The Elimination's effect of rare species on principle component Analysis (Case Study: Jamand district, Nowshahr)

Document Type : Scientific article

Authors

1 Natural Resources Faculty, Urmia University

2 Faculty of Natural Resources, University of Tarbiat Modarres, Nour

Abstract

Given the influence of number of species on the results of vegetation ordination, this study aims to evaluate how the removal of rare species will impact the results of Principle Component Analysis (PCA) in beech communities in the Hyrcanian Forests of northern Ian. Using random-systematic sampling method with 100*200 m network dimension, 26 sample plots were equally established in managed and control compartments. Percentage cover of each plant species was recorded using Braun-Blanquet scale. PCA was applied to determine the relationship between vegetation and environmental variables. Data were analyzed using Procrustan analysis in the R software. The results showed that the removal of rare species with less than 10% abundance in each sample plot will increases the variance but decrease the Eigen value and that there is a high correlation between the raw datasets and rare species removal. But no significant difference was seen for rare species having less than 5% abundance in each sample plot. It might therefore be concluded that the elimination of rare species with <5% abundance does not significantly influence on the loss of information. Results of this research can be used for forest communities with similar structure and composition but caution should be made in applying these results for rangelands.

Keywords