LANDSCAPE MODIFICATION AND PATCH SELECTION: THE

advertisement
TheAuk 114(3):443-455,1997
LANDSCAPE
MODIFICATION
AND PATCH SELECTION:
THE
DEMOGRAPHY
OF TWO SECONDARY
CAVITY NESTERS
COLONIZING
CLEARCUTS
RACHEL E HOLT TM AND KATHY MARTIN 2'3
•Divisionof LifeSciences,
Universityof Toronto,
Toronto,
OntarioM5S 1A1, Canada;
2Centre
for AppliedConservation
Biology,
Universityof BritishColumbia,
Vancouver,
British ColumbiaV6T 1Z4, Canada;and
3Canadian
WildlifeService,Delta,BritishColumbiaV4K 3Y3, Canada
ABSTRACT
.--Currentforestrypractices
radicallymodifyhabitat,particularlybyincreasing
the frequencyof earlyseralstages.Weexaminedthedemography
of two secondary
cavity
nesters,
theMountainBluebird(Sialiacurrucoides)
andtheTreeSwallow(Tachycineta
bicolor),
with respectto landscape
changes.
Thestudyareacontained
a mosaicof clearcuts
ranging
from 3 to 27 yearsin age.The initial distributionof both specieswasunevenwith respect
to patchage.Experimental
additionof nestboxes:(1) increased
thenumberof nestingpairs
of bothspecies,
and (2) expandedtheagerangeof occupied
patches.
Nest-site
availability
wastheprimarylimitingfactorin youngpatches,
but vegetation
structurewasincreasingly
importantin determiningoccupation
in olderpatches.
MountainBluebirdpatchoccupation
andTreeSwallowdensitywerebestpredictedby theproportionof trees1 to 3 m in height.
Althoughsomepatchescontainedmultiplepairsof MountainBluebirds
in naturalcavities,
experimental
patchestypicallycontainedno morethanonepair,eventhoughexcess
nest
boxeswere available.The presenceof nonbreedingfloaterswas an indicationof nest-site
limitationandsuggested
thatMountainBluebirds
sometimes
deferbreeding
attempts
rather
than nestcloseto conspecifics.
At low densities,Tree Swallowsalsowere over-dispersed,
but pairs saturatedavailablenestboxesas densityincreased.Clutchsize and numberof
chicks
fledgedfromsuccessful
nestsdidnotvarywithpatchageforeitherspecies,
although
swallowsinitiatedclutches
later in youngercomparedwith olderpatches.Nestpredation
variedconsistently
with patchage;meannestpredationfor bothspecies
was86%(n = 30
pairs)in the youngestpatches,
37%(n = 34) in mid-agepatches,
and 25%(n = 22) in the
oldestpatches.
The low densitiesof bothspeciesin olderpatcheswasnot associated
with
low productivity,
suggesting
thathabitatselection
occurredindependently
of potentialproductivityin thisnovelhabitat.Received
21 June1996,accepted
10 February
1997.
LANDSCAPE MODIFICATION can have various
effectson bird populations.For example,habitat loss may reduce overall population size
(LynchandWhigham1984)and decrease
population resilienceafter stochasticevents(Freemark and Merriam 1986).Habitat fragmentation hasthe potentialto createsource/sinkdynamicsthat causeindependentpopulationsto
becomedependenton sourcepopulations(Pulliam 1988).Edgeeffectscanincreasenestpredation for passerines(Wilcove 1985, Paton
1994,Yahnerand Mahan 1996),thus creating
ecologicaltraps (Gates and Gysel 1978, Ratti
and Reese1988,Terborgh1992).Most studies
of the effectsof landscapemodificationon
birds have focusedon speciesin fragmented
4E-mail:rholt@unixg.ubc.ca
443
originalhabitat,ratherthanon speciescolonizing the newly createdhabitat.
Clearcuts
in mature
forest increase the fre-
quencyof early seralstagesin the landscape.
Forest managementmay mimic forest landscapepatterns traditionally produced in firedominated
ecosystems
(BunnellandKremsater
1990).Speciesthatnaturallyinhabitburnsmay
colonize clearcuts as an alternate habitat because the role of fire has been reduced or erad-
icated in many forest regions (Hutto 1995).
However,habitatssuchasclearcutsmaysuperficially resembletraditionalhabitatsbut differ
in subtle attributes (Bunnell and Kremsater
1990) that can influencepopulationdemography (e.g. Zwickel and Bendell1972,Dunning
and Watts1990).
Habitat quality cannot always be assessed
from populationdensitybecausereproductive
successcan differ independentof density(Van
444
HOLTANDMARTIN
Horne 1983, Vickey et al. 1992). In particular,
speciesmay be poorly adapted to a novel or
modified environment.An understandingof
individual
[Auk,Vol. 114
assessed
solelyby observedcavityuse,because
natural cavitiesmay be unevenlydistributedin
the environment
andmayvaryin quality.Ac-
habitat selection decisions and the
cordingly,we added nest boxesto 15 clearcut
subsequentfitness consequences
associated patchesin orderto standardizethe availability,
with thesedecisions
is necessary
to assess
the distribution,and qualityof nestingsites.Using
potential influenceof habitat modificationon data from natural cavities and nest boxes, we
individuals, populations,and ultimately on tested: (1) preferencesfor, and productivity
differences between natural cavities and nest
biodiversity (Martin 1992).
Habitat
selection in variable environments
boxes;(2) factorslimiting breedingdensityfor
hasbeenmodeledundervariousassumptions;speciescolonizingthis novelhabitat;(3) proxin particularthe Ideal FreeDistribution(Fret- imate factors influencing selection among
well andLucas1970),theDespoticDistribution patchesof differentagesandwithin individual
(Fretwell and Lucas 1970, Fretwell 1972), and patches;
and (4) the overalldemographic
conthe Pre-emptiveModel (Pulliam1988,Pulliam sequences
of habitat selectionand nest-site
and Danielson1991).An assumptioncommon preferences.
to these models is that the observed distribu-
tion of individualshas a density-dependent
STUDY AREA AND METHODS
component.Interpretationof thedistributionof
individualsin newlycreatedhabitatsrequires Studysite.--The study site was in the Beaverfoot
knowledgeof whetherthe populationdensity Valleyin the RockyMountainssoutheastof Golden,
approaches
carryingcapacity.If the habitatis British Columbia (50øN, 116øW,1,400 m elevation;
not saturated,thenall individualsmayoccupy Fig. 1). The valleyoriginallycomprisedcontiguous
goodhabitat.At highpopulationdensity,how- montanespruceforest(white spruce[Piceaglauca],
andhybrids)and
ever,someindividualsmayoccupylower-qual- Engelmannspruce[P.engelmannii],
ity habitat.Thepresence
ofnonbreeding
"float- lodgepolepine (Pinuscontorta),with a fire-dominated ecology.Loggingbeganin 1968 and continuesto
er" maleshasbeendemonstrated
in manybird thepresenttime,resultingin a mosaicof 70clearcuts
populations(Brown1969,Martin 1989)and im-
(10 to 85 ha each)surroundedby matureforest(Fig.
plies that breedingpopulationsize is limited 1). Mostof theclearcutshavebeenreplantedandare
by a shortageof femalesand/or territories. at variousstagesof regeneration,ranging from bare
However, the demonstrationof floatersof both
patches(loggedin 1993)to standsof 4-m-talllodge-
sexes,necessaryto infer habitatsaturation,has pole pine and hybrid spruce.
been demonstrated less often (Power 1975,
Thecavity-nesting
assemblage.--The
clearcutswere
Hannon and Zwickel 1979, Village 1983, colonizedby two speciesof primary cavity nesters
(NorthernFlicker[Colaptes
auratus]
andHairy WoodStutchburyand Robertson1985).
villosus])and four speciesof secondWe studiedhabitatselectionand productiv- pecker[Picoides
ary cavity nesters(Mountain Bluebird,Tree Swallow,
ity of two secondarycavitynesters,the Moun-
Black-cappedChickadee[Parusatricapillus],
and Boreal Chickadee[P.hudsonicus]).
Bothprimary cavity
Swallow (Tachycineta
bicolor),that colonized nestersoccurredin low numbers,with approximateclearcutsin BritishColumbia.Cavity-nesting ly 15 pairsof eachspeciesin the studyarea.Mostof
tain Bluebird (Sialia currucoides)
and the Tree
birds often are limited by availabilityof suit-
the natural cavities in the Beaverfoot were excavated
able nest sites (Scott 1979, Brawn and Balda by Northern Flickers(50%of 42 natural cavitiesused
1988, Newton 1994), and secondary cavity- during the study)and Hairy Woodpeckers
(7%) in
nestersmay be particularly vulnerablebecause low stumps(0.3 to 2 m) createdby winter logging.
alsoexthey are unableto excavatetheir own cavities. NorthernFlickersand Hairy Woodpeckers
Nest-sitelimitationmay be a particularproblem in plantationforests,wheresuitable"wildlife trees"(maturedeador dying trees)often
are targeted for removal (Hunter 1990). Additionally,cavity nestersmay be limited by habitat preferences
(Welshand Capen1992)and
competition (Rosenzweig1985). Nest-site selectionby secondarycavity nesterscannotbe
cavatedcavitiesin aspen(14%)at heightsbetween3
m and 12 m. We couldnot determinethe origin of
17% of the excavated cavities. Natural holes created
by fire in the centerof treestumps(12%)were0.3 to
1 m high and were usedfor nestingonly by Mountain Bluebirds.Two speciesof mammalsused the
boxesor cavities(deermouse[Peromyscus
maniculatus] and northernflying squirrel[Glaucomys
sabrinusl).
July1997]
CavityNesters
in Clearcuts
British
445
Columbia
'*•:••Study
FIc. 1. Satelliteimage of the BeaverfootValley study area in the RockyMountains,BritishColumbia.
Clearcutpatches3 to 27 yearsold are highlighted.
Mountain
Bluebirds.--Mountain
Bluebirds
arrived
on the study area in late April, malesabouta week
before females.We found nestsby observingpairs
andby systematically
checkingcavitiesandboxes.In
1993,datacollectionbeganlatein the season,andwe
Chicksspent18 to 22 daysin the nest,and werefed
by bothparents.Wechecked
swallownestswith the
sameprotocolusedfor bluebirdnests.
Habitatdescription.--We
classifiedpatches(clearcuts) into four categories:"very open,.... open,"
were unable to determine whether nests were first at-
"treed," and"youngforest."Veryopenpatches(n =
10) had no vegetationmorethan 1 m tall and mostly
were coveredwith shrubs(but with patchesof grass
and bareground).Openpatches(n = 30) were dominatedby low vegetation(forbsandshrubs),butwith
up to 20%coverof trees1 to 3 m high.Treedpatches
(n = 21) had a predominantcoverof youngtrees(40
to 70%)between1 and3 m high.Youngforestpatches(n = 5) werecomprisedpredominantlyof conifers
natural cavities were checked, but data were not ob3 to 4 m high.Werecordedvegetationcharacteristics
tained for 11 natural nests(25%) that were inacces- in eachpatchalongtwoparalleltransects
150m long
sible due to nest height or tree instability.Arrival and 75 m apart. Transectswere adjacentto nestboxdates of bluebirds were similar in 1994 and 1995. The
es,or naturalcavitieswherepossible,or wereplaced
incubationperiod was 12 days, and chicksleft the randomly in the patch if no cavitieswere present.
nest after 18 to 21 days (see Power and Lombardo Alongthe entirelengthof eachtransect,we recorded
1996).Adults were capturedin a nesttrap designed the predominant vegetationcovering each 1-m
by R. E Holt. We banded68 adults and 130 chicks square (+_ 5%). Categoriesmeasuredwere: bare
with Canadian Wildlife Service numbered alumiground,forbsand grasses,shrubs,saplings<1 m
num legbandsand a uniquecombinationof two col- tall, saplings1 to 2 m tall ("medtree"), and trees >2
temptsor renests.The statusof 86% of 121 nestsin
1994 and 1995was known becausenew pairs were
discoveredbeforethe onsetof laying. We checked
nestseverythreedaysto determineclutchinitiation,
clutchsize,nestfate,fledgingdates,and chickproduction,usingeithera flashlightand a dentalmirror
(for deep cavities)or a small mirror attachedto a
stick(for high boxes).All nestboxesand mostof the
ored leg bands.
m ("talltree"). Arcsine transformationwas used to
normalizeproportionaldata,and factoranalysiswas
erfootValley in the first week of May, 7 to 14 days usedto determinethe principalvegetationvariables
laterthanbluebirds.Theyinitiallyremainedin flocks and to reduce the number of variablesif possible.
and examined many potential nesting sites. Nest The first three factorsexplained98.7% of the varibuilding generallybeganabouta weekbeforeclutch ancein the data.Factor1 explained44.7%of thevariinitiation,and no pairs attempteda secondclutch. anceand wasmade up of forb (varimaxrotatedfacTree Swallows.--Tree
Swallows arrived in the Beav-
446
HOLTANDMARTIN
tor loading = -0.94) and talltree (varimax rotated
factorloading = 0.92). Factor2 (medtree)explained
27.2% of the variance and had a rotated factor load-
[Auk, Vol. 114
for high densitiesof EasternBluebirds(Sialiasialis;
Gowaty and Bridges1991).To standardizethe position of interior boxesin patchesof variablesize,the
cornerbox wasplaced75 m from the patchedgein
all patches.Thecavityentrancediameterwas62mm,
ing -0.98, and Factor3 (shrub)explained27.9%of
the varianceand had a rotated factor loading of
-0.98. The other two categories,
bare groundand which was the median diameter for natural cavities
small trees, were insignificantand were dropped in thestudyareain 1993(Holt 1997).Throughoutthe
from further analyses.The rotatedfactor loadings study,we monitoreduse of natural cavitiesby both
and correlationanalysesindicatedthat forb and tall species.
treeswerehighlynegativelycorrelated(r2 = -0.79,
n = 38, P < 0.0001). Becausethe talltree variable
RESULTS
probablyhas more relevanceto perchinginsectivorousbirds thandoesthe forb variable,we dropped
Naturalcavities
versusnestboxes.--Despite
the
forb from further analyses.Bothmedtreeand shrub
were negativelycorrelatedwith talltree, so these presenceof nestboxes,the numberof birds of
thatnestedin naturalcavitiesdid
datawerenot useddirectlyin furtheranalyses(see bothspecies
Zar 1984).However,the residualsof the regressions not changeduring the study;however,use of
between talltree/shrub
and talltree/medtree
were
nestboxesincreasedwith increasingdensityof
uncorrelatedwith the originaltalltreedataandwere boxes (Table 1). For Mountain Bluebirds, initiusedin furtheranalysesasrepresentation
of factors ationdatesfor first clutches
were statistically
2 and 3.
In all furtherdiscussion,
we use"patchage"asan
equivalentfor "patchsuccessional
age,"andthecategoriesrepresentincreasingsuccessional
agesfrom
very open to youngforest.We used a categoricalexperimentaldesignto addressquestionsof patchoccupationand variationsin breedingsuccess
in relation to patchage.Weusedthe variablestalltree,and
residualsof medtreeand shrub,to predictoccupation and densityof pairsin patchesusingdiscriminant functionand multipleregressionanalyses.StatisticalanalyseswereperformedusingSystatversion
5.0 (1992).
equalfor nestsin boxesand natural sites(nested ANOVA, F = 0.81, df = 1 and 48, P = 0.37),
suggestingthat bluebirdsdid not prefer one
nesttype overthe other.We couldnot conduct
a similar analysisfor Tree Swallowsbecause
only five pairs of swallowsnestedin natural
cavitiesin eachyear.
Measuresof reproductivesuccessdid not
differ between natural cavities and nest boxes,
pooledacrossyears(clutchsize:f = 5.3 + SE
of 0.1 in boxes and in cavities, n = 23 for each;
nests:œ
Experimental
design.--Priorto the 1994 breeding numberyoungfledgedfrom successful
season,eightnestboxeswere addedto eachof five = 4.7 --- 0.4, n = 12 in boxes vs. œ = 4.8 + 0.5,
experimentalpatchesin eachof three patch-succes- n = 13 in natural cavities).Becausenest boxes
sional age categories(very open, open, and treed),
giving a total of 15 experimentalpatches.In 1995,
did not differ from cavitiesin preferenceor in
reproductive success,we combined data in
boxeswere addedto anotherfour patchesunoccupied in 1993and 1994in theyoungforestcategoryto
determinewhether pairs would be excludedfrom
nestingin thisoldersuccessional
stage.Todetermine
whetherfloaterswerepresent,we addedeightboxes
to four patchesthat were unoccupiedin 1993after
subsequent
analyseswhereappropriate.
Nest-siteavailabilityand patchage.--The dis-
the initiation
of first clutches in both 1994 and 1995.
tribution
of natural
cavities
was uneven
with
respectto patchage.Mostof the cavitieswere
in openand treedpatches,with only a few in
youngforestpatchesand nonein very open
patches(Table1). In 1993,prior to the addition
To addressnest-sitepreferencesand variation in
offspring production within patches,four boxes of nest boxes, the distribution of the 23 Moun-
wereplacedon the edgeand four in the interiorof tain Bluebirdand 5 Tree Swallowpairs largely
each patch, with two boxesplaced 1.5 m above reflected the distribution of natural cavities
groundand two placed3 m aboveground.We used
1.5 m for low boxes because this was the median
heightof naturalcavitiesexcavated
in stumpsonthe
study area,and 3 m for high boxesbecausethis was
thetallestboxwe couldeasilyerectandthenmonitor
in large numbers.We erected a total of 120 boxes,
with 60 placedin the interiorand60 on theedgeof
patches.Half of eachcategorywasplacedhigh and
the otherhalf wasplacedlow.Boxeswereplaced75
(Table1). During 1994and 1995,the numberof
bluebird and swallow pairs increasedfollowing nest-boxaddition, and the distributionof
thesepairs expandedin relationto patchage
(Table 1). Both speciesoccupiedvery open
patchesonly after the addition of nestboxes,
and the number
of bluebird
and swallow
nests
increasedtwo-fold and almost eight-fold,re-
m apart basedon mean nearest-neighbor
distance spectively,
in theopenpatches.
Densityalsoin-
July1997]
CavityNesters
in Clearcuts
447
TABLE
1. Numberof pairsof MountainBluebirds
andTreeSwallows,andnumberof clearcutpatches
with
nestingcavities(naturalor nestbox)in relationto seralstageand year.Patches
increasein agefromvery
open to youngforest.
Seralstage(patchage)a
Experimentalpatches
Year
1993
1994
1995
Very
open
Open
Treed
0
5
5
0
6
6
0
3
5
0
6
14
0
8
14
0
6
8
0
5
5
0
5
5
0
5
5
Non-experimental
patches
Young
forest
Very
open
Mountain
0
0
0
Young Total no.
forest
pairs
Open
Treed
19
15
18
4
3
6
0
0
0
3
2
3
2
2
1
0
0
0
22
22
22
11
11
11
4
4
4
Bluebird
0
0
0
23
32
40
Tree Swallow
1993
1994
1995 b
0
0
2
0
0
0
5
24
42
No. of patcheswithnest site
1993
1994
1995
0
0
4
0
0
0
Experimentalpatchescontainednestboxesin 1994and 1995;non-experimental
patchescontainednaturalcavitiesonly.
Distributionof TreeSwallownestsvariedsignificantlywith patchage(seetext).
creasedin treed patches,althoughnot all
patcheswere occupied(Table1). Only Tree
Swallows occupied boxes in young forest
patchesin 1995,and they did not occupyall
availablepatches(Table1). In both 1994and
1995,the distributionof bluebirdpairsdid not
differ from that expectedif pairs showedno
preferencefor patchage (bothyears:Kolmogorov-Smirnovtest, dmax
= 1, n = 14, P > 0.5).
In 1994, Tree Swallow distribution did not dif-
patchage(dma
X= 9, n = 36, P < 0.01),with a
higher frequencyof pairs in very open and
openpatches,and a lowerfrequencythanexpectedin treedandyoungforestpatches
(Table
1).
Patchsharingand intraspecific
competition.-
Numbersof bothspecies
increased
withtheaddition of nest boxes,but within experimental
patches
only23%(1994)and25%(1995)of the
original60 boxeswere occupiedby Mountain
fer from the null hypothesis
of no preference Bluebirds(Table2). In 26 of 28 occupiedexperfor patchage(dm•x
= 1, n 20,P > 0.05).In 1995, imentalpatches(overtwoyears),onlyonebluehowever, swallow distribution varied with bird pair occupied
eachpatchat onetime (TaTABLE
2. Fecunditymeasures(œ+ SE,n in parentheses)
andexperimentalboxoccupation
patterns(no.of
boxesoccupied,no. availablein parentheses)
for MountainBluebirdsand TreeSwallows.
Mountain
Bluebird
1994
Tree Swallow
1995
1994
1995
Fecundity measures
Mean dateof first egg
12 May
19 May
11 June
11 June
Clutch size
5.5 + 0.14 (13)
4.9 ___
0.25 (14)
5.2 + 0.22 (17)
5.2 _+0.06 (19)
Chicksfledged/paira
4.6 +_0.37 (8)
4.8 _+0.26 (7)
4.7 + 0.43 (12)
4.6 + 0.86 (8)
Secondclutches(%)b
4 (32)
No. of edgeboxes
No. of interior boxes
No. of high boxes
0 (60)
13 (60)
7 (30)
0 (40)
0 (24)
0 (40)
0 (60)
20 (47)
15 (23)
5 (24)
0 (60)
36 (44)
22 (22)
14 (22)
Experimental box occupation
No. of low boxes
6 (30)
Forneststhatfledgedat leastonechick.
Proportionof pairs renestingafter a successful
first brood.
0 (60)
16 (60)
8 (30)
8 (30)
448
HOtT ANDMARTIN
[Auk, Vol. 114
TABLE3. Intraspecific
patchsharingby Mountain TABLE
4. Summaryofintraspecific
patchsharingby
Bluebirdsand Tree Swallowsin patcheswith nest
boxesadded(experimental)
and patcheswith natural cavities only (non-experimental).Data are
number of patchesoccupiedby the number of
nestingpairsindicatedin boxhead(i.e.0, 1,2, etc.).
Mountain Bluebirds and Tree Swallows. Only
patcheswith multiplenestsitesareincluded.Sample size (total numberof patchesthat couldhave
contained>1 pair or number of pairs nesting in
patcheswith >1 nestsiteavailable)is shownin parentheses.
No. of pairs in
No. of
No. of pairsin
non-experimental
experimental
patches
patches
4 to
Year
0
1
2
3
Mountain
1993
1994
1995
2
0
12
14
1
1
4
1
2
3
6
patches
with
Year
Mountain
Bluebird
0
0
0
0
7
5
11
0
3
2
0
0
0
1
1
1
1993
1994
1995
1 (8)
5 (24)
4 (29)
0
0
0
1993
1994
1995
0 (2)
6 (19)
14 (20)
2
0
7
1
5
7
1
7
0
0
%
patch
%
6 (13)
13 (30)
12 (37)
46
43
32
0 (2)
13 (24)
35 (41)
0
54
85
Bluebird
12
20
14
Tree Swallow
Tree Swallow
1993
1994
1995
>1 pair
No. of pairs
sharinga
2
4
5
0
0
0
0
0
0
0
31
70
nest-sitelimited, the analysisincludedonly 15
ble 3). We testedthe goodness
of fit of the ob- experimental
and 14non-experimental
patches
serveddistributionof patchoccupation
against with multiplenestsitesoccupiedat leastonce
a Poissondistribution (Zar 1984). In 1994 and during the study.The percentageof patches
1995, the bluebird distribution differed from
sharedby MountainBluebirdpairs remained
random (1994: G = 11.83, df = 1, P < 0.001; relativelyconstantoverthe threeyears(f = 15
1995:G = 11.9,df = 1, P < 0.001).The popu- + 2%;Table4). Thepercentage
of pairssharing
lationmeanandvariancein bothyearsshowed patchesalsowasrelativelyconstant(f = 40 +
that the observed distribution was more dis4%;Table4). Boththepercentage
of TreeSwalpersed than random. A different pattern of low pairssharingpatchesand the percentage
patchsharingwasobservedin the non-experi- of sharedpatchesincreasedoverthe studypementalpatches(Table3). In patcheswith mul- riod (Table 4).
Themedianpatchsizein the studyareawas
tiple nestsites,mostwere occupiedby a single
pair of bluebirds.However,threepatcheswere 22 ha. Patches
with a singlebluebirdpair did
occupiedby two pairs of bluebirds,and one not differ in size(f = 27 +--61 ha) frompatches
patchhad five and six pairs nestingin close with more than onepair (f = 26.2 + 88 ha).
proximity in natural cavitiesin 1994and 1995, Consideringonly patchesthat containedmore
respectively(Table3).
than one nest, patch size was not correlated
Tree Swallowsoccupied43% of the 47 boxes with the maximurnnumberof pairs observed
left unoccupiedafter bluebird settlementin in a patch(MountainBluebird:rs = 0.08, n =
1994 (Table 2). The distribution of swallow 30, P > 0.50; Tree Swallow: rs = -0.26, n = 15,
nestsdiffered from randomin 1994(G = 6.0, df
P > 0.20).
= 2, P < 0.05),and comparisonof the popula-
Nonbreeding
fioaters.--Todeterminewhether
nonbreeding
MountainBluebirds
inhabitedthe
tion mean and variance showed that nests were
more dispersedthan random.In 1995,swal- study area, we added nest boxesto four prelowsoccupied
81% of theremainingboxes,ef- viouslyunoccupiedpatchesin 1994and 1995
fectivelysaturatingthe habitatevenly.
after for all known Mountain Bluebird pairs
We examinedthe influenceof intraspecific had startedincubating.All four patcheswere
competitionon patch occupationas: (1) the occupiedin 1994(onebluebirdpair per patch),
numberof patchesoccupiedby morethanone and 50% of patcheswere occupiedwithin six
pair of eachspecies,
and (2) the proportionof daysin 1995(onepair in eachof two patches).
thetotalnumberof nestingpairsof thatspecies All of the new pairs were unbanded,and no
thatshareda patchwith oneor morepairs(Ta- pair disappearedfrom known nesting sites
ble 4). Becausebluebirds and swallows were duringthis period.Therefore,we assumethat
July1997]
CavityNesters
in Clearcuts
449
thesebirdsconstituted
a floaterpopulation
that and one pair of bluebirds,presumablygiving
not have nested if the extra boxes had
bluebirdsa free choiceof box height.Again,
not beenavailable.This experimentcouldnot bluebirdsexhibitedno preferencefor box
addresswhetherthe TreeSwallowpopulation height(X2 = 0.2,df = 1, P = 0.63),althoughour
was saturated, because most of the swallows samplesizewassmall.
Mountain
had not startedlaying when the experiment Clutchandbroodsize.--Considering
was conducted.
Bluebirdsnesting in experimentalpatches,
would
Proximatefactorslimiting breedingdensity.Analysisof nest-boxoccupationin relationto
mean
clutch size and the mean
number
of
df = 1 and 36, P < 0.0001; medtree: F = 1.8, df
= 1 and 35, P = 0.18; shrub: F = 0.002, df = 1
14,P = 0.75)or patchage(clutchsize:F = 0.39,
cupationstatusfor 92%of thepatches.
Tree Swallowsoccupiedalmostall of the experimentalpatches,so we couldnot compare
occupied
versusunoccupied
patches.
However,
densityof swallowsvariedamongpatches,allowingusto examinetherelationship
between
pair densityand vegetationin a multiple re-
Clutchinitiationdate.--Thetiming of clutch
initiation couldbe influencedby traits of nest-
gression. We used maximum Tree Swallow
patchesin 1994 and 1995. Clutchinitiation
fledglingsfrom successfulnestsdid not vary
patchagesuggested
thatsettlement
in experi- betweenyears(two-wayANOVA;clutchsize:F
mental patcheswas nonrandom (Table 1). = 1.68,df = 1 and 21, P = 0.21;fledglings:F =
Basedondiscriminantfunctionanalysisofveg- 0.24,df = 1 and 11,P = 0.63)or amongpatches
etationstructurewithin occupiedversusun- of differentages(clutchsize:F = 1.16,df = 2
occupied
patches,
themostimportantpredictor and 21, P = 0.33; fledglings:F = 0.24, df = 1
ofpatchoccupation
byMountainBluebirds
was and 11,P = 0.63;Table5). Meanclutchsizeand
the talltree variable. The shrub and medium
mean number of chicksfledgedalso did not
treevariableswerenotsignificantpredictorsof varywith year(clutchsize:F = 0.07,df = 1 and
occupation(univariate tests:talltree F = 22.9, 29, P = 0.79;fledglings:F = 0.11, df = 1 and
df = 2 and 29, P = 0.68;fledglings:F = 0.31, df
and 36, P = 0.9).The overallmodelwassignif- = 2 and 14, P = 0.74; Table 5) for Tree Swalicant (Wilks' lambda = 0.56, F = 8.8, df = 3 and lows.Therewerenosignificant
yearandpatch34,P < 0.001),andit correctlyclassified
theoc- age interactions.
ing birdsor by externalfactorssuchasdateof
snow melt or habitat quality.To determine
whetherclutchinitiationdatewasindependent
in individualpatchesacrossyears,we tested
for a correlation between initiation dates within
densityasthefactorvariableto avoidsampling dates were not correlated between years in
the distribution
before saturation.
Tree Swal-
low densitywas affectedby vegetation(multiple r = 0.72,F = 6.16,df = 3 and 17,P < 0.005);
specifically,
it was inverselyrelatedto the tall
tree variable(partial r = -0.65, t = -4.3, P <
0.001). Shrub and medium tree variableswere
unrelatedto Tree Swallowdensity,and swallow density was predicted accuratelyusing
only the tall tree variablein the overallmodel.
Fourinteriorboxesandfouredgeboxeswere
availablein eachexperimental
patch.Mountain
BluebirdsandTreeSwallowsoccupiedonlyinteriorboxesin both years(Table2). Of the 60
interiorboxes,pairscouldchoosebetweenan
equalnumberof highandlow boxes.Stratifying databy years(Table2), occupation
ofboxes
by Mountain Bluebirdswas unaffectedby box
height (X2 = 0.05, df = 1, P = 0.83), whereas
Tree Swallowswere more prevalentin high
boxes (X2 = 17.7, df = 1, P < 0.0001). In 1994,
six patcheshad only singlepairsof swallows
MountainBluebirds(rs= 0.09,n = 13, P > 0.50)
or in Tree Swallows(rs= 0.20,n = 16,P > 0.20).
We thenuseda two-wayANOVAto determine
that clutch initiation
date in Mountain
Blue-
birds was not significantlyaffectedby either
yearor patchage(year:F = 2.98,df = 1 and25,
P = 0.09;patchage:F = 2.22,df = 2 and25, P
= 0.13). Clutch initiationdatesof Tree Swallowsalsodid not vary with year (F = 0.07,df
= 1 and38, P = 0.78),but theydid vary with
patchage (F = 8.57,df = 2 and 38, P = 0.001;
Table 5). This differencewas due to the later
clutchinitiationof clutchesin very opencomparedwith the open(P = 0.0001)andtreed(P
= 0.04)patches(Tukeytests).Interactions
betweenyear and patchage were not significant
for eitherspecies.
Nestpredation
andproductivity.--Because
nest
predation rates may be influencedby nest
height,we classifiednaturalcavitiesinto low
(<3 m) andhigh(>3 m) nests.Theproportion
450
HOLTANDMARTIN
[Auk, Vol. 114
of nestsdepredateddid not differ by height
(two-way ANOVA; F = 0.003,df = 1 and 2, P
= 0.96) or by year (F = 1.55,df = 2 and 2, P =
0.39).Predationratesat high andlow nestboxes also did not vary with heightclass(F =
0.001,df = 1 and 1, P = 0.99)or with year (F =
0.04,df = 1 and 1, P = 0.87).Includingonly experimentalpatches,predationratedid not differ betweenspecies(three-wayANOVA;F =
0.98, df = 1 and 5, P = 0.37),but did differ with
patchage(F = 41.99,df = 2 and5, P = 0.001)
and year (F = 12.75,df = 1 and 5, P = 0.02;
Table5). Predationrateswere higherin very
open comparedwith open (P = 0.008) and
treed(P = 0.001)patches
(Tukeytests;Table5).
We estimatedpotential productivity using
meanclutchsizeforeachspecies
andmeanpredationprobabilityin eachpatch-age
category.
Potentialproductivitywas2.5 to 3.5chicksper
pairlowerin theveryopencompared
with the
olderpatches
(Fig.2, Table5).
DISCUSSION
Nest-siteavailabilityand patchsuccession.Nest-site availability often limits secondary
cavity nesters(Lack 1954, Scott 1979, Brawn
and Balda 1988, Newton 1994; but see Welsh
andCapen1992).In ourstudyarea,thelackof
cavitiesprecludedTree Swallowsand Mountain Bluebirdsfrom nestingin youngpatches
and limited breeding densitiesin mid-age
patches.Vegetationstructuremay determine
avianspeciespresence
andcomposition
(James
1971,AndersonandShugart1974,Martin 1992,
Hansenet al. 1995),and both speciesdisappearedfromourstudyareaaftermorethan20
years of vegetationalsuccession,
despitethe
presence
of availablenestsites.Patchoccupation may be relatedto a numberof factors,includingchangesin prey abundance
or availabilitywith vegetationstructureandvariation
in theprobabilityof predation(YahnerandMahan 1996).
Patchoccupation
andpatchsharing.---Bluebirds
occupiedabout90% of the availableexperimentalpatches,
but lessthan15%of the pairs
nestedconcurrently
with conspecifics,
resultingin a "onepairperpatch"pattern.Thepresenceof floaterssuggested
thatbluebirdssometimesdeferbreedingratherthannestin close
proximityto conspecifics.
Floaterpairscanbe
presentdespitethe availabilityof apparently
July1997]
CavityNesters
in Clearcuts
451
[] mean
chicks
perpair The existence of floaters allowed us to inter[] mean
pairs
perpatchpret the observeddistributionpatternswith
Mountain Bluebird
more confidence.For instance,in occupied
Very Open
Open
Treed
patchesbluebird pairs were over-dispersed
ratherthan uniformlydistributed,whichsimply maybe a functionof low pair density.However,the presenceof floaterpairs strengthens
the possibilitythat intraspecificcompetition
limitedthenestingdensityofbluebirds.
Floater
pairs have previouslybeen demonstratedexperimentally for Mountain Bluebirds(Power
1975),and in manyamateurnest-boxprograms
bluebird densitiesincreaseimmediately after
theadditionof nestboxes(Newton1994).Floater pairsmaybe a generalphenomenon
in populations that are nest-site limited.
Tree Swallow
Behavioralavoidance
of conspecifics
is much
strongerfor bluebirds,perhapslimiting bluebird density at low levels,whereasswallow
density increased until the available boxes
weresaturated.The differentresponse
in density betweentheseotherwisesimilar species
stresses
the importanceof includingspeciesspecificlife-historyinformationin modelsthat
aim to assess
the implicationsof resourcemanagementon biodiversity (Westman1990, Hansen and Urban 1992, Hansen et al. 1992).
Nest-siteselection.-•Ourexperimentaldesign
allowedpairs to choosebetweenboxesplaced
Treed
Very Open
Open
on the edge versusthe interior of clearcut
patches.
Althoughbothspecies
appearedto be
Habitat Age Category
limitedby the availabilityof nestsites,neither
FIe. 2. Potential productivity (+_ SE) and mean speciesoccupiedboxeson the edgeof experinumberof pairs per patchfor MountainBluebirds mentalpatches.Rendelland Robertson(1990)
and Tree Swallowsin relationto successional
ageof also found that Tree Swallowsavoidededge
patches.
habitat. They explainedthis behavior as an
avoidance
suitablehabitat(Power1975,Stutchburyand
Robertson1985).At low density,swallowswere
over-dispersed,
but both the percentageof
pairssharinga patch,andthepercentof patches shared,increasedthroughoutthe study.
of interference
from
House
Wrens
(Troglodytes
aedon),which nest preferentially
within 30 m of the forestedge.In the same
study,however,69%of EasternBluebirdsnested within 30 m of the edgeand presumably
also lost nests to House Wrens. Unlike other
Muldal et al. (1985) also observedTree Swal- studies(Zarnowitz and Manuwa11985,Rendell
lowsspacingthemselves
throughoutclustersof and Robertson1990, Li and Martin 1991), our
assemblage
had low speciesdinest boxes,but intrapair distancesdecreased cavity-nesting
when densityincreased.Nestingdispersion versityand presumablyhad lowerlevelsof inalso was observed in Eastern Bluebirds that
terspecificcompetitionfor nest sites.An alternestedwith a meannearest-neighbor
distance native explanationfor avoidanceof edges is
of 70 m, eventhoughboxeswere 30 m apart nest-predationavoidance.Predatorabundance
(Gowaty and Bridges1991).Tree Swallowsoc- may increaseat forestedges(seePaton1994),
may avoidthe forestedgeas
casionallynest1 to 2 m apart (Harris 1979),al- andbothspecies
though15 m is a morecommondistance(Ren- a result of predationrisk.
dell and Robertson1990).
Heightof nestcavityaffectsnestingsuccess
452
HOLTANDMARTIN
[Auk, Vol. 114
in some cavity-nestingbirds, apparently be- openhabitatsimplyare making"the bestof a
causeof the increasedpredationin low nest bad job" (sensuKrebsand Davies1987),with
sites (Nilsson 1984,Li and Martin 1991). In our the alternativebeing to foregonestingor to
experimentaldesign,birds could choosebetweenboxes1.5 and 3 m aboveground.Mountain Bluebirdsshowedno preferencefor nestbox heightwithin this range.Tree Swallows,
however,clearlypreferredhighboxes,andall
naturalcavitiesusedby nestingswallowswere
in highstumps(>3 m) or standingaspens(>4
m). The shortageof natural cavitieshigh in
treesmay explainthe smallnumberof nesting
nestin a patchwith conspecifics.
Wecannotrejectthishypothesis,
althoughit seemsunlikely
that the costof patchsharingexceeds
thatof an
swallows before boxes were available. We have
the similar predationrates observedin these
two typesof nests(Holt 1997).The mostparsimoniousexplanationof the apparentdilem-
no evidencethat predationdifferedbetween
low and highboxes,or betweenlow andhigh
naturalcavities.
Theheightrangein ourstudy,
however,may havebeeninsufficientto deter
predatorsevenif nestshigherthan 3 m were
less vulnerable
than lower nests.
87% nest-predationrate.Additionally,it is pos-
siblethat dataobtainedfrom birdsnestingin
nest boxes does not reflect natural predation
rates.In our study,nestboxeswere similar in
heightand cavitydiameterto mostof the natural cavities in the area, which was reflected in
ma betweenpatchoccupation
andpotentialfitnessis that thesebirdshad no a prioriknowledgeof predationriskin thisnovelhabitatand
thus did not includepredationcostsin their
Habitatselection
andconsequences
for reproduc- habitat assessment.
tive success.--Habitat
selectionhas important
Publishedexamplesof speciesthat apparfitnessconsequences
and thereforeis expected entlyare poorlyadaptedto theirenvironment
to be understrongselection(Cody1985).The- are scarce.In a study of nest-siteselectionand
ories that relate habitat-selection decisions to
nestpredationin small drainageson the Colguttahabitatqualityand conspecifics
haveassumed oradoPlateau,Hermit Thrush(Catharus
that individualshave a good knowledgeof tus)and MacGillivray'sWarbler(Oporornis
toltheir habitatand can assessrisk (Fretwelland miei)nestsin smallfirs had lowerreproductive
Lucas 1970, Pulliam and Danielson 1991). Fe-
cundityof bothMountainBluebirdsand Tree
Swallowsdid notvarywith patchage,butproductivitywashighlyskewedasa resultofhigh
nestlingpredationin youngpatches.
Nevertheless, both bluebirds and swallows nested in the
youngpatchesat leastasfrequentlyasin older
patches.This resultis inconsistent
with our expectationthat individualsshouldoptimizefitness.One explanation
is that patternsof patch
occupationare age specific.Young or less-experiencedbirds may be moreprevalentin the
very openhabitat,consequently
loweringproductivity in thesepatches(Wheelwrightand
Schultz1994,Martin 1995).Tree Swallowsinitiatedclutches
laterin very openthanin older
patches,whichgivestentativesupportfor this
hypothesis.However,we also used artificial
eggs to examine patterns of nest predation in
the studyarea,whicheffectivelycontrolled
for
variationin adult quality. Resultsfrom these
experimentsshowedsimilar patternsof nest
predationwith patchage(Holt 1997),suggestingthatthepatternswe observed
did notresult
from age-related
phenomena.
Another explanationis that pairs in very
success(owing to high nest predation)than
nestsin surroundingvegetation(Martin and
Roper 1988, Martin 1993).For both species,
small firs may be more appropriatenesting
sitesin otherpartsof theirrange,andalthough
theyconferredlow reproductivesuccess
onthe
ColoradoPlateau,theirusemaybe maintained
by birdsoriginatingfrom areaswherethe behavior results in high reproductivesuccess
(Martin 1993).A similarphenomenon
may be
occurringwith MountainBluebirdsand Tree
Swallowscolonizingthe novelhabitatcreated
by clearcutting.Both speciesseemedwell
adaptedto colonizingtransienthabitats,but
neitherspeciesappearedto adjustpatch-selection decisionsto encompass
variationin the
patternsof nestpredation.
Traditionally,MountainBluebirdsand Tree
Swallowsinhabitopenhabitatssuchas farmland,parkland,andephemeralopenareasproduced after fires (Zeleny 1976).Bluebirdpopulationsin NorthAmericadeclinedin theearly
half of this century,apparentlyas a resultof
large-scale removal of suitable trees and
stumpsfor nesting(Zeleny1976,Newton1994)
as well as competitionwith introducedEuro-
July1997]
CavityNesters
in Clearcuts
453
pean Starlings (Sturnusvulgaris) and House and EvansLogging Companyprovided maps and
Sparrows (Passerdomesticus).
Nest-box trails emergencyradiosand were cooperativethroughout
were initiated to counterthis declineand ap- thestudy.RudyBoonstra,BobElner,JohnEadie,and
throughpear to have been successfulin many areas MyrnaPearmanprovidedusefulcomments
(Kibler 1969, Backhouse1986, Robertsonet al. out the study. Wendy Easton,Sally Hejl, and an
1992, Power and Lombardo 1996). More recently,bluebirdsand Tree Swallows,aswell as
primary cavity nesters,have colonizedclearcuts (Conner 1974, Bennett 1994), and many
studieshave suggestedmethodologies
to increasethe numbersof cavitynestersin clearcuts(e.g.Conner1978,Schreiber
anddeCalesta
1992,Bennett1994).However,thedemographic
consequences
of nestingin clearcutsareuncertain. The overallhigh predationratesthat we
foundsuggestthat the Beaverfoot
Valleyis less
productivethan it might initially appear,and
it probablyis lessproductivethan mosttraditional habitats. However, the net effect of cre-
atingpatchesof earlyseralstagesmaybe positive for thesespecies.
MountainBluebirdsand
TreeSwallowsproducedoffspringin thestudy
area,despitehigh ratesof nestpredation.The
neteffectof newlycreatedclearcutsonthepopulationdynamicsof thesespecies
will depend
onwhetherless-productive
areasarecolonized
in additionto (or asan alternativeto) moreproductive traditional
habitats.
anonymousreviewergreatly improvedthe manuscript.Leah Macfadyenprovidedexcellentsecretarial services.R.EH. appreciatesthe invaluablesupport of a Commonwealth
GraduateScholarship,
and
researchgrantsfrom the JamesL. Baillie Memorial
Fund,theNorth AmericanBluebirdSociety,andEllis
Bird Farm, Alberta. Natural Sciencesand Engineering ResearchCouncilof Canada,EnvironmentCanada Greenplan, and Canadian Wildlife Service
Grantsto K.M. alsosupportedthis research.
LITERATURE CITED
ANDERSON, S. H., AND H. H. SHUGART, JR. 1974.
Habitat selectionof breeding birds in an east
Tennessee
deciduousforest.Ecology55:828-837.
BACKHOUSE, E
1986. Bluebird revival. A heart-
warmingcomebackfrom thebrink of extinction.
CanadianGeographic106:33-39.
BENNETT, S. N. 1994. Initiatives in wildlife tree man-
agement:an evaluationof thehigh-cutstumping
technique.
M.S. thesis,SimonFraserUniversity,
Burnaby,BritishColumbia.
BRAWN,J. D., AND R. P.BALDA. 1988. Populationbi-
ologyof cavitynestersin northernArizona:Do
nest sites limit breeding densities?Condor 90:
We havedemonstrated
that breedingdensi61-71.
tiesof cavitynestersare limitedby a hierarchy BROWN,J. L. 1969. Territorialbehaviourand popuof species-specific
factors, including nest-site
lation regulation in birds. Wilson Bulletin 81:
availability,vegetationstructure,and intraspe293-329.
cificcompetition.
Consequently,
the additionof BUNNELL,E L., AND L. L. KREMSATER.1990. Sustainnest sitesto managedlandscapeswill not necing wildlife in managedforests.NorthwestEnvironmental Journal 6:243-269.
essarilyresult in globalincreasesof all cavity
nesters.Managementstrategiesshould con- COD¾,M. L. 1985. An introduction to habitat selection in birds. Pages3-56 in Habitat selectionin
sider variation in species-specific
traits to debirds (M. L. Cody, Ed.). AcademicPress,New
terminethe placementand physicalattributes
York.
of wildlife trees(e.g.snags)necessaryto mainCONNER,R. N. 1974. Eastern Bluebirds nesting in
tain productivityin a forestmatrix. The lifeclearcuts.Journal of Wildlife Management 38:
historyconsequences
of wildlife reservepatch934-935.
es and snagsfor cavity-nesting
speciesshould CONNER,
R. N. 1978. Managementfor cavitynesting
be evaluatedbefore embarkingon ambitious
birds.Pages120-128in Managementof southern
and perhapscounterproductive
wildlife enforestsfor non-gamebirds (R. M. DeGraaf, Ed.).
United States Forest Service General Technical
hancementin managedforests.
ACKNOWLEDGMENTS
Report SE-14.
DUNNING,J. B., AND B. D. WATTS.1990. Regional
differencesin habitatoccupancyof Bachman's
Sparrow.Auk 107:463-472.
K. E., AND H. G. MERRIAM.1986. Imporwood, SimoneRunyan, and Lea Elliott for tireless FREEMARK,
tance of area and habitat heterogeneityto bird
field work assistance.Yoho and KootenayNational
assemblages
in temperateforest fragments.BiParksprovidedinvaluablelogisticalsupport,in particular the use of the Ice Creek Cabin in the BeaverologicalConservation36:115-141.
S. 1972. Populationsin a seasonalenvifoot Valley.The Ministry of ForestsOffice(Golden) FRETWELL,
We thank Peter Post, Bruce Bennett, Debbie Well-
454
HOLTANDMARTIN
ronment. Princeton University Press,Princeton,
New Jersey.
FRETWELL,
S., AND H. J. LucAs. 1970. On territorial
[Auk, Vol. 114
and nesting successof cavity-nestingbirds in
high elevationforest drainages.Auk 108:405418.
behaviourand otherfactorsinfluencinghabitat LYNCH, J. E, AND D. E WHIGHAM. 1984. Effects of
distribution
in birds. Acta Biotheoretica
19:16-
36.
GATES,G. E., AND L. W. GYSEL.1978. Avian nest dis-
persionand fledgingsuccess
in field forestecotones.Ecology59:871-883.
GOWATY, P. D., AND W. C. BRIDGES. 1991. Nestbox
forest fragmentationon breeding bird communities in Maryland, U.S.A.BiologicalConservation 28:287-324.
MARTIN,K. 1989. Pairingand adoptionof offspring
by replacement
maleWillowPtarmigan:Behaviour, costsand consequences.
Animal Behaviour
37:569-578.
availabilityaffectsextra-pairfertilizationsand
conspecific
nestparasitismin EasternBluebirds. MARTIN,K. 1995. Patternsand mechanismsfor ageAnimal Behaviour 41:661-675.
dependentreproductionand survivalin birds.
HANNON, S. J., AND E C. ZW•½KEL. 1979. Probable
non-breedersamong female Blue Grouse.Con-
AmericanZoologist35:340-348.
MARTIN,t. E. 1992. Breedingproductivity consid-
erations:What are the appropriatefeaturesfor
dor 81:78-82.
HANSEN, A. J., W. C. MCCOMB, R. VEGA, M. G. RA-
PHAEL,AND M. HUNTER. 1995. Bird habitat re-
lationshipsin natural and managedforestsin
the WestCascades
of Oregon.Ecological
Applications 5:555-569.
HANSEN, A. J., AND D. L. URBAN. 1992. Avian re-
management?Pages 455-473 in Ecology and
conservationof Neotropicalmigrant landbirds
(J. M. Hagan, III and D.W. Johnston,Eds.).
SmithsonianInstitutionPress,Washington,
D.C.
MARTIN,t. E. 1993. Nest predation and nest sites:
New perspectives
onold patterns.BioScience
43:
523-532.
sponses
to landscapepattern:Theroleof species
MARTIN,t. E., ANDJ.J. ROPER.1988. Nest predation
life histories.Landscape
Ecology7:163-180.
HANSEN, A. J., D. L. URBAN, AND B. MARKS. 1992.
and nest site selectionof a westernpopulation
of Hermit Thrush. Condor 90:51-57.
Avian communitydynamics:The interplayof
landscape
trajectories
andspecies
life histories. MULDAL, A., H. L. GIBBS,AND g. J. ROBERTSON.1985.
Pages 107-195 in Landscapeboundaries:ConPreferrednest spacingof an obligatecavitynestingbird, the Tree Swallow.Condor87:356sequencesfor biotic diversity and ecological
flows. (E diCastri and A. J. Hansen, Eds.).
363.
NEWTON,I. 1994. The role of nest sitesin limiting
Springer-Verlag,
New York.
the numbersof hole-nestingbirds:A review.BiHARRIS,
R. N. 1979.Aggression,
superterritoriesand
ologicalConservation70:265-276.
reproductivesuccessin Tree Swallows.CanaNILSSON,S. G. 1984. The evolution of nest-site sedian Journalof Zoology57:2072-2078.
lectionamonghole-nestingbirds: The imporHOLT,R. E 1997.Habitatselection,
demography
and
conservationimplications for a cavity-nesting
tanceof nest predationand competition.Ornis
Scandinavica 15:167-175.
communityin a managedlandscape.
Ph.D.thePATON,P. W. C. 1994. The effect of edge on avian
sis,Universityof Toronto,Toronto,Ontario.
HUNTER,M. L. 1990. Wildlife, forestsand forestry:
nestsuccess:
How strongis the evidence?ConservationBiology8:17-26.
Principlesof managing forests.Prentice-Hall,
EnglewoodCliffs, New Jersey.
POWER,
H. W. 1975. Mountain Bluebirds:Experimental evidenceagainstaltruism. Science189:
HUTTO,R. L. 1995. Compositionof bird communi142-143.
ties followingstand-replacement
fire in northern Rocky Mountain (U.S.A.) conifer forests.
ConservationBiology9:1041-1058.
JAMES,E C. 1971. Ordination of habitat relation-
POWER,H. W., AND M.P. LOMBARDO. 1996. Moun-
shipsamongbreedingbirds.WilsonBulletin83:
tain Bluebird(Sialiacurrucoides).
In The birds of
North America, no. 222 (A. Poole and E Gill,
Eds.). Academyof Natural Sciences,Philadel-
215-236.
phia, and American Ornithologists'Union,
KIBLER,J. R. 1969. The establishmentand mainte-
Washington,D.C.
nanceof a bluebirdnest-boxproject.A review PULLIAM,H. R. 1988. Sources,sinksand population
and commentary.Bird-Banding40:114-129.
KREBS,J. R., AND N. B. DAVIES. 1987. Alternative
strategies.Pages153-169 in An introductionto
animalbehaviour,2nd ed. (J.R. KrebsandN. B.
Davies, Eds.). Blackwell ScientificPublications,
London.
sinks and habitat selection:A landscapeperspective on population dynamics. American
Naturalist
137:S50-S66.
RATTI,J.t., ANDK. P. REESE.1988. Preliminarytest
LACK,D. 1954. The natural regulationof animal
numbers. Clarendon Press,Oxford.
El, P., AND t. E. MARTIN.
regulation.AmericanNaturalist 132:652-661.
PULLIAM,H. R., AND B. J. DANIELSON. 1991. Sources,
1991. Nest-site selection
of the ecologicaltrap hypothesis.Journalof
Wildlife Management52:484-491.
RENDELL, W. B., AND R. J. ROBERTSON.1990. Influ-
July1997]
CavityNesters
in Clearcuts
enceof forestedgeon nestsite selectionby Tree
Swallows.
Wilson Bulletin
102:634-644.
ROBERTSON,
R. J., B. J. STUTCHBURY,AND R. R. CO-
HEN. 1992. Tree Swallow (Tachycineta
bicolor)In
The birds of North America,no. 11 ( A. Poole,P.
Stettenheim,andE Gills,Eds.).Academyof Nat-
455
ing density of kestrels.Journalof Animal Ecology 52:635645.
WELSH,C. J.E., ANDD. E. CAPEN.1992. Availability
of nestingsitesas a limit to woodpeckerpopulations.ForestEcologyand Management48:3141.
ural Sciences,
Philadelphia,and AmericanOr- WESTMAN,W. E. 1990. Managing for biodiversity.
BioScience 40:26-33.
nithologists'
Union,Washington,
D.C..
N. t., ANDC. B. SCHULTZ.1994. Age
ROSENZwEIG,
M. L. 1985.Sometheoreticalaspects
of WHEELWRIGHT,
and reproductionin SavannahSparrowsand
habitatselection.Pages517-540in HabitatselecTree Swallows.Journalof Animal Ecology63:
tion in birds (M. L. Cody,Ed.). AcademicPress,
New York.
SCHREIBER,
B., AND D. S. DECALESTA.1992. The re-
686-702.
W•LCOVE,
D. S. 1985. Nest predationin foresttracts
andthe declineof migratorysongbirds.
Ecology
lationship between cavity-nestingbirds and
66:1211-1214.
snags on clearcutsin western Oregon. Forest
YAHNER,R. H., ANDC. G. MAHAN. 1996. DepredaEcologyand Management50:299-316.
tion of artificialgroundnestsin a managedforSCOTT,
V. E. 1979. Bird responseto snagremovalin
estedlandscape.ConservationBiology10:285ponderosapine. Journalof Forestry77:26-28.
STUTCHBURY,
B. J., AND R. J. ROBERTSON.
1985. Float-
ing populationsof femaleTree Swallows.Auk
102:651-654.
SYSTAT,
INC. 1992. Statistics,version5 edition. Systat Incoporated,Evanston,Illinois.
TERBORGH,
J. 1992. Why American songbirdsare
vanishing.ScientificAmerican266:98.
VAN HORNE,B. 1983. Density as a misleading indicator of habitat quality. Journalof Wildlife
Management 47:893-901.
VICKEY, P. D., M. L. HUNTER, AND J. V. WELLS. 1992.
IS densityan indicatorof breedingsuccess?
Auk
109:706-710.
VILLAGE,g. 1983. The role of nestsite availability
and territorialbehaviouron limiting the breed-
288.
ZAR,J.H. 1984. Biostatistical
analysis,2nd ed. Prentice-Hall, EnglewoodCliffs, New Jersey.
ZARNOWITZ,J. E., AND D. A. MANUWAL. 1985. Ef-
fects of forest managementon cavity nesting
birds in N.W. Washington.Journalof Wildlife
Management 49:255-263.
ZELENY,
L. 1976.Thebluebird:How youcanhelpits
fight for survival.AudubonNaturalistLibrary,
Indiana University Press,Bloomington.
ZW•CKEL,E C., AND J.E BENDELL.1972. Blue Grouse,
habitat and populations.Pages150-169in ProceedingsXV InternationalOrnithologicalCongress(K. H. Voous,Ed.). The Hague,1970.E. J.
Brill, Leiden, The Netherlands.
Associate Editor: T. E. Martin
Download