Interspecific Competition and Social Behavior in Violet-Green Swallows

606 ShortCommunications [Auk, Vol. 107

Interspecific Competition and Social Behavior in Violet-green Swallows

JEFFREYD. BRAWN' Departmentof BiologicalSciences, Box 5640, Northern Arizona University, Flagstaff,Arizona 86011 USA

Resourcecompetition can lead to aggressiveor ter- swallowaggression toward female Western Bluebirds ritorial behavioramong conspecifics or heterospecif- aswell (e.g. Prescott1982). Encounters commonly in- ics(Orians and Willson 1964).The demographiccon- volved one or more swallowsflying towards,but not sequencesof these interactions may also influence striking, a bluebird perchedon or near a box. Blue- other life-history traits(Thornhill 1987).Competition birds remained perched, made frequent "rushing" for nest sitesamong cavity-nestingbirds is a potent flightsat oncomingswallows, or left the vicinity. In selective force that often prevents sexually mature 9 instances, as swallows "harassed" a bluebird, other individuals from breeding (Morse 1980). In popula- swallowsentered disputedboxes and depositednest tions of cavity nestersthat breed in north-centralAr- material(pine needles)or removedbluebird nestma- izona (see Brawn and Balda 1988), I observed inter- terial (grass).I never observedswallow aggression specificinteractions between Violet-green Swallows toward bluebirds away from nest boxes. (Tachycinetathallasina) and Western Bluebirds (Sialia Swallows were not marked individually, but I es- mexicana),whose abundancesare limited by nest-site timated groups to range from 4 to 12 individuals. availability.Here I describethese encounters and dis- Thesegroups always included two or more brightly cussthe possibleinfluence of interspecificresource colored males and two or more dully colored indi- competitionon socialbehavior in Violet-green Swal- viduals(probably females, but the phenologyof swallows. low plumagecharacteristics is not reliablyknown [Pyle In 1980, I installed 60 nest boxes on each of two et aL 1987]).Solitary swallows were all brightly col- 8.5-haareas in ponderosapine (Pinusponderosa) forest ored males. habitat where natural cavities were scarce(Brawn and Solitary swallowsnever displacedbluebirds (n = Balda1988). These areas are locatedsouth of Flagstaff, 15), but groupsdrove bluebirdsaway from a box in Arizona, in the U.S. Forest Service Coconino National 61% of the observedinteractions (n = 59). The pro- Forest, Coconino County. Swallows and bluebirds portionof groupinteractions that displacedbluebirds within each area used nest boxes almost exclusively was significantlygreater than that for bluebird vs. throughout the study (1981-1984, see Brawn 1988). solitaryswallow encounters(Chi-square test for pro- Neither speciesis a permanentresident, but bluebirds portions, X2 = 24.2, df = 1, P < 0.001). Of the 33 arrive and begin to nestabout a month beforeswal- disputednest boxes, 11 (33%)were usedby swallows, lows. bluebirds remained and nested in 18 (55%), and 4 I observedinterspecific interactions along transects (12%)were not subsequentlyused by either species. (5 per area) that permitted close observation of all One memberof sixdisplaced bluebird pairs was color- boxesduring each visit to an area.These observations banded,and two of thesepairs nested in other boxes. were made during the morning (0530-1100)or late The rate of interspecificinteractions per hour of afternoon(1500-1830) from early April through mid- field observationincreased from 1981through 1984. June. Upon detecting an interspecificencounter, I This trend was concomitant with annual increases in observeda disputed nest box until only one species numbersof boxesoccupied by bluebirdswhen swal- was present for longer than 15 min. Beginning in lows arrived at the study areas (Spearmanrank test, early May, I periodically inspectedall nest boxesto p = 0.64, df = 6, P < 0.10, samplesize based on number determinenesting activity. Boxes with nestswere in- of breedingseasons per areawhen both specieswere spectedtwice weekly throughoutthe nestingcycle. present).At least40% of the boxeson eacharea were I observed 74 interactions between swallows and unusedeach year, but thesewere probablyunsuitable bluebirds at 33 different nest boxes. Bluebirds had for swallowsbecause they were either too low or in begunnesting in all the disputedboxes and therefore densevegetation (see Brawn 1985). A relationship alwayshad priority. Fifty-nine of the encounters(80%) between variation in availability of contestablere- involved a single adult male bluebird and four or sourcesand frequencyof interspecificinteractions has more swallows; the remainder were between a male been reported in other cavity nesters(Minot and Per- bluebird and one swallow. Others have observed rins1986), among nectarivorous birds (Stiles and Wolf 1970),and among scorpionflies(Thornhill 1987). Within a breeding season,swallow aggressionwas ' Presentaddress: Smithsonian Tropical Research commonlydirected at bluebird pairs in the early or Institute,Box 2072, Balboa, Republic of Panama.Cor- preincubationstages of the breeding cycle. On av- respondenceaddress: Smithsonian Tropical Research erage,30% of the bluebirdnests were moreadvanced Institute, APO, Miami 34002-0011 USA. when swallowsbegan to selectnest sites.I never ob- July1990] ShortCommunications 607 served attemptsto usurp these boxes(but seePrescott Slagsvoid1978) would allow experimentalevaluation 1982).Field experimentswith other specieshave also of the relative importance of intra- vs. interspecific demonstrateda positive associationbetween pheno- resourcecompetition in avian social behavior. logical overlap in breeding activity and frequencyof I thank D. Enstrom,K. Johnson,B. G. Murray Jr., interspecificinteractions at nestsites (Slagsvoid 1978, T. D. Pitts, S. K. Robinson, K. Sieving, and B. Stutch- Gowaty 1981). Gwinner et at. (1987) suggesteda prox- bury for constructivecomments on earlier drafts. R. imate explanation; simple physical associationof a P. Baldaprovided usefuladvice throughout the study. male cavity nesterwith a nest site can increasetestes sizeand plasmalevels of luteinizing hormonewhich, LITERATURE CITED in birds, is associatedwith aggressiveor territorial behavior.The magnitudeof this effect increasesas an BRAWN,J.D. 1985. Population biology, community individual is associated with a nest site. Thus, swallow structure, and habitat selection of birds in pon- aggressiontowards bluebirds may be moresuccessful derosa pine forest habitat. Ph.D. dissertation, againstrelatively late breedersthat offer lower re- Flagstaff,Northern Arizona Univ. sistence. 1988. Selectivity and ecological conse- I suggestthat socialbehavior in Violet-green Swal- quencesof cavity nestersusing natural vs. arti- lows enhancesaccess to a vital resource.Interspecific ficial nest sites. Auk 105: 789-791. dominanceis often determinedby body size (Murray --, & R. P. BALDA. 1988. Population biology of 1981)and the averagemass of male Violet-green Swat- cavitynesters in northern Arizona:Do nestsites lows (œ= 14 g) is approximatelyhalf that of male limit breeding densities?Condor 90: 61-71. Western Bluebirds (œ= 29; Dunning 1984). Forming DVNNING,J. B., JR. 1984. Bodyweights of 686 species groupsto overcomeresource defense by larger-sized of North American birds. West. Bird Band. Assoc. heterospecificshas been observedamong other bird Monogr. 1: 1-37. speciesthat competefor accessto food (Murray 1981) GOWAT¾,P.A. 1981. Aggressionof breeding Eastern and amongcoral-reef fish (Robertsonet al. 1976).Ap- Bluebirds (Sialia sialis) toward their mates and proximately25% of all swallow nests(n = 47) were modelsof intra- and interspecificintruders. Anita. in boxespreviously occupied by bluebirds. Thus, so- Behav. 29: 1013-1027. cial aggressiontowards heterospecificsis not man- GWINNER,H., E. GWINNER,& J. DITTAMI. 1987. Effects datory to obtain a nest site. Nonetheless,interspecific of nestboxes on LH, testosterone, testicular size, competition(exploitative and interference)appears and the reproductive behaviour of male Euro- particularlystrong for swallowsin north-centralAr- pean Starlingsin spring. Behaviour 103: 68-82. izona becausethey arrive at breeding habitat and LOMBARDO, M. 1986. Attendants at Tree Swallow commence reproduction after all other sympatric nests.I. Are attendantshelpers at the nest?Con- speciesof cavity nesters(n = 5; see Brawn and Balda dor 88: 297-303. 1988). MINOT, E. O., & C. M. PERRINS.1986. Interspecific I frequently observed2 or more individuals enter interference competition-nestsites for Blue and active swallow nests, but I could not ascertain wheth- Great tits. J. Anita. Ecol. 55: 331-350. er or not all birds actually assistedthe nest attempt. MORSE,D. H. 1980. Behavioral mechanisms in ecol- The question of whether the swallow groups that ogy. Cambridge, Massachusetts,Harvard Univ. engaged in aggressiveinteractions were temporary Press. aggregationsor persistentsocial units is unanswered. MURRAY,B. G., JR. 1981. The origins of adaptive If Violet-green Swallows are true cooperativebreed- interspecificterritorialism. Biol. Rev. 56: 1-22. ers, my observationsmay indicate a rare example of OR•^NS,G. H., & M. F. WILLSON.1964. Interspecific interspecificresource competition that affected the territories in birds. Ecology45: 736-745. evolution of a resource-based social system (see ROBERTSON,D. R., H. P. A. SWEATMAN,E. A. FLETCHER, Thornhill 1987). Alternatively, Tree Swallows (T. bi- •r M. G. CLELAND.1976. Schooling as mecha- color)at active nests include a breeding pair and un- nism for circumventing the territoriality of com- related nest attendants that do not assist in repro- petitors.Ecology 57: 1208-1220. duction, but seek to usurp nests from conspecifics PRESCOTT,H.W. 1982. Using paired nesting boxes (Lombardo1986). Thus, socialbehavior in Violet-green to reduceswallow-bluebird competition. Sialia 4: Swallows could be functionally cooperativein the 3-7. contextof interspecificcompetition but, if successful, PYLE, P., S. N. G. HOWELL, R. P. YUNICK, & D. F. DE- can lead to competitive interactions among conspe- SANTE. 1987. Identification guide to North cifics. Data on the identity, relatedness,and role of American passerines.Bolinas, California, Slate individuals at nestscompared with thosethat partic- Creek Press. ipate in encounters with bluebirds would discrimi- SL^CSVOLD,T. 1978. Competition between the Great nate between these possibilities.Furthermore, ma- Tit Parusmajor and the Pied FlycatcherFicedula nipulating the availability of usable nest sites and hypoleuca:an experiment. Ornis Scandinavica9: local abundancesof interspecific competitors (sensu 46-50. 608 ShortCommunications [Auk,Vol. 107

STILES,F. G., & L. L. WOLF. 1970. Hummingbirdter- tra- and interspecificcompetition in scorpionfly ritoriality at a tropical flowering tree. Auk 87: mating systems.Am. Nat. 130: 711-729. 467-491. THORNHILL,R. 1987. The relativeimportance of in- Received2 October1989, accepted 17 January1990.

Hurricane Damage to Red-cockadedWoodpecker (Picoides borealis) Cavity Trees

R. TODD ENGSTROM • AND GREGORY W. EVANS 2 Departmentof BiologicalScience, Florida State University, Tallahassee,Florida 32306 USA

The Red-cockadedWoodpecker (Picoides borealis) is ern ThomasCounty, Georgia, contains one of the larg- a federally endangeredspecies, which inhabitspine est remaining old-growth longleaf pine populations. forestsin the southeasternUnited States(Ligon 1970). As part of a long-term study of life history dynamics Red-cockaded Woodpeckers excavate cavities for (Platt et al. 1988),all pines taller than breastheight roostingand nestingin mature,living pine trees(Ba- (1.5 m) on a 46-ha section of the Wade Tract were ker 1971), preferably longleaf pine (Pinuspalustris) given identification numbers on metal tags in 1979. infectedby redheartdisease (Fomes pini) (Lennartzet The diameter at breast height (DBH) of all tagged al. 1983).Excavation of a single cavity may require a treeswas recordedin 1979,and the survivingtagged year or more, and it has been suggestedthat cavity trees were remeasuredin 1984. Tagged trees were treesare the primaryecological constraint that shaped censusedannually for mortality through 1987. the evolution of the cooperativebreeding systemof Red-cockadedWoodpeckers are common in the the Red-cockadedWoodpecker (Lennartz et al. 1987, Wade Tract (Engstrom 1982). Some Red-cockaded Walters et al. 1988). Woodpecker nest trees were identified in 1979, but Natural disturbanceis an important component no systematicinventory of cavity trees was done be- of the southeasternpine forestsinhabited by the Red- fore the stormsin autumn 1985.In the spring of 1986, cockadedWoodpecker (Christensen 1977). Frequent Engstromvisually inspectedtrees in the Wade Tract surfacefires maintain the open habitat favored by during a study of hurricane damage to the entire Red-cockadedWoodpeckers (Ligon et al. 1986).High forest. All Red-cockadedWoodpecker cavity trees in winds and lightning associatedwith frequent sum- the 46-hastudy site were locatedat this time. Among mer thunderstormsand, less frequently, tornadoes nearly 7,800 marked and measuredtrees that were and hurricanes are important sourcesof mortality standing before the stormsin the autumn of 1985, 33 among larger pines (Platt et al. 1988).In late summer were Red-cockadedWoodpecker cavity trees. These and autumn 1985, three major storms struck the cavity trees ranged in size from 30 to 69.2 cm DBH northeasternGulf of Mexico: tropical storm Juan,and (œ= 49.0 cm; Fig. 1). hurricanes Elena and Kate. The combination of heavy Of 33 Red-cockadedWoodpecker cavity trees, 8 (24%) rain that saturatedthe soil and strongwinds damaged were killed during autumn 1985 (4 during hurricane some forests in northern Florida and southern Geor- Kate). Trunks of six of the eight trees snappedoff at gia. Trees were blown over and tree trunks snapped. the site of the woodpecker cavity; trunks of the re- During hurricane Kate, downbursts of wind gusting maining two trees snapped below the cavity. Five of to 160 km per hour were estimatedfor Leon County, the eight Red-cockadedWoodpecker cavity treesthat Florida (Clark 1986). We report the damage caused were snappedoff had rotten heartwood most likely by these storms(mostly hurricane Kate) to Red-cock- causedby redheart disease.The remaining three trees adedWoodpecker cavity trees in an old-growthlong- did not have obviouslyrotten heartwood at the point leaf pine forest in southern Georgia. of the snapped trunk. Four additional cavity trees The Wade Tract, an 80-ha conservation easement suffereda lossof large limbs or a gash in the trunk managed by Tall Timbers ResearchStation in south- that may increase the chance of mortality over the next decade. On the Wade Tract, age (measuredwith increment • Present address: Tall Timbers Research Station, cores)and size (DBH) of 399 randomly selectedlong- Route 1, Box 678, Tallahassee, Florida 32312 USA. leaf pine treesare highly correlated(Platt et al. 1988). 2Present address: Department of PublicHealth Sci- Trees of approximately 30 cm DBH in this random ences,The BowmanGray Schoolof Medicine,Wake sample ranged from just under 40 to over 80 years ForestUniversity, 300 SouthHawthorne Road, Win- old, and treeswith diametersof approximately60 cm ston-Salem, North Carolina 27103 USA. DBH ranged from roughly 180 to 240 years old. A1-