Nom du corpus

Corpus Systématique Végétale

Titre du document

Some applications of principal components analysis in vegetation: Ecological research of aquatic communities

Lien vers le document
Éditeur
Springer (journals)
Langue(s) du document
Anglais
Type de document
Research-article
Mots-clés d'auteur
  • Association matrix
  • Classification
  • Data reduction
  • Ordination
  • PCA
  • Water plant communities
Nom du fichier dans la ressource
Syst_veg6_v2_015570
Auteur(s)
  • Gerhard Wiegleb
Affiliation(s)
  • Fachbereich IV der Universität, Ammerländer Heerstr. 67-99, 2900, Oldenburg, West-Germany
Résumé

Three examples for the application of PCA to vegetationecological data are given. 1. Vegetation data concerning 33 species from 79 stagnant waters were replaced by the varimax rotated prineipal components of a Q-analysis and subsequently correlated to measured hydrochemical parameters according to a linear model. The correlations were calculated and statistically tested. The hydrogen carbonate proved to be a very important ecological parameter for the composition of the vegetation. 2. 16 principal components were extracted from an association matrix of the 47 most frequent macrophytes of Central Europe (according to 3000 relevés taken from publications) and varimax rotated. Only eigenvectors >0.3 were taken into account for interpretation. The method produces a classification of overlapping species groups. The groups related to the first 10 prineipal components are very well interpretable ecologically and correspond largely to former classification attempts. 3. The frequency of the 40 most frequent water plant species in 14 drainage areas was estimated in a 5-step scale. The result of the ordination was presented as a seatter diagram of the eigenvectors along the first and second unrotated prineipal axes. This showed that a gradient influences the composition of vegetation, which can be explained by the contrast between rhitral and potamal running waters. Finally, two other items are emphasized: 1. An appropriate scaling should be used. 2. Varimax rotated solutions are often superior to unrotated ones, particularly in eases of a complex matrix.

Catégories INIST
  • 1 - sciences appliquees, technologies et medecines
  • 2 - sciences biologiques et medicales
  • 3 - sciences biologiques fondamentales et appliquees. psychologie
  • 4 - ethologie animale
Catégories Scopus
  • 1 - Life Sciences ; 2 - Agricultural and Biological Sciences ; 3 - Plant Science
  • 1 - Physical Sciences ; 2 - Environmental Science ; 3 - Ecology
Identifiant ISTEX
DC57A617A47387739C2AE85F9FC355EB89752032
Revue

Vegetatio

Année de publication
1980
Présence de XML structuré
Non
Score qualité du texte
8.632
Version PDF
1.3
Type de publication
Journal
ark:/67375/1BB-0LB7XFGF-J
Powered by Lodex 9.3.8