Estimating soil P pools and desorption rates using flow‐through cells

Understanding and predicting dissolved phosphorus release from soils is critical to proper land management for maximizing P uptake by agronomic crops and minimizing losses to surface waters. In this study, we investigate the amount and rate of P release for 22 diverse soils using flow cells operating at low and high velocity. P release data were fit to a one‐dimensional advection‐dispersion equation assuming two P pools ( Q 1 and Q 2 ) each with their own kinetic desorption rate. The model provided exceptionally good fits to the P release data at both flow velocities. The high‐flow velocity desorbed P at a higher rate, while the slow‐flow velocity released more P and at higher concentrations; although when normalized for residence time, the rate of P desorption was greater in the slow‐flow velocity columns. Fitted values of the readily available P pool ( Q 1 ) for the two flow velocities were well correlated, with Q 1 values for the slow flow velocity being consistently greater than the high flow velocity. Conversely, kinetic desorption parameters between the two flow velocities were not as well correlated. Fitted Q 1 values were strongly correlated with oxalate‐extractable P (P Ox ), Mehlich‐3 P (P M3 ), and water‐extractable P (P W ). Q 1 values were much lower than P Ox and P M3 but similar to P W . We only found weak correlations between the kinetic desorption rates and measured soil properties. Our results show that the use of flow‐through cells can provide useful information on P release from soils, but results will be dependent on flow velocity.