Predicting soil test phosphorus decrease in non‐P‐fertilized conditions

Monitoring the availability of phosphorus (P) in soil under continuous cropping facilitates finding deficiency in crops and contributes to improving crop growth and nutrient management models. Soil P availability for crops is usually estimated by soil test P (STP), such as Bray‐1. This is widely used in the Americas. The relationship between the decrease of STP Bray‐1 and cumulative removal of P was evaluated in non‐P‐fertilized areas in long‐term studies. This removal was the sum of annual P removal over the study period as P exported in grains/crop outside the soil. The objectives were to: (a) quantify changes in STP as a function of cumulative P removal, (b) assess the relationship between relative decrease rate of STP and soil variables as well as annual removal of P by crops, and (c) develop a model to predict decrease of STP Bray‐1. Exponential decay functions were used to describe annual cumulative removal of P and STP from soil over time for 12 long‐term studies where no addition of P fertilizer was carried out. Changes in the relative rate of decrease of STP, relative to the initial STP Bray‐1 value at the onset of the experiment, were predicted by the ratio of soil organic matter to clay and silt and the average annual P removal by exponential decay (R²ₐdⱼ = 0.64; RMSE = 3.2 mg kg⁻¹). We propose this predictive model as suitable to provide estimates of the relative decrease rate of STP by Bray‐1 and thereby improve management of P for optimizing crop yield. HIGHLIGHTS: STP Bray‐1 decrease and cumulative P removal were related by exponential decay functions. Relative decrease rate of STP Bray‐1 was related to SOM/(clay+silt) ratio and annual P removal. A predictive model of the relative decrease rate of STP Bray‐1 was fitted and validate. Our model is a useful tool to help predict soil P availability and nutrient management.