Partition and Fate of Phosphorus in Two Ohio Soils.

Soil samples were collected from two Ohio soils (Epiagualfs and Agriaquolls) at seven sites (five cropland) ranging from low to extremely high phosphorus (P) Bray P₁,  plus grass and forested soils at a depth of 24 cm in the Grand Lake St. Marys Watershed.  Four soil samples at each site were split into 0-2, 2-6, 6-12, 12-18, and 18-24 cm and tested for soluble reactive P (SRP), iron/aluminum P (FeP), residual organic P (ResP), and total phosphorus (TP) and soil organic matter content (SOM). Statistical analysis showed the distribution was different than predicted (P <0.0001) for ResP and TP for P level, soil type, and depth interactions.  ResP and TP had means (Root MSE) of 512.4 (118.0) and 718.6 (140.8) and R² (.74 and .82) respectively. SOM and AC means for P level and interactions with P level*Depth and Soil Type* Depth were different than predicted (P<.0001) and for soil type (P<.0015 and P<.0018 respectively).  SOM and AC had means (Root MSE) of 3.7 (1.0) and 720.3 (75.2) and R² (.86 and .81) respectively. FeP had the highest R² (.89) with a174.3 mean and 51.5 Root MSE.  FeP mean distribution was different than predicted (P<.0001) for P level and interactions with P level*Depth and Soil Type* Depth and for soil type (P<.0062). Iron oxides under anaerobic and saturated soil conditions convert from ferric (Fe³⁺) to the ferrous (Fe²⁺) form, releasing SRP.  As P levels increase on these two soils, the ResP fraction appears to become saturated and FeP absorbs the excess P.  However, the FeP fraction is unstable in saturated soils and may be a major cause of eutrophication in Grand Lake St. Marys. In future studies FeP samples will be alternatively wet and dried to quantify the release of SRP from this soil fraction.