Feral pigs facilitate hyperpredation by golden eagles and
indirectly cause the decline of the island fox
Gary W. Roemer1, Timothy J. Coonan2, David K. Garcelon3, Jordi Bascompte4 and Lyndal Laughrin5
1
2
3
4
5
Department of Organismic Biology, Ecology and Evolution, University of California, Los Angeles, California 90095, USA
Channel Islands National Park, 1901 Spinnaker Drive, Ventura, CA 93003, USA
Institute for Wildlife Studies, PO Box 1104, Arcata, CA 95518, USA
Estación Biológica de Doñana, CSIC Apdo. 1056, E-41080 Sevilla, Spain
Santa Cruz Island Reserve, Marine Science Institute, University of California, Santa Barbara, CA 93106, USA
Abstract
Introduced species can compete with, prey upon or transmit disease to native forms, resulting in devastation of indigenous communities. A more subtle but equally severe effect of exotic species is as a
supplemental food source for predators that allows them to increase in abundance and then overexploit
native prey species. Here we show that the introduction of feral pigs (Sus scrofa) to the California
Channel Islands has sustained an unnaturally large breeding population of golden eagles (Aquila
chrysaetos), a native predator. The resulting increase in predation on the island fox (Urocyon littoralis) has caused the near extirpation of three subspecies of this endemic carnivore. Foxes evolved
on the islands over the past 20,000 years, pigs were introduced in the 1850s and golden eagles, historically, were only transient visitors. Although these three species have been sympatric for the past
150 years, this predator–prey interaction is a recent phenomenon, occurring within the last decade.
We hypothesize that this interaction ultimately stems from human-induced perturbations to the island,
mainland and surrounding marine environments.
INTRODUCTION
After habitat loss, biological invasions are now considered to be the most destructive force driving the loss of
global biodiversity (Vitousek et al., 1997). Invasions of
islands have been particularly detrimental, causing high
rates of extinction of indigenous fauna world-wide
(Atkinson, 1989). Theory predicts species turnover rates
to be naturally higher on nearshore islands because of
higher colonization rates by a continental biota that
is more closely linked (MacArthur & Wilson, 1967).
Anthropogenic impacts to continental ecosystems could
also influence nearshore islands, causing a higher rate of
species extinctions than expected under more natural conditions. Consequently, the biota of continental islands
may actually face a more diverse array of threats than
does the biota of oceanic islands because of regional connectivity with a continental ecosystem that is being continually disrupted by anthropogenic forces.
Here we describe an unexpected tri-tropic interaction
on a continental archipelago that appears to be driving
All correspondence to: Gary W. Roemer, Department of Fishery
and Wildlife Sciences, New Mexico State University, Las Cruces,
New Mexico 88003-8003, USA. E-mail: groemer@nmsu.edu
an insular carnivore towards extinction. We provide evidence that the presence of feral pigs (Sus scrofa) has
facilitated hyperpredation by golden eagles (Aquila
chrysaetos) and indirectly contributed to the decline of
three subspecies of the island fox (Urocyon littoralis), a
carnivore endemic to six of the eight California Channel
Islands. A form of apparent competition (Holt &
Lawton, 1994; Holt, 1997), hyperpredation occurs when
an introduced prey, well adapted to high predation pressure, indirectly facilitates the extinction of a native prey
by enabling a shared predator to increase in population
size (Courchamp, Langlais & Sugihara, 1999). We suggest that feral pigs acted as an abundant food that
enabled mainland golden eagles to colonize the Channel
Islands, increase in population size and overexploit the
unwary island fox. The increase in predation pressure
caused rapid population declines in three subspecific
populations of the fox. This premise is supported by: (1)
comparative demographic data that show that fox populations were at high density prior to eagle colonization
and declined thereafter; (2) a decrease in the survival of
foxes that coincided with an increase in eagle presence
on the islands; (3) physical evidence amassed at 28 fox
carcasses that shows that predation by eagles was the
principal cause of fox mortality; (4) the lack of other
potential mortality agents including micro- and
macroparasites; and (5) a mechanistic model that links
the pig population to the decline in foxes. We further
hypothesize that the ultimate cause of this interaction
stems from historic, human-induced perturbations to the
islands, to the mainland and to the surrounding marine
environments that collectively contributed to the decline
of the endemic island fox.
Table 1. Estimates of the parameters used to calculate time to
extinction, Te(n0), for each of three island fox populations found on
San Clemente, San Miguel and Santa Cruz Islands, California
METHODS
1
Study area
Island foxes are found on the six largest of the eight
California Channel Islands located off the coast of
southern California (Moore & Collins, 1995). Basically
a westward extension of the Santa Monica Mountains,
the northern Channel Islands are clustered together and
are separated from the mainland (30 to 44 km) by the
Santa Barbara Channel. The three southern islands lie
at greater distances from the mainland (32 to 98 km),
and are separated from each other by deepwater
channels.
The California Channel Islands are characterized by
a Mediterranean climate, with warm, dry summers and
mild, wet winters. Differences in microclimate among
islands and sites can be substantial. The prevailing northwest winds exert a cool, wet marine influence on northwest coasts, and the lower temperate zone location can
result in sunny, dry southern slopes. Annual precipitation is variable, owing to the influence of the El Niño–
Southern Oscillation phenomenon, that brings occasional
periods of heavy winter rains, or alternately, periods of
drought. Mean annual precipitation at Stanton Ranch,
Santa Cruz Island, for 1904–93 was 501 mm (Junak et
al., 1995). The southern islands are typically drier; mean
annual precipitation on San Clemente Island is 160 mm
(Kasaty, 1978, cf. Keegan, Coblentz & Winchell, 1994).
Population monitoring and time to extinction
We annually censused fox populations on large trapping
grids on two northern islands, San Miguel and Santa
Cruz (1993–99), and on two southern islands, Santa
Catalina (1989 and 1990) and San Clemente (1988–97)
using a capture–mark–recapture approach (Roemer,
1999; Coonan et al., 2000). Population size was estimated by first assessing fox density on the trapping grids
using program CAPTURE (White et al., 1982) and then
multiplying average density times island area (Roemer
et al., 1994). Foxes were also trapped along road transects on all islands in 1998 to compare nightly capture
success [(# of foxes captured/# of available traps) ×
100%].
We estimated the time to extinction [Te(n0)] and the
probability of population persistence for three fox populations with adequate data using the approach of Foley
(1994) (Table 1).
Parameter
Te (n0 )1
K
N0
rd
vr
vre
ρ
San Clemente
San Miguel
Santa Cruz
381
1264
682
–0.017
0.027
0.018
0.463
5
577
17
–0.562
0.314
0.443
0.17
13
1540
232
–0.433
0.037
0.037
0
Te(n0), in years was estimated with the following equation (Foley, 1994):
Te ( n 0 ) =
1
e 2sk 1 − e −2sn − 2sn 0
2srd
( (
)
)
where n = log N, k = log K, r = log R and N, K and R are population size, carrying capacity and growth rate, respectively. n0 is calculated from the last estimate of
population size for each subspecies (N0), K is estimated by multiplying the largest
density recorded for a subspecific population times the total area of the island where
the subspecies is found and rd is estimated as the mean of rt. s = rd/vre, where vre =
vr(1 + ρ)/(1 – ρ), ρ being the Pearson correlation between rt and rt+1. For details see
Roemer (1999).
Vital rate estimation and identification of mortality
agents
Annual survival probabilities were estimated using the
program MARK (White & Burnham, 1997). Island foxes
are extremely easy to capture, so we assumed that trapping returns would be representative of the sex and age
class distribution of the population at large (Crooks,
1994; Roemer et al., 1994). Fox populations were
censused on each trapping grid following a birth-pulse
and the total number of pups captured during each census was considered to be the reproductive output for that
year. Because island foxes are territorial (Roemer et al.,
2001), we estimated fertility by assigning pups to
particular females based on their respective capture
locations, on the female’s reproductive condition (i.e.
lactating) and on whether females and pups were captured in the same trap simultaneously (Garcelon et al.,
1999). Where available, paternity assessments made
through molecular genetic analyses were also used to
estimate fertility (Roemer et al., 2001).
We used the estimates of survival and fertility to parameterize a Lefkovitch matrix, a projection model that
can be used to predict changes in population growth over
time (Lefkovitch, 1965), for each of three island fox
populations (Table 2). Survival and fertility data were
collected for five definable age classes: Age class 0
(pups), Age class 1 (juveniles) and Age classes 2–4
(young adults to old adults). Each matrix, however, was
reduced to a two-sex, stage-based model. That is, both
sexes and all adult age classes were combined, yielding
three stages: pups, juveniles and adults (Table 3). The
matrices were reduced because survival did not differ
between the sexes, nor did it differ among the adult age
classes, but it was significantly different among pups,
juveniles and adults combined (Roemer, 1999). Further,
island foxes are predominantly monogamous and fertility was significantly different between juveniles and
adults (Roemer, 1999; Roemer et al., 2001). We then
conducted an analytical sensitivity analysis of the vital
Table 2. Minimum, mean and maximum estimates of stage-specific vital rates (VR)1 for three island fox populations. Mean vital rates for
the San Miguel and Santa Cruz fox populations were derived from estimates measured prior to the population declines (1993 and 1994),
whereas for the San Clemente fox population these means were estimated from data collected across all years (1988–97). Sample sizes are
given in parentheses
VR
min
P0
P1
PA
f1
f2
fA
0.40
0.55
0.49
0
0
0
San Clemente
mean
max
0.53
0.71
0.69
0.13
0.30
0.37
1.00
0.87
0.80
0.35
0.50
0.89
min
(150)
(263)
(521)
(105)
(172)
(265)
San Miguel
mean
0
0
0.30
0
0
0
0.53
0.73
0.75
0.30
1.41
0.92
max
0.74
0.82
1.00
0.31
2.33
3.00
min
(64)
(73)
(93)
(32)
(21)
(17)
0
0.2
0.16
0
0
0
Santa Cruz
mean
0.45
0.92
0.69
0.84
1.08
1.38
max
0.75
1.00
0.94
1.00
1.16
3.00
(37)
(41)
(27)
(8)
(17)
(5)
1
Survival probabilities are for pups (P0; Age class 0), yearlings (P1; Age class 1) and adults (PA; Age classes 2–4). The fertilities are for yearlings (f1; Age class 1), young
adults (f2; Age class 2) and adults (fA; Age classes 3 and 4). Fertilities (Fx) used in all matrices were calculated as follows [Fx = (fx)(Px–1)].
Table 3. Lefkovitch matrices for each of three island fox populations: San Clemente, San Miguel and Santa Cruz. Fx = (fx)(Px–1),
where fx is the fertility of stage x and Px–1 is the survival probability of the preceding stage. See Table 2 for stage-specific estimates
of fertility and survival
growth rate; (3) to determine the vital rates that contributed most to the observed declines (Caswell, 1989;
Horvitz, Schemske & Caswell, 1997; Roemer, 1999).
We implemented two radiotelemetry studies to eluci-
Lefkovitch Matrix
F1
F2
FA
P0
0
0
0
P1
PA
San Clemente
0.07
0.21
0.53
0
0
0.71
0.26
0
0.69
San Miguel
0.16
1.03
0.53
0
0
0.73
Santa Cruz
0.38
0.99
0.45
0
0
0.92
0.95
0
0.69
Santa Cruz Island, 32 foxes were radio-collared and captured or located every 2 or 3 days from January to
December 1994 (Roemer, 1999). After December 1994,
foxes were located twice each month until the end of the
study in September 1995. On San Miguel Island, 15
foxes were radio-collared and monitored daily from
November 1998 to July 1999. Physical evidence at the
carcass sites and field necropsy were used to infer the
cause of death of 21 foxes on Santa Cruz. Seven foxes
found dead on San Miguel were necropsied at the
Veterinary Medical Teaching Hospital, University of
California, Davis. Survival of all radio-collared foxes
was estimated with the Kaplan–Meier procedure with
staggered entry (Pollock et al., 1989).
We serologically assayed for five potentially lethal
canine diseases when the fox populations were in decline
and compared seroprevalence with a previous assay conducted when all fox populations were at high density
(Garcelon, Wayne & Gonzales, 1992). We also serologically assayed for heartworm (Dirofilaria immitis), a
parasite with potentially lethal consequences to canids
(Strickland, 1998), and conducted a fecal parasite survey across the entire range of the fox in 1998. The disease and fecal parasite assays were conducted at the
Washington Animal Disease and Diagnostic Laboratory.
Serological assays for heartworm antigen were performed using a commercially available enzyme-linked
immunosorbent assay (PetChek®, Idexx Laboratories,
Westbrook, Maine, USA; Roemer et al., 2000).
0.69
0
0.75
rates composing the matrix elements and compared
elasticities among populations.
Elasticities represent the influence of proportional
changes in life stage parameters on population growth
rate, λ (Caswell, 1989). When elasticities are calculated
for the elements of a projection matrix they sum to one
(de Kroon et al., 1986). These elasticities represent the
proportional contribution of each matrix element to the
total elasticity of λ and provide valuable information
regarding the relative effect of different life stage parameters on population growth (Caswell, 1989). However,
vital rates are embedded throughout the matrix elements
of a projection matrix (Noon & Sauer, 1992). For
example, stage-specific fertilities (Fx) are the product of
the average reproductive output of that stage multiplied
by the transition probability from the previous stage [Fx
= (fx)(Px–1)]. Furthermore, reproductive output may be a
complex composite of many life stage parameters (Doak,
Kareiva & Klepetka, 1994; Wisdom & Mills, 1997). To
elucidate the effects of stage-specific vital rates on population growth, we calculated the elasticity of λ to the
vital rates themselves rather than to the elements of the
matrix. Although the elasticities of vital rates are calculated on the same proportional scale, and like the elasticities of matrix elements can be directly compared
(Wisdom & Mills, 1997), the sum of the elasticities of
the vital rates does not equal one (Caswell, 1989). The
objectives of the sensitivity analysis were: (1) to compare demography of three island fox populations, two that
had declined (San Miguel and Santa Cruz) with one that
had not (San Clemente); (2) to identify the vital rates predicted to have the greatest influence on fox population
Distribution of eagle sightings on the Channel Islands
We explored the temporal and spatial distribution of
eagle sightings on the Channel Islands using observations
of golden eagles collected over the past 35 years. These
data represent the absolute number of sightings reported
by knowledgeable individuals who were conducting
research or management activities on the islands unrelated to our study. These anecdotal sightings are not
biased by research specifically aimed at determining the
distribution and abundance of golden eagles which, if
included, could have falsely suggested an increase in
golden eagle presence on the islands when none may
have occurred. Thus we feel these sightings represent
actual changes in the presence of golden eagles over time.
Modelling the effects of predation by golden eagles
Golden eagles have been observed feeding on freshly
killed piglets (Roemer, 1999), have been implicated in
the mortality of foxes (Roemer, 1999; see below), and
both piglet and fox remains were found in a golden eagle
nest on Santa Cruz Island (see below). Because of these
observations, we investigated whether predation by
eagles alone could have caused the fox population
declines and whether the presence of feral pigs may have
facilitated predation on foxes by sustaining an unnaturally large eagle population. To explore this potential
tri-trophic interaction we estimated predation rates of
golden eagles from a time–energy budget and incorporated these estimates into a simple, but realistic predator–prey model of hyperpredation.
We first estimated the standard metabolic rate (SMR;
W kg–1) of a 4.0 kg golden eagle from environmental
data measured on Santa Cruz Island from June 1995
through July 1998 (Hayes & Gessaman, 1980):
SMR = 1.163 (6.1168 – 0.6 W – 0.0793 Ta – 0.955 ×
10–3 IR + 0.1284 U)
where W is body mass (kg), Ta is ambient temperature
(°C), IR is incident radiation (W m–2) and U is wind speed
(m s–1). These estimates were then incorporated into a
time–energy budget to determine daily food consumption
(DFC; g bird–1 day–1) (Stalmaster & Gessaman, 1984;
Collopy & Edwards, 1989). Wet-matter intake was 54.6
g kg–1 day–1 based on the consumption of captive golden
eagles (Fevold & Craighead, 1958) and wet-matter
energy of a mammal was assumed to be 6.03 Kj g–1
(Collopy, 1986). Average DFC was 461 g bird–1 day–1 (±
15) for an actively foraging 4.0 kg golden eagle resident
on Santa Cruz Island. This estimate compares favourably
with DFC estimates (range 250 to 570 g bird–1 day–1) for
other eagles, including golden eagles, where daily intake
was directly measured (Fevold & Craighead, 1958; Love,
1979; Stalmaster, 1987). Island foxes on Santa Cruz
Island average 1.93 kg (± 0.25, n = 279; G. W. Roemer,
unpublished data) and assuming 30% wastage (Brown &
Watson, 1963) a 4.0 kg golden eagle could consume
124.6 foxes yr–1 provided it was feeding exclusively on
foxes. A single breeding golden eagle would require an
additional 7.6 foxes based on adding the average energetic requirements of a nestling eagle (Collopy, 1986).
We then used these estimates of predation rate in the
following hyperpredation model (Courchamp et al., 1999):
dF
F
= rF F 1 −
KF
dt
dP
P
− = rP P 1 −
KP
dt
F
−
(µ E + µb Eb )
αP + F nb nb
αP
αP + F
(µ nbE nb + µb Eb )
Where dF/dt, dP/dt, rF, rP, F, P, KF and KP represent the
change in population size, intrinsic growth rate, initial
population size and carrying capacity of the fox and pig
populations, respectively. µnb, µb, Enb and Εb are the predation rates and number of non-breeding and breeding
eagles, respectively. Parameter estimates and initial conditions are given in Table 4. Because we had no data on
relative prey consumption by eagles, but we knew that
eagles ate both foxes and piglets, we modelled prey preference (α) of eagles for pigs by setting this value at 3,
1 or 0.33. In other words, if the relative abundance of
prey is equal, eagles kill three pigs for every fox when
α = 3 and kill three foxes for every pig when α = 0.33.
Fox populations grow in a single pulse in April of each
year. Pigs can produce piglets in any month of the year
but, on average, do not reproduce for 3 months each year
on the Channel Islands (Baber & Coblentz, 1986).
Therefore, we assumed that for 3 months each year
piglets were unavailable as a food source. The eagle population was modelled in a simple, step-wise fashion. We
started with two breeding eagles and every year they
produce a single offspring. The eagle population grows
to a total of seven eagles, two breeding and five nonbreeding individuals, which is the minimum number of
eagles observed on Santa Cruz Island in 1999. Because
we assume this is the maximum number of eagles the
island can support, and that only a single pair breeds,
our estimates are probably conservative.
RESULTS AND DISCUSSION
Fox population declines
In 1993 all subspecific populations varied in size from
several hundred to over 1000 foxes (Roemer et al.,
1994). Fox density on San Miguel and Santa Cruz averaged 7.1 (± 1.1) foxes/km2, capture success (mean =
25.7% ± 3.7) was high, and estimated population sizes
were 350 and 1312 adult foxes, respectively (Fig. 1(a)).
Table 4. Parameter estimates and initial conditions1 for the
hyperpredation model used to simulate the effects of predation by
golden eagles on two prey, island foxes and feral pigs, on Santa
Cruz Island, California.
Parameter
Value
Reference
rF
rP
µnb
µb
F
P
KF
KP
Enb
Eb
0.32
0.78
124.6 foxes/year
132.2 foxes/year
1312 foxes
3517 pigs
1544 foxes
5000 pigs
0 to 5
2
Coonan et al., 2000
Caley, 1993
See Methods
See Methods
Roemer, 1999
Sterner, 1990
See footnote
See footnote
See footnote
See footnote
1
rF was estimated from serial estimates of population size when the San Miguel fox
population showed positive growth. Predation rates (µnb and µb) were calculated from
energetic estimates of food consumption. F was the capture–recapture estimate of
fox population size for Santa Cruz Island in 1993 and P was a capture–recapture
estimate of the pig population size on Santa Cruz Island in 1988. KF was determined
by multiplying the highest recorded density of foxes (6.2 foxes/km2) on Santa Cruz
times island area (249 km2). KP was set to 5000. The number of non-breeding eagles
(Enb) was initially set to 0 and limited to 5. The number of breeding eagles (Eb) was
held constant at 2.
Fig. 1. (a) Temporal trend in population size and (b) the probablity of population persistence for each of three island fox populations: San Clemente (SCL), Santa Cruz (SCR) and San Miguel (SMI). The estimates of T e(n0) used to generate the population persistence probabilities are given in Table 1.
By 1998, mean density had dropped to 0.8 (± 1.0)
foxes/km2, capture success (mean = 4.3% ± 1.9) had
decreased sixfold and population size on both islands
had plummeted (Fig. 1(a)). Only 15 adults were known
to be alive on San Miguel in 1999, with an estimated
133 foxes remaining on Santa Cruz. Capture success in
1998 on nearby Santa Rosa Island was also low, suggesting that fox populations had declined on all three
northern Channel Islands (Table 5).
In contrast, fox populations on the southern islands
remained relatively stable. Mean density on Santa
Catalina and San Clemente averaged 5.5 foxes/km2
(± 2.0) and mean capture success in 1998 on all three
southern islands was nearly eight times greater than that
observed on the northern islands (Table 5). San
Clemente’s fox population has decreased over the past
10 years, but the decline has not been steep (Fig. 1(a)).
Table 5. Total number of trap nights, number of individuals
captured (N), total captures and capture success (%) in 1998 on all
six California Channel Islands harbouring fox populations
Island
Northern Channel Islands
San Miguel
Santa Rosa
Trap
nights
N
Total
captures
Capture
success
876
21
49
5.6
132
9
10
4.8
Santa Cruz
Southern Channel Islands
San Nicolas
San Clemente
Santa Catalina
756
17
22
2.9
80
80
76
33
26
20
33
29
20
41.3
36.3
26.3
The San Miguel and Santa Cruz fox populations are
estimated to have a 50% probability of persistence
within the next decade, whereas the San Clemente fox
population has an estimated 80% probability of persistence for the next 100 years (Fig. 1(b)). These data show
that the declines are currently restricted to the northern
Channel Island fox populations and that these three subspecies meet the criteria of ‘critically endangered’ (Mace
& Lande, 1991).
Vital rate estimates and demographic modelling of
the fox populations
Survival rather than fertility is predicted to have the
greatest influence on λ (Fig. 2(a)). From 1993 to 1997,
annual survival of foxes on San Clemente (mean = 0.67
± 0.14) did not differ (Mann–Whitney U-test, U = 8.0,
P = 0.62) from years prior to 1993 (mean = 0.64 ± 0.03),
but was higher (Wilcoxon matched-pairs test, Z = 1.85,
P = 0.07) than annual survival of foxes on the northern
islands (mean = 0.38 ± 0.23; Fig. 2(b)). Survival of foxes
of all stages declined precipitously over time on both
San Miguel and Santa Cruz Islands (Maximum
Likelihood test, P < 0.05, Fig. 2(b)). Average annual fertility was also low (range = 0.13 to 1.9 pups/female) in
all island fox populations (Roemer, 1999). Change
in fertility over time approached statistical significance
in the San Clemente (Kruskal–Wallis test, H = 16.2, d.f.
= 9, n = 40, P = 0.06) and the Santa Cruz (H = 10.6,
d.f. = 5, n = 20, P = 0.06) fox populations, but not in
the San Miguel fox population (H = 1.8, d.f. = 4, n =
16, P = 0.76). In general, survival had changed more
than fertility had during the period of decline for both
the San Miguel (P0 > PA > P1 > FA > F2 > F1) and Santa
Cruz (P0 > PA > FA > P1 > F1 > F2) populations. These
results show that the northern island fox populations
(a)
1
0.9
0.8
Elasticity
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
f1
f2
fA
P0
P1
PA
Vital rate
Fig. 2. (a). The elasticity of λ to different vital rates for the San Clemente (black bars), Santa Cruz (hatched bars) and San
Miguel (white bars) island fox populations. Stage-specific vital rates are defined as in Table 2. (b) Average (± SD) apparent
survival of foxes of all age classes for the same islands.
have experienced a simultaneous decline in survival
across all age classes that has precipitated the declines
in population size.
Mortality agents of radio-collared foxes
Survival of radio-collared foxes on Santa Cruz Island
dropped from 1.0 in the first 9 months of study to 0.21
(± 0.08) in the next 17 months (Fig. 3). A total of 21
fox carcasses were discovered. Evidence from the carcasses suggests golden eagles were the primary cause of
mortality. Typically eagles leave talon holes with contusions in the integument and surrounding tissue, they
frequently deglove limbs to expose muscle tissue, and
often eviscerate their prey. Talon holes were found in
13 (61.9%) carcasses. Seven carcasses (30%) had talon
holes with associated muscle hematomas, indicating the
animals were alive at the time the talon holes were made,
and the other six carcasses (28.6%) had holes in their
skulls. These latter wounds would have been fatal if
inflicted while the fox was alive. Of those carcasses that
had limbs (n = 17), 15 (88%) had been degloved and
all carcasses but one were eviscerated. Feeding eagles
damage fragile bones such as the vertebrae, ribs and
scapulae, and may break the diaphyses of limb bones
(Hockett, 1989, 1991). Ninety percent of the carcasses
exhibited damage to bones typical of a raptor feeding
pattern (Roemer, 1999). Golden eagle feathers were
found at six (28.5%) different carcasses. Finally, all
foxes died when investigators were absent from the
study site. This pattern would be expected if the mortalities were a result of a wary predator disturbed by
human presence (Isbell & Young, 1993; Roemer, 1999).
Cumulative survival of the radio-collared foxes on
San Miguel Island was also low (0.23 ± 0.08) and
necropsy evidence suggested that a large raptor killed
five of seven foxes found dead. Golden eagle feathers
were found at two of the carcass sites and fox mortality
was clustered in the fall when eagles were observed.
In sum, of 28 foxes found dead on two islands, 24
(86%) showed signs of being eaten by a large raptor, 18
(64%) showed signs of being killed by a large raptor,
and golden eagle feathers were found near eight (29%)
carcasses. Feathers from other species were not found at
any carcass site.
Disease and parasite surveys
There was no concordance between pathogen prevalence
and the geographical pattern of fox population declines.
Most notably, we did not detect canine distemper virus
in any of the fox populations assayed, and parvovirus
decreased between sampling periods (Table 6). Of the
Table 6. Seroprevalence (%) of microparasites assayed in six island
fox populations in 1988 (Garcelon et al., 1992)1 and three fox
populations assayed between 1994 and 1997 (this study). N is the
number of individuals sampled
Year/Island
1988
San Miguel
Santa Rosa
Santa Cruz
San Nicolas
San Clemente
Santa Catalina
1994–97
San Miguel
Santa Cruz
San Clemente
N
Pathogen2
CDV CPV CAV LVC LVI
23
34
29
46
42
20
0
0
0
0
0
0
30
35
59
7
50
5
96
97
0
72
88
0
0
0
0
0
0
0
0
0
14
0
0
0
57
50
50
0
0
0
4
0
0
95
58
76
0
0
0
0
2
2
1
Reproduced by permission from the Journal of Wildlife Diseases.
CDV = canine distemper virus, CPV = canine parvovirus, CAV = canine adenovirus, LVC = Leptospira interrogans serovar canicola, LVI = L. i. serovar icterohaemorrhagiae.
2
Fig. 3. Mean cumulative survival (thick line) and upper and lower 95% confidence intervals (thin lines) of adult radio-collared
foxes from August 1993 to September 1995 on the west end of Santa Cruz Island, California.
pathogens assayed, these two would be expected to have
the most devastating impact on island fox populations
(Nicholson & Hill, 1984; Williams et al., 1988; Garcelon
et al., 1992). Canine adenovirus, which primarily affects
pup survival, appears to be enzootic (Garcelon et al.,
1992). Conceivably, rabies could have caused the
declines, but no rabid or otherwise sick foxes were
observed during the study period, as might be expected
(Roelke-Parker et al., 1996; Sillero-Zubiri, King &
Macdonald, 1996).
We detected heartworm antigen in four of the six populations, including all three northern islands, and on one
southern island (San Nicolas) where the fox population
has not declined (Roemer et al., 2000). Overall seroprevalence among adults from the four populations that
tested positive was 72% (n = 124) in 1988 and 78%
(n = 80) in 1997–98; heartworm was not detected in any
of the pups (n = 33) assayed. Although heartworm may
have contributed to the mortality of older foxes, it was
not detected in pups, seroprevalence did not change over
time, and one apparently infected fox population did not
decline (Roemer et al., 2000).
We found little evidence of gastrointestinal parasites in
our fecal surveys (Table 7), and of the parasites discovered
none would be expected to cause high mortality in adults
(Bowman & Lynn, 1995). These results suggest that neither micro- nor macroparasites were a major factor in the
recent fox population declines on the northern islands
(Roemer, 1999; Roemer et al., 2000).
In conclusion, predation by golden eagles is the only
mortality factor that is consistent with the simultaneous
decline in fox populations on three islands spanning
85 km of ocean.
Eagle sightings and the recent colonization of the
northern Channel Islands
Golden eagle sightings have increased on the northern
islands where foxes are in decline, but have not changed
Table 7. Percent occurrence of different parasites, or their eggs, in
the faeces of island foxes collected from all six populations in July
and August 1998. N is the number of individuals sampled
N
ISO
Northern Channel Islands
33
San Miguel
12
Santa Rosa
4
100
Santa Cruz
25
20
Southern Channel Islands
San Nicolas
22
32
San Clemente 18
28
Santa Catalina 15
46
Pathogen1
ANC TOX MES SAR NEO
42
0
0
25
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5
0
0
0
6
20
9
0
0
1
ISO = Isospora sp., probably I. canis, ANC = Ancylostoma sp., TOX = Toxascaris
sp., MES = Mesocestoides sp., SAR = Sarcocystis sp., NEO = Neospora sp.
on the southern islands where fox populations are comparatively stable (χ2 = 22.1, P < 0.0001; Fig. 4). We
believe the recent increase in golden eagle sightings on
the northern islands and the difference in the number of
eagle sightings between the southern and northern
islands is a consequence of golden eagles colonizing
Santa Cruz Island sometime in the early 1990s.
Evidence that eagles were eating foxes first surfaced
in 1994 on Santa Cruz (Fig. 3). In November 1997 a
group of two adult golden eagles and two juveniles was
observed flying together on Santa Cruz. One of the adults
transferred a snake, in mid-flight, to a begging juvenile
that was presumed to be its offspring (C. Collins, pers.
comm.). This observation was later corroborated in
September 1999 when a large, recently active golden
eagle nest was discovered. Nest measurements and
eggshell analysis confirmed that it was a golden
eagle nest (B. Latta and S. Sumida, pers. comm.).
Observations of resident raptors have been recorded for
the Channel Islands since 1895 (Kiff, 1980) and the 1999
nest represents the first breeding record for golden
eagles. Nest remains included the bones of two island
foxes, five feral piglets and three bird species.
Concurrent helicopter and ground surveys identified 7 to
60
Number of eagle sightings
50
40
30
22
20
10
10
8
6
2
2
1
1
0
Pre–1970
1970–1979
1980–1989
1990–1994
0
1995–1999
Period
Fig. 4. Number of golden eagle sightings on the northern (white bars) and southern (black bars) Channel Islands from before
1970 to December 1999.
15 different golden eagles, with the possibility for four
breeding pairs on Santa Cruz Island (GWR & LL, pers.
obs; B. Latta, pers. comm.).
The hyperpredation model
The hyperpredation model predicts that in the presence
of a large pig population, an asymptotic population of
seven eagles could cause the extinction of the Santa Cruz
fox population in 6.7 to 11.5 years depending on prey
preference (Fig. 5). If foxes were the only prey item
available, the model predicts that five eagles could have
extirpated the fox population in only 4 years. Thus island
foxes alone could not have sustained an eagle population as large as that observed on Santa Cruz Island. Pigs,
or some other abundant food, would have been necessary to explain both the number of eagles observed and
their duration on the island.
Because there are no pigs on San Miguel, we used a
single predator–single prey model to explore eagle–fox
dynamics on this island. This model suggested that a
single eagle could have caused the extinction of the San
Miguel fox population in 6.2 years. This population was
driven to near extirpation in 6 years (Fig. 1(a)). There
was no evidence that golden eagles were breeding on
San Miguel and given that San Miguel is much smaller
than Santa Cruz (38 km2 vs. 249 km2) it is likely that
breeding eagles would have been detected. Eagles were
observed on San Miguel only in the fall, when juvenile
eagles would be dispersing, and foxes were killed in the
fall during periods when eagles were sighted. These
observations support the premise that foxes alone could
not support a breeding eagle population and that some
other abundant food, like pigs, was necessary. The model
results are consistent with eagles being the sole cause of
the near extinction of the San Miguel fox population
whereas pigs are necessary to explain why eagles could
reproduce on Santa Cruz Island and subsequently cause
the decline of all northern island fox populations.
Although there are alternative prey species on Santa
Cruz (e.g., seabirds) these species are only seasonally
abundant. Feral pigs are the only prey on Santa Cruz
that represents a year-round food source for eagles that
would also be abundant during the eagle reproductive
season.
Our interdisciplinary approach supports the premise
that hyperpredation on foxes by eagles was facilitated
by the presence of feral pigs and caused the decline in
foxes on the northern Channel Islands. A question that
remains unanswered is whether eagles were regularly
8
5000
Eagle
4500
7
4000
Pig
6
5
3000
4
2500
2000
3
Number of eagles
Numbers of foxes or pigs
3500
1500
2
1000
Fox
1
500
0
0
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Year
Fig. 5. Trend in the fox, pig and eagle populations on Santa Cruz Island predicted from the hyperpredation model. Our time
unit is a day and we plotted population size every 90 days. The regular peaks in fox population size are due to modelling
growth as a single pulse each year. Ratios (pigs:foxes) of predator preference (α) for the prey are 3, 1 and 0.33. Time to extinction for the fox populations given these preferences was 11.5 years, 8.7 years, and 6.7 years, respectively.
commuting from the mainland, in a sense spilling over
from a larger region with all sorts of prey (Holt, 1984),
or whether most of the eagles were resident on the island.
The discovery of the first golden eagle nest on Santa
Cruz confirms that at least some eagles were breeding
there, but because golden eagles historically never bred
on the islands (Kiff, 1980) they must have come from
the nearby mainland. Despite this fact, the number of
eagles on Santa Cruz Island alone was considerable.
From November 1999 to June 2000 a total of 13 golden
eagles were live-captured and translocated from Santa
Cruz to distant locales as part of a conservation strategy
initiated by Channel Islands National Park (Coonan,
2001). In August 2000 one of us (GWR) observed three
golden eagles, two adults and one subadult, in territorial display on the western end of Santa Cruz Island. Thus
within 10 months’ time, during a period when the fox
population was extremely low (Fig. 1(a)), 16 golden
eagles had resided on Santa Cruz Island, more than twice
as many eagles as we modelled (n = 7; Fig. 5). It seems
likely that eagles are moving from the mainland to the
islands and those that colonize the islands are moving
among them.
Island foxes evolved on the Channel Islands over the
past 20,000 years (Wayne et al., 1991), pigs were introduced to the islands in the 1850s (Junak et al., 1995)
and, at least within historic times, golden eagles were
only transient visitors (Fig. 4). If these three species have
existed in sympatry for the last 150 years why has this
interaction occurred now? We hypothesize that the current perturbation is unique and due to historical anthropogenic factors that degraded the health of the island,
mainland and marine ecosystems. This complex interaction ultimately increased the vulnerability of island
foxes to extinction.
Island foxes are extremely docile and unlike most
canids are active during the day (Laughrin, 1977; Crooks
& Van Vuren, 1995). As with other island species that
have been decimated by novel predators (Diamond,
1989), island foxes evolved without high predation
pressure and therefore are most probably not vigilant
towards avian predators. This lack of predator vigilance
may have been exacerbated by a reduction in vegetative
cover on the islands resulting from 150 years of
European agricultural practices that included overgrazing by livestock and feral herbivores (Minnich, 1980;
Power, 1980; Van Vuren & Coblentz, 1987; Junak et
al., 1995). This reduction in vegetative cover probably
made island foxes especially vulnerable to avian
predation.
Environmental degradation of the marine environment
also impacted the islands. Pesticide manufacturers based
in Los Angeles released large amounts of organochlorine contaminants into the nearshore environment off the
Palos Verdes peninsula in southern California (Stull,
Swift & Niedoroda, 1996). This source of contaminants
contributed to the extirpation of the bald eagle
(Haliaeetus leucocephalus) on the Channel Islands by
the late 1950s (Kiff, 1980) and its influence continues
today. Bald eagles reintroduced to Santa Catalina Island
cannot reproduce without human intervention and some
adult birds have even died as a direct result of
organochlorine poisoning (Garcelon et al., 1989;
Garcelon & Thomas, 1997). Bald eagles are primarily
piscivorous and mostly forage over aquatic habitats
(Stalmaster, 1987), thus they are not expected to be a
significant predator of the island fox. Moreover, bald
eagles are territorial (Stalmaster, 1987), are aggressive
towards conspecifics and other raptors (Garcelon, 1990;
Perkins, Phillips & Garcelon, 1996) and would probably compete with golden eagles for food (Halley &
Gjershaug, 1998) and nest sites. With at least 24 breeding pairs of bald eagles previously resident on the
Channel Islands (Kiff, 1980) it is highly probable that
bald eagles acted as a deterrent to golden eagles and that
their extirpation paved the way for colonization of the
islands by golden eagles.
Subsequent to the extirpation of the bald eagle on the
Channel Islands, protection of the golden eagle was conferred in 1962 by an amendment to the Bald Eagle
Protection Act (US Code Service, 1999). Golden eagle
populations are now increasing or are stable across most
of the western US, except in coastal southern California
where increasing urbanization has reduced golden eagle
habitat (Harlow & Bloom, 1989). Increasing numbers of
golden eagles in the western US, together with increasing encroachment in southern California, may have displaced mainland eagles to the Channel Islands.
CONCLUSIONS
Although the Channel Islands were never physically
connected to North America (Junger & Johnson, 1980),
the islands and the mainland are inextricably linked.
Anthropogenic impacts to the islands, the mainland and
the surrounding marine environment appear to have contributed to the current conservation crisis on the islands.
As human pressure reinforces the connectivity between
ecosystems, novel impacts will continue to arise, disrupting ecosystem processes and causing new conservation concerns (Estes et al., 1998). If resource managers
are to avert the impending extinctions of the fox and
enhance the ecological integrity of the Channel Islands
a regional approach to conservation will probably be a
necessity (Soulé & Terborgh, 1999). This approach will
minimally need to include golden eagle translocation,
feral pig eradication, bald eagle reintroduction and the
captive propagation and release of island foxes (Coonan,
2001).
Acknowledgements
We thank all of the dedicated field assistants, especially
J. Howarth and K. Rutz. B. A. Sabo and P. Trail
identified the feathers found at fox carcasses and
L. Munson performed the necropsies on fox carcasses
from San Miguel. S. Spaulding and R. Lauston found
the eagle nest and B. Latta led its exploration. T. Boyle
provided the environmental data for Santa Cruz Island.
We thank P. Collins for suggesting the potential inter-
action between bald eagles and golden eagles. The
National Geographic Society, Canon Inc., the National
Park Service and the Institute for Wildlife Studies provided funding. At the time of this work, J.B. was a postdoctoral research fellow at the National Center for
Ecological Analysis and Synthesis, a center funded by
NSF (Grant DEB-94-21535), the University of
California at Santa Barbara and the State of California.
J. A. Estes, M. Gompper, R. D. Holt, A. J. MacInnis,
C. Vilà and an anonymous reviewer improved earlier
versions of the manuscript. G.W.R. wishes to thank
R. K. Wayne for his support and guidance throughout
this research and in the genesis of the final manuscript.
REFERENCES
Atkinson, I. (1989). Introduced animals and extinctions. In
Conservation for the twenty-first century: 54–75. Western, D. &
Pearl, M. C. (Eds). New York: Oxford University Press.
Baber, D. W. & Coblentz, B. E. (1986). Density, home range,
habitat use, and reproduction in feral pigs on Santa Catalina
Island. J. Mammal. 67: 512–525.
Bowman, D. D. & Lynn, R. C. (1995). Parasitology for veterinarians (6th edn). Philadelphia: W. B. Saunders.
Brown, L. H. & Watson, A. (1963). The golden eagle in relation
to its food supply. Ibis 106: 78–99.
Caley, P. (1993). Population dynamics of feral pigs (Sus
scrofa) in a tropical riverine habitat complex. Wildl. Res. 20:
625–636.
Caswell, H. (1989). Matrix population models: construction,
analysis and interpretation. Sunderland, MA: Sinauer
Associates.
Collopy, M. (1986). Food consumption and growth energetics of
nestling golden eagles. Wilson Bull. 98: 445–458.
Collopy, M. W. & Edwards, Jr., T. C. (1989). Territory size, activity budget, and role of undulating flight in nesting golden eagles.
J. Field Ornithol. 60: 43–51.
Coonan, T. J. (2001). Recovery plan for island foxes (Urocyon
littoralis) on the northern Channel Islands. Ventura, CA:
National Park Service, Channel Islands National Park.
Coonan, T. J., Schwemm, C. A., Roemer, G. W. & Austin, G.
(2000). Population decline of island foxes (Urocyon littoralis)
on San Miguel Island. In Proceedings of the Fifth Channel
Islands Symposium: 289–297. Browne, D. R., Mitchell, K. L. &
Chaney, H. W. (Eds). Camarillo, CA: US Department of the
Interior, Minerals Management Service.
Courchamp, F., Langlais, M. & Sugihara, G. (1999). Control of
rabbits to protect island birds from cat predation. Biol. Conserv.
89: 219–225.
Crooks, K. (1994). Demography and status of the island fox and
island spotted skunk on Santa Cruz Island, California.
Southwest. Nat. 39: 257–262.
Crooks, K. R. & Van Vuren, D. (1995). Resource utilization by
two insular endemic mammalian carnivores, the island fox and
island spotted skunk. Oecologia 104: 301–307.
de Kroon, H., Plaisier, A., van Groenendael, J. & Caswell, H.
(1986). Elasticity: the relative contribution of demographic parameters to population growth rate. Ecology 67: 1427–1431.
Diamond, J. (1989). Overview of recent extinctions. In
Conservation for the twenty-first century: 37–41. Western, D. &
Pearl, M. C. (Eds). New York: Oxford University Press.
Doak, D., Kareiva, P. & Klepetka, B. (1994). Modeling population viability for the desert tortoise in the western Mojave Desert.
Ecol. Applic. 4: 446–460.
Estes, J. A., Tinker, M. T., Williams, T. M. & Doak, D. F. (1998).
Killer whale predation on sea otters linking oceanic and
nearshore ecosystems. Science 282: 473–476.
Fevold, H. R. & Craighead, J. J. (1958). Food requirements of
the golden eagle. Auk 75: 312–317.
Foley, P. (1994). Predicting extinction times from environmental
stochasticity and carrying capacity. Conserv. Biol. 8: 124–137.
Garcelon, D. K. (1990). Observations of aggressive interactions
by bald eagles of known age and sex. The Condor 92:
532–534.
Garcelon, D. K., Risebrough, R. W., Jarman, W. M., Chartrand,
A. B. & Lettrell, E. E. (1989). Accumulation of DDE by bald
eagles Haliaeetus leucocephalus reintroduced to Santa Catalina
Island in Southern California. In Raptors in the modern world:
491–494. Meyburg, B.-U. & Chancellor, R. D. (Eds). Berlin:
World Working Group for Birds of Prey and Owls.
Garcelon, D. K., Roemer, G. W., Phillips, R. B. & Coonan, T. J.
(1999). Food provisioning by island foxes, Urocyon littoralis,
to conspecifics caught in traps. Southwest Nat. 44: 83–86.
Garcelon, D. K. & Thomas, N. J. (1997). DDE poisoning in an
adult bald eagle. J. Wildl. Dis. 33: 299–303.
Garcelon, D. K., Wayne, R. K. & Gonzales, B. J. (1992). A serologic survey of the island fox (Urocyon littoralis) on the Channel
Islands, California. J. Wildl. Dis. 28: 223–229.
Halley, D. J. & Gjershaug, J. O. (1998). Inter- and intra-specific
dominance relationships and feeding behaviour of golden eagles
Aquila chrysaetos and sea eagles Haliaeetus albicilla at carcasses. Ibis 140: 295–301.
Harlow, D. L. & Bloom, P. H. (1989). Buteos and the golden
eagle. In Proceedings of the Western Raptor Mangement
Symposium and Workshop: 102–110. Pendleton, B. G. (Ed.).
Washington, DC: National Wildlife Federation.
Hayes, S. R. & Gessaman, J. A. (1980). The combined effects of
air temperature, wind and radiation on the resting metabolism
of avian raptors. J. Therm. Biol. 5: 119–125.
Hockett, B. S. (1989). Archaeological significance of rabbit–raptor interactions in southern California. N. Am. Archaeologist 10:
123–139.
Hockett, B. S. (1991). Toward distinguishing human and raptor
patterning on leporid bones. Am. Antiquity 56: 667–679.
Holt, R. D. (1977). Predation, apparent competition and the structure of prey communities. Theor. Pop. Biol. 12: 197–229.
Holt, R. D. (1984). Spatial heterogenity, indirect interactions and
the coexistence of prey species. Am. Nat. 124: 377–406.
Holt, R. D. & Lawton, J. H. (1994). The ecological consequences
of shared natural enemies. Ann. Rev. Ecol. Syst. 25: 495–520.
Horvitz, C., Schemske, D. W. & Caswell, H. (1997). The relative
‘importance’ of life-history stages to population growth:
prospective and retrospective analyses. In Structured-population
models in marine, terrestrial. and freshwater systems: 247–271.
Tuljapurkar, S. & Caswell, H. (Eds). New York: Chapman &
Hall.
Isbell, L. A. & Young, T. P. (1993). Human presence reduces predation in a free-ranging vervet monkey population in Kenya.
Anim. Behav. 45: 1233–1235.
Junak, S., Ayers, T., Scott, R., Wilken, D. & Young, D. (1995).
A flora of Santa Cruz Island. Santa Barbara, CA: Santa Barbara
Botanic Gardens.
Junger, A. & Johnson, D. L. (1980). Was there a quarternary land
bridge to the northern Channel Islands. In The California
islands: proceedings of a multidisciplinary symposium: 33– 39.
Power, D. M. (Ed.). Santa Barbara, CA: Santa Barbara Museum
of Natural History.
Kasaty, P. (1978). 1978 conservation award report, March 28,
1978. Naval Ocean Systems Technical Document 150. San
Diego, CA: Natural Resource Program Office, Naval Ocean
Systems Center.
Keegan, D. R., Coblentz, B. E. & Winchell, C. S. (1994). Feral
goat eradication on San Clemente Island, California. Wildl. Soc.
Bull. 22: 56–61.
Kiff, L. F. (1980). Historical changes in resident populations of
California Islands raptors. In The California islands: proceed-
ings of a Multidisciplinary Symposium: 651–673. Power, D. M.
(Ed.). Santa Barbara, CA: Santa Barbara Museum of Natural
History.
Laughrin, L. L. (1977). The island fox: a field study of its behavior and ecology. Ph.D. thesis, University of California, Santa
Barbara.
Lefkovitch, L. P. (1965). The study of population growth in organisms grouped by stages. Biometrics 21: 1–18.
Love, J. A. (1979). The daily food intake of captive white-tailed
eagles. Bird Study 26: 64–66.
MacArthur, R. H. & Wilson, E. O. (1967). The theory of island
biogeography. New York: Princeton University Press.
Mace, G. M. & Lande, R. (1991). Assessing extinction threats:
toward a reevaluation of IUCN threatened species categories.
Conserv. Biol. 5: 148–157.
Minnich, R. A. (1980). Vegetation of Santa Cruz and Santa
Catalina Islands. In The California islands: proceedings of a
multidisciplinary symposium: 123–137. Power, D. M. (Ed.).
Santa Barbara, CA: Santa Barbara Museum of Natural History.
Moore, C. M. & Collins, P. W. (1995). Urocyon littoralis.
Mammalian Species 489. The American Society of
Mammologists, Department of Zoology, Brigham Young
University, Provo, UT.
Nicholson, W. S. & Hill, E. P. (1984). Mortality in gray foxes
from east-central Alabama. J. Wildl. Mgmt. 48: 1429–1432.
Noon, B. R. & Sauer, J. R. (1992). Population models for passerine birds: structure, parameterization, and analysis. In Wildlife
2001: populations: 441–464. McCullough, D. R. & Barrett,
R. B. (Eds). London: Elsevier Applied Science.
Perkins, D. W., Phillips, D. M. & Garcelon, D. K. (1996).
Predation on a bald eagle nestling by a red-tailed hawk. J. Raptor
Res. 30: 249.
Pollock, K. H., Winterstein, S. R., Bunck, C. M. & Curtis, P. D.
(1989). Survival analysis in telemetry studies: the staggered
entry design. J. Wild. Mgmt. 53: 7–15.
Power, D. M. (Ed.) (1980). The California islands: proceedings
of a multidisciplinary symposium. Santa Barbara, CA: Santa
Barbara Museum of Natural History.
Roelke-Parker, M. E., Munson, L., Packer, C. et al. (1996). A
canine distemper virus epidemic in Serengeti lions. Nature 379:
441–445.
Roemer, G. W. (1999). The ecology and conservation of the island
fox (Urocyon littoralis). Ph.D. thesis, University of California,
Los Angeles.
Roemer, G. W., Coonan, T. J., Garcelon, D. K., Starbird, C. H.
& McCall, J. W. (2000). Spatial and temporal variation in the
seroprevalence of canine heartworm antigen in the island fox.
J. Wild. Dis. 36: 723–728.
Roemer, G. W., Garcelon, D. K., Coonan, T. J. & Schwemm, C.
(1994). The use of capture-recapture methods for estimating,
monitoring, and conserving island fox populations. In The fourth
California islands symposium: update on the status of resources:
387–400. Halvorson, W. L. & Maender, G. J. (Eds). Santa
Barbara, CA: Santa Barbara Museum of Natural History.
Roemer, G. W., Smith, D. A., Garcelon, D. K. & Wayne, R. K.
(2001). The behavioural ecology of the island fox. J. Zool.,
Lond. 255: 1–14.
Sillero-Zubiri, C., King, A. A. & Macdonald, D. W. (1996).
Rabies and mortality in Ethiopian wolves (Canis simensis). J.
Wildl. Dis. 32: 80–86.
Soulé, M. E. & Terborgh, J. (1999). Continental conservation.
Washington, DC: Island Press.
Stalmaster, M. V. (1987). The bald eagle. New York: Universe
Books.
Stalmaster, M. V. & Gessaman, J. A. (1984). Ecological energetics and foraging behavior of overwintering bald eagles. Ecol.
Monogr. 54: 407–428.
Sterner, D. (1990). Population characteristics, home range, and
habitat use of feral pigs on Santa Cruz Island, California. M.Sc.
thesis, University of California, Berkeley.
Strickland, K. N. (1998). Canine and feline caval syndrome.
Seminars Vet. Medical Surgery (Small Animal) 13: 88–95.
Stull, J. K., Swift, D. J. P. & Niedoroda, A. W. (1996).
Contaminant dispersal on the Palos Verdes continental margin:
I. Sediments and biota near a major California wastewater discharge. The Science of the Total Environment 179: 73–90.
US Code Service. (1999). 16 USCS 668. Title 16, Conservation.
Chapter 5A, Protection and conservation of wildlife, Protection
of Bald and Golden eagles.
Van Vuren, D. & Coblentz, B. E. (1987). Some ecological effects
of feral sheep on Santa Cruz Island, California, USA. Biol.
Conserv. 41: 253–268.
Vitousek, P. M., Mooney, H. A., Lubchenco, J. & Melillo, J. M.
(1997). Human domination of earth’s ecosystems. Science 277:
494–499.
Wayne, R. K., George, S. B., Gilbert, D., Collins, P. W., Kovach,
S. D., Girman, D. & Lehman, N. 1991. A morphologic and
genetic study of the island fox, Urocyon littoralis. Evolution 45:
1849–1868.
White, G. C., Anderson, D. R., Burnham, K. P. & Otis, D. L.
(1982). Capture–recapture and removal methods for sampling
closed populations. Los Alamos, New Mexico: Los Alamos
National Laboratory.
White, G. C. & Burnham, J. P. (1997). Program Mark – survival
estimation from populations of marked animals. Online at
http://www.cnr.colostate.edu/~gwhite/software.html.
Williams, E. S., Thorne, E. T., Appel, M. J. G. & Belitsky,
D. W. (1988). Canine distemper in black-footed ferrets (Mustela
nigripes) from Wyoming. J. Wildl. Dis. 24: 385–398.
Wisdom, M. J. & Mills, L. S. (1997). Sensitivity analysis to guide
population recovery: prairie-chickens as an example. J. Wildl.
Mgmt. 61: 302–312.