Comparing tests for soil fertility: 1. Conversion equations between olsen and mehlich 3 as phosphorus extractants for 120 soils of south italy

Abstract

The multiple‐element extractant Mehlich 3 (M3) has not been tested extensively in Europe. In this Land, soil‐P test recommendations are based, since decades, on the evaluation of the Olsen‐extractable P, and the optimal soil‐P levels have been established to range between 1.5 and 3.0 mg of Olsen‐P per 100 g of soil. A research programme was started in order to assess the suitability of M3 as routine soil‐P test in European laboratories. As a first approach, we develop conversion equations from Olsen‐P to M3‐P, in order to assess the agreement and the consistency of the measurements under a wide range of chemical and physical soil properties. To this aim, 120 samples with drastically contrasting features were selected within 206 soils collected from all the regions of South Italy. Soil‐P ranges were 0.07–60.53 for M3, and 0.08–21.47 mg/100 g for Olsen. The results showed that M3‐P was a P extractant more efficient than Olsen. The amounts of M3‐P were, on the average, twice as large as the Olsen‐P ones, with mean values of 5.70 and 2.75 mg/100g, respectively. The soil properties exerted a great influence, as well as showed a contrasting effect, on the extraction efficiency of each method. For neutral‐alkaline calcareous soils, the average M3‐P/Olsen‐P ratio increased to 2.52, and the efficiency of M3 poorly varied according to soil pH and CaCO₃ content. On the contrary, in CaCO₃‐free acidic soils, the M3‐P/Olsen‐P ratio decreased to 1.63. In particular, anomalous ratio values less than 1.0 were observed for acidic soils with high content of organo‐mineral complexes with be shifted to 3.7–7.7 for calcareous soils, and to 2.7–4.9 for CaCO₃‐free soils. Field calibrations would give more information to establish the proper values according to either the soil properties and plant requirements. The results encourage the introduction of M3 as routine soil‐P test in our Countries. One must take into account, however, that some soil properties, in particular the CaCO₃ content, migth be considered for a more precise comparative evaluation with existing laboratory data.