The risk of P leaching from topsoil based on the change‐point estimated via a split‐line model between Olsen P and leachable P extracted by 0.01 M CaCl2 has been reported. However, little information is available for the assessment of P leaching from soil profiles. In this study, samples were collected at three depth profiles (0–20 cm, topsoil; 20–40 cm, subsoil; 40–60 cm, third‐layer soil) at each of 74 sites under agriculture and forest in an agroforestry area. A cascade extraction method was proposed to determine the leachable P in the subsoil, extracted by the topsoil extraction solution; a similar extracted process was followed in the third‐layer soil, and in the topsoil, it was still extracted by 0.01 M CaCl2. A positive linear correlation was found between the subsoil leachable P extracted by the topsoil extraction solution and the accumulated P obtained from the subsoil leached by topsoil leachates, and so on. Therefore, the cascade extraction method for determining leachable P from topsoils and underlying soils could be valuable for predicting the potential of P leaching from soil profiles. Approximately 81, 73, and 73% of the agricultural sampling sites were at or above the change‐points for the soil depths of 0 to 20, 20 to 40, and 40 to 60 cm (30.4, 32.9, and 18.2 mg kg−1 respectively); these values were higher than those for the forest site, implying a high risk of P leaching from agricultural soil profiles in the study area. Core Ideas A cascade extraction method for evaluating P leaching from soil profiles was proposed. The correlation of soil leachable P and accumulated P in soil leachates was confirmed. The risk of P leaching from soil profiles in an agroforestry area was calculated.

No abstract