MARCH 2005 SHORT COMMUNICATIONS 97

E. van der Vorst Roncero translated the abstract into rique occidentale: adaptations6cologiques aux fluc- Spanish. tuations de production des 6cosystames.Terre Vie 32 89-133. LITERATURE CITED Voous, K.H. 1957. The birds of Aruba, Curagao, and Bo- BOSQUE,C. ANDM. LENTINO.1987. The passageof North naire. Stud. Fauna Curafao other Caribb. Isl. 29:1-260 American migratory land birds through xerophitic 1982. Straggling to islands--South American habitatson the westerncoast of Venezuela.Biotrop. 19: birds in the islands of Aruba, Cnracao, and Bonaire, 267-273. South Caribbean.J. YamashinaLnst. Ornithol. 14:171- FE}•C.USON-LEEs, J. XrqDD.A. CHInSTIn.2001. Raptors of 178. the world. Christopher Helm, London, U.K. --. 1983. Birds of the Netherlands Antilles. De Wal- FFRENCH,R. 1973. A guide to the birds of Trinidad and burg Press, Utrecht, Netherlands. Tobago. Livingstone Publishing Company, Wynne- ß 1985. Additions to the avifauna of Aruba, Cura- wood, OK U.S.A. qao and Bonaire, South Caribbeanß Ornithol.Monogr MLODINOW, S.G. 2004. First records of Little Egret, 36:247-254. Green-winged Teal, Swallow-tailedKite, Tennessee Z>a•I•ES,J.I. AND K.L. BraDSTEIN(EDS.). 2000. Raptor Warbler, and Red-breasted Blackbird from Aruba. N. watch: a global directory of raptor migration sites Am. Birds 57:559-561. BirdLife Conservation Series 9. BirdLife Internation- NIJMaN,V. 2001. Spatial and temporal variation in mi- al, Cambridge, U.K. and Hawk Mountain Sanctuary, grant raptorson Java,Indonesia. Emu 101:259-263. Kempton, PA UßS.A. PRINS, T.G. AND A.O. DEBROT. 1996. First record of the Canada Warbler for Bonaire, Netherlands Antilles. Caribb.J. Sci. 32:248-249. Received8 January 2004; accepted3 June 2004 TmOLL4¾,J.-M. 1978. Les migrationsdes rapacesen Af- AssociateEditor: James R. Belthoff

j RaptorRes.39(1):97-101 ¸ 2005 The Raptor ResearchFoundation, Inc.

THE RELATIONSHIPOF FORAGINGHABITAT TO THE DIET OF BARN OWLS (TYTO ALBA) FROM CENTRAL CHILE

SABINE BEGALL • Departmentof GeneralZoology, Faculty of Biologyand Geography,University of Duisbu,g-Essen, D-45117 Essen,Germany

KEYWORDS: Barn Owl;Tyto alba; diet;,Chile, feeding ecology; (Schambergerand Fulk 1974, Fulk 1976, Jaksi• et al. anthropogenicdisturbance. 1993b), the Chinchilla National Reserve (Jaksi• et al 1992), or the Atacama Desert (Jaksi• et al. 1999). D•s- Severalstudies on the diet of the Barn Owl (Tyto alba) turbancesof hunting habitatsby human activitiessuch as related to broader investigationsof predator-prey rela- agriculture or pollution may differentially affect the oc- uonshipshave been conducted in Chile, most of which currence or abundance of small mammals, and thus, the •nvolved data collected in the semi-arid north-central diet of the Barn Owl. zone of this country (e.g., Schambergerand Fulk 1974, In this study, I examine the diet of Barn Owls occu- Jaksie et al. 1992, 1993b). The majority of these studies pying two ecologically-dissimilarstudy areas. Although dealt with predator-preyinteractions, and little attention both sitesare affacted by humans, their extent of distur- was paid to human influences on the Barn Owl's diet bance differs. One study site (El Alamo), located in the (e.g., Schlatteret al. 1980, Simeone1995). Indeed, most south of central Chile, consistedmainly of meadowsand of these studies were conducted in areas of reduced hu- wasonly slightlyinfluenced by human settlement.In con- man activities such as the Fray Jorge National Park trast, the second habitat (Los Maitenes) appeared to be noticeablypolluted by the copper-processingindustry as indicated by substantiallyreduced vegetative cover. Fur- • Email: [email protected] thermore, the polluted habitat lies within the geograph- 98 SHORT COMMUNICATIONS VOL. 39, NO. 1

•cal range of previousdiet studies(Jaksi• and Y•tfiez1979, RESUI.TS Herrera and Jaksi• 1980), and thus, allowscomparisons Altogether 689 individual prey were identified from between data sets. skullsfound in 557 owl pellets,resulting in 1.24 preyper pellet. However, estimates of the minimal number of MATERIALS AND METHODS specimens based on cranial and postcranial bones ac- Owl pelletswere collectedfrom two localitiesin central counted for 1.81 animals per pellet. The Barn Owl fed Chile (El Alamo and Los Maitenes) during field studies mainly on mammals(98.1%), or more preciselyon ro- conductedin January and March 1998. At both localities, dents (96.4%); the remaining 3.6% consistedof marsu- Barn Owls were observed using the roosts from which pials (1.15%), lagomorphs(0.6%), birds (1.15%), and the pellets were collected. The study site at El Alamo insects(0.7%). A total of 12 mammal specieswas iden- (36ø11'S,72ø24'W) comprisedmainly open vegetation tified, of which 10 were rodents (Table 1). Furthermore, (meadowsgrazed by cattle, goats,and sheep) with many remains of eight birds, and elytra and chitin shellsof five shrubs and relatively complete vegetation cover (80- beetles (Coleoptera) were found in the prey. Diet 100%) and scattered fields. The landscapewas inter- breadthscalculated for the broad prey categoriesmam- spersedwith forestfragments (<20%). A total of 318 pel- mals, birds, and insectswere comparablefor both study lets were collected in abandoned barns at six sites near sites (Bobs = 1.0.5 and 1.04 for E1 Alamo and Los Maite- settlements within an area of ca. 80 km 2. nes, respectively),and thus, yielded low standard niche At Los Maitenes (32ø46'S,71ø27'W), 239 pelletswere breadths (B•, = 0.026 and 0.021 for E! Alamo and Los collected in and nearby roosts that were located in the walls of a gorge, ca. 3 km from the nearest village. The Maitenes, respectively). five sampling siteswere located within a radius of ca. 5 Mammalian prey composition differed between the km. This area was contaminatedby sulphur dioxide and two localities (Table 1). Differences in the consumed arsenic emitted by an adjacent copper-processingrefin- prey acrossthe respectiverodent families were significant ery (J. Bustamente,Empresa Nacional de Minera Ven- (X2-test;P < 0.001). Whereas the cricetine rodents (Au- tanas,pers. comm.) ca. 6 km from the samplingsite. The liscomysmicropus, Oligoryzomys longicaudatus, and Phyllot•s vegetationwas poorly developedwith only a few shrubs, darwini) were more abundant in the owl's diet from Los a reduced vegetativeground cover (<20%), and the soil Maitenes than in pellets originating from El Alamo, the thd not contain geophytes(Begall and Gallardo 2000). reversewas true for Abrothrixlongipilis (Table 1). Because The land has not been used for agriculture fbr at least the body massof the respectivespecies ranged between two decades. 30-70 g, the distributionof prey accordingto bodymass I dissected pellets after soaking them in water, and (Fig. 1) showsa slightlyhigher amount of medium-sized mammalian skullswere identified using the key by Reise prey (<70 g) for Los Maitenes.Barn Owlsfrom El Alamo (1973). In the case of uncertainty, skulls (especially seemedto take heavier prey like Rattusrattus or Octodon teeth) were compared with specificationsand drawings bridgesi,with the latter being absent in Los Maitenes.The by Osgood (1943) and for cricetine rodentswith an on- hne key (Huifia-pukiosLimitada 2002). Scientificnomen- fossorial coruro (Spalacopuscyanus) was found more of- clature followedJaksie (1997) for vertebrates. ten in pellets from Los Maitcnes (7.8%) than in the Prey per pellet wasestimated by summingup the num- southern locality El Alamo (2.0%). The differencesof bers of identified individual prey animals divided by the mammalian prey compositionin the diet of the Barn Owl total number of pelletsanalyzed. Calculations of the min- resulted in significantly lower GMWP at Los Maitenes imum number of specimensper pellet were derivedfrom (54.4 _+25.3 g) in comparisonto El Alamo (66.1 +_36.4, the maximum number of either left or right elementsof P < 0.001). cranial and postcranialbones (Lyman et al. 2003). Diet breadth (Bobs)was calculated using the term DISCUSSION (E•n=••)-1 with p• being the fractionof prey categoryi One of the samplinglocalities (Los Maitenes) lieswith- (Levins 1968). This index can yield values ranging be- in the geographicalrange of sampling localitiesof two tween 1-3 (according to N = three prey categories:mam- previous studiesaddressing the diet of the Barn Owl •n mals, birds, and insects). In addition, I calculated stan- central Chile (Jaksieand Yafiez 1979, Herrera andJaks•e dard niche breadths according to the equation Bst = 1980). in thesestudies, a slightlydifferent Barn Owl prey (Bobs -- B.... )/(Bmax -- Bmin) (Colwell and Futuyma spectrum was found compared to the present findings 1971). Geometricmean weight of prey (GMWP) wascal- Nevertheless,both this study and the previousones re- culatedusing the formula GMWP = antilog (EL, p, log vealed that the most important prey (•50%) consisted m,) where p• is defined as above and m, is the massof of Phyllotisdarwini and Oligoryzomyslongicaudatus. While prey category i. Data on body massof mammalian prey were obtained from Begall et al. (1999) fbr Spalacopus Herrera and Jaksi• (1980) fbund Abrocomabennetti and cyanus,from Redford and Eisenberg (1992) ibr Auliscomys Octodondegus in the diet of Chilean Barn Owlswith pro- micropusand Octodonsp., and from Herrera and Jaksie portions of 4% and 3%, respectively,these rather large (1980) for the remaining taxa. Means are given as • -+ octodontoids (body massof adults > 200 g) were absent SD (standard deviation), and mean prey sizeswere com- in pellets that I had analyzed. Also chiropteran skulls, pared using t-test. that Herrera and Jaksi• (1980) found in low frequencies M•Rcu 2005 SUORT COMMUNICATIONS 99

Table I. Prey of Barn Owls at two study sitesfrom central Chile (El Alamo and Los Maitenes). Data on body mass of mammals were taken froln Begall et al. (1999) for Spalacopuscyanus, from Redford and Eisenberg (1992) for Auliscomysmicropus and Octodonsp., and from Herrera and Jaksi• (1980) for the remaining taxa.

BODY MASS EL ALAMO Los MAITENES SPECIES (g) N (%) N (%) Thylamyselegans (elegant fat-tailed opossum) 40 7 (2.0) 1 (0.3) Oryctotaguscuniculus (juv.) (old-worldrabbit) 180 3 (0.8) 1 (0.3) Mus musculus(house mouse) 17 6 (1.7) 16 (4.8) Rattus rattus (black rat) 158 34 (9.6) 3 (0.9) Abrothrixlongipilis (long-haired grass mouse) 76 39 (11.0) 2 (0.6) Abrothrixolivaceus (olive-grass mouse) 40 38 (10.7) 23 (6.8) Auliscomysmicropus (southern big-eared mouse) 73 2 (0.6) 18 (5.4) Oligoryzomyslongicaudatus (long-tailed rice rat) 45 78 (22.0) 120 (35.8) Phyllotisdarwini (Darwin's leaf-earedmouse) 66 72 (20.3) 113 (33.7) Octodonbridgesi (bridge's degu) 93 60 (17.0) 0 (0.0) Octodonlunatus (moon-toothed degu) 233 0 (0.0) 2 (0.6) Spalacopuscyanus (coruro) 90 7 (2.0) 26 (7.8) (Rodents, unidentified) 2 (0.6) 3 (0.9) (Birds, unidentified) 4 (1.1) 4 (1.2) (Insects, unidentified) 2 (0.6) 3 (0.9) Total 354 335

(<1%) were not present in the samplesfrom E1 Alamo prey availability.However, I did not assessavailability of or Los Maitenes, a difference which may be related to a prey by trapping except for Spalacopuscyanus (Begall and larger sample size in the earlier study. On the other Gallardo 2000; see below). Nevertheless, different fre- hand, the cricetid rodent Auliscomysmicropus was present quencies of Oligoryzomyslongicaudatus, Phyllotis darwim, at Los Maitenes (5.4%) aswell as at El Alamo (0.6%), yet and Abrothrixolivaceus found in the pellets from the two absentin previousstudies (Jaksi• and Yf•fiez1979, Her- studysites may be related to prey availability.Rodent out- rera and Jaksifi 1980). breaks in the above mentioned speciesoccur at regular The differences between prey compositionsfound at intervalsand may depend on bamboo flowering, rainfall E1Altorio and Los Maitenes may be due to differencesin peaks, but also on density-dependentfactors (Gallardo and Mercado 1999, Lima and Jaksifi 1999, Jaksifi and Lima 2003). 9O Other striking differences between the two sampling >, 8O localities were found with Octodonbridgesi (El Alamo' m 70 17.0%; Los Maitenes:0%) and Abrothrixlongipilis (11.0% - 60 versus0.6%). While Octodonbridgesi is absent at Los Mal- • 5O tenes (and the surrounding area), Abrothrixlongipilis IS • 4O most common in moister areas (like E1 Alamo) with a • $0 high proportion of shrub and litter cover (Redford and • 2O Eisenberg 1992)--conditions that are absent at Los Mm- tenes. Furthermore, Abrothrixlongipilis was also reported

0 to exhibit cyclicpopulations (Murfia et al. 2003) that may < 30g 30- 69 g 70- 100 g > 100 g have contributed to differing frequenciesof this species between the study sites.Higher amounts of Rattus rattus Body Mass Classes remainsin pelletsfrom E1Alamo (9.6% versus0.9%) cer- Figure 1. Percentage of prey in body mass token by tainly indicated the close associationof the rat with hu- Barn Owls from two localities from central Chile (white mans. bars represent samplesfrom Los Maitenes, black bars Skullsof Spalacopuscyanus in pellets analyzedby Her- samplesfrom E1 Alamo). Data on body masswere taken rera and Jaksifi (1980) accountedfor 0.8% of the Barn froln Begall et al. (1999) for Spalacopuscyanus, from Red- Owl diet even a smaller fraction than at E1 Alamo (2%) ford and Eisenberg (1992) for Auliscomysmicropus and during the current study.In contrast to these findings, Octodonsp., and from Herrera andJaksie (1980) for the Spalacopuscyanus accounted for 7.8% of the Barn Owl remaining taxa. diet at Los Maitenes. This pattern wassurprising because 100 SHORT COMMUNICATIONS VOL. 39, NO. 1 at both studysites (Los Maitenesand El Alamo), at least mente los resultados obtenidos con las caracteristicas de five coloniesof Spalacopuscyanus were found in imme- ambos hftbitats.La regi6n de E1 Alamo consisti6princi- diate vicinity (<2 km) of sampled roosts.Furthermore, palmente en camposcultivados y prados,mientras que la mean colonysizes were similar at both localities(Begall regi6n de Los Maitenes se caracteriz6por presentaruna et al. 1999, Begalland Gallardo 2000) suggestingsimilar cubierta vegetalescasa, encontrftndose fuertemente afec- population densities.Perhaps low vegetation cover and tada por la poluci6n generada por la presencia de in- absenceof geophytesat Los Maitenesforce Spalacopusto dustrias de cobre en las cercanias. En ambas fireas de forage aboveground (on leavesof Convolvulusarvensis) estudio, T. alba mostr6 una amplitud de dieta estrecha, more frequently,and thus, exposingthem to greaterpre- representandolos mamfferosun 98% del total de presas. dation pressure.In contrast, at E1 Alamo Spalacopuscy- Sin embargo, la composici6nde la dieta vari6 marcada- anusspent most time undergroundforaging on bulbsof mente entre los dos sitlos de estudio, encontr•ndose una the geophyte Dioscorealongipes (Begall and Gallardo mayor frecuenciade roedoresfosoriales de gran tamafio 2000). Despite the absent underground-food resourceat en fireasalteradas queen aquellasareas que presentaban Los Maitenes,Spalacopus cyanus maintains its typicalcom- una cubierta vegetal densa. Finalmente, los tamafiosde plex burrow network presumablyto avoid even higher las presasen la regi6n de Los Maitenes fueron menores levelsof predation. que los obtenidosen la regi6n de E1Alamo. Although there were severalnoticeable differences be- [Traducci6n de los autores] tween the two foraging habitats,Barn Owls seemedto be equally abundant in both areas and exploited a similar ACKNOWLEDGMENTS mammalian-prey-based diet with comparable diet I thank Julio Montes who helped collect pellets.Fredy breadths. This finding indicated that sufficient mamma- Mondaca (UniversidadAustral de Chile, Valdivia) helped lian prey was also availableat the disturbedarea of Los in identifying skullsof Chilean small mammals.I greatly Maitenes for Barn Owls and corroboratesthat this species appreciated help by Oliver Erdmann and Julia Baron specializeson mammalianprey. Long-term studies byJak- (University of Duisburg-Essen)for laboratory assistance, si• et al. (1992, 1993a) revealed that if mammalian prey and Hynek Burda, Ana Trejo, and two anonymousre- decreases, Barn Owls tend to abandon the area rather viewersfor comments on the manuscript. than shift toward non-mammalian prey as seen in some LITERATURE CITED other predator species (e.g., Bednarz and Dinsmore 1985, Selas 2001). BEDNARZ,J.C. ANDJJ. DINSMORE.1985. Flexible dietary The results of this study showedthat Chilean Barn responseand feeding ecologyof the Red-shouldered Owls took prey of medium body mass.Even the mean Hawk, Buteo lineatus, in Iowa. Can. Field-Nat. 99:262- value calculated for the study site at Los Maitenes 264. (GMWP = 54.4 g) was almost twice as high as the value BEGALL,S. ANDM.H. GALIARDO.2000. Spalacopuscyanus calculatedfor prey of Barn Owls from Chilean Patagonia (Octodontidae, Rodentia): an extremist in tunnel (Iriarte et al. 1990). Higher abundancesof large animals constructing and food storing among subterranean hke Rattusrattus and Octodonbridgesi at E1 Alamo may mammals.J. Zool.(Lond.) 251:53-60. have led to significantlyhigher mean prey massat this --, H. BUP,DA, AND M.H. GALL,•mDO.1999. Reproduc- locality (GMWP = 66.1 g). Studiesof White-tailedKites tion, postnataldevelopment, and growth of socialco- (Elanus leucurus)in central Chile revealed a similar pat- ruros, Spalacopuscyanus (Octodontidae, Rodenrta) tern (Schlatter et al. 1980). They found that prey taken from Chile.J. Mammal.80:210-217. in the undisturbedhabitat was significantlyheavier than COLWELL,R.K. ANDDJ. FUTUYMA.1971. On the measure- prey taken at the disturbedarea. Schlatterand colleages ment of niche breadth and overlap. Ecology52:567- related the higher proposed that kites from anthropo- 576. genically-disturbedareas may have had accessto more FULK,G.W. 1976. Owl predation and rodent mortality' a prey due to greater prey abundanceor higher prey vul- casestudy. Mammalia 40:423-427. nerability. Whereas, I suggestthere was not higher prey GALLAm)O,M.H. ANDC.L. MERCADO.1999. Mast seeding abundance in Los Maitenes due to lower habitat quality. of bamboo shrubs and mouse outbreaks in southern Rather, I assumethat higher prey vulnerability at Los Chile. J. Neotrop.Mammal. 6:103-111. Maitenes may be supportingthe current Barn Owl pop- HERRERA,C.M. ANDF.M. JAICSI&1980. Feeding ecologyof ulation. the Barn Owl in central Chile and southern Spain. a comparativestudy. Auk 97:760-767. ECOLOG•A DE ALIMENTACION DE TYTO ALBA EN EL CENTRO HUIiqA-PUKIOS LIMITADA. 2002. Clave de indentificaci6n DE CHILE Y SU RELACION CON EL IJt/•BITAT DE CAZA anat6rnica 6seo-cranealpara g•neros de cricetidos www'geøcities'cøm/biødiversidadchile/cricetch'htm RESUMEN.--Secompar6 la composici6nde la dieta de IRIARTE,J.A., W.L. FRANKLIN,AND w.m. JOHNSON.1990. Tyroalba en dos 1ocalidadesde la regi6n central de Chile Diets of sympatricraptors in southern Chile. J. Raptor (El Alamo y Los Maitenes), relacion•ndose posterior- Res. 24:41-46. MARCH 2005 SHORT COMMUNICATIONS 101

JAICSI&,F.M. 1997. Ecologiade los vertebradosde Chile. LVMAN,R.L., E. POWER,AND RJ. LYMAN.2003. Quantifi- EdicionesUniv. Cat61icade Chile, Santiago,Chile. cation and samplingof faunal remainsin owl pellets --AND M. LIMA. 2003. Myths and facts on ratadas: J. Taphonomy1: 3-14. bambooblooms, rainfall peaksand rodent outbreaks MUROA, R., L.A. GONZALEZ,AND M. L•MA. 2003. Second- in South America. Aust. Ecol. 28:237-251. order feedback and climatic effectsdetermine the dy- --AND J.L. YANEZ.1979. The diet of the Barn Owl namics of a small rodent population in a temperate in central Chile and its relation to the availabilityof forest of South America. Popul.Ecol. 45:19-24. prey.Auk 96:619-621. OSGOOD, W.H. 1943. The mammals of Chile. Field Mu- , J.E. JIMI•NEZ,S.A. CASTRO,AND P. FEINSINGER. seum of Natural History,Chicago, IL U.S.A. 1992. Numerical and functional responseofpredators REDFORD,K.H. ANDJ.F. EISENBERG.1992. Mammalsof the to a long-termdecline in mammalianprey at a semi- Neotropics:the southern cone. Vol. 2. Univ. Chicago arid Neotropical site. Oecologia89:90-101. Press,Chicago, IL U.S.A. , P. FEINSINGER,AND J.E. JIMtNEZ. 1993a. A long- REISE,D. 1973. Clave para la determinacitn de los cr•- term studyon the dynamicsof guild structureamong neos de marsupiales y roedotes chilenos. Gayana predatoryvertebrates at a semi-aridNeotropical site. (Zool.) 27:1-20. Oikos 67:87-96. SCHAMBERGER,M. AND G. FULK. 1974. Mamiferos del , P.L. MESERVE,J.R. GUTIERREZ,AND E.L. TABILO. parquenacional Fray Jorge. Idesia 3:166-179. 1993b. The componentsof predation on small mam- SCHLATTER,R.P., B. TORO,J.L. YANEZ,AND F.M. JAICS•C mals in semiarid Chile: preliminary results.Rev. Chil. 1980. Prey of the White-tailedKite in central Chile Hist. Nat. 66:305-321. and its relation to the hunting habitat. Auk 97:186- --, J.C. TORRES-MURA,C. CORNELIUS,AND P.A. MAR- 190. QUET. 1999. Small mammals of the Atacama Desert SEEAS,V. 2001. Predation on reptiles and birds by the (Chile). J. Arid Environ.42:129-135. Common Buzzard, Buteobuteo, in relation to changes LEVINS,R. 1968. Evolution in changing environments: in its main prey,voles. Can. J. Zool.79:2086-2093. some theoretical explorations. Princeton Univ. Press, SlMEONE,A.C. 1995. Ecologia trtfica del bailaNn Elanus Princeton, NJ U.S.A. leucurusy la lechuza Tytoalba y su relacitn con la in- LIMA, M. ANDF.M. JAICSIC.1999. Population dynamicsof tervencitn humana en el sur de Chile. Tesis de licen- three Neotropical small mammals:time seriesmodels ciatura, Univ. Austral de Chile, Valdivia, Chile. and the role of delayed density-dependencein pop- ulation irruptions. Aust.J. Ecol.24:25-34. Received11 February 2004; accepted21 November2004