Behavioral Interactions of Red-Winged Blackbirds and Common Grackles on a Common Breeding Ground

Home , Common grackle, Icterid, Quiscalus, Redwing

BEHAVIORAL INTERACTIONS OF RED-WINGED BLACKBIRDS AND COMMON GRACKLES ON A COMMON BREEDING GROUND

Jo• A. W•s

ECOLOGmTShave long been engagedin the study of relationshipsbetween species,but ethologistshave all too often ignoredinterspecies relationships,with the result that the behavioralaspects of community ecologyor competitionare poorlyknown. The few recentdetailed studies of interspecificbehavior (Simmons, 1951; Rand, 1954; Lanyon, 1957; Selanderand Giller, 1959, 1961; Moynihan, 1962, 1963; Orians and Collier, 1963; Orians and Willson, 1964) have exposednew approaches to investigationsof ecologicalrelationships (see Wynne-Edwards,1962). This study was undertakento examinethe nature and effects of behavioral interactionsbetween two icterids, the Red-wingedBlackbird (Agelaiusphoeniceus) and the CommonGrackle (Quiscalusquiscula), in a situationwhere the two specieswere breedingtogether in a small cattail marshand were as a consequencepotentially in competition. Red-wingsnest in a wide range of habitats, but usually are closely associatedwith marshesand, indeed,are often the "dominantspecies" of the Nearcftc marsh avifauna (Allen, 1914). Grackles,while being quite adaptablein their nestinghabits and utilizing a wide range of nesting substrates(see Bent, 1958), apparentlyonly sporadicallynest in marshes. Indeed, Beecher(1942), in analyzingnesting substrates of severalspecies in relation to available edge growth, concludedthat cattails were "pessi- mum" substratefor gracklesand "supra-optimum"for Red-wings.Thus in a situation in which both speciesutilize marsh flora for nestingsub- strate, any interpretationof the relationshipsbetween the speciesmust recognize that the habitat is typical for Red-wings and unusual for grackles.

T• STUDY AREA The study was conductedduring the spring and summer of 1962 and 1963 in a 2.4 acre (1 hectare) cattail (Typl•a latifolia) marsh on the east of Lake Wingra, in the University of WisconsinArboretum, Madison. This marsh is borderedby red osier dogwood(Cornus stolonifera)and black willow (Salix nigra), particularly on the side adjoining the lake, and is borderedon three sidesby fairly densestands of mixed hardwoods. Figure 1 illustratesthe generalfeatures of the study area. It is the same marsh usedby Nero (1956) and by Beer and Tibbitts (1950) in their studies of Red-wings. Grackles have nested in small numbers in this marsh since at least 1946 (J. Beer, pers. comm.). These first nests were

356 The Auk, 82: 356-374. July, 1965 July1965 ] WlEI•S,Behavior o!Red-wings andGrackles 357

LAKE WINGRA

April25, 1962 LEGEND Ope.Woter c•.oils Observution Tower Shrubbery RedwingTerritory GrockleNest Site Elevoted Song •rch

50ft.

April 12• 1963

Figure1. Gracklenest sites and Red-wingterritory configurations in the East WingraMarsh at the beginningof gracklenesting, 1962, 1963. in deaddogwoods in the westend of themarsh, but by 1948grackles werenesting in cattailsin severalplaces in the marsh. Throughmost of the two breedingseasons studied there were about 20 to 25 malegrackles and 18 to 22 femalegrackles directly associated withthe marsh; other small groups of bothsexes often frequented the marshedges, especially early in the season,but thesebirds wandered in and out of the areaand did not nestabout the marsh. In 1962there were 45 to 55 femaleRed-wings nesting in the marsh,while in 1963about 50 to 60 femaleswere present. In bothyears 21 maleRed-wings held territories in the marsh.

METttODS Observationswere made with 7 X 35 binocularsfrom the edgeof the marsh, fromthe roofof a carparked on the roadbordering the marsh,and, mostly, from a 20-footwooden tower located roughly in the centerof the marsh.During the 1963season a smallbattery-operated tape recorder was used to recordobservations, a techniqueparticularly useful for recordingthe detailsof rapidbehavioral sequences. 358 wiE•s,Behavior of Red-wings and Grackles [ Vol.Auk 82

Extensive searches through the entire marsh were made rather often to locate nests. When nests were found their position was plotted on a map of the marsh, a numbered cloth or plastic tag was affixed to nearby vegetation, and subsequent visits were made two or three times per week.

INTRASPECIFIC BEHAVIOR

DISPLAYS Both specieshave several well-defined displays which are regularly used in intraspecificsituations, and since these displaysare also seen in encountersbetween the speciesit may be helpful to describe several briefly. Similaritiesbetween the displaysof the two speciessuggest homologousrelationships and permit a parallel categorization.A detailed analysisof Red-wing displaysis presentedby Nero (1956), and Ficken (1963) has analyzed grackle displays. Song-spread.--Song-spreadin both speciesis a fairly mild threat display and servesintraspecifically as a territorial or self-advertisementdisplay. It is not uncommonlygiven by solitary birds, but it is usually delivered with greater intensity and frequencyin the presenceof other individuals of the same sex. In male Red-wingssong-spread is usually associated with the oak-a-leesong; the wings and tail are spread and the contour feathersraised during the song(Figure 2, A). Song-spreadin femaleRed- wings is similar to that of the male exceptin the accompanyingvocaliza- tion; the female gives a seriesof shrill, high, rapidly descendingtrills, something on the order of skraw-skraw-a-skree-skree-skree,with each display. In male gracklesthe song-spreaddisplay is quite similar to the above, but the vocalization is less elaborate and the raising of the head and body at the climaxof the songcontrast with the crouchingposition of the Red-wing(Figure 2, B). A similar displayis givenby femalegrackles, althoughthe visual and vocal componentsare typically lesspronounced. Bill-tilt.--This display, or displayssimilar to it, is found in many of the Icteridae and in a number of other passerines(Nero, 1963; An- drew, 1961; Marler, 1961), and it apparently functionsin maintaining a "critical distance." The display, which is given by both sexes,occurs only in the presenceof other birds, usuallymembers of the samesex. It is associatedwith territorial defensein the Red-wing (Figure 2, C), and with courtshipand (infrequently) territorial defensein the grackle (Fig- ure 2, D). Dive.--In both speciesdiving is a highly aggressiveaction, usually taking the form o.f a direct swoopingflight toward another individual. Diving is much more commonin Red-wingsthan in grackles,perhaps becauseof the smaller territory size of the latter, but in both speciesit July1965 ] WIEI•S,Behavior ofRed-wings andGrackles 359

½ D

Figure 2. Song-spreadand bill-tilt displays in Red-wing and grackle males. A, Red-wing song-spread; B, grackle song-spread; C, Red-wing bill-tilt; D, grackle bill-tilt. A and C after Nero (1956), B and D after Ficken (1963). is relatively uncommon in normal intraspecific activities. This motor pattern may have given rise to suchcourtship displays as "sexualchases" in the Red-wing (Nero, 1956), and "leader flights," "chases," and "together flights" in the grackle (Ficken, 1963), all of which are character- ized by pursuit of a female by severalmales. Tail-llicking.--This display is given by both Red-wingsand grackles of both sexesin responseto a strangesituation or a disturbance. In both speciesit consistsof repeatedupward flicks of the closedtail and is often accompaniedby sharp call notes.

Ba•m•G B•oLoG¾ Although I studied the breedingbiology of each speciesfairly inten- sively, only those aspectswhich bear on interspecificbehavior are dis- cussed here. Gracklesare not typically marshnesters, are strictly monogamous,and defendonly a small territory directly about the nest site. Red-wings,on the other hand, are characteristicmarsh nesters,are polygynous,and 360 W•E•vs,Behavior o)• Red-wings and Grackles [ Vol.Auk 82 defend clearly circumscribedareas of from .06 to 2.0 acres against other males, femalesother than their mates, and, at times, fledgling Red-wings. Adult male Red-wingsusually arrive on the East Wingra Marsh about the secondweek in March, the first arrivals usually being birds which have previouslybred on the marsh (Nero, 1956). Femalesarrive a week or two later. Adult male grackles generally arrive in the area two or three days after the male Red-wings,and are followed 5 to 10 days later by the adult females. Territories are establishedby male Red-wingsalmost immediately upon their return, but the boundariesmay undergo considerablefluctuations and may not becomefirmly establisheduntil nestingbegins five or six weekslater. Red-wing territoriesvary in position,size, and shape,with respect to the availability of elevated singing perches, the extent of vegetationaledge growth (cattail-open water interface), and population density. That territorial boundariesof Red-wings do fluctuate early in the seasonis important, becausethis affects the easewith which grackles may penetratethe marshand establishtheir own small territories. Red-wingcourtship typi.cally takes place in the territoriesafter they have been initially established. Grackle courtship, on the other hand, takes place about the edgesof the nestingarea, and territories are formed only after the nest site has been selectedby flights of the paired birds into the nesting area (the marsh). Thus grackle territories are nest- centered, with territorial aggressivenessdecreasing with increasing distance from the nest, while Red-wing territories are large, with no definite center of activity, and defenseis intense to the very edge of the clearly delineated area. Grackles began nesting activities in the marsh 5 to 15 days before Red-wings(see Figures 3 and 4). Gracklestook about a week more to construct their nests than Red-wings, and incubated the eggs a day or so longer (12-13 days). Grackleslaid an averageof 4.4 eggsper clutch in the marsh,about one o.rmore eggper clutch than Red-wings. Grackle nestlingsstayed in the nest two to five days longer than Red-wing nestlings,but left the marshwithin a day or two after fledging,while Red-wing fledglingsremained in the home territory for as long as two weeksafter leaving the nest. Both speciesare usually single-broodedin the Madison area. Nest constructionand incubationare performedentirely by the femalein both species,but male gracklesparticipate in feedingthe nestlings,while male Red-wingstake such a role only occasionally.In both speciesthe fledg- lingsare fed by both sexes,the male usuallypredominating. Ecologically,the two speciesappeared to utilize the marsh in much the same manner. Nests of both specieswere usually on the edgesof 1965July ] WIEldS,Behavior o)• Red-wings and Grackles 361

1962 I'• GRACKLE(14nests) ] REDWING(44nests)

APRIL MAY dUNE dULY DATE

Figure 3. Chronology of Red-wing and grackle nesting in the East Wingra Marsh, 1962. A, nests being laid in; B, nests being incubated; C, nests with young. cattail clumpsbordering areas of open water and very few differencesin the characterof the substrateselected were detected. Both speciescon- structed a cup-shapednest with a bulky framework of interwoven cattails, a cup of mud or wet cattails, and a cup lining of slendergrasses, rootlets,or hair. Gracklenests, however, were usually bulkier and heavier than Red-wing nests, and were normally built in somewhatdenser cattail clumps than Red-wing nests. The average distance from water to nest rim for 34 grackle nestsin the marsh was 20.0 inches (range, 9-27 inches), while 138 Red-wing nests averaged 23.7 inches up (range, 14- 33 inches). Red-wing nests were usually exposedand well-lighted, while grackle nestsoften were somewhathidden and usually shaded,perhaps as a result of their slightly lower position in denseclumps of cattails. 362 WIENS,Behavior of Red-wings and Grackles [ Vol.Auk 82

1963

rFmGRACKLE 17in A)(20 nests;

I i i i i i i i i i • I I I I I I I I I I I 65 11-15 21-25 I-5 11-15 21-25 I-5 11-15 21-25 I-5 11-15 APRIL MAY dUNE dULY DATE

Figure4. Chronologyof Red-wingand gracklenesting in the East WingraMarsh, 1963. A, nestsbeing built; B, nestsbeing laid in; C, nestsbeing incubated; D, nests with young.

Both gracklesand Red-wingsappear to obtain the bulk of their food (Lepidopteranlarvae) in the hardwoodforests bordering the marsh,but both fed to a limitedextent on emergingdamselflies and dragonfliesin the marsh.

INTERSPECIFIC BEttAVIOR Behavioralinteractions between Red-wingsand gracklesoccurred throughoutthe breedingseason, but weremore frequent during the early part of courtshipthan after the initiationof nesting(Figure 5). Most July1965 ] WRENS,Behavior ofRed-wings andGrackles 363

A. RED-WING INTRASPECIFIC RESPONSES B. RED-WING INTERSPECIFIC RESPONSES BILL- • NO SONG- BILL' TILT DIVE :•,q,RESPONSESPREAD TILT ....

--4-o' q"l'"o'" q'+ 0' "o' '

C, GRACKLE RESPONSES TO RED-WINGS RETREATgg•D,SPLAY ATTAC. o -H- 0 ++ 0 -•+

C• - MALESARRIVE ß •ESNESTINO BEGINSARRIVE ß EGGS LAID ß INC•TION BEGINS ß YOUNG HATCH ß Y•G FLEDGE ß BIRDS DEPART ß •SE ABSENT ß • ABUNDANT

•-GRACKLE __ I I II I II I rl I I 't' 0 -f'i' 0 •' 0 '1-1- 0

Figure 5. Approximate frequenciesof basic interspecificresponses with respectto Red-wing and grackle breeding cyclesin 1963. A, Red-wing responsesto territory intrusions by other Red-wings; B, Red-wing responsesto territory intrusions by grackles; C, grackle responsesto Red-wing displays. Frequenciesfor a response indicate the relative frequency through the breeding seasonof that responseonly, and do not relate to its frequencyin the total responserepertoire of the species.

interactionsbetween the two specieswere initiated by the Red-wingsand were apparentlythe resultof grackletrespasses into Red-wingterritories. Suchintrusions occurred most often in connectionwith gracklecourtship and with their effortsto reachtheir nestsor feed fledglinggrackles. Grackle courtship in the marsh involved two characteristicbehavior patterns,the "mutualperch" displays and "pair formationflights." The "mutualperch" displays were usually performed in smallmixed groups in the tall willowsand dogwoodsabout the marshedges. Male grackles showeda distincttendency to givethese displays from prominent perches in thesetrees (: songperches), but individualmales or pairs showedno attachmentto particularsong perches, and therewas no evidenceof any grackleterritoriality connected with theseperches. Many of thesetrees were alsoused as songor displayperches by the resident(i.e., territory- holding)Red-wing males, and thus were includedin Red-wingterritories 364 W•Ns,Behavior of Red-wings and Grackles [ Vol.^uk 82 and stronglydefended intraspecifically by the Red-wings. Gracklesdis- playing in thesetrees, however, almost never elicitedany responsefrom residentRed-wings (Table 1), even when the two birds were only a few feet apart. Grackle "pair formation flights" followed the typical pattern, one or more femalesleading a group of males out over the marsh. Often these flights endedwith the gracklesalighting in the cattails (and thus in Red- wing territories). Early flights into the marsh were apparently made at random with respect to Red-wing territories, but as courtship pro- gressed,and as Red-wing territories became increasinglywell defined, grackle flights became directed into areas along the edgesof Red-wing territories. Detailed records were kept for 42 grackle "pair formation flights" into the marsh which occurred between the arrival of females and the initiation of nestingactivities: of these flights 13 (31 per cent) ended in areas outside Red-wing territories (although such areas com- prisedless than 15 per cent of the usablemarsh area); 15 (36 per cent) ended fewer than 5 feet into a territory; 9 (21 per cent) ended 5 to 10 feet inside a territory; and only 5 (12 per cent) ended more than 10 feet into a territory. Deep penetrationsby grackles into Red-wing territories always elicited a response(usually a dive), while penetrations of only a few feet were met with a wider range of responses,and not uncommonlythe resident Red-wing male made no observableresponse (Table 1). By a week or 10 days after the beginningof grackle courtship the flights showed a distinct orientation to Red-wing territorial edgesor areas of poorly defined territories,and no penetrationsof more than 10 feet were observed. Grackle flights connectedwith nest site selectionwere similarly oriented, and apparently the subsequentplace- ment of most grackle nests dependedon the location of Red-wing territory boundariesas well as vegetationalcharacteristics (Figure 1). Some Red-wing territories continued to fluctuate after the establishmentof grackle nest sites, however, so a few grackle nests eventually became located well within Red-wing territories. With the initiation of nest constructionthe activities of grackles became more closelyassociated with their nest sites, and flights out into the cattails were usually made by a single female or pair. The response of Red-wing males to these grackleswas rather variable. On numerous occasionsgrackles approaching their nestselicited no observableresponse from the residentRed-wings, and the frequencyof such seeming"indif- ference" increasedwith the beginning of Red-wing nesting activity and the concurrentincrease in Red-wing intraspecificterritorial activity (Fig- ure 5). In some casesthe behavior of the resident Red-wingsactually suggested"avoidance" of the grackles;the Red-wingmale droppeddown July1965 ] WIENS,Behavior ofRed-wings andGrackles 365

TABLE 1 RED-WINO DISPLAY RESPONSES TO GRACKLES DURING GRACKLE COURTSHIP AND NESTING, 1962, 1963

Instancesof display responseof Red-wing male Targetor situation Bill- Song- Tail- No Dive tilt spread flick* response*

Grackle male 34 17 16 unusual usual Grackle female 55 25 16 unusual usual Grackle group 1 5 6 usual nearly always Position of grackles: on elevated song perch 2 6 8 unusual nearly always flying overhead 1 0 1 rare nearly always above Red-wing 1 22 14 not recorded not recorded below Red-wing 88 23 20 not recorded not recorded high in cattails 77 30 20 not recorded not recorded low in cattails 10 15 12 not recorded not recorded not in territory 1 2 0 unusual nearly always less than 5 ft. into terr. 32 16 15 usual usual 5-10 ft. into territory 48 27 16 usual unusual more than 10 ft. into terr. 6 1 1 unusual never grackle approachingits nest 68 27 19 usual usual

Not quantified. into the cattails or flew to another part of the territory when grackles made shallow intrusions. On other occasions,however, Red-wing males respondedaggressively to gracklesapproaching their own nests. By far the mostprevalent expression of aggressionat this time was a dive (Table 1), althoughdives were often mixed with bill-tilts, song-spreads,and tail- flicks in complexdisplay sequences. When younggrackles fledged they rapidly movedaway from the nest site, staying in the marsh only a day or two. Red-wing males were never seento respondto gracklefledglings even thoughthey were often well within Red-wingterritories and only a few feet from Red-wingmales. Adult gracklesarriving to feed such young, however,were almost always vigorouslydisplayed to or chasedfrom the territory by the Red-wing. Female Red-wingsare somewhatunusual among passerines in that they establish nest-centered "sub-territories" within their mate's territory, from which they repel all other Red-wingfemales (Nero, 1956). Female Red-wingswere observed to give bill-tilts to femalegrackles near the Red- wings'nest site on severaloccasions, but away from the nest site female Red-wingsrarely respondedto gracklesunless their matesdid, in which casethey movedabout nearbygiving disturbancecalls and tail-flicking. Grackleswere rarely aggressiveto Red-wingsalthough several times they dived at Red-wings(of either sex) two or three feet from the grackle nest. Generally the grackles tolerated Red-wings (and other gracklesas well) when they were more than four feet from the nest. 366 WIENS,Behavior of Red-wings and Grackles [ Vol.Auk 82

TABLE 2 GRACKLE RESPONSES TO RED-WING INTERSPECIFIC DISPLAYS GIVEN DURING THF. GRACKLECOURTSHIP AND NESTING PERIOD, 1962, 1963

Red-wing male display Grackle response Dive Bill-tilt Song-spread Tail-flick*

Dives at Red-wing 9 1 0 never Bill-tilt 2 19 2 never Song-spread 0 1 5 rare Drops into cattails 48 6 2 rare Flies away 26 6 5 rare Tail-flick* never unusual unusual unusual None 6 11 14 usual

Not quantified.

Gracklesgenerally responded to Red-wingaggression with evasiveactions or, occasionally,displays. They usually made no observableresponse to Red-wing song-spreads,but bill-tilts often elicited a grackle bill-tilt in return. Red-wing dives usually resulted in the grackles' dropping down low into the cattails or flying away (Table 2).

EFFECTS OF INTERACTIONS OF REPRODUCTION

Behavioralinteractions between species might be expectedto affect variousaspects of their reproduction,since they require energynormally availablefor intraspecificactivities. The initiation of breedingactivities couldbe retarded,nest placement or territory configurationaltered, pair- ing or nestingpatterns changed,feeding and broodingrates reduced,or nesting successaffected. Several instances have been recorded where interspecificinteractions apparently have reduced the reproductiveeffi- ciencyor loweredthe populationdensity of a species(Rip]ey, 1959, 1961; Oriansand Co]Her,1963; Wynne-Edwards,1962).

EFFECTS OF THE PRESENCE OF GRACKLES ON RED-WING REPRODUCTION Many authors (includingBent, 1958; Davis, 1944; Poor, 1946) have notedpredatory habits in the grackle. Indeed, Roberts (1932: 321) has suggestedthat grackle predation in some areas may be so great as to "prevent most other speciesmaking their home in the vicinity of a colony," and Meanley and Webb (1963) proposedthat grackles were probably responsiblefor most of the predation on Red-wing eggs in Chesapeaketidal marshes.On the Wingra marsh there was no evidence that grackleswere predatorson Red-wingeggs or young; their responses to Red-wingswere usually non-aggressive(Figure 5, Table 2). The presenceof grackles in the marsh had no observableeffects on the nature of Red-wing pair formation or nesting procedure. Displays July1965 ] WIENS,Behavior o!Red-wings andGrackles 367 used in courtshipor territorial behavior were also indistinguishablefrom those observedin "grackle-free" Red-wing coloniesor reported in the literature. There were slight modificationsof Red-wing territory bound- ariesto accommodatethe smallgrackle territories, but I doubt that grackle activity had any major effectson Red-wing territory configuration,or on the exact placementof Red-wing nests. Figures3 and 4 showthat Red-wingbreeding cycles in 1962 and 1963 were very similar with respectto timing. In the sametwo years the breeding cyclesof the grackleswere markedly dissimilar,even though female gracklesarrived (and courtshipbegan) about the same time each year. The constancyof Red-wing breedingin relation to the varying period of contactwith gracklesin the two years indicatesthat interactionshad little effect on the timing of Red-wing breeding, a conclusionfurther supportedby comparisonof the 1963 breedingchronology in the Wingra marsh with that in a "grackle-free"marsh a few miles away. In both marshes the various aspects of Red-wing nesting activity began and reachedpeaks at about the same time. A more complextype of reproductiveinterference could result from the releaseof excessiveaggressive behavior of Red-wingsin the presence of grackles,leading to a neglectof the nest or youngby the Red-wings. Such behavior has been termed "aggressiveneglect" by Hutchinsonand MacArthur (1959) and Ripley (1961). In the marsh female Red-wings were often attracted to Red-wing--grackleinteractions, and in 1962 several femalesnesting in areasof grackleactivity repeatedlyleft their nests to join in the interspecificresponses of their mates. In three of these nests the eggs becameaddled or the nestlingsdied. This suggeststhe operation of aggressiveneglect on at least a small scale. Breeding successin the entire marsh,however, was not markedlylower than that found for Red-wingsbreeding in the absenceof grackle activity in this region (Table 3). Such comparisonsmust be made cautiouslythough, because Red-wingnesting success may vary considerablywith regard to locality, season,vegetative substrate,and nest height (Case and Hewitt, 1963; Meanley and Webb, 1963). The low nestingsuccess of Red-wingsin the marshin 1963 is at least partially the result of two violent windstormsin early June.

EFFECTS OF THE PRESENCE OF RED-WINGS ON GRACKLE REPRODUCTION

The courtshipand pair formationbehavior of grackleswere apparently unchangedby Red-wing activity in the marsh, perhapsbecause most of the displayswere given in the treesbordering the marsh where Red-wing interferencewas slight. Gracklesshowed a definite tendency,however, to orient their courtship and nest site selectionflights to areas not in- 368 WIENs,Behavior o)• Red-wings and Grackles [ Vol.Auk 82

TABLE 3 COl•œPARISOlq'O1• RED-WING AND GRACKLE NESTING SUCCESSON THE EAST WlNGRA MARSH WITH NESTING SUCCESSESO1 • EACH SPECIES IN UNM'IXED COLONIES

Nests• Eggs• Year Hatched Fledged Young No. eggs young Laid Hatched ]ledged

Red-winged Blackbirds Grackles 1962 62 51 (82.3) 38 (61.2) 221 168 (76.0) 109 (49.3) present: 1963 76 52 (68.4) 36 (47.4) 255 162 (63.5) 92 (36.0) Grackles 19412 ------563 418 (74.3) 335-7 (59.7) absent: 19412 ------577 405 (70.2) 340-4 (59.3) 19478 91 78 (85.7) 57 (63.6) 325 258 (79.4) 170 (52.3) 19594 238 -- 83 (35.0) 730 393 (53.8) 204 (28.0) 19604 280 -- 67 (24.0) 902 402 (44.6) 162 (18.0)

Common Grackles Red-wings 1962 15 13 (86.6) 8 (53.3) 65 56 (86.1) 33 (50.6) present: 1963 18 14 (77.8) 12 (66.7) 81 61 (75.3) 44 (54.3) Red-wings 1947• 26 -- 12 (45.0) 117 84 (71.8) 51 (44.0) absent: 19485 15 -- 8 (53.3) 73 53 (72.6) 31 (42.5) 19495 21 -- 14 (66.7) 98 72 (73.4) 53 (54.1)

Percentagesgiven in parentheses. Smith, 1943; Cook County, Illinois. Beer and Tibbitts, 1950; Madison. Young, 1963 Stoddard,Wisconsin. Petersen and Young, 1950; Mad son. cluded in Red-wing territories, and the placementof grackle nests showed a correspondingcorrelation with the edgesof Red-wing territories (Fig- ure 1). This pattern of nesting distribution was probably a result of Red-wing aggressionin the more central portions of their territories. The manner of nest approachby marsh-nestinggrackles was considerably modified from that seenin "normal" colonies.In normal situations gracklesapproach the nest by flying directly to the nest area and quietly move to the nest site at nest level. In the Wingra marsh grackleswould alight in the cattails 10 to 25 feet from the nest and move slowly through the cattails toward the nest, approachingwell below nest level with re- peated tail-flicks. Such a slow, "inconspicuous"pattern of nest approach might functionto reduceinterspecific interference. Red-wing aggressivenessapparently had little effect on the timing of gracklenesting. Gracklenesting activity on the marshbegan considerably earlier in 1963 than in 1962 (Figures 3 and 4), although the birds ar- rived on the marsh and started courtship at about the same time each year. The reasonsfor the inconsistenttiming of gracklebreeding activity and the contrastingconsistency of Red-wing breeding activity are not dear, but it is probablethat the two speciesresponded to environmental conditionsin different ways. In 1963 gracklenesting chronology at the July1965 ] WlENS,Behavior o/ Red-wings andGrackles 369

Wingra marsh correspondedclosely with that of a colony of about 70 pairs nesting closeby in a plantation of 20-foot red cedars (Juniperus virginiana) where Red-wingswere absent. Grackle nestingsuccess in the marshagrees fairly well with the success of tree-nestinggrackle coloniesin other years at Madison (Table 3). In 1963 the gracklescompleted nesting before the two violent windstorms which affected Red-wing breedingsuccess.

DISCUSSION

Interspecificcompetition is usually if not always resolved,in the long run, by means other than interspecificaggression (see, for example, Crombie, 1947; Sv•irdson,1949; Wynne-Edwards, 1962; Hamilton, 1962; Moynihan, 1963). Where ecologicaloverlap is slight, interspecificaggression may involvean "undesirable"expenditure of time and energy (Dix- on, 1961; Orians and Collier, 1963) and perhapsgive rise to aggressive neglect (Udvardy, 1951; Ripley, 1961). Yet if ecologicalcompetition is great, interspecificaggression may lead to the establishmentof mutually exclusiveterritories between the species. Through such interspecific territorialismaggressive contacts between the speciesmay be reduced to the frequency of intraspecificcontacts. Even a slight reduction in the amount of interspecific competition through territorial segregation could be selectivelyadvantageous in enabling the speciesto retain both their ecologicalattributes and their sympatric distribution. Numerous examplesof interspecificterritorialism have been noted in birds (Simmons,1951; Johnson,1963; Orians and Willson, 1964). Often these involve speciesbreeding in structurally simple vegetation (such as marshesor grasslands)where there are few opportunitiesfor diversifica- tion of foraging techniquesand, hence, presumably few ways to avoid competition (Orians and Willson, 1964). Whether or not true interspecific territorialism existed between Red-wings and marsh-nesting grackleswas rather difficult to determine,because of the very small and vaguely defined territories of the grackles. Interspecific territorial aggressionwas often noted in both species,however, and presumably even partial territorial exclusionbetween the speciesmight contributeto their successful cohabitation of the marsh habitat. Selander and Giller (1963) have proposedthat in certain casesinter- specificterritorial aggressionmay be the fortuitousresult of the release of territorial behavior by morphologicalor behavioralsimilarities between the species.With such "mistaken identity" the aggressoris presumably respondingto sign stimuli which have only a partial resemblanceto the "optimal stimulus"presented by a memberof the samespecies (Tinber- gen, 1939). Such aggressionis especiallylikely when the thresholdof 370 W•E•s,Behavior of Red-wingsand Grackles [ Vol.Auk 82 the fighting responseis low or aggressive"motivation" is high, as is the caseduring the breedingseason. The interspecificresponses of the Red- wings and gracklesobserved in this study, however,indicated a more complex cause than mistaken identity, for Red-wing aggressionwas directed at grackleswith a fairly high frequencyand consistency(Figure 5, Table 1). In addition,Red-wings often respondedto gracklesby diving (Table 1), and activity which was generallyreplaced by ritualized displays in intraspecificdashes. Hinde (1952) has noted that diving is rare in intraspecificcontacts in severalEuropean tits (Parus), but occurs fairly often in interspecificencounters. Socialdisplays of animalsare rarely releasedby a singlesign stimulus, but rather are the result of simultaneouspresentation of a number of species-specificsocial releasers, a "heterogenoussummation" (Tinbergen, 1948). Simmons(1951) has advocatedthat mannerismsand general out- line or shapeare of prime importancein releasinginterspecific responses, while plumagecolor and pattern are of secondarysignificance. Interspe- cific territorial aggressionis by no means restricted to those species which are morphologicallysimilar (arians and Collier, 1963), but in general,species that hold mutually exclusiveterritories have similar pos- tures and/or vocalizationswhich enable them to elicit mutual aggressive responses(Johnson, 1963). Presumablya grackle intruding in a Red-wing territory does not pro- vide as completea set of releasersfor the residentRed-wing as does another Red-wing. Still, gracklesseem to produce a reasonablyclose facsimileof the aggregateof socialreleasers of Red-wing threat displays, for the responsesof Red-wingsto grackleswere not observablydifferent, qualitatively, from those induced by other Red-wings. In the closely- related Great-tailedand Boat-tailed grackles(Cassidix mexicanus and C. major) males regularly hold mutually exclusive territories, and all of the displaysnormally used in intraspecificinteractions appear in interspecificencounters as well (Selander and Giller, 1961). On the other hand, Gordon arians informs me (pers. comm.) that where Red-wings and Yellow-headedBlackbirds (Xanthocephalusxanthocephalus) breed together in Washington species-specificterritorial displays are only occasionally used in interspecificencounters, most contacts involving simple supplanting attacks. Hinde (1952) has noted that the displays usually given in intraspecific situations by European tits do not occur in encountersbetween species. Hinde thinks that the fighting drive is much more strongly activated than the fleeingdrive in interspecificskirmishes of tits, and that in order to activateboth drivesto approximatelyequal levelsthe particular stimulus conformationpresented by conspecificindividuals is required. Thus July1965 ] WIEN$,Behavior o]Red-wings andGrackles 371 displayssuch as "head up" (analogousto icterid bill-tilt), which seem to dependon suchan equilibriumbetween the two drives,do not appear in encountersbetween species. The fleeing or withdrawal tendencyis noted often in Red-wingsengaging in interspecificencounters with grackles;its activation (perhapsdue to the resemblanceof the grackle stimulusconfiguration to that of the Red-wing), interacting with the predominatingattack tendency,may create the "equilibrium"necessary for suchdisplays as bill-tilting to appear. In discussingthe natureof the responsesof SongSparrows (Melospiza melodia) to other specieson their territories,Nice (1943) noted an in- creasingtolerance of specificintruders after an initial periodof aggres- siveness,and suggestedthat this was the result of a habituation of the aggressiveimpulse after repeatedexercise on a harmlessvisitor. The decreasein frequencyor intensityof Red-wingresponses to grackleswith seasonaladvancement (Figure 5) could be the result of seasonalvaria- tions in aggressivenessrather than an indication of habituation. It is perhapsworth noting, however,that two bandedRed-wing males, resi- dents on the marsh in former years and thus birds with "previousex- perience"with grackles,were among the least aggressiveto intruding grackles.Orians and Collier (1963: 454) have noted that aggressivere- sponsesof male Red-wingsto Tricolored Blackbirds (Agelaiustricolor) invading their territorieswere intensebut ineffectual,and gradually subsided. Habituation would seemto be selectivelydisadvantageous when inter- specificecological competition was intense and where environmentalre- sourceswere limited, but could be advantageousin so far as it reduced distractionsto nestingactivity as visualizedin the conceptof aggressive neglect. As already noted, however,interspecific aggression may be ad- vantageouswhere it operatesas a dispersiveor density-limitingmechanism.

ACKNOWLEDGMENTS I wish to expressappreciation to Dr. John T. Emlen (for guidance,suggestions, and timely encouragementthroughout the study), to Dr. Gordon It. Orians (for suggestionson the manuscript), and to John Dallman (for helpful suggestionsre- garding the illustrations). The University of WisconsinArboretum Committee pro- vided researchfacilities in the University of WisconsinArboretum. This study was made in partial fulfillment of the requirementsfor the degree of Master of Science at the University of Wisconsin,and was supportedin part by a grant from the National ScienceFoundation. This is Journal Paper No. 62, University of Wisconsin Arboretum.

SUMMARY During the spring and summerof 1962 and 1963 the nature and effects of behavioralinteractions between Red-winged Blackbirds (Agelaius 372 W•E•s,Behavior of Red-wingsand Grackles [ Vol.Auk 82 phoeniceus)and Common Grackles (Quiscalusquiscula) were studied in a small cattail marsh at Madison, Wisconsin,where the two speciesnested in close association. The vegetation substrate constituted typical and apparently optimal nesting habitat for Red-wings,unusual and marginal habitat for grackles. Most interactionsbetween the two specieswere initiated by the Red- wings. Grackles were generally tolerated on the Red-wing singing perches about the edges of the marsh, but were attacked or displayed to when they penetratedthe marsh, especiallyearly in the season. Direct attack or diving was the most common Red-wing response. Bill- tilting and song-spread(threat displays), and tail-flicking (a "disturbance" display) occurredless often. Not infrequently Red-wings made no definite counter-responseto shallowpenetrations of gracklesinto their territories, and occasionallytheir behavior indicated avoidance of the grackles. Female Red-wingsrarely respondedto gracklesin the marsh unlesstheir mates did, but on severaloccasions they attacked or bill-tilted to female gracklesnear the Red-wing nest sites. Grackleswere rarely aggressivetoward Red-wings,but a few instances of interspecificdisplay about gracklenests were observed. Gracklestypi- cally respondedto Red-wing dives or threats with evasiveactions or oc- casionallyspecies-characteristic displays. The presenceof gracklesin the marsh had no apparent effects on Red- wing courtship,nesting patterns, or nest placement,and only slight effects on Red-wingterritory configuration.No predationby grackleswas observed. The initiation of breeding activities by Red-wings was not ob- servably retarded by the presenceof the grackles,and breeding success was similar to that found in Red-wing coloniesnot exposedto grackles. Severalnest failures, however,may have been due to Red-wing participa- tion in persistentinterspecific conflicts. The aggressivebehavior of Red-wings to grackles had several observ- able effects on grackle reproductivebehavior. Grackle courtship flights and nest sites becameoriented with respectto Red-wing territory edges. The pattern of nest approachby grackleswas considerablymodified from that seen in "normal" grackle colonies. Grackle courtship and pair formation, breeding schedules,and breeding successwere apparently unaf- fected by Red-wing activities. Territories of both specieswere to some extent mutually exclusive. Species-characteristicdisplays were used in interspecificas well as intraspecific territorial defense,indicating that each speciesproduced a rea- sonably closeapproximation of the releasersof the threat displaysof the other species. July1965 ] WIENS,Behavior o/ Red-wings and Grackles 373

LITERATURE CITED

ALLEN,A.A. 1914. The Redwinged Blackbird: a study in the ecology of a cattail marsh. Proc. Linnaean Soc. New York, 24-25: 43-128. ANDREW,R. J. 1961. The displays given by passefinesin courtship and reproduc- tive fighting: a review. Ibis, 103a: 313-348, 549-579. BEECriER,W. J. 1942. Nesting birds and the vegetation substrate. Chicago, Chi- cago Ornith. Soc., 69 pp. BEER,J. R., ANYD. TmBITTS. 1950. Nesting behavior of the Redwinged Blackbird. Flicker, 22: 61-77. BENT, A. C. 1958. Life histories of North American blackbirds, orioles, tanagers, and allies. U.S. Natl. Mus., Bull. 211: 1-549. CASE,N. A., Arm O. H. HEWITT. 1963. Nesting and productivity of the Red-winged Blackbird in relation to habitat. Living Bird, 2: 7-20. CROMB•E,A. C. 1947. Interspecific competition. J. Anim. Ecol., 16: 44-73. DAWS,M. 1944. Purple Grackle kills English Sparrow. Auk, t51: 139-140. Dixon, K. L. 1961. Habitat distribution and niche relationships in North Amer- ican speciesof Parus. Pp. 179-216 in Vertebrate speciation (W. F. Blair, ed.). Austin, Univ. Texas Press. FiCKE•T,R. W. 1963. Courtship and agonistic behavior of the Common Grackle, Quiscalusquiscula. Auk, 80: 52-72. HA•aIL•O•, T. YI. 1962. Speciesrelationships and adaptations for sympatry in the avian genus Vireo. Condor, 64: 40-68. HINDE, R. A. 1952. The behaviour of the Great Tit (Parus major) and some other related species. Behaviour Suppl. 2. Leiden, E. J. Brill. HV•C•NSON, G. E., ANDR. MAcART•VR. 1959. Appendix: on the theoretical sig- nificance of aggressiveneglect in interspecificcompetition. Amer. Nat., 93: 133-134. JomqsoN,N. K. 1963. Biosystematicsof sibling speciesof flycatchersin the Em- pidonax hammondii-oberholseri-wrightiicomplex. Univ. Calif. Publs. Zo51., 1515: 79-238. LANYON,W. E. 1957. The comparative biology of the meadowlarks (Sturnella) in Wisconsin. Publs. Nuttall Ornith. Club, No. 1: 1-67. MARKER,P. 1961. The evolution of visual communication. Pp. 96-121 in Vertebrate speciation (W. F. Blair, ed.). Austin, Univ. Texas Press. MEANLEY,B., ANDJ. S. WEBB. 1963. Nesting ecologyand reproductiverate of the Red-winged Blackbird in tidal marshes of the upper ChesapeakeBay region. Chesapeake Sci., 4: 90-100. MOY•AN, M. 1962. The organization and probable evolution of some mixed speciesflocks of neotropicalbirds. SmithsonianMisc. Coils., 143(7): 1-140. MOYm•AN, M. 1963. Inter-specific relations between some Andean birds. Ibis, 105: 327-339. NERO,R. W. 1956. A behavior study of the Red-winged Blackbird. Wilson Bull., 611: 5-37, 129-150. NERO, R. W. 1963. Comparative behavior of the Yellow-headed Blackbird, Red- winged Blackbird, and other icterids. Wilson Bull., 75: 376-413. NICE, M. M. 1943. Studiesin the life history of the Song Sparrow. II. The behavior of the Song Sparrow and other passerines.Trans. Linn. Soc. New York, 15: 1-328. OPaANS,G. H., A•D G. CO•IER. 1963. Competitionand blackbird social systems. Evolution, 17: 449-459. 374 WreNs,Behavior o! Red-wingsand Grackles [ Vol.Auk 82

Ore^Ns, G. H., ^m) M. F. W•LLSOS. 1964. Interspecific territories of birds. Ecology, 45: 736-745. PgTERSE•,A., ^•D H. YO•JNG. 1950. A nestingstudy of the Bronzed Grackle. Auk, 67: 466-476. PooR, H. 1946. Predation by grackles. Proc. Linn. Soc. New York, 54--57: 54-55. R^m), A. L. 1954. Social feeding behavior of birds. Fieldiana: Zoology, 36(1): 1-71. RxPLg¾,S. D. 1959. Competition between sunbird and honeyeater speciesin the Moluccan Islands. Amer. Nat., 93: 127-132. R•?LEY, S. D. 1961. Aggressiveneglect as a factor in interspecific competition in birds. Auk, 711: 366-371. ROBZXTS,T. S. 1932. The birds of Minnesota. Vol. 2. Minneapolis,Univ. Minn. Press. SEZ^•D•R,R. K., A•D D. R. GrazeR. 1959. Interspecificrelations of woodpeckersin Texas. Wilson Bull., 71: 107-124. SEz^m)•x,R. K., ^•) D. R. Gr•L•R. 1961. Analysisof sympatry of Great-tailed and Boat-tailed grackles. Condor, 63: 29-86. S•g^•u•R, R. K., Asu D. R. GrunER. 1963. Specieslimits in the woodpeckergenus Centurus (Aves). Bull. Amer. Mus. Nat. Hist., 124: 213-274. S•MMo•s, K. E. L. 1951. Interspecificterritorialism. Ibis, 93: 407-413. S•xr•, H.M. 1943. Size of breedingpopulations in relation to egg laying and re- productive successin the Eastern Red-wing (Agelaius p. phoeniceus). Ecology, 24: 183-207. SvXRusos,G. 1949. Competitionand habitat selectionin birds. Oikos, 1: 157-174. T•B•RG•, N. 1939. The behavior of the Snow Bunting in spring. Trans. Linn. Soc. New York, 5: 1-95. T•R½•, N. 1948. Social releasersand the experimentalmethod required for their study. Wilson Bull., 60: 6-51. Uuv^RD¾,M. D.F. 1951. The significanceof interspecificcompetition in bird life. Oikos, 3: 98-123. W¾•g-Euw^vd•s, V. C. 1962. Animal dispersion in relation to social behavior. New York, Hafner. You•½, H. 1963. Age-specificmortality in the eggs and nestlingsof blackbirds. Auk, 80: 145-155. Departmentof Zoology,University o] Wisconsin,Madison, Wisconsin.