The ecodist package for dissimilarity-based analysis of ecological data

Authors: GOSLEE, SARAH

Submitted to: Book Chapter
Publication Type: Book / Chapter
Publication Acceptance Date: 6/17/2025
Publication Date: N/A
Citation: N/A

Interpretive Summary: "Distance apart" is a key concept in ecology, whether it refers to travel distance, or to differences in the abundances of plants found on multiple sites. Sophisticated statistical methods have been developed to work with distances, allowing researchers to address complex questions about the relative influence of space and environment on plant and animal communities. Statistical software such as R is an important tool for performing these analyses. This chapter demonstrates a workflow, including both concepts and code, to perform a dissimilarity-based analysis on sample data from the Rocky Mountains. Major spatial and community methods are shown, making this a valuable educational example for both community and landscape ecologists.

Technical Abstract: The concepts of similarity and difference have been intrinsic to community ecology since its early origins. Many ecological questions can only be expressed in terms of distances: spatial structure, but also community similarity/dissimilarity. The clearest and most direct approach to these kinds of questions requires working with these symmetrical dissimilarity matrices directly. This chapter presents a general workflow within the dissimilarity-based analysis framework. The key analyses — correlation, ordination, and spatial correlation — are demonstrated with R code using the ecodist package for a sample dataset from the Rocky Mountains. The likely driving variables in this region are geographic distance and elevation: do changes in these variables relate to changes in plant community composition? Key decision points within the workflow are highlighted, making it possible to adapt this framework for other systems and other research questions.