200 journal of soil and water conservation may/june 2010—vol. 65, no. 3 David M. Butler is a research soil scientist at the US Horticultural Research Laboratory, USDA Agricultural Research Service (ARS), Fort Pierce, Florida. Dorcas H. Franklin is a soil scientist at the J. Phil Campbell, Jr. Natural Resource Conser- vation Center, USDA ARS, Watkinsville, Georgia. Miguel L. Cabrera is a professor in the Depart- ment of Crop and Soil Sciences, L. Mark Risse is a professor in the Department of Biological and Agricultural Engineering, and David E. Rad- cliffe is a professor in the Department of Crop and Soil Sciences, University of Georgia, Athens, Georgia. Larry T. West is the national leader for Soil Survey Research and Laboratory, USDA Nat- ural Resources Conservation Service, Lincoln, Nebraska. Julia W. Gaskin is a land application specialist in the Department of Biological and Agricultural Engineering, University of Georgia, Athens, Georgia. Assessment of the Georgia Phosphorus Index on farm at the feld scale for grassland management D.M. Butler, D.H. Franklin, M.L. Cabrera, L.M. Risse, D.E. Radcliffe, L.T. West, and J.W. Gaskin Abstract: In order to better manage agricultural phosphorus (P), most states in the United States have adopted a “P indexing” approach that ranks fields according to potential P losses. In Georgia, the Georgia P Index was developed to estimate the risk of bioavailable P loss from agricultural land to surface waters, considering sources of P, transport mechanisms, and management practices. Nine farm fields, managed as pasture or hay systems, were outfitted with 28 small in-field runoff collectors. Runoff P, soil P, and field management practices were monitored from 2004 to 2007. Fields varied from those rich in P (broiler litter or dairy slurry) to those without P amendments (inorganic nitrogen [N] or no amendments). Data relating to nutrient applications, soil properties, soil P, and management were used as input values to determine a Georgia P Index value estimating the risk of P export from each field. Results indicated that the Georgia P Index underrated the risk in only 2% of the cases when considering loads or mass losses of P, partly due to the influence of small annual runoff vol- umes and thus greater flow-weighted concentrations from some fields. While measured P export was generally low to moderate (<7.5 kg P ha –1 y –1 [<6.7 lb ac –1 yr –1 ]) from fields rated as a low or medium risk of P export, findings from this study indicated that the Georgia P Index, at times, overestimated the risk of P losses for hay systems and underestimated the risk of P losses for pastures when no amendments were applied. Key words: forages—phosphorus—phosphorus index—runoff—water quality In beef cattle–poultry grassland sys- tems typical of the Southern Piedmont (United States), nutrient imports in the form of fertilizers and poultry feed tend to be far greater than nutrient exports at both farm and regional scales (Sharpley et al. 2009). Most generated poultry litter or manure is applied to pasture land as good agronomic practice. However, when applied to meet crop nitrogen (N) requirements, applied phosphorus (P) is disproportionately high, leading to an accumulation of P in the soil and making it vulnerable to transport during storm events either in particulate or dissolved form. To better manage agricultural P, 47 states have adopted a “P indexing” approach, which ranks fields according to potential vul- nerability for P losses (Sharpley et al. 2003). Generally, the P indexing approach to P management works by accounting for both P-source and P-transport factors. Source fac- tors include soil test P (STP) and fertilizer or manure applications, whereas transport factors include runoff, erosion, leaching, and proximity to streams. Other modifying fac- tors may include soil texture, pH, P sorption capacity, flooding frequency, and various conservation factors (Sharpley et al. 2003). The Georgia P Index (Cabrera et al. 2002; Gaskin et al. 2005) was developed to estimate the risk of bioavailable P loss from agricultural land to surface waters considering P-source factors, P-transport factors, as well as man- agement practices that may impact P export. Potential export of P is divided into three loss pathways: (1) soluble P in runoff, (2) par- ticulate P in surface runoff, and (3) soluble P in leachate. Within each pathway, P-source factors (STP, inorganic fertilizer P, or organic fertilizer P) are adjusted according to man- agement practices and are summed to obtain a source-risk rating. Similarly, transport risk ratings are determined using factors such as risk of runoff, risk of sediment loss, and risk of leaching, respectively, for the three path- ways. Additional factors, such as vegetative buffer widths and depth to water table are also considered. Risk ratings for each trans- port pathway are summed, and a numerical P index value is determined in order to place a field within four categories (low, medium, high, or very high) of risk for export of P to surface waters. As is the case for the P indices of Arkansas and North Carolina, the Georgia P Index uses a quantitative approach to esti- mate P losses, whereas the majority of P indices in the United States use a qualitative approach (Osmond et al. 2006). Generally, there has not been extensive field-scale validation of P indices, although a few studies have examined the performance of the various P indices against various measures. Leytem et al. (2003) calculated P index ratings for 272 fields in Delaware and reported that the majority of fields were in the low-risk category, with only 28% of fields indicating that P-based management would be required. The authors contrasted this with the finding that 55% of all fields evaluated had “excessive” STP values and that approximately 14% of fields in the low rat- ing category had unsustainable levels of soil erosion.Veith et al. (2005) compared P index values computed using the Pennsylvania P Index for 22 agricultural fields under field crop management to total P loss calculated by SWAT (Soil and Water Assessment Tool). The two assessment tools were correlated (p = 0.002) with 73% of fields ranked similarly by the two tools. Of the remaining fields, half were underpredicted and half were overpredicted. doi:10.2489/jswc.65.3.200 Copyright © 2010 Soil and Water Conservation Society. All rights reserved. www.swcs.org 65(3):200-210 Journal of Soil and Water Conservation