Files
Abstract
Humans and wildlife are increasingly coming into contact with each other in interactions that are often adverse. Understanding such interactions requires ascertaining how patterns and processes operate at multiple scales. Conservation science’s understanding of broad ecological patterns and theories is constructed by examining local patterns and principles. Consequently, it is imperative wildlife coexistence be investigated at multiple scales to identify the most appropriate spatial and temporal resolutions for local contexts. In the region south of Victoria Falls, Zimbabwe, humans and elephants contend for space and resources amid seasonal fluctuations. Erratic precipitation and harsh growing conditions magnify this competition. The overarching goal of this research is to ameliorate human-elephant relations using geospatial analysis. This work uses geospatial analysis with scale acting as the pivotal nexus and conceptual framework. Following a review of methods for scaling remotely sensed data, land cover is classified at multiple spatial resolutions and different seasons to understand the spatial and temporal scale necessary for documenting changes in vegetation condition. Machine learning is used to model where conflict occurs and to determine which environmental and anthropogenic factors are most important for human-elephant conflict modeling in this region. In contrast to regions such as Hwange National Park in Zimbabwe and protected areas in Botswana and Kenya, little has been published about human-elephant conflict in this specific region. Findings engender greater precision in conflict mitigation efforts and enhance our scientific understanding of scaling processes and human-elephant relations.