Field evaluation of three phosphorus indices on new application sites in Texas

ABSTRACT:

Phosphorus (P) indices were developed to address nonpoint source P losses from agricultural fields; however, only limited information on P index performance at the field- and watershed-scale is available. Evaluation of P indices is necessary to provide the basis of modification and improvement of their usefulness as P management tools. In this study, the ability of the Texas P index to estimate P loss potential was evaluated by comparison with measured annual P loads over three years on four new pasture and six new cultivated litter application sites in the Texas Blackland Prairie. The Arkansas and lowa P indices were also evaluated. The Texas and lowa versions were able to provide reasonable estimates of P loss potential as illustrated with significant linear relationships (p < 0.01) between P index values and measured annual P loads. In general, the P index values, Mehlich 3 soil test P, and poultry litter application rate were better correlated with dissolved P concentrations and loads (r² ranged from 0.12 to 0.91) than with total P and particulate P loads (r² ranged from 0.00-0.31). A major source of error in P index load estimations was their inability to capture variability in annual soil erosion. This source of error was dramatically reduced by using measured erosion instead of estimated annual average erosion (average r² values increased from 0.24 to 0.58). While these results illustrate a potential for the P indices to make relative P loss assessments, research on incentives to prevent buildup of soil P levels, linkages between P levels in soils and receiving waters, and other important issues related to the use of P indices is warranted.