sign in sign up
<<
Home , Common buzzard, Hooded crow

J. Raptor Res. 28(2):100-107 ¸ 1994 The Raptor ResearchFoundation, Inc.

BEHAVIOR AND ACTIVITY OF REHABILITATED ( buteo) RELEASED IN NORTHERN

DAVIDE CSERMELY AND CARLO VITTORIO CORONA 1 Dipartimentodi Biologiae FisiologiaGenerali, Universitd di Parma, Viale delle Scienze,1-43100 Parma, Italy

AI•STI•CT.--The behaviorand habitat huntingof 16 rehabilitatedcommon buzzards (Buteo buteo) released in northernItaly were analyzed.The buzzardswere releasedindividually in differentseasons, and their activitywas recorded continuously for at leastthe first 3 d afterrelease and intermittently thereafter until they dispersedfrom the releasesite. The remainedin the surroundingarea for more than 100 d, showinga progressiveacclimation to the new environment.The releasedbuzzards interacted frequently with wild territorialconspecifics and were attackedby severalspecies of corvids,especially the hooded crow (Corvuscorone). Nevertheless, such interactions were not the directcause of dispersal.Some birds defendeda territoryadjacent to or insidethat of a wild .Prey capture was almost normal, although certainlyunderestimated. Small mammalsand reptileswere mostoften caught. Although the area chosen for this studyhad high human population,this was not a major sourceof interferencewith the releases. Thus,the buzzards appeared to be ableto copewith theirnew environment being minimally influenced by havingbeen in captivity. KEY WORDS: hawks;Buteo buteo; behavior; captivity; dispersal,' rehabilitation; behavioral ecology.

Comportamientoy actividad de los Buteobuteo rehabilitados y deyadosen libertaden el nordede Italia

RESUMEN.--Seanaliz6 la conductay el habitatde caza de 16 individuosrehabilitados de la especieButeo buteo liberados en el norte de Italia. Los individuos de B. buteo fueron liberados en el firea de estudio individualmenteyen diferentesestaciones; su actividadfue registradacontinuamente por al menostres dias despugsde su liberaci6ne intermitentementehasta el momentode abandonarel sitio de liberaci0n. Las avespermanecieron en los alrededoresdel firea por mils de 100 dias,mostrando una progresiva aclimataci0nal nuevoambiente. Los individuos liberdos interactuaban frecuentemente con conespecificos territorialessilvestres y fueronatacados por variasespecies de c6vidos,especialmente corone. Sin embargo,tales interacciones no fueronla causadirecta de su dispersion.Algunas aves defendieron territoriosvecinos o al interiorde losdefendidos por individuossilvestres. La capturade presafue casi normal,aunque ciertamente subestimada. Tanto pequefiosmamiferos como fueron a menudo capturados.Aunque las fireasescogidas para esteestudio tenlan una alta poblaci6nhumana, este factor no constituy6una gran fuente de interferenciassobre las liberaciones. En slntesis,B. buteoparece ser capazde incertarseen sunuevo mediambiente siendo escasamente influenciado por su cautividad. [Traducci6n de Ivan Lazo]

Severalprograms for the rehabilitationof raptors still neglected.In fact, it is almostimpossible to get have been developedin recentyears by institutions information of the fate of released birds from the devotedto the protectionof birds. Standard proce- literature, because most data refer to survival rate dures for raptor rehabilitation have been developed and recoverydistance from the releasesite (Servheen for severalspecies (Nelson 1977, Llewellynand Brain and English 1979, Duke et al. 1981, Ingram 1983, 1983, Pendletonet al. 1987, Weaver and Cade 1991) Hamilton et al. 1988). Moreover, little precisein- as well as the techniquesfor successfulrelease (Shet- formation has been compiledon the behaviorof in- rod et al. 1982, Llewellyn 1991). Nevertheless,the dividuals after release. adaptationof birds back into the natural habitat is The objectiveof this study was to fill that gap, investigatingin detail the behavior,activity, and in- tra- and interspecificinteractions in a group of com- • Present address: Via P. Petronia 89, 1-00136 Roma, mon buzzards(Buteo buteo) immediately after re- Italy. lease following rehabilitation until they dispersed

100 JUNE 1994 BEHAVIOROF RELEASEDBUZZARDS 101

from the area. Such an investigationis likely to be The daylight period was equally dividedinto three sec- of interestto raptor rehabilitators(Meyers and Mil- tions that were variable during the year based upon the ler 1992). photoperiod.The proportionof time spentin eachhabitat for each one-third of the daylight period and in every METHODS seasonwas arcsintransformed for comparison.The exact time of sunriseand sunsetfor the geographicalcoordinates The buzzardsused in this study were all wild birds, of the area were calculatedevery 2 wk. The days after housedtemporarily for rehabilitation at the Raptor Re- releasewere countedconsidering the day of releaseas day habilitationCentre (RRC) managedby the Italian Society one. The days of the year were indicatedconsidering the for the Protectionof Birds (LIPU) near Parma. They all spring equinox as day zero. originatedin northernor centralItaly within 100-200 km We usedthe Mann-Whitney U-test to comparemeans, of the release area. When released the birds were all in the Kruskal-Wallis one-way ANOVA (Siegel 1956) to perfect physical condition and flying fitness, and were evaluate time durations between seasons or between thirds chosenrandomly amongthose ready for releasing.Those of the day, the Spearman rank correlationto ascertain possiblyimprinted to humans were not considered. possiblecorrelations, and the Chi-square test to compare A total of 16 buzzards were studied. Six were adult frequencies.The means are given ___SE, and the proba- (two males and four females) and 10 were sub-adults (five bility is always given as two-tailed. malesand five females).They were releasednear the end of each season,from April 1990 to November 1991. Five RESULTS were studied between winter and spring, four between spring and summer, three between summer and autumn, Most buzzardsdid notdisappear quickly from the and four between autumn and winter. Thus, we avoided releasesite in 426.6 hr of observation(Table 1). The the most stressfulclimatic conditionsthat occur in Janu- area used by the birds was about 2730 ha in size, ary-February and in July-August (Kostrzewaand Kos- almost centered around the release site. One-half of trzewa 1991). The duration in captivity was variable, the sampledbuzzards left the studyarea within three ranging from a few days to severalmonths, depending on the seriousnessof the injury or illness.The mean duration days. Three departed within a few hours and one for nine buzzards was 295.5 + 109.2 d. We did not know on day three. This occurredin early spring and the period for the other five, but it was certainly within autumn. the same range. Mortality. Three buzzardsdied in the studyarea The release site was located within a waterfowl sanc- by electrocutionafter perching on medium-tension tuary managedby LIPU about 15 km north of Parma and 5 km from the Po River. The area is flat and without pylonswhich are widely distributedin the plains of extensivewoodlands but with a high human population Italy, and unfortunatelyare a seriousproblem for density. The site was chosenbecause of the necessityto otherspecies too. One diedof a gunshotwound observeclosely and track the buzzard behavior precisely, receivedduring the night or at dawn beforewe start- even for long distancesif necessary.Wild buzzards are regularly presentparticularly during winter. Severaltaxa edour observationsession, and anotherdied in spring of invertebratesand terrestrialvertebrates offered an easy for unknownreasons during a late snowfall.Finally, and variable food source. one buzzard was recaptured closeto starvation. The area surroundingthe releasesite containedseveral Habitat Use. Habitat typeswithin the studyarea biotopes,with rather differing vegetalcover. The habitat typeswere evaluatedusing the methoddescribed by Emlen were small. Their distributionwas almostregular, (1956). Several watercourses--the Parma River and sev- andthe buzzardsmoved very easilyfrom onehabitat eral streams--run within the studyarea. Most of the trees to another.We foundgreat individual variability in are concentratedalong them. habitat use (H = 31.79, N = 450, P < 0.001, mea- The buzzardswere releasedindividually between 0900- suredas minutes spentby eachbird in eachhabitat). 1500 H on days without precipitation.Beforehand they were kept on locationin an outdoor aviary for 1-2 wk in Birdsthat changedhabitat frequently had beenkept order for them to habituateto the environment.A radiotag in captivity for the least time (Z = 2.18, N = 279, (9 g two stage,BIOTRACK, Wareham, U.K.) was at- P < 0.05). During winter, buzzardsstayed in one tached some hours before release (Kenward 1987, White habitat longer than in other seasons,both during and Garrott 1990). days 1-3 (H = 10.28, N = 390, P < 0.05) and in The buzzardswere followed virtually continuouslyfrom dawn to dusk each day if weather permitted for the first all-days (H = 17.58, N = 485, P < 0.001). 3 d after release, hereafter referred as "days 1-3." If a The buzzards remained longer in open habitats bird did not leavethe area, it was subsequentlymonitored than in areaswith thick vegetationin every period intermittentlywith the sameschedule at 1-4 d intervals, considered(Kostrzewa 1989). The tendencyto ex- until the bird disappearedfrom the surroundings.That period, including the first 3 d, is hereafter indicated as plore different habitats immediately after release, "all-days." Observationswere carried out using 8 x bin- i.e., the minutesspent in eachhabitat beforemoving ocularsand a 10-40 x zoom spottingscope. to another one, decreasedwith time (r• = 0.10, N = 102 DAVIDE CSERMELYAND CARLO VITTORIO CORONA VOL. 28, No. 2

Table 1. The historyof commonbuzzards released following rehabilitation near Parma, Italy.

BUZZARD IDENTIFI- DAYS REMAINING DURATION OF CATION SEX, WITHIN THE OBSERVATION CAUSEOF CODE AGE DATE OF RELEASE STUDY AREA (hr) OBSERVATIONEND VR-340 M JU 25 Oct 1990 1 1.5 Abandonment a AV-670 FAD 6 Dec 1990 1 2.8 Abandonment a RN-700 F JU 27 Mar 1991 1 5.2 Abandonment a V-525 FAD 11 Apr 1990 3 17.4 Abandonmenta 0-790 F JU 15 Jun 1990 4 27.0 Abandonmenta VA-425 F JU 19 Jun 1991 4 25.7 Abandonmenta NM-920 MAD 11 Sep 1991 4 28.2 Abandonmenta VN-355 M JU 13 Apr 1991 7 30.4 Death RN-670 F JU 17 Nov 1990 11 37.7 Death VP-960 M JU 20 Sep 1991 13 29.0 Death AN-690 FAD 16 Mar 1991 14 45.2 Abandonment a B-355 F JU 30 Apr 1991 18 31.3 Abandonmenta A-260 M JU 9 Jun 1990 29 33.9 Abandonmenta GM-440 MAD 3 Nov 1990 39 46.5 Recapture RS-1150 FAD 15 Sep 1991 65 34.7 Abandonmenta RA-450 FAD 23 Jun 1991 103 30.2 Abandonmenta Buzzard left the release area.

485, P < 0.05). Habitats with treeswere usedmost spp.;X2(1) : 111.90,P < 0.001) and willows(Salix (74.8%of time), particularlytree rows (55.9%). The alba;X2<•> = 12.99, P < 0.001), while in summer(N time spentin suchhabitats was inversely correlated = 221) they restedin (Quercusspp.; X2<•>= with tree distance(rs = -0.12, N = 310, P < 0.05). 18.11,P < 0.001) andagain in poplars(X2<•> = 42.75, The birds preferredthose areas in spring,particu- P < 0.001) and willows X2<•>= 15.02, P < 0.001). larly the "irregular" woods(H -- 11.29, N = 147, In autumn (N = 237) they preferredopen habitats P < 0.02) and poplar plantations(H = 11.02, N = and either perchedon pylons(X2(•> = 21.72, P < 147, P < 0.02), while in summerthey stayedmostly 0.001), poplars(X2<•> = 27.16, P < 0.001) or de- •n woods with trees in rows (H -- 12.54, N = 146, scendedto the ground,but in winter (N = 46) they P < 0.01). Sucha preferencechanged dramatically returnedto a preferencefor trees,again principally in autumn and winter, when the birds choseprin- poplars(X2<•> = 40.45, P < 0.001). The perching cipally open/cultivatedareas (H = 20.55, N = 147, duration was unaffectedby the type of perchingsite P < 0.001). In contrast,they appearedto avoidthe and averaged30.70 _+ 1.22 min (N = 652). vicinity of buildingsor other areaswhere human Perch height was negativelycorrelated with the presencewas evident. Only sixbirds frequented such perchingduration in both days 1-3 (rs = -0.10, N areas, perchingclose to human settlementsand = 486, P < 0.05) and in subsequenttime periods spendingno more than 20% of theobservation period (rs= -0.09, N = 652, P < 0.02). Height was strong- there.We foundno relationshipbetween time of day ly influenced(H = 37.14, N = 652, P < 0.001) by and habitat preference.The birds remaining for a seasonin either periodranging from 5.26 + 0.26 m long time within the study area were also able to (all-days)in springto 3.27 + 0.17 m (all-days)in occupya territory adjacentto or within a territory summer and in winter. defendedby a wild conspecific,but behavedas sub- Flight Performance. The buzzardsflew some ordinate to the latter. distanceaway immediately after release, but re- Perching Sites.The buzzardsperched most fre- mainedwithin a rangeof 400-5000 m. The distance quentlyin treebranches, but alsooften used pylons, from the releasesite increased progressively to 1295.0 poles,or simplystood on the ground.In spring(N + 217.4 m on day three.These values were greatly = 148) theyperched most often on poplars(Populus affectedby season(H = 18.25,N = 551, P < 0.001), JUNE1994 BEHAVIOROF RELEASED BUZZARDS 103

Table2. Someparameters of flappingor glidingflights and soaring for eachreleased buzzard. (It wasnot possible to ascertainthe actual heightof all flights.)

BUZZARD MEAN No. MEAN No. OF IDENTIFI- TOTAL OF MINUTES HOURS MEAN (2 SE) MEAN (+ SE) CATION NO. OF BETWEENTWO BETWEENTWO HEIGHT (m) LENGTH (m) CODE FLIGHTS FLIGHTS SOARINGFLIGHTS OF FLIGHT (N) OF FLIGHT (N) AV-670 6 28.3 -- 11.5 _ 1.5 (2) 441.7 q- 141.8 (6) RN-700 8 38.8 5.167 -- 491.9 + 204.6 (8) VR-340 3 30.0 1.5 -- 150.0 + 57.7 (3) V-525 11 95.0 8.733 75.0 + 0.0 (2) 336.4 q- 134.7 (11) 0-790 25 64.8 -- -- 173.0 q- 17.8 (25) GM-440 51 49.1 46.633 -- 187.4 + 18.2 (51) NM-920 48 35.2 28.167 9.1 q- 1.9 (26) 176.6 + 17.6 (48) VA-425 45 34.2 -- 8.2 q- 0.7 (26) 161.1 q- 14.9 (45) VN-355 76 27.2 15.217 14.8 q- 2.8 (29) 185.0 + 20.1 (76) RN-670 90 24.8 -- -- 186.7 _+ 18.3 (90) AN-690 54 50.2 -- 18.5 _+11.5 (52) 160.6 q- 15.0 (54) VP-960 47 37.0 5.793 18.3 _+7.1 (19) 281.4 _+41.9 (47) B-355 42 25.4 7.829 22.1 q- 5.1 (16) 245.8 + 37.7 (42) A-260 62 33.4 11.305 -- 219.4 q- 22.1 (62) RS-1150 73 28.5 34.683 7.5 q- 1.4 (51) 186.3 + 20.7 (73) RA-450 51 35.5 30.183 7.0 + 0.9 (43) 132.8 q- 12.1 (51) with longerdistances in autumn (1043 + 206 m) 0.5 hr vs.one attempt every 17.0 + 4.1 hr, Z = 2.12, and shorter in winter (536 + 93 m). The time of N = 12, P < 0.05). More attemptswere recorded day did not haveany influence. duringautumn (N = 43, one attemptevery 2.9 hr) Most flights involvedflapping and gliding with andspring (N = 32, oneattempt every 4.1 hr) than soaringbeing recorded only at thebeginning of spring in winter (N = 3, occurringevery 12.5 hr). The and autumn. The frequencyof flights was highly interval betweentwo prey capture attemptswas very variable between individuals (H = 25.11, N = 126, variable (H = 20.09, N -- 48, P < 0.05). Moreover, P < 0.05), with intervalsbetween two flightsranging the frequencyof the attemptsincreased in relation from 24.8-95.0 min/bird (Table 2). High frequency to daysafter release(rs = 0.35, N = 48, P < 0.05). of flightswas associatedwith short rehabilitation Buzzardsgenerally hunted from perches(87.0% period(Z = 2.65, N = 11, P < 0.01). The longest of total attempts).Only 12 hunts were performed flightswere in autumnand spring(221.7 + 14.5 m by walkingor standingon theground and only three [N = 218] and 223.3 + 15.0 m [N = 234], respec- birdsdisplayed these patterns. Range of prey taken tively,in all-days),showing a significantdifference was variable,being mainly comprisedof , amongseasons (H = 12.46, N = 692, P < 0.01 in reptiles,and insects.Although the commonbuzzard all-days).Flight lengthincreased with distancefrom is well able to capturebirds (Tubbs 1974), these releasesite (rs= 0.13, N = 692, P < 0.001). Flight were not included in our hunting observationsin height and lengthwere positivelycorrelated (rs = contrastto observationsby Lovari (1974). When 0.31, N = 219, P < 0.001). huntingfrom perches,buzzards started from a mean Predatory Behavior. We recorded92 predation heightof 4.36 + 0.26 m. Neither the substratenor attempts,55 of which occurredduring days 1-3. the outcomewere related to the height. The quarry Twelve birdsout of the 16 studiedattempted to catch was caughtat a mean distanceof 13.06 + 1.11 m preyat leastonce (7.67 + 2.13 attempts/bird),with from the perch (range 2-60 m). much individual variation (one attempt every 1.3 hr The predationangle, i.e., the anglebetween the to one every33.9 hr). The four buzzardsthat died verticalfrom the perchto the groundand the path or were recapturedhad higher mean frequencies fromthe perchto the prey,supposing a linearglide, than the survivingbirds (one attempt every 2.7 + covereda wide range (0-85ø). This angle was af- 104 DAVlDE CSERMELYAND CARLOVITTORIO CORONA VOL. 28, No. 2 fectedby habitat substrateduring days 1-3 (H = specifics:317 interactionsinvolving the hoodedcrow 9.92, N = 50, P < 0.05). Uncultivated and grass (Corvuscotone), 63 involvedthe (Picapica), fields accountedfor the highest percentageof suc- and 50 thejay (Garrulusglandarius). The mean fre- cessfulattempts (N = 10). Banks of watercourses quency of interactionwith the hoodedcrow was accountedfor 42.9% of uncertain successes(N -- 14), highestin spring and lowestin autumn. The inter- but the percentageof successrelated to the grass actionswith the magpie were most frequent in au- fields was only 7.1%. Uncultivated or plowed fields tumn and very rare in winter, and thosewith the produced intermediate results. Unsuccessful at- were rare in winter but similar in the other tempts(N = 26) weremainly recordedin grassfields seasons.Interactions without regardto the bird spe- (38.4%), watercoursebanks (34.6%), and plowed ciesmost often occurred among rows of trees,ranging fields (19.2%). from 92.0% for jays to 54.6% for hoodedcrow. The The seasonstrongly influenced both the type of latter speciesalso frequently mobbedbuzzards in perchused for predationattempts (X2(3) = 42.65, P open areas (27.4%) and in other types of woods < 0.001) and the type of habitat substratewhere (17.2%). The number of mobbingindividuals was the attemptwas performed (X2o) = 8.61,P < 0.05). highly variablewith the maximum by the hooded In fact, buzzards preferred to hunt from rows of crow (up to 12 birds and up to eight in the magpie trees in spring (P < 0.001) and from pylons in and three in the jay). The corvidsinvolved in mob- autumn (P < 0.001). Most predation attempts oc- bing often performedtrue attackson the buzzard. curred on the grassfields during the cold season. The latter, however,generally paid no apparentat- Interactions with Conspecificsand Other Bird tentionto them. The mobbingrate, without regard .A totalof 29 interactionswith residentwild to the corvidspecies, varied between the seasons(X2(6) buzzardswas recordedinvolving five birdsout of the = 19.30, P < 0.01). The attacks were continuous 16 released. Most interactions occurred in summer (morethan oneattack/10 sec)in spring,at intervals and autumn (24.1% and 62.1%, respectively;cf. Kos- (lessthan one attack/10 sec)in autumn, and rare trzewa 1991), and we did not recordany interaction (lessthan one attack/60 sec)in winter. The season •n winter. Such interactions occurred soon after re- greatlyaffected both the numberof attackingbirds lease; in fact, approximately one-half occurredin and the total duration of the interaction (Table 3). days1-3. The intervalbetween two interactionsde- The maximum number of mobbingindividuals was creasedmarkedly with days (rs = -0.56, N -- 20, P much higher in spring and summer in all corvid < 0.05), reachingthe maximum value betweenday species(H = 12.85, N = 317, P < 0.001 in the 10-30 post-release.The interactionsoccurred mostly hooded crow;/-/= 9.93, N = 63, P < 0.05 in the when the releasedbuzzard was perchedand were magpie;H = 16.59, N = 50, P < 0.001 in the jay). rather variable in duration (range: 5 sec-35 min.), Corvidsmobbed longer in springand summer(jay and negativelycorrelated with perch height (rs = and magpie).The frequencyof interactiondecreased -0.66, N = 15, P < 0.05). An interaction between with number of days post-releasefor hoodedcrow two soaring birds was recordedonly once. Vocali- r -- 0.22, N -- 92, P < 0.05) but not for the magpie zations were very frequent during interactions,as andjay. Similarly, the frequencyof vocalizationsby observedalso by Tubbs (1974). the mobbinghooded crows was affectedby the season Wild buzzards attacked first in 55.1% of inter- (X2(6)= 87.48,P < 0.001). Vocalizationswere almost actions and the releasedbird attacked first only in continuous(more than one vocalization/5 sec) in 13.7% of times.Fighting, althoughof shortduration, spring,less frequent (less than one vocalization/5 occurred in 6.8% of observations. In these cases nei- sec)in winter, and virtually absentin autumn. ther buzzardshowed a tendencyto leave.Attacks by Severalother bird speciesinteracted with released the released buzzard never occurred on day one. buzzards, but these were too infrequent to allow Released buzzards that interacted with wild ones statistical evaluation. Seventeen interactions oc- scoredhigher in predationfrequency than thosenot curred with common kestrels (Falco tinnunculus), interacting(Z = 2.11, N = 12, P < 0.05). The three mostlyduring springnear the kestrels'nests. A few birdsthat interactedmost frequently eventually died interactionsoccurred with the marsh harrier (Circus or were recaptured. aeruginosus)and (Circuscyaneus) during The buzzardsin this studyinteracted with several autumn and winter in open habitats. Interactions corvidspecies much more frequentlythan with con- with sparrowhawks( nisus) occurred at the JUNE 1994 BEHAVIOROF RELEASEDBUZZARDS 105

Table 3. The mean (+SE) duration (minutes) of interactionsand the mean (+SE) number of attacking birds in three corvid speciesin each season.

HOODEDCROW MAGPIE .JAY

SEASON DURATION NO. BIRDS DURATION NO. BIRDS DURATION NO. BIRDS

Spring 5.62 ñ 0.59 1.50 ñ 0.16 4.38 4. 1.13 0.83 ñ 0.32 4.47 ñ 0.86 1.08 ñ 0.23 Summer 9.89 4. 1.21 1.56 ñ 0.16 4.37 _ 0.98 0.29 4. 0.24 3.82 _ 0.95 0.69 ñ 0.24 Autumn 9.42 ñ 1.74 0.91 4. 0.13 1.88 _ 0.31 0.06 ñ 0.04 1.05 ñ 0.21 0.05 4- 0.05 Winter 14.54 ñ 3.22 0.84 _+ 0.18 1.85 _+ 0.35 0.00 _+ 0.00 2.70 ñ 0.30 0.00 ñ 0.00

end of spring.Finally, severalnon-corvid Passeri- associatedwith a migratory behavior. Moreover, formesand two speciesof Columbidaeinteracted quick departure from the releasesite was recorded occasionally. only during the migration period. The frequencyof huntingattempts by our released DISCUSSION birds was high and possiblyunderestimated. Our Many of the buzzardswere able to survivefor data do not supportthe hypothesisby Hamilton et severalweeks around the releasesite. Although vary- al. (1988) that in red-tailed hawks (Buteojamaicen- ing in timing and direction,the abandonmentof the sis) and broad-winged hawks (Buteo lineatus) un- release site was similar to what has been recorded familiarity with the new area or captivefeeding neg- for rehabilitated congeneric American species atively affect hunting behavior. Even other (Hamilton et al. 1988). However, we recorded a parametersrelated to prey catchingability, attack greaterdistance than reportedfor buzzardsreleased glide (Wakeley 1978), and the prey attack angle in woodedhabitat (Llewellyn and Brain 1983) sug- (Janes1985) were similar to thoseof congenericwild gestingthat areaslacking large woods are likely not birds. attractivefor long-termoccupation, possibly because Predatoryproficiency of our birdslikely improved of the lack of hidingplaces. The survivalof buzzards with repetition.Prolonged captivity did not seemto for prolongedtime in this studyshows that release be detrimentalto hunting ability from perches,as in areas heavily populatedby humans is not very previously suggested in laboratory conditions detrimentalto the birds as was claimedby Hamilton (Csermelyet al. 1991). Suchan ability is shownby et al. (1988). the wide range of taxa taken as prey by our reha- On the other hand, lack of muscletone just after bilitated birds, a range very similar to the diet of releaselikely reducesthe readinessto dispersefrom wild Mediterranean populations (Lovari 1974, the releasesite (Servheenand English 1979). Low Manzi and Pellegrini 1989, Mafiosa and Cordero muscletone is certainlycaused by prolongedcaptiv- 1992). The increasedsuccess of prey capture with ity that in turn is correlatedwith the frequencyof days post-releasewas possiblyconnected to an in- flights and quick dispersal.Nevertheless, it is un- creasedknowledge of the environment.The increase likely that it inducesgreat vulnerability to the bird in hunting attemptsin migratory periodsmay have asclaimed by Duke et al. (1981). Althoughrepeated been due to increases in metabolism connected with flights in training aviaries at the RRC were very migration. Retaliation to a wild buzzard attack was important,they seemedto be inadequatefor long rare in the early post-releasedays but the frequency distanceflights soon after release.Nonetheless, good of interactions increased with time after release. In- muscletone and enduranceappeared to be achieved teractions,although frequent during reproductive and in a very few days. migration periods(Brown 1989), did not causebuz- Similar to American species(Duke et al. 1981), zards to leave the area which was oppositeof the the seasonthat releaseoccurred in clearly affected case for red-tailed hawks (Hamilton et al. 1988). the time of dispersaland the type of flight. In fact, Mobbing by corvidsseemed to causeonly the buz- althoughthe Italian populationis basicallynon-mi- zard's abandonmentof perches.This was true in gratory,the type of flightsperformed in springand spring and summer, when the corvidshave greater autumn (higher, longer,and frequentsoaring) are parental motivationtoward antipredatorybehavior 106 DAVIDE CSERMELYAND CARLO VITTORIO CORONA VOL. 28, No. 2

(R/3ell 1982). The hoodedcrow, due to its large size KENWARD,R. 1987. Wildlife radio tagging.Academic and great sociality,is the specieswith the greatest Press, London, U.K. ability to chasebuzzards. However, antipredatory KOSTRZEWA,A. 1989. Nest habitat separationin three European raptors: Accipitergentilis, Buteo buteoand behaviorby the corvidsdid not seemto havea very apivorus--A multivariate analysis. Pages 553- detrimental effect on buzzard releases. 559 in B.-U. Meyburg and R.D. Chancellor[EDS.], In conclusion,released buzzards showed a ready Raptors in the modernworld. World Working Group abilityto copewith theenvironment and to acclimate Birds Prey, , . to the wild. Preywas captured quite easily even after 1991. Interspecificinterference competition in prolongedcaptivity, although a certainlevel of train- three European raptor species.Ethol. Ecol. Evol. 3:127- ing wasevident. Moderate human presence around 143. the releasesite did not appeardetrimental. A greater KOSTRZEWA, R. AND m. KOSTRZEWA. 1991. Winter sourceof interferencelikely came from mobbing weather, spring and summerdensity, and subsequent corvids that sometimes forced buzzards to move from breeding successof Eurasian kestrels, common buz- zards, and northern goshawks.Auk 108:342-347. perches.From an appliedpoint of view we can say LLWELLYN,P.J. 1991. Assessingadult raptors prior to that the rehabilitationtechnique was basicallycor- release.Pages 33-47 in Raptor rehabilitation work- rect, becausenone of the buzzards showedevident shop.London Zoo, The Hawk Trust, The Hawk Board, behavioralmodifications related to the captivity pe- London, U.K. riod. -- AND P.F. BRAIN. 1983. Guidelines for the re- habilitationof injured raptors.Int. Zoo Year&23:121- ACKNOWLEDGMENTS 125. We are greatlyindebted to Daniele Ghillani, Maurizio Lovealii,S. 1974. The feedinghabits of four raptorsin Ravasini,and Marco Lambertini of the Italian Societyfor central Italy. Raptor Res. 8:45-57. the Protection of Birds, as well as the whole staff of the MANZI, A. AND M. PELLEGRINI. 1989. Dati sulla bio- Raptor RehabilitationCentre of Parma, for the help pro- logia riproduttiva della poiana (Buteobuteo) in un'area vided and the permissionto use the RRC facilities and della fascia collinare abruzzese. Avocetta 13:109-114. the birds housed there. We thank also the Parma Provin- cial Administrationfor the permissionto releasethe buz- MAiqOSA,S. AND P.J. CORDERO. 1992. Seasonaland zards on its territory. The commentsand criticismsby sexual variation in the diet of the common buzzard in Achim Kostrzewa, Santi Mafiosa, Gary Duke and an northeasternSpain. J. Raptor Res. 26:235-238. anonymousreferee offered many valuableinsights to im- MEYERS,J.M. AND D.L. MILLER. 1992. Post-release prove an earlier draft of the manuscript. The basicstim- activityof captive-and wild-reared bald . J. Wildl ulus for developingthis studycame from discussionswith NicolantonioAgostini. The researchwas supportedby the Manage. 56:744-749. Italian Ministero Universitfi e Ricerca Scientifica e Tec- NELSON,R.W. 1977. On the diagnosisand "cure" of nologicaand the ConsiglioNazionale delle Ricerche. imprinting in falconswhich fail to breedin captivity Pages39-49 in J.E. Cooperand R.E. Kenward[EDs.], LITERATURE CITED Paperson the veterinarymedicine and domesticbreed- BROWN,L. 1989. British birds of prey. Bloomsbury ing of diurnal birds of prey. British Falconers'Club, Oxford, U.K. Books, London, U.K. CSERMELY,D., N. AGOSTINI AND D. MAINARDI. 1991. PENDLETON, B.A.G., B.A. MILLSAP AND K.W. CLINE Predatory behaviourin captive wild Buzzard. Birds 1987. Raptor managementtechniques manual. Nat. Prey Bull. 4:133-142. Wildl. Fed., Sci. Tech. Ser. No. 10, Washington, DC U.S.A. DUKE, G.E., P.T. REDIGAND W. JONES. 1981. Recov- eries and resightingsof releasedrehabilitated raptors. R6ELL, A. 1982. A comparisonof nestdefence by jack- RaptorRes. 15:97-107. daws,rooks, and crows.Behar. Ecol. Sociobiol EMLEN,J.T. 1956. A methodfor describingand com- 11:1-6. paring avian habitats.Ibis 98:565-576. SERVHEEN,C. AND W. ENGLISH. 1979. Movements of HAMILTON,L.L., P.J. ZWANKAND G.H. OLSEN. 1988. rehabilitatedbald eaglesand proposedseasonal move- Movements and survival of released, rehabilitated mentspatterns of bald eaglesin the Pacific northwest. hawks. RaptorRes. 22:22-26. Raptor Res. 13:79-88. INGRAM,K.A. 1983. Releaseand survivalof a one-eyed SHERROD,S.K., W.R. HEINRICH, W.A. BURNHAM,J.H. goldeneagle. Ann. Proc.Am. Assoc.Zoo Vet. 1983:160. BARCLAYAND T.J. CADE. 1982. Hacking: a method JANES,S.W. 1985. Habitat selectionin raptorial birds. for releasingperegrine and otherbirds of prey. Pages159-188 in M.L. Cody [ED.], Habitat selection The PeregrineFund, Boise,ID U.S.A. in birds. Academic Press, London, U.K. SIEGEL,S. 1956. Nonparametricstatistics for the be- JUNE 1994 BEHAVIOROF RELEASEDBUZZARDS 107

havioral sciences.McGraw-Hill Book Co., New York, WEAVER,J.D AND T.J. CADE (EDS.). 1991. NY U.S.A. propagation.The Peregrine Fund, Boise,ID U.S.A. TUBBS,C.R. 1974. The buzzard. David & Charles, Lon- WHITE, G.C. AND R.A. GARROTT. 1990. Analysis of don, U.K. wildlife radio-trackingdata. AcademicPress, San Di- W^KELEY,J.S. 1978. Hunting methodsand factor af- ego, CA U.S.A. fectingtheir useby ferruginoushawks. Condor 80:327- 333. Received8 July 1993; accepted8 February 1994