A PLANT FREQUENCY METHOD FOR DETERMINING RANGE CONDITION (INVENTORY, EVALUATION).

The purposes of this study were to compare Soil Conservation Service and Forest Service methods for determining range condition, to develop a method for determining range condition using plant frequency data and to develop range condition guides using Sorensen's coefficient of similarity. The final objective was to compare condition scores based on frequency data to Soil Conservation Service and Forest Service condition scores. Loamy upland and sandy loam upland range sites in the Empire Valley and Santa Rita Experimental Range of southeastern Arizona were sampled. Three prominent soil series of upland range sites within the study area were selected for study. Within each soil series, areas of apparent high, medium and low levels of past and present livestock use were sampled at three locations. Soil profile descriptions were made for each of the 27 sample areas and soil characteristics were used to confirm soil series sampled. Plant communities were sampled by weight estimates, Parker 3-step transects and 40 x 40 cm frequency plots. Forest Service and Soil Conservation Service condition scores did not correlate well within the loamy upland or sandy loam upland range sites. Condition scores by both methods, especially for Forest Service data, are confounded with site potential. Cluster and factor analysis procedures identified plant groups associated with soil series, range site and sample location. Range condition was not identified as a criterion of plant group association. Sorensen's coefficient of similarity scores for frequency data based on a specific "high condition" reference area had a high coefficient of determination value when correlated with Soil Conservation Service condition scores for loamy upland range sites. Forest Service condition scores had a very low coefficient of determination value for the same comparison. Soil Conservation Service and Forest Service condition scores compared to similarity scores for frequency data for the sandy loam upland range site had positive but low coefficient of determination. Sorensen's coefficient of similarity, using frequency data as an index of condition, can be used to develop a range site condition guide for a homogeneous range site