|
Activity Number:
|
266
|
|
Type:
|
Invited
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
WNAR
|
| Abstract - #307864 |
|
Title:
|
Searching for Ecological-Environmental Relationships
|
|
Author(s):
|
Eric P. Smith and Samantha Bates Prins*+
|
|
Companies:
|
Virginia Polytechnic Institute and State University and James Madison University
|
|
Address:
|
Department of Mathematics and Statistics, Harrisonburg, VA, 22807,
|
|
Keywords:
|
environmetrics ; clustering ; stressors ; voronoi ; multivariate ; redundancy analysis
|
|
Abstract:
|
Biological monitoring of aquatic sites is used for trend assessment, site assessment and for evaluation of the restoration of sites. Despite the vast amount of information that is collected the analysis of this data remains difficult, especially when attempting to assign causes to changes in biological measures. One approach that we have investigated is to search for strong relationships between biological data and potential stressors and covariates. We use a search algorithm that is based on Voronoi tessellations to find irregular shapes that contain a set of spatially connected sites. Regression and multivariate methods are used to describe the relationships between biological metrics and stressors or between counts of taxa and stressors. The methods are illustrated on data sets from West Virginia and on data from the mid-Atlantic Highlands.
|
