Possible phosphorus losses from the top layer of agricultural soils by rainfall simulations in relation to multi-nutrient soil tests

The objective of the study was to examine a possibility of predicting phosphorus leaching from the top layer of agricultural soils by rainfall simulations by means of three multi-nutrient soil tests: Mehlich 3, NH4-acetate extraction and water extraction (1:5, w/v). Another objective was to determine parameters of maximum phosphorus losses after an extreme load of rainfall on the top layer. Forty soils from different localities of the Czech Republic were used for the experiment. A leaching experiment was conducted in pedological cylinders with a soil layer of about 1 cm and with the bottom from a glass microfibre filter with pores 1.2 mu m in size. Within 15 days the soils were flooded ten times with 25 mm of simulated rainfall in a minimum interval of 1 day. The closest regression between the soil test and phosphorus leaching was computed for NH4-acetate soil test (R-2 = 0.8831) and Mehlich 3 test (R-2 = 0.8572) after the first application of 25 mm of rainfall. In water extraction it was for the mean of 10 simulated rainfalls (R-2 = 0.8674). As leaching proceeded, the closeness of regression diminished due to fluctuations of P concentration in leachates (increases and decreases), mainly in soils with higher P-test. The increase in P concentration could be caused by the activation of phosphorus from Fe-phosphates under anaerobic conditions in wet soils. The steepest decrease in P concentration in leachates was observed in light soils with low CEC value and higher initial P-test.