Abstract
It is now well-recognized that urbanization strongly impacts wildlife communities and populations. However, we typically do not know which feature(s) affect individual species most strongly, and this lack of understanding impedes theory development and effective planning for conservation and management goals. To address this knowledge gap, we evaluated how the abundance of ten mammal species responded to six different features of urbanization quantified at five spatial scales using data from 112 camera traps deployed for two years across a gradient of urbanization in New Hampshire, USA. We fit Bayesian abundance models to measure response to each feature and scale. There was no singular urban feature or spatial scale in the best model for all species. Rather, species responded uniquely to features across scales, and the scale of urban features in the best model also varied. Within a species, the magnitude and direction of response varied across features and scales, with only black bear (Ursus americanus), gray fox (Urocyon cinereoargenteus), and Virginia opossum (Didelphis virginiana) exhibiting a consistently significant unidirectional relationship with a single feature across all scales. Our results emphasize that species respond to specific urban features, thus a failure to include certain features can cause misleading inference about wildlife response to “urbanization”. Therefore, researchers must carefully justify the choice of urban feature and spatial scale at which it is represented for each species of interest. An expanded inclusion of multiple urban features in wildlife research will inform management decisions and help attain conservation goals for species impacted by urbanization.