1010 SHORT COMMUNICATIONS

LITERATURE CITED GILMER,D. S., AND R. E. STEWART. 1983. Ferrugi- nous Hawk populations and habitat use in North BECHARD,M. J., R. L. KNIGHT, D. G. SMITH, AND R. Dakota. J. Wildl. Manage. 47: 146-157. E. FITZNER. 1990. Nest sites of sympatric hawks JANES,S. W. 1984. Influences of territory composi- (Buteo spp.) in Washington. J. Field Omith. 61: tion and interspecific competition on Red-tailed 159-170. Hawk reproductive success.Ecology 65:862-870. CLARK, P. J., AND F. C. EVANS. 1954. Distance to PIANKA, E. R. 1973. The structure of lizard com- nearest neighbor as a measure of spatial relation- munities. Annu. Rev. Ecol. Syst. 4153-74. ships in populations. Ecology 35:445-453. POOLE,K. G., AND R. G. BROMLEY. 1988. Interre- CODY, M. L. 1974. Competition and the structureof lationshipswithin a raptor guild in the central Ca- bird communities. Princeton Univ. Press, Prince- nadian Arctic. Can. J. Zool. 66~2275-2282. ton, NJ. SCHMUTZ,J. K., S. M. SCHMUTZ,AND D. A. BOAG. COTTRELL,M. J. 1981. Resourcepartitioning and re- 1980. Coexistenceof three speciesof hawks (Bu- productive successofthree speciesofhawks (Buteo teo spp.) in the prairie-parkland ecotone. Can. J. spp.)in an Oregonprairie. M.S.thesis,Oregon State Zool. 50:1075-1089. Univ., Corvallis, OR. SCHOENER,T. W. 1974, Resourcepartitioning in eco- ENSIGN,J. T. 1983. Nest site selection, productivity, logical communities. Science 185:27-39. and food habits of Ferruginous Hawks in south- easternMontana. M.S.the&, Montana StateUniv., Bozeman, MT.

TheCondor93:1010-1013 0 The CooperOrnithologic.4 Society 199 1

AGGRESSION AND DOMINANCE IN HOUSE FINCHES ’

JAMES R. BELTHOFF AND SIDNEY A. GAUTHREAUX, JR. Department of Biological Sciences,Clemson University,Clemson, SC 29634

Key words: Aggression;reversed sexual dominance; size, mass, subcutaneousfat reserves or other mor- House Finch; Carpodacus mexicanus; eastern popu- phological or physiologicalparameters). In this study, lation. our objective was to examine phenotypic correlatesof dominance among eastern House Finches. In contrast Among wintering birds, males generally dominate fe- to previous stud& of eastern populations, we used males and adults dominate immatures (Gauthreaux wild-caught. individuallv-marked House Finches that 1978). Reversed sexual dominance, when females we maintained in captivity. This approach allowed us dominate males during the nonbreeding season,is un- to measure morphological and physiological parame- common. Notable examples include speciesof raptors ters and control for group size, sex ratio, and hunger. where females are larger than males (Amadon 1975) It also allowed a comparison of field and laboratory and shorebirdsthat have undergonesexual role rever- results. sals,e.g., Spotted Sandpipers(Actitis macularia; Max- MATERIALS AND METHODS son and Oring 1980). Among passerines,reversed sex- ual dominance is rare and occurs mainly among the We capturedHouse Finchesthroughout Pickens Coun- Carpodacusfinches (Thompson 1960a, Samson 1977). ty, South Carolina during earlv Januarv 1989 with mist Females dominate males during winter in native, nets. Birds were housedin individuai cages(visually western populations of House Finches (Carpodacus isolated from each other) with unlimited accessto sun- mexicanus;Thompson 1960a,Kalinoski 1975).Among flower seeds,millet, and water until dominance trials recently introduced easternpopulations, studiesof un- began. Light on/off cycles simulated the natural pho- marked birds at sunflower feederssuggest that females toperiod. We banded finches with uniquely numbered also dominate males (Brown and Brown 1988, Shedd aluminum leg bands and color bands, and measured 1990). but these studies offer little insight into factors (to the nearest 0.5 mm) relaxed wing chord length, otheiihan sex (and associatedplumage)That are known tarsuslength, bill length (exposedculmen), medial rec- to influence dominance behavior in birds (e.g., body trix length,abdominal fat class(Helms and Drury 1960), and body mass (to the nearest0.1 g). We scored males for plumagebrightness on the forehead,eyebrow, crown and nape, malar, throat, breast, abdomen, back, wrist I Received 18 February, 1991. Final acceptance2 1 (lessersecondary and marginalcoverts), and rump. Body May 1991. regionsreceived a maximum of 3 points in each region SHORT COMMUNICATIONS 1011

TABLE 1. Comparison of mean (*SE) morphological and physiologicalcharacteristics between captive male and female House Finches.

Males Females Variable (n = 18) (n = 18) t-statistic P

Mass (g)’ 19.7 f 0.33 20.0 + 0.30 0.706 >0.484 Wing length (mm) 78.2 * 0.46 76.3 t- 0.38 3.187 <0.004 Tarsus length (mm) 20.1 + 0.19 19.9 * 0.19 0.517 >0.608 Bill length (mm) 10.2 * 0.09 10.3 * 0.08 0.902 >0.373 Tail length (mm) 59.1 rt 0.45 57.7 + 0.50 2.093 co.044 Net changein mass2(g) -0.4 + 0.15 -0.7 * 0.11 1.550 >0.130 Captivity mass change(g) - 1.8 & 0.30 - 1.9 + 0.27 0.083 >0.934

’ Mean of threepretrial masses. 2Mean of differencesbetween three pretrial and post-trial masses

as follows:O-no pigmentation; 1-only small amounts beak extended.Beak fencingand combat resultedwhen of color, predominantly orange,gold or yellow; 2-red opponents failed to yield to attackersfollowing either or pink color with only a few yellow or gold feathers; of the above displays. Combatants pecked vigorously 3 -red feathers only. Scores for the 10 regions were at each other’s beaks while rapidly calling chit-chit- summed, and males receiving high and low scoreswere chit. In each of the above interactions, we called the considered bright and dull, respectively. Because all individual that moved away the loser. We constructed captive finches possessedfully pneumatized skulls,we win/loss matrices by minimizing the number of re- could not accuratelyage individuals. versals beneath the diagonal (Brown 1975) and as- We determined group size (six) based on observa- signed dominance ranks to birds within groups. tions of wild flocks and on space constraints in the We compared morphological and physiological arena. We randomly combined three male and three charactersof males and females usingStudent ’s t-tests. female House Finches until six groupswere assembled Fat classes,which were not normally distributed, were (n = 36). We withheld food from individuals the eve- compared using Mann-Whitney U-tests. For analyses ning before trials in an attempt to simulate naturally of mass, we used an averageof the three pretrial mea- occurring conditions of reduced food during which surements. We estimated effects of captivity by sub- finches would have to compete. tracting mean pretrial mass from mass at time of cap- In January 1989 we conducted dominance trials in ture. We used analysis of covariance (ANCOVA) to a hexagonalarena constructedof woven wire ( 130 cm examine effectsof sex, wing length, and mass on dom- in diameter and 31 cm high) complete with a perch, inance rank with group as a blocking factor (Sokal and water jar, and food jar containing sunflower seeds.Al- Rohlf 1981). We also conducted separate one-way though spacein the arena was limited, density of birds analysesof variance (ANOVA) to determine effectsof was no greater than that observed on outdoor feeders dominance rank on weight changeand number of seeds during winter (pers. observ.). We modified the food jar obtained. To examine the potential effect of plastic leg so that only one finch could feed at a time. The arena band color on dominance status(see Burley 198l), we floor was constructedof hardwarecloth so that dropped performed a Kruskal-Wallis single factor -analysis of seedswould passthrough it. Thus, feedingwas possible variance (Sokal and Rohlf 198 1). All statistical tests only at the feeder. Prior to trials, flocks occupied the were no&directional (two-tailed) and rejection levels arena for up to 30 min. Trials consisted of a 5-min were set at (Y= 0.05. Means and their standard errors pretrial period, a 1-hr observation period, and a 5-min are given as K * SE throughout the text. post-trial period. From a blind located 5 m away, we counted the number of aggressiveinteractions for each RESULTS group member for two randomly selected 5-min pe- Males had significantly longer wings and tails than fe- riods using 10 x 40 binoculars. We observed each males, but mean bill length, tarsuslength, body mass, group on alternate days for a total of three trials (180 changein massduring trials, and changein massduring min) and returned birds to their individual cagesbe- captivity did not differ between the sexes (Table 1). tween trials. We weighed birds before and after trials Similarly, males and females did not differ significantly to determine changesin mass. in mean-fat index (2.3 + 0.45 and 2.4 +- 0.09, respec- Wins were scored for supplanting attacks, high in- tivelv: Mann-Whitnev U-test: U = 121.5. P > 0.20). tensity head-forward displays, and beak fencing or ac- Ofthe 2,629 interactions we observed,‘875 (33.3%) tual combat at or away from the food jar (see Thomp- were between females, and 379 (14.4%) were between son 1960b for complete description of displays). males. Of 1,375 (52.3%) intersexual interactions, fe- Supplanting attacks occurred when one bird flew to- males won 1,163 (85%) and males 212 (15%). Rela- wards another, causingthe bird approached to move tionships among group members were transitive, with away. We called the individual that moved away the few reversals. Female finches occupied the top three loser. High intensity head-forward displays,usually ac- positions in five of six groups. In the remaining group companied by “chit” calls, were the most common a male dominated all three females; these females in form of aggression.Winners lunged at or motioned turn dominated the remaining two males. Although toward opponents and displaced them with head and not independent, females (n = 18) won 113 k 30.0 1012 SHORT COMMUNICATIONS interactions out of 174 * 26.3 interactions per female and they may or may not defend the immediate area (65%). In contrast, the average male won 33 +- 18.4 surroundingthe nest site (Thompson 1960b, Samson interactionsout of 122 + 77.4 interactions(27%). Most 1976, Bjijrklund 1990). Instead, females choose the interactions were head-forward displays, beak fencing nest site and often defend it without assistancefrom and combat, but 18 interactions (0.7% of bouts) es- their mates (Thompson 1960b, Samson 1976, Bjork- calated to biting, wing pulling, and feather pulling. In lund 1990). Bjorklund (1990) concluded that female each instance, a female performed the behavior (17 reproductive successwas largely determined by choice times towards other females). of nest site (which was unrelated to male behavior) Sex had a significanteffect on dominance rank (F = among Common Rosefinches(Curpoducus erythn ’nus) 42.92, P < O.OOl),while neither wing chord length nor in Sweden, and the probability of nesting successwas mass significantly affected rank (F = 0.01, P > 0.90 independent of male physical characteristicsor behav- for each). Among females there was a significanteffect ior. Thus, the reproductive successof a given Curpo- of mass (F = 10.79, P < O.Ol), with lighter females dacusfemale may be dependent on her ability to secure achieving higher rank than heavier females and males. resourcesfor nesting (e.g., high quality nest sites) and Thus sex alone, rather than larger size correlated with perhapsdefend them againstother individuals seeking sex, appeared important in determining dominance in resources.If so, selectionfor aggressivefemales is pos- House Finches. We also observed no relationship be- sible, and suchaggression may be reflectedduring win- tween dominance rank and changein massduring cap- ter. Female dominance among Curpoducusfinches may tivity (F = 0.13, P > 0.72) or change in mass during therefore be a product of selection for increased ag- trials (F = 1.03, P > 0.32). There was no effect of gressivenessin females during the breeding season. plastic leg band color on dominance rank (Kruskal- Are studiesof captive House Finches representative Wallis test; H = 2.667, P > 0.50). of behavior in the field? To examine this relationship Among the 18 males, plumage scoresaveraged 19.8 we compared our results of captive House Finches to t 1.2 (ranae: 6-26). Males were briahtest (i.e.. the those of both Brown and Brown (1988) and Shedd greatestme&‘ plumage score) on the forehead (2.4 f (1990) who examined free-living easternHouse Finch- 0. l), nape (2.4 t 0.2) and rump (2.3 * 0.2) and least es. The present study corroborates results from field colorful on the abdomen (1.8 t 0.2) and wrist (0.5 * studiesin several respects.First, all three studiescon- 0.2). Brighter males did not consistently achieve high eluded that females dominated males. Next, all indi- dominance rank. In fact, the dullestmale in four groups cated that females interacted with females approxi- became most dominant among the males. mately twice as often as males interacted with other males:In fact, reported percentagesof suchinteractions DISCUSSION were remarkably similar (30.7% vs. 16.1%, II = 1,653 On the basis of scoring wins and lossesduring inter- interactions.Brown and Brown 1988:34.1% vs. 15.7%. actions, linear dominance hierarchies (Appleby 1983) n = 612, Shedd 1990; 33.3% vs. 14.4%, n = 2,629; were developed for eachgroup of captive House Finch- this study). Finally, both Brown and Brown (1988) and es. As in studiesof free-living, unmarked easternpop- the present study found that dull males frequently ulations, females achieved higher status than males. dominated brighter males. We conclude,therefore, that What attributes of females influence dominance in studies of dominance among captive House Finches House Finches?Because size seemsunimportant (i.e., can accuratelyreflect behavior in the field and, because smaller females dominated larger males), some other investigatorscan control many variables and individ- asymmetry favoring females may exist, such as sexual ually tag and measure experimental subjects,the lab- differencesin fightingabilities or expectedbenefits from oratory approach may be as valuable as field studies winning (Maynard Smith and Parker 1976). Whatever of behavioral dominance. the asymmetry, it appearsthat dominance patterns in We thank Jon Plissner and Steve Wagner for assis- House Finches are mediated proximately through tance in the field, and Linda Eldredgeand Linda Wang greaterfemale aggressiveness.Our resultsindicated that for aiding in maintaining laboratory birds. Dale Droge, females engagedin intrasexual bouts twice as often as Jack Dumbacher, Patty Gowaty, Sue Haig, Geoff Hill, males,and femalesexhibited higherlevels ofaggression Kathrine Luhring, Stuart Reitz, Jerry Waldvogel and (suggestedby behaviors such as biting and wing pull- other members of Clemson University’s Behavioral ing). Although Brown and Brown (1988) reported that Ecology Research Group made suggestionson the male and female House Finches differed little in ag- manuscript.The comments of two reviewersimproved gressivenesswhen interacting with House Sparrows the final draft. Financial support was provided by Sig- (Passerdomesticus), studies of intraspecificaggression ma Xi, The Scientific Research Society; the Eastern in House Finches (e.g., Brown and Brown 1988, Shedd Bird BandingAssociation; and a Paul A. StewartAward 1990, this study) showed that female-female interac- from the Wilson Ornithological Society. tions substantiallyoutnumber male-male interactions. Thus, although both sexes are equally aggressiveto LITERATURE CITED members of other species, female House Finches are more aggressivethan males intraspecifically. AMADON, D. 1975. Why are female birds of prey Selection pressures favoring increased female ag- larger than males?Raptor Res. 9: l-l 1. gressionand its ultimate consequencesremain unclear APPLEBY,M. C. 1983. The probability of linearity in but may relate to their mating system (Smith 1980). hierarchies. Anim. Behav. 31:600-608. In contrast to territorial species,male House Finches BJ&KL~, M. 1990. Mate choice is not important (and other Curpoducusmales) do not defend territories, for female reproductive successin the Common SHORT COMMUNICATIONS 1013

Rosefinch (Carpodacuserythrinus). Auk 107:35- MAYNARDSMITH, J., AND G. A. PARKER. 1976. The 44. logic of asymmetric contests. Anim. Behav. 24: BROU~N,J. L. 1975. The evolution of behavior. W. 159-175. W. Norton, New York. SAMSON,F. B. 1976. Territory, breeding density, and BROWN,M. B., AND C. M. BROWN. 1988. Access to fall departure in Cassin’s Finch. Auk 93:477-497. winter food resourcesby bright- versus dull-col- SAMSON,F-B. 1977. Socialdominance in winter flocks ored House Finches. Condor 90:729-73 1. of Cassin’s Finch. Wilson Bull. 89:57-66. BURLEY, N. 1981. Sex ratio manipulation and selec- SHEDD,D. H. 1990. Aggressiveinteractions in win- tion for attractiveness.Science-21 1~721-722. tering House Finches and Purple Finches. Wilson GAUTHREAUX,S. A., JR. 1978. The ecological sig- Bull. 102:174-178. nificance of behavioral dominance, p. 17-54. In SMITH, S. M. 1980. Henpecked males: the general P.P.G. Batesonand P. H. Klopfer teds.], Perspec- pattern in monogamy?J. Field Omithol. 5 1:55- tives in etholoav. Vol. 3. Plenum Press.New York. 63. HELMS,C. W., AND-H. W. DRURY, JR. 1960. Winter SOKAL,R. R., AND F. J. ROHLF. 198 1. Biometry. 2nd and migratory weight and fat field studieson some ed. W. H. Freeman, New York. North American buntings. Bird-Banding 3 1:l-40. THOMPSON.W. L. 1960a. Aaonistic behavior in the KALINOSKI, R. 1975. Intra- and interspecific aggres- House’Finch. Part II: factors in aggressivenessand sion in House Finches and House Sparrows.Con- sociality. Condor 62:378-402. dor 771375-384. THOMPSON,W. L. 1960b. Agonistic behavior in the MAXSON, S. J., AND L. W. ORING. 1980. Breeding House Finch. Part I: annual cycle and display pat- seasontime and energybudgets of the polyandrous terns. Condor 62~245-271. Spotted Sandpiper. Behaviour 74:200-263.

The Condor 93:1013-1016 0 The Cooper Ornithological Society 1991

A RETROSPECTIVE ANALYSIS OF RED-BAND EFFECTS ON RED-WINGED BLACKBIRDS ’

PATRICKJ. WEATHERHEAD,DREW J. HOYSAK,~KAREN J. METZ AND CHRISTOPHERG. ECKERT~ Department of Biology, Carleton University,Ottawa, Ontario KlS 5B6, Canada

Key words: Red- wingedBlackbird; color bands;ter- been confounded by other factors (Holder 1990). In a ritoriality; reproduction;survival. field experiment manipulating band color in Red- winged Blackbirds (Agelaius phoeniceus),Metz and Color bands are probably the most common technique Weatherhead (1991) found that red bands (matching used by ornithologiststo identify individual birds at a the males’ epaulets)provoked increased aggression from distance. Burlev’s (ea.. 1981. 1985. 1986) studies of neighbors, which was associatedwith a higher rate of captive Zebra Finches iPoepiila guitata) showed that territory loss. Collectively these studies raise the con- colored bands influenced several important aspectsof cern that much of ornithological research could be reproduction and survival. Recent field studies have compromised by the use of colored bands. Here we shown that bands matching the color of prominent investigate this possibility using data from our own secondary sexual traits had a negative impact on re- researchon Red-winged Blackbirds. production in Red-cockaded Woodpeckers (Picoides BecauseRed-winged Blackbirds have bright red ep- borealis)(Hagan and Reed 1988) and a positive effect aulets, the size of which appearsto be associatedwith on mate attraction in Rock Ptarmigan (Lagopusmutus) aspectsof social dominance (Eckert and Weatherhead (Brodsky 1988) although the latter result may have 1987a) and parental behavior (Eckert and Weather- head 1987b), Metz and Weatherhead (199 1) chose red bands as their experimental color. Black bands were ’ Received 6 March 1991. Final acceptance3 July used on controls becausethey matched the color of the 1991. rest of the birds’ bodies. They found that the red-band- 2Present address:Department of Zoology, Univer- ed males that lost their territories had larger epaulets sity of Western Ontario, London, Ontario N6A 5B7, than thosethat retained their territories, indicating that Canada. the red bands somehow interacted with the natural 3 Presentaddress: Department of Botany, University epaulet.A further experiment usingred, blue and black of Toronto, Toronto, Ontario M5S 3B2, Canada. bands confirmed the importance of red bands (Metz