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
Delineating cohesive ecological boundaries across scales.
Fortin, Marie-Josee*,1, Csillag, Ferenc, 1 Department of Zoology, Toronto, Ontario, Canada
ABSTRACT- Any ecological data set can be viewed as a realization of some stochastic process. Stochastic processes, with their set of parameters, define what kind of spatial patterns are more likely than others. Interpretations of the spatial pattern of the data are particularly sensitive to changes in one or more spatial characteristics. When the parameters vary in space, i.e., the process is not stationary, global spatial statistics cannot be applied. One constructive approach to such analytical problems is to partition the data set into stationary subsets. Several spatial partitioning methods are available ranging from local to global. Using the oribatid mite data set, we compare patch-growing and edge-detection methods in their ability to find and characterize homogeneous areas which can be delineated by cohesive boundaries. We seek boundaries that are persistent across multiple variables and across multiple scales. We conclude by presenting linkages between the detection of boundaries (as local heterogeneity in discretized fields) and the interpretation of boundaries (as elements of ecological entities) to provide some guidance to researchers and practitioners.
Key words: Local spatial statistics, Boundary detection, Multiple scales, Patch growing