Phosphorus Availability on Two Soils as Determined by Several Methods1

Abstract Alfalfa (Medicago sativa L.) was grown in the greenhouse on two Connecticut soils. Different residual P levels were established in the soils prior to seeding. The effectiveness of five soil testing procedures (Morgan, modified Morgan, modified Morgan plus fluoride, North Carolina, and Bray No. 1) in measuring P availability and predicting alfalfa yield was determined. When both soils were considered together, the Morgan and modified Morgan methods predicted P availability (as measured by P uptake) more reliably than the other three extractants. There were no differences among the five methods in measuring P availability when the soils were considered separately. The Morgan and modified Morgan procedures predicted yield from the first cutting more reliably than the other methods when both soils were considered together. The five extractants were equally effective in predicting total yield (two cuttings) when the soils were considered together or individually. Changes in Morgan and North Carolina‐extractable P effected by cropping correlated more highly with total P uptake from both soils than did changes in extractable P measured by the three other procedures. Phosphorus uptake from one of the soils was highly correlated (r = .92) with the log of the change in North Carolina‐extractable P.