Home > Research > Publications & Outputs > Simultaneous Quantification of Soil Phosphorus ...

Links

Text available via DOI:

View graph of relations

Simultaneous Quantification of Soil Phosphorus Labile Pool and Desorption Kinetics Using DGTs and 3D-DIFS

Research output: Contribution to journalJournal article

Published
Close
<mark>Journal publication date</mark>18/06/2019
<mark>Journal</mark>Environmental Science and Technology
Issue number12
Volume53
Number of pages11
Pages (from-to)6718-6728
Publication statusPublished
Early online date14/05/19
Original languageEnglish

Abstract

The buffering of phosphorus concentrations in soil solution by the soil-solid phase is an important process for providing plant root access to nutrients. Accordingly, the size of labile solid phase-bound phosphorus pool and the rate at which it can resupply phosphorous into the dissolved phase can be important variables in determining when the plant availability of the nutrient may be limited. The phosphorus labile pool (P-labile) and its desorption kinetics were simultaneously evaluated in 10 agricultural UK soils using the diffusive gradients in thin-films (DGT) technique. The DGT-induced fluxes in the soil and sediments model (DIFS) was fitted to the time series of DGT deployments (1-240 h), which allowed the estimation of P-labile, and the system response time (T-c). The P-labile concentration was then compared to that obtained by several soil P extracts including Olsen P, FeO-P, and water extractable P, in order to assess if the data from these analytical procedures can be used to represent the labile P across different soils. The Olsen P concentration, commonly used as a representation of the soil labile P pool, overestimated the desorbable P concentration by 6-fold. The use of this approach for the quantification of soil P desorption kinetic parameters found a wide range of equally valid solutions for T-c. Additionally, the performance of different DIFS model versions working in different dimensions (1D, 2D, and 3D) was compared. Although all models could provide a good fit to the experimental DGT time series data, the fitted parameters showed a poor agreement between different model versions. The limitations of the DIFS model family are associated with the assumptions taken in the modeling approach and the three-dimensional (3D) version is here considered to be the most precise among them.