The Journal of
Raptor Research

Volume 34 Number 1 March 2000

•r> A'r

■J'-

■L

JL^a

3 "^

i

v-^

Published by
he Raptor Research Foundation, Inc.

Contents

First-year Movements by Juvenile Mexican Spotted Owls in the Canyonlands of

Utah. David W. Willey and Charles van Riper III 1

Distribution, Population Size and Habitat Use of the Reunion Marsh Harrier,

Circus M. MAILIARDI. Vincent Bretagnolle, Thomas Ghestemme, Jean-Marc Thiollay and Carole
Attie 8

Nest-site Selection and Reproductive Success of Urban Red-shouldered Hawks
IN Central California. Stephen c. Rottenbom 18

Habitat Use by Red-tailed Hawks Wintering in the Delta Region of Arkansas.

Jleath D. Gamer and James C. Bednarz 26

An Evaluation of the Andean Condor Population in Northern Ecuador. Marcus

T Koenen, Sarah Gale Koenen and Norma Yanez 33

Short Communications

Summary of Philippine Eagle Reproductive Success, 1978-98. Hector C. Miranda, Jr., Dennis I.

Salvador, Jayson C. Ibanez and Gliceria A. Balaquit-Ibahez 37

Diet of Autumn Migrating Northern Saw-whet Owls on the Eastern Shore of Virginia. David M.

Whalen, Bryan D. Watts and David W. Johnston 42

Breeding Biology of the Eurasian Kestrel in the Steppes of Southwestern Spain. J.M. Aviles, J.M.

Sanchez and A. Sanchez 45

Breeding Densities and Habitat Attributes of Golden Eagles in Southeastern Spain. Martina

Carrete, Jose Antonio Sanchez-Zapata and Jose Francisco Calvo 48

Trapping Eshmates for Saker and Peregrine Falcons Used for Falconry in the United Arab Emirates.

Nigel W.H. Barton 53

First Confirmed Breeding Records and Other Incidental Sightings of Northern Harriers in

Labrador. Tony E. Chubbs, Bmce Mactavish, Keith Oram and Perry G. Trimper 56

Letters 58

Commentary 62

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.

THE RAPTOR RESEARCH FOUNDATION, INC.

(Founded 1966 )

OFFICERS

PRESIDENT: Michael N. Kochert SECRETARY: Patricia A. Hall

VICE-PRESIDENT: Keith L. Bildstein TREASURER: Jim Fitzpatrick

BOARD

NORTH AMERICAN DIRECTOR #1:

Philip Detrich

NORTH AMERICAN DIRECTOR #2:

Petra Bohat .l Wood
NORTH AMERICAN DIRECTOR #3:

Robert Lehman

INTERNATIONAL DIRECTOR #1:

Eduardo Inigo-Elias
INTERNATIONAL DIRECTOR #2:

Reuven Yosef

OF DIRECTORS

INTERNATIONAL DIRECTOR #3:

Beatriz Arroyo

DIRECTOR AT LARGE #1: Jemima ParryJones
DIRECTOR AT LARGE #2: Robert Kenward
DIRECTOR AT LARGE #3: Michael W. Gollopy
DIRECTOR AT LARGE #4; Miguel Ferrer
DIRECTOR AT LARGE #5: JOHN A. Smallwood
DIRECTOR AT LARGE #6: Brian A. Milsap

>I< >!<• *4.* •d** *1* *1# >Lr >!<

^ ^ ^ •T* •T* •T' ^ ^

EDITORIAL STAFF

EDITOR: MarcJ. Bechard, Department of Biology, Boise State University, Boise, ID 83725 U.S.A.

ASSOCIATE EDITORS

Aixen M. Fish Fabian Jaksic

Juan Jose Negro Daniel E. Variand

Chari.es J. Henny Cole Crocker-Bedford

Ian G. Warkentin James C. Bednarz

BOOK REVIEW EDITOR: Jeffrey S. Marks, Montana Cooperative Research Unit, University of Montana,

Missoula, MT 59812 U.S.A.

SPANISH EDITOR: Cesar Marquez Reyes, Institute Humboldt, Colombia, AA. 094766, Bogota 8, Colombia
EDITORIAL ASSISTANTS: Joan Clark, Keleigh Hague-Bechard, Elise Vernon Schmidt

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 (814 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 (7th ed., 1998) 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,” y. Raptor Res., Vol. 33(4), and are available from the editor.

COVER: Philippine Eagle {Pithecophaga jefferyi) . Painting by Dana Gardner.

THE JOURNAL OF RAPTOR RESEARCH

A QUARTERLY PUBLICATION OF THE RAPTOR RESEARCH FOUNDATION, INC.

VoL. 34 March 2000 No. 1

J Raptor Res. 34(1):1— 7

© 2000 The Raptor Research Foundation, Inc.

FIRST-YEAR MOVEMENTS BY JUVENILE MEXICAN SPOTTED
OWLS IN THE CAOTONLANDS OF UTAH

David W. Willey^ and Charles van Riper III

USGS-BRD Forest and Rangeland Ecosystem Science Center, Colorado Plateau Field Station, Northern Arizona University,

Box 5614, Flagstaff, AZ 86011 U.S.A.

Abstract. — We studied first-year movements of Mexican Spotted Owls {Strix occidentalis lucida) during
natal dispersal in canyonlands of southern Utah. Thirty-one juvenile Mexican Spotted Owls were cap-
tured and radio tracked during 1992-95 to examine behavior and conduct experiments related to the
onset of natal dispersal. Juvenile Spotted Owls dispersed from their nest areas during September to
October each year, with 85% leaving in September. The onset of movements was sudden and Juveniles
dispersed in varied directions. The median distance from nest area to last observed location was 25.7
km (range = 1.7-92-3 km). Three of 26 Juveniles tracked (11%) were alive after one year, although
none were observed with mates. We conducted a feeding experiment, using Mongolian gerbils (Meriones
unguicuculatus) , to test the influence of increased food supply on dispersal onset. The mean dispersal
date of five owls that received supplemental food (Julian day no. 255 ± 2.6 SD) was significantly different
than a control group (day no. 273 ± 12.3).

Key Words: Mexican Spotted Owl, Strix occidentalis lucida; natal dispersal, telemetry, dispersal timing, can-

yonlands.

Movimientos del primer ano de Juveniles de Strix occidentalis lucida durante la dispersion en tierras del
Canon al sur de Utah

Resumen. — Estudiamos los movimientos del primer ano de Strix occidentalis lucida durante la dispersion
natal en tierras del Canon al sur de Utah. Treinta y un buhos Juveniles fueron capturados y dotados de
radiotransmisores durante 1992—95 con el fin de examinar su comportamiento y conducir experimentos
relacionados con la iniciacion de la dispersion natal. Los buhos Juveniles se dispersaron de sus areas de
anidacion durante Septiembre a Octubre de cada ano, con un 85% partiendo en Septiembre. La ini-
ciacion de los movimientos fue repen tina y los Juveniles se dispersaron en varias direcciones. La distancia
media del area del nido a la ultima localizacion observada fue de 25.7 (rango = 1.7-92.3 km). Tres de
los 26 Juveniles monitoreados (11%) estaban vivos despues de un ano, aunque ninguno fue observado
con companero. Conducimos un experimento de alimentacion utilizando Meriones unguicuculatus, para
probar la influencia del incremento en comida con relacion a la dispersion. La fecha de dispersion
media de cinco buhos que recibieron suplemento alimenticio (dfa no. 255 ± 2.6 SD) fue significativa-
mente diferente que el grupo de control (dfa no. 273 ± 12.3).

[Traduccion de Cesar Marquez]

Study of the dispersal movements by raptors us-
ing radiotelemetry allows researchers to track in-

^ Present address: Department of Biology and Wildlife,
University of Alaska, Fairbanks, 211 Irving I, Fairbanks,
AK 99775 U.S.A.

dividuals over prolonged time periods (e.g., Ken-
ward et al. 1993). Using telemetry in Arizona,
Ganey et al. (1998) showed that Juvenile Mexican
Spotted Owls {Strix occidentalis lucida) began dis-
persal each year during September, and that early
movements were abrupt and in random directions

1

2

Willey and van Riper

VoL. 34, No. 1

from natal areas. In the Pacific Northwest, Spotted
Owls (5. o. caurina and occidentalis) showed similar
first-year movement patterns (Gutierrez et al.
1995).

Although numerous studies of Spotted Owls
have reported on early dispersal movements (Gu-
tierrez et al. 1995), few individuals have been ob-
served to settle and find mates. This is likely due
to the expense and difficulty of tracking birds over
long distances and time periods, and due to high
mortality rates. Because natal dispersal is defined
as the movement of an individual from its birth site
to the place where it reproduces (Howard 1960),
few studies have examined true dispersal in raptors
(Koenig et al. 1996). Only studies of Eastern
Screech-Owls {Otus asio) and Burrowing Owls
{Athene cunicularia) have used an experimental ap-
proach to assess proximate factors responsible for
the onset of dispersal (Ritchison et al. 1992, Belt-
hoff and Dufty 1998, King 1996).

In southern Utah, Mexican Spotted Owls occupy
arid rocky canyon habitat within a matrix of des-
ertscrub vegetation communities (Brown 1982,
Willey 1995). Our goal was to examine aspects of
Mexican Spotted Owl natal dispersal timing and
preliminary movements in Utah, and attempt to
track the owls using radiotelemetry from aircraft as
long as possible. Our objectives were to describe
first-year movement patterns by juvenile Spotted
Owls and conduct field experiments to examine
proximate mechanisms that influence the timing
of natal dispersal by owls.

Study Area and Methods

Field work was conducted at four study areas in south-
ern Utah: Zion, Capitol Reef and Canyonlands National
Parks, and on the Manti LaSal National Forest (Fig. 1).
The Zion National Park study area was located 1 km
north of Springdale, in southwestern Utah. Capitol Reef
was located 25 km northeast of Torrey in southcentral
Utah. Canyonlands National Park was located 45 km
southwest of Moab, Utah and the Manti LaSal was located
on Elk Ridge 35 km east of Blanding in southeastern
Utah. All areas were characterized by steep sandstone
canyons isolated among alluvial valleys and uplifted pla-
teaus (Thornbury 1965). Elevations ranged from 1109-
3960 m. Total annual precipitation averaged 17 cm per
yr and temperatures ranged seasonally from <0->40°C
(U.S. Weather Btireau, Climate and Precipitation Sum-
maries, Utah). Pinyonjuniper {Pinus edulis-Juniperus utah-
ensis) woodland and desertscrub vegetation dominated
the landscape matrix (Brown 1982). Canyon vegetation
included small patches of ponderosa pine {Pinus ponde-
rosa), Douglas-fir {Pseudotsuga menziesii), white fir {Abies
concolor), pinyon pine {Pinus monticula) and Utah juniper
{funiperus osteosperma) . Box elder {Acer negundo) , bigtooth

Figure 1. Location of four primary study areas where
Mexican Spotted Owl natal dispersal was investigated in
Utah during 1992-95.

maple {A. grandidentatum) , Fremont cottonwood {Populus
fremontii) and willows {Salix spp.) were present along me-
sic canyon bottoms and cliff seeps.

Juvenile Spotted Owls were trapped during 1-15 Au-
gust each year of the study using Bal-Chatri traps or
noose poles (Forsman 1983). Radiotransmitters (model
R1-2C, Holohil Inc., Ontario, Canada), weighing 5. 5-6.0
g with signal life ranging from 12-24 mo, were attached
to the two central rectrices using quick-set epoxy and un-
waxed dental floss (Reid et al. 1996).

We received radio signals using TR-4 and TR-2 receiv-
ers with handheld or airplane mounted H-antennas (Te-
lonics Inc., Mesa, Arizona). Locations were based on ae-
rial fixes or visual observation of roosting owls. We
recorded the Julian date, time and Universal Trans Mer-
cator coordinates for each owl location. From early Au-
gust until the owls dispersed, we attempted to relocate
owls every day. Once the owls began large movements
outside of the natal area, we relocated them every 3-7 d.
When owls could not be located by ground tracking, we

March 2000

Dispersal Movements by Spotted Owls

3

searched out 100 km in all directions from the last known
location with fixed-wing aircraft.

We defined the beginning of Spotted Owl natal dis-
persal as the time when an individual initiated movement
that led away from the nest site a distance equal to 1.68
km, the radius of a circle equal to an average adult home
range (Willey 1998). We assigned the dispersal day to the
Julian date halfway between the last known day in the
natal area and the first day confirmed outside of the natal
area. We assumed that dispersal movements ended when
a juvenile settled on a home range and found a mate.

Total distance moved was defined as the sum of dis-
tances moved between successive relocations from the
time a juvenile began dispersal to the last known loca-
tion. Final distance was defined as the straight-line dis-
tance between the nest location and last known location.
Movement rate (m/d) was defined as the distance be-
tween successive locations divided by the time interval.
We determined azimuth from the nest site to final loca-
tion for each individual and then estimated the dispersal
direction and the angular dispersion around the mean
direction (Batschelet 1981) for all individuals and each
cohort (i.e., each year’s group of juveniles) . Rayleigh’s
test (Zar 1974) was used to determine if movements were
random with respect to direction for individuals, among
and within cohorts.

We conducted field experiments to examine hypothe-
ses about proximate factors that influence the onset of
natal dispersal. Our first hypothesis was that juveniles dis-
persed from the parental home range to search for food,
an “Economic Hypothesis.” The null hypothesis was that
there would be no difference in the timing of dispersal
between juvenile Spotted Owls that were given supple-
mental food and juvenile Spotted Owls that were not sup-
plemented. An alternative hypothesis was that owls with
food supplements would show a later dispersal date than
owls not fed that leave to search for food outside the
natal area (H^j: > Ug). If dispersal onset was controlled

by body condition rather than food availability, we pos-
tulated that the additional food would cause the treat-
ment owls to leave sooner than controls (H^^: < Vg) .

An alternate mechanism was that juveniles were driven
from the parental home range by adults (Howard 1960).
Therefore, a second null hypothesis was that there is no
difference in adult aggressive behavior in the vicinity of
juveniles during the week preceding and the week fol-
lowing dispersal. We assumed that if adults drove young
from the area, then the number of aggressive calls by
adults would be greater in pre- versus post-dispersal ob-
servation periods (H^: Vi > U 2 ).

For the experiments, we located 10 juvenile Spotted
Owls in August 1994 and attached radiotransmitters to
their tail feathers. We were not able to determine the sex
of the juvenile owls in the field; however, the age (Fors-
man 1983, Moen et al. 1991) and initial condition (ster-
nal palpation) of the 10 juveniles were estimated at the
time of capture and determined to be quite similar
among the owlets. Therefore, five owls were randomly
selected for the treatment group and five for the control
group. The random allocation of owls to experimental
groups resulted in two sibling pairs selected for the fed
group. The treatment diet was two Mongolian gerbils per
d {Meriones unguicuculatus, 350—500 g/d) fed to each

treatment owl starting on 20 August at the start of the
predispersal period. Juveniles were fed two gerbils each
day (one each morning and evening) . Juveniles were re-
located each day and fed at the location where they were
found by placing the gerbils on the ground within 5 m
of the juveniles. The owls were fed each day until dis-
persal began and we conducted direct observation to
confirm each feeding. Members of the control group had
no food supplement. We used a two-sample Kest to com-
pare the mean dispersal dates between treatment and
control groups (Wilkinson 1990). Because the potential
correlation in dispersal timing by siblings could bias our
results, we repeated our analysis using only independent
owls in the control group (i.e., we selected the juvenile
that left first in two sibling pairs). Significance for the
tests was set at a = 0.10 (Steidl et al. 1997).

In testing Howard’s (1960) “social hypothesis,” we
considered calls an index to adult agonistic behavior.
Based on our interpretation of vocalization studies by
Forsman et al. (1984) and Ganey (1990), we classified
following “agitation” calls as agonistic; the bark call,
screeches and sharp whistles and the agitated series call
These calls are only produced by subadults and adults,
and not by juveniles. For the field test, the treatment
group remained the same as the previous experiment.
We selected this group because they were relatively more
accessible in the rugged terrain; however, selection of
this group may have biased our results because we did
not know the influence of supplemental feeding on adult
calls (i.e., we did not monitor calls of adults where juve-
niles were not fed gerbils) .

We assumed that if adults chased juveniles from their
natal areas we would observe aggressive vocalizations dur-
ing crepuscular and nocturnal time periods when vocal-
izations appeared to be most common (Ganey 1990).
Therefore, from mid-August-mid-September, when juve-
niles typically began dispersal, observers were positioned
within 100 m of the juveniles and recorded calls of adults
during 2-hr observation periods spaced throughout the
night beginning 30 min before sunset and ending 30 min
after sunrise. We sampled evenly across the night in both
pre- and post-dispersal periods to account for nighttime
variation in call rates (Ganey 1990). We contrasted the
number of agonistic calls per hr by adults recorded each
night during 1 wk pre-dispersal-1 wk post-dispersal. We
tested the null hypothesis using a one-tailed paired Mest.
Significance for this test was set at a = 0.10.

Results

Thirty-one juveniles were captured and radio-
tracked during 1992-95 in southern Utah (Table
1). Juveniles remained relatively close to their
hatching sites until the onset of dispersal (range of
locations from nest site = 67-520 m). Five owlets
died prior to dispersal and were removed from the
sample, leaving 26 that successfully dispersed from
their natal areas. Mortality after initial dispersal was
high for the 26 juveniles and only three (11.5%)
were alive after 1 yr. The three owls that survived
their first year appeared to settle in specific loca-

4

Willey and van Riper

VoL. 34, No. 1

Table 1. Dispersal date, distance (km) and fate of radio-tagged juvenile Mexican Spotted Owls. Owls with no dis-
persal date, or distance, died in the vicinity of the nest.

Owl

Code

Dispersal

Date

Last Live
Location

Fate‘S

Final

Distance’’

Total

Distance‘S

Ech474

1 Sep 92

15 Apr 93

Predation

21.2

189.9

Twi873

5 Oct 92

5 Feb 93

Predation

10.6

39.7

Cam784

19 Sep 92

21 Nov 93

Signal lost

26.5

187.8

Cam824

15 Sep 92

6 Feb 93

Exposure

43.6

97.1

Pea514

29 May 93

Exposure

Ele915

26 Sep 92

24 Aug 93

Signal lost

8.01

112.1

Ele574

5 Sep 92

29 Jun 93

Signal lost

1.68

136.6

Twi374

20 Sep 93

3 Mar 94

Exposure

15.1

40.9

Twi354

21 Nov 93

Exposure

Spr034

11 Sep 93

4 Mar 94

Exposure

21.2

27.9

Sprl23

15 Sep 93

Exposure

Ech414

11 Sep 93

3 Mar 94

Predation

28.8

58.0

Ech683

11 Sep 93

15 Dec 93

Exposure

24.3

25.3

Ham623

10 Sep 93

8 Nov 93

Predation

47.3

54.9

Ham262

10 Sep 93

6 Feb 94

Predation

90.0

106.1

Peal 75

10 Sep 93

6 Nov 93

Exposure

36.6

68.6

Bcd824

1 1 Sep 94

3 Nov 94

Radio molt

26.8

52.0

Bcd873

15 Sep 94

15 Nov 94

Exposure

24.8

32.2

Cam944

14 Sep 94

25 Feb 95

Predation

43.0

92.1

Fiv803

21 Sep 94

3 Nov 94

Predation

5.1

12.1

Usp983

14 Oct 94

Predation

Dar903

21 Sep 94

1 Oct 94

Signal lost

27.7

53.3

Bsp924

15 Oct 94

21 Oct 94

Radio molt

8.9

20.2

Sco783

14 Sep 94

5 Oct 94

Signal lost

17.1

31.3

Sco854

8 Sep 94

1 Oct 94

Signal lost

51.1

58.7

Fir054

17 Sep 94

5 Oct 94

Signal lost

31.5

40.3

Usp473

12 Sep 95

11 Oct 95

Signal lost

20.1

28.0

Usp564

5 Oct 95

15 Feb 96

Exposure

92.3

110.6

Fiv034

17 Sep 95

15 Nov 95

Predation

6.0

8.3

Bur354

10 Dec 95

Exposure

Tex262

24 Sep 95

27 Oct 95

Signal lost

30.8

85.0

Fate indicated the outcome of tracking: signal loss occurred when owl transmitters could not be relocated after extensive search,
radio molt refers to owls that dropped transmitters due to feather molt; exposure refers to owls that apparently starved to death,
predation refers to owls that appeared to have been killed by predators, although scavenging may have occurred.

Final distance was estimated as the straight-line distance from the nest to the last observed location of an owl.

Total distance was estimated as the sum of the distance between all successive locations observed during dispersal.

tions and their radio signals stopped transmitting
during the following summer. None of the juve-
niles that survived were found with mates, despite
extensive surveys, before their signals terminated.

The final distance dispersed by owlets averaged
29.2 ± 22.48 km (±SD) but varied substantially
(range = 1.68-92.3 km). Movement rate of indi-
viduals also varied, with periods of no movement
followed by periods of rapid movement in a few
days (Table 2). Four individuals averaged >7 km/
d during brief periods. The total distances traveled
by owls was positively correlated with number of

relocations/ owl (Spearman’s rank correlation, r =
0.723). Final distance traveled was not correlated
with the number of months the owls were tracked
(r = -0.093).

The mean dispersal direction (Batschelet 1981)
of all owls varied greatly and no significant direc-
tional pattern was determined within any cohort
(Rayleigh’s test, P — 0.687, 0.462, 0.104 and 0.371,
1992-95, respectively). Rayleigh’s test was signifi-
cant (P < 0.05) for five individual owls and sug-
gested that these owls exhibited directional, versus
random, movement during the portion of the dis-

March 2000

Dispersal Movements by Spotted Owls

5

Table 2. Moment rates (m/d) estimated for 20 radio-
tagged Mexican Spotted Owls with N > 4, where N =
number of relocations during the dispersal period.

Code

Year

N

Mean

SE

Min

(m/d)

Max

(m/d)

Ech474

1992

29

902

290

11

7476

Twi873

1992

27

301

88

12

2095

Cam784

1992

52

910

206

5

7530

Cam824

1992

15

709

180

2

2632

Ele915

1992

53

464

77

0

3641

Ele574

1992

46

619

171

0

7738

Twi374

1993

9

195

76

2

765

Spr034

1993

. 8

286

76

27

698

Ech414

1993

7

583

244

11

2057

Ech683

1993

4

718

320

22

1589

Ham623

1993

9

277

179

8

424

Ham262

1993

11

500

271

0

3192

Peal 75

1993

15

1382

504

0

6787

Bcd824

1994

6

3209

1463

185

9367

Cam944

1994

5

6692

2073

771

2232

Fiv803

1994

5

569

384

2

2273

Soc783

1994

11

631

117

258

1451

Sco854

1994

4

3698

860

338

6038

Tex262

1995

11

1189

279

57

2844

Usp564

1995

7

738

333

48

271

persal period that we monitored. The directional
movement of the remaining juveniles was appar-
ently random and most owls appeared to wander
the landscape, switching direction frequently.
Three juveniles that survived for at least 1 yr even-
tually reversed their direction of travel back toward
the natal area and two (Ele915 and Ele574, Table
1) were last observed within 8.01 and 1.68 km of
their nest sites when radio signals were lost.

We observed movements by five sibling pairs in
southern Utah. Dispersal date of siblings was pos-
itively correlated (r = 0.73) and suggested that sib-
lings begin dispersal at similar times. When we cal-
culated the azimuth between the nest site and final
location, siblings showed a tendency to disperse at
least 100° apart, suggesting they selected different
travel paths during dispersal. The angle between
sibling paths was even greater when the estimate
“mean direction of travel” (Batschelet 1981) was
used for this comparison (i.e., sibling paths were
180-209° apart).

We estimated to within 1 d when individual owls
dispersed from their natal area for 26 individuals
during the study (Table 1). During 1992-95, dis-
persal began during a 6-wk period from 1 Septem-

ber-15 October. Eighty-five percent of the juveniles
dispersed during September and 15% in October.
The mean day of dispersal did not differ among
the four cohorts (ANOVA, P = 0.063, F = 2.68, df
= 35), with 1992-95 cohorts dispersing during 1
September— 5 October, 10—20 September, 8 Sep-
tember-15 October and 12 September-5 October,
respectively. The overall mean day of dispersal was
17 September (±9.8 d) for all owls tracked.

During the field experiments, owls given supple-
mental food dispersed significantly earlier (x = 12
September ± 2.6 d) than the control owls (x = 30
September ± 12.3 d) as predicted by our second
alternative hypothesis (two-sample Mest, t= —2.75,
df = 8, P = 0.025). When we repeated the test with
independent observations (i.e., we eliminated 2
siblings), the difference in mean dispersal date re-
mained significant (two-sample Rest, t = —2.92, df
= 5, P = 0.033) . For the social hypothesis, we failed
to reject the null hypothesis of no difference
(paired Rest, t — —0.954, df = 4, P = 0.44) in the
number of agonistic calls by adults during the pre-
dispersal period (192 hr of observation during 21
nights) and the post-dispersal period (202 hr of
observation during 21 nights). On average, 0.66
(±1.63) agonistic calls/hr were recorded during
pre-dispersal and 0.97 (±1.74) calls/hr during
post-dispersal in the vicinity of three adult pairs.
Retrospective power analysis indicated that the
power of the test to reject the null hypothesis did
not exceed 0.20. Therefore, these results are in-
conclusive with respect to adult behavior toward
juveniles prior to dispersal.

Discussion

Factors that influence the onset of natal dispers-
al include endogenous mechanisms, where control
is governed by genetically-based systems, and ex-
ogenous mechanisms, in which timing is a re-
sponse to local factors, including food availability,
body condition and social interactions (Ferrer
1993). During our study of Mexican Spotted Owls,
we observed a narrow window for the onset of dis-
persal among four cohorts (similar to other stud-
ies, e.g., Gutierrez et al. 1985, 1995, Ganey et al.
1998). This provided strong evidence that onto-
genetic processes control the season when Spotted
Owls begin dispersal.

On the other hand, the results of our feeding
experiment, which showed a significantly earlier
departure period for food-supplemented owls, im-
plied that physical condition may supply the prox-

6

Willey and van Riper

VoL. 34, No. 1

imate signal for owls to leave the natal area. Belt-
hoff and Dufty (1998) reported that dominant
screech-owls dispersed earlier than subordinates
and, based on body size and mass evidence, they
assumed that dominant owlets were in better phys-
ical condition than the subordinates. Ellsworth and
Belthoff (1999) showed that dominant juvenile
screech-owls dispersed first (in five of seven
broods) and they concluded that dominance influ-
enced the timing of dispersal but not the distances
traveled to overwinter sites. Therefore, among
some owls, variation in timing of natal dispersal
may also be related to competitive ability.

Juvenile owls that leave the natal area in good
condition may avoid use of Critical muscle protein
and may be more likely to survive during dispersal.
Nilsson (1989) suggested that Juveniles that dis-
persed relatively early may also have a better
chance to settle in suitable habitat. Selection could
favor individuals that disperse as soon as they are
fit if vacant territories are limited; however, field
studies of Spotted Owls do not support a relation-
ship between dispersal date and age at first breed-
ing (Gutierrez et al. 1995). Furthermore, Spotted
Owls do not establish territories until at least 1-yr
old and few survive their first year.

Primarily due to low power and bias due to un-
known effects of supplemental feeding, the result
of our “Social Hypothesis” experiment was incon-
clusive with respect to whether juveniles were
forced out of natal areas by adults. In a study using
radiotelemetry to examine adult home ranges, Wil-
ley (1998) showed that during August-September
adults typically moved out of natal areas to other
portions of their home ranges and were far from
juveniles during the onset of dispersal. Northern
Spotted Owl adults also stop feeding juveniles in
mid- to late August and show no evidence of ag-
gression toward young before dispersal (E.D. Fors-
man pers. comm.). Thus, we believe that social in-
teractions between adults and juveniles were not
responsible for the onset of dispersal.

Results from our analysis of dispersal behavior
were similar to Spotted Owls reported from a va-
riety of landscapes. For example, Ganey et al.
(1998) reported a median dispersal distance of
16.9 km (range = 0.6-72.1 km) for juveniles
tracked in mixed-conifer forests versus our median
of 25.7 km (range = 1.68-92.3 km) in Utah. Ganey
et al. (1998) observed that Mexican Spotted Owls
initiated dispersal abruptly during September and
October and showed no obvious directional pat-

terns. We observed similar patterns for four co-
horts in the canyonlands of Utah.

In Arizona, New Mexico and Utah, Mexican
Spotted Owls were observed moving across open
low desert landscapes between islands of suitable
breeding habitat (Ganey et al 1998, Stacey pers.
comm., Willey 1998). Gutierrez et al. (1996) ob-
served movement of Mexican Spotted Owls be-
tween “Sky Island” mountain ranges in New Mex-
ico. Observation of long-distance movement by
dispersing juveniles provides evidence that widely
spaced enclaves have potential connectivity and
isolated populations may have genetic significance
to the owl’s conservation (Keitt et al. 1995, Gutier-
rez and Harrison 1996, Seamans et al. 1999).

In Utah and elsewhere across their range, pre-
dation and death caused by exposure (e.g., star-
vation, dehydration and disease) appeared to be
the primary causes of juvenile mortality during dis-
persal (Gutierrez et al. 1985, Ganey et al. 1998,
Miller et al. 1997, Willey 1998) . In Utah, only three
owls survived their first year (90% mortality) and
we found it interesting that two of the three settled
relatively close (1. 7-8.0 km) to their natal areas.
We also observed juveniles that died while moving
back toward natal areas in early spring. Therefore,
studies that do not follow juveniles to the place of
first reproduction (i.e., recruitment into the pop-
ulation) may bias their estimates of mean and me-
dian dispersal distances. This bias could be non-
trivial with regard to population modeling. In
addition, knowledge of each individual’s sex is im-
portant because dispersal may be sex-biased. In the
Pacific Northwest, female Northern Spotted Owls
have been observed traveling much further than
males (E.D. Forsman pers. comm.). Therefore, we
strongly urge that future studies of owl dispersal
collect blood samples to determine sex of each ju-
venile.

Acknowledgments

We thank the many dedicated biologists and field as-
sistants who supported this study. Special thanks to L.
Wells, Red-Tail Aviation, Moab, Utah, for his skill and
dedication during numerous tracking missions. We thank
E. Forsman, J. Belthoff, R. Gutierrez, J. Ganey and P. Sta-
cey for reviewing earlier drafts of this paper. We thank J.
Armetta and S. van Riper for support during the project.
Funding for this project was provided by the Utah Divi-
sion of Wildlife Resources and the National Park Service.

Literature Cited

Batschelet, E. 1981. Circular statistics in biology. Aca-
demic Press, New York, NY U.S.A.

March 2000

Dispersal Movements by Spotted Owls

7

Belthoff, J.R. AND A.M. Dufty, Jr. 1998. Corticosterone,
body condition and locomotor activity: a model for
dispersal in screech-owls. Anim. Behav. 55:405-415.

Brown, D.E. 1982. Biotic communities of the American
southwest-United States and Mexico. Des. Plants 1-4.

Ellsworth, E.A. and J.R. Belthoff. 1999. Effects of so-
cial status on the dispersal behavior of juvenile West-
ern Screech-Owls. Anim. Behav. 57:883-892.

Ferrer, M. 1993. Juvenile dispersal behavior and natal
philopatry of a long-lived raptor, the Spanish Imperial
Eagle Aquila adalberti. Ibis 135:132-138.

Forsman, E.D. 1983. Methods and materials for locating
and studying Spotted Owls. Gen. Tech. Rep. PNW-
162, For. Serv. Pac. Northwest Forest and Range Ex-
per. Sta., Portland, OR U.S.A.

, E.C. Meslow and H.M. Wight. 1984. Distribu-
tion and biology of the Spotted Owl in Oregon. Wildl.
Monogr. No. 87.

Ganey, J.L. 1990. Calling behavior of Spotted Owls in
northern Arizona. Cowdor 92:485-490.

, W.M. Block, J.K. Dwyer, B.E. Strohmeyer and

J.S. Jenness. 1998. Dispersal movements and survival
rates of juvenile Mexican Spotted Owls in Northern
Arizona. Wilson Bull. 110:206-217.

Gutierrez, R.J., A.B. Franklin, W.S. LaHaye, VJ. Mer-
etskyandJ.P. Ward. 1985. Juvenile Spotted Owl dis-
persal in Northwestern California: preliminary results.
Pages 39-49 in R.J. Gutierrez and A.B. Carey [Eds.],
Ecology and management of the Spotted Owl in the
Pacific Northwest. Gen. Tech. Rep. PNW-185, U.S.
For. Serv., Portland, OR U.S.A.

, A.B. Franklin and W.S. LaHaye. 1995. Spotted

Owl. The Birds of North America, No. 179. The Acad-
emy of Natural Sciences, Philadelphia, PA and the
American Ornithologists’ Union, Washington, DC
U.S.A.

AND S. Harrison. 1996. Applying metapopulation

theory to Spotted Owl management: a history and cri-
tique. Pages 167-185 in D.R. McCullough [Ed.], Me-
tapopulations and wildlife conservation. Island Press,
Washington DC U.S. A.

, M.E. Seamans and M.Z. Peery. 1996. Intermoun-
tain movement by Mexican Spotted Owls. Great Basin
Nat. 56:87-89.

Howard, W.E. 1960. Innate and environmental dispersal
of individual vertebrates. Amer. Midi. Nat. 63:152-161.

Keitt, T.H., A.B. Franklin and D.L. Urban. 1995. Land-
scape analysis and metapopulation structure. Pages 1-
16 in USDl 1995 Mexican Spotted Owl Recovery Plan,

Vol. II. USDI Fish and Wildl. Serv., Albuquerque, NM
U.S.A.

Kenward, R.E., V. Marcstrom and M. Karlbom. 1993.
Post-nestling behavior in goshawks, I. the causes of
dispersal. Anim. Behav. 46:365-370.

King, R.A. 1996. Post-fledging dispersal and behavioral
ecology of Burrowing Owls in southwestern Idaho
M.S. thesis, Boise State Univ., Boise, ID U.S.A.

Koenig, W.D., D. van Vuren and P.N. Hooge. 1996. De-
tectability, philopatry, and the distribution of dispersal
distances in vertebrates. Trends E col. &' Evol. 11:514-
517.

Miller, G.S., RJ. Small and E.C. Meslow. 1997. Habitat
selection by Spotted Owls during natal dispersal in
western Oregon./. Wildl. Manage. 61:140-150.

Moen, C.A., A.B. Franklin and R.J. Gutierrez. 1991. Age
determination of subadult Northern Spotted Owls in
Northwest California. Wildl. Soc. Bull. 19:489—493.

Nilsson, J. A. 1989. Causes and consequences of natal dis-
persal in the March Tit. J. Anim. Ecol. 58:619—636.

Reid, A.R., R.B. Horn and E.D. Forsman. 1996. A meth-
od for replacing tail-mounted radio transmitters on
birds. / Field Ornithol. 67:177-180.

Ritchison, G., J.R. Belthoff and EJ. Sparks. 1992. Dis-
persal restlessness: evidence for innate dispersal byju-
venile Eastern Screech-owls? Anim. Behav. 43:57-65.

Seamans, M.E., R.J. Gutierrez, C.A. May and M. Zacha-
riah Peery. 1999. Demography of two Mexican Spot-
ted Owl populations. Conserv. Biol. 13:744—754.

STEIDL, R.J.,J.P. Hayes and E. Schauber. 1997. Statistical
power analysis in wildlife research, f. Wildl. Manage.
61:270-279.

Thornbury, W. 1965. Regional geomorphology of the
United States. John Wiley and Sons, New York, NY
U.S.A.

Wilkinson, L. 1990. SYSTAT: the system for Statistics.
SYSTAT, Inc., Evanston, IL U.S.A.

Willey, D.W. 1995. Mexican Spotted Owls in canyonlands
of the Colorado Plateau. Pages 330-331 in E.T. La-
Roe, G.S. Farris, C.E. Puckett, P.D. Doran and M.J.
Mac [Eds.], Our living resources: a report to the na-
tion on the distribution, abundance, and health of
U.S. plants, animals, and ecosystems. USDI, National
Biological Service, Washington, DC U.S.A.

. 1998. Movements and habitat utilization by Mex-
ican Spotted Owls in canyonlands of southern Utah.
Ph.D. dissertation, Northern Arizona Univ., Flagstaff,
AZ U.S.A.

Zar, J.H. 1974. Biostatistical analysis. Prentice-Hall, Inc.,
Englewood Cliffs, NJ U.S.A.

Received 30 July 1999; accepted 20 November 1999

J Raptor Res. 34(l):8-l7

© 2000 The Raptor Research Foundation, Inc.

DISTRIBUTION, POPULATION SIZE AND HABITAT USE OF THE
REUNION MARSH HARRIER, CIRCUS M. MAILLABDI

Vincent Bretagnolle

CEBCrCNRS, 79360 Beauvoir sur Niort, France
Thomas Ghestemme

SEOR, Museum de Saint-Denis, 1, rue Poivre, 97400 Saint-Denis, La Reunion, France

Jean-Marc Thiollay

Laboratoire d’Ecologie, ENS, 46, rue d’Ulm, 75005 Paris, France

Carole Attie

79360 Vaubalier, France

Abstract. — On Reunion Island, Indian Ocean, the only surviving resident raptor is the endemic Re-
union Marsh Harrier, Circus m. maillardi. It has been considered conspecific with the Madagascar Marsh
Harrier ( C. m. macrosceles) but a reanalysis of the classification of both is needed. A survey was carried
out on the island in 1997-98 and the breeding population was found to be no larger than 100 pairs,
mostly concentrated in forested areas at mid-elevations (300-700 m) and down to sea level, but rarely
above 1200 m. The distribution of pairs was clumped with as many as 8 pairs within a 10 km^ patch,
and mostly concentrated in low degraded native woodlands on steep slopes. Foraging habitats were
more diversified and were widely distributed from coastal wetlands to cultivated fields, forest and upper
montane vegetation. Breeding was apparently not synchronized among pairs but egg laying occurred
primarily from December-March. The Reunion Marsh Harrier must be considered as Threatened, al-
though its distribution has apparently not declined between 1976-98.

Key Words: Reunion Marsh Harrier, Circus m. maillardi; Indian Ocean-, population size, habitat use.

Distribucion, tamano de poblacion y uso de habitat de Circus m. maillardi

Resumen. — En las Islas Reunion, Oceano Indico, la unica ave rapaz residente es Circus m. maillardi, la
cual ha sido considerada como C. m. macrocesles, al respecto, una revision de la clasificacion de estas dos
es requerida. La investigacion realizada en la isla entre 1997—98 encontro que la poblacion reproductiva
no era mayor de 100 pares, concentrados en areas boscosas en elevaciones medias (300-700 m) des-
cendiendo hasta el nivel del mar, raras veces sobre los 1200 m. La distribucion de las parejas fue
aglomerada, se encontraron hasta 8 parejas dentro de un parche de 10 km^, casi siempre concentradas
en areas de bosque native degradadas en pendientes inclinadas. Los habitos de forr^eo fueron mas
diversos, ampliamente distribuidos desde los humedales costeros hasta areas cultivadas, bosques y ve-
getacion montana de altura. La reproduccion no fue sincronizada entre parejas, esta ocurrio desde
Diciembre a Marzo. C. m. macrosceles debe ser considerado como amenazado, aunque su distribucion
aparentemente no ha declinado entre 1976-98.

[Traduccion de Cesar Marquez]

Within the group of large harriers (i.e., the
Marsh Harrier aeruginosus complex, sensu Sibley
and Monroe 1990), detailed information on their
ecology is only available for aeruginosus in Europe
(revietv in Clarke 1995), in Australasia

(review in Marchant and Higgins 1993) and to a
lesser extent for ranivorus in Africa (see Simmons

1997). The biology of the Reunion Harrier is very
poorly known, except for its morphology and molt
(Nieboer 1973) and some aspects of its breeding
biology (Clouet 1978).

Reunion Island (2515 km^, maximum 50 X 70
km) is larger, more forested and has a lower hu-
man population density than the other two Mas-

8

March 2000

Ecology of the Reunion Marsh Harrier

9

carene Islands (Mauritius and Rodrigues) in the
Western Indian Ocean. Both Mauritius and Re-
union Islands have only one raptor still surviving,
a kestrel and a harrier, respectively (Sinclair and
Langrand 1998). The Madagascar Marsh Harrier
(Circus maillardi), hereafter called the Reunion
Harrier, is now the only raptor breeding on Re-
union. There is some controversy with regard to its
taxonomic status and it is currently divided into
two well-marked subspecies, C. m. maillardi (Re-
union) and C. m. macrosceles (Madagascar and Com-
oros) (Howard and Moore 1980), but the Reunion
population may well deserve full specific status.
The population size of the Reunion Harrier on Re-
union was previously estimated at 200-300 pairs
(Clouet 1978, Barre et al. 1996), but no standard-
ized surveys were used to obtain these estimates.
In earlier times, the species was considered to be
abundant (Dubois 1672, in Barre et al. 1996), al-
though no measure of abundance or relative abun-
dance was provided.

We undertook a survey to assess the current sta-
tus, population size and trends of the Reunion
Harrier and to identify its conservation needs. This
paper provides the first comprehensive account of
the harrier on Reunion. Based on censuses carried
out in 1997-98, we provide a provisional distribu-
tion map and a current population estimate for the
whole island. We also summarize available data on
the habitat use and breeding period, based on un-
published information gathered between 1987-98.

Study Area

Reunion Island (21°15'S, 55°30'E) is a volcanic island,
2-3 million years old. It is 165 km from Mauritius, the
nearest island, and 700 km east of Madagascar (Fig. 1).
Most of the island is mountainous and steep. Three large
and deep cirques (caldeiras), Mafate, Cilaos and Salazie,
surround the highest peak (Piton des Neiges, 3069 m).
An active volcano (Piton de la Fournaise, 2631 m) in the
southeast is the second highest summit. Coastal lowlands
are now totally cultivated or urbanized. Below 1300 m,
degraded native forest is restricted to the steepest slopes.
The mean annual rainfall is highest on the eastern (wind-
ward) side of the island, increasing from 3 m on the coast
to 8-12 m between 1300-1900 m elevation and decreas-
ing above 2000 m elevation (Barcelo 1996). Precipitation
declines markedly on the eastern (leeward) side, down
to 1 m on the southwestern coast. The mean annual tem-
perature decreases from 24-26°C in the lowlands to 12°C
around 2000 m elevation where frost is frequent in winter
(June-August). Major cyclones occur every 5-10 years
during the hot rainy season (December-April), but trop-
ical storms with heavy rainfalls are of almost yearly oc-
currence.

The island was discovered in the 16th century and was

heavily deforested with human population growth. Mas-
sive extinctions, including 22 bird species, occurred from
the early stages of colonization due to hunting pressure,
introduced mammals (rats, cats, goats and pigs) and hab-
itat destruction (Barre et al. 1996). Intensive agriculture,
sprawling urbanization and construction of a dense road
network are still affecting the remaining natural habitats
The native flora comprises 750 species, but 1100 addi-
tional taxa have been introduced, some of which are in-
vading natural forest remnants (MacDonald et al. 1991).
At most 55 000 ha of forest remain (Cadet 1980, Dou-
menge and Renard 1989, Dupont 1990) on 22% of the
island area and consist of almost none of the former west-
ern dry woodlands, <1% of the original lowland mixed
forest, 60% of the montane forest and 80% of the high
altitude vegetation.

Five major natural habitats have been recognized (Ri-
vals 1952, Cadet 1980, Barre et al. 1996; Fig. 1): (1) dry
savanna woodland and semi-sclerophylous forest on the
coastal lowlands has almost entirely been replaced by cul-
tivation, urbanization and introduced vegetation; (2) hu-
mid lowland mixed evergreen forest (“Bois de Couleur des
Bas"), originally covered the eastern lowlands up to 800-
900 m and the western side from 750-1100 m, is now
largely degraded with remnants having a dense understo-
ry, dominated by an open canopy 6—15 m high; (3) up-
land wet mixed evergreen forest (“Bois de Couleur des
Hauls"), from 800-1900 m in the east and from 1100-
2000 m in the west, is richer in epiphytes and tree ferns;

(4) montane forest, between 1600-2000 m, is dominated
by large Acacia heterophylla (“Tamarins de Hauls") that are
taller (15-20 m) and larger (<1.5 m dbh) than trees m
mixed evergreen stands; the understory is locally rich in
native bamboos (Nastus borbonicus), giant heath (Philippia
monlana) and stands of screwpines (Pandanus montanus ) ,

(5) high heath and shrubs above the tree line are 1-2 m
tall and can be very dense. We also recognized the fol-
lowing five heavily modified habitats that are dominated
by exotic vegetation; (6) monospecific exotic tree plan-
tations of Cryplomeria spp., Pinus spp. and Eucalyplusspp.,
(7) dry derived savannas and shrubby areas mostly in the
western lowlands on abandoned fields; (8) cultivated ar-
eas, often large fields of sugarcane but sometimes with
more diversified crops with tree rows, woodlots and or-
chards; (9) urban and suburban areas, including associ-
ated gardens, roads and tourist resorts; and (10) wetlands
such as coastal ponds and marshes or lakes in the moun-
tains that are highly restricted and modified by exotic
vegetation and deforested surroundings.

Methods

We conducted a comprehensive survey aimed at locat-
ing every territorial pair of Reunion Marsh Harrier. Be-
cause of the landscape heterogeneity and a perceived un-
even distribution of the species, we avoided making a
population estimate that relied on extrapolations from a
limited number of sample areas. Instead, we surveyed
most areas of natural and semi-natural vegetation and
searched, as far as possible, most potentially suitable
breeding habitats for harriers. This was based on our pre-
vious experiences on Reunion and the dense network of
roads, forest tracks, mountain trails and viewpoints that
allowed an adequate coverage of otherwise seemingly in-

10

Bretagnolle et al.

VoL. 34, No. 1

accessible areas. Observations on harriers were made
during surveys from a network of 111 lookouts sampled
under good weather conditions between 0700-1800 H
from 23 October 1997-15 May 1998, and reported on a
map using a grid of 2 X 2-km squares with each square
searched for >1 hr and, if possible, >2 hr. Additionally,
two extensive surveys (14 December and 29 March) were
conducted by as many as 14 observers who simultaneous-
ly recorded the number, sex, movements and behavior
of all individual harriers seen within contiguous areas
from separate lookouts over a full day. The first survey
method covered the most suitable areas of habitat. From
each site, an area of about 3 km^ was surveyed using 10
X 42 binoculars. Because of the low-flying behavior of
harriers, the movements of all individuals were drawn on
1:25 000 topographical maps, using prominent land-
marks. Surveys were conducted most often with two peo-
ple. Therefore, we obtained an estimate of the minimum
number of pairs present within a well-defined area
around each lookout. In order to avoid possible double
counts of the same pair on adjacent squares, we only
counted those pairs whose core territory fell within the
survey square.

We also used a survey method based on a grid of 577
(4 km^) squares that covered the entire island. From 8
December 1997—22 January 1998, we sampled as many
squares as possible. Observations were made from the
most suitable viewpoint and crossed by foot or car. We
considered a square to be completely surveyed if at least
60 min of continuous observation time were spent at a
central point in the square, or if the total observation
time for the square was ^1 hr with periods of >20 min
of continuous observation in the same area. Many of
these squares were sampled for >2 hr. Also, during the
same breeding season (September 1997-March 1998),
we surveyed an additional 105 squares, some of them not
previously sampled. Using such criteria, 331 (2X2 km)
squares were surveyed and the minimum number of
pairs in each of them was assessed. Two pairs were con-
sidered to be different if they were seen at least once

simultaneously. Many other squares were crossed oppor-
tunistically but they were not adequately or fully surveyed
because of a lack of time, inappropriate weather condi-
tions or because they were unsuitable for breeding har-
riers (totally urbanized, cultivated or above treeline) .

Because we focused on the identification and locali-
zation of breeding or potentially-breeding pairs, we ex-
cluded observations of individual foraging birds. The fol-
lowing criteria were thus used in all census methods to
classify the degree of breeding evidence: a possible
breeding pair was a pair of adults flying together that
showed no particular breeding behavior or a single adult
that performed nuptial displays; a probable breeding
pair was a pair that showed territorial defense behavior,
usually an adult chasing an individual of the same sex,
or even talon grasping, two adults of opposite sexes that
displayed together, or one in the presence of the other
or adult females that gave solicitation to passing adult
males; a certain or confirmed breeding pair was one with
one or more young following an adult carrying a prey
with persistent begging calls, an adult bringing prey or
nest material to a potential nest site, prey transfer be-
tween males and females, the occupied nest was found
or an empty nest was found with a pair nearby.

All observations (irrespective of sampling methodolo-
gy) were plotted on a digitized map of Reunion Island
using a Geographical Information System (ARCVIEW 3.b
software; Environmental Systems Research Institute Inc.
1996). A grid was superimposed on the island GIS map
dividing it into 577 (2 X 2 km) squares whose limits
crossed habitat types randomly. Among them, 85 squares
overlapped the coastline and included a variable propor-
tion of sea area. When a nest site was not precisely locat-
ed and the record overlapped the limits of two squares,
it was assigned to a single and most appropriate square.

Data sets obtained by the two survey methods were
treated separately, but used together for estimating the
distribution of population on the island. Observations
from lookouts were first mapped on the island grid. We
then lumped all data to produce two kinds of maps. The
first summarized searching/sampling effort by giving
each square a sampling status (i.e., not sampled, sampled
by the first method, sampled by the second method or
sampled by both methods) . The second map was a gen-
eral distribution map of the species. Using these maps,
we determined the breeding status of harriers in each
square using the highest recorded status by any of the
methods.

For abundance estimates, we reported data from the
two methods in each square. When harriers were ob-
served in a given square, we summed the number of pos-
sible, probable and confirmed territorial pairs. When
there were two different pairs in the same or adjacent
squares, they had to have been seen simultaneously to be
tallied. This meant that, if a square was surveyed using
two different methods, we used the maximum number
of pairs given by one method.

Results

Surveys and Coverage of Reunion Island. A total
of 384 squares (66.6%) were adequately surveyed.
Because 43 additional squares were unsuitable for

March 2000

Ecology of the Reunion Marsh Harrier

11

V

/

'•■ ••>’.- V'-.-V ^^>^'J/:V‘' 'fe' •••'/;■■;•• ;:• ,;-^

(ir> «MKKiland<(
loMUnd wri fnrrU
upland wrt r<»rr«l
Acacia hetrrophyUa forrsi
r~ Tn hiKh ahiludr hralHrr

'i- ■ Uf- ■! ■ \

\

\

N

t

0 <

'-a

'«:;^ ■ f,,,; ^

//

\

le km

B

SI lU MS

[ ] dr> MiKKlIamiv

[:I:^X; lowland wrI fortwl

['2^71 upland wcl forrvl

t-b? ] Acacia hetrrophyUa forrvl

\

p \ ' I high altiludr hraihrr
I I flrldv sreond growih

''\si HI son

\

I

\Mr \SSI

\

V

KtIM

aai 1 1

5 16 Km ■■'^

" ’ SimMH^

Figure 1. Continued. (B) Reconstructed distribution of natural vegetation before human colonization. (C) Current
distribution of main vegetation types.

harriers, our combined survey covered 75% of Re-
union Island. Surveys based on 2-hr observations
from lookouts provided data on 108 squares {N =
111 lookouts) , while the method based on surveys

of squares resulted in 111 squares with >1 hr of
observation and an additional 196 squares with a
shorter census (Fig. 2). The least surveyed areas
were two of the three cirques which were known

12

Bretagnolle et al.

VoL. 34, No. 1

? not sanpled

• <1 hour observation

• >1 hour observation

• lookout and <1 hour observation

• lookout (2 hours)

not sampled, but inappropriate for harriers

Figure 2. Harrier survey sampling sites.

to contain very few harriers and a heavily forested
area in the east of the island where harriers were
probably scarce due to the vegetation structure.

Distribution Pattern of Reunion Harriers. Over-
all, harriers were distributed throughout the is-
land, with no marked preference for any region
(Fig. 3). Pairs were irregularly spaced with large
areas where they were absent as breeders. There
were six areas where as many as 7-8 pairs were
aggregated (e.g., within a 8-km segment of a valley
with a density of 5-7 pairs over 16 km^ [Fig. 3, see
Clouet 1978 for density estimates]). Surprisingly,
harriers were rare within the three large cirques
with no more than 2-3 breeding pairs in each al-
though they were abundant at the entrances of the
cirques. The bottom of these cirques were rather

flat and heavily populated but their surrounding
steep slopes could provide suitable breeding sites
for harriers. Overall, most pairs were concentrated
in the lower valleys, mostly along gullies, canyons
and other steep areas, a tendency which is also ap-
parent in some breeding seabirds.

Little or no data were available for 150 squares
but many of them were unsuitable for harriers.
Breeding harriers were absent from another 343
squares (60% of the island area). Possible breeders
were found in 20 squares, probable breeders in 46
squares and confirmed breeders in 18 squares (Fig.
4) . The rather low numbers in the latter category
were due to our late searching effort during the
fledging and postfledging periods (>3 mo, Clouet
1978). During this period, we observed many pairs

March 2000

Ecology of the Reunion Marsh Harrier

13

Figure 3.
square.

♦ harrier absent
0 confirmed pair

o probable pair
possible pair
? not sampled

Harrier breeding distribution on Reunion Island according to the highest breeding status recorded per

feeding flying young and possibly outside their
breeding territories. We treated these as probable
breeders.

Population Size. Lumping possible, probable
and certain pairs and using data from the two
methods, 62 squares had a single pair, 19 squares
had two pairs and 3 squares had three pairs for a
total of 109 pairs. They were divided into 21 con-
firmed breeding pairs, 28 probable pairs, 33 family
groups not belonging to any of the two former cat-
egories and 27 possible pairs (in most cases single
males displaying). Nearly half of the 150 squares
not adequately surveyed were in cirques or in
dense natural forests and were unlikely to be used
by harriers. Using the mean density estimate in
squares actually sampled, we estimated that <20
pairs would have been found in the 80 unsurveyed
squares. Therefore, the total estimated population
could be as high as 125-130 pairs (21 confirmed,
28 probable, 33 additional families and up to 50
unconfirmed pairs) . A more conservative estimate

would place the current breeding population of
Reunion Harriers at about 100 pairs.

Breeding Habitat. Confirmed breeding pairs of
harriers were distributed from 0-1200 m elevation,
and up to 1800 m if probable breeding pairs and
possible dispersing families were added. Overall,
about 75% of pairs were below 800 m elevation
and 25% were between 800-1600 m (Table 1).
Most were concentrated between 300-700 m (Fig.
5), with a median altitude of 500 m {N = 49, x =
650 ± 396 m ± SD). We could not test for a sta-
tistical preference for mid- to low-elevation breed-
ing ranges because of difficulties associated with
assessing the availability of areas at each elevation.
However, harriers clearly avoided high altitudes
(>1200 m) for breeding, but there was no evi-
dence that they selected a particular level below
this limit.

Because most habitat types had a limited altitu-
dinal distribution, there was a strong correlation
between habitat choice and elevation range (Table

14

Bretagnolle et al.

VoL. 34, No. 1

« harrier absent

• 1 pair

• 2 pairs

0 3 pairs

? not sampled

Figure 4. Harrier abundance map on Reunion Island during the 1997—98 breeding season. Number of pairs re
corded per 2 X 2 km squares.

Table 1 . Distribution of breeding pairs of the Reunion Harrier among habitat types.

Habitat

Habitat

Reference

Elevation

Range

Number of Breeding Pairs
Confirmed Probable

Dry savanna woodland

1

<500

0

2

Lowland evergreen forest^

2

200-800

10

15

Upland wet evergreen foresU

3

800-1600

4

8

Montane Acacia forest

4

1600-2000

0

0

High altitude heathland

5

2000-3000

0

0

Exotic tree plantations

6

200-1400

0

0

Derived shrubland

7

100-600

4

2

Cultivated areas and pastures

8

0-1500

1

1

Urban and suburban areas

9

0-1300

0

0

Wetlands

10

0-600

2

0

Total

21

28

* Including secondary and degraded forests.

March 2000

Ecology of the Reunion Marsh Harrier

15

IiIiIIImI

Ami. 4*

Figure 5. Altitudinal range of certain and probable
pairs of Reunion Harriers during the 1997-98 breeding
season.

1). Except for lowland swamp and shrubby areas,
all known harrier breeding sites were in wooded
areas in low, open, degraded native woodlands and
were never in taller, dense forest. Nests that we
found were never under a tree canopy but in open
patches within forests {N = 6), in savanna with
shrubs {N— 12) and in vegetation on cliffs or steep
grassy slopes (N = 10, see Clouet 1978). No nests
have ever been found or suspected to be in the
widespread sugarcane fields or under primary
dense forest cover.

Foraging Habitat. We pooled all independent
observations of foraging harriers (predominantly
males) obtained during census sessions (N = 447) .
Foraging habitats were much less restricted than
nesting habitats and included almost any habitat
type, except urban and suburban areas which were
apparently used in historical times (Barre et al.

Table 2. Foraging habitat of the Reunion Harrier.

1996). A minor proportion (15%) of foraging har-
riers were recorded over open grasslands, derived
savannas, roadsides, airfields and a golf course (Ta-
ble 2). The majority of them used various forested
areas including low native forests, open tree plan-
tations, shrubby clearings, heathland above tree-
line and woody vegetation on steep slopes (65%).
The third important foraging habitat type was sug-
arcane fields and pastures which accounted for
20% of the records. Few harriers foraged over
dense unbroken native forests, gardens, dense ma-
ture tree plantations, coastline, stony estuaries of
rivers, ponds and industrial areas (<2%). A set of
randomly encountered birds, including nonforag-
ing individuals, gave an even broader distribution.
In this sample, almost any habitat was occasionally
flown over, except towns and mountain slopes
above 2600 m, yet low or open woodlands were still
favored.

Discussion

According to our observations and those of
Clouet (1978), males begin to perform display
flights in August-September, nest building occurs
from October-November onward, egg laying from
January up to April and fledglings often follow psu"-
ents up to October. Our surveys (October-May)
covered only part of the breeding season but, be-
cause pairs of Reunion Harriers are sedentary on
their territories all year round, this survey period
may have only affected the proportion of con-
firmed compared to possible breeding pairs and
not their distribution or numbers. The time we
spent in survey squares was variable from l->4 hr.
Observation time (1 vs. >2 hr) significantly affect-

Habitat Category

Time Spent
(Hours)

Number of
Contacts

% OF

Contacts

Records
per Hour

Dry savanna or shrubby woodlands

32

24

5.4

0.75

Wet mixed evergreen forest

31

38

8.5

1.23

Secondary native forest

91

88

19.7

0.97

Montane Acacia forest

11

0

0

0.00

High altitude heathland

>10

2

0.4

0.20

Exotic tree plantations^

61

199

44.5

3.26

Cultivated areas (including pastures)

57

90

20.1

1.58

Urban and suburban areas

>10

0

0

0.00

Wetlands

11

6

1.3

0.55

Total

>314

447

100

® Including heavily degraded forests.

16

Bretagnolle et al.

VoL. 34, No. 1

ed the probability of detecting the presence of har-
riers in a given square (x^ = 81.7, df = \, P <
0.001), but there was no relationship between ob-
servation time and number of pairs detected on
squares (x^ = 0.5, df = 1, P> 0.05). This suggested
that observation time had little effect on our sur-
veys. Therefore, we are confident that potential
bias due to our methods is limited, and that the
data provide a realistic estimate of both distribu-
tion and population size.

Overall, harriers were distributed throughout
the island, with no marked concentration in any
part of it. The current distribution probably re-
flected more the influence of human disturbance
than true habitat preferences. Their preferred hab-
itat occurred on rather steep and forested slopes,
away from human settlements. Such characteristics
are most often found along rivers, especially at
cirque entrances. This suggested a currently signif-
icant level of human persecution, despite the har-
rier’s fully protected status.

The total population in 1998 was estimated to
consist of 400-600 individuals with <100 breeding
pairs, although this was a conservative estimate.
Uncertainties about the true population size re-
sulted from the large proportion of birds in adult
plumage which did not exhibit territorial or sexual
behavior. A more accurate estimate of the breeding
population size might be made if we had a better
understanding of the social and breeding behavior
of this species.

None of the previous estimates of the Reunion
Harrier population were based on systematic
searches. Therefore, estimates of the population
ranged from 130-300 breeding pairs (Clouet 1978,
Cheke 1987, Barre 1988, Barre et al. 1996). Our
population estimate cannot be compared to pre-
vious ones because it was derived from the first
complete and reliable survey of the island. Never-
theless, based on the map given by Cheke (1987)
and an unpublished map (M. Clouet pers. comm.),
neither the overall distribution of the harrier, nor
the areas with highest densities have shown any de-
tectable change between 1978-98. Based on 17th-
century accounts (Dubois, in Barre et al. 1996) , the
Reunion Harrier was probably much more abun-
dant than today. Habitat losses and faunal impov-
erishment following human colonization most like-
ly caused its early decline although shooting
probably played a significant role. There is no con-
clusive evidence showing that the harrier popula-
tion has declined significantly during the last 25 yr.

despite higher population estimates at a time when
its recently protected status was enforced very little.
Human persecution still takes place but may be
currently decreasing, and formerly cultivated or
grazed areas are now abandoned, thus increasing
areas of suitable habitat. Urbanization and road
construction are still increasing but they occur
mostly in lowlands and areas that have long been
deforested and already out of the harrier’s range.

Aggregation of nesters in a few areas may be due
to a lack of territorial behavior of harriers away
from the immediate vicinity of their nests. It exhib-
its a broad range of foraging habitats, a variety of
hunting techniques and an eclectic prey choice,
including birds, introduced mammals and some
reptiles, amphibians and grasshoppers (Clouet
1978, Cheke 1987, pers. obs.). This wide niche
breadth and adaptability are typical of most island
birds when compared to their continental counter-
parts, including tropical raptors (Thiollay 1993,
1997). Nevertheless, prey abundance and accessi-
bility are likely to be major determinants of hunt-
ing habitat selection.

Until the 16th century, the island was almost
completely covered with forest. The relatively short
rounded wings of the Reunion Harrier probably
are an adaptation to hunt in rather dense vegeta-
tion and its relatively long middle toe is typical of
a bird specialist, probably a necessity when terres-
trial mammals were absent and medium-sized low-
er vertebrates were uncommon (Nieboer 1973).
Today, large birds have disappeared but intro-
duced rodents {Rattus, Mus), insectivores {Tenrec,
Suncus ) , reptiles ( Calotes, Chameleo, Phelsuma ) , toads
(Bufo) and frogs (Ptychadena) are abundant (Probst
1997). Most of them are probably more difficult to
find and catch in forests than were the once nu-
merous pigeons and parrots (Barre et al. 1996).
This may explain why this harrier tends to avoid
closed canopy forest and to favor lower vegetation
and more open woodlands in spite of its apparent
morphological adaptation to forest. Nevertheless,
it remains much more of a forest bird than any of
the 12 other species of harriers in the world (del
Hoyo et al. 1994).

The population of Reunion Harriers is precari-
ously small by genetic, demographic and conser-
vation standards. It is currently the smallest popu-
lation of any native and nonmarine bird species on
Reunion. It is also one of the rarest raptor species
in the world and its population is now even lower
than that of the fast-recovering population of the

March 2000

Ecology of the Reunion Marsh Harrier

17

Mauritius Kestrel {Falco punctatus) , once the most
endangered raptor in the world (Collar et al. 1994,
del Hoyo et al. 1994). There is good evidence to
upgrade its taxonomic status to the full specific lev-
el, distinct from the Madagascar Harrier. As a tax-
on of its own, and according to lUCN criteria (Col-
lar et al. 1994), it may well deserve Endangered
status because of its small population size, small
range and current factors threatening its long-term
survival. Although the population appears to be
currently stable, this does not mean that it is at full
carrying capacity and/or that carrying capacity is
not declining through habitat loss or disturbance
and degradation. Human population growth and
economic development are very high on Reunion
and the species is threatened by human persecu-
tion both from shooting and loss of breeding and
foraging habitats, by increased urbanization and
road construction and frequent cyclones, heavy
rains and wildfires during the breeding season.

Acknowledgments

We are grateful to M. Clouet for providing many un-
published notes and maps. We also thank J.M. Probstwho
contributed additional information, E. Fortier who
shared fieldwork, M. Lecorre and members of the Socie-
te d’Etudes Ornithologiques de la Reunion for organiz-
ing simultaneous harrier counts over large areas and to
F. Thiollay who assisted with most of the surveys and
helped with typing successive drafts of the manuscript.
We thank M. Bechard, M. Pandolfi and an anonymous
referee for improving a previous draft. This work was sup-
ported by a grant from Reunion (Regional Council) , Eu-
ropean Community (FEDER) and French Ministry of
Environment (DIREN-Reunion), for the study and con-
servation of terrestrial birds on Reunion.

Literature Cited

Barcelo, a. 1996. Analyse des mecanismes hydrolo-
giques en domaine volcanique insulaire tropical a re-
lief jeune. Apports la connaissance du Bilan Hy-
drique. Massif du Piton de la Fournaise (He de la
Reunion). These de Doctorat Hydrologie, Universite
de Montpellier II, France.

Barre, N. 1988. Une avifaune menacee: les oiseaux de
La Reunion. Pages 167-196 in ].C. Thibault and J.
Guyot [Eds.], Livre rouge des oiseaux menaces des
regions fran^aises d’Outremer. International Council
for Bird Preservation, Saint-Cloud, France.

, A. Barau and C. Jouanin. 1996. Oiseaux de la

Reunion. Editions du Pacifique, Paris, France.

Cadet, T. 1980. La vegetation de ITsle de la Reunion.

Imprimerie Cazal, Saint-Denis, Reunion.

Cheke, A.S. 1987. The ecology of the surviving native
land birds of Reunion. Pages 301—358 in A.W. Dia-

mond [Ed.], Studies of Mascarene Island birds. Cam-
bridge Univ. Press, Cambridge, U.K.

Clarke, R. 1995, The Marsh Harrier. Hamlyn, London,
U.K.

Clouet, M. 1978. Le busard de maillard, Circus aerugi-
nosus, de ITsle de la Reunion. O.R.F.O. 48:95-106.

Collar, N.J., M.J. Crosby and A.J. Strattersfield. 1994
Birds to watch 2. The world list of threatened birds
Birdlife International, Cambridge, U.K.

DEL Hoyo, J., A. Elliott and J. Sarcatal. 1994. Hand-
book of the birds of the world. Vol. 2. Lynx Edicions,
Barcelona, Spain.

Doumenge, C. and Y. Renard. 1989. La conservation des
systemes forestiers de ITsle de la Reunion. lUCN,
Cambridge, U.K.

Dupont, J. 1990. Cartes de vegetation de la Reunion. Sre-
pen, Saint-Denis, Reunion.

Environmental Systems Research Institute, Inc. 1996
Redlands, CA U.S.A.

Howard, R. and A. Moore. 1980. A complete checklist
of the birds of the world. Academic Press, London,
U.K.

MacDonald, I.A.W., C. Thebaud, W.A. Strahm and D
Strasberg. 1991. Effects of alien plant invasions on
native vegetation remnants on La Reunion (Masca-
rene Islands, Indian Ocean). Environ. Conserv. 18:51-
61.

Marchant, S. and P.J. Higgins. 1993. Handbook of Aus-
tralia, New Zealand and Antarctic birds. Vol. 2. Ox-
ford Univ. Press, Melbourne, Australia.

Nieboer, E. 1973. Geographical and ecological differen-
tiation in the genus Circus. Dissertation, Free Univ.,
Amsterdam, Netherlands.

Probst, J.M. 1997. Animaux de la Reunion. Azalees Edi-
tions, Saint-Denis, Reunion.

Rivals, P. 1952. Etude sur la vegetation naturelle de File
de la Reunion. Travaux du Laboratoire Forestier de
Toulouse, V, 3 sections. Vol. I. Faculte des Sciences de
Toulouse, Toulouse, France.

Sibley, C.G. and B. Monroe. 1990. Distribution and tax-
onomy of birds of the world. Yale Univ. Press, New
Haven, CT U.S.A.

Simmons, R.E. 1997. The African Marsh Harrier. Pages
236-237 mJ.A. Harrison, D.G. Allan, L.G. Underhill,
M. Herremans, A.J. Tree, V. Parker and C.J. Brown
[Eds.], The atlas of southern African birds. Birdlife
South Africa, Johannesburg, South Africa.

Sinctwir, I. and O. Langrand. 1998. Birds of the Indian
Ocean islands. Struik, Cape Town, South Africa.

Thiollay, J.-M. 1993. Habitat segregation and the insular
syndrome in two congeneric raptors in New Caledo-
nia, the White-bellied Goshawk, Acdpiter haplochrous,
and the Brown Goshawk, A. fasciatus. Ibis 135:237-246.

. 1997. Distribution patterns and conservation of

bird community and raptor populations in the An-
daman archipelago. Ecography 20:67-82.

Received 1 April 1999; accepted 7 October 1999

/. Raptor Res. 8-25

© 2000 The Raptor Research Foundation, Inc.

NEST-SITE SELECTION AND REPRODUCTIVE SUCCESS OF
URBAN RED-SHOULDERED HAWKS IN
CENTRAL CALIFORNIA

Stephen C. Rottenborn

/f.T. Harvey &' Associates, 3150 Almaden Expressway, Suite 145, San Jose, CA 95118 U.S.A.

Abstract. — Fledging success was determined at nests of urban Red-shouldered Hawks (Buteo lineatus)
in California. Fourteen of 27 nests in 1994 and 38 of 58 nests in 1995 were in exotic trees, predominantly
eucalyptus {Eucalyptus spp.). Nesting and fledging success were higher in exotic trees than in native
trees in both years, owing in part to greater stability and protective cover. Most nest trees in upland
areas (>100 m from water) were exotics, and even in riparian habitats, where tall native cottonwoods
{Populus fremontii) and sycamores {Platanus racemosa) were available. Red-shouldered Hawks selected
eucalyptus more often than expected based on their availability. Of the habitat and nest-tree variables
measured at each nest, only nest-tree height and diameter were significantly associated with reproductive
success, suggesting that large, sturdy trees provided the best nest sites. Red-shouldered Hawk populations
in the study area have likely benefited from the planting of exotic eucalyptus and fan palms. Repro-
ductive success was not affected by the degree of urbanization around nest sites, as many successful
nests were found in heavily urbanized areas close to human activity.

Key Words: Red-shouldered Hawk; Buteo lineatus; reproductive success; riparian; exotic trees; eucalyptus.

Seleccion del sitio del nido y exito reproductivo de Buteo lineatus urbanos en el centro de California

Resumen. — El exito de crianza de pichones fue determinado en los nidos de Buteo lineatus urbanos en
California. Catorce de 27 nidos en 1994 y 38 de 58 nidos en 1995 fueron en arboles exoticos, predom-
inantemente eucaliptus {Eucalyptus spp.). El exito de anidacion y de crianza fue mayor en arboles
exoticos que en arboles nativos en ambos anos, debido en parte a la mayor estabilidad y cobertura de
proteccion. Casi todos los arboles con nidos en areas superiores (>100 m del agua) fueron exoticos,
inclusive en habitats riberenos en donde Populus fremontii y Platanus racemosa estaban disponibles, Buteo
lineatus selecciono a los eucaliptus con mas frecuencia que lo esperado con base en su disponibilidad.
De las variables de habitat y de arboles con nido medidas para cada nido, solo la altura del arbol y el
diametro fueron significativamente asociados con el exito reproductivo, lo cual sugiere que los arboles
grandes y fuertes proveen los mejores sitios de anidacion. Las poblaciones de Buteo lineatus en el area
de estudio aparentemente se han beneficiado de la siembra de eucaliptus exoticos y palmas de abanico.
El exito reproductivo no fue afectado por el grado de urbanizacion alrededor de los sitios de anidacion,
pues muchos nidos exitosos se encontraron en areas altamente urbanizadas cerca de la actividad hu-
mana.

[Traduccion de Cesar Marquez]

Populations of Red-shouldered Hawks {Buteo li-
neatus) have declined throughout much of the spe-
cies’ range during this century (Henny et al. 1973,
Titus et al. 1989, Bednarz et al. 1990). Most of this
decline has probably resulted from the destruction
and fragmentation of hardwood forests, particular-
ly riparian woodlands, upon which Red-shouldered
Hawks are largely dependent for breeding over
most of their range (Cohen 1970, Henny et al.
1973, Bednarz and Dinsmore 1981). Urbanization
may also have contributed to the decline of this
species, as Red-shouldered Hawks in some areas

avoid nesting near roads and buildings (Bednarz
and Dinsmore 1981, Bosakowski et al. 1992).

Populations of the western race B. 1. elegans, re-
siding primarily in California, have been affected
by human activities to some degree. Breeding pri-
marily in riparian habitats, B. 1. elegans declined in
parts of its range early this century as a result of
severe degradation of California’s riparian wood-
lands (Willett 1912, Grinnell and Wythe 1927),
which reduced these habitats to <5% of their orig-
inal levels prior to European settlement of the state
(Katibah 1984). After this decline. Red-shouldered

18

March 2000

Urban Red-shouldered Hawk Nest Success

19

Hawk populations appeared to be fairly stable (but
low) in California from about 1950 until the late
1960s (Cohen 1970, Brown 1971, Wilbur 1973).

In the past three decades. Red-shouldered Hawk
populations have increased in some parts of Cali-
fornia. Breeding Bird Survey data indicate signifi-
cant statewide increases since 1966 (Sauer et al.
1997), and local increases and range expansions
have been noted in a number of locations since
then (Harlow and Bloom 1989, Roberson 1993,
S. A. Laymon^dg Shuford 1993). Because riparian
woodlands are still greatly reduced from their his-
toric extent, these population increases have not
resulted from recovery of the hawk’s native ripari-
an habitats, and the reasons for this recent in-
crease are not entirely clear.

Historically, California Red-shouldered Hawks
bred primarily in riparian areas, nesting in tall, na-
tive riparian trees and foraging in adjacent marsh-
es and grasslands (Willett 1912, Grinnell and
Wythe 1927, Grinnell and Miller 1944). Studies
conducted in the midwestern and eastern U.S.
have reported that Red-shouldered Hawks are sen-
sitive to human disturbance and that they usually
nest far from developed areas (Bednarz and Dins-
more 1981, Bosakowski et al. 1992). However, re-
cent studies in southern California have found a
number of Red-shouldered Hawks nesting in ex-
otic trees, such as eucalyptus {Eucalyptus spp.) and
palms {Washingtonia spp.), often in urban and sub-
urban areas surrounded by development (Bloom
et al. 1993, Bloom and McCrary 1996). Some Red-
shouldered Hawks nest where these tall, sturdy ex-
otic trees have been planted in upland areas, some-
times far from riparian habitats (Palmer 1988).

Bloom and McCrary (1996) have suggested that
planting of eucalyptus has contributed to the ex-
pansion of this species’ range and populations in
California by providing suitable nest trees in up-
land areas (including urban areas) that were pre-
viously unsuitable for breeding Red-shouldered
Hawks. However, no published studies have yet de-
termined the breeding success of Red-shouldered
Hawks in native versus exotic trees or in riparian
versus upland areas. This study investigated the de-
gree to which urban Red-shouldered Hawks nest
in exotic trees and/or upland areas, the breeding
success of Red-shouldered Hawks nesting in native
versus exotic trees and upland versus riparian ar-
eas, and relationships between breeding success
and habitat and nest tree variables related to ur-
banization and habitat quality.

Study Area and Methods

In 1994 and 1995, I located Red-shouldered Hawk
nests in the northern Santa Clara Valley at the southern
end of the San Francisco Bay in Santa Clara County, Cal-
ifornia. Formerly dominated by oak spp.) savan-

na (Clarke 1952), the northern Santa Clara Valley was
covered primarily by agricultural land at the turn of the
20th century but now is dominated by residential and
industrial development. These residential and industrial
lands are sparsely vegetated, with relatively few trees large
enough to support Red-shouldered Hawk nests. However,
scattered large eucalyptus (most commonly Eucalyptus
globulus) , palms and various coniferous trees provide po-
tential nest trees within these upland areas.

Two major streams, Coyote Creek and the Guadalupe
River, as well as a number of smaller streams, flow
through the urban Santa Clara Valley and from the foot-
hills of the Santa Cruz Mountains and Diablo Ranges to
San Francisco Bay. The narrow riparian corridors along
the portions of these streams flowing through the Santa
Clara Valley are dominated by Fremont cottonwood {Pop-
ulus fremontii) and several species of willows ( Salix luci-
dum, S. laevigata, S. lasiolepis and S. exigua) along their
lower reaches, and by oaks and western sycamore {Pla-
tanus racemosa) along their upper reaches. In addition,
scattered tall eucalyptus provide additional potential
nesting sites within these riparian areas.

I used several methods for locating Red-shouldered
Hawk nests (“nest” defined as the site of an actual nest-
ing attempt) . During late winter and spring of each year,
I searched for Red-shouldered Hawks and their nests m
riparian habitats along most of the streams in the study
area and in upland areas. Recordings of Red-shouldered
Hawk vocalizations were played frequently during surveys
to elicit a territorial response from any hawks that might
be present. Because Red-shouldered Hawks prefer tall
trees for nesting (Bednarz and Dinsmore 1982, Dijak et
al. 1990), surveys in upland areas focused on areas with
tall trees. Some nests were found using data from the
Santa Clara County Breeding Bird Atlas project and from
locations provided by members of the Santa Clara Valley
Audubon Society. Surveys were more extensive in 1995
than in 1994, particularly in upland areas.

After a nest was found, I periodically monitored it with
a spotting scope until the young had fledged or until the
adults had abandoned the nest (after nest failure). After
young were visible in nests, I visited nests every 3-4 d
until the young were near fledging age, at which point
visits were made every 2-3 d. The number of young fledg-
ing from nests was estimated by the number of young
seen in nests within three days of fledging. Nests aban-
doned before they were completed and “alternate nests”
not used for nesting were excluded from analyses if it was
thought that they might be located within the territory
of a pair whose completed nest was included in these
analyses.

Data were analyzed separately for each year. The mean
number of young fledged per nest and the number
fledged per successful nest (i.e., a nest from which at
least one young fledged) were compared between 1994
and 1995 using Mann-Whitney G-tests. Kruskal-Wallis tests
were used to compare the mean number of young fledg-
ing from each nest and the mean number fledging from

20

Rottenborn

VoL. 34, No. 1

Table 1. Tree species used for nesting by Red-shouldered Hawks in 1994. The number of nests found in each tree
species, number of nests that successfully fledged young and mean number of young per nest and per successful
nest are given separately for nests in riparian areas (<100 m from the nearest stream) and upland areas {>100 m
from the nearest stream).

Tree Species

Total Nests

Successful

Nests

Young Fledged
PER Nest

Young Fledged
PER Successful
Nest

Riparian

Platanus racemosa

6

4

1.5 ± 0.5

2.3 ± 0.3

Populus fremontii

7

4

1.3 ± 0.4

1.8 ± 0.2

Washingtonia spp.

2

1

1.0 ± 1.0

2.0

Eucalyptus spp.

6

5

1.8 ± 0.5

2.2 ± 0.4

Upland

Washingtonia spp.

2

2

3.0 ± 0.0

3.0 ± 0.0

Eucalyptus spp.

4

4

2.8 ± 0.3

2.8 ± 0.3

each successful nest among nests in different tree species.
Mann-Whitney U-tests were used to compare fledging
success between nests in native and exotic trees and be-
tween eucalyptus nests in upland and riparian areas, and
they were also used to compare fledging success between
nests attended by two adults and those attended by a pair
that included a subadult (second-year) bird. All Mann-
Whitney Latests and Kruskal-Wallis tests were two-tailed.

In each year, the number of successful nests and the
mean number of young fledging from nests was deter-
mined for each tree species, or, in the case of eucalyptus,
genus. Nests were also categorized by their location rel-
ative to the nearest stream, “riparian” nests being locat-
ed <100 m from the nearest stream and “upland” nests
being >100 m from a stream; few riparian-associated
trees (e.g., cottonwoods and sycamores) are found >100
m from a stream channel within the study area.

In a previous study of riparian tree communities (Rot-
tenborn 1997), 1 had measured the diameters of all trees
on 68 randomly located, 35 m-radius plots along the low-
er reaches of Coyote Creek, Guadalupe River and Los
Gatos Creek. Using these data, 1 determined the number
of stems of native and exotic trees on each of the 68 plots
that exceeded the diameter of the smallest tree used for
nesting by Red-shouldered Hawks in the present study. A
Gtest was used to compare the mean proportion of the
large-diameter stems comprised of exotic species to the
proportion of hawk nests from Coyote Creek, Guadalupe
River and Los Gatos Creek that were in exotic trees to
determine whether or not hawks used native and exotic
trees in proportion to their availability within this subset
of the study area. For this comparison, 1 used all hawk
nests found along these streams over both years, count-
ing nests used in both 1994 and 1995 only once.

Habitat and nest-tree variables thought to be related
to nesting success, including nest-tree height (TREEHT) ,
nest-tree diameter at breast height (TREEDBH), nest
height (NESTHT), elevation of the nest site (ELEV) and
distance from the nest to the nearest body of water
(DISTWAT) were measured at each nest. Within radii of
200 and 500 m of each nest, the percent cover by artifi-
cial surfaces ( ART1F200, ART1F500) , including buildings

and pavement, was estimated from aerial photos, and the
distance from each nest to the nearest building (DIST-
BUIL) and paved road (DISTPAVE) was measured.
Differences in the mean values of these variables at suc-
cessful and unsuccessful nests were tested using Mann-
Whitney U-tests. In addition, simple linear regressions of
each of these habitat variables versus the number of
young fledging from each nest were conducted. These
analyses were conducted separately for 1994 and 1995
nests.

Results

Twenty-seven and 58 Red-shouldered Hawk nests
were located in 1994 and 1995, respectively. Thir-
teen of the 1994 nests were used again in 1995, so
these two datasets were not independent and were
therefore analyzed separately. Twenty-one of the 27
nests found in 1994 (77.8%) and 46 of the 58 nests
found in 1995 (79.3%) were successful in fledging
at least one young. When all nests were combined
for each year, the mean number of fledged young
per nest was 1.8 ± 0.2 (±SD) in 1994 and 1.6 ±
0.1 in 1995 (t = 0.9, P > 0.05), while the mean
number of fledged young per successful nest was
2.3 ± 0.1 in 1994 and 2.0 ± 0.1 in 1995 (t = 0.7,
P> 0.05).

Use of Upland Versus Riparian Habitats. Six of

the nests found in 1994 (22.2%) and 18 of those
found in 1995 (31.0%) were in upland areas >100
m from the nearest stream (Tables 1 and 2). Nests
were located ^1.7 km from the nearest body of
water. When all nests were pooled, the mean dis-
tance between nests and the nearest body of water
was 143 ± 69.8 m in 1994 and 110 ± 33.0 m in
1995. The mean distance from each upland nest

March 2000

Urban Red-shouldered Hawk Nest Success

21

Table 2. Tree species used for nesting by Red-shouldered Hawks in 1995. The number of nests found in each tree
species, number of nests that successfully fledged young and mean number of young per nest and per successful
nest are given separately for nests in riparian areas (<100 m from the nearest stream) and upland areas (>100 m
from the nearest stream) .

Tree Species

Total Nests

Successful

Nests

Young Fledged
per Nest

Young Fledged
per Successful
Nest

Riparian

Platanus racemosa

14

11

1.6 ± 0.3

2.0 ± 0.2

Populus fremontii

4

3

1.0 ± 0.4

1.3 ± 0.3

Alnus rhombifolia

1

0

0.0

—

Washingtonia spp.

1

0

0.0

—

Eucalyptus spp.

20

17

1.9 ± 0.2

2.2 ± 0.1

Upland

Quercus agrifolia

1

1

0.0

—

Washingtonia spp.

3

2

1.3 ± 0.7

2.0 ± 0.0

Eucalyptus spp.

14

12

1.8 ± 0.3

2.1 ± 0.2

to water was 590 ± 268 m in 1994 and 302 ±91.7
m in 1995.

With the exception of a single nest in a coast live
oak (Quercus agrifolia) in 1995, all upland nests
were in exotic trees, predominantly eucalyptus.
Nest success (i.e., the percentage of nests that
fledged at least one young) differed little between
eucalyptus nests in riparian and upland areas (Ta-
bles 1 and 2) . The mean number of young fledged
from each eucalyptus nest in riparian and upland
areas was 1.8 ± 0.5 and 2.8 ± 0.3, respectively, in
1994, and 1.9 ± 0.2 and 1.8 ± 0.3 in 1995. The
mean number of young fledged from each suc-
cessful eucalyptus nest in riparian and upland ar-
eas was 2-2 ± 0.4 and 2.8 ± 0.3, respectively, in
1994, and 2.2 ± 0.2 and 2.1 ± 0.2 in 1995. Neither
of these measures of fledging success differed sig-
nificantly between riparian and upland nests in ei-
ther year.

Nest Tree Species and Use of Native Versus Ex-
otic Trees. Red-shouldered Hawk nests were found
in four tree species in 1994 (Table 1) and six tree
species in 1995 (Table 2). Native trees used for
nesting were Fremont cottonwood, western syca-
more, white alder {Alnus rhombifolia) and coast live
oak, while exotics included eucalyptus and fan
palms ( Washingtonia filifera and W. robusta) . Of the
27 nests found in 1994, 13 (48.1%) were in native
trees and 14 (51.9%) were in exotics. Of the 58
nests monitored in 1995, 20 (34.5%) were in native
trees and 38 (65.5%) were in exotics.

Eight of 21 nests found in riparian areas in 1994
(38.1%) and 21 of 40 riparian nests in 1995

(52.5%) were in exotics, primarily eucalyptus (Ta-
bles 1 and 2). Furthermore, the percentage of
nests in exotic trees along the lower reaches of
Coyote Creek, Los Gatos Creek and Guadalupe
River was much higher than the percentage of
large trees that were exotics. Only 18% of the trees
having a diameter greater than the most slender
Red-shouldered Hawk nest tree were exotics,
whereas 56% of the nests detected in the same ar-
eas were in exotic trees (G = 11.44, P < 0.01). A
number of nests were found in lone eucalyptus (N
= 3 in 1994, 8 in 1995) or small eucalyptus groves
(N= 3 in 1994, 12 in 1995) surrounded by mature
cottonwoods or sycamores; generally, these euca-
lyptus were taller than surrounding native trees.
Nest success seemed somewhat lower for nests in
native trees than for those in exotic trees (Tables
1 and 2) in both 1994 (61.5% in natives, 85.7% in
exotics) and 1995 (75.0% in natives, 81.6% in ex-
otics), although z-tests revealed no significant dif-
ferences in these percentages {P > 0.05) .

Because neither nesting nor fledging success dif-
fered significantly between riparian and upland
nests in eucalyptus and fan palms, riparian and up-
land nests were pooled for comparison of fledging
success among the nest tree species. In both years,
the mean number of fledged young per nest was
highest in eucalyptus, intermediate in sycamores
and fan palms and lowest in cottonwoods (Fig. 1).
The mean number of young that fledged per nest
differed among tree species in 1994 {H = 8.43, P
< 0.05) and in 1995 {H = 8.57, P < 0.05), pri-
marily due to the great differences in nest success

22

Rottenborn

VoL. 34, No. 1

. j N«

3 -

2 . 5 -

Nest tree

Figure 1 . Mean number of fledged young per nest and
per successful nest in four tree species used for nesting
by Red-shouldered Hawks in 1994 (a) and 1995 (b).

between eucalyptus and cottonwood nests. Because
fledging success did not differ significandy be-
tween cottonwoods and sycamores or between eu-
calyptus and fan palms, fledging success was
pooled for all natives and compared to fledging
success pooled for all exotics. The mean number
of fledged young per nest was higher in nests in
exotic trees than in nests in native trees in 1994 {U
= 132, P < 0.05), although the difference was not
significant in 1995 {t = 1.61, P > 0.05; Fig. 2).

The mean number of fledged young per suc-
cessful nest did not differ significandy among tree
species (Fig. 1). However, the mean number of
fledged young per successful nest was significandy
higher in nests in exotic trees than in nests in na-
tive trees in 1994 {U= 78, P< 0.05; Fig. 2). There-
fore, even though the lower nest success in native
trees (due to nest failures) was not a factor in these
comparisons, fledging success was still found to be
higher in nests in exotic trees in 1994. There was
no significant difference in 1995 (? = 1.45, P >
0.05; Fig. 2).

In 1994, one of the 14 nests in exotics (7.1%)
and five of the 13 nests in natives (38.5%) were
attended by a subadult (second year) individual
paired to an adult. Three of the four unsuccessful

I Native □ Exotic

2 . 5 -

1994 1995

■o

1994 1995

Year

Figure 2. Mean number of fledged young per nest (a)
and per successful nest (b) in native and exotic trees
used for nesting by Red-shouldered Hawks.

nests in native trees in 1994 were attended by a
pair that included a subadult, and the mean num-
ber of fledged young per nest in native trees was
significantly lower for those pairs containing a sub-
adult than for pairs with two adults in 1994 {U =
38.5, P< 0.01). In 1995, three of 20 nests in natives
(15.0%) and three of 38 nests in exotic trees
(7.9%) had a subadult attending. Although two of
the five unsuccessful native-tree nests in 1995 were
attended by a subadult, differences in fledging suc-
cess between pairs that did and did not include a
subadult were not significant in that year. Nest suc-
cess was also low for subadults that nested in exotic
trees; the single pair with a subadult that nested in
a fan palm in 1994 was unsuccessful, and two of
the three pairs including a subadult that nested in
eucalyptus in 1995 failed to fledge any young.

Relationship of Habitat Variables to Reproduc-
tive Success. The habitat and nest-tree variables
measured at each nest differed little between suc-
cessful and unsuccessful nests (Table 3). Nest-tree
height and nest-tree diameter were the only vari-

March 2000

Urban Red-shouldered Hawk Nest Success

23

Table 3. Comparison of nest tree and habitat variables between successful and unsuccessful Red-shouldered Hawk
nests in 1994 and 1995.

1994 Nests

1995 Nests

Successful

Unsuccessful

Successful

Unsuccessful

Variable

Nests (A - 21)

Nests (A = 6)

Nes'FS (A = 48)

Nests (A = 10)

DISTWAT

180.1 ± 86.7

14.0 ± 6.4

127.7 ± 40.7

43.1 ± 12.9

TREEHT

24.4 ± 1.2

18.3 ± 0.9**

23.8 ± 0.8

21.6 ± 1.5

TREEDBH

101.3 ± 5.5

76.5 ± 2.7*

97.6 ± 2.9

89.5 ± 5.2

NESTHT

17.3 ± 1.1

13.5 ± 0.9

15.9 ± 0.8

15.2 ± 1.3

ELEV

183.9 ± 27.4

179.7 ± 53.74

258.6 ± 22.6

186.4 ± 36.4

ARTIF200

0.43 ± 0.06

0.47 ± 0.15

0.41 ± 0.04

0.46 ± 0.09

ARTIF500

0.45 ± 0.06

0.49 ±0.16

0.43 ± 0.04

0.49 ± 0.08

DISTBUIL

109.1 ± 26.8

75.0 ± 31.1

93.0 ± 18.3

108.0 ± 35.5

DISTPAVE

80.6 ± 25.0

59.5 ± 24.3

65.5 ± 14.1

70.3 ± 19.2

* P < 0.05.
** P< 0.01.

ables having significant effects on nest success (and
then only in 1994), with successful nests being tall-
er and having a greater diameter than unsuccessful
nests. The simple linear regressions between these
habitat variables and reproductive success indicat-
ed that the number of young fledging per nest in-
creased with increasing nest-tree height in both
years and with increasing tree diameter in 1995 (P
< 0.05 for these individual regressions), but nei-
ther nest success nor fledging success was related
significantly to variables representing the degree of
urbanization around the nests. In fact, many of the
nests found in this study were in heavily urbanized
areas, and some trees supporting successful nests
were located as little as 2 m from a building and 1
m from the nearest road.

Discussion

Studies of nesting Red-shouldered Hawks in east-
ern North America have found few nests in exotic
trees (Bent 1937, Henny et al. 1973, Titus and
Mosher 1981, Dijak et al. 1990), probably because
they have been conducted primarily in heavily for-
ested areas. Even in California, most studies of
Red-shouldered Hawks have reported few nests in
exotic trees, owing in some degree to the rural or
natural study areas used (Dixon 1928, Wiley 1975).
However, 37.7% of the nests found in a study in
southern California, which included some urban
and suburban areas, were in exotic trees (Bloom
and McCrary 1996). My results showed that Red-
shouldered Hawks nesting in the heavily urbanized

south San Francisco Bay area commonly use exotic
trees.

Although Red-shouldered Hawks in some parts
of California nest in native oaks and pines well re-
moved from riparian areas (Roberson 1993, RH.
Bloom pers. comm.), only one nest >100 m from
a stream in this study was in a native tree. Large
native oaks and California bays {Umbellularia cali-
fornica) were present in some upland portions of
the study area, but on the floor of the Santa Clara
Valley, few native trees suitably large for nesting
were present outside riparian areas. Therefore, the
planting of tall, sturdy exotic trees in upland areas
has provided a number of suitable nesting sites
where few existed previously, greatly expanding the
extent of suitable breeding habitat and increasing
the number of territories within the study area.

Many pairs seemed to prefer eucalyptus even in
riparian habitats where mature cottonwoods and
sycamores were available. Numerous riparian nests
were found in lone eucalyptus or small eucalyptus
groves surrounded by mature cottonwoods and syc-
amores, and exotic tree species were used for nest-
ing more often than was expected based on their
availability. Hawks in the genus Buteo generally se-
lect nest trees in proportion to the availability of
trees suitably large for nesting, rather than seeking
out specific tree species (Dixon 1928, Bent 1937,
Bednarz and Dinsmore 1982). The frequent selec-
tion of eucalyptus for nesting in this study likely
reflects the taller, broader-limbed nature of many
eucalyptus nest trees compared to the native trees
available.

24

Rottenborn

VoL. 34, No. 1

The higher nest and fledging success in exotic
trees may reflect in part their greater resistance to
wind damage. Damage to unstable nests by winds
is a potentially important cause of nest failure or
abandonment, and successful Red-shouldered
Hawks tend to build their nests on large diameter
branches that provide stability in high winds (Bed-
narz and Dinsmore 1982, Dijak et al. 1990). In my
study, only one nest in an exotic tree was damaged
significantly by wind in each year, whereas in native
trees, four nests in 1994 and two in 1995 were
heavily damaged by wind. Both nesting and fledg-
ing success tended to increase with tree height and
diameter, probably due in part to the greater sta-
bility of large trees. In fan palms, nests were sup-
ported very well by a large number of thick peti-
oles. In eucalyptus, and to a lesser extent in
sycamores, nests were usually well supported by
thick branches, providing a sturdy platform for
nests; nesting success was similar in these two spe-
cies. However, nests in cottonwoods, which had the
lowest success, were usually supported by fewer, of-
ten more slender, branches than nests in other spe-
cies.

It is possible that the evergreen nature of euca-
lyptus and fan palms provided some additional
protection from predators and weather, as most
nests were initiated before deciduous natives leafed
out. Also, mammalian predators may have difficul-
ty reaching nests in eucalyptus trees given these
trees’ height and slick, thick trunks, while the
“skirts” of dead fan palm leaves may protect nests
in palms from mammalian predators (P.H. Bloom
pers. comm.).

The percentage of nests occupied by a subadult
parent, which tend to have low nest success (Palm-
er 1988) , was higher for native nests than for exotic
nests in both years. Eight of 12 pairs with a sub-
adult failed to fledge any young, and fledging suc-
cess in native trees was significantly lower for pairs
that included a subadult than for those including
two adults in 1994. If eucalyptus trees confer some
advantage to nesting hawks, then older, more ex-
perienced birds would be expected to prefer these
trees, perhaps relegating younger birds to more
marginal territories without tall eucalyptus.

The large number of Red-shouldered Hawk
nests found in exotics and the concomitant high
nest success seems to have bolstered Red-shoul-
dered Hawk populations in California. The pres-
ence of these trees in riparian areas may have par-
tially offset the loss of riparian woodland for this

species, and the addition of suitable nest trees in
upland areas that previously lacked appropriate
trees has likely augmented hawk populations.

Studies in some areas have found that this spe-
cies is generally sensitive to disturbance, usually
nesting far from human activity (Bednarz and
Dinsmore 1981, Bosakowski et al. 1992). In my
study, many nests were very close to homes, offices
and busy roads, and neither the percent cover by
developed areas around nests nor the proximity of
nests to buildings and roads influenced nest suc-
cess or fledging success. As reported by Bloom and
McCrary (1996), urban-nesting Red-shouldered
Hawks are well-adapted to urban environments,
and many are quite tolerant of human activity.

Results of my study indicated that California
populations of Red-shouldered Hawks have actu-
ally benefited from the planting of exotic trees in
urban areas. However, these results should not be
extrapolated to all Red-shouldered Hawk popula-
tions or to riparian-associated bird species in gen-
eral. Populations of Red-shouldered Hawks in oth-
er parts of the species’ range are still low or
declining, and it appears that persistence of those
populations is dependent upon the preservation of
large tracts of native woodland (Bednarz and Dins-
more 1981, Peterson and Crocoll 1992). Also, the
results of my study are not meant to encourage the
planting of exotic vegetation. Exotic plants often
lack the resources required by many native animal
species (Anderson et al. 1977, Mills et al. 1989).
Unlike the Red-shouldered Hawk, many riparian-
associated bird species have not recovered from
the effects of widespread riparian habitat degra-
dation and do not nest frequently in exotic vege-
tation.

Acknowledgments

This work was supported financially by grants from the
Santa Clara Valley Audubon Society and D. and K. Blau,
and by P.R. Ehrlich. I thank R. Strubbe for assistance in
finding nests, W.G. Bousman for providing hawk loca-
tions from Santa Clara County Breeding Bird Atlas data,
and members of the Santa Clara Valley Audubon Society
for providing some nest locations. The manuscript ben-
efited from comments by P.R. Ehrlich, D. Plumpton, P.H.
Bloom and an anonymous reviewer.

Literature Cited

Anderson, B.W., A.E. Higgins and R.D. Ohmart. 1977.

Avian use of saltcedar communities in the lower Col-
orado River valley. Pages 128-136 m R.R. Johnson and

D. Jones [Eds.], Importance, preservation and man-

March 2000

Urban Red-shouldered Hawk Nest Success

25

agement of riparian habitats. USDA For. Serv. Gen.
Tech. Rep. RM-43.

Bednarz,J.C. andJ.J. Dinsmore. 1981. Status, habitat use,
and management of Red-shouldered Hawks in Iowa.
J. Wildl. Manage. 45:236-241.

AND . 1982. Nest-sites and habitat of Red-
shouldered and Red-tailed Hawks in Iowa. Wilson Bull.
94:31-45.

, D. Klem, Jr., LJ. Goodrich and S.E. Senner.

1990. Migration counts of raptors at Hawk Mountain,
Pennsylvania, as indicators of population trends,
1934-1986. Auk 107:96-109.

Bent, A.C. 1937. Life histories of North American birds
of prey. Part 1. U.S. Natl. Mus. Bull. 167.

Bloom, P.H., M.D. McCrary and MJ. Gibson. 1993. Red-
shouldered Hawk home-range and habitat use in
southern California./. Wildl. Manage. 57:258-265.

Bloom, P.H. and M.D. McCrary. 1996. The urban buteo:
Red-shouldered Hawks in southern California. Pages
31-39 in D.M. Bird, D.E. Varland and J.J. Negro
[Eds.], Raptors in human landscapes: adaptations to
built and cultivated environments. Academic Press,
London, U.K.

Bosakowski, T, D.G. Smith and R. Speiser. 1992. Status,
nesting density, and macrohabitat selection of Red-
shouldered Hawks in northern New Jersey. Wilson
Bull. 104:434-446.

Brown, W.H. 1971. Winter population trends in the Red-
shouldered Hawk. Am. Birds 25:813-817.

Clarke, W.C. 1952. The vegetation cover of the San Fran-
cisco Bay region in the Early Spanish Period. M.S. the-
sis, Univ. California, Berkeley, CA U.S.A.

Cohen, S.E. 1970. The distribution of the western Red-
shouldered Hawk {Buteo lineatus elegans Cassin). M.S.
thesis, California State College, Long Beach, CA
U.S.A.

Dijak, W.D., B. Tannenbaum and M.A. Parker. 1990.
Nest-site characteristics affecting success and reuse of
Red-shouldered Hawk nests. Wilson Bull. 102:480-486.

Dixon, J.B. 1928. Life history of the red-bellied hawk.
Condor 30:228-236.

Grinnell, J. and A.H. Miller. 1944. The distribution of
the birds of California. Pacific Coast Avifauna No. 27.

and M.W. Wythe. 1927. Directory of the bird-life

of the San Francisco Bay region. Pacific Coast Avifau-
na No. 18.

Harlow, D.L. and P.H. Bloom. 1989. Buteos and the
Golden Eagle. Pages 102-110 in Proc. Western Raptor
Manage. Symp. and Workshop. Natl. Wildl. Eed., Natl.
Wildl. Fed. Sci. and Tech. Ser. 12.

Henny, C.J., F.C. Schmid, E.D. Martin and L.L. Hood
1973. Territorial behavior, pesticides, and the popu-
lation ecology of the Red-shouldered Hawk in central
Maryland, 1943-1971. Ecology 54:545-554.

Katibah, E.F. 1984. A brief history of riparian forests in
the Central Valley of California. Pages 23-29 m R E
Warner and K.M. Hendrix [Eds,], California riparian
systems: ecology, conservation, and productive man-
agement. Univ. California Press, Berkeley, CA U.S.A

Mills, G.S., J.B. Dunning, Jr. andJ.M. Bates. 1989. Ef-
fects of urbanization on breeding bird community
structure in southwestern desert habitats. Condor 91 •
416-429.

Palmer, R.S. 1988. Handbook of North American birds.
Vol. 5. Yale Univ. Press, New Haven, CT U.S.A.

Peterson, J.M.C. and S.T. Crocoll. 1992. Red-shoul-
dered Hawk, Buteo lineatus. Pages 333-351 in K.J.
Schneider and D.M. Pence [Eds.], Migratory non-
game birds of management concerns in the north-
east. USDI, Fish and Wildlife Serv., Newton Corner,
MA U.S.A.

Roberson, D. 1993. Red-shouldered Hawk. Pages 96-97
in D. Roberson and C. Tenney [Eds.], Atlas of the
breeding birds of Monterey County, California. Mon-
terey Peninsula Audubon Society, Monterey, CA
U.S.A.

Sauer, J.R., J.E. Hines, G. Gough, I. Thomas and B.G.
Peterjohn. 1997. The North American breeding bird
survey results and analysis. Version 96.3. Patuxent
Wildlife Research Center, Laurel, MD U.S.A.

Shuford, W.D. 1993. Red-shouldered Hawk. Pages 144-
146 m W.D. Shuford [Ed.], The Marin County breed-
ing bird atlas: a distributional and natural history of
coastal California birds. California. Avifauna Series 1.
Bushtit Books, Bolinas, CA U.S.A.

Titus, K., M.R. Fuller, D.F. Stauffer and J.R. Sauer.
1989. Buteos. Pages 53-64 in Proc. Northeast Raptor
Manage. Symp. and Workshop. Natl. Wildl. Fed ,
Washington, DC U.S.A.

AND J.A. Mosher. 1981. Nest-site habitat selected

by woodland hawks in the central Appalachians. Auk
98:270-281.

Wilbur, S.R. 1973. The Red-shouldered Hawk in the
western United States. West. Birds 4:15-22.

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

Willett, G. 1912. Birds of the Pacific Slope of southern
California. Pacific Coast Avifauna No. 7.

Received 25 February 1999; accepted 30 October 1999

J. Raptor Res. 34(l):26-32
© 2000 The Raptor Research Foundation, Inc.

HABITAT USE BY RED-TAILED HAWKS WINTERING IN THE

DELTA REGION OF ARKANSAS

Heath D. Garner^ and James C. Bednarz

Department of Biological Sciences, Arkansas State University, State University, AR 72467 US. A.

Abstract. — ^We examined use and avoidance of specific habitat types by Red-tailed Hawks (Buteojamai-
censis) during the winter in Arkansas. We conducted biweekly road surveys to determine the relative
number of hawks present throughout the winter season and to record which habitat types hawks used
for hunting purposes. Available cover types were used significantly (P < 0.05) out of proportion to their
availability. A total of 55.6% of Red-tailed Hawks used rice fields, which made up 49.3% of the available
habitat. Additionally, 12-1% of Red-tailed Hawks were in forest patches that comprised only 0.9% of our
study area. Hawk numbers were less than expected in soybean fields with only 26.5% of red-tails observed
in 39% of the available habitat. Likewise, 5.8% of Red-tailed Hawks were in wheat fields that made up
10.8% of the habitat available. Juveniles and adults were not observed to use the various cover types
differently (x^ = 2.221, P = 0.528). Numbers of Red-tailed Hawks over the winter season were signifi-
cantly correlated with rodent numbers (r = 0.618, P = 0.05). Our results suggested that Red-tailed
Hawks were both spatially and temporally affected by variations in rodent abundance during the winter
in the Delta region of Arkansas.

Key Words: Red-tailed Hawk, Buteo jamaicensis; habitat use, prey abundance, winter, Arkansas.

Uso del habitat de Buteo jamaicensis que pasan el invierno en la region del Delta del Arkansas

Resumen. — Examinamos el uso y el rechazo de tipos de habitat por parte de Buteo jamaicensis durante
el invierno en Arkansas. Llevamos a cabo conteos de carretera dos veces por semana para determinar
el numero relativo de gavilanes presentes a lo largo de la estacion invernal para registrar que tipos de
habitat utilizaban para cazar. Los tipos de cobertura disponibles fueron utilizados significativamente (P
< 0.05) por fuera de las proporciones disponibles. Un total de 55.6% de los gavilanes utilizaron cultivos
de arroz, equivalentes al 49.3% del habitat disponible. Adicionalmente, 12.1% de los gavilanes utilizaron
parches de bosque los cuales representaban tan solo el 0.9% del area de estudio. Los numeros de
gavilanes fueron menores de lo esperado en cultivos de soya con tan solo 26.5% de los gavilanes ob-
servados en el 39% del habitat disponible. El 5.8% de los gavilanes fue encontrado en cultivos de trigo,
los cuales representaban el 10.8% del habitat disponible. Los juveniles y adultos no utilizaron las co-
berturas indistintamente (x^ = 2.221, P = 0.528). Los numeros de gavilanes durante la estacion de
invierno fueron significativamente correlacionados con los numeros de roedores (r = 0.618, P = 0.05).
Nuestros resultados sugieren que los gavilanes cola roja, fueron ambos espacialmente y temporalmente
afectados por las variaciones de abundancia de los roedores durante el invierno en la region del Delta
de Arkansas.

[Traduccion de Cesar Marquez]

The Red-tailed Hawk {Buteo jamaicensis) is prob-
ably the most common raptor in North America
occupying almost every region except the arctic
circle (Preston and Beane 1993). It inhabits open
areas interdigitated with trees throughout its range
(Bednarz and Dinsmore 1982). Despite its abun-
dance, relatively little is known about the wintering
ecology and biology of the Red-tailed Hawk. A few

^ Present address: Southern Ecological Services, 1204 Ar-
rowhead Farm Road, Jonesboro, AR 72401 U.S.A.

studies addressing the topic have been completed
in recent years, including work by Preston (1990)
in western Arkansas and Lish and Burge (1995) in
western Oklahoma. However, neither of these stud-
ies addressed wintering habitat use by red-tails in
agricultural areas.

Wintering hawks arrive in Arkansas in October
from the northern prairie region in the northcen-
tral states and Great Lakes region (James and Neal
1986). These hawks aggregate in the Delta region
of Arkansas which, perhaps, supports the largest

26

March 2000

Habitat Use by Wintering Red-tails

27

HIGHWAY GRAV£L RGAD survey ROUTE DRAINAGE/CREEK

Figure 1. Map of the study area (101 km^) showing the raptor survey route (59 km). Each square represents 2.59
km‘^. The study area was located approximately 17 km south of Jonesboro, Arkansas, in northwestern Poinsett County
(insert) .

population of migratory Red-tailed Hawks in North
America (Garner 1997). We initiated this study to
increase our understanding of what factors con-
tribute to the Red-tailed Hawk’s winter abundance
and their winter ecology in this region. Our pri-
mary objective was to determine whether Red-
tailed Hawks use certain habitats or cover types
more frequently and avoid others in the Delta re-
gion during the winter season. Our null hypothesis
was that the hawks use all cover types in proportion

to their availability. The second objective of our
research was to determine what factors may be re-
sponsible for certain preferences or avoidances of
these habitat types.

Study Area and Methods

The study area was located in northern Poinsett Coun-
ty, Arkansas, approximately 17 km south of Jonesboro,
Arkansas (Fig. 1). It was rectangular in shape including
101.4 km^ of agricultural land. Sections marking the four
corners of the study area were: northwest — T12N, R2E,

28

Garner and Bednarz

VoL. 34, No. 1

S7, northeast — T12N, R3E, S7; southeast — TllN, R3E,
S19; and southwest — TllN, R2E, S19. Weiner, Arkansas
was just west of the center of the study area. The area
was selected because large numbers of Red-tailed Hawks
were observed in the area during years previous to our
study (Hanebrink et al. 1978).

Historically, this area supported large stands of bottom-
land hardwood forests including dominant species such
as white oak ( Quercus alba) , southern red oak ( Q. falcata ) ,
hickory (Carya spp.), sweetgum {Liquidambar styraciflua) ,
sycamore {Platanus ocddentalis) and baldcypress {Taxo-
dium distichum). Understory species consisted of red ma-
ple (Acer rubra), elms (Ulmus spp.), green ash (Fraxinus
pennsylvanica) (Preston 1989) and a variety of grasses and
forbs. These communities were supported by periodic
flooding of natural creeks and rivers in the area such as
the Cache and L’Anguille River systems, as well as nearby
Bayou DeView. All but a few of the historic bottomland
communities were cleared for agriculture and timber
production during the early 20th century (Wayne and
Gatewood 1993). Today, this area, as well as most of the
Arkansas Delta region, is primarily monoculture farm-
land producing rice, soybeans and wheat. These row crop
fields are usually bordered by drainage ditches, small
woodlots or occasionally fallow fields. Spillage of these
crops (rice, soybeans and wheat) from trucks during
transport provides food for rodents using grassy roadside
habitats.

During the winter of 1995-96, cover types and wood-
lots were mapped by driving roadways in the study area
and cumulative Red-tailed Hawk observations from all
surveys were overlaid onto the cover map. All hawks re-
corded on these surveys were observed within 500 m of
the roadway and the amount of available habitat was
quantified for each cover type within this distance of the
road. All hawk locations were categorized into one of the
four cover types: rice fields, soybean fields, wheat fields
and forested areas. Narrow ruderal roadside areas were
not considered a cover type because all of the observa-
tions were made from roads. Therefore, all cover types
where Red-tailed Hawks were observed were adjacent to
roadside areas.

A cover map was also created using a geographic in-
formation system (GIS) data base. Roads, vegetation cov-
er, human habitation areas and waterways were entered
as data layers into a GIS. Data were digitized from Land-
sat 5 satellite imagery taken during May of 1992 and from
topographic maps. The GIS image allowed an accurate
measurement of cover types in the study area which did
not change between 1992 and the time fieldwork was
complete in 1996. A Bayesian classification program was
used to delineate habitat types by statistically comparing
reflectance values and combining them if no significant
difference was detected. The changing crop types from
year to year prevented accurate classification of row crop
habitats based on the 1992 data. Therefore, all agricul-
tural areas were lumped into one category, “row crops.”
The forested areas identified in the 1992 satellite image
remained the same and were present in the study area
during the 1995-96 winter season.

Cover types within the study area were as follows:

Rice fields. Rice was grown in the majority of the fields
within the study area and were normally harvested by

October and the stubble was left standing throughout
most of the winter until the planting season began in
March. Many of these fields were intentionally flooded
in winter to attract waterfowl. Rice fields were typically
divided by several dikes which corresponded to a 2.54 cm
change in elevation. These small dikes were constructed
so that a constant water level was maintained within each
partition of the rice field. Eields varied in size, ranging
from <50-246 ha and were sometimes bordered by any
of the other cover types.

Bean fields. Soybeans were the second most common
cover type in the study area. Beans were usually harvested
by the end of September or early October. These fields
were seldom left with stubble after harvest and some were
plowed under for the planting of winter wheat. After har-
vest, these fields contained virtually no cover and wildlife
were seldom observed in them.

Wheat fields. These were relatively few in number.
Wheat plants usually were not visible until December, at
which time seedlings developed and provided some cover
for wildlife. Wheat fields matured in May and were har-
vested in June, well after the winter study period. How-
ever, the seed was transported and planted during the
winter season and spillage may have influenced the num-
ber of rodents using roadways during this period.

Forests /woodlots. Forested areas were fragmented and
limited throughout the study area. Woodlots ranged in
size from 0.5-80 ha and varied in age throughout the
study area. These woodlots were composed primarily of
hardwood species such as red oak, white oak, sweetgum,
sycamore and hickory, as well as a variety of understory
species. Woodlots typically followed drainage areas such
as creeks and run-off ditches. Some smaller patches of
forest, however, were in the middle of agricultural fields
or along roadways. There were 21 separate woodlots pres-
ent in the study area (excluding Bayou DeView Wildlife
Management Area) that comprised a total area of 485 ha
(x = 23.1 ha).

Fallow areas. Fallow fields were the rarest cover type in
the study area and were found within the city limits of
Weiner, adjacent to grain elevators and between a rail-
road track and U.S. Highway 49. During the study period,
crops were not planted in these areas or fields. Fallow
fields supported native grass communities and early-suc-
cessional plants including bluestem (Andropogon spp.),
broomsedge (Andropogon virginicus), Indian grass (Sor-
ghastrum nutans), goldenrods (Solidago spp.), prairie
three-awn (Aristida oligantha) and sumacs (/^A'U.s spp.).

Water bodies. A total of 17 water bodies was in the study
area ranging in size from 3-61 ha (x = 14.3 ha). Many
of these ponds were constructed for fish farming or as
irrigation reservoirs.

We established a systematic road survey (59 km) that
ran throughout the study area (Fig. 1). During these sur-
veys, all red-tails and other raptor species observed were
tallied. Soaring hawks were also counted but were not
included in the data analysis because no association with
one habitat type could be determined. Two observers
(one driver and one observer) were present during all
surveys which were conducted between 0800-1200 H
Weather conditions varied among surveys; however, if
there was steady rainfall or winds greater than 32 km/hr,
we postponed surveys until the following morning.

March 2000

Habitat Use by Wintering Red-tails

29

Weather conditions were determined by reports supplied
by the local airport via the National Weather Service. This
same route was run approximately every 2 wk and the
starting points of each survey were alternated in order to
minimize a time-of-day bias. Subspecies or color morph
(as described by Clark and Wheeler 1987), age (juvenile
or adult) and location along the route were recorded for
each hawk observed. Finally, we recorded the habitat type
where each individual Red-tailed Hawk was observed.

The relative rodent abundance in each mcyor habitat
type was measured by trapping with Sherman live traps
(Preston 1990). Traplines, consisting of 50 traps, each
placed 10 m apart, were located in a rice field, woodland
edge, roadside area, soybean field and wheat field. In the
rice field, woodland edge habitat and roadside area, two
replicate samples were collected on consecutive days (to-
tal sample = 100 trap-nights) twice during two field sea-
sons (1994-95 and 1995-96). Data for only the 1995—96
field season are presented but results were consistent for
both field seasons (Garner 1997) . Rodent abundance was
sampled (100 trap-nights) once in December-January
and once in March within all habitat types and replicates.
Soybean fields and wheat fields were sampled once dur-
ing the 1994-95 field season but no rodents were trapped
in either habitat type. Because of the low numbers of
rodents using these areas, further trapping in these hab-
itat types was stopped. Only the roadside and woodland
edge sites were sampled for rodents in early March 1996
because the rice field habitat site was tilled under by late
February.

Statistical analyses were conducted using Statistical
Analysis System (SAS 1985) programs. Habitat availability
and use were analyzed according to Neu et al. (1974).
This analysis is performed using the distribution of hawks
observed in certain cover types and comparing this to the
number of hawks expected to use the cover types based
on availability. Significance of use or avoidance was based
on inclusion of the proportion of available habitat type
within a confidence interval based on actual use of that
habitat. Proportional use of specific habitat and cover
types by juvenile and adult hawks was tested with a chi-
square analysis. Pearson correlation analyses were also
conducted to examine possible temporal relationships
between observed hawks and rodent abundance. These
correlations were based on rodent numbers sampled and
the observed hawk numbers recorded on surveys con-
ducted within 10 d of rodent sampling.

Results

During the 1995-96 winter surveys, we recorded
275 Red-tailed Hawk sightings. We observed 153
(55.6%) hawks in rice fields and 73 hawks in soy-
bean fields (26.5%) . Rice fields were the most com-
mon cover type available (49.3% of the habitat sur-
veyed) followed by soybean fields (39%; Table 1).
We recorded 33 and 16 sightings in woodlots (for-
est edge) and wheat fields, respectively. Wheat
fields comprised 10.8% of the cover types, and for-
ests made up 0.9%. A slight preference was ob-
served for rice fields (use = 55.6% vs. availability

O

C

_o

'be

4;

u

M

Q

V

.C

4D

Oi

I

in

CTi

<U

Ui

3

O

■ 4 _>

o,

cb

Ui

be

'C

S-i

OJ

y>

0
y}

1

X

u

T)

O

'.M

3

b

V

"d

3

rt

yi

<U

IS

I

H

o in 00
CO o 00 in

nH Csl I— ^ i-H

H

nJ

(M

on

II

o in csi 00

J> ^

00 1> 00 00

z

0

U

m

z

o

g

g

CL,

(Z

P

tM

o

X 00 J> o
cn cn oi i>

40 cn p rH

2 ? ? ?

4^ CM cn o

i> I

l-H o o
o o o ci

z

0

W

*

*

H

CO

m

X

o

oi

m

CO

m

CM

o

0

m

CM

o

1-^

p

Ph

0

d

0^

0

d

d

d

d

on CO 40 cn m
in ^ cn J>

,-H

p cn j>

m oi (M m
cn o CM u-

nH , — I CM

O

H

cn o X o

O i

cn ,-H o o

d d d d P

I

o o o o o

cn CO in on cn

m CM cn CM

r— , ,-H cn

§

a

V <ri
O

V

Vi

*

of habitat was significantly different than available (a = 0.05) .

30

Garner and Bednarz

VoL. 34, No. 1

Table 2. Relative abundance indices for rodents (rodents per 100 trap-nights) and species composition sampled in
different habitats during the winter of 1995-96 and spring of 1996 in the Delta region of Arkansas.

Habitat Type

Date

Rodent

Index

Percent Composition per Sample Period

Sigmodon

hispidus

Oryzomys

palustris

Microtus

ochrogaster

Rattus

norvegicus

Mus

musculus

Roadside

29-30 Dec

59/100

69.5

16.9

5.1

1.7

6.8

Rice field

6-7 Jan

13/100

7.7

38.5

15.4

0

38.5

Woodland edge

4— 5 Jan

7/100

42.9

57.1

0

0

0

Mid-winter Means

26.3/100

40

37.5

6.8

0.6

15.1

Roadside

2-3 Mar

34/100

52.9

20.6

17.6

0

8.8

Woodland edge

4-5 Mar

1/100

0

100

0

0

0

Early Spring Means

17.5/100

26.5

60.3

8.8

0

4.4

= 49.3%), but this pattern was not significant {P
> 0.05) . Soybean and wheat fields were avoided {P
< 0.05) by red-tails (Table 1). Hawks significantiy
favored woodlots (P < 0.05), as these areas were
overused (12%) compared to their availability
(0.9%; Table 1). We observed mostly adults in all
cover types (80-87%; Table 1). Juvenile numbers
were slightly higher (20.5%) in soybean fields than
in other habitats. The proportion of adults and ju-
veniles observed among cover types, however, was
not significantly different than expected (x^ =
2.221, df = 3, P = 0.528).

Roadsides were sampled first on 29—30 Decem-
ber 1995 resulting in a capture of 59 rodents dur-
ing 100 trap-nights. This was the highest estimate
of rodent relative abundance recorded during the
early trapping period (December-January) (Table
2). Rice fields were sampled 6-7 January 1996,
when we recorded an abundance index of 13 in-
dividuals per 100 trap-nights. The woodlot site reg-
istered a relatively low rodent relative abundance
of 7 rodents/ 100 trap-nights. No rodents were cap-
tured during 100 trap-nights in bean and wheat
fields sampled. Cotton rats {Sigmodon hispidus,
57%), rice rats {Oryzomys palustris, 24%) and house
mice (Mus musculus, 11%) were the three most
common species sampled during the 1995—96 mid-
winter sampling period.

During the early spring sampling period, rodent
abundance decreased in the two habitat types sam-
pled with the woodland edge site showing the big-
gest decrease (7 to 1 rodent/ 100 trap-nights; Table
2). The roadside habitat produced a total of 34
individuals per 100 trap-nights; cotton rats were
most abundant making up 52.9% of the sample.

Pearson correlation analysis of road census data
with rodent sampling indices showed that Red-

tailed Hawk numbers and rodent numbers were
positively correlated (r = 0.618, N = 5, P = 0.05;
Fig. 2) , suggesting that Red-tailed Hawk use of hab-
itats in the Delta may also have been correlated
temporally with rodent abundance.

Discussion

Based on 2 yr of winter survey data, we obtained
an overall mean Red-tailed Hawk abundance index
of 5.02 hawks observed per 10 km (winters of
1994—95 and 1995-96). Our findings surpass the
highest densities previously reported by Lish and
Burge (1995) in Oklahoma. They compared their
findings of 3.78 hawks per 10 km to 23 other stud-
ies around the U.S. As far as we can determine,
our data in Arkansas represent the highest winter
density of Red-tailed Hawks ever reported based on
extensive or repeat road surveys.

Hawk numbers were highest (55.6% of hawk ob-
servations) in rice fields along the road survey
route; however, rice fields were also the most abun-
dant cover type occupying 1590 ha (49.3%) of area
sampled. The second most abundant field type
along the route was soybeans (1260 ha). The sec-
ond highest frequency of hawks (26.5%) was
counted in bean fields but hawk numbers observed
in these fields were less than expected when con-
sidering the availability of bean fields (39% of the
study area) . Hawks were observed using woodlots
significantly more often (12.1%) than expected
(0.9%) based on availability but used wheat fields
less than expected.

Preston (1990) reported that a combination of
rodent availability and vegetative structure influ-
enced the distribution of foraging hawks during
the winter in western Arkansas. He found that Red-
tailed Hawks used corn stubble and “old fields”

March 2000

Habitat Use by Wintering Red-tails

31

“vi

4->

.c

50

T3

O)

C

ni

E

40

(A

a

(0

n

o

X

o

T“

30

0

L.

0)

0)

Q.

4 Jan 95 -10 Jan 95 17 Mar 95 - 23 Mar 95 29 Mar 95 - 31 Mar 95 29 Dec 95 - 3 Jan 96 2 Mar 96 - 3 Mar 96

Time Periods

Figure 2. Total number of Red-tailed Hawks recorded during raptor surveys compared to rodent abundance indices
m rice fields and roadsides sampled within the same time periods.

for foraging more than bare ground or tall corn.
Similarly, Baker and Brooks (1981) found that Red-
tailed and Rough-legged Hawk {Buteo lagopus) den-
sities were relatively low in the habitat types with
the highest meadow vole (Microtus pennsylvanicus)
densities in Ontario. Based on this pattern, they
suggested that vegetative structure of each habitat
affected the vulnerability of the vole population.
Habitat types with high densities of voles and high
prey vulnerability (e.g., little cover) supported the
highest densities of hawks. In our study, Red-tailed
Hawks seemed to respond primarily to rodent
abundance. This may have been attributed to the
relatively similar vegetation structure found in
both rice fields and roadside areas. Bean fields had
little to no cover and supported few or no rodents.
However, when cover was removed from rice fields
by tilling, red-tails were observed to congregate in
these fields and feed on fleeing rodents. This rep-
resents a case where the prey cover was substan-
tially reduced, dramatically increasing vulnerability
of rodents using those habitats. We observed at
least 50 Red-tailed Hawks in one field while a trac-
tor was discing rice stubble on 23 February 1995.
Thus, in this aspect, our observations agree with
Preston’s (1990) that cover affects the prey vulner-
ability, and thus, availability of rodents. However,
in the Delta region of Arkansas, winter cover nor-
mally does not vary among key habitats, including

rice fields until spring planting occurs and Red-
tailed Hawks seem to respond to prey abundance
in most circumstances.

Preston (1990) also suggested that perch avail-
ability influenced red-tail numbers using certain
cover types. The abundance of utility poles along
almost every roadway in our study area probably
provided ample perch sites. In addition, woodlots
provided essentially a continuous line of perch
sites along the forest edge for red-tails that were
likely foraging in adjacent cover types. The abun-
dance of perch sites may explain the preference of
Red-tailed Hawks for forests (woodlots) despite
their lower rodent availability. The presence of
hawks within or next to bean fields probably rep-
resented situations in which hawks were foraging
in adjacent ruderal roadside habitats. Alternatively,
some of these hawks may have been resting and
not actually foraging. One notable difference be-
tween our study and Preston’s (1990) study was the
presence of roadside habitats. An abundance of
roadways throughout our study area provided a
substantial amount of ruderal roadside habitats,
which may have supported higher numbers of ro-
dents than the stubble fields sampled by Preston
(1990).

The observations of overall higher rodent abun-
dance along roadside habitats and that red-tails
counted during surveys were within 500 m of road-

32

Garner and Bednarz

VoL. 34, No. 1

sides suggested that hawks were associated with
roadside habitats. Rodent abundance along road-
sides and rice fields was higher than in other avail-
able cover types and we suggest that this factor may
explain the high hawk numbers observed adjacent
to, or in, these habitats. Our sampling showed dra-
matic differences in rodent numbers between ro-
dent-rich rice fields and roadsides and rodent-poor
bean and wheat fields. Our casual observations
suggested that these differences held true through-
out our study area and throughout the winter; we
frequently observed rodents during the day in
grassy roadsides, but never saw a rodent in a bean
or wheat field. However, we only sampled one ex-
ample of each habitat with traps; most of these sites
were sampled four times over two years with con-
sistent results. These data supported the hypothesis
that hawks more often occur near roadsides and
rice fields because they support high densities of
rodents, but this hypothesis requires further testing
with replicate spatial sampling.

Acknowledgments

Special thanks are extended to L. Childers for helping
with the CIS element of this project. We would also like
to thank N. Stinnett Garner, T. Chen, J. Williams and T.
Klotz for volunteering many hours in the field to help
collect data. We especially thank the Northeastern Arkan-
sas Audubon Society and the Arkansas State University
(ASU Faculty Grant to J. Bednarz) for providing partial
funding for this project. In addition, we thank V.R.
McDaniel and B. Bennett for their helpful advice during
this study. Thanks to R. Perry, W. Stout, B. Millsap and
D Smith for their reviews of earlier versions of this man-
uscript. We appreciate the efforts of J. Smith for recov-
ering the remains of bird “N/O.” The authors are es-
pecially indebted to A. Ruswald who recovered and
returned a lost field notebook.

Literature Cited

Baker, J.A. and R.J. Brooks. 1981. Distribution patterns
of raptors in relation to density of meadow voles. Con-
dor 83:42-47.

Bednarz, J.C. and J.J. Dinsmore. 1982. Nest-sites and hab-
itat of Red-shouldered and Red-tailed Hawks in Iowa.
Wilson Bull. 94:236-241.

Clark, W.S. and B.K. Wheeler. 1987. A field guide to
hawks. Houghton Mifflin Co., Boston, MA U.S.A.

Garner, H.D. 1997. Dynamics and stability of a popula-
tion of wintering Red-tailed Hawks in the Delta region
of Arkansas. M.S. thesis, Arkansas State Univ., State
University, AR U.S.A.

Hanebrink, E.L., K. Sutton and A.F. Posey. 1978. Spe-
cies composition and diversity of hawk populations in
northeastern Arkansas. Proc. Ark. Acad. Sci. 32:51-54.

James, D.A. and J.C. Neal. 1986. Arkansas birds, their
distribution and abundance. Univ. Arkansas Press,
Fayetteville, AR U.S.A.

Lish, J.W. and L.J. Burge. 1995. Population characteris-
tics of Red-tailed Hawks wintering in tallgrass prairies
in Oklahoma. Southwest Nat. 40:174-179.

Neu, C.W., C.R. Byers and J.M. Peek. 1974. A technique
for analysis of utilization-availability data. J. Wildl
Manage. 38:541-545.

Preston, C.R. 1990. Distribution of raptor foraging in
relation to prey biomass and habitat structure. Condor

92:107-112.

AND R.D. Beane. 1993. Red-tailed Hawk {Buteoja-

maicensis). In A. Poole and F. Gill [Eds.], The birds of
North America, No. 52. The Academy of Natural Sci-
ences, Philadelphia, PA and The American Ornithol-
ogists’ Union, Washington, DC U.S.A.

Preston, R.J., JR. 1989. North American trees: exclusive
of Mexico and tropical Florida. 4th Ed. Iowa State
Univ. Press, Ames, lA U.S.A.

SAS. 1985. SAS user’s guide: statistics. 5th Ed. SAS Insti-
tute, Inc., Cary, NC U.S.A.

Wayne, J. and W.B. Gatewood [Eds.]. 1993. The Arkan-
sas Delta: land of paradox. Univ. Arkansas Press, Fay-
etteville, AR U.S.A.

Received 1 April 1999; accepted 7 October 1999

J. Raptor Res. 34(l):33-36
© 2000 The Raptor Research Foundation, Inc.

AN EVALUATION OF THE ANDEAN CONDOR POPULATION IN

NORTHERN ECUADOR

Marcus T. Koenen^ and Sarah Gale Koenen^

Peace Corps, Casilla 157, Otavalo, Ecuador

Norma Yanez^

Universidad Technica del Norte, Quito, Ecuador

Abstract,— We evaluated the population structure of Andean Condors {Vultur gryphus) in the Cotacachi-
Cayapas and Cayambe-Coca Ecological Reserves, Ecuador. We conducted 1298 hr of fieldwork and made
496 condor observations. Age class and sex could be determined in 127 and 48 observations, respectively.

The population consisted of 1:1 female to male ratio yet only 20% of our observations were of juveniles
and subadiilts. The apparent skewed population structure suggested that the population may be declin-
ing.

Key Words: Andean Condor, Vultur gryphus; paramo; Ecuador, pcrpulation structure, vulture.

Evaluation de la poblacion del condor andino en el norte del Ecuador

Resumen. — Evaluamos la estructura de poblacion del condor andino {Vultur gryphus) en las reservas
ecologicas de Cotacachi-Cayapas y Cayambe-Coca, Ecuador. LLevamos a cabo un estudio de campo de
1298 horas e hicimos 496 observaciones de condor. Las clases por edad y sexo pudieron ser determi-
nadas en 127 y 48 observaciones, respectivamente. La poblacion consistio de 1:1 entre machos y hem-
bras, tan solo el 20% de las observaciones fueron juveniles y subadultos. La aparentemente sesgada
estructura poblacional sugiere que la poblacion puede estar declinando.

[Traduction de Cesar Marquez]

The Andean Condor ( Vultur gryphus) ranged his-
torically from Venezuela to Tierra del Fuego (Mur-
phy 1936). Its present range is greatly reduced
(McGahan 1972, Lieberman et al. 1993) and the
Andean Condor is now listed as Endangered over
its entire range due to its precipitous decline (U.S.
Fish and Wildlife Service 1986) and considered
critically imperiled in Ecuador (Granizo et al.
1997). In Ecuador, condors inhabit the paramos
above 3000 m (Josse and Anhalzar 1996). Carrion
consisting of wild and domestic ungulates make up
their primary diet (McGahan 1972). Given the
patchy distribution of their food resource, group
foraging behavior increases the probability of lo-

^ Present address: Wings of the Americas^ — The Nature
Conservancy, 4245 N. Fairfax Dr. Suite 100, Arlington, VA
22203-1606 U.S. A.

‘^Present address: The Wildlife Center of Virginia, P.O.
Box 1557, Waynesboro, VA 22980 U.S.A.

^ Present address: Garcia Moreno 522 Oficina 2, Ma-
chachi, Ecuador.

eating food over individual searching by lone birds
(Wallace and Temple 1988a).

The paramo habitat is under heavy human pres-
sures from agriculture, intensive livestock manage-
ment and tourism (Caberle et al. 1989, Luteyn
1992). Numerous condors have been found dead
in recent years but a systematic population study
has not been conducted. This study was undertak-
en to evaluate the Andean Condor population in
northern Ecuador.

Study Area and Methods

We conducted fieldwork from August 1996— March
1998 in the paramos above 3000 m of the Cotacachi-Cay-
apas Ecological Reserve (CCER; 0°25'N, 78°20'W) and
the Cayambe-Coca Ecological Reserve (CAER; 0°08’N,
78°00’W) (Fig. 1). Additionally, we conducted fieldwork
at Lake Mojanda, a proposed protected area (0°08'N,
78°17W). All areas lie in northern Ecuador. The CCER is
within the western cordillera of Imbabura Province and
the CAER is in the eastern cordillera of Pichincha Prov-
ince. Lake Mojanda lies between the cordilleras.

Paramo is an equatorial alpine grassland ecosystem
dominated by bunchgrasses (Festuca spp.) and character-
ized by shrubs {Polylepis incana, Brachyotum alpinum, B.

33

34

Koenen et al.

VoL. 34, No. 1

Figure 1. Approximate area (shaded) covered by con-
dor surveys in 1996—98 in the Cotacachi-Cayapas Ecolog-
ical Reserve (CCER), Cayambe-Coca Ecological Reserve
(CAER) and Lake Mojanda area, Ecuador.

kdifolium, Chuquiraga jussieui) , ground level plants {Hu-
perzia crassa, Valeriana rigida, Lupinus sp.) and giant ro-
sette-plants (Pwyaspp.) (Luteyn et al. 1992, Josse and An-
halzar 1996). The paramos cover approximately 25 000
ha (Josse and Anhalzer 1996), or roughly a little over 2%
of Ecuador’s landcover. Mean temperature is 8°C and to-
tal annual rainfall ranges from 900-2600 mm ( Josse and
Anhalzar 1996).

Surveys were conducted on week-long trips to the study
areas. Day-long surveys were made from 19 ridgetops that
were identified as condor viewing areas by park rangers
and local residents who knew the study areas well. Ten
of these ridgetops were located in CCER and known lo-
cally as: La Cienega, Pulumbura, El Campanario, Haci-
enda Chinchivi, Cerro Pilabo, Cerro Nagharo, Cerro
Quilili, Pantavi Grande, Pantavi Chico and Las Antenas.
Eight sites were used in CAER: Rasochupa, Turupamba
Chico, Turupamba Grande, Quebrada de Mirlos, Que-
brada Chimborazo, Ancholas, La Dormida and El Verde.
In addition, the area adjacent to Lake Mojanda was used.

Fieldwork was generally conducted during the day be-
tween dawn and dusk. Any surveys made while traveling
through paramo by horse, foot, four-wheel-drive vehicles,
or when stationary were counted toward hours in the
field except during heavy rains when condors were less
likely to be observed. All observations were made using
8X and 10 X binoculars and 20-60 X spotting scopes.
Photographs of condors were taken with 280 mm and
500 mm lenses when possible from 13 August 1996-22
January 1997.

We distinguished individual condors based on feather

wear and molt patterns and identified sex when possible
(Snyder and Johnson 1985). We noted condor behavior
for evidence of breeding status (McGahan 1972, Palmer
1988). Condors were grouped into adult (^8 yr old) and
immature (<8 yr old) age classes based on plumage char-
acteristics described by McGahan (1972). We assumed
that different age groups and sexes had equal chances of
being observed in the field.

Results

We spent 1298 hr in the field and made 496 con-
dor observations from 19 ridgetops. Of these ob-
servations, 298 condor sightings were in the CCER,
186 in the CAER and 12 at Lake Mojanda. For each
hour spent in the field, we were able to make 1.3
min of condor observations. We were able to de-
termine the sex in 48 observations and estimate
the condor’s age in 127 observations. Our sample
revealed a male to female ratio of 1:1. Adults com-
prised 80% of condors aged giving us a 4:1 ratio
of adults to immatures. The maximum number of
condors simultaneously in view was eight birds in
the CCER. We were able to identify at least seven
adult and four immature condors for all study sites
based on a comparison of feather characteristics
and molt patterns.

One nest was reported to us in May 1996 and
was situated 100 m above an active tunnel construc-
tion site adjacent to the CAER (INEFAN 1997). A
second pair of adult condors was seen copulating
in September 1997 near Lake Cuicocha of the
CCER. The pair was seen repeatedly in the area
through March 1998 but no nest was located. We
found no signs of additional breeding pairs.

Discussion

The results of our study suggested Andean Con-
dors may be undergoing a population decline in
northern Ecuador. Temple and Wallace (1989) de-
termined in Peru that Andean Condors >6 yr old
had a 94% survival rate. Independent juvenile con-
dors between 1-6 yr old had a 90% survival rate,
whereas dependent juvenile condors <1 yr old had
a 74% survival rate. Breeding once every 3 yr is
required to maintain a stable population. Wallace
and Temple (1988b) calculated that a 1:1 ratio of
adult to juvenile indicated a population with pairs
that were breeding about once every 2 yr. The
adult male to female ratio (1:1) found in our study
suggested that pairing should occur. However, the
4:1 ratio of adult to immature suggested that the
population was declining because the adult cohort
was not being replaced.

March 2000

Condor Population Evaluation

35

This skewed age structure in the population may
be the result of high mortality. Five Andean Con-
dors were reported dead in 1987 by Cayambe res-
idents. INEFAN park rangers also reported six
dead condors near the Antisana Ecological Reserve
m 1988. Five more were found killed by unknown
factors in CCER in the early 1990s. Causes of mor-
tality were not reported. One condor, however, was
reportedly killed by Compound 1080 in 1990 and
a car collision was suspected as the cause of a mor-
tality in 1996 (L. Martinez pers. comm.). Whereas
we expected that immatures had a much higher
mortality rate than adults (Wallace and Temple
1988b), we were unable to determine the degree
that mortality factors affected the immature con-
dor population in Ecuador. Given the low proba-
bility of finding or reporting dead condors, it was
likely that these incidents represented a small frac-
tion of the condors that presumably died since
1987.

The skewed age structure of this population
might also have been due to the fact that these
condors were not breeding at their full potential.
We found no evidence of breeding pairs except for
one nest and several copulation attempts by anoth-
er pair of condors. Wallace and Temple (1988b)
suggested that condors with food stress do not
breed regularly. Mdiile food shortages were not im-
plicated with the California Condor ( Gymnogyps cal-
ifornianus) population decline (Johnson et al.
1983, Ogden 1985, Snyder and Snyder 1989), food
availability shaped the condors’ distribution pat-
terns (Wilbur 1977). In Peru, Temple and Wallace
(1983), found that 26% of released juvenile An-
dean Condors died of disease, starvation or un-
known causes. The effect of food stress on the pop-
ulation in our study area was unknown.

Some factors associated with the California Con-
dor decline have not yet been noted in Ecuador.
Ogden (1985) and Wallace (1989), for example,
reported tbat human disturbance around Califor-
nia Condor nests can be problematic. While evi-
dence of human impact on Andean Condors is lim-
ited, we found that one young fledged successfully
from a nest situated 100 m above a noisy construc-
tion site (INEFAN 1997). While this may have been
an exceptional occurrence, it suggests that human
disturbance to nesting condors may be tolerated.

High mortality rates or low reproductive rates
will lead to a population that can only be sustained
by immigration. Survival rates and reproductive
rates of the condor population in Ecuador are not

known at present. We do know, however, that im-
migration occurs because at least two adult con-
dors in the Cayambe-Coca Ecological Reserve were
tagged with patagial markers and released in Co-
lombia from captive reared stock (Lieberman et al.
1993).

It should be noted that our results may also be
artifacts of a skewed or small sample. Although we
assumed that both sexes and all age classes had
equal chances of being observed in the field, this
may not have been the case. Given its distinctive
head shape, for example, adult males may be easier
to identify than females, especially at a distance.
We were able to positively identify the sex of only
10% of our observations leaving us a relatively
small sample size. Behavioral differences between
adult and immature condors may explain also
some of the skewed age ratio. Daily activity patterns
among California Condors, for example, were un-
predictable (Wilbur 1980). Snyder and Johnson
(1985) noted that movement patterns among Cal-
ifornia Condors were divided among breeding ter-
ritories and foraging areas. Adults stayed close to
their respective breeding areas while all age groups
appeared to mix thoroughly in foraging areas (Sny-
der and Snyder 1989) . If Andean Condors had sim-
ilar behavior to California Condors and our obser-
vation sites were limited to breeding territories,
then we would expect to see few Juveniles. Al-
though we attempted to address this issue by con-
ducting studies at 19 sites, surveys at additional for-
aging areas may contribute additional results.

The current Andean Condor population of
northern Ecuador is small and critically imperiled
(Granizo et al. 1997). Further field studies are
needed to determine the causes of condor mortal-
ity, the effect of food shortages on the population
and to develop and implement a successful man-
agement plan. Ecuador’s human population is ex-
pected to double in 28 yr (Caberle et al. 1989) and
condor habitat will likely diminish greatly as peo-
ple move higher into the paramos where condors
live.

Acknowledgments

The project was coordinated with local assistance by
INEFAN (Instituto Ecuatoriano Forestal y de Areas Na-
turales y Vida Silvestre), CECIA (Consejo Ecuatoriano
por la Conservacion y Investigacion de las Aves), and
Peace Corps/Ecuador. Appreciation is extended to the
director of the Cotacachi-Cayapas Ecological Reserve, G
Rosales, and director of the Cayambe-Coca Ecological Re-
serve, L. Martinez, for their gracious support. We greatly

36

Koenen et al.

VoL. 34, No. 1

appreciate comments to an earlier draft of this paper

from J. Jackson, J. and E. Sipple, S. Wilbur andj. Wiley.

Funding for this study was provided by US AID and the

World Nature Association. The Manomet Observatory

Birder’s Exchange Program provided field equipment.

Literature Cited

Caberle, B.J., M. Crespi, C. Dodson, C. Luzuriaga, D,
Rosse and j. Shores. 1989. An assessment of biolog-
ical diversity and tropical forests for Ecuador. Center
for International Development and Environment of
World Resources Institute, Washington, DC U.S.A.

Granizo, T., M. Guerrero, C. Pacheco, R. Phillips, M.B.
Ribadeneira and L. Suarez. 1997. Lista de aves amen-
azadas de extincion en el Ecuador. UICN-Sur, CECIA,
INEFAN, EcoCiencia and Birdlife International. Qui-
to, Ecuador.

INEFAN. 1997. Los condores de Papallacta. INEFAN.
Quito, Ecuador.

Johnson, E.V., D.L. Aulman, D.A. Clendenen, G. Gulia-
si, L.M. Morton, P.I. Principe and G.M. Wegener.
1985. California Condor; activity patterns and age
composition in a foraging area. Am. Birds 37:941-945.

JossE, C.M. and j. Anhalzer. 1996. A guide for the pa-
ramos of the Ecuadorian national protected areas sys-
tem. Institute Ecuatoriano Forestal y de Areas Natur-
ales y Vida Silvestre — INEFAN. Quito, Ecuador.

Lieberman, a., J.V. Rodriquez, J.M. Paez and J. Wiley.
1993. The reintroduction of the Andean Condor into
Colombia, South America: 1989-1991. Oryx 27:83-90.

Luteyn, J.L. 1992. Paramos; why study them? Pages 1-14
in H. Balslev and J.L. Luteyn [Eds.], Paramo; an An-
dean ecosystem under human influence. Academic
Press, New York, NY U.S.A.

, A.M. Cleef and O. Rangel. 1992. Plant diversity

in paramo: towards a checklist of paramo plants and
a generic flora. Pages 71-84 in H. Balslev and J.L.
Luteyn [Eds.], Paramo: an Andean ecosystem under
human influence. Academic Press, New York, NY
U.S.A.

McGahan, J. 1972. The behavior and ecology of the An-
dean Condor. Ph.D. dissertation, Univ. Wisconsin,
Madison, WI U.S.A.

Murphy, R.C. 1936. Oceanic birds of South America.
American Museum of Natural History, New York, NY
U.S.A.

Ogden, J. 1985. The California Condor. Pages 388-399
in R.L. Di Silvestro [Ed.], Audubon wildlife report
1985. Natl. Audubon Soc., New York, NY U.S.A.

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

Snyder, N.F.R. and E.R. Johnson. 1985. Photographic
censusing of the 1982-1983 California Condor pop-
ulation. Condor 87:1-13.

AND H.A. Snyder. 1989. Biology and conservation

of the California Condor. Pages 175-267 in D.M. Pow-
er [Ed,], Condor biology and conservation. Current
Ornithol. 6. Plenum Press, New York, NY U.S.A.

Temple, S.A. and M.P. Wallace. 1983. A study of tech-
niques for releasing hand-reared Andean Condors to
the wild: final report on contract FWS 14-16-009-78-
923. Dept. Wildlife Ecology, Univ. Wisconsin-Madi-
son, Madison, WI U.S.A.

AND . 1989. Survivorship patterns in a pop-
ulation of Andean Condors (Vultur gryphus). Pages
247-251 in B-U. Meyburg and R.D. Chancellor [Eds.],
Raptors in the modern world. WWGBP, Berlin, Ger-
many.

U.S. Fish and Wildlife Service. 1986. Endangered and
threatened wildlife and plants. USDI, U.S. Fish and
Wildlife Service. Washington, DC U.S.A.

Wallace, M.P. 1989. Andean Condor experimental re-
leases to enhance California Condor recovery. Endan-
gered Species Update 6:1-4.

and S.A. Temple. 1988a. A comparison between

raptor and vulture hacking techniques. Pages 75-81
in D.K. Garcelon and G.W. Roemer [Eds.], Proceed-
ings of the international symposium on raptor rein-
troduction, 1985. Inst, for Wildlife Studies, Areata, CA
U.S.A.

AND . 1988b. Impacts of the 1982-1983 El

Nino on population dynamics of Andean Condors in
Peru. Biotropica 211:14:4— IbQ.

Wilbur, S. 1977. Supplemental feeding of California
Condors. Pages 134-140 in S.A. Temple [Ed.], Man-
agement techniques for preserving threatened spe-
cies. Univ. Wisconsin Press, Madison, WI U.S.A.

. 1980. Estimating the size and trend of the Cali-
fornia Condor population, 1965-1978. Calif. Fish and
Game 66:40-48.

Received 1 May 1999; accepted 5 October 1999

Short Communications

J. Raptor Res. 34(1):37-41
© 2000 The Raptor Research Foundation, Inc.

Summary of Philippine Eagle Reproductive Success, 1978-98

Hector C. Miranda, Jr.

Philippine Eagle Foundation, Garnet cor. Diamond Street, Marfori Heights, Davao City, 8000 Philippines and University
of the Philippines-Mindanao, Ladislawa Avenue, Buhangin, Davao City, 8000 Philippines

Dennis I. Salvador, Jayson C. Ibanez and Gliceria A. Balaquit-Ibanez
Philippine Eagle Foundation, Garnet cor. Diamond Street, Marfori Heights, Davao City, 8000 Philippines

Keywords; Philippine Eagle, Pithecophagajefferyi; breed-
ing success; population decline, forest fragmentation.

The Philippine Eagle {Pithecophagajefferyi) is one of the
rarest eagles in the world. Its present status has always
been regarded as critically endangered. Previous esti-
mates of the Philippine Eagle population have been spec-
ulative (Alvarez 1970, Gonzales 1971, Rabor 1971). The
most recent estimate suggests that the total population
consists of between 300-500 individuals (Kennedy 1977,
1985). Philippine Eagles lay a single egg and have a 2-yr
cycle between successive breedings when pairs breed suc-
cessfully, but in cases when breeding attempts fail, adults
breed the following year. Since work began in earnest on
this species, a large amount of information on nesting
successes and failures in Mindanao have been amassed.
Here, we report this information based on records col-
lected over the past 20 yr, and provide insights as to the
key reason for the decline of the population.

Methods

We compiled all existing information on the reproduc-
tive success of Philippine Eagles based on published
(Kennedy 1985) and unpublished documents gathered
by the Philippine Eagle Foundation (PEF) between
1978-98. We defined a successful breeding attempt as
those with young eagles that survived until fledging.
Nests were located by daily surveillance from vantage
points, usually along mountain ridges and in areas where
eagle presence was reported by local settlers. Observers
stayed in these areas for about a week during the breed-
ing season between August-December and from 0600-
1500 H. Blinds were built in trees adjacent to nest trees,
usually about 50-100 m away. Life history information
was obtained and daily activities were recorded. A reward
system for reporting occupied nests was initiated in 1981.
From 1985 to the present time, the reward system was
intensified and coupled with other on-site programs such
as the development of community-based initiatives and
conservation education activities. Reports of sightings
were improved further by forging partnership arrange-
ments with broadcasting stations in Mindanao Island.

Results and Discussion

Prior to 1970, only one nesting pair of Philippine Ea-
gles was located (Gonzales 1971). From 1978-83, several
nesting pairs were intensively studied within the logging
concession of the Paper Industries Corporation of the
Philippines (PICOP) in Surigao del Sur and Davao Ori-
ental provinces, and within the Mount Apo National Park
(Kennedy 1981, 1985). This was a period when intensive
logging operations occurred on Mindanao Island and
many nesting areas were logged or altered by slash-and-
burn farmers. Eight breeding attempts by 6 pairs failed
(72.7%) out of a total of 11 attempts during this period
(Fig. 1). One nestling was retrieved from a nest at Mount
Apo National Park and is currently being kept at the Phil-
ippine Eagle Center in Davao City.

From 1984—88, the PEF and the Department of Envi-
ronment and Natural Resources (DENR) monitored the
breeding population. Of the eight pairs monitored, there
were 11 breeding attempts. Four failed (36.4%) and four
young (36.4% nesting success) were produced. This rep-
resented an 18.2% increase in fledging success compared
to the previous period.

From 1989-93, 11 breeding attempts by nine pairs re-
sulted in eight fledglings (72.7% nesting success) and,
from 1994—98, 17 breeding attempts by 12 pairs resulted
in a higher success rate (88.2%). The increase in breed-
ing pairs was mainly due to an increased awareness by
local people and increased observer effort and was not
indicative of the recovery of the population. Other strat-
egies such as the reward system and media-based infor-
mation campaign have also been widely used by the PEF
since the early 1990s to increase the information on the
number of breeding pairs in the population. The in-
crease in breeding pairs during the last decade may also
have been due to increasing fragmentation of lowland
dipterocarp rainforest that result in increased contact
with settlers.

Breeding success based on eight pairs with >1 nesting
attempt was estimated at 0.38 ± 0.14 (±SD) young/pair/

37

Table 1. Summary of Philippine Eagle nesting records from 1978-98. All information taken from Philippine Eagle Foundation records unless otherwise
stated.

38

Short Communications

VoL. 34, No. 1

c/5

H

2

W

S

S

o

u

0

2

3

c/5

c/5

w

u

U

D

05

z

o

l-H

U

o

a

i-i

o

-w

cn

C<5

<u

T3

1/

"p

2

w

a

■p

1/)

c^r

15

tS

"P

L<

Pj

15

&,

t3

• ^

flj

15

t3

a

PS

.2

i/i

-w

ts

-P

2

cyi

2

p

2

bo

be

15

'■B

P

• •pM

be

be

be

P

bo

P

P

P

’B

Cu

P

C«

15

z

P

P

P

zn

15

Z

P

T3

U

C

u

3

Si

V

<u

c/5

1 )

c

o

be

be

T3

D

c a

o o

■p 'p

W5 C/T

*-G

rt rt

u u

<j

q=l S

• *4 • PM

u u
0 0

tG

U U

‘C

15 15

> >

s s

>

Lh L>
0 0

p p

u

0

<P <P

15 15

"P P!

> >
15 15

TJ

15 15

V

P P

* pH ••pH

'B h

v: cT)

15 15

’r)

S-, U

15 15

•W •4-1

15

L> U

15 15

S_

o o

be be
n ^

T3
15
k.1
rt

Cl, ^

a Ai

cfl

'/5

T3
t 3 rt

be be
o bo be

;z; w w

- ^

a

r/h

c/5

15

u

u

p

tC

o

T5

1>

T5

15

T3

D

05 05 fiH fiH |Jh li(

"P 'd "P

G

o

p

p p

15 "O ’P

U

P

■a «

P

P

Ztl

c/5

15

c/5

15

cH

c/5

C/5

15

5

zn

zn

15

o

S3

P

zn

zn

15

l3

(4H

(Z)

C«

i3

IZ»

eft

cH

Cft

Cft

V

1

Cft

(L)

c2

Cft

cft

H

15

15

15

U

o

15

O

U

U

U

Gh

rG

U

U

15

15

G

1 ^

15

u

U

U

U

u

P

P

p

P

P

G

G

G

G

G

03

03

03

03

0

03

c/:)

(/5

C/!)

C/D

7!)

be

P

be

P

be

be

bc

P

bC

bc

P

be

bC

be

P

• pH

•S bc

be

P

♦ pH

G

« pH

P

* pH

p

• pH

w

be

be

15

e

be

15

be

be

be

P

G

be

p

» pH

P

P

P

P

P

15

15 P

G

p

15

•til

15

15

15

U

1/5

P 15

'Z3

Cft

* pH

be

zn

0

Cft

QJ

p

zn

0

P

1

P

1

P

03

biD

4~5

zn

^ P

<U

pH

be

P

bo

P.

G

Cft

Oh

Cft

•M

Cft

-JLi

Cft

0

2

w

c/3

G

» >»H

0

a,

a 2

P

O

G

• pH

15

P

G

• pH

0

2

bo

bo bc

he

0

Cl

2

0

& be

0

P be

0

Cl

H

PQ

cft

0

0 p

i-‘

Cft

P

zn

U

Cft

P

a

_P

c

G

G

P P

• pH -pH

_P

0

13

0 .a

0 .3

0

ce

0

V

w

0

P.

V

0

V

O

0

‘ho

0

p

0

bo ho

'3d

•IH

0

be

be

G

be

bo -p

S bo

C

bo

Cl

G

ho ‘p

‘p

be ’P

'•w

'•HJ

P P

p

be

be

• pH

bC Pi

be P

be

z

Q

p

p rt

t .a

Lp

o

P

Q

n

P

rS

_P

15

Ct

15 15

15

_p

G

P ^

p ^

_p

<4-J

«

P S

» pH

• pH

p^H

S

jo

s

P)

P

P P

P

• pH

P 'p

p

bo Cft Cft .-r

w Z Z 5^

G Cft
0 V

O z

be

be

W

Cft

15

z

bo

W

be

be

be

be

W

•W

Cft

15

z

P

o

P

l-H

P

15

P

l-H

•w

Cft

OJ

Z

P

zn

0

Ph

p

15

p

HH

P

15

P

HH

P

u

p

HH

Ji) 4-)

Cft Cft

0 c

Ph Gh

P

zn

0

P

p

c/5

15

z

Cft

15

z

G

u

G

HH

z fS

z f£

zn

15

z

Q

00

05

05

05

(M

p

0

%

I>

J>

00

00

00

00

00

05

05

oi

w

J>

oo"

00 "

00 "

rH

CO

00

00

00 '

oo'"

p

in'"

p

p

p

p

t£

00

00

05 ''

0

o'

CM

CM

(M

CM

CM

c/5

00

00

00

00

00

00

00

00

00

00

00

00

00

00

00

00

00

00

00

05

05

05

05

05

05

05

CQ

05

05

05

05

eji

05

05

05

05

05

05

05

05

05

05

05

05

05

05

05

05

05

05

05

05

05

05

05

05

05

0

1 — 1

I — (

T— 1

1-H

f-H

l-H

1-H

t-H

t-H

t-H

t-H

t-H

t-H

1-H

l-H

T— H

rH

l-H

l-H

^H

1-H

l-H

^H

t-H

^H

t-H

l-H

OOP

.2

^ ' p
^3 u

Table 1. Continued.

March 2000

Short Communications

39

c/5

H

Z

W

S

S

0

U

:z

o

HH

g

o

<u

a

u

rt

u

pO

B T3

■e

Sh

•p-p

O

c

<u

k.

t3

dJ

PI

Sm

S

X!

§ ti
Z ^

w

O

T3

be

he

W

T3

(U

V

W

P P P

P P P

<.w <4-( 4-1

c/) [/)

V3 y) yi

U

O

3

u u
u u
3 3

u

o

3

u u
w o

3 3

Jj

rS

<5

2

a

^2

V3

VI

V3

v:

v:

V3

»3

«3

V3

yi

V]

yj

v:

v:

V!

»3

v:

<U

V

U

<U

(U

<U

dJ

dJ

<LI

u

u

<J

u

U

4)

CJ

Cl

<J

u

u

<J

u

o

U

u

o

<J

3

3

3

3

3

3

3

3

3

1/1

1/)

1/1

/I

/I

CO

CO

CO

CO

ho

ho

be

be

be

_S

_S

_S

3

_S

bc

3

'he

'ho

'ho

• pM

be

be

‘to

be

C

t

0

t 3

P

be

(H

P

G

• pM

P

C

• p4

u

be

•a

dJ

V

<u

V

be

dj

be

t 3

re

G

Cp

C

be

t 3

u

(U

pC

X

X

3

X

•u

3

X

(U

a

'v

3

CJ

V3

0

V3

0

V3

o

VI

(U

V

0

■ p 4

he

V

O

'v

iTl

w

cn

3

Gp

Oh

a

3

Gp

t 3

Gp

X

V

o

'z

Bi

bi

c/3

0

Cl-

be

be

bo

• pN

0

0

.w

0

2

0

pj

he

2

0

SIX

be

dJ

'5? be

O

p-i

o

Gp

p

BQ

o

_3

_3

_3

4.J

3

3

3

3

3

• PN

3

Q

3

» W

X re

3

0

ce

0

‘be

‘be

‘be

o<

0

0

0

0

ho

O

w

be

o he

O

PpH

bl

be

be he

T3

'3

T3

• tW

•a

•a

•a

•a

■3

•a

be

p

Gp "3

*a

bo

z

3

3 3

di

dj

<u

2

re

re

re

re

(U

re

c

<u

o .2

re

3

« pM ■ pM

'P

XI

pO

P

pO

pQ

• p^

2 53

pG

w

yi

OJ

Z

1^3 V3

dj dj

z z

Vi

0

pL,

X

V3

0

Ph

X

V)

0

pin

3

0

u

3

CJ

3

HH

3

u

3

1— 1

3

u

3

1— (

3

<J

3

hH

X

o

ex

3

u

3

H4

2

X

V3

o

ex

be V
be 0
w ex

3

u

3

h-H

V

Z

be

3

he

X

yi

O

0.

pfi

3

u

3

o

III

CO

CO

ca

00

00

os

05

03

OS

(03

(M

eo

so

TtH

KS

VO

so

S3

VO

S3

S3

dO

(3

t^

i>

t>

00

t>

00

00

oe

<y>

3>

Cb

O)

CD

OS

OS

OS

OS

OS

OS

OS

OS

03

03

03

03

03

03

03

oa

os

Cb

Cb

OS

OS

OS

Os

OS

OS

OS

OS

OS

OS

OS

03

03

03

03

d?3

03

0

rH

r-H

1 — 1

i-H

pH

• pM

u

O

2

Ki

-D

iS

0

U

3 pd
- 3 S

0 3?^
*j- PQ

!J

• ^

(U

.2^3
Q ^

^

^ - rt

s s i

‘-D
S di

^ re p^
^

pS X5

— ' re -

Q

<j:>

Sh

3

1/1

'll

Tl

VI

3

Bt'

u

O

re

Ph

be

• pM

*2

ct

<

u

o

be

re

G

ex

u

re

Q

J

C/D

• p 4

dj

Ch

C

be

.0

s

3

H

® Entry for location generally proceeds as a combination of the locality (barangay, mountain range or a road marker [e.g., Rd. 6]) and the municipality (or district), followed
by the province (or city) .

‘’Kennedy (1985).

40

Short Communications

VoL. 34, No. 1

16 n

197B-83 1964-88 1989-93 1994-98

(6 pairs) (8 pairs) (9 pairs) (12 pairs)

■ Successful
B Failure
Q Unknown

Relative number of successful vs. failed nesting attempts

Figure 1 . Summary of the success and failure of Phil-
ippine Eagle breeding attempts based on 7-yr intervals.

yr and nesting success averaged 76.3% (Table 2). This
was a conservative estimate since we did not take into
account that eagles may have nested in following years
after previous breeding attempts failed, instead of their
typical 2-yr cycle. Many pairs had only one nesting record
and these were excluded in calculating percent breeding
success to minimize bias. One such pair at Rd. 6P PICOP,
Surigao del Sur was documented to have bred three
times during which one nesting attempt failed and the
fate of the other two attempts were unknown. Some
pairs, especially those within the Bukidnon province, had
100% breeding success rates while others like the pair at
Mount Apo, Toril, Davao City had a 33.3% success rate
and a productivity of 0.17 young/pair/yr. These differ-
ences may have been due to variation in food supply be-
tween the areas, differences in the ages of the breeding
birds (Newton 1979), or simply an artifact of the small
sample size. The overall success of Philippine Eagles av-
eraged about 58.0% for 50 breeding attempts by 29 pairs
from 1978-98. Based on the assumption that each breed-
ing attempt had equal probability of success or failure,
and that no regional differences existed among different
pairs or subpopulations, we considered this productivity
to be high and not indicative of a population suffering
from breeding failures.

Precise assessment of the causes of breeding failure is
difficult. Birds exposed to food shortages and distur-
bances during critical periods of the nesting cycle may
abandon eggs or nests (Newton 1979). Our summary of
causes of nesting failures (Table 1) was not complete be-
cause field methods varied over the years. Moreover, our
results showed that many of the breeding pairs were dis-
turbed by logging operations, slash-and-burn farming
and by the observers themselves. Three of 15 failures
(20.0%) were due to removal of young from nests or fell-
ing of nest trees with young. Most individuals currently
kept at the Philippine Eagle Center in Malagos, Davao
City were either confiscated or surrendered as juveniles.
There were also three cases (23.5%) wherein eggs were
addled and/or abandoned, but the causes of nest aban-
donment were unknown.

The information we obtained may also have been
based on the most conspicuous or accessible breeding
pairs and, therefore, it may not be indicative of the true
productivity of the population of Philippine Eagles. Some
Philippine Eagle pairs may be more experienced breed-
ers and may also be overrepresented in our sample which
could account for the high reproductive success we re-
corded. Also, the high breeding success may also reflect
the diminishing persecution of Philippine Eagles hy the
local people. Despite the limitations of the data we col-
lected, we believe that it provides important baseline in-
formation to help focus future research and conservation
efforts on the Philippine Eagle.

The current status of the Philippine Eagle as Critically
Endangered is based mainly on the fact that this is a
large-sized bird requiring a large territory and adapted
to a tropical rainforest ecosystem that is fast disappearing
in the Philippine archipelago. Theoretically, the assess-
ment of raptor population stability involves integration
of reproductive data with survival data for various age
classes (Henny et al. 1970), but the lack of information
on survival of Philippine Eagles after fledging limits the
precise assessment of their population status. Although it
is clear that the major threat to tropical birds of prey is
forest destruction (Thiollay 1985, 1989, 1992), it was un-

Table 2. Breeding rates of Philippine Eagle pairs with more than one recorded nesting attempt.

Breeding Pair Location

No. Breeding
Attempts

% Success

Breeding Rate
(Y ouNG/PAiR/yr)

Dalwangan, Malaybalay, Bukidnon

5

60

0.30

Minlanga Range, La Paz, Agusan del Sur

2

100

0.50

Freedom, Cabanglasan, Bukidnon

2

100

0.50

Guilang-guilang, Manolo Fortich, Bukidnon

3

100

0.50

Mount Apo, Toril, Davao City

6

33.3

0.17

Amabel, Magpet, North Cotabato

2

50

0.25

Mount Sinaka, Arakan Valley, North Cotabato

2

100

0.50

Salaysay, Marilog District, Davao City

6

66.7

0.33

Mean

76.3

0.38

March 2000

Short Communications

41

clear whether the population decline of the Philippine
Eagle is mainly due to reproductive failures or to in-
creased mortality of juveniles, subadults and/or adults. A
high rate of nesting failures could explain the population
decline because Philippine Eagles lay a single egg and
normally breed once every two years. Our data indicate
that it is not nesting failures that are responsible for the
population decline but that decreased survival of juve-
niles and subadults and their inability to disperse be-
tween forest islands to establish eventual breeding terri-
tories may be limiting the number of breeding pairs in
this population. Past studies have suggested that the sta-
bility of breeding raptor populations is not related to pro-
longed good production of young but could be main-
tained by immigration or dispersal (Mebs 1964, Ratcliffe
1972, Newton 1979, Grier 1980). Nevertheless, forest
fragmentation has untold effects on large tropical forest
raptors such as the Philippine Eagle. Future research
should focus on aspects of metapopulation dynamics
such as survival and dispersal studies in a highly frag-
mented habitat, continued monitoring of reproductive
performance of known breeding pairs in Mindanao and
initiation of basic population ecology studies in other is-
lands of the archipelago where Philippine Eagles are his-
torically known.

Resumen. — El exito reproductivo total del aguila de las
Filipinas {Pithecophaga jefferyi) promedio 58.0% de los in-
tentos reproductivos por 29 parejas desde 1978-98. El
exito reproductivo con base en ocho parejas con mas de
un intento reproductivo fue estimado en 0.38 ± 0.14
(±SD) juvenil/pareja/aho y el exito reproductivo pro-
medio 76.3%. Hubo 15 fracasos reproductivos, tres de los
cuales se debieron a la remocion de juveniles del nido o
caidas del nido, y en tres casos los huevos fueron infer-
tiles o abandonados. Nuestro analisis sugiere que los fra-
casos reproductivos no son un factor mayor en la decli-
nacion poblacional del aguila de las Filipinas y apunta a
un decrecimiento de la sobrevivencia de los juveniles y
subadultos y su inhabilidad para dispersarse entre los
fragmentos de bosque como la causa de la declinacion
numerica de esta especie.

[Traduccion de Cesar Marquez]

Acknowledgments

We would like to thank R.S. Kennedy and the members
of F.R.E.E. Ltd. whose pioneering work laid the founda-
tion for the conservation research of this magnificent
raptor. We are grateful to the other members of the Field
Research component of the Philippine Eagle Founda-
tion; Donald S. Afan, Glenn Lovell, L. Bueser, Kharina
M. Gatil, Adriano M. Oxales and the rest of the staff. We
thank the Peregrine Fund, John D. and Catherine T.
MacArthur Foundation, Oxbow Power Corporation, Mar-

co Polo-Davao, Standard Chartered Bank, Cincinnati Mu-
seum Center, Regional Eagle Watch Teams, Parks and
Wildlife Bureau of the Department of Environment and
Natural Resources of the Republic of the Philippines,
and Philippine Eagle Working Group for their unwaver-
ing support. We are grateful to W.A. Burnham and J.-M.
Thiollay for their valuable comments on the manuscript.
We also thank the long-time supporters of the Philippine
Eagle Foundation, R.S. Kennedy, W.A. Burnham and J.W.
Grier. Lastly, we thank the many participants and sup-
porters who over the years have contributed in one way
or another to the conservation efforts for the Philippine
Eagle.

Literature Cited

Alvarez, J.B., Jr. 1970. A report on the 1969 status of the
monkey-eating eagle of the Philippines. Pages 68-73
in Papers and Proc. Int. Union Conserv. Nat. Resour.,
Vol. 11. 11th Technical Meeting, New Delhi, India.
Gonzales, R.B. 1971. Report on the 1969 status of the
monkey-eating eagle on Mindanao Island, Philip-
pines. Bull. Int. Counc. Bird Preser. 11:154—168.

Grier, J.W. 1980. Modeling approaches to Bald Eagle
population dynamics. Wildl. Soc. Bull. 8:316-322.
Henny, C.J., W.S. Overton and H.M. Wright. 1970. De-
termining parameters for populations by using struc-
tural models. J. Wildl. Manage. 34:690—793.

Kennedy, R.S. 1977. Notes on the biology and population
status of the monkey-eating eagle of the Philippines.
Wilson Bull. 89:1-20.

. 1981. Saving the Philippine Eagle. Natl. Geogr.

159:847-856.

. 1985. Conservation research of the Philippine

Eagle. Natl. Geogr. Soc. Res. Rep. 18:401-414.

Mebs, T. 1964. Zur biologie und populationsdynamik des
mausebussard {Buteo buteo).J Omithol. 105:247-306.
Newton, I. 1979. Population ecology of raptors. T. & A.D.
Poyser, London, U.K.

Rabor, D.S. 1971. The present status of conservation of
the monkey-eating eagle of the Philippines. Philipp
Geogr. J. 15:90-103.

Ratcliffe, D.A. 1972. The peregrine population of Great
Britain in 1971. Bird Study 19:117-156.

Thiollay, J.-M. 1985. Falconiformes of tropical rainfor-
est: a review. Pages 155—165 in I. Newton and R.D.
Chancellor [Eds.], Conservation studies on raptors.
ICBP, Cambridge, U.K.

. 1989. Area requirements for the conservation of

rainforest of rainforest raptors and game birds m
French Guiana. Conserv. Biol. 3:128-137.

. 1992. Influence of selective logging on bird spe-
cies diversity in a Guianan rainforest. Conserv. Biol. 6:
47-63.

Received 5 December 1998; accepted 28 October 1999

J. Raptor Res. 34(1) :42-44
© 2000 The Raptor Research Foundation, Inc.

Diet of Autumn Migrating Northern Saw-whet Owls on the Eastern Shore

OF Virginia

David M. Whalen^ and Bryan D. Watts

Center for Conservation Biology, College of William and Mary, Williamsburg, VA 23187 U.S.A.

David W. Johnston

3219 Concordia Street, Fairfax, VA 22032 U.S.A.

Key Words: Northern Saw-whet Owl; Aegolius acadicus;

diet, migration; pellet analysis; stomach-content analysis.

Although numerous studies have been conducted on
the diet of Northern Saw-whet Owls {Aegolius acadicus) ,
these accounts are almost exclusively limited to the
spring breeding season (e.g., Cannings 1987, 1993, Marks
and Doremus 1988) or the winter (e.g.. Grove 1985,
Swengel and Swengel 1992, Holt and Leroux 1996). It is
well-known that many saw-whet owls make annual migra-
tory movements to lower latitudes in the eastern and cen-
tral U.S. (Mueller and Berger 1967, Holroyd and Woods
1975, Weir et al. 1980, Duffy and Kerlinger 1992, Brinker
et al. 1997). However, little is known about saw-whet owl
diets during autumn migration. The objectives of our
study were to determine the diet of autumn migrating
saw-whet owls, compare these results with data available
from previous diet studies, and compare results obtained
from pellet versus stomach-content analyses.

Study Area and Methods

Since 1994, migrating saw-whet owls have been trapped
and banded each autumn just south of the city of Cape
Charles (37°10'N, 75°50'W) on the eastern shore of Vir-
ginia (Whalen et al. 1997). Banding operations com-
menced during late October and ended in mid-Decem-
ber each year. Captured owls were kept in holding boxes
before being banded and released. Although individuals
were only kept in holding boxes for a short period of
time, owls occasionally regurgitated pellets while in the
boxes. From 1995-97, 1236 individual saw-whet owls were
captured, from which a total of 53 pellets was obtained.
These pellets were later dissected and prey remains were
identified to the most precise taxonomic group possible.
In addition to prey data collected from pellets, stomach
contents were determined from 15 road-killed saw-whet
owls found within 24 km of Cape Charles, mostly on U.S.
Route 13. All road-kills were collected during autumn or
early winter.

Results and Discussion

We identified 89 prey items from 53 saw-whet owl pel-
lets (Table 1). Based on pellet analysis, lepidopterans

^ Present address: Department of Biology, University of
New Mexico, Albuquerque, NM 87131 U.S.A.

comprised 42.7% of all individual prey items taken by
saw-whet owls. However, 78.9% of these insects were
found in only two owl pellets containing 20 and 10 adult
lepidopterans, respectively. Only six out of 53 pellets
(11.3%) were found to contain insect remains and most
of these also contained vertebrate prey. Thus, based on
pellets alone, only a small fraction of saw-whet owls
showed evidence of consuming insect prey.

Vertebrate prey remains were identified in 51 out of
53 pellets. Rodents and shrews comprised 96.1% of ver-
tebrate prey. The most common prey species were white-
footed mice (Peromyscus leucopus) and house mice {Mus
musculus). Other prey identified to species included
short-tailed shrews (Blarina breuicauda) , a southern flying
squirrel {Glaucomys volans), a silver-haired bat {Lasionyc-
teris noctivagans) and a Vesper Sparrow {Pooecetes grami-
neus).

Out of 15 road-killed saw-whet owl stomachs analyzed,
eight were empty and seven contained prey remains. In
contrast to diet evidence from pellet analysis, five out of
seven stomachs with prey remains contained insects. Ten
individual prey items were identified from stomach con-
tents, eight of which were insects and two of which were
rodents (Table 1).

Although rodents probably comprise most of the prey
biomass consumed by saw-whet owls migrating through
the lower Delmarva Peninsula, the number of insect prey
items is notable. More than 40% of prey items identified
in pellets were lepidopterans. This stands in sharp con-
trast to findings from numerous breeding and winter diet
studies (but see Hobson and Sealy 1991). Cannings
(1993) summarized prey items taken by saw-whet owls in
eastern North America, reporting that insects represent-
ed less than 1 % of all prey items. More insects were iden-
tified from 53 pellets in our study than from several thou-
sand pellets among all studies included in Cannings’
account.

Our results from stomach analysis of road-killed owls
provides even greater evidence that insects comprise an
important part of the diet of migrating saw-whet owls.
Since softer, more easily digested prey items such as in-
vertebrates are often under represented in pellets, stud-
ies that rely exclusively on pellet analysis may poorly re-
flect actual diet composition. Although our sample size
is small, 80% of prey items from stomach contents were

42

March 2000

Short Communications

43

Table 1. Prey remains of Northern Saw-whet Owls migrating through the eastern shore of Virginia during autumn
1995-97. Data are presented for prey items obtained from pellets {N = 53) regurgitated by captured owls and from
stomach contents of road-killed owls {N = 15).

Pellet Remains

Stomach Contents

Prey Type

No. Prey
Items“

Percent

No. Prey
Items^^

Percent

Mammals
Short-tailed shrew
{Blarina brevicauda)

3

3.4

1

10.0

White-footed mouse
{Peromyscus leucopus)

11

12.4

0

0.0

House mouse
{Mus musculus)

7

7.9

1

10.0

Southern flying squirrel
{Glaucomys volans)

1

1.1

0

0.0

Unidentified rodent'^

27

30.3

0

0.0

Silver-haired bat
{Lasionycteris noctivagans)

1

1.1

0

0.0

Birds

Vesper Sparrow
{Pooecetes gramineus)

1

1.1

0

0.0

Insects

Lepidoptera spp.

38

42.7

6

60.0

Homoptera spp.

0

0.0

1

10.0

Orthoptera spp.

0

0.0

1

10.0

Total

10

® Some individual pellets and stomachs contained multiple prey items.

Most unidentified rodent remains consisted of post-cranial bones and bone fragments.

insects and one-third of road-killed owls contained in-
sects in their stomachs. This finding was comparable to
that of Hobson and Sealy (1991), who found substantial
quantities of invertebrate prey remains in the stomachs
of resident saw-whet owls on the Queen Charlotte Islands
during the nonbreeding season. In their study, the ma-
jority of stomachs of road-killed owls contained marine
invertebrate prey. Thus, small mammals probably repre-
sent the chief prey type for saw-whet owls throughout
most of the year. However, invertebrate prey items may
be taken opportunistically, especially during times when
vertebrate prey abundance is low or unpredictable.

In recent years, the number of saw-whet owls trapped
during migration in the Northeast has varied dramatically
(Brinker et al. 1997, Whalen et al. 1997). The magnitude
and dynamics of migratory movements by this species
may be closely linked to the availability of prey relative
to regional saw-whet owl population sizes. To the best of
our knowledge, our study represents the first published
account of the food habits of saw-whet owls during au-
tumn migration. More information is needed on the diet
composition of saw-whet owls, especially during summer
and fall. In addition, information on small mammal and
insect abundances along known saw-whet owl migration

routes could help to clarify the relationship between prey
levels and the migratory behavior of saw-whet owls.

Resumen. — Los habitos alimenticios de Aegolius acadius
durante su migracion de otono en la costa este de Vir-
ginia fueron documentados utilizando un analisis de
egragropilas y contenidos estomacales. La mayoria de las
presas (96.1%) encontradas en las egagropilas fueron
roedores, especialmente Peromyscus leucopus y Mus mus-
culus. Una larga proporcion de insectos lepidopteros
(42.7%) fueron tambien encontrados, pero en menor
numero de egragopilas. Otros tipos de presas en las ega-
gropilas incluyeron Blarina brevicauda, Glaucomys volans,
Lasionycteris noctivagans y Poocetes gramineus. Los estoma-
gos de 7 de 15 buhos muertos en las carreteras contenian
comida. En contraste a los resultados de los analisis de
egagropilas, 80 de los items de presas identificadas de los
estomagos fueron insectos. For lo tanto el analisis de ega-
gropilas y contenidos estomacales pueden conducir hacia
diferentes conclusiones acerca de la dieta de los buhos.
En resumen estos resultados sugieren que los insectos
juegan un papel importante en la dieta de la migracion
de Aegolius acadicus aunque los roedores representan la

44

Short Communications

VoL. 34, No. 1

mayoria de la biomasa de presas consumidas por esta
especie.

[Traduccion de Cesar Marquez]

Acknowledgments

This manuscript benefited from the comments of R.
Cannings, K. Duffy and J. Marks. We thank C. Handley,
Jr. for assistance with small mammal identification, R.
Laybourne for identifying the bird remains and A. Solis
for identification of insects. We also thank D. Bradshaw,
S. Flickinger, D. Schwab, B. Sullivan, G. Williamson, M.
Wilson and J. Wood for collecting road-killed saw-whet
owls. Funding for the owl banding project was provided
by the Center for Conservation Biology at the College of
William and Mary.

Literature Cited

Brinker, D.F., K.E. Duffy, D.M. Whalen, B.D. Watts and
K.M. Dodge. 1997. Autumn migration of Northern
Saw-whet Owls {Aegolius acadicus) in the Middle Atlan-
tic and Northeastern United States: what observations
from 1995 suggest. Pages 74-89 mJ.R. Duncan, D.H.
Johnson and T.H. Nicholls [Eds.], Biology and con-
servation of owls of the Northern Hemisphere: Sec-
ond International symposium. Gen. Tech. Rep. NC-
190. USDA, For. Ser., North Central Research Station,
St. Paul, MN U.S.A.

Cannings, RJ. 1987. The breeding biology of Northern
Saw-whet Owls in southern British Columbia. Pages
193-198 in R.W. Nero, R.J. Clark, R.J. Knapton and
R.H. Hamre [Eds.], Biology and conservation of
northern forest owls. Gen. Tech. Rep. RM-142. USDA,
For. Ser., Rocky Mountain Forest and Range Experi-
ment Station, Fort Collins, CO U.S.A.

. 1993. Northern Saw-whet Owl {Aegolius acadicus).

In A. Poole and F. Gill [Eds.], The birds of North

America, No. 42. Acad. Nat. Sci., Philadelphia, PA,
and Am. Ornithol. Union, Washington, DC U.S.A.

Duffy, K. and P.A. Kerlinger. 1992. Autumn owl migra-
tion at Cape May Point, New Jersey. Wilson Bull. 104:
312-320.

Grove, R.A. 1985. Northern Saw-whet Owl winter food
and roosting habits in north-central Washington. Mur-
relet 66:21-24.

Hobson, K.A. and S.G. Sealy. 1991. Marine protein con-
tributions to the diet of Northern Saw-whet Owls on
the Queen Charlotte Islands: a stable-isotope ap-
proach. Auk 108:437-440.

Holroyd, G.L. and J.G. Woods. 1975. Migration of the
saw-whet owl in Eastern North America. Bird-Banding
46:101-105.

Holt, D.W. and L.A. Leroux. 1996. Diet of Northern
Pygmy-Owls and Northern Saw-whet Owls in west-cen-
tral Montana. Wilson Bull. 108:123-128.

Marks, J.S. and J.H. Doremus. 1988. Breeding-season
diet of Northern Saw-whet Owls in southwestern Ida-
ho. Wilson Bull. 100:690-694.

Mueller, H.C. and D.D. Berger. 1967. Observations on
migrating saw-whet owls. Bird-Banding 28:120-125.

Swengel, S.R. and A.B. Swengel. 1992. Diet of Northern
Saw-whet Owls in southern Wisconsin. Condor 0A:I0I-
711.

Weir, R.D., F. Cooke, M.H. Edwards and R.B. Stewart.
1980. Fall migration of saw-whet owls at Prince Ed-
ward Point, Ontario. Wilson Bull. 92:475-488.

Whalen, D.M., B.D. Watts, M.D. Wilson and D.S. Brad-
shaw. 1997. Magnitude and timing of the fall migra-
tion of Northern Saw-whet Owls through the Eastern
Shore of Virginia, 1994—1996. Raven 08:01-104.

Received 1 April 1999; accepted 31 October 1999

/. Raptor i?^5. 34 ( 1 ) :45-48
© 2000 The Raptor Research Foundation, Inc.

Breeding Biology of the Eurasian Kestrel in the Steppes of Southwestern Spain

J. M, Aviles and J. M. SAnchez

Conservation Research Group, Department of Zoology, University of Extremadura, Badajoz E-06071, Spain

A. Sanchez

Eorestry Agency of Extremadura, Merida, Spain

Key Words; Eurasian Kestrel', Falco tinnunculus; laying
date, clutch size, reproductive rates', steppe habitat, Spain.

The Eurasian Kestrel {Falco tinnunculus) breeds in all
countries of Europe. Its recent decline in the Palearctic
has been linked to agricultural intensification (Tucker
and Heath 1994), persecution, pesticides and climate
change (Shrubb 1993). Several studies have shovm large
annual fluctuations in numbers of breeding pairs due to
weather effects (Cramp and Simmons 1980, Kostrzewa
and Kostrzewa 1990, Kostrzewa and Kostrzewa 1991) and
rodent availability (Cave 1968, Cramp and Simmons
1980, Dijkstra et al. 1982, Bonin and Strenna 1986, Vil-
lage 1990, Korpimaki and Wiehn 1998).

While population declines have been documented for
the Palearctic, Tucker and Heath (1994) estimated that
the breeding population of Eurasian Kestrels in Spain
was relatively stable at 25 000 and 30 000 pairs. Aparicio
(1997) postulated that the breeding population has been
stable since the 1970s. Studies of the Eurasian Kestrel in
Iberia have focused on the species’ diet (Garzon 1974,
Veiga 1982), nesding growth (Veiga 1985) and other eco-
logical topics (Aparicio 1994a, 1994b, 1998). Data on the
productivity of the breeding population in Spain are very
scarce. The only published account to date is by Gil-Del-
gado et al. (1995), who reported on production in the
Castellon province (eastern Spain).

Here, we report on the productivity of Eurasian Kes-
trels nesting in the steppe habitat of southern Iberia,
Spain in 1989—90. We compare our results with data from
other populations throughout the species’ breeding
range in the western Palearctic.

Study Area and Methods

We studied the breeding population of Eurasian Kes-
trels at La Serena, Extremadura, in southwestern Spain
(38°50'N, 5°20'W). This area is characterized by dry pas-
tures (64.2%) and cereal crops (28.4%; mainly wheat,
barley and oats). There are also small areas of scrub
(3.1%; mainly Retama sphaerocarpa) and areas with Holm-
oaks (1.1%; Quercus rotundifolia) as well as fruit trees
(mainly almond [Prunus dulcis]; Sanchez and Sanchez
1991). The study area has a mesomediterranean climate
(Rivas-Martinez 1987). More information on the study
area and its climate are reported by Aviles and Sanchez
(1998).

In 1986, the Forestry Agency of Extremadura began a
conservation project in the area that consisted of placing

900 wooden nest boxes in steppe habitat on electric pow-
er line stanchions (Sanchez and Sanchez 1991, Sanchez
et al. 1996). In 1989 and 1990, 76 nest boxes were mon-
itored weekly from the first stages of breeding. Visits were
increased to 3— 4-d intervals during the nestling period to
more accurately determine factors influencing final
breeding success. Laying date for each nest was deter-
mined by subtracting the incubation period of the spe-
cies (28 d; Cramp and Simmons 1980) from the hatching
date. Hatching date was determined by experienced ob-
servers, who took into account that all eggs hatch in 4 d
(Cramp and Simmons 1980). We measured percent
hatching success as the percentage of eggs within each
clutch that hatched, percent nestling mortality as the per-
centage of young hatched that died in the nest, the num-
ber of fledglings per successful nest, breeding success as
the number of fledglings per pair that laid at least one
egg and percent egg productivity as the number of fledg-
lings in each nest as a proportion of the total number of
eggs laid in each nest.

Temperature and rainfall in the month previous to the
onset of egg-laying were obtained from the meteorolog-
ical station of Orellana, which is in the study area.

We tested for normality of the data with Kolmogorov-
Smirnov tests. Only data on fledglings per successful nest
and breeding success were normally distributed. For
these two variables, analyses of variance and covariance
were used for statistical analysis following Sokal and Rohlf
(1979). A chi-square test was used to compare data on
clutch size distribution to a Poisson distribution. Distri-
bution of data on hatching success, percent nestling mor-
tality and percent egg productivity were compared to a
binomial distribution, also with a chi-square test. In no
case was the null hypothesis rejected (P > 0.05; data dis-
tribution did not differ from a Poisson or binomial dis-
tribution). Therefore, we ran a General Linear Model
(GLM) for data with a Poisson distribution to study be-
tween-year and seasonal variation in clutch size, and a
GLM for data with binomial distribution to study varia-
tion in hatching success, percent nestling mortality and
percent egg productivity (Crawley 1993). We used the
statistical software package “S-PLUS 4” to test GLMs with
chi-square tests (MathSoft 1997). To avoid small sample
sizes when clutch size was used as a factor, the extremely
low clutch size of one egg was removed {N = 2) from
analyses. A Mann-Whitney test was used to test for be-
tween-year differences in laying date.

Results

Kestrels initiated egg laying on 10 April in 1990 and
23 April in 1989. The majority of pairs started laying dur-

45

46

Short Communications

VoL. 34, No. 1

Table 1. Comparisons of the reproductive performance of Eurasian Kestrels in southwestern Spain, 1989-90.

1989

Mean ± SD
(N)

1990

Mean ± SD
(TV)

Statistic

P

Both Wars
Mean ± SD
(AO

Laying date

5 May ± 15.5

3 May ± 16.5

U = 633.5

3 May ± 16.1

(26)

(49)

0.856^

(75)

Clutch size

4.6 ±1.9

3.9 ± 1.3

x\i = 85.8

4.2 ± 1.3

(26)

(49)

0.171^

(75)

Percent hatching success

84.4 ± 16.9

60.6 ± 37.5

x2289 = 375.2

68.8 ± 33.4

(24)

(43)

0.0009^

(67)

Percent nestling mortality

0.0 ± 0.0

4.8 ± 15.0

X^206 ~ 139.3

2.9 ± 11.9

(22)

(37)

0.491b

(59)

Fledglings per successful nest

4.0 ± 1.0

3.1 ± 1.4

7^1.55 = 8.2

3.4 ± 1.4

(22)

(34)

0.006<=

(56)

Breeding success

3.9 ± 1.3

2.4 ± 1.8

7^1,65 = 11.8

2.9 ± 1.8

(23)

(43)

O.OOT

(66)

Percent egg productivity

80.4 ± 24.3

57.4 ± 37.4

X^284 = 368.6

65.2 ± 34.8

(23)

(43)

0.009b

(66)

^ Mann-Whitney test.

^ General Linear Model.
Analyses of the Variance.

ing the first and second weeks of April. Mean clutch size
was 4.2 ± 1.3 eggs (±SD, range = 1-6 eggs, N = 75;
Tables 1 and 2). The most common clutches consisted
of 4 (27.6%) and 5 eggs (33.5%; Table 2). A clear sea-
sonal decline in clutch size was observed both in 1989 (b
= —0.048 eggs/day; x ^24 ~ 52.6; P< 0.001) and 1990 (b
= —0.036 eggs/day; x^ 4 s = 91.1; P< 0.001). We did not
detect between-year differences in the seasonal trend in
clutch size (interaction of year x laying date; x ^72 = 71.3;
P = 0.43).

We detected no effects on reproductive performance
when we evaluated the interactions among year, laying
date and clutch size (x^ test and analysis of covariance; P
> 0.05 in all cases).

Discussion

The onset of egg laying, percentage of eggs laid in
April, clutch size and other reproductive rates observed
in our study were within the range described for this spe-
cies (Cave 1968, Shrubb 1970, Glutz 1971, Gordon and

Ridley 1979, Cramp and Simmons 1980, Bonin and
Strenna 1986, Beukeboom et al. 1988, Hasenclever et al.
1989, Gil-Delgado et al. 1995). However, since our study
was restricted only to pairs breeding in nest boxes, re-
productive success may have been higher due to the po-
tential for lower predation rates.

Some studies have shown large annual variation in lay-
ing dates, clutch sizes and reproductive rates in Eurasian
Kestrels (Dijkstra et al. 1982, Beukeboom et al. 1988).
Some of these have related variation in reproductive per-
formance to weather conditions before egg laying (Kos-
trzewa and Kostrzewa 1990, Kostrzewa and Kostrzewa
1991). In our study, mean temperature in the month pre-
vious to the onset of egg laying did not vary between
years and was 14°C in 1989 and 13.5°C in 1990 (P> 0.05,
A = 30 in both years). However, in 1989, the month pre-
vious to the onset of reproduction was very rainy (84.1
mm in 1989 vs. 14.1 mm in 1990). Mean egg laying date
and mean clutch size did not vary between years. Kestrel
breeding performance, as measured by fledglings per

Table 2. Clutch sizes of Eurasian Kestrels in southwestern Spain, 1989-90.

Clutch Size

Year

1

2

3

4

5

6

1989

0

1

3

7

9

6

1990

2

8

5

14

18

3

Both years

2

9

8

21

27

9

(2.6%)

(11.9%)

(10.5%)

(27.6%)

(35.5%)

(11.9%)

March 2000

Short Communications

47

successful nest and breeding success, was higher in 1989
when more rain fell. Grasshoppers {Dodostaurus marro-
canus) are the primary prey of Eurasian Kestrels during
the nesting season in Spain (Veiga 1985, J.M. Aviles and
D. Parejo pers. obs.). Higher rainfall in the spring did
result in an increase in grasshopper availability in our
study area (Arias et al. 1993) . Our study was not designed
to evaluate potential relationships among rainfall, food
supply and breeding performance so we could not be
certain of the relationship, if any, among these factors.

We detected a clear seasonal decline in clutch size and
fledglings per successful nest and breeding success in the
La Serena population of Eurasian Kestrels. This trend has
been found for other raptors (Newton and Marquiss
1984, Picozzi 1984, Hornfeldt and Eklund 1990, Korpi-
maki and Hakkarainen 1991), including Palearctic pop-
ulations of the Eurasian Kestrel (Cramp and Simmons
1980, Dijkstra et al. 1982, Meijer et al. 1988, Beukeboom
et al. 1988).

The productivity of this population of Eurasian Kes-
trels was within the range described previously for this
species in other areas. Although we found some evidence
of improved reproductive output related to rainfall, more
focused research is required to confirm our observations.

Resumen. — Se ha estudiado dos anos la biologia repro-
ductora del cernicalo vulgar {Falco tinnunculus) nidifican-
do en nidales artificiales en zona esteparias del sudoeste
en Espana. El inicio de la reproduccion, la fecha media
de puesta, el tamano de puesta y el resto de tasas re-
productoras estuvieron dentro del rango descrito para la
especie en la region Paleartica. La especie mostro un des-
censo estacional significativo de su valor reproductivo. El
numero de polios volados por nido exitoso y el exito re-
productor fueron mayores el ano en que existieron ma-
yo res precipitaciones primaverales, sin emabrgo, la escasa
serie temporal disponible no permite establecer una re-
lacion entre el nivel de precipitaciones y el exito de la
especie.

[Traduccion de Autores]

Acknowledgments

We thank D. Parejo who read the manuscript, checked
the English and made many useful suggestions. We also
thank K.L. Bildstein, D.E. Varland and an anonymous ref-
eree for their valuable comments. Statistical advice from
JJ- Sanz and M. Gonzalez was essential for the authors.

Literature Cited

Aparicio, J.M. 1994a. The seasonal decline in clutch size:
an experiment with supplementary food in the kes-
trel, Falco tinnunculus. Oikos 71:451—458.

. 1994b. The effect of variation in the laying inter-
val on proximate determination of clutch size in the
European Kestrel. J. Avian Biol. 25:275-280.

. 1997. Cernicalo vulgar. Pages 130-131 in F. Pur-

roy [Ed.], Atlas de la aves de Espana: (1975-1995).
Lynx, Madrid, Spain.

. 1998. Individual optimization may explain differ-
ences in breeding time in the European Kestrel. J.
Avian Biol. 29:121-128.

Arias, A., C. Alvez, F. Garcia, D. Martinez de Velasco,
J. Olivera, a. Prieto and R. Santos. 1993. La lucha
contra la langosta marroqui (Dodostaurus marrocanus
Thunb.) en Extremadura durante el decenio 1983—
1992. Bol. San. Veg. Plagas 19:425-453.

Aviles, J.M. and A. Sanchez. 1998. Crecimiento de los
polios de carraca (Coradas garrulus) en medios este-
parios del sudoeste de la Peninsula Iberica: influencia
de las precipitaciones. Misc. Zool. 21:1—7.

Beukeboom, L., C. Dijkstra, S. Daan and T Meijer. 1988.
Seasonality of clutch size in the kestrel Falco tinnun-
culus-. an experimental approach. Ornis Scand. 19:41-
48.

Bonin, B. and L. Strenna. 1986. Sur la biologie du fau-
con crecerelle Falco tinnunculus en Auxois. Alauda 53:
241-262.

Cave, A.J. 1968. The breeding of the kestrel, Falco tm-
nunculusL.., in the reclaimed area Oostelijk Flevoland.
Neth.J. Zool. 18:313-407.

Cramp, S. and K.E.L. Simmons. 1980. Handbook of the
birds of Europe, the Middle East and North Africa.
Oxford Univ. Press, Oxford, U.K.

Crawley, M.J. 1993. GLIM for ecologists. Blackwell Sci-
entific Publications, Oxford, U.K.

Dijkstra, C., D. Vuursteen, S. Daan and D. Masman.
1982. Clutch size and laying date in the kestrel (Falco
tinnunculus)-. effect of supplementary food. Ibis 124:
210-213.

Garzon, j. 1974. Contribucion al estudio del status, ali-
mentacion y proteccion de las falconiformes en Es-
pana central. Ardeola 19:279-330.

Gil-Delgado, J.A., J.A. Verdejo and E. Barba. 1995. Nest-
ling diet and fledgling production of Eurasian Kes-
trels (Falco tinnunculus) in eastern Spain./. Raptor Res.
29:240-244.

Glutz, V.B. 1971. Handbuch der Vogel Mitteleuropas, 4,
Falconiformes. Akademishe Verlagsgesellschaft,
Frankfurt, Germany.

Gordon, S. and G. Ridley. 1979. The kestrel in Ayrshire
1970-1978. Scott. Birds 10:201-216.

Hasenclever, H., a. Kostrzewa and R. Kostrzewa.
1989. Brutbiologie des turmfalken (Falco tinnunculus):
16-jahrige Untersuchungen in Westfalen 1972-1987.
J. Omithol. 130:229-237.

Hornfeldt, B. and U. Eklund. 1990. The effect of food
on laying date and clutch size in Tengmalm’s Owl Ae-
golius funereus. Ibis 132:395-406.

KorpimAki, E. and H. Hakkarainen. 1991. Fluctuating
food supply affects the clutch size of Tengmalm’s Owl
independent of laying date. Oecologia 85:543-552.

AND J. WiEHN. 1998. Clutch size of kestrels: sea-
sonal decline and experimental evidence for food lim-
itation under fluctuating food conditions. Oikos 83:
259-272.

48

Short Communications

VoL. 34, No. 1

Kostrzewa, a. and R. Kostrzewa. 1990. The relationship
of spring and summer weather with density and
breeding performance of the buzzard Buteo buteo, gos-
hawk Accipiter gentilis and kestrel Falco tinnunculus. Ibis
132:550-559.

Kostrzewa, R. and A. Kostrzewa. 1991. Winter weather,
spring and summer density, and subsequent breeding
success of Eurasian Kestrels, Common Buzzards, and
Northern Goshawks. Auk 108:342-347.

MathSoft. 1997. S-PLUS 4 Guide to Statistics. Data anal-
ysis Products Division, Seattle, WA U.S.A.

Meijer, X, S. Daan and C. Dijkstra. 1988. Female con-
dition and reproduction; effects of food manipulation
in free-living and captive kestrels. Ardea 76:141-154.

Newton, I. and M. Marquiss. 1984. Seasonal trends in
breeding performance of Sparrowhawks./. Anim. EcoL
53:809-831.

Picozzi, N. 1984. Breeding biology of polygynous Hen
Harriers Circus c. cyaneus in Orkney. Ornis Scand. 15:
1 - 10 .

Rivas-Martinez, S. 1987. Memoria del mapa de series de
vegetacion de Espafia. ICONA, Madrid, Spain.

SAnchez, a. and J.M. SAnchez. 1991. Resultados de ocu-
pacion de cajas anidaderas en tendidos electricos en
Extremadura (Oeste de Espaha): 1986-1990. Ecolo^a
5:375-381.

, P. Munoz, J, Caldera and J.M. SAnchez. 1996.

Resultados de ocupacion de cajas anidaderas en li-
neas electricas en ambientes esteparios en Extrema-
dura (1986-1994). Pages 246-247 inj. Fernandez and
J. Sanz-Zuasti [Eds.], Conservacion de las aves este-
parias y sus habitats. Junta de Castilla y Leon, Valla-
dolid, Spain.

Shrubb, M. 1970. The present status of the kestrel in
Sussex. Bird Study 17:1-25.

. 1993. The kestrel. Hamlyn Species Guides 2,

London, U.K.

SOKAL, R.R. AND F.J. Rohlf. 1979. Biometria. Ediciones
H. Blume, Madrid, Spain.

Tucker, G.M. and M.F. Heath. 1994. Birds in Europe:
their conservation status. Birdlife Internal., Cambridge,
U.K.

Veiga, J.P, 1982. Ecologia de las rapaces de un ecosistema
Mediterraneo de montaha: aproximacion a su estudio
comunitario. M.S. thesis, Universidad Complutense,
Madrid, Spain.

. 1985. Crecimiento de los polios de Falco tinnun-
culus en el centre de Espaha. Aspectos energeticos y
ecologicos. Ardeola 32:187-201.

Village, A. 1990. The kestrel. T. & A.D. Poyser, London,
U.K.

Received 5 December 1998; accepted 28 October 1999

J. Raptor Res. 34(l):48-52
© 2000 The Raptor Research Foundation, Inc.

Breeding Densities and Habitat Attributes of Golden Eagles in Southeastern Spain

Martina Carrete, Jose Antonio Sanchez-Zapata and Jose Francisco Calvo
Departamento de Ecologia e Hidrologia, Facultad Biologia, Universidad de Murcia, Campus de Espinardo,

301 00 Espinardo, Murcia, Spain

Key Words: Golden Eagle, Aquila chrysaetos; habitat,

Mediterranean .

Predictions on how animals respond to habitat changes
are the primary aim of many conservation studies. De-
velopment of easy wildlife habitat models is an important
tool for conservation and ecosystem management (Gon-
zalez et al. 1992, Don^ar et al. 1993). Progress has been
made using Generalized Linear Models (GLMs) (Dobson
1983, McCullagh and Nelder 1989) to summarize the re-
lationships between species distributions and environ-
mental variables (Vincent and Haworth 1983, Nicholls
1989, Donazar et al. 1993).

It is known that patterns and processes in nature are
sensitive to the scale at which they are viewed (Cody

1985, Wiens et al. 1987, Wiens 1989, Levin 1992, Lima
and Zollner 1996). The scale at which systems are studied
has a powerful influence on final conclusions and spe-
cies-habitat relationships determined at one scale may
not apply to others. Populations are influenced by the
complex arrangement of habitat patches within land-
scapes and multiscaled studies seem to be the proper way
to approach their study (Wiens 1989, Levin 1992).

The Golden Eagle {Aquila chrysaetos) is a raptor with a
widespread distribution in the northern hemisphere. In
North America, Steenhof et al. (1997) showed an impor-
tant interaction between jackrabbit {Lepus californicus)
abundance and weather on eagle reproduction and more
recent work using radiotracking data (Marzluff et al.
1997) has noted the preference of Golden Eagles for
some habitat types, particularly shrub and open lands. In

March 2000

Short Communications

49

Europe, different qualitative descriptions around nest
sites have been published (Tjernberg 1983, Watson
1997), and McGrady et al. (1997) constructed a model
that delineated the area over which eagle pairs range and
habitats of particular importance. More general ap-
proaches should be used as comparative parameters to
consider entire populations and to reduce individual var-
iability (White and Garrot 1990, Aebischer et al. 1993).

The purpose of this paper is to make a mathematical
description of Golden Eagle breeding sites in southeast-
ern Spain, where one of the highest densities of this spe-
cies has been reported (Sanchez-Zapata et al. 1995). We
propose a two-scale approach, considering at first re-
sponses around nest sites, and then a landscape-level
analysis to evaluate the influence of the matrix on breed-
ing territories.

Study Area and Methods

The study area covered the Murcia region, a 11317
km^ area located in southeastern Spain with numerous
mountains ranging from 0-2000 m elevation. The cli-
mate is Mediterranean arid and semiarid with a mean
annual rainfall of 300 mm. Vegetation has a mosaic struc-
ture with cultivated lands (54%), grasslands and shrub-
lands (28%), forest (15%) and open lands (3%) (Alcaraz
et al. 1991).

All the territories known to be occupied by Golden
Eagles at least once during the period 1985-97 were con-
sidered (Sanchez-Zapata et al. 1995). The location of
breeding territories was incorporated into a Geographic
Information System (IDRISI, Eastman 1992) using the
UTM grid of 1 km^ cells. For the first small-scale land-
scape approach the 1 km^ cells were aggregated into 9
km^ (3X3 km) cells, so the regional map of 11 317 km^
cells was transformed into a map with 1381 cells of 9 km^.
The large-scale landscape analysis was focused on 88 cells
of 100 km2 (10 X 10 km).

The same GIS was used to characterize the breeding
sites using the following variables (Table 1): (1) SLOPE
and LAND USE — slope (° from horizontal) was calculat-
ed from a Digitized Land Model 1:100 000 (Servicio Car-
tografico Espahol) by comparing the altitude of each ba-
sic cell (200 X 200 m) with that of neighboring cells to
the north, south, east and west. An average value for the
different 200 X 200 m cells was calculated. Slope for larg-
er cells (3X3 km and 10 X 10 km) was obtained as the
mean value of 200 X 200 m subcells. These values ranged
from 0—24.2 at the 9 km^ scale and from 0.2—13.5 at the
100 km^ scale. Different land-use classes were obtained
from maps of the Ministerio de Agricultura (1:200 000)
as proportions of cell area (9 km^ and 100 km^) covered
by each. New categories were formed by combining re-
lated land-use cover categories (e.g., lemon, orange and
other fruit trees were combined to give a single arbore-
ous intensive agriculture category). (2) EDGE — edge was
measured as the length (km) of edges between different
land uses using the digitalized land-use map and ATLAS
GIS software. (3) STRUCTURE — number and size (ha)
of the different patches of natural vegetation obtained
from maps of the Direccion General de Produccion
Agraria (1:200 000).

Table 1 . Variables used to characterize the breeding ar-
eas of Golden Eagles in southeastern Spain.

Variables Used in General Linear Model
Land Use Categories

AINTA — % of cell covered by arboreous intensive ag-
riculture, such as lemon and orange trees.

HINTA — % of cell covered by herbaceous intensive ag-
riculture, such as vegetable crops.

AEXTA — % of cell covered by arboreous extensive ag-
riculture, such as olive and almond trees.

HEXTA — % of cell covered by herbaceous extensive ag-
riculture, such as cereal crops.

SHRUB — % of cell covered by shrubland.

FOREST — % of cell covered by forest, mainly Pinus ha-
lepensis.

SHF — % of cell covered by mixed shrubforest.

SLOPE — topographic irregularity index.

Edges

EAEA — length (km) of edges between intensive and ex-
tensive agriculture.

FOIA — length (km) of edges between intensive agri-
culture and forest.

lASH-length (km) of edges between intensive agricul-
ture and shrubland.

lASF — length (km) of edges between intensive agricul-
ture and mixed shrubforest.

FOEA — length (km) of edges between forest and ex-
tensive agriculture.

EASH — length (km) of edges between extensive agri-
culture and shrubland.

EASE — length (km) of edges between extensive agri-
culture and mixed shrubforest.

FOSH — length (km) of edges between forest and
shrubland.

FOSF — length (km) of edges between forest and mixed
shrubforest.

SHSF — length (km) of edges between shrubland and
mixed shrubforest.

Structure

PATCH — number of land-use patches per cell.

RICHNESS — number of different land-use patches per
cell.

DIVERSITY — diversity (Shannon-Weiner) of land uses

NFOREST — number of forest patches per cell.

SFOREST — mean size (ha) of forest patches per cell

NSHRUB — number of shrubland patches per cell.

SSHRUB — mean size (ha) of shrubland patches per
cell.

NSHF — number of mixed shrub-forest patches per cell

SSHF — mean size (ha) of mixed shrub-forest patches
per cell.

NNAT — number of natural vegetation patches per cell.

SNAT — mean size (ha) of natural vegetation patches
per cell.

50

Short Communications

VoL. 34, No. 1

Table 2. Response of Golden Eagles to the different habitat variables considered in southeastern Spain. % dev:
deviance explained (ns — not significant, * P< 0.05, ** p< 0.01, *** P< 0.001. Responses: + s-shaped function, + +
bell-shaped function, — s-shaped function.

Scale

3X3

10 X 10

% Dev

Response

% Dev

Response

Land LFse

AINTA

ns

ns

HINTA

9.18*

—

11.37*

—

AEXTA

ns

ns

HEXTA

7.65**

—

ns

SHRUB

ns

ns

FOREST

13.09***

-1-

22.81***

+

SHF

7.60*

21.88***

+

SLOPE

28.89**

+ +

42.12***

+

Edges

EAIA

6.76**

—

9.61**

—

FOIA

ns

ns

lASH

ns

ns

lASF

ns

ns

FOEA

5.46*

+

14.45*

+ +

EASH

3.98**

—

ns

EASE

ns

ns

FOSH

2.88**

+

ns

FOSF

4.95*

+ +

26.94***

+

SHSF

9.59***

+

ns

Structure

PATCH

2.58**

—

ns

RICHNESS

3.11*

+ +

ns

DIVERSITY

1.88*

+ +

7.33*

+ +

NFOREST

7 80***

+ +

ns

SFOREST

11.34***

+ +

17.84*

+ +

NSHRUB

ns

ns

SSHRUB

ns

ns

NSHF

9.80**

+ +

8.31**

+

SSHF

5.24*

+ +

ns

NNAT

1.96**

+

ns

SNAT

11.60***

-h-h

15.64*

+ +

Due to the effect of increasing sizes of cells, many
of them included large areas of sea and adjacent re-
gions that were not censused. Therefore, these cells
were excluded from the data analysis making fewer
breeding territories in the large-scale study (76 vs. 40
territories).

We used Generalized Linear Models (GLMs) to con-
struct models of the breeding density of Golden Eagles
(Dobson 1983, McCullagh and Nelder 1989, Nicholls
1989). For density response variables (number of breed-
ing territories) , the Poisson distribution was an adequate
error function (Vincent and Haworth 1983) and the dis-
crete Poisson function an appropiate link function (L =
g(a+bixi+. . ,+bkXi,)y This meant that the number of breeding

sites in an area was a discrete, s-shaped function when
the linear predictor was the first order polynomial or a
bell-shaped function for second order polynomials (San-
chez-Zapata and Calvo 1999).

For regression analysis, we used the program STATIS-
TIX (Analytical Software 1992) following a forward step-
wise analysis (Donazar et al. 1993). Each explanatory var-
iable was tested for significance in turn. The variable
contributing to the largest significant change in deviance
from the null model was then selected and fit to the mod-
el. Once a variable was fit to the model, we tested if the
addition of a second variable significantly improved the
model. We chose a 5% level of significance to include a
variable in a model.

March 2000

Short Communications

51

Table 3. General Linear Model for Golden Eagle breeding density in southeastern Spain. % dev: deviance explained.

Coefficient

SE

P

% Dev

9 km^ Scale

Constant

-6.55778

0.62282

0.0000

30.16

SLOPE

0.60749

0.12321

0.0000

SLOPE2

-0.0167

0.00562

0.0000

SHSF

2.034e“4

8.296e-5

0.0142

Constant

-6.64018

0.63324

0.0000

30.32

SLOPE

0.60065

0.12638

0.0000

SLOPE2

-0.01693

0.0058

0.0035

SHF

0.23580

0.10192

0.0207

Constant

-6.86474

0.65345

0.0000

31.43

SLOPE

0.54579

0.13264

0.0000

SLOPE2

-0.01445

0.00601

0.0162

SSHF

8.415e

3.826e-’

0.0279

SSHF2

-1.447e-i3

7.055e-i4

0.0402

100 km2 Scale

Constant

-0.75439

0.72879

0.0000

43.70

SLOPE

0.38313

0.05857

0.0000

SHRUB

0.06663

0.03237

0.0396

Results and Discussion

SLOPE was the most important variable at both scales
explaining a higher percentage at the 100 km^ scale
(28.89% and 42-12%, respectively). When scale changed,
the response of eagles changed from quadratic to linear.
Pinus halepensis forests (FOREST) was the second most
important variable (13.09% and 22.81% for each scale)
with a similar percentage of deviance explained by mixed
shrubforest at the 10 X 10 km scale (21.88%). Intensive
agriculture was negatively associated with eagles (HINTA
9.18 and 11.37%, respectively), while cereal crops were
negatively correlated only at the smaller scale (HEXTA
7.65%) (Table 2).

Edges between land uses were also important at the
larger scale, but explained low percentages of deviance
at the 3X3 km scale. Edges between forest and mixed
shrubforest (FOSF 26.94%) and edges between forest
and extensive agriculture (FOEA 14.45%) were the most
explanatory edge variable. Eagles responded negatively
to edges between extensive and intensive agriculture at
both scales (EAIA 6.76% and 9.61%, respectively). There
was a negative relationship between eagle densities and
edge between shrub and extensive agriculture only at the
9 km^ scale (EASH 3.98%) (Table 2).

Percentages of deviance explained by landscape struc-
ture were generally low, except for natural vegetation and
forest patch sizes (SNAT 11.60%, 15.64% and SFOREST
11.34%, 17.84% for each scale). At the 9 km^ scale, the
number of patches of natural vegetation and eagle den-
sities seemed to be negatively related (PATCH 2.58%)
(Table 2).

Because slope accounted for the higher percentages of

deviance, models were constructed entering SLOPE as
the first variable. Only a small reduction in deviance was
obtained by including other variables. At the larger scale,
the model was more explanatory than at the 3X3 km
scale (Table 3).

Because most Golden Eagles in Murcia nest on cliffs,
slope was the most important variable in predicting its
breeding densities. The linear response at the larger
scale suggested that eagles preferred the bigger moun-
tain systems of the region.

The primary factor influencing Golden Eagle breeding
success is food availability (Steenhof et al. 1997). Eagles
prey on medium-sized mammals such as jackrabbits {Le-
pus spp.) and rabbits {Oryctolagus cuniculus) (Steenhof et
al. 1988, 1997, Watson 1997), which are very common in
shrublands in Mediterranean areas (Moreno and Villa-
fuerte 1995, Palomares and Delibes 1997). These open
lands, where vegetation structure favors prey detection
and hunting success (Tjernberg 1983, Marzluff et al.
1997, McGrady et al. 1997), were the second most im-
portant variable after slope in modeling breeding densi-
ties. Although forests seemed to be an important factor,
this could have been a consequence of their distribution
in the mountain systems of the area (Chaparro 1996).

Eagle densities were negatively correlated with irrigat-
ed crops, possibly because of the high number of people
working in fields and their low prey populations. The
increase in power lines associated with irrigated land
could have also been an important negative factor for
Golden Eagles, as electrocution is the main cause of mor-
tality for many eagle species (Gonzalez et al. 1990, Ferrer
and Hiraldo 1992, Sanchez-Zapata et al. 1995).

52

Short Communications

VoL. 34, No. 1

Resumen. — Utilizando Generalized Linear Models
(GLMs) examinamos la densidad reproductiva del Aguila
real Aquila chrysaetos en relacion con los usos del suelo,
los hordes entre usos y la estructura del paisaje en el
sureste de Espana. Las respuestas se compararon a dos
escalas de pais^ye. La pendiente fue la variable mas im-
portante para predecir la densidad reproductiva. Las
manchas de vegetacion natural se relacionaron positiva-
mente con las aguilas mientras que la agricultura inten-
siva se correlaciono de manera negativa. El matorral pa-
rece tener efectos positives importantes, probablemente
al incrementar la disponibilidad de alimento.

[Traduccion de Autores]

Acknowledgments

We would like to thank Sergio Eguia, Jose E. Martinez,
Miguel A. Sanchez and Javier Royo for their assistance in
the fieldwork.

Literature Cited

Aebischer, N.J., RA. Robertson and R.E. Kenward,
1993, Compositional analysis of habitat use from ani-
mal radio-tracking data. Ecology 74:1313-1325.
Alcaraz, R, R Sanchez-Gomez, A. de la Torre, S. Rios
and J. Alvarez Rogel. 1991. Datos sobre la vegeta-
cion de Murcia (Espana) . Gufa Geobotanica de la Ex-
cursion de las XI Jornadas de Fitosociologia. DM-
PPU. Madrid, Spain.

Analytical Software. 1992. STATISTIX. St. Paul, MN
U.S.A.

Chaparro, J. 1996. Distribucion potencial del bosque y
de sus especies arboreas en zonas mediterraneas semi-
aridas: modelos y aplicaciones. Ph.D. dissertation,
Universidad de Murcia, Murcia, Spain.

Cody, M.L. [Ed.]. 1985. Habitat selection in birds. Aca-
demic Press, San Diego, CA U.S.A.

Dobson, A.J. 1983. Introduction to statistical modeling.

Chapman and Hall, London, U.K.

Donazar, J.A., F. Hiraldo and J. Bustamante. 1993. Fac-
tors influencing nest site selection, breeding density
and breeding success in the Bearded Vulture ( Gypae-
tus barbatus).J. Appl. Ecol. 30:504-514.

Eastman, J.R. 1992. IDRISI 4.0. Worcester, MA U.S.A.
Ferrer, M. and F. Hiraldo. 1992. Man-induced sex-bi-
ased mortality in the Spanish Imperial Eagle. Biol.
Conserv. 60:57-60.

Gonzalez, L.M., J, Bustamante and F. Hiraldo. 1990.
Factors influencing the present distribution of the
Spanish Imperial Eagle Aquila adalberti. Biol. Conserv.
51:311-319.

, AND . 1992. Nesting habitat selec-
tion by the Spanish Imperial eagle Aquila adalberti.
Biol. Conserv. 59:45-50.

Levin, S.A. 1992. The problem of pattern and scale in
ecology. Ecology 73:1943-1967.

Lima, S.L. and RA. Zollner. 1996. Towards a behavioral
ecology of ecological landscapes. TREE 11:131-135.

Marzluff, J.M., S.T. Knick, M.S. Vekasy, L.S. Schueck
AND T.J. Zarriello. 1997. Spatial use and habitat se-
lection of Golden Eagles in southwestern Idaho. Auk
114:673-686.

McCullagh, P. AND J.A. Nelder. 1989. Generalized lin-
ear modeling. Chapman and Hall, London, U.K.

McGrady, M.J., D.R.A. McLeod, SJ. Petty, J.R. Grant
AND I.P. Bainbridge. 1997. Golden Eagles and forest-
ry, Res. Inform. Note 292. Forestry Commission Re-
search Agency. Aberdeen, U.K.

Moreno, S. and R. Villafuerte. 1995. Traditional man-
agement of shrubland for the conservation of rabbits
Oryctolagus cuniculus and their predators in Donana
National Park, Spain. Biol. Conserv. 73:81—85.

Nicholls, A.O. 1989. How to make biological survey go
further with generalized linear models. Biol. Conserv.
50:51-75.

Palomares, F. and M. Delibes. 1997. Predation upon Eu-
ropean rabbits and their use of open and closed
patches in Mediterranean habitats. Oikos 80:407-410.

Sanchez-Zapata, J.A., M.A. Sanchez, J.R Calvo and M.A.
Esteve. 1995. Ecologia de las aves de presa en la re-
gion de Murcia. Cuadernos de Ecologia y Medioam-
biente. Universidad de Murcia, Murcia, Spain.

and J.F. Calvo. 1999. Raptor distribution in rela-
tion to landscape composition in semi-arid Mediter-
ranean habitats./. Appl. Ecol. 36:254-262.

Steenhof, K., M.N. Kochert and J.H. Doremus. 1988.
Nesting of subadult Golden Eagles in southwestern
Idaho. Auk 100:743-747.

, AND T.L. McDonald. 1997. Interactive ef-
fects of prey and weather on Golden Eagle reproduc-
tion. J. Anim. Ecol. 66:350-362.

Tjernberg, M. 1983. Population density of Golden Eagle
in relation to nest-size and food availability. In Breed-
ing ecology of the Golden Eagle, Aquila chrysaetos (L,),
in Sweden. Report 10. Swedish Univ. Agric. Sci., Dept.
Wildlife Ecology. Uppsala, Sweden.

Vincent, J. and M. Haworth. 1983. Poisson regression
models of species abundance./. Biogeogr. 10:153-160.

Watson, J. 1997. The Golden Eagle. T. 8c A.D. Poyser,
London, U.K.

White, G.C. and R.A. Garrot. 1990. Analysis of wildlife
radio-tracking data. Academic Press, London, U.K.

Wiens, J.A. 1989. Spatial scaling in ecology. Fund. Ecol. 3:
385-397.

, J.T. Rotenberry and B. van Horne. 1987. Habi-
tat occupancy patterns of North American shrubsteppe
birds: the effects of spatial scale. Oikos 48:132-147.

Received 25 February 1999; accepted 28 October 1999

J. Raptor Res. 34(l):53-55

© 2000 The Raptor Research Foundation, Inc.

Trapping Estimates For Saker and Peregrine Falcons Used for Fai.conry in the United

Arab Emirates

Nigel W.H. Barton^

Dubai Falcon Hospital, P. O. Box 2391 9, Dubai, United Arab Emirates

Keywords: trapping, Saker Falcon', Peregrine Falcon] Falco

cherrug; Falco peregrinus; Middle East; falconry.

Falconry in the Middle East depends on large numbers
of wild-caught Saker {Falco cherrug) and Peregrine {F. per-
egrinus) Falcons (Allen 1980, Platt 1988, Remple 1988,
Riddle and Remple 1994, Cade 1997). The larger female
falcons are used primarily to hunt Houbara Bustard
( Chlamydotis undulata ) , but some Saker Falcons are used
for hunting Arabian hares {Lepus capensis). Smaller male
Peregrine Falcons are more suited to catch Stone Cur-
lews {Burhinus oedicnemus). Falcons are trapped from as
far as Eastern China to the Red Sea coast. Although in-
formation might be gathered from trappers or from mar-
kets, figures for the number of falcons being sold to the
Middle East are difficult to collect and their reliability is
questionable. Other species used in smaller numbers in-
clude Lanner Falcons {F. biarmicus), Laggar Falcons {F.
jugger), Barbary Falcons (F. p. pelegrinoides) , Red-headed
Falcons {F. p. babylonicus) , Black Shaheens (F p. peregri-
nator), Gyrfalcons {F. rusticolus) and captive-bred Gyrfal-
con hybrids.

Since the early 1980s, professional veterinary care has
been provided for these falcons with well-established clin-
ics in Dubai, Abu Dhabi, Qatar, Bahrain and more re-
cently in Saudi Arabia. At the Dubai Falcon Hospital
(DFH), comprehensive computer records have been
maintained since 1983. As many as 60 falcons are treated
each day during the peak season. The majority of falcons
admitted are from the United Arab Emirates with small
numbers from other parts of the Middle East. An in-
crease has been seen in the number of visits falconers
make to the veterinary clinic. This is largely due to in-
creased awareness amongst falconers and encourage-
ment from the service being sponsored by H.H. Sheikh
Hamdan bin Rashid A1 Maktoum as a free service to the
people. Many of the falcons are brought immediately
upon purchase for routine checks and vaccinations and
are therefore healthy on their first visit. The hospital is
the most accurate source for information on falcon num-
bers and species within the region. In recent years, the
majority of falcons seen at the Abu Dhabi Falcon Hospital
have been hybrids. Hybrids are much less common in
Saudi Arabia and the recent establishment of a falcon
hospital in Riyadh should provide data, particularly on

1 Present address: Penllynin Farm, College Road, Car-
marthen, Carmarthenshire, SA33 5EH, U.K.

the numbers of wild-caught Saker Falcons being sold to
that region. This paper examines data collected on fal-
con numbers in the DFH from 1983-98. It is intended
to estimate the minimum numbers of wild-caught falcons
being used for falconry in the United Arab Emirates,
mortality during the year, the proportion of captive-bred
falcons being used and the species used.

Methods

Each falcon brought to the hospital was routinely al-
located a sequential case number and accompanying vet-
erinary record which included information on the spe-
cies, sex, age, mass, relevant health problems and the
owner’s background. On its first visit, each falcon was
implanted with a permanent microchip making it iden-
tifiable for subsequent visits. Any veterinary problems
were treated and results of treatments, surgeries, micro-
biology, hematology and pathology were added to the
computerized veterinary record.

A distinction was made between new falcons, total fal-
cons and total visits. New falcons were those which had
no microchip and were, therefore, assumed never to have
been in the hospital before, or was implanted elsewhere
but had never been to the hospital before. Total falcons
were new falcons and falcons seen during the year which
had been recorded at the DFH in previous years. Total
visits were the sum of the number of visits made by each
falcon during the year. No account is taken of length of
stay in the hospital. The annual figures were calculated
for the year beginning June 1, the peak falcon season
being October-February.

Saker and Peregrine Falcons are the most popular spe-
cies and they accounted for most of the falcons seen. For
this reason, the data were analyzed separately and the
data were for new falcons only each year. Very few captive-
bred Saker or Peregrine Falcons are sold to the Middle
East. Therefore, the figures were a good estimate for the
number of falcons trapped.

Results

From 1983-84, the total number of falcons of all spe-
cies brought to the hospital for treatment or routine
checks was 73 (Fig. 1). This number increased to 2594
in 1997-98. During the 5-yr period from 1993-98, the
numbers remained relatively constant despite the fact
that very few falcons were not brought to the hospital.

Since 1993, a record was kept of the number of visits
made by each falcon and hence a total number of visits
for all falcons each year. The number of new falcons and
the total number of falcons remained relatively constant
(Fig. 2) averaging 1900 and 2400, respectively, each year.

53

54

Short Communications

VoL. 34, No, 1

3000

|To4« 'm nt,»

Figure 1. Total falcons admitted to the Dubai Falcon
Hospital each year since 1983.

However, falcons were brought more frequently than in
previous years with the total number of visits exceeding
5000 in 1995-96. No figure was available for 1997-98.
Since 1993, approximately 25% of the falcons brought to
the hospital in any one year had been to the hospital the
previous year or before, an indication that most falcons
were held during the summer months and used again
the following hunting season. Of the other 75%, some
were lost, released at the end of the season, died or were
not brought to the hospital in subsequent years.

Both Saker and Peregrine Falcons were categorized by
age and sex in 1993-98 (Table 1). The numbers were
calculated as a percentage of new falcons (all species in-
cluded) to indicate the changes within each species, sex,
and age category. The total number of Saker and Pere-
grine Falcons remained fairly constant at about 1500 in-
dividuals, down from a maximum of 1953 falcons during
the 1993-94 season. The total number of Peregrine Fal-
cons increased by 12.5% over the 5-yr period, whereas
Saker Falcon numbers decreased by 43%. The largest de-
crease was in juvenile female Saker Falcons.

There was a gradual decline in the percentage of Saker
Falcons and an increase in the percentage of Peregrine
Falcons, During 1993-94, Saker and Peregrine Falcons
accounted for 94% of all new falcons seen. In 1997-98,
they accounted for 74%. One reason for this decrease in
Saker and Peregrine Falcons may have been the resur-
gence in popularity of captive-bred Gyrfalcons and Gyr-
falcon hybrids. From a total of 58 hybrids seen in 1993—
94, the number increased to 274 during 1997—98.
However, in absolute numbers, Saker Falcons decreased
from 1292 to 732 over the 5-yr period.

Discussion

Within the United Arab Emirates, Saker Falcon num-
bers are decreasing whereas the numbers of captive-bred
hybrids and Peregrine Falcons seems to be increasing.
Historically, Peregrine Falcons were used more than Sa-
ker Falcons in the Gulf States. In the 1970s, there was a
surge of captive-bred falcons into the Middle East, but
their numbers subsequently decreased because falconers
preferred wild-caught falcons. This attitude was largely

Figure 2. New falcons, total falcons and total falcon
visits to the Dubai Falcon Hospital from 1993-98. No val-
ue was available for total visits in 1997-98.

due to falconers applying the same methods which they
used to train wild falcons to training captive-bred falcons
Captive-bred falcons require different training methods
and the Arabs did not get the best out of the falcons and
were disappointed in them. There has recently been a
resurgence in the number of captive-bred falcons used,
pardy because falconers have modified their training

Table 1. Numbers of new Saker and Peregrine Falcons
admitted to the Dubai Falcon Hospital from 1993-98 cal-
culated as a percentage of the total of new falcons ad-
mitted. All species included in parentheses.

Peregrine

Saker Falcon Falcon

Male Female Male Female

1993-94

Adult

26

(1.3)

225 (10.8)

19

(0.9) 81 (3.9)

Juvenile

112

(5.4)

929 (44.7)

189

(9.1) 372 (17.9)

Total

1292

(62.2)

661 (31.8)

1994-95

Adult

27

(1.4)

242 (13.1)

23

(1.3) 107 (5.8)

Juvenile

73

(4.0)

623 (33.8)

171

(9.2) 358 (19.4)

Total

965

(52.3)

659 (35.7)

1995-96

Adult

12

(0.7)

222 (13.1)

18

(1.1) 69 (4.1)

Juvenile

53

(3.1)

473 (27.8)

222

(13.1) 367 (21.6)

Total

760

(44.7)

676 (39.9)

1996-97

Adult

15

(0.8)

208 (10.9)

16

(0.8) 105 (5.5)

Juvenile

77

(4.0)

563 (29.5)

199

(10.4) 313 (16.4)

Total

863

(45.2)

633 (33.1)

1997-98

Adult

17

(0.8)

217 (11.0)

36

(1.8) 133 (6.7)

Juvenile

72

(3.6)

426 (21.5)

199

(10.1) 376 (19.0)

Total

732

(36.9)

744 (37.6)

March 2000

Short Communications

55

methods and because of the increasing interest in hy-
brids. Large, light-colored falcons have always been high-
ly valued in Arabia. As such, pale colored Gyrfalcons and
Saker Falcons were previously sought after. Survival of
pure Gyrfalcons in the desert environment and Middle
East climate requires special handling and fatalities are
frequent. Large, pale, captive hybrids such as Gyrfalcon/
Saker Falcon crosses combine the advantages of a desert-
adapted falcon with a falcon of exceptional power and
beauty.

Saker Falcon numbers, especially juvenile females,
have fallen dramatically. During the 1997-98 season, 426
were admitted to the hospital compared to 929 in 1993.
Data were collected at one hospital in Dubai, a rapidly
developing country. In recent years, some sheikhs within
Dubai have bought only large numbers of captive-bred
hybrids. It is possible, therefore, that the market for Sa-
ker Falcons has concentrated on other regions of the
Middle East. However, unless female Peregrine Falcons
are preferred to female Saker Falcons, a similar decrease
might be expected in Peregrine Falcon numbers. On the
contrary, Peregrine Falcons have shown a recent in-
crease.

Is the decrease in the number of Saker Falcons admit-
ted to the hospital a real indication of fewer Sakers being
trapped? If so, is this because of reduced demand or de-
creasing population numbers? Although juvenile num-
bers decreased, the number of adult Saker Falcons re-
mained constant. Does this indicate that Sakers have
experienced several consecutive poor breeding seasons?
Almost every falcon is brought to the hospital at the start
of the season for a routine check. Therefore, the differ-
ence in numbers between the species should not be at-
tributed to some species being more susceptible to dis-
ease and therefore more often seen in the hospital.

The falcon species flown vary in different regions of
the Middle East. For example in Taif, the plateau region
in southern Hejaz, Saudi Arabia and on the Tihama
plains adjacent to the Red Sea, Barbary Falcons are pre-
ferred. They are more suited to hunt Stone Curlews,
Francolins {Francolinus pondicerianus) and Partridge {Am-
moperdix heyi) found in this region. They are also cheaper
to buy and more within the price range for people from
this area than large falcons which sell for higher prices
in the larger cities and which are used for Houbara Bus-
tard and hares (Cade 1997). Laggar Falcons, Lanner Fal-
cons and other small falcons are also trapped, but they
have little financial value.

From the Dubai data, there is no evidence that the
number of falcons being used for falconry is increasing.
The number of falcons seen each year since 1993 re-
mained relatively constant. Approximately 1500 Saker
and Peregrine Falcons were brought to the DFH each
year. In addition to these, falcons were trapped for sale
in other parts of the U.A.E., Saudi Arabia, Kuwait, Bah-

rain and Qatar. Riddle and Remple (1994) estimated that
there might be as many as 8600 Saker and Peregrine Fal-
cons in captivity in the Middle East. It appears that there
is considerable movement of falcons within the Middle
East countries with falcons arriving in the Emirates and
subsequently being moved to Saudi Arabia. If this is the
case, then this figure would overestimate the actual num-
ber of falcons being used. Additional data from other
falcon hospitals in the Middle East, especially Abu Dhabi
and Riyadh, in conjunction with ongoing population
studies should enable us to determine whether Saker and
Peregrine Falcon populations can support this trapping
pressure and, if not, what approach should be taken to
regulate it.

ReSUMEN. — El halcon sacre {Falco cherrug) y el peregrino
{Falco peregrinus), son los halcones mas utilizados en la
cetreria en el Medio Oriente. Los registros veterinarios
en el hospital de halcones de Dubai proveen una base
para estimar el numero de halcones atrapados anual-
mente. Durante el periodo de 1993-98, el numero de
halcones peregrinos atendidos en el hospital se incre-
mento en un 12.5%, mientras que los halcones sacre dis-
minuyeron en un 43%. Los hibridos de halcon gerifalte
{Falco rusticolus) aumentaron en un 13.8% del total de
halcones observados en la clinica. Las hembras juveniles
de halcon sacre disminuyeron en un 54%.

[Traduccion de Cesar Marquez]

Acknowledgments

I would like to thank the Dubai Falcon Hospital, T.
Cade, D. Ellis and an anonymous referee for providing
helpful comments on the manuscript.

Literature Cited

Allen, M.A. 1980. Falconry in Arabia. Orbis, London,
U.K.

Cade, TJ. 1997. Pages 135-142 in H.H. Roth and G. Merz
[Eds.], Wildlife resources: a global account of eco-
nomic use. Springer-Verlag, Berlin, Germany.

Platt, J.B. 1988. The genus Falco in Arabia. Pages 307-
312 m TJ. Cade [Ed.], Peregrine Falcon populations,
their management and recovery. The Peregrine Fund
Inc., Boise, ID U.S.A.

Remple, J.D. 1988. An overview of Arab falconry, its med-
ical lore and the introduction of avian medicine m
the Arabian Gulf. Pages 825-830 in T.J. Cade [Ed.] ,
Peregrine Falcon populations, their management and
recovery. The Peregrine Fund Inc., Boise, ID U.S.A
Riddle, K.E. and J.D. Remple. 1994. Use of the Saker and
other large falcons in Middle East falconry. Pages
415-420 in B.-U. Meyburg and R.D. Chancellor
[Eds.], Raptor conservation today. WWGBP, Berlin,
Germany.

Received 29 August 1998; accepted 6 October 1999

J. Raptor Res. 34(l):56-57
© 2000 The Raptor Research Foundation, Inc.

First Confirmed Breeding Records and Other Incidental Sightings of Northern

Harriers in Labrador

TonyE. Chubbs

Department of National Defence, 5 Wing Goose Bay, Box 7002, Postal Station A, Happy Valley — Goose Bay, Labrador,

NF AOP ISO Canada

Bruce Mactavish

Jacques Whitford Environment Limited, 607 Torbay Road, St. John's, NF AlA 4Y6 Canada

Keith Oram and Perry G. Trimper

Jacques Whitford Environment Limited, Box 274, Postal Station C, Happy Valley — Goose Bay, Labrador,

NF AOP ICO Canada

Key Words: Northern Harrier, Circus cyaneus; nest, breed-

ing, distribution] Labrador.

The Northern Harrier ( Circus cyaneus) is a globally dis-
tributed raptor of upland grasslands and fresh- and salt-
water marshes (MacWhirter and Bildstein 1996). Popu-
lations in northeastern North America have declined
precipitously over the past few decades (Kirk and Hyslop
1998). Through the early 1960s, Northern Harriers were
recorded along the north shore of the Gulf of St.
Lawrence near Natashquan (Todd 1963), but none were
recorded as occurring in Labrador (Godfrey 1986). Re-
cent range descriptions indicate that the northern extent
of the breeding range falls south of the Labrador border
(Macwhirter and Bildstein 1996) but, unlike most parts
of its range, the Northern Harrier is extending its breed-
ing range northeast into Labrador.

Herein, we describe the first known nest sites of North-
ern Harriers in Labrador and provide information on
other recent sightings.

Methods

Most sightings were recorded during surveys conduct-
ed as part of the Environmental Mitigation Program for
the Department of National Defence and for the Lower
Churchill River Labrador Hydro Project. Other inciden-
tal observations were made opportunistically by the au-
thors. Surveys have included the majority of Labrador
south of 57°N and most of northeastern Quebec, al-
though no sightings were recorded in the latter region.
Both nesting records were recorded during brief helicop-
ter stops at the airport in Churchill Falls, Labrador.

Results

On 11 July 1998, a male Northern Harrier was seen
flying over black spruce {Picea mariana) forest north of
the airport runway at Churchill Falls, Labrador (53°34'N,
64°08'W). Shordy thereafter, a female Northern Harrier
appeared and the male transferred prey to her in midair,
and then flew away. The female flew low over an alder

{Alnus spp.) thicket and dropped to an unseen point on
the ground.

We searched the area until we flushed the female. The
nest was located on the ground in a band of alders 25 m
wide on the north side of the runway. The alder thicket
ranged in height from 1—3 m and was bordered by black
spruce forest 30—50 m to the north. The nest was 15 m
from the open edge of the runway and separated by a 1-
m wide, water-filled ditch. The nest was a shallow bowl of
grass {Carex spp.) approximately 30 cm in outside diam-
eter with base material consisting of fallen alder leaves.
The nest was on gravel with some mosses and grasses
growing nearby. It contained one young <2-d old and
two white eggs.

The nest site was revisited on 18 August 1998. Since
our initial visit, the area had been clearcut of alders and
all brush had been removed. There was no sign of the
Northern Harriers or the nest. However, a low helicopter
flight over the same area flushed a buff-colored juvenile
Northern Harrier from the edge of the spruce forest less
than 30 m from the site of the nest. The 38-d period
between the approximate hatch date and this observa-
tion appeared to have been sufficient time for the young
harrier to fledge from the nest.

On 10 May 1999, an adult male was flushed on the
north side of the airfield and on 21 May a female was
observed on the south side, an area with some remaining
alders. On 25 May, we found a nest at the base of a group
of alders near small spruce trees that contained one
white egg. The base of the nest was constructed of dead
alder stems <1.5 cm in diameter and was lined with dry
grasses. The nest dimensions were approximately 45 cm
outside diameter, 20 cm inside diameter and 8 cm deep.
Ground cover within 3 m of the nest was comprised
mainly of mosses {Sphagnum spp.), Labrador tea {Ledum
groenlandicum) and sheep laurel {Kalmia angustifolia) .The
site was damp and interspersed with small pools of stand-
ing water from spring flooding. A nest check on 6 June
showed the nest contained a clutch of 5 eggs. On 10 July,
five young were observed and, on 29 July, three young

56

March 2000

Short Communications

57

were banded and two had already fledged. The three
banded young fledged on 30 July. These fledging dates
were consistent with those detailed by Harrison (1984).

Most observations of Northern Harriers in Labrador
were made in the coastal marsh areas of Lake Melville,
the lower Churchill River valley and the south Labrador
coast. The northernmost observation was an adult near
Snegamook Lake (54°35'N, 61°20'W). Several sightings
have been made of Northern Harriers in the Goose Bay
area (53°20'N, 60°20'W) in each summer from 1996-99.
Other sightings include a female carrying food at Ter-
rington Basin on 26 July 1998, a juvenile flying near Lac
Brule (52°20'N, 63°50'W) on 18 August 1998 and a pair
flying near the Churchill River at the outflow of the Eliz-
abeth River (53°14'N, 63°18'N) on 22 May 1999.

In the last decade, the Northern Harrier has expanded
its breeding range on insular Newfoundland and is now
a fairly common widespread breeder, especially on the
Great Northern Peninsula. Regional increases in obser-
vations may be correlated with coinciding peaks in mi-
crotine populations (Hamerstrom 1979, Hamerstrom et
al. 1985, Simmons et al. 1986). The species is now fairly
common on the Labrador side of the Strait of Belle Isle
where it is presumably breeding. Shrubland heath and
alder thickets are common vegetation types in insular
Newfoundland and Labrador and may represent a signif-
icant amount of suitable breeding habitat for Northern
Harriers. Nest-site locations in Labrador have similar hab-
itat to that used by Northern Harriers in other parts of
their range (Hamerstrom 1969, Simmons and Smith
1985). Little information is available on the distribution
and breeding range of Northern Harriers in northeast-
ern Canada, specifically Labrador. If estimates in other
northern regions by Potts (1998) are accurate, a signifi-
cant proportion of the northeastern North American
population may be unaccounted for due to lack of survey
effort. Examination of potential suitable habitat in Lab-
rador may reveal additional breeding localities.

Resumen. — Circus cyaneus es un ave poco comun en el
sur y centro de Labrador sin ningun registro previo de
anidacion. Detallamos observaciones recientes y la pri-
mera anidacion exitosa de Circus cyaneus en Labrador.

[Traduccion de Cesar Marquez]

Acknowledgments

Remote sightings reported here were made during sur-
veys for other species as part of the Environmental Miti-
gation Program for the Department of National Defence
(DND), Maj. G. Humphries, project manager. We thank

all those involved in the collection of information. Spe-
cial thanks are extended to D. Lemon and L. Grattan of
the Labrador Hydro Project for their collaboration and
the use of their data. W. Price and R. Neville were ob-
servers on many of the flights. G. Baikie of CEL Co.
searched for and found the nest site in 1999 and provid-
ed resighting information. G. Goodyear of Universal He-
licopters Newfoundland Limited conducted the banding
of the nestlings and was the pilot during many of the
surveys from which our observations were made. Com-
ments by A.J. Erskine, K.L. Bildstein and an anonymous
reviewer improved the manuscript.

Literature Cited

Godfrey, W.E. 1986. The birds of Canada, revised edi-
tion. National Museum of Natural Sciences, National
Museum of Canada, Ottawa, ON Canada.
Hamerstrom, F. 1969. A harrier population study. Pages
367-383 in].]. Hickey [Ed.], Peregrine Falcon popu-
lations, their biology and decline. Univ. Wisconsin
Press, Madison, WI U.S.A.

. 1979. Effect of prey on predator: voles and har-
riers. Auk 96:370-374.

, FN. Hamerstrom and CJ. Burke. 1985. Effect

of voles on mating systems in a central Wisconsin pop-
ulation of harriers. Wilson Bull. 97:332-346.

Harrison, C. 1984. A field guide to the nests, eggs and
nestlings of North American Birds. Collins Publishers,
Toronto, ON Canada.

Kirk, D.A. and C. Hyslop. 1998. Population status and
recent trends in Canadian raptors — a review. Biol. Con-
serv. 83:91—118.

MacWhirter, R.B. and K.L. Bildstein. 1996. Northern
Harrier {Circus cyaneus). In A. Poole and F. Gill [Eds.],
The birds of North America, No. 210. The Academy
of Natural Sciences, Philadelphia, PA and the Amer-
ican Ornithologists’ Union, Washington, DC U.S.A
Potts, G.R. 1998. Global dispersion of nesting Hen Har-
riers Circus cyaneus; implications for grouse moors in
the U.K. Ibis 140:76-88.

Simmons, R.E. and P.C. Smith. 1985. Do Northern Har-
riers choose nest sites adaptively? Can.J. Zool. 63:494—
498.

, P. Barnard, B. MacWhirter and G.L. Hansen.

1986. The influence of micro tines on polygyny, pro-
ductivity, age, and provisioning of breeding Northern
Harriers: a 5-year study. Can.J. Zool. 64:2447-2456.
Todd, W.E.C. 1963. Birds of the Labrador Peninsula and
adjacent areas. Univ. Toronto Press, Toronto, ON
Canada.

Received 19 July 1999; accepted 20 November 1999

Letters

J. Raptor Res. 34(l);58-59
© 2000 The Raptor Research Foundation, Inc.

First North American Record oe a Melanistic Female Northern Harrier

Melanism, or dark coloration that results from the excessive deposition of melanin pigments (Sage 1962, Brit. Birds
55:201-225), occurs in a diverse array of birds (Sage 1962; Gross 1965, 36:240-242). Melanism is espe-

cially common in birds of prey, many of which display a well-documented dark-morph plumage (Brown and Amadon
1968, Eagles, hawks and falcons of the world, Vols. 1 and 2, McGraw-Hill Book Company, New York, NYU.S.A.). In
these polychromatic species, dark morphs do not appear to be any less successful than their light-morph counterparts.
In addition, melanistic birds reproduce successfully with both light and dark conspecifics (Palmer 1988, Handbook
of North American birds, Vol. 5, Part 2, Yale Univ. Press, New Haven, CT U.S.A.).

Interestingly, given their prevalence in a variety of raptor species (e.g., Rough-legged Hawk \_Buteo lagopus]. Fer-
ruginous Hawk [Buteo regalis], Eleonora’s Falcon [Falco eleonorae]), melanistic individuals appear to be exceedingly
rare in certain birds of prey. Despite the fact that melanism has been documented for several Palearctic harriers
(Brown and Amadon 1968; Cramp and Simmons 1980, Handbook of the birds of Europe, the Middle East and North
Africa, Vol. 2, Oxford Univ. Press, Oxford, U.K.; Clark 1987, Brit. Birds 80:61-72) , including the Hen Harrier (Circus
cyaneus cyaneus) (Watson 1977, The Hen Harrier, T. & A.D. Poyser, Berkhamsted, U.K.), only one record exists of a
melanistic Northern Harrier (Circus cyaneus hudsonius), a male observed in California (Howell et al. 1992, West. Birds
23:79-80). Here, we report the first North American record of a melanistic female Northern Harrier.

At 1334 H MST on 12 November 1998, we observed a dark harrier coursing over a field 5 km northwest of Charlo,
Lake County, Montana (47°27'30"N, 114°13’30"W). We observed the harrier for approximately 40 min. Observations
were made in good light with 10 X 50 binoculars and a 20-60 zoom spotting scope at distances ranging from 100-
400 m. Despite its dark coloration, the harrier was easily identifiable by its characteristic flight and shape, its under-
wing markings, facial disk and hunting behavior. The bird’s most notable feature was the absence of the characteristic
white uppertail coverts (i.e., rump patch). Its head, back, uppertail coverts, breast and belly appeared to be black
On the upper wing, the marginal coverts were a dark brown with a hint of rufous. The location of this rufous area
matched that of the tawny coloration on the upper marginal coverts of typical-morph birds. The dorsal side of the
remiges were brownish-black giving them a slightly lighter appearance than the body. Although the upperside of the
tail was darker overall, it was very similar to that of a typical female harrier and consisted of 3-4 bold gray and black
bands and uniformly dark central rectrices.

The melanistic harrier’s underwing coverts were uniformly blackish-brown, whereas ventrally the remiges were light
grayish-silver with fairly dark barring on the secondaries and primaries, including the outer primaries. The underside
of the tail was similar to the dorsal side in that it consisted of bold gray and black bands. Finally, the bird had a
yellow cere and hides, the latter confirming this bird’s status as an adult. Typical adult female Northern Harriers
show dark barring on the underside of the remiges, particularly on the outer primaries, whereas typical adult males
exhibit inky-black wing tips on the underside and a light underwing with a bold subterminal band along the second-
aries (Clark and Wheeler 1987, A field guide to hawks, Houghton Mifflin Company, New York, NYU.S.A.). The
overall pattern, particularly the barring of the underwing, of the melanistic individual that we observed helped
confirm that it was a female. Howell et al. (1992) also determined the age and sex of a melanistic male Northern
Harrier by eye color, underwing pattern and size. The melanistic female interacted with a juvenile Northern Harrier
that was noticeably smaller and believed to be a male. Several minutes after finding the melanistic harrier, it stooped
twice on the juvenile, which turned and presented its talons each time and then left the area. A few minutes later
we watched the melanistic harrier catch and consume what appeared to be a vole (Microtus spp.). In the following 5
mm, it made two more capture attempts (both on small mammals), one of which was successful. Overall, during our
brief observation period, the melanistic harrier appeared to be hunting successfully, free of harassment by other
harriers.

Despite daily visits to the area where we first observed this melanistic harrier and continued surveys over the next
six months, we were unable to locate it again.

Unlike albinism, melanism is a heterozygous dominant trait (Sage 1962). However, Clark (1998, Wilson Bull. 110.
289-290) suggested that in monochromatic species, the spread of melanism may be inhibited by the inability of
melanistic individuals to acquire mates and produce offspring. Although the occurrence of melanism in both female

58

March 2000

Letters

59

and male Northern Harriers has now been documented, whether such individuals do in fact attract mates and
reproduce successfully remains unknown. — Chad V. Olson, Montana Cooperative Wildlife Research Unit, University
of Montana, Missoula, MT 59812 U.S.A. and Sophie A.H. Osborn, Division of Biological Sciences, University of
Montana, Missoula, MT 59812 U.SA.

J. Raptor Res. 34(l):59-60
© 2000 The Raptor Research Foundation, Inc.

Observations at an Ayres’ Hawk-Eagle Nest in Kibale National Park, Uganda

The Ayres’ Hawk-Eagle {Hieraaetus ayresii), an inhabitant of forested and wooded areas of sub-Saharan Africa, has
long been considered inexplicably rare throughout its range (Brown et al. 1982, Birds of Africa, Vol. 1. Academic
Press, London, U.K.). Its breeding behavior has been observed at no more than six nests in Kenya (Brown 1955, Ibts
97:38-64, 183-221; Brown 1966, Ibis 108:531-572; Dewhurst et al. 1988, Gaftar 3:85-93) and Zimbabwe (Phillips 1978,
Honeyguide 94::27-S9] Steyn 1982, Birds of prey of southern Africa, Tanager Books, Dover, U.K.), and some aspects of
its biology and ecology are still considered unknown (Virani and Watson 1998,/. Raptor Res. 32:28-39) . Here, I present
information on the nest site, breeding phenology, and behavior of a nesting pair of Ayres’ Hawk-Eagles in western
Uganda.

In August 1997, I located an Ayres’ Hawk-Eagle nest in the unlogged K-30 forestry compartment of Kibale National
Park (0°13'-0°41'N, 30°19'— 30°32'E), an area of moist, evergreen forest in western Uganda. A complete description
of habitat and climate are provided by Struhsaker (1997, Ecology of an African rainforest, Univ. Presses Florida,
Gainesville, FL U.S.A.). I monitored the nest through November 1997, checking it at least twice every month. Using
10 X 40 binoculars, I observed the nest from the ground about 30 m upslope of the nest. I observed the nest for
35.5 hr during 7 d of the incubation period and 50.5 hr during 7 d of the nestling period. Adult gender was initially
determined by size dimorphism, but once confirmed, the darker female was easily distinguished from the lighter
male.

Located on an 18° slope with a WSW aspect, the nest tree was a live Lovoa brownii, 38 m tall and 98 cm diameter
at breast height. The tree had no vines, few epiphytes, and the crown was completely isolated from the surrounding
canopy. The stick nest was positioned at a height of 32 m, at a four-way fork of a major branch. Located near the
base of the crown, the nest was plainly visible from the ground, but well-shaded from the sun. I estimated from the
ground that the nest measured 1.25 m across and 0.75 m tall. When first observed on 21 August 1997, the female
was already incubating. Based on the behavior of the adults, I estimated that one nestling hatched 31-33 d later,
between 30 September-2 October 1997. The nestling was first observed approximately three weeks later, when its
head was visible above the rim of the nest. Observations were discontinued on 24 November 1997, approximately 54
d after hatching, when the nestling was nearly fully feathered and it spent most of its time standing at the edge of
the nest.

During the incubation period, the female spent most of her time at the nest, either incubating (88%) or perched
nearby (5%). The female was out of sight for only 7% of this time, for periods ranging from 2—32 min. In contrast,
the male never incubated, made only short visits to the nest (2-14 min) and nest tree (15-19 min), and was out of
sight for 97% of this time. I observed no deliveries of prey or nesting material to the nest during incubation; possibly
some prey exchanges occurred away from the nest, where the female fed before returning.

After the nestling hatched, the female remained the primary nest attendant spending 61% of the observation
hours brooding or feeding the nestling. When not tending the nestling, the female was perched at or near the nest
16% of the time and out of sight of the observer for 23% of the time. The amount of time the female spent away
from the nest increased as the nestling grew. As during the incubation period, the male was out of sight for most
(98%) of the observation hours. The male made seven visits to the nest during the nestling period delivering nesting
material twice and prey once. Possibly some prey exchanges during the nestling period also occurred at some distance
from the nest. Most prey arrived at the nest apparently plucked of fur or feathers; the female spent little time handling
prey prior to feeding the nestling. I checked for prey remains beneath the nest tree throughout the incubation and
nestling periods, but found none. Except for an unidentified, pigeon-sized bird, I was unable to identify prey fed to
the nestling.

These observations showed the breeding behavior of Ayres’ Hawk-Eagle in the Kibale National Park was similar to

60

Letters

VoL. 34, No. 1

other regions of Africa. Although I was unahle to document the complete duration of the nesting cycle, I estimated
that it fell well within the 45 d incubation and 75 d nestling periods reported by Brown et al. (1982). Similar pairings
of dark females with light males have been recorded at at least two nests in Kenya (Brown 1966, Dewhurst et al.
1988). There are no records of nests containing more than a single nestling and records of two-egg clutches from
South Africa are considered suspect by most authors (Brown et al. 1982, Steyn 1982). Adult behavior at nests in
Kenya was consistent with my observations; incubation and brooding were conducted primarily or exclusively by the
female, while the male made brief visits to the nest to deliver prey and nesting material (Brown 1955, Dewhurst et
al. 1988).

E. Aliganyiri, A. Ermosi, T. Lawrence, J. Mugenyi, and K. Patrick assisted in the collection of data. 1 express my
gratitude to the Department of Zoology, Makerere University Biological Field Station, C. Chapman, and L. Chapman
for providing logistical support in Kibale National Park. A. Baker, C. Chapman, A. Kemp, and R. Watson provided
constructive comments on the manuscript. — Nathaniel E. Seavy, 223 Bartram HaU, P.O. Box 118525, Department of
Zoology, University of Florida, Gainesville, FL 32611-8525 U.SjI.

J Raptor Res. 34(1) :60

© 2000 The Raptor Research Foundation, Inc.

A Case of Nest Predation on Turkey Vultures Nesting in Argentina

The Turkey Vulture {Cathartes aura) is widely distributed along the American continent (May, J.B. 1935. The Hawks
of North America. National Association of Audubon Societies, New York, NY U.S.A.) but its breeding range in Ar-
gentina is not well-known. Only three records on Turkey Vulture nesting have been published up to now. Martin De
La Pena (1992, Guia de las Aves Argentinas. Literature of Latin America [Eds.], Buenos Aires, Argentina) found two
nests in northeastern Argentina and Jerome Jackson refers to another nest found on the Falkland Islands (1983,
Nesting phenology, nest-site selection and reproductive success of Black and Turkey Vultures. Pages 245-270 in S.
Wilbur and J. Jackson [Eds.], Vulture biology and management. Univ. California Press, Berkeley, CAU.S.A.). Turkey
Vultures have a very diversihed selection of nest sites, including hollow logs, prostrate trees and stumps, sides of steep
cliffs, abandoned buildings or just on the ground (Coles, V. 1938, Studies in the life history of the Turkey Vulture
Cathartes aura septentrionalis"WieA. Ph.D. dissertation, Cornell Univ., Ithaca, NYU.S.A.).

We found a Turkey Vulture nest in Lihue Calel National Park, La Pampa Province, Argentina in 1998. Lihue Calel
National Park is located in the central region of Argentina (38°00'S, 65°35°W) and contains bare rock hills (elevation
589 m) surrounded by flat semiarid desert (elevation 300 m). Vegetation in the area is a fine-grained mosaic of open
patches and scrubs, including Larrea cuneifolia, L. divaricata, L. nitida, Prosopis alpataco and Condalia microphylla. Each
year from September-February, large flocks of Turkey Vultures are commonly observed soaring in the park but no
nests have been recorded.

We found the nest on 21 November and the female was incubating two eggs in the nest. The vulture flew away as
we approached the site. The eggs were weighed and measured (68.7 X 49.05 mm, mass = 82 g; 69.05 X 48.65 mm,
mass = 82 g). The nest was on the north side of a hill approximately 40-m high. The nest was near the top of the
hill on the ground in a circular (8 m in diameter) patch of bushes {Geoffroea decorticans, Larrea nitida and Lycium
gilliesianum) . Height of the bushes averaged 2.3 m and they were surrounded by grass and rocks. Two rocks 70-cm
high protected the nest on the east and south sides. The nest was not in a depression and there were only 5 or 6
small twigs and some feathers next to the eggs.

On 26 December, we revisited the nest and found two young that we weighed and measured (950 g, total length
345 mm, wingspan 753 mm, wing chord 147 mm and culmen 17.15 mm; 1030 g, total length 358 mm, wingspan 842
mm, wing chord 172 ram and culmen 19.15 mm). Their breasts and backs were covered with down but all sheaths
of primaries and secondaries extended through the down layer and remiges emerged out from their sheaths about
2-4 cm. On 12 January 1999, the young were not found in the nest and the remains of one of their wings was about
6 m from the nest. No signs or footprints of predators were found at the nest site but the nest may have been
depredated by a felid, fox or reptile.

We wish to thank National Parks Administration of Argentina, staff of the Lihue Calel National Park, and special
thanks to Jose Antonio Sarasola, Daniel Munoz, Rene Barquin, Agustin Lanusse and Maria Cristina Martin. — Jose
Hernan Sarasola, Ramon Alberto Sosa and Juan Jose Maceda, Facultad de Ciencias Fxactas y Naturales, Universidad
Nacional de La Pampa, Uruguay 151, (6300) Santa Rosa, La Pampa, Argentina.

March 2000

Letters

61

J. Raptor Res. 34(1) :61

© 2000 The Raptor Research Foundation, Inc.

Decline of the Egyptian Vulture {Neophron percnopterus) in the Canary Islands

The Canary Islands lie off the northwest coast of Africa 100 km (Fuerteventura) north of the Tropic of Cancer.
There are seven islands and six islets of volcanic origin with mountainous terrain. The climate is subtropical. Since
the beginning of the 20th century the Canarian Black Oystercatcher {Haematopus meadewaldoi) and the Red Kite
{Milvus milvus) , as well as two endemic subspecies, Phylloscopus collybita exsul and Saxicola dacotiae murielae (Emmerson
et al. 1994, Catalogo y bibliografia de la avifauna canaria, Cabildo de Tenerife, Santa Cruz de Tenerife, Canary Islands)
have become extinct as breeding birds. Here, we report on the population trend of the Egyptian Vulture {Neophron
percnopterus) in the Canaries since the beginning of this century.

Our methods included a literature survey of all published and unpublished works referring to the species. We also
conducted a census of Fuerteventura and neighboring islets, the only area where there is still a breeding population,
and collected information on the breeding success of this island’s population in 1998.

In the Macaronesian region, the Egyptian Vulture is found only in the Canary Islands and Cape Verde. In both
cases, it is sedentary and does not migrate to the nearby African continent. The species is also considered accidental
m the Azores and Madeira (Le Grand 1983, Archipeligo 4:49-58; Bannerman and Bannerman 1965, Birds of the
Atlantic Islands, Vol. II, A history of the birds of Madeiri, the Desertas and the Porto Santo Islands, Oliver and Boyd,
London, U.K.). It was relatively abundant in the past in all the Canary Islands, with the exception of La Palma. Since
the 1950s, populations have been declining both in the Canary Islands and Cape Verde.

The first census of the population of Egyptian Vultures in the Canary Islands was conducted in 1987. It showed
that 31-37 pairs inhabited the islands with 26-31 pairs concentrated in Fuerteventura (Delgado et al. 1993, Bol. Mus.
Munic. Funchal. 2:77-78). The census we conducted in 1998 located only 18 nesting pairs of Egyptian Vultures m
Fuerteventura with the possibility of two additional pairs that also probably bred on the island. There were 1-2 pairs
in Lanzarote and only one pair on the islet of Alegranza making the total breeding population in the Canary Islands
at about 22 breeding pairs. This represented a 30% reduction in the breeding population in 10 yr.

Of the 18 pairs monitored in Fuerteventura, 10 did not breed successfully. Five of these failures occurred when
nests were deserted prior to egg laying, three occurred because the eggs were infertile and two occurred when young
were stolen from the nests. Only nine young fledged from eight successful breeding attempts for a productivity of
only 0.50 young per breeding pair. In other studies carried out in different European regions, the productivity was
twice as high averaging 1.00 {N = 42) young per breeding pair in Provence (Bergier 1985, World Working Group for
Birds of Prey, Bull. 2:77-78), 1.10 {N = 117) in the French Pyrenees (Braillon 1979, Le Percnoptere dans les Pyrennes
francais, La grande faune pyreneenne et des montagnes d’Europe, Universite de pau et F.I.E.P., Pau, France) and
0.81 {N = 117) in Navarra (Donazar and Ceballos 1988, Ardeola 35:3-14).

The theft of eggs and young from nests as well as poisoning and direct disturbances from people near nesting sites
are probably factors causing the low reproductive rate of this population. The increasing importance of tourism on
the island has caused the gradual abandonment of the traditional economy based on livestock grazing and agriculture
and the opening of new roads and the growth of urban areas. Many vultures also die each year because of collisions
or electrocution from overhead cables. A study carried out by Lorenzo (1995, Ecologia 9:403-407) found six dead
adult and juvenile Egyptian Vultures along 45.3 km of powerlines.

The Department of Territorial Policy and Environment of the Cabildo of Fuerteventura funded this study and
personnel helped with the fieldwork. We also thank Antonio Palacios, Domingo Trujillo and Ruben Barone. Miguel
Fernandez del Castillo and Tony Clarke translated this paper into English. The contribution of data made by Domingo
Concepcion and Eduardo Castilla has been very useful. Jose Antonio Donazar guided the study and reviewed the
draft, making useful suggestions. — Cesar-Javier Palacios, CaUe Valencia, l-2*-J, 35600 Puerto del Rosario, Fuerteven-
tura, Spain.

Commentary

J. Raptor Res. 34(l):62-63
© 2000 The Raptor Research Foundation, Inc.

How Bearded Vultures ( Gypaetus barbatus) Acquire Their Orange Coloration;

A Comment on Xirouchakis (1998)

Juan Jose Negro

Estacion Biologica de Donana, Department of Applied Biology, Spanish Council for Research (CSIC), Apdo. 1056,

41080 Sevilla, Spain

Antoni Margalida

Group of Study and Protection of Bearded Vulture (GEPT), Apdo. 43, 25520 El Pont de Suert (Lleida), Spain

Bearded Vultures {Gypaetus barbatus) are unique
among raptors in that they have cosmetic coloration in
their plumage. The typical bright orange coloration of
adults is due to iron-oxide particles that coat the other-
wise pure white feathers in the head, neck and ventral
parts (Cramp and Simmons 1980, del Hoyo et al. 1994).
The initial debate on the origin of the pigmentation of
Bearded Vultures was settled long ago by conducting
chemical and physical analyses of the feathers (Berthold
1967, Brown and Bruton 1991), but one fundamental
question remained. Do Bearded Vultures actively stain
their feathers by seeking sources of iron oxide, or does
the plumage accidentally acquire its rufous color from
the rocks where the birds roost or from the water in
which they bathe? In other birds where cosmetic colors
have been reported from soils, it is possible that the pig-
ments are acquired passively (Kennard 1918). Among wa-
terfowl, for instance, stained individuals are common in
species that dig for their food among the iron-rich mud
in shallow waters, while those species that are unstained
usually forage in deeper waters or in dry fields.

Until recently, it was thought that the adventitious pig-
mentation of Bearded Vultures was passively acquired
from stained rock ledges because no birds had ever been
seen bathing in red soils or rust-colored springs. Brown
and Bruton (1991) contradicted their own hypothesis of
accidental coloring which should have no adaptive ad-
vantage and proposed a functional hypothesis, based on
camouflage and feather protection against abrasion or
mallophaga, to explain the coating of plumage with iron
oxides. In a recent letter to this journal, Xirouchakis
(1998) provided a detailed account on how a Bearded
Vulture seemed to produce red dust for bathing by peck-
ing on an eroded limestone rock (terra rosa), but again
contradicted himself concluding that feather coloration
IS passively acquired.

We believe there is now enough evidence to support
the hypothesis that Bearded Vultures stain their plumage

deliberately. Particularly imposing is the body of work
conducted in Austria by Frey and Roth-Callies (1994).
They offered damp soils of different colors and prove-
nances to 33 caged birds of different ages, as well as 12
captive-raised individuals at a release site in the Alps. Ex-
cept for one blind captive bird that remained pure white,
all birds always preferred damp red soils containing iron
oxides when they were offered with other mud types. All
individuals showed an active coloration behavior consist-
ing of a species-specific bathing display that led to the
characteristic rufous coloration. Bathing in damp red
soils was a much more elaborate behavior than regular
bathing in clean water that is commonly observed in cap-
tivity (Frey and Roth-Callies 1994) and in the wild (Brown
and Bruton 1991). Water bathing, on the other hand,
seems to be a common behavior in most vulture species
that often gather to wash and loaf (Mundy et al. 1992).

Baths in red damp soil by Bearded Vultures lasted for
up to one hour. First, the birds crouched on the red
mud, then spread the damp mud that was adhered in
their undersides to their shoulders and upper backs with
their beaks and talons. Finally, the head was swung re-
peatedly between the head and shoulders. This type of
behavior has even been filmed for the documentary
“The Lammergeier is back,” directed by Michel Terrasse.
Bathing in red soils seems to be an innate behavior, as
young captive birds raised in isolation showed this be-
havior as soon as they were presented with damp red
soils.

Recent observations in the wild, including Xirouchak-
is’s, also demonstrate that the cosmetic coloration of
Bearded Vultures is deliberate. An adult Bearded Vulture
was observed bathing in a rust-colored spring in the
French Pyrenees in 1995 (Gaussimont et al. 1995), in
what may be the first reported case of a Bearded Vulture
coloring itself in the wild. A similar observation has been
recorded at a Bearded Vulture breeding territory in the
Spanish Pyrenees. In January 1998, one adult Bearded

62

March 2000

Commentary

63

Vulture was seen landing on the ground, near the bottom
of a cliff. It walked toward the cliff bottom and went out
of sight, but soon reappeared with its breast wet and dirty.
It went there once more, and still moved to a second spot
located a few meters away, coming out with its breast
soaking wet. When these places were inspected, the first
one had a rusty soil dampened by a slight trickle of water,
and the second had a small (<I m^), transparent pool
with iron oxides at the bottom. There were orange
splashes in the surrounding snow and some feathers that
were collected. We visited this place three more times in
the following months (February, April and October) and
found new feathers each time, indicating that Bearded
Vultures had later returned to bathe. These observations
suggest that color intensity, and thus the great individual
variation observed in some populations, such as the one
m Crete (Xirouchakis 1998), southern Africa (Brown and
Bruton 1991) and the population that we are studying in
the Spanish Pyrenees, is at least partly related to the fre-
quency of baths and/or chromatic variations that are in
different substrates such as in rust-colored springs. Suit-
able sources of iron oxides for plumage staining may be
limiting for Bearded Vultures (Houston et al. 1993), and
thus not all individuals in a population would have ready
access to color sources, particularly if they are within the
defended territories of conspecifics.

Despite many hours of systematic observations, includ-
ing the tracking of radio-marked birds in southern Africa,
the Alps and Pyrenees, the only reported observations of
Bearded Vultures bathing to color their feathers are the
ones that we mention. As suggested by Houston et al.
(1993) , coloration of feathers by bathing in wild Bearded
Vultures is a very secretive behavior, quite different from
the often communal and conspicuous bathing of other
vulture species (Mundy et al. 1992). Nevertheless, it is a
sophisticated behavior. The next question that needs to
be explored is why do Bearded Vultures coat their plum-
age to such an extent all across their breeding range in
Eurasia and Africa. Now that solid evidence of deliberate
staining of the feathers in this species is accumulating, it
IS legitimate to look for an adaptive explanation.

Acknowledgments

This work was supported by Fundacio Territori i Pait-

sage, and grant PB97-1264 from the Spanish Ministry of

Education and Culture.

Literature Cited

Berthold, P. 1967. liber Haftfarben bei Vogeln: Rostfar-
bung durch Eisenoxid beim Bartgeier {Gypaetus bar-
batus) un bei anderen Arten. Zool. Jahrb. Syst. 93:507-
595.

Brown, C.J. and A.G. Bruton. 1991. Plumage color and
feather structure of the Bearded Vulture {Gypaetus
barbatus).]. Zool. Lond. 223:627-640.

Caussimont, G., M. Hunot and P. Mariette. 1995. A
bathing Lammergeier. Page 53 in H. Frey, J. Kurzweil
and M. Bijleveld [Eds.], Bearded Vulture. Annual re-
port 1998. Foundation for the Conservation of the
Bearded Vulture, Wassenaar, Netherlands.

Cramp, S. and K.E.L. Simmons [Eds.]. 1980. Handbook
of the birds of the western Palearctic. Vol. 2. Oxford
Univ. 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 Editions, Barce-
lona, Spain.

Frey, H. and N. Roth-Callies. 1994. Zur Genese der
Haftfarbe (Rostfarbung durch Eisenoxid) beim Bart-
geier, Gypaetus barbatus. Egretta 37:1—22.

Houston, D.C., A. Hall and H. Frey. 1993. The char-
acteristics of the cosmetic soils used by Bearded Vul-
tures Gypaetus barbatus. Bull. Br. Omithol. 113:260-
263.

Kennard, F.H. 1918. Ferruginous stains on waterfowl.
Auk 85:123-132.

Mundy, R, D. Butchart, J. Ledger and S. Piper. 1992.
The vultures of Africa. Academic Press, London, U.K

Xirouchakis, S. 1998. Dust bathing in the Bearded Vul-
ture {Gypaetus barbatus) . J. Raptor Res. 32:322.

Received 30 July 1999; accepted 7 October 1999

J. Raptor Res. 34 ( 1 ) ; 64

© 2000 The Raptor Research Foundation, Inc.

Manuscript Referees

The following people reviewed manuscripts for the Journal of Raptor Research in 1999. Peer review plays a vital role
in the publishing process and in improving the quality of the Journal. The names of individuals who reviewed two or
more manuscripts are followed with an asterisk.

David E. Andersen, Beatriz Arroyo*, James C. Bednarz, Maria Isabel Bellocq, Keith L, Bildstein*, Richard O
Bierregaard, Jr., Peter H. Bloom*, Maria Susana Bo, Eugene S. Botelho, Thomas Bosakowski*, Vincent Bretagnolle,
Joseph B. Buchanan, Javier Bustamante*, TomJ. Cade*, Richard J. Cannings, William M. Clark, Jack Clinton-Eitniear,
Gary E. Duke, Katharine Duffy, Wade L. Eakle, James H. Enderson*, Anthony J. Erskine, Eric D. Forsman*, David K.
Garcelon, Frederick R. Gehlbach*, James A. Gessaman, Laurie J. Goodrich, Teryl Grubb, Ralph J. Gutierrez, Susan
M. Haig, Linnea S. Hall, J. Christopher Haney, Alan R. Harmata*, C. Stuart Houston*, David Houston, W. Grainger
Hunt*, Ronald E. Jackman, Jerome A. Jackson*, Andrew Jenkins*, Alan C. Kemp, Michael Kochert, Donald E. Kroods-
ma, Robert N. Lehman, Johnathan L. Longmire, Scott Lutz, Kateryna Makova, Santi C. Manosa, Jeffrey S. Marks*,
Carl D. Marti*, John M. Marzluff, Michael D. McCrary, Michael McGrady*, Carol L. McIntyre*, Heimo Mikkola, Brian
A. Millsap*, Joan L. Morrison, Helmut C. Mueller*, Peter Mundy*, Juan Jose Negro*, Ian Newton*, Massimo Pan-
dolfi*, Vincenzo Penteriani*, Steve Petty, Charles R. Preston*, Steve Redpath*, Gary Ritchison, Ricardo Rodriguez-
Estrella, Robert N. Rosenfield, Josef Schmutz, Terry A. Schultz, Fabrizio Sergio, Dwight Smith*, Karen Steenhof*,
William E. Stout, Theodor Swem, Ethan J. Temeles*, Jean-Marc Thiollay*, Russell Thorstrum, Kimberly Titus*, Ale-
jandro Travaini*, Ian G. Warkentin, Richard Watson, David Whitacre, Clayton M. White, Sanford R. Wilbur*, James
W. Wiley, Neil D. Woffinden, Petra Bohall Wood, Stavros Xirouchakis, Reuven Yosef*.

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. E-mail: alien®
buteobooks.com

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.

Great Horned Owl (372). C. Stuart Houston, Dwight G. Smith, and Christoph Rohner. 1998. 28 pp.

Gyrfalcon (114). Nancy J. Glum 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 1. Warkentin. 1993. 20 pp.

Mississippi Kite (402). James W. Parker. 1999. 28 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.

Northern Hawk Owl (356). James R. Duncan and Patricia A. Duncan. 1998. 28 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). PW. 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.

Turkey Vulture (339). David A, Kirk and Michael J. Mossman. 1998. 32 pp.

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

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

Buteo Books stocks all published species accounts, not only those covering raptors. The current list in taxo-
nomic order may be viewed at: http:/ /www. buteobooks.com

Buteo Books stocks the Handbook of the Birds of the World. The first five volumes of this projected 12-volume
work have been published including: Volume 2: New World Vultures to Guineafowl (1994) covering the
diurnal raptors and Volume 5: Barn Owls to Hummingbirds (1999) covering owls. These volumes are priced
at $185 each plus shipping and handling.

Usually available from Buteo Books, the classic reference on diurnal birds of prey: Brown, Leslie and Dean
Amadon. Eagles, Hawks and Falcons of the World. Country Life Books, 1968. Two volumes. First English
edition in brown cloth. Fine in slipcase. $300.00 and other editions at lesser prices.

2000 ANNUAL MEETING

The Raptor Research Foundation, Inc. 2000 annual meeting will be held on 8-11 November in
Jonesboro, Arkansas, U.S.A. For information about the meeting contact James C. Bednarz, Depart-
ment of Biological Sciences, Arkansas State University, RO. Box 599, State University, AR 72467
U.S.A. Telephone 501-972-3082, E-mailjbednarz@navajo.astate.edu.

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
1 17th Street South, Hastings, Minnesota 55033, U.S.A. (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.S.A.
Printed by Allen Press, Inc., Lawrence, Kansas, U.S.A.

Copyright 2000 by The Raptor Research Foundation, Inc. Printed in U.S.A.

© This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).

Raptor Research Foundation, Inc., Awards
Recognition for Significant Contributions^

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 WIDB, Department of Zoology,
Brigham Yoimg University, Provo, UT 84602 U.S.A. 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.S.A. 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 FolweU 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: Patricia A. HaU, 5937 E. Abbey
Road, Flagstaff, AZ 86004 U.S.A.

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.S.A. Deadline: Deadline established for meeting paper

abstracts.

Grants^

The Stephen R. TuUy 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.S.A. 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, 1220 Rosecrans St. #315, San Diego, CA 92106 U.S.A. Deadline: September 15.

^Nominations should include: (1) the name, title 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.

^Send 5 copies of a proposal (s5 pages) describing the applicant’s background, study goals and methods,
anticipated budget, and other funding.