Classification of related vegetation with some physical and chemical properties of soil in Sistan plain rangelands using multivariate regression

To conduct this research, first by preparing the land use map using Landsat 8 OLI sensor images and its combination with geological map, homogeneous units were determined and sampling was done in randomly-systematically in the representative areas of each plant community. Sampling was done along 3 transects of 500m which were vertical, in each site. In each transect placed 10 plots of 3×3m2 at a distance of 50 meters from each other And characteristics related to any plant including vegetative form, scientific name, genus, canopy percentage were harvested. Also, 48 soil samples were taken from a depth of 0-20cm and properties such as soil texture, organic matter, salinity, pH and adsorption ratio sodium were measured. To classification plant communities and investigate the relationships between these communities and soil properties using cluster analysis and principal components analysis respectively in Rstudio software. According to the cluster analysis results, four plant communities were identified and tamarix stricta and Suaeda fruticosa plant species with shrub-shrub vegetative form had the highest frequency with 37.2% and 26.5%, respectively. Results of ANOVA indicated that there were significant differences between some soil properties in identified plant communities. There is also a relationship between the plant communities distribution and environmental factors (p<0.05). The results also showed that the most important factors affecting the establishment and distribution of plant communities in the study area are the percentage of sand, electrical conductivity, sodium absorption ratio and percentage of organic matter