Ordination

Ordination refers to a collection of tools for summarizing continuous trends in multivariate data sets. These are venerable tools that have informed community ecology for many decades and which are now being adopted more generally in “big data” applications. There are three general lineages, corresponding to underlying assumptions about the data (e.g., linear versus nonlinear models). The focus here is on indirect ordinations as tools for summarizing and communicating trends in species or environmental data sets, an application that would arise naturally from an inventory and monitoring program. The workflow for ordination is illustrated using principal components analysis with environmental variables and, in parallel, nonmetric multidimensional scaling with species abundance data. In both illustrations, the process flows from exploratory data analysis to generating the ordination, to post-processing the ordination for interpretation and reporting. In the following chapter, we develop the complementary application of classifying discrete groups in data sets. We return to ordination in Chap. 10 , where it provides a framework for ecological assessment.