ASSTR•CT.--Two song dialects are described for montane White-crowned
Sparrows (Z0~ notrichia leucophrys oriantha) in the Sierra Nevada of California,
one ...
SONG
DIALECTS
MONTANE
AND
MATE
WHITE-CROWNED
SELECTION
IN
SPARROWS
Luis F. BAPTISTA 1 AND MARTIN L. MORTON Departmentof Biology,OccidentalCollege,LosAngeles,California90041 USA
ASSTR•CT.--Twosongdialectsare describedfor montaneWhite-crownedSparrows(Z0~ notrichialeucophrys oriantha)in the SierraNevada of California,one at Tioga Passand one at Gardisky Lake.3 km to the north. The dialectsdiffer from each other in syllabicmorphologyand in rhythmic(temporal)and pitch characteristics. Two malessangforeigndialects, one typical of the Canadian Rockiesand the other typical of Z. 1. gambeliiof Alaska and Canada.Both"songmisimprinted"birdswerepairedandat leastonebredsuccessfully. A third male,singingsyllablesfrom Mount Lassen,was alsopaired. Most femalesinduced to sing with testosteroneinjectionsdid not matchthe dialectsof their consorts,suggesting that song dialectsdid not influencemate choice.One female sanga song similar to those from Mount Lassento the north but was paired to a male singing a Gardisky song. Fourpossibleexplanations for mis-matchingof dialectsare discussed. The mostprobable explanationat this time is that mating is randomaccordingto dialectand that differential dispersaldistancesbetween the sexesaccountfor sex differencesin song types. Received 18 September 1981,accepted 25 January1982. Song
dialects
are well documented
in avian
In this study we test the assortativemating theory of song-dialectfunction by recording and comparingsongsof mated pairs of Whitecrowned Sparrows(Z. I. oriantha).
species (Krebs and Kroodsma 1980), but the
selective advantage of learning a dialect is poorly understood. Song dialects in North American White-crowned Sparrows (Zonotrichialeucophrys)were discussedin the early
MATERIALS
literature (Dawson 1923, Blanchard 1941, Pe-
AND METHODS
terson1941), but it was not until the study of
Montane White-crowned Sparrows are migratory and typically breed in subalpinemeadowsor ripar-
Marler and Tamura (1962) that these dialects
ian woodland at altitudes of about 2,000-3,000 m
were analyzed spectrographically and subject- (DeWolfe and DeWolfe 1962, Morton 1976). The ed to statistical analysis. Konishi (1965) first breeding biology (review in Morton 1978)and song demonstrated
dialectsof the westernpopulationshave been studied in great detail (Orejuela and Morton 1975, Baptista and King 1980). Our study area is near Tioga Pass,Mono County,
that some female White-crowned
Sparrowstreated with testosteronesang their native dialects, although females are seldom observedsinging in the wild (Blanchard1941, Kern and King 1972, Baptista 1975). Marler and Tamura (1962), Marler (1970), and Konishi
California
at latitude
37ø55 ' and an elevation
of about
3,000 m. In 1978we discovereda songdialectat Gar-
to: (1) induce birds to breed near their natal
diskyLake(Fig. 1) at the northernedgeof our study areathat was markedlydistinctfrom the TiogaPass dialect described earlier by Orejuela and Morton
areas, becausethey are attracted to their fa-
(1975).
(1965) suggestedthat dialectsmight function
Birds at all study localitiesbred in willows and
miliar natal dialect; and (2) influence females
to selectas matesmalesthat sing dialectsof the small pines. GardiskyLake (3,196 m) is separated female'snatal area, thus promoting assortative from Lee Vining Creekby a slopewith habitat unmating. Nottebohm (1969,1972, 1975)expand- suitable to breeding White-crowned Sparrows. The ed on the above ideas with data based on stud-
ies of the congenericZ. capensis.
slope is rocky and barren for the most part, with
scatteredpine trees. Birds at Lee Vining Creek oc-
cupyriparian habitat.The populationat Tioga Pass inhabits a meadow and surrounding mountain slopes.Most of our banding activitieswere concentrated at these three localities(gridded areasin Fig. • Presentaddress:Ornithologyand Mammalogy, 1), which supportedgoodnumbersof birds. A few CaliforniaAcademyof Sciences,GoldenGate Park, birds were also banded in smaller isolated tracts of San Francisco, California 94118 USA.
meadow between Tioga Passand Lee Vining Creek: 537
The Auk 99: 537-547. July 1982
538
BAPTISTA AND MORTON
[Auk, Vol. 99
Gardisk a•
Tioga Pk 3509
rn
Ellery Lake 2892
m
Dana
',•Gaylor
:)Lake 2942m •
,,XPk3354m
'z.' Yosemite Notional Park
'-
Piõ. ]. Map of studyarea.Hatchedareas(TioõaPassand GardiskyLake)and cross-hatched area(Lee Vininõ Creek)arelocalitiesof mostintensebandinõactivities.At LeeVininõ Creekthe two dialectsoccur sympatrically.
with only one exception(a bird at Mine Creek, Fig. 1, that sangthe GardiskyLake dialect),all birds in the isolatedmeadowssangTioga Passthemes. We color banded birds to study breeding success,
tive matingaccordingto dialect.Songsof maleswere recordedin the field on Nagra 4.2 and Nagra E tape
recordersusinga Dan Gibsonparabolicmicrophone. Songswere analyzedon a Kay electricsoundspecdispersalbetween the two dialectalpopulations, trograph(Model 7029A)usinghigh shapeand wide populationturnover,and the possibilityof assorta- band filters. Because individual variation in White-
TABLE1. Number of birds singingsongtypesat the threerecordinglocalities. Song type Tioga
Recordinglocalities TiogaPass Lee Vining Creek
GardiskyLake Total
Tioga
variant
Bilinguals
Gardisky
Other
Total
14 7
4 1
2 2
1 4
1a --
22 14
2b
12
3
48
1
--
--
9
22
5
4
14
Songtypicalof Z. I. orianthain the CanadianRockies(seeFig. 3). One songtypicalof Z. I. gambeliiand oneof Z. I. orianthaat MountLassen(seeFig. 3).
1982]
Zonotrichia SongDialects andMateSelection
539
I
I
2
$
4 5 6
kHz
•)
sec
•I
Fig. 2. A. Song typical of Gardisky Lake. Numbers correspondto charactersmeasuredand treated statisticallyin Table 2. B. Syllablesoccurringin songsrecordedat Tioga Passand Gardisky Lake. 1. Complex syllabletypical of GardiskyLake songs.2. Two complexsyllablestypicalof TiogaPasssongs.Thesesyllables typically occurin pairs. 3. Complex syllablefrom variant song recordedat Tioga Pass.4. Simple syllable precedingterminaltrill in GardiskyLake songs.5. Simplesyllablefrom terminaltrill in GardiskyLakesongs. 6. Simple syllablefrom terminal trill in Tioga Passsongs.
crown song is slight (review in Baptista1977),we selectedthe deanestspectrogramfrom eachbird to measure18 parameters(Table2, Fig. 2).
rounding hillsidessang the Gardisky dialect. One bird sang the dialect of Tioga Pass, one bird sangthe songof anotherpopulationto the At the end of the 1979 breeding season,we cap- north, and one bird sangthe songof another turedfemalesat GardiskyLakeand LeeVining Creek subspecies(seebeyond).The complexsyllable mated to malesfrom which songshad been recorded during the breeding seasonand induced them to typical of Gardiskywas of type 1 (Fig. 2B), and sing in the laboratorywith 0.2 ml intramuscularin- the trills were alwaysprecededby syllabletype jectionsof testosterone (SearleSC-16148,50 mg/cc). 4, which is absent in Tioga songs.The trills themselvesconsistedof syllabletype 5. Songs from Tioga Pass were identical to RESULTS AND DISCUSSION those of Gardisky Lake in the sequencingof INDIVIDUAL VARIATION IN SONG elementsbut differed in the morphologyof the Four of 48 malesin this study sangtwo song types (Table 1). Two of these bilingual birds were recorded at Tioga Pass and two on Lee Vining Creek(Fig. 1). The restof the birds sang one theme each, with minor variation typical of White-crownedSparrowsong(Orejuelaand Morton 1975, Baptista 1977). THE SONG DIALECTS
The main theme recordedat Gardisky Lake began with a whistle and was followed by a buzz, a complex syllable, a trill, and a lowpitched terminal buzz (Fig. 2A). Nine of the 12 birds recorded
at the lake shore or on the sur-
syllables(Fig. 2B). Complexsyllableswere of types2 and 3, and simple syllableswere type 6. Songscontainingcomplexsyllable3 (=Tioga variant of Table 1) were less common than
songscontainingcomplexsyllable2 and were utteredby 6 of 22 birdsrecorded.Two of these six birds were bilingual, using themes with both syllabletypes.Themeswith syllabletype 3 have also been recorded at Mount
San Gor-
gonio, San BernardinoMountains,450 km to the south (Baptistaand King 1980).The introductorybuzz in Tiogasongswas alsopreceded by a note(Fig. 3E), absentin Gardiskythemes. At LeeVining Creek4 of 14birds (28%) sampled sang the Gardisky Lake dialect and 10
540
[Auk, Vol. 99
BAPTISTAAND MORTO•
TABLE 2. Characters of GardiskyandTiogadialectsof Z. I. oriantha, .• + SD (n). Gardisky a
Tiogab,½
Duration (s)
1. Entiresong 2. Introductory whistle 3. Introductory buzz
1.87+ 0.12(13) 0.45+ 0.02(13) 0.25+ 0.02(13)
4. Terminal buzz
0.26 + 0.05 (13)
1.81 0.49 0.23 0.31
+ + + +
0.16 (21) 0.02 (24)* 0.02 (24)* 0.06 (21)*
6.13 2.17 3.95 3.69 6.00 3.92 2.65 6.03 3.30 2.74 2.55 2.17 0.47
+ + + + + + + + + + + + +
0.21 (22)* 0.34 (22) 0.35 (22)* 0.32 (24)* 0.28 (22)* 0.40 (22)* 0.57 (22)* 0.40 (22)* 0.21 (22)* 0.52 (22)* 0.10 (21) 0.34 (22) 0.15 (18)
Frequency(kHz)
5. 6. 7. 8. 9.
Maximumof song Minimumof song Frequency envelope of song Meanof introductory whistle Maximumof introductory buzz
10. Minimum of introductory buzz
11. Frequency envelope of introductory buzz 12. Maximum of trill 13. Minimum of trill
14. Frequency envelope of trill
7.00+ 2.10 + 4.81+ 3.25+ 4.85+ 3.15 + 1.67+
0.54(12) 0.19(13) 0.48(12) 0.10(13) 0.22(13) 0.19(13)
0.34(13)
6.33 + 0.43 (13) 3.12 + 0.26 (13)
3.21+ 0.58(13)
15. Maximum of terminal buzz 16. Minimum of terminal buzz
2.52 + 0.19 (13) 2.10 + 0.16 (13)
17. Frequency envelope of terminalbuzz
0.42+ 0.12(13)
Elements
18. Number of
10.85 + 1.14 (13)
10.00 + 1.48 (22)
Gardisky songs fromGardisky Lake,LeeViningCreek,andGlacier Canyon arepooled. Tiogasongs fromTiogaPassandLeeViningCreekarepooled.
ß = p < 0.005(two-tailedt-tests)whenGardiskyand Tiogaare compared.
birds (72%) sangTiogasongs.We regardthis these differences are real, as we could hear as an area of secondarycontactbetween the them even before examining the sound spectwo dialectal populations. trograms.Trill syllablesalso differ in tonal In addition to differencesin syllabic mor- quality (comparesyllabletypes 5 and 6, Fig. phology, the Tioga and Gardisky dialectsalso 2B). There is more energy distributed in the differ from eachotherin a numberof frequen- lower frequenciesin Tioga syllables(type 6). cy and temporal characteristics.Both song The pitch of the songis higher in the Gardisky types contain the same number of elements population (character5). (character18, Table2). Songduration alsodoes In summary,Tioga and Gardiskysongsdifnot differ between the two populations(char- fer in 12 of the 18 charactersquantified. Three acter1). The Tioga songs,however, appearto charactersare temporal, and nine pertain to haveslightlylongerintroductorywhistlesand pitch. terminal buzzes (characters2 and 4). IntroducMarie Mans sent us songsof eight Whitetory buzzesare slightlyshorterin Tiogasongs, crowned Sparrowsrecordedat Virginia Lakes, and, moreover, each buzz is preceded by a Mono County (elevation2,926 m), 16 km north note absentin Gardiskysongs(e.g. a note pre- of Tioga Pass. Spectrograms of these record-
cedesthe buzz in songE, but not in songsA
ings were immediatelyrecognizableby us as
to D, Fig. 4).
the dialect recorded at Gardisky Lake. Marie (pers.comm.)notedthat otherbirds in the area
The pitch of the introductorywhistleand introductory buzz are higher in Tioga songs (characters 8, 9, and10).Thefrequencyenvelope (width) of the buzz is also greaterin Tioga songs(character11). Trills in Gardisky songs are slightlyhigher pitchedand have a greater frequencyenvelopethan thosein Tioga songs (characters12, 13, and 14). Although slight,
alsosangthe samedialect.Thesedata indicate that the songs at Gardisky Lake (this study) representa southern extensionof the more northerly Virginia Lakesdialect. Playbackstudiesandassortative mating.--Data on femaleresponsesto playbackare confusing. In a laboratorystudy,Bakeret al. (1981a)found
1982]
ZonotrichiaSongDialectsandMate Selection
541
other with complexsyllable3. This individual would increase the proportion of one or the other theme in its reply bouts to matchwhich-
1979
•980
everthemewe playedto it. Thesedataindicate thatbirdsmaydistinguishbetweensongscontaining syllabletypes2 and 3. It is thus likely that theycanalsodistinguishbetweenthe GardiskyversusTiogadialects,which differ in the morphologyof their syllablesaswell asin pitch and tempo.This shouldnot be surprising,becausecoastalWhite-crowned Sparrowsare ca-
pable of distinguishing between songs of neighborsversus strangerssinging the same dialect (Bakeret al. 1981b).Moreover, the differ-
encesbetweenresponses to songsof neighbors versus strangers (t-test, P = 0.03) appear as
great as those to songsof home versusalien dialects(t-test, P = 0.032) (Baker et al. 1981c). Petrinovich and Patterson (1981) showed
6___ E
that female responsesto the playbackof an alien (adjacent)dialect were strongerthan to that of the homedialectin all parametersmeasured.Males,however,sangsignificantlymore
kHz
sec
•
Fig. 3. A, B. Song of Z. l. orianthamale at Gar-
disky Lake, which is similar to Z. l. gambeliithat
to the playback of the home dialect but trilled and fluttered more often in responseto the playbackof an alien dialect.
If the femaleresponses are regardedas sexmigratethroughthe area, recordedin 1979and 1980, respectively.C. Song typical of Canadian Rockies ual, then the above data support a disassortasungby bird breedingat GardiskyLako.D. Typical tive mating hypothesis.We agree with PetriGardiskyLake songfor comparison. E. TiogaPass novich and Patterson(1981), however, that the songrecordedat GardiskyLake.The syllabletype in responsesare very likely aggressiveand that the terminal trill is unique to this individual.
no meaningfulconclusionwith regardto mate choice based on dialect may be drawn from playback studies. Exceptionalsongs.--A male recorded at Garthat females respondedmore to the playback disky Lake in 1979sanga songtypicalof Z. I. of the home dialect than to an alien dialect. In gambelii, a subspeciesthat breeds in Canada a field study, Petrinovichand Patterson(1981) and Alaska(Fig. 3A). The complexsyllablefolfound that femalesrespondedmore stronglyto lowing the introductorywhistle in this bird's song was identical with warble type 6 of the playbackof an alien dialect. DeWolfe et al. (1974).We capturedand marked Do the songsdescribedhereinrepresentdialects
in
the
functional
sense? Are
White-
this individual
and noted its reddish
brown
bill and black lores, which distinguish Z. I. crowned Sparrows capable of discriminating orianthafrom Z. I. gambelii(the latter having between Gardisky and Tioga songsbased on yellowish-pink bill and white lores).This same the differencesdescribed?We have not yet bird waspresentat GardiskyLakein 1980(Fig. conductedplaybackexperimentsutilizing the 3B). It was mated, but we were unable to locate two dialects;the followingobservations,how- its nest and thus cannot comment on its breedever, suggestthat birds may distinguishbe- ing success.Morton et al. (1973, this study) tween them. have documentedflocksof Z. l. gambeliicomIn the 1981breedingseasonwe recordeda ing throughTioga Passin Septemberand minbird at Tioga Passsingingtwo themes,one glingwith the localbirds.Thesemigrantsoften containingcomplexsyllable2 (Fig. 2) and the singin passage,especiallyduring violationsof
542
BAPTISTA ANDMORrO•
individual distance (pers. obs.). We have ob-
served fledgling Z. I. orianthabeing fed by their parents as late as September.Such fledglings could possibly learn alien dialectsfrom
birds passingthrough.Alien songscouldalso be learnedin their wintering grounds,where several subspeciescome in contact (Baptista and King 1980). A secondbird recordedat Gardisky Lake in 1979 sang a song that contained most of the characteristicsof Tioga songs(Fig. 3E), including the note precedingthe introductory buzz and complexsyllable2. The terminal trill and
[Auk, Vol. 99
quavering and not crispas in typical male song (songsA-G). Nonetheless,with only one exception(songB1, Fig. 4), all femalesongswere unequivocally assignableto the Tioga or Gar-
disky dialectalpopulations.FemalesongsElG1 (Fig. 4) clearlycontainsyllabletype 1 (Fig. 2) of the Gardisky population. SongsA1, C1, and D1 contain syllabletype 2 of Tioga Pass. Two males from Gardisky Lake that sang the
local dialect were mated to femalessinging the samedialect (songsF and F1, G and G1 in Fig. 4). One male from GardiskyLake that sangthe Tioga dialect (song E in Fig. 4) was mated to buzz, however, were unlike that of any local a female that sang the Gardisky dialect (song bird
but matched
those of birds
recorded
at
El). One male singing the Gardisky dialect (songB, Fig. 4) was matedto a female (B1)that sang a song foreign to both dialectal areasin this study. The syllablesof her trill are similar to simple syllable 22 of Baptista and King (1980), thus far known only from Mr. Lassen oriantha in the Rocky Mountains of Alberta amongZ. I. orianthapopulations(Orejuelaand (Lein 1979, Fig. 3C this study; this bird was Morton 1975). The other males from Gardisky banded at Tioga Pass as an adult in August Lake that sang the local dialect (songsA, C, 1979).The four whistles,alternatingin low and and D) were mated to femalessinging the Tiohigh frequency followed by two buzzy ele- ga dialect (songsA1, C1, and D1). The three mentsin descending pitch,,aretypicalofsongs femalesfrom Lee Vining Creeksangthe Tioga recordedby Lein. Despiteits aberrantsong(for dialect,whereastheir matessangthe Gardisky Tioga Pass),this bird mated and fledgedthree dialect (Fig. 5). It is noteworthy that the only offspring.The male paired with a different fe- male singing a Tioga dialect (Fig. 4E) was matmale in 1981and fledgedthree more offspring. ed to a female singing a Gardisky dialect, deDespite singing a foreigndialect,it was paired spite the fact that the Tioga dialect was the very early in the breedingseason. more predominant among the females samWhat is the relativefitnessof this alien sing- pled. er compared to normal songstersin the popuIf data from GardiskyLake and Lee Vining lation? Fledging successat Tioga Passbetween Creekare pooled, then only 2 of the 10 females 1968and 1970rangedfrom 1.4 to 2.0 fledglings sangthe samedialectastheir mates,indicating per pair (Morton et al. 1972).If fledgingsuccess that mating may be disassortative(two-tailed may be regardedas an index of fitness,it is binomial probability, P = 0.11). If Gardisky Mount Lassenby Orejuela and Morton (1975). This bird was also mated, but again the mate was not captured and the nest not located. In 1980we recordeda Tioga Passbird singing a song similar to those recorded for Z. I.
noteworthy that this individual's unusual song did not render it lessproductivethan othersin the population. Songsof matedpairs.--Songs of 10 females were analyzed, 3 from Lee Vining Creek and
7 from GardiskyLake. The quality of female songvaried individually and alsoas a function of the amount of testosteroneinjected. Some femalessangfrequentlyabout 3 days after one injection. Others required two or three injec-
tions, each 3 or 4 days apart, before large enoughnumbersof "quality" songswere sung
Lake data are treated separately,then two pairs matchdialectsand five pairs do not (two-tailed
binomial probability, P = 0.45); mating may thus be random with respectto dialects. Both disassortativeand random mating hypotheses are treated at greaterlength below. Assortativemating.--Prerequisitesof the assortativemating theory are that (1) maleslearn and sing the song of their natal area, and (2) femalesalsolearn that songand are selectively attracted to it during pair formation (Nottebohm 1969: 312). Our recordingsfrom testos-
to permit adequate sampling. Some females terone-injectedZ. I. orianthado not support seldomsang songsof a quality comparableto the secondprecondition: only 2 of 10 females thoseof males; e.g. female song F1 (Fig. 4) is sang songs_matchingthose of their mates.
J•L¾1982]
Zonotrichia SongDialects andMateSelection
c
543
cI
El
F
FI
Fig. 4. A to G. Songsof malesrecordedat GardiskyLake. Exceptfor songE, which is a TiogaPasssong, all malesillustratedsangtypical Gardisky Lake songs.A1 to G1. Songsof femalesfrom GardiskyLake. A is matedto A1, B to B1, etc. Note that malesF and G are matedto femalessingingsimilar(GardiskyLake) songs.The malesinginga TiogaPasssong(songE) is matedto a femalesinginga GardiskyLakesong(El). One male singinga GardiskyLake song(B) is matedto a femalesingingsyllablestypicalof Mount Lassen songs.The othermalessingingGardiskyLakesongs(A, C, and D) are matedto femalessingingTiogaPass songs(A1, C1, and D1).
Moreover, 3 males and 1 female singing alien dialects were mated, and at least 1 bred suc-
mating is known in birds and has been describedin the congenericWhite-throated Spar-
cessfully.A parallel study with Z. I. nuttalli yielded similar results(Baptista1974;Petrinovich et al. 1981, in prep.). Female WhitecrownedSparrowsmust be using cues other
row (Z. albicollis; Lowther 1961, Lowther and Falls 1968). The term "isassortative," how-
than natal dialectsin selectingconsorts.
her natal area. While this may be the case,we feel that, until the pair formation processand
How might one explainthe mismatchingof dialectsin matedpairsof White-crownedSparrows?Four possibilitiescometo mind.
ever, implies a deliberate choiceby a female of a matesinginga dialectdissimilarto that of
mate choice are studied in detail, the term "disassortative"
is best not used for the White-
1. Dialects promote disassortative mating and thus ensure outcrossingbetween demes, i.e. femalesselectmalessinging songsdissim-
crowned Sparrow. Simpler alternate explanations are feasible, as developedbelow.
ilar to that
singing foreign dialectswere paired suggests
of their
natal
area. Disassortative
2. The fact that three males and one female
•4
BAPTISTA ANDMORTON
[Auk, Vol. 99
AI
[]
BI
cI
kHz. 2 • Fig. 5. Songsof threematedpairs of White-crownedSparrowsfrom Lee Vining Creek.Male songsare illustrateddown the left columnand thoseof their matesdown the right column.Note that femaleC1 (paired to male C) usessyllabletype 3 of the Tioga dialect, the only female to do so.
thatfemalesmay not pay attentionto songtype during pair formation,so that matingmay be random with respectto dialect. If so, the mis-
formedpart of this sample.Fromthe cohortof 325 nestlings,15 malesand 4 femalessubsequently returnedas breeders(Table 3). From the cohortof 409dispersingjuveniles,36 males
matchingof dialectsbetweenmatedpairs may be explainedby differentialdispersaldistances and 17 femalesreturned.All of the remaining between the sexes. That is, males settle in their
recruits, 28 males and 72 females, were ob-
first breeding seasonat a short distancefrom their birthplace,whereasfemalesdispersefarther, possiblyinto anotherdialectalarea. We recovereda breedingfemaleat Gardisky Lake that was hatched at Tioga Pass 3 km away.The femalesangthe Tiogadialectof her natalpopulationbut was matedto a malesinging the foreign(Gardisky)dialectof her breeding area. Differential dispersaldistancesbe-
servedfor the first time on the study area as adults.Thesedata are dearly consistentwith a dispersalpattern wherein femalesemigrate farther than males.
3. Mismatchingof dialectscouldbe due to
unequal sexratiosof breedingrecruitsas a resuitof differentialmortalitybetweenthe sexes. Assumingphilopatryand greaterannualmortality of females than males, recruitment of tween the sexes are well known in birds. breedingfemalesmust be from outside,again Femalesusually dispersefarther than males resulting in mismatched dialects between (Greenwood 1980, Van Hecke 1981). Indeed, membersof a breedingpair. Thus far, there Bakerand Mewaldt (1978)have presenteddata have been 172 recruitsto our breedingpopulation, 79 males and 93 females. These numbers suggestingthat this is so in Z. I. nuttalli. Resultsof our banding studieson Z. I. or-
iantha also suggestthis relationship. Since 1978, we have banded all adults on our study areaand asmany of their nestlingsaswe could find, a total of 325. A secondgroup of immature birds was also banded.
TABLE 3. Sources and numbers of birds recruited
as
breedersinto the study population.
These were inde-
pendentjuvenilesmovingontoor throughthe Sources(n) study areain late summer,a total of 409. Membersof this group, althoughfrom off the study Nestlings(325) Dispersingjuveniles (409) area,are thoughtto be localin origin. For ex- Adults (100) ample, nestlings that we banded on Mine Total Creek and in Glacier Canyon (see Fig. 1)
Recruits
Males
Females
15 36
4 17
28
72
79
93
JVL¾ 1982]
Zonotrichia Song Dialects andMateSelection
do not deviatesignificantly from parity (X2 = 1.13, P > 0.05). Each year, however, there seem to be a few more female than male re-
cruits, and we shouldnot yet rule out the possibility that differential mortality is occurring. 4. Mismatching of dialectscould be due to sexual differences in dispositions to learn
545
ever, that song may function as a behavioral characterexternalizingindividualfitness.Correlationsbetweenbreedingsuccessor territory quality and songcharacteristics have been reported in severalspecies(Kok 1972, Howard 1974,Payneand Payne1977).Kroodsma(1979) documented leader-follower roles during
songsof neighborsat sitessettled.We have countersingingin the Long-billedMarshWren palustris).He demonstratedthat presenteddataindicatingthatmostfemales(72 (Cistothorus of 93 or 77.4%) were recruitedfrom outside the study population and were never seen as ju-
leader-followerroles are ritualized expressions
veniles. They might have brought with them the dialect of their birthplaceand thus do not sing the local dialect.Of 79 of the malesre-
duelling, togetherwith loudnessof delivery, is a reflectionof ageand size and may be used by femalesascuesto assess vigor and potential fitness.It hasbeenpostulatedthat youngbirds (1) learn selectively the songs of fit males (Payne and Payne 1977, Jacobs1979) or (2) matchthe songsof territorialmalesat sitessettled, becauselearningtheir songbestowssome advantagein gaining and holding a territory (Payne 1978). Payne (in press) has recently demonstratedthat mating and nesting success of yearlingIndigo Buntings(Passerina cyanea) was higher in individuals that shared songs with olderneighborsthan in thosethat did not share songs.He suggestedthat their success may be due to somedeceptivemimicry directed at other malescompetingfor breedingterritories.Baptista(1975)hasnotedthat songsof nearestneighborsaremoresimilarto eachoth-
cruited, however, 28 (35.4%) were also seen
only as adults.Why do mostof thesesing the local dialect?
Males of some specieslearn dialects after dispersal, acquiring the songsof territorial males at sites settled (Kroodsma 1974, Verner
1975,Payne1978).Payne(1981)hastermedthis "socialadaptation,"assumingthat malesbetter their social statusby learning new songs. Males learn songsfrom their natal area, disperse,and learn new songsfrom localterritorial malesin the areasettledduring male-male interaction. Dispersed propagules soon countersingwith local males using the local dialect; the song learned earlier in their place of birth falls into disuse and goesundetected
by the field observer.Femalesalso disperse afterhaving learnedthe dialectof their birthplace.Because femalesare usuallysubordinate to males, countersingingbetween sexes seldom occurs,so that mostdispersedfemalesretain the songsbrought with them. It is also possible that femaleshave shorter sensitive periodsthan males,sothatfemalesceaselearning songsat an earlierage than do males. If dialects in White-crowned Sparrows do
of dominance/subordinance
and
that
vocal
er than to those of more distant individuals
in
Z. I. nuttalli, forming "subdialect" groups. This could be due either to birds learning
songsin their natal area and dispersingshort distancesor to birds dispersingfrom without and learningsongsof territorialmalesat sites settled. ACKNOWLEDGMENTS
We thank Paul Meade, Maria Elena Pereyra,May Wakamatsu, and Eileen Zerba for assistance in the
not promoteassortativemating, why do birds field. Marie Mans sharedwith us her unpublished learn them? Bird song is thought to function data on songs from the Virginia Lakes area. Judy in territorial advertisement and in attracting a
mate. Brown (1975:674)has suggestedthat females of monogamousspecieschoosemates more on the basis of territory quality than on the quality of individual males, in which case the role of songin territorydefensewould be more important than its role in attracting mates. Vocal copying would then be selected if the more complex culturally transmitted songs were the most effective in repelling males.
A number of authorshave suggested,how-
Merrill, Maria ElenaPereyra,and Alida Solishelped prepare the figures. R. B. Payne and an anonymous reviewer made valuable comments on an earlier draft
of this paper. Lewis Petrinovich advised and assisted us with some of our statistical tests. Our field studies
were supportedby NSF Grant DEB-77-12980.Searle Laboratorieskindly provided us with the testosterone used in this study. LITERATURE CITED
BAKER,M. C., & L. R. MEWALDT.1978. Song dialects as barriers to dispersalin White-crowned Sparrows, Zonotrichialeucophrysnuttalli. Evolution 32: 712-722.
546
BAPTISTAAND MOUTON
--,
K. J. SPITLER-NABORS,& D.C.
BV,•DLE¾. KREBS,J. R., & D. E. KROODSMA.1980. Repertoires
1981a.Earlyexperiencedeterminessongdialect responsiveness of femalesparrows.Science214: 819-821.
--,
[Auk, Vol. 99
D. B. THOMPSON,& G. L. SHERMANß1981b.
and geographicalvariation in bird song. Advß Study Behav. 11: 143-177. KROODSMA, D. E. 1974. Songlearning, dialects,and dispersalin the Bewick'sWren. Zß Tierpsychol.
Neighbor/stranger songdiscrimination in White-
35: 352-380.
crowned Sparrows.Condor 83: 265-267ß
1979. Vocal dueling among male Marsh Wrens: evidence for ritualized expressionsof
--,
& M. A. CUNNINGHAM.
19•1C. The'role ofmale versus male int•cactions dominance/subordinance. in maintaining populationdialectstruct•lre.Behav. Ecol. Sociobiol.
8: 65-69.
Auk 96: 506-515.
LEIN, M. R. 1979. Songpattern of the CypressHills population of White-crowned Sparrows. Can. Field-Natur.
BAPTISTA, L. F. 1974. The effectsof songsof winter-
93: 272-275.
ing White-crownedSparrowson songdevelop- LOWTHER, J. K. 1961. Polymorphismin the White-
--.
ment in sedentarypopulationsof the species.Z. Tierpsychol.34: 147-171. ß 1975. Songdialectsand demesin sedentary populationsof the White-crownedSparrow(Zonotrichia leucophrysnuttalli). Univ. California
throated Sparrow, Zonotrichiaalbicollis(Gme-
Publ.
whees, finches, sparrowsand allies (A. C. Bent
Zool.
row. Pp. 1364-1392in Life historiesof North American cardinals, grosbeaks,buntings, to-
105: 1-52.
1977. Geographicvariationin songand dialectsof the PugetSoundWhite-crownedSparrow. Condor
--,
lin). Can. J. Zoolß 39: 281-292. , & J. B. FALLSß1968. White-throated Spar-
79: 356-370.
& J. R. KING. 1980. Geographicalvariation in song and song dialectsof montaneWhitecrownedSparrows.Condor82: 267-284.
and collaborators). U.S. Natl. Mus. Bull. 237.
MARLER,P. 1970. A comparativeapproachto vocal learning: song developmentin White-crowned Sparrows.J. Comp. Physiol. Psychol.Monogr. 71: 1-25.
--,
BLANCHARD,B. D. 1941. The White-crowned Spar-
rows (Zonotrichialeucophrys) of the PacificSeaboard: environment and annual cycleßUniv. California
Publ. Zool. 46: 1-178.
BROWN,J. L. 1975. The evolutionof behaviorßNew
& M. TAMURA. 1962. Song "dialects" in three populationsof White-crownedSparrows. Condor
64: 368-377.
MORTON,M. L. 1976. Adaptive strategiesof Zonotrichiabreedingat high latitude or high altitude. Proc. 16th Intern. Ornithol. Congr., Canberra:
York, Norton Publ. Co.
322-336.
DAWSON,W. L. 1923. The birds of California. Los Angeles,SouthMoulton Co. DEWOLFE,B. B., & R. H. DEWoLFE. 1962. Mountain White-crownedSparrowsin California. Condor
1978.
row. Condor
nent variationsin the songsof Gambel'sWhitecrownedSparrows.Bird-Banding45: 224-252. GREENWOOD, P. J. 1980. Mating systems,philoparry and dispersalin birds and mammalsß Anim. Be-
bird (Mimuspolyglottos) song.Evolution28:428-
and the onset
of
80: 285-289.
Reproductivecycle and nesting successof the Mountain White-crowned Sparrow (Zonotrichia leucophrysoriantha)in the centralSierra Nevada.
, D. D. KASKA,& L. J. PEYTON.1974ßPromi-
hav. 28: 1140-1162.
conditions
, J. L. HORSTMANN,& J. M. OSBORN. 1972.
64: 378-389ß
HOWARD,g. D. 1974. The influence of sexualselection and interspecificcompetitionon Mocking-
Snow
breeding in the Mountain White-crowned Spar-
Condor
74: 152-163.
, --, & C. CAR•¾.1973.Bodyweightsand lipids of summeringMountain White-crowned Sparrowsin California. Auk 90: 83-93. NOTTEBOHM, F. 1969. The song of the Chingolo, Zonotrichiacapensis,in Argentina: description and evaluationof a systemof dialects.Condor
438.
71: 299-315.
JACOBS, C. H. 1979. Factorsinvolved in mate selec-
ß 1972. The origins of vocal learning. Amer. Natur.
tion and vocaldevelopmentin the Village Weaverbird (Ploceuscucullatus).Unpublished Ph.D.
--.
106: 116-140.
1975. Continental patternsof song variabil-
ity in Zonotrichiacapensis: some possibleecologicalcorrelates.Amer. Natur. 109:605-624. J.E., & M. L. MORTON.1975. Song diaducedsingingin femaleWhite-crownedSpar- OREJUELA, rows. Condor 74: 204-209. lectsin severalpopulationsof Mountain WhitecrownedSparrows(Zonotrichialeucophrys orianKo•(, O. B. 1972. Breedingsuccessand territorial dissertation.LosAngeles,Univ. California.
KERN, M.D.,
behavior
& J. R. KING. 1972ßTestosterone-in-
of male Boat-tailed
Gracklesß Auk 89:
528-540.
KONISHI,M. 1965. The role of auditory feedbackin the control of vocalization in the White-crowned
Sparrow.ZßTierpsychol.22: 770-783.
tha) in the Sierra Nevada. Condor 77: 145-153.
PAYNE,g. B. 1978ßMicrogeographicvariation in songsof SplendidSunbirdsNectariniacoccinigaster:populationphenetics,habitats,and song dialects.
Behaviour
65: 282-308.
JvL¾1982] --.
ZonotrichiaSongDialectsandMate Selection
547
1981. Populationstructureand socialbehav- PETRt•OVtC}t, L., & T. L. PATrEaSOn. 1981. The reior: modelsfor testingthe ecologicalsignificance sponsesof White-crownedSparrowsto songsof of songdialectsin birds. Pp. 108-120in Natural different dialectsand subspecies.Z. Tierpsyselection and social behavior (R. D. Alexander and D. W. Tinkle, Eds.). New York, Chiron Press.
chol. 57: 1-14.
ß--., & L. F. BAP'ris'r•,.1981.Songdialects as barriers to dispersal:a reevaluation.Evolu-
tion 35: 180--188. In press. Ecologicalconsequencesof song matching: breeding successand intraspecific VAN HECI•E, P. 1981. Ortstreue, Altersaufbau und songmimicry in Indigo Buntings.Ecology. Mortaliffit einer Population des Baumpiepers (Anthust. trivialis). J. Ornithol. 122: 23-35. --, & K. PAYNE.1977. Socialorganizationand mating successin local songpopulations of VilVgmqga,J. 1975. Complex song repertoire of male Long-billed Marsh Wrens in eastern Washinglage Indigobirds, Vidua chalybeata.Z. Tierpsychol. 45: 113-173. ton. Living Bird 14: 263-300. PE'rERSON, R. T. 1941.A fieldguideto westernbirds. Boston, Houghton Mifflin.
--.
