Modeling 31P-NMR P forms in Canadian grasslands using compositional analysis with balances
Abdi, D., B.J. Cade-Menun, N, Ziadi, Y. Shi, G. Bélanger, J. Lajeunesse, J. Lafond. 2016. Modeling 31P-NMR P forms in Canadian grasslands using compositional analysis with balances. Affiche scientifique présentée au Organic Phosphorus Workshop 2016, 5-9 septembre 2016, Lake District, England.
Phosphorus (P) fertilization has been shown to influence soil P forms. The preferred
technique to characterize soil P forms is 31P nuclear magnetic resonance (NMR) spectroscopy
and compositional data analysis using balances is appropriate for modeling the interactive P
forms . Our objective was to model changes in soil 31P-NMR P forms in response to
mineral P fertilization at two grassland sites in Québec, Canada, using compositional analysis
with balances. Four P rates (0, 10, 20, and 40 kg P ha-1), replicated four times, were applied
each year since 2010 to a previously established timothy (Phleum pratense L.) sward at two
sites, one located at Lévis on a Kamouraska clay and the other at Normandin on a Labarre
clay loam. However, only three replications and three P rates (0, 20, and 40 kg P ha-1) were
considered in this study. Soil samples to a 10-cm depth were collected in fall 2013 and were
analysed for different parameters including available P as extracted by Mehlich-3 solution
(PM3, the recommended method for soil test P in Québec) and P forms as extracted with
solution 31P-NMR. We used isometric log-ratio coordinates to compute balances between
31P-NMR P forms. At the two sites, PM3 was greater with the highest rate of P fertilization
(31.2 and 76.0 mg kg-1 at Lévis and Normandin, respectively) than without P (13.1 mg kg-1
and 16.8 mg kg-1). Total inorganic P increased over total organic P as P fertilizer was added,
and inorganic orthophosphate accumulated relative to pyrophosphate and polyphosphate
suggesting an increase of P availability but also an increased risk for P loss in runoff.
Phosphonate slightly increased over the other organic P forms with 20 kg P ha-1 at Lévis and
with 20 and 40 kg P ha-1 at Normandin. The P fertilization also increased the monoesters
relative to the diesters with 20 kg P ha-1 at Lévis and with 20 and 40 kg P ha-1 at Normandain.
Modeling changes in P forms using balances improved understanding of P dynamics in
grasslands that could result in improved P management strategies.