Estimating Dissolved Phosphorus Concentrations in Runoff from Three Physiographic Regions of Virginia

The relationship between STP (soil test P) and DRP (dissolved reactive P) in runoff has been shown to vary with soil type due to differences in soil properties. The purpose of this study was to determine if soil tests could indirectly take into account differences in soil properties and thus provide one relationship with STP and runoff DRP among a variety of soil types. Nine different soil types were sampled from four different fields to provide a range in STP. Unamended soils were packed into runoff boxes for use in a rainfall simulation study. All soils were analyzed for P, Al, and Fe using various extractions, and correlated to runoff DRP concentrations from all soils when grouped based on soil type and physiographic region. Slopes and intercepts were unique to soil type for the STP vs. runoff DRP relationship among all soil P tests except for WSP (water-soluble P) slopes, resulting in an overall universal slope of 0.0125 mg DRP L, which agrees with past studies. The y intercepts from the soil WSP vs. runoff DRP relationship were well related to soil clay content (negative relationship) in addition to the fact that Coastal Plain soils had a significantly greater y intercept than Piedmont and Ridge and Valley soils when grouped by physiographic region. This suggests that Coastal Plain soils possess a greater potential for releasing DRP into runoff compared with Ridge and Valley and Piedmont soil at equivalent WSP values. This study provides a means with which to estimate potential runoff DRP concentrations by use of soil WSP and clay content. INCREASED P concentrations to surface waters typically stimulate algal production and eutrophication (Kenney et al., 2002). The degradation of water quality in the Chesapeake Bay and other surface waters in the MidAtlantic region is attributed in part to increased loading of P (Coale et al., 2002). In addition, agriculture is considered an important nonpoint P source for both particulate (erosion) and dissolved forms to surface waters. Although erosion control can significantly reduce bioavailable P (Uusitalo et al., 2003) and total P (Sharpley and Kleinman, 2003) losses, significant losses of dissolved P can still occur when erosion is kept to a minimum (Sharpley, 1995; Daniel et al., 1994). Dissolved P losses in runoff have been shown to be well related to soil P concentrations and the degree of P saturation onto soil Fe and Al (Sims et al., 2002; Pote et al., 1996; Sharpley et al., 1977). Thus, soils with high concentrations of extractable P are considered to be at a greater risk of causing nonpoint dissolved P losses than low-P soils. Because of the increased awareness in regard to P losses from agricultural soils to surface waters, various P loss prediction models and P index tools have been developed (Karpinets et al., 2004; Vadas and Sims, 2002; Lemunyon and Gilbert, 1993) and often incorporate the relationship between extractable soil P and runoff dissolved P concentrations. The closeness of the relationship between P in solution, P leaching, and P in runoff with STP depends on the range of soils studied as well as the extractants used. These correlations are usually strong when the studied soils are fairly homogeneous in soil properties such as texture and pH (Sims et al., 1998; Yuan and Lavkulich, 1995; Lookman et al., 1996; Pote et al., 1996). As a result, recent studies have concluded that the relationship between extractable soil P and runoff dissolved P will vary with soil type (Torbert et al., 2002; Pote et al., 1999; Sharpley, 1995). Penn et al. (2005) suggested that among soils not amended with P, variability in the soil P vs. dissolved P relationship could be due in part to differences in clay mineral type and quantity. This observation is attributed to the fact that phosphate sorption is primarily a result of ligand exchange reactions between hydroxyls exposed on the surfaces of minerals and the phosphate molecule in soil solution. Although they showed that the relationship between soil WSP and runoff DRP among nine soil types could be reduced to two regression equations based on soil HIV (hydroxy-interlayered vermiculite) and kaolinite content, analysis of soil mineralogy for this purpose would not be practical. Therefore, a “universal” soil extractant capable of creating one relationship between soil P and runoff dissolved P regardless of soil type would be useful as a component in P prediction models. In theory, a universal soil extractant would indirectly take into account different P forms and soil mineralogy by primarily removing P from soil minerals that retain P with the least strength during a runoff event. The objectives of this study were to evaluate the use of various soil P extractants in predicting runoff P losses from selected agricultural soils of the Piedmont, Coastal Plain, and Ridge and Valley physiographic regions of Virginia. C.J. Penn, Dep. of Plant and Soil Sciences, Oklahoma State Univ., 367 Agricultural Hall, Stillwater, OK 74078; G.L. Mullins, Dep. of Plant and Environmental Sciences, New Mexico State Univ., Las Cruces, NM 88003; L.W. Zelazny, Dep. of Crop, Soil, and Environmental Sciences, Virginia Tech, 330 Smyth Hall, Blacksburg, VA 24061; A.N. Sharpley, USDA ARS, PSRMRU, 3702 Curtin Rd., State College, PA 16802. Received 18 Jan. 2006. *Corresponding author (chad.penn@