PREDICTING LIME-INDUCED CHANGES IN SOIL-PH FROM EXCHANGEABLE ALUMINUM, SOIL-PH, TOTAL EXCHANGEABLE CATIONS AND ORGANIC-CARBON VALUES MEASURED ON UNLIMED SOILS

The 'Lime-it' model is a decision support system for graziers wanting to lime acid soils. In this study we used field experimental data to t est, improve and validate the model's ability to predict changes in so il pH due to variable rates of lime. Data from 13 field experiments, i n which soil parameters were measured 1 year after liming acid soils, were used to derive an index of pH responsiveness to lime (LRI) at eac h site. Multivariate analysis was used to derive a predictive model: L RI was found to be significantly correlated (P < 0.0001) with hydrogen ion concentration ([H+]x 10(5)), exchangeable aluminium (Al), exchang eable cations (TEC) and percent organic carbon (C) data of the unlimed soils. The multivariate equation was then tested against an independe nt data set by comparing the predicted pH change with the measured pH change for eight soils. This evaluation, though generally acceptable, showed a small but significant deviation from the desired 1:1 ratio be tween observed and predicted pH change. We re-calibrated the model for the combined data to derive the model: LRI = 0.764+0.042[H+] -0.016 T EC -0.097 Al -0.016 C. When this model was tested over the whole data set for predicted upsilon, measured pH changes, the following result w as found: measured pH change = 1.01 (predicted pH change) -0.05 (R(2) = 0.85, n = 308). The implications of the predictive equation are cons idered with regard to the mechanisms that are thought to be associated with pH buffering.