Sulfur cycling in Swedish arable soils : a chemical perspective

Abstract: Sulfur (S) is an essential plant nutrient. Decreased S deposition in combination with a switch to high-analysis N/P-fertilizers has increased the need for S fertilization. Thus, soil research directed at understanding soil sulfur properties and processes has intensified. However, the methodology at hand has been insufficient for determining relationships between soil properties, S cycling and S availability to crops. In this thesis, recently developed methods were used to study the effect on soil S by two management systems, livestock production and arable crop production, at five different locations within a Swedish long-term fertility field experimental series. In an open incubation study, and a pot trial, where isotopic labeling (35S) was used to trace S transformations, S cycling rates were higher in the livestock system, especially in one soil (Orup). The S delivering capacity of all soils was too low to avoid S deficiency in ryegrass without mineral S application. Observed differences in S cycling patterns could not be satisfactorily explained by soil properties; however, multivariate analyses indicated net S mineralization was negatively related to C/N-ratios and SO42- content. The extent of organic S stabilization through organomineral association and physical protection within microaggregates was investigated by an extraction/dispersion method. The relative distribution between the pools varied between soils, with the residual (non-extractable) pool always being largest; however, only the physically protected fraction was negatively related to plant S uptake. All soil organic S pools were involved in S transformations, although the residual pool was less active than the other pools. Chemical speciation of S in soils and soil fractions was determined by S K-edge X-Ray Absorption Near-Edge Structure (XANES) spectroscopy. A new method for fitting spectra provided reliable quantification of S species by using internally calibrated spectra of dilute (30mM) model compounds. The response of S speciation to management system differed between soils, but highly oxidized S dominated in the organomineral fractions, and intermediate forms of oxidized S in the residual fraction. In conclusion, soil organic S speciation can be accurately quantified by S K-edge XANES spectroscopy. The speciation differs between organomineral associated S and residual S. Treatment effects are dependent on soil type, but S cycling is stimulated by long-term farmyard manure application, as seen in the livestock system.