PARENT SESSION
Special Session 8: Spatial statistics at multiple scales
Organized by: P Legendre and M Fortin
Tuesday, August 9, 8:00 AM - 11:30 AM, Exhibit Hall 210a-e, Level 2, Palais des congrès de Montréal

Direct multiscale ordination.

Wagner, Helene^*,1, ¹ WSL Swiss Federal Research Institute, Birmensdorf, Switzerland

ABSTRACT- Spatial effects in ecological response data may arise from different sources. First, organisms may respond to a spatially structured environment. Second, the species-environment relationship may change with scale. Third, even in a homogeneous environment, spatial processes such as dispersal or interspecific interactions may create spatial structure. What we observe as spatial autocorrelation in the response data may therefore be due to a combination of various processes. This spatial autocorrelation is often considered a nuisance, as it violates important assumptions for non-spatial analysis of the species-environment relationship, e.g., using regression, RDA, or CCA. The spatial partitioning of regression or ordination results with direct multiscale ordination provides a diagnostic tool for identifying problems of scale-dependent environmental response and residual autocorrelation, including signficance tests for the component patterns. The method is illustrated using simulated data and the oribatid mite data from Borcard et al (1992).

Key words: spatial autocorrelation, scale, multivariate analysis, oribatide mites