j RaptorRes. 34(3):196-202 ¸ 2000 The Raptor ResearchFoundation, Inc.

THE FOOD HABITS OF SYMPATRIC FOREST- DURING THE BREEDING SEASON IN NORTHEASTERN GUATEMALA

RUSSELL THORSTROM ThePeregrine Fund, 5666 WestFlying Lane, Boise,ID 83709 U.S.A.

ABSTRACT.--Thefood habitsof Barred (Micrasturruficollis) and CollaredForest-Falcons (M. semitorquatus) were studiedin Tikal NationalPark, Guatemala.On a numericalbasis for 405 identifiedprey for Barred Forest-Falcons,lizards (Anolis spp., Ameiva or Cnemidophorusspp., Laemanctus spp., and Corytophanesspp.) were the most numerous prey type comprising 61.5% of the diet. For Collared Forest-Falcons,on a numericalbasis of 170 identifiedprey, represented the greatestproportion at 45.9%. On a biomassbasis, lizards (37.3%) and (36.8%) were equallyimportant in the diet of Barred Forest- Falconsbut, for CollaredForest-Falcons, mammals (47%) and birds (45.4%) were the mostimportant prey.Food-niche overlap was 0.49 betweenthe two forest-falconsand prey that overlappedwere mice, rats,bats, birds (Momotusspp., Dendrocincla spp.), and lizards ( Corytophanesspp.). The wider food breadth of the Collared Forest-Falconwas probably attributable to the greaterdiversity of speciesin its diet. The Collared Forest-Falconis approximately3 times the size of Barred Forest-Falconsbut the mean weightof its prey (MWP) was10 timesgreater (• = 239 g) than that of BarredForest-Falcons (/= 24 g). KEYWORDS: BarredForest-; Micrastur ruficollis; Collared Forest-Falcon; Micrastur semitorquatus; food habits;niche overlap; niche breadth.

H•tbitos alimenticiosde dos halconesde bosquesimpatricos durante la estaci6nreproductiva en el noreste de Guatemala

RESUMEN.--Losh/tbitos alimenticios de Micrasturruficollis y Micrastursemitorquatus rueton estudiadosen Parque Nacional Tikal, Guatemala.En una base num6rica de 405 presasidentificadas para Micrastur ruficollis,las lagartijas(Anolis spp., Ameivao Cnemidaphorusspp., Laemanctusspp., y Corytaphanesspp.) rueton el tipo de presam/rs numeroso o seael 61.5% de la dieta. Para Micrastursemitorquatus, en una basenum6rica de 170 presasidentificadas, los mamiferosrepresentaron la proporci6nmayor con el 45.9%. En relaci6na la biomasa,las lagartijas(37.3%) y aves(36.8%) fueron igualmenteimportantes en la dieta de Micrasturruficollis, pero para Micrastursemitorquatus, los mamiferos(47%) y aves(45.4%), fueron las presasmgs importantes.El traslapedel nicho alimenticio rue de 0.49 entre los dos halcones de bosquey las presasque se traslaparonrueton ratones,ratas, murci•lagos, aves (Momotus spp., Den- drocinclaspp.), y lagartijas (Corytophanesspp.). El espectrom/rs amplio de la dieta de Micrastursemitor- quatusfue probablemente atribuible a la mayor diversidad de especiesde aves en su dieta. Micrastur semitorquatuses 3 vecesel tamafio de Micrasturruficollis pero su peso medio fue 10 vecesmayor (• = 239g) que el de Micrasturruficollis (/= 24 g). [Traducci6n de C•sar M/trquez]

Neotropical birds of prey are poorly known, es- its of this genus is given by Robinson (1994), but pecially the forest-dependentspecies which are in- it too is limited to incidental observations. conspicuousin their habits. The secretive forest The Barred Forest-Falcon(Micrastur ruficollis) is raptors of the genusMicrastur are among the least- perhaps the most common raptor in Neotropical studied raptors and most accountsof their diets forests.It has the widest distribution of any forest- come from stomach contents of museum speci- falcon, occurring from southeastern Mexico to mens or incidental observations(Dickey and van northern , ,and east through Rossem 1938, Friedmann 1948, Smith 1969, Izawa and the Guianas (Brown and Amadon 1989, 1978, Mader 1981, Willis et al. 1983, Mays1985, del Hoyo et al. 1994). It rangesfrom humid low- Trail 1987, Rappole et al. 1989, Thorstrom et al. land and foothill foreststo higher subtropicaland 1990). The most detailed account of the food hab- montane forestsreaching its limit near 2500 m. In-

196 SEPTEMBER2000 FOOD HABITS OF FOREST-FALCONS 197 formation on the diet of the Barred Forest-Falcon were made using 7-10X binocularsat distancesof 25-50 suggeststhat it feedsmainly on lizards (Thorstrom m. During the breeding season, observationsof prey items were recorded during prey deliveries and away et al. 1990, Thorstrom 1993, del Hoyo et al. 1994). from nests during radiotracking sessions.All prey was The Collared Forest-Falcon (M. semitorquatus) identified to the most accurate taxonomic level possible also has a broad distribution, ranging from central with the exceptionof amphibiansand insects,which were Mexico to eastern Bolivia, northern Argentina, and not identifiable to the specieslevel and were assignedto larger taxonomic groupings. The resulting tabulauon Paraguay (Brown and Amadon 1989). It occupies produced a total of 37 prey categoriesfor both species. dense primary and secondaryforests from sealevel Only observedprey delivered and captured were includ- to 2500 m. A recent sightingin Texas (Lasleyet al. ed in biomassestimates to avoid possiblebias from prey 1994) extended its northern distribution to the found in nests(Snyder and Wiley 1976, Wiley and Wiley southwestern U.S. Food of the Collared Forest-Fal- 1981, Marti 1987). Anolislizards were separatedin small (<20 cm) and large catagories(>20 cm). con includes birds, mammals, lizards, snakes, and To estimatemean weight of prey (MWP), I multiplied insects (Brown and Amadon 1989, Thorstrom each prey item by its averageweight (Table 1), summed 1993). the products and divided the sum by the total number In this paper, I compare the diet of Barred For- of prey observed. weightsfollow Emmons and est-Falcons and Collared Forest-Falcons based on Feer (1997), bird weightscome from Smithe (1966) and Dunning (1993), and reptile weightswere taken in the severalyears of nest observationsof prey deliveries, field. and direct observationsat and awayfrom nestsdur- Food-nichebreadths (FNB) were calculatedusing Lev- ing breeding seasonsfrom 1988-92 in northeast- ins' (1968) equation:FNB = 1/EP,j 2, where P• is the proportion of the ith prey categoryof speciesj. For com- ern Guatemala. My objectiveswere to compare parisonamong raptorswith different number of prey cat- prey frequency and biomass and to assessthe egories,a standardizedniche breadth value (FNBs) was amount of overlap in diet among the two species also calculated as follows: FNBs = (FNB - 1)/(n - 1), and compare food-nicheparameters and differenc- where n is the number of prey categories(Levins 1968) es as potential mechanismsfor coexistenceof these Niche overlapwas calculated using Schoener's (1970) in- dexof symmetricaloverlap: overlap = 1 - («)(•[P•j - two forest-falcons. P•a[),where P• is the proportionof the ith preycategory for speciesj and k. Linton et al. (1981) found this overlap STUDY AREA AND METHODS formula to be the only index that accuratelymeasures I studied Barred and Collared Forest-Falcons in Tikal real overlapbetween 7-85%. National Park, Pet6n, Guatemalafrom 1988-92. The park The Collared Forest-Falconis the largest of the two encompasses576 km9 in northeasternGuatemala and its specieswith a body massof 467-511 g for males (Dickey center lies at 17ø13'N,89ø36'W. Vegetation in the park is and van Rossem1938) and 556-750 g for unknownsexes semideciduoustropical forest with lowland rolling hills (Haverschmidt1968). Males I weighed averaged587 ñ ranging from 200-450 m elevation. 17.6 g (ñSD, range = 563-605 g, N = 4) and females Schulze and Whitacre (1999) described several forest averaged869 g ñ 63 g (range = 792-940 g, N= 6) typesthat occur along topographicaldrainage, soil type, Barred Forest-Falconsaveraged 167.8 _+10.6 g (range = and moisture gradients within the park. The two ex- 144-184 g, N = 25) for malesand 233.2 g -+ 23.9 (range tremes of this forest-typecontinuum are upland or high- = 200-322 g, N = 34) for females. ground forests (tall, semi-evergreenforests on well- RESULTS drained, shallowsoils) and "bajo" forests(low in stature, with open canopy and dense understory,occurring in Barred Forest-Falcon. I recorded lizards (Anohs low-lyingsites of deep, clay-richsoils subject to seasonal spp., Ameivaspp. or Cnemidophorusspp., Laemanctus flooding and drought). Tikal National Park is covered spp., and Corytophanesspp.), birds (Momotusspp., mostlyby unbroken primary forest,except for someareas where light selectivelogging occurred prior to 1969. Aulacorhynchusspp., Turdusspp., Leptotilaspp., Den- The climate has pronouncedwet and dry seasonswith drocinclaspp., Thryothorusspp., and Tyrannidae), rains usually beginning in May or June and ending by amphibians, mammals, snakes,and insects (Blatti- December. Between 1989-95, monthly precipitation dae) in the diet of Barred Forest-Falconsduring ranged from 1.0 mm in April to 302.5 mm during Sep- tember with an annual mean rainfall of 1309 mm (pers. the nesting season. obs.). Mean monthly temperaturesranged from a low of I observeda total of 600 prey items being deliv- 15øCin January to a high of 35øCin May. ered to females, nesdings, and fledglings from The forest and known forest-falcon territories were 1988-92. On a numerical basis,reptiles were the searcheddaily from February through August to docu- predominantprey comprising61.5% of the diet ment nestingactivity and potential breeding pairs.Nests of Barred Forest-Falconswere observed primarily from (249 prey items), followedby birds 22% (89), in- the ground and thoseof the CollaredForest-Falcon were sects8.2% (33), mammals5.9% (24), and amphib- occasionallyobserved from tree platforms. Observations ians 2.5% (10) (Fig. 1). Nearly one third (195) of 198 THORSTROM VOL. 34, NO. 3

Table 1. Weights used to estimateprey biomassof Barred and Collared Forest-Falconsat Tikal National Park, Guatemala.

WEIGHT PREY (g) SOURCE

Insects Blattaria 1.5 This study Reptiles Anolis<20 cm 3.9 This study Anolislarge >20 cm 13.8 This study Ameivaor Cnemidophorus 25 This study Laemanctus 15 This study Corytophanes 45 This study Birds Crypturellus 440 Smithe 1966 Penelope 600 Smithe 1966 Crax 500 Smithe 1966 Ortalis 450 Smithe 1966 Agriocharis 3000 Smithe 1966 Odontophorus 300 Smithe 1966 Leptotila 160 Smithe 1966 Ciccaba 240 Smithe 1966 Momotus 133 Dunning 1993 Ramphastos 350 Dunning 1993 Pteroglossus 220 Dunning 1993 Aulacorhynchus 150 Smithe 1966 Melanerpes 81 Dunning 1993 Celeus 85 Dunning 1993 Tyrannidae 15 Smithe 1966 Cyanocorax 200 Dunning 1993 Troglodytidae 15 Smithe 1966 Muscicapidae 75 Smithe1966 Mammals Sciurus small 205 Emmons and Feer 1997 Sciuruslarge 400 Emmons and Feer 1997 Artibeus 50 Emmons and Feer 1997 Unidentified bat 20 This study Unidentified mouse (Heteromys) 76 This study,Emmons and Feer 1997 Unidentified rat ( Rattus,Oryzomys, Sigmodon) 150 This study,Emmons and Feer 1997 the items were unidentified, especiallylate in the ly Ameivaspp. or Cnemidophorusspp.), 11 Laemanc- nestling period, becausemale forest-falconsflew ms spp., 5 Corytophanesspp., and 49 unidentified secretivelyinto their nests without calling their lizards. Snakes included 1 coral snake or mimic mates to receive prey, and females flew into the (Lampropeltissp. or Micrurus sp.) and 2 other nestsquickly and directly without vocalizations.It snakes.Eleven of the 267 identified prey (4%) was unlikely, however, that the unidentified prey were frogs (Rana spp. and/or Hyla spp.). Only 21 items differed from those actuallyidentified. The arthropods (8 cockroachesand 13 other items in- most detailed dietary information was obtained cludingspiders and beetles,8% of the diet) were during 1989 when 267 of 380 items delivered to identified.Birds contributed 52 preyitems (19.5 % nestswere identified. Again, most (64.0%, N = of the diet) and included five Blue-crowned Mot- 171) were lizardsand were representedby 57 small mots (Momotusmomota), two flycatchers (Tyranni- Anolisspp., 21 large Anolisspp., 28 teiids (mostlike- dae), two Emerald Toucanets (Aulacorhynchuspra- SEPTEMBER 2000 FOOD I-IABITS OF FOREST-F,•LCONS 199

a)

BarredForest-Falcon (n:267) CollaredForest-Falcon (n=170) ..,•'•:'•'•

'[• Mammals ':Z •. ß Birds -. [] Reptiles ii[] Amphibians [] Insects

Barred Forest-Falcon biomass Collared Forest-Falcon biomass -. , ,•,••' ...... 5,,...•, •,:. .•' ,.' : :. .•'• • Mammals :• ' ß Birds ...••...? ...... - • [] Reptiles .:• '•=• '"•:;•'"'"""•'--•'•,'•-&'7='=':'='=':':'-. • Amphibians .,:•.•:+?=? , ,,•=:4• -..-.=.- ß [] Insects

• ,.,• ,' "•!=••x•='• ...... •..•ii•?=" ..... ; :•..•;;::.:•.•.:.• •,..

Figure 1. Comparisonof the diets of Barred Forest-Falconsand Collared Forest-Falconsas (a) the percent prey of individualsand (b) the biomasscomposition (% weightof prey individuals).

sinus), one Gray-fronted Dove (Leptotilarufaxilla), prey delivered in 1989, five males brought in 3.8 one woodcreeper (Dendrocinclasp.), one Spot- kg (75.7%) and five females delivered 1.2 kg breasted Wren (Thry0th0rusmaculipectus), and one (24.3%) of the biomassduring the breeding sea- Clay-colored Robin (Turdus grayi). Birds taken son. ranged in size from an unidentified warbler (Den- Collared Forest-Falcon. I found squirrels (Sciu- droicasp.) at 9 g to a Gray-frontedDove at 160 g rus spp.), bats (Artibeusspp.), rats (Sigmodonspp.), (Smithe 1966, Dunning 1993). The nine mammals mice (Heter0mysspp.), birds (Crypturellusspp., Pe- I identified represented only 3% of the diet. nelopespp., Crax spp., Ortalisspp., Agriocharisspp., Among them were sevenrodents, one bat, and one Odontophorusspp., Leptotilaspp., Ciccabaspp., Mom- other mammal. The were possiblymem- otusspp., Ramphastosspp., Pteroglossusspp., Aulaco- bers of the genera Heteromysand Oryzomys.Snakes rhynchusspp., Melanerpesspp., Celeusspp., Cyanocor- accountedfor 3 prey itemsor 1.1% of the diet. ax spp., Dendrocolaptidae),snakes (Colubersp.), Biomass estimates were made for 267 identified and lizards (C0ryt0phanesspp.). prey items delivered during the 1989 breeding sea- From 1990-92, 222 prey items were delivered to son. On a biomassbasis, reptiles (37.3%), birds females,nestlings, and fledglingsand 170 of these (36.8%), and mammals(20.2%) comprised94.3% were identified. On a numerical basis, 45.9% were of the estimatedbiomass (Fig. 1). Males delivered mammals(78 preyitems), 34.7% birds (59), 18.8% more prey items and prey biomassthan females reptiles (13 lizardsand 19 snakes),and 0.6% am- during the breeding season.Of the 267 identified phibians (1 frog) (Fig. 1). The 52 unidentified 200 THORSTROM VO•.. 34, NO. 3 prey items were presumed to have been similar to Table 2. Food-nichebreadth, dietary overlap, and esti- those that were identified. In addition, 36 items mated mean weights (g) of prey (MWP) and of birds were given to two fledglings by an extra adult be- (MW) of Barred and Collared Forest-Falconsduring the lieved to be a male. This male specializedin catch- nestingseason. All calculationsbased on prey at the ge- neric or family level. Mean _+SE (N). ing toucansso I calculatedthe diet of Collared For- est-Falcons both with and without this male's contribution. BARRED COLLARED FOOD-NICHE FOREST- FOREST- Prey of Collared Forest-Falconsranged in size PARAMETERS FALCON FALCON from a frog estimatedat 20 g to an OcellatedTur- key (Ag•iocharisocellata) weighing about 3 kg. The Total identified prey two largestprey were the adult female turkey and items 267 170 a young Crested Guan (Penelopepurpurascens). Of Mammal species richness 3 6 the 13 lizards taken, 12 were in speciesbelonging Bird speciesrichness 7 15 to the genus Corytophanes.The 19 snakes I ob- Lizard speciesrichness 5 1 served were most likely colubrids.The 78 mam- 23.7 +-- 2.5 238.9 _+ 18.9 malsidentified included 42 Deppe'ssquirrels (Sciu- (267) (170) rus deppei;190-220 g), 11 Yucatan squirrels (S. MW birds 62.1 _+ 15.3 373.4 +_ 49.5 yucatanensis;420 g), two fruit bats (Artibeusspp.), (52) (59) 14 unidentifiedbats, 7 rat-sizedrodents including FNB 7.9 13.8 the hispid cotton rat ( Sigmodonhispidus), and 2 FNBs 0.33 0.49 mice believed to be spiny pocket mice (Heteromys Dietary overlap 0.49 spp.). Among the 59 birds identified, the most nu- merous were Collared Aracari (Pteroglossustorqua- tus, N = 9), Plain Chachalaca ( Ortalis vetula, N = spp., dominated the Barred Forest-Falcondiet and, 7), Great Curassow(Crax rubra, N = 7), Keel-billed as a result, it had a narrower niche breadth than Toucans (Ramphastossulfuratus, N = 6), Ruddy did the Collared Forest-Falcon. Collared Forest-Fal- Woodcreepers (Dendrocinclahomochroa, N = 4), cons took a higher richnessof bird and mammal Tinamous (Crypturellusspp., N -- 3) , and Brown species (Table 2). The standardized FNB of the Jays (Cyanocoraxmorio, N = 3). Barred Forest-Falcon was lower (0.33) than the In 1990, a third adult forest-falcon,probably a Collared Forest-Falcon(0.49). Dietary overlapbe- male, began deliveringprey items to two young, 4 tween the two forest-falcons was 0.49. Estimated wk after they fledged. We observedthis adult de- MWP captured by Collared Forest-Falconswas sig- liver 36 prey items until 11 weeksafter fledging.It nificantly heavier than that of Barred Forest-Fal- appeared to prefer Keel-billedToucans delivering cons (Table 2). The larger Collared Forest-Falcon 27 toucans, two Collared Aracari, two unidentified capturedlarger avian (/= 373.9 + 49.5 g, +SE, N birds, four squirrels (S. deppeO,and one unidenti- = 59) and mammalian (/= 179 + 12.5, N = 78) fied prey item. Sometimes it delivered two Keel- prey than did the Barred Forest-Falconwhich took billed Toucansa day. When this contributionwas mostlylizards (œ= 13.8 + 0.6, N = 122) and birds included in the overall diet of Collared Forest-Fal- (œ = 62.1 + 4.9, N = 52). cons, the diet was dominated by birds (43.9%, 90 individuals) followed by mammals (40.0%, 82), DISCUSSION reptiles (15.6%, 32), and amphibians(0.5%, 1). In Barred and Collared Forest-Falcons are moder- terms of biomass,this extra adult delivered12.6 kg ately dimorphic with Collared Forest-Falcons3-4 of prey during the post-fledgingperiod. times larger than Barred Forest-Falcons.Optimal Biomass estimates were based on the 170 iden- foraging theory predicts that larger predators tified prey itemsdelivered during the breedingsea- should have a wider food niche than smaller ones sons.On this basis,47.0% of the prey were mam- (Schoener 1970). I found this to be true for these mals, 45.4% birds and 6.5% reptiles (Fig. 1). two forest-falcons. Collared Forest-Falconscap- Squirrels represented66.7% of the mammalian tured a higher proportion of medium-sizedmam- biomass.Males delivered 11.4 kg (65.7%) and fe- mals, especiallysquirrels, and they had a greater males 5.9 kg (34.3%) of the biomass. diversityof birds in their diet givingthem a broad- Food-nicheParameters. Lizards, especially Anolis er food-nichebreadth (13.8) comparedto Barred SEPTEMBER 2000 FOOI• HABITS OF FOREST-FALCONS 201

Forest-Falcons(7.9). Barred Forest-Falconspreyed foot. Collared Forest-Falcons were observed chas- predominantlyon lizards, mainly Anolisspp., con- ing prey by running on the ground, around tree tributing to its narrower food-niche breadth, and trunks, and along large branches,whereas Barred birds were of secondaryimportance in their diet. Forest-Falconsusually attacked prey by surprise Collared Forest-Falconspreyed on a wider range of from concealed perches. sizes,ranging from a small frog (20 g) to The information provided here is limited to ob- large birds (3 kg) whereasBarred Forest-Falcons servationsduring the nesting seasonand may not caught prey ranging in size from insects(1.5 g) to accuratelyreflect the overall diet of these two spe- a dove (160 g). cies. There may be seasonalshifts in the diet of In terms of biomass,Barred Forest-Falconscap- these forest-falconsor certain prey types may be tured nearly equal proportionsof lizards (37.3%) taken preferentially due to experience or ability as and birds (36.8%) during the breeding season. observed in the extra adult Collared Forest-Falcon This wasattributed to the smaller mean weight of that delivered75% of its prey as Keel-billedTou- lizards (13.8 g) vs. the mean weight of birds (93.5 cans. This particular bird apparently had a special g). Birds were approximatelyseven times heavier ability or learned behavior for capturing toucans. but three times fewer in numbers. Prey biomassof More information is needed from other regionsin Collared Forest-Falcons was distributed nearly the Neotropics and during the nonbreeding sea- equally between mammals (47%) and birds son to determine the extent of niche breadth and (45.4%), but the mean weightof birds (368 g) was dietary overlap between these two species. twice that of mammals (179 g). However,fewer birds (59) than mammals (78) were delivered dur- ACKNOWLEDGMENTS ing the nesting season,contributing to the nearly This studywas part of The Peregrine Fund's Maya Pro- equal frequency of prey biomassof Collared For- ject, in cooperationwith the InstitutoNacional de Antro- est-Falcons. pologia y Historia (IDAEH), Centro de EstudiosConser- The food-niche overlap was relativelyhigh be- vationistas(CECON), Guatemala, and ConsejoNacional de Areas Protegidas (CONAP), Guatemala. A specml tween these two congeners and almost near the thanksto B. Burnham,J.P. Jenny, L. Kiff, and D. Whitacre competition threshold level of 0.6 which was pro- of The Peregrine Fund. Thanks to BoiseState University posedas biologicallysignificant by Zaret and Rand for providingassistance. I kindly thank the staff of Tikal (1971). Schoener(1984) and Temeles(1985) pre- National Park, Guatemala for their assistance. For assist- dicted that similar morphologicalfeatures of rap- ing in the field I would like to thank E.M. Ramirez,J.D. Ramos, C.M. Morales,J.M. Castillo, H. de J.G. Manzane- tors can be found among congenerswhich affect ro, and C.S. Mateo. their hunting ability and food habits.However, Bo- sakowskiand Smith (1992) showed that larger dif- LITERATURE CITED ferencesin body size limit food overlap below the BOS•KOWSKI,t. ANDD.G. SMITH.1992. Comparative diets competitionthreshold. Thus, while the two forest- of sympatricnesting raptors in the easterndeciduous falconsexhibited overlap on a few prey species,I forest biome. Can.J. Zool.70:984-992. suspectthat the effect on overall prey availability BROWN,L. ANDD. AMADON.1989. Eagles,, and fal- was probably insignificant. Both species have a cons of the world. Wellfleet Press, Seacaucus, NJ broad diet with Barred Forest-Falconsrelying more U.S.A. on lizards and Collared Forest-Falconspreying DICKEY,D.R. ANDAJ. vAN ROSSEM.1938. The birds of E1 mainly on squirrels. Salvador.Zool. Ser. 23, Field Mus. Nat. Hist., Chicago, The Barred Forest-Falconis dependent on ma- IL U.S.A. ture forests while the Collared Forest-Falcon oc- DUNNING,J.B., J}•. 1993. CRC handbook of avian body masses. CRC Press Inc., Boca Raton, FL U.S.A. cupies mature forests,forest edge, and secondary EMMONS,L.H. ANDF. FEER.1997. Neotropical rainforest woodlandsand thickets. Both speciesuse a short mammals.Univ. ChicagoPress, Chicago, IL U.S.A. stay"perch-hunting" technique, a common meth- FreEDMANN,H. 1948. Birds collected by the National Geo- od found in forest or -adaptedspecies graphic Society'sExpedition to northern Brazil and (Kenward 1982, Newton 1986). The higher con- southern Venezuela. Proc. U.S. Natl. Mus. 97:373-569 sumptionof avian prey by the Collared Forest-Fal- HAVERSCHMIDT,F. 1968. Birds of Surinam. Oliver & Boyd, con may be enhancedby its great maneuverability, London, U.K. owing to its long legs and long-arched tail which DEE HOYO,J., A. ELIOT, ANDJ. SAP,AG^TAL. 1994. Hand- are morphologicaladaptations for chasingprey by book of the Birds of the World. Vol. 2. New World 202 THORSTROM VOL. 34, No. 3

vultures to guineafowl. Lynx Edicions, Barcelona, [EDS.], Ecologicalcommunities conceptual issues and Spain. the evidence. Princeton Univ. Press, Princeton, NJ IZAWA,K. 1978. A field studyof the ecologyand behavior U.S.A. of the black-mantle tamarin (Sanguinusnigricollis). Fo- SCHULZE, M. AND D.F. WHITACRE. 1999. A classification lia Primates 19:241-274. and ordination of the tree community of Tikal Na- K•NWAm),R.E. 1982. Goshawkhunting behaviour and tional Park, Petfin, Guatemala. Bull. Fla. Mus. Nat. range size as a function of food and habitat availabil- Hist. 41:169-297. ity.J. Anim. Ecol.51:69-80. SNYDER,N.ER. ^NDJ.W. WILEY.1976. Sexualsize dimor- L•SLEY, G.W., C. SEXTON, )•ND G.D. LUCKNER.1994. Win- phism in hawksand owlsof North America. Ornithol ter season,December 1, 1993-February 28, 1994, Tex- Monogr.20. as Region. Natl. AudubonSoc. Field Notes 48:224-228. SMITH,N.G. 1969. Provokedrelease of mobbing--a hunt- LEVINS,R. 1968. Evolution in changing environments. ing technique of Micrasturfalcons. Ib/s 111:241-243. Princeton Univ. Press,Princeton, NJ U.S.A. SMITHE, F.B. 1966. The birds of Tikal. The Natural His- LINTON, L.R., R.W. DAVIES,AND FJ. WRONA.1981. Re- tory Press,Garden City, NY U.S.A. sourceutilization indices:an assessment.J. Anim. Ecol. TEMELES,EJ. 1985. Sexual sizedimorphism of bird-eating 50:283-292. hawks and the effect of prey vulnerability. Am. Nat. M•ER, W. 1981. Notes of nesting raptors in the llanos 125:485-499. of Venezuela. Condor 83:48-51. THORSTROM,R.K. 1993. Breeding ecologyof two species M•TI, C.D. 1987. Raptor food habitsstudies. Pages 67- of forest-falcons (Micrastur) in northeastern Guate- 79 in B.G. Pendleton, B.A. Millsap, K.W. Kline, and mala. M.S. thesis, Boise State Univ., Boise, ID U.S.A. D.A. Bird [EDS.], Raptor management techniques --, C.W. TURLEY, F.G. RAMIREZ, AND B.A. GILROY. manual. Natl. Wildl. Fed. Sci. Tech. Ser. No. 10. Na- 1990. Description of nest, eggs, and young of the tional Wildlife Federation, Washington, DC U.S.A. Barred Forest-falcon (Micrastur ruficollis)and of the M•s, N.M. 1985. Ants and foraging behavior of the Col- Collared Forest-falcon ( Micrastursemitorquatus). Con- lared Forest-Falcon. Wilson Bull. 97:231-232. dor 90:237-239. NEWTON,I. 1986. The Sparrowhawk.T. & A.D. Poyser, TRAIL,P.W. 1987. Predationand antipredatorbehavior at Calton, U.K. Guiana. Auh 104:495-507. R•POLE, J.H, M.A. P,•d•os,•qI) K. WINKER.1989. Win- WILEY,J.W. ANDB.N. WILEY.1981. Breeding seasonecol- tering wood thrush movements and mortality in ogy and behavior of Ridgway'sHawk (Buteoridgwayi). southern Veracruz. Auk 106:401-410. Condor 83:132-151. ROBINSON,S.K. 1994. Habitat selectionand foragingecol- WILLIS, E.O., D. WECHSLE•, ANI) F.G. STILES.1983. Forest- ogy of raptors in Amazonian Peru. Biotropica26:443- falcons, hawks, and pygmy-owlas ant followers. Rev. 458. Bras. Biol. 43:23-28. SCHOENE•,T.W. 1970. Nonsynchronousspatial overlap of ZARET,T.M. ANDA.S. RAND.1971. Competition in tropical lizardsin patchy habitats.Ecology 51:408-418. stream fishes: support for the competitive exclusion ß 1984. Size differencesamong sympatricbird-eat- principle. Ecology52:336-342. ing hawks:a worldwide survey.Pages 254-281 in D.R. Strong, D. Simberloff, L.G. Abele, and A.B. Thistle Received 19 July 1999; accepted 19 March 2000