nest morphology and body size of ross' geese and lesser snow geese
TheAuk 114(4):610-618, 1997
NEST MORPHOLOGY
AND BODY SIZE OF ROSS' GEESE AND LESSER SNOW GEESE
KEVIN G. MCCRACKENaTMALAN D. AFTON,1 AND RAY T. ALISAUSKAS2'3 U.S.Geological Survey,Biological Resources Division,LouisianaCooperative FishandWildlifeResearch Unit, LouisianaStateUniversity,BatonRouge,Louisiana70803,USA; 2PrairieandNorthernWildlifeResearch Centre,CanadianWildlifeService, 115 Perimeter Road,Saskatoon, SaskatchewanS7N OX4, Canada;and
3Department ofBiology, University ofSaskatchewan, Saskatoon, Saskatchewan S7NOWO,Canada ABSTRACT.--Arctic-nesting geesebuild large, insulatedneststo protectdevelopingembryosfrom cold ambienttemperatures. Ross'Geese(Chenrossii)are abouttwo-thirdsthe massof LesserSnowGeese(C.caerulescens caerulescens), havehighermass-specific metabolic rate,andmaintainlowernestattentiveness, yet theyhatchgoslings with morefunctionally maturegizzardsandmoreproteinfor theirsizethando LesserSnowGeese.Wecompared nestsize(a reflectionof nestinsulation)in fourdistincthabitatsin a mixedbreedingcolony of Ross' Geese and Lesser Snow Geese at Karrak Lake, Northwest Territories, Canada. After
adjustingmeasurements for nest-specific eggsizeand clutchsize,we foundthatoverallnest
morphology differedbetweenspecies andamonghabitats.Nestsizeincreased progressively amongheath,rock, mixed,and mosshabitats.When nestingmaterialswere not limiting, nestsweresmallerin habitatsthatprovidedcoverfromwind andprecipitation thanin habitatsthat did not providecoverRoss'Geeseconstructed relativelylarger,moreinsulated neststhan did LesserSnowGeese,whichmay hastenembryonicdevelopment, minimize energyexpenditureduringincubation,andminimizeembryoniccoolingduringrecesses. We suggestthat relativedifferences in nestmorphologyreflectgreaterselectionfor Ross'Geese
to improvenestinsulationbecause of theirsmallersize(adultsandembryos), highermassspecificmetabolicrate, and lower incubationconstancy.Received 13 May 1996,accepted 18 March 1997.
IN BIRDS, HEATISSUPPLIED to developingembryosprimarilyby bodywarmthof thebrooding parent (Afton and Paulus1992).Heat loss from developingembryosand metaboliccosts to parentscanbe minimizedby increasingnest insulation.Nestinsulationaffectsgrowthrates of nestlings(Winkler 1993),and probablyaffectsembryonicgrowthaswell,particularlyin precocialspecies.Nest morphologyand insulative propertiesof nestsprobablyare influencedby body size,metabolicrates,incubation constancy, ambientthermalconditions,availabilityof nestingmaterials,and risk of predation (Moller 1984).Mass-specific metabolicrate and heat-transfer rate generallyincreasewith decliningbody mass(Brody1945,Lasiewski and Dawson 1967, Templeton1970, Hill and Wyse 1989). Consequently,smaller species mustassimilatenutrientsat greaterratesthan largerspeciesto maintainhighmetabolicrates. In addition,smallerspecieshavelesscapacity to storeendogenous nutrientsthan do larger 4 E-mail: [email protected]
610
species.Thus,body sizeis of profoundconsequenceto incubatingparents,whichmustfast duringincubation unlesstheyarefedby mates or auxiliaries (von Haartman 1958, Skutch
1962, Afton and Paulus1992). If endogenous nutrients are not sufficientto meet energy requirementsduringincubationsessions, parents may recessfrom incubationin order to feed,
therebydecreasingincubationconstancyand exposingneststo predatorsand heatloss.Afton and Paulus(1992)reportedthat incubation constancyis positively related to body mass amongall speciesof waterfowl(Anatidae),particularly amonggeeseand swans. Within species, nest-building behaviormay be adaptedto maximizeefficiencyof nest insulationin responseto cumulativeembryonic metabolicrates,whichare productsof clutch sizeand embryobodysize.Because of the relationshipbetweensurfacearea and volume, smalleggsand clutcheshave greaterratesof heattransferthanlargereggsand clutches, all elsebeingequal.Thesefactorsare particularly importantunderthewindyconditions thatpre-
October1997]
NestMorphology ofArcticGeese
611
by tundrameadowsand vail in arcticlatitudes(Thompson andRaveling The area is characterized exposedPrecambrianbedrock.Shallow lakes and
1988).
Nestmorphology andinsulationmaybe in- streams are numerous. Karrak Lake contains the fluencedfurtherby localvariationin tempera- largestcolonyofRoss'Geesein NorthAmerica(Kerbes
ture, wind, and availabilityof nestingmateri- 1994)andhasnearlyequalnumbersof LesserSnow Geese(Slatteryand Alisauskas1992,1993).In 1993, als. Comparativestudies of hummingbirds about364,000Ross'and LesserSnowgeesenestedat
(Oreotrochilus spp.;Pearson1953,CorleySmith KarrakLake(Slatteryet al. 1994). 1969), Village Weavers(Ploceus cucullatus; Col-
We examinednestson the largestisland near the
lias and Collias1971),and HelmetedHoney- centerof thelake.Wearbitrarilyselected19circular eaters (Lichenostomus melanops; Franklin 1995)
plots (20-m radius) every 100 m along a transect
demonstrate that birds nestingin cold areas whereRoss'andLesserSnowgeesenestedtogether. construct better insulated nests than those We then selectedeveryRoss'(n = 54) and Lesser clutch nestingunder warmer conditions.Ryder (1964, Snowgoose(n = 51) nestin theplots.Species, 1967, 1972) found that nests of Ross' Geese size,egg length,and eggbreadthwere recordedfor
all nests.Embryoagewasestimated(Weller1956)to calculateclutchinitiation date, assumingfor both tatsthatprovidedshelter. Experimental studies speciesa layingrate of 1.3 daysper eggand an incubationperiodof 23 days(Ryder1967,Ankneyand
(Chenrossii)werelargerin habitatsthatdid not provide shelterfrom wind than thosein habi-
of egg coolingratesin three otherspeciesof
Macinnes 1978). Nestswere consideredsuccessfulif
arctic-nestinggeese(Chencanagica, Brantaber- at leastoneegg hatched.Nest habitatwas classified nicla nigricans,Brantacanadensis minima)indi- as heath,rock, moss,or mixed accordingto domicatethat,acrossspecies, nestinsulationandin- nantfeaturesof theimmediatenesthabitat(seeRycubation constancy are inversely related der 1967, 1972; McLandress 1983). Nests in heath (Thompson andRaveling1988).Finally,intra- habitatswerebuilt directlyinto patchesof Labrador decumbens), white heather(Cassiope tetraspecificand interspecific competitionfor high- tea (Ledurn or willow qualitynestsitesand nestmaterialsalsomay gona),dwarf birch (Betulaglandulosa), (Salixspp.).Nestsin rockhabitatwerebuilt oneither influencenestmorphologyand insulation. We measured
nests of Ross' Geese and Lesser
SnowGeese(Chencaerulescens caerulescens) in a mixedbreedingcolonyin arcticCanada.Weas-
gravelor bedrocksubstrateandcomposed of gravel, soil, or twigs of dwarf birch and willow. Nestsin mosshabitatswere composedof both living and dead Sphagnum speciesas well as nonliving heath fragmentsand birch or willow branches.Nests in mixedhabitatpossessed any combination of char-
sumedthatnestmorphometrics aredirectlyrelated to nestinsulationaccordingto the heatflow equation: actersfound in heath, rock, and mosshabitats. = k. at. (ax)-',
(1)
We measured(_+1 cm) outer diameter,wall thick-
ness,circumference, rim height,bowl depth,and inwhereH represents heatflow,k represents ther- ner diameterof eachnest during initial nestvisits (Appendices1 and 2). After the eggshatched,nests mal conductivity,T representstemperature, wereweighedon a portableelectronicscale(_+1 g).
andx represents thickness of insulatingmate-
Masses of 10 Lesser Snow Goose nests were estimat-
rial (Serway1990).We predictedthat nestsize ed using stepwisemultiple regression(PROCREG;
wouldvary inverselywith body size and in- SAS 1990) becausenest markers were lost between cubationconstancyand would vary positively initial and final nest visits. Best fit was obtained with with mass-specific metabolic rate.Wealsopre- separateequationsfor eachhabitat: dictedthatnestsizeshouldbe greaterin habih = --599.1 + 4.9(circumference) tatsthataremoreexposed to wind andprecip- masshoar itation,whenadequatenestmaterialsareavail+ 14.4(outerdiameter) able(Ryder 1964,1967,1972). METHODS
We measured
(r2 = 0.74, P < 0.0001);
massrock = -2073 + 170.8(wallthickness) (r2 = 0.76, P < 0.0046); and
nests of Ross' Geese and Lesser
Snow Geese at Karrak Lake, Northwest Territories,
Canada(67ø15'N,100ø15'W)between19 Juneand 15 July 1994.Karrak Lake is locatedon the west tribu-
tary of the SimpsonRiver in the centralCanadian Arctic,southof theQueenMaudGulf (Ryder1972).
(2)
(3)
massm•xe d = -782.1 + 29.5(outer diameter)
(r2 = 0.33, P < 0.0001). Total
clutch volume
was calculated
(4) for each nest
612
MCCRACKEN, AFTON,AND ALISAUSKAS
from length(_+0.1 ram) and breadth(_+0.1 ram) of eacheggin a nestfollowingHoyt (1979;Kv= 0.51). To estimate
the relative
difference
in overall nest
[Auk, Vol. 114
(PROCCATMOD;SAS1990)to testwhethernestsuccess varied between speciesand among habitat types.
morphology betweenspecies andtoaccount forvariationdue to differencesin egg size and clutchsize, we dividednestmeasurements by the squarerootof the total clutchvolumefor eachnest,exceptfor nest
RESULTS
The averagedatethatfirst eggswerelaid was 3 June,and this did not differ betweenspecies analysisof covariance (ANCOVA)with clutchvolor among habitatsand plots (all Ps > 0.12). ume as a covariate,or using residual nest measurementscorrectedfor clutchvolume,would be inap- Clutch size was inverselyrelated to first-egg propriatebecauseof significantheterogeneity in date (F = 1.44, df = 1 and 101, P = 0.0001)but clutchvolumebetweenspecies (F = 45.26,df = 1 and did not differ betweenspeciesor vary in relamass,which was divided by clutchvolume.Using
date,or plot88, P < 0.0001,r2 = 0.34);i.e. significantinterspecific tion to plot, species-by-initiation colinearitywaspresentamongpredictorvariablesin by-species-by-initiation date (all Ps > 0.23). the model.We subsequently determinedthat the squarerootof clutchvolumeratherthanclutchvolume or the cuberoot of clutchvolumeis the appropriatedenominator for linearmeasurements. Using clutch volume and the cube root of clutch volume as
thedenominator did not removethespecies effecton inner diameter(P < 0.05), whereasdividing by the squarerootof clutchvolumerenderedspecies effects insignificant,as would be expectedif nestinner diametersmatchedthebody form of eachspecies. We usedmultivariateanalysisof variance(MANOVA) to test whether nest measurementsdiffered amongspeciesandhabitats(PROCGLM; SAS1990). We useda split-plotdesignto accountfor variance
Mean nest success was 92.6 --- SE of 2.8% and
did not differ between species(X2 = 0.06, n = 87, P = 0.805)or amonghabitats(X2 = 0.02, n = 87, P = 0.992).
Variationbetween species andamonghabitats.The overall nest morphologycorrectedfor clutchvolumedifferedbetweenspecies(MANOVA, F = 3.90, df = 7 and 78, P = 0.001) and
amonghabitats(F = 5.87,df = 14 and156,P = 0.0001).Ploteffects,andspecies-by-habitat and plot-by-species-by-habitat interactionswere not significant(all Ps > 0.10).The outer diameter, wall thickness, circumference, and rim
due to plot effect,whichpotentiallycontainscom- height of Ross' Goosenests were relatively ponentsofbothsamplingerrorandnaturalvariation due to density and nonrandom distributionsof geese.In the full model, speciesand habitatwere fixed effects,whereasplot was a randomeffect.Fvalues reported from MANOVA were determined usingWilks' lambda.Beginningwith thelargest-order interaction,nonsignificanteffectswereiterative-
larger than thoseof LesserSnowGoosenests (Table1). Nest-bowl depth and nest inner diameterdid not differ betweenspecies,probably becausethey are a functionof femalebody size,eggsize,andclutchsize(all of whichwere accountedfor by the correctionfactor).Despite
ly removedfrom the modeltmtil we obtainedthe differences in relative size of nests, nest mass most parsimonious modelcontainingonly signifi- did not differ between species,perhaps becant effects. Following a significantMANOVA, we
cause nests were not dried, and small amounts
usedanalysisof variance(ANOVA)to testwhether of soil, pebbles,and old nest fragmentscould
individual nest measurementsdiffered among significant effects.We comparedleast squaresmeans amonghabitattypeswith t-tests(PDIFFoption;SAS 1990).We did not find any Ross'Goosenestsin rock habitats;consequently, onlyheath,moss,and mixed habitat nests were included in models comparing
not be separatedfrom nests.Outer diameter, wall thickness,circumference, rim height,and nestmassincreasedprogressively acrossheath, mixed, and mosshabitats,i.e. from more protectedto lessprotectedhabitats.Bowl depth species andhabitats. Wesubsequently compared the and inner diameterdid not differ amonghabfour habitatsin a separatemodelfor LesserSnow itats (Table2). Geese. In this model, we did not divide nest mea-
surements by the squareroot of clutchvolumebut ratherincludedthe squareroot of clutchvolumeas a covariate because clutch volume and habitat were
not related (F = 0.96, df = 7 and 40, P = 0.47). We
Variation among habitats in Lesser Snow Geese.--The overallnestmorphologyof Lesser Snow Geesediffered amonghabitats(MANCOVA, F = 3.97, df = 21 and 118 P = 0.0001)
but was unrelated to clutchvolume (F = 1.02, usedANOVA to testwhetherfirst-eggdatediffered betweenspecies,and amonghabitatsand plots.We df = 7 and 40, P = 0.43). Outer diameter, wall rim height,and nest usedANCOVAto comparemeanclutchsizebetween thickness,circumference, mass generally increased amongheath,rock, species andamongplots,with initiationdateincluded as a covariate.We used categoricaldata analyses mixed, and mosshabitats,i.e. from more pro-
October 1997]
NestMorphology ofArcticGeese
613
TABLE1. Least-squares means(œ_+SE)of nestdimensions and nestmassadjustedfor clutchvolumefor Ross'Geeseand LesserSnowGeesenestingat Karrak Lake,1994. Ross' Goose
Measurementa
Lesser Snow Goose
(n = 54)
(n = 36)
pb
Outer diameter Wall thickness Circumference
0.0925 _+ 0.0024 0.0347 _+ 0.0012 0.0310 --- 0.0086
0.0838 --- 0.0031 0.0302 _+ 0.0016 0.0265 -+ 0.0110
0.031 0.028 0.002
Rim height Bowldepth
0.0127_+0.0006 0.0129_+0.0003
0.0105-+ 0.0008 0.0120-+0.0004
0.032 0.072
Inner diameter Nest mass
0.0287 0.0030
0.0281 -+ 0.0006 0.0025 +- 0.0003
0.501 0.200
_+ 0.0005 -+ 0.0002
Measurements dividedby thesquarerootof clutchvolumeexceptfornestmass,whichwasdividedby clutchvolume. ANOVA for specieseffect(df = 1 and 84 for eachtest).
tectedto lessprotectedhabitats(Table3). Nest bowls were shallowest
Like most waterfowl, Ross' and Lesser Snow
in rock habitats and dif-
geese line their nests with considerable fered from those in heath, mixed, and moss amountsof down and covertheir eggsbefore habitats.As expected,the inner diameterof recessingfrom incubation.Down providesexLesserSnowGoosenestsdid not differ among cellent insulation, and the amount of down in habitats. a nest may vary betweenspeciesand among habitats.We did not quantifythe amountof down or the specificcompositionof materials DISCUSSION in eachnest,but downmaybeparticularlyimThompsonand Raveling(1987) suggested portantfor geesenestingin habitatswhere that predationinteractswith body size in de- nestingmaterialsare scarce.Differenttypesof termining incubation constancy in geese, nestingmaterials(e.g. heath and moss)may whichin turnmayberelatedtonestinsulation. differ in their insulativeproperties.Although In areaswith highratesof predation,theuseof we did not identify nestmaterialsspecifically, feathersin nestsfor insulation(Mailer 1984) nest materials are related to the distribution of may be counteractedby selectionfor incon- materialsin a givenhabitattype,i.e.Ross'and spicuousnestsif feathersare not cryptic.We LesserSnow geesedo not use differentmatebelievethat concealment of eggsand nestsof rialswithin a habitat(pers.obs.).ConsequentRoss' Geese and Lesser Snow Geese at Karrak ly, inclusionof thehabitateffectin ouranalyses Lakeisnotinfluenced by strongselective forces should account for variation in the insulative of predation.Ross'and LesserSnowgeeseare propertiesof differentnest materials. non-cryptic,colonial and nest in extremely Nestmorphology andhabitatvariation.--Ryder densenumbers.Our claimis furthersupported (1964,1967,1972)foundthat Ross'Goosenests by high nestingsuccess (Slatteryand Alisaus- were largestin mosshabitats,intermediatein mixed habitats, and smallest in rock habitats.
kas 1992, 1993).
TABLE 2. Least-squares means(œ_+SE)of nestdimensions and nestmassadjustedfor clutchvolumefor Ross'Gooseand LesserSnowGoosenestscombinedby habitatat KarrakLake,1994. Nestinghabitat Measurement• Outer
Heath (n = 18)
Mixed (n = 43)
diameter
0.0664 _+ 0.0041 ^
0.0948
Wall thickness
0.0218 _+ 0.0021 ^
Circumference
0.2252 _+ 0.0145 ^
Rim height Bowl depth Inner diameter Nest mass
-+ 0.0025 B
Moss (n = 29)
pb
0.1032 _+ 0.0034 •
0.0001
0.0356 -+ 0.0013 B
0.0399 _+ 0.0017 B
0.0001
0.3006 -+ 0.0092 •
0.3376 _+ 0.0121 c
0.0001
0.0086___ 0.0011^ 0.0121_+0.0005
0.0130-+ 0.0007B 0.0124-+ 0.0003
0.0133_+0.0009B 0.0127ñ 0.0004
0.001 0.727
0.0287 -+ 0.0008 0.0019 _+ 0.0004 ^
0.0276 -+ 0.0005 0.0025 -+ 0.0002 ^
0.0289 _+ 0.0007 0.0038 -+ 0.0004 B
0.274 0.002
• Measurements dividedby the squarerootof clutchvolumeexceptfor nestmass,whichwasdividedby clutchvolume. bANOVA for habitateffect(df = 2 and 84 for eachtest).Within rows,valueswith differentsuperscripts aresignificantlydifferent(P < 0.05) basedon t-testsfollowinga significanteffectof habitat.
614
MCCRACKEN, AFTON,ANDALISAUSKAS
[Auk,Vol. 114
TABLE3. Least-squares means(• _+SE)of nestdimensionsand nestmassfor LesserSnowGoosenestsby habitat at Karrak Lake, 1994.
Nestinghabitat Measurement a Outer diameter Wall thickness Circumference
Rim height Bowldepth Inner diameter
Nest mass
Heath (n = 7)
Rock (n = 15)
38.7 +- 2.9 ^ 12.1 + 1.5 ^ 121.0 _+ 11.2 ^
49.7 +- 2.0 B 17.5 -+ 1.0 B 164.3 q- 7.6 B
4.7 + 1.0^ 7.4 _+0.4^ 18.1 -+ 0.7
563.7 + 177.0 ^
5.7 -+0.7^.B 6.2 +_0.3B 18.0 _+ 0.5
910.7 +_ 120.9 ^
Mixed (n = 20) 58.5 + 1.7 c 21.6 q- 0.9 c 177.2 + 6.6 B
7.3 + 0.6B 7.6 + 0.3^ 17.6 -+ 0.4
921.7 + 104.7 ^
Moss (n = 9) 60.8 + 2.6 c 23.1 q- 1.3 c 201.2 +_ 9.8 c
7.8 -+ 0.9B 7.7 + 0.4^ 17.6 q- 0.6
1,379 + 156.1B
Pa 0.0001 0.0001 0.0001
0.043 0.003 0.827
0.011
• ANOVA for habitateffect(df = 3 and 47 for eachtest).Within rows,valueswith differentsuperscripts are significantlydifferent(P < 0.05) basedon t-testsfollowinga significanteffectof habitat.
He suggestedthat thesedifferencesresulted Competition.--Competition for high-quality from differencesin availabilityof nestmaterial nest sites and nest materials may influence and exposureto weather.Our data for Ross' nest-site selection such that access to the best and LesserSnow geeseclearly are consistent sitesvariesamonggeeseat boththemacrohabwith Ryder'sobservations. Wealsoconcurwith itat andmicrohabitat levels.This,in turn,may his interpretationof the adaptivesignificance influence nest size and nest success. In rock and of variationin nest size amonghabitats.Nest other habitatswhere nest materialsare limited, materials are abundant in heath, mixed, and all availablenest materials are incorporated mosshabitats(Ryder1967,pers.obs.),but rock into nests(Afton unpubl. data). Moreover,we habitatstypicallypossess littlenestingmaterial haveobservedgeesefighting(within and beotherthansmalltwigsandgravel.Nestsbuilt tween species)over nest materialsduring the in heathtypicallyare shelteredby low (1 to 6 prelayingand earlylayingperiods.In light of cm) vegetationthat could decreaseconvective exponentialpopulation growth and habitat heat loss from nests. Nests built in mixed habdegradationat Karrak Lake and elsewhere itats often are constructedof vegetationor in (Slatteryet al. 1994,Batt1997),somegeesemay rocksthatmayprovidesheltering effects.Nests be nestingin poorerhabitatwith the net effect built in openmossappearto be the leastshel- that globalnest success may have declined tered.Althoughnestsin rocktypicallyoccurin from previousyearswhen nestdensitieswere unvegetatedsiteswhere exposureis greatest, lower Macrohabitat and microhabitat selection.--Macwe suspectthatplacement of nestswithincrevicesor in theleeof largerockscandramatically rohabitatselectionprobablyis a functionof decreaseexposureand convectiveheat loss. habitatavailabilityduring nestinitiationand The outer nest dimensions,includingouter exposureto weathercausedby habitattopogdiameter, wall thickness, circumference, and raphy. Ryder (1967, 1972) found that Ross' rim height,providethe bestestimatesof the Goosenestdensitieswerehighestin mixedand size of the nest wall. These measurements heathhabitatsand lowestin the moreexposed should be a direct reflection of nest insulation rock and moss habitats.McLandress(1983) accordingto the heat-flowequation.These demonstrated that Lesser Snow Geese show a measurementsdiffered significantly among markedpreference for rockhabitats.This may habitatsand generallyincreasedacrossheath, be becauserockhabitatsareusuallythe firstto rock, mixed, and mosshabitats. Inner nest mea- clear of snow and dry out each spring, and surements also should influence nest insulaLesserSnow Geese typically initiate nesting tion,butshouldconformmorecloselyto female earlier than Ross'Geeseat Karrak Lake (Slatbody shape,egg size,and clutchsize than to tery and Alisauskas1993). We found no Ross' habitat.Inner diameterdid not differ among Goosenestsin rock habitatsor on the upperhabitats.Bowldepthdid not vary amonghab- most parts of eskers,althoughLesserSnow itatsexceptin rock,whereit was smallest;the Geesewere commonlyfound nestingin such impenetrabilityof rock and gravel substrates locations. We arguethatRoss'Geesedo not ocprobablyexplainsthis finding. cupy such habitatsbecauseof their smaller
October1997]
NestMorphology ofArcticGeese
615
body size,highermass-specific metabolicrate, lower incubation constancyrelative to Lesser lower incubationconstancy,and smaller egg Snow Geese. andclutchsize.Furthermore, Ross'GeeseprobBy constructing relativelylargerand probaably incur energeticadvantagesby initiating bly better-insulatednests,female Ross'Geese nestsa few daysafter snowhas clearedfrom can minimize energyexpenditureduring inthe more sheltered heath and mixed habitats. cubationand embryoniccoolingduring recesBecause of theirlargerbodysize,LesserSnow ses.Slatteryand Alisauskas(1995)found that Geeseprobablyare better able to toleratecon- neonate Ross' Geese hatch with more functionditionsin themoreexposedrockhabitats.Con- ally maturegizzardsandmoreproteinfor their sequently, theytendto nestin rockhabitatsout size than did Lesser Snow Geese. In addition, Ross'Goosebroodsdispersefartherfrom nestof proportionto its availability. Microhabitatnest-siteselectionprobablyoc- ing areasthan do LesserSnowGoosebroods cursaftergeesehavesetup territories.Within (Slattery1994).Moreover,Ross'Goosegoslings a territory, there is considerablevariation fledge two weeks earlier than LesserSnow amongpossiblenest sites,and femalessome- Goose goslings(Owen 1980). The relatively timesexhibitnest-buildingbehaviorin several larger size and greater insulation of Ross' locations withintheterritorypriorto layingthe Goosenests probably buffer developingemfirst egg (Afton unpubl.data). A small dwarf bryos against fluctuationsin ambient temperdevelopment, andulbirch,patchof Labradortea, or pile of small atures,hastenembryonic boulders,lessthan a meter in diameter,may timatelypromotea greaterpropensityfor acyield a nestsite that providesexcellentprotec- tivity during the brood-rearingperiod (Slattion from wind, whereasan adjacentarea of tery and Alisauskas1995). In light of the physiological differences bebaregroundmayofferlittle protectionfromthe tween Ross' Geese and Lesser Snow Geese, we wind. Small-scale factorssuchasproximityto are adjacentterritory holdersmay influencenest- suggestthat overall fitnessconsequences site selection further
associated with nest construction and selection
Nest morphology and variationbetweenspe- of nestinghabitatin arctic-nestinggeese.We
hopethatthispaperwill inspiremorecomprehensivestudiesof fitnessconsequences resulting from nest morphology and site selection. specificbasalmetabolicrate that is 1.18kcal ß day-• ß g • higher than that of LesserSnow Experimentsinvolving manipulation of nest Geese (Aschoffand Pohl 1970, Alisauskasand materials (particularly down) coupled with of incubationtemperaturesand Ankney 1992). FemaleRoss'and LesserSnow measurements hatchling body condition wouldbeparticularly geese exhibit high incubation constancy cies.--Ross' Geese are about two-thirds the mass of Lesser Snow Geese and exhibit a mass-
throughoutthe 23-dayincubationperiodand rely heavily upon endogenous nutrientsfor both egg laying and incubation(Ryder 1967,
informative.
ACKNOWLEDGMENTS
1970, Ankney and Macinnes 1978, Afton and We thankI. J.Ball,B. Batt,G. Gauthier,W. Koenig, Paulus 1992, Alisauskas and Ankney 1992). T. Martin, M. R. McLandress,J.Ryder,B. Wilson,and
However,Ross'Geesehaveslightlylowerincubationconstancythan LesserSnowGeese (Krechemarand Syroechkovsky 1978, Afton and Paulus 1992, LeSchacket al. 1998) and spendmore time foragingduring incubation (Aftonunpubl.data).Whencorrected forclutch volume(whichis a functionof body size,egg size, and clutchsize), nestsof Ross'Geeseare largerthannestsof LesserSnowGeese.These resultsprobablyreflectgreaterselectionfor ef-
two anonymous reviewersfor helpful commentson the manuscript,and C. LeSchack,D. McCracken,N. McCracken,J.Ryder,S. Slattery,D. Stern,V. Wright, and all thosewho spentthe summerof 1994at Karrak Lake for ideas and assistance. This research was
supported by theCaliforniaDepartmentof Fishand Game; the Arctic GooseJointVenture; the Canadian
Wildlife Service,PolarContinentalShelfProject;the Schoolof ForestryWildlife and Fisheries,Collegeof Agriculture, and Agricultural Center of Louisiana StateUniversity;theLouisianaBoardof Regents; the
ficient nest insulationby Ross'Geesebecause LouisianaDepartmentof WildlifeandFisheries; and of their smallerbody size (adultsand embry- the LouisianaCooperativeFish and Wildlife Reos), greater mass-specific metabolicrate, and search Unit.
616
MCCRACKEN, AFTON,ANDALISAUSKAS LITERATURE CITED
[Auk, Vol. 114
itat selectionand nest spacingin a colonyof Ross'and LesserSnowgeese.Auk 100:335-343.
AFTON, A.D., AND $. L. PAULUS. 1992. Incubation
and brood care. Pages62-108 in Ecologyand managementof breedingwaterfowl (B. D. J.Batt, A.D. Afton, M. G. Anderson,C. D. Ankney, D.
H. Johnson, J.A. Kadlec,andG. L. Krapu,Eds.). University of MinnesotaPress,Minneapolis. ALISAUSKAS, g. t., AND C. D. ANKNEY. 1992. The cost
of egglayingand its relationshipto nutrientreservesin waterfowl. Pages30-61 in Ecologyand managementof breedingwaterfowl(B. D. J.Batt, A. D. Afton, M. G. Anderson,C. D. Ankney,D.
MOLLER,A. P. 1984. On the use of feathers in birds nests: Predictions
and tests. Ornis
Scandinavica
15:38-42.
OWEN,M. 1980.Wild geeseof theworld.B.T. Batesford, London.
PEARSON, O. P. 1953.Useof cavesbyhummingbirds and otherspeciesat high altitudesin Peru.Condor 55:17-20.
RYDER, J.P. 1964. A preliminarystudyof the breeding biology of Ross'Goose.Wildfowl 15:127137.
J.P. 1967. The breedingbiologyof Ross' H. Johnson, J.A. Kadlec,andG. L. Krapu,Eds.). RYDER, Goosein thePerryRiverregion,NorthwestTerUniversity of MinnesotaPress,Minneapolis. ritories.CanadianWildlife ServiceReport 3:1ANKNEY, C. D., AND C. D. MACINNES. 1978. Nutrient 56. reservesand reproductiveperformanceof female Lesser Snow Geese. Auk 95:459-471. RYDER, J.P. 1970. A possiblefactorin the evolution of clutch size in Ross' Goose. Wilson Bulletin 82: ASCHOFF, J., ANDH. POHL.1970. Rhythmicvaria4-13.
tionsin energymetabolism.FederationProceedRYDER, J.P. 1972.Biologyof nestingRoss'Geese.Arings 29:1541-1552. dea 60:185-215. BATT,B. D. J.(Ed.). 1997. Arcticecosystems in peril: INC. 1990. SAS user'sguide: StatisReport of the Arctic GooseHabitat Working $A$ INSTITUTE,
Group.U.S.FishandWildlifeService, Washington, D.C., and Canadian Wildlife Service, Otta-
tics,version6. SASInstitute,Inc., Cary, North Carolina.
SERWAY, g. A. 1990. Physicsfor scientistsand engineers.Saunders Publishing, Philadelphia. BRODY, S. 1945. Bioenergetics and growth.ReinSKUTCH, A. E 1962. The constancyof incubation. hold, New York. wa, Ontario.
COLLIAS,N., AND E. C. COLLIAS. 1971. Some obser-
Wilson
Bulletin
74:115-152.
S.M. 1994.Neonatereserves, growthand vationson behaviorenergeticsin the Village SLATTERY, Weaverbird.
survival in Ross' and LesserSnow goosegos-
Auk 88:124-143.
CORI•E¾ SMITH,G. H. 1969.A high altitudehummingbirdon thevolcanoCotopaxi.Ibis 111:1722.
FRANKLIN,D.C. 1995. Helmeted Honeyeatersbuild bulkier nests in cold weather. Auk 112:247-248.
HILL,g. W., ANDG. A. WYSE.1989. Animal physiology. Harper and Row, New York.
lings.M. S. thesis,Universityof Saskatchewan, Saskatoon.
SLATTERY,S. M., AND g. t. ALISAUSKAS. 1992. Stud-
iesonthe nutritionalecologyandpopulationbiologyof Ross'and LesserSnowgeesenestingat Karrak Lake, N.W.T. ProgressReport to Califor-
nia Departmentof Fish and Game.Canadian
Wildlife Service,Saskatoon,Saskatchewan. HOYT,D. F. 1979. Practicalmethodsof estimating SLATTERY,S. M., AND g. T. ALISAUSKAS.1993. Stud-
volumeand freshweightof bird eggs.Auk 96: 73-77.
KERBES,R. H. 1994. Colonies and numbers of Ross' Geese and Lesser Snow Geese in the Queen
Maud Gulf Migratory Bird Sanctuary.Canadian Wildlife ServiceOccasionalPapersNo. 81. KRECHEMAR, A. V., AND E. V. SYROECHKOVSKY.1978.
Ecology of incubationin Anser caerulescens on
the WrangelIsland.Zoologicheski Zhurual 57: 899-910.
iesonthenutritionalecologyandpopulation biology of Ross' Geese nesting at Karrak Lake, N.W.T. ProgressReport to California Department of Fish and Game. Canadian Wildlife
Ser-
vice, Saskatoon,Saskatchewan. SLATTERY, S. M., AND g. t. ALISAUSKA$.1995. Egg
characteristics and body reservesof neonate Ross'and LesserSnow geese.Condor97:970984.
$LATTERY,$. g., g. t. ALISAUSKAS,AND G. GENTLE.
LASIEWSKI,g., AND W. g. DAWSON. 1967. A re-ex-
1994. Standard operating proceduresfor re-
amination of the relation between standard met-
search activities
abolicrateandbodyweightin birds.Condor69:
Northern Research Centre, Canadian Wildlife Service,Saskatoon,Saskatchewan.
13-23.
LE$CHACK, C. R., g. D. AFTON, AND R. t. ALISAUSKAS. 1998. Effects of male removal
on female re-
productivebiology in Ross' and LesserSnow geese.WilsonBulletin110:in press. MCLANDRESS, R. M. 1983.Temporalchanges inhab-
at Karrak
Lake. Prairie
and
TEMPLETON, J.g. 1970.Reptiles.Comparative physiologyof thermoregulation, vol. 1 (G. C. whirtow, Ed.). Academic Press,New York. THOMPSON, $. C., AND D. G. RAVELING. 1987. Incu-
bationbehaviorof EmperorGeesecomparedto
October 1997]
NestMorphology ofArcticGeese
617
M. W. 1956. A simple field candlerfor waothergeese:Interactions of predation,bodysize, WELLER, and energetics.Auk 103:707-716. terfowl eggs.Journalof Wildlife Management THOMPSON, S.C., AND D. G. RAVELING. 1988. Nest
insulation and incubation constancyof arctic geese.Wildfowl 39:124-132.
VONHAARTMAN, L. 1958.Theincubationrhythmof the femalePiedFlycatcher(Ficedula hypoleuca) in thepresence andabsence of themale.OrnisFennica 35:71-76.
20:111-113.
WINKLER, D. W. 1993.Useand importanceof feathers as nest lining in Tree Swallows(Tachycineta bicolor).Auk 110:29-36.
AssociateEditor: I. ]. Ball
APPENDIX 1. Nestdimensions(cm)and nestmass(g) for Ross'Goosenestsby habitatat KarrakLake,1994. Valuesare ϖ SE,with rangein parentheses. Measurement Outer diameter Wallthickness Circumference
Rim height Bowl depth Inner diameter Nest mass
Heath (n = 11) 37.6 _+2.8 (27-52) 12.7 ñ 1.4 (7-19) 133.4 _+9.5 (94-180)
5.2 ñ 0.6 (1-8) 6.5 _+0.2 (6-8) 15.2 ñ 0.4 (13-18) 585.0 -+ 92.3 (218-1,050)
Mixed (n = 23) 51.6 ñ 1.0 (36-59) 19.7 ñ 0.5 (13-24) 171.5 _+4.3 (120-222)
7.6 -+ 0.4 (4-10) 7.0 _+0.2 (6-10) 14.7 ñ 0.3 (12-19) 755.1 ñ 66.3 (262-1,642)
Moss (n = 20) 55.4 ñ 1.4 21.7 ñ 0.8 179.8 _+4.1
(43-66) (16-28) (146-206)
7.2 _+0.4 (3-10) 6.7 _+0.2 (5-8) 15.1 ñ 0.4 (12-18) 935.9 _+100.3 (295-1,790)
618
MCCRACKEN, AFTON,ANDALISAUSKAS
[Auk, Vol. 114
NEST MORPHOLOGY
AND BODY SIZE OF ROSS' GEESE AND LESSER SNOW GEESE
KEVIN G. MCCRACKENaTMALAN D. AFTON,1 AND RAY T. ALISAUSKAS2'3 U.S.Geological Survey,Biological Resources Division,LouisianaCooperative FishandWildlifeResearch Unit, LouisianaStateUniversity,BatonRouge,Louisiana70803,USA; 2PrairieandNorthernWildlifeResearch Centre,CanadianWildlifeService, 115 Perimeter Road,Saskatoon, SaskatchewanS7N OX4, Canada;and
3Department ofBiology, University ofSaskatchewan, Saskatoon, Saskatchewan S7NOWO,Canada ABSTRACT.--Arctic-nesting geesebuild large, insulatedneststo protectdevelopingembryosfrom cold ambienttemperatures. Ross'Geese(Chenrossii)are abouttwo-thirdsthe massof LesserSnowGeese(C.caerulescens caerulescens), havehighermass-specific metabolic rate,andmaintainlowernestattentiveness, yet theyhatchgoslings with morefunctionally maturegizzardsandmoreproteinfor theirsizethando LesserSnowGeese.Wecompared nestsize(a reflectionof nestinsulation)in fourdistincthabitatsin a mixedbreedingcolony of Ross' Geese and Lesser Snow Geese at Karrak Lake, Northwest Territories, Canada. After
adjustingmeasurements for nest-specific eggsizeand clutchsize,we foundthatoverallnest
morphology differedbetweenspecies andamonghabitats.Nestsizeincreased progressively amongheath,rock, mixed,and mosshabitats.When nestingmaterialswere not limiting, nestsweresmallerin habitatsthatprovidedcoverfromwind andprecipitation thanin habitatsthat did not providecoverRoss'Geeseconstructed relativelylarger,moreinsulated neststhan did LesserSnowGeese,whichmay hastenembryonicdevelopment, minimize energyexpenditureduringincubation,andminimizeembryoniccoolingduringrecesses. We suggestthat relativedifferences in nestmorphologyreflectgreaterselectionfor Ross'Geese
to improvenestinsulationbecause of theirsmallersize(adultsandembryos), highermassspecificmetabolicrate, and lower incubationconstancy.Received 13 May 1996,accepted 18 March 1997.
IN BIRDS, HEATISSUPPLIED to developingembryosprimarilyby bodywarmthof thebrooding parent (Afton and Paulus1992).Heat loss from developingembryosand metaboliccosts to parentscanbe minimizedby increasingnest insulation.Nestinsulationaffectsgrowthrates of nestlings(Winkler 1993),and probablyaffectsembryonicgrowthaswell,particularlyin precocialspecies.Nest morphologyand insulative propertiesof nestsprobablyare influencedby body size,metabolicrates,incubation constancy, ambientthermalconditions,availabilityof nestingmaterials,and risk of predation (Moller 1984).Mass-specific metabolicrate and heat-transfer rate generallyincreasewith decliningbody mass(Brody1945,Lasiewski and Dawson 1967, Templeton1970, Hill and Wyse 1989). Consequently,smaller species mustassimilatenutrientsat greaterratesthan largerspeciesto maintainhighmetabolicrates. In addition,smallerspecieshavelesscapacity to storeendogenous nutrientsthan do larger 4 E-mail: [email protected]
610
species.Thus,body sizeis of profoundconsequenceto incubatingparents,whichmustfast duringincubation unlesstheyarefedby mates or auxiliaries (von Haartman 1958, Skutch
1962, Afton and Paulus1992). If endogenous nutrients are not sufficientto meet energy requirementsduringincubationsessions, parents may recessfrom incubationin order to feed,
therebydecreasingincubationconstancyand exposingneststo predatorsand heatloss.Afton and Paulus(1992)reportedthat incubation constancyis positively related to body mass amongall speciesof waterfowl(Anatidae),particularly amonggeeseand swans. Within species, nest-building behaviormay be adaptedto maximizeefficiencyof nest insulationin responseto cumulativeembryonic metabolicrates,whichare productsof clutch sizeand embryobodysize.Because of the relationshipbetweensurfacearea and volume, smalleggsand clutcheshave greaterratesof heattransferthanlargereggsand clutches, all elsebeingequal.Thesefactorsare particularly importantunderthewindyconditions thatpre-
October1997]
NestMorphology ofArcticGeese
611
by tundrameadowsand vail in arcticlatitudes(Thompson andRaveling The area is characterized exposedPrecambrianbedrock.Shallow lakes and
1988).
Nestmorphology andinsulationmaybe in- streams are numerous. Karrak Lake contains the fluencedfurtherby localvariationin tempera- largestcolonyofRoss'Geesein NorthAmerica(Kerbes
ture, wind, and availabilityof nestingmateri- 1994)andhasnearlyequalnumbersof LesserSnow Geese(Slatteryand Alisauskas1992,1993).In 1993, als. Comparativestudies of hummingbirds about364,000Ross'and LesserSnowgeesenestedat
(Oreotrochilus spp.;Pearson1953,CorleySmith KarrakLake(Slatteryet al. 1994). 1969), Village Weavers(Ploceus cucullatus; Col-
We examinednestson the largestisland near the
lias and Collias1971),and HelmetedHoney- centerof thelake.Wearbitrarilyselected19circular eaters (Lichenostomus melanops; Franklin 1995)
plots (20-m radius) every 100 m along a transect
demonstrate that birds nestingin cold areas whereRoss'andLesserSnowgeesenestedtogether. construct better insulated nests than those We then selectedeveryRoss'(n = 54) and Lesser clutch nestingunder warmer conditions.Ryder (1964, Snowgoose(n = 51) nestin theplots.Species, 1967, 1972) found that nests of Ross' Geese size,egg length,and eggbreadthwere recordedfor
all nests.Embryoagewasestimated(Weller1956)to calculateclutchinitiation date, assumingfor both tatsthatprovidedshelter. Experimental studies speciesa layingrate of 1.3 daysper eggand an incubationperiodof 23 days(Ryder1967,Ankneyand
(Chenrossii)werelargerin habitatsthatdid not provide shelterfrom wind than thosein habi-
of egg coolingratesin three otherspeciesof
Macinnes 1978). Nestswere consideredsuccessfulif
arctic-nestinggeese(Chencanagica, Brantaber- at leastoneegg hatched.Nest habitatwas classified nicla nigricans,Brantacanadensis minima)indi- as heath,rock, moss,or mixed accordingto domicatethat,acrossspecies, nestinsulationandin- nantfeaturesof theimmediatenesthabitat(seeRycubation constancy are inversely related der 1967, 1972; McLandress 1983). Nests in heath (Thompson andRaveling1988).Finally,intra- habitatswerebuilt directlyinto patchesof Labrador decumbens), white heather(Cassiope tetraspecificand interspecific competitionfor high- tea (Ledurn or willow qualitynestsitesand nestmaterialsalsomay gona),dwarf birch (Betulaglandulosa), (Salixspp.).Nestsin rockhabitatwerebuilt oneither influencenestmorphologyand insulation. We measured
nests of Ross' Geese and Lesser
SnowGeese(Chencaerulescens caerulescens) in a mixedbreedingcolonyin arcticCanada.Weas-
gravelor bedrocksubstrateandcomposed of gravel, soil, or twigs of dwarf birch and willow. Nestsin mosshabitatswere composedof both living and dead Sphagnum speciesas well as nonliving heath fragmentsand birch or willow branches.Nests in mixedhabitatpossessed any combination of char-
sumedthatnestmorphometrics aredirectlyrelated to nestinsulationaccordingto the heatflow equation: actersfound in heath, rock, and mosshabitats. = k. at. (ax)-',
(1)
We measured(_+1 cm) outer diameter,wall thick-
ness,circumference, rim height,bowl depth,and inwhereH represents heatflow,k represents ther- ner diameterof eachnest during initial nestvisits (Appendices1 and 2). After the eggshatched,nests mal conductivity,T representstemperature, wereweighedon a portableelectronicscale(_+1 g).
andx represents thickness of insulatingmate-
Masses of 10 Lesser Snow Goose nests were estimat-
rial (Serway1990).We predictedthat nestsize ed using stepwisemultiple regression(PROCREG;
wouldvary inverselywith body size and in- SAS 1990) becausenest markers were lost between cubationconstancyand would vary positively initial and final nest visits. Best fit was obtained with with mass-specific metabolic rate.Wealsopre- separateequationsfor eachhabitat: dictedthatnestsizeshouldbe greaterin habih = --599.1 + 4.9(circumference) tatsthataremoreexposed to wind andprecip- masshoar itation,whenadequatenestmaterialsareavail+ 14.4(outerdiameter) able(Ryder 1964,1967,1972). METHODS
We measured
(r2 = 0.74, P < 0.0001);
massrock = -2073 + 170.8(wallthickness) (r2 = 0.76, P < 0.0046); and
nests of Ross' Geese and Lesser
Snow Geese at Karrak Lake, Northwest Territories,
Canada(67ø15'N,100ø15'W)between19 Juneand 15 July 1994.Karrak Lake is locatedon the west tribu-
tary of the SimpsonRiver in the centralCanadian Arctic,southof theQueenMaudGulf (Ryder1972).
(2)
(3)
massm•xe d = -782.1 + 29.5(outer diameter)
(r2 = 0.33, P < 0.0001). Total
clutch volume
was calculated
(4) for each nest
612
MCCRACKEN, AFTON,AND ALISAUSKAS
from length(_+0.1 ram) and breadth(_+0.1 ram) of eacheggin a nestfollowingHoyt (1979;Kv= 0.51). To estimate
the relative
difference
in overall nest
[Auk, Vol. 114
(PROCCATMOD;SAS1990)to testwhethernestsuccess varied between speciesand among habitat types.
morphology betweenspecies andtoaccount forvariationdue to differencesin egg size and clutchsize, we dividednestmeasurements by the squarerootof the total clutchvolumefor eachnest,exceptfor nest
RESULTS
The averagedatethatfirst eggswerelaid was 3 June,and this did not differ betweenspecies analysisof covariance (ANCOVA)with clutchvolor among habitatsand plots (all Ps > 0.12). ume as a covariate,or using residual nest measurementscorrectedfor clutchvolume,would be inap- Clutch size was inverselyrelated to first-egg propriatebecauseof significantheterogeneity in date (F = 1.44, df = 1 and 101, P = 0.0001)but clutchvolumebetweenspecies (F = 45.26,df = 1 and did not differ betweenspeciesor vary in relamass,which was divided by clutchvolume.Using
date,or plot88, P < 0.0001,r2 = 0.34);i.e. significantinterspecific tion to plot, species-by-initiation colinearitywaspresentamongpredictorvariablesin by-species-by-initiation date (all Ps > 0.23). the model.We subsequently determinedthat the squarerootof clutchvolumeratherthanclutchvolume or the cuberoot of clutchvolumeis the appropriatedenominator for linearmeasurements. Using clutch volume and the cube root of clutch volume as
thedenominator did not removethespecies effecton inner diameter(P < 0.05), whereasdividing by the squarerootof clutchvolumerenderedspecies effects insignificant,as would be expectedif nestinner diametersmatchedthebody form of eachspecies. We usedmultivariateanalysisof variance(MANOVA) to test whether nest measurementsdiffered amongspeciesandhabitats(PROCGLM; SAS1990). We useda split-plotdesignto accountfor variance
Mean nest success was 92.6 --- SE of 2.8% and
did not differ between species(X2 = 0.06, n = 87, P = 0.805)or amonghabitats(X2 = 0.02, n = 87, P = 0.992).
Variationbetween species andamonghabitats.The overall nest morphologycorrectedfor clutchvolumedifferedbetweenspecies(MANOVA, F = 3.90, df = 7 and 78, P = 0.001) and
amonghabitats(F = 5.87,df = 14 and156,P = 0.0001).Ploteffects,andspecies-by-habitat and plot-by-species-by-habitat interactionswere not significant(all Ps > 0.10).The outer diameter, wall thickness, circumference, and rim
due to plot effect,whichpotentiallycontainscom- height of Ross' Goosenests were relatively ponentsofbothsamplingerrorandnaturalvariation due to density and nonrandom distributionsof geese.In the full model, speciesand habitatwere fixed effects,whereasplot was a randomeffect.Fvalues reported from MANOVA were determined usingWilks' lambda.Beginningwith thelargest-order interaction,nonsignificanteffectswereiterative-
larger than thoseof LesserSnowGoosenests (Table1). Nest-bowl depth and nest inner diameterdid not differ betweenspecies,probably becausethey are a functionof femalebody size,eggsize,andclutchsize(all of whichwere accountedfor by the correctionfactor).Despite
ly removedfrom the modeltmtil we obtainedthe differences in relative size of nests, nest mass most parsimonious modelcontainingonly signifi- did not differ between species,perhaps becant effects. Following a significantMANOVA, we
cause nests were not dried, and small amounts
usedanalysisof variance(ANOVA)to testwhether of soil, pebbles,and old nest fragmentscould
individual nest measurementsdiffered among significant effects.We comparedleast squaresmeans amonghabitattypeswith t-tests(PDIFFoption;SAS 1990).We did not find any Ross'Goosenestsin rock habitats;consequently, onlyheath,moss,and mixed habitat nests were included in models comparing
not be separatedfrom nests.Outer diameter, wall thickness,circumference, rim height,and nestmassincreasedprogressively acrossheath, mixed, and mosshabitats,i.e. from more protectedto lessprotectedhabitats.Bowl depth species andhabitats. Wesubsequently compared the and inner diameterdid not differ amonghabfour habitatsin a separatemodelfor LesserSnow itats (Table2). Geese. In this model, we did not divide nest mea-
surements by the squareroot of clutchvolumebut ratherincludedthe squareroot of clutchvolumeas a covariate because clutch volume and habitat were
not related (F = 0.96, df = 7 and 40, P = 0.47). We
Variation among habitats in Lesser Snow Geese.--The overallnestmorphologyof Lesser Snow Geesediffered amonghabitats(MANCOVA, F = 3.97, df = 21 and 118 P = 0.0001)
but was unrelated to clutchvolume (F = 1.02, usedANOVA to testwhetherfirst-eggdatediffered betweenspecies,and amonghabitatsand plots.We df = 7 and 40, P = 0.43). Outer diameter, wall rim height,and nest usedANCOVAto comparemeanclutchsizebetween thickness,circumference, mass generally increased amongheath,rock, species andamongplots,with initiationdateincluded as a covariate.We used categoricaldata analyses mixed, and mosshabitats,i.e. from more pro-
October 1997]
NestMorphology ofArcticGeese
613
TABLE1. Least-squares means(œ_+SE)of nestdimensions and nestmassadjustedfor clutchvolumefor Ross'Geeseand LesserSnowGeesenestingat Karrak Lake,1994. Ross' Goose
Measurementa
Lesser Snow Goose
(n = 54)
(n = 36)
pb
Outer diameter Wall thickness Circumference
0.0925 _+ 0.0024 0.0347 _+ 0.0012 0.0310 --- 0.0086
0.0838 --- 0.0031 0.0302 _+ 0.0016 0.0265 -+ 0.0110
0.031 0.028 0.002
Rim height Bowldepth
0.0127_+0.0006 0.0129_+0.0003
0.0105-+ 0.0008 0.0120-+0.0004
0.032 0.072
Inner diameter Nest mass
0.0287 0.0030
0.0281 -+ 0.0006 0.0025 +- 0.0003
0.501 0.200
_+ 0.0005 -+ 0.0002
Measurements dividedby thesquarerootof clutchvolumeexceptfornestmass,whichwasdividedby clutchvolume. ANOVA for specieseffect(df = 1 and 84 for eachtest).
tectedto lessprotectedhabitats(Table3). Nest bowls were shallowest
Like most waterfowl, Ross' and Lesser Snow
in rock habitats and dif-
geese line their nests with considerable fered from those in heath, mixed, and moss amountsof down and covertheir eggsbefore habitats.As expected,the inner diameterof recessingfrom incubation.Down providesexLesserSnowGoosenestsdid not differ among cellent insulation, and the amount of down in habitats. a nest may vary betweenspeciesand among habitats.We did not quantifythe amountof down or the specificcompositionof materials DISCUSSION in eachnest,but downmaybeparticularlyimThompsonand Raveling(1987) suggested portantfor geesenestingin habitatswhere that predationinteractswith body size in de- nestingmaterialsare scarce.Differenttypesof termining incubation constancy in geese, nestingmaterials(e.g. heath and moss)may whichin turnmayberelatedtonestinsulation. differ in their insulativeproperties.Although In areaswith highratesof predation,theuseof we did not identify nestmaterialsspecifically, feathersin nestsfor insulation(Mailer 1984) nest materials are related to the distribution of may be counteractedby selectionfor incon- materialsin a givenhabitattype,i.e.Ross'and spicuousnestsif feathersare not cryptic.We LesserSnow geesedo not use differentmatebelievethat concealment of eggsand nestsof rialswithin a habitat(pers.obs.).ConsequentRoss' Geese and Lesser Snow Geese at Karrak ly, inclusionof thehabitateffectin ouranalyses Lakeisnotinfluenced by strongselective forces should account for variation in the insulative of predation.Ross'and LesserSnowgeeseare propertiesof differentnest materials. non-cryptic,colonial and nest in extremely Nestmorphology andhabitatvariation.--Ryder densenumbers.Our claimis furthersupported (1964,1967,1972)foundthat Ross'Goosenests by high nestingsuccess (Slatteryand Alisaus- were largestin mosshabitats,intermediatein mixed habitats, and smallest in rock habitats.
kas 1992, 1993).
TABLE 2. Least-squares means(œ_+SE)of nestdimensions and nestmassadjustedfor clutchvolumefor Ross'Gooseand LesserSnowGoosenestscombinedby habitatat KarrakLake,1994. Nestinghabitat Measurement• Outer
Heath (n = 18)
Mixed (n = 43)
diameter
0.0664 _+ 0.0041 ^
0.0948
Wall thickness
0.0218 _+ 0.0021 ^
Circumference
0.2252 _+ 0.0145 ^
Rim height Bowl depth Inner diameter Nest mass
-+ 0.0025 B
Moss (n = 29)
pb
0.1032 _+ 0.0034 •
0.0001
0.0356 -+ 0.0013 B
0.0399 _+ 0.0017 B
0.0001
0.3006 -+ 0.0092 •
0.3376 _+ 0.0121 c
0.0001
0.0086___ 0.0011^ 0.0121_+0.0005
0.0130-+ 0.0007B 0.0124-+ 0.0003
0.0133_+0.0009B 0.0127ñ 0.0004
0.001 0.727
0.0287 -+ 0.0008 0.0019 _+ 0.0004 ^
0.0276 -+ 0.0005 0.0025 -+ 0.0002 ^
0.0289 _+ 0.0007 0.0038 -+ 0.0004 B
0.274 0.002
• Measurements dividedby the squarerootof clutchvolumeexceptfor nestmass,whichwasdividedby clutchvolume. bANOVA for habitateffect(df = 2 and 84 for eachtest).Within rows,valueswith differentsuperscripts aresignificantlydifferent(P < 0.05) basedon t-testsfollowinga significanteffectof habitat.
614
MCCRACKEN, AFTON,ANDALISAUSKAS
[Auk,Vol. 114
TABLE3. Least-squares means(• _+SE)of nestdimensionsand nestmassfor LesserSnowGoosenestsby habitat at Karrak Lake, 1994.
Nestinghabitat Measurement a Outer diameter Wall thickness Circumference
Rim height Bowldepth Inner diameter
Nest mass
Heath (n = 7)
Rock (n = 15)
38.7 +- 2.9 ^ 12.1 + 1.5 ^ 121.0 _+ 11.2 ^
49.7 +- 2.0 B 17.5 -+ 1.0 B 164.3 q- 7.6 B
4.7 + 1.0^ 7.4 _+0.4^ 18.1 -+ 0.7
563.7 + 177.0 ^
5.7 -+0.7^.B 6.2 +_0.3B 18.0 _+ 0.5
910.7 +_ 120.9 ^
Mixed (n = 20) 58.5 + 1.7 c 21.6 q- 0.9 c 177.2 + 6.6 B
7.3 + 0.6B 7.6 + 0.3^ 17.6 -+ 0.4
921.7 + 104.7 ^
Moss (n = 9) 60.8 + 2.6 c 23.1 q- 1.3 c 201.2 +_ 9.8 c
7.8 -+ 0.9B 7.7 + 0.4^ 17.6 q- 0.6
1,379 + 156.1B
Pa 0.0001 0.0001 0.0001
0.043 0.003 0.827
0.011
• ANOVA for habitateffect(df = 3 and 47 for eachtest).Within rows,valueswith differentsuperscripts are significantlydifferent(P < 0.05) basedon t-testsfollowinga significanteffectof habitat.
He suggestedthat thesedifferencesresulted Competition.--Competition for high-quality from differencesin availabilityof nestmaterial nest sites and nest materials may influence and exposureto weather.Our data for Ross' nest-site selection such that access to the best and LesserSnow geeseclearly are consistent sitesvariesamonggeeseat boththemacrohabwith Ryder'sobservations. Wealsoconcurwith itat andmicrohabitat levels.This,in turn,may his interpretationof the adaptivesignificance influence nest size and nest success. In rock and of variationin nest size amonghabitats.Nest other habitatswhere nest materialsare limited, materials are abundant in heath, mixed, and all availablenest materials are incorporated mosshabitats(Ryder1967,pers.obs.),but rock into nests(Afton unpubl. data). Moreover,we habitatstypicallypossess littlenestingmaterial haveobservedgeesefighting(within and beotherthansmalltwigsandgravel.Nestsbuilt tween species)over nest materialsduring the in heathtypicallyare shelteredby low (1 to 6 prelayingand earlylayingperiods.In light of cm) vegetationthat could decreaseconvective exponentialpopulation growth and habitat heat loss from nests. Nests built in mixed habdegradationat Karrak Lake and elsewhere itats often are constructedof vegetationor in (Slatteryet al. 1994,Batt1997),somegeesemay rocksthatmayprovidesheltering effects.Nests be nestingin poorerhabitatwith the net effect built in openmossappearto be the leastshel- that globalnest success may have declined tered.Althoughnestsin rocktypicallyoccurin from previousyearswhen nestdensitieswere unvegetatedsiteswhere exposureis greatest, lower Macrohabitat and microhabitat selection.--Macwe suspectthatplacement of nestswithincrevicesor in theleeof largerockscandramatically rohabitatselectionprobablyis a functionof decreaseexposureand convectiveheat loss. habitatavailabilityduring nestinitiationand The outer nest dimensions,includingouter exposureto weathercausedby habitattopogdiameter, wall thickness, circumference, and raphy. Ryder (1967, 1972) found that Ross' rim height,providethe bestestimatesof the Goosenestdensitieswerehighestin mixedand size of the nest wall. These measurements heathhabitatsand lowestin the moreexposed should be a direct reflection of nest insulation rock and moss habitats.McLandress(1983) accordingto the heat-flowequation.These demonstrated that Lesser Snow Geese show a measurementsdiffered significantly among markedpreference for rockhabitats.This may habitatsand generallyincreasedacrossheath, be becauserockhabitatsareusuallythe firstto rock, mixed, and mosshabitats. Inner nest mea- clear of snow and dry out each spring, and surements also should influence nest insulaLesserSnow Geese typically initiate nesting tion,butshouldconformmorecloselyto female earlier than Ross'Geeseat Karrak Lake (Slatbody shape,egg size,and clutchsize than to tery and Alisauskas1993). We found no Ross' habitat.Inner diameterdid not differ among Goosenestsin rock habitatsor on the upperhabitats.Bowldepthdid not vary amonghab- most parts of eskers,althoughLesserSnow itatsexceptin rock,whereit was smallest;the Geesewere commonlyfound nestingin such impenetrabilityof rock and gravel substrates locations. We arguethatRoss'Geesedo not ocprobablyexplainsthis finding. cupy such habitatsbecauseof their smaller
October1997]
NestMorphology ofArcticGeese
615
body size,highermass-specific metabolicrate, lower incubation constancyrelative to Lesser lower incubationconstancy,and smaller egg Snow Geese. andclutchsize.Furthermore, Ross'GeeseprobBy constructing relativelylargerand probaably incur energeticadvantagesby initiating bly better-insulatednests,female Ross'Geese nestsa few daysafter snowhas clearedfrom can minimize energyexpenditureduring inthe more sheltered heath and mixed habitats. cubationand embryoniccoolingduring recesBecause of theirlargerbodysize,LesserSnow ses.Slatteryand Alisauskas(1995)found that Geeseprobablyare better able to toleratecon- neonate Ross' Geese hatch with more functionditionsin themoreexposedrockhabitats.Con- ally maturegizzardsandmoreproteinfor their sequently, theytendto nestin rockhabitatsout size than did Lesser Snow Geese. In addition, Ross'Goosebroodsdispersefartherfrom nestof proportionto its availability. Microhabitatnest-siteselectionprobablyoc- ing areasthan do LesserSnowGoosebroods cursaftergeesehavesetup territories.Within (Slattery1994).Moreover,Ross'Goosegoslings a territory, there is considerablevariation fledge two weeks earlier than LesserSnow amongpossiblenest sites,and femalessome- Goose goslings(Owen 1980). The relatively timesexhibitnest-buildingbehaviorin several larger size and greater insulation of Ross' locations withintheterritorypriorto layingthe Goosenests probably buffer developingemfirst egg (Afton unpubl.data). A small dwarf bryos against fluctuationsin ambient temperdevelopment, andulbirch,patchof Labradortea, or pile of small atures,hastenembryonic boulders,lessthan a meter in diameter,may timatelypromotea greaterpropensityfor acyield a nestsite that providesexcellentprotec- tivity during the brood-rearingperiod (Slattion from wind, whereasan adjacentarea of tery and Alisauskas1995). In light of the physiological differences bebaregroundmayofferlittle protectionfromthe tween Ross' Geese and Lesser Snow Geese, we wind. Small-scale factorssuchasproximityto are adjacentterritory holdersmay influencenest- suggestthat overall fitnessconsequences site selection further
associated with nest construction and selection
Nest morphology and variationbetweenspe- of nestinghabitatin arctic-nestinggeese.We
hopethatthispaperwill inspiremorecomprehensivestudiesof fitnessconsequences resulting from nest morphology and site selection. specificbasalmetabolicrate that is 1.18kcal ß day-• ß g • higher than that of LesserSnow Experimentsinvolving manipulation of nest Geese (Aschoffand Pohl 1970, Alisauskasand materials (particularly down) coupled with of incubationtemperaturesand Ankney 1992). FemaleRoss'and LesserSnow measurements hatchling body condition wouldbeparticularly geese exhibit high incubation constancy cies.--Ross' Geese are about two-thirds the mass of Lesser Snow Geese and exhibit a mass-
throughoutthe 23-dayincubationperiodand rely heavily upon endogenous nutrientsfor both egg laying and incubation(Ryder 1967,
informative.
ACKNOWLEDGMENTS
1970, Ankney and Macinnes 1978, Afton and We thankI. J.Ball,B. Batt,G. Gauthier,W. Koenig, Paulus 1992, Alisauskas and Ankney 1992). T. Martin, M. R. McLandress,J.Ryder,B. Wilson,and
However,Ross'Geesehaveslightlylowerincubationconstancythan LesserSnowGeese (Krechemarand Syroechkovsky 1978, Afton and Paulus 1992, LeSchacket al. 1998) and spendmore time foragingduring incubation (Aftonunpubl.data).Whencorrected forclutch volume(whichis a functionof body size,egg size, and clutchsize), nestsof Ross'Geeseare largerthannestsof LesserSnowGeese.These resultsprobablyreflectgreaterselectionfor ef-
two anonymous reviewersfor helpful commentson the manuscript,and C. LeSchack,D. McCracken,N. McCracken,J.Ryder,S. Slattery,D. Stern,V. Wright, and all thosewho spentthe summerof 1994at Karrak Lake for ideas and assistance. This research was
supported by theCaliforniaDepartmentof Fishand Game; the Arctic GooseJointVenture; the Canadian
Wildlife Service,PolarContinentalShelfProject;the Schoolof ForestryWildlife and Fisheries,Collegeof Agriculture, and Agricultural Center of Louisiana StateUniversity;theLouisianaBoardof Regents; the
ficient nest insulationby Ross'Geesebecause LouisianaDepartmentof WildlifeandFisheries; and of their smallerbody size (adultsand embry- the LouisianaCooperativeFish and Wildlife Reos), greater mass-specific metabolicrate, and search Unit.
616
MCCRACKEN, AFTON,ANDALISAUSKAS LITERATURE CITED
[Auk, Vol. 114
itat selectionand nest spacingin a colonyof Ross'and LesserSnowgeese.Auk 100:335-343.
AFTON, A.D., AND $. L. PAULUS. 1992. Incubation
and brood care. Pages62-108 in Ecologyand managementof breedingwaterfowl (B. D. J.Batt, A.D. Afton, M. G. Anderson,C. D. Ankney, D.
H. Johnson, J.A. Kadlec,andG. L. Krapu,Eds.). University of MinnesotaPress,Minneapolis. ALISAUSKAS, g. t., AND C. D. ANKNEY. 1992. The cost
of egglayingand its relationshipto nutrientreservesin waterfowl. Pages30-61 in Ecologyand managementof breedingwaterfowl(B. D. J.Batt, A. D. Afton, M. G. Anderson,C. D. Ankney,D.
MOLLER,A. P. 1984. On the use of feathers in birds nests: Predictions
and tests. Ornis
Scandinavica
15:38-42.
OWEN,M. 1980.Wild geeseof theworld.B.T. Batesford, London.
PEARSON, O. P. 1953.Useof cavesbyhummingbirds and otherspeciesat high altitudesin Peru.Condor 55:17-20.
RYDER, J.P. 1964. A preliminarystudyof the breeding biology of Ross'Goose.Wildfowl 15:127137.
J.P. 1967. The breedingbiologyof Ross' H. Johnson, J.A. Kadlec,andG. L. Krapu,Eds.). RYDER, Goosein thePerryRiverregion,NorthwestTerUniversity of MinnesotaPress,Minneapolis. ritories.CanadianWildlife ServiceReport 3:1ANKNEY, C. D., AND C. D. MACINNES. 1978. Nutrient 56. reservesand reproductiveperformanceof female Lesser Snow Geese. Auk 95:459-471. RYDER, J.P. 1970. A possiblefactorin the evolution of clutch size in Ross' Goose. Wilson Bulletin 82: ASCHOFF, J., ANDH. POHL.1970. Rhythmicvaria4-13.
tionsin energymetabolism.FederationProceedRYDER, J.P. 1972.Biologyof nestingRoss'Geese.Arings 29:1541-1552. dea 60:185-215. BATT,B. D. J.(Ed.). 1997. Arcticecosystems in peril: INC. 1990. SAS user'sguide: StatisReport of the Arctic GooseHabitat Working $A$ INSTITUTE,
Group.U.S.FishandWildlifeService, Washington, D.C., and Canadian Wildlife Service, Otta-
tics,version6. SASInstitute,Inc., Cary, North Carolina.
SERWAY, g. A. 1990. Physicsfor scientistsand engineers.Saunders Publishing, Philadelphia. BRODY, S. 1945. Bioenergetics and growth.ReinSKUTCH, A. E 1962. The constancyof incubation. hold, New York. wa, Ontario.
COLLIAS,N., AND E. C. COLLIAS. 1971. Some obser-
Wilson
Bulletin
74:115-152.
S.M. 1994.Neonatereserves, growthand vationson behaviorenergeticsin the Village SLATTERY, Weaverbird.
survival in Ross' and LesserSnow goosegos-
Auk 88:124-143.
CORI•E¾ SMITH,G. H. 1969.A high altitudehummingbirdon thevolcanoCotopaxi.Ibis 111:1722.
FRANKLIN,D.C. 1995. Helmeted Honeyeatersbuild bulkier nests in cold weather. Auk 112:247-248.
HILL,g. W., ANDG. A. WYSE.1989. Animal physiology. Harper and Row, New York.
lings.M. S. thesis,Universityof Saskatchewan, Saskatoon.
SLATTERY,S. M., AND g. t. ALISAUSKAS. 1992. Stud-
iesonthe nutritionalecologyandpopulationbiologyof Ross'and LesserSnowgeesenestingat Karrak Lake, N.W.T. ProgressReport to Califor-
nia Departmentof Fish and Game.Canadian
Wildlife Service,Saskatoon,Saskatchewan. HOYT,D. F. 1979. Practicalmethodsof estimating SLATTERY,S. M., AND g. T. ALISAUSKAS.1993. Stud-
volumeand freshweightof bird eggs.Auk 96: 73-77.
KERBES,R. H. 1994. Colonies and numbers of Ross' Geese and Lesser Snow Geese in the Queen
Maud Gulf Migratory Bird Sanctuary.Canadian Wildlife ServiceOccasionalPapersNo. 81. KRECHEMAR, A. V., AND E. V. SYROECHKOVSKY.1978.
Ecology of incubationin Anser caerulescens on
the WrangelIsland.Zoologicheski Zhurual 57: 899-910.
iesonthenutritionalecologyandpopulation biology of Ross' Geese nesting at Karrak Lake, N.W.T. ProgressReport to California Department of Fish and Game. Canadian Wildlife
Ser-
vice, Saskatoon,Saskatchewan. SLATTERY, S. M., AND g. t. ALISAUSKA$.1995. Egg
characteristics and body reservesof neonate Ross'and LesserSnow geese.Condor97:970984.
$LATTERY,$. g., g. t. ALISAUSKAS,AND G. GENTLE.
LASIEWSKI,g., AND W. g. DAWSON. 1967. A re-ex-
1994. Standard operating proceduresfor re-
amination of the relation between standard met-
search activities
abolicrateandbodyweightin birds.Condor69:
Northern Research Centre, Canadian Wildlife Service,Saskatoon,Saskatchewan.
13-23.
LE$CHACK, C. R., g. D. AFTON, AND R. t. ALISAUSKAS. 1998. Effects of male removal
on female re-
productivebiology in Ross' and LesserSnow geese.WilsonBulletin110:in press. MCLANDRESS, R. M. 1983.Temporalchanges inhab-
at Karrak
Lake. Prairie
and
TEMPLETON, J.g. 1970.Reptiles.Comparative physiologyof thermoregulation, vol. 1 (G. C. whirtow, Ed.). Academic Press,New York. THOMPSON, $. C., AND D. G. RAVELING. 1987. Incu-
bationbehaviorof EmperorGeesecomparedto
October 1997]
NestMorphology ofArcticGeese
617
M. W. 1956. A simple field candlerfor waothergeese:Interactions of predation,bodysize, WELLER, and energetics.Auk 103:707-716. terfowl eggs.Journalof Wildlife Management THOMPSON, S.C., AND D. G. RAVELING. 1988. Nest
insulation and incubation constancyof arctic geese.Wildfowl 39:124-132.
VONHAARTMAN, L. 1958.Theincubationrhythmof the femalePiedFlycatcher(Ficedula hypoleuca) in thepresence andabsence of themale.OrnisFennica 35:71-76.
20:111-113.
WINKLER, D. W. 1993.Useand importanceof feathers as nest lining in Tree Swallows(Tachycineta bicolor).Auk 110:29-36.
AssociateEditor: I. ]. Ball
APPENDIX 1. Nestdimensions(cm)and nestmass(g) for Ross'Goosenestsby habitatat KarrakLake,1994. Valuesare ϖ SE,with rangein parentheses. Measurement Outer diameter Wallthickness Circumference
Rim height Bowl depth Inner diameter Nest mass
Heath (n = 11) 37.6 _+2.8 (27-52) 12.7 ñ 1.4 (7-19) 133.4 _+9.5 (94-180)
5.2 ñ 0.6 (1-8) 6.5 _+0.2 (6-8) 15.2 ñ 0.4 (13-18) 585.0 -+ 92.3 (218-1,050)
Mixed (n = 23) 51.6 ñ 1.0 (36-59) 19.7 ñ 0.5 (13-24) 171.5 _+4.3 (120-222)
7.6 -+ 0.4 (4-10) 7.0 _+0.2 (6-10) 14.7 ñ 0.3 (12-19) 755.1 ñ 66.3 (262-1,642)
Moss (n = 20) 55.4 ñ 1.4 21.7 ñ 0.8 179.8 _+4.1
(43-66) (16-28) (146-206)
7.2 _+0.4 (3-10) 6.7 _+0.2 (5-8) 15.1 ñ 0.4 (12-18) 935.9 _+100.3 (295-1,790)
618
MCCRACKEN, AFTON,ANDALISAUSKAS
[Auk, Vol. 114
