A quantitative assessment of phosphorus forms in Australian soils

Abstract

Solution 31P nuclear magnetic resonance (NMR) spectroscopy is by far the most widely-used spectroscopic technique for the speciation of soil organic P, but is yet to be used to characterise a wide range of Australian soils. Therefore, using this technique we analysed the NaOH-EDTA extracts of 18 diverse Australian soils. The majority of 31P NMR signal was assigned to orthophosphate, representing 46 to 90% of total NaOHEDTA extractable P. Orthophosphate diesters and pyrophosphate were present in all soil extracts, their concentrations ranging from 5 to 87 mg/kg (1-5% of total NaOH-EDTA extractable P) and up to 62 mg/kg (5% of total NaOH-EDTA extractable P) respectively. Up to 12 well-resolved orthophosphate monoesters resonances were identified (α- & β- glycerophosphate, myo-inositol hexakisphosphate (phytate), adenosine- 5´-monophosphate (AMP), scyllo-Inositol hexakisphosphate). Orthophosphate monoesters were dominated by α- & β-glycerophosphate and phytate. All three compounds were assigned in all spectra with concentrations of α- & β glycerophosphate ranging from 1–5% and phytate up to 9% of total NaOH-EDTA extractable P. However, phytate concentrations were considerably lower than values determined previously for other soils. As well as numerous sharp resonances in the monoester region which we attributed to specific P-containing compounds, our results showed a large proportion of monoester P (24–65%) could be assigned to a single broad feature. We suggest that this broad signal is due to organic P found in large molecules such as humic acids.