Why Are Nesting Marsh Wrens and Yellow-Headed Blackbirds Spatially Segregated?

WHY ARE NESTING MARSH WRENS AND YELLOW-HEADED BLACKBIRDS SPATIALLY SEGREGATED?

MARTY L. LEONARDAND JAROSLAVPICMAN Departmentof Biology,University of Ottawa,Ottawa, Ontario KIN 6N5,Canada

ABSTR^CT.--Theactivity centers of MarshWrens (Cistothorus palustris) and Yellow-headed Blackbirds(Xanthocephalus xanthocephalus) are spatiallysegregated. This segregationmay oc- curbecause (1) one speciesexcludes the otheror (2) the two speciesprefer different habitats. Thesehypotheses were testedby documentingchanges in the sizeand locationof Marsh Wren territoriesthroughout the breedingseason, and by the removalof conestingYellow- headedBlackbirds. The expansionof Marsh Wren territoriesinto blackbirdbreeding areas after boththe naturaldeparture and the removalof blackbirdsbest supports the explanation basedon the activeexclusion of Marsh Wrensby Yellow-headedBlackbirds. Received 8 Feb- ruary1985, accepted 21 August1985.

ECONOMICdefensibility of resources(Brown Red-winged Blackbirds(Nero 1956, Burt 1970, 1964)necessary for reproductionshould favor Runyan 1979, Picman 1980). Yellow-headed the evolution of breeding territories in birds. Blackbirds also nest in marshes, and limited Resourcesare defended most often against con- evidencesuggests that yellow-headsmay force specificsbecause these individuals are gener- compressionof Marsh Wren territories (Orians ally more similar in their requirementsthan and Willson 1964, Verrter 1975). heterospecifics.However, territories also may Spatial segregation between Marsh Wrens be defended against individuals of closely re- and Yellow-headed Blackbirds can be ex- lated speciesand morphologically or ecologi- plained by the following hypotheses.(1) One callysimilar species (e.g. Cody 1969,Reed 1982). speciesexcludes the other from mutually pre- Of particular interest are incidencesof inter- ferred habitat, either directly through behav- specificaggression and territoriality between ioral interactions (interference competition) or markedlydissimilar species (Willson 1967;Ver- indirectly through resourcedepletion (exploi- ner 1975; Picman 1980, 1982). tation competition) (Ricklefs 1979). (2) Each Marsh Wrens (Cistothoruspalustris) puncture speciesprefers a different habitat. These hy- eggsand kill nestlingsof other marsh-nesting pothesesmake the following predictions.First, passerines(Orians and Willson 1964,Burt 1970, if one speciesexcludes the other, then the re- Picman 1977, Bump 1983). Aggressive interac- moval of the dominant speciesshould result in tions have been noted between Marsh Wrens the movement of the subordinate speciesinto and Swamp Sparrows (Melospiza georg•ana; that area. This expansionshould not occur in Willson 1967), Song Sparrows (M. melodia; areasinhabited only by the subordinatespecies. Willson 1967),Red-winged Blackbirds (Agelaius If, however,spatial segregation results from dif- phoeniceus;Orians and Willson 1964, Willson ferencesin habitatpreference, then both species 1967, Picman 1983), and Yellow-headed Black- should remain in the samearea throughout the birds (Xanthocephalusxanthocephalus; Orians and breeding season,regardless of the presenceor Willson 1964, Verner 1975, Bump 1983). Pre- absenceof the other species,and barring any vious evidencesuggests that egg pecking has major seasonalchanges in the habitat. Alter- evolvedprimarily asa meansof reducingcom- natively, if one speciesactively excludesthe petition and not of obtaining food (Picman other, then aggressiveinteractions should be 1984). Marsh Wren nest locationsand feeding observed between them. The absence of ag- areasare often spatiallysegregated from other gressiveinteractions would supportthe expla- co-occurringpasserines (Willson 1967, Verner nation based on indirect exclusion. 1975, Picman 1980). The purposeof our study was to determine The most intensive studies have concentrat- which hypothesisbest explains spatial segre- ed on interactions between Marsh Wrens and gation between Marsh Wrens and Yellow-

135 The Auk 103: 135-140. January 1986 136 LEONARDAND PICMAN [Auk, Vol. 103

TABLE1. Dates of sampling periods for 1983 and 1984 and the maximum number of Yellow-headed Blackbird nests active during those intervals at Site 2.

No. (%) of active yellow- Year Period Date head nests 1983 1 22 May to 5 June 197(87%) 2 5-19 June 131 (58%) 3 19 June to 3 July 19 (8.4%) 4 3-15 July 5 (2.2%) 1984 1 24 May to 7 June 194 (88%) 2 7-21 June 180 (82%) 3 21 Juneto 5 July 14 (6.4%) 4 5-17 July 2 (0.9%) headed Blackbirds. Two types of data were collected to test the above predictions:(1) descriptive data on interactions between the speciesand on seasonalchanges of wren territories in one area with and another area with- out yellow-heads;and (2) experimentaldata on the responsesof Marsh Wrens to the removal of nesting yellow-heads.

METHODS

Studysites.--The descriptive study was conducted from 22 May to 16 July 1983and 24 May to 17 July 1984 in two marsh sites in Delta, Manitoba, Canada. Site 1 wasa relativelydry, homogeneouscattail (Ty- phaspp.) marsh, approximately 6 ha in area,that sup- Fig. 1. Map of Marsh Wren territories at Site 2 for ported 13 Marsh Wren territories in 1984. In addi- Period 2 when yellow-headswere present (dotted tion, 2 pairs of Common Yellowthroats (Geothlypis lines)and Period3 when yellow-headswere absent trichas)and 3 pairs of Red-wingedBlackbirds nested (solidlines), 1984.Shaded areas represent water. The along the drier edgesof this site. Site 2 was 4 ha in boundary of the yellow-head colony is represented areaand consistedof phragmites(Phragmites australis) by a dashedline. Resultswere similarfor 1983. and cattailsalong the edges,and bulrush(Scirpus sp.) and cattail patchessurrounding two central ponds. In 1983 this marsh contained 67 Yellow-headed Territory movements.--Maleyellow-heads in the Blackbird and 6 Marsh Wren territories; in 1984, there Delta area arrive approximatelythree weeks before were 63 yellow-headand 10 Marsh Wren territories. male Marsh Wrens (pets. obs.). Yellow-head males No other passetinesnested in this area. The vegeta- were capturedand markedindividually with colored tion at both studysites became denser as the season leg bands. Territory boundarieswere mapped by progressed(i.e. with the growth of new vegetation). plotting songperches and flight pathsof individual However, water depth and the compositionand dis- maleson a map of the studysite. Each male was ob- tribution of vegetation within each site remained servedfor a 30-min interval twice a week throughout constant.Both siteswere divided into a grid of 20 x the breedingseason. 20-m squares,marked by 2-m-high wooden stakes. Twenty-six of 29 male Marsh Wrens were trapped Water depth was measuredat each stakeusing a 100- and marked. Territory boundarieswere mapped on cm-high stick divided into 5-cm intervals. The ex- a weekly basisthroughout the study period by ob- perimentalstudy was conductedfrom 31 May to 11 serving the movementsof individual males during June 1984 at a third site in Delta. This site was ap- 1-h intervals in 1983 and 30-min intervals in 1984. proximately 1.5 ha in size and composedof a rela- Songperches, courtship nests (Verner 1963),and flight tively homogeneouscattail stand surrounding a small pathswere plotted on mapsof the studysites for each pond.The areacontained 21 yellow-headand 4 Marsh interval.Territory areas were calculatedby the min- Wren territories. imum convex polygon method for four 2-week in- January1986] SpatialSegregation Between Wrens and Blackbirds 137

T^BI•E2. Mean water depth (cm, + SD) measured in Marsh Wren and Yellow-headed Blackbird territories at Site 2 in 1983 and 1984 and mean water depth in Marsh Wren territories at Site I in 1984.

Site 2 Site 1

1983 1984 1984

Marsh Wren 23.0 + 10.9 17.0 + 11.6 8.98 + 5.42 n = 28 n = 26 n = 76 Yellow-head 39.2 + 8.9 33.9 + 9.7 n = 35 n = 60 Statistical t = 6.31 t = 7.63 comparison P < 0.001 P < 0.001 tervals. These intervals represent two periods with in 1984). Marsh Wren and yellow-head terri- respectto yellow-heads(Table 1): when yellow-heads tories did not overlap except for one case in were presentin the study area (PeriodsI and 2) and 1983 and one in 1984. Yellow-heads at Site 2 when yellow-headswere essentiallyabsent (Periods defended territories surrounding the central 3 and 4). Although there were no yellow-headsat ponds, whereas Marsh Wrens establishedter- Site 1, territory sizes for the same intervals were recorded as controls. ritories along the edges of the marsh (Fig. 1). Behavioral interactions.--Behavioral interactions be- The water was significantlyhigher in yellow- head territories than in Marsh Wren territories tween Marsh Wrensand yellow-headswere recorded during time-budgetstudies. In 1983 a maximum of 4 at Site 2 in both 1983 and 1984 (Table 2). Water territorial male wrens/day were observedfor I h each. depth measuredat Site 1 in 1984 was signifi- In 19848 males/daywere observedfor 30 min each. cantly lower than water depth measuredin the The observationperiod for both yearswas between Marsh Wren territories at Site 2 in 1984 (Table 0530 and 0930, correspondingto the birds' mostac- 2; t = -3.59, P < 0.001). tive period (pets.obs.). The duration,type, and num- Territorymovements.--There were no signifi- ber of separateinteractions were recorded for each cant differencesamong the four periods (ran- period. Any of the following activitieswas consid- domized block ANOVA; F = 1.23; df = 3, 26; ered to be an interaction:(1) a chaseinvolving in- dividualsof the two species,(2) a yellow-headflying P > 0.10) in the territory sizesof Marsh Wrens directly to a wren perchedhigh in the vegetation, nesting at Site 1 (Fig. 2). There were, however, and (3) a yellow-headperched directly abovea wren significant differencesin territory sizes across that had been chaseddown into the vegetation. the season at Site 2 [randomized block ANO- Removalexperiment.--A project conducted by A. Is- VA; F = 13.2;df = 3, 15; P < 0.001for 1983 (Fig. abelleinvolving the removalof yellow-headsgave 3); F = 9.03; df = 3, 27; P < 0.001 for 1984 (Fig. us an opportunity to directly test our predictions. 4)]. Marsh Wren territories increased in size Individual yellow-head territoriesin the removal area following the departureof yellow-heads(Figs. were located as described above, and Marsh Wren 3 and 4). This increasein territory size was di- territory boundarieswere determinedby flushing rectedinto areasformerly occupiedby yellow- (Wiens1969). Thirteen of 21 male yellow-headsand all activeyellow-head nests were removedon 31 May heads (Fig. 1), and toward the deeper portions and 1 June 1984. Further abandonment in the follow- of the marsh. ing two daysleft a total of 3 activeyellow-head ter- Behavioralinteractions.--During 72 sampling ritories at this third site. Marsh Wren territories were periods (49.2 h) in 1983, we observed 31 inter- mapped I day before and 10 days after the yellow- actions,totaling 18 min, between Yellow-head- head removal. ed Blackbirds and Marsh Wrens. This included wrens with territories that overlapped with yellow-head territories (n = 2). In 1984 there RESULTS were 32 interactions, totaling 17.2 min, during 82 sampling periods (38.8 h). These interac- Most Yellow-headed Blackbirds in Delta tions involved the arrival of either male (in Marsh established territories before the first 36/63 caseswhere sex was noted) or female (in malewrens arrived. They finished breeding and 23/63 caseswhere sexwas noted) yellow-heads left the marsh at least two months before the at a Marsh Wren singing perch, followed by a Marsh Wrens (only 2.2% of all yellow-head dive by the wren into the vegetation. The yel- nests were active in Period 4 in 1983 and 0.9% low-heads chased individual wrens into the 138 LEONARDAND PICMAN [Auk, Vol. 103

Site I, 1984 N-6 •2 14

•. x 24 fSite 2,198:5N- 6 g 2o

'E [2 N-6

0 I 2 • 4 0 Period 2 S 4 Fi•. 2. Mea• te•ito• sizes (•SD) •o• Marsh Period W•e•s at Site ] du•i• •e•iods [-&, [98&. Fig, 3, Mean territoT sizes (_+SD)for Marsh Wrens at Site 2 during Periods1-4, 1983.Significant vegetationin 10 of 31 altercationsin 1983 and differencesbetween means, as determined by a Stu- 10 of 32 in 1984. In the remainder of the inter- dent-Newman-KeulsMultiple Range Test (P = 0.05), actions the yellow-heads remained perched are representedby different letters (i.e. bar "b" is above the wrens. Yellow-headsnever ignored significantlydifferent from bar "c"). wrens that were present either in or near their territories. DISCUSSION Removalexperiment.--Marsh Wren and yellow-head territories were mutually exclusiveat In both the descriptiveand the experimental the removal-experiment site. Yellow-heads studies, increasesin Marsh Wren territory size formed territories around the pond, while coincided with the departure or removal of Marsh Wrens were located near the edgesof Yellow-headed Blackbirds. Wrens expanded the marsh. their territories into areas formerly occupied In the 10-day period following the removal by yellow-heads.Territory size did not change of yellow-head males,all Marsh Wren territo- significantlyin a controlarea that did not con- ries (n = 4) increased in size. The mean size of tain yellow-heads.Yellow-heads also actively these territories more than doubled (mean ter- chased and harassed Marsh Wrens whenever ritory size + SD beforeyellow-head removal = they were encountered. Our results best sup- 231.1 + 73.3 m; after = 585.8 + 259.9 m). Pre- port the hypothesisthat the spatialsegregation sumablybecause of the small samplesize, this of Marsh Wrensand yellow-headsis a resultof differencewas not significant(paired t-test, t = direct interactions between these passetines. 1.31, P > 0.05). As at Site 2, territories expand- Verner (1975) reported that yellow-heads ed in the direction of the formerly occupied both chasewrens and destroy wren eggs and yellow-head territories. In addition, after the nests.The almosttenfold sizeadvantage of the yellow-headremoval 2 wrens moved into this yellow-headsmay facilitate these behaviors. area and formed territories in areaspreviously Furthermore, the earlier arrival of the yellow- inhabited by yellow-heads.There was no sig- headsshould help theat exclude Marsh Wrens nificant changein territory size nor were there because(1) the yellow-heads should be more any newly arrived wrens at Sites 1 and 2 dur- familiar with the area they are defending, and ing this time. (2) interactionswith conspecificsshould be re- January1986] SpatialSegregation Between Wrens and Blackbirds 139 ducedbecause male yellow-headsalready have 28 - Site 2,1984 acquired territories and females. Previous studies (Orians and Willson 1964, Catchpole1978, Cody 1978)indicate that species N=10 may be interspecificallyterritorial becausethey x competefor limiting resources.Marsh Wrens and yellow-headspotentially could compete for food or suitable nest sites (Verner 1975). Yel- • 16 low-heads at Delta Marsh forage away from N their territories (Andre Isabelle pets. comm.), and thus competition for food seemsunlikely. N-I0 However, Marsh Wrens forage on their terri- o toriesand, although they may not competedi- *F 8 rectly with yellow-headsfor food, they are ex- cluded from high-quality feeding sites along the water's edge (see Orians 1980). Nest sites protected from predators [i.e. in 0 the deeper,central portionsof the marsh(Horn I 4 1968, Hoogland and Sherman 1976, Richter 1984)] may be limited. Marsh Wrens at Site 1, Period the drier site, sufferedhigher predation rates Fig. 4. Mean territory sizes (___SD)for Marsh than thoseat Site 2 [50%(14/28) were depre- Wrens at Site 2 during Periods 1-4, 1984. Significant dated at Site 1 and 19%(7/37) at Site 2]. Yellow- differencesbetween means,as determined by a Stu- headswith nestsin the central part of the col- dent-Newman-KeulsMultiple Range Test (P = 0.05), ony had fewer nest lossesdue to predatorsand are representedby different letters (i.e. bar "a" is higher overall nesting successthan individuals significantlydifferent from bar "b"). that nestedin the drier peripheral areas(Andre Isabellepets. comm.).Therefore, yellow-heads may increase their fitness by defending nest low-head eggs and young, then the segrega- sitesin the deeper parts of the marsh. tion may alsobe the result of exploitationcom- Intense interference competition between petition (Ricklefs 1979). That is, yellow-heads Marsh Wrens and Yellow-headed Blackbirds is may be more effectivecompetitors, depleting suggestedby the nest-destroyingbehavior of resourcesso that they are unavailable to wrens. Marsh Wrens.At Site 2, 50.4%of all activeyel- In this case,interference competition and ex- low-head nests were depredated in 1983 and ploitation competition need not be mutually 35% in 1984. The proportion of lossesdue to exclusive.Unfortunately, our data do not allow Marsh Wrens could not be assesseddirectly. us to rule out exploitationcompetition as a vi- However, there was a significant correlation able hypothesis.However, this explanation is betweenthe numberof depredatedyellow-head weakened by the evidence for interference nests and the distance to the nearest active competition and by the fact that Marsh Wrens Marsh Wren nest (r = -0.90, P < 0.025; Andre quickly moved into yellow-head territories af- Isabellepets. comm.).Only 15%of the yellow- ter yellow-heads left or were removed, sug- headsin 1983 and 23% of the yellow-headsin gesting that resourceswere not depleted. 1984 renested. Therefore, the nest-destroying Our data suggestthat yellow-heads exclude behaviorof the wrens may drive many yellow- Marsh Wrens from high-quality feeding and heads from the marsh and increase the wrens' nesting sites. Becausethe spatial segregation accessto optimal foraginggrounds and nesting between these speciesappears to result from sites. aggressiveinteractions (i.e. interference com- If yellow-heads chase Marsh Wrens to ex- petition), we suggestthat this is a case of in- clude them from mutually preferredhabitat, as terspecific territoriality. In addition, the we have argued,then the observedsegregation compressionof Marsh Wren territories by yel- could be a result of interferencecompetition low-heads may lower wren pairing success. (Ricklefs1979). However, if yellow-headschase Thus, the behavioral interactions between these Marsh Wrens to prevent the destructionof yel~ speciesmay influence not only their spatialdis- 140 LEONARDAND PICM^N [Auk,Vol. 103 tributions, but also their reproductive strate- ORIANS,G. H. 1980ß Adaptationsof marsh-nesting gies. blackbirds. Monogr. Population Biol. No. 14. Princeton, New Jersey,Princeton Univ. Press.

ACKNOWLEDGMENTS , & M. F. WILLSON.1964. Interspecificterri- toriesof birds. Ecology45: 736-745. We thank JudyGeggie, Mary Hill, and Andre Isa- PICMAN,J. 1977. Destructionof eggsby the Long- belle for assistance in the field. We also thank Andre billed Marsh Wren (Telrnatodytespalustris). Can. Isabelle for his data on the yellow-headsand J.-C. J. Zool. 55: 1914-1920. Belies-Isles,S. Bump, A. Horn, W. Richter, K. Teath- ß 1980. Behavioral interactions between Red- er, and an anonymousreviewer for making many winged Blackbirdsand Long-billedMarsh Wrens valuable suggestions.Field accommodationswere and their role in the evolution of the redwing provided by the University of Manitoba Field Sta- polygynousmating system.Unpublished Ph.D. tion. This project was funded by The Canadian Na- dissertation, Vancouver, Univ. British Columbia. tional Sportsmen'sFund and a National Scienceand 1982. Impact of Red-wingedBlackbirds on Engineering ResearchCouncil (NSERC) Operating singing activities of Long-billed Marsh Wrens. Grant to J. Picmanand a NSERCPostgraduate Schol- Can. J. Zool. 60: 1683-1689. arship to M. L. Leonard. 1983. Aggression by Red-winged Black- birds towards Marsh Wrens. Can. J. ZooL 61: 1896-1899. LITERATURE CITED ß 1984ßExperimental study on the role of in- BROWN,J. L. 1964. The evolution of diversity in traspecificand interspecificcompetition in the arian territorial systems.Wilson Bull. 76: 160- evolution of nest-destroyingbehavior in Marsh 169. Wrens. Can. J. Zool. 62: 2353-2356. BUMP, S. R. 1983. Yellow-headed Blackbird nest de- REED,T. g. 1982. Interspecificterritoriality in the fense:aggressive responses to Marsh Wrens. Un- Chaffinch and Great Tit on islands and the main- publishedM.Sc. thesis,Logan, Utah StateUniv. land of Scotland:playback and removal experi- BURT,DE V. E. 1970. Habitat selection and species ments. Anita. Behav. 30: 171-181. interactionsof some marsh passetines.Unpub- RICHTER,W. 1984. Nestling survival and growth in lished M.Sc. thesis, Ames, Iowa State Univ. the Yellow-headed Blackbird, Xanthocephalus CATCHPOLE,C.K. 1978. Interspecificterritorialism xanthocephalus.Ecology 65: 597-608. and competitionin Acrocephaluswarblers as re- RICKLEFS,R. E. 1979. Ecology. New York, Chiron yealedby playbackexperiments in areasof sym- Press. patry and allopatry. Anim. Behav.26: 1072-1080. RUN¾^N,C. S. 1979. Comparative habitats of, and CODY, M. L. 1969. Convergent characteristicsin competition between, the Long-billed Marsh sympatricspecies: a possiblerelation to interspe- Wren and the Red-winged Blackbird at Pitt cific competitionand aggression.Condor 71: 223- Meadows,British Columbia. Unpublished M.Sc. 239ß thesis, Vancouver, Univ. British Columbia. ß 1978. Habitat selectionand interspecificter- VERNER,J. 1963. Song rates and polygamy in the ritoriality among the sylviid warblers of En- Long-billed Marsh Wren. Proc. 13th Intern. Or- gland and SwedenßEcol. Monogr. 48: 351-396. nithol. Congr.:299-307. HOOGLAND,J. L., • P. W. SHERMAN. 1976. Advan- ß 1975. Interspecificaggression between Yel- tages and disadvantagesof Bank Swallow (Ri- low-headed Blackbirdsand Long-billed Marsh pariariparia) coloniality. Ecol.Monogr. 46: 33-58. Wrens. Condor 77: 328-331. Ho•'•, H. S. 1968. The adaptive significanceof co- WIENS,J.A. 1969. An approach to the study of eco- lonial nesting in the Brewer'sBlackbird (Eupha- logical relationshipsamong grassland birds. Or- guscyanocephalus). Ecology 49: 682-686. nithol. Monogr. 8: 1-93. NERO,R. W. 1956. A behavior study of the Red- WILLSON,M.F. 1967. Notes on the interspecificbe- winged Blackbird.II. Territoriality. Wilson Bull. havioral relationshipsof marsh-nestingpasser- 68: 129-150. ines. Auk 84: 118-120.