Natal Dispersal of Eastern Screech-Owls’

The Condor91:254-265 0 TheCooper Ornithological Society I989

NATAL DISPERSAL OF EASTERN SCREECH-OWLS’

JAMES R. BELTHOFF~ AND GARY RITCHISON Department of BiologicalSciences, Eastern Kentucky University,Richmond, KY 40475

Abstract. Using radiotelemetrywe monitoreddispersing juvenile EasternScreech-Owls (Otus asio) in centralKentucky during 1985 and 1986.Juvenile owls (n = 16) from seven familiesremained on natal territoriesfor an average(*SE) of 55 + 1.3 daysafter fledging. The meandispersal date was 15 July, rangingfrom 8 to 2 1 July. The mean numberof days betweendispersal of the first and last membersof a brood was 4.3, rangingfrom 0 to 9 days.Juveniles (n = 17) disperseda mediandistance of 2.3 km from their nest(X = 4.4 k 1.11 km), rangingfrom 0.4 to 16.9 km, including one juvenile that continued to useportions of its natal home range. Dispersal distance was not significantly correlated with either dispersal date or the number of days that juveniles remained on natal territories. Mean dispersaldirection was 2 10 * 99. l”, and the distribution of dispersalangles did not differ significantlyfrom random. After departing from natal areas,individuals (n = 7) settled after an average of 5.6 days, ranging from 2 to 11 days. Mortality of juvenile owls was 18.2% during the period prior to dispersalbut increasedto 67% after dispersal. Key words: Natal dispersal;Eastern Screech-Owl;Otus asio;juvenile mortality; central Kentucky.

INTRODUCTION em Screech-Owlshave been restricted to the re- Dispersal, the movement from natal to first covery of birds banded as nestlings in popula- breeding site (natal dispersal) or between breed- tions that use artificial nest boxes (VanCamp and ing sites (breeding dispersal), has received rela- Henny 1975, Gehlbach 1986). Such studies pro- tively little attention from field investigators, pri- vide information concerning dispersal distances marily because of the practical difficulties and directions but offer little insight into the be- involved in following individuals as they dis- havior of dispersing individuals. Further, infor- perse. Nevertheless, the importance of dispersal mation based on band recoveries can be biased to populations has long been recognized. For ex- (Miller and Meslow 1985). ample, Gadgil (1971, p. 253) stated that, “dis- The objective of our study was to monitor dispersal is one of the most important and among persingjuvenile Eastern Screech-Owlsin central the least understood factors of population biol- Kentucky. We summarize the timing, duration, ogy,” while Horn (1983) noted that dispersaland rate, distance, and direction of natal dispersal its patterns affect all aspectsof a species’ ecology movements, and report the causes,extent, and and behavior. Although defined differently by timing of mortality. various authors, we defined natal dispersalas the STUDY AREA AND METHODS permanent movement of individuals to a new location irrespective of whether or not they re- We conducted observations between May 1985 produced after dispersal (i.e., gross dispersal; and February 1987 in and near the 680-ha Cen- Greenwood 1980). tral Kentucky Wildlife Management Area Eastern Screech-Owls (Otus asio) are widely (CKWMA), located 17 km southeast of Rich- distributed and common throughout much of the mond, Madison County, Kentucky. The man- eastern United States. Most populations appear agement area consisted of small deciduous to be nonmigratory with little movement among woodlots and thickets interspersed with culti- adult owls. Previous studies of dispersalby East- vated fields and old fields (seeBelthoff [ 19871 for a description of habitat types and dominant plant species).Areas surrounding the CKWMA were mainly agricultural, with extensively wooded 1Received 16 September1988. Final acceptance7 December1988. tracts and mountainous terrain located to the 2 Presentaddress: Department of BiologicalSci- east and southeast. ences,Clemson University, Clemson,SC 29634. We captured adult Eastern Screech-Owlseither

w41 EASTERN SCREECH-OWL DISPERSAL 255 from artificial nest boxes and natural tree cavi- engine airplane from which we either held the ties, or by luring them into mist nets by broad- two-element yagi antenna from an open window castingbounce songs(Ritchison et al. 1988). Nests or secured it to a wheel support. We traversed were located by following radio-taggedadults and areas surrounding the study area at altitudes of by examining suitable tree cavities. We removed approximately 700 m. After determining the gen- nestlings from nests and equipped each with a eral location of dispersingjuveniles in this man- radio transmitter (Wildlife Materials Inc., Car- ner, we attempted to locate juvenile owls from bondale, IL) and a U.S. Fish and Wildlife Service the ground. We tried to locate dispersing owls aluminum band several days before fledging. Ra- daily but were sometimes unsuccessful. Loca- dio transmitters were attached backpack style tions of dispersingjuveniles in rugged, inacces- (Smith and Gilbert 198 1) with woven nylon cord sible terrain were determined by the triangula- (approximately 0.25 cm in diameter). Complete tion of at least two compass bearings; owls in packages,transmitter plus harness, weighed less more accessibleterrain were visually located. We than 8 g. Entire broods were fitted with radio measureddispersal distances and directions from transmitters with one exception, the 1986 Trap the nest to either the site of first breeding attempt, Range family. Initially, we radio-tagged only two the center of an animal’s activity range during of four young in this family, but one of the radio- the fall and winter, or to the site of an individual’s taggedyoung was killed the night it fledged. We death using aerial photographs and U.S. Geo- captured and radio-tagged an additional 1986 logical Survey topographical maps of the study Trap Range juvenile 27 days later by luring it area. The sex of juvenile owls could not be de- into a mist net during playback of bounce songs termined prior to an individual’s first breeding on its natal territory. We refer to individual owls attempt, at which time we determined sex by the by either the last three digits of their U.S. Fish presenceof a brood patch or by behavior. and Wildlife Service bands or by their respective We used Mann-Whitney U-tests to examine family names. differencesin dispersal distances and the timing After young owls fledged, we located the diur- of dispersal between years and Spearman’s rank nal roost sites of each adult and juvenile an av- correlation coefficients to examine the effect of erage of four times per week in 1985 and daily dispersal date on dispersal distance (Zar 1974). in 1986 until dispersal. Thus, we were able to Linear means and their standard errors are re- determine the number of days that juveniles reported as K -t SE. The degree to which distri- mained on natal territories prior to dispersal.We butions of dispersal distances were skewed (sk) defined the date of dispersal as that date when was calculated by subtracting the median from an owl no longer roosted on its natal territory the mean, multiplying the difference by three, (juvenile owls actually dispersed the previous and dividing by the standard deviation (Glass night). Becauseowls were radio-tagged, the date and Hopkins 1984, p. 68). Mean dispersalangles of dispersalwas easily detectedand was typically (a) were calculated after Zar (1974). We calcu- unambiguous. However, three young owls de- lated angular deviation (s), a statistic analogous parted from natal territories but subsequently to standard deviation for linear data, by using returned to roost. We termed such movements the following formula (Zar 1974, p. 3 16): “exploratory” movements and, because owls s = 180/a [d-4.60517 log r] degrees, generally remained on natal territories for only a day or so following exploratory movements, where r is a measure of concentration that has we considereddispersal date as the date of initial no units but can vary from 0, when there is so departure. much dispersion that a mean angle cannot be After dispersal, we searchedfor owls in adja- described,to 1.O, when all data are concentrated cent and surrounding woodlots on foot using a in the same direction (see Zar 1974, p. 3 13 for hand-held receiver (Model TR-24, Telonics Inc., calculation of r). Mean angles and angular de- Mesa, AZ) and a two-element yagi antenna (Te- viation are given as d ? s. We applied Rayleigh’s lonics Inc.). If an owl could not be located in this test (Rayleigh’s z) to determine if significant mean manner, we drove roads surrounding the study directions occurredwithin the sampled dispersal area in an automobile with a bumper-mounted distributions or whether dispersaldirections were omni-directional antenna (Telonics Inc.). If still distributed randomly (Zar 1974). Finally, we ap- unable to locate dispersingbirds, we used a single plied the nonparametric Watson’s test to deter- 256 JAMES R. BELTHOFF AND GARY RITCHISON

TABLE 1. Timing of juvenile natal dispersalmovements in sevenfamilies of EasternScreech-Owls in central Kentucky.

Family n Days PF’ (X k SE) Range(days) Mean date Range(dates) Off-property : 62.7 + 1.45 60-65 19 July 16-21 July 1985 Muddy Creek 57.7 -c 1.20 56-60 10 July 8-12 July 1985 Trap Range‘85 3 49.0 f 2.31 45-53 13 July 9-18 July 1985 Stream 2 55.0 + 0.00 55 15 July 15 July 1986 GoosePen 22 53.0 f 0.00 53 20.5 July 20-21 July 1986 Hilltop 1 58.0 f 0.00 58 16 July 16 July 1986 Trap Range‘86 2 53.5 f 0.50 53-54 13.5 July 13-14 July 1986 Overall 16 55.4 2 1.26 45-65 15 July 8-21 July I Number of dayspostfledging when youngowls first left natal territories. 2An additional GoosePen juvenile remainedcm portions of its natal territory so we could not determinedispersal date for this individual. mine differencesin dispersal directions between ond juvenile roosted 0.5 km from its natal ter- years (Zar 1974). ritory on the first day after dispersing(owl a; Fig. 1). This individual returned to roost on its natal RESULTS territory the next day but left again that evening. We failed to locate this owl the next day but TIMING OF DISPERSAL found it roosting 1.3 km from its natal territory We captured and radio-tagged 21 young Eastern the following day (day 4 postdispersal;Fig. 1). A Screech-Owls from seven families. The mean third owl (owl g) dispersed on 13 July but was fledging date was 2 1 May (n = 2 I), ranging from not located until 15 July, when it roosted 1.1 km 14 to 30 May. Siblings either fledged during the from its natal territory (day 3; Fig. 1). We located same night or over a f-night period. Young this bird roosting 0.3 km farther from its natal screech-owlsthat escapedpredation (n = 16) re- territory 2 days later (day 5; Fig. 1). After roost- mained on natal territories for an average of 55 ing in this same general area for several days, we + 1.3 days after fledging, ranging from 45 to 65 found this owl roosting in its natal territory on days (Table 1). The mean number of days that 21 July (day 9 postdispersal). After roosting in young owls remained on natal territories did not its natal territory for two additional days, on 24 differ between years (Mann-Whitney U-test, P > July this owl returned to the same general area O.OS),averaging 56 f 2.2 days in 1985 and 54 where it had roosted on day 5 postdispersal(owl + 0.8 days in 1986. The mean dispersaldate was g; Fig. 1). 15 July (n = 16), ranging from 8 to 21 July. One juvenile (Goose Pen 344) continued to The date of dispersal varied among siblings. roost in its natal territory until 3 October, nearly For example, young owls (n = 3) in the 1985 11 weeks after its two siblings had dispersed. Trap Range family dispersed from their natal Although still in an area used by the entire family territory over a 9-day period, while young in the during the postfledging period, this young owl Goose Pen family (n = 3) dispersedover a 2-day appeared to be independent of its parents. That period (Table 1). Overall, the mean number of is, the adult male was radio-taggedand was never days between dispersal of the first and last sib- found roosting closer than about 0.4 km away lings in a family was 4 * 1.2 (n = 6 families), during the period from 19 July (about the time with no significant difference between years when this owl’s two siblings dispersed) through (Mann-Whitney U-test, P > 0.10). 10 October (when the adult male’s transmitter We recorded exploratory movements by three was removed). Although the adult female was young owls. Two owls left natal territories for not radio-tagged, she was not observed roosting one day before returning. Although the roost site with the juvenile after 28 July. On 3 October we of one of these owls could not be located the day recaptured the juvenile and replaced its trans- after dispersal, it was found back in its natal mitter. The following day we were unable to lo- territory the following day. This individual once cate a signal from this bird’s transmitter. We did, again left its natal territory that evening and was however, observe an owl roosting in a site that located at a roost site 0.8 km from its natal ter- had often been used by the juvenile, suggesting ritory the following day (owl j; Fig. 1). The sec- failure of the new transmitter rather than move- EASTERN SCREECH-OWL DISPERSAL 257

ment from the area. Subsequentattempts to capture owls in the area failed and, therefore, we are not certain if this owl made any further movements. Becausewe are uncertain if or when this individual dispersed,we excluded this owl from analysesof the timing of dispersaland calculated its “dispersal” direction and distance based on

the location of its recapture. 270'

RATE AND DURATION OF DISPERSAL Three young screech-owlswere located at roost sites the first day after dispersal from natal territories. These birds traveled 0.5, 1.1, and 3.2 km respectively (owls a, o, and k; Fig. 1). The individual that traveled 0.5 km returned to its natal territory on the following night, and was subsequentlylocated 1.3 km from its nest 3 days after its initial “exploratory” movement (owl a; Fig. 1). Three additional owls located at roost sites 2 days after dispersal were 0.8, 1.5, and 2.4 km from their respective nests (owls b, h, and 1; Fig. 1). Other owls were first located 3 to 5 days after leaving natal territories (owls c, d, e, g, and j; Fig. 1). Owls dispersing longer distances had to be located from the air and, therefore, were first located much later (21 to 54 days postdispersal) (owls f, i, p, and n; Fig. 1). 270' Many young owls settled in areas within a relatively short time after dispersing. The movements of seven owls were well-documented and the mean duration of dispersal for these individuals was 5.6 t- 1.10 days, ranging from 2 to 11 days. One owl was located 2 days after dispersal from its natal territory and was still in the same l&loo general area 98 days later when its transmitter FIGURE 1. Dispersal movements of individual ju- failed (owl 1; Fig. 1). Other owls were first located venile Eastern Screech-Owlsin central Kentucky. Let- 11,4, 5, and 6 days after dispersal and remained ters refer to individuals identified in Table 2. The cen- in the same general areas for 167, 188, 174, and tral dots represent the respective nest sites of young owls. Numbers indicate the number of days after dis- 187 days, respectively, when they were either persal that individual owls were first found at that lo- found dead or their transmitter failed (owls a, e, cation (Table 2 lists the last date that individuals were g, and h; Fig. 1). Another bird (owl i) was known known to be at that location). Outside the 4-km circle, to have moved nearly 3 km some time between refer to Table 2 for actual dispersaldistances. Individ- days 22 and 46 postdispersal(Fig. 1). After this ual m dispersedon 2 1 July 1986 and was located only once, at the position indicated, on 15 February 1987. movement, the owl apparently remained in the same general area until it was found dead nearly 6 months later. Another owl was first located 2.8 areas exhibited further movements some time km from its nest site 4 days after dispersingfrom later. One young male was first located 2 days its natal area. This owl was located 16.9 km from after dispersaland remained in the same general its nest site 74 days later (owl c; Fig. l), but we area for over 5 months. In mid-February, this made no further observations of this bird prior young owl (owl b) moved about 0.8 km from this to failure of its transmitter. area and paired with a female (Fig. 1). Another Two young owls that had apparently estab- young owl (sex unknown) had apparently settled lished ranges shortly after dispersal from natal in an area 3 days after its initial dispersal flight. 258 JAMES R. BELTHOFF AND GARY RITCHISON

-_ DISTANCE OF DISPERSAL I I s 1985 (3 Median dispersal distances for 1985 and 1986 l 1986 I i I were 1.8 km (n = 9) and 3.0 km (n = 8), re- 8 . spectively (Fig. 2). Dispersal distances were B slightly skewed (sk = 1.39) resulting in mean ._t: 10 0 1 dispersaldistances that were consistently greater than medians, averaging 4.2 * 1.60 km and 4.7 -t 1.50 km for 1985 and 1986, respectively (Ta- H.3 ; ble 2). Overall, juveniles (n = 17) moved a median distance of 2.3 km (X = 4.4 ? 1.11 km). 0 : I I I 40 50 60 70 These statistics, however, probably underesti- Days Post-fledging mate actual dispersal distances because some birds may not have settled permanently when FIGURE 2. Dispersal distancesas a function of the number of days that juvenile Eastern Screech-Owls transmitter failure occurred. Dispersal distances remained on natal territories in central Kentucky dur- did not differ significantly between years (Mann- ing 1985 (n = 9) and 1986 (n = 7). There was no Whitney U-test, P > 0.05). Becausethere were significantcorrelation between the number ofdays owls no significant differencesbetween years in either remained on natal territories and dispersaldistance (I, = 0.20, P > 0.23). Equation of plotted line is: y = number of days on the natal territory or dispersal -7.14 + 0.21~. Note: n = 2 observations at 60 days distance, we pooled data for both years to ex- postfledgingand 1.7 km. amine the relationship between the timing of dispersal and dispersal distance. We predicted a priori that early disperserswould settle closer to However, this owl moved nearly 1.4 km in early natal territories than late dispersers(i.e., Murray November, after spending over 100 days in the 1967, Waser 1985) but found no significant cor- same general area (owl o; Fig. 1). Recapture ef- relation between the number of days that indi- forts failed and transmitter failure prevented fur- viduals remained on natal territories after fledg- ther observations of this individual. ing and the distance of dispersal (r, = 0.20, P >

TABLE 2. Dispersal distancesand fates of 17 juvenile Eastern Screech-Owlsin central Kentucky. Individuals with fates listed as unknown were located only by aircraft or made additional movements and radio transmitters could not be replaced prior to their failure. Stable dispersersappeared to be settled prior to transmitter failure (after Gutitrrez et al. 1985).

Dispersal Family oE_Yd SC3 distance(km) Final sighting Fate Off-property 867 a U 2.25 1-14-86 Dead 868 b M 1.68 4-86 Nesting2 869 c U 16.90 10-4-85 Unknown Muddy Creek 870 d U 1.76 lo-lo-85 Stable 871 e U 1.80 1-14-86 Stable 872 f U 3.80 9-15-85 Stable Trap Range (1985) 877 g U 1.43 l-6-86 Dead3 878 h U 1.16 1-16-86 Dead4 879 i U 6.59 3-28-86 Dead5 Stream 347 j U 2.30 10-10-86 Stable 348 k U 7.40 8-4-86 Unknown Goose Pen 343 1 U 2.42 10-26-86 Stable 344 - U 0.40 10-3-86 Stable 345 m M? 1.05 2-15-87 Possibly paired Hilltop 342 o U 3.54 1l-3-86 Unknown Trap Range (1986) 324 p U 7.16 10-15-86 Stable 325 n U 13.00 8-4-86 Unknown I Lettersused to identify individuals in Figure 1. 2Acquired a mate and nested.Four younghatched but nonefledged. aRemains of carcassand radio transmitterfound in unidentifiedunderground mammalian burrow. 4 Scatteredfeathers and radio transmitterfound in openfield; suspectedGreat Homed Owl predation. * Radio transmitterfound on groundwith no signsof carcass. EASTERN SCREECH-OWL DISPERSAL 259

0.23, one-tailed test) (Fig. 2). Similarly, there was no significant relationship between dispersaldate and distance (rs= 0.11, P > 0.34, one-tailed test).

DIRECTION OF DISPERSAL The mean dispersal direction (n = 17) was 2 11 * 99.1”. The observed distribution of dispersal directions did not differ significantly from random (Rayleigh’s test, z = 0.86, P > 0.05, r = 0.22). There was, however, a significant difference in dispersal direction between years (Wat- son’s test, I? = 0.26, P < 0.02). Mean dispersal direction was approximately 187 f 55.0” and 332 f 85.3“ for 1985 (n = 9) and 1986 (n = 8), respectively(Fig. 3). In 1985, dispersaldirections differed significantly from random (Rayleigh’s test, z = 3.58, P -c 0.05, r = 0.63), with most individuals dispersing to the southeast and southwest (Fig. 3). Dispersal directions of the 00

1986 cohort, however, did not differ significantly 20 ,my\ from random (Rayleigh’s test, z = 0.87, P > 0.05, r = 0.33). 15 We located several owls more than once prior 10 to their apparent settling. Observation of these 5 individuals suggestthat dispersing owls move consistently in the same general direction (Fig. 270’ 900 1). One owl did reverse direction, moving back to an area occupied earlier. This individual was observed 0.8 km from its nest site 3 days postdispersal(Fig. lj). Four days later this bird moved approximately 1.6 km to another area (see day 7 location; Fig. lj) where it remained for 5 days 1986 u before returning to the area where first located 3 C@lSOO days postdispersal.After remaining in this area FIGURE 3. Dispersal directions and distancesof the for 11 days the owl returned to the area first 1985 (n = 9) and 1986 (n = 8) cohorts of Eastern occupied 7 days postdispersaland still occupied Screech-Owlsin central Kentucky. that area when its transmitter failed 88 days postdispersal. the young throughout the postfledging period. Five days after fledging, we found the first ju- MORTALITY -TIMING, CAUSES AND EXTENT venile’s radio transmitter and remains on the Mortality ofjuvenile owls after fledging and prior ground; the cause of death was unknown. The to dispersal from natal territories was 18.2%, with other owl survived for 41 days, being killed 6 four of 22 young dying. A 1986 Trap Range ju- days after injuring its left eye. The remains of venile was killed by what appeared to be a Great this owl were discovered in a tree surrounded by Homed Owl (Bubo virginianus) on the night that “whitewash,” suggestingGreat Homed Owl pre- it fledged, while a Muddy Creek juvenile suc- dation. cumbed to an unidentified mammalian predator Only one juvenile from each cohort was known 3 days after fledging. The remaining two deaths to have survived to the next breeding season. occurred in the Hilltop family after a Great Four of five juveniles known to be alive in De- Homed Owl killed the adult female 3 days before cember 1985 either starved or were killed by her young fledged,leaving the adult male to care predatorsprior to the end of March 1986 (Table for three young alone. We did not observe a re- 2). We were able to recapture and replace the placement female on the territory or near any of radio transmitter on only one 1986 juvenile but, 260 JAMES R. BELTHOFF AND GARY RITCHISON

as noted previously, the transmitter apparently ing. Moreover, one young owl in the presentstudy failed. We did, however, locate Goose Pen ju- continued to roost in its natal territory for nearly venile 345 roosting in a cavity within the terri- 11 weeks after the dispersal of its two siblings. tory of the Muddy Creek family during February These observations suggestthat young screech- 1987, possiblypaired with the adult female, which owls are not forced from parental territories due roosted nearby. to parental aggression. Previous investigators have noted an absence DISCUSSION of parent-offspring aggressionin other speciesas well. Weise and Meyer (1979) provided evidence TIMING OF DISPERSAL that parental aggressionwas not the driving or Young Eastern Screech-Owlsin the present study initiating force behind dispersal of young Black- remained on natal territories for an average of capped Chickadees (Parus atricapillus), but 55 days after fledging, with owls dispersing from suggestedthat interspecificcompetition could in- 8 through 21 July. VanCamp and Henny (1975, fluence the total distances traveled by young p. 19) suggested that young screech-owls in chickadees.Similarly, Nilsson and Smith (1985) northern Ohio “begin dispersingfrom their natal found no evidence that parental aggressionforced area in late summer and early fall . . .” Eckert young Marsh Tits (P. palustris) to initiate dis- (1974) suggestedthat young screech-owls dispersal. Furthermore, no evidence of parental perse by late August. Gehlbach (1986) reported aggressionhas been reported in the Chaffinch that juvenile screech-owls begin to disperse in (Fringih coelebs,Marler 1956) Great Tit (P. late summer in Texas. He further noted that all major; Royama 1962, Saitou 1979) or Northern young screech-owlsseem to be dispersingby Au- Harrier (Circus cyaneus, Beske 1982). In con- gust or September, as suggestedby an increase trast, Bunn et al. (1982) suggestedthat while the in the number of descendingtrills (whinny songs) oldestCommon Barn-Owls (Tyto alba) in a brood uttered by resident males. Ritchison et al. (1988) gradually drift away from natal territories, the also reported increased use of whinny songsby youngest owls in a brood are forced away by the screech-owlsduring the period of natal dispersal growing aggression of the adults. Similarly, in in central Kentucky. Hough (1960) observed an contrast to Weise and Meyer (1979) Holleback increase in the use of whinny songsbeginning in (1974) reported that the breakup of Black-capped July in New York, suggestingthe initiation of Chickadee broods may be the result of parental natal dispersal. Thus, it appears that young aggressiontoward the young or of young toward screech-owls disperse from natal territories in each other. Thus, although parent-offspring July and August throughout their range. aggressionmay be important in initiating dis- In the weeksprior to dispersal,juveniles roost- persal ofthe young in some cases,such aggression ed with their parents less often and the distance appears to be of little importance in initiating between the roost sites of young owls and their dispersal in many species. Nilsson and Smith parents increased (Belthoff 1987). In addition, (1985) suggestedthat parental aggressionis not based on calculations of biweekly home-range neededto initiate dispersalof young becausethere overlap between adult owls and their offspring, may be selectivepressure on juveniles to disperse juvenile owls began to move more widely outside as early as possible. That is, dominant individ- the home rangesof their parents during this same uals are likely to obtain better quality territories period (Belthoff 1987). Such behavior suggests or ranges and early established dispersers may that there may be little interaction between young be dominant over later dispersers. owls and their parents immediately prior to dis- The mean number of days between dispersal persal. Similarly, Gehlbach (1986) reported that of the first and last siblings for six families in the as the postfledging period continued young present study was 4.3. Similarly, siblings are re- screech-owls became increasingly independent ported to initiate dispersal over a period of sev- but were tolerated within parental territories for eral days in Eurasian Sparrowhawks (Accipiter most of the summer. In addition, three young nisus, Wyllie 1985), 0 to 31 days in Red-tailed owls in the present study returned to roost in Hawks (Buteojamaicensis, Johnson 1973), and parental territories following brief dispersal up to 15 days in Ferruginous Hawks (Buteo re- movements, with one individual roosting within galis, Konrad and Gilmer 1986). Nilsson and 20 m of its parents for 3 days before finally leav- Smith (1985) observed similar individual differ- EASTERN SCREECH-OWL DISPERSAL 261 ences in the timing of dispersal among siblings young birds to determine whether surrounding in Marsh Tits and suggestedthat such variation areasare occupiedand, if not, whether such areas might be the result of dominance status. These represent suitable habitat. Or, perhaps individ- authors found that larger individuals (assumed uals exhibiting these movements are unsuccess- to be dominants) tended to dispersebefore small- ful at locating suitable cover or adequate prey er individuals (assumedto be subdominants) and and return to natal territories where the location suggestedthat subdominants remained in paren- of both are more familiar. tal territories longer to increasetheir self-feeding ability. This increased ability is needed because RATE AND DURATION OF DISPERSAL subdominants may have to move farther before The rate of movement by young Eastern Screech- they manage to settle (as a result of losing inter- Owls during dispersal was difficult to determine actionswith more dominant individuals) and they in the present study because owls could not be are likely to end up in poorer habitats (Nilsson located every day. Even owls located relatively and Smith 1985). Other factors that could con- close to natal areas may have remained unde- tribute to variation in dispersal date among sib- tected for prolonged time periods becauseof low lings in screech-owlsinclude age (first young to sampling effort in areas with rugged topography hatch may be first to disperse), sex (males may and limited access,or perhaps such individuals disperse prior to females becausethey must es- wandered outside the study area and returned tablish territories to gain accessto females), and later. Nevertheless, our data suggestthat actively physical condition (individuals must attain the dispersingscreech-owls may travel up to 3.2 km physical condition necessaryto disperse). per night, although most appear to travel much All but one of the juvenile Eastern Screech- shorter distances (0.5-2.4 km per night). Gu- Owls in the present study dispersed from natal titrrez et al. (1985) reported that young Spotted territories, and this individual may have dis- Owls dispersingfrom natal territories traveled at persedas well. After dispersal of its two siblings an average rate of 8 km per day, with a range of in mid-July, this young owl consistently roosted 1.6 to 17 km per day. Forsman et al. (1984) about 0.4 km from the adult male. The roost reported that a young Spotted Owl left its natal sites of this young owl after mid-July were in an territory and moved 4.4 km within 2 days. area that had been used for roosting by the entire The dispersalmovements of seven young owls family during the postfledgingperiod. However, were well-documented in the present study and the family used this area on only five occasions these individuals settled in areas an average of (20 June-24 June). Thus, this apparent nondis- 5.6 days after first leaving natal territories. These perser may simply have “dispersed” to an area data, however, must be viewed with caution be- on the periphery of its parents’ territory. Similar causethey come largely from short-distance dis- behavior has been reported in other species. persers.Screech-owls dispersing longer distances Greenwood et al. (1979) observed that nearly may take longer to reach their final destination. 25% of male Great Tits establishedterritories in For example, one owl in the present study was or adjacent to their natal one. Dunstan (1970) known to have traveled nearly 3 km sometime reported a juvenile Great Horned Owl in South between days 22 and 46 postdispersal.Thus, the Dakota that apparently did not disperse, rather earliest that this bird could have settled would its parents abandoned their territory and moved have been day 23 postdispersal.Furthermore, all approximately 1.6 km to another area. Such be- but two owls in the present study either died or havior appearsto support Murray’s (1967) rule- were lost due to transmitter failure prior to es- move to the first uncontested site you find and tablishment of breeding territories. It is possible, no farther. therefore, that the young owls may have made Three young owls in the present study made further movements. In fact, two owls in the pres- exploratory movements away from natal terri- ent study did make additional movements after tories prior to final dispersal. Similar behavior spending several months in one area. Nightly has been reported in young Great Horned Owls home ranges of Eastern Screech-Owls are re- (Dunstan 1970) and Red-tailed Hawks (Johnson portedly largestduring winter (Smith and Gilbert 1973). Such behavior has not been observed in 1984, pers. observ.), so it is possible that young Spotted Owls (Strix occidentalis,Gutietrez et al. owls are forced into making further movements 1985). Exploratory movements may permit by adjacent adults (or dominant juveniles) that 262 JAMES R. BELTHOFF AND GARY RITCHISON

expand their territorial boundaries, possibly in termining dispersal distances (Murray 1967, responseto decreasedprey abundance during this Gauthreaux 1978, Moore and Ali 1984, Liberg time. Alternatively, young owls may make ad- and von Schantz 1985, Waser 1985). For ex- ditional movements during late winter and early ample, Waser (1985) suggestedthat early dis- spring in search of mates. persersmay settle closer to their natal territories Although limited, available data suggestthat than later dispersers, filling vacant areas pro- the dispersalmovements of young birds are often gressively farther from natal territories as the of relatively short duration. For example, Weise seasonprogresses. In other words, dispersingju- and Meyer (1979) noted that the dispersal phase veniles are prevented from settling in areaswhere for young Black-capped Chickadeespersisted for adults or earlier dispersingjuveniles are present. only a few weeks.Matthysen (1987) reported that Assuming that early broods dispersebefore later juvenile European Nuthatches (S&a europaeu) ones, a correlation would be expected between may settle and begin territorial defense of areas hatching date and dispersaldistance, or between lessthan 4 to 6 days after leaving natal territories hatching date and the percentageof young from (or lessthan 2 weeks after fledging). Band recov- early vs. late broods returning to breed in natal eries suggest that young sparrowhawks move areas. Such a pattern has been observed in Song chiefly during their first few weeks of indepen- Sparrows(Melospiza melodia, Arcese and Smith dent life, and are relatively sedentary thereafter 1985) House Wrens (Troglodytesaedon, Drilling (Newton and Marquiss 1983). Finally, Bowman and Thompson 1988) Marsh Tits (Nilsson and and Robe1 (1977) found that two young Greater Smith 1985) and Great Tits (Belgian and Dutch Prairie Chickens (Tympanuchus cupido) settled populations) (Dhondt and Huble 1968, Dhondt only 1 day after leaving their natal areas and two 197 1, Dhondt and Olaerts 198 1). In contrast, late others settled after only 3 days. However, these broods of Eastern Bluebirds (Siafia sialis) settle birds were not monitored throughout the fall and closer to natal areas than early broods (P. A. into the next spring, when young prairie chickens Gowaty, pers. comm.). Hatching date reportedly may make additional movements (Bowman and has no significant effect on dispersal distance in Robe1 1977). House Sparrows, Passer domesticus (Lowther 1979) Eurasian Sparrowhawks (Newton and DISTANCE OF DISPERSAL Marquiss 1983) Great Tits (Swedish and British Young Eastern Screech-Owlsin the present study populations) (Dhondt 1979, Greenwood et al. disperseda median distance of 2.3 km, ranging 1979) European Nuthatches (Matthysen and from 0.4 to 16.9 km. Similarly, Gehlbach (1986) Schmidt 1987) and Boreal Owls, Aegolius fi- reported that young Eastern Screech-Owls in nereus(Korpimaki and Lagerstrom 1988). Sim- Texas dispersedup to 14.5 km from their natal ilarly, resultsofthe present study (although based territories, although most dispersed less than 2 on dispersal dates rather than hatching dates) km. VanCamp and Henny (1975) reported that suggestno significant relationship between dis- Eastern Screech-Owlsin the northeastern United persal date and dispersal distance in young East- States dispersed a mean distance of 32 km, al- em Screech-Owls. though most moved much shorter distances. The optimal strategy for a dispersingjuvenile Other investigators have reported similar obser- may be to settle in the first unoccupied site en- vations of other species,with most individuals countered in suitable habitat, thereby limiting dispersing relatively short distances and a few costsassociated with dispersing (Murray 1967). dispersingmuch greater distances(e.g., Adamcik Although it appearsthat most young owls in the and Keith 1978, Baker and Mewaldt 1978, present study could have used such a strategy, Greenwood et al. 1979, Weise and Meyer 1979, the longer-distance dispersers almost certainly Newton and Marquiss 1983). Some investigators passedover unoccupied sitesin suitable habitats. have suggestedthat certain individuals may pos- Further, at least two young owls dispersed rela- sess innate tendencies to disperse longer dis- tively long distances (>2 km) in 1 or 2 days, tances (Howard 1960, Lidicker 1962) but such suggestingthat they were not inspecting all avail- hypotheseshave proved difficult to test (Green- able habitat for either suitability or occupancy. wood et al. 1979). Instead, contemporary hy- Similarly, Weise and Meyer (1979) observed that potheses have focused on the possible role of young Black-capped Chickadees did not simply competition and behavioral dominance in de- drift gradually from parental territories to adja- EASTERN SCREECH-OWL DISPERSAL 263 cent territories, but exhibited movements of It is also possible that many young screech-owls longer distance than would be necessarysimply are forced into suboptimal habitats (with re- for feeding or escapingthe aggressivenessof oth- ducedprey availability) by more dominant adults. er birds. It also appearsthat young House Wrens Either or both of these factors may increase the pass through unoccupied sites before settling susceptibility of young owls to starvation or in- (Drilling and Thompson 1988). Such behavior creased predation by Great Homed Owls and suggeststhat genetic factors may be influencing other predators.In any case,our results,although dispersalbehavior. Newton and Marquiss (1983) based on a small sample, suggestthat the winter speculatedthat certain aspectsof dispersal (dis- months are critical in terms of survival of young tance as well as lack of directional bias and the screech-owls. timing of main movement) may be under some High mortality rates have been reported for genetic control in sparrowhawks. first-year individuals in other species of owls. Mortality rates of first-year Tawny Owls (Strix DIRECTION OF DISPERSAL aluco) average between 47 and 66% (Southern The mean dispersal direction of young Eastern 1970). Barrowclough and Coats (1985) calculat- Screech-Owlsin the present study did not differ ed an expected first-year survivorship for Spot- significantly from random. Similarly, VanCamp ted Owls of 19%. Gutitrrez et al. (1985) radio- and Henny (1975) examined dispersaldirections tagged and followed seven dispersing Spotted of 158 screech-owlsbanded as nestlings in the Owls and all of these owls died during dispersal. northeastern United States and concluded that Miller and Meslow (1985) radio-tagged and fol- dispersaldirection was random. Our data further lowed 18 dispersingSpotted Owls and only three suggestthat young screech-owlsmoved consis- were still alive the following May. Larsen et al. tently in the same general direction during dis- (1987) reported that 78% (seven of nine) ofyoung persal. Similar behavior has been reported in Eagle Owls (Bubo bubo) releasedin Norway were dispersingSpotted Owls in California (Gutitrrez found dead within 12 weeks. Korpimaki and et al. 198 5). Weise and Meyer (1979) noted that Lagerstrom(1988) banded 4,3 11 fledgling Boreal a dispersing Black-capped Chickadee also con- Owls, 53 (1.2%) of which were recovered after tinued moving in the same direction. surviving at least their first winter. These authors further reported that the proportion of fledglings MORTALITY OF YOUNG OWLS recovered was significantly larger during the in- Two of six young owls whose fates were known crease phase of vole (Microtus and Clethriono- in the present study survived into the next breed- mys spp.) cycles than during peak, decreasing, ing season, a mortality rate of 67%. VanCamp and low phases,suggesting that food conditions and Henny (1975) estimated that the mortality during the postfledging and independence pe- rate for young screech-owlsin the northeastern riods are crucial for the survival of young Boreal United States was 69.5%. Three of the young Owls. owls in the present study were known to have died in January, while the fourth died in March. ACKNOWLEDGMENTS We recovered only one of the carcasses.An au- We thank T. D. Ritchison, E. J. Sparks, and D. T. topsy revealed an emaciated owl, with no sub- Towles for assistancein the field and J. Campbell and cutaneous or intra-abdominal fat and an empty R. Williamson for volunteeringtheir servicesas pilots. proventriculus. Although the small intestine had J. L. Ganey, P. A. Gowaty, R. J. Gutitrrez, J. H. Pliss- ner, and members of the Behavioral EcologyResearch a moderate infestation of Coccidiasp. and nema- Group at Clemson University provided constructive todes were observed in the ventriculus, the ab- comments on earlier drafts of this paper. Financial senceof fat suggeststhat the owl starved to death. assistancewas provided by Sigma Xi, The Scientific During the summer and fall months, Eastern ResearchSociety, and by EasternKentucky University. Screech-Owlsprey largely on invertebrates (Du- ley 1979, pers. observ.). As winter progressesthe LITERATURE CITED availability of invertebrates declinesand screech- ADAMCIK,R. S., AND L. B. KEITH. 1978. Regional owls begin to prey on small mammals and birds movements and mortality of Great Homed Owls in relation to Snowshoe Hare fluctuations. Can. (Duley 1979, pers. observ.). Such prey may be Field Nat. 921228-234. more difficult to capture than invertebrate prey, ARCESE,P., AND J.N.M. SMITH. 1985. Phenotypic especiallyfor young, inexperienced screech-owls. correlates and ecological consequencesof domi- 264 JAMES R. BELTHOFF AND GARY RITCHISON

nance in Song Sparrows. J. Anim. Ecol. 54:817- methods in education and psychology. 2nd ed. 830. Prentice-Hall, Englewood Cliffs, NJ. BAKER, M. C., AND L. R. MEWALDT. 1978. Song di- GREENWOOD,P. J. 1980. Mating systems,philopatry alects as barriers to dispersal in White-crowned and dispersal in birds and mammals. Anim. Be- Sparrows, Zonotrichia leucophrysnuttalli. Evo- hav. 28:1140-l 162. lution 321712-722. GREENWOOD,P. J., P. H. HARVEY,AND C. M. PERRINS. BARROWCLOUGH,G. F., AND S. L. COATS. 1985. The 1979. The role of dispersalin the Great Tit (Parus demographyand population geneticsof owls, with major): the causes,consequences and heritability specialreference to the conservationofthe Spotted of natal dispersal.J. Anim. Ecol. 48: 123-142. Owl (Strix occidentalis),p. 74-85. In R. J. Gu- GUTIBRREZ,R. J., A. B. FRANKLIN,W. LAHAYE, V. J. titrrez and A. B. Carey [eds.], Ecology and man- MERETSKY,AND J. P. WARD. 1985. Juvenile agement of the Spotted Owl in the Pacific North- Spotted Owl dispersalin northwesternCalifornia: west. Gen. Tech. Rep. PNW-185, USDA, Forest preliminary results, p. 60-65. In R. J. Gutierrez Service, Pacific Northwest For. and Range Exp. and A. B. Carey [eds.], Ecology and management Station, Portland, OR. of the Spotted Owl in the Pacific Northwest. Gen. BELTHOFF,J. R. 1987. Post-fledgingbehavior of the Tech. Rep. PNW- 185, USDA, Forest Service, Pa- EasternScreech-Owl (Ohs asio). M.S.thesis. East- cific Northwest For. and Range_ Exp.- Station. Port- ern Kentucky Univ., Richmond. land, OR. BESKE.A. E. 1982. Local and miaratorv movements HOLLEBACK.M. 1974. Behavioral interactions and of radio-taggedjuvenile Harriers. Raptor Res. 16: the dispersalof the family in Black-cappedChick- 39-53. adees. Wilson Bull. 86:466-468. BOWMAN,T. J., AND R. J. ROBEL. 1977. Brood break- HORN,H. S. 1983. Some theories about dispersal,p. up, dispersal, mobility, and mortality of juvenile 54-59. In I. R. Swingland and P. J. Greenwood Prairie~Chickens.J. Wild]. Manage. 4 1:27-34. [eds.], The ecology ofanimal movement. Oxford BUNN, D. S., A. B. WARBURTON,AND R.D.S. WILSON. Univ. Press. Oxford. 1982. The Barn Owl. Buteo Books, Vermillion, HOUGH, F. 196b. Two significant calling periods of SD. the ScreechOwl. Auk 771227-228. DHONDT,A. A. 1971. Some factors influencing ter- HOWARD,W. E. 1960. Innate and environmental dis- ritory in the Great Tit, Parus major L. Giervalk persal of individual vertebrates. Am. Midl. Nat. 61:125-135. 63:152-161. DHONDT,A. A. 1979. Summer dispersaland survival JOHNSON,S. J. 1973. Post-fledgingactivity of the Red- of juvenile Great Tits in southern Sweden. Oeco- tailed Hawk. Raptor Res. 7:43-48. logia 42:139-157. KONRAD,P. M., AND-D. S. GILMER. 1986. Post fledg- DHONDT,A. A., AND J. HUBLE. 1968. Fledging-date ing behavior of FerruginousHawks in North Da- and sex in relation to dispersal in young Great kota. Raptor Res. 20:35-39. Tits. Bird Study 15:127-l 34. KORPIMAKI,E., ANDM. LAGERSTROM.1988. Survival DHONDT, A. A., AND G. OLAERTS. 1981. Variations and natal dispersal of flednlinas of Tenamalm’s in survivaland dispersalwith ringingdate asshown Owl in relation to fluctuating food condifion and by recoveries of Belgian Great Tits Parus major. hatching date. J. Anim. Ecol. 57:433-441. Ibis 123:96-98. LARSEN,R. S., G. A. SONERUD,AND 0. H. STENSRUD. DRILLING,N. E., AND C. F. THOMPSON.1988. Natal 1987. Dispersal and mortality of juvenile Eagle and breeding dispersal in House Wrens (Troglo- Owls releasedfrom captivity in southeastNorway dytesaedon). Auk 105:480-49 1. as revealed bv radio telemetrv. v. 2 15-2 19. In R. DULEY,L. J. 1979. Life history aspectsof the Screech- W. Nero, R. J. Clark, R. J. I?napton, and R. H. Owl (Otus asio) in Tennessee.M.S.thesis. Univ. Hamre [eds.], Biology and conservationof north- of Tennessee,Knoxville. em forest owls: symposium proceedings. Gen. DUNSTAN,T. C. 1970. Post-fledgingactivities of ju- Tech. Rep. RM-142, USDA, ForestService, Rocky venile Great Homed Owls asdetermined by radio- Mountain For. and Range Exp. Station, Fort Col- telemetry. Ph.D.diss. Univ. of South Dakota, Ver- lins, CO. million. LIBERG,O., AND T. VON SCHANTZ. 1985. Sex-biased ECKERT,A. W. 1974. The owls of North America. philopatry and dispersal in birds and mammals: Doubleday and Co., Garden City, NY. the Oedipus hypothesis.Am. Nat. 126:129-135. FORSMAN,E. D., E. C. MESLOW,AND H. M. WIGHT. LIDICKER,W. Z., JR. 1962. Emigration as a possible 1984. Distribution and biology of the Spotted mechanism permitting the regulation of popula- Owl in Oregon. Wildl. Monogr. 87:1-64. tion density below carryingcapacity. Am. Nat. 96: GADGIL, M. 1971. Dispersal: population conse- 29-33. auencesand evolution. Ecoloav 52:253-26 1. LOWTHER,P. E. 1979. Growth and dispersalof nest- GAU-~HREAUX,S. A., JR. 1978. -The ecological sig- ling House Sparrows: sexual differences. Inland nificance of behavioral dominance, p. 17-54. In Bird Banding 5 1:23-29. P.P.G. Batesonand P. H. Klopfer [eds.], Perspec- MARLER,P. 1956. Behaviour of the Chaffinch Frin- tives in ethology.Vol. 3. Plenum Press,New York. gilla coelebs.Behaviour Suppl. 5: l-l 84. GEHLBACH,F. R. 1986. Odd couplesofsuburbia. Nat. MATTHYSEN,E. 1987. Territory establishmentof ju- Hist. 6:56-66. venile Nuthatches after fledging.Ardea 75:53-57. GLASS,G. V., AND K. D. HOPKINS. 1984. Statistical MATTHYSEN,E., AND K.-H. SCHMIDT. 1987. Natal EASTERN SCREECH-OWL DISPERSAL 265

dispersalin the Nuthatch. Omis. Stand. 18:313- SAITOU,T. 1979. Ecological study of social organi- 316. sation in the Great Tit, Pam major L. III. Home MILLER,G. S., AND E. C. MESLOW. 1985. Dispersal range of the basic flocks and dominance relation- data for juvenile Spotted Owls: the problem of ship of the members in a basic flock. Misc. Rep. small sample size, p. 69-73. In R. J. Gutierrez and Yamashina Inst. Omithol. 11:137-148. A. B. Carey leds.1. Ecologv and management of SMITH,D. G., ANDR. GILBERT. 1981. Backpackradio the Spotted owl in the Pacific Northwest. Gen. transmitter attachment successin Screech Owls Tech. Reu. PNW- 185. USDA. Forest Service. Pa- (Otus usio). N. Am. Bird Bander 6: 142-143. cific NoAhwest For. and Range Exp. Sta., Port- SMITH,D. G., A& R. GILBERT.1984. EasternScreech- land, OR. Owl home range and use of suburbanhabitats in MOORE,J., AND R. ALI. 1984. Are dispersal and in- southern Connecticut. J. Field Omithol. 55:322- breedingavoidance related? Anim. Behav. 32:94- 329. 112. - SOUTHERN,H. N. 1970. The natural control of a pop- MURRAY,B. G., JR. 1967. Dispersal in vertebrates. ulation of Tawny Owls (Strix uhco). J. Zool. 162: Ecologv 48:975-978. 197-285. NEWTON,i.i AND M. MARQUISS. 1983. Dispersal of VANCAMP,L. F., ANDC. J. HENNY. 1975. The Screech Sparrowhawks between birthplace and breeding Owl: its life history and population ecology in place. J. Anim. Ecol. 521463-477. northern Ohio. N. Am. Fauna 71:1-65. NILSSON,J. A., AND H. G. SMITH. 1985. Early fledg- WASER,P. M. 1985. Does competition drive dis- ling mortality and the timing of juvenile dispersal persal?Ecology 66:1171-l 175. in the Marsh Tit Puruspalustris.Omis. Stand. 16: WEISE,C. M., AND J. R. MEYER. 1979. Juvenile dis- 293-298. persaland developmentof site-fidelityin the Black- RITCHISON,G., P. M. CAVANAGH,J. R. BELTHOFF,AND capped Chickadee. Auk 96:40-55. E. J. SPARKS.1988. The singingbehaviorofEast- WYLLIE,I. 1985. Post-fledgingperiod and dispersal em Screech-Owls: seasonaltiming and response of youngSparrowhawks Accipiter nisus. Bird Study to playback of conspecificsong. Condor 90:648- 32:196-198. 652. ZAR, J. H. 1974. Biostatisticalanalysis. Prentice-Hall, ROYAMA,T. 1962. A study of the breeding behavior Englewood Cliffs, NJ. and ecology of the Great Tit, Purus major. Ph.D.diss. Tokyo Univ., Tokyo, Japan.