The Auk 113(4):849-857, 1996
PLUMAGE
COLORATION
EXPERIMENTS
AND
WITH
CONSPICUOUSNESS
THE PIED
IN BIRDS:
FLYCATCHER
SVEIN DALE • AND TORE SLAGSVOLD
Universityof Oslo,Department
of Biology,
P.O. Box1050Blindern,N-0316 Oslo,Norway
ABstRAcT.--Hypotheses
concerningpredation, sexualselection,and communicationin
birds often assumethat individuals differ in conspicuousness.
However, few studieshave
testedthisby measuringthe response
of individualsthatreceivethe signals.We investigated
the effectof plumage colorationon conspicuousness
by presentingcagedPied Flycatchers
(Ficedulahypoleuca)
to unmated territorial malesand measuringthe time until a response
(either courtshipor aggression)from the territorial male wasobserved.In the firstexperiment,
we presentedcagedmalesand femalesat variousdistances
and degreesof habitatopenness.
Cagedbirds elicited responsesmore rapidly in open than in closedhabitats,and response
time increased with distance. Black-and-white-coloredmales seemed to trigger a response
morerapidlythanbrown-colored
females,at leastin openhabitats.In the secondexperiment,
we presentedbright-coloredmales,dull males,and bright-painteddull males.All trialswere
madeat the samedistance.Cagedbirds againelicited more rapid responses
in open than in
closedhabitats.Bright-coloredmalesseemedto trigger a responsemore quickly than dullcoloredmales,but only in closedhabitats.Paintedbirdselicitedresponse
timesintermediate
to thoseof bright and dull males.The differencein interactionbetweencolorand habitatin
the two experimentsis difficult to explain but may have been causedby differencesin
backgroundrelated to seasonaldevelopment of vegetation.One possibleproblem with experimentson conspicuousness
is that the responsemay be delayed after the receiverhas
detectedthe signal.Field observations
of the behaviorof territorialbirdssuggested
that this
probablydid not confoundthe resultsof our study.Thus,the resultssupportthe intuitive
but previouslyuntestedhypothesisthat a bright and contrastingcolorationmakesbirdsmore
conspicuousto conspecifics.
Received
30 December
1995,accepted
10 April 1996.
GREAT VARIATION exists in animal
coloration
Testingwhether variousplumagecolorsand
patternsdiffer in conspicuousness
is a difficult
evolved in response to predation (Cott 1940; task. Birds of different color may differ in deBakerand Parker 1979;Harvey and Paxton1981; tectabilityto humans(G6tmarkand Unger 1994,
both amongand within species.This may have
Guilford 1990; Endler 1991; Dumbacher et al.
G6tmark and Hohlf•ilt 1995), but bird vision is
different from and perhapsbetter than that of
humans(Goldsmith1990).Ideally, responseby
1991, Sartre et al. 1994; review in Andersson
receiversof the signal (e.g. color), such as a
1994), intraspecific communication (Rohwer conspecificor a predator, should be measured
1975,1982;Whitfield 1986;Slagsvoidand Lifjeld (Bennett et al. 1994).Experimentsalsomust be
1988; Butcher and Rohwer 1989), or for species designedto ensurethat a responsecan be obrecognition (Rowland 1979, Endler 1983, An- servedoncethe subjecthasreceivedthe signal.
dersson 1994). Several of these hypothesesasPied Flycatchers (Ficedulahypoleuca)vary in
sume that an individual with bright or con- plumage color. Males range from black-andtrastingcolorationis conspicuous(i.e. more eas- white (i.e. "bright") to brown (i.e. "dull"),
ily discovered)than one with a dull and cryptic whereas all females are brown (Drost 1936).
coloration (Cott 1940, Parker 1983, Slagsvoid Males benefit from a bright plumage because
and Lifjeld 1988,Butcherand Rohwer 1989,En- they are preferredby femalesas mates(Sartre
dlet 1991).
et al. 1994),but bright malesseemto pay a cost
1992; G6tmark 1992, 1993, 1995), sexual selection (Hamilton and Zuk 1982, Parker 1983, Hill
in termsof a higher predationrisk (Slagsvoid
et al. 1995; but see G6tmark 1992, 1993, 1995).
• Presentaddress:Agricultural University of Nor- The benefit of bright plumagemay increaseif
way, Departmentof Biologyand Nature Conserva- bright individuals are more easily discovered
tion,P.O. Box5014,N-1432ks, Norway.E-mail: by intruding males and prospectingfemales
[email protected]
(Slagsvoidand Lifjeld 1988).However,the cost
849
850
DALE
AND$LAGSVOLD
of bright plumagemay increaseat the sametime
if bright malesalsoare discoveredmore easily
by predators.An unresolved issueis whether
plumage coloractually influencesconspicuousness to other
birds.
The purposeof this studywasto testwhether
bright and dull Pied Flycatchersdiffer in conspicuousness
to conspecifics.We testedthis by
measuringthe elapsedtime before wild males
respondedto cagedbirds of different colors.In
addition,
we varied distance between sender
and receiver of signalsand habitat opennessto
see whether
these factors
affected
the time
to
response.
An important question is whether response
time really reflectsthe time it takesthe receiver
to detect the signal. A receiver could delay a
responseafter detecting a signal. In order to
addressthis question we also examined the behavior of the receiversafter they had responded
to caged birds. We predicted that if a signal is
respondedto immediately,then intensity of the
responseshould not vary with habitat or distance.However, if responseintensitydecreases
with decreasinghabitat opennessor longer distances, this could indicate that the receiver of
the signal also may have delayed the response
time.
STUDY AREA AND METHODS
Studyarea.--The study was done in two woodland
areas(Sinober and Brennaplots) near Oslo in southeasternNorway during the breedingseasons
of 1991
and 1993. The study areasconsistedof a mixture of
deciduousand coniferousforest of varying degrees
of openness.Only maleswere tested,and they were
unmated
and defended
one or more nest boxes. If a
[Auk,Vol. 113
originally had a dull plumage(colorscores5 and 6.5)
but were dyed black with Nyanzol dissolvedin 10%
hydrogenperoxidein orderto resemblebright males.
The painting madebrown partsof the plumageblack,
but becausethe whitish patchesof dull malesare not
as large and pure white as thoseof bright males,the
black-painted
maleshadlesscontrasting
plumagethan
naturally bright males.
The 10 cagedbirds were capturedshortly after arrival on the breeding areas.They were held in captivity during the periodof presentations
and released
afterwards.During mostof the time in captivity, the
birds were kept in wooden cageswith a front of wire
netting, and they were given unlimited accessto
mealwormsand water. The birdswere put in the wire
netting cagesduring presentationsand were provided with mealworms.During presentationsthey behaved calmly and gave no callsbut moved around in
the cage.As far as we couldjudge,all birds behaved
similarly in the cage.
Presentations.--The
cagewas fixed to the trunk of a
tree about 1.5 m abovethe ground and was covered
with a blanket connectedto a string. The observer
was positionedat a distanceof about 20 m from both
the cage and the nest box of the male to be tested.
After fixing the cagein place,the observerwaited at
least3 min beforestartingthe experimentby pulling
the string and exposingthe cage. We required that
the test male was in normal songactivity and close
to the nest box (less than about 10 m away) for the
experimentto start.The time elapsinguntil the male
respondedto the cagedbird was recorded.Unmated
male Pied Flycatchersrespond to caged birds with
bright plumage (bright malesor black-paintedbirds)
with aggressivebehaviorsthat include flying to the
cage,whereasthey respondto dull birds (femalesor
dull males)with courtshipbehaviorsthat include flying to the nestboxopeningand giving enticingcalls
(Slagsvoid
andS•tre 1991,S•tre andSlagsvoid1992).
We used either of thesetwo responsesas evidence
that the male had discoveredthe cagedbird. The
male defended more than one nest box, the trial was
duration
done at the preferrednestbox of that maleasjudged
by daily observations.
Cagedbirds.--Totestthe influenceof plumagecolor,
we presentedPied Flycatchersin a wire netting cage
(24 x 33 x 35 cm; netting diameter 1.6 mm, spacing
1.3 cm) at varying distancesfrom a nestbox and recordedthe time elapsinguntil responseby the resi-
respondedwithin that time, then we assumedthat
the cagedbird had not beendiscovered.We excluded
trials where other Pied Flycatchersappeared.
In the first experiment(1991),the cagewasplaced
dent male. In 1991, two different males, both of color
score3 (Drost 1936), and two different females were
of each trial was 30 min. If the male had not
at four different
distances from the nest boxes of males
(i.e. 5, 10, 20 and 40 m). We used three kinds of hab-
itats that differed in vegetationdensity in the area
between the cageand the nest box: (1) open (cage
wasfully visiblefrom the nestboxwith no interfering
vegetation);(2) partly open (cagewasvisible from the
usedin the cage.The Drostcolorscoreclassifies
male
plumagecolor on a scalefrom 1 (bright) to 7 (dull). nest box but somebushesor treesobstructedthe view);
Males with scoresof 1-3 can be describedas bright and (3) closed(cagewasdifficult to seefrom the nest
with contrastingblack-and-whiteplumage,whereas box due to many treesand bushesbut could be seen
males with scores 5-7 are dull brownish and femalemore easily from other directions).In the secondexlike. In 1993,two bright males(both of color score periment (1993), all trials were done at a distance of
1.5) and two dull males (color scores 5 and 7) were
10 m from the nestbox,and in only two habitats(open
presented.In addition, we presentedtwo malesthat or closed).Trials were conductedfrom 18 May to 17
October1996]
Coloration
andConspicuousness
851
June 1991, from 0710-1500, and from 1 to 10 June
The use of only a limited number of cagedbirds in
presentationsalso could be a sourceof pseudoreplisunshineand in cloudyweather, but not during rain. cation. We restrictedthe number of cagedbirds for
Experiment1 consistedof 85 trialsat the nestboxes ethical reasons,becausekeeping them in captivity
of 30 different malesin the Sinober and Brenna plots. sometimespreventedthem from breedingduring that
Individual males were tested 1-6 times (median of 3).
year. Using different cagedbirds in every trial would
Resident males tested more than once always were
have required 175 caged birds, whereas we used a
exposedto different combinationsof the color (sex) total of 10 birds. However, as stated above, there was
of the cagedbird, distance,and habitat,and they were no evidence that the resident males reacted differnot testedmore than onceper day. Experiment2 con- ently to individual cagedbirds.Therefore,this kind
sisted of 90 trials at the nest boxes of 15 different
of pseudoreplicationprobably did not bias our remales in the Brenna plot (all males present). Each suits.
male was tested with each of the six different comIn the 1991experiment,we controlledfor the effect
binations of color (bright, dull, and black-painted) of habitatwhen analyzingthe effectof distance(and
and habitat (open and closed).This designpermitted vice versa)by using Kendall partial rank-ordercorpairwisecomparisons
of the responseof malesto each relation (Siegel and Castellan 1988). Although we
wished to control for the effect of distance and habitat
type of presentation.The order of presentationsdiffered among malesso that one-third of them were simultaneously to assessthe independent effect of
first presentedwith a bright bird, one-third with a plumage color, no nonparametrictest is available for
dull bird, and one-third with a black-paintedbird. suchtests.Instead,we performeda three-factorANOHalf of the trials of each color type were first done VA to obtainan idea of the relative effectsof plumage
in openhabitat,the other half in closedhabitat.Trials color, distance,and habitat on the time to response.
alternated between open and closedhabitat for each Becauseof unequal sample sizes and differencesin
male. Twelve different orders of presentation were
distribution of variables,the data violate the assumpused for the 15 males; thus, three orders were used
tions of ANOVA, but they may nonethelessprovide
on two maleseach.The design ensuredthat all types someuseful insight. To avoid missingcells, the disof presentationswould have about an equal proba- tance variable was grouped into two levels, short (5
bility of being done first, second,and so on in the and I0 m) and long (20 and 40 m) distance.If the
sequenceof six trials for each male. Finally, as in residentmale did not respondto the cagedbird durexperiment I, no males were tested more than once ing the trial (30 min), the time to responsewas asper day. Analysesrevealedthat studyarea,date, time signed an arbitrary value of 31 min. The time to reof day of the trial, order of trial, and the identity of sponsewaslog-transformedto approacha normal disthe caged bird had no significant effects in either tribution. We alsoperformedANOVAs with arbitrary
values of 60 min instead of 31 min for trials with no
experiment (P > 0.11 in all tests;data not shown).
Response
intensity.--To test whether the response response,but the resultswere similar.ANOVAs using
time waslikely to reflecttime to detectthe cagedbird, each male only once (to avoid pseudoreplication)
we measuredthe intensity of the responseduring a against caged males and females, respectively,pro1-min period immediately after the time of response. duced results similar to the total data set.
This was done in 1991on all trials in the Brennaplot
and in trials after 21 May in the Sinober plot. We
recordedthe time spenton or inside the nest boxand
RESULTS
the time spent on the cage.The total of these times
was usedas a measureof responseintensity.
Statisticalanalyses.--Weused one-tailed tests be- RESPONSE INTENSITY
causeall existing theory predictsthat the cagedbird
In the 1991 experiment, the responseintenshould be discoveredsooner if it is bright, in open
1993, from 0625-1315. Trials were conducted both in
habitat, or at a short distance than if it is dull, in
closed habitat, or at a long distance.In testswhere
the value of the z-statisticis given, negativez-values
(and thus low P-values) indicate that the trend was
in the predicteddirection,and positivez-values(and
thus high P-values) indicate that the trend was opposite the predicted direction. All other tests were
two-tailed, and this is statedexplicitly in every case.
In the 1991 experiment, some males were tested
more than once (see above). To avoid pseudoreplication, we also performed the correlation analyses
with a reducedsamplesize, including eachmale only
once for each sex of the caged bird (the first trial).
sity of resident males to cagedmaleswas unaffected by both habitat (Kendall partial rankorder correlation,T, = -0.05, n = 21, P > 0.25;
distanceheld constant)and distance(T, = 0.07,
n = 21, P > 0.25; habitat held constant). When
the cagedbird was a female, habitat again had
no effecton responseintensity (T, = -0.02, n
= 35, P > 0.25; distance held constant), whereas
responseintensity decreasedwith distance(T,
= -0.27, n = 35, P = 0.01; habitat held constant).
This suggeststhat responsetime reflectsdetection
time.
852
DALEANDSLAGSVOLD
[] B•ght
[] Dull
6
[Auk,Vol. 113
[] Bdght
12
8
15
[] Dull
9
14
2O
10
14
10
28
o
o
11
E
0,1
0,1
5
20
40
Open
Cloaed
Partly
open
Olatance (m)
FiG. I. Time (logarithmic scale)elapsinguntil responseto the cagedbird by resident male Pied Flycatchersin relation to distancebetweenthe cagedbird
and the nest box of the male, and the plumage color
of the cagedbird (bright male or dull female). Bars
indicate median values, lines the 25-75% interval.
Habitat
FIG.2. Time (logarithmic scale)elapsinguntil responseto the cagedbird by residentmale Pied Flycatchersin relation to habitat and the plumage color
of the cagedbird (bright male or dull female). Bars
Trials in which there was no responseto the caged indicate median values, lines the 25-75% interval.
bird were assigneda value of 31 min. The upper 75% Trials in which there was no responseto the caged
value for cagedmalesat I0, 20 and 40 m distance,and bird were assigneda value of 31 min. The median
for females at 40 m, was 31 rain (i.e. no response). time was 31 min (i.e. no response)for both colorsin
Numbers at top indicate samplesizes.Data from the the closed habitat. Numbers at top indicate sample
sizes.Data from the 1991 experiment.
1991 experiment.
EFFECT OF DISTANCE AND HABITAT
Experiment/.--Both distanceand habitat had
significanteffectson responsetimes in 1991trials with male and female cagedbirds. Response
times were quicker when the caged bird was
presentedcloserto the nestbox than when farther away (Kendall rank-order correlation,T =
0.42, n = 85, P < 0.0001; Fig. 1), and when it
was in open versus closedhabitats (T = 0.59, n
= 85, P < 0.0001;Fig. 2). Distancewas still significant when controlling for habitat openness
samplesize, including each male only once for
each sex of the caged bird. The results were
similar to thosewith full samplesize (data not
shown). Second,we comparedthe responsesof
males that were tested at least twice in one kind
of habitat or at one distance, and with the same
sex of the caged bird in both trials. Response
times
were
fastest for the shortest
distance
in
12 of 12 males (trials in the same kind of habitat;
Wilcoxon matched-pairssigned-rankstest, z =
-3.06, P = 0.001). Response times also were
more rapid in the most open habitat in five of
(Kendallpartialrank-ordercorrelation,T, = 0.24, six males (trials at the same distance; z = - 1.57,
n = 85, P < 0.001), and the effect of habitat
P = 0.058, same test). In conclusion,analyses
opennesswasstill very strongwhen controlling
showed
that both
habitat
and distance
had in-
for distance(T, = 0.51,n = 85, P < 0.001).Fur- dependent effects on the time elapsing until
resident males respondedto a cagedbird.
Experiment
2.--Responsetime to a cagedbird
effect of color in addition to habitat or distance.
was again more rapid in open than closedhabThe effectsof distance(cagedmales:T, = 0.22, itatsin 1993trials with only males(bright, dull,
n = 28, P < 0.05;cagedfemales:T, = 0.24,n = and black-painted;Fig. 3). Responsetimeswere
57, P < 0.01) and habitat (cagedmales:T, = faster in open than in closedhabitat for 12 of
0.65,n = 28, P < 0.001;cagedfemales:T• = 0.47, 15 males tested on bright males (Wilcoxon
n = 57, P < 0.001) were significant for each sex. matched-pairssigned-rankstest, z = -2.33, P
thermore,we analyzed the trials with male and
female cagedbirds separatelyto control for the
Because
some
males
were
tested
more
than
= 0.010), 13 of 15 males tested on dull males (z
once, we analyzed the data in two additional = -2.39, P = 0.008), and 13 of 15 males tested
ways to avoid pseudoreplication.First, we per- on black-painted dull males (z = -2.56, P =
formed the correlation analyseswith a reduced 0.005).Distancedid not vary in this experiment.
October
1996]
Coloration
andConspicuousness
EFFECT OF PLUMAGE COLOR
853
[] Bdght
[] Dull
[] Painted
Experiment/.--We compared responsetimes
when the cagedbird was a bright male versus
a dull female. The color of the cagedbird had
no effect (Mann-Whitney U-test, z = -0.99, n
= 85, P = 0.16). Color still had no significant
effectwhen we analyzedthe dataseparatelyfor
each distance (5 m: z = -1.64, n = 20, P = 0.051;
10 m: z = 0.03, n = 26, P = 0.51; 20 m: z = -0.15,
n=21, P=0.44;40m:z=-0.10,
n= 18, P=
0,1
Open
Closed
0.46; Fig. 1). However, an interaction between
habitatopennessand the colorof the cagedbird
Hsbltet
seemedapparent (Fig. 2). Bright-coloredmales
F•G. 3. Time (logarithmicscale)elapsinguntil reelicited more rapid responsesthan did dull-colored females in open habitat (z = -2.58, n = sponseto the cagedbird by residentmalePied Fly39, P = 0.005) but not in partly open habitat (z catchers(n = 15 for each type of presentation) in
= -1.04, n = 23, P = 0.15) or in closed habitats
relationto habitatandthe plumagecolorof the caged
bird (bright male, dull male, or black-painteddull
(z = 1.44,n = 23, P = 0.93). Splitting of data to male). Barsindicate median values, lines the 25-75%
perform testsof each combinationof distance interval. Data from the 1993experiment.
and habitat was not possible becausesample
sizes became
too small.
The datawere further analyzedby usingeach
male testedonly oncefor eachsexof the caged
bird in order to avoid pseudoreplication.The
results were similar to those with full sample
size (data not shown). Only four caseswere
available where trials were made with both male
and female cagedbirds at the samedistancein
the same habitat with
the same test male. Test
the open habitat(only 8 of 15 malesresponded
to the bright male sooner;z = -0.80, P = 0.21;
Fig. 3).
Black-painteddull maleswere expectedto be
discoveredmore rapidly than dull males.This
was the case in 10 of 15 comparisonsin the
closedhabitat (Wilcoxon matched-pairssignedranks test, z = -1.42, P = 0.078) but in only 6
of 15 comparisons
in the open habitat(z = 0.23,
P = 0.59;Fig. 3). We expectedthat paintedmales
cases.
would take somewhatlonger than bright males
An ANOVA (see Methods) showed that the to be discoveredbecausepainted maleshad less
time elapsinguntil the residentmale responded contrast in their plumage. Responsetime to
to the cagedbird was most strongly influenced bright maleswas fasterfor 10 of 15 malesin
by the kind of habitat (F = 35.9, df = 2 and 73, closedhabitat (z = -2.33, P = 0.010) and for 11
P < 0.0001). Bright males elicited faster re- of 15 males in open habitat (z = -2.22, P =
sponsesthan dull females(F = 5.02, df = 1 and 0.013; Fig. 3).
73, P = 0.014),indicatingan independenteffect
of the color of the cagedbird. Distancealsohad
DISCUSSION
a significantindependenteffect(F = 3.75, df =
Animal sensesdiffer markedly between dif1 and 73, P = 0.029).The ANOVA alsosuggested
that there was an interaction between habitat
ferent systematicgroups.For example,mamand plumage color (F = 3.59, df = 2 and 73, P mals and birds exhibit clear differences in vi= 0.017).Thus, the ANOVA supportedthe con- sion (Goldsmith 1990). In order to assessthe
effect of distance, habitat, and other factors on
clusionsfrom the nonparametricanalyses.
responsesof the actual reExperiment
2.--In this experiment,the effect conspicuousness,
of plumage color on responsetime was inves- ceiversof the signalmustbe measured,and tests
tigated by pairwise comparisons.Bright males mustreflectthe rangeof variationthat is releelicited quicker responsesthan dull males by vant to the receiver (Bennett et al. 1994).
13 of 15 males tested in the closed habitat (WilConspicuousness
or motivation?--Weattempted
coxon matched-pairs signed-ranks test, z = to study how distance,habitat, and plumage
-2.61, P = 0.005), but this was not the casein color influencedconspicuousness
of Pied Flymalesrespondedquickerto cagedmalesin two
casesand to caged females in the other two
854
DALEAND$LAGSVOLD
[Auk,VoL 113
catchersto conspecifics.To assessconspicuous- should delay a responseto either sex.Birdspernesswe recordedhow quickly the cagedbirds ceived as females should be courted immediwere detected. We used the first observable reately, whereas birds perceived as intruding
sponse(courtshipor aggression)
asevidencethat males should be chasedaway as soon as possithe caged bird had been discovered,but such ble. If anything, a prospectingfemalemight be
responsesdo not necessarilycorrespondto the more important for male fitness than a male
actual time that the resident male detected the
believe that using the time of responseas a
intruder. The female may becomethe mate of
the male, whereas intruding males are rarely
able to displace a resident male. If males are
more motivated to respond to caged females
and therefore delay responsesto caged males,
measure
then
caged bird. A receiver of a signal could delay
the responsefor a variety of reasons.However,
there are two lines of evidence
of the time of detection
that lead us to
is reasonable.
females
should
be "discovered"
sooner
First, field observations of resident males dur-
than males.The oppositewas found in the presing trials suggestedthat they often first re- ent study, however. In conclusion,we think it
spondedafterhavingmovedto a positionwhere is unlikely that the resultsof the presentstudy
the cagedbird was easier to see,and then the were biased by variation in motivation to reresponsefollowed almost immediately.It was spond.
quite rare that a residentmale respondedafter
Effectof distance
and habitat.--Theresultsaphaving perched in one place for some time. peared to confirm the prediction that distance
However, we did not record and quantify this and habitat affected conspicuousness
to consystematically,which shouldbe done in future specifics;birds respondedfaster in more open
studies.
habitat and when caged birds were placed at
Second,the resultsshowedthat the response closerdistances.Thesemay seemto be obvious
intensity toward caged males was not related results, but such effects nevertheless need to be
to distanceor habitat, which might have been verified experimentally.The resultsindicated
expectedif motivation differed.The resultscon- that the absolute values of differences were
cerning response intensity toward caged fe- rather large and may have important consemales showed that habitat had no effect, wherequencesfor severalaspectsof avianecologyand
as responseintensity decreasedwith distance. behavior (see below). Furthermore, the demThus, the effectof distanceon time to response onstration of predicted effectsof distanceand
with cagedfemalesmay have been causedby a habitat suggeststhat the experimental design
low motivation to respond to females at long worked and should be appropriate for testing
distances.
the hypothesisthat plumage color affectsconIt is more difficult to assess whether
the respicuousness.
ceiver's motivation to respond varied with
Effectof plumagecolor.--In 1991,bright indiplumagecolorof the cagedbird. Residentmales viduals were discovered sooner than dull inrespond to bright birds with aggression,even dividuals, at least in open habitat. Presentation
if they are black-paintedfemales,whereasdull
of a bright male and a dull female to test the
effect of plumage color might be considered
courted, even when they are males (Slagsvoid ambiguous because differences in responses
birds are treated as females and, hence, are
and Sartre 1991, Sartre and Slagsvoid 1992).
could be an effect of sex rather than color. How-
Thus,differencesin responsein relationto color
could be causedby differencesin motivation to
respond to a bird that is perceived as a male
versusa bird that is perceivedas a female. We
cannotcompletelyrule out the possibilitythat
ever, previous studieshave shown that plumage color functions as a cue for sex determination (Slagsvoid and Sartre 1991, Sartre and
Slagsvoid 1992). Territorial males respond to
it is important to respond more quickly to an
both females and dull males with courtship,
whereas bright males, painted females, and
painted dull males are respondedto with agfield observationssuggestedthat responsesto gression.Thus, differencesin responseprobaall types of cagedbirds were rapid and strong bly are an effectof color, not of the real sexof
once the test male was in a position to discover the cagedbird.
the caged bird. Furthermore, it is difficult to
Evenso,we usedonly cagedmalesin the 1993
understand why a territorial, unmated male experiment to avoid possiblesexdifferencesin
intruder
of one sex than the other. However,
October
1996]
Coloration
andConspicuousness
the behavior of the cagedbird. We usedboth
bright and dull males,and in addition used a
black-painteddull male to control for any behavioral differencesbetween bright and dull
males. The results showed that bright males
elicited faster responsesthan dull ones. However, in this experiment this differencewas evident only in the closedhabitat, in contrastto
the first experiment where the difference was
found only in the open habitat.
The surprisingdifferencein interactionbetween plumage color and habitat in the two
experiments is difficult to explain. One possibility is that there was an interactionbetween
plumagecolorand the colorpatternof the background that had different effectsin open and
closedhabitats.The only obviousdifferencebetween the two experimentswas in time of season;trials were done on average10 daysearlier
in 1991 than in 1993. In addition, the seasonal
developmentof the vegetationwas at leastone
week later in 1991than in 1993becauseof very
low temperaturesin the end of April and in
May. Therefore, trials done in 1991 were more
against a backgroundof dead leaves on the
ground and naked branchesand tree trunks,at
leastin the early trials. In contrast,mosttrials
done in 1993 were against a predominantly
greenbackground;the leavesof the treeswere
well developedand the ground vegetationhad
emerged.The color pattern of the background
is expectedto affectthe conspicuousness
of an
855
Flycatchers
themselves
dependson whetheran
individual is on the ground or in a tree.
Implications
of differential
conspicuousness
in relationto color.--In birds,brightly coloredmales
seemto be at an advantagerelativeto dull males
becausethey are preferredby femalesas mates
(Hill 1991, S•tre et al. 1994). However, females
only visit a restrictedsubsetof malesin an area
(Benschand Hasselquist1992,Dale et al. 1992).
Males will therefore benefit from any character
that increasestheir chancesof being discovered
by females(Parker1983).Songis known to attract females (Eriksson and Wallin 1986) and
may be the mostimportantmeansfor malesof
signaling their presenceto females.However,
a brightplumagecoloralsomayhelpto increase
detectabilityas shown by the presentstudy,
though the relative importancemay be small
comparedwith song.This providesa mechanism for the evolution of bright coloration
through intersexualselectioneven in species
where femalesdo not prefer bright males(Parker 1983).
Malesalsomaybenefitfrom a bright plumage
colorin the contextof territory defensebecause
the bright plumagemay signal the presenceof
territory owners.This would be advantageous
if intrudersoften retreatoncethey discoverthat
a territory is already occupied(Slagsvoidand
Lifjeld 1988). In contestsover territory owner-
ship, residentsoften will have an advantage
(Maynard Smith and Parker 1976),and owners
individual (Endler 1984, Marchetti 1993, G6tmay thereforeoften avoid conflictssimply by
mark and Unger 1994,G6tmark and Hohlffilt signalingtheir presenceto intruders.
The resultsalsomayhaveimplicationsfor the
1995) and may have played a role in our experiments, though we did not quantify the relation between plumage colorationand predation.Slagsvoidet al. (1995)foundthat among
backgroundcoloration.
Nevertheless, in both experiments individPied Flycatchers,predationrisk from Eurasian
uals with a bright plumage color had an in- Sparrowhawks (Accipiternisus)is highest on
creasedconspicuousness
relative to dull indi- bright males.This makessenseif a bright plumto the sparrowviduals.Another study of detectabilityof Pied age increasesconspicuousness
Flycatchers
with humansasobservers
(G6tmark hawk as well as to conspecifics.Bright males
and Hohlffilt 1995) showed that humans de- may suffermorepredationsimplybecausethey
tected mounts of male Pied Flycatchersmore are more easilydetected.A rigoroustestof this
rapidly than mountsof femaleswhen mounts hypothesisshould measurehow quickly the
were placedon the ground,but not when they sparrowhawk itself discoversbright and dull
were placed in trees. G6tmark and Hohlffilt males.However, studieswith stuffed Pied Fly(1995) suggested that the black-and-white catchersshowed that sparrowhawksand other
plumageof malesare an exampleof disruptive predatorsattack dull mounts more often than
coloration that makes males difficult to see
bright ones(G6tmark1992, 1993,1995),assugagainsta backgroundof contrastingpatchesof gestedby the unprofitableprey hypothesis(Balight and shadowin trees.Further studiesare ker and Parker 1979).In theseexperiments,the
neededto testwhether conspicuousness
to Pied bright and dull modelswere placedon exposed
856
DALE
ANDSLAGSVOLD
[Auk,VoL 113
J.P., B. M. BEEHLER,
T. F. SPANDE,H. M.
sitesand/or near eachother sothat the predator DUMBACHER,
GARRAFFO,
ANDJ. W. DALY. 1992. Homobatrachmay have discoveredboth models very easily
otoxin in the genusPitohui:Chemicaldefensein
and almost simultaneously (see G6tmark 1995
birds? Science 258:799-801.
for alternative interpretations of his experiENDLER,J. A. 1983. Natural and sexual selection on
ments). Thus, it is still possiblethat live bright
color patternsin poeciliid fishes.Environmental
males are more conspicuous to sparrowhawks
Biology of Fishes9:173-190.
than are dull males under natural conditions,
ENDLER,
J.A. 1984. Progressivebackgroundmatch-
but this requiresfurther study.
ing in moths,and a quantitativemeasureof crypIn conclusion,the resultsof our studysuggest
sis.BiologicalJournalof the LinneanSociety22:
187-231.
that a bright and contrastingcolorationincreasENDLER,
J. A. 1991. Interactionsbetween predators
esconspicuousness
to conspecifics.
This concluand prey. Pages169-196 in Behaviouralecology
sionhasimplicationsfor the evolutionof bright
(J. R. Krebs and N. B. Davies, Eds.). Blackwell
plumage and/or sexual dichromatismin relaScientific Publications, Oxford.
tion to mate attraction, territory defense, and
ERIKSSON,
D., AND L. WALLIN. 1986. Male bird song
perhapspredation.However, we need more deattracts females--a field experiment. Behavioral
tailed studiesto be able to explainthe difference
Ecologyand Sociobiology19:297-299.
in interaction
between
color
and habitat
ob-
GOLDSMITH,T. H.
1990. Optimization, constraint,
served in the two experiments.There also is a
and history in the evolution of eyes. Quarterly
need for similar studiesusing a predatorsuch
Review of Biology 65:281-322.
asthe CurasianSparrowhawkasthe receiver of G(•TMARK,F. 1992. Anti-predator effect of conspicuous plumage in a male bird. Animal Behaviour
the signal.
44:51-56.
G(•TMARK,
F. 1993. Conspicuouscolorationin male
birds is favoured by predation in some species
ACKNOWLEDGMENTS
and disfavoured in others. Proceedingsof the
Royal Societyof London SeriesB 253:143-146.
We are grateful to J. Allen, Frank G6tmark, Geoff
Hill, JanT. Lifjeld, Karen Marchetti, and Glenn-Peter C•TMARK,F. 1995. Black-and-whiteplumagein male
Pied Flycatchers(Ficedulahypoleuca)
reducesthe
S•etre for comments on the manuscript. Financial
risk of predationfrom Sparrowhawks(Accipiter
support was provided by the Norwegian Research
Council and from the Nansen Foundation.
nisus)during the breeding season.Behavioral
Ecology6:22-26.
C_•TMARK,
F., ANDA. HOHLF.KLT.
1995. Brightmale
LITERATURE CITED
plumageand predationrisk in passefinebirds:
Are males easier to detect than females? Oikos
ANDERS$ON, M.
1994. Sexual selection. Princeton
74:475-484.
G•TMARK,F., ANDU. UNGER.1994. Are conspicUous
BAKER,R. R., AND G. A. PARKER. 1979. The evolution
birds profitable prey? Field experiments with
hawksand stuffedprey species.Auk 111:251-262.
of bird coloration.PhilosophicalTransactionsof
T. 1990. The evolution of aposematism.
the Royal Societyof London SeriesB 287:63-130. GUILFORD,
BENNETT,A. T. D., I. C. CUTHILL,AND K. J. NORRIS.
Pages23-61 in Insect defense:Adaptive mecha1994. Sexual selection and the mismeasure of
nismsand strategiesof prey and predators(D. L.
color. American Naturalist 144:848-860.
Evans and J. O. Schmidt, Eds.). State University
University Press,Princeton, New Jersey.
BENSCH,S., AND D. HASSELQUIST.1992. Evidence for
active female choice in a polygynouswarbler.
Animal
Behaviour
44:301-312.
BUTCHER,G. S., AND S. ROHWER. 1989. The evolution
of New York Press, New York.
HAMILTON, W. D., AND M. ZUK. 1982. Heritable true
fitnessand bright birds:A role for parasites?
Science 218:384-387.
of conspicuous and distinctive coloration for
HARVEY,P. H., AND R. J. PAXTON. 1981. The evolu-
communicationin birds. Current Ornithology
COTT,H. B. 1940. Adaptive coloration in animals.
tion of aposematiccoloration.Oikos37:391-396.
HILL, G. E. 1991. Plumage coloration is a sexually
selectedindicator of male quality. Nature 350:
Methuen, London.
DALE, S., H. RINDEN, AND T. SLAGSVOLD. 1992. Com-
MARCHETTI,
K. 1993. Dark habitatsand bright birds
6:51-108.
337-339.
illustrate the role of the environment in species
divergence.Nature 362:149-152.
MAYNARDSMITH, J., AND G. A. PARKER. 1976. The
DROST,
R. 1936. Ober das Brutkleidm•innlicher
logic of asymmetriccontests.Animal Behaviour
petition for a mate restrictsmatesearchof female
Pied Flycatchers.BehavioralEcologyand Sociobiology 30:165-176.
Trauenfliegenf'finger,Muscicapahypoleuca.Vogelzug 6:179-186.
24:159-175.
PARKER,
G.A.
1983. Mate quality and mating deci-
October1996]
Coloration
andConspicuousness
sions. Pages141-166 in Mate choice (P. Bateson,
Ed.). CambridgeUniversity Press,Cambridge.
RO•WER,S. 1975. The social significanceof avian
winter plumagevariability.Evolution29:593-610.
RO•WER, S. 1982. The evolution of reliable and un-
reliablebadgesof fighting ability. AmericanZoologist 22:531-546.
Row•lr•, W.J. 1979. The use of color in intraspecific communication.Pages379-421 in The behavioral significance of color (E. H. Burtt, Ed.).
Garland Press, New York.
SI_•GSVOLD,
T., AI•V J. T. LII•JœLV.1988. Plumage colour and sexual selectionin the Pied Flycatcher
Ficedulahypoleuca.
Animal Behaviour36:395-407.
S•svoLv,
T., •v
G.-P. S•ETRE. 1991. Evolution of
plumage color in male Pied Flycatchers(Ficedula
hypoleuca):
Evidencefor female mimicry. Evolution 45:910-917.
S•EtRE,G.-P., S. DALE, •D
T. S•SVOLV.
1994. Fe-
male Pied Flycatchersprefer brightly coloured
males. Animal
Behaviour
48:1407-1416.
S•EtRE, G.-P., At•V T. SLA•SVOLV. 1992. Evidence for
SIEGEL,
S., A•D N.J. CASTELr•a•.1988. Nonparametric statistics for the behavioral
857
sciences. McGraw-
Hill, New York.
SLAGSVOLD,T., S. DALE, AND A. KRUSZEWlCZ. 1995.
Predationfavourscrypticcolorationin breeding
malePied Flycatchers.
Animal Behaviour50:11091121.
sex recognition by plumage colour in the Pied
Flycatcher,Ficedula
hypoleuca.
Animal Behaviour
44:293-299.
W•rr•ELr•, D. P. 1986. Plumagevariability and territoriality in breeding Turnstone Arenaria in-
terpres:Statussignallingor individual recognition? Animal
Behaviour
34:1471-1482.
Associate Editor: T. E. Martin
(continued
fromp. 848)
Jehl, Jr., Paul A. Johnsgard,L. ScottJohnson, JamesM. Ramakka,JohnH. Rappole,PamelaC.
David W. Johnston*, Richard F. Johnston*,
Rasmussen,Jeremy M.V. Raynet, Roland L.
Donald W. Kaufman, Ellen D. Ketterson, Rolf
Redmond,Kerry P. Reese,Mary L. Reid, J. V.
Reinsen*,WallaceB.Rendell,JohnD. Reynolds,
RobertE. Ricklefs*,JamesD. Rising,Chandler
R. Koford,CareyW. Krajewski*,Gary L. Krapu,
Donald E. Kroodsma*,David B. Lank, Wesley
E. Lanyon, Mark S. Laska,David S. Lee, Robert
E. Lemon,Marty L. Leonard,C. M. Lessells,Jan
T. Lifjeld, StevenL. Lima*, BradleyC. Livezey,
JonathanL. Longmire*,BruceLyon, RichardE.
MacMillen, Mark L. Mallory, Karen Marchetti,
Jeffrey S. Marks*, A. G. Marshall, Monica B.
S. Robbins*, Mark B. Robbins, Robert J. Robel,
Julie R. Roberts, Scott K. Robinson*, Frank C.
Rohwer, Sievert A. Rohwer*, Robert N. Rosen-
field, Eivin Roskaft,StephenI. Rothstein*, Stephen M. Russell,PeterG. Ryan,JohnP. Ryder,
Joel A. Schmutz,JosefK. Schmutz, Stephan J.
Martella, Carl D. Marti, Thomas E. Martin, Brian
Schoech,John I-I. Schulz*, David M. Scott, Wil-
A. Maurer, D. Archibald McCallurn, Kevin J.
McGowan, Frank McKinney, Raymond Mc-
liam A. Searcy,JamesS. Sedinger*,Gilles Seutin, SteveK. Sherrod,ThomasW. Sherry,Gerald
F. Shields*,TheodoreR. Simons*,Tore Slags-
Neil*, Sue McRae, Donald B. Miles, Edward I-I.
Miller, Michael R. Miller, Anders P. Moller,
void, I-Ienrik Smith, W. John Smith, Paul R.
William A. Montevecchi,RobertD. Montgo- Sotherland, Colleen C. St. Clair, Peter B. Stamerle, Frank R. Moore, Juan Moreno, R. I.G.
Morrison, D. JamesMountjoy, Robert S. Mulvihill*,
Ronald L. Mumme*,
Edward C. Mur-
phy, Michael T. Murphy, Brent W. Murray,
Kenneth A. Nagy, Douglas A. Nelson*, Ian
Newton*,Ian C.T. Nisbet,ErikaNol, R.•. Norberg, ChristopherJ. Norment, ThomasNudds,
Nadav Nur, John P. O'Neill, Harry M. Ohlendorf, RobertD. Ohmart, LewisW. Oring, Ralph
S. Palmer,KennethC. Parkes*,JeffreyParrish*,
RobertB. Payne*, Craig M. Pease,SusanPeters,
Lisa J. Petit, JaroslavPicman, Raymond Pierotti*, PamelaJ. Pietz, Walter I-I. Piper, Darrell
W. Pogue,Alan Poole,Pilai Poonswad,Charles
R. Preston, Trevor
Price, Daniel
Promislow,
RichardO. Prum, Peter Pyle, JamesS. Quinn*,
Robert J. Raikow*, Ralph J. Raitt, C. J. Ralph,
cey*, J. M. Starck, Peter Stettenheim*, F. Gary
Stiles*, Douglas F. Stotz, Philip C. Stouffer,
BridgetJ. Stutchbury,BarbaraTaborsky,Kevin
L. Teather,John W. Terborgh,ScottB. Terrill,
Brian G. Thomas,Max C. Thompson,Terrence
R. Tiersch,Rodger Titman, Bret W. Tobalske,
Roy E. Tomlinson, G. I-Ienu Visser, Carol M.
Vleck*, RobertB. Waide, John Warham, Bryan
D. Watts, Patrick J. Weatherhead,Wesley W.
Weathers, Jon I-I. Wetton, Nathaniel T. Wheel-
wright, ChristopherJ.Whelan,KarenL. Wiebe*,
John A. Wiens, David S. Wilcove, JosephWilliams,John C. Wingfield, Hans Winkler, Peter
Woodall, Glen E. Woolfenden, Ayaka Yamaguchi, Ken Yasukawa,Ronald C. Ydenberg*,
Robert M. Zink*, Marlene Zuk*, Richard L. Zusi,
Fred C. Zwickel.