Evaluating Multi-Species Occupancy Models for Rare and Elusive Species

• Activity Number: 140
• Type: Contributed
• Date/Time: Monday, August 4, 2014 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics and the Environment
• Author(s): Kristin M. Broms*+ and John Tipton and Viviana Ruiz and Mevin Hooten
• Companies: Colorado State University and Colorado State University and Colorado State University and Colorado State University
• Keywords: binary data ; hierarchical Bayesian ; species distribution modeling

Abstract

In ecological surveys, presence-absence (or more appropriately, detection-nondetection) data are often collected for multiple species at once on each sampling occasion. However, analyses are typically conducted through single-species models or simple conditional models which do not allow for learning about rare species by borrowing strength from more common species to inform detection and occupancy probabilities. More sophisticated jointly specified multi-species occupancy models have been described in the literature, but do not appear to be widely used. We examined existing multi-species models in terms of their ability to provide accurate inference and then describe a new model using hierarchical Bayesian methods to estimate species richness based on a mechanistic understanding of community ecology. We then fit the model to survey data of plains fish communities in eastern Colorado to uncover the patterns of variability in species compositions across the landscape and to assess future sampling issues.