j. Raptor Res. 27(3):143-148 ¸ 1993 The Raptor ResearchFoundation, Inc.

OBSERVATIONS ON THE COMPARATIVE BEHAVIORAL ECOLOGY OF HARRIS' IN CENTRAL CHILE

JAIME E. JIMgNEZ• Departmentof Wildlifeand RangeSciences, University of Florida, Gainesville, FL 32611 U.S.A.

FABIANM. JAKSI• Departamentode Ecologfa,Universidad Catdlica de Chile,Casilla I Id-D, Santiago,Chile

ABsTR•CT.--Throughoutone year we observedthe behavioralecology of Harris' ( unicinctus)in central Chile. The hawks' activity period lacked diel or annual variation. Their most commonflight modewas soaringin thermalsand wind updrafts,rarely usingflapping flight, and never hovering.Harris' Hawks appearedto selectphysiographic features that favoredthe presenceof updrafts, particularlynorth- and west-facing slopes and ridgetops, but werealso commonly seen flying over ravines (wherethey perched frequently). Prey were primarily small- and medium-sized mammals, and secondarily medium-sizedbirds. Although not aggressive,Harris' Hawks were nonethelessattacked by two other sympatricraptors, Black-chested ( melanoleucus) and Red-backedHawks (Buteopo- lyosoma).These three ,which were studiedover the sameperiod and with the sametechniques, were similar in activityand behavior,although Harris' Hawks were slightlymore differentthan the latter were between themselves.

Observacionessobre la ecologiaconductual comparativa de peucosen Chile central RESUMEN.--Porun afio observamosla ecologiaconductual del peuco(Parabuteo unicinctus) en Chile central.E1 periodode actividadde los peucosno difiri6 ni dentrodel dia ni a lo largo del afio. Su modo de vuelomils comfn fue el planeoen corrientest•rmicas y de obstrucci6n,raramento usando el vuelo batidoy nuncael revoloteo.Los peucos parecieron seleccionar caracteristicas fisiogrificas que favoreclan la presenciade corrientesascendentes, particularmente las laderas de exposici6nnorte y oestey las cimas, pero tambi•n se los veia volandosobre quebradas (en dondetambi•n se posaban).Sus presaseran principalmentemamlferos de tamafiospequefio y mediano,y secundariamenteaves de tamafiomediano. Aunquelos peucoseran pocoagresivos, eran atacadospor otrasdos especies simpltridas de rapaces,el iguila (Geranoaetusmelanoleucus) y el aguilucho( polyosoma). Estas tres especies,que fuerones- tudiadasen el mismoperlodo y conlas mismat•cnicas, eran similaresen actividady conducta,aunque los peucoseran ligeramentemils diferentesque las otrasdos entre elias. [Traducci6n Autores]

The Harris' Hawk, Parabuteo unicinctus,is dis- 1980), food-nicherelationships with othersympatric tributed from the southwestern United States in the predators(Jaksifi et al. 1981), and conservationsta- north throughCentral and SouthAmerica to central tus (Jaksifiand Jimgnez 1986). Here we report on Chile and Argentina (Brown and Amadon 1968). aspectsof the behavioralecology of the Harris' Hawk The specieshas been relatively well studiedin North nearits southernmostdistributional limit. This study America (e.g., Mader 1975a, Bednarz et al. 1988 is basedon one year of observationsmade in central and referencestherein). By contrast,little has been Chile concurrentwith thosemade on two other sym- publishedabout the specieselsewhere in Central and patric raptors: Black-chestedEagle (Geranoaetus South America. To our knowledge,the only reports tnelanoleucus;Jimgnez and Jaksic 1989) and Red- come from Chile and cover its diet (Jaksifi et al. backedHawk (Buteopolyosoma; Jimgnez and Jaksifi 1991). Becausethis is the last report of this series, we take the opportunity to make comparisonsbe- • Presentaddress: Department of Biology,Utah StateUni- tween the behavioralecology of the Harris' Hawk versity, Logan, UT 84322 U.S.A. and the two other species.

143 144 JAIMEF•. JIMgNEZ AND FABIAN M. JAI•SI• VOL. 27, NO. 3

STUDY SITE AND METHODS titled with standardprocedures (Marti 1987), as exem- The studysite is describedin Jim•nezand Jaksifi (1989). plified by Jim•nez and Jaksifi(1989) and Pavez et al. Briefly,San Carlos de Apoquindo(33ø23'S 70ø31'W) is a (1992) for sympatricBlack-chested Eagles. ruggedarea 20 km E of Santiagoin the Andeanfoothills with elevationsranging from 1050-1915 m. The physi- RESULTS AND DISCUSSION ographyincludes both flat areasand numerousridges dis- sectedby deepravines. The climateis mediterranean,with Activity levels(measured in minutes) did not dif- cool,rainy wintersand dry, hotsummers. The wind blows fer amongtimes of the day (F = 1.74, df = 5, P > westwardlyfrom the valley to the mountainsduring the 0.12), nor with season(F = 1.73, df = 3, P > 0.16); daytime.The dominantvegetation is an evergreenscrub (very similar to the California chaparral)that changes the interaction betweenthese variables was signifi- physiognomydepending on topographyand orientation. cant (F = 1.91, df = 15, P < 0.03). Indeed, Harris' Based on the number and duration of sightings,the Hawks were observedthroughout the day with no Black-chestedEagle was the mostcommon raptor at the clear peaksof activityin any interval of the day, or site, having been sampled(timed with a stopwatch)for nearly93 hr over a calendaryear (Jim•nez and Jaksifi in any particular season.Thus, Harris' Hawks dif- 1989); the secondmost common raptor was the Red-backed fered from sympatricBlack-chested Eagles, which Hawk, observedfor nearly 29 hr timed (Jim•nez and had a bimodal activity period throughoutthe year, Jaksifi1991). Harris' Hawk rankedthird in this respect, and from Red-backed Hawks, which were more ac- having been monitoredfor 13 hr over the samesample tive during summer and least active during winter. period. Following the same protocol used for the other two Thermal soaringwas the mostcommon flight mode species,for each Harris' Hawk observedwe recorded:1) throughoutthe year, ranging from 47% (winter) to time of observationand duration of activity, 2) activity 16% (summer) of the observedtime (Table 1). Wind type, and 3) habitat beneathbird. We recognizedthe fol- soaringwas the secondmost prevalent flight mode, lowing activity types (cf. Jim•nez and Jaksifi 1989 for detaileddescription): thermal soaring,wind soaring,cruis- accountingfor 37% (fall) to 16% (summer) of the ing, hovering,harassing, perching, and miscellaneousbe- activity time. Harris' Hawks spentmost of their time havior. We recognizedthe followinghabitat types(cf. Ji- perchedduring summer(64%), but very little during m•nez andJaksifi 1989 for detaileddescription): fiatlands, winter (4%). The remainingactivities (Table 1) ac- ravines, ridgetops,and slopes(east-, west-, south-, and countedat the most (during winter) for 19% of the north-facing).We mapped thesehabitats and calculated their surfaceareas (slope-corrected) with a digital planim- daily activity period, and for lessthan 7% during eter from a high-resolutionaerial photograph.The phys- the remaining seasons.It should be noted that our iognomy of each habitat type, in terms of the vegetative samplingtechnique (observing from a hilltop) may coverrepresented by trees,shrubs, herbs, bare ground,and have biasedour detectionstoward soaringbirds and rocks, as well as estimatedprey numbers,were reported in Jim•nez and Jaksifi(1989). away from perchingHarris' Hawks and thosemak- We made observationswith binoculars from the top of ing shortlow flights.Unlike sympatricBlack-chested a hill, from dawn to dusk, with the day divided into six Eagles(Jim•nez and Jaksi• 1989) and Red-backed equal-timeintervals (see Jim•nez and Jaksifi1989). Ob- Hawks (Jim•nez and Jaksifi1991), Harris' Hawks servationstook place during one entire day every other apparently did not hover, and spent little time in week between 1 August 1984 and 1 August 1985. We pooleda total of 780 min of samplinginto four seasons: agonisticinteractions (Table 1). spring (1 August to 30 October), summer (1 November Observed occurrencesof Harris' Hawks (in min- to 31 January), fall (1 February to 30 April), and winter utes) differed both amonghabitat types(F = 18.06, (1 May to 31 July). We usedone- or two-way ANOVA df = 6, P < 0.0001) and amongseasons (F = 3.53, for unequalsample sizes (Sokal and Rohlf 1981:210,360), with PROC GLM in SAS (1985), and the Student-New- df = 3, P < 0.02); there was a significantinteraction man-Keuls test for a posterioricontrasts (SAS 1985:444). between these variables (F = 2.13, df -- 18, P < For analyzing frequency data, we used the G statistic 0.004). Throughout the year, Harris' Hawks flew (Sokal and Rohlf 1981:695). It should be noted that the more often over north-facingslopes (from 59% in assumptionof independenceof data is violatedin behav- summer to 19% in winter), secondarilyover west- ioral studies,because what a doesone minute is very hkely to influencewhat it will be doing the next minute. facing slopes,ravines, and ridgetops(from 35% to However, this should affect more strongly comparisons 3% in different seasons),and very little over flat- made within a season than between seasons. Still, the lands, south- and east-facingslopes (Table 2). Ex- P-values calculated should be consideredimprecise and cept for the latter three, there were clear seasonal our interpretations of differencesand similarities taken shiftsin the use of the remaining four habitat types cautiously. We identified prey remains found under perchesand (SNK a posterioritest, P < 0.05 in all cases).West- one nest, and from regurgitated pellets. Prey were iden- facingslopes were usedmore during fall and winter SEPTEMBER 1993 HARRIS' HAWK BEHAVIOR IN CHILE 145

Table 1. Time (in minutes) spentby Harris' Hawks in different activitiesin central Chile, 1984-1985.

SPRING a SUMMER FALL WINTER ACTIVITY N = 12 d N = 12 d N = 6 d N = 7 d Thermal soaring 8.3 + 4.8b 3.5 _+1.9 8.4 _+7.7 5.2 + 3.9 % time 34.4 15.8 34.3 47.3 Wind soaring 5.8 + 3.3 3.5 + 2.1 9.1 + 4.1 3.3 _+ 1.9 % time 24.2 15.8 37.1 30.0 Cruising 1.5 + 1.0 0.8 _+0.5 1.3 + 1.0 2.1 + 1.3 % time 6.2 3.6 5.3 19.1 Hovering 0.0 + 0.0 0.0 + 0.0 0.0 _+ 0.0 0.0 + 0.0 % time 0.0 0.0 0.0 0.0 Harassing 0.0 + 0.0 0.1 + 0.1 0.1 + 0.1 0.0 + 0.0 % time 0.0 0.4 0.4 0.0 Perching 8.3 + 5.9 14.3 + 14.5 5.4 + 8.1 0.4 + 0.5 % time 34.4 64.4 22.1 3.6 Other 0.2 _+ 0.2 0.0 + 0.0 0.2 + 0.2 0.0 _+ 0.0 % time 0.8 0.0 0.8 0.0 Total time/period 24.1 ñ 10.6 22.2 + 13.9 24.5 _+ 16.0 11.0 + 5.2 Total time/season 289.2 266.4 147.0 77.0 See Methods sectionfor periodsinvolved. Mean numberof minutesthis activity was observed per day + two standarderrors. than in other seasons,while the reversehappened facingslope (19%), ridgetop (18%), west-facing slope with regard to north-facingslopes (Table 2). (14%), flatland(9%), ravine (7%), and east-facing Jim•nez and Jaksi• (1989) showedthat the areal slope(7%). Harris' Hawks thus appearedto avoid representationof differenthabitat types in the study flying over south-facingslopes and fiatlands,while sitewas as follows:north-facing slope (26%), south- they concentratedon ravines (Table 2). The near

Table 2. Time (in minutes)spent by Harris' Hawks on sevenhabitat types in centralChile, 1984-1985.

SPRINGa SUMMER FALL WINTER ACTIVITY N = 12 d N-- 12 d N = 6 d N = 7 d

Flatland 0.1 + 0.2 b 0.4 + 0.5 0.4 + 0.5 0.7 + 1.3 % time 0.4 1.8 1.6 6.4 Ravine 7.9 + 3.9 4.7 + 4.3 3.5 + 2.6 0.7 + 0.6 % time 32.8 21.2 14.3 6.4 Ridgetop 4.0 + 3.7 2.1 + 2.1 5.6 + 5.8 1.8 + 2.2 % time 16.6 9.4 22.9 16.3 North slope 9.2 + 5.5 13.1 + 14.9 4.8 + 4.9 2.1 + 1.6 % time 38.2 59.0 19.6 19.1 South slope 0.5 + 0.4 1.2 + 1.2 2.9 + 1.9 1.1 + 1.8 % time 2.1 5.4 11.8 10.0 East slope 0.0 q- 0.0 0.0 q- 0.0 0.1 q- 0.1 0.8 q- 1.0 % time 0.0 0.0 0.4 7.3 West slope 2.4 + 3.1 0.7 q- 0.8 7.2 + 9.9 3.8 q- 2.6 % time 9.9 3.2 29.4 34.5 Total time/period 24.1 + 10.6 22.2 + 13.9 24.5 __+16.0 11.0 q- 5.2 Total time/season 289.2 266.4 147.0 77.0 See Methods sectionfor periodsinvolved. Meannumber of minutesthis activity was observed per day + twostandard errors. 146 JAIMe.E. JIMRNEZAND FABIAN M. JAKSI• VOL. 27, No. 3

Table 3. Prey of Harris' Hawks in central Chile, basedon prey remains and pelletsfound under severalperches and one nest, 1984-1985. Integer numbersare the absolutenumerical representation of eachprey item; numbersin parenthesesare the percentnumerical representationof major prey classes.Prey weightsin grams.

PREY WEIGHT PELLETS REMAINS TOTAL

Mammals -- (67.4) (66.7) (67.3) Rodentia Abrocomabennetti (Bennett's chinchilla rat) 231 5 1 6 Akodonlongipilis (Hairy field mouse) 63 1 0 1 Akodonolivaceus (Olivaceous field mouse) 44 1 0 1 Octodondegus (Fence degu rat) 184 20 3 23 Oryzomyslongicaudatus (Long-tailed rice rat) 36 3 0 3 Lagomorpha Oryctolaguscuniculus (European rabbit)a 1300 18 12 30 Mammals unidentified -- 47 0 47

Birds -- (19.8 (29.1) (21.2) Tinamiformes Nothoproctaperdicaria (Chilean Tinamou) 400 1 1 2 Columbiformes Zenaidaauriculata (Eared-dove) 125 0 1 1 Galliformes Callipeplacalifornica (California Quail) 200 0 1 1 Piciformes Unidentified 100 1 0 1 Passeriformes Mimus thenca(Chilean Mockingbird) 65 1 2 3 Curaeuscuraeus (Austral Blackbird) 90 1 1 2 Sturnellaloyca (Red-breasted Meadowlark) 110 1 0 1 Anairetesparulus (Tufted Tit-tyrant) 11 1 0 1 unidentified -- 21 1 22 Bird egg -- 1 0 1 Reptiles -- (7.8) (4.2) (7.3) Sauria Liolaemussp. (Unidentified lizard) 15 4 0 4 Callopistespallurea (Chilean racerunner) 65 4 1 5 Serpentes Philodryaschamissonis (Long-tailed snake) 150 3 0 3 Insects (5.0) (0.0) (4.2) Coleoptera Buprestidae <1 1 0 1 Scarabaeidae <1 5 0 5 Insects unidentified <1 1 0 1 Total number of prey -- 141 24 165 Both adults and juveniles;the weight indicatedis for adults.

proportionaluse of north-facingslopes and ridgetops sweptby the prevailingwesterly wind (Jim•nez and may be explainedbecause these habitat types receive Jaksifi 1989). In contrast,south-facing slopes and more direct solar radiation, and thus probablygen- fiatlandsreceived relatively low amountsof incident erate more thermal drafts. Similarly, west-facing radiation and were shelteredfrom prevailingwinds. slopesreceive radiation from the settingsun, and are Harris' Hawks behavedsimilarly to Black-chested SEPTEMBER1993 HARRIS' HAWK BEHAVIORIN CHILE 147

Eagles(Jimgnez and Jaksifi 1989) andRed-backed Jim6nezand Jaksifi1991). Harris' Hawks were in- Hawks (Jimgnezand Jaksifi 1991), exceptfor flying frequentlyharassed by conspecifics(23% of the at- much more often over the cool and windless ravines. tacks)and primarily by Red-backedHawks (46%) The amountof time perchingin differenthabitat and Black-chestedEagles (31%) which are larger. typeswas not related to theirrelative availability (G None of these agonisticencounters involved prey, = 20.0, df = 4, P < 0.001). In 42% of 45 sightings, unlike the situationreported by Bildstein(1987) for Harris' Hawks were perchedin north-facingslopes, sympatricRed-tailed Hawks (Buteojamaicensis)and 33% were perchedin ravines,11% in ridgetops,9% Rough-leggedHawks (B. lagopus)in Ohio. in west-, and 5% in south-facingslopes. We never Becausepellets found under perchesand remains saw Harris' Hawks perchingin flatlandsor east- found under the one nest sampled were collected facingslopes. These findingsare similar to those in the samehabitat and time of the year, a compar- reportedfor Black-chestedEagles (Jimgnezand ison is warranted. Mammalian prey were equally Jaksi•1989) and Red-backed Hawks (Jim•nez and representedamong pellets and nest remains, avian Jaksi• 1991), exceptfor perchingin ravines.Note prey were comparativelyunderrepresented among that order of preferencefor habitat types used for pellets,and the reverseoccurred with both reptilian perchingby Harris' Hawks was the sameas for and insectprey (Table 3). Overall, the differences flyingover. The apparentpreference for perching in prey compositionwere not dramaticallydifferent on north-facingslopes may be becausethey afford betweenpellets and nestremains. Harris' Hawks in easyaccess to updrafts.The preferencefor perching this locality consumedprimarily small- (rodents) in ravinesmay be associatedto the higherdensities and medium-sized mammals (rabbits), and second- of mammalianprey in thoseareas (cf. data in Ji- arily medium-sizedbirds (severalorders). Lizards m•nez and Jaksi• 1989). and snakeswere minor componentsof their diet, and Perchsubstrates were not usedhomogeneously (G insects were almost insignificant in biomass. No = 48.0, df = 8, P < 0.001). In decreasingorder, comparabledata are available for sympatric Red- perchesused were standingdead trees (33%), live backedHawks (Jim•nez and Jaksi• 1991), but for treesof Quillajasaponaria (29%), live treesof Lith- Black-chestedEagles Jim6nez and Jaksifi (1989) and raea caustica(16%), boulders (9%), standingdead Pavez et al. (1992) showedthat their diet in the same bromeliads (7%), live trees of Kageneckiaoblonga localityis similar to that reportedhere for Harris' (2%), of Porlieriachilensis (2%), and live columnar Hawks. Major differencesare the lack of insectcon- cacti(2%). SympatricBlack-chested Eagles (Jim•- sumption by the eagles,the relatively higher con- nez and Jaksi• 1989) and Red-backedHawks (Ji- sumptionof reptiles,and the lower consumptionof mgnezand ,Jaksi• 1991) displayedroughly the same birds. perchingpreferences. All thesestructures were above It is noteworthythat throughoutthe studyperiod, the generallevel of the scrubcanopy, and probably no evidencewas observedof cooperativebreeding providedgood visibility and easyaccess to updrafts. and/or hunting, as is commonin North America Aggressiveencounters (harassment) were ob- (Mader 1975b, 1979, Bednarz 1987, 1988). servedonly duringsummer and fall, and occupied In summary,Harris' Hawks were slightlymore lessthan 1% of the Harris' Hawks' time (Table 1). differentin termsof activityand behaviorthan were Red-backedHawks (weight = 975 g) initiated 53 Black-chested Eagles and Red-backed Hawks. attacks on Harris' Hawks and received 48 attacks Among the most remarkable differencesbetween (Jim•nez and Jaksi• 1991), Black-chestedEagles Harris' Hawks and the otherswere: (a) they lacked (weight = 2378 g) perpetrated49 and received62 diel or annual differencesin activitylevels, (b) they attacks(Jimgnez and Jaksi• 1989), and conspecificsdid not usehovering as a flight mode,(c) they spent (weight= 876 g) initiated13 and received7 attacks moretime flying overravines, (d) they usedperches throughoutthe same observationperiod. Black- in ravinesmore often, and (e) they showedlesser chestedEagles were aboutas frequently harassed by aggressiveness. conspecificsas by other species(49% of attackson ACKNOWLEDGMENTS Black-chestedEagles were allospecific;cf. Jimgnez We thank Eduardo Pavez for field support. The re- andJaksi• 1989), whereasRed-backed Hawks were searchwas originally fundedby grant DIUC 83-202 from slightlymore frequentlyharassed by other species the Universidad Cat61ica de Chile and INT 83-08032 than by conspecifics(55% of allospecificattacks; cf. from the U.S. National ScienceFoundation, and completed 148 JAIMEE. JIMgNEZAND FABIAN M. JAKSI• VOL. 27, NO. 3 undertenure of grantsDIUC 87-094, INT 88-02054, and , H.W. GREENEAND J.L. YAI•EZ. 1981. The Ghile's FONDEGYT 90-0725. James Bednarz, Peter guild structureof a communityof predatoryvertebrates Bloom,Gary Bortolotti,David Ellis, and two anonymous in central Chile. Oecologia49:21-28. reviewersmade cogent criticisms that helpedimprove our JIMgNEZ,J.E. ANDF.M. JAKSIe. 1989. Behavioralecol- paper. ogy of Grey -,Geranoaetus melanoleucus, in central Chile. Condor 91:913-921. LITERATURE CITED -- AND-- 1991. Behavioralecology of Red- BEDNARZ,J.C. 1987. Pair and group reproductivesuc- backed Hawks in central Chile. Wilson Bull. 103:132- cess,polyandry, and cooperativebreeding in Harris' 137. Hawks. Auk 104:393-404. MADER, W.J. 1975a. Biology of the Harris' Hawk in 1988. Gooperativehunting in Harris' Hawks southern Arizona. Living Bird 14:59-85. (Parabuteounicinctus). Science 239:1525-1527. 1975b. Extra adults at Harris' Hawk nests. , J.W. DAWSONAND W.H. WHALEY. 1988. The Condor 77:482-485. status of the Harris' Hawk in the southwestern United 1979. Breeding behavior of a polyandroustrio States.Pages 71-82 in R. Glinsky,B. Millsap, S. Hoff- of Harris' Hawks in southern Arizona. Auk 96:776- man and B. PendletonlEDs.], Proc. SouthwestRaptor 788. Manage. Symp. and Workshop. Nat. Wildl. Fed., MARTI, C.D. 1987. Raptor food habits studies.Pages Washington, DC U.S.A. 67-80 in B.A. Giron Pendleton,B.A. Millsap, K.W. BILDSTEIN,K.L. 1987. Behavioralecology of Red-tailed Cline and D.M. Bird lEDS.],Raptor managementtech- Hawks (Buteojamaicensis),Rough-legged Hawks (Bu- niques manual. Nat. Wildl. Fed., Washington, DC teolagopus), Northern Harriers (Circuscyaneus), and U.S.A. American Kestrels(Falco sparverius) in south central PAVEZ,E.F., C.A. GONZALEZAND J.E. JIM•NEZ. 1992. Ohio. Ohio Biol. Surv., Biol. Notes 18. Diet shifts of Black-chestedEagles (Geranoaetusmela- BROWN,L. ANDO. AMADON. 1968. Eagles,hawks and noleucus)from nativeprey to Europeanrabbits. J. Rap- falconsof the world. Country Life Books, Felthorn, tor Res. 26:27-32. Middlesex, U.K. SAS. 1985. SAS user'sguide: statistics, fifth edition. SAS JAKSI•, F.M. ANDJ.E. JIMENEZ. 1986. The conserva- Institute, Inc., Cary, NC U.S.A. tion statusof raptors in Chile. Birds Prey Bull. 3:95- SOKAL,R.R. AND F.J. ROHLF. 1981. Biometry. 2nd 104. edition, W.H. Freeman and Co., San Francisco, CA , J.L. YAiqEZAND R.P. SCHLATTER.1980. Prey U.S.A. of the Harris' Hawk in central Chile. Auk 97:196- 198. Received28 August 1992; accepted1 April 1993