Defining and Modeling Two Interpretations of Perception in Removal-Distance Models of Point-Count Surveys

• Activity Number: 298 - Ecology and Environmental Policy
• Type: Contributed
• Date/Time: Tuesday, August 1, 2017 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics and the Environment
• Author(s): Adam Martin-Schwarze* and Jarad Niemi and Philip M Dixon
• Companies: Iowa State University and Iowa State University and Iowa State University
• Keywords: abundance ; distance sampling ; removal sampling ; Bayesian ; N-mixture

Abstract

Removal and distance modeling are two common methods to adjust counts for imperfect detection in point-count surveys. Removal modeling uses first detection times to estimate how often animals are available to be detected. Distance modeling uses detection distances to estimate perceptibility, the ability of an observer to detect available animals. Several recent articles have formulated models to combine the approaches into a single removal-distance framework. We observe that these models employ but do not distinguish between two distinct interpretations of perceptibility - one based on perceiving available individuals, the other on perceiving availability cues. Both are correct in certain situations. We apply Bayesian analysis of a hierarchical N-mixture model to simulated and actual avian point counts. We show that the choice of perceptibility model affects bias and coverage in abundance estimation, especially when animals are frequently available but hard to perceive. We introduce a three-stage model for detection that incorporates availability and both kinds of perceptibility. Our model is unbiased with nominal coverage for data simulated from either perceptibility model.