Question: In vegetation science, the compositional dissimilarity among two or more groups of plots is usually tested with dissimilarity-based multivariate analysis of variance (db-MANOVA), whereas the compositional characterization of the different groups is performed by means of indicator species analysis. Although db-MANOVA and indicator species analysis are apparently very far from each other, the question we address here is: can we put both approaches under the same methodological umbrella? Methods: We will show that for a specific class of dissimilarity measures, the partitioning of variation used in one-factor db-MANOVA can be additively decomposed into species-level values allowing us to identify the species that contribute most to the compositional differences among the groups. Results: The proposed method, for which we provide a simple R function, is illustrated with one small data set on alpine vegetation sampled along a successional gradient. Conclusion: The species that contribute most to the compositional differences among the groups are preferentially concentrated in particular groups of plots. Therefore, they can be appropriately called indicator species. This connects multivariate analysis of variance with indicator species analysis.

A new method for indicator species analysis in the framework of multivariate analysis of variance / Ricotta, Carlo; Pavoine, Sandrine; Cerabolini, Bruno E. L.; Pillar, Valério D.. - In: JOURNAL OF VEGETATION SCIENCE. - ISSN 1100-9233. - (2021). [10.1111/jvs.13013]

A new method for indicator species analysis in the framework of multivariate analysis of variance

Ricotta, Carlo
;
2021

Abstract

Question: In vegetation science, the compositional dissimilarity among two or more groups of plots is usually tested with dissimilarity-based multivariate analysis of variance (db-MANOVA), whereas the compositional characterization of the different groups is performed by means of indicator species analysis. Although db-MANOVA and indicator species analysis are apparently very far from each other, the question we address here is: can we put both approaches under the same methodological umbrella? Methods: We will show that for a specific class of dissimilarity measures, the partitioning of variation used in one-factor db-MANOVA can be additively decomposed into species-level values allowing us to identify the species that contribute most to the compositional differences among the groups. Results: The proposed method, for which we provide a simple R function, is illustrated with one small data set on alpine vegetation sampled along a successional gradient. Conclusion: The species that contribute most to the compositional differences among the groups are preferentially concentrated in particular groups of plots. Therefore, they can be appropriately called indicator species. This connects multivariate analysis of variance with indicator species analysis.
2021
dissimilarity-based analysis of variance; Euclidean distance; randomization test; standardized effect size; within-group sum of squared dissimilarities
01 Pubblicazione su rivista::01a Articolo in rivista
A new method for indicator species analysis in the framework of multivariate analysis of variance / Ricotta, Carlo; Pavoine, Sandrine; Cerabolini, Bruno E. L.; Pillar, Valério D.. - In: JOURNAL OF VEGETATION SCIENCE. - ISSN 1100-9233. - (2021). [10.1111/jvs.13013]
File allegati a questo prodotto
File Dimensione Formato  
Ricotta_New-method_2021.pdf

accesso aperto

Tipologia: Versione editoriale (versione pubblicata con il layout dell'editore)
Licenza: Creative commons
Dimensione 1.07 MB
Formato Adobe PDF
1.07 MB Adobe PDF

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11573/1540868
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 5
  • ???jsp.display-item.citation.isi??? 3
social impact