Urban Wildlife Through Space and Time

Much of my work is focused on developing techniques and methods to get the most ecologically relevant information as possible from observational data. First, I illustrate common approaches used to analyze camera trap data with the analysis of a single species, the Virginia opossum (Didelphis virginiana), and show that this species has different habitat requirements throughout Chicago as urbanization increases. My next study shows how to estimate species associations and co-occurrence rates with camera trap data, which I then validate with extensive use of simulations. Following this, I apply the model I developed to estimate rates of co-occurrence between coyote (Canis latrans), raccoon (Procyon lotor), and the Virginia opossum. While I predicted coyote would negatively influence the two smaller mesocarnivores, I found no evidence of this relationship in the data. In my next chapter, I develop an approach to estimate periodic trends in the spatiotemporal distribution of species by incorporating Fourier series into dynamic occupancy models. Overall, this approach accounts for between 30-73% of the temporal variability in the colonization rates of the species I analyzed. This approach also outperforms other more commonly used approaches that estimate temporal dynamics. In my final chapter, I develop an approach to quantify values and perceptions towards wildlife through comments made on social media. Collectively, I see my dissertation as setting a foundation for future empirical research through the creation of generalizable and robust statistical methods that can be used to answer both basic and applied problems in ecology.