Soil Phosphorus Fraction Dynamics during 18 Years of Cultivation on a Typic Paleudult

Traditional soil P test methods estimate plant available inorganic P but ignore the less available inorganic and organic P pools. In low‐input systems where fertilizer P additions are very low to nil, these less available P pools may be a better measure of potential plant available P, particularly in highly weathered soils. The objectives of this study were to determine the size and changes in soil P pools in the nonfertilized and fertilized treatments of a long‐term continuous cultivation experiment established on a Typic Paleudult in Yurimaguas, Peru. A modified version of the Hedley et al. procedure was used to sequentially fractionate soil P into increasingly recalcitrant organic and inorganic pools. The use of path analysis highlights the interactions among P pools and the different roles of the pools in P cycling between the nonfertilized and fertilized system. For the fertilized system, the NaOH‐extractable inorganic P pool acts as a sink for fertilizer P but desorption is rapid enough to maintain high levels of plant available P. For this system, inorganic P pools explain 96% of the variation in the level of available P. Organic P is the primary source of plant‐available P in the nonfertilized system and explains 44% of the variation in available P. Available P, measured by anion exchange resin, is dependent on crop residue ( b values ± standard error = 4.98 ± 3.57), whereas yield depends most strongly on the available P (0.16 ± 0.11) and on NaOH‐extractable organic P (−0.17 ± 0.11). The lack of stability in organic P levels in the first 10 yr illustrates the need for long‐term experiments. The presented results support the notion that traditional soil P tests are inadequate for low‐ to no‐input systems.