THE VARIABILITY IN SOIL CHARACTERIZATION DATA FROM LABORATORIES IN THE NORTHEASTERN UNITED STATES

Abstract

Fourteen soil samples, representing 9 soil series, were analyzed by 7 characterization laboratories. The resulting data were compiled and statistically analyzed to estimate the variability in analytical results within and between the characterization laboratories. When a soil sample is analyzed by several different laboratories, the results in the sand, silt, and clay data varied by .+-.1.9%, .+-.2.9%, and .+-.2.5%, respectively. The organic Ca results varied by .+-.0.13%, and 15 atm moisture data varied by .+-.1.3%. The between-laboratory variability in the determination of extractable cations and acidity resulted in a variation of CEC [cation exchange capacity] of .+-.2.9% meq/100 g and .+-.8.5% base saturation. The pH values had a standard deviation from the mean of 0.2, 0.1, and 0.2 pH units when determined in H2O, 1 N KCl, and 0.01 M CaCl2, respectively. These levels of between-laboratory variability are generally about 3 times greater than the levels of analytical variability within the individual laboratories. Significant relationships existed between the amount of laboratory variability and the levels of clay, extractable H, Ca, Mg, Na, and K, and CEC in the various samples analyzed. Due to the inherent laboratory variability included in soils data, decisions based on these data should be somewhat flexible when the determining soil characteristic is near a soil series or taxonomic class limit.