DORIS J. WATT, C. JOHN RALPH, 2 AND CARTER T. ATKINSON 3

THE ROLE OF PLUMAGE POLYMORPHISM IN DOMINANCE RELATIONSHIPS OF THE WHITE-THROATED SPARROW DORIS J. WATT, C. JOHN RALPH, 2 AND CARTER T. ATKINSON 3 Department f Bilgy, Dickinsn Cllege, Carlisle, Pennsylvania 17013 USA ABSTRACT.--We studied dminance behavirs f captive winter flcks f White-thrated Sparrws (Zntrichialbicllis) in central Pennsylvania. Preliminary tests with a small grup f birds indicated that there might be differences in behavirs between clr mrphs if age and sexual differences were cntrlled. Studies f tw larger grups prduced cnsiderable variatin in results between the grups. One grup displayed striking crrelatins between dminance ability and clr. In this grup white adult males dminated thers mre than tan adult males did, whereas tan immature males, adult females, and immature females were mre ften dminant than were white immature and female birds. The secnd grup did nt display such striking differences. Tan adult and immature females were mre dminant than white females, as in the first grup, but nt significantly. Immature males displayed the ppsite trend frm the first grup--white mrphs were mre dminant than tans. Adult males in the secnd grup shwed n clear trend. We als fund differences in dminance between tan and white females that appeared t depend n seasn, white birds dminating tan birds in a small grup in the spring, a reversal f relatinships dcumented in the fall. Within age-sex classes, dminant females in bth large grups tended t be duller in plumage brightness (scaled as an index) than were subrdinate females. In ne f the tw large grups, duller immature males were mre dminant than brighter nes, while brighter adult males tended t be mre dminant than duller nes. We suggest that the relatinships between immature male plumage and dminance may be influenced by the mrph f adult males f specific dminance status in the flck. Spring plumage indices were mre ften crrelated with fall dminance differences than were fall plumage cnditins, suggesting a genetic r causal relatinship between dminance and plumage rather than a prximate (status signaling) functin. Recrded mrph frequencies frm this study and the literature supprt the hypthesis that selectin is balanced between the tw mrphs and is dependent n the advantages r disadvantages f aggressiveness within a sex. Received 9 March 1983, accepted 25 July 1983. LOWTHER (1961) described tw clr mrphs f the White-thrated Sparrw (Zntrichia albicllis), a white-striped and a tan-striped frm. He shwed that bth mrphs included males and females. In field bservatins f 110 mated pairs, he bserved negative assrtative mating: white-striped males mated with tan-striped females, and tan-striped males mated with whitestriped females (Lwther and Falls 1968). Var- Present address: Bilgy Department, Saint Mary's Cllege, Ntre Dame, Indiana 46556 USA. Requests fr reprints shuld be sent t this address. 2 Present address: U.S.D.A. Frest Service, Redwd Sciences Labratry, 1700 Bayview Drive, Arcata, Califrnia 95521 USA. Present Address: Department f Preventive Medicine, Cllege f Veterinary Medicine, University f Flrida, Bx J-136, Gainesville, Flrida 32610 USA. 11 dy (1971) disagreed that clr variatin was due t tw mrphs; she fund clr types t have a cntinuus distributin. Mre recent studies have supprted Lwther's interpretatin, hwever: Thrneycrft (1966, 1975) dcumented a chrmsmal dimrphism in the species that was always related t the plumage dimrphism; Atkinsn and Ralph (1980) fund birds in breeding plumage t be bimdally distributed with regard t quantitative plumage measurements; and Rising and Shields (1980) de- scribed ther mrphlgical differences between the clr mrphs. Several f these authrs suggested that there might be behaviral differences between the clr mrphs n the wintering grunds. We examined differences in dminance be- havirs between the tw clr mrphs in winter flcks. Several attempts t find differences in dminance behavir between the mrphs The Auk 101: 110-120. January 1984

January 1984] White-thrated Sparrw Dminance during the nnbreeding seasn have been made (Harringtn 1973, Hailman 1975, Ficken et al. 1978), but nne f these investigatrs cnsidered the sex r age f the birds. We studied the interrelatinships amng sex, age, and clr mrph and fund that a cmbinatin f characters was a gd predictr f dminance rank. The purpse f this paper is t demnstrate the relatinships f these variables with scial dminance behavir f White-thrated Sparrws in captive, nnbreeding flcks. METHODS Birds were caught in mist nets thrughut Octber 1975 at Carlisle, Pennsylvania. Birds were individually marked at capture with clred leg bands and weighed. We determined age by the amunt f skull ssificatin and sex by the unflattened wing chrd at capture (Atkinsn and Ralph 1980) r by lapartmy the fllwing spring. Birds were kept in three indr aviaries (2 x 2 x 2 m) under cntrlled artificial lights, initially n a winter light schedule. The daylength was increased n 1 April t 12 hurs light and 12 hurs dark t stimulate prenuptial mlt. By 16 April the birds were cming int breeding cnditin, as indicated by mlt and singing. We investigated scial relatinships within fur grups f birds: a small grup f 16 males and females n 28 Nvember 1975; tw large grups f 51 and 54 birds in Nvember and December 1975; and a small grup f 17 females n 16 April 1976. Determinatin f clr mrph.--even thugh Whitethrated Sparrws in breeding plumage can be divided int tw distinct clr mrphs (Thrneycrft 1975, Atkinsn and Ralph 1980), cnsiderable individual variatin exists in plumage characteristics (Vardy 1971). Atkinsn and Ralph (1980) demnstrated that variatin in plumage in the fall ranged frm a bright t a dull extreme and that cmbinatins f varius plumage characteristics int an index resulted in a nrmal distributin. In additin, plumage characteristics were crrelated with sex, age, and seasn: males were generally brighter than females, adults brighter than immature birds, and all birds became brighter in the spring (Atkinsn and Ralph 1980). In this study, we calculated an index f plumage brightness similar t Atkinsn and Ralph's (1980), using fur f their five plumage characteristics: percentage black in lateral crwn stripes, thrat pattern (as classified by Lwther 1961), median crwn-stripe clr, and superciliary stripe clr. We did nt include chest streaking because f its lw variability between seasns. The fur plumage characteristics were measured at capture in the fall and again in May after prenuptial mlt. We used the median value f the indices t divide birds arbitrarily int "white" r "tan" grups: birds with index values greater than the median were called "white"; thse with less were "tan." Because such plumage indices prduced a relative cmparisn f brightness amng birds and because we did nt prepare karytypes, ur designatins f mrph may nt cincide with Thrneycrft's genetically determined "clr mrph." Atkinsn and Ralph (1980), as well as Thrneycrft (1966, 1975), hwever, fund that, althugh the tw mrphs are difficult t distinguish in autumn, they are much mre distinct in the spring. Therefre, we used spring-plumage index values t apprximate the mrph class mst similar t Thrneycrft's genetically determined "clr mrph," and, in fact, nly 9 f the 105 birds wuld have differed in their mrph classificatin between fall and spring plumages. Behaviral bservatins.--behaviral interactins were bserved thrugh a ne-way windw verlking a feeding platfrm in each f the three aviaries. During bservatin perids, fd was restricted t a small dish n the platfrm. We recrded winners and lsers f dminance encunters at the feeding dish. Winning birds supplanted, pecked, r chased the lsing individuals. Using the methds f Sabine (1959) and Brwn (1975: 86), we determined dminance hierarchies fr the tw small grups f birds by cn- structing dminance matrices f encunters. Large grups did nt frm clear linear hierarchies, precluding ranking f individual birds. Instead, we tallied dminance dyads fr each mrph-age-sex class. A dminance dyad between tw birds [i.e. A > B: the "pair relatins" f Sabine (1959)] was designated when ne bird was dminant in mre than ne-half f the encunters between the pair members, regardless f the number f interactins. The percentage f dyads in which a bird was dminant was defined as its "percentage dminance," r the number f individuals it cnsistently dminated in knwn relatins. In the tw large grups, the numbers f dminant dyads were cmbined fr members f each age- sex-mrph class t give the percentage dminance f each class. Statistics.--Because the percentage data were tested and fund t be nrmal, they were nt transfrmed. Statistical tests included a test fr differences be- tween tw percentages, prduct-mment crrelatin, and Mann-Whitney U-Tests (Skal and Rhlf 1969). RESULTS SMALL GROUP (NOVEMBER) We cnducted a preliminary investigatin f a small grup (16 birds) in Nvember. The grup exhibited an essentially linear and stable hierarchy (Fig. 1). Inspectin f physical characteristics crrelated with dminance ranks f

112 WATT, RALPH, AND ATKINSON [Auk, Vl. 101 BIRD RANK LOSER 2,l li113 21 ii i 2 I 4 i! 4! :5 :5 i I t 2 I : 2 I :5 I 4 z 9 I 2 2 12 / 4 3 I 2 3 3 g. 1. Dminance a r r 16 hre4hra ed parrws n ve ber. The h erarchy r ve ns ances(b rd ranked 8 was d nan b rd ranked 2 e c.). Tw reversals (R) ccurred where a subrdinate b rd a acked a d nan b rd. these individuals (Table 1) indicated that sex was the best single predictr f dminance. Other measures, such as mrph, age, wing length, r weight, did nt explain dminance rank simply. Amng females, hwever, tan birds were mre dminant than white birds. The relatinship amng males with respect t mrph was nt clear. The results f this prelim- inary study led us t believe that differences in dminance behavirs between white and tan mrphs might be fund in larger grups within sex and/r age classes. LARGE GROUPS Apprpriateness f encunter-frequency analysis.--we wished t investigate the rle that mrph, age, and sex played in determining dminance in the tw large grups. Measures invlving the percentage f encunters f ne class with anther class wuld be a cnvenient means f analysis. Befre we culd prceed with this analysis, hwever, it was necessary t determine hw the prprtin f birds in each f the tw clr mrphs in a class might affect any measurement f aggressin, as Hailman (1975, pers. cmm.) has suggested. That is, if birds f ne class encuntered birds f anther class at randm, then it wuld be pssible t use encunter frequencies as a measure f dminance. If encunter frequencies were nt randm, then cmparisns f mrphs summed ver the varius age-sex classes wuld nt be valid. Therefre, we calculated the number f expected encunters frm the numbers f individuals in each mrph-age-sex class in the tw large grups (Table 2) and cmpared these t bserved values (Tables 3 and 4). Nne f the eight classes, in either f the tw grups, encuntered the ther classes randmly, that is,

January 1984] White-thrated Sparrw Dminance 113 TABLE 1. Relatinships between dminance rank and physical characteristics f individual White-thrated Sparrws represented in Fig. 1. Dminance Wing length Weight Weight b rank Sex Mrph Age (mm) (g) (g) 1 M White Adult 74 26.8 30.4 2 M Tan Adult 72 23.0 26.2 3 M White Adult 72 25.0 27.0 4 M White Immature 70 24.4 23.8 5 F Tan Immature 67 23.5 22.6 6 F Tan Immature 65 23.6 24.5 7 F Tan Immature 65 22.1 20.5 8 F Tan Adult 69 25.6 28.5 9 F Tan Immature 68 23.5 21.7 10 F White Adult 65 24.0 24.8 11 M Tan Adult 72 23.8 29.1 12 F White Adult 67 23.2 25.3 13 F White Adult 70 23.8 25.1 14 F White Adult 68 22.8 23.9 15 F White Adult 70 26.6 26.3 16 F White Immature 69 22.9 24.6 Weight at capture. Weight at testing, 28 Nvember. in prprtin t class size. In general, males were bserved in aggressive encunters mre frequently, and females less frequently, than expected (rw ttals in Tables 3 and 4, bserved vs. expected). Therefre, encunters were nt ccurring randmly, and cmparisns f mrphs summed ver classes wuld nt be an apprpriate analysis. Similarly, an analysis f the number f wins (the "attacks" f Hailman 1975) wuld nt be apprpriate, because they wuld be a functin f encunter frequency. Als, an analysis f percentage wins r lsses, based n ttal encunters, wuld be inapprpriate, because a particularly aggressive bird might inflate the percentage wins fr its class. Apprpriateness f analysis f dyad frmatin.- Anther methd f cmparing dminance repercentage dminance values f white adult males with thse f tan adult males, the number f dyads they frm with ther mrph-agesex classes shuld be equal. We tested these values (Tables 5 and 6) and fund that fr all eight cmparisns f age-sex classes there was n significant difference between white and tan mrphs in the prbability f frming dyadic relatinships with the ther classes, with a single exceptin (significant at slightly greater than the P < 0.05 level). Because the assumptin f equal prbability was met, we summarized percentage dminance values fr each mrph-agesex class in each grup (Table 7) fr testing differences between the mrphs within a given age-sex class. Dminance as measured by dyad frmatin.- Striking differences in dminance between the latinships f the tw mrphs within age-sex classes invlves the use f percentage dminance, the percentage f dyads in which a given class TABLE 2. Numbers f birds in each mrph-age-sex was dminant ver anther class. The use f class fr tw large grups. percentages alleviates the prblem f nnrandm encunter frequencies, because nly ne dyadic relatinship is recrded, regardless f the number f interactins between the tw birds. Given that a bird is invlved in a set f dyadic relatinships, the percentage f thse relatinships in which it is dminant is independent f encunter frequency. The nly remaining prblem is the distributin f dyadfrmatin frequencies: in rder t cmpare Grup Grup Class 1 2 Tan adult males (TAM) 5 3 White adult males (WAM) 7 8 Tan immature males (TIM) 5 4 White immature males (WIM) 9 6 Tan adult females (TAF) 12 11 White adult females (WAF) 3 4 Tan immature females (TIF) 6 7 White immature females (WIF) 7 8

114 watt, RALPH, AND ATKINSON [Auk, Vl. 101 ;> '- ;> c ;> c,. c,.

January 1984] White-thrated Sparrw Dminance 115

116 WATT, RALPH, AND ATKINSON [Auk, Vl. 101 TABLE 7. Percentage dminance in tw grups f each mrph-age-sex class ver classes ther than its wn. Grup 1 Grup 2 n a % t b n a % t b Tan White adult males 224 162 74.6 79.6 } 1.16 ns 196 85 51.8 68.9 } 2.71'* Tan White immature males 264 151 39.1 53.0 } 2.74** 148 187 69.6 36.4 } 6.16'* Tan adult females 288 35.4 1.39 ns 4.23** White adult females 37 24.3 69 34.8 tan White immature females 190 142 38.4 35.2 } 0.60 176 43.75 } ns 192 22.9 4.29** n - number f dminance relatinships (dyads) cnsidered in the cmputatin f percentage dminance. b Statistical cmparisns were made within age-sex class between white and tan birds. ** = a < 0.01; ns = nt significant. tw mrphs, within each age-sex class, were fund in Grup 2 (Table 7). In encunters between adult males, white mrphs were mre ften dminant. In encunters between im- mature males, adult females, r immature females, hwever, tan mrphs were mre ften dminant. Grup 1 shwed less striking differences. In encunters between adult females r immature females, tan mrphs were als mre dminant than white mrphs, but nt significantly. In encunters between adult males r immature males, tan mrphs were mre dminant, a trend ppsite t that Shwn by Grup 2 birds. Only the relatinship between immature males was significant, hwever. DIFFERENCES WITHIN AGE-SEx CLASS Because we fund instances f differences in dminance between mrphs within each agesex class, we investigated pssible differences invlving mre cntinuus plumage variatin within Grups 1 and 2. Specifically, we examined whether r nt whiter birds were mre dminant within each age-sex class (e.g. in adult males) by analyzing crrelatins f percentage dminance with spring-plumage index values and with fall-plumage index values. If birds were using prximate cues t evaluate prbable dminance relatinships befre interacting, i.e. the status signaling f Rhwer (1975), fall values shuld have the highest crrelatins with dminance. If genetic differences in dminance ability were crrelated with clr mrph (apparent nly in spring), hwever, then spring values wuld be mst highly crrelated. A third and nnexclusive pssibility was that winter dminance behavirs in sme way influence the expressin f clr mrph in the spring. We culd nt distinguish between cause and effect under this last hypthesis. We did assess relatinships between fall dminance and degree f change in plumage frm fall t spring, hwever. We fund that spring values f plumage brightness crrelated mre ften with winter dminance behavirs than did fall values (Table 8). T demnstrate the nature f the crrelatins, we pltted the spring-plumage index f individual birds in each f the tw grups against their measured percentage dminance (Fig. 2). A significant psitive crrelatin was fund fr adult males in Grup 2 (whiter birds were mre dminant). Significant negative relatinships were fund fr immature males in Grup 2, adult females in Grup 1, and ttal females (immature females varied in the apprpriate directin but were nt quite significant when cnsidered alne) (Table 8). While mst birds increase in plumage-index value frm fall t spring (Atkinsn and Ralph 1980), sme birds d s mre than thers. We fund that the de- gree f plumage brightening was negatively crrelated with fall dminance behavir f adult females in Grup 2 and f all females cmbined (Table 8). Females that were mre dminant in fall had less plumage brightening in spring. FEMALE MORPH DIFFERENCES IN TWO SEASONS Thrneycrft (1975) nted that white females appeared t be mre aggressive than tan females n the breeding grunds. We fund that tan females usually dminated white females and were mre dminant t ther birds

January 1984] White-thrated Sparrw Dminance 117 TABLE 8. Crrelatins f percentage dminance measures with plumage measures taken during the fall and spring and degree f plumage brightening frm fall t spring. Crrelatins between plumage and dminance were mre ften significant in spring than in fall. Adult Samples males n Crrelatin (r) Fall plumage Spring plumage Plumage brightening Grup 1 12-0.142-0.064 0.181 Grup 2 11 0.490 0.606 0.172 Ttal 23 0.142 0.201 0.144 Immature males Grup 1 14 0.410 0.243-0.159 Grup 2 10-0.801'* 0.781'* -0.327 Ttal 24-0.193-0.254-0.201 Ttal males 47 0.067-0.018 0.170 Adult females Grup 1 15 0.531' -0.539* -0.146 Grup 2 15 0.101-0.490-0.617' Ttal 30-0.242-0.376' - 0.260 Immature females Grup 1 13 0.062 0.210-0.281 Grup 2 15-0.202-0.347-0.368 Ttal 28-0.143-0.286-0.305 Ttal females 58-0.209-0.358** -0.307* ** = significant at a < 0.01; * - < 0.05. than were white females. Because ur results cntradicted previus suggestins, we hypthesized that female dminance relatinships might change with seasn. Tan females culd be mre dminant than white females in fall but becme mre subrdinate in spring. T test this hypthesis, we assembled a small grup f 17 females frm the larger hierarchies and bserved their dminance behavir in April. Fllwing prenuptial mlt, white females were significantly mre dminant than tan females (Table 9; Mann-Whitney U-Test; U = 57, P < 0.05). Mrever, a significant psitive crrelatin was fund between individual springplumage indices and percentage dminance in April (r = 0.58, P < 0.05). Thus, it appears that behaviral relatinships between mrphs, at least fr females, may change with seasn. Because male birds were nt present in this last test, we cannt rule ut the influence males might have had n female dminance. DISCUSSION Dminance and plumage mrph.--lwther and Falls (1968) and Falls (1969) fund that white White-thrated Sparrws are mre aggressive and are mre persistent and frequent singers than are tan sparrws. Thrneycrft (1975) suggested that white females might be mre aggressive than their tan cunterparts. Other investigatrs (Harringtn 1973, Hailman 1975, Ficken et al. 1978) have suggested the same fr nnbreeding grups--white mrphs are mre aggressive than tan mrphs. Hailman (1975) reanalyzed much f Harringtn's data and cncluded that the data suggesting that white mrphs were mre aggressive than tan mrphs were incnclusive, because the relative prprtins f mrphs were nt knwn. Taking int accunt the prprtins f the mrphs present during the encunters, Ficken et al. (1978) cncluded that white mrphs are mre frequently the aggressrs than are tan mrphs. They suggested that there is a behaviral difference between the tw mrphs. Amng the interpretatins they discussed was the pssibility f a differential sex r age bias in the mrph ppulatins present. Mre imprtant, they als suggested that, like its clratin, the tan mrph's relative lack f aggressiveness represents a netenic cnditin, in that in bth behavir and mrphlgy it retains characteristics f yung birds in adulthd: "tan head-

118 WATT, RALPH, AND ATKINSON [Auk, Vl. 101 i 7 ADULT FEMALES 0 [] e 0 ß 0 0 0 25 ADULT MALES i.ij z z i IMMATURE FEMALES [] IMMATURE MALES I-- 7 Z J. J. 5 O O O ß O [] [] [] 25. ß eo ß, ß [] ß O0 ß I0 20 30 40 I 0 0 PLUMAGE BRIGHTNESS INDEX Fig. 2. Plt f percentage dyads in which individuals were dminant in Nvember against their plumagebrightness index as measured after spring mlt (see text fr explanatin). Circles (O) and squares ([ ) represent birds in Grups 1 and 2, respectively. stripe and relative unaggressiveness are characteristics f birds in their first fall f life." Our finding f relatively greater dminance in tan females in the fall is the reverse f the abve reprts fr the spring and breeding seasns, and it disagrees with the netenic hypthesis f Ficken et al. (1978). We feel that amng females, in particular, a reversal in relative dminance ccurs between seasns. It is pssible that harassment f white females by dminant males culd frce them t lwer levels in the fall hierarchy, and, subsequently in the spring, their levels might rise as males became less aggressive tward ptential mates. It is als pssible, hwever, that white female birds actually becme mre aggressive, relative t tan females, in the spring. This culd be due t hrmnal changes affecting each mrph differently. Future studies shuld be able t determine whether the presence f males r a change in physilgy is respnsible fr these changes in relative dminance levels. At this time, we d nt knw if adult males exhibit such a seasnal change in dminance; in ne grup we fund white adult males were mre dminant than tan adult males. Immature

January 1984] White-thrated Sparrw Dminance 119 males in ne grup were mre dminant when white, in the ther grup when tan. This bviusly requires further testing, especially f all-male grups in bth spring and fall. Sexual differences may exist between the clr mrphs in their relatinship t dminance rank, and further studies require assessment f sexual effects in the analysis. Crrelatins between fall dminance behav- ir and the "spring" rather than the "fall" clr mrph supprt the hypthesis (Ficken et al. 1978) that behaviral differences between the clr mrphs are genetic rather than being based n appearances alne (i.e. status signals). We cannt rule ut the pssibility, hwever, that behaviral differences in the fall affect the expressin f clr in spring plumages. Sex differences in mrph ratis.--thrneycrft (1975) fund amng yung females the rati f tan t white was 1:1, while amng males it was 43:57. By breeding seasn, the ratis were 74:26 in females and 32:68 in males. In ur study, 36% f the males were tan, and 64% were white. Amng females, 62% were tan, and 38% were white (Table 2). These values are intermediate between Thrneycrft's ratis f yung birds and thse f his adult breeding birds and prbably reflect a nrmal autumn rati, at least fr central Pennsylvania. TABLE 9. Relatinships between dminance rank, percentage dminance, and clr characteristics in a grup f 17 female White-thrated Sparrws in April. Spring- Dminance Clr plumage Percentage rank mrph index dminance i White 36.84 93 2 White 26.15 69 3 White 35.51 67 4 Tan 19.80 76 5 White 37.42 75 6 Tan 24.15 63 7 Tan 23.13 69 8 White 36.83 69 9 White 34.33 44 10 Tan 24.96 40 11 White 35.66 29 12 White 35.52 40 13 Tan 13.63 19 14 Tan 8.78 29 15 Tan 21.79 20 16 Tan 11.26 15 17 Tan 16.12 13 mrphs in the fall. [Wessel and Leigh (1941) studied fall birds, but they sexed mst f their birds by plumage differences, a methd nw knwn t be inapprpriate.] Plumage brightness within age-sex class.--we Bdy size and dminance.--rising and Shields (1980) fund that white males were slightly larger than tan males. Althugh they had small samples f females, they shwed that in 8 f 9 measurements, white females were larger than tan females. Based n bdy size alne, ne might have predicted that white birds shuld fund that, within age-sex classes, springplumage brightness was a gd indicatr f dminance in females, but nt in males in tw f fur grups. The pattern f significance fund in ne grup leads us t hypthesize that dminance in immature males might be influenced by the relatins f the adult males dminate tan birds in bth sexes. We fund in in their grup. Similarly t a speculatin menthe analysis f the small hierarchy (Table 1), hwever, that bdy size was nt a gd predictr f dminance rank, within either agesex classes r clr mrphs. Effects f seasn.--in the studies f Harringtn (1973) and Ficken et al. (1978), data were cllected frm 22 April t 4 May, when the majrity f birds are in nuptial plumage (Lwther and Falls 1968). Therefre, their resuits can best be cmpard t thse f the breedtined abve regarding female rank, adult males may persecute individuals f their wn mrph class selectively, s that in grups where white males are strngly dminant, immature white males are strngly submissive. Studies f behavirs in grups with different sex, age, and mrph ratis culd clarify these questins. With regard t Rhwer's (1975) hypthesis that plumage variability in birds is used as a status signal, we dubt that it is being used as ing studies (Lwther and Falls 1968, Falls 1969). such by White-thrated Sparrws. Plumage it- Our results fr a grup f females in nuptial plumage agree with these previus studies; self is nt a gd predictr f dminance in a mixed-sex grup f this species. Fr example, a white birds were mre dminant than tan tan female in the autumn culd nt predict the birds in late spring. relative status f a white bird withut knwl- Our study is the first t investigate and dc- edge f its sex--if it were a male, it wuld be ument dminance differences between clr dminant t her; if a female, subrdinate. Ad-

120 WATTß RALPH, AND ATKINSON [Auk, Vl. 101 ditinally, changes in signals with seasn decrease the predictive value f plumages. Plumage variatin may be imprtant in facilitating individual recgnitin, hwever (Shields 1977). Evlutin.--Linkage between clr mutatins and aggressive levels has been dcumented in mammals by Keeler et al. (1968, 1970). The interactin between plumage, sex, and hrmnal cntrl f aggressin may als be a cnsiderable influence n White-thrated Sparrws. As an evlutinary strategy fr females, it may be advantageus t be dminant t ther females in fall fraging flcks and then t becme mre subrdinate in rder t pair mre easily with aggressive males. This is the pattern seen in tan females. The same trend (mre aggressive in fall but less in the spring), hwever, wuld be a disadvantage fr tan males when cmpeting with white males fr territries in the spring. The apparent selective advantages f being tan, then, culd be balanced thrugh ppsing pressures n males and females. This hypthesis is supprted by the trend f decreasing relative frequencies f tan males frm autumn t spring and the increasing relative frequency f tan females (Thrneycrft 1975). White females wuld appear t be at a disad- vantage in bth seasns (assuming subrdinatin in winter and aggressiveness in spring are selectively disadvantageus fr females), and, in fact, white females are prprtinately the least numerus f any sex-mrph class by breeding seasn (Thrneycrft 1975). The white mrph, then, culd be maintained by its general advantage t adult males at all seasns. ACKNOWLEDGMENTS We thank J. B. Falls, J.P. Hailman, D. James, S. Rhwer, W. Shields, and R. Zink fr help during the curse f this study and fr helpful cmments n the manuscript. What frbearance we received frm the Reineman Sanctuary was appreciated. Bill Jefferies and Nel Ptter prvided helpful supprt fr ur research at Dickinsn Cllege. LITERATURE CITED ATKINSON, C. T., & C. J. RALPH. 1980. Acquisitin f plumage plymrphism in White-thrated Sparrws. Auk 97: 245-252. BROWNß J.L. 1975. The evlutin f behavir. New Yrkß W. W. Nrtn and C., Inc. FALLS, J.B. 1969. Functins f territrial sng in the White-thrated Sparrw. Pp. 207-232 in Bird vcalizatins (R. A. Hinde, Ed.). Cambridgeß Cambridge Univ. Press. FICKEN, R. W., M. $. FICKEN, & J.P. HAILMAN. 1978. Differential aggressin in genetically different mrphs f the White-thrated Sparrw (Zn- trichia albicllis). Z. Tierpsychl. 46: 43-57. HAILMANß J.P. 1975. Analysis f aggressin in Whitethrated Sparrw types f different prprtins. Bird-Banding 46: 236-240. HARRINGTON, B.A. 1973. Aggressin in winter resident and spring migrant White-thrated Sparrws in Massachusetts. Bird-Banding 44: 314-315. KEELER, C., S. RIDGEWAYß L. LIPSCOMBß & E. FROMM. 1968. The genetics f adrenal size and tameness in clrphase fxes. J. Hered. 59: 82-84. ß T. MELLINGER, E. FROMM, & L. WADE. 1970. Melaninß adrenalin and the legacy f fear. J. Hered. 61: 81-88. LOWTHER, J. K. 1961. Plymrphism in the Whitethrated Sparrw, Zntrichia albicllis (Gmelin). Can. J. Zl. 39: 281-292. ß & J. B. FALLS. 1968. White-thrated Sparrw. Pp. 1364-1392 in Life histries f Nrth American cardinalsß grsbeaks, buntingsß twhees, finchesß sparrws and allies (A. C. Bent and cllabratrs). U.S. Natl. Mus. Bull. 237. RISING, J. D., & G. F. SHIELDS. 1980. Chrmsmal and mrphlgical crrelates in tw New Wrld sparrws (Emberizidae). Evlutin 34: 654-662. ROHWER, S.A. 1975. The scial significance f avian winter plumage variability. Evlutin 29: 593-610. SABINE, W. S. 1959. The winter sciety f the Oregn Junc: intlerance, dminance, and the pecking rder. Cndr 61: 110-135. SHIELDS, W. M. 1977. The scial significance f avian winter plumage variability: a cmment. Evlutin 31: 905-907. SOKAL, R. R., & F. J. ROHLF. 1969. Bimetry. San Francisc, W. H. Freeman and C. THORNEYCROFT, H. B. 1966. Chrmsmal plymrphism in the White-thrated Sparrw, Zntrichia albicllis (Gmelin). Science 154: 1571-1572. 1975. A cytgenetic study f the Whitethrated Sparrw, Zntrichialbicllis (Gmelin). Evlutin 29: 611-621. VARD¾, L. E. 1971. Clr variatin in the crwn f the White-thrated Sparrw, Zntrichialbicllis. Cndr 73: 401-414. WESSEL, J.P., & W. H. LEIGH. 1941. Studies f the flck rganizatin f the White-thrated Sparrw. Wilsn Bull. 53: 222-230.