Relationship between soil test phosphorus and phosphorus release to solution

Continued fertilizer applications in excess of those required for optimum p lant growth can increase soil phosphorus (P) concentration and the potentia l for P movement to surface waters, which can contribute to freshwater eutr ophication, Although soil test methods were developed for soil fertility as sessment and fertilizer recommendations, they are frequently used for envir onmental risk assessment because of a lack of consensus on what constitutes a technically defensible environmental soil P test. Several studies have f ound soil test P (STP) is related to the concentration or release of P into soil solution-overland, subsurface, or drainage flow-by two linear relatio nships of significantly different slopes (P < 0.05) on either side of a cha nge point for a limited number of soils. Thus, we investigated the existenc e and behavior of a change point in soil P release for a wide range of vari ously managed soils from the United Kingdom, New Zealand, and the United St ates, The soils varied in pH (3.0-8.2), organic C (1-172 g kg(-1)), and P ( 2-173 mg kg(-1) as 0.5M NaHCO3 extractable P (Olsen P) and 21-553 mg kg(-1) as Mehlich-3 P), Soil P release was determined by CaCl2 extraction (5:1 so lution to soil ratio for 30 min). For all soils, CaCl2-P increased with STP as either Olsen or Mehlich-3 P (representing a quantity/intensity relation ship typical of sorption-desorption isotherms). Statistically significant ( P < 0.05) change points for Olsen P occurred in most soils (20-112 mg P kg( -1)) and for Mehlich-3 P for the United States soils (120-190 mg P kg(-1)). Soil P release (CaCl2-P) increased at a greater rate per unit STP increase above these change points than below, Where no change point was detected, it was found that sampled soils were either of low or high P saturation and , thus, were grouped below or above the change point. The change point coul d be estimated to within 40% of the determined value with a minimum of eigh t randomly selected samples (4 on either side of the change point).