طبقه بندي وضعيت پوشش گياهي و وضعيت مرتع با استفاده از داده هاي فراواني

Vegetation and Range Condition Classification, Using Frequency Data

Frequency data in 12 range sites from 3 vegetation types were used to determine vegetation and range condition classes. U.S. Forest Service method was employed for condition classification of the same sites. Results of this method were considered as control for evaluating the results of frequency method. To determine vegetation and range condition, percent composition in each species was calculated from frequency data. Species were divided into 3 groups of desirable, intermediate and least desirable, and percent compositions in each group were calculated. Using SPSS software, 2 regression models were constructed. First model with 2 independent variables included percent compositions of desirable and intermediate range species. In this model, dependent variable was vegetation condition score. The other model with 3 independent variables included two groups of species percent compositions as previous model plus soil scores as the third variable. Dependent variable in this model was therefore range condition score. Vegetation and range condition classification of the 12 range sites resulted from first and second models had 91% and 100% accuracy respectively, when compared with the results obtained from U.S. Forest Service method. Quadrat size (l*l m) proved to be the most appropriate size for frequency data collection on studied range sites. Frequency data in these sites were used to create a condition classification guide. Entering percent composition of desirable and intermediate species in vertical and horizontal axes of this guide respectively, class of vegetation condition for a given range site will be determined.