The estimation and application of the spatial distribution of lime requirement for strongly acidic soils

• Main Authors: Chen-Chih Wu, 吳振記
• Other Authors: 陳仁炫
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/80574603797411205316

博士 === 中興大學 === 土壤環境科學系所 === 95 === Strongly acidic soil is the main problem soil in Taiwan, liming has been suggested to be the most economical and useful practice to reclaim strongly acidic soils. However, the effect is determined by the kinds and the amount of lime materials used. The suitable liming practices are very important for managing strongly acidic soils. The objectives of this study were to determine the lime requirements of twenty-four strongly acid soils of Taiwan using the methods of incubation, reducing Al saturation to 10%, and three buffer methods (SMP-single, Woodruff-original and Woodruff-improved method, respectively). In addition, aregression analysis was also assessed to compare the effects among four different lime materials (agricultural lime, dolomite, oyster-shell powder, and silicate slag). Result showed that the suggested lime requirement of the strongly acidic soils can be calculated by using regression models developed by correlation between lime requirement(y) determined by reducing Al saturation to 10% and lime requirement determined by buffer method(x), such as y = 0.42 x1 -2.16 (R2 = 0.87, p<0.001); y = 1.05 x2 -2.33 (R2 = 0.84, p<0.001) and y = 0.73 x3 - 1.63 (R2 = 0.88, p<0.001), x1, x2, and x3 are the lime requirement determined by SMP-single method, Woodruff-original method and Woodruff-improved method, respectively. The capacity of rising soil pH is generally agricultural lime>dolomite>oyster-shell powder>silicate slag. In the other study in Ho-li, soil samples that took by grid sampling were used to study for lime requirement spatial distribution. The results showed that the least sampling size of this study area was 0.5-ha griding sampling. Furthermore, the spatial distribution of LR could be predicted by using ordinary kriging that the data of LR get from regression models. The application of this regression models was sampled from Da-du-shan soil with 0.5-ha griding sampling. The spatial distribution of LR in Da-du-shan soil was measured by regression models and predicted by using ordinary kriging. Afterward, the LR spatial distribution map was combined with the electronic map by geographic information system. According to the map, we can get the liming of the actual position which wanted to ameirolate and achieve the purpose of improvement quickly.