Phosphorus turnover characteristics of soils: Comparison of pot experiment results and modelling by stepwise regression analyses

Abstract

The phosphorus retention ability of soils depends on several factors and influences the effectiveness of fertilization as well as the release of P from soil to water. In the present study the phosphorus supplying and/or retention ability of soils were estimated by two approaches: biological approach (pot experiments) and modelling (by regression analyses). In the course of the biological approach pot experiments were carried out with soils showing significant differences in total and available P contents. Soil samples were collected from selected plots of 9 sites of the National Long-Term Fertilization Trials (NLFT) after 20 years of fertilization, which represents different agro-ecological regions of Hungary. Site characteristics covered a wide range in pH, carbonate and P content, representing typical soil types of the country. With the statistical approach (modelling), the most important soil properties were included and the role of these factors was evaluated by stepwise regression analyses. From the equations, the contribution of important soil parameters to phosphorus supplying and retention ability could be quantified. The objective of the present study was to find a simple way to compare and evaluate the two approaches in P nutrient turnover of soils. Results of the two approaches were correlated. From these results, a rank correlation was also made from the experimental and calculated results. A very close relationship was observed for the P supply and retention of soils (r value was 0.918 for the N 0 P 0 K 0 unfertilized control and 0.927 for the N 200 P 200 K 100 fertilization level). Values obtained with rank correlation were 0.87 and 0.866, respectively, verifying that both methodologies are reliable for estimating the nutrient dynamics in soils and to predict P dynamics in a diverse range of soils.