Feast or famine: How food deserts, wealth and greenspace correlate with coyote diets in Seattle, Washington

Abstract Urban environments create unique challenges for wildlife, where dietary habits may reflect both ecological heterogeneity and societal disparities. In this study, we explored how social factors such as household income and human population density, and ecological factors such as tree canopy correlate with coyote ( Canis latrans ) diets in Seattle, Washington. We hypothesized that coyotes in lower‐income and food insecure neighbourhoods (i.e. ‘food deserts’) would consume more anthropogenic corn‐ and grain‐rich foods as a result of less access to proper waste disposal services and trash containing fewer whole foods (e.g. fresh fruits/vegetables/meats). On the other hand, we hypothesized that coyotes in wealthier, greener areas would have diets richer in non‐anthropogenic foods. We used non‐invasive faecal metabarcoding from scats collected in city parks to obtain a high‐resolution dataset on population‐level coyote diet. Using generalized linear mixed models and multivariate analyses, we found significant variation in diet among city parks. Contrary to the food desert hypothesis, coyotes in less affluent neighbourhoods were not consuming higher proportions of corn‐ and carb‐rich foods. Most food consumption categories were not well explained by socio‐economic factors, with the exception that reported illegal dumping incidents had strong correlation with the amount of non‐native rodent and rabbit detected in scats. Diet diversity was negatively associated with tree canopy cover, adding to a growing body of literature that suggests diet diversity and breadth increases with urbanization. Household income did not significantly correlate with dietary diversity. These findings suggest that coyote diet may be influenced by individual behaviour and localized prey availability, and effects of human‐driven socio‐economic patterns can vary considerably within and among cities. High‐resolution spatial and dietary data can thus reveal the ways that ecological and socio‐economic factors combine to shape urban wildlife behaviour. Read the free Plain Language Summary for this article on the Journal blog.