924 ShortCommunications andCommentaries [Auk,Vol. 110

The Auk 110(4):924-927, 1993

Variation in BlackBib of the EurasianSiskin (CardueIisspin us) and Its Role as a Reliable Badgeof Dominance

J. C. $ENAR,• M. CAMERINO,'J. L. COPEWE,* AND N. B. METCALFE2 •Museude Zoologia, Ap. 593,08080 Barcelona, Spain; and 2AppliedOrnithology Unit, Department of Zoology,University of Glasgow, GlasgowG12 8QQ, UnitedKingdom

Rohwer (1975, 1982) proposedthat the variation were trapped at baited feeders using platform and extentof colorpatches in the plumageof win- trapsand clap nets,and were marked with numbered tering birdscould work asbadges of socialstatus. The aluminum bands. Trapping was carried out at least major advantageof these signals would be that in- twice weekly, but the food was available continually dividualsof unequalstatus competing for limited re- so that Eurasian Siskins used the area heavily. For sourceswould not need to risk accidentalinjury or eachmale siskin we recorded,whenever possible, age wasteenergy assessing the relative fighting ability of and the maximumlength and breadthof the badge. potentialopponents (Rohwer 1982).The status-sig- Badge size was measuredby tilting the 's head nallinghypothesis has been tested in severalspecies backin line with the body. Followingthe procedure with variableplumage (reviewed by Whitfield 1987; of Molter (1987a), we determined the relationship seealso Amat 1986,Watt 1986a,b, Fugleand Rothstein between badge area, length (L, ram) and breadth (B, 1987,Motter 1987a,b, 1989,Jackson et al. 1988).These ram) from 12 museum skins. Thus, badge area (A, studieshave, however,produced contradictory re- mm 2) was calculated as: sults.For plumagedifferences to act as true badges A = 8.38 + 0.448 LB (1) of dominance,they should correlatewithin and not just between sexand age classes.This has been dem- (r = 0.83, F = 22.042,P < 0.001).Age was determined onstratedin onlya few species(Great Tit [Parusmajor], accordingto Svensson(1984) and Cooper and Burton J•irviand Bakken 1984; Yellow Warbler [Dendroica pe- (1988). We trapped and measured638 yearling and techia],Studd and Robertson1985; House Sparrow 311 adult males. [Passerdomesticus], Moller 1987a;Whitfield 1987; see Field observationson agonistic interactions be- however Ritchison 1985, P6ys•i1988, Wilson 1992). tween large- and small-badgedmale birds were con- The EurasianSiskin ( ) shows great ducted from a blind located a few meters from a bird variability in the extentof blackishplumage in the feeder. Data were collected on 11 and 12 March 1991, bib (Newton1972). This chin patch, as in HouseSpar- and we only recorded interactionsin which one of rows (Molter 1987a),is presentonly in males,which the birds clearly had a larger badge than his oppo- are dominant over females. Dominance, however, is nent. An individual was considered to have won an unaffected by age (Senar 1985). This lack of a rela- encounterif its opponent gave a submissiveposture tionship between age and dominance,and the ab- or withdrew (Senar et al. 1990a). In order to avoid senceof the badge in femalessimplifies analysis of biasin deciding which interactionsto focuson at any the significanceof the chinpatch in statussignalling. one time, we designed a long bird table (8 x 100 cm) Other featuresin the EurasianSiskin plumagealso that constrainedthe birds into forming a "linear" couldbe usedfor dominancesignalling. For instance, flock; observationsalways commencedat the right- yellow wing stripeshave been found to work as a hand end by noting all interactionsoccurring within badgeof statusin its Americancounterpart, the Pine the focal field of the binoculars (four to five birds at Siskin(Carduelis pinus; Balph and Balphunpubl. pa- a time) and, once a maximum of three interactions per); they could have an important role. However, (maximumtwo per individual) had taken place with- sincewings are displayed in only 25% of the agonistic in each set of birds, observations switched to the next displaysused by the EurasianSiskin (Senar 1990), the group of siskinsto the left. Usually, there were about chin is always visible, and most of the encounters 30 to 40 siskinsfeeding on the feeder at any one time, between unfamiliar birds in wild flocks are between and there were often more than 50 additional birds males(i.e. the badgedindividuals; Senar et al. 1990b), waiting in a tree 1 m from the feeder, with birds we restricted this analysis to the black bib. We de- continuouslyflying from one placeto another.Band- scribevariation in the size of the chin patch in male ing operationsshowed that there wasa high turnover EurasianSiskins and test (both in captivityand in the in the population, both from week to week (Senar et field) whether this variation is related to dominance al. 1992),and within a given hour, with birds arriving status. and leaving the feeding patch continuously.There- Methods.--Plumagevariability and dominancere- fore, althoughthe animalswere not individually col- lations in the field were studied in a suburban area or banded and the possibility of pseudoreplication of Barcelona(NE Spain)during the 1990-1991winter. alwaysremains, we think it is unlikely that individual October 1993] ShortCommunications and Commentaries 925

3O

ß Yeirllng (• Adult

ß Venttings [] Adulte

0 Y-19.G2S+ 0268X r= 0.71p< 0,01

0-$ 10-15 30-35 40-45 SO-SS 5 10 15 20 25 30 35 40 45 50 Bibsize) (nm•) Bib81Z8 (mn•)

Fig. 1. Frequencydistribution of male Eurasian Fig. 2. Relationshipin captive EurasianSiskins Siskinbib-size classes in relation to age (n = 638year- between bib size and dominance score. ling and 311 adult males). had a similar range of badge sizes,there were more birds contributedmore than one or two data points adults than expectedwith large badges(Fig. 1). to the analyses. We recorded39 agonisticinteractions in which both The laboratoryflock was composed of six adult and contestants were males and one of the individuals six yearling males. These were selectedso that indi- clearly had a larger badge than the other. Larger- vidualsexhibiting the full rangeof badgesizes were badgedbirds won 77% of these interactions,signifi- representedin both age classes.For each captive in- cantly more than expectedby chance(binomial test, dividualwe measured(on capture)badge size, wing two-tailed Z = 0.0014),and so were appreciablydom- length, third primary length, and mass.Birds were inant over smaller-badgedbirds. The large-badged color banded for individual identification and housed individuals also initiated most of the interactions in an outdooraviary (100 x 100 x 60 cm; for details (79%),a trait typical of dominant siskins(Senar 1985). of feeding schedule,see Senar et al. 1990a).All ob- The correlationin the captiveflock between badge servationswere madewithin 20 daysof initial capture size and social status(i.e. dominance scorefrom CA) (i.e. lessthan the time needed by siskinsto become was0.71(F= 10.0,df = 1,10,P < 0.01;Fig.2). Multiple- familiar with one another;see Senar et al. 1990a).The regressionanalysis of socialstatus on badgesize, age, experiment,therefore, was designedso that interac- mass,wing and third-primary length showedthat the tions were recordedin a non-establishedgroup (see badgesize was the only variable that explained the Watt 1986a). We recorded 1,806 contests over re- observed variation in social status when all the other sources(food, water or perches)in 11.5 h of obser- variables were held constant (Table 1). vation. In any interaction we recordedthe behavior Discussion.--Since dominance rank within male sis- usedby the actor(the initiating bird) and the response kin flocksis positivelyassociated with the sizeof the by the reactor.The behaviorsused by siskinsin ag- badge,irrespective of age(and by definition,sex), the gressive encounters are described in Senar (1990) and chin patchcan potentiallybe usedfor signallingsta- Senaret al. (1990a).The resulting data were used to tus (Whitfield 1987).It hasbeen suggestedthat such constructa matrix relating each of the 12 individuals badgesof statusshould be particularlyselected for in with the number of times that they either used or specieswith flocksthat havean unstablemembership, receivedthe differentbehavior patterns. Correspon- since they would reduce the cost of assertingstatus denceanalysis (CA; Heijden et al. 1990)based on this each time birds joined a new flock (Rohwer 1982, matrix produced a dominance scorefor each individ- Whitfield 1987). Lemel and Wallin (1993) have shown ual (see Senar et al. 1990a), which was used in sub- that badgeshave more impact on the outcomeof con- sequentanalyses relating dominanceto the morpho- flicts between unfamiliar birds rather than be- logical variables. tween birds with prior experience.This is not com- Results.--Badgesize in male siskins ranged be- pletely the casein siskins,which appear to live in tween 0.0 and 66.4 mm 2, with a mean of 20.04 ñ SE subflocksthat are stableand highly integrated(Senar of 0.43 mm2 (n = 949).Adults had on averagelarger et al. 1990a).However, given the high mobility of the badgesthan yearlings(adult œ= 23.43 + 0.76 mm2, species,these subflocks can interactwith many other n = 311; yearling œ = 18.39 ñ 0.50 mm2, n = 638; groups during the course of a winter (Senar et al. Mann-Whitney test, U = 78,449.5, P < 0.001). This 1992), so that badgesmay be used primarily in inter- was due to the fact that, althoughboth age classes actions between the members of different subflocks. 926 ShortCommunications and Commentaries [Auk, Vol. 110

TABLE1. Multiple-regressionmodel fitting data from es having high numbers of individuals at both ex- captive flock, with dominancescore as dependent tremes of the distribution. These two points suggest variable. that the greater mean badge size of adult birds may be due to dominant birds being more likely to survive Independent to adulthood, rather than to adults being dominant variable B Partial T per se. Nevertheless, further work is required before Bib size 0.71 -- 3.16' either hypothesis can be convincingly rejected. Third-primary length 0.19 0.27 0.85 Acknowledgments.--Weare grateful to Doris J. Watt, Wing length 0.19 -0.27 0.83 D. Philip Whitfield and Anders P. Mailer for helpful Mass 0.09 0.87 0.37 comments,and to L. M. Copete, J. Domenech and D. Age -0.02 0.03 0.08 Valera for their help during the fieldwork. This study **, P < 0.01; all others ns (P > 0.05). is a contribution to DGICYT project PB92-0044-C02- 02.

Sincethe FurasianSiskin chin patch is only present LITERATURE CITED in males, it could be argued that this is a sexual or- nament, more important in the breeding seasonand AMAT,J. A. 1986. A possible significanceof the oc- only used incidentally in the nonbreeding seasonas curence of chin-spots in male ducks. Gerfaut 76: a means of assessingand establishingstatus. We do 31-35. not have the data to test this hypothesis.However, COOPER,J. E. S., AND P. J. K. BURTON. 1988. An ad- the clear bimodality in the size of the black bib, char- ditional age criterion for Siskins.Ringing & Migr. acteristic of several sexual ornaments in different an- 9:93-94. imal species(Gadgil 1972),supports this view. Badges FUGLE,G. N., AND S. I. ROTHSTEIN.1987. Experiments used to signal dominance statusin agonisticencoun- on the control of deceptive signals of status in tersin other species(Great Tit [J•irviand Bakken 1984, White-crowned sparrows.Auk 104:188-197. PiSys•i1988, Wilson 1992, Lemel and Wallin 1993]; GAGDIL,M. 1972. Male dimorphism as a conse- House Sparrow [Mailer 1987a])have alsobeen shown quence of sexual selection. Am. Nat. 106:574-580. to be clearly and perhaps more significantly related HEIJDEN, P. G. M., VAN DER, H. DE VRIESAND J. A. R. to sexual selection (Great Tit [Norris 1990a, b]; House A.M. vAN HOOF. 1990. Correspondenceanal- Sparrow[Mailer 1990]).Nonetheless, our resultsshow ysis of transition matrices, with special attention that even if the character had evolved as a sexual to missing entries and asymmetry.Anim. Behav. ornament, it is now clearly related to dominance sta- 40:49-64. tus. JACKSON,W. M,, S, ROHWER,AND g. L. WINNEGRAD. Extensivefield sampling has shown that adult male 1988. Statussignaling is absentwithin age-and- siskins have, on average, larger badges than year- sexclasses of Harris' Sparrows.Auk 105:424-427. lings. This could be a consequenceof either: (1) an J•RVI, T., AND M. BAKKEN. 1984. The function of the increase in badge size after the first total molt (i.e. variation in the breaststripe of the Great Tit (Par- when birds molt into an adult ); or (2) dif- us major).Anim. Behav. 32:590-596. ferential survival favoring more dominant (and, hence, LEMEL,J., AND K. WALLIN. 1993. Status signalling, larger-badged) birds, so that the adult classcontains motivational condition and dominance: An ex- more larger-badged birds than would be expected if perimental study in the Great Tit, Parusmajor L. there were no directional selection. A direct test of Anim. Behav. 45:549-558. thesetwo possibilitieswould be to determine changes MOLLER,A.P. 1987a. Variation in badge size in male in badge size from yearling to adult age classes.Un- House Sparrows Passerdomesticus: Evidence for fortunately, the low number of recapturesfrom one statussignalling. Anim. Behav. 35:1637-1644. winter to the next in this species(Newton 1972) pre- MOLLER,A. P. 1987b. Social control of deception cludesthis analysis.Comparison of the mortality rates among status signalling House Sparrows Passer of yearling male siskinsin relation to badge size also domesticus. Behav. Ecol. Sociobiol. 20:307-311. is not feasible due to the high mobility and nomadic MOLLER, A.P. 1989. Natural and sexual selection on tendency of siskins(Senar et al. 1992). The remaining a plumage signal of statusand on morphology available data are highly contradictory.On the one in House Sparrows, Passerdomesticus. J. Evol. Biol. hand, the fact that only adults were found to have 2:125-140. bib sizesabove 50 mm 2 (Fig. 1) suggeststhat individ- MOLLER, A. P. 1990. Sexual behavior is related to uals do increasetheir bib size with age. However, this badge size in the House Sparrow Passerdomesti- explanation implies a relationship between domi- cus. Behav. Ecol. Sociobiol. 27:23-29. nance and age that is not supported by the available NEWTON, I. 1972. . Collins, London. data (note that Senar [1985] also found no correlation NORRIS,K.J. 1990a. Female choice and the evolution between age and dominance in 13 captive Furasian of the conspicuous plumage coloration of mo- Siskins). Figure 1 also shows that there is no simple nogamousmale Great Tits. Behav. Ecol. Sociobiol. relationship between age and bib size, both age class- 26:129-138. October1993] ShortCommunications and Commentaries 927

NORRIS,K.J. 1990b. Femalechoice and the quality of social stability in flocking Siskins (Carduelis of parentalcare in the Great Tit Parusmajor. Be- spinus).Ethology 85:13-24. hav. Ecol. Sociobiol. 27:275-281. SENAR,J. C., J. L. COPEWE,AND N. B. METCALFE.1990b. P•¾s.•, H. 1988. Feeding consequencesof the dom- Dominance relationshipsbetween resident and inance statusin Great Tit Parusmajor groups. Or- transientwintering Siskins.Ornis Scand.21:129- nis Fenn. 65:69-75. 132. RIT½I-•ISON,G. 1985. Plumage variability and social STUDD,g. V., AND g. J. ROBERTSON.1985. Evidence statusin captivemale House Sparrows.Kentucky for reliable badgesof statusin territorial Yellow Warbler 61:39-42. Warblers (Dendroicapetechia). Anim. Behav. 33: ROHWER,S. 1975. The social significanceof avian 1102-1113. winter plumage variability. Evolution 29:593-610. SVENSSON,L. 1984. Identification guide to European ROHWER, S. 1982. The evolution of reliable and un- .Svensson, Stockholm. reliable badgesof fighting ability. Am. Zool. 22: WATT, D. J. 1986a. A comparative study of status 531-546. signalling in sparrows( Zonotrichia). Anim. SENAR,J. C. 1985. Interactional rules in captive Sis- Behav. 34:1-15. kins (Carduelisspinus). Misc. Zool. 9:347-360. WATT,D. J. 1986b. Relationship of plumage vari- SENAR,J.C. 1990. Agonisticcommunication in social ability, size and sexto socialdominance in Harris' species:What is communicated?Behaviour 112: Sparrows.Anim. Behav. 34:16-27. 270-283. WHITFIELD,D. P. 1987. Plumage variability, status SENAR,J. C., 2P.H. K. BURTON,AND N. g. METCALFE. signalling and individual recognition in avian 1992. Variation in the nomadic tendency of a flocks. Trends Ecol. & Evol. 2:13-18. wintering Carduelisspinus and its relation- WILSON,J. D. 1992. A re-assessmentof the signifi- ship with body condition. Ornis Scand. 23:63- canceof statussignalling in populationsof wild 72. GreatTits, Parus major. Anim. Behav.43:999-1009. SENAR,J. C., M. CAMERINO,AND N. B. METCALFE. 1990a. Familiarity breeds tolerance:The development Received6 March 1992, accepted22 November1992.

The Auk 110(4):927-930, 1993

ShorebirdPredation on Eggsof HorseshoeCrabs During Spring Stopoveron Delaware Bay

GONZALOCASTRO • AND J.P. MYERS2 •Wetlandsfor the Americas,P.O. Box 1770, Manomet, Massachusetts02345, USA; and Alton JonesFoundation, 232 E. High Street,Charlottesville, Virginia 22901, USA

Most speciesof shorebirdsmigrate long distances ter and Botton 1985). As many as 420,000 shorebirds and stop at only a few selectedsites (stopovers)to (peak count of all species combined for a 160-km replenishfat depositsthat will fuel the remainderof shoreline)have been recordedin a given year (Dunne the migratoryflight (Myers 1983,Davidson 1984, Sen- et al. 1982).A thorough descriptionof Delaware Bay ner and Howe 1984,Myers et al. 1987,Piersma 1987). can be found in Clark et al. (1993). Sincegreat numbers of individualsutilize thesestop- We calculatedthe total energy consumptionby oversfor fattening,it is likely that the collectiveen- shorebirdsduring their spring stopoverat Delaware ergy consumptionby shorebirdsat these stopovers Bayby estimatingthe total energy requirements(en- mustbe large. Delaware Bay located between the states ergy expenditureand fat deposition),and correcting of Delaware and New Jerseyin the easternUnited for the assimilationefficiency of ingestedfood. Basal States (38ø47'N to 39ø20'N and 74ø50'W to 75ø30'W; metabolicrates (BMR) were calculatedusing the equa- Clark et al. 1993) is one of thesesites in spring.It tions for shorebirds of Kersten and Piersma (1987), hosts hundreds of thousands of shorebirds that feed which take into accountthe relatively high BMRs of primarily on the eggsof the horseshoecrab (Limulus shorebirds (Castro 1987, Kersten and Piersma 1987, polyphemus;Myers 1986). Birds arrive in synchrony Mathiu et al. 1989). Total energy expenditure was with the horseshoecrabs, which emergefrom the sea calculatedby assumingthat the daily energy expen- and lay their eggsin beachesduring high tide (Shus- diture is equivalent to 2.5 BMR. This is the average