scispace - formally typeset
Search or ask a question
Author

Olivia Linson

Bio: Olivia Linson is an academic researcher from University of Michigan. The author has contributed to research in topics: Predation & Species richness. The author has co-authored 1 publications.

Papers
More filters
Journal ArticleDOI
TL;DR: In this paper , a meta-analysis was conducted to quantify the effects of urban environments on three components of trophic ecology in predators: dietary species richness, dietary evenness and stable isotopic ratios (IRs).
Abstract: Predation is a fundamental ecological process that shapes communities and drives evolutionary dynamics. As the world rapidly urbanizes, it is critical to understand how human perturbations alter predation and meat consumption across taxa. We conducted a meta-analysis to quantify the effects of urban environments on three components of trophic ecology in predators: dietary species richness, dietary evenness and stable isotopic ratios (IRs) (δ13C and δ15N IR). We evaluated whether the intensity of anthropogenic pressure, using the human footprint index (HFI), explained variation in effect sizes of dietary attributes using a meta-regression. We calculated Hedges’ g effect sizes from 44 studies including 11 986 samples across 40 predatory species in 39 cities globally. The direction and magnitude of effect sizes varied among predator taxa with reptilian diets exhibiting the most sensitivity to urbanization. Effect sizes revealed that predators in cities had comparable diet richness, evenness and nitrogen ratios, though carbon IRs were more enriched in cities. We found that neither the 1993 nor 2009 HFI editions explained effect size variation. Our study provides, to our knowledge, the first assessment of how urbanization has perturbed predator–prey interactions for multiple taxa at a global scale. We conclude that the functional role of predators is conserved in cities and urbanization does not inherently relax predation, despite diets broadening to include anthropogenic food sources such as sugar, wheat and corn.

3 citations

Posted ContentDOI
21 Dec 2020-bioRxiv
TL;DR: In this paper, the authors conducted a meta-analysis to quantify the effects of urban environments on three components of trophic ecology in predators: dietary species richness (DSR), dietary evenness (DEV), and stable isotopic ratios (d13C and d15N IR).
Abstract: Predation is a fundamental ecological process that shapes communities and drives long-term evolutionary dynamics. As the world rapidly urbanizes, it is critical to understand how the built environment and other human perturbations alter predation across taxa. We conducted a meta-analysis to quantify the effects of urban environments on three components of trophic ecology in predators: dietary species richness (DSR), dietary evenness (DEV), and stable isotopic ratios (d13C and d15N IR). We then evaluated whether intensity of anthropogenic pressure, using the human footprint index (HFI), explained variation in the effect sizes of dietary attributes using a meta-regression. We calculated Hedges9 g effect sizes from 44 studies including 11,986 samples across 40 predatory species in 39 cities globally. The direction and magnitude of effect sizes varied between predator taxonomic groups with reptile diets exhibiting the most sensitivity to urbanization. Effect sizes revealed that predators in cities had comparable DSR, DEV, and nitrogen ratios, though carbon consumption was significantly higher. We found that HFI did not explain variation in effect sizes, a result consistent between the 1993 and 2009 editions of this metric. Our study provides the first assessment of how urbanization has perturbed predator-prey interactions for multiple taxa at a global scale, revealing that the functional role of predators is conserved in cities and urbanization does not inherently relax predation.

3 citations


Cited by
More filters
Journal ArticleDOI
TL;DR: In this paper , a meta-analysis was conducted to quantify the effects of urban environments on three components of trophic ecology in predators: dietary species richness, dietary evenness and stable isotopic ratios (IRs).
Abstract: Predation is a fundamental ecological process that shapes communities and drives evolutionary dynamics. As the world rapidly urbanizes, it is critical to understand how human perturbations alter predation and meat consumption across taxa. We conducted a meta-analysis to quantify the effects of urban environments on three components of trophic ecology in predators: dietary species richness, dietary evenness and stable isotopic ratios (IRs) (δ13C and δ15N IR). We evaluated whether the intensity of anthropogenic pressure, using the human footprint index (HFI), explained variation in effect sizes of dietary attributes using a meta-regression. We calculated Hedges’ g effect sizes from 44 studies including 11 986 samples across 40 predatory species in 39 cities globally. The direction and magnitude of effect sizes varied among predator taxa with reptilian diets exhibiting the most sensitivity to urbanization. Effect sizes revealed that predators in cities had comparable diet richness, evenness and nitrogen ratios, though carbon IRs were more enriched in cities. We found that neither the 1993 nor 2009 HFI editions explained effect size variation. Our study provides, to our knowledge, the first assessment of how urbanization has perturbed predator–prey interactions for multiple taxa at a global scale. We conclude that the functional role of predators is conserved in cities and urbanization does not inherently relax predation, despite diets broadening to include anthropogenic food sources such as sugar, wheat and corn.

3 citations

Journal ArticleDOI
TL;DR: In this article , the authors conducted a quantitative literature review to select studies focused on red fox diet, published before or during 2018 and found that invertebrates were the most frequently reported food of red fox.
Abstract: A key defining feature of a species' niche and ecological roles is its diet (Pocheville 2015). What an animal species eats influences its trophic position, how it moves around, and which other species it interacts with, among many other behaviours. While diets are flexible, predators have evolved to forage optimally on certain prey types or species (Hayward et al. 2011, 2016). The mechanisms that determine optimal foraging behaviour include adaptative responses to prey availability, food quality, energy required to handle prey (Pyke et al. 1977, Sundell et al. 2003), and predation risk from larger co-occurring predators (Haswell et al. 2018). Ecological and climatic conditions determine animal distributions and, in turn, diet composition, so the study of intraspecific diet variation using large biogeographical datasets provides a powerful means to understand the ecology of Carnivora. For instance, regional- and continental-scale studies of dietary variation have been conducted for feral cats Felis catus (Doherty et al. 2015), wildcats Felis silvestris (Lozano et al. 2006), badgers Meles meles (Goszczyński et al. 2000), polecats Mustela putorius (Lodé 1997), common genets Genetta genetta (Virgós et al. 1999), otters Lutra lutra (Clavero et al. 2003), martens Martes spp. (Zhou et al. 2011), and dingoes Canis dingo (Doherty et al. 2019). Knowledge about spatial differences in feeding behaviour can contribute to understanding the foraging strategies used by generalist predators to exploit a wide range of food resources optimally. Among generalist medium-sized carnivores, the red fox Vulpes vulpes (hereafter ‘fox’) is a prime example of a species with adaptive foraging behaviour that allows it to exploit alternative prey when the abundance of its main prey decreases (Kjellander & Nordström 2003). In addition, foxes are able to survive in a range of environments, including highly modified urban and agricultural areas where they exploit anthropogenic foods (Harris 1981, Saunders et al. 1993, Contesse et al. 2004, Bateman & Fleming 2012) and domestic poultry and pets (Lewis et al. 1993). As prey availability varies with habitat and environmental factors, it is not surprising that fox diet composition varies with geographic location. A regional-scale review of red fox diet in the Iberian Peninsula found that invertebrates were the most frequently reported food of foxes followed by fruit/seeds, small mammals, lagomorphs, carrion/garbage, birds, and reptiles (Díaz-Ruiz et al. 2013). Throughout Europe, rodents are the principal food of foxes, followed by plants, invertebrates, birds, lagomorphs, reptiles, and amphibians (Soe et al. 2017). In Australia, where the fox has been introduced, its diet mainly comprises small and medium-sized mammals, livestock, reptiles, birds, invertebrates, and vegetation (Fleming et al. 2021). Notwithstanding these regional- and continental-scale studies of fox diet, quantitative studies describing worldwide biogeographical patterns in fox diet composition have not been undertaken. Understanding of geographic variation in red fox diet at a global scale can be used to predict how this widespread species will respond and adapt to future land use and climate change. ↑ Small mammals, birds, fruit ↓ Medium-sized mammals, invertebrates ↓ Diet richness ↑ Small mammals ↓ Medium-sized mammals ↓ Diet richness ↑ Invertebrates, fruit ↓ Medium-sized mammals ↑ Invertebrates ↑ Diet richness ↑ Fruit ↑ Diet richness ↓ Birds, invertebrates ↓ Diet richness ↓ Large mammals ↑ Birds, fruit ↑ Large mammals ↓ Small and medium-sized mammals, birds ↑ Large mammals ↓ Small and medium-sized mammals, fruit We carried out a quantitative literature review to select studies focused on fox diet, published before or during 2018. We used (“Vulpes vulpes” OR fox) AND (diet OR predation OR ecology) as keywords in ISI Web of Science, JSTOR, and Google Scholar. For each study, we downloaded the title, abstract, authors, year, and journal name. Additionally, we examined the reference lists of all articles that were identified in our initial dataset to ensure key literature was not missed. We selected studies reporting frequency of occurrence (FO, i.e. the number of individual samples where a food item was present as a percentage of the total number of samples) of food items consumed by foxes, from scats, stomachs, or both scats and stomachs (Appendices S1 and S2). FO was selected for comparison of studies as other metrics have not been used as consistently or as widely. To ensure comparability in dietary metrics, we excluded 40 studies that only reported diet composition in relation to total diet contents, e.g. percentage or relative weight/volume, relative FO, percentage volume, prey found around dens, or percentage of hair sampled. We included studies with ≥16 samples that collected fox diet data either in a single year or over many years, as well as either in a single location or in a small geographical region (e.g. county or district). To limit pseudoreplication, when several locations, years, or seasons were sampled in the same study, we pooled FO of each food category across sites <80 km apart, or for the same sites sampled across several years or seasons. We chose this site distance based on fox home range size (median = 3.25 km2; Main et al. 2019), distances travelled by foxes (11 km, Coman et al. 1991; up to 8 km, Tsukada 1997), and maximum dispersal distance (>80 km; Trewhella et al. 1988, Newsome et al. 2017), to reduce the probability that a fox could frequently travel between study sites. This resulted in a total of 217 fox diet studies included in our analyses (Fig. 1). We used a consistent set of 13 food categories to report on fox diet: 1) small mammals (adult weight <500 g), 2) medium-sized mammals (500–6999 g), 3) large mammals (≥7000 g), 4) unidentified mammals, 5) birds, 6) invertebrates, 7) reptiles, 8) amphibians, 9) fish, 10) fruit, 11) vegetation, 12) garbage (i.e. human-related materials and discarded food), and 13) unidentified food. These food categories were chosen because they have been widely reported as foods consumed by foxes in many parts of the world (Abe 1975, Catling 1988, Jankowiak et al. 2008, Drygala et al. 2014). If the authors of a study pooled data for multiple mammal sizes in only one food category, we classed this category as ‘unidentified mammals’. In those instances, we recorded ‘not applicable’ against the individual mammal size categories. Also, if a primary source reported on food types that encompassed more than one of our categories (e.g. ‘fruit and vegetation’ or ‘amphibians and reptiles’), we considered them as ‘unidentified food’. Values reported as <1% and <0.01% were included in our dataset as 0.5% and 0.005%, respectively. If a value or comment was not provided for a food category in the study, we coded it as ‘not applicable’ in our dataset. When the food category was mentioned in the text or accounted for but not recorded in the diet, we included a zero value in our dataset. Where food categories used in primary sources differed from the set of categories we adopted, we used combinatorial probabilities (for more details, see Murphy et al 2019) to pool FO of food categories (i.e. number of individuals of each food or food occurrences over sample size). We created a 20 km circular buffer around each study location to estimate mean annual temperature and precipitation, elevation, and HFI. Mean annual temperature, precipitation, and elevation at 5 m resolution were sourced from the WorldClim dataset (www.wordclim.org). We quantified anthropogenic influence using the HFI layer version 2, 1995–2004 (Wildlife Conservation Society – WCS 2005); this database represents a global spatial dataset of the HFI normalised by biome and realm. Global HFI is estimated using population density, human land use, infrastructure (e.g. built-up areas, night-time lights), and human access (e.g. coastlines, roads, railroads). Given the range in publication dates among papers included in our study (1935–2018), we fitted for each food category one model using the full dataset (n = 217 fox diet studies) and another only including papers carried out between 1995 and 2004 (n = 59 fox diet studies) to match the temporal resolution of the HFI layer (Appendix S3). We found similar patterns between those models; thus, we used the full dataset for subsequent analyses (Appendix S4). Predictor variables were not highly correlated with each other (r ≤ 0.62). To confirm that our choice of a 20 km buffer was appropriate, we tested the degree of correlation between three different buffer radius values (i.e. 5, 10, and 20 km) and found that correlation coefficients were very high (temperature: r > 0.99; precipitation: r > 0.99; elevation: r > 0.96; and HFI: r > 0.90). This indicates that our inferences are unlikely to be influenced greatly by our choice of buffer distance. We assessed patterns of fox diet composition (excluding unidentified birds, unidentified mammals, and unidentified food) among and within continents using the analysis of similarity (ANOSIM) in the ‘vegan’ package (version 2.5-7; Oksanen et al. 2020) in R (version 4.1.2; R Core Team 2021). We excluded Africa from these analyses because only two studies from Africa met our criteria. ANOSIM provided a measure of dissimilarity (R) among and within continents. Dissimilarity (R) values range between −1 (i.e. low dissimilarity) and +1 (i.e. high dissimilarity between groups). Because ANOSIM requires a complete dataset, we considered food category absences as genuine absences. We used Euclidean distances to ordinate fox diet composition in two dimensions using 300 random starts. We performed Monte Carlo randomisation to determine the significance of the final stress values and used ANOSIM to test the hypothesis of no difference between two or more groups, against 999 random permutations of the data, followed by pairwise ANOSIMs. We modelled the relationship between fox dietary variables as dependent variables (i.e. FO of each food category, as well as total diet richness, diversity, and equitability) in separate analyses using the predictor variables of absolute latitude, mean elevation, mean annual temperature, mean annual precipitation, HFI, and sampling method (stomach to scat ratio: all scats = 0 and all stomachs = 1). We tested for multicollinearity among our predictor variables using variance inflation factors calculated in the ‘car’ package (version 3.0.11; Fox & Weisberg 2019). There was high collinearity between absolute latitude and temperature (variance inflation factor >5); thus, we did not include those two variables in the same models. We fitted generalised linear models with predictors of either temperature or absolute latitude, plus precipitation, elevation, and the HFI. That is, we fitted two models for each response variable. We used the ‘lme4’ package (version 1.1.27.1; Bates et al. 2015) in R to fit generalised linear models. All predictor variables were mean-standardised before they were added to the model. We assessed model fit using the quartile–quartile plot function provided in the ‘DHARMa’ package (Hartig 2021). This analysis indicated overdispersion of residuals in all food categories; thus, a Tweedie generalised linear model was fitted. The alpha value was set in each model to maximise the normality of the residuals as indicated using the ‘Tweedie’ package in R (Dunn & Smyth 2005, 2008, Dunn 2017). For each food category, models of all combinations of variables were assessed using dredge in the ‘MuMIn’ package in R (Barton 2020), which were then weighted according to the Akaike Information Criterion corrected for small sample size (AICc; Burnham & Anderson 2002) or Tweedie-AIC value (t-AIC; Dunn 2017). We averaged estimates across models that were within two units of the best model and selected predictor variables included in those models to carry out model predictions. We present means and 95% confidence intervals for significant covariates (P < 0.05). At the global scale, the most commonly reported food categories in fox diets were small mammals (mean FO: 45 ± 4%) and invertebrates (FO: 41 ± 4%; Fig. 2, Appendix S5). The composition of fox diet differed significantly among continents (R = 0.090; P = 0.001), and there were different degrees of dietary overlap in pairwise comparisons (Table 2). Key differences between continents (excluding Africa) included a lower occurrence of small mammals in Australia (mean FO: 23 ± 4%), lower occurrence of medium-sized mammals in Asia (mean FO: 4 ± 7%), higher occurrence of birds (Europe: mean FO: 36 ± 5%; North America: mean FO: 40 ± 15%), fruit (Europe: mean FO: 37 ± 6%; North America: mean FO: 39 ± 18%), and garbage (Europe: mean FO: 18 ± 5%; North America: mean FO: 8 ± 10%) in Europe and North America, as well as a higher occurrence of reptiles in Australia (mean FO: 10 ± 3%; Fig. 2, Appendix S3). Three pairwise comparisons between continents were significant, with Australia and Europe having the highest similarity (R = 0.092; P = 0.001; Appendix S6) and Australia and North America the lowest (R = 0.386; P = 0.001; Table 2, Appendix S6). In the latitude model, fox diet richness decreased with increasing precipitation (t = −2.174, P = 0.031; Fig. 3, Appendix S7). In the temperature model, fox diet richness decreased with both increasing temperature (t = −3.302, P = 0.001) and precipitation (t = −3.450, P < 0.001; Appendices S8 and S9). In latitude models, the FO of small mammals (t = 8.662, P < 0.001), large mammals (t = 3.720, P < 0.001), and birds (t = 7.302, P < 0.001) in fox diet increased with increasing absolute latitude (Fig. 4a–c, Appendix S7). The FO of invertebrates (t = 2.545, P = 0.011) and fruit (t = 3.308, P = 0.001) increased with increasing elevation (Fig. 4e,f), while the FO of medium-sized mammals decreased (t = −2.300, P = 0.022; Fig. 4d). In the temperature model, the FO of birds decreased with increasing elevation (t = −3.059, P = 0.002; Appendices S8 and S10). In the latitude model, bird (t = −2.920, P = 0.004) and invertebrate (t = −3.113, P = 0.002) FO decreased with increasing precipitation (Fig. 4g,h, Appendix S7). There was a lower incidence of large mammals (t = −2.805, P = 0.005) with increasing HFI (Fig. 4i), while bird (t = 2.259, P = 0.025) and fruit (t = 4.727, P < 0.001) FO in fox diet increased (Fig. 4j,k, Appendix S7). These results were similar in temperature models (Appendices S8 and S10). The FO of small mammals (t = −8.067, P < 0.001), large mammals (t = −2.531, P = 0.012), and birds (t = −6.090, P < 0.001) decreased with increased temperature (Appendices S8 and S10). In latitude models, the FO of some food categories in fox diet was influenced by the sampling method. Small mammals (t = −2.824, P = 0.005), medium-sized mammals (t = −2.219, P = 0.027), and fruit (t = −2.867, P = 0.005) were more likely to be recorded in studies that analysed scats rather than stomach contents (Fig. 4l,m,o; Appendix S7). By contrast, large mammals (t = 2.326, P = 0.021) were more likely to be recorded in studies that analysed stomach contents rather than scats (Fig. 4n, Appendix S7). Similar results were found for temperature models (Appendices S8 and S10). Based on the collation of a large dataset of comparable dietary studies from most of the global range of the fox, we assessed geographic variation in fox diet and tested predictions in relation to environmental and anthropogenic drivers. We found that fox diet composition varied among continents and that geographic, climatic, and anthropogenic variables influenced fox diet richness. These results are likely to reflect differences in both prey availability and anthropogenic influences. Moreover, the sample type used in studies (stomach-to-scat ratio) influenced the occurrence of mammals and fruit in fox diet samples. Globally, fox diets were dominated by the occurrence of small mammals and invertebrates, and we found similarities in fox diet composition among some continents (e.g. Australia and Europe). Our results for global fox diet composition accord with previous studies at smaller continental or regional scales, showing that small mammals and invertebrates are principal food items in terms of FO, including in Europe (Soe et al. 2017), the Iberian Peninsula (Díaz-Ruiz et al. 2013), and Australia (Fleming et al. 2021). Consumption of these food categories may be related to their relative availability, as has been demonstrated at local scales (Pavey et al. 2008, Cupples et al. 2011, Spencer et al. 2014), and to prey preferences exhibited by predators (Randa et al. 2009, Spencer et al. 2017). Less abundant prey may be preferred by foxes over more abundant prey if less abundant prey species are naïve to fox predation, making them easier to capture (Graham et al. 2017), or if foxes have evolved adaptations to prey more successfully on those species (optimal foraging). This is especially true within the introduced range of foxes, where foxes can represent a direct (Salo et al. 2007) and indirect (Molsher et al. 2017) threat to naïve prey (Woinarski et al. 2015, Radford et al. 2018). One of the most widely recognised phenomena in ecology is the decline in species diversity with increasing absolute latitude (Hillebrand 2004). Many climatic and other ecological variables, which also vary with absolute latitude, are likely to drive these species distribution patterns, and therefore the prey available to carnivores such as the fox. Fox diet richness, however, did not change with absolute latitude. This result accords with Díaz-Ruiz et al. (2013) who found no relationship between fox diet richness and latitude in the Iberian Peninsula. In contrast, throughout Europe, fox diet diversity has been shown to decrease with latitude in cold but not in warm periods (Soe et al. 2017). Relationships between diet composition and latitude have also been found for other medium-sized carnivores such as American martens Martes americana in the Holarctic region (Zhou et al. 2011). In support of our predictions, small mammal and bird FO in fox diet increased with increasing absolute latitude, which may be due to cooler temperatures away from the equator. Indeed, the biological and ecological processes of endotherms such as small mammals and birds are somewhat independent of ambient temperature, while those of ectotherms (i.e. amphibians, reptiles, and invertebrates) are directly linked with ambient temperature (e.g. Caldwell et al. 2017, Brandt et al. 2018, Jara et al. 2019). Also, the diversity and possibly abundance of reptiles decrease away from the equator (Roll et al. 2017). Thus, it may be that endothermic prey (birds and mammals) are more likely to be consumed by foxes than ectothermic prey (e.g. reptiles) at higher latitudes. Similar latitudinal patterns have been found throughout Europe, where mammals (mainly rodents) and birds were more common in fox diet with increasing latitude (Soe et al. 2017). Also, at the regional scale in the Iberian Peninsula, small mammal FO in fox diet was higher at northern than at southern latitudes (Díaz-Ruiz et al. 2013). We also found that large mammal FO in fox diet increased with increasing absolute latitude. The weight of prey in this food category (≥7000 g) suggests that it is mostly consumed by foxes as carrion. However, the fox use of prey in this food category could also be related in part to the higher densities (>250 head km−2) of large domestic mammals (livestock) away from the equator (Robinson et al. 2014). Like latitude, elevation also has strong effects on ecological community composition and climatic variables (Heaney 2001, McCain 2009), which may in turn influence what carnivores eat. For instance, the diet of Eurasian otters Lutra lutra is characterised by a greater proportion of some food categories (e.g. amphibians) at higher elevations (Remonti et al. 2009). At a regional level on the Iberian Peninsula, medium-sized mammal FO in fox diet decreased with increasing elevation (Díaz-Ruiz et al. 2013). We found the same results at the global scale, and also found that invertebrate and fruit FO in fox diet increased with increasing elevation. The altitudinal trend of consumption of invertebrates and fruit by foxes has also been described at a local scale in the Sumava Mountains in the Czech Republic (Hartová-Nentvichová et al. 2010). The decrease in bird FO in fox diet with increasing elevation that we demonstrated is likely to reflect the decrease in bird species richness with increasing elevation (McCain 2009). Our results highlight the importance of considering the effects of both temperature and precipitation on carnivore diets across large spatial scales. The composition of carnivore diets has been assessed in different seasons (Díaz-Ruiz et al. 2013), cold and warm periods (Soe et al. 2017), Mediterranean and non-Mediterranean climates (Lozano et al. 2006), and across different bioclimatic regions (Doherty et al. 2015, 2019, Fleming et al. 2021). Temperature and precipitation can have different effects (sometimes even counteracting each other) depending on the ecology of predators and prey. We found that fox diet richness decreased with both temperature and precipitation, and the occurrence of many food categories in fox diets varied with precipitation but not with temperature. The finding that diet richness decreased as both temperature and precipitation increased was contrary to our predictions. This result suggests that, under climate change scenarios, fox diet composition could shift with changing prey availability, particularly given the behavioural adaptability of this species. Diet shifts due to climate change have been demonstrated for other carnivores such as the Endangered eastern quoll Dasyurus viverrinus in Australia (Fancourt et al. 2018) and the polar bear Ursus maritimus in Canada (Gormezano & Rockwell 2013). Broader knowledge of the influence of climate variables (e.g. temperature and precipitation) on the relative abundance of prey types and hence carnivore diet composition will enhance our understanding of climate change effects on predator–prey interactions. The activity of some prey is influenced by temperature, which may in turn influence their vulnerability to predation. Consequently, the thermoregulatory strategy of prey can be an important determinant of spatial and temporal variation in carnivore diets. For example, throughout Europe, reptiles (and amphibians) and invertebrates are consumed more by foxes in the warm period than in the cold period (Soe et al. 2017), and in Australia, there is an increasing occurrence of reptiles and frogs in fox diet with increasing temperature (Fleming et al. 2021). However, we found no support for our predictions of relationships between invertebrate FO in fox diet and temperature. Our results suggest that the consumption of small and large mammals and birds by foxes decreases with increasing temperature. This contrasts with results at the continental level in Australia, where the occurrence of mammals (total of all taxa) in fox diet increases with increasing temperature (Fleming et al. 2021), and in the Iberian Peninsula where the occurrence of lagomorphs in fox diet increased with increasing temperature (Díaz-Ruiz et al. 2013). This suggests that there may be continent-specific relationships between some biogeographical variables and the dietary occurrence of specific food categories, which warrants further investigation in future studies. Our results also suggest that the consumption of birds by foxes decreases with increasing precipitation. This could be explained by a reduction in the abundance of the main bird families consumed by foxes (Passeridae and Columbidae; Serafini & Lovari 1993, Contesse et al. 2004, Balestrieri et al. 2011), but further studies are needed to identify the precise drivers of change. Foxes can benefit from living in highly modified areas (Stepkovitch et al. 2019). At the global level, fox home range size shows a strong negative relationship with human population density (Main et al. 2019), probably because foxes benefit from food subsidies in urban areas (Bateman & Fleming 2012, Dawson et al. 2016). Nevertheless, we found no support for our prediction of a relationship between overall fox diet richness and HFI. This accords with Gámez et al. (2020), who found that diet richness of vertebrate predators is unrelated to HFI. At the European level, however, annual fox diet diversity increases with HFI, although this was not replicated during all seasons and was related to the spatial scale used to assess HFI (Soe et al. 2017). We found that large mammal FO in fox diet decreased with increasing HFI, while bird and fruit FO increased. The increase in consumption of large mammals by foxes away from highly anthropogenic areas is likely to be due to the greater presence of wild (Underwood & Kilheffer 2016) and domestic (i.e. livestock) large mammals away from cities and built-up areas. Foxes can consume juveniles of large domestic mammals (Saunders et al. 1993, Gentle 2006) and adults when they are vulnerable (e.g. when giving birth), but large wild and domestic mammals are most likely to be taken as carrion (Catling 1988, Fleming et al. 2016). In highly modified areas such as cities, anthropogenic features influence diversity and relative richness of birds (Aronson et al. 2014). Our results suggest that the proportion of birds in the diet of foxes increases with increasing HFI, probably as a consequence of the greater abundance of some birds closer to cities (Clergeau et al. 1998). In addition, the increase in bird prey in cities and built-up areas could reflect the greater vulnerability of ground-foraging or ground-roosting birds (Woinarski et al. 2021), increased road strike of birds around cities (Fleming et al. 2021), or common incidence of waterbirds around towns. The increase in fruit consumption by foxes with increasing HFI is likely to be due to the abundance of horticultural, ornamental, and pest plant species around cities and built-up areas. Fruit is consumed by foxes in Europe, North Africa, and Australia (Doncaster et al. 1990, Dell'Arte & Leonardi 2005, 2009, Rosalino & Santos-Reis 2009), and foxes are effective seed dispersal agents in natural habitats (Koike et al. 2008, Rosalino et al. 2010). In anthropogenic habitats such as agricultural landscapes, consumption by foxes of domestic fleshy fruit (e.g. figs, grapes, melons, apples, olives, and cherries; Lowe 1989, Dawson et al. 2016) also makes them a potential disperser of those domestic plant species. Knowledge about the seed dispersal role of foxes in anthropogenic habitats, especially in urbanised ones where ornamental and pest plant species are present (McKinney 2006), could be used to determine the role of foxes in the invasive potential of these plant species. Digestibility of different food types taken by the fox varies considerably, which influences the results of diet composition analyses. For mammal prey, there is a direct relationship between prey body size and digestibility (Ferreras & Fernandez-de-Simon 2019). Therefore, the detected occurrence of larger mammals in diet samples derived from scats could be lower than in those derived from stomachs, and the opposite could be true of small mammals. In the present study, small and medium-sized mammals had lower FO in studies that analysed stomach contents rather than scats, while large mammals had greater FO in studies that analysed stomach contents. These findings support continental-scale results in Australia (Fleming et al. 2021). It is likely that foxes consuming small and medium-sized mammals eat a greater proportion of solid parts (e.g. bones and hairs, which are indigestible) than those consuming large mammals. We also found greater FO of fruit in studies that analysed scat contents rather than stomachs. This could also reflect the fact that some authors failed to consider fruit as possible dietary item for foxes (Brunner et al. 1975, Kirkwood et al. 2000, Rosalino & Santos-Reis 2009, Dawson et al. 2016). For instance, in our dataset 17% of studies based on stomachs and 10% of studies based on scats did not report data for fruits. Indeed, contrary to our result, Fleming et al. (2021) found that plant material (including fruits and other vegetation parts) in Australian fox diets was more often recorded in studies using stomach contents than in studies based on scat analysis, probably due to the presence of highly digestible plant material (i.e. fruit pulp). Consequently, when possible, fruits should be considered as a distinct food category, separate from other vegetation parts, in order to improve the evaluation of any temporal (seasonal) variation or spatial variation (e.g. related to the distribution of cultivated and wild fruiting plants). T.S.D. was supported by a Discovery Early Career Research Award from the Australian Research Council (DE200100157). Appendix S1. Data extraction procedure. Appendix S2. List of studies included in this review. Appendix S3. Effects of the anthropogenic variable HFI (Human Footprint Index) on frequency of occurrence (%) of food categories estimated by generalised linear models. Appendix S4. Data used in the analyses. Appendix S5. Mean ± 95% confidence intervals (CI) of frequency of occurrence (%) of food categories in red fox diet, globally and in the five continents. Appendix S6. Boxplots of among-group and within-group dissimilarities in red fox diet. Appendix S7. Parameter estimates and standard errors from the best latitude generalised linear models to describe variation in frequency of occurrence (%) of food categories in red fox diet. Appendix S8. Parameter estimates and standard error from the best temperature generalised linear models to describe variation in frequency of occurrence (%) of food categories in red fox diet. Appendix S9. Effect of temperature models, including significant climatic variables on dietary richness estimated by generalised linear models. Appendix S10. Effects of temperature models, including significant geographic, climatic, anthropogenic, and type-of-sample variables, on frequency of occurrence of food categories estimated by generalised linear models. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

3 citations

Journal ArticleDOI
TL;DR: In this article , the authors compared the temporal response of a small carnivore, the raccoon (Procyon lotor), to the larger coyote (Canis latrans) in four study areas across Michigan that represented a gradient of pressure from humans.
Abstract: Abstract Animals exhibit variation in their space and time use across an urban–rural gradient. As the top‐down influences of apex predators wane due to human‐driven declines, landscape‐level anthropogenic pressures are rising. Human impacts can be analogous to apex predators in that humans can drive increased mortality in both prey species and carnivores, and impact communities through indirect fear effects and food subsidies. Here, we evaluate the time use of a common mesocarnivore across an urban–rural gradient and test whether it is influenced by the intensity of the use of a larger carnivore. Using multiple camera‐trap surveys, we compared the temporal response of a small carnivore, the raccoon (Procyon lotor), to the larger coyote (Canis latrans) in four study areas across Michigan that represented a gradient of pressure from humans. We found that raccoon time use varied by study area and was most unique at the rural extreme. Raccoons consistently did not shift their activity pattern in response to coyotes in the study area with the highest anthropogenic pressures despite the considerable interannual variation, and instead showed stronger responses to coyotes in more rural study areas. Temporal shifts were characterized by raccoons being more diurnal in areas of high coyote activity. We conclude that raccoons may shift time use in the presence of coyotes, dependent on the level of anthropogenic pressure. Our results highlight that the variation in raccoon time use across the entirety of the urban–rural gradient needed to be considered, as anthropogenic pressures may dominate and obscure the dynamics of this interaction.

1 citations