Abstract
Three techniques were evaluated as soil P tests for western Canadian soils: anion‐exchange membrane (AEM), water, and bicarbonate extraction. The AEM, water, and bicarbonate‐extractable total P represented novel approaches to compare to the widely used bicarbonate‐extractable inorganic P (traditional Olsen) soil test. In a range of Saskatchewan soils, similar trends in predicted relative P availability were observed for AEM, water extraction, bicarbonate‐extractable total P, and bicarbonate‐extractable organic P. Correlations between soil test values revealed AEM and water‐extractable P to be most closely correlated, consistent with the similar manner of P removal in the two tests.
Phosphorus availability, as predicted by the tests, was compared to actual P uptake by canola and wheat grown on 14 soils in a growth chamber experiment. P uptake by canola was highly correlated with AEM (r2 = 0.86–0.90), water (0.87 ‐0.94), and bicarbonate‐extractable total (0.91) and inorganic (0.92) P. Uptake of P by wheat was not quite as highly correlated with test‐predicted values: AEM (r2 = ‐0.73–0.78), water (0.72–0.77), bicarbonate total (0.82), bicarbonate‐inorganic P (0.75).
The similarity in coefficients of determination among test methods indicated nearly identical abilities of the tests to predict soil P availability in the range of soils examined. The AEM and water extractions, unlike bicarbonate, are largely independent of soil type and may prove superior when a wider range of soils is being tested. Bicarbonate‐extractable total P and water‐extractable P suffer limitations in analytical simplicity and cost. In testing for P alone, AEM was considered superior to the other methods due to low cost, simplicity, independence of soil type, and high correlation with plant uptake.