DECEMBER 1997 SHORT COMMUNICATIONS 387 presa m•tsfrecuente que se identific6 en las egagr6pilas lution of body size:consequences of an energetic def- es Tylomysnudicaudus, una rata grande semiarb6rea. inition of fitness. Am. Nat. 142:573-584. Otras presas incluyen insectosde diferentes tamafios, CHAVEZ TAPIA, C., L. ESPINOZA,A. DESUCRE, P. RAMIREZ, mamiferos medianosy pequefios (incluyendo un murciO- F. LOPEZMARTINEZ, V. CORTES,N. CASTILLO,V. QUIN- Iago), avesmedianas y probablementecrusfftceos. La pre- TANA,P. GUZMAN,A. VAZQUEZ,C. CHAVEZ2•ND J F sa principal probablemente es demasiadogrande para MACiAS. 1993. Estudio mastofaunistico de Cerro de que se alimenten de ella los otros bfihosque se encuen- Oro, Tuxtepec, Oaxaca. Memorias XVII Simposiode tran en la mayor parte del •trea de distribuci6n del Bfiho Biologiasde Campo y X Coloquio EstudiantilTercera Gorjiblanco. Etapa. UNAM CampusIztacala MOxico,D.E, Mexico. EMERSON, S.B., H.W. GREENE 2•ND E.L. CHARNOV. 1994. ACKNOWLEDGMENTS Allometric aspects of predator-prey interactions. The mammalremains were identified by the third author, Pages 123-139 in P.C. Wainwright and S.M. Reilly the insectremains by GuillerminaOrtega (tettigonids)and [EDS.],Ecological morphology: integrative organismal Santiago Zaragoza (Coleoptera), and the possiblecrusta- biology.Univ. Chicago Press,Chicago, IL U.S.A. cean remainsby Jos6Luis Villalobos,all from the Instituto EMMONS,L.H. 1990. Neotropical rainforest mammals' a de Biologia,UNAM, Mexico City. Bird remainswere iden- field guide. Univ. Chicago Press,Chicago, IL U.S.A. ufied by the first author with the help of Adolfo Navarro HOWELL,S.N.G. 2•NDS. WEBB.1995. A guide to the birds and EsperanzaAlvarez Mondrag6n, using the referencecol- of Mexico and northern Central America. Oxford lection of the Museo de Zoologia, Facultad de Ciencias, Univ. Press, Oxford, UI• UNAM. We are grateful to Tizoc Altarniranoof the Museo MARTI,C.D. 1974. Feeding ecologyof four sympamc de Zoologia,UNAM•Campus Iztacala,for partial support owls. Condor 76:45-61. of our fieldworkin Cerro de Oro. DalcioDacol helped find informationon the diet of thisspecies. SICK, H. 1993. Birds in Brazil. Princeton Univ. Press, Princeton, NJ U.S.A. LITERATURE CITED STILES,F.G. 2•NDA.F. SKUTCH.1989. A guide to the birds of Costa Rica. Comstock/Cornell Univ. Press, Ithaca, •tLVAREZDEL TORO, M. 1980.Las Aves de Chiapas,2nd NY U.S.A. ed. UniversidadAut6noma de Chiapas,Mexico. TWENTE,J.W. 1954. Predation on batsby hawksand owls. BAI•R, J.I• 1962. The manner and efficiency of raptor Wilson Bull. 66:135-136. depredationson bats. Condor64:500-504. WILSON,D.S. 1975. The adequacyof body sizeas a niche BOWLES,J.H. 1916. Notes on the feeding habits of the difference. Am. Nat. 109:769-784. Dusky Horned Owl. Oologist33:151-152. BROWN,J.H., P.A. MARQUETAND M.L. TAPER.1993. Evo- Received13 October 1996; accepted24July 1997.

J. RaptorRes. 31 (4):387-389 ¸ 1997 The Raptor ResearchFoundation, Inc.

NOTES ON A NEST OF THE TAWNYFISH-OWL (KETUPAFLAVIPES) AT SAKATANGSTY, T•dw•'q

YUAN-HSUN SUN I Departmentof Wildlifeand Fisheries Sciences, Texas A &M University,College Station, TX 77843 U.S.A.

YING WANG Departmentof Biology,National TaiwanNormal University,Taipei, Taiwan 117

KEITH A. ARNOLD Departmentof Wildlifeand FisheriesSciences, Texas A&M University,College Station, TX 77843 U.S.A.

KEY WORDS: TawnyFish-Owl; Ketupa flavipes; breeding; parts of Ospreys(Pandion haliaetus), fish-eagles (Ichthy- Taiwan ophagaspp.) and sea-eagles(Haliaeetus spp.). There are four speciesin the Asian genusKetupa and three species Fish-owlsare often regarded as the nocturnal counter- in the African genus Scotopelia(Fogden 1973). Of the Asian species,we know the least about the Tawny Fish- Owl (Ketupafiavipes).Kou (1986) reported one instance • Present address: Department of Forest Resource of mating on Taiwan and Voous (1988) described nest Management and Technology,National Pingtung Poly- locations and clutch sizes in India. Herein, we document technic Institute, Pingtung, Taiwan 902. the diet of a pair of TawnyFish-Owls and attempt to dem- 388 SHORT COMMUNICATIONS VOL. 31, NO. 4 onstratehow availabilityof amphibianprey speciesaffects RESULTS AND DISCUSSION the compositionof the diet on Taiwan. We flushed a Tawny Fish-Owl from the Sakatang

STUDY AREA AND METHODS Stream bed while searchingfor owl pellets in January 1994. On 7 March, a fish-owl, later identified as the male, We conductedour studyfrom January-December1994 flew into a riparian forest near where we found an oc- at SakatangStream on the easternside of Taiwan, 90 km cupied nest on 23 March. The nest was on top of an south of Taipei. The area consistsof a gorge with bluffs epiphytic bird's nest fern (Pseudorynariacoronans) on a and hills ranging up to 2400 m elevation,with tropical big large-leafedNanmu tree (Machiluskusanoi), about 80 rainforest vegetation dominated by Ficusand Lauraceae cm in dbh and about 70 m from the stream.Indigenous along streams(Taiwan Forestry Bureau 1995). A few ab- original villagesare scatteredthroughout the area (Hsu huntersindicated that theyhad previouslyseen two other 1984). owl nestsin theseferns. Gerhardt et al. (1994) reported We monitored breeding behavior at the nest from 23 that the nettropical Black-and-whiteOwl (Strixnigrolinea- March-15 May 1994. To evaluate selection of prey, we ta) would nest in the epiphytic orchid Trigonidiumeger- identified the numbersand typesof prey that the pair of tonianumin large, live trees. The bird's nest fern also TawnyFish-Owls brought to their nest.The nestwas mon- occurs in southern China, Burma, Nepal and Malaya itored using three automatic camerasat the nest and a within the distribution of other Asian fish-owls(except blind 20 m away.We kept separateobservations on prey for Blakiston'sFish-Owl, Ketupa blakistoni). Nevertheless, delivered by each of the adultsbecause we could easily these speciesdo not use bird's nest fern but insteadnest distinguishthe male by the white patch on the top of its in tree cavities, holes in river banks, caves on cliffs, forks head. Observationson the female ended on 8 May when shewas trapped and instrumentedwith a radiotransmit- of trees and abandoned nests of fish-eagles (Fogden ter. Observationson the male ended on 15 May when it 1973, Voous 1988). was killed by an indigenous hunter. Subsequently,we The nest was 10 m above ground and contained one took the owlet to an aviary on the Taiwan Normal Uni- white egg in a shallow unlined depression.To reduce versitycampus in Taipei. disturbance,we did not visitthe nest againuntil 14 April To evaluateselection of amphibianprey, we used the when we found a 2-3-wk-old downy owlet in the nest. It percentof the total of each prey speciesthat wasbrought weighed 650 g, or about 24-32% of an adult fish-owl's to the nestas an index of preyuse. We usedthe numbers weight (2050-2650 g). Subsequentobservations indicat- of individualsof each prey speciesseen per hr within 3 ed that the female brooded the young for severalhours m of either side of 5 transectsmeasuring a total of I km at a time during daylighthours. Brooding stopped on 20 as an index of prey availability.We calculatedindices for each stream habitat type along Sakatang Stream using April. over five nightsfrom 16 April-10 May 1994. Countsalong The adultsoften duetted in the vicinityof the nest0.5- transectswere usuallyfinished within 1 hr, but were al- 1 hr prior to onset of foraging. Most duets began prior wayscompleted within 3 hr after dark on nightswith little to sunset and lasted 23-70 min, but duets were also some- or no rain (<5 mm). times heard during the hr before sunrise. The female To estimatethe overall abundanceof a given amphib- gave a mewing "hew" and the male a deep "who-hot" ian along SakatangStream, we multiplied the prey avail- call. The nestling'sbegging calls, "whe," resembledthat ability index for each habitat type by the percent of the of the female's mewing, but at a lower pitch, and were total stream habitat that it comprised. Percentagesof given at night in responseto the calling parents. each habitat type were calculatedusing 5-km stream sec- Food deliveries to the nest occurred exclusivelyat tions and establishing imaginary perpendicular lines night. Here, too, deliveriescoincided with sunriseand acrossthe stream every 10 m. Along each line, we re- sunset,occurring more frequently 1 hr after dark and corded habitat type directly under the transect at I m intervals.The percent total of the points in each habitat before daybreak.Southern owl populationsusually show type was calculated as the ratio of area of each habitat this bimodal feeding periodicity in contrastto the uni- type for the 5-km stream section. Stream habitat types modal pattern of their northern conspecifics(Mikkola included: (1) low-gradient riffle, current <0.5 m/sec, 1983). Gehlbach (1994) suggestedthat in cold regions w•th rocksabove water surface<5 m apart; (2) high-gra- with shorter nightsand lower prey densitiesowls will for- dient riffle, current >0.5 m/sec, with rocks above water age intensivelyat night and are forced to hunt by day. surface<5 m apart; (3) run, current >0.5 m/sec, with Basedon 80 photographsmade of the adultsdeliver- rocksabove water surface>5 m apart; and (4) pool, cur- ing prey to the nest (Table 1), each parent'scontribution rent <0.5 m/sec, water >30 cm in depth, with rocks to the total food supplied to the nest did not change abovewater surface>5 m apart. during the nestingseason (X 2 = 3.70, df = 2, P = 0.16). We used a Chi-squareContingency Table (Neu et al. 1974, Conover 1980) to test for associationsbetween food Nevertheless,there wasa tendencyfor the female to de- deliveryrates, sex of the adult owl and ageof the nestling liver more food items during the postbrooding stage, to determine prey selection.Data were managedand an- whereas the male provided all of the food when the alyzedwith the StatisticalAnalysis System (SAS Institute young owl wasonly 1-wkold. 1989). Amphibians, followed by fish, comprised the bulk DECEMBER 1997 SHORT COMMUNICATIONS 389

Table 1. Number (percent) of feeding trips made by a Table 2. Number (percent) of prey items deliveredby nestingfemale and male TawnyFish-Owl from April-May a female and male Tawny Fish-Owl to their nest from 1994. April-May 1994.

NESTLING STAGE FEMALE MALE PREY FEMALE MALE

Brooding (14-20 d) 6 (40.0) 9 (60.0) Brown tree frog (Buergeriarobus- Early postbrooding(21-28 d) 14 (58.3) 10 (41.7) ta) and tip-nosed frog (Rana Late postbreeding(29-35 d) 28 (68.3) 13 (31.7) nadia swinhosana) 10 (47.6) 6 (28.5) Taiwan common toad (Bufo bufo gargarizans) 3 (14.3) 8 (38.1) (>75%) of the prey deliveries(Table 2). Of the amphib- Unidentified amphibians 5 (23.8) 2 (9.5) Fish (Va•icorhinusbarbatulus) 3 (14.3) 2 (9.5) ians, 11 toads (Bufo bufogarga•izans) were delivered more frequently than expected basedon their estimatedavail- Spinous country-rat (Niviventer ability (96 toads/km, Bonferroni, P < 0.05). We estimat- coxinga) 0 1 (4.8) ed that there were 291 brown tree (Buergeriarobusta) and Freshwatercrabs (Candioptamon tip-nosed frogs (Rana naria swinhosana)available along spp.) 0 1 (4.8) our 1 km transects,but only 16 were observedto be de- Freshwatershrimps (Macrobrachi- livered to the nest. As the nestling grew, the adults in- urn spp.) 0 I (4.8) creased the frequency of their food deliveries, nearly doubling their deliveriesduring the postbroodingperiod this study.Two reviewers,F. Gehlbach and H. Mikkola, (X2 = 10.6, df = 2, P = 0.005). Thereafter, deliveries provided a number of suggestionsfor improvement of declined. In addition, the proportion of larger prey items this paper. such as toadsincreased in the late postbroodingperiod. Other than an amphibian leg bone or passerinefeath- LITERATURE CITED er, we rarely found prey remains or pellets in the nest. CONOVER,WJ. 1980. Practical nonparametric statistics. However, two female toads were left in the nest twice in 2nd ed. John Wiley & Sons,New York, NY U.S.A. late April, indicating that the owlet was sensitiveto their FOGDEN,M. 1973. Fish owls,eagle owls,and the Snowy toxic glands. Owl. Pages 53-85 in J.A. Burton [ED.], Owls of the world: their evolution, structure, and ecology.A&W RESUMEN.--Nosotrosobservamos un nido de le Ketupa Visual Library, New York, NY U.S.A. flavipespara documentar su conducta de cria y sus cos- GEHLBACH, F. 1994. The Eastern Screech Owl. Texas tumbres de comer durante el tiempo de cria. Un huevo A&M Univ. Press,College Station, TX U.S.A. singular fue ponido en el nido. La hembra aparente- GERHARDT,R.P., N.B. GONZALEZ,D.M. GERHARDTAND CJ. mente hizo toda la incubaci6n.E1 par canto mas antesla FLATTEN.1994. Breeding biologyand home range of puesta del sol y a la salida del sol y cantando coincidio two Ciccaba owls. Wilson Bull. 106:629-639. con actividadde forraje. Un total de 80 entregasde presa Hsu, K. 1984. The flora of the Tarako National Park. fueron fotograficadosen el nido. E1 macho entrego mas Build. Plan. Service,Taiwan Interior Ministry (in Chi- presa cuando empezo el tiempo de poner, pero la hem- nese), Taipei, Taiwan. bra entregopresa durante el ultimo parte del tiempo de Kou, T. 1986. Spring of Tawny Fish Owl. Taiwan Bzrds poner. La presa consisti6de Bufo bufogargarizans que 1986:56-57. aparentemente no fueron tornadosen su proporci6n de MIKKOLA,H. 1983. Owls of Europe. Buteo Books, Ver- disponibilidad.La frecuenciay el tamafio de la presaen- million, SD U.S.A. tregada para el nido aumento durante el tiempo de pon- NEU, C.W., C.R. BYERSAND J.M. PEEK.1974. A technique er. for analysisof utilization-availabilitydata. J. Wzldl. [Traducci6n de Rafil De La Garza,Jr.] Manage.38:541-545.

ACKNOWLEDGMENTS SASINSTITUTE. 1989. SASuser's guide: statistics.SAS In- stitute, Cary, NC U.S.A. The following individualsassisted in the collection of TAIWAN FORESTRY BUREAU. 1995. The third forest re- the field data: P. Chiang, C. Fang, T. Fu, L. Hsiao, T. Hsu, H. Lee, L. Liao, Y. Liao, H. Mai, P. Mark, Y. Sun, T. Tin, sourceand land use inventory in Taiwan. Taiwan For- T. Wang, H. Wu, S. Wu and Y. Wu. Without their assis- estry Bureau, Taipei (in Chinese), Taipei, Taiwan. tance, this studycould not have been completed.We are Voous, K.H. 1988. Owls of the northern hemisphere. very grateful to the staff of the Tarako National Park for The MIT Press,Cambridge, MS U.S.A. their excellentservices throughout the study.The Coun- cil of Agriculture, Taiwan, provided financial support for Received6 December 1996; accepted12 August 1997