Online Program Home
  My Program

Abstract Details

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
Abstract #323625
Title: Bayesian Variance Selection in Analysis of Species Distribution
Author(s): Jing Zhang* and Baini Li and Thomas Crist
Companies: Miami University and Miami University and Miami University
Keywords: Variable Selection ; Bayesian ; Species Distribution ; Diversity Partition ; Shrinkage Priors
Abstract:

In ecological research, it is of interest to study the mechanism of how environmental variables are related to species abundance and richness. The observed species abundances often contain excessive number of zeroes due to limitation of field sampling; hence the observed species richness is often smaller than the actual species richness in the study region. Even though a lot of environmental variables, e.g. climate, geography feature land use and land cover, are observed in such study with the species abundance measurements, the species abundance might respond to only a small portion of these environmental variables. Identification of these variables would be crucial to study the mechanism of how species abundance is related to environment change and how diversity partition in a region responds to these changes. This study aims at developing a Bayesian hierarchical approach, which can handle the zero-inflation and sparsity simultaneously. Bayesian shrinkage priors are used to detect signals (relevant environmental variables) and avoid picking up the redundant predictors. The proposed approach is used to analyze the butterfly occurrence data collected in O


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

Back to the full JSM 2017 program

 
 
Copyright © American Statistical Association