Promoting human-wildlife coexistence through ecological connectivity modelling on the University of British Columbia Vancouver campus

Within these networks, humans and wildlife coexist, living and moving in close proximity with the potential for conflict. Remotely sensed data provides accurate knowledge of landscape cover.

Circuit-based ecological connectivity models can simulate the movement of wildlife across complex urban landscapes. The resistance surfaces needed to perform this modelling were derived using random forest machine learning land cover classification, on combined Planet SkySat multispectral satellite imagery, and light detection and ranging-derived digital surface model datasets with 78.2% overall accuracy. By comparing movement models for wildlife species and humans, sites of overlapping movement were identified primarily along roads and between forested patches in the south of campus.