Amelioration of subsurface acidity in the south-west of Western Australia: downward movement and mass balance of surface-incorporated lime after 2-15 years

Twenty-one lime trial sites in the south-western agricultural region of Western Australia (WA) ranging in duration from 2 to 15 years were sampled in 1995 to a depth of 40 cm in 10-cm increments. In each case, samples were taken from both control plots and those with highest lime rates; additionally, samples were taken from plots with intermediate lime rates at 3 trial locations. Lime had been incorporated at rates of 2-12 t/ha within the top 0-10 cm layer at 20 of the sites, and at 15 t/ha within 0-20 cm at the remaining site. The following measurements were made: all samples, pH (1 : 5 0.01 M CaCl2) and exchangeable cations (1 : 10 0.1 M BaCl2); selected depths, undissolved carbonate; selected samples, total Ca (1 : 10 0.2 M HCl) and exchangeable Ca (1 : 10 0.1 M BaCl2 buffered to pH 8.2). The pH of the unlimed subsurface was between 4.0 and 4.5 at 10-20 cm for 15 sites, at 20-30 cm for 7 sites, and at 30-40 cm for 2 sites. In addition, 3 of the unlimed sites had a pH 4.0 at all of these depths. Where lime had been applied, the pH was significantly (P < 0.05) higher than in the controls by at least 0.2 units for 13 sites at depths of 10 cm below the depth of lime incorporation. The same significantly elevated pH was found for 5 sites at depths of 20 cm and for 4 sites at depths of 30 cm. At 5 sites that had also been sampled 3 years previously, the subsurface pH had increased within 4-7 years of liming with 2.5-5 t/ha. Where added lime could be accounted for, the amount of lime added explained 65% of the variation in the residual lime effect at a depth of 10-40 cm. Over all of the wheatbelt trials this represented an average downward movement of lime to a maximum depth of 40 cm of at least 8.3 kg/ha. year. t CaCO3 applied (range 0-22.5). Increased exchangeable Ca corresponded with a decrease in other base cations in only 2 trials. Decreases in exchangeable Al corresponded with increases in both pH and exchangeable Ca, but were less than the Ca increases. Precipitation of exchangeable Al represented an average of about one-sixth of the pH buffering capacity. Mass balance for Ca was achieved in 11 out of 20 cases in the wheatbelt, if Ca loss as nitrate was assumed to occur at rates of leaching published for WA. Lime recovery on the basis of pH increase was lower. The quantity of lime accounted for by Ca in the added and the residual undissolved lime, plus the increase in exchangeable Ca, ranged from <10% to >90%, with <50% in 9 of 19 cases at 15 trials. Nonexchangeable Ca in HCl extracts of a subset of surface samples was from 1% to 60% of Ca not accounted for. In 6 cases the Ca deficit exceeded plausible rates of acidification, and at 2 of these marked decreases of about 1 t/ha undissolved CaCO3 corresponded with episodes of high wind speed and observations of soil erosion. The implications of possible losses of lime by erosion are discussed.