oleic) 1.05% 

Moisture 0.34% 

Insoluble matter 0. 05% 

Iodine number (Wijs) 132.5 

Stearin at 70° F. (21.1° C.) 1.04% 

Unsaponifiable matter 0. 49% 

Titer 70. 2° F. 

(21.2° C.) 
Lovibond color using a 1-inch column : 

Yellow 20 

Red 3. 5 

553006 0—62 5 



60 OTHER FEATURES OF THE SURFACE TOPOGRAPHY 

Clegg (1951) reported on the characteristics of oil from cold- 
rendered fur-seal blubber as follows : 

Free fatty acid (AOAC) 1. 58% 

Moisture and other volatile matter (AOAG) not measurable 

Iodine number (AOAC, Hanus) 108 

Unsaponifiable matter (AOAC) 0.64% 

Saponification number (AOAC) 196.3 

Specific gravity 25° C./25° C 0. 917 

Index of refraction at 25° C 1. 4743 

Vitamin A content (1894XE328) 1 306 units/g. 

1 Determined on the whole oil without saponification. 

Minato (1949) gave additional data on "body oil" of a fur seal 
taken off Japan. Miyauchi and Sanford (1947) found little vitamin 
A in two samples of fur-seal blubber oil : 490 and 493 units per gram. 

In a sample of fresh seal blubber, Wilber (1952) found total 
phospholipids 4.3 percent and total cholesterol 0.24 percent. The 
phospholipid content is high — as in perhaps all marine mammal f ats — 
and is conceivably "associated with a high rate of fat turnover at the 
depots." 

The sample analyzed by Wilber was reddish orange. The carcass 
of a seal with orange blubber, called by natives of the Pribilof 
Islands a "salmon-eater", may occasionally be seen on the killing 
field among carcasses with normal, creamy white blubber. Orange 
individuals include fewer than 1 percent of the total. Wilber con- 
cluded that the orange color is, in all probability, a carotenoid 
pigment. It may, in fact, be astaxanthin, a pigment found by 
Baalsrud (1956) in an abnormal, reddish-orange cod. Baalsrud 
stated that "There is as yet no obvious explanation for this sporadic 
appearance of astaxanthin in cod flesh, but somewhat similar ob- 
servations have been made on whales. Occasionally red (asta- 
xanthin-containing) whale-body oils and whale-liver oils are encoun- 
tered; in this case pigmentation is taken to indicate an unspecified 
pathological condition." 

Experts of the Fouke Fur Company say that an indistinct dark 
band can be seen along the dorsal region of the finished, dyed 
sealskin from old female animals. This is not present in male skins. 
It is referred to as an "oily band," though its real nature is unknown. 
Actually, it may be correlated with age rather than with sex; that 
is, only younger males commonly enter the commercial kill, as against 
females of all ages. 



SUMMARY 

The midsummer population of northern fur seals Callorhirms 
ursinus is estimated at 1,978,000. Of this number, 1,800,000, or 91 
percent, originate on the Pribiliof Islands. The Pribilof herd is 
capable of yielding 80,000 to 100,000 sealskins a year. 

The epidermis of the sealskin has a cornified layer about 15 microns 
thick and a generative layer about 25 microns thick. The dermis 
or leather is 3 to 4 mm. thick, thickest on old males. Regardless 
of age, the follicular (hair-rooot) stratum of the dermis does not 
vary appreciably from 2.3 to 2.4 mm. in thickness. An apocrine 
sweat gland is associated with each hair bundle, rising from depths 
of 2 to 3 mm. Sweat glands are not functional in the black pup. 

The skin is pigmented (dark gray) on the following naked parts: 
nose, lips, rim of the eye, inside of the ear pinna, penial opening 
scrotum, vestibular opening, teats, anus, and flippers. It is also 
pigmented (light gray) beneath the pelage of the ear pinna and 
tail. 

The pelage consists of bundles, rather evenly distributed, about 
15 per sq. mm., each bundle with a guard hair that slopes, rooflike, 
above, and anterior to, a group of 35 to 40 underfur fibers. At the 
skin surface, the bundle is surrounded by a thick sheath. The guard 
hair is flanked by a pair of sebaceous glands seated 0.8 to 1.0 mm. 
below the surface ; functional even in the black pup. There is no hair- 
erecting muscle. The guard hair is most deeply rooted. Behind it, 
the roots of the individual fur fibers are arranged in stairstep fashion, 
the most posterior fibers nearest the surface. The total number of 
hair and fur fibers per sq. mm. is about 570 (370,000 per sq. in.). 
The pelage is highly efficient as a thermal insulator ; the underfur is 
water repellent. The haired surface of the body of the adult male has 
an area about 2.5 times that of the female. 

Guard hairs may attain a length of 33 mm. on the male (mane 
hairs to 70 mm.) and 20 mm. on the female. Underfur hairs may 
attain a length of 14 mm. on the male and 13 mm. on the female. 
The pelage of the belly is shorter than that of the back. The individ- 
ual guard hairs and underfur hairs resemble those of mink; less 
closely, those of otter. The guard hair of the seal is stiff and flattened, 
lanceolate, medullated, heavily pigmented as a rule, with diamond- 
petal scale-pattern along most of the shaft. The underfur is much 

553006 O— 62 6 61 



62 SUMMARY 

finer, distinctly wavy, without medulla, with very little pigment, and 
with pectinate scale-pattern along most of the shaft. 

The fur-seal embryo implants in early November. Hair primordia 
are visible on a midwinter fetus of y 2 oo mean newborn weight, and 
erupted hairs on a midwinter fetus of y 20 mean newborn weight. 
Finer, paler hairs erupt in advance of coarser, blacker ones; the 
early fetus is gray. Heavy black guard hairs are aligned in a streak 
along the back of the fetus of 1/3 mean newborn weight. Waves 
of hair-growth move backward and downward from the head and, 
slightly later, forward and downward from the rump. The top of 
the fetal flipper is covered with fine hairs which disappear at birth, 
though sweat glands remain. A fetus of 2.7 kg. (y 2 mean newborn 
weight) is well covered, except on the belly, with black hair. 

The birthcoat is jet black, and its fibers are mature. It consists 
of 75 to 80 percent underhairs and 20 to 25 percent guard hairs. The 
arrangement of hair bundles is quite unlike that in the adult coat. 
In the birthcoat, there are 40 to 45 small bundles per sq. mm., each 
containing 1, 2, or 3 hairs. When the bundle includes 2 or 3 hairs, 
the posterior ones are underhairs. 

The pup guard hairs are about 15 to 20 mm. in length. To a 
certain extent, they intergrade in size and structure with the under- 
hairs. The largest ones resemble adult guard hairs. The pup un- 
derhairs are 6 to 15 mm. in length. They differ from adult underfur 
fibers in being shorter, coarser, less wavy, with more pigment, and 
often with medulla. In the birthcoat, adult-type underfur follicles 
are already taking form, deep in the dermis. 

The height of the pupping season is 15 July. Two weeks later, 
the black birthcoat has started visibly to molt and has taken on a 
ragged appearance. By the end of September, it has largely been 
replaced by the adult-type pelage of the silver pup. The silver 
pelage persists for about 11 months, or until the end of August of 
the second summer. 

The annual molt of the seal is believed to be a prolonged affair, 
4 to 5 months from start to finish. The first adult-type molt (to 
silver pup) centers in September; the second in August (to yearling 
of autumn) ; the third in September (to 2-year-old of autumn) ; the 
fourth and subsequent molts in late September or October. The 
fibers are shed individually, with the result that the pelage is always 
intact and ready to meet the demands of an amphibious life. One 
rarely sees a "molt line." Molting is not accompanied by sloughing 
of the epidermis as in certain phocids. In autumn, the coat is duller 
and browner; in spring (at sea) it is brighter and more silvery. 
In late adolescence, ages 3 to 5 years, the whitish rump patches dis- 
appear, mane and wig hairs of the male start to lengthen, and pig- 



SUMMARY 63 

ment ceases to form at the roots of the whiskers (vibrissae) in both 
sexes. 

At birth, there are 2 superciliary (above-the-eye) and 20 to 23 
mystacial (snout) vibrissae on each side of the head. The number 
does not vary with age or sex. Nearly all of the vibrissae have 
erupted on a fetus of y 2 oo mean newborn weight, far in advance 
of the body hairs. Like sheep-wool fibers, the vibrissae are ever- 
growing, even in spring when the body-hair follicles are at rest. 
They continue to grow, so far as we know, throughout life. On a 
seal approaching the end of the first year of life, the longest vibrissa 
is growing at the rate of 0.2 mm. per day. The vibrissae on the 
adult are 4 to 5 times longer than those on the newborn. The longest 
vibrissa of record has a maximum length on the male of 334 mm. 
(13.1 in.), and on the female 220 mm. (8.66 in.). The vibrissae 
are the only fur-seal hairs that taper all the way from base to tip. 

The most commonly observed pelage anomalies are the following: 
(1) Genetic color- freaks such as albino, piebald, and chocolate. (2) 
"Rubbed" pelage with guard hair absent from extensive areas; 
etiology unknown. (3) Naked, scabby patches on lousy pups. (4) 
Thick-skin, or pachy derma; etiology unknown. (5) Warty nodules, 
thus far seen only on the flippers; etiology unknown. (6) Female- 
like pelage on males with infantile testes. (7) Algal and barnacle 
growths on the pelage, especially in winter. 

The Pribilof sealskin industry is now about 175 years old. "Island 
operations" include killing, skinning, blubbering (flensing), and cur- 
ing. "Factory operations" include unhairing, leathering, dyeing, and 
finishing. The skins receive a chamois, rather than chemical tannage. 
Field measurements of seals show that larger individuals tend to 
arrive on land in summer slightly ahead of smaller seals of the same 
age. However, as a result of selection by the killing crew, the com- 
position of the take (by size of skin) remains constant throughout 
the sealing season. The weight of the blubbered, subadult, male pelt 
varies from 3.2 lb (on a seal of 38-inch field length) to 8.5 lb. (on a 
49-inch seal). Commercial sealskins range in length and width as 
follows (inches): Raw, salted: length 28^4.5, width 19-30. Fin- 
ished, dyed: length 35.5-51.5, width 19-29. Thus, the finished skin 
gains considerably in length. Raw, salted sealskins are graded in five 
main categories, the smallest ("small medium") representing an area 
of about 718 sq. in. and the largest ("extra extra large") 1,261 sq. in. 
Breaking-strength tests applied by the Bureau of Standards reveal 
that sealskin leather compares favorably with the leather of light 
calfskin and sheepskin. 

On the fetus, the ear pinna is flattened and somewhat flexible, like 
that of many land carnivores; on the adult, it is more cylindroid, 



64 SUMMARY 

stiffer, and more specialized for submarine life. It ceases to grow in 
about the 8th year of life, with a length (from notch) of 50 mm. on 
males and 45 mm. on females. 

The 4 mammary teats (abnormally 5) are abdominal and equal in 
importance. On males and young females, they are hidden beneath 
the pelage. The mammary gland-complex is an extensive apron cov- 
ering the lower thorax, abdomen, and sides of the body. It attains a 
thickness of 2 cm. and an area of more than 2,000 sq. cm. (300 sq. in.). 
It has a milk capacity of 2 to 3 liters. A sample of milk contained 46 
percent fat. 

The testes descend in the third or fourth year. The tail is insig- 
nificant ; throughout life it scarcely doubles in length. On each fetal 
flipper, five claw primordia are clearly visible. However, only the 
three middle claws of each hind flipper become functional ; these are 
used exclusively for scratching the body. On an adult male, the 
middle claw (third digit) is about 30 mm. in length. 

The blubber may attain a thickness of 6 cm., considerably less than 
that of phocid seals of comparable size. The crude blubber on a sub- 
adult male weighs 12 to 15 lb., and yields roughly 0.6 gal. of oil. 
The free fatty acid component of blubber oil is high (1.05 to 1.58 
percent), as is the phospholipid content (4.3 percent). Reddish- 
orange blubber, occasionally seen, may contain astaxanthin. 

In appendix A, Munsell colors are given for 8 seals of selected age 
and sex. The middle value in an array of 39 colors recorded for 
fur-seal pelage is light brown. Seals from American and Asian 
waters are inseparable on the basis of color. Even to the trained ob- 
server, there are no sex distinctions in color pattern up to age 2 or 3 
years. By age 4, the male is beginning to show a grayish mane and 
wig, and to lose his rump patches. 






LITERATURE CITED 

Abegglen, C. E., A. Y. Roppel, and F. Wilke. 

1956-1958. Alaska fur seal investigations, Pribilof Islands, Alaska. 

Report of field activities June-September . . . Published annually by U.S. 

Fish and Wildlife Service, Seattle. Processed. 
American Medical Association Journal. 

1943. Can hair turn white overnight? 121 : 161-162. 
Aoki, T., and M. Wad a. 

1951. Functional activity of the sweat glands in the hairy skin of the dog. 
Science, 114 : 123-124. 

Aubeb, L. 

1952. The anatomy of follicles producing wool-fibres, with special reference 
to keratinization. Royal Society of Edinburgh, Transactions, 62 : 191-254, 
5 pis. (1 in color). 

Baalsrud, K. 

1956. Astaxanthin in the muscle of cod. Nature, 178 : 1182-1183. 
Bachrach, M. 

1946. Fur ; a practical treatise. Rev. ed. Prentice-Hall, New York 672 p. 
Baker, R. C. 

1957. Fur seals of the Pribilof Islands. U.S. Fish and Wildlife Service, 
Conservation in Action No. 12. 24 p. 

Bartholomew, G. A., Jr. 

1951. Summary of observations made by . . . Bartholomew ... on the 
social and reproductive behavior of the Alaska fur seal during June, 
July, and August, 1951. U.S. Fish and Wildlife Service, Seattle, MS 
report, 4 September, 3 p. 

Bartholomew, G. A., Jr., and P. G. Hoel. 

1953. Reproductive behavior of the Alaska fur seal, Callorhinus ursinus. 
Journal of Mammalogy, 34 : 417-436. 

Bassett, C. F., and L. M. Llewellyn. 

1948. The molting and fur growth pattern in the adult silver fox. Ameri- 
can Midland Naturalist, 39 : 597-601. 

1949. The molting and fur growth pattern in the adult mink. American 
Midland Naturalist, 42 : 751-756. 

Bassett, C. F., O. P. Pearson, and F. Wilke. 

1944. The effect of artificially increased length of day on molt, fur growth, 
and priming of silver fox pelts. Journal of Experimental Zoology, 96 : 77- 
83. 

Bergersen, B. 

1931. Beitrage zur Kenntnis der Haut einiger Pinnipedier . . . Skrifter 
Norske Videnskaps-Akademi i Oslo, Mat. -Naturvidensk. Klasse, 1931 : 
1-179, 22 pis. 

BlSSONNETTE, T. H. 

1935. Relations of hair cycles in ferrets to changes in the anterior hypo- 
physis and to light cycles. Anatomical Record, 63 : 159-168. 

1942. Anomalous seasonal coat-color-change in a small male Bonaparte's 
weasel . . . American Midland Naturalist, 28 : 327-333. 

65 



66 LITERATURE CITED 

Bissonnette, T. H., and E. E. Bailey. 

1944. Experimental modification and control of molts and changes in coat- 
color in weasels by controlled lighting. New York Academy of Sciences, 
Annals, 45 : 221-260. 
Bowers, D. E. 

1956. A study of methods of color determination. Systematic Zoology, 
5 : 147-160, 182. 
Bowker, R. O. 

1931. Some physical properties of fur-seal skins. Technical Association 
of the Fur Industry, Journal, 2 : 34-43. 

BR0NDSTED, H. V. 

1931. Bygninger af Snuden og Ansigtsmuskulaturen hos nogle Pinnipedier 
. . . K. Danske Vidensk. Selsk. Skrifter, Naturvidensk. og Mathem. Afd. 
9. Raekke, IV. 2, p. 41-85, 12 tbls. 

Carlisle, D. B. 

1954. On the relationship between mammary, sweat, and sebaceous glands. 
Quarterly Journal of Microscopical Science, 95 : 79-83. 
Carter, H. B. 

1939. A histological technique for the estimation of follicle population per 
unit area of skin in the sheep. Council for Scientific and Industrial 
Research, Australia, Journal, 12 : 250-258. 
Chase, H. B. 

1954. Growth of the hair. Physiological Reviews, 34: 113-126 (incl. bibl. 
of 78 titles). 
Clegg, W. 

1951. Characteristics of oil from cold-rendered fur seal blubber. U.S. Fish 
and Wildlife Service, Commercial Fisheries Review, February, p. 30-31. 
Danforth, C. H. 

1925. Hair. Journal of the American Medical Association, Chicago, 152 p. 
Fortune magazine. 

1930. The seal and the U.S. Treasury. November, p. 70-72, 122, 124. 
Fouke, P. B. 

1949. St. Louis: fur sealskin market of the world. American National 
Fur and Market Journal, 27 : 13, 76-77. 
Fouke Fur Company. 

1958. The romance of the Alaska fur seal. Published by the company, 
St. Louis. 48 p. 
Fur Trade Review. 

1916. First batch of seals dyed in St. Louis. May, p. 80-81. 
Green, D. D. 

1947. Albino coyotes are rare. Journal of Mammalogy, 28 : 63. 
Gunn, C. K. 

1932. Phenomena of primeness. Canadian Journal of Research, 6: 387- 
397, 2 pis. 

Hall, E. R. 

1951. American weasels. University of Kansas Publications, Museum of 
Natural History, 4 : 1-466, 41 pis. 
Hamilton, J. B. (Editor) 

1951. The growth, replacement, and types of hairs. Annals of the New 
York Academy of Sciences, 53 : 461-752 (27 articles, 27 authors) . 
Hardy, J. I. 

1935. A practical laboratory method of making thin cross sections of 
fibers. U.S. Department of Agriculture, Circular 378. 10 p. 



LITERATURE CITED 67 

Habdy, J. I., and Thora M. Plitt 

1940. An improved method for revealing the surface structure of fur fibers. 
U.S. Fish and Wildlife Service, Wildlife Circular 7. 10 p. 
Hausman, L. A. 

1939. Furs under the microscope. Nature Magazine, November, p. 501-503. 
1944. Applied microscopy of hair. Scientific Monthly, 59 : 195-202. 
Jordan, D. S., and others 

1898. Observations on the fur seals of the Pribilof Islands, 1872-1897 . . . 
[part 2, p. 250-606]. In The fur seals and fur-seal islands of the North 
Pacific Ocean . . . edited by David Starr Jordan. Govt. Print. Off., 
Washington, Treasury Department Document 2017, 4 parts, 1898-99. 
Krumbiegel, I. 

1954. Korperbedeckung [vol. 1, p. 39-68]. In Biologie der Saugetiere. Agis 
Verlag GmbH, Krefeld. 2 vols. 
Maerz, A., and M. R. Paul 

1950. A dictionary of color. Ed. 2, McGraw-Hill, New York, 208 p. Color 
plates. 
Mathiak, H. A. 

1938. A rapid method of cross-sectioning mammalian hairs. Journal of 
Wildlife Management, 2 : 162-164. 
Mathur, B. N. 

1927. Theory of oil tannage with special reference to seal oil. Journal of 
the American Leather Chemists' Association, 22 : 2-44. 
McLaren, I. A. 

1958. The economics of seals in the eastern Canadian arctic. Fishery 
Research Board of Canada, Circular 1. 94 p. 
Menges, R. W., and Lucille K. Georg 

1957. Survey of animal ringworm in the United States. U.S. Public Health 
Service, Public Health Reports, 72 : 503-509. 

Miller, M. E. 

1952. Guide to the dissection of the dog. Ed. 3., reprinted 1955. Published 
by the author, Ithaca, N.Y. 427 p. 
Miller, R. S. 

1958. The Munsell system of color notation. Journal of Mammalogy, 
39: 278-286. 

MlNATO, A. 

1949. Constituents of body oil from marine animals. Journal of the Phar- 
maceutical Society of Japan, 69 : 68-101. 
Miyauchl, D. T., and F. B. Sanford 

1947. Vitamin A content of fur seal oils. U.S. Fish and Wildlife Service, 
Commercial Fisheries Review, 9 : 5-8. 
Montagna, W. 

1956. The structure and function of skin. Academic Press, New York, 356 p. 
Montagna, W., and R. J. Harrison. 

1957. Specializations in the skin of the seal (Phoca vitulina). American 
Journal of Anatomy, 100 : 81-101, 6 plates. 

Munsell Color Company 

1954. Munsell soil color charts. Pub. by the company, Baltimore, 6 p., 9 
color plates, all loose-leaf. 
Murray, M. D. 

1958. Ecology of the louse Lepidophthirus macrorhini Ehderlein 1904, on 
the elephant seal Mirounga leonina (L) . Nature, 182 : 404-405. 



68 LITERATURE CITED 

Nakai, J., and T. Shida 

1948. Sinus-hairs of the sei-whale (Balaenoptera borealis). Scientific Re- 
ports, Whales Research Institute, Tokyo, No. 1, p. 41-47. 
National Bureau of Standards 

1955. The ISCC-NBC method of designating colors and a dictionary of 
color names. National Bureau of Standards Circular 553, 158 p. 
Noback, C. R. 

1951. Morphology and phylogeny of hair. Annals of the New York Academy 
of Sciences, 53 : 476-492. 
Odland, G. F. 

1954. Skin and epidermal derivatives [p. 428-457]. In Histology, edited 
by Roy O. Greep, with 13 contributors. Blakiston, New York, 12 + 953 p. 
Parnell, J. P. 

1951. Hair pattern and distribution in mammals. Annals of the New York 
Academy of Sciences, 53 : 493-497. 
Partridge, R. A. 

1938. A study of the lipids of fresh seal skin. Journal of the American 
Leather Chemists' Association, 33 : 144-156. 
Pearson, Anita K., and R. K. Enders 

1951. Further observations on the reproduction of the Alaskan fur seal. 
Anatomical Record, 111 : 695-712. 
Pocock, R. I. 

1914. On the facial vibrissae of Mammalia. Proceedings of the Zoological 
Society of London, 1914, p. 889-912. 
Poland, H. 

1892. Fur-bearing animals in nature and commerce. Gurney and Jackson, 
London, 66 4- 392 p. 
Rabsch, B. 

1953. Die Tranendrusen der Siiugetiere. Wiss. Z. Martin-Luther-Univ. 
Halle- Wittenberg, Jg. 2, Heft 8, Math.-natur. Reihe Nr. 4, p. 477-508. 
Rand, R. W. 

1956. The Cape fur seal Arctocephalus pusillus (Schreber) : its general char- 
acteristics and moult. Union of South Africa, Department of Commerce 
and Industries, Division of Fisheries, Investigational Report 21. 52 p. 
Roddy, W. T. 

1956. Histology of animal skins [p. 4-40]. In The chemistry and tech- 
nology of leather. Vol. 1. Preparation for tannage. Edited by Fred 
O'Flaherty, William T. Roddy, and Robert M. Lollar. Reinhold, New 
York, 2 vols. 

Rothschild, Miriam, and C. Lane 

1957. Note on change of pelage in the stoat (Mustela erminea L.). Pro- 
ceedings of the Zoological Society of London, 128 : 602. 

Samet, A. 

1950. Pictorial encyclopedia of furs/from animal land to furtown. Pub- 
lished by the author, New York. 474 p. 
Scheffer, V. B. 

1949. The clitoris bone in two pinnipeds. Journal of Mammalogy, 30 : 269- 

270. 
1950a. Growth layers on the teeth of Pinnipedia as an indication of age. 

Science, 112 :309-311. 
1950b. Experiments in the marking of seals and sea-lions. U.S. Fish 
and Wildlife Service, Special Scientific Reports— Wildlife, No. 4. 33 p. 






LITERATURE CITED 69 

1950c. The food of the Alaska fur seal. U.S. Fish and Wildlife Service, 

Wildlife Leaflet 329. 16 p. Processed. 
1950d. Growth of the testes and baculum in the fur seal, CallorMnm 

ursinus. Journal of Mammalogy, 31 : 384-394. 

1951. Cryptorchid fur seals. American Midland Naturalist, 46:646-648. 
1955. Body size with relation to population density in mammals. Journal 

of Mammalogy, 36 : 493-515. 
1958. Seals, sea lions, and walruses ; a review of the Pinnipedia. Stan- 
ford University Press, 179 p., 32 plates. 
Scheffee, V. B., and K. W. Kenyon 

1952. The fur seal herd comes of age. National Geographic Magazine, 
101 :491-512 (10 color plates in text). 

Scheffer, V. B., and F. Wilke 

1953. Relative growth in the northern fur seal. Growth, 17:129-145. 
Sohops, P. (In collaboration with Rudolf Fritzsche). 

1938. Pelze. J. J. Weber, Leipzig. 52 p. 16 color plates. 
Shanks, C. E. 

1948. The pelt-primeness method of aging muskrats. American Midland 
Naturalist, 39 :179-187. 
Stevenson, C. H. 

1904. The skins of fur-seals [p. 298-308, pi. 31]. In Utilization of the skins 
of aquatic animals. U.S. Fish Commission, Report of Commissioner for 
1902, p. 281-352, 13 plates. 
1905a. Leather from seal skins. Scientific American Supplement, 59: 

24334-24335. 
1905b. The skins of fur seals. Scientific American Supplement, 59 : 24502- 
24504. 
Stoves, J. L. 

1958. Fibre microscropy/its technique and application. D. Van Nostrand, 
Princeton, 286 p. 
Taylor, F. H. C, M. Fujinaga, and F. Wilke 

1955. Distribution and food habits of the fur seal of the North Pacific 
Ocean, U.S. Fish and Wildlife Service, Washington. 86 p. 
Terao, A. 

1940. Microscopical examination of the leather of aquatic animals. Jap- 
anese Society of Scientific Fisheries (Tokyo), Bulletin, vol. 8, No. 6, 
p. 343-346. 
Thompson, S. H. 

1950. Seal fisheries [p. 716-732]. In Marine products of commerce, by 
Donald K. Tressler and James McW. Lemon. Reinhold, New York, 782 p. 
U.S. Bureau of Fisheries 

1916. Dyeing and dressing Government fur-seal pelts. Fisheries Service 
Bulletin, No. 11 (April 1916), p. 4. 

1917. Alaska fisheries and fur industries in 1916. Bur. Fish. Doc. 838, 
118 p. 

1920. Fur-seal flippers as a source of glue. Fisheries Service Bulletin, 
No. 62 (July 1,1920), p. 2. 

1922. Alaska fishery and fur-seal industries in 1921. Document 933. 85 p. 

1923. Alaska fishery and fur-seal industries in 1922. Document 951. 118 p. 
1938. Preparation of Pribilof Islands fur-seal skins for market. 3 p. 

Processed. 



70 LITERATURE CITED 

U.S. Department of Agriculture 

1954. Hyperkeratosis (X-disease) of cattle. Department Leaflet 355. 6 p. 
U.S. Fish and Wildlife Service 

1952-57. Alaska fishery and fur-seal industries . . . Fish and Wildlife 
Service, Statistical Digests, Nos. 23, 26 29, 31, 33, 35, 37, and 40 [annual 
reports for calendar years 1948-55]. 
Wilber, C. G. 

1952. Fur seal blubber. Journal of Mammalogy, 33 : 483-485. 

WlLDMAN, A. B. 

1954. The microscopy of animal textile fibres, including methods for the 
complete analysis of fibre blends. Wool Industries Research Association, 
Torridon, Headingley, Leeds, 209 p. Illustrations in text, 235 halftones, 
11 color photographs, 88 line drawings. 

WlLKE, F. 

1959. Fat content of fur-seal milk. Murrelet 39 :40. 
Woollard, H. H. 

1930. The cutaneous glands of man. Journal of Anatomy, 64 : 415-421. 

Zobell, C. E. 

1946. Marine microbiology. Chronica Botanica, Waltham, 240 p. 



TABLES 



Table 1. — Length and weight of male fetal seals collected off the North American 
coast, by 10-day periods, 1951-52 

[16-25 January 1951, off Sitka, Alaska (Ford Wilke, MS); 15 February to 29 June 1952, California to Gulf 
of Alaska (Taylor and others, 1955, table 27); summer, full-term fetal and newborn seals collected selec- 
tively on Pribilof Islands, Alaska, 19 June to 11 August 1940-50 (Scheffer and Wilke, 1953, tables 1 and 2)] 



Midpoint of 10-day period 


Length in centimeters 


Weight in kilograms 




N 


Mean 


Range 


SD 


N 


Mean 


Range 


SD 


20 Jan 


41 


19.3 


14-24 


2.5 


41 


0.21 


0. 08-0. 39 


0.08 


30 


9 Feb 


















19 
29 


11 


28.8 


25-34 


2.6 


11 


.58 


.43- .83 


.16 


10 Mar.. 


15 
14 
28 
14 
19 

1 

45 

3 


35.6 
36.2 
42.4 
47.6 
50.1 

60.5 
64.4 
64.7 


27-42 
33-45 
37-48 
44-53 
46-57 


3.6 
2.9 
2.7 
2.7 
3.0 


15 
14 
28 
14 
19 

2 
61 

4 


1.01 
1.31 

1.59 
2.20 
2.42 

5.05 
4.89 
5.61 


. 53-1. 37 

. 80-1. 65 

1. 17-2. 04 

1. 79-2. 89 

1.96-3.35 

4. 99-5. 10 
4. 05-7. 03 
4. 54-6. 80 


.72 
.67 
.90 
.95 
.36 


20 
30 
9 Apr... 


19 
8 June 


18 

28 


60-72 
59-69 


2.4 
5.1 


.67 
.83 


Total... 


191 








209 






















Summer 


23 


65.9 


54-75 


4.6 


23 


5.4 


4. 08-7. 14 


.9 





Table 2. — Length and weight of female fetal seals collected off the North 
American coast, by 10-day periods, 1951-52 

[16-25 January 1951, off Sitka, Alaska (Ford Wilke, MS); 15 February to 29 June 1952, California to Gulf 
of Alaska (Taylor and others, 1955, table 27); summer, full-term fetal and newborn seals collected selec- 
tively on Pribilof Islands, Alaska, 19 June to 11 August 1940-50 (Scheffer and Wilke, 1953, tables 1 and 2)] 



Midpoint of 10-day period 


Length in centimeters 


Weight in kilograms 




N 


Mean 


Range 


SD 


N 


Mean 


Range 


SD 


20 Jan... 


37 


18.7 


14-25 


2.5 


38 


0.19 


0. 08-0. 37 


0.07 


30 


9 Feb 


















19 

29 


12 


27.3 


22-33 


3.3 


12 


.46 


. 29- . 79 


.52 


10 Mar 


15 
6 

22 
9 

20 
1 

1 

45 

7 


34.1 
38.7 
40.4 
43.1 
47.3 
48.7 

56.6 
61.9 
63.4 


30-39 
37-40 
31-45 
4<M5 
43-54 


2.7 
1.2 
3.5 

1.9 
2.7 


15 
6 

22 
9 

20 
1 

1 

58 

7 


.88 
1.21 
1.40 
1.66 
2.15 
2.30 

3.66 

4.87 
5.55 


. 63-1. 25 
1. 11-1. 28 

. 68-1. 90 
1.39-1.96 
1. 81-2. 64 


.54 
.06 
.89 
.62 
.89 


20 
30 
9 Apr.... 


19 
29 


8 June 










18 


57-67 
54-69 


2.6 
5.0 


3. 18-6. 12 
4. 99-6. 80 


.60 
.97 


28 




Total 


175 








189 






















Summer 


16 


63.1 


54-69 


4.4 


16 


4.8 


3. 31-6. 01 


.7 





71 



72 



TABLES 






Table 3. — Mean lengths of underfur and guard-hair fibers, by age and sex 

of seal 

[Measured on 114 tanned pelts at neck (10-25 em. behind ears), back (mid-dorsum at level of fore flippers) , 
and belly (ahead of navel) . Sexes are lumped through age 2 . Figures in parentheses represent N, or num - 
ber in sample.] 



Age and sex 



Black pup, newborn 2 
Black pup, molting. . 
Silver pup 



Yearling, pelagic. . 
Yearling, autumn.. 

Two-year-old 

Three-year-old male. 
Three-year-old 

female. 
Four-year-old male.. 
Four-year-old female . 
Five-year-old male. _ 
Five-year-old female. 

Six-year-old male 

Six-year-old female.. 
Seven-year-and- 

older males. 
Seven-year-and- 

older females. 

Maximum length of 
fibers on male 

Youngest age group 
in which maxi- 
mum length is 
attained 

Maximum length of 
fibers on female 

Youngest age group 
in which maxi- 
mum length is 
attained 



Dates taken 



11 June-22July. 
11 Aug.-25 Sept 
29 Sept.-17 Nov 



16-25 Apr 

24 Sept.-5 Nov. 
20 Aug.-21 Sept 

22 June-19 July 

23 Mar.-15 Sept 



Mean date 



16 June-22 July. 
2 Apr.-17 Aug... 
28 June-2 July 

14-24 July 

18-24 June 

17June-16July. 
13 June-5 July . 

13 Mar.-3 Oct- 



Length in millimeters ' 



Neck 



Under- 
fur 



6 July (4) — 

4 Sept. (4)- 

25 Oct. (7)- 

22 Apr. (4). 

26 Oct. (20). 
31 Aug. (7). 

5 July (10) - 
8 July (3).. 

4 July (12).. 
2 July (9) - 

1 July (6) . 
21 July (2) 
21 June (2). 

2 July (2) - 
25 June (9). 

10 July (13) .. 



8.5 

(2) 

9.7 

(3) 

11.8 

(5) 

12.7 

12.8 

12.0 

13.2 

12.7 

13.1 

12.6 
14.3 
12.3 
15.0 
12.0 
14.7 

12.0 



S yr. 
14 



Guard 
hairs 



10+ 



19.0 

(2) 

18.7 

(3) 

18.2 

(5) 

18.3 

19.4 

16.3 

20.2 

18.3 

20.1 

18.9 
21.0 
19.0 
26.5 
18.5 
46.8 

18.5 



8yr. 
22 



Ayr. 



Back 



Under- Guard 
fur hairs 



8.5 

(4) 

8.7 

(3) 

11.3 

(7) 

11.5 

11.3 

11.3 

11.4 

10.7 

11.9 
10.9 
11.8 
11.0 
11.5 
11.0 
12.7 

11.0 



7 yr. 
13 



Zyr. 



17.3 

(4) 

18.7 

(3) 

17.3 

(7) 

17.7 

17.5 

15.3 

16.9 

16.0 

17.9 
16.3 
17.8 
17.3 
18.5 
18.0 
23.0 

17.0 



33 



8yr. 
20 



lyr. 



Belly 



Under- Guard 
fur hairs 



6.0 
(2) 
5.5 
(2) 
7.0 
(5) 
7.7 
8.0 
6.3 
7.2 
7.7 

7.1 
7.2 
7.5 
7.3 
7.5 
7.0 
7.0 

7.0 



lyr. 
10 



Adult 



12.5 

(2) 
11.0 

(2) 
10.6 

(5) 
12.3 
12.8 

9.4 
11.5 
12.0 

11.3 
11.3 
11.8 
12.3 
12.0 
12.0 
13.1 

10.9 



8yr. 
15 



Adult 



• "Length" is equivalent to "depth" of underfur or guard hair; with fibers in natural, slightly bent, or 
wavy attitude. 

2 The black pup has no underfur, only a sparse underhair coat. Members of all age classes have an inter- 
mediate coat of short guard hairs, difficult to see without a lens, and ignored in the present table. 



Table 4. — Length of longest vibrissa, by age and sex 

[The longest mystacial vibrissa is normally the posterior bristle in row 4 or 5 (counting 6 horizontal rows 
from top of snout to lip). Its length is measured from surface of skin to tip. Since the tip may be worn 
or broken, maximum length is more important than minimum] 



Age» 


Number o: 


specimens 


Range 


Mean 


Male 


Female 


Male 


Female 


Male 


Female 




17 
1 
5 

20 

24 

28 

22 

6 

4 

1 

1 

4 


14 

8 
9 
10 
10 
4 
3 
2 
1 
1 
5 


51- 75 


52- 63 


63 
104 
118 
124 
133 
149 
163 
185 
199 
306 
286 
259 


57 








95-134 
111-144 

96-155 
110-222 
115-212 
148-204 
132-255 


77-120 
80-115 
95-142 
79-141 
106-127 
124-164 
118-168 


98 




94 




111 




113 




117 




150 




143 




120 








120 




217-334 


104-220 


145 







" Males, totaling 133, were taken June- August (except 1 yearling on 13 September); females, totaling 67, 
were taken June-September. 



TABLES 



73 



Table 5. — Change in color of mystacial vibrissae with age, female seals 
[Adapted from Abegglen and others (1957, p. 97; 1958, p. 186); based on 14.457 female seals] 



Age ' 


Black 


Black and white 


White 




Number 


Percent 


Number 


Percent 


Number 


Percent 




1,753 

1,067 

394 

57 

6 

2 


86 

43 

14 

3 

1 


282 

1,329 

1,901 

886 

234 

92 

26 

7 

27 


14 

54 

67 

44 

20 

11 

4 

2 

2 


6 

74 

531 

1,072 

919 

747 

657 

428 

1,960 






3 




19 




53 




79 




89 






96 








98 


10+ 






98 











i Estimated from tooth-ridge counts. 

Table 6. — Sizes of grading boards for raw, salted skins 

[Measured from outline tracings of boards provided by Fouke Fur Company in 1958; see figure 2] 



Size classification 


Length 


Width 


Area 




Inches 
30 

34 U 

35 >4 
39 
39% 


Inches 

21% 
24 H 
26 
27 
30 


Square inches 
718 




877 




946 




1,039 




1,261 







Table 7. — Sizes of male sealskins taken in early season 

[Size classification of raw, salted skins at St. Louis factory; skins originally stripped and blubbered on 
St. Paul Island from latter part of June through 15 July] 





1938 


1939 


1940 


1941 


4 years 


Size classification 










Total 
number 


Mean 




Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Small medium 

Medium 


6,507 

10, 971 

2,308 

163 

7 


32.6 

54.9 

11.6 

.9 


13, 818 

11,450 

1,636 

287 

21 


50.8 

42.1 

6.0 

1.0 

.1 


8,919 

14,058 

3,294 

155 

10 


33.7 

53.2 

12.5 

.6 


9,342 
15,368 
5,553 
1,258 
U24 


29.5 

48.7 

17.5 

3.9 

.4 


38,586 

51,847 

12,791 

1,863 

162 


9,646 

12, 962 

3,198 

466 

40 


36.7 

49.3 




12.1 


Extra large 


1.8 


Extra extra large. .. 


.1 


Total 


19, 956 


100.0 


27, 212 


100.0 


26, 436 


100.0 


31,645 


100.0 


105, 249 


26, 312 


100.0 







1 Including 4 wigs. 



74 



TABLES 



Table 8. — Sizes of male sealskins taken in late season 

[Size classification of raw, salted skins at St. Louis factory; skins originally stripped and blubbered on St - 
Paul Island from 16 July to end of July or early August] 





1938 


1939 


1940 


1941 


4 years 


Size classification 










Total 
number 


Mean 




Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Smnll medium 


5,897 

9,139 

2,088 

120 

3 


34.1 
53.1 
12.1 

.7 


8,457 

8,529 

1,170 

106 

2 


46.3 

46.7 

6.4 

.6 


5,774 

8, 558 

2,038 

167 

9 


34.9 

51.7 

12.3 

1.1 


13, 893 

13, 551 

3,750 

910 

i 116 


43.1 

42.1 

11.6 

2.8 

.4 


34, 021 

39, 777 

9,046 

1,303 

130 


8,505 

9,944 

2,261 

326 

32 


40.4 
47.2 


Large 


10.7 




1.6 


Extra extra large... 


.1 


Total 


17, 247 


100.0 


18, 264 


100.0 


16, 546 


100.0 


32. 220 


100.0 


84,277 121.069 


100.0 











1 Including 7 wigs. 

Table 9. — Comparison of body weights of female seals arriving on land in early 
slimmer and in late summer 

[From a kill of approximately 500 seals, none with full-term fetus, on hauling grounds and rookeries of St. 
Paul Island, 15 June-4 September 1953. (Ford Wilke, MS, 1953)] 



Age 



4 years 2 

5 years-. 

6 years.. 

7 years.. 



Earliest 25 

seals, mean 

weight in 

pounds 



62.9 
71.9 

77.7 
78.6 



Latest 25 

seals, mean 

weight in 

pounds 



55.2 
62.1 
70.8 
71.2 



Percent 
difference 



13.9 
15.8 
9.7 
10.4 



1 Estimated from tooth-ridge counts. 

2 Total sample 34, rather than 50, in this age class; thus, total number seals measured 184. 



Table 10. — Weight of fresh male sealskin with relation to field length of seal 

[Sample of 558 skins from subadult males, mostly ages 3 and 4 years, taken in regular commercial kill, St. 
Paul Island, 17 June to 27 July 1949. Length is "field length" or approximate length from snout to tip 
of tail on unskinned animal. Weight is of pelt, freshly blubbered and wrung, without mask and flippers. 



Length of seal 


N 


Weight of pelt in pounds 


V 




Range 


Mean 


SD 




38 inches 


8 
48 
49 
54 
46 
54 
44 
45 
49 
48 
46 
41 
15 
11 


3. 2-4. 2 
3. 1-4. 6 
3. 3-5. 8 

3. 7-5. 
3. 9-6. 1 
4. 0-6. 7 
4. 0-7. 5 

4. 6-7. 2 
5. 1-7. 3 
5.4-7.1 

5. 5-7. 7 

5. 2-8. 5 
6. 0-8. 

6. 2-8. 5 


3.7 
3.9 
4.3 
4.4 
4.7 
5.0 
5.5 
5.7 
6.0 
6.2 
6.4 
6.7 
7.0 
7.3 


0.33 
.33 
.48 
.31 
.50 
.62 
.75 
.59 
.50 
.40 
.51 
.62 
.63 
.69 


Percent 

9.1 




8.6 


40 inches 


11.3 




7.0 




10.7 


43 inches 


12.4 


44incheS-__ ._ 


13.6 




10.3 


46 inches 


8.3 


47 inches 


6.5 




8.0 


49 inches 


9.3 


50 inches . 


9.0 


51 inches. 


9.5 







TABLES 



75 



Table 11. — Trade classification of raw, salted, male sealskin with relation to 

field length of seal 

[Based on pelts of 523 subadult male seals sampled at random between 18 June and 20 July 1946, on St. Paul 
Island; classified by Harry Gladson in 1947] 



Field length 


Small 
medium 


Medium 


Large 


Extra large 


Extra extra 
large 


All classes 


and quarter 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


41 inches: 


2 
9 
10 
11 




2 
2 
10 

6 




1 




1 








6 
11 

22 
17 














Third 




2 














Fourth 
































Season -.. 


32 


57.1 


20 


35.7 


3 


5.4 


1 


1.8 






56 


100 








42 inches: 


1 

5 
6 
10 




4 
6 
11 
9 




1 












6 
11 
19 

19 


















Third 




2 
















































22 


40.0 


30 


54.5 


3 


5.5 










55 


100 














43 inches: 
First 


1 

1 




3 

8 
19 
5 








1 








5 

12 
22 
14 








3 
2 
1 












Third 




1 












8 




























10 


18.9 


35 


66.0 


6 


11.3 


2 


3.8 






53 


100 













44 inches: 
First 






2 
7 
13 

17 




1 




2 








5 

10 
16 
21 






3 

1 
2 












Third 




2 

2 












































6 


11.6 


39 


75.0 


5 


9.6 


2 


3.8 






52 


100 










45 inches: 
First 






3 
3 
10 

7 




2 
6 
5 
3 












5 
11 
20 
16 






1 

1 
6 






1 
4 










Third 










Fourth 


























8 


15.4 


23 


44.2 


16 


30.8 


5 


9.6 






52 


100 










46 inches: 
First 






3 

5 
2 
2 




11 
8 

11 
4 




5 
8 
2 
1 








19 
21 
16 

7 


















Third 


1 












Fourth 
























Season 


1 


1.6 


12 


19.0 


34 


54.0 


16 


25.4 






63 


100 


47 inches: 
First 






3 

8 
2 
2 




6 
7 
9 
4 




6 
6 
2 








15 

21 

15 

6 


















Third 

Fourth 


1 






1 
























Season 


1 


1.7 


15 


26.3 


26 


45.6 


14 


24.6 


1 


1.8 


57 


100 


48 inches: 
First - 






2 
1 
4 
3 




6 
7 
5 
3 




7 
15 
3 




2 
1 




17 

24 

12 

6 












Third 








Fourth 






































10 


16.9 


21 


35.6 


25 


42.4 


3 


5.1 


59 


100 










49 inches: 
First 






1 
1 

1 




5 
7 




8 
8 




3 




17 

16 

1 

5 












Third 














Fourth 








3 




2 




























Season 






3 


7.7 


15 


38.5 


18 


46.1 


3 


7.7 


39 


100 











1— 



76 



TABLES 



Table 11. — Trade classification of raw, salted, male sealskin with relation to 
field length of seal — Continued 



Field length 


Small 
medium 


Medium 


Large 


Extra large 


Extra extra 
large 


All classes 


and quarter 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


50 inches: 
First 






1 




6 




7 
5 
2 




3 

1 




17 
6 
2 




Second 








Third 
















Fourth 
















































Season 






1 


4.0 


6 


24.0 


14 


56.0 


4 


16.0 


25 


100 










51 inches: 
First 










2 




3 

3 
1 

1 




1 




6 
3 
2 
1 
















Third... 










1 












Fourth 








































Season 










3 


25.0 


8 


66.7 


1 


8.3 


12 


100 














All lengths: 

First 

Second 

Third 

Fourth 


4 
19 
20 
37 




24 
41 
72 
51 




41 
38 
39 
20 




40 

46 

15 

4 




9 
2 
1 




118 
146 
147 
112 














Season 


80 


15.3 


188 


35.9 


138 


26.4 


105 


20.1 


12 


2.3 


523 


100 



Table 12. — Trade classification of finished, dyed, male sealskin with relation to 

field length of seal 

[Based on pelts of 523 subadult male seals samnled at random between 18 June and 20 July 1946, on St. 
Paul Island; classified by Harry Oladson in 1947] 



Field length 


Small 
medium 


Medium 


Large 


Extra large 


Extra extra 
large 


All classes 


and quarter 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


41 inches: 
First 






3 

7 

16 
13 




2 
3 
2 




1 








6 
11 

22 
17 




Second 


1 

4 
4 












Third 














Fourth 
































Season 


9 


16.1 


39 


69.6 


7 


12.5 


1 


1.8 






56 


100 










42 inches: 
First 






3 
9 
11 

18 




3 
2 
6 
1 












6 
11 

19 
19 




Second 


















Third 


2 
















Fourth 
































Season 


2 


3.6 


41 


74.6 


12 


21.8 










55 


100 














43 inches: 
First 






2 
6 
17 
11 




3 
5 
5 
3 












5 
12 
22 
14 




Second 








1 










Third 














Fourth 




































Season 






36 


67.9 


16 


30.2 


1 


1.9 






53 


100 















44 inches: 
First 








2 
7 
9 
9 




3 








5 
10 
16 
21 




Second 






3 

7 
12 












Third.. 


















Fourth 




































Season 






22 


42.3 


27 


51.9 


3 


5.8 






52 


100 




== 


== 









TABLES 



77 



Table 12. 



-Trade classification of finished, dyed, male sealskin with relation to 
field length of seal — Continued 



Field length 


Small 
medium 


Medium 


Large 


Extra large 


Extra extra 
large 


All classes 


and quarter 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


Num- 
ber 


Per- 
cent 


15 inches: 
First. 










3 

6 
10 
5 




2 
3 
3 








5 
11 
20 
16 










2 
6 
11 












Third 








1 






Fourth 
































19 


36.5 


24 


46.2 


8 


15.4 


1 


1.9 


52 


100 










46 inches: 
First 






2 
1 

4 




11 
10 
9 

5 




6 
10 
3 
2 








19 
21 

16 

7 


















Third. 














Fourth 






































7 


11.1 


35 


55.6 


21 


33.3 






63 


100 














47 inches: 
First 










8 
13 

7 
5 




7 
7 
5 

1 








15 

21 

15 

6 
















1 

1 






Third 






2 






Fourth 
































2 


3.5 


33 


57.9 


20 


35.1 


2 


3.5 


57 


100 










48 inches: 
First 










6 
13 

1 

5 




10 

10 

9 

1 




1 
1 
1 




17 
24 
12 

6 
















Third 






1 






Fourth 
































1 


1.7 


25 


42.4 


30 


50.8 


3 


5.1 


59 


100 










49 inches: 
First 










6 
5 
1 




10 
9 




1 
1 




17 

16 

1 

5 




Second 






1 






Third . 








Fourth . 












3 




2 




























1 


2.6 


12 


30.7 


22 


56.4 


4 


10.3 


39 


100 










50 inches: 
First 










4 




10 
5 
2 




3 

1 




17 
6 
2 




Second 












Third 
















Fourth 
















































Season 










4 


16.0 


17 


68.0 


4 


16.0 


25 


100 














51 inches: 
First.. 










1 




4 
3 
1 
1 




1 




6 
3 
2 
1 
















Third.. 
















1 






Fourth 






































Season 










1 


8.3 


9 


75.0 


2 


16.7 


12 


100 














All lengths: 
First 






10 
29 

64 
65 




49 
64 
50 
33 




53 
48 
23 

8 




6 
4 
4 
2 




118 
146 
147 
112 




Second 

Third.. 

Fourth 


1 
6 
4 






Season 


11 


2.1 


168 


32.1 


196 


37.5 


132 


25.2 


16 


3.1 


523 


100 



663006 0—62- 



78 



TABLES 



Table 13. — Trade classification of raw, salted, male sealskin with relation to 

over-all dimensions 

[Based on pelts of 523 subadult male seals sampled at random between 18 June and 20 July 1948, on St. Paul 
Island; classified by Harry Qladson in 1947. Dimensions are approximate] 



Trade classification (size) 



Length in inches 



Minimum Maximum 



Width in inches 



Minimum Maximum 



Small medium.. 

Medium 

Large 

Extra large 

Extra extra large 



28 

303^ 

33 

35}^ 

38 



34^ 

37 

39^ 

42 

44M 



19 

22 

23^ 

25 



24 

25^ 

27 

28^ 

30 



Table 14. — Trade classification of finished, dyed, male sealskin with relation 

to over-all dimensions 

[Based on pelts of 523 subadult male seals sampled at random between 18 June and 20 July 1946, on St. Paul 
Island; classified by Harry Gladson in 1947. Dimensions are approximate] 



Trade classification (size) 



Length in inches 



Minimum Maximum 



Width in inches 



Minimum Maximum 



Small medium... 

Medium 

Large 

Extra large 

Extra extra large 



35J-3 
38J^ 
41^ 
44^ 
47^ 



39^ 
42}^ 
45>2 
48^ 
51J-S 



19 

20^ 
22 
23H 
25 



23 

24^ 
26 

27H 
29 



Table 15. — Length of ear from notch, by age and sex 



Age 



Number of specimens 



Male 



Female 



Range 



Male 



Female 



Mean 



Male 



Female 



years 

1 year 

2 years 

3 years 

4 years 

5 years 

6 years 

7 years 

8 years 

9 years 

10 years 

"adult" 

Total 



134 



Mm. 
33-41 



Mm. 
32-37 



Mm. 



36-45 


35-42 


40-46 


35-45 


40-49 


37-46 


42-49 


37-45 


42-51 


38-41 


45-50 


39-46 


49-57 


44-46 







48-53 



42-46 



67 



Mm. 



35 



TABLES 



79 



Table 16. — Yield of oil from fur seals killed on St. Paul Island, Alaska, 191$ 

to 1958 

[Source: U.S. Fish and Wildlife Service (1952-57) and unpublished data] 





From carcasses 


From skins 


Year 


Number 
of car- 
casses 

handled 


Yield of 

carcass 

oil 


Yield of 
carcass 
oil per 
carcass 


Number 

of sealskins 

handled 


Yield of 

blubber 

oil 


Yield of 

blubber 

oil per 

skin 


1958 


59, 570 
57, 315 
93,300 
49, 700 
49,850 
54,297 
50,935 
49, 565 
47, 800 
55, 232 


Gallons 
18, 035 
i 18, 000 
27, 813 
12, 771 
11, 352 
15, 180 
8,267 
12, 165 
10, 056 
14, 830 


Gallons 
0.33 
.31 
.30 
.26 
.23 
.28 
.16 
.25 
.21 
.27 


60, 080 
61, 244 
93, 966 
50, 803 
50, 239 
54, 995 
51, 560 
50,573 
48, 696 
57,445 


Gallons 

33, 400 

i 32, 000 

60,464 

29, 821 

30, 044 

31, 620 
28, 138 
28,000 
31, 957 
34, 420 


Gallons 
0.56 


1957 


.52 


1956.. 


.64 


1955 .- 


.59 


1954.. 


.60 


1953 


.57 


1952 


.55 


1951 


.55 


1950 


.65 


1949 


.60 







1 Estimated. 



APPENDIX A— COLOR NOTES 

Reproductions of the natural colors of seals or their pelts have 
rarely been published (Fortune, 1930; Schops and Fritzsche, 1938; 
Scheffer and Kenyon, 1952). Thus, it seems desirable to place on 
record the fresh colors of the fur-seal pelage. The simple black 
pattern of the pup has already been described. In the following 
pages, color notes will be given for males and females representing 
the following classes: silver pup, yearling in autumn, 3-year-old 
(adolescent) , and adult. 

Colors were compared on St. Paul Island, unless otherwise noted, 
from living or freshly killed seals, pelage clean and dry (or moist), 
in sunlight whenever possible, with a Munsell Color Company (1954) 
soil color chart as reference. Bowers (1956) and Miller (1958) 
have pointed out the advantages of the Munsell system of color 
notation. The names used in the present paper, however, are not 
those given on the soil color chart, but rather the more widely used 
ISCC-NBS names (National Bureau of Standards, 1955, p. 15-31). 

The pelage hues of the northern fur seal range from 5 YR to 10 
YR, with two exceptions : (1) On tanned pelts, a dark reddish brown 
color (2.5 YR 2/4 or 3/4) may appear. This is the color labeled 
by Maerz and Paul (1950, pi. 8) as "seal." It is next to "chocolate." 
"Seal" as a color name probably originated in the fur market from 
examination of tanned pelts. The color itself can often be seen in 
life near the bases of the flippers and around the teats and vestib- 
ular mucosa. As a matter of fact, the names "seal" and "seal brown" 
have been applied by artists and professional colorists to at least 
seven colors (National Bureau of Standards, 1955, p. 144-145). 
These colors range from dark reddish gray through moderate brown 
to dark olive brown. (2) The second exception is pale yellow (2.5 
Y 8/4), seen on adult vibrissae and usually referred to as "white." 
The middle value in an array of 39 colors recorded for fur-seal 
pelage is light brown (7.5 YR 5/4) — an interesting but perhaps 
unimportant fact. 

As a result of contact with rookery soil as well as excrement, 
urine, and regurgitated bile and milk, the coat of the seal invariably 
becomes stained. (Colors imparted by algal growths have been 
mentioned.) After patiently watching seals from a blind, Barthol- 
omew (1951, p. 3) concluded: 

When the pregnant females first arrive on the breeding grounds they are 
pale silvery gray. Within 4 days of coming ashore they turn yellow-brown. 
Each time they go to sea they regain some of their grayness, and females who 

81 



82 APPENDIX A — COLOR NOTES 

have made three trips to sea cannot, on the basis of color, be definitely dis- 
tinguished from females who have just come ashore for the first time. 

Similar color changes take place in the males, although their coloration is 
much more variable than that of the females. When they first come ashore, 
individual males will vary from light gray to almost black, but with each 
succeeding day ashore they become progressively more brown. One male, which 
when it was marked [tagged] was almost black, with a virtually white mane, 
after a week ashore became tan, with a yellow-brown mane. 

In the present study, I have tried to select clean specimens for 
observation, though for certain individuals, especially silver pups, it 
has been difficult to distinguish between stain and true pelage color. 

There is no strong evidence of graying (interference in the synthesis 
of melanin) with increasing age, except in the vibrissae, which are 
permanent hairs. As previously mentioned, these begin to turn white 
at the base, near the time of sexual maturity or a little later. 

Up to the time of writing (1959), no diagnostic features of pelage 
or internal anatomy that might be used to identify American, as 
against Asian, fur seals have been discovered. The three main popu- 
lations of Callorhinus ursinus breed, respectively, in eastern Bering 
Sea, western Bering Sea, and Sea of Okhotsk. (Perhaps 3,000 breed 
in the northern Kuriles, on the rim of the North Pacific.) Repro- 
ductive isolation in the three groups is rather complete, as indicated by 
the strong homing instinct of individual seals to the land of their 
birth. However, important numbers of seals from each group are 
known to mingle at sea in winter ; American seals have been recovered 
on Soviet grounds, and vice versa. After careful study, Taylor and 
others (1955, p. 61-65) could find no evidence of pelage differences 
among the members of the three main populations. 

Silver Pup, Male 

I can distinguish a male from a female silver pup on the basis of 
genitalia, less surely on the basis of size and shape of canine teeth, 
and even less surely on the basis of body size. (In a sample of 173 
pups weighed on 4 October 1947, the mean weight of males was 13.9 
kg., the mean weight of females 12.0 kg.) I cannot see any differ- 
ences in the color pattern of male and female silver pups, or even of 
autumn yearlings, in the second adult-type pelage. Nevertheless, in 
view of the possibility that slight differences do exist, I have 
reported separately on the color pattern for each sex. 

DORSAL ASPECT 

Top of snout light grayish brown (10 YR 7/3) ; upper lip also 
light grayish browm; forehead brownish gray (10 YR 4/1) ; cheeks 
light grayish brown (10 YR 7/3) ; region around eyes brownish gray 



SILVER PUP, FEMALE 83 

(10 YR 3/1), in strong contrast to paler cheek stripes; crown brown- 
ish gray (10 YR 4/1) ; ears brownish gray (10 YR 3/1) with slightly 
worn tips; neck, shoulders, back, rump, and tail brownish gray (10 
YR 4/1) ; no mane; rump patches (prominent) and flanks brownish 
pink (7.5 YR 7/2) ; bases of nippers, dorsal and ventral, dark grayish 
brown (5 YR 2/2). Light-brown color of belly extends upward and 
along sides into armpit, visible from dorsal aspect. 

VENTRAL ASPECT 

Lower lip and chin brownish pink (7.5 YR 7/2) ; lower lip stained 
brownish, probably from bile; throat brownish gray (10 YR 4/1). 
(Some silver pups have a continuous bright silver throat and anterior 
chest region, without the dark band of the throat as in the present 
specimen.) Chest, anterior region, brownish pink (7.5 YR 7/2) ; 
chest, posterior region between flippers, and belly, anterior region, 
brownish gray (10 YR 3/1) ; belly, posterior region, light brown (5 
YR 6/3) ; around penial opening grayish brown (5 YR 4/2) ; armpits 
very bright, brownish pink (5 YR 7/2) . 

Specimens : Principal specimen, age 8-10 weeks ; killed 24 September 
(24-9-58 B). Three others, killed 13 October-17 November; BDM 
187, BDM 188, BDM 184. 

Silver Pup, Female 

DORSAL ASPECT 

Top of snout light yellowish brown (10 YR 6/3) ; upper lip dark 
yellowish brown (10 YR 3/3), fading to color of cheeks, which are 
light grayish brown (10 YR 7/3) ; region around eyes brownish gray 
(10 YR 3/1) ; ears dark grayish yellowish brown (10 YR 3/2), slight- 
ly worn at tips; forehead, crown, back of neck, shoulders, back and 
rump brownish gray (10 YR 4/1) ; rump patches prominent, light 
brownish gray (10 YR 6/1) ; tail (dorsal) brownish gray (10 YR 
4/1); flanks mostly like belly, light grayish brown (10 YR 7/3); 
bases of flippers (dorsal and ventral) dark grayish yellowish brown 
(10 YR2/2). 

VENTRAL ASPECT 

Lower lips and chin at corners of mouth like top of snout, slight 
yellowish brown (10 YR 6/3), but at anterior tip stained darker, 
grayish brown (7.5 YR 3/2) ; throat brownish gray (10 YR 4/1) ; 
chest, anterior region, light grayish yellowish brown (10 YR 7/2) ; 
chest, posterior region between flippers, grayish yellowish brown (10 
YR 4/2) ; armpits brownish pink (7.5 YR 7/2), except for narrow 



84 APPENDIX A — COLOR NOTES 

zone near flippers, where moderate yellowish brown (10 YR 5/3) ; 
belly, anterior region, light brown (5 YR 6/3) ; belly posterior region, 
light grayish yellowish brown (10 YR 7/2) ; tail, ventral surface, 
brownish gray (10 YR 4/1) ; location of mammary teats not visible. 

Specimens : Principal specimen, age 8-10 weeks ; killed 28 September 
(28-9-58 A). Three others, killed 13 October-17 November; BDM 
185;BDM186;BDM189. 

Yearling, Autumn, Male 

DORSAL ASPECT 

Top of snout light brownish gray (10 YR 6/1) ; upper lip light 
grayish yellowish brown ( 10 YR 7/2) ; forehead light brownish gray 
(10 YR 5/1) ; cheeks light grayish yellowish brown (10 YR 7/2) ; 
region around eyes brownish gray (10 YR 4/1) ; crown light brownish 
gray (10 YR 5/1) ; ears grayish yellowish brown (10 YR 4/2) ; back 
of neck (no mane), shoulders, back, and rump light brownish gray 
(10 YR 5/1) ; rump patches not conspicuous (though conspicuous on 
certain other yearling males), light yellowish brown (10 YR 6/2) ; 
tail, dorsal, dark grayish yellowish brown (10 YR 2/1) ; flanks 
shading into color of belly, visible well up along sides; bases of 
flippers, dorsal and ventral, grayish brown (5 YR 3/2). 

VENTRAL ASPECT 

Lower lip and chin light yellowish brown (10 YR 6/3) ; throat 
brownish gray (10 YR 4/1) ; chest, anterior region, light grayish 
yellowish brown (10 YR 7/2) ; chest, posterior region between 
flippers, grayish brown (7.5 YR 4/2) ; arm pits moderate brown (5 
YR 4/3) ; belly, anterior region, grayish brown (7.5 YR 4/2) ; belly, 
posterior region, light grayish yellowish brown (10 YR 7/2), stained 
brownish posterior to penial opening; tail, ventral, grayish brown 
(5 YR 3/2), with black margins. 

Specimens: Principal specimen killed 26 September (26-9-58 A). 
Sixteen others killed 13 September-5 November (27-9-58 B, BDM 
nos. 7, 8, 14, 15, 21-24, 290, and 512-517) . 

Yearling, Autumn, Female 

DORSAL ASPECT 

Top of snout and upper lip light brown (7.5 YR 5/4) ; cheeks light 
yellowish brown (7.5 YR 7/4) ; narrow region around eyes grayish 
brown (7.5 YR 3/2) ; outward of this brownish gray (10 YR 3/1) ; 
ears moderate brown (7.5 YR 4/4), rubbed bare at tip; forehead, a 
circular region of grayish brown (7.5 YR 3/2) ; crown, back of neck, 



THREE-YEAR-OLD, ADOLESCENT MALE (BACHELOR) 85 

shoulders, back and rump brownish gray (10 YE. 3/1) ; rump with 
conspicuous patches of light brownish gray (10 YR 6/1) ; tail, dorsal, 
dark grayish yellowish brown (10 YR 2/1) ; flanks intermediate color 
between back and belly, appearing light-colored from above; bases 
of flippers, dorsal and ventral, dark grayish yellowish brown (10 
YE 3/2). 

VENTRAL ASPECT 

Lower lip light brown (7.5 YR 5/4) ; chin and throat brownish 
gray (10 YR 4/1) ; chest light grayish yellowish brown (10 YR 7/2) ; 
belly brownish pink (7.5 YR 7/2) ; tail, ventral, dark grayish yellow- 
ish brown (10YR2/1). 

Specimens: Principal specimen killed 3 October (3-10-58 A). 
Four others killed 27 October-23 November (BDM nos. 16, 25, 26, 
and 29) . 

Three-year-old, Adolescent Male (Bachelor) 

DORSAL ASPECT 

Top of snout and upper lip light yellowish brown (10 YR 6/3) ; 
upper lip palest at corner of mouth, under eye, and darker toward 
muzzle ; cheeks also light yellowish brown ; region around eyes grayish 
yellowish brown (10 YR 5/2) ; ears grayish yellowish brown (10 
YR 4/2), not worn at tips, followed posteriorly by faded streak; 
forehead to rump brownish gray (10 YR 4/1) ; crown with an area 
about 6 cm. in diameter in which the guard hairs are longer (28 mm.) 
than those surrounding (15 mm.) and are erect or slightly recurved; 
rump patches faintly suggested; tail, dorsal, dark grayish yellowish 
brown (10 YR 2/1) ; flanks, transition color between back and pos- 
terior region of belly; bases of flippers, dorsal and ventral, dark 
grayish yellowish brown ( 10 YR 2/2) . 

VENTRAL ASPECT 

Lower lip and chin moderate yellowish brown (10 YR 5/3); 
throat brownish gray (10 YR 4/1) ; chest, anterior region, light 
yellowish brown (10 YR 6/3) ; chest, posterior region between flip- 
pers, grayish brown (7.5 YR 3/2) ; armpits moderate brown (5 YR 
3/4); belly, anterior region, grayish brown (5 YR 3/2), shading 
gradually into color of belly, posterior region, light yellowish brown 
(10 YR 6/3) ; tail, ventral, brownish gray (5 YR 3/1). 

Specimens: Principal specimen killed 27 September (27-9-58A). 
Ten others killed 22 June-19 July (BDM nos. 60, 67, 70, 72, 73, 77-79, 
83, and 87). 



86 APPENDIX A COLOR NOTES 

Three-year-old, Adolescent Female (Young Cow) 

DORSAL ASPECT 

Top of snout grayish brown (7.5 YR 4/2) ; upper lip light yellow- 
ish brown (10 YR 6/3) ; forehead brownish gray (10 YR 4/1) ; 
cheeks moderate yellowish brown (10 YR 5/3) ; region around eyes 
brownish gray (10 YR 3/1); crown brownish gray (10 YR 4/1); 
ears grayish yellowish brown (10 YR 4/2), worn and blackish at 
tips; back of neck, shoulders, back and rump brownish gray (10 YR 
4/1 ) . Flanks are colors in transition from back to belly. Tail, dorsal, 
and bases of all flippers, all surfaces, dark grayish yellowish brown 
(10 YR2/2). 

VENTRAL ASPECT 

Lower lip and chin moderate yellowish brown (10 YR 5/3) ; 
throat brownish gray (10 YR 4/1) ; chest, anterior region, light 
yellowish brown (10 YR 6/3) ; chest, posterior region between 
flippers and belly, anterior region, grayish brown (7.5 YR 3/2) ; 
belly, posterior region, moderate brown (7.5 YR 4/4) ; location of 
mammary teats not visible ; vestibular mucosa dark brownish gray to 
black, with narrow rim of dark reddish brown hair (2.5 YR 2/4) ; 
tail, ventral, dark grayish yellowish brown (10 YR 2/2). 

Specimens: Principal specimen killed 24 September (24-9-58 A). 
Three others killed 23 March-15 September (NWC 52-3048, BDM 
287, and BDM 410). 



Adult Male (Bull) 



DORSAL ASPECT 



Top of snout and upper lip grayish yellowish brown (10 YR 5/2) ; 
forehead, cheeks, and region around eyes grayish yellowish brown 
(10 YR 4/2) ; crown and ears dark grayish yellowish brown (10 YR 
3/2) ; crown with longer hairs ("wig") same color as surroundings; 
ear tips worn bare, nearly black; back of neck (mane) light yellowish 
brown (10 YR 7/4), faintly parted (divergence line) into right and 
left sides by the paired neck muscles, longest hairs 70 mm. ; shoulders 
and back gradually changing from grayish yellowish brown (10 YR 
4/2) to dark grayish yellowish brown (10 YR 3/2) ; rump and upper 
surface of tail dark grayish brown (5 YR 2/2) ; flanks grayish yel- 
lowish brown (10 YR 4/2); bases of all flippers, upper and lower 
surfaces, dark reddish brown (2.5 YR 2/4) . 



ADULT FEMALE (OLD COW) 87 

VENTRAL ASPECT 

Lower lip and chin grayish yellowish brown (10 YR 4/2) ; throat, 
chest, and belly grayish brown (7.5 YR 3/2) ; penial opening not 
marked by color change ; tail, ventral surface, nearly naked, grayish. 

Specimens: Principal specimen at least 15 years old, teeth worn 
or missing, killed 19 September (19-9-58 A) . Nine other bulls killed 
25 June-5 July, all 7-year-olds or older (age 7, BDM nos. 251-255; 
age 8, BDM nos. 302-303; age 9, BDM 319; "adult", BDM 75). 
Many other known-age males between ages 3 and 7 have been 
examined. 

Adult Female (Old Cow) 

DORSAL ASPECT 

Top of snout light brown (7.5 YR 5/4) ; upper lip moderate brown 
(7.5 YR 4/4) ; forehead grayish brown (7.5 YR 3/2) ; cheeks moder- 
ate brown (7.5 YR 4/4) ; region around eyes grayish brown (7.5 YR 
3/2) ; crown grayish brown (7.5 YR 4/2) ; ears light brown (7.5 YR 
5/4) along two-thirds of length, bare and blackish on tip; back of 
neck grayish brown (7.5 YR 4/2) with a suggestion of a lighter, 
grayer color on crown and mane; shoulders, back and rump grayish 
brown (7.5 YR 3/2) ; tail, dorsal, dark grayish yellowish brown (10 
YR 2/2) ; flanks light brown (7.5 YR 5/4), lightening toward armpit 
and darkening toward hind flipper ; bases of flippers, dorsal and ven- 
tral, dark grayish brown (5 YR 2/2) . 

VENTRAL ASPECT 

Lower lip, chin, and throat grayish brown (7.5 YR 3/2) ; chest, 
anterior region, light grayish brown (7.5 YR 5/2) ; chest, posterior 
region, grayish brown (7.5 YR 3/2) ; armpits moderate brown (5 
YR 4/4) ; belly, grayish brown (5 YR 3/2) ; mammary teats not 
visible, their location marked by a few gray hairs; vestibular mu- 
cosa wrinkled, dark brownish black, surrounded by a thin line of 
dark reddish brown (2.5 YR 2/4) ; tail, ventral, also dark reddish 
brown. 

Specimens: Principal specimen killed 25 September (25-9-58 B). 
Eight others killed 13 March-11 September; all 7-year-olds or older 
(age 7, BDM nos. 276, 279; age 8, BDM nos. 324 and 349; age 10, 
BDM 404; age over 10, SITKA 50-25, SITKA 50-34, NWC 52- 
3029). Many other females between ages 3 and 7 have been exam- 
ined ; also many females recorded simply as "adult." 



APPENDIX B— GLOSSARY 

For textbook-style illustrations of the structure of hair and skin, 
see Auber (1952), Chase (1954), Hamilton (1951), Hausman (1939, 
1944), Miller (1952), Montagna (1956), Odland (1954), Parnell 
(1951), and Wildman (1954). 

Awn. — (See guard hair.) 

Bachelor.— Colloquially, a male seal of any age between 2 and 6 years, in- 
clusive. Thus, a pup born in summer 1950 became a "yearling" on 1 January 
1951, a "bachelor" on 1 January 1952, and a "bull" on 1 January 1957. 

Birthcoat. — Pelage of the black pup, newborn. 

Blastocyst (blastula).— An early stage of the embryo when the cells are ar- 
ranged in a single layer to form a hollow sphere, barely visible to the naked 

eye. 

Blubber (panniculus adiposus) .— The thick stratum of yellowish or whitish, 
fatty connective tissue which underlies the skin of most marine mammals. 

Bulb.— Swollen base where hair root and hair follicle are indistinguishable. 

Bull.— Colloquially, a male seal older than 6 years. (See bachelor.) 

Cast.— A negative impression, in a sheet of plastic film or gelatin, of part of 
the surface of a fiber. 

Club hair.— An inactive, mature hair with characteristic shrunken, rather 
than bulbous, base. 

Connective tissue sheath.— An especially heavy sheath surrounding the outer 
root sheath of the vibrissa. The connective tissue sheath serves for attach- 
ment of erectile muscles. 

Cortex.— Main substance of the hair, situated between the outer cuticle and 
the central medulla; usually pigmented; consisting of dead, keratinized cells. 

Cow.— Colloquially, a female seal older than a yearling. Thus, a pup born in 
summer 1950 became a "yearling" on 1 January 1951 and a "2-year-old cow" 
on 1 January 1952. 

Cuticle.— A single layer of translucent cells on the surface of the hair. The 
cells are attached at one end, with their thin free margins pointing toward 
the tip of the hair. In the fur seal, the cuticle is less than 1 micron thick. 

Cuticular-scale pattern.— Various nomenclatures have been proposed, depend- 
ing upon shape of the visible portion of the scale, degree of overlap, and form 
of external margins. See Wildman (1954) for diagrams and photographs of 
coronal, diamond-petal, and other patterns. 

Dermis (derma, corium).— The thickest and most elaborate stratum of the 
skin (representing, in the fur seal, about 99 percent of its thickness), lying 
between the epidermis and the blubber, consisting largely of connective tissue 
surrounding the hair roots and the sweat and sebaceous glands. 

Epidermis. — The thin surface layer of skin, consisting of two main parts: a 
superficial stratum corneum of dead, translucent cells and a deeper stratum 
malpighii of active, deeply staining cells. At the site of each hair follicle, the 
two layers dip deeply into the dermis. 

89 



90 APPENDIX B — GLOSSARY 

Fiber. — As used in this work, a hair of any kind, including a vibrissa. 

Follicle. — A cylindrical sleeve or pouch, representing an invagination of the 
outer skin, in which the hair grows. It is swollen at the base into a bulb. 

Follicular bundle (common follicular bundle, common hair bundle). — A group 
of follicles, each distinct at its base (bulb) but coalescing near the surface of 
the skin. Near the surface of the adult skin, the follicular bundle contains a 
guard hair, 35-40 underfur fibers, sweat duct, and 2 sebaceous ducts, all sur- 
rounded by a thick, conspicuous sheath (the outer root sheath). 

Germ (germ plate, hair germ).— Matrix cells which remain below the hair 
follicle, at the tip of the papilla, in the resting stage between molts. They 
initiate growth of a new hair at regular intervals. 

Guard hair (shield hair, awn hair, overhair). — A fiber of the outer coat of 
the juvenile and adult; largest, thickest, and most deeply rooted of the pelage 
fibers ; distinctly pigmented and medullated along most of the shaft ; expanded 
toward the tip into a blade. (Certain of the largest ones, appearing widely 
spaced in longitudinal rows on the fetus, might be called guide hairs, or 
"Leithaare".) 

Hair. — The hair is divided regionally into root and shaft ; structurally into 
cuticle, cortex, and medulla (which see). So far as known, all hairs of the 
adult fur seal, except the vibrissae, are replaced annually by new hairs origi- 
nating on the sites of their predecessors. 

Horizontal section (tangential section). — A section cut in a plane parallel to 
the surface of the skin. Since the pelage fibers emerge at a slant, they rarely 
appear as true circles in a horizontal section of skin. 

Lanugo. — The fine hair on the body of the fetus. 

Matrix. — Base of the root bulb where the cells are most actively dividing to 
form the hair. 

Median section. — A section made in the plane which divides the body sym- 
metrically into right and left halves. In a true median section of hair from 
the middle of the back, for example, the entire length of the hair can be seen. 

Medulla. — Pith of the hair ; a series of gas-filled cells along the axis of the 
shaft ; present only in guard hairs and largest underhairs. See Wildman 
(1954) for classification of medullary patterns. 

Melanocyte. — A cell in the root of the hair follicle, or elsewhere, which manu- 
factures a yellowish brown pigment ; melanin. Concentrated melanin appears 
black. 

Micron. — A unit of measurement : 0.001 mm. or 0.000039 inch. 

Molt. — Replacement of an older crop by a newer crop of fur and hair fibers. 
In the fur seal, the first molt (in late summer) results in a qualitatively 
different population of fibers. Each subsequent annual molt, prolonged over 
a period of 4-5 months in autumn, results in a similar, though newer, population. 

Munsell color notation. — (See Appendix A and Literature Cited.) 

Otariid (eared seal). — Anglicized name for any fur seal or sea lion of the 
family Otariidae, including 7 genera (Scheffer, 1958). 

Palmar. — Pertaining to the palm. 

Panniculus adiposus. — (See blubber.) 

Panniculus carnosus. — A discontinuous thin sheet of muscle underneath the 
blubber. When a sealskin is stripped forcibly from the body, large patches of 
the panniculus carnosus remain with the skin. 

Papilla. — A cone of connective tissue, continuous with the dermis, which rises 
into the bulb of the follicle. 

Parchment cure. — Method- of preserving a sealskin by stretching it on a hoop 
and allowing it to dry. 






APPENDIX B — GLOSSARY 91 

Parous. — Having borne one or more young. 

Phocid (hair seal, earless seal, true seal). — Anglicized name for any seal 
of the family Phocidae, including 13 genera. 

Pilary system. — The pelage, including the hidden roots and nourishing struc- 
tures of the fibers as well as their visible shafts. 

Pilosebaceous. — Pertaining to the hair and its sebaceous gland or glands. In 
the fur seal, the guard hair with follicle and pair of sebaceous glands compose 
a pilosebaceous unit. The hair emerges through a pilosebaceous funnel and 
orifice. A bundle of underfur fibers emerges through the same funnel and 
orifice. 

Plantar. — Pertaining to the sole. 

Pore. — On hairless skin, the opening of a sweat gland; on unhaired skin or 
leather, the pilosebaceous orifice. 

Prime. — A pelt is prime when molt is complete, that is, when new fibers have 
ceased to grow in length; pigment is no longer being formed in the root; and 
old fibers have been pushed out or shed. 

Primordium. — An embryonic hair follicle first visible as a thickening of the 
epidermis. 

Pup. — A young seal up to about age 6 months, or arbitrarily to 31 December of 
the year of birth. In the present report, pups are classified as "black pup, new- 
born" ; "black pup, molting" ; and "silver pup." 

Rhinarium. — Naked area of roughened skin at the tip of the snout. 

Roadskin. — Colloquial term applied on the Pribilof Islands to a seal in shock 
from overheating or exhaustion ; also to its skin. 

Root. — The basal (proximal) portion of the hair which is buried in the skin 
and is surrounded by 1-3 sheaths. Near the deepest part of the root, the hair 
matrix (living) becomes the hair (dead). 

Root sheath.- — The outer root sheath is a pouch-like continuation of the epi- 
dermis which surrounds the follicle, except at its extreme base, where the 
papilla enters. The inner root sheath originates in the bulb of the follicle and 
extends part way up the root of the hair, interlocking with the cuticle of the 
hair. ( See also connective tissue sheath. ) 

Scale. — (See cuticle.) 

Sebaceous gland.— A gland secreting an oily hairdressing known as sebum. 
Twin glands are associated with each follicular bundle in the juvenile as well 
as in the adult pelage. Each gland secretes independently through an exit 
near the base of the guard-hair shaft. 

Shaft. — The free portion of the hair. At its basal (proximal end, the shaft 
may free itself of the root sheath a short distance below the level of the 
epidermis. At its terminal (distal) end, it tapers to a sharp tip. 

Stage. — Colloquially, a pelt is said to be "stagy", or going through a stage, 
when it is molting ; especially when dark pigment flecks can be seen on the 
buffed side of the pelt. 

Stratum. — (See epidermis.) 

Sweat gland. — A sweat gland of the apocrine type originates beneath and 
beside each guard hair, and empties into the pilary funnel above the exit 
of the sebaceous glands. See Woolard (1930) for distinction between the 
small, ordinary, superficial eccrine glands of man and the large, deep apocrine 
glands of man and lower animals. 

Tela subcutanea. — A filmy layer of connective tissue, often difficult to dis- 
tinguish, which binds the panniculus adiposus or panniculus carnosus to the 
body, 



92 APPENDIX B GLOSSARY 

Tooth ridge. — Dentin is deposited unequally in winter and summer on the 
root of the fur-seal tooth. The result is a series of concentric, alternate, 
ridges and valleys which may be counted on the surface of the root up to 
about age 10 to 12 years. 

Underfur fiber (underfur hair). — One of the fibers of the undercoat of the 
adult-type pelage ; first seen in mature form when the pup has reached an age of 
about 6 months ; always in bundles of 35 to 40 ; nonpigmented ; nonmedullated, 
wavy ; varying but little in cross-section shape and diameter along its length. 

Underhair. — A fiber of the undercoat of the pup; shortest, thinnest, and most 
superficially rooted of the juvenile pelage fibers ; faintly pigmented ; usually 
nonmedullated ; varying but little in cross-section shape and diameter along 
its length. 

Vellus. — A coating of fine, temporary hairs such as those on top of the flippers 
of the fur-seal fetus. 

Vibrissa (sinus hair, tactile hair, whisker). — An elaborate sensory bristle, 
much larger than any body hair, situated on either side of the snout (mys- 
tacial) and above each eye (superciliary). 

Yearling. — ( See bachelor. ) 



PLATES 




Plate 1.— Wet pelage of subadult male in San Diego Zoo. Water parts the tips of 
the guard hairs into silvery streaks but does not penetrate to the base of the 
underfur. ( San Diego Zoo photo by R. Van Nostrand ) 



95 



553006 0-62-8 




^mm&im 



yww* 










Plate 2. — Cross section of subadult male showing pelage relations; posterior face 
of section at level of bronchi; flesh frozen; pelage damp; X V-i (above) and X fl 
(below). GH — guard hairs; UF — underfur hairs; E — epidermis; D — dermis of 
two indistinct zones, the darker one containing the hair roots; PA — panniculus 
adiposus; PC — panniculus carnosus; TC — tela subcutanea. (3037-9) 



96 




Plate 3.— Tanned pelt of subadult male showing pepper-and-salt effect of white- 
tipped guard hairs. (Above) Outer surface; anterior end at left; X 4. (Below) 
Inner surface, with guard hairs extending beyond wooly underfur fibers ; leather 
at bottom of photo; X 4. (1285 and 1292 A) 



97 









'0f 








mm 







Plate 4-A.— Strip of parchment-cured pelt from back of subadult male; 12 July; 
natural size. (3978) 




Plate 4-B. — Freshly cut section of pelt of yearling with drop of water placed on 
underfur to demonstrate its water-repellent nature; 7 September; x 4. (4063) 




Plate 5. — Scaly appearance of epidermis from back of subadult male ; 12 July ; skin 
sample parcbment-cured ; later macerated 20 days in warm water; plucked; 
hardened in formalin; illuminated from anterior end (top) ; photographed in moist 
condition: x 40. (4212) 



99 




* f * 4 









* »' > •/ 



Plate 6. — Surface of suede-tanned leather from back of neck of 5-year-old female, 
showing distribution of pilosebaceous orifices, or "pores." A bundle including 1 
guard hair and 35 to 40 underfur hairs was withdrawn in the tanning process 
from each pore. Anterior end of body toward top of photo; X 40. (4075) 



100 




\\\ 






Plate 7-A. — Small square of parchment-cured pelt from back of subadult male. 
Lateral view; X 4. (3913) 



■*. \.W7«k':-. 










Plate 7-B.— Transverse section of pelt from rump of adult female showing natural 
waves in underfur ; fibers held together with tritolyl phosphate ; anterodorsal view ; 
X 3. (3985) 

101 




Plate 8. — Silhouette of bit cut from tanned skin of neck of 9-year-old male, showing 
underfur fibers, ordinary guard hairs, and long guard hairs (mane hairs) ; 1 July* 
X 2. (41. -.2) 



102 



M / ■ 



M 







•^& 






Plate 0.— Median section through skin of back of 7-year-old male ; 30 September ; 
pelage not fully prime ; pigment cells still active in bulbs of underfur follicles and 
guard-hair follicles; X 30. B— bundle of underfur hairs; E— epidermis ; GH— 
guard-hair bulb ; PF— pilosebaceous funnel ; SBG— sebaceous gland: SWG— sweat 
gland ; UF— underfur-hair bulb. * 417 °) 



103 







•'-J* /'(. 



Plate 10. — Median section through skin of back of 7-year-old female; 30 September; 
X 30. (4171) 



104 




r ♦ ; • 


•1 1 


• # 




% .» 




• • * • 


« 


t # # 


■% 


«§». w 


* 


# • 


* u 




» 1! 




* 




• 


• /.■•*• 


,• **■$ 




~~* \ 






* - 

• ; » 






• * 


• * •# -.* 


• A 



Plate 11.— (Left) Duct of sweat gland rising from secretory portion ; median section 
of skin from back of 5-year-old female; 23 September; X 120. (Right) Similar; 
secretory portion of sweat gland of 3-year-old female ; 5 September X 400. 

(4176 and 4180) 



105 



w& 



Plate 12. — Sweat droplets appearing on palmar surface of fore flipper freshly 
severed from body of 3-year-old female seal; exposed to heat lamp; X 4. (4055) 



106 



\JT' 




is- ! 'Pi 


















# 










h 



r 



»' > ."lex '>' » 



/ 



■*S 





+S2**ar-7*^ i**: 






Plate 13. — Horizontal section, at depth of about 0.1 mm., from back of neck of 
3-year-old female; pelage fully prime; 7 February; posterior end at top; X 200. 
Note 8 entire follicular bundles, each with a single, large, round, translucent 
section of guard-hair root and 35 to 40 underfur roots. (4209) 

107 




Plate 14. — A single follicular bundle; X 800. Note (above) bundle of 40 underfui 
fibers, (left) sweat duct, and (below) guard hair flanked by 2 sebaceous ducts. 

(4213) 



108 



$& 



w*\ <. 









l ». 




Plate 15. — (See pi. 13.) Section at depth of about 0.4 mm., where sebaceous glands 
are largest in cross section ; X 200. Note sweat duct between guard hair and fur 
bundle. (4214) 

109 



5V# 



i.h 

^18 







41 






Plate 16. — (See pi. 13.) Section at depth of about 1.0 mm., near base of underfur 
roots ; X 640. Note follicles of underfur fibers at various levels ; some being 
more superficial, appearing as dark bulb-sections. At lower right, the sweat duct 
rises through the base of a large sebaceous gland. (4215) 



110 










^5! 











Plate 17. — (See pi. 13.) Section at depth of about 1.2 mm., showing lumens of 
sweat glands; X 200. Note (center) the dome of a gland, with its duct, beside 
a guard-hair follicle. (On another section cut nearer the surface, a bundle of 
underfur fibers is situated vertically above this dome.) (4216) 



553006 0-62-9 



111 



If 

i' 




I 


' ' iHV fi 


1 v 


w 







> 




Plate 18.— Bundles of fibers rising from surface of skin of subadnlt male : antero- 
dorsal view. (Above) Parchment-cured skin. (Below) Plucked, chamois-tanned, 
black-dyed skin. (3986 and 3939) 

112 




*late 19. — Bundles of fibers rising from surface of skin of subadult male ; antero- 
dorsal view; X 40. (Left) Hyrax mount; each prominent, white-dotted column 
is the medulla of a guard hair. A tuft of underfur fibers, nonmedullated, rises 
from the pilosebaceous orifice behind the guard hair. (Right) Polyvinyl-acetate 
mount, showing portion of root as well as shaft ; medulla here appearing black 
rather than white. (3929 and 4114) 



113 






£ 












-» * 



Plate 20. — Root of guard-hair follicle in median section of skin from back of 
4-year-old male; 2 September; X 80 (above) and X 400 (below). B — bulb; 
IRS— inner root sheat; ORS— outer root sbeat; P— papilla. (4182 and 4183) 



114 










Plate 21. — Vibrissa of 144 g. fetus ; 25 January ; cross section of upper portion of 
follicle; X 200. Primordia of nonspecialized body bairs appear at top of pboto. 
CTS — connective tissue sheath (much thicker around vibrissa than around guard 
hair) ; IRS — inner root sheat; ORS — outer root sheath; V — vibrissa, mainly pig- 
mented cortex with thin, dark cuticle and suggestion of central medulla. (4163) 



115 




Plate 22.— Stumps of underfur fibers and guard hairs from parchment-cured seal- 
skin ; in tri-n-butyl phosphate ; fibers cut at approximately halfway point ; antero- 
dorsal view showing flat side of each guard hair ; mountant has partly invaded 
medulla and has pushed original gas upward through cut tips of guard hair; 
X 50. (4118) 



116 




Plate 23.— Cross sections (by Hardy device) of adult-type pelage; X 300. (Left) 
Prime silver pup ; 13 October ; tanned skin. Smallest bodies are underf ur fibers ; 
largest are guard bairs of various sizes. (Right) Subadult male; 12 July; 
parcbment-cured skin. Note that guard bairs are larger and darker b*»re. 

(4153 and 4112) 



117 




Plate 24. — Cross section of coarse mane hairs ; some white and some dark brown ; 
cut about 1 cm. from surface of skin; 10-year-old male; (left) X 120; (right) 
X 800. (4099 and 4098 A) 



118 



, f)X 



Mi 




Plate 25. — Cross section of underfill- fibers of finished, brown-dyed, subadult male 
sealskin ; X 1000. (4110 C) 

119 



•V 






W f 













.* . 

Plate 26-A. — Pigment granules in blade of large guard hair from mane of 10-year-oU 
male; field includes about one-third of cross section; medulla not open at this 
level; X 1000. (4160) 




Plate 26-B. — Underfur fibers from back of subadult male; parchment-cured skin 
thermoplastic cast ; X 100. Note smooth attenuated root of one fiber. (4089 A) 



120 






'late 27. — Guard hairs against background of tine underfur fibers ; first adult-type 
foelage; silver pup; 13 October. (Only the medulla or pith of the hair stands out 
in this mountant.) (Above) Hairs of various sizes; X 100. (Below) Portions 
of two medium-size hairs: X 500. (4102 A and 4102 B) 



121 




Plate 28.— Cuticular-scale pattern on basal region of shaft of guard hair ; parchment- 
cured skin; thermoplastic casts; X 100 and X 500. (4092 O and 4094) 



122 




yifrsff 



»j^i^ i- 11 





'■ kg 



Plate 29. — Cuticular-scale pattern on blade of shaft of guard hair ; parchment-cured 
skin; thermoplastic cast ; X 500. (4091 B) 



123 



[I 



y 






[I 

Plate 30. — Cuticular-scale patterns on underfur fibers from pelt of subadult male ; 
parchment-cured; gelatin easts; X 500. (Left to right) basal region, middle 
region (3 fibers), and tip. (4088 A, B, C, D, F) 



124 



L^m%£i 



'*!> 




Plate 31.— Medulla of guard hair of albino in first adult-type pelage ; 2 December ; 
basal region of shaft in benzol ; X 500. Gas-filled chambers of the medulla reflect 
light and cause the pelage to appear white. (4185) 



125 



^ «► 



* 






* 



*r + 



-' 






Plate 32-A. — Priniordiuni of hair follicle developing as thickening of epidermis in 
skin of back of 23.7 g. fetus; X 500. (4158) 











%* ^ "'^' 



Plate 32-B. — Three early follicles pushing downward into dermis of 161 g. fetus; 
X 500. (416| 



126 









.r 






W <w 4 



I 



tf jT^ 



•* 



t ' ^r # — • ^^^^ 

Mt # '• -* * a*. 

■ fe ^ A *f* * 

; , ... 



_ ^i* «• - — 






* 



% 



• «*" a 



I* 



Plate 33. — Surface of skin from back of 144 g. fetus ; each bump the site of a hair 
about to erupt; specimen removed from formalin to alcohol; now partly dried; 
posterior end of body at top of photo; X 40. (3910) 



127 
553006O-62-10 




Plate 34. — Fetus of 260 g. ; 19 January ; when first, almost invisible, hairs are 
appearing on face and head; superciliary and niystacial vibrissae well developed. 

(38581 



128 




fe^ •" ./ ;, "V.;, 4 - T»/i-^ tfjf .•" '."■ ' .- 




Plate 35. — Earliest pelage, on cheek-skin of 260 g. fetus. (Above) Median section 
showing roots; X 100. (Below) Surface view showing shafts of several hairs 
and many whitish bumps indicating sites of other hairs about to erupt ; posterior 
end at top; X 25. (4168 and 4123) 

129 





f. 




Plate 36. — Body of fetus, weight 372 g. ; standard length 255 nun. ; female ; photo- 
graphed in fresh condition out of mother killed 1C February and held on ice until 
23 February. (1802 A) 



130 



• 



im 



§4*'* 












I 



i 

m 






.>: 



* •& / 'fit 



i 



<*5 


















I * , »S.f:i§ r 



li 



\ 



& 



Plate 37.— Horizontal section, at depth of about 0.1 mm., from back of neck of 
575 g. fetus ; posterior end at top ; X 200. The 5 large objects are hair follicles ; 
the numerous small dark ones may be (?) underhair primordia. (4191) 



131 




Plate 38-A. — Underhairs erupting on back of 660 g. fetus ; X 2~>. 



i ' ' \\m\ • 









(3926 



| 



I 





: l/ll/l ! I IS III/. 



Plate 38-B. — Large, coarse guard hairs beginning to appear in field of smaller 
guard-hair tips and underhairs; on back of 1.93 kg. fetus; X 10. (4134)j 

132 



illfll 




Plate 39. — Developing underhairs and guard hairs on back of neck of 1.45 kg. fetus ; 
earlier growth at anterior end (bottom of photo) ; X 10. (4139) 

133 




Plate 40. — Pelage developing on back of 2.21 kg. fetus ; 2 May ; anterodorsal view ; 
X 50. (3931) 



134 



^ 



Plate 41-A. — Head of 1.7 kg. fetus ; natural size. 



(4132) 




Plate 41-B. — Three male fetuses showing pelage development; 1.4 kg. (27 April) ; 
2.7 kg. (28 April) ; 3.7 kg. (9 May). (KWK 50-630) 



135 




Plate 42. — Bases of developing underhairs and guard hairs on back of 2.44 kg. 
fetus (0.5 mean newborn weight) ; fibers shaved; partly dried; X 40. (3974) 



136 







Plate 43.— Twin fetuses of 9 May; removed from uterus preserved in formalin; 
X 0.33. Each is a female in nearly mature birtheoat ; one is 54 cm. and 3.43 kg. ; 
the other 53 cm. and 3.49 kg. (4017) 



137 




Plate 44. — Full-term fetus ; female ; weight 3.3 kg., length 54 cm. ; 6 July ; pelage 
glossy black. (1682 and 1683) 



138 




Plate 45. — Black pup, newborn female ; anterodorsal view of hair shafts rising from 
skin ; formalin specimen ; X 40. Note coarse guard hairs and fine underhairs 
rising more or less independently. (3937) 



139 



«*r, 



iw 
















mm 



m^S^W^ii^ 



Plate 46. — Horizontal section, at depth of about 0.1 mm., from back of neck of 
full-term fetus (black pup); posterior end at top; X 200. Follicular bundles 
include 1, 2, or 3 hairs; the anterior one a guard hair. (4193) 

140 




Plate 47-A. — ( See pi. 46. ) Section at depth of about 0.2 mm. ; through one follicular 
bundle ; X 400. Note guard hair at right, underhair at left, and sebaceous glands 
above and below. (4195) 



.£&B$m&*i 





■Ml 



9 







Plate 47-B. — (See pi. 46.) Section at depth of about 0.8 mm.: bulb of guard-hair 
follicle at right ; bulb of underhair follicle at left, surrounded by primordia of 
adult-type underfur follicles ; X 400. (4196) 



141 




Plate 48. — Black pups, molting, on 29 July. Loose birthcoat fibers are blowing 
about on the sandy rookeries at this time of year. (2250A) 



142 



til'/il 



■J 



■ ■ 



1 



Plate 49. — Black pup, molting ; approaching silver stage ; 15 September ; wet pelage 
of back of neck ; x 2.3. Pup had. come out of ocean and shaken itself, and was 
sprawled on a rock. Guard hairs 20 mm. in length ; underhairs 11 mm. ( See 
plate 50.) (4036) 



553006O-62-11 



143 




Plate 50. — Black pup, molting; 15 September; posteroventral view of fresh, dry 
pelage from back ; X 8. (Same specimen as shown in plate 49.) Pigmented roots 
of many adult-type guard hairs may be seen in the skin; adult-type underfur 
fibers are beginning to dominate the surface pelage. (4142) 



144 





B 





Plate 51. — Tanned pelts showing transition from birthcoat to first adult-type pelage. 
A. — black pup; newborn female; 12 July; 4.42 kg. (9.75 lb.). B. — black pup; 
molting female; 11 August; 9.30 kg. (20.5 lb.). C. — black pup; molting male; 
29 September; 14.7 kg. (32.5 lb.). D. — silver pup; autumn female; 13 October; 
15.2 kg. (33.5 lb.). (4190) 



145 




Plate 52-A. — Head of silver pup; male; 24 September; entire weight of animal 
12.7 kg. (28 1b.). (4056) 




Plate 52-B.— Similar ; female; 28 September; weight 11.8 kg. (26 lb.). (4066) 



146 




Plate 53. — Tanned pelt of yearling, pelagic; male; 25 April; X 0.2. Compare 
with silver pup in pi. 52-A. Dark streak along back is an artifact, a result of 
folding. (4000 NWC 52-3656) 



147 





^^^^s^ 






i 




V 




\ 








-*m 



Plate 54. — Tanned pelt of yearling, autumn; female; 31 October; in second adult- 
type pelage; X 0.2. (4000 BDM 25) 



148 



Plate 55. — Head of yearling, autumn; male; 26 September. 



(4060) 



149 




Plate 56. — Four-year-old male on 26 July, entering its 5th year of life and its 5th 
molt. This molt centers in October. Length 141 cm. (55.5 in.) ; weight 51.2 kg. 
( 113 lb. ) ; scar on neck represents hot-iron brand applied in first summer ; dorsal 
view. (1721) 



150 




Plate 57. — Similar to preceding figure; ventral view. 



(1722) 



151 




Plate 58. — Eight-year-old male on 2 July; length 191 cm. (75 in.) ; weight 1S4 kg. 
( 405 lb. ) . Note metal tag, applied 8 years previously, on right fore flipper. ( 2571) 



152 




Plate 59.— Nine-year-old male on 27 June; length 196 cm. (77 in.) ; weight 188 kg. 
(415 1b.). Note hot-iron brand, dark spot on lower edge of mane. (2564) 



153 





Pl^te 60. — Tanned pelt of adult female (age 10+) taken on breeding ground on 
30 October ; in dingy old pelage ; X 0.14. Compare plates 61 and 62. 

(4000 USNM 28603F) 



154 



(Plate 61. — Same pelt as in preceding figure; reverse side showing roots of replace- 
ment guard hairs; a typical unprime skin; X 0.14. (4000 USNM 2S6032L) 



155 





Plate G2. — Tanned pelt of adult female (age 10+ ) taken at sea on 28 March ; in brighi 
new pelage; X 0.14. Compare plates 60 and 61. (4000 SITKA 50-25) 



156 




5rSG 






Ik' Wv 






it*' 



jftft&b . 




Plate 63. — Horizontal section, at depth of about 0.1 mm., from back of neck of 
yearling during molt ; 26 September ; posterior end at top ; X 200. In two of the 
larger follicular bundles the wide blade of a new guard hair is erupting. (4204) 

157 




Plate 64.— (See pi. 63.) One follicular bundle ; X 600. Note (from top to bottom) 
bundle of underfur fibers, with newer, strap-shaped fibers above and older, cylindri- 
cal fibers below; sweat duct (left); wide, pigmented blade of erupting guard 
hair; round, translucent root of old guard hair. (4207) 



158 




Platb 65. — Molt line on rump of yearling, autumn ; female ; 3 October ; X 0.7. 
Second adult-type pelage is progressing toward rear (toward bottom of figure). 

(4070) 



553006O-62-12 



159 




Plate 66. — Tanned pelt of 2-year-old female; 21 September; in third molt; pelage 
short and lacklustre; X 0.2. (4000 BDM 239) 



160 




Plate 67. — Reverse side of unprime pelt which was shown in plate 66 ; anterior end 
toward bottom of photo ; X 40. Note black stumps of replacement guard hairs, 
cut off on 21 September before they had reached the surface. Insert is X 4. 

(4010 and 4008) 



161 




Plate 68-A. — Two mystacial vibrissae from 6-year-old female, showing massive 
connective-tissue sheath around root; specimen cleaned by marine amphipods; 
X 6. ( 413 °) 




Plate 6S-B— Vibrissa of 10-year-old male showing smooth surface; X 50. (3951) 







Plate 69.— Cross section of vibrissa of black pup (full-term fetus) showing: open 
medulla ; broad cortex, pigmented except for narrow peripheral band ; thin cuticle 
appearing dark under this illumination; X 400. (4137) 



163 




Plate 70-A. — Face of 103 g. fetus with full set of 20 mystacial vibrissae; X 6. 

(4126) 







Plate 70-B. — Vibrissae of adult male turned outward and forward in threat reaction 
during breeding season: l."> July: vibrissae white at this age. (2647) 



164 




Plate 71. — Vibrissae of subadult males, ages 3 and 4 years, plucked on killing field by 
small boy. In these year-classes the vibrissae are beginning to turn white at base, 
and to present a mottled appearance. (2221) 



165 




Plate 72.— Four pelage phases of the pup on 11 October (reading clockwise) : pup 
born late in summer and termed "black pup, molting" ; pup in first adult-type pelage 
termed "silver pup" ; and two freaks termed "chocolate" and "albino," respectively. 

(2323) 



166 




Plate 73-A. — Albino pup in first adult-type pelage, corresponding to "silver" pelage 
of normal pup in autumn; male; 10 October. (2321) 




Plate 73-B. — Head of female albino pup ; Seattle Zoo ; 2 December ; dark areas are 
stained. (4131) 



167 



/^K 




Plate 74. — Albino adult female collected on rookery, 15 August, in stained coat which 
she had been wearing for about 10 months. (2871) 



168 




■KKBSHhBHIkSHK 






Plate 75-A. — Piebald subadult male; 8 August. 



(1044) 




Plate 75-B. — Partial albino pup in birthcoat (molting) ; 11 August ; eyes and flippers 
pinkish white; underhair white; guard hair grayish brown. (2436) 



169 




Plate 76-A. — Two adult bulls, exemplifying range in pelage color from light to dark ; 
22 July. (2056) 




Plate 76-B — Pelt of subadult male; pale phase; 5 August. (4000 BDM 350) 



170 




Plate 77-A. — Adult cows with blotchy or "rubbed" pelage ; guard hair absent in 
patches; 23 July. (1004) 




Plate 77-B. — Similar ; body of 6-year-old cow ; washed and nearly dried ; 12 
September; dorsal view. (4031) 



171 





Plate 78. — Body of subadult male with "rubbed" back ; 26 July. 



172 








Plate 79. — Top of head of subaclult male shown in preceding plate. 



(1872) 



173 




Plate 80. — Fresh, clamp pelt of 2-year-old female lacking most of guard hair ; guard 
hair present only in patches on face and on bases of flippers; 2 July. (2821) 



174 




Plate 81. — Tuft of fur from "rubbed" area on back of subadult male, showing absence 
of guard hairs but fairly normal appearance of underfur fibers ; 27 July ; postero- 
ventral view: X 15. (3977) 



553006O-62-13 



175 



*y^% 



Plate 82-A. — Bare, scabby areas on rump of black pup, molting ; 16 July. Note louse 
at right of center. (1832) 




Plate 82-B.— About 130 lice feeding on penial opening of black pup, molting: 11 
August; X 4. (1908) 



176 



.*" 




*m*> 



£***"** 



«4r-««P 



Plate 83. — Pachyderma ; subadult male skin rejected from commercial take on St. 
Paul Island ; 3 July ; anterior edge of section cut from back ; alcohol ; X 8. (4167) 



177 




Plate 84. — Pelt of old female exemplifying poor fur quality. Killed 10 August; 
unhaired in St. Louis on 28 October of following year ; dried on hoop. 

(3996 ex Harry May) 



178 




Plate 85. — Right fore flipper of subadult male, showing blisters on dorsal surface ; 
16 July ; fresh ; about X 0.6. On this individual, both surfaces of all flippers were 
affected. ( 2859 > 






179 




Plate 86. — Cryptorchid killed on 7 July ; length 188 cm. (74 in.) ; weight 101 kg. (222 
lb.). Note slender, ungainly appearance of trunk and limbs; absence of wig and 
mane. (2000) 



180 




Plate 87-A. — Cryptorchid in right foreground, treated as "female" by harem bull at 
left; 15 July. (1708) 




Plate 87-B. — Tanned pelt of same cryptorchid, original body weight 101 kg. (222 
lb.) ; length of tanned skin, snout to tip of tail, 200 cm. (4000 BDM 86) 



181 










'"'H Y >'^ 







Plate 88. — Brown alga Ectocarpus on left flank and belly of subadult female shot at 
sea off central California; 12 December; (below) enlarged to natural size. 

(2464) 



182 






*, *•> * 



m 



Plate 89. — Barnacles Lepas attached to damp pelage on rump of subadult male about 
3 years of age; St. Paul Island; 18 July; X %. Pelt has been removed and blub- 
bered; many barnacles have been crushed. (1850) 

183 




Plate 90. — Blubbering or defatting a sealskin after it has been washed for 24 hours 
in cold running seawater ; St. Paul Island; 16 July. (2031) 



184 



. 



Plate 91. — Freshly blubbered sealskin. 



(2828) 
185 




Plate 92. — Commercial sealskin taken on 27 July from subadult nnile about 3 years 
of age; skin blubbered, wrung, shaken, and dried under fan for 2 hours: weight 
1.93 kg. (4.25 lb.) ; X %. , v-420) 



186 




Plate 9.3. — Sealskin at processing plant, having 
preparation for unhairing. 



been washed, hooped, and dried in 
(Fouke Fur Co.) 



187 




Plate 94-A. — Unhairing. 



(FoukeFur Co.) 




Plate 04-B. — Dyeing. 



(Fouke Fur Co.) 



188 






*y V ** &™f k 




,*, 



V 






V ' / 



Plate 95-A. — Fragment of untanned sealskin immediately after the unhairing 
process; lateral view; anterior end toward viewer's left; X 7. (2923 A) 




Plate 95-B. — Fragment of finished sealskin after all processing, including tanning, 
dehairing, and brown-dyeing; near posterolateral view; X 7. (2924) 



189 







-■Mate 





Plate 96. — Looking down on fragment of tanned, unhaired pelt from neck of 4-year-old 
male ; killed 22 July ; anterior end at top : X 10. Insert is natural size. 

(4006 and 4002) 



190 




Plate 97. — Demonstration pelt showing three stages of processing (from top to 
bottom) : "tanned in the hair," with all fibers intact ; "unhaired," with guard hairs 
removed ; and "finished," with fur fibers straightened and dyed brown. Specimen 
courtesy Fouke Fur Co. (2718) 



553006O-62-14 



191 




Plate 98-A. — Snout of yearling male ; 26 September ; about X 1.8. The "cowlick" 
above the rhinarium is double on certain individuals. (4062) 




Plate 08-B. — Profile of head of yearling female showing mouth ; 3 October. (4067) 



192 




Plate 99. — Eye of 3-year-old female; Seattle Zoo; 20 December; X 2. (3SS4) 



193 




Plate 100. — "Tears" on the cheek of a female seal on a warm, quiet day ; 15 July. 

(2625) 



194 




5 late 101-A.— Left ear of 1.19 kg. fetus; X 10. 



(4137) 




5 late 101-B. — Ear of 9-year-old female ; edges held apart by a pin ; 17 September ; X 3. 

(4048) 



195 




Plate 102. — The four mammary teats and the navel (center) on an adult female: 
20 July. (1848) 



196 



^?s" 







Plate 103-A. — Right anterior teat, forcibly extended and hardened in formalin, of 
old, parous female; 2 September; natural size. (4218) 





Plate 103-B. — Inside of two skins showing one posterior teat on a nulliparous (left) 
and on a parous, lactating individual (right). Seals killed 27 September; skins 
held in salt until 12 May ; blubbered and photographed. (2735 A) 



197 




Plate 104.— White discs show location of the navel and 4 rudimentary teats on a 
6-year-old male; 9 September; about X V\- (4028 A) 



198 



Plate 105. — Leather side of tanned, buffed pelt of yearling male killed about 1 
November, showing location of four rudimentary teats; X Vg- (4000 BDM 516) 



199 




Plate 106-A— Belly of adult female seal after 500 ml. embalming fluid has been 
injected in each teat. (4025 and 4026) 




Plate 106-B.— Portion of mammary gland peeled down (by knife) from smooth 
tissue which connects it to body; X 3. (4034) 



200 







Plate 107-A. — Tail of 6-year-old male ; scrotum relaxed ; 9 September ; natural size. 

(4029) 




Plate 107-B. — Tail of old bull ; scrotum pulled forward in order to reveal anus : 20 
September; natural size. (4049) 



201 




Plate 108. — Lower abdomen of female, weight 22.6 kg. (50 lb.) ; probably nulliparous ; 
4 August. Light-colored region is vestibule, with clitoris in front and anal opening 
behind and hidden. (1898) 



202 




Plate 109. — Surface of right heel of silver pup; 28 September; X 10. (4065) 



203 





Plate 110 —Right flippers of old male (above) and 7-year-old female (below) ; dorsal 
view; X0.18. (2413 and 2425) 



204 




Plate Ill-A. — Claws of 6-year-old male; digits 3 (in background) and 4 (in fore- 
ground) of left hind flipper; X 1.3. (4030) 




I 







Plate 111-B. — Roentgenogram of hind flipper of 3-year-old male, showing 3 functional 
claws; X 1.2. (4188) 



205 




Plate 112. — Adult female scratching herself, 
flippers are functional. 



Only the three middle claws of the hind 
(4140 ex Karl W. Kenyon) 



206 



U.S. GOVERNMENT PRINTING OFFICE : 1962 O — 553006