Can We Improve the Kansas P-Index?.

The phosphorus index (PI) is a commonly used tool to assess the risk of phosphorus (P) loss from agricultural fields. However, concerns have been raised about the effectiveness of P indices in limiting the P loss from agricultural fields and improving water quality. Due to limited or lack of measured P loss data, P indices have not been updated or evaluated rigorously. A process-based computer model such as the Agricultural Policy/Environmental Extender (APEX) can be used as an alternative to generating actual P loss datasets to evaluate and update the P indices. The objectives of the study were to evaluate and update the Kansas PI. Average annual runoff, sediment, and P losses were estimated for 588 management scenarios at two locations in eastern Kansas, including watershed and management variables of soil series, slope, cropping system, tillage practice, soil test P concentration, P source, P application rate, P application timing, and P application method. Nonlinear regression was used to adjust the weighting factors in the Kansas PI and improve the correlation between the P loss risk ratings and estimated P loss. A new PI was developed based on the component PI structure proposed by Bolster et al. (2012) and selecting weighting factors for each component with multiple regression techniques. The KS-PI rating explained 57% of the variability in estimated average annual P loss (r²=0.57, p<0.001). Although there was a relatively good correlation, the PI lacked sensitivity to cropping system impacts on P loss.  Using the component index format substantially improved the correlation between average annual P loss and PI ratings (r²=0.74, p<0.001). The component PI could be further improved by refining methods for estimating cropping system impacts on runoff and including more locations during development.