Estimation models from soil pH with a solid-to-liquid ratio of 1:2.5 to pH measured by other methods using soils in Japan

Abstract

ABSTRACT In Japan, soil pH is primarily measured using water with a solid-to-liquid (SL) ratio of 1:2.5, whereas soil pH measurement methods vary by country. Although some soil pH estimation models for each country have been developed to convert soil pH measured by other methods, suitable models for soils in Japan remain unidentified. Therefore, we propose models for estimating the various soil pH values from those measured by water with an SL ratio of 1:2.5. We collected a dataset of 192 topsoil properties in paddy-upland rotation fields in eastern Japan. The dataset comprised soil pH, electrical conductivity (EC, SL ratio of 1:5), particle density, silt and clay content, and total carbon. Soil pH was measured using distilled water (DW), 0.01 M calcium chloride (CaCl2), and 1 M potassium chloride (KCl) with an SL ratio of 1:1 (pH-1DW), 1:2.5 (pH-2.5DW, pH-2.5CaCl2, pH-2.5KCl), and 1:5 (pH-5DW, pH-5CaCl2, pH-5KCl). The comparison of linear regression models showed that the models comprising pH-2.5DW and intercept were sufficiently accurate to estimate pH-1DW and 5DW. However, to estimate pH-CaCl2 and pH-KCl, the models comprising pH-2.5DW, EC, and intercept were sufficiently accurate. From the study results, we proposed the following equations: pH-1DW = 1.10 × pH-2.5DW − 0.69 (R2 = 0.92, RMSE = 0.19); pH-5DW = 0.80 × pH-2.5DW + 1.32 (R2 = 0.86, RMSE = 0.16); pH-2.5CaCl2 = 0.93 × pH-2.5DW + 1.75 × EC − 0.44 (R2 = 0.94, RMSE = 0.14); pH-5CaCl2 = 0.83 × pH-2.5DW + 2.19 × EC + 0.29 (R2 = 0.88, RMSE = 0.20); pH-2.5KCl = 0.92 × pH-2.5DW + 2.86 × EC − 0.97 (R2 = 0.93, RMSE = 0.15); pH-5KCl = 0.88 × pH-2.5DW + 2.94 × EC − 0.57 (R2 = 0.93, RMSE = 0.15).