Content uploaded by Sharon A Poessel
Author content
All content in this area was uploaded by Sharon A Poessel on Apr 08, 2017
Content may be subject to copyright.
ARTICLE
Coyote (Canis latrans) diet in an urban environment: variation
relative to pet conflicts, housing density, and season
S.A. Poessel, E.C. Mock, and S.W. Breck
Abstract: Coyotes (Canis latrans Say, 1823) are highly successful in urbanized environments, but as they populate cities, conflict
can occur and often manifests in the form of incidents with pets. To better understand whether coyotes view pets as prey or,
alternatively, as competitors or a threat, we conducted a diet analysis of coyotes in the Denver metropolitan area (DMA) by
analyzing scats. We also examined differences in diet between high- and low-density housing and among seasons. We found only
small percentages of trash and domestic pets in the coyote diet. The presence of pets in the diet did not coincide with the increase
of pet conflicts in the DMA in December and January, supporting the hypothesis that coyote conflict with pets is primarily driven
by competition or a threat response. Coyotes relied mostly on native plant and animal species, and rodents and lagomorphs were
the most prevalent diet items. Coyotes consumed rodents and non-native plants more often in high-density housing and deer,
corn, and native plants more often in low-density housing. Coyotes also consumed more fruits and invertebrates during summer
and autumn and more mammals and birds in winter and spring. As human–coyote conflicts increase in urban areas, under-
standing how coyotes and other urban-adapted carnivores use anthropogenic resources may provide insight that can be used to
promote coexistence between humans and wildlife.
Key words: anthropogenic, Canis latrans, coyote, Denver, food habits.
Résumé : Les coyotes (Canis latrans Say, 1823) ont beaucoup de succès dans les milieux urbanisés, mais leur établissement dans
les villes peut entraîner des conflits qui se manifestent souvent par des incidents impliquant des animaux de compagnie. Afin
de mieux comprendre si les coyotes considèrent les animaux de compagnie comme des proies ou comme des concurrents ou
menaces, nous avons analysé les régimes alimentaires de coyotes dans la région métropolitaine de Denver (DMA) en analysant
des excréments. Nous avons également examiné les différences de régime alimentaire entre des secteurs résidentiels de forte
et de faible densité et d’une saison a
`l’autre. Nous n’avons décelé que de faibles pourcentages de déchets et d’animaux de
compagnie dans le régime alimentaire des coyotes. La présence d’animaux de compagnie dans l’alimentation des coyotes ne
coïncidait pas avec l’augmentation des conflits avec ces derniers dans la DMA en décembre et janvier, ce qui appuie l’hypothèse
voulant que les conflits entre coyotes et animaux de compagnie découlent principalement de la concurrence ou de réactions
a
`la menace. Les coyotes avaient principalement recours a
`des espèces de plantes et d’animaux indigènes, les rongeurs et
lagomorphes étant les éléments prédominants de leur régime alimentaire. Les coyotes consommaient plus souvent des rongeurs
et plantes non indigènes dans les secteurs résidentiels de forte densité, et des cerfs, du maïs et des plantes indigènes dans les
secteurs résidentiels de faible densité. Les coyotes consommaient également plus de fruits et d’invertébrés durant l’été et
l’automne et plus de mammifères et d’oiseaux a
`l’hiver et au printemps. Avec l’augmentation du nombre de conflits entre
humains et coyotes dans les zones urbaines, la compréhension de l’utilisation des ressources anthropiques par les coyotes et
d’autres carnivores adaptés au milieu urbain pourrait fournir des renseignements utiles pour les efforts visant a
`promouvoir la
coexistence des humains et des espèces sauvages. [Traduit par la Rédaction]
Mots-clés : anthropique, Canis latrans, coyote, Denver, habitudes alimentaires.
Introduction
Urbanization can lead to habitat loss and fragmentation and is
one of the primary causes of species endangerment (Czech et al.
2000;McKinney 2002;Markovchick-Nicholls et al. 2008). However,
some wildlife species, including certain mammalian carnivores,
can thrive in urban environments (Ditchkoff et al. 2006;Bateman
and Fleming 2012). Carnivores that adapt to urban environments
generally tend to be small to medium-sized, have high reproduc-
tive potential, can tolerate people, and are dietary generalists
(Fuller et al. 2010). The diet of these urban carnivores usually
includes some anthropogenic food, such as cultivated plants, pets,
garbage, and roadkill (Bateman and Fleming 2012).
Coyotes (Canis latrans Say, 1823) live in nearly every major metro-
politan area in the United States (Poessel et al. 2017), and they
exemplify the characteristics of urban-adapted carnivores (Morey
et al. 2007;Gehrt and Riley 2010). Diet studies of coyotes have been
conducted in several urban areas, revealing that coyotes use both
natural food items (e.g., deer, rabbits, small mammals, and wild
Received 5 February 2016. Accepted 15 January 2017.
S.A. Poessel.* Department of Wildland Resources, Utah State University, Logan, UT 84322, USA.
E.C. Mock.
†
Department of Fish, Wildlife, and Conservation Biology, Colorado State University, Fort Collins, CO 80521, USA.
S.W. Breck. USDA Wildlife Services, National Wildlife Research Center, Fort Collins, CO 80526, USA.
Corresponding author: S.A. Poessel (email: sharpoes@gmail.com).
*Present address: Forest and Rangeland Ecosystem Science Center, U.S. Geological Survey, Boise, ID 83706, USA.
†
Present address: Denver Zoological Foundation, Denver, CO 80205, USA.
This work is free of all copyright and may be freely built upon, enhanced, and reused for any lawful purpose without restriction under copyright or database
law. The work is made available under the Creative Commons CC0 1.0 Universal Public Domain Dedication (CC0 1.0).
287
Can. J. Zool. 95: 287–297 (2017) dx.doi.org/10.1139/cjz-2016-0029 Published at www.nrcresearchpress.com/cjz on 26 February 2017.
Can. J. Zool. Downloaded from www.nrcresearchpress.com by BOISE STATE UNIVERSITY on 04/07/17
For personal use only.
fruits) and anthropogenic foods (e.g., garbage, domestic pets, pet
food, and cultivated plants) (McClure et al. 1995;Quinn 1997;
Fedriani et al. 2001;Morey et al. 2007;Gehrt and Riley 2010;
Lukasik and Alexander 2012). Some coyote diet studies have
recognized temporal fluctuations in diet (e.g., Litvaitis and Shaw
1980;Crimmins et al. 2012), but no studies have evaluated diet
relative to urban characteristics, such as housing density.
Increased coyote presence in more urban areas can lead to a rise in
encounters and conflicts with humans (Baker and Timm 1998;
Curtis and Hadidian 2010;Gehrt and Riley 2010). Reports of coyote
conflicts with humans and especially domestic pets occur throughout
the United States (Poessel et al. 2017). For example, in the Denver
metropolitan area (DMA), conflicts with pets are now common,
and behaviors that include stalking and, in rare cases, attacks on
humans have become more prevalent (Poessel et al. 2013). Con-
flicts have been found to be particularly high in the winter months in
both the DMA and Chicago (Gehrt and Riley 2010;Poessel et al.
2013), although increased conflicts during the summer months
have been reported in other North American urban areas (White
and Gehrt 2009;Lukasik and Alexander 2011). Questions about
why pet conflicts have increased and why they fluctuate season-
ally are common among natural resource managers. One question
is whether coyote–pet conflict is driven primarily by the coyote’s
desire to eat pets (i.e., predation) or whether conflict is primarily
a product of coyotes viewing pets as competitors or as a threat.
The answer to this question should allow wildlife managers to
better understand coyote behavior and motivations in highly
urbanized areas.
The principal goal of our study was to identify patterns in the
anthropogenic portion of the coyote diet in the DMA. Our primary
objectives were to (i) examine how dogs (Canis lupus familiaris L.,
1758) and cats (Felis catus L., 1758) vary seasonally in the coyote diet
and how this variation coincides with seasonal trends in conflicts
and (ii) determine how other anthropogenic food items (e.g.,
trash) in the diet vary between areas of high- and low-density
housing (hereafter, “high-density” and “low-density”). Our sec-
ondary objective was to describe the food habits of coyotes in the
DMA and to determine how coyote diet varied in high- and low-
density areas, as well as across seasons. If coyotes viewed pets
primarily as prey rather than as competitors or a threat, then we
expected pet consumption to be higher in the winter months,
mirroring documented patterns in conflict. We also expected that
consumption of trash would be more apparent in high-density
housing areas, reflecting a greater influence of anthropogenic
food sources in higher density areas.
Materials and methods
Study sites
We defined the DMA as the “Denver urban area” as delineated
by the U.S. Census Bureau (United States Census Bureau 2015). The
DMA comprises over 35 municipalities in north-central Colorado
and all or parts of seven counties (Adams, Arapahoe, Boulder,
Broomfield, Denver, Douglas, and Jefferson). It has a semiarid
climate and, during the study, had annual precipitation of 43 cm
and monthly temperatures ranging from a mean low of −10 °C in
December to a mean high of 31 °C in August (Weather Underground
2015). The DMA had a human population size of almost 2.4 million
in 2010 and an approximate size of 1760 km
2
. Located along the
front range of the Rocky Mountains of Colorado, the DMA lies
between the foothills of the Rocky Mountains to the west and
agricultural fields and grasslands to the east. Historically, the
DMA was primarily dominated by grassland habitat, but now
incorporates various land-cover types, including urban develop-
ment, woodlands, agriculture, and grasslands (Poessel et al. 2013).
Before selecting our study sites, we first defined areas of high-
and low-density housing within the DMA. Designation of these
areas was based on housing-density data obtained from the Spa-
tially Explicit Regional Growth Model (SERGoM version 3; Theobald
2005), which depicts housing density for the coterminous United
States at 100 m resolution. We defined high-density areas as those
with a housing density of less than 0.5 acre per unit and low-
density areas as those with a housing density of greater than
5 acres per unit. High-density areas contained smaller amounts of
open space with housing developments extending to the borders
of such open space, whereas low-density areas contained larger
amounts of open space with housing developments nearby. We
chose to include only high- and low-density areas (excluding
medium-density areas) to compare differences in coyote diet be-
tween two distinctive types of urban habitat. We then selected
11 parks or open-space areas known to contain coyotes near neigh-
borhoods, 5 in areas of high-density and 6 in areas of low-density
housing (Fig. 1). These open-space areas consisted of natural hab-
itat and contained both paved and dirt trails. Within each open-
space area, we then selected trails that we walked to collect coyote
scats. We selected trails used by coyotes (as indicated by the pres-
ence of coyote scats) that covered most parts of the open-space
area.
Scat collection and identification
We collected coyote scats every month for 1 year, beginning in
June 2013. We initially walked each transect to clear all scat and
then subsequently walked each transect once per month to collect
scat, although we only used scat from every other collection pe-
riod in the analyses. We received heavy snow in February that
made collecting scat difficult, but also degraded scat when the
snow melted, leaving few if any samples to be collected; thus, we
were unable to analyze scat for the month of February. We were
also unable to conduct genetic analyses during our study; hence,
we distinguished coyote scats from dog scats based on color,
shape, texture, and the amount of hair. Coyote scats tend to be
12–30 mm in diameter; red, dark brown, or gray in color; contain
bones, hair, grasses, or seeds; and possess thick, segmented cords
and tapered ends (Lukasik and Alexander 2012). Dog scats can be
many sizes, are brown or yellow in color, tend to be smoother in
shape, and lack hair, bones, and seeds. We removed any scats that
looked similar to dog scats; thus, we possibly may have discarded
coyote scats that may have contained a diet similar to a domesti-
cated dog. We did not observe any foxes or bobcats (Lynx rufus
(Schreber, 1777)) in our study sites (so we assumed they were rare),
and raccoon (Procyon lotor (L., 1758)) scats have a different shape
than coyote scats, so we were not concerned about differentiating
coyote scats with those of these species. We stored scats in a
laboratory at room temperature until analysis.
Scat analysis
We randomly chose up to 10 scats per site for the months of
August, October, December, March, April, and June and combined
these scats into a single fecal sample. If we did not have 10 scats for
a particular site (which frequently occurred), then we used all of
the scats available. We thus had one fecal sample per site per
month for analysis purposes. We chose this analysis method not
only because of limited resources to process each scat individu-
ally, but also because it allowed us to address our primary ques-
tion of whether the importance of anthropogenic items varied
between seasons or between housing densities.
We first brushed off each scat to remove any external vegeta-
tion or rocks that may have attached to the scat after defecation.
After combining scats into fecal samples, we then placed them on
a tray and dried them in an incubator at 50 °C for 24 h. We
recorded the total dry mass and then placed the fecal sample in a
soap bath for 24 h to soften the feces for analysis.
Once softened, we broke up and washed the fecal sample
through a sieve that retained all of the macroscopic material of
hair, feathers, exoskeletons, bones, seeds, vegetation, and trash.
We discarded the microscopic material (more than 50% of the
288 Can. J. Zool. Vol. 95, 2017
Published by NRC Research Press
Can. J. Zool. Downloaded from www.nrcresearchpress.com by BOISE STATE UNIVERSITY on 04/07/17
For personal use only.
fecal sample), containing unidentifiable organic matter but also
potentially earthworm chaetae and feather fragments (Reynolds
and Aebischer 1991). We washed and dried the remaining material
(i.e., macroscopic material) two more times after placing material
in a large bucket with water and mixing it with a paint mixer. Our
goal was to separate bones, seeds, and trash, which tended to sink
from the hair, feathers, vegetation, and exoskeletons that tended
to float. We then separated bones, hair, seeds, and trash and re-
corded the dry mass of each of these items. We then divided the
masses of each of these components by the total mass of the fecal
samples to calculate the percent composition. These percentages
provided a relative index of vertebrates (bones, which could in-
clude mammals, birds, reptiles, and amphibians), mammals (hair),
plants (seeds), and anthropogenic food (trash) in the coyote diet.
We subsampled hairs from each combined fecal sample to iden-
tify the mammals consumed by coyotes. For each fecal sample, we
randomly selected 100 hairs using a system of gridded cells on a
tray. We then laid the selected hairs parallel to one another on a
slide for analysis under a microscope. Using a hair identification
key, we identified each subsampled hair according to banding,
medulla, and cuticle patterns and placed each hair into the following
categories: rodent, lagomorph, deer, raccoon, red fox (Vulpes vulpes
(L., 1758)), coyote, cat, dog, soricomorph, and unidentified (Moore
et al. 1974). We further condensed these categories by combining
Fig. 1. Map of the western portion of the Denver metropolitan area (DMA) with high- and low-density housing. White space includes public
land, undeveloped private land, and areas of housing density between 0.5 and 5 acres per unit. Black crosses represent the five high-density
transects and black stars represent the six low-density transects used to collect coyote (Canis latrans) scats from 2013 to 2014.
Poessel et al. 289
Published by NRC Research Press
Can. J. Zool. Downloaded from www.nrcresearchpress.com by BOISE STATE UNIVERSITY on 04/07/17
For personal use only.
raccoon, red fox, and coyote into a “carnivores” category and cat
and dog into a “pets” category. We recorded the frequency of each
mammal category in fecal samples and calculated the percentage of
hairs in each mammal category by dividing the number of hairs in
each group by the total hairs analyzed for each sample (i.e., 100 hairs).
We then computed mean percentages for each housing density
across months and for each month across housing densities.
Because we used hairs to identify mammal species (the most
common vertebrate group in fecal samples), we did not identify
the bones in the fecal samples to species or use the mass of bones
in analyses. We identified birds by analyzing feathers (as de-
scribed below). We only found reptile skin in one scat; thus, we
did not include reptiles and amphibians as a separate category in
the coyote diet.
We separated all seeds by type based on size and appearance
and submitted them to the Colorado State Seed Laboratory (Colo-
rado State University, Fort Collins, Colorado, USA) for identifica-
tion. After identification of the seeds to plant species, we then
recorded the masses of seeds by species and the frequency of each
species in fecal samples. We then computed mean seed masses of
each species for each housing density across months and for each
month across housing densities. We further identified each plant
species as either native or non-native.
Finally, in each fecal sample, we visually estimated the amount
of feathers and exoskeletons, which were not removed from the
samples. We did not identify feathers to bird species; however,
exoskeletons were mostly intact and were easily identified as
grasshoppers or beetles. We assigned a separate score for both
feathers and exoskeletons in each sample. For feathers, we as-
signed a 0 when no feathers were visible, 1 when roughly 1%–20%
of the sample was composed of feathers, and 2 when >20% of the
sample was composed of feathers. For exoskeletons, we assigned a
0 to fecal samples with no exoskeletons present, 1 when
roughly 1%–10% of the sample was composed of exoskeletons,
and 2 when >10% of the sample contained exoskeletons. We inad-
vertently used a 10% break for exoskeletons rather than 20%; how-
ever, we proceeded with this scoring rather than simply reporting
presence or absence so that we could present patterns similar to
what we demonstrated for feathers. For both feathers and exo-
skeletons, we then calculated a weighted average of scores for
each housing density across months and for each month across
housing densities. To do so, for each housing density, we multi-
plied the number of sites within a month with a particular score
by the value of that score (i.e., 0, 1, or 2), added these three prod-
ucts, then divided this sum by the total number of sites in a month
for which we had fecal samples. These scores reflect a rough qual-
itative assessment of the mean consumption of birds and inverte-
brates by coyotes (i.e., a score of 0 reflects no consumption, a score
of 1 reflects low consumption, and a score of 2 reflects high con-
sumption).
We note that, for each food-item group described above (mam-
mals, plants, and birds or invertebrates), we used a different anal-
ysis method to determine variation in the coyote diet between
housing densities and across months. Our goal was to determine
not only how frequently a food item occurred in the diet, but also
how important each item was in the diet by using masses when
possible. These methods may vary from other coyote diet studies.
Statistical analysis
We focused our statistical analysis on our primary objective of
whether coyotes kill pets because of predation or because of com-
petition or a threat response. We conducted an analysis of vari-
ance (ANOVA) test in R version 0.99.446 (R Core Team 2015)to
determine if the amount of pet hair in the coyote diet varied by
Fig. 2. Mean (±SE) percent composition of coyote (Canis latrans) fecal samples (mass of each item divided by mass of the fecal sample) in four
diet groups (bones, seeds, hair, and trash) in high- and low-density sites in the Denver metropolitan area, from 2013 to 2014.
290 Can. J. Zool. Vol. 95, 2017
Published by NRC Research Press
Can. J. Zool. Downloaded from www.nrcresearchpress.com by BOISE STATE UNIVERSITY on 04/07/17
For personal use only.
month. Poessel et al. (2013) previously found that coyote–pet con-
flicts were more frequent during the winter months (specifically,
December and January). If coyotes pursue pets primarily for con-
sumption, then pet hair in scats should increase in December. If
no such pattern occurs, then the hypothesis that coyotes pursue
pets primarily because they perceive them as competitors or a
threat is supported.
We found very little trash in the coyote diet in both housing-
density areas (see Results), so statistical tests to address our second
primary objective, i.e., to determine how anthropogenic food
items vary between high- and low-density housing areas, were not
possible or necessary. We also did not statistically analyze specific
food items and how they varied by housing density or month (i.e.,
our secondary objective). Instead, we present these results in a
Table 1. Frequencies and percentages of mammal hairs, as well as frequencies, masses, and percentages of seeds of
plant species and trash, found in 64 coyote (Canis latrans) fecal samples (consisting of up to 10 scats per site per month)
in the Denver metropolitan area, from 2013 to 2014.
Group Species Frequency Mass (g) Percentage (%)
Hair
Rodents NA 64 43.02
Lagomorphs NA 61 25.73
Deer Mule deer, Odocoileus hemionus 35 10.29
Carnivores Raccoon, Procyon lotor 40 8.98
Carnivores Red fox, Vulpes vulpes 34 5.14
Carnivores Coyote 23 1.58
Pets Domestic cat, Felis catus 24 2.12
Pets Domestic dog, Canis lupus familiaris 25 1.21
Soricomorphs NA 10 1.39
Unknown NA 14 0.54
Total by group
Rodents 64 43.02
Lagomorphs 61 25.73
Deer 35 10.29
Carnivores 56 15.70
Pets 41 3.33
Soricomorphs 10 1.39
Unknown 14 0.54
Seeds
Non-native Russian olive, Elaeagnus angustifolia 17 30.05 6.48
Native Cactus, Opuntia spp. 11 209.28 45.16
Non-native Corn, Zea mays 8 24.47 5.28
Non-native Field bindweed, Convolvulus arvensis 7 0.32 0.07
Non-native Sorghum grain, Sorghum spp. 6 1.44 0.31
Native American plum, Prunus americana 6 24.11 5.20
Native Sunflower, Helianthus annuus 6 0.28 0.06
Non-native Grapevine, Vitis spp. 5 16.60 3.58
Native Manzanita/bearberry, genus Arctostaphylos Adans. 5 2.89 0.62
Non-native Wheat, genus Triticum L. 5 0.40 0.09
Native Chokecherry, Prunus virginiana 4 139.96 30.20
Non-native Watermelon, Citrullus lanatus (Thunb.) Matsum. & Nakai 3 0.16 0.03
Non-native Oat, Avena sativa L. 3 1.96 0.42
Native Virginia groundcherry, Physalis virginiana Mill. 2 0.45 0.10
Native Raspberry, genus Rubus L. 2 0.08 0.02
Native Flax, Linum usitatissimum L. 2 1.42 0.31
Non-native Safflower, Carthamus tinctorius L. 1 0.02 0.01
Non-native Dandelion, genus Taraxacum F.H. Wigg 1 0.06 0.01
Non-native Pigweed, genus Amaranthus L. 1 0.01 0.00
Non-native Black bindweed, Fallopia convolvulus (L.) Á. Löve 1 0.47 0.10
Non-native Hairy vetch, Vicia villosa Roth 1 0.02 0.01
Non-native Alfalfa, Medicago sativa L. 1 3.93 0.85
Unknown NA 19 5.05 1.09
Total mass 463.43 100.00
Total by group
Native 27 378.47 81.67
Non-native 35 79.91 17.24
Unknown 19 5.05 1.09
Trash
NA NA 27 21.0 0.31
Note: Group represents condensed categories for mammals and species origin for plants. Frequency represents the number of fecal
samples (out of 64) containing that food item. Mass represents the total mass of each plant species and group and of all trash found in
fecal samples. For hair, percentage is the mean percentage of hairs for each mammal species or group, calculated as the number of
hairs for that species or group divided by total hairs analyzed (i.e., 100) in each fecal sample. For seeds, percentage is the mass for each
plant species or group divided by the total mass for all plant species (463.43 g). For trash, percentage is the mass of all trash divided by
the total mass of all fecal samples analyzed (6788.66 g). NA, not available or not identified.
Poessel et al. 291
Published by NRC Research Press
Can. J. Zool. Downloaded from www.nrcresearchpress.com by BOISE STATE UNIVERSITY on 04/07/17
For personal use only.
descriptive manner by graphically evaluating patterns in the data;
the error bars in these graphs indicate the amount of variation
between housing densities and across months.
Results
Across our 11 sites, we sampled 26.2 km of trails (high-density
housing: 12.5 km, mean ± SD = 2.5 ± 1.4 km; low-density housing:
13.7 km, mean ± SD = 2.3 ± 1.3 km). We analyzed 64 fecal samples,
29 in high-density sites and 35 in low-density sites. These fecal
samples consisted of 424 scats, 210 in high-density sites (mean ±
SD = 7.2 ± 3.0) and 214 in low-density sites (mean ± SD = 6.1 ± 2.9).
Among diet items, hair had the highest percentage by mass in
combined coyote fecal samples (mean ± SE = 26.2% ± 1.7%), fol-
lowed by bones (10.4% ± 1.0%) and seeds (5.7% ± 1.5%). Percentage of
trash by mass in fecal samples was negligible (0.3% ± 0.1%). Hair
percentages by mass were higher in fecal samples in high-density
sites and seed percentages were higher in low-density sites (Fig. 2).
Based on the hair analysis, the most common mammal group
found in fecal samples was rodents (43.0% ± 3.0%), followed by
lagomorphs (25.7% ± 2.7%) and carnivores (15.7% ± 1.8%; raccoon:
9.0% ± 1.5%; fox: 5.1% ± 1.0%; coyote: 1.6% ± 0.3%; Table 1). Carnivore
consumption could have been from scavenging, predation, or
in the case of coyotes, self-grooming. Fecal samples contained a
small percentage of pets (i.e., cats and dogs; 3.3% ± 0.6%; Table 1).
Percentages of rodents and pets in fecal samples were higher in
high-density sites and percentages of deer were higher in low-
density sites (Fig. 3a). Mammals were more common in the diet in
December, March, April, and June than in August or October and
more common in the diet in high-density sites than low-density
sites throughout the year, especially in August, October, and
March (Fig. 3b). Pet hairs were least common in the diet in December
(Fig. 4), the time of year when coyote–pet conflicts were highest in
the DMA (see Fig. 3bin Poessel et al. 2013). Month was not a
significant variable in the ANOVA testing for a relationship be-
Fig. 3. Mean (±SE) percentage of hairs of mammal groups found in coyote (Canis latrans) fecal samples in high- and low-density sites (a) and
mean (±SE) mass of all hairs found in coyote fecal samples by month in high- and low-density sites (b) in the Denver metropolitan area, from
2013 to 2014. The “unknown” category in panel aincludes all unidentified mammal hairs found in coyote fecal samples during the study.
292 Can. J. Zool. Vol. 95, 2017
Published by NRC Research Press
Can. J. Zool. Downloaded from www.nrcresearchpress.com by BOISE STATE UNIVERSITY on 04/07/17
For personal use only.
tween the amount of pet hair in the diet and month (F
[5,58]
= 1.8,
P= 0.123).
Fecal samples contained 33 different species of seeds that coy-
otes either fed on directly or were present in prey items of coyotes.
Russian olive (Elaeagnus angustifolia L.), cactus (species of the genus
Opuntia Mill.), corn (Zea mays L.), field bindweed (Convolvulus arvensis L.),
sorghum (species of the genus Sorghum Moench), American plum
(Prunus americana Marshall), and sunflower (Helianthus annuus L.)
were most commonly present in fecal samples by frequency
(Table 1). Russian olive and grapevine (species of the genus Vitis L.)
were more common in high-density sites, whereas cactus, choke-
cherry (Prunus virginiana L.), and corn were more prevalent in low-
density sites by mass (Fig. 5a). The presence of seeds in fecal
samples also reflected seasonal availability, with more plants con-
sumed in August and October than in December, March, April, or
June. Fecal samples also contained a higher amount of seeds in
August and October in low-density sites than in high-density sites
(Fig. 5b), especially of native plants.
Ground beetles and grasshoppers made up the majority of exo-
skeletons found in scats. The presence of exoskeletons was high-
est in August and October and lowest in March and April (Fig. 6a).
Exoskeletons were also more common in low-density sites than
in high-density sites for most of the year (Fig. 6a). The amounts of
feathers in scats varied across months in both high- and low-
density sites and were most common in December and April
(Fig. 6b). Feathers were also more common in high-density sites
than in low-density sites for most of the year (Fig. 6b).
Discussion
Coyotes appeared to have consumed anthropogenic food items
in low amounts in our study area. Pets comprised only 3% of
mammal hairs in the coyote diet, and they were more commonly
found in high-density sites (Fig. 3a). This finding corroborates
other studies that have determined that coyotes consume domes-
tic cats and dogs in low quantities (Fedriani et al. 2001;Morey et al.
2007;Gehrt and Riley 2010;Lukasik and Alexander 2012). Murray
et al. (2015) also found that coyotes in higher density (urban) areas
consumed more pets than did coyotes in lower density (rural)
areas. In Tucson, Arizona, Grubbs and Krausman (2009) observed
19 cats that were killed by coyotes, 18 of which were consumed.
However, they did not report a dietary analysis for these coyotes,
so the importance of cats in the coyote diet in this study area is
unknown.
In the DMA, over 92% of coyote–human conflicts reported be-
tween 2003 and 2010 were incidents with pets (471; Poessel et al.
2013). These included injuries and deaths of cats and dogs caused
by coyotes, although we did not have information on how many of
these pets were killed. Hence, coyotes can be a serious threat to
the safety of domestic pets. Furthermore, coyote conflicts with
pets were more frequent during the winter months (December
and January; see Fig. 3bin Poessel et al. 2013). However, pets com-
prised only a very small percentage of the coyote diet in our study,
and pet hair did not increase in the diet during winter, possibly
indicating that coyotes usually do not consume the cats and dogs
that they kill. These results support the hypothesis that coyote
conflict with pets is primarily driven by competition or a threat
response, rather than predation. Coyotes may view cats as com-
petitors because cats will hunt rodents, a primary component of
the coyote diet, similar to the finding that coyotes also perceive
other canids, such as dogs and foxes, as competitors for food and
habitat (Gosselink et al. 2003). Winter months coincide with coy-
ote breeding season, so during this time of year, coyotes may view
dogs as potential competitors for mates, leading to increased con-
flicts with pets (Poessel et al. 2013). Additionally, pets increased in
the coyote diet in March, the time of year when births of pups
Fig. 4. Mean (±SE) percentage of hairs from pets found in coyote (Canis latrans) fecal samples in high- and low-density sites by month in the
Denver metropolitan area, from 2013 to 2014.
Poessel et al. 293
Published by NRC Research Press
Can. J. Zool. Downloaded from www.nrcresearchpress.com by BOISE STATE UNIVERSITY on 04/07/17
For personal use only.
begin, so coyotes may perceive dogs as a threat to their pups and
pets killed during this time also may be consumed.
We found only a small amount of trash in the coyote diet. Only
0.3% of fecal sample masses consisted of trash in both high- and
low-density sites. However, 27 of 64 fecal samples (42%) contained
trash items (Table 1), although we do not know how many individual
scats contained trash. Hence, although coyotes do not appear to be
consuming large amounts of garbage, they may be accessing it fre-
quently, even though during our study we did not observe coyotes
eating curbside trash or entering trash dumpsters.
We used trash to measure anthropogenic food in the coyote diet
because this human-associated item was indigestible and easily
recovered in fecal samples. However, coyotes could have con-
sumed other anthropogenic food items, such as pet food or hu-
man food that is digested and cannot be measured in scats. Thus,
fecal analysis has limitations for examining the diet of wildlife
species in urban environments. Stable isotope analysis is an alter-
native method that has been used in urban systems to examine
coyote diets (Murray et al. 2015;Newsome et al. 2015). This tech-
nique is better able to discriminate between natural prey items
and anthropogenic resources by estimating the isotopic composi-
tion of each food item and of coyotes, measured in hair or vibris-
sae. Previous studies using this method have shown that some
coyotes in urban areas will consume anthropogenic food (Murray
et al. 2015;Newsome et al. 2015), and stable isotope analysis can
report a higher prevalence of anthropogenic foods in coyote diets
than scat analysis. Therefore, if assessing anthropogenic food re-
sources in the urban coyote diet is a primary objective, then fu-
ture studies should include this technique or others like it (e.g.,
fatty acid analyses) in conjunction with fecal analysis to obtain a
more accurate assessment of coyote diets in urban areas.
The diet of coyotes in our study area varied both spatially and
temporally, and coyotes showed evidence of opportunistic forag-
ing. Generally, coyote diets consisted of more mammals in high-
density sites and more plants in low-density sites, although some
individual animal and plant species varied from this pattern. Like-
Fig. 5. Mean (±SE) mass of seeds found in coyote (Canis latrans) fecal samples in high- and low-density sites of the most common plant
species (a) and all plants by month (b) in the Denver metropolitan area, from 2013 to 2014. The “other” category in panel aincludes all other
plant species consumed by coyotes during the study.
294 Can. J. Zool. Vol. 95, 2017
Published by NRC Research Press
Can. J. Zool. Downloaded from www.nrcresearchpress.com by BOISE STATE UNIVERSITY on 04/07/17
For personal use only.
wise, mammals and birds were more common in winter and
spring (December–June), whereas plants and invertebrates were
more prevalent in late summer and autumn (August–October),
likely due to seasonal food availability. Coyotes in our study area
preferred natural habitat over developed landscapes, although
they did use residential areas, especially at night (Poessel et al.
2016); however, we do not know where coyotes were primarily
foraging.
Specifically for animals, mammal hair, particularly that of ro-
dents and rabbits, was the most prevalent diet item in coyote fecal
samples. Coyotes consumed rodents more often in high-density
sites and deer more often in low-density sites (Fig. 3a). Mule deer
(Odocoileus hemionus (Rafinesque, 1817)) require large amounts of
space, are dependent on natural areas, and therefore were more
readily available in the low-density sites than in the high-density
sites. Thus, coyotes may have increased rodents in their diet in
high-density areas because deer were not as available.
Coyote consumption of rodents followed the seasonal pattern
of higher prevalence in winter and spring, although rodents were
common in the diet throughout the year. Because every one of the 64
fecal samples contained rodent hair, this mammal group influ-
enced the overall seasonal pattern for all mammals. Deer hair was
most common in fecal samples in June, after the birthing period
of fawns when they are vulnerable to coyote predation. However,
coyotes fed on deer throughout the year, likely available as car-
rion from roadkill. Lagomorphs were more prevalent in the diet in
late autumn and early winter than in spring and summer. Morey
et al. (2007) also found that lagomorphs occurred in coyote diets in
lower frequencies during the summer, suggesting that rabbits are
better able to avoid capture by coyotes during a time of increased
vegetation growth. Furthermore, fruit is more available during
summer, so coyotes may be switching from animals to plants
during this time.
Specifically for plants, coyotes consumed Russian olive (non-
native) and grapevine (non-native) more often in high-density
sites and cactus (native), chokecherry (native), and corn (non-
native) more often in low-density sites (Fig. 5a). Russian olive, a
highly invasive species, has become naturalized throughout the
western United States and is primarily found in riparian areas
(Shafroth et al. 1995). The high-density sites in our study area were
located near riparian areas, and coyotes used riparian areas fre-
quently (Poessel et al. 2016), leading to a high prevalence of this
Fig. 6. Weighted average scores of exoskeletons (a) and feathers (b) found in coyote (Canis latrans) fecal samples by month in high- and
low-density sites in the Denver metropolitan area, from 2013 to 2014.
Poessel et al. 295
Published by NRC Research Press
Can. J. Zool. Downloaded from www.nrcresearchpress.com by BOISE STATE UNIVERSITY on 04/07/17
For personal use only.
plant in coyote scats at these sites. Grapevines are widely culti-
vated in gardens, likely resulting in a large number of these plants
in residential areas and a high occurrence of these seeds in coyote
scats at high-density sites. Cactus and chokecherry are both native
species to Colorado and were common in low-density sites where
coyotes were more likely to consume native fruits. Finally, corn
may have been more common in low-density sites because of the
higher presence of deer in these areas; we speculate that residents
in these sites may be placing corn outside to feed deer, which can be
accessed by coyotes. Corn fields also are present near the low-
density sites, and we occasionally observed coyotes in these fields.
Coyote consumption of most plant species followed the same
seasonal pattern of high prevalence in summer or autumn. Russian
olive was found in the coyote diet in higher masses in August,
October, and December. However, corn was most commonly
found in the diet in early winter (December), a time of year when
coyotes might be more food-stressed (Bekoff and Wells 1981).
Corn, as well as sorghum, is found in different brands of birdseed,
and residents of the DMA may be leaving out dried or cracked corn
and birdseed in bird feeders. Thus, coyotes likely were utilizing
this food source to supplement their diet in the winter. Although
we can only speculate, the presence of corn in the coyote diet is
likely anthropogenically driven, whether from residents feeding
deer, placing corn in bird feeders, or cultivating corn plants.
Russian olive was the most frequently consumed plant, occur-
ring in 17 of 64 fecal samples (27%; Table 1). Although coyotes
consumed 14 different non-native plant species, the largest vol-
ume of seeds in the coyote diet consisted of native species (82% of
total seed mass, although the majority of this mass was from only
two species; Table 1). Hence, coyotes may be an important dis-
perser of not only Russian olive, but also a variety of native plant
species in this highly urbanized area.
Conclusions
Our scat analysis revealed that pets consisted of only a small
percentage of the coyote diet, and pet consumption decreased
when coyote–pet conflicts increased, which supports the hypoth-
esis that coyote conflict with pets is primarily driven by competi-
tion or a threat response, rather than predation. Coyotes in urban
areas have a wide variety of foods available to them, so they do not
need to consume pets. However, territoriality in coyotes is strong
and they will remove perceived threats to their ability to survive
and reproduce. Hence, pet owners should be diligent in their
efforts to protect their pets by not letting pets outside unsuper-
vised, including in their yards, by keeping pets on leash when
walking them in natural areas, and by being especially aware of
coyotes during winter. Our work further revealed that trash did
not constitute a significant portion of the diet of coyotes, but our
analysis technique was not sensitive to detecting certain anthro-
pogenic food sources that are difficult to detect with scat analysis.
Urban residents can reduce the amount of such food available to
coyotes by securing trash and removing pet food from outdoor
areas. As reports of wildlife conflicts in urban areas continue to
increase, understanding how urban-adapted species, such as coy-
otes, use anthropogenic resources may provide insight that can be
used to promote coexistence between humans and wildlife.
Acknowledgements
We thank the many technicians and volunteers who assisted
with scat collection and laboratory work, including R. Much,
S. Peterson, M. Annis, A. Pepper, K. Loos, R. Maison, T. Greeley,
S. Smith, R. Galvan, L. Bowie, M. Casey, J. Hawkins, C. Mitchell,
B. Tomlin, P. Mills, R. Deangelis, and C. Kerr. Funding and logisti-
cal support were provided by the U.S. Department of Agriculture,
Wildlife Services, National Wildlife Research Center.
References
Baker, R.O., and Timm, R.M. 1998. Management of conflicts between urban
coyotes and humans in southern California. In Proceedings of the 18th Ver-
tebrate Pest Conference, Costa Mesa, Calif., 2–5 March 1998. Edited by
R.O. Baker and A.C. Crabb. University of California, Davis. pp. 299–312.
Bateman, P.W., and Fleming, P.A. 2012. Big city life: carnivores in urban envi-
ronments. J. Zool. 287: 1–23. doi:10.1111/j.1469-7998.2011.00887.x.
Bekoff, M., and Wells, M.C. 1981. Behavioural budgeting by wild coyotes: the
influence of food resources and social organization. Anim. Behav. 29: 794–
801. doi:10.1016/S0003-3472(81)80013-9.
Crimmins, S.M., Edwards, J.W., and Houben, J.M. 2012. Canis latrans (coyote)
habitat use and feeding habits in central West Virginia. Northeast Nat. 19:
411–420. doi:10.1656/045.019.0304.
Curtis, P.D., and Hadidian, J. 2010. Responding to human–carnivore conflicts in
urban areas. In Urban carnivores: ecology, conflict, and conservation.
Edited by S.D. Gehrt, S.P.D. Riley, and B.L. Cypher. The Johns Hopkins University
Press, Baltimore, Md. pp. 201–211.
Czech, B., Krausman, P.R., and Devers, P.K. 2000. Economic associations among
causes of species endangerment in the United States. BioScience, 50: 593–
601. doi:10.1641/0006-3568(2000)050[0593:EAACOS]2.0.CO;2.
Ditchkoff, S.S., Saalfeld, S.T., and Gibson, C.J. 2006. Animal behavior in urban
ecosystems: modifications due to human-induced stress. Urban Ecosyst. 9:
5–12. doi:10.1007/s11252-006-3262-3.
Fedriani, J.M., Fuller, T.K., and Sauvajot, R.M. 2001. Does availability of anthro-
pogenic food enhance densities of omnivorous mammals? An example with
coyotes in southern California. Ecography, 24: 325–331. doi:10.1111/j.1600-0587.
2001.tb00205.x.
Fuller, T.K., DeStefano, S., and Warren, P.S. 2010. Carnivore behavior and ecol-
ogy, and relationship to urbanization. In Urban carnivores: ecology, conflict,
and conservation. Edited by S.D. Gehrt, S.P.D. Riley, and B.L. Cypher. The Johns
Hopkins University Press, Baltimore, Md. pp. 13–19.
Gehrt, S.D., and Riley, S.P.D. 2010. Coyotes (Canis latrans). In Urban carnivores:
ecology, conflict, and conservation. Edited by S.D. Gehrt, S.P.D. Riley, and
B.L. Cypher. The Johns Hopkins University Press, Baltimore, Md. pp. 79–95.
Gosselink, T.E., van Deelen, T.R., Warner, R.E., and Joselyn, M.G. 2003. Temporal
habitat partitioning and spatial use of coyotes and red foxes in east-central
Illinois. J. Wildl. Manage. 67: 90–103. doi:10.2307/3803065.
Grubbs, S.E., and Krausman, P.R. 2009. Observations of coyote–cat interactions.
J. Wildl. Manage. 73: 683–685. doi:10.2193/2008-033.
Litvaitis, J.A., and Shaw, J.H. 1980. Coyote movements, habitat use, and food
habits in southwestern Oklahoma. J. Wildl. Manage. 44: 62–68. doi:10.2307/
3808351.
Lukasik, V.M., and Alexander, S.M. 2011. Human–coyote interactions in Calgary,
Alberta. Hum. Dimens. Wildl. 16: 114–127. doi:10.1080/10871209.2011.544014.
Lukasik, V.M., and Alexander, S.M. 2012. Spatial and temporal variation of coyote
(Canis latrans) diet in Calgary, Alberta. Cities Environ. 4: Article 8. doi:10.15365/
cate.4182011.
Markovchick-Nicholls, L., Regan, H.M., Deutschman, D.H., Widyanata, A.,
Martin, B., Noreke, L., and Hunt, T.A. 2008. Relationships between human
disturbance and wildlife land use in urban habitat fragments. Conserv. Biol.
22: 99–109. doi:10.1111/j.1523-1739.2007.00846.x. PMID:18254856.
McClure, M.F., Smith, N.S., and Shaw, W.W. 1995. Diets of coyotes near the
boundary of Saguaro National Monument and Tuscon, Arizona. Southwest.
Nat. 40: 101–104.
McKinney, M.L. 2002. Urbanization, biodiversity, and conservation. BioScience,
52: 883–890. doi:10.1641/0006-3568(2002)052[0883:UBAC]2.0.CO;2.
Moore, T.D., Spence, L.E., Dugnolle, C.E., and Hepworth, W.P. 1974. Identifica-
tion of the dorsal guard hairs of some mammals of Wyoming. Wyoming
Game and Fish Department Bulletin No. 14.
Morey, P.S., Gese, E.M., and Gehrt, S. 2007. Spatial and temporal variation in the
diet of coyotes in the Chicago metropolitan area. Am. Midl. Nat. 158: 147–161.
doi:10.1674/0003-0031(2007)158[147:SATVIT]2.0.CO;2.
Murray, M., Cembrowski, A., Latham, A.D.M., Lukasik, V.M., Pruss, S., and
St. Clair, C.C. 2015. Greater consumption of protein-poor anthropogenic food
by urban relative to rural coyotes increases diet breadth and potential for
human–wildlife conflict. Ecography, 38: 1235–1242. doi:10.1111/ecog.01128.
Newsome, S.D., Garbe, H.M., Wilson, E.C., and Gehrt, S.D. 2015. Individual vari-
ation in anthropogenic resource use in an urban carnivore. Oecologia, 178:
115–128. doi:10.1007/s00442-014-3205-2. PMID:25669449.
Poessel, S.A., Breck, S.W., Teel, T.L., Shwiff, S., Crooks, K.R., and Angeloni, L.
2013. Patterns of human–coyote conflicts in the Denver Metropolitan Area.
J. Wildl. Manage. 77: 297–305. doi:10.1002/jwmg.454.
Poessel, S.A., Breck, S.W., and Gese, E.M. 2016. Spatial ecology of coyotes in the
Denver metropolitan area: influence of the urban matrix. J. Mammal. 97:
1414–1427. doi:10.1093/jmammal/gyw090.
Poessel, S.A., Gese, E.M., and Young, J.K. 2017. Environmental factors influencing
the occurrence of coyotes and conflicts in urban areas. Landsc. Urban Plan.
157: 259–269. doi:10.1016/j.landurbplan.2016.05.022.
Quinn, T. 1997. Coyote (Canis latrans) food habits in three urban habitat types of
western Washington. Northwest Sci. 71:1–5.
R Core Team. 2015. R: a language and environment for statistical computing.
Version 0.99.446 [computer program]. R Foundation for Statistical Comput-
ing, Vienna, Austria. Available from https://www.r-project.org.
296 Can. J. Zool. Vol. 95, 2017
Published by NRC Research Press
Can. J. Zool. Downloaded from www.nrcresearchpress.com by BOISE STATE UNIVERSITY on 04/07/17
For personal use only.
Reynolds, J.C., and Aebischer, N.J. 1991. Comparison and quantification of carni-
vore diet by faecal analysis: a critique, with recommendations, based on a
study of the Fox Vulpes vulpes. Mammal Rev. 21: 97–122. doi:10.1111/j.1365-2907.
1991.tb00113.x.
Shafroth, P.B., Auble, G.T., and Scott, M.L. 1995. Germination and establishment
of the native plains cottonwood (Populus deltoides Marshall subsp. monilifera)
and the exotic Russian-olive (Elaeagnus angustifolia L.). Conserv. Biol. 9: 1169–
1175. doi:10.1046/j.1523-1739.1995.9051159.x-i1.
Theobald, D.M. 2005. Landscape patterns of exurban growth in the U.S.A. from
1980 to 2020. Ecol. Soc. 10: 32.
United States Census Bureau. 2015. 2010 census urban and rural classification
and urban area criteria. Available from https://www.census.gov/geo/
reference/ua/urban-rural-2010.html [accessed 13 May 2016].
Weather Underground. 2015. Weather history for Centennial, Colo. The
Weather Channel, LLC. Available from http://www.wunderground.com/
history/airport/KAPA/2013/6/1/MonthlyHistory.html [accessed 28 December
2015].
White, L.A., and Gehrt, S.D. 2009. Coyote attacks on humans in the United States and
Canada. Hum. Dimens. Wildl. 14: 419–432. doi:10.1080/10871200903055326.
Poessel et al. 297
Published by NRC Research Press
Can. J. Zool. Downloaded from www.nrcresearchpress.com by BOISE STATE UNIVERSITY on 04/07/17
For personal use only.