I

(ISSN 1)892-1 01 6)

Volume 32 September 1998 Number 3

Contents

Breeding Distribution and Nest-Site Habitat of Northern Goshawks in

WISCONSIN. Robert N. Rosenfield, John Bielefeldt, Dale R. Trexel and Thomas C.J. Doolittle.. 189

Solitary and Social Hunting in Pale Chanting Goshawk ( Meuerax canorus)
Families: Why Use Both Strategies? Gerard Maian 195

Forest Management Effects on Nesting Habitat Selected by Eurasian
Black Vultures (Aegypius monachus) in Central Spain. Juan a. Fargaiio,

Guillermo Blanco and Eduardo Soto-Largo 202

Selection of Settlement Areas by Juvenile Bonelli’s Eagle in Catalonia.

Santi Manosa, Joan Real andjordi Codina 208

Winter Foraging Ecology of Bald Eagles on a Regulated River in

SOUTHWEST Idaho. Gregory S. Kaltenecker, Karen Steenhof, Marc J. Bechard and James C.
Munger 215

Urban, Suburban and Rural Red-tailed Hawk Nesting Habitat and

Populations in Southeast Wisconsin. William e. stout, Raymond k. Anderson and
Joseph M. Papp 221

Highway Mortality of Barn Owls in Northeastern France. Sylvie Massemin

and Thierry Zorn 229

The Effect of Burrow Site Use on the Reproductive Success of a
Partially Migratory Population of Western Burrowing Owls (Speotyto

CUNICULARIA HYPUGAEA ) . Eugene S. Botelho and Patricia C. Arrowood 233

Breeding-Season Food Habits of Burrowing Owls ( Athene cunicularia) in

Southwestern Dominican Republic. James w. Wiley 241

Short Communications

Rates of Open-Field Foraging by the Mississippi Kite (Ictinia mississippiensis) .

E. William Wischusen 246

Evaluation of Neck-Mounted Radio Transmitters for Use with Juvenile

OSPREYS. Lauren N. Gilson 247

Organochlorines and Mercury in Peregrine Falcon Eggs from Western

NORTH Carolina. Tom Augspurger and Allen Boynton 251

Importance of Birds and Potential Bias in Food Habit Studies of
Montagu’s Harriers (Circus pygargus) in Southeastern Spain. Jose a.

Sanchez-Zapata and Jose F. Calvo 254

Prey Brought to Red-shouldered Hawk Nests in the Georgia Piedmont.

Doug L. Howell and Brian R. Chapman 257

Peregrine Falcons (Falco peregrinus) Nest in a Quarry and on Highway

CUTBANKS IN Alaska. Robert j. Ritchie, Terry J. Doyle and John M. Wright 261

Lice (Phthiraptera: Amblycera, Ischnocera) of Raptors in Hungarian Zoos
and Rehabilitation Centers. Szaboics Sok 264

Letters 267

THE JOURNAL OF RAPTOR RESEARCH

A QUARTERLY PUBLICATION OF THE RAPTOR RESEARCH FOUNDATION, INC.

Vol. 32 Sept 1998 No. 3

J. Raptor Res. 32(3):189-194
© 1998 The Raptor Research Foundation, Inc.

BREEDING DISTRIBUTION AND NEST-SITE HABITAT OF
NORTHERN GOSHAWKS IN WISCONSIN

Robert N. Rosenfield

Department of Biology, University of Wisconsin, Stevens Point, WI 54481 U.S.A.

John Bielefeldt

Park Planning, Racine County Public Works Division, Sturtevant, WI 53177 U.S.A.

Dale R. Trexel

Department of Ecology, Evolution, and Behavior, 100 Ecology Building, 1987 Upper Buford Circle,

University of Minnesota, St. Paul, MN 55108 U.S.A.

Thomas C. J. Doolittle

Bad River Band of Lake Superior Chippewa, PO. Box 39, Odanah, WI 54861 U.S.A.

Abstract. — We found Northern Goshawks ( Accipiter gentilis ) nesting widely throughout the northern
two-thirds of Wisconsin during 1996-97, with no evidence of range contraction as might be expected
as one index of changing status if the state’s breeding population were declining. During 1977-97,
habitat was sampled on 0.04 ha circular plots at 37 goshawk nests, of which 78% were in deciduous
trees, especially trembling aspen ( Populus tremuloides) . Mean nest-tree height, mean nest-tree diameter-
at-breast-height (dbh), and mean tree density were 25 m, 41 cm, and 423 stems/ha, respectively. A
comparison of these and 20 other habitat features at nest sites found by unbiased vs. potentially-biased
methods failed to detect statistically significant differences between these two data sets. Goshawks nested
in a broad array of forest types, including pine plantations and forest fragments in agriculturally-dom-
inated landscapes.

Key WORDS: Northern Goshawk ; Accipiter gentilis; nesting distribution; breeding range, nest-site habitat.

Distribucion del habitat de anidacion de Accipiter gentilis en Wisconsin

RESUMEN. — Encontramos a Accipiter gentilis anidando a traves de los dos tercios del norte de Wisconsin
durante 1996-97, sin evidencias de que su rango de distribucion disminuya como se podria esperar de
su declinacion en su poblacion reproductiva. Durante 1977-97, se hicieron muestras de habitat en
parcelas circulares de 0.04 ha en 37 nidos de azor de los cuales 78% se encontraban en arboles cadu-
cifolios ( Populus tremuloides ) . La media de altura, la media de diametro a la altura del pecho y la media
de la densidad de los arboles fue de 25m, 41 cm y 423 troncos/ha respectivamente. Una comparacion
de estas y otras 20 caracteristicas de habitat en los sitios de los nidos encontrados entre metodos sin
sesgos y potencialmente sesgados no reporto diferencias estadisticamente significativas entre estos dos
grupos de datos. Los azores anidaron en una amplia variedad de tipos de bosques incluyendo planta-
ciones de pinos y fragmentos de bosques en paisajes agricolas.

[Traduccion de Cesar Marquez]

189

190

Rosenfield et al.

Vol. 32, No. 3

There is concern that populations of the North-
ern Goshawk ( Accipiter gentilis, hereafter goshawk)
may be declining in North America because of hu-
man-induced habitat alterations (Braun etal. 1996,
Kennedy 1997, Squires and Reynolds 1997). There
is a consequent need for baseline information on
key population attributes that may, or may not,
yield evidence of population declines (Block et al.
1994). Kennedy (1997) investigated the possibility
of goshawk population declines in North America,
in part through a literature review of its breeding
distribution for evidence of range contractions. Be-
cause the bulk of studies she examined were geo-
graphically limited, she was unable to provide in-
formation from mid-continental regions.
Moreover, the majority of studies on nest-site hab-
itat in North America have focused on goshawks in
the western U.S., where most investigations have
used biased searching techniques to locate nests
(Squires and Reynolds 1997). Apfelbaum and Seel-
bach (1983) have reported nest tree species at 22
goshawk nests in the midwestern U.S., but their
tally included Pennsylvania as a supposedly mid-
western state and did not associate nest-tree data
with specific locales at state or sub-state levels. As
with distributional data mentioned above, pub-
lished information on nest-site habitat for the gos-
hawk in mid-continental North America thus re-
mains very limited.

Here, we show that the goshawk currently
(1996-97) has a wide breeding distribution in Wis-
consin with no sign of range contraction and de-
scribe nest-site habitat for 37 nests sampled from
1977-97, including a comparison of habitat
features at goshawk nests found by unbiased vs. po-
tentially biased methods.

Study Area and Methods

We found goshawk nests in Wisconsin from 1996-97
by using three methods. First, we searched four quadrats,
each about 3885 ha in size, that were objectively estab-
lished (i.e., without past or present knowledge of forest
serai stages or use of these sites by goshawks) within pre-
dominately wooded habitats in the northern third of Wis-
consin. Second, we searched historic goshawk nesting ar-
eas and, third, we obtained nest-site information from
nonproject personnel including staff of the Wisconsin
Department of Natural Resources (WDNR) and other
agencies, falconers, and others.

Nest-site habitat data were collected in 0.04 ha circular
plots (11.3 m radius) centered on the nest tree, using
procedures described by James and Shugart (1970) as
modified by Titus and Mosher (1981). We sampled hab-
itat at four occupied nests (i.e., eggs laid), one in each
of the four quadrats during 1996-97, and at 33 occupied

Figure 1. The distribution of Northern Goshawk nests
(1977-97) in Wisconsin at which nest-site habitat data
were collected.

nests elsewhere in Wisconsin during 1977-97 (Fig. 1). Of
these 37 geographically separate goshawk nest sites, 23
(62%) were sampled in the year of breeding, while four
and seven other nest sites were sampled one or two years,
respectively, following nesting. Habitat at three nest sites
unaltered by logging or other human activities was sam-
pled five to seven years after discovery of nests. Table 1
describes vegetation and physical measurements ob-
tained at each nest site.

In addition, we arbitrarily divided the nest-site habitat
sample into those nests that we regarded as found by
unbiased means (N = 21) and those found by potentially
biased means ( N — 16). We compared values of habitat
features in these two categories on the premise that nests
found by potentially biased searching techniques may not
be representative of nest-site habitats used by goshawks
(Siders and Kennedy 1996, Squires and Reynolds 1997).
Nests found by unbiased means were characterized as
those found on the quadrats ( N = 4) or detected during
activities other than searching for goshawks, such as song-
bird inventories, botanical surveys, recreational hiking,
and other incidental discoveries, excluding cruising for-
ests for timber ( N = 17). Nests were classified as being
found by potentially biased methods when searches for
goshawk nests were conducted in habitat presumed suit-
able for nesting, such as mature, late serai northern hard-
wood-conifer forests in Wisconsin ( N = 9) or when nests
were found during timber cruising (N = 7), an activity
that may not equally represent all potential nesting hab-
itats (Hayward and Escano 1989). The majority of habitat
variables did not exhibit normal distributions in Lilliefors
Tests so nonparametric Mann-Whitney U tests were used
to assess the potential significance of differences between

Sept 1998

Goshawk Distribution and Habitat

191

Table 1 . Vegetational and physical features measured at
Northern Goshawk nest sites in Wisconsin.

Feature

Description

Dist. to Water

Distance to nearest permanent wa-
ter source (pacing or USGS 7.5'
quadrangle)

Dist. to Opening

Distance to nearest forest opening
>5 ha in size (pacing or USGS
7.5' quadrangle)

Nest Tree Height

Height of nest tree (Haga altime-
ter)

Nest Tree dbh

Diameter at breast height of nest
tree

Nest Height

Height of nest (meter tape or
Haga altimeter)

Nest Percent

(Nest height/Tree height) (100)

Degree Slope

Maximum slope by altimeter or cli-
nometer

Tall Shrub Index

Index of tall shrubs <3 cm dbh
and >shoulder height 3

Low Shrub Index

Index of low shrubs <3 cm dbh
between knee and shoulder
height 3

Under. Density

Number of understory trees >9
cm dbh per ha

Canopy Height

Mean height of five canopy trees
in study plot (Haga altimeter)

Total Canopy

Percent of area over study plot oc-
cluded by overstory foliage 5

Decid. Canopy

Percent of area over plot (not of
total canopy) occluded by decid-
uous overstory foliage 5

Conif. Canopy

Percent of area over plot occluded
by evergreen overstory foliage 5

Total Understory

Percent of area over plot occluded
by understory foliage 5

Decid. Understory

Percent of area over plot occluded
by deciduous understory foliage 5

Conif. Understory

Percent of area over plot occluded
by coniferous understory foli-
age 5

Total Ground

Percent of ground in plot covered
by ground-layer foliage 5

Decid. Ground

Percent of ground in plot covered
by ground-layer deciduous foli-
age 5

Conif. Ground

Percent of ground in plot covered
by ground-layer coniferous foli-
age 5

Tree Density

Number of canopy trees >9 cm
dbh per ha

Basal Area

Basal area in m 2 /ha of canopy
trees

Mean dbh

Mean dbh of canopy trees in study
plot

11 Sum of stems on 4 plot radii.
b 40 ocular tube readings.

Figure 2. Known past and present distributions and
1996—97 nest-site locations of the Northern Goshawk in
Wisconsin. 1 = resident range (Gromme 1963); 2 = res-
ident range (Robbins 1991). Unshaded circles = coun-
ties with nests before 1958; shaded circles = 1996 nest
locations; squares = 1997 nest locations.

these two nest-site categories. Because of the number of
multiple univariate comparisons, we calculated that an
alpha of 0.002 was the appropriate level of significance
for statistical inference (Sokal and Rohlf 1981). Statistical
analyses were performed on SYSTAT (Wilkinson 1992)

Results and Discussion

In 1996-97, goshawk nests (N = 34) were widely
distributed in the northern two-thirds of Wisconsin
(Fig. 2). The historic breeding distribution of the
goshawk in Wisconsin in the mid-1800s, before tim-
ber harvests became pervasive in the northern part
of the state, is unknown. The first known nests in
the state were found in the 1890s (Robbins 1991).
Through 1958, only about 12 nests had been doc-
umented, all in five northern counties. As late as
1964 (Scott 1964), the northern half of the state
held only 15% of the in-state members of the Wis-
consin Society for Ornithology. During the same
time period, no goshawk nests were reported in
the southern half of the state in counties that held
the large majority of society members engaged in
ornithological activity (Scott 1964). Gromme
(1963) probably based his map of the resident
range of the goshawk (Fig. 2) on the above nests
plus summer sight records. Robbins (1991) pic-

192

Rosenfiet.d et al.

Vol. 32, No. 3

tured a somewhat larger resident range for the gos-
hawk. We interpret their maps of resident range as
a presumed but perhaps imperfectly documented
breeding range. Squires and Reynolds (1997) ex-
plicitly interpreted Robbins’ map as breeding
range in the state. Several additional sources por-
tray a similar resident or breeding distribution in
northern Wisconsin (Palmer 1988, Johnsgard
1990, Rosenfield et al. 1991).

The current nesting range of the goshawk in
Wisconsin is much larger than previously indicated
by maps of resident and breeding ranges. There-
fore, there is no current evidence for a contraction
of the goshawk’s nesting range in the state, as
might be expected as one index of changing status
if the state’s breeding population were declining
(Kennedy 1997). S. Postupalsky (pers. comm.) has
suggested that the 1996-97 nesting range may ac-
tually indicate that there has been an expansion in
the breeding distribution, as has occurred in Mich-
igan (Postupalsky 1991) and possibly other eastern
states (Kennedy 1997).

The southern distributional limit mapped by
Robbins (1991) corresponds approximately with
the present limit of extensively forested lands in
northern Wisconsin (Wisconsin Department of
Natural Resources 1995). Many of the nests found
in 1996-97 south of Robbins’ line were located in
woodlands in the predominately agricultural land-
scape that characterizes much of the southern half
of the state.

We tallied the nest tree species used by goshawks
at 37 nests in 1977-97, a sample including 25 nests
from 1996-97 and 12 nests from prior years (Table
2). Of these nests, 29 (78%) were built in decidu-
ous trees. Of these, 12 (41%) were in trembling
aspens ( Populus tremuloides) . Goshawks nested in a
broad array of other deciduous trees at ^17% of
the nest sites. Nest trees occurred in woodlands at
both early (e.g., trembling aspen and white birch
[Betula papyriferd]) and late serai stages (e.g., sugar
maple [A. saccharum ] and eastern hemlock [Tsuga
canadensis ] ) . Forest stands used for nesting includ-
ed such varied habitats as pine plantations, upland
maple ( Acer spp.) and maple-oak ( Quercus spp.)
woodlands, black ash ( Fraxinus nigra) swamps, and
aspen monotypes, as well as forest fragments in
southern Wisconsin. It is not surprising that the
goshawk appears to use diverse woodland habitats
for nesting in Wisconsin, given its wide breeding
distribution over two-thirds of the state and the
breadth of its nesting habitats throughout North

Table 2. Tree species used for nesting (%) by North-
ern Goshawks in Wisconsin, 1977—97.

Tree Species

No. Nest
Trees
(N = 37)

Trembling aspen ( Populus tremuloides)

12 (32)

Sugar maple ( Acer saccharum)

5(14)

Yellow birch ( Betula alleghaniensis)

5(14)

White pine ( Pinus strobus)

5(14)

Eastern hemlock ( Tsuga canadensis )

3(8)

Northern red oak ( Quercus rubra)

2(5)

Black ash (Fraxinus nigra)

2(5)

Red maple (Acer rubrum)

1 (3)

White birch (Betula papyrifera)

1 (3)

Basswood (Tilia americana )

1 (3)

America (Braun et al. 1996, Squires and Reynolds
1997). Elsewhere in the western Great Lakes re-
gion, trembling aspen was also used for nesting in
a majority (10 of 14) of recent (1994-96) goshawk
nest sites in Minnesota (Martell and Dick 1996). In
Michigan, aspens ( P. grandidentata and P. tremu-
loides) were again used more frequently than other
tree species by breeding goshawks (S. Postupalsky
pers. comm.).

Of the 37 goshawk sites sampled, four were in
pine plantations. Nests were built in white pine (Fi-
nns strohus, N — 3) and a trembling aspen (N = 1)
within the plantation. These four plantation nests,
all found by unbiased means, were located in
northeasternmost (N = 1 ) and southcentral ( N —
3) parts of the state. At least three of these four
plantation nests fledged young. S. Postupalsky
(pers. comm.) also reports that pine plantations
have recently been used as nest sites by successfully
breeding goshawks in Michigan. Squires and Rey-
nolds’ (1997) review of nest-site habitats used by
goshawks in North America did not report pine
plantations as occupied nesting habitat.

Our comparison of habitat features at goshawk
nest sites found by unbiased vs. potentially biased
means failed to detect statistically significant differ-
ences ( P > 0.002) between these two data sets for
any of the 23 analyzed features with the sample
sizes available (Table 3) . The overall statistical sim-
ilarity between habitat features at goshawk nests
found by unbiased vs. potentially biased methods
thus appeared to uphold the utility of a pooled
sample as a descriptor of goshawk nest-site habitat
in Wisconsin, a sample derived from a wide geo-

Sept 1998

Goshawk Distribution and Habitat

193

Table 3. Northern Goshawk habitat features at nests found by unbiased methods, potentially biased methods, and
pooled methods. Data are reported as mean values ±SE (95% confidence interval). P = exact probability value of
test between unbiased vs. biased categories.

Feature

Unbiased
N= 21

Biased
N= 16

Pooled
N= 37

P

Dist. to Water (m)

193.8 ± 43.3

241.1 ± 75. 9 a

213.5

40. 0 b (132.3-294.7)

0.75

Dist. to Opening (m)

184.0 ± 43.0

133.5 ± 21. 0 a

163.3

-1-

26. 6 b

(109.3-217.1)

0.89

Nest Tree Height (m)

23.7 ± 0.6

25.7 ± 0.8

24.6

±

0.5

(23.6-25.6)

0.05

Nest Tree dbh (cm)

35.7 ± 2.4

47.4 ± 3.3

40.8

H-

2.2

(36.4-45.2)

0.01

Nest Height (m)

14.1 ± 0.4

15.5 ± 0.6

14.7

+

0.4

(14.0-15.4)

0.12

Nest Percent (%)

59.9 ± 1.8

60.3 ± 2.1

60.1

1.3

(57.4-62.8)

0.85

Degree Slope (°)

6.2 ± 1.6

6.2 ± 1.6

6.2

1.2

(3.9-8.6)

0.79

Tall Shrub Index

14.4 ± 4.1

33.4 ± 10.2

22.6

-h

5.1

(12.2-33.0)

0.21

Low Shrub Index

46.9 ± 8.1

49.6 ± 10.7

48.0

-h

6.4

(35.1-61.0)

0.84

Under. Density (trees/ha)

283.3 ± 42.7

268.8 ± 42.5

277.0

30.0

(216.1-338.0)

0.78

Canopy Height (m)

23.9 ± 0.7

25.5 ± 0.7

24.6

0.5

(23.6-25.6)

0.10

Total Canopy (%)

78.7 ± 4.4

85.5 ± 3.1

81.6

+

2.9

(75.8-87.4)

0.20

Decid. Canopy (%)

61.5 ± 7.5

72.3 ± 7.7

66.2

H-

5.4

(55.3-77.1)

0.12

Conif. Canopy (%)

17.2 ± 6.5

13.1 ± 6.8

15.4

-h

4.7

(6.0-24.9)

1.00

Total Understory (%)

45.4 ± 6.0

48.1 ± 4.8

46.6

3.9

(38.6-54.6)

0.74

Decid. Understory (%)

40.3 ± 6.2

43.1 ± 5.0

41.5

±

4.1

(33.2-49.8)

0.65

Conif. Understory (%)

5.1 ± 3.3

5.0 ± 2.6

5.1

+

2.1

(0. 7-9.4)

0.94

Total Ground (%)

58.4 ± 4.3

43.9 ± 5.0

52.1

+

3.4

(45.2-59.1)

0.03

Decid. Ground (%)

56.4 ± 4.3

42.0 ± 5.3

50.2

+

3.5

(43.1-57.3)

0.04

Conif. Ground (%)

2.0 ± 1.1

1.9 ± 1.1

2.0

0.8

(0.4— 3.5)

0.75

Tree Density (trees/ha)

458.3 ± 63.4

376.6 ± 42.3

423.0

+

40.4

(341.0-505.0)

0.69

Basal Area (m 2 /ha)

28.2 ± 3.7

33.0 ± 3.2

30.3

■+*

2.5

(25.2-35.4)

0.16

Mean dbh (cm)

27.0 ± 1.4

32.9 ± 1.8

29.4

Ar

1.2

(27.0-31.9)

0.03

a N = 15 due to missing data.
b N = 36 due to missing data.

graphic area of the state. For this pooled sample,
mean nest-tree height was 25 m and mean nest-
tree dbh was 41 cm (Table 3) . Mean canopy height
(25 m) was identical to mean nest-tree height, but
mean tree dbh within 0.04-ha sample plots sur-
rounding nest trees was substantially less than
mean nest-tree dbh (29 vs. 41 cm). Mean canopy
closure was 82%. Squires and Reynolds (1997)
have suggested that such a high degree of canopy
closure is one of the most uniform aspects of hab-
itat at goshawk nest sites in North America.

Acknowledgments

We gratefully acknowledge the support and coopera-
tion of the many individuals who made this study possi-
ble, especially the assistance provided by S. Adams, J. Ax-
tell, A. Belleman, T. Booms, C. Hutler, B. Isenring, J.
Mills, D. Noble, L. Petersen, B. Pierce, and B. Smith. This
manuscript was greatly improved by the comments of D.
Andersen, W. Gould, M. Martell, J. Marzluff, and S. Pos-
tupalsky. We also thank the staff of the Necedah National
Wildlife Refuge of the U.S. Fish and Wildlife Service, and
the Chequamegon National Forest of the U.S. Forest Ser-

vice for their help. We thank The Nature Conservancy
and the Wisconsin Department of Natural Resources for
providing some nest locations. This project was funded
in part by the Bureau of Endangered Resources and the
Bureau of Forestry in the Wisconsin Department of Nat-
ural Resources, and the Wisconsin Falconers Association
The Personnel Development Committee at University of
Wisconsin at Stevens Point also provided research funds
and sabbatical support to the senior author.

Literature Cited

Apfelbaum, S.I. and P. Seelbach. 1983. Nest tree, habitat
selection, and productivity of seven North American
raptor species based on the Cornell University nest
record card program. Raptor Res. 17:97-113.

Block, W.M., M.L. Morrison and M.H. Reiser. 1994
The Northern Goshawk: ecology and management.
Stud. Avian Biol. 16:1-2.

Braun, C.E., J.H. Enderson, YB. Linhart, C.D. Marti
and M.R. Fuller. 1996. Northern Goshawk and forest
management in the southwestern United States.
Wildl. Soc., Tech. Rev. 96-2.

Gromme, O.J. 1963. Birds of Wisconsin. Univ. Wisconsin
Press, Madison, WT U.S.A.

194

Rosenfield et al.

Vol. 32, No. 3

Hayward, G.D. and R.E. Escano. 1989. Goshawk nest-site
characteristics in western Montana and northern Ida-
ho. Condor 91:476-479.

James, C.F. and H.H. Shugart. 1970. A quantitative meth-
od of habitat description. Audubon Field Notes 24:727-
736.

Johnsgard, P.A. 1990. Hawks, eagles, and falcons of
North America. Smithsonian Inst. Press, Washington,
DC U.S.A.

Kennedy, P.L. 1997. The Northern Goshawk {Accipiter gen-
tilis atricapillus ) : is there evidence of a population de-
cline? J. Raptor Res. 31:95-106.

Martell, M. and T. Dick. 1996. Nesting habitat charac-
teristics of the Northern Goshawk (Accipiter gentilis) in
Minnesota. Final Rep., Minnesota Dep. Nat. Res.,
Nongame Wildl. Program, Project No. 9407382. St.
Paul, MN U.S.A.

PALMER, R.S. 1988. Handbook of North American birds,
Volume 4. Yale Univ. Press, New Haven, CT U.S.A.

POSTUPALSKY, S. 1991. Northern Goshawk Accipiter gentilis.
Page 168 in R. Brewer, G.A. McPeek and R.J. Adams,
Jr. [Eds.] The atlas of breeding birds of Michigan.
Mich. State Univ., East Lansing, MI U.S.A.

Robbins, S.D., JR. 1991. Wisconsin birdlife, population
and distribution, past and present. Univ. Wis. Press,
Madison, WI U.S.A.

Rosenfield, R.N.,J. Bielefeldt, R.K. Anderson andJ.M.

Papp. 1991. Status reports: accipiters. Pages 42-49 in
B.E. Pendleton and D.L. Krahe [Eds.], Proceedings
midwest raptor management symposium and work-
shop. Natl. Wildl. Fed. Sci. Tech. Ser. 15, Washington
DC U.S.A.

SCOTT, W.E. 1964. Quarter-century history of the society.
Pass. Pigeon 26:28-62.

Siders, M.S. and P.L. Kennedy. 1994. Nesting of accipiter
hawks: is body size a consistent predictor of nest hab-
itat characteristics? Stud. Avian Biol. 16:92-96.

Sokal, R.R. and F.J. Rohlf. 1981. Biometry. 2nd Ed.
W.H. Freeman and Co., San Francisco, CA U.S.A.

Squires, J.R. and R.T. Reynolds. 1997. Northern Gos-
hawk (Accipiter gentilis) . In A.. Poole and F. Gill [Eds.],
The birds of North America, No. 298. The Academy
of Natural Sciences, Philadelphia, PA U.S.A.

Titus, K. and J.A. Mosher. 1981. Nest-site habitat select-
ed by woodland hawks in the central Appalachians.
Auk 98:270-281.

Wisconsin Department of Natural Resources. 1995.
Wisconsin’s biodiversity as a management issue. A re-
port to the department of natural resources manag-
ers. Wis. Dep. Nat. Res., Madison, WI U.S.A.

Wilkinson, L. 1992. SYSTAT; The system for statistics.
SYSTAT Inc. Evanston, IL U.S.A.

Received 29 December 1997; accepted 17 May 1998

J. Raptor Res. 32(3):195-201
© 1998 The Raptor Research Foundation, Inc.

SOLITARY AND SOCIAL HUNTING IN PALE CHANTING
GOSHAWK ( MELIERAX CANORUS) FAMILIES:

WHY USE BOTH STRATEGIES?

Gerard Malan 1

Percy FitzPatrick Institute, University of Cape Town, Rondebosch, 7700 South Africa

Abstract. — I observed Pale Chanting Goshawks ( Melierax canorus) using solitary and social hunting strat-
egies. Most goshawks hunted predominantly alone, but if an individual was unable to flush and catch a
cornered rodent from a shrub, other family members joined in a social hunt. Goshawks perched near
or on the tops of shrubs and repeatedly struck at rodents until they were caught. Other family members
did not pursue the goshawk that caught prey, even if it did not make the initial hunt. During social
hunts, there was no evidence of a dominance hierarchy in families when they were not hunting. I found
hunting success of individual goshawks to be low (11-12%) for both solitary and social hunts. Only
large rodents were caught during social hunts, whereas smaller vertebrates (lizards and birds), and
invertebrates, were caught during solitary hunts. It appeared that dominant breeders did not klepto-
parasitize or dominate subordinate family members during social hunts to maximize their individual
hunting success. Juveniles were significantly less successful than adults in capturing rodent prey, but
may have increased their foraging efficiency and survival by participating in social hunts. Dominant Pale
Chanting Goshawks that allowed offspring to partake in social hunts may, therefore, behave selfishly to
increase their inclusive fitness.

Key Words: Pale Chanting Goshawk, Melierax canorus; social hunting, juvenile survival; prey size; energy

intake.

Caza individual y social de Melierax canorus: por que utilizar ambas estrategias?

Resumen. — Observe a Melierax canorus utilizar estrategias de caza individual y social. La mayoria de los
azores cazan principalmente en forma individual, pero si un individuo no es capaz de capturar a un
roedor acorralado en un rastrojo, otros miembros de la familia se pueden unir en una caceria social.
Los miembros restantes de la familia no persiguen al azor que ha capturado la presa. Durante la caceria
social, no hubo evidencia de dominancia jerarquica la cual existe cuando no estan cazando. Encontre
que el exito individual de caza fue menor al 11-12% en ambas modalidades individual y social. Los
grandes roedores fueron capturados solo en cacerias sociales, mientras que los vertebrados mas pe-
quenos (lagartijas y aves), asi como tambien los invertebrados fueron capturados durante la caza indi-
vidual. Sugiero que los reproductores dominantes no practican el kleptoparasitismo o domiman a miem-
bros subordinados de la familia durante la caza social con el fin de maximizar el exito de la caza
individual. Los juveniles fueron menos exitosos que los adultos en capturar roedores pero pudieron
haber aumentado su eficiencia de forrajeo y sobreviviencia al participar en la caza social. Los dominantes
Melierax canorus que permitieron a sus hijos participar en la caza social pudieron haber actuado en
forma autosuficiente con el fin de aumentar su vigor.

[Traduccion de Cesar Marquez]

Predators can use various hunting strategies to
increase their individual foraging success. They
can hunt alone or in association with related or
unrelated conspecifics, or even with heterospecif-
ics (Packer and Ruttan 1988, Ellis et al. 1993). In
such hunting associations, they can pursue strat-

1 Present address: Department of Zoology, University of
Durban-Westville, PBX54001, Durban 4000, South Africa.

egies ranging from active participation, where all
individuals participate fully and benefit from so-
cial hunts, to kleptoparasitism (Hector 1986,
Scheel and Packer 1991, Heinsohn and Packer
1995, Steele and Hockey 1995). Predators may
adopt one or more of these strategies if their in-
dividual hunting success is low or if prey is large
and difficult to catch (Packer and Ruttan 1988).
The optimal combination of strategies should

195

196

Malan

Vol. 32, No. 3

maximize their net energy return (Hansen 1986,
Bednarz 1988),

The Pale Chanting Goshawk ( Melierax canorus ) is
a large, common raptor that inhabits the arid
regions of southern Africa. In one study in the Lit-
de Karoo, South Africa, Pale Chanting Goshawks
were found to live in family groups consisting of a
breeding unit of either a polyandrous trio (a pair
plus an additional cobreeding male) or a monog-
amous pair, with or without nonbreeders (up to
two) and juveniles (up to four) (Malan et al. 1996).
Cobreeders participated fully in reproductive activ-
ities, including copulations, but nonbreeders were
actively excluded from the nesting area during the
breeding season. Whereas polyandrous trios were
recorded only in broken veld, delayed dispersal by
nonbreeders and juveniles was the norm in all veg-
etation types. A dominance hierarchy existed in
families with the female breeder on top followed
by the male breeder and cobreeder and then the
nonbreeders and juveniles (Malan and Jenkins
1996). Although the Pale Chanting Goshawk is a
generalist feeder, relatively large rodent prey (45-
124 g, Otomys unisulcatus, Parotomys brantsii, and
Rhabdomys pumilio) that forage near vegetation or
in the open make up most of the biomass in its
diet (Malan and Crowe 1996). Other prey taxa in-
clude a range of other vertebrates as well as inver-
tebrates. Pale Chanting Goshawks are obligate
perch hunters and hunt from natural (trees or
shrubs) or artificial (fence posts and telephone
poles) perches from which they gently swoop to
the ground (Malan and Crowe 1997).

This study tests the hypothesis that Pale Chant-
ing Goshawks use solitary and social hunting to
maximize their individual hunting success in cap-
turing large and difficult to catch rodent prey I
observed the methods used by Pale Chanting Gos-
hawks during social hunts, as well as the size of the
prey caught during solitary and social hunts. Sec-
ondly, the solitary and social hunting strategies of
large families ( x — 5.5 goshawks) in one habitat
were compared with small families (x = 3.4) in an-
other habitat. Thirdly, I compared the hunting tac-
tics of juvenile Pale Chanting Goshawks with those
of adults as well as foraging fledglings, still depen-
dent on their parents for food.

Study Area and Methods

The 146 km 2 study area was located near Calitzdorp
(Little Karoo, 33°32'S, 21°48'E) in South Africa. It re-
ceives an average annual rainfall of 20 cm and the to-

pography is generally flat. It is utilized for extensive Os-
trich ( Struthio camelus ) farming.

Two semi-arid vegetation types occurred in the study-
area, broken veld (Karroid Broken Veld vegetation type ,
Acocks 1988) in the north and dwarf shrubland (Succu-
lent Karoo) in the south. Broken veld consisted of small
trees and shrubs (1-3 m high) scattered in a matrix of
low shrubs. Dwarf shrubland consisted of a sparse layer
of prostrate succulents and herbs. In dwarf shrubland,
Pale Chanting Goshawks only occupied areas with a high
availability of perches (mostly fenceposts) whereas bro-
ken veld with its abundant trees and shrubs was probably
saturated with Pale Chanting Goshawk families (Malan
1995).

I defined hunts as flights by goshawks from perches to
attack prey on the ground or in the air. During each
hunt, I aged the participating goshawk(s) as follows:
adults, juveniles or goshawks in immature plumage, and
fledglings or offspring still fed by their parents for up to
80 d after leaving nests (Malan 1995). For adults, hunting-
data of bl eeders and nonbreeders were combined. Three
hunt outcomes were recognized: successful hunts or
hunts that ended when goshawks landed on the ground
and caught prey, unsuccessful hunts or hunts that ended
when goshawks landed on the ground but failed to catch
prey, and abandoned hunts or hunts that ended when
goshawks flew down from perches and, upon reaching
the point of impact, briefly hovered about 1 m above
potential prey, then flew off without the prey. Hunting
of termites was not analyzed because they were not
chased (Malan and Crowe 1996).

Using instantaneous sampling (Lehner 1979), I fol-
lowed a focal Pale Chanting Goshawk by vehicle and re-
corded aspects of its hunting behavior every 60 sec. Dur-
ing each observation period, the focal goshawk was
followed from 60-300 min and, when it was out of sight,
the observation period was terminated. The hunting be-
havior of all other family members within 100 m of the
focal goshawk was also recorded. The study was conduct-
ed from February 1988-March 1989, but the hunting be-
havior of mated adults was only studied in the nonbreed-
ing and prelaying (from first copulation until egg-laying)
periods. The hunting behavior of goshawks was also re-
corded during casual observations during the summer
breeding seasons of 1989-95. Solitary and social hunts
were recorded during 64 observation periods (total ob-
servation time = 11 139 min, x = 174 min, SD = 67 min),
and solitary hunts during an additional 17 observation
periods (2074 min; x = 123, SD = 54 min). The hunting
behavior of 15 adults were studied for 57 observation pe-
riods (9398 min), five juveniles for 16 observation peri-
ods (2651 min), and two fledglings for eight observation
periods (1164 min). Capture rates per hour were calcu-
lated for each observation period and compared between
observations periods for single and social hunts, goshawk
age classes, and hunt outcomes.

A solitary hunt is defined as only the focal Pale Chant-
ing Goshawk hunting. A social hunt involved either the
focal goshawk hunting and being joined on the ground
by family members, or the focal goshawk joining family
members in a hunt. I termed these hunts “social” be-
cause family members hunted together in a nonaggres-
sive and cooperative manner. A social hunt was successful

Sept 1998

Pale Chanting Goshawk Hunting Strategies

197

Table 1. A comparison of solitary and social striking rates (per hour) by adult, juvenile, and feldgling Pale Chanting
Goshawks. During successful hunts, prey was caught. During unsuccessful hunts, goshawks landed on the ground but
failed to catch prey. During abandoned hunts, goshawks briefly hovered about 1 m above potential prey but flew off
without prey.

Adults

Juveniles

Fledglings

Kruskal-

Wai.lis

df

Solitary hunts

Successful

0.15 ± 0.24 1

0.07 ± 0.13

0.04 ± 0.10

2.66 ns

2

Unsuccessful

1.00 ± 1.05

1.74 ± 1.02

1.35 ± 1.45

8.49*

2

Abandoned

0.12 ± 0.24

0.21 ± 0.26

0.11 ± 0.31

4.24 ns

2

All solitary hunts

1.26 ± 1.13

2.02 ± 1.21

1.49 ± 1.69

6.99*

2

Social hunts

Successful

0.04 ±0.11

0.01 ± 0.06

0.00

1.51 ns

2

Unsuccessful

0.08 ± 0.19

0.02 ± 0.07

0.00

3.24 ns

2

Abandoned

0.05 ± 0.13

0.00

0.00

3.90 ns

2

All social hunts

0.17 ± 0.28

0.03 ± 0.09

0.00

6.01*

2

1 = mean ± 1 SD.
* = P < 0.05.

if any of the participating Pale Chanting Goshawks
caught prey. Due to my small sample size, only social
hunts involving two goshawks were analyzed. In all social
hunts analyzed, only adult Pale Chanting Goshawks
joined the focal adult or juvenile.

Pale Chanting Goshawk families are strictly territorial
and unrelated conspecifics were not tolerated within ter-
ritories (Malan and Jenkins 1996). Pale Chanting Gos-
hawks thus always hunted in association with family mem-
bers. This association was compared between the
significantly larger polyandrous families in broken veld
( x = 5.5 goshawks) and smaller monogamous families in
dwarf shrubland (x = 3.4; Malan 1995). The presence of
family members within a 100 m radius of the focal animal
was compared between three families each from broken
veld and dwarf shrubland. For each observation period,
I calculated the proportion of time spent alone or in
close proximity with one or more family members, either
adults, juveniles, or fledglings. Data from 21 focal indi-
viduals were analyzed for 63 observation periods(10 055
min; x = 160, SD = 66 min) and arcsine transformed to
improve normality (Zar 1984).

Results

Prey was attacked on the ground in 99% (N =
397) of all hunts. When prey was pursued on the
ground, it was chased actively, very often with wings
aloft and flapping. If vertebrate prey, such as an
otomyinid rodent, was cornered under a shrub and
a family member joined the focal goshawk on the
ground, the Pale Chanting Goshawks would sur-
round the shrub and/ or perch on top. Individuals
would then repeatedly strike at the rodent by
jumping into the shrub (flush-and-ambush strate-
gy; Bednarz 1988). In four hunts, all unsuccessful,

a Pale Chanting Goshawk attacked a bird from a
perch and actively chased the bird in horizontal
flapping flight.

The frequency of successful solitary hunts by
adults in dwarf shrubland (14%; N — 86) was not
significantly different from the frequency of suc-
cessful solitary hunts in broken veld (10%, N =
107, Log-likelihood Ratio with Yates correction: G c
= 0.31, P > 0.50) . The frequency of successful so-
cial hunts by adults also did not differ significantly
between broken veld (25%, N = 8) and dwarf
shrubland (21%, N = 14; G c = 0.12, P > 0.70).
The rates per hour that adults participated in suc-
cessful, unsuccessful, or abandoned solitary hunts
did not differ significantly (Rest, P > 0.05,) be-
tween large and small families. Likewise, the rates
per hour that adults participated in successful, un-
successful, or abandoned social hunts did not dif-
fer significantly (Rest, P > 0.05,) between large
and small families. My sample size prevented a
comparison between the hunting rates of success-
ful, unsuccessful, or abandoned hunts for both ju-
veniles and fledglings from large and small fami-
lies.

When the solitary and social striking rates of all
age classes were considered, the ratio of solitary to
social hunts by adults was significantly less than for
juveniles and fledglings (Table 1). Juveniles (4%,
N = 100) and fledglings (4%, N = 25) were equally
successful in solitary hunts, but were significantly
less successful than adults (11%, N = 193; G c =

198

Malan

Vol. 32, No. 3

Table 2. Percent time per observation period Pale
Chanting Goshawks hunted within 100 m of other family
members in large (x = 5.5 goshawks) and small families
(x - 3.4).

Small Large

Number Families Families

present

n = 3

n = 3

F

df

Zero

84 ± 44

85 ± 54

0.01 ns

1,46

One

15 ± 44

11 ± 38

0.36 ns

1,46

Two

0 ± 4

1 ± 4

2.62 ns

1, 46

Three

0 ± 0

1 ± 6

4.23*

1, 46

Four

0 ± 0

0 ± 3

2.04 ns

1,46

* = P < 0.05.

4.23, P < 0.05) . Juveniles engaged in social hunts
(2%, N = 102) significantly less often than adults
(10%, N = 215, G c = 6.75, P < 0.01), whereas
fledglings did not participate in social hunts at all.
During social hunts, once prey was caught, gos-
hawks flew off with the item and they were not
pursued by the remaining family members. Adults
caught prey during social hunts in 23% (N = 22)
of hunts. The adult that initiated the social hunt
caught the prey in 20% ( N = 10) of instances,
whereas the focal adult that joined the hunt caught
the prey in 27% (N = 12) of instances (G c = 0.05,
P > 0.75). The frequency of success of the solitary
hunts by adults (11%, N — 193) did not differ sig-
nificantly from their success in social hunts (23%,
N = 22, G c = 1.40, P> 0.10). Juveniles participated
in two social hunts, joining the hunt in both in-
stances, and in one of these hunts, the juvenile was
successful. During casual observations, juveniles
that participated in social hunts caught rodent
prey in four instances.

During all successful hunts only vertebrates (ro-
dents, lizards and birds) were caught. Hunts for

Table 3. Percent time per observation period adult, ju-
venile, and fledgling Pale Chanting Goshawks spent with-
in 100 m radius of other family members.

Number Fledg-

Present Adults Juveniles lings F df

Zero

+1

00

41

99

13

99

16

7.30**

2, 69

One

13 ±

38

1

+

13

1

-1-

16

7.07**

2, 69

Two

0 ±

16

0

0

2.18 ns

2, 69

Three

0 ±

10

0

0

0.97 ns

2, 69

Four

0 ±

6

0

0

0.49 ns

2, 69

** = p < 0.01.

arthropods were probably so quick, and in the low-
er vegetation layer, that they were not seen. During
solitary hunts, fledglings caught one lizard (Sauna
spp.), juveniles caught three lizards and one bird,
and adults caught 18 rodents (86%), two lizards
and one bird. During social hunts, only rodents
were caught with adults capturing five rodents and
one juvenile catching one rodent.

When the association of family members be-
tween large and small families was investigated,
adults of large families spent significantly more
time in close proximity (<100 m) to three family
members than did adults from small families (Ta-
ble 2). The proportion of time spent alone (x =
99, SD = 13%) or in close proximity to one family
member (x = 1%, SD = 13%, ANOVA, all P >
0.05) did not differ significantly between juveniles
of large and small families. Adults, compared to
juveniles and fledglings, spent significantly less
time alone and significantly more time in close
proximity to one family member (Table 3) .

Discussion

Despite the potential advantages associated with
hunting in groups, such as an increase in individ-
ual hunting success and energy return (Bednarz
1988), Pale Chanting Goshawks still predominandy
hunt alone. The average hunting success of indi-
vidual adult goshawks in social hunts was only
11.5%, half of the 23% success of social hunts in
which two goshawks participated. Nevertheless, it
was similar to the 1 1 % hunting success of individ-
uals in solitary hunts.

Why would Pale Chanting Goshawks follow two
hunting strategies that contribute the same to an
individual’s hunting success? For social hunting to
be a viable option, the individual benefits of this
hunting strategy must equal or exceed that of
hunting singly (Hansen 1986). First, such benefit
could only result if prey captured in family pursuits
is, on average, larger than that caught in solitary
pursuits (Steele and Hockey 1995). Pale Chanting
Goshawks preyed mostly on relatively large oto-
myinid rodents (mean body mass = 124 g), as well
as the smaller Khabdomys pumilio (mean body mass
= 45 g) (Malan and Crowe 1996). In broken veld
and dwarf shrubland, these rodents contributed
87% or 22 682 g of biomass and 68% or 249 in-
dividuals to the vertebrate diet. Pale Chanting Gos-
hawks also preyed on smaller mammals, small
birds, hatchling tortoises, small snakes and lizards,
as well as sunspiders, harvester termites, grasshop-

Sept 1998

Pale Chanting Goshawk Hunting Strategies

199

pers and beetles (Malan and Crowe 1996). The av-
erage mass of rodents caught in the two vegetation
types was 90 ± 40 (±1 SD) g, birds 70 ± 34 g, and
reptiles 12 ± 9 g (Malan and Crowe 1996). Thus,
because only rodents were captured in social
hunts, the average size of prey captured in this way
was indeed larger than those caught in solitary
hunts. In terms of hunting socially, it was also the
prey biomass obtained during these hunts, and not
only the relative success or hunting technique
used, that was important to each individual. Al-
though the hunting success in solitary and social
hunts was equal, the energy returns from hunting
large animals in social hunts may have surpassed
the returns from hunting smaller prey in solitary
hunts.

A second reason why Pale Chanting Goshawks
may use two hunting strategies is that their individ-
ual hunting success in catching vertebrate prey in
solitary hunts is low. Solitary adult hunting success
of 11% was substantially lower than the mean of
59% (range - 31-72%) for 11 raptor species that
hunt ground-dwelling prey (Toland 1986). It is
even lower than the 19-33% success (x = 27%) for
raptors that hunt other birds in the air, a technique
generally thought to be less successful than search-
ing for prey on the ground (Toland 1986) . The low
success of Pale Chanting Goshawks highlights the
difficulty they experience in catching vertebrate
prey in a shrub-rich substrate. Since solitary hunt-
ers have low success of catching large vertebrate
prey. Pale Chanting Goshawks may adopt a social
hunting strategy to supplement their solitary hunt-
ing and thus increase their overall hunting success.

In spite of the apparent benefits of hunting so-
cially, adult Pale Chanting Goshawks did not ha-
bitually hunt together. Adults spent only 15% of
time within 100 m of family members, compared
with 71% of the time that Harris’ Hawks ( Parabuteo
unicinctus) spent within 50 m of group members
(Bednarz 1988). Pale Chanting Goshawks could,
however, visually monitor each other’s movements
by perching on the highest available perch. The
flapping wing motions during a pursuit may act as
a signal to other family members that a hunt is in
progress. If the prey animal was cornered, a soli-
tary Pale Chanting Goshawk probably cannot act
as a hunter and a beater, and would thus fail in its
solitary attack strategy. If the hunting goshawk
could attract family members, however, it would
have some chance of obtaining the prey. The hunt-
ing behavior of the initiator and the goshawks that

subsequently join hunts, therefore, appear to be
selfish. If individuals behaved selfishly during social
hunts, then why did dominant Pale Chanting Gos-
hawks not attempt to increase their hunting suc-
cess by kleptoparasitizing subordinate family mem-
bers and why was there no aggressive behavior
observed between family members during a social
hunt? Pale Chanting Goshawks do kleptoparasitize
Booted Eagles ( Hieraaetus pennatus ) with rodent
prey (unpubl. data) and at nesting sites, subordi-
nate cobreeding Pale Chanting Goshawks do trans-
fer prey to the dominant female and male breeder,
but not vice versa (Malan and Jenkins 1996). Pack-
er and Ruttan (1988) predicted that if single prey
items are hunted, but not shared amongst partici-
pants, group members will always cooperate fully
in hunts. Dominant Pale Chanting Goshawk breed-
ers may not kleptoparasitize subordinate members
because the initial benefit (suckers payoff, Axelrod
and Hamilton 1981) of hunting prey not to be
shared may result in defection by subordinates,
with a subsequent decrease in the rate of social
hunts. Likewise, if dominant breeders exert their
dominance on subordinate members upon arrival
at cornered prey, prey may escape and again no
benefits can be gained by the goshawks participat-
ing. By displaying no obvious aggression towards
each other during social hunts, not perceived to
be a common trait among raptors (Faaborg and
Bednarz 1990), each family member may increase
its individual hunting success.

Given the increase in hunting success when com-
bining social and solitary hunting, one would ex-
pect not only the hunting success per Pale Chant-
ing Goshawk in bigger families to be greater, but
also individuals from bigger families would be ex-
pected to spend more time hunting socially. Sur-
prisingly few differences were found between the
hunting strategies of large and small families (5.5
vs. 3.4 goshawks) . Emlen (1994) suggested that the
benefits of social activities such as social hunting
may be secondarily derived after families formed
because goshawks were constrained through a fac-
tor such as a lack of territorial space, from dis-
persing, and breeding in pairs. Even if the benefits
of social hunting are secondarily derived, I suggest
that hunting in families may hold fitness benefits
for participants.

First, the participation by rapacious juveniles in
social hunts may hold additional benefits associat-
ed with hunting relatively large prey that are not
shared by family members (Stacey and Ligon

200

Malan

Vol. 32, No. 3

1987). Prior to independence, young raptors ex-
perience a high mortality rate (Newton 1995),
partly because of their low foraging efficiency
(Heinsohn 1991). If they are raised in a social fam-
ily and delay dispersal from that family, the benefits
of philopatry may include participation in social
hunts (Heinsohn et al. 1988). Juveniles were in-
volved in social hunts, albeit at a very low rate
(0.03/hr). Furthermore, they were only able to
catch lizards (estimated mass 10 g; Malan and
Crowe 1996) in solitary hunts, but caught rodents
in social hunts. The benefits of hunting relatively
large rodent prey may increase their foraging ef-
ficiency and survival during the critical 12—16 mo
of their life. If this was the case, it is difficult to
explain why these juveniles engaged in social hunts
significantly less often than adults did. The fact
that juveniles only occupied a segment of the ter-
ritory in close proximity to the nesting site (un-
publ. data), may have made them less able to de-
tect family members hunting in other segments of
the territory.

Second, breeders may also gain fitness benefits
from hunting with their offspring. The success
from hunting socially, measured in terms of surviv-
al fitness, may be higher than if determined di-
rectly from the hunting success of individuals in
the family (Packer and Ruttan 1988, Koenig and
Mumme 1990). The act of allowing other family
members to partake in social hunts may thus ben-
efit the individual that cornered the prey indirect-
ly, as the loss in direct fitness is compensated by a
gain in indirect fitness. Individual Pale Chanting
Goshawks that allowed other family members to
partake in social hunts may be therefore behaving
selfishly to increase their inclusive fitness.

Acknowledgments

This research was supported in part by grants from the
Foundation for Research Development (M.Sc. bursary),
the Frank M. Chapman Memorial Fund (American Mu-
seum of Natural History), the Bob Blundell Memorial
Scholarship, the Leslie Brown Memorial Grant, the Uni-
versity of Cape Town Equipment and Research Commit-
tee, and the Percy FitzPatrick Institute. I thank Grant
Benn, Tim Crowe, Phil Hockey and David Jacobs for valu-
able comments of an earlier draft. I also thank the land-
owners at Calitzdorp for access to their properties.

Literature Cited

Acocks, J.P.H. 1988. Veld types of South Africa, Mem. Bot.
Surv. S. Afr. No. 57.

Axelrod, R. and W.D. Hamilton. 1981. The evolution
of cooperation. Science 211:1390— 1396.

Bednarz, J.C. 1988. Cooperative hunting in the Harris’
Hawk ( Parabuteo unicinctus). Science 239:1525—1527.

Ellis, D.H.,J.C. Bednarz, D.G. Smith and S.P. Flemming.
1993. Social foraging classes in raptorial birds. Bio-
Science 43:14—20.

Emlen, S.T. 1994. Benefits, constrains and the evolution
of the family. Trends Ecol. Evol. 9:282-284.

Faaborg, J. and J.C. Bednarz. 1990. Galapagos and Har-
ris’ Hawks: divergent causes of sociality in two raptors.
Pages 359-383 in P.B. Stacey and W.D. Koenig [Eds.
Cooperative breeding in birds. Cambridge Univ.
Press, Cambridge, U.K.

Hansen, A.J. 1986. Fighting behavior in Bald Eagles: a
test of game theory. Ecology 67:787-797.

Hector, D.P. 1986. Cooperative hunting and its relation-
ship to foraging success and prey size in an avian
predator. Ethology 73:247-257.

Heinsohn, R.G. 1991. Slow learning of foraging skills and
extended parental care in cooperative breeding
White-winged Choughs. Am. Nat. 137:874-881.

, A. Cockburn and R.B. Cunningham. 1988. For-
aging, delayed maturation, and advantages of coop-
erative breeding in White-winged Choughs, Carcorax
melanorhmphos. Ethology 77:177-186.

and C. Packer. 1995. Complex cooperative strat-
egies in group-territorial African lions. Science 269:
1260-1262.

Koenig, W.D. and R.L. Mumme. 1990. Population ecology
of the cooperative breeding Acorn Woodpecker.
Princeton Univ. Press, Princeton, NJ U.S.A.

Lehner, P.N. 1979. Handbook of ethological methods.
Garland STMP Press, New York, NY U.S.A.

Malan, G. 1995. Cooperative breeding and delayed dis-
persal in the Pale Chanting Goshawk Melierax canorus.
Ph.D. dissertation, University of Cape Town, Ronde-
bosch, South Africa.

and T.M. Crowe. 1996. The diet and conservation

of monogamous and polyandrous Pale Chanting Gos-
hawks in the Little Karoo, South Africa, S. Afr. J. Wildl.
Res. 26:1-10.

and A.R. Jenkins. 1996. Territory and nest de-
fense in the Pale Chanting Goshawks: do the co-
breeders help? S. Afr. J. Zool. 31:170-176.

and T.M. Crowe. 1997. Perch availability and prey

visibility: factors that may constitute habitat quality in
the Pale Chanting Goshawk. S. Afr. J. Zool. 32:14—20.

, T.M. Crowe, R. Biggs and J.J. Herholdt. 1997.

The social system of the Pale Chanting Goshawk Me-
lierax canorus: monogamy versus polyandry and de-
layed dispersal. Ibis 139:313-321.

Newton, I. 1995. The contribution of some recent re-
search on birds to ecological understanding./. Animal
Ecol. 64: 675-696.

Packer, C. and L. Ruttan. 1988. The evolution of co-
operative hunting. Am. Nat. 132:159-198.

Scheel, D. AND C. Packer. 1991. Group hunting behavior

Sept 1998

Pale Chanting Goshawk Hunting Strategies

201

of lions: a search for cooperation. Anim. Behav. 41:
697-709.

Stacey, P.B. and J.D. Ligon. 1987. Territory quality and
dispersal options in the Acorn Woodpecker, and a
challenge to the habitat-saturation model of cooper-
ative breeding. Am. Nat. 130:654—676.

Steele, W.K. and P.A.R. Hockey. 1995. Factors influenc-
ing rate and success of intraspecific kleptoparasitism

among Kelp Gulls ( Larus dominicanus) . Auk 112:847-
859.

Toland, B. 1986. Hunting success of some Missouri rap-
tors. Wilson Bull. 98:116-125.

Zar, J.H. 1984. Biostatistical analysis. Prentice-Hall, Inc.,
Princeton, NJ U.S.A.

Received 20 May 1996; accepted 9 May 1998

BUTEO BOOKS

The following Birds of North America Species Accounts are available through Buteo Books, 3130 Laurel Road,
Shipman, VA 22971. TOLL-FREE ORDERING: 1-800-722-2460; FAX: (804) 263-4842.

Barn Owl (1). Carl D. Marti. 1992. 16 pp.

Boreal Owl (63). G.D. Hayward and P.H. Hayward. 1993. 20 pp.

Broad-winged Hawk. (218). L.J. Goodrich, S.C. Crocoll and S.E. Senner. 1996. 28 pp.

Burrowing Owl (61). E.A. Haug, B.A, Millsap and M.S. Martell. 1993. 20 pp.

Common Black-hawk (122). Jay H. Schnell. 1994. 20 pp.

Cooper’s Hawk (75). R.N. Rosenfield and J. Bielefeldt. 1993. 24 pp.

Crested Caracara (249). Joan L. Morrison. 1996. 28 pp.

Eastern Screech-owl (165). Frederick R. Gehlbach. 1995. 24 pp.

Ferruginous Hawk (172). MarcJ. Bechard and Josef K. Schmutz. 1995. 20 pp.

Flammulated Owl (93). D. Archibald McCallum. 1994. 24 pp.

Great Gray Owl (41). Evelyn L. Bull and James R. Duncan. 1993. 16 pp.

Gyrfalcon (114). Nancy J. Clum and Tom J. Cade. 1994. 28 pp.

Harris’ Hawk (146). James C. Bednarz. 1995. 24 pp.

Long-eared Owl (133). J.S. Marks, D.L. Evans and D.W. Holt. 1994. 24 pp.

Merlin (44). N.S. Sodhi, L. Oliphant, R James and I. Warkentin. 1993. 20 pp.

Northern Saw-whet Owl (42). Richard J. Cannings. 1993. 20 pp.

Northern Goshawk (298). John R. Squires and Richard T. Reynolds. 1997. 32 pp.

Northern Harrier (210). R. Bruce MacWhirter and Keith L. Bildstein. 1996. 32 pp.

Red-shouldered Hawk (107). Scott T. Crocoll. 1994. 20 pp.

Red-tailed Hawk (52). C.R. Preston and R.D. Beane. 1993. 24 pp.

Short-eared Owl (62). D.W. Holt and S.M. Leasure. 1993. 24 pp.

Snail Kite (171). P.W. Sykes, Jr., J. A. Rodgers, Jr. and R.E. Bennetts. 1995. 32 pp.

Snowy Owl (10). David F. Parmelee. 1992. 20 pp.

Spotted Owl (179). R.J. Gutierrez, A.B. Franklin and W.S. Lahaye. 1995. 28 pp.

Swainson’s Hawk (265). A. Sidney England, MarcJ. Bechard and C. Stuart Houston. 1997. 28 pp.
Swallow-tailed Kite (138). Kenneth D. Meyer. 1995. 24 pp.

White-tailed Hawk (30). C. Craig Farquhar. 1992. 20 pp.

White-tailed Kite (178). Jeffrey R. Dunk. 1995. 16 pp.

J. Raptor Res. 32(3):202-207
© 1998 The Raptor Research Foundation, Inc.

FOREST MANAGEMENT EFFECTS ON NESTING HABITAT
SELECTED BY EURASIAN BLACK VULTURES
(. AEGYPIUS MONACHUS) IN CENTRAL SPAIN

Juan A. Fargallo

Departamento de Ecologia Evolutiva, Museo Nacional de Ciencias Naturales ( CSIC), J. Gutierrez Abascal 2,

E-28006 Madrid, Spain

Guillermo Blanco

Departamento de Biologia Animal, Universidad de Alcala de Henares, 28871 Madrid, Spain

Eduardo Soto-Largo

Pza. Mariano de Cavia 1, 28007 Madrid, Spain

Abstract. — We studied two recently established colonies of Eurasian Black Vultures ( Aegypius monachus) .
One was located in an abandoned maritime pine (Finns pinaster) plantation formerly used for resin
production. The other colony was in a Scots pine ( P silvestris) plantation currently used for timber
production. The vultures used nest sites with mature trees in forest openings and on steep slopes in
the medium-upper portions of mountains. These openings had few roads. Differences in tree density,
distance to nearest neighbor’s nests, and tolerance to high road density were observed between the nest
sites used by the two colonies. Sylvicultural practices in either forests explained differences in nest-site
selection between the colonies.

Key Words: Eurasian Black Vulture; Aegypius monachus; forest management, nest-site habitat.

Efecto del manejo del bosque sobre la seleccion del habitat de nidificacion del Buitre Negro ( Aegypius
monachus ) en Espana central

Resumen. — Se ha estudiado el efecto del manejo del bosque sobre los requerimientos del habitat de
nidificacion del Buitre Negro ( Aegypius monachus ) en dos colonias de reciente ocupacion. Las colonias
se establecieron en dos plantaciones de pino autoctono del Sistema Central espanol. Ambos bosques se
encuentran sometidos a distinto tipo de explotacion. En uno de ellos ( Pinus silvestris ) se realiza una
explotacion maderera y el otro (Pinus pinaster) se encuentra en la actualidad en desuso tras haber estado
dedicado a la extraccion de resina. En ambos bosques los buitres seleccionaron lugares de nidificacion
con arboles maduros para colocar los nidos, en areas mas desprovistas de vegetacion, con una mayor
pendiente y situados en el tercio medio-superior de las montanas. En estas areas tambien se encontro
un menor numero de caminos. Se observaron diferencias en la densidad de arboles, en la distancia al
nido ocupado mas proximo y en la tolerancia a los caminos entre las dos colonias. El diferente uso del
bosque explica estas diferencias.

[Traduccion Autores]

The Eurasian Black Vulture (Aegypius monachus)
is a widespread Palearctic species that has suffered
a marked decline in Europe in the last century,
disappearing from 15 countries (Bijleveld 1974,
Tucker and Heath 1994) and experiencing a
marked reduction in range and abundance in
Greece, Russia, Spain, Ukraine, and also likely in
Turkey (Blanco and Gonzalez 1992, Tucker and
Heath 1994). About 200 pairs remained in Spain
from 1966-72 (Bernis 1966, Hiraldo 1974). Nev-

ertheless, in the last 20 yr, the population has in-
creased to an estimated 774 pairs in 1990 (Gon-
zalez 1990) and 900-1000 pairs in 1992 accounting
for between 67-90% of the breeding European
population (Gonzalez in Tucker and Heath 1994).
Few studies have focused on the analysis of habitat
requirements of this species with only cursory men-
tion of some colonies (Hiraldo 1974, Torres et al.
1980, Bermejo 1990, Donazar 1993).

Loss and alteration of habitat are the most seri-

202

Sept 1998

Habitat Selected by Eurasian Black Vultures

203

ous threats for raptor populations (Bijleveld 1974,
Tucker and Heath 1994). Their large territories,
home ranges, and historical persecution by hu-
mans have made populations of forest breeding
raptors sensitive to forest management and habitat
change (Fuller 1996, Niemi and Hanowski 1997).
Indeed, declines in Eurasian Black Vulture popu-
lations had been previously attributed to altera-
tions of nesting habitats (Hiraldo 1974, Bermejo
1990, Blanco and Gonzalez 1992, Donazar 1993).
The spread of human settlements and intensive ag-
riculture in lowlands and pastures in highlands
have caused forest fragmentation and loss of for-
ested areas. Further, intensive forestry has created
mosaics of mature stands mixed with successional
stands of various ages (e.g., Moorman and Chap-
man 1996, Niemi and Hanowski 1997).

The Eurasian Black Vulture is closely associated
with forests during the breeding season (7-8 mo)
and afterwards because it uses trees as roosts and
shelters (Bernis 1966, Hiraldo and Donazar 1989).
Understanding the interaction between forest
management practices and habitat selection by the
Eurasian Black Vulture is crucial in planning con-
servation measures for this threatened species. Fur-
thermore, recent colonization of some areas ena-
bles us to determine its habitat requirements
because individuals select optimal vacant nesting
sites, unlike older colonizations where individuals
have been forced to occupy lower quality nesting
sites (Brown 1969, Kadmon 1993, Ferrer and Don-
azar 1996).

Study Area and Methods

Two recently established colonies of Eurasian Black
Vultures were studied in two pine forests of the Spanish
Central Range (Sierra de Guadarrama, 40°30'N). The
first colony (colony A) was located in a young maritime
pine ( Firms pinaster) plantation (forest A) in the province
of Madrid (mean elevation = 1150 m, range = 930-
1387). This pine species is native of the Spanish Central
Range (Mallada 1892). The old pine forest was felled
during the Spanish Civil War (1936-39) and the post-war
period, and only a few old pines remained. After the war,
the area was reforested with a new plantation of pines,
making the new forest a mixture of mostly trees about
60-yr old interspersed with few mature trees. Resin ex-
traction was the main use of this forest. The first pair of
Eurasian Black Vultures became established in this area
in 1985 and, by 1994, the colony consisted of 11 pairs.
The second colony (colony B) was in a Scots pine ( Pinus
sylvestris) plantation (forest B) in the province of Segovia
(mean elevation = 1600 m, range = 1315-2196). Scots
pine is also native of the Spanish Central Range (Galera
and Martin 1990). Nowadays, wood exploitation for tim-
ber is the main use of this forest. In past years, trees were

Table 1. Variables used to characterize Eurasian Black
Vulture nesting and random plots. Variables were mea-
sured from a point in the center of the plot.

DT — Stem diameter tree at breast height in cm
HT — Height of tree in m

DNN — Distance of nearest nest of same species in m
SCH — Height of scrubland in plot in m
DTT — Mean distance to three nearest trees in m
PCC — Percent canopy cover in plot
PUC — Percent understory cover in plot
PSC — Percent of scrubland in plot
PPL — Percent of pastureland in plot
PRL — Percent of rockyland in plot
NT1 — Number of trees in plot with a stem diameter
<15 cm

NT2 — Number of trees in plot with a stem diameter 16-
30 cm

NT3 — Number of trees in plot with a stem diameter 31-
50 cm

NT4 — Number of trees in plot with a stem diameter
>51 cm

NOT — Total number of overstory trees in plot
ATT — Elevation of plots in meters

1 11 — Topographic irregularity index (number of 20 m
contours cut by four 500 m lines from nest tree in
the four cardinal compass directions)

NPM — Nest position on mountain by mountain division
(1 = upper third, 2 = medium third, and 3 =
lower third)

DNV — Distance to nearest village in m
MOR — Meters of paved and unpaved roads within plots
with a radius of 500 m from nest trees

felled in large areas and newly reforested. Presently, this
forest is intensively managed and only individual mature
trees are selectively cut leaving some protected areas for
natural regeneration. These areas are located in the up-
per third of the mountains. Vultures became established
in this forest in 1989, and, by 1994, the colony consisted
of seven pairs. The two colonies are 32 km apart.

We used procedures suggested by Titus and Mosher
(1981), Andrew and Mosher (1982), Bosakowski et al.
(1992), Donazar et al. (1993), and Moorman and Chap-
man (1996) to study habitat selection by these vultures.
Nest sites were located in the spring of 1994 and plotted
on a 1:25 000 topographic map. Characteristics of nest
sites were measured in the field during the winter of
1994-95. A nest site was defined as the nest tree plus a
circular plot of 25 m radius around it (Titus and Mosher
1981). Vegetation cover within these 25-m radius plots
was visually estimated from the center of each subplot
(Prodon 1976, Rubio and Carrascal 1994). We also noted
the number of trees in different diameter categories and
the heights of nest trees. Measurements on other habitat
variables (Table 1) studied were obtained from the top-
ographic map.

Random sampling was used to estimate available nest-

204

Fargallo et al.

Vol. 32, No. 3

Table 2. Sample means ± SD of nest sites and random plots in each colony and forest. ANOVA F values are given
for differences between random and nest site plots, between nest sites of each colony, and between random plots of
each forest.

Colony A

Colony B

Var.

Nest Sites
(16V

Random

(31)

F

(df =

P

1,44)

Nest Sites
(13)

Random

(37)

F

(df =

P

1,47)

DT

57.6 ± 21.0

32.8 ± 10.7

27.5

0.00

62.9 ± 13.5

51.4 ± 12.7

7.6

0.00

HT

13.3 ± 1.7

10.0 ± 2.4

25.1

0.00

13.8 ± 2.2

15.2 ± 3.5

1.1

0.31

DNN

379 ± 205

582 ± 347

5.1

0.02

1362 ± 930

731 ± 478

1.9

0.17

SCH

1.6 ± 0.3

1.8 ± 0.6

1.2

0.28

1.3 ± 1.0

1.8 ± 0.9

2.4

0.26

DTT

8.9 ± 5.6

4.0 ± 2.7

14.1

0.00

7.6 ± 3.1

5.4 ± 5.3

4.3

0.04

PCC

42.0 ± 20.7

64.3 ± 24.5

7.1

0.01

40.6 ± 23.8

62.5 ± 31.4

3.4

0.07

PUC

22.4 ± 19.4

5.5 ± 10.6

20.1

0.00

10.6 ± 12.6

11.9 ± 14.9

0.0

0.90

PSC

73.1 ± 26.7

68.7 ± 28.9

0.5

0.48

35.6 ± 26.4

43.1 ± 29.4

0.2

0.60

PPL

1.3 ± 5.0

19.0 ± 25.6

8.0

0.00

17.8 ± 18.4

24.4 ± 24.7

0.1

0.72

PRL

25.6 ± 24.1

12.3 ± 16.4

3.3

0.07

21.1 ± 25.3

7.6 ± 13.2

2.1

0.15

NT1

12.8 ± 22.6

24.6 ± 24.3

1.5

0.23

20.2 ± 13.6

38.6 ± 56.5

0.0

0.84

NT2

7.8 ± 4.6

23.5 ± 25.9

11.3

0.00

21.6 ± 17.2

25.8 ± 17.4

0.6

0.44

NT3

2.8 ± 2.4

7.8 ± 8.3

6.8

0.01

17.3 ± 10.5

31.4 ± 22.8

2.5

0.12

NT4

0.7 ± 0.9

0.4 ± 0.8

1.1

0.29

11.0 ± 9.9

11.0 ± 10.3

0.1

0.71

NOT

24.3 ± 24.0

57.0 ± 45.5

10.4

0.00

70.1 ± 24.0

107 ± 66

1.1

0.29

ATT

1158 ± 43.7

1135 ± 93.6

1.3

0.26

1761 ± 46.0

1591 ± 42

15.9

0.00

TII

20.6 ± 3.3

14.7 ± 4.7

17.5

0.00

19.3 ± 3.0

15.9 ± 4.8

3.6

0.06

NPM

1.3 ± 0.4

1.8 ± 0.8

6.2

0.01

1.4 ± 0.5

2.1 ± 0.9

6.3

0.02

DNV

1521 ± 542

1774 ± 715

0.7

0.40

4075 ± 869

4317 ± 1393

0.1

0.78

MOR

235 ± 217

501 ± 333

3.8

0.05

133 ± 216

408 ± 425

7.0

0.01

a Sample size.

ing habitat. In each colony, random points were plotted
on the map with a numbered grid (Titus and Mosher
1981, Hubert 1993). Random points were always includ-
ed within forested areas, excluding those habitats in
which vultures did not nest (e.g., young pine planta-
tions). Once random points were located, the nearest
tree was randomly selected as the center of the plot.

To compare nearest-neighbor distances, only nests occu-
pied during the 1994 breeding season were considered.
As in other colonies (Bernis 1966, Hiraldo 1983), some
pairs did not breed during the year of our study, but the
vultures remained close to their nests during the breed-
ing season. For this reason, we considered these nest sites
as occupied in order to determine the distance between
neighboring nests sites.

We considered all the nests found in the winter of
1994-95, even though some were likely old nesting sites
of the same vulture pairs. An analysis of variance (STA-
TISTICS 1993) using planned comparisons was designed
to analyze for differences between nests and random
plots, and to analyze for differences between habitats.
This approach allowed us to include all nest sites in the
analysis, avoiding potential pseudoreplication that may
have resulted from including nest sites of the same birds.
Variables deviating from normality were logarithmic
transformed, and percentages were arcsin-transformed.

Results

All the nests found in the two colonies were in
pines. In both colonies, we detected a tendency to
nest in trees with the greatest diameter (DT, Table
2). Nest trees also were farther from neighboring
trees (DTT) and, consequently, had lower canopy
covers (PCC). These results indicated a tendency
for Eurasian Black Vultures to nest in openings.
Percentage and height of scrubland (PSC and
SCH, respectively) did not affect nest-site selection.

In terms of nest-site topography, vultures fre-
quentiy selected areas located in the middle-upper
third of the mountains (NPM), with steep slopes
(TII, Table 2) . Both colonies occurred at high el-
evations (>1000 m). The minimum elevation
(ATT) for a nest was 1090 m (colony A) and the
maximum was 1880 m (colony B)

In relation to human disturbance, Eurasian
Black Vultures nested in areas with fewer meters of
road (MOR) in the surrounding area but the prox-
imity of human habitation (DNV) did not appear

Sept 1998

Habitat Selected by Eurasian Black Vultures

205

Table 2. Extended .

Diff.

A-B

Diff.

A-B

F

(df

Nests

P

= 1,27)

Random
F P

(df = 1.66)

0.0

0.84

41.3

0.00

0.2

0.69

46.8

0.00

5.9

0.03

1.0

0.32

1.6

0.23

0.0

0.91

0.4

0.54

1.1

0.31

0.8

0.40

0.0

0.98

3.6

0.07

5.5

0.02

8.6

0.00

15.5

0.00

15.6

0.00

3.1

0.09

0.7

0.18

1.5

0.23

1.4

0.25

1.0

0.32

1.9

0.18

0.1

0.72

12.5

0.00

46.9

0.00

19.2

0.00

78.0

0.00

6.3

0.02

13.2

0.00

136.2

0.00

253.3

0.00

0.1

0.78

1.2

0.27

2.3

0.14

1.2

0.29

39.5

0.00

88.1

0.00

5.9

0.03

2.0

0.17

to affect nest-site selection. The closest nest was
800 m from a village.

Forest B had a higher tree density (NT3, NT4,
and NOT) with taller (HT) and thicker (DT) trees
than those in forest A (Table 2) . Forest B also had
a smaller percentage of scrubland (PSC), higher
elevation, and a greater distance from the closest
village.

Nest trees used by vultures in colony A were
taller than in random plots but, in colony B, they
were not. Furthermore, in colony A, nest sites had
a higher proportion of understory cover (PUC)
with lower overstory tree density (NOT) . There was
also a trend to nest in sites with a higher percent-
age of rocky land (PRL) . Vultures in colony B nest-
ed in areas with tree densities similar to those of
random plots and with higher densities than those
of colony A. This was due to the different structur-
al characteristics of the two plantations, plantation
B having a higher average tree density. Also, in col-
ony A but not in B, the percentage of pastureland

(PPL) was smaller in nest plots than in random
plots.

We found a mean distance of 637 m between
adjacent nest sites (DNN) of different pairs of vul-
tures in the two colonies. The minimum distance
was 150 m (colony A) and the maximum was 2325
m (colony B). The distance between neighboring
nests was significantly shorter in colony A (379 m)
than in B (1362 m).

Between colony differences in the nest-site selec-
tion in regard to DNV were due to the differences
in the distance between both plantations, colony A
being significantly closer to a village. MOR was sim-
ilar in both forests but nest plots in colony A had
significandy more meters of road than those in col-
ony B.

Discussion

Management practices resulted in a modifica-
tion of the pine forests we studied including alter-
ation of structural diversity, sizes of trees, and spe-
cies composition. These effects were more
pronounced when forests were used for timber
production than when they were used for resin col-
lection. However, the current synthetic elaboration
of glue has provoked a decreasing demand of resin
in home and international markets causing the
end of traditional management for resin extraction
(Gil et al. 1990) . Furthermore, the damage caused
to the trees by the stem cuts results in a low quality
of wood timber, which is hardly marketable (Gil et
al. 1990). In this management type, the size of tree
is less important and tree felling does not occur.

Eurasian Black Vultures, like other large vultures
such as the Lappet-faced Vulture ( Torgos tracheli-
otus ) , White-headed Vulture ( Trigonoceps occipitalis ) ,
and White-backed Vulture ( Gyps africanus ), build
their nests in the tops of trees (Houston 1974,
Mundy 1982, Husder and Howells 1988). Eurasian
Black Vultures build large nests with a diameter of
nearly 160 cm and 93 cm in height (Torres et al.
1980). These large nests frequendy fall to the
ground because of their weight (Bernis 1966).
Most nest trees used in both colonies were older
than 60 yr, indicating that vultures prefer mature
trees in which to build their nests. This preference
is probably due to the fact that large trees are nec-
essary to hold their nests. Because there were only
a few mature trees in colony A and these were dis-
tributed homogeneously on the upper and middle
third of the mountain, their preference for older
trees restricted the distribution of vultures in forest

206

Fargallo et al.

Vol. 32, No. 3

A. In colony B, vultures used old trees restricted to
the middle-upper third of the mountains where
the forest had been protected from timber har-
vesting. There, trees located next to the crest of
mountains are more exposed to hard climatic con-
ditions and were shorter than expected in relation
to the diameter of the stem.

Habitat around nests was characterized by open
areas with pronounced slopes. Hiraldo and Dona-
zar (1989) found Eurasian Black Vultures breeding
in steep areas increase the amount of time avail-
able for searching for food and vultures breeding
in plains are more dependent on thermals for fly-
ing (Hiraldo and Donazar 1989). Open areas with
steep slopes may also provide greater visibility of
predators, favor an easy nest access, and limit hu-
man disturbance (Titus and Mosher, 1981, Speiser
and Bosakowski 1988, Donazar et al. 1993, Moor-
man and Chapman 1996). In colony B, vultures
had fewer options to select nest sites because ma-
ture trees were only in protected areas. The larger
percentage of pastureland in the random plots in
colony A was due to the fact that a higher propor-
tion of habitat was devoted to cattle grazing on the
lower and middle thirds of the mountains.

The mean distance to the nearest-neighbor nest
in colony A (379 m) was within the range (175-
521 m) described by Hiraldo (1974), and signifi-
cantly less than distances between random plots in-
dicating a tendency for vultures to aggregate in this
forest. For colony B this distance is much greater
and did not differ from random plots. In forest B,
vultures are forced to nest in protected areas while
the apparently more natural conditions in forest A
allowed them to select optimal nesting sites.

Human disturbance has been suggested as the
main factor limiting Eurasian Black Vulture pro-
ductivity (Garzon 1973, Hiraldo 1977, Bermejo
1990). Vultures in our study seemed to avoid hu-
man activity by nesting in areas with less paved and
unpaved roads. There were no significant differ-
ences in road density between both forests, how-
ever, nesting areas in colony A contained a higher
number of roads. This higher tolerance to roads
was associated with the large number of old and
abandoned roads left after resin extraction ended
in the forest. The distance to the nearest village
did not seem to be a factor that influenced nest-
site selection, probably because the village was far
enough away to be tolerated by this species.

This study demonstrates differences in habitat
selection of Eurasian Black Vultures between areas

subjected to different forestry practices. In forest
A, where resin collection was carried out until
1985, traditional management afforded a relatively
natural structure of the forest with few modifica-
tions in size class distribution and tree density. Af-
ter the end of sylvicultural management, rapid oc-
cupation by vultures and homogeneous nest
distribution was possible in this forest. In contrast,
sylvicultural management for pine harvesting in
forest B resulted in a greater modification of the
size class distribution and tree density. Such forest
management has forced the vultures to nest in pro-
tected areas located in the upper third of the
mountains where there are mature trees resulting
in a less homogeneous distribution of this vulture
colony.

Acknowledgments

We thank to L.M. Carrascal and J. Moreno for useful
comments on earlier drafts. C. Cornelius, J. Torres, and
F. Jaksic improved the final manuscript. Fernando Go-
mez, Mario Herrera and Ana Dimas participated in the
held work. We also thank Stephie Michalsky for improv-
ing the English.

Literature Cited

Andrew, J.M. andJ.A. Mosher. 1982. Bald Eagle nest-site
selection and nesting habitat in Maryland. J. Wild 'l
Manage. 46:383-390.

Bermejo, C. 1990. Situacion actual del Buitre Negro ( Ae -
gypius monachus) en la provincia de Avila. I Congreso
Internacional sobre Aves Carroneras, Cuenca. Aeden-
at-CODA. ICONA, Madrid, Spain.

Bernis, F. 1966. El Buitre Negro ( Aegypius monachus) en
Iberia. Ardeola 12:45-99.

Bijleveld, M. 1974. Birds of prey in Europe. Macmillan
Press Ltd, London and Basingstore, U.K.

Blanco, J.C. andJ.L. Gonzalez. 1992. Libro rojo de los
vertebrados de Espana. MAPA-ICONA, Madrid, Spain.
Bosakowski, T., D.G. Smith and R. Speiser. 1992. Niche
overlap of two sympatric-nesting hawks Acdpiter spp.
in the New Jersey-New York Highlands. Ecography 15 -
358-372.

Brown, J.L. 1969. Territorial behavior and population
regulation in birds. Wilson. Bull. 81:293-329.
Donazar, J.A. 1993. Los buitres ibericos: biologia y con-
servacion, J.M, Reyero [Ed,], Madrid, Spain.

, F. Hiraldo and J. Bustamante. 1993. Factors in-
fluencing nest-site selection, breeding density and
breeding success in the Bearded Vulture ( Gypeatus bar-
batus).J. Appt Ecol. 30:504-514.

Ferrer, M. and J.A. Donazar. 1996. Density-dependent
fecundity by habitat heterogeneity in an increasing
population of Spanish Imperial Eagles. Ecology 77:69-
74.

Fuller, M.R. 1996. Forest raptor population trends in

Sept 1998

Habitat Selected by Eurasian Black Vultures

207

North America. Pages 167-208 in R.M. DeGraaf and
R.I. Miller [Eds.], Conservation of faunal diversity in
forested landscapes. Chapman-Hall, London, U.K.

Galera, R.M. and S. Martin. 1990. Regiones de proce-
dencia de Pinus sylvestris L. en Espana: su uso en las
repoblaciones. Ecologia (special issue) 1:527-540.

Garzon, J. 1973. Contribution al estudio del estatus, al-
imentation y protection de las falconiformes de Es-
pana central. Ardeola 19:280-330.

Gil, L., J. Gordo, R. Alia, G. Catalan andJ.A. Pardos.
1990. Pinus pinaster Aiton en el paisaje vegetal de la
Peninsula Iberica. Ecologia (special issue) 1:469-496.

GonzAlez, L.M. 1990. Censo de las poblaciones reprod-
uctoras de Aguila Imperial y Buitre Negro en Espana.
Quercus 58:16-22.

HlRALDO, F. 1974. Colonias de criay censo de los Buitres
Negros ( Aegypius monachus ) en Espana. Naturalia His-
panica 2.

. 1977. El Buitre Negro ( Aegypius monachus) en la

Peninsula Iberica. Ph.D. dissertation. Universidad de
Sevilla, Sevilla, Spain.

. 1983. Breeding biology of the Cinereous Vulture.

Pages 193-203 in S.R. Wilbur and J. A. Jackson [Eds.],
Vulture biology and management. Univ. Cal i fornia
Press, Berkeley, CA U.S.A.

andJ.A. Donazar. 1989. Foraging time in the Ci-
nereous Vulture Aegypius monachus : seasonal and local
variations and influence of weather. Bird Study 37:128-
132.

Houston, D.C. 1974. Breeding of the White-backed and
Rtippell’s Griffon vultures, Gyps africanus and G. rup-
pellii. Ibis 118:14—40.

Hubert, C. 1993. Nest-site habitat selected by Common
Buzzard ( Buteo buteo) in southeastern France. J! Raptor
Res. 27:102-105.

Hustler, K. and W.W. Howells. 1988. Breeding biology
of the Whitehead Vulture in Hwange National Park,
Zimbabwe. Ostrich 59:21-24.

Kadmon, R. 1993. Population dynamics consequences of

habitat heterogeneity: an experimental study. Ecology
74:826-825.

Mallada, L. 1892. Catalogo general de las especies fosiles
encontradas en Espana. Boletin de la Comision del
Mapa Geologico de Espana 18:1-253.

Moorman, C.E. and B.R. Chapman. 1996. Nest-site selec-
tion of Red-shouldered and Red-tailed hawks in a
managed forest. Wilson Bull. 108:357-368.

Mundy, J.P. 1982. The comparative biology of southern
African vultures. Vulture Study Group, Johannesburg,
South Africa.

Niemi, G.J. and J.M. Hanowski. 1997. Concluding re-
marks on raptor responses to forest management, a
Holarctic perspective./. Raptor Res. 31:191—196.

Prodon, R. 1976. Le substrate, facteur ecologique de la
vie en eau courante: observations et experiences sur
les larves de Microptera testacea et Cordulogaster annu-
latus. Ph.D. dissertation, Univ. Lyon, Lyon, France.

Rubio, J.L. and L.M. Carrascal. 1994. Habitat selection
and conservation of an endemic Spanish Lizard Al-
gyroides marchi (Reptilia. Lacertidae) . Biol. Conserv. 70
245-250.

Speiser, R.T. and T. Bosakowski. 1988. Nest-site prefer-
ences of Red-tailed Hawks in the Highlands of south-
eastern New York and northern New Jersey. J. Field
Ornithol. 59:361-368.

STATISTICA. 1993. Zoologiska Institute. Uppsala Univ,
Uppsala, Sweden.

Titus, K. andJ.A. Mosher. 1981. Nest-site habitat select-
ed by woodland hawks in the central Appalachians.
Auk 98:270-281.

Torres, J.A., P. Jordano andJ. Villasante. 1980. Estruc-
tura y dinamica temporal de una colonia de Buitre
Negro, Aegypius monachus , en Sierra Morena central
(Cordoba). Bol. Est. Cent. Ecol. 9:67-71.

Tucker, G.M. and M.F. Heath, Eds. 1994. Birds in Eu-
rope: their conservation status. BirdLife Internation-
al, Cambridge, U.K.

Received 12 October 1997; accepted 15 May 1998

J Raptor Res. 32(3):208-214
© 1998 The Raptor Research Foundation, Inc.

SELECTION OF SETTLEMENT AREAS BY JUVENILE
BONELLI’S EAGLE IN CATALONIA

Santi Manosa, Joan Real and Jordi Codina

Departament de Biologia Animal, Facultat de Biologia, Universitat de Barcelona, Avinguda Diagonal, 645,

08028 Barcelona, Catalonia, Spain

Abstract. — We observed Bonelli’s Eagles (Hieraaetus fasciatus) outside their breeding range in Catalonia
(northeastern Spain) to identify the main dispersal areas of juvenile eagles in the Central Catalan Basin.
Autumn counts were conducted to determine the size of the nonbreeding population in the main
dispersal area and to analyze factors leading to the selection of settlement areas by juveniles. The
permanent eagle population in the area was estimated at about 18 >l-yr-old eagles for an average density
of 0.90 eagles/100 km 2 . Number of eagles in the area <l-yr old fluctuated seasonally between 17-94
eagles with the number reaching a maximum in early autumn followed by a rapid decline during winter.
Juvenile settlement areas appeared to be selected based on gamebird and rabbit abundance, rather than
on landscape variables. As a consequence, game management measures appeared to be key in the
conservation of Bonelli’s Eagles.

KEYWORDS: Bonelli’s Eagle, Hieraaetus fasciatus; population density, conservation ; habitat selection', juvenile

dispersal', Spain.

Selection de zonas de asentamiento juvenil por el aguila perdicera en Cataluna

Resumen. — Describimos la localization y caracteristicas ambientales de la principal zona de dispersion
juvenil de las aguilas perdiceras Hieraaetus fasciatus en Cataluna (Espana). Se realizaron conteos otonales
en coche con objeto de evaluar el tamano de la poblacion de aguilas no reproductoras que utilizan esta
zona, y se analizan los factores que conducen a la selection de las areas de asentamiento dentro de esta
region. La poblacion de aguilas dispersantes esta formada por una poblacion permanente de unos 18
ejemplares de mas de 1 ano de edad, a una densidad de 0.90 aves/100 km 2 , y una poblacion estacion-
almente fluctuante de 17-94 aves de menos de un ano de edad, maxima a principios de otono y que
disminuye rapidamente a lo largo del invierno. Las aguilas seleccionan sus areas de asentamiento en
funcion de la abundancia de perdiz y conejo, mientras que las caracteristicas del paisaje aparecen como
un factor mucho menos importante. La gestion correcta de la caza menor constituye una pieza clave
para la supervivencia de las aguilas perdiceras jovenes en sus areas de asentamiento temporal.

[Traduction Autores]

In long-lived birds of prey, the acquisition of sex-
ual maturity is delayed for several years which
means that nonbreeding individuals, whose behav-
ior and ecology may differ from that of breeding
birds, form an important fraction of the total pop-
ulation (Newton 1979, Ferrer 1993a, Omland and
Hoffman 1996, Bustamante et al. 1997). In some
species, nonadult birds (those not in definitive
plumage) tend to settle in areas outside the breed-
ing range in juvenile dispersal or temporary settle-
ment areas where they stay for variable periods be-
fore moving to another settlement area or joining
breeding populations (Gonzalez et al. 1989, Ferrer
1993b, 1993c). Lack of appropriate setdement ar-
eas or reduction in their habitat quality may result
in a reduction in nonadult survival and have seri-

ous effects on population stability. Full understand-
ing of the habitat requirements of long-lived rap-
tors during the nonadult stages of their lives is,
therefore, necessary to design appropiate conser-
vation measures for threatened or declining pop-
ulations.

The Bonelli’s Eagle ( Hieraaetus fasciatus) is such
a long-lived raptor. It does not mature sexually un-
til 2-4 yr of age (Newton 1979, del Hoyo et al.
1994, Real and Manosa 1997) and it is an Endan-
gered Species whose populations have declined
across Europe (Rocamora 1994, Real and Manosa
1997). The current Spanish breeding population
of 675-751 pairs represents about 80% of the Eu-
ropean population (Arroyo 1991, Real et al. 1994,
Real et al. 1996). The nonbreeding population is

208

Sept 1998

Juvenile Bonelli’s Eagle Settlement Areas

209

SPAIN

T*

r-

(

i

•

1

•

•

•

•

•

•

•

•

•

•

•

•

•

r

r

Jr

r

i /

EmporcM^S

Delta del
Llobregat

10 km

Delta de
■ 7^ I’Ebre

Figure 1. Breeding range (shaded area) and dispersal
areas (dotted squares) of Bonelli’s Eagles in Catalonia
(northeastern Spain). The area depicted in Fig. 2 is out-
lined.

estimated at approximately 700 eagles (Real and
Manosa 1997). These nonbreeders are commonly
seen in areas unsuitable for nesting (Cramp and
Simmons 1980, Cugnasse and Cramm 1990, Real
et al. 1990, Cheylan et al. 1996), but little is known
concerning the exact locations and physical char-
acteristics of these settlement areas (Gonzalez et al.
1989, Ferrer 1990, Cugnasse and Cramm 1990,
Real and Manosa 1992, Real et al. 1994). The aims
of this study were to describe the locations and
general characteristics of juvenile dispersal areas of
Bonelli’s Eagles in Catalonia (northeastern Spain),
to estimate the size, age structure, and temporal
dynamics of the nonbreeding population in these
areas, and to describe factors leading to the selec-
tion of these settlement areas by nonadult Bonelli’s
Eagles.

Study Area and Methods

In Catalonia, the Bonelli’s Eagle breeds along the
coastal and pre-pyrenean mountain ranges (Fig. 1 ). A to-
tal of 70-80 pairs, or about 10% of the European breed-
ing population, breeds in this area (Real and Manosa
1992). From 1980-94, we compiled observations from as

Montclar 429 km 37.3

[Agramunt 1907 km

32.0

769 km

Sant

Plans de Gulm
SI6 197 kmj~]185 km

Granyena 1001 km

31.0 rJ I— I

Prelxana

Mas de 352 km [ 2 ^

Melons 1324 km , ,

~ 1 l_®^jArbeca 629 km

I®*® ICastelldans
I — 1 431 km

Surd
304 km

10 km

Figure 2. Areas surveyed during autumn eagle counts.
The number of km driven and the linear index of abun-
dance of eagles/ 1000 km are given for each sector.

many ornithologists as possible on Bonelli’s Eagles out-
side their breeding range. Results were plotted on 10 X

10 km Universal Transverse Mercator (UTM) squares
showing the presence or absence of nonbreeding eagles
in each square of the grid (Fig. 1).

We also conducted counts of Bonelli’s Eagles in the
autumn and winter in central Catalonia in the main non-
breeding area where eagles were more seen. Counts were
conducted for 3 yr from August-March 1991-94. They
were made to identify the areas used by eagles and to
estimate the number of eagles in the region. We selected

11 sectors (Fig. 2), each one exhibiting uniform land-
scape in terms of topography, land use, and vegetation
cover and patterns. Sectors were selected in order to rep-
resent all types of landscape patterns available to eagles.
Each sector was driven by car once a week at a speed not
exceeding 40 km/hr. Because of the large area to be cov-
ered on each sampling day, counts were conducted con-
tinuously from sunrise to sunset. In order to obtain an
equal sample for each sector at all times of the day, the
sampling sequence for sectors was changed each day, but
travel constraints prevented perfect randomization. To
analyze potential biases introduced by different sampling
schedules between sectors, counts were grouped into
four time intervals: early morning (before 0700 H),
morning (0700-1200 H), afternoon (1200-1700 H), and
evening (after 1700 H).

Transects were driven by teams of two observers with
each team including at least one trained observer and a
driver. Each team counted eagles in all of the sectors so
bias among observers was reduced. Each time a sector
was sampled, time of day, number of km driven, number
of Bonelli’s Eagles, and number of prey species (Red-
legged Partridge [Alectoris rufa ] and European rabbits
[Oryctolagus cuniculus ]) were recorded. When possible,
eagles were classified as juvenile (<1 yr), immature (1-2
yr), subadult (2-3 yr), or adult (>3 yr) according to
plumage criteria (Parellada 1986). Although some prac-
tice is needed to sex Bonelli’s Eagles in the field, the
most probable gender of the eagles was determined
based on size and color, females being larger and darker
than males (Parellada 1986). When possible, eagles were
classified as having full or empty crops. From these data.

210

Manosa et al.

Vol. 32, No. 3

linear indices of abundance for partridges, rabbits, and
each age class of eagles were computed for each sector.

Distance sampling (Buckland et al. 1993) was used to
estimate the abundance and density of eagles within the
sectors where they were observed. Each eagle observation
was plotted on a 1:50000 map and the perpendicular
distance to the transect was measured. An overall detec-
tion function was fitted to these data using the DIS-
TANCE software (Laake et al. 1993). Because most con-
tacts (93%) involved single birds, the model was fit by
considering each eagle as a single object. To obtain a
more robust estimate of the detection function, obser-
vations were truncated at 1.0 km after visual inspection
of the distance histogram (Buckland et al. 1993). The
detection function allowed the estimation of the Ef fective
Strip Width for the counts and an overall density estimate
was computed from sector density estimates weighted by
the area of each sector.

To determine habitat selection, each sector was divided
on 3 X 3 km UTM squares (900 ha or 9 km 2 each) on
which landscape variables (Table 1) were measured from
a CORINE Land Cover 1:250 000 habitat digital database
processed by means of Arc/Info software, and 1:50 000
military maps. The landscape of every sector was then
described using the same variables averaged over each
sector. The area variables were measured on the 900 ha
unit squares from each sector and could be considered
as proportions, so they were better analysed by convert-
ing them to log-ratios (Table 1) to make them indepen-
dent from one another (Robertson et al. 1993). Spear-
man rank correlation coefficients between eagle
abundance indices and partridge indices, rabbit indices,
habitat variables, and land cover log-ratios were comput-
ed to show the cause of spatial variation in eagle abun-
dance between sectors.

Results

Location and General Characteristics of the Ju-
venile Dispersal Areas. We compiled 81 indepen-
dent observations of Bonelli’s Eagles outside their
usual breeding range in Catalonia. Some occurred
along the coast, in wetland areas such as Aigua-
molls de l'Emporda, Delta del Llobregat, and Del-
ta de l'Ebre, but most observations came from the
Central Catalan Basin. Overall, 14 records were re-
ported in coastal areas (2.8 records per 100 km 2 )
and 6*7 in inland areas (3.19 records per 100 km 2 )
(Fig. 1 ) . The main juvenile dispersal area in Cata-
lonia was a belt of extensive dry farmland, bound-
ed by the Pyrenees, the litoral mountains, and the
irrigated lands of central Catalonia. This is a nearly
flat region, lacking cliffs and large forests, and is
mainly devoted to cereal crops and to a lesser ex-
tent olive, almond and vineyards.

Origin, Sex-ratio, Age-ratio and Abundance of
Birds. Within the 11 sectors, 355 autumn counts
were conducted along 7528 km. Eagles were ob-
served on only 322 counts from nine sectors over

Table 1. Names, definition, units, and source for habitat
variables recorded in juvenile Bonelli’s Eagle settlement
areas.

Name

Description, Units and Source

ROAD

Paved roads (km/9 km 2 ); CORINE

TRAC

Unpaved roads and tracks (km/9 km 2 );
CORINE

URBA

Urban, residential, and other developed ar-
eas (ha/9 km 2 ); CORINE

CROP

Arable land (ha/9 km 2 ); CORINE

BUSH

Low bush and dry grassland (ha/9 km 2 );
CORINE

WOOD

Woodland (ha/9 km 2 ) (m 2 ); CORINE

MIXE

Land occupied by agriculture with areas of
natural vegetation (ha/9 km 2 ); CORINE

ALTI

Average elevation (m) (Minimum elevation
+ Maximum elevation) /2; 1:50 000 map

RELI

Relief index (total number of 20 m isolines
crossed by the diagonals of the 3X3 km
square [A?]); 1:50 000 map

IKARUFA

Number of partridges/ 1000 km; Linear
counts

IKACUNI

Number of rabbits/1000 km; Linear counts

IKAFASC

Number of Bonelli’s Eagles/1000 km; Lin-
ear counts

LR-WC

Log-ratio woodland-crop: In (wood/
(crop + 1) + 1)

LR-BC

Log-ratio bush-crop: In (bush/ (crop+1) + 1)

LR-MC

Log-ratio mixed-crop: In (mixed/
(crop + 1) + 1)

LR-BW

Log-ratio bush-wood: In (bush/ (wood + 1 ) + 1 )

LR-MW

Log-ratio mixed-wood: In (mixed/
(wood+1) + 1)

LR-MB

Log-ratio woodland-crop: In (mixed/
(bush+1) + 1)

7039 km. In these counts, Bonelli’s Eagles were ob-
served on 196 occasions totalling 211 birds. One
eagle was observed on 183 (93%), two on 11 (6%),
and three on two (1%) occasions. Over 25% of
eagles had wing tags that had been attached in
1989-93 while they were nestlings in the nearby
Catalan breeding population. Forty (27%) of the
juvenile, six (18%) of the immature, and four
(28%) of the subadult eagles had wing tags. No
eagles tagged outside Catalonia (24 eagles in
southeastern France and 60 eagles in Murcia-Ala-
cant, southeastern Spain) were seen in our survey.

For eagles whose sexes were determined (24%),
32 were male (63%) and 19 female (37%; Binomial
test, P = 0.093). For wing-tagged birds of known
sex, six were male and three female.

Sept 1998

Juvenile Bonelli’s Eagle Settlement Areas

211

Table 2. Summary statistics for the variables recorded in sectors where Bonelli’s Eagles were detected ( N = 9) or
not detected ( N - 2). Mean ± SD (minimum-maximum)

Eagles Present Eagles Absent

(N =9) (N = 2)

ROAD

2.1

0.7

(0.9-3. 3)

2.3

+

2.5 (0.5-4. 1)

TRAC

23.3

6.1

(11.4-34)

26.3

1.0 (25.6-27.1)

URBA

2.3

2.3

(0-6.5)

2.0

+

1.5 (0.9-3. 0)

CROP

640

195

(274-898)

618

H-

151 (510-725)

BUSH

49

70

(0-223)

7

-f-

7 (2-11)

WOOD

23

28

(0-67)

45

■+■

2 (43-46)

MIXE

186

182

(213-567)

229

162 (114-343)

ALTI

402

65

(307-544)

676

+

18 (663-689)

RELI

35

+

10

(21-48)

50

-+-

10 (43-56)

IKARUFA

2582

1900 (182-5470)

1115

-h

362 (859-1372)

IKACUNI

78

90

(8-289)

39

+

1 (38-39)

IKAFASC

26

17

(3-55)

—

No significant difference in the linear index of
abundance of eagles was found between periods
when the six sectors visited in every period were
included (Kruskall-Wallis test, a? = 0.05, df = 2, P
= 0.98), or when the 10 sectors visited in both the
last two periods were compared (Kruskall-Wallis
test, x 2 = 0.006, df = 1, P = 0.94). There was also
no significant difference between the linear indi-
ces of abundance of nonjuvenile (immatures, sub-
adults, and adults) eagles between periods (Krus-
kall-Wallis test, a 2 = 1.377, df = 2, P = 0.50 or a 2
= 0.006, df = 1, P = 0.94). Therefore, periods
were considered together.

Of the 198 individuals we aged, 145 (73%) were
juvenile, 33 (17%) immature, 14 (7%) subadult,

O

Eu 50

O'

Q :

Z> 40

o

o

O 30

LL

O

>- 20
O

w 10

o

UJ

cc

0

K

x

\

PERPENDICULAR DISTANCE (m)

Figure 3. The number of eagles observed in relation to
the perpendicular distance to the transect line. The solid
line shows the shape of the adjusted detection model.

and six (3%) adult eagles. Two (1%) were only
identified as >1 yr of age. The observation rate for
nonjuveniles remained fairly stable around 0.75 ea-
gles/100 km during autumn and winter, while the
observation rate of yearlings reached a maximum
of 3.2 eagles/ 100 km in early autumn and declined
to 0.72 eagles/100 km in December-March (Fig.
4) . As a consequence, the age ratio of nonjuveniles
to juveniles increased from 0.37 in August to 1.00
after November. The seasonal variation on eagle
abundance correlated to a similar variation in Red-
legged Partridge abundance (r s = 0.73; P = 0.015,
N = 10).

The age ratio also varied between sectors (0.91
in Mas de Melons, N = 42 eagles; 0.67 in Almen-
ara, N = 40 eagles; 0.45 in Montclar, N= 16 eagles;
0.16 in Granyena, N = 29 eagles; and 0.07 in Agra-
munt, N = 59 eagles), but was not significantly
related to eagle abundance (r s = 0.68; P = 0.205,
N = 5) or prey abundance (r s = —0.50; P = 391,
N = 5). The linear index of abundance of young
eagles was positively correlated with the linear in-
dex of abundance of mature eagles (r s = 0.63; P =
0.04, N = 11).

A uniform function with two cosine adjustments
was found to be the best model to fit the perpen-
dicular distance data (Fig. 3). After truncation at
1.0 km, only 196 birds remained in the analysis,
and the Effective Strip Width was estimated at 416
m (95% C.I. = 370—469). For the sectors where
eagles were observed, we obtained an overall en-
counter rate of 2.7 eagles/100 km (95% C.I.= 1.9-
3.8), and an average density estimate of 3.2 eagles/

212

Manosa et al.

Vol. 32, No. 3

Figure 4. The number of Bonelli’s Eagles in the central
Catalonia dispersal area according to the time of the year
expressed as the number of individuals seen/ 100 km of
transect.

100 km 2 (95% C.I.= 2.3-4.5). Maximum eagle den-
sity was attained in late September (5.2 eagles/ 100
km 2 ), and minimum density inJanuary-March (1.5
eagles/ 100 km 2 ). The number of mature eagles re-
mained fairly stable at about 0.90 eagles/ 100 km 2 ,
while the number of young showed large variation
from late September (3.9 eagles/100 km 2 ) to late
winter (0.73 eagles/100 km 2 , Fig. 4). If these den-
sity estimates are extrapolated to the entire poten-
tial dispersal area in central Catalonia (dotted area
in Fig. 1 = 2000 km 2 ), an estimate of 18 nonjuven-
ile eagles and 17-94 yearlings would be obtained
for the entire area.

Selection of Settlement Areas, Eagle Behavior
and Habitat Use. Although the sequence of sam-
pling was changed on each sampling day, differ-
ences in the average sampling time between sec-
tors were detected (Kruskall-Wallis test, P< 0.000).
However, a two-way ANOVA of the square-root
transformed ( x + %) 1/2 index of eagle abundance
(Zar 1984) showed no significant relationship be-
tween eagle abundance indices and time of day
(F 3 3 23 = 0.759; P = 0.518) and a significant effect
of sector (F 10) 323 = 5.537; P< 0.000), so differences
in eagle abundance between sectors were not the
result of a different sampling pattern (Fig. 2). A
large variation was found in the habitat character-
istics of the sectors where eagles were observed
(Table 2) . The linear index of abundance of Bo-
nelli’s Eagles was positively correlated with the lin-
ear index of abundance of Red-legged Partridge (r s
= 0.74; P = 0.01, N = 11), and to the log-ratio of
bush-woodland (r s = 0.73; P = 0.01, N = 11). If

only the nine sectors where eagles were observed
were considered, the linear index of abundance of
Bonelli’s Eagles was found to be positively corre-
lated with the Red-legged Partridge linear index of
abundance (r s = 0.72; P = 0.03, N= 9), the rabbit
linear index of abundance (r s = 0.77; P = 0.02, N
= 9), the log-ratio mixed-crop (r s = 0.87; P =
0.002, N— 9), and the log-ratio bush-woodland (r s
= 0.71; P = 0.031, N = 9).

Within the settlement areas, eagles were ob-
served feeding on Red-legged Partridge on one oc-
casion, twice on European rabbits, and once on
Quarry Guineafowl ( Numida meleagris). Unsuccess-
ful attacks were observed once on feral pigeon ( Co-
lumba livia) , once on Wood Pigeon ( Columba pal-
umbus), once on pigeon ( Columba sp.), twice on
Red-legged Partridges, and once on an unidenti-
fied bird. Full crops were observed in 43% of the
eagles whose crop contents could be determined
( N = 120, 57% of all eagles). This was probably an
overestimate since empty crops were probably re-
corded as unnoticed more than full crops. For
those sectors in which crop contents were estimat-
ed for >5 eagles, those with a higher linear index
of eagle abundance had a larger proportion of ea-
gles with full crops (r s = 0.98; P ~ 0.02, N — 4).
In some sectors, eagles were more often observed
perching (Plans de Sio 100%, Almenara 69%, Mas
de Melons 50%) than in others (Montclar 25%,
Granyena 29%, Agramunt 31%), but this was not
related to eagle abundance (r s = —0.20; P = 0.70,
N = 6). Of 97 perches observed, trees were the
most frequendy used (52%) followed by large
rocks (19%), the ground (18%), large transport
power poles (6%), buildings (4%), and small dis-
tribution power poles (1%).

Discussion

Eagle abundance was related to partridge and
rabbit abundance in the dispersal area. In sectors
with larger eagle concentrations, eagles with full
crops were more frequendy encountered and these
areas had more bush, dry grassland or cropland
mixed with natural habitats relative to woodland or
homogeneous farmland. These open habitats were
likely the most suitable for eagles because of their
prey abundance and because their open habiat
maximized eagle foraging success (Bohall and Col-
lopy 1984, Janes 1985, Preston 1990). Therefore,
given the large variation in habitat pattern between
sectors where eagles were observed, we concluded
that nonadult Bonelli’s Eagles selected setdement

Sept 1998

Juvenile Bonelli’s Eagle Settlement Areas

213

areas mainly on food availability rather than their
topographic and landscape patterns. A similar hab-
itat selection strategy has been described in other
species of birds, in which food is a proximate factor
for habitat selection (Hilden 1965, Hutto 1985,
Gonzalez et al. 1989, Ferrer 1990, Gerrard et al.
1990, Heredia et al. 1991, Bustamante et al. 1997).

Although yearlings form the bulk of the non-
breeding Bonelli’s Eagle population in central Cat-
alonia, older birds form the stable fraction of the
population. Yearlings arrive in the region in late
summer, following independence (Real et al.
1989), and probably occupy only those areas left
available by older birds which die, move to other
areas, or are recruited into the breeding popula-
tion. In these dispersal areas, young eagles find suf-
ficient food and probably avoid competition with
breeders. The seasonal decline of the dispersal
population after October probably results from a
combination of factors including further innate
dispersal (Horn 1983), competitive exclusion
(Waser 1985), juvenile mortality (Real et al. 1996,
Real and Manosa 1997), or a decline in prey avail-
ability (Newton 1979, Ferrer 1993b, Brodeur et al.

1996) which was supported by the fact that the sea-
sonal decline of eagle abundance matched that of
Red-legged Partridges. A continuous turnover of
eagles with other dispersal areas, temporary re-
turns to natal areas, or movements to nearby
breeding areas may also occur as has been de-
scribed in the Spanish Imperial Eagle (Ferrer
1993c).

According to our wing-tagging information, the
Catalan breeding population is the main source of
Bonelli’s Eagles that come to the dispersal area in
central Catalonia. A higher number of males were
observed in the area than females. This suggested
that there may be female-biased juvenile dispersal
in this species, as is commonly found in raptors
(Clarke et al. 1997) or, alternatively, a sex-biased
mortality (Ferrer and Hiraldo 1992). Given the size
and demographic parameters (Real and Manosa

1997) of the breeding population in Catalonia, ev-
ery year 55 independent juveniles, 22 immatures,
and 9 subadults enter this eagle population. The
dispersal area in central Catalonia can only give
permanent refuge to about 35 of these birds, so
many must disperse farther away. Although the
mechanisms of dispersal and recruitment in Bo-
nelli’s Eagles are only partially understood, in our
opinion increasing the carrying capacity of dis-
persal areas adjacent to breeding areas would con-

tribute to reduced juvenile dispersal, decrease pre-
adult mortality, and improve recruitment rates in
nearby breeding areas (Gonzalez et al. 1989),
which would be particularly helpful to stop the de-
cline of isolated subpopulations or those found on
the edge of the species range. Sensible game man-
agement is an essential tool to achieve this objec-
tive. However, reduced dispersal will only be advan-
tageous if the main mortality factors for eagles in
the area (Real et al. 1996, Real and Manosa 1997)
are also eliminated.

Acknowledgments

We are indebted to Rodrigo del Amo, Joan Manel
Baques, Dani Diaz, Joan Estrada, Daniel Gonzalez, Ferran
Gonzalez-Prat, Gloria Laviga, Montse Lopez, Marc No-
guera, Vittorio Pedrocchi and Montserrat Tortosa for
contributing to field work. We also thank Dr. J. Loman,
Dr. H. Kallander, Dr. G.R. Bortolotti and Dr. M. McGrady
for their useful comments on earlier versions of the
manuscript. Josep Garriga, Cap del Servei d ’Agents Rur-
als del Departament d’Agricultura, Ramaderia i Pesca de
la Generalitat de Catalunya, provided access to the COR-
INE database, and Emili Ponsa helped in the extraction
of data from it. We especially thank Fundacio Miquel Tor-
res for continous financial support of Bonelli’s Eagle con-
servation in Catalonia. We also thank Caixa de Sabadell
for financial support of the fieldwork on which this paper
is based. The Fundacio Bosch i Gimpera cooperated m
the administration of part of the project. The first author
received a postdoctoral grant from Comissio Interdepar-
tamental de Recerca i Innovacio Tecnologica (CIRIT) de
la Generalitat de Catalunya (BPOST-9316).

Literature Cited

Arroyo, B. 1991. Resultados del censo nacional de aguila
perdicera. Quercus 70: 17.

Bohall, P.G. and M.W. Golloy. 1984. Seasonal abun-
dance, habitat use, and perch sites of four raptor spe-
cies in north-central Florida, J. Field Ornithol. 55:181-
189.

Brodeur, S., R. Decarie, D.M. Bird and M. Fuller. 1996.
Complete migration cycle of Golden Eagles breeding
in northern Quebec. Condor 98:293-299.

Buckland, S.T., D.R. Anderson, K.P. Burnham and J.L
Laake. 1993. Distance sampling. Chapman and Hall,
London, U.K.

Bustamante, J., J.A. Donazar, F. Hiraldo, O. Ceballos
and A. Travaini. 1997. Differential habitat selection
by immature and adult Grey Eagle-Buzzards Geranoae-
tus melanoleucus. Ibis 139:322-330.

Cheylan, G., A. Ravairol, J.M. Cugnasse, J.M. Billet
and C. Joulot. 1996. Dispersion des aigles de Bonelli
Hieraaetus fasciatus juveniles bagues en France. Alauda
64:413-419.

Clarke, A.L., B.E. Szether and E. Roskaft. 1997. Sex bi-
ases in avian dispersal: a reappraisal. Oikos 79:429-
438.

214

Manosa et al.

Vol. 32, No. 3

Cugnasse, J.M. and P. Cramm. 1990. L’erratisme de
l’aigle de Bonelli Hieraaetus fasciatus en France. Alau-
da 58:59-66.

Craig, E.H., T.H. Craig and L.R. Powers. 1986. Habitat
use by wintering Golden Eagles and Rough-legged
Hawks in southern Idaho./. Raptor Res. 20:69-71.

Cramp, S. and K.E.L. Simmons [Eds.]. 1980. The birds of
the western Palearctic. Vol. 2. Hawks to Bustards. Ox-
ford Uniy. Press, Oxford, U.K.

del Hoyo, J., A. Elliot and J. Sargatal [Eds.]. 1994.
Handbook of the birds of the world. Vol. 2. New
World Vultures to Guineafowl. Lynx Edicions, Barce-
lona, Spain.

Ferrer, M. 1990. Dispersion juvenil de la poblacion de
aguilas imperiales del Parque Nacional de Donana.
Ph.D. dissertation, Universidad de Sevilla, Sevilla,
Spain.

. 1993a. El aguila imperial. Eurofauna, Madrid,

Spain.

. 1993b. Reduction in hunting success and settle-
ment strategies in young Spanish Imperial Eagles.
Anim. Behav. 45:406-408.

. 1993c. Juvenile dispersal behavior and natal phil-

opatry of a long-lived raptor, the Spanish Imperial Ea-
gle Aquila adalberti. Ibis 135:132-138.

and F. Hiraldo. 1992. Man-induced sex-biased

mortality in the Spanish Imperial Eagle. Biol. Con-
serv. 60:57-60.

Gerrard, J.M., G.R. Bortolotti, E.H. Dzus, P.N. Ger-
rard and D.W.A. Whitefield. 1990. Boat census of
Bald Eagles during the breeding season. Wilson Bull.
102:720-726.

GonzAlez, L.M., B. Heredia, J.L. Gonzalez and J.C.
Alonso. 1989. Juvenile dispersal of Spanish Imperial
Eagles. / Field Ornithol. 60:369-379.

Heredia, B., J.C. Alonso and F. Hiraldo. 1991. Space
and habitat use by Red Kites Milvus milvus during win-
ter in the Guadalquivir marshes: a comparison be-
tween resident and wintering populations. Ibis 133:
374-381.

Hilden, O. 1965. Habitat selection in birds: a review.
Ann. Zool. Fenn. 2:53-75.

Horn, S.H. 1983. Some theories about dispersal. Pages
55-62 in R.S. Swingland and P.J. Greenwood [Eds.],
The ecology of animal movement. Oxford Univ. Press,
Oxford, U.K.

Hutto, R.L. 1985. Habitat selection by non-breeding mi-
gratory land birds. Pages 455-476 in M.L. Cody [Ed.],
Habitat selection in birds. Academic Press, Orlando,
FL U.S.A.

Janes, S.W. 1985. Habitat selection in raptorial birds.
Pages 159-188 in M.L. Cody [Ed.], Habitat selection
in birds. Academic Press, Orlando, FL U.S.A.

Laake, J.L., S.T. Buckland, D.R. Anderson and K.P.
Burnham. 1993. Distance user’s guide, Version 2.0.

Colorado Coop. Fish Wildl. Res. Unit, Colorado State
Univ., Fort Collins, CO U.S.A.

Newton, I. 1979. Population ecology of raptors. T. & A.D.
Poyser, Berkhamsted, U.K.

Omland, K.S. and S.W. Hoffman. 1996. Seasonal, diet,
and spatial dispersion patterns of Golden Eagles mi-
gration in southern Montana. Condor 98:633-636.

Parellada, X. 1986. Variacio del plomatge i identificacio
de l’aliga cuabarrada Hieraaetus fasciatus fasciatus.
Pages 87-96 in Rapinyaires Mediterranis, II. Centre
de Recerca i Proteccio de Rapinyaires, Barcelona,
Spain.

Preston, C.R. 1990. Distribution of raptor foraging in
relation to prey biomass and habitat structure. Condor
92:107-112.

Real, J. and S. Manosa. 1992. La Conservacio de l’aliga
perdiguera a Catalunya. Univ. Barcelona, Barcelona,
Spain.

and . 1997. Demography and conservation

of western European Bonelli’s Eagle Hieraaetus fascia-
tus populations. Biol. Conserv. 79:59-66.

, S. Manosa, J. Bertran, P. Romero, A. Sorolla,

J. FernAndez and L. Balaguer. 1989. El proces
d’independitzacio en els joves d’aliga perdiguera, Hi-
eraaetus fasciatus, al Parc Natural de Sant Lloreny del
Munt i l’Obac. Pages 43-47 in I Trobada d’estudiosos
de Sant Lloreny del Munt i l’Obac. Diputacio de Bar-
celona, Servei de Parcs Naturals, Barcelona, Spain.

, S. Manosa, R. del Amo, J.A. Sanchez Zapata,

M.A. Sanchez, D. Carmona and J.E. Martinez. 1990
La regresion del aguila perdicera: una cuestion de
demografia. Quercus 70:6-12-

, S. Manosa and J. Codina. 1996. Estatus, demo-
grafia y conservacion del aguila perdicera ( Hieraaetus
fasciatus) en el Mediterraneo. Pages 83-89 in J. Mun-
taner andj. Mayol [Eds.], Biologia y conservacion de
las rapaces mediterraneas, 1994. Monograflas, No. 4
SEO, Madrid, Spain.

S. Manosa, J. Codina and R. del Amo. 1994. Es-

tado de varias poblaciones de aguila perdicera en Eu-
ropa occidental durante 1993. Quercus 98:8-11.

Robertson, P.A., M.I.A. Woodburn, W. Neutel and C.E
Bealey. 1993. Effects of land use on breeding pheas-
ant density. /. Appl. Ecol. 30:465-477.

Rocamora, G. 1994. Bonelli’s Eagle Hieraaetus fasciatus.
Pages 184-185 in G.M. Tucker and M.F. Heath [Eds.],
Birds in Europe, their conservation status. Birdlife
Conserv. Ser. No. 3. BirdLife International, Cam-
bridge, U.K,

Waser, P.M. 1985. Does competition drive dispersal? Ecol-
ogy 66:1171-1175.

Zar, J.H. 1984. Biostatistical analysis. Prentice Hall,
Princeton, NJ U.S.A.

Received 6 June 1997; accepted 20 May 1998

J. Raptor Res. 32(3):215-220
© 1998 The Raptor Research Foundation, Inc.

WINTER FORAGING ECOLOGY OF BALD EAGLES ON A
REGULATED RIVER IN SOUTHWEST IDAHO

Gregorys. Kaltenecker

Department of Biology, Raptor Research Center, Boise State University, Boise, ID 83725 U.S.A.

Karen Steenhof

USGS Forest and Rangeland Ecosystem Science Center, Snake River Field Station, 970 Lusk St., Boise, ID 83706 U.S.A

Marc J. Bechard and James C. Munger

Department of Biology, Raptor Research Center, Boise State University, Boise, ID 83725 U.S.A.

Abstract. — We studied Bald Eagle foraging ecology on the South Fork Boise River, Idaho, during the
winters of 1990-92. We compared habitat variables at 29 foraging sites, 94 perch sites, and 131 random
sites. Habitat variables included river habitat (pool, riffle, run), distance to the nearest change in river
habitat, distance to nearest available perch, number and species of surrounding perches, and average
river depth and flow. Eagles foraged more at pools than expected, and closer (<15 m) to changes in
river habitat than expected. Where eagles foraged at riffles, those riffles were slower than riffles where
they perched or riffles that were available at random. Where eagles foraged at runs, those runs were
shallower than runs at either perch or random sites. Eagles perched less at riffles and more at sites
where trees were available than expected. Changes in river habitat represent habitat edges where river
depth and flow change, making fish more vulnerable to eagle predation. Fish are more susceptible to
predation at shallower river depths and slower flows. Slower river flows may be related to decreased
surface turbulence, which also increases vulnerability of fish to aerial predation.

Key Words: Bald Eagle, Haliaeetus leucocephalus; wintering ecology, foraging ecology, dams; rivers; Idaho.

Ecologia del forrajeo de invierno de aguilas Calvas en un rio regulado del suroeste de Idaho

Resumen. — Estudiamos la ecologia de forrajeo de aguilas Calvas en el Rio South Fork Boise en Idaho,
durante los inviernos de 1990-1992. Comparamos las variables de habitat en 29 sitios de forrajeo, 93
perchas y 131 sitios al azar. Las variables de habitat incluyeron habitats del rio (pozos, escorrentias,
otros), la distancia al cambio de habitat mas cercano del rio, la distancia mas cercana a una percha
disponible, el numero y especies de perchas alrededor y el promedio de profundidad y escorrentia. Las
aguilas forrajearon mas en los pozos de lo esperado, y mas cerca (15 m) a los cambios de habitat en el
rio de lo esperado. En los sitios poco profundos en donde las aguilas forrajearon, estos fueron mas
lentos que aquellos en donde las aguilas utilizaron perchas disponibles al azar. En los sitios en donde
las aguilas se percharon en escorrentias, estas fueron menos profundas que las de las perchas o sitios
al azar. Las aguilas utilizaron menos perchas en sitios de escorrentias y mas en sitios en donde los
arboles estaban mas disponibles de lo esperado. Los cambios de habitat en el rio estaban representados
por las orillas en donde la profundidad y el flujo variaban, haciendo a los peces mas vulnerables a la
depredacion de las aguilas. Los peces son mas susceptibles a la depredacion en los niveles menos
profundos y en escorrentias mas lentas los cuales pueden estar relacionados con la disminucion de la
turbulencia en la superficie, lo que aumenta la vulnerabilidad de los peces a la depredacion aerea.

[Traduccion de Cesar Marquez]

The winter diets of Bald Eagles (Haliaeetus leu-
cocephalus) differ depending on locale, habitat,
weather conditions, and prey availability, but fish
are selected most often when available (Stalmaster
1987). Bald Eagles may concentrate during winter
near dams where open water and fish are readily

available (Steenhof et al. 1980). Dams can keep
downstream areas from freezing and can provide
a reliable source of fish that have been killed or
stunned while passing through dam turbines
(Steenhof 1978, Brown et al. 1989). In rivers, ben-
thic-feeding fish are the most commonly taken

215

216 Kaltenecker et al. Vol. 32, No. 3

Figure 1. South Fork Boise River, Idaho, studied during winters 1990-92; Anderson Ranch Dam to Trail Creek.

prey (Dunstan and Harper 1975, McEwan and
Hirth 1980, Todd et al. 1982, Haywood and
Ohmart 1986, Hunt et al. 1992), but eagles may
also take rainbow trout ( Oncorhynchus mykiss ) if
available (Brown et al. 1989, Spahr 1990, Brown
1993). The Boise River, a tributary of the Snake
River, is a major drainage containing free-flowing
and regulated river reaches and three reservoirs.
Mammal carrion and fish are the main prey of
Bald Eagles wintering in the Boise River System
(Kaltenecker and Bechard 1995, Kaltenecker
1997). We studied foraging ecology of Bald Eagles
on the South Fork Boise River during the winters
of 1990-92, and present results which identify and
describe foraging and perching habitat.

Study Area and Methods

The South Fork Boise River flows from the Sawtooth
Mountain Range in southwestern Idaho and drains an
area of approximately 1568 km 2 (Gebhards 1964). An-
derson Ranch Dam, a U.S. Bureau of Reclamation power-
generating and irrigation facility, is 19 km downstream
from the town of Pine. Our study area included approx-
imately 20 km of river located between Anderson Ranch
Dam and Trail Creek and was easily accessible by vehicle
along U.S. Forest Service road #113 (Fig. 1). Both the
river and Bald Eagles perching along it could be seen
from our observation points on the road. The South Fork
flowed through a steep-sided valley dominated by shrub-
steppe vegetation consisting of sagebrush ( Artemisia tri-
dentata) , bitterbrush ( Purshia tridentata) , native perennial

grasses ( Poa secunda, Pseudoregnaria spicata, Aristida longi-
seta ), and exotic annuals cheatgrass ( Bromus tectorum ) and
medusahead rye ( Taeniatherum caput-medusae ). Cotton-
wood/willow riparian vegetation ( Populus trichocarpa,
Salix spp., Betula spp., Alnus spp.) dominated the river
bottom and other riparian areas. Some mixed-conifer
stands ( Pinus ponderosa, Pseudotsuga menziesii) were pres-
ent on north-facing slopes. Elevations ranged from 1100-
1220 m, and temperature extremes varied from — 30°C
to 16°C during December-March. River flows were reg-
ulated by Anderson Ranch Dam and were maintained at
the standard winter minimum flow (approx. 91 m s /s)
throughout both winters. Drought conditions prevailed
during both years of the study.

We conducted a total of 224 hr of foraging observa-
tions on 28 d between 15 December-1 March (12 days
during the first winter, and 16 days during the second
winter). Observations were conducted by one person
from a vehicle using 8 X 30 binoculars and 45 X spotting
scope. We began observing at dawn, and continued
throughout the day until all eagles left the river or re-
turned to night roosts. Observation points were selected
so that perched or flying eagles and the river were in full
view between 150-500 m away from the observer. We re-
corded foraging activity as successful or unsuccessful at-
tempts at fish prey. A foraging site was defined as the
exact point in the river where a foraging attempt was
made. Foraging attempts were initiated either from the
wing or nearby perch locations. We identified fish species
taken by eagles from observation of prey captures or
feeding, or by analysis of prey remains collected from
feeding sites immediately after eagles departed. Remains
used to identify fish species included scales, opercular
bones, and mandibles. During observations, we also re-

Sept 1998

Winter Foraging of Bald Eagles

217

corded all perches used by eagles within 75 m of the
river. Perch sites were defined as any tree, cliff, or rock
outcrop where we observed eagles perching. Perches
from which prey strikes were initiated were included in
the sample of perch sites.

Once a foraging site had been identified, we returned
to It during late February or March of the same winter
and measured surrounding habitat. Because river flows
were regulated at a constant level throughout both win-
ters, we assumed that surrounding habitat did not
change significantly between observation of prey cap-
tures and measurement of habitat. We used a line-tran-
sect method modified from Bovee (1982) and Platts et
al. (1983) to measure physical habitat parameters asso-
ciated with each foraging site. At each foraging site, we
recorded predominant river habitat (three categories:
pool, riffle, run), distance to nearest change in river hab-
itat, and distance to the nearest perch. Furthermore, a
transect was established across the river perpendicular to
flow, and river depth, stream flow, and bottom substrate
were recorded at five equidistant points (verticals) along
the transect. At each vertical, we measured depth and
flow using a Price AA flowmeter. At each end of the tran-
sect, we recorded the number of surrounding perches
and predominant species of tree within a 75 m arc. At all
eagle perch sites located within 75 m of the river, we
measured surrounding habitat as described above for for-
aging sites. Lasdy, we selected an additional sample of
sites by converting random numbers into distances (m)
downstream from Anderson Ranch Dam. After locating
random sites using a metric tape, we measured surround-
ing habitat similar to foraging and perch sites.

Habitat variables associated with foraging, perch, and
random sites were analyzed using logistic regression (LO-
GIST procedure, SAS 1990), which determines the ef-
fects of several different independent variables on a sin-
gle dependent variable (Harrell 1986, Trexler and Travis
1993). The dependent variable in our analyses was site
type (three categories: foraging, perch, or random). We
conducted three separate logistic regression analyses,
comparing foraging to random sites, foraging to perch
sites, and perch to random sites. We used stepwise logistic
regression, with the significance level to enter the model
and to remain in the model set at 0.15. Independent
variables entered into the analyses were river habitat
(pool, riffle, run), distance to nearest change in river
habitat, presence of available perches, distance to nearest
perch, and number of surrounding perches. River habi-
tat is a nominal variable, and was therefore transformed
into a set of 0,1 variables that were used in the analysis.
To prevent over specification of the model, we consid-
ered the variable “run” as the base state and did not
include “run” in the model; thus, we determined if be-
ing a pool or riffle increased the chance of, for example,
being a foraging site. Distance to the nearest perch was
placed in a category of 1 to 6, with 1 = 0-10 m, 2 = 11-
25 m, 3 = 26-50 m, 4 = 51-75 m, and 5 — >75 m.
Number of perches was placed in a category of 0-5, with
0 = no surrounding perches, 1 = less than 5, 2 = 6-10,
3 = 10-20, and 4 = >20 perches available within 75 m.
We further explored relationships of variables contrib-
uting significantly to logistic regression models using Chi-
square goodness-of-fit tests (Zar 1984; FREQ procedure,

SAS 1990). We calculated average stream flow and depth
for each transect, and compared means using analysis of
variance (AN OVA; GLM procedure, SAS 1990) to deter-
mine if flow or depth characteristics varied significantly
between foraging, perch, or random sites by river habitat
type.

Results

Counts of Bald Eagles from 18 aerial surveys
conducted every two weeks during both winters of
our study ranged from 0-17 (x = 7.8) eagles. We
observed 31 attempted prey captures of fish (17
successful) at 29 different sites, identified 94 eagle
perch sites, and collected habitat data from 131
random sites. Fish species taken by eagles included
largescale suckers ( Catostomus macrocheilus, N =
10), mountain whitefish ( Prosapium williamsoni, N
= 4) , and rainbow trout (N — 3) .

Due to low sample sizes, and because stream
flows were similar during both winters, we lumped
habitat data collected during both years of the
study for analyses. Significant differences existed
between foraging and random sites, foraging and
perch sites, and perch and random sites. Foraging
sites differed from random sites with regard to riv-
er habitat and distance to the nearest change in
river habitat (Table 1 ) . Further analysis using Chi-
square goodness-of-fit tests revealed that eagles for-
aged at pools more than expected (number ex-
pected - 3.63, actual number = 8, X 2 — 6.1, P —
0.013, df = 1), and that foraging sites were closer
(<15 m) to changes in river habitat more than ex-
pected (number expected = 5.6, actual number —
12, x 2 = 9.5, P = 0.002, df = 1). Where eagles
foraged at runs, those runs were shallower than
runs available at random (Table 2) . Where eagles
foraged at riffles, those riffles had slower stream
flows than riffles available at random.

Perch sites were similar to foraging sites, but dif-
fered with regard to distance to the nearest change
in river habitat and the number of surrounding
perches (Table 1). As with foraging sites compared
to random, foraging sites were closer to changes in
river habitat (x 2 = 9.5, P — 0.002, df = 1) than
perch sites. Foraging sites also had fewer surround-
ing perches than perch sites. Perch sites differed
from random with regard to the presence, number
of, and distance to surrounding perches. Not all
random sites had potential eagle perches available
within 75 m. No differences existed between perch
and random sites with regard to either river depth
or flows (Table 2).

218

Kaltenecker et al.

Vol. 32, No. 3

Table 1. Results from three separate stepwise logistic regression procedures comparing habitat between foraging
and random, foraging and perching, and perching and random sites for Bald Eagles. Sample sizes are in parentheses.

Parameter

Estimate

Standard

Error

Wald

Chi-Square

P Value

Odds

Ratio

FORAGING (29) vs. RANDOM (131) ab

Intercept

-1.682

0.583

8.331

0.004

Pool

1.423

0.549

6.768

0.009 e

4.l75 f

Distance to Habitat Change

-0.018

0.007

5.736

0.0l7 e

0.983

Perch

0.967

0.548

3.115

0.078

2.630

FORAGING (29) vs. PERCHING (94) c

Intercept

4.751

1.584

8.999

0.003

Distance to Habitat Change

-0.015

0.007

4.514

0.034 e

0.985

Perch

-1.987

1.234

2.591

0.108

0.137

Number of Surrounding Perches

-2.463

0.631

15.194

0.0001 e

0.085

PERCHING (94) vs. RANDOM (131) d

Intercept

-8.833

1.469

36.15

0.0001

Pool

0.955

0.606

2.486

0.115

2.599

Distance to Nearest Perch

0.801

0.334

5.731

0.017 e

2.227

Perch

2.978

1.031

8.332

0.004 e

19.65

Number of Surrounding Perches

3.812

0.643

35.15

0.0001 e

45.24

a The first of the two listed site types was the modeled state, thus the sign (+ or -) of the parameter estimate indicates whether an
increase in the independent variable was associated with a higher (if +) or lower (if — ) probability of being a site of the modeled
state.

b Model statistics: overall G = 19.74 with 3 df (P = 0.0002); concordance/discordance: 73.6%/22.8%. Concordance is determined as
follows. All possible pairings of foraging and random sites are created. A pair of sites is defined as concordant if the foraging site of
that pair is also the site predicted by the logistic regression model (based on predictor variables, e.g., habitat) to be the site more
likely to be the foraging site. A pair is discordant if the model predicts (incorrectly) that the random site is more likely to be the
foraging site. Percents of the total number of pairs that are concordant or discordant are presented. Ties are not presented.
c Model statistics: overall G = 45.22 with 3 df (P = 0.0001); concordance/discordance: 85.6%/ 12.2%.
d Model statistics: overall G = 153.86 with 4 df (P = 0.0001); concordance/discordance: 85.1%/3.7%.

e Denotes variables contributing significantly to stepwise logistic regression models at P < 0.05 level (analyses performed using SAS,
procedure logist).

f A pool site has an approximately four-fold greater probability of being a feeding site than does a non-pool site.

Discussion

Fish species captured by Bald Eagles can influ-
ence foraging behavior and foraging site selection
in rivers. Many authors have discussed increased
vulnerability of bottom-feeding fish to avian pred-
ators (Swenson 1979, Todd et al. 1982, Haywood
and Ohmart 1986). In our study, eagles took more
benthic-dwelling than pelagic fish. River habitat
also may influence Bald Eagle foraging site selec-
tion. In our study, eagles foraged more from pools
than other river habitats. Hunt et al. (1992) re-
ported that eagles foraged more from pools than
other habitats in California’s Pit River. In Arizona,
nesting Bald Eagles also foraged most at pools
(Haywood and Ohmart 1986). On the Boise River,
Spahr (1990) reported that eagles were observed
at pools more than expected. During winter, most
fish species, especially salmonids, seek pools or

other areas of low stream velocity to maintain po-
sition with minimal energy expenditure (Allen
1969, Cunjak and Power 1986, 1987, Hillman et al.
1987). Because of winter temperatures and low
stream flows on the South Fork Boise River, it is
likely that during our study, pools were areas of
high fish abundance.

Changes in river habitat, especially from pool to
riffle or pool to run, usually indicate decreasing
water depth and a change in stream flow, both
found to be important parameters at foraging sites
during our study. We found that eagles foraged at
sites which were closer to river habitat changes
than were random sites. This suggests that changes
in river habitat may be important to foraging Bald
Eagles as habitat edges. The edges of habitats con-
taining higher prey densities may represent areas
where fish become vulnerable to predation due to

Sept 1998

Winter Foraging of Bald Eagles

219

Table 2. Results from ANOVA procedures on average depth and average velocity by river habitat type. P-values are
from ANOVA, means with different letters are different by Tukey’s Studentized Range (HSD) tests at P < 0.05 level.

Mean Depth (m)

Mean Flow (m/s)

Habitat Type

N

X

SE

N

X

SE

POOLS

Foraging

8

0.80a

0.10

8

0.32a

0.05

Perch

13

0.84a

0.10

13

0.33a

0.05

Random
F, (df), P Value

12

0.92a

0.33, (2), 0.72

0.10

12

0.32a

0.10, (2), 0.90

0.05

RIFFLES

Foraging

8

0.41a

0.04

8

0.62a

0.05

Perch

16

0.40a

0.04

16

0.80b

0.05

Random
F, (df), P Value

38

0.41a

0.02, (2), 0.98

0.04

38

0.71b

3.54, (2), 0.04

0.05

RUNS

Foraging

13

0.47a

0.03

13

0.57a

0.03

Perch

64

0.57ab

0.03

64

0.54a

0,03

Random
F, (df), P Value

81

0.61b

4.76, (2), 0.01

0.03

81

0.50a

2.96, (2), 0.06

0.03

decreasing water depth. Haywood and Ohmart
(1986) reported that eagles foraged from pools
bounded by shallows or riffles where benthic feed-
ing fish were vulnerable to predation. Hunt et al.
(1992) also showed that eagles foraged from shal-
low areas of pools. Wintering Bald Eagles in Grand
Canyon, Arizona, foraged more in creeks (i.e.
smaller, shallower streams) than rivers (Brown
1993) . We found that eagles foraged at runs which
were shallower than those available at random.
Though water depth influences fish vulnerability,
foraging site selection by Bald Eagles also may be
influenced by stream flow. We found that eagles
foraged at riffles with lower stream flows. Water
turbulence is related to stream flow; the faster the
flow, the greater the turbulence. Low surface tur-
bulence may be an important component of Bald
Eagle foraging sites (Hunt et al. 1992), enabling
eagles to better detect fish.

We concur with other authors that physical hab-
itat parameters of rivers or streams are important
to Bald Eagle foraging site selection and foraging
success. Eagles commonly took prey from habitats
where fish were likely most abundant, but concen-
trated foraging efforts at the edges of those habi-
tats where water was shallower and slower, suggest-
ing that vulnerability of prey also may be
important.

Acknowledgments

Funding and vehicles for this study were provided by
the U.S. Forest Service, Boise Nadonal Forest. Housing
was arranged by Idaho Department of Fish and Game
and the U.S. Bureau of Reclamation. Necessary equip-
ment was provided by Boise State University, Idaho De-
partment of Fish and Game, and the U.S. Fish and Wild-
life Service. We thank L.L. Donohoo, U.S. Forest Service,
for logistic support, advice on design, and help in the
field. This manuscript benefitted from reviews by L.L.
Donohoo, M.N. Kochert, G. Bortolotti, R. Rnight, and
J.H. Kaltenecker. Maps were prepared by E. Holzer and
M. Spencer. Invaluable help in the field was provided by
R. Moore, J. Hilty, J. Weaver, R. Garwood, L. Spain, and
B. Zoellick.

Literature Cited

Allen, K.R. 1969. Limitations on production in salmonid
populations in streams. Pages 3-18 in T.G. Northgate
[Ed.], Symposium on salmon and trout in streams.
Univ. British Columbia, Vancouver Canada.

Bovee, K.D. 1982. A guide to stream habitat analysis us-
ing the Instream Flow Incremental Methodology. In-
stream flow information paper 12. U.S. Fish and Wild.
Serv. FWS/OBS-82/86.

Brown, B.T., R. Mesta, L.E. Stevens and J. Weisheit.
1989. Changes in winter distribution of Bald Eagles
along the Colorado River in Grand Canyon, Arizona.
J. Raptor Res. 23:11 0-1 1 3.

Brown, B.T. 1993. Winter foraging ecology of Bald Eagles
in Arizona. Condor 95:132-138.

Cunjak, R.A. AND G. Power. 1986. Winter habitat utili-

220

Kaltenecker et al.

Vol. 32, No. 3

zation by stream resident brook trout (Salvelinus fon-
tinalis) and brown trout ( Salmo trutta). Can J. Fish.
Aquat. Sci. 43:1970—1981.

Cunjak, R.A. and G. Power. 1987. Cover use by stream-
resident trout in winter: a field experiment. N. Am. J.
of Fisheries Manage. 7:539-544.

Dunstan, T.C. and J.F. Harper. 1975. Food habits of Bald
Eagles in North-Central Minnesota, j. Wild. Manage.
39:140-143.

Gebhards, S.V. 1964. Federal aid to fish restoration. Job
performance report, Project No. F-51-R-1. Idaho
Dept Fish and Game, Boise, ID U.S.A.

Harrell, F.E. 1986. The LOGIST procedure. Sugi Sup-
plemental Library Guide, Version 5. SAS Institute,
Cary, NC U.S.A.

Haywood, D.D. and R.D. Ohmart. 1986. Utilization of
benthic-feeding fish by inland breeding Bald Eagles.
Condor 88:35-42.

Hillman, T.W.,J.S. Griffith and W.S. Platts. 1987. Sum-
mer and winter habitat selection by juvenile chinook
salmon in a highly sedimented Idaho stream. Trans.
Am. Fish. Soc. 116:185-195. '

Hunt, W.G., B.S. Johnson and R.E. Jackman. 1992. Car-
rying capacity for Bald Eagles wintering along a north-
western river. J. Raptor Res. 26:49-60.

Kaltenecker, G.S. and M.J. Bechard. 1995. Bald Eagle
wintering habitat study, upper Boise River Drainage,
Idaho. Raptor Res. Ser. No. 9. Boise State Univ., Boise,
ID U.S.A.

Kaltenecker, G.S. 1997. Winter ecology of Bald Eagles
in the upper Boise River Drainage, Idaho. M.S. thesis,
Boise State Univ., Boise, ID U.S.A.

McEwan, L.C. and D.H. Hirth. 1980. Food habits of the

Bald Eagle in north-central Florida. Condor 82:229-
231.

Platts, W.S., W.F. Megahan and G.W. Minspiall. 1983.
Methods for evaluating stream, riparian, and biotic
conditions. U.S. Dept. Ag., Forest Service, Gen. Tech.
Report INT-138. Intermountain Forest and Range Ex-
periment Station, Ogden, UT U.S.A.

SAS, 1990. SAS/STAT user’s guide, version 6, fourth edi-
tion. SAS institute, Cary, NC U.S.A.

Spahr, R. 1990. Factors affecting the distribution of Bald
Eagles and effects of human activity on Bald Eagles
wintering along the Boise River. M.S. thesis, Boise
State Univ., Boise, ID U.S.A.

Stalmaster, M.V. 1987. The Bald Eagle. Universe Books,
New York, NY U.S.A.

Steenhof, K. 1978. Management of wintering Bald Ea-
gles. U.S. Fish and Wild. Serv. Rep., FWS/OBS^78/79,
Harper’s Ferry, WV U.S.A.

, S.S. Berlinger and L.H. Fredrickson. 1980.

Habitat use by wintering Bald Eagles in South Dakota.
J. Wild. Manage. 44:798-805.

SWENSON, J.E. 1979. The relationship between prey spe-
cies ecology and dive success in Ospreys. Auk 96:408-
412.

Todd, C.S., L.S. Young, R.B. Owen, Jr. and FJ. Gram-
lich. 1982. Food habits of Bald Eagles in Maine. J
Wild. Manage. 46:636-645.

Trexler, J.C. and J. Travis. 1993. Nontraditional regres-
sion analyses. Ecology 74:1629-1637.

Zar, J. H. 1984. Biostatistical analysis. Prentice-Hall, Inc.,
Englewood Cliffs, NJ U.S.A.

Received 6 August 1997; accepted 23 June 1998

J Raptor Res. 32(3):221-228
© 1998 The Raptor Research Foundation, Inc.

URBAN, SUBURBAN AND RURAL RED-TAILED HAWK NESTING
HABITAT AND POPULATIONS IN SOUTHEAST WISCONSIN

William E. Stout

W2364 Heather Street, Oconomowoc, WI 53066 U.S.A.

Raymond K. Anderson

University of Wisconsin-Stevens Point, Stevens Point, WI 54481 U.S.A.

Joseph M. Papp

Rt. 1 Box 158 A, Drummond, WI 54832 U.S.A.

Abstract. — Nesting Red-tailed Hawks (Buteo jamaicensis) are becoming increasingly common in urban
environments. We described Red-tailed Hawk nesting habitat and reproductive success and compared
urban, suburban, and rural nesting locations in southeast Wisconsin. Nest sites were classified as urban,
suburban or rural if >70%, 30-70%, or <30% of the area (706.9 ha, 1.5-km radius) around nests was
used for industrial or residential purposes, respectively. Mean success and productivity of breeding Red-
tailed Hawks in the metropolitan Milwaukee area from 1989-94 ( N = 426) was 81.9% (range = 75.3-
92.7%) and 1.43 young/breeding pair (range = 1.13-1.91), respectively. Brood size averaged 1.75
young/successful nest (range = 1.61-2.06). Productivity was variable and was significandy higher in
1994 than each of the preceding yr (P < 0.001). Based on internest distances, the density of the Red-
tailed Hawk nesting population for rural locations was greater than in suburban areas and lowest in
urban locations. The amount of natural microhabitat cover around nests (19.6 ha, 0.25-km radius) did
not differ for urban, suburban, or rural nest sites ( P = 0.967) indicating that cover was an important
component of the nesting habitat of Red-tailed Hawks. Natural cover comprised about 16% of the
landscape area of urban sites and 40% of this area was wooded with the remaining 60% consisting of
herbaceous cover. Urban planning should consider the amount of natural cover to allow Red-tailed
Hawks and other wildlife to coexist with humans in an urban environment.

Key Words: Red-tailed Hawk, Buteo jamaicensis; urban; suburban ; rural; nesting habitat, nesting density.

Habitat de anidacion urbano, suburbano y rural de Buteo jamaicensis en el sureste de Wisconsin

Resumen. — La anidacion en areas de Buteo jamaicensis es cada vez mas comun en ambientes urbanos.
Describimos el habitat de anidacion de Buteo jamaicensis y su exito reproductive y comparamos las
localidades urbanas, suburbanas y rurales de anidacion en el sureste de Wisconsin. Los sitios de los
nidos fueron clasificados como urbanos, rurales y suburbanos si ^70%, <30%, y 30-70% del area (706.9
ha, 1.5 km de radio) alrededor del nido eran utilizadas para propositos industrial o residencial (desar-
rollo) respectivamenmte. La media del exito en la productividad de los nidos ocupados por Buteo ja-
maicensis en el area metropolitana de Milwakee entre 1989-94 (N = 426) fue de 81.9% (rango = 75.3-
92.7%) y 1 .43 juveniles/nido ocupado (rango = 1.13-1.91). Tamano de la nidada promedio de 1.75
juveniles/ nido exitoso (rango = 1.61-2.06). La productividad fue variable y significativamente mas alta
en 1994 que en cada uno de los anos precedentes (p < 0.0001). Con base en la distancia entre nidos
se observo que la densidad de la poblacion reproductiva de las localidades rurales, fue mayor que en
las areas suburbanas y fue menor en areas urbanas. La cantidad de cobertura de microhabitat natural
alrededor de los nidos (19.6 ha, 0.25 km de radio) no fue diferente entre los sitios de los nidos urbanos,
suburbanos y rurales ( P = 0.967) lo cual indica que la cobertura es un componente importante del
habitat de anidacion de Buteo jamaicensis. La cobertura natural incluyo el 16% del microhabitat de los
sitios urbanos, 40% de esta area eran bosques y el 60% restante eran cobertura de pastizales. La pla-
neacion urbana debe considerar la cantidad de cobertura natural requerida para que Buteo jamaicensis
y la vida silvestre puedan coexistir con los humanos en un ambiente urbano.

[Traduccion de Cesar Marquez]

221

222

Stout et al.

Vol. 32, No. 3

Red-tailed Hawks ( Buteo jamaicensis) nest in ur-
ban environments, yet no comprehensive studies
have been published on their urban nesting habi-
tat. Two reports in Michigan document the suc-
cessful nesting of red-tails in urban settings (Val-
entine 1978, Hull 1980), and urban nesting also
has been reported in Puerto Rico (Santana et al.

1986) and New York (Minor et al. 1993).

Three studies of rural Red-tailed Hawk popula-
tions have previously been conducted in Wisconsin
(Orians and Kuhlman 1956, Gates 1972, Petersen
1979). Howell et al. (1978) correlated nesting hab-
itat structure and productivity at rural nest sites in
Ohio and found that highly productive sites had
more than twice as much fallow land, less than half
as much cropland, and less than half the number
of woodlots than did sites with low productivity.
Other studies of red-tails conducted in rural areas
throughout North America have described other
aspects of red-tail ecology (e.g., Wiley 1975, Fitch
and Bare 1978, Adamcik et al. 1979).

Our objectives were to describe Red-tailed Hawk
nesting habitat and reproductive success, and to
compare urban, suburban, and rural nesting loca-
tions in southeast Wisconsin. We determined rela-
tive nesting population densities for all three lo-
cations based on internest distances and identified
important physical components of the nesting hab-
itat.

Study Area

Our study area covered approximately 1100 km 2 locat-
ed in the metropolitan Milwaukee area in southeast Wis-
consin (43°N, 88°W). It included Milwaukee county and
parts of Waukesha, Washington, and Ozaukee counties.
Milwaukee and Ozaukee counties are bordered by Lake
Michigan to the east. Milwaukee county covers an area
of 626.5 km 2 . The city of Milwaukee covers an area of
248.5 km 2 with a human population of 629 554 (1994
population estimate; 2533 people per km 2 ). Human pop-
ulation density decreases radially from the city of Milwau-
kee to suburban communities and to rural areas. Two
interstate highways transect the study area. Land use
within the study area included agricultural, natural, in-
dustrial/commercial, and residential areas.

Methods

Red-tailed Hawk nests were located from a vehicle
from 1 February-30 April, 1987-94 (Craighead and
Craighead 1956) and visited at least twice (once within
10 d after the onset of incubation and again when nest-
lings were 20—35 d of age) during each nesting season to
determine productivity (Postupalsky 1974, Steenhof

1987) . Woodlots that were not entirely visible from the
road early in the season before leaf-out were checked by
foot. A breeding pair (i.e., eggs were laid) was considered

successful if >1 nestling survived to a bandable age (20-
35 d). Intrayear internest distances for 1989 and 1990
were measured to determine the nearest breeding pair
of Red-tailed Hawks (nearest neighbor; Clark and Evans
1954). These data were used as an index for population
nesting density and to compare urban, suburban, and
rural densities (Clark and Evans 1954, McGovern and
McNurney 1986). We believe that all nests were found in
urban and suburban areas and, therefore, the distances
between nests in these locations are accurate.

To describe Red-tailed Hawk nesting habitat and to
compare urban, suburban, and rural locations, we char-
acterized features of 1989 and 1990 nest sites on four
different spatial scales: 1) nest site, 2) habitat, 3) mac-
rohabitat and 4) landscape (Titus and Mosher 1981, Mo-
sher et al. 1986, 1987, Adamus 1995, Stout 1995; Table
1). The nest-site scale described the nest and nest tree
and data were collected when nestlings were 20-35-d old
Nest exposure (i.e., the open side of the nest) was as-
signed one of the following values: total access/ exposure,
N, NE, E, SE, S, SW, W, or NW. The nest tree was clas-
sified as being in a woodlot interior (the tree crown did
not touch a woodlot edge), on the edge of an interior
woodlot clearing (clearing was 2:0.1 ha), savannah (not
on an edge), woodlot edge, hedgerow, lone tree, pow-
erline tower, or billboard.

The habitat scale described vegetation within a 0.04-ha
circular plot (11.3 m radius) centered on the nest tree
and data were collected after fledging through Septem-
ber for 1989 and 1990 nest sites. Canopy, understory,
shrub, ground cover, and slope of the plot were de-
scribed according to Titus and Mosher (1981) and Mo-
sher et al. (1986, 1987). Shrub structure was classified by
shrub density, shrub index and density board (Mosher et
al. 1986). Slope and slope aspect were determined for
sites with a slope 22% using a compass and clinometer.

The landscape scale described land use within a 1.5-
km radius (706.9 ha) of the nest tree. Data were collected
for 1989 and 1990 nest sites, and used for analysis and
nest site classification (i.e., as urban, suburban, or rural).
The amount of land with natural, agricultural, residen-
tial, and industrial cover types within the landscape area
was determined from 1990 aerial photos (1 cm = 48 m)
with a compensating polar planimeter. The number of
individual areas of each cover type was recorded. Natural
habitat included woodlots, tree and shrub savannahs,
shrublands, herbaceous cover (grasses and forbs, fallow
fields, and inactive pastures) , and open water. The mean
area of open water was <1% (7 ha; maximum = 6.2%,
43.8 ha) and primarily consisted of pothole ponds and,
therefore, was included in the natural category. For man-
agement recommendations, natural habitat was subdivid-
ed into grassland and forest habitat. Agricultural land in-
cluded row crops (e.g., corn), cover crops (e.g., alfalfa
and clover), actively grazed pastures, tree nurseries and
orchards. Residential land included human dwellings
and other buildings and land associated with them. In-
dustrial land included nonresidential industrial and com-
mercial buildings, pavement, roads, graded land (e.g ,
gravel pits), mowed land (e.g., cemeteries, airports,
mowed land surrounding industrial buildings), and non-
mowed land associated with human activity (e.g., freeway
intersections, nonmowed land surrounding industrial

Sept 1998

Urban Red-tailed Hawks

223

buildings) . Each area was measured separately and com-
bined for analysis. Industrial and residential areas were
considered developed. Natural and agricultural areas
were considered undeveloped because they are devoid of
any buildings or pavement. A nest site was classified as
urban if >70% of the landscape area (706.9 ha) was de-
veloped, rural if ^30%, and suburban if 30-70% was de-
veloped (Stout et al. 1996). Hedgerow length was mea-
sured within the landscape area. The Baxter- Wolfe
interspersion index was determined from the changes in
cover type along the north-south and east-west median
lines within the landscape area (Baxter and Wolfe 1972,
Mosher et al. 1987). The area and perimeter of woodlots
containing nests were measured. Distances to the nearest
residence, industrial building and road were recorded
and mean distance to buildings was determined by using
a point-quarter method of measuring the distance to the
nearest building in each of four quadrants; a buffer area
(circular area surrounding the nest without buildings)
was calculated by using the mean distance to buildings
as the radius of a circle (Stout 1995). The macrohabitat
scale described land use within a 0.25 km radius (19.6
ha) of the nest for a comparison of land use patterns
closer to the nest site. The same variables that were mea-
sured at the landscape scale also were determined at the
macrohabitat scale.

Nest-site data were collected for all known breeding
pairs of Red-tailed Hawks in the metropolitan Milwaukee
area for 1989 and 1990. Nest sites that were used in both
1989 and 1990 (in either the same or a different nest
tree or structure) were included in the analysis only
once. Macrohabitat and landscape-scale data were col-
lected on all urban sites and at least as many suburban
and rural sites. According to our definitions, 15 urban
nest sites were found. For the urban, suburban, and rural
comparison, 22 suburban and 18 rural nest sites were
identified. Nest-site and habitat data were collected for
these sites where access (landowner permission) was
granted.

Categorical data were tested with a Chi-square good-
ness of fit. Urban, suburban, and rural nest sites were
compared using univariate statistics. Frequency distribu-
tions were used to determine variables with normal dis-
tributions. Log transformations were used when applica-
ble. Quantitative variables with normal distributions were
treated with parametric methods (one-way ANOVA). The
TUKEY multiple range test was used to identify different
groups. Nonparametric methods (Kruskal-Wallis test,
Chi-square approximation; Sokal and Rohlf 1981) were
used for nonparametric variables. All tests were consid-
ered significant when P < 0.05. The Statistical Package
for the Social Sciences (SPSS; Nie et al. 1975) was used
for statistical analyses.

Results

Productivity did not differ among urban, subur-
ban, and rural nest sites used by breeding Red-
tailed Hawks (Table 1). Mean nesting success for
Red-tailed Hawks in the Milwaukee metropolitan
area from 1989-94 (N = 426) was 81.9% (range =
75.3—92.7%; Table 2). Productivity of breeding

pairs for the same 6-yr period averaged 1.43
young/breeding pair (range — 1.13-1.91), and
1.75 young/successful nest (range — 1.61-2.06).
Productivity was significantly higher in 1994 than
each of the preceding years (P < 0.001). Mean in-
ternest distance for urban sites was greater than in
suburban and rural sites ( P = 0,004, P < 0.001,
respectively) , and mean internest distance was
greater for suburban than rural sites (P — 0.018;
Table 1).

In 1989 and 1990, we found 89 breeding Red-
tailed Hawks nesting in 18 species of trees. Four
were on high voltage transmission towers and one
was on a billboard. Nests constructed in trees and
on unnatural structures occurred in urban, sub-
urban, and rural areas (Stout 1995, Stout et al.
1996) . Only one nest-site variable, nest-tree height,
was different for urban, suburban, and rural loca-
tions indicating behavioral consistency in nest
building (Stout 1995, Table 1). Nest structures
w T ere in woodlots or on edges of woodlots more
often than in hedgerows, totally exposed lone
trees, or human-made structures (x 2 = 23.273, df
= 2 , P < 0.001). Nests had a northw r est exposure
more often than other directions ( N = 88; Fig. 1;
X 2 = 35.955, df = 8, P< 0.001). Sloped sites (N =
41) were not used more often than nonsloped sites
(N - 38; x 2 - 0.114, df = 1 , P = 0.736). When
sloped, red-tails used a southeast slope more often
than other directions (Fig. 1; x 2 = 19.293, df = 7,
P= 0.007).

At the habitat scale, the percent slope of plots
was greater for suburban sites than for rural sites,
the number of shrub species at suburban sites was
greater than at both urban and rural sites, and the
number of small understory saplings (dbh — 1-4
cm) at suburban sites was greater than at rural sites
(Table 1).

At the landscape scale, total hedgerow length
within the landscape area, mean building distance,
buffer area, nearest residence, industrial structure,
building, road, the Baxter-Wolfe interspersion in-
dex, and the amount of natural, agricultural, in-
dustrial and residential land were different for ur-
ban, suburban, and rural sites (Table 1). At the
macrohabitat scale, agricultural, industrial, and res-
idential land use were different, but the amount of
natural cover (total grassland and forest cover) did
not differ among the three sites (Table 1). Natural
cover within the macrohabitat area averaged 10.3
ha for all three locations while natural habitat with-
in the larger landscape area averaged 111.3 ha

224

Stout et al.

Vol. 32, No. 3

Table 1. Comparison of productivity, nest site, habitat (0.04-ha circular plot, 11.3-m radius), macrohabitat (19.6-ha,
0.25-km radius) and landscape (706.9 ha, 1.5 km radius) for urban, suburban and rural Red-tailed Hawk nest sites.
Nest site and habitat results do not include nests on artificial substrates. Productivity, macrohabitat and landscape
results include nests on artificial substrates.

Variable

Urban Nest Sites

Suburban
Nest Sites

Rural Nest Sites

One-way ANOVA b
Kriiskai.-Wat.tis TesV

F/x 2 P

Mean ± SE
Range ( N)

Mean ± SE
Range ( N)

Mean ± SE
Range ( N)

Productivity

1.27 ± 0.25

1.50 ± 0.19

1.44 ± 0.22

0.593 c

0.744

0-3 (15)

0-3 (22)

0-3 (18)

Nest Site

Nest tree height (m)

20.09 ± 1.00 x

23.33 ± 0.67?

21.09 ± 0.99 x ?

3.699 b

0.033

14.10-26.30 (11)

18.50-28.96 (20)

14.17-28.65 (16)

Habitat (0.04-ha circular plot, 11.3-m radius)

% Slope

2.7 ± 1.75 x ?

3.6 ± 0.86 x

1.0 ± 0.46?

6.076 c

0.048

0-10 (7)

0-16 (21)

0-6 (15)

No. shrub species

4.6 ± 0.95 x

7.4 ± 0.56?

4.5 ± 0.84 x

5.640 b

0.007

1-8 (7)

4-12 (21)

0-11 (15)

No. small saplings

48.3 ± 9.61 x ?

72.6 ± 8.67 x

40.9 ± 9.32?

6.420 c

0.040

0-78 (7)

15-183 (21)

0-113 (15)

Macrohabitat Area (19.6-ha, 0.25-km radius)

Grassland (ha)

4.77 ± 1.38

4.37 ± 0.76

4.53 ± 1.34

0.143 c

0.813

0.0-17.2 (15)

0.0-13.3 (22)

0.0-18.5 (18)

Forest (ha)

4.83 ± 1.19

6.06 ± 0.72

5.91 ± 1.05

1.528 c

0.466

0.0-13.3 (15)

0.0-15.2 (22)

0.3-14.1 (18)

Natural (ha)

9.76 ± 1.68

10.62 ± 0.97

10.44 ± 1.49

0.067 c

0.967

0.0-17.7 (15)

2.6-18.8 (22)

1.3-19.6 (18)

Agricultural (ha)

0.17 ± 0.1 7 X

3.89 ± 1.05?

7.76 ± 1.53 z

21.17U

<0.001

0.0-2. 6 (15)

0.0-15.6 (22)

0.0-18.2 (18)

Industrial (ha)

5.24 ± 1.68 x

3.14 ± 0.96 x

0.46 ± 0.25?

10.263 c

0.006

0.0-18.9 (15)

0.0-16.2 (22)

0.0-3. 5 (18)

Residential (ha)

4.43 ± 1.16*

1.96 ± 0.56?

0.93 ± 0.45?

15.160 c

0.001

0.7-19.0 (15)

0.0-9. 1 (22)

0.0-6. 4 (18)

Landscape

Woodlot area 3 (ha)

9.93 ±4.19

8.53 ± 1.27

9.39 ± 2.99

0.1 64 b

0.850

0.3-45.4 (11)

2.5-20.2 (20)

0.3-39.5 (15)

Woodlot perimeter 3 (m)

1550 ± 403.0

1425 ± 137.0

1715 ± 440.9

0.348 b

0.708

288-3936 (11)

768-2688 (20)

307-6816 (15)

Mean building dis. (m)

224 ± 17.7 X

322 ± 35.4?

455 ± 29. 9 Z

12.607 b

<0.001

68-341 (15)

79-759 (22)

150-692 (18)

Buffer area 3 (ha)

17.10 ± 2.35 x

40.89 ± 8.85 x

62.18 ± 7.27?

9.004 b

<0.001

1.5-36.5 (15)

2.0-181.0 (22)

7.1-127.5 (18)

Nearest residence 3 (m)

117 ± 10. 6 X

240 ± 26.1?

289 ± 34.3?

11.327 b

<0.001

30-178 (15)

86-533 (22)

67-571 (18)

Nearest industry 3 (m)

348 ± 69. 8 X

397 ± 72. l x

743 ± 97.2?

6.915 b

0.002

48-1080 (15)

62-1166 (22)

187-1375 (17)

Nearest building 3 (m)

106 ± 11.3*

180 ± 15.0?

272 ± 33. 0 Z

12.620 b

<0.001

30-178 (15)

62-293 (22)

67-571 (18)

Nearest road 3 (m)

114 ± 14.9 X

218 ± 28.5?

322 ± 48.5?

8.292 b

0.001

24-197 (15)

53-518 (22)

38-878 (18)

Mean internest dis. 3 (m)

2743 ± 319. 3 X

1780 ± 120.9?

1316 ± 165.5 Z

11.322 b

<0.001

1327-4968 (15)

799-2904 (20)

403-2246 (15)

Sept 1998

Urban Red-tailed Hawks

225

Table 1. Continued.

Variable

Urban Nest Sites

Suburban
Nest Sites

Rural Nest Sites

One-way ANOVA b
Kruskal-Wallis Test 0

F/x 2 P

Mean ± SE
Range ( N)

Mean ± SE
Range (N)

Mean ± SE
Range (N)

Landscape Area (706.9-ya, 1.5-km radius)

Baxter-Wolfe Index

18.3 ± 1.36 x

28.8 ± 1.03?

26.2 ± 1.26 z

19.304 b

<0.001

8-27 (15)

21-40 (21)

19-37 (18)

Hedgerow length (m)

7619 ± 1087 x

10 506 ± 995 x ?

12 053 ± 981?

4.258 b

0.019

2208-16080 (15)

1920-18 432 (22)

3984-18 720 (18)

Grassland (ha)

67.20 ± 11.14*

137.23 ± 8.57?

141.18 ± 22.77?

6.707 b

0.003

0.0-146.3 (15)

70.0-231.9 (22)

24.7-312.5 (18)

Forest (ha)

39.30 ± 6.26 x

77.82 ± 7.56?

103.80 ± 9.70 z

14.007 b

<0.001

0.0-94.0 (15)

31.1-178.9 (22)

43.1-187.3 (18)

Natural (ha)

111.27 ± 13.52 x

221.07 ± 10.68?

253.11 ± 29.22?

13.166 b

<0.001

16.3-190.2 (15)

123.7-329.4 (22)

81.3-457.4 (18)

Agricultural (ha)

11.69 ± 4.05 x

128.05 ± 14.75?

309.74 ± 30.76 z

40.587 c

<0.001

0.0-48.8 (15)

20.5-310.3 (22)

108.2-534.4 (18)

Industrial (ha)

273.85 ± 35.34 x

180.45 ± 18.94?

53.57 ± 9.77 z

23.1l7 b

<0.001

56.6-499.1 (15)

39.6-354.2 (22)

0.0-123.0 (18)

Residential 3 (ha)

310.00 ± 31.28 x

177.27 ± 15.65?

90.68 ± 9.60 z

25.905 b

<0.001

153.4-537.2 (15)

21.9-331.5 (22)

25.5-173.2 (18)

a Variables log-transformed for one-way Analysis of Variance (one-way ANOVA) .
b One-way ANOVA F values.

c Kruskal-Wallis test x 2 values (x 2 approximation) .

^Values followed by the same superscript letter x , r or z , are not significantly different at the P < 0.05 level (TUKEY multiple range
test b or Mann-Whitney U test c ) .

(15.7%) for urban nest sites only, and this natural
habitat was interspersed among developed land in
an average of 16.4 different tracts.

Discussion

Reproductive success and productivity of breed-
ing Red-tailed Hawks during our 6-yr study was
comparable to that of previous studies in Wiscon-
sin (Orians and Kuhlman 1956, Gates 1972, Peter-
son 1979; Table 2) and an urban/suburban area
in New York (Minor et al. 1993). Red-tailed Hawk
nest success estimates for North America range
from 58-93% (Preston and Beane 1993).

The distance between breeding pairs of Red-
tailed Hawks was used as an index of nesting den-
sity (McGovern and McNurney 1986). Our mean
internest distance of 1.9 km was comparable to
other studies (Fitch et al. 1946, Orians and Kuhl-
man 1956, Gates 1972, Petersen 1979, McGovern
and McNurney 1986). Rural nests were significant-
ly closer together than suburban and urban nests,
and suburban nests were closer together than ur-
ban nests which indicated that nesting density de-

creased from rural to urban areas. We found rural
nests adjacent to suburban nests at the perimeter
of our study area. As a result, the nearest breeding
pair of red-tails may not have been found in all
rural areas making rural nests even closer than our
data indicated. Peterson (1979) found a mean in-
ternest distance of 1.51 km in rural Wisconsin. Our
mean internest distance of 1.32 km between rural
nests may indicate that the density of nesting Red-
tailed Hawks may have increased in rural southeast
Wisconsin over the past 25 yr, possibly because of
increased availability of nesting habitat resulting
from changes in agricultural practices such as the
conservation reserve program (CRP).

The microclimate surrounding nest structures is
important in the selection of nest sites by raptors.
We found Red-tailed Hawk nests had predomi-
nantly northern exposures (primarily NW and NE)
and sloped sites had southeast aspects. Speiser and
Bosakowski (1988) also found Red-tailed Hawk
nests to have southeast facing slope exposures.
They suggested that a southeast slope maximizes
insulation to the nest on cold mornings and min-

226

Stout et at.

Vol. 32, No. 3

Table 2.

Red-tailed Hawk

reproductive

success from 1989-

-94 for the metropolitan

Milwaukee, Wisconsin

area.

Breeding

Nest

Nf.st

Young/Nest

Reproductive Success

Yr

Pairs

Failures Success (%)

1

2

3

A a

B b

1989

59

11

81.4

20

24

4

1.36

1.67

1990

85

21

75.3

19

39

6

1.35

1.80

1991

92

16

82.6

33

40

3

1.33

1.61

1992

83

9

89.2

24

45

5

1.55

1.74

1993

52

16

69.2

16

17

3

1.13

1.64

1994

55

4

92.7

13

22

16

1.91

2.06

Total

426

77

81.9

125

187

37

1.43

1.75

a Young/breeding pair.
b Young/ successful nest.

imizes the possibility of heat stress in the after-
noon. Southeast slopes may help to keep nestlings
dry by minimizing the effects of predominantly
northwest storm winds in Wisconsin while north-
ern nest accesses may provide more shade and re-
duce heat stress. Several studies also found that
nest sites usually have unobstructed access and a
commanding view of the surrounding area (Peter-
sen 1979, Bednarz and Dinsmore 1982, Santana et
al. 1986, Speiser and Bosakowski 1988, Bechard et
al. 1990, Toland 1990, Preston and Beane 1993).
Sloped nest sites probably provide this type of nest
orientation.

Red-tailed Hawks used similar types of nest sites
in urban, suburban, and rural locations, however,
suburban nest sites tended to be located on sloped
sites and in wetlands, probably because upland
sites are developed first. Suburban areas also had
the highest land use diversity (Baxter-Wolfe inter-
spersion index) while urban locations had the least
amount of land use diversity. Woodlot area and pe-
rimeter remained relatively constant for urban,

Nest

Exposure

N-6

Slope

Aspect

N-2

E-3

Figure 1. Nest exposure (N = 84) and slope aspect ( N
= 41) at Red-tailed Hawk nest sites in southeast Wiscon-
sin. Sample size is indicated for each direction.

suburban, and rural nesting locations indicating
that 9 ha may represent an ideal size woodlot for
Red-tailed Hawk nesting sites. Other studies have
found that red-tails selected smaller woodlots,
open stands, and woodlot edges compared to larg-
er woodlots or closed canopy woodlot interiors
(Orians and Kuhlman 1956, Gates 1972, Petersen
1979). Speiser and Bosakowski (1988) found that
red-tails nested closer to forest openings than ran-
dom sites and Howell et al. (1978) reported that
the most productive pairs of Red-tailed Hawks used
small woodlots.

Landscape variables (e.g., nearest road, industry,
residence) varied significantly and increased from
urban to suburban and rural areas. The amount of
natural and agricultural land within the landscape
scale decreased as the amount of industrial and
residential land increased. While the amount of ag-
ricultural land increased and residential and in-
dustrial land decreased at the macrohabitat scale
from rural through suburban and urban areas, the
amount of natural cover within the macrohabitat
remained consistent for all three areas averaging
10.3 ha indicating that natural cover constitutes an
important nesting habitat component for Red-
tailed Hawks.

For the purposes of urban planning and devel-
opment, we believe that managing for important
habitat components such as natural cover will en-
hance the availability of nesting habitat for Red-
tailed Hawks in urban areas. Based on our find-
ings, we recommend that at least 16% of urban
land be left in natural habitat with approximately
40% wooded and 60% herbaceous cover. This nat-
ural habitat should be distributed among residen-
tial and industrial land in approximately 16 sepa-

Sept 1998

Urban Red-tailed Hawks

227

rate tracts within the landscape area (706.9 ha).
Wooded areas should be approximately 9 ha to
provide suitable nesting woodlots.

Acknowledgments

J.W. Hardin and R.W. Miller provided guidance and
editorial assistance. D. Hebbert helped collect nest site
and habitat data. R.D. Beane, S. Postupalsky, K. Titus and
two anonymous reviewers provided comments and sug-
gestions that greatly improved this manuscript. J.A. Rei-
nartz and R. Rogers provided assistance with statistical
analyses. The senior author’s wife, Vicki, daughter, Jen-
nifer, and son, Tim, provided continual support, patience
and assistance in all areas of this research project. This
project was partially funded by the U.W.-Stevens Point
Graduate School Student Research Fund, and V.W. De-
broux. An earlier version of this manuscript was submit-
ted to the University of Wisconsin-Stevens Point, College
of Natural Resources, in April 1995 for partial fulfillment
of a Master of Science degree in Natural Resources
(Wildlife).

Literature Cited

Adamcik, R.S., A.W. Todd and L.B. Keith. 1979. Demo-
graphic and dietary responses of Red-tailed Hawks
during a snowshoe hare fluctuation. Can. Field-Nat. 93:
16-27.

Adamus, P.R. 1995. Validating a habitat evaluation meth-
od for predicting avian richness. Wildl Soc. Bull. 23:
743-749.

Baxter, W.L. and C.W. Wolfe. 1972. The interspersion
index as a technique for evaluation of Bobwhite Quail
habitat. Pages 158-165 mJ.A. Morrison andJ.C. Lewis
[Eds.], Proc. First Natl. Bobwhite Quail Symp. Okla.
State Univ., Res. Found. Stillwater, OK U.S.A.
Bechard, M.J., R.L. Knight, D.G. Smith and R.E. Fitzner.
1990. Nest sites and habitats of sympatric hawks ( Buteo
spp.) in Washington. / Field Ornithol. 61:159-170.
Bednarz, J.C. and J.J. Dinsmore. 1982. Nest sites and hab-
itat of Red-shouldered and Red-tailed Hawks in Iowa.
Wilson Bull. 94:31-45.

Clark, P-J- and F.C. Evans. 1954. Distance to nearest
neighbor as a measure of spatial relationships in pop-
ulations. Ecology 35:445—453.

Craighead, J-J- and F.C. Craighead. 1956. Hawks owls
and wildlife. The Stackpole Co., Harrisburg, PA and
Wildlife Manage. Inst., Washington DC U.S.A.

Fitch, H.S., F. Swenson and D.F. Tillotson. 1946. Be-
havior and food habits of the Red-tailed Hawk. Condor
48:205-257.

and R.O. Bare. 1978. A field study of the Red-
tailed Hawk in eastern Kansas. Trans. Kansas Acad . Sci.
81:1-13.

Gates, J.M. 1972. Red-tailed Hawk populations and ecol-
ogy in east-central Wisconsin. Wilson Bull. 84:421-433.
Howell, J., B. Smith, J.B. Holt and D.R. Osborne. 1978.
Habitat structure and productivity in the Red-tailed
H awk. Bird Banding 49 : 1 62-1 7 1 .

Hull, C.N. 1980. Additional successful nesting of a Red-
tailed Hawk in an urban subdivision. Jack Pine Warbler
58:30.

McGovern, M. and J.M. McNurney. 1986. Densities of
Red-tailed Hawk nests in aspen stands in the Piceance
Basin, Colorado./. Raptor Res. 20:43—45.

Minor, W.F., M. Minor and M.F. Ingraldi. 1993. Nesting
of Red-tailed Hawks and Great Horned Owls in a cen-
tral New York urban/suburban area. / Field Ornithol.
64:433-439.

Mosher, J.A., K. Titus and M.R. Fuller. 1986. Devel-
oping a practical model to predict nesting habitat of
woodland hawks. Pages 31-35 in J. Verner, M.L. Mor-
rison and C.J. Ralph [Eds.], Wildlife 2000: modeling
habitat relationships of terrestrial vertebrates. Univ. of
Wisconsin Press, Madison, WI U.S.A.

, and . 1987. Habitat sampling, mea-
surement and evaluation. Pages 81—97 in B.A. Giron
Pendleton, B.A. Millsap, K.W. Cline and D.M. Bird
[Eds.], Raptor management techniques manual. Nat.
Wildl. Fed., Washington, DC U.S.A.

Nie, N.H., C.H. Hull, J.G. Jenkins, K. Steinbrenner and
D.H. Bent [Eds.]. 1975. Statistical package for the so-
cial sciences. McGraw Hill, Inc., New York, NY U.S.A.

Orians, G. and F. Kuhlman. 1956. Red-tailed Hawk and
horned owl populations in Wisconsin. Condor 58:371-
385.

Peterson, L. 1979. Ecology of Great Horned Owls and Red-
tailed Hawks in southeastern Wisconsin. Wis. Dept. Nat.
Res. Tech. Bull. No. Ill, Madison, WI U.S.A.

Postupalsky, S. 1974. Raptor reproductive success: some
problems with methods, criteria, and terminology.
Pages 21-31 in F.N. Hamerstrom, B.E. Harrell and
R.R. Olendorff [Eds.], Management of raptors. Rap-
tor Res. Rep. 2.

Preston, C.R. and R.D. Beane. 1993. Red-tailed Hawk
(Buteo jamaicensis) . In A. Poole and F. Gill [Eds.], The
birds of North America., No. 52. Philadelphia: Acad
Nat. Sci.; Am. Ornithol. Union, Washington, DC
U.S.A.

Santana, E.C., E.N. LaBoy, J.A. Mosher and S.A. Tem-
ple. 1986. Red-tailed Hawk nest sites in Puerto Rico.
Wilson Bull. 98:561-570.

Sokal, R.R. AND F.J. Rohlf. 1981. Biometry. W.H. Free-
man and Co., New York, NY U.S.A.

SPEISER, R. AND T. BOSAKOWSKI. 1988. Nest site prefer-
ences of Red-tailed Hawks in the highlands of south-
eastern New York and northern New Jersey. /. Field
Ornithol. 59:361-368.

Steenhof, K. 1987. Assessing raptor reproductive success
and productivity. Pages 157-170 in B.A. Giron Pen-
dleton, B.A. Milsap, K.W. Cline, D.M. Bird [Eds.],
Raptor management techniques manual. Natl. Wildl.
Fed., Washington, DC U.S.A.

Stout, W.E. 1995. An urban, suburban, rural Red-tailed
Hawk nesting habitat comparison in southeast Wis-

228

Stout et al.

Yol. 32, No. 3

consin. M.S. thesis, Univ. Wisconsin, Stevens Point, WI
U.S.A.

, R.K. Anderson and J.M. Papp. 1996. Red-tailed

Hawks nesting on man-made and natural structures
in southeast Wisconsin. Pages 77-86 in D.M. Bird,
D.E. Varland andJ.J. Negro [Eds.], Raptors in human
landscapes. Academic Press, London, U.K.

Titus, K. andJ.A. Mosher. 1981. Nest-site habitat select-
ed by woodland hawks in the central Appalachians.
Auk 98:270-281.

Toland, B.R. 1990. Nesting ecology of Red-tailed Hawks
in central Missouri. Trans. Miss. Acad. Sci. 24:1-16.

Valentine, A.E. 1978. The successful nesting of a Red-
tailed Hawk in an urban subdivision . Jack Pine Warbler
56:209-210.

Wiley, J.W. 1975. Nesting and reproductive success of
Red-tailed Hawks and Red-shouldered Hawks in Or-
ange County, California, 1973. Condor 77:133-139.

Received 21 November 1997; accepted 19 May 1998

/. Raptor Res. 32(3) :229-232
© 1998 The Raptor Research Foundation, Inc.

HIGHWAY MORTALITY OF BARN OWLS IN
NORTHEASTERN FRANCE

Sylvie Massemin 1

Centre d’Ecologie et de Physiologie Energetiques, CNRS, 23 Rue Becquerel, 67087 Strasbourg , Cedex 2, France

Thierry Zorn

Office National de la Chasse, 1 7 Avenue de Wagram, 7501 7 Paris, Cedex, France

Abstract. — We found a total of 187 road-killed raptors along a 150 km stretch of highway in north-
eastern France between 1990-94. Of these, 148 were Barn Owls (Tyto alba), 15 Long-eared Owls ( Asio
otus), and 10 Tawny Owls ( Strix aluco). We analyzed different variables including the topography of
highways, the types of habitats crossed by highways, and the types of vegetation along highways to
determine why so many Barn Owls were killed. Most mortalities (64%) occurred along embanked
stretches that crossed open fields and lacked hedges on either side. We concluded that the local pop-
ulation density and flight behavior of Barn Owls were probably related to such high mortality.

Key Words: Barn Owl\ Tyto Alba; highway mortality.

Mortalidad de Tyto alba en autopistas del noreste de Francia

Resumen. — Encontramos un total de 187 aves rapaces atropelladas a lo largo de un segmento de 150
km de autopista en el noreste de Francia entre 1990-94. De estas, 148 fueron Tyto alba, 15 Asio otusy
10 Strix aluco, para un total de 173 aves rapaces nocturnas. Analizamos las diferentes variables incluyendo
la topografla de las autopistas, el tipo de habitats atravezados por la autopista y tipos de vegetacion con
el fin de determinar las causas de mortalidad de Tyto alba. La mayorfa de las muertes (64%), ocurrieron
a lo largo de terraplenes angostos que cruzaban sitios abiertos sin arbustos a los lados. Concluimos que
la densidad poblacional local y el comportamiento de vuelo de Tyto alba estaban probablemente rela-
cionados con esta alta tasa de mortalidad.

[Traduccion de Cesar Marquez]

Studies indicate that large numbers of raptors,
especially Barn Owls (Tyto alba) are killed along
highways in Europe (Baudvin et al. 1991, de Bruijn
1994, Taylor 1994). In France, there is a predomi-
nance of owls killed and, although the numbers
vary according to region, Barn Owls ( Tyto alba ) are
most commonly killed (Bourquin 1983, Joveniaux
1986, Athanaze 1992). Most Barn owls are killed in
autumn and winter (Joveniaux 1986, Athanaze
1992, de Bruijn 1994, Taylor 1994), but there is
also a high mortality along highways during the
post-fledging period. Whereas the temporal varia-
tion in Barn Owl mortality has been shown, little
information is available on the spatial variation in
mortalities or the causes of death (Joveniaux 1986,
Athanaze 1992). Here, we present the results of a
study designed to show how such variables as the

1 Present Address: Section of Ecology, Department of Bi-
ology, University of Turku, 20500 Turku, Finland.

landscape crossed by the highway, the topography,
and the vegetation along the roadway affect Barn
Owl mortality.

Material and Methods

Raptors killed by vehicles were collected along a 150-
km section of a highway between Strasbourg and Metz
(northeastern France) over a 5-yr period (1990-94). In
the case of dead Barn Owls, the location of carcasses on
the highway was noted as either in the emergency stop-
ping lane, the traffic lanes, or the median strip. We also
noted the landscape crossed by the highway (forest, open
field including cultivated fields, wasteland, bogs, and con-
crete) , the topography, and the type of vegetation along
the sides of the highway. Because Barn Owls fly at an
average height of 5 m (Baudvin 1986), we classified this
section of embanked and excavated highway (Fig. 1) into
the following classes: highly embanked (>5 m elevation
on at least one side), shallow embanked (1-4 m elevation
on at least one side) , level highway, shallow excavated (1-
4 m excavation on at least one side), deeply excavated
(>5 m excavation on at least one side), and embanked/
excavated (height not distinguished). Vegetation along

229

230

Massemin and Zorn

Vol. 32, No. 3

Flight of the B. O.

Embanked road

Excavated road

Figure 1. Schematic representation of two highway to-
pographies. The arrow indicates the possible direction of
flight of Barn Owls killed.

the side of the highway was classified according to the
amount of hedge present (present, absent, and present/
absent) .

Factorial Correspondence Analysis (FCA) was used to
determine which variables explained Barn Owl mortality.
A variable highly correlated with one of the principal
axes explains a large part of the inertia of this axis (STAT-
ITCF, Dervin 1988). Variables identified using this meth-
od were verified using x 2 analysis. For this analysis, the
proportion of birds killed in each class of the variable
was compared to the kilometric proportion of the class
along the highway (proportion of killed birds expected) .
Therefore, if a variable had no influence on Barn Owl
mortality, the number of killed birds collected was the
same as the expected number.

Results

A total of 187 road-killed raptors was found rep-
resenting three owl and one buzzard species. Of
the owls, 148 were Barn Owls (86%), 15 Long-
eared Owls (Asio otus), and 10 Tawny Owls ( Strix
aluco) .

Although dead Barn Owls were not distributed
evenly along the section of highway studied, on av-

erage we found about 1 Barn Owl/5 km of high-
way. The largest number (59%) was found along a
small segment of the highway (23%) (Fig. 2). The
direction of traffic did not seem to affect mortality
but the majority of owls (55%) were found in the
emergency stopping lane. Only a few individuals
were found in the median strip (18%).

The results of FCA on the different variables
measured indicated that the topography of the
highway (mostly the excavated parts) and vegeta-
tion along the sides (present or absent) were im-
portant variables contributing 99.9% of the first
principal component (Al, Fig. 3). Likewise, topog-
raphy (principally level) and landscape (mainly
forest) contributed to 96.4% of the second prin-
cipal component (A2) . Most of the owls were killed
along embanked stretches of the highway that
lacked roadside hedges and crossed open fields.

Comparison of the number of observed mortal-
ities vs. those expected showed that mortalities did
in fact increase along embanked highway stretches
(X 2 = 13.78, P < 0.05) that crossed open fields (x 2
= 26.99, P < 0.05). Our findings were most signif-
icant when the highway stretch both lacked a
hedge and was highly embanked (x 2 = 4.82, P <
0.05). The stretch of highway with the highest rap-
tor mortality (59%) had the highest embankment
and lacked hedges (x 2 = 7.39, P < 0.05). Fewer
raptors (24%) were found dead in a stretch of the
highway with both a high excavation and a hedge
(X 2 = 8.4, P< 0.05).

Discussion

Unlike Bourquin (1983) and Joveniaux (1986),
who found most Barn Owls killed along highways

4 •
3 ■

o

2

i .

. .

ii

.

III

1

' 1 II 1

ii mu uni i I inn m i

i ii Inn ill

in 1 1

Illl II

III

111

_UUI

Number of birds

F orest/Openfield

Forest,

ns Min Bi i Hii 1 _i

Openfield

. i ,

Excavated road

>111111111 III Ii

III HR II III

lllllllllll IBI II III 111 III

1 1 in mu in ii 1 1 ilium

unt

iii iii mi min muni ii

S* III IfriTii fnl

Embanked road

ii n nr

TOirnfm

iMmW

lllllllllll

i ir i inriiFrr nnf liTrT FiKuMtai*

0

25

50

75

100

125 150

Landscape

Topography

km

Figure 2. Spatial distribution of Barn Owl mortalities on the highway studied (Strasbourg-Metz) . Number of raptors
killed, topography, and landscape features are shown. The largest number of dead Barn Owls was found between
kilometers 114-149.

Sept 1998

Road-killed Barn Owls

231

A2

1 1

.

.m '

1

>

c

El

v- c

>

[em

CJ a

[ex]* °

op^ hO

O Individuals
• Classes of va

i i

°-o.

n" 1 n-50

riables

i i

- 2-1 0 1 2

Figure 3. Biplot of the first two principal components of variables recorded at sites where Barn Owls were found
killed by vehicles. The first (Al) and the second (A2) principal components explained 33% and 28% of the variance,
respectively. Many individuals are on the same coordinates. To achieve a good representation of FCA, numbers are
proportional to the diameter of the point (see scale in lower right corner). Forest/open field class (14% of total
sector) is included in the open field class, embanked/excavated class (14% of total sector) is included in the em-
banked class, and the presence/ absence of hedge class (15% of total sector) is included in the class presence of
hedge. Symbols are as follows: landscape (forest [f| and open field [op]), togography (excavated road [ex], level [1],
and embanked road [em]), and hedge (absence [hO] and presence [hi]).

crossing through forests, we found that Barn Owls
were principally killed along sections of highway
that crossed open fields. This may have been due
to the fact that Barn Owls primarily hunt in open
landscapes (Michelat and Giraudoux 1992) but it
could also be related to the fact that Barn Owls
killed along highways crossing through forests may
have been attracted to excavated highway stretches
that supported hedges on either side. These
stretches support numerous small mammals that
take refuge at the sides of highways and are undis-
turbed by agriculture (Spitz 1977, van der Reest
1992). Barn Owls typically prey on small mammals
such as common voles ( Microtus arvalis ) , woodmice
(Apodemus sylvaticus ), and common shrews ( Sorex
araneus) (Bourquin 1981, Baudvin et al. 1991). Be-
cause the sides of highways favor high winter den-
sities of many of these species and make them
more available because they are frequently snow-
free (Bourquin 1983), these areas could potentially
become traps for Barn Owls that are killed by pass-
ing vehicles.

Portions of embanked highway that crossed
open fields also accounted for many of the Barn
Owl mortalities. In this type of a situation, Barn

Owls typically fly 2-5 m above the ground while
hunting (Baudvin 1986), a height which corre-
sponds to the normal height of passing trucks and
cars. Barn Owls may cross embanked highways
without climbing increasing the likelihood of im-
pacts. Since we found most dead owls in the emer-
gency lane of the highway, it appears that impacts
probably occurred at the edges of the highway,
when owls first started to cross the road. The
amount of traffic probably had little effect on in-
creasing the mortality (Canteneur 1964, Illner
1992) but the high speed of the traffic (>80 km/
hr) probably did increase the danger for owls. It
appeared that many of the owls were not killed by
direct impact with vehicles but by impact with the
ground after they were projected up into the air
by turbulence behind vehicles.

We feel that few Tawny and Long-eared Owls
were killed along this stretch of highway because
they were simply less common in the area. The
Tawny Owl is a woodland species and there was
little forest habitat along this segment of highway.
Likewise, few Long-eared Owls were known to oc-
cur in the area.

232

Massemin and Zorn

Vol. 32, No. 3

Acknowledgments

This study was supported by a grant from the Societe
des Autoroutes du Nord et de l’Est de la France (SANEF)
and by a grant from the French Ministry of the Environ-
ment. We are very grateful to the FIR (Fonds
d’lntervention pour les Rapaces) for their administrative
help. We are indebted to Drs. E. Challet and D. Currie
for constructive comments on this manuscript.

Literature Cited

Athanaze, P. 1992. Etude de l’impact de l’autoroute A6
sur les populations de rapaces nocturnes. Rapport du
Centre Ornithologique Rhone-Alpes (CORA). Univ.
Lyon 1, Villeurbanne, France.

Baudvin, H. 1986. La reproduction de la Chouette Ef-
fraie ( Tyto alba ) . Le Jean le Blanc XXV.

, J.C. Genot and Y. Muller. 1991. Les rapaces

nocturnes. Sang de la Terre, Paris, France.

Bourquin, J,D. 1981. Les petits mammiferes vivant le
long des autoroutes. Strasse und Verkehr 2A3— 47.

. 1983. Mortalite des rapaces le long de

l’autoroute Geneve-Lausanne. Nos oiseaux 37:1 49-1 69.
Canteneur, R. 1964. Les oiseaux sauvages victimes de la
circulation routiere dans l’Est de la France. L’Oiseau
et R.F.O. 34:252-267.

de Bruijn, O. 1994. Population ecology and conservation
of the Barn Owl (Tyto alba ) in farmland habitats in
Liermers and Achterhoek (the Netherlands). Ardea
82:1-109.

Dervin, C. 1988. STAT-ITCF (institut Technique des Cer-
eales et des Fourrages). Paris, France.

Illner, H. 1992. Road deaths of Westphalian owls: meth-
odological problems, influence of road type and pos-
sible effects of population levels. Pages 94—100 mC.A,
Galbraith, I. Taylor, and S. Percival [Eds.], The ecol-
ogy and conservation of European owls. Joint Nature
Conservation Committee, Peterborough, U.K.

J oveniaux, A. 1986. Influence de la realisation d’une au-
toroute sur les populations de Chouettes Effraies.
Rapport d’Environnement Participation et Amenage-
ment (EPA) 13. Lons le Saunier, France.

Michelat, D. and P. Giraudoux. 1992. Activite nocturne
et strategic de recherche de nourriture de la Chouet-
te Effraie ( Tyto alba) a partir du site de nidification.
Alauda 60:3-8.

Muselet, D. 1985. Etude du comportement de l’avifaune
lors de la traversee de voies autoroutieres, autoroute
A10 au Nord d’ Orleans, Loiret. Route et faune sau-
vage. Conseil de l’Europe, Strasbourg, France.

Spitz, F. 1977. Le campagnol des champs ( Microtus ar-
valis) en Europe. Bull. OEPP 7:165-175.

TAYLOR, I. 1994. Predator-prey relationships and conser-
vation. University Press, Cambridge, U.K.

van der Reest, P.J. 1992. Kleine zoogdieren in neder-
landse wagbermen: oecologie en beheer. Lutra 35:1-
27.

Received 6 August 1997; accepted 14 February 1998

J Raptor Res. 32(3):233-240
© 1998 The Raptor Research Foundation, Inc.

THE EFFECT OF BURROW SITE USE ON THE REPRODUCTIVE
SUCCESS OF A PARTIALLY MIGRATORY POPULATION OF

WESTERN BURROWING OWLS
(SPEOTYTO CUNI CULARIA HYPUGAEA)

Eugene S. Botelho 1 and Patricia C. Arrowood 2

P.0. Box 30001 /Dept. 3 At] Department of Biology, New Mexico State University, Las Cruces, NM 88003-0001 U.S.A

Abstract. — We compared the number of nestlings produced by pairs of Burrowing Owls ( Speotyto cun-
icularia hypugaea) using burrows in different types of nest sites, use of different types of burrows by
resident and migrant males, and burrow type use by returning migrant males and females and the
productivity of individuals that switched burrows. The number of nesdings and fledglings produced by
pairs nesting in artificial burrows was also compared to the productivity of pairs in natural burrows. We
determined that pairs in undisturbed areas used burrows located in or at the base of cliff walls more
often than any other burrow type, while pairs in disturbed areas used burrows on flat ground more
often. Both resident and migrant males used burrows in or at the base of cliff walls more often in
undisturbed areas but, in disturbed areas, they used burrows in flat ground more often. Most males and
females that switched burrows from one year to the next produced more nestlings in burrows they left
than in new burrows. Pairs which nested in artificial burrows produced significantly more nestlings than
those that used natural burrows, but pairs in natural burrows produced significantly more fledglings.

Our results suggest the importance of determining burrow sites favored by nesting owls prior to initia-
tion of conservation plans which require protection of areas containing nest holes or installation of
artificial burrows.

Key Words: Burrowing Owl ; Speotyto cunicularia hypugaea; nest type use, artificial burrows ; conservation.

El efecto de los sitios de madriguera en el exito reproductive de una poblacion parcialmente migratoria
de Speotyto cunicularia hypugaea

Resumen. — Gomparamos el numero de pichones producidos por pares de Speotyto cunicularia hypugaea
que utilizaron madrigueras en distintos tipos de sitios de anidacion, el uso de distintos tipos de mad-
rigueras por machos residentes y migratorios, el uso de distintos tipos de madrigueras por machos y
hembras que retornaron al mismo lugar y la productividad de los individuos que cambiaron madri-
gueras. El numero de pichones producidos por pares que anidaron en las madrigueras artificiales fue
comparado con la productividad de los pares que anidaron en las madrigueras artificiales. Determina-
mos que los pares en areas no perturbadas utilizaron madrigueras localizadas en la base de paredes en
precipicios en mas ocasiones que otro tipo de madrigueras, dnientr as que los pares en areas perturbadas
utilizaron madrigueras en el suelo con mas frecuencia. Los machos residentes y migratorios utilizaron
madrigueras en la base o en los precipicios con mayor frecuencia en las areas no perturbadas, pero en
las areas perturbadas utilizaron el suelo con mayor frecuencia. Los machos y hembras que cambiaron
madrigueras de un ano a otro, produjeron mas pichones en la madriguera que dejaron que en la nueva.
Los pares que anidaron en madrigueras artificiales produjeron significativamente mas pichones. Nues-
tros resultados resaltan la importancia de la determinacion de sitios de madrigueras para anidacion de
buhos antes de la iniciacion de planes de conservacion, los cuales pueden requerir de la proteccion de
areas que contengan cavidades de nidos o la instalacion de madrigueras artificiales.

[Traduccion de Cesar Marquez]

1 Present Address: 58 Ohio Street, New Bedford, MA 02745 U.S.A.

2 Present Address: Department of Wildlife and Fisheries Sciences, New Mexico State University, Las Cruces, NM 88003
U.S.A.

233

234

Rotelho and Arrowood

Vol. 32, No. 3

The western Burrowing Owl ( Speotyto cunicularia
hypugaea, from here on referred to as the Burrow-
ing Owl) nests in underground burrows usually
dug by other animals (Coulombe 1971, Thomsen
1971, Haug et al. 1993). Its requirement for un-
derground nests may leave it with few choices, de-
pending on the biology of animals that excavate
burrows in a particular location (e.g., colonial vs.
dispersed fossorial mammals) . Conversely, in
regions with diverse physiography (e.g., cavities in
the cliff faces of dry creeks or rivers), Burrowing
Owls may encounter a variety 7 of possible nest site
possibilities. Understanding the relationship be-
tween the use of different burrow site types and
reproductive success in burrowing owls is impor-
tant in light of recent conservation plans for the
species throughout much of its range (Haug et al.
1993).

This study was conducted on a population of
Burrowing Owls which nested on the campus of
New Mexico State University (NMSU). Partial mi-
gration occurs in this population with all females
and fledglings migrating from the study area each
year. The majority of males, however, reside on the
study area throughout the year (resident) but a few
migrate (migrants). Resident and migrant males
use nesting burrows (either retaining the previous-
ly-used one or switching to a new one) and defend
the area surrounding their burrows prior to the
arrival of females each year. Females begin to ar-
rive in the area in February and immediately
choose a mate.

Variation in burrow use by this population led
us to consider what factors might affect the pro-
ductivity 7 of pairs which nested in different types of
burrows. Burrow sites with no or low grass cover
and high elevation should offer the most protec-
tion from predators, thus, males should more com-
monly use burrows in cliff walls. Compared to flat
ground, burrows at the base of cliffs should offer
more protection from predators, so they should be
used more often than those on flat ground but less
than those in cliffs. An abundance of lights in our
study area attracted insects, bats and nighthawks
( Chordeiles spp.), all of which the owls ate. In some
lighted areas, the only type of burrow available was
on flat ground; in these cases the increased prey
availability should have offset increased predation
risk. Thus, based on food availability 7 , pairs in dis-
turbed areas should be more successful than those
in natural, nonlighted areas. Pairs with a resident
male should produce more nestlings due to the

increased experience of males in these areas and
their opportunity throughout the winter to assess
different burrows. We also felt that the use of ar-
tificial burrows should enhance reproductive suc-
cess since they are not susceptible to collapse, are
protected from flooding, and are impossible for
larger predators to dig out and enter.

In order to test these predictions, we evaluated
the reproductive success of Burrowing Owls that
used burrows of different types. We also compared
the types of burrows used by resident males with
that of migrants. To determine the effect of switch-
ing nest burrows from one year to the next, we
compared the number of nestlings produced by in-
dividuals at their current and previous burrow
sites. We also compared the number of nestlings
produced by pairs which used natural versus arti-
ficial burrows. Our results will help to determine
if the types of burrows used by Burrowing Owls
should be considered as part of future conserva-
tion plans, especially plans wilich involve installa-
tion of artificial burrows in areas without natural
burrows.

Study Area and Methods

Our study area on the NMSU campus encompassed a
triangle of approximately 364 ha. The campus included
irrigated pastures at its lowest elevation (—3900 m) and
Chihuahuan Desert vegetation of approximately 121 ha
at its highest elevation (—4100 m). Campus buildings oc-
cupied the central part of the triangle. The abundance
of rock squirrels ( Spermophilus variegatus ) throughout the
campus resulted in a large number of available burrows.
Sometimes the squirrels dug shallow burrows which the
owls enlarged. Spotted ground squirrels (Spermophilus spi-
losoma) dug smaller burrows which may have been en-
larged by the owls. In natural areas, rock squirrels, cot-
tontail rabbits ( Sylvilagus spp.) and jackrabbits ( Lepus
spp.) also dug burrows. The rabbit population was large
in the natural areas. Naturally-occurring crevices, abun-
dant throughout the campus, were also used or enlarged
by the owls. Since there were hundreds of shallow and
deep burrows present at any one time, the owls had an
abundant supply of burrow opportunities.

A total of 59 pairs nested in natural burrows located
in two natural and two disturbed areas on the campus of
NMSU. No pair was used more than once in this study.
We did, however, include different pairs which used the
same burrow in different years and we have repeated
data for some pairs which switched burrows from year to
year. We define a “natural” burrow as any existing cavity
either above or below ground that had not been modi-
fied by us. We do not mean to suggest that a “natural”
burrow was located in a natural (i.e., undisturbed) set-
ting, although some of the burrows used in this study did
fall into this category.

We used two natural areas in our study. The first con-
sisted of an abandoned landfill (4.1 ha) inoperative for

Sept 1998

Burrowing Owl Nest Site Use

235

at least 10 yr prior to this study. Its initial contents had
been covered with soil and its base was overgrown with
native vegetation. The second area consisted of an earth-
en dam at one end of a flood control basin (8 ha, Bo-
telho and Arrowood 1996). Both natural areas were lo-
cated in the remote southeast part of campus and rarely
visited by people. They were actually large depressions
surrounded by cliffs up to 10 m high. Burrows were lo-
cated in several different sites including flat ground,
above ground and 3-10 m up in the sides or at the base
of cliff walls. Vegetation consisted of typical arid Chihua-
huan Desert vegetation, dominated by Creosote Bush
( Larra tridentata ) and Mesquite (Prosopsis spp.) .

Disturbed areas consisted of the university quadrangle
(quad) and football stadium (stadium). Owls nested
among closely-spaced buildings separated by walkways,
lawns, parking lots and other buildings on the quad and
at the top of small hills and at the base of cement walls
behind each endzone in the stadium (Botelho and Ar-
rowood 1996). Pairs typically used burrows under cement
walkways or curbs, especially those in the vicinity of street
lights or other types of artificial lighting. Soil was rich
loamy topsoil; more durable burrows occurred in this soil
type because it was less susceptible to collapse during
rainstorms. Vegetation consisted of irrigated cultivated
grass on well-manicured lawns with some trees and
shrubs (Botelho and Arrowood 1996). In addition, bur-
rows were located at the base of light posts in a large
parking lot and in large pipes above ground.

.Some burrows in cliff walls in natural areas were locat-
ed high off the ground, but within human reach from
the top of the cliff. Burrows in cliff walls had very little
space at their entrances for nestlings to congregate dur-
ing feedings. As a consequence, nestlings sometimes fell
from the front of their burrows and either found shelter
in a burrow close to the ground or fell victim to preda-
tion. In disturbed areas, burrows in cliff walls were similar
to those in natural areas. Burrows in cliff walls in dis-
turbed areas were only available behind each endzone
inside the stadium. Owls which used these burrows
perched on the tops of bleachers and on fences.

Burrows at the base of cliff walls in natural areas were
dug at ground level into the sides of cliff walls. Because
of their location at ground level, these burrows had more
space at their entrances for nestlings to congregate dur-
ing feedings and there was no danger of nestlings falling
from their burrows. Because of their close proximity to
both the ground and a cliff wall, these burrows could be
blocked when loose dirt from the cliff poured over their
entrances during heavy rains. Burrows were located be-
neath stone walls behind each endzone in the stadium
and at the base of buildings in the quad in disturbed
areas. Unlike burrows at the base of cliff walls in natural
areas, in disturbed areas some burrows were dug under
concrete sidewalks and abutments. Owls which used bur-
rows at the base of cliff walls used buildings or cement
walls as perches.

Burrows in flat ground in natural areas were dug di-
rectly into the desert floor and were surrounded by
sparse vegetation. These burrows had few elevated perch
sites and were resistant to erosion but lacked a cliff face
which may have increased vulnerability to predation be-
cause predators could approach the burrow from all di-

rections. Burrows in flat ground, however, had ample
space at their entrances for nestlings to congregate dur-
ing feedings without the danger of nestlings falling from
the burrow. Burrows in flat ground in disturbed areas
were located on lawns (often at the base of chain link
fences), or under curbs. Owls regularly used man-made
perch sites (e.g., fences, walls, and buildings) when nest-
ing in these burrows.

All above ground nesting attempts occurred in dis-
turbed areas. These nest sites consisted of large metal
pipes located on flat ground. In one case a pair nested
in a drainpipe located in the side of a building.

We constructed 24 artificial burrows. Eight of 24 nat-
ural burrows were situated in such a way that we could
replace them with artificial burrows. The remaining 16
burrows were left in place and artificial burrows were in-
stalled in the vicinity of and adjacent to them. We re-
placed natural burrows with artificial burrows in winter
when breeding was not in progress. Natural burrows were
excavated in the evening after we placed a one-way door
(this door allowed owls to leave the burrow but not reen-
ter) over the burrow entrance for at least 48 h to ensure
that no owls were present inside the burrow during ex-
cavation. We oriented the chambers and tunnels of our
artificial burrows as close as possible to that of original
burrows.

Artificial burrows were completely self contained and
consisted of a nesting chamber (a 19 1 covered plastic
bucket) located at the end of a tunnel made of two 2 5
m X 10 cm PVC pipes (with 2 cm holes drilled every 6
cm for drainage) connected by a right angle PVC con-
nector. A single 10 cm hole was cut into the side of the
plastic bucket about two cm from the bottom to allow
insertion of the PVC tunnel pipe. A 10 cm hole also was
cut into the cover of the bucket; we could insert a hand
through this hole to gain access to the nest for weighing
and measuring nestlings without removing the entire
bucket lid. The cover hole was capped with a PVC lid
We drilled three to four holes (each 2 cm in diameter)
in the bottom of the plastic bucket for drainage. During
installation we placed dirt in the bottom of the bucket
and inside the tunnel pipes. To avoid human distur-
bance, the entire burrow (including covers) was buried.
Artificial burrows were not buried under mounds as in
Trullio (1995) and Collins and Landry (1977) because
some of our early, more obvious artificial burrows were
stolen (probably for the PVC pipe) before any owls had
begun to use them.

All of the owl pairs used in this study and any young
they produced were trapped using either a cage and one-
way door trap (Banuelos 1993, PVC tube trap (Botelho
and Arrowood 1995), or captured by hand in the artifi-
cial nest cavity. Captured owls were banded with USGS
aluminum bands and a unique combination of colored
plastic bands. We insured that all nestlings were captured
by repeated observation and trapping at each burrow un-
til all nestlings were marked on three consecutive obser-
vation periods. Because we did not excavate natural bur-
rows, we cannot rule out that some nestlings may have
gone undetected. We feel, however, that undetected nest-
lings, if they did occur, were rare.

Because our data is nonnormal, we used nonparamet-

236

Botelho and Arrowood

Vol. 32, No. 3

Table 1 . Types of burrows used by nesting Burrowing Owls and the numbers of nestlings they produced in undis-
turbed and disturbed areas.

Undisturbed Areas

Disturbed Areas

No. Pairs

No. Nestlings Nestlings/Pair

No. Pairs

No. Nestlings Nestlings/Pair

Burrow Type

(%)

(%)

(x ± 1 SE)

(%)

(%)

(x ± 1 SE)

Vertical cliff

15 (47)

25 (45)

1.7 ± 1.8

4(15)

0(0)

0

Base of cliff

11 (34)

14 (25)

1.3 ± 1.7

7 (26)

25 (38)

3.6 ± 2.1

Flat ground

6(19)

17 (30)

2.8 ± 4.1

12 (44)

41 (62)

3.4 ± 1.8

Above ground

0(0)

0(0)

0

4(15)

0(0)

0

Total

32

87

27

66

ric statistics. Our alpha level for significance is 0.05.
Means are reported with standard errors (x ± 1 SE).

Results

In undisturbed areas, pairs used burrows in cliff
walls more often than burrows on flat ground but
the difference was not significant (x 2 = 3.80, df =

2, 0.10 > P > 0.02; Table 1). In disturbed areas,
pairs used burrows in flat ground more than bur-
rows in cliff walls and above ground but, here also,
the difference was not significant (x 2 = 6.33, df =

3, 0.10 > P > 0.05). Burrows in flat ground in
disturbed areas were very common and potential
sites in cliff walls were less common than in natural
areas because they only occurred in the stadium
and banks of the irrigation canal. However, there
were numerous burrows in the stadium and along
the canal that were dug by squirrels. Burrows in
cliff walls in disturbed areas that appeared suitable
for nesting were not used. Sites at the base of cliffs
were common under the concrete edges of build-
ings and walls. Even though burrows in culverts
and pipes appeared to be common throughout dis-

Table 2. Types of burrow sites used by resident and mi-
grant male Burrowing Owls in undisturbed and disturbed
areas.

Burrow

Location

Undisturbed Areas

Disturbed Areas

No.

Resi-

dent

Males

(%)

No.

Migrant

Males

(%)

No.

Resi-

dent

Males

(%)

No.

Migrant

Males

(%)

Vertical cliff

11 (52)

5 (100)

0(0)

1 (10)

Base of cliff

8 (38)

0(0)

6 (43)

1 (10)

Flat ground

2(10)

0(0)

7 (50)

7 (70)

Above ground

0(0)

0(0)

1 (7)

1 UO)

turbed areas, only four pairs utilized them. These
nesting attempts failed.

In undisturbed areas, pairs that used burrows in
flat ground produced significantly more nestlings
than pairs in the other types of burrows (Kruskal-
Wallis test, F = 13.52, df = 2, P< 0.005; Table 1).
In disturbed areas, pairs that nested in burrows in
cliff wall and above ground sites produced no nest-
lings. Pairs which used burrows at the base of cliff
walls and in flat ground produced significantly
more nestlings than their counterparts in natural
areas (F = 11.40, df = 3, P < 0.005; Table 1).

In undisturbed areas, the distribution of breed-
ing resident males was more equal among available
burrow types than was the distribution of breeding
migrant males (Table 2). Migrant males exclusively
used burrows in cliff walls although the highest
percentage of resident males also used burrows in
cliff walls. The lowest percentage of males in nat-
ural areas used burrows in flat ground. In contrast
to undisturbed areas, very few migrant and resi-
dent males used burrows in cliff walls in disturbed
areas. Instead, they mostly used burrows in flat
ground with resident males using burrows at the
base of cliff walls more often than migrants.

Among migrants that bred in 1993 and returned
to breed in 1994 (N = 15), 60% changed burrow
site types with 67% of males and 50% of females
using burrows in different site types in 1994 (Table
3). Among those migrants that bred in 1994 and
returned to breed in 1995 (N = 12), 58% changed
burrow site types. Males that returned in 1995 over-
whelmingly used burrows of the same site type
(75%), the reverse of what happened in 1994. Sev-
enty-one percent of females, however, used bur-
rows in sites different from those used in 1994. Use
of same (N — 11) and different (N — 16) burrow
types over both years by males and females did not
differ significantly (x 2 — 0.926, df = 1, P > 0.05).

Sept 1998

Burrowing Owl Nest Site Use

237

Table 3. Burrow switching by migrant male and female Burrowing Owls between 1993-94 and 1994—95.

Burrow Type

1993-1994

1994-1995

Males

Females

Total

Males

Females

Total

From base of cliff to base of cliff

1

0

1

0

0

0

to cliff wall

2

1

3

0

1

1

to flat ground

0

0

0

0

0

0

to artificial burrow

0

0

0

0

0

0

From cliff wall to cliff wall

1

1

2

1

0

1

to base of cliff

1

1

2

0

1

1

to flat ground

1

0

1

0

1

1

to artificial burrow

0

1

1

0

0

0

From flat ground to flat ground

1

1

2

2

2

4

to base of cliff

1

0

1

0

0

0

to cliff wall

1

0

1

0

1

1

to artificial burrow

0

0

0

0

0

0

From artificial burrow to artificial burrow

0

1

1

0

0

0

to flat ground

0

0

0

1

2

3

to base of cliff

0

0

0

0

0

0

to cliff wall

0

0

0

0

0

0

Total nestings

9

6

15

4

7

12

Total number switches between years

6

3

9

1

5

7

Among those individuals that used burrows on
different sites the following year, two (one male
and one female) moved from a burrow in or at the
base of a cliff wall to a burrow in flat ground (Ta-
ble 3). Of the remainder, 18% moved from bur-
rows in cliff walls to burrows at the base of cliffs.
Only 1 1 % of the owls that moved from burrows in
flat ground to those in or at the base of cliff walls
moved to a different site type.

The number of nestlings produced by pairs that
bred in our study area in one year and returned
to breed again the following year did not differ
significantly regardless of burrow type used (Wil-
coxon Signed Ranks Test, T = — 13, P = 0.343, N
— 6 for males and T = 10, P = 0.0635, N = 4 for
females). On average, females that switched bur-
row types from one year to the next, produced
more nestlings in the burrows they left rather than
in their new burrows (Table 4). Males switching
burrows from one year to the next produced equal
numbers of nestlings in the two sites.

Eight pairs which nested in artificial burrows
produced an average of 8.3 ± 3.5 eggs per pair
(Table 5). The number of nestlings ranged from
0-8 (x = 3.5 ± 2.9). Clutches in all but two artifi-
cial burrows partially hatched; the two clutches
which failed to hatch were abandoned prior to
hatching because the mates died. Of the 28 nest-

lings produced in artificial burrows, only 12 or
43% fledged. In all but one burrow where all nest-
lings hatched synchronously, one nestling hatched
much later (2—4 d) than the rest and always died.
These smaller nestlings usually disappeared from
the burrow overnight either through predation or
cannibalism. One female was videotaped feeding
her youngest nestling to the surviving young. Old-
er nesdings which failed to fledge also disappeared
quickly from burrows without a trace. Owls in ar-
tificial burrows produced an average of 3.5 ± 2.9
nesdings (N — 8 nests) which is significandy higher
than production in natural burrows (2.2 ± 1.9
nesdings, N = 59 nests; Mann-Whitney U test, Z =
— 2.07, N = 67, P < 0.02). When pairs abandoning
their burrows prior to hatching were removed
from the analysis, owls which used artificial bur-
rows still produced significandy more nestlings (x
= 3.3 ±1.3 nesdings for natural and x = 4.7 ±
2.3 nesdings for artificial burrows, Z = —1.68, N =
44, P < 0.05). However, when we compared the
number of fledglings produced by the two types of
burrows, the number produced by pairs in natural
burrows was significandy greater for natural than
for artificial burrows (x = 1.9 ± 1.9 nesdings for
natural and x = 1.5 ± 1.5 nestlings for artificial
burrows, Z = —2.81, N = 67, P < 0.003). The av-
erage number of fledglings produced by pairs in

238

Botelho and Arrowood

Vol. 32, No. 3

Table 4. Number of nestlings produced by migrant Burrowing Owls that returned to the same and different burrow
types between 1993—94 and 1994—95.

Males

Females

Current

Previous

Current

Previous

Year

Year

Total

Year

Year

Total

From base of cliff to base of cliff

0

4

4

0

0

0

to vertical cliff

1

6

7

6

7

13

to flat ground

0

0

0

0

0

0

to artificial burrow

0

0

0

0

0

0

From vertical cliff to vertical cliff

1

0

1

0

0

0

to base of cliff

2

1

3

0

4

4

to flat surface

0

0

0

3

8

11

to artificial burrow

0

0

0

3

3

6

From flat surface to flat surface

3

3

6

6

6

12

to base of cliff

3

3

6

0

0

0

to vertical cliff

3

2

5

3

3

6

to artificial burrow

0

0

0

0

0

0

From artificial to artificial burrow

0

0

0

3

3

6

to flat ground

6

7

13

9

10

19

to base of cliff

0

0

0

0

0

0

to cliff wall

0

0

0

0

0

0

Average for migrants returning to
same type of burrow
Average for migrants returning to

1.0 ± 1.4

1.8 ± 2.1

2.3 ± 2.9

2.3 ± 2.9

different type of burrow

1.3 ± 1.9

1.6 ± 2.5

2.0 ± 3.0

2.9 ± 3.6

Table 5. Hatching and fledging success of eight pairs of
Burrowing Owls nesting in artificial burrows form 1993-
95. Fledglings are defined as young that were observed
flying in their natal territories.

Pair

Clutch

Size

No. Eggs
Hatching
(%)

No.

Fledglings

(%)

1

6

0(0)

0 (0) a

2

7

0(0)

0 (0) a

3

7

3 (43)

2(67)

4

7

3 (27)

2(67)

5

8

7(64)

4(57)

6

9

3 (33)

0(0)

7

11

4 (36)

3(75)

8

11

8 (73)

1 (13)

Total

66

28

12

Mean

8.3

3.5

1.5

SE

1.9

2.9

1.5

Total b

53

28

12

Mean b

8.8

4.6

2.0

SE b

1.8

2.3

1.4

a Nest abandoned.

b Abandoned burrows have been omitted.

natural burrows where adults did not abandon was
2.9 ± 1.5, significantly higher than that produced
by pairs in artificial burrows (2 0 ±1.4 fledglings,
Z = -2.97, N = 43, P < 0.002).

Discussion

Our prediction that males would use burrows lo-
cated in sites with high elevation and low grass cov-
er in cliffs more often due to decreased predation
was supported for resident and migrant males in
natural areas but not disturbed areas. Only one
male in a disturbed area used a burrow in a cliff
wall despite the apparent availability of cliff sites.
One possible reason for not using burrows in cliffs
may have been that they were located in the ter-
ritories of other males not using cliff burrows. An-
other reason may have been the possible high mor-
tality of fledglings when they fell from their
burrows although this seemed unlikely because
pairs using burrows in cliff walls were as productive
as pairs that used burrows at the base of cliffs in
natural areas.

Burrows in cliff walls appeared to be safer from
predators because of their height and approach by

Sept 1998

Burrowing Owl Nest Site Use

239

predators was possible in only one direction. Mac-
Cracken et al. (1985) and Green and Anthony
(1989) have shown that Burrowing Owls use bur-
rows located in sites on mounds of dirt with low
grass cover, but our study shows for the first time
that Burrowing Owls can also use sites associated
with cliffs. The presence of depressions surround-
ed by steep cliffs coupled with the tendency of rock
squirrels and rabbits to colonize these areas and
dig holes in and at the bases of the cliff walls can
provide an unusual type of nest site for Burrowing
Owls. The only other case we know of where Bur-
rowing Owls have been shown to use burrows in or
at the bases of cliffs is in Albuquerque, New Mex-
ico, 155 km north of our study site (Kendall pers.
comm.).

Our prediction that pairs should use burrows at
the bases of cliffs more often than on flat ground
was supported for undisturbed areas but not dis-
turbed areas. Also, pairs that nested at the bases of
cliffs in undisturbed areas produced fewer nest-
lings on average than pairs which nested either in
cliff walls or in flat ground. In disturbed areas,
however, pairs that nested in the bases of cliffs pro-
duced more nestlings than all other burrow types.
Larger broods in disturbed areas may have been
due to increased prey availability attributed to ar-
tificial lighting, especially in the stadium. Also, the
larger amount of space at the entrances to burrows
in the bases of cliffs better accommodated larger
broods and restricted the approach routes of pred-
ators. Decreased risk of nestling predation in dis-
turbed areas may have contributed to this trend
but we have no data on the effect of predation on
the reproductive success of this population.

Most studies of Burrowing Owls have found
them occupying burrows in relatively flat ground
although some elevation near the burrow is im-
portant. Burrowing Owls in Oregon (Green and
Anthony 1989), in South Dakota (MacCracken et
al. 1985), and in Colorado (Plump ton and Lutz
1993) preferred burrows on high ground with low
mean shrub volume or low grass cover, possibly to
gain an elevated unobstructed view. Females in this
study monitored their surroundings from an ele-
vated site with a clear view and gave alarm calls to
which the nestlings responded by running into the
burrow. For flightless nestlings to respond quickly,
females must produce alarm calls well in advance
of a predator’s approach making a clear view of
the area surrounding the burrow important. In un-
disturbed areas, pairs using burrows in flat ground

were most productive; in disturbed areas such pairs
produced only slightly fewer nestlings than pairs at
the bases of cliffs.

The lower overall productivity of pairs in undis-
turbed areas may have been due in part to preda-
tion. A pair of Barn Owls ( Tyto alba ) used a burrow
located in a cliff wall in the landfill only 2-3 m
away from an occupied Burrowing Owl nest and
within easy striking distance of up to 13 other
nests. Burrowing Owls actively mobbed the Barn
Owls as they left their burrow but we are unaware
of any predation by the Barn Owls on Burrowing
Owls. Also, lack of an available food supply close
to their burrows may have lowered productivity, es-
pecially among those pairs which nested in areas
without the benefit of insects attracted by artificial
lighting. Violent storms, which passed through the
study area in late summer, may have also resulted
in the deaths of small nestlings caught outside
their burrows.

Most females and males which returned to a dif-
ferent burrow type from one year to the next pro-
duced fewer nestlings in their second breeding at-
tempt than in their first. Decreased reproductive
success in new burrows may explain why owls
switched burrows infrequently and never accepted
artificial burrows installed in the vicinity of their
nesting burrow.

An average hatching and fledging success of
42% and 18%, respectively, by pairs which nested
in artificial burrows was lower than that found in
other studies where artificial burrows have been
used (Landry 1979, Olenick 1987). Pairs that nest-
ed in artificial burrows produced significantly
more nestlings than pairs that used natural bur-
rows even if pairs that failed to hatch any eggs were
included in the analysis. In fact, pairs which nested
in artificial burrows produced almost one nestling
more on average than their counterparts in natural
burrows. The opposite was true for fledglings. Pairs
that nested in natural burrows produced signifi-
cantly more fledglings than pairs that used artifi-
cial burrows regardless of whether pairs failing to
hatch any eggs were included in the analysis. After
removing pairs that failed to hatch any eggs, pairs
nesting in natural burrows produced almost one
more fledgling on average than pairs which used
artificial burrows. These results were unexpected
because we thought the antipredator advantages of
artificial burrows would enhance fledgling produc-
tion. Nestlings in artificial burrows were captured
inside the nest chamber and weighed three to four

240

Botelho and Arrowood

Vol. 32, No. 3

times per week during the nestling period to de-
termine growth rates for another study. Artificial
burrows, however, were not disturbed once clutch-
es were complete and incubation began. Thus, one
reason for the observed trend in nestling and
fledgling production by pairs in natural and arti-
ficial burrows could have been human disturbance
during the nestling period and the lack of it during
the incubation period.

We suggest that conservation plans for Burrow-
ing Owls involving the use of artificial burrows in
areas without natural nesting burrows should con-
sider the characteristics of burrow sites previously
used by the owls for nesting. Because some of the
owls that switched burrows from year to year suf-
fered decreased nesting success, there may be se-
lection against year to year movement among bur-
rows. Given their nest site fidelity (Haug et al.
1993), disturbance of nest sites could have a dev-
astating impact on Burrowing Owl populations,
even if artificial burrows are installed nearby.

This study demonstrates the importance of in-
stalling artificial burrows in sites most favored by
nesting pairs. Owls in this study nesting in undis-
turbed areas used burrows located in and at the
bases of cliff walls where artificial burrows could
not be installed. On average, pairs in artificial bur-
rows produced significantly more nestlings than
pairs in natural burrows, indicating that artificial
burrows did not contribute to decreased nestling
productivity. Furthermore, human disturbance
may have played a role in lower fledgling produc-
tion by pairs in artificial burrows.

Acknowledgments

We are grateful to the physical plant at NMSU for al-
lowing access to our field sites on campus. Dan Howard
provided helpful comments on the manuscript. Dennis
Clason provided help with statistical analyses. Betsy Bo-
telho helped with field work. We thank Marc Bechard,
Jim Belthoff, Elizabeth Haug, and Paul James for help
with revision of this manuscript and Phil Zwank for ar-
ranging the loan of night vision equipment from the
USFWS Wildlife Cooperative Research Unit on the cam-
pus of NMSU. This research was supported by the NMSU
Department of Biology, a Frank M. Chapman Award from

the American Museum of Natural History, and grants

from the New Mexico Council for Higher Education and

the Chicago Zoological Society to ESB.

Literature Cited

Banuelos, G.H.T. 1993. An alternative trapping method
for Burrowing Owls . /. Raptor Res. 27:85-86.

Botelho. E.S, and PC. Arrowood. 1995. A novel, sim-
ple, safe and effective trap for Burrowing Owls and
other fossorial animals./. Field Ornithol. 66:380-384.

and . 1996. Nesting success of western Bur-
rowing Owls ( Speotyto cunicularia hypugaea) on natural
and human-altered sites. Pages 61—68 in D.M. Bird,
D.J. Varland andJ.J. Negro [Eds.], Raptors in human
landscapes. Academic Press, London, U.K.

Collins, C.T. and R.E. Landry. 1977. Artificial nest bur-
rows for Burrowing Owls. N. Am. Bird Bander 2:151-
154.

Coulombe, H.N. 1971. Behavior and population ecology
of the Burrowing Owl ( Speotyto cunicularia ) in the Im-
perial Valley of California. Condor 73:162-176.

Green, G.A. and R.G. Anthony. 1989. Nesting success
and habitat relationships of Burrowing Owls in the
Columbia Basin, Oregon. Condor 91:347-354.

Haug, E.A., B.A. Millsap and M.S. Martell. 1993. Bur-
rowing Owl ( Speotyto cunicularia). Pages 1-20 in A.F.
Poole and F. Gill [Eds.], The Birds of North America,
No. 61. Academy of Natural Sciences, Philadelphia,
PA U.S.A. and the American Ornithologists’ Union,
Washington, DC U.S.A.

Landry, R.E. 1979. Growth and development of the Bur-
rowing Owl, Athene cunicularia. M.A. thesis, California
State University, Long Beach, CA U.S.A.

MacCracken, J.G., D.W. Uresk and R.M. Hansen. 1985.
Vegetation and soils of Burrowing Owl nest sites in
Conata Basin, South Dakota. Condor 87:152-154.

Olenick, B. 1987. Reproductive success of Burrowing
Owls using artificial burrows in southeast Idaho. Eyas
10:38.

Plumpton, D.L. and S.R. Lutz. 1993. Nesting habitat use
by Burrowing Owls in Colorado./. Raptor Res. 27:175-
179.

Thomsen, L. 1971. Behavior and ecology of Burrowing
Owls on the Oakland Municipal Airport. Condor 73:
177-192.

Trullio, L.A. 1995. Passive relocation: a method to pre-
serve Burrowing Owls on disturbed sites. /. Field Or-
nithol. 66:99-106.

Received 3 August 1996; accepted 8 May 1998

J. Raptor Res. 32(3):241-245
© 1998 The Raptor Research Foundation, Inc.

BREEDING-SEASON FOOD HABITS OF
BURROWING OWLS ( ATHENE CUNICULARIA) IN
SOUTHWESTERN DOMINICAN REPUBLIC

James W. Wiley 1

Biological Research Division, U.S. Geological Survey, Reston, VA 22092 U.S.A.

Abstract. — Diet data from 20 Burrowing Owl ( Athene cunicularia) nests were collected in southwestern
Dominican Republic in 1976, 1982, and 1996. Invertebrates (53.3%) comprised the most numerous
prey items (N = 396) delivered to nests by adult owls, but vertebrates (46.7%) were much better rep-
resented than in other studies of Burrowing Owl diet. Among vertebrates, birds (28.3% of all items)
and reptiles (14.9%) were most important, whereas mammals (1.0%) and amphibians (2.5%) were less
commonly delivered to nests. Vertebrates, however, comprised more than twice (69.2%) of the total
biomass as invertebrates (30.8%), with birds (50.4%) and reptiles (12.8%) the most important of the
vertebrate prey classes. A positive relationship was observed between bird species abundance and num-
ber of individuals taken as prey by Burrowing Owls.

Key WORDS: Athene cunicularia; Burrowing Owl, diet, Dominican Republic, ecology.

Habitos alimenticios durante la epoca reproductiva de Athene cunicularia en el suroeste de la Republica
Dominicana

Resumen. — Datos de la dieta de 20 nidos de Athene cunicularia fueron colectados en el suroeste de la
Republica Dominicana durante 1976, 1982 y 1996. Los invertebrados (53.3%), fueron los items mas
numerosos ( N = 396) entregados en los nidos por los buhos adultos, pero los vertebrados (46.7%)
fueron mucho mejor representados que en otros estudios sobre habitos alimenticios de Athene cunicu-
laria. Entre los vertebrados, las aves (28.3% de todos los items) y los reptiles (14.9 %) fueron los mas
importantes mientras que los marmferos (1.0%) y los anfibios (2.5%) fueron menos comunes. Los
vertebrados sin embargo, constituyeron mas del doble (69.2%) del total de la biomasa. Los invertebrados
(30.8%), aves (50.4%) y los reptiles (12.8%) fueron las clases mas importantes de presas. Existio una
relation positiva entre la abundancia de especies de aves y el numero de individuos capturados como
presas por Athene cunicularia.

[Traduction de Cesar Marquez]

Burrowing Owls ( Athene cunicularia) occur from
western Canada south through the western U.S.,
Mexico, Central and South America, and irregu-
larly in Florida and the West Indies. In the Carib-
bean islands they presendy inhabit most of the Ba-
hamas; Cuba, including several cays of the
Sabana-Camaguey Archipielago and the Isla de la
Juventud (previously Isla de Pinos) ; and the west-
ern half of Hispaniola, including Gonave and Bea-
ta islands (AOU 1998). The species formerly oc-
curred throughout the Greater Antilles and several
of the Lesser Antilles. Populations disappeared
from Jamaica (Olson and Steadman 1977, Morgan

1 Present Address: Grambling Cooperative Wildlife Pro-
ject, P.O. Box 841, Grambling State University, Gram-
bling, LA 71245 U.S.A.

1993), the Cayman Islands (Morgan 1977, 1994),
and Puerto Rico (Pregill and Olson 1981), possibly
as a result of changing climate and habitat condi-
tions, and predation by introduced mammals (Pre-
gill 1981, Pregill and Olson 1981, Wiley 1986a). In
the Lesser Antilles, Burrowing Owls recendy oc-
curred in St. Kitts, Nevis, Antigua, Redonda, and
Marie Galante (AOU 1983). However, the two rac-
es endemic to the Lesser Antilles are thought to
be extinct: A. c. guadeloupensis from Marie Galante,
and A. c. amaura from Antigua, Nevis, Redonda,
and St. Kitts, again partly as a result of predation
by exotic animals (Greenway 1967).

In spite of the losses of several populations from
former ranges and current concern for the species’
survival in the Caribbean, little has been reported
on the biology of Burrowing Owls in the region

241

242

Wiley

Vol. 32, No. 3

Table 1. Prey brought to nests by breeding Burrowing Owls, foothills of the Sierra de Bahoruco, southwestern
Dominican Republic, 1976, 1982, and 1996.

Prey Species

Number of Prey Items (%)

Observed
Brought
to Nest Prey

(1976) Remains Pellets Total

Biomass (%)

. Mean
Prey
Bio-
mass

(g) Total

Mammals

House mouse Mus musculus

1(2.8)

1(0.8)

1(0-4)

3(0.8)

21.0

63.0(2.1)

Black rat Rattus rattus (young)

1(0.4)

1(0.3)

80.0

80.0(2.6)

Total mammals

1(2.8)

1(0.8)

2(0.9)

4(1.0)

143.0(4.7)

Birds

Common Ground-Dove Columbina pas-

2(1.6)

2(0.9)

4(1.0)

28.9

115.6(3.8)

senna

Antillean Mango Anthracothorax domin-

3(8.3)

5(3.9)

7(3.0)

15(3.8)

5.2

78.0(2.6)

% C ' Lf/S

Hispaniolan Emerald Chlorostilbon swain-

1(0.8)

1(0.3)

4.6

4.6(0. 2)

sonii

Broad-billed Tody Todus subulatus

2(5.6)

6(4.7)

10(4.3)

18(4.6)

8.3

149.4(4.9)

Unidentified tody Todus sp.

2(1.6)

3(1.3)

5(1.3)

8.0

40.0(1.3)

Stolid Flycatcher Myiarchus stolidus

1(2.8)

1(0.4)

2(0.5)

24.1

48.2(1.6)

Hispaniolan Pewee Contopus hispani-

2(1.6)

2(0.5)

11.5

23.0(0.8)

olensis

Northern Mockingbird Mimus polyglottos

Adult

1(0.4)

1(0.3)

42.3

42.3(1.4)

Fledgling

1(2.8)

1(0.8)

2(0.5)

35.0

70.0(2.3)

Red-legged Thrush Turdus plumbeus

1(0.8)

1(0.4)

2(0.5)

74.0

148.0(4.9)

Black-whiskered Vireo Vireo altiloquus

3(2.3)

3(0.8)

19.1

57.3(1.9)

Flat-billed Vireo Vireo nanus

1(0.4)

1(0.3)

10.7

10.7(0.4)

American Redstart Setophaga americana

1(2.8)

1(0.4)

2(0.5)

8.7

17.4(0.6)

Green-tailed Warbler Microligea

1(2.8)

5(3.9)

9(3.9)

15(3.8)

12.5

187.5(6.2)

palustris

Ovenbird Seiurus aurocapillus

1(0.4)

1(0.3)

18.7

18.7(0.6)

Bananaquit Coereba flaveola

1(2.8)

4(3.1)

12(5.2)

17(4.3)

8.7

147.9(4.9)

Black-crowned Palm-Tanager Phaenico-

1(0.8)

6(2.6)

7(1.8)

30.5

213.5(7.0)

philus palmarum

Greater Antillean Bullfinch Loxigilla vio-

2(1.6)

2(0.5)

22.3

44.6(1.5)

lacea

Yellow-faced Grassquit Tiaris olivacea

1(0.8)

1(0.4)

2(0.5)

8.0

16.0(0.5)

Unidentified bird

10(4.3)

10(2.5)

9.6

96.0(3.2)

Total birds

10(27.8)

36(28.1)

66(28.4)

112(28.3)

1528.7(50.4)

Amphibians

Eleutherodactylus abbotti

1(2.8)

3(2.3)

6(2.6)

10(2.5)

4.0

40.0(1.3)

Total amphibians

1(2.8)

3(2.3)

6(2.6)

10(2.5)

40.0(1.3)

Reptiles

Ameiva chrysolaema

1(2.8)

5(2.2)

6(1.5)

8.9

53.4(1.8)

Anolis distichus

2(5.6)

4(3.1)

15(6.5)

21(5.3)

6.7

140.7(4.6)

Anolis semilineatus

3(8.3)

7(5.5)

9(3.9)

19(4.8)

6.4

121.6(4.0)

Sphaerodactylus cryphius

1(2.8)

4(3.1)

5(2.2)

10(2.5)

4.3

43.0(1.4)

Typhlops hectus

1(0.8)

1(0.3)

9.0

9. 0(0. 3)

Uromacer frenatus

1(0.8)

1(0.4)

2(0.5)

10.5

21.0(0.7)

Total reptiles

7(19.4)

17(13.3)

35(15.1)

59(14.9)

388.7(12.8)

Total vertebrates

19(52.7)

57(44.5)

109(47.0)

185(46.7)

2100.4(69.2)

Sept 1998

Burrowing Owl Diet

243

Table 1. Continued.

Biomass (%)

Number of Prey Items (%)

Prey Species

Observed
Brought
to Nest
(1976)

Prey

Remains

Pellets

Total

Prey

Bio-

mass

(G)

Total

Invertebrates
unidentified locusts

5(13.9)

21(16.4)

38(16.4)

64(16.2)

2.0

128.0(4.2)

unidentified beetles

8(22.2)

28(21.9)

57(24.6)

93(23.5)

3.1

288.3(9.5)

unidentified tarantulas

3(8.3)

17(13.3)

19(8.2)

39(9.9)

9.7

378.3(12.5)

unidentified centipedes

1(2.8)

5(3.9)

9(3.9)

15(3.8)

9.3

139.5(4.6)

Total invertebrates

17(47.2)

71(55.5)

123(53.0)

211(53.3)

934.1(30.8)

Totals

36

128

232

396

3034.5(100.0)

(Wiley 1986a, 1986b). Here, I report on the high
incidence of avian prey I observed in the diet of
Burrowing Owls in southwestern Dominican Re-
public based on direct observations, prey remains,
and regurgitated pellets collected at nests.

Methods

Data were collected in southwestern Dominican Re-
public from March-June 1976, June-July 1982, and
March 1996 from occupied Burrowing Owl nests along
the lower slopes of the Sierra de Bahoruco, from Cruce
de Limon near Lago Enriquillo (elevation 30 m) south
to El Naranjo west of Puerto Escondido (elevation 350
m). The area is in the subtropical dry woodland zone
(Union Panamericana 1967) characterized by acacia-cac-
tus woodland that becomes more luxuriant with increas-
ing elevation (Durland 1922). Typical vegetation include
cacti (guasabara pilotera Opuntia antillana, cagiiey Neoab-
bottia paniculata, cayuco Pilosocereus polygonus), palmera
yarey ( Copernicia berteroana) , bayahonda ( Prosopis juliflo-
ra), Capparis spp., baitoa ( Phyllostylon brasiliensis) , aroma
(Acacia farnesiana) , guayacan ( Guajacum officinale), guay-
acancillo ( Guajacum sanctum), almacigo ( Bursera simaru-
ba), guano ( Coccothrinax argentata), and doncella ( Byrson -
ima ludda) . Annual rainfall averages about 455 mm, with
peaks in January, April-May, and August-November.

Owls were common in the area and nested wherever
suitable substrate was available. I systematically surveyed
the owl population in my study area by foot on several
occasions, paying special attention to the presence of
nest sites. All possible nests were revisited a minimum of
four times to confirm their occupancy. In 1976, one 9-ha
study plot contained 18 occupied owl nests (x = 1 nesting
pair per 0.5 ha), concentrated in three colonies of 3, 8,
and 7 pairs (mean nearest-neighbor distance = 22.5 ±
6.7 m SD; range = 15-35 m).

All old pellets (i.e., regurgitated castings) and prey re-
mains were cleared from study nests the day before I be-
gan each period of data collection. I spent 42 hr observ-
ing two occupied nests from blinds placed 4 m from the
burrows on 14-15 April (1600-2300 H), 15-16 April

(1930-0600 H), 30 April-1 May (2200-0530 H), 1-2 May
(1600-2300 H), and 15-16 May (1630-0230 H), 1976.
Although observations are preferable to prey remains
and pellets for analyzing diets of birds of prey (Snyder
and Wiley 1976), prey remains and regurgitated pellets
do provide traditional materials for examination of rap-
tor food habits (Errington 1930). I collected remains and
pellets at eight nests in 1976, five in 1982, and five in
1996. Observations from blinds were made at different
nests than those from which I collected pellets and prey
remains. All data were collected during the nestling stage
before young emerged from nests. Prey remains and
items in pellets were identified by comparison with spec-
imens in the Museo Nacional de Historia Natural (Santo
Domingo; MNHN) , using a dissecting microscope when
needed. Prey biomass was determined from animals cap-
tured in or near my study area, from specimen data labels
in the MNHN, and from data provided by Anabelle Dod.
I surveyed bird populations for relative abundance of
species using fixed transects (Emlen 1971, 1977) and
mist nets on or adjacent to the Burrowing Owl study area.

Results

A total of 396 prey items was identified at least
to order (Table 1). Invertebrates (53.3% of all prey
items) made up the most numerous items brought
to nests by adult owls. Beetles (23.5% of all items),
locusts (16.2%), and tarantulas (9.9%) were the
most commonly delivered prey. Birds (28.3%) and
reptiles (14.9%) were also important items, where-
as mammals (1.0%) and amphibians (2.5%) made
up only a minor portion of the prey brought to
nests. Among birds, Broad-billed Todies ( Todus sub-
ulatusr, 4.6% of items), Bananaquits (Coereba flav-
eola\ 4.3%), Antillean Mangos ( Anthracothorax dom-
inicus; 3.8%), and Green-tailed Warblers ( Microligea
palustrisr, 3.8%) were the most common items.
Among reptiles, anole lizards ( Anolis distichus, 5.3%

244

Wiley

Vol. 32, No. 3

and A. semilineatus, 4.8%) were the most frequently
delivered species.

At the two nests watched from blinds in 1976,
the prey delivery rate averaged 0.86 ± 0.68 (±1
SD) items per hour. A young rat ( Rattus rattus ) ob-
served brought to a nest by an adult was estimated
to weigh about 80 g and represented the largest
prey item (53% of mean adult Burrowing Owl bio-
mass) in the sample. The largest avian prey items
delivered were Red-legged Thrushes ( Turdus plum-
beusr, x — 74 ± 2.53 g, N = 17; 49% of mean adult
owl mass).

Whereas invertebrates were the most numerous
items brought to nests by adult owls, vertebrates
comprised more than twice (69.2%) as much of
the total biomass as invertebrates (30.8%) in the
combined samples, with birds (50.4%) and reptiles
(12.8%) the most important of the vertebrate prey
classes. Mean weights of prey species ranged from
2.0-80.0 g, with vertebrates averaging 18.7 ± 19.0
g and invertebrates 6.0 ± 4.0 g. The greatest range
within prey categories occurred among birds,
which varied from 4.6-74.0 g.

I found a positive relationship between bird spe-
cies abundance and numbers of individuals taken
as prey by Burrowing Owls (Spearman Rank Cor-
relation, Z — 2.1, P = 0.04, N — 17). I did not
evaluate relative abundance of populations of oth-
er prey categories but, based on my casual obser-
vations, at least reptiles showed some degree of
correlation between abundance and numbers
taken as Burrowing Owl prey.

Discussion

Burrowing Owls in North America feed primar-
ily on invertebrates (90.9%) and only occasionally
eat mammals (6.9%), reptiles and amphibians
(2.0%), and birds (0.3%) (summarized in Earhart
and Johnson 1970 and Snyder and Wiley 1976).
The diet of Burrowing Owls in the West Indies has
been reported as consisting of small rodents, small
reptiles, frogs, and, especially, large insects, includ-
ing crickets, grasshoppers, and beetles (e.g., Bru-
denell-Bruce 1975, Campbell 1978, Dod 1978, Gar-
rido 1992, Kirkconnell et al. 1992). Danforth
(1929) reported the contents of two Burrowing
Owl stomachs, one of which contained mouse fur,
whereas the other contained beetle and centipede
parts. Abbott {in Wetmore and Swales 1931) found
one Burrowing Owl stomach contained one lizard,
one scorpion, one mouse, and several insects.

Although avian prey were particularly well-rep-

resented in my samples, only occasional mention
has been made of Burrowing Owls preying on
birds in the West Indies (Brudenell-Bruce 1975,
Dod 1978, Kirkconnell et al. 1992), including two
cases of cannibalism or scavenging (Abbott in Wet-
more and Swales 1931, Brudenell-Bruce 1975).
Gnatcatchers ( Polioptila sp.) were among the con-
tents of five Burrowing Owl stomachs collected by
Regalado (1975) in Cuba. “R.H.L.” (1883) report-
ed the remains of Black-cowled Orioles {Icterus
dominicensis ) , Greater Antillean Grackles ( Quiscalus
niger), and Common Ground-Doves {Columbine,
passerina) at Burrowing Owl nests in Haiti. In the
southern Bahama Islands, Buden (1974) noted a
high proportion of avian prey in Barn Owl {Tyto
alba) food remains compared with continental
samples. He suggested that these results reflected
low T er abundance of rodents on islands.

Generally, direct observations of prey delivered
by raptors to their nests reveal a higher proportion
of smaller and more delicate items, such as some
arthropods, than do examinations of prey remains
and regurgitated pellets (Snyder and Wiley 1976).
Thus, small items generally are underrepresented
in analyses of remains and pellets. However, the
proportion of invertebrates (47.2%) I observed
brought to the nests was slightly lower than that
represented by remains (55.5%) and pellet con-
tents (53.0%). Collectively, vertebrates (52.8%)
were observed brought to the nests as often as in-
vertebrates. The combined data from prey remains
and pellet contents revealed higher incidence of
invertebrates (53.9%), but vertebrates (46.1%)
were not far outnumbered by arthropods. Based
on prey biomass, however, vertebrates clearly rep-
resented a far more important food source than
did invertebrates during the breeding season.

All of the prey species brought to nests by Bur-
rowing Owls were common and, among birds, all
but the American Redstart {Setophaga americana)
and Ovenbird {Seiurus aurocapillus) are residents in
Hispaniola.

Acknowledgments

For their assistance in the field, I thank Beth Wiley and
Julio E. Cardona. Annabelle Dod, Jose A. Ottenwalder,
and Sixto J, Inchaustegui generously assisted with collec-
tions at the Museo Nacional de Historia Natural in Santo
Domingo. I thank James R. Belthoff and two anonymous
reviewers for their help in improving the manuscript.

Literature Cited

American Ornithologists’ Union. 1998. Check-list of
North American birds, 7th Ed. A.O.U., Washington,
DC U.S.A.

Sept 1998

Burrowing Owl Diet

245

Brudenell-Bruce, P.G.C. 1975. The birds of the Baha-
mas. New Providence and the Bahama Islands. Tap-
linger Publ. Co., New York, NYU.S.A.

Buden, D.W. 1974. Prey remains of Barn Owls in the
southern Bahama Islands, Wilson Bull 86:336-343.

Campbell, D.G. 1978. The ephemeral islands. A natural
history of the Bahamas. McMillan Education Limited,
London, U.K.

Danforth, S.T. 1929. Notes on the birds of Hispaniola.
Auk 46:358-375.

Dod, A.S. de. 1978. Aves de la Republica Dominicana.
Mus. Nac. Hist. Nat., Santo Domingo.

Durland, W.D. 1922. The forests of the Dominican Re-
public. Geogr. Rev. 1922:206-222.

Earhart, C.M. and N.K. Johnson. 1970. Size dimor-
phism and food habits of North American owls. Con-
dor 72:251-264.

Emlen, J.T. 1971. Population densities of birds derived
from transect counts. Auk 88:323-341.

. 1977. Estimating breeding season bird densities

from transect counts. Auk 94:455-468.

Errington, P.L. 1930. The pellet analysis method of rap-
tor food habits study. Condor 32:292-296.

Garrido, O.H. 1992. Conozca las rapaces. Editorial Gen-
te Nueva, La Habana, Cuba.

Greenway, J.C., Jr. 1967. Extinct and vanishing birds of
the world, 2nd revised ed. Dover Publ., Inc., New
York, NYU.S.A.

Kirkconnell, A., O.H. Garrido, R.M. Posada Rodri-
guez and S.O. Cubillos. 1992. Los grupos troficos en
la avifauna cubana. Poeyana No. 415:1-21.

Morgan, G.S. 1977. Late Pleistocene fossil vertebrates
from the Cayman Islands, British West Indies. M.S.
thesis, Univ. Florida, Gainesville, FL U.S.A.

. 1993. Quaternary land vertebrates of Jamaica.

Pages 417-442 in R.M. Wright and E. Robinson
[Eds.] , Biostratigraphy of Jamaica. Geological Society
of Marica Memoir 182, Boulder, CO U.S.A.

. 1994. Late Quaternary fossil vertebrates from the

Cayman Islands. Pages 465-508 in M.A. Brunt and J.E.
Davies [Eds.], Monographiae biologicae volume 7
Kluwer Academic Publishers, Dordrecht, The Neth-
erlands.

Olson, S.L. and D.W. Steadman. 1977. A new genus of
flighdess ibis (Threskiornithidae) and other fossil
birds from cave deposits in Jamaica. Proc. Biol. Soc.
Wash. 90:447-457.

Pregill, G.K, 1981. Late Pleistocene herpetofaunas from
Puerto Rico. Univ. Kansas Mus. Nat. Hist. Misc. Publ.
71:1-72.

and S.L. Olson. 1981. Zoogeography of West In-
dian vertebrates in relation to Pleistocene climatic cy-
cles. Annu. Rev. Ecol. Syst. 12:75-98.

Regalado Ruiz, P. 1975. Primer hallazgo de Speotyto cun-
icularia (Molina) anidando en Cuba. Revista Baracoa
No. 1-2:36-56.

R.H.L. 1833. On the burrowing owl. Field Naturalist 1.
459-461.

Snyder, N.F.R. and J.W. Wiley. 1976. Sexual size dimor-
phism in hawks and owls of North America. AOU
Monogr. No. 20.

Union Panamericana. 1967. Reconocimiento y evalua-
cion de los recursos naturales de la Republica Dom-
inicana. Estudio para su desarrollo y planificacion.
Organization de los Estados Americanos, Washing-
ton, DC U.S.A.

Wetmore, A. and B.H. Swales. 1931. The birds of Haiti
and the Domninican Republic. U.S. Nat. Mus. Bull
155.

Wiley, J.W. 1986a. Habitat change and its effects on Puer-
to Rican raptors. Birds of Prey Bull. No. 3:51-56.

. 1986b. Status and conservation of raptors in the

West Indies. Birds of Prey Bull. No. 3:57-70.

Received 21 November 1997; accepted 12 May 1998

Short Communications

/. Raptor Res. 32(3):246-247
© 1998 The Raptor Research Foundation, Inc.

Rates of Open-field Foraging by the Mississippi Kite ( Ictinia mississippiensis )

E. William Wischusen

Department of Biological Sciences, 104 Life Sciences Building, Baton Rouge, LA 70803 U.S.A.

Key Words: Mississippi Kite; Ictinia mississippiensis; for-

aging success; foraging habitat.

During the past 50 yr, populations of the Mississippi
Kite ( Ictinia mississippiensis ) appear to have recovered
from earlier declines (Levy 1971, Parker 1977, Parker
and Ogden 1977, Glinski and Ohmart 1983). Although
several hypotheses have been proposed to explain the
recovery, the reason or reasons remain unclear. The hy-
potheses propose that habitat changes on the breeding
or wintering grounds have allowed for this recovery (Par-
ker and Ogden 1977, Glinski and Ohmart 1983). These
hypotheses stress changes in the habitats that either in-
crease the nesting or foraging opportunities for kites.
While several studies describe nesting habitats used by
Mississippi Kites (Sutton 1939, Jackson 1945, Glinski and
Ohmart 1983, Cely 1987), only one documents foraging
success (Skinner 1962). Skinner found that kites aver-
aged six successful captures of insects per 40 min interval
while foraging over an open field in Alabama. I report
on the results of observations of Mississippi Kites foraging
over a large open field in southern Louisiana.

Study Area and Methods

The study area was located in the Sherburne Wildlife
Management Area approximately 5 km southeast of the
town of Krotz Springs, in southern Louisiana. The area
consisted of an approximately 1000 ha old field sur-
rounded by bottomland hardwood forest on three sides
and bordered by Big Alabama Bayou on the fourth. For-
aging by Mississippi Kites was observed using 7 X 35 bin-
oculars and a 20 X spotting scope. Individual birds were
observed for 2-min intervals which was long enough to
include one or more forage attempts and short enough
to insure that birds would remain in suitable viewing
range for the entire period. During each foraging inter-
val, I recorded the number of successful prey captures.
Prey capture success or failure was determined by ob-
serving whether or not the individual fed following a prey
capture attempt. After each foraging interval, a different
individual was chosen to observe for the next foraging
interval.

During July-August 1995 and May-August 1996, obser-
vations of Mississippi Kite foraging were conducted on 10
d between 1000-1530 H. All observations were made un-
der partly cloudy to sunny and calm weather conditions.

Results and Discussion

The foraging behavior observed was similar to that de-
scribed by Skinner (1962). Kites soared at heights of 50—
100 m and made steep stoops to capture insects. Once
prey was captured, the birds would level off and eat the
prey while soaring over the field (Skinner 1962). The
number of kites observed foraging ranged from 1 to >50,
with an average of 4-6 individuals observed each day
These numbers are similar to the average of 10 reported
by Skinner (1962).

Mississippi Kites were observed for 248 2-min foraging
intervals during this period. The kites caught an average
of 1.18 ± 0.076 (± 1 SE, range = 0-5) prey items per 2-
min foraging interval for a total of 292 total prey items.
This is a much higher rate of prey capture than previ-
ously reported. Skinner (1962) reported an average open
field foraging rate of 6 prey items per 40-min interval or
0.3 prey items per 2-min foraging interval. His observa-
tions were limited to a 5-d period in July (Skinner 1962)

On a monthly basis there were significant differences
between the rate of prey item capture (ANOVA, F =
3.036, df = 3, 244, P = 0.03). The highest success rates
were observed during May and the lowest during August
(Fig. 1). Some of the differences in foraging rates be-
tween months may be explained by juvenile birds forag-
ing in July and August. Although the ages of all foraging
birds were not determined, some juveniles were included
in the July and August samples. All observed prey cap-
tures were insects. Dragonflies and large grasshoppers
were very abundant in this field and have been reported
as main prey for this species (Wayne 1906, Sutton 1939,
Jackson 1945, Skinner 1962).

The results of this study suggest that Mississippi Kites
can be much more efficient at foraging over open fields
than reported by Skinner (1962). Most importantly, the
results point out the need for comparative data of kite
foraging in other habitats and in other locations. Togeth-
er these data would allow for a better understanding of
the foraging abilities of this species and might eventually
lead to a better understanding of its recovery.

RESUMEN. — Las poblaciones de Ictinia mississippiensis en
America del Norte han aumentado recientemente
Aunque las modificaciones del habitat son consideradas

246

Sept 1998

Short Communications

247

Figure 1. Monthly differences in the mean number of
prey items captured by Mississippi Kites ( Ictinia mississip-
piensis) during 2-min foraging intervals.

como las causas de este aumento, existen pocos estudios
que puedan cuantificar esto. Las observaciones sobre el
forrajeo de Ictinia mississippiensis en ambientes abiertos
en Louisiana sugieren que esta especie es muy eficiente
en su forrajeo en este habitat. Ictinia mississipiensisc aptura
un promedio de 1.18 presas por cada 2 minutos de in-
tervalo de forrajeo. Esta es una tasa exitosa mucho mayor

que las anteriores en ambientes abiertos. Se hace nece-
saria la comparacion de datos de forrajeo colectados en
distintos habitats y localidades.

[Traduccion de Cesar Marquez]

Acknowledgments

I am indebted to Dr. John C. Ogden and two anony-
mous reviewers for their comments and suggestions on
an earlier draft of this manuscript and to the Louisiana
Department of Wildlife and Fisheries for their assistance

Literature Cited

Cely, J.E. 1987. American Swallow-tailed Kite uses Missis-
sippi Kite nest . J. Raptor Res. 21:124.

Glinski, R.L. and R.D. Ohmart. 1983. Breeding ecology
of the Mississippi Kite in Arizona. Condor 85:200-207.
Jackson, A.S. 1945. Mississippi Kite. Texas Game and Fish
3:6-7.

Levy, S.H. 1971. The Mississippi Kite in Arizona. Condor
73:476.

Parker, J.W. 1977. Second record of the Mississippi Kite
in Guatemala. Auk 94:168-169.

and J.W. Ogden. 1977. The recent history and

status of the Mississippi Kite. Am. Birds 33:119-129.
Skinner, R.W. 1962. Feeding habits of the Mississippi
Kite. Auk 79:273-274.

Sutton, G.M. 1939. The Mississippi Kite in spring. Condor
41:41-53.

Wayne, A.T. 1906. A contribution to the ornithology of
South Carolina, chiefly the coast region. Auk 23:56-
68 .

Received 12 October 1997; accepted 14 May 1998

J. Raptor Res. 32(3):247-250
© 1998 The Raptor Research Foundation, Inc.

Evaluation of Neck-Mounted Radio Transmitters for Use With Juvenile Ospreys

Lauren N. Gilson 1

Raptor Research Center, Department of Biology, Boise State

Key Words: Osprey, Pandion haliaetus; transmitters ; neck-

lace, retention.

Transmitters on necklaces, originally used on game
birds, resulted from a modification of neck-mounted
markers developed in 1970 in response to selective pre-

1 Present Address: P.O. Box 179 Wakkerstroom, 2480
South Africa.

University, 1910 University Drive, Boise, ID 83725 U.S.A.

dation on individuals with back-mounted markers (Pyrah
1970, Amstrup 1980). For larger birds, neck-mounted
transmitters are used infrequently; backpack-style har-
nesses are preferred for their tenacity and durability in
long-term research (Day et al. 1980, Marion and Shamis
1977, Young and Kochert 1987). For short-term research
using short-lived radio transmitters, mounting methods
must be highly reliable for the length of the study but
need not be permanent. In a study of fledgling behavior,
I mounted radio transmitters around the necks of Os-

248

Short Communications

Vol. 32, No. 3

A.

A.

Figure 1. Rubber band style transmitter mount. A)
Front view. B) As worn by fledglings.

preys (Pandion haliaetus ) using standard rubber bands
which have been used with Ospreys (C.R Schaadt pers.
comm.) as well as with raptors manned for falconry, and
crimped nylon-wound elastic mounts designed and tested
in this study. I describe and review the merits and draw-
backs of each.

Methods

In June-August 1993, I attached eight modified neck-
lace-style transmitters (ATS Model 2032: 5. 6-5. 9 g, 90 d
battery life) to nestling Ospreys of age 35-45 d at Cas-
cade Reservoir, Valley County, ID. I stitched the radio to
a 3 X 4 cm patch of 100% nylon pack cloth rolled into
a sleeve and sewn around a size 34 rubber band (0.4 cm
width, 12.5-13 cm unstretched circumference; herein re-
ferred to as RB, Fig. 1 ) . In July-August 1 994, I attached
16 pendant-style transmitters (Merlin Systems: 7. 4-7. 7 g,
90 d battery life) to Osprey nestlings at Cascade Reser-
voir. I hung the unit around an Osprey’s neck on an
adjustable loop of nylon-wrapped elastic (Stretchrite
Round Cord Elastic, Rhode Island Textile Company;
herein referred to as NWE, Fig. 2). The necklace con-
sisted of two elastic segments with looped ends rejoined
by cotton thread, fed through a 1 .5 cm segment of metal
tubing (Archer Butt Connector, No. 64-3036) and
crimped to size.

Figure 2. Nylon-wound elastic style transmitter mount.
A) Rear view. B) Side view. C) As worn by fledglings
(mounting bracket and antenna toward the Osprey’s
breast) .

Results

Ospreys shed or removed seven of nine RB mounts in

1993. I recovered five of these. On two units, the rubber
bands were broken and the three others were intact. All
recovered transmitters remained firmly attached to the
pack cloth.

RB mounts were lost at an average of 35 d (range 21—
44 d) after application. In 115 hr of observations, I saw
only one juvenile Osprey pull at its transmitter. Its sibling
also preened around the necklace, both on the second
day after application. Transmitter positions occasionally
shifted, indicating that the unit moved freely about the
Osprey’s neck.

Ospreys shed or removed eight of 16 NWE mounts in

1994. The elastic pulled out of the crimping on one unit,
two separated at the break-away loops, the elastic was not
recovered with four units, and the eighth unit was not
recovered. NWE mounts failed at an average of 26 d
(range 18-37 d). In 504 hr of observation, I observed no
Osprey young pulling at their own or their nest mates’

Sept 1998

Short Communications

249

transmitters. All shed or removed transmitters were lost
18-44 d after application. The two mounting styles dif-
fered significantly in mean length of retention (Wilcoxon
Rank Sum Test, S - 73.5, Z = 1.97, P = 0.04; NWE x =
25.7 d, RB x = 34.8 d) but not in variance in retention
(F 6 7 = 1.63, P = 0.53; NWE s = 2.18, RB s = 2.98). The
locations of transmitter losses were not random: 13 of 15
were recovered at or below nest or perch sites (x 2 = 6.2,
P 0.016).

Two Ospreys, one in each year, pulled the necklace
material into their mouths. Whereas the 1993 Osprey fed
normally despite transmitter position and shed the unit
without incident at the reservoir shore two days later, the
1994 Osprey fed itself with difficulty. I trapped this fledg-
ling at its natal nest and removed the transmitter. The
elastic caused only minor abrasion of tissue at the corners
of the mandibles. Necklaces did not obstruct feeding of
any other juvenile Ospreys.

Transmitters that were retained through dispersal from
the breeding area (one in 1993, four in 1994) were worn
by Ospreys for an average of 34 d (range 28-43 d). Four
shed units that were replaced and subsequently retained
through dispersal provided an additional 6—18 d of in-
formation before the Ospreys dispersed.

Discussion

The main benefits of necklace style transmitter mounts
are their low cost, easy construction, rapid application in
the field, and minimal physical impact to their recipients.
I prepared both styles of necklace mounts ahead of time,
then simply slipped them over the heads and worked
them under the feathers in the field. This greatly re-
duced the length of disturbance and amount of stress
incurred by juveniles during marking and measurement.
Per unit, both styles required under 30 min to prepare,
cost less than $1.00 to construct, and took only minutes
to attach. However, I found drawbacks to both styles I
tested.

RB mounts were faster to attach than NAVE mounts,
but could not be adjusted to fit snugly. Rubber bands
should have outlasted the battery life of the transmitter
and dropped off after the unit was defunct. However, the
elasticity of rubber bands apparently enabled fledglings
to pull the transmitters off without breaking the bands.
Although this behavior was never observed in the field,
seven of nine (one remounted) units were either shed
prematurely or successfully removed, three with the
rubber band unbroken.

NWE cord had less stretch than rubber bands, permit-
ting a tighter fit to each individual. However, nestlings
scratched at their necks more often in 1994 than in 1993.
I observed two Ospreys scratching at their necks 1 wk
before both shed their transmitters: one separated at the
loops and the other was missing its elastic. Preening re-
sulted in two Ospreys being bridled by their necklaces,
indicating that beak preening was not strong enough to
break the mounts, but also that neither style fit sufficient-

ly snugly. Were a talon to become hooked under the elas-
tic, the downward force of an Osprey’s leg was probably
sufficient to pull the elastic out of the crimping or snap
the threads at the break-away loops. I recovered no units
on which the nylon cord was broken, in contrast to three
of nine broken RB mounts. Transmitter removals in 1994
may thus have been attributable to scratching or preen-
ing in response to irritation, possibly caused by the
crimped metal tubing. The recovery of most shed units
below platforms or trees supports the notion that trans-
mitter loss was associated with a behavior that is per-
formed while perched.

All methods of transmitter mountings vary in physical
impact to their bearers and in retention. Many studies
that have employed neck-mounted transmitters or visual
markers have encountered loss of markers, injury to the
bearer, or death by starvation or predation (Hawkins and
Simpson 1985, Small and Rusch 1985, Marks and Marks
1987, Maclnnes and Dunn 1987, Pekins 1987, Sorenson
1989, Ely 1990, Samuels et al. 1990), thus leading most
researchers to avoid neck mounts. However, Marcstrom
et al. (1989) found significantly higher survival of pheas-
ants with neck-mounts than with backpacks. Necklace
mounts require a minimum of skill and time to attach in
the field and cannot damage developing wing and tail
feathers of young birds.

Tail mounts have been used effectively with many adult
raptors including Ospreys (Kenward 1985a, Hagan and
Walters 1990, Phelps 1993) but have resulted in damaged
or loss of the retrices to which they were attached (Sa-
muels and Fuller 1994). Backpacks are widely recom-
mended and widely used for raptor studies, yet improp-
erly fitted backpacks have entangled feet (Nicholls and
Warner 1968) and can damage growing body feathers of
young birds (Kenward 1985b). Backpacks also sometimes
affect behavior which can result in selective predation on
their bearers (Small and Rusch 1985, Marcstrom et al.
1989). Although backpacks are more costly and require
more time and skill to attach properly, the benefits of
greater retention and reduced impact on behavior may
outweigh the costs. Kenward (1985b) recommended tail
mounts as best for raptors “. . . unless the retrices of
young birds are not yet fully grown,” which precludes
their use with pre-fledging juveniles. He also recom-
mended anklet mounts over backpacks for juveniles, but
found the transmission range of anklet-mounted radios
was more readily reduced by low or ground perching, as
is often observed in young Ospreys in this population.

In addition to cost-effectiveness and rapid deployment
in the field, I used neck-mounted transmitters to reduce
the risks to juveniles of damage to developing wing and
tail feathers, interference with normal development of
flight and hunting skills, selective predation on already
vulnerable juveniles, or possible electrocution of young
Ospreys with tail or backpack mounts. In general, while
the neck-mounted transmitters in this study did not ap-
pear to cause damage to their bearers, antennae may

250

Short Communications

Vol. 32, No. 3

have been annoying, as juvenile Ospreys were observed
biting at antennae while feeding and shaking their heads
to move the antennae out of the way. Whereas necklace-
style transmitter mounts were easy to construct and apply
in the field, their retention was generally low, and ap-
peared to be obtrusive in several instances. Neither neck-
mounting method proved sufficiently durable for use
with juvenile Ospreys. Both styles were shed early in the
post-fledging period, which lasts 30-35 d at this study
area.

Four possible means of improving NWE mounts are:
(1) use tubing with “teeth” that could effectively bite
into the elastic to prevent it from pulling out, (2) secur-
ing the NWE inside of the tubing with a drop of cyano-
acrylate glue at each end, (3) knotting the end of the
NWE against the ends of the crimped tubing, or (4)
stitching through the elastic rather than crimping it. I
was unable to locate “toothed” tubing small enough to
effectively crimp the 1/8” NWE cord. Although gluing
and/or stitching through might damage the elastic itself,
it may secure the mount better than crimping alone, ex-
tending the effectiveness while still remaining a tempo-
rary mounting method.

Resumen. — El uso de accesorios para sujetar radio-trans-
misores temporalmente son aconsejables para las neces-
idades de investigation en el corto plazo, sin embargo
los transmisores deben ser retenidos durante la duration
del estudio. Con el fin de valorar la eficiencia de transmi-
sores de collar en las aves rapaces, evalue la retention de
dos disenos en individuos juveniles de Pandion haliaetus :
una banda de caucho y un cordon elastico de nylon. Los
transmisores montados con bandas de caucho fueron sig-
nificativamente retenidos por mas tiempo que los de cor-
don elastico de nylon. Sinembargo ninguno de los dos
fue retenido hasta el periodo requerido para ser reco-
mendados para realizar investigaciones de campo.

[Traduction de Cesar Marquez]

Literature Cited

Amstrup, S.C. 1980. A radio collar for game birds. /
Wildl. Manage. 44:214-217.

Day, G.I., S.D. Schemnitz and R.D. Taber. 1980. Captur-
ing and marking wild animals. Pages 61-88 in S.D.
Schemnitz [Ed.], Wildlife management techniques
manual, 4th ed. Wildl. Soc., Inc., Washington, DC
U.S.A.

Ely, C.R. 1990. Effects of neck bands on the behavior of
wintering Greater White-fronted geese. J. Field Orni-
thol. 61:249-253.

Hagan, J.M., III and J.R. Walters. 1992. Foraging be-

havior, reproductive success, and colonial nesting in
Ospreys. Auk 107:506—521.

Hawkins, L.L. and S.G. Simpson. 1985. Neckband a hand-
icap in an aggressive encounter between Tundra
Swans . /. Field Ornithol. 56:182-184.

Kenward, R.E. 1985a. Radio transmitters tail-mounted on
hawks. Ornis Scand. 9:220-223.

. 1985b. Raptor radio-tracking and telemetry, ICBP

Tech. Publ. 5:409-420.

MacInnes, C.D. and E.H. Dunn. 1987. Effects of neck
bands of Canada Geese nesting at the McConnell Riv-
er. J. Field Ornithol. 59:239-246.

Marcstrom, V., R.E. Kenward and M. Karlbom. 1989.
Survival of Ring-necked Pheasants with backpacks,
necklaces, and leg bands. / Wildl. Manage. 53:808-
810.

Marion, W.R. and J.D. Shamis. 1977. An annotated bib-
liography of bird marking techniques. Bird-Banding
48:42-61.

Marks, S.J. and V.S. Marks. 1987. Influence of radio col-
lars on survival of Sharp-tailed Grouse. J. Wildl. Man-
age. 51:468-471.

Nicholls, T.H. and D.W. Warner. 1968. A harness for
attaching radio transmitters to large owls. Bird Band-
ing 39:209-214.

Pekins, P.J. 1987. Effects of poncho-mounted radios on
Blue Grouse. J. Field Ornithol. 59:46-50.

Phelps, J.M., III. 1993. Factors affecting foraging and
productivity of Ospreys ( Pandion haliaetus) at Cascade
Reservoir, Idaho. M.S. thesis, Boise State Univ., Boise,
ID U.S.A.

Pyrah, D. 1970. Poncho markers for game birds./. Wildl.
Manage. 34:466-467.

Samuels, M.D. and M.R. Fuller. 1994. Wildlife radiote-
lemetry. Pages 370-416 in T.H. Bookhout [Ed.], Re-
search and management techniques for wildlife and
habitats. Wildl. Soc., Bethesda, MD U.S.A.

, D.H. Rusch and S.R. Craven. 1990. Influence of

neck bands on recovery and survival rates of Canada
Geese./. Wildl. Manage. 54:45-53.

Small, R.J. and D.H. Rusch. 1985. Backpacks versus pon-
chos: survival and movements of Ruffed Grouse.
Wildl. Soc. Bull. 13:163-165.

Sorenson, M.D. 1989. Effects of neck collar radios on
female Redheads. / Field Ornithol. 60:523-528.

Young, L.S. and M.N. Kochert. 1987. Marking Tech-
niques. Pages 125-156 in B.A. Giron Pendleton, B.A
Millsap, K.W. Cline and D.M. Bird [Eds.], Raptor
management techniques manual. Natl. Wildl. Fed.,
Washington, DC U.S.A.

Received 10 December 1996; accepted 9 May 1998

Sept 1998

Short Communications

251

J Raptor Res. 32(3):251-254
© 1998 The Raptor Research Foundation, Inc.

Organochlorines and Mercury in Peregrine Falcon Eggs from

Western North Carolina

Tom Augspurger

U.S. Fish and Wildlife Service, RO. Box 33726, Raleigh, NC 27636-3726 U.S.A.

Allen Boynton 1

North Carolina Wildlife Resources Commission, Nongame and Endangered Wildlife Program,

Morganton, NC 28655 U.S.A.

Key Words: Peregrine Falcon; Falco peregrinus; mercury;

organochlorine pesticides; North Carolina.

In North Carolina, Peregrine Falcons ( Falco peregrinus
anatum) were historically regarded as an uncommon
breeding bird in the western mountains (Pearson et al.
1919, 1942) prior to their contaminant-induced extirpa-
tion in the eastern U.S. (Hickey 1969). Surveys in the
early 1970s indicated peregrines no longer bred in North
Carolina or at any historical nest sites east of the Missis-
sippi River (Fyfe et al. 1976). Efforts to restore peregrines
to the breeding bird fauna of North Carolina included
protection of nesting habitat and the introduction of 81
young peregrines, produced in captivity by the Peregrine
Fund, Inc. and private breeders (Barclay 1988), at seven
sites between 1984-91. Introduced birds first bred in
1987 and successfully so in 1988. One to five pairs have
bred each year producing a mean of 0.0-2. 5 young per
occupied territory and an average of 0.81 young per oc-
cupied territory from 1987-95 (Boynton and Currie
1993, C. McGrath pers. comm.).

Peregrine Falcons in western North Carolina occupy a
portion of the Southern Appalachian recovery region; an
objective of the recovery plan is the establishment of 20-
25 nesting pairs in this region (U.S. Fish and Wildlife
Service [hereafter USFWS] 1979, 1991), a level that has
not been attained. Although contaminants were not sus-
pected to be limiting productivity, no analyses had been
performed on this new population prior to this assess-
ment. Our objectives were to quantify organochlorine
and mercury concentrations and shell thickness from
western North Carolina peregrine eggs and to evaluate
their significance relative to reproduction.

Methods

Between 1990-93, five Peregrine Falcon eggs were col-
lected from four clutches in three breeding territories
after they were either incubated past term or abandoned.

1 Present address: Virginia Department of Game and In-
land Fisheries, Route 1, Box 107, Marion, Virginia 24354
U.S.A.

Eggs were stored frozen until harvested into acid-rinsed
and solvent-rinsed glass jars for analyses.

The USFWS Patuxent Analytical Control Facility ana-
lyzed for 20 organochlorine compounds and mercury.
Organochlorine analyses were by gas-liquid chromatog-
raphy with peak confirmation of p,p'-DDE by gas chro-
matography/ mass spectrometry under methods adapted
from Cromartie et al. (1975). Mercury determination was
performed by cold vapor atomic absorption spectropho-
tometry as described by the Joint Mercury Residues Panel
(1961).

The lower limit of detection was 0.01 ppm wet weight
for organochlorines and 0.04 ppm wet weight for mer-
cury. A procedural blank indicated no background con-
tamination of analytical equipment or reagents. Results
of duplicate analyses and spike recoveries for mercury
(104%), DDT metabolites (54.7-107%), lindane
(69.6%), and chlordane metabolites (65.1-89.8%) were
within acceptable ranges for method precision and ac-
curacy. Residues reported here were not adjusted for re-
coveries.

We used a regression equation for American Kestrel
{Falco sparverius) eggs to estimate egg volume (Wiemeyer
et al. 1986). Because samples had dehydrated from ex-
posure and refrigeration, we adjusted all residue concen-
trations for moisture loss using egg weight to volume ra-
tios and assuming a specific gravity of 1.0 (Stickel et al.
1973). All contaminant concentrations are reported as
parts per million (ppm) fresh wet weight. Contaminant
concentrations for the two eggs collected in 1991 from
Whiterock Cliff were averaged prior to calculating geo-
metric means. Consequently, geometric means are based
on clutches {N = 4) rather than individual eggs (N = 5).

Eggshell thicknesses of the five specimens and shell
fragments from four additional nests were determined
with a Federal Model 35 bench comparator thickness
gauge at the Western Foundation for Vertebrate Zoology.
At least 10 measurements of each sample were made for
each mean reported.

Results and Discussion

DDE, a metabolite of DDT, was considered causative in
the extirpation of the Peregrine Falcon after it was found
that DDE-induced eggshell thinning of around 20% was
invariably associated with declining populations (Rise-
brough and Peakall 1988). Corresponding DDE residues

252

Short Communications

Vol. 32, No. 3

Table 1. Concentrations of mercury (Hg) and organochlorines measured in eggs of North Carolina Peregrine
Falcons. Concentrations are ppm fresh wet weight and corresponding shell thicknesses are in mm.

Location and Year

Shell

Thickness

Hg

P-P'

DDE

Total

DDT a

Total

CHLORDANE b

Dieldrin

Lindane

Chimney Rock
Rutherford County
1993

0.332

0.14

1.47

1.70

0.50

0.12

0.02

Looking Glass Rock
Transylvania County
1990

0.329

0.07

5.72

6.13

1.95

0.42

0.03

Looking Glass Rock
Transylvania County
1992

0.386

0.11

3.96

4.26

1.04

0.18

0.01

Whiterock Cliff
Madison County
1991

0.328

0.05

5.14

5.70

1.32

0.54

0.02

Whiterock Cliff
Madison County
1991

0.321

0.10

2.64

3.03

1.71

0.66

0.01

a Total DDT defined here as summed o,p'— DDD, o,p'— DDE, o,p'-DDT, p,p'-DDD, p,p'— DDE, and p,p'— DDT.

b Total chlordane defined here as sum of alpha chlordane, cis-nonachlor, gamma chlordane, heptachlor epoxide, oxychlordane, and
trans-nonachlor.

of 15-20 ppm have been associated with population level
average eggshell thinning of this magnitude (Peakall and
Kiff 1988). The geometric mean concentration of p,p'-
DDE in North Carolina Peregrine Falcon clutches was
3.37 ppm (Table 1). Total metabolites of DDT ranged
from 1.70-6.13 ppm with a geometric mean concentra-
tion of 3.73 ppm. These DDE concentrations are well
below those associated with population declines (Peakall
and Kiff 1988).

Arithmetic mean eggshell thicknesses were 0.339 mm
for the five whole eggs and 0.334 mm when the shell
fragments from four additional nests were included. As-
suming a pre-1947 eggshell thickness for eastern Pere-
grine Falcons of 0.360 mm (Burns et al. 1994), the values
reported here are approximately 7% thinner than nor-
mal, pre-DDT era eggshells. Peakall and Kiff s (1988)
summary of thinning data indicate that extirpated or de-
clining Peregrine Falcon populations occurred whenever
mean population-level eggshell thinning exceeded 17%
(except for intensively managed populations). Fyfe et al.
(1988) suggested 14.5% thinning as a minimum estimate
of the threshold below which there is no appreciable ef-
fect on peregrine productivity. Since the stock for the
restored eastern Peregrine Falcon population was de-
rived from several sources, eggshell thickness compari-
sons must necessarily be approximate.

The geometric mean concentration of total chlordane
components and metabolites (1.11 ppm) was composed
primarily by oxychlordane (0.30 ppm) and heptachlor
epoxide (0.22 ppm). Peakall et al. (1990) considered

>1.5 ppm heptachlor epoxide to be critical for produc-
ing adverse effects in peregrine eggs.

Dieldrin levels in raptors were reviewed by Peakall et
al. (1990); they derived an egg screening value for deter-
mining adverse effects in Peregrine Falcons of 1-4 ppm.
Peakall (1996) reported a “no-observed-effect” level of
dieldrin at 0.7 ppm. North Carolina Peregrine Falcon
eggs were well below this range with a geometric mean
of 0.27 ppm.

Lindane has been used to control balsam wooly adel-
gid ( Adelges piceae ) infestation of fraser fir ( Abies fraseri )
in western North Carolina. The geometric mean concen-
tration of lindane in our samples was 0.02 ppm. Lindane
has not been associated with avian impairment in the
wild, and levels well above those reported here have been
detected in bird eggs without any apparent effects (Wie-
meyer 1996).

Mercury has been shown to cause mortality and repro-
ductive impairment in wild birds (Eisler 1987). The geo-
metric mean mercury concentration of 0.10 ppm in
North Carolina Peregrine Falcon eggs is below estimates
of 0.5— 1.0 ppm used by others as screening values for
reproductive effects on peregrines (Peakall et al. 1990)
and other raptors (Bowerman et al. 1995).

The geometric mean concentration of DDE in pere-
grine eggs from the Southern Appalachian recovery re-
gion was approximately half that reported for eggs from
the mid-Atlantic and northeastern U.S. (Gilroy and Bar-
clay 1988, Burns et al. 1994). The lower DDE concentra-
tions in peregrine eggs from the Southern Appalachian

Sept 1998

Short Communications

253

recovery region may be due to different prey prefer-
ences. Preliminary data indicate that western North Car-
olina peregrines depend largely on resident birds, partic-
ularly Rock Doves ( Columba livid ) , Mourning Doves
( Zenaida macroura ), Northern Flickers ( Colaptes auratus),
and Blue Jays ( Cyanocitta cristata ) (Boynton and Currie
1993). Additional documentation of prey preferences is
ongoing.

Factors possibly limiting productivity include nest pre-
dation, inclement weather, inexperience, poor food sup-
ply, and human disturbance (Boynton and Currie 1993).
While a larger data set is advisable before ruling out pol-
lutant effects, our current data suggest that environmen-
tal contaminants are not limiting productivity of pere-
grines in the Southern Appalachian recovery region. The
USFWS (1995) has indicated its intention to remove the
Peregrine Falcon from the list of Endangered and
Threatened Species. These data may serve as a baseline
in future monitoring of the Southern Appalachian pop-
ulation which may be a requirement following delisting.

Resumen. — Concentraciones de pesticidas organoclora-
dos y mercurio fueron encontrados en huevos colectados
de cinco halcones peregrinos de una poblacion reesta-
blecida en el noroeste de Carolina. Las concentraciones
de la media geometrica de p,p' DDE (3.37 ppm en peso
fresco), total de metabolitos de chlordane (1.11 ppm),
dieldrin (0.27 ppm), lindano (0.02 ppm), y mercurio
(0.10 ppm) estaban generalmente por debajo de los ni-
veles asociados al fracaso reproductive. Los 0.334 mili-
metros de la media aritmetica del grueso de la cascara
de los huevos, fue del 7% mas delgada de lo normal
encontrado en la “pr-era” del DDT de la poblacion orig-
inal del este de Estados Unidos.

[Traduccion de Cesar Marquez]

Acknowledgments

We thank John Moore of Patuxent Analytical Control
Facility for coordinating chemical analyses. The data
were part of USFWS Regional Study ID#93-4PER. Stanley
Wiemeyer, Chuck Henny, Tom Cade, Robert Currie,
Chris McGrath, and Randy Wilson provided helpful re-
views of earlier versions of the manuscript.

Literature Cited

Barclay, J.H. 1988. Peregrine restoration in the eastern
United States. Pages 549-558 in T.J. Cade, J.H. En-
derson, C.G. Thelander and C.M. White [Eds.], Per-
egrine Falcon populations: their management and re-
covery. The Peregrine Fund, Inc., Boise, ID U.S.A.
Bowerman, W r W 7 ., J.P. Giesy, D.A. Best and V.J. Kramer.
1995. A review of factors affecting productivity of Bald
Eagles in the Great Lakes region: implications for re-
covery. Environ. Health Perspect. 103:51-59.

Boynton, A.C. and R. Currie. 1993. Productivity of re-
introduced Peregrine Falcons in western North Car-
olina. Proc. Annu. Conf. Southeast. Assoc. Fish and Wildl.
Agencies 47:386—393.

Burns, S.A., W.M. Jarman, T.J. Cade, L.F. Kiff and B J
WALTON. 1994. Organochlorines and eggshell thin-
ning in Peregrine Falcon Falco peregrinus eggs from
the eastern United States, 1986-1988. Pages 709-716
in B.-U. Meyburg and R.D. Chancellor [Eds.], Raptor
conservation today. Proceedings of the IV world con-
ference on birds of prey and owls. Pica Press, London,
U.K.

Cromartie, E., W.L. Reichel, L.N. Locke, A. A. Belisle,

T. E. Kaiser, T.G. Lamont, B.M. Mulhern, R.M. Prou-
ty and D.M, Swineford. 1975. Residues of organo-
chlorine pesticides and polychlorinated biphenyls and
autopsy data for Bald Eagles, 1971-72. Pest. Momt J.
9:11-14.

Eisler, R. 1987. Mercury hazards to fish, wildlife, and
invertebrates: a synoptic review. USFWS Biol. Rep.
85(1.10). U.S. Fish Wildl. Serv., Laurel, MD U.S.A

Fyfe, R.W., S.A. Temple and T.J. Cade. 1976. The 1975
North American Peregrine Falcon survey. Can. Field
Nat. 90:228-273.

, R.W. Risebrough, J.G. Monk, W.M. Jarman, D.W.

Anderson, L.F. Kiff, J.L. Linger, I.C.T. Nisbet, W.
Walker, II and B.J. Walton. 1988. DDE, productivity,
and eggshell thickness relationships in the genus Fal-
co. Pages 319-335 in T.J. Cade, J.H. Enderson, C.G.
Thelander and C.M. White [Eds.], Peregrine Falcon
populations: their management and recovery. The
Peregrine Fund, Inc., Boise, ID U.S. A.

Gilroy, M.J. and J.H. Barclay. 1988. DDE residues and
eggshell characteristics of reestablished peregrines in
the eastern United States. Pages 403-411 mT.J. Cade,
J.H. Enderson, C.G. Thelander and C.M. White
[Eds.], Peregrine Falcon populations: their manage-
ment and recovery. The Peregrine Fund, Inc., Boise
ID U.S.A.

Hickey, J.J. 1969. Peregrine Falcon populations: their bi-
ology and decline. Univ. Wisconsin Press, Madison,
WI U.S.A.

Joint Mercury Residue Panel. 1961. Recommended
methods of analysis of pesticide residues in food
stuffs. Analyst 86:608-614.

Peakall, D.B. 1996. Dieldrin and other cyclodiene pes-
ticides in wildlife. Pages 73-97 in W.N. Beyer, G.H.
Heinz and A.W. Redmon-Norwood [Eds.], Environ-
mental contaminants in wildlife: interpreting tissue
concentrations. Lewis Publishers, Boca Raton, FL

U. S. A.

and L.F. Kiff. 1988. DDE contamination in pere-
grines and American Kestrels and its effect on repro-
duction. Pages 337-350 in T.J. Cade, J.H. Enderson,
C.G. Thelander and C.M. White [Eds.], Peregrine Fal-
con populations: their management and recovery.
The Peregrine Fund, Inc., Boise, ID U.S.A.

, D.G. Noble, J.E. Elliot, J.D. Somers and G. Er-
ickson. 1990. Environmental contaminants in Cana-
dian Peregrine Falcons, Falco peregiinus: a toxicologi-
cal assessment. Can. Field Nat. 104:244-254.

254

Short Communications

Vol. 32, No. 3

Pearson, T.G., C.S. Brimley and H.H. Brimley. 1919.
Birds of North Carolina. North Carolina geologic and
economic survey. Vol. IV. Edwards and Broughton
Printing Co., Raleigh, NC U.S.A.

. 1942. Birds of North Carolina. North Carolina

Department of Agriculture. Bynum Printing Co., Ra-
leigh, NC U.S.A.

Risebrough, R.W. and D.B. Peakall. 1988. The relative
importance of several organochlorines in the decline
of Peregrine Falcon populations. Pages 449-462 in
T.J. Cade, J.H. Enderson, C.G. Thelander and C.M.
White [Eds.], Peregrine Falcon populations: their
management and recovery. The Peregrine Fund, Inc.,
Boise, ID U.S.A.

Stickel, L.F., S.N. Wiemeyer and L.J. Blus. 1973. Pesti-
cide residues in eggs of wild birds: adjustment for loss
of moisture and lipid. Bull. Environ. Contam. Toxicol. 9:
193-196.

U.S. Fish and Wildlife Service. 1979. Eastern Peregrine
Falcon recovery plan. Washington, DC U.S.A.

. 1991. First update of Peregrine Falcon ( Falcoper -

egrinus), eastern population, revised recovery plan.
Newton Corner, MA U.S.A.

. 1995. Endangered and threatened wildlife and

plants; advance notice of a proposal to remove the
American Peregrine Falcon from the list of endan-
gered and threatened wildlife. Federal Register 60:
34406-34409.

Wiemeyer, S.N. 1996. Other organochlorine pesticides in
birds. Pages 99-115 in W.N. Beyer, G.H. Heinz and
A.W. Redmon-Norwood [Eds.], Environmental con-
taminants in wildlife: interpreting tissue concentra-
tions. Lewis Publishers, Boca Raton, FL U.S.A.

, R.D. Porter, G.L. Hensler and J.R. Maestrelli.

1986. DDE, DDT + dieldrin: residues in American
Kestrels and relations to reproduction. USFWS, Fish
Wildl. Tech. Rep. 6. Washington, DC U.S.A.

Received 23 January 1997; accepted 11 May 1998

J. Raptor Res. 32(3):254-256
© 1998 The Raptor Research Foundation, Inc.

Importance of Birds and Potential Bias in Food Habit Studies of
Montagu’s Harriers (Circus pygargus) in Southeastern Spain

Jose A. SAnchez-Zapata and Jose F. Calvo
Departamento de Ecologia e Hidrologia, Facultad Biologia, Universidad de Murcia,
Campus de Espinardo, 301 00 Espinardo, Murcia, Spain

Key Words: Montagu’s Harrier, Circus pygargus; diet,

mediterranean.

Methods used to study the diets of raptors include the
analysis of pellets, stomach contents, prey remains in
nests or under perches, and direct observation of prey
delivered to nests. In many species, including harriers
( Circus spp.), analysis of prey remains only appears to
underestimate the proportion of smaller prey and over-
estimate the occcurence of large prey (Schipper 1973,
Simmons et al. 1991, Manosa 1994, Real 1996). Several
researchers have studied the diet of Montagu’s Harrier
{Circus pygargus) during the breeding season (Perez-Chis-
cano and Fernandez-Cruz 1971, Perez-Chiscano 1974,
Corbacho et al. 1995, Thiollay 1968, Helmich 1986), on
migration (Castroviejo 1969), and while wintering
(Cramp and Simmons 1980, Cormier and Baillon 1991).
Extensive studies of their breeding diet in Spain (Hiraldo
et al. 1975, Arroyo 1997), Holland and France (Schipper
1973), and Britain (Underhill-Day 1993) indicate that

small birds and mammals are important prey in northern
and central Europe, whereas in southern Europe inver-
tebrates appear to be numerically important, together
with small birds (Underhill-Day 1993, Hiraldo et al.
1975). The goal of our study was to assess the importance
of birds in the diet of Montagu’s Harriers and to test how
different study methods affect the results of such a food
habits study.

Study Area and Methods

The diets of three pairs of Montagu’s Harriers breed-
ing in a wadi or “rambla” in a mediterranean semiarid
region in southeastern Spain (Suarez et al. 1996) were
studied during 1995. Ajauque is a small wetland located
in the most arid sector (average annual rainfall = 30 cm)
of Murcia in southeastern Spain. Ajauque rambla drains
an impermeable watershed of sedimentary marls. In arid
and semiarid lands of the region, ramblas are more pro-
ductive than surrounding lands owing to their vegetation
that consists mainly of reeds {Phragmites australis) and hal-
ophytic plants. The Ajauque rambla is part of a protected

Sept 1998

Short Communications

255

Table 1. Montagu’s Harrier prey during courtship (April), incubation (May), nestling (June), and fledging (July)
periods, and all data pooled (All). Proportion of each prey and prey group by number (%) and by weight (%W).

Prey

April

%

May

%

June

%

July

%

All

%

%W

BIRDS

22

59

31

68

15

44

13

59

81

58

86

Carduelis chloris

1

3

0

0

1

3

0

0

2

1

3

Carduelis carduelis

0

0

1

2

1

3

0

0

2

1

2

Carduelis spp.

2

5

0

0

2

7

1

5

5

4

5

Galerida theklae

2

5

0

0

0

0

0

0

2

1

3

Galerida sp.

1

3

1

2

1

3

2

9

5

4

8

Calandrella rufescens

1

3

0

0

0

0

0

0

1

1

1

Alaudidae ind.

1

3

9

20

1

3

0

0

11

8

14

Cisticola juncidis

0

0

0

0

1

3

0

0

1

1

1

Saxicola torquata

0

0

0

0

3

9

3

13

6

4

4

Unidentified passerine

14

38

20

44

5

15

7

32

46

33

45

REPTILES

4

11

7

15

4

12

1

5

16

11

7

Psammodromus algirus

2

5

2

4

0

0

0

0

4

3

2

Unidentified lizzard

2

5

5

11

4

12

1

5

12

8

6

MAMMALS

0

0

2

4

0

0

0

0

2

2

5

Mus spp.

0

0

1

2

0

0

0

0

1

1

1

Oryctolagus cuniculus

0

0

1

2

0

0

0

0

1

1

4

INVERTEBRATES

11

30

6

13

15

44

8

36

40

29

2

Anacridium aegyptium

7

19

0

0

5

15

0

0

12

9

2

Unidentified Orthoptera

0

0

2

4

1

3

2

9

5

3

0

Orictes nasicornus

1

3

0

0

0

0

0

0

1

1

0

Unidentified Coleoptera

0

0

4

9

7

21

5

22

16

12

0

Unidentified invert.

3

8

0

0

2

6

1

5

5

4

0

TOTAL

37

46

34

22

139

area that includes 69.4 ha of wetland where 3-5 pairs of
Montagu’s Harriers regularly breed on dense salty-shrub
(. Sarcocornia fruticosa and Scirpus spp .) and hunt in sur-
rounding shrubsteppes and cropland.

We divided the breeding season into four periods of
approximately 30 d each: courtship period (April), in-
cubation period (May), nestling period (June), and
fledging period (July) . We used three different methods
to assess food habits: (1) we collected pellets found un-
der perches used by male harriers, (2) we collected prey
remains found under perches used by both males and
females, and (3) we identified prey during aerial trans-
fers by males to females and young. Perches were visited
weekly during the breeding period and all prey remains
were identified and removed. Prey were identified using
reference collections. Weights of prey were assigned fol-
lowing Hiraldo et al. (1975) and Donazar (1988). We
used percentage uniformity tests (Sokal and Rohlf 1969)
to compare the different diets determined using the dif-
ferent study methods.

Results and Discussion

When we pooled all the data, small birds were the most
numerous prey (58%) followed by invertebrates (29%),
lizards (12%), and mammals (1%) (Table 1). Overall,
small birds including small land, shrub and edge-nesting
passerines (Thela lark, Galerida theklae, Lesser Short-toed
Lark, Calandrella rufescens ; Common Stonechat, Saxicola

torquata; Citting Cisticola, Cisticola juncidis; and finches,
Carduelis spp.) made up to 84% of the prey by weight.
During the courtship and incubation periods, nestling
passerines and eggs accounted for 60-86% of the birds
eaten. Orthopterans and coleopterans were the second
prey group by number, followed by a small lizard Psam-
modromus algirus which is the most common lizard species
in salty shrub-steppes in southeastern Spain (Hernandez
et al. 1993). Small mammals occurred infrequently in the
diet probably because of their low densities and noctur-
nal nature, especially in arid and semiarid Mediterranean
regions (Herrera and Hiraldo 1976, Sanchez 1994).

The proportion of birds found in pellet samples (48%,
N = 83) was smaller than in samples of prey remains
(78%, N = 37) and aerial transfers (72%, N = 19),
though there were only significant differences between
pellets and prey remains ( t = 3.23, df = 2, P = 0.0012).
Several authors have suggested that pellets provide a less
biased source of information on raptor diets because
many prey species found in pellets are seldom found in
prey remains at nests or under perches (Schipper 1973,
Simmons et al. 1991, Real 1996). Because pellet samples
underestimated the number of small birds in our study,
it seems that the opposite appears to be true in the diet
of Montagu’s Harriers. Most of bird prey recorded were
nesdings or juveniles which are easy to digest and diffi-

256

Short Communications

Vol. 32, No. 3

cult to find in pellets (Underhill-Day 1993), whereas in-
vertebrate and lizard remains such as scales in pellets
were easily detected. Birds have also been shown to be
the main prey of Montagu’s Harriers both in terms of
numbers and weight in most of Europe during the breed-
ing season (Schipper 1973, Hiraldo et al. 1975, Under-
hill-Day 1993, Corbacho et al. 1995, Arroyo 1997). Small
mammals such as voles ( Mictrotus spp.) and young hares
( Lepus spp.) have been shown to be only locally or tem-
porally important (Thiollay 1968, Arroyo 1997) even
though most studies have only considered pellets.

RESUMEN. — Se analiza la dieta del Aguilucho cenizo en
un humedal del Sureste de Espana durante el periodo
reproductor utilizando tres metodos diferentes; egagro-
pilas, restos de presas en posaderos y transferencia de
presas. Los pajaros se revelan como el componente prin-
cipal de la dieta tanto en numero como en biomasa, se-
guidos por los invertebrados y reptiles. Se observaron ses-
gos en la proportion de aves que aparecen en la dieta
segun el metodo de estudio.

[Traduction Autores]

Acknowledgments

This study had financial support from the Direction
General del Medio Natural of Murcia Region and LIFE
projects (UE). Sergio Eguia, Andres Gimenez, Javier
Royo, Jose Maria Caballero, Jose Joaquin Hernandez,
Asuncion Andreu and Trino Ferrandez helped during
the field work. Andres Millan and Eulalia Clemente
helped to identify invertebrate remains. Jose Antonio
Donazar revised an early draft and Beatriz Arroyo, Keith
L Bildstein and an anonymous referee made valuable
comments on the manuscript.

Literature Cited

Arroyo, B.E. 1997. Diet of Montagu’s Harrier Circus py-
gargus in central Spain: analysis of temporal and geo-
graphic variation. Ibis 139:664—672.

Castroviejo, J. 1969. Sobre paso y alimentacion de Circus
pygargus en el NW de Espana. Ardeola 14:216—217.
Cormier, J.P. and F. Baillon. 1991. Concentracion de
busards cendres Circus pygargus (L.) dans la region de
’bour (Senegal) durante l’hiver 1988-1989: utilisation
du milieu et regime alimentaire. Alauda 59:163-168.
Corbacho, C., A. Munoz and P. Bartolome. 1995. Es-
pectro trofico del aguilucho cenizo ( Circus pygargus
L.) en Extremadura. Alytes: 441-448.

Cramp, S. and K.E. Simmons. 1980. The birds of western
Paleartic, Vol. II. Oxford Univ. Press, Oxford, U.K.
Donazar, J.A. 1988. Variaciones en la alimentacion entre

adultos reproductores y polios en el Buho Real ( Bubo
bubo). Ardeola 35:278-284.

Helmich, J. 1986. Notas sobre el ritmo de actividad y la
alimentacion del aguilucho cenizo en Agosto y Sep-
tiembre en Extremadura. Alytes 5:69-79.

HernAndez, V.F. Dicenta, F. Robledano, M. Garcia,
M.A. Esteve and L. Ramirez. 1993. Anfibios y reptiles
de la Region de Murcia. Cuadernos de Ecologia y Me-
dioambiente, No. 1. Univ. Murcia, Murcia, Spain.

Herrera, C.M. and F. Hiraldo. 1976. Food niche and
trophic relationships among European owls. Ornis
Scand. 7:29-41.

Hiraldo, F., F. Fernandez and F. Amores. 1975. Diet of
the Montagu’s Harrier {Circus pygargus) in southwest-
ern Spain. Donana Acta Vertebrata 2:25—55.

Manosa, S. 1994. Goshawk diet in a Mediterranean area
of northeastern Spain. J. Raptor. Res. 28:84—92.

Perez-Chiscano, J.L. 1974. Sumario informe sobre ali-
mentacion de rapaces en el NE de Badajoz. Ardeola
19:331-336.

AND M. Fernandez-Cruz. 1971. Sobre Grus grusy

Circus pygargus en Extremadura. Ardeola, Vol. Especial 1
509-574.

Real, J. 1996. Biases in diet study methods in the Bo-
nelli’s Eagle./. Wildl. Manage. 60:632-638.

Sanchez, J.A. 1994. Ecologia de las aves de presa de la
region de Murcia. M.S. thesis, Univ. Murcia, Murica,
Spain.

Schipper, W.J.A. 1973. A comparison of prey selection in
sympatric harriers ( Circus) in western Europe. Le Ger-
faut 63:17-120.

Simmons, R.E., D.M. Avery and G. Avery. 1991. Biases in
diets determined from pellets and remains: correc-
tion factors for a mammal and bird-eating raptor. J
Raptor Res. 25:63-67.

Sokal, R.R. and F.J. Rohlf. 1969. Biometry. Freeman and
Co., San Francisco, CA U.S.A.

Suarez, M.L., M.R. Vidal, J.F. Calvo, M.A. Esteve, R.
Gomez, A. Gimenez, J.A. Pujol, J.A. SAnchez, M. Par-
do, J. Contreras and L. Ramirez. 1996. Zone humide
d’Ajauque-Rambla Salada. Pages 40-52 in C. Morillo
and J.L. Gonzalez [Eds.], Management of Mediterra-
nean wetlands. III. Ministerio de Medio Ambiente,
Gran Via San Francisco, Madrid.

Thiollay, J.-M. 1968. La pression de predation estivale
du busard cendre Circus pygargus L. sur les popula-
tions de Microtus arvalis en Vendee. Terre Vie 22:321-
342.

Underhill-Day, J.C. 1993. The foods and feeding rates
of Montagu’s Harriers Circus pygargus breeding in ar-
able farmland. Bird Study 40:74—80.

Received 6 August 1997; accepted 19 May 1998

J. Raptor Res. 32(3):257-260
© 1998 The Raptor Research Foundation, Inc.

Prey Brought to Red-Shouldered Hawk Nests in the Georgia Piedmont

Doug L. Howell 1 and Brian R. Chapman

Daniel B. Warnell School of Forest Resources, University of Georgia, Athens, Georgia 30602-2152 U.S.A.

Key Words: Red-shouldered Hawk, Buteo lineatus; prey ;

nests ; food habits', Georgia s; Piedmont.

The Red-shouldered Hawk ( Buteo lineatus ) is a com-
mon breeding species throughout the southeastern U.S.
Despite its wide distribution, information on its food hab-
its in the Southeast is largely anecdotal (Burleigh 1958,
Janik and Mosher 1982). Although food habit studies in
several geographic regions have documented the breadth
the Red-shouldered Hawk’s diet (Craighead and Craig-
head 1956, Snyder and Wiley 1976, Bednarz and Dins-
more 1985, Parker 1986), they differ with respect to the
importance of certain prey classes in the diet. While this
could represent variation within and between regions, it
may be an artifact of the methods used to quantify prey
(Marti 1987), or due to the failure to report results in
terms of biomass (Steenhof 1983). Our objectives were
to quantify the prey brought to nests of the Red-shoul-
dered Hawk in Georgia and to compare food habits in
Georgia with those reported elsewhere.

Study Area and Methods

The study was conducted on the 5718-ha Bishop F.
Grant Memorial Forest (BGF), located in Putnam and
Morgan counties, approximately 14 km north of Eaton-
ton (83°28'N, 33°25'W), in east-central Georgia. The area
lies within the Piedmont physiographic province, a pe-
neplain dissected by numerous streams to form a rolling
topography (Brender 1973). Elevation ranges from 120—
220 m above sea level. Average annual rainfall is approx-
imately 120 cm, with peak precipitation occurring in win-
ter (USDA-SCS 1965, 1976).

Over 60% of BGF consists of natural or planted stands
of loblolly pine ( Pinus taeda) . Bottomland hardwood for-
ests (7%) exist along the area’s major drainages. These
include Big Indian Creek, Glady Creek, and Little River.
Dominant vegetation includes green ash ( Fraxinus penn-
sylvanicus) , sweetgum (Liquidambar styraciflua) , box elder
(Acer' negundo), sycamore ( Platanus occidentalis) , overcup
oak ( Quercus lyrata) , water oak ( Q. nigra ) , and willow oak
(Q. phellos). Upland hardwood stands (23%) consisting
of mixed oaks ( Quercus spp.) and hickories ( Carya spp.),
blackgum ( Nyssa sylvatica), sweetgum, and winged elm
( Ulmus alata) lie adjacent to bottomland corridors, or are
associated with major drainage basins. The remainder of
BGF is maintained as pasture for cattle grazing and hay
production, or is planted as wildlife food plots. Several

1 Present address: Directorate of Public Works and the
Environment, AFZA-PW-DW, Wildlife Branch, Fort Bragg,
NC 28307-5000 U.S.A.

small reservoirs provide irrigation, public fishing, and wa-
terfowl habitat.

We monitored prey deliveries to eight occupied Red-
shouldered Hawk nests within (N — 6) and around ( N =
2) BGF from 3 April-14 July 1994. Old nests were located
prior to leaf-out, and then rechecked for signs of occu-
pancy. Observations totaling 103 hr were made with a 20-
45 X spotting scope and 8X binoculars from a ground
blind placed within 20 m the base of the nest tree. Ob-
servation periods were normally 4-6 hr and were allocat-
ed randomly to cover all daylight hours (0600-1800 H)
Most nests were observed over one time interval at least
once each week from early in incubation until the young
had fledged. Nest sites were checked periodically for re-
mains of prey beneath the nest. We compared the obser-
vational data with those of prey remains to insure that we
counted only those prey remains that could not have
been seen during observations from blinds. Regurgitated
pellets normally contained only hair or feathers, and
were excluded from the analysis. Prey items were identi-
fied to species or the lowest possible taxonomic category.

We calculated the percent frequency of each prey item
from the total number of items delivered to nests and
collected from prey remains. The percent biomass con-
tribution of each prey item was calculated by multiplying
the frequency of occurrence of each prey item by its
mean body mass. When possible, we derived biomass di-
rectly from prey collected on the study area. Otherwise,
we estimated prey biomass (Marti 1987) from the litera-
ture (Golley 1962, Steenhof 1983, Dunning 1984). Large
insects were assumed to weigh 1 g, the average mass ob-
tained from representative samples collected on the
study area. The masses of unidentified prey were as-
sumed to be similar to the mean mass of the most closely
related, identified taxa.

Results and Discussion

All Red-shouldered Hawk nests were located in bot-
tomland forests, or in upland hardwood stands adjacent
to the bottomland corridor (Moorman and Chapman
1996), Mean nest height was 17.6 m (range = 12.2-21.3
m, N = 8) . Six of eight nests fledged at least one young
(range = 1-2) for an average of 1.8 young per successful
nest.

A total of 181 prey items (Table 1) was identified by
observations made from blinds ( N = 144) and remains
collected beneath nests ( N — 37) . Prey delivered to nests
averaged 36.1 g (range = 1-487 g). Vertebrates repre-
sented 76.2% of the prey by numbers and 97.2% of prey
biomass. Vertebrate prey included nine species of mam-
mals, nine species of birds, eight species of reptiles, and
four species of amphibians. Invertebrates represented
23.8% of the prey by numbers, but were insignificant in

257

258

Short Communications

Vol. 32, No. 3

Table 1. Food habits of Red-shouldered Hawks ( Buteo lineatus ) in the Bishop F. Grant Memorial Forest, Putnam and
Morgan counties, Georgia in 1994. Prey items from eight nests identified from visual observations in blinds (N =
144) and prey remains beneath nests (N — 37). Prey listed taxonomically by class. N = number of individuals, %N
= percent occurrence of prey, Mass = mean prey biomass in grams, and %B = percent of total biomass.

Prey Species N % N Mass %B

Oligochaeta

Unidentified earthworms
Crustacea

Unidentified crayfish ( Cambarus spp.)
Insecta

Unidentified beetles
Unidentified grasshoppers and crickets
Unidentified caterpillars
Amphibia

Spotted salamander ( Ambystoma maculatum )
Two-lined salamander ( Eurycea bislineata )
Unidentified salamanders
Southern toad ( Bufo terrestris )

Unidentified toads ( Bufo spp.)

Southern leopard frog ( Rana utricularia )
Unidentified frogs ( Rana spp.)

Reptilia

Snapping turtle ( Chelydra serpentina)

Green anole ( Anolis carolinensis)

Eastern fence lizard ( Sceloporus undulatus )
Unidentified skinks ( Eumeces spp.)

Black racer ( Coluber constrictor)

Black rat snake (E lap he obsoleta)

Eastern kingsnake ( Lampropeltis getulus )
Rough green snake ( Opheodrys aestivus )
Unidentified water snakes ( Nerodia spp.)
Eastern garter snake ( Thamnophis sirtalis)
Aves

Mourning Dove ( Zenaida macroura)
Carolina Wren ( Thryothorus ludovicianus)
American Robin ( Turdus migratorius )

Pine Warbler ( Dendroica pinus )
Unidentified warbler ( Dendroica sp.)
Common Yellowthroat ( Geothlypis trichas)
Kentucky Warbler ( Oporonis formosus)
Hooded Warbler ( Wilsonia citrina )
Northern Cardinal ( Cardinalis cardinalis)
Indigo Bunting ( Passerina cyanea)
Unidentified passerines
Mammalia

Short-tailed shrew ( Blarina brevicauda )
Eastern mole ( Scalopus aquaticus )

Eastern cottontail ( Sylvilagus floridanus)
Eastern chipmunk ( Tamias striatus )

Eastern gray squirrel ( Sciurus carolinensis)
White-footed mouse ( Peromyscus leucopus)
Unidentified mice ( Peromyscus spp.)

Golden mouse ( Ochrotomys nuttalli)

Hispid cotton rat ( Sigmodon hispidus)

Pine vole ( Microtus pinetorum)

Unidentified voles ( Microtus spp.)
Unidentified rodents
TOTAL

8

4.4

6

0.7

17

9.4

7

1.8

5

2.8

1

0.1

8

4.4

1

0.1

5

2.8

1

0.1

1

0.6

12

0.2

5

2.9

7

0.5

10

5.5

9

1.4

3

1.7

20

0.9

4

2.2

20

1.2

10

5.5

38

5.8

13

7.2

38

7.7

11

6.1

24 a

4.0

3

1.7

15

0.7

3

1.7

17

0.8

4

2.2

18

1.1

3

1.7

77

3.5

2

1.1

190

5.8

1

0.5

190

2.9

1

0.5

15

0.2

5

2.9

125 a

9.7

10

5.5

64

9.8

1

0.5

119

1.8

1

0.5

21

0.3

1

0.5

78

1.2

1

0.5

12

0.2

1

0.5

12

0.2

1

0.5

10

0.1

1

0.5

14

0.2

2

1.1

11

0.3

3

1.7

45

2.1

1

0.5

15

0.2

2

1.1

17

0.5

2

1.1

13

0.4

1

0.5

40

0.6

1

0.5

230 a

3.5

4

2.2

110

6.7

1

0.5

487

7.5

3

1.7

18

0.8

2

1.1

18

0.6

2

1.1

18

0.6

6

3.3

85

7.8

4

2.2

26

1.6

3

1.7

26

1.2

5

2.9

33

2.6

181

100.0

100.0

Specimens represent subadult animals.

Sept 1998

Short Communications

259

terms of prey biomass (2.8%; Table 1). All of the crus-
taceans identified were crayfish (Cambaridae), and the
oligochaetes were earthworms (Lumbricidae).

Amphibians (25.6%) were the most frequently deliv-
ered prey items to Red-shouldered Hawk nests. Frogs
(. Rana spp.) were numerically important as prey and, col-
lectively, they represented 41.4% of amphibian prey de-
livered to nests. Reptiles (38.5%) and mammals (33.9%)
contributed most to total prey biomass (Table 1), fol-
lowed by amphibians (17.7%), birds (7.1%), crustaceans
(1.8%), oligochaetes (0.7%), and insects (0.3%). The
species contributing most to total prey biomass were east-
ern garter snakes ( Thamnophis sir tails) and water snakes
( Nerodia spp.).

Red-shouldered Hawks in the Georgia Piedmont
preyed upon a variety of food items. Although their food
habits were similar to those reported in previous studies
in which the majority of prey taken were amphibians,
reptiles, mammals, and crayfish (Craighead and Craig-
head 1956, Snyder and Wiley 1976, Bednarz and Dins-
more 1985, Parker 1986), the importance of certain prey
classes such as amphibians and reptiles differed. Craig-
head and Craighead (1956) and Bednarz (1979), for in-
stance, reported food habits based largely on percentages
of prey occurring in pellets and found that small mam-
mals were the preferred food of nesting Red-shouldered
Hawks in Michigan and Iowa, respectively. Snyder and
Wiley (1976) reported that invertebrates were the domi-
nant foods found in Red-shouldered Hawk stomachs. Par-
ker (1986) used visual observations to identify food items
delivered to nests in Missouri, and found that amphibians
were the most frequently delivered prey. Overall, frogs
{Rana spp.) were the most frequently delivered prey to
Red-shouldered Hawk nests in our study, and contributed
most to total prey biomass. Craighead and Craighead
(1956), Bednarz (1979), and Parker (1986) all found
frogs to be important foods of Red-shouldered Hawks.
However, studies relying on pellet analysis alone may un-
derestimate the propoifions of amphibians in the diet
because amphibians are often completely digested and
leave little osseous remains in pellets (Errington 1932).
Snyder and Wiley (1976) and Portnoy (1974) both re-
ported a higher incidence of frogs in the diet than sug-
gested by pellet analysis.

Eastern garter snakes, unidentified water snakes and
hispid cotton rats ( Sigrnodon hispidus ) also were important
prey, in terms of both numbers and biomass. Garter
snakes were reported in Red-shouldered Hawk diets in
Michigan (Craighead and Craighead 1956) and Iowa
(Bednarz 1979). Cotton rats have never been reported as
important prey, but distribution of cotton rats does not
extend into the northern portion of the Red-shouldered
Hawk’s range, where voles (Microtus spp.) and mice ( Pero -
myscus spp.) are taken more frequently (Craighead and
Craighead 1956, Bednarz 1979). Snapping turtles ( Chel -
ydra serpentina) were numerically important prey items,

but contributed little to overall prey biomass because in-
dividuals taken by the hawks were small.

Red-shouldered Hawks in our study rarely brought
birds to nests and birds contributed little to total prey
biomass (7.1%). Only Craighead and Craighead (1956)
found that birds were important prey items based on the
frequency of occurrence of avian species within pellets

Although invertebrates, particularly crayfish, were fre-
quently delivered to Red-shouldered Hawk nests in our
study, they contributed little to total prey biomass (2.8%).
Snyder and Wiley (1976) found that 55.6% of a Red-
shouldered Hawk’s diet included invertebrates. The in-
vertebrate component of their study probably was over-
estimated because they did not examine their data in
terms of biomass. In addition, their study was based large-
ly on analysis of stomach contents, which could contain
items of secondary origin, particularly insects, ingested
incidentally as stomach contents of prey.

Of the prey delivered to Red-shouldered Hawk nests
in this study, 60% (108 of 181) were those frequently
associated with bottomland forests, marshes, or wet
meadows. Red-shouldered Hawks we equipped with radio
transmitters (Howell and Chapman 1997) were located
most often foraging within bottomland forests close to
water, small beaver ( Castor canadensis ) ponds, wet mead-
ows, or areas containing many seasonally or permanently
flooded pools. Other researchers also have demonstrated
the importance of these habitats as foraging sites for Red-
shouldered Hawks (Henny et al. 1973, Portnoy 1974,
Bednarz 1979, Parker 1986, Bloom et al. 1993). Red-
shouldered Hawks in our study foraged in the bottom-
land forest habitat and used the variety of foods within
it, rather than specializing on particular prey species, a
result consistent with Bednarz and Dinsmore (1985). The
most important foods of Red-shouldered Hawks during
the nesting season were reptiles and amphibians, partic-
ularly snakes and frogs, associated with the bottomland
forest. Small mammals may become more important dur-
ing the winter months, given the seasonality of the pre-
ferred prey (Craighead and Craighead 1956) and also
may increase in importance as buffer foods during ex-
tremely dry conditions (Bednarz and Dinsmore 1985)

Resumen. — Las presas de Buteo lineatus fueron estudiadas
durante la estacion reproductiva en tin area de pinos de
manejo intensivo en la region fisiogeografica del piede-
monte de Georgia. Un total de 1881 items fueron entre-
gados a los pichones (N = 144) y, colectados como restos
de presas debajo de los nidos (N = 37) . Los vertebrados
representaron el 76.2% de las presas en numeros y el
97.2% de la biomasa, incluyendo nueve especies de aves,
ocho especies de reptiles y cuatro especies de anfibios.
Los invertebrados representaron el 23.8 % de las presas
en numeros pero fueron insignificantes en terminos de
biomasa (2.8%). Serpientes, ranas y roedores fueron las
presas mas frecuentemente entregadas y de mayor con-
tribution a la bioinasa total de presas. Sesenta por ciento

260

Short Communications

Vol. 32, No. 3

de las presas entregadas en los nidos fueron aquellas aso
ciadas a habitats del sotobosque, lo cual sugiere que Buteo
hneatus forrajea extensivamente en este habitat.

[Traduccion de Cesar Marquez]

Acknowledgments

Funding was provided by the Daniel B. Warnell School
of Forest Resources at the University of Georgia and Mc-
Intire Stennis Project No. GEO-0074-MS. We thank Kim
A James for assistance with data collection and species
identification. Special recognition goes to the staff of
Bishop F. Grant Memorial Forest, particularly A. Harris
and J. Gallagher, for housing and on-site logistical sup-
port. We thank C.E. Moorman, who provided field assis-
tance throughout the project, and D.E. Andersen, M.J.
Bechard, J.C. Bednarz, J.J. Dinsmore, A.S. Johnson, D.G.
Krementz, C.D. Marti, and K. Titus who reviewed earlier
drafts of the manuscript.

Literature Cited

Bednarz, J.C. 1979. I. Productivity, nest sites, and habitat
of Red-shouldered and Red-tailed Hawks in Iowa. II.
Status, habitat utilization, and management of Red-
shouldered Hawks in Iowa. M.S. thesis, Iowa State
Univ., Ames, IA U.S.A.

and J.J. Dinsmore. 1985. Flexible dietary response

and feeding ecology of the Red-shouldered Hawk, Bu-
teo Hneatus, in Iowa. Can. Field-Nat. 99:262-264.
Bloom, P.H., M.D. McCrary and M.J. Gibson. 1993. Red-
shouldered Hawk home-range and habitat use in
southern California./. Wildl. Manage. 57:258-265.
Brender, E.V. 1973. Silviculture of loblolly pine in the
Georgia Piedmont. Georgia Forest Research Council
Rep. No. 33. Atlanta, GA U.S.A.

Burleigh, T.D. 1958. Georgia birds. Univ. Oklahoma
Press, Norman, OK U.S.A.

Craighead, JJ- and F.C. Craighead, Jr. 1956. Hawks,
owls, and wildlife. Stackpole Co., Harrisburg, PA
U.S.A.

Dunning, J.B., Jr. 1984. Body weights of 686 species of
North American birds. Monog. No. 1. West. Bird

Banding Assoc., Eldon Publ. Co., Cave Creek, AZ
U.S.A.

Errington, P.L. 1932. Technique of raptor food habits
study. Condor 35:19-29.

Golley, F.B. 1962. Mammals of Georgia. Univ. Georgia
Press, Athens, GA U.S.A.

Henny, C.J., F.S. Schmid, E.L. Martin and L.L. Hood.
1973. Territorial behavior, pesticides, and the popu-
lation ecology of Red-shouldered Hawks in central
Maryland, 1943-1971. Ecology 54:545-554.

Howell, D.L. and B.R. Chapman. 1997. Home range and
habitat use of Red-shouldered Hawks in Georgia. Wil-
son Bull. 109:131-144.

Janik, C.A. and J.A. Mosher. 1982. Breeding biology of
raptors in the Appalachians. Raptor Res. 16:18-24.

Marti, C.D. 1987. Raptor food habits studies. Pages 67-
80 in B.A.G. Pendleton, B.A. Millsap, K.W. Cline and
D.M. Bird [Eds.], Raptor management techniques
manual. Nat. Wildl. Fed., Sci. Tech. Ser. 10. Washing-
ton, DC U.S.A.

Moorman, C.E. and B.R. Chapman. 1996. Nest-site selec-
tion of Red-shouldered Hawks and Red-tailed Hawks
in a managed forest. Wilson Bull. 108:357-368.

Parker, M.A. 1986. The foraging behavior and habitat
use of breeding Red-shouldered Hawks ( Buteo linea-
tus) in southeastern Missouri. M.A. thesis, Univ. Mis-
souri, Columbia, MO U.S.A.

Portnoy, J.W. 1974. Some ecological and behavioral as-
pects of a nesting population of Red-shouldered
Hawks ( Buteo Hneatus Hneatus). M.S. thesis, Univ. Mas-
sachusetts, Amherst, MA U.S.A.

Snyder, N.F.R. and J.W. Wiley. 1976. Sexual size dimor-
phism in hawks and owls of North America. Ornith.
Monogr. 20:1-96.

Steenhof, K. 1983. Prey weights for computing percent
biomass in raptor diets. Raptor Res. 17:15-27.

USDA-SCS. 1965. Soil survey of Morgan County, Georgia.
USDA Soil Conservation Service.

USDA-SCS. 1976. Soil survey of Baldwin, Jones, and Put-
nam counties, Georgia. USDA Soil Conservation Ser-
vice.

Received 10 April 1997; accepted 11 May 1998

Sept 1998

Short Communications

261

J. Raptor Res. 32(3):261-264
© 1998 The Raptor Research Foundation, Inc.

Peregrine Falcons {Falco peregrjnus) Nest in a Quarry and on

Highway Cutbanks in Alaska

Robert J. Ritchie

ABR, Inc. Environmental Research and Services, RO. Box 80410, Fairbanks, AK 99708 U.S.A.

Terry J. Doyle

Tetlin National Wildlife Refuge, RO. Box 779, Tok, AK 99780 U.S.A.

John M. Wright

Alaska Department of Fish and Game, 1300 Mile College Road, Fairbanks, AK 99701 U.S.A.

Key Words: Peregrine Falcon; Falco peregrinus; nesting,

artificial nests', disturbance; Alaska.

In Alaska, the Peregrine Falcon {Falco peregrinus) nests
primarily on cliffs, although nesting on low relief dirt
banks does occur (Cade 1960). As peregrine populations
have recovered and increased in Alaska, sites that once
may have been described as suboptimal for nesting have
been occupied. Further, man-made structures and al-
tered habitats might be expected to attract at least oc-
casional use as densities of these falcons, as well as altered
habitats, continue to increase. Examples of human de-
velopments in Alaska that might attract nesting pere-
grines include towers, quarries, and road cutbanks. Some
of these artificial nesting habitats have been used else-
where in the breeding range of the species (White et al.
1988, Cade et al. 1996), and have been used in Alaska by
Gyrfalcons {Falco rusticolus; White and Roseneau 1970,
Ritchie 1992). In this paper, we describe recent Peregrine
Falcon nesting at two well-trafficked highway cutbanks
along the Alaska Highway in east central Alaska and at a
quarry site on the Seward Peninsula in northwestern
Alaska.

Highway Sites

In 1995, we located a pair of Peregrine Falcons at a
quarried road cut along the Alaska Highway in east-cen-
tral Alaska. The nest (EN) was located on a broad, rocky
ledge (approximately 1 m X 1 m), approximately 10 m
from the base of the cut (Fig. 1). There was no rock
overhang above the nest, but steep rock sides adjacent to
the scrape sheltered the site partially from severe weath-
er. Subsequent visits that year revealed the female incu-
bating two eggs, both of which hatched and both young
fledged. In 1996, an adult, presumably the female, was
observed incubating two eggs but the nest eventually
failed. In 1997, a pair of falcons again occupied this site,
four eggs were incubated, and three young were present
on 31 July.

The second nest (WN) was located along the highway
on a cutbank approximately 300 km west of EN. Al-
though the nest occurred on a well-shaded, natural
ledge, approximately 50 m above the highway (Fig. 2),
the lower third of the cliff had been quarried during
widening of the road. This location was first identified as
a possible nesting location by biologists during a general
bird survey in the area; a pair was observed in May 1997
(M. Ambrose and C. McIntyre pers. comm.). Our obser-
vations that summer revealed that a pair nested on the
bluff and that one young fledged in mid-August.

Peregrine Falcons are a common breeding bird along
the Tanana River and some of its major tributaries adja-
cent to the Alaska Highway. More than 25 pairs occur on
riparian cliffs along the main channel (Bente and Wright
1995). Off-river sites on bluffs bordering older portions
of the floodplain or in upland areas (like EN) are present
but more limited than are sites fronting the main river
channel. Some of these off-river sites have been occu-
pied, but after most cliff habitats along the river had
been used (B. Ritchie, unpubl. data).

Although initial sightings of falcons at these two sites
were not associated with rigorous Peregrine Falcon sur-
veys in the area, we think that it is unlikely that pere-
grines occupied these sites during the previous decade.
The EN site had been checked by one of the authors
since 1988. The WN site is along a route frequented by
biologists interested in peregrine use of cliffs in the area.

Quarry Site

The third site (QN) was located east of Nome, north-
western Alaska, in a rock quarry fronted by a well-main-
tained, two-lane gravel road. The quarry was cut in grad-
uated 6-8 m steps to an elevation of approximately 110
m, on a 185 m tall headland on the coast of the Bering
Sea. Peregrines were first reported there in 1988, when
a pair of adults with one young was observed in late July
(T. Booth and B. Nelson pers. comm.). From 1989-91, a
pair of defensive peregrines was observed each year, but

262

Short Communications

Vol. 32, No. 3

Figure 1. A view of a quarried road cut used by Peregrine Falcons for nesting (EN) along the Alaska Highway.

no young were found during intensive searches. Two
empty nest scrapes were found near the top of the quarry
in niches in vertical step faces of the blasted rock in 1989.
In 1996 and 1997 members of birding tours occasionally
observed single peregrines at the site (P. Bente pers.
comm.) and Common Ravens (Corvus corax) nested in
the quarry.

As of 1991, more than 35 peregrine nest sites were
known for the northwest coast of Alaska, on sea cliffs and
dirt bluffs from the mouth of the ’Yukon River to Cape
Prince of Wales (Bente and Wright 1992) . The closest
neighboring sites were approximately 60 km of our quar-
ry site. Peregrines may have been at the quarry before
1988, but it is unlikely. Birding tours regularly drove by
the site to reach premier birding spots and the quarry
itself was checked occasionally by agency representatives.

Atypical Sites

We think that at least two factors make these sites of
special interest. First, all three sites were located in what
can be described as moderate to high disturbance zones.
Both sites in interior Alaska were along busy highways.
Traffic records for EN identified over 600 vehicles/ day
during May-August (U.S. Customs, ALCAN Station, un-
publ. data), the period during which peregrines would
have courted, laid and incubated eggs, and raised young
(Cade 1960). Traffic at WN undoubtedly was much great-
er due to regular commuter activity associated with near-
by communities. In addition, heavy equipment (e.g., rock
crusher) operated behind EN in the years it was success-
ful.

The road beneath QN was used daily by numerous ve-
hicles after mid-May (1988-91, 1996-97; J. Wright, un-

publ. data). The quarry was in operation periodically
during our visits. In July 1988, a rock crusher was oper-
ating at the base of the quarry within 200 m of the nest
with young. From 1989—91, the quarry was not in oper-
ation when we stopped 2—4 times each summer to check
for peregrines. In 1996 and 1997, blasting and crushing
operations regularly occurred.

While it is true that a few other peregrine nests in Alas-
ka have been found close to roads, the volume of and
proximity to traffic are substantially less at these than lev-
els and distances recorded at EN and WN. Numerous
cases of peregrines using high-traffic areas (e.g., occu-
pied buildings and traffic-laden bridges) have been de-
scribed in urban areas (Cade et al. 1996, Bell et al. 1996).
However, it is likely that many of these birds (especially
in eastern North America) were captive-reared and re-
leased in urban areas (Cade and Bird 1990). These “ur-
ban” birds may have been more tolerant of human activ-
ity. To our knowledge, this phenomenon of using human
altered habitats, within high disturbance areas, has not
been reported for remote and wild populations in North
America.

Confirmed records for Peregrine Falcons nesting on
man-modified structures in Alaska are lacking. Pere-
grines were reported defending a large microwave tower
at a Dewline Site on Barter Island in northern Alaska (D.
Nigro pers. comm.), but proof of nesting was not estab-
lished. In earlier years, Peregrine Falcons had successfully
nested on a coastal dirt bluff within 2 km of the tower
site (F. Mauer pers. comm.). Quarried sites undoubtedly
provide suitable habitat for nesting and may be more at-
tractive where natural cliffs are limited or occupied. As
peregrine populations have recovered elsewhere from

Sept 1998

Short Communications

263

Figure 2. A view from the nest scrape on a quarried
road cut used by Peregrine Falcons for nesting (WN)
along the Alaska Highway.

the low numbers in the 1970s and prime habitats are
reoccupied, quarries have been used (Australia: White et
al. 1988, Olsen 1995; Britain: Ratcliffe 1988; U.S.: Cade
et al. 1996). It is important to note, however, that while
peregrines were successful in nesting at all of our sites
during some years, they only definitely nested in one of
six years at the QN site and failed during incubation in
one of three years at EN. We cannot rule out disturbance
as a factor in unsuccessful nesting.

The three sites described herein are close analogs of
natural cliff habitats in their respective regions. With the
exception of the levels of human disturbance, each pro-
vided basic requirements for successful nesting: good
ledges and protection from ground predators. Addition-
ally, each appeared to be proximal to enough habitats
used by common prey species for successful hunting.
Prey remains gathered at WN and EN were typical of prey
for peregrines in interior Alaska (Cade 1960). For ex-
ample, shorebirds (53%), ducks and gulls (18%) and pas-
serines (29%) comprised the prey items found at the EN
site ( N =17). Similar species have been gathered at Tan-
ana River nests (B. Ritchie, unpubl. data).

As populations of the Peregrine Falcon continue to re-
cover in Alaska, they appear to be increasing in areas
which once were thought to contain suboptimal nesting
sites (e.g., off-river areas, subalpine cliffs). As reported
here, some pairs have been able to use “suboptimal”
(i.e., disturbed areas) cliffs modified by humans. It will
be interesting to monitor the continued recovery of per-
egrines in Alaska as man-modified habitats increase in
distribution, abundance, and diversity. We suspect that a
growing number of birds will tolerate human activity and
attempt to use these areas for nesting.

Resumen. — Falco peregrinus anido exitosamente en tres
habitats alterados de Alaska. Dos de los sitios fueron en
canteras a lo largo de congestionadas autopistas en el
interior de Alaska. El tercer sitio ocupado fue una de las
caras de una cantera activa en el noreste de Alaska. Estas
son las primeras observaciones registradas de halcones
peregrinos anidando en este tipo de habitats en lugares
remotos como Alaska. Aun mas, todos los sitios estaban
cercanos a carreteras con bastante trafico. De continuar
en aumento el numero de peregrinos y de sitios alterados
en Alaska, sospechamos que mas halcones peregrinos
utilizaran estos habitats y estructuras.

[Traduccion de Cesar Marquez]

Acknowledgments

The authors would like to thank Clayton White, Ted
Swem and Dan Varland for reviewing a draft of this paper
and making insightful comments for its improvements.
We would also like to thank a number of people who
provided their observations of Peregrine Falcons at sites
described in our paper: Michelle Ambrose; Peter Bente
and Bob Nelson, Alaska Department of Fish and Game;
Carol McIntyre, National Park Service; Tony Booth, Fran
Mauer, and Henry Timm, U.S. Fish and Wildlife Service;
and Debbie Nigro, ABR, Inc.

Literature Cited

Bell, D.A., D.P. Gregoire and B.J. Walton. 1996. Bridge
use by Peregrine Falcons in the San Francisco Bay
Area. Pages 15-24 in D.M. Bird, D.E. Varland and J.J.
Negro [Eds.], Raptors in human landscapes. Academ-
ic Press, London, U.K.

Bente, PJ- and J.M. Wright. 1992. Documentation of ac-
tive Peregrine Falcon nest sites. Alaska Dept. Fish and
Game, Fed. Aid in Wildl. Restor., Final Rep., Proj. SE
2-6, Juneau, AK U.S.A.

and . 1995. Documentation of Peregrine

Falcon nest sites in relation to state land use propos-
als. Alaska Dept. Fish and Game, Fed. Aid in Wildl.
Restor., Final Rep. Proj. SE 2-8, Part II, Juneau, AK
U.S. A.

Cade, T.J. 1960. The ecology of peregrines and Gyrfal-
cons in Alaska. Univ. Calif. Publ. Zool. 63:151-290.
and D.M. Bird. 1990. Peregrine Falcons, Fa Ico per-
egrinus, nesting in an urban environment: a review.
Can. Field-Nat. 104:209—218.

264

Short Communications

Vol. 32, No. 3

, M. Martell, P. Redig, G. Septon and H. Tor-

doff. 1996. Peregrine Falcons in urban North Amer-
ica. Pages 3-13 in D.M. Bird, D.E. Varland and J.J.
Negro [Eds.], Raptors in human landscapes. Academ-
ic Press, London, U.K.

Olsen, P. 1995. Australian birds of prey: the biology and
ecology of raptors. John Hopkins Univ. Press, Balti-
more, MD U.S.A.

Ratcliffe, D.A. 1988. The peregrine population of Great
Britain and Ireland, 1965-1985. Pages 147-157 mT.J.
Cade, J.H. Enderson, C.G. Thelander and C.M. White
[Eds.], Peregrine Falcon populations: their manage-

ment and recovery. The Peregrine Fund, Inc., Boise,
ID U.S.A.

Ritchie, R.J. 1992. Effects of oil development on provid-
ing nesting opportunities for Gyrfalcons and Rough-
legged Hawks in northern Alaska. Condor 93:180-184

White, C.M. and D.G. Roseneau. 1970. Observations on
food, nesting, and winter populations of large North
American falcons. Condor 72:13—115.

, W.B. Emison and W.M. Bren. 1988. Atypical nest-
ing habitat of the Peregrine Falcon in Victoria, Aus-
tralia. /. Raptor Res. 22:37-43.

Received 29 December 1997; accepted 17 May 1998

J. Raptor Res. 32(3):264-266
© 1998 The Raptor Research Foundation, Inc.

Lice (Phthiraptera: Amblycera, Ischnocera) of Raptors in Hungarian Zoos and

Rehabilitation Centers

SZABOLCS SOLT

H-9083 Ecs, Petofi u. 60., Hungary

Key WORDS: Phthiraptera', louse infestations', rehabilitation

centers', injured raptors, Hungary.

Here, I describe louse (Phthiraptera: Amblycera, Is-
chnocera) infestations of raptors kept in zoos and reha-
bilitation centers in Hungary and conclude that injuries
increase the frequency and extent of such infestations.
Fifty-five individuals of 18 species of raptors from the fam-
ilies Accipitridae, Falconidae, Tytonidae, and Strigidae
were examined in 1995-96 at the Zoological Parks of
Gyor and Veszprem (14 and 12 birds, respectively), the
raptor rehabilitation center of Ferto-Hansag National
Park at Koszeg (11 birds), and the rehabilitation center
of Hortobagy National Park at Gorestanya (18 birds).
There was no regular use of insecticides to control louse
infestations at any of these sites. Many of the raptors were
badly injured by electrocution from high voltage trans-
mission lines or by illegal shooting. Injuries often result-
ed in extensive damage to wings and legs.

Lice were collected using forceps during 10-min visual
examinations while the birds were immobilized by assis-
tants. Twenty birds (36%) were found to carry lice re-
sulting in a total of 373 lice (86 males, 196 females, and
91 nymphs. Table 1) collected. Eight species office were
found, two of which were typical parasites of galliform
hosts and presumably originated from dead chickens sup-
plied as food.

Avian grooming partially serves to control the spread
of ectoparasites (Marshall 1981). Grooming, such as foot

scratching, eliminates lice on the head and bill preening
removes lice from other body parts (Clayton 1991, Rozsa
1993) . Since I assumed that birds with major limb injuries
were presumed to preen less frequently, I compared the
numbers of lice on 11 injured and 28 healthy raptors.
Because of the aggregated distribution of lice on differ-
ent individuals (Rekasi et al. 1997), I used a one-tailed
Mann-Whitney U-test as a nonparametric statistic. Statis-
tical analyses were carried out by InStat 2.01.

Avian lice can be viewed as representatives of a single
ecological guild of ectoparasites. Thus, their abundance
can be expressed as total louse numbers (belonging to
any species) living on the same individual bird (Rozsa
1997). When comparisons were made between raptors
with damaged limbs versus intact limbs, there was a sig-
nificant difference in total louse abundance (U = 48.5,
P< 0.001, Fig. 1). Lice also show considerable site-spec-
ificity on their hosts (e.g., Perez et al. 1996); therefore
different louse taxa should show different responses to
decreased grooming abilities of injured raptors. Species
of the genus Craspedorrhynchus, for instance, are typically
distributed on the head and nape of raptors and birds
scratch using their feet to remove them (Gallego et al
1987). Limb-damaged birds naturally show much less
foot scratching, either because they lack the use of one
leg which prohibits them from reaching their heads with
the other one, or because they have broken wings which
distorts normal foot scratching movements at least on
one side of the body. In fact, the abundance of Craspe-

Sept 1998

Short Communications

265

Table 1. Lice collected from healthy and injured raptors in Hungarian zoos and rehabilitation centers.

Raptor Species

No. Healthy
Birds

No. Injured
Birds

Louse Species

No. From
Healthy
Birds

No. From
Injured
Birds

Accipiter gentilis

1

1

Menopon gallinae 1

0

2

Buteo buteo

11

6

Craspedorrhynchus

8

102

platystomus

Colpocephalum

0

31

buteonis

Degeeriella fulva

7

35

Menopon gallinaF-

0

4

Lipeurus caponis*

0

1

Buteo rufinus

1

0

Craspedorrhynchus

8

0

platystomus

Degeeriella fulva

20

0

Circus aeruginosus

3

0

—

0

0

Circaetus gallicus

1

0

—

0

0

Falco peregrinus

1

0

—

0

0

Falco subbuteo

0

1

—

0

0

Falco tinnunculus

8

1

Degeeriella rufa

40

27

Aquila heliaca

0

1

Degeeriella fulva

0

2

Aquila pomarina

0

1

Craspedorrhynchus

0

73

naevius

Hieraetus pennatus

1

0

Degeeriella fulva

12

0

Theratopius ecaudatus

1

0

—

0

0

Asio otus

2

0

—

0

0

Athene noctua

4

0

—

0

0

Bubo bubo

1

0

—

0

0

Nyctea scandiaca

2

0

—

0

0

Stryx aluco

2

0

—

0

0

Tyto alba

5

0

Kurodaia

1

0

subpachygaster

a Lice specific to galliforms.

60

50 -

o

3 40-

30

<D

o

c

to

TJ

C

D

n

<

W 20-

□ limb-damaged birds
■ intact birds

10 -

1

X.

a)

u

3

-C

O

c

>>

-C

h_

o

XJ

Si

tfl

2

a

3

<J

>N

O

T3

SL

0)

<D

O)

o

a)

a)

co

v

a

Figure 1. Mean number (±SD) of all lice and only lice belonging to the genera Craspedorrhynchus and Degeeriella on
limb-damaged (11 birds) and healthy raptors (28 birds).

266

Short Communications

Vol. 32, No. 3

dorrhynchus lice was significantly greater in injured versus
healthy raptors (U = 101.5, P< 0.05).

Contrastingly, lice of the genus Degeeriella are typically
distributed on the wings, especially the primary and sec-
ondary feathers. Raptors typically preen with their beaks
to eliminate them (Clay 1958). Limb-damaged birds usu-
ally exhibit normal preening activities except for preen-
ing of broken wings. Degeeriella lice appeared to be more
abundant on limb-damaged raptors in my examination,
though this difference was not significant (U = 119.5, P
> 0 . 10 ).

The correlation between major injuries of raptor limbs
and increased louse infestations seemed to be related to
the impairment of their grooming behavior. Site-specific
lice such as those that infest the heads of raptors were
found to be common. Since wing- or leg-damaged raptors
are at times important for raptor breeding and repatria-
tion programs, it is important that they be routinely ex-
amined for ectoparasite infestations and may need spe-
cial attention to control their lice.

Resumen. — Cincuenta y cinco individuos de 18 especies
de aves rapaces fueron examinados con el fin de encon-
trar infestaciones de piojos (Phthiraptera: Amblycera, Is-
chnocera) en Hungria en 1995 y 1996. Un total de 373
piojos de 8 especies fueron encontrados en 20 aves. Los
danos mayores causados en alas y patas como electrocu-
taciones, aparentemente lograron disminuir la habilidad
de acicalarse de las aves aumentando la frecuencia de los
piojos. Debido a que estas aves fueron mas propensas a
ser infestadas por ectoparasitos, especialmente aquellas
con problemas en alas y patas, hubo que prestarles mayor
atencion para el control de piojos.

[Traduccion de Cesar Marquez]

Acknowledgments

I thank the staff of Zoos and National Parks for pro-
viding access to captive raptors, Jozsef Rekasi for his help

in louse identification and Lajos Rozsa for comments on

the manuscripts.

Literature Cited

Clay, T. 1958. Revisions of Mallophaga genera. Degeeriella
from the Falconiformes. Bull. Brit. Mus. (Nat. Hist.)
Entomol. 7:5-207.

Clayton, D.H. 1991. Coevolution of avian grooming and
ectoparasite avoidance. Pages 258-289 in J. E. Loye
and M. Zuk [Eds.], Bird-parasite interactions. Oxford
Univ. Press, Oxford, U.K.

Gallego, J., M.P.M. Mateo and J.M. Aguirre. 1987. Mal-
ofagos de rapaces Espanolas. II. Las especies del ge-
nero Craspedorrhynchus Keler, 1938 parasitas de falcon-
iformes, con descripcion de tres especies nuevas. Eos
63:31-66.

Marshall, A.G. 1981. The ecology of ectoparasitic in-
sects. Academic Press, London, U.K.

Perez, J.M, I. Ruiz-Martinez and J.E. Cooper. 1996. Oc-
currence of chewing lice on Spanish raptors. Ardeola
43:129-138.

Rekasi, J., L. Rozsa and J.B. Kiss. 1997. Patterns in the
distribution of avian lice (Phthiraptera: Amblycera, Is-
chnocera)./ Avian Biol. 28:150-156.

Rozsa, L. 1993. An experimental test of the site-specificity
of preening to control lice in feral pigeons./. Parasitol.
79:968-970.

. 1997. Patterns in the abundance of avian lice

(Phthiraptera: Amblycera, Ischnocera) . J. Avian Biol
28:249-254.

Received 21 November 1997; accepted 12 May 1998

Letters

J. Raptor Res. 32(3):267

© 1998 The Raptor Research Foundation, Inc.

Evidence of Spotted Kestrel ( Falco moluccensis ) Nesting in the Roofs of

Sumba’s Traditional Houses

On the Island of Sumba, Indonesia, the persistent, animist religion of ancestor worship has widely preserved the
traditional thatched house structure. These houses have a striking roof that is low-sided but high-peaked. The houses
associated with clan ancestors have higher roofs and are preferably placed on hilltops. Traditional villages are found
scattered throughout the rolling country, much of which is used for extensive cattle and horse raising, but ancestral
houses are also in small towns.

During a tour of Sumba, from 10-14 August 1997, I made short visits to nine traditional villages. In five of these,
and also near traditional houses in the center of Waikabubak (the second largest town of Sumba with a population
of about 15 000), I found Spotted Kestrels ( Falco moluccensis). I found no more than one pair at each location and
the kestrels called from trees close to the houses or hunted in adjacent fields. In the village of Praigoli, southeast of
Waikabubak, I also observed an adult kestrel enter the top of a roof and emerge after some seconds to stand near
the presumed nesting cavity. About 10 km to the west, I saw similar behavior in a village near Morossi Beach; on this
occasion the kestrel entered the roof top with a large orthopteran in its bill and it stayed inside longer. Both houses
were occupied, but the upper part of the Sumbanese houses are undisturbed because the owners believe them to be
places reserved for the spirits of their ancestors. I was intrigued by some possible link between this belief and the
presence of kestrels, but an apparently well-informed local guide was unable to give me any relevant information on
this subject.

Although the Spotted Kestrel is already listed among the raptors that are attracted to towns by opportunities to
nest in buildings (Brown 1976, Birds of prey, Ross International Books, Ltd., London, U.K.), finding it nesting in
traditional villages suggests much older association with human dwellings. The distribution of this Indonesian en-
demic is centered on the biogeographic region called Wallacea. The extensive grassland that is found in this region,
which is especially vast on Sumba, may be the result of human activity but has existed long enough to support a
distinctive bird fauna (White and Bruce 1986, The Birds of Wallacea, B.O.U. Checklist No. 7, London, U.K.) . Although
it is rather an opportunistic species, the Spotted Kestrel has shown its preference for open habitats in Wallacea in
the past (e.g., Rensch 1931, Mitt. Zool, Mus. Berlin 17:451-637) when the region was more forested.

During my tour, Sumba’s grassland seemed especially rich in orthopteran prey. However, the Spotted Kestrel
probably has to cope with a scarcity of prominent rocks to nest on Sumba, especially where I saw evidence of nesting
birds. Also, trees with hollows are scarce in these grasslands. High-peaked thatched roofs situated on hilltops probably
make up for this deficiency. Unfortunately for kestrels, the thatching practice is being challenged by longer-lasting,
though thermally less insulating, sheet-iron roofs.

I found no previous report of this apparently common nesting habit. This may be because the Spotted Kestrel
looks similar to the Common Kestrel {Falco tinnunculus) and did not draw the attention of early field naturalists,
Sumba is still visited infrequently by tourists who might tend to report these birds; Sumba’s traditional villages offer
many other attractions that might distract people from the species, and birders avoid human settlements with Spotted
Kestrels to search for more “memorable” species. In fact, the Spotted Kestrel remains a little known species over all
of its range.

I thank T. Cade, D.E. Varland, and an anonymous reviewer for text improvement. — Tiziano Londei, Dipartimento
di Biologia, Universita di Milano, Via Celoria 26, 20133 Milano, Italy.

267

268

Letters

Yol. 32, No. 3

J. Raptor Res. 32(3) :268

© 1998 The Raptor Research Foundation, Inc.

Caribou Antlers as Nest Materials for Golden Eagles in Northwestern Alaska

There are few published records of antlers in Golden Eagle ( Aquila chrysaetos ) nests (three for Scotland [L.
MacNally, 1977, The ways of an eagle, Collins and Harvill Press, London, U.K. and S. Gordon, 1955, The golden
eagle: king of birds, Citadel Press, New York, NY U.S.A.] and an unspecified number for Colorado [R. Olendorff,
1975, Golden eagle country, Alfred A. Knopf, New York NYU.S.A.]). We found many caribou (Rangifer tarandus)
antlers in three Golden Eagle nests in the Cape Kruzenstern region (67-68°N, 163-164°W) of northwestern Alaska.
At one nest that was situated at 270 m elevation on a cliff face, three young and two adults were present in 1986.
This nest contained numerous male and female antlers composing 10-15% of the matrix of the nest. The second
nest in which antlers were found was situated 240 m elevation atop a 2-m cliff in 1988. We counted 21 antlers, some
of which were very large male antlers, all in the perimeter of the nest (ca 10-15% of the bulk). The third nest was
situated at 300 m elevation on a cliff face in 1988 and it contained numerous antlers, mostly of female caribou (ca
10% of nest bulk). We did not find antlers in one other Golden Eagle nest on the Cape and in five other nests along
the Noatak River immediately to the east.

Because all three nests were within 13 km of each other and no two were simultaneously occupied, there was a
possibility that all were alternate nests of the same pair of eagles with a preference for antlers. Another explanation
for the local importance of antlers is that the lack of sizable sticks, the similarity of antlers to sticks, and the abundance
of antlers, especially the much smaller female caribou antlers, lead the eagles to substitute antlers for sticks. C.M.
White and J.R. Haugh (pers. comm.) saw parts of antlers in a nest in northwestern Alaska and Y. Potopov (pers.
comm.) found a small caribou antler in a tree nest along the Kolyma River in northeastern Siberia. We also found
a single deer (probably Odocoileus hemionus) antler in a nest in Montana. — David H. Ellis, USGS Patuxent Wildlife
Research Center, HCR 1 Box 4420, Oracle, A Z 85623 U.S A. and Richard L. Bunn, AFZC-ECM-NR (Wildlife), Building
302, Fort Carson, CO 80913-5000 U.S.A.

THE RAPTOR RESEARCH FOUNDATION, INC.

(Founded 1966 )

OFFICERS

PRESIDENT: Michael N. Kochert
VICE-PRESIDENT: David E. Andersen

SECRETARY: Patricia A. Hall
TREASURER: Jim Fitzpatrick

BOARD OF DIRECTORS

NORTH AMERICAN DIRECTOR #1:
Brian A. Millsap

NORTH AMERICAN DIRECTOR #2:
Petra Bohall Wood

INTERNATIONAL DIRECTOR #3:

Michael McGrady

DIRECTOR AT LARGE #1: Patricia L. Kennedy
DIRECTOR AT LARGE #2: John A. Smallwood
DIRECTOR AT LARGE #3: James C. Bednarz
DIRECTOR AT LARGE #4: Cesar MArquez Reyes
DIRECTOR AT LARGE #5: Lloyd Kiff
DIRECTOR AT LARGE #6: Robert Ken ward

NORTH AMERICAN DIRECTOR #3:
Karen Steenhof

INTERNATIONAL DIRECTOR #1:
Massimo Pandqlfi

INTERNATIONAL DIRECTOR #2:
Reuven Yosef

EDITORIAL STAFF

EDITOR: MarcJ. Bechard, Department of Biology, Boise State University, Boise, ID 83725 U.S.A.

BOOK REVIEW EDITOR: Jeffrey S. Marks, Montana Cooperative Research Unit, University of Montana,

Missoula, MT 59812 U.S.A.

SPECIAL PUBLICATIONS EDITOR: Daniel E. Varland, Rayonier, 3033 Ingram Street, Hoquiam, WA

98550

SPANISH EDITOR: Cesar Marquez Reyes, Instituto Humbolist Colombia, AA. 094766, Bogota 8, Colombia

The Journal of Raptor Research is distributed quarterly to all current members. Original manuscripts
dealing with the biology and conservation of diurnal and nocturnal birds of prey are welcomed from
throughout the world, but must be written in English. Submissions can be in the form of research articles,
letters to the editor, thesis abstracts and book reviews. Contributors should submit a typewritten original
and three copies to the Editor. All submissions must be typewritten and double-spaced on one side of 216
X 278 mm (8 Vi X 11 in.) or standard international, white, bond paper, with 25 mm (1 in.) margins. The
cover page should contain a title, the author’s full name(s) and address (es). Name and address should be
centered on the cover page. If the current address is different, indicate this via a footnote. A short version
of the tide, not exceeding 35 characters, should be provided for a running head. An abstract of about
250 words should accompany all research articles on a separate page.

Tables, one to a page, should be double-spaced throughout and be assigned consecutive Arabic numer-
als. Collect all figure legends on a separate page. Each illustration should be centered on a single page
and be no smaller than final size and no larger than twice final size. The name of the author(s) and figure
number, assigned consecutively using Arabic numerals, should be pencilled on the back of each figure.

Names for birds should follow the A.O.U. Checklist of North American Birds (6th ed., 1983) or another
authoritative source for other regions. Subspecific identification should be cited only when pertinent to
the material presented. Metric units should be used for all measurements. Use the 24-hour clock (e.g.,
0830 H and 2030 H) and “continental” dating (e.g., 1 January 1990).

Refer to a recent issue of the journal for details in format. Explicit instructions and publication policy
are oudined in “Information for contributors,” / Raptor Res., Vol. 27(4), and are available from the editor.

ASSOCIATE EDITORS

Allen M. Fish
Gary R. Bortolotti
Charles J. Henny

Fabian Jaksic
Daniel E. Varland

The Raptor Research Foundation, Inc. gratefully acknowledges a grant and logistical support
provided by Boise State University to assist in the publication of the journal.

Persons interested in predatory birds are invited to join The Raptor Research Foundation, Inc. Send
requests for information concerning membership, subscriptions, special publications, or change of address
to OSNA, P.O. Box 1897, Lawrence, KS 66044-8897, U.S.A.

The Journal of Raptor Research (ISSN 0892-1016) is published quarterly and available to individuals for $33.00
per year and to libraries and institutions for $50.00 per year from The Raptor Research Foundation, Inc., 14377
117th Street South, Hastings, Minnesota 55033, U.SA. (Add $3 for destinations outside of the continental United
States.) Periodicals postage paid at Hastings, Minnesota, and additional mailing offices. POSTMASTER: Send
address changes to The Journal of Raptor Research, OSNA, P.O. Box 1897, Lawrence, KS 66044-8897, U.SA.

Printed by Allen Press, Inc., Lawrence, Kansas, U.SA

Copyright 1998 by The Raptor Research Foundation, Inc. Printed in U.SA

® This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).

Raptor Research Foundation, Inc., Awards
Recognition for Significant Contributions 1

The Dean Amadon Award recognizes an individual who has made significant contributions in the field of
systematics or distribution of raptors. Contact: Dr. Clayton White, 161 WIDE, Department of Zoology,
Brigham Young University, Provo, UT 84602 U.SA. Deadline August 15.

The Tom Cade Award recognizes an individual who has made significant advances in the area of captive
propagation and reintroduction of raptors. Contact: Dr. Brian Walton, Predatory Bird Research Group,
Lower Quarry, University of California, Santa Cruz, CA 95064 U.SA. Deadline: August 15.

The Fran and Frederick Hamerstrom Award recognizes an individual who has contributed significantly to
the understanding of raptor ecology and natural history. Contact: Dr. David E. Andersen, Department
of Fisheries and Wildlife, 200 Hodson Hall, 1980 Folwell Avenue, University of Minnesota, St. Paul,
MN 55108 U.S.A Deadline: August 15.

Recognition and Travel Assistance

The James R. Koplin Travel Award is given to a student who is the senior author of the paper to be
presented at the meeting for which travel funds are requested. Contact: Dr. Petra Wood, West Vir ginia
Cooperative Fish and Wildlife Research Unit, P.O. Box 6125, Percival Hall, Room 333, Morgantown,
WV 26506-6125 U.SA. Deadline: established for conference paper abstracts.

The William C. Andersen Memorial Award is given to the student who presents the best paper at the annual
Raptor Research Foundation Meeting. Contact: Ms. Laurie Goodrich, Hawk Mountain Sanctuary, Rural
Route 2, Box 191, Kempton, PA 19529-9449 U.SA. Deadline: Deadline established for meeting paper
abstracts.

Grants 2

The Stephen R. Tully Memorial Grant for $500 is given to support research, management and conservation
of raptors, especially to students and amateurs with limited access to alternative funding. Contact: Dr.

Kimberly Titus, Alaska Division of Wildlife Conservation, P.O. Box 20, Douglas, AK 99824 U.SA. Dead-
line: September 10.

The Leslie Brown Memorial Grant for $500-$l,000 is given to support research and/or the dissemination
of information on raptors, especially to individuals carrying out work in Africa. Contact: Dr. Jeffrey L.
Lincer, 15644 Kingman Rd., Poway, CA 92064 U.SA. Deadline: September 15.

1 Nominations should include: (1) the name, tide and address of both nominee and nominator, (2) the
names of three persons qualified to evaluate the nominee’s scientific contribution, (3) a brief (one page)
summary of the scientific contribution of the nominee.

2 Send 5 copies of a proposal (^5 pages) describing the applicant’s background, study goals and methods,
anticipated budget, and other funding.