Ornithological Monographs 18

BREEDING

BEHAVIOR

BIOLOGY

OF THE

AND

OLDSQUAW

(CLANG ULA t-]YEAdrALIS L. )

BY

ROBERT

ORNITHOLOGICAL

M.

ALISON

MONOGRAPHS
PUBLISHED

THE AMERICAN

BY

ORNITHOLOGISTS'

1975

NO.

UNION

18


BREEDING

BIOLOGY

AND

BEHAVIOR
OF THE OLDSQUAW
(CLANG {_/L•I HYœA4t•ILIS L. )

BY

ROBERT

ORNITHOLOGICAL

M.

ALISON

MONOGRAPHS
PUBLISHED

THE AMERICAN

BY

ORNITHOLOGISTS'

1975

NO.

UNION

18


ORNITHOLOGICAL

MONOGRAPHS

This series,publishedby the AmericanOrnithologists'
Union, has been
established
for major paperstoo long for inclusionin the Union'sjournal,
The Auk. Publication
hasbeenmadepossible
throughthe generosity

of Mrs.
CarllTuckerandtheMarciaBradyTuckerFoundation,Inc.
Correspondence
concerning
manuscripts
for publicationin the seriesshould
be addressed
to the Editor,Dr. JohnWilliamHardy, Departmentof Natural
Sciences,The Florida State Museum, University of Florida, Gainesville,
Florida 32611.

Copiesof OrnithologicalMonographsmay be orderedfrom the Assistant
Treasurerof theAOU, GlenE. Woolfenden,Departmentof Biology,University
of SouthFlorida,Tampa, Florida 33620.
OrnithologicalMonographs,No. 18, vi + 52 pp.
Editor-in-chief: John William Hardy
SpecialAssociateEditor for this issue: Milton W. Weller
Author's address: Wildlife Branch, Ministry of Natural
Resources,Whitney Block, Queen's Park,
Toronto, Ontario.

Issued: September23, 1975

Price: $3.50 prepaid ($2.50 to AOU Members)
Library of CongressCatalogue Card Number 75-20649

Printed by the Allen Press,Inc., Lawrence, Kansas66044
Copyright @ by American Ornithologists' Union, 1975

TABLE
TABLE

of CONTENTS

INTRODUCTION

MATERIALS

OF

CONTENTS

...........................................

..................................................

AND

METHODS

......................................

iii
1

1

The Study Area ..............................................................................

1

Weather During the Study Period ...............................................

3

Vegetationin the Study Area ........................................................

4

Census Methods

5

Territorial

Winter
RESULTS

.............................................................................

Behavior

Behavior

........................................................................

...........................................................................

..........................................................................................

Population Characteristics.............................................................

7

8
8

8

Homing ......................................................................................... 10
Nesting ..........................................................................

11

The Eggs.....................................................................................

16

The Nest and Incubation

17

...........................................................

The Young ..........................................................................

23

Territoriality .................................................................................. 24

Displays ......................................................................................

32

Other Aspectsof ReproductiveBiology........................................... 43
Major FactorsInfluencingProduction............................................. 46
ACKNOWLEDGMENTS
SUMMARY
LITERATURE

.................................................................

..............................................................
CITED

.....................................................

iii

48
48
50


LIST

OF FIGURES

Figure 1. Location of study area .....................................

2

2.

Map of studyareashowingtraditionalnestinglocations............ 3

3.

Habitat of the study area .....................................

4

4.

Head plumagesof Oldsquaws............................................

6

5.

"Hummock"habitatin early June suitablefor nesting............ 7

6.

Oldsquawnestcup ....................................................

7.

Variationin headplumages
of downyyoungOldsquaws

.......... 23

8.

Plumagedevelopmentin youngOldsquaws........................

25

9.

Plumagedevelopment
in youngOldsquaws ...................

26

14

10.

Fall and winter plumagesof immature Oldsquaws................ 27

11.

Fall and winter plumagesof adult Oldsquaws....................... 28

12.

Characteristic
posturesof maleOldsquaws................................ 33

13.

Adult male OldsquawsperformingBreastDisplay.................... 35


LIST

Table

OF

TABLES

1. Historiesof bandedOldsquawsat Churchill,Manitoba ........... 11

2. Locationsof Oldsquawnestsin the studyarea, 1968 to 1971 .... 12
3. Concealmentrankingfor Oldsquawnestsin the study
area. Data from nestsof previousyears is also
presentfor comparison.....................................

13

4. Distancesbetween Oldsquaw nests and the nearest
open water ...........................................................

15

5. Nest associationbetweenOldsquawsand Arctic Terns
in the study area. Only active nestsof known
historyare included.....................................

15

6. Clutchsizesof Oldsquawsin the studyarea, 1968 to 1971 ....... 17

7. Avian and mammalianpredatorsand nestdestruction
in the studyarea .........................................................

19

8. Oldsquawnestlossesdueto predationin the studyarea........... 20
9. Summaryof the fate of Oldsquawnestsin the study
area, 1968 to 1971 ..............................................

21

10. Hatchingdatesof Oldsquaws
in the studyarea, 1968 to 1971 .... 22

11. Responses
of drake Oldsquawsto stuffeddecoysand
recorded vocalizations .........................................

29

12. The orderof displaysperformedby courtingmales.................... 37


INTRODUCTION

The Oldsquaw
(Clangulahyemalis),or Long-tailed
Duck, is probably
more numerousand widespread
than any other speciesof arctic waterfowl
(Delacour1959). It breedsacrossnorthernEurasiaandfrom PointBarrow,

Alaska,alongtheArcticcoastof NorthAmerica,
including
thewestern
coasts
of HudsonandJamesBaysandin northeast
Labrador(Godfrey1966).
Breedingindividualshave been found on almost all the Arctic islands of

Canadato a latitudeof 83ø north, and the speciesreportedlybreedson
Baffinand Southampton
Islands(Phillips1925). Althoughthe Oldsquaw
hasbredalongthe eastcoastof Labradorand in the Gulf of St. Lawrence,
thereis no evidence
that it breedstherenow. The bird is circumpolar
in
distribution
andcommonly
breedsin Greenland(to 83ø 23' N), Iceland,the
Faroes,the north coastsof Norway,Sweden,Finland,Russia,on the New

Siberian
Islands
andall alongthecoastof theArcticOceanto theBeringSea

(Millais 1913).

The winterrangeof the species
is well known(Millais 1913, Bent 1925,
Phillips1925). However,winteringareasof specificnorthernpopulations
are unknown,primarilybecause
of the lack of bandingdata. A largeproportionof the Icelandpopulationapparently
wintersin southwest
Greenland
(Salomonsen
1950) whilesomeindividuals
of certainRussianpopulations
winteron the westcoastof Europe(Boyd 1957).
The breeding
biologyof the Oldsquaw
is poorlyknown.Drury (1961)
mentioned
certaindisplaysthat he saw on the breedinggrounds.Millais
( 1913), Bent( 1925), andPhillips( 1925) described
nestlocation,clutchsize
and aspects
of adultbehavior,but no studiesof individuallyrecognizable
breedingOldsquawshave been published.
Themainobjectives
of myresearch
reported
in thispaperwereto determine:

(1) Oldsquawpopulationlevelsand breedingpair distribution,
(2) nest site tenacity and homing,

(3) nestdistributionand location,nestingsuccess,
and the effectof weather
upon nesting and renesting,
(4) territorial behavior, and
(5) to describespecies-specific
displaysof this speciesin sufficientdetail
to permitcomparison
with thoseof otherspecies.
MATERIALS
THE

AND

METHODS

STUDY AREA

Thispapercontains
the resultsof studiesof Oldsquaws
conducted
in four
consecutive
breedingseasons,1968 to 1971, at Churchill,Manitoba (58 ø
45' N, 99ø 5' W). I observedOldsquaws
on a 400.0ha area about5 km east


ORNITHOLOGICAL

HUDSON

59 ø
05'

•urchill

NO.

MONOGRAPHS

18

BAY

STUDY

AREA
58 ø

•)

Churchill

Stygge

45'

Lake

Churchill

58 ø

River

25'

•

5000

5000

58ø
05'

METERS

94ø 15'

93 ø 55'

Figure1. Map showing
locationof the studyarea.

ofChurchill
(Fig.1). TheOldsquaw
breeding
population
in thisareaaverages
45 pairs,themajorityarriving

in thefirstweekof June.Pairsaremost
commonly
observed
onsmallinlandponds
butindividuals
andsmallgroups
mayoccasionally
be seenamong
thechunks
of packice in openwateron
HudsonBay, especially
at the mouthsof rivers.

Thestudy
areaisabout14kmeastofthemouth
of theChurchill
Riveron
thewestcoast
ofHudson
Bayandiscontained
within58ø43'N to 58ø46'N
and93ø 52' W to 94ø 01' W (Fig. 2). The areaincludes
sparsely
forested

tundra,dry upland,marshland
andscrubland,
andextends
almost10 km
along

theHudson
Baycoast.JehlandSmith(1970) described
theareain
detail.Thereare90 pondsandlakesranging
in areafrom0.1 ha to about
20.3ha andaveraging
2.8 ha. Noneof thesepondsandlakeswasreduced
tomud-flatconditions
duringthestudy.Mostof thesmaller
pondsareice-free

by10June,
andallthelarger
ponds
andlakes(except
Stygge
Lake)areclear
of iceby 20 June.Remnant
icemayoccuralongthecoastof Hudson
Bay
until late July.

Oldsquaws
consistently
outnumber
allother
species
ofwaterfowl
inthestudy
area.However,

theCommon
Eider(Somateria
mollissima)
andtheCanada


1975

ALISON:

OLDSQUAW BIOLOGY

3

94ø 00'

I

93ø 53'

I

I

I

•

I

I

I

I

HUDSON
BAY L

Figure 2. Map of the study area showing traditional nesting location. Minimum
densitiesof one nest per 2.3 areas are shaded.

Goose (Branta canadensis)are quite common. The following speciesalso
occurmoreor lessregularly: GreaterScaup(Aythya marila), Green-winged
Teal ( Anas carolinensis
), Pintail ( Anas acuta), American Widgeon ( A nas
americana), Red-breastedMerganser (Mergus serrator), Mallard (Anas
platyrhynchos) and Common Scoter (Melanitta nigra).
In orderthat breedinggrounddata couldbe supplemented
by observations
of winter behavior,Oldsquawswere observedduring the winters of 1967
to 1969 at Toronto Harbor, Ontario (43 ø 40' N, 79 ø 30' W). The average
winteringpopulationof this speciesat Torontocomprises
about7,000 individuals. In addition,observationsof the behavior of six captive Oldsquaws
supplemented
field data.
WEATHER

DURING

THE STUDY PERIOD

In 1968, daily temperaturesin the last 10 days of May were 3.$øC above
normal (-2.1 øC), whereasJunetemperatures
were $.1øC and 1.9øC, respectively,below normal (6.1 øC). Most pondswere not clear of ice until 20' June
that year. In 1970, temperatures
in the final week of May averaged2.2øC
abovenormal, June temperaturesaveraged3.0øC abovenormal, and ponds
were clear of ice by 1 June.
Juneprecipitationfrom 1968 through1970 wasbelownormal ($1.4 ram),
but in 1971, abovenormal rainfall was recordedfor that month. Wind velocities throughoutthe study period averaged21 kmph and the mean daily


4

ORNITHOLOGICAL

MONOGRAPHS

NO

18

Figure 3. Habilat typical of the study area showingBlack Spruceand open tundra
association.

maximumwind velocitywas 38.8 kmph. In general,the 1968 breeding
period at Churchillwas dry and moderate,the 1969 seasonwas dry and
cold, the 1970 periodwas dry and warm, and the 1971 seasonwas wet and
cold.

VEGETATION

IN THE STUDY AREA

Despite an overall impressionof monotonoussamencss(Fig. 3), the vegetationin the studyareavariedconsiderablylocally. On the basisof component
plant types, it was possibleto subdividethe study area into the following
categories: tundra, marsh,dry upland and scrubland.
(a) Tumlra.--Tundra includes those typically grassy expanseslying
beyondand occasionallywithin the limit of rank forest. The study area is
mainly tundra and containsextensivelowland areas dominated by sedges
includingHudsonBay Sedge(Carex amplyorhyncha)andNorthernBogSedge
(C. gynocrates). Grasses(Poa spp.) and dwarf willows (Salix spp.) commonlyoccur.Frequentlybryophytes
andlichenspredominate,
thewholeforming a continuous
growthtypicallylessthan 35 cm in height. Isolateddwarf
birches (Betula spp.) may occasionallyoccur.
(b) Marsh.--As a result of the relatively poor drainage large expanses
of shallow surfacewater generallyoccur. In such marshlandareas, the dom-


1975

ALISON: OLDSQUAW BIOLOGY

5

inant vegetationincludeshygrophyticsedges,suchas Water Sedge(Carex
aquatilis),grasses
(Arctagrostis

spp.) aswell ashygrophilous
willows(Salix
candidaand $. myrtillifolia) (Scoggan1959). Occasionally
"hummocks",
averaging
25 to 50 cm highoccurin marshland
tracts,the resultingformation
beingtermed"hillocktundra" (Polunin1967). Conditionson hummocks
are relativelydry, and lichensand ground-shrubs
commonlygrow. Depressionsbetweenadjacenthummocksform boglets.
(c) Upland.--Rockscommonlyprojectabovethe surfacecontour.The
vegetation
in thesetractsis usuallylessdensethan in moistareas. Xerophiloussedges,
grasses
(Hierochloe
spp.)andwillowsfrequently
occur.Although
lichensand bryophytesare plentiful,and in mostareasappearto dominate,
the vegetationcommonlybarely covers the ground.

(d) Scrubland.•Frequentlywithin the tundra proper, especiallyalong
watercourses
and in damp areas,densegrowthsof dwarf willows and/or
birchesappear.In mostinstances,
the overallgrowthdoesnot exceed0.9 m
in heightbut rarely may attain heightsof more than 1.8 m. Scrub areas
are usuallysodenseandtangledthat few otherplantsoccur. However,where
the dominantwillowsand birchesare lesstightlypacked,a dense,rich underflora of lichensand/or mossesmay develop.
CENSUS METHODS

I conducted
total censuses
of adultand subadultOldsquaws
throughoutthe
study. The censusroute requiredabout three hours to complete,and I
alwaystook censuses
between07:00 and 10:00 C.S.T. I madecountsdaily
from I June to. 30 June. By 1 July the entire breedingpopulationhad
arrived and daily censuses
were no longer necessary.In every instanceage
(adult or subadult) and sex of each individualwere readily determinedby
plumage.Individualvariationincludingvariationin plumage(Fig. 4) and
vocalizationmade recognitionof most adults possiblewithout marking
(dyeing)andmadeit possible
for meto tracethedailymovements
of individual
pairs.
In order to examinenest site tenacity and homing in Oldsquaws,it was
necessary
to captureand band severaladults. Althoughadult femaleswere

relativelyeasilycapturedwith mistnetson theirnests,adultdrakesweremore
elusive.All femalesbandedat the nestsitesubsequently
returnedand resumed
incubation.In capturingotherindividuals,fourqnchmeshmistnetswereplaced
horizontallyover holes in the ice on inland lakes and ponds. Oldsquaws
attemptingto dive into theseholeswere invariablycaughtin the nets. Whenever a female was capturedin this manner, one or more drakeswould fly
from nearbyopenwater and land near the entangledbird. Frequentlythese
individualsbecamecaughtin the net as well. Pairs were often capturedin
this manner. By 1 July 1971, 45 Oldsquawshad beenbanded.

6

ORNITHOLOGICAL

MONOGRAPHS

NO.

18

A

G

H

Figure 4. Head plumagesof Oldsquaws.Male: A. Immature about 8 weeks old;
B. Immature about 16 weeks old; C. Adult. Female: D. Immature about 8 weeks

old; E. Immature about 16 weeks old; F. Adult. Head plumagesof breeding adults.
G.

Male; H.

Female.

Figure 5 showstypicalnestingsubstratein early June. Most activenests
were found by careful censusesof traditionalOldsquawnestingcolonies.

However, extensivesearchingwas required before certain mainland and
isolated island nests were located. As a result I did not discover several nests

until the femalehad commenced
egglaying,and thesewere not includedin
the data. The location of each nest was noted in relation to the nearest open

water and the nearestactive Oldsquawnest. All active nestswere photographed.Nest surveys
frequentlyyieldednot only activenestsbut alsonest
cupsof previousyears (duringthe study,I found 259 nestsitesof previous

years). All activenestswerecensused
daily but if the femalewas present
the nest contents were not examined.


1975

Figure 5.

ALISON:

OLDSQUAW

BIOLOGY

"Hummock" habitat in early lune suitable for Oldsquaw nesting.

The degreeof concealment
of eachnestwassubjectively

ratedas follows:
(1) Well concealed.--Incubatingduck concealedfrom aboveand all sides.
(2) Open from above.--Duck concealedfrom all sides.
(3) Partially concealed.-•Duckpoorly concealedfrom sidesand not at all
from above, and
(4) Poorly concealed.•Duck not concealedfrom any side or from above.
Predatorcensuses
were frequentlyconductedalthoughno attemptwasmade
to locatenestsor burrowsof predators.
TERRITORIAL

BEHAVIOR

I examinedthe territorial behaviorof 10 adult drake Oldsquaws.Five of
these individuals were exposedto each of several combinationsof decoys
(one combination per bird for a period of 5 minutes every 2 days). Five
otherswere exposedto the samecombinations,but in each instancea 1-minute
recorded sequencecomprising 24 "ahr-ahr-ahroulit" vocalizations (Alison
1970) was playedfrom a hidden cassetterecordercapableof producingabout
2

watts.

Oldsquawdecoyswere preparedfrom skinstaken from a seriesof specimens
caughtin fishermen'snets near Picton, Ontario, in May 1969. Each skin was
filled with "Styrofoam" and subsequentlyshaped. Glass eyes were placed
in each decoy and faded bills were painted appropriately.


8

ORNITHOLOGICAL

MONOGRAPHS

NO.

18

The decoys
wereplacedin eachterritory,10 m offshore,
whilethe defending
drakewassubmerged
(feeding)andthuspresumably
unawareof my presence.
Eachexperiment
requiredabout15minutes
to complete,
andI remained
hidden
behindsmallspruceclumpsduringthe period.
Aggressive
behavior,including"ahr-ahr-ahroulit"
vocalizations,
bill toss,
threatpostures,
or attackcomprised
a positiveresponse.Any otherbehavior
or no responseat all within 5 minutescompriseda negativeresponse.
WINTER

BEHAVIOR

Althoughday-to-dayrecognitionof an individualbird was difficult,it was
nonetheless
possibleon a daily basisto identifyin flockscertainindividuals
possessing
uniqueplumage.As a resultbehaviorof individualmalesor females
was sampledat randomon separateoccasions.Initiation and development
of epigamicbehaviorwere observedin two marked adult males betweenOctober 1968 and May 1969.
RESULTS
POPULATION

CHARACTERISTICS

Adultandsubadultpopulations.--Oldsquaws
generallyarrivedon thestudy
areain June.Individualswerefirst notedon 3 June1968, 3 June1969, 4 June
1970,and31 May 1971. I wasableto studyin detailthearrivalof thebreeding

population
eachyear. In 1968therewere47 pairsof adults;8 pairsof these
arrivedon 3 June,4 on 6 June,20 on 1! Juneand 15 on 16 June. In 1969,
therewere43 pairsof breedingadults;2 pairsarrivedon 3 June,3 on 4 June,
10 on 8 June,4 on 11 Juneand24 on 23 June. In 1970, 47 pairswereresidenton the studyarea;5 pairsarrivedon 4 June,4 on 5 June,11 on 9 June,
11 on 11 June,10 on 12 Juneand6 on 20 June. In 1971,therewere44 pairs
of adultbreeders;
2 pairsarrivedon 31 May, 2 on 1 June,5 on 4 June,21 on
10 Juneand 14 on 16 June. Thus, in the studyperiodthe entirebreeding
populationinvariablyhad arrivedby 23 June. Of the non-breeding

adultOldsquawsthat were occasionally
observedon the studyarea, nine unpaired
femalesremainedthroughoutonebreedingperiod.
From 1968 through1971, 8 (all females),9 (8 females),11 (all females)
and5 (all females)subadults,
respectively,
remained
in the studyarea. None
of thesebirdscouldbe identifiedas homingindividuals,as iramatureswere
notbanded.Commonly
subadults
remainedin smallflocksalooffrombreeding
pairs.

Invariablythe entirebreedingpopulationarrivespairedand only a few
surplusindividuals,
mostlyfemalesandsubadults,
are presentthroughout
one
breedingseason.Bent (1925), Salomonsen
(1950), Millais (1913) and
othershave reportedin contrastthat pairingusuallydoesnot occurin this


1975

ALISON: OLDSQUAW BIOLOGY

9

species
untilthebirdsarriveat thebreeding
grounds.I am convinced
thatthis
conclusion
wasbasedupona confused
recognition
of pair-forming
and pairmaintainingdisplays(Alison 1972). It has beennoted (Alison 1970) that
pair-formation
requires
at least30 daysandthatthemaximum
possible
breedingperiodat Churchill
wouldprobably
notbesufficiently
longfor pair-forming
to occur on the breedinggrounds.Drury (1961) reportedthat on Bylot
IslandOldsquaws
appearedto be pairedon arrivaland that therewere only
a few surplusdrakes.Althoughit has generallybeenreportedthat subadult
Oldsquaws
summeron the Arcticcoasts,this doesnot appearto be strictly
true. Thirty-threesubadults,
mostlyfemales,remainedon the studyarea in
JuneandJuly (varyingnumbersin eachyear). Thesebirdsmayhavebeen
rearedin the Churchillarea or evenin the studyarea in previousbreeding
seasons.

An apparentinconsistency

recursconcerning
the relativeproportionsof
the sexesin migratingflocksof this species.Drury (1961) and Salomonsen
(1950) reporteda surplusof drakeson the breedinggrounds.Ekblaw (in
Bent 1925), stated that males outnumber females in the first flocks but

later the femalesappearto be as numerousas the males. However, records
of Femalesmight have resultedfrom erroneousidentificationby previous
authorsof immaturefemalesand subadultsas males. Migrating flocks of
Oldsquawsgenerallycompriseseveralhundredindividuals(Bergman and
Donner 1964). At Churchill the breedingpopulationdoes not arrive en
masse;rather,severalpairsarrive at intervalsof a few days. It is likely therefore that as a migratingflock passesa certainbreedinglocality,thosefemales
returningto natal ponds leave the flock taking their respectivemates with
them. Thus, homing femalesand pairs are encounteredat the southernextremity of the breedinggrounds. As more homingfemalesand pairs leave
a migratingflock, the relativeproportionof adult drakesinevitablyriseswith
increasinglatitude. This would explainthe notationin Bent (1925) that in
the far north males outnumber

females in the first flocks.

There is also some

evidencethat the mainbody of subadultsis aboutone week behindthe adults
in migratingnorth. The later arrival of the subadultsin the far north would
appearsimplyas an increasedproportionof female-likeindividualsin observed
flocks--an increasewhichmay have beenmisinterpreted
by authorscited by
Bent (1925) and othersas constituting
the arrival of the as yet unpairedadult
females.

Adult and subadultmortality.--Nest surveys,censuses
and other activities
associated
with this studyresultedin regularintensivecoverageof the entire
studyarea. The remainsof deadOldsquaws
that appearedto havebeenkilled
and eaten by predatorswere occasionally
found. An unknownnumberof
thesemay have succumbed
first to diseaseand subsequently
been eaten by
scavengers.
The numberof adultdrakesin the studyareain 1968 was47 and


10

ORNITHOLOGICAL

MONOGRAPHS

NO. 18

of thesetwowerekilled(in June)by unknown
avianpredators.In 1969and
1971, no adultdrakeswerefounddead. In 1970, oneadultdrake (out of 47)
waskilledby a maleMarshHawk (Circuscyaneus).In everyyear of the
studyfemalemortalitywas nil, and no subadultswere found dead. On 12
June! 971, an apparently

unhealthyadultfemaleOldsquaw(whichcouldnot
fly) was repeatedlyattackedand finally killed by an Arctic Loon (Gavia
arctica) outsidethe studyarea.
Residence.--Adultpopulations
generallyincreasedin Juneand decreased
in JulyandAugust.Althoughresidentbreeders
hadinvariablyarrivedby 23
June,non-resident
migratingindividuals
wereoccasionally
foundin the study
areauntil ! July. Eachbreedingpair occupieda smallpond and sometimes
twoor threepairsoccupied
eachof thelargerlakes.Of the90 pondsandlakes
in the studyarea,only 27 were occupiedannuallywhereas49 were never
occupiedby breedingOldsquaws.The attractiveness
of certainpondsor lakes
may havebeena functionof nestingcover (i.e. the presenceof long-isolated
peninsulasor islands).

Drakesweregenerallysedentary
throughout
Juneevenafterincubationhad
begun,butby 10 Julyseveraldrakeshaddeparted
thestudyareaandgathered
in flockson the HudsonBay coast.By 30 Julyno adultdrakesremainedin
the studyarea. Smallflockscomprising
adultdrakesandsubadults
of both
sexesoccuralongthecoastsof Greenlandin earlyAugust(Salomonsen

1950).
Drury (1961) observedflocksof Oldsquaws,predominantlydrakes,in the
seaoff Bylot Island in late July. The adult femalepopulationin my study
remainedconstantthroughoutJuly,but by mid-Augusta slowbut continuous
declinewasnotedas femalesand broodsdesertedthe inlandponds. By 31
Augustno adultfemalesremainedon the studyarea.
HOMING

Threeduckswereeachcaughtat their respective
nestsitesin two consecutive yearsand one other nestedat the samelake in two consecutive
years
(Table 1). Adult femalesshowa strongtendencyto returnto previousnest
sites: 3 of 13 individuals
returnedto a specificsiteusedpreviously;
8 females
returnedto the samepond as in a previousyear; 2 ducksbandedon nests
movedto nearbyponds. Additionalfemalesmay have movedto nearbyponds
locatedoutsidethestudyareaandremainedthereundetected.Two drakeswere
recapturedat the samelake in at leasttwo consecutive
years;the seconddrake
was capturedin four consecutive
years at the samelake.
Of threeduckscaughtat the samenestsitein two consecutive
years,all had
hadsuccessful
nestsin theprevious
year. Fourfemalesnestedat thesamepond
in two or more consecutive
years,and of thesetwo had had successful
nestsin

the previousyearswhereastwo had not. Of the latter two, one nestedon the


1975

ALISON:

OLDSQUAW
TABLE

BIOLOGY

I1

1

HISTORIESOF BANDEDOLDSQUAWS
AT CHURCHILL•MANITOBA
Band Number

CPF

2635

History of Individual

Adult

female, banded on nest 1969; rebanded 596-64602 on same nest
1970; at same lake 1971.

596-64605

Adult

female, banded on nest 1970; on same nest 1971.

CPF

Adult

female, banded on nest 1969; at same lake 1971.

596-64606

Adult

female, banded on nest 1970; at same lake 1971.

CPF

2625

Adult female, banded on nest 1969; rebanded 596-64608 on same nest
1970; at same lake 1971.

CPF

2629

CPF

2608

Adult female, banded on nest 1969; at same lake 1971.
Adult female, banded on nest 1969; at same lake 1970; rebanded

CPF

2618

CPF

2628

Adult female, banded on nest 1969; rebanded 596-64676 at lake 75 m

CPF

2638

Adult female, banded on nest 1969; rebanded 596-64677 at same lake

2605

596-64674

at same lake

1971.

Adult female, banded on nest 1969; rebanded 596-64675 at lake 100 m
east

1971.

southwest

1971.

1971.

596-64601
CPF

2640

Adult male, bandedas adult 1970; at same lake 1971.
Adult male, banded as adult 1968; at same lake 1969; rebanded 59664609 at same lake 1970; at same lake 1971.

sameislandas in a previousyear (whenthe nesthad beendestroyed);the
other moved from an island nest site to the mainland.
NESTING

Nestlocationanddistribution.•Phillips
(! 925) notedthat althoughmany
nestsof this speciesoccurredsingly,somefemaleswere found nestingin
colonies.In this studymanyOldsquawneststendedto be clustered.Only
oneinstance
of parasitism

of onefemaleOldsquaw
on anotherwasnoted,and
in thatinstance
bothfemaleswereseemingly
unawarethat the nestwasbeing
shared.Onefemalesubsequently
incubated
the clutchwhilethe otherpermanently desertedthe nest but remainednearby.
Excludingthis instanceof parasitismthe minimumdistancebetweenany
two active Oldsquawnestswas 7.5 cm. In 1971, three active nestswere
observed
withinthe sameareaof 0.3 m2. Invariablywhenevermorethanone
activenestoccurred
at a particulartundrapond,theneststendedto.beclumped.
Obviouslywheneverthe chosennest site was on an island, the maximum
possible
distancebetweenanytwo activenestswasdetermined
by the sizeof
the island. However,mainlandnestsweregenerallyclusteredas well. Around


12

ORNITHOLOGICAL

MONOGRAPHS

TABLE

NO.

18

2

LOCATIONS
OF OLDSQUAW
NESTSIN THE STUDYAREA,1968 TO 1971
Number

Nest Location

of Active

Nests

1968

1969

1970

1971

Tundra

3(12.7)*

5(19.1)

2(8.4)

2(6.9)

SpruceForest

3(12.7)

2(9.5)

5(20.8)

2(6.9)

Mainland

Dry Upland

0

0

0

2(6.9)

Marsh

3(12.7)

2(9.5)

1(4.1)

3(10.3)

! (4.8)

1(4.8)

Islands

Scrubland

11(57.1)

11(57.1)

16(66.7)

0

18(62.1)

2 (6.9)

Total

2!

2!

24

29

Figures in parentheses are percentages.

onethree-hectare
tundralake, all four activenestsoccurredon oneisolatedpeninsula,thatmeasured
about12 x 3 m. Furtherinvestigation
revealed24 other
nestcupsof previousyearson the samepeninsulawhereasno additionalnest
cupswerefoundanywhereelsearoundthat lake. In one instancetwo mainland
nestsat a 0.4 ha pondwere 1.8 m apart. The maximumdistancebetweenany
two activemainlandnestsat the samepondor lake was 409.4 m. Nonetheless,
36 of 95 nests(37.9%) occurredsingly(i.e. no other activenestwas within
60 m).

Oldsquawnestshave been found on offshoreislands,on mainlandtundra,
alongthe edgesof freshwaterpondsand on islandsin theseponds (Millais
1913, Phillips 1925, Bent 1925, Salomonsen1950). In the study area, 56
of 95 nests(58.9%) of known historywere locatedon islandsin freshwater
pondsor lakes (Table 2). Twenty-fournests (25.3%) were on mainland
tundra whereasnine (9.5%) were in marshy areas. Four (4.2%) nestswere
foundin scrublandand only two (2.1%) were in dry upland. There was no
significantdifferencebetweenthe number of individualnestson islandsas
comparedwith mainland nests (P •<0.05).

Thestudyareacontains

79 islands
in freshwater
pondsor lakesandof these
seven
hadat leastoneactiveOldsquaw
nestin fourconsecutive
years.Active
nestswere found on 16 islands(20.2%) whereas45 islandsshowedno evi-

denceof formernests.Thosenestingplaceson islandsor on the mainlandin
whichnestswerelocatedin two or moreconsecutive
yearsare referredto as

traditional
nesting
areas
in thispaper.Fewnests
werefoundatponds
occupied
by pairsof HerringGulls(Larusargentatt•s)
andonlytwo nestswerediscovered
in lakesfrequented
byArcticLoons.Oldsquaws
didnotnestatponds


1975

ALISON: OLDSQUAW BIOLOGY
TABLE

13

3

CONCEALMENTRANKING FOR OLDSQUAWNESTS IN THE STUDY AREA*
Number
Active

Concealment

Well concealed
Open from above
Partially concealed
Poorly concealed

Islands

10(17.9)
46(82.1)
0
0

Nests

Mainland

16(41.0)
15(38.5)
6(15.4)

2 (5.1)

of Nests
Previous

Islands

Nests

Mainland

18(17.2) 27(24.3)
110(82.8) 61(54.9)
0
19( 17.1)
0
4 (3.7)

* Data from nests of previous years is also present for comparison. Figures in parenthesesare

percentages.

or lakesat whichCommonEider colonieswere found. In additionbreeding
pairsalwaysavoidedcertainpondsfor no apparentreason.Thirty-sixof 50
islandnestswerelocatedon largeislands(over9.8 x 10-2 ha). The maximum
numberof activenestson any islandwasseven.Three activenestswere found
on one 0.7 sq. m. islandin 1970. In 1970 and 1971, a singleactive nest was
located on the same small (28 cm x 41 cm) island.

Ekblaw (in Bent 1925) notedeachnestwas a smallcuplikedepression

linedwith dry plantsor leavesand coveredwith varyingquantitiesof down.
On the studyarea,nestswerealwayslinedwith driedleavesof dwarfbirch or
dwarf willow to a depthof about2.5 to 5.0 cm. The first egg was always
buriedin thisplantmaterial. In eachinstance,it wasnot until afterthe second
egg had been depositedthat down appearedon any nest. In 43 (45.3%)
instances,
no downwas found until after the third egghad been laid. Subsequentto the depositionof the third egg, increasingquantitiesof down were
observedon eachnestand after the fifth egg,the downlayer appearedcomplete. In no instancewas the nest lined with down. Occasionallythe down
layer was coveredwith or mixed with plant debris,especiallyon mainland
tundra

nests.

Twenty-six (27.3%) of 95 nestswere well concealedfrom above (Table 3)
by BlackSpruce(Piceamariana)boughsandfrom the sidesby variousgrasses,
dwarfwillows(Fig. 6) or dwarfbirches.Sixty-one(64.1%) nestswere open
from abovebut wereconcealedfrom all sidesby ! 0 to ! 5 cm of plant growth.
Only 6 (6.3%) nestswere partially concealedfrom the sidesand open from
above while 2 (2.3%) were poorly concealed. Mainland nests were more
completelyconcealed(z • 2.14, P •< 0.05) than nestslocated on islandsin
freshwaterponds or lakes.
Oldsquaw
nestsarereportedly
generally
foundcloseto water(Phillips1925,
Bent 1925). However,Bent (1925) noted a nest 7 m from the water'sedge
andEvans(1970) observed
nestsat least200 m from the nearestopenwater.

14

ORNITHOLOGICAL

Figure 6.

MONOGRAPHS

NO.

18

Oldsquaw nest cup located beneath concea•ng dwarf willow.

In this studyaveragedistancesbetweennestson islandsand the nearestopen
water were minimaland were determinedmostlyby the limiting size of each
island. Average distancesand correspondingstandarddeviationsbetween
nestson islandsand the nearestopen water were 0.8 -+ 3.4 m (n: 56), but
mean distanceswere greater in tundra and marshland areas 10.4 _+9.5 m
(n: 12), and 9.5 -+ 3.4 m (n: 9). Two nestson dry uplandwere 143 m
and 52.5 m, respectively,
[rom the nearestpond. Evans (1970) foundaverage
distances from the nearest water for nests on islands to be about 2.1 m whereas

nestson mainlandtundra and on mainlandbeachaveragedabout 28.6 m and
9.0 m, respectively. Table 4 showsthat about 65% of all active nestswere
within 1.4 m o[ openwaterwhereasonly 10% weremore than 14 m from the
edge of a pond or lake.
Severalspeciesof waterfowl,includingthe Mallard, Pintail, Gadwall (Anas
strepera), LesserScaup (Aythya aliihis) and white-wingedScoter (Melanitta

deglandi) are knownto nestin high dcnsitiesamonglarids (Vermeer 1958).
Evans (1970) presentedevidencesuggesting
that someOldsquawsat Churchill
nest in associationwith Arctic Terns (Sterna paradisaea). It has been suggestedthat speciesof waterfowlnestingamongcertainlarids may deriveprotection from the well-organizeddefensetactics apparently effective against
somepredators. Fifty-nine Oldsquawnestswere found in associationwith at
leastone Arctic Tern nest (at the samelake or pond). Of these,44 (46.39)


1975

ALISON:

15

OLDSQUAW BIOLOGY
TABLE

4

DISTANCESBETWEEN OLDSQUAWNESTS AND THE NEARESTOPEN WATER
Number

in meters

of

Active Nests

Distance

Percentages

1968

1969 1970

1971

Cumulative Percentages

1968

1969

1970

1971

1968

1969

1970

1971

0.0-

0.7

11

11

12

15

57.1

52.8

50.0

51.7

57.1

52.8

50.0

51.7

0.8-

1.4

3

2

2

5

14.3

9.4

8.3

17.2

71.4

62.2

58.3

68.9

1.5-

2.1

2

2

0

3

9.4

9.4

0

10.3

80.8

71.6

58.3

79.2

2.2- 3.5

1

2

4

1

4.8

9.4

16.6

3.5

85.6

81.0

74.9

82.7

3.6-

0

1

1

1

0

4.8

4.2

3.5

85.6

85.8

79.1

86.2

7.1-14.0

1

0

3

1

4.8

0

12.5

3.5

90.4

85.8

91.6

89.7

14.1-28.0

1

1

1

1

4.8

4.8

4.2

3.5

95.2

90.6

96.8

93.2

28.1+

1

2

1

2

4.8

9.4

4.2

6.8

Total

7.0

20

21

24

100.0 100.0 100.0 100.0

29

werelocatedon islands(Table 5) whereas15 (15.7%) wereon the mainland.
But 12 Oldsquawnests(12.7%) on islandsand 24 (25.3%) mainlandnests
werelocatedat lakesor pondsnot inhabitedby terns.On islandsin freshwater
pondstherewasa significant
association
betweenOldsquawnestsandArctic
Ternnests(x2 = 12.6,P •<0.01). On themainlanda significant
disassociation
of the nestsof thetwospecies
wasobserved
(x2 = 12.6,P •<0.01). In 1968,
1970and1971,Oldsquaw
nestswereinvariablyinitiatedpriorto the arrivalof
ternson the breedinggrounds.In 1969, 12 femaleOldsquaws
failedto nest
until afterseveralpairsof ternshad arrivedin the studyarea.
TABLE

5

NEST ASSOCIATIONBETWEEN OLDSQUAWSAND ARCTIC TERNS IN THE
STUDY AREA*
Number of Active Oldsquaw Nests
On Islands

On Mainland

Terns

Terns not

Terns

Terns not

Year

Resident

Resident

Resident

Resident

1968

9

2

4

1969

9

2

4

6

1970

13

3

3

5

1971

13

5

4

7

Total

44

12

15

24

* Only nests of known history are included.

6


16

ORNITHOLOGICAL

TaE

MONOGRAPHS

NO.

18

EGGS

Egg laying.--The nestsitewas alwaysselectedby the female. The male
invariably

followedthefemaleto theshoreof thepondor lakebut nonewas
observed to follow her as she walked about on islands or on the mainland

searchingfor a desirablelocation. I found no evidencethat femalesselected

andprepared
a nestsitepriorto layingthefirstegg.Femalespresumably
about
to lay thefirsteggwerefrequenfiy
observed
flyingfrompondto pondapparentlysearching
eachfor a suitablenestsite. One femalerequiredover an hour
to walkaboutthesurface
of a 0.39hapeninsula,
apparently
checking
undereach
Black Sprucefor a possiblesite. Seeminglyfindingnone, the bird then flew
immediately
to a nearbypond,climbedupontoa smallisland,anddisappeared
beneatha shortsprucetree. Forty-eight
minuteslatershereappeared,
entered
thewater,swamabout10 m awayfrom the island,andthenflew to a nearby
pondto join her mate. Two hourspriorto thisincident,I hadvisitedthe same
islandandfoundonlyonenestcup (from a previous
year). At thattimeit had
containeda few wet leavesand severalegg shells. When I returnedto the
islandsubsequent
to the female'sdeparture,the nestcup containeda single

eggburiedbeneathabouttwo inchesof dry leaves.The wet leavesand egg
shellshad apparentlybeenremovedby the female. The followingyear I observedthe sameseriesof eventsinvolvinganotherfemale. In neitherinstance
hadthefemalesbredat thatparticularnestsitein the previousbreedingseason.
In two instancesfemales,having laid the first egg, depositedthe second
(and the subsequent
clutch) at anothersiteon the samesmallisland. Usually
oneeggwasdeposited
perday. The firsteggwasmostoftenlaidabout1630,the
secondeggabout1100 the followingday, the third eggabout 1300 C.S.T. the
third day andeachsubsequent
eggat intervalsof about26 hoursuntil the clutch
wascompleted.Therewereno instances
of interruptedegglayingsuchasthose
mentionedby Cooch (1965) in Common Eiders. Incubation in Oldsquaws
alwaysbegansubsequent
to clutchcompletion. During the laying period femalesdid not returnto their nests,exceptto depositeggs,evenif the temperature fell below 0øC. After the clutch was completefemales generallyleft
their nestsunattendedonly twicedaily to feed,preen,or undertakesomeother
maintenanceactivity. Most females were absent from nests between 0900
and 1000 and again between 1400 and 1630 C.S.T.
Clutchsize.-•In this speciesclutchsizehasbeenvariouslyrecordedas from
5 or 6 to 9 or 10 (Millais 1913), 5 to 7 (Bent 1925) and 5 to 8 (Phillips
1925). Salomonsen(1950) recordedthe averageclutch size of Oldsquaws
in Greenland as 5.7 eggs. Two nestscontaining 11 eggs (Salomonsen1950)
and 17 eggs(Phillips1925) havebeenreported,althoughthesemay havebeen
the combined clutches of two females.

Clutchsizesof nestslocatedon islandsdid not differsignificantly
(F = 2.00,
df = 7/87. P •< 0.1) from those of mainland nests (Table 6). In five in-

1975

17

ALISON: OLDSQUAW BIOLOGY
TABLE

6

CLUTCHSIZESOF OLDSQUAWS
IN THE STUDYAREA, 1968
Mean

Clutch

TO 1971'

Size

Nest

Location

Islands

1968

1969

1970

1971

in

Freshwater

6.8 ñ 0.6

6.9 ñ 0.7

6.7 ñ 0.5

6.8 ñ 0.4

6.8 ñ 0.5

6.8 ñ 0.4

7.0 ñ 0.8

6.8 ñ 0.8

Ponds

Mainland

* Figures are mean clutch sizes and standard deviations.

stancesnestsof bandedfemaleswerelocatedin two consecutive
years. The
sampleis small;nonetheless,
four out of five femaleshad clutchescomprising
thesamenumberof eggs(7, 7, 7, and6 eggs,respectively)
in twoconsecutive
years.The fifth individualalsolaidthesamenumberof eggsin bothyears,but
in thesecond
yearthefirsteggwasabandoned
andtheremainderof the clutch
waslaid at a nearbynestsite. In two of theseinstances
the nestwaslocatedon
an islandin the firstyear andon the mainlandin the secondyear.
In waterfowlclutchandeggsizesareusuallyinverselyrelated(Lack 1966a).
In certainspecies
of waterfowl(i.e. Arctic species)largereggsprobablyaid
in thermoregulation,
providenewlyhatchedyoungwith largefoodreserves,
andapparently
offsettheadvantage
of smalleggsize.Theclutchsizein a given
species
mustevolvesothata particular
paircanmaximize
itsgenetic
impacton
subsequent
generations.
This mayresult,especially
in precocialspecies,

in a
clutchsizeequivalentto thelargestnumberof eggsthat a femalecanefficiently
incubate,
or thelargestnumberof eggswhichshecanproduce
in a shortperiod
of time. In thestudyareaOldsquaw
clutchsizewassignificantly
larger(6.8)
thanthatreportedfor the samespecies
in previousstudies(Salomonsen
1950,
and others) conductedelsewhere.This apparentlyinconsistent
clutch size
cannothoweverbe examineduntil a detailedcomparativestudyinvolving
severalbreedinglocalitieshas been conducted.
Viability of eggs.-•Non-viableeggsincludedthosewhichfailed to hatch in

successful
nestsdueto infertilityor deathof the embryo. In this study75 of
383 eggs(19.3%) were non-viable.The rate of non-viabilitywas constant
from year to year. In someother speciesof waterfowlegg lossdue to nonviabilityaverages7 to 9% (Earl 1950, Anderson1956 and 1957, Steelet al.
1956, Keith 1961). Perhapsthe low viabilityof Oldsquaweggsis relatedto
the extendedexposureof incompleteclutchesto severeweather.
THE

NEST

AND INCUBATION

Nestinitiation.--In this studyonly nestsof knownhistorywereincludedin

the data (i.e. exactdateof initiationof eachclutchwasknown). Oldsquaws


18

ORNITHOLOGICAL

MONOGRAPHS

NO.

18

tendedto breedat traditionallocations.Consequently,
by surveying
eachof
thesesitesdailyit waspossible
to discover
the deposition
dateof the first and
eachsubsequent
eggin 95 clutches.In thestudyperiodtheearliestactivenest
wasfoundon 3 June and the latestnest (not a renest) on 28 June. The mean
datesof clutchinitiationwere 12 June,23 June, 11 June and 9 Junein 1968
through1971, respectively.In 1969, the observeddelayin nestinitiationwas

apparentlydueto ice conditions
on the breedingponds.Egglayinggenerally
extendedover a period of about 15 to 20 days.
In 1971, whennestingoccurredearliest,the warmestMay and the coldest

Juneof the entirestudyperiodwererecorded.Whennestingoccurredlatest,
in 1969,bothMay andJunetemperatures
werebelownormal.Aboveaverage
temperatures
in bothMay andJuneof 1970did notsignificantly
advance
nest
initiation.Junesnowfallapparently
did not affectnesting,
probablybecause
no accumulationresulted. The effect of heavy rainfall upon Oldsquawnest
initiationisunknown.Keith (1961) foundno correlationbetweenprecipitation

andmeandateof clutchinitiationin severalspecies
of prairiewaterfowl.
The average
dateof clutchinitiationin Oldsquaws
wassignificantly
delayed

onlybycoldMaytemperatures
(b = -1.23, t = 4.69,df = 2, P • 0.05). Thus,
in yearswithbelowaverage
May temperatures,
snowandicedisappeared
from
thestudyarealaterthanin yearswhenMay temperatures
wereaverage
or above

average.Nestingwasconsequently
delayedbecause
potentialnestinglocations
werecoveredby snow.Furthermore
otherwise
isolatedislandswerestillconnected
to themainland
byicebridges.Ryder(1969) foundthatmostnestsof
Ross'Goose(Anser rossii) and "Lesser"SnowGooselocatedon offshore
islands
thatwerejoinedto themainland
byicebridges
weredestroyed
byArctic
Foxes (Alopex lagopus). While remnantice remainedin breedingponds,
Oldsquaw
nestsmightbe similarlyavailableto mainlandpredators(Alison
1970).

Effectof weatheruponincubation.--Lowtemperatures,
snow,andmoderate
rainfalldo not seemto affectegglossthroughdesertionin this species.The
meandatesfor clutchinitiationfrom 1968 through1971 were previouslygiven
as 12, 12, 11 and9 June,respectively.
In 1968, from 8 to 16 June,the mean

dailytemperature
was5.3øC,overnight
lowsaveraged
-0.6øCand20 cmof

snowfell. In 1969, from 19 to 27 June,dailytemperatures
andovernightlows
averaged
4.4øCand1.1øC,respectively.
In 1970,from7 to 15 June,themean
dailytemperature
was6.3øCandtheaverage
overnight
lowswereabout0.1øC.
On 10 June1971, 18.6 mm of rain fell and on 17 to 19 Juneof the sameyear,
19.6 mm of rain wererecorded.Oldsquaws
wereneverobservedto desertnests
undertheseconditions.No nestsof this specieswere lost due to flooding.
Predation on nests.•Most recent waterfowl studieshave shown that predationwas the main causeof nestloss. In this study also a larger proportion

of unsuccessful
nestsresultedfrom predationthan from any other cause.


1975

ALISON: OLDSQUAW BIOLOGY
TABLE

AVIAN

AND MAMMALIAN

19

7

PREDATORS AND NEST DESTRUCTION IN

THE

STUDY AREA
Potential
Nest
Predator

Number
1968

1969

of Predators
1970

1971

Parasitic Jaeger

2

2

4

4

Common

Crow

2

2

2

2

Common

Raven

1

1

I

1

16
2

14
2

14
0

16
3

Herring Gull
Long-tailed Jaeger
Arctic

Fox

2

3

2

3

0

1

0

2

Total

25

25

23

31

% of Nests Destroyed
by Avian Predators

16.7

16.7

20.0

34.6

% of Nests Destroyed
by Mammalian Predators

11. I

11.1

5.0

3.9

Total % Destroyed

27.8

27.8

25.0

38.5

Red Fox

Thepotentialaviannestpredators
in thestudyareaincluded
Parasitic
Jaeger
(Stercorariusparasiticus),Long-tailedJaeger (S. longicaudus),Common
Raven (Corvuscorax), CommonCrow (C. brachyrhynchos)
and Herring
Gull. Of theseonlythe ParasiticJaegerwasa knownpredatorof Oldsquaw
nests.In 1968 and 1969, ParasiticJaegersaccounted
for all avian-destroyed
nests.AlthoughLong-tailedJaegers,CommonRavens,CommonCrowsand
HerringGullswerenotobserved
destroying
Oldsquaw
nestsin thisstudy,they
havereportedly
takenaheavytollof waterfowl

eggselsewhere
(Kortright1953,
Sutton 1932, Lokemoen1966, Larson 1960).
The ArcticFox andRedFox (Vulpes[ulva) werethe onlytwo species
of
mammalian
predators
observed
in thestudyarea. Neitherof thesewasactually
observeddestroying
nests,but severalOldsquawnestswere apparentlydestroyed
by foxes(a characteristic
fox odorpersisted
at eachdestroyed
nest).
Predatordensitywas determinedby directcount. Each numberin Table 7
representsan individualpredatorwhichwas observedat the samelocationin
at least10 differentcensuses.
Thusthe data probablyrepresenta minimum
numberof predatorsas otherindividualsthat infrequentlyenteredthe study
areafrom otherareaswerenot included.Furthermore,in eachyear the nests
of all recordedParasiticJaegers,Herring Gulls, and CommonCrowswere
located.At leastoneactivefox denwasfoundin eachyear (two in 1971).
In 1971,predatordensity
wasgreaterthanin anyotheryearof thestudyanda