Comparison of Four Methods for Interpreting the Shoemaker‐McLean‐Pratt (SMP) Lime Requirement Test

Abstract

AbstractUsing a lime requirement (LR) test not calibrated under field conditions or on soils which differ greatly from those used to calibrate the procedure may lead to erroneous results. The objectives of this research were, (i) to compare under field conditions the accuracy of four methods of interpreting the SMP LR test, and (ii) to quantify the random, constant, and proportional error associated with each method. Lime applications ranging from 0 to 14.6 Mg ha−1 were made to field plots on two acid soils in northwest Oregon. Surface soil samples taken prior to liming were analyzed according to four methods of interpreting the SMP lime requirement test, namely the original SMP single buffer (SB), McLean's double buffer (DB) and two modified SMP procedures currently used in Oregon and Washington (Modified SB and Modified SB + initial pH, respectively). Statistical models from these four methods were algebraically solved to permit calculation of predicted pH values due to lime amendment which were compared to soil pH measured one and two years after liming. The proportional errors associated with the Original SB, DB, Modified SB and Modified SB + initial pH procedures were 1.2, 1.3, 0.19, and 0.06, respectively. The magnitudes of the constant and random errors associated with these same four methods were 7.8, 8.6, 0.64, and 0.26 pH units; and 0.83; 0.77, 0.19, and 0.25 pH units, respectively. All methods not calibrated under field conditions under‐predicted soil pH values resulting from applications of lime by an average of 0.5 to 0.7 pH units. These results emphasize the importance of employing field conditions and soils from similar geographical areas when calibrating LR tests.