variation in antiparasitic behaviors of red-winged blackbirds in response to simulated brown

13 Pages·2014·0.23 MB·English

TheWilsonJournalofOrnithology126(3):488–499,2014
VARIATION IN ANTIPARASITIC BEHAVIORS OF RED-WINGED
BLACKBIRDS IN RESPONSE TO SIMULATED BROWN-
HEADED COWBIRDS
CAROL S. HENGER1,2 AND MARK E. HAUBER1
ABSTRACT.—Red-wingedBlackbirds(Agelaiusphoeniceus)areafrequentlyparasitizedhostspeciesofobligatebrood
parasitic Brown-headed Cowbirds (Molothrus ater). Yet, this common host does not reject parasite eggs or young. A
hypothesistoexplainthelackofcowbirdeggandnestlingrejectionisthatredwingsfrontloadtheirresponsesbyinvesting
intoantiparasiticnestdefensebehaviorstowardfemalecowbirds,relativetonon-layingmalecowbirdsandnon-threatening
otherspecies.Areviewofpriorstudiesusingcowbird-mountpresentationswithorwithoutacousticplaybackssupported
somebutnotallpredictionsofthishypothesis.Weconductedanewstudyattwogeographicallyseparatesites,inNew
YorkCity,NYandinIthaca,NY,USA,whereRed-wingedBlackbirdswerepresentedwithtaxidermicmodelsonatripod
and species- and sex-specific vocalization playbacks of female or male Brown-headed Cowbirds, female Northern
Cardinals (Cardinalis cardinalis), or a tripod and silence (control). Reactions of both sexes to mount and playback
presentationsnearknown,activenests,conductedrepeatedlythroughoutthebreedingcycle,weresupplementedwithmount
andplaybackpresentationsto maleswithoutknownnests.Red-wingedBlackbirdsrespondedmoreaggressively toward
presentationsofbothsexesoftheparasitethantothecardinalandthetripodtreatments.Inaddition,theyrespondedmore
aggressivelyatknownneststomountspresentedcloserratherthanfarther,andatnestswithnestlingsratherthaneggs.
Furthermore,aggressivereactionsweremorefrequentinIthaca,NYthaninNewYorkCity,NY.Finally,thereactionsof
femaleandmaleredwingsnearknownactivenestsandofmaleswithoutknownnestsshowedstatisticallysimilarpatterns,
implyingthatdatafromthesetwotypesofexperimentalcontextscouldbecombinedinouranalyses,andperhapsinfuture
studies, to more efficiently characterize this host’s responses to parasite presentations. Received 16 November 2013.
Accepted1March2014.
Keywords: aggression,broodparasitism,mountpresentation,nestdefense,sexdifferences.
Avian obligate brood parasites lay their eggs young. Nest desertion releases the hosts from
intootherspecies’nests,anddependonthehosts extended provisioning for the parasites (Langmore
to provision and protect their offspring (Davies etal.2003,Grim2007,Hauberetal.2013),whileit
2000).Parasitism,bydefinition,negativelyaffects also avoids the cognitive costs and errors of
the fitness of the host species through the initial discriminatingownandforeigneggs(Lotem1993,
exchange of host eggs for parasitic eggs, de- MoskatandHauber2007,Moskatetal.2014).
creased hatching success of host eggs, and the TheparasitismfrequencyofRed-wingedBlack-
competition for resources between unrelated birdsAgelaius phoeniceus (hereafter: redwing)by
chicks (Rothstein 1990, Hauber 2003a). Brown-headed Cowbirds Molothrus ater (hereaf-
Hostsemploymanystrategiestopreventbrood ter: cowbird) is exceedingly variable among
parasitism. Building nests in cavities with small different geographic locations and within popula-
entrances (Hoover and Robinson 2007), in tions across North America, and even between
enclosed spaces (Hauber 2001) or in dense cover differentyears(OrtegaandCruz1988,Freemanet
reduces the risk of being located by the parasites al. 1990, Clotfelter and Yasukawa 1999). Parasit-
(Clotfelter 1998, Hauber and Russo 2000). ism of redwings by cowbirds also varies at the
Abandoning parasitized nests (Hosoi and Roth- microhabitat scale with vegetation characteristics,
stein 2000), burying parasitic eggs (Valera et al. including decreasing parasitism rates with more
1997), and ejecting parasitic eggs from the nest distance from potential cowbird perch sites (Clot-
felter 1998), and lower parasitism rates of marsh-
(Lahti 2006) are also potential means of foregoing
nesting, densely colonial versus upland-nesting,
continuedcare for the unrelated parasitic eggs and
morelooselycolonialredwings(Peeretal.2005);
thelatterpatternisthoughttobebecauseofmore
1Animal Behavior and Conservation Program, Depart- successfulmobbing and nest defensebyredwings
mentofPsychology,HunterCollegeoftheCityUniversity
againstcowbirdswithinthemoredenselypopulat-
of New York, 695 Park Avenue, New York, NY 10065,
edcolonies(Strausberger2001).
USA.
2Correspondingauthor; Despite this variation in the local likelihood of
e-mail:[email protected] cowbird parasitism, redwings are so numerous
488
HengerandHauberNANTIPARASITICNESTDEFENSE 489
throughout North America that they are consid- harmless heterospecific birds (Robertson and
eredthehostspeciesresponsibleforproducingthe Norman 1976, Ortega and Cruz 1991, Prather et
mostcowbirdfledglings(Lowther1993).Surpris- al.1999).Critically,the responsesof redwingsto
ingly,however,redwingsdonotpreventthemany predatory birds vs. cowbirds are also quantita-
costs of cowbird parasitism through the rejection tivelydifferent(NeudorfandSealy1992;STable
of cowbird eggs or nestlings (Rothstein 1975, 1), implying that even though cowbirds do
Roskaft et al. 1990, Clotfelter and Yasukawa occasionally prey on host eggs and nestlings,
1999). This is unexpected because redwings are parasiticcowbirdsareconsideredadifferentclass
indeed able to remove egg-sized objects experi- of threat compared to other avian nest predators
mentally placed in their nests, but the birds (Yasukawa et al. 1992; S Table 1). The specific
evidently do not respond to real or experimental hypothesis, therefore, for the evolution of the
cowbird eggs by ejecting them (Ortega et al. specific patterns and timing of antiparasitic
1993). Moreover, cowbirds do not appear to use defenses (and the lack thereof) by redwings
egg mimicry or crypsis to avoid the detection of against cowbirds is that this host species front-
theparasiticegg(Ortegaetal.1993).Itispossible loads its defenses by predominately reducing the
that redwings have not yet evolved the necessary cost of parasitism through the prevention of
adaptations to eject eggs because of evolutionary parasitic egg laying in the first place rather than
lag (Wardetal.1996)or itmaybetoo costlyfor by recognizing and rejecting parasitic eggs or
redwings to evolve an ejection behavior because young (Røskaft et al. 1990, Grim 2011, Feeney
of the potential cost of damaging one or more of et al. 2012).
their own eggs while attempting to reject the Our study used experimental mount-presenta-
relativelythicker-shelledcowbirdegg(Rowheret tions to examine aggressive nest-defense behav-
al. 1989). Redwings do sometimes desert parasit- iorsemployedbyredwingstowardcowbirds.The
ized nests (Clotfelter and Yasukawa 1999), but it specific aim of the study was to determine
canbecostlyintermsoftimelostbuildinganew whether patterns of host behaviors represent
nest(ClotfelterandYasukawa1999),andthereis specific antiparasitic behaviors that may function
also a greater chance of predation later in the so as to prevent female cowbirds from approach-
season (Capper et al. 2012). ingandparasitizing redwingnests. Ourthreesets
Theredwings’primarycostofparasitismisthe of comparisons are described in the following
decrease in clutch size as the cowbird removes a paragraphs.
redwing egg and lays its own into the nest Comparison1.—Therehasbeenlittlepublished
(Hoover et al. 2006); in turn redwings suffer researchtodatereportingontherateandvariance
minimal additional costs to raise a cowbird chick of cowbird parasitism and the potential antipara-
inthenestalongsidetheirownyoung(Lorenzana siticresponsesofredwingsinthesespecies’north-
and Sealy 1999, Hauber 2003a). A hypothesized eastern range of sympatry, including New York
strategy to reduce the initial and non-recoverable State, USA (S Table 1). Previous research in
cost of cowbird parasitism is for the redwings to Ithaca, NY (Armstrong 2002) found cowbird
frontload their defenses (Welbergen and Davies parasitism of redwings to be ,1%, and during
2009, Kilner and Langmore 2011, Feeney et al. 5 years of research with 20+ redwing nests
2012) to prevent the cowbird from approaching monitored per year within 50 km of Ithaca, NY,
the nest, removing a redwing egg, and laying its Hauber (2003b) found no cowbird eggs in
own egg in the nest. For example, in a different redwing nests in either marsh or upland sites.
host-parasite system, Welbergen and Davies Examining the specificity of redwings’ nest
(2009) found that Eurasian Reed Warblers’ defenses against parasitic cowbirds in such
(Acrocephalus scirpaceus) aggression toward populations may be informative for future com-
parasitic Common Cuckoos (Cuculus canorus) parisons and meta-analyses as to the cognitive
reduced rates of parasitism. basis and ecological efficiency of antiparasitic
Observations at nests and experiments with strategies in local populations. The present study
mount-presentations (reviewed in Supplementary therefore explored the parasite/host relationship
Materials:STable1)haverevealedthatredwings between redwings and cowbirds in New York
vigorously defend their nests from nearby cow- State by examining nest defense behaviors in
birds, by alarm calling, mobbing, and physically response to presentations of taxidermic models
attacking the parasitic intruders, relative to accompanied by playbacks of the respective
490 THEWILSONJOURNALOFORNITHOLOGYNVol.126,No.3,September2014
vocalizations of female and male cowbirds, and from redwings defending nests to those without
female Northern Cardinals (Cardinalis cardina- known nests. Moreover, our study is one of the
lis), as a non-parasitic, non-predatory sympatric few to use principal components to describe
stimulus species. We hypothesized that the patterns of antiparasitic behaviors in an avian
redwings would engage in significantly more host species.
aggressive behaviors toward the female cowbird
mount than the other bird models, implying that METHODS
parasitically-laying female cowbirds are consid- Study Site and Species.—We located adult
eredauniqueclassofthreattothenestingattempt redwings at various parks in New York City and
of redwings. Ithaca, NY, USA, during the spring and summer
Comparison 2.—We compared the redwings’ of 2010. The search was focused on the typical
reactions to the mounts between New York City, redwing breeding habitats of fresh and saltwater
NY and Ithaca, NY. This was done initially in marshes, as well as upland meadows (Searcy and
order to increase the sample size of redwings for Yasukawa1995).Bothredwingsandcowbirdslay
our study but it also allowed us to compare the eggs from April to July (Lowther 1993, Searcy
consistency and variation of host behaviors and Yasukawa 1995). The female Northern
between these two relatively distant sites of Cardinal (hereafter: cardinal) was used as a
,270 km apart (Rand McNally 2004), with a control bird species because cardinals are sym-
different ecological context (urban vs. rural patric with redwings and they are approximately
setting),withinthegeneralregionofnortheastern the same size as cowbirds, but they are neither
North America. Our study is the first that we are parasitic nor do they prey upon redwing nests
aware of regarding the context of obligate brood (Halkin and Linville 1999).
parasitism of redwings within the urban draw of Redwings were not banded or otherwise
New York City (hereafter NYC). individually identifiable as part of this study. To
Comparison 3.—Finally, we aimed to deter- reducepseudoreplication,wewalkedatleast50m
mine whether it was possible to study redwings’ betweenpresumedredwingterritories(Searcyand
antiparasitic behavior from the behaviors of Yasukawa 1995).Inaddition, eachmount/control
individuals without having to locate active nests (see section below on Presentation of Mounts)
in the vicinity of the parasitic mount’s presenta- was presented only once at each territory with a
tion. Locating redwing, and other cowbird-host, known active nest and whereas only 1 of the 4
nests is potentially time consuming and can be mount/control treatments were presented at sites
logistically difficult or detrimental for the out- withnoknownnests.Wedefinedanestasactive
come of the nesting attempt, especially in the iftherewasatleastoneegginthenestandifwe
urban setting of highly controlled public parks of observed both a male and female redwing within
New York City with many potential nearby 5mofthenest.Atsiteswithnoknownnests,the
observers. If urban researchers were able to mounts were presented only when at least one
feasiblycombinethedatafromredwingterritories redwingmalewasobservedperchedinsubmerged
without known nests with data taken from vegetation at a likely nesting territory.
responsesatknownactivenests,itwouldproduce In and around Ithaca, NY, a total of three
a larger data set and a potentially more econom- nesting pairs of redwings and 16 redwing males
ical and/or robust account of local antiparasitic without known nests were exposed to the mount
behaviors. presentations. The presentations occurred in and
To complement the already large literature on around the Cornell University’s Experimental
using mount-presentations in the study of avian Aquatic Ponds Units 1 and 2, located in Lansing,
host-parasiteinteractions(reviewedinGrim2005) NY, USA (42u509330N, 76u469580W). The
and in cowbird-redwing interactions (as cited Cornell Pond Units are described in detail by
above, Strausberger and Horning 1998; S Table Westneat (1992) and Hauber (2000). All of the
1), our study was designed to provide novel sites in Ithaca were situated in rural areas, and
experimentaldatainthatwestudiedparasitismin sitesbehindfences,notopentothegeneralpublic.
urbanareaswithinNYC,weconductedplaybacks InNYC,weconductedexperimentsinQueens,
of sex- and species-specific vocalizations associ- NY,whereatotalofnineactivenestsofredwings
ated with each mount of parasite sex and control and 18 presumed territories without known nests
species, and we compared antiparasitic behaviors were exposed to the mount presentations. The
HengerandHauberNANTIPARASITICNESTDEFENSE 491
study localities in Queens included Flushing Pro 1.3 Sound Editing Software was used to
Meadows Corona Park (40.73uN, 73.84uW), generate 5-min sound files for each of the
Conselyea’s Pond at Brookfield Park (40.65uN, vocalizations.Eachsoundfileconsistedof10sec
73.74uW),BaisleyPond(40.67uN,73.78uW)and ofsoundfollowedby20secofsilence.The10sec
Kissena Park (40.74uN, 73.80uW). We also ofsoundwerecomprisedof 1secofvocalization
conducted experiments in the Bronx, NY, and followed by 2 sec of silence, repeated for 10 sec.
theexperiments includedthreeredwingnestsand To generate ‘‘silent’’ playbacks, the intensity of
nine territories without known nests. The sites in the sound was decreased to 1026 of its original
and around the Bronx, NY were Van Cortland intensity. Thus, the control included the tripod
Park (40.88uN, 73.89uW), Scout Park in Bronx- presentation, accompanied by 5 min and one of
ville, NY (40.93uN, 73.83uW), Soundview Park two exemplars of ‘‘silent’’ playback.
(40.76uN, 73.86uW) and Shore Road (40.86uN, The combined 10 sec of vocalization followed
73.80uW). Finally, we studied redwings in by 20 sec of silence was duplicated 10 times to
Brooklyn, NY, where the experiments included a generate a 5-min sound file. One min of silence
totalofeightterritorieswithoutknownnests.The (sound intensity 1025 of original sound) was
sites in Brooklyn, NY were Prospect Park Lake addedtothebeginningandendofeachsoundfile,
(40.65uN, 73.96uW) and Marine Park (40.60uN, allowingtimefortheexperimentertopositionthe
73.93uW). All of the presentations done in NYC mountatthenestatthebeginningofthetrialand
took place in public parks. to retrieve the mount after the trial was over
Presentation of Mounts.—We used taxidermic without exposing the birds to a different vocali-
avian models for the experiment (courtesy of B. zation. A final band pass filter between 500–
Strausberger Field Museum, Chicago, USA). 10,000 Hz was applied for all sound files. All
Logistical limitations meant that only one exem- playbacks were standardized to a uniform peak
plarmodelwasusedperstimulustype.Studiesthat amplitude and were played back at the maximal
have used replicate specimens for antiparasitic volume setting. Playbacks were played from an
mount presentation experiments have found no iPod Nano device connected to two portable
differencebetweenreplicatedummytypes(Davies RadioshackTM Model #40-1434 Portable Folding
andWelbergen2008,Grim2011).Aphotographic Speakers.
tripod served as a base upon which to mount and The study sites were chosen by looking for
presenteachmodel.Thetripod(height0.8m)was suitable redwing nesting microhabitats in parks
painted brown in order to provide some camou- with bodies of water surrounded by vegetation
flage.Thepanheadwasremoved fromthe top of (Searcy and Yasukawa 1995). Once a suitable
the tripod and in its place was attached a square locality was found, we searched for male
piece of Styrofoam, also painted brown. Each redwings since they are easier to locate than
modelbirdwasattachedtoametalrodthatwould females with their larger bodies and deep black
fitintotheStyrofoam,whichallowedthemodelsto coloration, and they find and defend territories
bepositionedupright.Theemptytripoditselfwas (Searcy and Yasukawa 1995). After male red-
alsousedasatreatmentcontroltotestwhetherthe wings were seen or heard, we approached them
redwings would show aggression to a foreign untilwewerenomorethan2mfromthebirds.At
objectinorneartheirnestandterritories. this distance, males defending nests will begin to
Many birds recognize and respond differently produce alarm vocalizations or hover towards
to heterospecific vocalizations (Hauber et al. humans who approach the nest (Searcy and
2002,ForsmanandMartin2009),andsospecies- Yasukawa 1995). If the male redwings directed
and sex-specific playbacks were used simulta- aggressivebehaviorstowardus,wewouldlookin
neously with the different mounts’ presentations the immediate area for nests. Redwing females
to elicit responses from the redwings. Two generallybuildtheirnests,1mfromtheground
exemplars of species- and sex-specific vocaliza- in marsh vegetation (Beletsky 1996), and male
tions were used each for the male cowbird, the redwings will also intensify their aggressive
femalecowbird,andthecardinalmount,toreduce behavior as humans approach closer to the nest
pseudoreplication concerns (Hauber et al. 2001). (Searcy and Yasukawa 1995), which sometimes
ThevocalizationswereobtainedfromTheDigital aids in locating the nest. If a nest was not found,
Resource Commons of Ohio University (drc. only one of the mounts or the tripod, chosen in a
ohiolink.edu) and www.findsounds.com. Raven random balanced manner, was presented to each
492 THEWILSONJOURNALOFORNITHOLOGYNVol.126,No.3,September2014
maleredwing,regardlessofwhetherthemalewas was determined in a random balanced manner.
solitary or there were females present. All 4 Most of the trials were videotaped using a
mounts/tripod were not presented to the males handheld video camera. Some trials were re-
without known nests, because it was not possible viewed during data analysis to determine which
toknowwhetherthesamebirdswouldbepresent sex of blackbird performed a certain behavior.
at the site during subsequent presentation. Data Collection and Analysis.—We used an
When a nest was found, the numbers of electronicbehavioraldataloggerforthisstudyby
redwing and cowbird eggs and/or nestlings were utilizing a recently developed iPhone/iPad touch
recorded for the nest. We would then retreat to a applicationnamed‘‘WhatISee.’’(T.Heuser2009;
distanceatleast10mawayfromthenestandwait hereafter: App or application; described in detail
fortheredwingstostopalarmcalling.Duringthat byHengeretal.2012).Wegeneratedanethogram
time, the mount apparatus would be set up in template for redwings by using the vocalizations
preparation to present the taxidermic models. and behaviors recorded in previous experiments
Once the birds were silent, one of the mounts or of redwing nest defense that were found in the
the tripod was presented to the birds. In order to literature (Supplementary Materials: S Table 2).
keepthebirdmodelhiddenuntilitwaspresented A Principal Components Analysis was used in
tothebirdsabrownclothwasdrapedontopofthe order to reduce the high number of possibly
model. correlated behavioral responses into a smaller
Themountapparatuswaspresentedascloseas numberofindependentvariables(Kachigan1991,
possibletothenest.Thedistancefromthemount Grim 2005, see also Honza et al.2006). The data
apparatus to the nest varied between 1–4 m, fromtheNestandNoKnownNestSiteswerefirst
because the nests were usually built in dense combined and then Principal Components (PC)
vegetation.Themountapparatuswasplacedatthe scoresweregeneratedseparatelyfrombehavioral,
nearestedgeofthevegetation.Itwaspresentedat vocal, and spatial distribution data types. The
the same height, 0.8 m from the ground, at each Principal Components were divided into these
presentation. The taxidermic bird mount was three groups in order to characterize information
alwaysdisplayedfacingthenest.Thebrowncloth about the physical and vocal reactions to the
covering the model was removed as soon as the modelsaswellastheproximityoftheredwingsto
apparatus was in position. The iPod and speakers the models and the number of redwings that
wereplacedinsidethebrownclothtocamouflage responded. Only Principal Components with Ei-
them. They were then positioned on the ground genvalues .1.0 were used in the statistical tests
just beside the mount apparatus. (Kachigan 1991). The Eigenvectors are shown in
At each presentation site, the apparatus was theSupplementaryMaterial(STables3–5).
positioned at the same location during all four All statistical tests were conducted in JMP 8.0
presentations. We started recording the behaviors Software. A backward elimination protocol was
foreachtrialassoonasthevocalizationsbeganto used in a General Linear Mixed Model (GLMM)
play.Sincetheplaybackofthesilenttrackdidnot (O’Hara 2009), with all of the Principal Compo-
have audible components with which we could nents as dependent variables and Nest ID as the
synchronize recording, the behaviors for each random effect. The Principal Components are
control trial were recorded as soon as we were AggressiveResponse,AggressiveMaleVocaliza-
positioned 10 m away from the model. Each tions, Aggressive Female Vocalizations, Female
recording was played at the maximum volume. Strikes, Intimidation, Close Multiple Males and
The birds’ behaviors were recorded for 5 min, Females, Close Male, Distant Female, Close
even if the birds flew away. One-zero sampling Male, Female Any Distance, and Other Birds.
scanswere used and the behaviors were recorded The experimentally controlled predictor variables
every15sec.Thistypeofsamplingwaschosenin were Location, Type of Mount, Mount Presenta-
ordertorecordrapidbehaviors,such asvocaliza- tion Order, and Nest Present or Absent; correla-
tions and strikes, which are difficult to capture tive predictors included Eggs or Nestlings, and
withothersamplingmethods(MartinandBateson Distance from Model.
2009). Mount presentations were performed All of the predictors were added to the model
between 30–45 min apart to allow the redwings and non-significant predictors were removed one
to return to their normal behavior (Honza et al. at a time. We stopped removing predictors when
2006). The order of presentation of the mounts all of the remainders in the model showed a
HengerandHauberNANTIPARASITICNESTDEFENSE 493
FIG.1. Differencesintheresponses(leastsquaremean6SE)ofRed-wingedBlackbirdsacrossalltreatments(PC
score:‘‘AggressiveMaleVocalizations’’)dependingonwhethertherewasaknownactivenestinthevicinityofthemount-
playbackpresentationexperiment.ThisdifferencewassignificantatP,0.05.
significant effect at a 5 0.05. The GLMM for the PC score ‘‘Close Male, Female Any
approach provides a numerical approximation of Distance’’(pooleddata:r250.76,F 53.22,
3,53.8
degrees of freedom through the Kenward-Roger P50.030,Fig. 2c).Responsestowardthefemale
method,sothereportedvaluesarenotnecessarily cowbird produced significantly lower values than
wholenumbers(KenwardandRoger1997).There responses toward the three other models for the
wereno aprioripredictions forinteractions.Two PCscore‘‘Intimidation’’(pooleddata:r250.83,
Principal Components (‘‘Close Multiple Males F 5 5.13, P 5 0.004, Fig. 2d).
3,49.22
andFemales’’and‘‘OtherBirdSpecies’’)didnot There was no significant difference between
meet the assumptions of parametric tests and so reactions to the stimulus presentations for the
they were log (x + 10) transformed prior to following PCs when all data were combined:
10
analyses. ‘‘FemaleStrike’’(r250.11,F 51.34,P5
3,91.74
0.27),‘‘AggressiveFemaleVocalizations,’’(r25
RESULTS 0.47, F 5 1.10, P 5 0.35) ‘‘Close Multiple
3,75.74
Comparison 1.—For four of the PC scores, Males & Females’’ (r2 5 0.43, F3,76.66 5 2.40,
there were significant differences between the P50.076)‘‘CloseMale,DistantFemales’’(r25
redwings’ reactions to the different types of 20.36, F3,92.61 5 2.17, P 5 0.097), and ‘‘Other
stimulus presentations. Considering the PC score Bird Species’’ (r2 5 0.14, F3,80.67 5 1.17, P 5
‘‘Aggressive Response,’’ redwings showed sig- 0.33).
nificantlyhigherPCscoresinreactiontoboththe Comparison2.—ThePCscores‘‘MaleAggres-
female cowbird and male cowbird than the siveVocalizations’’werehigherinIthacathanin
cardinal or the tripod (pooled data: r2 5 0.48, NYC(KnownNestdataonly:r250.70,F 5
1,9.56
F 55.98,P50.001,Fig. 2a).Significantly 17.243, P 5 0.002, Fig. 3a).
3,82.46
higher values of the PC score ‘‘Aggressive Male Comparison 3.—The PC score ‘‘Aggressive
Vocalizations’’ were recorded when redwings Male Vocalizations’’ was the only PCA metric
were reacting to the male and female cowbirds that was significantly different depending on
comparedtothecardinalorthetripod(NoKnown whetherredwingswerepresentedwiththemount-
Nest data only: r2 5 0.24, F 5 4.98, P 5 ed model(s) and playback(s) with or without a
3,15.39
0.004, Fig. 2b). Responses toward the female known active nest; the values were higher near
parasite produced significantly higher PC scores knownnests(r250.74,F 55.80,P50.02,
1,42.45
thanresponsestothemaleparasiteandthetripod Fig. 1); all other PC scores were statistically
494 THEWILSONJOURNALOFORNITHOLOGYNVol.126,No.3,September2014
FIG.2. a–d.Differencesintheresponses(leastsquaremean6SE)ofRed-wingedBlackbirds(PCscores:‘‘Aggressive
Response,’’ ‘‘Aggressive Male Vocalizations (sites with no known nests),’’ ‘‘Close Male, Female Any Distance,’’ and
‘‘Intimidation’’) to all treatments of the mount and playback presentations. Mounts with different letters are post hoc
significantlydifferentfromeachotherintheGLMManalysis.F:female,M:male.
similar between Nest and No Known Nest sites SingleMale,SingleFemaleAnyDistance’’(r25
(P . 0.05). 0.24, F 5 14.21, P 5 0.004).
1,9.39
CorrelativePredictors.—ThePCscores‘‘Male
Aggressive Vocalizations’’ were higher when DISCUSSION
redwings were defending known nests with
Comparison 1.—Redwings displayed consis-
nestlingsthanwitheggs(KnownNestDataOnly:
tently more aggression toward both the male and
r2 5 0.70, F1,9.41 5 10.035, P 5 0.011, Fig. 3b). female cowbird mounts than the female cardinal
The following PCs had significantly higher mount and the tripod. This suggests that the
scores at sites where the tripod with the model, redwings respond to both male and female
andtheplaybackspeakerwereplacedclosertothe cowbirds as a different class of threat, relative to
nest than sites with model presentation farther innocuous sympatric non-parasitic bird species
away from the known nest (data not shown): and inanimate objects. Several previous studies
‘‘Aggressive Response’’ (r2 5 0.24, F1,14.41 5 (Robertson and Norman 1976, Strausberger and
7.07,P 5 0.018), ‘‘Aggressive Female Vocaliza- Horning 1998, Gill et al. 2008; summarized in S
tions’’ (r2 5 0.38, F 5 4.79, P 5 0.047), Table 1) found that behaviors of cowbird hosts
1,13.38
‘‘CloseMultipleMalesandMultipleFemales’’(r2 towardmaleandfemaleparasiticmountswerenot
5 0.26, F 5 8.41, P 5 0.012), and ‘‘Close significantly different, implying that either the
1,14.08
HengerandHauberNANTIPARASITICNESTDEFENSE 495
FIG. 3. a–b. Differences in the responses (least square mean 6 SE) of Red-winged Blackbirdsto all treatments of
mount and playback presentations (PC scores: ‘‘Aggressive Male Vocalizations’’) between Ithaca and New York City
(NYC),NYandbetweenthosedefendingnestsandeggs.Onlydatafromknownnestsiteswasusedintheseanalysesand
plots.ThesedifferencesweresignificantatP,0.05.
redwings cannot distinguish the twosexes or that thoughcowbirdparasitismrateswereconsistently
thesehostsdidnotperceivethefemalecowbirdas .20% per annum in other sympatric hosts
a uniquethreat relative to males. Thisconclusion (Hauber 2003b). In NYC, the first redwing nest
is in line with the observation that even male inour12totalnestsamplewasparasitized(8.3%)
Brown-headed Cowbirds may participate in nest in2010,butallothernestswerenot.Althoughwe
searchingbehaviorsattributedtypicallytofemales found that redwings produced significantly more
of this obligate parasitic species (Strausberger aggressive vocalizations in Ithaca than in NYC
1998, Igl 2003). (Fig. 1)inresponsetocowbirdmounts,itremains
However, there also were several cowbird sex- unclear whether and how this may be related to
specific differences in our host behavioral re- variationintheriskofparasitismorthesuccessto
sponsedata,similartoasseeninadifferentsetof in fact reduce incidents of parasitism. Alterna-
previous studies (Folkers 1982), in that redwing tively, it is possible that the redwings with
males positioned themselves closer to the female territories inside public parks in NYC are
cowbird mount than toward the male cowbird habituated to a certain amount of disturbance
mount and the two controls. These consistent and therefore do not react to all types of intruder
mount-sex specific behavioral patterns imply that stimuli as strongly as do these same hosts with
hosts respond to the different sexes of the brood territories in and around the private and rural
parasite with different defensive strategies. Con- Cornell Pond Units. Future studies should exam-
versely, the PC ‘‘Intimidation’’ score was signif- ine habitat and site-specific variation with larger
icantly less in response to the female cowbird sample sizes of redwing nests and their cowbird
mount than to the other two mounts and the parasitismstatusacrossseveralurban-ruralgradi-
tripod. This PC is positively correlated with the ents (following Scales et al. 2011).
behaviorsDive,Hover,andNoReaction(STable Comparison 3.—The only PC that was signif-
3). A potential factor for why these behaviors icantly different depending on whether there was
were performed less often toward the female a known active nest in the vicinity of the mount
cowbird mount is perhaps that redwings were in presentation was the ‘‘Aggressive Male Vocali-
fact more likely to physically strike the female zation’’ PC. This PC was comprised of positive
cowbird mount than to dive and hover near it contributions from the vocalizations ‘‘check,’’
(Fig. 2d). ‘‘whistle,’’and‘‘ring’’(STable5);andredwings
Comparison 2.—In Ithaca, throughout 5 years performed these behaviors significantly more
of research between 1997–2010, no parasitized often when defending known nests. Researchers
redwing nests were found (n . 100 nests), even looking to combine behavioral nest defense data
496 THEWILSONJOURNALOFORNITHOLOGYNVol.126,No.3,September2014
fromredwingswithorwithout knownactivenest tween two sites in New York State, and found
sites would need to be aware of and account for more aggression in redwings in rural Ithaca than
this variability. Overall, however, we found that in urban New York City. Finally, we compared
the behaviors from males without known nests thebehaviorsoftheredwingswithactiveneststo
were statistically similar to behaviors from redwings with no known nests and documented
redwingswithknownactivenests.Ourconclusion that only one Principal Component, out of nine
isthatitispossibletocombinedatafromnesting total PCs, differed significantly depending on
redwing pairs and redwing males without known whether there was a known active nest in the
nests to study these hosts’ responses in cowbird vicinity.Thissuggeststhatinaninitialsurvey,the
mount and sound presentation experiments. presence of an active host nest is not necessarily
CorrelativePredictorsofRedwingResponses.— required to provide a suitable experimental
Although not as the outcome of an experimental analysisofredwings’responsestowardtaxidermic
design, we found statistical patterns indicating mounts and vocalization playbacks of brood
that redwings displayed more aggression toward parasitic cowbirds.
the mounts when defending nestlings than eggs.
ACKNOWLEDGMENTS
Both female and male cowbirds are known to
destroy host eggs and nestlings (Granfors et al. For funding, we are grateful to the Provost’s Office of
2001, Igl 2003), perhaps as part of a farming or HunterCollegeandtheViceChancellor’sOfficeoftheCity
mafia strategy (Arcese et al. 1996, Hoover and University of New York, the PSC CUNY faculty grant
Robinson2007,Hauber2009)andso,toexamine program,theNationalGeographicSociety,andtheHuman
FrontierScienceProgram.WethankB.M.Strausbergerfor
whether redwings perceive male cowbirds as
loaning the taxidermic mounts used in this study, and B.
specific threats to nesting success, future mount
Croston,Z.Aidala,D.Lahti,S.Chase,T.Grim,C.Moskat,
presentation experiments should include larger andJ.Hyland-Brunoforhelpfulassistance,comments,and
sample sizes of redwing nests occupied by discussions.
nestlings and compare reactions toward both
sexes of cowbirds across nesting stages. LITERATURE CITED
Inaddition,wefoundthataggressiveresponses ARCESE,P.,J.N.M.SMITH,ANDM.I.HATCH.1996.Nest
were greater as the mounts and playbacks were predation by cowbirds, and its consequences for
displayedclosertothenests.Clotfelter(1998)also passerine demography. Proceedings of the National
foundthatredwingsweremoreaggressivetowards AcademyofSciencesoftheUnitedStatesofAmerica.
93:4608–4611.
a female cowbird mount as the distance from the
ARMSTRONG, M. A. 2002. Defenses of Red-winged
mounttothenestdecreased.Femalecowbirdsmust
Blackbirds against brood parasites and predators: the
be physically at the nest in order to remove host
acquisitionofthreatrecognitionandthedynamicsof
eggsandlayparasiticeggs,andsoredwingswould group defense. Dissertation. State University of New
benefitfromincreasingtheiraggressivebehaviors York,Binghamton,USA.
as female cowbirds approach so that the brood BAZIN,R.C.ANDS.G.SEALY.1993.Experimentsonthe
parasitesdonotland,accessthenest,removehost responsesofarejectorspeciestothreatsofpredation
andcowbirdparasitism.Ethology94:326–338.
eggs,andlayeggsoftheirowninthenest.
BELETSKY, L. D. 1996. The Red-winged Blackbird: the
Conclusions.—Red-winged Blackbirds re-
biologyofastronglypolygynoussongbird.Academic
sponded more aggressively toward mount and Press,SanDiego,California,USA.
playback presentations of both sexes of the BRISKIE, J. V. AND S. G. SEALY. 1989. Changes in nest
obligate brood parasite, the Brown-headed Cow- defense against a brood parasite over the breeding
bird,thantothefemaleNorthernCardinal,andan cycle.Ethology82:61–67.
empty tripod (as controls). In addition, male BRISKIE, J. V., S. G. SEALY, AND K. A. HOBSON. 1992.
Behavioraldefensesagainstavianbroodparasitismin
redwings approached the female cowbird presen-
sympatric and allopatric host populations. Evolution
tation more closely relative to the male cowbird. 46:334–340.
These findings are in line with the theory that BURHANS, D. E. 2001. Enemy recognition by Field
redwings mount nest defenses to reduce the cost Sparrows.WilsonBulletin113:189–193.
of parasitism; indeed, Strausberger (2001) found BURHANS,D.E.,B.M.STRAUSBERGER,ANDM.D.CAREY.
2001.RegionalvariationinresponseofFieldSparrows
an inverse association between host aggression
to the threat of Brown-headed Cowbird parasitism.
toward a female cowbird mount and cowbird
Auk118:776–780.
parasitism rates. Our study also compared anti- CAPPER,C.L.,M.F.GUIGUENO,ANDS.G.SEALY.2012.
parasitic aggressive behaviors of redwings be- Acceptance of simulated cowbird parasitism in a
HengerandHauberNANTIPARASITICNESTDEFENSE 497
northernpopulationofRed-wingedBlackbirds.Amer- GRIM, T. 2005. Host recognition of brood parasites:
icanMidlandNaturalist167:127–135. implications for methodology in studies of enemy
CLOTFELTER, E. D. 1998. What cues do Brown-headed recognition.Auk122:530–543.
Cowbirds use to locate Red-winged Blackbird hosts GRIM, T. 2007. Experimental evidence for chick discrim-
nests?AnimalBehaviour55:1181–1189. ination without recognition in a brood parasite host.
CLOTFELTER, E. D. AND K. YASUKAWA. 1999. Impact of ProceedingsoftheRoyalSocietyB274:373–381.
broodparasitismbyBrown-headedCowbirdsonRed- GRIM, T. 2011. Ejecting chick cheats: a changing
winged Blackbird reproductive success. Condor paradigm?FrontiersinZoology8:14.
101:105–114. HALKIN,S.L.ANDS.U.LINVILLE.1999.NorthernCardinal
CRUZ, A. AND L. COOPER. 2001. Behavioral defenses (Cardinalis cardinalis). The birds of North America.
against brood parasitism in the American Robin Number440.
(Turdus migratorius). Final Report submitted to the HAUBER, M. E. 2000. Nest predation and cowbird
City of Boulder Open Space Department, Colorado, parasitism in Song Sparrows. Journal of Field
USA, www-static.bouldercolorado.gov/docs/4177_ Ornithology71:389–398.
Cruz_Alexander_Behavioral-1-201307091311.pdf HAUBER, M. E. 2001. Site selection and repeatability in
(accessed4May2010). Brown-headedCowbird(Molothrusater)parasitismof
DAVIES,N.B.2000.Cuckoos,cowbirds,andothercheats. Eastern Phoebe (Sayornis phoebe) nests. Canadian
London:T&ADPoyser,London,UK. JournalofZoology79:1518–1523.
DAVIES,N.B.ANDJ.A.WELBERGEN.2008.Cuckoo-hawk HAUBER, M. E. 2003a. Hatching asynchrony, nestling
mimicry? Proceedings of the Royal Society B competition, and the cost of interspecific brood
275:1817–1822. parasitism.BehavioralEcology14:224–235.
D’ORAZIO,K.A.ANDD.L.NEUDORF.2008.Nestdefense HAUBER, M. E. 2003b. Lower begging responsiveness of
by Carolina Wrens. Wilson Journal of Ornithology host versus parasitic Brown-headed Cowbird (Molo-
thrusater) nestlingsis related to species identitybut
120:467–472.
nottoearlysocialexperience.JournalofComparative
FEENEY,W.E.,J.A.WELBERGEN,ANDN.E.LANGMORE.
Psychology117:24–30.
2012. The frontline of avian brood parasite-host
coevolution.AnimalBehaviour84:3–12. HAUBER, M. E. 2009. Does the removal of avian brood
parasiteeggsincreasehostproductivity?Acasestudy
FOLKERS,K.L.1982.Hostbehavioraldefensestocowbird
with Brown-headed Cowbirds Molothrus ater and
parasitism. Kansas Ornithological Society Bulletin
Song Sparrows Melospiza melodia near Ithaca, New
33:32–34.
York,USA.ConservationEvidence6:83–88.
FOLKERS, K. L. AND P. E. LOWTHER. 1985. Responses of
HAUBER,M.E.,H.E.PEARSON,A.REH,ANDA.MERGES.
nestingRed-wingedBlackbirdsandYellowWarblers
2002. Discrimination between host songs by brood
toBrown-headedCowbirds.JournalofFieldOrnithol-
parasitic Brown-headed Cowbirds (Molothrus ater).
ogy56:175–177.
AnimalCognition5:129–137.
FORSMAN,J.T.ANDT.E.MARTIN.2009.Habitatselection
HAUBER, M. E. AND S. A. RUSSO. 2000. Perch proximity
forparasite-freespacebyhostsofparasiticcowbirds.
correlates with higher rates of cowbird parasitism of
Oikos118:464–470.
ground nesting Song Sparrows. Wilson Bulletin
FREEMAN, S., D. F. GORI, AND S. ROWHER. 1990. Red- 112:150–153.
winged Blackbirds and Brown-headed Cowbirds:
HAUBER,M.E.,S.A.RUSSO,ANDP.W.SHERMAN.2001.A
some aspects of a host-parasite relationship. Condor
password for species recognition in a brood parasite
92:336–340.
bird. Proceedings of the Royal Society of London B
GILL,S.A.,D.L.NEUDORF,ANDS.G.SEALY.1997.Host 268:1041–1048.
responses to cowbirds near the nest: cues for
HAUBER,M.E.,P.SAMAS,M.G.ANDERSON,J.RUTILA,J.
recognition.AnimalBehaviour53:1287–1293.
LOW, P. CASSEY, AND T. GRIM. 2013. Life-history
GILL, S. A.,D.L. NEUDORF, ANDS. G.SEALY.2008.Do theory predicts host behavioural responses to experi-
hostsdiscriminatebetweensexuallydichromaticmale mental brood parasitism. Ethology Ecology and
and female Brown-headed Cowbirds? Ethology Evolution26:(online).
114:548–556. HENGER,C.S.,S.B.WALLACE,ANDM.E.HAUBER.2012.
GILL,S.A.ANDS.G.SEALY.1996.NestdefensebyYellow Comparison of two field data collection methods in
Warblers:recognitionofabroodparasiteandanavian recordingavianbehavior.Kingbird62:198211.
nestpredator.Behaviour133:263–282. HOBSON, K. A. AND S. G. SEALY. 1989. Responses of
GILL,S.A.ANDS.G.SEALY.2004.Functionalreferencein YellowWarblerstothethreat ofcowbirdparasitism.
analarmsignalgivenduringnestdefense:seetcallsof AnimalBehaviour38:510–519.
YellowWarblersdenotebrood-parasiticBrown-head- HONZA,M.,V.SˇICHA,P.PROCHA´ZKA,ANDR.LEZˇALOVA´.
ed Cowbirds. Behavioral Ecology and Sociobiology 2006. Host nest defense against a color-dimorphic
56:71–80. brood parasite: Great Reed Warblers (Acrocephalus
GRANFORS, D. A., P. J. PIETZ, AND L. A. JOYAL. 2001. arundinaceus) versus Common Cuckoos (Cuculus
Frequencyof egg and nestling destruction by female canorus).JournalofOrnithology147:629–637.
Brown-headed Cowbirds at grassland nests. Auk HOOVER,J.P.,K.YASUKAWA,ANDM.E.HAUBER.2006.
118:765–769. Spatially and temporally structured avian brood