Skip to main content

Full text of "North American fauna"

See other formats 


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   ,lflJfc«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  0 

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  wTas  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  wrere  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  wTas  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  swrans  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  wrrote 
(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)  wTas  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 

0 
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  wTay  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  nowT 
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  lowrer  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  C02.  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  wThile  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 
Lave  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  7V2 
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  7y2 
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 
0 
3 

2 
4 
1 

0 
0 
0 
0 

3 
3 

5 
2 
4 

1 
4 
0 
2 
0 

2 
3 
0 
3 

4 

0 
4 
0 
2 
0 

2 
3 
3 
3 
4 
0 
4 
0 
0 
0 

2 
3 
3 
3 
2 
0 
4 
0 
2 
0 

2 
3 
3 
3 
1 
1 
4 
0 
2 
0 

0 
3 
3 

3 
0 
1 
3 

0 
2 
0 

2 
3 
3 
0 
1 
1 
2 
0 
3 
0 

2 
3 
3 
3 
1 
1 
2 
0 
0 
0 

2 

No.  2      

3 

No  3         

2 

No.  4    

3 

No  5         --- 

1 

No.  6    

1 

No.  7         

2 

No.  8 

0 

No.  9        

3 

No  10           

0 

Total 

31 

17 

Cygnet   loss   since    previous 

5 
3 

0 
0 

1 
1 

2 
1 

2 
1 

1 
1 

1 
1 

1 
1 

0 
0 

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  7y2 '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 

0 

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 

0 

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 

0 

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 ' 


0 

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 

0 

0 
0 

0 
0 
0 

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  vTas  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  wras  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 
0 
0 
1,3 
0 
3 
0 


Year 


1952 

1953 

1954 

1955 

1956 

1957 

Total 


Pairs 

Cygnets 

nested 

initially 

observed 

1 

3 

1 

o 

2 

2,2 

2 

3.3 

(') 

0 

1 

1 

14 

34 

Cygnets 
raised  to 
flying  age 


3 

2 

2,1 

3,3 

0 
0 


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-breeders1 

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  breeding1  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— 0 ;  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— 0  ;  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 — 0 ;  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 — 3V2  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) — liy2  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 — 0 ;  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  y16)  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  22  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  0  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}7pe  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  0  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  0  0 

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: 
0  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  y2  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) 
uThe  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 


0 


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 


0 


♦  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  S8y2 inches;  medium,  S8y2  to  4Qy2;  large,  42^ 
to  45% ;  extra  large,  45V&  to  4tSy2 ;  extra  extra  large,  481/£  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 


0 


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  y2oo  mean  newborn  weight,  and 
erupted  hairs  on  a  midwinter  fetus  of  y20  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.  (y2  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  y2oo  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. 0 
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. 0 

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 


0  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 


'"'HY>'^ 


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  **&™fk 


,*, 


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 


5late  101-A.— Left  ear  of  1.19  kg.  fetus;  X  10. 


(4137) 


5late  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