Online Program Home
My Program

Abstract Details

Activity Number: 123
Type: Contributed
Date/Time: Monday, August 1, 2016 : 8:30 AM to 10:20 AM
Sponsor: Section on Statistics and the Environment
Abstract #320536
Title: Modeling Nonconstant Detection Rates in Removal-Sampled Point-Count Surveys
Author(s): Adam Martin-Schwarze* and Philip Dixon and Jarad Niemi
Companies: and Iowa State University and Iowa State University
Keywords: abundance ; N-mixture model ; survival analysis ; Bayesian
Abstract:

The standard model for estimating detection from removal-sampled point-count surveys assumes that organisms are detected at a constant rate; however, this assumption is often not justified. Detection rates can be influenced by organismal behaviors, such as responses to observer presence, and also by survey methods affecting observer effort. We model detection rates in continuous time via a time-to-detection distribution embedded within a hierarchical N-mixture framework using Bayesian methods. Our model is thus a combination of survival time-to-event analysis with unknown-N, unknown-p abundance estimation. We apply this model to Ovenbird counts and to datasets simulated under various time-to-detection patterns. Models assuming constant detection produce biased estimates when true detection varies with time, whereas models allowing for variable detection produce less biased estimates and nominal credible interval coverage. Models ignoring detection heterogeneity yield biased estimates when such heterogeneity exists, whereas models accounting for it return reasonable coverage rates and can outperform heterogeneity-ignorant models even when there is no heterogeneity.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2016 program

 
 
Copyright © American Statistical Association