Statistical Evaluation of the Taxonomic Composition of Three Soil Map Units in Virginia

Abstract

Abstract A consociation map unit is usually characterized by its central concept selected on the basis of mapping experience, by defining the central concept using morphological and laboratory data, and by assuming that ranges of soil properties defined for the taxon that encompasses the central concept and taxa that includes similar soils represent ranges delineated by the map unit. This study was conducted to statistically evaluate the dominant taxon or central concept as an indicator of map unit purity. Results indicate that the central concept of one of three map units studied had confidence intervals (CIs) that did not exceed limits defined for more than one taxon. Numbers of samples needed for 50% assurance of estimating means within 10% of their true value exceeded cost and time constraints for current soil characterization programs. Of the three map units studied, the highest probability of observing the stated composition was about 0.3 or 30% of the time. Therefore, descriptions of the patterns of occurrence of soils and observed ranges of properties within map units appear to provide more appropriate bases for predicting responses of soils to use and management for the three map units studied.