The Condor 95:288-296 0 The Cooper Ornithological Society 1993

THE NEST DEFENSE BEHAVIOR OF EASTERN SCREECH-OWLS: EFFECTS OF NEST STAGE, SEX, NEST TYPE AND PREDATOR LOCATION’

THOMAS MCKJZLL SPROAT AND GARY RITCHISON Department of BiologicalSciences, Eastern Kentucky University,Richmond, KY 40475

Abstract. The responsesof male and female EasternScreech-Owls (Otus asio)to a human approachingthe nest were examined. Intensity of nest defense as measured by distance of approach, number of flights and dives, and number of vocalizations was more pronounced during the nestling stagethan during the incubation stage. These results generally support both the age-investment and positive-reinforcement hypotheses. In addition, male owls defended nestlingsmore vigorously than did females. As in many other owls, male screech- owls are smaller than females (reversed sexual dimorphism), and the increased maneuver- ability of the smaller males may have contributed to this difference. Other factors (e.g., the use of a relatively large predator), however, could also be involved. Males also responded with greater intensity as the human intruder moved closer to the nest, supporting the hypothesisthat “well-armed” parents should respond vigorously as a predator gets close to a nest. Such a response informs the potential predator of the direction it must move to reduce the likelihood of injury. Key words: EasternScreech-Owl; nest defense; Otus asio; age-investment;reversedsexual dimorphism.

INTRODUCTION unlike passerines,the value of young raptors to Nest defense behavior has been defined as “. . . predators may not increase throughout the nest- behavior that decreasesthe probability that a ling period becausetheir ability to defend them- predator will harm the contents of the nest but selves increaseswith age (Wallin 1987). Finally, which simultaneously increases the probability most raptors exhibit reversed sexual dimor- of injury or death to the parent” (Montgomerie phism (Mueller and Meyer 1985, Mueller 1986). and Weatherhead 1988). Previous studies have Most owls exhibit these characteristicsand, as a revealed that parent birds typically increase the result, their nest defense behavior might differ intensity of such defense during the period from from that of passerines. clutch initiation through fledging (Montgomerie Much of the information available concerning and Weatherhead 1988) and, further, that males the nest defense behavior of owls is qualitative, often defend nests more vigorously than females with little or no supporting data (e.g., Mikkola (Buitron 1983, Regelmann and Curio 1986, 1983). To date, only two studies have quantified Breitwisch 1988, Knight and Temple 1988, the nest defense behavior of owls (Wiklund and Westneat 1989; but see Hobson et al. 1988, Stigh 1983, Wallin 1987). Our objectives were Weatherhead 1989). to examine nest defense behavior of Eastern Most studies of nest defense have been con- Screech-Owls(Otus asio) and, where possible, to ducted with passerines. Several factors suggest compare their behavior to that of passerines.We that the nest defense behavior of raptors may specifically sought to (1) examine intensity of differ from that of passerines.For example, rap- nest defense by screech-owls throughout the tors are relatively longer-lived than passerines nesting period, (2) compare nest defense behav- (Andersen 1990) and also have the ability to at- ior of male and female screech-owls,and (3) ex- tack and injure potential nest predators (Mont- amine possible effectsof nest type and predator gomerie and Weatherhead 1988). In addition, location on nest defense behavior. METHODS I Received 17 June 1992. Accepted 30 November Nest defense behavior was examined during the 1992. breeding seasonsof 1990 and 199 1 at the Central NEST DEFENSE BY SCREECH-OWLS 289

Kentucky Wildlife Management Area, 17 km proximately 40 m of the nest (estimated using southeastof Richmond, Madison Co., Kentucky. radio-telemetry). To permit observation of Screech-owlsin central Kentucky begin egg-lay- screech-owlsin the vicinity of nests(within about ing in mid- to late March and young typically 10 m), two lanterns (each equipped with a dark leave the nest in mid- to late May (Belthoff 1987). red lens) were mounted on poles about 5 m apart Thus, experiments began in March and contin- and about 5 m in front of the nest tree. Lanterns ued until fledging of the young. were set up prior to sunset to minimize distur- Screech-owls were captured either by taking bance. All trials were conducted within 1 hr of them from nest boxes and natural cavities or by sunsetand trials were not conducted during pre- luring them into mist nets by broadcastingbounce cipitation, fog, or winds over 20 km/hr. songs(Cavanagh and Ritchison 1987, Ritchison During each trial, we noted the following. (1) et al. 1988). Captured owls were fitted with a Minimum approach distance, which is the clos- U.S. Fish and Wildlife Service aluminum band est approachby owls to the predator. (2) Number and a radio-transmitter (Wildlife Materials, Inc., of flights. All observed flights were noted. Sus- Carbondale, IL). pected flights, such as a possible change in lo- Sex was determined by observing behavior cation as determined by radio-telemetry, were during nesting. Females incubate eggsand brood not counted. (3) Number of dives. A dive was young (Gehlbach 1986, pers.observ.), while males defined as any break in horizontal flight directed typically roost in nearby trees (pers. observ.). at the predator (Knight and Temple 1988). (4) Nests were not examined prior to hatching be- Number of hits, which was the number of times cause screech-owlsdisturbed during incubation an owl struck the predator. (5) Number and type often abandon the nest (VanCamp and Henny of vocalizations. The vocal repertoire of Eastern 1975). Date of hatching was determined by plac- Screech-Owlsincludes bounce songsand whinny ing microphones in the cavities and noting the songs(Cavanagh and Ritchison 1987, Ritchison date when young were first heard. The date on et al. 1988, Sproat 1992), bark calls, screeches, which incubation began was estimated by back- and bill-claps (Gehlbach 1986, Voous 1988, dating 26 days, the average duration of incuba- Sproat 1992). Vocalizations were recordedby the tion (Sherman 1911). After hatching, each nest individual acting as the predator using a Uher was inspected twice, once during the first week 4000 Report Monitor tape recorder with a Dan post-hatching to determine the number of nest- Gibson parabolic microphone. lings and a secondtime about 2 1 to 28 days post- In addition to the pairs of screech-owlstested hatching to determine the number of fledglings. several times during the nesting cycle, six pairs Seven pairs of screech-owls(four in 1990 and (four in 1990 and two in 199 1) were tested only three in 1991) were tested four times and one once. Four of these trials were conducted during pair (199 1) was tested three times. Trials were the incubation period (two each year) and two conducted at 12 to 14 day intervals during the during the nestling period (both in 1990). Al- approximately eight-weeknesting cycle. For pairs though the limited number of such “non-repet- tested four times, Trial 1 was conducted during itive” pairs precluded comparisons with pairs the early incubation stage,Trial 2 during the late tested repetitively, data from these non-repeti- incubation stage,Trial 3 during the early nestling tive pairs were combined with those from the stage,and Trial 4 during the late nestling stage. repetitive pairs to examine the effectsof predator For the pair tested three times, no trial was con- location on nest defense behavior. ducted during the early incubation stage. Two Paired comparisons (males versus females and people were involved in each trial, with TMS egg stageversus nestling stage)were made using remaining 10 m from the nest as an observer and Wilcoxon tests (SAS Institute 1989). Multiple the secondperson approachingthe nest as a pred- comparisons (among predator locations) were ator. During each trial, the predator (equipped made using repeated measures analysis of vari- with helmet and jacket) spent 8 min at a point ance with a least-squarespost-hoc test (SAS In- 8 m in front of the nest tree, 4 min at the base stitute 1989). Spearman tests(SAS Institute 1989) of the nest tree, 4 min at a location about half were used to examine the possible relationship way between the ground and the cavity (using a between the intensity of nest defense and the ladder), and a final 4 min at the 8 m location. number of eggsand nestlings. All values are pre- Trials began when both adults were within ap- sented as mean f standard error. 290 THOMAS McKELL SPROAT AND GARY RITCHISON

TABLE 1. Comparisonof the responsesof male and femaleEastern Screech-Owls during the nestlingstage.

Minimum approachdistance (m) 11 3.36 0.93 7 6.70 1.15 2.19 0.0288 Flights 11 2.20 0.51 7 1.07 0.45 1.73 0.0840 Dives 7 0.73 0.25 1 0.20 0.20 2.22 0.0267 Hits 1 0.07 0.07 : 0.00 - 0.93 0.3506 Bouncesongs 2 0.60 0.53 0.00 1.39 0.1644 Whinny songs 10 4.67 1.42 4 1.00 0.63 2.43 0.0149 Bark calls 6 3.40 2.41 2 0.73 0.61 1.49 0.1354 Screechcalls 2 0.13 0.09 1 0.20 0.20 0.48 0.6326 a Basedon resultsof all tnals (n = 15) conductedwth eight malesand eight females(seven trials during the early nestlingstage and eight during the late nestlingstage). DNumber of individuals respondingduring the 15 trials. LStandard error.

RESULTS calls during Trial 4 (Fig. 1). Approaching females averaged about one flight per trial, made only RESPONSESDURING INCUBATION one dive, and uttered few vocalizations during All female screech-owls(seven for Trial 1 and the nestling stage(Table 1). No significantchanges eight for Trial 2) were incubating at the start of (Wilcoxon, P > 0.05) were observed between the trials. Two females left nest cavities during Trials 3 and 4 for either males or females. the trials, and only one approachedthe predator. The responsesof male screech-owlswere more During Trial 1, no males (n = 7) approachednest pronounced than those of females during the sites. Two of eight males approached the nest nestling stage(Table 1). The responsesof males during Trial 2, both to within 5 m. Males made were significantly greater for three measures only two flights during the 15 trials while females (minimum approach distance, number of dives, made only one. Neither males nor females made and number of whinny songs), and approached any dives or hits. Few vocalizations were uttered significancefor one additional measure (number during Trials 1 and 2. Females uttered only whin- of flights). ny songs(one by one female during Trial 1 and 15 by one female during Trial 2) while males uttered bounce songs (one by one male during COMPARISON OF RESPONSES Trial 1 and 12 by two males during Trial 2) Male owls responded with greater intensity dur- whinny songs(one by one male during Trial l), ing the nestling stage than during incubation, and bark calls (one by one male during Trial 1). making significantly more flights (Z = 3.05, P = No significant changes(Wilcoxon, P > 0.05) in 0.0023) and dives (z = 2.92, P = 0.0036) and behavior were observed between Trials 1 and 2 uttering significantly more whinny songs (z = for either males or females. 3.43, P = 0.0006) and bark calls (z = 2.16, P = 0.0309). Female screech-owls exhibited little RESPONSESDURING THE NESTLING STAGE change in behavior, with only the number of flights increasing significantly (z = 2.46, P = Six males approached the predator during both 0.0 139) during the nestling stages. Trials 3 and 4 (n = 8 for Trial 3 and n = 7 for Trial 4), with mean minimum approach dis- tances of 3.6 & 1.8 m and 3.1 * 1.2 m, respec- NEST DEFENSE AND NUMBER OF tively. Three females approached the predator EGGS AND NESTLINGS during both Trials 3 (n = 8) and 4 (n = 7) with The mean clutch size for screech-owls in this mean minimum approach distancesof 6.5 f 1.6 study was 3.2 ? 0.9 (n = 8, range = 2-5). The m and 7.0 f 2.1 m, respectively. Approaching mean number of nestlings was 2.9 + 1.1 (n = 8, males averaged about two flights and less than range = l-5). No significant correlations were one dive per trial (Table 1). Vocal responsesof found between any measure of nest defense and male screech-owls consisted mainly of whinny either clutch size or number of nestlings (Spear- songsduring Trial 3 and whinny songsand bark man, P > 0.05). NEST DEFENSE BY SCREECH-OWLS 291

n Bounce Songs 10.0 E!a Whinny Songs a Bark Calls E q Screech Calls p?; 6.0 &

I33 6.0

!Z % 4.0

2 2.0

FIGURE 1. Number (*SE) of vocalizations given by male Eastern Screech-Owlsduring nest defense trials. Trial 1 = early incubation, Trial 2 = late incubation, Trial 3 = early nestling, and Trial 4 = late nestling.

NEST DEFENSE AND NEST TYPE that approached significance. The single hit was Screech-owlsused both nest boxes (n = 4) and also performed by a male when the predator was natural cavities (n = 4) and the behavior of owls at the closest stop. In contrast, the behavior of using these two types of sites did not differ sig- female screech-owls did not vary significantly nificantly (Wilcoxon, P > 0.05 for all variables) (ANOVA, P > 0.05 for all variables) with pred- during the incubation stage.During the nestling ator location. stage,males using nest boxes approached signif- icantly (z = 1.97, P = 0.0484) closer to the pred- DISCUSSION ator and made significantly (z = 2.16, P = 0.0309) more flights than males using natural cavities. VARIATION IN BEHAVIOR DURING NESTING Although only approachingsignificance (z = 1.88, P = 0.0608), males using nest boxes also made The intensity of nest defense by adult Eastern more dives. No significant differences were ob- Screech-Owlsincreased during the nesting peri- served in the behavior of female screech-owls od, with males and, to a lesser degree, females using the two types of nest sites. However, two exhibiting greater responsesduring the nestling variables approached significance, i.e., females period than during incubation. Similar results using nest boxes approached closer (z = 1.67, P have been reported for several species of birds = 0.0952) to the predator and gave more (z = (reviewed by Montgomerie and Weatherhead 1.74, P = 0.08 15) whinny songs. 1988), including other raptors (Wallin 1987, Fer- rer 1990, Wiklund 1990, Andersen 1990). Fac- NEST DEFENSE AND PREDATOR tors that may contribute to an increasedintensity LOCATION of nest defenseduring the nestling period include Males performed significantly (F = 5.86, df = 3, declining renesting potential (Barash 1975, 140, P = 0.0008) more dives when the predator Weatherhead 1979), increasing nest conspicu- was closestto the nest (stop 3). Males also per- ousness and value of the young to predators formed more flights (F = 2.58, df = 3, 140, P = (Harvey and Greenwood 1978), the increasing 0.059)when the predator was closest,a difference value of the young to parents (Andersson et al. 292 THOMAS McKELL SPROAT AND GARY RITCHISON

1980) and positive reinforcement (Knight and ing the nesting period because chances of the Temple 1986a, 1986b). young surviving increasewith age and, thus, they The breeding cycle of Eastern Screech-Owlsin become increasingly valuable. This may be par- central Kentucky lasts approximately four ticularly true for Eastern Screech-Owls because months, with egg-layingbeginning in mid-March the chances of nest predation may decline after and young owls leaving parental territories be- hatching (Gehlbach 1986) and, in addition, young ginning in mid-July (Belthoff 1987, Belthoff and raptors are better able to defend themselves as Ritchison 1989). As a result of this long breeding they grow older (Wallin 1987). In agreement with cycle, screech-owls have limited renesting po- Andersson et al.‘s (1980) age-investment hy- tential (Johnsgard 1988, pers. observ.), with op- pothesis, Eastern Screech-Owls took relatively portunities for renestingprobably limited to pairs low risks during incubation and relatively greater that losenests early in the incubation period (pers. risks during the nestling period. observ.). This suggeststhat changesobserved in Andersson et al. (1980:538) also predicted that “ the intensity of nest defenseby screech-owlsdur- . . . the optimal level of parental defense in- ing the nesting period were probably not related creases during any well-defined stage of the to changesin renesting potential. breeding cycle, for example the incubation and The increasing conspicuousnessof nests (e.g., nestling periods. . . .” Eastern Screech-Owls ex- beggingcalls of nestlings may become louder and hibited little or no nest defense during the in- adults may make more trips to and from the nest cubation period, and no significant changes in as feeding rates increase)and the increasing mass behavior were observed between the early in- of nestlings (and, therefore, increasing value to cubation and late incubation trials. In addition, predators) may require an increase in the inten- disturbance of screech-owlsduring the incuba- sity of nest defense by parent birds during the tion period often causes nest abandonment nestling period (Harvey and Greenwood 1978, (VanCamp and Henny 1975). Similarly, Wallin Greig-Smith 1980). These factors seem unlikely (1987) found no evidence of nest defenseby adult to apply to Eastern Screech-Owls.Newton (1979) Tawny Owls (Strix &co) during the incubation suggestedthat most predation in raptors occurs period. In contrast, nest defense during incuba- during incubation or early in the nestling stage. tion has been reported in several passerine Predation rates may decline during the nesting species(Bjerke et al. 1985, Knight and Temple period becausenest sites may become less con- 1986b, Breitwisch 1988, Hobson et al. 1988, spicuous as trees leaf out (Montgomerie and Weatherhead 1989, Westneat 1989, Rytkonen et Weatherhead 1988). In central Kentucky, leaf al. 1990). Interspecific differences in behavior out typically begins in April and screech-owleggs may be due in part to differencesin average life- usually hatch in mid- to late April. Another fac- span. The probability of survival until the next tor influencing predation rates is nest height. breeding seasonis lower for shorter-lived species Nilsson (1984) found a negative relationship be- than for longer-lived speciesand, therefore, the tween nest height and predation rate for several cost of nest defense (i.e., loss of future repro- speciesof cavity-nesting birds, with very few high ductive successdue to injury or death) is greater nests (higher that 4.1 m) lost to predators. Belt- for long-lived species(Montgomerie and Weath- hoff and Ritchison (1990) examined nest site se- erhead 1988). Thus, relatively long-lived species lection by Eastern Screech-owlsand reported a like Eastern Screech-Owlsand Tawny Owls may mean cavity height of 6.5 m. Nests located in be less willing to take risks defending eggs.An- high cavities may be less vulnerable because dersen (1990) suggestedthat Red-tailed Hawks ground-based predators may be less likely to ei- (Buteojamaicensis) may avoid high risk behav- ther hear the begging calls of young or observe ior during nest defense because adult survival adults visiting the nest site. Finally, the growth and probability of renesting in subsequent years of young screech-owlsor other young raptorsmay are high compared to many passerines. not increase their value to predators. As young EasternScreech-Owls also exhibited no changes raptors grow older their ability to defend them- in behavior between the early nestling and late selves improves and, thus, their value to poten- nestling stages.Similarly, Andersen (1990) found tial predators may decline (Wallin 1987). that while the calling rates of Red-tailed Hawks Andersson et al. (1980) suggestedthat parent increased during the nestling stage, other mea- birds should increase levels of nest defense dur- suresof defense (dive rate and closestapproach) NEST DEFENSE BY SCREECH-OWLS 293 were not correlated with the age of nestlings. In Weatherhead 1988). Nest defense may be more contrast, the intensity of nest defense has been risky for females than males if females are weak- found to increase throughout the nestling stage, ened by the rigors of nesting (Reid and Mont- peaking at the time young leave the nest, in a gomerie 1985, Montgomerie and Weatherhead variety of passerines (Bjerke et al. 1985, 1988). Although Wallin (1987) found that the Weatherhead 1989, Rytkonen et al. 1990). The physical condition of female Tawny Owls influ- failure of Eastern Screech-Owls and Red-tailed encedthe intensity of nest defense,even the light- Hawks to increase levels of nest defense during est females showed high defense levels. Further, the nestling stage may be due to the increasing although the body mass of female owls typically ability of young raptors to either defend them- declines during brooding, such losses may be selves or flee from predators as they get older. adaptive by permitting females to serve a passive Andersen (1990) suggestedthat defending older storage role. Mass loss therefore may not be a nestlings who could hide or flee might be an in- true indicator of condition (Korpimaki 1990). appropriate responsefor adult Red-tailed Hawks. Thus, the possible effectsof nesting on the con- Similarly, Wallin (1987) suggestedthat young dition of female birds, including female screech- Tawny Owls were better able to defend them- owls, and how such effects might influence the selves as they grew older. In addition, as young intensity of nest defense remain unclear. How- raptors or other relatively large speciesincrease ever, females are the primary nest defenders in in size, the chancesthat a predator will take the severalspecies (e.g., Wallin 1987, Andersen 1990, entire brood may decrease. Ferrer 1990) suggestingthat, at least in those Knight and Temple (1986a, 1986b) suggested species,the rigors of nesting do not weaken fe- that increased levels of nest defense during the males sufficiently to preclude nest defense. nesting period may be due to positive reinforce- In populations with biased sex ratios the less ment. Recent studies,however, suggestthat such abundant sex should take fewer risks in nest de- reinforcement appeared to have little or no effect fense becausethe lifetime costs in fitness would on levels of nest defense by Northern Mocking- be higher than for the abundant sex (Montgom- birds (Mimuspolyglottos; Breitwisch 1988) Song erie and Weatherhead 1988). VanCamp and Sparrows (Melospiza melodia; Weatherhead Henny (1975) reported data suggesting that 1989) Mourning Doves (Zenaida macrouru; screech-owl populations in the northeastern Westmoreland 1989), Indigo Buntings (Passeri- United States exhibit a balanced sex ratio. Sim- na cyanea; Westneat 1989), and Willow Tits ilarly, Duley (1979) reported a balanced sex ratio (Parus montanus; Rytkonen et al. 1990). Con- for Eastern Screech-Owls in Tennessee. Al- trary to the predictionsof this hypothesis,screech- though not conclusive, such data suggestthat owls in our study exhibited no significant changes differencesin the nest defense behavior of male in behavior between trials conducted during the and female screech-owlsare not the result of a same breeding stage (i.e., during the incubation biased sex ratio. stageor nestling stage).If birds are rewarded for Differences in the ability of males and females heightened aggressiveness,then a continuous in- to raise offspring unaided may also contribute to crease in response levels might be evident be- sexual differences in nest defense behavior. A tween successivetrials. However, these results parent whose mate would be unable to raise a are only suggestive,and our limited sample of brood on its own should take less risk in nest pairs tested only once does not permit a direct defense (Montgomerie and Weatherhead 1988). test of the positive reinforcement hypothesis. Male Eastern Screech-Owlsdo not incubate eggs Thus, both the age-investment hypothesis and or brood young (Bent 1938; Voous 1988, pers. the positive reinforcement hypothesis are sup- observ.). Further, nestling screech-owls cannot ported by our results. thermoregulate until 14-l 6 days post-hatching (Lohrer 1985) suggestingthat female screech- SEX DIFFERENCES IN NEST owls must brood young for about two weeksafter DEFENSE BEHAVIOR hatching. Thus, male screech-owlswould prob- Male screech-owlsdefended nestlings more vig- ably not be able to raise young unaided until orously than did females. Several factors may about halfway through the four-week nestling pe- contribute to such sexual differences, including riod. This may have contributed to the low levels differencesin risk perception (Montgomerie and of nest defenseexhibited by female screech-owls 294 THOMAS McKELL SPROAT AND GARY RITCHISON during the incubation and early brooding peri- cavity) may help explain the greater responseof ods. However, the ability of males and females male screech-owls.Clearly, additional studiesare to raise offspring unaided would seem to be sim- needed to determine how predator size, habitat ilar during the late nestling stage. Despite this, characteristics, and predator location influence the responses of male screech-owls remained the responses of males and females in species greater than those of females. Similar observa- exhibiting sexual size dimorphism. tions have been reported for Great Tits (Purus major; Regelmann and Curio 1986). Such results NEST DEFENSE AND NEST TYPE suggestthat factors other than the ability to raise The increasedintensity of nest defenseexhibited offspring unaided are responsible for differences by screech-owlsusing nest boxes may have been in the nest defense behavior of males and fe- due in part to the greater visibility of such nest males. sites. Montgomerie and Weatherhead (1988) Like many raptors, Eastern Screech-Owls ex- suggestedthat crypticity could influence the in- hibit reversed sexual dimorphism (Henny and tensity of nest defensebehavior, with individuals VanCamp 1979) and this difference in relative in cryptic nests responding less than those with size could influence the nest defensebehavior of more exposed nests. In support of this hypoth- males and females. Flights and dives were im- esis, Ricklefs (1977) reported that Panamanian portant components of the nest defensebehavior birds exhibiting the most vigorous defense were of screech-owls, and the smaller size and, per- those with conspicuous nests. Similarly, Amer- haps, greater maneuverability of males may re- ican Robins (Turdus migratorius) with exposed duce the risk involved in performing such be- nests responded to potential predators with sig- haviors. Such differencesin maneuverability may nificantly more swoopsand hits than those with also contribute to sexual differences in the nest cryptic nests (McLean et al. 1986). defense behavior of Snowy Owls (Nyctea scan- diaca; Wiklund and Stigh 1983) and Rough-leg- NEST DEFENSE AND PREDATOR ged Hawks (Buteo lagopus;Andersson and Wik- LOCATION lund 1987). However, the larger size of many Montgomerie and Weatherhead (1988: 183) sug- female raptors may predisposethem to nest de- gested that “well armed” (raptorial) parents fense (Andersson and Norberg 198 1). Results in should respondvigorously as a predator getsclose support of this hypothesis have been reported to the nest because such a responseinforms the (Wallin 1987, Andersen 1990, Ferrer 1990). Sev- predator of a willingness to attack. If the intensity eral factors may contribute to interspecific of response increases as a potential predator differences in the behavior of male and female moves closer to the nest, then the direction that raptors. For example, Montgomerie and Weath- the predator must move to reduce the likelihood erhead (1988) suggestedthat responding to a rel- of injury should be clear (Montgomerie and atively small predator may be less risky for a Weatherhead 1988). In support of this hypoth- larger parent while respondingto a relatively large esis, we found that the intensity of response by predator may be less risky for a smaller, more male screech-owls increased as the predator maneuverable parent. Habitat characteristics moved closer to the nest. could also be a factor. Larger parents may re- spond to a potential predator with flights and dives in open habitats, while smaller, more ma- ACKNOWLEDGMENTS neuverable parents may respond in habitats with We thank D. D. Able, A. Bowen, T. Buck, B. Moore more obstructions (e.g., woodlots or forests).The and, especially, P. H. Klatt for assistancein the field. location of a predator relative to a nest site could We also thank P. J. Weatherhead and an anonymous reviewer for helpful comments on the manuscript. Fi- also influence responselevels. Smaller, more ma- nancial support was provided by Eastern Kentucky neuverable parents may use flights and dives to University. distract approaching predators, while larger par- ents may use flights, dives, and even hits to deter LITERATURE CITED predators in the immediate vicinity of a nest. ANDERSEN,D. E. 1990. Nest-defensebehavior of Red- Thus, our choices of study site (woodlots) and tailed Hawks. Condor 92:991-997. methodology (using a relatively large predator, ANDERSSON,D. E., AND R. A. NORBERG.198 1. Evo- humans, that did not go all the way to the nest lution of reversed size dimorphism and role par- NEST DEFENSE BY SCREECH-OWLS 295

titioning among predatory birds, with a size scal- KNIGHT, R. L., AND S. A. TEMPLE. 1986b. Nest de- ing of flight performance. Biol. J. Linn. Sot. 15: fence in the American Goldfinch. Anim. Behav. 105-l 30. 34:887-897. ANDERSSON,M., C. G. WIKLUND,AND H. RUNDGREN. KNIGHT,R. L., ANDS. A. TEMPLE. 1988. Nest-defense 1980. Parental defence of offspring:a model and behavior in the Red-winged Blackbird. Condor 90: an example. Anim. Behav. 281536-542. 193-200. ANDERSSON,S., AND C. G. WIKLUND. 1987. Sex role KORPIMAKI,E. 1990. Body mass of breeding Teng- partitioning during offspring protection in the malm’s Owl Aegoliusfunereus: seasonal, between Rough-legged Buzzard Buteo lagopus. Ibis 129: year, site and age related variation. Omis Stand. 103-107. 21:169-178. BARASH,D. P. 1975. Evolutionary aspectsofparental LQHRER,F. E. 1985. Ontogeny of thermoregulation behavior: distraction behavior of the Alpine Ac- in the Eastern Screech-Owl.J. Field Omithol. 56: centor. Wilson Bull. 871367-373. 65-66. BELTHOFF,J. R. 1987. Post-fledgingbehavior of the MCLEAN, I. G., J.M.N. SMITH, AND K. G. STEWART. Eastern Screech-Owl. M.Sc.thesis. Eastern Ken- 1986. Mobbing behaviour, nest exposure, and tucky Univ., Richmond, KY. breeding successin the American Robin. Behav- BELTHOFF,J. R., ANDG. RITCHISON.1989. Natal dis- iour 96:171-186. nersal of Eastern Screech-Owls. Condor 9 1:254- MIKKOLA, H. 1983. Owls of Europe. Buteo Books, i65. Vermillion, SD. BELTHOFF,J. R., AND G. RITCHISON. 1990. Nest-site MONTGOMERIE,R. D., ANDP. J. WEATHERHEAD.1988. selectionby Eastern Screech-Owlsin central Ken- Risks and rewardsof nest defenceby parent birds. tucky. Condor 92:982-990. Ouart. Rev. Biol. 63:167-187. BENT,A. C. 1938. Life histories of North American MUELLER,H. C. 1986. The evolution of reversedsex- birds of prey, part 2. U.S. Natl. Mus. Bull. 170, ual dimorphism in owls: an empirical analysis of Washington, DC. oossibleselective factors. Wilson Bull. 98:387-406. BJERKE,T., Y. ESPMARK,ANDT. FONSTAD.1985. Nest MUELLER,H. C., AND K. MEYER. 1985. The evolution defence and parental investment in the Redwing, of reversed sexualdimorphism in size: a compar- Turdus iliacus. Omis Stand. 16:14-l 9. ative analysisof the Falconiformes of the western BREITWISCH,R. 1988. Sex differences in defence of Palearctic,p. 65-l 0 1. In R. F. Johnston[ed.], Cur- eggsand nestlingsby Northern Mockingbirds, Mi- rent ornithology, Vol. 2. Plenum Press,New York. mus polyglottos.Anim. Behav. 36~62-72. NEWTON,I. 1979. Population ecologyof raptors. Bu- BUITRON,D. 1983. Variability in the response of teo Books, Vermillion, SD. Black-billed Magpies to natural predators. Behav- NILSSON,S. G. 1984. The evolution of nest-site se- iour 781209-236. lection by hole-nesting birds: the importance of CAVANAGH,P. M., AND G. RITCHISON. 1987. Vari- nest predation and competition. Omis Stand. 15: ation in the bounce and whinny songsof the East- 167-175. em Screech-Owl. Wilson Bull. 99:620-627. REGELMANN,K., ANDE. CURIO. 1986. Why do Great DULEY,L. J. 1979. Life history aspectsof the screech Tit (Parus major) males defend their brood more owl (&us asio) in Tennessee.M.Sc.thesis, Univ. than females ho? Anim. Behav. 34:1206-1214. of Tennessee,Knoxville, TN. REID. M. L.. AND R. D. MONTGOMERIE.1985. Sea- FERRER,M. 1990. Nest defense by male and female sonal patternsof nest defence by Baird’s Sandpip- SpanishImperial Eagles.J. Raptor Res. 24:77-79. ers. Can. J. Zool. 63:2207-22 11. GEHLBACH,F. R. 1986. Odd couplesofsuburbia.Nat. RICKLEFS,R. E. 1977. Reactions of some Panamani- Hist. 95(6):56-66. an birds to human intrusion at the nest. Condor GREIG-SMITH,P. W. 1980. Parental investment in 791376-379. nest defence by Stonechats (Saxicola torquata). RITCHISON,G., P. M. CAVANAGH,J. R. BELTHOFF,AND Anim. Behav. 28:604-6 19. E. J. SPARKS.1988. The singingbehavior of East- HARVEY,P. H., AND P. J. GREENWOOD.1978. Anti- em Screech-Owls: seasonaltiming and response predator defence strategies: some evolutionary to playback of conspecificsong. Condor 90:648- problems, p. 129-15 1. In J. R. Krebs and N. B. 652. Davies [eds.],Behavioral ecology:an evolutionary RYTKONEN,S., K. KOIVULA, AND M. ORELL. 1990. approach. Blackwell Scientific, Oxford, England. Temporal increasein nest defence intensity of the HENNY, C. J., AND L. F. VANCAMP. 1979. Annual Willow Tit (Parusmontanus): parental investment weight cvcle of wild screech-owls.Auk 96:795- or methodological artifact? Behav. Ecol. Socio- 796: . biol. 27:283-286. HOBSON,K. A., M. L. BOUCHART,AND S. G. SEALY. SAS INSTITUTE. 1989. SAS user’s guide: statistics. 1988. Responsesof naive Yellow Warblers to a 1989 ed. SAS Institute, Cary, NC. novel nest predator. Anim. Behav. 36: 1823-l 830. SHERMAN,A. R. 1911. Nest life of the screechowl. JOHNSGARD,P. A. 1988. North American owls: bi- Auk 28:155-168. ology and natural history. Smithsonian Institute SPROAT,T. M. 1992. The nest defense behavior of Press, Washington, DC. Eastern Screech-Owls. M.S.thesis, Eastern Ken- KNIGHT,R. L., AND S. A. TEMPLE. 1986a. Why does tucky Univ., Richmond, KY. intensity of nest defense increaseduring the nest- VANCAMP,L. F., ANDC. J. HENNY. 1975. The screech ing cycle?Auk 103:318-327. owl: its life history and population ecology in 296 THOMAS McKELL SPROAT AND GARY RITCHISON

northern Ohio. N. Amer. Fauna 7 1, U.S. Fish and fence in Mourning Doves: a test oftwo hypotheses. Wildl. Serv., Washington, DC. Anim. Behav. 38:1062-1066. Voous, K. H. 1988. Owls of the northern hemi- WESTNEAT, D. F. 1989. Kntensityof nest defense in sphere. Wm. Collins Sons and Co., London. Indigo Buntingsincreases with stageand not num- WALLIN.K. 1987. Defence as narental care in Tawnv ber of visits. Auk 106~747-749. Owls (Strix aluco). Behavibur 102:213-230. . WIKLUND, C. G. 1990. The adaptive significanceof WEATHERHEAD,P. J. 1979. Do Savannah Sparrows nest defence by Merlin, Falco columbarius,males. commit the Concord fallacy?Behav. Ecol. Socio- Anim. Behav. 40:244-253. biol. 5:373-38 1. WIKLUND, C. G., AND J. STIGH. 1983. Nest defence WEATHERHEAD,P. J. 1989. Nest defence by Song and evolution of reversedsexual size dimorphism Sparrows:methodological and life history consid- in Snowy Owls Nycteascandiaca. Omis Stand. 14: erations. Behav. Ecol. Sociobiol. 25: 129-l 36. 58-62. WESTMORELAND,D. 1989. Offspring age and nest de-