The<b>ecodist</b>Package for Dissimilarity-based Analysis of Ecological Data

TLDR

Ecologists study how species composition relates to environmental and spatial structure, and ecodist provides flexible dissimilarity methods—though the commonly used partial Mantel test is limited by its linearity assumption for complex spatial patterns. The paper introduces ecodist’s features, especially simple and partial Mantel tests, and proposes a modified Mantel correlogram to handle complex nonlinear spatial structures. Ecodist implements simple and partial Mantel tests, a modified Mantel correlogram, and partial multivariate correlograms to test relationships between variables across spatial scales. The methods are illustrated with artificial data and a real study of plant communities in northeastern U.S.

Abstract

Ecologists are concerned with the relationships between species composition and environmental framework incorporating space explicitly is an extremely flexible tool for answering these questions. The R package ecodist brings together methods for working with dissimilarities, including some not available in other R packages. We present some of the features of ecodist, particularly simple and partial Mantel tests, and make recommendations for their effective use. Although the partial Mantel test is often used to account for the effects of space, the assumption of linearity greatly reduces its effectiveness for complex spatial patterns. We introduce a modification of the Mantel correlogram designed to overcome this restriction and allow consideration of complex nonlinear structures. This extension of the method allows the use of partial multivariate correlograms and tests of relationship between variables at different spatial scales. Some of the possibilities are demonstrated using both artificial data and data from an ongoing study of plant community composition in grazinglands of the northeastern United States.