The ecodist package for R provides a suite of dissimilarity-based methods for analyzing ecological data, including simple and partial Mantel tests. These methods allow ecologists to investigate relationships between species composition and environmental factors while explicitly incorporating spatial structure. The package includes functions for calculating dissimilarities, performing Mantel tests, and generating correlograms to assess spatial autocorrelation. The authors discuss the assumptions and limitations of these methods, emphasizing the importance of selecting appropriate dissimilarity metrics and handling spatial autocorrelation. They introduce a novel extension to the Mantel correlogram that overcomes the linearity assumption, allowing for the analysis of complex nonlinear spatial structures. The package is demonstrated using both artificial data and real-world ecological data from grazinglands in the northeastern United States. The article concludes with recommendations for effective use of the methods and highlights the flexibility and power of dissimilarity-based approaches in ecological research.The ecodist package for R provides a suite of dissimilarity-based methods for analyzing ecological data, including simple and partial Mantel tests. These methods allow ecologists to investigate relationships between species composition and environmental factors while explicitly incorporating spatial structure. The package includes functions for calculating dissimilarities, performing Mantel tests, and generating correlograms to assess spatial autocorrelation. The authors discuss the assumptions and limitations of these methods, emphasizing the importance of selecting appropriate dissimilarity metrics and handling spatial autocorrelation. They introduce a novel extension to the Mantel correlogram that overcomes the linearity assumption, allowing for the analysis of complex nonlinear spatial structures. The package is demonstrated using both artificial data and real-world ecological data from grazinglands in the northeastern United States. The article concludes with recommendations for effective use of the methods and highlights the flexibility and power of dissimilarity-based approaches in ecological research.