Modeling of species richness by macro ecological methods (MEM) and Stacked Species Distribution Models (S-SDM) in Rangelands

Species richness is a simple and interpretable indicator of biodiversity and is one of the most important and and practical measurements in the rangeland ecosystems. Experimental models of richness are used to determine the points of richness of suitable and unsuitable species, and the factors affecting it. The aim of this study is to compare two methods of modeling species richness S-SDM and MEM in the Sorkh-Griveh rangelands in Mazandaran province. For sampling of vegetation and soil, random-systematic stratification method was used. In each sampling unit, 10 plots of 1m2 and a total of 260 plots of one square meter were located. Species list, presence and absence of species were recorded in each plot. Also, three soil samples were taken in each sampling unit from a depth of 0-30 cm. Data related to a number of soil physical and chemical factors were collected in the measurement laboratory, topographic variables were determined with the help of GIS software and 15-year climatic data were collected from meteorological stations in nearby areas. Plant species richness was predicted by amplified regression tree (BRT) model with macroecological methods (MEM) and aggregated distribution model (S-SDM). Data analysis was performed using R software version 3.1.1. The results showed that the species richness observed with the richness predicted by both methods and the richness predicted by MEM method had a significant correlation with the richness predicted by the combination of MEM and S-SDM methods. The results obtained from Spearman's correlation between different methods of predicting the richness of plant species indicated that the S-SDM method with a value of -0.184 and then the MEM method with a value of 0.177 were more consistent with the observed species richness. The results show that by modeling species richness, information about plant community composition, species identification and their prediction in space and time can be used to investigate biodiversity changes.