Anda di halaman 1dari 8

Contributed Paper

Influence of Care of Domestic Carnivores on

Their Predation on Vertebrates
Department of Wildlife Ecology and Conservation and School of Natural Resources and the Environment, University of Florida, 110
Newins-Ziegler Hall, Gainesville, FL 32611-0430, U.S.A.
Instituto de Ciencia Animal & Programa de Investigacion Aplicada en Fauna Silvestre, Facultad de Ciencias Veterinarias,
Universidad Austral de Chile, Casilla 567, Valdivia, Chile

Abstract: Domestic dogs ( Canis familiaris) and cats ( Felis catus) are the most abundant mammalian
carnivores worldwide. Given that domestic carnivores rely on human-provided food, their densities are usually
independent of prey densities. Nevertheless, underfed pets may need to hunt to meet their energetic and
nutritional requirements. We explored the effects of different levels of care (provision of food) of dogs and cats
on their predation rates on wild vertebrates in 2 areas of southern Chile. We interviewed cat and dog owners
and analyzed prey remains in scats of pets to examine how domestic dogs and cats were managed and to
gather information on the wild vertebrates killed and harassed by pets. We used logistic regression to examine
the association between pet care and the frequency of wild vertebrate remains in scats. The probability of a
dog preying on vertebrates was higher for poorly fed than for adequately fed dogs (odds ratio = 3.7) and for
poorly fed than for adequately fed cats (odds ratio = 4.7). Domestic dogs and cats preyed on most endemic
and threatened mammals present in the study sites. Our results provide support for the hypothesis that the
less care domestic animals receive from owners the higher the probability those animals will prey on wild

Keywords: Chile, domestic dog, house cat, invasive species, management, trophic subsidy
Influencia del Cuidado de Carnvoros Domesticos Sobre Sus Tasas de Depredacion en Vertebrados
Resumen: Los perros (Canis familiaris) y gatos domesticos (Felis catus) son los carnvoros mas abundantes
globalmente. Puesto que los carnvoros domesticos dependen de alimentos provistos por el ser humano, sus
densidades son usualmente independientes de las densidades de sus presas. Sin embargo, las mascotas mal
alimentadas podran necesitar cazar para suplir sus requerimientos energeticos y nutricionales. Nosotros
exploramos los efectos de diferentes niveles de cuidado (provision de alimento) de perros y gatos sobre sus
tasas de depredacion en vertebrados silvestres en 2 areas del sur de Chile. Nosotros entrevistamos propietarios
de perros y gatos y analizamos los restos de presas en heces para examinar como eran manejados los perros
y gatos domesticos y para obtener informacion sobre vertebrados perseguidos y matados por mascotas.
Utilizamos regresiones logsticas para evaluar la asociacion entre el cuidado de las mascotas y la frecuencia
de vertebrados silvestres en heces. La probabilidad de depredar sobre vertebrados fue mas alta para perros
mal alimentados que para perros bien alimentados (razon de disparidad = 3.7) y para gatos mal alimentados
que para gatos bien alimentados (razon de disparidad = 4.7). Los perros y gatos domesticos depredaron sobre
la mayora de los mamferos endemicos y amenazados presentes en las areas de estudio. Nuestros resultados
proveen soporte para la hipotesis de que mientras menor sea el cuidado dado a los animales domesticos por
sus propietarios mayor es la probabilidad de que depreden sobre vertebrados silvestres.

Palabras Clave: Chile, perro domestico, gato domestico, especie invasora, manejo, subsidios troficos

Department of Wildlife Ecology and Conservation and School of Natural Resources and the Environment, University of Florida, 110 Newins-
Ziegler Hall, Gainesville, FL 32611-0430, U.S.A., email
Paper submitted October 15, 2010; revised manuscript accepted January 10, 2011.
Conservation Biology, Volume 25, No. 4, 808815

C 2011 Society for Conservation Biology
DOI: 10.1111/j.1523-1739.2011.01690.x
Silva-Rodrguez & Sieving 809

Introduction less likely it is to prey on wild animals (Kays & DeWan

2004). If this hypothesis holds, then the per capita effects
Domestic cats (Felis catus) and dogs (Canis familiaris) of pets may be reduced by improving the quality and
are the most abundant and ubiquitous carnivores in the quantity of their diet. Few researchers have tested this
world. Worldwide there are over 400 million cats and hypothesis in field settings, and no one has found an
500 million dogs (Jarvis 1990; Wanderler et al. 1993). association between predation rates and management of
Both species are often cited as causes of declines and lo- domestic carnivores (e.g., cats) (Barrat 1998).
cal extinctions of various taxa (e.g., mammals, Vazques- We examined whether differences in management of
Dominguez et al. 2004; birds, Taborsky 1988; reptiles, domestic cats and dogs (hereafter pets) are associated
Iverson 1978). Domestic cats and dogs kill (Woods et al. with differences in their predation rates on vertebrates.
2003; Manor & Saltz 2004), transmit disease to (Fiorello We compared predation on vertebrates by poorly and
2004: Cleaveland et al. 2006), and compete with (Butler adequately fed pets. We conducted our study in rural
& du Toit 2002; Vanak & Gompper 2010) wild animals. areas, where we expected to observe variation in pet
The direct effects of domestic carnivores on wild verte- management and the probability of interactions between
brates can be extreme. For example, in New Zealand a wild animals and pets. We also sought to identify cases of
single dog killed over 500 individuals (more than half the predation on or harassment of endemic and threatened
population) of the endangered North Island Brown Kiwi species.
(Apteryx mantelli) (Taborsky 1988). Similarly, a single
cat was responsible for the extinction of a subspecies Methods
of the deer mouse (Peromyscus guardia) (Vazques-
Study Site
Dominguez et al. 2004). These are extreme examples,
and rates of predation on vertebrates by individual do- The temperate rainforests in southern Chile and Ar-
mestic carnivores vary widely (Barrat 1998; Woods et al. gentina have high levels of endemism (Myers et al. 2000).
2003). A high proportion of forests in this area have been con-
The human-domestic carnivore relation is defined by verted to agriculture and to plantations of non-native trees
the services these animals provide to their human own- (Armesto et al. 1998; Van Holt 2009). In this region dogs
ers, including protection, companionship, and pest con- commonly prey on endemic species such as the south-
trol. In return, people provide their pets with varying ern pudu (Pudu puda) (Silva-Rodrguez et al. 2010b) and
levels of food, shelter, and health care (Serpell 1995). affect the use of space of native carnivores such as chilla
The food owners provide results in maintenance of densi- foxes (Lycalopex griseus) (Silva-Rodrguez et al. 2010a).
ties of domestic carnivores that are far higher than those Catsalthough unstudiedare also suspected to be a
of wild carnivores (Vanak & Gompper 2009), yet food threat to some endemic species (Willson et al. 2005).
provision does not necessarily stop dogs and cats from We conducted our study between July and December
killing wild vertebrates they encounter (Turner & Meis- 2008 in 2 rural landscapes of southern Chile. Chaihun
ter 1988). Thus, it follows that high predator densities (39 57 S; 73 34 W) is in Los Ros district at the northern
may cause frequent mortality of prey species at poten- limit of the largest protected area in the Chilean Coastal
tially unsustainable levels, even when prey abundances Range. This area of 800 ha was inhabited by approxi-
are low. In this case even inefficient or infrequent preda- mately 70 families engaged primarily in the fishing and
tion on endangered or rare prey may reduce probabilities tourism trades. Chaihun is surrounded by large extents of
of persistence of such species (Kays & DeWan 2004). native forest and has annual precipitation > 2100 mm and
The management of domestic carnivores ranges from a mean annual temperature of 11.3 C (Delgado 2010).
complete control (i.e., dogs and cats receive all basic re- Centinela (40 14 S; 73 04 W) is a rural area north of the
quirements from owners and are kept inside houses or city of La Union. The area (5000 ha) was inhabited by
are always restrained) to feral domestic carnivores that re- approximately 100 families engaged primarily in subsis-
ceive no human care and search for their own food (Kays tence farming. Annual average precipitation is 1267 mm
& DeWan 2004; Vanak & Gompper 2009). In the mid- and mean annual temperature is 11.6 C (Luebert & Plis-
dle of this range of management are animals that largely coff 2005). The landscape is dominated by pastures and
range free outside of homes and receive some food from eucalyptus plantations, and native forest covers approx-
people (Kays & DeWan 2004). No reliable data exist for imately 25% of the area. To our knowledge there are no
estimating the relative frequency of these various man- feral dogs or cats in these areas. Dogs in both areas were
agement conditions as they apply to domestic carnivores of mixed breeds and ranged in size from 6 to 10 kg (fox
globally, but both feral and owned but free-ranging ani- terrier size) to 40 to 60 kg (Rottweiler size). Most dogs
mals are common (Butler & Bingham 2000; Kitala et al. were of medium size (1020 kg). The body weight of
2001; Fiorello 2004). cats ranged between 1.5 and 6 kg. Five threatened (IUCN
It has been hypothesized that the more food a free- 2010) or endemic mammals occupy the region: vulner-
ranging domestic carnivore receives from humans the able pudus (Cervidae) and guignas (Leopardus guigna,

Conservation Biology
Volume 25, No. 4, 2011
810 Domestic Carnivore Predation Behavior

Felidae), endangered southern river otters (Lontra provo- owners property and cages contained food, water, and
cax, Mustelidae [only in Chaihun]), and endemic moni- sand to reduce stress and to facilitate defecation. When
tos del monte (Dromiciops gliroides, Microbiotheridae) the owner did not want to restrain cats, and in all cases
and Chilean climbing mice (Irenomys tarsalis, Criceti- with dogs, we collected fresh scats that we were certain
dae). Seven birds, 4 rhinocryptids, and 3 furnariids, are belonged to the examined pet (e.g., we or the owner
endemic to the South American temperate forests and observed the animal defecating). Although this sampling
present in both sites. protocol implies intense monitoring of animals and a po-
tential compromise in terms of sample size, it ensured
Domestic Carnivore Management and Demography that the origin of scats could be identified to individuals.
We obtained scats from 33 cats and 41 dogs in Chaihun
We obtained information regarding number, age, sex,
and 56 cats and 51 dogs in Centinela.
management, and veterinary care of pets by interview-
We used the methods of Reynolds and Aesbischer
ing homeowners in Chaihun and Centinela. We based
(1991) for scat analyses. We identified mammals that
our questions on those used in similar research (But-
were eaten to genus (and species when possible) through
ler & Bingham 2000; Fiorello 2004). On the basis of
the use of keys based on the scale pattern and medulla
an a priori power analysis, we sought to survey a min-
arrangement of hairs and teeth; keys were specific to the
imum of 40 households each in Chaihun and Cen-
geographic region (Chehebar & Martin 1989; Pearson
tinela so that the sample would include at least 68 pets
1995). We identified bird feathers with available keys
of each species. We used images from Google Earth
(Day 1966; Reyes 1992). Other prey items were identi-
( and data collected in previous
fied to order. From the scat analyses, we calculated the
research (Silva-Rodrguez et al. 2010a) to identify house-
proportion of animals that consumed each prey species.
hold locations. Houses in each landscape were selected
We obtained additional information on predation from
for interviews through random sampling. We asked peo-
interviews. We asked about vertebrate species their pets
ple why they kept pets, how they managed them, what
frequently killed or harassed, including specific questions
food they fed them, and whether they vaccinated and
about rare species (Supporting Information). We asked
treated their pets for parasites (Supporting Information).
whether individual cats had killed or harassed animals
We calculated a body condition score (BCS) for each
in the last day and whether individual dogs had killed
pet. A BCS is based on a clinical assessment of the amount
or harassed medium and large mammals within the last
of fat covering the ribs, base of the tail, and abdominal
year. We separated events that referred to predation from
contour (Lund et al. 1999; Burkholder 2000). The score
those that referred to harassment.
ranges from 1.0, for an extremely thin animal, to 5.0, for
an obese animal (Lund et al. 1999; Stafford 2006). The BCS Data Analyses
is not based on animal size, which varies widely within
species and breeds (Burkholder 2000). Nevertheless, its To assess the effects of human-provided food on the prob-
use as a single criterion with which to assess the quality of ability of detecting vertebrate remains in pet scats, we
food provided by specific owners may be questionable. used logistic regression (Agresti 2002). We included food
For example, an animal may be in poor body condition if supply (poor or adequate), site (Chaihun, Centinela), age
it is ill (Stafford 2006) or may have good body condition (juvenile [<1 year old], adult), and sex as covariates in the
despite being underfed by owners because it obtains ad- analysis of cats. For dogs we obtained predation data from
ditional food elsewhere. Thus, we classified all pets we interviews because only a small proportion of scats con-
examined as adequately or poorly fed on the basis of the tained animal remains. We only analyzed data on animals
following criteria. Pets with BCS <3.0 were underfed killed by pets. Food supply, site, age, and total number of
independent of owner-reported feeding regime, and pets dogs owned by the household were covariates. Sex was
with BCS 3.0 were adequately fed only if the owner not included as a predictor variable for dogs because in
provided it with commercial or prepared food on a daily both areas there were very few females (Table 1). The
basis. If the pet was fed on inadequate food items, such number of dogs per household was included as a covari-
as wheat bran or household scraps, we classified it as ate because dogsunlike catsmay hunt in packs. We
poorly fed independent of its body condition. With this assessed model fit with the unweighted sum of squares
approach some pets in poor health could be classified test (Hosmer et al. 1997). Analyses were conducted in R
as poorly fed on the basis of a low BCS. If such misclas- (version 2.9.2; R Development Core Team 2009).
sifications occurred, however, then our results would
underestimate predation by poorly fed animals. Results
Diet Domestic Carnivore Management
We collected scats of cats after keeping the cats inside We conducted 37 and 54 interviews in Chaihun and Cen-
transport cages for 24 h. Animals were kept on their tinela respectively. All people we interviewed owned at

Conservation Biology
Volume 25, No. 4, 2011
Silva-Rodrguez & Sieving 811

Table 1. Data on pet dogs and cats obtained through interviews of households in Chaihun (n = 37) and Centinela (n = 54) in southern Chile.

Dog Cat
Chaihun Centinela Chaihun Centinela

Households with pets (%) 89.2 94.4 83.8 79.6

Mean pets per household 1.4 2.7 1.3 1.6
Motivation to own pet (%)
companionship 100.0 100.0 90.3 65.1
theft prevention 87.9 98.0 0.0 0.0
prevention of predation on farm animals 12.1 31.4 0.0 0.0
rat control 0.0 0.0 64.5 86.1
Sex ratio (males per female) 16.3 7.5 1.4 0.8
Management of pet
free roaming 90.9 84.3 100.0 100.0
restricted (leash or fence) 9.1 15.7 0.0 0.0
commercial or prepared food 66.7 33.3 54.8 23.3
wheat bran 15.1 54.9 0.0 9.3
household scraps 18.2 11.8 45.2 65.1
not fed 0.0 0.0 0.0 4.7
vaccinated against rabies 33.3 3.9 0.0 0.0
other vaccines 0.0 0.0 0.0 0.0
treated for parasites 18.2 7.8 0.0 0.0
Reproductive control 0.0 0.0 3.2 0.0
Motivations to own a dog or cat are not mutually exclusive.

least one dog or cat. In both sites the populations of pet would prey on vertebrates were 3.7 times higher
dogs were strongly biased toward males (89.8% males), for poorly fed than for adequately fed dogs ( 2 = 5.26,
whereas sex biases were not apparent for cats (49.5% df = 1, p < 0.05; Table 2).
males). The strong bias toward male dogs was explained
by cultural factors. Dogs were owned for companion-
ship, household protection, and prevention of livestock (a)
theft or predation. Cats were owned for companionship
and rat control. The general management of dogs and
cats was similar in both areas. Most owners allowed
their pets to roam free (dogs, 86.9%; cats, 100%), did
not provide them with basic health care (e.g., no vac-
cinations for diseases other than rabies), and fed their
pets on food items such as wheat bran (dogs, 39.3%;
cats, 5.4%) or household scraps (dogs, 14.3%; cats, 56.8%; (b)
Table 1).

Predation of Pets on Wild Vertebrates According to


In both areas people we interviewed mentioned their

dogs had killed (14%) or harassed (25%) native species
such as chilla foxes, guignas, and pudus during the pre-
vious year. Respondents also mentioned their dogs ha-
rassed (58%) or killed (40%) non-native European hares
(Lepus europaeus) during the same timeframe (Fig. 1).
Adult dogs, poorly fed dogs, and dogs from households
that owned multiple dogs preyed on vertebrates more of- Figure 1. Percentage of owners of dogs who reported
ten than juvenile, adequately fed, and singly owned dogs their dogs had killed or harassed medium-sized wild
(likelihood ratio 2 = 28.75, df = 4, p < 0.01; unweighted mammals in (a) Chaihun and (b) Centinela in
sum of squares test z = 1.21; p = 0.23). The odds a southern Chile within a year of being interviewed.

Conservation Biology
Volume 25, No. 4, 2011
812 Domestic Carnivore Predation Behavior

Table 2. Logistic regression model of variables associated with Scat Analyses

predation of dogs on wild vertebrates during the previous year as
reported by their owners in Chaihun and Centinela in southern Chile. We examined the scats of 92 dogs and 89 cats. Farm ani-
mals and European hares were found in scats of 8.7% and
Likelihood ratio
5.4% of dogs, respectively. Rare and endemic species
Odds ratio 95% CI 2 df p were not detected in dog scats. Rodents and passerine
species of birds were in scats of 25.8% and 15.7% of
Intercept 0.00 0.000.07 19.73 1.00 0.00
cats, respectively. The genus of most rodents we could
Food supply
poor 3.75 1.2113.03 5.26 1.00 0.02 identify in scats was non-native Rattus (5.6%) and na-
adequate 1.00 tive Abrothrix (7.9%). The most common bird species in
Site scat was the House Wren (4.5%). Monito del monte and
Chaihun 1.73 0.536.15 0.81 1.00 0.37 Chilean climbing mouse were detected in scats (1.1%
Centinela 1.00
respectively, Table 3). Vertebrate remains were found
adult 6.78 1.1152.61 4.32 1.00 0.04 more often in scats from cats of Centinela than in scats
juvenile 1.00 from cats of Chaihun and in scats of poorly fed cats
Number of dogs 3.73 1.928.17 18.09 1.00 0.00 than in scats of adequately fed cats (likelihood ratio
2 = 17.5, df = 4, p< 0.01; unweighted sum of squares
test z = 0.67; p = 0.51). The odds that vertebrate re-
Most cat owners (89.2%) reported their cats preyed mains were present in the scats of poorly fed cats were
on birds, small mammals, and lizards. One interviewee at 4.7 times higher than in the scats of adequately fed cats
Chaihun and 4 at Centinela observed their cats bringing ( 2 = 9.35, df = 1, p < 0.01; Table 4).
prey home in the last 24 h. We identified 4 individual dead
birds brought home by cats: 2 House Wrens (Troglodytes
aedon), 1 Chilean Swallow (Tachycineta meyeni), and 1 Discussion
House Sparrow (Passer domesticus). Interviewees iden-
tified several other species killed by cats, but none of Kays and DeWan (2004) hypothesize that the more food
the endemic species were mentioned. It was not possi- a pet receives from people the less likely it is to prey
ble to verify the identity of small mammals brought into on wild vertebrates because it is less driven by hunger
owners homes by cats because the small mammals were to kill. Our results are consistent with this hypothesis.
quickly disposed of and people did not identify them. Poorly fed dogs and cats preyed on vertebrates more

Table 3. Vertebrates detected in scats of pet cats in Chaihun (n = 33 cats) and Centinela (n = 56 cats) in southern Chile.

Scat (%)
Class Order Species Land cover /land usea Chaihun Centinela

Mammals Microbiotheria Dromiciops gliroidesb F 0.0 1.8

Chiroptera Unidentified 0.0 1.8
Rodentia Rattus spp.c F, H, S 3.0 7.1
Abrothrix spp. F, S 6.1 8.9
Oligoryzomys longicaudatus F, S 0.0 3.6
Irenomys tarsalisb F 0.0 1.8
Mus musculusc H 3.0 3.6
unidentified 6.1 5.4
Lagomorfa Lepus europaeusc P, S 0.0 3.6
subtotal 18.2 37.5
Birds Tinamiformes Nothoprocta perdicaria S 0.0 3.6
Pelecaniformes unidentified 3.0 0.0
Charadriiformes Vanellus chilensis P, W 0.0 1.8
Apodiformes Sephanoides sephaniodes F, H, S 0.0 1.8
Passeriformes Troglodytes aedon F, H, S 3.0 5.4
Carduelis barbata F, H, S 0.0 1.8
unidentified 6.1 10.7
subtotal 12.1 26.8
Reptiles Squamata unidentified 0.0 1.8
subtotal 0.0 1.8
Total prey 27.3 57.1
a Land-cover and land-use associations of prey species (Jaramillo 2003; Iriarte 2008): F, forest; H, human settlements; P, pastures; S, shrubs; W,
b Endemic to South American temperate forests.
c Non-native species.

Conservation Biology
Volume 25, No. 4, 2011
Silva-Rodrguez & Sieving 813

Table 4. Logistic regression model of variables associated with the exposure and attacks on prey (including kills) is nega-
detection of vertebrate remains in the scats of pet cats in Chaihun and tively associated with time since last meal which, given
Centinela in southern Chile.
realistic conditions, might allow prey to escape from a
Likelihood ratio recently fed cat but not from a food-deprived one (Biben
1979). Our results are consistent with these findings in
Parameter Odds ratio 95% CI 2 df p
that we found underfed animals killed vertebrates signif-
Intercept 0.10 0.020.54 4.70 1.00 0.03 icantly more often than adequately fed animals.
Food supply
poor 4.74 1.7313.85 9.35 1.00 0.00
adequate 1.00 Potential Effects of Dogs and Cats on Rare Vertebrates
Chaihun 0.34 0.11- 0.93 4.43 1.00 0.04 Owned pets preyed on most rare and endemic mam-
Centinela 1.00 malian species present in the region. Nevertheless, en-
Sex demic mammals were detected in only 2 cat scats
female 0.68 0.251.80 0.60 1.00 0.44 (Table 3) and interactions between dogs and rare and en-
male 1.00 demic mammals were only detected through interviews.
adult 2.10 0.548.71 1.14 1.00 0.29 Low predation rates on endemic mammals likely reflect
juvenile 1.00 their low abundances (Beckerman et al. 2007). For ex-
ample, in southern Chile pudu densities range between
3.85 and 6.25 individuals/km2 (Simonetti & Mella 1997),
often than adequately fed animals in both study areas whereas dog densities in rural areas are > 7.0 individu-
(Tables 2 & 4, Supporting Information). We believe the als/km2 (e.g., Silva-Rodrguez et al. 2010a). Thus, frequent
lack of association between food supply and predation in predation by any given dog on pudus is impossible. Be-
a study conducted in Australia (Barrat 1998), where most cause dogs may be more abundant than endangered prey;
animals were well fed, may be a reflection of the study however, infrequent predation by an individual dog may
site. Nevertheless, even in the United Kingdom thinner decrease the probability of persistence of relatively rare
cats bring prey home more often (Woods et al. 2003), prey species. Dog attack is one of the most frequent
which may result from differences in amount of food causes of death for pudus in Chile, including in our study
provided by humans. sites and areas nearby (Silva-Rodrguez et al. 2010b). Fur-
Although few researchers have directly compared pre- thermore, our estimation of the potential effect of dogs
dation rates between poorly and adequately fed animals, and cats on threatened and endemic vertebrates is conser-
studies on the diets of animals at the extreme ranges of vative because we sampled only pets. Pets are expected
management (i.e., feral vs. well-fed, owned pets) show to prey on what is available near houses (e.g., cats preyed
that animals that receive less food from humans hunt on Mus musculus and Rattus spp., Table 3). Feral dogs
more. For example, well-managed (limited range) dogs and cats were absent in our study sites, but they thrive
in urban areas rely on food provided by people and free- elsewhere (Iriarte 2008), where they are not given as
ranging dogs in rural areas rely on food provided by peo- much food as or as often as pets and are not associated
ple and wild caught prey, whereas feral dogs rely heavily with houses or yards. We believe that feral animals prey
on wild prey (Vanak & Gompper 2009). Scats of feral on wild vertebrates more often than pets and feed more
cats (i.e., no food provided by humans) usually contain frequently than pets in areas located far from houses.
more than one prey individual per scat (e.g., Liberg 1984; In addition to predation, pets affect vertebrate com-
Matias & Catry 2008; Peck et al. 2008). In contrast, stud- munities through nonlethal interactions. People often
ies of owned, well-fed cats report predation rates of < 5 reported their pets harassed vertebrates (Fig. 1). Such
vertebrates/cat/month (e.g., Barrat 1998; Woods et al. interactions do not directly cause the death of the un-
2003; Kays & DeWan 2004). Our data also are consistent domesticated animals, but they may affect their use of
with the hypothesis that feeding of owned domestic ani- space and even their fecundity. For example, in one of
mals influences predation rates of those pets on wild ani- our study sites we found previously that domestic dogs
mals. Therefore, we hypothesize that an improvement in constrain the use of space by chilla foxes through harass-
the quantity and quality of food will decrease incidences ment (Silva-Rodrguez et al. 2010a). Similar findings have
of pets preying on vertebrates and improve the body been reported in other regions where the occurrence
condition of pets. of carnivores is negatively associated with dog activity
Controlled experiments with cats also suggest an as- (George & Crooks 2006; Lacerda et al. 2009; Vanak &
sociation between food deprivation and predation. Al- Gompper 2010). Sublethal effects of pets may also affect
though satiated cats hunt, the amount of time since the fitness of vertebrates. For example, snowshoe hares (Le-
last meal is positively associated with the probability of pus americanus) that were experimentally exposed to
killing or playing with prey (or toys) (Biben 1979; Hall dogs had lower birth rates and offspring body weights
& Bradshaw 1998). Furthermore, the time between prey relative to hares in the control group (Sheriff et al. 2009).

Conservation Biology
Volume 25, No. 4, 2011
814 Domestic Carnivore Predation Behavior

Small decreases in fecundity may cause major decreases Armesto, J. J., R. Rozzi, C. Smith-Ramirez, and M. T. K. Arroyo. 1998.
in the abundance of prey species. In some cases, sub- Conservation targets in South American temperate forests. Science
lethal effects of domestic carnivores could have greater
Barratt, D. G. 1998. Predation by house cats, Felis catus (L.), in Can-
effects than predation (Beckerman et al. 2007). Consid- berra, Australia. II. Factors affecting the amount of prey caught
ering that free-ranging pets are the most common carni- and estimates of the impact on wildlife. Wildlife Research 25:475
vores in many landscapes around the world (Wandeler 487.
et al. 1993; Vanak & Gompper 2009), including southern Beckerman, A. P., M. J. Boots, and K. J. Gaston. 2007. Urban bird de-
clines and the fear of cats. Animal Conservation 10:320325.
Chile (Silva-Rodrguez et al. 2010b), these sublethal ef-
Biben, M. 1979. Predation and predatory play behaviour of domestic
fects may be ubiquitous and their mitigation may depend cats. Animal Behavior 27:8194.
on restricting the movement of pets (Vanak & Gompper Burkholder, W. J. 2000. Use of body condition scores in clinical assess-
2010). ment of the provision of optimal nutrition. Journal of the American
In the developing world, where most owned domestic Veterinary Medical Association 5:650654.
Butler, J. R. A., and J. Bingham. 2000. Demography and dog-human re-
carnivores are free ranging and do not receive healthcare
lationships of the dog population in Zimbabwean communal lands.
or adequate food (Butler & Bingham 2000; Kitala et al. Veterinary Record 147:442446.
2001; Fiorello 2004; this study), the effects of domestic Butler, J. R. A., and J. T. du Toit. 2002. Diet of free-ranging domestic
carnivores are not limited to predation. Thus, we suggest dogs (Canis familiaris) in rural Zimbabwe: implications for wild
that better food should be accompanied by better health scavengers on the periphery of wildlife reserves. Animal Conserva-
tion 5:2937.
care (e.g., to control diseases such as canine distemper
Cleaveland, S., M. Kaare, D. Knobel, and M. K. Laurenson. 2006. Ca-
virus and rabies that also affect wild animals) (Cleaveland nine vaccinationproviding broader benefits for disease control.
et al. 2006), and, when possible, restriction of movement Veterinary Microbiology 117:4350.
of pets. Improving nutrition could lead to an increase in Chehebar, C., and S. Martin. 1989. Gua para el reconocimiento mi-
reproductive success of pets; thus, reproductive control croscopico de los pelos de los mamferos de la Patagonia. Donana
Acta Vertebrata 16:247291.
also is fundamental to prevent population growth. Bet-
Day, M. G. 1966. Identification of hair and feather remains in gut and
ter food, healthcare, and reproductive control are impor- faeces of stoats and weasels. Journal of Zoology 148:201217.
tant not only for the conservation of several endangered Delgado, C. 2010. Plan de manejo de la Reserva Costera Valdiviana
species, but also for animal welfare. [Management plan of the Valdivian Coastal Reserve]. The Nature
Conservancy Valdivia, Arlington, Virginia.
Fiorello, C. V. 2004. Disease ecology of wild and domestic carnivores
in Bolivia. PhD dissertation. Department of Ecology, Evolution, and
Acknowledgments Environmental Biology, Columbia University, New York.
George, S. L., and K. R. Crooks. 2006. Recreation and large mam-
We thank local people for letting us work with their mal activity in an urban nature reserve. Biological Conservation
pets. The staff of Reserva Costera Valdiviana provided 133:107117.
invaluable support. J. Rau provided the key literature. Hall, S. L., and J. W. S. Bradshaw. 1998. The influence of hunger on ob-
M. Christman provided statistical advice. W. Chaves, E. ject play by adult domestic cats. Applied Animal Behaviour Science
Fleishman, J. Frechette, C. Heatherington, F. Hua, P. 58:143150.
Hosmer, D. W., T. Hosmer, S. le Cessie, and S. Lemeshow. 1997. A
Huang, M. Main, R. Johnson, J. Simonetti, A. Vanak, T. comparison of goodness-of-fit tests for the logistic regression model.
Van Holt, and 3 anonymous reviewers provided valuable Statistics in Medicine 16:965980.
comments on this manuscript. The study was approved Iriarte, A. 2008. Mamferos de Chile [Mammals of Chile]. Lynx editions,
by the Non-Regulatory Animal Research System (IFAS Barcelona.
ARC#016-08WEC-ADD001) and funded by a Field Re- IUCN (International Union for Conservation of Nature). 2010. IUCN Red
List of threatened species. Version 2010.4. IUCN, Cambridge, United
search grant from the Tropical Conservation and Devel- Kingdom. Available from (accessed De-
opment program, University of Florida. E.A.S. was funded cember 2010).
by a FulbrightConicyt fellowship. Iverson, J. B. 1978. Impact of feral cats and dogs on populations of
West-Indian rock iguana, Cyclura carinata. Biological Conservation
Supporting Information Jaramillo, A. 2003. Birds of Chile. Princeton University Press, Princeton,
The questionnaire (Appendix S1) and the results of the New Jersey.
Jarvis, P. J. 1990. Urban cats as pests and pets. Environmental Conser-
analysis of the data with a multimodel framework (Ap- vation 17:169171.
pendix S2) are available online. The authors are solely Kays, R. W., and A. A. DeWan. 2004. Ecological impact of inside/outside
responsible for the content and functionality of these ma- house cats around a suburban nature preserve. Animal Conservation
terials. Queries (other than the absence of the material) 7:273283.
should be directed to the corresponding author. Kitala, P., J. McDermott, M. Kyule, J. Gathuma, B. Perry, and A. Wan-
deler. 2001. Dog ecology and demography information to support
the planning of rabies control in Machakos District, Kenya. Acta
Literature Cited Tropica 78:217230.
Lacerda, A. C. R., W. M. Tomas, and J. Marinho-Filho. 2009. Domestic
Agresti, A. 2002. Categorical data analysis. 2nd edition. John Wiley and dogs as an edge effect in the Brasilia National Park, Brazil: Interac-
Sons, Hoboken, New Jersey. tions with native mammals. Animal Conservation 12:477487.

Conservation Biology
Volume 25, No. 4, 2011
Silva-Rodrguez & Sieving 815

Liberg, O. 1984. Food-habits and prey impact by feral and house-based Silva-Rodrguez, E. A., G. R. Ortega-Sols, and J. E. Jimenez.
domestic cats in a rural area in southern Sweden. Journal of Mam- 2010a. Conservation and ecological implications of habitat pref-
malogy 65:424432. erences by chilla foxes and free-ranging dogs in a human-
Luebert, F., and P. Pliscoff. 2005. Bioclimas de la cordillera de la dominated landscape in southern Chile. Austral Ecology 35:765
costa del centro-sur de Chile [Bioclimates of the coastal range 777
of south central Chile]. Pages 6072 in C. Smith-Ramirez, J. J. Silva-Rodriguez, E. A., C. Verdugo, O. A. Aleuy, J. G. Sanderson, G.
Armesto, and C. Valdovinos, editors. Historia, biodiversidad y R. Ortega-Solis, F. Osorio-Zuniga, and D. Gonzalez-Acuna. 2010b.
ecologa de los bosques costeros de Chile [History, biodiversity Evaluating mortality sources for the vulnerable pudu Pudu puda in
and ecology of the Chilean coastal forests]. Editorial Universitaria, Chile: implications for the conservation of a threatened deer. Oryx
Santiago. 44:97103.
Lund, E. M., P. J. Armstrong, C. A. Kirk, L. M. Kolar, and J. S. Klausner. Simonetti, J. A., and J. E. Mella. 1997. Park size and the conservation
1999. Health status and population characteristics of dogs and cats of Chilean mammals. Revista Chilena de Historia Natural 70:213
examined at private veterinary practices in the United States. Journal 220
of the American Veterinary Medical Association 214:13361341. Stafford, K. 2006. The welfare of dogs. Springer-Verlag, Dordrecht,
Manor, R., and D. Saltz. 2004. The impact of free-roaming dogs on Netherlands.
gazelle kid/female ratio in a fragmented area. Biological Conserva- Taborsky, M. 1988. Kiwis and dog predation; observations at Waitangi
tion 119:231236. State Forest. Notornis 35:197202.
Matias, R., and P. Catry. 2008. The diet of feral cats at New Island, Turner, D. C., and O. Meister. 1988. Hunting behaviour of the domestic
Falkland Islands, and impact on breeding seabirds. Polar Biology cat. Pages 111121 in D. C. Turner and P. Bateson, editors. The
31:609616. domestic cat: the biology of its behaviour. Cambridge University
Myers, N., R. A. Mittermeier, C. G. Mittermeier, G. A. B. da Fonseca, Press, Cambridge, United Kingdom.
and J. Kent. 2000. Biodiversity hotspots for conservation priorities. Van Holt, T. 2009. Influence of landscape change on the nearshore
Nature 403:853858. fisheries in southern Chile. Ph.D. dissertation. University of Florida,
Pearson, O. P. 1995. Annotated keys for identifying small mammals Gainesville.
living in or near Nahuel-Huapi National Park or Lanin National Park, Vanak, A. T., and M. E. Gompper. 2009. Dogs Canis familiaris as car-
southern Argentina. Mastozoologia Neotropical 2:99148. nivores: their role and function in intraguild competition. Mammal
Peck, D. R., L. Faulquier, P. Pinet, S. Jaquemet, and M. Le Corre. 2008. Review 39:265283.
Feral cat diet and impact on sooty terns at Juan de Nova Island, Vanak, A. T., and M. E. Gompper. 2010. Interference competition at the
Mozambique Channel. Animal Conservation 11:6574. landscape level: the effect of free-ranging dogs on a native mesocar-
R Development Core Team. 2009. R: a language and environment nivore. Journal of Applied Ecology 47:12251232
for statistical computing. R Foundation for Statistical Computing, Vazquez-Dominguez, E., G. Ceballos, and J. Cruzado. 2004. Extirpation
Vienna. of an insular subspecies by a single introduced cat: the case of
Reyes, C. F. 1992. Clave para la identificacion de los ordenes de aves the endemic deer mouse Peromyscus guardia on Estanque Island,
chilenas: microestructura de los nodos de las barbulas [Key for the Mexico. Oryx 38:347350.
identification of orders of Chilean birds: microstructure of the nodes Wandeler, A. I., H. C. Matter, A. Budde, and A. Kappeler. 1993. The
of the barbules]. Instituto Profesional de Osorno, Osorno, Chile. ecology of dogs and canine rabies: a selective review. Revue Sci-
Reynolds, J. C., and N. J. Aebischer. 1991. Comparison and quantifica- entifique et Technique Office International des Epizooties 12:51
tion of carnivore diet by fecal analysisa critique, with recommen- 71.
dations, based on a study of the fox Vulpes vulpes. Mammal Review Willson, M. F., K. E. Sieving, and T. L. De Santo Daz. 2005. Aves
21:97122. del bosque de Chiloe: diversidad, amenazas y conservacion. Pages
Serpell, J. 1995. The domestic dog: its evolution, behaviour, and interac- 468476 in C. Smith-Ramirez, J. J. Armesto, and C. Valdovinos, ed-
tions with people. Cambridge University Press, Cambridge, United itors. Historia, biodiversidad y ecologa de los bosques costeros de
Kingdom. Chile. Editorial Universitaria, Santiago.
Sheriff, M. J., C. J. Krebs, and R. Boonstra. 2009. The sensitive hare: Woods, M., R. A. McDonald, and S. Harris. 2003. Predation of wildlife
sublethal effects of predator stress on reproduction in snowshoe by domestic cats Felis catus in Great Britain. Mammal Review
hares. Journal of Animal Ecology 78:12491258. 33:174188.

Conservation Biology
Volume 25, No. 4, 2011