THE ROLE OF PLUMAGE DOMINANCE POLYMORPHISM RELATIONSHIPS WHITE-THROATED IN OF THE SPARROW DORISJ. WATT,• C. JOHN RALPH,2 AND CARTERT. ATKINSON3 Departmentof Biology,Dickinson College,Carlisle,Pennsylvania 17013USA ABSTRACT.--We studied dominancebehaviorsof captive winter flocksof White-throated Sparrows(Zonotrichia albicollis) in centralPennsylvania.Preliminarytestswith a smallgroup of birdsindicatedthat there might be differencesin behaviorsbetweencolormorphsif age and sexualdifferenceswere controlled.Studiesof two larger groupsproducedconsiderable variationin resultsbetweenthe groups.One group displayedstriking correlationsbetween dominanceability and color. In this group white adult malesdominatedothersmore than tan adult males did, whereas tan immature males, adult females, and immature females were more often dominant than were white immatureand female birds. The secondgroup did not displaysuchstriking differences.Tan adult and immaturefemaleswere more dominant than white females,as in the first group, but not significantly.Immature malesdisplayed the oppositetrend from the firstgroup--white morphswere moredominantthan tans.Adult malesin the secondgroup showed no clear trend. We also found differencesin dominance between tan and white femalesthat appearedto depend on season,white birds dominating tan birds in a small group in the spring,a reversalof relationshipsdocumentedin the fall. Within age-sexclasses,dominantfemalesin both large groupstendedto be duller in plumage brightness(scaledas an index) than were subordinatefemales.In one of the two large groups,duller immature maleswere more dominant than brighter ones,while brighter adult males tended to be more dominant than duller ones. We suggestthat the relationships betweenimmaturemale plumageand dominancemay be influencedby the morph of adult malesof specificdominancestatusin the flock. Spring plumage indiceswere more often correlatedwith fall dominancedifferencesthan were fall plumage conditions,suggestinga genetic or causalrelationshipbetween dominanceand plumage rather than a proximate (statussignaling) function. Recordedmorph frequenciesfrom this study and the literature supportthe hypothesisthat selectionis balancedbetweenthe two morphsand is dependent on the advantagesor disadvantagesof aggressiveness within a sex. Received9 March 1983, accepted25 July 1983. LOWTHER (1961) described two color morphs of the White-throated Sparrow (Zonotrichiaalbicollis),a white-striped and a tan-stripedform. He showed that both morphs included males and females. In field observations of 110 mated pairs, he observednegative assortativemating: white-striped males mated with tan-striped females,and tan-striped malesmated with whitestriped females(Lowther and Falls 1968). Var- • Present address: Biology Department, Saint Mary'sCollege,Notre Dame,Indiana46556USA.Requestsfor reprintsshouldbe sentto this address. 2 Present address: U.S.D.A. Forest Service, Red- wood SciencesLaboratory,1700 Bayview Drive, Ar- dy (1971)disagreedthat colorvariationwasdue to two morphs;she found color typesto have a continuous distribution. • PresentAddress:Departmentof PreventiveMedicine, Collegeof Veterinary Medicine, University of Florida, Box J-136, Gainesville, Florida 32610 USA. 11o recent studies chromosomal dimorphism in the speciesthat was always related to the plumage dimorphism; Atkinson and Ralph (1980) found birds in breeding plumage to be bimodally distributed with regard to quantitative plumage measurements;and Rising and Shields (1980) described other morphological differences between the color morphs.Severalof theseau- thorssuggestedthat there might be behavioral differences between the color morphs on the wintering grounds. We examined cata, California 95521 USA. More havesupportedLowther'sinterpretation,however: Thorneycroft(1966, 1975) documenteda differences in dominance be- haviors between the two color morphs in winter flocks. Several attempts to find differences in dominance behavior between the morphs The Auk 101: 110-120. January 1984 January1984] White-throated SparrowDominance during the nonbreedingseasonhave been made (Harrington 1973, Hailman 1975, Ficken et al. 1978), but none of these investigatorsconsidered the sexor age of the birds. We studiedthe interrelationshipsamong sex, age, and color morph and found that a combination of characterswas a goodpredictor of dominancerank. The purposeof this paper is to demonstratethe relationships of these variables with social dominance behavior of White-throated Sparrows in captive, nonbreeding flocks. METHODS Birdswere caughtin mist netsthroughoutOctober 1975 at Carlisle, Pennsylvania. Birds were individually marked at capturewith colored leg bandsand weighed. We determinedage by the amount of skull ossificationand sexby the unflattenedwing chord at capture (Atkinson and Ralph 1980) or by laparotomy the following spring.Birdswere kept in three indoor aviaries (2 x 2 x 2 m) under controlled artificial lights, initially on a winter light schedule. The daylength was increasedon 1 April to 12 hours light and 12 hours dark to stimulate prenuptial molt. By 16 April the birds were coming into breeding condition, as indicated by molt and singing. We investigated social relationshipswithin four groups of birds: a smallgroupof 16 malesand femaleson 28 November 1975;two largegroupsof 51 and 54 birds in November and December1975;and a small group of 17 females on 16 April 1976. Determinationof colormorph.--Even though Whitethroated Sparrows in breeding plumage can be divided into two distinct color morphs (Thorneycroft 1975, Atkinson and Ralph 1980), considerable individual variation exists in plumage characteristics (Vardy 1971).Atkinson and Ralph (1980) demonstrated that variation in plumage in the fall ranged from a bright to a dull extreme and that combinationsof various plumage characteristicsinto an index resulted in a normal distribution. In addition, plumage characteristicswere correlated with sex,age, and sea- birds arbitrarily into "white" or "tan" groups: birds with index valuesgreaterthan the medianwere called "white"; those with less were "tan." Because such plumage indices produced a relative comparisonof brightnessamong birds and becausewe did not prepare karyotypes,our designationsof morph may not coincidewith Thorneycroft'sgeneticallydetermined "color morph." Atkinson and Ralph (1980), as well as Thorneycroft (1966, 1975), however, found that, although the two morphsare difficult to distinguish in autumn,they are much more distinctin the spring. Therefore,we used spring-plumageindex valuesto approximate the morph class most similar to Thorneycroft'sgeneticallydetermined "color morph," and, in fact, only 9 of the 105 birds would have differed in their morph classification betweenfall and spring plumages. Behavioral observations.--Behavioral interactions were observedthrough a one-way window overlook- ing a feeding platform in eachof the three aviaries. During observationperiods, food was restricted to a small dish on the platform. We recordedwinners and losersof dominanceencountersat the feeding dish. Winning birds supplanted,pecked, or chasedthe losing individuals. Using the methods of Sabine (1959) and Brown (1975: 86), we determined dominance hierarchiesfor the two smallgroupsof birds by constructingdominancematricesof encounters.Large groups did not form clear linear hierarchies, precluding ranking of individual birds. Instead,we tallied dominancedyads for each morph-age-sexclass. A dominance dyad between two birds [i.e. A > B: the "pair relations" of Sabine (1959)] was designated when one bird was dominant in more than one-half of the encountersbetweenthe pair members,regardlessof the numberof interactions.The percentageof dyads in which a bird was dominant was defined as its "percentage dominance," or the number of individuals it consistently dominated in known relations. In the two large groups,the numbersof dominant dyadswere combinedfor membersof eachage- sex-morphclassto give the percentagedominanceof each class. Statistics.--Because the percentagedata were tested son: males were generally brighter than females, and found to be normal, they were not transformed. adults brighter than immature birds, and all birds Statistical tests included a test for differences bebecamebrighter in the spring (Atkinson and Ralph tween two percentages,product-momentcorrelation, and Mann-Whitney U-Tests(Sokal and Rohlf 1969). 1980). In this study, we calculatedan index of plumage brightness similar to Atkinson and Ralph's (1980), using four of their five plumagecharacteristics: percentageblack in lateral crown stripes,throat pattern (asclassifiedby Lowther 1961),median crown-stripe color, and superciliary stripe color. We did not include chest streaking becauseof its low variability between seasons.The four plumage characteristics were measuredat capture in the fall and again in May after prenuptial molt. We used the median value of the indices to divide RESULTS SMALLGROUP(NOVEMBER) We conducteda preliminary investigationof a small group (16 birds) in November. The group exhibited an essentiallylinear and stable hierarchy (Fig. 1). Inspectionof physical characteristics correlated with dominance ranks of 112 WATT,RALPH, ANDATKINSON [Auk,Vol. 101 LOSER BIRD RANK 2 •,l 4 i li113 ! z 9 12 • I I 4 •2 • • • • • • •2 4 3 2 3 3 •or 16 •hRe4hroa•ed •parrows •n •ove•ber. The h•erarchy •or •ve •ns•ances(b•rd ranked •8 was do•nan• b•rd a•acked i 2 / I• a subordinate iI 4 ! :5 :5i I t 2 I : 2 I :5 I •g. 1. Dominance•a•r• 21 a do•nan• •o b•rd ranked •2• e•c.). Two reversals (R) occurred where b•rd. these individuals (Table 1) indicated that sex class with another class would be a convenient was the best single predictor of dominance. meansof analysis.Beforewe couldproceedwith this analysis,however, it was necessaryto deOther measures, such as morph, age, wing length, or weight, did not explain dominance termine how the proportionof birds in eachof rank simply. Among females, however, tan the two colormorphsin a classmight affectany birds were more dominant than white birds. measurementof aggression,as Hailman (1975, That is, if birds of The relationshipamong maleswith respectto pers.comm.)hassuggested. morphwasnot clear.The resultsof thisprelim- one class encountered birds of another class at inary studyled us to believethat differencesin random, then it would be possibleto use endominance behaviors between white and tan counter frequenciesas a measure of domimorphsmight be found in largergroupswithin nance. If encounter frequencieswere not random,then comparisons of morphssummedover sex and/or age classes. the variousage-sexclasseswould not be valid. LARGE GROUPS Appropriateness of encounter-frequency analysis.--We wished to investigate the role that morph, age, and sex played in determining dominance in the two large groups. Measures involving the percentageof encountersof one Therefore, we calculated the number of ex- pectedencountersfrom the numbersof individualsin eachmorph-age-sexclassin the two large groups(Table 2) and comparedtheseto observed values (Tables 3 and 4). None of the eight classes,in either of the two groups,encountered the other classesrandomly, that is, January1984] White-throated Sparrow Dominance 113 TABLE1. Relationshipsbetween dominancerank and physicalcharacteristics of individual White-throated Sparrowsrepresentedin Fig. 1. Dominance rank Sex 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Morph M M M M F F F F F F M F F F F F Age White Tan White White Tan Tan Tan Tan Tan White Tan White White White White White Adult Adult Adult Immature Immature Immature Immature Adult Immature Adult Adult Adult Adult Adult Adult Immature Wing length Weight• (mm) (g) Weightb (g) 74 72 72 70 67 65 65 69 68 65 72 67 70 68 70 69 26.8 23.0 25.0 24.4 23.5 23.6 22.1 25.6 23.5 24.0 23.8 23.2 23.8 22.8 26.6 22.9 30.4 26.2 27.0 23.8 22.6 24.5 20.5 28.5 21.7 24.8 29.1 25.3 25.1 23.9 26.3 24.6 Weight at capture. Weight at testing,28 November. in proportion to class size. In general, males were observedin aggressiveencountersmore frequently, and females less frequently, than expected(row totals in Tables3 and 4, observed vs. expected). Therefore, encounters were not occurring randomly, and comparisons of morphs summed over classeswould not be an appropriate analysis. Similarly, an analysisof the number of wins (the "attacks" of Hailman 1975) would not be percentage dominance values of white adult males with those of tan adult males, the num- ber of dyads they form with other morph-agesex classesshould be equal. We tested these values (Tables 5 and 6) and found that for all eight comparisonsof age-sexclassesthere was no significantdifferencebetween white and tan morphs in the probability of forming dyadic relationshipswith the other classes,with a single exception(significantat slightlygreaterthan the P < 0.05 level). Becausethe assumptionof equalprobability was met, we summarizedpercentagedominancevaluesfor eachmorph-agesex classin each group (Table 7) for testing differencesbetweenthe morphswithin a given age-sexclass. Dominanceas measuredby dyad formation.Striking differencesin dominance between the appropriate,becausethey would be a function of encounter frequency. Also, an analysis of percentage wins or losses,based on total encounters, would be inappropriate, becausea particularly aggressivebird might inflate the percentagewins for its class. Appropriateness of analysisof dyadformation.Another method of comparing dominance relationshipsof the two morphs within age-sex classesinvolves the useof percentage dominance, the percentageof dyads in which a given class TABLE2. Numbers of birds in each morph-age-sex was dominant over another class. The use of percentagesalleviates the problem of nonrandom encounter frequencies,becauseonly one dyadic relationship is recorded, regardlessof the number birds. Given of interactions that a bird between is involved the two in a set of dyadic relationships,the percentage of those relationships in which it is dominant is independent of encounter frequency. The only remaining problem is the distribution of dyadformation frequencies: in order to compare classfor two large groups. Group Group Class Tan adult males (TAM) White adult males (WAM) Tan immature males (TIM) White immature males (WIM) Tan adult females (TAF) White adult females (WAF) Tan immature females (TIF) White immature females (WIF) 1 2 5 7 5 9 12 3 6 7 3 8 4 6 11 4 7 8 114 watt, RALPH,AND ATKINSON ;> '-o ;> co ;> co,.o co,.o [Auk, Vol. 101 January1984] White-throated SparrowDominance 115 116 WATT,RALPH, ANDATKINSON [Auk,Vol. 101 TABLE7. Percentagedominancein two groupsof each morph-age-sexclassover classesother than its own. Group 1 na Group 2 % tb Tan adult males White immature males Tan immature males 224 74.6 } 162 79.6 264 53.0 } 151 39.1 1.16ns Tan adultfemales White adult females 288 37 2.74** 1.39ns Whiteadultmales 35.4 24.3 142 35.2 } tan immature females 190 38.4 Whiteimmature females na % tb 85 68.9 51.8 } 2.71'* 187 36.4 148 69.6 } 6.16'* 196 69 4.23** 34.8 22.9 } 4.29** 176 43.75 0.60ns 192 • n - numberof dominancerelationships(dyads)consideredin the computationof percentagedominance. bStatisticalcomparisons were madewithin age-sexclassbetweenwhite and tan birds.** = a < 0.01;ns = not significant. two morphs, within each age-sexclass,were found in Group 2 (Table 7). In encountersbetween adult males, white morphs were more often dominant. In encounters between im- the expressionof color morph in the spring. We could not distinguishbetween causeand effectunder this last hypothesis.We did assess relationships between fall dominance and de- mature males, adult females, or immature fe- gree of changein plumagefrom fall to spring, males, however, tan morphs were more often dominant. Group 1 showedlessstriking differ- however. ences. In encounters between adult females or immature females, tan morphs were also more dominant than white morphs, but not significantly. In encountersbetween adult males or immature males, tan morphs were more dom- inant, a trend oppositeto that Shownby Group 2 birds. Only the relationship between immature males was significant, however. DIFFERENCES WITHIN Because we found dominance behaviors than did fall values (Table 8). To demonstrate the nature of the cor- relations,we plotted the spring-plumageindex of individual birds in each of the two groups againsttheir measuredpercentagedominance (Fig. 2). A significant positivecorrelation was found for adult malesin Group 2 (whiter birds were more dominant). Significant negative relationships were found for immature males in AGE-SEx CLASS instances We found that spring values of plumage brightnesscorrelated more often with winter of differences in Group 2, adult females in Group 1, and total females(immature femalesvaried in the appropriate direction but were not quite significant dominance between morphs within each agesex class,we investigated possible differences involving more continuous plumage variation within Groups 1 and 2. Specifically, we exam- birds increasein plumage-indexvalue from fall to spring (Atkinson and Ralph 1980),somebirds ined do so more than others. We found whether or not whiter birds were more dominantwithin eachage-sexclass(e.g.in adult males)by analyzing correlationsof percentage dominance with spring-plumage index values and with fall-plumage index values. If birds were using proximatecuesto evaluateprobable dominancerelationshipsbefore interacting,i.e. the statussignaling of Rohwer (1975), fall values should have the highest correlations with dominance. If genetic differences in dominance ability were correlated with color morph (apparentonly in spring),however,then spring valueswould be mosthighly correlated.A third and nonexclusive possibility was that winter dominance behaviors in some way influence when considered alone) (Table 8). While most that the de- gree of plumage brightening was negatively correlated with fall dominance behavior of adult females in Group 2 and of all females combined (Table 8). Females that were more dominant in fall had lessplumagebrightening in spring. FEMALE MORPH DIFFERENCES IN TWO SEASONS Thorneycroft (1975) noted that white femalesappearedto be more aggressivethan tan females on the breeding grounds. We found that tan females usually dominated white females and were more dominant to other birds January1984] White-throated Sparrow Dominance 117 TABLE 8. Correlationsof percentagedominancemeasureswith plumagemeasurestakenduring the fall and spring and degree of plumage brightening from fall to spring. Correlationsbetween plumage and dominancewere more often significantin spring than in fall. Correlation (r) Samples n Fall plumage• Spring plumage• Group 1 Group 2 12 11 -0.142 0.490 -0.064 0.606 Total 23 0.142 0.201 Group 1 Group 2 14 10 0.410 -0.801'* Total 24 - 0.193 47 0.067 -0.018 Group 1 Group 2 15 15 0.531' 0.101 -0.539* -0.490 -0.146 -0.617' Total 30 -0.242 - 0.376' - 0.260 13 15 0.062 -0.202 0.210 -0.347 -0.281 -0.368 28 - 0.143 - 0.286 - 0.305 58 -0.209 -0.358** -0.307* Adult Plumagebrightening• males Immature 0.144 males Total males Adult 0.181 0.172 0.243 0.781'* - 0.254 -0.159 -0.327 - 0.201 0.170 females Immature females Group 1 Group 2 Total Total females • ** = significant at a < 0.01; * - • < 0.05. than were white females. Because our results contradictedprevioussuggestions,we hypothesized that female dominance relationships might change with season.Tan females could be more dominant than white females in fall but becomemore subordinatein spring. To test this hypothesis,we assembleda small group of 17 femalesfrom the larger hierarchiesand observed their dominance behavior in April. Following prenuptial molt, white females were significantly more dominant than tan females (Table 9; Mann-Whitney U-Test; U = 57, P < 0.05). Moreover, a significant positive correlation was found between individual springplumage indicesand percentagedominancein April (r = 0.58, P < 0.05). Thus, it appearsthat behavioral relationships between morphs, at leastfor females,may changewith season.Becausemale birds were not present in this last test, we cannot rule out the influence males might have had on female dominance. DISCUSSION Dominance andplumagemorph.--Lowtherand Falls (1968) and Falls (1969) found that white White-throated Sparrows are more aggressive and are more persistent and frequent singers than are tan sparrows.Thorneycroft(1975) suggested that white females might be more aggressivethan their tan counterparts.Other investigators (Harrington 1973, Hailman 1975, Ficken et al. 1978) have suggestedthe samefor nonbreeding groups--white morphs are more aggressivethan tan morphs. Hailman (1975) reanalyzedmuch of Harrington's data and concluded that the data suggesting that white morphs were more aggressivethan tan morphs were inconclusive, becausethe relative proportions of morphs were not known. Taking into accountthe proportions of the morphs present during the encounters,Ficken et al. (1978) concluded that white morphs are more frequently the aggressorsthan are tan morphs. They sug- gestedthat there is a behavioral difference between the two morphs. Among the interpretations they discussedwas the possibility of a differential sex or age bias in the morph pop- ulations present. More important, they also suggested that, like its coloration, the tan morph's relative lack of aggressivenessrepresents a neotenic condition, in that in both be- havior and morphology it retains characteris- tics of young birds in adulthood: "tan head- 118 WATT,RALPH, ANDATKINSON [Auk, Vol. 101 io•7 ADULT o FEMALES o o •e o 0 ß o• 0 o o 0 o 0 [] o 0 25 ADULT MALES i.iJ o z z ioo IMMATURE IMMATURE FEMALES I-- [] 7• MALES [] Z •J •.• [] [] •J •. 5o ß O O O O ß eO o 25. ß ,o o I0 20 [] ß oo ß O0 ß 30 40 PLUMAGE I BRIGHTNESS 0 0 INDEX Fig. 2. Plot of percentagedyads in which individuals were dominant in November againsttheir plumagebrightnessindex asmeasuredafter spring molt (seetext for explanation). Circles (O) and squares([•) represent birds in Groups 1 and 2, respectively. stripe and relative unaggressivenessare characteristics of birds in their first fall of life." Our finding of relatively greater dominance in tan females in the fall is the reverse of the above reports for the spring and breeding seasons, and it disagrees with the neotenic hypothesis of Ficken et al. (1978). We feel that among females, in particular, a reversal in relative dominance occurs between seasons. It is possiblethat harassmentof white females by dominant males could force them to lower becamelessaggressivetoward potential mates. It is also possible, however, that white female birds actually becomemore aggressive,relative to tan females, in the spring. This could be due to hormonal changesaffecting each morph differently. Future studiesshould be able to determine whether the presence of males or a change in physiology is responsible for these changesin relative dominance levels. At this time, we do not know if adult males exhibit such a seasonalchange in dominance; levels in the fall hierarchy, and, subsequently in one group we found white adult maleswere in the spring, their levels might rise as males more dominant than tan adult males. Immature January1984] White-throated Sparrow Dominance malesin one group were more dominant when white, in the other group when tan. This obviously requires further testing, especially of all-malegroupsin both springand fall. Sexual differencesmay existbetweenthe colormorphs in their relationship to dominance rank, and further studiesrequire assessment of sexualeffectsin the analysis. Correlations between fall dominance behav- ior and the "spring" rather than the "fall" color morph support the hypothesis(Ficken et al. 1978) that behavioral differences between the color morphs are genetic rather than being basedon appearances alone(i.e. statussignals). We cannot rule out the possibility, however, that behavioral differences in the fall affect the expressionof color in spring plumages. Sex differences in morphratios.--Thorneycroft (1975)found amongyoung femalesthe ratio of tan to white was 1:1, while among malesit was 43:57. By breeding season, the ratios were 74:26 in females and 32:68 in males. 119 TABLE9. Relationships between dominance rank, percentagedominance, and color characteristicsin a group of 17 female White-throatedSparrowsin April. Dominance Color Springplumage Percentage rank morph index dominance White White White Tan White Tan Tan White White Tan White White Tan Tan Tan Tan Tan 36.84 26.15 35.51 19.80 37.42 24.15 23.13 36.83 34.33 24.96 35.66 35.52 13.63 8.78 21.79 11.26 16.12 93 69 67 76 75 63 69 69 44 40 29 40 19 29 20 15 13 i 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 In our study,36%of the maleswere tan, and 64%were white. Among females,62%were tan, and 38% were white (Table 2). These values are intermediatebetween Thorneycroft'sratiosof young morphs in the fall. [Wessel and Leigh (1941) studied fall birds, but they sexed most of their birds and thoseof his adult breeding birds and birds by plumage differences,a method now probablyreflecta normal autumn ratio, at least known to be inappropriate.] for central Pennsylvania. Plumagebrightnesswithin age-sexclass.--We Bodysizeand dominance.--Rising and Shields found that, within age-sex classes,spring(1980) found that white males were slightly plumage brightnesswas a good indicator of larger than tan males.Although they had small dominance in females, but not in males in two samplesof females,they showed that in 8 of 9 of four groups. The pattern of significance measurements, white femaleswere larger than found in one group leads us to hypothesize tan females. Based on body size alone, one that dominancein immature males might be might have predicted that white birds should influenced by the relations of the adult males dominate tan birds in both sexes. We found in in their group. Similarly to a speculationmenthe analysis of the small hierarchy (Table 1), tionedaboveregardingfemalerank, adult males however, that body size was not a good pre- may persecuteindividualsof their own morph dictor of dominance rank, within either age- classselectively,so that in groupswhere white sex classesor color morphs. males are strongly dominant, immature white Effectsof season.--Inthe studiesof Harring- males are strongly submissive.Studies of beton (1973) and Ficken et al. (1978), data were haviorsin groupswith different sex, age, and collected from 22 April to 4 May, when the morph ratioscould clarify thesequestions. majority of birds are in nuptial plumage With regard to Rohwer's (1975) hypothesis (Lowther and Falls 1968). Therefore, their re- that plumage variability in birds is used as a suitscan bestbe compardto thoseof the breed- statussignal, we doubt that it is being usedas ing studies(Lowtherand Falls1968,Falls1969). such by White-throated Sparrows.Plumage itOur results for a group of females in nuptial self is not a good predictor of dominance in a plumage agree with these previous studies; mixed-sexgroup of this species.For example,a white birds were more dominant than tan tan female in the autumn could not predict the relative status of a white bird without knowlbirds in late spring. Our study is the first to investigateand doc- edge of its sex--if it were a male, it would be ument dominance differences between color dominant to her; if a female, subordinate. Ad- 120 WATTß RALPH, ANDATKINSON [Auk,Vol. 101 ditionally, changesin signalswith seasonde- FALLS,J.B. 1969. Functionsof territorial song in the White-throated Sparrow. Pp. 207-232 in Bird creasethe predictivevalue of plumages.Plumvocalizations (R. A. Hinde, Ed.). Cambridgeß age variation may be important in facilitating Cambridge Univ. Press. individual recognition,however (Shields1977). FICKEN, R. W., M. $. FICKEN, & J.P. HAILMAN. 1978. Evolution.--Linkagebetween color mutations Differential aggressionin genetically different and aggressivelevels has been documentedin morphs of the White-throated Sparrow (Zonomammalsby Keeler et al. (1968, 1970).The intrichiaalbicollis).Z. Tierpsychol. 46: 43-57. teractionbetweenplumage,sex,and hormonal HAILMANß J.P. 1975. Analysisof aggressionin Whitecontrol of aggressionmay also be a considerthroated Sparrow types of different proportions. able influence on White-throated Sparrows. As an evolutionary strategy for females,it may be advantageousto be dominant to other femalesin fall foraging flocksand then to become more subordinate in order to pair more easilywith aggressivemales.This is the pattern Bird-Banding46: 236-240. HARRINGTON, B.A. 1973. Aggressionin winter resident and spring migrant White-throated Sparrows in Massachusetts. Bird-Banding44: 314-315. KEELER,C., S. RIDGEWAYßL. LIPSCOMBß& E. FROMM. 1968. The geneticsof adrenal size and tameness in colorphasefoxes.J. Hered. 59: 82-84. seenin tan females.The sametrend (more agß T. MELLINGER, E. FROMM, & L. WADE. 1970. gressivein fall but lessin the spring),however, Melaninßadrenalin and the legacyof fear. J. Hewould be a disadvantagefor tan males when red. 61: 81-88. competingwith white malesfor territoriesin LOWTHER,J. K. 1961. Polymorphism in the Whitethroated Sparrow, Zonotrichiaalbicollis(Gmelin). the spring. The apparent selectiveadvantages Can. J. Zool. 39: 281-292. of being tan, then, could be balancedthrough ß & J. B. FALLS. 1968. White-throated Sparopposingpressureson malesand females.This row. Pp. 1364-1392 in Life histories of North hypothesisis supportedby the trend of deAmerican cardinalsßgrosbeaks,buntingsßtowcreasingrelative frequenciesof tan malesfrom hees, finchesßsparrows and allies (A. C. Bent autumn to spring and the increasing relative and collaborators). U.S. Natl. Mus. Bull. 237. frequency of tan females (Thorneycroft 1975). RISING, J. D., & G. F. SHIELDS. 1980. Chromosomal White females would appear to be at a disadand morphological correlatesin two New World vantagein both seasons(assumingsubordinasparrows(Emberizidae).Evolution 34: 654-662. tion in winter and aggressivenessin spring ROHWER, S.A. 1975. The socialsignificanceof avian winter plumage variability. Evolution 29: 593are selectively disadvantageousfor females), 610. and, in fact, white femalesare proportionately SABINE, W. S. 1959. The winter society of the Orethe least numerousof any sex-morphclassby gon Junco: intolerance, dominance, and the breedingseason(Thorneycroft1975).The white pecking order. Condor 61: 110-135. morph, then, could be maintainedby its gen- eral advantage to adult males at all seasons. SHIELDS, W. M. 1977. The social significanceof avian winter plumage variability: a comment.Evo- ACKNOWLEDGMENTS SOKAL,R. R., & F. J. ROHLF. 1969. Biometry. San lution We thank J. B. Falls, J.P. Hailman, D. James, S. 31: 905-907. Francisco, W. H. Freeman and Co. THORNEYCROFT, H. B. 1966. Chromosomalpolymorphism in the White-throatedSparrow,Zonotrich- Rohwer, W. Shields,and R. Zink for help during the courseof this studyand for helpful commentson the ia albicollis(Gmelin). Science 154: 1571-1572. manuscript. What forbearancewe received from the 1975. A cytogenetic study of the WhiteReineman Sanctuarywas appreciated.Bill Jefferies throated Sparrow, Zonotrichiaalbicollis(Gmelin). and Noel Potter provided helpful support for our Evolution 29: 611-621. researchat Dickinson College. VARD¾, L. E. 1971. Color variation in the crown of the White-throated Sparrow, Zonotrichiaalbicollis. LITERATURE CITED ATKINSON,C. T., & C. J. RALPH. 1980. Acquisition of plumagepolymorphismin White-throatedSparrows. Auk 97: 245-252. BROWNß J.L. 1975. The evolution of behavior. New Yorkß W. W. Norton and Co., Inc. Condor 73: 401-414. WESSEL,J.P., & W. H. LEIGH. 1941. Studies of the flock organization of the White-throated Sparrow. Wilson Bull. 53: 222-230.