Infiltration and Soil Loss of Three Gypsum-Amended Ultisols under Simulated Rainfall^†

Abstract

By-product phosphogypsum (PG) was applied to the surface of topsoil samples from three sandy Georgia Ultisols at a rate of 5 Mg ha⁻¹ in 0.3-m² runoff trays to test the efficacy of this amendment in enhancing infiltration of these soils. Simulated rainfall was applied at 50 mm h⁻¹ for 2.5 h in an initial 1-h (dry) event and three subsequent 0.5-h (wet) events applied 24 h later. In the absence of gypsum, infiltration rate stabilized at 10 mm h⁻¹ for the Cecil soil (Typic Hapludults), 5 mm h⁻¹ for the Wedowee (Typic Hapludults), and 1 mm h⁻¹ for the Worsham (Typic Ochraquults), with surface crusting and sand sorting apparent on all the soils. Surface-applied PG roughly doubled the final infiltration rate of the Cecil and Wedowee soils, and increased infiltration of the Worsham to a lesser degree. Cumulative and average infiltration was significantly higher on all of the PG-amended soils. Soil loss on the Worsham and Wedowee soils was decreased by 50% with the use of PG, and was reduced by 30% on the Cecil. Sediment concentrations were somewhat lower during the wet events with PG treatment for the Worsham and Wedowee soils, showing that less detached soil was available for transport. Although sediment particle sizes for control soils were 15 to 30% clay and 60 to 75% silt, PG-amended soils showed no clay-sized sediment in the runoff, but higher silt contents. Electrical conductivity measurements of the runoff suggest the very low electrolyte concentrations in the untreated soils promote dispersion, resulting in pore blockage and high sediment transportability, whereas PG maintained a high enough electrolyte level (0.5–1.3 dS m⁻¹) to keep soil clays flocculated, thereby reducing crusting and sediment availability.