Landscape ecology is one of theyoungest branches of ecology science, and especially deals on the study of landscape composition, structure and function.  Landscape elements are such as patches, corridors, matrixalong the border and ecotons. Landscape metrics are algorithms for quantifying thespatial characteristics of patches, classes or entire landscape mosaics, and thebest way to compare different landscape requirements. Due to this, the aim of this study was to quantifylandscape patterns and the possibility of breakdown of plant communities usinglandscape metrics. Metrics that were used in this study thenumber of patch, class area proportion in the landscape, mean patch size, edge density, mean shape index, mean patch fractal dimension and total edgein the class level and Shannon diversity index in the landscape level. Metrics listed for vegetationmap and morphology units were calculated using the patch analyst extension for Poshtkooh rangelands of Yazd. Then, using the principal component analysis were determined the important metrics for the separation of plant communities. The resultsof the principal component analysis for the separation of class in morphology units map showed that edge density metric have thegreatest impact but in the vegetation map landscape metrics, were equally important and more importance metric could not be determined and compared to separation the units of land shape in the morphology units map were less separation. Finally, with using of discriminate analysis, wasdetermined thedifferentiation function and metrics that create separation between the plant communities and land classes. Discriminate analysis showed that thenumber of patch metric, edge density and mean patch fractal dimension can beseparated plant communities and classes in the map, units are less resolution.