Knowing how many species or individuals occur at a given place and time is fundamental to many questions in wildlife ecology, conservation and management. Enumerating wildlife, however, is complicated by our imperfect and varying (with method, species, habitat, etc) ability to detect animals. In this seminar, I introduce hierarchical statistical models (HSM) as a tool to deal with imperfect detection. I present two case studies using HSM of different levels of occurrence is estimated from species detection/non-detection data. We use this framework to evaluate how land use and climate change driven conversion to savannah habitat will likely affect terrestrial mammals in the Southern Brazilian Amazon. The second study uses open population spatial capture-recapture (opSCR), a much more complex framework that uses spatially explicit individual-level data collected over multiple surveys to estimate population density and demographic rates. We use opSCR to investigate the sex and age-specific expansion of the Pyrenean brown bear population. These case studies highlight the flexibility of HSM as a tool to account for different sampling scenarios and address a wide range of ecological questions.
[more]