||Explanatory Models for Ecological Response Surfaces.
Jager, H. I. ;
Overton, W. S. ;
||Oak Ridge National Lab., TN. Environmental Sciences Div. ;Oregon State Univ., Corvallis. Dept. of Statistics.;Corvallis Environmental Research Lab., OR.;Department of Energy, Washington, DC.
Mathematical models ;
Regression analysis ;
Spatial dependencies ;
Spatial distribution ;
Environmental surveys ;
Case studies ;
New York ;
Water pollution ;
EMAP(Environmental Monitoring and Assessment Program) ;
Response variables ;
ANC(Acid neutralizing capacity) ;
ELS(Eastern Lake Survey) ;
||Most EPA libraries have a fiche copy filed under the call number shown. Check with individual libraries about paper copy.
It is often spatial patterns in environmental and ecological variables that arouse interest and demand explanation. The spatial organization of ecological variables, such as species abundance, is often viewed as a collection of individual species responses to variation in the physical environment. In this chapter, the authors use a regression model to predict lake acid neutralizing capacity (ANC) based on environment predictor variables over a large region. These predictions are used to produce model-based population estimates. Two key features of the authors' modeling approach are that it honors the spatial context and the design of the sample data. The spatial context of the data is brought into the analysis of model residuals through the interpretation of residual maps and semivariograms. The sampling design is taken into account by including stratification variables from the design in the model. This ensures that the model applies to a real population. (Copyright (c) 1993 Oxford University Press.)