January1984] ShortCommunications 193

Avian Methods of Feeding on Burserasirearuba () Fruits in Panama

JILL M. TRAINER• AND TOM C. WILL Museumof Zoology,University of Michigan,Ann Arbor,Michigan 48109 USA

We observed 26 species of birds and noted the The 12-m-tall tree grew in a cleared field l0 m from feeding method usedby each when taking fruits of dry deciduousforest. The fruits, fleshy egg-shaped a single Burserasimaruba tree in Panama(Table 1). The capsulesabout 1 cm in length, had a leatherygreen lcterinae, representedby five species,showed the skin (pericarp) over a thin layer of bright pink aril greatestdiversity in feeding methodsand differed covering one or two one-seeded nutlets. The abun- from one another in feeding rate, in handling time dant fruits were conspicuousbecause the tree we ob- (Table 2), and probably in dispersalconsequences for served had lost most of its leaves. B. sirearuba.The two oriole species,which depend Birds handled fruits in three ways, each having least on fruit in the diet (Bent 1958), also fed least different consequencesfor feeding efficiencyand seed efficiently and probably provided the poorestseed dispersal: (1) fruit skin removed, nutlet with aril dispersal. swallowed, and nutlet defecatedor regurgitated lat- Burserasimaruba, or gumbo-limbo tree, is a wide- er; (2) fruit skin removed, fruit picked and manipu- spreadtropical species common in advancedsecond- lated in the mouth to remove the aril, and nutlet ary growth. It flowersin Panamafrom mid-March to dropped beneath the tree; and (3) fruit skin removed, mid-June;fruits are presentthroughout the year, but aril eaten, and nutlet left on the tree. mature fruits are most abundant in the early dry sea- Most visitorsswallowed the fruits, but somespecies son around January(Croat 1978: 943). We made ob- consistentlydropped the nutlets under the tree or servationsbetween 0645 and 0830 on 4-8 February left them on the tree. During this study five bird 1982 near Kobbe Beach, Panama (60 m elevation). speciesappeared to be the most effective seed dis- perserson the basis of feeding method, number of •Presentaddress: 3000 SpargerRoad, Durham, North Carolina 27705 individualsvisiting the tree, number of daysseen at USA. the tree, and foraging rate:Zarhynchus wagleri, Cacicus

TABLE1. Bird speciesand methodsof feeding on fruits of Burserasimaruba in Panama.

Feeding method (seetext) Species 1 2 3 Crotophagaani (Smooth-billedAni) Pteroglossustorquatus (Collared Aracari) Melanerpesrubricapillus (Red-crowned Woodpecker) Myiarchuspanamensis (Panama Flycatcher) Megarhynchuspitangua (Boat-billed Flycatcher) Myiozetetessimilis () Myiodynastesmaculatus (Streaked Flycatcher) Tyrannusmelancholicus () Tityra semifasciata() Chiroxiphialanceolata (Lance-tailed Manakin) Turdusgrayi (Clay-coloredRobin) Vireofiavifrons (Yellow-throated ) Vireoolivaceus (Red-eyed Vireo) Vermivoraperegrina (Tennessee Warbler) Dendroicacastanea (Bay-breasted Warbler) Protonotariacitrea (Prothonotary Warbler) Thraupisepiscopus (Blue-gray Tanager) Tangarainornata (Plain-colored Tanager) Dacniscayana (Blue Dacnis) Cyanerpescyaneus (Red-legged Honeycreeper) Pheucticusludovicianus (Rose-breasted Grosbeak) X Zarhynchuswagleri (Chestnut-headed Oropendola) Cacicuscela (Yellow-rumped Cacique) Icteruschrysater (Yellow-backed Oriole) X Icterusgalbula (Northern Oriole) X Scaphiduraoryzivora (Giant Cowbird) X 194 ShortCommunications [Auk,Vol. 101

TABLE2. Handling times and feeding ratesfor five icterine speciesfeeding on Burserasimaruba fruits in Panama.

œbody weight (g) œhandling time œfeeding rate Species (s) (fruits / min) Male Female Zarhynchuswagleri 1.5 5.7 211.4 115.3 Cacicus cela 1.4 5.6 75.9 65.5 Scaphiduraoryzivora 9.4 2.4 204.5 171.5 Icterusgalbula 16.6 1.3 35.2 33.9 Icteruschrysater 18.2 1.4 53.5 48.3

cela,Megarhynchus pitangua, Vireo olivaceus, and Myi- votes, because adults eat some insects and feed in- odynastesmaculatus. A squirrel, Sciurusvariegatoides, sectsto their young.Obligate frugivores are believed also took fruits but may have eaten the seeds. to differ from opportunisticfrugivores in depending The five icterine speciesdiffered from one another more completely on fruit in the diet and, as a result in feeding method,feeding rate, handling time (Ta- of coevolution,in providing more reliable dispersal ble 2), and probablyin quality of dispersalprovided. of seeds(McKey 1975,Howe 1977).When applied to Oropendolasand caciques(Z. wagleriand C. cela) oropendolasand caciques,the term "opportunistic swallowedthe fruits and had the shortesthandling frugivore" is somewhatmisleading, becausethese times and the highest feeding rates.At least someof speciesrequire and actively seek large quantitiesof the intact nutlets passfreely through the guts of Z. fruit in their diet. This dependenceon fruit, even wagleriand C. cela,which defecatethe nutlets when though not complete,has probablyinfluenced the caughtin mist nets. Icterusgalbula removed the pulp foragingadaptations and dispersion patterns of adults. from the nutlet and usuallydropped the nutlets un- They meet their dietary requirementsby efficient der the tree, while Icteruschrysater usually left them handling of fruit such as B. simaruba.The orioles, on the tree. Theseorioles had the longesthandling however,do not seemto depend on large quantities times. Their feeding rateswere the lowest,partly be- of fruit (Bent 1958). It may be that they meet their causethey were subordinateto the other icterines requirementsby opportunisticallyeating small quan- and were often supplantedbefore getting a chance tities of fruit as encountered.The low feeding rates to eat the fruit. Although the two orioles probably on B. simaruba,if they also reflect how other typesof provided little dispersalof seeds,their exploitation fruit are handled,suggest that efficienthandling of of fruit had relatively little impacton the tree due to fruit has not been important for orioles. their low feeding rates and subordinatestatus. Sca- The difference between "opportunistic" and "ob- phiduraoryzivora usually dropped the nutlets under ligate" frugivoresin the degree to which they are the tree and had intermediate handling times and specialistson fruit and the quality of seeddispersal feeding rates. Swallowing fruits whole rather than they provide may not be as great as once believed removing the pulp from the nutlet allowed birds to (Wheelwrightand Orians1982). Oropendolas and ca- forage more efficientlyon B. simarubafruit and may ciquesmay differ lessin these respectsfrom "obli- incidentally have increasedthe dispersalquality pro- gate" frugivores than they do from the orioles. vided. We were supportedin this work by a grant from Feeding method, handling time, and feeding rate the H. H. Rackham School of Graduate Studies, Uni- in the Icterinaewere relatedto body size;the largest versity of Michigan, while we were visitors of the birds swallowedfruits and had the lowest handling SmithsonianTropical ResearchInstitute. We thank times and the highest feeding rates. The failure of RobertB. Payne, Peter R. Grant, and Barry E. Ham- the smallerIcterus species to swallow fruits, however, mel for commentingon the manuscript. was probablynot becausethe fruits were too big, as many even smaller birds swallowedthe fruits (e.g. LITERATURE CITED Vireoolivaceus). The different methodsused by icter- ines to handle Burserasirnaruba are probably not de- BEECHER,W. J. 1951. Adaptationsfor food-getting termined by body size, but insteadreflect differences in the American blackbirds. Auk 68: 411-440. amongthe speciesin how food is handledin general. BENT, A. C. 1958. Life histories of North American Members of the Icterinae exhibit particularly high blackbirds,orioles, tanagers, and allies.Bull. U.S. diversity in feeding adaptationsand diet (Beechef Natl. Mus. 211: 28-466. 1951). CROAT, T.B. 1978. Flora of Barro Colorado Island. Within the Icterinaethe largestdifference in feed- Stanford, California, Stanford Univ. Press. ing efficiencyand dispersalquality occurredbetween HOWE,H. F. 1977. Bird activity and seed dispersal the oropendolas/caciques and the orioles.All of these of a tropicalwet foresttree. Ecology58: 539-550. species are often considered opportunistic frugi- McKE¾,D. 1975. The ecologyof coevolvedseed dis- January1984] ShortCommunications 195

persal systems.Pp. 159-191 in Coevolution of persal, problems of terminology, and constraints anirnals and (L. E. Gilbert and P. H. Ra- on coevolution. Amer. Natur. 119: 402-413. ven, Eds.). Austin, Texas, Univ. Texas Press. WHEELWRIGHT, N. t., & G. H. ORIANS. 1982. Seed Received17 January1983, accepted 6 September1983. dispersal by animals: contrastswith pollen dis-

Evidence of Aggressive Behavior in Female Blue Grouse

RICHARD A. LEWIS Departmentof Zoology,University of Alberta,Edmonton, Alberta T6G 2E9,Canada

Historically, studiesof territoriality and other forms landed in the same area. One cackled, but as I ap- of spacing behavior have emphasized interactions proached the calling stopped, and the unbanded fe- between males, with little attention being given to male flushed far down a hill when I disturbed her. I the study of similar behaviors in females.Recently, could not relocate the marked fernale. however, work with some tetraonids has demonstrat- The banded female was a known adult and had ed that femalesdo respondaggressively towards one been seen in the area where the interaction occurred another and that these behaviors may relate to the three times before 30 April; she had nested nearby spacingof individuals(Stirling 1968,Herzog and Boag as a yearling in 1981. Later in 1982 she was seen in 1977)and/or the regulationof breedingdensities and the samearea with a brood, and the age of the chicks production (Robel 1970). indicated that her nest hatched on 13 June. The in- Results from laboratory experiments with Blue teractionI observed,therefore, occurred about 13 days Grouse (Dendragapusobscurus) indicate that females before she began laying, that is, at a time when she will attacktheir mirror images(Stirling 1968). In ad- would have been establishinga hornerange and pre- dition, indirect evidence from field studiessuggests paring to breed (Hannon et al. 1979, Harmon 1980). that females may spacethemselves on the breeding An unbanded female in this area began cackling range (Hannon et al. 1982), and Hannon (1980) hy- immediately when I played taped cacklesto a nearby pothesized that the "cackle" call is an aggressivevo- territorial male on 25 April. Most females (over 85%) calization that mediatesthis spacing.Some females on my study area were marked and therefore the have cackled at, and in a few instances have ap- unbanded females seen on 25 and 30 April were proached, tape recordings of cackles(Hannon 1978); probably the same individual. The interaction I ob- at present, however, no documentedcases of females servedon the latter date was possiblyin an area where chasing or attacking other females are available. In the home ranges of these two females overlapped, this note I describethe details of an aggressivein- or, alternatively, the unbanded female could have teraction between two females, which I observed been trying to establish a prenesting home range while conducting studies of Blue Grouse on Hard- (Hannon et al. 1982). No unbanded brood females wicke Island, British Columbia in 1982. were later seen in this immediate area. On 30 April I flushed an unidentified female at Hens did not attack female models when Hannon 1705 and she flew to an area of tall trees approxi- (1980) conductedplayback experimentsin the field. mately 75 m away. Twenty minutes later two females Therefore, she postulated that the mechanism for began uttering "whinny" calls (Stirling and Bendell spacingof femalesis a combinationof warning calls 1970) in the area where the female landed. The fe- and rnutual avoidance rather than one of overt ag- malescontinued calling vigorously,with most of the gression.Although the encounter I observeddid not calls being whinnies; a few cackle calls also were involve direct contact, it does demonstrate that fe- given. As I approached,one flew in my direction and males do interact aggressively and that both the landed in a tree a few meters away. Within seconds whinny and cacklecall are usedin suchinteractions. the other female flew toward the first hen and landed Presently,the relative importanceof mutual avoid- in a tree 10 m from her. The second female was band- ance and overt aggressionin spacing fernalescannot ed, but the first was not. Both began cacklingat each be evaluated, however, becausefemale spacing be- other, with the banded one appearing to be the ag- havior has been irnplicated only recently as being gressor.The unmarked hen cackled softly and infre- irnportant in regulating breeding densities of Blue quently and walked slowly along a branch. The Grouse (Hannon 1980, Hannon et al. 1982), and as banded female cackled, flew to within 5 m of the yet researchon this problern has been limited. other female, and walked toward her. When the I gratefully acknowledgethe Bendicksonand Mur- banded female was 3-4 rn away the unmarked hen ray families for the hospitality they extended me flew 40 m to the northwest and landed in another during my studieson Hardwicke Island. Permission tall tree. Again the banded female flew after her and to work there was granted by Crown Zellerbach of