Skip to main content
Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Age-dependent relationships among diet, body condition, and Echinococcus multilocularis infection in urban coyotes

  • Scott Sugden ,

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing

    Affiliations Department of Biological Sciences, University of Alberta, Edmonton, Alberta, Canada, Department of Natural Resource Sciences, McGill University, Sainte-Anne-de-Bellevue, Quebec, Canada

  • Deanna K. Steckler,

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing

    Affiliation Department of Biological Sciences, University of Alberta, Edmonton, Alberta, Canada

  • Dana Sanderson,

    Roles Investigation, Methodology, Writing – review & editing

    Affiliation Department of Biological Sciences, MacEwan University, Edmonton, Alberta, Canada

  • Bill Abercrombie,

    Roles Investigation, Resources, Writing – review & editing

    Affiliation Animal Damage Control, Bushman Inc., Sherwood Park, Alberta, Canada

  • Duncan Abercrombie,

    Roles Investigation, Writing – review & editing

    Affiliation Animal Damage Control, Bushman Inc., Sherwood Park, Alberta, Canada

  • M. Alexis Seguin,

    Roles Formal analysis, Investigation, Resources, Writing – review & editing

    Affiliation IDEXX Laboratories, Inc., Westbrook, Maine, United States of America

  • Kyra Ford,

    Roles Investigation, Writing – review & editing

    Affiliation Department of Biological Sciences, University of Alberta, Edmonton, Alberta, Canada

  • Colleen Cassady St. Clair

    Roles Conceptualization, Funding acquisition, Methodology, Resources, Supervision, Writing – original draft, Writing – review & editing

    Affiliation Department of Biological Sciences, University of Alberta, Edmonton, Alberta, Canada


Urban coyotes (Canis latrans) in North America increasingly exhibit a high prevalence of Echinococcus multilocularis, a cestode of recent and rising public health concern that uses rodents as intermediate hosts and canids as definitive hosts. However, little is known about the factors that drive the high urban prevalence of this parasite. We hypothesized that the diet of urban coyotes may contribute to their higher E. multilocularis infection prevalence via either (a) greater exposure to the parasite from increased rodent consumption or (b) increased susceptibility to infection due to the negative health effects of consuming anthropogenic food. We tested these hypotheses by comparing the presence and intensity of E. multilocularis infection to physiological data (age, sex, body condition, and spleen mass), short-term diet (stomach contents), and long-term diet (δ13C and δ15N stable isotopes) in 112 coyote carcasses collected for reasons other than this study from Edmonton, Alberta and the surrounding area. Overall, the best predictor of infection status in this population was young age, where the likelihood of infection decreased with age in rural coyotes but not urban ones. Neither short- nor long-term measures of diet could predict infection across our entire sample, but we found support for our initial hypotheses in young, urban coyotes: both rodent and anthropogenic food consumption effectively predicted E. multilocularis infection in this population. The effects of these predictors were more variable in rural coyotes and older coyotes. We suggest that limiting coyote access to areas in which anthropogenic food and rodent habitat overlap (e.g., compost piles or garbage sites) may effectively reduce the risk of infection, deposition, and transmission of this emerging zoonotic parasite in urban areas.


Generalist species thrive in urban environments due to their ability to exploit anthropogenic resources [1,2], but urban habitat use can also alter ecological relationships in ways that increase the transmission and spread of parasites. For example, the resources in urban environments, including available habitat and anthropogenic food subsidies, are often concentrated in space or time, causing hosts or vectors to aggregate [35]. In addition, nutrient-poor diets and chemical pollutants in urban areas can decrease host health and immune function, thereby increasing susceptibility to parasites [6]. Because urban areas harbor unique assemblages of species that may not coexist elsewhere, urban-dwelling hosts may also be exposed to new parasites, and, conversely, parasites may encounter novel hosts such as humans or pets [7,8]. Many of the parasites that occur in urban wildlife are zoonotic, such as sarcoptic mange [9,10], Toxoplasmosis [11], and the raccoon roundworm [12]. Increasing populations of urban generalists presents new public health risks because they necessarily increase human exposure to the zoonotic parasites of these species [6,13,14].

One such zoonotic parasite of increasing concern in Canada is the trophically transmitted tapeworm Echinococcus multilocularis [15]. This tapeworm uses rodents (e.g., voles and mice) and canids (e.g., foxes and coyotes) as intermediate and definitive hosts, respectively [16]. Humans may become infected by accidentally ingesting parasite eggs, which most commonly occurs either via transmission from pets that have consumed infected feces or rodents or from the consumption of unwashed produce contaminated with infected feces [17,18]. In humans, E. multilocularis infections can lead to alveolar echinococcosis, a potentially fatal zoonosis [19]. Multiple strains of E. multilocularis occur throughout the northern hemisphere [20], and while the historical North American strain was not particularly infectious to humans [21], the more virulent European strain has recently arrived in Canada, where it has become widespread among wolves, foxes, and coyotes [22,23]. The increasing prevalence of this virulent E. multilocularis strain raises significant human health concerns because some definitive hosts, especially coyotes, are common inhabitants of urban areas across the continent.

The majority of locally acquired human E. multilocularis infections have been reported from the province of Alberta [24], where the parasite appears to be highly prevalent among coyotes in urban areas. Indeed, several recent studies by us and other authors have reported that over half of Edmonton’s urban coyote population carries E. multilocularis [2527], an infection prevalence 50% higher than that of coyotes in the surrounding area [27]. However, the reasons for this uniquely high prevalence of E. multilocularis in Edmonton’s coyote population remain unknown. One potential driver is age: juveniles are generally more susceptible to parasites than adults across a range of host and parasite taxa [2830], and our previous urban study population was younger, on average, than rural coyotes [27]. Several studies have indeed found that E. multilocularis parasitism can be age-dependent [3134]. However, another potentially important driver of urban infections is diet: Edmonton’s urban coyotes consume more anthropogenic food and rodents and less overall protein than their rural counterparts [27,35,36], which may alter the transmission dynamics of a trophically transmitted parasite like E. multilocularis.

The first diet-related mechanism that may contribute to the high prevalence of E. multilocularis in Edmonton coyotes is greater exposure to the parasite via increased consumption of infected rodents. Studies of urban foxes in Europe have shown that E. multilocularis infection prevalence parallels rodent availability [37,38]; other European studies have found that E. multilocularis is less prevalent in urban foxes than in semi-urban or rural populations [32,39,40] because urban areas provide less rodent habitat and, given the availability of anthropogenic food subsidies, urban predators are less dependent on rodents [41,42]. However, Edmonton is unlike European urban areas because the city contains an abundance of rodent habitat stemming from its large, contiguous river valley, low human density, and high percentage of undeveloped land [43]. Coyotes in Edmonton correspondingly appear to specialize on rodents [35,36], especially voles [44], which are common intermediate hosts for E. multilocularis [45]. Edmonton’s unique setting and the known dietary habits of its coyote population suggest that Edmonton coyotes may have unusually frequent exposure to E. multilocularis.

A second diet-related mechanism that could drive the prevalence of E. multilocularis in Edmonton coyotes is higher susceptibility to the parasite due to the negative health effects of consuming nutrient-poor anthropogenic food [27,36,46]. Consumption of anthropogenic food has been associated with increased parasite burdens in several wildlife species due to suppressed immune function [47], poor nutrition [48], and aggregation of infected hosts [46]. Our previous work showed that protein-poor diets can alter the coyote gut microbiome and reduce body condition [27], which might increase susceptibility to infection [5]. We have also shown that coyotes in Edmonton aggregate at large, unsecured compost piles, which can contain immune-suppressing mycotoxins [49]; coyote scats collected from these compost piles were 10 times more likely to contain tapeworm eggs than scats collected elsewhere in the city [49]. However, coyotes that deposited parasite-rich feces at these compost sites could have exhibited high infections either because of the immune-suppressing effects of consuming compost, as we hypothesize here, or because they were exploiting an abundance of rodents that were also attracted to compost, as we hypothesize above. Disentangling these two mechanisms requires a more detailed analysis of the dietary factors that drive E. multilocularis infections in urban coyotes, which can then inform targeted strategies for reducing infection rates and the likelihood of human exposure.

Here, we assessed which of these two proposed dietary mechanisms (i.e., greater exposure to infected prey or greater susceptibility to infection) best explains the uniquely high prevalence of E. multilocularis infections in Edmonton’s urban coyote population in the context of other factors (e.g., age, sex, and body condition) that can also moderate infection dynamics. We measured both the presence and intensity of E. multilocularis infections in the intestines of coyotes collected within and outside the City of Edmonton. To quantify both short- and long-term diet, we used stomach contents and carbon (δ13C) and nitrogen (δ15N) stable isotopes. Based on these measures, our first hypothesis (increased exposure) would be supported by increased rodent consumption (indicated via stomach contents or δ15N values) in infected coyotes, and our second hypothesis (greater susceptibility) would be supported by increased anthropogenic food consumption (indicated via stomach contents or δ13C values). Our results target key dietary and contextual factors underlying the prevalence of an emerging zoonotic parasite in Alberta and provide an important foundation for wildlife managers looking to mitigate the public health risks associated with E. multilocularis in urban wildlife.

Materials and methods

Study area

Coyote carcasses were collected between 2017 and 2020 from areas within (“urban”) and outside (“rural”) the City of Edmonton (Fig 1; S1 Table in S1 Appendix). Located in central Alberta, Edmonton has a population of approximately 1,000,000 residents and is bisected by the North Saskatchewan River, allowing for an extensive network of urban green spaces along the river valley and associated ravine network. These urban green spaces provide abundant habitat for coyotes as well as many of their small mammal prey species. Our rural study area was located south of Edmonton and encompassed several lakes and natural areas around the municipalities of Leduc and Beaumont. Climatically, the Edmonton region is characterized by a prairie climate with warm summers (high temperatures of 21°C to 23°C) and cold winters (average high temperatures of −3°C to −7°C).

Fig 1. Map of the study area, with inset showing the location of Edmonton in central Alberta, Canada.

Samples were collected from areas within (“urban”) and outside (“rural”) the city, as highlighted in the map. Roads (purple) and water bodies (blue) are shown for reference.

Sample collection and necropsy

All coyote carcasses were donated by outside sources following one of population management, roadkill, or conflict with humans. Carcasses were stored at -80°C for at least five days to neutralize the zoonotic risk of E. multilocularis [50] and then transferred to -20°C until necropsy. The carcasses used in this study include the carcasses used by Sugden et al. [27] as well as additional samples collected in subsequent years. We therefore followed the same necropsy procedures described by Sugden et al. [27]: we measured body size (mass, length, and girth), calculated the kidney fat index (KFI) as a metric for body condition [51], and divided spleen mass (in grams) by body mass (in kilograms) to obtain a coarse measure of immune system function [52]. We removed the small intestine for parasite analysis, the stomach for diet analysis, the outer hind claws for stable isotope analysis, and the lower jaw for age determination. Age was later determined by counting the cementum annuli of the lower canine teeth; teeth were fixed, decalcified, sectioned, stained, and visualized following published methods [53] as described by Sugden et al. [27].

E. multilocularis detection and quantification

We used a combination of molecular and morphological methods to detect the presence and intensity of E. multilocularis infections in coyote intestines. Following the scraping-filtration-counting approach recommended by Gesy et al. [54], we first divided the small intestine into four equal lengths. Because E. multilocularis is less abundant in the anterior small intestine [55], we removed a ~0.5 g mucosal scraping from the second quarter for molecular analysis before rinsing and filtering each of the four intestinal quarters. To ensure that our molecular and morphological tests were spatially paired, we did not combine the filtrates from each intestinal quarter. Intestinal contents trapped on a 150 μm sieve were resuspended in distilled H2O (10–20 ml) and then fixed in 70% ethanol to a total volume of 60 ml. The sample from the second intestinal quarter was divided into four 15-ml aliquots for morphological parasite analysis.

For molecular analysis, we used a commercially available real-time polymerase chain reaction (qPCR) assay (Echinococcus RealPCRTM Panel, IDEXX Laboratories, Inc.) to detect the presence of E. multilocularis in the small intestine. Mucosal scrapings were stored in sterile microtubes at -20°C and transported to IDEXX Laboratories, Inc. (West Sacramento, CA). Real-time PCR was performed using the LightCycler 480 system (Roche) with proprietary forward and reverse primers and hydrolysis probes. The E. multilocularis qPCR assay targets a ribosomal RNA sequence between the cox1 and cox2 genes. Real-time PCR was performed with seven quality controls as described in S1 Appendix.

For morphological parasite analyses, we counted worms in intestinal samples to determine the presence and intensity of E. multilocularis infections. We used a dissection microscope to observe one 15-ml aliquot of the contents from the second quarter of the intestine. All parasite scolexes were morphologically identified and counted [56]. For three samples that contained >1,000 worms in the first 1 ml of the aliquot, we estimated the total number of worms in the remaining sample by multiplying the count by 15 ml. All the scolexes we identified appeared to be E. multilocularis; however, because coyotes can also be infected by the related parasite E. canadensis and scolexes of the two species are difficult to visually discriminate [57], we only recorded worm counts for carcasses that tested positive for E. multilocularis via qPCR.

Conversely, due to the high sensitivity of the qPCR test, 22 samples tested positive for E. multilocularis but contained no visible worms in the second quarter of the intestine. We attributed this to either the presence of an exceptionally light infection (e.g., a single worm) or remnant DNA from a new or recently concluded infection. To ensure that our analyses targeted only animals with active infections (i.e., significant enough to be shedding eggs), we defined “biologically active” infections as those that were both qPCR-positive for E. multilocularis and had visual evidence of Echinococcus spp. scolexes. Based on these criteria, the 22 samples that were qPCR-positive but contained no visible worms were considered “uninfected” for all statistical analyses. A more detailed analysis of these samples is provided in S2 Appendix.

Coyote diet analysis

We used measurements of short- and long-term diet to test our hypotheses that diet influences E. multilocularis infection patterns. Short-term diet was assessed using stomach contents, which provide evidence of the last meal(s) eaten prior to death. We removed all contents from the stomach and then measured the total stomach content volume (in ml) via water displacement [58]. Contents were then rinsed and sorted into nine mutually exclusive diet components: four prey items (ungulates, rodents, meso-mammals, and birds), two types of anthropogenic food (digestible and indigestible), vegetation, insects, and native fruit. We defined anthropogenic food as items that originated from human food production or waste processes; items of at least minimal nutritive value (e.g., apples, dog kibble, chicken bones, birdseed) were considered digestible, and non-nutritive trash items (e.g., food wrappers, plastic, leather scraps) were considered indigestible. The volume of each component was quantified by water displacement, as before. Diet components with non-zero volumes less than our minimum measurement sensitivity (0.5 ml) were assigned a volume of 0.1 ml.

To assess diet over a longer term, we measured carbon (δ13C) and nitrogen (δ15N) stable isotope signatures in claw samples, which represent diet over the past 6–8 months [59,60]. In general, higher δ13C values are associated with corn-based (i.e., anthropogenic) food consumption [61]. Higher δ15N values indicate increased protein (i.e., prey) consumption. While this measure does not specifically test our hypothesis about rodents, we assumed that higher δ15N values for urban coyotes would reflect increased rodent consumption because rodents are the most common prey item for coyotes in urban areas [36]. Claw samples were prepared for stable isotope analysis following previously described methods [27].

Statistical analysis

All statistical analyses were conducted using R version 3.6.3 [62] and are described fully in S1 Appendix. Because worm counts were highly right-skewed (skewness 6.60, kurtosis 50.27), we truncated worm counts at 10,000, added a pseudo-count of 1, and then natural log-transformed the counts, producing a response variable that approximated a negative binomial distribution (S1 Fig in S1 Appendix). To reduce collinearity among our measures of body condition, we limited our assessment of body condition to (i) axis scores on the first principal component of a principal components analysis (PCA) performed on mass, body size, girth, and KFI (S2 Fig, S2 Table in S1 Appendix), and (ii) spleen mass (normalized by body mass), which was retained separately because it was not correlated with the other measures (S2 Fig in S1 Appendix). Insects were excluded from further analysis due to their low occurrence rate in our sample (n = 7 individuals), and we measured diet diversity by calculating the Shannon diversity index for each coyote’s stomach contents. Finally, we distinguished “juvenile” from “adult” coyotes based on whether they were younger or older, respectively, than the median age of our sample. This distinction was only used for the convenience of data visualization; numerical ages were used for all statistical analysis.

We then used univariate regressions to test for broad differences in infection status (presence/absence of a biologically active infection) and intensity (natural log-transformed worm counts) as a function of location, age, sex, body condition, stomach contents, diet diversity, and stable isotope values. Infection status and intensity were modeled with logistic and negative binomial regressions, respectively. For each response/predictor pair, we assessed the significance of their univariate relationship by comparing each regression to the corresponding null model using a likelihood ratio test, with significance defined at p < 0.05. All predictors were scaled to mean zero and unit variance prior to analysis, and we weighted urban samples proportionally to their abundance to account for their smaller population size in our sample. Results from the univariate regression models were further verified using chi-squared tests, Wilcox rank-sum tests, and Spearman’s rank correlations, as appropriate.

Because coyote diet and infection status have been shown to vary with both location and age in this sample [27,35] and elsewhere [3133,63], we also tested interaction effects to determine whether the effect of a variable differed between urban and rural areas or across different ages, as described in S1 Appendix. In brief, we ran three additional models for each response/predictor pair, allowing each predictor to interact with (i) age, (ii) location, and (iii) the interaction between age and location. We determined the significance of each interaction using a likelihood ratio test, as before, where each model was compared to the model in which the focal predictor did not appear.

To identify the overall best dietary predictors of infection status and intensity, we extended our interaction models by allowing multiple measures of diet to appear in the same model. For these models, we chose only the dietary predictors that explicitly tested our hypotheses about rodent and/or anthropogenic food consumption. Given the exploratory nature of our questions, we chose to evaluate all model subsets using the R package MuMIn [64] and averaged predictor coefficients across all models with a delta Akaike information criterion (ΔAICC) < 2. Before averaging, model coefficients were standardized by their partial standard deviation and adjusted based on model weight [65]. We quantified the predictive accuracy of the final models using Cohen’s kappa (for the logistic regression) and root mean-squared error (for the negative binomial regression) calculated from k-fold cross-validation with k = 5.

We lastly used ordination approaches to assess whether infection status or intensity were better predicted by a cluster of diet components than by individual diet items. For these analyses, we excluded ten coyotes with empty stomachs, converted diet component volumes to relative abundances, and then ordinated the stomach content data using the Bray-Curtis distance. Differences between diet composition and each of infection prevalence and intensity were assessed using permutational multivariate analyses of variance (PERMANOVAs) that also included age and location as predictors. All ordination analyses were implemented using the R package vegan [66].


We obtained 112 coyote carcasses from population control programs in rural areas (n = 66), roadkill in the City of Edmonton (n = 35), and lethal management due to conflict with humans (urban: n = 6; rural = 5) (S1 Table in S1 Appendix:), for a total of n = 71 rural and n = 41 urban samples. Coyotes ranged in age from 0.26 to 11.43 years (mean: 2.47±2.32, median: 1.78), with a similar age distribution in both urban and rural areas (S3 Fig in S1 Appendix). Rodents and anthropogenic food were each present in approximately 50% of stomachs, with voles being the most common rodent prey item (S3 Table in S1 Appendix). As in previous studies of Edmonton coyotes [27,35,46], stomach contents and stable isotope values indicated that urban coyotes consumed significantly more digestible anthropogenic food and smaller prey items (e.g., rodents) than rural coyotes (S3-S5 Tables in S1 Appendix). Conversely, rural coyotes consumed significantly more total prey, including meso-mammals and ungulates, as well as significantly more indigestible anthropogenic food (S3-S5 Tables in S1 Appendix). In both urban and rural areas, juvenile coyotes consumed a more diverse variety of smaller prey items, whereas older coyotes consumed more ungulates (S3-S5 Tables in S1 Appendix).

E. multilocularis infection status and intensity

We detected biologically active E. multilocularis infections in 48.2% of the coyotes we examined, though 69.9% tested positive via qPCR (see S2 Appendix). The mean parasite load (± SD) was 1,351 ± 6,391 worms (range: 0–55,000); three urban individuals each carried >10,000 worms and accounted for 72% of all E. multilocularis scolexes we counted. Infection prevalence and intensity varied with both location and age: urban coyotes had a 28% higher infection prevalence than rural coyotes and carried an average of eight times more worms (Fig 2), though neither difference was significant in univariate regression models (Table 1). Infection prevalence and intensity were also significantly higher in younger coyotes and coyotes in poorer condition (Fig 2; Table 1), though we note that body condition was closely correlated with age (S2 Fig in S1 Appendix). We found no significant sex-based differences in either measure of infection (Table 1), and no single measure of diet—stomach content volumes, diet diversity, or stable isotope values—was able to predict infection status or intensity across our entire sample (Table 1; S1 Table in S3 Appendix). Similarly, neither infection status nor intensity were associated with multivariate representations of stomach contents in ordination analyses (S1 Fig, S2 Table in S3 Appendix).

Fig 2.

E. multilocularis prevalence and intensity compared between sampling locations (urban vs. rural, left column), age classes (juvenile vs. adult, center column), and health classes (above or below the median health score, right column). Prevalence is represented as the percentage of the sampled population, and infection intensity is represented as worm counts with a log-transformed y-axis. Juvenile and adult coyotes were distinguished as being younger or older, respectively, than the median age of our sample (1.78 yr). Significant differences (p < 0.05 in regression models; see Methods and Table 1) are represented with asterisks (*).

Table 1. Results from likelihood ratio tests assessing relationships between contextual and dietary variables and each of infection status and intensity.

A significant two-way interaction between location and age suggested that infection status and intensity declined with age in rural coyotes but not urban coyotes (Fig 3A and 3B; Table 1). We excluded our composite measure of body condition from our interaction analyses because of its correlation with age; we instead assessed spleen size alone, which did not exhibit this correlation. The relationship between spleen size and each of infection status and intensity exhibited a significant three-way interaction with location and age (Table 1). Specifically, young animals (< 1 yr) with larger spleens were less likely to be infected and carried fewer worms, but older animals with larger spleens were more likely to be infected and carried more worms (Fig 3C; Table 1). This reversing trend with age was consistent between urban and rural areas, though the age at which the relationship reversed was higher in urban coyotes (Fig 3C).

Fig 3. Interactions between age, location, and spleen mass (normalized by body mass, g kg-1) as predictors of E. multilocularis infection status and intensity.

Infection status (a) and intensity (b) both declined significantly with age in rural coyotes but less so in urban coyotes. (c) Spleen mass exhibited a significant three-way interaction with age and location. The figure shows predicted relationships between spleen size and infection for urban and rural coyotes at three different ages (the mean age and ±1 standard deviation). Infection status (probability of infection) and intensity (natural log-transformed worm counts) were both adjusted to appear on the same y-axis.

Diet and infection status

Despite the lack of clear univariate associations between E. multilocularis infection and diet, several dietary variables became significant predictors of infection status when accounting for their two- and three-way interactions with location and/or age. With respect to our initial hypotheses that either rodent or anthropogenic food consumption might drive infections, we found that rodent consumption was positively associated with infection in young (< 1 yr) urban coyotes, but negatively associated with infection in older (> 1 yr) urban coyotes and rural coyotes of all ages (Figs 4A and 5; Table 1). Digestible anthropogenic food consumption was positively associated with infection in younger coyotes from both urban and rural areas, but negatively associated with infection in older individuals (Figs 4A and 5; Table 1). Carbon stable isotope values (δ13C) followed the same trend as rodents, whereby long-term anthropogenic food consumption was positively associated with infection in younger, urban coyotes but negatively associated with infection in older coyotes and rural coyotes (Figs 4A and 5; Table 1).

Fig 4. Dietary drivers of E. multilocularis infection status.

(a) Three-way interactions between age, location, and each focal diet item (stomach contents [solid lines, measured in ml] and stable isotope values [dashed lines]). As in Fig 3, relationships are plotted separately for urban and rural coyotes at three different ages. All diet components were rescaled so that the x-axis extends from the minimum to maximum value for each item (digestible anthropogenic food, 0–360 ml; indigestible anthropogenic food, 0–160 ml; rodents, 0–325 ml; δ13C, -24.3–-19.2, δ15N, 5.2–10.8). Refer to Table 1 for the results of likelihood ratio tests indicating the significance of these relationships. (b) Model-averaged coefficients for the top models predicting infection intensity. Each main effect is plotted followed by its two- and three-way interactions with other variables (indicated with an ‘x’). The coefficient for location reflects urban relative to rural coyotes. Coefficients were standardized by the partial standard deviation and weighted based on model weight prior to averaging. Thick and thin lines indicate 50% and 95% confidence intervals, respectively.

Fig 5. Coyote diet and stable isotope measures in relation to coyote infection status, age, and location (urban or rural).

Note the different scales on each panel. In the bar graphs, error bars indicate the standard error. In the boxplots, the box indicates the median and interquartile range, the line represents the 95th percentile, and the dots represent outliers. Coyotes were classified as juvenile or adult based on whether they were younger or older, respectively, than the median sample age (1.78 yr).

These predictors and their interactions were retained in the nine top-ranked models predicting infection status from a suite of dietary variables (Fig 4B; S3 Table in S3 Appendix). Although neither indigestible anthropogenic food consumption nor δ15N values had significantly predicted infection in two- or three-way interaction models, these predictors and some of their associated interactions were retained in the top model set (Fig 4B). In these models, indigestible anthropogenic food consumption was associated with infection at all ages and in both locations, and long-term protein consumption (δ15N) was associated with infection in rural coyotes and young, urban coyotes but not older, urban coyotes (Fig 4A). Based on model-averaged coefficient values, digestible and indigestible anthropogenic food consumption had a stronger relationship with infection status than rodent consumption or stable isotope values (Fig 4B). The accuracy of the top-ranked models ranged from 65.1–73.1% in cross-validation tests, with Cohen’s kappa of 0.35–0.46 and McFadden’s pseudo-R2 of 0.34–0.39, suggesting that the models had a fair to moderate level of accuracy when predicting infection status from dietary variables and their interactions with age and location (S3 Table in S3 Appendix).

Diet and infection intensity

Dietary drivers of infection intensity largely mirrored the dietary drivers of infection status: digestible anthropogenic food and rodent consumption both predicted infection intensity in significant three-way interactions with location and age. As with models for infection status, both measures were strong predictors of infection intensity in younger, urban coyotes but had reduced or opposite effects in older coyotes and rural coyotes (Fig 6A). Indigestible anthropogenic food consumption was also a significant predictor of infection intensity as part of a three-way interaction (Fig 6A; Table 1); it most strongly predicted infection intensity in urban coyotes, though this was largely because one of the three highly infected urban coyotes (>10,000 worms) had consumed a small amount (3.5 ml) of indigestible anthropogenic food and most other urban coyotes had not consumed any (S4 Table in S1 Appendix; S2 Fig in S3 Appendix). δ13C exhibited a significant interaction with age (Table 1), whereby higher values were negatively associated with worm counts at young ages but positively associated with worm counts at older ages. All five dietary predictors and most two- and three-way interactions were preserved in the thirteen top-ranked models predicting infection intensity (Fig 6B; S4 Table in S3 Appendix); however, these models had root mean-squared errors from 2.86–4.03 (measured on a log-transformed scale of worm counts) and R2 values less than 0.14 (S4 Table in S3 Appendix), suggesting that models predicting infection intensity were generally less accurate than models predicting infection status.

Fig 6. Dietary drivers of E. multilocularis infection intensity.

(a) Three-way interactions between age, location, and each focal diet item (stomach contents [solid lines, measured in ml] and stable isotope values [dashed lines]). As in Figs 3 and 4, relationships are plotted separately for urban and rural coyotes at three different ages. All diet components were rescaled so that the x-axis extends from the minimum to maximum value for each item (see Fig 4). Refer to Table 1 for the significance of these relationships. (b) Model-averaged coefficients for the top models predicting infection intensity. Each main effect is plotted followed by its two- and three-way interactions with other variables (indicated with an ‘x’). The coefficient for location reflects urban relative to rural coyotes. Coefficients were standardized by the partial standard deviation and weighted based on model weight prior to averaging. Thick and thin lines indicate 50% and 95% confidence intervals, respectively.


Mitigating the spread and public health risk of a trophically transmitted parasite like E. multilocularis requires understanding the dietary drivers of E. multilocularis infection. We hypothesized that coyote diets would relate to increased E. multilocularis infection via two potential mechanisms: greater exposure to the parasite, indicated by greater consumption of rodent intermediate hosts, or greater susceptibility to infection, indicated by greater consumption of poor-quality anthropogenic food. Overall, we found limited evidence that these mechanisms act consistently across our entire sample pool: neither short- nor long-term measures of diet could universally discriminate between infected and uninfected coyotes, and the best overall predictor of infection status and intensity was a younger age. Nevertheless, when accounting for the effects of age, short-term measures of rodent and anthropogenic food consumption were the best dietary predictors of infection, though not always in the ways we expected. We found support for our initial hypotheses in young, urban coyotes: digestible and indigestible anthropogenic food consumption, rodent consumption, and δ13C and δ15N values all predicted E. multilocularis infection in this population. However, the effects of these predictors were more variable in rural coyotes and older coyotes and for models predicting infection intensity, suggesting that the relationship between diet and infection may be a complex function of habitat use, age and/or body condition, and other variables that we could not measure.

Our estimate of the overall prevalence of active E. multilocularis infections in Edmonton’s urban coyote population (56.1%) generally matches the previously published estimates of infection prevalence based on subsamples of our study population (65.2% by Luong et al. [26], 53% by Sugden et al. [27]), though we observed a higher prevalence among rural coyotes than in our previous study (43% here vs. 35% by Sugden et al. [27]). Some of these discrepancies could be associated with the different methods used to test for infections: in this study, we used a sensitive molecular test (qPCR) on samples taken from the jejunum, where E. multilocularis is most expected to occur [55,56]. Nevertheless, the prevalence of E. multilocularis in both the urban and rural populations studied here is notably higher than in other Canadian coyote populations and much higher than other Canadian urban areas for which data is available (Calgary, 21–29%; Winnipeg, 7%; see (S6 Table in S1 Appendix for data and references). Moreover, our qPCR results indicated that over 80% of urban coyotes carried E. multilocularis DNA (compared to 64% of rural coyotes). Although the qPCR-positive intestines with no visible worms presumably reflect an infection that is not producing viable eggs, the high DNA detection rate underscores the frequency with which urban coyotes in Edmonton may encounter, acquire, and potentially transmit E. multilocularis infections.

Age-dependent immunity

With respect to contextual drivers of infection, we found that young age was the best overall predictor of E. multilocularis infections. This result agrees with previous reports of more E. multilocularis infections among young coyotes [33,67] and foxes [31,32], though not all studies have found an equivalent relationship between age and E. multilocularis infection [25,26]. Increased parasitism in juveniles is common across a range of host and parasite taxa because juveniles generally have naïve immune systems and expend more energy on growth than immunity [29,68]. However, by adulthood, many animals produce a measurable immune response to gastrointestinal parasites; for example, older sheep infected with gastrointestinal helminths produce more lymphocytes than infected younger sheep [69], and adult dogs infected with E. multilocularis have increased levels of serum immunoglobulin G against adult worms [70]. It is unlikely that canids become entirely immune to E. multilocularis infection [71], but a study of domestic dogs experimentally reinfected with E. multilocularis five times over one year showed a tremendous reduction in infection intensity, with ~90% fewer adult worms than never-exposed individuals [72,73].

Our results provide several lines of indirect evidence for increasing immunity with age. First, not only were young coyotes more likely to be infected, but coyotes that were qPCR-positive for E. multilocularis DNA without visible worms were, on average, 13.5 months older than coyotes that tested positive via both approaches, but 4.5 months younger than coyotes with no evidence of infection (S2 Fig in S2 Appendix). Second, infection intensities peaked at young ages: 73% of the coyotes with >1,000 scolexes in their intestines were less than 1 year old, and the three individuals with >10,000 scolexes were all less than 6 months old. Finally, the relationship that we observed between immune system investment (measured as spleen mass) and infection status at different ages may reflect age-dependent resistance or immunity to E. multilocularis. In young coyotes, larger spleens were associated with a lower probability of infection; for these individuals, a larger spleen may indicate greater investment in immune development [52] and therefore a greater ability to resist infection. Conversely, larger spleens were associated with a higher probability of infection in older animals, which could indicate the effect of a current infection: spleen size swells as the animal mounts an immune response, an effect that has been observed in several other species including geese [74] and various rodents [75].

The evidence of increased immunity with age in our coyote sample was not evident in urban coyotes, suggesting that urban coyotes were less able to develop or mount an effective age-dependent immune response. Whereas the likelihood and intensity of infection declined significantly with age in rural coyotes, urban coyotes were equally likely to be infected at all ages. We suspect that this absent signal of increasing immunity with age stems from the overall greater immune stress experienced by urban animals [5]. For example, genes for inflammatory pathways, antioxidant production, and toxicant neutralization have all been shown to be upregulated in urban wildlife compared to rural conspecifics [76,77], and these stresses may make an animal less adept at responding to an infection or parasite [78,79]. In addition, urban coyotes consumed more anthropogenic food than rural coyotes, which can further reduce immune function by leading to malnutrition or directly suppressing immune activity [80,81].

Dietary drivers of infection

With respect to our diet-based hypotheses, we initially predicted that the higher E. multilocularis prevalence in urban coyotes could be a result of either increased exposure to the parasite via consumption of infected hosts or increased susceptibility to the parasite due to the negative health effects of consuming anthropogenic food. After accounting for the interactions with age driven by potential age-dependent immunity, we found support for both these hypotheses in young (< 1 yr) urban coyotes: both short- and long-term measures of rodent and anthropogenic food consumption predicted infection status in this population, and short-term measures of these items also predicted infection intensity. We suspect that urban coyotes that forage at compost sites (as described by Murray et al. [49]), which provide ideal habitat for voles and other rodents, consume high quantities of both rodents and anthropogenic food, thus increasing both their exposure and potential susceptibility to infection. Contrary to our expectations, our two hypothesized mechanisms could not be clearly disentangled and may act synergistically. Actually, similar concerns about the overlap between rodent habitat and anthropogenic food sources in relation to E. multilocularis have been voiced in Europe owing to the fact that both rodents and anthropogenic food attract foxes [82].

Intriguingly, neither rodent nor digestible anthropogenic food consumption continued to predict infection status or intensity in urban coyotes as they aged, even though the likelihood of infection did not decline with age. This finding opposes an age-independent positive correlation between rodent consumption and infection intensity in foxes [42]. Given that the maximum lifespan of adult E. multilocularis worms is approximately 7 months [83], older coyotes would still need to be consuming infected rodents to maintain active, detectable infections. Liccioli et al. [84] estimated that over 57% of urban coyotes in Calgary become reinfected following an initial infection, and Liccioli et al. [44] estimated that coyotes consume, on average, 1.05 infected intermediate hosts during the winter (our sample collection period). Our results suggest that older urban coyotes still consume enough infected intermediate hosts to maintain ongoing infections, even if their overall diet shifts to larger prey with age and increased rodent consumption no longer directly predicts infection. However, longitudinal studies using non-invasive or non-destructive approaches to track infection and diet over time would be needed to resolve this complex interplay between rodent consumption and age-related E. multilocularis exposure and infection.

We found more mixed evidence for our diet-based hypotheses in rural coyotes: short-term anthropogenic food consumption from stomach contents predicted infection only at the youngest (<0.5 yr) ages, and short-term rodent consumption did not predict infection at any age. Instead, indigestible anthropogenic food consumption predicted infection at all ages, though these relationships were not all statistically significant. The rural coyotes with the greatest amounts of indigestible anthropogenic food in their stomachs were not only more likely to be infected at higher worm burdens, but they were also younger coyotes captured closest to the Leduc regional landfill (S3 Fig in S3 Appendix). In this case, the relationship between digestible and indigestible anthropogenic food consumption and E. multilocularis infection in young coyotes may parallel the relationship observed in urban coyotes: young individuals that congregate and/or forage in areas where rodent habitat and anthropogenic food sources overlap (e.g., compost piles or garbage dumps) are more likely to be infected. It is not entirely clear why infected, young, rural coyotes would not also have consumed more rodents than uninfected coyotes, but it is possible that the overall reliance of juvenile coyotes on small prey items obscured this effect. In addition, plastic consumption has been associated with poorer nutrition, intestinal inflammation, and exposure to immune-compromising toxins [85], all of which may drive increased susceptibility to E. multilocularis infection without increased rodent consumption.


We acknowledge that our study is limited in its ability to determine how specific dietary components contribute to parasite presence or intensity in a population of wild carnivores. First, our method of sample collection may bias our data towards a younger overall age distribution because younger animals are more prone to being trapped, hit by cars, and coming into conflict with humans. The abundance of young coyotes in our study may therefore have inflated our population-wide measures of parasite prevalence and intensity [86,87], though we note that a previous sample of road-killed coyotes in Edmonton was not biased toward younger animals [88]. Second, we note that the likelihood of infection depends not only on the number of intermediate hosts consumed but also the prevalence of E. multilocularis among intermediate hosts. Given the paucity of available data on this topic—the most recent estimate of E. multilocularis prevalence in rodents comes from 1970, when Holmes et al. [89] observed a prevalence of 25% among rodents near Edmonton—our study design assumes a consistent prevalence across all habitats we sampled. Nevertheless, given the generally high prevalence of E. multilocularis among urban carnivores, we suggest that future studies should focus in parallel on the prevalence of E. multilocularis among intermediate hosts.

Finally, because E. multilocularis requires up to 60 days to mature in coyote intestines [44], our measures of diet do not directly align with timescale of infection. Stomach contents only reflect the last meal before death and could overrepresent food items that are less digestible [90]; conversely, stable isotopes are too coarse to reflect specific diet components, such as the distinction between rodents and other prey. This may partly explain why our models for infection intensity had limited predictive power. However, previous studies in Edmonton and elsewhere suggest that individual coyotes and other canids specialize over time and have reasonably consistent diets [35,60], in which case stomach contents may serve as an appropriate proxy for diet at the time of infection. Moreover, our two methods of diet analysis are some of the few methods available for assessing diet in wild carnivores, making them representative of the approaches that would also be taken by wildlife managers.

Management implications

Although previous studies have considered coyote culling [91], rodent poisons [92], and anthelminthic coyote baits [93] as strategies for controlling parasite spread, these approaches are often laborious, expensive, and have achieved only mixed success at controlling E. multilocularis infections in foxes in Europe [9499]. Moreover, culling programs can reduce the age distribution of coyote populations and increase litter sizes [100], which might worsen the prevalence of this parasite. Our results instead suggest that effective management of urban coyotes and E. multilocularis should focus on identifying the resources and habitats preferred by young coyotes, because this would correspond to where eggs are more likely to be shed in the environment. In addition, managerial approaches designed to prevent coyotes from accessing communal anthropogenic food sources (e.g., compost piles or garbage facilities) could reduce the spatial overlap between coyotes and rodents, which is the most likely reason for why both anthropogenic food and rodent consumption were associated with infection in young urban coyotes. Because the extended asymptomatic period of alveolar echinococcosis creates challenges in confirming sources of human exposure to E. multilocularis [24,101], mitigation strategies should also include public education that alerts citizens to the presence of E. multilocularis, the need to wash items that potentially come in contact with viable eggs (e.g., hands, toys, tools, garden produce), and the value of deworming dogs, especially those that consume rodents [102]. Overall, evidence that the more virulent European strain of E. multilocularis is now widespread in Canada and has infected over 20 people in Alberta [24] speaks to the need for more research on how best to limit infections in wildlife and minimize the risk of transmission to humans.

Supporting information

S1 Appendix. Supplementary methods and sample overview.

Additional information on the statistical methods used for data analysis and the coyote sample used in this study. This appendix includes coyote demographic data (age, sex, year of collection, etc.), diet information (percent occurrence of difference diet items), and the full results of statistical tests related to these variables.


S2 Appendix. Analyses reproduced for qPCR-positive coyotes.

For the analyses presented in the main manuscript text, we focused on coyotes that both (i) tested positive for E. multilocularis via qPCR and (ii) had visible evidence of Echinococcus scolexes in their intestines. To demonstrate that this decision did not substantively alter our conclusions, this appendix reproduces our analyses, except in these cases we considered all qPCR-positive coyotes as positive for E. multilocularis, regardless of whether scolexes were observed in the intestine.


S3 Appendix. Predictors of E. multilocularis infection.

Full numerical results for (i) statistical tests for differences in infection status or intensity as a function of contextual variables or diet and (ii) generalized linear models predicting infection status or intensity as a function of these same variables. These data support Figs 3, 4 and 6 in the main manuscript text.




The authors would like to thank Malcolm Abercrombie for providing coyote carcasses collected as part of the fur-trapping and population management program with Animal Damage Control, Inc., as well as multiple employees at the City of Edmonton Animal Care and Control Centre who collected and saved road-killed carcasses for us. We also thank Arya Horon for assistance counting E. multilocularis scoleces, Alvin Kwan for assistance with stable isotope analysis, and Arlene Oatway for assistance with cementum aging. Darcy Visscher provided valuable critical feedback on the data analysis and manuscript.


  1. 1. Fischer JD, Cleeton SH, Lyons TP, Miller JR. Urbanization and the predation paradox: the role of trophic dynamics in structuring vertebrate communities. BioScience. 2012;62(9): 809–18.
  2. 2. Sorace A, Gustin M. Distribution of generalist and specialist predators along urban gradients. Landsc Urban Plan. 2009;90: 111–8.
  3. 3. Moyers SC, Adelman JS, Farine DR, Thomason CA, Hawley DM. Feeder density enhances house finch disease transmission in experimental epidemics. Philos Trans R Soc Lond B Biol Sci. 2018;373: 20170090. pmid:29531145
  4. 4. Werner CS, Nunn CL. Effect of urban habitat use on parasitism in mammals: a meta-analysis. Proc Biol Sci. 2020;287: 20200397. pmid:32396800
  5. 5. Murray MH, Sánchez CA, Becker DJ, Byers KA, Worsley‐Tonks KEL, Craft ME. City sicker? A meta‐analysis of wildlife health and urbanization. Frontiers in Ecology and the Environment. 2019;17(10): 575–83.
  6. 6. Becker DJ, Streicker DG, Altizer S. Using host species traits to understand the consequences of resource provisioning for host-parasite interactions. J Anim Ecol. 2018;87(2): 511–25. pmid:29023699
  7. 7. Brearley G, Rhodes J, Bradley A, Baxter G, Seabrook L, Lunney D, et al. Wildlife disease prevalence in human-modified landscapes. Biol Rev Camb Philos Soc. 2013;88(2): 427–42. pmid:23279314
  8. 8. Schatz AM, Park AW. Host and parasite traits predict cross-species parasite acquisition by introduced mammals. Proc Biol Sci. 2021;288(1950): 20210341. pmid:33947240
  9. 9. Reddell CD, Abadi F, Delaney DK, Cain JW 3rd, Roemer GW. Urbanization’s influence on the distribution of mange in a carnivore revealed with multistate occupancy models. Oecologia. 2021;195(1): 105–16. pmid:33538881
  10. 10. Scott DM, Baker R, Tomlinson A, Berg MJ, Charman N, Tolhurst BA. Spatial distribution of sarcoptic mange (Sarcoptes scabiei) in urban foxes (Vulpes vulpes) in Great Britain as determined by citizen science. Urban Ecosystems. 2020;23(5): 1127–40.
  11. 11. Barros M, Cabezon O, Dubey JP, Almeria S, Ribas MP, Escobar LE, et al. Toxoplasma gondii infection in wild mustelids and cats across an urban-rural gradient. PLoS One. 2018;13(6): e0199085. pmid:29924844
  12. 12. Jardine CM, Pearl DL, Puskas K, Campbell DG, Shirose L, Peregrine AS. The impact of land use, season, age, and sex on the prevalence and intensity of Baylisascaris procyonis infections in raccoons (Procyon lotor) from Ontario, Canada. J Wildl Dis. 2014;50(4): 784–91. pmid:25098302
  13. 13. Strand TM, Lundkvist A. Rat-borne diseases at the horizon. A systematic review on infectious agents carried by rats in Europe 1995–2016. Infect Ecol Epidemiol. 2019;9(1): 1553461. pmid:30834071
  14. 14. Mackenstedt U, Jenkins D, Romig T. The role of wildlife in the transmission of parasitic zoonoses in peri-urban and urban areas. Int J Parasitol PAR. 2015;4(1): 71–9. pmid:25830108
  15. 15. Massolo A, Liccioli S, Budke C, Klein C. Echinococcus multilocularis in North America: the great unknown. Parasite. 2014;21: 73. pmid:25531581
  16. 16. Deplazes P, Eckert J. Veterinary aspects of alveolar echinococcosis—a zoonosis of public health significance. Vet Parasitol. 2001;98: 65–87. pmid:11516580
  17. 17. Kern P, Ammon A, Kron M, Sinn G, Sander S, Peterson LR, et al. Risk factors for alveolar echinococcosis in humans. Emerging Infectious Diseases. 2004;10(12): 2088–93. pmid:15663843
  18. 18. Bastien M, Vaniscotte A, Combes B, Umhang G, Germain E, Gouley V, et al. High density of fox and cat faeces in kitchen gardens and resulting rodent exposure to Echinococcus multilocularis and Toxoplasma gondii. Folia Parasitol (Praha). 2018;65. pmid:29528299
  19. 19. Craig PS, Rogan MT, Campos-Ponce M. Echinococcosis: disease, detection and transmission. Parasitology. 2004;127: S5–S20.
  20. 20. Snabel V, Antolova D, Cavallero S, D’Amelio S. On the geographic genetic variants of the cestode Echinococcus multilocularis with reference to the original descriptions from Bowles et al. (1992) and Bowles and McManus (1993), and their use. Parasitol Int. 2020;75: 102039. pmid:31843686
  21. 21. Jenkins EJ, Peregrine AS, Hill JE, Somers C, Gesy K, Barnes B, et al. Detection of European strain of Echinococcus multilocularis in North America. Emerg Infect Dis. 2012;18(6): 1010–2. pmid:22608114
  22. 22. Gesy K, Hill JE, Schwantje H, Liccioli S, Jenkins EJ. Establishment of a European-type strain of Echinococcus multilocularis in Canadian wildlife. Parasitology. 2013;140(9): 1133–7. pmid:23714582
  23. 23. Gesy KM, Schurer JM, Massolo A, Liccioli S, Elkin BT, Alisauskas R, et al. Unexpected diversity of the cestode Echinococcus multilocularis in wildlife in Canada. Int J Parasitol Parasites Wildl. 2014;3(2): 81–7. pmid:25161905
  24. 24. Houston S, Belga S, Buttenschoen K, Cooper R, Girgis S, Gottstein B, et al. Epidemiological and clinical characteristics of alveolar echinococcosis: an emerging infectious disease in Alberta, Canada. Am J Trop Med Hyg. 2021;00(0): 1–7. pmid:33755579
  25. 25. Catalano S, Lejeune M, Liccioli S, Verocai GG, Gesy KM, Jenkins EJ, et al. Echinococcus multilocularis in urban coyotes, Alberta, Canada. Emerg Infect Dis. 2012;18(10): 1625–8.
  26. 26. Luong LT, Chambers JL, Moizis A, Stock TM, St Clair CC. Helminth parasites and zoonotic risk associated with urban coyotes (Canis latrans) in Alberta, Canada. J Helminthol. 2018: 1–5. pmid:30588908
  27. 27. Sugden S, Sanderson D, Ford K, Stein LY, St Clair CC. An altered microbiome in urban coyotes mediates relationships between anthropogenic diet and poor health. Sci Rep. 2020;10(1): 22207. pmid:33335116
  28. 28. Worsley-Tonks KEL, Gehrt SD, Anchor C, Escobar LE, Craft ME. Infection risk varies within urbanized landscapes: the case of coyotes and heartworm. Parasit Vectors. 2021;14(1): 464. pmid:34503566
  29. 29. Ashby B, Bruns E. The evolution of juvenile susceptibility to infectious disease. Proc Biol Sci. 2018;285(1881). pmid:29925619
  30. 30. Combrink L, Glidden CK, Beechler BR, Charleston B, Koehler AV, Sisson D, et al. Age of first infection across a range of parasite taxa in a wild mammalian population. Biol Lett. 2020;16(2): 20190811. pmid:32070234
  31. 31. Yimam AE, Nonaka N, Oku Y, Kamiya M. Prevalence and intensity of Echinococcus multilocularis in red foxes (Vulpes vulpes schrencki) and raccoon dogs (Nyctereutes procyonoides albus) in Otaru city, Hokkaido, Japan. Jpn J Vet Res. 2002;49(4): 287–96.
  32. 32. Fischer C, Reperant LA, Weber JM, Hegglin D, Deplazes P. Echinococcus multilocularis infections of rural, residential and urban foxes (Vulpes vulpes) in the canton of Geneva, Switzerland. Parasite. 2005;12(4): 339–46. pmid:16402566
  33. 33. Liccioli S, Catalano S, Kutz SJ, Lejeune M, Verocai GG, Duignan PJ, et al. Gastrointestinal parasites of coyotes (Canis latrans) in the metropolitan area of Calgary, Alberta, Canada. Can J Zool. 2012;90(8): 1023–30.
  34. 34. Losson B, Kervyn T, Detry J, Pastoret PP, Mignon B, Brochier B. Prevalence of Echinococcus multilocularis in the red fox (Vulpes vulpes) in southern Belgium. Vet Parasitol. 2003;117(1–2): 23–8. pmid:14597275
  35. 35. Sugden S, Murray M, Edwards MA, St. Clair CC. Inter-population differences in coyote diet and niche width along an urban–suburban–rural gradient. Journal of Urban Ecology. 2021;7(1).
  36. 36. Murray M, Cembrowski A, Latham ADM, Lukasik VM, Pruss S, St Clair CC. Greater consumption of protein-poor anthropogenic food by urban relative to rural coyotes increases diet breadth and potential for human-wildlife conflict. Ecography. 2015;38(12): 1235–42.
  37. 37. Guerra D, Hegglin D, Bacciarini L, Schnyder M, Deplazes P. Stability of the southern European border of Echinococcus multilocularis in the Alps: evidence that Microtus arvalis is a limiting factor. Parasitology. 2014;141: 1593–602. pmid:24932666
  38. 38. Baudrot V, Perasso A, Fritsch C, Raoul F. Competence of hosts and complex foraging behavior are two cornerstones in the dynamics of trophically transmitted parasites. J Theor Biol. 2016;397: 158–68. pmid:26992573
  39. 39. Stieger C, Hegglin D, Schwarzenbach G, Mathis A, Deplazes P. Spatial and temporal aspects of urban transmission of Echinococcus multilocularis. Parasitology. 2002;124: 631–40. pmid:12118719
  40. 40. Raoul F, Hegglin D, Giraudoux P. Trophic ecology, behaviour and host population dynamics in Echinococcus multilocularis transmission. Vet Parasitol. 2015;213: 162–71. pmid:26276578
  41. 41. Hegglin D, Bontadina F, Contesse P, Gloor S, Deplazes P. Plasticity of predation behaviour as a putative driving force for parasite life-cycle dynamics: the case of urban foxes and Echinococcus multilocularis tapeworm. Funct Ecol. 2007;21(3): 552–60.
  42. 42. Robardet E, Giraudoux P, Caillot C, Boue F, Cliquet F, Augot D, et al. Infection of foxes by Echinococcocus multilocularis in urban and suburban areas of Nancy, France: influence of feeding habits and environment. Parasite. 2008;15(1): 77–85. pmid:18416250
  43. 43. City of Edmonton. Biodiversity Report. Local Action for Biodiversity Project. 2008: 1–100.
  44. 44. Liccioli S, Bialowas C, Ruckstuhl KE, Massolo A. Feeding ecology informs parasite epidemiology: prey selection modulates encounter rate with Echinococcus multilocularis in urban coyotes. PLoS One. 2015;10(3): e0121646. pmid:25768437
  45. 45. Hofer S, Gloor S, Muller U, Mathis A, Hegglin D, Deplazes P. High prevalence of Echinococcus multilocularis in urban red foxes (Vulpes vulpes) and voles (Arvicola terrestris) in the city of Zurich, Switzerland. Parasitology. 2000;120 135–42. pmid:10726275
  46. 46. Murray M, Edwards MA, Abercrombie B, St Clair CC. Poor health is associated with use of anthropogenic resources in an urban carnivore. Proc Biol Sci. 2015;282: 20150009. pmid:25876843
  47. 47. Cummings CR, Hernandez SM, Murray M, Ellison T, Adams HC, Cooper RE, et al. Effects of an anthropogenic diet on indicators of physiological challenge and immunity of white ibis nestlings raised in captivity. Ecol Evol. 2020;10(15): 8416–28. pmid:32788990
  48. 48. Ezenwa VO. Interactions among host diet, nutritional status and gastrointestinal parasite infection in wild bovids. Int J Parasitol. 2004;34(4): 535–42. pmid:15013743
  49. 49. Murray MH, Hill J, Whyte P, St Clair CC. Urban compost attracts coyotes, contains toxins, and may promote disease in urban-adapted wildlife. Ecohealth. 2016;13(2): 285–92. pmid:27106524
  50. 50. Veit P, Bilger B, Schad V, Schafer J, Frank W, Lucius R. Influence of environmental factors on the infectivity of Echinococcus multilocularis eggs. Parasitology. 1995;110: 79–86.
  51. 51. Huot J, Poulle ML, Crete M. Evaluation of several indices for assessment of coyote (Canis latrans) body composition. Can J Zool. 1995;73: 1620–4.
  52. 52. Corbin E, Vicente J, Martin-Hernando MP, Acevedo P, Perez-Rodriguez L, Gortazar C. Spleen mass as a measure of immune strength in mammals. Mammal Rev. 2008;38(1): 108–15.
  53. 53. Stewart REA, Stewart BE, Stirling I, Street E. Counts of growth layer groups in cementum and dentine in ringed seals (Phoca hispida). Mar Mam Sci. 1996;12(3): 383–401.
  54. 54. Gesy K, Pawlik M, Kapronczai L, Wagner B, Elkin B, Schwantje H, et al. An improved method for the extraction and quantification of adult Echinococcus from wildlife definitive hosts. Parasitol Res. 2013;112(5): 2075–8. pmid:23471781
  55. 55. Karamon J, Sroka J, Dabrowska J, Bilska-Zajac E, Skrzypek K, Rozycki M, et al. Distribution of parasitic helminths in the small intestine of the red fox (Vulpes vulpes). Pathogens. 2020;9(6): 1–16. pmid:32560133
  56. 56. Thompson RCA. Biology and systematics of Echinococcus In: Thompson RCA, Deplazes P, Lymbery AJ, editors. Advances in Parasitology. 95. London, UK: Elsevier Ltd.; 2017.
  57. 57. Santa MA, Pastran SA, Klein C, Duignan P, Ruckstuhl K, Romig T, et al. Detecting co-infections of Echinococcus multilocularis and Echinococcus canadensis in coyotes and red foxes in Alberta, Canada using real-time PCR. Int J Parasitol Parasites Wildl. 2018;7(2): 111–5. pmid:29988802
  58. 58. Wolfert DR, Miller TJ. Age, growth, and food of northern pike in eastern Lake Ontario. Trans Am Fish Soc. 1978;107(5): 696–702. <696:Agafon>2.0.Co;2.
  59. 59. McLaren AAD, Crawshaw GJ, Patterson BR. Carbon and nitrogen discrimination factors of wolves and accuracy of diet inferences using stable isotope analysis. Wildlife Society Bulletin. 2015;39(4): 788–96.
  60. 60. Newsome SD, Garbe HM, Wilson EC, Gehrt SD. Individual variation in anthropogenic resource use in an urban carnivore. Oecologia. 2015;178(1): 115–28. pmid:25669449
  61. 61. Newsome SD, Ralls K, Van Horn Job C, Fogel ML, Cypher BL. Stable isotopes evaluate exploitation of anthropogenic foods by the endangered San Joaquin kit fox (Vulpes macrotis mutica). J Mammal. 2010;91(6): 1313–21.
  62. 62. R Development Core Team. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2020.
  63. 63. Poessel SA, Mock EC, Breck SW. Coyote (Canis latrans) diet in an urban environment: variation relative to pet conflicts, housing density, and season. Can J Zool. 2017;95(4): 287–97.
  64. 64. Barton K. MuMIn: Multi-model inference. 2019.
  65. 65. Cade BS. Model averaging and muddled multimodel inferences. Ecology. 2015;96(9): 2370–82. pmid:26594695
  66. 66. Oksanen J, Blanchet FG, Friendly M, Kindt R, Legendre P, McGlinn D, et al. vegan: Community ecology package. 2.5–6 ed2019.
  67. 67. Kotwa JD, Isaksson M, Jardine CM, Campbell GD, Berke O, Pearl DL, et al. Echinococcus multilocularis infection, southern Ontario, Canada. Emerg Infect Dis. 2019;25(2): 265–72. pmid:30666935
  68. 68. Acerini CI, Morris S, Morris A, Kenyon F, McBean D, Pemberton JM, et al. Helminth parasites are associated with reduced survival probability in young red deer. Parasitology. 2022;149(13): 1702–8. pmid:36052566
  69. 69. Colditz IG, Watson DL, Gray GD, Eady SJ. Some relationships between age, immune responsiveness and resistance to parasites in ruminants. International Journal for Parastiology. 1996;26(8/9): 869–77. pmid:8923136
  70. 70. Foster N, Elsheikha HM. The immune response to parasitic helminths of veterinary importance and its potential manipulation for future vaccine control strategies. Parasitol Res. 2012;110(5): 1587–99. pmid:22314781
  71. 71. Torgerson PR. Canid immunity to Echinococcus spp.: impact on transmission. Parasite Immunol. 2006;28(7): 295–303. pmid:16842266
  72. 72. Kouguchi H, Irie T, Matsumoto J, Nakao R, Sugano Y, Oku Y, et al. The timing of worm exclusion in dogs repeatedly infected with the cestode Echinococcus multilocularis. J Helminthol. 2016;90(6): 766–72. pmid:26750177
  73. 73. Kouguchi H, Furuoka H, Irie T, Matsumoto J, Nakao R, Nonaka N, et al. Adult worm exclusion and histological data of dogs repeatedly infected with the cestode Echinococcus multilocularis. Data Brief. 2020;29: 105353. pmid:32181312
  74. 74. Shutler D, Alisauskas RT, McLaughlin JD. Mass dynamics of the spleen and other organs in geese: measures of immune relationships to helminths? Canadian Journal of Zoology. 1998;77: 351–9.
  75. 75. Ponlet N, Chaisiri K, Claude J, Morand S. Incorporating parasite systematics in comparative analyses of variation in spleen mass and testes sizes of rodents. Parasitology. 2011;138(13): 1804–14. pmid:21554839
  76. 76. Capilla-Lasheras P, Dominoni DM, Babayan SA, O’Shaughnessy PJ, Mladenova M, Woodford L, et al. Elevated Immune Gene Expression Is Associated with Poor Reproductive Success of Urban Blue Tits. Frontiers in Ecology and Evolution. 2017;5.
  77. 77. Salmon P, Watson H, Nord A, Isaksson C. Effects of the Urban Environment on Oxidative Stress in Early Life: Insights from a Cross-fostering Experiment. Integr Comp Biol. 2018;58(5): 986–94. pmid:30052952
  78. 78. Dupjan S, Dawkins MS. Animal Welfare and Resistance to Disease: Interaction of Affective States and the Immune System. Front Vet Sci. 2022;9: 929805. pmid:35774975
  79. 79. Bailly J, Scheifler R, Belvalette M, Garnier S, Boissier E, Clement-Demange VA, et al. Negative impact of urban habitat on immunity in the great tit Parus major. Oecologia. 2016;182(4): 1053–62. pmid:27646717
  80. 80. Hwang J, Kim Y, Lee SW, Kim NY, Chun MS, Lee H, et al. Anthropogenic food provisioning and immune phenotype: Association among supplemental food, body condition, and immunological parameters in urban environments. Ecol Evol. 2018;8(5): 3037–46. pmid:29531715
  81. 81. Strandin T, Babayan SA, Forbes KM. Reviewing the effects of food provisioning on wildlife immunity. Philos Trans R Soc Lond B Biol Sci. 2018;373(1745). pmid:29531143
  82. 82. Duscher G, Steineck T, Günter P, Prosl H, Joachim A. Echinococcus multilocularis in foxes in Vienna and surrounding territories. Wiener Tierarztliche Monatsschrift. 2005;92(1): 16–20.
  83. 83. Romig T, Deplazes P, Jenkins D, Giraudoux P, Massolo A, Craig PS, et al. Ecology and life cycle patterns of Echinococcus species. Adv Parasitol. 2017;95: 213–314. pmid:28131364
  84. 84. Liccioli S, Rogers S, Greco C, Kutz SJ, Chan F, Ruckstuhl KE, et al. Assessing individual patterns ofEchinococcus multilocularisinfection in urban coyotes: non-invasive genetic sampling as an epidemiological tool. Journal of Applied Ecology. 2015;52(2): 434–42.
  85. 85. Puskic PS, Lavers JL, Bond AL. A critical review of harm associated with plastic ingestion on vertebrates. Sci Total Environ. 2020;743: 140666. pmid:32758824
  86. 86. Van Deelen TR, Gosselink TE. Coyote survival in a row-crop agricultural landscape. Can J Zool. 2006;84(11): 1630–6.
  87. 87. Kreling SES, Gaynor KM, Coon CAC. Roadkill distribution at the wildland‐urban interface. J Wildl Manage. 2019;83(6): 1427–36.
  88. 88. Murray MH, St. Clair CC. Individual flexibility in nocturnal activity reduces risk of road mortality for an urban carnivore. Behav Ecol. 2015;26(6): 1520–7.
  89. 89. Holmes JC, Mahrt JL, Samuel WM. The occurrence of Echinococcus multilocularis Leuckart, 1863 in Alberta. Canadian Journal of Zoology. 1970;49: 575–6.
  90. 90. Amundsen PA, Sanchez-Hernandez J. Feeding studies take guts—critical review and recommendations of methods for stomach contents analysis in fish. J Fish Biol. 2019;95(6): 1364–73. pmid:31589769
  91. 91. Hegglin D, Bontadina F, Deplazes P. Human-wildlife interactions and zoonotic transmission of Echinococcus multilocularis. Trends Parasitol. 2015;31(5): 167–73. pmid:25599832
  92. 92. Serieys LEK, Lea AJ, Epeldegui M, Armenta TC, Moriarty J, VandeWoude S, et al. Urbanization and anticoagulant poisons promote immune dysfunction in bobcats. Proc Biol Sci. 2018;285: 20172533. pmid:29343604
  93. 93. Tackmann K, Loschner U, Mix H, Staubach C, Thulke HH, Ziller M, et al. A field study to control Echinococcus multilocularis-infections of the red fox (Vulpes vulpes) in an endemic focus. Epidemiol Infect. 2001;127(3): 577–87. pmid:11811893
  94. 94. Mosnier A, Boisjoly D, Courtois R, Ouellet JP. Extensive predator space use can limit the efficacy of a control program. J Wildl Manage. 2008;72(2): 483–91.
  95. 95. Mitchell BR, Jaeger MM, Barrett RH. Coyote depredation management: current methods and research needs. Wildl Soc Bull. 2004;32(4): 1209–18.
  96. 96. Morin DJ, Kelly MJ. The dynamic nature of territoriality, transience and biding in an exploited coyote population. Wildl Biol. 2017: wlb.00335.
  97. 97. Comte S, Raton V, Raoul F, Hegglin D, Giraudoux P, Deplazes P, et al. Fox baiting against Echinococcus multilocularis: contrasted achievements among two medium size cities. Prev Vet Med. 2013;111: 147–55. pmid:23642656
  98. 98. Comte S, Umhang G, Raton V, Raoul F, Giraudoux P, Combes B, et al. Echinococcus multilocularis management by fox culling: an inappropriate paradigm. Prev Vet Med. 2017;147: 178–85. pmid:29254718
  99. 99. König A, Romig T, Janko C, Hildenbrand R, Holzhofer E, Kotulski Y, et al. Integrated-baiting concept against Echinococcus multilocularis in foxes is successful in southern Bavaria, Germany. Eur J Wildl Res. 2008;54(3): 439–47.
  100. 100. Gese E, editor Demographic and spatial responses of coyotes to changes in food and exploitation. Wildlife Damage Management Conference; 2005.
  101. 101. Massolo A, Klein C, Kowalewska-Grochowska K, Belga S, MacDonald C, Vaughan S, et al. European Echinococcus multilocularis identified in patients in Canada. N Engl J Med. 2019;381(4): 384–5. pmid:31340100
  102. 102. Deplazes P, van Knapen F, Schweiger A, Overgaauw PA. Role of pet dogs and cats in the transmission of helminthic zoonoses in Europe, with a focus on echinococcosis and toxocarosis. Vet Parasitol. 2011;182: 41–53. pmid:21813243