A useful statistical tool for inferring patterns from biological atlases

What is it about?

Biological atlases typically compile coarse-scale observational data on observed presences (e.g., checklists of species) across large areas to provide helpful snapshots of distributions at one or more periods in time. Unfortunately, these atlases may offer only limited insights into factors driving changes in distributions. Atlases also may fail to adequately address the imperfect detectability of species; i.e., the possibility that failure to detect a species was driven by observation processes (e.g., survey effort, observer experience) rather than ecological processes. We address both these shortcomings using the statistical approach of occupancy estimation and modelling with a bird species breeding in New York state, USA over two breeding bird atlas periods (1980s and 2000s).

Featured Image