People and wildlife are living in an increasingly urban world, replete with unprecedented human densities, sprawling built environments, and altered landscapes. Such anthropogenic pressures can affect multiple processes within an ecological community, from spatial patterns to interspecific interactions. We tested two competing hypotheses, human shields versus human competitors, to characterize how humans affect the carnivore community using multi-species occupancy models. From 2017-2020, we conducted the first camera survey of city parks in Detroit, Michigan, and collected spatial occurrence data of the local native carnivore community. Our 12,106-trap night survey captured detected data for coyotes (Canis latrans), red foxes (Vulpes vulpes), gray foxes (Urocyon cinereoargenteus), raccoons (Procyon lotor), and striped skunks (Mephitis mephitis). Overall occupancy varied across species (Ψcoyote=0.40, Ψ raccoon=0.54, Ψred fox =0.19, Ψstriped skunk =0.09). Contrary to expectations, humans did not significantly affect individual occupancy for these urban carnivores. However, co-occurrence between coyote and skunk only increased with human activity. The observed positive spatial association between an apex and subordinate pair supports the human shield hypothesis. Our findings demonstrate how urban carnivores can exploit spatial refugia and coexist with humans in the cityscape. 

Multi-species occupancy model of Detroit carnivores following Rota 2016

ReadMe file uploaded as a supplemental .docx file defining variables as well as describing the data frames for users.