Statistical Analysis Of Spatial Structure In Microbial Communities

Abstract

This chapter provides a review of the basic statistical techniques used to detect and quantify spatial structure in ecological data as they can be applied to the analysis of microbial communities. It also discusses the general implications of spatial structure in data analysis, including the inappropriate use of parametric statistical tests with spatially autocorrelated data, and suggests possible alternative procedures. Methods discussed include geostatistics and variogram analysis, kriging, correlograms, Mantel and partial Mantel tests, and time-series analysis.