Thresholds of target phosphorus fertility classes in European fertilizer recommendations in relation to critical soil test phosphorus values derived from the analysis of 55 European long-term field experiments
Résumé
Phosphorus (P) fertilizer recommendations of individual countries may differ in many aspects, but often the main principle is to reach or maintain a target range of plant-available P in soil. Within this target P fertility class, the soil is expected to supply enough P to the crop, while P fertilization replaces what is exported by the harvested crop. However, the threshold values of the target P fertility classes are based on a multitude of different soil test P (STP) methods and vary by a factor of up to three, even for countries using the same STP method. This study aimed to provide a comparison of the thresholds of target P fertility classes of different European countries and critical soil test P values (Pcrit; STP below which the average relative yield falls below 95% due to P insufficiency) derived from the analysis of data from 55 long-term field experiments in eight European countries. To overcome the issue of diverging STP methods, all values were converted to Olsen-P using empirically based conversion equations from the literature. Converted threshold values varied by a factor of up to five. For the experimental data, we fitted multi-level Mitscherlich-type models to determine Pcrit values of unfertilized soils corresponding to 95% of maximum yield. We found an average Olsen-Pcrit value of 15 mg P kg−1 (adj. R² = 0.37; RMSE = 14.1% relative yield; n = 2368; 55 experiments), which lies far below several country-specific thresholds of target P fertility classes. Crop-specific analyses resulted in higher Olsen-Pcrit values for sugar beet (22 mg P kg−1), potato (19 mg P kg−1) and winter rapeseed (18 mg P kg−1). Among the texture classes (loam, sand, silt and clay), sandy soils exhibited the highest average Olsen-Pcrit value (22 mg P kg−1). We consider a reevaluation of extraordinarily high country-specific thresholds as well as an inclusion of crop type and soil texture (where not already implemented) to be a reasonable measure towards more cost-effective and environment-friendly P fertilization.