Nest Success and Nesting Habits of Eastern Kingbirds and Other Flycatchers

Home , Eastern kingbird, Kingbird, Western kingbird

Condor85:20X-2 19 0 The CooperOrnithological Society 1983

NEST SUCCESS AND NESTING HABITS OF EASTERN KINGBIRDS AND OTHER FLYCATCHERS

MICHAEL T. MURPHY

ABSTRACT.-Patterns of nest placement and its relationship to nest successin the Eastern Kingbird (T_JJT~Y~Y~UStyrannus) were studied in populations breeding in New York and Kansas. Data were augmented with information on nest placement in other open-nesting tyrannids in order to examine the hypothesis that these flycatchers place their nests chiefly so as to conceal them from predators. Nesting successwas significantly greater in New York than in Kansas but was relatively high in both populations, as is apparently true of North American breeding flycatchers in general. Geographic variation in nest placement in the Eastern Kingbird was relatively small and statistical comparisons of failed and successfulnests indicated that nests placed at mid-heights in the tree, and about midway between the center of the tree and canopy edge, were most successful. These nest sites were also the most commonly used sites. Increased vegetative cover around the nest and a greater number of supporting branches for the nest were also associatedwith success,and it is their interaction with nest height and distance from the canopy edge that is apparently most important in determining success.Most tyrant flycatchers nest several meters or more above the ground, and interspecific variation in nest placement matches the range of sites used by Eastern Kingbirds. Aggressive nest defense is apparently characteristic of the family, and average nest height is also positively correlated with length of the nestlingperiod. These facts support the predator-avoidance hypothesisand suggest that nest placement habits and aggressivenest defense are means by which open- nesting birds can reduce rates of nestling mortality.

Nest site selectionin birds can be an important exposed in the nest for up to 39 days. I there- determinant of reproductive successby affect- fore predicted that nest placement would be ing losses caused by predators and weather, important in determining the reproductive the two most important causesof nest failure successof Eastern Kingbirds. Furthermore, (Nolan 1963, Ricklefs 1969a). Except for the since slow nestling growth and long periods of well-known differencesin the breeding biology nest occupancy are typical of most tyrannids of cavity- and open-nesting birds (von Haart- (Ricklefs 1976; Murphy, in press), nest place- man 1957, Lack 1968) few studies of tem- ment should be especially important in this perate-zone breeding passerineshave related family and reflect the need to reduce nest loss- aspectsof nest placement to the species’ over- es. I tested these hypothesesusing data on nest all breeding habits. If specieswith long nesting placement and successin Eastern Kingbirds, periods placed their nestsin cryptic, easily de- and comparative data from the literature. fended or inaccessiblesites, they could coun- teract the increased probability of predators STUDY AREA AND METHODS finding the nest before the young fledge. This I studied the breeding biology of Eastern King- hypothesis has yet to be examined for tem- birds in western New York and eastern Kansas perate-zone breeding passerines, yet is pre- from May through August, 1979 and 1980, sumably the basis for the great diversity of respectively. The center of the New York re- nesting habits in tropical breeding land birds search site was located midway between the (Ricklefs 1969a). towns of Eden and Angola, Erie Co. (42”40’N, As part of a geographic comparison of the 78”57’W) and was approximately 16 km from breeding biology of the Eastern Kingbird (Ty- Lake Erie. In Kansas, I concentrated research rannus tyrunnus), I collected data on nest in an area 6.5 km west of Lawrence, Douglas placement and success.Eastern Kingbird nests Co., near Clinton Lake (38”57’N, 95’19’W). are relatively conspicuous and are typically The study site in New York encompassedthe placed on horizontal branchesclose to the can- adjacent lands 0.5 km on either side of 45 km opy edge (Davis 1941, Bent 1942, Pettingill of country roads, plus several larger isolated 1973). Nestling kingbirds develop slowly areas(range of 2-4 ha). My study site in Kansas (Murphy, in press) and eggsand nestlingsare included the adjacent lands 0.5 km on either

12081 KINGBIRD NESTING HABITS 209 side of country roads (30 km), plus an inten- site were made using the Student’s t-test. Rel- sively studied area measuring 1.6 X 0.8 km. ative nest height and relative distancefrom the My methods preclude accurate estimates of center of the tree to the canopy were arc-sine density, but breeding pairs seemed about 1.5 transformed before testing. The homogeneity to 2 times as common in Kansas as in New of variances was tested usingthe F,,,-test (So- York. kal and Rohlf 198 1). If variances were heter- At each nest I recorded the species of the oscedastic,an approximate t-test based on the nest tree, and six nest placement variables: nest assumption of unequal variances was used to height (m); distance from top of tree to nest test for significant differences between means (m); distance from nest to center of tree (m); (Sokal and Rohlf 198 1: 408). horizontal distance from nest to outer perim- If a nest’s successdepends upon its place- eter of canopy (m); number of branchesor twigs ment, it is likely that numerous variables, of supportingthe nest;and vegetative cover with- which I have measured or generated eight, in- in an imaginary sphere20 cm in radius around teract to ultimately determine the nest’s fate. the nest. Relative nest height (nest height/tree In order to improve ecologicalinterpretability, height) and relative distance from the center I reduced the number of variables to a few of the tree (distancefrom center of tree to nest/ mutually independent linear combinations of distance from center of tree to canopy edge) the original correlated variables using princi- were also calculated. Cover was estimated by pal component analysis (PCA; BMDP4M, placing a compass on the nest surface and Dixon and Brown 1979). Because of the di- ranking the quality of the cover for the four verse nature of the data (i.e., counts, lengths, cardinal compassdirections and directly over etc.), variables were standardizedby extracting the nest. Cover quality was scored as follows: the principal components from a correlation O-25% cover, scored0.0; 25-75% cover, scored matrix. However. because standardization 0.5; 75-100% cover, scored 1.0. Scores were based on the total variance tends to de-em- then summed to give an overall estimate of phasizecharacters with high variance, I further the degree of nest concealment (maximum of standardized variables on the basis of the 5.0). Inaccessible nests were examined using pooled? weighted within-group variances for an extensible mirror-and-pole apparatus. Dis- successfuland failed nests (Rohwer and Kil- tances and heights, later converted to meters, gore 1973). This procedureis appropriate when were measured to the nearest 0.5 foot using a PCA is being used to separate groups where pole marked off in feet, by climbing the tree between-group variances may be large com- and measuring distances directly, or using a pared to within-group variances, since the for- range-finder and clinometer. mer variables are those best able to differen- A successfulnest was one that fledgedat least tiate between groups (see Rohwer and Kilgore one nestling. At both sites I found more than 1973). one-half of the nestsbefore incubation started, An adaptively based predator avoidance and for these I divided the number of suc- theory for nest site selection predicts a direct cessful nests by the total number (S/n) to es- relationship between duration of the nestling timate success.I also calculated nest success period and nest height. To test this hypothesis using the entire sample and correcting for ex- I used literature sourcesto generate a corre- posuretime (Mayfield 196 1). The statusof most lation between nestling period length and nest nestswas checked every two to three days. The site characteristics for species with informa- standard error of the estimated successrate tion available. In this analysis,the duration of and 95% confidencelimits were also calculated the nestling period used was the most com- (Johnson 1979). Nest failures were categorized monly reported period. For nest height, I cal- as being due to either predation, weather, de- culated the weighted average of all reports of sertion, failure of eggsto hatch, or loss of an nest height for each speciesfor the correlation adult. If an entire nest and its contents dis- with nestling period length. appeared I assumed that it was destroyed by weather. If nest contents were missing but the RESULTS nest was relatively undisturbed I assumedpre- dation had occurred. Death of nestlingsdue to NEST SUCCESS overexposure to the sun also left the nest un- Estimates of nest successat the two sites in- disturbed. In these cases,where it was known dicated that breeding successof Eastern King- or strongly suspected,losses were attributed to birds was relatively high for an open-nesting weather. passerine(Table 1; seeRicklefs 1969a for com- Univariate comparisons of nest placement parative data). Accurate geographic compari- variables between New York and Kansas and sonswithout estimates of annual variation in between successfuland failed nestswithin each successare probably not possible,but nest suc- 210 MICHAEL T. MURPHY

TABLE 1. Estimates of nest success,daily rates of nest or death of nestlingsdue to direct exposure to mortality, and sourcesof nest failure in Eastern Kingbirds intense sunlight. The latter phenomenon was breeding in New York and Kansas. Sample sizes in pa- limited to nestlingsthat were unable to regu- rentheses. late their temperature effectively (Murphy, un-

New York KEUlsas publ. data)-nestlings that were lessthan seven days old (n = all nestlingsin three nests). Sel- Nests found prior to the dom was I able to attribute nest lossesto spe- start of incubationa 58.8% (34) 31.4% (35) Mayheld estimate” 51.1% (57) 36.3% (52) cific predators, but the majority of losseswere 95% confidence intervd’ 50.7-65.7% 30.2-43.5% most likely the result of avian or mammalian Sourcesof failure predators. Known instances included preda- Predation 72.2% (13) 57.9% (22) tion on a brood by an American Kestrel (Falco Weather 16.7% (3) 31.6%(12) sparverius) in New York and destruction of a Failure to hatch 5.6% (1) 2.6% (1) clutch of eggsby a fox squirrel (Sciurus niger) Loss of an adult 5.6% (I) 0.0% (0) in Kansas. I suspectthat squirrels are impor- Desertion 0.0% (0) 1.9% (3) tant nest predators for aboreal-nestingspecies Daily nest mortality in Kansas, since I also observed predation on rate” 1.59 3.08 a Northern Oriole (Zcterusgalbula) nest by a ilPercent success = (number of successful nests/total number of nests) x 100. fox squirrel. Other nestsprobably were lost to DSuccess calculated usmg Mayfield’s (1961) correctmn for exposure. American Crows (Corvus brachyrhynchos), c confidence ~ntenals calculated u~lng Johnson’s (1979) techmque. d Mortabty rate = -(log PV): Rlcklefs (1969~1). since I have observed crows searchingthrough vegetation in trees in the vicinity of kingbird nestsin apparent attempts to locate the nests. cess was 59% greater in New York than in Similarly, nest successamong other north Kansas (for nestsfound before the start of in- temperate breeding tyrannids also appears to cubation, t = 2.32, df = 67, P < 0.05). A pos- be relatively high (Table 2). The mortality rate sible causefor the lower successin Kansasmay that I recorded in Kansas in 1980 (3.08%/day) have been the more frequent nest visits that I was next to the highest of the 14 records for made there (nearly every day for many nests). open-nesting species.The average daily mor- Frequent nest visits have been known to cause tality rate for the nine open-nesting species increased lossesto predators (Bart 1978). listed in Table 2 (X = 1.77, SD = 0.793, n = Predation on Eastern Kingbird nestswas the 14) is intermediate with respect to the average major cause of failure in both New York and daily mortality rates for hole- and niche-nest- Kansas,followed by lossesdue to weather (Ta- ing birds (.x = 1.02, y1= 7; Ricklefs 1969a) as ble 1). Losses due to weather resulted from compared to above-ground-nestingbirds (K = nestsbeing blown out of trees by strongwinds, 2.27, y1= 8, Ricklefs 1969a [mortality rate for

TABLE 2. Nest successand daily mortality rates of tyrant flycatchersbreeding in North America.

Nesting Nestling Sample cycle length period length Percent Mortality Species sire (days) (days) success rate (%/day) Reference Scissor-tailedFlvcatcher 16 33 16 81.2 0.63 Fitch 1950 (T_Vx7nnusf&jilficalus) Western Kingbird 46 35 16.5 56.5 1.65 Goldberg 1979 (T. r~rticalls) 23 35 16.5 56.7 1.64 Cuesta I974 Cassin’s Kingbird 36 35 16.5 47.2 2.18 Goldberg 1919 (7.’ l~oc$vYuls) Eastern Kingbird 57 35.5 17 57.7a 1.59 This study, New York 52 33 16 37.3’ 3.08 This study, Kansas Eastern Wood-Pewee 6 34 17 74.p 0.89 Best and Stauffer 1980 ((ontopus‘ vircns) Acadian Flycatcher 138 30 13 64.5 1.46 Walkinshaw 1966a, (Empid0na.u viwscens) Mumford 1964 Traill’s Flycatcher 33 31.5 13.5 44.6 2.56 King 1955 (I?. trurllii) 92 31.5 13.5 65.2 1.36 Berger 1967 207 31.5 13.5 69.5 1.16 Walkinshaw 1966b 91 28 12 39.5 3.32 Holcomb 1972 Least Flvcatcher 101 31.5 14 51.9 2.08 Walkinshaw 1966~ (E. I&lII?1LIS) Western Flycatcher 28 34.5 16.5 61.5a 1.14 Davis et al. 1963 (E. d!ficilis)

* Nest success m these studies corrected for exposure tune

212 MICHAEL T. MURPHY

New York axis indicated that cover and number of supporting branches tended to decrease as absolute nest height and distance from the top of the tree and canopy edge increased (Fig. 1). Principal Component 3 explained 18.8 and 20.3% of the variation in nest placement in New York and Kansas, respectively, and in New York contrasted distance from the canopy edge with relative distance from the tree center (Fig. 1). Principal Component 3 in Kan- sascontrasted nestsplaced absolutely and relatively high in a tree, and absolutely and relatively far from the tree center, with nests placed low and close to the tree center. The contrast of the horizontal and vertical components of nest placement by PC1 was similar to the species-specificpatterns of nest placement found within other flycatchers (Ta- ble 4). With the exception of the Yellow-bellied Flycatcher, all open-nestingtyrannids place their nests in trees or bushes, 2 m or more above the ground. Those species nesting at lower heights usually place their nestsclose to the canopy edge (e.g., Acadian Flycatcher and Scissor-tailed Flycatcher; Table 4). The alternative, that of placing nestsnear the top of the tree but close to the main trunk, is used by the NEST PLACEMENT VARIABLES Least Flycatcher, Western Kingbird and Cas- FIGURE 1. Principal component loadings from the sin’s Kingbird (Table 4). The wide range of analysis of placement of nests in New York and Kansas. nest heights of speciesin the genus Tyrannus, PCl, PC2 and PC3 refer to Principal Components 1, 2 especially the Eastern Kingbird, indicates that and 3, respectively.Abbreviations for nestplacement variables are: nh, nest height: dt, distanceof nest from top of nest height depends to some extent upon the tree; rh, relative nest height; dc, distance of nest from availability of suitable nest trees (Table 4; D. center of tree; de, distance of nest from canopy edge; rd, I. MacKenzie, pers.comm.). However, the four relative distance of nest from center of tree, bu, number Tyrannus species(included at the top of Table of branches supportingnest; cv, vegetative cover around 4) clearly demonstrate the inverse relationship nest. between nest height and distancefrom the center of the tree. pattern of nest placement was usually also ac- NEST PLACEMENT AND SUCCESS companied by placement of the nest farther Nest successwas significantly correlated with from the tree center, in both a relative and nest placement variables in New York only absolute sense(Fig. 1). I found a similar pat- (Table 5). However, in all casescomparisons tern for Kansas. As relative nest height de- between failed and successfulnests in Kansas creased, distance from the top of the tree in- tended to be in the same direction as in New creasedand both absoluteand relative distance York. Successfulnests were placed at lower from the center of the tree increased (Fig. 1). absolute and relative heights than failed nests, Thus, for both populations, I interpreted PC1 and also had significantly more surrounding as a contrast of the horizontal and vertical cover and supporting branches in New York components of nest placment. Principal Com- (Table 5). Thus, lower successwas associated ponent 2 in New York explained 20.8% of the with greater exposure of the nest, due either original variation, and contrasted relative and to less vegetative cover around the nest, or absolutenest height with the microhabitat nest placement in increasinglyextreme positions in placement variables (i.e., number of support- the tree (i.e., closer to the peak). Relative dis- ing branchesand vegetative cover). It indicat- tance from the center of the tree and relative ed that nestsplaced lower in the tree generally nest height in Kansas were apparently not as- had more cover and more supportingbranches sociated with success(Table 5). than nests placed higher (Fig. 1). Principal I combined samplesfrom both sitesin order Component 2 explained 23.9% of the original to further examine the relationship between variation in Kansas nest placement data. This successand relative height and distance from

KINGBIRD NESTING HABITS 215

TABLE 5. Comparison of the placement of successfuland failed Eastern Kingbird nests in New York and Kansas. - New York (mean[SD]) Kansas (mean[SD]) Vanable Successful Faded I” Successful Failed F

Nest height (m) 5.9 (4.10) 7.8 (4.48) 1.46 6.1 (2.63) 7.9 (5.56) 1.55 Distance from top (m) 2.9 (2.40) 2.4 (1.63) 0.83 3.0 (1.37) 3.6 (3.96) 0.77 Relative height 0.68 (0.131) 0.77 (0.125) 2.46* 0.66 (0.152) 0.69 (0.197) 0.58 Distance from center (m) 1.7 (0.85) 1.9 (0.86) 0.57 2.3 (1.26) 3.0 (1.84) 1.71 Distance from canopy (m) 1.O (0.42) 0.9 (0.55) 0.63 1.7 (0.58) 2.0 (1.69) 0.91 Relative distance 0.61 (0.183) 0.68 (0.141) 1.67 0.55 (0.176) 0.60 (0.206) 0.97 Number of branches 6.0 (0.92) 5.0 (0.76) 3.90*** 5.2 (1.23) 4.6 (0.88) 1.70 Cover 3.4 (0.99) 2.5 (0.83) 3.40*** 3.0 (1.36) 2.7 (1.10) 0.80 Sample size 42 15 20 31

* Value of t-test comparmg mean scores between faded and successful nests. * P < 0.05: *** P < 0.001

five relatively low nests again nearly estab- two additional species(Western Wood-Pewee, lished a significantdifference between the two Co&opus sordidulus, and Olive-sided Fly- groups (t = 1.85, df = 44, P = 0.075) sug- catcher, C. borealis) that have long nestling gestingthat high nestsclose to the canopy edge periods (2 17 days). Both typically place their were more likely to fail. Although significant nestshigh in trees, usually on horizontal limbs differences were found in only one of the two (Bent 1942). Peewees (Contopus spp.) also sites, and for only one component, the results commonly build their nests on dead or dying of the principal components analyses were branches, and cover parts of the nest exterior consistent with univariate comparisons in as- with lichens, apparently in order to cryptically sociatinglower relative height, greateramounts pattern the nest (Bent 1942; M. T. Murphy, of vegetative cover, and increased distance pers. observ.). from the canopy edge with increased proba- bilities of success. DISCUSSION The duration ofthe nestlingperiods and nest Inaccessible, secure nest sites should be at a heightsused to test the predator avoidance the- premium when time constraints limit oppor- ory for nest site selection were taken from Ta- tunities for breeding and environmental fac- bles 2 and 4, respectively. A significant posi- tors limit rates of growth. Tyrant flycatchers tive relationship exists between length of the breed relatively later (Robins 1970) and have nestlingperiod and nest height (r = 0.606, df = longer nesting cycles than most other north 11, P < 0.05). However, nestling period du- temperate breeding, open-nesting passerines ration and adult body weight are also corre- (Table 2). Their breeding seasonsare therefore lated, though not significantly in these species relatively short and only three species of (Y = 0.487, df = 11, ns). The correlation of phoebes (Sayornis spp.) are able to raise two nestling period duration with nest height may broods per year (Bent 1942, Robins 1970). In therefore be an artifact of a relationship be- the Eastern Kingbird, neststhat fail at the nest- tween body weight and nest height. However, ling stageare most often not replaced (Murphy, the latter correlation is not significant (Y = in press). Specialized nesting habits are there- 0.467, df = 11). These results support the fore to be expected in this family and infor- predator avoidance hypothesis, and strongly mation on nest placement presentedhere sup- suggestthat flycatcherscompensate for greater ports this argument, the strongest support risks of predation as the length of exposure coming from the significant relationship be- increases by making nests less accessible to tween nest height and duration of the nestling ground predators, and most mammalian pred- period. The relatively high nest successap- ators. I could not find quantitative data for parently characteristic of the family in North

TABLE 6. Mean factor scoreson Principal Components 1, 2 and 3 for successfuland failed nests in New York and Kansas.

New York (mean [SD]) Kansas (mean [SD]) PC1 PC2 PC3 n PCI PC2 PC3 n

Successful 0.02 (1.042) 0.26 (0.958) -0.14 (0.918) 42 -0.08 (0.730) -0.28 (0.839) -0.23 (0.835) 20 Failed -0.06 (0.901) -0.74 (0.714) 0.40 (1.139) 15 0.05 (1.150) 0.18 (1.164) 0.15 (1.081) 31 f-value 0.29 3.71*** 1.85 0.48 1.66 1.32

***P < 0.001 216 MICHAEL T. MURPHY

New York New York

3 I I I 1 3

0 . 2 . 2 hl . 0 0 (3 . . . 1 I- . 00 1 0

z 1 . z‘ ’ 0 l l ./ 1;‘ = . z . . . 0 : 0 0 . e-0 do- 8. Oi: T ::;_.L 0 0.0 0 0 1 . 00 0 8. . 0 0 2 8g -1 0 00 0 Oio. . .** 0 -1 00 l 0 0 . . . 0. . 0 . . . -2 P -2 0 . I

-2 -1 0 1 2 3 -2 -1 0 1 2 3

COMPONENT 1 COMPONENT 2

Kansas Kansas

3 01 I I I 3 I I I I

2- . 0 0 0 0 0 0 0 m 3 0 0 . + 0 c . 0 0 z -‘ . 0 0 0 . : 0 . 8* l 0” 0 0 a.* O 0. . .O . - _.A l -• 0 0. . 0 0. l 0 ca ar . 000 cJez -l o” 0 OQo a. 00 -1 - .g‘ i=0 l 0o 0 b l O -2 - 0

I I I I I I I I -2 -1 0 1 2 3 -2 -1 0 1 2 3

COMPONENT 1 COMPONENT 2 FIGURE 4. Bivariate plots of component scoresfor successful(closed circles) and unsuccessful(open circles) nests on Principal Components 1 and 2, and 2 and 3 in New York and Kansas. Mean values for both classesof nests on all three components are given in Table 6.

America suggeststhat the nest placement be- the larger lower limbs offered more support for haviors adopted by these speciesare effective their movements. In addition, nestsplaced ex- means of reducing nest losses. tremely high in trees usually also failed. Sim- Patterns observed in my study suggestthat ilarly, except for the low success(20%, n = 5) predation is the driving force behind nest site of nestsplaced near the trunk in Kansas, nest selection. First, it is the major cause of nest successwas inversely related to relative dis- failures in the EasternKingbird. Kingbird nests tance from the center of the tree. Exposure to placed relatively low had the least probability physical factors (high winds and intense sun- of fledging young, due almost solely to the ac- light) was probably more severeat higher levels tivities of predators (80% of losses below a in trees or at the ends of brancheswhere vege- relative height of 0.41). Low nests were pre- tation and supportingbranches are thinner and sumably more accessible,possibly because they weaker. If predators were absent, nestsplaced were more visible to ground predators, and/or relatively low or close to the trunk would pre- KINGBIRD NESTING HABITS 217

sumably be most successful.The complete ab- (Bent 1942) and both Davis (1941) and sence of nests below a relative height of 0.41 MacKenzie and Sealy (198 1) stated that the in New York and the low percentagein Kansas most violent Eastern Kingbird interactions oc- (10%) suggeststhat these nest sites are suscep- cur intraspecificallyat the nest site. MacKenzie tible to predation. Low sites are apparently and Sealy (198 1) also suggestedthat Eastern shunnedby Eastern Kingbirds. Maximum suc- Kingbird aggressionoccurs in relation to com- cessthus occurred at mid-heights in the tree petition for nest sites.However, kingbirds (Ty- and about midway between the tree center and rarzrzusspp.) also aggressivelydefend their con- canopy edge, indicating that those sites most spicuous nests against predators. Ricklefs heavily used were also generally the most suc- (1977a) detected a strong correlation between cessful. Nest placement behavior therefore nest conspicuousnessand intensity of nest de- probably represents a compromise between fense in tropical passerines.Conspicuous nests balancing lossesto predators and weather. It are generally placed peripherally, and as Rick- seems to act in a manner analogous to stabi- lefs (1977a) suggested,the extra defense re- lizing selectionsince at one extreme (low nests, quired by conspicuousnests is probably aided closeto trunk) predatorsare able to easily reach by this peripheral placement since it affords nests, but at the other (high nests or near the air spacein which aerial foragerswith powerful canopy edge) damage or loss of neststo phys- flight abilities can maneuver. Eastern King- ical factors increases. birds watch vigilantly over their nests (Smith However, results of the principal compo- 1966; M. T. Murphy, pers. observ.), regularly nents analyses of Eastern Kingbird nests in- intercepting potential nest predators over 100 dicated that successwas equally high among m from the nest and forcing them from the nestsplaced low in trees when placed on long area. I have seen flying Blue Jays (Cyanocitta horizontal branches, and those placed at the cristata) driven to the ground and, on one oc- tops of trees, but relatively closer to the main casion, knocked off a perch by an Eastern trunk. Both positions are apparently difficult Kingbird. Crows and squirrels are similarly for nonflying predators to reach, probably be- harassed.Western Kingbirds are nearly as ag- causeof the weaknessof the branchesat those gressive and both have repeatedly struck me sites (for an example see Walkinshaw 1966a: while I weighed nestlings at nests. While nu- 255). Vegetative cover around the nest, and merous species are occasionally attacked, the number of supporting branchesare appar- corvids and other avian predators are the ently important determinants of successand speciesmost frequently the objects of the most it is their interaction with relative and absolute persistent and aggressiveinterspecific attacks nest heightsand distancefrom the canopy edge by Eastern Kingbirds. that are most important in determining suc- Reproductive effort, of which nest defense cess. These results suggestthat the different is a componen+.t,is predicted to vary inversely nest sites used by these flycatchers provide with the potential for future breeding attempts equally effective means of hiding from pred- (e.g., as individuals age or as the end of the ators. The actual site where each speciesplaces breeding seasonapproaches; Pianka and Par- its nest probably reflects the range of suitable ker 1975, Ricklefs 1977b). Short-lived indi- sitesavailable to it within the tree speciesnor- viduals (e.g., passerinebirds; Farner 1955) with mally found in the habitats and geographic restricted breeding seasons(tyrant flycatchers; ranges where each species evolved. For ex- Robins 1970; Murphy, in press)should invest ample, within the genus Tyrannus, the Scissor- relatively greater amounts of time and energy tailed Flycatcher has the lowest average nest in each breeding attempt than should individ- height and the Western and Cassin’s kingbirds ualswith more than one opportunity for breed- the highest.The former speciesoccupies open, ing per season. The intense aggressionover grasslandhabitats (Smith 1966; M. T. Mur- nest sites and the presumed high risks asso- phy, pers. observ.) where average tree height ciated with nest defense in tyrannids would is low. The latter two speciesoften nest along therefore be predicted from the limited breed- river coursesin the west where tree height av- ing opportunities that most specieshave. East- eragesconsiderably higher (Smith 1966). The ern Kingbirds differ in their intensity of nest Eastern Kingbird is perhaps the most catholic defenseand more aggressivepairs have higher in its choice of nest sites, possibly becauseof nest success(Blancher and Robertson 1982). the great diversity of habitats encounteredover The ability of kingbirds and other flycatchers its broad geographic range (see Fig. 1 in to engagein aggressivenest defenseis probably MacKenzie and Sealy 198 1). aided by the peripheral placement of the nest Adult defense of the nest may also contrib- and their powerful flight abilities. ute to high nest successin these species.Tyr- The hypothesisthat tyrannid aggressiveness annids are well known for their aggressiveness is related to nest defense is supported by one 218 MICHAEL T. MURPHY case of its complete absence. Adult Yellow- LITERATURE CITED bellied Flycatchers, which place their nestson BART, J. 1978. Impact of human visitation on avian the ground (Bent 1942), are quiet and unob- nesting success.Living Bird 16:187-l 92. trusive around the nest, and nests are appar- BENT, A. C. 1942. Life histories of North American fly- ently difficult to find (Walkinshaw 1967). Since catchers, larks, swallows and their allies. U.S. Natl. these birds would have little chance of deter- Mus. Bull. 179. BERGER,A. J. 1967. Traill’s Flycatcher in Washtenaw ring most terrestrial predators, it seemslikely County, Michigan. Jack-Pine Warbler 45:117-123. that they have evolved an alternative habit of BEST,L. B., AND D. F. STAUFFER.1980. Factors affecting anonymity in the nesting area so as to reduce nestingsuccess in riparian bird communities. Condor the chance of attracting predators to the nest. 82:149-158. Behavior at the nest and nest placement are BLANCHER,P. J., AND R. J. ROBERTSON.1982. Kingbird aggression:does it deter predation?Anim. Behav. 30: therefore likely coevolved traits, and are the 929-930. primary means of reducing nest losses. CUESTA,L. R. 1974. Comparative breeding ecology of These findings indicate that even within the the Western Kingbird (Tyrannus verticalis) in three same habitats, not all speciesare necessarily habitat tyoes. M. SC.thesis, New Mexico State Univ.. Las Cru&. exposed to the same risks of predation. Selec- DAVIS, D. E. 1941. The belligerency of the kingbird. tion of particularly safe nest sites and/or ag- Wilson Bull. 53: 157-168. gressive nest defense are methods by which DAVIS,D. E. 1954. The breedingbiology of Hammond’s rates of nest loss can be reduced, suggestinga Flycatcher. Auk 7 1:164-l 7 1. basis for the intense intraspecific competition DAVIS, J., G. F. FISLER,AND B. S. DAVIS. 1963. The breeding biology of the Western Flycatcher. Condor for nest sites by Eastern Kingbirds. However, 651337-382. the efficacy of nest site selection to reduce nest DIXON, W. J., AND M. B. BROWN [EDS.]. 1979. BMPD- lossesis not always realized. Hence, it is not 79: biomedical computer programs, P-series. Univ. justified to assume, as has sometimes been of California Press, Berkeley. FARNER.D. S. 1955. Birdbandinn in the studv of DOD- done, that all things are equal among open- ulaiion dynamics, p. 397-4491 In A. Wolfson fed:], nesting passerines.Different risks of nest loss Recent studiesin avian biology.Univ. of Illinois Press, cast doubt on models of life history evolution Urbana. based on this premise, an example being the FITCH, F. W. 1950. Life history and ecologyof the Scis- hypothesis that growth rates of nestling pas- sor-tailedFlycatcher, Muscivoraforficata. Auk 67: 145- 168. serines are pushed to the physiological maxi- FITZPATRICK,J. W. 1981. Searchstrategies of tyrant fly- mum becauseof the intensity of predator pres- catchers.Anim. Behav. 29:810-821. sure (Ricklefs 1969b). Unquestionably, high GOLDBERG, N. H. 1979. Behavioral flexibility and for- rates of nest predation will select for faster aging strategies in Cassin’s and Western kingbirds nestling growth and early fledging, but other (Tyrannus vociferans and T. verticalis) breeding sym- patricallyin riparian habitatsin centralArizona. Ph.D. constraintson the rate of nestlingdevelopment diss., Univ. of Illinois, Champaign. (e.g., foraging mode; O’Connor 1978) may not HOLCOMB,L. C. 1972. Traill’s Flycatcher breeding bi- allow such increasesin growth rate. The spe- ology. Nebr. Bird Rev. 40:50-68. cialized foraging techniquesof most tyrant fly- JOHNSON, N. K. 1963. Biosystematicsof sibling species catchers generally result in low rates of en- of flycatchersin the Empidonax hammondi-oberhol- seri-wrightii complex. Univ. Calif. Publ. Zool. 66:79- counters with prey (Fitzpatrick 1981), which 238. may be further worsenedby inclement weather JOHNSON,D. H. 1979. Estimating nest success:the May- (Mahan 1964). The prediction that nestling field method and an altematiie. Auk 96:65 l-66 1._ growth rates should drop under such condi- KING. J. R. 1955. Notes on the life historv of Traill’s F&catcher (Empidonax traillii) in southeastern tions (O’Connor 1978) is in general supported Wishington. Auk 72: 148-l 73. by the available information (Ricklefs 1976; LACK. D. 1968. Ecological adaotations for breeding in Murphy 198 1, in press).Nest site selectionap- Girds. Methuen Press, Londdn. parently provides a mechanism to allow such MACKENZIE,D. I., AND S. G. SEALY. 1981. Nest site reductions in the rate of nestling development. selection in Eastern and Western kingbirds: a multi- variate approach. Condor 83:310-321. ACKNOWLEDGMENTS MAHAN, H. D. 1964. The effectsof the environment on growth, development and temperature regulation in I thank M. Gaines, R. F. Johnston, P. E. Lowther and 0. nestling EasternPhoebes. Ph.D. diss., Michigan State R. Taylor for reading and commenting on earlier drafts Univ., East Lansing. of this paper. Constrictive criticisms were also given by MAYFIELD,H. F. 1961. Nesting successcalculated from R. C. Fleischer. M. Holmaren. and W. B. MacGillivrav. exposure. Wilson Bull. 73:255-26 1. The late David MacKenzieprovided a prepublicationcopy MUMFORD,R. E. 1964. The breedingbiology of the Aca- of his paper and additional information on nest placement dian Flycatcher. Misc. Publ. Mus. Zool. Univ. Mich., in Eastern Kingbirds from his study area. His comments 125. on an earlier draft are also appreciated.The Northeastern MURPHY, M. T. 1981. Growth and aging of nestling Bird-Banding Association &d the Bird Division at the Eastern Kingbirds and Eastern Phoebes.J. Field Or- Museum of Natural History at the University of Kansas nithol. 52:309-3 16. provided funds. The help of my wife, Karmel Licata, MURPHY,M. T. In press. The breedingbiology ofEastern throughout this researchwas invaluable. Kingbirds: geographiccomparisons. Ecology. KINGBIRD NESTING HABITS 219

NOLAN, V., JR. 1963. Reproductive successin birds in ROHWER, S. A., AND D. L. KILGORE, JR. 1973. Inter- a deciduousforest scrubhabitat. Ecology44:305-3 13. breeding in the arid-land foxes, Vulpesvelox and V. O’CONNOR, R. J. 1978. Growth strategiesof nestling macrotis.Syst. Zool. 22: 157-l 65. passerines.Living Bird 16:209-238. SMITH, W. J. 1966. Communications and relationships OHLENDORF,H. M. 1974. Competitive relationships in the genus Tyrannus.Publ. Nuttall Ornithol. Club. among kingbirds (Tyrunnus) in Trans-Pecos Texas. 6. Wilson Bull. 86:357-373. SOKAL, R. R., AND F. J. ROHLF. 198 1. Biometry. W. H. PETTINGILL,0. S., JR. 1973. Exceptional nest sitesof the Freeman, San Francisco. Eastern Kingbird. Jack-Pine Warbler 5 1:124-l 26. VON HAARTMAN, L. 1957. Adaptations in hole-nesting PIANKA,E. R., AND W. S. PARKER. 1975. Age-specific birds. Evolution 11:339-347. reproductive tactics. Am. Nat. 109:453-464. WALKINSHAW, L. C. 1966a. Studies of the Acadian Fly- RICKLEFS,R. E. 1969a. An analysisof nesting mortality catcher in Michigan. Bird-Banding 37:227-257. in birds. Smithson. Contrib. Zool. 9. WALKINSHAW, L. C. 1966b. Summerbiology of the Traill’s RICKLEFS,R. E. 196913. Preliminary models for growth Flycatcher. Wilson Bull. 78:3146. rates of altricial birds. Ecology 50:1031-10401 WALKI~SHAW, L. C. 1966~. Summer observationsof the RICKLEFS.R. E. 1976. Growth ratesofbirds in the humid Least Flycatcher in Michigan. Jack-Pine Warbler 44: New World tropics. Ibis 118:179-207. 150-168. RICKLEFS,R. E. 1977a. Reactions of some Panamanian WALKINSHAW, L. C. 1967. The Yellow-bellied Flycatcher birds to human intrusion at the nest. Condor 79:376- in Michigan. Jack-Pine Warbler 45:2-9. 379. RICKLEFS,R. E. 1977b. On the evolution of reproductive strategiesin birds: reproductive effort. Am.Nat. 111: Museum of Natural History, Department of Systematics 453-478. and Ecology, University of Kansas, Lawrence, Kansas ROBINS,J. D. 1970. The relationship of food supply to 66045. Received 1 February 1982. Final acceptance20 the timing of breeding in aerial foragers.Kansas Or- December 1982. nithol. Sot. Bull. 21:9-15.

Cbndor85:219 0 The CooperOrnithological Society 1983

RECENT PUBLICATIONS

Habitat Use and Productivity of Canvasbacks in South- cover. No price given. This study of native grasslandbird western Manitoba, 1961-72.-Jerome H. Stoudt. 1982. populations and their habitat follows standard formats, U.S. Fish and Wildlife Service. Snecial Scientific Renort- resultingin useful data that are analyzed with appropriate Wildlife, No. 248. 3 1 p. Paper’cdver. Representingtwelve statistical methods. The taxonomy and nomenclature of yearsof observationwhich ended more than a decadeago, soils is difficult at best, and the number of independent this paper is largely a historical study. Thirty-three tables variables that make up a given soil type is large. The of observationaldata compose the bulk of the report. An authorsprovide an excellentsummary of the first problem, editorial footnote on current conditions in the Minnedosa and ignore the second. The resulting discussion of the area would have been valuable. Appendix A is a list of interrelationshipsamong native grassland,birds, soils,and common and scientific names of plants and animals. Ap- grazing is interesting, and supported whenever possible. pendix B is a phenologyof the breedingseason for selected The evident optimum breedinghabitat presentedfor each organisms in the study area. Photographs, and a list of of 29 bird speciesis garnered from literature and obser- references(but not all citations used).-J. Tate. vational sources,and from generalized results supported by the data. The conclusionsare, for the most part, mean- Effects of Soils and Grazing on Breeding Birds of Uncul- ingful statementsof the relationship between each grass- tivated Upland Grasslands of the Northern Great Plains.- land bird and its habitat. A small but valuable paper for H. A. Kantrud and R. L. Kologiski. 1982. U.S. Fish and anyoneinterested in the birds of our Northern Great Plains. Wildlife Service.Wildlife ResearchReport 15. 33 p. Paper References.- J. Tate.