Investigation of rangeland communities composition using multivariate nonparametric permutation

Multivariate data analysis in ecology and biodiversity research is very important. Ecologists often need to test hypotheses about the effects of experimental factors on the entire community composition. To analyze multivariate data, the use of classical statistical methods is based on assumptions such as the normal distribution of data that are not usually observed in ecological data. To analyze multivariate data, the use of classical statistical methods is based on assumptions such as the normal distribution of data that are not usually observed in ecological data. Therefore, in recent years, nonparametric tests, based on permutation tests and distance or dissimilarity matrix, have been widely used to test the existence of differences in species composition in ecology sciences. The purpose of this paper is to introduce and familiarization with new ecological non-parametric multivariate tests related to ecology sciences such as SIMPER, ANOSIM, PERMANOVA and PERMDISP, with the aim of analyzing the composition of plant communities. In order to introduce these analyzes, vegetation data of six sites of rangelands located in the surrounding area of ​​Tehran province were used and the composition of the plant communities of the mentioned areas was analyzed. SIMPER analysis showed that Stipa hohenackeriana and Bromus tomentellus, respectively, had the largest role in differentiating among the sites studied in arid and semi-arid regions. ANOSIM and PERMANOVA analyzes showed a significant difference of plant composition among the sites. According to the results of these tests, Firouzkooh-Alborz, Damavand-Semnan and Saveh-Salafchegan sites had more similarity in terms of composition of vegetation. PERMDISP analysis showed that heterogeneity and multivariate dispersion of species coverage were significantly higher in Salafchegan and Saveh sampling sites. Therefore, according to the results, it can be stated that in order to preserve biodiversity in the study sites, at least three separate management plans are needed. Also, according to SIMPER analysis results, management plans to preserve the biodiversity of the areas studied can be supported by the distinct species identified in each site.