Phosphorus Speciation in Riparian Soils: A Phosphorus-31 Nuclear Magnetic Resonance Spectroscopy and Enzyme Hydrolysis Study

In the Lake Champlain Basin, phosphorus (P) loading from streambank erosion and cropland are both important P sources, and a better understanding of the factors affecting riparian P loss is needed to help prioritize riparian restoration efforts. We utilized solution phosphorus-31 nuclear magnetic resonance (NMR) spectroscopy and an enzyme hydrolysis method to characterize P and assess bioavailability in 14 commonly mapped riparian soils from northwestern Vermont. Surface horizons were sampled from distinct series at two riparian restoration sites to capture a range of soil properties. Samples were extracted with sodium hydroxide–ethylenediaminetetra-acetic acid (NaOH-EDTA) and analyzed by solution 31P NMR to speciate and quantify P compounds, and commercially available phosphatase enzymes were used to fractionate water-extractable molybdate unreactive P (MUP) into labile orthophosphate monoesters and orthophosphate diesters. Phosphorus extracted by NaOH-EDTA ranged from 74 to 510 mg P kg−1 (representing 14.2 to 31.9% of total soil P), of which 58 ± 13% was identified as organic P. Phosphorus compounds identified in all samples included myo-inositol hexakisphosphate (myo-IHP), scyllo-IHP, neo-IHP, chir-IHP, glycerophosphate, glucose 6-phosphate, mononucleotides, choline phosphate, glucose 1-phosphate, DNA, pyrophosphate, and orthophosphate. Orthophosphate monoesters accounted for 53.7 ± 12.3% of total NaOH-EDTA extractable P and 93 ± 3% of the NaOH-EDTA organic P, indicating the importance of organic P in these soils. Stereoisomers of IHP accounted for 29 ± 7% of NaOH-EDTA extractable Po. For the water extractions, 78 ± 13% of total P was MUP, of which 18 ± 6% was labile orthophosphate monoesters and 31 ± 15% was orthophosphate diesters. Results suggest that analytical indices of riparian P loss potential should consider both organic and inorganic P.