WHY
DOES INTENSITY
INCREASE
RICHARD
DURING
OF AVIAN
THE
NEST
NESTING
DEFENSE
CYCLE?
L. KNIGHT • AND STANLEY A. TEMPLE
Departmentof WildlifeEcology,Universityof Wisconsin,
Madison,Wisconsin
53706USA
ABSTRACT.--Theoretical
modelspredict that intensity of avian nestdefenseshouldincrease
with age of the offspring.Empiricalobservationsconformingwith this prediction have been
taken as support for these models. We here present a new explanation for the observed
correlationsbetween offspring age and level of nest defense. We propose that increased
intensityof nest-defensebehavioris largely a resultof the methodsusedby the researchers
who made the observations.We suggestthat when an observerrepeatedlyvisits or brings a
potential nest predatorto a nest, nest-defensebehavior of parentsis modified by positive
reinforcement
andlossoffear.Wetested
thishypothesis
bymeasuring
nest-defense
behavior
of female American Robins(Turdusmigratorius)
and male and female Red-winged Blackbirds
(Agelaiusphoeniceus).
The testswere performed at nestsvisited only once during different
stagesof the nesting cycle and at a group of nests visited repeatedly from initiation of
incubation to fledging. Nest-defenseintensity (as measuredby call rates, closestapproach
to the predator, and numbers of dives and strikes) of robins and blackbirds at nestsvisited
once,in most cases,did not increasethrough the nestingcycle.At nestsvisited repeatedly
through the nesting cycle,intensity of nest defenseby both robins and blackbirds,in many
cases,increasedsignificantly.In addition to intensityof nestdefenseincreasingat multiplevisit nests,we observedhigher proportionsof birds at these nestsperforming nest-defense
behaviorsthan at single-visit nests.Received19 March 1985,accepted
20 October1985.
MOST studies of nest defense in altricial birds
survival of parents and their young decreases
have revealedincreasesin the intensity of nest as the nestlings grow older, making parents
defense by parent birds through the nesting willing to defend older nestlingsmore aggrescycle (Smith 1950; Erpino 1968; Barash1975; sively.
Curio 1975; D'Arms 1978; Weatherhead 1979,
The secondexplanation is basedon the hy1982;Anderssonet al. 1980;Greig-Smith 1980; pothesisthat older nestlingsare more conspicPatterson et al. 1980; Biermann and Robertson
uous to predators. The nest and its contents
1981; East 1981; Blancher and Robertson 1982;
becomemore conspicuousas the nesting cycle
RtSell and Bossema 1982; Merritt 1984; Shields
progresses,
necessitatingan increasein the in1984).Two explanationshave been offeredfor tensity of nest defenseto counteractthe nest's
to predators(Skutch
the temporalchangesin the intensity of nest increasedconspicuousness
defense.The first, and currently most popular, 1949, Harvey and Greenwood 1978).
We offer a third, alternativeexplanationfor
explanationis derived from the theory of parental investment (Trivers 1972). Barash (1975) increasesin nest defense through the nesting
extended the theory to include parental de- cycle.We proposethat the increasesare largely
fense of eggs and young. Whether basedon the results of the methods used by the recumulative past parental investment (Trivets searcherswho made the observations.We sug1972, Barash1975) or future expectedbenefits gestthat when an observerrepeatedlyvisits or
minus expectedcosts(Dawkins and Carlisle brings a potential predator to a nest and rec1976,Boucher1977,Maynard Smith 1977),this ords the parent birds' responses,the nest-dehypothesispredictsan increasein nestdefense fense behavior is gradually modified by posiand loss of fear. Positive
as the young near fledging. In a similar fash- tive reinforcement
ion, Anderssonet al. (1980) hypothesizedthat reinforcement is involved becausethe parent
the relative difference between the expected birds have been rewarded repeatedlyfor their
nest-defensebehavior.After being attacked,the
• Present
address: Institute
for Environmental
observeror other potential nestpredatorleaves
Studies,Universityof Washington,Seattle,Washing- without harming the nest. A lossof fear is inton 98195 USA.
volved becausethe defending birds gradually
318
The Auk 103: 318-327. April 1986
April 1986]
Changes
inNest-defense
Intensity
learn that the observer or other potential nest
predator is not dangerousto them; however,
they still view the objectas a threat to the nest's
contents.Both positive reinforcementand loss
of fear could explain much of the reported increases in nest-defense
behavior.
We examined these explanationsby measuring nest-defensebehavior in experimentswith
American Robins (Turdusmigratorius)
and Red-
winged Blackbirds(Agelaius
phoeniceus).
319
blackbirdsand robins was assumedto have begun
the day beforethe lasteggwaslaid. For single-visit
nests, 7 blackbird and 8 robin nests were chosen for
each of the 6 time intervals. Multiple-visit nestswere
visitedevery 3 days,beginningwith the initiation of
incubationand continuing until fledging.
Nest defense in blackbirdswas quantified by mea-
suringthe responsesof the male and female parents
either to an observer (always the same individual)
standingby the nest or to a mounted raccoon(Procyonlotor)on top of an adjustablealuminumpole, in
both caseswithin 0.5 m of the nest. Responsesof
individual
STUDY AREA AND METHODS
Nests of blackbirds
and robins were located within
the city of Madison,Wisconsin.Somenestswere visited every 3 days,whereas others were visited only
once.
Blackbird
nests visited
once were
located
in
Nielsen, University Bay, Redwing, and Odana
marshes;nests visited repeatedly were located in
Kettle Marsh. These areasare open water, cattail (Typha spp.) marshes (Bedford et al. 1975). At no time
during the studyperiodwere humansseenin any of
these marshes.
All
robin
nests were
located
on the
blackbirds
to the human
and the raccoon
(balancedfor order of presentation)were measured
on the same day, ! h apart. A cloth attached to a
string coveredthe raccoonand was removed after
the observer
had retreated
at least outside
of the
male's territory and the blackbirds had resumed
"normal" behavior. The raccoon mount was placed
so that the head was level with and facing the nest.
During a 3-min period, the types of calls, the total
number of calls, dives, and strikes, and the nearest
distancethe birds approachedthe potential predator
were recorded by two observers (one observer
watched the male and the other watched the female).
University of Wisconsincampusand adjacentneigh-
When the potential nest predator was a human, that
borhoods.
individual
Nests
at which
nest defense
was measured
once
during the incubationperiod were located during
the nest-buildingor egg-layingstages,thereby ensuring that we knew the date of initiation of incubation. Visits to nestsduring theseperiodsconsisted
of walking past the nest without pausing. Nestsat
which nest defensewas measuredonce during the
nestling period were nestswe found with nestlings.
Measurements
of tarsusand wing chordimmediately
after nest-defensetrials allowed us to age nestlings
on the basisof comparisonswith growth curves(S.
K. Knight pets. comm.).BecauseRed-wingedBlackbirdsarepolygynous,only one nestwasusedin each
male's territory, thereby ensuring that the response
of a male blackbird
Blackbird
was not measured
nests where
nest defense
twice.
was measured
once were selectedrandomly from a larger set of
known nests.Blackbird nestsvisited repeatedly were
in a different marsh (Kettle Marsh) and were chosen
randomly from nests within each male's territory.
Robin nestswere assignedrandomly as single-visit
or multiple-visit nests.One-visit nestswere at least
100 m from multiple-visit nests. For multiple-visit
nests,we began with a larger number of nests,but
in our analyseswe used only data collected from
blackbird (n = 7) and robin (n = 8) nests that successfullyfledged at least one young.
The combinedincubationand nestling periodsfor
both
robins
and blackbirds
are between
22 and 24
days(Stokes1979).We thereforedivided the nesting
cycle into 6 four-day intervals,beginning with the
initiation
of incubation.
Initiation
of incubation
for
distant
recorded
observer
tances were
the female's
recorded
estimated
behavior
the male's.
to the nearest
while
the
Closest
dis-
0.5 m. Dives
were any break in horizontal flight that was directed
at the predator. Call types recorded were those describedby Orians and Christman (1968).
Nest defense in robins was measuredby record-
ing the responses
of femalesto a humanat the nest.
Femaleswere sexedby their lighter headand breast,
particularly when comparedwith mates,and by the
fact that only females incubate (Stokes 1979). Call
types (Stokes1979) and number, dives, strikes,and
the closestapproachto the potential predator at the
nest were recorded during a 3-min period. We did
not measure
nest-defense
behavior
of male robins
or
responsesof female robins to a mounted raccoon.
Linear regressionanalysis(Ryan et al. 1976) was
used to assessthe degree to which either the stage
of the nesting cycleor the number of previousvisits
could be used to predict nest-defenseresponse.We
used analysisof variance to test the null hypothesis
that regressioncoefficientsdid not differ significantly from zero; all testswere one-tailed. Results were
consideredstatisticallysignificantat P < 0.05.
For multiple-visit nests the assumptionof independencewould be violated if we usedthe responses
from all visits to each nest when performing linear
regressionanalysis.Accordingly, for each response
variable for each of the 8 nest visits, we randomly
selected(with replacement) one nest to provide the
datumfor the regressionanalysis.Five independent
regressionanalyseswere performed for each responsevariable.Residualsfrom the regressionanal-
320
KNIGHT
ANDTEMPLE
MALE BLACKBIRD
FEMALE
E 20
BLACKBIRD
20
DAYS
[Auk,Vol. 103
FEMALE
ROBIN
20
SINCE
INITIATION
OF
INCUBATION
Fig. 1. Mean _+2 SE of nest-defenseresponsesby blackbirdsand robinsto a human (open circles)and
raccoon(closedcircles)at nestsvisitedonly once.n = 7 nestsfor blackbirdsand n = 8 nestsfor robins.
yseswere plottedagainstthe independentvariables our resultsgave a trend oppositeto what existand the predictedy-valuesto determinewhetherdata ing theory predicted (e.g. as the nesting cycle
transformationswere necessary.The variableswere
log or square-roottransformed when necessaryto
meet the assumptionof linearity. Becauseexisting
theory on increasesin nest-defenseintensity do not
necessarilypredict a linear increase over time, we
also performed ANOVA tests on the data from single-visit nests to see whether there were stepwise
progressed,male blackbirds increasedtheir
increases.
the raccoon;in 3 of the 11 regressionequations
We calculated
correlation
coefficients
between
the
distancefrom the predator model). Coefficients
of determination were extremely low for each
of the 11 regressionequations(Table 1). The
greatestamountof variationexplainedby stage
of the nestingcyclewas 9% for female callsto
stageof the nestingcycleexplainednone of the
various nest-defenseresponsesat all nests.Because
variation.
dives and strikes (when there were strikes) were
We found similar trends for the responsesof
female robins at single-visit nests.For none of
the three responses
wasthe null hypothesisthat
nest-defenseintensity was independent of the
stageof the nestingcyclerejected(Table 1). As
highly correlated(P < 0.001),we usedthe sumof the
dives and strikesin our analyses.
RESULTS
in blackbirds, the coefficients of determination
The first hypothesiswe examined was that were low. Stageof the nestingcycleexplained
nest-defenseintensity was related to the stage none of the observed variation in calls or nearof the nestingcycle.To test this hypothesiswe est approachand only 1% of the variation in
analyzed resultsfrom only single-visit nestsso dives and strikes (Table 1).
that stageof the nesting cycle was not confounded with the number of previous visits.
Results from the ANOVA
tests were no dif-
ferent from those of the regressionanalyses,
exceptthat femalecallsto the raccoonno longer showeda significantincreaseover time (AN-
The evidence from single-visit nests only
weakly supported this hypothesis and more
stronglysupportedthe null hypothesisthat in- OVA, P > 0.25).
The second hypothesis examined was that
tensity of nest defenseis independent of stage
of the nestingcycle (Fig. 1, Table 1). Nest-de- changesin nest-defenseintensitywere depenfenseintensityby both male and female black- dent on the number of previous nest visits by
birdsdid not changethrough the nestingcycle a potentialnestpredatorthat doesnot harm the
in the direction predicted by existing theory nest or the defending parents.Our null hyfor 10 of the 11 responses(Table 1). Only for pothesis was that nest-defenseintensity was
calls by femalesto the raccoonwas the null independent of number of previous nest visits.
For blackbirdsat multiple-visit nests,regreshypothesisrejected.Indeed,for four variables,
April 1986]
Changes
in Nest-defense
Intensity
321
T^BI•E1. Contributionof stageof the nestingcyclein explainingvariancein predator-elicitedresponsesof
Red-winged Blackbirdsand American Robinsto potential nest predatorsat nestsvisited once.
Coefficient
Speciesand sex and
responsevariable
of
Resultsof analysis of variancec
Regression
coefficient(B)a
determination
(r2)b
Calls/3 min to human
Calls/3 min to raccoon
2.71
0.05
Closestapproach (m) to human
Closestapproach(m) to raccoon
0.60
0.48
0.00
0.02
0.03
0.08
0.00
0.03
1.14
1.90
2.07
4.37
0.03
2.10
>0.25
>0.10
>0.10
<0.05
>0.25
>0.10
0.09
-0.09
0.03
0.09
0.00
0.01
2.19
4.44
0.08
1.51
>0.10
<0.05
>0.25
>0.10
0.08
0.00
0.77
>0.25
4.25
0.00
0.00
0.01
0.64
0.10
1.44
>0.25
>0.25
>0.10
F
P
Red-winged Blackbirdmales
Dives and strikes/3 min to human
Dives and strikes/3 min to raccoon
0.03
-0.78
Red-winged Blackbirdfemales
Calls/3 min to human
Calls/3 min to raccoon
18.00
0.24
Closestapproach (m) to human
Closestapproach(m) to raccoon
Dives and strikes/3 min to human a
Dives and strikes/3 min to raccoon
American
Robin females
Calls/3 min to human
Closestapproach(m) to human
Dives and strikes/3 min to human
-0.05
0.09
In all casesn = 42 for Red-winged Blackbirdsand n = 48 for American Robins.
Adjustedfor degreesof freedom.
Analysisof variancetestedwhetherthe regressioncoefficients
were significantlydifferentfrom zero.
No dives or strikes.
sion coefficientsfor each of the 12 responses
differedsignificantlyfrom 0 in at least1 of the
5 regressionanalysesperformed on each responsevariable.For 7 of the responsevariables
all regressionanalyseswere significant (Table
2, Fig. 2). For all 12 variables,the resultswere
uniformly in the directionpredictedby our alternativehypothesis.Theseresultsindicatethat
blackbirds called more often, dove and struck
more often, and approached potential nest
predatorscloserasthe numberof previousnest
visits increased.
Coefficients
of determination
were consistentlymuch higher for number of
previousvisits than for stageof the nesting
intruder, whereasnumber of previousnest visits explained as much as 44% of the variation.
Although all blackbirds and robins called
while defendingtheir nests,therewere marked
differencesin the proportionof birdsthat dove
at or hit the potential predator at single- and
multiple-visitnests.For nestswith nestlings912 days old, very few parents at single-visit
nestseither dove at or hit the potential predator, whereasat multiple-visitnestsover half of
the parentsdid (Table 3). Additionally, there
were many more dives and strikes by parents
at multiple-visitneststhanby parentsat single-
visit nests.For example,when young were between 9 and 12 daysold, there were 1,244dives
Nest-defense
intensityby robinsat multiple- and strikes by parents at multiple-visit nests
visit nests also increasedin responseto the but only 21 dives and no strikesby parentsat
number of previousvisits.As with blackbirds, single-visit nests.
cycle (Tables 1 and 2).
the number of calls, dives, and strikes in-
Differencesin predator-elicitedresponses
by
creasedwith repeatedvisits,while closestdistance approachedby robins decreased(Table
2). For all responsevariables,coefficientsof determination were much higher for number of
previousvisitsthan for stageof nestingcycle
(Tables1 and 2). For example,stageof the nest-
birds at nests visited once and nests visited re-
ing cycle explained none of the variation in
ward humans
closestdistancerobinsapproachedthe human
peatedlycould be due to multiple-visit nests
having parentsthat were inherently more aggressivethan parentsat nestsvisitedonly once.
Blackbirds in Kettle Marsh, where nests were
visited repeatedly,may be more aggressivetoand raccoons than blackbirds in
the marshes where we measured nest defense
322
KNIGHTAND TEMPLE
[Auk, Vol. 103
TABLE
2. Contributionof stageof the nestingcyclein explainingvariancein predator-elicited
responses
of
Red-winged
Blackbirds
andAmericanRobinsto potentialnestpredators
at nestsvisitedevery3 days.
Speciesand sexand
response
variable
Regression
coefficient
(B)a
CoefficientResults
ofanalysis
ofvariance
½
of determination(r2)b
F
P
Red-winged Blackbirdmales
Calls/3 min to human
Calls/3 min to raccoon
Closestapproach(m) to human
Closestapproach(m) to raccoon
Dives and strikes/3 min to human
Dives and strikes/3 min to raccoon
0.11-0.20
0.11-0.20
-2.98 to -1.49
-0.74 to -0.53
0.32-2.34
6.91-16.40
0.26-0.74
0.46-0.64
0.48-0.53
0.26-0.43
0.70-0.78
0.16-0.55
3.46-20.52
7.00-13.32
7.40-8.88
3.50-6.35
16.97-25.50
2.31-7.02
0.0025
0.01
0.01
0.025
0.001
0.025
<
<
<
<
<
<
P
P
P
P
P
P
<
<
<
<
<
<
0.25
0.05
0.05
0.25
0.01
0.25
0.25-38.52
0.17-32.85
-0.28 to -0.25
-0.38 to -0.28
1.61-1.94
2.37-14.67
0.49-0.58
0.40-0.51
0.49-0.53
0.20-0.41
0.65-0.68
0.52-0.74
7.73-10.50
5.71-8.29
7.73-8.88
2.80-5.86
14.21-15.52
8.47-20.70
0.01
0.025
0.01
0.05
0.005
0.0025
<
<
<
<
<
<
P
P
P
P
P
P
<
<
<
<
<
<
0.05
0.10
0.05
0.25
0.01
0.05
0.24-4.32
-1.97 to -1.61
1.71-1.75
0.34-0.71
0.39-0.44
0.39-0.75
4.58-17.72
5.57-6.50
5.57-22.09
Red-wingedBlackbirdfemales
Calls/3 min to human
Calls/3 min to raccoon
Closestapproach(m) to human
Closestapproach(m) to raccoon
Dives and strikes/3 min to human
Dives and strikes/3 min to raccoon
American
Robin females
Calls/3 min to human
Closestapproach(m) to human
Dives and strikes/3 min to human
0.0025 < P < 0.10
0.025 < P < 0.10
0.001 < P < 0.10
aForeachresponse
variableon eachof the 8 nestvisits,we randomlyselected(with replacement)
onenest
toprovidedatafor theregression
analyses.
Rangeof valuespresented
from5 independent
regression
analyses
performedfor eachresponsevariable.
bAdjustedfor degreesof freedom.
cAnalysisof variancetestedwhetherthe regression
coefficients
weresignificantly
differentfromzero.
only once.This couldbe due to proximityand
natural exposureto humansor raccoons.Such
an explanationis unlikely for robin nestsbecausethey were assignedrandomly to treatments. We examinedthis possibilityfor both
blackbirdsand robinsby comparingthe initial
responsesamong nestsvisited once and nests
visited repeatedly.There were no significant
predator,becausethe proximity and intensity
of their nest-defense
behavior
increased. Con-
versely,if parentsor their nestswere harmed
(negativereinforcement),then a declinewould
be expectedin the intensity of particularcomponents(e.g.divesand strikes)of nest-defense
behavior, depending upon the frequency and
outcome of these previous encounters (e.g.
Knight 1984,Knight and Templein pressa).
We are not the first to proposethat increased
intensity of responsesto predatorsmight be
the result of repeatedpresentationsand withDISCUSSION
drawals of potential predators.Verplanck (in
Our evidencesuggeststhat stageof the nest- FIinde 1954) suggestedthat the withdrawal of
ing cycle has only a weak influence on nest- an owl that is being mobbedmight reinforce
defenseintensity.Our resultsstronglysupport distraction behavior. Gramza (1967) proposed
the hypothesisthat increasedintensityin nest- that withdrawal of a human intruder might
defensebehaviorthrough the nestingcycleoc- positively reinforce distraction behavior in
micurs when an observer repeatedly visits or nesting Common Nighthawks (Chordeiles
placesa potential nest predatorat a nest and nor). Buitron (1983: 232) declared, "Studies
then leaves or withdraws the stimulus without
showing an increasein parental defensewith
the ageof the young have generallyinvolved
harming the birds.
Even thoughnestswere not harmed,parent the samepredator,human or model, throughbirds still viewed us or the predator model as out the breedingseason.In somecasesthe same
a potentialthreat.The continuednestdefense, pairs were repeatedlytested,often every day,
however,impliesthat parentslosefear of the and increasedfamiliarity with a particular
differences (Mann-Whitney U-test, all P values > 0.10) in any of thesecomparisons.
April 1986]
Changes
inNest-defense
Intensity
DAYS SINCE INITIATION
323
OF INCUBATION
Fig.2. Mean+ 2 SEof nest-defense
responses
byblackbirds
androbins
toa human
(opencircles)
and
raccoon
(closed
circles)
at nestsvisitedeverythreedays.n = 7 nestsforblackbirds
andn = 8 nestsforrobins.
predatorand situationsmayhaveaffectedthe
sponses,
we found strikingdifferencesin adult
behaviorat single-vs. multiple-visitnests.We
be disadvantageous,
the repeatedwithdrawal measurednest defenseonly during the incuresponse.While habituationto predatorswould
ofthehumanormodelpredator
withnoinjury bationandnestlingstages
of thenestingcycle.
to youngor adultsmay reinforcethe vigor of We wouldnot be surprisedto find significant
the response."AmericanGoldfinches(Carduelis differences
througha nestingcyclethatincludtristis)showedsignificantincreasesin call rates edadditional
stages,
suchasnestbuilding,eggfrom the incubationto the nestlingperiodat laying,and postfledging.Birdsrespondto the
nestsvisitedevery3 days(Knightand Temple presenceof a predatordifferentlyduring the
in pressb). At nestsvisitedonly onceduring earlypart of the nestingcycle,asis evidenced
the incubationor nestlingperiods,however, by their greatertendencyto desertduringthe
therewere no significantdifferencesin call rates nest-buildingand egg-layingstages(Skutch
eitherto humansor to modelsof a BlueJay 1976).Likewise,parentsmightdefendfledged
(Cyanocitta
cristata)or an AmericanKestrel(Fal- younglessintenselythan eggsor nestlingsbecosparverius).
causefledgedyoungare dispersedand candeThe large amount of variation around the fend themselves
(e.g.by flying or hiding).
calculated mean values in our work indicates
substantial individual variation in nest-defense
Intensity of nest defense has been shown to
be positivelycorrelated
with nestsuccess
(e.g.
behavior(Figs.1,2).Despitesmallsamplesizes Greig-Smith 1980,Blancherand Robertson1982,
and the large variations in nest-defensere-
Knight and Temple in pressb). Variouscom-
TABLE
3. Comparisons
of predator-elicited
responses
of Red-winged
Blackbirds
andAmerican
Robinsat
nestsvisitedonceand at nestsvisitedevery3 days.
Proportion of parents
that dove or struck a
Speciesand sexof parent Potential predator
Male blackbird
Femaleblackbird
Femalerobin
Total number of
dives and strikes a
Single-visit Multiple-visit Single-visitMultiple-visit
nests
nests
nests
nests
Human
2/7
7/7
8
151
Raccoon
2/ 7
7/ 7
7
586
Human
0/ 7
3/ 7
0
15
Raccoon
0/ 7
4/ 7
0
302
Human
1/8
8/8
6
190
Responses
duringnestvisitswhennestlings
were9-12daysold.
324
KNIGHT
ANDTEMPLE
ponents of nest-defensebehavior, suchas calls,
may function in ways that minimize a predator's ability to locate and destroy a nest (e.g.
Greig-Smith 1980,Knight and Temple in press
b). This suggeststhat nest-defensebehavior is
adaptive.The existing theory on nest defense
assumesthat the reproductive value of the nest
contents increaseswith age, that the nest contents become more conspicuouswith increasing age, that nest defense incurs a cost (i.e. is
risky or energeticallyexpensive),and that the
level of nest defense performed is a measureof
either parental investment or the conspicuousnessof the young (Skutch 1949, Barash 1975,
Harvey and Greenwood 1978, Andersson et al.
1980). Of these assumptions,the first two are
axiomatic. The value and conspicuousnessof
nestsinevitably increaseswith age. Becauseour
resultssuggestthat existingtheory doesnot adequately explain patterns of nest-defenseintensity, closescrutiny of the last two assumptions
would
be instructive.
Previousnest-defense
studies.--Eachof 12 previous studies that examined
nest defense of al-
tricial birds in response to a human intruder
reported increases in nest-defense intensity
through the nesting cycle (Smith 1950;Erpino
1968; Barash 1975; Searcy 1979; Weatherhead
1979, 1982; Andersson et al. 1980; Greig-Smith
[Auk,Vol. 103
the particular model). Curio (1975: 29) stated
that "As in many birds, responsestrength(calls
per minute) exhibits a marked increaseduring
the breeding cycle." Later, however, (p. 31) he
statedthat "... differencesbetween stages(of
the nestingcycle)becomegenerallysignificant
only if covering the data of at least three subsequentones."Regelmannand Curio (1983)and
Curio et al. (1984) found that nest-defensein-
tensityof GreatTits (Parusmajor),as measured
by time (latency of approach)and distanceapproached,increasedwith nestling age;they did
not report data for the incubation period. Additionally, they reportedinconsistencies
in their
data basedon such factorsas year and study
area and attributed the large amount of unexplained variation in their responsevariables to
individualvariationamongthe parents.D'Arms
(1978) found no differences between the incu-
bation and nestling stagesin male Red-winged
Blackbird nest defense to avian predators.Female Red-winged Blackbirds with nestlings
were more aggressivethan femaleswith eggs.
Gottfried
(1979) measured nest-defense re-
sponseof Northern Cardinals (Cardinaliscardinalis),American Robins, Gray Catbirds (Dumetella carolinensis), and
Brown
Thrashers
studiesusinga human asthe stimulus.Several
(Toxostoma
rufum)to predator modelsand concluded:"My data do not clearly tell if the intensity of aggressiontoward nest predators
changesduring the nestingcycle." Pattersonet
al. (1980) found that White-crowned Sparrows
(Zonotrichialeucophrys)
were no more aggressive to a mounted Scrub Jay (Aphelocoma
coerulescens)
during the nestling period than during the incubation period. They did report
significantincreasesin nest defenseto models
of an American Kestrel and a Dark-eyed Junco
(Juncohyemalis)and to a live terrestrial garter
snake (Thamnophiselegans).During May and
June, Biermann and Robertson(1981) reported
that female Red-winged Blackbirdsresponded
no more aggressivelyto a rubber snakein their
nestsduring the nestling than the incubation
period. During July, however, blackbirdswere
more aggressivewhen they had nestlingsthan
when they had eggs.Finally, R•Selland Bossema (1982)reporteda statisticallynonsignificant
trend for a higher proportion of naive Blackbilled magpies(Picapica)to attack a live avian
predatorat nestsduring the nestling stagethan
during incubation.
of these presented potential predators to naive
birds (i.e. birds that had not previously seen
responsesof magpiesand Common Jackdaws
1980; East 1981; Blancher and Robertson 1982;
Merritt 1984; Shields 1984). In each study individual nests were visited more than once; in
at leasteight of the studiesnestswere visited
at 1-3-day intervals. Additionally, becauseparent birds may learn to recognize individuals
(e.g. the researcher)from previous visits (Buitron 1983,Merritt 1984,Knight and Temple in
pressa), nest defense can be even more intense
to familiar
individuals
than to novel
intruders
(Knight and Temple in pressa). We know of
no published studies where observers visited
nests repeatedly and failed to obtain an increase in nest defense.
Nest-defense studies that measured parent
responses
of altricial birds through the nesting
cycleto either a taxidermicmount (Curio 1975,
Gottfried 1979, Patterson et al. 1980, Biermann
and Robertson1981, Shields 1984) or a living
animal (D'Arms 1978, R•Selland Bossema1982,
Regelmannand Curio 1983, Curio et al. 1984)
have obtained
results
far less consistent
than
R•Selland Bossema(1982) also measured the
April1986]
Changes
inNest-defense
Intensity
(Corvusmonedula)
to weekly presentationsof the
predator. None of the eight pairs of magpies
showed consistent changes in responses
through the nesting cycle, although jackdaws
did show increasedscolding. Shields (1984)
measured nest defense in Barn Swallows (Hi-
rundorustica)to a predatormodelat nestsand
found that intensity of mobbing was significantly greater during the nestling stage than
during the incubation stage. He placed the
predatormodel near nestsat intervalsthat varied from twice weekly to twice during the
breeding season.
Two possibleexplanationsexist for the contradictoryfindings of the studieswith models
or live predatorsother than man. First, because
researchershave to place the predator at the
nestand then withdraw beforemeasuringnest
defense,they have, in effect,presentedthemselvesfirst as a potentialpredator(Kruuk 1964:
325
free-rangingpredatorsprobablywould influence the responseto the predator model (Pugesek1983,Smith et al. 1984).This is particularly important becausemany of the species
usedaspredatorscanbe successfully
driven off
by parents(Biermannand Robertson1981,Ri3ell
and Bossema1982), thereby providing parents
with positive reinforcement.
Our resultsprovide a mechanismfor Pugesek's (1983) observations that older California
Gulls(Laruscalifornicus)
attackhumanintruders
more often than middle-aged and young parents do. Older birds may defend their nests
moreaggressivelybecausethey have had more
previous"successful"
nest-defenseinteractions
with
human
intruders.
In studies where
nest-
defenseintensity and nest successwere corre-
lated (e.g. Anderssonet al. 1980,Greig-Smith
1980,Blancherand Robertson1982,Knight and
Templein pressb), it would be interestingto
75-76). Unless the researcher is hidden from
know
whether
the birds
that
nested
success-
the parents'view, is a sufficientdistanceaway
from the nest, or allows the parents to habituate to the observer'spresence,at leastpart of
the nest-defensebehavior measuredmay be attributed to the researcher'spresence. Curio
(1975:26-27), for example,found that Pied Flycatchers(Ficedulahypoleuca)continued calling
for some time after a potential predator was
removedfrom the area. In only 2 of the 9 studies listed above was it apparent that the observer was concealed.In one study the parent
birds were captured and banded before measuring nest defense. In this and four other
fully were the older birds.
We know of only one study that quantified
nest-defense behavior of an altricial bird to flee-
ranging wild nest predators. Buitron (1983)
measuredBlack-billedMagpienest-defense
behavior
to a number
of avian
and mammalian
predatorsof magpie adults, nestlings,or eggs.
Magpiesrespondedto a variety of live predators with similar intensity and frequency regardlessof the stageof the nestingcycle.
It is clear that caution
must be used in ac-
cepting various explanationsfor increasesin
nest-defenseintensity. Many previous studies
studies it was not stated whether the observers
that attempted to test hypothesesexplaining
were visible, whether the birds were allowed
changesin nest-defenseintensity lacked adeto return to "normal" behavior before obserquatecontrols,which made it difficult to attribvations began, or how far the observerswere ute changesin nestdefenseto a particularcause.
from the nest.In one studythere were at least Future nest-defensestudiesshould employ dethree observerspresentduring the trials. We signsthat minimize severalof the potentially
suggestthat the inconsistentfindings in these confoundingvariableswe have discussed.
studiesmay be attributedto the presenceof the
researcher,particularlyif the researchervisited
ACKNOWLEDGMENTS
the nest repeatedlythrough the nestingcycle.
For valuable assistance in the field we thank ReSecond,studiesusing models or live predators as stimuli to study nest-defensebehavior becca Abel and Alan Crossley. Our research was
have chosen speciesthat occur naturally in funded by grantsfrom the GraduateSchooland the
Friends of the Arboretum, University of Wisconsin,
nesting areasand that are suspectedof being and
from the Max McGraw Wildlife
Foundation.
Perfrequent nest predators.We have demonstratmission to work in the marsheswas provided by
ed that nest-defenseintensity increasesafter Katherine Bradley, Director of the Arboretum, Uni"successful" defenses of the nest (i.e. the re-
versity of Wisconsin,and by Si Widstrand, Madison
searcherwithdraws or removes the predator ConservationParksDepartment.Commentsfrom M.
model).Therefore,the frequencyand outcome Andersson, D. E. Blockstein, D. Buitron, J.P. Hailof previous encountersbetween parents and man, S. K. Knight, G. H. Orians, M. Samuel,and K.
326
KNIGHT
ANDTEMPLE
Yasukawaclarified our thinking on this subjectand
are deeply appreciated.We alsothank refereesW. M.
Shieldsand L. K. Southernfor suggestingimprovements on the manuscript.
[Auk,Vol. 103
R. Krebs and N. B. Davies, Eds.). Sunderland,
Massachusetts, Sinauer Assoc.
HINDE, R.A. 1954. Factorsgoverning the changes
in strength of a partially inborn response,as
shown by the mobbing behaviour of the Chaf-
finch (Fringillacoelebs).
II. Proc. Royal Soc.London, Sect. B, 142: 331-358.
LITERATURE CITED
ANDERSSON,
M., G. WIKLUND, & H. RUNDGREN. 1980.
KNIGHT,R.L. 1984. Responsesof nesting ravens to
human beings in areasof different human den-
Parental defenceof offspring:a model and an
example. Anim. Behav. 28: 536-542.
BARASH,
D.P. 1975. Evolutionary aspectsof parental behavior: distractionbehavior of the Alpine
Accentor.
Wilson
Bull. 87: 367-373.
Wetlandsof Dane County, Wisconsin.Madison,
Wisconsin,Dane Co. Regional Planning Comm.
BIERMANN,G. C., & R. J. ROBERTSON.1981. An in-
creasein parental investmentduring the breeding season.Anim. Behav. 29: 487-489.
BLANCHER,
P. J., & R. J. ROBERTSON.
1982. Kingbird
aggression:does it deter predation? Anim. Behav. 30: 929-930.
111: 786-788.
78: 209-236.
115.
K. REGELMANN,& U. ZIMMERMAN. 1984. The
defenceof first and secondbroodsby Great Tit
(Parusmajor) parents: a test of predictive sociobiology. Z. Tierpsychol.66: 101-127.
D'ARMS,E. H.H. 1978. Red-wingedBlackbirdmobbing behavior in the context of changing investmentpatterns.Unpublished M.S. thesis,Seattle, Univ. Washington.
DAWKINS, R., & T. R. CARLISLE. 1976.
Parental in-
vestment, mate desertion and a fallacy. Nature
262: 131-133.
EAST,M. 1981. Alarm calling and parental investment in the robin Erithacus rubecula. Ibis 123: 223230.
1968. Nest-related activities of Black-
billed Magpies.Condor 70: 154-165.
GOTTFRIED,
B. M. 1979. Anti-predator aggressionin
birds nestingin old field habitats:an experimental analysis.Condor 81: 251-257.
GRAMZA,A. 1967. Responsesof brooding nighthawks to a disturbance
GREIG-SMITH,P.W.
stimulus.
Auk 84: 72-86.
1980. Parental investment in nest
defenceby Stonechats(Saxicola
torquata).Anim.
Behav.
In pressb. Nest defence in the
Goldfinch.
Anim.
Behav.
KRUUK,H. 1964. Predatorsandanti-predatorbehaviour of the Black-headedGull Larusridibundus
(L.).
Behaviour Suppl. 11: 1-129.
MAYNARD SMITH, J. 1977. Parental investment: a
prospectiveanalysis.Anim. Behav.25: 1-9.
MERRITT,
P. G. 1984. Observerrecognitionby the
Northern Mockingbird.J.Field Ornithol. 55:252ORIANS, G. H., & G. M. CHRISTMAN. 1968.
A com-
Tricolored, and Yellow-headed
Calif.
Publ. Zool.
blackbirds. Univ.
84: 1-81.
PATTERSON,
T. L., L. PETRINOVICH,
& D. K. JAMES.1980.
CuRIo,E. 1975. The functional organizationof antipredatorbehaviourin the PiedFlycatcher:
a study
of avian visual perception.Anim. Behav. 23: 1-
ERPINO,M.J.
& --.
American
parative study of the behavior of Red-winged,
BUITRON,D. 1983. Variability in the responsesof
Black-billed Magpies to natural predators. Be-
--,
Behav.
253.
BOUCHER,D. H. 1977. On wasting parental invest-
haviour
86: 345-346.
& S. A. TEMPLE.In pressa. Methodological
problems in studiesof avian nest defence.Anim.
--,
BEDFORD,
B. L., E. H. ZIMME, • J. H. ZIMMERMAN. 1975.
ment. Amer. Natur.
sities. Condor
--,
28: 604-619.
HARVEY,P. H., & P. J. GREENWOOD.1978. Antipre-
dator defence strategies: some evolutionary
problems.Pp. 129-151 in Behaviouralecology(J.
Reproductivevalue and appropriatenessof response to predators by White-crowned Sparrows. Behav. Ecol. Sociobiol.
7: 227-231.
PUGESEK,
B. H. 1983. The relationshipbetween parental age and reproductive effort in the California Gull (Laruscalifornicus).
Behav. Ecol. Sociobiol.
13: 161-171.
REGELMANN,K., & E. CURIO. 1983. Determinants of
brooddefencein the GreatTit ParusmajorL. Behav. Ecol. Sociobiol.
13: 131-145.
R•ELL,A., & I. BOSSEMA.
1982. A comparisonof nest
defenseby Jackdaws,
Rooks,magpiesand crows.
Behav. Ecol. Sociobiol.
11: 1-6.
RYAN, T. A., JR., B. L. JOINER,& B. F. RYAN. 1976.
MINITAB
student handbook. North Scituate,
Massachusetts,
Duxbury Press.
SEARCY,
W. A. 1979. Female choiceof mates:a general model for birds and its application to Redwinged Blackbirds (Agelaiusphoeniceus).
Amer.
Natur.
114: 77-100.
SHIELDS,
W. M. 1984. Barn Swallow mobbing:selfdefence, collateral kin defence, group defence,
or parental care?Anim. Behav. 32: 132-148.
SKUTCH,
A. F. 1949. Do tropical birds rear as many
young as they can nourish?Ibis 91: 430-455.
--.
1976. Parent birds and their young. Austin,
Univ.
Texas Press.
SMITH, J. ]N•.M., P. ARCESE,& I. G. McLEAN. 1984.
Age, experience,and enemy recognitionby wild
Song Sparrows.Behav. Ecol. Sociobiol.14: 101106.
April1986]
Changes
inNest-defense
Intensity
SMITH,$. 1950. The Yellow Wagtail. London, Collins Press.
327
WEATHERHEAD,
P. J. 1979. Do Savannah Sparrows
commit the Concorde fallacy? Behav. Ecol. So-
STOKES,
D.W. 1979. A guide to the behavior of com-
ciobiol.
mon birds. Boston, Little, Brown & Co.
TRIVERS, R.L. 1972. Parental investment and sexual
1982. Risk-taking by Red-winged Blackbirds and the Concordefallacy. Z. Tierpsychol.
selection.Pp. 139-179in Sexualselectionand the
descentof man (B. Campbell,Ed.). Chicago,A1dine.
5: 373-381.
60: 199-208.