ORGANIZATION OF SINGING IN HOUSE WRENS

advertisement
J. Field Ornithol., 58(2):190-197
ORGANIZATION
MEREDITH
OF
SINGING
IN
HOUSE
E. PLATT • AND MILLICENT
WRENS
S. FICKEN
Departmentof BiologicalSciences
Universityof Wisconsin-Milwaukee
Milwaukee, Wisconsin 53207 USA
Abstract.--House Wren songsare complexand the repertoiresof individualsare large.
Most of the songsin this study were composed
of nine syllablesand they were most
frequentlymade up of only two different syllabletypes.The numberof repetitionsof a
particularsyllabletype varied.Syllablearrangementwas alsovariable;particularsyllables
usuallywere not foundin only one specificpart of the song.There was frequentsyllable
sharingamongbirdsof this population,but little songtype sharingoccurred.Often, part
of the songwas sharedbut the presence
of uniquesyllablesproducedmany differentsong
types and, consequently,distinctsongrepertoiresamong birds.
A numberof theorieshavebeenproposed
to explainthe evolutionof complexsongsand
large repertoiresin birds.In HouseWrens large repertoiresmay be involvedin enhancing
territorial defensealthoughone hypothesisrelated to territorial defense,the Beau Geste
hypothesis,was rejected.Also, large repertoiresmay have evolvedthroughfemale choice.
ORGANIZACI(•NDEL CANTO DEL REYEZUELO(TROGLODYTES
AEDON)
Sinopsis.--E1cantodc Troglodytes
aedoncscomplcjoy conun ampliorcpcrtorio.La mayorcadc las cancioncs
tichen9 silabas,aunqucs61odosdc cstassondifcrcntcs.La rcpctici6n
dc una silabaparticularvaria entreindividuos.E1 arrcglodc las silabastambigncsvariable.
No sc cncontr6silabasparticulatesquc cstuvicrancolocadas
cn partescspcc[ficas
dc la
canci6n.Aunqucpartcdc la canci6ncscompartida,
la prcscncia
dc s•labasparticulates
da
origen a muchascancioncsdifcrcntcsy pot cndc a un gran rcpcrtorio.Sc hah propucsto
variastcoriaspara cxplicarla cvoluciondc complcjos
dc canclones
y rcpcrtorioscn aves.
En cl ave cstudiadacs posiblcquc cl gran rcpcrtoriopucda scr dc utilidad cn la dcfcnza
dcl tcrritorio, aunqucla hipotcsisBeau Gcstccsrcchazada.Un gran rcpcrtoriopucdchabcr
cvolucionado
comorcspucstaa la sclccci6ndc parcja pot partc dc hcmbras.
Compared to most other North American wren species,the House
Wren (Troglodytes
aedon)has fairly complexsongsand individualshave
large songrepertoires(Kroodsma1977). The objectives
of this studyare
to (1) determinesongorganizationand repertoire size of individualsin
a local population,(2) determinethe degreeof syllableand songtype
sharingin a population,and (3) test the predictionof the Beau Geste
hypothesis
(Krebs1977) that songtypechanges
occurconcomitantly
with
perch switches.
METHODS
Songsof four male HouseWrens were recordedat the Universityof
Wisconsin-MilwaukeeField Station,Saukville(OzaukeeCounty) Wisconsinbetween7 June and 23 July, 1982. Songswere recordedat various
timesof the day (0530-1600, C.D.T.) for any particular individual. Two
of the birds had territoriesin an upland apple orchard,one was located
• Presentaddress.'
Invertebrate
ZoologySection,MilwaukeePublicMuseum,800 W. WellsSt.,
Milwaukee, Wisconsin53233 USA.
190
Vol.58,No.2
Singing
inHouse
Wrens
[ 19 1
in a low brushy area adjacentto an open field, and one occupiedan
upland area of primarily aspen (Populustrernuloides).
The territoriesof
Birds I, III, and IV were contiguous.Bird II was locatedapproximately
100 m from the other three. All were probablywithin hearingrange of
one another and these were the only House Wrens in the immediate
area. Althoughthe birds were not banded,individualswere easilyidentified by territorial behavior;they tended to sing at or near their own
nestsitesand only rarely enteredanotherbird's territory.
A Uher 4200 Reporttape recorderand an Electrovoice644 directional
microphonewere usedto recordthe songs.Informationdescribingsinging
locationswas recordedand distancesbetweensongpercheswere measured, enabling comparisonsto be made betweensongtypes given and
distancesmovedbetweensongs.Intervals betweensuccessive
songstarts
were also recorded.Sonagramswere made with a Kay 6061 B SonaGraph at an intermediateband width setting(150 Hz).
Identificationof syllableswasprimarily by shape,usingfrequencyand
temporal characteristics.Letter symbolswere assignedto syllablesto
facilitateanalysis.Sequences
of identicalsyllabletypes,regardlessof the
numberof repetitionsof eachtype, characterizedunique songtypes.For
example, AAAAABBBBBKKKK and AAAAABBBKKKKKKK were
consideredto be the same song type but they were not the same as
AAAAAJJJJKKKKKKK.
RESULTS
Songdescription
and organization.--HouseWren songconsistsof two
parts.The introductorypart is madeup of very low amplitude"chattery"
notes,but the terminal sectioncontainsclearer,louder, "bubbly" notes.
As the introductorypart of the songin our recordingswas too faint to
be analyzedaccurately,this report deals only with the louder, terminal
portion.Severalsongcharacteristics
(e.g.,total syllablenumberper song,
numberof differentsyllablesper song,syllableorder)variedconsiderably
amongbirds. Figure 1 illustratesthe syllablesin the study population
and Figure 2 showsseveralrepresentativesongsof four different individuals,demonstratingthe ways in which syllablesare combined.
Of the 1993 songsanalyzed,the total number of syllablesper song
ranged from 3 to 22, with a mean of 11.49 (SD = 12.06). The mode
was 9. The number of different syllabletypesper songrangedfrom 1 to
6, with a mean of 2.92 (SD = 3.10) and a mode of 2.
Order and numberof repetitionsof eachsyllabletype oftenvaried in
different songs,even in the same individual. Most syllableswere not
specificto a particularpart of the song.For example,if a certainsyllable
occurredin the beginningof a particular songand was also presentin
anothersong,it did not necessarilyoccurin the beginningof that other
song.
Syllablesranged in frequencyfrom 1.5 to 8.0 KHz. Song duration
rangedfrom 0.33 to 2.48 s (N = 220) with a mean of 1.25 (SD = 0.35).
192]
A
B
D
M. E. PlattandM. S.Ficken
E
F
G
J
K
O
P
T
j. Field
Ornithol.
Spring 1987
U
AB CDE F H I J KL i T U W•Z
A
B
C
E
H
I
J
K
O
Q
X
Y
FIGURE
1.
Syllablesin the songs(terminal, "bubbly" sectionsonly) of House Wrens at
the Universityof Wisconsin-MilwaukeeField Station.(a) Bird I (b) Bird II (c) Bird
III (d) Bird IV.
Generally,malessangmostrapidly when a femalewas closeto or in the
nest.
Amountof syllablesharing.--Examination of sonagramsrevealedan
extensiveamount of syllablesharing.Sevenof the 26 syllablesusedby
the four birds were commonto all. Six other syllableswere sharedby
three out of the four birds. Twelve syllablesmay have been unique to
individuals.Bird I shared l0 of its 12 syllableswith at least one other
individual, Bird II shared13 of its 16 syllables,Bird III shared14 of its
19 syllables,and Bird IV shared 11 of its 13 syllables.Coincidence
indices(Dice 1945) were calculatedto show degreeof similarity of syllable typesbetweenpairs of birds. In the equation
Coincidenceindex = 2h(a + b)
h = number of sharedsyllables,a --- number of syllablesof Bird A, and
b -- numberof syllablesof Bird B. Values of the coincidence
index range
from 0, indicatingcompletelack of sharingof songsyllablesof the two
individuals,to 1.0, indicating completesharing. In this study, values
FIGURE2. Examplesof House Wren songsshowingthe songtype variability that may
occurboth amongand within individualbirds. (a) Bird I (b) Bird II (c) Bird III (d)
Bird
IV.
Vol.58,No.2
Singing
inHouse
Wrens
[ 193
8
'
0
,,"
DE
D
T
U
-
J
K
;
Time
Time
(sec)
(sec)
8
,
•.
T
U
L
B
1
Time
T•me
(sec)
(sec)
8
.•.
•
K
'
• m
I I
•
I
I
B
.
Time
(sec)
B
Time
(8ec)
s
T
K
ø•'
1½I'i,!.•,
•2 I
K
B
J
T•me
K
(sec)
Time
0
Time (sec)
B
DE
Time
(sec)
194]
m. E. PlattandM. S.Ficken
T^•Lw
Bird
I (12) a
IX (16)
III (19)
IV (13)
1.
J.Field
Ornithol.
Spring 1987
Coincidenceindicesfor syllablesharing.
II
0.71
III
0.65
0.74
IV
0.56
0.69
0.69
Total number of different syllables.
rangedfrom 0.56 to 0.74 (Table 1), with a mean value of 0.67 (SD =
0.063), indicatinga relativelyhigh degreeof similarity betweenindividuals with regard to syllabletypes.
Unique syllableswere generallylocatedat either of the two endsof
the songs.Both of Bird I's two unique syllableswere locatedat one end
or the other of the 15 songsin which they occurred.Bird II had three
unique syllables.Of the 71 songsin which they occurred,they were at
the beginningor end in 47 of them. In the other 24 songs,they were
near the end and each time they were followedby a syllabletype (B)
which terminatedthe song.Bird III had five unique syllables,one of
which initiated the songin 39 of the 62 songsin which they occurred.
In the other 23, they were near the beginningof the song,beingpreceded
by only one syllable (Q). Bird IV had two unique syllables.Of the 10
songsin which they occurred,they were within the bodyof the songonly
twice, and in both casesa unique syllablealsobeganthe song.
Amountof songtype sharing.--Sharingof songtypes (unique syllable
combinations)occurredinfrequently. Of the 130 songtypes that were
usedby the four birds, 120 were unique. Only one songtype was used
by all of them and only two were sharedby three out of the four individuals.More specifically,Bird I shared7 of its 33 songtypes,Bird II
shared9 of its 37 songtypes,Bird III shared5 of its 52 songtypes,and
Bird IV shared3 of its 22 songtypes.Coincidenceindicesof the degree
of sharingof songtypesbetweenpairs of birdsrangedfrom 0.03 to 0.17
(Table 2), with a mean value of 0.09 (SD = 0.047). These indices,
however,do not take into accountsongsthat were similar but not identical. Many songsdiffered by the interjectionor deletionof only one
syllable-typebut were still classifiedas different songtypes.
Specificrepertoirecomparisons.--Repertoires
of four birds were compared. Sevenof the twelve more frequentlyusedsongtypeswere specific
to particular individuals. For example, AJKB was given 91 times, but
only by Bird III, and CODE was given 105 times, 103 times by Bird
III. The most commonsong,AB, was given by all of the birds but the
percent of the entire repertoire devotedto this song varied from one
individual'with 6% to others with 12, 24, and 36%.
Possible
influences
on songtype.--Changein perch site,changein song
type, and interval betweensuccessive
songstartswere testedfor interactionusinga log-likelihoodratio test(Sokaland Rohlf 1981) on a three-
TABLE
Bird
I
II
III
IV
[195
Singingin HouseWrens
Vol. 58, No. 2
(33) a
(37)
(52)
(22)
2.
Coincidenceindicesfor songtype sharing.
II
III
IV
0.17
0.07
0.09
0.07
0.10
0.03
aTotal number of different songtypes.
way contingency
table (Table 3). No significantthree-way interaction
occurred(G[PST] = 0.056, P > 0.50).
All combinations
of two of the factors were tested for interaction
with
the effectsof the third factor removed.There was no significantrelationship between whether perch site changedand whether song type
changedbetweentwo successive
songswhen the effect of time between
songswas removed (G[PS(T}] = 0.220, P > 0.50), where T = time
intervalbetweensongs,P -- change/nochangein perchsite,$ = change/
no changein songtype. Also, neither changesin perch sitesnor song
type changesshoweda significantdependence
on lengthof time between
songswhen the effectof the otherfactorwasremoved(G[PT(S}] = 0.570,
P > 0.50 and G[ST(P•] = 0.056, P > 0.50 respectively).
DISCUSSION
Most North American wren speciesorganize their songboutsinto a
pattern that Hartshorne (1973) termed "eventualvariety" (AAA ...
BBB... where A and B are differentsongtypes),singingthe samesong
type severaltimesbeforeswitching.A few species,
includingthe Marsh
Wren (Cistothoruspalustris) and Sedge Wren (Cistothorusplatensis)
(Kroodsma1977), haveboutpatternsthat Hartshornecalled"immediate
variety"(ABCD ...), in whichthereis little or no immediatesongtype
repetition.Kroodsmaclassifiesthe House Wren as an AAA... BBB...
songster.Sinceeachunique arrangementwas considered
to be a distinct
songtype,theseHouseWrens had severaldissimilarsongtypesin a row,
resultingin songbouts that were made up of sequences
of immediate
variety interspersedwithin the eventualvariety pattern.
House Wrens in this study sharedmany syllableswith at least one
otherbird, but eachof the birds alsohad severalunique syllables.Song
typeswere sharedinfrequently.In many avianspecies,suchasthe Tufted
Titmouse(Parusbicolor),individualssharemostor all of their songtypes
with their neighbors(Schroederand Wiley 1983). Verner (1975) found
that male Marsh Wrens in easternWashingtonsharedmostof their song
typeswith eachother.
Many hypotheses
havebeensuggested
for large repertoires(reviewed
by Krebs and Kroodsma1980). In additionto the possibilityof song
matchingwith territorial neighbors,large songrepertoiresmay enhance
territorial defensethroughthe proposedBeau Gestephenomenon(Krebs
196]
m. E. PlattandM. S.Fieken
J.Field
Ornithol.
Spring1987
TABLE3. Three-way contingencytable of the relationshipsamonginterval, changeof
perch,and changeof songtype.
Time interval
Changeof
perch
Change in songtype
yes
Short
(0.0-6.0 s)
Long
(•6.0 s)
Totals
no
yes
17
12
no
43
37
29
80
60
49
109
yes
18
14
32
no
39
32
71
57
46
103
117
95
212
Totals
1977) involving deceit of incoming males by the diversity of songsby
territory holders.A predictionof this hypothesisis that a bird will switch
songtypesas it movesfrom one songperchto another.A studyof Chaffinches(Fringillacoelebs)
by Dawsonand Jenkins(1983) did not support
the Beau Geste hypothesis.Yasukawa (1981) demonstratedthat RedwingedBlackbirds(Agelaius
phoeniceus)
do switchsongsat differentperch
sitesbut he and Searcy(1985) showedthat the number of songtypesis
not usedto assess
male densities.Our studyof wrens fails to supportthe
prediction.Krebs (1977) developedthe Beau Gestehypothesisto account
for repertoire evolution in specieswith repertoire sizes of 2-20 song
types,so it was not surprisingthat the songswitchingpredictiondid not
hold true for House Wrens which have very large repertoiresizes.
Thus, althoughthis studydoesnot supportone of the key predictions
of the Beau Gestehypothesis,that songtypeschangewith perchswitches,
large songrepertoiresmay still be involvedin territorial defensethrough
advantagesof songmatchingby territorial neighborsor through reduction of habituation. Large repertoiresin House Wrens may also have
evolvedthroughtheir effecton females.For example,the mostreasonable
explanationof the very low amplitudesyllablesat the beginningof songs
is that theseare directedat the mate who is probablyin closeproximity,
rather than at a territorial neighborthat is probablymuch more distant.
In addition,there is the possibilitythat large repertoiresare attractiveto
femalesand stimulatethem more physiologicallythan would repetitions
of a singlesongtype (e.g., Kroodsma1976).
ACKNOWLEDGMENTS
We thank J. D. Buntin, L. S. Johnson,J. A. Reinartz, C. V. Sommer,and C. M. Weise
for helpful commentson the manuscript.This is publicationnumber80 of the University
of Wisconsin-Milwaukee
Field Station.
LITERATURE
CITED
DAWSON,
S. M., ANDP. F. JENKINS. 1983. Chaffinchsongrepertoiresand the Beau Geste
hypothesis.Behaviour87:256-269.
Vol.58,No.2
Singing
inHouse
Wrens
[ 197
DICE, L. R. 1945. Measures of the amount of ecologicalassociationbetween species.
Ecology 26:297-302.
HARTSHORNE,
C. 1973. Born to sing.Indiana University Press,Bloomington,Indiana.
KREBS,
J.R. 1977. The significance
of songrepertoires:the Beau Gestehypothesis.Anim.
Behav. 25:475-478.
KREBS,J. R., AND D. E. KROODSMA.1980. Repertoiresand geographicalvariation in
bird song.Pp. 143-177, in J. Rosenblattet al., eds.Advancesin the studyof behavior.
Academic Press, New York.
KROODSMA,
D. E. 1976. Reproductivedevelopmentin a female songbird:differential
stimulationby quality of male song.Science192:574-575.
1977. Correlatesof songorganizationamong North American wrens. Am. Nat.
11• :995-1008.
SCHROEDER,
D. J., AND R. H. WILEY. 1983. Communicationwith sharedsongthemes
in Tufted
Titmice.
Auk
100:414-424.
SOKAL,R. R., ANDF. J. ROHLF. 1981. Biometry.W. H. Freemanand Company,San
Francisco, California.
VERNER,J. 1975. Complex songrepertoireof male Long-billed Marsh Wrens in eastern
Washington. Living Bird 14:263-300.
Y^SUK^W^,
K. 1981. Songrepertoiresin the Red-wingedBlackbird(Agelaius
phoeniceus):
a testof the Beau Gestehypothesis.
Anim. Behav.29:114-125.
--,
^NDW. A. SE^RC¾.1985. Songrepertoiresanddensityassessment
in Red-winged
Blackbirds:further testsof the Beau Gestehypothesis.Behav. Ecol. Sociobiol.16:171175.
Received23 June 1986; accepted2 Dec. 1986.
Download