Objectives Long-term phosphate (P) fertilization leads to P accumulation in soil, decreasing yield and P fertilizer use efficiency. We studied the release kinetic characteristics of the accumulated P to provide a scientific basis for P fertilizer management.

Methods The long-term fertilization experiment was located in Harbin, Heilongjiang Province. Soil samples from different treatments were marked P1, P2, P3, P4, and P5, corresponding the soil Olsen-P contents were 15.5, 20.2, 93.2, 199, and 216 mg/kg, respectively. The five soil samples were extracted continuously with 0.5 mol/L NaHCO₃ for the determination of the release of available P, and the release dynamics were fitted or described using different kinetic equations. The soil labile P fraction (Resin-P, NaHCO₃-P_i, NaHCO₃-P_o), moderate labile P fraction (NaOH-P_i, NaOH-P_o, DilHCl-P), and stable P fraction (ConcHCl-P_i, ConcHCl-P_o, Residual-P) before and after the extraction were analyzed using the Hedley method.

Results In all the five treatments, the available P was released fast at beginning and then slowed down, and reached equilibrium after 10 repeated extraction in 1400 minutes. The dynamics of available P release were illustrated by Michaelis equation, first-order equation, Elovich equation, parabola equation and power function equation, with the coefficient of determination (R²) above 0.964. Michaelis and first-order equation were preferable as the R² values were 0.9955 and 0.9922, respectively. From the Michaelis equation, the maximum available P releases of the five treatments were 59.8, 60.9, 332.6, 589.7, and 717.0 mg/kg, every 1 mg/kg increase of Olsen-P produced 3.12 mg/kg more P release. The maximum release proportion of soil total P was increased with the Olsen-P content which were 13.68%, 14.86%, 45.44%, 60.67%, and 66.45% for the five treatments, respectively, however, the difference between total P and extractable P in treatments was about 373.99 mg/kg, indicating that the stability of extractive residual P in soil. The first-order equation and the Elovich equation showed that the release of available P was a kinetic process dominated by diffusion. After continuous extraction, soil labile P fraction decreased by 82.9%, moderate labile P fraction decreased by 34.6%, and stable P fraction did not change significantly. In the labile P fraction, Resin-P and NaHCO₃-P_i decreased by 89.8% and 84.5%, respectively. The kinetic equation parameters of soil active P pool, stable P pool and available P release were significantly and positively correlated with organic matter, Olsen-P and total P while negatively correlated with pH, indicating that the release process of soil accumulated P was mainly affected by soil labile P, moderate labile P, soil organic matter, pH, and total P content.

Conclusions The stable P in black soil was less affected by fertilization, most of the accumulated P was released, with the highest release proportion in fraction of Resin-P and NaHCO₃-P_i. The release dynamics of P fitted well by Michaelis equation and first-order equation, and the release amount of P increased with the increase of Olsen-P content in black soil. The main factors driving the release process are soil organic matter, pH, and total P content.