J. Field Ornithol., 62(2):239-244 RESPONSE OF MOTTLED OWLS TO BROADCAST OF CONSPECIFIC CALL RICHARD P. GERHARDT Raptor ResearchCenter BoiseState University Boise, Idaho 83725 USA Abstract.--Mottled Owls (Ciccabavirgata) were exposedto broadcastsof tape recordingsof conspecificcalls during a 6owk period roughly spanningthe nestingseason of this species in Tikal National Park, Guatemala. Their responseswere counted and the relationship between calling and certain parameters (such as weather, light and time of night) was evaluated.The owlswere quite responsiveto broadcastof conspecificvocalizations through- out the study period. Wind was the only factor, of thosetested, that affectedthe potential to hear an owl call. The high rate of response(40%) under a variety of conditionssuggests that broadcastof tapedcalls can be usedas a censustool for this species.Additionally, results indicatethat the Mottled Owl is commonin Tikal and suggeststhat this little-known species is apparently territorial. RESPUESTA POR PARTE DE INDIVIDUOS DE CICCABA VIRGATA A GRABACIONES DE LLAMADAS DE CONESPEC•FICOS Sinopsis.--Enel estudioque se 11ev6a caboen el Parque Tikal de Guatemala,individuos de buhosmoteados (Ciccaba virgata) rueton expuestospot 6 scmanas(durante el periodo que en gran mcdida correspondea la •poca de reproduci6nde la especic)a grabacioncsde 11amadasde miembrosde su propia espccie.Ademfis dc tomarsccn cuentasus rcspuestas, se evalu0 de igual manera, la relaci6n entre las 11amadasy cicrtosparfimctros talcs como condici6nclimato16gica, luz y hora dc la nochc.Los buhosmostraton gran rcspucstaalas grabacionesdc conespecfficosa travis dcl periodo dc cstudio. E1 vicnto fuc el finico factor, de los tomadosen consideraci6n,que afect6la potencialidadde escucharla 11amadadc las aves.La alta frecuenciade respuestas(40%) bajo una gran varicdadde condiciones,sugicre que el usar grabacionespuede set muy fitil comoherramicnta para haccr ccnsosde cstas aves.Los resultadosdel trabajotambi•n indicanquc estebuho cs cornfincn Tikal y quc la especieaparentementc es territorialista. Broadcastsof tapedvocalizations have been used to studyseveral aspects of owl behavior,ecology and distribution(Eastern ScreechOwl, Otusasio (Cink 1975, Nowicki 1974), Western ScreechOwl, O. kennicottiiand Elf Owl, Microthenewhitneyi (Johnson et al. 1979, 1981), SpottedOwl, Strix occidentalis(Forsman et al. 1977), and Burrowing Owl, Athenecunicularia (Martin 1973)). The Mottled Owl (Ciccabavirgata) is widely distributedthrough the Neotropics,from Mexico to northeasternArgentina (Peterson and Chalif 1973). It is the most abundant of the larger tropical owls in Mexico (Blake 1953) and probably the most commonforest owl of the humid lowlands of Honduras (Monroe 1968). Conversely,the Mottled Owl seeminglyexists in very small numbersin Guyana and Trinidad (Bu- chanan 1971) and Surinam (Hayerschmidt 1968). Little is known of the foodhabits and nestingbehavior of this nocturnalowl. Indeed,even the taxonomicposition of this and other tropical woodland speciesremains in somedoubt (Norberg 1977, Peters1938, Voous 1964). 239 240] R.P. Gerhardt J.Field Ornithol. Spring 1991 2000 •Z ! sec FIGURE1. SonGgramof male Mottled Owl (Ciccabavirgata) from Tikal National Park, Guatemala. Mottled Owls producea wide variety of calls. The territorial call is a seriesof deep hoots,described as bru bru and bu bu bu (Wetmore 1968) or keeooweeyoor cowooawoo(Ridgeley 1976). A whistledscreech has also been ascribedto this species(Eisenmann 1955). In an interestingana- tomicalstudy, Miller (1963) discoveredthat the voicebox is moreenlarged and more specializedthan that of other owls. This enablesthe Mottled Owl to producean especiallylow-pitched note for a bird of its size. In Tikal National Park, Guatemala, Mottled Owls typically utter a 4-6-note call, with one or two low, muffled preliminary notesfollowed by three higher, boomingnotes (Fig. 1). Another low, muffled note may or may not completethe series.Occasionally, a singleor doublehoot may be heard;this, however,seems never to be associatedwith a calling bout and may representan alarm call. The female also producesa cat-like yowl which seemsto be used as a food solicitationcall. No calling surveyshave been reported for Mottled Owls. In fact, Blake (1953) and Meyer de Schauenseeand Phelps (1978) provide the only previous suggestionof decoyingthis speciesby imitating its call, and Kricher (1989) statesthat this speciesdoes not tend to be attractedto tapes. I initiatedthis studyto determineif Mottled Owls respondto broadcasts of conspecificcalls and to evaluate factorsthat might influence respon- siveness. STUDY AREA AND METHODS Tikal National Park is located in the Peten forest of northern Gua- temala. This semi-deciduous,lowland forest is classifiedas part of the tropicaldry life zone(Holdridge 1957). Annual rainfall rangesfrom 1136 to 1761 mm, with distinctwet and dry seasons,the latter beingDecember- April (Smithe1966). This regionis oneof the mostornithologically diverse of Central America (Land 1970). A cassettetape was made by recordinga pair of Mottled Owls in the field at a distanceof approximately20 m. A Marantz PMD 221 recorder and SennheiserME 80 directional microphonewere used. This tape, usedfor the entire study,consisted of 4 min of a male'shoot call repeated Vol.62, No. 2 MottledOwl Response [241 13 times,followed by 1 min of a female'ssolicitation call repeated4 times. Calls were spacedat intervalsof 15-20 s, which allowed the observerto hear responsesduring the broadcastperiod. Surveyswere conductedalong 14.4 km of road in the southernhalf of the Park. Eighteenpoints were spacedat intervalsof 0.8 km along this transect(Fuller and Mosher 1987). A singlesurvey consisted of a trial at each of these 18 points. Data were collectedbetween 11 April and 22 May 1989. Three time periodswere surveyed:1930-2200 (eightsurveys), 2300-0130 (eight surveys),and 0230-0500 (sevensurveys), to determine if time of night has an effecton responsivenessof Mottled Owls. Eastern ScreechOwls have been shown to respondmore towards dawn (Beatty 1977), while nestingBarred Owls, Strix varia, are more responsivenear the middle of the night (Smith 1978). I conducteda total of 23 surveys.During 18 of the surveys,recorded Mottled Owl calls were played. Five control surveys(during which no call was played)were conductedrandomly throughout the studyperiod. Broadcastingwas done using a CassetteGame Caller (Johnny Stewart Game Calls of Waco, Texas). The speakerin this systemhas an effective frequencyrange of 275-14,000 Hz, which includesthe frequenciesfound in the owl calls (Fig. 1). Playbacklevels were measuredwith a Realistic soundlevel meter. At 1 m from the speaker,on axis, the SPL was 88 + 2 dB impulse, with settingsat c-weighting and fast response.The tape was played for 5 min at each point, after which I movedon to the next point. During controltrials, I listenedfor 5 min at each point without broadcastinga call. Upon hearing an owl (or owls) respond,the elapsedtime (since the onsetof playback) was recorded.Also recordedfor each trial point was temperature,the positionof the moonrelative to the horizon (moonphase was recordedfor each survey date), and cloud cover and wind indexed on a 0, 1, 2 scale.For wind, this correspondedto no wind, slight breeze, and wind noisein the trees,respectively. The road createda break in the canopy,which allowed me to determinecloud coverand moon position. In addition, light (including moon phase, cloud cover, and positionof moon) was indexed on a 0, 1, 2, 3 scale,with 3 being a clear, moonlit night. No surveyswere conductedduring rain. Data were analyzed using stepwiselogistic regression,which deter- minesthe effectsof severalindependent variables (which may be either categoricalor continuous)on a singledependent variable (SAS procedure Logist;Hartell 1986). Presenceor absenceof responsewas the dependent variable. Independentvariables analyzed were date, location, time of night, moonphase, cloud cover, wind and temperature.A similar analysis was also performed in which the light index was substitutedfor moon phase and cloud cover. Becausetwo different calls (male and female) were used, I could not say whether a bird respondingduring the fifth minute of playbackwas respondingto the male's call belatedlyor to the female'scall. Therefore, to avoid the possibilityof confoundingthe results,only resultsfrom the 242] R. P. Gerhardt J. Field Ornithol. Spring 1991 TABLE1. Resultsof stepwiselogistic regression. All other variables(date, time of night, temperature,moon phase and light) were not significantat the levelof P = 0.1. (Analysis performedusing SAS procedureLogist.) Variable x 2 P R Stimulus 26.46 0.0001 0.216 Wind 7.36 0.0067 - 0.101 Location 4.90 0.0268 0.074 first 4 min of the broadcast(during the broadcastof the male'svocaliza- tion) or the controltrials were usedin the analysis. RESULTS AND DISCUSSION There were 414 observations,324 trials with tapedcalls and 90 control trials. A responsewas obtainedon 136 of the 414 total trials (33%). Three independentvariables had significanteffects on the potentialfor owlsto be heard(Table 1). This was the caseboth when the regression analysiswas performed with all independentvariables and when the light variablewas substitutedfor moonphase and cloudcover. Broadcastingthe taped calls was the mostimportant factor for hearing an owl (Table 1). Owls respondedto the taped calls40% (N = 128) of the time, while owls were heard during only 9% (N = 8) of the control trials.This indicatesthat MottledOwls