1 Scientific Agricultural Research Center of Tartous, General Commission for Scientific Agricultural Research – GCSAR, Tartous, Syria
(*Corresponding author: Dr. Ibrahim MUBARAK, Email: imubarak1976@gmail.com Mob:0932328681).

Received: 12/ 2/ 2026     Accepted: 16/ 5/ 2026

Abstract: 

Soil available phosphorus (av.P, ppm) is often determined using laborious and costly laboratory methods. Therefore, this study aimed to develop a pedotransfer function (PTF) to predict av.P in greenhouse soils in Tartous, based on easily measurable chemical properties: soil pH, organic matter (OM, %), electrical conductivity (EC, dS m⁻¹), and available potassium (av.K, ppm). The study relied on a database containing 1158 soil samples, randomly divided into 80% for model calibration and 20% for independent validation. Multiple linear regression analysis revealed the optimal mathematical model as: av.P = 53.919 – 7.059 (pH) + 1.254 (OM) + 3.651 (EC) + 0.017 (av.K) The model achieved high explanatory efficiency with a coefficient of determination (R2=0.838, p<0.001), with a total absence of multicollinearity as the maximum variance inflation factor (VIF) reached only 1.54. Furthermore, both the t-test and the criteria of root mean square error (RMSE = 4.843) and mean absolute error (MAE = 3.817) in the validation data proved the model’s accuracy and lack of bias. The model reflects the chemical rules governing soil phosphorus dynamics, with a negative correlation with pH and positive correlations with the other properties driven by the intensive fertilization legacy. The study recommends adopting the model as a low-cost digital tool to support rational fertilization decision-making in greenhouses.

Keywords: Pedotransfer function, Available phosphorus, Multicollinearity, Greenhouse, Tartous.

 Full paper in Arabic: PDF