Download effects of olfactory cues on artificial

The Auk 111(4):945-952, 1994
EFFECTS
OF OLFACTORY
ARTIFICIAL-NEST
CUES ON
EXPERIMENTS
CHRISTOPHERJ. WHELAN, MICHAEL L. DILGER,DAVE RONSON,
NATHALIE HALLYN, AND STEVEN DILGER
The Morton Arboretum, Lisle, Illinois 60532, USA
ABSTRACT.--We
examined the effectsof two potential olfactory cueson the outcome of
experimentsusing artificial neststo assaypredationon open-cupnestingsongbirds.In two
experimentslasting 15 dayseachand replicatedtwo timesat three sites,bamboonestsbaited
with Japanese
Quail (Corturnixjaponica)
eggswere placedon the groundin a 5 x 4 grid array
with 50 m between adjacentnests.Egg survivorshipwas monitored every five days, and
remote cameraswere used to take photographsof animalsremoving bait eggsfrom nests.
One experimentconsistedof four treatmentsthat varied the type and amount of olfactory
informationprovided by the investigator:(1) human scent;(2) "no scent";(3) "deer scent";
and (4) perfume. Artificial scentswere applied to shoes,clothing, and skin during the experiment set-upand monitoring.For the first experiment,ratesof nestlosswere greaterfor
the human-scent treatment than deer-scent and/or no-scent treatments in one-half of the
spatial/temporalreplicates.Rain appearsto have complicatedthe effectsof the scenttreatments in some of the spatial/temporal replicates.Final levels of nest loss, however, were
greater for perfume and human-scenttreatmentsthan deer-scentand no-scenttreatments
regardlessof spatial/temporalreplicate.The secondexperimentconsistedof two treatments,
one in which eggswere replacedperiodically,and the otherin which eggswere not replaced
for the duration of the experiment.Both treatmentshad similar ratesand final levels of nest
loss.Approximately90%of 51 photographs
were of olfactory-searching,
predominantlynocturnal mammals.We concludethat olfactory-searching
predatorscan cue on human odors
left in the areaof artificialnests,but that either rain or scents(e.g. the deer scent)canbe
usedto decreasethis potentialbias.In contrast,changingegg qualitiesover the time scale
of our experiments
do notprovideadditionalcuesto predators.
Received
20April1992,accepted
25 November 1992.
ALTHOUGH
NESTPREDATION
haslong been recognized as a major sourceof reproductive failure in bird species(e.g. Ricklefs 1969), only recently hasnestpredation been hypothesizedto
be an important determinant of bird community structure (Martin 1988a, b). Predation on
eggsand nestlingsmay regulate bird populations (George 1987),influence habitat selection
(Blancherand Robertson1985), and affect patternsof nestdispersionand speciescoexistence
(Martin 1988a,b). Perhapsmore importantly,
nest predation also may be important for the
conservationof many bird species(Ambuel and
Temple 1983,Wilcove 1985).In the increasingly
fragmentedlandscapesof North America, nest
predation may be an important contribution to
population declinesof many open-nestingbird
species,particularly long-distanceNeotropical
migrants(e.g. Terborgh 1989).
One major methodologicalproblem in studies of nest predation is finding enough nestsof
a given bird speciesto have meaningful sample
sizesand to control possiblyconfoundingvari-
ables (e.g. see Reitsmaet al. 1990). Largely becauseof these methodologicaldifficulties, investigatorshave been using artificial nestsbaited with quail eggsand have measuredegg survivorship over someperiod of time simulating
a normal incubationperiod (e.g.Anderssonand
Wiklund 1978, Loiselle and Hoppes 1983, Wilcove 1985, Martin
1987, Yahner
1989, Reitsma
et al. 1990).Despite the growing useof artificial
nests and the admission
that some results from
using thesenestsdo not mimic resultsfrom real
nests(e.g.Reitsmaet al. 1990,S.A. Temple pers.
comm.), few studies have examined in detail
methodologicalissuesinvolving the use of artificial nests. Martin (1987), for instance, reported that artificial wicker-basket nests had
lower rates of nest predation than artificially
baited real nests when both nest types were
placedoff the ground,but that artificial and real
nestshad similarratesof predationwhen placed
on the ground.He concludedthat the off-ground
nestsare found primarily by visual-searching
predatorsandthatthesepredatorspossiblyavoid
945
946
WHELAN
ETAL.
artificial nests(T. E. Martin pets. comm.);also,
he judgedthat groundnestsare found primarily
by olfactory-searchingpredators who do not
respond to nest appearance.
Other investigatorshave questioned the extent to which olfactory-searchingpredatorsfind
artificial nests,thus questioning the biological
relevance of the results of experiments using
artificial nests(Willebrand and Marcstr6m 1988).
If artificial nestsare seldomfound by olfactorysearchingpredators,their reflection of predation pressureson real nestsmay be minimal.
When a large olfactory-searching
predator(e.g.
a raccoon,Procyonlotor) finds a songbird nest,
there is little possibilityfor the parentsto drive
[Auk, Vol. 111
the Elburn ForestPreserveis surroundedby agricultural fields (mostly corn or soybeans).Olfactorysearchingpredatorsat these sitesinclude the southern flying squirrel (Glaucomys
volans),Virginia opossum (Didelphismarsupialis),
raccoon,coyote (Canislatrans),red fox (Vulpesvulpes),and stripedskunk (Mephitismephitis),
as well asdomesticdogs(Canisfamiliaris) and cats (Feliscattus).
Effectof humanscenttrails (experiment/).--Human
scentassociated
with artificial nestsand eggscould
serveasan olfactorycuethat couldlead to extremely
biasedestimatesof nestpredation(Reitsmaet al. 1990).
This experiment consisted of four treatments de-
signedto differ in the extentof olfactoryinformation
supplied to potential predators. Each treatment con-
sistedof 20 bamboo,wicker nestsarrayedon a 5 x 4
grid system,with 50 m betweenadjacentnests(den-
offthepredator(seePettingill 1976).Evensmall sity 6.67 nests/ha). Eachnestwas lined with leaf litter
songbirds,however, can sometimesdrive away
small, visual-searchingpredatorslike Blue Jays
(Cyanocittacristata) and eastern chipmunks
(Tamiasstriatus;C. J. Whelan pers. obs., R. T.
Holmes pets. comm.).
Artificial nestsprovide an attractive, nondestructive test system for examining nest pre-
and placed in a hole in the ground with the top of
the nest flush with the soil surface. Nests were baited
with two JapaneseQuail (Corturnixjaponica)eggs,and
were examinedfor eggsurvivorshipat days5, 10,and
15 following placementin the field, conformingto a
monitoring scheduleused by Reitsma et al. (1990).
The experiment was conducted twice, once in late
May and once in early July. Scenttreatmentswere
dation. While use of artificial nests remains conswitched among grid locations at each site for each
troversial, they do allow experimental control
time replicate. A single replicate of each treatment
over numerous confounding factors like nest was conductedsimultaneouslyat each experimental
shape,height,density,positionwithin a plant, site, for a total of three replicatesof each treatment
and number of eggs.Therefore,the usefulness both times the experiment was conducted.
of artificial nests in understanding the imporTreatments consistedof: (1) human scent; (2) "deer
tance of nest predation needs to be evaluated. scent" (Tink's # 69 Doe-in-Rut Lure); (3) "no scent"
Here we report the resultsof experimentsde- (ScentShield and Scent Walker• Boot Pads);and (4)
signedto test specificallythe potentialfor olfactory-searchingnest predatorsto find artifi-
perfume (Primo). The human-scenttreatment served
(human-scenttrails and egg age/odor).
decrease
as a basisfor comparisonwith previousstudiesand
was
establishedby putting nestsout on the grid in a
cial nestswhen thesenestsare carefully hidden
on the ground under vegetation.We used two standardfashion, walking from grid intersectionto
grid intersection,digging a slight depressionin the
complementaryexperimentaltestsystemsthat: ground, placing the nest in the hole, lining the nest
(1) clearly indicatedwhether or not olfactory- with leaves,and baiting the nestwith two quail eggs.
searchingpredatorsfind artificial nests;and (2) Following Reitstoaet al. (1990), we rubbed our hands
examinedtwo possiblemethodologicalbiases with leaf litter prior to handling nests or eggs to
STUDY SITES AND METHODS
Experimentstook place from May through August
1991at three separatesitesabout 25 to 65 km west of
Chicago, Illinois: (1) the Morton Arboretum (Lisle,
Illinois); (2) Hawthorn Hill (WoodridgePark District,
Woodridge, Illinois); and (3) the Elburn ForestPreserve (Elburn, Illinois). All three sites consistedof
the amount
of human
scent left on or near
the nest.Deer scentsupposedlyconsistsof white-tailed
deer (Odocoileus
virginianus)
doe hormones,bodily secretions, and urine (but see Trost 1989); it can be
simulatedusingammoniaand water (Trost1989).Scent
Shield and Boot Pads (used in the no-scent treatment)
are designed to eliminate or to mask human scent.
For each nonhuman-scent treatment, shoes, clothes,
and skin were treatedwith the animalscent,masking
scent,or perfume. Otherwise, nestswere treated simisolated tracts of eastern deciduous forest dominated
ilarly to the human-scent treatment. Becauseof seby variousoaks(Quercus
spp.),sugarmaple(Acersac- verely degraded habitat, part of the Hawthorn Hill
charurn),white ash (Fraxinusamericana),and a variety site was unusable. Therefore, at Hawthorn Hill we
of shrubsand understorytree species(e.g. ironwood, used only three scent treatments:human scent, no
Ostryavirginiana).
The Morton Arboretumand Haw- scent,and perfume.
Two types of complementarydata were collected
thorn Hill are surrounded by suburban homes,but
October
1994]
Olfactory
Cues
andArtificial
Nests
and analyzed:ratesof egg survivorship(i.e. speedat
which eggsdisappeared)over the 15-daysimulated
947
electricallyactivateda solenoidmountedon the camera. When a bait egg was disturbed,the shutterwas
"incubation period," and the final level of overall
released.Cameraswere placedalongtransects
at the
predation(i.e. numberof eggsremainingat lastcen-
Arboretum and at Elburn, but not at Hawthorn Hill.
sus). Rates of predation at each site for each time
replicatewere analyzed with survival analysis(Be-
The cameraswere not placed on any of the experimentalgrids.Instead,transects
were establishedaway
nedetti et al. 1990). Final levels of predation (number
fromthe experimental
gridsusingoneof threescent
of nestssurvivingatendof 15-dayincubationperiod)
were analyzedwith a log-linearmodel (Brown 1990)
treatments(human scent,no scent,or perfume), and
five cameraswere placed along each of these tran-
that examinedsimultaneouslythe effectsof time rep-
sects.Cameraswere moved5 m or morefollowing a
licate, site, and scent treatment on the number of nests
depredation event.
surviving(for an exampleof thesestatisticalanalyses,
see Whelan et al. 1991).
The expectationfor this experimentwas that if viRESULTS
sual-searching
predatorsare the major predatorson
artificialnests,then ratesand final levelsof predation
Experiment1.--In the first time replicate there
shouldnot differ amongtreatments.We assumedthat were significantdifferencesin ratesof nest loss
olfactory-searching
predatorslike raccoons
are more amongscenttreatmentsonly at the Elburn Forlikely to associatethe presenceof food with humans est Preserve. At this site, rates of nest loss inthan with deer. Consequently,if thesepredatorsare
more important, we assumedthat the human-scent
andperfumetreatments
shouldhavegreateramounts
of predation than the deer-scentand no-scenttreatments.
Effectof egg age and odor(experiment2).--Unrefrigerated eggs that are not incubated can develop a
rather strongodor with increasingage.Therefore,it
is possiblethat the longer baited eggs survive, the
greaterwill be their susceptibilityto olfactory-searching predators(a concoctionincluding rotten chicken
eggsis a frequent bait usedby mammaltrappers;Joel
S. Brown pers.comm.).Egg susceptibilityshould not
changewith egg age for visual-searchingpredators.
Experiment2 testedwhether egg susceptibilityin-
creasedfrom deer scent,to perfume, to no scent,
and to human scent (Gehan-Wilcoxon test, X 2
= 8.79, df = 3, P = 0.032;Fig. 1). In the second
time replicate, ratesof nest lossdiffered significantly amongthe scenttreatmentsat all three
sites. Differences
the Arboretum
in rates of nest loss at both
and Hawthorn
Hill
were close
to our expectationif olfactory-searching
predatorswere primarily responsiblefor nest losses.
At the Arboretum, rates of nest loss increased
from deer scent, to no scent, to human scent,
and to perfume(Gehan-WilcoxonX 2= 12.52,df
= 3, P < 0.006;Fig. 2). At Hawthorn Hill, rates
creases
asa functionof eggage.The experimentcon-
increased from no scent, to human scent, and
sisted of two treatments, each of which had the same
to perfume (Gehan-WilcoxonX 2 = 7.415, df =
2, P < 0.03; Fig. 2). At the Elburn Forest Preserve, rates of nest loss differed among scent
treatments,but not in the way predicted if olfactory-searchingmammalsare the important
predatorson these nests.Rates of nest loss at
basic design as experiment 1 and was replicated at
each experimental site. In the no-replacementtreatment, nests were baited on the first day of the experiment, and surviving eggswere not replacedfor
the entire 15-day incubation period. In the replacement treatment, surviving eggswere replaced with
fresheggson eachcensusday.Experiment2 wasconductedtwice in a staggeredfashionwith experiment
1. Statisticalanalyseswere the sameasfor experiment
1. In experiment2, if visual-searchingpredatorsare
the most important predatorson artificial nests,we
expectedno differencein ratesand levels of predation
between the two treatments.If olfactory-searching
predators are the most important predators on artificial nests,we expectedgreater rates and levels of
predationon the no-replacementtreatmentneststhan
the replacementtreatment nests.
Remote-camera
system.--In addition to the aboveexperiments, remote cameras modified from Picman
(1987; M. L. Dilger et al. unpubl. manuscript)were
usedto obtainpicturesof predatorstaking eggsfrom
the artificialnests.Eachcamerajnestsetupconsisted
of a microswitch
attached
to an artificial
nest that
Elburn
increased from human
scent, to deer
scent and perfume, and to no scent (GehanWilcoxon X 2 = 8.789, df = 3, P < 0.04; Fig. 2).
We found that on average49.5%of nests(range
31-75%) in all treatments were chewed, were
moved, or disappeared.
The log-linear model that best fits the multidimensionalcontingencytable of time replicate, experimental site, scent treatment, and
number of nestssurviving included two interactions of main effects (G = 32.41, df = 26, P =
0.180). The interaction of scent treatment and
number nestssurviving (G = 12.12,df = 3, P =
0.007) indicates that final survivorship decreased from deer scent, to no scent, to human
scentand perfume (Fig. 3). The three-way in-
948
WH•LANEl'AL.
lOO
[Auk, Vol. 111
teractionof time replicate,experimentalsite and
number nestssurviving (G = 43.72, df = 2, P <
Arboretum,
Rep
1
0.001) indicates that the levels of overall mor-
8o
Deer Scent
No Scent
HumanScent
8o
4o
2o
i
i
•
5
lO
15
lOO
8O
8O
4O
2O
Elburn,
Rep I
tality differed amongthe sitesdepending upon
time replicate (Fig. 4). For instance,at the Arboretum, survivorship decreasedmoderately in
the secondtime replicate.At Elburn, survivorship decreaseddramaticallyin the secondtime
replicateand, at Hawthorn Hill, there was virtually no difference in survivorship between
time replicates.
Experiment2.--There were no differences in
rates of nest lossbetween the no-egg-replacement and egg-replacement treatments at any
experimental site in either time replicate. The
log-linear analysis of the multidimensional
contingency table of experimental site, time
replicate, replacement treatment, and number
of nestssurviving showed no significantmain
effects,nor any significantinteractionsof main
effects.Indeed, virtually no nests survived at
each site for each time replicate. On average
46.24%of nests(range43-80%) in all treatments
were chewed, were moved, or disappeared.
Remote-camera
photographs.--Fifty-onephotographswere taken showing eight different animal species.Most photographs (36) were of
raccoons,regardlessof scenttransect.In several
of thesephotographs,up to three different raccoons could be seen in the frame. The next most
0
I
I
5
10
15
IO0
8O
8O
4O
2O
i
i
ß
5
10
15
Day
abundantanimal photographed(7) was the Virginia opossum.Photographsof two other mammals,the easternchipmunk and the white-tailed
deer, were taken only on the perfume transect.
Although the eastern chipmunk is primarily a
visual-searchingpredator,it is alsopossiblethat
it sometimeslocatesfood using olfaction (D.
Jedlicka pers. comm.). Whether the deer was
actuallypreying uponthe egg,or simplytripped
the cameraaccidentally,is not evident from the
photograph. The three photographsof a nonolfactorypredator,the Blue Jay,were associated
with the no-scenttransects.Three of the photographs are of animals that probably tripped
the cameras accidentally, an American Robin
(Turdusmigratorius),a Black-cappedchickadee
(Parusatricapillus),and an unidentified bird.
Fig. 1. Percent nestssurviving over 15-day simulated incubationperiod for nestsin four scenttreatments for first-time replicate of experiment (n = 20
nests/treatmen0.
October1994]
Olfactory
CuesandArtificial
Nests
lOO
o•
Arboretum,
Rep 2
6o
3O
;>
8o
•
--m--o-+
949
Deer
Scent
No Soent
Human
Scent
Perfume
•
20
c
lO
a.
o
I
4o
1
Deer
r
•
No Scent
I-
Human
1
[--]
Perfume
Fig. 3. For different treatments,percentnestssurviving at completionof eachtime replicatefor both
time replicatescombined(n = 60 nests/treatment,except for deer-scenttreatment in which n = 40 nests).
2o
o
,5
lO
15
DISCUSSION
lOO
Our resultsindicatethat olfactory-searching
Elburn, Rep 2
mammalswere the primary predatorson arti-
ficial nestsbaitedwith quail eggs.In experiment1 for one-halfof the site/timereplicates,
m
60
•'
40
•
the ratesof predationwere greatestfor nests
associated
with human or perfume scent,and
least for nests associated with no scent or deer
scent. Nests in the deer-scent treatment had al-
mosttwicethe final survivorshipasthosein the
humanand perfumetreatments.Finally,about
20
90%of the photographs
of animalsattacking
artificial nestswere of olfactory-searching
mammals.Of these,the raccoonis clearlythe
0
5
10
15
lOO
most important.
In contrast,therewere no differencesamong
treatmentsin experiment2. Accordingto original expectations,this would have been inter-
pretedasevidenceagainstolfactory-searching
predatorsplaying a majorrole in depredation
of artificialnests.However,giventhe resultsof
the remote cameras,coupledwith the large
8o
number of neststhat were chewed, were moved,
6o
or disappeared
in bothexperiments
1 and2, our
originalexpectationfor this experimentseems
mistaken.
If changingodorhasnoeffecton egg
susceptibility,
or if egg odor doesnot change
duringthe 15-daysimulated
incubation
period,
then there should have been no great differences between the two experimentaltreat-
4o
2o
0
•
[
ß
5
10
15
Day
Fig. 2. Percentnestssurvivingover 1S-daysimulatedincubationperiodfor nestsin four scenttreat-
mentsfor second-time
replicateof experiment.Symbol for perfumescentis hidden by symbolfor no
scentin that their survival curveswere identical (n
= 20 nests/treatment).
950
W•EL*NL•r^L.
60 I
4O
[Auk,Vol. 111
Thus, it is plausiblethat the confusingresults
of the first time replicateand the Elburn study
site in the secondtime replicatecan be attributed to confounding effectsof rain.
If our resultsare interpreted to indicatethat
olfactory-searching
predatorsare morelikely to
find artificial neststhan visual-searchingpredators, it may be argued that these nests adequately assayfor effectsof olfactory-searching
-1
predators.However, our findings also suggest
that olfactory-searchingpredatorscan use olfactoryinformation suppliedby the investiga1
2
1
2
1
2
tor and, thus, results from these experiments
Arboretum
Elburn
Hawthorn Hill
mayactuallyoverestimatethe importanceof olpredators.There may be two
Fig. 4. Percentnestssurviving accordingto both factory-searching
experimentalsite and time replicate(n = 80 nests. ways to prevent suchoverestimation.First, be[time replicate]-•.[site]-z exceptHawthorn Hill, for
which n = 60 nests.[time replicate]-•.[site]-•).
cause rain obscures or confounds
whatever
ol-
factoryinformationis left in the neighborhood
of the artificial nest by the investigator,one
possibleway to decreasethe ability of olfactoryments, regardlessof whether predation is by searchingpredatorsto cue in to that informaolfactory-searchingor visual-searchingpreda- tion is to set up the experimentduring a raintots.
storm. Second,in dry-weather conditions,anOne striking result is the amount of vari- imal or shielding scents(asused in our experability in the rate of nest depredationamong iments)couldeffectivelymaskthe odor left by
the sites and between time replicates for ex- the investigator.
periment 1. A likely explanation for this variPerhapsthe most important implication of
ability seemsto be related to weather condi- our findings is that uniform methodology is
tions, especiallythose conditionsat the times essential for results of different studies to be
the experimentswere setup. At the time when comparable(e.g.Knight and Temple 1986).Difthe first time replicatewassetup, northeastern ferent methodologies(e.g. frequency of nest
Illinois (including each of the three sites) ex- monitoring) are likely to introduce different
perienceda large, regional rain storm.The veg- suitesof biases,making comparisonsproblemetation and ground at eachsite were extremely atic. In the future, if someinvestigatorsfollow
wet both when the nestswere placed on the our lead in using animal or shielding scentsto
grid and, again, when they were baited with reducethe effectsof investigatorodor,they will
eggs.It seemspossiblethat the rain decreased need to use cautionwhen comparingtheir rethe effectivenessof the scentsapplied to shoes, suits with studies that used no scents.
clothing, and skin.
Justhow olfactory-searching
predatorsarereThis interpretationis supportedby the results spondingto the scenttreatmentsof experiment
of the secondtime replicate.Two of the sites 1 is not clear,but two alternatives(not mutually
(the Arboretum and Hawthorn Hill) remained exclusive) seem likely possibilities.First, the
dry during the courseof the experiment. The predatormay detecta trail of scentleft by the
Elburn ForestPreserve,however, experienced investigator, learn to associatethis scent with
localizedbut extremelysevererainstorms,both an egg reward and, thus, "trapline" from nest
when the nestswere placed on the grids and to nest. Second, the scent treatments we used
when they were baited with eggs.At the two may simply increasethe "radius of detection"
siteswith dry-weather conditions,ratesof nest of a nest (i.e. the extent of the area around a
loss were least for the deer-scent and no-scent
nest associatedwith a particular scentcue). If
treatments, and greatest for the human-scent predatorslike raccoonsare attractedto unusual
and perfumetreatments.In contrast,at Elburn, scents(neophilia rather than neophobia),the
which was deluged by rain, there were signif- greater the radius of detection, the higher the
icant differencesamong the scent treatments, likelihood of nest depredation.Casualinspecbut not in a way that is easily interpretable. tionsof the temporaland spatialpatternsof nest
October
1994]
Olfactory
Cues
andArtificial
Nests
lossesin theseexperimentssuggestthat, in our
systems,the secondalternative is the more like-
ly. Determiningwhich, if either,of thesealternativesis correctmaybe an importantnext step
951
ing scents,and RonaldDilger for donatingcameras
fordocumenting
nestpredators.
Quaileggsweresupplied by Henry MarksandVivian Pattenof the SoutheastPoultryResearch
Laboratoryof the Agricultural
Research
Service,SouthAtlanticArea, U.S. Department of Agriculture.We alsothank the Woodridge
in understandingmethodologicalintricaciesof
the experimentaltechnique.
Park District and the Forest Preserve District of Kane
Our experimentsindicatethat it maybe pos- Countyfor permissionto work on their grounds.This
sibleto vary experimentallythe typesof olfac- work wassupportedby a FrankM. ChapmanGrant
tory cuesassociated
with nestsand, thus,begin from the AmericanMuseumof Natural History to
to determine which cuesvarious predators are
C.J.W.,a Steenbock
Grantfromthe WisconsinSociety
at least potentially able to use when finding for Ornithologyto M.L.D., and the Emily Rodgers
nests.Experimentalmanipulationof visualcues, Davis Endowment to the Morton Arboretum.
as well as various combinations of visual/olfac-
tory cues,alsomay be possible.Suchmanipulationscould reveal much about the sensory
LITERATURE CITED
stimuli importantto the success
of nestpredators,andthisinformationcouldin turn suggest AMBUEL,B.,ANDS. A. TEMPLE.1983. Area-dependent
changesin the bird communitiesand vegetation
habitatmanipulationsthat coulddecreasetheir
of southernWisconsinforests.Ecology64:1057impacton cup-nestingbird species.
1068.
Investigators
studyingaviannestingsuccess ANDERSSON,
M., AND C. G. WIKLUND. 1978. Clumphavelong been concernedthat their presence
ing versusspacingout: Experimentson nest preand monitoringof nestscould have negative
dation in Fieldfares(Turduspilaris).Anim. Behav.
26:1207-1212.
consequences
for individualbirdsunderstudy
(..e.g.
Bart1977,Lenington1979,G6tmarkand
Ahlund 1984). Predators could be attracted to
BART,J. 1977. Impact of human visitationson avian
nesting success.Living Bird 16:187-192.
agitatedparents,follow investigatorsvisually,
BENEDETTI,J., K. YUEN, AND L. YOUNG. 1990. Life
cue in to markers left near nests, or follow scent
trailsto the nest.Our resultssupportthe suggestionthatolfactorypredators
potentiallycould
use human scent to locate nests. Various ideas
tables and survival functions. Pages739-768 in
BMDP statisticalsoftware manual, vol. 2. (W. J.
Dixon, Ed.). Univ. California Press,Berkeley.
BLANCHER,
P. J., AND R. J. ROBERTSON.1985. Site con-
sistencyin kingbird breeding performance:Im-
have been offeredto counteractthe potential
plicationsfor sitefidelity. J.Anim. Ecol.54:1017biascausedby the presenceof humans,includ1027.
ing repellents (e.g. mustardoil and kerosene BROWN,M. B. 1990. Part 3. Analysis of multiway
[Hammond and Forward 1956]; naphthalene
tables:Log-linearmodels.Pages277-310in BMDP
statisticalsoftware,vol. 2. (W. J. Dixon, Ed.). Univ.
mothballs[Hamerstrom1970,Gawlik et al. 1988]
and mechanical devices [Post and Greenlaw
1989]).As suggestedabovefor artificial nosts,
the useof animalor shieldingscentsto mask
humanodor couldbe an importanttool in stud-
iesof actualnestingsuccess.
Our findingsin-
California Press,Berkeley.
GAWLIK, D. E., M. E. HOSTETLER,AND K. L. BILDSTEIN.
1988. Naphthalenemoth ballsdo not deter mammalian predatorsat Red-wingedBlackbirdnest.
J. Field. Ornithol. 59:189-191.
GEORGE,T. L. 1987. Greater land bird densities on
dicate that any steps taken to decreasethe
island vs. mainland:Relation to nest predation
amountof humanscentin theproximityof nests
level. Ecology68:1393-1400.
shouldhelp to decrease
the potentialnegative CdSTM.a•RK,
F., ANDM. J(HLUND.1984. Do field obinfluenceof observerpresence.
serversattract nest predatorsand influence nesting success
of CommonEiders?J.Wildl. Manage.
48:381-387.
ACKNOWLEDGMENTS
HAMERSTROM,
F. 1970. Think with a goodnosenear
a nest. Raptor Res. News 4:79-80.
We thank Joel Brown, Marcy DeMauro, Richard HAMMOND, M. C., AND W. F. FORWARD. 1956. ExHolmes, BobMorgan, Len Reitsma,ScottRobinson,
periments on causesof duck nest predation. J.
and StanleyTemple for helpful consultationon exWildl. Manage. 20:243-247.
perimental design and/or interpretation.We also KNIGHT,R. L., ANDS. A. TEMPLE.1986. Methodologthank three reviewerswhosecomments
greatlyimical problems in studiesof avian nest defence.
provedthe manuscript.
We areespeciallygratefulto
JoeRothfor suggesting
the useof animalandshield-
Anim.
Behav.
34:561-566.
LENINGTON, S. 1979. Predators and blackbirds: The
952
wrm•u
"uncertaintyprinciple" in field biology.Auk 96:
190-192.
LOISELLE,
B. A., AND W. G. HOPEES.1983. Nest predation in insular and mainland
est in Panama.
Condor
lowland
rain for-
85:93-95.
MARTIN,T. E. 1987. Artificial nest experiments:Effects of nest appearanceand type of predator.
Condor
89:925-928.
MARTIN, T. E. 1988a. Processesorganizing opennesting bird assemblages:Competition or nest
predation?Evol. Ecol.2:37-50.
MARTIN,T. E. 1988b. On the advantageof being
different:Nest predationand the coexistence
of
bird species.Proc.Natl. Acad. Sci. USA 85:21962199.
[Auk, Vol. 111
Effectsof removal of red squirrels, Tamiasciurus
hudsonicus,
and easternchipmunks, Tamiasstria-
tus,on nest predationin a northernhardwoods
forest: An artificial nest experiment. Oikos 57:
375-380.
RICKL•S,R.E. 1969. Ananalysisofnestingmortality
in birds. Smithson. Contr. Zool. 9:1-48.
TERBORGH,
J. 1989. Where have all the birds gone?
Princeton Univ. Press,Princeton, New Jersey.
TROST,
F. 1989. Sexscentshowdown. Outdoor Digest
September/October:16-17.
Wh•L•N, C. J., M. F. WILLSON,
C. A. TUMA,ANDI.
SOUZA-PINTO.1991. Spatial and temporal patternsof postdispersal
seedpredation.Can.J.Bot.
69:428-436.
PETTINGILL,
O. S., JR. 1976. Observedactsof predation on birds in northern lower Michigan. Living
Bird
ET AI,.
15:33-41.
PICM•N, J. 1987. An inexpensivecamerasetup for
the study of predation at artificial nests.J. Field
Ornithol.
POST,W., •D
58:372-382.
J. S. GREENLAW.1989. Metal barriers
WILCOVE,
D. S. 1985. Nest predationin foresttracts
and the declineof migratorysongbirds.Ecology
66:1211-1214.
WILLEBRAND,
T., AND V. MARCSTR•M.1988. On the
dangerofusingdummyneststostudypredation.
Auk 105:378-379.
YAHNER,R. H. 1989. Effectsof nest-siteselectionon
protectnear-groundnestsfrom predators.J.Field
depredationof artificialnests.J. Wildl. Manage.
Ornithol.
53:21-25.
60:102-103.
RErrSMA,L. R., R. T. HOLMES,•ND T. W. SHERRY. 1990.