Gila Woodpecker Agonistic Behavior

GILA WOODPECKER AGONISTIC BEHAVIOR

GENE L. BRENOWITZ

ABSTR•CT.--Agonisticbehavior of Gila Woodpeckers,including vocalizations, visual displays, and other related behaviors,is described.Interactions with both con- and heterospecificswere analyzedby stochasticprocesses, and it is shownthat the timing of aggressiontoward a species coincidedwith the time during which that specieswas searchingfor nest sitesor cavities.The behaviorshown toward Flickersand Starlingsappears to fulfill criteriafor interspecificterritoriality, and evidenceis presentedto supportthe contentionthat the functionof this behavioris protection of nest sites and nest holes from competitors.The effects of intra- and interspecific territorialityon the distributionof the speciesinvolved were found to be significantlydifferent. Simultaneousprotection of differentsized territories is discussed.--DepartmentofBiology, Univer- sity of New Mexico, Albuquerque87131. Present address: Department of Zoology,Michigan State University, East Lansing, 48824. Accepted21 June 1976.

GILA Woodpeckers(Melanerpes uropygialis), which are commonlyassociated with the giant cactus(Carnegiea gigantea) of Arizona'sdesert lowland, inhabit cottonwood (Populusdeltoides) groves along the Gila River in southwestNew Mexico. While the behavior of its congeners,the Golden-fronted (M. aurifrons) and Red-bellied (M. carolinus) Woodpeckers have been studied thoroughly (Kilham 1958, 1961, 1963, Selanderand Giller 1959, Stickel 1965), details of Gila Woodpeckerbehavior in the literature are limited to largely anecdotalaccounts. The most striking feature of these reportsis that Gila Woodpeckersare consistentlydescribed as beinghighly aggressive toward both conspecificsand a broad range of other speciesincluding Bronzed Cow- birds (Tangaviusaeneus) and thrashers(Toxostoma spp.) (Gilman 1915, Bent 1939). My preliminary observationsindicated that they are distinctly more aggressive than other speciesin the region they inhabit. The main purposeof this study was to determine the ecologicalfunctions of Gila Woodpeckers' agonistic behavior. Apparently they are territorial toward conspecifics, Common Flickers (Colaptesauratus), and Starlings (Sturnus vulgaris). Several lines of evidence suggestthat protection of nest sites and nest holes from competitorsis the main function of their interspecific aggression.In this paper descriptionsof agonisticbehaviors are presentedfirst. Then stochasticprocesses, which have proved useful in demonstratingchanges in behavior over time (Stokes 1962), are used to describechanges in levels of agonisticbehavior shown toward other pairs of conspecificsand membersof other speciesover the courseof the breeding season. Following this quantitative analysis, details of intra- and interspecific territoriality are presented. The functions of Gila Woodpecker territoriality are considerednext. The hypothesisthat speciestoward which they are territorial should overlap more in their use of nest sites than other species,is tested and confirmed by comparing nest sitesfor thosespecies. Direct evidenceof competitionfor nest holesis also presented. Finally the effectsof intra- and interspecificterritoriality on the distribution of the speciesinvolved are compared by examining distancesbetween Gila Woodpecker nest cavities and their nearest conspecificand Flicker nest cavities.

STUDY AREA AND METHODS I studiedGila Woodpeckersat Red Rock on the Gila River in Grant County,New Mexico, on 17-18 March and from 17 April to 4 July, 1973, and from 16 to 20 March 1974. The studytract consistedof severalcottonwood groves on both banksof the river: eachgrove contained 1-3 pairsof Gila Woodpeck- 49 The Auk 95: 49-58. January 1978 50 GENEL. Bm•NOWXTZ [Auk, Vol. 95

Fig. 1. Territoriesof Gila Woodpeckerpairs 1, 2, and 9. All threeterritories, as well as thoseof all other pairs studied, were within the limits of cottonwoodgroves. Solid lines representboundaries of in- traspecificallydefended areas, and dottedlines represent boundaries of interspecificallydefended areas. ers.The canopywas almost exclusively cottonwoods, and the groundwas covered with sandand debris, from recentflooding. This riparianzone extended less than 200 m backfrom the river and adjoineda narrow strip of irrigated farmland. With few exceptions,no trees were found beyondthis, and desert formed the remainder of the local habitat (Fig. 1). Six pairs of Gila Woodpeckerswere followedclosely through the breedingseason, and threeadditional pairs were watchedoccasionally. Total observationtime was in excessof 400 hours. Althoughno birds were banded,I couldidentify three males(designated p3M, p7M, and p8M) and two females(p7F and p8F) by their distinctivecalls. Each of thesebirds remainedin a particularterritory throughthe 1973 breedingseason. Common Flickers, LadderbackedWoodpeckers (Picoides scalaris), and Starlingswere also studied becauseof their interactions with Gila Woodpeckers. I recordedvocalizations and drumming on a Uher 4000 Report-L tape recorderat a tape speedof 19 cm per sec with use of a Uher M514 microphone and, sometimes, a 61-cm diameter parabolic reflector. Vocalizations were analyzed by use of a Kay Electric Company Sonagraph model 6061B, with a model 6076 amplitude display and scale magnifier unit. A detailed analysis of Gila Woodpecker vocalizations will be presentedelsewhere (Brenowitz 1977). Stochastic processes.--I recorded frequencies of behaviors of other pairs of conspecifics,Flickers, LadderbackedWoodpeckers, and Starlings(designated preceding events 1-19; seeTables 1, 2, and 3) that sometimeswere followed by Gila Woodpeckeragonistic behaviors (designated following eventsA-G). For each repetition of a behavior I recordedwhether or not a Gila Woodpeckeragonistic behavior followed and, if so, which one (A-G) it was. In situationswhere birds exchangedseveral behaviors back and forth, onlythe initial behaviorswere included. No otherpreceding event was counted until the prior exchange concludedand the pair being watched was again in a position to undertake a following event. Visual displayswere excludedfrom this analysisbecause they can not be detectedreliably in the densecotton- wood foliage. Over 300 hours of observationswere made for this analysis. Probabilitiesfor each preceding event-followingevent sequence,as well as precedingevent-no- following event sequences,were calculatedas follows: frequency of a following event (or no-following event)/frequencyof a precedingevent. The total probability of a precedingevent beingfollowed by Gila Woodpeckeragonistic behavior also was calculated. The study was divided into three time blocks:the prenestlingperiod (23 April-9 May 1973), the early nestling period (15-29 May 1973), and the late nestling-fledglingperiod (30 May-4 July 1973).A separateset of probabilitieswas calculatedfor eachtime block (seeTables 1, 2, and 3). January1978] Gila WoodpeckerAgonistic Behavior 5 1

TABLE 1 RESULTS OF STOCHASTIC PROCESS FOR PRENESTLING PERIOD

Following event A B C D E No Call Call Drum- Gravel Sup- F G following Sum of Precedingevent: 1 2 ming call plant Chase Attack event A-G Flicker 1) Intrusion 0.00 0.09 0.00 0.00 0.46 0.27 0.09 0.09 0.91 (n = 11) (0)• (1) (0) (0) (5) (3) (1) (1) (10) 2) Drumming 0.18 0.00 0.00 0.00 0.00 0.00 0.00 0.82 0.18 (n = 34) (6) (0) (0) (0) (0) (0) (0) (28) (6) 3) "Kheer"• 0.13 0.13 0.00 0.00 0.00 0.00 0.00 0.75a 0.25 (n = 24) (3) (3) (0) (0) (0) (0) (0) (18) (6) 4) "Wicks.." 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.00 0.00 (n = 5) (0) (0) (0) (0) (0) (0) (0) (5) (0) 5) "Gila-like" 0.35 0.07 0.00 0.00 0.00 0.00 0.00 0.58 0.42 (n = 43) (15) (3) (0) (0) (0) (0) (0) (25) (18) 6) "Eh, eh.." 0.00 0.06 0.00 0.00 0.00 0.00 0.00 0.94 0.06 (n = 18) (0) (1) (0) (0) (0) (0) (0) (17) (1) c• Gila 7) Intrusion 0.00 0.00 0.00 0.00 0.00 0.33 0.67 0.00 1.00 (n = 3) (0) (0) (0) (0) (0) (1) (2) (0) (3) 8) Drumming 0.20 0.00 0.60 0.00 0.00 0.00 0.00 0.20 0.80 (n = 10) (2) (0) (6) (0) (0) (0) (0) (2) (8) 9) "Call 1" 0.18 0.00 0.00 0.00 0.00 0.00 0.00 0.82 0.18 (n = 51) (9) (0) (0) (0) (0) (0) (0) (42) (9) 10) "Call 2" 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.00 0.00 (n = 8) (0) (0) (0) (0) (0) (0) (0) (8) (0) 9 Gila 11) Intrusion ...... (n = 0) ...... 12) Drumming 0.00 0.00 1.00 0.00 0.00 0.00 0.00 0.00 1.00 (n = 1) (0) (0) (1) (0) (0) (0) (0) (0) (1) 13) "Call 1"0.00 0.00 0.00 0.0œ 0.00 0.00 0.00 1.00 0.00 (n = 7) (0) (0) (0) (0) (0) (0) (0) (7) (0) 14) "Call 2" 0.17 0.00 0.00 0.00 0.00 0.00 0.00 0.83 0.17 (n = 12) (2) (0) (0) (0) (0) (0) (0) (10) (2) Ladderbacked 15) Intrusion 0.07 0.00 0.00 0.00 0.71 0.07 0.00 0.14 0.86 (n = 14) (1) (0) (0) (0) (10) (1) (0) (2) (12) 16) Drumming 1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.00 (n = 1) (1) (0) (0) (0) (0) (0) (0) (0) (1) 17) "Pik.." 0.58 0.08 0.00 0.00 0.00 0.00 0.00 0.33 0.67 (n = 12) (7) (1) (0) (0) (0) (0) (0) (4) (8) Starling 18) Intrusion 0.00 1.00 0.00 0.00 0.00 0.00 0.00 0.00 1.00 (n = 2) (0) (2) (0) (0) (0) (0) (0) (0) (2) 19) "Calls" 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.00 0.00 (n = 2) (0) (0) (0) (0) (0) (0) (0) (2) (0) Numbersin parenthesesare the "n's"associated with probabilitiesabove them. All precedingevents in quotationmarks are vocalizations. For a precedingevent, probabilities should add to 1.00;discrepancies are dueto rounding errors.

Comparisonof nest sites.--Sites of 6 Gila Woodpecker, 5 Flicker, and 3 LadderbackedWoodpecker nest cavities were scoredin the following 2 categoriesaccording to diameter: (1) trunks and limbs, (2) branches. Nest sites for these three specieswere then compared by Fisher exact probability tests. Intra- and interspecificdistributions of nestcavities.--Distances from Gila Woodpeckernest cavities to the nearestGila Woodpeckerand Flicker nestcavities were measuredand the distributionscompared by a Mann-Whitney U-test.

DESCRIPTIONS OF AGONISTIC BEHAVIORS Agonisticbehavior includesa broad range of activitiesused in conflict situations. In this study aggressionwas defined operationally as supplantings, chases, and 52 GENE L. BRENOWITZ [Auk, Vol. 95

TABLE 2 RESULTS OF STOCHASTIC PROCESS FOR EARLY NESTLING PERIOD

Following event

A B C D E No Call Call Drum- Gravel Sup- F G following Sum of Preceding event: 1 2 ming call plant Chase Attack event A-G Flicker 1) Intrusion 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.00 0.00 (n = 1) (0)• (0) (0) (0) (0) (0) (0) (1) (0) 2) Drumming 0.00 0.17 0.00 0.00 0.00 0.00 0.00 0.83 0.17 (n = 6) (0) (1) (0) (0) (0) (0) (0) (5) (1) 3) "Kheer"2 0.01 0.03 0.00 0.00 0.00 0.00 0.00 0.95a 0.04 (n = 87) (1) (3) (0) (0) (0) (0) (0) (83) (4) 4) "Wicka.." 0.00 0.06 0.00 0.00 0.00 0.00 0.00 0.94 0.06 (n = 18) (0) (1) (0) (0) (0) (0) (0) (17) (1) 5) "Gila-like" -- (n = 0) 6) "Eh, eh . ." 0.08 0.00 0.00 0.00 0.00 0.00 0.00 0.92 0.08 (n = 50) (4) (0) (0) (0) (0) (0) (0) (46) (4) d Gila 7) Intrusion (n = 0) 8) Drumming 0.30 0.03 0.03 0.00 0.00 0.00 0.00 0.65 0.35 (n = 37) (11) (1) (1) (0) (0) (0) (0) (24) (13) 9) "Call 1" 0.16 0.02 0.02 0.01 0.00 0.00 0.00 0.79 0.21 (n = 288) (46) (6) (6) (3) (0) (0) (0) (227) (61) 10) "Call 2" 0.12 0.12 0.04 0.00 0.00 0.00 0.00 0.73 0.27 (n = 26) (3) (3) (1) (0) (0) (0) (0) (19) (7) 9 Gila 11) Intrusion (n = 0) 12) Drumming (n = 0) 13) "Call 1" 0.03 0.03 0.00 0.00 0.00 0.00 0.00 0.95 0.05 (n = 40) (1) (1) (0) (0) (0) (0) (0) (38) (2) 14) "Call 2" 0.11 0.00 0.00 0.00 0.00 0.00 0.00 0.89 0.11 (n = 75) (8) (0) (0) (0) (0) (0) (0) (67) (8) Ladderbacked 15) Intrusion 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.00 0.00 (n = 1) (0) (0) (0) (0) (0) (0) (0) (1) (0) 16) Drumming 0.50 0.00 0.00 0.00 0.00 0.00 0.00 0.50 0.50 (n = 4) (2) (0) (0) (0) (0) (0) (0) (2) (2) 17) "Pik . ." 0.04 0.01 0.00 0.00 0.00 0.00 0.00 0.95 0.05 (n = 209) (8) (2) (0) (0) (0) (0) (0) (199) (10) Starling 18) Intrusion 0.00 0.06 0.00 0.00 0.26 0.10 0.00 0.58 0.42 (n = 31) (0) (2) (0) (0) (8) (3) (0) (18) (13) 19) "Calls" 0.00 0.50 0.00 0.00 0.00 0.00 0.00 0.50 0.50 (n = 2) (0) (1) (0) (0) (0) (0) (0) (1) (1) Numbersin parenthesesare the "n's" associatedwith probabilitiesabove them. All precedingevents in quotationmarks are vocalizations. For a precedingevent, probabilities should add to 1.00;discrepancies are due to roundingerrors.

attacks. In Gila Woodpeckersaggression is not the culmination of a seriesof agonistic behaviors. In virtually all supplantings,chases, and attacks no previousagonistic behavior was observed.I rarely observedaggression in associationwith other behaviors, although call 2 sometimesaccompanied supplantings of Starlings. Call I is composedof uniform vibrato notesand soundsvery much like the "Cha- aa-ah" territorial call of Red-bellied Woodpeckers(see Kilham 1961). It is used as a contact call by which members of a pair signal their location and is the Gila Wood- pecker'sprimary territorial advertisementdisplay. Males usethis call more often than do females. January 1978] Gila WoodpeckerAgonistic Behavior 53

TABLE 3 RESULTS OF STOCHASTIC PROCESS FOR LATE NESTLING-FLEDGLING PERIOD

Following event A B C D E No Call Call Drum- Gravel Sup- F G following Sum of Precedingevent: 1 2 ming call plant Chase Attack event A-G

Flicker 1) Intrusion 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.00 0.00 (n = 2) (0)' (0) (0) (0) (0) (0) (0) (2) (0) 2) Drumming 0.06 0.00 0.06 0.00 0.00 0.00 0.00 0.88 0.12 (n = 16) (1) (0) (1) (0) (0) (0) (0) (14) (2) 3) "Kheer"2 0.04 0.00 0.00 0.00 0.00 0.00 0.00 0.96 0.04 (n = 53) (2) (0) (0) (0) (0) (0) (0) (51) (2) 4) "Wicka.." 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.94 0.06 (n = 17) (1) (0) (0) (0) (0) (0) (0) (16) (1) 5) "Gila-like" -- ...... (n = o) ...... 6) "Eh, eh . ." 0.04 0.00 0.00 0.00 0.00 0.00 0.00 0.96 0.04 (n = 95) (4) (0) (0) (0) (0) (0) (0) (91) (4) c• Gila 7) Intrusion 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.00 0.00 (n = 1) (0) (0) (0) (0) (0) (0) (0) (1) (0) 8) Drumming 0.20 0.00 0.00 0.00 0.00 0.00 0.00 0.80 0.20 (n = 5) (1) (0) (0) (0) (0) (0) (0) (4) (1) 9) "Call 1" 0.11 0.00 0.02 0.00 0.00 0.00 0.00 0.87 0.13 (n = 119) (13) (0) (2) (0) (0) (0) (0) (104) (15) 10) "Call 2" 0.14 0.00 0.00 0.00 0.00 0.00 0.00 0.86 0.14 (n = 14) (2) (0) (0) (0) (0) (0) (0) (12) (2) 9 Gila 11) Intrusion 0.25 0.00 0.00 0.00 0.00 0.00 0.00 0.75 0.25 (n = 4) (1) (0) (0) (0) (0) (0) (0) (3) (1) 12) Drumming ...... (n = O) ...... 13) "Call 1" 0.06 0.00 0.00 0.00 0.00 0.00 0.00 0.94 0.06 (n = 31) (2) (0) (0) (0) (0) (0) (0) (29) (2) 14) "Call 2" 0.00 0.05 0.00 0.00 0.00 0.00 0.00 0.95 0.05 (n = 22) (0) (1) (0) (0) (0) (0) (0) (21) (1) Ladderbacked 15) Intrusion 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.00 0.00 (n = 1) (0) (0) (0) (0) (0) (0) (0) (1) (0) 16) Drumming ...... (n = O) ...... 17) "Pik . ." 0.50 0.00 0.00 0.00 0.00 0.00 0.00 0.50 0.50 (n = 2) (1) (0) (0) (0) (0) (0) (0) (1) (1) Starling 18) Intrusion 0.04 0.08 0.00 0.04 0.21 0.13 0.04 0.46 0.54 (n = 24) (1) (2) (0) (1) (5) (3) (1) (11) (13) 19) "Calls" -- ...... (n = O) ...... Numbers in parenthesesare the "n's" associatedwith probabilitiesabove them. All precedingevents in quotationmarks are vocalizations.

Call 2 is a seriesof sharp"pip, pip" notes.It is mostfrequently given in responseto disturbanceby humans, and the presenceof intruders of other species.Call 2 is the only vocalizationgiven in conjunctionwith visual displays(head bobbing and head shaking) and serves as a general alarm call. Drumming is a meansof soundproduction in which the bill rapidly strikesa resonantobject such as a deadbranch. Drumming is considereda form of territorial proclamation (Lawrence 1967). Gravel call consistsof harsh rasping notes spectrographicallysimilar to those of 54 GI•NI• L. BRENOWITZ [Auk, Vol. 95

call 1. It was heard infrequently, in situations where Gila Woodpeckers appeared highly agitated by birds of other species. Bill pointing display.--The bill, head, and neck are extended parallel to the substrate on which an individual is perched, and the pose is held rigidly. The displayingbird orientsitself so that its bill points directly at the individual eliciting the display. This display seemssimilar to the rigid poseof Red-bellied Woodpeckers (Kilham 1961). Head bobbingand head shaking displays.--The head is thrust in a forward and downward arc from a position with the bill parallel to the substrate, to a position with the bill almost perpendicularto the substrate.The return motion of raising the head is lessexaggerated. In someinstances the arc is increasedby bringing the head back so that the bill pointsabove a positionparallel to the substrate.The orientation of the displayingbird varies with respectto the individual eliciting the display. Head shaking is similar to head bobbing but with lateral movement accompanyingthe up-down movement. Displays similar to these have been describedfor the following genera of woodpeckers: Picoides (Lawrence 1967, Kilham 1969, Short 1971), Sphyrapicus(Lawrence 1967), and Colaptes(Lawrence 1967). Supplanting is replacing an individual at a given place. In supplanting a Gila Woodpecker flies and/or hitches along a branch to the place where another individual is perched or foraging. Commonly the latter leaves when the Gila Wood- pecker arrives. Chasingis pursuinga secondindividual alsoin motion. Suchpursuits were toward the boundaries of a Gila Woodpecker's territory and always away from the nest cavity. Chasing and supplanting often occurred together. On one occasion p9M supplanteda Starling 39 times with chasesbetween successivesupplantings. Attacking is movement toward an individual that culminates in forceful contact between individuals.

RESULTS STOCHASTIC PROCESSES

As male Gila Woodpeckersplay a more active role in agonisticencounters with both con- and heterospecifics,only resultsfor malesare presentedhere. Tables 1, 2, and 3 show the probabilitiesof Flicker, Gila Woodpecker,Ladderbacked Wood- pecker, and Starling behaviors (precedingevents 1-19) being followed by Gila Woodpeckeragonistic behaviors (following events A-G) for the prenestling,early nestling,and late nestling-fledglingperiods, respectively. Results are presentedac- cordingto categoriesof precedingevents (intrusion, vocalizations,and drumming). Intrusion .--The activities of intruders in the territory of the pair being watched were groupedtogether in the "intrusion"category. For membersof otherspecies this categorywas their presencein the spacewithin the circumferenceof the crownof the nest cavity tree. One can predict with considerablecertainty that during the prenestlingperiod, intrusionsof the speciesstudied will be followed by Gila Woodpeckeragonistic behavior (see total probabilities, Table 1). Furthermore for all speciesexcept Star- lings, this behavior will most likely be some form of aggression(supplant, chase, or attack). In no caseis it possibleto predict with certainty which specificbehavior will follow. In the early nestlingperiod, intrusionsby speciesother than Starlingswere rare and were not followed by Gila Woodpeckeragonistic behavior. While the number of January 1978] Gila WoodpeckerAgonistic Behavior 55

Starling intrusionsincreased, it was not possibleto predict that Gila Woodpecker agonisticbehavior would follow. When followingevents do occur,they will likely be some form of aggression. In the late nestling-fledglingperiod aggressioncontinued to follow Starling intru- sionsbut, once again, not predictablyso. Also, a greater variety of behaviorsfol- lowed theseintrusions so that even when following events are observed,that they will be aggressionis lesspredictable. It is interestingto note that followingintrusions by other species,Gila Woodpeck- ers were more likely to supplant individuals than chase or attack them. In- traspecificallythis trend was reversed.Also, the periodswhen aggressiontoward other speciesoccurred coincided with the times that those specieswere actively seekingnest sites. Drumming.--One can predict that during the prenestingperiod most male Gila Woodpecker drummings will be followed by agonistic behavior (see Table 1). For Flicker drummingsone can predict with some certainty that no events will follow. Ladderbacked Woodpecker and female Gila Woodpeckerdrumming occurredtoo infrequentlyto make reasonablepredictions. Predictions regarding Flicker drumming are the samefor the remainder of the study, and for other speciestotal probabilities decrease together with the number of instances of drumming. Vocalizations.--Precedingevents in quotationmarks (seeTables 1, 2, and 3) are vocalizations.For descriptionsof Flicker and LadderbackedWoodpecker vocalizations seeLawrence (1967) and Short (1971), respectively.The most striking observa- tion is that, except for Ladderbacked Woodpecker "Pik..'s," one can predict that most vocalizationsof all speciesstudied will not be followed by agonisticbehavior. This is true for all time blocks.

TERRITORIALITY Most aggressivebehavior of Gila Woodpeckersappears to be relatedto defenseof an area, territoriality. By March pairs were regularly spacedalong the river (mean distancebetween cavities: 120 m, SE -+ 7, n = 6) and most intraspecific encounters were limited to exchangesof vocalizations.Their territories(Fig. 1) were multipurpose (type A, Hinde 1956), and virtually all disputed boundariescoincided with localized food sources;e.g. a newly ripened raspberry bush (Rubus sp.) was the location of one encounter. Gila Woodpeckerspersistently defended space up to 40-50 m from their nest cavitiesfrom Flickers and Starlingsfor at leastpart of the breedingseason (see results of stochasticprocesses). Furthermore all typesof agonisticbehavior they displayed toward conspecificsthey also displayed toward Flickers, and they used all types except visual displays in interactions with Starlings. Thus behavior shown toward these speciesmeets criteria establishedfor interspecific territoriality (cf. Simmons 1951, Lanyon 1956). In contrast, they did not defend space from Ladderbacked Woodpeckers,and the only aggressionseen between these species involved one pair of LadderbackedWoodpeckers and one male Gila Woodpeckeron a singleday. Althoughthis studyprimarily concernsGila Woodpeckerbehavior, some mention of other species'responses to their aggressionseems appropriate. The commonest responseof both Flickers and Starlingsto supplantings,chases, and attacksby Gila Woodpeckerswas to retreat toward the periphery of the resident'sterritory. In several instancesStarlings held their ground despite Gila Woodpeckers'efforts to supplant them. Once a male (p3M) attacked a Starling perched in its nest cavity tree 56 GENE L. BRENOWlTZ [Auk, Vol. 95 only to be attacked itself and driven from that tree by the same bird. This was the only time I witnesseda Starling clearly dominating a Gila Woodpecker. Although Flickers were far less tenaciousthan Starlings, I recordedone instancein which a pair of Flickers supplanteda male (plM) from the branch containingits nest cavity. Neither Flickers nor Starlingsattempted to exclude Gila Woodpeckersfrom space around their nestcavities, so apparentlyneither speciesreciprocates the territoriality the Gila Woodpeckers show toward them. Evidence for the function of interspecific territoriality.--Nest sites for Flickers, Gila Woodpeckers,and LadderbackedWoodpeckers were comparedby useof Fisher exactprobability tests.Gila Woodpeckersused trunks and limbs for nestsites significantly more than did Ladderbacked Woodpeckers (P -- 0.047), whereas they did not differ from Flickers in this regard (P -- 0.545). Starlings were excludedfrom this analysis becausea trunk or limb must already contain a cavity to be a potential nest site for them. These resultsindicate a greater potential for competitionfor nest sites between Gila Woodpeckers and Flickers than between Gila and Ladderbacked Woodpeckers. There is also direct evidence of nest hole competition. On 30 April 1973 a pair of Flickers took a cavity from a male Gila Woodpecker (plM). They subsequently enlargedits entranceand succeededin rearingyoung there. Starlingsare well known as nesthole competitorsof other speciesof woodpeckers(Kessel 1957, Kilham 1959). They moved into Red Rock in early April, but did not break up into pairs and begin nestinguntil the early nestlingperiod of Gila Woodpeckers.Three pairs of Gila Woodpeckers(pl, p2, p3) lost cavitiesto Starlingsduring the early nestlingand late nestling-fledglingperiods, and one pair (p2) lost three successivecavities, providing additional evidencethat Flickers and Starlingsare sometimesdominant over Gila Woodpeckers. No cavities were lost to Ladderbacked Woodpeckers. Intra- and interspecificdistribution of nest cavities.--The distancebetween Gila Woodpecker nest cavities (X = 120 m, SE --_ 7.4, n = 6) was significantly larger than the distance between Gila Woodpecker and Flicker nest cavities (X = 76 m, SE -+ 16.6, n = 9; Mann-Whitney U = 30, P < 0.05). Once again, Starlingswere excluded from this analysis for reasonsstated above.

DISCUSSION

Although most reported casesof interspecificterritoriality involve closelyrelated species(Johnson 1963, Cody 1969, Cody and Brown 1970, Murray 1971, Rohwer 1973),arians and Willson (1964)suggested that it might be profitablefor a speciesto exclude dissimilar potential competitorsas well. Gila Woodpeckersappear to be an example of such a species.While protectionof nest holeshas been largely excluded from considerationsof interspecificterritoriality (Simmons1951, arians and Willson 1964), severallines of evidencestrongly implicate it as a function of Gila Woodpecker interspecificaggression. First, the timing of aggressiontoward a species(as shown by stochasticprocesses) coincided with the periodduring which that specieswas actively seekingnest sites or cavities. Second,it has been shownthat Gila Woodpeckersare territorial toward speciesthat overlap with them in use of nest sites and are not territorial toward speciesthat usesignificantly different nest sites. Direct evidenceof competition for nest cavities also was presented. arians and Willson (1964) suggestthat interspecificterritoriality should serve fewer functionsthan intraspecificterritoriality, and the evidencepresented here sup- January1978] Gila WoodpeckerAgonistic Behavior 5 7 ports this view. It seemed reasonable to expect that differences in functions of territories might be reflectedby differencesin the distribution of the speciesinvolved. A comparison of the distancesbetween Gila Woodpecker nest cavities and the nearest conspecificand Flicker nest cavities showsthat intraspecific distancesbetween nest cavities were significantlylarger than interspecificdistances. A territory of given character can relate specificallyto intruders of a particular category. Here speciesthat compete with or threaten Gila Woodpeckersin different ways form these categories. Defense against Flickers and Starlings, which compete for nest sites and cavities, and conspecificsdiffer. Theoretically then, a speciescan simultaneouslymaintain more than one territory. In the present case,a spacearound the nest cavity is defendedfrom con- and heterospecifics,and a larger surrounding spaceis protectedonly from conspecifics.Thus, despitea differencein size of spaces defendedintra- and interspecifically,both can be consideredterritories. Murray (1971) contendsthat known casesof interspecificterritoriality are consis- tent with the assumptionthat it is misdirectedintraspecific territoriality and, there- fore, a nonadaptive characteristic,but he does suggestcriteria for recognizingcases of adaptive interspecific territoriality. Observations of competition for cavities and closeregulation of aggressionboth spatially and temporally suggestthat Gila Wood- peckers can recognize the behavior of Flickers and Starlings. Responseto these speciesprobably is not misdirectedintraspecific aggressionand appears to be highly adaptive. Also, this behavior fulfills several of Murray's (1971) criteria for adaptive interspecific territoriality.

ACKNOWLEDGMENTS

I am grateful to J. D. Ligon for his help in all phasesof this study. M. Balaban, L. S. Demski, K. B. Domino, J.P. Gluck, R. J. Harris, and J. A. King helped in the analysisof data and preparationof this manuscript, as well as in many other ways. I would also like to thank Mr. and Mrs. G. Harper, Mr. and Mrs. G. Hightower, C. Patton, and R. Wright for use of their land and facilitiesat Red Rock. This study was supportedby grantsfrom the JosselynVan Tyne Fund of the AmericanOrnithologists' Union and the Student ResearchAllocations Committee of the Graduate Student Associationof the University of New Mexico.

LITERATURE CITED

BENT, A. C. 1939. Life historiesof North Americanwoodpeckers. U.S. Natl. Mus. Bull. 174. BRENOWITZ,G. L. 1977. An analysisof Gila Woodpeckervocalizations. Wilson Bull. in press. COPY,M. L. 1969. Convergentcharacteristics in sympatricspecies: a possiblerelation to interspecific competitionand aggression.Condor 71: 222-239. , ANDJ. H. BROWN. 1970. Characterconvergence in MexicanFinches. Evolution 24: 304-310. GILMAN, M. F. 1915. Woodpeckersof the Arizona lowlands. Condor 17: 151-163. HINDE, R. A. 1956. The biologicalsignificance of the territoriesof birds. Ibis 98: 340-369. JOHNSON,N. K. 1963. Biosystematicsof sibling species of flycatchersin the Empidonaxhammondii- oberholseri-wrightiicomplex. Univ. California Publ. Zool. 66: 79-238. mESSEL,B. 1957. A studyof the breedingbiology of the EuropeanStarling in North America.Amer. Midl. Naturalist 58: 257-331. KILHAM,L. 1958. Pair formation,mutual tapping and nesthole selection of Red-belliedWoodpeckers. Auk 75: 318-329. 1959. Territorial behavior of wintering Red-headedWoodpeckers. Wilson Bull. 70: 347-358. 1961. Reproductivebehavior of Red-belliedWoodpeckers. Wilson Bull. 73: 237-254. 1963. Food storingof Red-belliedWoodpeckers. Wilson Bull. 75: 227-234. 1969. Reproductivebehavior of Hairy Woodpeckers,part 3. Agonisticbehavior in relation to courtshipand territow. Wilson Bull. 81: 169-183. LANYON, W. E. 1956. Territow in meadowlarks. GenusSturnella. Ibis 98: 485-489. 58 GENEL. BSrNOWIWZ [Auk,Vol. 95

LAWRENCE,L. DE K. 1967. A comparativelife-history of four speciesof woodpeckers.Ornithol. Monogr. 5: 1-156. MURRAY,B. G. 1971. The ecologicalconsequences of interspecificterritorial behavior in birds. Ecology 52: 414-423. ORIANS,G. H., AND m. F. WILLSON. 1964. Interspecificterritories of birds. Ecology45: 736-745. ROHWER,S.A. 1973. Significanceof sympatryto behavior and evolutionof Great Plains Meadowlarks. Evolution 27: 44-57. SELANDER,R. K., AND D. R. GILLER. 1959. Interspecificrelations of woodpeckersin Texas. Wilson Bull. 71: 107-124. SHORT, t. t. 1971. Systematicsand behavior of some North American woodpeckersgenus Picoides (Aves). Bull. Amer. Mus. Nat. Hist. 145: 1-118. SIMMONS,m. E. t. 1951. Interspecificterritorialism. Ibis 93: 407-413. STICKEL,D. W. 1965. Territorial and breedinghabits of Red-belliedWoodpeckers. Amer. Midi. Natu- ralist 74: 110-118. STOKES,A. W. 1962. Agonistic behavior among Blue Tits at a winter feeding station. Behaviour 19: 118-138.