2021 年 65 巻 2 号 p. 110-121
This study assessed the performance of soil order, land-use, and landform variable classifications to estimate soil properties, including pH, organic matter content (OM), available phosphorus (AP), and available potassium (AK). We analyzed the properties of surface soil samples (0–20 cm depth) from 30,471 sites across northeast Thailand. The differences in the properties among the classification groups were tested using a multiple comparison analysis. Our results showed that pH, OM, AP, and AK differed significantly (at the 5% level) between the different soil orders, land-use, and landform categories. In the regional soils, landform classification well-explained the pH, OM, and AK variability and for AP, land use was the best parameter. We propose three alternative means: (1) modify the simple soil classification information transfer, (2) develop stronger prediction models, and (3) provide sequential soil analysis for farms.